diff --git "a/checkpoint-21000/trainer_state.json" "b/checkpoint-21000/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-21000/trainer_state.json"
@@ -0,0 +1,147033 @@
+{
+ "best_metric": null,
+ "best_model_checkpoint": null,
+ "epoch": 67.09265175718849,
+ "eval_steps": 500,
+ "global_step": 21000,
+ "is_hyper_param_search": false,
+ "is_local_process_zero": true,
+ "is_world_process_zero": true,
+ "log_history": [
+ {
+ "epoch": 0.003194888178913738,
+ "grad_norm": 9.75,
+ "learning_rate": 0.0005,
+ "loss": 3.5217,
+ "step": 1
+ },
+ {
+ "epoch": 0.006389776357827476,
+ "grad_norm": 48.75,
+ "learning_rate": 0.0005,
+ "loss": 10.8705,
+ "step": 2
+ },
+ {
+ "epoch": 0.009584664536741214,
+ "grad_norm": 14.6875,
+ "learning_rate": 0.0005,
+ "loss": 3.0211,
+ "step": 3
+ },
+ {
+ "epoch": 0.012779552715654952,
+ "grad_norm": 23.75,
+ "learning_rate": 0.0005,
+ "loss": 4.3266,
+ "step": 4
+ },
+ {
+ "epoch": 0.01597444089456869,
+ "grad_norm": 5.25,
+ "learning_rate": 0.0005,
+ "loss": 1.9879,
+ "step": 5
+ },
+ {
+ "epoch": 0.019169329073482427,
+ "grad_norm": 12.0625,
+ "learning_rate": 0.0005,
+ "loss": 2.4791,
+ "step": 6
+ },
+ {
+ "epoch": 0.022364217252396165,
+ "grad_norm": 10.0625,
+ "learning_rate": 0.0005,
+ "loss": 2.2867,
+ "step": 7
+ },
+ {
+ "epoch": 0.025559105431309903,
+ "grad_norm": 1.265625,
+ "learning_rate": 0.0005,
+ "loss": 1.8691,
+ "step": 8
+ },
+ {
+ "epoch": 0.02875399361022364,
+ "grad_norm": 11.125,
+ "learning_rate": 0.0005,
+ "loss": 2.289,
+ "step": 9
+ },
+ {
+ "epoch": 0.03194888178913738,
+ "grad_norm": 13.875,
+ "learning_rate": 0.0005,
+ "loss": 2.5878,
+ "step": 10
+ },
+ {
+ "epoch": 0.03514376996805112,
+ "grad_norm": 9.4375,
+ "learning_rate": 0.0005,
+ "loss": 2.2933,
+ "step": 11
+ },
+ {
+ "epoch": 0.038338658146964855,
+ "grad_norm": 1.375,
+ "learning_rate": 0.0005,
+ "loss": 1.9492,
+ "step": 12
+ },
+ {
+ "epoch": 0.04153354632587859,
+ "grad_norm": 6.03125,
+ "learning_rate": 0.0005,
+ "loss": 2.0396,
+ "step": 13
+ },
+ {
+ "epoch": 0.04472843450479233,
+ "grad_norm": 7.4375,
+ "learning_rate": 0.0005,
+ "loss": 2.1731,
+ "step": 14
+ },
+ {
+ "epoch": 0.04792332268370607,
+ "grad_norm": 5.90625,
+ "learning_rate": 0.0005,
+ "loss": 2.0859,
+ "step": 15
+ },
+ {
+ "epoch": 0.051118210862619806,
+ "grad_norm": 3.328125,
+ "learning_rate": 0.0005,
+ "loss": 1.833,
+ "step": 16
+ },
+ {
+ "epoch": 0.054313099041533544,
+ "grad_norm": 3.71875,
+ "learning_rate": 0.0005,
+ "loss": 1.8686,
+ "step": 17
+ },
+ {
+ "epoch": 0.05750798722044728,
+ "grad_norm": 3.453125,
+ "learning_rate": 0.0005,
+ "loss": 1.8999,
+ "step": 18
+ },
+ {
+ "epoch": 0.06070287539936102,
+ "grad_norm": 2.34375,
+ "learning_rate": 0.0005,
+ "loss": 1.8458,
+ "step": 19
+ },
+ {
+ "epoch": 0.06389776357827476,
+ "grad_norm": 0.34375,
+ "learning_rate": 0.0005,
+ "loss": 1.7944,
+ "step": 20
+ },
+ {
+ "epoch": 0.0670926517571885,
+ "grad_norm": 1.625,
+ "learning_rate": 0.0005,
+ "loss": 1.8099,
+ "step": 21
+ },
+ {
+ "epoch": 0.07028753993610223,
+ "grad_norm": 1.4296875,
+ "learning_rate": 0.0005,
+ "loss": 1.7969,
+ "step": 22
+ },
+ {
+ "epoch": 0.07348242811501597,
+ "grad_norm": 0.1318359375,
+ "learning_rate": 0.0005,
+ "loss": 1.7794,
+ "step": 23
+ },
+ {
+ "epoch": 0.07667731629392971,
+ "grad_norm": 1.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.7922,
+ "step": 24
+ },
+ {
+ "epoch": 0.07987220447284345,
+ "grad_norm": 0.6640625,
+ "learning_rate": 0.0005,
+ "loss": 1.7673,
+ "step": 25
+ },
+ {
+ "epoch": 0.08306709265175719,
+ "grad_norm": 0.314453125,
+ "learning_rate": 0.0005,
+ "loss": 1.7476,
+ "step": 26
+ },
+ {
+ "epoch": 0.08626198083067092,
+ "grad_norm": 0.73828125,
+ "learning_rate": 0.0005,
+ "loss": 1.758,
+ "step": 27
+ },
+ {
+ "epoch": 0.08945686900958466,
+ "grad_norm": 0.50390625,
+ "learning_rate": 0.0005,
+ "loss": 1.7467,
+ "step": 28
+ },
+ {
+ "epoch": 0.0926517571884984,
+ "grad_norm": 0.3515625,
+ "learning_rate": 0.0005,
+ "loss": 1.7427,
+ "step": 29
+ },
+ {
+ "epoch": 0.09584664536741214,
+ "grad_norm": 0.63671875,
+ "learning_rate": 0.0005,
+ "loss": 1.7317,
+ "step": 30
+ },
+ {
+ "epoch": 0.09904153354632587,
+ "grad_norm": 0.115234375,
+ "learning_rate": 0.0005,
+ "loss": 1.7203,
+ "step": 31
+ },
+ {
+ "epoch": 0.10223642172523961,
+ "grad_norm": 0.82421875,
+ "learning_rate": 0.0005,
+ "loss": 1.7163,
+ "step": 32
+ },
+ {
+ "epoch": 0.10543130990415335,
+ "grad_norm": 0.37109375,
+ "learning_rate": 0.0005,
+ "loss": 1.7097,
+ "step": 33
+ },
+ {
+ "epoch": 0.10862619808306709,
+ "grad_norm": 0.5390625,
+ "learning_rate": 0.0005,
+ "loss": 1.7059,
+ "step": 34
+ },
+ {
+ "epoch": 0.11182108626198083,
+ "grad_norm": 0.546875,
+ "learning_rate": 0.0005,
+ "loss": 1.6942,
+ "step": 35
+ },
+ {
+ "epoch": 0.11501597444089456,
+ "grad_norm": 0.47265625,
+ "learning_rate": 0.0005,
+ "loss": 1.6987,
+ "step": 36
+ },
+ {
+ "epoch": 0.1182108626198083,
+ "grad_norm": 0.5390625,
+ "learning_rate": 0.0005,
+ "loss": 1.6947,
+ "step": 37
+ },
+ {
+ "epoch": 0.12140575079872204,
+ "grad_norm": 0.287109375,
+ "learning_rate": 0.0005,
+ "loss": 1.6771,
+ "step": 38
+ },
+ {
+ "epoch": 0.12460063897763578,
+ "grad_norm": 1.0390625,
+ "learning_rate": 0.0005,
+ "loss": 1.664,
+ "step": 39
+ },
+ {
+ "epoch": 0.12779552715654952,
+ "grad_norm": 0.703125,
+ "learning_rate": 0.0005,
+ "loss": 1.6817,
+ "step": 40
+ },
+ {
+ "epoch": 0.13099041533546327,
+ "grad_norm": 0.609375,
+ "learning_rate": 0.0005,
+ "loss": 1.6801,
+ "step": 41
+ },
+ {
+ "epoch": 0.134185303514377,
+ "grad_norm": 0.408203125,
+ "learning_rate": 0.0005,
+ "loss": 1.6438,
+ "step": 42
+ },
+ {
+ "epoch": 0.13738019169329074,
+ "grad_norm": 1.71875,
+ "learning_rate": 0.0005,
+ "loss": 1.6109,
+ "step": 43
+ },
+ {
+ "epoch": 0.14057507987220447,
+ "grad_norm": 1.0390625,
+ "learning_rate": 0.0005,
+ "loss": 1.6345,
+ "step": 44
+ },
+ {
+ "epoch": 0.14376996805111822,
+ "grad_norm": 0.6171875,
+ "learning_rate": 0.0005,
+ "loss": 1.5955,
+ "step": 45
+ },
+ {
+ "epoch": 0.14696485623003194,
+ "grad_norm": 36.25,
+ "learning_rate": 0.0005,
+ "loss": 1.9916,
+ "step": 46
+ },
+ {
+ "epoch": 0.1501597444089457,
+ "grad_norm": 12.8125,
+ "learning_rate": 0.0005,
+ "loss": 1.708,
+ "step": 47
+ },
+ {
+ "epoch": 0.15335463258785942,
+ "grad_norm": 0.8515625,
+ "learning_rate": 0.0005,
+ "loss": 1.5948,
+ "step": 48
+ },
+ {
+ "epoch": 0.15654952076677317,
+ "grad_norm": 0.490234375,
+ "learning_rate": 0.0005,
+ "loss": 1.5718,
+ "step": 49
+ },
+ {
+ "epoch": 0.1597444089456869,
+ "grad_norm": 0.306640625,
+ "learning_rate": 0.0005,
+ "loss": 1.5491,
+ "step": 50
+ },
+ {
+ "epoch": 0.16293929712460065,
+ "grad_norm": 0.18359375,
+ "learning_rate": 0.0005,
+ "loss": 1.5168,
+ "step": 51
+ },
+ {
+ "epoch": 0.16613418530351437,
+ "grad_norm": 0.345703125,
+ "learning_rate": 0.0005,
+ "loss": 1.494,
+ "step": 52
+ },
+ {
+ "epoch": 0.16932907348242812,
+ "grad_norm": 0.734375,
+ "learning_rate": 0.0005,
+ "loss": 1.4803,
+ "step": 53
+ },
+ {
+ "epoch": 0.17252396166134185,
+ "grad_norm": 0.8984375,
+ "learning_rate": 0.0005,
+ "loss": 1.4731,
+ "step": 54
+ },
+ {
+ "epoch": 0.1757188498402556,
+ "grad_norm": 1.375,
+ "learning_rate": 0.0005,
+ "loss": 1.4745,
+ "step": 55
+ },
+ {
+ "epoch": 0.17891373801916932,
+ "grad_norm": 0.447265625,
+ "learning_rate": 0.0005,
+ "loss": 1.4506,
+ "step": 56
+ },
+ {
+ "epoch": 0.18210862619808307,
+ "grad_norm": 0.271484375,
+ "learning_rate": 0.0005,
+ "loss": 1.4113,
+ "step": 57
+ },
+ {
+ "epoch": 0.1853035143769968,
+ "grad_norm": 0.67578125,
+ "learning_rate": 0.0005,
+ "loss": 1.4002,
+ "step": 58
+ },
+ {
+ "epoch": 0.18849840255591055,
+ "grad_norm": 0.53515625,
+ "learning_rate": 0.0005,
+ "loss": 1.4037,
+ "step": 59
+ },
+ {
+ "epoch": 0.19169329073482427,
+ "grad_norm": 0.15234375,
+ "learning_rate": 0.0005,
+ "loss": 1.3735,
+ "step": 60
+ },
+ {
+ "epoch": 0.19488817891373802,
+ "grad_norm": 0.1630859375,
+ "learning_rate": 0.0005,
+ "loss": 1.3766,
+ "step": 61
+ },
+ {
+ "epoch": 0.19808306709265175,
+ "grad_norm": 0.1318359375,
+ "learning_rate": 0.0005,
+ "loss": 1.3479,
+ "step": 62
+ },
+ {
+ "epoch": 0.2012779552715655,
+ "grad_norm": 0.2392578125,
+ "learning_rate": 0.0005,
+ "loss": 1.3584,
+ "step": 63
+ },
+ {
+ "epoch": 0.20447284345047922,
+ "grad_norm": 0.326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.3482,
+ "step": 64
+ },
+ {
+ "epoch": 0.20766773162939298,
+ "grad_norm": 0.326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.3185,
+ "step": 65
+ },
+ {
+ "epoch": 0.2108626198083067,
+ "grad_norm": 0.298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.3195,
+ "step": 66
+ },
+ {
+ "epoch": 0.21405750798722045,
+ "grad_norm": 0.59765625,
+ "learning_rate": 0.0005,
+ "loss": 1.3362,
+ "step": 67
+ },
+ {
+ "epoch": 0.21725239616613418,
+ "grad_norm": 2.6875,
+ "learning_rate": 0.0005,
+ "loss": 1.3411,
+ "step": 68
+ },
+ {
+ "epoch": 0.22044728434504793,
+ "grad_norm": 0.98828125,
+ "learning_rate": 0.0005,
+ "loss": 1.3137,
+ "step": 69
+ },
+ {
+ "epoch": 0.22364217252396165,
+ "grad_norm": 1.421875,
+ "learning_rate": 0.0005,
+ "loss": 1.3276,
+ "step": 70
+ },
+ {
+ "epoch": 0.2268370607028754,
+ "grad_norm": 2.09375,
+ "learning_rate": 0.0005,
+ "loss": 1.3308,
+ "step": 71
+ },
+ {
+ "epoch": 0.23003194888178913,
+ "grad_norm": 0.2021484375,
+ "learning_rate": 0.0005,
+ "loss": 1.2752,
+ "step": 72
+ },
+ {
+ "epoch": 0.23322683706070288,
+ "grad_norm": 1.75,
+ "learning_rate": 0.0005,
+ "loss": 1.2899,
+ "step": 73
+ },
+ {
+ "epoch": 0.2364217252396166,
+ "grad_norm": 1.171875,
+ "learning_rate": 0.0005,
+ "loss": 1.2757,
+ "step": 74
+ },
+ {
+ "epoch": 0.23961661341853036,
+ "grad_norm": 0.52734375,
+ "learning_rate": 0.0005,
+ "loss": 1.2816,
+ "step": 75
+ },
+ {
+ "epoch": 0.24281150159744408,
+ "grad_norm": 0.1337890625,
+ "learning_rate": 0.0005,
+ "loss": 1.2628,
+ "step": 76
+ },
+ {
+ "epoch": 0.24600638977635783,
+ "grad_norm": 0.490234375,
+ "learning_rate": 0.0005,
+ "loss": 1.2807,
+ "step": 77
+ },
+ {
+ "epoch": 0.24920127795527156,
+ "grad_norm": 0.84375,
+ "learning_rate": 0.0005,
+ "loss": 1.2723,
+ "step": 78
+ },
+ {
+ "epoch": 0.2523961661341853,
+ "grad_norm": 1.53125,
+ "learning_rate": 0.0005,
+ "loss": 1.2665,
+ "step": 79
+ },
+ {
+ "epoch": 0.25559105431309903,
+ "grad_norm": 0.5390625,
+ "learning_rate": 0.0005,
+ "loss": 1.2679,
+ "step": 80
+ },
+ {
+ "epoch": 0.25878594249201275,
+ "grad_norm": 0.302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.2648,
+ "step": 81
+ },
+ {
+ "epoch": 0.26198083067092653,
+ "grad_norm": 0.95703125,
+ "learning_rate": 0.0005,
+ "loss": 1.2625,
+ "step": 82
+ },
+ {
+ "epoch": 0.26517571884984026,
+ "grad_norm": 2.265625,
+ "learning_rate": 0.0005,
+ "loss": 1.2852,
+ "step": 83
+ },
+ {
+ "epoch": 0.268370607028754,
+ "grad_norm": 0.494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.2501,
+ "step": 84
+ },
+ {
+ "epoch": 0.2715654952076677,
+ "grad_norm": 3.578125,
+ "learning_rate": 0.0005,
+ "loss": 1.3461,
+ "step": 85
+ },
+ {
+ "epoch": 0.2747603833865815,
+ "grad_norm": 2.5625,
+ "learning_rate": 0.0005,
+ "loss": 1.3066,
+ "step": 86
+ },
+ {
+ "epoch": 0.2779552715654952,
+ "grad_norm": 2.578125,
+ "learning_rate": 0.0005,
+ "loss": 1.3007,
+ "step": 87
+ },
+ {
+ "epoch": 0.28115015974440893,
+ "grad_norm": 2.3125,
+ "learning_rate": 0.0005,
+ "loss": 1.299,
+ "step": 88
+ },
+ {
+ "epoch": 0.28434504792332266,
+ "grad_norm": 1.2578125,
+ "learning_rate": 0.0005,
+ "loss": 1.2455,
+ "step": 89
+ },
+ {
+ "epoch": 0.28753993610223644,
+ "grad_norm": 1.1953125,
+ "learning_rate": 0.0005,
+ "loss": 1.2684,
+ "step": 90
+ },
+ {
+ "epoch": 0.29073482428115016,
+ "grad_norm": 2.625,
+ "learning_rate": 0.0005,
+ "loss": 1.2668,
+ "step": 91
+ },
+ {
+ "epoch": 0.2939297124600639,
+ "grad_norm": 2.421875,
+ "learning_rate": 0.0005,
+ "loss": 1.2911,
+ "step": 92
+ },
+ {
+ "epoch": 0.2971246006389776,
+ "grad_norm": 0.8125,
+ "learning_rate": 0.0005,
+ "loss": 1.2512,
+ "step": 93
+ },
+ {
+ "epoch": 0.3003194888178914,
+ "grad_norm": 0.93359375,
+ "learning_rate": 0.0005,
+ "loss": 1.2686,
+ "step": 94
+ },
+ {
+ "epoch": 0.3035143769968051,
+ "grad_norm": 1.296875,
+ "learning_rate": 0.0005,
+ "loss": 1.2513,
+ "step": 95
+ },
+ {
+ "epoch": 0.30670926517571884,
+ "grad_norm": 0.63671875,
+ "learning_rate": 0.0005,
+ "loss": 1.2451,
+ "step": 96
+ },
+ {
+ "epoch": 0.30990415335463256,
+ "grad_norm": 1.4296875,
+ "learning_rate": 0.0005,
+ "loss": 1.2626,
+ "step": 97
+ },
+ {
+ "epoch": 0.31309904153354634,
+ "grad_norm": 0.74609375,
+ "learning_rate": 0.0005,
+ "loss": 1.235,
+ "step": 98
+ },
+ {
+ "epoch": 0.31629392971246006,
+ "grad_norm": 1.296875,
+ "learning_rate": 0.0005,
+ "loss": 1.2582,
+ "step": 99
+ },
+ {
+ "epoch": 0.3194888178913738,
+ "grad_norm": 0.85546875,
+ "learning_rate": 0.0005,
+ "loss": 1.2426,
+ "step": 100
+ },
+ {
+ "epoch": 0.3226837060702875,
+ "grad_norm": 2.1875,
+ "learning_rate": 0.0005,
+ "loss": 1.2789,
+ "step": 101
+ },
+ {
+ "epoch": 0.3258785942492013,
+ "grad_norm": 1.8359375,
+ "learning_rate": 0.0005,
+ "loss": 1.2725,
+ "step": 102
+ },
+ {
+ "epoch": 0.329073482428115,
+ "grad_norm": 0.7578125,
+ "learning_rate": 0.0005,
+ "loss": 1.2357,
+ "step": 103
+ },
+ {
+ "epoch": 0.33226837060702874,
+ "grad_norm": 0.7265625,
+ "learning_rate": 0.0005,
+ "loss": 1.2328,
+ "step": 104
+ },
+ {
+ "epoch": 0.3354632587859425,
+ "grad_norm": 0.34375,
+ "learning_rate": 0.0005,
+ "loss": 1.2328,
+ "step": 105
+ },
+ {
+ "epoch": 0.33865814696485624,
+ "grad_norm": 0.59765625,
+ "learning_rate": 0.0005,
+ "loss": 1.2448,
+ "step": 106
+ },
+ {
+ "epoch": 0.34185303514376997,
+ "grad_norm": 0.2177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.2384,
+ "step": 107
+ },
+ {
+ "epoch": 0.3450479233226837,
+ "grad_norm": 0.439453125,
+ "learning_rate": 0.0005,
+ "loss": 1.2179,
+ "step": 108
+ },
+ {
+ "epoch": 0.34824281150159747,
+ "grad_norm": 0.22265625,
+ "learning_rate": 0.0005,
+ "loss": 1.2237,
+ "step": 109
+ },
+ {
+ "epoch": 0.3514376996805112,
+ "grad_norm": 0.353515625,
+ "learning_rate": 0.0005,
+ "loss": 1.2199,
+ "step": 110
+ },
+ {
+ "epoch": 0.3546325878594249,
+ "grad_norm": 0.30859375,
+ "learning_rate": 0.0005,
+ "loss": 1.2015,
+ "step": 111
+ },
+ {
+ "epoch": 0.35782747603833864,
+ "grad_norm": 0.16015625,
+ "learning_rate": 0.0005,
+ "loss": 1.2006,
+ "step": 112
+ },
+ {
+ "epoch": 0.3610223642172524,
+ "grad_norm": 0.30859375,
+ "learning_rate": 0.0005,
+ "loss": 1.1853,
+ "step": 113
+ },
+ {
+ "epoch": 0.36421725239616615,
+ "grad_norm": 0.1669921875,
+ "learning_rate": 0.0005,
+ "loss": 1.2047,
+ "step": 114
+ },
+ {
+ "epoch": 0.36741214057507987,
+ "grad_norm": 0.265625,
+ "learning_rate": 0.0005,
+ "loss": 1.2196,
+ "step": 115
+ },
+ {
+ "epoch": 0.3706070287539936,
+ "grad_norm": 0.2060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.1936,
+ "step": 116
+ },
+ {
+ "epoch": 0.3738019169329074,
+ "grad_norm": 0.1455078125,
+ "learning_rate": 0.0005,
+ "loss": 1.2152,
+ "step": 117
+ },
+ {
+ "epoch": 0.3769968051118211,
+ "grad_norm": 0.2001953125,
+ "learning_rate": 0.0005,
+ "loss": 1.2134,
+ "step": 118
+ },
+ {
+ "epoch": 0.3801916932907348,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.2019,
+ "step": 119
+ },
+ {
+ "epoch": 0.38338658146964855,
+ "grad_norm": 0.12109375,
+ "learning_rate": 0.0005,
+ "loss": 1.2046,
+ "step": 120
+ },
+ {
+ "epoch": 0.3865814696485623,
+ "grad_norm": 0.12109375,
+ "learning_rate": 0.0005,
+ "loss": 1.2143,
+ "step": 121
+ },
+ {
+ "epoch": 0.38977635782747605,
+ "grad_norm": 0.0859375,
+ "learning_rate": 0.0005,
+ "loss": 1.2057,
+ "step": 122
+ },
+ {
+ "epoch": 0.3929712460063898,
+ "grad_norm": 0.11962890625,
+ "learning_rate": 0.0005,
+ "loss": 1.2077,
+ "step": 123
+ },
+ {
+ "epoch": 0.3961661341853035,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.2256,
+ "step": 124
+ },
+ {
+ "epoch": 0.3993610223642173,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.2145,
+ "step": 125
+ },
+ {
+ "epoch": 0.402555910543131,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.197,
+ "step": 126
+ },
+ {
+ "epoch": 0.4057507987220447,
+ "grad_norm": 0.0947265625,
+ "learning_rate": 0.0005,
+ "loss": 1.1964,
+ "step": 127
+ },
+ {
+ "epoch": 0.40894568690095845,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.1948,
+ "step": 128
+ },
+ {
+ "epoch": 0.41214057507987223,
+ "grad_norm": 0.11279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.1961,
+ "step": 129
+ },
+ {
+ "epoch": 0.41533546325878595,
+ "grad_norm": 0.08203125,
+ "learning_rate": 0.0005,
+ "loss": 1.1928,
+ "step": 130
+ },
+ {
+ "epoch": 0.4185303514376997,
+ "grad_norm": 0.11767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.196,
+ "step": 131
+ },
+ {
+ "epoch": 0.4217252396166134,
+ "grad_norm": 0.10986328125,
+ "learning_rate": 0.0005,
+ "loss": 1.2027,
+ "step": 132
+ },
+ {
+ "epoch": 0.4249201277955272,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.1935,
+ "step": 133
+ },
+ {
+ "epoch": 0.4281150159744409,
+ "grad_norm": 0.1328125,
+ "learning_rate": 0.0005,
+ "loss": 1.1768,
+ "step": 134
+ },
+ {
+ "epoch": 0.43130990415335463,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.1959,
+ "step": 135
+ },
+ {
+ "epoch": 0.43450479233226835,
+ "grad_norm": 0.1279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.2068,
+ "step": 136
+ },
+ {
+ "epoch": 0.43769968051118213,
+ "grad_norm": 0.1259765625,
+ "learning_rate": 0.0005,
+ "loss": 1.1837,
+ "step": 137
+ },
+ {
+ "epoch": 0.44089456869009586,
+ "grad_norm": 0.0810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.2078,
+ "step": 138
+ },
+ {
+ "epoch": 0.4440894568690096,
+ "grad_norm": 0.17578125,
+ "learning_rate": 0.0005,
+ "loss": 1.1934,
+ "step": 139
+ },
+ {
+ "epoch": 0.4472843450479233,
+ "grad_norm": 0.0888671875,
+ "learning_rate": 0.0005,
+ "loss": 1.188,
+ "step": 140
+ },
+ {
+ "epoch": 0.4504792332268371,
+ "grad_norm": 0.11328125,
+ "learning_rate": 0.0005,
+ "loss": 1.1848,
+ "step": 141
+ },
+ {
+ "epoch": 0.4536741214057508,
+ "grad_norm": 0.109375,
+ "learning_rate": 0.0005,
+ "loss": 1.1918,
+ "step": 142
+ },
+ {
+ "epoch": 0.45686900958466453,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.1872,
+ "step": 143
+ },
+ {
+ "epoch": 0.46006389776357826,
+ "grad_norm": 0.103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.1884,
+ "step": 144
+ },
+ {
+ "epoch": 0.46325878594249204,
+ "grad_norm": 0.1171875,
+ "learning_rate": 0.0005,
+ "loss": 1.1711,
+ "step": 145
+ },
+ {
+ "epoch": 0.46645367412140576,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.1978,
+ "step": 146
+ },
+ {
+ "epoch": 0.4696485623003195,
+ "grad_norm": 0.1484375,
+ "learning_rate": 0.0005,
+ "loss": 1.1813,
+ "step": 147
+ },
+ {
+ "epoch": 0.4728434504792332,
+ "grad_norm": 0.11181640625,
+ "learning_rate": 0.0005,
+ "loss": 1.1796,
+ "step": 148
+ },
+ {
+ "epoch": 0.476038338658147,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.1908,
+ "step": 149
+ },
+ {
+ "epoch": 0.4792332268370607,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.1841,
+ "step": 150
+ },
+ {
+ "epoch": 0.48242811501597443,
+ "grad_norm": 0.09326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.1825,
+ "step": 151
+ },
+ {
+ "epoch": 0.48562300319488816,
+ "grad_norm": 0.10009765625,
+ "learning_rate": 0.0005,
+ "loss": 1.1762,
+ "step": 152
+ },
+ {
+ "epoch": 0.48881789137380194,
+ "grad_norm": 0.111328125,
+ "learning_rate": 0.0005,
+ "loss": 1.1857,
+ "step": 153
+ },
+ {
+ "epoch": 0.49201277955271566,
+ "grad_norm": 0.1787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.1803,
+ "step": 154
+ },
+ {
+ "epoch": 0.4952076677316294,
+ "grad_norm": 0.30078125,
+ "learning_rate": 0.0005,
+ "loss": 1.18,
+ "step": 155
+ },
+ {
+ "epoch": 0.4984025559105431,
+ "grad_norm": 0.5859375,
+ "learning_rate": 0.0005,
+ "loss": 1.1833,
+ "step": 156
+ },
+ {
+ "epoch": 0.5015974440894568,
+ "grad_norm": 1.6171875,
+ "learning_rate": 0.0005,
+ "loss": 1.2157,
+ "step": 157
+ },
+ {
+ "epoch": 0.5047923322683706,
+ "grad_norm": 1.171875,
+ "learning_rate": 0.0005,
+ "loss": 1.1896,
+ "step": 158
+ },
+ {
+ "epoch": 0.5079872204472844,
+ "grad_norm": 0.51953125,
+ "learning_rate": 0.0005,
+ "loss": 1.1791,
+ "step": 159
+ },
+ {
+ "epoch": 0.5111821086261981,
+ "grad_norm": 0.98046875,
+ "learning_rate": 0.0005,
+ "loss": 1.1843,
+ "step": 160
+ },
+ {
+ "epoch": 0.5143769968051118,
+ "grad_norm": 1.375,
+ "learning_rate": 0.0005,
+ "loss": 1.2064,
+ "step": 161
+ },
+ {
+ "epoch": 0.5175718849840255,
+ "grad_norm": 0.60546875,
+ "learning_rate": 0.0005,
+ "loss": 1.2178,
+ "step": 162
+ },
+ {
+ "epoch": 0.5207667731629393,
+ "grad_norm": 0.8359375,
+ "learning_rate": 0.0005,
+ "loss": 1.2047,
+ "step": 163
+ },
+ {
+ "epoch": 0.5239616613418531,
+ "grad_norm": 1.0078125,
+ "learning_rate": 0.0005,
+ "loss": 1.1627,
+ "step": 164
+ },
+ {
+ "epoch": 0.5271565495207667,
+ "grad_norm": 0.79296875,
+ "learning_rate": 0.0005,
+ "loss": 1.2085,
+ "step": 165
+ },
+ {
+ "epoch": 0.5303514376996805,
+ "grad_norm": 0.40234375,
+ "learning_rate": 0.0005,
+ "loss": 1.1637,
+ "step": 166
+ },
+ {
+ "epoch": 0.5335463258785943,
+ "grad_norm": 0.6328125,
+ "learning_rate": 0.0005,
+ "loss": 1.1917,
+ "step": 167
+ },
+ {
+ "epoch": 0.536741214057508,
+ "grad_norm": 0.8359375,
+ "learning_rate": 0.0005,
+ "loss": 1.1973,
+ "step": 168
+ },
+ {
+ "epoch": 0.5399361022364217,
+ "grad_norm": 0.48046875,
+ "learning_rate": 0.0005,
+ "loss": 1.1796,
+ "step": 169
+ },
+ {
+ "epoch": 0.5431309904153354,
+ "grad_norm": 0.333984375,
+ "learning_rate": 0.0005,
+ "loss": 1.1863,
+ "step": 170
+ },
+ {
+ "epoch": 0.5463258785942492,
+ "grad_norm": 0.62890625,
+ "learning_rate": 0.0005,
+ "loss": 1.1672,
+ "step": 171
+ },
+ {
+ "epoch": 0.549520766773163,
+ "grad_norm": 0.28125,
+ "learning_rate": 0.0005,
+ "loss": 1.1634,
+ "step": 172
+ },
+ {
+ "epoch": 0.5527156549520766,
+ "grad_norm": 0.3984375,
+ "learning_rate": 0.0005,
+ "loss": 1.1817,
+ "step": 173
+ },
+ {
+ "epoch": 0.5559105431309904,
+ "grad_norm": 0.255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.1746,
+ "step": 174
+ },
+ {
+ "epoch": 0.5591054313099042,
+ "grad_norm": 0.279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.18,
+ "step": 175
+ },
+ {
+ "epoch": 0.5623003194888179,
+ "grad_norm": 0.1806640625,
+ "learning_rate": 0.0005,
+ "loss": 1.1703,
+ "step": 176
+ },
+ {
+ "epoch": 0.5654952076677316,
+ "grad_norm": 0.236328125,
+ "learning_rate": 0.0005,
+ "loss": 1.1728,
+ "step": 177
+ },
+ {
+ "epoch": 0.5686900958466453,
+ "grad_norm": 0.1650390625,
+ "learning_rate": 0.0005,
+ "loss": 1.1784,
+ "step": 178
+ },
+ {
+ "epoch": 0.5718849840255591,
+ "grad_norm": 0.1298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.1595,
+ "step": 179
+ },
+ {
+ "epoch": 0.5750798722044729,
+ "grad_norm": 0.09375,
+ "learning_rate": 0.0005,
+ "loss": 1.143,
+ "step": 180
+ },
+ {
+ "epoch": 0.5782747603833865,
+ "grad_norm": 0.1708984375,
+ "learning_rate": 0.0005,
+ "loss": 1.1713,
+ "step": 181
+ },
+ {
+ "epoch": 0.5814696485623003,
+ "grad_norm": 0.087890625,
+ "learning_rate": 0.0005,
+ "loss": 1.1782,
+ "step": 182
+ },
+ {
+ "epoch": 0.5846645367412141,
+ "grad_norm": 0.197265625,
+ "learning_rate": 0.0005,
+ "loss": 1.1569,
+ "step": 183
+ },
+ {
+ "epoch": 0.5878594249201278,
+ "grad_norm": 0.09423828125,
+ "learning_rate": 0.0005,
+ "loss": 1.1721,
+ "step": 184
+ },
+ {
+ "epoch": 0.5910543130990416,
+ "grad_norm": 0.134765625,
+ "learning_rate": 0.0005,
+ "loss": 1.1435,
+ "step": 185
+ },
+ {
+ "epoch": 0.5942492012779552,
+ "grad_norm": 0.09228515625,
+ "learning_rate": 0.0005,
+ "loss": 1.1515,
+ "step": 186
+ },
+ {
+ "epoch": 0.597444089456869,
+ "grad_norm": 0.083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.1584,
+ "step": 187
+ },
+ {
+ "epoch": 0.6006389776357828,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.1592,
+ "step": 188
+ },
+ {
+ "epoch": 0.6038338658146964,
+ "grad_norm": 0.091796875,
+ "learning_rate": 0.0005,
+ "loss": 1.1505,
+ "step": 189
+ },
+ {
+ "epoch": 0.6070287539936102,
+ "grad_norm": 0.10498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.1591,
+ "step": 190
+ },
+ {
+ "epoch": 0.610223642172524,
+ "grad_norm": 0.08251953125,
+ "learning_rate": 0.0005,
+ "loss": 1.1742,
+ "step": 191
+ },
+ {
+ "epoch": 0.6134185303514377,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.1524,
+ "step": 192
+ },
+ {
+ "epoch": 0.6166134185303515,
+ "grad_norm": 0.09716796875,
+ "learning_rate": 0.0005,
+ "loss": 1.1746,
+ "step": 193
+ },
+ {
+ "epoch": 0.6198083067092651,
+ "grad_norm": 0.09814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.1648,
+ "step": 194
+ },
+ {
+ "epoch": 0.6230031948881789,
+ "grad_norm": 0.11865234375,
+ "learning_rate": 0.0005,
+ "loss": 1.1738,
+ "step": 195
+ },
+ {
+ "epoch": 0.6261980830670927,
+ "grad_norm": 0.060791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.1687,
+ "step": 196
+ },
+ {
+ "epoch": 0.6293929712460063,
+ "grad_norm": 0.11669921875,
+ "learning_rate": 0.0005,
+ "loss": 1.1626,
+ "step": 197
+ },
+ {
+ "epoch": 0.6325878594249201,
+ "grad_norm": 0.083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.1629,
+ "step": 198
+ },
+ {
+ "epoch": 0.6357827476038339,
+ "grad_norm": 0.111328125,
+ "learning_rate": 0.0005,
+ "loss": 1.1569,
+ "step": 199
+ },
+ {
+ "epoch": 0.6389776357827476,
+ "grad_norm": 0.08984375,
+ "learning_rate": 0.0005,
+ "loss": 1.1499,
+ "step": 200
+ },
+ {
+ "epoch": 0.6421725239616614,
+ "grad_norm": 0.08935546875,
+ "learning_rate": 0.0005,
+ "loss": 1.1733,
+ "step": 201
+ },
+ {
+ "epoch": 0.645367412140575,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.1627,
+ "step": 202
+ },
+ {
+ "epoch": 0.6485623003194888,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.1507,
+ "step": 203
+ },
+ {
+ "epoch": 0.6517571884984026,
+ "grad_norm": 0.08642578125,
+ "learning_rate": 0.0005,
+ "loss": 1.1558,
+ "step": 204
+ },
+ {
+ "epoch": 0.6549520766773163,
+ "grad_norm": 0.10791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.1458,
+ "step": 205
+ },
+ {
+ "epoch": 0.65814696485623,
+ "grad_norm": 0.12158203125,
+ "learning_rate": 0.0005,
+ "loss": 1.1474,
+ "step": 206
+ },
+ {
+ "epoch": 0.6613418530351438,
+ "grad_norm": 0.13671875,
+ "learning_rate": 0.0005,
+ "loss": 1.1791,
+ "step": 207
+ },
+ {
+ "epoch": 0.6645367412140575,
+ "grad_norm": 0.0810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.1428,
+ "step": 208
+ },
+ {
+ "epoch": 0.6677316293929713,
+ "grad_norm": 0.11669921875,
+ "learning_rate": 0.0005,
+ "loss": 1.1496,
+ "step": 209
+ },
+ {
+ "epoch": 0.670926517571885,
+ "grad_norm": 0.0830078125,
+ "learning_rate": 0.0005,
+ "loss": 1.1508,
+ "step": 210
+ },
+ {
+ "epoch": 0.6741214057507987,
+ "grad_norm": 0.1484375,
+ "learning_rate": 0.0005,
+ "loss": 1.173,
+ "step": 211
+ },
+ {
+ "epoch": 0.6773162939297125,
+ "grad_norm": 0.1865234375,
+ "learning_rate": 0.0005,
+ "loss": 1.166,
+ "step": 212
+ },
+ {
+ "epoch": 0.6805111821086262,
+ "grad_norm": 0.25390625,
+ "learning_rate": 0.0005,
+ "loss": 1.1448,
+ "step": 213
+ },
+ {
+ "epoch": 0.6837060702875399,
+ "grad_norm": 0.51171875,
+ "learning_rate": 0.0005,
+ "loss": 1.147,
+ "step": 214
+ },
+ {
+ "epoch": 0.6869009584664537,
+ "grad_norm": 1.171875,
+ "learning_rate": 0.0005,
+ "loss": 1.1662,
+ "step": 215
+ },
+ {
+ "epoch": 0.6900958466453674,
+ "grad_norm": 2.203125,
+ "learning_rate": 0.0005,
+ "loss": 1.1905,
+ "step": 216
+ },
+ {
+ "epoch": 0.6932907348242812,
+ "grad_norm": 0.60546875,
+ "learning_rate": 0.0005,
+ "loss": 1.1664,
+ "step": 217
+ },
+ {
+ "epoch": 0.6964856230031949,
+ "grad_norm": 3.21875,
+ "learning_rate": 0.0005,
+ "loss": 1.2114,
+ "step": 218
+ },
+ {
+ "epoch": 0.6996805111821086,
+ "grad_norm": 2.296875,
+ "learning_rate": 0.0005,
+ "loss": 1.2394,
+ "step": 219
+ },
+ {
+ "epoch": 0.7028753993610224,
+ "grad_norm": 1.4609375,
+ "learning_rate": 0.0005,
+ "loss": 1.1962,
+ "step": 220
+ },
+ {
+ "epoch": 0.7060702875399361,
+ "grad_norm": 1.8125,
+ "learning_rate": 0.0005,
+ "loss": 1.2004,
+ "step": 221
+ },
+ {
+ "epoch": 0.7092651757188498,
+ "grad_norm": 1.8984375,
+ "learning_rate": 0.0005,
+ "loss": 1.1989,
+ "step": 222
+ },
+ {
+ "epoch": 0.7124600638977636,
+ "grad_norm": 0.6953125,
+ "learning_rate": 0.0005,
+ "loss": 1.1755,
+ "step": 223
+ },
+ {
+ "epoch": 0.7156549520766773,
+ "grad_norm": 0.95703125,
+ "learning_rate": 0.0005,
+ "loss": 1.1922,
+ "step": 224
+ },
+ {
+ "epoch": 0.7188498402555911,
+ "grad_norm": 0.4296875,
+ "learning_rate": 0.0005,
+ "loss": 1.1809,
+ "step": 225
+ },
+ {
+ "epoch": 0.7220447284345048,
+ "grad_norm": 1.0390625,
+ "learning_rate": 0.0005,
+ "loss": 1.1795,
+ "step": 226
+ },
+ {
+ "epoch": 0.7252396166134185,
+ "grad_norm": 0.447265625,
+ "learning_rate": 0.0005,
+ "loss": 1.1676,
+ "step": 227
+ },
+ {
+ "epoch": 0.7284345047923323,
+ "grad_norm": 0.3515625,
+ "learning_rate": 0.0005,
+ "loss": 1.1767,
+ "step": 228
+ },
+ {
+ "epoch": 0.731629392971246,
+ "grad_norm": 0.310546875,
+ "learning_rate": 0.0005,
+ "loss": 1.1672,
+ "step": 229
+ },
+ {
+ "epoch": 0.7348242811501597,
+ "grad_norm": 0.1953125,
+ "learning_rate": 0.0005,
+ "loss": 1.1661,
+ "step": 230
+ },
+ {
+ "epoch": 0.7380191693290735,
+ "grad_norm": 0.2431640625,
+ "learning_rate": 0.0005,
+ "loss": 1.1504,
+ "step": 231
+ },
+ {
+ "epoch": 0.7412140575079872,
+ "grad_norm": 0.15625,
+ "learning_rate": 0.0005,
+ "loss": 1.1709,
+ "step": 232
+ },
+ {
+ "epoch": 0.744408945686901,
+ "grad_norm": 0.181640625,
+ "learning_rate": 0.0005,
+ "loss": 1.164,
+ "step": 233
+ },
+ {
+ "epoch": 0.7476038338658147,
+ "grad_norm": 0.123046875,
+ "learning_rate": 0.0005,
+ "loss": 1.1453,
+ "step": 234
+ },
+ {
+ "epoch": 0.7507987220447284,
+ "grad_norm": 0.138671875,
+ "learning_rate": 0.0005,
+ "loss": 1.1583,
+ "step": 235
+ },
+ {
+ "epoch": 0.7539936102236422,
+ "grad_norm": 0.11669921875,
+ "learning_rate": 0.0005,
+ "loss": 1.1439,
+ "step": 236
+ },
+ {
+ "epoch": 0.7571884984025559,
+ "grad_norm": 0.1298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.1535,
+ "step": 237
+ },
+ {
+ "epoch": 0.7603833865814696,
+ "grad_norm": 0.091796875,
+ "learning_rate": 0.0005,
+ "loss": 1.1706,
+ "step": 238
+ },
+ {
+ "epoch": 0.7635782747603834,
+ "grad_norm": 0.173828125,
+ "learning_rate": 0.0005,
+ "loss": 1.1639,
+ "step": 239
+ },
+ {
+ "epoch": 0.7667731629392971,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.1548,
+ "step": 240
+ },
+ {
+ "epoch": 0.7699680511182109,
+ "grad_norm": 0.1357421875,
+ "learning_rate": 0.0005,
+ "loss": 1.1484,
+ "step": 241
+ },
+ {
+ "epoch": 0.7731629392971247,
+ "grad_norm": 0.0908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.1398,
+ "step": 242
+ },
+ {
+ "epoch": 0.7763578274760383,
+ "grad_norm": 0.08935546875,
+ "learning_rate": 0.0005,
+ "loss": 1.1482,
+ "step": 243
+ },
+ {
+ "epoch": 0.7795527156549521,
+ "grad_norm": 0.08251953125,
+ "learning_rate": 0.0005,
+ "loss": 1.15,
+ "step": 244
+ },
+ {
+ "epoch": 0.7827476038338658,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.1677,
+ "step": 245
+ },
+ {
+ "epoch": 0.7859424920127795,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.1551,
+ "step": 246
+ },
+ {
+ "epoch": 0.7891373801916933,
+ "grad_norm": 0.0986328125,
+ "learning_rate": 0.0005,
+ "loss": 1.1487,
+ "step": 247
+ },
+ {
+ "epoch": 0.792332268370607,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.1525,
+ "step": 248
+ },
+ {
+ "epoch": 0.7955271565495208,
+ "grad_norm": 0.1103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.1437,
+ "step": 249
+ },
+ {
+ "epoch": 0.7987220447284346,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.1541,
+ "step": 250
+ },
+ {
+ "epoch": 0.8019169329073482,
+ "grad_norm": 0.1005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.1628,
+ "step": 251
+ },
+ {
+ "epoch": 0.805111821086262,
+ "grad_norm": 0.07861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.157,
+ "step": 252
+ },
+ {
+ "epoch": 0.8083067092651757,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.1532,
+ "step": 253
+ },
+ {
+ "epoch": 0.8115015974440895,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.1421,
+ "step": 254
+ },
+ {
+ "epoch": 0.8146964856230032,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.1445,
+ "step": 255
+ },
+ {
+ "epoch": 0.8178913738019169,
+ "grad_norm": 0.05859375,
+ "learning_rate": 0.0005,
+ "loss": 1.1463,
+ "step": 256
+ },
+ {
+ "epoch": 0.8210862619808307,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.1502,
+ "step": 257
+ },
+ {
+ "epoch": 0.8242811501597445,
+ "grad_norm": 0.059326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.1399,
+ "step": 258
+ },
+ {
+ "epoch": 0.8274760383386581,
+ "grad_norm": 0.08935546875,
+ "learning_rate": 0.0005,
+ "loss": 1.1536,
+ "step": 259
+ },
+ {
+ "epoch": 0.8306709265175719,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.1419,
+ "step": 260
+ },
+ {
+ "epoch": 0.8338658146964856,
+ "grad_norm": 0.08837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.1397,
+ "step": 261
+ },
+ {
+ "epoch": 0.8370607028753994,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.14,
+ "step": 262
+ },
+ {
+ "epoch": 0.8402555910543131,
+ "grad_norm": 0.08447265625,
+ "learning_rate": 0.0005,
+ "loss": 1.1598,
+ "step": 263
+ },
+ {
+ "epoch": 0.8434504792332268,
+ "grad_norm": 0.058349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.1507,
+ "step": 264
+ },
+ {
+ "epoch": 0.8466453674121406,
+ "grad_norm": 0.0986328125,
+ "learning_rate": 0.0005,
+ "loss": 1.1413,
+ "step": 265
+ },
+ {
+ "epoch": 0.8498402555910544,
+ "grad_norm": 0.059326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.1515,
+ "step": 266
+ },
+ {
+ "epoch": 0.853035143769968,
+ "grad_norm": 0.0888671875,
+ "learning_rate": 0.0005,
+ "loss": 1.137,
+ "step": 267
+ },
+ {
+ "epoch": 0.8562300319488818,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.1624,
+ "step": 268
+ },
+ {
+ "epoch": 0.8594249201277955,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.1487,
+ "step": 269
+ },
+ {
+ "epoch": 0.8626198083067093,
+ "grad_norm": 0.060791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.1423,
+ "step": 270
+ },
+ {
+ "epoch": 0.865814696485623,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.1492,
+ "step": 271
+ },
+ {
+ "epoch": 0.8690095846645367,
+ "grad_norm": 0.10009765625,
+ "learning_rate": 0.0005,
+ "loss": 1.154,
+ "step": 272
+ },
+ {
+ "epoch": 0.8722044728434505,
+ "grad_norm": 0.1103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.1443,
+ "step": 273
+ },
+ {
+ "epoch": 0.8753993610223643,
+ "grad_norm": 0.08251953125,
+ "learning_rate": 0.0005,
+ "loss": 1.1401,
+ "step": 274
+ },
+ {
+ "epoch": 0.8785942492012779,
+ "grad_norm": 0.083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.1357,
+ "step": 275
+ },
+ {
+ "epoch": 0.8817891373801917,
+ "grad_norm": 0.080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.1592,
+ "step": 276
+ },
+ {
+ "epoch": 0.8849840255591054,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.1389,
+ "step": 277
+ },
+ {
+ "epoch": 0.8881789137380192,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.1415,
+ "step": 278
+ },
+ {
+ "epoch": 0.8913738019169329,
+ "grad_norm": 0.060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.1409,
+ "step": 279
+ },
+ {
+ "epoch": 0.8945686900958466,
+ "grad_norm": 0.08544921875,
+ "learning_rate": 0.0005,
+ "loss": 1.1409,
+ "step": 280
+ },
+ {
+ "epoch": 0.8977635782747604,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.1351,
+ "step": 281
+ },
+ {
+ "epoch": 0.9009584664536742,
+ "grad_norm": 0.1123046875,
+ "learning_rate": 0.0005,
+ "loss": 1.1404,
+ "step": 282
+ },
+ {
+ "epoch": 0.9041533546325878,
+ "grad_norm": 0.11376953125,
+ "learning_rate": 0.0005,
+ "loss": 1.1374,
+ "step": 283
+ },
+ {
+ "epoch": 0.9073482428115016,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.1445,
+ "step": 284
+ },
+ {
+ "epoch": 0.9105431309904153,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.1367,
+ "step": 285
+ },
+ {
+ "epoch": 0.9137380191693291,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.1306,
+ "step": 286
+ },
+ {
+ "epoch": 0.9169329073482428,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.149,
+ "step": 287
+ },
+ {
+ "epoch": 0.9201277955271565,
+ "grad_norm": 0.08544921875,
+ "learning_rate": 0.0005,
+ "loss": 1.1514,
+ "step": 288
+ },
+ {
+ "epoch": 0.9233226837060703,
+ "grad_norm": 0.080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.1292,
+ "step": 289
+ },
+ {
+ "epoch": 0.9265175718849841,
+ "grad_norm": 0.08203125,
+ "learning_rate": 0.0005,
+ "loss": 1.1252,
+ "step": 290
+ },
+ {
+ "epoch": 0.9297124600638977,
+ "grad_norm": 0.1318359375,
+ "learning_rate": 0.0005,
+ "loss": 1.1296,
+ "step": 291
+ },
+ {
+ "epoch": 0.9329073482428115,
+ "grad_norm": 0.140625,
+ "learning_rate": 0.0005,
+ "loss": 1.1361,
+ "step": 292
+ },
+ {
+ "epoch": 0.9361022364217252,
+ "grad_norm": 0.095703125,
+ "learning_rate": 0.0005,
+ "loss": 1.1491,
+ "step": 293
+ },
+ {
+ "epoch": 0.939297124600639,
+ "grad_norm": 0.10693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.1432,
+ "step": 294
+ },
+ {
+ "epoch": 0.9424920127795527,
+ "grad_norm": 0.140625,
+ "learning_rate": 0.0005,
+ "loss": 1.1411,
+ "step": 295
+ },
+ {
+ "epoch": 0.9456869009584664,
+ "grad_norm": 0.1005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.1343,
+ "step": 296
+ },
+ {
+ "epoch": 0.9488817891373802,
+ "grad_norm": 0.1376953125,
+ "learning_rate": 0.0005,
+ "loss": 1.1524,
+ "step": 297
+ },
+ {
+ "epoch": 0.952076677316294,
+ "grad_norm": 0.1328125,
+ "learning_rate": 0.0005,
+ "loss": 1.1362,
+ "step": 298
+ },
+ {
+ "epoch": 0.9552715654952076,
+ "grad_norm": 0.166015625,
+ "learning_rate": 0.0005,
+ "loss": 1.1309,
+ "step": 299
+ },
+ {
+ "epoch": 0.9584664536741214,
+ "grad_norm": 0.2041015625,
+ "learning_rate": 0.0005,
+ "loss": 1.1412,
+ "step": 300
+ },
+ {
+ "epoch": 0.9616613418530351,
+ "grad_norm": 0.2158203125,
+ "learning_rate": 0.0005,
+ "loss": 1.1393,
+ "step": 301
+ },
+ {
+ "epoch": 0.9648562300319489,
+ "grad_norm": 0.1083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.1483,
+ "step": 302
+ },
+ {
+ "epoch": 0.9680511182108626,
+ "grad_norm": 0.09326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.1301,
+ "step": 303
+ },
+ {
+ "epoch": 0.9712460063897763,
+ "grad_norm": 0.08935546875,
+ "learning_rate": 0.0005,
+ "loss": 1.1317,
+ "step": 304
+ },
+ {
+ "epoch": 0.9744408945686901,
+ "grad_norm": 0.08251953125,
+ "learning_rate": 0.0005,
+ "loss": 1.1346,
+ "step": 305
+ },
+ {
+ "epoch": 0.9776357827476039,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.1479,
+ "step": 306
+ },
+ {
+ "epoch": 0.9808306709265175,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.1333,
+ "step": 307
+ },
+ {
+ "epoch": 0.9840255591054313,
+ "grad_norm": 0.09326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.1342,
+ "step": 308
+ },
+ {
+ "epoch": 0.987220447284345,
+ "grad_norm": 0.1357421875,
+ "learning_rate": 0.0005,
+ "loss": 1.1399,
+ "step": 309
+ },
+ {
+ "epoch": 0.9904153354632588,
+ "grad_norm": 0.1259765625,
+ "learning_rate": 0.0005,
+ "loss": 1.1393,
+ "step": 310
+ },
+ {
+ "epoch": 0.9936102236421726,
+ "grad_norm": 0.08056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.1327,
+ "step": 311
+ },
+ {
+ "epoch": 0.9968051118210862,
+ "grad_norm": 0.16796875,
+ "learning_rate": 0.0005,
+ "loss": 1.1437,
+ "step": 312
+ },
+ {
+ "epoch": 1.0,
+ "grad_norm": 0.30859375,
+ "learning_rate": 0.0005,
+ "loss": 1.133,
+ "step": 313
+ },
+ {
+ "epoch": 1.0031948881789137,
+ "grad_norm": 0.466796875,
+ "learning_rate": 0.0005,
+ "loss": 1.146,
+ "step": 314
+ },
+ {
+ "epoch": 1.0063897763578276,
+ "grad_norm": 1.1328125,
+ "learning_rate": 0.0005,
+ "loss": 1.1581,
+ "step": 315
+ },
+ {
+ "epoch": 1.0095846645367412,
+ "grad_norm": 2.28125,
+ "learning_rate": 0.0005,
+ "loss": 1.1834,
+ "step": 316
+ },
+ {
+ "epoch": 1.012779552715655,
+ "grad_norm": 0.90625,
+ "learning_rate": 0.0005,
+ "loss": 1.147,
+ "step": 317
+ },
+ {
+ "epoch": 1.0159744408945688,
+ "grad_norm": 3.96875,
+ "learning_rate": 0.0005,
+ "loss": 1.2474,
+ "step": 318
+ },
+ {
+ "epoch": 1.0191693290734825,
+ "grad_norm": 3.9375,
+ "learning_rate": 0.0005,
+ "loss": 1.27,
+ "step": 319
+ },
+ {
+ "epoch": 1.0223642172523961,
+ "grad_norm": 1.640625,
+ "learning_rate": 0.0005,
+ "loss": 1.1838,
+ "step": 320
+ },
+ {
+ "epoch": 1.0255591054313098,
+ "grad_norm": 0.98046875,
+ "learning_rate": 0.0005,
+ "loss": 1.1858,
+ "step": 321
+ },
+ {
+ "epoch": 1.0287539936102237,
+ "grad_norm": 1.5078125,
+ "learning_rate": 0.0005,
+ "loss": 1.1884,
+ "step": 322
+ },
+ {
+ "epoch": 1.0319488817891374,
+ "grad_norm": 0.482421875,
+ "learning_rate": 0.0005,
+ "loss": 1.1729,
+ "step": 323
+ },
+ {
+ "epoch": 1.035143769968051,
+ "grad_norm": 1.6015625,
+ "learning_rate": 0.0005,
+ "loss": 1.1909,
+ "step": 324
+ },
+ {
+ "epoch": 1.038338658146965,
+ "grad_norm": 0.32421875,
+ "learning_rate": 0.0005,
+ "loss": 1.1666,
+ "step": 325
+ },
+ {
+ "epoch": 1.0415335463258786,
+ "grad_norm": 1.03125,
+ "learning_rate": 0.0005,
+ "loss": 1.165,
+ "step": 326
+ },
+ {
+ "epoch": 1.0447284345047922,
+ "grad_norm": 0.6953125,
+ "learning_rate": 0.0005,
+ "loss": 1.1821,
+ "step": 327
+ },
+ {
+ "epoch": 1.0479233226837061,
+ "grad_norm": 0.31640625,
+ "learning_rate": 0.0005,
+ "loss": 1.1584,
+ "step": 328
+ },
+ {
+ "epoch": 1.0511182108626198,
+ "grad_norm": 0.48828125,
+ "learning_rate": 0.0005,
+ "loss": 1.1574,
+ "step": 329
+ },
+ {
+ "epoch": 1.0543130990415335,
+ "grad_norm": 0.1650390625,
+ "learning_rate": 0.0005,
+ "loss": 1.1654,
+ "step": 330
+ },
+ {
+ "epoch": 1.0575079872204474,
+ "grad_norm": 0.365234375,
+ "learning_rate": 0.0005,
+ "loss": 1.1488,
+ "step": 331
+ },
+ {
+ "epoch": 1.060702875399361,
+ "grad_norm": 0.16796875,
+ "learning_rate": 0.0005,
+ "loss": 1.1434,
+ "step": 332
+ },
+ {
+ "epoch": 1.0638977635782747,
+ "grad_norm": 0.34375,
+ "learning_rate": 0.0005,
+ "loss": 1.1481,
+ "step": 333
+ },
+ {
+ "epoch": 1.0670926517571886,
+ "grad_norm": 0.2265625,
+ "learning_rate": 0.0005,
+ "loss": 1.1385,
+ "step": 334
+ },
+ {
+ "epoch": 1.0702875399361023,
+ "grad_norm": 0.3828125,
+ "learning_rate": 0.0005,
+ "loss": 1.1563,
+ "step": 335
+ },
+ {
+ "epoch": 1.073482428115016,
+ "grad_norm": 0.13671875,
+ "learning_rate": 0.0005,
+ "loss": 1.1546,
+ "step": 336
+ },
+ {
+ "epoch": 1.0766773162939298,
+ "grad_norm": 0.251953125,
+ "learning_rate": 0.0005,
+ "loss": 1.1441,
+ "step": 337
+ },
+ {
+ "epoch": 1.0798722044728435,
+ "grad_norm": 0.1103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.1606,
+ "step": 338
+ },
+ {
+ "epoch": 1.0830670926517572,
+ "grad_norm": 0.1552734375,
+ "learning_rate": 0.0005,
+ "loss": 1.1365,
+ "step": 339
+ },
+ {
+ "epoch": 1.0862619808306708,
+ "grad_norm": 0.12060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.1541,
+ "step": 340
+ },
+ {
+ "epoch": 1.0894568690095847,
+ "grad_norm": 0.134765625,
+ "learning_rate": 0.0005,
+ "loss": 1.1354,
+ "step": 341
+ },
+ {
+ "epoch": 1.0926517571884984,
+ "grad_norm": 0.12255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.1594,
+ "step": 342
+ },
+ {
+ "epoch": 1.095846645367412,
+ "grad_norm": 0.0849609375,
+ "learning_rate": 0.0005,
+ "loss": 1.1354,
+ "step": 343
+ },
+ {
+ "epoch": 1.099041533546326,
+ "grad_norm": 0.1064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.1505,
+ "step": 344
+ },
+ {
+ "epoch": 1.1022364217252396,
+ "grad_norm": 0.0947265625,
+ "learning_rate": 0.0005,
+ "loss": 1.141,
+ "step": 345
+ },
+ {
+ "epoch": 1.1054313099041533,
+ "grad_norm": 0.130859375,
+ "learning_rate": 0.0005,
+ "loss": 1.1335,
+ "step": 346
+ },
+ {
+ "epoch": 1.1086261980830672,
+ "grad_norm": 0.11083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.1194,
+ "step": 347
+ },
+ {
+ "epoch": 1.1118210862619808,
+ "grad_norm": 0.08251953125,
+ "learning_rate": 0.0005,
+ "loss": 1.1339,
+ "step": 348
+ },
+ {
+ "epoch": 1.1150159744408945,
+ "grad_norm": 0.0966796875,
+ "learning_rate": 0.0005,
+ "loss": 1.1428,
+ "step": 349
+ },
+ {
+ "epoch": 1.1182108626198084,
+ "grad_norm": 0.09423828125,
+ "learning_rate": 0.0005,
+ "loss": 1.141,
+ "step": 350
+ },
+ {
+ "epoch": 1.121405750798722,
+ "grad_norm": 0.10693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.1415,
+ "step": 351
+ },
+ {
+ "epoch": 1.1246006389776357,
+ "grad_norm": 0.0947265625,
+ "learning_rate": 0.0005,
+ "loss": 1.1244,
+ "step": 352
+ },
+ {
+ "epoch": 1.1277955271565494,
+ "grad_norm": 0.12158203125,
+ "learning_rate": 0.0005,
+ "loss": 1.1296,
+ "step": 353
+ },
+ {
+ "epoch": 1.1309904153354633,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.125,
+ "step": 354
+ },
+ {
+ "epoch": 1.134185303514377,
+ "grad_norm": 0.10693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.1438,
+ "step": 355
+ },
+ {
+ "epoch": 1.1373801916932909,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.1404,
+ "step": 356
+ },
+ {
+ "epoch": 1.1405750798722045,
+ "grad_norm": 0.10205078125,
+ "learning_rate": 0.0005,
+ "loss": 1.1492,
+ "step": 357
+ },
+ {
+ "epoch": 1.1437699680511182,
+ "grad_norm": 0.083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.1389,
+ "step": 358
+ },
+ {
+ "epoch": 1.1469648562300319,
+ "grad_norm": 0.10693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.1334,
+ "step": 359
+ },
+ {
+ "epoch": 1.1501597444089458,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.1426,
+ "step": 360
+ },
+ {
+ "epoch": 1.1533546325878594,
+ "grad_norm": 0.09814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.1413,
+ "step": 361
+ },
+ {
+ "epoch": 1.156549520766773,
+ "grad_norm": 0.08203125,
+ "learning_rate": 0.0005,
+ "loss": 1.1436,
+ "step": 362
+ },
+ {
+ "epoch": 1.159744408945687,
+ "grad_norm": 0.0888671875,
+ "learning_rate": 0.0005,
+ "loss": 1.13,
+ "step": 363
+ },
+ {
+ "epoch": 1.1629392971246006,
+ "grad_norm": 0.09130859375,
+ "learning_rate": 0.0005,
+ "loss": 1.1325,
+ "step": 364
+ },
+ {
+ "epoch": 1.1661341853035143,
+ "grad_norm": 0.0849609375,
+ "learning_rate": 0.0005,
+ "loss": 1.1237,
+ "step": 365
+ },
+ {
+ "epoch": 1.1693290734824282,
+ "grad_norm": 0.07861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.1327,
+ "step": 366
+ },
+ {
+ "epoch": 1.1725239616613419,
+ "grad_norm": 0.11865234375,
+ "learning_rate": 0.0005,
+ "loss": 1.1276,
+ "step": 367
+ },
+ {
+ "epoch": 1.1757188498402555,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.1338,
+ "step": 368
+ },
+ {
+ "epoch": 1.1789137380191694,
+ "grad_norm": 0.1298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.1378,
+ "step": 369
+ },
+ {
+ "epoch": 1.182108626198083,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.1294,
+ "step": 370
+ },
+ {
+ "epoch": 1.1853035143769968,
+ "grad_norm": 0.1279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.1354,
+ "step": 371
+ },
+ {
+ "epoch": 1.1884984025559104,
+ "grad_norm": 0.060791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.1367,
+ "step": 372
+ },
+ {
+ "epoch": 1.1916932907348243,
+ "grad_norm": 0.1611328125,
+ "learning_rate": 0.0005,
+ "loss": 1.1185,
+ "step": 373
+ },
+ {
+ "epoch": 1.194888178913738,
+ "grad_norm": 0.0576171875,
+ "learning_rate": 0.0005,
+ "loss": 1.1187,
+ "step": 374
+ },
+ {
+ "epoch": 1.1980830670926517,
+ "grad_norm": 0.11279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.1427,
+ "step": 375
+ },
+ {
+ "epoch": 1.2012779552715656,
+ "grad_norm": 0.08056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.1403,
+ "step": 376
+ },
+ {
+ "epoch": 1.2044728434504792,
+ "grad_norm": 0.11181640625,
+ "learning_rate": 0.0005,
+ "loss": 1.1422,
+ "step": 377
+ },
+ {
+ "epoch": 1.207667731629393,
+ "grad_norm": 0.057373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.1158,
+ "step": 378
+ },
+ {
+ "epoch": 1.2108626198083068,
+ "grad_norm": 0.060302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.1157,
+ "step": 379
+ },
+ {
+ "epoch": 1.2140575079872205,
+ "grad_norm": 0.06103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.1254,
+ "step": 380
+ },
+ {
+ "epoch": 1.2172523961661341,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.1207,
+ "step": 381
+ },
+ {
+ "epoch": 1.220447284345048,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.1219,
+ "step": 382
+ },
+ {
+ "epoch": 1.2236421725239617,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.104,
+ "step": 383
+ },
+ {
+ "epoch": 1.2268370607028753,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.1238,
+ "step": 384
+ },
+ {
+ "epoch": 1.230031948881789,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.1156,
+ "step": 385
+ },
+ {
+ "epoch": 1.233226837060703,
+ "grad_norm": 0.07861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.1117,
+ "step": 386
+ },
+ {
+ "epoch": 1.2364217252396166,
+ "grad_norm": 0.08642578125,
+ "learning_rate": 0.0005,
+ "loss": 1.1277,
+ "step": 387
+ },
+ {
+ "epoch": 1.2396166134185305,
+ "grad_norm": 0.059326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.1377,
+ "step": 388
+ },
+ {
+ "epoch": 1.2428115015974441,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.1317,
+ "step": 389
+ },
+ {
+ "epoch": 1.2460063897763578,
+ "grad_norm": 0.05810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.1299,
+ "step": 390
+ },
+ {
+ "epoch": 1.2492012779552715,
+ "grad_norm": 0.0810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.1398,
+ "step": 391
+ },
+ {
+ "epoch": 1.2523961661341854,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.1369,
+ "step": 392
+ },
+ {
+ "epoch": 1.255591054313099,
+ "grad_norm": 0.1044921875,
+ "learning_rate": 0.0005,
+ "loss": 1.1316,
+ "step": 393
+ },
+ {
+ "epoch": 1.2587859424920127,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.1407,
+ "step": 394
+ },
+ {
+ "epoch": 1.2619808306709266,
+ "grad_norm": 0.0947265625,
+ "learning_rate": 0.0005,
+ "loss": 1.1283,
+ "step": 395
+ },
+ {
+ "epoch": 1.2651757188498403,
+ "grad_norm": 0.061279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.1225,
+ "step": 396
+ },
+ {
+ "epoch": 1.268370607028754,
+ "grad_norm": 0.08837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.1405,
+ "step": 397
+ },
+ {
+ "epoch": 1.2715654952076676,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.1114,
+ "step": 398
+ },
+ {
+ "epoch": 1.2747603833865815,
+ "grad_norm": 0.0927734375,
+ "learning_rate": 0.0005,
+ "loss": 1.1427,
+ "step": 399
+ },
+ {
+ "epoch": 1.2779552715654952,
+ "grad_norm": 0.09814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.1354,
+ "step": 400
+ },
+ {
+ "epoch": 1.281150159744409,
+ "grad_norm": 0.099609375,
+ "learning_rate": 0.0005,
+ "loss": 1.1348,
+ "step": 401
+ },
+ {
+ "epoch": 1.2843450479233227,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.1264,
+ "step": 402
+ },
+ {
+ "epoch": 1.2875399361022364,
+ "grad_norm": 0.09912109375,
+ "learning_rate": 0.0005,
+ "loss": 1.1284,
+ "step": 403
+ },
+ {
+ "epoch": 1.29073482428115,
+ "grad_norm": 0.083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.1314,
+ "step": 404
+ },
+ {
+ "epoch": 1.293929712460064,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.1393,
+ "step": 405
+ },
+ {
+ "epoch": 1.2971246006389776,
+ "grad_norm": 0.10546875,
+ "learning_rate": 0.0005,
+ "loss": 1.1229,
+ "step": 406
+ },
+ {
+ "epoch": 1.3003194888178915,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.1309,
+ "step": 407
+ },
+ {
+ "epoch": 1.3035143769968052,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.1327,
+ "step": 408
+ },
+ {
+ "epoch": 1.3067092651757188,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.136,
+ "step": 409
+ },
+ {
+ "epoch": 1.3099041533546325,
+ "grad_norm": 0.0908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.1465,
+ "step": 410
+ },
+ {
+ "epoch": 1.3130990415335464,
+ "grad_norm": 0.13671875,
+ "learning_rate": 0.0005,
+ "loss": 1.1324,
+ "step": 411
+ },
+ {
+ "epoch": 1.31629392971246,
+ "grad_norm": 0.09521484375,
+ "learning_rate": 0.0005,
+ "loss": 1.1218,
+ "step": 412
+ },
+ {
+ "epoch": 1.3194888178913737,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.1221,
+ "step": 413
+ },
+ {
+ "epoch": 1.3226837060702876,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.1293,
+ "step": 414
+ },
+ {
+ "epoch": 1.3258785942492013,
+ "grad_norm": 0.05908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.1326,
+ "step": 415
+ },
+ {
+ "epoch": 1.329073482428115,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.1219,
+ "step": 416
+ },
+ {
+ "epoch": 1.3322683706070286,
+ "grad_norm": 0.08203125,
+ "learning_rate": 0.0005,
+ "loss": 1.1322,
+ "step": 417
+ },
+ {
+ "epoch": 1.3354632587859425,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.1243,
+ "step": 418
+ },
+ {
+ "epoch": 1.3386581469648562,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.1255,
+ "step": 419
+ },
+ {
+ "epoch": 1.34185303514377,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.1145,
+ "step": 420
+ },
+ {
+ "epoch": 1.3450479233226837,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.118,
+ "step": 421
+ },
+ {
+ "epoch": 1.3482428115015974,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.1124,
+ "step": 422
+ },
+ {
+ "epoch": 1.351437699680511,
+ "grad_norm": 0.10498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.1155,
+ "step": 423
+ },
+ {
+ "epoch": 1.354632587859425,
+ "grad_norm": 0.13671875,
+ "learning_rate": 0.0005,
+ "loss": 1.1265,
+ "step": 424
+ },
+ {
+ "epoch": 1.3578274760383386,
+ "grad_norm": 0.126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.1144,
+ "step": 425
+ },
+ {
+ "epoch": 1.3610223642172525,
+ "grad_norm": 0.134765625,
+ "learning_rate": 0.0005,
+ "loss": 1.129,
+ "step": 426
+ },
+ {
+ "epoch": 1.3642172523961662,
+ "grad_norm": 0.0810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.1216,
+ "step": 427
+ },
+ {
+ "epoch": 1.3674121405750799,
+ "grad_norm": 0.138671875,
+ "learning_rate": 0.0005,
+ "loss": 1.1473,
+ "step": 428
+ },
+ {
+ "epoch": 1.3706070287539935,
+ "grad_norm": 0.1259765625,
+ "learning_rate": 0.0005,
+ "loss": 1.1088,
+ "step": 429
+ },
+ {
+ "epoch": 1.3738019169329074,
+ "grad_norm": 0.12109375,
+ "learning_rate": 0.0005,
+ "loss": 1.1146,
+ "step": 430
+ },
+ {
+ "epoch": 1.376996805111821,
+ "grad_norm": 0.1650390625,
+ "learning_rate": 0.0005,
+ "loss": 1.1216,
+ "step": 431
+ },
+ {
+ "epoch": 1.3801916932907348,
+ "grad_norm": 0.08251953125,
+ "learning_rate": 0.0005,
+ "loss": 1.1308,
+ "step": 432
+ },
+ {
+ "epoch": 1.3833865814696487,
+ "grad_norm": 0.1513671875,
+ "learning_rate": 0.0005,
+ "loss": 1.1287,
+ "step": 433
+ },
+ {
+ "epoch": 1.3865814696485623,
+ "grad_norm": 0.0947265625,
+ "learning_rate": 0.0005,
+ "loss": 1.1215,
+ "step": 434
+ },
+ {
+ "epoch": 1.389776357827476,
+ "grad_norm": 0.10400390625,
+ "learning_rate": 0.0005,
+ "loss": 1.1217,
+ "step": 435
+ },
+ {
+ "epoch": 1.3929712460063897,
+ "grad_norm": 0.15625,
+ "learning_rate": 0.0005,
+ "loss": 1.1154,
+ "step": 436
+ },
+ {
+ "epoch": 1.3961661341853036,
+ "grad_norm": 0.08642578125,
+ "learning_rate": 0.0005,
+ "loss": 1.1161,
+ "step": 437
+ },
+ {
+ "epoch": 1.3993610223642172,
+ "grad_norm": 0.09716796875,
+ "learning_rate": 0.0005,
+ "loss": 1.1182,
+ "step": 438
+ },
+ {
+ "epoch": 1.4025559105431311,
+ "grad_norm": 0.087890625,
+ "learning_rate": 0.0005,
+ "loss": 1.1179,
+ "step": 439
+ },
+ {
+ "epoch": 1.4057507987220448,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.1243,
+ "step": 440
+ },
+ {
+ "epoch": 1.4089456869009584,
+ "grad_norm": 0.0849609375,
+ "learning_rate": 0.0005,
+ "loss": 1.1176,
+ "step": 441
+ },
+ {
+ "epoch": 1.4121405750798721,
+ "grad_norm": 0.11669921875,
+ "learning_rate": 0.0005,
+ "loss": 1.1187,
+ "step": 442
+ },
+ {
+ "epoch": 1.415335463258786,
+ "grad_norm": 0.15625,
+ "learning_rate": 0.0005,
+ "loss": 1.1455,
+ "step": 443
+ },
+ {
+ "epoch": 1.4185303514376997,
+ "grad_norm": 0.1591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.1254,
+ "step": 444
+ },
+ {
+ "epoch": 1.4217252396166133,
+ "grad_norm": 0.1533203125,
+ "learning_rate": 0.0005,
+ "loss": 1.1242,
+ "step": 445
+ },
+ {
+ "epoch": 1.4249201277955272,
+ "grad_norm": 0.1220703125,
+ "learning_rate": 0.0005,
+ "loss": 1.1102,
+ "step": 446
+ },
+ {
+ "epoch": 1.428115015974441,
+ "grad_norm": 0.130859375,
+ "learning_rate": 0.0005,
+ "loss": 1.1355,
+ "step": 447
+ },
+ {
+ "epoch": 1.4313099041533546,
+ "grad_norm": 0.09326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.1216,
+ "step": 448
+ },
+ {
+ "epoch": 1.4345047923322682,
+ "grad_norm": 0.1669921875,
+ "learning_rate": 0.0005,
+ "loss": 1.1275,
+ "step": 449
+ },
+ {
+ "epoch": 1.4376996805111821,
+ "grad_norm": 0.1123046875,
+ "learning_rate": 0.0005,
+ "loss": 1.1256,
+ "step": 450
+ },
+ {
+ "epoch": 1.4408945686900958,
+ "grad_norm": 0.11669921875,
+ "learning_rate": 0.0005,
+ "loss": 1.1145,
+ "step": 451
+ },
+ {
+ "epoch": 1.4440894568690097,
+ "grad_norm": 0.1640625,
+ "learning_rate": 0.0005,
+ "loss": 1.1269,
+ "step": 452
+ },
+ {
+ "epoch": 1.4472843450479234,
+ "grad_norm": 0.091796875,
+ "learning_rate": 0.0005,
+ "loss": 1.1189,
+ "step": 453
+ },
+ {
+ "epoch": 1.450479233226837,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.1288,
+ "step": 454
+ },
+ {
+ "epoch": 1.4536741214057507,
+ "grad_norm": 0.10400390625,
+ "learning_rate": 0.0005,
+ "loss": 1.1285,
+ "step": 455
+ },
+ {
+ "epoch": 1.4568690095846646,
+ "grad_norm": 0.1494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.1184,
+ "step": 456
+ },
+ {
+ "epoch": 1.4600638977635783,
+ "grad_norm": 0.10986328125,
+ "learning_rate": 0.0005,
+ "loss": 1.1027,
+ "step": 457
+ },
+ {
+ "epoch": 1.4632587859424921,
+ "grad_norm": 0.11376953125,
+ "learning_rate": 0.0005,
+ "loss": 1.1125,
+ "step": 458
+ },
+ {
+ "epoch": 1.4664536741214058,
+ "grad_norm": 0.09814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.1206,
+ "step": 459
+ },
+ {
+ "epoch": 1.4696485623003195,
+ "grad_norm": 0.123046875,
+ "learning_rate": 0.0005,
+ "loss": 1.1288,
+ "step": 460
+ },
+ {
+ "epoch": 1.4728434504792332,
+ "grad_norm": 0.271484375,
+ "learning_rate": 0.0005,
+ "loss": 1.1247,
+ "step": 461
+ },
+ {
+ "epoch": 1.476038338658147,
+ "grad_norm": 0.3828125,
+ "learning_rate": 0.0005,
+ "loss": 1.1149,
+ "step": 462
+ },
+ {
+ "epoch": 1.4792332268370607,
+ "grad_norm": 0.62890625,
+ "learning_rate": 0.0005,
+ "loss": 1.1178,
+ "step": 463
+ },
+ {
+ "epoch": 1.4824281150159744,
+ "grad_norm": 1.421875,
+ "learning_rate": 0.0005,
+ "loss": 1.1315,
+ "step": 464
+ },
+ {
+ "epoch": 1.4856230031948883,
+ "grad_norm": 1.859375,
+ "learning_rate": 0.0005,
+ "loss": 1.149,
+ "step": 465
+ },
+ {
+ "epoch": 1.488817891373802,
+ "grad_norm": 0.162109375,
+ "learning_rate": 0.0005,
+ "loss": 1.1193,
+ "step": 466
+ },
+ {
+ "epoch": 1.4920127795527156,
+ "grad_norm": 1.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.1289,
+ "step": 467
+ },
+ {
+ "epoch": 1.4952076677316293,
+ "grad_norm": 1.7265625,
+ "learning_rate": 0.0005,
+ "loss": 1.146,
+ "step": 468
+ },
+ {
+ "epoch": 1.4984025559105432,
+ "grad_norm": 0.609375,
+ "learning_rate": 0.0005,
+ "loss": 1.1236,
+ "step": 469
+ },
+ {
+ "epoch": 1.5015974440894568,
+ "grad_norm": 2.859375,
+ "learning_rate": 0.0005,
+ "loss": 1.19,
+ "step": 470
+ },
+ {
+ "epoch": 1.5047923322683707,
+ "grad_norm": 2.234375,
+ "learning_rate": 0.0005,
+ "loss": 1.1704,
+ "step": 471
+ },
+ {
+ "epoch": 1.5079872204472844,
+ "grad_norm": 1.484375,
+ "learning_rate": 0.0005,
+ "loss": 1.1629,
+ "step": 472
+ },
+ {
+ "epoch": 1.511182108626198,
+ "grad_norm": 1.2734375,
+ "learning_rate": 0.0005,
+ "loss": 1.1456,
+ "step": 473
+ },
+ {
+ "epoch": 1.5143769968051117,
+ "grad_norm": 1.3203125,
+ "learning_rate": 0.0005,
+ "loss": 1.1474,
+ "step": 474
+ },
+ {
+ "epoch": 1.5175718849840254,
+ "grad_norm": 0.91015625,
+ "learning_rate": 0.0005,
+ "loss": 1.1277,
+ "step": 475
+ },
+ {
+ "epoch": 1.5207667731629393,
+ "grad_norm": 1.78125,
+ "learning_rate": 0.0005,
+ "loss": 1.1545,
+ "step": 476
+ },
+ {
+ "epoch": 1.5239616613418532,
+ "grad_norm": 1.1484375,
+ "learning_rate": 0.0005,
+ "loss": 1.153,
+ "step": 477
+ },
+ {
+ "epoch": 1.5271565495207668,
+ "grad_norm": 1.1875,
+ "learning_rate": 0.0005,
+ "loss": 1.1688,
+ "step": 478
+ },
+ {
+ "epoch": 1.5303514376996805,
+ "grad_norm": 1.03125,
+ "learning_rate": 0.0005,
+ "loss": 1.1409,
+ "step": 479
+ },
+ {
+ "epoch": 1.5335463258785942,
+ "grad_norm": 1.546875,
+ "learning_rate": 0.0005,
+ "loss": 1.1553,
+ "step": 480
+ },
+ {
+ "epoch": 1.5367412140575079,
+ "grad_norm": 0.79296875,
+ "learning_rate": 0.0005,
+ "loss": 1.1528,
+ "step": 481
+ },
+ {
+ "epoch": 1.5399361022364217,
+ "grad_norm": 1.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.1609,
+ "step": 482
+ },
+ {
+ "epoch": 1.5431309904153354,
+ "grad_norm": 0.76953125,
+ "learning_rate": 0.0005,
+ "loss": 1.1525,
+ "step": 483
+ },
+ {
+ "epoch": 1.5463258785942493,
+ "grad_norm": 1.390625,
+ "learning_rate": 0.0005,
+ "loss": 1.16,
+ "step": 484
+ },
+ {
+ "epoch": 1.549520766773163,
+ "grad_norm": 0.85546875,
+ "learning_rate": 0.0005,
+ "loss": 1.1406,
+ "step": 485
+ },
+ {
+ "epoch": 1.5527156549520766,
+ "grad_norm": 1.1640625,
+ "learning_rate": 0.0005,
+ "loss": 1.1462,
+ "step": 486
+ },
+ {
+ "epoch": 1.5559105431309903,
+ "grad_norm": 0.859375,
+ "learning_rate": 0.0005,
+ "loss": 1.1415,
+ "step": 487
+ },
+ {
+ "epoch": 1.5591054313099042,
+ "grad_norm": 1.140625,
+ "learning_rate": 0.0005,
+ "loss": 1.1492,
+ "step": 488
+ },
+ {
+ "epoch": 1.5623003194888179,
+ "grad_norm": 0.78125,
+ "learning_rate": 0.0005,
+ "loss": 1.1426,
+ "step": 489
+ },
+ {
+ "epoch": 1.5654952076677318,
+ "grad_norm": 1.0078125,
+ "learning_rate": 0.0005,
+ "loss": 1.1638,
+ "step": 490
+ },
+ {
+ "epoch": 1.5686900958466454,
+ "grad_norm": 0.5703125,
+ "learning_rate": 0.0005,
+ "loss": 1.1434,
+ "step": 491
+ },
+ {
+ "epoch": 1.571884984025559,
+ "grad_norm": 0.703125,
+ "learning_rate": 0.0005,
+ "loss": 1.1368,
+ "step": 492
+ },
+ {
+ "epoch": 1.5750798722044728,
+ "grad_norm": 0.6484375,
+ "learning_rate": 0.0005,
+ "loss": 1.1421,
+ "step": 493
+ },
+ {
+ "epoch": 1.5782747603833864,
+ "grad_norm": 0.70703125,
+ "learning_rate": 0.0005,
+ "loss": 1.1492,
+ "step": 494
+ },
+ {
+ "epoch": 1.5814696485623003,
+ "grad_norm": 0.52734375,
+ "learning_rate": 0.0005,
+ "loss": 1.1375,
+ "step": 495
+ },
+ {
+ "epoch": 1.5846645367412142,
+ "grad_norm": 0.453125,
+ "learning_rate": 0.0005,
+ "loss": 1.1213,
+ "step": 496
+ },
+ {
+ "epoch": 1.5878594249201279,
+ "grad_norm": 0.5859375,
+ "learning_rate": 0.0005,
+ "loss": 1.1385,
+ "step": 497
+ },
+ {
+ "epoch": 1.5910543130990416,
+ "grad_norm": 0.296875,
+ "learning_rate": 0.0005,
+ "loss": 1.1238,
+ "step": 498
+ },
+ {
+ "epoch": 1.5942492012779552,
+ "grad_norm": 0.419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.1291,
+ "step": 499
+ },
+ {
+ "epoch": 1.5974440894568689,
+ "grad_norm": 0.1953125,
+ "learning_rate": 0.0005,
+ "loss": 1.1283,
+ "step": 500
+ },
+ {
+ "epoch": 1.6006389776357828,
+ "grad_norm": 0.36328125,
+ "learning_rate": 0.0005,
+ "loss": 1.1199,
+ "step": 501
+ },
+ {
+ "epoch": 1.6038338658146964,
+ "grad_norm": 0.279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.1211,
+ "step": 502
+ },
+ {
+ "epoch": 1.6070287539936103,
+ "grad_norm": 0.1826171875,
+ "learning_rate": 0.0005,
+ "loss": 1.1335,
+ "step": 503
+ },
+ {
+ "epoch": 1.610223642172524,
+ "grad_norm": 0.291015625,
+ "learning_rate": 0.0005,
+ "loss": 1.1285,
+ "step": 504
+ },
+ {
+ "epoch": 1.6134185303514377,
+ "grad_norm": 0.1494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.1359,
+ "step": 505
+ },
+ {
+ "epoch": 1.6166134185303513,
+ "grad_norm": 0.2294921875,
+ "learning_rate": 0.0005,
+ "loss": 1.1188,
+ "step": 506
+ },
+ {
+ "epoch": 1.619808306709265,
+ "grad_norm": 0.2392578125,
+ "learning_rate": 0.0005,
+ "loss": 1.139,
+ "step": 507
+ },
+ {
+ "epoch": 1.623003194888179,
+ "grad_norm": 0.09326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.1236,
+ "step": 508
+ },
+ {
+ "epoch": 1.6261980830670928,
+ "grad_norm": 0.19140625,
+ "learning_rate": 0.0005,
+ "loss": 1.107,
+ "step": 509
+ },
+ {
+ "epoch": 1.6293929712460065,
+ "grad_norm": 0.17578125,
+ "learning_rate": 0.0005,
+ "loss": 1.1296,
+ "step": 510
+ },
+ {
+ "epoch": 1.6325878594249201,
+ "grad_norm": 0.10791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.1347,
+ "step": 511
+ },
+ {
+ "epoch": 1.6357827476038338,
+ "grad_norm": 0.1806640625,
+ "learning_rate": 0.0005,
+ "loss": 1.1081,
+ "step": 512
+ },
+ {
+ "epoch": 1.6389776357827475,
+ "grad_norm": 0.1318359375,
+ "learning_rate": 0.0005,
+ "loss": 1.1228,
+ "step": 513
+ },
+ {
+ "epoch": 1.6421725239616614,
+ "grad_norm": 0.095703125,
+ "learning_rate": 0.0005,
+ "loss": 1.1132,
+ "step": 514
+ },
+ {
+ "epoch": 1.645367412140575,
+ "grad_norm": 0.146484375,
+ "learning_rate": 0.0005,
+ "loss": 1.1234,
+ "step": 515
+ },
+ {
+ "epoch": 1.648562300319489,
+ "grad_norm": 0.158203125,
+ "learning_rate": 0.0005,
+ "loss": 1.1235,
+ "step": 516
+ },
+ {
+ "epoch": 1.6517571884984026,
+ "grad_norm": 0.09130859375,
+ "learning_rate": 0.0005,
+ "loss": 1.1173,
+ "step": 517
+ },
+ {
+ "epoch": 1.6549520766773163,
+ "grad_norm": 0.162109375,
+ "learning_rate": 0.0005,
+ "loss": 1.1272,
+ "step": 518
+ },
+ {
+ "epoch": 1.65814696485623,
+ "grad_norm": 0.080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.1173,
+ "step": 519
+ },
+ {
+ "epoch": 1.6613418530351438,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.1116,
+ "step": 520
+ },
+ {
+ "epoch": 1.6645367412140575,
+ "grad_norm": 0.09326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.1293,
+ "step": 521
+ },
+ {
+ "epoch": 1.6677316293929714,
+ "grad_norm": 0.08056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.1161,
+ "step": 522
+ },
+ {
+ "epoch": 1.670926517571885,
+ "grad_norm": 0.0595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.1211,
+ "step": 523
+ },
+ {
+ "epoch": 1.6741214057507987,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.1292,
+ "step": 524
+ },
+ {
+ "epoch": 1.6773162939297124,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.1373,
+ "step": 525
+ },
+ {
+ "epoch": 1.680511182108626,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.1212,
+ "step": 526
+ },
+ {
+ "epoch": 1.68370607028754,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.1206,
+ "step": 527
+ },
+ {
+ "epoch": 1.6869009584664538,
+ "grad_norm": 0.056396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.1081,
+ "step": 528
+ },
+ {
+ "epoch": 1.6900958466453675,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.1107,
+ "step": 529
+ },
+ {
+ "epoch": 1.6932907348242812,
+ "grad_norm": 0.05810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.113,
+ "step": 530
+ },
+ {
+ "epoch": 1.6964856230031948,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.1136,
+ "step": 531
+ },
+ {
+ "epoch": 1.6996805111821085,
+ "grad_norm": 0.054931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.1147,
+ "step": 532
+ },
+ {
+ "epoch": 1.7028753993610224,
+ "grad_norm": 0.055908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.1197,
+ "step": 533
+ },
+ {
+ "epoch": 1.706070287539936,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.1073,
+ "step": 534
+ },
+ {
+ "epoch": 1.70926517571885,
+ "grad_norm": 0.060791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.1195,
+ "step": 535
+ },
+ {
+ "epoch": 1.7124600638977636,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.1119,
+ "step": 536
+ },
+ {
+ "epoch": 1.7156549520766773,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.1153,
+ "step": 537
+ },
+ {
+ "epoch": 1.718849840255591,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.1124,
+ "step": 538
+ },
+ {
+ "epoch": 1.7220447284345048,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.1056,
+ "step": 539
+ },
+ {
+ "epoch": 1.7252396166134185,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.1052,
+ "step": 540
+ },
+ {
+ "epoch": 1.7284345047923324,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.1061,
+ "step": 541
+ },
+ {
+ "epoch": 1.731629392971246,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.1321,
+ "step": 542
+ },
+ {
+ "epoch": 1.7348242811501597,
+ "grad_norm": 0.062255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.1162,
+ "step": 543
+ },
+ {
+ "epoch": 1.7380191693290734,
+ "grad_norm": 0.060302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.1158,
+ "step": 544
+ },
+ {
+ "epoch": 1.741214057507987,
+ "grad_norm": 0.08154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.1391,
+ "step": 545
+ },
+ {
+ "epoch": 1.744408945686901,
+ "grad_norm": 0.058837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.1136,
+ "step": 546
+ },
+ {
+ "epoch": 1.7476038338658149,
+ "grad_norm": 0.1064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.1076,
+ "step": 547
+ },
+ {
+ "epoch": 1.7507987220447285,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0934,
+ "step": 548
+ },
+ {
+ "epoch": 1.7539936102236422,
+ "grad_norm": 0.08935546875,
+ "learning_rate": 0.0005,
+ "loss": 1.1357,
+ "step": 549
+ },
+ {
+ "epoch": 1.7571884984025559,
+ "grad_norm": 0.1005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.1302,
+ "step": 550
+ },
+ {
+ "epoch": 1.7603833865814695,
+ "grad_norm": 0.06201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.1258,
+ "step": 551
+ },
+ {
+ "epoch": 1.7635782747603834,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.1141,
+ "step": 552
+ },
+ {
+ "epoch": 1.766773162939297,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.1075,
+ "step": 553
+ },
+ {
+ "epoch": 1.769968051118211,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.1321,
+ "step": 554
+ },
+ {
+ "epoch": 1.7731629392971247,
+ "grad_norm": 0.057373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.1174,
+ "step": 555
+ },
+ {
+ "epoch": 1.7763578274760383,
+ "grad_norm": 0.09716796875,
+ "learning_rate": 0.0005,
+ "loss": 1.1265,
+ "step": 556
+ },
+ {
+ "epoch": 1.779552715654952,
+ "grad_norm": 0.08642578125,
+ "learning_rate": 0.0005,
+ "loss": 1.1166,
+ "step": 557
+ },
+ {
+ "epoch": 1.7827476038338657,
+ "grad_norm": 0.09619140625,
+ "learning_rate": 0.0005,
+ "loss": 1.1181,
+ "step": 558
+ },
+ {
+ "epoch": 1.7859424920127795,
+ "grad_norm": 0.07861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.1187,
+ "step": 559
+ },
+ {
+ "epoch": 1.7891373801916934,
+ "grad_norm": 0.08544921875,
+ "learning_rate": 0.0005,
+ "loss": 1.118,
+ "step": 560
+ },
+ {
+ "epoch": 1.792332268370607,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.1275,
+ "step": 561
+ },
+ {
+ "epoch": 1.7955271565495208,
+ "grad_norm": 0.08935546875,
+ "learning_rate": 0.0005,
+ "loss": 1.1194,
+ "step": 562
+ },
+ {
+ "epoch": 1.7987220447284344,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.1198,
+ "step": 563
+ },
+ {
+ "epoch": 1.8019169329073481,
+ "grad_norm": 0.10302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.1126,
+ "step": 564
+ },
+ {
+ "epoch": 1.805111821086262,
+ "grad_norm": 0.0908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.1087,
+ "step": 565
+ },
+ {
+ "epoch": 1.8083067092651757,
+ "grad_norm": 0.1455078125,
+ "learning_rate": 0.0005,
+ "loss": 1.1056,
+ "step": 566
+ },
+ {
+ "epoch": 1.8115015974440896,
+ "grad_norm": 0.09033203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0968,
+ "step": 567
+ },
+ {
+ "epoch": 1.8146964856230032,
+ "grad_norm": 0.09716796875,
+ "learning_rate": 0.0005,
+ "loss": 1.1102,
+ "step": 568
+ },
+ {
+ "epoch": 1.817891373801917,
+ "grad_norm": 0.0908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.1165,
+ "step": 569
+ },
+ {
+ "epoch": 1.8210862619808306,
+ "grad_norm": 0.09521484375,
+ "learning_rate": 0.0005,
+ "loss": 1.1026,
+ "step": 570
+ },
+ {
+ "epoch": 1.8242811501597445,
+ "grad_norm": 0.0966796875,
+ "learning_rate": 0.0005,
+ "loss": 1.1145,
+ "step": 571
+ },
+ {
+ "epoch": 1.8274760383386581,
+ "grad_norm": 0.1025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.1089,
+ "step": 572
+ },
+ {
+ "epoch": 1.830670926517572,
+ "grad_norm": 0.15234375,
+ "learning_rate": 0.0005,
+ "loss": 1.1052,
+ "step": 573
+ },
+ {
+ "epoch": 1.8338658146964857,
+ "grad_norm": 0.115234375,
+ "learning_rate": 0.0005,
+ "loss": 1.1242,
+ "step": 574
+ },
+ {
+ "epoch": 1.8370607028753994,
+ "grad_norm": 0.1494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.1105,
+ "step": 575
+ },
+ {
+ "epoch": 1.840255591054313,
+ "grad_norm": 0.09326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.1091,
+ "step": 576
+ },
+ {
+ "epoch": 1.8434504792332267,
+ "grad_norm": 0.12890625,
+ "learning_rate": 0.0005,
+ "loss": 1.1171,
+ "step": 577
+ },
+ {
+ "epoch": 1.8466453674121406,
+ "grad_norm": 0.08056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.1056,
+ "step": 578
+ },
+ {
+ "epoch": 1.8498402555910545,
+ "grad_norm": 0.1103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.1006,
+ "step": 579
+ },
+ {
+ "epoch": 1.8530351437699681,
+ "grad_norm": 0.0947265625,
+ "learning_rate": 0.0005,
+ "loss": 1.1273,
+ "step": 580
+ },
+ {
+ "epoch": 1.8562300319488818,
+ "grad_norm": 0.1416015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0995,
+ "step": 581
+ },
+ {
+ "epoch": 1.8594249201277955,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.1045,
+ "step": 582
+ },
+ {
+ "epoch": 1.8626198083067091,
+ "grad_norm": 0.1123046875,
+ "learning_rate": 0.0005,
+ "loss": 1.1111,
+ "step": 583
+ },
+ {
+ "epoch": 1.865814696485623,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.1084,
+ "step": 584
+ },
+ {
+ "epoch": 1.8690095846645367,
+ "grad_norm": 0.08154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.1151,
+ "step": 585
+ },
+ {
+ "epoch": 1.8722044728434506,
+ "grad_norm": 0.1083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.1073,
+ "step": 586
+ },
+ {
+ "epoch": 1.8753993610223643,
+ "grad_norm": 0.07861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.1145,
+ "step": 587
+ },
+ {
+ "epoch": 1.878594249201278,
+ "grad_norm": 0.09375,
+ "learning_rate": 0.0005,
+ "loss": 1.1143,
+ "step": 588
+ },
+ {
+ "epoch": 1.8817891373801916,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.1067,
+ "step": 589
+ },
+ {
+ "epoch": 1.8849840255591053,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.1075,
+ "step": 590
+ },
+ {
+ "epoch": 1.8881789137380192,
+ "grad_norm": 0.061279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.1174,
+ "step": 591
+ },
+ {
+ "epoch": 1.891373801916933,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.116,
+ "step": 592
+ },
+ {
+ "epoch": 1.8945686900958467,
+ "grad_norm": 0.06005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.1124,
+ "step": 593
+ },
+ {
+ "epoch": 1.8977635782747604,
+ "grad_norm": 0.08349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.1093,
+ "step": 594
+ },
+ {
+ "epoch": 1.900958466453674,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.1099,
+ "step": 595
+ },
+ {
+ "epoch": 1.9041533546325877,
+ "grad_norm": 0.0947265625,
+ "learning_rate": 0.0005,
+ "loss": 1.1142,
+ "step": 596
+ },
+ {
+ "epoch": 1.9073482428115016,
+ "grad_norm": 0.08740234375,
+ "learning_rate": 0.0005,
+ "loss": 1.1234,
+ "step": 597
+ },
+ {
+ "epoch": 1.9105431309904153,
+ "grad_norm": 0.10693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.1114,
+ "step": 598
+ },
+ {
+ "epoch": 1.9137380191693292,
+ "grad_norm": 0.11962890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0996,
+ "step": 599
+ },
+ {
+ "epoch": 1.9169329073482428,
+ "grad_norm": 0.130859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0932,
+ "step": 600
+ },
+ {
+ "epoch": 1.9201277955271565,
+ "grad_norm": 0.150390625,
+ "learning_rate": 0.0005,
+ "loss": 1.1229,
+ "step": 601
+ },
+ {
+ "epoch": 1.9233226837060702,
+ "grad_norm": 0.2060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.1163,
+ "step": 602
+ },
+ {
+ "epoch": 1.926517571884984,
+ "grad_norm": 0.11767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.1171,
+ "step": 603
+ },
+ {
+ "epoch": 1.9297124600638977,
+ "grad_norm": 0.1884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0972,
+ "step": 604
+ },
+ {
+ "epoch": 1.9329073482428116,
+ "grad_norm": 0.1435546875,
+ "learning_rate": 0.0005,
+ "loss": 1.1159,
+ "step": 605
+ },
+ {
+ "epoch": 1.9361022364217253,
+ "grad_norm": 0.09814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.1188,
+ "step": 606
+ },
+ {
+ "epoch": 1.939297124600639,
+ "grad_norm": 0.1787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.1041,
+ "step": 607
+ },
+ {
+ "epoch": 1.9424920127795526,
+ "grad_norm": 0.08935546875,
+ "learning_rate": 0.0005,
+ "loss": 1.1229,
+ "step": 608
+ },
+ {
+ "epoch": 1.9456869009584663,
+ "grad_norm": 0.1611328125,
+ "learning_rate": 0.0005,
+ "loss": 1.1111,
+ "step": 609
+ },
+ {
+ "epoch": 1.9488817891373802,
+ "grad_norm": 0.177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.1249,
+ "step": 610
+ },
+ {
+ "epoch": 1.952076677316294,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.102,
+ "step": 611
+ },
+ {
+ "epoch": 1.9552715654952078,
+ "grad_norm": 0.1201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.1271,
+ "step": 612
+ },
+ {
+ "epoch": 1.9584664536741214,
+ "grad_norm": 0.091796875,
+ "learning_rate": 0.0005,
+ "loss": 1.1141,
+ "step": 613
+ },
+ {
+ "epoch": 1.961661341853035,
+ "grad_norm": 0.0908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.1024,
+ "step": 614
+ },
+ {
+ "epoch": 1.9648562300319488,
+ "grad_norm": 0.1474609375,
+ "learning_rate": 0.0005,
+ "loss": 1.1173,
+ "step": 615
+ },
+ {
+ "epoch": 1.9680511182108626,
+ "grad_norm": 0.08251953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0982,
+ "step": 616
+ },
+ {
+ "epoch": 1.9712460063897763,
+ "grad_norm": 0.1611328125,
+ "learning_rate": 0.0005,
+ "loss": 1.108,
+ "step": 617
+ },
+ {
+ "epoch": 1.9744408945686902,
+ "grad_norm": 0.08447265625,
+ "learning_rate": 0.0005,
+ "loss": 1.1048,
+ "step": 618
+ },
+ {
+ "epoch": 1.9776357827476039,
+ "grad_norm": 0.0966796875,
+ "learning_rate": 0.0005,
+ "loss": 1.1046,
+ "step": 619
+ },
+ {
+ "epoch": 1.9808306709265175,
+ "grad_norm": 0.099609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0982,
+ "step": 620
+ },
+ {
+ "epoch": 1.9840255591054312,
+ "grad_norm": 0.11572265625,
+ "learning_rate": 0.0005,
+ "loss": 1.1089,
+ "step": 621
+ },
+ {
+ "epoch": 1.9872204472843449,
+ "grad_norm": 0.0927734375,
+ "learning_rate": 0.0005,
+ "loss": 1.1159,
+ "step": 622
+ },
+ {
+ "epoch": 1.9904153354632588,
+ "grad_norm": 0.126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.1207,
+ "step": 623
+ },
+ {
+ "epoch": 1.9936102236421727,
+ "grad_norm": 0.08544921875,
+ "learning_rate": 0.0005,
+ "loss": 1.1135,
+ "step": 624
+ },
+ {
+ "epoch": 1.9968051118210863,
+ "grad_norm": 0.1123046875,
+ "learning_rate": 0.0005,
+ "loss": 1.1214,
+ "step": 625
+ },
+ {
+ "epoch": 2.0,
+ "grad_norm": 0.09228515625,
+ "learning_rate": 0.0005,
+ "loss": 1.1099,
+ "step": 626
+ },
+ {
+ "epoch": 2.0031948881789137,
+ "grad_norm": 0.09228515625,
+ "learning_rate": 0.0005,
+ "loss": 1.1135,
+ "step": 627
+ },
+ {
+ "epoch": 2.0063897763578273,
+ "grad_norm": 0.08447265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0994,
+ "step": 628
+ },
+ {
+ "epoch": 2.009584664536741,
+ "grad_norm": 0.0869140625,
+ "learning_rate": 0.0005,
+ "loss": 1.1072,
+ "step": 629
+ },
+ {
+ "epoch": 2.012779552715655,
+ "grad_norm": 0.1064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.1049,
+ "step": 630
+ },
+ {
+ "epoch": 2.015974440894569,
+ "grad_norm": 0.0888671875,
+ "learning_rate": 0.0005,
+ "loss": 1.1061,
+ "step": 631
+ },
+ {
+ "epoch": 2.0191693290734825,
+ "grad_norm": 0.1025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.117,
+ "step": 632
+ },
+ {
+ "epoch": 2.022364217252396,
+ "grad_norm": 0.166015625,
+ "learning_rate": 0.0005,
+ "loss": 1.103,
+ "step": 633
+ },
+ {
+ "epoch": 2.02555910543131,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.1046,
+ "step": 634
+ },
+ {
+ "epoch": 2.0287539936102235,
+ "grad_norm": 0.142578125,
+ "learning_rate": 0.0005,
+ "loss": 1.1001,
+ "step": 635
+ },
+ {
+ "epoch": 2.0319488817891376,
+ "grad_norm": 0.09033203125,
+ "learning_rate": 0.0005,
+ "loss": 1.1277,
+ "step": 636
+ },
+ {
+ "epoch": 2.0351437699680512,
+ "grad_norm": 0.10595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.1081,
+ "step": 637
+ },
+ {
+ "epoch": 2.038338658146965,
+ "grad_norm": 0.115234375,
+ "learning_rate": 0.0005,
+ "loss": 1.1007,
+ "step": 638
+ },
+ {
+ "epoch": 2.0415335463258786,
+ "grad_norm": 0.1298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0982,
+ "step": 639
+ },
+ {
+ "epoch": 2.0447284345047922,
+ "grad_norm": 0.107421875,
+ "learning_rate": 0.0005,
+ "loss": 1.1092,
+ "step": 640
+ },
+ {
+ "epoch": 2.047923322683706,
+ "grad_norm": 0.177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.1067,
+ "step": 641
+ },
+ {
+ "epoch": 2.0511182108626196,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.1124,
+ "step": 642
+ },
+ {
+ "epoch": 2.0543130990415337,
+ "grad_norm": 0.12060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.1018,
+ "step": 643
+ },
+ {
+ "epoch": 2.0575079872204474,
+ "grad_norm": 0.09814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.1093,
+ "step": 644
+ },
+ {
+ "epoch": 2.060702875399361,
+ "grad_norm": 0.1953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0927,
+ "step": 645
+ },
+ {
+ "epoch": 2.0638977635782747,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0753,
+ "step": 646
+ },
+ {
+ "epoch": 2.0670926517571884,
+ "grad_norm": 0.2421875,
+ "learning_rate": 0.0005,
+ "loss": 1.1134,
+ "step": 647
+ },
+ {
+ "epoch": 2.070287539936102,
+ "grad_norm": 0.08056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.1163,
+ "step": 648
+ },
+ {
+ "epoch": 2.073482428115016,
+ "grad_norm": 0.134765625,
+ "learning_rate": 0.0005,
+ "loss": 1.1096,
+ "step": 649
+ },
+ {
+ "epoch": 2.07667731629393,
+ "grad_norm": 0.11181640625,
+ "learning_rate": 0.0005,
+ "loss": 1.1065,
+ "step": 650
+ },
+ {
+ "epoch": 2.0798722044728435,
+ "grad_norm": 0.09716796875,
+ "learning_rate": 0.0005,
+ "loss": 1.1059,
+ "step": 651
+ },
+ {
+ "epoch": 2.083067092651757,
+ "grad_norm": 0.1357421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0991,
+ "step": 652
+ },
+ {
+ "epoch": 2.086261980830671,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0946,
+ "step": 653
+ },
+ {
+ "epoch": 2.0894568690095845,
+ "grad_norm": 0.09521484375,
+ "learning_rate": 0.0005,
+ "loss": 1.112,
+ "step": 654
+ },
+ {
+ "epoch": 2.0926517571884986,
+ "grad_norm": 0.09130859375,
+ "learning_rate": 0.0005,
+ "loss": 1.1156,
+ "step": 655
+ },
+ {
+ "epoch": 2.0958466453674123,
+ "grad_norm": 0.08154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.1043,
+ "step": 656
+ },
+ {
+ "epoch": 2.099041533546326,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.1184,
+ "step": 657
+ },
+ {
+ "epoch": 2.1022364217252396,
+ "grad_norm": 0.10302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.1268,
+ "step": 658
+ },
+ {
+ "epoch": 2.1054313099041533,
+ "grad_norm": 0.05908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.1149,
+ "step": 659
+ },
+ {
+ "epoch": 2.108626198083067,
+ "grad_norm": 0.08251953125,
+ "learning_rate": 0.0005,
+ "loss": 1.121,
+ "step": 660
+ },
+ {
+ "epoch": 2.1118210862619806,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.1181,
+ "step": 661
+ },
+ {
+ "epoch": 2.1150159744408947,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.112,
+ "step": 662
+ },
+ {
+ "epoch": 2.1182108626198084,
+ "grad_norm": 0.08349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.1181,
+ "step": 663
+ },
+ {
+ "epoch": 2.121405750798722,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0927,
+ "step": 664
+ },
+ {
+ "epoch": 2.1246006389776357,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.1004,
+ "step": 665
+ },
+ {
+ "epoch": 2.1277955271565494,
+ "grad_norm": 0.09375,
+ "learning_rate": 0.0005,
+ "loss": 1.1157,
+ "step": 666
+ },
+ {
+ "epoch": 2.130990415335463,
+ "grad_norm": 0.09912109375,
+ "learning_rate": 0.0005,
+ "loss": 1.1159,
+ "step": 667
+ },
+ {
+ "epoch": 2.134185303514377,
+ "grad_norm": 0.07861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.1085,
+ "step": 668
+ },
+ {
+ "epoch": 2.137380191693291,
+ "grad_norm": 0.08544921875,
+ "learning_rate": 0.0005,
+ "loss": 1.1092,
+ "step": 669
+ },
+ {
+ "epoch": 2.1405750798722045,
+ "grad_norm": 0.1396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.1073,
+ "step": 670
+ },
+ {
+ "epoch": 2.143769968051118,
+ "grad_norm": 0.11181640625,
+ "learning_rate": 0.0005,
+ "loss": 1.1158,
+ "step": 671
+ },
+ {
+ "epoch": 2.146964856230032,
+ "grad_norm": 0.130859375,
+ "learning_rate": 0.0005,
+ "loss": 1.113,
+ "step": 672
+ },
+ {
+ "epoch": 2.1501597444089455,
+ "grad_norm": 0.244140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0854,
+ "step": 673
+ },
+ {
+ "epoch": 2.1533546325878596,
+ "grad_norm": 0.11865234375,
+ "learning_rate": 0.0005,
+ "loss": 1.1174,
+ "step": 674
+ },
+ {
+ "epoch": 2.1565495207667733,
+ "grad_norm": 0.15234375,
+ "learning_rate": 0.0005,
+ "loss": 1.1185,
+ "step": 675
+ },
+ {
+ "epoch": 2.159744408945687,
+ "grad_norm": 0.2275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.1158,
+ "step": 676
+ },
+ {
+ "epoch": 2.1629392971246006,
+ "grad_norm": 0.10400390625,
+ "learning_rate": 0.0005,
+ "loss": 1.1196,
+ "step": 677
+ },
+ {
+ "epoch": 2.1661341853035143,
+ "grad_norm": 0.1240234375,
+ "learning_rate": 0.0005,
+ "loss": 1.1127,
+ "step": 678
+ },
+ {
+ "epoch": 2.169329073482428,
+ "grad_norm": 0.16015625,
+ "learning_rate": 0.0005,
+ "loss": 1.102,
+ "step": 679
+ },
+ {
+ "epoch": 2.1725239616613417,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.1114,
+ "step": 680
+ },
+ {
+ "epoch": 2.1757188498402558,
+ "grad_norm": 0.1416015625,
+ "learning_rate": 0.0005,
+ "loss": 1.1197,
+ "step": 681
+ },
+ {
+ "epoch": 2.1789137380191694,
+ "grad_norm": 0.1142578125,
+ "learning_rate": 0.0005,
+ "loss": 1.1024,
+ "step": 682
+ },
+ {
+ "epoch": 2.182108626198083,
+ "grad_norm": 0.111328125,
+ "learning_rate": 0.0005,
+ "loss": 1.1105,
+ "step": 683
+ },
+ {
+ "epoch": 2.1853035143769968,
+ "grad_norm": 0.1708984375,
+ "learning_rate": 0.0005,
+ "loss": 1.1268,
+ "step": 684
+ },
+ {
+ "epoch": 2.1884984025559104,
+ "grad_norm": 0.0810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.1141,
+ "step": 685
+ },
+ {
+ "epoch": 2.191693290734824,
+ "grad_norm": 0.09716796875,
+ "learning_rate": 0.0005,
+ "loss": 1.1169,
+ "step": 686
+ },
+ {
+ "epoch": 2.194888178913738,
+ "grad_norm": 0.166015625,
+ "learning_rate": 0.0005,
+ "loss": 1.1006,
+ "step": 687
+ },
+ {
+ "epoch": 2.198083067092652,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.1142,
+ "step": 688
+ },
+ {
+ "epoch": 2.2012779552715656,
+ "grad_norm": 0.107421875,
+ "learning_rate": 0.0005,
+ "loss": 1.1105,
+ "step": 689
+ },
+ {
+ "epoch": 2.2044728434504792,
+ "grad_norm": 0.09619140625,
+ "learning_rate": 0.0005,
+ "loss": 1.1078,
+ "step": 690
+ },
+ {
+ "epoch": 2.207667731629393,
+ "grad_norm": 0.083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.11,
+ "step": 691
+ },
+ {
+ "epoch": 2.2108626198083066,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0965,
+ "step": 692
+ },
+ {
+ "epoch": 2.2140575079872207,
+ "grad_norm": 0.10595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.109,
+ "step": 693
+ },
+ {
+ "epoch": 2.2172523961661343,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.1204,
+ "step": 694
+ },
+ {
+ "epoch": 2.220447284345048,
+ "grad_norm": 0.1572265625,
+ "learning_rate": 0.0005,
+ "loss": 1.1108,
+ "step": 695
+ },
+ {
+ "epoch": 2.2236421725239617,
+ "grad_norm": 0.0888671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0983,
+ "step": 696
+ },
+ {
+ "epoch": 2.2268370607028753,
+ "grad_norm": 0.1083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.1063,
+ "step": 697
+ },
+ {
+ "epoch": 2.230031948881789,
+ "grad_norm": 0.1650390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0993,
+ "step": 698
+ },
+ {
+ "epoch": 2.2332268370607027,
+ "grad_norm": 0.10791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.1072,
+ "step": 699
+ },
+ {
+ "epoch": 2.236421725239617,
+ "grad_norm": 0.09423828125,
+ "learning_rate": 0.0005,
+ "loss": 1.1085,
+ "step": 700
+ },
+ {
+ "epoch": 2.2396166134185305,
+ "grad_norm": 0.1318359375,
+ "learning_rate": 0.0005,
+ "loss": 1.1114,
+ "step": 701
+ },
+ {
+ "epoch": 2.242811501597444,
+ "grad_norm": 0.1513671875,
+ "learning_rate": 0.0005,
+ "loss": 1.1156,
+ "step": 702
+ },
+ {
+ "epoch": 2.246006389776358,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.1025,
+ "step": 703
+ },
+ {
+ "epoch": 2.2492012779552715,
+ "grad_norm": 0.087890625,
+ "learning_rate": 0.0005,
+ "loss": 1.1019,
+ "step": 704
+ },
+ {
+ "epoch": 2.252396166134185,
+ "grad_norm": 0.08837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.1024,
+ "step": 705
+ },
+ {
+ "epoch": 2.255591054313099,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.1084,
+ "step": 706
+ },
+ {
+ "epoch": 2.258785942492013,
+ "grad_norm": 0.08056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.1102,
+ "step": 707
+ },
+ {
+ "epoch": 2.2619808306709266,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0935,
+ "step": 708
+ },
+ {
+ "epoch": 2.2651757188498403,
+ "grad_norm": 0.0869140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0946,
+ "step": 709
+ },
+ {
+ "epoch": 2.268370607028754,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.1144,
+ "step": 710
+ },
+ {
+ "epoch": 2.2715654952076676,
+ "grad_norm": 0.0859375,
+ "learning_rate": 0.0005,
+ "loss": 1.1079,
+ "step": 711
+ },
+ {
+ "epoch": 2.2747603833865817,
+ "grad_norm": 0.08447265625,
+ "learning_rate": 0.0005,
+ "loss": 1.1062,
+ "step": 712
+ },
+ {
+ "epoch": 2.2779552715654954,
+ "grad_norm": 0.11865234375,
+ "learning_rate": 0.0005,
+ "loss": 1.1197,
+ "step": 713
+ },
+ {
+ "epoch": 2.281150159744409,
+ "grad_norm": 0.08544921875,
+ "learning_rate": 0.0005,
+ "loss": 1.098,
+ "step": 714
+ },
+ {
+ "epoch": 2.2843450479233227,
+ "grad_norm": 0.126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.1077,
+ "step": 715
+ },
+ {
+ "epoch": 2.2875399361022364,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.1042,
+ "step": 716
+ },
+ {
+ "epoch": 2.29073482428115,
+ "grad_norm": 0.0947265625,
+ "learning_rate": 0.0005,
+ "loss": 1.1175,
+ "step": 717
+ },
+ {
+ "epoch": 2.2939297124600637,
+ "grad_norm": 0.10107421875,
+ "learning_rate": 0.0005,
+ "loss": 1.1048,
+ "step": 718
+ },
+ {
+ "epoch": 2.297124600638978,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0994,
+ "step": 719
+ },
+ {
+ "epoch": 2.3003194888178915,
+ "grad_norm": 0.1240234375,
+ "learning_rate": 0.0005,
+ "loss": 1.1014,
+ "step": 720
+ },
+ {
+ "epoch": 2.303514376996805,
+ "grad_norm": 0.0849609375,
+ "learning_rate": 0.0005,
+ "loss": 1.1071,
+ "step": 721
+ },
+ {
+ "epoch": 2.306709265175719,
+ "grad_norm": 0.1357421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0929,
+ "step": 722
+ },
+ {
+ "epoch": 2.3099041533546325,
+ "grad_norm": 0.11962890625,
+ "learning_rate": 0.0005,
+ "loss": 1.1184,
+ "step": 723
+ },
+ {
+ "epoch": 2.313099041533546,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0994,
+ "step": 724
+ },
+ {
+ "epoch": 2.31629392971246,
+ "grad_norm": 0.09130859375,
+ "learning_rate": 0.0005,
+ "loss": 1.1076,
+ "step": 725
+ },
+ {
+ "epoch": 2.319488817891374,
+ "grad_norm": 0.08642578125,
+ "learning_rate": 0.0005,
+ "loss": 1.112,
+ "step": 726
+ },
+ {
+ "epoch": 2.3226837060702876,
+ "grad_norm": 0.12060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.1032,
+ "step": 727
+ },
+ {
+ "epoch": 2.3258785942492013,
+ "grad_norm": 0.09765625,
+ "learning_rate": 0.0005,
+ "loss": 1.1009,
+ "step": 728
+ },
+ {
+ "epoch": 2.329073482428115,
+ "grad_norm": 0.126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.1108,
+ "step": 729
+ },
+ {
+ "epoch": 2.3322683706070286,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0917,
+ "step": 730
+ },
+ {
+ "epoch": 2.3354632587859427,
+ "grad_norm": 0.0859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0892,
+ "step": 731
+ },
+ {
+ "epoch": 2.3386581469648564,
+ "grad_norm": 0.11083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0988,
+ "step": 732
+ },
+ {
+ "epoch": 2.34185303514377,
+ "grad_norm": 0.08544921875,
+ "learning_rate": 0.0005,
+ "loss": 1.1011,
+ "step": 733
+ },
+ {
+ "epoch": 2.3450479233226837,
+ "grad_norm": 0.10546875,
+ "learning_rate": 0.0005,
+ "loss": 1.1031,
+ "step": 734
+ },
+ {
+ "epoch": 2.3482428115015974,
+ "grad_norm": 0.09033203125,
+ "learning_rate": 0.0005,
+ "loss": 1.1172,
+ "step": 735
+ },
+ {
+ "epoch": 2.351437699680511,
+ "grad_norm": 0.09130859375,
+ "learning_rate": 0.0005,
+ "loss": 1.1016,
+ "step": 736
+ },
+ {
+ "epoch": 2.3546325878594248,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0967,
+ "step": 737
+ },
+ {
+ "epoch": 2.357827476038339,
+ "grad_norm": 0.142578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0997,
+ "step": 738
+ },
+ {
+ "epoch": 2.3610223642172525,
+ "grad_norm": 0.09814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0896,
+ "step": 739
+ },
+ {
+ "epoch": 2.364217252396166,
+ "grad_norm": 0.09716796875,
+ "learning_rate": 0.0005,
+ "loss": 1.1012,
+ "step": 740
+ },
+ {
+ "epoch": 2.36741214057508,
+ "grad_norm": 0.0830078125,
+ "learning_rate": 0.0005,
+ "loss": 1.104,
+ "step": 741
+ },
+ {
+ "epoch": 2.3706070287539935,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.1165,
+ "step": 742
+ },
+ {
+ "epoch": 2.373801916932907,
+ "grad_norm": 0.1162109375,
+ "learning_rate": 0.0005,
+ "loss": 1.1077,
+ "step": 743
+ },
+ {
+ "epoch": 2.376996805111821,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0971,
+ "step": 744
+ },
+ {
+ "epoch": 2.380191693290735,
+ "grad_norm": 0.09375,
+ "learning_rate": 0.0005,
+ "loss": 1.106,
+ "step": 745
+ },
+ {
+ "epoch": 2.3833865814696487,
+ "grad_norm": 0.099609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0973,
+ "step": 746
+ },
+ {
+ "epoch": 2.3865814696485623,
+ "grad_norm": 0.126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0794,
+ "step": 747
+ },
+ {
+ "epoch": 2.389776357827476,
+ "grad_norm": 0.07861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.1059,
+ "step": 748
+ },
+ {
+ "epoch": 2.3929712460063897,
+ "grad_norm": 0.0966796875,
+ "learning_rate": 0.0005,
+ "loss": 1.1153,
+ "step": 749
+ },
+ {
+ "epoch": 2.3961661341853033,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.1157,
+ "step": 750
+ },
+ {
+ "epoch": 2.3993610223642174,
+ "grad_norm": 0.08154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.109,
+ "step": 751
+ },
+ {
+ "epoch": 2.402555910543131,
+ "grad_norm": 0.0810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.1029,
+ "step": 752
+ },
+ {
+ "epoch": 2.405750798722045,
+ "grad_norm": 0.1025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.1105,
+ "step": 753
+ },
+ {
+ "epoch": 2.4089456869009584,
+ "grad_norm": 0.1005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.1121,
+ "step": 754
+ },
+ {
+ "epoch": 2.412140575079872,
+ "grad_norm": 0.140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0969,
+ "step": 755
+ },
+ {
+ "epoch": 2.415335463258786,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0973,
+ "step": 756
+ },
+ {
+ "epoch": 2.4185303514377,
+ "grad_norm": 0.0859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0998,
+ "step": 757
+ },
+ {
+ "epoch": 2.4217252396166136,
+ "grad_norm": 0.0810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.1165,
+ "step": 758
+ },
+ {
+ "epoch": 2.4249201277955272,
+ "grad_norm": 0.09228515625,
+ "learning_rate": 0.0005,
+ "loss": 1.1114,
+ "step": 759
+ },
+ {
+ "epoch": 2.428115015974441,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.1019,
+ "step": 760
+ },
+ {
+ "epoch": 2.4313099041533546,
+ "grad_norm": 0.08203125,
+ "learning_rate": 0.0005,
+ "loss": 1.102,
+ "step": 761
+ },
+ {
+ "epoch": 2.4345047923322682,
+ "grad_norm": 0.08642578125,
+ "learning_rate": 0.0005,
+ "loss": 1.1138,
+ "step": 762
+ },
+ {
+ "epoch": 2.437699680511182,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.1089,
+ "step": 763
+ },
+ {
+ "epoch": 2.440894568690096,
+ "grad_norm": 0.0810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.1051,
+ "step": 764
+ },
+ {
+ "epoch": 2.4440894568690097,
+ "grad_norm": 0.09228515625,
+ "learning_rate": 0.0005,
+ "loss": 1.1015,
+ "step": 765
+ },
+ {
+ "epoch": 2.4472843450479234,
+ "grad_norm": 0.0830078125,
+ "learning_rate": 0.0005,
+ "loss": 1.1008,
+ "step": 766
+ },
+ {
+ "epoch": 2.450479233226837,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.1112,
+ "step": 767
+ },
+ {
+ "epoch": 2.4536741214057507,
+ "grad_norm": 0.08642578125,
+ "learning_rate": 0.0005,
+ "loss": 1.1064,
+ "step": 768
+ },
+ {
+ "epoch": 2.4568690095846644,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.1084,
+ "step": 769
+ },
+ {
+ "epoch": 2.460063897763578,
+ "grad_norm": 0.08251953125,
+ "learning_rate": 0.0005,
+ "loss": 1.1064,
+ "step": 770
+ },
+ {
+ "epoch": 2.463258785942492,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0965,
+ "step": 771
+ },
+ {
+ "epoch": 2.466453674121406,
+ "grad_norm": 0.0888671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0932,
+ "step": 772
+ },
+ {
+ "epoch": 2.4696485623003195,
+ "grad_norm": 0.0859375,
+ "learning_rate": 0.0005,
+ "loss": 1.1037,
+ "step": 773
+ },
+ {
+ "epoch": 2.472843450479233,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.1097,
+ "step": 774
+ },
+ {
+ "epoch": 2.476038338658147,
+ "grad_norm": 0.09814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0997,
+ "step": 775
+ },
+ {
+ "epoch": 2.479233226837061,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0958,
+ "step": 776
+ },
+ {
+ "epoch": 2.4824281150159746,
+ "grad_norm": 0.08935546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0845,
+ "step": 777
+ },
+ {
+ "epoch": 2.4856230031948883,
+ "grad_norm": 0.1220703125,
+ "learning_rate": 0.0005,
+ "loss": 1.1118,
+ "step": 778
+ },
+ {
+ "epoch": 2.488817891373802,
+ "grad_norm": 0.09912109375,
+ "learning_rate": 0.0005,
+ "loss": 1.1262,
+ "step": 779
+ },
+ {
+ "epoch": 2.4920127795527156,
+ "grad_norm": 0.09130859375,
+ "learning_rate": 0.0005,
+ "loss": 1.1232,
+ "step": 780
+ },
+ {
+ "epoch": 2.4952076677316293,
+ "grad_norm": 0.0908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0943,
+ "step": 781
+ },
+ {
+ "epoch": 2.498402555910543,
+ "grad_norm": 0.09521484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0979,
+ "step": 782
+ },
+ {
+ "epoch": 2.501597444089457,
+ "grad_norm": 0.08056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0945,
+ "step": 783
+ },
+ {
+ "epoch": 2.5047923322683707,
+ "grad_norm": 0.099609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0987,
+ "step": 784
+ },
+ {
+ "epoch": 2.5079872204472844,
+ "grad_norm": 0.1513671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0964,
+ "step": 785
+ },
+ {
+ "epoch": 2.511182108626198,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0983,
+ "step": 786
+ },
+ {
+ "epoch": 2.5143769968051117,
+ "grad_norm": 0.2236328125,
+ "learning_rate": 0.0005,
+ "loss": 1.1001,
+ "step": 787
+ },
+ {
+ "epoch": 2.5175718849840254,
+ "grad_norm": 0.263671875,
+ "learning_rate": 0.0005,
+ "loss": 1.1036,
+ "step": 788
+ },
+ {
+ "epoch": 2.520766773162939,
+ "grad_norm": 0.212890625,
+ "learning_rate": 0.0005,
+ "loss": 1.095,
+ "step": 789
+ },
+ {
+ "epoch": 2.523961661341853,
+ "grad_norm": 0.146484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0935,
+ "step": 790
+ },
+ {
+ "epoch": 2.527156549520767,
+ "grad_norm": 0.09423828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0946,
+ "step": 791
+ },
+ {
+ "epoch": 2.5303514376996805,
+ "grad_norm": 0.2021484375,
+ "learning_rate": 0.0005,
+ "loss": 1.1,
+ "step": 792
+ },
+ {
+ "epoch": 2.533546325878594,
+ "grad_norm": 0.2099609375,
+ "learning_rate": 0.0005,
+ "loss": 1.1093,
+ "step": 793
+ },
+ {
+ "epoch": 2.536741214057508,
+ "grad_norm": 0.11181640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0955,
+ "step": 794
+ },
+ {
+ "epoch": 2.539936102236422,
+ "grad_norm": 0.1259765625,
+ "learning_rate": 0.0005,
+ "loss": 1.1129,
+ "step": 795
+ },
+ {
+ "epoch": 2.543130990415335,
+ "grad_norm": 0.275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.1045,
+ "step": 796
+ },
+ {
+ "epoch": 2.5463258785942493,
+ "grad_norm": 0.29296875,
+ "learning_rate": 0.0005,
+ "loss": 1.116,
+ "step": 797
+ },
+ {
+ "epoch": 2.549520766773163,
+ "grad_norm": 0.1728515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0834,
+ "step": 798
+ },
+ {
+ "epoch": 2.5527156549520766,
+ "grad_norm": 0.09521484375,
+ "learning_rate": 0.0005,
+ "loss": 1.1062,
+ "step": 799
+ },
+ {
+ "epoch": 2.5559105431309903,
+ "grad_norm": 0.248046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0941,
+ "step": 800
+ },
+ {
+ "epoch": 2.559105431309904,
+ "grad_norm": 0.212890625,
+ "learning_rate": 0.0005,
+ "loss": 1.1071,
+ "step": 801
+ },
+ {
+ "epoch": 2.562300319488818,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.1003,
+ "step": 802
+ },
+ {
+ "epoch": 2.5654952076677318,
+ "grad_norm": 0.220703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0943,
+ "step": 803
+ },
+ {
+ "epoch": 2.5686900958466454,
+ "grad_norm": 0.1728515625,
+ "learning_rate": 0.0005,
+ "loss": 1.1001,
+ "step": 804
+ },
+ {
+ "epoch": 2.571884984025559,
+ "grad_norm": 0.0888671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0928,
+ "step": 805
+ },
+ {
+ "epoch": 2.5750798722044728,
+ "grad_norm": 0.1767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.1016,
+ "step": 806
+ },
+ {
+ "epoch": 2.5782747603833864,
+ "grad_norm": 0.08837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0947,
+ "step": 807
+ },
+ {
+ "epoch": 2.5814696485623,
+ "grad_norm": 0.11328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0942,
+ "step": 808
+ },
+ {
+ "epoch": 2.584664536741214,
+ "grad_norm": 0.1748046875,
+ "learning_rate": 0.0005,
+ "loss": 1.1072,
+ "step": 809
+ },
+ {
+ "epoch": 2.587859424920128,
+ "grad_norm": 0.1513671875,
+ "learning_rate": 0.0005,
+ "loss": 1.1104,
+ "step": 810
+ },
+ {
+ "epoch": 2.5910543130990416,
+ "grad_norm": 0.12890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0865,
+ "step": 811
+ },
+ {
+ "epoch": 2.594249201277955,
+ "grad_norm": 0.244140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0935,
+ "step": 812
+ },
+ {
+ "epoch": 2.597444089456869,
+ "grad_norm": 0.173828125,
+ "learning_rate": 0.0005,
+ "loss": 1.1089,
+ "step": 813
+ },
+ {
+ "epoch": 2.600638977635783,
+ "grad_norm": 0.162109375,
+ "learning_rate": 0.0005,
+ "loss": 1.1105,
+ "step": 814
+ },
+ {
+ "epoch": 2.6038338658146962,
+ "grad_norm": 0.1669921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0776,
+ "step": 815
+ },
+ {
+ "epoch": 2.6070287539936103,
+ "grad_norm": 0.1376953125,
+ "learning_rate": 0.0005,
+ "loss": 1.1075,
+ "step": 816
+ },
+ {
+ "epoch": 2.610223642172524,
+ "grad_norm": 0.185546875,
+ "learning_rate": 0.0005,
+ "loss": 1.1121,
+ "step": 817
+ },
+ {
+ "epoch": 2.6134185303514377,
+ "grad_norm": 0.1259765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0886,
+ "step": 818
+ },
+ {
+ "epoch": 2.6166134185303513,
+ "grad_norm": 0.0966796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0999,
+ "step": 819
+ },
+ {
+ "epoch": 2.619808306709265,
+ "grad_norm": 0.154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.1047,
+ "step": 820
+ },
+ {
+ "epoch": 2.623003194888179,
+ "grad_norm": 0.11865234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0815,
+ "step": 821
+ },
+ {
+ "epoch": 2.626198083067093,
+ "grad_norm": 0.08154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.1172,
+ "step": 822
+ },
+ {
+ "epoch": 2.6293929712460065,
+ "grad_norm": 0.138671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0992,
+ "step": 823
+ },
+ {
+ "epoch": 2.63258785942492,
+ "grad_norm": 0.091796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0826,
+ "step": 824
+ },
+ {
+ "epoch": 2.635782747603834,
+ "grad_norm": 0.11083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0995,
+ "step": 825
+ },
+ {
+ "epoch": 2.6389776357827475,
+ "grad_norm": 0.1494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.1072,
+ "step": 826
+ },
+ {
+ "epoch": 2.642172523961661,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0954,
+ "step": 827
+ },
+ {
+ "epoch": 2.6453674121405752,
+ "grad_norm": 0.1005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.1016,
+ "step": 828
+ },
+ {
+ "epoch": 2.648562300319489,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0983,
+ "step": 829
+ },
+ {
+ "epoch": 2.6517571884984026,
+ "grad_norm": 0.0849609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0962,
+ "step": 830
+ },
+ {
+ "epoch": 2.6549520766773163,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.1043,
+ "step": 831
+ },
+ {
+ "epoch": 2.65814696485623,
+ "grad_norm": 0.062255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0986,
+ "step": 832
+ },
+ {
+ "epoch": 2.661341853035144,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0971,
+ "step": 833
+ },
+ {
+ "epoch": 2.6645367412140573,
+ "grad_norm": 0.0810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.1011,
+ "step": 834
+ },
+ {
+ "epoch": 2.6677316293929714,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0994,
+ "step": 835
+ },
+ {
+ "epoch": 2.670926517571885,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.1145,
+ "step": 836
+ },
+ {
+ "epoch": 2.6741214057507987,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0972,
+ "step": 837
+ },
+ {
+ "epoch": 2.6773162939297124,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.1172,
+ "step": 838
+ },
+ {
+ "epoch": 2.680511182108626,
+ "grad_norm": 0.1552734375,
+ "learning_rate": 0.0005,
+ "loss": 1.1025,
+ "step": 839
+ },
+ {
+ "epoch": 2.68370607028754,
+ "grad_norm": 0.1015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0958,
+ "step": 840
+ },
+ {
+ "epoch": 2.686900958466454,
+ "grad_norm": 0.130859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0992,
+ "step": 841
+ },
+ {
+ "epoch": 2.6900958466453675,
+ "grad_norm": 0.2060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0977,
+ "step": 842
+ },
+ {
+ "epoch": 2.693290734824281,
+ "grad_norm": 0.169921875,
+ "learning_rate": 0.0005,
+ "loss": 1.1144,
+ "step": 843
+ },
+ {
+ "epoch": 2.696485623003195,
+ "grad_norm": 0.08349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.097,
+ "step": 844
+ },
+ {
+ "epoch": 2.6996805111821085,
+ "grad_norm": 0.099609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0832,
+ "step": 845
+ },
+ {
+ "epoch": 2.702875399361022,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.1095,
+ "step": 846
+ },
+ {
+ "epoch": 2.7060702875399363,
+ "grad_norm": 0.08740234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0931,
+ "step": 847
+ },
+ {
+ "epoch": 2.70926517571885,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0993,
+ "step": 848
+ },
+ {
+ "epoch": 2.7124600638977636,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0948,
+ "step": 849
+ },
+ {
+ "epoch": 2.7156549520766773,
+ "grad_norm": 0.08251953125,
+ "learning_rate": 0.0005,
+ "loss": 1.12,
+ "step": 850
+ },
+ {
+ "epoch": 2.718849840255591,
+ "grad_norm": 0.0986328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0868,
+ "step": 851
+ },
+ {
+ "epoch": 2.722044728434505,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0945,
+ "step": 852
+ },
+ {
+ "epoch": 2.7252396166134183,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0869,
+ "step": 853
+ },
+ {
+ "epoch": 2.7284345047923324,
+ "grad_norm": 0.09716796875,
+ "learning_rate": 0.0005,
+ "loss": 1.1022,
+ "step": 854
+ },
+ {
+ "epoch": 2.731629392971246,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.095,
+ "step": 855
+ },
+ {
+ "epoch": 2.7348242811501597,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0948,
+ "step": 856
+ },
+ {
+ "epoch": 2.7380191693290734,
+ "grad_norm": 0.062255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0977,
+ "step": 857
+ },
+ {
+ "epoch": 2.741214057507987,
+ "grad_norm": 0.0810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0972,
+ "step": 858
+ },
+ {
+ "epoch": 2.744408945686901,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.1033,
+ "step": 859
+ },
+ {
+ "epoch": 2.747603833865815,
+ "grad_norm": 0.08251953125,
+ "learning_rate": 0.0005,
+ "loss": 1.1016,
+ "step": 860
+ },
+ {
+ "epoch": 2.7507987220447285,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.1081,
+ "step": 861
+ },
+ {
+ "epoch": 2.753993610223642,
+ "grad_norm": 0.10595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0859,
+ "step": 862
+ },
+ {
+ "epoch": 2.757188498402556,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.1026,
+ "step": 863
+ },
+ {
+ "epoch": 2.7603833865814695,
+ "grad_norm": 0.11376953125,
+ "learning_rate": 0.0005,
+ "loss": 1.1045,
+ "step": 864
+ },
+ {
+ "epoch": 2.763578274760383,
+ "grad_norm": 0.09912109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0882,
+ "step": 865
+ },
+ {
+ "epoch": 2.7667731629392973,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.1007,
+ "step": 866
+ },
+ {
+ "epoch": 2.769968051118211,
+ "grad_norm": 0.08154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0997,
+ "step": 867
+ },
+ {
+ "epoch": 2.7731629392971247,
+ "grad_norm": 0.0869140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0893,
+ "step": 868
+ },
+ {
+ "epoch": 2.7763578274760383,
+ "grad_norm": 0.099609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0894,
+ "step": 869
+ },
+ {
+ "epoch": 2.779552715654952,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.102,
+ "step": 870
+ },
+ {
+ "epoch": 2.7827476038338657,
+ "grad_norm": 0.1005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.104,
+ "step": 871
+ },
+ {
+ "epoch": 2.7859424920127793,
+ "grad_norm": 0.0888671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0934,
+ "step": 872
+ },
+ {
+ "epoch": 2.7891373801916934,
+ "grad_norm": 0.09814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0995,
+ "step": 873
+ },
+ {
+ "epoch": 2.792332268370607,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.111,
+ "step": 874
+ },
+ {
+ "epoch": 2.7955271565495208,
+ "grad_norm": 0.10205078125,
+ "learning_rate": 0.0005,
+ "loss": 1.1073,
+ "step": 875
+ },
+ {
+ "epoch": 2.7987220447284344,
+ "grad_norm": 0.119140625,
+ "learning_rate": 0.0005,
+ "loss": 1.1075,
+ "step": 876
+ },
+ {
+ "epoch": 2.801916932907348,
+ "grad_norm": 0.1513671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0898,
+ "step": 877
+ },
+ {
+ "epoch": 2.8051118210862622,
+ "grad_norm": 0.10009765625,
+ "learning_rate": 0.0005,
+ "loss": 1.1046,
+ "step": 878
+ },
+ {
+ "epoch": 2.8083067092651754,
+ "grad_norm": 0.17578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0817,
+ "step": 879
+ },
+ {
+ "epoch": 2.8115015974440896,
+ "grad_norm": 0.11669921875,
+ "learning_rate": 0.0005,
+ "loss": 1.1064,
+ "step": 880
+ },
+ {
+ "epoch": 2.8146964856230032,
+ "grad_norm": 0.203125,
+ "learning_rate": 0.0005,
+ "loss": 1.1053,
+ "step": 881
+ },
+ {
+ "epoch": 2.817891373801917,
+ "grad_norm": 0.10205078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0955,
+ "step": 882
+ },
+ {
+ "epoch": 2.8210862619808306,
+ "grad_norm": 0.08642578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0985,
+ "step": 883
+ },
+ {
+ "epoch": 2.8242811501597442,
+ "grad_norm": 0.0947265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0953,
+ "step": 884
+ },
+ {
+ "epoch": 2.8274760383386583,
+ "grad_norm": 0.095703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0971,
+ "step": 885
+ },
+ {
+ "epoch": 2.830670926517572,
+ "grad_norm": 0.126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.1059,
+ "step": 886
+ },
+ {
+ "epoch": 2.8338658146964857,
+ "grad_norm": 0.125,
+ "learning_rate": 0.0005,
+ "loss": 1.0895,
+ "step": 887
+ },
+ {
+ "epoch": 2.8370607028753994,
+ "grad_norm": 0.1025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0987,
+ "step": 888
+ },
+ {
+ "epoch": 2.840255591054313,
+ "grad_norm": 0.0859375,
+ "learning_rate": 0.0005,
+ "loss": 1.1066,
+ "step": 889
+ },
+ {
+ "epoch": 2.8434504792332267,
+ "grad_norm": 0.1748046875,
+ "learning_rate": 0.0005,
+ "loss": 1.1002,
+ "step": 890
+ },
+ {
+ "epoch": 2.8466453674121404,
+ "grad_norm": 0.12158203125,
+ "learning_rate": 0.0005,
+ "loss": 1.091,
+ "step": 891
+ },
+ {
+ "epoch": 2.8498402555910545,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0863,
+ "step": 892
+ },
+ {
+ "epoch": 2.853035143769968,
+ "grad_norm": 0.1015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0996,
+ "step": 893
+ },
+ {
+ "epoch": 2.856230031948882,
+ "grad_norm": 0.0986328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0997,
+ "step": 894
+ },
+ {
+ "epoch": 2.8594249201277955,
+ "grad_norm": 0.061767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0896,
+ "step": 895
+ },
+ {
+ "epoch": 2.862619808306709,
+ "grad_norm": 0.09228515625,
+ "learning_rate": 0.0005,
+ "loss": 1.1062,
+ "step": 896
+ },
+ {
+ "epoch": 2.8658146964856233,
+ "grad_norm": 0.083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0952,
+ "step": 897
+ },
+ {
+ "epoch": 2.8690095846645365,
+ "grad_norm": 0.056884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.1112,
+ "step": 898
+ },
+ {
+ "epoch": 2.8722044728434506,
+ "grad_norm": 0.08251953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0973,
+ "step": 899
+ },
+ {
+ "epoch": 2.8753993610223643,
+ "grad_norm": 0.10546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0963,
+ "step": 900
+ },
+ {
+ "epoch": 2.878594249201278,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0943,
+ "step": 901
+ },
+ {
+ "epoch": 2.8817891373801916,
+ "grad_norm": 0.0986328125,
+ "learning_rate": 0.0005,
+ "loss": 1.1071,
+ "step": 902
+ },
+ {
+ "epoch": 2.8849840255591053,
+ "grad_norm": 0.09814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.1058,
+ "step": 903
+ },
+ {
+ "epoch": 2.8881789137380194,
+ "grad_norm": 0.1240234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0809,
+ "step": 904
+ },
+ {
+ "epoch": 2.891373801916933,
+ "grad_norm": 0.1083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.1034,
+ "step": 905
+ },
+ {
+ "epoch": 2.8945686900958467,
+ "grad_norm": 0.08203125,
+ "learning_rate": 0.0005,
+ "loss": 1.106,
+ "step": 906
+ },
+ {
+ "epoch": 2.8977635782747604,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.088,
+ "step": 907
+ },
+ {
+ "epoch": 2.900958466453674,
+ "grad_norm": 0.0966796875,
+ "learning_rate": 0.0005,
+ "loss": 1.1068,
+ "step": 908
+ },
+ {
+ "epoch": 2.9041533546325877,
+ "grad_norm": 0.11474609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0917,
+ "step": 909
+ },
+ {
+ "epoch": 2.9073482428115014,
+ "grad_norm": 0.0869140625,
+ "learning_rate": 0.0005,
+ "loss": 1.1022,
+ "step": 910
+ },
+ {
+ "epoch": 2.9105431309904155,
+ "grad_norm": 0.08349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0893,
+ "step": 911
+ },
+ {
+ "epoch": 2.913738019169329,
+ "grad_norm": 0.1123046875,
+ "learning_rate": 0.0005,
+ "loss": 1.1024,
+ "step": 912
+ },
+ {
+ "epoch": 2.916932907348243,
+ "grad_norm": 0.099609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0825,
+ "step": 913
+ },
+ {
+ "epoch": 2.9201277955271565,
+ "grad_norm": 0.12060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0911,
+ "step": 914
+ },
+ {
+ "epoch": 2.92332268370607,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0865,
+ "step": 915
+ },
+ {
+ "epoch": 2.9265175718849843,
+ "grad_norm": 0.10400390625,
+ "learning_rate": 0.0005,
+ "loss": 1.077,
+ "step": 916
+ },
+ {
+ "epoch": 2.9297124600638975,
+ "grad_norm": 0.087890625,
+ "learning_rate": 0.0005,
+ "loss": 1.1067,
+ "step": 917
+ },
+ {
+ "epoch": 2.9329073482428116,
+ "grad_norm": 0.0947265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0932,
+ "step": 918
+ },
+ {
+ "epoch": 2.9361022364217253,
+ "grad_norm": 0.10400390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0902,
+ "step": 919
+ },
+ {
+ "epoch": 2.939297124600639,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0961,
+ "step": 920
+ },
+ {
+ "epoch": 2.9424920127795526,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.092,
+ "step": 921
+ },
+ {
+ "epoch": 2.9456869009584663,
+ "grad_norm": 0.0830078125,
+ "learning_rate": 0.0005,
+ "loss": 1.1035,
+ "step": 922
+ },
+ {
+ "epoch": 2.9488817891373804,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.1038,
+ "step": 923
+ },
+ {
+ "epoch": 2.952076677316294,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0801,
+ "step": 924
+ },
+ {
+ "epoch": 2.9552715654952078,
+ "grad_norm": 0.10400390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0986,
+ "step": 925
+ },
+ {
+ "epoch": 2.9584664536741214,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.1021,
+ "step": 926
+ },
+ {
+ "epoch": 2.961661341853035,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.1058,
+ "step": 927
+ },
+ {
+ "epoch": 2.9648562300319488,
+ "grad_norm": 0.123046875,
+ "learning_rate": 0.0005,
+ "loss": 1.092,
+ "step": 928
+ },
+ {
+ "epoch": 2.9680511182108624,
+ "grad_norm": 0.1162109375,
+ "learning_rate": 0.0005,
+ "loss": 1.095,
+ "step": 929
+ },
+ {
+ "epoch": 2.9712460063897765,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0921,
+ "step": 930
+ },
+ {
+ "epoch": 2.97444089456869,
+ "grad_norm": 0.0859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0984,
+ "step": 931
+ },
+ {
+ "epoch": 2.977635782747604,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.1027,
+ "step": 932
+ },
+ {
+ "epoch": 2.9808306709265175,
+ "grad_norm": 0.080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.1098,
+ "step": 933
+ },
+ {
+ "epoch": 2.984025559105431,
+ "grad_norm": 0.1669921875,
+ "learning_rate": 0.0005,
+ "loss": 1.087,
+ "step": 934
+ },
+ {
+ "epoch": 2.987220447284345,
+ "grad_norm": 0.1337890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0834,
+ "step": 935
+ },
+ {
+ "epoch": 2.9904153354632586,
+ "grad_norm": 0.0869140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0829,
+ "step": 936
+ },
+ {
+ "epoch": 2.9936102236421727,
+ "grad_norm": 0.111328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0968,
+ "step": 937
+ },
+ {
+ "epoch": 2.9968051118210863,
+ "grad_norm": 0.111328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0981,
+ "step": 938
+ },
+ {
+ "epoch": 3.0,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0997,
+ "step": 939
+ },
+ {
+ "epoch": 3.0031948881789137,
+ "grad_norm": 0.10595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0923,
+ "step": 940
+ },
+ {
+ "epoch": 3.0063897763578273,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0984,
+ "step": 941
+ },
+ {
+ "epoch": 3.009584664536741,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0866,
+ "step": 942
+ },
+ {
+ "epoch": 3.012779552715655,
+ "grad_norm": 0.08349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.1114,
+ "step": 943
+ },
+ {
+ "epoch": 3.015974440894569,
+ "grad_norm": 0.173828125,
+ "learning_rate": 0.0005,
+ "loss": 1.1098,
+ "step": 944
+ },
+ {
+ "epoch": 3.0191693290734825,
+ "grad_norm": 0.267578125,
+ "learning_rate": 0.0005,
+ "loss": 1.1078,
+ "step": 945
+ },
+ {
+ "epoch": 3.022364217252396,
+ "grad_norm": 0.3046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0974,
+ "step": 946
+ },
+ {
+ "epoch": 3.02555910543131,
+ "grad_norm": 0.40234375,
+ "learning_rate": 0.0005,
+ "loss": 1.1002,
+ "step": 947
+ },
+ {
+ "epoch": 3.0287539936102235,
+ "grad_norm": 0.66796875,
+ "learning_rate": 0.0005,
+ "loss": 1.1038,
+ "step": 948
+ },
+ {
+ "epoch": 3.0319488817891376,
+ "grad_norm": 1.5859375,
+ "learning_rate": 0.0005,
+ "loss": 1.1155,
+ "step": 949
+ },
+ {
+ "epoch": 3.0351437699680512,
+ "grad_norm": 1.1875,
+ "learning_rate": 0.0005,
+ "loss": 1.1169,
+ "step": 950
+ },
+ {
+ "epoch": 3.038338658146965,
+ "grad_norm": 0.72265625,
+ "learning_rate": 0.0005,
+ "loss": 1.1095,
+ "step": 951
+ },
+ {
+ "epoch": 3.0415335463258786,
+ "grad_norm": 0.330078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0951,
+ "step": 952
+ },
+ {
+ "epoch": 3.0447284345047922,
+ "grad_norm": 0.369140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0994,
+ "step": 953
+ },
+ {
+ "epoch": 3.047923322683706,
+ "grad_norm": 0.404296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0989,
+ "step": 954
+ },
+ {
+ "epoch": 3.0511182108626196,
+ "grad_norm": 0.224609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0864,
+ "step": 955
+ },
+ {
+ "epoch": 3.0543130990415337,
+ "grad_norm": 0.279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.1107,
+ "step": 956
+ },
+ {
+ "epoch": 3.0575079872204474,
+ "grad_norm": 0.201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0921,
+ "step": 957
+ },
+ {
+ "epoch": 3.060702875399361,
+ "grad_norm": 0.2470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0972,
+ "step": 958
+ },
+ {
+ "epoch": 3.0638977635782747,
+ "grad_norm": 0.203125,
+ "learning_rate": 0.0005,
+ "loss": 1.1043,
+ "step": 959
+ },
+ {
+ "epoch": 3.0670926517571884,
+ "grad_norm": 0.365234375,
+ "learning_rate": 0.0005,
+ "loss": 1.1,
+ "step": 960
+ },
+ {
+ "epoch": 3.070287539936102,
+ "grad_norm": 0.1533203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0784,
+ "step": 961
+ },
+ {
+ "epoch": 3.073482428115016,
+ "grad_norm": 0.2001953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0842,
+ "step": 962
+ },
+ {
+ "epoch": 3.07667731629393,
+ "grad_norm": 0.11376953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0985,
+ "step": 963
+ },
+ {
+ "epoch": 3.0798722044728435,
+ "grad_norm": 0.1669921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0918,
+ "step": 964
+ },
+ {
+ "epoch": 3.083067092651757,
+ "grad_norm": 0.1318359375,
+ "learning_rate": 0.0005,
+ "loss": 1.1076,
+ "step": 965
+ },
+ {
+ "epoch": 3.086261980830671,
+ "grad_norm": 0.1103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0999,
+ "step": 966
+ },
+ {
+ "epoch": 3.0894568690095845,
+ "grad_norm": 0.1240234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0946,
+ "step": 967
+ },
+ {
+ "epoch": 3.0926517571884986,
+ "grad_norm": 0.08984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0923,
+ "step": 968
+ },
+ {
+ "epoch": 3.0958466453674123,
+ "grad_norm": 0.11328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0943,
+ "step": 969
+ },
+ {
+ "epoch": 3.099041533546326,
+ "grad_norm": 0.1123046875,
+ "learning_rate": 0.0005,
+ "loss": 1.1088,
+ "step": 970
+ },
+ {
+ "epoch": 3.1022364217252396,
+ "grad_norm": 0.1396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0903,
+ "step": 971
+ },
+ {
+ "epoch": 3.1054313099041533,
+ "grad_norm": 0.107421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0981,
+ "step": 972
+ },
+ {
+ "epoch": 3.108626198083067,
+ "grad_norm": 0.10498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0764,
+ "step": 973
+ },
+ {
+ "epoch": 3.1118210862619806,
+ "grad_norm": 0.0908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0944,
+ "step": 974
+ },
+ {
+ "epoch": 3.1150159744408947,
+ "grad_norm": 0.1005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0875,
+ "step": 975
+ },
+ {
+ "epoch": 3.1182108626198084,
+ "grad_norm": 0.0947265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0897,
+ "step": 976
+ },
+ {
+ "epoch": 3.121405750798722,
+ "grad_norm": 0.08251953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0933,
+ "step": 977
+ },
+ {
+ "epoch": 3.1246006389776357,
+ "grad_norm": 0.10498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0982,
+ "step": 978
+ },
+ {
+ "epoch": 3.1277955271565494,
+ "grad_norm": 0.08203125,
+ "learning_rate": 0.0005,
+ "loss": 1.1015,
+ "step": 979
+ },
+ {
+ "epoch": 3.130990415335463,
+ "grad_norm": 0.1474609375,
+ "learning_rate": 0.0005,
+ "loss": 1.1006,
+ "step": 980
+ },
+ {
+ "epoch": 3.134185303514377,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0719,
+ "step": 981
+ },
+ {
+ "epoch": 3.137380191693291,
+ "grad_norm": 0.11669921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0759,
+ "step": 982
+ },
+ {
+ "epoch": 3.1405750798722045,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0928,
+ "step": 983
+ },
+ {
+ "epoch": 3.143769968051118,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.097,
+ "step": 984
+ },
+ {
+ "epoch": 3.146964856230032,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0966,
+ "step": 985
+ },
+ {
+ "epoch": 3.1501597444089455,
+ "grad_norm": 0.09033203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0989,
+ "step": 986
+ },
+ {
+ "epoch": 3.1533546325878596,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0948,
+ "step": 987
+ },
+ {
+ "epoch": 3.1565495207667733,
+ "grad_norm": 0.08056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0975,
+ "step": 988
+ },
+ {
+ "epoch": 3.159744408945687,
+ "grad_norm": 0.08056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0979,
+ "step": 989
+ },
+ {
+ "epoch": 3.1629392971246006,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0755,
+ "step": 990
+ },
+ {
+ "epoch": 3.1661341853035143,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0803,
+ "step": 991
+ },
+ {
+ "epoch": 3.169329073482428,
+ "grad_norm": 0.10791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0944,
+ "step": 992
+ },
+ {
+ "epoch": 3.1725239616613417,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0879,
+ "step": 993
+ },
+ {
+ "epoch": 3.1757188498402558,
+ "grad_norm": 0.09716796875,
+ "learning_rate": 0.0005,
+ "loss": 1.1092,
+ "step": 994
+ },
+ {
+ "epoch": 3.1789137380191694,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0953,
+ "step": 995
+ },
+ {
+ "epoch": 3.182108626198083,
+ "grad_norm": 0.12158203125,
+ "learning_rate": 0.0005,
+ "loss": 1.1004,
+ "step": 996
+ },
+ {
+ "epoch": 3.1853035143769968,
+ "grad_norm": 0.1513671875,
+ "learning_rate": 0.0005,
+ "loss": 1.1067,
+ "step": 997
+ },
+ {
+ "epoch": 3.1884984025559104,
+ "grad_norm": 0.09716796875,
+ "learning_rate": 0.0005,
+ "loss": 1.099,
+ "step": 998
+ },
+ {
+ "epoch": 3.191693290734824,
+ "grad_norm": 0.201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.103,
+ "step": 999
+ },
+ {
+ "epoch": 3.194888178913738,
+ "grad_norm": 0.080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.087,
+ "step": 1000
+ },
+ {
+ "epoch": 3.198083067092652,
+ "grad_norm": 0.212890625,
+ "learning_rate": 0.0005,
+ "loss": 1.1029,
+ "step": 1001
+ },
+ {
+ "epoch": 3.2012779552715656,
+ "grad_norm": 0.12060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0902,
+ "step": 1002
+ },
+ {
+ "epoch": 3.2044728434504792,
+ "grad_norm": 0.146484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0848,
+ "step": 1003
+ },
+ {
+ "epoch": 3.207667731629393,
+ "grad_norm": 0.142578125,
+ "learning_rate": 0.0005,
+ "loss": 1.1095,
+ "step": 1004
+ },
+ {
+ "epoch": 3.2108626198083066,
+ "grad_norm": 0.1640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0871,
+ "step": 1005
+ },
+ {
+ "epoch": 3.2140575079872207,
+ "grad_norm": 0.13671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0954,
+ "step": 1006
+ },
+ {
+ "epoch": 3.2172523961661343,
+ "grad_norm": 0.130859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0858,
+ "step": 1007
+ },
+ {
+ "epoch": 3.220447284345048,
+ "grad_norm": 0.1279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0962,
+ "step": 1008
+ },
+ {
+ "epoch": 3.2236421725239617,
+ "grad_norm": 0.10009765625,
+ "learning_rate": 0.0005,
+ "loss": 1.1005,
+ "step": 1009
+ },
+ {
+ "epoch": 3.2268370607028753,
+ "grad_norm": 0.1357421875,
+ "learning_rate": 0.0005,
+ "loss": 1.1042,
+ "step": 1010
+ },
+ {
+ "epoch": 3.230031948881789,
+ "grad_norm": 0.0927734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0863,
+ "step": 1011
+ },
+ {
+ "epoch": 3.2332268370607027,
+ "grad_norm": 0.125,
+ "learning_rate": 0.0005,
+ "loss": 1.0911,
+ "step": 1012
+ },
+ {
+ "epoch": 3.236421725239617,
+ "grad_norm": 0.08447265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0738,
+ "step": 1013
+ },
+ {
+ "epoch": 3.2396166134185305,
+ "grad_norm": 0.11181640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0901,
+ "step": 1014
+ },
+ {
+ "epoch": 3.242811501597444,
+ "grad_norm": 0.0859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0953,
+ "step": 1015
+ },
+ {
+ "epoch": 3.246006389776358,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0857,
+ "step": 1016
+ },
+ {
+ "epoch": 3.2492012779552715,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0983,
+ "step": 1017
+ },
+ {
+ "epoch": 3.252396166134185,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0806,
+ "step": 1018
+ },
+ {
+ "epoch": 3.255591054313099,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0903,
+ "step": 1019
+ },
+ {
+ "epoch": 3.258785942492013,
+ "grad_norm": 0.1103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.1002,
+ "step": 1020
+ },
+ {
+ "epoch": 3.2619808306709266,
+ "grad_norm": 0.083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0899,
+ "step": 1021
+ },
+ {
+ "epoch": 3.2651757188498403,
+ "grad_norm": 0.0947265625,
+ "learning_rate": 0.0005,
+ "loss": 1.098,
+ "step": 1022
+ },
+ {
+ "epoch": 3.268370607028754,
+ "grad_norm": 0.080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0935,
+ "step": 1023
+ },
+ {
+ "epoch": 3.2715654952076676,
+ "grad_norm": 0.09130859375,
+ "learning_rate": 0.0005,
+ "loss": 1.096,
+ "step": 1024
+ },
+ {
+ "epoch": 3.2747603833865817,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0903,
+ "step": 1025
+ },
+ {
+ "epoch": 3.2779552715654954,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0915,
+ "step": 1026
+ },
+ {
+ "epoch": 3.281150159744409,
+ "grad_norm": 0.0859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0846,
+ "step": 1027
+ },
+ {
+ "epoch": 3.2843450479233227,
+ "grad_norm": 0.10791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.1001,
+ "step": 1028
+ },
+ {
+ "epoch": 3.2875399361022364,
+ "grad_norm": 0.12255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.1052,
+ "step": 1029
+ },
+ {
+ "epoch": 3.29073482428115,
+ "grad_norm": 0.11279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0882,
+ "step": 1030
+ },
+ {
+ "epoch": 3.2939297124600637,
+ "grad_norm": 0.09326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.1,
+ "step": 1031
+ },
+ {
+ "epoch": 3.297124600638978,
+ "grad_norm": 0.1318359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0888,
+ "step": 1032
+ },
+ {
+ "epoch": 3.3003194888178915,
+ "grad_norm": 0.095703125,
+ "learning_rate": 0.0005,
+ "loss": 1.1017,
+ "step": 1033
+ },
+ {
+ "epoch": 3.303514376996805,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0935,
+ "step": 1034
+ },
+ {
+ "epoch": 3.306709265175719,
+ "grad_norm": 0.1142578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0804,
+ "step": 1035
+ },
+ {
+ "epoch": 3.3099041533546325,
+ "grad_norm": 0.08251953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0969,
+ "step": 1036
+ },
+ {
+ "epoch": 3.313099041533546,
+ "grad_norm": 0.1494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0896,
+ "step": 1037
+ },
+ {
+ "epoch": 3.31629392971246,
+ "grad_norm": 0.12890625,
+ "learning_rate": 0.0005,
+ "loss": 1.1082,
+ "step": 1038
+ },
+ {
+ "epoch": 3.319488817891374,
+ "grad_norm": 0.09326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.1004,
+ "step": 1039
+ },
+ {
+ "epoch": 3.3226837060702876,
+ "grad_norm": 0.0986328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0938,
+ "step": 1040
+ },
+ {
+ "epoch": 3.3258785942492013,
+ "grad_norm": 0.12353515625,
+ "learning_rate": 0.0005,
+ "loss": 1.1027,
+ "step": 1041
+ },
+ {
+ "epoch": 3.329073482428115,
+ "grad_norm": 0.0888671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0971,
+ "step": 1042
+ },
+ {
+ "epoch": 3.3322683706070286,
+ "grad_norm": 0.1142578125,
+ "learning_rate": 0.0005,
+ "loss": 1.094,
+ "step": 1043
+ },
+ {
+ "epoch": 3.3354632587859427,
+ "grad_norm": 0.099609375,
+ "learning_rate": 0.0005,
+ "loss": 1.103,
+ "step": 1044
+ },
+ {
+ "epoch": 3.3386581469648564,
+ "grad_norm": 0.10693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0929,
+ "step": 1045
+ },
+ {
+ "epoch": 3.34185303514377,
+ "grad_norm": 0.13671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0831,
+ "step": 1046
+ },
+ {
+ "epoch": 3.3450479233226837,
+ "grad_norm": 0.0859375,
+ "learning_rate": 0.0005,
+ "loss": 1.1027,
+ "step": 1047
+ },
+ {
+ "epoch": 3.3482428115015974,
+ "grad_norm": 0.1318359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0957,
+ "step": 1048
+ },
+ {
+ "epoch": 3.351437699680511,
+ "grad_norm": 0.1767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0983,
+ "step": 1049
+ },
+ {
+ "epoch": 3.3546325878594248,
+ "grad_norm": 0.138671875,
+ "learning_rate": 0.0005,
+ "loss": 1.1046,
+ "step": 1050
+ },
+ {
+ "epoch": 3.357827476038339,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0912,
+ "step": 1051
+ },
+ {
+ "epoch": 3.3610223642172525,
+ "grad_norm": 0.10009765625,
+ "learning_rate": 0.0005,
+ "loss": 1.1045,
+ "step": 1052
+ },
+ {
+ "epoch": 3.364217252396166,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0931,
+ "step": 1053
+ },
+ {
+ "epoch": 3.36741214057508,
+ "grad_norm": 0.15234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0911,
+ "step": 1054
+ },
+ {
+ "epoch": 3.3706070287539935,
+ "grad_norm": 0.12890625,
+ "learning_rate": 0.0005,
+ "loss": 1.107,
+ "step": 1055
+ },
+ {
+ "epoch": 3.373801916932907,
+ "grad_norm": 0.146484375,
+ "learning_rate": 0.0005,
+ "loss": 1.083,
+ "step": 1056
+ },
+ {
+ "epoch": 3.376996805111821,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0829,
+ "step": 1057
+ },
+ {
+ "epoch": 3.380191693290735,
+ "grad_norm": 0.0966796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0863,
+ "step": 1058
+ },
+ {
+ "epoch": 3.3833865814696487,
+ "grad_norm": 0.1103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0923,
+ "step": 1059
+ },
+ {
+ "epoch": 3.3865814696485623,
+ "grad_norm": 0.0859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0783,
+ "step": 1060
+ },
+ {
+ "epoch": 3.389776357827476,
+ "grad_norm": 0.095703125,
+ "learning_rate": 0.0005,
+ "loss": 1.09,
+ "step": 1061
+ },
+ {
+ "epoch": 3.3929712460063897,
+ "grad_norm": 0.09619140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0954,
+ "step": 1062
+ },
+ {
+ "epoch": 3.3961661341853033,
+ "grad_norm": 0.080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0941,
+ "step": 1063
+ },
+ {
+ "epoch": 3.3993610223642174,
+ "grad_norm": 0.111328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0724,
+ "step": 1064
+ },
+ {
+ "epoch": 3.402555910543131,
+ "grad_norm": 0.08056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0881,
+ "step": 1065
+ },
+ {
+ "epoch": 3.405750798722045,
+ "grad_norm": 0.0947265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0953,
+ "step": 1066
+ },
+ {
+ "epoch": 3.4089456869009584,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0959,
+ "step": 1067
+ },
+ {
+ "epoch": 3.412140575079872,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.1014,
+ "step": 1068
+ },
+ {
+ "epoch": 3.415335463258786,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0808,
+ "step": 1069
+ },
+ {
+ "epoch": 3.4185303514377,
+ "grad_norm": 0.08056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0939,
+ "step": 1070
+ },
+ {
+ "epoch": 3.4217252396166136,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0935,
+ "step": 1071
+ },
+ {
+ "epoch": 3.4249201277955272,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0952,
+ "step": 1072
+ },
+ {
+ "epoch": 3.428115015974441,
+ "grad_norm": 0.12890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0859,
+ "step": 1073
+ },
+ {
+ "epoch": 3.4313099041533546,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0842,
+ "step": 1074
+ },
+ {
+ "epoch": 3.4345047923322682,
+ "grad_norm": 0.1142578125,
+ "learning_rate": 0.0005,
+ "loss": 1.1117,
+ "step": 1075
+ },
+ {
+ "epoch": 3.437699680511182,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0999,
+ "step": 1076
+ },
+ {
+ "epoch": 3.440894568690096,
+ "grad_norm": 0.095703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0924,
+ "step": 1077
+ },
+ {
+ "epoch": 3.4440894568690097,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.1006,
+ "step": 1078
+ },
+ {
+ "epoch": 3.4472843450479234,
+ "grad_norm": 0.10107421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0783,
+ "step": 1079
+ },
+ {
+ "epoch": 3.450479233226837,
+ "grad_norm": 0.13671875,
+ "learning_rate": 0.0005,
+ "loss": 1.1062,
+ "step": 1080
+ },
+ {
+ "epoch": 3.4536741214057507,
+ "grad_norm": 0.1767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0941,
+ "step": 1081
+ },
+ {
+ "epoch": 3.4568690095846644,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0975,
+ "step": 1082
+ },
+ {
+ "epoch": 3.460063897763578,
+ "grad_norm": 0.111328125,
+ "learning_rate": 0.0005,
+ "loss": 1.1022,
+ "step": 1083
+ },
+ {
+ "epoch": 3.463258785942492,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.089,
+ "step": 1084
+ },
+ {
+ "epoch": 3.466453674121406,
+ "grad_norm": 0.09033203125,
+ "learning_rate": 0.0005,
+ "loss": 1.088,
+ "step": 1085
+ },
+ {
+ "epoch": 3.4696485623003195,
+ "grad_norm": 0.08154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0936,
+ "step": 1086
+ },
+ {
+ "epoch": 3.472843450479233,
+ "grad_norm": 0.091796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0952,
+ "step": 1087
+ },
+ {
+ "epoch": 3.476038338658147,
+ "grad_norm": 0.0859375,
+ "learning_rate": 0.0005,
+ "loss": 1.1019,
+ "step": 1088
+ },
+ {
+ "epoch": 3.479233226837061,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0959,
+ "step": 1089
+ },
+ {
+ "epoch": 3.4824281150159746,
+ "grad_norm": 0.09033203125,
+ "learning_rate": 0.0005,
+ "loss": 1.1036,
+ "step": 1090
+ },
+ {
+ "epoch": 3.4856230031948883,
+ "grad_norm": 0.11181640625,
+ "learning_rate": 0.0005,
+ "loss": 1.096,
+ "step": 1091
+ },
+ {
+ "epoch": 3.488817891373802,
+ "grad_norm": 0.10009765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0851,
+ "step": 1092
+ },
+ {
+ "epoch": 3.4920127795527156,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0878,
+ "step": 1093
+ },
+ {
+ "epoch": 3.4952076677316293,
+ "grad_norm": 0.0888671875,
+ "learning_rate": 0.0005,
+ "loss": 1.1152,
+ "step": 1094
+ },
+ {
+ "epoch": 3.498402555910543,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0921,
+ "step": 1095
+ },
+ {
+ "epoch": 3.501597444089457,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.1226,
+ "step": 1096
+ },
+ {
+ "epoch": 3.5047923322683707,
+ "grad_norm": 0.095703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0895,
+ "step": 1097
+ },
+ {
+ "epoch": 3.5079872204472844,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0905,
+ "step": 1098
+ },
+ {
+ "epoch": 3.511182108626198,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0622,
+ "step": 1099
+ },
+ {
+ "epoch": 3.5143769968051117,
+ "grad_norm": 0.1103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0842,
+ "step": 1100
+ },
+ {
+ "epoch": 3.5175718849840254,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0992,
+ "step": 1101
+ },
+ {
+ "epoch": 3.520766773162939,
+ "grad_norm": 0.0908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0919,
+ "step": 1102
+ },
+ {
+ "epoch": 3.523961661341853,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.1071,
+ "step": 1103
+ },
+ {
+ "epoch": 3.527156549520767,
+ "grad_norm": 0.1064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0731,
+ "step": 1104
+ },
+ {
+ "epoch": 3.5303514376996805,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0861,
+ "step": 1105
+ },
+ {
+ "epoch": 3.533546325878594,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0905,
+ "step": 1106
+ },
+ {
+ "epoch": 3.536741214057508,
+ "grad_norm": 0.08642578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0888,
+ "step": 1107
+ },
+ {
+ "epoch": 3.539936102236422,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0886,
+ "step": 1108
+ },
+ {
+ "epoch": 3.543130990415335,
+ "grad_norm": 0.0966796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0919,
+ "step": 1109
+ },
+ {
+ "epoch": 3.5463258785942493,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0819,
+ "step": 1110
+ },
+ {
+ "epoch": 3.549520766773163,
+ "grad_norm": 0.11767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0752,
+ "step": 1111
+ },
+ {
+ "epoch": 3.5527156549520766,
+ "grad_norm": 0.08837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0952,
+ "step": 1112
+ },
+ {
+ "epoch": 3.5559105431309903,
+ "grad_norm": 0.119140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0946,
+ "step": 1113
+ },
+ {
+ "epoch": 3.559105431309904,
+ "grad_norm": 0.11962890625,
+ "learning_rate": 0.0005,
+ "loss": 1.1049,
+ "step": 1114
+ },
+ {
+ "epoch": 3.562300319488818,
+ "grad_norm": 0.2490234375,
+ "learning_rate": 0.0005,
+ "loss": 1.1064,
+ "step": 1115
+ },
+ {
+ "epoch": 3.5654952076677318,
+ "grad_norm": 0.1572265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0848,
+ "step": 1116
+ },
+ {
+ "epoch": 3.5686900958466454,
+ "grad_norm": 0.11669921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0876,
+ "step": 1117
+ },
+ {
+ "epoch": 3.571884984025559,
+ "grad_norm": 0.1455078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0959,
+ "step": 1118
+ },
+ {
+ "epoch": 3.5750798722044728,
+ "grad_norm": 0.203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0942,
+ "step": 1119
+ },
+ {
+ "epoch": 3.5782747603833864,
+ "grad_norm": 0.1806640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0909,
+ "step": 1120
+ },
+ {
+ "epoch": 3.5814696485623,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0942,
+ "step": 1121
+ },
+ {
+ "epoch": 3.584664536741214,
+ "grad_norm": 0.2236328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0987,
+ "step": 1122
+ },
+ {
+ "epoch": 3.587859424920128,
+ "grad_norm": 0.427734375,
+ "learning_rate": 0.0005,
+ "loss": 1.099,
+ "step": 1123
+ },
+ {
+ "epoch": 3.5910543130990416,
+ "grad_norm": 0.609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0918,
+ "step": 1124
+ },
+ {
+ "epoch": 3.594249201277955,
+ "grad_norm": 1.0859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0946,
+ "step": 1125
+ },
+ {
+ "epoch": 3.597444089456869,
+ "grad_norm": 1.5546875,
+ "learning_rate": 0.0005,
+ "loss": 1.1187,
+ "step": 1126
+ },
+ {
+ "epoch": 3.600638977635783,
+ "grad_norm": 0.345703125,
+ "learning_rate": 0.0005,
+ "loss": 1.1087,
+ "step": 1127
+ },
+ {
+ "epoch": 3.6038338658146962,
+ "grad_norm": 2.375,
+ "learning_rate": 0.0005,
+ "loss": 1.1524,
+ "step": 1128
+ },
+ {
+ "epoch": 3.6070287539936103,
+ "grad_norm": 0.94140625,
+ "learning_rate": 0.0005,
+ "loss": 1.1178,
+ "step": 1129
+ },
+ {
+ "epoch": 3.610223642172524,
+ "grad_norm": 1.5703125,
+ "learning_rate": 0.0005,
+ "loss": 1.1704,
+ "step": 1130
+ },
+ {
+ "epoch": 3.6134185303514377,
+ "grad_norm": 1.03125,
+ "learning_rate": 0.0005,
+ "loss": 1.1532,
+ "step": 1131
+ },
+ {
+ "epoch": 3.6166134185303513,
+ "grad_norm": 1.25,
+ "learning_rate": 0.0005,
+ "loss": 1.1493,
+ "step": 1132
+ },
+ {
+ "epoch": 3.619808306709265,
+ "grad_norm": 1.390625,
+ "learning_rate": 0.0005,
+ "loss": 1.149,
+ "step": 1133
+ },
+ {
+ "epoch": 3.623003194888179,
+ "grad_norm": 0.96484375,
+ "learning_rate": 0.0005,
+ "loss": 1.1278,
+ "step": 1134
+ },
+ {
+ "epoch": 3.626198083067093,
+ "grad_norm": 1.46875,
+ "learning_rate": 0.0005,
+ "loss": 1.1392,
+ "step": 1135
+ },
+ {
+ "epoch": 3.6293929712460065,
+ "grad_norm": 0.48828125,
+ "learning_rate": 0.0005,
+ "loss": 1.1379,
+ "step": 1136
+ },
+ {
+ "epoch": 3.63258785942492,
+ "grad_norm": 0.482421875,
+ "learning_rate": 0.0005,
+ "loss": 1.1125,
+ "step": 1137
+ },
+ {
+ "epoch": 3.635782747603834,
+ "grad_norm": 0.6015625,
+ "learning_rate": 0.0005,
+ "loss": 1.1258,
+ "step": 1138
+ },
+ {
+ "epoch": 3.6389776357827475,
+ "grad_norm": 0.25390625,
+ "learning_rate": 0.0005,
+ "loss": 1.1064,
+ "step": 1139
+ },
+ {
+ "epoch": 3.642172523961661,
+ "grad_norm": 0.365234375,
+ "learning_rate": 0.0005,
+ "loss": 1.1162,
+ "step": 1140
+ },
+ {
+ "epoch": 3.6453674121405752,
+ "grad_norm": 0.427734375,
+ "learning_rate": 0.0005,
+ "loss": 1.1229,
+ "step": 1141
+ },
+ {
+ "epoch": 3.648562300319489,
+ "grad_norm": 0.1796875,
+ "learning_rate": 0.0005,
+ "loss": 1.1053,
+ "step": 1142
+ },
+ {
+ "epoch": 3.6517571884984026,
+ "grad_norm": 0.37890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0912,
+ "step": 1143
+ },
+ {
+ "epoch": 3.6549520766773163,
+ "grad_norm": 0.1611328125,
+ "learning_rate": 0.0005,
+ "loss": 1.108,
+ "step": 1144
+ },
+ {
+ "epoch": 3.65814696485623,
+ "grad_norm": 0.36328125,
+ "learning_rate": 0.0005,
+ "loss": 1.1046,
+ "step": 1145
+ },
+ {
+ "epoch": 3.661341853035144,
+ "grad_norm": 0.498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.1064,
+ "step": 1146
+ },
+ {
+ "epoch": 3.6645367412140573,
+ "grad_norm": 0.201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.11,
+ "step": 1147
+ },
+ {
+ "epoch": 3.6677316293929714,
+ "grad_norm": 0.291015625,
+ "learning_rate": 0.0005,
+ "loss": 1.107,
+ "step": 1148
+ },
+ {
+ "epoch": 3.670926517571885,
+ "grad_norm": 0.26953125,
+ "learning_rate": 0.0005,
+ "loss": 1.103,
+ "step": 1149
+ },
+ {
+ "epoch": 3.6741214057507987,
+ "grad_norm": 0.171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0955,
+ "step": 1150
+ },
+ {
+ "epoch": 3.6773162939297124,
+ "grad_norm": 0.27734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0845,
+ "step": 1151
+ },
+ {
+ "epoch": 3.680511182108626,
+ "grad_norm": 0.138671875,
+ "learning_rate": 0.0005,
+ "loss": 1.1003,
+ "step": 1152
+ },
+ {
+ "epoch": 3.68370607028754,
+ "grad_norm": 0.416015625,
+ "learning_rate": 0.0005,
+ "loss": 1.101,
+ "step": 1153
+ },
+ {
+ "epoch": 3.686900958466454,
+ "grad_norm": 0.27734375,
+ "learning_rate": 0.0005,
+ "loss": 1.112,
+ "step": 1154
+ },
+ {
+ "epoch": 3.6900958466453675,
+ "grad_norm": 0.185546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0973,
+ "step": 1155
+ },
+ {
+ "epoch": 3.693290734824281,
+ "grad_norm": 0.345703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0981,
+ "step": 1156
+ },
+ {
+ "epoch": 3.696485623003195,
+ "grad_norm": 0.10791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0964,
+ "step": 1157
+ },
+ {
+ "epoch": 3.6996805111821085,
+ "grad_norm": 0.373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0961,
+ "step": 1158
+ },
+ {
+ "epoch": 3.702875399361022,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.1143,
+ "step": 1159
+ },
+ {
+ "epoch": 3.7060702875399363,
+ "grad_norm": 0.31640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0948,
+ "step": 1160
+ },
+ {
+ "epoch": 3.70926517571885,
+ "grad_norm": 0.14453125,
+ "learning_rate": 0.0005,
+ "loss": 1.089,
+ "step": 1161
+ },
+ {
+ "epoch": 3.7124600638977636,
+ "grad_norm": 0.1845703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0966,
+ "step": 1162
+ },
+ {
+ "epoch": 3.7156549520766773,
+ "grad_norm": 0.1201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0903,
+ "step": 1163
+ },
+ {
+ "epoch": 3.718849840255591,
+ "grad_norm": 0.162109375,
+ "learning_rate": 0.0005,
+ "loss": 1.1036,
+ "step": 1164
+ },
+ {
+ "epoch": 3.722044728434505,
+ "grad_norm": 0.1162109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0784,
+ "step": 1165
+ },
+ {
+ "epoch": 3.7252396166134183,
+ "grad_norm": 0.1376953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0866,
+ "step": 1166
+ },
+ {
+ "epoch": 3.7284345047923324,
+ "grad_norm": 0.09375,
+ "learning_rate": 0.0005,
+ "loss": 1.0934,
+ "step": 1167
+ },
+ {
+ "epoch": 3.731629392971246,
+ "grad_norm": 0.11767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0878,
+ "step": 1168
+ },
+ {
+ "epoch": 3.7348242811501597,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0978,
+ "step": 1169
+ },
+ {
+ "epoch": 3.7380191693290734,
+ "grad_norm": 0.111328125,
+ "learning_rate": 0.0005,
+ "loss": 1.097,
+ "step": 1170
+ },
+ {
+ "epoch": 3.741214057507987,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0969,
+ "step": 1171
+ },
+ {
+ "epoch": 3.744408945686901,
+ "grad_norm": 0.08447265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0908,
+ "step": 1172
+ },
+ {
+ "epoch": 3.747603833865815,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0803,
+ "step": 1173
+ },
+ {
+ "epoch": 3.7507987220447285,
+ "grad_norm": 0.09375,
+ "learning_rate": 0.0005,
+ "loss": 1.0884,
+ "step": 1174
+ },
+ {
+ "epoch": 3.753993610223642,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.078,
+ "step": 1175
+ },
+ {
+ "epoch": 3.757188498402556,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0809,
+ "step": 1176
+ },
+ {
+ "epoch": 3.7603833865814695,
+ "grad_norm": 0.07861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0798,
+ "step": 1177
+ },
+ {
+ "epoch": 3.763578274760383,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0986,
+ "step": 1178
+ },
+ {
+ "epoch": 3.7667731629392973,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.1035,
+ "step": 1179
+ },
+ {
+ "epoch": 3.769968051118211,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.103,
+ "step": 1180
+ },
+ {
+ "epoch": 3.7731629392971247,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0927,
+ "step": 1181
+ },
+ {
+ "epoch": 3.7763578274760383,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0875,
+ "step": 1182
+ },
+ {
+ "epoch": 3.779552715654952,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.1001,
+ "step": 1183
+ },
+ {
+ "epoch": 3.7827476038338657,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0963,
+ "step": 1184
+ },
+ {
+ "epoch": 3.7859424920127793,
+ "grad_norm": 0.058349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.1002,
+ "step": 1185
+ },
+ {
+ "epoch": 3.7891373801916934,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0827,
+ "step": 1186
+ },
+ {
+ "epoch": 3.792332268370607,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0993,
+ "step": 1187
+ },
+ {
+ "epoch": 3.7955271565495208,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0742,
+ "step": 1188
+ },
+ {
+ "epoch": 3.7987220447284344,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0919,
+ "step": 1189
+ },
+ {
+ "epoch": 3.801916932907348,
+ "grad_norm": 0.11083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0963,
+ "step": 1190
+ },
+ {
+ "epoch": 3.8051118210862622,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.1043,
+ "step": 1191
+ },
+ {
+ "epoch": 3.8083067092651754,
+ "grad_norm": 0.1337890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0765,
+ "step": 1192
+ },
+ {
+ "epoch": 3.8115015974440896,
+ "grad_norm": 0.08642578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0922,
+ "step": 1193
+ },
+ {
+ "epoch": 3.8146964856230032,
+ "grad_norm": 0.12890625,
+ "learning_rate": 0.0005,
+ "loss": 1.091,
+ "step": 1194
+ },
+ {
+ "epoch": 3.817891373801917,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0865,
+ "step": 1195
+ },
+ {
+ "epoch": 3.8210862619808306,
+ "grad_norm": 0.1298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.1056,
+ "step": 1196
+ },
+ {
+ "epoch": 3.8242811501597442,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0924,
+ "step": 1197
+ },
+ {
+ "epoch": 3.8274760383386583,
+ "grad_norm": 0.107421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0776,
+ "step": 1198
+ },
+ {
+ "epoch": 3.830670926517572,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0862,
+ "step": 1199
+ },
+ {
+ "epoch": 3.8338658146964857,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0743,
+ "step": 1200
+ },
+ {
+ "epoch": 3.8370607028753994,
+ "grad_norm": 0.130859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0978,
+ "step": 1201
+ },
+ {
+ "epoch": 3.840255591054313,
+ "grad_norm": 0.09619140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0888,
+ "step": 1202
+ },
+ {
+ "epoch": 3.8434504792332267,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0954,
+ "step": 1203
+ },
+ {
+ "epoch": 3.8466453674121404,
+ "grad_norm": 0.1044921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0937,
+ "step": 1204
+ },
+ {
+ "epoch": 3.8498402555910545,
+ "grad_norm": 0.0830078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0663,
+ "step": 1205
+ },
+ {
+ "epoch": 3.853035143769968,
+ "grad_norm": 0.11767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0895,
+ "step": 1206
+ },
+ {
+ "epoch": 3.856230031948882,
+ "grad_norm": 0.1123046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0832,
+ "step": 1207
+ },
+ {
+ "epoch": 3.8594249201277955,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0909,
+ "step": 1208
+ },
+ {
+ "epoch": 3.862619808306709,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0857,
+ "step": 1209
+ },
+ {
+ "epoch": 3.8658146964856233,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.1005,
+ "step": 1210
+ },
+ {
+ "epoch": 3.8690095846645365,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0949,
+ "step": 1211
+ },
+ {
+ "epoch": 3.8722044728434506,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0881,
+ "step": 1212
+ },
+ {
+ "epoch": 3.8753993610223643,
+ "grad_norm": 0.09130859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0879,
+ "step": 1213
+ },
+ {
+ "epoch": 3.878594249201278,
+ "grad_norm": 0.058349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0928,
+ "step": 1214
+ },
+ {
+ "epoch": 3.8817891373801916,
+ "grad_norm": 0.0849609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0886,
+ "step": 1215
+ },
+ {
+ "epoch": 3.8849840255591053,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0945,
+ "step": 1216
+ },
+ {
+ "epoch": 3.8881789137380194,
+ "grad_norm": 0.09619140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0906,
+ "step": 1217
+ },
+ {
+ "epoch": 3.891373801916933,
+ "grad_norm": 0.1083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0868,
+ "step": 1218
+ },
+ {
+ "epoch": 3.8945686900958467,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0953,
+ "step": 1219
+ },
+ {
+ "epoch": 3.8977635782747604,
+ "grad_norm": 0.1142578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0878,
+ "step": 1220
+ },
+ {
+ "epoch": 3.900958466453674,
+ "grad_norm": 0.0595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0872,
+ "step": 1221
+ },
+ {
+ "epoch": 3.9041533546325877,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0858,
+ "step": 1222
+ },
+ {
+ "epoch": 3.9073482428115014,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0824,
+ "step": 1223
+ },
+ {
+ "epoch": 3.9105431309904155,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0745,
+ "step": 1224
+ },
+ {
+ "epoch": 3.913738019169329,
+ "grad_norm": 0.0810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0852,
+ "step": 1225
+ },
+ {
+ "epoch": 3.916932907348243,
+ "grad_norm": 0.09619140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0864,
+ "step": 1226
+ },
+ {
+ "epoch": 3.9201277955271565,
+ "grad_norm": 0.0810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0926,
+ "step": 1227
+ },
+ {
+ "epoch": 3.92332268370607,
+ "grad_norm": 0.09130859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0807,
+ "step": 1228
+ },
+ {
+ "epoch": 3.9265175718849843,
+ "grad_norm": 0.080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0882,
+ "step": 1229
+ },
+ {
+ "epoch": 3.9297124600638975,
+ "grad_norm": 0.083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0911,
+ "step": 1230
+ },
+ {
+ "epoch": 3.9329073482428116,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0917,
+ "step": 1231
+ },
+ {
+ "epoch": 3.9361022364217253,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0793,
+ "step": 1232
+ },
+ {
+ "epoch": 3.939297124600639,
+ "grad_norm": 0.08154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.1017,
+ "step": 1233
+ },
+ {
+ "epoch": 3.9424920127795526,
+ "grad_norm": 0.06103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0775,
+ "step": 1234
+ },
+ {
+ "epoch": 3.9456869009584663,
+ "grad_norm": 0.06201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0904,
+ "step": 1235
+ },
+ {
+ "epoch": 3.9488817891373804,
+ "grad_norm": 0.060791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0777,
+ "step": 1236
+ },
+ {
+ "epoch": 3.952076677316294,
+ "grad_norm": 0.061279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.098,
+ "step": 1237
+ },
+ {
+ "epoch": 3.9552715654952078,
+ "grad_norm": 0.060791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0779,
+ "step": 1238
+ },
+ {
+ "epoch": 3.9584664536741214,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0903,
+ "step": 1239
+ },
+ {
+ "epoch": 3.961661341853035,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.1053,
+ "step": 1240
+ },
+ {
+ "epoch": 3.9648562300319488,
+ "grad_norm": 0.059814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0825,
+ "step": 1241
+ },
+ {
+ "epoch": 3.9680511182108624,
+ "grad_norm": 0.057373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0918,
+ "step": 1242
+ },
+ {
+ "epoch": 3.9712460063897765,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0922,
+ "step": 1243
+ },
+ {
+ "epoch": 3.97444089456869,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0747,
+ "step": 1244
+ },
+ {
+ "epoch": 3.977635782747604,
+ "grad_norm": 0.0849609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0952,
+ "step": 1245
+ },
+ {
+ "epoch": 3.9808306709265175,
+ "grad_norm": 0.056396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0877,
+ "step": 1246
+ },
+ {
+ "epoch": 3.984025559105431,
+ "grad_norm": 0.08056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0786,
+ "step": 1247
+ },
+ {
+ "epoch": 3.987220447284345,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0967,
+ "step": 1248
+ },
+ {
+ "epoch": 3.9904153354632586,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0849,
+ "step": 1249
+ },
+ {
+ "epoch": 3.9936102236421727,
+ "grad_norm": 0.0615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0847,
+ "step": 1250
+ },
+ {
+ "epoch": 3.9968051118210863,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0949,
+ "step": 1251
+ },
+ {
+ "epoch": 4.0,
+ "grad_norm": 0.09619140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0956,
+ "step": 1252
+ },
+ {
+ "epoch": 4.003194888178914,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.106,
+ "step": 1253
+ },
+ {
+ "epoch": 4.006389776357827,
+ "grad_norm": 0.10400390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0781,
+ "step": 1254
+ },
+ {
+ "epoch": 4.0095846645367414,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0783,
+ "step": 1255
+ },
+ {
+ "epoch": 4.012779552715655,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0922,
+ "step": 1256
+ },
+ {
+ "epoch": 4.015974440894569,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0874,
+ "step": 1257
+ },
+ {
+ "epoch": 4.019169329073482,
+ "grad_norm": 0.1171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0783,
+ "step": 1258
+ },
+ {
+ "epoch": 4.022364217252396,
+ "grad_norm": 0.0830078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0811,
+ "step": 1259
+ },
+ {
+ "epoch": 4.02555910543131,
+ "grad_norm": 0.11376953125,
+ "learning_rate": 0.0005,
+ "loss": 1.089,
+ "step": 1260
+ },
+ {
+ "epoch": 4.0287539936102235,
+ "grad_norm": 0.119140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0875,
+ "step": 1261
+ },
+ {
+ "epoch": 4.031948881789138,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0924,
+ "step": 1262
+ },
+ {
+ "epoch": 4.035143769968051,
+ "grad_norm": 0.11376953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0991,
+ "step": 1263
+ },
+ {
+ "epoch": 4.038338658146965,
+ "grad_norm": 0.09716796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0876,
+ "step": 1264
+ },
+ {
+ "epoch": 4.041533546325879,
+ "grad_norm": 0.056396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.1018,
+ "step": 1265
+ },
+ {
+ "epoch": 4.044728434504792,
+ "grad_norm": 0.08984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0803,
+ "step": 1266
+ },
+ {
+ "epoch": 4.047923322683706,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.083,
+ "step": 1267
+ },
+ {
+ "epoch": 4.05111821086262,
+ "grad_norm": 0.08056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0852,
+ "step": 1268
+ },
+ {
+ "epoch": 4.054313099041534,
+ "grad_norm": 0.0830078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0946,
+ "step": 1269
+ },
+ {
+ "epoch": 4.057507987220447,
+ "grad_norm": 0.1083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0825,
+ "step": 1270
+ },
+ {
+ "epoch": 4.060702875399361,
+ "grad_norm": 0.08642578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0773,
+ "step": 1271
+ },
+ {
+ "epoch": 4.063897763578275,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0831,
+ "step": 1272
+ },
+ {
+ "epoch": 4.067092651757188,
+ "grad_norm": 0.11572265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0993,
+ "step": 1273
+ },
+ {
+ "epoch": 4.0702875399361025,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0949,
+ "step": 1274
+ },
+ {
+ "epoch": 4.073482428115016,
+ "grad_norm": 0.1220703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0877,
+ "step": 1275
+ },
+ {
+ "epoch": 4.07667731629393,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0837,
+ "step": 1276
+ },
+ {
+ "epoch": 4.079872204472843,
+ "grad_norm": 0.1435546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0796,
+ "step": 1277
+ },
+ {
+ "epoch": 4.083067092651757,
+ "grad_norm": 0.095703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0887,
+ "step": 1278
+ },
+ {
+ "epoch": 4.086261980830671,
+ "grad_norm": 0.0966796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0964,
+ "step": 1279
+ },
+ {
+ "epoch": 4.0894568690095845,
+ "grad_norm": 0.1025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0959,
+ "step": 1280
+ },
+ {
+ "epoch": 4.092651757188499,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0802,
+ "step": 1281
+ },
+ {
+ "epoch": 4.095846645367412,
+ "grad_norm": 0.08056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0868,
+ "step": 1282
+ },
+ {
+ "epoch": 4.099041533546326,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0798,
+ "step": 1283
+ },
+ {
+ "epoch": 4.102236421725239,
+ "grad_norm": 0.09130859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0988,
+ "step": 1284
+ },
+ {
+ "epoch": 4.105431309904153,
+ "grad_norm": 0.1640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0877,
+ "step": 1285
+ },
+ {
+ "epoch": 4.108626198083067,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0849,
+ "step": 1286
+ },
+ {
+ "epoch": 4.111821086261981,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0835,
+ "step": 1287
+ },
+ {
+ "epoch": 4.115015974440895,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0778,
+ "step": 1288
+ },
+ {
+ "epoch": 4.118210862619808,
+ "grad_norm": 0.08447265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0616,
+ "step": 1289
+ },
+ {
+ "epoch": 4.121405750798722,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.1029,
+ "step": 1290
+ },
+ {
+ "epoch": 4.124600638977636,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0856,
+ "step": 1291
+ },
+ {
+ "epoch": 4.127795527156549,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0894,
+ "step": 1292
+ },
+ {
+ "epoch": 4.1309904153354635,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0952,
+ "step": 1293
+ },
+ {
+ "epoch": 4.134185303514377,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0842,
+ "step": 1294
+ },
+ {
+ "epoch": 4.137380191693291,
+ "grad_norm": 0.08349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0889,
+ "step": 1295
+ },
+ {
+ "epoch": 4.140575079872204,
+ "grad_norm": 0.1357421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0857,
+ "step": 1296
+ },
+ {
+ "epoch": 4.143769968051118,
+ "grad_norm": 0.1005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0889,
+ "step": 1297
+ },
+ {
+ "epoch": 4.146964856230032,
+ "grad_norm": 0.08447265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0853,
+ "step": 1298
+ },
+ {
+ "epoch": 4.1501597444089455,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0912,
+ "step": 1299
+ },
+ {
+ "epoch": 4.15335463258786,
+ "grad_norm": 0.10107421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0843,
+ "step": 1300
+ },
+ {
+ "epoch": 4.156549520766773,
+ "grad_norm": 0.08349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0768,
+ "step": 1301
+ },
+ {
+ "epoch": 4.159744408945687,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0908,
+ "step": 1302
+ },
+ {
+ "epoch": 4.1629392971246,
+ "grad_norm": 0.107421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0841,
+ "step": 1303
+ },
+ {
+ "epoch": 4.166134185303514,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0833,
+ "step": 1304
+ },
+ {
+ "epoch": 4.169329073482428,
+ "grad_norm": 0.10205078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0839,
+ "step": 1305
+ },
+ {
+ "epoch": 4.172523961661342,
+ "grad_norm": 0.12255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0901,
+ "step": 1306
+ },
+ {
+ "epoch": 4.175718849840256,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0857,
+ "step": 1307
+ },
+ {
+ "epoch": 4.178913738019169,
+ "grad_norm": 0.0966796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0784,
+ "step": 1308
+ },
+ {
+ "epoch": 4.182108626198083,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0874,
+ "step": 1309
+ },
+ {
+ "epoch": 4.185303514376997,
+ "grad_norm": 0.125,
+ "learning_rate": 0.0005,
+ "loss": 1.0909,
+ "step": 1310
+ },
+ {
+ "epoch": 4.18849840255591,
+ "grad_norm": 0.1103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0724,
+ "step": 1311
+ },
+ {
+ "epoch": 4.1916932907348246,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0806,
+ "step": 1312
+ },
+ {
+ "epoch": 4.194888178913738,
+ "grad_norm": 0.09765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0739,
+ "step": 1313
+ },
+ {
+ "epoch": 4.198083067092652,
+ "grad_norm": 0.08837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0926,
+ "step": 1314
+ },
+ {
+ "epoch": 4.201277955271565,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0841,
+ "step": 1315
+ },
+ {
+ "epoch": 4.204472843450479,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0814,
+ "step": 1316
+ },
+ {
+ "epoch": 4.207667731629393,
+ "grad_norm": 0.056884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.078,
+ "step": 1317
+ },
+ {
+ "epoch": 4.210862619808307,
+ "grad_norm": 0.11083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0843,
+ "step": 1318
+ },
+ {
+ "epoch": 4.214057507987221,
+ "grad_norm": 0.09033203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0799,
+ "step": 1319
+ },
+ {
+ "epoch": 4.217252396166134,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0721,
+ "step": 1320
+ },
+ {
+ "epoch": 4.220447284345048,
+ "grad_norm": 0.060791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0789,
+ "step": 1321
+ },
+ {
+ "epoch": 4.223642172523961,
+ "grad_norm": 0.0908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0955,
+ "step": 1322
+ },
+ {
+ "epoch": 4.226837060702875,
+ "grad_norm": 0.08154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0892,
+ "step": 1323
+ },
+ {
+ "epoch": 4.2300319488817895,
+ "grad_norm": 0.061767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0807,
+ "step": 1324
+ },
+ {
+ "epoch": 4.233226837060703,
+ "grad_norm": 0.09228515625,
+ "learning_rate": 0.0005,
+ "loss": 1.079,
+ "step": 1325
+ },
+ {
+ "epoch": 4.236421725239617,
+ "grad_norm": 0.1123046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0925,
+ "step": 1326
+ },
+ {
+ "epoch": 4.23961661341853,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0841,
+ "step": 1327
+ },
+ {
+ "epoch": 4.242811501597444,
+ "grad_norm": 0.1123046875,
+ "learning_rate": 0.0005,
+ "loss": 1.1085,
+ "step": 1328
+ },
+ {
+ "epoch": 4.246006389776358,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0863,
+ "step": 1329
+ },
+ {
+ "epoch": 4.2492012779552715,
+ "grad_norm": 0.058349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0892,
+ "step": 1330
+ },
+ {
+ "epoch": 4.252396166134186,
+ "grad_norm": 0.09033203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0727,
+ "step": 1331
+ },
+ {
+ "epoch": 4.255591054313099,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0866,
+ "step": 1332
+ },
+ {
+ "epoch": 4.258785942492013,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0947,
+ "step": 1333
+ },
+ {
+ "epoch": 4.261980830670926,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0902,
+ "step": 1334
+ },
+ {
+ "epoch": 4.26517571884984,
+ "grad_norm": 0.062255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0902,
+ "step": 1335
+ },
+ {
+ "epoch": 4.268370607028754,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0862,
+ "step": 1336
+ },
+ {
+ "epoch": 4.271565495207668,
+ "grad_norm": 0.06201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0858,
+ "step": 1337
+ },
+ {
+ "epoch": 4.274760383386582,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0674,
+ "step": 1338
+ },
+ {
+ "epoch": 4.277955271565495,
+ "grad_norm": 0.08203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0841,
+ "step": 1339
+ },
+ {
+ "epoch": 4.281150159744409,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0694,
+ "step": 1340
+ },
+ {
+ "epoch": 4.284345047923322,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0823,
+ "step": 1341
+ },
+ {
+ "epoch": 4.287539936102236,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0727,
+ "step": 1342
+ },
+ {
+ "epoch": 4.2907348242811505,
+ "grad_norm": 0.123046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0769,
+ "step": 1343
+ },
+ {
+ "epoch": 4.293929712460064,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0757,
+ "step": 1344
+ },
+ {
+ "epoch": 4.297124600638978,
+ "grad_norm": 0.115234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0814,
+ "step": 1345
+ },
+ {
+ "epoch": 4.300319488817891,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0798,
+ "step": 1346
+ },
+ {
+ "epoch": 4.303514376996805,
+ "grad_norm": 0.0947265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0973,
+ "step": 1347
+ },
+ {
+ "epoch": 4.306709265175719,
+ "grad_norm": 0.0927734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0773,
+ "step": 1348
+ },
+ {
+ "epoch": 4.3099041533546325,
+ "grad_norm": 0.1103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0787,
+ "step": 1349
+ },
+ {
+ "epoch": 4.313099041533547,
+ "grad_norm": 0.1298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0902,
+ "step": 1350
+ },
+ {
+ "epoch": 4.31629392971246,
+ "grad_norm": 0.0810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0867,
+ "step": 1351
+ },
+ {
+ "epoch": 4.319488817891374,
+ "grad_norm": 0.201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0615,
+ "step": 1352
+ },
+ {
+ "epoch": 4.322683706070287,
+ "grad_norm": 0.14453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0855,
+ "step": 1353
+ },
+ {
+ "epoch": 4.325878594249201,
+ "grad_norm": 0.09033203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0905,
+ "step": 1354
+ },
+ {
+ "epoch": 4.329073482428115,
+ "grad_norm": 0.1181640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0809,
+ "step": 1355
+ },
+ {
+ "epoch": 4.332268370607029,
+ "grad_norm": 0.115234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0824,
+ "step": 1356
+ },
+ {
+ "epoch": 4.335463258785943,
+ "grad_norm": 0.08984375,
+ "learning_rate": 0.0005,
+ "loss": 1.087,
+ "step": 1357
+ },
+ {
+ "epoch": 4.338658146964856,
+ "grad_norm": 0.11181640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0819,
+ "step": 1358
+ },
+ {
+ "epoch": 4.34185303514377,
+ "grad_norm": 0.1376953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0774,
+ "step": 1359
+ },
+ {
+ "epoch": 4.345047923322683,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0935,
+ "step": 1360
+ },
+ {
+ "epoch": 4.348242811501597,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0754,
+ "step": 1361
+ },
+ {
+ "epoch": 4.3514376996805115,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0747,
+ "step": 1362
+ },
+ {
+ "epoch": 4.354632587859425,
+ "grad_norm": 0.1025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0992,
+ "step": 1363
+ },
+ {
+ "epoch": 4.357827476038339,
+ "grad_norm": 0.11181640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0764,
+ "step": 1364
+ },
+ {
+ "epoch": 4.361022364217252,
+ "grad_norm": 0.08203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0876,
+ "step": 1365
+ },
+ {
+ "epoch": 4.364217252396166,
+ "grad_norm": 0.142578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0815,
+ "step": 1366
+ },
+ {
+ "epoch": 4.36741214057508,
+ "grad_norm": 0.10009765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0718,
+ "step": 1367
+ },
+ {
+ "epoch": 4.3706070287539935,
+ "grad_norm": 0.08349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.09,
+ "step": 1368
+ },
+ {
+ "epoch": 4.373801916932908,
+ "grad_norm": 0.0947265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0777,
+ "step": 1369
+ },
+ {
+ "epoch": 4.376996805111821,
+ "grad_norm": 0.095703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0828,
+ "step": 1370
+ },
+ {
+ "epoch": 4.380191693290735,
+ "grad_norm": 0.09423828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0993,
+ "step": 1371
+ },
+ {
+ "epoch": 4.383386581469648,
+ "grad_norm": 0.095703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0847,
+ "step": 1372
+ },
+ {
+ "epoch": 4.386581469648562,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0923,
+ "step": 1373
+ },
+ {
+ "epoch": 4.389776357827476,
+ "grad_norm": 0.1005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0828,
+ "step": 1374
+ },
+ {
+ "epoch": 4.39297124600639,
+ "grad_norm": 0.10107421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0704,
+ "step": 1375
+ },
+ {
+ "epoch": 4.396166134185304,
+ "grad_norm": 0.10205078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0889,
+ "step": 1376
+ },
+ {
+ "epoch": 4.399361022364217,
+ "grad_norm": 0.1630859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0849,
+ "step": 1377
+ },
+ {
+ "epoch": 4.402555910543131,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0898,
+ "step": 1378
+ },
+ {
+ "epoch": 4.405750798722044,
+ "grad_norm": 0.10791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0925,
+ "step": 1379
+ },
+ {
+ "epoch": 4.4089456869009584,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0504,
+ "step": 1380
+ },
+ {
+ "epoch": 4.412140575079873,
+ "grad_norm": 0.11865234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0794,
+ "step": 1381
+ },
+ {
+ "epoch": 4.415335463258786,
+ "grad_norm": 0.0947265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0854,
+ "step": 1382
+ },
+ {
+ "epoch": 4.4185303514377,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0791,
+ "step": 1383
+ },
+ {
+ "epoch": 4.421725239616613,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0757,
+ "step": 1384
+ },
+ {
+ "epoch": 4.424920127795527,
+ "grad_norm": 0.0966796875,
+ "learning_rate": 0.0005,
+ "loss": 1.083,
+ "step": 1385
+ },
+ {
+ "epoch": 4.428115015974441,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0864,
+ "step": 1386
+ },
+ {
+ "epoch": 4.431309904153355,
+ "grad_norm": 0.060791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0926,
+ "step": 1387
+ },
+ {
+ "epoch": 4.434504792332269,
+ "grad_norm": 0.059326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0806,
+ "step": 1388
+ },
+ {
+ "epoch": 4.437699680511182,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0801,
+ "step": 1389
+ },
+ {
+ "epoch": 4.440894568690096,
+ "grad_norm": 0.080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0823,
+ "step": 1390
+ },
+ {
+ "epoch": 4.444089456869009,
+ "grad_norm": 0.1044921875,
+ "learning_rate": 0.0005,
+ "loss": 1.099,
+ "step": 1391
+ },
+ {
+ "epoch": 4.447284345047923,
+ "grad_norm": 0.2451171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0935,
+ "step": 1392
+ },
+ {
+ "epoch": 4.4504792332268375,
+ "grad_norm": 0.2333984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0855,
+ "step": 1393
+ },
+ {
+ "epoch": 4.453674121405751,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0871,
+ "step": 1394
+ },
+ {
+ "epoch": 4.456869009584665,
+ "grad_norm": 0.1044921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0916,
+ "step": 1395
+ },
+ {
+ "epoch": 4.460063897763578,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0766,
+ "step": 1396
+ },
+ {
+ "epoch": 4.463258785942492,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0788,
+ "step": 1397
+ },
+ {
+ "epoch": 4.466453674121405,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0806,
+ "step": 1398
+ },
+ {
+ "epoch": 4.4696485623003195,
+ "grad_norm": 0.07861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0828,
+ "step": 1399
+ },
+ {
+ "epoch": 4.472843450479234,
+ "grad_norm": 0.09130859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0752,
+ "step": 1400
+ },
+ {
+ "epoch": 4.476038338658147,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0933,
+ "step": 1401
+ },
+ {
+ "epoch": 4.479233226837061,
+ "grad_norm": 0.0908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0868,
+ "step": 1402
+ },
+ {
+ "epoch": 4.482428115015974,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0893,
+ "step": 1403
+ },
+ {
+ "epoch": 4.485623003194888,
+ "grad_norm": 0.087890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0704,
+ "step": 1404
+ },
+ {
+ "epoch": 4.488817891373802,
+ "grad_norm": 0.091796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0821,
+ "step": 1405
+ },
+ {
+ "epoch": 4.492012779552716,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0902,
+ "step": 1406
+ },
+ {
+ "epoch": 4.49520766773163,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0859,
+ "step": 1407
+ },
+ {
+ "epoch": 4.498402555910543,
+ "grad_norm": 0.08056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0867,
+ "step": 1408
+ },
+ {
+ "epoch": 4.501597444089457,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0835,
+ "step": 1409
+ },
+ {
+ "epoch": 4.50479233226837,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0859,
+ "step": 1410
+ },
+ {
+ "epoch": 4.507987220447284,
+ "grad_norm": 0.09375,
+ "learning_rate": 0.0005,
+ "loss": 1.0932,
+ "step": 1411
+ },
+ {
+ "epoch": 4.511182108626198,
+ "grad_norm": 0.08740234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0929,
+ "step": 1412
+ },
+ {
+ "epoch": 4.514376996805112,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0856,
+ "step": 1413
+ },
+ {
+ "epoch": 4.517571884984026,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0868,
+ "step": 1414
+ },
+ {
+ "epoch": 4.520766773162939,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0917,
+ "step": 1415
+ },
+ {
+ "epoch": 4.523961661341853,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0777,
+ "step": 1416
+ },
+ {
+ "epoch": 4.527156549520766,
+ "grad_norm": 0.0927734375,
+ "learning_rate": 0.0005,
+ "loss": 1.091,
+ "step": 1417
+ },
+ {
+ "epoch": 4.5303514376996805,
+ "grad_norm": 0.1279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.1019,
+ "step": 1418
+ },
+ {
+ "epoch": 4.533546325878595,
+ "grad_norm": 0.080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0882,
+ "step": 1419
+ },
+ {
+ "epoch": 4.536741214057508,
+ "grad_norm": 0.12890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0899,
+ "step": 1420
+ },
+ {
+ "epoch": 4.539936102236422,
+ "grad_norm": 0.08984375,
+ "learning_rate": 0.0005,
+ "loss": 1.089,
+ "step": 1421
+ },
+ {
+ "epoch": 4.543130990415335,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0754,
+ "step": 1422
+ },
+ {
+ "epoch": 4.546325878594249,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0694,
+ "step": 1423
+ },
+ {
+ "epoch": 4.549520766773163,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.1054,
+ "step": 1424
+ },
+ {
+ "epoch": 4.552715654952077,
+ "grad_norm": 0.111328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0674,
+ "step": 1425
+ },
+ {
+ "epoch": 4.555910543130991,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.083,
+ "step": 1426
+ },
+ {
+ "epoch": 4.559105431309904,
+ "grad_norm": 0.11083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0708,
+ "step": 1427
+ },
+ {
+ "epoch": 4.562300319488818,
+ "grad_norm": 0.1513671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0866,
+ "step": 1428
+ },
+ {
+ "epoch": 4.565495207667731,
+ "grad_norm": 0.1640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0847,
+ "step": 1429
+ },
+ {
+ "epoch": 4.568690095846645,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0736,
+ "step": 1430
+ },
+ {
+ "epoch": 4.571884984025559,
+ "grad_norm": 0.212890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0841,
+ "step": 1431
+ },
+ {
+ "epoch": 4.575079872204473,
+ "grad_norm": 0.146484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0881,
+ "step": 1432
+ },
+ {
+ "epoch": 4.578274760383387,
+ "grad_norm": 0.08349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0874,
+ "step": 1433
+ },
+ {
+ "epoch": 4.5814696485623,
+ "grad_norm": 0.1494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0893,
+ "step": 1434
+ },
+ {
+ "epoch": 4.584664536741214,
+ "grad_norm": 0.1494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0923,
+ "step": 1435
+ },
+ {
+ "epoch": 4.587859424920127,
+ "grad_norm": 0.06201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0948,
+ "step": 1436
+ },
+ {
+ "epoch": 4.5910543130990416,
+ "grad_norm": 0.12353515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0834,
+ "step": 1437
+ },
+ {
+ "epoch": 4.594249201277956,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.072,
+ "step": 1438
+ },
+ {
+ "epoch": 4.597444089456869,
+ "grad_norm": 0.119140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0829,
+ "step": 1439
+ },
+ {
+ "epoch": 4.600638977635783,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0876,
+ "step": 1440
+ },
+ {
+ "epoch": 4.603833865814696,
+ "grad_norm": 0.08154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0748,
+ "step": 1441
+ },
+ {
+ "epoch": 4.60702875399361,
+ "grad_norm": 0.08056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0716,
+ "step": 1442
+ },
+ {
+ "epoch": 4.6102236421725244,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0774,
+ "step": 1443
+ },
+ {
+ "epoch": 4.613418530351438,
+ "grad_norm": 0.09765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0788,
+ "step": 1444
+ },
+ {
+ "epoch": 4.616613418530352,
+ "grad_norm": 0.08935546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0745,
+ "step": 1445
+ },
+ {
+ "epoch": 4.619808306709265,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0869,
+ "step": 1446
+ },
+ {
+ "epoch": 4.623003194888179,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0887,
+ "step": 1447
+ },
+ {
+ "epoch": 4.626198083067092,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0847,
+ "step": 1448
+ },
+ {
+ "epoch": 4.6293929712460065,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0803,
+ "step": 1449
+ },
+ {
+ "epoch": 4.63258785942492,
+ "grad_norm": 0.08203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0863,
+ "step": 1450
+ },
+ {
+ "epoch": 4.635782747603834,
+ "grad_norm": 0.0576171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0927,
+ "step": 1451
+ },
+ {
+ "epoch": 4.638977635782748,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0597,
+ "step": 1452
+ },
+ {
+ "epoch": 4.642172523961661,
+ "grad_norm": 0.060302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0776,
+ "step": 1453
+ },
+ {
+ "epoch": 4.645367412140575,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0845,
+ "step": 1454
+ },
+ {
+ "epoch": 4.6485623003194885,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.069,
+ "step": 1455
+ },
+ {
+ "epoch": 4.651757188498403,
+ "grad_norm": 0.08935546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0834,
+ "step": 1456
+ },
+ {
+ "epoch": 4.654952076677317,
+ "grad_norm": 0.1201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0793,
+ "step": 1457
+ },
+ {
+ "epoch": 4.65814696485623,
+ "grad_norm": 0.0888671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0799,
+ "step": 1458
+ },
+ {
+ "epoch": 4.661341853035144,
+ "grad_norm": 0.09130859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0721,
+ "step": 1459
+ },
+ {
+ "epoch": 4.664536741214057,
+ "grad_norm": 0.08642578125,
+ "learning_rate": 0.0005,
+ "loss": 1.1068,
+ "step": 1460
+ },
+ {
+ "epoch": 4.667731629392971,
+ "grad_norm": 0.083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0843,
+ "step": 1461
+ },
+ {
+ "epoch": 4.6709265175718855,
+ "grad_norm": 0.083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0827,
+ "step": 1462
+ },
+ {
+ "epoch": 4.674121405750799,
+ "grad_norm": 0.08544921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0816,
+ "step": 1463
+ },
+ {
+ "epoch": 4.677316293929713,
+ "grad_norm": 0.08740234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0976,
+ "step": 1464
+ },
+ {
+ "epoch": 4.680511182108626,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0833,
+ "step": 1465
+ },
+ {
+ "epoch": 4.68370607028754,
+ "grad_norm": 0.1376953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0872,
+ "step": 1466
+ },
+ {
+ "epoch": 4.686900958466453,
+ "grad_norm": 0.1337890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0825,
+ "step": 1467
+ },
+ {
+ "epoch": 4.6900958466453675,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0914,
+ "step": 1468
+ },
+ {
+ "epoch": 4.693290734824281,
+ "grad_norm": 0.125,
+ "learning_rate": 0.0005,
+ "loss": 1.0884,
+ "step": 1469
+ },
+ {
+ "epoch": 4.696485623003195,
+ "grad_norm": 0.087890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0956,
+ "step": 1470
+ },
+ {
+ "epoch": 4.699680511182109,
+ "grad_norm": 0.12890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0858,
+ "step": 1471
+ },
+ {
+ "epoch": 4.702875399361022,
+ "grad_norm": 0.1826171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0882,
+ "step": 1472
+ },
+ {
+ "epoch": 4.706070287539936,
+ "grad_norm": 0.1513671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0809,
+ "step": 1473
+ },
+ {
+ "epoch": 4.7092651757188495,
+ "grad_norm": 0.08056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0837,
+ "step": 1474
+ },
+ {
+ "epoch": 4.712460063897764,
+ "grad_norm": 0.11962890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0912,
+ "step": 1475
+ },
+ {
+ "epoch": 4.715654952076678,
+ "grad_norm": 0.119140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0799,
+ "step": 1476
+ },
+ {
+ "epoch": 4.718849840255591,
+ "grad_norm": 0.10302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0742,
+ "step": 1477
+ },
+ {
+ "epoch": 4.722044728434505,
+ "grad_norm": 0.10302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0782,
+ "step": 1478
+ },
+ {
+ "epoch": 4.725239616613418,
+ "grad_norm": 0.060302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0918,
+ "step": 1479
+ },
+ {
+ "epoch": 4.728434504792332,
+ "grad_norm": 0.05419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0749,
+ "step": 1480
+ },
+ {
+ "epoch": 4.731629392971246,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0739,
+ "step": 1481
+ },
+ {
+ "epoch": 4.73482428115016,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0953,
+ "step": 1482
+ },
+ {
+ "epoch": 4.738019169329074,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0752,
+ "step": 1483
+ },
+ {
+ "epoch": 4.741214057507987,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0705,
+ "step": 1484
+ },
+ {
+ "epoch": 4.744408945686901,
+ "grad_norm": 0.091796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0876,
+ "step": 1485
+ },
+ {
+ "epoch": 4.747603833865814,
+ "grad_norm": 0.06201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0898,
+ "step": 1486
+ },
+ {
+ "epoch": 4.7507987220447285,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0865,
+ "step": 1487
+ },
+ {
+ "epoch": 4.753993610223642,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0805,
+ "step": 1488
+ },
+ {
+ "epoch": 4.757188498402556,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0853,
+ "step": 1489
+ },
+ {
+ "epoch": 4.76038338658147,
+ "grad_norm": 0.08544921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0906,
+ "step": 1490
+ },
+ {
+ "epoch": 4.763578274760383,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0897,
+ "step": 1491
+ },
+ {
+ "epoch": 4.766773162939297,
+ "grad_norm": 0.08203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0894,
+ "step": 1492
+ },
+ {
+ "epoch": 4.7699680511182105,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.092,
+ "step": 1493
+ },
+ {
+ "epoch": 4.773162939297125,
+ "grad_norm": 0.11279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0844,
+ "step": 1494
+ },
+ {
+ "epoch": 4.776357827476039,
+ "grad_norm": 0.058349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0813,
+ "step": 1495
+ },
+ {
+ "epoch": 4.779552715654952,
+ "grad_norm": 0.0830078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0815,
+ "step": 1496
+ },
+ {
+ "epoch": 4.782747603833866,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.078,
+ "step": 1497
+ },
+ {
+ "epoch": 4.785942492012779,
+ "grad_norm": 0.091796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0883,
+ "step": 1498
+ },
+ {
+ "epoch": 4.789137380191693,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0705,
+ "step": 1499
+ },
+ {
+ "epoch": 4.792332268370607,
+ "grad_norm": 0.1083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0751,
+ "step": 1500
+ },
+ {
+ "epoch": 4.795527156549521,
+ "grad_norm": 0.134765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0921,
+ "step": 1501
+ },
+ {
+ "epoch": 4.798722044728435,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0836,
+ "step": 1502
+ },
+ {
+ "epoch": 4.801916932907348,
+ "grad_norm": 0.10888671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0748,
+ "step": 1503
+ },
+ {
+ "epoch": 4.805111821086262,
+ "grad_norm": 0.08740234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0825,
+ "step": 1504
+ },
+ {
+ "epoch": 4.8083067092651754,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0804,
+ "step": 1505
+ },
+ {
+ "epoch": 4.81150159744409,
+ "grad_norm": 0.08935546875,
+ "learning_rate": 0.0005,
+ "loss": 1.081,
+ "step": 1506
+ },
+ {
+ "epoch": 4.814696485623003,
+ "grad_norm": 0.10888671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0846,
+ "step": 1507
+ },
+ {
+ "epoch": 4.817891373801917,
+ "grad_norm": 0.146484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0817,
+ "step": 1508
+ },
+ {
+ "epoch": 4.821086261980831,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0852,
+ "step": 1509
+ },
+ {
+ "epoch": 4.824281150159744,
+ "grad_norm": 0.1533203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0749,
+ "step": 1510
+ },
+ {
+ "epoch": 4.827476038338658,
+ "grad_norm": 0.08203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0757,
+ "step": 1511
+ },
+ {
+ "epoch": 4.830670926517572,
+ "grad_norm": 0.1005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0746,
+ "step": 1512
+ },
+ {
+ "epoch": 4.833865814696486,
+ "grad_norm": 0.10986328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0924,
+ "step": 1513
+ },
+ {
+ "epoch": 4.8370607028754,
+ "grad_norm": 0.10400390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0697,
+ "step": 1514
+ },
+ {
+ "epoch": 4.840255591054313,
+ "grad_norm": 0.14453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0666,
+ "step": 1515
+ },
+ {
+ "epoch": 4.843450479233227,
+ "grad_norm": 0.1455078125,
+ "learning_rate": 0.0005,
+ "loss": 1.078,
+ "step": 1516
+ },
+ {
+ "epoch": 4.84664536741214,
+ "grad_norm": 0.12109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0903,
+ "step": 1517
+ },
+ {
+ "epoch": 4.8498402555910545,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0801,
+ "step": 1518
+ },
+ {
+ "epoch": 4.853035143769968,
+ "grad_norm": 0.1025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0932,
+ "step": 1519
+ },
+ {
+ "epoch": 4.856230031948882,
+ "grad_norm": 0.09716796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0741,
+ "step": 1520
+ },
+ {
+ "epoch": 4.859424920127795,
+ "grad_norm": 0.08154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0789,
+ "step": 1521
+ },
+ {
+ "epoch": 4.862619808306709,
+ "grad_norm": 0.0888671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0796,
+ "step": 1522
+ },
+ {
+ "epoch": 4.865814696485623,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0808,
+ "step": 1523
+ },
+ {
+ "epoch": 4.8690095846645365,
+ "grad_norm": 0.10546875,
+ "learning_rate": 0.0005,
+ "loss": 1.075,
+ "step": 1524
+ },
+ {
+ "epoch": 4.872204472843451,
+ "grad_norm": 0.08544921875,
+ "learning_rate": 0.0005,
+ "loss": 1.084,
+ "step": 1525
+ },
+ {
+ "epoch": 4.875399361022364,
+ "grad_norm": 0.09521484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0828,
+ "step": 1526
+ },
+ {
+ "epoch": 4.878594249201278,
+ "grad_norm": 0.10498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0884,
+ "step": 1527
+ },
+ {
+ "epoch": 4.881789137380192,
+ "grad_norm": 0.09521484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0777,
+ "step": 1528
+ },
+ {
+ "epoch": 4.884984025559105,
+ "grad_norm": 0.11376953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0916,
+ "step": 1529
+ },
+ {
+ "epoch": 4.888178913738019,
+ "grad_norm": 0.255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0964,
+ "step": 1530
+ },
+ {
+ "epoch": 4.891373801916933,
+ "grad_norm": 0.2197265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0782,
+ "step": 1531
+ },
+ {
+ "epoch": 4.894568690095847,
+ "grad_norm": 0.12060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0872,
+ "step": 1532
+ },
+ {
+ "epoch": 4.897763578274761,
+ "grad_norm": 0.09033203125,
+ "learning_rate": 0.0005,
+ "loss": 1.088,
+ "step": 1533
+ },
+ {
+ "epoch": 4.900958466453674,
+ "grad_norm": 0.1787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0834,
+ "step": 1534
+ },
+ {
+ "epoch": 4.904153354632588,
+ "grad_norm": 0.2255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0728,
+ "step": 1535
+ },
+ {
+ "epoch": 4.907348242811501,
+ "grad_norm": 0.1552734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0846,
+ "step": 1536
+ },
+ {
+ "epoch": 4.9105431309904155,
+ "grad_norm": 0.0986328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0762,
+ "step": 1537
+ },
+ {
+ "epoch": 4.913738019169329,
+ "grad_norm": 0.10302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0875,
+ "step": 1538
+ },
+ {
+ "epoch": 4.916932907348243,
+ "grad_norm": 0.1083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0696,
+ "step": 1539
+ },
+ {
+ "epoch": 4.920127795527156,
+ "grad_norm": 0.09619140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0731,
+ "step": 1540
+ },
+ {
+ "epoch": 4.92332268370607,
+ "grad_norm": 0.1669921875,
+ "learning_rate": 0.0005,
+ "loss": 1.082,
+ "step": 1541
+ },
+ {
+ "epoch": 4.926517571884984,
+ "grad_norm": 0.09814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0791,
+ "step": 1542
+ },
+ {
+ "epoch": 4.9297124600638975,
+ "grad_norm": 0.08251953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0805,
+ "step": 1543
+ },
+ {
+ "epoch": 4.932907348242812,
+ "grad_norm": 0.0830078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0882,
+ "step": 1544
+ },
+ {
+ "epoch": 4.936102236421725,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0683,
+ "step": 1545
+ },
+ {
+ "epoch": 4.939297124600639,
+ "grad_norm": 0.08154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0766,
+ "step": 1546
+ },
+ {
+ "epoch": 4.942492012779553,
+ "grad_norm": 0.1416015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0787,
+ "step": 1547
+ },
+ {
+ "epoch": 4.945686900958466,
+ "grad_norm": 0.1484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0844,
+ "step": 1548
+ },
+ {
+ "epoch": 4.94888178913738,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0893,
+ "step": 1549
+ },
+ {
+ "epoch": 4.952076677316294,
+ "grad_norm": 0.1875,
+ "learning_rate": 0.0005,
+ "loss": 1.0878,
+ "step": 1550
+ },
+ {
+ "epoch": 4.955271565495208,
+ "grad_norm": 0.1552734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0755,
+ "step": 1551
+ },
+ {
+ "epoch": 4.958466453674122,
+ "grad_norm": 0.09716796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0895,
+ "step": 1552
+ },
+ {
+ "epoch": 4.961661341853035,
+ "grad_norm": 0.0849609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0919,
+ "step": 1553
+ },
+ {
+ "epoch": 4.964856230031949,
+ "grad_norm": 0.0908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0889,
+ "step": 1554
+ },
+ {
+ "epoch": 4.968051118210862,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0862,
+ "step": 1555
+ },
+ {
+ "epoch": 4.9712460063897765,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0849,
+ "step": 1556
+ },
+ {
+ "epoch": 4.97444089456869,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0803,
+ "step": 1557
+ },
+ {
+ "epoch": 4.977635782747604,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0832,
+ "step": 1558
+ },
+ {
+ "epoch": 4.980830670926517,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0746,
+ "step": 1559
+ },
+ {
+ "epoch": 4.984025559105431,
+ "grad_norm": 0.091796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0783,
+ "step": 1560
+ },
+ {
+ "epoch": 4.987220447284345,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0776,
+ "step": 1561
+ },
+ {
+ "epoch": 4.9904153354632586,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0758,
+ "step": 1562
+ },
+ {
+ "epoch": 4.993610223642173,
+ "grad_norm": 0.1103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0913,
+ "step": 1563
+ },
+ {
+ "epoch": 4.996805111821086,
+ "grad_norm": 0.09814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0838,
+ "step": 1564
+ },
+ {
+ "epoch": 5.0,
+ "grad_norm": 0.1044921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0967,
+ "step": 1565
+ },
+ {
+ "epoch": 5.003194888178914,
+ "grad_norm": 0.11474609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0758,
+ "step": 1566
+ },
+ {
+ "epoch": 5.006389776357827,
+ "grad_norm": 0.130859375,
+ "learning_rate": 0.0005,
+ "loss": 1.086,
+ "step": 1567
+ },
+ {
+ "epoch": 5.0095846645367414,
+ "grad_norm": 0.08935546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0873,
+ "step": 1568
+ },
+ {
+ "epoch": 5.012779552715655,
+ "grad_norm": 0.1123046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0868,
+ "step": 1569
+ },
+ {
+ "epoch": 5.015974440894569,
+ "grad_norm": 0.099609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0909,
+ "step": 1570
+ },
+ {
+ "epoch": 5.019169329073482,
+ "grad_norm": 0.083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0824,
+ "step": 1571
+ },
+ {
+ "epoch": 5.022364217252396,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0762,
+ "step": 1572
+ },
+ {
+ "epoch": 5.02555910543131,
+ "grad_norm": 0.1123046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0871,
+ "step": 1573
+ },
+ {
+ "epoch": 5.0287539936102235,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0812,
+ "step": 1574
+ },
+ {
+ "epoch": 5.031948881789138,
+ "grad_norm": 0.10205078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0906,
+ "step": 1575
+ },
+ {
+ "epoch": 5.035143769968051,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0716,
+ "step": 1576
+ },
+ {
+ "epoch": 5.038338658146965,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0773,
+ "step": 1577
+ },
+ {
+ "epoch": 5.041533546325879,
+ "grad_norm": 0.10302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0868,
+ "step": 1578
+ },
+ {
+ "epoch": 5.044728434504792,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0823,
+ "step": 1579
+ },
+ {
+ "epoch": 5.047923322683706,
+ "grad_norm": 0.109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0992,
+ "step": 1580
+ },
+ {
+ "epoch": 5.05111821086262,
+ "grad_norm": 0.08837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0803,
+ "step": 1581
+ },
+ {
+ "epoch": 5.054313099041534,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0918,
+ "step": 1582
+ },
+ {
+ "epoch": 5.057507987220447,
+ "grad_norm": 0.08984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0881,
+ "step": 1583
+ },
+ {
+ "epoch": 5.060702875399361,
+ "grad_norm": 0.11865234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0888,
+ "step": 1584
+ },
+ {
+ "epoch": 5.063897763578275,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0856,
+ "step": 1585
+ },
+ {
+ "epoch": 5.067092651757188,
+ "grad_norm": 0.1259765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0718,
+ "step": 1586
+ },
+ {
+ "epoch": 5.0702875399361025,
+ "grad_norm": 0.1435546875,
+ "learning_rate": 0.0005,
+ "loss": 1.082,
+ "step": 1587
+ },
+ {
+ "epoch": 5.073482428115016,
+ "grad_norm": 0.08740234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0715,
+ "step": 1588
+ },
+ {
+ "epoch": 5.07667731629393,
+ "grad_norm": 0.09521484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0806,
+ "step": 1589
+ },
+ {
+ "epoch": 5.079872204472843,
+ "grad_norm": 0.11865234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0817,
+ "step": 1590
+ },
+ {
+ "epoch": 5.083067092651757,
+ "grad_norm": 0.08203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0778,
+ "step": 1591
+ },
+ {
+ "epoch": 5.086261980830671,
+ "grad_norm": 0.08544921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0926,
+ "step": 1592
+ },
+ {
+ "epoch": 5.0894568690095845,
+ "grad_norm": 0.142578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0793,
+ "step": 1593
+ },
+ {
+ "epoch": 5.092651757188499,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0796,
+ "step": 1594
+ },
+ {
+ "epoch": 5.095846645367412,
+ "grad_norm": 0.10791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0796,
+ "step": 1595
+ },
+ {
+ "epoch": 5.099041533546326,
+ "grad_norm": 0.09814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0788,
+ "step": 1596
+ },
+ {
+ "epoch": 5.102236421725239,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0878,
+ "step": 1597
+ },
+ {
+ "epoch": 5.105431309904153,
+ "grad_norm": 0.08984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0792,
+ "step": 1598
+ },
+ {
+ "epoch": 5.108626198083067,
+ "grad_norm": 0.08349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0746,
+ "step": 1599
+ },
+ {
+ "epoch": 5.111821086261981,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0827,
+ "step": 1600
+ },
+ {
+ "epoch": 5.115015974440895,
+ "grad_norm": 0.1083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.084,
+ "step": 1601
+ },
+ {
+ "epoch": 5.118210862619808,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0764,
+ "step": 1602
+ },
+ {
+ "epoch": 5.121405750798722,
+ "grad_norm": 0.08349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0818,
+ "step": 1603
+ },
+ {
+ "epoch": 5.124600638977636,
+ "grad_norm": 0.0830078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0773,
+ "step": 1604
+ },
+ {
+ "epoch": 5.127795527156549,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0817,
+ "step": 1605
+ },
+ {
+ "epoch": 5.1309904153354635,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0722,
+ "step": 1606
+ },
+ {
+ "epoch": 5.134185303514377,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.084,
+ "step": 1607
+ },
+ {
+ "epoch": 5.137380191693291,
+ "grad_norm": 0.09326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0705,
+ "step": 1608
+ },
+ {
+ "epoch": 5.140575079872204,
+ "grad_norm": 0.09619140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0711,
+ "step": 1609
+ },
+ {
+ "epoch": 5.143769968051118,
+ "grad_norm": 0.060791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.089,
+ "step": 1610
+ },
+ {
+ "epoch": 5.146964856230032,
+ "grad_norm": 0.09765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0641,
+ "step": 1611
+ },
+ {
+ "epoch": 5.1501597444089455,
+ "grad_norm": 0.11865234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0859,
+ "step": 1612
+ },
+ {
+ "epoch": 5.15335463258786,
+ "grad_norm": 0.09619140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0791,
+ "step": 1613
+ },
+ {
+ "epoch": 5.156549520766773,
+ "grad_norm": 0.205078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0807,
+ "step": 1614
+ },
+ {
+ "epoch": 5.159744408945687,
+ "grad_norm": 0.19140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0714,
+ "step": 1615
+ },
+ {
+ "epoch": 5.1629392971246,
+ "grad_norm": 0.10302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0788,
+ "step": 1616
+ },
+ {
+ "epoch": 5.166134185303514,
+ "grad_norm": 0.1181640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0773,
+ "step": 1617
+ },
+ {
+ "epoch": 5.169329073482428,
+ "grad_norm": 0.201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0944,
+ "step": 1618
+ },
+ {
+ "epoch": 5.172523961661342,
+ "grad_norm": 0.1630859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0745,
+ "step": 1619
+ },
+ {
+ "epoch": 5.175718849840256,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0805,
+ "step": 1620
+ },
+ {
+ "epoch": 5.178913738019169,
+ "grad_norm": 0.1904296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0964,
+ "step": 1621
+ },
+ {
+ "epoch": 5.182108626198083,
+ "grad_norm": 0.2021484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0761,
+ "step": 1622
+ },
+ {
+ "epoch": 5.185303514376997,
+ "grad_norm": 0.1416015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0773,
+ "step": 1623
+ },
+ {
+ "epoch": 5.18849840255591,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0693,
+ "step": 1624
+ },
+ {
+ "epoch": 5.1916932907348246,
+ "grad_norm": 0.12109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0795,
+ "step": 1625
+ },
+ {
+ "epoch": 5.194888178913738,
+ "grad_norm": 0.05859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0838,
+ "step": 1626
+ },
+ {
+ "epoch": 5.198083067092652,
+ "grad_norm": 0.09423828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0807,
+ "step": 1627
+ },
+ {
+ "epoch": 5.201277955271565,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0787,
+ "step": 1628
+ },
+ {
+ "epoch": 5.204472843450479,
+ "grad_norm": 0.11083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0801,
+ "step": 1629
+ },
+ {
+ "epoch": 5.207667731629393,
+ "grad_norm": 0.13671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0757,
+ "step": 1630
+ },
+ {
+ "epoch": 5.210862619808307,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0777,
+ "step": 1631
+ },
+ {
+ "epoch": 5.214057507987221,
+ "grad_norm": 0.10205078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0879,
+ "step": 1632
+ },
+ {
+ "epoch": 5.217252396166134,
+ "grad_norm": 0.177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0796,
+ "step": 1633
+ },
+ {
+ "epoch": 5.220447284345048,
+ "grad_norm": 0.142578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0719,
+ "step": 1634
+ },
+ {
+ "epoch": 5.223642172523961,
+ "grad_norm": 0.1181640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0792,
+ "step": 1635
+ },
+ {
+ "epoch": 5.226837060702875,
+ "grad_norm": 0.244140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0802,
+ "step": 1636
+ },
+ {
+ "epoch": 5.2300319488817895,
+ "grad_norm": 0.185546875,
+ "learning_rate": 0.0005,
+ "loss": 1.086,
+ "step": 1637
+ },
+ {
+ "epoch": 5.233226837060703,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0816,
+ "step": 1638
+ },
+ {
+ "epoch": 5.236421725239617,
+ "grad_norm": 0.10205078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0859,
+ "step": 1639
+ },
+ {
+ "epoch": 5.23961661341853,
+ "grad_norm": 0.142578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0733,
+ "step": 1640
+ },
+ {
+ "epoch": 5.242811501597444,
+ "grad_norm": 0.125,
+ "learning_rate": 0.0005,
+ "loss": 1.0831,
+ "step": 1641
+ },
+ {
+ "epoch": 5.246006389776358,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0833,
+ "step": 1642
+ },
+ {
+ "epoch": 5.2492012779552715,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0673,
+ "step": 1643
+ },
+ {
+ "epoch": 5.252396166134186,
+ "grad_norm": 0.12451171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0657,
+ "step": 1644
+ },
+ {
+ "epoch": 5.255591054313099,
+ "grad_norm": 0.10888671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0777,
+ "step": 1645
+ },
+ {
+ "epoch": 5.258785942492013,
+ "grad_norm": 0.099609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0673,
+ "step": 1646
+ },
+ {
+ "epoch": 5.261980830670926,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0696,
+ "step": 1647
+ },
+ {
+ "epoch": 5.26517571884984,
+ "grad_norm": 0.1064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.06,
+ "step": 1648
+ },
+ {
+ "epoch": 5.268370607028754,
+ "grad_norm": 0.1181640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0717,
+ "step": 1649
+ },
+ {
+ "epoch": 5.271565495207668,
+ "grad_norm": 0.08056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0851,
+ "step": 1650
+ },
+ {
+ "epoch": 5.274760383386582,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0738,
+ "step": 1651
+ },
+ {
+ "epoch": 5.277955271565495,
+ "grad_norm": 0.07861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0738,
+ "step": 1652
+ },
+ {
+ "epoch": 5.281150159744409,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0885,
+ "step": 1653
+ },
+ {
+ "epoch": 5.284345047923322,
+ "grad_norm": 0.0810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0739,
+ "step": 1654
+ },
+ {
+ "epoch": 5.287539936102236,
+ "grad_norm": 0.1025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0948,
+ "step": 1655
+ },
+ {
+ "epoch": 5.2907348242811505,
+ "grad_norm": 0.0615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0742,
+ "step": 1656
+ },
+ {
+ "epoch": 5.293929712460064,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0911,
+ "step": 1657
+ },
+ {
+ "epoch": 5.297124600638978,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0895,
+ "step": 1658
+ },
+ {
+ "epoch": 5.300319488817891,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0765,
+ "step": 1659
+ },
+ {
+ "epoch": 5.303514376996805,
+ "grad_norm": 0.056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0776,
+ "step": 1660
+ },
+ {
+ "epoch": 5.306709265175719,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0788,
+ "step": 1661
+ },
+ {
+ "epoch": 5.3099041533546325,
+ "grad_norm": 0.0830078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0758,
+ "step": 1662
+ },
+ {
+ "epoch": 5.313099041533547,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0777,
+ "step": 1663
+ },
+ {
+ "epoch": 5.31629392971246,
+ "grad_norm": 0.09521484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0804,
+ "step": 1664
+ },
+ {
+ "epoch": 5.319488817891374,
+ "grad_norm": 0.08251953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0824,
+ "step": 1665
+ },
+ {
+ "epoch": 5.322683706070287,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0703,
+ "step": 1666
+ },
+ {
+ "epoch": 5.325878594249201,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0735,
+ "step": 1667
+ },
+ {
+ "epoch": 5.329073482428115,
+ "grad_norm": 0.1376953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0683,
+ "step": 1668
+ },
+ {
+ "epoch": 5.332268370607029,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0778,
+ "step": 1669
+ },
+ {
+ "epoch": 5.335463258785943,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0868,
+ "step": 1670
+ },
+ {
+ "epoch": 5.338658146964856,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0868,
+ "step": 1671
+ },
+ {
+ "epoch": 5.34185303514377,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0674,
+ "step": 1672
+ },
+ {
+ "epoch": 5.345047923322683,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0739,
+ "step": 1673
+ },
+ {
+ "epoch": 5.348242811501597,
+ "grad_norm": 0.080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0813,
+ "step": 1674
+ },
+ {
+ "epoch": 5.3514376996805115,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.093,
+ "step": 1675
+ },
+ {
+ "epoch": 5.354632587859425,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0678,
+ "step": 1676
+ },
+ {
+ "epoch": 5.357827476038339,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0863,
+ "step": 1677
+ },
+ {
+ "epoch": 5.361022364217252,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0764,
+ "step": 1678
+ },
+ {
+ "epoch": 5.364217252396166,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0715,
+ "step": 1679
+ },
+ {
+ "epoch": 5.36741214057508,
+ "grad_norm": 0.061279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0981,
+ "step": 1680
+ },
+ {
+ "epoch": 5.3706070287539935,
+ "grad_norm": 0.103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0766,
+ "step": 1681
+ },
+ {
+ "epoch": 5.373801916932908,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0784,
+ "step": 1682
+ },
+ {
+ "epoch": 5.376996805111821,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0754,
+ "step": 1683
+ },
+ {
+ "epoch": 5.380191693290735,
+ "grad_norm": 0.08935546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0934,
+ "step": 1684
+ },
+ {
+ "epoch": 5.383386581469648,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0833,
+ "step": 1685
+ },
+ {
+ "epoch": 5.386581469648562,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0793,
+ "step": 1686
+ },
+ {
+ "epoch": 5.389776357827476,
+ "grad_norm": 0.07861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0866,
+ "step": 1687
+ },
+ {
+ "epoch": 5.39297124600639,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0765,
+ "step": 1688
+ },
+ {
+ "epoch": 5.396166134185304,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0773,
+ "step": 1689
+ },
+ {
+ "epoch": 5.399361022364217,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.069,
+ "step": 1690
+ },
+ {
+ "epoch": 5.402555910543131,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0737,
+ "step": 1691
+ },
+ {
+ "epoch": 5.405750798722044,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0714,
+ "step": 1692
+ },
+ {
+ "epoch": 5.4089456869009584,
+ "grad_norm": 0.1240234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0641,
+ "step": 1693
+ },
+ {
+ "epoch": 5.412140575079873,
+ "grad_norm": 0.08056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0736,
+ "step": 1694
+ },
+ {
+ "epoch": 5.415335463258786,
+ "grad_norm": 0.1953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0917,
+ "step": 1695
+ },
+ {
+ "epoch": 5.4185303514377,
+ "grad_norm": 0.1845703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0771,
+ "step": 1696
+ },
+ {
+ "epoch": 5.421725239616613,
+ "grad_norm": 0.173828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0631,
+ "step": 1697
+ },
+ {
+ "epoch": 5.424920127795527,
+ "grad_norm": 0.1669921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0697,
+ "step": 1698
+ },
+ {
+ "epoch": 5.428115015974441,
+ "grad_norm": 0.10205078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0764,
+ "step": 1699
+ },
+ {
+ "epoch": 5.431309904153355,
+ "grad_norm": 0.37109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0824,
+ "step": 1700
+ },
+ {
+ "epoch": 5.434504792332269,
+ "grad_norm": 0.5859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0744,
+ "step": 1701
+ },
+ {
+ "epoch": 5.437699680511182,
+ "grad_norm": 1.2734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0899,
+ "step": 1702
+ },
+ {
+ "epoch": 5.440894568690096,
+ "grad_norm": 1.9921875,
+ "learning_rate": 0.0005,
+ "loss": 1.1113,
+ "step": 1703
+ },
+ {
+ "epoch": 5.444089456869009,
+ "grad_norm": 0.38671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0945,
+ "step": 1704
+ },
+ {
+ "epoch": 5.447284345047923,
+ "grad_norm": 3.140625,
+ "learning_rate": 0.0005,
+ "loss": 1.154,
+ "step": 1705
+ },
+ {
+ "epoch": 5.4504792332268375,
+ "grad_norm": 1.625,
+ "learning_rate": 0.0005,
+ "loss": 1.101,
+ "step": 1706
+ },
+ {
+ "epoch": 5.453674121405751,
+ "grad_norm": 3.1875,
+ "learning_rate": 0.0005,
+ "loss": 1.1798,
+ "step": 1707
+ },
+ {
+ "epoch": 5.456869009584665,
+ "grad_norm": 2.046875,
+ "learning_rate": 0.0005,
+ "loss": 1.1776,
+ "step": 1708
+ },
+ {
+ "epoch": 5.460063897763578,
+ "grad_norm": 1.0390625,
+ "learning_rate": 0.0005,
+ "loss": 1.1488,
+ "step": 1709
+ },
+ {
+ "epoch": 5.463258785942492,
+ "grad_norm": 1.21875,
+ "learning_rate": 0.0005,
+ "loss": 1.119,
+ "step": 1710
+ },
+ {
+ "epoch": 5.466453674121405,
+ "grad_norm": 1.4375,
+ "learning_rate": 0.0005,
+ "loss": 1.1469,
+ "step": 1711
+ },
+ {
+ "epoch": 5.4696485623003195,
+ "grad_norm": 0.275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.1197,
+ "step": 1712
+ },
+ {
+ "epoch": 5.472843450479234,
+ "grad_norm": 0.51171875,
+ "learning_rate": 0.0005,
+ "loss": 1.1213,
+ "step": 1713
+ },
+ {
+ "epoch": 5.476038338658147,
+ "grad_norm": 0.234375,
+ "learning_rate": 0.0005,
+ "loss": 1.1018,
+ "step": 1714
+ },
+ {
+ "epoch": 5.479233226837061,
+ "grad_norm": 0.369140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0925,
+ "step": 1715
+ },
+ {
+ "epoch": 5.482428115015974,
+ "grad_norm": 0.349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.1124,
+ "step": 1716
+ },
+ {
+ "epoch": 5.485623003194888,
+ "grad_norm": 0.32421875,
+ "learning_rate": 0.0005,
+ "loss": 1.09,
+ "step": 1717
+ },
+ {
+ "epoch": 5.488817891373802,
+ "grad_norm": 0.2578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0873,
+ "step": 1718
+ },
+ {
+ "epoch": 5.492012779552716,
+ "grad_norm": 0.310546875,
+ "learning_rate": 0.0005,
+ "loss": 1.1011,
+ "step": 1719
+ },
+ {
+ "epoch": 5.49520766773163,
+ "grad_norm": 0.546875,
+ "learning_rate": 0.0005,
+ "loss": 1.103,
+ "step": 1720
+ },
+ {
+ "epoch": 5.498402555910543,
+ "grad_norm": 0.66015625,
+ "learning_rate": 0.0005,
+ "loss": 1.1051,
+ "step": 1721
+ },
+ {
+ "epoch": 5.501597444089457,
+ "grad_norm": 0.703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0829,
+ "step": 1722
+ },
+ {
+ "epoch": 5.50479233226837,
+ "grad_norm": 0.76171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0966,
+ "step": 1723
+ },
+ {
+ "epoch": 5.507987220447284,
+ "grad_norm": 0.73828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0965,
+ "step": 1724
+ },
+ {
+ "epoch": 5.511182108626198,
+ "grad_norm": 0.33984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0896,
+ "step": 1725
+ },
+ {
+ "epoch": 5.514376996805112,
+ "grad_norm": 0.330078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0867,
+ "step": 1726
+ },
+ {
+ "epoch": 5.517571884984026,
+ "grad_norm": 0.5625,
+ "learning_rate": 0.0005,
+ "loss": 1.0807,
+ "step": 1727
+ },
+ {
+ "epoch": 5.520766773162939,
+ "grad_norm": 0.19921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0694,
+ "step": 1728
+ },
+ {
+ "epoch": 5.523961661341853,
+ "grad_norm": 0.345703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0899,
+ "step": 1729
+ },
+ {
+ "epoch": 5.527156549520766,
+ "grad_norm": 0.30859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0768,
+ "step": 1730
+ },
+ {
+ "epoch": 5.5303514376996805,
+ "grad_norm": 0.20703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0826,
+ "step": 1731
+ },
+ {
+ "epoch": 5.533546325878595,
+ "grad_norm": 0.40625,
+ "learning_rate": 0.0005,
+ "loss": 1.0867,
+ "step": 1732
+ },
+ {
+ "epoch": 5.536741214057508,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0907,
+ "step": 1733
+ },
+ {
+ "epoch": 5.539936102236422,
+ "grad_norm": 0.318359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0739,
+ "step": 1734
+ },
+ {
+ "epoch": 5.543130990415335,
+ "grad_norm": 0.091796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0754,
+ "step": 1735
+ },
+ {
+ "epoch": 5.546325878594249,
+ "grad_norm": 0.26953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0721,
+ "step": 1736
+ },
+ {
+ "epoch": 5.549520766773163,
+ "grad_norm": 0.08642578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0766,
+ "step": 1737
+ },
+ {
+ "epoch": 5.552715654952077,
+ "grad_norm": 0.271484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0849,
+ "step": 1738
+ },
+ {
+ "epoch": 5.555910543130991,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0652,
+ "step": 1739
+ },
+ {
+ "epoch": 5.559105431309904,
+ "grad_norm": 0.25,
+ "learning_rate": 0.0005,
+ "loss": 1.0792,
+ "step": 1740
+ },
+ {
+ "epoch": 5.562300319488818,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0795,
+ "step": 1741
+ },
+ {
+ "epoch": 5.565495207667731,
+ "grad_norm": 0.2392578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0912,
+ "step": 1742
+ },
+ {
+ "epoch": 5.568690095846645,
+ "grad_norm": 0.10302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.083,
+ "step": 1743
+ },
+ {
+ "epoch": 5.571884984025559,
+ "grad_norm": 0.279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0878,
+ "step": 1744
+ },
+ {
+ "epoch": 5.575079872204473,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0857,
+ "step": 1745
+ },
+ {
+ "epoch": 5.578274760383387,
+ "grad_norm": 0.2373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0813,
+ "step": 1746
+ },
+ {
+ "epoch": 5.5814696485623,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0914,
+ "step": 1747
+ },
+ {
+ "epoch": 5.584664536741214,
+ "grad_norm": 0.205078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0864,
+ "step": 1748
+ },
+ {
+ "epoch": 5.587859424920127,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0706,
+ "step": 1749
+ },
+ {
+ "epoch": 5.5910543130990416,
+ "grad_norm": 0.1630859375,
+ "learning_rate": 0.0005,
+ "loss": 1.069,
+ "step": 1750
+ },
+ {
+ "epoch": 5.594249201277956,
+ "grad_norm": 0.0986328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0786,
+ "step": 1751
+ },
+ {
+ "epoch": 5.597444089456869,
+ "grad_norm": 0.146484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0686,
+ "step": 1752
+ },
+ {
+ "epoch": 5.600638977635783,
+ "grad_norm": 0.1123046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0937,
+ "step": 1753
+ },
+ {
+ "epoch": 5.603833865814696,
+ "grad_norm": 0.1162109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0699,
+ "step": 1754
+ },
+ {
+ "epoch": 5.60702875399361,
+ "grad_norm": 0.10009765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0742,
+ "step": 1755
+ },
+ {
+ "epoch": 5.6102236421725244,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.088,
+ "step": 1756
+ },
+ {
+ "epoch": 5.613418530351438,
+ "grad_norm": 0.087890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0786,
+ "step": 1757
+ },
+ {
+ "epoch": 5.616613418530352,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0689,
+ "step": 1758
+ },
+ {
+ "epoch": 5.619808306709265,
+ "grad_norm": 0.09326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0719,
+ "step": 1759
+ },
+ {
+ "epoch": 5.623003194888179,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.08,
+ "step": 1760
+ },
+ {
+ "epoch": 5.626198083067092,
+ "grad_norm": 0.09716796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0758,
+ "step": 1761
+ },
+ {
+ "epoch": 5.6293929712460065,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0774,
+ "step": 1762
+ },
+ {
+ "epoch": 5.63258785942492,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0848,
+ "step": 1763
+ },
+ {
+ "epoch": 5.635782747603834,
+ "grad_norm": 0.08740234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0838,
+ "step": 1764
+ },
+ {
+ "epoch": 5.638977635782748,
+ "grad_norm": 0.11865234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0783,
+ "step": 1765
+ },
+ {
+ "epoch": 5.642172523961661,
+ "grad_norm": 0.10888671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0656,
+ "step": 1766
+ },
+ {
+ "epoch": 5.645367412140575,
+ "grad_norm": 0.162109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0854,
+ "step": 1767
+ },
+ {
+ "epoch": 5.6485623003194885,
+ "grad_norm": 0.142578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0817,
+ "step": 1768
+ },
+ {
+ "epoch": 5.651757188498403,
+ "grad_norm": 0.2177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0788,
+ "step": 1769
+ },
+ {
+ "epoch": 5.654952076677317,
+ "grad_norm": 0.10986328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0618,
+ "step": 1770
+ },
+ {
+ "epoch": 5.65814696485623,
+ "grad_norm": 0.201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0753,
+ "step": 1771
+ },
+ {
+ "epoch": 5.661341853035144,
+ "grad_norm": 0.0908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0677,
+ "step": 1772
+ },
+ {
+ "epoch": 5.664536741214057,
+ "grad_norm": 0.232421875,
+ "learning_rate": 0.0005,
+ "loss": 1.077,
+ "step": 1773
+ },
+ {
+ "epoch": 5.667731629392971,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.071,
+ "step": 1774
+ },
+ {
+ "epoch": 5.6709265175718855,
+ "grad_norm": 0.1806640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0925,
+ "step": 1775
+ },
+ {
+ "epoch": 5.674121405750799,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0887,
+ "step": 1776
+ },
+ {
+ "epoch": 5.677316293929713,
+ "grad_norm": 0.140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0653,
+ "step": 1777
+ },
+ {
+ "epoch": 5.680511182108626,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.079,
+ "step": 1778
+ },
+ {
+ "epoch": 5.68370607028754,
+ "grad_norm": 0.11767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0723,
+ "step": 1779
+ },
+ {
+ "epoch": 5.686900958466453,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0795,
+ "step": 1780
+ },
+ {
+ "epoch": 5.6900958466453675,
+ "grad_norm": 0.130859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0639,
+ "step": 1781
+ },
+ {
+ "epoch": 5.693290734824281,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0715,
+ "step": 1782
+ },
+ {
+ "epoch": 5.696485623003195,
+ "grad_norm": 0.08740234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0754,
+ "step": 1783
+ },
+ {
+ "epoch": 5.699680511182109,
+ "grad_norm": 0.09228515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0773,
+ "step": 1784
+ },
+ {
+ "epoch": 5.702875399361022,
+ "grad_norm": 0.09765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0737,
+ "step": 1785
+ },
+ {
+ "epoch": 5.706070287539936,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0913,
+ "step": 1786
+ },
+ {
+ "epoch": 5.7092651757188495,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0693,
+ "step": 1787
+ },
+ {
+ "epoch": 5.712460063897764,
+ "grad_norm": 0.08837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.079,
+ "step": 1788
+ },
+ {
+ "epoch": 5.715654952076678,
+ "grad_norm": 0.107421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0647,
+ "step": 1789
+ },
+ {
+ "epoch": 5.718849840255591,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.075,
+ "step": 1790
+ },
+ {
+ "epoch": 5.722044728434505,
+ "grad_norm": 0.12255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0717,
+ "step": 1791
+ },
+ {
+ "epoch": 5.725239616613418,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0774,
+ "step": 1792
+ },
+ {
+ "epoch": 5.728434504792332,
+ "grad_norm": 0.1328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0789,
+ "step": 1793
+ },
+ {
+ "epoch": 5.731629392971246,
+ "grad_norm": 0.0869140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0721,
+ "step": 1794
+ },
+ {
+ "epoch": 5.73482428115016,
+ "grad_norm": 0.12158203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0749,
+ "step": 1795
+ },
+ {
+ "epoch": 5.738019169329074,
+ "grad_norm": 0.09375,
+ "learning_rate": 0.0005,
+ "loss": 1.0875,
+ "step": 1796
+ },
+ {
+ "epoch": 5.741214057507987,
+ "grad_norm": 0.146484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0953,
+ "step": 1797
+ },
+ {
+ "epoch": 5.744408945686901,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0744,
+ "step": 1798
+ },
+ {
+ "epoch": 5.747603833865814,
+ "grad_norm": 0.10693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0787,
+ "step": 1799
+ },
+ {
+ "epoch": 5.7507987220447285,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0727,
+ "step": 1800
+ },
+ {
+ "epoch": 5.753993610223642,
+ "grad_norm": 0.09130859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0855,
+ "step": 1801
+ },
+ {
+ "epoch": 5.757188498402556,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0842,
+ "step": 1802
+ },
+ {
+ "epoch": 5.76038338658147,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.071,
+ "step": 1803
+ },
+ {
+ "epoch": 5.763578274760383,
+ "grad_norm": 0.11328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0634,
+ "step": 1804
+ },
+ {
+ "epoch": 5.766773162939297,
+ "grad_norm": 0.08154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0821,
+ "step": 1805
+ },
+ {
+ "epoch": 5.7699680511182105,
+ "grad_norm": 0.0849609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0864,
+ "step": 1806
+ },
+ {
+ "epoch": 5.773162939297125,
+ "grad_norm": 0.10595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0792,
+ "step": 1807
+ },
+ {
+ "epoch": 5.776357827476039,
+ "grad_norm": 0.058837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0811,
+ "step": 1808
+ },
+ {
+ "epoch": 5.779552715654952,
+ "grad_norm": 0.08447265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0803,
+ "step": 1809
+ },
+ {
+ "epoch": 5.782747603833866,
+ "grad_norm": 0.0869140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0787,
+ "step": 1810
+ },
+ {
+ "epoch": 5.785942492012779,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0873,
+ "step": 1811
+ },
+ {
+ "epoch": 5.789137380191693,
+ "grad_norm": 0.08056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.067,
+ "step": 1812
+ },
+ {
+ "epoch": 5.792332268370607,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0609,
+ "step": 1813
+ },
+ {
+ "epoch": 5.795527156549521,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0794,
+ "step": 1814
+ },
+ {
+ "epoch": 5.798722044728435,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0884,
+ "step": 1815
+ },
+ {
+ "epoch": 5.801916932907348,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0824,
+ "step": 1816
+ },
+ {
+ "epoch": 5.805111821086262,
+ "grad_norm": 0.06201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0632,
+ "step": 1817
+ },
+ {
+ "epoch": 5.8083067092651754,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0901,
+ "step": 1818
+ },
+ {
+ "epoch": 5.81150159744409,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0801,
+ "step": 1819
+ },
+ {
+ "epoch": 5.814696485623003,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0781,
+ "step": 1820
+ },
+ {
+ "epoch": 5.817891373801917,
+ "grad_norm": 0.146484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0941,
+ "step": 1821
+ },
+ {
+ "epoch": 5.821086261980831,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0749,
+ "step": 1822
+ },
+ {
+ "epoch": 5.824281150159744,
+ "grad_norm": 0.169921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0806,
+ "step": 1823
+ },
+ {
+ "epoch": 5.827476038338658,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0672,
+ "step": 1824
+ },
+ {
+ "epoch": 5.830670926517572,
+ "grad_norm": 0.1318359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0722,
+ "step": 1825
+ },
+ {
+ "epoch": 5.833865814696486,
+ "grad_norm": 0.0849609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0764,
+ "step": 1826
+ },
+ {
+ "epoch": 5.8370607028754,
+ "grad_norm": 0.09130859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0719,
+ "step": 1827
+ },
+ {
+ "epoch": 5.840255591054313,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0901,
+ "step": 1828
+ },
+ {
+ "epoch": 5.843450479233227,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0879,
+ "step": 1829
+ },
+ {
+ "epoch": 5.84664536741214,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.065,
+ "step": 1830
+ },
+ {
+ "epoch": 5.8498402555910545,
+ "grad_norm": 0.06103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0872,
+ "step": 1831
+ },
+ {
+ "epoch": 5.853035143769968,
+ "grad_norm": 0.08447265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0717,
+ "step": 1832
+ },
+ {
+ "epoch": 5.856230031948882,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0871,
+ "step": 1833
+ },
+ {
+ "epoch": 5.859424920127795,
+ "grad_norm": 0.08837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0763,
+ "step": 1834
+ },
+ {
+ "epoch": 5.862619808306709,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0743,
+ "step": 1835
+ },
+ {
+ "epoch": 5.865814696485623,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0783,
+ "step": 1836
+ },
+ {
+ "epoch": 5.8690095846645365,
+ "grad_norm": 0.051513671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0784,
+ "step": 1837
+ },
+ {
+ "epoch": 5.872204472843451,
+ "grad_norm": 0.080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0758,
+ "step": 1838
+ },
+ {
+ "epoch": 5.875399361022364,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.085,
+ "step": 1839
+ },
+ {
+ "epoch": 5.878594249201278,
+ "grad_norm": 0.08349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0742,
+ "step": 1840
+ },
+ {
+ "epoch": 5.881789137380192,
+ "grad_norm": 0.052978515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0776,
+ "step": 1841
+ },
+ {
+ "epoch": 5.884984025559105,
+ "grad_norm": 0.1259765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0868,
+ "step": 1842
+ },
+ {
+ "epoch": 5.888178913738019,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0716,
+ "step": 1843
+ },
+ {
+ "epoch": 5.891373801916933,
+ "grad_norm": 0.107421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0835,
+ "step": 1844
+ },
+ {
+ "epoch": 5.894568690095847,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0761,
+ "step": 1845
+ },
+ {
+ "epoch": 5.897763578274761,
+ "grad_norm": 0.08349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0827,
+ "step": 1846
+ },
+ {
+ "epoch": 5.900958466453674,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.077,
+ "step": 1847
+ },
+ {
+ "epoch": 5.904153354632588,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0731,
+ "step": 1848
+ },
+ {
+ "epoch": 5.907348242811501,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0949,
+ "step": 1849
+ },
+ {
+ "epoch": 5.9105431309904155,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0667,
+ "step": 1850
+ },
+ {
+ "epoch": 5.913738019169329,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0715,
+ "step": 1851
+ },
+ {
+ "epoch": 5.916932907348243,
+ "grad_norm": 0.09619140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0721,
+ "step": 1852
+ },
+ {
+ "epoch": 5.920127795527156,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0771,
+ "step": 1853
+ },
+ {
+ "epoch": 5.92332268370607,
+ "grad_norm": 0.09814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0807,
+ "step": 1854
+ },
+ {
+ "epoch": 5.926517571884984,
+ "grad_norm": 0.056884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0742,
+ "step": 1855
+ },
+ {
+ "epoch": 5.9297124600638975,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0712,
+ "step": 1856
+ },
+ {
+ "epoch": 5.932907348242812,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0677,
+ "step": 1857
+ },
+ {
+ "epoch": 5.936102236421725,
+ "grad_norm": 0.1181640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0753,
+ "step": 1858
+ },
+ {
+ "epoch": 5.939297124600639,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0804,
+ "step": 1859
+ },
+ {
+ "epoch": 5.942492012779553,
+ "grad_norm": 0.125,
+ "learning_rate": 0.0005,
+ "loss": 1.076,
+ "step": 1860
+ },
+ {
+ "epoch": 5.945686900958466,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0844,
+ "step": 1861
+ },
+ {
+ "epoch": 5.94888178913738,
+ "grad_norm": 0.10888671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0723,
+ "step": 1862
+ },
+ {
+ "epoch": 5.952076677316294,
+ "grad_norm": 0.109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0787,
+ "step": 1863
+ },
+ {
+ "epoch": 5.955271565495208,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0684,
+ "step": 1864
+ },
+ {
+ "epoch": 5.958466453674122,
+ "grad_norm": 0.1298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0849,
+ "step": 1865
+ },
+ {
+ "epoch": 5.961661341853035,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0722,
+ "step": 1866
+ },
+ {
+ "epoch": 5.964856230031949,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0733,
+ "step": 1867
+ },
+ {
+ "epoch": 5.968051118210862,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0585,
+ "step": 1868
+ },
+ {
+ "epoch": 5.9712460063897765,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0876,
+ "step": 1869
+ },
+ {
+ "epoch": 5.97444089456869,
+ "grad_norm": 0.061767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0816,
+ "step": 1870
+ },
+ {
+ "epoch": 5.977635782747604,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0758,
+ "step": 1871
+ },
+ {
+ "epoch": 5.980830670926517,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0983,
+ "step": 1872
+ },
+ {
+ "epoch": 5.984025559105431,
+ "grad_norm": 0.061767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0682,
+ "step": 1873
+ },
+ {
+ "epoch": 5.987220447284345,
+ "grad_norm": 0.09228515625,
+ "learning_rate": 0.0005,
+ "loss": 1.078,
+ "step": 1874
+ },
+ {
+ "epoch": 5.9904153354632586,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0749,
+ "step": 1875
+ },
+ {
+ "epoch": 5.993610223642173,
+ "grad_norm": 0.060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0726,
+ "step": 1876
+ },
+ {
+ "epoch": 5.996805111821086,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0664,
+ "step": 1877
+ },
+ {
+ "epoch": 6.0,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0812,
+ "step": 1878
+ },
+ {
+ "epoch": 6.003194888178914,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0872,
+ "step": 1879
+ },
+ {
+ "epoch": 6.006389776357827,
+ "grad_norm": 0.07861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0844,
+ "step": 1880
+ },
+ {
+ "epoch": 6.0095846645367414,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.068,
+ "step": 1881
+ },
+ {
+ "epoch": 6.012779552715655,
+ "grad_norm": 0.0830078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0765,
+ "step": 1882
+ },
+ {
+ "epoch": 6.015974440894569,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0845,
+ "step": 1883
+ },
+ {
+ "epoch": 6.019169329073482,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0758,
+ "step": 1884
+ },
+ {
+ "epoch": 6.022364217252396,
+ "grad_norm": 0.0849609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0799,
+ "step": 1885
+ },
+ {
+ "epoch": 6.02555910543131,
+ "grad_norm": 0.09619140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0821,
+ "step": 1886
+ },
+ {
+ "epoch": 6.0287539936102235,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0907,
+ "step": 1887
+ },
+ {
+ "epoch": 6.031948881789138,
+ "grad_norm": 0.08154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0729,
+ "step": 1888
+ },
+ {
+ "epoch": 6.035143769968051,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0731,
+ "step": 1889
+ },
+ {
+ "epoch": 6.038338658146965,
+ "grad_norm": 0.09375,
+ "learning_rate": 0.0005,
+ "loss": 1.0751,
+ "step": 1890
+ },
+ {
+ "epoch": 6.041533546325879,
+ "grad_norm": 0.061279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.071,
+ "step": 1891
+ },
+ {
+ "epoch": 6.044728434504792,
+ "grad_norm": 0.09130859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0913,
+ "step": 1892
+ },
+ {
+ "epoch": 6.047923322683706,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0775,
+ "step": 1893
+ },
+ {
+ "epoch": 6.05111821086262,
+ "grad_norm": 0.05908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0792,
+ "step": 1894
+ },
+ {
+ "epoch": 6.054313099041534,
+ "grad_norm": 0.09033203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0836,
+ "step": 1895
+ },
+ {
+ "epoch": 6.057507987220447,
+ "grad_norm": 0.05810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0956,
+ "step": 1896
+ },
+ {
+ "epoch": 6.060702875399361,
+ "grad_norm": 0.08837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0622,
+ "step": 1897
+ },
+ {
+ "epoch": 6.063897763578275,
+ "grad_norm": 0.058837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0776,
+ "step": 1898
+ },
+ {
+ "epoch": 6.067092651757188,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0749,
+ "step": 1899
+ },
+ {
+ "epoch": 6.0702875399361025,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0713,
+ "step": 1900
+ },
+ {
+ "epoch": 6.073482428115016,
+ "grad_norm": 0.08544921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0693,
+ "step": 1901
+ },
+ {
+ "epoch": 6.07667731629393,
+ "grad_norm": 0.08984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0729,
+ "step": 1902
+ },
+ {
+ "epoch": 6.079872204472843,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.061,
+ "step": 1903
+ },
+ {
+ "epoch": 6.083067092651757,
+ "grad_norm": 0.0908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0752,
+ "step": 1904
+ },
+ {
+ "epoch": 6.086261980830671,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0828,
+ "step": 1905
+ },
+ {
+ "epoch": 6.0894568690095845,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0762,
+ "step": 1906
+ },
+ {
+ "epoch": 6.092651757188499,
+ "grad_norm": 0.08154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0843,
+ "step": 1907
+ },
+ {
+ "epoch": 6.095846645367412,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0727,
+ "step": 1908
+ },
+ {
+ "epoch": 6.099041533546326,
+ "grad_norm": 0.057373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.091,
+ "step": 1909
+ },
+ {
+ "epoch": 6.102236421725239,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0762,
+ "step": 1910
+ },
+ {
+ "epoch": 6.105431309904153,
+ "grad_norm": 0.05908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0837,
+ "step": 1911
+ },
+ {
+ "epoch": 6.108626198083067,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0726,
+ "step": 1912
+ },
+ {
+ "epoch": 6.111821086261981,
+ "grad_norm": 0.08154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0796,
+ "step": 1913
+ },
+ {
+ "epoch": 6.115015974440895,
+ "grad_norm": 0.05126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.096,
+ "step": 1914
+ },
+ {
+ "epoch": 6.118210862619808,
+ "grad_norm": 0.060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0875,
+ "step": 1915
+ },
+ {
+ "epoch": 6.121405750798722,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0652,
+ "step": 1916
+ },
+ {
+ "epoch": 6.124600638977636,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0813,
+ "step": 1917
+ },
+ {
+ "epoch": 6.127795527156549,
+ "grad_norm": 0.08203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0798,
+ "step": 1918
+ },
+ {
+ "epoch": 6.1309904153354635,
+ "grad_norm": 0.1015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0792,
+ "step": 1919
+ },
+ {
+ "epoch": 6.134185303514377,
+ "grad_norm": 0.08935546875,
+ "learning_rate": 0.0005,
+ "loss": 1.087,
+ "step": 1920
+ },
+ {
+ "epoch": 6.137380191693291,
+ "grad_norm": 0.146484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0785,
+ "step": 1921
+ },
+ {
+ "epoch": 6.140575079872204,
+ "grad_norm": 0.080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0704,
+ "step": 1922
+ },
+ {
+ "epoch": 6.143769968051118,
+ "grad_norm": 0.0849609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0727,
+ "step": 1923
+ },
+ {
+ "epoch": 6.146964856230032,
+ "grad_norm": 0.0830078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0677,
+ "step": 1924
+ },
+ {
+ "epoch": 6.1501597444089455,
+ "grad_norm": 0.08203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0905,
+ "step": 1925
+ },
+ {
+ "epoch": 6.15335463258786,
+ "grad_norm": 0.0595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0832,
+ "step": 1926
+ },
+ {
+ "epoch": 6.156549520766773,
+ "grad_norm": 0.09130859375,
+ "learning_rate": 0.0005,
+ "loss": 1.068,
+ "step": 1927
+ },
+ {
+ "epoch": 6.159744408945687,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0699,
+ "step": 1928
+ },
+ {
+ "epoch": 6.1629392971246,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0716,
+ "step": 1929
+ },
+ {
+ "epoch": 6.166134185303514,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0859,
+ "step": 1930
+ },
+ {
+ "epoch": 6.169329073482428,
+ "grad_norm": 0.09423828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0845,
+ "step": 1931
+ },
+ {
+ "epoch": 6.172523961661342,
+ "grad_norm": 0.0947265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0826,
+ "step": 1932
+ },
+ {
+ "epoch": 6.175718849840256,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0682,
+ "step": 1933
+ },
+ {
+ "epoch": 6.178913738019169,
+ "grad_norm": 0.10009765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0796,
+ "step": 1934
+ },
+ {
+ "epoch": 6.182108626198083,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0655,
+ "step": 1935
+ },
+ {
+ "epoch": 6.185303514376997,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0703,
+ "step": 1936
+ },
+ {
+ "epoch": 6.18849840255591,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0686,
+ "step": 1937
+ },
+ {
+ "epoch": 6.1916932907348246,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0786,
+ "step": 1938
+ },
+ {
+ "epoch": 6.194888178913738,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0774,
+ "step": 1939
+ },
+ {
+ "epoch": 6.198083067092652,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.076,
+ "step": 1940
+ },
+ {
+ "epoch": 6.201277955271565,
+ "grad_norm": 0.0830078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0727,
+ "step": 1941
+ },
+ {
+ "epoch": 6.204472843450479,
+ "grad_norm": 0.083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0686,
+ "step": 1942
+ },
+ {
+ "epoch": 6.207667731629393,
+ "grad_norm": 0.1103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0684,
+ "step": 1943
+ },
+ {
+ "epoch": 6.210862619808307,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0761,
+ "step": 1944
+ },
+ {
+ "epoch": 6.214057507987221,
+ "grad_norm": 0.1376953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0729,
+ "step": 1945
+ },
+ {
+ "epoch": 6.217252396166134,
+ "grad_norm": 0.08349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0709,
+ "step": 1946
+ },
+ {
+ "epoch": 6.220447284345048,
+ "grad_norm": 0.1630859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0724,
+ "step": 1947
+ },
+ {
+ "epoch": 6.223642172523961,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0876,
+ "step": 1948
+ },
+ {
+ "epoch": 6.226837060702875,
+ "grad_norm": 0.10400390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0782,
+ "step": 1949
+ },
+ {
+ "epoch": 6.2300319488817895,
+ "grad_norm": 0.0830078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0886,
+ "step": 1950
+ },
+ {
+ "epoch": 6.233226837060703,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0727,
+ "step": 1951
+ },
+ {
+ "epoch": 6.236421725239617,
+ "grad_norm": 0.08251953125,
+ "learning_rate": 0.0005,
+ "loss": 1.066,
+ "step": 1952
+ },
+ {
+ "epoch": 6.23961661341853,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0685,
+ "step": 1953
+ },
+ {
+ "epoch": 6.242811501597444,
+ "grad_norm": 0.12109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0797,
+ "step": 1954
+ },
+ {
+ "epoch": 6.246006389776358,
+ "grad_norm": 0.06103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0719,
+ "step": 1955
+ },
+ {
+ "epoch": 6.2492012779552715,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0552,
+ "step": 1956
+ },
+ {
+ "epoch": 6.252396166134186,
+ "grad_norm": 0.09765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0741,
+ "step": 1957
+ },
+ {
+ "epoch": 6.255591054313099,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0631,
+ "step": 1958
+ },
+ {
+ "epoch": 6.258785942492013,
+ "grad_norm": 0.1494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0713,
+ "step": 1959
+ },
+ {
+ "epoch": 6.261980830670926,
+ "grad_norm": 0.08837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0843,
+ "step": 1960
+ },
+ {
+ "epoch": 6.26517571884984,
+ "grad_norm": 0.1103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.078,
+ "step": 1961
+ },
+ {
+ "epoch": 6.268370607028754,
+ "grad_norm": 0.12255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0816,
+ "step": 1962
+ },
+ {
+ "epoch": 6.271565495207668,
+ "grad_norm": 0.08642578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0807,
+ "step": 1963
+ },
+ {
+ "epoch": 6.274760383386582,
+ "grad_norm": 0.1171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0887,
+ "step": 1964
+ },
+ {
+ "epoch": 6.277955271565495,
+ "grad_norm": 0.0966796875,
+ "learning_rate": 0.0005,
+ "loss": 1.087,
+ "step": 1965
+ },
+ {
+ "epoch": 6.281150159744409,
+ "grad_norm": 0.111328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0645,
+ "step": 1966
+ },
+ {
+ "epoch": 6.284345047923322,
+ "grad_norm": 0.10986328125,
+ "learning_rate": 0.0005,
+ "loss": 1.087,
+ "step": 1967
+ },
+ {
+ "epoch": 6.287539936102236,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0893,
+ "step": 1968
+ },
+ {
+ "epoch": 6.2907348242811505,
+ "grad_norm": 0.1162109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0642,
+ "step": 1969
+ },
+ {
+ "epoch": 6.293929712460064,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0715,
+ "step": 1970
+ },
+ {
+ "epoch": 6.297124600638978,
+ "grad_norm": 0.158203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0824,
+ "step": 1971
+ },
+ {
+ "epoch": 6.300319488817891,
+ "grad_norm": 0.09228515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0863,
+ "step": 1972
+ },
+ {
+ "epoch": 6.303514376996805,
+ "grad_norm": 0.10009765625,
+ "learning_rate": 0.0005,
+ "loss": 1.079,
+ "step": 1973
+ },
+ {
+ "epoch": 6.306709265175719,
+ "grad_norm": 0.1259765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0798,
+ "step": 1974
+ },
+ {
+ "epoch": 6.3099041533546325,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0589,
+ "step": 1975
+ },
+ {
+ "epoch": 6.313099041533547,
+ "grad_norm": 0.1533203125,
+ "learning_rate": 0.0005,
+ "loss": 1.057,
+ "step": 1976
+ },
+ {
+ "epoch": 6.31629392971246,
+ "grad_norm": 0.08837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.084,
+ "step": 1977
+ },
+ {
+ "epoch": 6.319488817891374,
+ "grad_norm": 0.1708984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0693,
+ "step": 1978
+ },
+ {
+ "epoch": 6.322683706070287,
+ "grad_norm": 0.0830078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0735,
+ "step": 1979
+ },
+ {
+ "epoch": 6.325878594249201,
+ "grad_norm": 0.11083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0827,
+ "step": 1980
+ },
+ {
+ "epoch": 6.329073482428115,
+ "grad_norm": 0.0908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0838,
+ "step": 1981
+ },
+ {
+ "epoch": 6.332268370607029,
+ "grad_norm": 0.08251953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0869,
+ "step": 1982
+ },
+ {
+ "epoch": 6.335463258785943,
+ "grad_norm": 0.0966796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0778,
+ "step": 1983
+ },
+ {
+ "epoch": 6.338658146964856,
+ "grad_norm": 0.083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0806,
+ "step": 1984
+ },
+ {
+ "epoch": 6.34185303514377,
+ "grad_norm": 0.083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0825,
+ "step": 1985
+ },
+ {
+ "epoch": 6.345047923322683,
+ "grad_norm": 0.09521484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0702,
+ "step": 1986
+ },
+ {
+ "epoch": 6.348242811501597,
+ "grad_norm": 0.09619140625,
+ "learning_rate": 0.0005,
+ "loss": 1.079,
+ "step": 1987
+ },
+ {
+ "epoch": 6.3514376996805115,
+ "grad_norm": 0.1474609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0939,
+ "step": 1988
+ },
+ {
+ "epoch": 6.354632587859425,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0808,
+ "step": 1989
+ },
+ {
+ "epoch": 6.357827476038339,
+ "grad_norm": 0.11962890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0815,
+ "step": 1990
+ },
+ {
+ "epoch": 6.361022364217252,
+ "grad_norm": 0.095703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0737,
+ "step": 1991
+ },
+ {
+ "epoch": 6.364217252396166,
+ "grad_norm": 0.083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0652,
+ "step": 1992
+ },
+ {
+ "epoch": 6.36741214057508,
+ "grad_norm": 0.11279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0783,
+ "step": 1993
+ },
+ {
+ "epoch": 6.3706070287539935,
+ "grad_norm": 0.0888671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0611,
+ "step": 1994
+ },
+ {
+ "epoch": 6.373801916932908,
+ "grad_norm": 0.12109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0897,
+ "step": 1995
+ },
+ {
+ "epoch": 6.376996805111821,
+ "grad_norm": 0.09619140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0852,
+ "step": 1996
+ },
+ {
+ "epoch": 6.380191693290735,
+ "grad_norm": 0.09716796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0873,
+ "step": 1997
+ },
+ {
+ "epoch": 6.383386581469648,
+ "grad_norm": 0.193359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0822,
+ "step": 1998
+ },
+ {
+ "epoch": 6.386581469648562,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0773,
+ "step": 1999
+ },
+ {
+ "epoch": 6.389776357827476,
+ "grad_norm": 0.1669921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0588,
+ "step": 2000
+ },
+ {
+ "epoch": 6.39297124600639,
+ "grad_norm": 0.10107421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0681,
+ "step": 2001
+ },
+ {
+ "epoch": 6.396166134185304,
+ "grad_norm": 0.09130859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0825,
+ "step": 2002
+ },
+ {
+ "epoch": 6.399361022364217,
+ "grad_norm": 0.083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0703,
+ "step": 2003
+ },
+ {
+ "epoch": 6.402555910543131,
+ "grad_norm": 0.08447265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0562,
+ "step": 2004
+ },
+ {
+ "epoch": 6.405750798722044,
+ "grad_norm": 0.107421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0575,
+ "step": 2005
+ },
+ {
+ "epoch": 6.4089456869009584,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0814,
+ "step": 2006
+ },
+ {
+ "epoch": 6.412140575079873,
+ "grad_norm": 0.09765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0752,
+ "step": 2007
+ },
+ {
+ "epoch": 6.415335463258786,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0767,
+ "step": 2008
+ },
+ {
+ "epoch": 6.4185303514377,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0829,
+ "step": 2009
+ },
+ {
+ "epoch": 6.421725239616613,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0861,
+ "step": 2010
+ },
+ {
+ "epoch": 6.424920127795527,
+ "grad_norm": 0.0888671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0651,
+ "step": 2011
+ },
+ {
+ "epoch": 6.428115015974441,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0757,
+ "step": 2012
+ },
+ {
+ "epoch": 6.431309904153355,
+ "grad_norm": 0.0830078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0848,
+ "step": 2013
+ },
+ {
+ "epoch": 6.434504792332269,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0823,
+ "step": 2014
+ },
+ {
+ "epoch": 6.437699680511182,
+ "grad_norm": 0.1142578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0793,
+ "step": 2015
+ },
+ {
+ "epoch": 6.440894568690096,
+ "grad_norm": 0.09521484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0716,
+ "step": 2016
+ },
+ {
+ "epoch": 6.444089456869009,
+ "grad_norm": 0.061279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0888,
+ "step": 2017
+ },
+ {
+ "epoch": 6.447284345047923,
+ "grad_norm": 0.09375,
+ "learning_rate": 0.0005,
+ "loss": 1.0709,
+ "step": 2018
+ },
+ {
+ "epoch": 6.4504792332268375,
+ "grad_norm": 0.08837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.065,
+ "step": 2019
+ },
+ {
+ "epoch": 6.453674121405751,
+ "grad_norm": 0.062255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0792,
+ "step": 2020
+ },
+ {
+ "epoch": 6.456869009584665,
+ "grad_norm": 0.10009765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0552,
+ "step": 2021
+ },
+ {
+ "epoch": 6.460063897763578,
+ "grad_norm": 0.0498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0528,
+ "step": 2022
+ },
+ {
+ "epoch": 6.463258785942492,
+ "grad_norm": 0.10400390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0692,
+ "step": 2023
+ },
+ {
+ "epoch": 6.466453674121405,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0776,
+ "step": 2024
+ },
+ {
+ "epoch": 6.4696485623003195,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0725,
+ "step": 2025
+ },
+ {
+ "epoch": 6.472843450479234,
+ "grad_norm": 0.0615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0733,
+ "step": 2026
+ },
+ {
+ "epoch": 6.476038338658147,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0747,
+ "step": 2027
+ },
+ {
+ "epoch": 6.479233226837061,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0776,
+ "step": 2028
+ },
+ {
+ "epoch": 6.482428115015974,
+ "grad_norm": 0.08251953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0637,
+ "step": 2029
+ },
+ {
+ "epoch": 6.485623003194888,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0682,
+ "step": 2030
+ },
+ {
+ "epoch": 6.488817891373802,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0777,
+ "step": 2031
+ },
+ {
+ "epoch": 6.492012779552716,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0717,
+ "step": 2032
+ },
+ {
+ "epoch": 6.49520766773163,
+ "grad_norm": 0.050537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0823,
+ "step": 2033
+ },
+ {
+ "epoch": 6.498402555910543,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0725,
+ "step": 2034
+ },
+ {
+ "epoch": 6.501597444089457,
+ "grad_norm": 0.05859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0863,
+ "step": 2035
+ },
+ {
+ "epoch": 6.50479233226837,
+ "grad_norm": 0.08935546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0829,
+ "step": 2036
+ },
+ {
+ "epoch": 6.507987220447284,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0844,
+ "step": 2037
+ },
+ {
+ "epoch": 6.511182108626198,
+ "grad_norm": 0.11669921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0793,
+ "step": 2038
+ },
+ {
+ "epoch": 6.514376996805112,
+ "grad_norm": 0.06005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0805,
+ "step": 2039
+ },
+ {
+ "epoch": 6.517571884984026,
+ "grad_norm": 0.1025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0936,
+ "step": 2040
+ },
+ {
+ "epoch": 6.520766773162939,
+ "grad_norm": 0.09326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0745,
+ "step": 2041
+ },
+ {
+ "epoch": 6.523961661341853,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0722,
+ "step": 2042
+ },
+ {
+ "epoch": 6.527156549520766,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0621,
+ "step": 2043
+ },
+ {
+ "epoch": 6.5303514376996805,
+ "grad_norm": 0.09619140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0849,
+ "step": 2044
+ },
+ {
+ "epoch": 6.533546325878595,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0836,
+ "step": 2045
+ },
+ {
+ "epoch": 6.536741214057508,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0781,
+ "step": 2046
+ },
+ {
+ "epoch": 6.539936102236422,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0837,
+ "step": 2047
+ },
+ {
+ "epoch": 6.543130990415335,
+ "grad_norm": 0.09423828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0807,
+ "step": 2048
+ },
+ {
+ "epoch": 6.546325878594249,
+ "grad_norm": 0.08740234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0753,
+ "step": 2049
+ },
+ {
+ "epoch": 6.549520766773163,
+ "grad_norm": 0.07861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0714,
+ "step": 2050
+ },
+ {
+ "epoch": 6.552715654952077,
+ "grad_norm": 0.11669921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0888,
+ "step": 2051
+ },
+ {
+ "epoch": 6.555910543130991,
+ "grad_norm": 0.061279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0759,
+ "step": 2052
+ },
+ {
+ "epoch": 6.559105431309904,
+ "grad_norm": 0.1025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0659,
+ "step": 2053
+ },
+ {
+ "epoch": 6.562300319488818,
+ "grad_norm": 0.10986328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0665,
+ "step": 2054
+ },
+ {
+ "epoch": 6.565495207667731,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0807,
+ "step": 2055
+ },
+ {
+ "epoch": 6.568690095846645,
+ "grad_norm": 0.12158203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0731,
+ "step": 2056
+ },
+ {
+ "epoch": 6.571884984025559,
+ "grad_norm": 0.0859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0759,
+ "step": 2057
+ },
+ {
+ "epoch": 6.575079872204473,
+ "grad_norm": 0.1025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0838,
+ "step": 2058
+ },
+ {
+ "epoch": 6.578274760383387,
+ "grad_norm": 0.1669921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0772,
+ "step": 2059
+ },
+ {
+ "epoch": 6.5814696485623,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0827,
+ "step": 2060
+ },
+ {
+ "epoch": 6.584664536741214,
+ "grad_norm": 0.09619140625,
+ "learning_rate": 0.0005,
+ "loss": 1.071,
+ "step": 2061
+ },
+ {
+ "epoch": 6.587859424920127,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0659,
+ "step": 2062
+ },
+ {
+ "epoch": 6.5910543130990416,
+ "grad_norm": 0.09716796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0751,
+ "step": 2063
+ },
+ {
+ "epoch": 6.594249201277956,
+ "grad_norm": 0.08837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.081,
+ "step": 2064
+ },
+ {
+ "epoch": 6.597444089456869,
+ "grad_norm": 0.08837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0781,
+ "step": 2065
+ },
+ {
+ "epoch": 6.600638977635783,
+ "grad_norm": 0.080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0732,
+ "step": 2066
+ },
+ {
+ "epoch": 6.603833865814696,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0747,
+ "step": 2067
+ },
+ {
+ "epoch": 6.60702875399361,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0813,
+ "step": 2068
+ },
+ {
+ "epoch": 6.6102236421725244,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0873,
+ "step": 2069
+ },
+ {
+ "epoch": 6.613418530351438,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0926,
+ "step": 2070
+ },
+ {
+ "epoch": 6.616613418530352,
+ "grad_norm": 0.103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0879,
+ "step": 2071
+ },
+ {
+ "epoch": 6.619808306709265,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0795,
+ "step": 2072
+ },
+ {
+ "epoch": 6.623003194888179,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0718,
+ "step": 2073
+ },
+ {
+ "epoch": 6.626198083067092,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0635,
+ "step": 2074
+ },
+ {
+ "epoch": 6.6293929712460065,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.078,
+ "step": 2075
+ },
+ {
+ "epoch": 6.63258785942492,
+ "grad_norm": 0.103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0723,
+ "step": 2076
+ },
+ {
+ "epoch": 6.635782747603834,
+ "grad_norm": 0.060791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0762,
+ "step": 2077
+ },
+ {
+ "epoch": 6.638977635782748,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0714,
+ "step": 2078
+ },
+ {
+ "epoch": 6.642172523961661,
+ "grad_norm": 0.056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0779,
+ "step": 2079
+ },
+ {
+ "epoch": 6.645367412140575,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.074,
+ "step": 2080
+ },
+ {
+ "epoch": 6.6485623003194885,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0718,
+ "step": 2081
+ },
+ {
+ "epoch": 6.651757188498403,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0824,
+ "step": 2082
+ },
+ {
+ "epoch": 6.654952076677317,
+ "grad_norm": 0.08447265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0796,
+ "step": 2083
+ },
+ {
+ "epoch": 6.65814696485623,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0902,
+ "step": 2084
+ },
+ {
+ "epoch": 6.661341853035144,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0734,
+ "step": 2085
+ },
+ {
+ "epoch": 6.664536741214057,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0535,
+ "step": 2086
+ },
+ {
+ "epoch": 6.667731629392971,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0668,
+ "step": 2087
+ },
+ {
+ "epoch": 6.6709265175718855,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.073,
+ "step": 2088
+ },
+ {
+ "epoch": 6.674121405750799,
+ "grad_norm": 0.05908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0728,
+ "step": 2089
+ },
+ {
+ "epoch": 6.677316293929713,
+ "grad_norm": 0.061279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0689,
+ "step": 2090
+ },
+ {
+ "epoch": 6.680511182108626,
+ "grad_norm": 0.0810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0638,
+ "step": 2091
+ },
+ {
+ "epoch": 6.68370607028754,
+ "grad_norm": 0.060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0724,
+ "step": 2092
+ },
+ {
+ "epoch": 6.686900958466453,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0824,
+ "step": 2093
+ },
+ {
+ "epoch": 6.6900958466453675,
+ "grad_norm": 0.056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0745,
+ "step": 2094
+ },
+ {
+ "epoch": 6.693290734824281,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0787,
+ "step": 2095
+ },
+ {
+ "epoch": 6.696485623003195,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.065,
+ "step": 2096
+ },
+ {
+ "epoch": 6.699680511182109,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0554,
+ "step": 2097
+ },
+ {
+ "epoch": 6.702875399361022,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0752,
+ "step": 2098
+ },
+ {
+ "epoch": 6.706070287539936,
+ "grad_norm": 0.05712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0745,
+ "step": 2099
+ },
+ {
+ "epoch": 6.7092651757188495,
+ "grad_norm": 0.0615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0733,
+ "step": 2100
+ },
+ {
+ "epoch": 6.712460063897764,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0782,
+ "step": 2101
+ },
+ {
+ "epoch": 6.715654952076678,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0749,
+ "step": 2102
+ },
+ {
+ "epoch": 6.718849840255591,
+ "grad_norm": 0.08251953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0788,
+ "step": 2103
+ },
+ {
+ "epoch": 6.722044728434505,
+ "grad_norm": 0.055419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0744,
+ "step": 2104
+ },
+ {
+ "epoch": 6.725239616613418,
+ "grad_norm": 0.05712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0692,
+ "step": 2105
+ },
+ {
+ "epoch": 6.728434504792332,
+ "grad_norm": 0.0615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0776,
+ "step": 2106
+ },
+ {
+ "epoch": 6.731629392971246,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0753,
+ "step": 2107
+ },
+ {
+ "epoch": 6.73482428115016,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0664,
+ "step": 2108
+ },
+ {
+ "epoch": 6.738019169329074,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0638,
+ "step": 2109
+ },
+ {
+ "epoch": 6.741214057507987,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0675,
+ "step": 2110
+ },
+ {
+ "epoch": 6.744408945686901,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.066,
+ "step": 2111
+ },
+ {
+ "epoch": 6.747603833865814,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0781,
+ "step": 2112
+ },
+ {
+ "epoch": 6.7507987220447285,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0675,
+ "step": 2113
+ },
+ {
+ "epoch": 6.753993610223642,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0715,
+ "step": 2114
+ },
+ {
+ "epoch": 6.757188498402556,
+ "grad_norm": 0.058837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0614,
+ "step": 2115
+ },
+ {
+ "epoch": 6.76038338658147,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0629,
+ "step": 2116
+ },
+ {
+ "epoch": 6.763578274760383,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0728,
+ "step": 2117
+ },
+ {
+ "epoch": 6.766773162939297,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0676,
+ "step": 2118
+ },
+ {
+ "epoch": 6.7699680511182105,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0621,
+ "step": 2119
+ },
+ {
+ "epoch": 6.773162939297125,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0651,
+ "step": 2120
+ },
+ {
+ "epoch": 6.776357827476039,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0783,
+ "step": 2121
+ },
+ {
+ "epoch": 6.779552715654952,
+ "grad_norm": 0.05615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0641,
+ "step": 2122
+ },
+ {
+ "epoch": 6.782747603833866,
+ "grad_norm": 0.10400390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0643,
+ "step": 2123
+ },
+ {
+ "epoch": 6.785942492012779,
+ "grad_norm": 0.0576171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0589,
+ "step": 2124
+ },
+ {
+ "epoch": 6.789137380191693,
+ "grad_norm": 0.10400390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0775,
+ "step": 2125
+ },
+ {
+ "epoch": 6.792332268370607,
+ "grad_norm": 0.08251953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0638,
+ "step": 2126
+ },
+ {
+ "epoch": 6.795527156549521,
+ "grad_norm": 0.1083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0734,
+ "step": 2127
+ },
+ {
+ "epoch": 6.798722044728435,
+ "grad_norm": 0.0986328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0791,
+ "step": 2128
+ },
+ {
+ "epoch": 6.801916932907348,
+ "grad_norm": 0.08203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0673,
+ "step": 2129
+ },
+ {
+ "epoch": 6.805111821086262,
+ "grad_norm": 0.11572265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0701,
+ "step": 2130
+ },
+ {
+ "epoch": 6.8083067092651754,
+ "grad_norm": 0.154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0852,
+ "step": 2131
+ },
+ {
+ "epoch": 6.81150159744409,
+ "grad_norm": 0.1162109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0668,
+ "step": 2132
+ },
+ {
+ "epoch": 6.814696485623003,
+ "grad_norm": 0.1572265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0623,
+ "step": 2133
+ },
+ {
+ "epoch": 6.817891373801917,
+ "grad_norm": 0.1005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0831,
+ "step": 2134
+ },
+ {
+ "epoch": 6.821086261980831,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0757,
+ "step": 2135
+ },
+ {
+ "epoch": 6.824281150159744,
+ "grad_norm": 0.103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0734,
+ "step": 2136
+ },
+ {
+ "epoch": 6.827476038338658,
+ "grad_norm": 0.10888671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0662,
+ "step": 2137
+ },
+ {
+ "epoch": 6.830670926517572,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0917,
+ "step": 2138
+ },
+ {
+ "epoch": 6.833865814696486,
+ "grad_norm": 0.09033203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0691,
+ "step": 2139
+ },
+ {
+ "epoch": 6.8370607028754,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0676,
+ "step": 2140
+ },
+ {
+ "epoch": 6.840255591054313,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0716,
+ "step": 2141
+ },
+ {
+ "epoch": 6.843450479233227,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0615,
+ "step": 2142
+ },
+ {
+ "epoch": 6.84664536741214,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0772,
+ "step": 2143
+ },
+ {
+ "epoch": 6.8498402555910545,
+ "grad_norm": 0.09228515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0553,
+ "step": 2144
+ },
+ {
+ "epoch": 6.853035143769968,
+ "grad_norm": 0.11767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.066,
+ "step": 2145
+ },
+ {
+ "epoch": 6.856230031948882,
+ "grad_norm": 0.08642578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0757,
+ "step": 2146
+ },
+ {
+ "epoch": 6.859424920127795,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0704,
+ "step": 2147
+ },
+ {
+ "epoch": 6.862619808306709,
+ "grad_norm": 0.1279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0711,
+ "step": 2148
+ },
+ {
+ "epoch": 6.865814696485623,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0889,
+ "step": 2149
+ },
+ {
+ "epoch": 6.8690095846645365,
+ "grad_norm": 0.09326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0893,
+ "step": 2150
+ },
+ {
+ "epoch": 6.872204472843451,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.056,
+ "step": 2151
+ },
+ {
+ "epoch": 6.875399361022364,
+ "grad_norm": 0.08740234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0779,
+ "step": 2152
+ },
+ {
+ "epoch": 6.878594249201278,
+ "grad_norm": 0.0888671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0696,
+ "step": 2153
+ },
+ {
+ "epoch": 6.881789137380192,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0736,
+ "step": 2154
+ },
+ {
+ "epoch": 6.884984025559105,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0779,
+ "step": 2155
+ },
+ {
+ "epoch": 6.888178913738019,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0606,
+ "step": 2156
+ },
+ {
+ "epoch": 6.891373801916933,
+ "grad_norm": 0.08203125,
+ "learning_rate": 0.0005,
+ "loss": 1.056,
+ "step": 2157
+ },
+ {
+ "epoch": 6.894568690095847,
+ "grad_norm": 0.0869140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0666,
+ "step": 2158
+ },
+ {
+ "epoch": 6.897763578274761,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0674,
+ "step": 2159
+ },
+ {
+ "epoch": 6.900958466453674,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0715,
+ "step": 2160
+ },
+ {
+ "epoch": 6.904153354632588,
+ "grad_norm": 0.06201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0732,
+ "step": 2161
+ },
+ {
+ "epoch": 6.907348242811501,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0703,
+ "step": 2162
+ },
+ {
+ "epoch": 6.9105431309904155,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0711,
+ "step": 2163
+ },
+ {
+ "epoch": 6.913738019169329,
+ "grad_norm": 0.10107421875,
+ "learning_rate": 0.0005,
+ "loss": 1.063,
+ "step": 2164
+ },
+ {
+ "epoch": 6.916932907348243,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0808,
+ "step": 2165
+ },
+ {
+ "epoch": 6.920127795527156,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0795,
+ "step": 2166
+ },
+ {
+ "epoch": 6.92332268370607,
+ "grad_norm": 0.08642578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0749,
+ "step": 2167
+ },
+ {
+ "epoch": 6.926517571884984,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0781,
+ "step": 2168
+ },
+ {
+ "epoch": 6.9297124600638975,
+ "grad_norm": 0.0615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0848,
+ "step": 2169
+ },
+ {
+ "epoch": 6.932907348242812,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0577,
+ "step": 2170
+ },
+ {
+ "epoch": 6.936102236421725,
+ "grad_norm": 0.062255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.078,
+ "step": 2171
+ },
+ {
+ "epoch": 6.939297124600639,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0725,
+ "step": 2172
+ },
+ {
+ "epoch": 6.942492012779553,
+ "grad_norm": 0.057373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0814,
+ "step": 2173
+ },
+ {
+ "epoch": 6.945686900958466,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.071,
+ "step": 2174
+ },
+ {
+ "epoch": 6.94888178913738,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.084,
+ "step": 2175
+ },
+ {
+ "epoch": 6.952076677316294,
+ "grad_norm": 0.08203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0612,
+ "step": 2176
+ },
+ {
+ "epoch": 6.955271565495208,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0736,
+ "step": 2177
+ },
+ {
+ "epoch": 6.958466453674122,
+ "grad_norm": 0.058837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0722,
+ "step": 2178
+ },
+ {
+ "epoch": 6.961661341853035,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0657,
+ "step": 2179
+ },
+ {
+ "epoch": 6.964856230031949,
+ "grad_norm": 0.057861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.068,
+ "step": 2180
+ },
+ {
+ "epoch": 6.968051118210862,
+ "grad_norm": 0.08154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0756,
+ "step": 2181
+ },
+ {
+ "epoch": 6.9712460063897765,
+ "grad_norm": 0.062255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0773,
+ "step": 2182
+ },
+ {
+ "epoch": 6.97444089456869,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0773,
+ "step": 2183
+ },
+ {
+ "epoch": 6.977635782747604,
+ "grad_norm": 0.0615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0724,
+ "step": 2184
+ },
+ {
+ "epoch": 6.980830670926517,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.073,
+ "step": 2185
+ },
+ {
+ "epoch": 6.984025559105431,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0792,
+ "step": 2186
+ },
+ {
+ "epoch": 6.987220447284345,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0758,
+ "step": 2187
+ },
+ {
+ "epoch": 6.9904153354632586,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0705,
+ "step": 2188
+ },
+ {
+ "epoch": 6.993610223642173,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0683,
+ "step": 2189
+ },
+ {
+ "epoch": 6.996805111821086,
+ "grad_norm": 0.060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0725,
+ "step": 2190
+ },
+ {
+ "epoch": 7.0,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0839,
+ "step": 2191
+ },
+ {
+ "epoch": 7.003194888178914,
+ "grad_norm": 0.0595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0782,
+ "step": 2192
+ },
+ {
+ "epoch": 7.006389776357827,
+ "grad_norm": 0.05859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0757,
+ "step": 2193
+ },
+ {
+ "epoch": 7.0095846645367414,
+ "grad_norm": 0.10595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.088,
+ "step": 2194
+ },
+ {
+ "epoch": 7.012779552715655,
+ "grad_norm": 0.1328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0796,
+ "step": 2195
+ },
+ {
+ "epoch": 7.015974440894569,
+ "grad_norm": 0.0810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.066,
+ "step": 2196
+ },
+ {
+ "epoch": 7.019169329073482,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0661,
+ "step": 2197
+ },
+ {
+ "epoch": 7.022364217252396,
+ "grad_norm": 0.07861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0702,
+ "step": 2198
+ },
+ {
+ "epoch": 7.02555910543131,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0588,
+ "step": 2199
+ },
+ {
+ "epoch": 7.0287539936102235,
+ "grad_norm": 0.1298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0789,
+ "step": 2200
+ },
+ {
+ "epoch": 7.031948881789138,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0952,
+ "step": 2201
+ },
+ {
+ "epoch": 7.035143769968051,
+ "grad_norm": 0.08056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0674,
+ "step": 2202
+ },
+ {
+ "epoch": 7.038338658146965,
+ "grad_norm": 0.0595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0539,
+ "step": 2203
+ },
+ {
+ "epoch": 7.041533546325879,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0618,
+ "step": 2204
+ },
+ {
+ "epoch": 7.044728434504792,
+ "grad_norm": 0.06103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0684,
+ "step": 2205
+ },
+ {
+ "epoch": 7.047923322683706,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0625,
+ "step": 2206
+ },
+ {
+ "epoch": 7.05111821086262,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0832,
+ "step": 2207
+ },
+ {
+ "epoch": 7.054313099041534,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0816,
+ "step": 2208
+ },
+ {
+ "epoch": 7.057507987220447,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0558,
+ "step": 2209
+ },
+ {
+ "epoch": 7.060702875399361,
+ "grad_norm": 0.06005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0668,
+ "step": 2210
+ },
+ {
+ "epoch": 7.063897763578275,
+ "grad_norm": 0.062255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.07,
+ "step": 2211
+ },
+ {
+ "epoch": 7.067092651757188,
+ "grad_norm": 0.060302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0841,
+ "step": 2212
+ },
+ {
+ "epoch": 7.0702875399361025,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0714,
+ "step": 2213
+ },
+ {
+ "epoch": 7.073482428115016,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0673,
+ "step": 2214
+ },
+ {
+ "epoch": 7.07667731629393,
+ "grad_norm": 0.1142578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0695,
+ "step": 2215
+ },
+ {
+ "epoch": 7.079872204472843,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.071,
+ "step": 2216
+ },
+ {
+ "epoch": 7.083067092651757,
+ "grad_norm": 0.10595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0701,
+ "step": 2217
+ },
+ {
+ "epoch": 7.086261980830671,
+ "grad_norm": 0.06103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0821,
+ "step": 2218
+ },
+ {
+ "epoch": 7.0894568690095845,
+ "grad_norm": 0.083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0905,
+ "step": 2219
+ },
+ {
+ "epoch": 7.092651757188499,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0721,
+ "step": 2220
+ },
+ {
+ "epoch": 7.095846645367412,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0745,
+ "step": 2221
+ },
+ {
+ "epoch": 7.099041533546326,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0727,
+ "step": 2222
+ },
+ {
+ "epoch": 7.102236421725239,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0811,
+ "step": 2223
+ },
+ {
+ "epoch": 7.105431309904153,
+ "grad_norm": 0.08740234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0828,
+ "step": 2224
+ },
+ {
+ "epoch": 7.108626198083067,
+ "grad_norm": 0.08984375,
+ "learning_rate": 0.0005,
+ "loss": 1.08,
+ "step": 2225
+ },
+ {
+ "epoch": 7.111821086261981,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0625,
+ "step": 2226
+ },
+ {
+ "epoch": 7.115015974440895,
+ "grad_norm": 0.1376953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0815,
+ "step": 2227
+ },
+ {
+ "epoch": 7.118210862619808,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0634,
+ "step": 2228
+ },
+ {
+ "epoch": 7.121405750798722,
+ "grad_norm": 0.099609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0634,
+ "step": 2229
+ },
+ {
+ "epoch": 7.124600638977636,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0863,
+ "step": 2230
+ },
+ {
+ "epoch": 7.127795527156549,
+ "grad_norm": 0.06005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0747,
+ "step": 2231
+ },
+ {
+ "epoch": 7.1309904153354635,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0609,
+ "step": 2232
+ },
+ {
+ "epoch": 7.134185303514377,
+ "grad_norm": 0.05712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0634,
+ "step": 2233
+ },
+ {
+ "epoch": 7.137380191693291,
+ "grad_norm": 0.060791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0684,
+ "step": 2234
+ },
+ {
+ "epoch": 7.140575079872204,
+ "grad_norm": 0.10400390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0613,
+ "step": 2235
+ },
+ {
+ "epoch": 7.143769968051118,
+ "grad_norm": 0.057861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0613,
+ "step": 2236
+ },
+ {
+ "epoch": 7.146964856230032,
+ "grad_norm": 0.08642578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0746,
+ "step": 2237
+ },
+ {
+ "epoch": 7.1501597444089455,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0594,
+ "step": 2238
+ },
+ {
+ "epoch": 7.15335463258786,
+ "grad_norm": 0.1171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0699,
+ "step": 2239
+ },
+ {
+ "epoch": 7.156549520766773,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0758,
+ "step": 2240
+ },
+ {
+ "epoch": 7.159744408945687,
+ "grad_norm": 0.0947265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0673,
+ "step": 2241
+ },
+ {
+ "epoch": 7.1629392971246,
+ "grad_norm": 0.1240234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0611,
+ "step": 2242
+ },
+ {
+ "epoch": 7.166134185303514,
+ "grad_norm": 0.08642578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0666,
+ "step": 2243
+ },
+ {
+ "epoch": 7.169329073482428,
+ "grad_norm": 0.087890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0616,
+ "step": 2244
+ },
+ {
+ "epoch": 7.172523961661342,
+ "grad_norm": 0.08203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0717,
+ "step": 2245
+ },
+ {
+ "epoch": 7.175718849840256,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0625,
+ "step": 2246
+ },
+ {
+ "epoch": 7.178913738019169,
+ "grad_norm": 0.125,
+ "learning_rate": 0.0005,
+ "loss": 1.0638,
+ "step": 2247
+ },
+ {
+ "epoch": 7.182108626198083,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0758,
+ "step": 2248
+ },
+ {
+ "epoch": 7.185303514376997,
+ "grad_norm": 0.1357421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0637,
+ "step": 2249
+ },
+ {
+ "epoch": 7.18849840255591,
+ "grad_norm": 0.13671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0725,
+ "step": 2250
+ },
+ {
+ "epoch": 7.1916932907348246,
+ "grad_norm": 0.09033203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0621,
+ "step": 2251
+ },
+ {
+ "epoch": 7.194888178913738,
+ "grad_norm": 0.193359375,
+ "learning_rate": 0.0005,
+ "loss": 1.078,
+ "step": 2252
+ },
+ {
+ "epoch": 7.198083067092652,
+ "grad_norm": 0.109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0652,
+ "step": 2253
+ },
+ {
+ "epoch": 7.201277955271565,
+ "grad_norm": 0.1298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.067,
+ "step": 2254
+ },
+ {
+ "epoch": 7.204472843450479,
+ "grad_norm": 0.17578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0848,
+ "step": 2255
+ },
+ {
+ "epoch": 7.207667731629393,
+ "grad_norm": 0.09521484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0735,
+ "step": 2256
+ },
+ {
+ "epoch": 7.210862619808307,
+ "grad_norm": 0.10888671875,
+ "learning_rate": 0.0005,
+ "loss": 1.081,
+ "step": 2257
+ },
+ {
+ "epoch": 7.214057507987221,
+ "grad_norm": 0.17578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0828,
+ "step": 2258
+ },
+ {
+ "epoch": 7.217252396166134,
+ "grad_norm": 0.08642578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0677,
+ "step": 2259
+ },
+ {
+ "epoch": 7.220447284345048,
+ "grad_norm": 0.173828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0729,
+ "step": 2260
+ },
+ {
+ "epoch": 7.223642172523961,
+ "grad_norm": 0.2109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0545,
+ "step": 2261
+ },
+ {
+ "epoch": 7.226837060702875,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0708,
+ "step": 2262
+ },
+ {
+ "epoch": 7.2300319488817895,
+ "grad_norm": 0.185546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0674,
+ "step": 2263
+ },
+ {
+ "epoch": 7.233226837060703,
+ "grad_norm": 0.1337890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0768,
+ "step": 2264
+ },
+ {
+ "epoch": 7.236421725239617,
+ "grad_norm": 0.061279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0682,
+ "step": 2265
+ },
+ {
+ "epoch": 7.23961661341853,
+ "grad_norm": 0.1298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0783,
+ "step": 2266
+ },
+ {
+ "epoch": 7.242811501597444,
+ "grad_norm": 0.1484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0767,
+ "step": 2267
+ },
+ {
+ "epoch": 7.246006389776358,
+ "grad_norm": 0.10302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0697,
+ "step": 2268
+ },
+ {
+ "epoch": 7.2492012779552715,
+ "grad_norm": 0.27734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0788,
+ "step": 2269
+ },
+ {
+ "epoch": 7.252396166134186,
+ "grad_norm": 0.171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0663,
+ "step": 2270
+ },
+ {
+ "epoch": 7.255591054313099,
+ "grad_norm": 0.0986328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0641,
+ "step": 2271
+ },
+ {
+ "epoch": 7.258785942492013,
+ "grad_norm": 0.1474609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0504,
+ "step": 2272
+ },
+ {
+ "epoch": 7.261980830670926,
+ "grad_norm": 0.0888671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0677,
+ "step": 2273
+ },
+ {
+ "epoch": 7.26517571884984,
+ "grad_norm": 0.1728515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0684,
+ "step": 2274
+ },
+ {
+ "epoch": 7.268370607028754,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0809,
+ "step": 2275
+ },
+ {
+ "epoch": 7.271565495207668,
+ "grad_norm": 0.09521484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0772,
+ "step": 2276
+ },
+ {
+ "epoch": 7.274760383386582,
+ "grad_norm": 0.08349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0709,
+ "step": 2277
+ },
+ {
+ "epoch": 7.277955271565495,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0687,
+ "step": 2278
+ },
+ {
+ "epoch": 7.281150159744409,
+ "grad_norm": 0.09912109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0789,
+ "step": 2279
+ },
+ {
+ "epoch": 7.284345047923322,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0675,
+ "step": 2280
+ },
+ {
+ "epoch": 7.287539936102236,
+ "grad_norm": 0.08642578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0813,
+ "step": 2281
+ },
+ {
+ "epoch": 7.2907348242811505,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0716,
+ "step": 2282
+ },
+ {
+ "epoch": 7.293929712460064,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0835,
+ "step": 2283
+ },
+ {
+ "epoch": 7.297124600638978,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0495,
+ "step": 2284
+ },
+ {
+ "epoch": 7.300319488817891,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0854,
+ "step": 2285
+ },
+ {
+ "epoch": 7.303514376996805,
+ "grad_norm": 0.10693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0785,
+ "step": 2286
+ },
+ {
+ "epoch": 7.306709265175719,
+ "grad_norm": 0.05517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0652,
+ "step": 2287
+ },
+ {
+ "epoch": 7.3099041533546325,
+ "grad_norm": 0.09423828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0636,
+ "step": 2288
+ },
+ {
+ "epoch": 7.313099041533547,
+ "grad_norm": 0.0927734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0684,
+ "step": 2289
+ },
+ {
+ "epoch": 7.31629392971246,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0744,
+ "step": 2290
+ },
+ {
+ "epoch": 7.319488817891374,
+ "grad_norm": 0.11669921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0581,
+ "step": 2291
+ },
+ {
+ "epoch": 7.322683706070287,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0616,
+ "step": 2292
+ },
+ {
+ "epoch": 7.325878594249201,
+ "grad_norm": 0.12255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0678,
+ "step": 2293
+ },
+ {
+ "epoch": 7.329073482428115,
+ "grad_norm": 0.08251953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0714,
+ "step": 2294
+ },
+ {
+ "epoch": 7.332268370607029,
+ "grad_norm": 0.08203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0824,
+ "step": 2295
+ },
+ {
+ "epoch": 7.335463258785943,
+ "grad_norm": 0.09765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0689,
+ "step": 2296
+ },
+ {
+ "epoch": 7.338658146964856,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0826,
+ "step": 2297
+ },
+ {
+ "epoch": 7.34185303514377,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0686,
+ "step": 2298
+ },
+ {
+ "epoch": 7.345047923322683,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0568,
+ "step": 2299
+ },
+ {
+ "epoch": 7.348242811501597,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0694,
+ "step": 2300
+ },
+ {
+ "epoch": 7.3514376996805115,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0721,
+ "step": 2301
+ },
+ {
+ "epoch": 7.354632587859425,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0799,
+ "step": 2302
+ },
+ {
+ "epoch": 7.357827476038339,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0757,
+ "step": 2303
+ },
+ {
+ "epoch": 7.361022364217252,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0729,
+ "step": 2304
+ },
+ {
+ "epoch": 7.364217252396166,
+ "grad_norm": 0.09033203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0834,
+ "step": 2305
+ },
+ {
+ "epoch": 7.36741214057508,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0533,
+ "step": 2306
+ },
+ {
+ "epoch": 7.3706070287539935,
+ "grad_norm": 0.1259765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0632,
+ "step": 2307
+ },
+ {
+ "epoch": 7.373801916932908,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0769,
+ "step": 2308
+ },
+ {
+ "epoch": 7.376996805111821,
+ "grad_norm": 0.111328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0689,
+ "step": 2309
+ },
+ {
+ "epoch": 7.380191693290735,
+ "grad_norm": 0.08740234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0779,
+ "step": 2310
+ },
+ {
+ "epoch": 7.383386581469648,
+ "grad_norm": 0.08349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.084,
+ "step": 2311
+ },
+ {
+ "epoch": 7.386581469648562,
+ "grad_norm": 0.10400390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0629,
+ "step": 2312
+ },
+ {
+ "epoch": 7.389776357827476,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0741,
+ "step": 2313
+ },
+ {
+ "epoch": 7.39297124600639,
+ "grad_norm": 0.083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0618,
+ "step": 2314
+ },
+ {
+ "epoch": 7.396166134185304,
+ "grad_norm": 0.08740234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0729,
+ "step": 2315
+ },
+ {
+ "epoch": 7.399361022364217,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0648,
+ "step": 2316
+ },
+ {
+ "epoch": 7.402555910543131,
+ "grad_norm": 0.09033203125,
+ "learning_rate": 0.0005,
+ "loss": 1.07,
+ "step": 2317
+ },
+ {
+ "epoch": 7.405750798722044,
+ "grad_norm": 0.0888671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0881,
+ "step": 2318
+ },
+ {
+ "epoch": 7.4089456869009584,
+ "grad_norm": 0.08203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0736,
+ "step": 2319
+ },
+ {
+ "epoch": 7.412140575079873,
+ "grad_norm": 0.1025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0725,
+ "step": 2320
+ },
+ {
+ "epoch": 7.415335463258786,
+ "grad_norm": 0.0810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0778,
+ "step": 2321
+ },
+ {
+ "epoch": 7.4185303514377,
+ "grad_norm": 0.1015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0698,
+ "step": 2322
+ },
+ {
+ "epoch": 7.421725239616613,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0701,
+ "step": 2323
+ },
+ {
+ "epoch": 7.424920127795527,
+ "grad_norm": 0.09619140625,
+ "learning_rate": 0.0005,
+ "loss": 1.083,
+ "step": 2324
+ },
+ {
+ "epoch": 7.428115015974441,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0674,
+ "step": 2325
+ },
+ {
+ "epoch": 7.431309904153355,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0652,
+ "step": 2326
+ },
+ {
+ "epoch": 7.434504792332269,
+ "grad_norm": 0.052734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0572,
+ "step": 2327
+ },
+ {
+ "epoch": 7.437699680511182,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0662,
+ "step": 2328
+ },
+ {
+ "epoch": 7.440894568690096,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0755,
+ "step": 2329
+ },
+ {
+ "epoch": 7.444089456869009,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0749,
+ "step": 2330
+ },
+ {
+ "epoch": 7.447284345047923,
+ "grad_norm": 0.08447265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0844,
+ "step": 2331
+ },
+ {
+ "epoch": 7.4504792332268375,
+ "grad_norm": 0.11376953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0668,
+ "step": 2332
+ },
+ {
+ "epoch": 7.453674121405751,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0749,
+ "step": 2333
+ },
+ {
+ "epoch": 7.456869009584665,
+ "grad_norm": 0.0966796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0614,
+ "step": 2334
+ },
+ {
+ "epoch": 7.460063897763578,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0758,
+ "step": 2335
+ },
+ {
+ "epoch": 7.463258785942492,
+ "grad_norm": 0.08447265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0697,
+ "step": 2336
+ },
+ {
+ "epoch": 7.466453674121405,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0708,
+ "step": 2337
+ },
+ {
+ "epoch": 7.4696485623003195,
+ "grad_norm": 0.1162109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0682,
+ "step": 2338
+ },
+ {
+ "epoch": 7.472843450479234,
+ "grad_norm": 0.08154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.081,
+ "step": 2339
+ },
+ {
+ "epoch": 7.476038338658147,
+ "grad_norm": 0.099609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0602,
+ "step": 2340
+ },
+ {
+ "epoch": 7.479233226837061,
+ "grad_norm": 0.1259765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0735,
+ "step": 2341
+ },
+ {
+ "epoch": 7.482428115015974,
+ "grad_norm": 0.0888671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0696,
+ "step": 2342
+ },
+ {
+ "epoch": 7.485623003194888,
+ "grad_norm": 0.0859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0779,
+ "step": 2343
+ },
+ {
+ "epoch": 7.488817891373802,
+ "grad_norm": 0.0927734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0754,
+ "step": 2344
+ },
+ {
+ "epoch": 7.492012779552716,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0639,
+ "step": 2345
+ },
+ {
+ "epoch": 7.49520766773163,
+ "grad_norm": 0.09375,
+ "learning_rate": 0.0005,
+ "loss": 1.0587,
+ "step": 2346
+ },
+ {
+ "epoch": 7.498402555910543,
+ "grad_norm": 0.09130859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0715,
+ "step": 2347
+ },
+ {
+ "epoch": 7.501597444089457,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0688,
+ "step": 2348
+ },
+ {
+ "epoch": 7.50479233226837,
+ "grad_norm": 0.0810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0628,
+ "step": 2349
+ },
+ {
+ "epoch": 7.507987220447284,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0754,
+ "step": 2350
+ },
+ {
+ "epoch": 7.511182108626198,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0606,
+ "step": 2351
+ },
+ {
+ "epoch": 7.514376996805112,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0745,
+ "step": 2352
+ },
+ {
+ "epoch": 7.517571884984026,
+ "grad_norm": 0.0576171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0759,
+ "step": 2353
+ },
+ {
+ "epoch": 7.520766773162939,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0667,
+ "step": 2354
+ },
+ {
+ "epoch": 7.523961661341853,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0739,
+ "step": 2355
+ },
+ {
+ "epoch": 7.527156549520766,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0646,
+ "step": 2356
+ },
+ {
+ "epoch": 7.5303514376996805,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0727,
+ "step": 2357
+ },
+ {
+ "epoch": 7.533546325878595,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0612,
+ "step": 2358
+ },
+ {
+ "epoch": 7.536741214057508,
+ "grad_norm": 0.056884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0669,
+ "step": 2359
+ },
+ {
+ "epoch": 7.539936102236422,
+ "grad_norm": 0.0615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0504,
+ "step": 2360
+ },
+ {
+ "epoch": 7.543130990415335,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0789,
+ "step": 2361
+ },
+ {
+ "epoch": 7.546325878594249,
+ "grad_norm": 0.050537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0662,
+ "step": 2362
+ },
+ {
+ "epoch": 7.549520766773163,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0694,
+ "step": 2363
+ },
+ {
+ "epoch": 7.552715654952077,
+ "grad_norm": 0.056396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0727,
+ "step": 2364
+ },
+ {
+ "epoch": 7.555910543130991,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0811,
+ "step": 2365
+ },
+ {
+ "epoch": 7.559105431309904,
+ "grad_norm": 0.0576171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0695,
+ "step": 2366
+ },
+ {
+ "epoch": 7.562300319488818,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0701,
+ "step": 2367
+ },
+ {
+ "epoch": 7.565495207667731,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0719,
+ "step": 2368
+ },
+ {
+ "epoch": 7.568690095846645,
+ "grad_norm": 0.09423828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0684,
+ "step": 2369
+ },
+ {
+ "epoch": 7.571884984025559,
+ "grad_norm": 0.09130859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0712,
+ "step": 2370
+ },
+ {
+ "epoch": 7.575079872204473,
+ "grad_norm": 0.0830078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0732,
+ "step": 2371
+ },
+ {
+ "epoch": 7.578274760383387,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0696,
+ "step": 2372
+ },
+ {
+ "epoch": 7.5814696485623,
+ "grad_norm": 0.05712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0699,
+ "step": 2373
+ },
+ {
+ "epoch": 7.584664536741214,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0756,
+ "step": 2374
+ },
+ {
+ "epoch": 7.587859424920127,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0616,
+ "step": 2375
+ },
+ {
+ "epoch": 7.5910543130990416,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0612,
+ "step": 2376
+ },
+ {
+ "epoch": 7.594249201277956,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0749,
+ "step": 2377
+ },
+ {
+ "epoch": 7.597444089456869,
+ "grad_norm": 0.052978515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0691,
+ "step": 2378
+ },
+ {
+ "epoch": 7.600638977635783,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0752,
+ "step": 2379
+ },
+ {
+ "epoch": 7.603833865814696,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0565,
+ "step": 2380
+ },
+ {
+ "epoch": 7.60702875399361,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0718,
+ "step": 2381
+ },
+ {
+ "epoch": 7.6102236421725244,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0721,
+ "step": 2382
+ },
+ {
+ "epoch": 7.613418530351438,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0733,
+ "step": 2383
+ },
+ {
+ "epoch": 7.616613418530352,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0713,
+ "step": 2384
+ },
+ {
+ "epoch": 7.619808306709265,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0603,
+ "step": 2385
+ },
+ {
+ "epoch": 7.623003194888179,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0838,
+ "step": 2386
+ },
+ {
+ "epoch": 7.626198083067092,
+ "grad_norm": 0.140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0783,
+ "step": 2387
+ },
+ {
+ "epoch": 7.6293929712460065,
+ "grad_norm": 0.08251953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0705,
+ "step": 2388
+ },
+ {
+ "epoch": 7.63258785942492,
+ "grad_norm": 0.2138671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0554,
+ "step": 2389
+ },
+ {
+ "epoch": 7.635782747603834,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0926,
+ "step": 2390
+ },
+ {
+ "epoch": 7.638977635782748,
+ "grad_norm": 0.1806640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0705,
+ "step": 2391
+ },
+ {
+ "epoch": 7.642172523961661,
+ "grad_norm": 0.1845703125,
+ "learning_rate": 0.0005,
+ "loss": 1.077,
+ "step": 2392
+ },
+ {
+ "epoch": 7.645367412140575,
+ "grad_norm": 0.1708984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0639,
+ "step": 2393
+ },
+ {
+ "epoch": 7.6485623003194885,
+ "grad_norm": 0.10546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0563,
+ "step": 2394
+ },
+ {
+ "epoch": 7.651757188498403,
+ "grad_norm": 0.1904296875,
+ "learning_rate": 0.0005,
+ "loss": 1.067,
+ "step": 2395
+ },
+ {
+ "epoch": 7.654952076677317,
+ "grad_norm": 0.193359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0752,
+ "step": 2396
+ },
+ {
+ "epoch": 7.65814696485623,
+ "grad_norm": 0.1298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0759,
+ "step": 2397
+ },
+ {
+ "epoch": 7.661341853035144,
+ "grad_norm": 0.12890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0678,
+ "step": 2398
+ },
+ {
+ "epoch": 7.664536741214057,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0801,
+ "step": 2399
+ },
+ {
+ "epoch": 7.667731629392971,
+ "grad_norm": 0.10693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0743,
+ "step": 2400
+ },
+ {
+ "epoch": 7.6709265175718855,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0703,
+ "step": 2401
+ },
+ {
+ "epoch": 7.674121405750799,
+ "grad_norm": 0.1142578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0578,
+ "step": 2402
+ },
+ {
+ "epoch": 7.677316293929713,
+ "grad_norm": 0.08251953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0791,
+ "step": 2403
+ },
+ {
+ "epoch": 7.680511182108626,
+ "grad_norm": 0.08642578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0725,
+ "step": 2404
+ },
+ {
+ "epoch": 7.68370607028754,
+ "grad_norm": 0.060302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0766,
+ "step": 2405
+ },
+ {
+ "epoch": 7.686900958466453,
+ "grad_norm": 0.08154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0562,
+ "step": 2406
+ },
+ {
+ "epoch": 7.6900958466453675,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0626,
+ "step": 2407
+ },
+ {
+ "epoch": 7.693290734824281,
+ "grad_norm": 0.054931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0593,
+ "step": 2408
+ },
+ {
+ "epoch": 7.696485623003195,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0788,
+ "step": 2409
+ },
+ {
+ "epoch": 7.699680511182109,
+ "grad_norm": 0.0849609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0666,
+ "step": 2410
+ },
+ {
+ "epoch": 7.702875399361022,
+ "grad_norm": 0.08642578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0738,
+ "step": 2411
+ },
+ {
+ "epoch": 7.706070287539936,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0627,
+ "step": 2412
+ },
+ {
+ "epoch": 7.7092651757188495,
+ "grad_norm": 0.09033203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0641,
+ "step": 2413
+ },
+ {
+ "epoch": 7.712460063897764,
+ "grad_norm": 0.09130859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0632,
+ "step": 2414
+ },
+ {
+ "epoch": 7.715654952076678,
+ "grad_norm": 0.08984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0739,
+ "step": 2415
+ },
+ {
+ "epoch": 7.718849840255591,
+ "grad_norm": 0.177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0737,
+ "step": 2416
+ },
+ {
+ "epoch": 7.722044728434505,
+ "grad_norm": 0.080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0695,
+ "step": 2417
+ },
+ {
+ "epoch": 7.725239616613418,
+ "grad_norm": 0.10107421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0577,
+ "step": 2418
+ },
+ {
+ "epoch": 7.728434504792332,
+ "grad_norm": 0.09619140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0743,
+ "step": 2419
+ },
+ {
+ "epoch": 7.731629392971246,
+ "grad_norm": 0.091796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0607,
+ "step": 2420
+ },
+ {
+ "epoch": 7.73482428115016,
+ "grad_norm": 0.0810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0711,
+ "step": 2421
+ },
+ {
+ "epoch": 7.738019169329074,
+ "grad_norm": 0.1513671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0775,
+ "step": 2422
+ },
+ {
+ "epoch": 7.741214057507987,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0787,
+ "step": 2423
+ },
+ {
+ "epoch": 7.744408945686901,
+ "grad_norm": 0.12060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0833,
+ "step": 2424
+ },
+ {
+ "epoch": 7.747603833865814,
+ "grad_norm": 0.1044921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0464,
+ "step": 2425
+ },
+ {
+ "epoch": 7.7507987220447285,
+ "grad_norm": 0.10595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0748,
+ "step": 2426
+ },
+ {
+ "epoch": 7.753993610223642,
+ "grad_norm": 0.10205078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0646,
+ "step": 2427
+ },
+ {
+ "epoch": 7.757188498402556,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.065,
+ "step": 2428
+ },
+ {
+ "epoch": 7.76038338658147,
+ "grad_norm": 0.083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0868,
+ "step": 2429
+ },
+ {
+ "epoch": 7.763578274760383,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0685,
+ "step": 2430
+ },
+ {
+ "epoch": 7.766773162939297,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0648,
+ "step": 2431
+ },
+ {
+ "epoch": 7.7699680511182105,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0714,
+ "step": 2432
+ },
+ {
+ "epoch": 7.773162939297125,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0635,
+ "step": 2433
+ },
+ {
+ "epoch": 7.776357827476039,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0605,
+ "step": 2434
+ },
+ {
+ "epoch": 7.779552715654952,
+ "grad_norm": 0.08203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0802,
+ "step": 2435
+ },
+ {
+ "epoch": 7.782747603833866,
+ "grad_norm": 0.07861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0697,
+ "step": 2436
+ },
+ {
+ "epoch": 7.785942492012779,
+ "grad_norm": 0.06201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0644,
+ "step": 2437
+ },
+ {
+ "epoch": 7.789137380191693,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0618,
+ "step": 2438
+ },
+ {
+ "epoch": 7.792332268370607,
+ "grad_norm": 0.06005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0669,
+ "step": 2439
+ },
+ {
+ "epoch": 7.795527156549521,
+ "grad_norm": 0.061767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0725,
+ "step": 2440
+ },
+ {
+ "epoch": 7.798722044728435,
+ "grad_norm": 0.0576171875,
+ "learning_rate": 0.0005,
+ "loss": 1.06,
+ "step": 2441
+ },
+ {
+ "epoch": 7.801916932907348,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.071,
+ "step": 2442
+ },
+ {
+ "epoch": 7.805111821086262,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0741,
+ "step": 2443
+ },
+ {
+ "epoch": 7.8083067092651754,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0606,
+ "step": 2444
+ },
+ {
+ "epoch": 7.81150159744409,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0666,
+ "step": 2445
+ },
+ {
+ "epoch": 7.814696485623003,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0661,
+ "step": 2446
+ },
+ {
+ "epoch": 7.817891373801917,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0611,
+ "step": 2447
+ },
+ {
+ "epoch": 7.821086261980831,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.068,
+ "step": 2448
+ },
+ {
+ "epoch": 7.824281150159744,
+ "grad_norm": 0.087890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0668,
+ "step": 2449
+ },
+ {
+ "epoch": 7.827476038338658,
+ "grad_norm": 0.0859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0651,
+ "step": 2450
+ },
+ {
+ "epoch": 7.830670926517572,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0624,
+ "step": 2451
+ },
+ {
+ "epoch": 7.833865814696486,
+ "grad_norm": 0.0849609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0847,
+ "step": 2452
+ },
+ {
+ "epoch": 7.8370607028754,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0637,
+ "step": 2453
+ },
+ {
+ "epoch": 7.840255591054313,
+ "grad_norm": 0.08056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0754,
+ "step": 2454
+ },
+ {
+ "epoch": 7.843450479233227,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0715,
+ "step": 2455
+ },
+ {
+ "epoch": 7.84664536741214,
+ "grad_norm": 0.06201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0715,
+ "step": 2456
+ },
+ {
+ "epoch": 7.8498402555910545,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0616,
+ "step": 2457
+ },
+ {
+ "epoch": 7.853035143769968,
+ "grad_norm": 0.08203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0755,
+ "step": 2458
+ },
+ {
+ "epoch": 7.856230031948882,
+ "grad_norm": 0.09130859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0588,
+ "step": 2459
+ },
+ {
+ "epoch": 7.859424920127795,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0497,
+ "step": 2460
+ },
+ {
+ "epoch": 7.862619808306709,
+ "grad_norm": 0.08251953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0731,
+ "step": 2461
+ },
+ {
+ "epoch": 7.865814696485623,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0686,
+ "step": 2462
+ },
+ {
+ "epoch": 7.8690095846645365,
+ "grad_norm": 0.12255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0642,
+ "step": 2463
+ },
+ {
+ "epoch": 7.872204472843451,
+ "grad_norm": 0.1015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0684,
+ "step": 2464
+ },
+ {
+ "epoch": 7.875399361022364,
+ "grad_norm": 0.08984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0659,
+ "step": 2465
+ },
+ {
+ "epoch": 7.878594249201278,
+ "grad_norm": 0.1259765625,
+ "learning_rate": 0.0005,
+ "loss": 1.062,
+ "step": 2466
+ },
+ {
+ "epoch": 7.881789137380192,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0656,
+ "step": 2467
+ },
+ {
+ "epoch": 7.884984025559105,
+ "grad_norm": 0.1181640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0633,
+ "step": 2468
+ },
+ {
+ "epoch": 7.888178913738019,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0693,
+ "step": 2469
+ },
+ {
+ "epoch": 7.891373801916933,
+ "grad_norm": 0.125,
+ "learning_rate": 0.0005,
+ "loss": 1.0832,
+ "step": 2470
+ },
+ {
+ "epoch": 7.894568690095847,
+ "grad_norm": 0.0859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0677,
+ "step": 2471
+ },
+ {
+ "epoch": 7.897763578274761,
+ "grad_norm": 0.10546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0715,
+ "step": 2472
+ },
+ {
+ "epoch": 7.900958466453674,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0644,
+ "step": 2473
+ },
+ {
+ "epoch": 7.904153354632588,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0616,
+ "step": 2474
+ },
+ {
+ "epoch": 7.907348242811501,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0701,
+ "step": 2475
+ },
+ {
+ "epoch": 7.9105431309904155,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0548,
+ "step": 2476
+ },
+ {
+ "epoch": 7.913738019169329,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0576,
+ "step": 2477
+ },
+ {
+ "epoch": 7.916932907348243,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0675,
+ "step": 2478
+ },
+ {
+ "epoch": 7.920127795527156,
+ "grad_norm": 0.059326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0619,
+ "step": 2479
+ },
+ {
+ "epoch": 7.92332268370607,
+ "grad_norm": 0.07861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0577,
+ "step": 2480
+ },
+ {
+ "epoch": 7.926517571884984,
+ "grad_norm": 0.107421875,
+ "learning_rate": 0.0005,
+ "loss": 1.074,
+ "step": 2481
+ },
+ {
+ "epoch": 7.9297124600638975,
+ "grad_norm": 0.087890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0678,
+ "step": 2482
+ },
+ {
+ "epoch": 7.932907348242812,
+ "grad_norm": 0.0859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0711,
+ "step": 2483
+ },
+ {
+ "epoch": 7.936102236421725,
+ "grad_norm": 0.087890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0661,
+ "step": 2484
+ },
+ {
+ "epoch": 7.939297124600639,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0461,
+ "step": 2485
+ },
+ {
+ "epoch": 7.942492012779553,
+ "grad_norm": 0.08349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0666,
+ "step": 2486
+ },
+ {
+ "epoch": 7.945686900958466,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0622,
+ "step": 2487
+ },
+ {
+ "epoch": 7.94888178913738,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0763,
+ "step": 2488
+ },
+ {
+ "epoch": 7.952076677316294,
+ "grad_norm": 0.08154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0638,
+ "step": 2489
+ },
+ {
+ "epoch": 7.955271565495208,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0796,
+ "step": 2490
+ },
+ {
+ "epoch": 7.958466453674122,
+ "grad_norm": 0.1064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0655,
+ "step": 2491
+ },
+ {
+ "epoch": 7.961661341853035,
+ "grad_norm": 0.1337890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0742,
+ "step": 2492
+ },
+ {
+ "epoch": 7.964856230031949,
+ "grad_norm": 0.060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.07,
+ "step": 2493
+ },
+ {
+ "epoch": 7.968051118210862,
+ "grad_norm": 0.12255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0749,
+ "step": 2494
+ },
+ {
+ "epoch": 7.9712460063897765,
+ "grad_norm": 0.0810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0647,
+ "step": 2495
+ },
+ {
+ "epoch": 7.97444089456869,
+ "grad_norm": 0.056884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0829,
+ "step": 2496
+ },
+ {
+ "epoch": 7.977635782747604,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.068,
+ "step": 2497
+ },
+ {
+ "epoch": 7.980830670926517,
+ "grad_norm": 0.09326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0753,
+ "step": 2498
+ },
+ {
+ "epoch": 7.984025559105431,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.071,
+ "step": 2499
+ },
+ {
+ "epoch": 7.987220447284345,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0616,
+ "step": 2500
+ },
+ {
+ "epoch": 7.9904153354632586,
+ "grad_norm": 0.08642578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0772,
+ "step": 2501
+ },
+ {
+ "epoch": 7.993610223642173,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0642,
+ "step": 2502
+ },
+ {
+ "epoch": 7.996805111821086,
+ "grad_norm": 0.109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0586,
+ "step": 2503
+ },
+ {
+ "epoch": 8.0,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0703,
+ "step": 2504
+ },
+ {
+ "epoch": 8.003194888178914,
+ "grad_norm": 0.119140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0656,
+ "step": 2505
+ },
+ {
+ "epoch": 8.006389776357828,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0673,
+ "step": 2506
+ },
+ {
+ "epoch": 8.00958466453674,
+ "grad_norm": 0.10888671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0713,
+ "step": 2507
+ },
+ {
+ "epoch": 8.012779552715655,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0738,
+ "step": 2508
+ },
+ {
+ "epoch": 8.015974440894569,
+ "grad_norm": 0.0966796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0722,
+ "step": 2509
+ },
+ {
+ "epoch": 8.019169329073483,
+ "grad_norm": 0.08642578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0638,
+ "step": 2510
+ },
+ {
+ "epoch": 8.022364217252397,
+ "grad_norm": 0.059326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0691,
+ "step": 2511
+ },
+ {
+ "epoch": 8.02555910543131,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0667,
+ "step": 2512
+ },
+ {
+ "epoch": 8.028753993610223,
+ "grad_norm": 0.060791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0626,
+ "step": 2513
+ },
+ {
+ "epoch": 8.031948881789138,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0673,
+ "step": 2514
+ },
+ {
+ "epoch": 8.035143769968052,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0743,
+ "step": 2515
+ },
+ {
+ "epoch": 8.038338658146964,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0725,
+ "step": 2516
+ },
+ {
+ "epoch": 8.041533546325878,
+ "grad_norm": 0.0947265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0675,
+ "step": 2517
+ },
+ {
+ "epoch": 8.044728434504792,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0662,
+ "step": 2518
+ },
+ {
+ "epoch": 8.047923322683706,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.083,
+ "step": 2519
+ },
+ {
+ "epoch": 8.05111821086262,
+ "grad_norm": 0.0966796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0677,
+ "step": 2520
+ },
+ {
+ "epoch": 8.054313099041533,
+ "grad_norm": 0.09423828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0598,
+ "step": 2521
+ },
+ {
+ "epoch": 8.057507987220447,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0764,
+ "step": 2522
+ },
+ {
+ "epoch": 8.060702875399361,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0644,
+ "step": 2523
+ },
+ {
+ "epoch": 8.063897763578275,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0698,
+ "step": 2524
+ },
+ {
+ "epoch": 8.06709265175719,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0617,
+ "step": 2525
+ },
+ {
+ "epoch": 8.070287539936102,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.061,
+ "step": 2526
+ },
+ {
+ "epoch": 8.073482428115016,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0609,
+ "step": 2527
+ },
+ {
+ "epoch": 8.07667731629393,
+ "grad_norm": 0.10986328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0612,
+ "step": 2528
+ },
+ {
+ "epoch": 8.079872204472844,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0573,
+ "step": 2529
+ },
+ {
+ "epoch": 8.083067092651758,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.055,
+ "step": 2530
+ },
+ {
+ "epoch": 8.08626198083067,
+ "grad_norm": 0.130859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0635,
+ "step": 2531
+ },
+ {
+ "epoch": 8.089456869009584,
+ "grad_norm": 0.15234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0532,
+ "step": 2532
+ },
+ {
+ "epoch": 8.092651757188499,
+ "grad_norm": 0.08984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0638,
+ "step": 2533
+ },
+ {
+ "epoch": 8.095846645367413,
+ "grad_norm": 0.0908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0621,
+ "step": 2534
+ },
+ {
+ "epoch": 8.099041533546325,
+ "grad_norm": 0.083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0794,
+ "step": 2535
+ },
+ {
+ "epoch": 8.10223642172524,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0758,
+ "step": 2536
+ },
+ {
+ "epoch": 8.105431309904153,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0529,
+ "step": 2537
+ },
+ {
+ "epoch": 8.108626198083067,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0645,
+ "step": 2538
+ },
+ {
+ "epoch": 8.111821086261982,
+ "grad_norm": 0.08349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0784,
+ "step": 2539
+ },
+ {
+ "epoch": 8.115015974440894,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0689,
+ "step": 2540
+ },
+ {
+ "epoch": 8.118210862619808,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0667,
+ "step": 2541
+ },
+ {
+ "epoch": 8.121405750798722,
+ "grad_norm": 0.06005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0626,
+ "step": 2542
+ },
+ {
+ "epoch": 8.124600638977636,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0722,
+ "step": 2543
+ },
+ {
+ "epoch": 8.12779552715655,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0641,
+ "step": 2544
+ },
+ {
+ "epoch": 8.130990415335463,
+ "grad_norm": 0.060302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0698,
+ "step": 2545
+ },
+ {
+ "epoch": 8.134185303514377,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0695,
+ "step": 2546
+ },
+ {
+ "epoch": 8.13738019169329,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.069,
+ "step": 2547
+ },
+ {
+ "epoch": 8.140575079872205,
+ "grad_norm": 0.1044921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0664,
+ "step": 2548
+ },
+ {
+ "epoch": 8.143769968051119,
+ "grad_norm": 0.107421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0595,
+ "step": 2549
+ },
+ {
+ "epoch": 8.146964856230031,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0495,
+ "step": 2550
+ },
+ {
+ "epoch": 8.150159744408946,
+ "grad_norm": 0.08642578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0785,
+ "step": 2551
+ },
+ {
+ "epoch": 8.15335463258786,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0739,
+ "step": 2552
+ },
+ {
+ "epoch": 8.156549520766774,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0761,
+ "step": 2553
+ },
+ {
+ "epoch": 8.159744408945686,
+ "grad_norm": 0.087890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0728,
+ "step": 2554
+ },
+ {
+ "epoch": 8.1629392971246,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0644,
+ "step": 2555
+ },
+ {
+ "epoch": 8.166134185303514,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0766,
+ "step": 2556
+ },
+ {
+ "epoch": 8.169329073482428,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0586,
+ "step": 2557
+ },
+ {
+ "epoch": 8.172523961661343,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0778,
+ "step": 2558
+ },
+ {
+ "epoch": 8.175718849840255,
+ "grad_norm": 0.08544921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0749,
+ "step": 2559
+ },
+ {
+ "epoch": 8.178913738019169,
+ "grad_norm": 0.10107421875,
+ "learning_rate": 0.0005,
+ "loss": 1.07,
+ "step": 2560
+ },
+ {
+ "epoch": 8.182108626198083,
+ "grad_norm": 0.0888671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0593,
+ "step": 2561
+ },
+ {
+ "epoch": 8.185303514376997,
+ "grad_norm": 0.181640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0835,
+ "step": 2562
+ },
+ {
+ "epoch": 8.188498402555911,
+ "grad_norm": 0.12255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0603,
+ "step": 2563
+ },
+ {
+ "epoch": 8.191693290734824,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0732,
+ "step": 2564
+ },
+ {
+ "epoch": 8.194888178913738,
+ "grad_norm": 0.1640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0718,
+ "step": 2565
+ },
+ {
+ "epoch": 8.198083067092652,
+ "grad_norm": 0.2294921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0764,
+ "step": 2566
+ },
+ {
+ "epoch": 8.201277955271566,
+ "grad_norm": 0.109375,
+ "learning_rate": 0.0005,
+ "loss": 1.071,
+ "step": 2567
+ },
+ {
+ "epoch": 8.204472843450478,
+ "grad_norm": 0.11376953125,
+ "learning_rate": 0.0005,
+ "loss": 1.083,
+ "step": 2568
+ },
+ {
+ "epoch": 8.207667731629392,
+ "grad_norm": 0.11181640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0668,
+ "step": 2569
+ },
+ {
+ "epoch": 8.210862619808307,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0757,
+ "step": 2570
+ },
+ {
+ "epoch": 8.21405750798722,
+ "grad_norm": 0.08447265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0639,
+ "step": 2571
+ },
+ {
+ "epoch": 8.217252396166135,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0718,
+ "step": 2572
+ },
+ {
+ "epoch": 8.220447284345047,
+ "grad_norm": 0.09521484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0673,
+ "step": 2573
+ },
+ {
+ "epoch": 8.223642172523961,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0524,
+ "step": 2574
+ },
+ {
+ "epoch": 8.226837060702875,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0584,
+ "step": 2575
+ },
+ {
+ "epoch": 8.23003194888179,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0622,
+ "step": 2576
+ },
+ {
+ "epoch": 8.233226837060704,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0597,
+ "step": 2577
+ },
+ {
+ "epoch": 8.236421725239616,
+ "grad_norm": 0.0615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0652,
+ "step": 2578
+ },
+ {
+ "epoch": 8.23961661341853,
+ "grad_norm": 0.060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0675,
+ "step": 2579
+ },
+ {
+ "epoch": 8.242811501597444,
+ "grad_norm": 0.056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0589,
+ "step": 2580
+ },
+ {
+ "epoch": 8.246006389776358,
+ "grad_norm": 0.0595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.058,
+ "step": 2581
+ },
+ {
+ "epoch": 8.249201277955272,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0657,
+ "step": 2582
+ },
+ {
+ "epoch": 8.252396166134185,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0696,
+ "step": 2583
+ },
+ {
+ "epoch": 8.255591054313099,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0548,
+ "step": 2584
+ },
+ {
+ "epoch": 8.258785942492013,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0528,
+ "step": 2585
+ },
+ {
+ "epoch": 8.261980830670927,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0666,
+ "step": 2586
+ },
+ {
+ "epoch": 8.26517571884984,
+ "grad_norm": 0.08544921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0602,
+ "step": 2587
+ },
+ {
+ "epoch": 8.268370607028753,
+ "grad_norm": 0.052001953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0783,
+ "step": 2588
+ },
+ {
+ "epoch": 8.271565495207668,
+ "grad_norm": 0.09912109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0686,
+ "step": 2589
+ },
+ {
+ "epoch": 8.274760383386582,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0663,
+ "step": 2590
+ },
+ {
+ "epoch": 8.277955271565496,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0791,
+ "step": 2591
+ },
+ {
+ "epoch": 8.281150159744408,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.073,
+ "step": 2592
+ },
+ {
+ "epoch": 8.284345047923322,
+ "grad_norm": 0.061767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0699,
+ "step": 2593
+ },
+ {
+ "epoch": 8.287539936102236,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0666,
+ "step": 2594
+ },
+ {
+ "epoch": 8.29073482428115,
+ "grad_norm": 0.08203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0706,
+ "step": 2595
+ },
+ {
+ "epoch": 8.293929712460065,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0757,
+ "step": 2596
+ },
+ {
+ "epoch": 8.297124600638977,
+ "grad_norm": 0.123046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0659,
+ "step": 2597
+ },
+ {
+ "epoch": 8.300319488817891,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0642,
+ "step": 2598
+ },
+ {
+ "epoch": 8.303514376996805,
+ "grad_norm": 0.10791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0665,
+ "step": 2599
+ },
+ {
+ "epoch": 8.30670926517572,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0648,
+ "step": 2600
+ },
+ {
+ "epoch": 8.309904153354633,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0707,
+ "step": 2601
+ },
+ {
+ "epoch": 8.313099041533546,
+ "grad_norm": 0.0849609375,
+ "learning_rate": 0.0005,
+ "loss": 1.065,
+ "step": 2602
+ },
+ {
+ "epoch": 8.31629392971246,
+ "grad_norm": 0.10546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0596,
+ "step": 2603
+ },
+ {
+ "epoch": 8.319488817891374,
+ "grad_norm": 0.10205078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0758,
+ "step": 2604
+ },
+ {
+ "epoch": 8.322683706070288,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0616,
+ "step": 2605
+ },
+ {
+ "epoch": 8.3258785942492,
+ "grad_norm": 0.095703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0618,
+ "step": 2606
+ },
+ {
+ "epoch": 8.329073482428115,
+ "grad_norm": 0.09619140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0686,
+ "step": 2607
+ },
+ {
+ "epoch": 8.332268370607029,
+ "grad_norm": 0.08984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0701,
+ "step": 2608
+ },
+ {
+ "epoch": 8.335463258785943,
+ "grad_norm": 0.12451171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0787,
+ "step": 2609
+ },
+ {
+ "epoch": 8.338658146964857,
+ "grad_norm": 0.1474609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0505,
+ "step": 2610
+ },
+ {
+ "epoch": 8.34185303514377,
+ "grad_norm": 0.0966796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0734,
+ "step": 2611
+ },
+ {
+ "epoch": 8.345047923322683,
+ "grad_norm": 0.2255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0727,
+ "step": 2612
+ },
+ {
+ "epoch": 8.348242811501597,
+ "grad_norm": 0.16796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0756,
+ "step": 2613
+ },
+ {
+ "epoch": 8.351437699680512,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0662,
+ "step": 2614
+ },
+ {
+ "epoch": 8.354632587859426,
+ "grad_norm": 0.12158203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0594,
+ "step": 2615
+ },
+ {
+ "epoch": 8.357827476038338,
+ "grad_norm": 0.12158203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0674,
+ "step": 2616
+ },
+ {
+ "epoch": 8.361022364217252,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0829,
+ "step": 2617
+ },
+ {
+ "epoch": 8.364217252396166,
+ "grad_norm": 0.1064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0499,
+ "step": 2618
+ },
+ {
+ "epoch": 8.36741214057508,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0684,
+ "step": 2619
+ },
+ {
+ "epoch": 8.370607028753994,
+ "grad_norm": 0.09716796875,
+ "learning_rate": 0.0005,
+ "loss": 1.059,
+ "step": 2620
+ },
+ {
+ "epoch": 8.373801916932907,
+ "grad_norm": 0.09375,
+ "learning_rate": 0.0005,
+ "loss": 1.0739,
+ "step": 2621
+ },
+ {
+ "epoch": 8.37699680511182,
+ "grad_norm": 0.060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0658,
+ "step": 2622
+ },
+ {
+ "epoch": 8.380191693290735,
+ "grad_norm": 0.0869140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0664,
+ "step": 2623
+ },
+ {
+ "epoch": 8.383386581469649,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0595,
+ "step": 2624
+ },
+ {
+ "epoch": 8.386581469648561,
+ "grad_norm": 0.11572265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0629,
+ "step": 2625
+ },
+ {
+ "epoch": 8.389776357827476,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0572,
+ "step": 2626
+ },
+ {
+ "epoch": 8.39297124600639,
+ "grad_norm": 0.1015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0784,
+ "step": 2627
+ },
+ {
+ "epoch": 8.396166134185304,
+ "grad_norm": 0.08837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0606,
+ "step": 2628
+ },
+ {
+ "epoch": 8.399361022364218,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0555,
+ "step": 2629
+ },
+ {
+ "epoch": 8.40255591054313,
+ "grad_norm": 0.06201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0604,
+ "step": 2630
+ },
+ {
+ "epoch": 8.405750798722044,
+ "grad_norm": 0.059326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0662,
+ "step": 2631
+ },
+ {
+ "epoch": 8.408945686900958,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0707,
+ "step": 2632
+ },
+ {
+ "epoch": 8.412140575079873,
+ "grad_norm": 0.08349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0731,
+ "step": 2633
+ },
+ {
+ "epoch": 8.415335463258787,
+ "grad_norm": 0.10009765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0618,
+ "step": 2634
+ },
+ {
+ "epoch": 8.418530351437699,
+ "grad_norm": 0.0576171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0651,
+ "step": 2635
+ },
+ {
+ "epoch": 8.421725239616613,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0734,
+ "step": 2636
+ },
+ {
+ "epoch": 8.424920127795527,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.072,
+ "step": 2637
+ },
+ {
+ "epoch": 8.428115015974441,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0612,
+ "step": 2638
+ },
+ {
+ "epoch": 8.431309904153355,
+ "grad_norm": 0.058837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0577,
+ "step": 2639
+ },
+ {
+ "epoch": 8.434504792332268,
+ "grad_norm": 0.09375,
+ "learning_rate": 0.0005,
+ "loss": 1.0817,
+ "step": 2640
+ },
+ {
+ "epoch": 8.437699680511182,
+ "grad_norm": 0.09326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0572,
+ "step": 2641
+ },
+ {
+ "epoch": 8.440894568690096,
+ "grad_norm": 0.1181640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0671,
+ "step": 2642
+ },
+ {
+ "epoch": 8.44408945686901,
+ "grad_norm": 0.11328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0555,
+ "step": 2643
+ },
+ {
+ "epoch": 8.447284345047922,
+ "grad_norm": 0.09521484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0686,
+ "step": 2644
+ },
+ {
+ "epoch": 8.450479233226837,
+ "grad_norm": 0.154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0599,
+ "step": 2645
+ },
+ {
+ "epoch": 8.45367412140575,
+ "grad_norm": 0.146484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0661,
+ "step": 2646
+ },
+ {
+ "epoch": 8.456869009584665,
+ "grad_norm": 0.080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0719,
+ "step": 2647
+ },
+ {
+ "epoch": 8.460063897763579,
+ "grad_norm": 0.10595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0622,
+ "step": 2648
+ },
+ {
+ "epoch": 8.463258785942491,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0748,
+ "step": 2649
+ },
+ {
+ "epoch": 8.466453674121405,
+ "grad_norm": 0.087890625,
+ "learning_rate": 0.0005,
+ "loss": 1.074,
+ "step": 2650
+ },
+ {
+ "epoch": 8.46964856230032,
+ "grad_norm": 0.08447265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0755,
+ "step": 2651
+ },
+ {
+ "epoch": 8.472843450479234,
+ "grad_norm": 0.0849609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0676,
+ "step": 2652
+ },
+ {
+ "epoch": 8.476038338658148,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0495,
+ "step": 2653
+ },
+ {
+ "epoch": 8.47923322683706,
+ "grad_norm": 0.09228515625,
+ "learning_rate": 0.0005,
+ "loss": 1.062,
+ "step": 2654
+ },
+ {
+ "epoch": 8.482428115015974,
+ "grad_norm": 0.060791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0761,
+ "step": 2655
+ },
+ {
+ "epoch": 8.485623003194888,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0716,
+ "step": 2656
+ },
+ {
+ "epoch": 8.488817891373802,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0767,
+ "step": 2657
+ },
+ {
+ "epoch": 8.492012779552716,
+ "grad_norm": 0.14453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0709,
+ "step": 2658
+ },
+ {
+ "epoch": 8.495207667731629,
+ "grad_norm": 0.11279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0736,
+ "step": 2659
+ },
+ {
+ "epoch": 8.498402555910543,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0729,
+ "step": 2660
+ },
+ {
+ "epoch": 8.501597444089457,
+ "grad_norm": 0.1376953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0689,
+ "step": 2661
+ },
+ {
+ "epoch": 8.504792332268371,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0671,
+ "step": 2662
+ },
+ {
+ "epoch": 8.507987220447284,
+ "grad_norm": 0.11865234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0763,
+ "step": 2663
+ },
+ {
+ "epoch": 8.511182108626198,
+ "grad_norm": 0.140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0577,
+ "step": 2664
+ },
+ {
+ "epoch": 8.514376996805112,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0635,
+ "step": 2665
+ },
+ {
+ "epoch": 8.517571884984026,
+ "grad_norm": 0.166015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0668,
+ "step": 2666
+ },
+ {
+ "epoch": 8.52076677316294,
+ "grad_norm": 0.0810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0653,
+ "step": 2667
+ },
+ {
+ "epoch": 8.523961661341852,
+ "grad_norm": 0.10888671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0748,
+ "step": 2668
+ },
+ {
+ "epoch": 8.527156549520766,
+ "grad_norm": 0.09228515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0571,
+ "step": 2669
+ },
+ {
+ "epoch": 8.53035143769968,
+ "grad_norm": 0.09130859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0598,
+ "step": 2670
+ },
+ {
+ "epoch": 8.533546325878595,
+ "grad_norm": 0.1279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0704,
+ "step": 2671
+ },
+ {
+ "epoch": 8.536741214057509,
+ "grad_norm": 0.09130859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0635,
+ "step": 2672
+ },
+ {
+ "epoch": 8.539936102236421,
+ "grad_norm": 0.1884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0606,
+ "step": 2673
+ },
+ {
+ "epoch": 8.543130990415335,
+ "grad_norm": 0.1171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0679,
+ "step": 2674
+ },
+ {
+ "epoch": 8.54632587859425,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0714,
+ "step": 2675
+ },
+ {
+ "epoch": 8.549520766773163,
+ "grad_norm": 0.1533203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0711,
+ "step": 2676
+ },
+ {
+ "epoch": 8.552715654952078,
+ "grad_norm": 0.1298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0617,
+ "step": 2677
+ },
+ {
+ "epoch": 8.55591054313099,
+ "grad_norm": 0.1220703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0785,
+ "step": 2678
+ },
+ {
+ "epoch": 8.559105431309904,
+ "grad_norm": 0.26953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0655,
+ "step": 2679
+ },
+ {
+ "epoch": 8.562300319488818,
+ "grad_norm": 0.20703125,
+ "learning_rate": 0.0005,
+ "loss": 1.071,
+ "step": 2680
+ },
+ {
+ "epoch": 8.565495207667732,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.069,
+ "step": 2681
+ },
+ {
+ "epoch": 8.568690095846645,
+ "grad_norm": 0.169921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0586,
+ "step": 2682
+ },
+ {
+ "epoch": 8.571884984025559,
+ "grad_norm": 0.111328125,
+ "learning_rate": 0.0005,
+ "loss": 1.073,
+ "step": 2683
+ },
+ {
+ "epoch": 8.575079872204473,
+ "grad_norm": 0.09033203125,
+ "learning_rate": 0.0005,
+ "loss": 1.071,
+ "step": 2684
+ },
+ {
+ "epoch": 8.578274760383387,
+ "grad_norm": 0.16796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0785,
+ "step": 2685
+ },
+ {
+ "epoch": 8.581469648562301,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0653,
+ "step": 2686
+ },
+ {
+ "epoch": 8.584664536741213,
+ "grad_norm": 0.091796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0717,
+ "step": 2687
+ },
+ {
+ "epoch": 8.587859424920127,
+ "grad_norm": 0.06103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0612,
+ "step": 2688
+ },
+ {
+ "epoch": 8.591054313099042,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.058,
+ "step": 2689
+ },
+ {
+ "epoch": 8.594249201277956,
+ "grad_norm": 0.140625,
+ "learning_rate": 0.0005,
+ "loss": 1.06,
+ "step": 2690
+ },
+ {
+ "epoch": 8.59744408945687,
+ "grad_norm": 0.11328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0714,
+ "step": 2691
+ },
+ {
+ "epoch": 8.600638977635782,
+ "grad_norm": 0.1328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0757,
+ "step": 2692
+ },
+ {
+ "epoch": 8.603833865814696,
+ "grad_norm": 0.158203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0524,
+ "step": 2693
+ },
+ {
+ "epoch": 8.60702875399361,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0701,
+ "step": 2694
+ },
+ {
+ "epoch": 8.610223642172524,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0692,
+ "step": 2695
+ },
+ {
+ "epoch": 8.613418530351439,
+ "grad_norm": 0.10595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0538,
+ "step": 2696
+ },
+ {
+ "epoch": 8.616613418530351,
+ "grad_norm": 0.08984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0682,
+ "step": 2697
+ },
+ {
+ "epoch": 8.619808306709265,
+ "grad_norm": 0.154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0589,
+ "step": 2698
+ },
+ {
+ "epoch": 8.62300319488818,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0709,
+ "step": 2699
+ },
+ {
+ "epoch": 8.626198083067093,
+ "grad_norm": 0.146484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0764,
+ "step": 2700
+ },
+ {
+ "epoch": 8.629392971246006,
+ "grad_norm": 0.193359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0629,
+ "step": 2701
+ },
+ {
+ "epoch": 8.63258785942492,
+ "grad_norm": 0.134765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0652,
+ "step": 2702
+ },
+ {
+ "epoch": 8.635782747603834,
+ "grad_norm": 0.095703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0616,
+ "step": 2703
+ },
+ {
+ "epoch": 8.638977635782748,
+ "grad_norm": 0.177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0731,
+ "step": 2704
+ },
+ {
+ "epoch": 8.642172523961662,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0509,
+ "step": 2705
+ },
+ {
+ "epoch": 8.645367412140574,
+ "grad_norm": 0.1787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.072,
+ "step": 2706
+ },
+ {
+ "epoch": 8.648562300319488,
+ "grad_norm": 0.11474609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0673,
+ "step": 2707
+ },
+ {
+ "epoch": 8.651757188498403,
+ "grad_norm": 0.10107421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0634,
+ "step": 2708
+ },
+ {
+ "epoch": 8.654952076677317,
+ "grad_norm": 0.12158203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0671,
+ "step": 2709
+ },
+ {
+ "epoch": 8.65814696485623,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0638,
+ "step": 2710
+ },
+ {
+ "epoch": 8.661341853035143,
+ "grad_norm": 0.0986328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0656,
+ "step": 2711
+ },
+ {
+ "epoch": 8.664536741214057,
+ "grad_norm": 0.09765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0777,
+ "step": 2712
+ },
+ {
+ "epoch": 8.667731629392971,
+ "grad_norm": 0.10693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0536,
+ "step": 2713
+ },
+ {
+ "epoch": 8.670926517571885,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.084,
+ "step": 2714
+ },
+ {
+ "epoch": 8.6741214057508,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0633,
+ "step": 2715
+ },
+ {
+ "epoch": 8.677316293929712,
+ "grad_norm": 0.08349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0484,
+ "step": 2716
+ },
+ {
+ "epoch": 8.680511182108626,
+ "grad_norm": 0.08544921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0735,
+ "step": 2717
+ },
+ {
+ "epoch": 8.68370607028754,
+ "grad_norm": 0.1142578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0662,
+ "step": 2718
+ },
+ {
+ "epoch": 8.686900958466454,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0627,
+ "step": 2719
+ },
+ {
+ "epoch": 8.690095846645367,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0733,
+ "step": 2720
+ },
+ {
+ "epoch": 8.69329073482428,
+ "grad_norm": 0.08544921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0735,
+ "step": 2721
+ },
+ {
+ "epoch": 8.696485623003195,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0594,
+ "step": 2722
+ },
+ {
+ "epoch": 8.699680511182109,
+ "grad_norm": 0.10107421875,
+ "learning_rate": 0.0005,
+ "loss": 1.072,
+ "step": 2723
+ },
+ {
+ "epoch": 8.702875399361023,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.072,
+ "step": 2724
+ },
+ {
+ "epoch": 8.706070287539935,
+ "grad_norm": 0.09130859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0594,
+ "step": 2725
+ },
+ {
+ "epoch": 8.70926517571885,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0747,
+ "step": 2726
+ },
+ {
+ "epoch": 8.712460063897764,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.064,
+ "step": 2727
+ },
+ {
+ "epoch": 8.715654952076678,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0709,
+ "step": 2728
+ },
+ {
+ "epoch": 8.718849840255592,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0664,
+ "step": 2729
+ },
+ {
+ "epoch": 8.722044728434504,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0554,
+ "step": 2730
+ },
+ {
+ "epoch": 8.725239616613418,
+ "grad_norm": 0.05859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0593,
+ "step": 2731
+ },
+ {
+ "epoch": 8.728434504792332,
+ "grad_norm": 0.08154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0614,
+ "step": 2732
+ },
+ {
+ "epoch": 8.731629392971247,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.077,
+ "step": 2733
+ },
+ {
+ "epoch": 8.73482428115016,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0611,
+ "step": 2734
+ },
+ {
+ "epoch": 8.738019169329073,
+ "grad_norm": 0.09033203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0788,
+ "step": 2735
+ },
+ {
+ "epoch": 8.741214057507987,
+ "grad_norm": 0.080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0681,
+ "step": 2736
+ },
+ {
+ "epoch": 8.744408945686901,
+ "grad_norm": 0.083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0738,
+ "step": 2737
+ },
+ {
+ "epoch": 8.747603833865815,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0466,
+ "step": 2738
+ },
+ {
+ "epoch": 8.750798722044728,
+ "grad_norm": 0.0888671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0759,
+ "step": 2739
+ },
+ {
+ "epoch": 8.753993610223642,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0702,
+ "step": 2740
+ },
+ {
+ "epoch": 8.757188498402556,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0507,
+ "step": 2741
+ },
+ {
+ "epoch": 8.76038338658147,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.072,
+ "step": 2742
+ },
+ {
+ "epoch": 8.763578274760384,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.064,
+ "step": 2743
+ },
+ {
+ "epoch": 8.766773162939296,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0559,
+ "step": 2744
+ },
+ {
+ "epoch": 8.76996805111821,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0681,
+ "step": 2745
+ },
+ {
+ "epoch": 8.773162939297125,
+ "grad_norm": 0.0986328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0757,
+ "step": 2746
+ },
+ {
+ "epoch": 8.776357827476039,
+ "grad_norm": 0.1220703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0631,
+ "step": 2747
+ },
+ {
+ "epoch": 8.779552715654953,
+ "grad_norm": 0.06201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0662,
+ "step": 2748
+ },
+ {
+ "epoch": 8.782747603833865,
+ "grad_norm": 0.10498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0566,
+ "step": 2749
+ },
+ {
+ "epoch": 8.78594249201278,
+ "grad_norm": 0.07861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0689,
+ "step": 2750
+ },
+ {
+ "epoch": 8.789137380191693,
+ "grad_norm": 0.138671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0529,
+ "step": 2751
+ },
+ {
+ "epoch": 8.792332268370608,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0571,
+ "step": 2752
+ },
+ {
+ "epoch": 8.795527156549522,
+ "grad_norm": 0.0859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0693,
+ "step": 2753
+ },
+ {
+ "epoch": 8.798722044728434,
+ "grad_norm": 0.09228515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0764,
+ "step": 2754
+ },
+ {
+ "epoch": 8.801916932907348,
+ "grad_norm": 0.083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0592,
+ "step": 2755
+ },
+ {
+ "epoch": 8.805111821086262,
+ "grad_norm": 0.130859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0586,
+ "step": 2756
+ },
+ {
+ "epoch": 8.808306709265176,
+ "grad_norm": 0.2578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0809,
+ "step": 2757
+ },
+ {
+ "epoch": 8.811501597444089,
+ "grad_norm": 0.2060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0536,
+ "step": 2758
+ },
+ {
+ "epoch": 8.814696485623003,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0675,
+ "step": 2759
+ },
+ {
+ "epoch": 8.817891373801917,
+ "grad_norm": 0.150390625,
+ "learning_rate": 0.0005,
+ "loss": 1.064,
+ "step": 2760
+ },
+ {
+ "epoch": 8.821086261980831,
+ "grad_norm": 0.130859375,
+ "learning_rate": 0.0005,
+ "loss": 1.079,
+ "step": 2761
+ },
+ {
+ "epoch": 8.824281150159745,
+ "grad_norm": 0.1337890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0578,
+ "step": 2762
+ },
+ {
+ "epoch": 8.827476038338657,
+ "grad_norm": 0.275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0808,
+ "step": 2763
+ },
+ {
+ "epoch": 8.830670926517572,
+ "grad_norm": 0.279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0725,
+ "step": 2764
+ },
+ {
+ "epoch": 8.833865814696486,
+ "grad_norm": 0.2451171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0644,
+ "step": 2765
+ },
+ {
+ "epoch": 8.8370607028754,
+ "grad_norm": 0.1025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0521,
+ "step": 2766
+ },
+ {
+ "epoch": 8.840255591054314,
+ "grad_norm": 0.1533203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0704,
+ "step": 2767
+ },
+ {
+ "epoch": 8.843450479233226,
+ "grad_norm": 0.1865234375,
+ "learning_rate": 0.0005,
+ "loss": 1.071,
+ "step": 2768
+ },
+ {
+ "epoch": 8.84664536741214,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0628,
+ "step": 2769
+ },
+ {
+ "epoch": 8.849840255591054,
+ "grad_norm": 0.232421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0538,
+ "step": 2770
+ },
+ {
+ "epoch": 8.853035143769969,
+ "grad_norm": 0.1044921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0652,
+ "step": 2771
+ },
+ {
+ "epoch": 8.856230031948883,
+ "grad_norm": 0.11572265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0565,
+ "step": 2772
+ },
+ {
+ "epoch": 8.859424920127795,
+ "grad_norm": 0.134765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0787,
+ "step": 2773
+ },
+ {
+ "epoch": 8.86261980830671,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0723,
+ "step": 2774
+ },
+ {
+ "epoch": 8.865814696485623,
+ "grad_norm": 0.09375,
+ "learning_rate": 0.0005,
+ "loss": 1.0705,
+ "step": 2775
+ },
+ {
+ "epoch": 8.869009584664537,
+ "grad_norm": 0.0908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0707,
+ "step": 2776
+ },
+ {
+ "epoch": 8.87220447284345,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0706,
+ "step": 2777
+ },
+ {
+ "epoch": 8.875399361022364,
+ "grad_norm": 0.09814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0679,
+ "step": 2778
+ },
+ {
+ "epoch": 8.878594249201278,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0781,
+ "step": 2779
+ },
+ {
+ "epoch": 8.881789137380192,
+ "grad_norm": 0.09619140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0512,
+ "step": 2780
+ },
+ {
+ "epoch": 8.884984025559106,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0551,
+ "step": 2781
+ },
+ {
+ "epoch": 8.888178913738018,
+ "grad_norm": 0.10400390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0801,
+ "step": 2782
+ },
+ {
+ "epoch": 8.891373801916933,
+ "grad_norm": 0.1201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0664,
+ "step": 2783
+ },
+ {
+ "epoch": 8.894568690095847,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0633,
+ "step": 2784
+ },
+ {
+ "epoch": 8.89776357827476,
+ "grad_norm": 0.0927734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0592,
+ "step": 2785
+ },
+ {
+ "epoch": 8.900958466453675,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0647,
+ "step": 2786
+ },
+ {
+ "epoch": 8.904153354632587,
+ "grad_norm": 0.08251953125,
+ "learning_rate": 0.0005,
+ "loss": 1.064,
+ "step": 2787
+ },
+ {
+ "epoch": 8.907348242811501,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0656,
+ "step": 2788
+ },
+ {
+ "epoch": 8.910543130990416,
+ "grad_norm": 0.059814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0646,
+ "step": 2789
+ },
+ {
+ "epoch": 8.91373801916933,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0778,
+ "step": 2790
+ },
+ {
+ "epoch": 8.916932907348244,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0636,
+ "step": 2791
+ },
+ {
+ "epoch": 8.920127795527156,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0589,
+ "step": 2792
+ },
+ {
+ "epoch": 8.92332268370607,
+ "grad_norm": 0.058837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0571,
+ "step": 2793
+ },
+ {
+ "epoch": 8.926517571884984,
+ "grad_norm": 0.0810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0582,
+ "step": 2794
+ },
+ {
+ "epoch": 8.929712460063898,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0625,
+ "step": 2795
+ },
+ {
+ "epoch": 8.93290734824281,
+ "grad_norm": 0.0595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0656,
+ "step": 2796
+ },
+ {
+ "epoch": 8.936102236421725,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0534,
+ "step": 2797
+ },
+ {
+ "epoch": 8.939297124600639,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0562,
+ "step": 2798
+ },
+ {
+ "epoch": 8.942492012779553,
+ "grad_norm": 0.08154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0627,
+ "step": 2799
+ },
+ {
+ "epoch": 8.945686900958467,
+ "grad_norm": 0.09912109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0521,
+ "step": 2800
+ },
+ {
+ "epoch": 8.94888178913738,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0605,
+ "step": 2801
+ },
+ {
+ "epoch": 8.952076677316294,
+ "grad_norm": 0.11083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0578,
+ "step": 2802
+ },
+ {
+ "epoch": 8.955271565495208,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0663,
+ "step": 2803
+ },
+ {
+ "epoch": 8.958466453674122,
+ "grad_norm": 0.1220703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0831,
+ "step": 2804
+ },
+ {
+ "epoch": 8.961661341853034,
+ "grad_norm": 0.10205078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0679,
+ "step": 2805
+ },
+ {
+ "epoch": 8.964856230031948,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0648,
+ "step": 2806
+ },
+ {
+ "epoch": 8.968051118210862,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0756,
+ "step": 2807
+ },
+ {
+ "epoch": 8.971246006389777,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0655,
+ "step": 2808
+ },
+ {
+ "epoch": 8.97444089456869,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0659,
+ "step": 2809
+ },
+ {
+ "epoch": 8.977635782747605,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.064,
+ "step": 2810
+ },
+ {
+ "epoch": 8.980830670926517,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0754,
+ "step": 2811
+ },
+ {
+ "epoch": 8.984025559105431,
+ "grad_norm": 0.154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.071,
+ "step": 2812
+ },
+ {
+ "epoch": 8.987220447284345,
+ "grad_norm": 0.080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0583,
+ "step": 2813
+ },
+ {
+ "epoch": 8.99041533546326,
+ "grad_norm": 0.09765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0632,
+ "step": 2814
+ },
+ {
+ "epoch": 8.993610223642172,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0663,
+ "step": 2815
+ },
+ {
+ "epoch": 8.996805111821086,
+ "grad_norm": 0.06201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0546,
+ "step": 2816
+ },
+ {
+ "epoch": 9.0,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0658,
+ "step": 2817
+ },
+ {
+ "epoch": 9.003194888178914,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0757,
+ "step": 2818
+ },
+ {
+ "epoch": 9.006389776357828,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0644,
+ "step": 2819
+ },
+ {
+ "epoch": 9.00958466453674,
+ "grad_norm": 0.062255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0669,
+ "step": 2820
+ },
+ {
+ "epoch": 9.012779552715655,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.073,
+ "step": 2821
+ },
+ {
+ "epoch": 9.015974440894569,
+ "grad_norm": 0.056884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0634,
+ "step": 2822
+ },
+ {
+ "epoch": 9.019169329073483,
+ "grad_norm": 0.0849609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0766,
+ "step": 2823
+ },
+ {
+ "epoch": 9.022364217252397,
+ "grad_norm": 0.09765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0563,
+ "step": 2824
+ },
+ {
+ "epoch": 9.02555910543131,
+ "grad_norm": 0.1875,
+ "learning_rate": 0.0005,
+ "loss": 1.0639,
+ "step": 2825
+ },
+ {
+ "epoch": 9.028753993610223,
+ "grad_norm": 0.080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0628,
+ "step": 2826
+ },
+ {
+ "epoch": 9.031948881789138,
+ "grad_norm": 0.1904296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0657,
+ "step": 2827
+ },
+ {
+ "epoch": 9.035143769968052,
+ "grad_norm": 0.087890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0718,
+ "step": 2828
+ },
+ {
+ "epoch": 9.038338658146964,
+ "grad_norm": 0.08642578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0585,
+ "step": 2829
+ },
+ {
+ "epoch": 9.041533546325878,
+ "grad_norm": 0.08251953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0455,
+ "step": 2830
+ },
+ {
+ "epoch": 9.044728434504792,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0723,
+ "step": 2831
+ },
+ {
+ "epoch": 9.047923322683706,
+ "grad_norm": 0.123046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0733,
+ "step": 2832
+ },
+ {
+ "epoch": 9.05111821086262,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0737,
+ "step": 2833
+ },
+ {
+ "epoch": 9.054313099041533,
+ "grad_norm": 0.12353515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0708,
+ "step": 2834
+ },
+ {
+ "epoch": 9.057507987220447,
+ "grad_norm": 0.08251953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0667,
+ "step": 2835
+ },
+ {
+ "epoch": 9.060702875399361,
+ "grad_norm": 0.09765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0646,
+ "step": 2836
+ },
+ {
+ "epoch": 9.063897763578275,
+ "grad_norm": 0.10498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0616,
+ "step": 2837
+ },
+ {
+ "epoch": 9.06709265175719,
+ "grad_norm": 0.10302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0537,
+ "step": 2838
+ },
+ {
+ "epoch": 9.070287539936102,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0578,
+ "step": 2839
+ },
+ {
+ "epoch": 9.073482428115016,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0614,
+ "step": 2840
+ },
+ {
+ "epoch": 9.07667731629393,
+ "grad_norm": 0.06201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0523,
+ "step": 2841
+ },
+ {
+ "epoch": 9.079872204472844,
+ "grad_norm": 0.09130859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0626,
+ "step": 2842
+ },
+ {
+ "epoch": 9.083067092651758,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.06,
+ "step": 2843
+ },
+ {
+ "epoch": 9.08626198083067,
+ "grad_norm": 0.091796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0617,
+ "step": 2844
+ },
+ {
+ "epoch": 9.089456869009584,
+ "grad_norm": 0.05029296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0596,
+ "step": 2845
+ },
+ {
+ "epoch": 9.092651757188499,
+ "grad_norm": 0.08837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0678,
+ "step": 2846
+ },
+ {
+ "epoch": 9.095846645367413,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0583,
+ "step": 2847
+ },
+ {
+ "epoch": 9.099041533546325,
+ "grad_norm": 0.10693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0716,
+ "step": 2848
+ },
+ {
+ "epoch": 9.10223642172524,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.06,
+ "step": 2849
+ },
+ {
+ "epoch": 9.105431309904153,
+ "grad_norm": 0.056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0767,
+ "step": 2850
+ },
+ {
+ "epoch": 9.108626198083067,
+ "grad_norm": 0.099609375,
+ "learning_rate": 0.0005,
+ "loss": 1.064,
+ "step": 2851
+ },
+ {
+ "epoch": 9.111821086261982,
+ "grad_norm": 0.05517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0588,
+ "step": 2852
+ },
+ {
+ "epoch": 9.115015974440894,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0662,
+ "step": 2853
+ },
+ {
+ "epoch": 9.118210862619808,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0615,
+ "step": 2854
+ },
+ {
+ "epoch": 9.121405750798722,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0753,
+ "step": 2855
+ },
+ {
+ "epoch": 9.124600638977636,
+ "grad_norm": 0.06201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0713,
+ "step": 2856
+ },
+ {
+ "epoch": 9.12779552715655,
+ "grad_norm": 0.08642578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0624,
+ "step": 2857
+ },
+ {
+ "epoch": 9.130990415335463,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0612,
+ "step": 2858
+ },
+ {
+ "epoch": 9.134185303514377,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0743,
+ "step": 2859
+ },
+ {
+ "epoch": 9.13738019169329,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0557,
+ "step": 2860
+ },
+ {
+ "epoch": 9.140575079872205,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0538,
+ "step": 2861
+ },
+ {
+ "epoch": 9.143769968051119,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0551,
+ "step": 2862
+ },
+ {
+ "epoch": 9.146964856230031,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.065,
+ "step": 2863
+ },
+ {
+ "epoch": 9.150159744408946,
+ "grad_norm": 0.0595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.074,
+ "step": 2864
+ },
+ {
+ "epoch": 9.15335463258786,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0657,
+ "step": 2865
+ },
+ {
+ "epoch": 9.156549520766774,
+ "grad_norm": 0.09228515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0618,
+ "step": 2866
+ },
+ {
+ "epoch": 9.159744408945686,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0643,
+ "step": 2867
+ },
+ {
+ "epoch": 9.1629392971246,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0523,
+ "step": 2868
+ },
+ {
+ "epoch": 9.166134185303514,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.054,
+ "step": 2869
+ },
+ {
+ "epoch": 9.169329073482428,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0476,
+ "step": 2870
+ },
+ {
+ "epoch": 9.172523961661343,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0702,
+ "step": 2871
+ },
+ {
+ "epoch": 9.175718849840255,
+ "grad_norm": 0.05908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0675,
+ "step": 2872
+ },
+ {
+ "epoch": 9.178913738019169,
+ "grad_norm": 0.061279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0712,
+ "step": 2873
+ },
+ {
+ "epoch": 9.182108626198083,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0534,
+ "step": 2874
+ },
+ {
+ "epoch": 9.185303514376997,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0628,
+ "step": 2875
+ },
+ {
+ "epoch": 9.188498402555911,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0596,
+ "step": 2876
+ },
+ {
+ "epoch": 9.191693290734824,
+ "grad_norm": 0.08349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.087,
+ "step": 2877
+ },
+ {
+ "epoch": 9.194888178913738,
+ "grad_norm": 0.10546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0652,
+ "step": 2878
+ },
+ {
+ "epoch": 9.198083067092652,
+ "grad_norm": 0.08642578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0712,
+ "step": 2879
+ },
+ {
+ "epoch": 9.201277955271566,
+ "grad_norm": 0.08203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0462,
+ "step": 2880
+ },
+ {
+ "epoch": 9.204472843450478,
+ "grad_norm": 0.0830078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0697,
+ "step": 2881
+ },
+ {
+ "epoch": 9.207667731629392,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0598,
+ "step": 2882
+ },
+ {
+ "epoch": 9.210862619808307,
+ "grad_norm": 0.08349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0531,
+ "step": 2883
+ },
+ {
+ "epoch": 9.21405750798722,
+ "grad_norm": 0.099609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0567,
+ "step": 2884
+ },
+ {
+ "epoch": 9.217252396166135,
+ "grad_norm": 0.0859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0701,
+ "step": 2885
+ },
+ {
+ "epoch": 9.220447284345047,
+ "grad_norm": 0.08203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0766,
+ "step": 2886
+ },
+ {
+ "epoch": 9.223642172523961,
+ "grad_norm": 0.0830078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0745,
+ "step": 2887
+ },
+ {
+ "epoch": 9.226837060702875,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0625,
+ "step": 2888
+ },
+ {
+ "epoch": 9.23003194888179,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.058,
+ "step": 2889
+ },
+ {
+ "epoch": 9.233226837060704,
+ "grad_norm": 0.05908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0629,
+ "step": 2890
+ },
+ {
+ "epoch": 9.236421725239616,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0564,
+ "step": 2891
+ },
+ {
+ "epoch": 9.23961661341853,
+ "grad_norm": 0.06005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0677,
+ "step": 2892
+ },
+ {
+ "epoch": 9.242811501597444,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0454,
+ "step": 2893
+ },
+ {
+ "epoch": 9.246006389776358,
+ "grad_norm": 0.080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.059,
+ "step": 2894
+ },
+ {
+ "epoch": 9.249201277955272,
+ "grad_norm": 0.0576171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0709,
+ "step": 2895
+ },
+ {
+ "epoch": 9.252396166134185,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0727,
+ "step": 2896
+ },
+ {
+ "epoch": 9.255591054313099,
+ "grad_norm": 0.056884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0597,
+ "step": 2897
+ },
+ {
+ "epoch": 9.258785942492013,
+ "grad_norm": 0.08154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0487,
+ "step": 2898
+ },
+ {
+ "epoch": 9.261980830670927,
+ "grad_norm": 0.0927734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0636,
+ "step": 2899
+ },
+ {
+ "epoch": 9.26517571884984,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0715,
+ "step": 2900
+ },
+ {
+ "epoch": 9.268370607028753,
+ "grad_norm": 0.142578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0479,
+ "step": 2901
+ },
+ {
+ "epoch": 9.271565495207668,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0654,
+ "step": 2902
+ },
+ {
+ "epoch": 9.274760383386582,
+ "grad_norm": 0.058837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0578,
+ "step": 2903
+ },
+ {
+ "epoch": 9.277955271565496,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0766,
+ "step": 2904
+ },
+ {
+ "epoch": 9.281150159744408,
+ "grad_norm": 0.06005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0685,
+ "step": 2905
+ },
+ {
+ "epoch": 9.284345047923322,
+ "grad_norm": 0.05224609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0616,
+ "step": 2906
+ },
+ {
+ "epoch": 9.287539936102236,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0594,
+ "step": 2907
+ },
+ {
+ "epoch": 9.29073482428115,
+ "grad_norm": 0.11083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0711,
+ "step": 2908
+ },
+ {
+ "epoch": 9.293929712460065,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0658,
+ "step": 2909
+ },
+ {
+ "epoch": 9.297124600638977,
+ "grad_norm": 0.099609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0726,
+ "step": 2910
+ },
+ {
+ "epoch": 9.300319488817891,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0667,
+ "step": 2911
+ },
+ {
+ "epoch": 9.303514376996805,
+ "grad_norm": 0.08984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0578,
+ "step": 2912
+ },
+ {
+ "epoch": 9.30670926517572,
+ "grad_norm": 0.08544921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0696,
+ "step": 2913
+ },
+ {
+ "epoch": 9.309904153354633,
+ "grad_norm": 0.09033203125,
+ "learning_rate": 0.0005,
+ "loss": 1.064,
+ "step": 2914
+ },
+ {
+ "epoch": 9.313099041533546,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0694,
+ "step": 2915
+ },
+ {
+ "epoch": 9.31629392971246,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0587,
+ "step": 2916
+ },
+ {
+ "epoch": 9.319488817891374,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0825,
+ "step": 2917
+ },
+ {
+ "epoch": 9.322683706070288,
+ "grad_norm": 0.08154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0682,
+ "step": 2918
+ },
+ {
+ "epoch": 9.3258785942492,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0691,
+ "step": 2919
+ },
+ {
+ "epoch": 9.329073482428115,
+ "grad_norm": 0.059814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0564,
+ "step": 2920
+ },
+ {
+ "epoch": 9.332268370607029,
+ "grad_norm": 0.09521484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0651,
+ "step": 2921
+ },
+ {
+ "epoch": 9.335463258785943,
+ "grad_norm": 0.08642578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0679,
+ "step": 2922
+ },
+ {
+ "epoch": 9.338658146964857,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0699,
+ "step": 2923
+ },
+ {
+ "epoch": 9.34185303514377,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0605,
+ "step": 2924
+ },
+ {
+ "epoch": 9.345047923322683,
+ "grad_norm": 0.10693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0584,
+ "step": 2925
+ },
+ {
+ "epoch": 9.348242811501597,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0687,
+ "step": 2926
+ },
+ {
+ "epoch": 9.351437699680512,
+ "grad_norm": 0.08349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0659,
+ "step": 2927
+ },
+ {
+ "epoch": 9.354632587859426,
+ "grad_norm": 0.087890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0536,
+ "step": 2928
+ },
+ {
+ "epoch": 9.357827476038338,
+ "grad_norm": 0.058349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0682,
+ "step": 2929
+ },
+ {
+ "epoch": 9.361022364217252,
+ "grad_norm": 0.11669921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0685,
+ "step": 2930
+ },
+ {
+ "epoch": 9.364217252396166,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0533,
+ "step": 2931
+ },
+ {
+ "epoch": 9.36741214057508,
+ "grad_norm": 0.12060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0666,
+ "step": 2932
+ },
+ {
+ "epoch": 9.370607028753994,
+ "grad_norm": 0.0966796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0619,
+ "step": 2933
+ },
+ {
+ "epoch": 9.373801916932907,
+ "grad_norm": 0.08837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0684,
+ "step": 2934
+ },
+ {
+ "epoch": 9.37699680511182,
+ "grad_norm": 0.130859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0701,
+ "step": 2935
+ },
+ {
+ "epoch": 9.380191693290735,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0681,
+ "step": 2936
+ },
+ {
+ "epoch": 9.383386581469649,
+ "grad_norm": 0.205078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0637,
+ "step": 2937
+ },
+ {
+ "epoch": 9.386581469648561,
+ "grad_norm": 0.09033203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0587,
+ "step": 2938
+ },
+ {
+ "epoch": 9.389776357827476,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0572,
+ "step": 2939
+ },
+ {
+ "epoch": 9.39297124600639,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0701,
+ "step": 2940
+ },
+ {
+ "epoch": 9.396166134185304,
+ "grad_norm": 0.08154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0691,
+ "step": 2941
+ },
+ {
+ "epoch": 9.399361022364218,
+ "grad_norm": 0.0830078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0697,
+ "step": 2942
+ },
+ {
+ "epoch": 9.40255591054313,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.053,
+ "step": 2943
+ },
+ {
+ "epoch": 9.405750798722044,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0754,
+ "step": 2944
+ },
+ {
+ "epoch": 9.408945686900958,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0503,
+ "step": 2945
+ },
+ {
+ "epoch": 9.412140575079873,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0523,
+ "step": 2946
+ },
+ {
+ "epoch": 9.415335463258787,
+ "grad_norm": 0.099609375,
+ "learning_rate": 0.0005,
+ "loss": 1.055,
+ "step": 2947
+ },
+ {
+ "epoch": 9.418530351437699,
+ "grad_norm": 0.1103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0565,
+ "step": 2948
+ },
+ {
+ "epoch": 9.421725239616613,
+ "grad_norm": 0.07861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0659,
+ "step": 2949
+ },
+ {
+ "epoch": 9.424920127795527,
+ "grad_norm": 0.09521484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0677,
+ "step": 2950
+ },
+ {
+ "epoch": 9.428115015974441,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0683,
+ "step": 2951
+ },
+ {
+ "epoch": 9.431309904153355,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0748,
+ "step": 2952
+ },
+ {
+ "epoch": 9.434504792332268,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0507,
+ "step": 2953
+ },
+ {
+ "epoch": 9.437699680511182,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0645,
+ "step": 2954
+ },
+ {
+ "epoch": 9.440894568690096,
+ "grad_norm": 0.103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0612,
+ "step": 2955
+ },
+ {
+ "epoch": 9.44408945686901,
+ "grad_norm": 0.08447265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0628,
+ "step": 2956
+ },
+ {
+ "epoch": 9.447284345047922,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0599,
+ "step": 2957
+ },
+ {
+ "epoch": 9.450479233226837,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0684,
+ "step": 2958
+ },
+ {
+ "epoch": 9.45367412140575,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0771,
+ "step": 2959
+ },
+ {
+ "epoch": 9.456869009584665,
+ "grad_norm": 0.09619140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0633,
+ "step": 2960
+ },
+ {
+ "epoch": 9.460063897763579,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0613,
+ "step": 2961
+ },
+ {
+ "epoch": 9.463258785942491,
+ "grad_norm": 0.083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0597,
+ "step": 2962
+ },
+ {
+ "epoch": 9.466453674121405,
+ "grad_norm": 0.056884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.069,
+ "step": 2963
+ },
+ {
+ "epoch": 9.46964856230032,
+ "grad_norm": 0.08544921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0657,
+ "step": 2964
+ },
+ {
+ "epoch": 9.472843450479234,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0609,
+ "step": 2965
+ },
+ {
+ "epoch": 9.476038338658148,
+ "grad_norm": 0.09326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0646,
+ "step": 2966
+ },
+ {
+ "epoch": 9.47923322683706,
+ "grad_norm": 0.1162109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0684,
+ "step": 2967
+ },
+ {
+ "epoch": 9.482428115015974,
+ "grad_norm": 0.09765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0694,
+ "step": 2968
+ },
+ {
+ "epoch": 9.485623003194888,
+ "grad_norm": 0.087890625,
+ "learning_rate": 0.0005,
+ "loss": 1.059,
+ "step": 2969
+ },
+ {
+ "epoch": 9.488817891373802,
+ "grad_norm": 0.08203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0653,
+ "step": 2970
+ },
+ {
+ "epoch": 9.492012779552716,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0597,
+ "step": 2971
+ },
+ {
+ "epoch": 9.495207667731629,
+ "grad_norm": 0.07861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0604,
+ "step": 2972
+ },
+ {
+ "epoch": 9.498402555910543,
+ "grad_norm": 0.060791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0447,
+ "step": 2973
+ },
+ {
+ "epoch": 9.501597444089457,
+ "grad_norm": 0.059814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0638,
+ "step": 2974
+ },
+ {
+ "epoch": 9.504792332268371,
+ "grad_norm": 0.056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0597,
+ "step": 2975
+ },
+ {
+ "epoch": 9.507987220447284,
+ "grad_norm": 0.05712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.066,
+ "step": 2976
+ },
+ {
+ "epoch": 9.511182108626198,
+ "grad_norm": 0.05615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0736,
+ "step": 2977
+ },
+ {
+ "epoch": 9.514376996805112,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0526,
+ "step": 2978
+ },
+ {
+ "epoch": 9.517571884984026,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0596,
+ "step": 2979
+ },
+ {
+ "epoch": 9.52076677316294,
+ "grad_norm": 0.061279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0568,
+ "step": 2980
+ },
+ {
+ "epoch": 9.523961661341852,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0653,
+ "step": 2981
+ },
+ {
+ "epoch": 9.527156549520766,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0682,
+ "step": 2982
+ },
+ {
+ "epoch": 9.53035143769968,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0684,
+ "step": 2983
+ },
+ {
+ "epoch": 9.533546325878595,
+ "grad_norm": 0.111328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0611,
+ "step": 2984
+ },
+ {
+ "epoch": 9.536741214057509,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0717,
+ "step": 2985
+ },
+ {
+ "epoch": 9.539936102236421,
+ "grad_norm": 0.0615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0663,
+ "step": 2986
+ },
+ {
+ "epoch": 9.543130990415335,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0576,
+ "step": 2987
+ },
+ {
+ "epoch": 9.54632587859425,
+ "grad_norm": 0.05517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0582,
+ "step": 2988
+ },
+ {
+ "epoch": 9.549520766773163,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0548,
+ "step": 2989
+ },
+ {
+ "epoch": 9.552715654952078,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0691,
+ "step": 2990
+ },
+ {
+ "epoch": 9.55591054313099,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0598,
+ "step": 2991
+ },
+ {
+ "epoch": 9.559105431309904,
+ "grad_norm": 0.059814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0533,
+ "step": 2992
+ },
+ {
+ "epoch": 9.562300319488818,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0618,
+ "step": 2993
+ },
+ {
+ "epoch": 9.565495207667732,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0553,
+ "step": 2994
+ },
+ {
+ "epoch": 9.568690095846645,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0624,
+ "step": 2995
+ },
+ {
+ "epoch": 9.571884984025559,
+ "grad_norm": 0.08154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0627,
+ "step": 2996
+ },
+ {
+ "epoch": 9.575079872204473,
+ "grad_norm": 0.05908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0626,
+ "step": 2997
+ },
+ {
+ "epoch": 9.578274760383387,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0703,
+ "step": 2998
+ },
+ {
+ "epoch": 9.581469648562301,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0639,
+ "step": 2999
+ },
+ {
+ "epoch": 9.584664536741213,
+ "grad_norm": 0.058837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0635,
+ "step": 3000
+ },
+ {
+ "epoch": 9.587859424920127,
+ "grad_norm": 0.06201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0594,
+ "step": 3001
+ },
+ {
+ "epoch": 9.591054313099042,
+ "grad_norm": 0.06103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.069,
+ "step": 3002
+ },
+ {
+ "epoch": 9.594249201277956,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0768,
+ "step": 3003
+ },
+ {
+ "epoch": 9.59744408945687,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0729,
+ "step": 3004
+ },
+ {
+ "epoch": 9.600638977635782,
+ "grad_norm": 0.061279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0659,
+ "step": 3005
+ },
+ {
+ "epoch": 9.603833865814696,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.073,
+ "step": 3006
+ },
+ {
+ "epoch": 9.60702875399361,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0611,
+ "step": 3007
+ },
+ {
+ "epoch": 9.610223642172524,
+ "grad_norm": 0.09716796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0783,
+ "step": 3008
+ },
+ {
+ "epoch": 9.613418530351439,
+ "grad_norm": 0.09228515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0689,
+ "step": 3009
+ },
+ {
+ "epoch": 9.616613418530351,
+ "grad_norm": 0.0986328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0553,
+ "step": 3010
+ },
+ {
+ "epoch": 9.619808306709265,
+ "grad_norm": 0.083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.061,
+ "step": 3011
+ },
+ {
+ "epoch": 9.62300319488818,
+ "grad_norm": 0.10009765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0589,
+ "step": 3012
+ },
+ {
+ "epoch": 9.626198083067093,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0424,
+ "step": 3013
+ },
+ {
+ "epoch": 9.629392971246006,
+ "grad_norm": 0.08984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0573,
+ "step": 3014
+ },
+ {
+ "epoch": 9.63258785942492,
+ "grad_norm": 0.10888671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0613,
+ "step": 3015
+ },
+ {
+ "epoch": 9.635782747603834,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0574,
+ "step": 3016
+ },
+ {
+ "epoch": 9.638977635782748,
+ "grad_norm": 0.12060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.064,
+ "step": 3017
+ },
+ {
+ "epoch": 9.642172523961662,
+ "grad_norm": 0.0546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0602,
+ "step": 3018
+ },
+ {
+ "epoch": 9.645367412140574,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.07,
+ "step": 3019
+ },
+ {
+ "epoch": 9.648562300319488,
+ "grad_norm": 0.05224609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0695,
+ "step": 3020
+ },
+ {
+ "epoch": 9.651757188498403,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0644,
+ "step": 3021
+ },
+ {
+ "epoch": 9.654952076677317,
+ "grad_norm": 0.060791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0531,
+ "step": 3022
+ },
+ {
+ "epoch": 9.65814696485623,
+ "grad_norm": 0.08837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0663,
+ "step": 3023
+ },
+ {
+ "epoch": 9.661341853035143,
+ "grad_norm": 0.10791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0628,
+ "step": 3024
+ },
+ {
+ "epoch": 9.664536741214057,
+ "grad_norm": 0.05810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0522,
+ "step": 3025
+ },
+ {
+ "epoch": 9.667731629392971,
+ "grad_norm": 0.1611328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0766,
+ "step": 3026
+ },
+ {
+ "epoch": 9.670926517571885,
+ "grad_norm": 0.181640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0587,
+ "step": 3027
+ },
+ {
+ "epoch": 9.6741214057508,
+ "grad_norm": 0.057373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0607,
+ "step": 3028
+ },
+ {
+ "epoch": 9.677316293929712,
+ "grad_norm": 0.12255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0768,
+ "step": 3029
+ },
+ {
+ "epoch": 9.680511182108626,
+ "grad_norm": 0.10888671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0643,
+ "step": 3030
+ },
+ {
+ "epoch": 9.68370607028754,
+ "grad_norm": 0.08251953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0495,
+ "step": 3031
+ },
+ {
+ "epoch": 9.686900958466454,
+ "grad_norm": 0.09814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.058,
+ "step": 3032
+ },
+ {
+ "epoch": 9.690095846645367,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0767,
+ "step": 3033
+ },
+ {
+ "epoch": 9.69329073482428,
+ "grad_norm": 0.126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0536,
+ "step": 3034
+ },
+ {
+ "epoch": 9.696485623003195,
+ "grad_norm": 0.134765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0632,
+ "step": 3035
+ },
+ {
+ "epoch": 9.699680511182109,
+ "grad_norm": 0.091796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0576,
+ "step": 3036
+ },
+ {
+ "epoch": 9.702875399361023,
+ "grad_norm": 0.169921875,
+ "learning_rate": 0.0005,
+ "loss": 1.078,
+ "step": 3037
+ },
+ {
+ "epoch": 9.706070287539935,
+ "grad_norm": 0.12060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0663,
+ "step": 3038
+ },
+ {
+ "epoch": 9.70926517571885,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0463,
+ "step": 3039
+ },
+ {
+ "epoch": 9.712460063897764,
+ "grad_norm": 0.1337890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0586,
+ "step": 3040
+ },
+ {
+ "epoch": 9.715654952076678,
+ "grad_norm": 0.134765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0716,
+ "step": 3041
+ },
+ {
+ "epoch": 9.718849840255592,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0594,
+ "step": 3042
+ },
+ {
+ "epoch": 9.722044728434504,
+ "grad_norm": 0.13671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0581,
+ "step": 3043
+ },
+ {
+ "epoch": 9.725239616613418,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0626,
+ "step": 3044
+ },
+ {
+ "epoch": 9.728434504792332,
+ "grad_norm": 0.1064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0653,
+ "step": 3045
+ },
+ {
+ "epoch": 9.731629392971247,
+ "grad_norm": 0.1796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0823,
+ "step": 3046
+ },
+ {
+ "epoch": 9.73482428115016,
+ "grad_norm": 0.1142578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0784,
+ "step": 3047
+ },
+ {
+ "epoch": 9.738019169329073,
+ "grad_norm": 0.1240234375,
+ "learning_rate": 0.0005,
+ "loss": 1.064,
+ "step": 3048
+ },
+ {
+ "epoch": 9.741214057507987,
+ "grad_norm": 0.1826171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0567,
+ "step": 3049
+ },
+ {
+ "epoch": 9.744408945686901,
+ "grad_norm": 0.09228515625,
+ "learning_rate": 0.0005,
+ "loss": 1.064,
+ "step": 3050
+ },
+ {
+ "epoch": 9.747603833865815,
+ "grad_norm": 0.1650390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0605,
+ "step": 3051
+ },
+ {
+ "epoch": 9.750798722044728,
+ "grad_norm": 0.12890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0736,
+ "step": 3052
+ },
+ {
+ "epoch": 9.753993610223642,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0645,
+ "step": 3053
+ },
+ {
+ "epoch": 9.757188498402556,
+ "grad_norm": 0.130859375,
+ "learning_rate": 0.0005,
+ "loss": 1.074,
+ "step": 3054
+ },
+ {
+ "epoch": 9.76038338658147,
+ "grad_norm": 0.11865234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0697,
+ "step": 3055
+ },
+ {
+ "epoch": 9.763578274760384,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0694,
+ "step": 3056
+ },
+ {
+ "epoch": 9.766773162939296,
+ "grad_norm": 0.142578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0649,
+ "step": 3057
+ },
+ {
+ "epoch": 9.76996805111821,
+ "grad_norm": 0.057861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0534,
+ "step": 3058
+ },
+ {
+ "epoch": 9.773162939297125,
+ "grad_norm": 0.1005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0629,
+ "step": 3059
+ },
+ {
+ "epoch": 9.776357827476039,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0657,
+ "step": 3060
+ },
+ {
+ "epoch": 9.779552715654953,
+ "grad_norm": 0.1044921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0601,
+ "step": 3061
+ },
+ {
+ "epoch": 9.782747603833865,
+ "grad_norm": 0.193359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0713,
+ "step": 3062
+ },
+ {
+ "epoch": 9.78594249201278,
+ "grad_norm": 0.1552734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0738,
+ "step": 3063
+ },
+ {
+ "epoch": 9.789137380191693,
+ "grad_norm": 0.1318359375,
+ "learning_rate": 0.0005,
+ "loss": 1.068,
+ "step": 3064
+ },
+ {
+ "epoch": 9.792332268370608,
+ "grad_norm": 0.189453125,
+ "learning_rate": 0.0005,
+ "loss": 1.063,
+ "step": 3065
+ },
+ {
+ "epoch": 9.795527156549522,
+ "grad_norm": 0.10205078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0646,
+ "step": 3066
+ },
+ {
+ "epoch": 9.798722044728434,
+ "grad_norm": 0.11474609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0644,
+ "step": 3067
+ },
+ {
+ "epoch": 9.801916932907348,
+ "grad_norm": 0.10107421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0469,
+ "step": 3068
+ },
+ {
+ "epoch": 9.805111821086262,
+ "grad_norm": 0.0615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0613,
+ "step": 3069
+ },
+ {
+ "epoch": 9.808306709265176,
+ "grad_norm": 0.09033203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0615,
+ "step": 3070
+ },
+ {
+ "epoch": 9.811501597444089,
+ "grad_norm": 0.08349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.06,
+ "step": 3071
+ },
+ {
+ "epoch": 9.814696485623003,
+ "grad_norm": 0.10546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0756,
+ "step": 3072
+ },
+ {
+ "epoch": 9.817891373801917,
+ "grad_norm": 0.1630859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0777,
+ "step": 3073
+ },
+ {
+ "epoch": 9.821086261980831,
+ "grad_norm": 0.216796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0625,
+ "step": 3074
+ },
+ {
+ "epoch": 9.824281150159745,
+ "grad_norm": 0.08984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0666,
+ "step": 3075
+ },
+ {
+ "epoch": 9.827476038338657,
+ "grad_norm": 0.1181640625,
+ "learning_rate": 0.0005,
+ "loss": 1.049,
+ "step": 3076
+ },
+ {
+ "epoch": 9.830670926517572,
+ "grad_norm": 0.0947265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0591,
+ "step": 3077
+ },
+ {
+ "epoch": 9.833865814696486,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0695,
+ "step": 3078
+ },
+ {
+ "epoch": 9.8370607028754,
+ "grad_norm": 0.08935546875,
+ "learning_rate": 0.0005,
+ "loss": 1.05,
+ "step": 3079
+ },
+ {
+ "epoch": 9.840255591054314,
+ "grad_norm": 0.08544921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0707,
+ "step": 3080
+ },
+ {
+ "epoch": 9.843450479233226,
+ "grad_norm": 0.1025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0695,
+ "step": 3081
+ },
+ {
+ "epoch": 9.84664536741214,
+ "grad_norm": 0.171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0636,
+ "step": 3082
+ },
+ {
+ "epoch": 9.849840255591054,
+ "grad_norm": 0.0986328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0755,
+ "step": 3083
+ },
+ {
+ "epoch": 9.853035143769969,
+ "grad_norm": 0.28125,
+ "learning_rate": 0.0005,
+ "loss": 1.0594,
+ "step": 3084
+ },
+ {
+ "epoch": 9.856230031948883,
+ "grad_norm": 0.1513671875,
+ "learning_rate": 0.0005,
+ "loss": 1.065,
+ "step": 3085
+ },
+ {
+ "epoch": 9.859424920127795,
+ "grad_norm": 0.138671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0549,
+ "step": 3086
+ },
+ {
+ "epoch": 9.86261980830671,
+ "grad_norm": 0.095703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0675,
+ "step": 3087
+ },
+ {
+ "epoch": 9.865814696485623,
+ "grad_norm": 0.146484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0655,
+ "step": 3088
+ },
+ {
+ "epoch": 9.869009584664537,
+ "grad_norm": 0.1787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.067,
+ "step": 3089
+ },
+ {
+ "epoch": 9.87220447284345,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0719,
+ "step": 3090
+ },
+ {
+ "epoch": 9.875399361022364,
+ "grad_norm": 0.26171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0657,
+ "step": 3091
+ },
+ {
+ "epoch": 9.878594249201278,
+ "grad_norm": 0.1904296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0534,
+ "step": 3092
+ },
+ {
+ "epoch": 9.881789137380192,
+ "grad_norm": 0.1298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0728,
+ "step": 3093
+ },
+ {
+ "epoch": 9.884984025559106,
+ "grad_norm": 0.330078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0592,
+ "step": 3094
+ },
+ {
+ "epoch": 9.888178913738018,
+ "grad_norm": 0.375,
+ "learning_rate": 0.0005,
+ "loss": 1.0623,
+ "step": 3095
+ },
+ {
+ "epoch": 9.891373801916933,
+ "grad_norm": 0.3359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0674,
+ "step": 3096
+ },
+ {
+ "epoch": 9.894568690095847,
+ "grad_norm": 0.166015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0618,
+ "step": 3097
+ },
+ {
+ "epoch": 9.89776357827476,
+ "grad_norm": 0.1259765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0584,
+ "step": 3098
+ },
+ {
+ "epoch": 9.900958466453675,
+ "grad_norm": 0.181640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0712,
+ "step": 3099
+ },
+ {
+ "epoch": 9.904153354632587,
+ "grad_norm": 0.07861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.06,
+ "step": 3100
+ },
+ {
+ "epoch": 9.907348242811501,
+ "grad_norm": 0.25390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0521,
+ "step": 3101
+ },
+ {
+ "epoch": 9.910543130990416,
+ "grad_norm": 0.2001953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0556,
+ "step": 3102
+ },
+ {
+ "epoch": 9.91373801916933,
+ "grad_norm": 0.09033203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0631,
+ "step": 3103
+ },
+ {
+ "epoch": 9.916932907348244,
+ "grad_norm": 0.294921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0511,
+ "step": 3104
+ },
+ {
+ "epoch": 9.920127795527156,
+ "grad_norm": 0.2734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0682,
+ "step": 3105
+ },
+ {
+ "epoch": 9.92332268370607,
+ "grad_norm": 0.055419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0747,
+ "step": 3106
+ },
+ {
+ "epoch": 9.926517571884984,
+ "grad_norm": 0.1875,
+ "learning_rate": 0.0005,
+ "loss": 1.0611,
+ "step": 3107
+ },
+ {
+ "epoch": 9.929712460063898,
+ "grad_norm": 0.0830078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0615,
+ "step": 3108
+ },
+ {
+ "epoch": 9.93290734824281,
+ "grad_norm": 0.1259765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0613,
+ "step": 3109
+ },
+ {
+ "epoch": 9.936102236421725,
+ "grad_norm": 0.1259765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0597,
+ "step": 3110
+ },
+ {
+ "epoch": 9.939297124600639,
+ "grad_norm": 0.09423828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0608,
+ "step": 3111
+ },
+ {
+ "epoch": 9.942492012779553,
+ "grad_norm": 0.111328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0641,
+ "step": 3112
+ },
+ {
+ "epoch": 9.945686900958467,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0711,
+ "step": 3113
+ },
+ {
+ "epoch": 9.94888178913738,
+ "grad_norm": 0.10400390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0726,
+ "step": 3114
+ },
+ {
+ "epoch": 9.952076677316294,
+ "grad_norm": 0.10205078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0664,
+ "step": 3115
+ },
+ {
+ "epoch": 9.955271565495208,
+ "grad_norm": 0.09423828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0535,
+ "step": 3116
+ },
+ {
+ "epoch": 9.958466453674122,
+ "grad_norm": 0.1572265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0782,
+ "step": 3117
+ },
+ {
+ "epoch": 9.961661341853034,
+ "grad_norm": 0.0849609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0528,
+ "step": 3118
+ },
+ {
+ "epoch": 9.964856230031948,
+ "grad_norm": 0.1181640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0504,
+ "step": 3119
+ },
+ {
+ "epoch": 9.968051118210862,
+ "grad_norm": 0.09521484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0681,
+ "step": 3120
+ },
+ {
+ "epoch": 9.971246006389777,
+ "grad_norm": 0.07861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0754,
+ "step": 3121
+ },
+ {
+ "epoch": 9.97444089456869,
+ "grad_norm": 0.1171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0671,
+ "step": 3122
+ },
+ {
+ "epoch": 9.977635782747605,
+ "grad_norm": 0.060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0672,
+ "step": 3123
+ },
+ {
+ "epoch": 9.980830670926517,
+ "grad_norm": 0.10498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0709,
+ "step": 3124
+ },
+ {
+ "epoch": 9.984025559105431,
+ "grad_norm": 0.09814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0759,
+ "step": 3125
+ },
+ {
+ "epoch": 9.987220447284345,
+ "grad_norm": 0.0908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.049,
+ "step": 3126
+ },
+ {
+ "epoch": 9.99041533546326,
+ "grad_norm": 0.1513671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0564,
+ "step": 3127
+ },
+ {
+ "epoch": 9.993610223642172,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0711,
+ "step": 3128
+ },
+ {
+ "epoch": 9.996805111821086,
+ "grad_norm": 0.1005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0556,
+ "step": 3129
+ },
+ {
+ "epoch": 10.0,
+ "grad_norm": 0.08642578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0704,
+ "step": 3130
+ },
+ {
+ "epoch": 10.003194888178914,
+ "grad_norm": 0.1845703125,
+ "learning_rate": 0.0005,
+ "loss": 1.071,
+ "step": 3131
+ },
+ {
+ "epoch": 10.006389776357828,
+ "grad_norm": 0.09033203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0499,
+ "step": 3132
+ },
+ {
+ "epoch": 10.00958466453674,
+ "grad_norm": 0.111328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0692,
+ "step": 3133
+ },
+ {
+ "epoch": 10.012779552715655,
+ "grad_norm": 0.10302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0583,
+ "step": 3134
+ },
+ {
+ "epoch": 10.015974440894569,
+ "grad_norm": 0.10498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0653,
+ "step": 3135
+ },
+ {
+ "epoch": 10.019169329073483,
+ "grad_norm": 0.10693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0566,
+ "step": 3136
+ },
+ {
+ "epoch": 10.022364217252397,
+ "grad_norm": 0.09228515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0465,
+ "step": 3137
+ },
+ {
+ "epoch": 10.02555910543131,
+ "grad_norm": 0.12255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0659,
+ "step": 3138
+ },
+ {
+ "epoch": 10.028753993610223,
+ "grad_norm": 0.0849609375,
+ "learning_rate": 0.0005,
+ "loss": 1.046,
+ "step": 3139
+ },
+ {
+ "epoch": 10.031948881789138,
+ "grad_norm": 0.10009765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0547,
+ "step": 3140
+ },
+ {
+ "epoch": 10.035143769968052,
+ "grad_norm": 0.1689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0673,
+ "step": 3141
+ },
+ {
+ "epoch": 10.038338658146964,
+ "grad_norm": 0.0888671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0699,
+ "step": 3142
+ },
+ {
+ "epoch": 10.041533546325878,
+ "grad_norm": 0.1650390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0633,
+ "step": 3143
+ },
+ {
+ "epoch": 10.044728434504792,
+ "grad_norm": 0.09228515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0645,
+ "step": 3144
+ },
+ {
+ "epoch": 10.047923322683706,
+ "grad_norm": 0.12255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0685,
+ "step": 3145
+ },
+ {
+ "epoch": 10.05111821086262,
+ "grad_norm": 0.1005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0643,
+ "step": 3146
+ },
+ {
+ "epoch": 10.054313099041533,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0525,
+ "step": 3147
+ },
+ {
+ "epoch": 10.057507987220447,
+ "grad_norm": 0.1103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0661,
+ "step": 3148
+ },
+ {
+ "epoch": 10.060702875399361,
+ "grad_norm": 0.11181640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0544,
+ "step": 3149
+ },
+ {
+ "epoch": 10.063897763578275,
+ "grad_norm": 0.1015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0565,
+ "step": 3150
+ },
+ {
+ "epoch": 10.06709265175719,
+ "grad_norm": 0.1279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0579,
+ "step": 3151
+ },
+ {
+ "epoch": 10.070287539936102,
+ "grad_norm": 0.083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0559,
+ "step": 3152
+ },
+ {
+ "epoch": 10.073482428115016,
+ "grad_norm": 0.1064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.054,
+ "step": 3153
+ },
+ {
+ "epoch": 10.07667731629393,
+ "grad_norm": 0.1123046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0646,
+ "step": 3154
+ },
+ {
+ "epoch": 10.079872204472844,
+ "grad_norm": 0.126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0598,
+ "step": 3155
+ },
+ {
+ "epoch": 10.083067092651758,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0564,
+ "step": 3156
+ },
+ {
+ "epoch": 10.08626198083067,
+ "grad_norm": 0.09814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0606,
+ "step": 3157
+ },
+ {
+ "epoch": 10.089456869009584,
+ "grad_norm": 0.0966796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0611,
+ "step": 3158
+ },
+ {
+ "epoch": 10.092651757188499,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0632,
+ "step": 3159
+ },
+ {
+ "epoch": 10.095846645367413,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0545,
+ "step": 3160
+ },
+ {
+ "epoch": 10.099041533546325,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.058,
+ "step": 3161
+ },
+ {
+ "epoch": 10.10223642172524,
+ "grad_norm": 0.059326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0649,
+ "step": 3162
+ },
+ {
+ "epoch": 10.105431309904153,
+ "grad_norm": 0.0966796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0568,
+ "step": 3163
+ },
+ {
+ "epoch": 10.108626198083067,
+ "grad_norm": 0.06005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0627,
+ "step": 3164
+ },
+ {
+ "epoch": 10.111821086261982,
+ "grad_norm": 0.0810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0575,
+ "step": 3165
+ },
+ {
+ "epoch": 10.115015974440894,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0529,
+ "step": 3166
+ },
+ {
+ "epoch": 10.118210862619808,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0551,
+ "step": 3167
+ },
+ {
+ "epoch": 10.121405750798722,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0748,
+ "step": 3168
+ },
+ {
+ "epoch": 10.124600638977636,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0644,
+ "step": 3169
+ },
+ {
+ "epoch": 10.12779552715655,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0519,
+ "step": 3170
+ },
+ {
+ "epoch": 10.130990415335463,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0591,
+ "step": 3171
+ },
+ {
+ "epoch": 10.134185303514377,
+ "grad_norm": 0.057373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0719,
+ "step": 3172
+ },
+ {
+ "epoch": 10.13738019169329,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0628,
+ "step": 3173
+ },
+ {
+ "epoch": 10.140575079872205,
+ "grad_norm": 0.059326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.048,
+ "step": 3174
+ },
+ {
+ "epoch": 10.143769968051119,
+ "grad_norm": 0.083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0706,
+ "step": 3175
+ },
+ {
+ "epoch": 10.146964856230031,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.062,
+ "step": 3176
+ },
+ {
+ "epoch": 10.150159744408946,
+ "grad_norm": 0.08935546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0644,
+ "step": 3177
+ },
+ {
+ "epoch": 10.15335463258786,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.059,
+ "step": 3178
+ },
+ {
+ "epoch": 10.156549520766774,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0705,
+ "step": 3179
+ },
+ {
+ "epoch": 10.159744408945686,
+ "grad_norm": 0.060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.058,
+ "step": 3180
+ },
+ {
+ "epoch": 10.1629392971246,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0613,
+ "step": 3181
+ },
+ {
+ "epoch": 10.166134185303514,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0623,
+ "step": 3182
+ },
+ {
+ "epoch": 10.169329073482428,
+ "grad_norm": 0.057373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0612,
+ "step": 3183
+ },
+ {
+ "epoch": 10.172523961661343,
+ "grad_norm": 0.05908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0532,
+ "step": 3184
+ },
+ {
+ "epoch": 10.175718849840255,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0635,
+ "step": 3185
+ },
+ {
+ "epoch": 10.178913738019169,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0723,
+ "step": 3186
+ },
+ {
+ "epoch": 10.182108626198083,
+ "grad_norm": 0.09716796875,
+ "learning_rate": 0.0005,
+ "loss": 1.065,
+ "step": 3187
+ },
+ {
+ "epoch": 10.185303514376997,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.063,
+ "step": 3188
+ },
+ {
+ "epoch": 10.188498402555911,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0647,
+ "step": 3189
+ },
+ {
+ "epoch": 10.191693290734824,
+ "grad_norm": 0.08349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0688,
+ "step": 3190
+ },
+ {
+ "epoch": 10.194888178913738,
+ "grad_norm": 0.06005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0611,
+ "step": 3191
+ },
+ {
+ "epoch": 10.198083067092652,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0507,
+ "step": 3192
+ },
+ {
+ "epoch": 10.201277955271566,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0671,
+ "step": 3193
+ },
+ {
+ "epoch": 10.204472843450478,
+ "grad_norm": 0.080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0586,
+ "step": 3194
+ },
+ {
+ "epoch": 10.207667731629392,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0609,
+ "step": 3195
+ },
+ {
+ "epoch": 10.210862619808307,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0595,
+ "step": 3196
+ },
+ {
+ "epoch": 10.21405750798722,
+ "grad_norm": 0.099609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0494,
+ "step": 3197
+ },
+ {
+ "epoch": 10.217252396166135,
+ "grad_norm": 0.059326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0463,
+ "step": 3198
+ },
+ {
+ "epoch": 10.220447284345047,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.056,
+ "step": 3199
+ },
+ {
+ "epoch": 10.223642172523961,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0587,
+ "step": 3200
+ },
+ {
+ "epoch": 10.226837060702875,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0595,
+ "step": 3201
+ },
+ {
+ "epoch": 10.23003194888179,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0678,
+ "step": 3202
+ },
+ {
+ "epoch": 10.233226837060704,
+ "grad_norm": 0.08203125,
+ "learning_rate": 0.0005,
+ "loss": 1.054,
+ "step": 3203
+ },
+ {
+ "epoch": 10.236421725239616,
+ "grad_norm": 0.0947265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0571,
+ "step": 3204
+ },
+ {
+ "epoch": 10.23961661341853,
+ "grad_norm": 0.16796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0546,
+ "step": 3205
+ },
+ {
+ "epoch": 10.242811501597444,
+ "grad_norm": 0.09423828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0657,
+ "step": 3206
+ },
+ {
+ "epoch": 10.246006389776358,
+ "grad_norm": 0.208984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0668,
+ "step": 3207
+ },
+ {
+ "epoch": 10.249201277955272,
+ "grad_norm": 0.1201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0565,
+ "step": 3208
+ },
+ {
+ "epoch": 10.252396166134185,
+ "grad_norm": 0.1005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0477,
+ "step": 3209
+ },
+ {
+ "epoch": 10.255591054313099,
+ "grad_norm": 0.11181640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0625,
+ "step": 3210
+ },
+ {
+ "epoch": 10.258785942492013,
+ "grad_norm": 0.08837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0532,
+ "step": 3211
+ },
+ {
+ "epoch": 10.261980830670927,
+ "grad_norm": 0.09716796875,
+ "learning_rate": 0.0005,
+ "loss": 1.062,
+ "step": 3212
+ },
+ {
+ "epoch": 10.26517571884984,
+ "grad_norm": 0.0615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0592,
+ "step": 3213
+ },
+ {
+ "epoch": 10.268370607028753,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0595,
+ "step": 3214
+ },
+ {
+ "epoch": 10.271565495207668,
+ "grad_norm": 0.1337890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0604,
+ "step": 3215
+ },
+ {
+ "epoch": 10.274760383386582,
+ "grad_norm": 0.14453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0617,
+ "step": 3216
+ },
+ {
+ "epoch": 10.277955271565496,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0637,
+ "step": 3217
+ },
+ {
+ "epoch": 10.281150159744408,
+ "grad_norm": 0.123046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0599,
+ "step": 3218
+ },
+ {
+ "epoch": 10.284345047923322,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0705,
+ "step": 3219
+ },
+ {
+ "epoch": 10.287539936102236,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0691,
+ "step": 3220
+ },
+ {
+ "epoch": 10.29073482428115,
+ "grad_norm": 0.062255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0674,
+ "step": 3221
+ },
+ {
+ "epoch": 10.293929712460065,
+ "grad_norm": 0.11376953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0611,
+ "step": 3222
+ },
+ {
+ "epoch": 10.297124600638977,
+ "grad_norm": 0.07861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0535,
+ "step": 3223
+ },
+ {
+ "epoch": 10.300319488817891,
+ "grad_norm": 0.08251953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0585,
+ "step": 3224
+ },
+ {
+ "epoch": 10.303514376996805,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.061,
+ "step": 3225
+ },
+ {
+ "epoch": 10.30670926517572,
+ "grad_norm": 0.059814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0606,
+ "step": 3226
+ },
+ {
+ "epoch": 10.309904153354633,
+ "grad_norm": 0.07861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0437,
+ "step": 3227
+ },
+ {
+ "epoch": 10.313099041533546,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0558,
+ "step": 3228
+ },
+ {
+ "epoch": 10.31629392971246,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0515,
+ "step": 3229
+ },
+ {
+ "epoch": 10.319488817891374,
+ "grad_norm": 0.125,
+ "learning_rate": 0.0005,
+ "loss": 1.0527,
+ "step": 3230
+ },
+ {
+ "epoch": 10.322683706070288,
+ "grad_norm": 0.10009765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0559,
+ "step": 3231
+ },
+ {
+ "epoch": 10.3258785942492,
+ "grad_norm": 0.10693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0652,
+ "step": 3232
+ },
+ {
+ "epoch": 10.329073482428115,
+ "grad_norm": 0.1083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0787,
+ "step": 3233
+ },
+ {
+ "epoch": 10.332268370607029,
+ "grad_norm": 0.10693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0678,
+ "step": 3234
+ },
+ {
+ "epoch": 10.335463258785943,
+ "grad_norm": 0.1474609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0469,
+ "step": 3235
+ },
+ {
+ "epoch": 10.338658146964857,
+ "grad_norm": 0.083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0524,
+ "step": 3236
+ },
+ {
+ "epoch": 10.34185303514377,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0592,
+ "step": 3237
+ },
+ {
+ "epoch": 10.345047923322683,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0638,
+ "step": 3238
+ },
+ {
+ "epoch": 10.348242811501597,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0616,
+ "step": 3239
+ },
+ {
+ "epoch": 10.351437699680512,
+ "grad_norm": 0.083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0405,
+ "step": 3240
+ },
+ {
+ "epoch": 10.354632587859426,
+ "grad_norm": 0.08984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0675,
+ "step": 3241
+ },
+ {
+ "epoch": 10.357827476038338,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0727,
+ "step": 3242
+ },
+ {
+ "epoch": 10.361022364217252,
+ "grad_norm": 0.107421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0589,
+ "step": 3243
+ },
+ {
+ "epoch": 10.364217252396166,
+ "grad_norm": 0.061767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0518,
+ "step": 3244
+ },
+ {
+ "epoch": 10.36741214057508,
+ "grad_norm": 0.1376953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0727,
+ "step": 3245
+ },
+ {
+ "epoch": 10.370607028753994,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0423,
+ "step": 3246
+ },
+ {
+ "epoch": 10.373801916932907,
+ "grad_norm": 0.12255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0555,
+ "step": 3247
+ },
+ {
+ "epoch": 10.37699680511182,
+ "grad_norm": 0.0810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0598,
+ "step": 3248
+ },
+ {
+ "epoch": 10.380191693290735,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0539,
+ "step": 3249
+ },
+ {
+ "epoch": 10.383386581469649,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0637,
+ "step": 3250
+ },
+ {
+ "epoch": 10.386581469648561,
+ "grad_norm": 0.0908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.06,
+ "step": 3251
+ },
+ {
+ "epoch": 10.389776357827476,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0823,
+ "step": 3252
+ },
+ {
+ "epoch": 10.39297124600639,
+ "grad_norm": 0.0869140625,
+ "learning_rate": 0.0005,
+ "loss": 1.064,
+ "step": 3253
+ },
+ {
+ "epoch": 10.396166134185304,
+ "grad_norm": 0.058837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0628,
+ "step": 3254
+ },
+ {
+ "epoch": 10.399361022364218,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0549,
+ "step": 3255
+ },
+ {
+ "epoch": 10.40255591054313,
+ "grad_norm": 0.08984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0645,
+ "step": 3256
+ },
+ {
+ "epoch": 10.405750798722044,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.053,
+ "step": 3257
+ },
+ {
+ "epoch": 10.408945686900958,
+ "grad_norm": 0.09521484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0585,
+ "step": 3258
+ },
+ {
+ "epoch": 10.412140575079873,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.057,
+ "step": 3259
+ },
+ {
+ "epoch": 10.415335463258787,
+ "grad_norm": 0.1455078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0673,
+ "step": 3260
+ },
+ {
+ "epoch": 10.418530351437699,
+ "grad_norm": 0.1748046875,
+ "learning_rate": 0.0005,
+ "loss": 1.056,
+ "step": 3261
+ },
+ {
+ "epoch": 10.421725239616613,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0543,
+ "step": 3262
+ },
+ {
+ "epoch": 10.424920127795527,
+ "grad_norm": 0.11181640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0623,
+ "step": 3263
+ },
+ {
+ "epoch": 10.428115015974441,
+ "grad_norm": 0.099609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0594,
+ "step": 3264
+ },
+ {
+ "epoch": 10.431309904153355,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0544,
+ "step": 3265
+ },
+ {
+ "epoch": 10.434504792332268,
+ "grad_norm": 0.10986328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0593,
+ "step": 3266
+ },
+ {
+ "epoch": 10.437699680511182,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0632,
+ "step": 3267
+ },
+ {
+ "epoch": 10.440894568690096,
+ "grad_norm": 0.10595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0512,
+ "step": 3268
+ },
+ {
+ "epoch": 10.44408945686901,
+ "grad_norm": 0.1357421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0666,
+ "step": 3269
+ },
+ {
+ "epoch": 10.447284345047922,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0618,
+ "step": 3270
+ },
+ {
+ "epoch": 10.450479233226837,
+ "grad_norm": 0.1640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0657,
+ "step": 3271
+ },
+ {
+ "epoch": 10.45367412140575,
+ "grad_norm": 0.1181640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0443,
+ "step": 3272
+ },
+ {
+ "epoch": 10.456869009584665,
+ "grad_norm": 0.09326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.067,
+ "step": 3273
+ },
+ {
+ "epoch": 10.460063897763579,
+ "grad_norm": 0.154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0484,
+ "step": 3274
+ },
+ {
+ "epoch": 10.463258785942491,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0619,
+ "step": 3275
+ },
+ {
+ "epoch": 10.466453674121405,
+ "grad_norm": 0.1708984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0771,
+ "step": 3276
+ },
+ {
+ "epoch": 10.46964856230032,
+ "grad_norm": 0.1728515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0545,
+ "step": 3277
+ },
+ {
+ "epoch": 10.472843450479234,
+ "grad_norm": 0.062255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0607,
+ "step": 3278
+ },
+ {
+ "epoch": 10.476038338658148,
+ "grad_norm": 0.17578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0643,
+ "step": 3279
+ },
+ {
+ "epoch": 10.47923322683706,
+ "grad_norm": 0.138671875,
+ "learning_rate": 0.0005,
+ "loss": 1.06,
+ "step": 3280
+ },
+ {
+ "epoch": 10.482428115015974,
+ "grad_norm": 0.099609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0731,
+ "step": 3281
+ },
+ {
+ "epoch": 10.485623003194888,
+ "grad_norm": 0.2060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0702,
+ "step": 3282
+ },
+ {
+ "epoch": 10.488817891373802,
+ "grad_norm": 0.1689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.081,
+ "step": 3283
+ },
+ {
+ "epoch": 10.492012779552716,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0551,
+ "step": 3284
+ },
+ {
+ "epoch": 10.495207667731629,
+ "grad_norm": 0.08349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0586,
+ "step": 3285
+ },
+ {
+ "epoch": 10.498402555910543,
+ "grad_norm": 0.1083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0721,
+ "step": 3286
+ },
+ {
+ "epoch": 10.501597444089457,
+ "grad_norm": 0.060791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0539,
+ "step": 3287
+ },
+ {
+ "epoch": 10.504792332268371,
+ "grad_norm": 0.083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0726,
+ "step": 3288
+ },
+ {
+ "epoch": 10.507987220447284,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0611,
+ "step": 3289
+ },
+ {
+ "epoch": 10.511182108626198,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0672,
+ "step": 3290
+ },
+ {
+ "epoch": 10.514376996805112,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0545,
+ "step": 3291
+ },
+ {
+ "epoch": 10.517571884984026,
+ "grad_norm": 0.060302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.066,
+ "step": 3292
+ },
+ {
+ "epoch": 10.52076677316294,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0631,
+ "step": 3293
+ },
+ {
+ "epoch": 10.523961661341852,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0543,
+ "step": 3294
+ },
+ {
+ "epoch": 10.527156549520766,
+ "grad_norm": 0.10009765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0584,
+ "step": 3295
+ },
+ {
+ "epoch": 10.53035143769968,
+ "grad_norm": 0.05517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0735,
+ "step": 3296
+ },
+ {
+ "epoch": 10.533546325878595,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0749,
+ "step": 3297
+ },
+ {
+ "epoch": 10.536741214057509,
+ "grad_norm": 0.087890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0674,
+ "step": 3298
+ },
+ {
+ "epoch": 10.539936102236421,
+ "grad_norm": 0.08056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0618,
+ "step": 3299
+ },
+ {
+ "epoch": 10.543130990415335,
+ "grad_norm": 0.06005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0592,
+ "step": 3300
+ },
+ {
+ "epoch": 10.54632587859425,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0537,
+ "step": 3301
+ },
+ {
+ "epoch": 10.549520766773163,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.05,
+ "step": 3302
+ },
+ {
+ "epoch": 10.552715654952078,
+ "grad_norm": 0.05908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.063,
+ "step": 3303
+ },
+ {
+ "epoch": 10.55591054313099,
+ "grad_norm": 0.058837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0539,
+ "step": 3304
+ },
+ {
+ "epoch": 10.559105431309904,
+ "grad_norm": 0.05078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0823,
+ "step": 3305
+ },
+ {
+ "epoch": 10.562300319488818,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0523,
+ "step": 3306
+ },
+ {
+ "epoch": 10.565495207667732,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0602,
+ "step": 3307
+ },
+ {
+ "epoch": 10.568690095846645,
+ "grad_norm": 0.05859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0762,
+ "step": 3308
+ },
+ {
+ "epoch": 10.571884984025559,
+ "grad_norm": 0.057861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0623,
+ "step": 3309
+ },
+ {
+ "epoch": 10.575079872204473,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0531,
+ "step": 3310
+ },
+ {
+ "epoch": 10.578274760383387,
+ "grad_norm": 0.062255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0693,
+ "step": 3311
+ },
+ {
+ "epoch": 10.581469648562301,
+ "grad_norm": 0.05517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0664,
+ "step": 3312
+ },
+ {
+ "epoch": 10.584664536741213,
+ "grad_norm": 0.058837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0549,
+ "step": 3313
+ },
+ {
+ "epoch": 10.587859424920127,
+ "grad_norm": 0.054931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.065,
+ "step": 3314
+ },
+ {
+ "epoch": 10.591054313099042,
+ "grad_norm": 0.0595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0572,
+ "step": 3315
+ },
+ {
+ "epoch": 10.594249201277956,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0443,
+ "step": 3316
+ },
+ {
+ "epoch": 10.59744408945687,
+ "grad_norm": 0.055419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0522,
+ "step": 3317
+ },
+ {
+ "epoch": 10.600638977635782,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0611,
+ "step": 3318
+ },
+ {
+ "epoch": 10.603833865814696,
+ "grad_norm": 0.05615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0648,
+ "step": 3319
+ },
+ {
+ "epoch": 10.60702875399361,
+ "grad_norm": 0.0615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0656,
+ "step": 3320
+ },
+ {
+ "epoch": 10.610223642172524,
+ "grad_norm": 0.062255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0591,
+ "step": 3321
+ },
+ {
+ "epoch": 10.613418530351439,
+ "grad_norm": 0.061767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0627,
+ "step": 3322
+ },
+ {
+ "epoch": 10.616613418530351,
+ "grad_norm": 0.057373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0569,
+ "step": 3323
+ },
+ {
+ "epoch": 10.619808306709265,
+ "grad_norm": 0.087890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0499,
+ "step": 3324
+ },
+ {
+ "epoch": 10.62300319488818,
+ "grad_norm": 0.056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0542,
+ "step": 3325
+ },
+ {
+ "epoch": 10.626198083067093,
+ "grad_norm": 0.0888671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0567,
+ "step": 3326
+ },
+ {
+ "epoch": 10.629392971246006,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.051,
+ "step": 3327
+ },
+ {
+ "epoch": 10.63258785942492,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0684,
+ "step": 3328
+ },
+ {
+ "epoch": 10.635782747603834,
+ "grad_norm": 0.08837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0639,
+ "step": 3329
+ },
+ {
+ "epoch": 10.638977635782748,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0576,
+ "step": 3330
+ },
+ {
+ "epoch": 10.642172523961662,
+ "grad_norm": 0.0859375,
+ "learning_rate": 0.0005,
+ "loss": 1.066,
+ "step": 3331
+ },
+ {
+ "epoch": 10.645367412140574,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0529,
+ "step": 3332
+ },
+ {
+ "epoch": 10.648562300319488,
+ "grad_norm": 0.09912109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0596,
+ "step": 3333
+ },
+ {
+ "epoch": 10.651757188498403,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0753,
+ "step": 3334
+ },
+ {
+ "epoch": 10.654952076677317,
+ "grad_norm": 0.09716796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0667,
+ "step": 3335
+ },
+ {
+ "epoch": 10.65814696485623,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0642,
+ "step": 3336
+ },
+ {
+ "epoch": 10.661341853035143,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0588,
+ "step": 3337
+ },
+ {
+ "epoch": 10.664536741214057,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0641,
+ "step": 3338
+ },
+ {
+ "epoch": 10.667731629392971,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0644,
+ "step": 3339
+ },
+ {
+ "epoch": 10.670926517571885,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.057,
+ "step": 3340
+ },
+ {
+ "epoch": 10.6741214057508,
+ "grad_norm": 0.08447265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0546,
+ "step": 3341
+ },
+ {
+ "epoch": 10.677316293929712,
+ "grad_norm": 0.13671875,
+ "learning_rate": 0.0005,
+ "loss": 1.06,
+ "step": 3342
+ },
+ {
+ "epoch": 10.680511182108626,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0618,
+ "step": 3343
+ },
+ {
+ "epoch": 10.68370607028754,
+ "grad_norm": 0.2001953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0481,
+ "step": 3344
+ },
+ {
+ "epoch": 10.686900958466454,
+ "grad_norm": 0.103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0653,
+ "step": 3345
+ },
+ {
+ "epoch": 10.690095846645367,
+ "grad_norm": 0.1396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0624,
+ "step": 3346
+ },
+ {
+ "epoch": 10.69329073482428,
+ "grad_norm": 0.19140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0585,
+ "step": 3347
+ },
+ {
+ "epoch": 10.696485623003195,
+ "grad_norm": 0.0849609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0628,
+ "step": 3348
+ },
+ {
+ "epoch": 10.699680511182109,
+ "grad_norm": 0.0849609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0465,
+ "step": 3349
+ },
+ {
+ "epoch": 10.702875399361023,
+ "grad_norm": 0.107421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0559,
+ "step": 3350
+ },
+ {
+ "epoch": 10.706070287539935,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0628,
+ "step": 3351
+ },
+ {
+ "epoch": 10.70926517571885,
+ "grad_norm": 0.2138671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0587,
+ "step": 3352
+ },
+ {
+ "epoch": 10.712460063897764,
+ "grad_norm": 0.1357421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0607,
+ "step": 3353
+ },
+ {
+ "epoch": 10.715654952076678,
+ "grad_norm": 0.11328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0526,
+ "step": 3354
+ },
+ {
+ "epoch": 10.718849840255592,
+ "grad_norm": 0.1923828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0651,
+ "step": 3355
+ },
+ {
+ "epoch": 10.722044728434504,
+ "grad_norm": 0.1728515625,
+ "learning_rate": 0.0005,
+ "loss": 1.064,
+ "step": 3356
+ },
+ {
+ "epoch": 10.725239616613418,
+ "grad_norm": 0.0556640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0482,
+ "step": 3357
+ },
+ {
+ "epoch": 10.728434504792332,
+ "grad_norm": 0.16015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0569,
+ "step": 3358
+ },
+ {
+ "epoch": 10.731629392971247,
+ "grad_norm": 0.08740234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0514,
+ "step": 3359
+ },
+ {
+ "epoch": 10.73482428115016,
+ "grad_norm": 0.0908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0789,
+ "step": 3360
+ },
+ {
+ "epoch": 10.738019169329073,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0689,
+ "step": 3361
+ },
+ {
+ "epoch": 10.741214057507987,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0639,
+ "step": 3362
+ },
+ {
+ "epoch": 10.744408945686901,
+ "grad_norm": 0.09423828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0569,
+ "step": 3363
+ },
+ {
+ "epoch": 10.747603833865815,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0473,
+ "step": 3364
+ },
+ {
+ "epoch": 10.750798722044728,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0519,
+ "step": 3365
+ },
+ {
+ "epoch": 10.753993610223642,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0657,
+ "step": 3366
+ },
+ {
+ "epoch": 10.757188498402556,
+ "grad_norm": 0.0810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0642,
+ "step": 3367
+ },
+ {
+ "epoch": 10.76038338658147,
+ "grad_norm": 0.10986328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0586,
+ "step": 3368
+ },
+ {
+ "epoch": 10.763578274760384,
+ "grad_norm": 0.1259765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0685,
+ "step": 3369
+ },
+ {
+ "epoch": 10.766773162939296,
+ "grad_norm": 0.07861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0618,
+ "step": 3370
+ },
+ {
+ "epoch": 10.76996805111821,
+ "grad_norm": 0.08154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0635,
+ "step": 3371
+ },
+ {
+ "epoch": 10.773162939297125,
+ "grad_norm": 0.11474609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0651,
+ "step": 3372
+ },
+ {
+ "epoch": 10.776357827476039,
+ "grad_norm": 0.060791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0586,
+ "step": 3373
+ },
+ {
+ "epoch": 10.779552715654953,
+ "grad_norm": 0.09521484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0582,
+ "step": 3374
+ },
+ {
+ "epoch": 10.782747603833865,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0527,
+ "step": 3375
+ },
+ {
+ "epoch": 10.78594249201278,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0658,
+ "step": 3376
+ },
+ {
+ "epoch": 10.789137380191693,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0527,
+ "step": 3377
+ },
+ {
+ "epoch": 10.792332268370608,
+ "grad_norm": 0.06103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0609,
+ "step": 3378
+ },
+ {
+ "epoch": 10.795527156549522,
+ "grad_norm": 0.05810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0686,
+ "step": 3379
+ },
+ {
+ "epoch": 10.798722044728434,
+ "grad_norm": 0.08349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0596,
+ "step": 3380
+ },
+ {
+ "epoch": 10.801916932907348,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0496,
+ "step": 3381
+ },
+ {
+ "epoch": 10.805111821086262,
+ "grad_norm": 0.1337890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0577,
+ "step": 3382
+ },
+ {
+ "epoch": 10.808306709265176,
+ "grad_norm": 0.08935546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0551,
+ "step": 3383
+ },
+ {
+ "epoch": 10.811501597444089,
+ "grad_norm": 0.09326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.061,
+ "step": 3384
+ },
+ {
+ "epoch": 10.814696485623003,
+ "grad_norm": 0.12158203125,
+ "learning_rate": 0.0005,
+ "loss": 1.062,
+ "step": 3385
+ },
+ {
+ "epoch": 10.817891373801917,
+ "grad_norm": 0.14453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0588,
+ "step": 3386
+ },
+ {
+ "epoch": 10.821086261980831,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0682,
+ "step": 3387
+ },
+ {
+ "epoch": 10.824281150159745,
+ "grad_norm": 0.08447265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0624,
+ "step": 3388
+ },
+ {
+ "epoch": 10.827476038338657,
+ "grad_norm": 0.10986328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0662,
+ "step": 3389
+ },
+ {
+ "epoch": 10.830670926517572,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.07,
+ "step": 3390
+ },
+ {
+ "epoch": 10.833865814696486,
+ "grad_norm": 0.140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0612,
+ "step": 3391
+ },
+ {
+ "epoch": 10.8370607028754,
+ "grad_norm": 0.061767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0649,
+ "step": 3392
+ },
+ {
+ "epoch": 10.840255591054314,
+ "grad_norm": 0.12353515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0614,
+ "step": 3393
+ },
+ {
+ "epoch": 10.843450479233226,
+ "grad_norm": 0.08837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0569,
+ "step": 3394
+ },
+ {
+ "epoch": 10.84664536741214,
+ "grad_norm": 0.05810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0617,
+ "step": 3395
+ },
+ {
+ "epoch": 10.849840255591054,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.051,
+ "step": 3396
+ },
+ {
+ "epoch": 10.853035143769969,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0408,
+ "step": 3397
+ },
+ {
+ "epoch": 10.856230031948883,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0611,
+ "step": 3398
+ },
+ {
+ "epoch": 10.859424920127795,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0505,
+ "step": 3399
+ },
+ {
+ "epoch": 10.86261980830671,
+ "grad_norm": 0.058837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0689,
+ "step": 3400
+ },
+ {
+ "epoch": 10.865814696485623,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0698,
+ "step": 3401
+ },
+ {
+ "epoch": 10.869009584664537,
+ "grad_norm": 0.10205078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0485,
+ "step": 3402
+ },
+ {
+ "epoch": 10.87220447284345,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0547,
+ "step": 3403
+ },
+ {
+ "epoch": 10.875399361022364,
+ "grad_norm": 0.0576171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0687,
+ "step": 3404
+ },
+ {
+ "epoch": 10.878594249201278,
+ "grad_norm": 0.05322265625,
+ "learning_rate": 0.0005,
+ "loss": 1.058,
+ "step": 3405
+ },
+ {
+ "epoch": 10.881789137380192,
+ "grad_norm": 0.061279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0476,
+ "step": 3406
+ },
+ {
+ "epoch": 10.884984025559106,
+ "grad_norm": 0.058837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0674,
+ "step": 3407
+ },
+ {
+ "epoch": 10.888178913738018,
+ "grad_norm": 0.107421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0663,
+ "step": 3408
+ },
+ {
+ "epoch": 10.891373801916933,
+ "grad_norm": 0.06005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0651,
+ "step": 3409
+ },
+ {
+ "epoch": 10.894568690095847,
+ "grad_norm": 0.0986328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0687,
+ "step": 3410
+ },
+ {
+ "epoch": 10.89776357827476,
+ "grad_norm": 0.09375,
+ "learning_rate": 0.0005,
+ "loss": 1.0615,
+ "step": 3411
+ },
+ {
+ "epoch": 10.900958466453675,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0645,
+ "step": 3412
+ },
+ {
+ "epoch": 10.904153354632587,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0665,
+ "step": 3413
+ },
+ {
+ "epoch": 10.907348242811501,
+ "grad_norm": 0.09375,
+ "learning_rate": 0.0005,
+ "loss": 1.0486,
+ "step": 3414
+ },
+ {
+ "epoch": 10.910543130990416,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0548,
+ "step": 3415
+ },
+ {
+ "epoch": 10.91373801916933,
+ "grad_norm": 0.1181640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0639,
+ "step": 3416
+ },
+ {
+ "epoch": 10.916932907348244,
+ "grad_norm": 0.11767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0516,
+ "step": 3417
+ },
+ {
+ "epoch": 10.920127795527156,
+ "grad_norm": 0.0849609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0509,
+ "step": 3418
+ },
+ {
+ "epoch": 10.92332268370607,
+ "grad_norm": 0.1826171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0573,
+ "step": 3419
+ },
+ {
+ "epoch": 10.926517571884984,
+ "grad_norm": 0.12255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0537,
+ "step": 3420
+ },
+ {
+ "epoch": 10.929712460063898,
+ "grad_norm": 0.1328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0676,
+ "step": 3421
+ },
+ {
+ "epoch": 10.93290734824281,
+ "grad_norm": 0.2890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0608,
+ "step": 3422
+ },
+ {
+ "epoch": 10.936102236421725,
+ "grad_norm": 0.26953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0542,
+ "step": 3423
+ },
+ {
+ "epoch": 10.939297124600639,
+ "grad_norm": 0.10302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.071,
+ "step": 3424
+ },
+ {
+ "epoch": 10.942492012779553,
+ "grad_norm": 0.158203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0624,
+ "step": 3425
+ },
+ {
+ "epoch": 10.945686900958467,
+ "grad_norm": 0.2490234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0618,
+ "step": 3426
+ },
+ {
+ "epoch": 10.94888178913738,
+ "grad_norm": 0.11962890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0554,
+ "step": 3427
+ },
+ {
+ "epoch": 10.952076677316294,
+ "grad_norm": 0.1103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0622,
+ "step": 3428
+ },
+ {
+ "epoch": 10.955271565495208,
+ "grad_norm": 0.18359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0653,
+ "step": 3429
+ },
+ {
+ "epoch": 10.958466453674122,
+ "grad_norm": 0.125,
+ "learning_rate": 0.0005,
+ "loss": 1.0624,
+ "step": 3430
+ },
+ {
+ "epoch": 10.961661341853034,
+ "grad_norm": 0.10595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0729,
+ "step": 3431
+ },
+ {
+ "epoch": 10.964856230031948,
+ "grad_norm": 0.181640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0675,
+ "step": 3432
+ },
+ {
+ "epoch": 10.968051118210862,
+ "grad_norm": 0.06201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.071,
+ "step": 3433
+ },
+ {
+ "epoch": 10.971246006389777,
+ "grad_norm": 0.11962890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0553,
+ "step": 3434
+ },
+ {
+ "epoch": 10.97444089456869,
+ "grad_norm": 0.1337890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0458,
+ "step": 3435
+ },
+ {
+ "epoch": 10.977635782747605,
+ "grad_norm": 0.05810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0642,
+ "step": 3436
+ },
+ {
+ "epoch": 10.980830670926517,
+ "grad_norm": 0.12890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0585,
+ "step": 3437
+ },
+ {
+ "epoch": 10.984025559105431,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0657,
+ "step": 3438
+ },
+ {
+ "epoch": 10.987220447284345,
+ "grad_norm": 0.09619140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0653,
+ "step": 3439
+ },
+ {
+ "epoch": 10.99041533546326,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0619,
+ "step": 3440
+ },
+ {
+ "epoch": 10.993610223642172,
+ "grad_norm": 0.10888671875,
+ "learning_rate": 0.0005,
+ "loss": 1.056,
+ "step": 3441
+ },
+ {
+ "epoch": 10.996805111821086,
+ "grad_norm": 0.15234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0537,
+ "step": 3442
+ },
+ {
+ "epoch": 11.0,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0601,
+ "step": 3443
+ },
+ {
+ "epoch": 11.003194888178914,
+ "grad_norm": 0.1396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0577,
+ "step": 3444
+ },
+ {
+ "epoch": 11.006389776357828,
+ "grad_norm": 0.1142578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0553,
+ "step": 3445
+ },
+ {
+ "epoch": 11.00958466453674,
+ "grad_norm": 0.193359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0651,
+ "step": 3446
+ },
+ {
+ "epoch": 11.012779552715655,
+ "grad_norm": 0.1181640625,
+ "learning_rate": 0.0005,
+ "loss": 1.049,
+ "step": 3447
+ },
+ {
+ "epoch": 11.015974440894569,
+ "grad_norm": 0.1171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0568,
+ "step": 3448
+ },
+ {
+ "epoch": 11.019169329073483,
+ "grad_norm": 0.21484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0662,
+ "step": 3449
+ },
+ {
+ "epoch": 11.022364217252397,
+ "grad_norm": 0.12353515625,
+ "learning_rate": 0.0005,
+ "loss": 1.059,
+ "step": 3450
+ },
+ {
+ "epoch": 11.02555910543131,
+ "grad_norm": 0.087890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0569,
+ "step": 3451
+ },
+ {
+ "epoch": 11.028753993610223,
+ "grad_norm": 0.162109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0621,
+ "step": 3452
+ },
+ {
+ "epoch": 11.031948881789138,
+ "grad_norm": 0.15234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0547,
+ "step": 3453
+ },
+ {
+ "epoch": 11.035143769968052,
+ "grad_norm": 0.09228515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0563,
+ "step": 3454
+ },
+ {
+ "epoch": 11.038338658146964,
+ "grad_norm": 0.15625,
+ "learning_rate": 0.0005,
+ "loss": 1.0639,
+ "step": 3455
+ },
+ {
+ "epoch": 11.041533546325878,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.064,
+ "step": 3456
+ },
+ {
+ "epoch": 11.044728434504792,
+ "grad_norm": 0.11083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0611,
+ "step": 3457
+ },
+ {
+ "epoch": 11.047923322683706,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0562,
+ "step": 3458
+ },
+ {
+ "epoch": 11.05111821086262,
+ "grad_norm": 0.087890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0501,
+ "step": 3459
+ },
+ {
+ "epoch": 11.054313099041533,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0526,
+ "step": 3460
+ },
+ {
+ "epoch": 11.057507987220447,
+ "grad_norm": 0.05615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0712,
+ "step": 3461
+ },
+ {
+ "epoch": 11.060702875399361,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0543,
+ "step": 3462
+ },
+ {
+ "epoch": 11.063897763578275,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0603,
+ "step": 3463
+ },
+ {
+ "epoch": 11.06709265175719,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0561,
+ "step": 3464
+ },
+ {
+ "epoch": 11.070287539936102,
+ "grad_norm": 0.09716796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0587,
+ "step": 3465
+ },
+ {
+ "epoch": 11.073482428115016,
+ "grad_norm": 0.0927734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0527,
+ "step": 3466
+ },
+ {
+ "epoch": 11.07667731629393,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0688,
+ "step": 3467
+ },
+ {
+ "epoch": 11.079872204472844,
+ "grad_norm": 0.12890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0488,
+ "step": 3468
+ },
+ {
+ "epoch": 11.083067092651758,
+ "grad_norm": 0.07861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0598,
+ "step": 3469
+ },
+ {
+ "epoch": 11.08626198083067,
+ "grad_norm": 0.09130859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0664,
+ "step": 3470
+ },
+ {
+ "epoch": 11.089456869009584,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0525,
+ "step": 3471
+ },
+ {
+ "epoch": 11.092651757188499,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0506,
+ "step": 3472
+ },
+ {
+ "epoch": 11.095846645367413,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0621,
+ "step": 3473
+ },
+ {
+ "epoch": 11.099041533546325,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0562,
+ "step": 3474
+ },
+ {
+ "epoch": 11.10223642172524,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0443,
+ "step": 3475
+ },
+ {
+ "epoch": 11.105431309904153,
+ "grad_norm": 0.091796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0692,
+ "step": 3476
+ },
+ {
+ "epoch": 11.108626198083067,
+ "grad_norm": 0.053955078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0781,
+ "step": 3477
+ },
+ {
+ "epoch": 11.111821086261982,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0673,
+ "step": 3478
+ },
+ {
+ "epoch": 11.115015974440894,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0655,
+ "step": 3479
+ },
+ {
+ "epoch": 11.118210862619808,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0601,
+ "step": 3480
+ },
+ {
+ "epoch": 11.121405750798722,
+ "grad_norm": 0.0810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0527,
+ "step": 3481
+ },
+ {
+ "epoch": 11.124600638977636,
+ "grad_norm": 0.054443359375,
+ "learning_rate": 0.0005,
+ "loss": 1.06,
+ "step": 3482
+ },
+ {
+ "epoch": 11.12779552715655,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0595,
+ "step": 3483
+ },
+ {
+ "epoch": 11.130990415335463,
+ "grad_norm": 0.0556640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0615,
+ "step": 3484
+ },
+ {
+ "epoch": 11.134185303514377,
+ "grad_norm": 0.0927734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0665,
+ "step": 3485
+ },
+ {
+ "epoch": 11.13738019169329,
+ "grad_norm": 0.06005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0679,
+ "step": 3486
+ },
+ {
+ "epoch": 11.140575079872205,
+ "grad_norm": 0.08154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0561,
+ "step": 3487
+ },
+ {
+ "epoch": 11.143769968051119,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0641,
+ "step": 3488
+ },
+ {
+ "epoch": 11.146964856230031,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0612,
+ "step": 3489
+ },
+ {
+ "epoch": 11.150159744408946,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0634,
+ "step": 3490
+ },
+ {
+ "epoch": 11.15335463258786,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0502,
+ "step": 3491
+ },
+ {
+ "epoch": 11.156549520766774,
+ "grad_norm": 0.138671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0528,
+ "step": 3492
+ },
+ {
+ "epoch": 11.159744408945686,
+ "grad_norm": 0.080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0649,
+ "step": 3493
+ },
+ {
+ "epoch": 11.1629392971246,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0715,
+ "step": 3494
+ },
+ {
+ "epoch": 11.166134185303514,
+ "grad_norm": 0.057861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0629,
+ "step": 3495
+ },
+ {
+ "epoch": 11.169329073482428,
+ "grad_norm": 0.0537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0562,
+ "step": 3496
+ },
+ {
+ "epoch": 11.172523961661343,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0575,
+ "step": 3497
+ },
+ {
+ "epoch": 11.175718849840255,
+ "grad_norm": 0.057861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0688,
+ "step": 3498
+ },
+ {
+ "epoch": 11.178913738019169,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0623,
+ "step": 3499
+ },
+ {
+ "epoch": 11.182108626198083,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0624,
+ "step": 3500
+ },
+ {
+ "epoch": 11.185303514376997,
+ "grad_norm": 0.166015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0556,
+ "step": 3501
+ },
+ {
+ "epoch": 11.188498402555911,
+ "grad_norm": 0.11669921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0719,
+ "step": 3502
+ },
+ {
+ "epoch": 11.191693290734824,
+ "grad_norm": 0.08837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.066,
+ "step": 3503
+ },
+ {
+ "epoch": 11.194888178913738,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0653,
+ "step": 3504
+ },
+ {
+ "epoch": 11.198083067092652,
+ "grad_norm": 0.080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0628,
+ "step": 3505
+ },
+ {
+ "epoch": 11.201277955271566,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0428,
+ "step": 3506
+ },
+ {
+ "epoch": 11.204472843450478,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0649,
+ "step": 3507
+ },
+ {
+ "epoch": 11.207667731629392,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0586,
+ "step": 3508
+ },
+ {
+ "epoch": 11.210862619808307,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.039,
+ "step": 3509
+ },
+ {
+ "epoch": 11.21405750798722,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0563,
+ "step": 3510
+ },
+ {
+ "epoch": 11.217252396166135,
+ "grad_norm": 0.056884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.059,
+ "step": 3511
+ },
+ {
+ "epoch": 11.220447284345047,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0551,
+ "step": 3512
+ },
+ {
+ "epoch": 11.223642172523961,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0572,
+ "step": 3513
+ },
+ {
+ "epoch": 11.226837060702875,
+ "grad_norm": 0.056396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0533,
+ "step": 3514
+ },
+ {
+ "epoch": 11.23003194888179,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0519,
+ "step": 3515
+ },
+ {
+ "epoch": 11.233226837060704,
+ "grad_norm": 0.0859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0442,
+ "step": 3516
+ },
+ {
+ "epoch": 11.236421725239616,
+ "grad_norm": 0.054443359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0668,
+ "step": 3517
+ },
+ {
+ "epoch": 11.23961661341853,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0594,
+ "step": 3518
+ },
+ {
+ "epoch": 11.242811501597444,
+ "grad_norm": 0.058837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0604,
+ "step": 3519
+ },
+ {
+ "epoch": 11.246006389776358,
+ "grad_norm": 0.0830078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0524,
+ "step": 3520
+ },
+ {
+ "epoch": 11.249201277955272,
+ "grad_norm": 0.062255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0623,
+ "step": 3521
+ },
+ {
+ "epoch": 11.252396166134185,
+ "grad_norm": 0.08056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.054,
+ "step": 3522
+ },
+ {
+ "epoch": 11.255591054313099,
+ "grad_norm": 0.09619140625,
+ "learning_rate": 0.0005,
+ "loss": 1.058,
+ "step": 3523
+ },
+ {
+ "epoch": 11.258785942492013,
+ "grad_norm": 0.054443359375,
+ "learning_rate": 0.0005,
+ "loss": 1.057,
+ "step": 3524
+ },
+ {
+ "epoch": 11.261980830670927,
+ "grad_norm": 0.09716796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0652,
+ "step": 3525
+ },
+ {
+ "epoch": 11.26517571884984,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.058,
+ "step": 3526
+ },
+ {
+ "epoch": 11.268370607028753,
+ "grad_norm": 0.10546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0653,
+ "step": 3527
+ },
+ {
+ "epoch": 11.271565495207668,
+ "grad_norm": 0.1640625,
+ "learning_rate": 0.0005,
+ "loss": 1.055,
+ "step": 3528
+ },
+ {
+ "epoch": 11.274760383386582,
+ "grad_norm": 0.15625,
+ "learning_rate": 0.0005,
+ "loss": 1.0565,
+ "step": 3529
+ },
+ {
+ "epoch": 11.277955271565496,
+ "grad_norm": 0.08349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0657,
+ "step": 3530
+ },
+ {
+ "epoch": 11.281150159744408,
+ "grad_norm": 0.099609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0564,
+ "step": 3531
+ },
+ {
+ "epoch": 11.284345047923322,
+ "grad_norm": 0.0810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0515,
+ "step": 3532
+ },
+ {
+ "epoch": 11.287539936102236,
+ "grad_norm": 0.083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0643,
+ "step": 3533
+ },
+ {
+ "epoch": 11.29073482428115,
+ "grad_norm": 0.08740234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0566,
+ "step": 3534
+ },
+ {
+ "epoch": 11.293929712460065,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0488,
+ "step": 3535
+ },
+ {
+ "epoch": 11.297124600638977,
+ "grad_norm": 0.1337890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0521,
+ "step": 3536
+ },
+ {
+ "epoch": 11.300319488817891,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0517,
+ "step": 3537
+ },
+ {
+ "epoch": 11.303514376996805,
+ "grad_norm": 0.111328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0558,
+ "step": 3538
+ },
+ {
+ "epoch": 11.30670926517572,
+ "grad_norm": 0.06201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0539,
+ "step": 3539
+ },
+ {
+ "epoch": 11.309904153354633,
+ "grad_norm": 0.09765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0574,
+ "step": 3540
+ },
+ {
+ "epoch": 11.313099041533546,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0538,
+ "step": 3541
+ },
+ {
+ "epoch": 11.31629392971246,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0539,
+ "step": 3542
+ },
+ {
+ "epoch": 11.319488817891374,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0618,
+ "step": 3543
+ },
+ {
+ "epoch": 11.322683706070288,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0611,
+ "step": 3544
+ },
+ {
+ "epoch": 11.3258785942492,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0661,
+ "step": 3545
+ },
+ {
+ "epoch": 11.329073482428115,
+ "grad_norm": 0.08447265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0467,
+ "step": 3546
+ },
+ {
+ "epoch": 11.332268370607029,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.071,
+ "step": 3547
+ },
+ {
+ "epoch": 11.335463258785943,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0631,
+ "step": 3548
+ },
+ {
+ "epoch": 11.338658146964857,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0709,
+ "step": 3549
+ },
+ {
+ "epoch": 11.34185303514377,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0622,
+ "step": 3550
+ },
+ {
+ "epoch": 11.345047923322683,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0434,
+ "step": 3551
+ },
+ {
+ "epoch": 11.348242811501597,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0577,
+ "step": 3552
+ },
+ {
+ "epoch": 11.351437699680512,
+ "grad_norm": 0.0830078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0677,
+ "step": 3553
+ },
+ {
+ "epoch": 11.354632587859426,
+ "grad_norm": 0.09130859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0583,
+ "step": 3554
+ },
+ {
+ "epoch": 11.357827476038338,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0572,
+ "step": 3555
+ },
+ {
+ "epoch": 11.361022364217252,
+ "grad_norm": 0.08251953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0599,
+ "step": 3556
+ },
+ {
+ "epoch": 11.364217252396166,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0606,
+ "step": 3557
+ },
+ {
+ "epoch": 11.36741214057508,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0644,
+ "step": 3558
+ },
+ {
+ "epoch": 11.370607028753994,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0624,
+ "step": 3559
+ },
+ {
+ "epoch": 11.373801916932907,
+ "grad_norm": 0.080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0686,
+ "step": 3560
+ },
+ {
+ "epoch": 11.37699680511182,
+ "grad_norm": 0.10205078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0726,
+ "step": 3561
+ },
+ {
+ "epoch": 11.380191693290735,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0583,
+ "step": 3562
+ },
+ {
+ "epoch": 11.383386581469649,
+ "grad_norm": 0.10302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0503,
+ "step": 3563
+ },
+ {
+ "epoch": 11.386581469648561,
+ "grad_norm": 0.08154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0644,
+ "step": 3564
+ },
+ {
+ "epoch": 11.389776357827476,
+ "grad_norm": 0.08251953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0585,
+ "step": 3565
+ },
+ {
+ "epoch": 11.39297124600639,
+ "grad_norm": 0.103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0582,
+ "step": 3566
+ },
+ {
+ "epoch": 11.396166134185304,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0607,
+ "step": 3567
+ },
+ {
+ "epoch": 11.399361022364218,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0631,
+ "step": 3568
+ },
+ {
+ "epoch": 11.40255591054313,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0568,
+ "step": 3569
+ },
+ {
+ "epoch": 11.405750798722044,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.051,
+ "step": 3570
+ },
+ {
+ "epoch": 11.408945686900958,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0539,
+ "step": 3571
+ },
+ {
+ "epoch": 11.412140575079873,
+ "grad_norm": 0.060791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0639,
+ "step": 3572
+ },
+ {
+ "epoch": 11.415335463258787,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0597,
+ "step": 3573
+ },
+ {
+ "epoch": 11.418530351437699,
+ "grad_norm": 0.1123046875,
+ "learning_rate": 0.0005,
+ "loss": 1.056,
+ "step": 3574
+ },
+ {
+ "epoch": 11.421725239616613,
+ "grad_norm": 0.055908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0556,
+ "step": 3575
+ },
+ {
+ "epoch": 11.424920127795527,
+ "grad_norm": 0.10205078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0599,
+ "step": 3576
+ },
+ {
+ "epoch": 11.428115015974441,
+ "grad_norm": 0.05712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0509,
+ "step": 3577
+ },
+ {
+ "epoch": 11.431309904153355,
+ "grad_norm": 0.10595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0591,
+ "step": 3578
+ },
+ {
+ "epoch": 11.434504792332268,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0637,
+ "step": 3579
+ },
+ {
+ "epoch": 11.437699680511182,
+ "grad_norm": 0.054931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0584,
+ "step": 3580
+ },
+ {
+ "epoch": 11.440894568690096,
+ "grad_norm": 0.0576171875,
+ "learning_rate": 0.0005,
+ "loss": 1.065,
+ "step": 3581
+ },
+ {
+ "epoch": 11.44408945686901,
+ "grad_norm": 0.052734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0656,
+ "step": 3582
+ },
+ {
+ "epoch": 11.447284345047922,
+ "grad_norm": 0.06103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.039,
+ "step": 3583
+ },
+ {
+ "epoch": 11.450479233226837,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.053,
+ "step": 3584
+ },
+ {
+ "epoch": 11.45367412140575,
+ "grad_norm": 0.05517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0551,
+ "step": 3585
+ },
+ {
+ "epoch": 11.456869009584665,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.073,
+ "step": 3586
+ },
+ {
+ "epoch": 11.460063897763579,
+ "grad_norm": 0.058349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0614,
+ "step": 3587
+ },
+ {
+ "epoch": 11.463258785942491,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0588,
+ "step": 3588
+ },
+ {
+ "epoch": 11.466453674121405,
+ "grad_norm": 0.057373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0593,
+ "step": 3589
+ },
+ {
+ "epoch": 11.46964856230032,
+ "grad_norm": 0.08837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.052,
+ "step": 3590
+ },
+ {
+ "epoch": 11.472843450479234,
+ "grad_norm": 0.09423828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0565,
+ "step": 3591
+ },
+ {
+ "epoch": 11.476038338658148,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0446,
+ "step": 3592
+ },
+ {
+ "epoch": 11.47923322683706,
+ "grad_norm": 0.11083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0576,
+ "step": 3593
+ },
+ {
+ "epoch": 11.482428115015974,
+ "grad_norm": 0.05517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0553,
+ "step": 3594
+ },
+ {
+ "epoch": 11.485623003194888,
+ "grad_norm": 0.09130859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0641,
+ "step": 3595
+ },
+ {
+ "epoch": 11.488817891373802,
+ "grad_norm": 0.09326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0665,
+ "step": 3596
+ },
+ {
+ "epoch": 11.492012779552716,
+ "grad_norm": 0.0859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0629,
+ "step": 3597
+ },
+ {
+ "epoch": 11.495207667731629,
+ "grad_norm": 0.12255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0539,
+ "step": 3598
+ },
+ {
+ "epoch": 11.498402555910543,
+ "grad_norm": 0.0556640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0703,
+ "step": 3599
+ },
+ {
+ "epoch": 11.501597444089457,
+ "grad_norm": 0.10009765625,
+ "learning_rate": 0.0005,
+ "loss": 1.066,
+ "step": 3600
+ },
+ {
+ "epoch": 11.504792332268371,
+ "grad_norm": 0.09228515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0604,
+ "step": 3601
+ },
+ {
+ "epoch": 11.507987220447284,
+ "grad_norm": 0.060302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0658,
+ "step": 3602
+ },
+ {
+ "epoch": 11.511182108626198,
+ "grad_norm": 0.08837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0578,
+ "step": 3603
+ },
+ {
+ "epoch": 11.514376996805112,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0685,
+ "step": 3604
+ },
+ {
+ "epoch": 11.517571884984026,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0648,
+ "step": 3605
+ },
+ {
+ "epoch": 11.52076677316294,
+ "grad_norm": 0.07861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.055,
+ "step": 3606
+ },
+ {
+ "epoch": 11.523961661341852,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0559,
+ "step": 3607
+ },
+ {
+ "epoch": 11.527156549520766,
+ "grad_norm": 0.055419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0646,
+ "step": 3608
+ },
+ {
+ "epoch": 11.53035143769968,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0551,
+ "step": 3609
+ },
+ {
+ "epoch": 11.533546325878595,
+ "grad_norm": 0.051025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0639,
+ "step": 3610
+ },
+ {
+ "epoch": 11.536741214057509,
+ "grad_norm": 0.062255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0547,
+ "step": 3611
+ },
+ {
+ "epoch": 11.539936102236421,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0739,
+ "step": 3612
+ },
+ {
+ "epoch": 11.543130990415335,
+ "grad_norm": 0.051513671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0681,
+ "step": 3613
+ },
+ {
+ "epoch": 11.54632587859425,
+ "grad_norm": 0.0576171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0521,
+ "step": 3614
+ },
+ {
+ "epoch": 11.549520766773163,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0585,
+ "step": 3615
+ },
+ {
+ "epoch": 11.552715654952078,
+ "grad_norm": 0.12158203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0676,
+ "step": 3616
+ },
+ {
+ "epoch": 11.55591054313099,
+ "grad_norm": 0.19140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0683,
+ "step": 3617
+ },
+ {
+ "epoch": 11.559105431309904,
+ "grad_norm": 0.10986328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0544,
+ "step": 3618
+ },
+ {
+ "epoch": 11.562300319488818,
+ "grad_norm": 0.2177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0655,
+ "step": 3619
+ },
+ {
+ "epoch": 11.565495207667732,
+ "grad_norm": 0.10693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0602,
+ "step": 3620
+ },
+ {
+ "epoch": 11.568690095846645,
+ "grad_norm": 0.134765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0785,
+ "step": 3621
+ },
+ {
+ "epoch": 11.571884984025559,
+ "grad_norm": 0.0908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0465,
+ "step": 3622
+ },
+ {
+ "epoch": 11.575079872204473,
+ "grad_norm": 0.09228515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0613,
+ "step": 3623
+ },
+ {
+ "epoch": 11.578274760383387,
+ "grad_norm": 0.11328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0626,
+ "step": 3624
+ },
+ {
+ "epoch": 11.581469648562301,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0626,
+ "step": 3625
+ },
+ {
+ "epoch": 11.584664536741213,
+ "grad_norm": 0.0869140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0612,
+ "step": 3626
+ },
+ {
+ "epoch": 11.587859424920127,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0348,
+ "step": 3627
+ },
+ {
+ "epoch": 11.591054313099042,
+ "grad_norm": 0.0615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0638,
+ "step": 3628
+ },
+ {
+ "epoch": 11.594249201277956,
+ "grad_norm": 0.0908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0646,
+ "step": 3629
+ },
+ {
+ "epoch": 11.59744408945687,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0477,
+ "step": 3630
+ },
+ {
+ "epoch": 11.600638977635782,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0484,
+ "step": 3631
+ },
+ {
+ "epoch": 11.603833865814696,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0536,
+ "step": 3632
+ },
+ {
+ "epoch": 11.60702875399361,
+ "grad_norm": 0.0908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0535,
+ "step": 3633
+ },
+ {
+ "epoch": 11.610223642172524,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0497,
+ "step": 3634
+ },
+ {
+ "epoch": 11.613418530351439,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0563,
+ "step": 3635
+ },
+ {
+ "epoch": 11.616613418530351,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0553,
+ "step": 3636
+ },
+ {
+ "epoch": 11.619808306709265,
+ "grad_norm": 0.09130859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0645,
+ "step": 3637
+ },
+ {
+ "epoch": 11.62300319488818,
+ "grad_norm": 0.1103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0596,
+ "step": 3638
+ },
+ {
+ "epoch": 11.626198083067093,
+ "grad_norm": 0.140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0672,
+ "step": 3639
+ },
+ {
+ "epoch": 11.629392971246006,
+ "grad_norm": 0.2236328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0502,
+ "step": 3640
+ },
+ {
+ "epoch": 11.63258785942492,
+ "grad_norm": 0.09375,
+ "learning_rate": 0.0005,
+ "loss": 1.061,
+ "step": 3641
+ },
+ {
+ "epoch": 11.635782747603834,
+ "grad_norm": 0.1298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0527,
+ "step": 3642
+ },
+ {
+ "epoch": 11.638977635782748,
+ "grad_norm": 0.12353515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0575,
+ "step": 3643
+ },
+ {
+ "epoch": 11.642172523961662,
+ "grad_norm": 0.1064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0524,
+ "step": 3644
+ },
+ {
+ "epoch": 11.645367412140574,
+ "grad_norm": 0.1796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0643,
+ "step": 3645
+ },
+ {
+ "epoch": 11.648562300319488,
+ "grad_norm": 0.1376953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0603,
+ "step": 3646
+ },
+ {
+ "epoch": 11.651757188498403,
+ "grad_norm": 0.1455078125,
+ "learning_rate": 0.0005,
+ "loss": 1.058,
+ "step": 3647
+ },
+ {
+ "epoch": 11.654952076677317,
+ "grad_norm": 0.23046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0558,
+ "step": 3648
+ },
+ {
+ "epoch": 11.65814696485623,
+ "grad_norm": 0.1298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0602,
+ "step": 3649
+ },
+ {
+ "epoch": 11.661341853035143,
+ "grad_norm": 0.057373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0647,
+ "step": 3650
+ },
+ {
+ "epoch": 11.664536741214057,
+ "grad_norm": 0.1279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0602,
+ "step": 3651
+ },
+ {
+ "epoch": 11.667731629392971,
+ "grad_norm": 0.138671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0712,
+ "step": 3652
+ },
+ {
+ "epoch": 11.670926517571885,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0515,
+ "step": 3653
+ },
+ {
+ "epoch": 11.6741214057508,
+ "grad_norm": 0.1669921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0464,
+ "step": 3654
+ },
+ {
+ "epoch": 11.677316293929712,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0552,
+ "step": 3655
+ },
+ {
+ "epoch": 11.680511182108626,
+ "grad_norm": 0.1044921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0682,
+ "step": 3656
+ },
+ {
+ "epoch": 11.68370607028754,
+ "grad_norm": 0.09765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0539,
+ "step": 3657
+ },
+ {
+ "epoch": 11.686900958466454,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0579,
+ "step": 3658
+ },
+ {
+ "epoch": 11.690095846645367,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0583,
+ "step": 3659
+ },
+ {
+ "epoch": 11.69329073482428,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0574,
+ "step": 3660
+ },
+ {
+ "epoch": 11.696485623003195,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0632,
+ "step": 3661
+ },
+ {
+ "epoch": 11.699680511182109,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0638,
+ "step": 3662
+ },
+ {
+ "epoch": 11.702875399361023,
+ "grad_norm": 0.058837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0557,
+ "step": 3663
+ },
+ {
+ "epoch": 11.706070287539935,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0539,
+ "step": 3664
+ },
+ {
+ "epoch": 11.70926517571885,
+ "grad_norm": 0.061767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0656,
+ "step": 3665
+ },
+ {
+ "epoch": 11.712460063897764,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0568,
+ "step": 3666
+ },
+ {
+ "epoch": 11.715654952076678,
+ "grad_norm": 0.0546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0556,
+ "step": 3667
+ },
+ {
+ "epoch": 11.718849840255592,
+ "grad_norm": 0.057373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.044,
+ "step": 3668
+ },
+ {
+ "epoch": 11.722044728434504,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0686,
+ "step": 3669
+ },
+ {
+ "epoch": 11.725239616613418,
+ "grad_norm": 0.0595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0804,
+ "step": 3670
+ },
+ {
+ "epoch": 11.728434504792332,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.051,
+ "step": 3671
+ },
+ {
+ "epoch": 11.731629392971247,
+ "grad_norm": 0.0869140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0568,
+ "step": 3672
+ },
+ {
+ "epoch": 11.73482428115016,
+ "grad_norm": 0.12060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0533,
+ "step": 3673
+ },
+ {
+ "epoch": 11.738019169329073,
+ "grad_norm": 0.083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0587,
+ "step": 3674
+ },
+ {
+ "epoch": 11.741214057507987,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0598,
+ "step": 3675
+ },
+ {
+ "epoch": 11.744408945686901,
+ "grad_norm": 0.0595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0586,
+ "step": 3676
+ },
+ {
+ "epoch": 11.747603833865815,
+ "grad_norm": 0.061767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0378,
+ "step": 3677
+ },
+ {
+ "epoch": 11.750798722044728,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0577,
+ "step": 3678
+ },
+ {
+ "epoch": 11.753993610223642,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0593,
+ "step": 3679
+ },
+ {
+ "epoch": 11.757188498402556,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0582,
+ "step": 3680
+ },
+ {
+ "epoch": 11.76038338658147,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0581,
+ "step": 3681
+ },
+ {
+ "epoch": 11.763578274760384,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.059,
+ "step": 3682
+ },
+ {
+ "epoch": 11.766773162939296,
+ "grad_norm": 0.058837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0685,
+ "step": 3683
+ },
+ {
+ "epoch": 11.76996805111821,
+ "grad_norm": 0.05810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0664,
+ "step": 3684
+ },
+ {
+ "epoch": 11.773162939297125,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0571,
+ "step": 3685
+ },
+ {
+ "epoch": 11.776357827476039,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0478,
+ "step": 3686
+ },
+ {
+ "epoch": 11.779552715654953,
+ "grad_norm": 0.083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0509,
+ "step": 3687
+ },
+ {
+ "epoch": 11.782747603833865,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0564,
+ "step": 3688
+ },
+ {
+ "epoch": 11.78594249201278,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0686,
+ "step": 3689
+ },
+ {
+ "epoch": 11.789137380191693,
+ "grad_norm": 0.10107421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0549,
+ "step": 3690
+ },
+ {
+ "epoch": 11.792332268370608,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0503,
+ "step": 3691
+ },
+ {
+ "epoch": 11.795527156549522,
+ "grad_norm": 0.1318359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0611,
+ "step": 3692
+ },
+ {
+ "epoch": 11.798722044728434,
+ "grad_norm": 0.09033203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0597,
+ "step": 3693
+ },
+ {
+ "epoch": 11.801916932907348,
+ "grad_norm": 0.10400390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0628,
+ "step": 3694
+ },
+ {
+ "epoch": 11.805111821086262,
+ "grad_norm": 0.1689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0502,
+ "step": 3695
+ },
+ {
+ "epoch": 11.808306709265176,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0609,
+ "step": 3696
+ },
+ {
+ "epoch": 11.811501597444089,
+ "grad_norm": 0.1708984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0621,
+ "step": 3697
+ },
+ {
+ "epoch": 11.814696485623003,
+ "grad_norm": 0.1376953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0629,
+ "step": 3698
+ },
+ {
+ "epoch": 11.817891373801917,
+ "grad_norm": 0.061279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0638,
+ "step": 3699
+ },
+ {
+ "epoch": 11.821086261980831,
+ "grad_norm": 0.1337890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0529,
+ "step": 3700
+ },
+ {
+ "epoch": 11.824281150159745,
+ "grad_norm": 0.10009765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0591,
+ "step": 3701
+ },
+ {
+ "epoch": 11.827476038338657,
+ "grad_norm": 0.11962890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0477,
+ "step": 3702
+ },
+ {
+ "epoch": 11.830670926517572,
+ "grad_norm": 0.16015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0567,
+ "step": 3703
+ },
+ {
+ "epoch": 11.833865814696486,
+ "grad_norm": 0.058349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0742,
+ "step": 3704
+ },
+ {
+ "epoch": 11.8370607028754,
+ "grad_norm": 0.10888671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0458,
+ "step": 3705
+ },
+ {
+ "epoch": 11.840255591054314,
+ "grad_norm": 0.09033203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0601,
+ "step": 3706
+ },
+ {
+ "epoch": 11.843450479233226,
+ "grad_norm": 0.1298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0566,
+ "step": 3707
+ },
+ {
+ "epoch": 11.84664536741214,
+ "grad_norm": 0.09228515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0537,
+ "step": 3708
+ },
+ {
+ "epoch": 11.849840255591054,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0646,
+ "step": 3709
+ },
+ {
+ "epoch": 11.853035143769969,
+ "grad_norm": 0.099609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0501,
+ "step": 3710
+ },
+ {
+ "epoch": 11.856230031948883,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0655,
+ "step": 3711
+ },
+ {
+ "epoch": 11.859424920127795,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.063,
+ "step": 3712
+ },
+ {
+ "epoch": 11.86261980830671,
+ "grad_norm": 0.10107421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0637,
+ "step": 3713
+ },
+ {
+ "epoch": 11.865814696485623,
+ "grad_norm": 0.06201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0653,
+ "step": 3714
+ },
+ {
+ "epoch": 11.869009584664537,
+ "grad_norm": 0.083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0596,
+ "step": 3715
+ },
+ {
+ "epoch": 11.87220447284345,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0583,
+ "step": 3716
+ },
+ {
+ "epoch": 11.875399361022364,
+ "grad_norm": 0.080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0532,
+ "step": 3717
+ },
+ {
+ "epoch": 11.878594249201278,
+ "grad_norm": 0.08544921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0657,
+ "step": 3718
+ },
+ {
+ "epoch": 11.881789137380192,
+ "grad_norm": 0.08447265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0603,
+ "step": 3719
+ },
+ {
+ "epoch": 11.884984025559106,
+ "grad_norm": 0.12060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0636,
+ "step": 3720
+ },
+ {
+ "epoch": 11.888178913738018,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0572,
+ "step": 3721
+ },
+ {
+ "epoch": 11.891373801916933,
+ "grad_norm": 0.169921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0615,
+ "step": 3722
+ },
+ {
+ "epoch": 11.894568690095847,
+ "grad_norm": 0.1318359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0571,
+ "step": 3723
+ },
+ {
+ "epoch": 11.89776357827476,
+ "grad_norm": 0.06201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0523,
+ "step": 3724
+ },
+ {
+ "epoch": 11.900958466453675,
+ "grad_norm": 0.0859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0514,
+ "step": 3725
+ },
+ {
+ "epoch": 11.904153354632587,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0575,
+ "step": 3726
+ },
+ {
+ "epoch": 11.907348242811501,
+ "grad_norm": 0.0859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0435,
+ "step": 3727
+ },
+ {
+ "epoch": 11.910543130990416,
+ "grad_norm": 0.1123046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0488,
+ "step": 3728
+ },
+ {
+ "epoch": 11.91373801916933,
+ "grad_norm": 0.0859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0609,
+ "step": 3729
+ },
+ {
+ "epoch": 11.916932907348244,
+ "grad_norm": 0.177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.05,
+ "step": 3730
+ },
+ {
+ "epoch": 11.920127795527156,
+ "grad_norm": 0.10400390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0437,
+ "step": 3731
+ },
+ {
+ "epoch": 11.92332268370607,
+ "grad_norm": 0.11474609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0574,
+ "step": 3732
+ },
+ {
+ "epoch": 11.926517571884984,
+ "grad_norm": 0.16015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0577,
+ "step": 3733
+ },
+ {
+ "epoch": 11.929712460063898,
+ "grad_norm": 0.06005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.052,
+ "step": 3734
+ },
+ {
+ "epoch": 11.93290734824281,
+ "grad_norm": 0.11669921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0628,
+ "step": 3735
+ },
+ {
+ "epoch": 11.936102236421725,
+ "grad_norm": 0.11181640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0537,
+ "step": 3736
+ },
+ {
+ "epoch": 11.939297124600639,
+ "grad_norm": 0.08837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0606,
+ "step": 3737
+ },
+ {
+ "epoch": 11.942492012779553,
+ "grad_norm": 0.126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0609,
+ "step": 3738
+ },
+ {
+ "epoch": 11.945686900958467,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0583,
+ "step": 3739
+ },
+ {
+ "epoch": 11.94888178913738,
+ "grad_norm": 0.1396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0655,
+ "step": 3740
+ },
+ {
+ "epoch": 11.952076677316294,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0446,
+ "step": 3741
+ },
+ {
+ "epoch": 11.955271565495208,
+ "grad_norm": 0.1455078125,
+ "learning_rate": 0.0005,
+ "loss": 1.066,
+ "step": 3742
+ },
+ {
+ "epoch": 11.958466453674122,
+ "grad_norm": 0.28515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0613,
+ "step": 3743
+ },
+ {
+ "epoch": 11.961661341853034,
+ "grad_norm": 0.2099609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0644,
+ "step": 3744
+ },
+ {
+ "epoch": 11.964856230031948,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0627,
+ "step": 3745
+ },
+ {
+ "epoch": 11.968051118210862,
+ "grad_norm": 0.2265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0662,
+ "step": 3746
+ },
+ {
+ "epoch": 11.971246006389777,
+ "grad_norm": 0.201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0639,
+ "step": 3747
+ },
+ {
+ "epoch": 11.97444089456869,
+ "grad_norm": 0.06103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0676,
+ "step": 3748
+ },
+ {
+ "epoch": 11.977635782747605,
+ "grad_norm": 0.1064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0569,
+ "step": 3749
+ },
+ {
+ "epoch": 11.980830670926517,
+ "grad_norm": 0.087890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0626,
+ "step": 3750
+ },
+ {
+ "epoch": 11.984025559105431,
+ "grad_norm": 0.1025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0447,
+ "step": 3751
+ },
+ {
+ "epoch": 11.987220447284345,
+ "grad_norm": 0.251953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0505,
+ "step": 3752
+ },
+ {
+ "epoch": 11.99041533546326,
+ "grad_norm": 0.279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0437,
+ "step": 3753
+ },
+ {
+ "epoch": 11.993610223642172,
+ "grad_norm": 0.08154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0652,
+ "step": 3754
+ },
+ {
+ "epoch": 11.996805111821086,
+ "grad_norm": 0.1513671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0483,
+ "step": 3755
+ },
+ {
+ "epoch": 12.0,
+ "grad_norm": 0.1279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0509,
+ "step": 3756
+ },
+ {
+ "epoch": 12.003194888178914,
+ "grad_norm": 0.0859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0588,
+ "step": 3757
+ },
+ {
+ "epoch": 12.006389776357828,
+ "grad_norm": 0.0908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0483,
+ "step": 3758
+ },
+ {
+ "epoch": 12.00958466453674,
+ "grad_norm": 0.05908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0637,
+ "step": 3759
+ },
+ {
+ "epoch": 12.012779552715655,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0678,
+ "step": 3760
+ },
+ {
+ "epoch": 12.015974440894569,
+ "grad_norm": 0.0556640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0425,
+ "step": 3761
+ },
+ {
+ "epoch": 12.019169329073483,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0681,
+ "step": 3762
+ },
+ {
+ "epoch": 12.022364217252397,
+ "grad_norm": 0.059326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0578,
+ "step": 3763
+ },
+ {
+ "epoch": 12.02555910543131,
+ "grad_norm": 0.057861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0677,
+ "step": 3764
+ },
+ {
+ "epoch": 12.028753993610223,
+ "grad_norm": 0.091796875,
+ "learning_rate": 0.0005,
+ "loss": 1.043,
+ "step": 3765
+ },
+ {
+ "epoch": 12.031948881789138,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0482,
+ "step": 3766
+ },
+ {
+ "epoch": 12.035143769968052,
+ "grad_norm": 0.10986328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0557,
+ "step": 3767
+ },
+ {
+ "epoch": 12.038338658146964,
+ "grad_norm": 0.0888671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0608,
+ "step": 3768
+ },
+ {
+ "epoch": 12.041533546325878,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0656,
+ "step": 3769
+ },
+ {
+ "epoch": 12.044728434504792,
+ "grad_norm": 0.109375,
+ "learning_rate": 0.0005,
+ "loss": 1.057,
+ "step": 3770
+ },
+ {
+ "epoch": 12.047923322683706,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0651,
+ "step": 3771
+ },
+ {
+ "epoch": 12.05111821086262,
+ "grad_norm": 0.0986328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0632,
+ "step": 3772
+ },
+ {
+ "epoch": 12.054313099041533,
+ "grad_norm": 0.1103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0603,
+ "step": 3773
+ },
+ {
+ "epoch": 12.057507987220447,
+ "grad_norm": 0.057373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0502,
+ "step": 3774
+ },
+ {
+ "epoch": 12.060702875399361,
+ "grad_norm": 0.087890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0599,
+ "step": 3775
+ },
+ {
+ "epoch": 12.063897763578275,
+ "grad_norm": 0.10888671875,
+ "learning_rate": 0.0005,
+ "loss": 1.054,
+ "step": 3776
+ },
+ {
+ "epoch": 12.06709265175719,
+ "grad_norm": 0.09765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0673,
+ "step": 3777
+ },
+ {
+ "epoch": 12.070287539936102,
+ "grad_norm": 0.08837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0611,
+ "step": 3778
+ },
+ {
+ "epoch": 12.073482428115016,
+ "grad_norm": 0.08837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0583,
+ "step": 3779
+ },
+ {
+ "epoch": 12.07667731629393,
+ "grad_norm": 0.12353515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0514,
+ "step": 3780
+ },
+ {
+ "epoch": 12.079872204472844,
+ "grad_norm": 0.240234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0698,
+ "step": 3781
+ },
+ {
+ "epoch": 12.083067092651758,
+ "grad_norm": 0.1865234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0612,
+ "step": 3782
+ },
+ {
+ "epoch": 12.08626198083067,
+ "grad_norm": 0.05615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0543,
+ "step": 3783
+ },
+ {
+ "epoch": 12.089456869009584,
+ "grad_norm": 0.140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0634,
+ "step": 3784
+ },
+ {
+ "epoch": 12.092651757188499,
+ "grad_norm": 0.08642578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0435,
+ "step": 3785
+ },
+ {
+ "epoch": 12.095846645367413,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0545,
+ "step": 3786
+ },
+ {
+ "epoch": 12.099041533546325,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0598,
+ "step": 3787
+ },
+ {
+ "epoch": 12.10223642172524,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0644,
+ "step": 3788
+ },
+ {
+ "epoch": 12.105431309904153,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0598,
+ "step": 3789
+ },
+ {
+ "epoch": 12.108626198083067,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0549,
+ "step": 3790
+ },
+ {
+ "epoch": 12.111821086261982,
+ "grad_norm": 0.0927734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0615,
+ "step": 3791
+ },
+ {
+ "epoch": 12.115015974440894,
+ "grad_norm": 0.0869140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0542,
+ "step": 3792
+ },
+ {
+ "epoch": 12.118210862619808,
+ "grad_norm": 0.107421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0391,
+ "step": 3793
+ },
+ {
+ "epoch": 12.121405750798722,
+ "grad_norm": 0.10400390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0569,
+ "step": 3794
+ },
+ {
+ "epoch": 12.124600638977636,
+ "grad_norm": 0.1123046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0516,
+ "step": 3795
+ },
+ {
+ "epoch": 12.12779552715655,
+ "grad_norm": 0.1806640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0592,
+ "step": 3796
+ },
+ {
+ "epoch": 12.130990415335463,
+ "grad_norm": 0.11328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0613,
+ "step": 3797
+ },
+ {
+ "epoch": 12.134185303514377,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.057,
+ "step": 3798
+ },
+ {
+ "epoch": 12.13738019169329,
+ "grad_norm": 0.1376953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0612,
+ "step": 3799
+ },
+ {
+ "epoch": 12.140575079872205,
+ "grad_norm": 0.138671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0421,
+ "step": 3800
+ },
+ {
+ "epoch": 12.143769968051119,
+ "grad_norm": 0.0869140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0706,
+ "step": 3801
+ },
+ {
+ "epoch": 12.146964856230031,
+ "grad_norm": 0.1787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0653,
+ "step": 3802
+ },
+ {
+ "epoch": 12.150159744408946,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0655,
+ "step": 3803
+ },
+ {
+ "epoch": 12.15335463258786,
+ "grad_norm": 0.09521484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0498,
+ "step": 3804
+ },
+ {
+ "epoch": 12.156549520766774,
+ "grad_norm": 0.11767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0622,
+ "step": 3805
+ },
+ {
+ "epoch": 12.159744408945686,
+ "grad_norm": 0.08544921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0644,
+ "step": 3806
+ },
+ {
+ "epoch": 12.1629392971246,
+ "grad_norm": 0.0859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0584,
+ "step": 3807
+ },
+ {
+ "epoch": 12.166134185303514,
+ "grad_norm": 0.08251953125,
+ "learning_rate": 0.0005,
+ "loss": 1.046,
+ "step": 3808
+ },
+ {
+ "epoch": 12.169329073482428,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0529,
+ "step": 3809
+ },
+ {
+ "epoch": 12.172523961661343,
+ "grad_norm": 0.087890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0504,
+ "step": 3810
+ },
+ {
+ "epoch": 12.175718849840255,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0416,
+ "step": 3811
+ },
+ {
+ "epoch": 12.178913738019169,
+ "grad_norm": 0.1005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0548,
+ "step": 3812
+ },
+ {
+ "epoch": 12.182108626198083,
+ "grad_norm": 0.05810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0652,
+ "step": 3813
+ },
+ {
+ "epoch": 12.185303514376997,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0558,
+ "step": 3814
+ },
+ {
+ "epoch": 12.188498402555911,
+ "grad_norm": 0.061767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0587,
+ "step": 3815
+ },
+ {
+ "epoch": 12.191693290734824,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0492,
+ "step": 3816
+ },
+ {
+ "epoch": 12.194888178913738,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0573,
+ "step": 3817
+ },
+ {
+ "epoch": 12.198083067092652,
+ "grad_norm": 0.08056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0654,
+ "step": 3818
+ },
+ {
+ "epoch": 12.201277955271566,
+ "grad_norm": 0.08203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0594,
+ "step": 3819
+ },
+ {
+ "epoch": 12.204472843450478,
+ "grad_norm": 0.09375,
+ "learning_rate": 0.0005,
+ "loss": 1.0636,
+ "step": 3820
+ },
+ {
+ "epoch": 12.207667731629392,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0597,
+ "step": 3821
+ },
+ {
+ "epoch": 12.210862619808307,
+ "grad_norm": 0.09326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0677,
+ "step": 3822
+ },
+ {
+ "epoch": 12.21405750798722,
+ "grad_norm": 0.08056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0604,
+ "step": 3823
+ },
+ {
+ "epoch": 12.217252396166135,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0572,
+ "step": 3824
+ },
+ {
+ "epoch": 12.220447284345047,
+ "grad_norm": 0.107421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0494,
+ "step": 3825
+ },
+ {
+ "epoch": 12.223642172523961,
+ "grad_norm": 0.10986328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0628,
+ "step": 3826
+ },
+ {
+ "epoch": 12.226837060702875,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0549,
+ "step": 3827
+ },
+ {
+ "epoch": 12.23003194888179,
+ "grad_norm": 0.0830078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0712,
+ "step": 3828
+ },
+ {
+ "epoch": 12.233226837060704,
+ "grad_norm": 0.08447265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0542,
+ "step": 3829
+ },
+ {
+ "epoch": 12.236421725239616,
+ "grad_norm": 0.1318359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0727,
+ "step": 3830
+ },
+ {
+ "epoch": 12.23961661341853,
+ "grad_norm": 0.09521484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0447,
+ "step": 3831
+ },
+ {
+ "epoch": 12.242811501597444,
+ "grad_norm": 0.0888671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0559,
+ "step": 3832
+ },
+ {
+ "epoch": 12.246006389776358,
+ "grad_norm": 0.1298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0535,
+ "step": 3833
+ },
+ {
+ "epoch": 12.249201277955272,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0466,
+ "step": 3834
+ },
+ {
+ "epoch": 12.252396166134185,
+ "grad_norm": 0.08837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0586,
+ "step": 3835
+ },
+ {
+ "epoch": 12.255591054313099,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0542,
+ "step": 3836
+ },
+ {
+ "epoch": 12.258785942492013,
+ "grad_norm": 0.154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0562,
+ "step": 3837
+ },
+ {
+ "epoch": 12.261980830670927,
+ "grad_norm": 0.1416015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0556,
+ "step": 3838
+ },
+ {
+ "epoch": 12.26517571884984,
+ "grad_norm": 0.08740234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0534,
+ "step": 3839
+ },
+ {
+ "epoch": 12.268370607028753,
+ "grad_norm": 0.09912109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0472,
+ "step": 3840
+ },
+ {
+ "epoch": 12.271565495207668,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0621,
+ "step": 3841
+ },
+ {
+ "epoch": 12.274760383386582,
+ "grad_norm": 0.0859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0572,
+ "step": 3842
+ },
+ {
+ "epoch": 12.277955271565496,
+ "grad_norm": 0.08154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0576,
+ "step": 3843
+ },
+ {
+ "epoch": 12.281150159744408,
+ "grad_norm": 0.0810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0591,
+ "step": 3844
+ },
+ {
+ "epoch": 12.284345047923322,
+ "grad_norm": 0.10986328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0591,
+ "step": 3845
+ },
+ {
+ "epoch": 12.287539936102236,
+ "grad_norm": 0.05615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0525,
+ "step": 3846
+ },
+ {
+ "epoch": 12.29073482428115,
+ "grad_norm": 0.0947265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0552,
+ "step": 3847
+ },
+ {
+ "epoch": 12.293929712460065,
+ "grad_norm": 0.0595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0588,
+ "step": 3848
+ },
+ {
+ "epoch": 12.297124600638977,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0624,
+ "step": 3849
+ },
+ {
+ "epoch": 12.300319488817891,
+ "grad_norm": 0.0576171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0649,
+ "step": 3850
+ },
+ {
+ "epoch": 12.303514376996805,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0649,
+ "step": 3851
+ },
+ {
+ "epoch": 12.30670926517572,
+ "grad_norm": 0.062255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0473,
+ "step": 3852
+ },
+ {
+ "epoch": 12.309904153354633,
+ "grad_norm": 0.060302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0513,
+ "step": 3853
+ },
+ {
+ "epoch": 12.313099041533546,
+ "grad_norm": 0.05908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0551,
+ "step": 3854
+ },
+ {
+ "epoch": 12.31629392971246,
+ "grad_norm": 0.060791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0662,
+ "step": 3855
+ },
+ {
+ "epoch": 12.319488817891374,
+ "grad_norm": 0.05712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.061,
+ "step": 3856
+ },
+ {
+ "epoch": 12.322683706070288,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0556,
+ "step": 3857
+ },
+ {
+ "epoch": 12.3258785942492,
+ "grad_norm": 0.055419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0712,
+ "step": 3858
+ },
+ {
+ "epoch": 12.329073482428115,
+ "grad_norm": 0.05859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0562,
+ "step": 3859
+ },
+ {
+ "epoch": 12.332268370607029,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0645,
+ "step": 3860
+ },
+ {
+ "epoch": 12.335463258785943,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0478,
+ "step": 3861
+ },
+ {
+ "epoch": 12.338658146964857,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0524,
+ "step": 3862
+ },
+ {
+ "epoch": 12.34185303514377,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.054,
+ "step": 3863
+ },
+ {
+ "epoch": 12.345047923322683,
+ "grad_norm": 0.052734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0617,
+ "step": 3864
+ },
+ {
+ "epoch": 12.348242811501597,
+ "grad_norm": 0.1083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0568,
+ "step": 3865
+ },
+ {
+ "epoch": 12.351437699680512,
+ "grad_norm": 0.057861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0601,
+ "step": 3866
+ },
+ {
+ "epoch": 12.354632587859426,
+ "grad_norm": 0.083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0731,
+ "step": 3867
+ },
+ {
+ "epoch": 12.357827476038338,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0593,
+ "step": 3868
+ },
+ {
+ "epoch": 12.361022364217252,
+ "grad_norm": 0.05419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0693,
+ "step": 3869
+ },
+ {
+ "epoch": 12.364217252396166,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0475,
+ "step": 3870
+ },
+ {
+ "epoch": 12.36741214057508,
+ "grad_norm": 0.06103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.067,
+ "step": 3871
+ },
+ {
+ "epoch": 12.370607028753994,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0577,
+ "step": 3872
+ },
+ {
+ "epoch": 12.373801916932907,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0628,
+ "step": 3873
+ },
+ {
+ "epoch": 12.37699680511182,
+ "grad_norm": 0.057373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0542,
+ "step": 3874
+ },
+ {
+ "epoch": 12.380191693290735,
+ "grad_norm": 0.0615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0537,
+ "step": 3875
+ },
+ {
+ "epoch": 12.383386581469649,
+ "grad_norm": 0.0966796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0589,
+ "step": 3876
+ },
+ {
+ "epoch": 12.386581469648561,
+ "grad_norm": 0.055419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0606,
+ "step": 3877
+ },
+ {
+ "epoch": 12.389776357827476,
+ "grad_norm": 0.08740234375,
+ "learning_rate": 0.0005,
+ "loss": 1.054,
+ "step": 3878
+ },
+ {
+ "epoch": 12.39297124600639,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.047,
+ "step": 3879
+ },
+ {
+ "epoch": 12.396166134185304,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0634,
+ "step": 3880
+ },
+ {
+ "epoch": 12.399361022364218,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0676,
+ "step": 3881
+ },
+ {
+ "epoch": 12.40255591054313,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.05,
+ "step": 3882
+ },
+ {
+ "epoch": 12.405750798722044,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0523,
+ "step": 3883
+ },
+ {
+ "epoch": 12.408945686900958,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0591,
+ "step": 3884
+ },
+ {
+ "epoch": 12.412140575079873,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0519,
+ "step": 3885
+ },
+ {
+ "epoch": 12.415335463258787,
+ "grad_norm": 0.056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0475,
+ "step": 3886
+ },
+ {
+ "epoch": 12.418530351437699,
+ "grad_norm": 0.061279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0609,
+ "step": 3887
+ },
+ {
+ "epoch": 12.421725239616613,
+ "grad_norm": 0.04931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0652,
+ "step": 3888
+ },
+ {
+ "epoch": 12.424920127795527,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0627,
+ "step": 3889
+ },
+ {
+ "epoch": 12.428115015974441,
+ "grad_norm": 0.0556640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0528,
+ "step": 3890
+ },
+ {
+ "epoch": 12.431309904153355,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0553,
+ "step": 3891
+ },
+ {
+ "epoch": 12.434504792332268,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0596,
+ "step": 3892
+ },
+ {
+ "epoch": 12.437699680511182,
+ "grad_norm": 0.10205078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0564,
+ "step": 3893
+ },
+ {
+ "epoch": 12.440894568690096,
+ "grad_norm": 0.1904296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0566,
+ "step": 3894
+ },
+ {
+ "epoch": 12.44408945686901,
+ "grad_norm": 0.1337890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0583,
+ "step": 3895
+ },
+ {
+ "epoch": 12.447284345047922,
+ "grad_norm": 0.055419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0637,
+ "step": 3896
+ },
+ {
+ "epoch": 12.450479233226837,
+ "grad_norm": 0.130859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0483,
+ "step": 3897
+ },
+ {
+ "epoch": 12.45367412140575,
+ "grad_norm": 0.083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0475,
+ "step": 3898
+ },
+ {
+ "epoch": 12.456869009584665,
+ "grad_norm": 0.154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0738,
+ "step": 3899
+ },
+ {
+ "epoch": 12.460063897763579,
+ "grad_norm": 0.2373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0568,
+ "step": 3900
+ },
+ {
+ "epoch": 12.463258785942491,
+ "grad_norm": 0.134765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0753,
+ "step": 3901
+ },
+ {
+ "epoch": 12.466453674121405,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0549,
+ "step": 3902
+ },
+ {
+ "epoch": 12.46964856230032,
+ "grad_norm": 0.12158203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0592,
+ "step": 3903
+ },
+ {
+ "epoch": 12.472843450479234,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.06,
+ "step": 3904
+ },
+ {
+ "epoch": 12.476038338658148,
+ "grad_norm": 0.1396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0561,
+ "step": 3905
+ },
+ {
+ "epoch": 12.47923322683706,
+ "grad_norm": 0.12109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0524,
+ "step": 3906
+ },
+ {
+ "epoch": 12.482428115015974,
+ "grad_norm": 0.0859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0567,
+ "step": 3907
+ },
+ {
+ "epoch": 12.485623003194888,
+ "grad_norm": 0.16796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0702,
+ "step": 3908
+ },
+ {
+ "epoch": 12.488817891373802,
+ "grad_norm": 0.17578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0599,
+ "step": 3909
+ },
+ {
+ "epoch": 12.492012779552716,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0617,
+ "step": 3910
+ },
+ {
+ "epoch": 12.495207667731629,
+ "grad_norm": 0.2216796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0574,
+ "step": 3911
+ },
+ {
+ "epoch": 12.498402555910543,
+ "grad_norm": 0.181640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0492,
+ "step": 3912
+ },
+ {
+ "epoch": 12.501597444089457,
+ "grad_norm": 0.10205078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0576,
+ "step": 3913
+ },
+ {
+ "epoch": 12.504792332268371,
+ "grad_norm": 0.271484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0581,
+ "step": 3914
+ },
+ {
+ "epoch": 12.507987220447284,
+ "grad_norm": 0.171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0669,
+ "step": 3915
+ },
+ {
+ "epoch": 12.511182108626198,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0533,
+ "step": 3916
+ },
+ {
+ "epoch": 12.514376996805112,
+ "grad_norm": 0.1396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0641,
+ "step": 3917
+ },
+ {
+ "epoch": 12.517571884984026,
+ "grad_norm": 0.1337890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0552,
+ "step": 3918
+ },
+ {
+ "epoch": 12.52076677316294,
+ "grad_norm": 0.09765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0482,
+ "step": 3919
+ },
+ {
+ "epoch": 12.523961661341852,
+ "grad_norm": 0.1396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0542,
+ "step": 3920
+ },
+ {
+ "epoch": 12.527156549520766,
+ "grad_norm": 0.11376953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0533,
+ "step": 3921
+ },
+ {
+ "epoch": 12.53035143769968,
+ "grad_norm": 0.1240234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0595,
+ "step": 3922
+ },
+ {
+ "epoch": 12.533546325878595,
+ "grad_norm": 0.1806640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0625,
+ "step": 3923
+ },
+ {
+ "epoch": 12.536741214057509,
+ "grad_norm": 0.09814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0593,
+ "step": 3924
+ },
+ {
+ "epoch": 12.539936102236421,
+ "grad_norm": 0.259765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0689,
+ "step": 3925
+ },
+ {
+ "epoch": 12.543130990415335,
+ "grad_norm": 0.275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0616,
+ "step": 3926
+ },
+ {
+ "epoch": 12.54632587859425,
+ "grad_norm": 0.1650390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0593,
+ "step": 3927
+ },
+ {
+ "epoch": 12.549520766773163,
+ "grad_norm": 0.1484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0558,
+ "step": 3928
+ },
+ {
+ "epoch": 12.552715654952078,
+ "grad_norm": 0.33203125,
+ "learning_rate": 0.0005,
+ "loss": 1.033,
+ "step": 3929
+ },
+ {
+ "epoch": 12.55591054313099,
+ "grad_norm": 0.265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0622,
+ "step": 3930
+ },
+ {
+ "epoch": 12.559105431309904,
+ "grad_norm": 0.08251953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0645,
+ "step": 3931
+ },
+ {
+ "epoch": 12.562300319488818,
+ "grad_norm": 0.330078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0542,
+ "step": 3932
+ },
+ {
+ "epoch": 12.565495207667732,
+ "grad_norm": 0.404296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0698,
+ "step": 3933
+ },
+ {
+ "epoch": 12.568690095846645,
+ "grad_norm": 0.353515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0471,
+ "step": 3934
+ },
+ {
+ "epoch": 12.571884984025559,
+ "grad_norm": 0.09375,
+ "learning_rate": 0.0005,
+ "loss": 1.0588,
+ "step": 3935
+ },
+ {
+ "epoch": 12.575079872204473,
+ "grad_norm": 0.2138671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0541,
+ "step": 3936
+ },
+ {
+ "epoch": 12.578274760383387,
+ "grad_norm": 0.1904296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0629,
+ "step": 3937
+ },
+ {
+ "epoch": 12.581469648562301,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0628,
+ "step": 3938
+ },
+ {
+ "epoch": 12.584664536741213,
+ "grad_norm": 0.232421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0595,
+ "step": 3939
+ },
+ {
+ "epoch": 12.587859424920127,
+ "grad_norm": 0.1396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0625,
+ "step": 3940
+ },
+ {
+ "epoch": 12.591054313099042,
+ "grad_norm": 0.1220703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0836,
+ "step": 3941
+ },
+ {
+ "epoch": 12.594249201277956,
+ "grad_norm": 0.177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0651,
+ "step": 3942
+ },
+ {
+ "epoch": 12.59744408945687,
+ "grad_norm": 0.062255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.049,
+ "step": 3943
+ },
+ {
+ "epoch": 12.600638977635782,
+ "grad_norm": 0.16015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0583,
+ "step": 3944
+ },
+ {
+ "epoch": 12.603833865814696,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0641,
+ "step": 3945
+ },
+ {
+ "epoch": 12.60702875399361,
+ "grad_norm": 0.158203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0549,
+ "step": 3946
+ },
+ {
+ "epoch": 12.610223642172524,
+ "grad_norm": 0.146484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0594,
+ "step": 3947
+ },
+ {
+ "epoch": 12.613418530351439,
+ "grad_norm": 0.1015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0512,
+ "step": 3948
+ },
+ {
+ "epoch": 12.616613418530351,
+ "grad_norm": 0.220703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0564,
+ "step": 3949
+ },
+ {
+ "epoch": 12.619808306709265,
+ "grad_norm": 0.1044921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0527,
+ "step": 3950
+ },
+ {
+ "epoch": 12.62300319488818,
+ "grad_norm": 0.150390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0533,
+ "step": 3951
+ },
+ {
+ "epoch": 12.626198083067093,
+ "grad_norm": 0.2099609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0656,
+ "step": 3952
+ },
+ {
+ "epoch": 12.629392971246006,
+ "grad_norm": 0.059814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0517,
+ "step": 3953
+ },
+ {
+ "epoch": 12.63258785942492,
+ "grad_norm": 0.1630859375,
+ "learning_rate": 0.0005,
+ "loss": 1.062,
+ "step": 3954
+ },
+ {
+ "epoch": 12.635782747603834,
+ "grad_norm": 0.0908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0508,
+ "step": 3955
+ },
+ {
+ "epoch": 12.638977635782748,
+ "grad_norm": 0.109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0448,
+ "step": 3956
+ },
+ {
+ "epoch": 12.642172523961662,
+ "grad_norm": 0.15234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0573,
+ "step": 3957
+ },
+ {
+ "epoch": 12.645367412140574,
+ "grad_norm": 0.06103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0589,
+ "step": 3958
+ },
+ {
+ "epoch": 12.648562300319488,
+ "grad_norm": 0.1376953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0595,
+ "step": 3959
+ },
+ {
+ "epoch": 12.651757188498403,
+ "grad_norm": 0.1572265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0598,
+ "step": 3960
+ },
+ {
+ "epoch": 12.654952076677317,
+ "grad_norm": 0.055908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0676,
+ "step": 3961
+ },
+ {
+ "epoch": 12.65814696485623,
+ "grad_norm": 0.1630859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0475,
+ "step": 3962
+ },
+ {
+ "epoch": 12.661341853035143,
+ "grad_norm": 0.080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0556,
+ "step": 3963
+ },
+ {
+ "epoch": 12.664536741214057,
+ "grad_norm": 0.123046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0446,
+ "step": 3964
+ },
+ {
+ "epoch": 12.667731629392971,
+ "grad_norm": 0.12451171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0402,
+ "step": 3965
+ },
+ {
+ "epoch": 12.670926517571885,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0562,
+ "step": 3966
+ },
+ {
+ "epoch": 12.6741214057508,
+ "grad_norm": 0.134765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0709,
+ "step": 3967
+ },
+ {
+ "epoch": 12.677316293929712,
+ "grad_norm": 0.05712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0529,
+ "step": 3968
+ },
+ {
+ "epoch": 12.680511182108626,
+ "grad_norm": 0.08203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0447,
+ "step": 3969
+ },
+ {
+ "epoch": 12.68370607028754,
+ "grad_norm": 0.058837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0402,
+ "step": 3970
+ },
+ {
+ "epoch": 12.686900958466454,
+ "grad_norm": 0.11962890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0448,
+ "step": 3971
+ },
+ {
+ "epoch": 12.690095846645367,
+ "grad_norm": 0.1435546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0629,
+ "step": 3972
+ },
+ {
+ "epoch": 12.69329073482428,
+ "grad_norm": 0.11328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0454,
+ "step": 3973
+ },
+ {
+ "epoch": 12.696485623003195,
+ "grad_norm": 0.2294921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0643,
+ "step": 3974
+ },
+ {
+ "epoch": 12.699680511182109,
+ "grad_norm": 0.1328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0617,
+ "step": 3975
+ },
+ {
+ "epoch": 12.702875399361023,
+ "grad_norm": 0.09619140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0538,
+ "step": 3976
+ },
+ {
+ "epoch": 12.706070287539935,
+ "grad_norm": 0.09521484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0632,
+ "step": 3977
+ },
+ {
+ "epoch": 12.70926517571885,
+ "grad_norm": 0.09521484375,
+ "learning_rate": 0.0005,
+ "loss": 1.065,
+ "step": 3978
+ },
+ {
+ "epoch": 12.712460063897764,
+ "grad_norm": 0.1015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0588,
+ "step": 3979
+ },
+ {
+ "epoch": 12.715654952076678,
+ "grad_norm": 0.08349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0455,
+ "step": 3980
+ },
+ {
+ "epoch": 12.718849840255592,
+ "grad_norm": 0.09375,
+ "learning_rate": 0.0005,
+ "loss": 1.0715,
+ "step": 3981
+ },
+ {
+ "epoch": 12.722044728434504,
+ "grad_norm": 0.1220703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0695,
+ "step": 3982
+ },
+ {
+ "epoch": 12.725239616613418,
+ "grad_norm": 0.2333984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0626,
+ "step": 3983
+ },
+ {
+ "epoch": 12.728434504792332,
+ "grad_norm": 0.15625,
+ "learning_rate": 0.0005,
+ "loss": 1.0805,
+ "step": 3984
+ },
+ {
+ "epoch": 12.731629392971247,
+ "grad_norm": 0.126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0431,
+ "step": 3985
+ },
+ {
+ "epoch": 12.73482428115016,
+ "grad_norm": 0.146484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0643,
+ "step": 3986
+ },
+ {
+ "epoch": 12.738019169329073,
+ "grad_norm": 0.091796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0597,
+ "step": 3987
+ },
+ {
+ "epoch": 12.741214057507987,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0555,
+ "step": 3988
+ },
+ {
+ "epoch": 12.744408945686901,
+ "grad_norm": 0.0986328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0588,
+ "step": 3989
+ },
+ {
+ "epoch": 12.747603833865815,
+ "grad_norm": 0.056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0595,
+ "step": 3990
+ },
+ {
+ "epoch": 12.750798722044728,
+ "grad_norm": 0.10400390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0602,
+ "step": 3991
+ },
+ {
+ "epoch": 12.753993610223642,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0583,
+ "step": 3992
+ },
+ {
+ "epoch": 12.757188498402556,
+ "grad_norm": 0.1162109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0552,
+ "step": 3993
+ },
+ {
+ "epoch": 12.76038338658147,
+ "grad_norm": 0.11474609375,
+ "learning_rate": 0.0005,
+ "loss": 1.058,
+ "step": 3994
+ },
+ {
+ "epoch": 12.763578274760384,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0562,
+ "step": 3995
+ },
+ {
+ "epoch": 12.766773162939296,
+ "grad_norm": 0.0849609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0623,
+ "step": 3996
+ },
+ {
+ "epoch": 12.76996805111821,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0538,
+ "step": 3997
+ },
+ {
+ "epoch": 12.773162939297125,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0587,
+ "step": 3998
+ },
+ {
+ "epoch": 12.776357827476039,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0438,
+ "step": 3999
+ },
+ {
+ "epoch": 12.779552715654953,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0631,
+ "step": 4000
+ },
+ {
+ "epoch": 12.782747603833865,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0614,
+ "step": 4001
+ },
+ {
+ "epoch": 12.78594249201278,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0559,
+ "step": 4002
+ },
+ {
+ "epoch": 12.789137380191693,
+ "grad_norm": 0.05859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0669,
+ "step": 4003
+ },
+ {
+ "epoch": 12.792332268370608,
+ "grad_norm": 0.058837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0526,
+ "step": 4004
+ },
+ {
+ "epoch": 12.795527156549522,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0691,
+ "step": 4005
+ },
+ {
+ "epoch": 12.798722044728434,
+ "grad_norm": 0.062255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0483,
+ "step": 4006
+ },
+ {
+ "epoch": 12.801916932907348,
+ "grad_norm": 0.05322265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0509,
+ "step": 4007
+ },
+ {
+ "epoch": 12.805111821086262,
+ "grad_norm": 0.05078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0565,
+ "step": 4008
+ },
+ {
+ "epoch": 12.808306709265176,
+ "grad_norm": 0.05224609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0506,
+ "step": 4009
+ },
+ {
+ "epoch": 12.811501597444089,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0572,
+ "step": 4010
+ },
+ {
+ "epoch": 12.814696485623003,
+ "grad_norm": 0.04931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0429,
+ "step": 4011
+ },
+ {
+ "epoch": 12.817891373801917,
+ "grad_norm": 0.080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0581,
+ "step": 4012
+ },
+ {
+ "epoch": 12.821086261980831,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0603,
+ "step": 4013
+ },
+ {
+ "epoch": 12.824281150159745,
+ "grad_norm": 0.08056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0562,
+ "step": 4014
+ },
+ {
+ "epoch": 12.827476038338657,
+ "grad_norm": 0.09716796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0521,
+ "step": 4015
+ },
+ {
+ "epoch": 12.830670926517572,
+ "grad_norm": 0.08544921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0526,
+ "step": 4016
+ },
+ {
+ "epoch": 12.833865814696486,
+ "grad_norm": 0.08154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0547,
+ "step": 4017
+ },
+ {
+ "epoch": 12.8370607028754,
+ "grad_norm": 0.060302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0629,
+ "step": 4018
+ },
+ {
+ "epoch": 12.840255591054314,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0518,
+ "step": 4019
+ },
+ {
+ "epoch": 12.843450479233226,
+ "grad_norm": 0.0546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0605,
+ "step": 4020
+ },
+ {
+ "epoch": 12.84664536741214,
+ "grad_norm": 0.062255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0574,
+ "step": 4021
+ },
+ {
+ "epoch": 12.849840255591054,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0683,
+ "step": 4022
+ },
+ {
+ "epoch": 12.853035143769969,
+ "grad_norm": 0.054443359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0427,
+ "step": 4023
+ },
+ {
+ "epoch": 12.856230031948883,
+ "grad_norm": 0.054443359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0592,
+ "step": 4024
+ },
+ {
+ "epoch": 12.859424920127795,
+ "grad_norm": 0.06103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0505,
+ "step": 4025
+ },
+ {
+ "epoch": 12.86261980830671,
+ "grad_norm": 0.05322265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0573,
+ "step": 4026
+ },
+ {
+ "epoch": 12.865814696485623,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0523,
+ "step": 4027
+ },
+ {
+ "epoch": 12.869009584664537,
+ "grad_norm": 0.053955078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0558,
+ "step": 4028
+ },
+ {
+ "epoch": 12.87220447284345,
+ "grad_norm": 0.10498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0365,
+ "step": 4029
+ },
+ {
+ "epoch": 12.875399361022364,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0494,
+ "step": 4030
+ },
+ {
+ "epoch": 12.878594249201278,
+ "grad_norm": 0.11767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0552,
+ "step": 4031
+ },
+ {
+ "epoch": 12.881789137380192,
+ "grad_norm": 0.0810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0602,
+ "step": 4032
+ },
+ {
+ "epoch": 12.884984025559106,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.066,
+ "step": 4033
+ },
+ {
+ "epoch": 12.888178913738018,
+ "grad_norm": 0.10205078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0472,
+ "step": 4034
+ },
+ {
+ "epoch": 12.891373801916933,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.052,
+ "step": 4035
+ },
+ {
+ "epoch": 12.894568690095847,
+ "grad_norm": 0.1025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0496,
+ "step": 4036
+ },
+ {
+ "epoch": 12.89776357827476,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0623,
+ "step": 4037
+ },
+ {
+ "epoch": 12.900958466453675,
+ "grad_norm": 0.08837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0535,
+ "step": 4038
+ },
+ {
+ "epoch": 12.904153354632587,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0483,
+ "step": 4039
+ },
+ {
+ "epoch": 12.907348242811501,
+ "grad_norm": 0.08642578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0751,
+ "step": 4040
+ },
+ {
+ "epoch": 12.910543130990416,
+ "grad_norm": 0.0986328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0595,
+ "step": 4041
+ },
+ {
+ "epoch": 12.91373801916933,
+ "grad_norm": 0.087890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0481,
+ "step": 4042
+ },
+ {
+ "epoch": 12.916932907348244,
+ "grad_norm": 0.1044921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0462,
+ "step": 4043
+ },
+ {
+ "epoch": 12.920127795527156,
+ "grad_norm": 0.1533203125,
+ "learning_rate": 0.0005,
+ "loss": 1.054,
+ "step": 4044
+ },
+ {
+ "epoch": 12.92332268370607,
+ "grad_norm": 0.2080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0597,
+ "step": 4045
+ },
+ {
+ "epoch": 12.926517571884984,
+ "grad_norm": 0.1064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0653,
+ "step": 4046
+ },
+ {
+ "epoch": 12.929712460063898,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0485,
+ "step": 4047
+ },
+ {
+ "epoch": 12.93290734824281,
+ "grad_norm": 0.123046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0627,
+ "step": 4048
+ },
+ {
+ "epoch": 12.936102236421725,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0549,
+ "step": 4049
+ },
+ {
+ "epoch": 12.939297124600639,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0558,
+ "step": 4050
+ },
+ {
+ "epoch": 12.942492012779553,
+ "grad_norm": 0.0888671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0599,
+ "step": 4051
+ },
+ {
+ "epoch": 12.945686900958467,
+ "grad_norm": 0.09130859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0583,
+ "step": 4052
+ },
+ {
+ "epoch": 12.94888178913738,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0562,
+ "step": 4053
+ },
+ {
+ "epoch": 12.952076677316294,
+ "grad_norm": 0.09765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0684,
+ "step": 4054
+ },
+ {
+ "epoch": 12.955271565495208,
+ "grad_norm": 0.1845703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0523,
+ "step": 4055
+ },
+ {
+ "epoch": 12.958466453674122,
+ "grad_norm": 0.1416015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0699,
+ "step": 4056
+ },
+ {
+ "epoch": 12.961661341853034,
+ "grad_norm": 0.1396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.071,
+ "step": 4057
+ },
+ {
+ "epoch": 12.964856230031948,
+ "grad_norm": 0.2177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0451,
+ "step": 4058
+ },
+ {
+ "epoch": 12.968051118210862,
+ "grad_norm": 0.08056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0656,
+ "step": 4059
+ },
+ {
+ "epoch": 12.971246006389777,
+ "grad_norm": 0.1123046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0473,
+ "step": 4060
+ },
+ {
+ "epoch": 12.97444089456869,
+ "grad_norm": 0.0888671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0504,
+ "step": 4061
+ },
+ {
+ "epoch": 12.977635782747605,
+ "grad_norm": 0.11181640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0519,
+ "step": 4062
+ },
+ {
+ "epoch": 12.980830670926517,
+ "grad_norm": 0.140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0662,
+ "step": 4063
+ },
+ {
+ "epoch": 12.984025559105431,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0589,
+ "step": 4064
+ },
+ {
+ "epoch": 12.987220447284345,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0624,
+ "step": 4065
+ },
+ {
+ "epoch": 12.99041533546326,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0479,
+ "step": 4066
+ },
+ {
+ "epoch": 12.993610223642172,
+ "grad_norm": 0.07861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0633,
+ "step": 4067
+ },
+ {
+ "epoch": 12.996805111821086,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0573,
+ "step": 4068
+ },
+ {
+ "epoch": 13.0,
+ "grad_norm": 0.08251953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0645,
+ "step": 4069
+ },
+ {
+ "epoch": 13.003194888178914,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0514,
+ "step": 4070
+ },
+ {
+ "epoch": 13.006389776357828,
+ "grad_norm": 0.0830078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0557,
+ "step": 4071
+ },
+ {
+ "epoch": 13.00958466453674,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0511,
+ "step": 4072
+ },
+ {
+ "epoch": 13.012779552715655,
+ "grad_norm": 0.056396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0459,
+ "step": 4073
+ },
+ {
+ "epoch": 13.015974440894569,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0624,
+ "step": 4074
+ },
+ {
+ "epoch": 13.019169329073483,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0596,
+ "step": 4075
+ },
+ {
+ "epoch": 13.022364217252397,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0415,
+ "step": 4076
+ },
+ {
+ "epoch": 13.02555910543131,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0549,
+ "step": 4077
+ },
+ {
+ "epoch": 13.028753993610223,
+ "grad_norm": 0.053955078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0606,
+ "step": 4078
+ },
+ {
+ "epoch": 13.031948881789138,
+ "grad_norm": 0.12451171875,
+ "learning_rate": 0.0005,
+ "loss": 1.058,
+ "step": 4079
+ },
+ {
+ "epoch": 13.035143769968052,
+ "grad_norm": 0.15234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0581,
+ "step": 4080
+ },
+ {
+ "epoch": 13.038338658146964,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0629,
+ "step": 4081
+ },
+ {
+ "epoch": 13.041533546325878,
+ "grad_norm": 0.12255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0369,
+ "step": 4082
+ },
+ {
+ "epoch": 13.044728434504792,
+ "grad_norm": 0.05712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0546,
+ "step": 4083
+ },
+ {
+ "epoch": 13.047923322683706,
+ "grad_norm": 0.0966796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0613,
+ "step": 4084
+ },
+ {
+ "epoch": 13.05111821086262,
+ "grad_norm": 0.0947265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0665,
+ "step": 4085
+ },
+ {
+ "epoch": 13.054313099041533,
+ "grad_norm": 0.06201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0499,
+ "step": 4086
+ },
+ {
+ "epoch": 13.057507987220447,
+ "grad_norm": 0.091796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0513,
+ "step": 4087
+ },
+ {
+ "epoch": 13.060702875399361,
+ "grad_norm": 0.11083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0658,
+ "step": 4088
+ },
+ {
+ "epoch": 13.063897763578275,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0635,
+ "step": 4089
+ },
+ {
+ "epoch": 13.06709265175719,
+ "grad_norm": 0.1015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0513,
+ "step": 4090
+ },
+ {
+ "epoch": 13.070287539936102,
+ "grad_norm": 0.054931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0554,
+ "step": 4091
+ },
+ {
+ "epoch": 13.073482428115016,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0598,
+ "step": 4092
+ },
+ {
+ "epoch": 13.07667731629393,
+ "grad_norm": 0.10791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0577,
+ "step": 4093
+ },
+ {
+ "epoch": 13.079872204472844,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0698,
+ "step": 4094
+ },
+ {
+ "epoch": 13.083067092651758,
+ "grad_norm": 0.10498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0525,
+ "step": 4095
+ },
+ {
+ "epoch": 13.08626198083067,
+ "grad_norm": 0.06103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0534,
+ "step": 4096
+ },
+ {
+ "epoch": 13.089456869009584,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0494,
+ "step": 4097
+ },
+ {
+ "epoch": 13.092651757188499,
+ "grad_norm": 0.0615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0594,
+ "step": 4098
+ },
+ {
+ "epoch": 13.095846645367413,
+ "grad_norm": 0.054931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0408,
+ "step": 4099
+ },
+ {
+ "epoch": 13.099041533546325,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0684,
+ "step": 4100
+ },
+ {
+ "epoch": 13.10223642172524,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0583,
+ "step": 4101
+ },
+ {
+ "epoch": 13.105431309904153,
+ "grad_norm": 0.08203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0568,
+ "step": 4102
+ },
+ {
+ "epoch": 13.108626198083067,
+ "grad_norm": 0.0830078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0508,
+ "step": 4103
+ },
+ {
+ "epoch": 13.111821086261982,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0622,
+ "step": 4104
+ },
+ {
+ "epoch": 13.115015974440894,
+ "grad_norm": 0.08154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0605,
+ "step": 4105
+ },
+ {
+ "epoch": 13.118210862619808,
+ "grad_norm": 0.05859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0418,
+ "step": 4106
+ },
+ {
+ "epoch": 13.121405750798722,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.061,
+ "step": 4107
+ },
+ {
+ "epoch": 13.124600638977636,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0458,
+ "step": 4108
+ },
+ {
+ "epoch": 13.12779552715655,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0492,
+ "step": 4109
+ },
+ {
+ "epoch": 13.130990415335463,
+ "grad_norm": 0.0615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.051,
+ "step": 4110
+ },
+ {
+ "epoch": 13.134185303514377,
+ "grad_norm": 0.0615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0595,
+ "step": 4111
+ },
+ {
+ "epoch": 13.13738019169329,
+ "grad_norm": 0.052734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0706,
+ "step": 4112
+ },
+ {
+ "epoch": 13.140575079872205,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0687,
+ "step": 4113
+ },
+ {
+ "epoch": 13.143769968051119,
+ "grad_norm": 0.109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0667,
+ "step": 4114
+ },
+ {
+ "epoch": 13.146964856230031,
+ "grad_norm": 0.05029296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0488,
+ "step": 4115
+ },
+ {
+ "epoch": 13.150159744408946,
+ "grad_norm": 0.10205078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0729,
+ "step": 4116
+ },
+ {
+ "epoch": 13.15335463258786,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.065,
+ "step": 4117
+ },
+ {
+ "epoch": 13.156549520766774,
+ "grad_norm": 0.08447265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0599,
+ "step": 4118
+ },
+ {
+ "epoch": 13.159744408945686,
+ "grad_norm": 0.0986328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0487,
+ "step": 4119
+ },
+ {
+ "epoch": 13.1629392971246,
+ "grad_norm": 0.056884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.07,
+ "step": 4120
+ },
+ {
+ "epoch": 13.166134185303514,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.061,
+ "step": 4121
+ },
+ {
+ "epoch": 13.169329073482428,
+ "grad_norm": 0.062255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0672,
+ "step": 4122
+ },
+ {
+ "epoch": 13.172523961661343,
+ "grad_norm": 0.0869140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0592,
+ "step": 4123
+ },
+ {
+ "epoch": 13.175718849840255,
+ "grad_norm": 0.058837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0643,
+ "step": 4124
+ },
+ {
+ "epoch": 13.178913738019169,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0612,
+ "step": 4125
+ },
+ {
+ "epoch": 13.182108626198083,
+ "grad_norm": 0.057861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0618,
+ "step": 4126
+ },
+ {
+ "epoch": 13.185303514376997,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0537,
+ "step": 4127
+ },
+ {
+ "epoch": 13.188498402555911,
+ "grad_norm": 0.09130859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0571,
+ "step": 4128
+ },
+ {
+ "epoch": 13.191693290734824,
+ "grad_norm": 0.0810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0494,
+ "step": 4129
+ },
+ {
+ "epoch": 13.194888178913738,
+ "grad_norm": 0.1103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.057,
+ "step": 4130
+ },
+ {
+ "epoch": 13.198083067092652,
+ "grad_norm": 0.099609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0608,
+ "step": 4131
+ },
+ {
+ "epoch": 13.201277955271566,
+ "grad_norm": 0.2333984375,
+ "learning_rate": 0.0005,
+ "loss": 1.072,
+ "step": 4132
+ },
+ {
+ "epoch": 13.204472843450478,
+ "grad_norm": 0.27734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0591,
+ "step": 4133
+ },
+ {
+ "epoch": 13.207667731629392,
+ "grad_norm": 0.17578125,
+ "learning_rate": 0.0005,
+ "loss": 1.051,
+ "step": 4134
+ },
+ {
+ "epoch": 13.210862619808307,
+ "grad_norm": 0.080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0412,
+ "step": 4135
+ },
+ {
+ "epoch": 13.21405750798722,
+ "grad_norm": 0.19921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0678,
+ "step": 4136
+ },
+ {
+ "epoch": 13.217252396166135,
+ "grad_norm": 0.1640625,
+ "learning_rate": 0.0005,
+ "loss": 1.05,
+ "step": 4137
+ },
+ {
+ "epoch": 13.220447284345047,
+ "grad_norm": 0.08740234375,
+ "learning_rate": 0.0005,
+ "loss": 1.059,
+ "step": 4138
+ },
+ {
+ "epoch": 13.223642172523961,
+ "grad_norm": 0.291015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0611,
+ "step": 4139
+ },
+ {
+ "epoch": 13.226837060702875,
+ "grad_norm": 0.30859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0491,
+ "step": 4140
+ },
+ {
+ "epoch": 13.23003194888179,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0547,
+ "step": 4141
+ },
+ {
+ "epoch": 13.233226837060704,
+ "grad_norm": 0.2333984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0539,
+ "step": 4142
+ },
+ {
+ "epoch": 13.236421725239616,
+ "grad_norm": 0.2138671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0688,
+ "step": 4143
+ },
+ {
+ "epoch": 13.23961661341853,
+ "grad_norm": 0.08740234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0574,
+ "step": 4144
+ },
+ {
+ "epoch": 13.242811501597444,
+ "grad_norm": 0.15625,
+ "learning_rate": 0.0005,
+ "loss": 1.0702,
+ "step": 4145
+ },
+ {
+ "epoch": 13.246006389776358,
+ "grad_norm": 0.2392578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0649,
+ "step": 4146
+ },
+ {
+ "epoch": 13.249201277955272,
+ "grad_norm": 0.208984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0545,
+ "step": 4147
+ },
+ {
+ "epoch": 13.252396166134185,
+ "grad_norm": 0.1083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0543,
+ "step": 4148
+ },
+ {
+ "epoch": 13.255591054313099,
+ "grad_norm": 0.35546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0594,
+ "step": 4149
+ },
+ {
+ "epoch": 13.258785942492013,
+ "grad_norm": 0.359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0513,
+ "step": 4150
+ },
+ {
+ "epoch": 13.261980830670927,
+ "grad_norm": 0.2255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0489,
+ "step": 4151
+ },
+ {
+ "epoch": 13.26517571884984,
+ "grad_norm": 0.09375,
+ "learning_rate": 0.0005,
+ "loss": 1.0584,
+ "step": 4152
+ },
+ {
+ "epoch": 13.268370607028753,
+ "grad_norm": 0.291015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0455,
+ "step": 4153
+ },
+ {
+ "epoch": 13.271565495207668,
+ "grad_norm": 0.1884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0514,
+ "step": 4154
+ },
+ {
+ "epoch": 13.274760383386582,
+ "grad_norm": 0.1669921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0514,
+ "step": 4155
+ },
+ {
+ "epoch": 13.277955271565496,
+ "grad_norm": 0.470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0554,
+ "step": 4156
+ },
+ {
+ "epoch": 13.281150159744408,
+ "grad_norm": 0.59765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0514,
+ "step": 4157
+ },
+ {
+ "epoch": 13.284345047923322,
+ "grad_norm": 0.671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0681,
+ "step": 4158
+ },
+ {
+ "epoch": 13.287539936102236,
+ "grad_norm": 0.89453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0592,
+ "step": 4159
+ },
+ {
+ "epoch": 13.29073482428115,
+ "grad_norm": 1.7578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0763,
+ "step": 4160
+ },
+ {
+ "epoch": 13.293929712460065,
+ "grad_norm": 0.49609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0576,
+ "step": 4161
+ },
+ {
+ "epoch": 13.297124600638977,
+ "grad_norm": 0.42578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0465,
+ "step": 4162
+ },
+ {
+ "epoch": 13.300319488817891,
+ "grad_norm": 1.4453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0712,
+ "step": 4163
+ },
+ {
+ "epoch": 13.303514376996805,
+ "grad_norm": 0.91015625,
+ "learning_rate": 0.0005,
+ "loss": 1.063,
+ "step": 4164
+ },
+ {
+ "epoch": 13.30670926517572,
+ "grad_norm": 0.298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.055,
+ "step": 4165
+ },
+ {
+ "epoch": 13.309904153354633,
+ "grad_norm": 0.349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.059,
+ "step": 4166
+ },
+ {
+ "epoch": 13.313099041533546,
+ "grad_norm": 0.408203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0603,
+ "step": 4167
+ },
+ {
+ "epoch": 13.31629392971246,
+ "grad_norm": 0.1826171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0473,
+ "step": 4168
+ },
+ {
+ "epoch": 13.319488817891374,
+ "grad_norm": 0.52734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0601,
+ "step": 4169
+ },
+ {
+ "epoch": 13.322683706070288,
+ "grad_norm": 0.2373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0505,
+ "step": 4170
+ },
+ {
+ "epoch": 13.3258785942492,
+ "grad_norm": 0.28125,
+ "learning_rate": 0.0005,
+ "loss": 1.0566,
+ "step": 4171
+ },
+ {
+ "epoch": 13.329073482428115,
+ "grad_norm": 0.47265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0574,
+ "step": 4172
+ },
+ {
+ "epoch": 13.332268370607029,
+ "grad_norm": 0.1845703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0665,
+ "step": 4173
+ },
+ {
+ "epoch": 13.335463258785943,
+ "grad_norm": 0.2080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.062,
+ "step": 4174
+ },
+ {
+ "epoch": 13.338658146964857,
+ "grad_norm": 0.16796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0559,
+ "step": 4175
+ },
+ {
+ "epoch": 13.34185303514377,
+ "grad_norm": 0.1337890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0529,
+ "step": 4176
+ },
+ {
+ "epoch": 13.345047923322683,
+ "grad_norm": 0.119140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0452,
+ "step": 4177
+ },
+ {
+ "epoch": 13.348242811501597,
+ "grad_norm": 0.123046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0485,
+ "step": 4178
+ },
+ {
+ "epoch": 13.351437699680512,
+ "grad_norm": 0.09033203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0588,
+ "step": 4179
+ },
+ {
+ "epoch": 13.354632587859426,
+ "grad_norm": 0.08056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0738,
+ "step": 4180
+ },
+ {
+ "epoch": 13.357827476038338,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0465,
+ "step": 4181
+ },
+ {
+ "epoch": 13.361022364217252,
+ "grad_norm": 0.080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0508,
+ "step": 4182
+ },
+ {
+ "epoch": 13.364217252396166,
+ "grad_norm": 0.11376953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0654,
+ "step": 4183
+ },
+ {
+ "epoch": 13.36741214057508,
+ "grad_norm": 0.1142578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0623,
+ "step": 4184
+ },
+ {
+ "epoch": 13.370607028753994,
+ "grad_norm": 0.0927734375,
+ "learning_rate": 0.0005,
+ "loss": 1.046,
+ "step": 4185
+ },
+ {
+ "epoch": 13.373801916932907,
+ "grad_norm": 0.0849609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0446,
+ "step": 4186
+ },
+ {
+ "epoch": 13.37699680511182,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0492,
+ "step": 4187
+ },
+ {
+ "epoch": 13.380191693290735,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0521,
+ "step": 4188
+ },
+ {
+ "epoch": 13.383386581469649,
+ "grad_norm": 0.060791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0595,
+ "step": 4189
+ },
+ {
+ "epoch": 13.386581469648561,
+ "grad_norm": 0.05712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0457,
+ "step": 4190
+ },
+ {
+ "epoch": 13.389776357827476,
+ "grad_norm": 0.0576171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0566,
+ "step": 4191
+ },
+ {
+ "epoch": 13.39297124600639,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0589,
+ "step": 4192
+ },
+ {
+ "epoch": 13.396166134185304,
+ "grad_norm": 0.080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.054,
+ "step": 4193
+ },
+ {
+ "epoch": 13.399361022364218,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0651,
+ "step": 4194
+ },
+ {
+ "epoch": 13.40255591054313,
+ "grad_norm": 0.09033203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0538,
+ "step": 4195
+ },
+ {
+ "epoch": 13.405750798722044,
+ "grad_norm": 0.10595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0594,
+ "step": 4196
+ },
+ {
+ "epoch": 13.408945686900958,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0494,
+ "step": 4197
+ },
+ {
+ "epoch": 13.412140575079873,
+ "grad_norm": 0.1455078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0557,
+ "step": 4198
+ },
+ {
+ "epoch": 13.415335463258787,
+ "grad_norm": 0.08837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0482,
+ "step": 4199
+ },
+ {
+ "epoch": 13.418530351437699,
+ "grad_norm": 0.099609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0453,
+ "step": 4200
+ },
+ {
+ "epoch": 13.421725239616613,
+ "grad_norm": 0.12451171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0444,
+ "step": 4201
+ },
+ {
+ "epoch": 13.424920127795527,
+ "grad_norm": 0.0966796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0547,
+ "step": 4202
+ },
+ {
+ "epoch": 13.428115015974441,
+ "grad_norm": 0.11767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0492,
+ "step": 4203
+ },
+ {
+ "epoch": 13.431309904153355,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0523,
+ "step": 4204
+ },
+ {
+ "epoch": 13.434504792332268,
+ "grad_norm": 0.0927734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0634,
+ "step": 4205
+ },
+ {
+ "epoch": 13.437699680511182,
+ "grad_norm": 0.09912109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0477,
+ "step": 4206
+ },
+ {
+ "epoch": 13.440894568690096,
+ "grad_norm": 0.11767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0542,
+ "step": 4207
+ },
+ {
+ "epoch": 13.44408945686901,
+ "grad_norm": 0.1005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0548,
+ "step": 4208
+ },
+ {
+ "epoch": 13.447284345047922,
+ "grad_norm": 0.09228515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0617,
+ "step": 4209
+ },
+ {
+ "epoch": 13.450479233226837,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0471,
+ "step": 4210
+ },
+ {
+ "epoch": 13.45367412140575,
+ "grad_norm": 0.10986328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0505,
+ "step": 4211
+ },
+ {
+ "epoch": 13.456869009584665,
+ "grad_norm": 0.0947265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0641,
+ "step": 4212
+ },
+ {
+ "epoch": 13.460063897763579,
+ "grad_norm": 0.1376953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0665,
+ "step": 4213
+ },
+ {
+ "epoch": 13.463258785942491,
+ "grad_norm": 0.056884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0585,
+ "step": 4214
+ },
+ {
+ "epoch": 13.466453674121405,
+ "grad_norm": 0.1572265625,
+ "learning_rate": 0.0005,
+ "loss": 1.054,
+ "step": 4215
+ },
+ {
+ "epoch": 13.46964856230032,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0621,
+ "step": 4216
+ },
+ {
+ "epoch": 13.472843450479234,
+ "grad_norm": 0.08837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0632,
+ "step": 4217
+ },
+ {
+ "epoch": 13.476038338658148,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0729,
+ "step": 4218
+ },
+ {
+ "epoch": 13.47923322683706,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0511,
+ "step": 4219
+ },
+ {
+ "epoch": 13.482428115015974,
+ "grad_norm": 0.061279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0504,
+ "step": 4220
+ },
+ {
+ "epoch": 13.485623003194888,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0438,
+ "step": 4221
+ },
+ {
+ "epoch": 13.488817891373802,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0521,
+ "step": 4222
+ },
+ {
+ "epoch": 13.492012779552716,
+ "grad_norm": 0.056884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0586,
+ "step": 4223
+ },
+ {
+ "epoch": 13.495207667731629,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0623,
+ "step": 4224
+ },
+ {
+ "epoch": 13.498402555910543,
+ "grad_norm": 0.058837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0479,
+ "step": 4225
+ },
+ {
+ "epoch": 13.501597444089457,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0526,
+ "step": 4226
+ },
+ {
+ "epoch": 13.504792332268371,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0584,
+ "step": 4227
+ },
+ {
+ "epoch": 13.507987220447284,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0574,
+ "step": 4228
+ },
+ {
+ "epoch": 13.511182108626198,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0564,
+ "step": 4229
+ },
+ {
+ "epoch": 13.514376996805112,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0524,
+ "step": 4230
+ },
+ {
+ "epoch": 13.517571884984026,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.055,
+ "step": 4231
+ },
+ {
+ "epoch": 13.52076677316294,
+ "grad_norm": 0.05859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0622,
+ "step": 4232
+ },
+ {
+ "epoch": 13.523961661341852,
+ "grad_norm": 0.053466796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0612,
+ "step": 4233
+ },
+ {
+ "epoch": 13.527156549520766,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0574,
+ "step": 4234
+ },
+ {
+ "epoch": 13.53035143769968,
+ "grad_norm": 0.057373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0531,
+ "step": 4235
+ },
+ {
+ "epoch": 13.533546325878595,
+ "grad_norm": 0.0859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0476,
+ "step": 4236
+ },
+ {
+ "epoch": 13.536741214057509,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0483,
+ "step": 4237
+ },
+ {
+ "epoch": 13.539936102236421,
+ "grad_norm": 0.0859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0603,
+ "step": 4238
+ },
+ {
+ "epoch": 13.543130990415335,
+ "grad_norm": 0.054443359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0548,
+ "step": 4239
+ },
+ {
+ "epoch": 13.54632587859425,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0493,
+ "step": 4240
+ },
+ {
+ "epoch": 13.549520766773163,
+ "grad_norm": 0.0498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0576,
+ "step": 4241
+ },
+ {
+ "epoch": 13.552715654952078,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0402,
+ "step": 4242
+ },
+ {
+ "epoch": 13.55591054313099,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0559,
+ "step": 4243
+ },
+ {
+ "epoch": 13.559105431309904,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0438,
+ "step": 4244
+ },
+ {
+ "epoch": 13.562300319488818,
+ "grad_norm": 0.059814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0528,
+ "step": 4245
+ },
+ {
+ "epoch": 13.565495207667732,
+ "grad_norm": 0.05712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0688,
+ "step": 4246
+ },
+ {
+ "epoch": 13.568690095846645,
+ "grad_norm": 0.057373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0554,
+ "step": 4247
+ },
+ {
+ "epoch": 13.571884984025559,
+ "grad_norm": 0.05224609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0458,
+ "step": 4248
+ },
+ {
+ "epoch": 13.575079872204473,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0574,
+ "step": 4249
+ },
+ {
+ "epoch": 13.578274760383387,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0497,
+ "step": 4250
+ },
+ {
+ "epoch": 13.581469648562301,
+ "grad_norm": 0.08251953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0554,
+ "step": 4251
+ },
+ {
+ "epoch": 13.584664536741213,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0539,
+ "step": 4252
+ },
+ {
+ "epoch": 13.587859424920127,
+ "grad_norm": 0.10205078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0581,
+ "step": 4253
+ },
+ {
+ "epoch": 13.591054313099042,
+ "grad_norm": 0.056884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0499,
+ "step": 4254
+ },
+ {
+ "epoch": 13.594249201277956,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0723,
+ "step": 4255
+ },
+ {
+ "epoch": 13.59744408945687,
+ "grad_norm": 0.05712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0693,
+ "step": 4256
+ },
+ {
+ "epoch": 13.600638977635782,
+ "grad_norm": 0.056396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0701,
+ "step": 4257
+ },
+ {
+ "epoch": 13.603833865814696,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0414,
+ "step": 4258
+ },
+ {
+ "epoch": 13.60702875399361,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0568,
+ "step": 4259
+ },
+ {
+ "epoch": 13.610223642172524,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0492,
+ "step": 4260
+ },
+ {
+ "epoch": 13.613418530351439,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0494,
+ "step": 4261
+ },
+ {
+ "epoch": 13.616613418530351,
+ "grad_norm": 0.0966796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0624,
+ "step": 4262
+ },
+ {
+ "epoch": 13.619808306709265,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0587,
+ "step": 4263
+ },
+ {
+ "epoch": 13.62300319488818,
+ "grad_norm": 0.08056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0439,
+ "step": 4264
+ },
+ {
+ "epoch": 13.626198083067093,
+ "grad_norm": 0.083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0551,
+ "step": 4265
+ },
+ {
+ "epoch": 13.629392971246006,
+ "grad_norm": 0.10009765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0532,
+ "step": 4266
+ },
+ {
+ "epoch": 13.63258785942492,
+ "grad_norm": 0.05712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0593,
+ "step": 4267
+ },
+ {
+ "epoch": 13.635782747603834,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.044,
+ "step": 4268
+ },
+ {
+ "epoch": 13.638977635782748,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0588,
+ "step": 4269
+ },
+ {
+ "epoch": 13.642172523961662,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0668,
+ "step": 4270
+ },
+ {
+ "epoch": 13.645367412140574,
+ "grad_norm": 0.09130859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0493,
+ "step": 4271
+ },
+ {
+ "epoch": 13.648562300319488,
+ "grad_norm": 0.08642578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0575,
+ "step": 4272
+ },
+ {
+ "epoch": 13.651757188498403,
+ "grad_norm": 0.09375,
+ "learning_rate": 0.0005,
+ "loss": 1.0612,
+ "step": 4273
+ },
+ {
+ "epoch": 13.654952076677317,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0646,
+ "step": 4274
+ },
+ {
+ "epoch": 13.65814696485623,
+ "grad_norm": 0.09228515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0531,
+ "step": 4275
+ },
+ {
+ "epoch": 13.661341853035143,
+ "grad_norm": 0.057861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.047,
+ "step": 4276
+ },
+ {
+ "epoch": 13.664536741214057,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0385,
+ "step": 4277
+ },
+ {
+ "epoch": 13.667731629392971,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0604,
+ "step": 4278
+ },
+ {
+ "epoch": 13.670926517571885,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0647,
+ "step": 4279
+ },
+ {
+ "epoch": 13.6741214057508,
+ "grad_norm": 0.12353515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0526,
+ "step": 4280
+ },
+ {
+ "epoch": 13.677316293929712,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.044,
+ "step": 4281
+ },
+ {
+ "epoch": 13.680511182108626,
+ "grad_norm": 0.08837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.05,
+ "step": 4282
+ },
+ {
+ "epoch": 13.68370607028754,
+ "grad_norm": 0.053955078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0671,
+ "step": 4283
+ },
+ {
+ "epoch": 13.686900958466454,
+ "grad_norm": 0.08642578125,
+ "learning_rate": 0.0005,
+ "loss": 1.052,
+ "step": 4284
+ },
+ {
+ "epoch": 13.690095846645367,
+ "grad_norm": 0.060791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0649,
+ "step": 4285
+ },
+ {
+ "epoch": 13.69329073482428,
+ "grad_norm": 0.055908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0499,
+ "step": 4286
+ },
+ {
+ "epoch": 13.696485623003195,
+ "grad_norm": 0.10009765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0658,
+ "step": 4287
+ },
+ {
+ "epoch": 13.699680511182109,
+ "grad_norm": 0.056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0655,
+ "step": 4288
+ },
+ {
+ "epoch": 13.702875399361023,
+ "grad_norm": 0.09423828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0484,
+ "step": 4289
+ },
+ {
+ "epoch": 13.706070287539935,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0485,
+ "step": 4290
+ },
+ {
+ "epoch": 13.70926517571885,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0505,
+ "step": 4291
+ },
+ {
+ "epoch": 13.712460063897764,
+ "grad_norm": 0.0927734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0466,
+ "step": 4292
+ },
+ {
+ "epoch": 13.715654952076678,
+ "grad_norm": 0.1240234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0495,
+ "step": 4293
+ },
+ {
+ "epoch": 13.718849840255592,
+ "grad_norm": 0.205078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0664,
+ "step": 4294
+ },
+ {
+ "epoch": 13.722044728434504,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0671,
+ "step": 4295
+ },
+ {
+ "epoch": 13.725239616613418,
+ "grad_norm": 0.1484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0664,
+ "step": 4296
+ },
+ {
+ "epoch": 13.728434504792332,
+ "grad_norm": 0.111328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0627,
+ "step": 4297
+ },
+ {
+ "epoch": 13.731629392971247,
+ "grad_norm": 0.095703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0508,
+ "step": 4298
+ },
+ {
+ "epoch": 13.73482428115016,
+ "grad_norm": 0.2392578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0679,
+ "step": 4299
+ },
+ {
+ "epoch": 13.738019169329073,
+ "grad_norm": 0.203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0629,
+ "step": 4300
+ },
+ {
+ "epoch": 13.741214057507987,
+ "grad_norm": 0.0947265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0572,
+ "step": 4301
+ },
+ {
+ "epoch": 13.744408945686901,
+ "grad_norm": 0.203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0634,
+ "step": 4302
+ },
+ {
+ "epoch": 13.747603833865815,
+ "grad_norm": 0.0849609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0435,
+ "step": 4303
+ },
+ {
+ "epoch": 13.750798722044728,
+ "grad_norm": 0.1796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0639,
+ "step": 4304
+ },
+ {
+ "epoch": 13.753993610223642,
+ "grad_norm": 0.1123046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0593,
+ "step": 4305
+ },
+ {
+ "epoch": 13.757188498402556,
+ "grad_norm": 0.09716796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0546,
+ "step": 4306
+ },
+ {
+ "epoch": 13.76038338658147,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0573,
+ "step": 4307
+ },
+ {
+ "epoch": 13.763578274760384,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0573,
+ "step": 4308
+ },
+ {
+ "epoch": 13.766773162939296,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0515,
+ "step": 4309
+ },
+ {
+ "epoch": 13.76996805111821,
+ "grad_norm": 0.056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0489,
+ "step": 4310
+ },
+ {
+ "epoch": 13.773162939297125,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0487,
+ "step": 4311
+ },
+ {
+ "epoch": 13.776357827476039,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0559,
+ "step": 4312
+ },
+ {
+ "epoch": 13.779552715654953,
+ "grad_norm": 0.05908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0575,
+ "step": 4313
+ },
+ {
+ "epoch": 13.782747603833865,
+ "grad_norm": 0.053955078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0669,
+ "step": 4314
+ },
+ {
+ "epoch": 13.78594249201278,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0506,
+ "step": 4315
+ },
+ {
+ "epoch": 13.789137380191693,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0499,
+ "step": 4316
+ },
+ {
+ "epoch": 13.792332268370608,
+ "grad_norm": 0.06005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0588,
+ "step": 4317
+ },
+ {
+ "epoch": 13.795527156549522,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0568,
+ "step": 4318
+ },
+ {
+ "epoch": 13.798722044728434,
+ "grad_norm": 0.059326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0495,
+ "step": 4319
+ },
+ {
+ "epoch": 13.801916932907348,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0508,
+ "step": 4320
+ },
+ {
+ "epoch": 13.805111821086262,
+ "grad_norm": 0.052978515625,
+ "learning_rate": 0.0005,
+ "loss": 1.067,
+ "step": 4321
+ },
+ {
+ "epoch": 13.808306709265176,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0591,
+ "step": 4322
+ },
+ {
+ "epoch": 13.811501597444089,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0483,
+ "step": 4323
+ },
+ {
+ "epoch": 13.814696485623003,
+ "grad_norm": 0.09228515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0642,
+ "step": 4324
+ },
+ {
+ "epoch": 13.817891373801917,
+ "grad_norm": 0.0595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.036,
+ "step": 4325
+ },
+ {
+ "epoch": 13.821086261980831,
+ "grad_norm": 0.11279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0473,
+ "step": 4326
+ },
+ {
+ "epoch": 13.824281150159745,
+ "grad_norm": 0.080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0445,
+ "step": 4327
+ },
+ {
+ "epoch": 13.827476038338657,
+ "grad_norm": 0.08349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0621,
+ "step": 4328
+ },
+ {
+ "epoch": 13.830670926517572,
+ "grad_norm": 0.0859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0621,
+ "step": 4329
+ },
+ {
+ "epoch": 13.833865814696486,
+ "grad_norm": 0.0927734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0601,
+ "step": 4330
+ },
+ {
+ "epoch": 13.8370607028754,
+ "grad_norm": 0.1328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0425,
+ "step": 4331
+ },
+ {
+ "epoch": 13.840255591054314,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0567,
+ "step": 4332
+ },
+ {
+ "epoch": 13.843450479233226,
+ "grad_norm": 0.1513671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0376,
+ "step": 4333
+ },
+ {
+ "epoch": 13.84664536741214,
+ "grad_norm": 0.05810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0439,
+ "step": 4334
+ },
+ {
+ "epoch": 13.849840255591054,
+ "grad_norm": 0.115234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0438,
+ "step": 4335
+ },
+ {
+ "epoch": 13.853035143769969,
+ "grad_norm": 0.0908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0485,
+ "step": 4336
+ },
+ {
+ "epoch": 13.856230031948883,
+ "grad_norm": 0.10302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0478,
+ "step": 4337
+ },
+ {
+ "epoch": 13.859424920127795,
+ "grad_norm": 0.091796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0478,
+ "step": 4338
+ },
+ {
+ "epoch": 13.86261980830671,
+ "grad_norm": 0.10302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0612,
+ "step": 4339
+ },
+ {
+ "epoch": 13.865814696485623,
+ "grad_norm": 0.150390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0511,
+ "step": 4340
+ },
+ {
+ "epoch": 13.869009584664537,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0524,
+ "step": 4341
+ },
+ {
+ "epoch": 13.87220447284345,
+ "grad_norm": 0.10791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0572,
+ "step": 4342
+ },
+ {
+ "epoch": 13.875399361022364,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0376,
+ "step": 4343
+ },
+ {
+ "epoch": 13.878594249201278,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0578,
+ "step": 4344
+ },
+ {
+ "epoch": 13.881789137380192,
+ "grad_norm": 0.05712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0659,
+ "step": 4345
+ },
+ {
+ "epoch": 13.884984025559106,
+ "grad_norm": 0.07861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0566,
+ "step": 4346
+ },
+ {
+ "epoch": 13.888178913738018,
+ "grad_norm": 0.0830078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0583,
+ "step": 4347
+ },
+ {
+ "epoch": 13.891373801916933,
+ "grad_norm": 0.1044921875,
+ "learning_rate": 0.0005,
+ "loss": 1.065,
+ "step": 4348
+ },
+ {
+ "epoch": 13.894568690095847,
+ "grad_norm": 0.0859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0486,
+ "step": 4349
+ },
+ {
+ "epoch": 13.89776357827476,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0702,
+ "step": 4350
+ },
+ {
+ "epoch": 13.900958466453675,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0597,
+ "step": 4351
+ },
+ {
+ "epoch": 13.904153354632587,
+ "grad_norm": 0.056884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0599,
+ "step": 4352
+ },
+ {
+ "epoch": 13.907348242811501,
+ "grad_norm": 0.0576171875,
+ "learning_rate": 0.0005,
+ "loss": 1.039,
+ "step": 4353
+ },
+ {
+ "epoch": 13.910543130990416,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0613,
+ "step": 4354
+ },
+ {
+ "epoch": 13.91373801916933,
+ "grad_norm": 0.05078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0599,
+ "step": 4355
+ },
+ {
+ "epoch": 13.916932907348244,
+ "grad_norm": 0.062255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0443,
+ "step": 4356
+ },
+ {
+ "epoch": 13.920127795527156,
+ "grad_norm": 0.056884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0607,
+ "step": 4357
+ },
+ {
+ "epoch": 13.92332268370607,
+ "grad_norm": 0.062255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0659,
+ "step": 4358
+ },
+ {
+ "epoch": 13.926517571884984,
+ "grad_norm": 0.054931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0504,
+ "step": 4359
+ },
+ {
+ "epoch": 13.929712460063898,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0488,
+ "step": 4360
+ },
+ {
+ "epoch": 13.93290734824281,
+ "grad_norm": 0.060302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0557,
+ "step": 4361
+ },
+ {
+ "epoch": 13.936102236421725,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.053,
+ "step": 4362
+ },
+ {
+ "epoch": 13.939297124600639,
+ "grad_norm": 0.059814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0554,
+ "step": 4363
+ },
+ {
+ "epoch": 13.942492012779553,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0441,
+ "step": 4364
+ },
+ {
+ "epoch": 13.945686900958467,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0507,
+ "step": 4365
+ },
+ {
+ "epoch": 13.94888178913738,
+ "grad_norm": 0.05908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0638,
+ "step": 4366
+ },
+ {
+ "epoch": 13.952076677316294,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0541,
+ "step": 4367
+ },
+ {
+ "epoch": 13.955271565495208,
+ "grad_norm": 0.0888671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0502,
+ "step": 4368
+ },
+ {
+ "epoch": 13.958466453674122,
+ "grad_norm": 0.055419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0559,
+ "step": 4369
+ },
+ {
+ "epoch": 13.961661341853034,
+ "grad_norm": 0.05126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0553,
+ "step": 4370
+ },
+ {
+ "epoch": 13.964856230031948,
+ "grad_norm": 0.0849609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0514,
+ "step": 4371
+ },
+ {
+ "epoch": 13.968051118210862,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0448,
+ "step": 4372
+ },
+ {
+ "epoch": 13.971246006389777,
+ "grad_norm": 0.142578125,
+ "learning_rate": 0.0005,
+ "loss": 1.05,
+ "step": 4373
+ },
+ {
+ "epoch": 13.97444089456869,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0603,
+ "step": 4374
+ },
+ {
+ "epoch": 13.977635782747605,
+ "grad_norm": 0.10595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0595,
+ "step": 4375
+ },
+ {
+ "epoch": 13.980830670926517,
+ "grad_norm": 0.0947265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0526,
+ "step": 4376
+ },
+ {
+ "epoch": 13.984025559105431,
+ "grad_norm": 0.058349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0452,
+ "step": 4377
+ },
+ {
+ "epoch": 13.987220447284345,
+ "grad_norm": 0.09423828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0457,
+ "step": 4378
+ },
+ {
+ "epoch": 13.99041533546326,
+ "grad_norm": 0.07861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0502,
+ "step": 4379
+ },
+ {
+ "epoch": 13.993610223642172,
+ "grad_norm": 0.09814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0577,
+ "step": 4380
+ },
+ {
+ "epoch": 13.996805111821086,
+ "grad_norm": 0.1494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0655,
+ "step": 4381
+ },
+ {
+ "epoch": 14.0,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.053,
+ "step": 4382
+ },
+ {
+ "epoch": 14.003194888178914,
+ "grad_norm": 0.171875,
+ "learning_rate": 0.0005,
+ "loss": 1.073,
+ "step": 4383
+ },
+ {
+ "epoch": 14.006389776357828,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0646,
+ "step": 4384
+ },
+ {
+ "epoch": 14.00958466453674,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0458,
+ "step": 4385
+ },
+ {
+ "epoch": 14.012779552715655,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0446,
+ "step": 4386
+ },
+ {
+ "epoch": 14.015974440894569,
+ "grad_norm": 0.10205078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0547,
+ "step": 4387
+ },
+ {
+ "epoch": 14.019169329073483,
+ "grad_norm": 0.057861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0581,
+ "step": 4388
+ },
+ {
+ "epoch": 14.022364217252397,
+ "grad_norm": 0.1259765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0647,
+ "step": 4389
+ },
+ {
+ "epoch": 14.02555910543131,
+ "grad_norm": 0.10546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0501,
+ "step": 4390
+ },
+ {
+ "epoch": 14.028753993610223,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0619,
+ "step": 4391
+ },
+ {
+ "epoch": 14.031948881789138,
+ "grad_norm": 0.11181640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0503,
+ "step": 4392
+ },
+ {
+ "epoch": 14.035143769968052,
+ "grad_norm": 0.05908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0402,
+ "step": 4393
+ },
+ {
+ "epoch": 14.038338658146964,
+ "grad_norm": 0.10400390625,
+ "learning_rate": 0.0005,
+ "loss": 1.058,
+ "step": 4394
+ },
+ {
+ "epoch": 14.041533546325878,
+ "grad_norm": 0.05322265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0531,
+ "step": 4395
+ },
+ {
+ "epoch": 14.044728434504792,
+ "grad_norm": 0.08984375,
+ "learning_rate": 0.0005,
+ "loss": 1.054,
+ "step": 4396
+ },
+ {
+ "epoch": 14.047923322683706,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0611,
+ "step": 4397
+ },
+ {
+ "epoch": 14.05111821086262,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0508,
+ "step": 4398
+ },
+ {
+ "epoch": 14.054313099041533,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0484,
+ "step": 4399
+ },
+ {
+ "epoch": 14.057507987220447,
+ "grad_norm": 0.111328125,
+ "learning_rate": 0.0005,
+ "loss": 1.062,
+ "step": 4400
+ },
+ {
+ "epoch": 14.060702875399361,
+ "grad_norm": 0.107421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0619,
+ "step": 4401
+ },
+ {
+ "epoch": 14.063897763578275,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0482,
+ "step": 4402
+ },
+ {
+ "epoch": 14.06709265175719,
+ "grad_norm": 0.10791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0506,
+ "step": 4403
+ },
+ {
+ "epoch": 14.070287539936102,
+ "grad_norm": 0.1494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0353,
+ "step": 4404
+ },
+ {
+ "epoch": 14.073482428115016,
+ "grad_norm": 0.09912109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0529,
+ "step": 4405
+ },
+ {
+ "epoch": 14.07667731629393,
+ "grad_norm": 0.1806640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0535,
+ "step": 4406
+ },
+ {
+ "epoch": 14.079872204472844,
+ "grad_norm": 0.10400390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0621,
+ "step": 4407
+ },
+ {
+ "epoch": 14.083067092651758,
+ "grad_norm": 0.150390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0626,
+ "step": 4408
+ },
+ {
+ "epoch": 14.08626198083067,
+ "grad_norm": 0.10693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0558,
+ "step": 4409
+ },
+ {
+ "epoch": 14.089456869009584,
+ "grad_norm": 0.11181640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0516,
+ "step": 4410
+ },
+ {
+ "epoch": 14.092651757188499,
+ "grad_norm": 0.1357421875,
+ "learning_rate": 0.0005,
+ "loss": 1.06,
+ "step": 4411
+ },
+ {
+ "epoch": 14.095846645367413,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0395,
+ "step": 4412
+ },
+ {
+ "epoch": 14.099041533546325,
+ "grad_norm": 0.1494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0642,
+ "step": 4413
+ },
+ {
+ "epoch": 14.10223642172524,
+ "grad_norm": 0.150390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0585,
+ "step": 4414
+ },
+ {
+ "epoch": 14.105431309904153,
+ "grad_norm": 0.1669921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0586,
+ "step": 4415
+ },
+ {
+ "epoch": 14.108626198083067,
+ "grad_norm": 0.1513671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0448,
+ "step": 4416
+ },
+ {
+ "epoch": 14.111821086261982,
+ "grad_norm": 0.171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0525,
+ "step": 4417
+ },
+ {
+ "epoch": 14.115015974440894,
+ "grad_norm": 0.17578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0591,
+ "step": 4418
+ },
+ {
+ "epoch": 14.118210862619808,
+ "grad_norm": 0.3046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0615,
+ "step": 4419
+ },
+ {
+ "epoch": 14.121405750798722,
+ "grad_norm": 0.1337890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0486,
+ "step": 4420
+ },
+ {
+ "epoch": 14.124600638977636,
+ "grad_norm": 0.486328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0615,
+ "step": 4421
+ },
+ {
+ "epoch": 14.12779552715655,
+ "grad_norm": 0.353515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0673,
+ "step": 4422
+ },
+ {
+ "epoch": 14.130990415335463,
+ "grad_norm": 0.4765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0674,
+ "step": 4423
+ },
+ {
+ "epoch": 14.134185303514377,
+ "grad_norm": 0.7265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0622,
+ "step": 4424
+ },
+ {
+ "epoch": 14.13738019169329,
+ "grad_norm": 1.5390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0771,
+ "step": 4425
+ },
+ {
+ "epoch": 14.140575079872205,
+ "grad_norm": 0.5703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0549,
+ "step": 4426
+ },
+ {
+ "epoch": 14.143769968051119,
+ "grad_norm": 0.5625,
+ "learning_rate": 0.0005,
+ "loss": 1.0602,
+ "step": 4427
+ },
+ {
+ "epoch": 14.146964856230031,
+ "grad_norm": 1.703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0888,
+ "step": 4428
+ },
+ {
+ "epoch": 14.150159744408946,
+ "grad_norm": 0.28515625,
+ "learning_rate": 0.0005,
+ "loss": 1.052,
+ "step": 4429
+ },
+ {
+ "epoch": 14.15335463258786,
+ "grad_norm": 1.2890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0697,
+ "step": 4430
+ },
+ {
+ "epoch": 14.156549520766774,
+ "grad_norm": 0.9921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0744,
+ "step": 4431
+ },
+ {
+ "epoch": 14.159744408945686,
+ "grad_norm": 0.333984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0424,
+ "step": 4432
+ },
+ {
+ "epoch": 14.1629392971246,
+ "grad_norm": 0.5859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0473,
+ "step": 4433
+ },
+ {
+ "epoch": 14.166134185303514,
+ "grad_norm": 0.4609375,
+ "learning_rate": 0.0005,
+ "loss": 1.05,
+ "step": 4434
+ },
+ {
+ "epoch": 14.169329073482428,
+ "grad_norm": 0.3125,
+ "learning_rate": 0.0005,
+ "loss": 1.0622,
+ "step": 4435
+ },
+ {
+ "epoch": 14.172523961661343,
+ "grad_norm": 0.5078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0618,
+ "step": 4436
+ },
+ {
+ "epoch": 14.175718849840255,
+ "grad_norm": 0.25390625,
+ "learning_rate": 0.0005,
+ "loss": 1.066,
+ "step": 4437
+ },
+ {
+ "epoch": 14.178913738019169,
+ "grad_norm": 0.337890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0658,
+ "step": 4438
+ },
+ {
+ "epoch": 14.182108626198083,
+ "grad_norm": 0.2001953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0593,
+ "step": 4439
+ },
+ {
+ "epoch": 14.185303514376997,
+ "grad_norm": 0.2236328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0569,
+ "step": 4440
+ },
+ {
+ "epoch": 14.188498402555911,
+ "grad_norm": 0.1259765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0452,
+ "step": 4441
+ },
+ {
+ "epoch": 14.191693290734824,
+ "grad_norm": 0.2158203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0605,
+ "step": 4442
+ },
+ {
+ "epoch": 14.194888178913738,
+ "grad_norm": 0.1171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0566,
+ "step": 4443
+ },
+ {
+ "epoch": 14.198083067092652,
+ "grad_norm": 0.1259765625,
+ "learning_rate": 0.0005,
+ "loss": 1.058,
+ "step": 4444
+ },
+ {
+ "epoch": 14.201277955271566,
+ "grad_norm": 0.0947265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0368,
+ "step": 4445
+ },
+ {
+ "epoch": 14.204472843450478,
+ "grad_norm": 0.11181640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0482,
+ "step": 4446
+ },
+ {
+ "epoch": 14.207667731629392,
+ "grad_norm": 0.09033203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0505,
+ "step": 4447
+ },
+ {
+ "epoch": 14.210862619808307,
+ "grad_norm": 0.1083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0716,
+ "step": 4448
+ },
+ {
+ "epoch": 14.21405750798722,
+ "grad_norm": 0.08056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0508,
+ "step": 4449
+ },
+ {
+ "epoch": 14.217252396166135,
+ "grad_norm": 0.1064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0623,
+ "step": 4450
+ },
+ {
+ "epoch": 14.220447284345047,
+ "grad_norm": 0.07861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0465,
+ "step": 4451
+ },
+ {
+ "epoch": 14.223642172523961,
+ "grad_norm": 0.0986328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0609,
+ "step": 4452
+ },
+ {
+ "epoch": 14.226837060702875,
+ "grad_norm": 0.060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0574,
+ "step": 4453
+ },
+ {
+ "epoch": 14.23003194888179,
+ "grad_norm": 0.06103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0579,
+ "step": 4454
+ },
+ {
+ "epoch": 14.233226837060704,
+ "grad_norm": 0.0849609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0607,
+ "step": 4455
+ },
+ {
+ "epoch": 14.236421725239616,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0473,
+ "step": 4456
+ },
+ {
+ "epoch": 14.23961661341853,
+ "grad_norm": 0.125,
+ "learning_rate": 0.0005,
+ "loss": 1.048,
+ "step": 4457
+ },
+ {
+ "epoch": 14.242811501597444,
+ "grad_norm": 0.095703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0532,
+ "step": 4458
+ },
+ {
+ "epoch": 14.246006389776358,
+ "grad_norm": 0.0966796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0497,
+ "step": 4459
+ },
+ {
+ "epoch": 14.249201277955272,
+ "grad_norm": 0.11865234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0518,
+ "step": 4460
+ },
+ {
+ "epoch": 14.252396166134185,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0623,
+ "step": 4461
+ },
+ {
+ "epoch": 14.255591054313099,
+ "grad_norm": 0.12109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0384,
+ "step": 4462
+ },
+ {
+ "epoch": 14.258785942492013,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0573,
+ "step": 4463
+ },
+ {
+ "epoch": 14.261980830670927,
+ "grad_norm": 0.09375,
+ "learning_rate": 0.0005,
+ "loss": 1.0417,
+ "step": 4464
+ },
+ {
+ "epoch": 14.26517571884984,
+ "grad_norm": 0.080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0553,
+ "step": 4465
+ },
+ {
+ "epoch": 14.268370607028753,
+ "grad_norm": 0.1240234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0538,
+ "step": 4466
+ },
+ {
+ "epoch": 14.271565495207668,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0531,
+ "step": 4467
+ },
+ {
+ "epoch": 14.274760383386582,
+ "grad_norm": 0.0859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0426,
+ "step": 4468
+ },
+ {
+ "epoch": 14.277955271565496,
+ "grad_norm": 0.0947265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0596,
+ "step": 4469
+ },
+ {
+ "epoch": 14.281150159744408,
+ "grad_norm": 0.10693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0573,
+ "step": 4470
+ },
+ {
+ "epoch": 14.284345047923322,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.067,
+ "step": 4471
+ },
+ {
+ "epoch": 14.287539936102236,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.042,
+ "step": 4472
+ },
+ {
+ "epoch": 14.29073482428115,
+ "grad_norm": 0.10595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0577,
+ "step": 4473
+ },
+ {
+ "epoch": 14.293929712460065,
+ "grad_norm": 0.0810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0531,
+ "step": 4474
+ },
+ {
+ "epoch": 14.297124600638977,
+ "grad_norm": 0.091796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0619,
+ "step": 4475
+ },
+ {
+ "epoch": 14.300319488817891,
+ "grad_norm": 0.09033203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0461,
+ "step": 4476
+ },
+ {
+ "epoch": 14.303514376996805,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0453,
+ "step": 4477
+ },
+ {
+ "epoch": 14.30670926517572,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0658,
+ "step": 4478
+ },
+ {
+ "epoch": 14.309904153354633,
+ "grad_norm": 0.05859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0482,
+ "step": 4479
+ },
+ {
+ "epoch": 14.313099041533546,
+ "grad_norm": 0.060791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0609,
+ "step": 4480
+ },
+ {
+ "epoch": 14.31629392971246,
+ "grad_norm": 0.06201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0671,
+ "step": 4481
+ },
+ {
+ "epoch": 14.319488817891374,
+ "grad_norm": 0.058349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0632,
+ "step": 4482
+ },
+ {
+ "epoch": 14.322683706070288,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0473,
+ "step": 4483
+ },
+ {
+ "epoch": 14.3258785942492,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0549,
+ "step": 4484
+ },
+ {
+ "epoch": 14.329073482428115,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0483,
+ "step": 4485
+ },
+ {
+ "epoch": 14.332268370607029,
+ "grad_norm": 0.0595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0572,
+ "step": 4486
+ },
+ {
+ "epoch": 14.335463258785943,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0408,
+ "step": 4487
+ },
+ {
+ "epoch": 14.338658146964857,
+ "grad_norm": 0.049560546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0621,
+ "step": 4488
+ },
+ {
+ "epoch": 14.34185303514377,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0621,
+ "step": 4489
+ },
+ {
+ "epoch": 14.345047923322683,
+ "grad_norm": 0.05419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0662,
+ "step": 4490
+ },
+ {
+ "epoch": 14.348242811501597,
+ "grad_norm": 0.0859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0525,
+ "step": 4491
+ },
+ {
+ "epoch": 14.351437699680512,
+ "grad_norm": 0.058349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0487,
+ "step": 4492
+ },
+ {
+ "epoch": 14.354632587859426,
+ "grad_norm": 0.054931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0485,
+ "step": 4493
+ },
+ {
+ "epoch": 14.357827476038338,
+ "grad_norm": 0.06201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0482,
+ "step": 4494
+ },
+ {
+ "epoch": 14.361022364217252,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0523,
+ "step": 4495
+ },
+ {
+ "epoch": 14.364217252396166,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.051,
+ "step": 4496
+ },
+ {
+ "epoch": 14.36741214057508,
+ "grad_norm": 0.058837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0567,
+ "step": 4497
+ },
+ {
+ "epoch": 14.370607028753994,
+ "grad_norm": 0.0576171875,
+ "learning_rate": 0.0005,
+ "loss": 1.05,
+ "step": 4498
+ },
+ {
+ "epoch": 14.373801916932907,
+ "grad_norm": 0.05615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0512,
+ "step": 4499
+ },
+ {
+ "epoch": 14.37699680511182,
+ "grad_norm": 0.058837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0516,
+ "step": 4500
+ },
+ {
+ "epoch": 14.380191693290735,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0607,
+ "step": 4501
+ },
+ {
+ "epoch": 14.383386581469649,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0634,
+ "step": 4502
+ },
+ {
+ "epoch": 14.386581469648561,
+ "grad_norm": 0.1005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0575,
+ "step": 4503
+ },
+ {
+ "epoch": 14.389776357827476,
+ "grad_norm": 0.08056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0535,
+ "step": 4504
+ },
+ {
+ "epoch": 14.39297124600639,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0555,
+ "step": 4505
+ },
+ {
+ "epoch": 14.396166134185304,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0507,
+ "step": 4506
+ },
+ {
+ "epoch": 14.399361022364218,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.05,
+ "step": 4507
+ },
+ {
+ "epoch": 14.40255591054313,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0491,
+ "step": 4508
+ },
+ {
+ "epoch": 14.405750798722044,
+ "grad_norm": 0.0556640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0423,
+ "step": 4509
+ },
+ {
+ "epoch": 14.408945686900958,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0475,
+ "step": 4510
+ },
+ {
+ "epoch": 14.412140575079873,
+ "grad_norm": 0.05810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0597,
+ "step": 4511
+ },
+ {
+ "epoch": 14.415335463258787,
+ "grad_norm": 0.052734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0541,
+ "step": 4512
+ },
+ {
+ "epoch": 14.418530351437699,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0499,
+ "step": 4513
+ },
+ {
+ "epoch": 14.421725239616613,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0538,
+ "step": 4514
+ },
+ {
+ "epoch": 14.424920127795527,
+ "grad_norm": 0.05517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0536,
+ "step": 4515
+ },
+ {
+ "epoch": 14.428115015974441,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0416,
+ "step": 4516
+ },
+ {
+ "epoch": 14.431309904153355,
+ "grad_norm": 0.09814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0644,
+ "step": 4517
+ },
+ {
+ "epoch": 14.434504792332268,
+ "grad_norm": 0.047119140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0688,
+ "step": 4518
+ },
+ {
+ "epoch": 14.437699680511182,
+ "grad_norm": 0.08740234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0503,
+ "step": 4519
+ },
+ {
+ "epoch": 14.440894568690096,
+ "grad_norm": 0.08984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0567,
+ "step": 4520
+ },
+ {
+ "epoch": 14.44408945686901,
+ "grad_norm": 0.057373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0582,
+ "step": 4521
+ },
+ {
+ "epoch": 14.447284345047922,
+ "grad_norm": 0.15625,
+ "learning_rate": 0.0005,
+ "loss": 1.0486,
+ "step": 4522
+ },
+ {
+ "epoch": 14.450479233226837,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0463,
+ "step": 4523
+ },
+ {
+ "epoch": 14.45367412140575,
+ "grad_norm": 0.10888671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0515,
+ "step": 4524
+ },
+ {
+ "epoch": 14.456869009584665,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0471,
+ "step": 4525
+ },
+ {
+ "epoch": 14.460063897763579,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0549,
+ "step": 4526
+ },
+ {
+ "epoch": 14.463258785942491,
+ "grad_norm": 0.0849609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0562,
+ "step": 4527
+ },
+ {
+ "epoch": 14.466453674121405,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0498,
+ "step": 4528
+ },
+ {
+ "epoch": 14.46964856230032,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0551,
+ "step": 4529
+ },
+ {
+ "epoch": 14.472843450479234,
+ "grad_norm": 0.09375,
+ "learning_rate": 0.0005,
+ "loss": 1.0638,
+ "step": 4530
+ },
+ {
+ "epoch": 14.476038338658148,
+ "grad_norm": 0.109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0566,
+ "step": 4531
+ },
+ {
+ "epoch": 14.47923322683706,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0523,
+ "step": 4532
+ },
+ {
+ "epoch": 14.482428115015974,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0548,
+ "step": 4533
+ },
+ {
+ "epoch": 14.485623003194888,
+ "grad_norm": 0.061767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0547,
+ "step": 4534
+ },
+ {
+ "epoch": 14.488817891373802,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0588,
+ "step": 4535
+ },
+ {
+ "epoch": 14.492012779552716,
+ "grad_norm": 0.08984375,
+ "learning_rate": 0.0005,
+ "loss": 1.064,
+ "step": 4536
+ },
+ {
+ "epoch": 14.495207667731629,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0426,
+ "step": 4537
+ },
+ {
+ "epoch": 14.498402555910543,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0449,
+ "step": 4538
+ },
+ {
+ "epoch": 14.501597444089457,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0566,
+ "step": 4539
+ },
+ {
+ "epoch": 14.504792332268371,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0453,
+ "step": 4540
+ },
+ {
+ "epoch": 14.507987220447284,
+ "grad_norm": 0.05419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0587,
+ "step": 4541
+ },
+ {
+ "epoch": 14.511182108626198,
+ "grad_norm": 0.0869140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0617,
+ "step": 4542
+ },
+ {
+ "epoch": 14.514376996805112,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0458,
+ "step": 4543
+ },
+ {
+ "epoch": 14.517571884984026,
+ "grad_norm": 0.10107421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0511,
+ "step": 4544
+ },
+ {
+ "epoch": 14.52076677316294,
+ "grad_norm": 0.1162109375,
+ "learning_rate": 0.0005,
+ "loss": 1.044,
+ "step": 4545
+ },
+ {
+ "epoch": 14.523961661341852,
+ "grad_norm": 0.119140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0544,
+ "step": 4546
+ },
+ {
+ "epoch": 14.527156549520766,
+ "grad_norm": 0.10498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0547,
+ "step": 4547
+ },
+ {
+ "epoch": 14.53035143769968,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0456,
+ "step": 4548
+ },
+ {
+ "epoch": 14.533546325878595,
+ "grad_norm": 0.1337890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0457,
+ "step": 4549
+ },
+ {
+ "epoch": 14.536741214057509,
+ "grad_norm": 0.08251953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0654,
+ "step": 4550
+ },
+ {
+ "epoch": 14.539936102236421,
+ "grad_norm": 0.061767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0544,
+ "step": 4551
+ },
+ {
+ "epoch": 14.543130990415335,
+ "grad_norm": 0.134765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0484,
+ "step": 4552
+ },
+ {
+ "epoch": 14.54632587859425,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0519,
+ "step": 4553
+ },
+ {
+ "epoch": 14.549520766773163,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0477,
+ "step": 4554
+ },
+ {
+ "epoch": 14.552715654952078,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0417,
+ "step": 4555
+ },
+ {
+ "epoch": 14.55591054313099,
+ "grad_norm": 0.056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0579,
+ "step": 4556
+ },
+ {
+ "epoch": 14.559105431309904,
+ "grad_norm": 0.060791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0554,
+ "step": 4557
+ },
+ {
+ "epoch": 14.562300319488818,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0553,
+ "step": 4558
+ },
+ {
+ "epoch": 14.565495207667732,
+ "grad_norm": 0.05810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0546,
+ "step": 4559
+ },
+ {
+ "epoch": 14.568690095846645,
+ "grad_norm": 0.059814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0437,
+ "step": 4560
+ },
+ {
+ "epoch": 14.571884984025559,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.055,
+ "step": 4561
+ },
+ {
+ "epoch": 14.575079872204473,
+ "grad_norm": 0.05322265625,
+ "learning_rate": 0.0005,
+ "loss": 1.053,
+ "step": 4562
+ },
+ {
+ "epoch": 14.578274760383387,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0385,
+ "step": 4563
+ },
+ {
+ "epoch": 14.581469648562301,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0632,
+ "step": 4564
+ },
+ {
+ "epoch": 14.584664536741213,
+ "grad_norm": 0.09423828125,
+ "learning_rate": 0.0005,
+ "loss": 1.065,
+ "step": 4565
+ },
+ {
+ "epoch": 14.587859424920127,
+ "grad_norm": 0.05615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0542,
+ "step": 4566
+ },
+ {
+ "epoch": 14.591054313099042,
+ "grad_norm": 0.0556640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0544,
+ "step": 4567
+ },
+ {
+ "epoch": 14.594249201277956,
+ "grad_norm": 0.057861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0476,
+ "step": 4568
+ },
+ {
+ "epoch": 14.59744408945687,
+ "grad_norm": 0.05712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0537,
+ "step": 4569
+ },
+ {
+ "epoch": 14.600638977635782,
+ "grad_norm": 0.056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0547,
+ "step": 4570
+ },
+ {
+ "epoch": 14.603833865814696,
+ "grad_norm": 0.051513671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0558,
+ "step": 4571
+ },
+ {
+ "epoch": 14.60702875399361,
+ "grad_norm": 0.057861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0468,
+ "step": 4572
+ },
+ {
+ "epoch": 14.610223642172524,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0458,
+ "step": 4573
+ },
+ {
+ "epoch": 14.613418530351439,
+ "grad_norm": 0.0849609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0539,
+ "step": 4574
+ },
+ {
+ "epoch": 14.616613418530351,
+ "grad_norm": 0.056396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0513,
+ "step": 4575
+ },
+ {
+ "epoch": 14.619808306709265,
+ "grad_norm": 0.058349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0444,
+ "step": 4576
+ },
+ {
+ "epoch": 14.62300319488818,
+ "grad_norm": 0.09765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0502,
+ "step": 4577
+ },
+ {
+ "epoch": 14.626198083067093,
+ "grad_norm": 0.0595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0389,
+ "step": 4578
+ },
+ {
+ "epoch": 14.629392971246006,
+ "grad_norm": 0.09033203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0547,
+ "step": 4579
+ },
+ {
+ "epoch": 14.63258785942492,
+ "grad_norm": 0.05419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0619,
+ "step": 4580
+ },
+ {
+ "epoch": 14.635782747603834,
+ "grad_norm": 0.0546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0482,
+ "step": 4581
+ },
+ {
+ "epoch": 14.638977635782748,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0571,
+ "step": 4582
+ },
+ {
+ "epoch": 14.642172523961662,
+ "grad_norm": 0.05859375,
+ "learning_rate": 0.0005,
+ "loss": 1.046,
+ "step": 4583
+ },
+ {
+ "epoch": 14.645367412140574,
+ "grad_norm": 0.056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0648,
+ "step": 4584
+ },
+ {
+ "epoch": 14.648562300319488,
+ "grad_norm": 0.054931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0648,
+ "step": 4585
+ },
+ {
+ "epoch": 14.651757188498403,
+ "grad_norm": 0.05859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0518,
+ "step": 4586
+ },
+ {
+ "epoch": 14.654952076677317,
+ "grad_norm": 0.055419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0585,
+ "step": 4587
+ },
+ {
+ "epoch": 14.65814696485623,
+ "grad_norm": 0.0576171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0523,
+ "step": 4588
+ },
+ {
+ "epoch": 14.661341853035143,
+ "grad_norm": 0.0869140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0482,
+ "step": 4589
+ },
+ {
+ "epoch": 14.664536741214057,
+ "grad_norm": 0.055908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0475,
+ "step": 4590
+ },
+ {
+ "epoch": 14.667731629392971,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0504,
+ "step": 4591
+ },
+ {
+ "epoch": 14.670926517571885,
+ "grad_norm": 0.056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0499,
+ "step": 4592
+ },
+ {
+ "epoch": 14.6741214057508,
+ "grad_norm": 0.09912109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0528,
+ "step": 4593
+ },
+ {
+ "epoch": 14.677316293929712,
+ "grad_norm": 0.05615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.052,
+ "step": 4594
+ },
+ {
+ "epoch": 14.680511182108626,
+ "grad_norm": 0.059814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.037,
+ "step": 4595
+ },
+ {
+ "epoch": 14.68370607028754,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0426,
+ "step": 4596
+ },
+ {
+ "epoch": 14.686900958466454,
+ "grad_norm": 0.09326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0428,
+ "step": 4597
+ },
+ {
+ "epoch": 14.690095846645367,
+ "grad_norm": 0.056396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0481,
+ "step": 4598
+ },
+ {
+ "epoch": 14.69329073482428,
+ "grad_norm": 0.0517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0586,
+ "step": 4599
+ },
+ {
+ "epoch": 14.696485623003195,
+ "grad_norm": 0.060791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0572,
+ "step": 4600
+ },
+ {
+ "epoch": 14.699680511182109,
+ "grad_norm": 0.051513671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0545,
+ "step": 4601
+ },
+ {
+ "epoch": 14.702875399361023,
+ "grad_norm": 0.052490234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0565,
+ "step": 4602
+ },
+ {
+ "epoch": 14.706070287539935,
+ "grad_norm": 0.052490234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0612,
+ "step": 4603
+ },
+ {
+ "epoch": 14.70926517571885,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0617,
+ "step": 4604
+ },
+ {
+ "epoch": 14.712460063897764,
+ "grad_norm": 0.0595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0569,
+ "step": 4605
+ },
+ {
+ "epoch": 14.715654952076678,
+ "grad_norm": 0.0947265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0509,
+ "step": 4606
+ },
+ {
+ "epoch": 14.718849840255592,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0684,
+ "step": 4607
+ },
+ {
+ "epoch": 14.722044728434504,
+ "grad_norm": 0.091796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0562,
+ "step": 4608
+ },
+ {
+ "epoch": 14.725239616613418,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0556,
+ "step": 4609
+ },
+ {
+ "epoch": 14.728434504792332,
+ "grad_norm": 0.060302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0547,
+ "step": 4610
+ },
+ {
+ "epoch": 14.731629392971247,
+ "grad_norm": 0.0966796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0434,
+ "step": 4611
+ },
+ {
+ "epoch": 14.73482428115016,
+ "grad_norm": 0.055908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0449,
+ "step": 4612
+ },
+ {
+ "epoch": 14.738019169329073,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0489,
+ "step": 4613
+ },
+ {
+ "epoch": 14.741214057507987,
+ "grad_norm": 0.059814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0626,
+ "step": 4614
+ },
+ {
+ "epoch": 14.744408945686901,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0505,
+ "step": 4615
+ },
+ {
+ "epoch": 14.747603833865815,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0509,
+ "step": 4616
+ },
+ {
+ "epoch": 14.750798722044728,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0471,
+ "step": 4617
+ },
+ {
+ "epoch": 14.753993610223642,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0393,
+ "step": 4618
+ },
+ {
+ "epoch": 14.757188498402556,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0514,
+ "step": 4619
+ },
+ {
+ "epoch": 14.76038338658147,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0485,
+ "step": 4620
+ },
+ {
+ "epoch": 14.763578274760384,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0398,
+ "step": 4621
+ },
+ {
+ "epoch": 14.766773162939296,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0489,
+ "step": 4622
+ },
+ {
+ "epoch": 14.76996805111821,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0533,
+ "step": 4623
+ },
+ {
+ "epoch": 14.773162939297125,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0514,
+ "step": 4624
+ },
+ {
+ "epoch": 14.776357827476039,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0569,
+ "step": 4625
+ },
+ {
+ "epoch": 14.779552715654953,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0494,
+ "step": 4626
+ },
+ {
+ "epoch": 14.782747603833865,
+ "grad_norm": 0.09033203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0499,
+ "step": 4627
+ },
+ {
+ "epoch": 14.78594249201278,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0453,
+ "step": 4628
+ },
+ {
+ "epoch": 14.789137380191693,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0515,
+ "step": 4629
+ },
+ {
+ "epoch": 14.792332268370608,
+ "grad_norm": 0.0849609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0528,
+ "step": 4630
+ },
+ {
+ "epoch": 14.795527156549522,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0672,
+ "step": 4631
+ },
+ {
+ "epoch": 14.798722044728434,
+ "grad_norm": 0.05615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0641,
+ "step": 4632
+ },
+ {
+ "epoch": 14.801916932907348,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0546,
+ "step": 4633
+ },
+ {
+ "epoch": 14.805111821086262,
+ "grad_norm": 0.06103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0489,
+ "step": 4634
+ },
+ {
+ "epoch": 14.808306709265176,
+ "grad_norm": 0.0849609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0463,
+ "step": 4635
+ },
+ {
+ "epoch": 14.811501597444089,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0558,
+ "step": 4636
+ },
+ {
+ "epoch": 14.814696485623003,
+ "grad_norm": 0.11181640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0483,
+ "step": 4637
+ },
+ {
+ "epoch": 14.817891373801917,
+ "grad_norm": 0.061279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0479,
+ "step": 4638
+ },
+ {
+ "epoch": 14.821086261980831,
+ "grad_norm": 0.111328125,
+ "learning_rate": 0.0005,
+ "loss": 1.052,
+ "step": 4639
+ },
+ {
+ "epoch": 14.824281150159745,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0501,
+ "step": 4640
+ },
+ {
+ "epoch": 14.827476038338657,
+ "grad_norm": 0.09130859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0516,
+ "step": 4641
+ },
+ {
+ "epoch": 14.830670926517572,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0583,
+ "step": 4642
+ },
+ {
+ "epoch": 14.833865814696486,
+ "grad_norm": 0.058349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0626,
+ "step": 4643
+ },
+ {
+ "epoch": 14.8370607028754,
+ "grad_norm": 0.057373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0508,
+ "step": 4644
+ },
+ {
+ "epoch": 14.840255591054314,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0569,
+ "step": 4645
+ },
+ {
+ "epoch": 14.843450479233226,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0392,
+ "step": 4646
+ },
+ {
+ "epoch": 14.84664536741214,
+ "grad_norm": 0.056884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0702,
+ "step": 4647
+ },
+ {
+ "epoch": 14.849840255591054,
+ "grad_norm": 0.0615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0459,
+ "step": 4648
+ },
+ {
+ "epoch": 14.853035143769969,
+ "grad_norm": 0.08154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0489,
+ "step": 4649
+ },
+ {
+ "epoch": 14.856230031948883,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0495,
+ "step": 4650
+ },
+ {
+ "epoch": 14.859424920127795,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0541,
+ "step": 4651
+ },
+ {
+ "epoch": 14.86261980830671,
+ "grad_norm": 0.0908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.056,
+ "step": 4652
+ },
+ {
+ "epoch": 14.865814696485623,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0585,
+ "step": 4653
+ },
+ {
+ "epoch": 14.869009584664537,
+ "grad_norm": 0.09814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0503,
+ "step": 4654
+ },
+ {
+ "epoch": 14.87220447284345,
+ "grad_norm": 0.057373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0545,
+ "step": 4655
+ },
+ {
+ "epoch": 14.875399361022364,
+ "grad_norm": 0.10009765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0585,
+ "step": 4656
+ },
+ {
+ "epoch": 14.878594249201278,
+ "grad_norm": 0.0615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0523,
+ "step": 4657
+ },
+ {
+ "epoch": 14.881789137380192,
+ "grad_norm": 0.11376953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0499,
+ "step": 4658
+ },
+ {
+ "epoch": 14.884984025559106,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0608,
+ "step": 4659
+ },
+ {
+ "epoch": 14.888178913738018,
+ "grad_norm": 0.1455078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0609,
+ "step": 4660
+ },
+ {
+ "epoch": 14.891373801916933,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0495,
+ "step": 4661
+ },
+ {
+ "epoch": 14.894568690095847,
+ "grad_norm": 0.1240234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0462,
+ "step": 4662
+ },
+ {
+ "epoch": 14.89776357827476,
+ "grad_norm": 0.09521484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0579,
+ "step": 4663
+ },
+ {
+ "epoch": 14.900958466453675,
+ "grad_norm": 0.1533203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0425,
+ "step": 4664
+ },
+ {
+ "epoch": 14.904153354632587,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0503,
+ "step": 4665
+ },
+ {
+ "epoch": 14.907348242811501,
+ "grad_norm": 0.12158203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0498,
+ "step": 4666
+ },
+ {
+ "epoch": 14.910543130990416,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0541,
+ "step": 4667
+ },
+ {
+ "epoch": 14.91373801916933,
+ "grad_norm": 0.11474609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0509,
+ "step": 4668
+ },
+ {
+ "epoch": 14.916932907348244,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0544,
+ "step": 4669
+ },
+ {
+ "epoch": 14.920127795527156,
+ "grad_norm": 0.11865234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0461,
+ "step": 4670
+ },
+ {
+ "epoch": 14.92332268370607,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0528,
+ "step": 4671
+ },
+ {
+ "epoch": 14.926517571884984,
+ "grad_norm": 0.115234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0529,
+ "step": 4672
+ },
+ {
+ "epoch": 14.929712460063898,
+ "grad_norm": 0.07861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0486,
+ "step": 4673
+ },
+ {
+ "epoch": 14.93290734824281,
+ "grad_norm": 0.1083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0519,
+ "step": 4674
+ },
+ {
+ "epoch": 14.936102236421725,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0448,
+ "step": 4675
+ },
+ {
+ "epoch": 14.939297124600639,
+ "grad_norm": 0.09130859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0451,
+ "step": 4676
+ },
+ {
+ "epoch": 14.942492012779553,
+ "grad_norm": 0.0517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0499,
+ "step": 4677
+ },
+ {
+ "epoch": 14.945686900958467,
+ "grad_norm": 0.08984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0473,
+ "step": 4678
+ },
+ {
+ "epoch": 14.94888178913738,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0599,
+ "step": 4679
+ },
+ {
+ "epoch": 14.952076677316294,
+ "grad_norm": 0.1279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0495,
+ "step": 4680
+ },
+ {
+ "epoch": 14.955271565495208,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0574,
+ "step": 4681
+ },
+ {
+ "epoch": 14.958466453674122,
+ "grad_norm": 0.1513671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0433,
+ "step": 4682
+ },
+ {
+ "epoch": 14.961661341853034,
+ "grad_norm": 0.060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0372,
+ "step": 4683
+ },
+ {
+ "epoch": 14.964856230031948,
+ "grad_norm": 0.2060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0566,
+ "step": 4684
+ },
+ {
+ "epoch": 14.968051118210862,
+ "grad_norm": 0.08984375,
+ "learning_rate": 0.0005,
+ "loss": 1.051,
+ "step": 4685
+ },
+ {
+ "epoch": 14.971246006389777,
+ "grad_norm": 0.138671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0469,
+ "step": 4686
+ },
+ {
+ "epoch": 14.97444089456869,
+ "grad_norm": 0.083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0553,
+ "step": 4687
+ },
+ {
+ "epoch": 14.977635782747605,
+ "grad_norm": 0.11376953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0423,
+ "step": 4688
+ },
+ {
+ "epoch": 14.980830670926517,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0631,
+ "step": 4689
+ },
+ {
+ "epoch": 14.984025559105431,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0483,
+ "step": 4690
+ },
+ {
+ "epoch": 14.987220447284345,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0608,
+ "step": 4691
+ },
+ {
+ "epoch": 14.99041533546326,
+ "grad_norm": 0.08544921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0503,
+ "step": 4692
+ },
+ {
+ "epoch": 14.993610223642172,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0545,
+ "step": 4693
+ },
+ {
+ "epoch": 14.996805111821086,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0554,
+ "step": 4694
+ },
+ {
+ "epoch": 15.0,
+ "grad_norm": 0.06201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0649,
+ "step": 4695
+ },
+ {
+ "epoch": 15.003194888178914,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0447,
+ "step": 4696
+ },
+ {
+ "epoch": 15.006389776357828,
+ "grad_norm": 0.054931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0592,
+ "step": 4697
+ },
+ {
+ "epoch": 15.00958466453674,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0478,
+ "step": 4698
+ },
+ {
+ "epoch": 15.012779552715655,
+ "grad_norm": 0.0576171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0526,
+ "step": 4699
+ },
+ {
+ "epoch": 15.015974440894569,
+ "grad_norm": 0.0595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0478,
+ "step": 4700
+ },
+ {
+ "epoch": 15.019169329073483,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0534,
+ "step": 4701
+ },
+ {
+ "epoch": 15.022364217252397,
+ "grad_norm": 0.061767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0429,
+ "step": 4702
+ },
+ {
+ "epoch": 15.02555910543131,
+ "grad_norm": 0.0537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0355,
+ "step": 4703
+ },
+ {
+ "epoch": 15.028753993610223,
+ "grad_norm": 0.0615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0515,
+ "step": 4704
+ },
+ {
+ "epoch": 15.031948881789138,
+ "grad_norm": 0.06005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0477,
+ "step": 4705
+ },
+ {
+ "epoch": 15.035143769968052,
+ "grad_norm": 0.05908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0568,
+ "step": 4706
+ },
+ {
+ "epoch": 15.038338658146964,
+ "grad_norm": 0.057373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0599,
+ "step": 4707
+ },
+ {
+ "epoch": 15.041533546325878,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0403,
+ "step": 4708
+ },
+ {
+ "epoch": 15.044728434504792,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0445,
+ "step": 4709
+ },
+ {
+ "epoch": 15.047923322683706,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.057,
+ "step": 4710
+ },
+ {
+ "epoch": 15.05111821086262,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0607,
+ "step": 4711
+ },
+ {
+ "epoch": 15.054313099041533,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0461,
+ "step": 4712
+ },
+ {
+ "epoch": 15.057507987220447,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.063,
+ "step": 4713
+ },
+ {
+ "epoch": 15.060702875399361,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0628,
+ "step": 4714
+ },
+ {
+ "epoch": 15.063897763578275,
+ "grad_norm": 0.058349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.048,
+ "step": 4715
+ },
+ {
+ "epoch": 15.06709265175719,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0572,
+ "step": 4716
+ },
+ {
+ "epoch": 15.070287539936102,
+ "grad_norm": 0.0537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.056,
+ "step": 4717
+ },
+ {
+ "epoch": 15.073482428115016,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0493,
+ "step": 4718
+ },
+ {
+ "epoch": 15.07667731629393,
+ "grad_norm": 0.0556640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0565,
+ "step": 4719
+ },
+ {
+ "epoch": 15.079872204472844,
+ "grad_norm": 0.05078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0462,
+ "step": 4720
+ },
+ {
+ "epoch": 15.083067092651758,
+ "grad_norm": 0.056884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0416,
+ "step": 4721
+ },
+ {
+ "epoch": 15.08626198083067,
+ "grad_norm": 0.05419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0632,
+ "step": 4722
+ },
+ {
+ "epoch": 15.089456869009584,
+ "grad_norm": 0.08447265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0353,
+ "step": 4723
+ },
+ {
+ "epoch": 15.092651757188499,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0522,
+ "step": 4724
+ },
+ {
+ "epoch": 15.095846645367413,
+ "grad_norm": 0.05078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0521,
+ "step": 4725
+ },
+ {
+ "epoch": 15.099041533546325,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0483,
+ "step": 4726
+ },
+ {
+ "epoch": 15.10223642172524,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0611,
+ "step": 4727
+ },
+ {
+ "epoch": 15.105431309904153,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0588,
+ "step": 4728
+ },
+ {
+ "epoch": 15.108626198083067,
+ "grad_norm": 0.08837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0447,
+ "step": 4729
+ },
+ {
+ "epoch": 15.111821086261982,
+ "grad_norm": 0.11279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0396,
+ "step": 4730
+ },
+ {
+ "epoch": 15.115015974440894,
+ "grad_norm": 0.057861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0548,
+ "step": 4731
+ },
+ {
+ "epoch": 15.118210862619808,
+ "grad_norm": 0.05322265625,
+ "learning_rate": 0.0005,
+ "loss": 1.063,
+ "step": 4732
+ },
+ {
+ "epoch": 15.121405750798722,
+ "grad_norm": 0.09228515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0474,
+ "step": 4733
+ },
+ {
+ "epoch": 15.124600638977636,
+ "grad_norm": 0.099609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0492,
+ "step": 4734
+ },
+ {
+ "epoch": 15.12779552715655,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0535,
+ "step": 4735
+ },
+ {
+ "epoch": 15.130990415335463,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0445,
+ "step": 4736
+ },
+ {
+ "epoch": 15.134185303514377,
+ "grad_norm": 0.10546875,
+ "learning_rate": 0.0005,
+ "loss": 1.059,
+ "step": 4737
+ },
+ {
+ "epoch": 15.13738019169329,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0437,
+ "step": 4738
+ },
+ {
+ "epoch": 15.140575079872205,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.046,
+ "step": 4739
+ },
+ {
+ "epoch": 15.143769968051119,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0558,
+ "step": 4740
+ },
+ {
+ "epoch": 15.146964856230031,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.058,
+ "step": 4741
+ },
+ {
+ "epoch": 15.150159744408946,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0582,
+ "step": 4742
+ },
+ {
+ "epoch": 15.15335463258786,
+ "grad_norm": 0.056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0586,
+ "step": 4743
+ },
+ {
+ "epoch": 15.156549520766774,
+ "grad_norm": 0.08984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0642,
+ "step": 4744
+ },
+ {
+ "epoch": 15.159744408945686,
+ "grad_norm": 0.09375,
+ "learning_rate": 0.0005,
+ "loss": 1.0571,
+ "step": 4745
+ },
+ {
+ "epoch": 15.1629392971246,
+ "grad_norm": 0.057373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0543,
+ "step": 4746
+ },
+ {
+ "epoch": 15.166134185303514,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0391,
+ "step": 4747
+ },
+ {
+ "epoch": 15.169329073482428,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0609,
+ "step": 4748
+ },
+ {
+ "epoch": 15.172523961661343,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.056,
+ "step": 4749
+ },
+ {
+ "epoch": 15.175718849840255,
+ "grad_norm": 0.09033203125,
+ "learning_rate": 0.0005,
+ "loss": 1.063,
+ "step": 4750
+ },
+ {
+ "epoch": 15.178913738019169,
+ "grad_norm": 0.11376953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0561,
+ "step": 4751
+ },
+ {
+ "epoch": 15.182108626198083,
+ "grad_norm": 0.060302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0622,
+ "step": 4752
+ },
+ {
+ "epoch": 15.185303514376997,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0559,
+ "step": 4753
+ },
+ {
+ "epoch": 15.188498402555911,
+ "grad_norm": 0.055419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0499,
+ "step": 4754
+ },
+ {
+ "epoch": 15.191693290734824,
+ "grad_norm": 0.056396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0584,
+ "step": 4755
+ },
+ {
+ "epoch": 15.194888178913738,
+ "grad_norm": 0.0859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0625,
+ "step": 4756
+ },
+ {
+ "epoch": 15.198083067092652,
+ "grad_norm": 0.0888671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0537,
+ "step": 4757
+ },
+ {
+ "epoch": 15.201277955271566,
+ "grad_norm": 0.10986328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0517,
+ "step": 4758
+ },
+ {
+ "epoch": 15.204472843450478,
+ "grad_norm": 0.08447265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0525,
+ "step": 4759
+ },
+ {
+ "epoch": 15.207667731629392,
+ "grad_norm": 0.1181640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0594,
+ "step": 4760
+ },
+ {
+ "epoch": 15.210862619808307,
+ "grad_norm": 0.087890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0329,
+ "step": 4761
+ },
+ {
+ "epoch": 15.21405750798722,
+ "grad_norm": 0.12353515625,
+ "learning_rate": 0.0005,
+ "loss": 1.054,
+ "step": 4762
+ },
+ {
+ "epoch": 15.217252396166135,
+ "grad_norm": 0.095703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0572,
+ "step": 4763
+ },
+ {
+ "epoch": 15.220447284345047,
+ "grad_norm": 0.11376953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0471,
+ "step": 4764
+ },
+ {
+ "epoch": 15.223642172523961,
+ "grad_norm": 0.059814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.045,
+ "step": 4765
+ },
+ {
+ "epoch": 15.226837060702875,
+ "grad_norm": 0.0869140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0557,
+ "step": 4766
+ },
+ {
+ "epoch": 15.23003194888179,
+ "grad_norm": 0.130859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0474,
+ "step": 4767
+ },
+ {
+ "epoch": 15.233226837060704,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0415,
+ "step": 4768
+ },
+ {
+ "epoch": 15.236421725239616,
+ "grad_norm": 0.1552734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0434,
+ "step": 4769
+ },
+ {
+ "epoch": 15.23961661341853,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.044,
+ "step": 4770
+ },
+ {
+ "epoch": 15.242811501597444,
+ "grad_norm": 0.1435546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0615,
+ "step": 4771
+ },
+ {
+ "epoch": 15.246006389776358,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0519,
+ "step": 4772
+ },
+ {
+ "epoch": 15.249201277955272,
+ "grad_norm": 0.1015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0509,
+ "step": 4773
+ },
+ {
+ "epoch": 15.252396166134185,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0473,
+ "step": 4774
+ },
+ {
+ "epoch": 15.255591054313099,
+ "grad_norm": 0.087890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0535,
+ "step": 4775
+ },
+ {
+ "epoch": 15.258785942492013,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0557,
+ "step": 4776
+ },
+ {
+ "epoch": 15.261980830670927,
+ "grad_norm": 0.09423828125,
+ "learning_rate": 0.0005,
+ "loss": 1.047,
+ "step": 4777
+ },
+ {
+ "epoch": 15.26517571884984,
+ "grad_norm": 0.09619140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0467,
+ "step": 4778
+ },
+ {
+ "epoch": 15.268370607028753,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0471,
+ "step": 4779
+ },
+ {
+ "epoch": 15.271565495207668,
+ "grad_norm": 0.0908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0497,
+ "step": 4780
+ },
+ {
+ "epoch": 15.274760383386582,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0567,
+ "step": 4781
+ },
+ {
+ "epoch": 15.277955271565496,
+ "grad_norm": 0.09228515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0597,
+ "step": 4782
+ },
+ {
+ "epoch": 15.281150159744408,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0624,
+ "step": 4783
+ },
+ {
+ "epoch": 15.284345047923322,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0502,
+ "step": 4784
+ },
+ {
+ "epoch": 15.287539936102236,
+ "grad_norm": 0.09814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0454,
+ "step": 4785
+ },
+ {
+ "epoch": 15.29073482428115,
+ "grad_norm": 0.09228515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0495,
+ "step": 4786
+ },
+ {
+ "epoch": 15.293929712460065,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0484,
+ "step": 4787
+ },
+ {
+ "epoch": 15.297124600638977,
+ "grad_norm": 0.0986328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0544,
+ "step": 4788
+ },
+ {
+ "epoch": 15.300319488817891,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0559,
+ "step": 4789
+ },
+ {
+ "epoch": 15.303514376996805,
+ "grad_norm": 0.083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.057,
+ "step": 4790
+ },
+ {
+ "epoch": 15.30670926517572,
+ "grad_norm": 0.08154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0471,
+ "step": 4791
+ },
+ {
+ "epoch": 15.309904153354633,
+ "grad_norm": 0.0869140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0365,
+ "step": 4792
+ },
+ {
+ "epoch": 15.313099041533546,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0439,
+ "step": 4793
+ },
+ {
+ "epoch": 15.31629392971246,
+ "grad_norm": 0.08984375,
+ "learning_rate": 0.0005,
+ "loss": 1.052,
+ "step": 4794
+ },
+ {
+ "epoch": 15.319488817891374,
+ "grad_norm": 0.060302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0561,
+ "step": 4795
+ },
+ {
+ "epoch": 15.322683706070288,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0516,
+ "step": 4796
+ },
+ {
+ "epoch": 15.3258785942492,
+ "grad_norm": 0.06005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0499,
+ "step": 4797
+ },
+ {
+ "epoch": 15.329073482428115,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0472,
+ "step": 4798
+ },
+ {
+ "epoch": 15.332268370607029,
+ "grad_norm": 0.06005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0437,
+ "step": 4799
+ },
+ {
+ "epoch": 15.335463258785943,
+ "grad_norm": 0.0927734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0602,
+ "step": 4800
+ },
+ {
+ "epoch": 15.338658146964857,
+ "grad_norm": 0.057373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0501,
+ "step": 4801
+ },
+ {
+ "epoch": 15.34185303514377,
+ "grad_norm": 0.0595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0525,
+ "step": 4802
+ },
+ {
+ "epoch": 15.345047923322683,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0417,
+ "step": 4803
+ },
+ {
+ "epoch": 15.348242811501597,
+ "grad_norm": 0.06005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0467,
+ "step": 4804
+ },
+ {
+ "epoch": 15.351437699680512,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0594,
+ "step": 4805
+ },
+ {
+ "epoch": 15.354632587859426,
+ "grad_norm": 0.107421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0447,
+ "step": 4806
+ },
+ {
+ "epoch": 15.357827476038338,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0524,
+ "step": 4807
+ },
+ {
+ "epoch": 15.361022364217252,
+ "grad_norm": 0.08154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.057,
+ "step": 4808
+ },
+ {
+ "epoch": 15.364217252396166,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.056,
+ "step": 4809
+ },
+ {
+ "epoch": 15.36741214057508,
+ "grad_norm": 0.053466796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0558,
+ "step": 4810
+ },
+ {
+ "epoch": 15.370607028753994,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0569,
+ "step": 4811
+ },
+ {
+ "epoch": 15.373801916932907,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0502,
+ "step": 4812
+ },
+ {
+ "epoch": 15.37699680511182,
+ "grad_norm": 0.061279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0561,
+ "step": 4813
+ },
+ {
+ "epoch": 15.380191693290735,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0558,
+ "step": 4814
+ },
+ {
+ "epoch": 15.383386581469649,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0534,
+ "step": 4815
+ },
+ {
+ "epoch": 15.386581469648561,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0587,
+ "step": 4816
+ },
+ {
+ "epoch": 15.389776357827476,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0514,
+ "step": 4817
+ },
+ {
+ "epoch": 15.39297124600639,
+ "grad_norm": 0.08056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.056,
+ "step": 4818
+ },
+ {
+ "epoch": 15.396166134185304,
+ "grad_norm": 0.062255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0418,
+ "step": 4819
+ },
+ {
+ "epoch": 15.399361022364218,
+ "grad_norm": 0.1083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0418,
+ "step": 4820
+ },
+ {
+ "epoch": 15.40255591054313,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0576,
+ "step": 4821
+ },
+ {
+ "epoch": 15.405750798722044,
+ "grad_norm": 0.08740234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0665,
+ "step": 4822
+ },
+ {
+ "epoch": 15.408945686900958,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0461,
+ "step": 4823
+ },
+ {
+ "epoch": 15.412140575079873,
+ "grad_norm": 0.099609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0417,
+ "step": 4824
+ },
+ {
+ "epoch": 15.415335463258787,
+ "grad_norm": 0.09765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0671,
+ "step": 4825
+ },
+ {
+ "epoch": 15.418530351437699,
+ "grad_norm": 0.05419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.056,
+ "step": 4826
+ },
+ {
+ "epoch": 15.421725239616613,
+ "grad_norm": 0.0576171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0616,
+ "step": 4827
+ },
+ {
+ "epoch": 15.424920127795527,
+ "grad_norm": 0.08984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0641,
+ "step": 4828
+ },
+ {
+ "epoch": 15.428115015974441,
+ "grad_norm": 0.05615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0372,
+ "step": 4829
+ },
+ {
+ "epoch": 15.431309904153355,
+ "grad_norm": 0.080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0638,
+ "step": 4830
+ },
+ {
+ "epoch": 15.434504792332268,
+ "grad_norm": 0.056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0431,
+ "step": 4831
+ },
+ {
+ "epoch": 15.437699680511182,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0443,
+ "step": 4832
+ },
+ {
+ "epoch": 15.440894568690096,
+ "grad_norm": 0.052978515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0573,
+ "step": 4833
+ },
+ {
+ "epoch": 15.44408945686901,
+ "grad_norm": 0.09033203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0423,
+ "step": 4834
+ },
+ {
+ "epoch": 15.447284345047922,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0615,
+ "step": 4835
+ },
+ {
+ "epoch": 15.450479233226837,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0535,
+ "step": 4836
+ },
+ {
+ "epoch": 15.45367412140575,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0471,
+ "step": 4837
+ },
+ {
+ "epoch": 15.456869009584665,
+ "grad_norm": 0.11474609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0557,
+ "step": 4838
+ },
+ {
+ "epoch": 15.460063897763579,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.052,
+ "step": 4839
+ },
+ {
+ "epoch": 15.463258785942491,
+ "grad_norm": 0.057861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0504,
+ "step": 4840
+ },
+ {
+ "epoch": 15.466453674121405,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0587,
+ "step": 4841
+ },
+ {
+ "epoch": 15.46964856230032,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0634,
+ "step": 4842
+ },
+ {
+ "epoch": 15.472843450479234,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.065,
+ "step": 4843
+ },
+ {
+ "epoch": 15.476038338658148,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0467,
+ "step": 4844
+ },
+ {
+ "epoch": 15.47923322683706,
+ "grad_norm": 0.09619140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0575,
+ "step": 4845
+ },
+ {
+ "epoch": 15.482428115015974,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0532,
+ "step": 4846
+ },
+ {
+ "epoch": 15.485623003194888,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0586,
+ "step": 4847
+ },
+ {
+ "epoch": 15.488817891373802,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0396,
+ "step": 4848
+ },
+ {
+ "epoch": 15.492012779552716,
+ "grad_norm": 0.10693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0458,
+ "step": 4849
+ },
+ {
+ "epoch": 15.495207667731629,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0504,
+ "step": 4850
+ },
+ {
+ "epoch": 15.498402555910543,
+ "grad_norm": 0.054931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0505,
+ "step": 4851
+ },
+ {
+ "epoch": 15.501597444089457,
+ "grad_norm": 0.049072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0425,
+ "step": 4852
+ },
+ {
+ "epoch": 15.504792332268371,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0541,
+ "step": 4853
+ },
+ {
+ "epoch": 15.507987220447284,
+ "grad_norm": 0.056884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0645,
+ "step": 4854
+ },
+ {
+ "epoch": 15.511182108626198,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.06,
+ "step": 4855
+ },
+ {
+ "epoch": 15.514376996805112,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0547,
+ "step": 4856
+ },
+ {
+ "epoch": 15.517571884984026,
+ "grad_norm": 0.08251953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0605,
+ "step": 4857
+ },
+ {
+ "epoch": 15.52076677316294,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0476,
+ "step": 4858
+ },
+ {
+ "epoch": 15.523961661341852,
+ "grad_norm": 0.0859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0615,
+ "step": 4859
+ },
+ {
+ "epoch": 15.527156549520766,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0515,
+ "step": 4860
+ },
+ {
+ "epoch": 15.53035143769968,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0631,
+ "step": 4861
+ },
+ {
+ "epoch": 15.533546325878595,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0465,
+ "step": 4862
+ },
+ {
+ "epoch": 15.536741214057509,
+ "grad_norm": 0.0576171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0561,
+ "step": 4863
+ },
+ {
+ "epoch": 15.539936102236421,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0426,
+ "step": 4864
+ },
+ {
+ "epoch": 15.543130990415335,
+ "grad_norm": 0.08740234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0669,
+ "step": 4865
+ },
+ {
+ "epoch": 15.54632587859425,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0478,
+ "step": 4866
+ },
+ {
+ "epoch": 15.549520766773163,
+ "grad_norm": 0.0947265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0611,
+ "step": 4867
+ },
+ {
+ "epoch": 15.552715654952078,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0567,
+ "step": 4868
+ },
+ {
+ "epoch": 15.55591054313099,
+ "grad_norm": 0.05908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0495,
+ "step": 4869
+ },
+ {
+ "epoch": 15.559105431309904,
+ "grad_norm": 0.058837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0615,
+ "step": 4870
+ },
+ {
+ "epoch": 15.562300319488818,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0449,
+ "step": 4871
+ },
+ {
+ "epoch": 15.565495207667732,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0535,
+ "step": 4872
+ },
+ {
+ "epoch": 15.568690095846645,
+ "grad_norm": 0.111328125,
+ "learning_rate": 0.0005,
+ "loss": 1.054,
+ "step": 4873
+ },
+ {
+ "epoch": 15.571884984025559,
+ "grad_norm": 0.08837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0496,
+ "step": 4874
+ },
+ {
+ "epoch": 15.575079872204473,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0581,
+ "step": 4875
+ },
+ {
+ "epoch": 15.578274760383387,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0578,
+ "step": 4876
+ },
+ {
+ "epoch": 15.581469648562301,
+ "grad_norm": 0.08740234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0334,
+ "step": 4877
+ },
+ {
+ "epoch": 15.584664536741213,
+ "grad_norm": 0.060791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0518,
+ "step": 4878
+ },
+ {
+ "epoch": 15.587859424920127,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0485,
+ "step": 4879
+ },
+ {
+ "epoch": 15.591054313099042,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0611,
+ "step": 4880
+ },
+ {
+ "epoch": 15.594249201277956,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.058,
+ "step": 4881
+ },
+ {
+ "epoch": 15.59744408945687,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0514,
+ "step": 4882
+ },
+ {
+ "epoch": 15.600638977635782,
+ "grad_norm": 0.080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0616,
+ "step": 4883
+ },
+ {
+ "epoch": 15.603833865814696,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0609,
+ "step": 4884
+ },
+ {
+ "epoch": 15.60702875399361,
+ "grad_norm": 0.126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0524,
+ "step": 4885
+ },
+ {
+ "epoch": 15.610223642172524,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.05,
+ "step": 4886
+ },
+ {
+ "epoch": 15.613418530351439,
+ "grad_norm": 0.09423828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0551,
+ "step": 4887
+ },
+ {
+ "epoch": 15.616613418530351,
+ "grad_norm": 0.0595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0462,
+ "step": 4888
+ },
+ {
+ "epoch": 15.619808306709265,
+ "grad_norm": 0.056396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.046,
+ "step": 4889
+ },
+ {
+ "epoch": 15.62300319488818,
+ "grad_norm": 0.053955078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0561,
+ "step": 4890
+ },
+ {
+ "epoch": 15.626198083067093,
+ "grad_norm": 0.061279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0525,
+ "step": 4891
+ },
+ {
+ "epoch": 15.629392971246006,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0416,
+ "step": 4892
+ },
+ {
+ "epoch": 15.63258785942492,
+ "grad_norm": 0.053955078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0435,
+ "step": 4893
+ },
+ {
+ "epoch": 15.635782747603834,
+ "grad_norm": 0.05859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0431,
+ "step": 4894
+ },
+ {
+ "epoch": 15.638977635782748,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0399,
+ "step": 4895
+ },
+ {
+ "epoch": 15.642172523961662,
+ "grad_norm": 0.05419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0556,
+ "step": 4896
+ },
+ {
+ "epoch": 15.645367412140574,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0429,
+ "step": 4897
+ },
+ {
+ "epoch": 15.648562300319488,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0484,
+ "step": 4898
+ },
+ {
+ "epoch": 15.651757188498403,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0444,
+ "step": 4899
+ },
+ {
+ "epoch": 15.654952076677317,
+ "grad_norm": 0.052734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0469,
+ "step": 4900
+ },
+ {
+ "epoch": 15.65814696485623,
+ "grad_norm": 0.0888671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0485,
+ "step": 4901
+ },
+ {
+ "epoch": 15.661341853035143,
+ "grad_norm": 0.09375,
+ "learning_rate": 0.0005,
+ "loss": 1.0623,
+ "step": 4902
+ },
+ {
+ "epoch": 15.664536741214057,
+ "grad_norm": 0.08544921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0611,
+ "step": 4903
+ },
+ {
+ "epoch": 15.667731629392971,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.056,
+ "step": 4904
+ },
+ {
+ "epoch": 15.670926517571885,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0545,
+ "step": 4905
+ },
+ {
+ "epoch": 15.6741214057508,
+ "grad_norm": 0.052734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0608,
+ "step": 4906
+ },
+ {
+ "epoch": 15.677316293929712,
+ "grad_norm": 0.060302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0518,
+ "step": 4907
+ },
+ {
+ "epoch": 15.680511182108626,
+ "grad_norm": 0.0595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0662,
+ "step": 4908
+ },
+ {
+ "epoch": 15.68370607028754,
+ "grad_norm": 0.061767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0533,
+ "step": 4909
+ },
+ {
+ "epoch": 15.686900958466454,
+ "grad_norm": 0.05712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0424,
+ "step": 4910
+ },
+ {
+ "epoch": 15.690095846645367,
+ "grad_norm": 0.1015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0432,
+ "step": 4911
+ },
+ {
+ "epoch": 15.69329073482428,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0448,
+ "step": 4912
+ },
+ {
+ "epoch": 15.696485623003195,
+ "grad_norm": 0.1298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.06,
+ "step": 4913
+ },
+ {
+ "epoch": 15.699680511182109,
+ "grad_norm": 0.05810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0499,
+ "step": 4914
+ },
+ {
+ "epoch": 15.702875399361023,
+ "grad_norm": 0.1025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0481,
+ "step": 4915
+ },
+ {
+ "epoch": 15.706070287539935,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.053,
+ "step": 4916
+ },
+ {
+ "epoch": 15.70926517571885,
+ "grad_norm": 0.08154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0472,
+ "step": 4917
+ },
+ {
+ "epoch": 15.712460063897764,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.052,
+ "step": 4918
+ },
+ {
+ "epoch": 15.715654952076678,
+ "grad_norm": 0.095703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0435,
+ "step": 4919
+ },
+ {
+ "epoch": 15.718849840255592,
+ "grad_norm": 0.0859375,
+ "learning_rate": 0.0005,
+ "loss": 1.06,
+ "step": 4920
+ },
+ {
+ "epoch": 15.722044728434504,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0485,
+ "step": 4921
+ },
+ {
+ "epoch": 15.725239616613418,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0447,
+ "step": 4922
+ },
+ {
+ "epoch": 15.728434504792332,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0504,
+ "step": 4923
+ },
+ {
+ "epoch": 15.731629392971247,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0497,
+ "step": 4924
+ },
+ {
+ "epoch": 15.73482428115016,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0525,
+ "step": 4925
+ },
+ {
+ "epoch": 15.738019169329073,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0572,
+ "step": 4926
+ },
+ {
+ "epoch": 15.741214057507987,
+ "grad_norm": 0.056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0486,
+ "step": 4927
+ },
+ {
+ "epoch": 15.744408945686901,
+ "grad_norm": 0.05859375,
+ "learning_rate": 0.0005,
+ "loss": 1.062,
+ "step": 4928
+ },
+ {
+ "epoch": 15.747603833865815,
+ "grad_norm": 0.047119140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0484,
+ "step": 4929
+ },
+ {
+ "epoch": 15.750798722044728,
+ "grad_norm": 0.060302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0501,
+ "step": 4930
+ },
+ {
+ "epoch": 15.753993610223642,
+ "grad_norm": 0.059326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.051,
+ "step": 4931
+ },
+ {
+ "epoch": 15.757188498402556,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0571,
+ "step": 4932
+ },
+ {
+ "epoch": 15.76038338658147,
+ "grad_norm": 0.0576171875,
+ "learning_rate": 0.0005,
+ "loss": 1.048,
+ "step": 4933
+ },
+ {
+ "epoch": 15.763578274760384,
+ "grad_norm": 0.060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0575,
+ "step": 4934
+ },
+ {
+ "epoch": 15.766773162939296,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0566,
+ "step": 4935
+ },
+ {
+ "epoch": 15.76996805111821,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0522,
+ "step": 4936
+ },
+ {
+ "epoch": 15.773162939297125,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0515,
+ "step": 4937
+ },
+ {
+ "epoch": 15.776357827476039,
+ "grad_norm": 0.060302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0468,
+ "step": 4938
+ },
+ {
+ "epoch": 15.779552715654953,
+ "grad_norm": 0.056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0519,
+ "step": 4939
+ },
+ {
+ "epoch": 15.782747603833865,
+ "grad_norm": 0.0869140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0392,
+ "step": 4940
+ },
+ {
+ "epoch": 15.78594249201278,
+ "grad_norm": 0.053955078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0596,
+ "step": 4941
+ },
+ {
+ "epoch": 15.789137380191693,
+ "grad_norm": 0.119140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0454,
+ "step": 4942
+ },
+ {
+ "epoch": 15.792332268370608,
+ "grad_norm": 0.05419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0474,
+ "step": 4943
+ },
+ {
+ "epoch": 15.795527156549522,
+ "grad_norm": 0.11767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0452,
+ "step": 4944
+ },
+ {
+ "epoch": 15.798722044728434,
+ "grad_norm": 0.09228515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0555,
+ "step": 4945
+ },
+ {
+ "epoch": 15.801916932907348,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0553,
+ "step": 4946
+ },
+ {
+ "epoch": 15.805111821086262,
+ "grad_norm": 0.1376953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0645,
+ "step": 4947
+ },
+ {
+ "epoch": 15.808306709265176,
+ "grad_norm": 0.050048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0662,
+ "step": 4948
+ },
+ {
+ "epoch": 15.811501597444089,
+ "grad_norm": 0.10693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0578,
+ "step": 4949
+ },
+ {
+ "epoch": 15.814696485623003,
+ "grad_norm": 0.05322265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0544,
+ "step": 4950
+ },
+ {
+ "epoch": 15.817891373801917,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0626,
+ "step": 4951
+ },
+ {
+ "epoch": 15.821086261980831,
+ "grad_norm": 0.05810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0545,
+ "step": 4952
+ },
+ {
+ "epoch": 15.824281150159745,
+ "grad_norm": 0.107421875,
+ "learning_rate": 0.0005,
+ "loss": 1.038,
+ "step": 4953
+ },
+ {
+ "epoch": 15.827476038338657,
+ "grad_norm": 0.047119140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0506,
+ "step": 4954
+ },
+ {
+ "epoch": 15.830670926517572,
+ "grad_norm": 0.08251953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0442,
+ "step": 4955
+ },
+ {
+ "epoch": 15.833865814696486,
+ "grad_norm": 0.044189453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0532,
+ "step": 4956
+ },
+ {
+ "epoch": 15.8370607028754,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0498,
+ "step": 4957
+ },
+ {
+ "epoch": 15.840255591054314,
+ "grad_norm": 0.05322265625,
+ "learning_rate": 0.0005,
+ "loss": 1.058,
+ "step": 4958
+ },
+ {
+ "epoch": 15.843450479233226,
+ "grad_norm": 0.06201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0436,
+ "step": 4959
+ },
+ {
+ "epoch": 15.84664536741214,
+ "grad_norm": 0.056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0533,
+ "step": 4960
+ },
+ {
+ "epoch": 15.849840255591054,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0531,
+ "step": 4961
+ },
+ {
+ "epoch": 15.853035143769969,
+ "grad_norm": 0.0576171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0587,
+ "step": 4962
+ },
+ {
+ "epoch": 15.856230031948883,
+ "grad_norm": 0.048095703125,
+ "learning_rate": 0.0005,
+ "loss": 1.037,
+ "step": 4963
+ },
+ {
+ "epoch": 15.859424920127795,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0482,
+ "step": 4964
+ },
+ {
+ "epoch": 15.86261980830671,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0525,
+ "step": 4965
+ },
+ {
+ "epoch": 15.865814696485623,
+ "grad_norm": 0.083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0394,
+ "step": 4966
+ },
+ {
+ "epoch": 15.869009584664537,
+ "grad_norm": 0.061767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0624,
+ "step": 4967
+ },
+ {
+ "epoch": 15.87220447284345,
+ "grad_norm": 0.06103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0573,
+ "step": 4968
+ },
+ {
+ "epoch": 15.875399361022364,
+ "grad_norm": 0.062255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0512,
+ "step": 4969
+ },
+ {
+ "epoch": 15.878594249201278,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0427,
+ "step": 4970
+ },
+ {
+ "epoch": 15.881789137380192,
+ "grad_norm": 0.08251953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0654,
+ "step": 4971
+ },
+ {
+ "epoch": 15.884984025559106,
+ "grad_norm": 0.09130859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0569,
+ "step": 4972
+ },
+ {
+ "epoch": 15.888178913738018,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0509,
+ "step": 4973
+ },
+ {
+ "epoch": 15.891373801916933,
+ "grad_norm": 0.1103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0453,
+ "step": 4974
+ },
+ {
+ "epoch": 15.894568690095847,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0559,
+ "step": 4975
+ },
+ {
+ "epoch": 15.89776357827476,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0492,
+ "step": 4976
+ },
+ {
+ "epoch": 15.900958466453675,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0424,
+ "step": 4977
+ },
+ {
+ "epoch": 15.904153354632587,
+ "grad_norm": 0.08837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0461,
+ "step": 4978
+ },
+ {
+ "epoch": 15.907348242811501,
+ "grad_norm": 0.130859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0552,
+ "step": 4979
+ },
+ {
+ "epoch": 15.910543130990416,
+ "grad_norm": 0.095703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0574,
+ "step": 4980
+ },
+ {
+ "epoch": 15.91373801916933,
+ "grad_norm": 0.185546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0676,
+ "step": 4981
+ },
+ {
+ "epoch": 15.916932907348244,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0661,
+ "step": 4982
+ },
+ {
+ "epoch": 15.920127795527156,
+ "grad_norm": 0.10693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0496,
+ "step": 4983
+ },
+ {
+ "epoch": 15.92332268370607,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.055,
+ "step": 4984
+ },
+ {
+ "epoch": 15.926517571884984,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0444,
+ "step": 4985
+ },
+ {
+ "epoch": 15.929712460063898,
+ "grad_norm": 0.08203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0609,
+ "step": 4986
+ },
+ {
+ "epoch": 15.93290734824281,
+ "grad_norm": 0.059814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0535,
+ "step": 4987
+ },
+ {
+ "epoch": 15.936102236421725,
+ "grad_norm": 0.1259765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0444,
+ "step": 4988
+ },
+ {
+ "epoch": 15.939297124600639,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0452,
+ "step": 4989
+ },
+ {
+ "epoch": 15.942492012779553,
+ "grad_norm": 0.06103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0519,
+ "step": 4990
+ },
+ {
+ "epoch": 15.945686900958467,
+ "grad_norm": 0.05517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0491,
+ "step": 4991
+ },
+ {
+ "epoch": 15.94888178913738,
+ "grad_norm": 0.057861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0441,
+ "step": 4992
+ },
+ {
+ "epoch": 15.952076677316294,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0547,
+ "step": 4993
+ },
+ {
+ "epoch": 15.955271565495208,
+ "grad_norm": 0.058837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0388,
+ "step": 4994
+ },
+ {
+ "epoch": 15.958466453674122,
+ "grad_norm": 0.0595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0534,
+ "step": 4995
+ },
+ {
+ "epoch": 15.961661341853034,
+ "grad_norm": 0.058837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.056,
+ "step": 4996
+ },
+ {
+ "epoch": 15.964856230031948,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0491,
+ "step": 4997
+ },
+ {
+ "epoch": 15.968051118210862,
+ "grad_norm": 0.05908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0535,
+ "step": 4998
+ },
+ {
+ "epoch": 15.971246006389777,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0536,
+ "step": 4999
+ },
+ {
+ "epoch": 15.97444089456869,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0512,
+ "step": 5000
+ },
+ {
+ "epoch": 15.977635782747605,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0493,
+ "step": 5001
+ },
+ {
+ "epoch": 15.980830670926517,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0415,
+ "step": 5002
+ },
+ {
+ "epoch": 15.984025559105431,
+ "grad_norm": 0.05712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0587,
+ "step": 5003
+ },
+ {
+ "epoch": 15.987220447284345,
+ "grad_norm": 0.056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0455,
+ "step": 5004
+ },
+ {
+ "epoch": 15.99041533546326,
+ "grad_norm": 0.08935546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0523,
+ "step": 5005
+ },
+ {
+ "epoch": 15.993610223642172,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0549,
+ "step": 5006
+ },
+ {
+ "epoch": 15.996805111821086,
+ "grad_norm": 0.07861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0496,
+ "step": 5007
+ },
+ {
+ "epoch": 16.0,
+ "grad_norm": 0.08740234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0591,
+ "step": 5008
+ },
+ {
+ "epoch": 16.003194888178914,
+ "grad_norm": 0.12890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0518,
+ "step": 5009
+ },
+ {
+ "epoch": 16.00638977635783,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0526,
+ "step": 5010
+ },
+ {
+ "epoch": 16.009584664536742,
+ "grad_norm": 0.1123046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0636,
+ "step": 5011
+ },
+ {
+ "epoch": 16.012779552715656,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0567,
+ "step": 5012
+ },
+ {
+ "epoch": 16.015974440894567,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.041,
+ "step": 5013
+ },
+ {
+ "epoch": 16.01916932907348,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0553,
+ "step": 5014
+ },
+ {
+ "epoch": 16.022364217252395,
+ "grad_norm": 0.058837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0525,
+ "step": 5015
+ },
+ {
+ "epoch": 16.02555910543131,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.047,
+ "step": 5016
+ },
+ {
+ "epoch": 16.028753993610223,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0604,
+ "step": 5017
+ },
+ {
+ "epoch": 16.031948881789138,
+ "grad_norm": 0.0888671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0496,
+ "step": 5018
+ },
+ {
+ "epoch": 16.03514376996805,
+ "grad_norm": 0.059326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0616,
+ "step": 5019
+ },
+ {
+ "epoch": 16.038338658146966,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0412,
+ "step": 5020
+ },
+ {
+ "epoch": 16.04153354632588,
+ "grad_norm": 0.057861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0465,
+ "step": 5021
+ },
+ {
+ "epoch": 16.044728434504794,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0393,
+ "step": 5022
+ },
+ {
+ "epoch": 16.047923322683705,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0578,
+ "step": 5023
+ },
+ {
+ "epoch": 16.05111821086262,
+ "grad_norm": 0.05029296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0571,
+ "step": 5024
+ },
+ {
+ "epoch": 16.054313099041533,
+ "grad_norm": 0.1328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0391,
+ "step": 5025
+ },
+ {
+ "epoch": 16.057507987220447,
+ "grad_norm": 0.09423828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0516,
+ "step": 5026
+ },
+ {
+ "epoch": 16.06070287539936,
+ "grad_norm": 0.11474609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0495,
+ "step": 5027
+ },
+ {
+ "epoch": 16.063897763578275,
+ "grad_norm": 0.1357421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0544,
+ "step": 5028
+ },
+ {
+ "epoch": 16.06709265175719,
+ "grad_norm": 0.08984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0553,
+ "step": 5029
+ },
+ {
+ "epoch": 16.070287539936103,
+ "grad_norm": 0.1337890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0507,
+ "step": 5030
+ },
+ {
+ "epoch": 16.073482428115017,
+ "grad_norm": 0.07861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0503,
+ "step": 5031
+ },
+ {
+ "epoch": 16.076677316293928,
+ "grad_norm": 0.14453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0549,
+ "step": 5032
+ },
+ {
+ "epoch": 16.079872204472842,
+ "grad_norm": 0.08642578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0566,
+ "step": 5033
+ },
+ {
+ "epoch": 16.083067092651756,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0521,
+ "step": 5034
+ },
+ {
+ "epoch": 16.08626198083067,
+ "grad_norm": 0.1005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0515,
+ "step": 5035
+ },
+ {
+ "epoch": 16.089456869009584,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0555,
+ "step": 5036
+ },
+ {
+ "epoch": 16.0926517571885,
+ "grad_norm": 0.134765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0559,
+ "step": 5037
+ },
+ {
+ "epoch": 16.095846645367413,
+ "grad_norm": 0.049560546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0492,
+ "step": 5038
+ },
+ {
+ "epoch": 16.099041533546327,
+ "grad_norm": 0.158203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0446,
+ "step": 5039
+ },
+ {
+ "epoch": 16.10223642172524,
+ "grad_norm": 0.0859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0539,
+ "step": 5040
+ },
+ {
+ "epoch": 16.105431309904155,
+ "grad_norm": 0.11962890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0522,
+ "step": 5041
+ },
+ {
+ "epoch": 16.108626198083066,
+ "grad_norm": 0.09033203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0457,
+ "step": 5042
+ },
+ {
+ "epoch": 16.11182108626198,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0527,
+ "step": 5043
+ },
+ {
+ "epoch": 16.115015974440894,
+ "grad_norm": 0.06103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0617,
+ "step": 5044
+ },
+ {
+ "epoch": 16.118210862619808,
+ "grad_norm": 0.07861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0495,
+ "step": 5045
+ },
+ {
+ "epoch": 16.121405750798722,
+ "grad_norm": 0.0830078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0477,
+ "step": 5046
+ },
+ {
+ "epoch": 16.124600638977636,
+ "grad_norm": 0.055419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0681,
+ "step": 5047
+ },
+ {
+ "epoch": 16.12779552715655,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0479,
+ "step": 5048
+ },
+ {
+ "epoch": 16.130990415335464,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0465,
+ "step": 5049
+ },
+ {
+ "epoch": 16.13418530351438,
+ "grad_norm": 0.0908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0464,
+ "step": 5050
+ },
+ {
+ "epoch": 16.13738019169329,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0462,
+ "step": 5051
+ },
+ {
+ "epoch": 16.140575079872203,
+ "grad_norm": 0.11181640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0487,
+ "step": 5052
+ },
+ {
+ "epoch": 16.143769968051117,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0454,
+ "step": 5053
+ },
+ {
+ "epoch": 16.14696485623003,
+ "grad_norm": 0.111328125,
+ "learning_rate": 0.0005,
+ "loss": 1.036,
+ "step": 5054
+ },
+ {
+ "epoch": 16.150159744408946,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0553,
+ "step": 5055
+ },
+ {
+ "epoch": 16.15335463258786,
+ "grad_norm": 0.0810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0533,
+ "step": 5056
+ },
+ {
+ "epoch": 16.156549520766774,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.054,
+ "step": 5057
+ },
+ {
+ "epoch": 16.159744408945688,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0516,
+ "step": 5058
+ },
+ {
+ "epoch": 16.162939297124602,
+ "grad_norm": 0.08056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0445,
+ "step": 5059
+ },
+ {
+ "epoch": 16.166134185303516,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0646,
+ "step": 5060
+ },
+ {
+ "epoch": 16.169329073482427,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0448,
+ "step": 5061
+ },
+ {
+ "epoch": 16.17252396166134,
+ "grad_norm": 0.0869140625,
+ "learning_rate": 0.0005,
+ "loss": 1.062,
+ "step": 5062
+ },
+ {
+ "epoch": 16.175718849840255,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0466,
+ "step": 5063
+ },
+ {
+ "epoch": 16.17891373801917,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0549,
+ "step": 5064
+ },
+ {
+ "epoch": 16.182108626198083,
+ "grad_norm": 0.0869140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0517,
+ "step": 5065
+ },
+ {
+ "epoch": 16.185303514376997,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0443,
+ "step": 5066
+ },
+ {
+ "epoch": 16.18849840255591,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0545,
+ "step": 5067
+ },
+ {
+ "epoch": 16.191693290734825,
+ "grad_norm": 0.05908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0634,
+ "step": 5068
+ },
+ {
+ "epoch": 16.19488817891374,
+ "grad_norm": 0.10009765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0481,
+ "step": 5069
+ },
+ {
+ "epoch": 16.19808306709265,
+ "grad_norm": 0.052001953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0395,
+ "step": 5070
+ },
+ {
+ "epoch": 16.201277955271564,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0461,
+ "step": 5071
+ },
+ {
+ "epoch": 16.20447284345048,
+ "grad_norm": 0.08642578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0492,
+ "step": 5072
+ },
+ {
+ "epoch": 16.207667731629392,
+ "grad_norm": 0.080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0607,
+ "step": 5073
+ },
+ {
+ "epoch": 16.210862619808307,
+ "grad_norm": 0.0537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0593,
+ "step": 5074
+ },
+ {
+ "epoch": 16.21405750798722,
+ "grad_norm": 0.061279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0586,
+ "step": 5075
+ },
+ {
+ "epoch": 16.217252396166135,
+ "grad_norm": 0.053466796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0537,
+ "step": 5076
+ },
+ {
+ "epoch": 16.22044728434505,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0492,
+ "step": 5077
+ },
+ {
+ "epoch": 16.223642172523963,
+ "grad_norm": 0.056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0443,
+ "step": 5078
+ },
+ {
+ "epoch": 16.226837060702877,
+ "grad_norm": 0.0927734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0447,
+ "step": 5079
+ },
+ {
+ "epoch": 16.230031948881788,
+ "grad_norm": 0.06103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0516,
+ "step": 5080
+ },
+ {
+ "epoch": 16.233226837060702,
+ "grad_norm": 0.052978515625,
+ "learning_rate": 0.0005,
+ "loss": 1.043,
+ "step": 5081
+ },
+ {
+ "epoch": 16.236421725239616,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0498,
+ "step": 5082
+ },
+ {
+ "epoch": 16.23961661341853,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0361,
+ "step": 5083
+ },
+ {
+ "epoch": 16.242811501597444,
+ "grad_norm": 0.059326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0503,
+ "step": 5084
+ },
+ {
+ "epoch": 16.24600638977636,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0458,
+ "step": 5085
+ },
+ {
+ "epoch": 16.249201277955272,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.059,
+ "step": 5086
+ },
+ {
+ "epoch": 16.252396166134186,
+ "grad_norm": 0.0947265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0532,
+ "step": 5087
+ },
+ {
+ "epoch": 16.2555910543131,
+ "grad_norm": 0.107421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0505,
+ "step": 5088
+ },
+ {
+ "epoch": 16.25878594249201,
+ "grad_norm": 0.1396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.037,
+ "step": 5089
+ },
+ {
+ "epoch": 16.261980830670925,
+ "grad_norm": 0.21484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0538,
+ "step": 5090
+ },
+ {
+ "epoch": 16.26517571884984,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0604,
+ "step": 5091
+ },
+ {
+ "epoch": 16.268370607028753,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0587,
+ "step": 5092
+ },
+ {
+ "epoch": 16.271565495207668,
+ "grad_norm": 0.08984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0519,
+ "step": 5093
+ },
+ {
+ "epoch": 16.27476038338658,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0554,
+ "step": 5094
+ },
+ {
+ "epoch": 16.277955271565496,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0476,
+ "step": 5095
+ },
+ {
+ "epoch": 16.28115015974441,
+ "grad_norm": 0.056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0635,
+ "step": 5096
+ },
+ {
+ "epoch": 16.284345047923324,
+ "grad_norm": 0.08203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0434,
+ "step": 5097
+ },
+ {
+ "epoch": 16.287539936102238,
+ "grad_norm": 0.053466796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0578,
+ "step": 5098
+ },
+ {
+ "epoch": 16.29073482428115,
+ "grad_norm": 0.09375,
+ "learning_rate": 0.0005,
+ "loss": 1.0528,
+ "step": 5099
+ },
+ {
+ "epoch": 16.293929712460063,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0401,
+ "step": 5100
+ },
+ {
+ "epoch": 16.297124600638977,
+ "grad_norm": 0.10693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0561,
+ "step": 5101
+ },
+ {
+ "epoch": 16.30031948881789,
+ "grad_norm": 0.08349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0399,
+ "step": 5102
+ },
+ {
+ "epoch": 16.303514376996805,
+ "grad_norm": 0.0927734375,
+ "learning_rate": 0.0005,
+ "loss": 1.049,
+ "step": 5103
+ },
+ {
+ "epoch": 16.30670926517572,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0581,
+ "step": 5104
+ },
+ {
+ "epoch": 16.309904153354633,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0516,
+ "step": 5105
+ },
+ {
+ "epoch": 16.313099041533548,
+ "grad_norm": 0.06103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0543,
+ "step": 5106
+ },
+ {
+ "epoch": 16.31629392971246,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0532,
+ "step": 5107
+ },
+ {
+ "epoch": 16.319488817891372,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0507,
+ "step": 5108
+ },
+ {
+ "epoch": 16.322683706070286,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0375,
+ "step": 5109
+ },
+ {
+ "epoch": 16.3258785942492,
+ "grad_norm": 0.09423828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0548,
+ "step": 5110
+ },
+ {
+ "epoch": 16.329073482428115,
+ "grad_norm": 0.06201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0534,
+ "step": 5111
+ },
+ {
+ "epoch": 16.33226837060703,
+ "grad_norm": 0.09521484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0398,
+ "step": 5112
+ },
+ {
+ "epoch": 16.335463258785943,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0377,
+ "step": 5113
+ },
+ {
+ "epoch": 16.338658146964857,
+ "grad_norm": 0.09375,
+ "learning_rate": 0.0005,
+ "loss": 1.0652,
+ "step": 5114
+ },
+ {
+ "epoch": 16.34185303514377,
+ "grad_norm": 0.059814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0521,
+ "step": 5115
+ },
+ {
+ "epoch": 16.345047923322685,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0484,
+ "step": 5116
+ },
+ {
+ "epoch": 16.3482428115016,
+ "grad_norm": 0.0849609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0463,
+ "step": 5117
+ },
+ {
+ "epoch": 16.35143769968051,
+ "grad_norm": 0.060302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0602,
+ "step": 5118
+ },
+ {
+ "epoch": 16.354632587859424,
+ "grad_norm": 0.0869140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0567,
+ "step": 5119
+ },
+ {
+ "epoch": 16.357827476038338,
+ "grad_norm": 0.083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0438,
+ "step": 5120
+ },
+ {
+ "epoch": 16.361022364217252,
+ "grad_norm": 0.08544921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0435,
+ "step": 5121
+ },
+ {
+ "epoch": 16.364217252396166,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0478,
+ "step": 5122
+ },
+ {
+ "epoch": 16.36741214057508,
+ "grad_norm": 0.06103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0675,
+ "step": 5123
+ },
+ {
+ "epoch": 16.370607028753994,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0574,
+ "step": 5124
+ },
+ {
+ "epoch": 16.37380191693291,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0588,
+ "step": 5125
+ },
+ {
+ "epoch": 16.376996805111823,
+ "grad_norm": 0.057861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0542,
+ "step": 5126
+ },
+ {
+ "epoch": 16.380191693290733,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0537,
+ "step": 5127
+ },
+ {
+ "epoch": 16.383386581469647,
+ "grad_norm": 0.054931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0496,
+ "step": 5128
+ },
+ {
+ "epoch": 16.38658146964856,
+ "grad_norm": 0.058837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0494,
+ "step": 5129
+ },
+ {
+ "epoch": 16.389776357827476,
+ "grad_norm": 0.059814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0612,
+ "step": 5130
+ },
+ {
+ "epoch": 16.39297124600639,
+ "grad_norm": 0.08349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0542,
+ "step": 5131
+ },
+ {
+ "epoch": 16.396166134185304,
+ "grad_norm": 0.052001953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0543,
+ "step": 5132
+ },
+ {
+ "epoch": 16.399361022364218,
+ "grad_norm": 0.087890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0413,
+ "step": 5133
+ },
+ {
+ "epoch": 16.402555910543132,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0515,
+ "step": 5134
+ },
+ {
+ "epoch": 16.405750798722046,
+ "grad_norm": 0.126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0422,
+ "step": 5135
+ },
+ {
+ "epoch": 16.408945686900957,
+ "grad_norm": 0.057373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0438,
+ "step": 5136
+ },
+ {
+ "epoch": 16.41214057507987,
+ "grad_norm": 0.10400390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0502,
+ "step": 5137
+ },
+ {
+ "epoch": 16.415335463258785,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.052,
+ "step": 5138
+ },
+ {
+ "epoch": 16.4185303514377,
+ "grad_norm": 0.1083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0452,
+ "step": 5139
+ },
+ {
+ "epoch": 16.421725239616613,
+ "grad_norm": 0.058349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0579,
+ "step": 5140
+ },
+ {
+ "epoch": 16.424920127795527,
+ "grad_norm": 0.10888671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0609,
+ "step": 5141
+ },
+ {
+ "epoch": 16.42811501597444,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.055,
+ "step": 5142
+ },
+ {
+ "epoch": 16.431309904153355,
+ "grad_norm": 0.08740234375,
+ "learning_rate": 0.0005,
+ "loss": 1.048,
+ "step": 5143
+ },
+ {
+ "epoch": 16.43450479233227,
+ "grad_norm": 0.056884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0539,
+ "step": 5144
+ },
+ {
+ "epoch": 16.437699680511184,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.061,
+ "step": 5145
+ },
+ {
+ "epoch": 16.440894568690094,
+ "grad_norm": 0.056396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0433,
+ "step": 5146
+ },
+ {
+ "epoch": 16.44408945686901,
+ "grad_norm": 0.08642578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0558,
+ "step": 5147
+ },
+ {
+ "epoch": 16.447284345047922,
+ "grad_norm": 0.0595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0578,
+ "step": 5148
+ },
+ {
+ "epoch": 16.450479233226837,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0542,
+ "step": 5149
+ },
+ {
+ "epoch": 16.45367412140575,
+ "grad_norm": 0.1337890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0385,
+ "step": 5150
+ },
+ {
+ "epoch": 16.456869009584665,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0608,
+ "step": 5151
+ },
+ {
+ "epoch": 16.46006389776358,
+ "grad_norm": 0.1484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0487,
+ "step": 5152
+ },
+ {
+ "epoch": 16.463258785942493,
+ "grad_norm": 0.08251953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0408,
+ "step": 5153
+ },
+ {
+ "epoch": 16.466453674121407,
+ "grad_norm": 0.2060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0388,
+ "step": 5154
+ },
+ {
+ "epoch": 16.46964856230032,
+ "grad_norm": 0.08203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0486,
+ "step": 5155
+ },
+ {
+ "epoch": 16.472843450479232,
+ "grad_norm": 0.1328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0537,
+ "step": 5156
+ },
+ {
+ "epoch": 16.476038338658146,
+ "grad_norm": 0.134765625,
+ "learning_rate": 0.0005,
+ "loss": 1.052,
+ "step": 5157
+ },
+ {
+ "epoch": 16.47923322683706,
+ "grad_norm": 0.171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0522,
+ "step": 5158
+ },
+ {
+ "epoch": 16.482428115015974,
+ "grad_norm": 0.09912109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0551,
+ "step": 5159
+ },
+ {
+ "epoch": 16.48562300319489,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0598,
+ "step": 5160
+ },
+ {
+ "epoch": 16.488817891373802,
+ "grad_norm": 0.08935546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0475,
+ "step": 5161
+ },
+ {
+ "epoch": 16.492012779552716,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0489,
+ "step": 5162
+ },
+ {
+ "epoch": 16.49520766773163,
+ "grad_norm": 0.09716796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0576,
+ "step": 5163
+ },
+ {
+ "epoch": 16.498402555910545,
+ "grad_norm": 0.1064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0524,
+ "step": 5164
+ },
+ {
+ "epoch": 16.501597444089455,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0428,
+ "step": 5165
+ },
+ {
+ "epoch": 16.50479233226837,
+ "grad_norm": 0.1357421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0562,
+ "step": 5166
+ },
+ {
+ "epoch": 16.507987220447284,
+ "grad_norm": 0.07861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0322,
+ "step": 5167
+ },
+ {
+ "epoch": 16.511182108626198,
+ "grad_norm": 0.1220703125,
+ "learning_rate": 0.0005,
+ "loss": 1.054,
+ "step": 5168
+ },
+ {
+ "epoch": 16.51437699680511,
+ "grad_norm": 0.09033203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0555,
+ "step": 5169
+ },
+ {
+ "epoch": 16.517571884984026,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0569,
+ "step": 5170
+ },
+ {
+ "epoch": 16.52076677316294,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0566,
+ "step": 5171
+ },
+ {
+ "epoch": 16.523961661341854,
+ "grad_norm": 0.057373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0564,
+ "step": 5172
+ },
+ {
+ "epoch": 16.527156549520768,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0509,
+ "step": 5173
+ },
+ {
+ "epoch": 16.53035143769968,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0552,
+ "step": 5174
+ },
+ {
+ "epoch": 16.533546325878593,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0582,
+ "step": 5175
+ },
+ {
+ "epoch": 16.536741214057507,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0423,
+ "step": 5176
+ },
+ {
+ "epoch": 16.53993610223642,
+ "grad_norm": 0.058349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.056,
+ "step": 5177
+ },
+ {
+ "epoch": 16.543130990415335,
+ "grad_norm": 0.062255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0423,
+ "step": 5178
+ },
+ {
+ "epoch": 16.54632587859425,
+ "grad_norm": 0.056884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0536,
+ "step": 5179
+ },
+ {
+ "epoch": 16.549520766773163,
+ "grad_norm": 0.06005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0559,
+ "step": 5180
+ },
+ {
+ "epoch": 16.552715654952078,
+ "grad_norm": 0.056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0425,
+ "step": 5181
+ },
+ {
+ "epoch": 16.55591054313099,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0572,
+ "step": 5182
+ },
+ {
+ "epoch": 16.559105431309906,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0409,
+ "step": 5183
+ },
+ {
+ "epoch": 16.562300319488816,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0438,
+ "step": 5184
+ },
+ {
+ "epoch": 16.56549520766773,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0472,
+ "step": 5185
+ },
+ {
+ "epoch": 16.568690095846645,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0433,
+ "step": 5186
+ },
+ {
+ "epoch": 16.57188498402556,
+ "grad_norm": 0.1083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0491,
+ "step": 5187
+ },
+ {
+ "epoch": 16.575079872204473,
+ "grad_norm": 0.08642578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0483,
+ "step": 5188
+ },
+ {
+ "epoch": 16.578274760383387,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0479,
+ "step": 5189
+ },
+ {
+ "epoch": 16.5814696485623,
+ "grad_norm": 0.058349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0443,
+ "step": 5190
+ },
+ {
+ "epoch": 16.584664536741215,
+ "grad_norm": 0.05224609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0506,
+ "step": 5191
+ },
+ {
+ "epoch": 16.58785942492013,
+ "grad_norm": 0.08203125,
+ "learning_rate": 0.0005,
+ "loss": 1.045,
+ "step": 5192
+ },
+ {
+ "epoch": 16.591054313099043,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0465,
+ "step": 5193
+ },
+ {
+ "epoch": 16.594249201277954,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0636,
+ "step": 5194
+ },
+ {
+ "epoch": 16.597444089456868,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0481,
+ "step": 5195
+ },
+ {
+ "epoch": 16.600638977635782,
+ "grad_norm": 0.058349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0481,
+ "step": 5196
+ },
+ {
+ "epoch": 16.603833865814696,
+ "grad_norm": 0.061279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0546,
+ "step": 5197
+ },
+ {
+ "epoch": 16.60702875399361,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0426,
+ "step": 5198
+ },
+ {
+ "epoch": 16.610223642172524,
+ "grad_norm": 0.05859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0405,
+ "step": 5199
+ },
+ {
+ "epoch": 16.61341853035144,
+ "grad_norm": 0.054443359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0447,
+ "step": 5200
+ },
+ {
+ "epoch": 16.616613418530353,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0587,
+ "step": 5201
+ },
+ {
+ "epoch": 16.619808306709267,
+ "grad_norm": 0.08056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0561,
+ "step": 5202
+ },
+ {
+ "epoch": 16.623003194888177,
+ "grad_norm": 0.049560546875,
+ "learning_rate": 0.0005,
+ "loss": 1.044,
+ "step": 5203
+ },
+ {
+ "epoch": 16.62619808306709,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0507,
+ "step": 5204
+ },
+ {
+ "epoch": 16.629392971246006,
+ "grad_norm": 0.06005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0369,
+ "step": 5205
+ },
+ {
+ "epoch": 16.63258785942492,
+ "grad_norm": 0.1220703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0522,
+ "step": 5206
+ },
+ {
+ "epoch": 16.635782747603834,
+ "grad_norm": 0.060791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0557,
+ "step": 5207
+ },
+ {
+ "epoch": 16.638977635782748,
+ "grad_norm": 0.10546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0426,
+ "step": 5208
+ },
+ {
+ "epoch": 16.642172523961662,
+ "grad_norm": 0.060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0496,
+ "step": 5209
+ },
+ {
+ "epoch": 16.645367412140576,
+ "grad_norm": 0.052978515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0337,
+ "step": 5210
+ },
+ {
+ "epoch": 16.64856230031949,
+ "grad_norm": 0.05224609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0386,
+ "step": 5211
+ },
+ {
+ "epoch": 16.6517571884984,
+ "grad_norm": 0.05615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0512,
+ "step": 5212
+ },
+ {
+ "epoch": 16.654952076677315,
+ "grad_norm": 0.046875,
+ "learning_rate": 0.0005,
+ "loss": 1.052,
+ "step": 5213
+ },
+ {
+ "epoch": 16.65814696485623,
+ "grad_norm": 0.08642578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0377,
+ "step": 5214
+ },
+ {
+ "epoch": 16.661341853035143,
+ "grad_norm": 0.048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.049,
+ "step": 5215
+ },
+ {
+ "epoch": 16.664536741214057,
+ "grad_norm": 0.10595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0493,
+ "step": 5216
+ },
+ {
+ "epoch": 16.66773162939297,
+ "grad_norm": 0.09716796875,
+ "learning_rate": 0.0005,
+ "loss": 1.052,
+ "step": 5217
+ },
+ {
+ "epoch": 16.670926517571885,
+ "grad_norm": 0.173828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0471,
+ "step": 5218
+ },
+ {
+ "epoch": 16.6741214057508,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0467,
+ "step": 5219
+ },
+ {
+ "epoch": 16.677316293929714,
+ "grad_norm": 0.1318359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0496,
+ "step": 5220
+ },
+ {
+ "epoch": 16.680511182108628,
+ "grad_norm": 0.11865234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0552,
+ "step": 5221
+ },
+ {
+ "epoch": 16.68370607028754,
+ "grad_norm": 0.09814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.062,
+ "step": 5222
+ },
+ {
+ "epoch": 16.686900958466452,
+ "grad_norm": 0.1826171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0554,
+ "step": 5223
+ },
+ {
+ "epoch": 16.690095846645367,
+ "grad_norm": 0.1376953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0448,
+ "step": 5224
+ },
+ {
+ "epoch": 16.69329073482428,
+ "grad_norm": 0.146484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0489,
+ "step": 5225
+ },
+ {
+ "epoch": 16.696485623003195,
+ "grad_norm": 0.326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.066,
+ "step": 5226
+ },
+ {
+ "epoch": 16.69968051118211,
+ "grad_norm": 0.18359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0567,
+ "step": 5227
+ },
+ {
+ "epoch": 16.702875399361023,
+ "grad_norm": 0.10986328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0597,
+ "step": 5228
+ },
+ {
+ "epoch": 16.706070287539937,
+ "grad_norm": 0.2119140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0467,
+ "step": 5229
+ },
+ {
+ "epoch": 16.70926517571885,
+ "grad_norm": 0.10400390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0436,
+ "step": 5230
+ },
+ {
+ "epoch": 16.712460063897765,
+ "grad_norm": 0.1298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0477,
+ "step": 5231
+ },
+ {
+ "epoch": 16.715654952076676,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0558,
+ "step": 5232
+ },
+ {
+ "epoch": 16.71884984025559,
+ "grad_norm": 0.1181640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0479,
+ "step": 5233
+ },
+ {
+ "epoch": 16.722044728434504,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0499,
+ "step": 5234
+ },
+ {
+ "epoch": 16.72523961661342,
+ "grad_norm": 0.1083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0544,
+ "step": 5235
+ },
+ {
+ "epoch": 16.728434504792332,
+ "grad_norm": 0.08447265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0621,
+ "step": 5236
+ },
+ {
+ "epoch": 16.731629392971247,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0602,
+ "step": 5237
+ },
+ {
+ "epoch": 16.73482428115016,
+ "grad_norm": 0.09765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0602,
+ "step": 5238
+ },
+ {
+ "epoch": 16.738019169329075,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0565,
+ "step": 5239
+ },
+ {
+ "epoch": 16.74121405750799,
+ "grad_norm": 0.08837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0532,
+ "step": 5240
+ },
+ {
+ "epoch": 16.7444089456869,
+ "grad_norm": 0.09033203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0492,
+ "step": 5241
+ },
+ {
+ "epoch": 16.747603833865814,
+ "grad_norm": 0.10595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.046,
+ "step": 5242
+ },
+ {
+ "epoch": 16.750798722044728,
+ "grad_norm": 0.057861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0525,
+ "step": 5243
+ },
+ {
+ "epoch": 16.75399361022364,
+ "grad_norm": 0.05029296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0535,
+ "step": 5244
+ },
+ {
+ "epoch": 16.757188498402556,
+ "grad_norm": 0.055419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0406,
+ "step": 5245
+ },
+ {
+ "epoch": 16.76038338658147,
+ "grad_norm": 0.062255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0567,
+ "step": 5246
+ },
+ {
+ "epoch": 16.763578274760384,
+ "grad_norm": 0.054931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0588,
+ "step": 5247
+ },
+ {
+ "epoch": 16.766773162939298,
+ "grad_norm": 0.05908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0412,
+ "step": 5248
+ },
+ {
+ "epoch": 16.769968051118212,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0489,
+ "step": 5249
+ },
+ {
+ "epoch": 16.773162939297123,
+ "grad_norm": 0.05078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0453,
+ "step": 5250
+ },
+ {
+ "epoch": 16.776357827476037,
+ "grad_norm": 0.0615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0543,
+ "step": 5251
+ },
+ {
+ "epoch": 16.77955271565495,
+ "grad_norm": 0.051025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0564,
+ "step": 5252
+ },
+ {
+ "epoch": 16.782747603833865,
+ "grad_norm": 0.053955078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0532,
+ "step": 5253
+ },
+ {
+ "epoch": 16.78594249201278,
+ "grad_norm": 0.059814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0571,
+ "step": 5254
+ },
+ {
+ "epoch": 16.789137380191693,
+ "grad_norm": 0.05078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0392,
+ "step": 5255
+ },
+ {
+ "epoch": 16.792332268370608,
+ "grad_norm": 0.056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0549,
+ "step": 5256
+ },
+ {
+ "epoch": 16.79552715654952,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0488,
+ "step": 5257
+ },
+ {
+ "epoch": 16.798722044728436,
+ "grad_norm": 0.05322265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0606,
+ "step": 5258
+ },
+ {
+ "epoch": 16.80191693290735,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.044,
+ "step": 5259
+ },
+ {
+ "epoch": 16.80511182108626,
+ "grad_norm": 0.052978515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0664,
+ "step": 5260
+ },
+ {
+ "epoch": 16.808306709265175,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0527,
+ "step": 5261
+ },
+ {
+ "epoch": 16.81150159744409,
+ "grad_norm": 0.0517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0556,
+ "step": 5262
+ },
+ {
+ "epoch": 16.814696485623003,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0531,
+ "step": 5263
+ },
+ {
+ "epoch": 16.817891373801917,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0472,
+ "step": 5264
+ },
+ {
+ "epoch": 16.82108626198083,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0477,
+ "step": 5265
+ },
+ {
+ "epoch": 16.824281150159745,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0384,
+ "step": 5266
+ },
+ {
+ "epoch": 16.82747603833866,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.04,
+ "step": 5267
+ },
+ {
+ "epoch": 16.830670926517573,
+ "grad_norm": 0.087890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0484,
+ "step": 5268
+ },
+ {
+ "epoch": 16.833865814696484,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.047,
+ "step": 5269
+ },
+ {
+ "epoch": 16.837060702875398,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0459,
+ "step": 5270
+ },
+ {
+ "epoch": 16.840255591054312,
+ "grad_norm": 0.08984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0435,
+ "step": 5271
+ },
+ {
+ "epoch": 16.843450479233226,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0581,
+ "step": 5272
+ },
+ {
+ "epoch": 16.84664536741214,
+ "grad_norm": 0.083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0544,
+ "step": 5273
+ },
+ {
+ "epoch": 16.849840255591054,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0621,
+ "step": 5274
+ },
+ {
+ "epoch": 16.85303514376997,
+ "grad_norm": 0.130859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0638,
+ "step": 5275
+ },
+ {
+ "epoch": 16.856230031948883,
+ "grad_norm": 0.17578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0466,
+ "step": 5276
+ },
+ {
+ "epoch": 16.859424920127797,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0443,
+ "step": 5277
+ },
+ {
+ "epoch": 16.86261980830671,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.045,
+ "step": 5278
+ },
+ {
+ "epoch": 16.86581469648562,
+ "grad_norm": 0.10009765625,
+ "learning_rate": 0.0005,
+ "loss": 1.043,
+ "step": 5279
+ },
+ {
+ "epoch": 16.869009584664536,
+ "grad_norm": 0.0849609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0448,
+ "step": 5280
+ },
+ {
+ "epoch": 16.87220447284345,
+ "grad_norm": 0.10498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0494,
+ "step": 5281
+ },
+ {
+ "epoch": 16.875399361022364,
+ "grad_norm": 0.0859375,
+ "learning_rate": 0.0005,
+ "loss": 1.057,
+ "step": 5282
+ },
+ {
+ "epoch": 16.878594249201278,
+ "grad_norm": 0.10693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0493,
+ "step": 5283
+ },
+ {
+ "epoch": 16.881789137380192,
+ "grad_norm": 0.08544921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0437,
+ "step": 5284
+ },
+ {
+ "epoch": 16.884984025559106,
+ "grad_norm": 0.1953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0503,
+ "step": 5285
+ },
+ {
+ "epoch": 16.88817891373802,
+ "grad_norm": 0.0859375,
+ "learning_rate": 0.0005,
+ "loss": 1.048,
+ "step": 5286
+ },
+ {
+ "epoch": 16.891373801916934,
+ "grad_norm": 0.189453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0531,
+ "step": 5287
+ },
+ {
+ "epoch": 16.894568690095845,
+ "grad_norm": 0.1865234375,
+ "learning_rate": 0.0005,
+ "loss": 1.03,
+ "step": 5288
+ },
+ {
+ "epoch": 16.89776357827476,
+ "grad_norm": 0.0810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0577,
+ "step": 5289
+ },
+ {
+ "epoch": 16.900958466453673,
+ "grad_norm": 0.2138671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0366,
+ "step": 5290
+ },
+ {
+ "epoch": 16.904153354632587,
+ "grad_norm": 0.10986328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0479,
+ "step": 5291
+ },
+ {
+ "epoch": 16.9073482428115,
+ "grad_norm": 0.150390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0503,
+ "step": 5292
+ },
+ {
+ "epoch": 16.910543130990416,
+ "grad_norm": 0.220703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0635,
+ "step": 5293
+ },
+ {
+ "epoch": 16.91373801916933,
+ "grad_norm": 0.054443359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0508,
+ "step": 5294
+ },
+ {
+ "epoch": 16.916932907348244,
+ "grad_norm": 0.2060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0444,
+ "step": 5295
+ },
+ {
+ "epoch": 16.920127795527158,
+ "grad_norm": 0.1162109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0578,
+ "step": 5296
+ },
+ {
+ "epoch": 16.923322683706072,
+ "grad_norm": 0.10791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0521,
+ "step": 5297
+ },
+ {
+ "epoch": 16.926517571884983,
+ "grad_norm": 0.1630859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0504,
+ "step": 5298
+ },
+ {
+ "epoch": 16.929712460063897,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0608,
+ "step": 5299
+ },
+ {
+ "epoch": 16.93290734824281,
+ "grad_norm": 0.1650390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0575,
+ "step": 5300
+ },
+ {
+ "epoch": 16.936102236421725,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0441,
+ "step": 5301
+ },
+ {
+ "epoch": 16.93929712460064,
+ "grad_norm": 0.10009765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0423,
+ "step": 5302
+ },
+ {
+ "epoch": 16.942492012779553,
+ "grad_norm": 0.05712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0543,
+ "step": 5303
+ },
+ {
+ "epoch": 16.945686900958467,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0576,
+ "step": 5304
+ },
+ {
+ "epoch": 16.94888178913738,
+ "grad_norm": 0.052490234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0468,
+ "step": 5305
+ },
+ {
+ "epoch": 16.952076677316295,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0437,
+ "step": 5306
+ },
+ {
+ "epoch": 16.955271565495206,
+ "grad_norm": 0.05029296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0365,
+ "step": 5307
+ },
+ {
+ "epoch": 16.95846645367412,
+ "grad_norm": 0.1162109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0477,
+ "step": 5308
+ },
+ {
+ "epoch": 16.961661341853034,
+ "grad_norm": 0.158203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0444,
+ "step": 5309
+ },
+ {
+ "epoch": 16.96485623003195,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0534,
+ "step": 5310
+ },
+ {
+ "epoch": 16.968051118210862,
+ "grad_norm": 0.10107421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0608,
+ "step": 5311
+ },
+ {
+ "epoch": 16.971246006389777,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0531,
+ "step": 5312
+ },
+ {
+ "epoch": 16.97444089456869,
+ "grad_norm": 0.055908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0464,
+ "step": 5313
+ },
+ {
+ "epoch": 16.977635782747605,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0483,
+ "step": 5314
+ },
+ {
+ "epoch": 16.98083067092652,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0519,
+ "step": 5315
+ },
+ {
+ "epoch": 16.984025559105433,
+ "grad_norm": 0.05517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0603,
+ "step": 5316
+ },
+ {
+ "epoch": 16.987220447284344,
+ "grad_norm": 0.062255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0576,
+ "step": 5317
+ },
+ {
+ "epoch": 16.990415335463258,
+ "grad_norm": 0.057861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0483,
+ "step": 5318
+ },
+ {
+ "epoch": 16.99361022364217,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0451,
+ "step": 5319
+ },
+ {
+ "epoch": 16.996805111821086,
+ "grad_norm": 0.0546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0646,
+ "step": 5320
+ },
+ {
+ "epoch": 17.0,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0489,
+ "step": 5321
+ },
+ {
+ "epoch": 17.003194888178914,
+ "grad_norm": 0.047607421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0429,
+ "step": 5322
+ },
+ {
+ "epoch": 17.00638977635783,
+ "grad_norm": 0.08251953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0601,
+ "step": 5323
+ },
+ {
+ "epoch": 17.009584664536742,
+ "grad_norm": 0.1083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0562,
+ "step": 5324
+ },
+ {
+ "epoch": 17.012779552715656,
+ "grad_norm": 0.212890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0448,
+ "step": 5325
+ },
+ {
+ "epoch": 17.015974440894567,
+ "grad_norm": 0.09912109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0639,
+ "step": 5326
+ },
+ {
+ "epoch": 17.01916932907348,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0389,
+ "step": 5327
+ },
+ {
+ "epoch": 17.022364217252395,
+ "grad_norm": 0.058837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0536,
+ "step": 5328
+ },
+ {
+ "epoch": 17.02555910543131,
+ "grad_norm": 0.08935546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0381,
+ "step": 5329
+ },
+ {
+ "epoch": 17.028753993610223,
+ "grad_norm": 0.07861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0514,
+ "step": 5330
+ },
+ {
+ "epoch": 17.031948881789138,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0602,
+ "step": 5331
+ },
+ {
+ "epoch": 17.03514376996805,
+ "grad_norm": 0.09716796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0402,
+ "step": 5332
+ },
+ {
+ "epoch": 17.038338658146966,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0546,
+ "step": 5333
+ },
+ {
+ "epoch": 17.04153354632588,
+ "grad_norm": 0.107421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0462,
+ "step": 5334
+ },
+ {
+ "epoch": 17.044728434504794,
+ "grad_norm": 0.07861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0502,
+ "step": 5335
+ },
+ {
+ "epoch": 17.047923322683705,
+ "grad_norm": 0.056396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0514,
+ "step": 5336
+ },
+ {
+ "epoch": 17.05111821086262,
+ "grad_norm": 0.08837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.049,
+ "step": 5337
+ },
+ {
+ "epoch": 17.054313099041533,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0524,
+ "step": 5338
+ },
+ {
+ "epoch": 17.057507987220447,
+ "grad_norm": 0.058837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0401,
+ "step": 5339
+ },
+ {
+ "epoch": 17.06070287539936,
+ "grad_norm": 0.07861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0402,
+ "step": 5340
+ },
+ {
+ "epoch": 17.063897763578275,
+ "grad_norm": 0.11669921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0441,
+ "step": 5341
+ },
+ {
+ "epoch": 17.06709265175719,
+ "grad_norm": 0.216796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0321,
+ "step": 5342
+ },
+ {
+ "epoch": 17.070287539936103,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0581,
+ "step": 5343
+ },
+ {
+ "epoch": 17.073482428115017,
+ "grad_norm": 0.2578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0516,
+ "step": 5344
+ },
+ {
+ "epoch": 17.076677316293928,
+ "grad_norm": 0.10498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.053,
+ "step": 5345
+ },
+ {
+ "epoch": 17.079872204472842,
+ "grad_norm": 0.1748046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0437,
+ "step": 5346
+ },
+ {
+ "epoch": 17.083067092651756,
+ "grad_norm": 0.1533203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0539,
+ "step": 5347
+ },
+ {
+ "epoch": 17.08626198083067,
+ "grad_norm": 0.0908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0578,
+ "step": 5348
+ },
+ {
+ "epoch": 17.089456869009584,
+ "grad_norm": 0.142578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0434,
+ "step": 5349
+ },
+ {
+ "epoch": 17.0926517571885,
+ "grad_norm": 0.059326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.055,
+ "step": 5350
+ },
+ {
+ "epoch": 17.095846645367413,
+ "grad_norm": 0.12109375,
+ "learning_rate": 0.0005,
+ "loss": 1.046,
+ "step": 5351
+ },
+ {
+ "epoch": 17.099041533546327,
+ "grad_norm": 0.056884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0528,
+ "step": 5352
+ },
+ {
+ "epoch": 17.10223642172524,
+ "grad_norm": 0.146484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0545,
+ "step": 5353
+ },
+ {
+ "epoch": 17.105431309904155,
+ "grad_norm": 0.05419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0532,
+ "step": 5354
+ },
+ {
+ "epoch": 17.108626198083066,
+ "grad_norm": 0.1787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0427,
+ "step": 5355
+ },
+ {
+ "epoch": 17.11182108626198,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0562,
+ "step": 5356
+ },
+ {
+ "epoch": 17.115015974440894,
+ "grad_norm": 0.1357421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0559,
+ "step": 5357
+ },
+ {
+ "epoch": 17.118210862619808,
+ "grad_norm": 0.103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0478,
+ "step": 5358
+ },
+ {
+ "epoch": 17.121405750798722,
+ "grad_norm": 0.109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0437,
+ "step": 5359
+ },
+ {
+ "epoch": 17.124600638977636,
+ "grad_norm": 0.1630859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0538,
+ "step": 5360
+ },
+ {
+ "epoch": 17.12779552715655,
+ "grad_norm": 0.05322265625,
+ "learning_rate": 0.0005,
+ "loss": 1.055,
+ "step": 5361
+ },
+ {
+ "epoch": 17.130990415335464,
+ "grad_norm": 0.12109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0563,
+ "step": 5362
+ },
+ {
+ "epoch": 17.13418530351438,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0538,
+ "step": 5363
+ },
+ {
+ "epoch": 17.13738019169329,
+ "grad_norm": 0.09814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0641,
+ "step": 5364
+ },
+ {
+ "epoch": 17.140575079872203,
+ "grad_norm": 0.10009765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0491,
+ "step": 5365
+ },
+ {
+ "epoch": 17.143769968051117,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.056,
+ "step": 5366
+ },
+ {
+ "epoch": 17.14696485623003,
+ "grad_norm": 0.1484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0694,
+ "step": 5367
+ },
+ {
+ "epoch": 17.150159744408946,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0579,
+ "step": 5368
+ },
+ {
+ "epoch": 17.15335463258786,
+ "grad_norm": 0.11376953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0386,
+ "step": 5369
+ },
+ {
+ "epoch": 17.156549520766774,
+ "grad_norm": 0.0546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0535,
+ "step": 5370
+ },
+ {
+ "epoch": 17.159744408945688,
+ "grad_norm": 0.1083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0559,
+ "step": 5371
+ },
+ {
+ "epoch": 17.162939297124602,
+ "grad_norm": 0.0556640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0474,
+ "step": 5372
+ },
+ {
+ "epoch": 17.166134185303516,
+ "grad_norm": 0.09521484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0435,
+ "step": 5373
+ },
+ {
+ "epoch": 17.169329073482427,
+ "grad_norm": 0.05517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0602,
+ "step": 5374
+ },
+ {
+ "epoch": 17.17252396166134,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0546,
+ "step": 5375
+ },
+ {
+ "epoch": 17.175718849840255,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0406,
+ "step": 5376
+ },
+ {
+ "epoch": 17.17891373801917,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0443,
+ "step": 5377
+ },
+ {
+ "epoch": 17.182108626198083,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0535,
+ "step": 5378
+ },
+ {
+ "epoch": 17.185303514376997,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0411,
+ "step": 5379
+ },
+ {
+ "epoch": 17.18849840255591,
+ "grad_norm": 0.05859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0523,
+ "step": 5380
+ },
+ {
+ "epoch": 17.191693290734825,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0603,
+ "step": 5381
+ },
+ {
+ "epoch": 17.19488817891374,
+ "grad_norm": 0.058349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0474,
+ "step": 5382
+ },
+ {
+ "epoch": 17.19808306709265,
+ "grad_norm": 0.055908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0444,
+ "step": 5383
+ },
+ {
+ "epoch": 17.201277955271564,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0531,
+ "step": 5384
+ },
+ {
+ "epoch": 17.20447284345048,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0431,
+ "step": 5385
+ },
+ {
+ "epoch": 17.207667731629392,
+ "grad_norm": 0.091796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0554,
+ "step": 5386
+ },
+ {
+ "epoch": 17.210862619808307,
+ "grad_norm": 0.10107421875,
+ "learning_rate": 0.0005,
+ "loss": 1.054,
+ "step": 5387
+ },
+ {
+ "epoch": 17.21405750798722,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0426,
+ "step": 5388
+ },
+ {
+ "epoch": 17.217252396166135,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0493,
+ "step": 5389
+ },
+ {
+ "epoch": 17.22044728434505,
+ "grad_norm": 0.09228515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0385,
+ "step": 5390
+ },
+ {
+ "epoch": 17.223642172523963,
+ "grad_norm": 0.1015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0466,
+ "step": 5391
+ },
+ {
+ "epoch": 17.226837060702877,
+ "grad_norm": 0.08154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0613,
+ "step": 5392
+ },
+ {
+ "epoch": 17.230031948881788,
+ "grad_norm": 0.06201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0409,
+ "step": 5393
+ },
+ {
+ "epoch": 17.233226837060702,
+ "grad_norm": 0.10400390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0474,
+ "step": 5394
+ },
+ {
+ "epoch": 17.236421725239616,
+ "grad_norm": 0.1015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0518,
+ "step": 5395
+ },
+ {
+ "epoch": 17.23961661341853,
+ "grad_norm": 0.10546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0538,
+ "step": 5396
+ },
+ {
+ "epoch": 17.242811501597444,
+ "grad_norm": 0.146484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0529,
+ "step": 5397
+ },
+ {
+ "epoch": 17.24600638977636,
+ "grad_norm": 0.0810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0463,
+ "step": 5398
+ },
+ {
+ "epoch": 17.249201277955272,
+ "grad_norm": 0.146484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0362,
+ "step": 5399
+ },
+ {
+ "epoch": 17.252396166134186,
+ "grad_norm": 0.08349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0472,
+ "step": 5400
+ },
+ {
+ "epoch": 17.2555910543131,
+ "grad_norm": 0.10107421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0379,
+ "step": 5401
+ },
+ {
+ "epoch": 17.25878594249201,
+ "grad_norm": 0.083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0595,
+ "step": 5402
+ },
+ {
+ "epoch": 17.261980830670925,
+ "grad_norm": 0.11767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0577,
+ "step": 5403
+ },
+ {
+ "epoch": 17.26517571884984,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0446,
+ "step": 5404
+ },
+ {
+ "epoch": 17.268370607028753,
+ "grad_norm": 0.115234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0662,
+ "step": 5405
+ },
+ {
+ "epoch": 17.271565495207668,
+ "grad_norm": 0.16015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0551,
+ "step": 5406
+ },
+ {
+ "epoch": 17.27476038338658,
+ "grad_norm": 0.05712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0344,
+ "step": 5407
+ },
+ {
+ "epoch": 17.277955271565496,
+ "grad_norm": 0.11279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0535,
+ "step": 5408
+ },
+ {
+ "epoch": 17.28115015974441,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0566,
+ "step": 5409
+ },
+ {
+ "epoch": 17.284345047923324,
+ "grad_norm": 0.09033203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0413,
+ "step": 5410
+ },
+ {
+ "epoch": 17.287539936102238,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.056,
+ "step": 5411
+ },
+ {
+ "epoch": 17.29073482428115,
+ "grad_norm": 0.0830078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0412,
+ "step": 5412
+ },
+ {
+ "epoch": 17.293929712460063,
+ "grad_norm": 0.057373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0356,
+ "step": 5413
+ },
+ {
+ "epoch": 17.297124600638977,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0477,
+ "step": 5414
+ },
+ {
+ "epoch": 17.30031948881789,
+ "grad_norm": 0.054443359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0522,
+ "step": 5415
+ },
+ {
+ "epoch": 17.303514376996805,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0477,
+ "step": 5416
+ },
+ {
+ "epoch": 17.30670926517572,
+ "grad_norm": 0.048583984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0348,
+ "step": 5417
+ },
+ {
+ "epoch": 17.309904153354633,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0516,
+ "step": 5418
+ },
+ {
+ "epoch": 17.313099041533548,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0435,
+ "step": 5419
+ },
+ {
+ "epoch": 17.31629392971246,
+ "grad_norm": 0.0595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.041,
+ "step": 5420
+ },
+ {
+ "epoch": 17.319488817891372,
+ "grad_norm": 0.05419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0526,
+ "step": 5421
+ },
+ {
+ "epoch": 17.322683706070286,
+ "grad_norm": 0.05908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0448,
+ "step": 5422
+ },
+ {
+ "epoch": 17.3258785942492,
+ "grad_norm": 0.099609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0417,
+ "step": 5423
+ },
+ {
+ "epoch": 17.329073482428115,
+ "grad_norm": 0.0576171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0486,
+ "step": 5424
+ },
+ {
+ "epoch": 17.33226837060703,
+ "grad_norm": 0.043701171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0422,
+ "step": 5425
+ },
+ {
+ "epoch": 17.335463258785943,
+ "grad_norm": 0.08203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0499,
+ "step": 5426
+ },
+ {
+ "epoch": 17.338658146964857,
+ "grad_norm": 0.052490234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0486,
+ "step": 5427
+ },
+ {
+ "epoch": 17.34185303514377,
+ "grad_norm": 0.10009765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0547,
+ "step": 5428
+ },
+ {
+ "epoch": 17.345047923322685,
+ "grad_norm": 0.052734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0412,
+ "step": 5429
+ },
+ {
+ "epoch": 17.3482428115016,
+ "grad_norm": 0.1064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0498,
+ "step": 5430
+ },
+ {
+ "epoch": 17.35143769968051,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.052,
+ "step": 5431
+ },
+ {
+ "epoch": 17.354632587859424,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.06,
+ "step": 5432
+ },
+ {
+ "epoch": 17.357827476038338,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0494,
+ "step": 5433
+ },
+ {
+ "epoch": 17.361022364217252,
+ "grad_norm": 0.05810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0474,
+ "step": 5434
+ },
+ {
+ "epoch": 17.364217252396166,
+ "grad_norm": 0.06103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0548,
+ "step": 5435
+ },
+ {
+ "epoch": 17.36741214057508,
+ "grad_norm": 0.083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0457,
+ "step": 5436
+ },
+ {
+ "epoch": 17.370607028753994,
+ "grad_norm": 0.138671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0519,
+ "step": 5437
+ },
+ {
+ "epoch": 17.37380191693291,
+ "grad_norm": 0.051513671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0632,
+ "step": 5438
+ },
+ {
+ "epoch": 17.376996805111823,
+ "grad_norm": 0.177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0567,
+ "step": 5439
+ },
+ {
+ "epoch": 17.380191693290733,
+ "grad_norm": 0.0869140625,
+ "learning_rate": 0.0005,
+ "loss": 1.045,
+ "step": 5440
+ },
+ {
+ "epoch": 17.383386581469647,
+ "grad_norm": 0.11865234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0402,
+ "step": 5441
+ },
+ {
+ "epoch": 17.38658146964856,
+ "grad_norm": 0.09765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0477,
+ "step": 5442
+ },
+ {
+ "epoch": 17.389776357827476,
+ "grad_norm": 0.09228515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0548,
+ "step": 5443
+ },
+ {
+ "epoch": 17.39297124600639,
+ "grad_norm": 0.10888671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0486,
+ "step": 5444
+ },
+ {
+ "epoch": 17.396166134185304,
+ "grad_norm": 0.08740234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0527,
+ "step": 5445
+ },
+ {
+ "epoch": 17.399361022364218,
+ "grad_norm": 0.10107421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0316,
+ "step": 5446
+ },
+ {
+ "epoch": 17.402555910543132,
+ "grad_norm": 0.1142578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0472,
+ "step": 5447
+ },
+ {
+ "epoch": 17.405750798722046,
+ "grad_norm": 0.212890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0546,
+ "step": 5448
+ },
+ {
+ "epoch": 17.408945686900957,
+ "grad_norm": 0.12890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0394,
+ "step": 5449
+ },
+ {
+ "epoch": 17.41214057507987,
+ "grad_norm": 0.1318359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0452,
+ "step": 5450
+ },
+ {
+ "epoch": 17.415335463258785,
+ "grad_norm": 0.2451171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0476,
+ "step": 5451
+ },
+ {
+ "epoch": 17.4185303514377,
+ "grad_norm": 0.0556640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0489,
+ "step": 5452
+ },
+ {
+ "epoch": 17.421725239616613,
+ "grad_norm": 0.171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0638,
+ "step": 5453
+ },
+ {
+ "epoch": 17.424920127795527,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0555,
+ "step": 5454
+ },
+ {
+ "epoch": 17.42811501597444,
+ "grad_norm": 0.1396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0518,
+ "step": 5455
+ },
+ {
+ "epoch": 17.431309904153355,
+ "grad_norm": 0.1064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0499,
+ "step": 5456
+ },
+ {
+ "epoch": 17.43450479233227,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0526,
+ "step": 5457
+ },
+ {
+ "epoch": 17.437699680511184,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0543,
+ "step": 5458
+ },
+ {
+ "epoch": 17.440894568690094,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0515,
+ "step": 5459
+ },
+ {
+ "epoch": 17.44408945686901,
+ "grad_norm": 0.087890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0306,
+ "step": 5460
+ },
+ {
+ "epoch": 17.447284345047922,
+ "grad_norm": 0.080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0423,
+ "step": 5461
+ },
+ {
+ "epoch": 17.450479233226837,
+ "grad_norm": 0.1181640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0529,
+ "step": 5462
+ },
+ {
+ "epoch": 17.45367412140575,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0438,
+ "step": 5463
+ },
+ {
+ "epoch": 17.456869009584665,
+ "grad_norm": 0.10986328125,
+ "learning_rate": 0.0005,
+ "loss": 1.055,
+ "step": 5464
+ },
+ {
+ "epoch": 17.46006389776358,
+ "grad_norm": 0.061279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0445,
+ "step": 5465
+ },
+ {
+ "epoch": 17.463258785942493,
+ "grad_norm": 0.11083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0457,
+ "step": 5466
+ },
+ {
+ "epoch": 17.466453674121407,
+ "grad_norm": 0.0546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0509,
+ "step": 5467
+ },
+ {
+ "epoch": 17.46964856230032,
+ "grad_norm": 0.1435546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0621,
+ "step": 5468
+ },
+ {
+ "epoch": 17.472843450479232,
+ "grad_norm": 0.04541015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0424,
+ "step": 5469
+ },
+ {
+ "epoch": 17.476038338658146,
+ "grad_norm": 0.125,
+ "learning_rate": 0.0005,
+ "loss": 1.0612,
+ "step": 5470
+ },
+ {
+ "epoch": 17.47923322683706,
+ "grad_norm": 0.052734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0427,
+ "step": 5471
+ },
+ {
+ "epoch": 17.482428115015974,
+ "grad_norm": 0.060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0399,
+ "step": 5472
+ },
+ {
+ "epoch": 17.48562300319489,
+ "grad_norm": 0.06201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0602,
+ "step": 5473
+ },
+ {
+ "epoch": 17.488817891373802,
+ "grad_norm": 0.04833984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0624,
+ "step": 5474
+ },
+ {
+ "epoch": 17.492012779552716,
+ "grad_norm": 0.060791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0479,
+ "step": 5475
+ },
+ {
+ "epoch": 17.49520766773163,
+ "grad_norm": 0.053466796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0566,
+ "step": 5476
+ },
+ {
+ "epoch": 17.498402555910545,
+ "grad_norm": 0.08349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0558,
+ "step": 5477
+ },
+ {
+ "epoch": 17.501597444089455,
+ "grad_norm": 0.1083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0422,
+ "step": 5478
+ },
+ {
+ "epoch": 17.50479233226837,
+ "grad_norm": 0.08740234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0425,
+ "step": 5479
+ },
+ {
+ "epoch": 17.507987220447284,
+ "grad_norm": 0.1455078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0488,
+ "step": 5480
+ },
+ {
+ "epoch": 17.511182108626198,
+ "grad_norm": 0.10693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0339,
+ "step": 5481
+ },
+ {
+ "epoch": 17.51437699680511,
+ "grad_norm": 0.1640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0322,
+ "step": 5482
+ },
+ {
+ "epoch": 17.517571884984026,
+ "grad_norm": 0.0830078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0513,
+ "step": 5483
+ },
+ {
+ "epoch": 17.52076677316294,
+ "grad_norm": 0.08251953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0443,
+ "step": 5484
+ },
+ {
+ "epoch": 17.523961661341854,
+ "grad_norm": 0.09130859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0536,
+ "step": 5485
+ },
+ {
+ "epoch": 17.527156549520768,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0483,
+ "step": 5486
+ },
+ {
+ "epoch": 17.53035143769968,
+ "grad_norm": 0.09228515625,
+ "learning_rate": 0.0005,
+ "loss": 1.04,
+ "step": 5487
+ },
+ {
+ "epoch": 17.533546325878593,
+ "grad_norm": 0.09814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0448,
+ "step": 5488
+ },
+ {
+ "epoch": 17.536741214057507,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0528,
+ "step": 5489
+ },
+ {
+ "epoch": 17.53993610223642,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.058,
+ "step": 5490
+ },
+ {
+ "epoch": 17.543130990415335,
+ "grad_norm": 0.05908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0455,
+ "step": 5491
+ },
+ {
+ "epoch": 17.54632587859425,
+ "grad_norm": 0.05859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0505,
+ "step": 5492
+ },
+ {
+ "epoch": 17.549520766773163,
+ "grad_norm": 0.05712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0418,
+ "step": 5493
+ },
+ {
+ "epoch": 17.552715654952078,
+ "grad_norm": 0.0576171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0486,
+ "step": 5494
+ },
+ {
+ "epoch": 17.55591054313099,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0463,
+ "step": 5495
+ },
+ {
+ "epoch": 17.559105431309906,
+ "grad_norm": 0.107421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0471,
+ "step": 5496
+ },
+ {
+ "epoch": 17.562300319488816,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0547,
+ "step": 5497
+ },
+ {
+ "epoch": 17.56549520766773,
+ "grad_norm": 0.09716796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0479,
+ "step": 5498
+ },
+ {
+ "epoch": 17.568690095846645,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0576,
+ "step": 5499
+ },
+ {
+ "epoch": 17.57188498402556,
+ "grad_norm": 0.1220703125,
+ "learning_rate": 0.0005,
+ "loss": 1.06,
+ "step": 5500
+ },
+ {
+ "epoch": 17.575079872204473,
+ "grad_norm": 0.058837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0377,
+ "step": 5501
+ },
+ {
+ "epoch": 17.578274760383387,
+ "grad_norm": 0.103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0501,
+ "step": 5502
+ },
+ {
+ "epoch": 17.5814696485623,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0526,
+ "step": 5503
+ },
+ {
+ "epoch": 17.584664536741215,
+ "grad_norm": 0.11669921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0466,
+ "step": 5504
+ },
+ {
+ "epoch": 17.58785942492013,
+ "grad_norm": 0.054931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0425,
+ "step": 5505
+ },
+ {
+ "epoch": 17.591054313099043,
+ "grad_norm": 0.109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0509,
+ "step": 5506
+ },
+ {
+ "epoch": 17.594249201277954,
+ "grad_norm": 0.061279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0519,
+ "step": 5507
+ },
+ {
+ "epoch": 17.597444089456868,
+ "grad_norm": 0.1298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0363,
+ "step": 5508
+ },
+ {
+ "epoch": 17.600638977635782,
+ "grad_norm": 0.150390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0507,
+ "step": 5509
+ },
+ {
+ "epoch": 17.603833865814696,
+ "grad_norm": 0.1728515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0649,
+ "step": 5510
+ },
+ {
+ "epoch": 17.60702875399361,
+ "grad_norm": 0.3203125,
+ "learning_rate": 0.0005,
+ "loss": 1.062,
+ "step": 5511
+ },
+ {
+ "epoch": 17.610223642172524,
+ "grad_norm": 0.185546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0444,
+ "step": 5512
+ },
+ {
+ "epoch": 17.61341853035144,
+ "grad_norm": 0.07861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0525,
+ "step": 5513
+ },
+ {
+ "epoch": 17.616613418530353,
+ "grad_norm": 0.1279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0386,
+ "step": 5514
+ },
+ {
+ "epoch": 17.619808306709267,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0503,
+ "step": 5515
+ },
+ {
+ "epoch": 17.623003194888177,
+ "grad_norm": 0.126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.046,
+ "step": 5516
+ },
+ {
+ "epoch": 17.62619808306709,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0425,
+ "step": 5517
+ },
+ {
+ "epoch": 17.629392971246006,
+ "grad_norm": 0.126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0588,
+ "step": 5518
+ },
+ {
+ "epoch": 17.63258785942492,
+ "grad_norm": 0.1357421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0441,
+ "step": 5519
+ },
+ {
+ "epoch": 17.635782747603834,
+ "grad_norm": 0.1025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0538,
+ "step": 5520
+ },
+ {
+ "epoch": 17.638977635782748,
+ "grad_norm": 0.193359375,
+ "learning_rate": 0.0005,
+ "loss": 1.048,
+ "step": 5521
+ },
+ {
+ "epoch": 17.642172523961662,
+ "grad_norm": 0.060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0559,
+ "step": 5522
+ },
+ {
+ "epoch": 17.645367412140576,
+ "grad_norm": 0.19140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0525,
+ "step": 5523
+ },
+ {
+ "epoch": 17.64856230031949,
+ "grad_norm": 0.115234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0217,
+ "step": 5524
+ },
+ {
+ "epoch": 17.6517571884984,
+ "grad_norm": 0.1376953125,
+ "learning_rate": 0.0005,
+ "loss": 1.043,
+ "step": 5525
+ },
+ {
+ "epoch": 17.654952076677315,
+ "grad_norm": 0.18359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0433,
+ "step": 5526
+ },
+ {
+ "epoch": 17.65814696485623,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0475,
+ "step": 5527
+ },
+ {
+ "epoch": 17.661341853035143,
+ "grad_norm": 0.1435546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0428,
+ "step": 5528
+ },
+ {
+ "epoch": 17.664536741214057,
+ "grad_norm": 0.1142578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0358,
+ "step": 5529
+ },
+ {
+ "epoch": 17.66773162939297,
+ "grad_norm": 0.27734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0431,
+ "step": 5530
+ },
+ {
+ "epoch": 17.670926517571885,
+ "grad_norm": 0.1298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0505,
+ "step": 5531
+ },
+ {
+ "epoch": 17.6741214057508,
+ "grad_norm": 0.1337890625,
+ "learning_rate": 0.0005,
+ "loss": 1.06,
+ "step": 5532
+ },
+ {
+ "epoch": 17.677316293929714,
+ "grad_norm": 0.12451171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0438,
+ "step": 5533
+ },
+ {
+ "epoch": 17.680511182108628,
+ "grad_norm": 0.1162109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0522,
+ "step": 5534
+ },
+ {
+ "epoch": 17.68370607028754,
+ "grad_norm": 0.193359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0244,
+ "step": 5535
+ },
+ {
+ "epoch": 17.686900958466452,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0604,
+ "step": 5536
+ },
+ {
+ "epoch": 17.690095846645367,
+ "grad_norm": 0.1416015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0476,
+ "step": 5537
+ },
+ {
+ "epoch": 17.69329073482428,
+ "grad_norm": 0.08203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0629,
+ "step": 5538
+ },
+ {
+ "epoch": 17.696485623003195,
+ "grad_norm": 0.19140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0454,
+ "step": 5539
+ },
+ {
+ "epoch": 17.69968051118211,
+ "grad_norm": 0.0966796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0625,
+ "step": 5540
+ },
+ {
+ "epoch": 17.702875399361023,
+ "grad_norm": 0.08544921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0496,
+ "step": 5541
+ },
+ {
+ "epoch": 17.706070287539937,
+ "grad_norm": 0.10302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0475,
+ "step": 5542
+ },
+ {
+ "epoch": 17.70926517571885,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0491,
+ "step": 5543
+ },
+ {
+ "epoch": 17.712460063897765,
+ "grad_norm": 0.14453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0438,
+ "step": 5544
+ },
+ {
+ "epoch": 17.715654952076676,
+ "grad_norm": 0.083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0495,
+ "step": 5545
+ },
+ {
+ "epoch": 17.71884984025559,
+ "grad_norm": 0.232421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0566,
+ "step": 5546
+ },
+ {
+ "epoch": 17.722044728434504,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0467,
+ "step": 5547
+ },
+ {
+ "epoch": 17.72523961661342,
+ "grad_norm": 0.1318359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0376,
+ "step": 5548
+ },
+ {
+ "epoch": 17.728434504792332,
+ "grad_norm": 0.050048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0588,
+ "step": 5549
+ },
+ {
+ "epoch": 17.731629392971247,
+ "grad_norm": 0.126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0456,
+ "step": 5550
+ },
+ {
+ "epoch": 17.73482428115016,
+ "grad_norm": 0.10400390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0468,
+ "step": 5551
+ },
+ {
+ "epoch": 17.738019169329075,
+ "grad_norm": 0.09130859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0576,
+ "step": 5552
+ },
+ {
+ "epoch": 17.74121405750799,
+ "grad_norm": 0.1376953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0525,
+ "step": 5553
+ },
+ {
+ "epoch": 17.7444089456869,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0537,
+ "step": 5554
+ },
+ {
+ "epoch": 17.747603833865814,
+ "grad_norm": 0.1533203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0425,
+ "step": 5555
+ },
+ {
+ "epoch": 17.750798722044728,
+ "grad_norm": 0.07861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0485,
+ "step": 5556
+ },
+ {
+ "epoch": 17.75399361022364,
+ "grad_norm": 0.12158203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0533,
+ "step": 5557
+ },
+ {
+ "epoch": 17.757188498402556,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0622,
+ "step": 5558
+ },
+ {
+ "epoch": 17.76038338658147,
+ "grad_norm": 0.12353515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0464,
+ "step": 5559
+ },
+ {
+ "epoch": 17.763578274760384,
+ "grad_norm": 0.08740234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0419,
+ "step": 5560
+ },
+ {
+ "epoch": 17.766773162939298,
+ "grad_norm": 0.15234375,
+ "learning_rate": 0.0005,
+ "loss": 1.054,
+ "step": 5561
+ },
+ {
+ "epoch": 17.769968051118212,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0523,
+ "step": 5562
+ },
+ {
+ "epoch": 17.773162939297123,
+ "grad_norm": 0.107421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0619,
+ "step": 5563
+ },
+ {
+ "epoch": 17.776357827476037,
+ "grad_norm": 0.08056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0504,
+ "step": 5564
+ },
+ {
+ "epoch": 17.77955271565495,
+ "grad_norm": 0.08251953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0515,
+ "step": 5565
+ },
+ {
+ "epoch": 17.782747603833865,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0461,
+ "step": 5566
+ },
+ {
+ "epoch": 17.78594249201278,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0468,
+ "step": 5567
+ },
+ {
+ "epoch": 17.789137380191693,
+ "grad_norm": 0.08984375,
+ "learning_rate": 0.0005,
+ "loss": 1.055,
+ "step": 5568
+ },
+ {
+ "epoch": 17.792332268370608,
+ "grad_norm": 0.05517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0395,
+ "step": 5569
+ },
+ {
+ "epoch": 17.79552715654952,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0396,
+ "step": 5570
+ },
+ {
+ "epoch": 17.798722044728436,
+ "grad_norm": 0.057861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.052,
+ "step": 5571
+ },
+ {
+ "epoch": 17.80191693290735,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.047,
+ "step": 5572
+ },
+ {
+ "epoch": 17.80511182108626,
+ "grad_norm": 0.1103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0375,
+ "step": 5573
+ },
+ {
+ "epoch": 17.808306709265175,
+ "grad_norm": 0.09716796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0574,
+ "step": 5574
+ },
+ {
+ "epoch": 17.81150159744409,
+ "grad_norm": 0.07861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.042,
+ "step": 5575
+ },
+ {
+ "epoch": 17.814696485623003,
+ "grad_norm": 0.047607421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0518,
+ "step": 5576
+ },
+ {
+ "epoch": 17.817891373801917,
+ "grad_norm": 0.083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0446,
+ "step": 5577
+ },
+ {
+ "epoch": 17.82108626198083,
+ "grad_norm": 0.053955078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0552,
+ "step": 5578
+ },
+ {
+ "epoch": 17.824281150159745,
+ "grad_norm": 0.05615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.059,
+ "step": 5579
+ },
+ {
+ "epoch": 17.82747603833866,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0606,
+ "step": 5580
+ },
+ {
+ "epoch": 17.830670926517573,
+ "grad_norm": 0.048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0619,
+ "step": 5581
+ },
+ {
+ "epoch": 17.833865814696484,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0441,
+ "step": 5582
+ },
+ {
+ "epoch": 17.837060702875398,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.051,
+ "step": 5583
+ },
+ {
+ "epoch": 17.840255591054312,
+ "grad_norm": 0.1357421875,
+ "learning_rate": 0.0005,
+ "loss": 1.045,
+ "step": 5584
+ },
+ {
+ "epoch": 17.843450479233226,
+ "grad_norm": 0.052978515625,
+ "learning_rate": 0.0005,
+ "loss": 1.045,
+ "step": 5585
+ },
+ {
+ "epoch": 17.84664536741214,
+ "grad_norm": 0.13671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0625,
+ "step": 5586
+ },
+ {
+ "epoch": 17.849840255591054,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0496,
+ "step": 5587
+ },
+ {
+ "epoch": 17.85303514376997,
+ "grad_norm": 0.1376953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0437,
+ "step": 5588
+ },
+ {
+ "epoch": 17.856230031948883,
+ "grad_norm": 0.056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0397,
+ "step": 5589
+ },
+ {
+ "epoch": 17.859424920127797,
+ "grad_norm": 0.12060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0617,
+ "step": 5590
+ },
+ {
+ "epoch": 17.86261980830671,
+ "grad_norm": 0.054443359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0457,
+ "step": 5591
+ },
+ {
+ "epoch": 17.86581469648562,
+ "grad_norm": 0.1181640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0517,
+ "step": 5592
+ },
+ {
+ "epoch": 17.869009584664536,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0534,
+ "step": 5593
+ },
+ {
+ "epoch": 17.87220447284345,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0501,
+ "step": 5594
+ },
+ {
+ "epoch": 17.875399361022364,
+ "grad_norm": 0.060791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0518,
+ "step": 5595
+ },
+ {
+ "epoch": 17.878594249201278,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0396,
+ "step": 5596
+ },
+ {
+ "epoch": 17.881789137380192,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0532,
+ "step": 5597
+ },
+ {
+ "epoch": 17.884984025559106,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0477,
+ "step": 5598
+ },
+ {
+ "epoch": 17.88817891373802,
+ "grad_norm": 0.049560546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0404,
+ "step": 5599
+ },
+ {
+ "epoch": 17.891373801916934,
+ "grad_norm": 0.091796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0425,
+ "step": 5600
+ },
+ {
+ "epoch": 17.894568690095845,
+ "grad_norm": 0.052734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0417,
+ "step": 5601
+ },
+ {
+ "epoch": 17.89776357827476,
+ "grad_norm": 0.060302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0468,
+ "step": 5602
+ },
+ {
+ "epoch": 17.900958466453673,
+ "grad_norm": 0.052490234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0356,
+ "step": 5603
+ },
+ {
+ "epoch": 17.904153354632587,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0561,
+ "step": 5604
+ },
+ {
+ "epoch": 17.9073482428115,
+ "grad_norm": 0.0810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0484,
+ "step": 5605
+ },
+ {
+ "epoch": 17.910543130990416,
+ "grad_norm": 0.0615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0351,
+ "step": 5606
+ },
+ {
+ "epoch": 17.91373801916933,
+ "grad_norm": 0.08154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0611,
+ "step": 5607
+ },
+ {
+ "epoch": 17.916932907348244,
+ "grad_norm": 0.10595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0512,
+ "step": 5608
+ },
+ {
+ "epoch": 17.920127795527158,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0477,
+ "step": 5609
+ },
+ {
+ "epoch": 17.923322683706072,
+ "grad_norm": 0.1201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0508,
+ "step": 5610
+ },
+ {
+ "epoch": 17.926517571884983,
+ "grad_norm": 0.06201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0465,
+ "step": 5611
+ },
+ {
+ "epoch": 17.929712460063897,
+ "grad_norm": 0.1171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0517,
+ "step": 5612
+ },
+ {
+ "epoch": 17.93290734824281,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0572,
+ "step": 5613
+ },
+ {
+ "epoch": 17.936102236421725,
+ "grad_norm": 0.099609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0388,
+ "step": 5614
+ },
+ {
+ "epoch": 17.93929712460064,
+ "grad_norm": 0.099609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0514,
+ "step": 5615
+ },
+ {
+ "epoch": 17.942492012779553,
+ "grad_norm": 0.11669921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0554,
+ "step": 5616
+ },
+ {
+ "epoch": 17.945686900958467,
+ "grad_norm": 0.054443359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0487,
+ "step": 5617
+ },
+ {
+ "epoch": 17.94888178913738,
+ "grad_norm": 0.09765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0432,
+ "step": 5618
+ },
+ {
+ "epoch": 17.952076677316295,
+ "grad_norm": 0.058837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0584,
+ "step": 5619
+ },
+ {
+ "epoch": 17.955271565495206,
+ "grad_norm": 0.1318359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0509,
+ "step": 5620
+ },
+ {
+ "epoch": 17.95846645367412,
+ "grad_norm": 0.052734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0526,
+ "step": 5621
+ },
+ {
+ "epoch": 17.961661341853034,
+ "grad_norm": 0.10400390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0468,
+ "step": 5622
+ },
+ {
+ "epoch": 17.96485623003195,
+ "grad_norm": 0.055419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0484,
+ "step": 5623
+ },
+ {
+ "epoch": 17.968051118210862,
+ "grad_norm": 0.08154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.046,
+ "step": 5624
+ },
+ {
+ "epoch": 17.971246006389777,
+ "grad_norm": 0.056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0501,
+ "step": 5625
+ },
+ {
+ "epoch": 17.97444089456869,
+ "grad_norm": 0.103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0461,
+ "step": 5626
+ },
+ {
+ "epoch": 17.977635782747605,
+ "grad_norm": 0.061279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0424,
+ "step": 5627
+ },
+ {
+ "epoch": 17.98083067092652,
+ "grad_norm": 0.11474609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0566,
+ "step": 5628
+ },
+ {
+ "epoch": 17.984025559105433,
+ "grad_norm": 0.126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0544,
+ "step": 5629
+ },
+ {
+ "epoch": 17.987220447284344,
+ "grad_norm": 0.095703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0399,
+ "step": 5630
+ },
+ {
+ "epoch": 17.990415335463258,
+ "grad_norm": 0.220703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0483,
+ "step": 5631
+ },
+ {
+ "epoch": 17.99361022364217,
+ "grad_norm": 0.08837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.04,
+ "step": 5632
+ },
+ {
+ "epoch": 17.996805111821086,
+ "grad_norm": 0.1484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0447,
+ "step": 5633
+ },
+ {
+ "epoch": 18.0,
+ "grad_norm": 0.1201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0686,
+ "step": 5634
+ },
+ {
+ "epoch": 18.003194888178914,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0397,
+ "step": 5635
+ },
+ {
+ "epoch": 18.00638977635783,
+ "grad_norm": 0.12060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0353,
+ "step": 5636
+ },
+ {
+ "epoch": 18.009584664536742,
+ "grad_norm": 0.056884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0515,
+ "step": 5637
+ },
+ {
+ "epoch": 18.012779552715656,
+ "grad_norm": 0.09326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0576,
+ "step": 5638
+ },
+ {
+ "epoch": 18.015974440894567,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0388,
+ "step": 5639
+ },
+ {
+ "epoch": 18.01916932907348,
+ "grad_norm": 0.1044921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0512,
+ "step": 5640
+ },
+ {
+ "epoch": 18.022364217252395,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0482,
+ "step": 5641
+ },
+ {
+ "epoch": 18.02555910543131,
+ "grad_norm": 0.11767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0558,
+ "step": 5642
+ },
+ {
+ "epoch": 18.028753993610223,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0585,
+ "step": 5643
+ },
+ {
+ "epoch": 18.031948881789138,
+ "grad_norm": 0.10009765625,
+ "learning_rate": 0.0005,
+ "loss": 1.046,
+ "step": 5644
+ },
+ {
+ "epoch": 18.03514376996805,
+ "grad_norm": 0.050048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0533,
+ "step": 5645
+ },
+ {
+ "epoch": 18.038338658146966,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0553,
+ "step": 5646
+ },
+ {
+ "epoch": 18.04153354632588,
+ "grad_norm": 0.058349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0496,
+ "step": 5647
+ },
+ {
+ "epoch": 18.044728434504794,
+ "grad_norm": 0.058349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0543,
+ "step": 5648
+ },
+ {
+ "epoch": 18.047923322683705,
+ "grad_norm": 0.05517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0492,
+ "step": 5649
+ },
+ {
+ "epoch": 18.05111821086262,
+ "grad_norm": 0.0556640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0566,
+ "step": 5650
+ },
+ {
+ "epoch": 18.054313099041533,
+ "grad_norm": 0.05029296875,
+ "learning_rate": 0.0005,
+ "loss": 1.042,
+ "step": 5651
+ },
+ {
+ "epoch": 18.057507987220447,
+ "grad_norm": 0.0556640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0521,
+ "step": 5652
+ },
+ {
+ "epoch": 18.06070287539936,
+ "grad_norm": 0.0498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0394,
+ "step": 5653
+ },
+ {
+ "epoch": 18.063897763578275,
+ "grad_norm": 0.05078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0401,
+ "step": 5654
+ },
+ {
+ "epoch": 18.06709265175719,
+ "grad_norm": 0.05419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0511,
+ "step": 5655
+ },
+ {
+ "epoch": 18.070287539936103,
+ "grad_norm": 0.08984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0477,
+ "step": 5656
+ },
+ {
+ "epoch": 18.073482428115017,
+ "grad_norm": 0.05419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.056,
+ "step": 5657
+ },
+ {
+ "epoch": 18.076677316293928,
+ "grad_norm": 0.052490234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0431,
+ "step": 5658
+ },
+ {
+ "epoch": 18.079872204472842,
+ "grad_norm": 0.061279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0483,
+ "step": 5659
+ },
+ {
+ "epoch": 18.083067092651756,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0394,
+ "step": 5660
+ },
+ {
+ "epoch": 18.08626198083067,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0497,
+ "step": 5661
+ },
+ {
+ "epoch": 18.089456869009584,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0582,
+ "step": 5662
+ },
+ {
+ "epoch": 18.0926517571885,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0376,
+ "step": 5663
+ },
+ {
+ "epoch": 18.095846645367413,
+ "grad_norm": 0.0830078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0557,
+ "step": 5664
+ },
+ {
+ "epoch": 18.099041533546327,
+ "grad_norm": 0.08935546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0509,
+ "step": 5665
+ },
+ {
+ "epoch": 18.10223642172524,
+ "grad_norm": 0.10693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0484,
+ "step": 5666
+ },
+ {
+ "epoch": 18.105431309904155,
+ "grad_norm": 0.09375,
+ "learning_rate": 0.0005,
+ "loss": 1.0525,
+ "step": 5667
+ },
+ {
+ "epoch": 18.108626198083066,
+ "grad_norm": 0.1025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0454,
+ "step": 5668
+ },
+ {
+ "epoch": 18.11182108626198,
+ "grad_norm": 0.109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0449,
+ "step": 5669
+ },
+ {
+ "epoch": 18.115015974440894,
+ "grad_norm": 0.1396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0463,
+ "step": 5670
+ },
+ {
+ "epoch": 18.118210862619808,
+ "grad_norm": 0.0595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.049,
+ "step": 5671
+ },
+ {
+ "epoch": 18.121405750798722,
+ "grad_norm": 0.103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0564,
+ "step": 5672
+ },
+ {
+ "epoch": 18.124600638977636,
+ "grad_norm": 0.0556640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0456,
+ "step": 5673
+ },
+ {
+ "epoch": 18.12779552715655,
+ "grad_norm": 0.1298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0437,
+ "step": 5674
+ },
+ {
+ "epoch": 18.130990415335464,
+ "grad_norm": 0.0927734375,
+ "learning_rate": 0.0005,
+ "loss": 1.055,
+ "step": 5675
+ },
+ {
+ "epoch": 18.13418530351438,
+ "grad_norm": 0.0986328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0425,
+ "step": 5676
+ },
+ {
+ "epoch": 18.13738019169329,
+ "grad_norm": 0.130859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0461,
+ "step": 5677
+ },
+ {
+ "epoch": 18.140575079872203,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.05,
+ "step": 5678
+ },
+ {
+ "epoch": 18.143769968051117,
+ "grad_norm": 0.1533203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0558,
+ "step": 5679
+ },
+ {
+ "epoch": 18.14696485623003,
+ "grad_norm": 0.0478515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0356,
+ "step": 5680
+ },
+ {
+ "epoch": 18.150159744408946,
+ "grad_norm": 0.1435546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0405,
+ "step": 5681
+ },
+ {
+ "epoch": 18.15335463258786,
+ "grad_norm": 0.0517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0569,
+ "step": 5682
+ },
+ {
+ "epoch": 18.156549520766774,
+ "grad_norm": 0.142578125,
+ "learning_rate": 0.0005,
+ "loss": 1.046,
+ "step": 5683
+ },
+ {
+ "epoch": 18.159744408945688,
+ "grad_norm": 0.08837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0443,
+ "step": 5684
+ },
+ {
+ "epoch": 18.162939297124602,
+ "grad_norm": 0.126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0387,
+ "step": 5685
+ },
+ {
+ "epoch": 18.166134185303516,
+ "grad_norm": 0.1318359375,
+ "learning_rate": 0.0005,
+ "loss": 1.059,
+ "step": 5686
+ },
+ {
+ "epoch": 18.169329073482427,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0556,
+ "step": 5687
+ },
+ {
+ "epoch": 18.17252396166134,
+ "grad_norm": 0.08251953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0467,
+ "step": 5688
+ },
+ {
+ "epoch": 18.175718849840255,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0477,
+ "step": 5689
+ },
+ {
+ "epoch": 18.17891373801917,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0483,
+ "step": 5690
+ },
+ {
+ "epoch": 18.182108626198083,
+ "grad_norm": 0.055419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0572,
+ "step": 5691
+ },
+ {
+ "epoch": 18.185303514376997,
+ "grad_norm": 0.080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.038,
+ "step": 5692
+ },
+ {
+ "epoch": 18.18849840255591,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0467,
+ "step": 5693
+ },
+ {
+ "epoch": 18.191693290734825,
+ "grad_norm": 0.08154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0443,
+ "step": 5694
+ },
+ {
+ "epoch": 18.19488817891374,
+ "grad_norm": 0.1181640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0602,
+ "step": 5695
+ },
+ {
+ "epoch": 18.19808306709265,
+ "grad_norm": 0.1572265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0376,
+ "step": 5696
+ },
+ {
+ "epoch": 18.201277955271564,
+ "grad_norm": 0.07861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0495,
+ "step": 5697
+ },
+ {
+ "epoch": 18.20447284345048,
+ "grad_norm": 0.08935546875,
+ "learning_rate": 0.0005,
+ "loss": 1.042,
+ "step": 5698
+ },
+ {
+ "epoch": 18.207667731629392,
+ "grad_norm": 0.08935546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0461,
+ "step": 5699
+ },
+ {
+ "epoch": 18.210862619808307,
+ "grad_norm": 0.08740234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0542,
+ "step": 5700
+ },
+ {
+ "epoch": 18.21405750798722,
+ "grad_norm": 0.04931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0548,
+ "step": 5701
+ },
+ {
+ "epoch": 18.217252396166135,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0537,
+ "step": 5702
+ },
+ {
+ "epoch": 18.22044728434505,
+ "grad_norm": 0.057373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0491,
+ "step": 5703
+ },
+ {
+ "epoch": 18.223642172523963,
+ "grad_norm": 0.05029296875,
+ "learning_rate": 0.0005,
+ "loss": 1.056,
+ "step": 5704
+ },
+ {
+ "epoch": 18.226837060702877,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0445,
+ "step": 5705
+ },
+ {
+ "epoch": 18.230031948881788,
+ "grad_norm": 0.0595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0495,
+ "step": 5706
+ },
+ {
+ "epoch": 18.233226837060702,
+ "grad_norm": 0.060791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0432,
+ "step": 5707
+ },
+ {
+ "epoch": 18.236421725239616,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0583,
+ "step": 5708
+ },
+ {
+ "epoch": 18.23961661341853,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0498,
+ "step": 5709
+ },
+ {
+ "epoch": 18.242811501597444,
+ "grad_norm": 0.05517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0543,
+ "step": 5710
+ },
+ {
+ "epoch": 18.24600638977636,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0528,
+ "step": 5711
+ },
+ {
+ "epoch": 18.249201277955272,
+ "grad_norm": 0.05615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0527,
+ "step": 5712
+ },
+ {
+ "epoch": 18.252396166134186,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0491,
+ "step": 5713
+ },
+ {
+ "epoch": 18.2555910543131,
+ "grad_norm": 0.1015625,
+ "learning_rate": 0.0005,
+ "loss": 1.027,
+ "step": 5714
+ },
+ {
+ "epoch": 18.25878594249201,
+ "grad_norm": 0.125,
+ "learning_rate": 0.0005,
+ "loss": 1.047,
+ "step": 5715
+ },
+ {
+ "epoch": 18.261980830670925,
+ "grad_norm": 0.13671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0547,
+ "step": 5716
+ },
+ {
+ "epoch": 18.26517571884984,
+ "grad_norm": 0.1650390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0584,
+ "step": 5717
+ },
+ {
+ "epoch": 18.268370607028753,
+ "grad_norm": 0.119140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0523,
+ "step": 5718
+ },
+ {
+ "epoch": 18.271565495207668,
+ "grad_norm": 0.10888671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0517,
+ "step": 5719
+ },
+ {
+ "epoch": 18.27476038338658,
+ "grad_norm": 0.12890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0492,
+ "step": 5720
+ },
+ {
+ "epoch": 18.277955271565496,
+ "grad_norm": 0.05859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0496,
+ "step": 5721
+ },
+ {
+ "epoch": 18.28115015974441,
+ "grad_norm": 0.083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0377,
+ "step": 5722
+ },
+ {
+ "epoch": 18.284345047923324,
+ "grad_norm": 0.0908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0448,
+ "step": 5723
+ },
+ {
+ "epoch": 18.287539936102238,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0441,
+ "step": 5724
+ },
+ {
+ "epoch": 18.29073482428115,
+ "grad_norm": 0.0986328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0512,
+ "step": 5725
+ },
+ {
+ "epoch": 18.293929712460063,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0448,
+ "step": 5726
+ },
+ {
+ "epoch": 18.297124600638977,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0438,
+ "step": 5727
+ },
+ {
+ "epoch": 18.30031948881789,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0595,
+ "step": 5728
+ },
+ {
+ "epoch": 18.303514376996805,
+ "grad_norm": 0.054443359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0353,
+ "step": 5729
+ },
+ {
+ "epoch": 18.30670926517572,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0521,
+ "step": 5730
+ },
+ {
+ "epoch": 18.309904153354633,
+ "grad_norm": 0.059814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0411,
+ "step": 5731
+ },
+ {
+ "epoch": 18.313099041533548,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0553,
+ "step": 5732
+ },
+ {
+ "epoch": 18.31629392971246,
+ "grad_norm": 0.1064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.059,
+ "step": 5733
+ },
+ {
+ "epoch": 18.319488817891372,
+ "grad_norm": 0.080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0538,
+ "step": 5734
+ },
+ {
+ "epoch": 18.322683706070286,
+ "grad_norm": 0.1337890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0573,
+ "step": 5735
+ },
+ {
+ "epoch": 18.3258785942492,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0452,
+ "step": 5736
+ },
+ {
+ "epoch": 18.329073482428115,
+ "grad_norm": 0.095703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0578,
+ "step": 5737
+ },
+ {
+ "epoch": 18.33226837060703,
+ "grad_norm": 0.05615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0473,
+ "step": 5738
+ },
+ {
+ "epoch": 18.335463258785943,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0447,
+ "step": 5739
+ },
+ {
+ "epoch": 18.338658146964857,
+ "grad_norm": 0.08447265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0457,
+ "step": 5740
+ },
+ {
+ "epoch": 18.34185303514377,
+ "grad_norm": 0.1064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.048,
+ "step": 5741
+ },
+ {
+ "epoch": 18.345047923322685,
+ "grad_norm": 0.080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0471,
+ "step": 5742
+ },
+ {
+ "epoch": 18.3482428115016,
+ "grad_norm": 0.1484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0475,
+ "step": 5743
+ },
+ {
+ "epoch": 18.35143769968051,
+ "grad_norm": 0.05859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0495,
+ "step": 5744
+ },
+ {
+ "epoch": 18.354632587859424,
+ "grad_norm": 0.1640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0502,
+ "step": 5745
+ },
+ {
+ "epoch": 18.357827476038338,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0586,
+ "step": 5746
+ },
+ {
+ "epoch": 18.361022364217252,
+ "grad_norm": 0.11083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0563,
+ "step": 5747
+ },
+ {
+ "epoch": 18.364217252396166,
+ "grad_norm": 0.09814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0575,
+ "step": 5748
+ },
+ {
+ "epoch": 18.36741214057508,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0474,
+ "step": 5749
+ },
+ {
+ "epoch": 18.370607028753994,
+ "grad_norm": 0.1181640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0362,
+ "step": 5750
+ },
+ {
+ "epoch": 18.37380191693291,
+ "grad_norm": 0.08447265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0483,
+ "step": 5751
+ },
+ {
+ "epoch": 18.376996805111823,
+ "grad_norm": 0.2099609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0382,
+ "step": 5752
+ },
+ {
+ "epoch": 18.380191693290733,
+ "grad_norm": 0.140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0519,
+ "step": 5753
+ },
+ {
+ "epoch": 18.383386581469647,
+ "grad_norm": 0.15234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0545,
+ "step": 5754
+ },
+ {
+ "epoch": 18.38658146964856,
+ "grad_norm": 0.248046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0487,
+ "step": 5755
+ },
+ {
+ "epoch": 18.389776357827476,
+ "grad_norm": 0.0849609375,
+ "learning_rate": 0.0005,
+ "loss": 1.037,
+ "step": 5756
+ },
+ {
+ "epoch": 18.39297124600639,
+ "grad_norm": 0.1572265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0442,
+ "step": 5757
+ },
+ {
+ "epoch": 18.396166134185304,
+ "grad_norm": 0.09228515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0579,
+ "step": 5758
+ },
+ {
+ "epoch": 18.399361022364218,
+ "grad_norm": 0.10302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0521,
+ "step": 5759
+ },
+ {
+ "epoch": 18.402555910543132,
+ "grad_norm": 0.08935546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0551,
+ "step": 5760
+ },
+ {
+ "epoch": 18.405750798722046,
+ "grad_norm": 0.10205078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0501,
+ "step": 5761
+ },
+ {
+ "epoch": 18.408945686900957,
+ "grad_norm": 0.18359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0504,
+ "step": 5762
+ },
+ {
+ "epoch": 18.41214057507987,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0402,
+ "step": 5763
+ },
+ {
+ "epoch": 18.415335463258785,
+ "grad_norm": 0.126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0506,
+ "step": 5764
+ },
+ {
+ "epoch": 18.4185303514377,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.031,
+ "step": 5765
+ },
+ {
+ "epoch": 18.421725239616613,
+ "grad_norm": 0.1669921875,
+ "learning_rate": 0.0005,
+ "loss": 1.061,
+ "step": 5766
+ },
+ {
+ "epoch": 18.424920127795527,
+ "grad_norm": 0.10009765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0363,
+ "step": 5767
+ },
+ {
+ "epoch": 18.42811501597444,
+ "grad_norm": 0.08837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0539,
+ "step": 5768
+ },
+ {
+ "epoch": 18.431309904153355,
+ "grad_norm": 0.09716796875,
+ "learning_rate": 0.0005,
+ "loss": 1.048,
+ "step": 5769
+ },
+ {
+ "epoch": 18.43450479233227,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0453,
+ "step": 5770
+ },
+ {
+ "epoch": 18.437699680511184,
+ "grad_norm": 0.1376953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0442,
+ "step": 5771
+ },
+ {
+ "epoch": 18.440894568690094,
+ "grad_norm": 0.0615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0575,
+ "step": 5772
+ },
+ {
+ "epoch": 18.44408945686901,
+ "grad_norm": 0.1142578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0477,
+ "step": 5773
+ },
+ {
+ "epoch": 18.447284345047922,
+ "grad_norm": 0.08349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0374,
+ "step": 5774
+ },
+ {
+ "epoch": 18.450479233226837,
+ "grad_norm": 0.1982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0612,
+ "step": 5775
+ },
+ {
+ "epoch": 18.45367412140575,
+ "grad_norm": 0.1298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0341,
+ "step": 5776
+ },
+ {
+ "epoch": 18.456869009584665,
+ "grad_norm": 0.1357421875,
+ "learning_rate": 0.0005,
+ "loss": 1.055,
+ "step": 5777
+ },
+ {
+ "epoch": 18.46006389776358,
+ "grad_norm": 0.2275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.044,
+ "step": 5778
+ },
+ {
+ "epoch": 18.463258785942493,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0431,
+ "step": 5779
+ },
+ {
+ "epoch": 18.466453674121407,
+ "grad_norm": 0.2138671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0567,
+ "step": 5780
+ },
+ {
+ "epoch": 18.46964856230032,
+ "grad_norm": 0.0859375,
+ "learning_rate": 0.0005,
+ "loss": 1.056,
+ "step": 5781
+ },
+ {
+ "epoch": 18.472843450479232,
+ "grad_norm": 0.11376953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0412,
+ "step": 5782
+ },
+ {
+ "epoch": 18.476038338658146,
+ "grad_norm": 0.05859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0413,
+ "step": 5783
+ },
+ {
+ "epoch": 18.47923322683706,
+ "grad_norm": 0.080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0313,
+ "step": 5784
+ },
+ {
+ "epoch": 18.482428115015974,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0434,
+ "step": 5785
+ },
+ {
+ "epoch": 18.48562300319489,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0413,
+ "step": 5786
+ },
+ {
+ "epoch": 18.488817891373802,
+ "grad_norm": 0.06103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0456,
+ "step": 5787
+ },
+ {
+ "epoch": 18.492012779552716,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0441,
+ "step": 5788
+ },
+ {
+ "epoch": 18.49520766773163,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0464,
+ "step": 5789
+ },
+ {
+ "epoch": 18.498402555910545,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0588,
+ "step": 5790
+ },
+ {
+ "epoch": 18.501597444089455,
+ "grad_norm": 0.05615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0531,
+ "step": 5791
+ },
+ {
+ "epoch": 18.50479233226837,
+ "grad_norm": 0.083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0463,
+ "step": 5792
+ },
+ {
+ "epoch": 18.507987220447284,
+ "grad_norm": 0.056396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0488,
+ "step": 5793
+ },
+ {
+ "epoch": 18.511182108626198,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0406,
+ "step": 5794
+ },
+ {
+ "epoch": 18.51437699680511,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0483,
+ "step": 5795
+ },
+ {
+ "epoch": 18.517571884984026,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.047,
+ "step": 5796
+ },
+ {
+ "epoch": 18.52076677316294,
+ "grad_norm": 0.054931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0564,
+ "step": 5797
+ },
+ {
+ "epoch": 18.523961661341854,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.047,
+ "step": 5798
+ },
+ {
+ "epoch": 18.527156549520768,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0565,
+ "step": 5799
+ },
+ {
+ "epoch": 18.53035143769968,
+ "grad_norm": 0.054931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0498,
+ "step": 5800
+ },
+ {
+ "epoch": 18.533546325878593,
+ "grad_norm": 0.057861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0392,
+ "step": 5801
+ },
+ {
+ "epoch": 18.536741214057507,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0412,
+ "step": 5802
+ },
+ {
+ "epoch": 18.53993610223642,
+ "grad_norm": 0.0927734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0393,
+ "step": 5803
+ },
+ {
+ "epoch": 18.543130990415335,
+ "grad_norm": 0.0595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0627,
+ "step": 5804
+ },
+ {
+ "epoch": 18.54632587859425,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0404,
+ "step": 5805
+ },
+ {
+ "epoch": 18.549520766773163,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0548,
+ "step": 5806
+ },
+ {
+ "epoch": 18.552715654952078,
+ "grad_norm": 0.09423828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0478,
+ "step": 5807
+ },
+ {
+ "epoch": 18.55591054313099,
+ "grad_norm": 0.052001953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0501,
+ "step": 5808
+ },
+ {
+ "epoch": 18.559105431309906,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0578,
+ "step": 5809
+ },
+ {
+ "epoch": 18.562300319488816,
+ "grad_norm": 0.053955078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0371,
+ "step": 5810
+ },
+ {
+ "epoch": 18.56549520766773,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0463,
+ "step": 5811
+ },
+ {
+ "epoch": 18.568690095846645,
+ "grad_norm": 0.0537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.055,
+ "step": 5812
+ },
+ {
+ "epoch": 18.57188498402556,
+ "grad_norm": 0.05859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0345,
+ "step": 5813
+ },
+ {
+ "epoch": 18.575079872204473,
+ "grad_norm": 0.05615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0462,
+ "step": 5814
+ },
+ {
+ "epoch": 18.578274760383387,
+ "grad_norm": 0.099609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0427,
+ "step": 5815
+ },
+ {
+ "epoch": 18.5814696485623,
+ "grad_norm": 0.125,
+ "learning_rate": 0.0005,
+ "loss": 1.0446,
+ "step": 5816
+ },
+ {
+ "epoch": 18.584664536741215,
+ "grad_norm": 0.08740234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0528,
+ "step": 5817
+ },
+ {
+ "epoch": 18.58785942492013,
+ "grad_norm": 0.1455078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0503,
+ "step": 5818
+ },
+ {
+ "epoch": 18.591054313099043,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0544,
+ "step": 5819
+ },
+ {
+ "epoch": 18.594249201277954,
+ "grad_norm": 0.10693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0508,
+ "step": 5820
+ },
+ {
+ "epoch": 18.597444089456868,
+ "grad_norm": 0.08740234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0517,
+ "step": 5821
+ },
+ {
+ "epoch": 18.600638977635782,
+ "grad_norm": 0.1455078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0436,
+ "step": 5822
+ },
+ {
+ "epoch": 18.603833865814696,
+ "grad_norm": 0.052490234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0413,
+ "step": 5823
+ },
+ {
+ "epoch": 18.60702875399361,
+ "grad_norm": 0.11669921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0443,
+ "step": 5824
+ },
+ {
+ "epoch": 18.610223642172524,
+ "grad_norm": 0.0595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0468,
+ "step": 5825
+ },
+ {
+ "epoch": 18.61341853035144,
+ "grad_norm": 0.1064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0379,
+ "step": 5826
+ },
+ {
+ "epoch": 18.616613418530353,
+ "grad_norm": 0.080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0406,
+ "step": 5827
+ },
+ {
+ "epoch": 18.619808306709267,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0504,
+ "step": 5828
+ },
+ {
+ "epoch": 18.623003194888177,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.048,
+ "step": 5829
+ },
+ {
+ "epoch": 18.62619808306709,
+ "grad_norm": 0.0615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0416,
+ "step": 5830
+ },
+ {
+ "epoch": 18.629392971246006,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0459,
+ "step": 5831
+ },
+ {
+ "epoch": 18.63258785942492,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0508,
+ "step": 5832
+ },
+ {
+ "epoch": 18.635782747603834,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0359,
+ "step": 5833
+ },
+ {
+ "epoch": 18.638977635782748,
+ "grad_norm": 0.083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0583,
+ "step": 5834
+ },
+ {
+ "epoch": 18.642172523961662,
+ "grad_norm": 0.09716796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0383,
+ "step": 5835
+ },
+ {
+ "epoch": 18.645367412140576,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0395,
+ "step": 5836
+ },
+ {
+ "epoch": 18.64856230031949,
+ "grad_norm": 0.0927734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0435,
+ "step": 5837
+ },
+ {
+ "epoch": 18.6517571884984,
+ "grad_norm": 0.05712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0501,
+ "step": 5838
+ },
+ {
+ "epoch": 18.654952076677315,
+ "grad_norm": 0.053466796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0595,
+ "step": 5839
+ },
+ {
+ "epoch": 18.65814696485623,
+ "grad_norm": 0.08935546875,
+ "learning_rate": 0.0005,
+ "loss": 1.039,
+ "step": 5840
+ },
+ {
+ "epoch": 18.661341853035143,
+ "grad_norm": 0.05322265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0528,
+ "step": 5841
+ },
+ {
+ "epoch": 18.664536741214057,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0411,
+ "step": 5842
+ },
+ {
+ "epoch": 18.66773162939297,
+ "grad_norm": 0.05810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.059,
+ "step": 5843
+ },
+ {
+ "epoch": 18.670926517571885,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0529,
+ "step": 5844
+ },
+ {
+ "epoch": 18.6741214057508,
+ "grad_norm": 0.05908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0351,
+ "step": 5845
+ },
+ {
+ "epoch": 18.677316293929714,
+ "grad_norm": 0.08544921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0441,
+ "step": 5846
+ },
+ {
+ "epoch": 18.680511182108628,
+ "grad_norm": 0.0546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0418,
+ "step": 5847
+ },
+ {
+ "epoch": 18.68370607028754,
+ "grad_norm": 0.053466796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0607,
+ "step": 5848
+ },
+ {
+ "epoch": 18.686900958466452,
+ "grad_norm": 0.05224609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0537,
+ "step": 5849
+ },
+ {
+ "epoch": 18.690095846645367,
+ "grad_norm": 0.0595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0467,
+ "step": 5850
+ },
+ {
+ "epoch": 18.69329073482428,
+ "grad_norm": 0.06201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0475,
+ "step": 5851
+ },
+ {
+ "epoch": 18.696485623003195,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0441,
+ "step": 5852
+ },
+ {
+ "epoch": 18.69968051118211,
+ "grad_norm": 0.05810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0519,
+ "step": 5853
+ },
+ {
+ "epoch": 18.702875399361023,
+ "grad_norm": 0.08056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0449,
+ "step": 5854
+ },
+ {
+ "epoch": 18.706070287539937,
+ "grad_norm": 0.07861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0381,
+ "step": 5855
+ },
+ {
+ "epoch": 18.70926517571885,
+ "grad_norm": 0.10595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0455,
+ "step": 5856
+ },
+ {
+ "epoch": 18.712460063897765,
+ "grad_norm": 0.0615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.053,
+ "step": 5857
+ },
+ {
+ "epoch": 18.715654952076676,
+ "grad_norm": 0.09521484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0394,
+ "step": 5858
+ },
+ {
+ "epoch": 18.71884984025559,
+ "grad_norm": 0.0517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0445,
+ "step": 5859
+ },
+ {
+ "epoch": 18.722044728434504,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0633,
+ "step": 5860
+ },
+ {
+ "epoch": 18.72523961661342,
+ "grad_norm": 0.054443359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0389,
+ "step": 5861
+ },
+ {
+ "epoch": 18.728434504792332,
+ "grad_norm": 0.052001953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0614,
+ "step": 5862
+ },
+ {
+ "epoch": 18.731629392971247,
+ "grad_norm": 0.05908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0497,
+ "step": 5863
+ },
+ {
+ "epoch": 18.73482428115016,
+ "grad_norm": 0.057373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0537,
+ "step": 5864
+ },
+ {
+ "epoch": 18.738019169329075,
+ "grad_norm": 0.052490234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0382,
+ "step": 5865
+ },
+ {
+ "epoch": 18.74121405750799,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0338,
+ "step": 5866
+ },
+ {
+ "epoch": 18.7444089456869,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.051,
+ "step": 5867
+ },
+ {
+ "epoch": 18.747603833865814,
+ "grad_norm": 0.05078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0558,
+ "step": 5868
+ },
+ {
+ "epoch": 18.750798722044728,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0523,
+ "step": 5869
+ },
+ {
+ "epoch": 18.75399361022364,
+ "grad_norm": 0.0478515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0477,
+ "step": 5870
+ },
+ {
+ "epoch": 18.757188498402556,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0628,
+ "step": 5871
+ },
+ {
+ "epoch": 18.76038338658147,
+ "grad_norm": 0.043212890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0436,
+ "step": 5872
+ },
+ {
+ "epoch": 18.763578274760384,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0428,
+ "step": 5873
+ },
+ {
+ "epoch": 18.766773162939298,
+ "grad_norm": 0.057373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0417,
+ "step": 5874
+ },
+ {
+ "epoch": 18.769968051118212,
+ "grad_norm": 0.0576171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0483,
+ "step": 5875
+ },
+ {
+ "epoch": 18.773162939297123,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0549,
+ "step": 5876
+ },
+ {
+ "epoch": 18.776357827476037,
+ "grad_norm": 0.049560546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0406,
+ "step": 5877
+ },
+ {
+ "epoch": 18.77955271565495,
+ "grad_norm": 0.0576171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0553,
+ "step": 5878
+ },
+ {
+ "epoch": 18.782747603833865,
+ "grad_norm": 0.05224609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0458,
+ "step": 5879
+ },
+ {
+ "epoch": 18.78594249201278,
+ "grad_norm": 0.06005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0525,
+ "step": 5880
+ },
+ {
+ "epoch": 18.789137380191693,
+ "grad_norm": 0.0615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.05,
+ "step": 5881
+ },
+ {
+ "epoch": 18.792332268370608,
+ "grad_norm": 0.0546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0482,
+ "step": 5882
+ },
+ {
+ "epoch": 18.79552715654952,
+ "grad_norm": 0.05908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0479,
+ "step": 5883
+ },
+ {
+ "epoch": 18.798722044728436,
+ "grad_norm": 0.062255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0409,
+ "step": 5884
+ },
+ {
+ "epoch": 18.80191693290735,
+ "grad_norm": 0.0595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0512,
+ "step": 5885
+ },
+ {
+ "epoch": 18.80511182108626,
+ "grad_norm": 0.061279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0484,
+ "step": 5886
+ },
+ {
+ "epoch": 18.808306709265175,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0418,
+ "step": 5887
+ },
+ {
+ "epoch": 18.81150159744409,
+ "grad_norm": 0.06103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0583,
+ "step": 5888
+ },
+ {
+ "epoch": 18.814696485623003,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0523,
+ "step": 5889
+ },
+ {
+ "epoch": 18.817891373801917,
+ "grad_norm": 0.09033203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0457,
+ "step": 5890
+ },
+ {
+ "epoch": 18.82108626198083,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0447,
+ "step": 5891
+ },
+ {
+ "epoch": 18.824281150159745,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0436,
+ "step": 5892
+ },
+ {
+ "epoch": 18.82747603833866,
+ "grad_norm": 0.0810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0484,
+ "step": 5893
+ },
+ {
+ "epoch": 18.830670926517573,
+ "grad_norm": 0.056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0415,
+ "step": 5894
+ },
+ {
+ "epoch": 18.833865814696484,
+ "grad_norm": 0.09716796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0575,
+ "step": 5895
+ },
+ {
+ "epoch": 18.837060702875398,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0573,
+ "step": 5896
+ },
+ {
+ "epoch": 18.840255591054312,
+ "grad_norm": 0.126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0471,
+ "step": 5897
+ },
+ {
+ "epoch": 18.843450479233226,
+ "grad_norm": 0.1904296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0316,
+ "step": 5898
+ },
+ {
+ "epoch": 18.84664536741214,
+ "grad_norm": 0.061279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0477,
+ "step": 5899
+ },
+ {
+ "epoch": 18.849840255591054,
+ "grad_norm": 0.107421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0602,
+ "step": 5900
+ },
+ {
+ "epoch": 18.85303514376997,
+ "grad_norm": 0.10205078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0472,
+ "step": 5901
+ },
+ {
+ "epoch": 18.856230031948883,
+ "grad_norm": 0.060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0482,
+ "step": 5902
+ },
+ {
+ "epoch": 18.859424920127797,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0424,
+ "step": 5903
+ },
+ {
+ "epoch": 18.86261980830671,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0405,
+ "step": 5904
+ },
+ {
+ "epoch": 18.86581469648562,
+ "grad_norm": 0.0556640625,
+ "learning_rate": 0.0005,
+ "loss": 1.047,
+ "step": 5905
+ },
+ {
+ "epoch": 18.869009584664536,
+ "grad_norm": 0.05615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.041,
+ "step": 5906
+ },
+ {
+ "epoch": 18.87220447284345,
+ "grad_norm": 0.057373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0493,
+ "step": 5907
+ },
+ {
+ "epoch": 18.875399361022364,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0536,
+ "step": 5908
+ },
+ {
+ "epoch": 18.878594249201278,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0475,
+ "step": 5909
+ },
+ {
+ "epoch": 18.881789137380192,
+ "grad_norm": 0.09033203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0369,
+ "step": 5910
+ },
+ {
+ "epoch": 18.884984025559106,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0476,
+ "step": 5911
+ },
+ {
+ "epoch": 18.88817891373802,
+ "grad_norm": 0.0888671875,
+ "learning_rate": 0.0005,
+ "loss": 1.052,
+ "step": 5912
+ },
+ {
+ "epoch": 18.891373801916934,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0375,
+ "step": 5913
+ },
+ {
+ "epoch": 18.894568690095845,
+ "grad_norm": 0.08935546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0586,
+ "step": 5914
+ },
+ {
+ "epoch": 18.89776357827476,
+ "grad_norm": 0.08935546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0416,
+ "step": 5915
+ },
+ {
+ "epoch": 18.900958466453673,
+ "grad_norm": 0.0849609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0492,
+ "step": 5916
+ },
+ {
+ "epoch": 18.904153354632587,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0461,
+ "step": 5917
+ },
+ {
+ "epoch": 18.9073482428115,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0357,
+ "step": 5918
+ },
+ {
+ "epoch": 18.910543130990416,
+ "grad_norm": 0.060791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0477,
+ "step": 5919
+ },
+ {
+ "epoch": 18.91373801916933,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0499,
+ "step": 5920
+ },
+ {
+ "epoch": 18.916932907348244,
+ "grad_norm": 0.058349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0475,
+ "step": 5921
+ },
+ {
+ "epoch": 18.920127795527158,
+ "grad_norm": 0.099609375,
+ "learning_rate": 0.0005,
+ "loss": 1.054,
+ "step": 5922
+ },
+ {
+ "epoch": 18.923322683706072,
+ "grad_norm": 0.103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0405,
+ "step": 5923
+ },
+ {
+ "epoch": 18.926517571884983,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0408,
+ "step": 5924
+ },
+ {
+ "epoch": 18.929712460063897,
+ "grad_norm": 0.1083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0434,
+ "step": 5925
+ },
+ {
+ "epoch": 18.93290734824281,
+ "grad_norm": 0.061279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0474,
+ "step": 5926
+ },
+ {
+ "epoch": 18.936102236421725,
+ "grad_norm": 0.09423828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0631,
+ "step": 5927
+ },
+ {
+ "epoch": 18.93929712460064,
+ "grad_norm": 0.11083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0594,
+ "step": 5928
+ },
+ {
+ "epoch": 18.942492012779553,
+ "grad_norm": 0.1962890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0545,
+ "step": 5929
+ },
+ {
+ "epoch": 18.945686900958467,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0567,
+ "step": 5930
+ },
+ {
+ "epoch": 18.94888178913738,
+ "grad_norm": 0.2080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0438,
+ "step": 5931
+ },
+ {
+ "epoch": 18.952076677316295,
+ "grad_norm": 0.11328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0471,
+ "step": 5932
+ },
+ {
+ "epoch": 18.955271565495206,
+ "grad_norm": 0.08154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0527,
+ "step": 5933
+ },
+ {
+ "epoch": 18.95846645367412,
+ "grad_norm": 0.1201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0373,
+ "step": 5934
+ },
+ {
+ "epoch": 18.961661341853034,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0526,
+ "step": 5935
+ },
+ {
+ "epoch": 18.96485623003195,
+ "grad_norm": 0.1142578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0464,
+ "step": 5936
+ },
+ {
+ "epoch": 18.968051118210862,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0434,
+ "step": 5937
+ },
+ {
+ "epoch": 18.971246006389777,
+ "grad_norm": 0.08984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0503,
+ "step": 5938
+ },
+ {
+ "epoch": 18.97444089456869,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0573,
+ "step": 5939
+ },
+ {
+ "epoch": 18.977635782747605,
+ "grad_norm": 0.07861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0447,
+ "step": 5940
+ },
+ {
+ "epoch": 18.98083067092652,
+ "grad_norm": 0.05078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0511,
+ "step": 5941
+ },
+ {
+ "epoch": 18.984025559105433,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0346,
+ "step": 5942
+ },
+ {
+ "epoch": 18.987220447284344,
+ "grad_norm": 0.080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0464,
+ "step": 5943
+ },
+ {
+ "epoch": 18.990415335463258,
+ "grad_norm": 0.0966796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0575,
+ "step": 5944
+ },
+ {
+ "epoch": 18.99361022364217,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0535,
+ "step": 5945
+ },
+ {
+ "epoch": 18.996805111821086,
+ "grad_norm": 0.061767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0436,
+ "step": 5946
+ },
+ {
+ "epoch": 19.0,
+ "grad_norm": 0.08251953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0424,
+ "step": 5947
+ },
+ {
+ "epoch": 19.003194888178914,
+ "grad_norm": 0.056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0399,
+ "step": 5948
+ },
+ {
+ "epoch": 19.00638977635783,
+ "grad_norm": 0.052001953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0425,
+ "step": 5949
+ },
+ {
+ "epoch": 19.009584664536742,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0602,
+ "step": 5950
+ },
+ {
+ "epoch": 19.012779552715656,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0489,
+ "step": 5951
+ },
+ {
+ "epoch": 19.015974440894567,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0436,
+ "step": 5952
+ },
+ {
+ "epoch": 19.01916932907348,
+ "grad_norm": 0.150390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0425,
+ "step": 5953
+ },
+ {
+ "epoch": 19.022364217252395,
+ "grad_norm": 0.08056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0445,
+ "step": 5954
+ },
+ {
+ "epoch": 19.02555910543131,
+ "grad_norm": 0.2041015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0508,
+ "step": 5955
+ },
+ {
+ "epoch": 19.028753993610223,
+ "grad_norm": 0.1396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0419,
+ "step": 5956
+ },
+ {
+ "epoch": 19.031948881789138,
+ "grad_norm": 0.0986328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0475,
+ "step": 5957
+ },
+ {
+ "epoch": 19.03514376996805,
+ "grad_norm": 0.150390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0446,
+ "step": 5958
+ },
+ {
+ "epoch": 19.038338658146966,
+ "grad_norm": 0.1044921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0372,
+ "step": 5959
+ },
+ {
+ "epoch": 19.04153354632588,
+ "grad_norm": 0.1904296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0635,
+ "step": 5960
+ },
+ {
+ "epoch": 19.044728434504794,
+ "grad_norm": 0.30859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0467,
+ "step": 5961
+ },
+ {
+ "epoch": 19.047923322683705,
+ "grad_norm": 0.1025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0382,
+ "step": 5962
+ },
+ {
+ "epoch": 19.05111821086262,
+ "grad_norm": 0.27734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0554,
+ "step": 5963
+ },
+ {
+ "epoch": 19.054313099041533,
+ "grad_norm": 0.30859375,
+ "learning_rate": 0.0005,
+ "loss": 1.056,
+ "step": 5964
+ },
+ {
+ "epoch": 19.057507987220447,
+ "grad_norm": 0.109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0526,
+ "step": 5965
+ },
+ {
+ "epoch": 19.06070287539936,
+ "grad_norm": 0.2216796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0463,
+ "step": 5966
+ },
+ {
+ "epoch": 19.063897763578275,
+ "grad_norm": 0.18359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0471,
+ "step": 5967
+ },
+ {
+ "epoch": 19.06709265175719,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0346,
+ "step": 5968
+ },
+ {
+ "epoch": 19.070287539936103,
+ "grad_norm": 0.2099609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0453,
+ "step": 5969
+ },
+ {
+ "epoch": 19.073482428115017,
+ "grad_norm": 0.16015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0335,
+ "step": 5970
+ },
+ {
+ "epoch": 19.076677316293928,
+ "grad_norm": 0.21484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0489,
+ "step": 5971
+ },
+ {
+ "epoch": 19.079872204472842,
+ "grad_norm": 0.431640625,
+ "learning_rate": 0.0005,
+ "loss": 1.046,
+ "step": 5972
+ },
+ {
+ "epoch": 19.083067092651756,
+ "grad_norm": 0.357421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0438,
+ "step": 5973
+ },
+ {
+ "epoch": 19.08626198083067,
+ "grad_norm": 0.1044921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0481,
+ "step": 5974
+ },
+ {
+ "epoch": 19.089456869009584,
+ "grad_norm": 0.361328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0486,
+ "step": 5975
+ },
+ {
+ "epoch": 19.0926517571885,
+ "grad_norm": 0.30078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0397,
+ "step": 5976
+ },
+ {
+ "epoch": 19.095846645367413,
+ "grad_norm": 0.053466796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0448,
+ "step": 5977
+ },
+ {
+ "epoch": 19.099041533546327,
+ "grad_norm": 0.248046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0482,
+ "step": 5978
+ },
+ {
+ "epoch": 19.10223642172524,
+ "grad_norm": 0.169921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0494,
+ "step": 5979
+ },
+ {
+ "epoch": 19.105431309904155,
+ "grad_norm": 0.12451171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0512,
+ "step": 5980
+ },
+ {
+ "epoch": 19.108626198083066,
+ "grad_norm": 0.1748046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0362,
+ "step": 5981
+ },
+ {
+ "epoch": 19.11182108626198,
+ "grad_norm": 0.10986328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0407,
+ "step": 5982
+ },
+ {
+ "epoch": 19.115015974440894,
+ "grad_norm": 0.1767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0378,
+ "step": 5983
+ },
+ {
+ "epoch": 19.118210862619808,
+ "grad_norm": 0.10302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0469,
+ "step": 5984
+ },
+ {
+ "epoch": 19.121405750798722,
+ "grad_norm": 0.20703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0446,
+ "step": 5985
+ },
+ {
+ "epoch": 19.124600638977636,
+ "grad_norm": 0.09521484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0384,
+ "step": 5986
+ },
+ {
+ "epoch": 19.12779552715655,
+ "grad_norm": 0.2158203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0417,
+ "step": 5987
+ },
+ {
+ "epoch": 19.130990415335464,
+ "grad_norm": 0.09033203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0407,
+ "step": 5988
+ },
+ {
+ "epoch": 19.13418530351438,
+ "grad_norm": 0.2275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0331,
+ "step": 5989
+ },
+ {
+ "epoch": 19.13738019169329,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.047,
+ "step": 5990
+ },
+ {
+ "epoch": 19.140575079872203,
+ "grad_norm": 0.1591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0554,
+ "step": 5991
+ },
+ {
+ "epoch": 19.143769968051117,
+ "grad_norm": 0.083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0531,
+ "step": 5992
+ },
+ {
+ "epoch": 19.14696485623003,
+ "grad_norm": 0.109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0499,
+ "step": 5993
+ },
+ {
+ "epoch": 19.150159744408946,
+ "grad_norm": 0.099609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0587,
+ "step": 5994
+ },
+ {
+ "epoch": 19.15335463258786,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0406,
+ "step": 5995
+ },
+ {
+ "epoch": 19.156549520766774,
+ "grad_norm": 0.0927734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0506,
+ "step": 5996
+ },
+ {
+ "epoch": 19.159744408945688,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0452,
+ "step": 5997
+ },
+ {
+ "epoch": 19.162939297124602,
+ "grad_norm": 0.1044921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0547,
+ "step": 5998
+ },
+ {
+ "epoch": 19.166134185303516,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0346,
+ "step": 5999
+ },
+ {
+ "epoch": 19.169329073482427,
+ "grad_norm": 0.061279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0284,
+ "step": 6000
+ },
+ {
+ "epoch": 19.17252396166134,
+ "grad_norm": 0.12158203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0399,
+ "step": 6001
+ },
+ {
+ "epoch": 19.175718849840255,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0527,
+ "step": 6002
+ },
+ {
+ "epoch": 19.17891373801917,
+ "grad_norm": 0.099609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0537,
+ "step": 6003
+ },
+ {
+ "epoch": 19.182108626198083,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.047,
+ "step": 6004
+ },
+ {
+ "epoch": 19.185303514376997,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0475,
+ "step": 6005
+ },
+ {
+ "epoch": 19.18849840255591,
+ "grad_norm": 0.060302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0536,
+ "step": 6006
+ },
+ {
+ "epoch": 19.191693290734825,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.049,
+ "step": 6007
+ },
+ {
+ "epoch": 19.19488817891374,
+ "grad_norm": 0.08837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0351,
+ "step": 6008
+ },
+ {
+ "epoch": 19.19808306709265,
+ "grad_norm": 0.057861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0543,
+ "step": 6009
+ },
+ {
+ "epoch": 19.201277955271564,
+ "grad_norm": 0.057373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0484,
+ "step": 6010
+ },
+ {
+ "epoch": 19.20447284345048,
+ "grad_norm": 0.059326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0512,
+ "step": 6011
+ },
+ {
+ "epoch": 19.207667731629392,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0546,
+ "step": 6012
+ },
+ {
+ "epoch": 19.210862619808307,
+ "grad_norm": 0.08447265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0369,
+ "step": 6013
+ },
+ {
+ "epoch": 19.21405750798722,
+ "grad_norm": 0.055908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0425,
+ "step": 6014
+ },
+ {
+ "epoch": 19.217252396166135,
+ "grad_norm": 0.0576171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0483,
+ "step": 6015
+ },
+ {
+ "epoch": 19.22044728434505,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0504,
+ "step": 6016
+ },
+ {
+ "epoch": 19.223642172523963,
+ "grad_norm": 0.05224609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0482,
+ "step": 6017
+ },
+ {
+ "epoch": 19.226837060702877,
+ "grad_norm": 0.05810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0477,
+ "step": 6018
+ },
+ {
+ "epoch": 19.230031948881788,
+ "grad_norm": 0.051513671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0481,
+ "step": 6019
+ },
+ {
+ "epoch": 19.233226837060702,
+ "grad_norm": 0.056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0644,
+ "step": 6020
+ },
+ {
+ "epoch": 19.236421725239616,
+ "grad_norm": 0.0498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0474,
+ "step": 6021
+ },
+ {
+ "epoch": 19.23961661341853,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0537,
+ "step": 6022
+ },
+ {
+ "epoch": 19.242811501597444,
+ "grad_norm": 0.06201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0508,
+ "step": 6023
+ },
+ {
+ "epoch": 19.24600638977636,
+ "grad_norm": 0.09228515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0451,
+ "step": 6024
+ },
+ {
+ "epoch": 19.249201277955272,
+ "grad_norm": 0.080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0467,
+ "step": 6025
+ },
+ {
+ "epoch": 19.252396166134186,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.043,
+ "step": 6026
+ },
+ {
+ "epoch": 19.2555910543131,
+ "grad_norm": 0.05517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0388,
+ "step": 6027
+ },
+ {
+ "epoch": 19.25878594249201,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.026,
+ "step": 6028
+ },
+ {
+ "epoch": 19.261980830670925,
+ "grad_norm": 0.052978515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0374,
+ "step": 6029
+ },
+ {
+ "epoch": 19.26517571884984,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0486,
+ "step": 6030
+ },
+ {
+ "epoch": 19.268370607028753,
+ "grad_norm": 0.05029296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0483,
+ "step": 6031
+ },
+ {
+ "epoch": 19.271565495207668,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0428,
+ "step": 6032
+ },
+ {
+ "epoch": 19.27476038338658,
+ "grad_norm": 0.052001953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0428,
+ "step": 6033
+ },
+ {
+ "epoch": 19.277955271565496,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0446,
+ "step": 6034
+ },
+ {
+ "epoch": 19.28115015974441,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0499,
+ "step": 6035
+ },
+ {
+ "epoch": 19.284345047923324,
+ "grad_norm": 0.123046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0429,
+ "step": 6036
+ },
+ {
+ "epoch": 19.287539936102238,
+ "grad_norm": 0.059326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0508,
+ "step": 6037
+ },
+ {
+ "epoch": 19.29073482428115,
+ "grad_norm": 0.11474609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0501,
+ "step": 6038
+ },
+ {
+ "epoch": 19.293929712460063,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0509,
+ "step": 6039
+ },
+ {
+ "epoch": 19.297124600638977,
+ "grad_norm": 0.0927734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0449,
+ "step": 6040
+ },
+ {
+ "epoch": 19.30031948881789,
+ "grad_norm": 0.1357421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0548,
+ "step": 6041
+ },
+ {
+ "epoch": 19.303514376996805,
+ "grad_norm": 0.0517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.049,
+ "step": 6042
+ },
+ {
+ "epoch": 19.30670926517572,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0468,
+ "step": 6043
+ },
+ {
+ "epoch": 19.309904153354633,
+ "grad_norm": 0.053955078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0564,
+ "step": 6044
+ },
+ {
+ "epoch": 19.313099041533548,
+ "grad_norm": 0.0947265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0494,
+ "step": 6045
+ },
+ {
+ "epoch": 19.31629392971246,
+ "grad_norm": 0.05322265625,
+ "learning_rate": 0.0005,
+ "loss": 1.049,
+ "step": 6046
+ },
+ {
+ "epoch": 19.319488817891372,
+ "grad_norm": 0.09326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0599,
+ "step": 6047
+ },
+ {
+ "epoch": 19.322683706070286,
+ "grad_norm": 0.0556640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0506,
+ "step": 6048
+ },
+ {
+ "epoch": 19.3258785942492,
+ "grad_norm": 0.0498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0346,
+ "step": 6049
+ },
+ {
+ "epoch": 19.329073482428115,
+ "grad_norm": 0.061279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0431,
+ "step": 6050
+ },
+ {
+ "epoch": 19.33226837060703,
+ "grad_norm": 0.052734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0465,
+ "step": 6051
+ },
+ {
+ "epoch": 19.335463258785943,
+ "grad_norm": 0.052978515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0408,
+ "step": 6052
+ },
+ {
+ "epoch": 19.338658146964857,
+ "grad_norm": 0.059814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0373,
+ "step": 6053
+ },
+ {
+ "epoch": 19.34185303514377,
+ "grad_norm": 0.05029296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0301,
+ "step": 6054
+ },
+ {
+ "epoch": 19.345047923322685,
+ "grad_norm": 0.0537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0508,
+ "step": 6055
+ },
+ {
+ "epoch": 19.3482428115016,
+ "grad_norm": 0.051513671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0534,
+ "step": 6056
+ },
+ {
+ "epoch": 19.35143769968051,
+ "grad_norm": 0.054443359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0442,
+ "step": 6057
+ },
+ {
+ "epoch": 19.354632587859424,
+ "grad_norm": 0.04931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.046,
+ "step": 6058
+ },
+ {
+ "epoch": 19.357827476038338,
+ "grad_norm": 0.049560546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0387,
+ "step": 6059
+ },
+ {
+ "epoch": 19.361022364217252,
+ "grad_norm": 0.08740234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0352,
+ "step": 6060
+ },
+ {
+ "epoch": 19.364217252396166,
+ "grad_norm": 0.0576171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0559,
+ "step": 6061
+ },
+ {
+ "epoch": 19.36741214057508,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0394,
+ "step": 6062
+ },
+ {
+ "epoch": 19.370607028753994,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0381,
+ "step": 6063
+ },
+ {
+ "epoch": 19.37380191693291,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0548,
+ "step": 6064
+ },
+ {
+ "epoch": 19.376996805111823,
+ "grad_norm": 0.055908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.045,
+ "step": 6065
+ },
+ {
+ "epoch": 19.380191693290733,
+ "grad_norm": 0.059326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0483,
+ "step": 6066
+ },
+ {
+ "epoch": 19.383386581469647,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0579,
+ "step": 6067
+ },
+ {
+ "epoch": 19.38658146964856,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0363,
+ "step": 6068
+ },
+ {
+ "epoch": 19.389776357827476,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0506,
+ "step": 6069
+ },
+ {
+ "epoch": 19.39297124600639,
+ "grad_norm": 0.052001953125,
+ "learning_rate": 0.0005,
+ "loss": 1.051,
+ "step": 6070
+ },
+ {
+ "epoch": 19.396166134185304,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0403,
+ "step": 6071
+ },
+ {
+ "epoch": 19.399361022364218,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0447,
+ "step": 6072
+ },
+ {
+ "epoch": 19.402555910543132,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0417,
+ "step": 6073
+ },
+ {
+ "epoch": 19.405750798722046,
+ "grad_norm": 0.10009765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0472,
+ "step": 6074
+ },
+ {
+ "epoch": 19.408945686900957,
+ "grad_norm": 0.1025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0496,
+ "step": 6075
+ },
+ {
+ "epoch": 19.41214057507987,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0487,
+ "step": 6076
+ },
+ {
+ "epoch": 19.415335463258785,
+ "grad_norm": 0.08154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0584,
+ "step": 6077
+ },
+ {
+ "epoch": 19.4185303514377,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0466,
+ "step": 6078
+ },
+ {
+ "epoch": 19.421725239616613,
+ "grad_norm": 0.08642578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0512,
+ "step": 6079
+ },
+ {
+ "epoch": 19.424920127795527,
+ "grad_norm": 0.050537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0585,
+ "step": 6080
+ },
+ {
+ "epoch": 19.42811501597444,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0562,
+ "step": 6081
+ },
+ {
+ "epoch": 19.431309904153355,
+ "grad_norm": 0.052490234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0461,
+ "step": 6082
+ },
+ {
+ "epoch": 19.43450479233227,
+ "grad_norm": 0.0830078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0585,
+ "step": 6083
+ },
+ {
+ "epoch": 19.437699680511184,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0559,
+ "step": 6084
+ },
+ {
+ "epoch": 19.440894568690094,
+ "grad_norm": 0.0859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0447,
+ "step": 6085
+ },
+ {
+ "epoch": 19.44408945686901,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0447,
+ "step": 6086
+ },
+ {
+ "epoch": 19.447284345047922,
+ "grad_norm": 0.1005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0364,
+ "step": 6087
+ },
+ {
+ "epoch": 19.450479233226837,
+ "grad_norm": 0.057373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0485,
+ "step": 6088
+ },
+ {
+ "epoch": 19.45367412140575,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0561,
+ "step": 6089
+ },
+ {
+ "epoch": 19.456869009584665,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0358,
+ "step": 6090
+ },
+ {
+ "epoch": 19.46006389776358,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0557,
+ "step": 6091
+ },
+ {
+ "epoch": 19.463258785942493,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0574,
+ "step": 6092
+ },
+ {
+ "epoch": 19.466453674121407,
+ "grad_norm": 0.058837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0422,
+ "step": 6093
+ },
+ {
+ "epoch": 19.46964856230032,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0582,
+ "step": 6094
+ },
+ {
+ "epoch": 19.472843450479232,
+ "grad_norm": 0.09228515625,
+ "learning_rate": 0.0005,
+ "loss": 1.036,
+ "step": 6095
+ },
+ {
+ "epoch": 19.476038338658146,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0485,
+ "step": 6096
+ },
+ {
+ "epoch": 19.47923322683706,
+ "grad_norm": 0.05908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0557,
+ "step": 6097
+ },
+ {
+ "epoch": 19.482428115015974,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0445,
+ "step": 6098
+ },
+ {
+ "epoch": 19.48562300319489,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0473,
+ "step": 6099
+ },
+ {
+ "epoch": 19.488817891373802,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0415,
+ "step": 6100
+ },
+ {
+ "epoch": 19.492012779552716,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0484,
+ "step": 6101
+ },
+ {
+ "epoch": 19.49520766773163,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0544,
+ "step": 6102
+ },
+ {
+ "epoch": 19.498402555910545,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0576,
+ "step": 6103
+ },
+ {
+ "epoch": 19.501597444089455,
+ "grad_norm": 0.05126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0465,
+ "step": 6104
+ },
+ {
+ "epoch": 19.50479233226837,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0417,
+ "step": 6105
+ },
+ {
+ "epoch": 19.507987220447284,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0452,
+ "step": 6106
+ },
+ {
+ "epoch": 19.511182108626198,
+ "grad_norm": 0.09375,
+ "learning_rate": 0.0005,
+ "loss": 1.0366,
+ "step": 6107
+ },
+ {
+ "epoch": 19.51437699680511,
+ "grad_norm": 0.0546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0437,
+ "step": 6108
+ },
+ {
+ "epoch": 19.517571884984026,
+ "grad_norm": 0.053955078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0381,
+ "step": 6109
+ },
+ {
+ "epoch": 19.52076677316294,
+ "grad_norm": 0.11083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0509,
+ "step": 6110
+ },
+ {
+ "epoch": 19.523961661341854,
+ "grad_norm": 0.056884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.031,
+ "step": 6111
+ },
+ {
+ "epoch": 19.527156549520768,
+ "grad_norm": 0.10595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0461,
+ "step": 6112
+ },
+ {
+ "epoch": 19.53035143769968,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0402,
+ "step": 6113
+ },
+ {
+ "epoch": 19.533546325878593,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0558,
+ "step": 6114
+ },
+ {
+ "epoch": 19.536741214057507,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0472,
+ "step": 6115
+ },
+ {
+ "epoch": 19.53993610223642,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0524,
+ "step": 6116
+ },
+ {
+ "epoch": 19.543130990415335,
+ "grad_norm": 0.056396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0441,
+ "step": 6117
+ },
+ {
+ "epoch": 19.54632587859425,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0506,
+ "step": 6118
+ },
+ {
+ "epoch": 19.549520766773163,
+ "grad_norm": 0.059326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0389,
+ "step": 6119
+ },
+ {
+ "epoch": 19.552715654952078,
+ "grad_norm": 0.1572265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0626,
+ "step": 6120
+ },
+ {
+ "epoch": 19.55591054313099,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.048,
+ "step": 6121
+ },
+ {
+ "epoch": 19.559105431309906,
+ "grad_norm": 0.115234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0335,
+ "step": 6122
+ },
+ {
+ "epoch": 19.562300319488816,
+ "grad_norm": 0.06005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0507,
+ "step": 6123
+ },
+ {
+ "epoch": 19.56549520766773,
+ "grad_norm": 0.0849609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0582,
+ "step": 6124
+ },
+ {
+ "epoch": 19.568690095846645,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0457,
+ "step": 6125
+ },
+ {
+ "epoch": 19.57188498402556,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0554,
+ "step": 6126
+ },
+ {
+ "epoch": 19.575079872204473,
+ "grad_norm": 0.08447265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0389,
+ "step": 6127
+ },
+ {
+ "epoch": 19.578274760383387,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0506,
+ "step": 6128
+ },
+ {
+ "epoch": 19.5814696485623,
+ "grad_norm": 0.1376953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0498,
+ "step": 6129
+ },
+ {
+ "epoch": 19.584664536741215,
+ "grad_norm": 0.080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0412,
+ "step": 6130
+ },
+ {
+ "epoch": 19.58785942492013,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0507,
+ "step": 6131
+ },
+ {
+ "epoch": 19.591054313099043,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0495,
+ "step": 6132
+ },
+ {
+ "epoch": 19.594249201277954,
+ "grad_norm": 0.12255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0478,
+ "step": 6133
+ },
+ {
+ "epoch": 19.597444089456868,
+ "grad_norm": 0.09326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.045,
+ "step": 6134
+ },
+ {
+ "epoch": 19.600638977635782,
+ "grad_norm": 0.1337890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0448,
+ "step": 6135
+ },
+ {
+ "epoch": 19.603833865814696,
+ "grad_norm": 0.12890625,
+ "learning_rate": 0.0005,
+ "loss": 1.046,
+ "step": 6136
+ },
+ {
+ "epoch": 19.60702875399361,
+ "grad_norm": 0.1728515625,
+ "learning_rate": 0.0005,
+ "loss": 1.057,
+ "step": 6137
+ },
+ {
+ "epoch": 19.610223642172524,
+ "grad_norm": 0.1279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0374,
+ "step": 6138
+ },
+ {
+ "epoch": 19.61341853035144,
+ "grad_norm": 0.11376953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0431,
+ "step": 6139
+ },
+ {
+ "epoch": 19.616613418530353,
+ "grad_norm": 0.2021484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0448,
+ "step": 6140
+ },
+ {
+ "epoch": 19.619808306709267,
+ "grad_norm": 0.0986328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0589,
+ "step": 6141
+ },
+ {
+ "epoch": 19.623003194888177,
+ "grad_norm": 0.1337890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0567,
+ "step": 6142
+ },
+ {
+ "epoch": 19.62619808306709,
+ "grad_norm": 0.1474609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0416,
+ "step": 6143
+ },
+ {
+ "epoch": 19.629392971246006,
+ "grad_norm": 0.130859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0597,
+ "step": 6144
+ },
+ {
+ "epoch": 19.63258785942492,
+ "grad_norm": 0.291015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0616,
+ "step": 6145
+ },
+ {
+ "epoch": 19.635782747603834,
+ "grad_norm": 0.10302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0553,
+ "step": 6146
+ },
+ {
+ "epoch": 19.638977635782748,
+ "grad_norm": 0.11669921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0479,
+ "step": 6147
+ },
+ {
+ "epoch": 19.642172523961662,
+ "grad_norm": 0.061767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0422,
+ "step": 6148
+ },
+ {
+ "epoch": 19.645367412140576,
+ "grad_norm": 0.08740234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0474,
+ "step": 6149
+ },
+ {
+ "epoch": 19.64856230031949,
+ "grad_norm": 0.0546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0473,
+ "step": 6150
+ },
+ {
+ "epoch": 19.6517571884984,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0436,
+ "step": 6151
+ },
+ {
+ "epoch": 19.654952076677315,
+ "grad_norm": 0.1259765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0429,
+ "step": 6152
+ },
+ {
+ "epoch": 19.65814696485623,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0447,
+ "step": 6153
+ },
+ {
+ "epoch": 19.661341853035143,
+ "grad_norm": 0.126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0451,
+ "step": 6154
+ },
+ {
+ "epoch": 19.664536741214057,
+ "grad_norm": 0.095703125,
+ "learning_rate": 0.0005,
+ "loss": 1.048,
+ "step": 6155
+ },
+ {
+ "epoch": 19.66773162939297,
+ "grad_norm": 0.10546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0435,
+ "step": 6156
+ },
+ {
+ "epoch": 19.670926517571885,
+ "grad_norm": 0.1025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0519,
+ "step": 6157
+ },
+ {
+ "epoch": 19.6741214057508,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0533,
+ "step": 6158
+ },
+ {
+ "epoch": 19.677316293929714,
+ "grad_norm": 0.158203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0527,
+ "step": 6159
+ },
+ {
+ "epoch": 19.680511182108628,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0562,
+ "step": 6160
+ },
+ {
+ "epoch": 19.68370607028754,
+ "grad_norm": 0.1103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0534,
+ "step": 6161
+ },
+ {
+ "epoch": 19.686900958466452,
+ "grad_norm": 0.054931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0559,
+ "step": 6162
+ },
+ {
+ "epoch": 19.690095846645367,
+ "grad_norm": 0.123046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0485,
+ "step": 6163
+ },
+ {
+ "epoch": 19.69329073482428,
+ "grad_norm": 0.05810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0534,
+ "step": 6164
+ },
+ {
+ "epoch": 19.696485623003195,
+ "grad_norm": 0.1552734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0538,
+ "step": 6165
+ },
+ {
+ "epoch": 19.69968051118211,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0462,
+ "step": 6166
+ },
+ {
+ "epoch": 19.702875399361023,
+ "grad_norm": 0.1044921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0343,
+ "step": 6167
+ },
+ {
+ "epoch": 19.706070287539937,
+ "grad_norm": 0.062255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0473,
+ "step": 6168
+ },
+ {
+ "epoch": 19.70926517571885,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0545,
+ "step": 6169
+ },
+ {
+ "epoch": 19.712460063897765,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0381,
+ "step": 6170
+ },
+ {
+ "epoch": 19.715654952076676,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0389,
+ "step": 6171
+ },
+ {
+ "epoch": 19.71884984025559,
+ "grad_norm": 0.053466796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0444,
+ "step": 6172
+ },
+ {
+ "epoch": 19.722044728434504,
+ "grad_norm": 0.057373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0407,
+ "step": 6173
+ },
+ {
+ "epoch": 19.72523961661342,
+ "grad_norm": 0.056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0467,
+ "step": 6174
+ },
+ {
+ "epoch": 19.728434504792332,
+ "grad_norm": 0.0576171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0422,
+ "step": 6175
+ },
+ {
+ "epoch": 19.731629392971247,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0658,
+ "step": 6176
+ },
+ {
+ "epoch": 19.73482428115016,
+ "grad_norm": 0.055419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0517,
+ "step": 6177
+ },
+ {
+ "epoch": 19.738019169329075,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0491,
+ "step": 6178
+ },
+ {
+ "epoch": 19.74121405750799,
+ "grad_norm": 0.05712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0452,
+ "step": 6179
+ },
+ {
+ "epoch": 19.7444089456869,
+ "grad_norm": 0.058837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0372,
+ "step": 6180
+ },
+ {
+ "epoch": 19.747603833865814,
+ "grad_norm": 0.0517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0369,
+ "step": 6181
+ },
+ {
+ "epoch": 19.750798722044728,
+ "grad_norm": 0.07861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0389,
+ "step": 6182
+ },
+ {
+ "epoch": 19.75399361022364,
+ "grad_norm": 0.052001953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0539,
+ "step": 6183
+ },
+ {
+ "epoch": 19.757188498402556,
+ "grad_norm": 0.0595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0509,
+ "step": 6184
+ },
+ {
+ "epoch": 19.76038338658147,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0492,
+ "step": 6185
+ },
+ {
+ "epoch": 19.763578274760384,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0394,
+ "step": 6186
+ },
+ {
+ "epoch": 19.766773162939298,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.053,
+ "step": 6187
+ },
+ {
+ "epoch": 19.769968051118212,
+ "grad_norm": 0.09814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0385,
+ "step": 6188
+ },
+ {
+ "epoch": 19.773162939297123,
+ "grad_norm": 0.09814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0475,
+ "step": 6189
+ },
+ {
+ "epoch": 19.776357827476037,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0558,
+ "step": 6190
+ },
+ {
+ "epoch": 19.77955271565495,
+ "grad_norm": 0.09033203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0523,
+ "step": 6191
+ },
+ {
+ "epoch": 19.782747603833865,
+ "grad_norm": 0.060791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0484,
+ "step": 6192
+ },
+ {
+ "epoch": 19.78594249201278,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0505,
+ "step": 6193
+ },
+ {
+ "epoch": 19.789137380191693,
+ "grad_norm": 0.05810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0448,
+ "step": 6194
+ },
+ {
+ "epoch": 19.792332268370608,
+ "grad_norm": 0.0595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0599,
+ "step": 6195
+ },
+ {
+ "epoch": 19.79552715654952,
+ "grad_norm": 0.059814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0411,
+ "step": 6196
+ },
+ {
+ "epoch": 19.798722044728436,
+ "grad_norm": 0.057861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0412,
+ "step": 6197
+ },
+ {
+ "epoch": 19.80191693290735,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0412,
+ "step": 6198
+ },
+ {
+ "epoch": 19.80511182108626,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0368,
+ "step": 6199
+ },
+ {
+ "epoch": 19.808306709265175,
+ "grad_norm": 0.054443359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0505,
+ "step": 6200
+ },
+ {
+ "epoch": 19.81150159744409,
+ "grad_norm": 0.05810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0424,
+ "step": 6201
+ },
+ {
+ "epoch": 19.814696485623003,
+ "grad_norm": 0.062255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0471,
+ "step": 6202
+ },
+ {
+ "epoch": 19.817891373801917,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0406,
+ "step": 6203
+ },
+ {
+ "epoch": 19.82108626198083,
+ "grad_norm": 0.06005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0603,
+ "step": 6204
+ },
+ {
+ "epoch": 19.824281150159745,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.051,
+ "step": 6205
+ },
+ {
+ "epoch": 19.82747603833866,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0402,
+ "step": 6206
+ },
+ {
+ "epoch": 19.830670926517573,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0568,
+ "step": 6207
+ },
+ {
+ "epoch": 19.833865814696484,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0386,
+ "step": 6208
+ },
+ {
+ "epoch": 19.837060702875398,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0453,
+ "step": 6209
+ },
+ {
+ "epoch": 19.840255591054312,
+ "grad_norm": 0.0888671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0513,
+ "step": 6210
+ },
+ {
+ "epoch": 19.843450479233226,
+ "grad_norm": 0.06201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0358,
+ "step": 6211
+ },
+ {
+ "epoch": 19.84664536741214,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.047,
+ "step": 6212
+ },
+ {
+ "epoch": 19.849840255591054,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0407,
+ "step": 6213
+ },
+ {
+ "epoch": 19.85303514376997,
+ "grad_norm": 0.05908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0393,
+ "step": 6214
+ },
+ {
+ "epoch": 19.856230031948883,
+ "grad_norm": 0.12060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0444,
+ "step": 6215
+ },
+ {
+ "epoch": 19.859424920127797,
+ "grad_norm": 0.10888671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0599,
+ "step": 6216
+ },
+ {
+ "epoch": 19.86261980830671,
+ "grad_norm": 0.09765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0574,
+ "step": 6217
+ },
+ {
+ "epoch": 19.86581469648562,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.033,
+ "step": 6218
+ },
+ {
+ "epoch": 19.869009584664536,
+ "grad_norm": 0.119140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0425,
+ "step": 6219
+ },
+ {
+ "epoch": 19.87220447284345,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0593,
+ "step": 6220
+ },
+ {
+ "epoch": 19.875399361022364,
+ "grad_norm": 0.09521484375,
+ "learning_rate": 0.0005,
+ "loss": 1.055,
+ "step": 6221
+ },
+ {
+ "epoch": 19.878594249201278,
+ "grad_norm": 0.09423828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0448,
+ "step": 6222
+ },
+ {
+ "epoch": 19.881789137380192,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0459,
+ "step": 6223
+ },
+ {
+ "epoch": 19.884984025559106,
+ "grad_norm": 0.08984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0504,
+ "step": 6224
+ },
+ {
+ "epoch": 19.88817891373802,
+ "grad_norm": 0.126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0463,
+ "step": 6225
+ },
+ {
+ "epoch": 19.891373801916934,
+ "grad_norm": 0.061767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.045,
+ "step": 6226
+ },
+ {
+ "epoch": 19.894568690095845,
+ "grad_norm": 0.1259765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0381,
+ "step": 6227
+ },
+ {
+ "epoch": 19.89776357827476,
+ "grad_norm": 0.080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0437,
+ "step": 6228
+ },
+ {
+ "epoch": 19.900958466453673,
+ "grad_norm": 0.2138671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0523,
+ "step": 6229
+ },
+ {
+ "epoch": 19.904153354632587,
+ "grad_norm": 0.0859375,
+ "learning_rate": 0.0005,
+ "loss": 1.05,
+ "step": 6230
+ },
+ {
+ "epoch": 19.9073482428115,
+ "grad_norm": 0.10400390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0373,
+ "step": 6231
+ },
+ {
+ "epoch": 19.910543130990416,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0437,
+ "step": 6232
+ },
+ {
+ "epoch": 19.91373801916933,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0428,
+ "step": 6233
+ },
+ {
+ "epoch": 19.916932907348244,
+ "grad_norm": 0.080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0444,
+ "step": 6234
+ },
+ {
+ "epoch": 19.920127795527158,
+ "grad_norm": 0.06201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0491,
+ "step": 6235
+ },
+ {
+ "epoch": 19.923322683706072,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0534,
+ "step": 6236
+ },
+ {
+ "epoch": 19.926517571884983,
+ "grad_norm": 0.087890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0501,
+ "step": 6237
+ },
+ {
+ "epoch": 19.929712460063897,
+ "grad_norm": 0.06103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0417,
+ "step": 6238
+ },
+ {
+ "epoch": 19.93290734824281,
+ "grad_norm": 0.09814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.039,
+ "step": 6239
+ },
+ {
+ "epoch": 19.936102236421725,
+ "grad_norm": 0.056884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.042,
+ "step": 6240
+ },
+ {
+ "epoch": 19.93929712460064,
+ "grad_norm": 0.09619140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0443,
+ "step": 6241
+ },
+ {
+ "epoch": 19.942492012779553,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.049,
+ "step": 6242
+ },
+ {
+ "epoch": 19.945686900958467,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0421,
+ "step": 6243
+ },
+ {
+ "epoch": 19.94888178913738,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0365,
+ "step": 6244
+ },
+ {
+ "epoch": 19.952076677316295,
+ "grad_norm": 0.060791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0557,
+ "step": 6245
+ },
+ {
+ "epoch": 19.955271565495206,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0529,
+ "step": 6246
+ },
+ {
+ "epoch": 19.95846645367412,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0483,
+ "step": 6247
+ },
+ {
+ "epoch": 19.961661341853034,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0441,
+ "step": 6248
+ },
+ {
+ "epoch": 19.96485623003195,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.041,
+ "step": 6249
+ },
+ {
+ "epoch": 19.968051118210862,
+ "grad_norm": 0.05810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0474,
+ "step": 6250
+ },
+ {
+ "epoch": 19.971246006389777,
+ "grad_norm": 0.09130859375,
+ "learning_rate": 0.0005,
+ "loss": 1.055,
+ "step": 6251
+ },
+ {
+ "epoch": 19.97444089456869,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0373,
+ "step": 6252
+ },
+ {
+ "epoch": 19.977635782747605,
+ "grad_norm": 0.1103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0473,
+ "step": 6253
+ },
+ {
+ "epoch": 19.98083067092652,
+ "grad_norm": 0.0556640625,
+ "learning_rate": 0.0005,
+ "loss": 1.044,
+ "step": 6254
+ },
+ {
+ "epoch": 19.984025559105433,
+ "grad_norm": 0.1103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0394,
+ "step": 6255
+ },
+ {
+ "epoch": 19.987220447284344,
+ "grad_norm": 0.044921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0334,
+ "step": 6256
+ },
+ {
+ "epoch": 19.990415335463258,
+ "grad_norm": 0.09326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.037,
+ "step": 6257
+ },
+ {
+ "epoch": 19.99361022364217,
+ "grad_norm": 0.054931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0363,
+ "step": 6258
+ },
+ {
+ "epoch": 19.996805111821086,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0589,
+ "step": 6259
+ },
+ {
+ "epoch": 20.0,
+ "grad_norm": 0.051513671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0544,
+ "step": 6260
+ },
+ {
+ "epoch": 20.003194888178914,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0513,
+ "step": 6261
+ },
+ {
+ "epoch": 20.00638977635783,
+ "grad_norm": 0.05712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0408,
+ "step": 6262
+ },
+ {
+ "epoch": 20.009584664536742,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0318,
+ "step": 6263
+ },
+ {
+ "epoch": 20.012779552715656,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0483,
+ "step": 6264
+ },
+ {
+ "epoch": 20.015974440894567,
+ "grad_norm": 0.08349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0514,
+ "step": 6265
+ },
+ {
+ "epoch": 20.01916932907348,
+ "grad_norm": 0.06005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0486,
+ "step": 6266
+ },
+ {
+ "epoch": 20.022364217252395,
+ "grad_norm": 0.08203125,
+ "learning_rate": 0.0005,
+ "loss": 1.04,
+ "step": 6267
+ },
+ {
+ "epoch": 20.02555910543131,
+ "grad_norm": 0.05810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0458,
+ "step": 6268
+ },
+ {
+ "epoch": 20.028753993610223,
+ "grad_norm": 0.1015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0476,
+ "step": 6269
+ },
+ {
+ "epoch": 20.031948881789138,
+ "grad_norm": 0.0546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0555,
+ "step": 6270
+ },
+ {
+ "epoch": 20.03514376996805,
+ "grad_norm": 0.1025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0383,
+ "step": 6271
+ },
+ {
+ "epoch": 20.038338658146966,
+ "grad_norm": 0.05126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0381,
+ "step": 6272
+ },
+ {
+ "epoch": 20.04153354632588,
+ "grad_norm": 0.08447265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0531,
+ "step": 6273
+ },
+ {
+ "epoch": 20.044728434504794,
+ "grad_norm": 0.054931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0315,
+ "step": 6274
+ },
+ {
+ "epoch": 20.047923322683705,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0487,
+ "step": 6275
+ },
+ {
+ "epoch": 20.05111821086262,
+ "grad_norm": 0.05322265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0557,
+ "step": 6276
+ },
+ {
+ "epoch": 20.054313099041533,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.055,
+ "step": 6277
+ },
+ {
+ "epoch": 20.057507987220447,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0548,
+ "step": 6278
+ },
+ {
+ "epoch": 20.06070287539936,
+ "grad_norm": 0.1171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0411,
+ "step": 6279
+ },
+ {
+ "epoch": 20.063897763578275,
+ "grad_norm": 0.055908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0421,
+ "step": 6280
+ },
+ {
+ "epoch": 20.06709265175719,
+ "grad_norm": 0.169921875,
+ "learning_rate": 0.0005,
+ "loss": 1.033,
+ "step": 6281
+ },
+ {
+ "epoch": 20.070287539936103,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0423,
+ "step": 6282
+ },
+ {
+ "epoch": 20.073482428115017,
+ "grad_norm": 0.130859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0341,
+ "step": 6283
+ },
+ {
+ "epoch": 20.076677316293928,
+ "grad_norm": 0.08544921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0551,
+ "step": 6284
+ },
+ {
+ "epoch": 20.079872204472842,
+ "grad_norm": 0.054443359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0517,
+ "step": 6285
+ },
+ {
+ "epoch": 20.083067092651756,
+ "grad_norm": 0.0537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0462,
+ "step": 6286
+ },
+ {
+ "epoch": 20.08626198083067,
+ "grad_norm": 0.047119140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0561,
+ "step": 6287
+ },
+ {
+ "epoch": 20.089456869009584,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0496,
+ "step": 6288
+ },
+ {
+ "epoch": 20.0926517571885,
+ "grad_norm": 0.0537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0451,
+ "step": 6289
+ },
+ {
+ "epoch": 20.095846645367413,
+ "grad_norm": 0.054443359375,
+ "learning_rate": 0.0005,
+ "loss": 1.049,
+ "step": 6290
+ },
+ {
+ "epoch": 20.099041533546327,
+ "grad_norm": 0.056396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0484,
+ "step": 6291
+ },
+ {
+ "epoch": 20.10223642172524,
+ "grad_norm": 0.052734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0543,
+ "step": 6292
+ },
+ {
+ "epoch": 20.105431309904155,
+ "grad_norm": 0.058349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0377,
+ "step": 6293
+ },
+ {
+ "epoch": 20.108626198083066,
+ "grad_norm": 0.048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0446,
+ "step": 6294
+ },
+ {
+ "epoch": 20.11182108626198,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0411,
+ "step": 6295
+ },
+ {
+ "epoch": 20.115015974440894,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0363,
+ "step": 6296
+ },
+ {
+ "epoch": 20.118210862619808,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0484,
+ "step": 6297
+ },
+ {
+ "epoch": 20.121405750798722,
+ "grad_norm": 0.10986328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0337,
+ "step": 6298
+ },
+ {
+ "epoch": 20.124600638977636,
+ "grad_norm": 0.060791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0402,
+ "step": 6299
+ },
+ {
+ "epoch": 20.12779552715655,
+ "grad_norm": 0.12060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0451,
+ "step": 6300
+ },
+ {
+ "epoch": 20.130990415335464,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.048,
+ "step": 6301
+ },
+ {
+ "epoch": 20.13418530351438,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0453,
+ "step": 6302
+ },
+ {
+ "epoch": 20.13738019169329,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0402,
+ "step": 6303
+ },
+ {
+ "epoch": 20.140575079872203,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0478,
+ "step": 6304
+ },
+ {
+ "epoch": 20.143769968051117,
+ "grad_norm": 0.09326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0341,
+ "step": 6305
+ },
+ {
+ "epoch": 20.14696485623003,
+ "grad_norm": 0.080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0418,
+ "step": 6306
+ },
+ {
+ "epoch": 20.150159744408946,
+ "grad_norm": 0.0615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0449,
+ "step": 6307
+ },
+ {
+ "epoch": 20.15335463258786,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.04,
+ "step": 6308
+ },
+ {
+ "epoch": 20.156549520766774,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0482,
+ "step": 6309
+ },
+ {
+ "epoch": 20.159744408945688,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0447,
+ "step": 6310
+ },
+ {
+ "epoch": 20.162939297124602,
+ "grad_norm": 0.09716796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0527,
+ "step": 6311
+ },
+ {
+ "epoch": 20.166134185303516,
+ "grad_norm": 0.05419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0583,
+ "step": 6312
+ },
+ {
+ "epoch": 20.169329073482427,
+ "grad_norm": 0.05615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0496,
+ "step": 6313
+ },
+ {
+ "epoch": 20.17252396166134,
+ "grad_norm": 0.05615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0391,
+ "step": 6314
+ },
+ {
+ "epoch": 20.175718849840255,
+ "grad_norm": 0.056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.04,
+ "step": 6315
+ },
+ {
+ "epoch": 20.17891373801917,
+ "grad_norm": 0.0576171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0472,
+ "step": 6316
+ },
+ {
+ "epoch": 20.182108626198083,
+ "grad_norm": 0.057861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0465,
+ "step": 6317
+ },
+ {
+ "epoch": 20.185303514376997,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0418,
+ "step": 6318
+ },
+ {
+ "epoch": 20.18849840255591,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0463,
+ "step": 6319
+ },
+ {
+ "epoch": 20.191693290734825,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0515,
+ "step": 6320
+ },
+ {
+ "epoch": 20.19488817891374,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.042,
+ "step": 6321
+ },
+ {
+ "epoch": 20.19808306709265,
+ "grad_norm": 0.060302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0341,
+ "step": 6322
+ },
+ {
+ "epoch": 20.201277955271564,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0446,
+ "step": 6323
+ },
+ {
+ "epoch": 20.20447284345048,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0567,
+ "step": 6324
+ },
+ {
+ "epoch": 20.207667731629392,
+ "grad_norm": 0.126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0587,
+ "step": 6325
+ },
+ {
+ "epoch": 20.210862619808307,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0465,
+ "step": 6326
+ },
+ {
+ "epoch": 20.21405750798722,
+ "grad_norm": 0.10009765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0573,
+ "step": 6327
+ },
+ {
+ "epoch": 20.217252396166135,
+ "grad_norm": 0.052734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0506,
+ "step": 6328
+ },
+ {
+ "epoch": 20.22044728434505,
+ "grad_norm": 0.0849609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0535,
+ "step": 6329
+ },
+ {
+ "epoch": 20.223642172523963,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0391,
+ "step": 6330
+ },
+ {
+ "epoch": 20.226837060702877,
+ "grad_norm": 0.11279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0412,
+ "step": 6331
+ },
+ {
+ "epoch": 20.230031948881788,
+ "grad_norm": 0.059326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0448,
+ "step": 6332
+ },
+ {
+ "epoch": 20.233226837060702,
+ "grad_norm": 0.11181640625,
+ "learning_rate": 0.0005,
+ "loss": 1.049,
+ "step": 6333
+ },
+ {
+ "epoch": 20.236421725239616,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.046,
+ "step": 6334
+ },
+ {
+ "epoch": 20.23961661341853,
+ "grad_norm": 0.0830078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0528,
+ "step": 6335
+ },
+ {
+ "epoch": 20.242811501597444,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0374,
+ "step": 6336
+ },
+ {
+ "epoch": 20.24600638977636,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0356,
+ "step": 6337
+ },
+ {
+ "epoch": 20.249201277955272,
+ "grad_norm": 0.09423828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0348,
+ "step": 6338
+ },
+ {
+ "epoch": 20.252396166134186,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0451,
+ "step": 6339
+ },
+ {
+ "epoch": 20.2555910543131,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0488,
+ "step": 6340
+ },
+ {
+ "epoch": 20.25878594249201,
+ "grad_norm": 0.062255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0296,
+ "step": 6341
+ },
+ {
+ "epoch": 20.261980830670925,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0553,
+ "step": 6342
+ },
+ {
+ "epoch": 20.26517571884984,
+ "grad_norm": 0.09375,
+ "learning_rate": 0.0005,
+ "loss": 1.0463,
+ "step": 6343
+ },
+ {
+ "epoch": 20.268370607028753,
+ "grad_norm": 0.10107421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0489,
+ "step": 6344
+ },
+ {
+ "epoch": 20.271565495207668,
+ "grad_norm": 0.057861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0447,
+ "step": 6345
+ },
+ {
+ "epoch": 20.27476038338658,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.039,
+ "step": 6346
+ },
+ {
+ "epoch": 20.277955271565496,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0476,
+ "step": 6347
+ },
+ {
+ "epoch": 20.28115015974441,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0403,
+ "step": 6348
+ },
+ {
+ "epoch": 20.284345047923324,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0423,
+ "step": 6349
+ },
+ {
+ "epoch": 20.287539936102238,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0409,
+ "step": 6350
+ },
+ {
+ "epoch": 20.29073482428115,
+ "grad_norm": 0.1103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.039,
+ "step": 6351
+ },
+ {
+ "epoch": 20.293929712460063,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0452,
+ "step": 6352
+ },
+ {
+ "epoch": 20.297124600638977,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0492,
+ "step": 6353
+ },
+ {
+ "epoch": 20.30031948881789,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0458,
+ "step": 6354
+ },
+ {
+ "epoch": 20.303514376996805,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0435,
+ "step": 6355
+ },
+ {
+ "epoch": 20.30670926517572,
+ "grad_norm": 0.08935546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0404,
+ "step": 6356
+ },
+ {
+ "epoch": 20.309904153354633,
+ "grad_norm": 0.111328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0322,
+ "step": 6357
+ },
+ {
+ "epoch": 20.313099041533548,
+ "grad_norm": 0.08154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0456,
+ "step": 6358
+ },
+ {
+ "epoch": 20.31629392971246,
+ "grad_norm": 0.146484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0563,
+ "step": 6359
+ },
+ {
+ "epoch": 20.319488817891372,
+ "grad_norm": 0.053955078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0476,
+ "step": 6360
+ },
+ {
+ "epoch": 20.322683706070286,
+ "grad_norm": 0.1103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.046,
+ "step": 6361
+ },
+ {
+ "epoch": 20.3258785942492,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0495,
+ "step": 6362
+ },
+ {
+ "epoch": 20.329073482428115,
+ "grad_norm": 0.052978515625,
+ "learning_rate": 0.0005,
+ "loss": 1.037,
+ "step": 6363
+ },
+ {
+ "epoch": 20.33226837060703,
+ "grad_norm": 0.061279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.042,
+ "step": 6364
+ },
+ {
+ "epoch": 20.335463258785943,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0326,
+ "step": 6365
+ },
+ {
+ "epoch": 20.338658146964857,
+ "grad_norm": 0.06201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.036,
+ "step": 6366
+ },
+ {
+ "epoch": 20.34185303514377,
+ "grad_norm": 0.060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0509,
+ "step": 6367
+ },
+ {
+ "epoch": 20.345047923322685,
+ "grad_norm": 0.056396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0377,
+ "step": 6368
+ },
+ {
+ "epoch": 20.3482428115016,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0566,
+ "step": 6369
+ },
+ {
+ "epoch": 20.35143769968051,
+ "grad_norm": 0.05517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0456,
+ "step": 6370
+ },
+ {
+ "epoch": 20.354632587859424,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0473,
+ "step": 6371
+ },
+ {
+ "epoch": 20.357827476038338,
+ "grad_norm": 0.0517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0454,
+ "step": 6372
+ },
+ {
+ "epoch": 20.361022364217252,
+ "grad_norm": 0.06103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0437,
+ "step": 6373
+ },
+ {
+ "epoch": 20.364217252396166,
+ "grad_norm": 0.08251953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0401,
+ "step": 6374
+ },
+ {
+ "epoch": 20.36741214057508,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0518,
+ "step": 6375
+ },
+ {
+ "epoch": 20.370607028753994,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0499,
+ "step": 6376
+ },
+ {
+ "epoch": 20.37380191693291,
+ "grad_norm": 0.052490234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0514,
+ "step": 6377
+ },
+ {
+ "epoch": 20.376996805111823,
+ "grad_norm": 0.0478515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0617,
+ "step": 6378
+ },
+ {
+ "epoch": 20.380191693290733,
+ "grad_norm": 0.054443359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0353,
+ "step": 6379
+ },
+ {
+ "epoch": 20.383386581469647,
+ "grad_norm": 0.052490234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0475,
+ "step": 6380
+ },
+ {
+ "epoch": 20.38658146964856,
+ "grad_norm": 0.056884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0554,
+ "step": 6381
+ },
+ {
+ "epoch": 20.389776357827476,
+ "grad_norm": 0.08837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0471,
+ "step": 6382
+ },
+ {
+ "epoch": 20.39297124600639,
+ "grad_norm": 0.0439453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0556,
+ "step": 6383
+ },
+ {
+ "epoch": 20.396166134185304,
+ "grad_norm": 0.056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.049,
+ "step": 6384
+ },
+ {
+ "epoch": 20.399361022364218,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0576,
+ "step": 6385
+ },
+ {
+ "epoch": 20.402555910543132,
+ "grad_norm": 0.05029296875,
+ "learning_rate": 0.0005,
+ "loss": 1.052,
+ "step": 6386
+ },
+ {
+ "epoch": 20.405750798722046,
+ "grad_norm": 0.054931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0472,
+ "step": 6387
+ },
+ {
+ "epoch": 20.408945686900957,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0496,
+ "step": 6388
+ },
+ {
+ "epoch": 20.41214057507987,
+ "grad_norm": 0.05859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0506,
+ "step": 6389
+ },
+ {
+ "epoch": 20.415335463258785,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0687,
+ "step": 6390
+ },
+ {
+ "epoch": 20.4185303514377,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0435,
+ "step": 6391
+ },
+ {
+ "epoch": 20.421725239616613,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0485,
+ "step": 6392
+ },
+ {
+ "epoch": 20.424920127795527,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0534,
+ "step": 6393
+ },
+ {
+ "epoch": 20.42811501597444,
+ "grad_norm": 0.057373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0573,
+ "step": 6394
+ },
+ {
+ "epoch": 20.431309904153355,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0516,
+ "step": 6395
+ },
+ {
+ "epoch": 20.43450479233227,
+ "grad_norm": 0.0537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0468,
+ "step": 6396
+ },
+ {
+ "epoch": 20.437699680511184,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0434,
+ "step": 6397
+ },
+ {
+ "epoch": 20.440894568690094,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0471,
+ "step": 6398
+ },
+ {
+ "epoch": 20.44408945686901,
+ "grad_norm": 0.0595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0354,
+ "step": 6399
+ },
+ {
+ "epoch": 20.447284345047922,
+ "grad_norm": 0.0546875,
+ "learning_rate": 0.0005,
+ "loss": 1.049,
+ "step": 6400
+ },
+ {
+ "epoch": 20.450479233226837,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0312,
+ "step": 6401
+ },
+ {
+ "epoch": 20.45367412140575,
+ "grad_norm": 0.08203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0395,
+ "step": 6402
+ },
+ {
+ "epoch": 20.456869009584665,
+ "grad_norm": 0.05517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0603,
+ "step": 6403
+ },
+ {
+ "epoch": 20.46006389776358,
+ "grad_norm": 0.126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0395,
+ "step": 6404
+ },
+ {
+ "epoch": 20.463258785942493,
+ "grad_norm": 0.058837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0536,
+ "step": 6405
+ },
+ {
+ "epoch": 20.466453674121407,
+ "grad_norm": 0.08056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.037,
+ "step": 6406
+ },
+ {
+ "epoch": 20.46964856230032,
+ "grad_norm": 0.053955078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0474,
+ "step": 6407
+ },
+ {
+ "epoch": 20.472843450479232,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0557,
+ "step": 6408
+ },
+ {
+ "epoch": 20.476038338658146,
+ "grad_norm": 0.054443359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0412,
+ "step": 6409
+ },
+ {
+ "epoch": 20.47923322683706,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0445,
+ "step": 6410
+ },
+ {
+ "epoch": 20.482428115015974,
+ "grad_norm": 0.08935546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0437,
+ "step": 6411
+ },
+ {
+ "epoch": 20.48562300319489,
+ "grad_norm": 0.0498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.046,
+ "step": 6412
+ },
+ {
+ "epoch": 20.488817891373802,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.038,
+ "step": 6413
+ },
+ {
+ "epoch": 20.492012779552716,
+ "grad_norm": 0.056884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0423,
+ "step": 6414
+ },
+ {
+ "epoch": 20.49520766773163,
+ "grad_norm": 0.060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0529,
+ "step": 6415
+ },
+ {
+ "epoch": 20.498402555910545,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0484,
+ "step": 6416
+ },
+ {
+ "epoch": 20.501597444089455,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0481,
+ "step": 6417
+ },
+ {
+ "epoch": 20.50479233226837,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0373,
+ "step": 6418
+ },
+ {
+ "epoch": 20.507987220447284,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0537,
+ "step": 6419
+ },
+ {
+ "epoch": 20.511182108626198,
+ "grad_norm": 0.059814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0372,
+ "step": 6420
+ },
+ {
+ "epoch": 20.51437699680511,
+ "grad_norm": 0.05419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0438,
+ "step": 6421
+ },
+ {
+ "epoch": 20.517571884984026,
+ "grad_norm": 0.054443359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0431,
+ "step": 6422
+ },
+ {
+ "epoch": 20.52076677316294,
+ "grad_norm": 0.049560546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0583,
+ "step": 6423
+ },
+ {
+ "epoch": 20.523961661341854,
+ "grad_norm": 0.05517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0454,
+ "step": 6424
+ },
+ {
+ "epoch": 20.527156549520768,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.047,
+ "step": 6425
+ },
+ {
+ "epoch": 20.53035143769968,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0509,
+ "step": 6426
+ },
+ {
+ "epoch": 20.533546325878593,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0383,
+ "step": 6427
+ },
+ {
+ "epoch": 20.536741214057507,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.055,
+ "step": 6428
+ },
+ {
+ "epoch": 20.53993610223642,
+ "grad_norm": 0.06201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0386,
+ "step": 6429
+ },
+ {
+ "epoch": 20.543130990415335,
+ "grad_norm": 0.0556640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0464,
+ "step": 6430
+ },
+ {
+ "epoch": 20.54632587859425,
+ "grad_norm": 0.055908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0472,
+ "step": 6431
+ },
+ {
+ "epoch": 20.549520766773163,
+ "grad_norm": 0.0556640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0441,
+ "step": 6432
+ },
+ {
+ "epoch": 20.552715654952078,
+ "grad_norm": 0.059326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0624,
+ "step": 6433
+ },
+ {
+ "epoch": 20.55591054313099,
+ "grad_norm": 0.059814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0375,
+ "step": 6434
+ },
+ {
+ "epoch": 20.559105431309906,
+ "grad_norm": 0.08544921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0558,
+ "step": 6435
+ },
+ {
+ "epoch": 20.562300319488816,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0479,
+ "step": 6436
+ },
+ {
+ "epoch": 20.56549520766773,
+ "grad_norm": 0.05419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0501,
+ "step": 6437
+ },
+ {
+ "epoch": 20.568690095846645,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0519,
+ "step": 6438
+ },
+ {
+ "epoch": 20.57188498402556,
+ "grad_norm": 0.05712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0491,
+ "step": 6439
+ },
+ {
+ "epoch": 20.575079872204473,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0557,
+ "step": 6440
+ },
+ {
+ "epoch": 20.578274760383387,
+ "grad_norm": 0.1083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0322,
+ "step": 6441
+ },
+ {
+ "epoch": 20.5814696485623,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.045,
+ "step": 6442
+ },
+ {
+ "epoch": 20.584664536741215,
+ "grad_norm": 0.061767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0427,
+ "step": 6443
+ },
+ {
+ "epoch": 20.58785942492013,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.04,
+ "step": 6444
+ },
+ {
+ "epoch": 20.591054313099043,
+ "grad_norm": 0.0947265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0474,
+ "step": 6445
+ },
+ {
+ "epoch": 20.594249201277954,
+ "grad_norm": 0.08544921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0449,
+ "step": 6446
+ },
+ {
+ "epoch": 20.597444089456868,
+ "grad_norm": 0.119140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0413,
+ "step": 6447
+ },
+ {
+ "epoch": 20.600638977635782,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0504,
+ "step": 6448
+ },
+ {
+ "epoch": 20.603833865814696,
+ "grad_norm": 0.1923828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0338,
+ "step": 6449
+ },
+ {
+ "epoch": 20.60702875399361,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0497,
+ "step": 6450
+ },
+ {
+ "epoch": 20.610223642172524,
+ "grad_norm": 0.0888671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0524,
+ "step": 6451
+ },
+ {
+ "epoch": 20.61341853035144,
+ "grad_norm": 0.08544921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0438,
+ "step": 6452
+ },
+ {
+ "epoch": 20.616613418530353,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0487,
+ "step": 6453
+ },
+ {
+ "epoch": 20.619808306709267,
+ "grad_norm": 0.055419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0377,
+ "step": 6454
+ },
+ {
+ "epoch": 20.623003194888177,
+ "grad_norm": 0.059814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0343,
+ "step": 6455
+ },
+ {
+ "epoch": 20.62619808306709,
+ "grad_norm": 0.06201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.05,
+ "step": 6456
+ },
+ {
+ "epoch": 20.629392971246006,
+ "grad_norm": 0.05712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0429,
+ "step": 6457
+ },
+ {
+ "epoch": 20.63258785942492,
+ "grad_norm": 0.05224609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0485,
+ "step": 6458
+ },
+ {
+ "epoch": 20.635782747603834,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0446,
+ "step": 6459
+ },
+ {
+ "epoch": 20.638977635782748,
+ "grad_norm": 0.058837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0421,
+ "step": 6460
+ },
+ {
+ "epoch": 20.642172523961662,
+ "grad_norm": 0.058349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0549,
+ "step": 6461
+ },
+ {
+ "epoch": 20.645367412140576,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0511,
+ "step": 6462
+ },
+ {
+ "epoch": 20.64856230031949,
+ "grad_norm": 0.0888671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0343,
+ "step": 6463
+ },
+ {
+ "epoch": 20.6517571884984,
+ "grad_norm": 0.1669921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0594,
+ "step": 6464
+ },
+ {
+ "epoch": 20.654952076677315,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0551,
+ "step": 6465
+ },
+ {
+ "epoch": 20.65814696485623,
+ "grad_norm": 0.16015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0459,
+ "step": 6466
+ },
+ {
+ "epoch": 20.661341853035143,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0407,
+ "step": 6467
+ },
+ {
+ "epoch": 20.664536741214057,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0614,
+ "step": 6468
+ },
+ {
+ "epoch": 20.66773162939297,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.036,
+ "step": 6469
+ },
+ {
+ "epoch": 20.670926517571885,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.038,
+ "step": 6470
+ },
+ {
+ "epoch": 20.6741214057508,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0511,
+ "step": 6471
+ },
+ {
+ "epoch": 20.677316293929714,
+ "grad_norm": 0.083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0481,
+ "step": 6472
+ },
+ {
+ "epoch": 20.680511182108628,
+ "grad_norm": 0.0556640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0509,
+ "step": 6473
+ },
+ {
+ "epoch": 20.68370607028754,
+ "grad_norm": 0.05322265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0401,
+ "step": 6474
+ },
+ {
+ "epoch": 20.686900958466452,
+ "grad_norm": 0.0595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0423,
+ "step": 6475
+ },
+ {
+ "epoch": 20.690095846645367,
+ "grad_norm": 0.08544921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0487,
+ "step": 6476
+ },
+ {
+ "epoch": 20.69329073482428,
+ "grad_norm": 0.083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0518,
+ "step": 6477
+ },
+ {
+ "epoch": 20.696485623003195,
+ "grad_norm": 0.0537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0557,
+ "step": 6478
+ },
+ {
+ "epoch": 20.69968051118211,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0348,
+ "step": 6479
+ },
+ {
+ "epoch": 20.702875399361023,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0386,
+ "step": 6480
+ },
+ {
+ "epoch": 20.706070287539937,
+ "grad_norm": 0.052490234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0553,
+ "step": 6481
+ },
+ {
+ "epoch": 20.70926517571885,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0378,
+ "step": 6482
+ },
+ {
+ "epoch": 20.712460063897765,
+ "grad_norm": 0.08154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0463,
+ "step": 6483
+ },
+ {
+ "epoch": 20.715654952076676,
+ "grad_norm": 0.08642578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0526,
+ "step": 6484
+ },
+ {
+ "epoch": 20.71884984025559,
+ "grad_norm": 0.062255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0503,
+ "step": 6485
+ },
+ {
+ "epoch": 20.722044728434504,
+ "grad_norm": 0.11572265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0371,
+ "step": 6486
+ },
+ {
+ "epoch": 20.72523961661342,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.055,
+ "step": 6487
+ },
+ {
+ "epoch": 20.728434504792332,
+ "grad_norm": 0.1357421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0407,
+ "step": 6488
+ },
+ {
+ "epoch": 20.731629392971247,
+ "grad_norm": 0.109375,
+ "learning_rate": 0.0005,
+ "loss": 1.035,
+ "step": 6489
+ },
+ {
+ "epoch": 20.73482428115016,
+ "grad_norm": 0.0908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0572,
+ "step": 6490
+ },
+ {
+ "epoch": 20.738019169329075,
+ "grad_norm": 0.09033203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0418,
+ "step": 6491
+ },
+ {
+ "epoch": 20.74121405750799,
+ "grad_norm": 0.05126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0362,
+ "step": 6492
+ },
+ {
+ "epoch": 20.7444089456869,
+ "grad_norm": 0.08349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0486,
+ "step": 6493
+ },
+ {
+ "epoch": 20.747603833865814,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0446,
+ "step": 6494
+ },
+ {
+ "epoch": 20.750798722044728,
+ "grad_norm": 0.1044921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0442,
+ "step": 6495
+ },
+ {
+ "epoch": 20.75399361022364,
+ "grad_norm": 0.10986328125,
+ "learning_rate": 0.0005,
+ "loss": 1.048,
+ "step": 6496
+ },
+ {
+ "epoch": 20.757188498402556,
+ "grad_norm": 0.099609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0368,
+ "step": 6497
+ },
+ {
+ "epoch": 20.76038338658147,
+ "grad_norm": 0.1533203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0432,
+ "step": 6498
+ },
+ {
+ "epoch": 20.763578274760384,
+ "grad_norm": 0.09326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0579,
+ "step": 6499
+ },
+ {
+ "epoch": 20.766773162939298,
+ "grad_norm": 0.162109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0427,
+ "step": 6500
+ },
+ {
+ "epoch": 20.769968051118212,
+ "grad_norm": 0.10888671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0496,
+ "step": 6501
+ },
+ {
+ "epoch": 20.773162939297123,
+ "grad_norm": 0.08056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0548,
+ "step": 6502
+ },
+ {
+ "epoch": 20.776357827476037,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.041,
+ "step": 6503
+ },
+ {
+ "epoch": 20.77955271565495,
+ "grad_norm": 0.06103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0383,
+ "step": 6504
+ },
+ {
+ "epoch": 20.782747603833865,
+ "grad_norm": 0.062255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0435,
+ "step": 6505
+ },
+ {
+ "epoch": 20.78594249201278,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0479,
+ "step": 6506
+ },
+ {
+ "epoch": 20.789137380191693,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0476,
+ "step": 6507
+ },
+ {
+ "epoch": 20.792332268370608,
+ "grad_norm": 0.10595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0401,
+ "step": 6508
+ },
+ {
+ "epoch": 20.79552715654952,
+ "grad_norm": 0.134765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0478,
+ "step": 6509
+ },
+ {
+ "epoch": 20.798722044728436,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0428,
+ "step": 6510
+ },
+ {
+ "epoch": 20.80191693290735,
+ "grad_norm": 0.1376953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0488,
+ "step": 6511
+ },
+ {
+ "epoch": 20.80511182108626,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0365,
+ "step": 6512
+ },
+ {
+ "epoch": 20.808306709265175,
+ "grad_norm": 0.107421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0549,
+ "step": 6513
+ },
+ {
+ "epoch": 20.81150159744409,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0521,
+ "step": 6514
+ },
+ {
+ "epoch": 20.814696485623003,
+ "grad_norm": 0.1005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0445,
+ "step": 6515
+ },
+ {
+ "epoch": 20.817891373801917,
+ "grad_norm": 0.10693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0535,
+ "step": 6516
+ },
+ {
+ "epoch": 20.82108626198083,
+ "grad_norm": 0.05859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0467,
+ "step": 6517
+ },
+ {
+ "epoch": 20.824281150159745,
+ "grad_norm": 0.0810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0324,
+ "step": 6518
+ },
+ {
+ "epoch": 20.82747603833866,
+ "grad_norm": 0.09423828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0475,
+ "step": 6519
+ },
+ {
+ "epoch": 20.830670926517573,
+ "grad_norm": 0.1279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0467,
+ "step": 6520
+ },
+ {
+ "epoch": 20.833865814696484,
+ "grad_norm": 0.0986328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0437,
+ "step": 6521
+ },
+ {
+ "epoch": 20.837060702875398,
+ "grad_norm": 0.2265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0456,
+ "step": 6522
+ },
+ {
+ "epoch": 20.840255591054312,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.037,
+ "step": 6523
+ },
+ {
+ "epoch": 20.843450479233226,
+ "grad_norm": 0.12353515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0481,
+ "step": 6524
+ },
+ {
+ "epoch": 20.84664536741214,
+ "grad_norm": 0.1298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0407,
+ "step": 6525
+ },
+ {
+ "epoch": 20.849840255591054,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0415,
+ "step": 6526
+ },
+ {
+ "epoch": 20.85303514376997,
+ "grad_norm": 0.1376953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0445,
+ "step": 6527
+ },
+ {
+ "epoch": 20.856230031948883,
+ "grad_norm": 0.12158203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0511,
+ "step": 6528
+ },
+ {
+ "epoch": 20.859424920127797,
+ "grad_norm": 0.27734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0436,
+ "step": 6529
+ },
+ {
+ "epoch": 20.86261980830671,
+ "grad_norm": 0.1064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0305,
+ "step": 6530
+ },
+ {
+ "epoch": 20.86581469648562,
+ "grad_norm": 0.18359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0429,
+ "step": 6531
+ },
+ {
+ "epoch": 20.869009584664536,
+ "grad_norm": 0.162109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0515,
+ "step": 6532
+ },
+ {
+ "epoch": 20.87220447284345,
+ "grad_norm": 0.125,
+ "learning_rate": 0.0005,
+ "loss": 1.0541,
+ "step": 6533
+ },
+ {
+ "epoch": 20.875399361022364,
+ "grad_norm": 0.1484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0447,
+ "step": 6534
+ },
+ {
+ "epoch": 20.878594249201278,
+ "grad_norm": 0.0966796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0386,
+ "step": 6535
+ },
+ {
+ "epoch": 20.881789137380192,
+ "grad_norm": 0.1611328125,
+ "learning_rate": 0.0005,
+ "loss": 1.05,
+ "step": 6536
+ },
+ {
+ "epoch": 20.884984025559106,
+ "grad_norm": 0.050048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0424,
+ "step": 6537
+ },
+ {
+ "epoch": 20.88817891373802,
+ "grad_norm": 0.1396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0573,
+ "step": 6538
+ },
+ {
+ "epoch": 20.891373801916934,
+ "grad_norm": 0.053955078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0501,
+ "step": 6539
+ },
+ {
+ "epoch": 20.894568690095845,
+ "grad_norm": 0.1533203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0382,
+ "step": 6540
+ },
+ {
+ "epoch": 20.89776357827476,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0325,
+ "step": 6541
+ },
+ {
+ "epoch": 20.900958466453673,
+ "grad_norm": 0.11181640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0363,
+ "step": 6542
+ },
+ {
+ "epoch": 20.904153354632587,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0533,
+ "step": 6543
+ },
+ {
+ "epoch": 20.9073482428115,
+ "grad_norm": 0.10595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0496,
+ "step": 6544
+ },
+ {
+ "epoch": 20.910543130990416,
+ "grad_norm": 0.099609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0513,
+ "step": 6545
+ },
+ {
+ "epoch": 20.91373801916933,
+ "grad_norm": 0.1328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0414,
+ "step": 6546
+ },
+ {
+ "epoch": 20.916932907348244,
+ "grad_norm": 0.2001953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0579,
+ "step": 6547
+ },
+ {
+ "epoch": 20.920127795527158,
+ "grad_norm": 0.060302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0429,
+ "step": 6548
+ },
+ {
+ "epoch": 20.923322683706072,
+ "grad_norm": 0.1201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.046,
+ "step": 6549
+ },
+ {
+ "epoch": 20.926517571884983,
+ "grad_norm": 0.080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.056,
+ "step": 6550
+ },
+ {
+ "epoch": 20.929712460063897,
+ "grad_norm": 0.0908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0399,
+ "step": 6551
+ },
+ {
+ "epoch": 20.93290734824281,
+ "grad_norm": 0.126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0531,
+ "step": 6552
+ },
+ {
+ "epoch": 20.936102236421725,
+ "grad_norm": 0.09375,
+ "learning_rate": 0.0005,
+ "loss": 1.0547,
+ "step": 6553
+ },
+ {
+ "epoch": 20.93929712460064,
+ "grad_norm": 0.2197265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0423,
+ "step": 6554
+ },
+ {
+ "epoch": 20.942492012779553,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.044,
+ "step": 6555
+ },
+ {
+ "epoch": 20.945686900958467,
+ "grad_norm": 0.271484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0457,
+ "step": 6556
+ },
+ {
+ "epoch": 20.94888178913738,
+ "grad_norm": 0.173828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0432,
+ "step": 6557
+ },
+ {
+ "epoch": 20.952076677316295,
+ "grad_norm": 0.126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0483,
+ "step": 6558
+ },
+ {
+ "epoch": 20.955271565495206,
+ "grad_norm": 0.279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0505,
+ "step": 6559
+ },
+ {
+ "epoch": 20.95846645367412,
+ "grad_norm": 0.103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0454,
+ "step": 6560
+ },
+ {
+ "epoch": 20.961661341853034,
+ "grad_norm": 0.1904296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0481,
+ "step": 6561
+ },
+ {
+ "epoch": 20.96485623003195,
+ "grad_norm": 0.2236328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0417,
+ "step": 6562
+ },
+ {
+ "epoch": 20.968051118210862,
+ "grad_norm": 0.08154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0422,
+ "step": 6563
+ },
+ {
+ "epoch": 20.971246006389777,
+ "grad_norm": 0.259765625,
+ "learning_rate": 0.0005,
+ "loss": 1.035,
+ "step": 6564
+ },
+ {
+ "epoch": 20.97444089456869,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0509,
+ "step": 6565
+ },
+ {
+ "epoch": 20.977635782747605,
+ "grad_norm": 0.1630859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0546,
+ "step": 6566
+ },
+ {
+ "epoch": 20.98083067092652,
+ "grad_norm": 0.053955078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0506,
+ "step": 6567
+ },
+ {
+ "epoch": 20.984025559105433,
+ "grad_norm": 0.1201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0431,
+ "step": 6568
+ },
+ {
+ "epoch": 20.987220447284344,
+ "grad_norm": 0.059814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0375,
+ "step": 6569
+ },
+ {
+ "epoch": 20.990415335463258,
+ "grad_norm": 0.10546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0466,
+ "step": 6570
+ },
+ {
+ "epoch": 20.99361022364217,
+ "grad_norm": 0.08349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0493,
+ "step": 6571
+ },
+ {
+ "epoch": 20.996805111821086,
+ "grad_norm": 0.1337890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0389,
+ "step": 6572
+ },
+ {
+ "epoch": 21.0,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0465,
+ "step": 6573
+ },
+ {
+ "epoch": 21.003194888178914,
+ "grad_norm": 0.1103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0433,
+ "step": 6574
+ },
+ {
+ "epoch": 21.00638977635783,
+ "grad_norm": 0.06201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0408,
+ "step": 6575
+ },
+ {
+ "epoch": 21.009584664536742,
+ "grad_norm": 0.0849609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0481,
+ "step": 6576
+ },
+ {
+ "epoch": 21.012779552715656,
+ "grad_norm": 0.05419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0432,
+ "step": 6577
+ },
+ {
+ "epoch": 21.015974440894567,
+ "grad_norm": 0.1083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.049,
+ "step": 6578
+ },
+ {
+ "epoch": 21.01916932907348,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0448,
+ "step": 6579
+ },
+ {
+ "epoch": 21.022364217252395,
+ "grad_norm": 0.095703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0504,
+ "step": 6580
+ },
+ {
+ "epoch": 21.02555910543131,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0567,
+ "step": 6581
+ },
+ {
+ "epoch": 21.028753993610223,
+ "grad_norm": 0.10888671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0542,
+ "step": 6582
+ },
+ {
+ "epoch": 21.031948881789138,
+ "grad_norm": 0.140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0414,
+ "step": 6583
+ },
+ {
+ "epoch": 21.03514376996805,
+ "grad_norm": 0.0966796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0457,
+ "step": 6584
+ },
+ {
+ "epoch": 21.038338658146966,
+ "grad_norm": 0.09521484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0415,
+ "step": 6585
+ },
+ {
+ "epoch": 21.04153354632588,
+ "grad_norm": 0.107421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0473,
+ "step": 6586
+ },
+ {
+ "epoch": 21.044728434504794,
+ "grad_norm": 0.09228515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0509,
+ "step": 6587
+ },
+ {
+ "epoch": 21.047923322683705,
+ "grad_norm": 0.11669921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0388,
+ "step": 6588
+ },
+ {
+ "epoch": 21.05111821086262,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0507,
+ "step": 6589
+ },
+ {
+ "epoch": 21.054313099041533,
+ "grad_norm": 0.1328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0368,
+ "step": 6590
+ },
+ {
+ "epoch": 21.057507987220447,
+ "grad_norm": 0.059326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0531,
+ "step": 6591
+ },
+ {
+ "epoch": 21.06070287539936,
+ "grad_norm": 0.1669921875,
+ "learning_rate": 0.0005,
+ "loss": 1.038,
+ "step": 6592
+ },
+ {
+ "epoch": 21.063897763578275,
+ "grad_norm": 0.06201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0417,
+ "step": 6593
+ },
+ {
+ "epoch": 21.06709265175719,
+ "grad_norm": 0.1279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0401,
+ "step": 6594
+ },
+ {
+ "epoch": 21.070287539936103,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0324,
+ "step": 6595
+ },
+ {
+ "epoch": 21.073482428115017,
+ "grad_norm": 0.11083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.049,
+ "step": 6596
+ },
+ {
+ "epoch": 21.076677316293928,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0407,
+ "step": 6597
+ },
+ {
+ "epoch": 21.079872204472842,
+ "grad_norm": 0.058837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0447,
+ "step": 6598
+ },
+ {
+ "epoch": 21.083067092651756,
+ "grad_norm": 0.06103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0514,
+ "step": 6599
+ },
+ {
+ "epoch": 21.08626198083067,
+ "grad_norm": 0.1103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0448,
+ "step": 6600
+ },
+ {
+ "epoch": 21.089456869009584,
+ "grad_norm": 0.12060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0441,
+ "step": 6601
+ },
+ {
+ "epoch": 21.0926517571885,
+ "grad_norm": 0.0859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0435,
+ "step": 6602
+ },
+ {
+ "epoch": 21.095846645367413,
+ "grad_norm": 0.0986328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0439,
+ "step": 6603
+ },
+ {
+ "epoch": 21.099041533546327,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0403,
+ "step": 6604
+ },
+ {
+ "epoch": 21.10223642172524,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0313,
+ "step": 6605
+ },
+ {
+ "epoch": 21.105431309904155,
+ "grad_norm": 0.059814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0275,
+ "step": 6606
+ },
+ {
+ "epoch": 21.108626198083066,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0479,
+ "step": 6607
+ },
+ {
+ "epoch": 21.11182108626198,
+ "grad_norm": 0.0830078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0412,
+ "step": 6608
+ },
+ {
+ "epoch": 21.115015974440894,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0416,
+ "step": 6609
+ },
+ {
+ "epoch": 21.118210862619808,
+ "grad_norm": 0.08154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0428,
+ "step": 6610
+ },
+ {
+ "epoch": 21.121405750798722,
+ "grad_norm": 0.09228515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0355,
+ "step": 6611
+ },
+ {
+ "epoch": 21.124600638977636,
+ "grad_norm": 0.08837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0533,
+ "step": 6612
+ },
+ {
+ "epoch": 21.12779552715655,
+ "grad_norm": 0.10595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0407,
+ "step": 6613
+ },
+ {
+ "epoch": 21.130990415335464,
+ "grad_norm": 0.10107421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0405,
+ "step": 6614
+ },
+ {
+ "epoch": 21.13418530351438,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0509,
+ "step": 6615
+ },
+ {
+ "epoch": 21.13738019169329,
+ "grad_norm": 0.08447265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0483,
+ "step": 6616
+ },
+ {
+ "epoch": 21.140575079872203,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0473,
+ "step": 6617
+ },
+ {
+ "epoch": 21.143769968051117,
+ "grad_norm": 0.0810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0635,
+ "step": 6618
+ },
+ {
+ "epoch": 21.14696485623003,
+ "grad_norm": 0.0888671875,
+ "learning_rate": 0.0005,
+ "loss": 1.05,
+ "step": 6619
+ },
+ {
+ "epoch": 21.150159744408946,
+ "grad_norm": 0.05712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0397,
+ "step": 6620
+ },
+ {
+ "epoch": 21.15335463258786,
+ "grad_norm": 0.08203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0362,
+ "step": 6621
+ },
+ {
+ "epoch": 21.156549520766774,
+ "grad_norm": 0.0859375,
+ "learning_rate": 0.0005,
+ "loss": 1.046,
+ "step": 6622
+ },
+ {
+ "epoch": 21.159744408945688,
+ "grad_norm": 0.1005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0396,
+ "step": 6623
+ },
+ {
+ "epoch": 21.162939297124602,
+ "grad_norm": 0.1357421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0464,
+ "step": 6624
+ },
+ {
+ "epoch": 21.166134185303516,
+ "grad_norm": 0.0615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0438,
+ "step": 6625
+ },
+ {
+ "epoch": 21.169329073482427,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0411,
+ "step": 6626
+ },
+ {
+ "epoch": 21.17252396166134,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0406,
+ "step": 6627
+ },
+ {
+ "epoch": 21.175718849840255,
+ "grad_norm": 0.08447265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0566,
+ "step": 6628
+ },
+ {
+ "epoch": 21.17891373801917,
+ "grad_norm": 0.130859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0334,
+ "step": 6629
+ },
+ {
+ "epoch": 21.182108626198083,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.052,
+ "step": 6630
+ },
+ {
+ "epoch": 21.185303514376997,
+ "grad_norm": 0.208984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0419,
+ "step": 6631
+ },
+ {
+ "epoch": 21.18849840255591,
+ "grad_norm": 0.057861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0557,
+ "step": 6632
+ },
+ {
+ "epoch": 21.191693290734825,
+ "grad_norm": 0.1689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0412,
+ "step": 6633
+ },
+ {
+ "epoch": 21.19488817891374,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0361,
+ "step": 6634
+ },
+ {
+ "epoch": 21.19808306709265,
+ "grad_norm": 0.12890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0461,
+ "step": 6635
+ },
+ {
+ "epoch": 21.201277955271564,
+ "grad_norm": 0.080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0561,
+ "step": 6636
+ },
+ {
+ "epoch": 21.20447284345048,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0498,
+ "step": 6637
+ },
+ {
+ "epoch": 21.207667731629392,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0516,
+ "step": 6638
+ },
+ {
+ "epoch": 21.210862619808307,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0481,
+ "step": 6639
+ },
+ {
+ "epoch": 21.21405750798722,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0379,
+ "step": 6640
+ },
+ {
+ "epoch": 21.217252396166135,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0482,
+ "step": 6641
+ },
+ {
+ "epoch": 21.22044728434505,
+ "grad_norm": 0.05810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0464,
+ "step": 6642
+ },
+ {
+ "epoch": 21.223642172523963,
+ "grad_norm": 0.054931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0475,
+ "step": 6643
+ },
+ {
+ "epoch": 21.226837060702877,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0453,
+ "step": 6644
+ },
+ {
+ "epoch": 21.230031948881788,
+ "grad_norm": 0.059814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0451,
+ "step": 6645
+ },
+ {
+ "epoch": 21.233226837060702,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0446,
+ "step": 6646
+ },
+ {
+ "epoch": 21.236421725239616,
+ "grad_norm": 0.04833984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0334,
+ "step": 6647
+ },
+ {
+ "epoch": 21.23961661341853,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0581,
+ "step": 6648
+ },
+ {
+ "epoch": 21.242811501597444,
+ "grad_norm": 0.0537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0539,
+ "step": 6649
+ },
+ {
+ "epoch": 21.24600638977636,
+ "grad_norm": 0.06005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0527,
+ "step": 6650
+ },
+ {
+ "epoch": 21.249201277955272,
+ "grad_norm": 0.057373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0328,
+ "step": 6651
+ },
+ {
+ "epoch": 21.252396166134186,
+ "grad_norm": 0.05615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0475,
+ "step": 6652
+ },
+ {
+ "epoch": 21.2555910543131,
+ "grad_norm": 0.06103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0391,
+ "step": 6653
+ },
+ {
+ "epoch": 21.25878594249201,
+ "grad_norm": 0.059326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0501,
+ "step": 6654
+ },
+ {
+ "epoch": 21.261980830670925,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0325,
+ "step": 6655
+ },
+ {
+ "epoch": 21.26517571884984,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0403,
+ "step": 6656
+ },
+ {
+ "epoch": 21.268370607028753,
+ "grad_norm": 0.057373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0313,
+ "step": 6657
+ },
+ {
+ "epoch": 21.271565495207668,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0462,
+ "step": 6658
+ },
+ {
+ "epoch": 21.27476038338658,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0546,
+ "step": 6659
+ },
+ {
+ "epoch": 21.277955271565496,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0442,
+ "step": 6660
+ },
+ {
+ "epoch": 21.28115015974441,
+ "grad_norm": 0.0498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0538,
+ "step": 6661
+ },
+ {
+ "epoch": 21.284345047923324,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0404,
+ "step": 6662
+ },
+ {
+ "epoch": 21.287539936102238,
+ "grad_norm": 0.0556640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0464,
+ "step": 6663
+ },
+ {
+ "epoch": 21.29073482428115,
+ "grad_norm": 0.10009765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0575,
+ "step": 6664
+ },
+ {
+ "epoch": 21.293929712460063,
+ "grad_norm": 0.0595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0445,
+ "step": 6665
+ },
+ {
+ "epoch": 21.297124600638977,
+ "grad_norm": 0.1298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0483,
+ "step": 6666
+ },
+ {
+ "epoch": 21.30031948881789,
+ "grad_norm": 0.048583984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0529,
+ "step": 6667
+ },
+ {
+ "epoch": 21.303514376996805,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.045,
+ "step": 6668
+ },
+ {
+ "epoch": 21.30670926517572,
+ "grad_norm": 0.055908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0549,
+ "step": 6669
+ },
+ {
+ "epoch": 21.309904153354633,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0389,
+ "step": 6670
+ },
+ {
+ "epoch": 21.313099041533548,
+ "grad_norm": 0.051513671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0381,
+ "step": 6671
+ },
+ {
+ "epoch": 21.31629392971246,
+ "grad_norm": 0.08349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0456,
+ "step": 6672
+ },
+ {
+ "epoch": 21.319488817891372,
+ "grad_norm": 0.056396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0454,
+ "step": 6673
+ },
+ {
+ "epoch": 21.322683706070286,
+ "grad_norm": 0.12890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0546,
+ "step": 6674
+ },
+ {
+ "epoch": 21.3258785942492,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0476,
+ "step": 6675
+ },
+ {
+ "epoch": 21.329073482428115,
+ "grad_norm": 0.08203125,
+ "learning_rate": 0.0005,
+ "loss": 1.047,
+ "step": 6676
+ },
+ {
+ "epoch": 21.33226837060703,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0386,
+ "step": 6677
+ },
+ {
+ "epoch": 21.335463258785943,
+ "grad_norm": 0.0615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0445,
+ "step": 6678
+ },
+ {
+ "epoch": 21.338658146964857,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0454,
+ "step": 6679
+ },
+ {
+ "epoch": 21.34185303514377,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0436,
+ "step": 6680
+ },
+ {
+ "epoch": 21.345047923322685,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0427,
+ "step": 6681
+ },
+ {
+ "epoch": 21.3482428115016,
+ "grad_norm": 0.09423828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0446,
+ "step": 6682
+ },
+ {
+ "epoch": 21.35143769968051,
+ "grad_norm": 0.08056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0465,
+ "step": 6683
+ },
+ {
+ "epoch": 21.354632587859424,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0513,
+ "step": 6684
+ },
+ {
+ "epoch": 21.357827476038338,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0458,
+ "step": 6685
+ },
+ {
+ "epoch": 21.361022364217252,
+ "grad_norm": 0.052001953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0393,
+ "step": 6686
+ },
+ {
+ "epoch": 21.364217252396166,
+ "grad_norm": 0.1142578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0463,
+ "step": 6687
+ },
+ {
+ "epoch": 21.36741214057508,
+ "grad_norm": 0.05517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0506,
+ "step": 6688
+ },
+ {
+ "epoch": 21.370607028753994,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0471,
+ "step": 6689
+ },
+ {
+ "epoch": 21.37380191693291,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0532,
+ "step": 6690
+ },
+ {
+ "epoch": 21.376996805111823,
+ "grad_norm": 0.056884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0433,
+ "step": 6691
+ },
+ {
+ "epoch": 21.380191693290733,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0472,
+ "step": 6692
+ },
+ {
+ "epoch": 21.383386581469647,
+ "grad_norm": 0.05615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0456,
+ "step": 6693
+ },
+ {
+ "epoch": 21.38658146964856,
+ "grad_norm": 0.053955078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0369,
+ "step": 6694
+ },
+ {
+ "epoch": 21.389776357827476,
+ "grad_norm": 0.0615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.056,
+ "step": 6695
+ },
+ {
+ "epoch": 21.39297124600639,
+ "grad_norm": 0.061767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.046,
+ "step": 6696
+ },
+ {
+ "epoch": 21.396166134185304,
+ "grad_norm": 0.052734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0477,
+ "step": 6697
+ },
+ {
+ "epoch": 21.399361022364218,
+ "grad_norm": 0.055908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0488,
+ "step": 6698
+ },
+ {
+ "epoch": 21.402555910543132,
+ "grad_norm": 0.0498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.048,
+ "step": 6699
+ },
+ {
+ "epoch": 21.405750798722046,
+ "grad_norm": 0.052734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0433,
+ "step": 6700
+ },
+ {
+ "epoch": 21.408945686900957,
+ "grad_norm": 0.08056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0472,
+ "step": 6701
+ },
+ {
+ "epoch": 21.41214057507987,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0415,
+ "step": 6702
+ },
+ {
+ "epoch": 21.415335463258785,
+ "grad_norm": 0.130859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0458,
+ "step": 6703
+ },
+ {
+ "epoch": 21.4185303514377,
+ "grad_norm": 0.0849609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0368,
+ "step": 6704
+ },
+ {
+ "epoch": 21.421725239616613,
+ "grad_norm": 0.091796875,
+ "learning_rate": 0.0005,
+ "loss": 1.029,
+ "step": 6705
+ },
+ {
+ "epoch": 21.424920127795527,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0389,
+ "step": 6706
+ },
+ {
+ "epoch": 21.42811501597444,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.045,
+ "step": 6707
+ },
+ {
+ "epoch": 21.431309904153355,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0572,
+ "step": 6708
+ },
+ {
+ "epoch": 21.43450479233227,
+ "grad_norm": 0.09130859375,
+ "learning_rate": 0.0005,
+ "loss": 1.036,
+ "step": 6709
+ },
+ {
+ "epoch": 21.437699680511184,
+ "grad_norm": 0.0576171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0399,
+ "step": 6710
+ },
+ {
+ "epoch": 21.440894568690094,
+ "grad_norm": 0.0947265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0465,
+ "step": 6711
+ },
+ {
+ "epoch": 21.44408945686901,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0391,
+ "step": 6712
+ },
+ {
+ "epoch": 21.447284345047922,
+ "grad_norm": 0.0537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0406,
+ "step": 6713
+ },
+ {
+ "epoch": 21.450479233226837,
+ "grad_norm": 0.10009765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0465,
+ "step": 6714
+ },
+ {
+ "epoch": 21.45367412140575,
+ "grad_norm": 0.05322265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0484,
+ "step": 6715
+ },
+ {
+ "epoch": 21.456869009584665,
+ "grad_norm": 0.08984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0411,
+ "step": 6716
+ },
+ {
+ "epoch": 21.46006389776358,
+ "grad_norm": 0.057861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0459,
+ "step": 6717
+ },
+ {
+ "epoch": 21.463258785942493,
+ "grad_norm": 0.061767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0507,
+ "step": 6718
+ },
+ {
+ "epoch": 21.466453674121407,
+ "grad_norm": 0.057861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.041,
+ "step": 6719
+ },
+ {
+ "epoch": 21.46964856230032,
+ "grad_norm": 0.0927734375,
+ "learning_rate": 0.0005,
+ "loss": 1.036,
+ "step": 6720
+ },
+ {
+ "epoch": 21.472843450479232,
+ "grad_norm": 0.09619140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0441,
+ "step": 6721
+ },
+ {
+ "epoch": 21.476038338658146,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0407,
+ "step": 6722
+ },
+ {
+ "epoch": 21.47923322683706,
+ "grad_norm": 0.057373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0534,
+ "step": 6723
+ },
+ {
+ "epoch": 21.482428115015974,
+ "grad_norm": 0.0888671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0437,
+ "step": 6724
+ },
+ {
+ "epoch": 21.48562300319489,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0458,
+ "step": 6725
+ },
+ {
+ "epoch": 21.488817891373802,
+ "grad_norm": 0.062255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0451,
+ "step": 6726
+ },
+ {
+ "epoch": 21.492012779552716,
+ "grad_norm": 0.06005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0449,
+ "step": 6727
+ },
+ {
+ "epoch": 21.49520766773163,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0404,
+ "step": 6728
+ },
+ {
+ "epoch": 21.498402555910545,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.046,
+ "step": 6729
+ },
+ {
+ "epoch": 21.501597444089455,
+ "grad_norm": 0.061279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0487,
+ "step": 6730
+ },
+ {
+ "epoch": 21.50479233226837,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0501,
+ "step": 6731
+ },
+ {
+ "epoch": 21.507987220447284,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0452,
+ "step": 6732
+ },
+ {
+ "epoch": 21.511182108626198,
+ "grad_norm": 0.054443359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0435,
+ "step": 6733
+ },
+ {
+ "epoch": 21.51437699680511,
+ "grad_norm": 0.058837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0529,
+ "step": 6734
+ },
+ {
+ "epoch": 21.517571884984026,
+ "grad_norm": 0.056884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0327,
+ "step": 6735
+ },
+ {
+ "epoch": 21.52076677316294,
+ "grad_norm": 0.05712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0452,
+ "step": 6736
+ },
+ {
+ "epoch": 21.523961661341854,
+ "grad_norm": 0.0576171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0433,
+ "step": 6737
+ },
+ {
+ "epoch": 21.527156549520768,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0494,
+ "step": 6738
+ },
+ {
+ "epoch": 21.53035143769968,
+ "grad_norm": 0.060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0417,
+ "step": 6739
+ },
+ {
+ "epoch": 21.533546325878593,
+ "grad_norm": 0.05517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0464,
+ "step": 6740
+ },
+ {
+ "epoch": 21.536741214057507,
+ "grad_norm": 0.1142578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0391,
+ "step": 6741
+ },
+ {
+ "epoch": 21.53993610223642,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0496,
+ "step": 6742
+ },
+ {
+ "epoch": 21.543130990415335,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0477,
+ "step": 6743
+ },
+ {
+ "epoch": 21.54632587859425,
+ "grad_norm": 0.087890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0543,
+ "step": 6744
+ },
+ {
+ "epoch": 21.549520766773163,
+ "grad_norm": 0.055908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0396,
+ "step": 6745
+ },
+ {
+ "epoch": 21.552715654952078,
+ "grad_norm": 0.08935546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0389,
+ "step": 6746
+ },
+ {
+ "epoch": 21.55591054313099,
+ "grad_norm": 0.049072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.056,
+ "step": 6747
+ },
+ {
+ "epoch": 21.559105431309906,
+ "grad_norm": 0.11474609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0454,
+ "step": 6748
+ },
+ {
+ "epoch": 21.562300319488816,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0454,
+ "step": 6749
+ },
+ {
+ "epoch": 21.56549520766773,
+ "grad_norm": 0.0576171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0531,
+ "step": 6750
+ },
+ {
+ "epoch": 21.568690095846645,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0385,
+ "step": 6751
+ },
+ {
+ "epoch": 21.57188498402556,
+ "grad_norm": 0.05615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0326,
+ "step": 6752
+ },
+ {
+ "epoch": 21.575079872204473,
+ "grad_norm": 0.054931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0543,
+ "step": 6753
+ },
+ {
+ "epoch": 21.578274760383387,
+ "grad_norm": 0.05419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0446,
+ "step": 6754
+ },
+ {
+ "epoch": 21.5814696485623,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0399,
+ "step": 6755
+ },
+ {
+ "epoch": 21.584664536741215,
+ "grad_norm": 0.08154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0507,
+ "step": 6756
+ },
+ {
+ "epoch": 21.58785942492013,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.04,
+ "step": 6757
+ },
+ {
+ "epoch": 21.591054313099043,
+ "grad_norm": 0.04541015625,
+ "learning_rate": 0.0005,
+ "loss": 1.046,
+ "step": 6758
+ },
+ {
+ "epoch": 21.594249201277954,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0506,
+ "step": 6759
+ },
+ {
+ "epoch": 21.597444089456868,
+ "grad_norm": 0.060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0477,
+ "step": 6760
+ },
+ {
+ "epoch": 21.600638977635782,
+ "grad_norm": 0.060302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.046,
+ "step": 6761
+ },
+ {
+ "epoch": 21.603833865814696,
+ "grad_norm": 0.05419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0331,
+ "step": 6762
+ },
+ {
+ "epoch": 21.60702875399361,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0393,
+ "step": 6763
+ },
+ {
+ "epoch": 21.610223642172524,
+ "grad_norm": 0.059326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0462,
+ "step": 6764
+ },
+ {
+ "epoch": 21.61341853035144,
+ "grad_norm": 0.08837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0289,
+ "step": 6765
+ },
+ {
+ "epoch": 21.616613418530353,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0508,
+ "step": 6766
+ },
+ {
+ "epoch": 21.619808306709267,
+ "grad_norm": 0.09521484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0488,
+ "step": 6767
+ },
+ {
+ "epoch": 21.623003194888177,
+ "grad_norm": 0.052978515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0493,
+ "step": 6768
+ },
+ {
+ "epoch": 21.62619808306709,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0418,
+ "step": 6769
+ },
+ {
+ "epoch": 21.629392971246006,
+ "grad_norm": 0.12255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0542,
+ "step": 6770
+ },
+ {
+ "epoch": 21.63258785942492,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0495,
+ "step": 6771
+ },
+ {
+ "epoch": 21.635782747603834,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0366,
+ "step": 6772
+ },
+ {
+ "epoch": 21.638977635782748,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0447,
+ "step": 6773
+ },
+ {
+ "epoch": 21.642172523961662,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0447,
+ "step": 6774
+ },
+ {
+ "epoch": 21.645367412140576,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0374,
+ "step": 6775
+ },
+ {
+ "epoch": 21.64856230031949,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.046,
+ "step": 6776
+ },
+ {
+ "epoch": 21.6517571884984,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0491,
+ "step": 6777
+ },
+ {
+ "epoch": 21.654952076677315,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0527,
+ "step": 6778
+ },
+ {
+ "epoch": 21.65814696485623,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0509,
+ "step": 6779
+ },
+ {
+ "epoch": 21.661341853035143,
+ "grad_norm": 0.060791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0396,
+ "step": 6780
+ },
+ {
+ "epoch": 21.664536741214057,
+ "grad_norm": 0.059326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.038,
+ "step": 6781
+ },
+ {
+ "epoch": 21.66773162939297,
+ "grad_norm": 0.05712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0585,
+ "step": 6782
+ },
+ {
+ "epoch": 21.670926517571885,
+ "grad_norm": 0.051025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0463,
+ "step": 6783
+ },
+ {
+ "epoch": 21.6741214057508,
+ "grad_norm": 0.057861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0456,
+ "step": 6784
+ },
+ {
+ "epoch": 21.677316293929714,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0559,
+ "step": 6785
+ },
+ {
+ "epoch": 21.680511182108628,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0471,
+ "step": 6786
+ },
+ {
+ "epoch": 21.68370607028754,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0482,
+ "step": 6787
+ },
+ {
+ "epoch": 21.686900958466452,
+ "grad_norm": 0.056884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0515,
+ "step": 6788
+ },
+ {
+ "epoch": 21.690095846645367,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0392,
+ "step": 6789
+ },
+ {
+ "epoch": 21.69329073482428,
+ "grad_norm": 0.054443359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0408,
+ "step": 6790
+ },
+ {
+ "epoch": 21.696485623003195,
+ "grad_norm": 0.0546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0434,
+ "step": 6791
+ },
+ {
+ "epoch": 21.69968051118211,
+ "grad_norm": 0.056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0458,
+ "step": 6792
+ },
+ {
+ "epoch": 21.702875399361023,
+ "grad_norm": 0.052978515625,
+ "learning_rate": 0.0005,
+ "loss": 1.034,
+ "step": 6793
+ },
+ {
+ "epoch": 21.706070287539937,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0552,
+ "step": 6794
+ },
+ {
+ "epoch": 21.70926517571885,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0487,
+ "step": 6795
+ },
+ {
+ "epoch": 21.712460063897765,
+ "grad_norm": 0.052978515625,
+ "learning_rate": 0.0005,
+ "loss": 1.053,
+ "step": 6796
+ },
+ {
+ "epoch": 21.715654952076676,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0503,
+ "step": 6797
+ },
+ {
+ "epoch": 21.71884984025559,
+ "grad_norm": 0.058349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0277,
+ "step": 6798
+ },
+ {
+ "epoch": 21.722044728434504,
+ "grad_norm": 0.054931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0563,
+ "step": 6799
+ },
+ {
+ "epoch": 21.72523961661342,
+ "grad_norm": 0.08251953125,
+ "learning_rate": 0.0005,
+ "loss": 1.036,
+ "step": 6800
+ },
+ {
+ "epoch": 21.728434504792332,
+ "grad_norm": 0.10546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0527,
+ "step": 6801
+ },
+ {
+ "epoch": 21.731629392971247,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0558,
+ "step": 6802
+ },
+ {
+ "epoch": 21.73482428115016,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0571,
+ "step": 6803
+ },
+ {
+ "epoch": 21.738019169329075,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0479,
+ "step": 6804
+ },
+ {
+ "epoch": 21.74121405750799,
+ "grad_norm": 0.05419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.048,
+ "step": 6805
+ },
+ {
+ "epoch": 21.7444089456869,
+ "grad_norm": 0.061767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0493,
+ "step": 6806
+ },
+ {
+ "epoch": 21.747603833865814,
+ "grad_norm": 0.048095703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0479,
+ "step": 6807
+ },
+ {
+ "epoch": 21.750798722044728,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0485,
+ "step": 6808
+ },
+ {
+ "epoch": 21.75399361022364,
+ "grad_norm": 0.060791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0525,
+ "step": 6809
+ },
+ {
+ "epoch": 21.757188498402556,
+ "grad_norm": 0.04833984375,
+ "learning_rate": 0.0005,
+ "loss": 1.04,
+ "step": 6810
+ },
+ {
+ "epoch": 21.76038338658147,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0406,
+ "step": 6811
+ },
+ {
+ "epoch": 21.763578274760384,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.034,
+ "step": 6812
+ },
+ {
+ "epoch": 21.766773162939298,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.035,
+ "step": 6813
+ },
+ {
+ "epoch": 21.769968051118212,
+ "grad_norm": 0.053955078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0297,
+ "step": 6814
+ },
+ {
+ "epoch": 21.773162939297123,
+ "grad_norm": 0.07861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.049,
+ "step": 6815
+ },
+ {
+ "epoch": 21.776357827476037,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0458,
+ "step": 6816
+ },
+ {
+ "epoch": 21.77955271565495,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0389,
+ "step": 6817
+ },
+ {
+ "epoch": 21.782747603833865,
+ "grad_norm": 0.0888671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0408,
+ "step": 6818
+ },
+ {
+ "epoch": 21.78594249201278,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0499,
+ "step": 6819
+ },
+ {
+ "epoch": 21.789137380191693,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0377,
+ "step": 6820
+ },
+ {
+ "epoch": 21.792332268370608,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.035,
+ "step": 6821
+ },
+ {
+ "epoch": 21.79552715654952,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0512,
+ "step": 6822
+ },
+ {
+ "epoch": 21.798722044728436,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0444,
+ "step": 6823
+ },
+ {
+ "epoch": 21.80191693290735,
+ "grad_norm": 0.059326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.035,
+ "step": 6824
+ },
+ {
+ "epoch": 21.80511182108626,
+ "grad_norm": 0.055908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0494,
+ "step": 6825
+ },
+ {
+ "epoch": 21.808306709265175,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0369,
+ "step": 6826
+ },
+ {
+ "epoch": 21.81150159744409,
+ "grad_norm": 0.1162109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0515,
+ "step": 6827
+ },
+ {
+ "epoch": 21.814696485623003,
+ "grad_norm": 0.0869140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0524,
+ "step": 6828
+ },
+ {
+ "epoch": 21.817891373801917,
+ "grad_norm": 0.10791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0493,
+ "step": 6829
+ },
+ {
+ "epoch": 21.82108626198083,
+ "grad_norm": 0.054931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0424,
+ "step": 6830
+ },
+ {
+ "epoch": 21.824281150159745,
+ "grad_norm": 0.0888671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0375,
+ "step": 6831
+ },
+ {
+ "epoch": 21.82747603833866,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0549,
+ "step": 6832
+ },
+ {
+ "epoch": 21.830670926517573,
+ "grad_norm": 0.055908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0466,
+ "step": 6833
+ },
+ {
+ "epoch": 21.833865814696484,
+ "grad_norm": 0.059326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0448,
+ "step": 6834
+ },
+ {
+ "epoch": 21.837060702875398,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.044,
+ "step": 6835
+ },
+ {
+ "epoch": 21.840255591054312,
+ "grad_norm": 0.047607421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0333,
+ "step": 6836
+ },
+ {
+ "epoch": 21.843450479233226,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0497,
+ "step": 6837
+ },
+ {
+ "epoch": 21.84664536741214,
+ "grad_norm": 0.08251953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0391,
+ "step": 6838
+ },
+ {
+ "epoch": 21.849840255591054,
+ "grad_norm": 0.058349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0467,
+ "step": 6839
+ },
+ {
+ "epoch": 21.85303514376997,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0452,
+ "step": 6840
+ },
+ {
+ "epoch": 21.856230031948883,
+ "grad_norm": 0.08349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0521,
+ "step": 6841
+ },
+ {
+ "epoch": 21.859424920127797,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0393,
+ "step": 6842
+ },
+ {
+ "epoch": 21.86261980830671,
+ "grad_norm": 0.057373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0457,
+ "step": 6843
+ },
+ {
+ "epoch": 21.86581469648562,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0531,
+ "step": 6844
+ },
+ {
+ "epoch": 21.869009584664536,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0437,
+ "step": 6845
+ },
+ {
+ "epoch": 21.87220447284345,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0468,
+ "step": 6846
+ },
+ {
+ "epoch": 21.875399361022364,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0463,
+ "step": 6847
+ },
+ {
+ "epoch": 21.878594249201278,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.038,
+ "step": 6848
+ },
+ {
+ "epoch": 21.881789137380192,
+ "grad_norm": 0.055908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0412,
+ "step": 6849
+ },
+ {
+ "epoch": 21.884984025559106,
+ "grad_norm": 0.05810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0479,
+ "step": 6850
+ },
+ {
+ "epoch": 21.88817891373802,
+ "grad_norm": 0.0537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0471,
+ "step": 6851
+ },
+ {
+ "epoch": 21.891373801916934,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0589,
+ "step": 6852
+ },
+ {
+ "epoch": 21.894568690095845,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0436,
+ "step": 6853
+ },
+ {
+ "epoch": 21.89776357827476,
+ "grad_norm": 0.08056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0518,
+ "step": 6854
+ },
+ {
+ "epoch": 21.900958466453673,
+ "grad_norm": 0.056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.05,
+ "step": 6855
+ },
+ {
+ "epoch": 21.904153354632587,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0481,
+ "step": 6856
+ },
+ {
+ "epoch": 21.9073482428115,
+ "grad_norm": 0.05908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0391,
+ "step": 6857
+ },
+ {
+ "epoch": 21.910543130990416,
+ "grad_norm": 0.11279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0397,
+ "step": 6858
+ },
+ {
+ "epoch": 21.91373801916933,
+ "grad_norm": 0.09912109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0345,
+ "step": 6859
+ },
+ {
+ "epoch": 21.916932907348244,
+ "grad_norm": 0.1044921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0364,
+ "step": 6860
+ },
+ {
+ "epoch": 21.920127795527158,
+ "grad_norm": 0.138671875,
+ "learning_rate": 0.0005,
+ "loss": 1.035,
+ "step": 6861
+ },
+ {
+ "epoch": 21.923322683706072,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0417,
+ "step": 6862
+ },
+ {
+ "epoch": 21.926517571884983,
+ "grad_norm": 0.2041015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0419,
+ "step": 6863
+ },
+ {
+ "epoch": 21.929712460063897,
+ "grad_norm": 0.11865234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0456,
+ "step": 6864
+ },
+ {
+ "epoch": 21.93290734824281,
+ "grad_norm": 0.11865234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0498,
+ "step": 6865
+ },
+ {
+ "epoch": 21.936102236421725,
+ "grad_norm": 0.15234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0504,
+ "step": 6866
+ },
+ {
+ "epoch": 21.93929712460064,
+ "grad_norm": 0.056396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0583,
+ "step": 6867
+ },
+ {
+ "epoch": 21.942492012779553,
+ "grad_norm": 0.1357421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0483,
+ "step": 6868
+ },
+ {
+ "epoch": 21.945686900958467,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.046,
+ "step": 6869
+ },
+ {
+ "epoch": 21.94888178913738,
+ "grad_norm": 0.13671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0515,
+ "step": 6870
+ },
+ {
+ "epoch": 21.952076677316295,
+ "grad_norm": 0.08251953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0348,
+ "step": 6871
+ },
+ {
+ "epoch": 21.955271565495206,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0398,
+ "step": 6872
+ },
+ {
+ "epoch": 21.95846645367412,
+ "grad_norm": 0.0546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0318,
+ "step": 6873
+ },
+ {
+ "epoch": 21.961661341853034,
+ "grad_norm": 0.1181640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0414,
+ "step": 6874
+ },
+ {
+ "epoch": 21.96485623003195,
+ "grad_norm": 0.0615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0444,
+ "step": 6875
+ },
+ {
+ "epoch": 21.968051118210862,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0299,
+ "step": 6876
+ },
+ {
+ "epoch": 21.971246006389777,
+ "grad_norm": 0.08056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0441,
+ "step": 6877
+ },
+ {
+ "epoch": 21.97444089456869,
+ "grad_norm": 0.0810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0419,
+ "step": 6878
+ },
+ {
+ "epoch": 21.977635782747605,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0438,
+ "step": 6879
+ },
+ {
+ "epoch": 21.98083067092652,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0499,
+ "step": 6880
+ },
+ {
+ "epoch": 21.984025559105433,
+ "grad_norm": 0.1279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0331,
+ "step": 6881
+ },
+ {
+ "epoch": 21.987220447284344,
+ "grad_norm": 0.055908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0404,
+ "step": 6882
+ },
+ {
+ "epoch": 21.990415335463258,
+ "grad_norm": 0.1083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0422,
+ "step": 6883
+ },
+ {
+ "epoch": 21.99361022364217,
+ "grad_norm": 0.05517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0314,
+ "step": 6884
+ },
+ {
+ "epoch": 21.996805111821086,
+ "grad_norm": 0.08935546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0443,
+ "step": 6885
+ },
+ {
+ "epoch": 22.0,
+ "grad_norm": 0.049072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0526,
+ "step": 6886
+ },
+ {
+ "epoch": 22.003194888178914,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0479,
+ "step": 6887
+ },
+ {
+ "epoch": 22.00638977635783,
+ "grad_norm": 0.056396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0302,
+ "step": 6888
+ },
+ {
+ "epoch": 22.009584664536742,
+ "grad_norm": 0.05810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0572,
+ "step": 6889
+ },
+ {
+ "epoch": 22.012779552715656,
+ "grad_norm": 0.0615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0447,
+ "step": 6890
+ },
+ {
+ "epoch": 22.015974440894567,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0482,
+ "step": 6891
+ },
+ {
+ "epoch": 22.01916932907348,
+ "grad_norm": 0.1337890625,
+ "learning_rate": 0.0005,
+ "loss": 1.05,
+ "step": 6892
+ },
+ {
+ "epoch": 22.022364217252395,
+ "grad_norm": 0.0927734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0283,
+ "step": 6893
+ },
+ {
+ "epoch": 22.02555910543131,
+ "grad_norm": 0.1865234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0468,
+ "step": 6894
+ },
+ {
+ "epoch": 22.028753993610223,
+ "grad_norm": 0.0810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0474,
+ "step": 6895
+ },
+ {
+ "epoch": 22.031948881789138,
+ "grad_norm": 0.1923828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0413,
+ "step": 6896
+ },
+ {
+ "epoch": 22.03514376996805,
+ "grad_norm": 0.08837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0451,
+ "step": 6897
+ },
+ {
+ "epoch": 22.038338658146966,
+ "grad_norm": 0.14453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0523,
+ "step": 6898
+ },
+ {
+ "epoch": 22.04153354632588,
+ "grad_norm": 0.150390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0341,
+ "step": 6899
+ },
+ {
+ "epoch": 22.044728434504794,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0397,
+ "step": 6900
+ },
+ {
+ "epoch": 22.047923322683705,
+ "grad_norm": 0.228515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0524,
+ "step": 6901
+ },
+ {
+ "epoch": 22.05111821086262,
+ "grad_norm": 0.248046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0449,
+ "step": 6902
+ },
+ {
+ "epoch": 22.054313099041533,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.033,
+ "step": 6903
+ },
+ {
+ "epoch": 22.057507987220447,
+ "grad_norm": 0.2275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0511,
+ "step": 6904
+ },
+ {
+ "epoch": 22.06070287539936,
+ "grad_norm": 0.08154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0318,
+ "step": 6905
+ },
+ {
+ "epoch": 22.063897763578275,
+ "grad_norm": 0.19921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0409,
+ "step": 6906
+ },
+ {
+ "epoch": 22.06709265175719,
+ "grad_norm": 0.158203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0539,
+ "step": 6907
+ },
+ {
+ "epoch": 22.070287539936103,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0548,
+ "step": 6908
+ },
+ {
+ "epoch": 22.073482428115017,
+ "grad_norm": 0.0947265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0383,
+ "step": 6909
+ },
+ {
+ "epoch": 22.076677316293928,
+ "grad_norm": 0.059814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0413,
+ "step": 6910
+ },
+ {
+ "epoch": 22.079872204472842,
+ "grad_norm": 0.08544921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0418,
+ "step": 6911
+ },
+ {
+ "epoch": 22.083067092651756,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0343,
+ "step": 6912
+ },
+ {
+ "epoch": 22.08626198083067,
+ "grad_norm": 0.11474609375,
+ "learning_rate": 0.0005,
+ "loss": 1.046,
+ "step": 6913
+ },
+ {
+ "epoch": 22.089456869009584,
+ "grad_norm": 0.0888671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0525,
+ "step": 6914
+ },
+ {
+ "epoch": 22.0926517571885,
+ "grad_norm": 0.1494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.058,
+ "step": 6915
+ },
+ {
+ "epoch": 22.095846645367413,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0538,
+ "step": 6916
+ },
+ {
+ "epoch": 22.099041533546327,
+ "grad_norm": 0.10546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0406,
+ "step": 6917
+ },
+ {
+ "epoch": 22.10223642172524,
+ "grad_norm": 0.10693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0391,
+ "step": 6918
+ },
+ {
+ "epoch": 22.105431309904155,
+ "grad_norm": 0.1318359375,
+ "learning_rate": 0.0005,
+ "loss": 1.047,
+ "step": 6919
+ },
+ {
+ "epoch": 22.108626198083066,
+ "grad_norm": 0.11328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0505,
+ "step": 6920
+ },
+ {
+ "epoch": 22.11182108626198,
+ "grad_norm": 0.15625,
+ "learning_rate": 0.0005,
+ "loss": 1.0434,
+ "step": 6921
+ },
+ {
+ "epoch": 22.115015974440894,
+ "grad_norm": 0.0810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0465,
+ "step": 6922
+ },
+ {
+ "epoch": 22.118210862619808,
+ "grad_norm": 0.150390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0496,
+ "step": 6923
+ },
+ {
+ "epoch": 22.121405750798722,
+ "grad_norm": 0.20703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0479,
+ "step": 6924
+ },
+ {
+ "epoch": 22.124600638977636,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0262,
+ "step": 6925
+ },
+ {
+ "epoch": 22.12779552715655,
+ "grad_norm": 0.240234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0368,
+ "step": 6926
+ },
+ {
+ "epoch": 22.130990415335464,
+ "grad_norm": 0.1044921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0317,
+ "step": 6927
+ },
+ {
+ "epoch": 22.13418530351438,
+ "grad_norm": 0.1416015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0516,
+ "step": 6928
+ },
+ {
+ "epoch": 22.13738019169329,
+ "grad_norm": 0.1396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0518,
+ "step": 6929
+ },
+ {
+ "epoch": 22.140575079872203,
+ "grad_norm": 0.07861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0503,
+ "step": 6930
+ },
+ {
+ "epoch": 22.143769968051117,
+ "grad_norm": 0.134765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0462,
+ "step": 6931
+ },
+ {
+ "epoch": 22.14696485623003,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0375,
+ "step": 6932
+ },
+ {
+ "epoch": 22.150159744408946,
+ "grad_norm": 0.16796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0546,
+ "step": 6933
+ },
+ {
+ "epoch": 22.15335463258786,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0511,
+ "step": 6934
+ },
+ {
+ "epoch": 22.156549520766774,
+ "grad_norm": 0.08837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0471,
+ "step": 6935
+ },
+ {
+ "epoch": 22.159744408945688,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0545,
+ "step": 6936
+ },
+ {
+ "epoch": 22.162939297124602,
+ "grad_norm": 0.109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0449,
+ "step": 6937
+ },
+ {
+ "epoch": 22.166134185303516,
+ "grad_norm": 0.1240234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0438,
+ "step": 6938
+ },
+ {
+ "epoch": 22.169329073482427,
+ "grad_norm": 0.10986328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0486,
+ "step": 6939
+ },
+ {
+ "epoch": 22.17252396166134,
+ "grad_norm": 0.19140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0497,
+ "step": 6940
+ },
+ {
+ "epoch": 22.175718849840255,
+ "grad_norm": 0.09619140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0417,
+ "step": 6941
+ },
+ {
+ "epoch": 22.17891373801917,
+ "grad_norm": 0.263671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0425,
+ "step": 6942
+ },
+ {
+ "epoch": 22.182108626198083,
+ "grad_norm": 0.11865234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0464,
+ "step": 6943
+ },
+ {
+ "epoch": 22.185303514376997,
+ "grad_norm": 0.1484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0526,
+ "step": 6944
+ },
+ {
+ "epoch": 22.18849840255591,
+ "grad_norm": 0.1357421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0395,
+ "step": 6945
+ },
+ {
+ "epoch": 22.191693290734825,
+ "grad_norm": 0.1005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0343,
+ "step": 6946
+ },
+ {
+ "epoch": 22.19488817891374,
+ "grad_norm": 0.15625,
+ "learning_rate": 0.0005,
+ "loss": 1.0425,
+ "step": 6947
+ },
+ {
+ "epoch": 22.19808306709265,
+ "grad_norm": 0.056396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0409,
+ "step": 6948
+ },
+ {
+ "epoch": 22.201277955271564,
+ "grad_norm": 0.193359375,
+ "learning_rate": 0.0005,
+ "loss": 1.045,
+ "step": 6949
+ },
+ {
+ "epoch": 22.20447284345048,
+ "grad_norm": 0.16796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0468,
+ "step": 6950
+ },
+ {
+ "epoch": 22.207667731629392,
+ "grad_norm": 0.08837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0371,
+ "step": 6951
+ },
+ {
+ "epoch": 22.210862619808307,
+ "grad_norm": 0.220703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0288,
+ "step": 6952
+ },
+ {
+ "epoch": 22.21405750798722,
+ "grad_norm": 0.055908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0388,
+ "step": 6953
+ },
+ {
+ "epoch": 22.217252396166135,
+ "grad_norm": 0.201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0477,
+ "step": 6954
+ },
+ {
+ "epoch": 22.22044728434505,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0544,
+ "step": 6955
+ },
+ {
+ "epoch": 22.223642172523963,
+ "grad_norm": 0.146484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0337,
+ "step": 6956
+ },
+ {
+ "epoch": 22.226837060702877,
+ "grad_norm": 0.10009765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0523,
+ "step": 6957
+ },
+ {
+ "epoch": 22.230031948881788,
+ "grad_norm": 0.16015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0384,
+ "step": 6958
+ },
+ {
+ "epoch": 22.233226837060702,
+ "grad_norm": 0.1142578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0421,
+ "step": 6959
+ },
+ {
+ "epoch": 22.236421725239616,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0571,
+ "step": 6960
+ },
+ {
+ "epoch": 22.23961661341853,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0431,
+ "step": 6961
+ },
+ {
+ "epoch": 22.242811501597444,
+ "grad_norm": 0.0546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0452,
+ "step": 6962
+ },
+ {
+ "epoch": 22.24600638977636,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0443,
+ "step": 6963
+ },
+ {
+ "epoch": 22.249201277955272,
+ "grad_norm": 0.095703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0363,
+ "step": 6964
+ },
+ {
+ "epoch": 22.252396166134186,
+ "grad_norm": 0.060791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.045,
+ "step": 6965
+ },
+ {
+ "epoch": 22.2555910543131,
+ "grad_norm": 0.09619140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0429,
+ "step": 6966
+ },
+ {
+ "epoch": 22.25878594249201,
+ "grad_norm": 0.087890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0407,
+ "step": 6967
+ },
+ {
+ "epoch": 22.261980830670925,
+ "grad_norm": 0.0986328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0412,
+ "step": 6968
+ },
+ {
+ "epoch": 22.26517571884984,
+ "grad_norm": 0.10302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0439,
+ "step": 6969
+ },
+ {
+ "epoch": 22.268370607028753,
+ "grad_norm": 0.1318359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0448,
+ "step": 6970
+ },
+ {
+ "epoch": 22.271565495207668,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0406,
+ "step": 6971
+ },
+ {
+ "epoch": 22.27476038338658,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0454,
+ "step": 6972
+ },
+ {
+ "epoch": 22.277955271565496,
+ "grad_norm": 0.057861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0522,
+ "step": 6973
+ },
+ {
+ "epoch": 22.28115015974441,
+ "grad_norm": 0.0908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0403,
+ "step": 6974
+ },
+ {
+ "epoch": 22.284345047923324,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0438,
+ "step": 6975
+ },
+ {
+ "epoch": 22.287539936102238,
+ "grad_norm": 0.1318359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0502,
+ "step": 6976
+ },
+ {
+ "epoch": 22.29073482428115,
+ "grad_norm": 0.119140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0381,
+ "step": 6977
+ },
+ {
+ "epoch": 22.293929712460063,
+ "grad_norm": 0.1669921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0384,
+ "step": 6978
+ },
+ {
+ "epoch": 22.297124600638977,
+ "grad_norm": 0.10595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0419,
+ "step": 6979
+ },
+ {
+ "epoch": 22.30031948881789,
+ "grad_norm": 0.0947265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0279,
+ "step": 6980
+ },
+ {
+ "epoch": 22.303514376996805,
+ "grad_norm": 0.1259765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0347,
+ "step": 6981
+ },
+ {
+ "epoch": 22.30670926517572,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0519,
+ "step": 6982
+ },
+ {
+ "epoch": 22.309904153354633,
+ "grad_norm": 0.08154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0482,
+ "step": 6983
+ },
+ {
+ "epoch": 22.313099041533548,
+ "grad_norm": 0.08056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0466,
+ "step": 6984
+ },
+ {
+ "epoch": 22.31629392971246,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.054,
+ "step": 6985
+ },
+ {
+ "epoch": 22.319488817891372,
+ "grad_norm": 0.08203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0437,
+ "step": 6986
+ },
+ {
+ "epoch": 22.322683706070286,
+ "grad_norm": 0.11279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0524,
+ "step": 6987
+ },
+ {
+ "epoch": 22.3258785942492,
+ "grad_norm": 0.060791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0451,
+ "step": 6988
+ },
+ {
+ "epoch": 22.329073482428115,
+ "grad_norm": 0.12353515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0443,
+ "step": 6989
+ },
+ {
+ "epoch": 22.33226837060703,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0505,
+ "step": 6990
+ },
+ {
+ "epoch": 22.335463258785943,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.048,
+ "step": 6991
+ },
+ {
+ "epoch": 22.338658146964857,
+ "grad_norm": 0.08837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0433,
+ "step": 6992
+ },
+ {
+ "epoch": 22.34185303514377,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0479,
+ "step": 6993
+ },
+ {
+ "epoch": 22.345047923322685,
+ "grad_norm": 0.09521484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0428,
+ "step": 6994
+ },
+ {
+ "epoch": 22.3482428115016,
+ "grad_norm": 0.10009765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0466,
+ "step": 6995
+ },
+ {
+ "epoch": 22.35143769968051,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0325,
+ "step": 6996
+ },
+ {
+ "epoch": 22.354632587859424,
+ "grad_norm": 0.0498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.033,
+ "step": 6997
+ },
+ {
+ "epoch": 22.357827476038338,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0496,
+ "step": 6998
+ },
+ {
+ "epoch": 22.361022364217252,
+ "grad_norm": 0.060302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0574,
+ "step": 6999
+ },
+ {
+ "epoch": 22.364217252396166,
+ "grad_norm": 0.05859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0392,
+ "step": 7000
+ },
+ {
+ "epoch": 22.36741214057508,
+ "grad_norm": 0.055908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0438,
+ "step": 7001
+ },
+ {
+ "epoch": 22.370607028753994,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0475,
+ "step": 7002
+ },
+ {
+ "epoch": 22.37380191693291,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0471,
+ "step": 7003
+ },
+ {
+ "epoch": 22.376996805111823,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.058,
+ "step": 7004
+ },
+ {
+ "epoch": 22.380191693290733,
+ "grad_norm": 0.10888671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0464,
+ "step": 7005
+ },
+ {
+ "epoch": 22.383386581469647,
+ "grad_norm": 0.12060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.048,
+ "step": 7006
+ },
+ {
+ "epoch": 22.38658146964856,
+ "grad_norm": 0.138671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0384,
+ "step": 7007
+ },
+ {
+ "epoch": 22.389776357827476,
+ "grad_norm": 0.107421875,
+ "learning_rate": 0.0005,
+ "loss": 1.046,
+ "step": 7008
+ },
+ {
+ "epoch": 22.39297124600639,
+ "grad_norm": 0.0869140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0413,
+ "step": 7009
+ },
+ {
+ "epoch": 22.396166134185304,
+ "grad_norm": 0.10791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0405,
+ "step": 7010
+ },
+ {
+ "epoch": 22.399361022364218,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0275,
+ "step": 7011
+ },
+ {
+ "epoch": 22.402555910543132,
+ "grad_norm": 0.05810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0486,
+ "step": 7012
+ },
+ {
+ "epoch": 22.405750798722046,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0337,
+ "step": 7013
+ },
+ {
+ "epoch": 22.408945686900957,
+ "grad_norm": 0.10546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0402,
+ "step": 7014
+ },
+ {
+ "epoch": 22.41214057507987,
+ "grad_norm": 0.09326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0381,
+ "step": 7015
+ },
+ {
+ "epoch": 22.415335463258785,
+ "grad_norm": 0.0859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0246,
+ "step": 7016
+ },
+ {
+ "epoch": 22.4185303514377,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0401,
+ "step": 7017
+ },
+ {
+ "epoch": 22.421725239616613,
+ "grad_norm": 0.1171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0383,
+ "step": 7018
+ },
+ {
+ "epoch": 22.424920127795527,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0502,
+ "step": 7019
+ },
+ {
+ "epoch": 22.42811501597444,
+ "grad_norm": 0.055419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0531,
+ "step": 7020
+ },
+ {
+ "epoch": 22.431309904153355,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0492,
+ "step": 7021
+ },
+ {
+ "epoch": 22.43450479233227,
+ "grad_norm": 0.053466796875,
+ "learning_rate": 0.0005,
+ "loss": 1.044,
+ "step": 7022
+ },
+ {
+ "epoch": 22.437699680511184,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0406,
+ "step": 7023
+ },
+ {
+ "epoch": 22.440894568690094,
+ "grad_norm": 0.05517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0362,
+ "step": 7024
+ },
+ {
+ "epoch": 22.44408945686901,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0305,
+ "step": 7025
+ },
+ {
+ "epoch": 22.447284345047922,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0302,
+ "step": 7026
+ },
+ {
+ "epoch": 22.450479233226837,
+ "grad_norm": 0.080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0393,
+ "step": 7027
+ },
+ {
+ "epoch": 22.45367412140575,
+ "grad_norm": 0.055419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0408,
+ "step": 7028
+ },
+ {
+ "epoch": 22.456869009584665,
+ "grad_norm": 0.083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0344,
+ "step": 7029
+ },
+ {
+ "epoch": 22.46006389776358,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.047,
+ "step": 7030
+ },
+ {
+ "epoch": 22.463258785942493,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0367,
+ "step": 7031
+ },
+ {
+ "epoch": 22.466453674121407,
+ "grad_norm": 0.0615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0506,
+ "step": 7032
+ },
+ {
+ "epoch": 22.46964856230032,
+ "grad_norm": 0.0986328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0573,
+ "step": 7033
+ },
+ {
+ "epoch": 22.472843450479232,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0395,
+ "step": 7034
+ },
+ {
+ "epoch": 22.476038338658146,
+ "grad_norm": 0.10595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0467,
+ "step": 7035
+ },
+ {
+ "epoch": 22.47923322683706,
+ "grad_norm": 0.09228515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0424,
+ "step": 7036
+ },
+ {
+ "epoch": 22.482428115015974,
+ "grad_norm": 0.134765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0343,
+ "step": 7037
+ },
+ {
+ "epoch": 22.48562300319489,
+ "grad_norm": 0.06005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.037,
+ "step": 7038
+ },
+ {
+ "epoch": 22.488817891373802,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0559,
+ "step": 7039
+ },
+ {
+ "epoch": 22.492012779552716,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0415,
+ "step": 7040
+ },
+ {
+ "epoch": 22.49520766773163,
+ "grad_norm": 0.056884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0471,
+ "step": 7041
+ },
+ {
+ "epoch": 22.498402555910545,
+ "grad_norm": 0.046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0413,
+ "step": 7042
+ },
+ {
+ "epoch": 22.501597444089455,
+ "grad_norm": 0.05712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0459,
+ "step": 7043
+ },
+ {
+ "epoch": 22.50479233226837,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0328,
+ "step": 7044
+ },
+ {
+ "epoch": 22.507987220447284,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0512,
+ "step": 7045
+ },
+ {
+ "epoch": 22.511182108626198,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0477,
+ "step": 7046
+ },
+ {
+ "epoch": 22.51437699680511,
+ "grad_norm": 0.083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0446,
+ "step": 7047
+ },
+ {
+ "epoch": 22.517571884984026,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0395,
+ "step": 7048
+ },
+ {
+ "epoch": 22.52076677316294,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0318,
+ "step": 7049
+ },
+ {
+ "epoch": 22.523961661341854,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0429,
+ "step": 7050
+ },
+ {
+ "epoch": 22.527156549520768,
+ "grad_norm": 0.0859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0451,
+ "step": 7051
+ },
+ {
+ "epoch": 22.53035143769968,
+ "grad_norm": 0.11376953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0422,
+ "step": 7052
+ },
+ {
+ "epoch": 22.533546325878593,
+ "grad_norm": 0.10595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0341,
+ "step": 7053
+ },
+ {
+ "epoch": 22.536741214057507,
+ "grad_norm": 0.08447265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0579,
+ "step": 7054
+ },
+ {
+ "epoch": 22.53993610223642,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.041,
+ "step": 7055
+ },
+ {
+ "epoch": 22.543130990415335,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.059,
+ "step": 7056
+ },
+ {
+ "epoch": 22.54632587859425,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0519,
+ "step": 7057
+ },
+ {
+ "epoch": 22.549520766773163,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0476,
+ "step": 7058
+ },
+ {
+ "epoch": 22.552715654952078,
+ "grad_norm": 0.0546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0412,
+ "step": 7059
+ },
+ {
+ "epoch": 22.55591054313099,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0544,
+ "step": 7060
+ },
+ {
+ "epoch": 22.559105431309906,
+ "grad_norm": 0.05908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0438,
+ "step": 7061
+ },
+ {
+ "epoch": 22.562300319488816,
+ "grad_norm": 0.05078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0489,
+ "step": 7062
+ },
+ {
+ "epoch": 22.56549520766773,
+ "grad_norm": 0.05419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0377,
+ "step": 7063
+ },
+ {
+ "epoch": 22.568690095846645,
+ "grad_norm": 0.059326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0396,
+ "step": 7064
+ },
+ {
+ "epoch": 22.57188498402556,
+ "grad_norm": 0.08447265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0428,
+ "step": 7065
+ },
+ {
+ "epoch": 22.575079872204473,
+ "grad_norm": 0.053955078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0462,
+ "step": 7066
+ },
+ {
+ "epoch": 22.578274760383387,
+ "grad_norm": 0.045166015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0427,
+ "step": 7067
+ },
+ {
+ "epoch": 22.5814696485623,
+ "grad_norm": 0.056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0469,
+ "step": 7068
+ },
+ {
+ "epoch": 22.584664536741215,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0471,
+ "step": 7069
+ },
+ {
+ "epoch": 22.58785942492013,
+ "grad_norm": 0.05029296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0498,
+ "step": 7070
+ },
+ {
+ "epoch": 22.591054313099043,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0446,
+ "step": 7071
+ },
+ {
+ "epoch": 22.594249201277954,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0453,
+ "step": 7072
+ },
+ {
+ "epoch": 22.597444089456868,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0382,
+ "step": 7073
+ },
+ {
+ "epoch": 22.600638977635782,
+ "grad_norm": 0.045654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0333,
+ "step": 7074
+ },
+ {
+ "epoch": 22.603833865814696,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0377,
+ "step": 7075
+ },
+ {
+ "epoch": 22.60702875399361,
+ "grad_norm": 0.051025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0476,
+ "step": 7076
+ },
+ {
+ "epoch": 22.610223642172524,
+ "grad_norm": 0.0615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0454,
+ "step": 7077
+ },
+ {
+ "epoch": 22.61341853035144,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0367,
+ "step": 7078
+ },
+ {
+ "epoch": 22.616613418530353,
+ "grad_norm": 0.107421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0471,
+ "step": 7079
+ },
+ {
+ "epoch": 22.619808306709267,
+ "grad_norm": 0.046142578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0324,
+ "step": 7080
+ },
+ {
+ "epoch": 22.623003194888177,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0397,
+ "step": 7081
+ },
+ {
+ "epoch": 22.62619808306709,
+ "grad_norm": 0.0615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0429,
+ "step": 7082
+ },
+ {
+ "epoch": 22.629392971246006,
+ "grad_norm": 0.0537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0302,
+ "step": 7083
+ },
+ {
+ "epoch": 22.63258785942492,
+ "grad_norm": 0.1083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0539,
+ "step": 7084
+ },
+ {
+ "epoch": 22.635782747603834,
+ "grad_norm": 0.05615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0479,
+ "step": 7085
+ },
+ {
+ "epoch": 22.638977635782748,
+ "grad_norm": 0.06201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.051,
+ "step": 7086
+ },
+ {
+ "epoch": 22.642172523961662,
+ "grad_norm": 0.056884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.047,
+ "step": 7087
+ },
+ {
+ "epoch": 22.645367412140576,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.046,
+ "step": 7088
+ },
+ {
+ "epoch": 22.64856230031949,
+ "grad_norm": 0.095703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0498,
+ "step": 7089
+ },
+ {
+ "epoch": 22.6517571884984,
+ "grad_norm": 0.1015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0414,
+ "step": 7090
+ },
+ {
+ "epoch": 22.654952076677315,
+ "grad_norm": 0.1142578125,
+ "learning_rate": 0.0005,
+ "loss": 1.033,
+ "step": 7091
+ },
+ {
+ "epoch": 22.65814696485623,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0427,
+ "step": 7092
+ },
+ {
+ "epoch": 22.661341853035143,
+ "grad_norm": 0.0869140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0478,
+ "step": 7093
+ },
+ {
+ "epoch": 22.664536741214057,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0514,
+ "step": 7094
+ },
+ {
+ "epoch": 22.66773162939297,
+ "grad_norm": 0.0908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0368,
+ "step": 7095
+ },
+ {
+ "epoch": 22.670926517571885,
+ "grad_norm": 0.059326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0455,
+ "step": 7096
+ },
+ {
+ "epoch": 22.6741214057508,
+ "grad_norm": 0.09228515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0545,
+ "step": 7097
+ },
+ {
+ "epoch": 22.677316293929714,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0523,
+ "step": 7098
+ },
+ {
+ "epoch": 22.680511182108628,
+ "grad_norm": 0.0849609375,
+ "learning_rate": 0.0005,
+ "loss": 1.033,
+ "step": 7099
+ },
+ {
+ "epoch": 22.68370607028754,
+ "grad_norm": 0.08935546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0523,
+ "step": 7100
+ },
+ {
+ "epoch": 22.686900958466452,
+ "grad_norm": 0.060791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0404,
+ "step": 7101
+ },
+ {
+ "epoch": 22.690095846645367,
+ "grad_norm": 0.053955078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0448,
+ "step": 7102
+ },
+ {
+ "epoch": 22.69329073482428,
+ "grad_norm": 0.0859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0345,
+ "step": 7103
+ },
+ {
+ "epoch": 22.696485623003195,
+ "grad_norm": 0.0908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0414,
+ "step": 7104
+ },
+ {
+ "epoch": 22.69968051118211,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.04,
+ "step": 7105
+ },
+ {
+ "epoch": 22.702875399361023,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0452,
+ "step": 7106
+ },
+ {
+ "epoch": 22.706070287539937,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0416,
+ "step": 7107
+ },
+ {
+ "epoch": 22.70926517571885,
+ "grad_norm": 0.08349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0451,
+ "step": 7108
+ },
+ {
+ "epoch": 22.712460063897765,
+ "grad_norm": 0.0859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0386,
+ "step": 7109
+ },
+ {
+ "epoch": 22.715654952076676,
+ "grad_norm": 0.0986328125,
+ "learning_rate": 0.0005,
+ "loss": 1.048,
+ "step": 7110
+ },
+ {
+ "epoch": 22.71884984025559,
+ "grad_norm": 0.07861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0453,
+ "step": 7111
+ },
+ {
+ "epoch": 22.722044728434504,
+ "grad_norm": 0.0810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0478,
+ "step": 7112
+ },
+ {
+ "epoch": 22.72523961661342,
+ "grad_norm": 0.08251953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0262,
+ "step": 7113
+ },
+ {
+ "epoch": 22.728434504792332,
+ "grad_norm": 0.053466796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0404,
+ "step": 7114
+ },
+ {
+ "epoch": 22.731629392971247,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0483,
+ "step": 7115
+ },
+ {
+ "epoch": 22.73482428115016,
+ "grad_norm": 0.0908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0432,
+ "step": 7116
+ },
+ {
+ "epoch": 22.738019169329075,
+ "grad_norm": 0.052490234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0382,
+ "step": 7117
+ },
+ {
+ "epoch": 22.74121405750799,
+ "grad_norm": 0.0849609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0458,
+ "step": 7118
+ },
+ {
+ "epoch": 22.7444089456869,
+ "grad_norm": 0.060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0444,
+ "step": 7119
+ },
+ {
+ "epoch": 22.747603833865814,
+ "grad_norm": 0.060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0431,
+ "step": 7120
+ },
+ {
+ "epoch": 22.750798722044728,
+ "grad_norm": 0.052490234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0509,
+ "step": 7121
+ },
+ {
+ "epoch": 22.75399361022364,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0361,
+ "step": 7122
+ },
+ {
+ "epoch": 22.757188498402556,
+ "grad_norm": 0.055419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0397,
+ "step": 7123
+ },
+ {
+ "epoch": 22.76038338658147,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0329,
+ "step": 7124
+ },
+ {
+ "epoch": 22.763578274760384,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0417,
+ "step": 7125
+ },
+ {
+ "epoch": 22.766773162939298,
+ "grad_norm": 0.054931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0237,
+ "step": 7126
+ },
+ {
+ "epoch": 22.769968051118212,
+ "grad_norm": 0.083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0542,
+ "step": 7127
+ },
+ {
+ "epoch": 22.773162939297123,
+ "grad_norm": 0.0478515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0482,
+ "step": 7128
+ },
+ {
+ "epoch": 22.776357827476037,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0415,
+ "step": 7129
+ },
+ {
+ "epoch": 22.77955271565495,
+ "grad_norm": 0.10205078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0489,
+ "step": 7130
+ },
+ {
+ "epoch": 22.782747603833865,
+ "grad_norm": 0.055419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.039,
+ "step": 7131
+ },
+ {
+ "epoch": 22.78594249201278,
+ "grad_norm": 0.0830078125,
+ "learning_rate": 0.0005,
+ "loss": 1.032,
+ "step": 7132
+ },
+ {
+ "epoch": 22.789137380191693,
+ "grad_norm": 0.0908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0476,
+ "step": 7133
+ },
+ {
+ "epoch": 22.792332268370608,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0549,
+ "step": 7134
+ },
+ {
+ "epoch": 22.79552715654952,
+ "grad_norm": 0.10107421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0395,
+ "step": 7135
+ },
+ {
+ "epoch": 22.798722044728436,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0482,
+ "step": 7136
+ },
+ {
+ "epoch": 22.80191693290735,
+ "grad_norm": 0.056396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0427,
+ "step": 7137
+ },
+ {
+ "epoch": 22.80511182108626,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0465,
+ "step": 7138
+ },
+ {
+ "epoch": 22.808306709265175,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0488,
+ "step": 7139
+ },
+ {
+ "epoch": 22.81150159744409,
+ "grad_norm": 0.0556640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0518,
+ "step": 7140
+ },
+ {
+ "epoch": 22.814696485623003,
+ "grad_norm": 0.099609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0536,
+ "step": 7141
+ },
+ {
+ "epoch": 22.817891373801917,
+ "grad_norm": 0.05517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0449,
+ "step": 7142
+ },
+ {
+ "epoch": 22.82108626198083,
+ "grad_norm": 0.08203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0499,
+ "step": 7143
+ },
+ {
+ "epoch": 22.824281150159745,
+ "grad_norm": 0.0615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0461,
+ "step": 7144
+ },
+ {
+ "epoch": 22.82747603833866,
+ "grad_norm": 0.055419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0526,
+ "step": 7145
+ },
+ {
+ "epoch": 22.830670926517573,
+ "grad_norm": 0.060302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0381,
+ "step": 7146
+ },
+ {
+ "epoch": 22.833865814696484,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0391,
+ "step": 7147
+ },
+ {
+ "epoch": 22.837060702875398,
+ "grad_norm": 0.0517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0382,
+ "step": 7148
+ },
+ {
+ "epoch": 22.840255591054312,
+ "grad_norm": 0.052978515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0476,
+ "step": 7149
+ },
+ {
+ "epoch": 22.843450479233226,
+ "grad_norm": 0.052490234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0487,
+ "step": 7150
+ },
+ {
+ "epoch": 22.84664536741214,
+ "grad_norm": 0.061767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0472,
+ "step": 7151
+ },
+ {
+ "epoch": 22.849840255591054,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0384,
+ "step": 7152
+ },
+ {
+ "epoch": 22.85303514376997,
+ "grad_norm": 0.047607421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0457,
+ "step": 7153
+ },
+ {
+ "epoch": 22.856230031948883,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0378,
+ "step": 7154
+ },
+ {
+ "epoch": 22.859424920127797,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.049,
+ "step": 7155
+ },
+ {
+ "epoch": 22.86261980830671,
+ "grad_norm": 0.05517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0489,
+ "step": 7156
+ },
+ {
+ "epoch": 22.86581469648562,
+ "grad_norm": 0.046630859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0407,
+ "step": 7157
+ },
+ {
+ "epoch": 22.869009584664536,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0478,
+ "step": 7158
+ },
+ {
+ "epoch": 22.87220447284345,
+ "grad_norm": 0.04541015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0224,
+ "step": 7159
+ },
+ {
+ "epoch": 22.875399361022364,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.041,
+ "step": 7160
+ },
+ {
+ "epoch": 22.878594249201278,
+ "grad_norm": 0.056396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0496,
+ "step": 7161
+ },
+ {
+ "epoch": 22.881789137380192,
+ "grad_norm": 0.12353515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0463,
+ "step": 7162
+ },
+ {
+ "epoch": 22.884984025559106,
+ "grad_norm": 0.1142578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0401,
+ "step": 7163
+ },
+ {
+ "epoch": 22.88817891373802,
+ "grad_norm": 0.11962890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0364,
+ "step": 7164
+ },
+ {
+ "epoch": 22.891373801916934,
+ "grad_norm": 0.1708984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0288,
+ "step": 7165
+ },
+ {
+ "epoch": 22.894568690095845,
+ "grad_norm": 0.11865234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0508,
+ "step": 7166
+ },
+ {
+ "epoch": 22.89776357827476,
+ "grad_norm": 0.228515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0271,
+ "step": 7167
+ },
+ {
+ "epoch": 22.900958466453673,
+ "grad_norm": 0.12353515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0453,
+ "step": 7168
+ },
+ {
+ "epoch": 22.904153354632587,
+ "grad_norm": 0.125,
+ "learning_rate": 0.0005,
+ "loss": 1.0538,
+ "step": 7169
+ },
+ {
+ "epoch": 22.9073482428115,
+ "grad_norm": 0.19140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0436,
+ "step": 7170
+ },
+ {
+ "epoch": 22.910543130990416,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0531,
+ "step": 7171
+ },
+ {
+ "epoch": 22.91373801916933,
+ "grad_norm": 0.2216796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0517,
+ "step": 7172
+ },
+ {
+ "epoch": 22.916932907348244,
+ "grad_norm": 0.2333984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0596,
+ "step": 7173
+ },
+ {
+ "epoch": 22.920127795527158,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0657,
+ "step": 7174
+ },
+ {
+ "epoch": 22.923322683706072,
+ "grad_norm": 0.1767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0529,
+ "step": 7175
+ },
+ {
+ "epoch": 22.926517571884983,
+ "grad_norm": 0.08056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0547,
+ "step": 7176
+ },
+ {
+ "epoch": 22.929712460063897,
+ "grad_norm": 0.09619140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0417,
+ "step": 7177
+ },
+ {
+ "epoch": 22.93290734824281,
+ "grad_norm": 0.062255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0527,
+ "step": 7178
+ },
+ {
+ "epoch": 22.936102236421725,
+ "grad_norm": 0.10107421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0489,
+ "step": 7179
+ },
+ {
+ "epoch": 22.93929712460064,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0485,
+ "step": 7180
+ },
+ {
+ "epoch": 22.942492012779553,
+ "grad_norm": 0.1142578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0508,
+ "step": 7181
+ },
+ {
+ "epoch": 22.945686900958467,
+ "grad_norm": 0.11767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0475,
+ "step": 7182
+ },
+ {
+ "epoch": 22.94888178913738,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0452,
+ "step": 7183
+ },
+ {
+ "epoch": 22.952076677316295,
+ "grad_norm": 0.09326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0401,
+ "step": 7184
+ },
+ {
+ "epoch": 22.955271565495206,
+ "grad_norm": 0.05810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0398,
+ "step": 7185
+ },
+ {
+ "epoch": 22.95846645367412,
+ "grad_norm": 0.08935546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0367,
+ "step": 7186
+ },
+ {
+ "epoch": 22.961661341853034,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0507,
+ "step": 7187
+ },
+ {
+ "epoch": 22.96485623003195,
+ "grad_norm": 0.11767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0454,
+ "step": 7188
+ },
+ {
+ "epoch": 22.968051118210862,
+ "grad_norm": 0.08837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0416,
+ "step": 7189
+ },
+ {
+ "epoch": 22.971246006389777,
+ "grad_norm": 0.1357421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0513,
+ "step": 7190
+ },
+ {
+ "epoch": 22.97444089456869,
+ "grad_norm": 0.091796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0395,
+ "step": 7191
+ },
+ {
+ "epoch": 22.977635782747605,
+ "grad_norm": 0.1484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0476,
+ "step": 7192
+ },
+ {
+ "epoch": 22.98083067092652,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0432,
+ "step": 7193
+ },
+ {
+ "epoch": 22.984025559105433,
+ "grad_norm": 0.1181640625,
+ "learning_rate": 0.0005,
+ "loss": 1.05,
+ "step": 7194
+ },
+ {
+ "epoch": 22.987220447284344,
+ "grad_norm": 0.083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0508,
+ "step": 7195
+ },
+ {
+ "epoch": 22.990415335463258,
+ "grad_norm": 0.08447265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0624,
+ "step": 7196
+ },
+ {
+ "epoch": 22.99361022364217,
+ "grad_norm": 0.07861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0416,
+ "step": 7197
+ },
+ {
+ "epoch": 22.996805111821086,
+ "grad_norm": 0.0810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0349,
+ "step": 7198
+ },
+ {
+ "epoch": 23.0,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0336,
+ "step": 7199
+ },
+ {
+ "epoch": 23.003194888178914,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0474,
+ "step": 7200
+ },
+ {
+ "epoch": 23.00638977635783,
+ "grad_norm": 0.058837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0388,
+ "step": 7201
+ },
+ {
+ "epoch": 23.009584664536742,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0398,
+ "step": 7202
+ },
+ {
+ "epoch": 23.012779552715656,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0332,
+ "step": 7203
+ },
+ {
+ "epoch": 23.015974440894567,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0557,
+ "step": 7204
+ },
+ {
+ "epoch": 23.01916932907348,
+ "grad_norm": 0.09375,
+ "learning_rate": 0.0005,
+ "loss": 1.0414,
+ "step": 7205
+ },
+ {
+ "epoch": 23.022364217252395,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0481,
+ "step": 7206
+ },
+ {
+ "epoch": 23.02555910543131,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0525,
+ "step": 7207
+ },
+ {
+ "epoch": 23.028753993610223,
+ "grad_norm": 0.11474609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0322,
+ "step": 7208
+ },
+ {
+ "epoch": 23.031948881789138,
+ "grad_norm": 0.1044921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0413,
+ "step": 7209
+ },
+ {
+ "epoch": 23.03514376996805,
+ "grad_norm": 0.0546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0498,
+ "step": 7210
+ },
+ {
+ "epoch": 23.038338658146966,
+ "grad_norm": 0.0830078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0423,
+ "step": 7211
+ },
+ {
+ "epoch": 23.04153354632588,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0384,
+ "step": 7212
+ },
+ {
+ "epoch": 23.044728434504794,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0433,
+ "step": 7213
+ },
+ {
+ "epoch": 23.047923322683705,
+ "grad_norm": 0.0986328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0501,
+ "step": 7214
+ },
+ {
+ "epoch": 23.05111821086262,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0495,
+ "step": 7215
+ },
+ {
+ "epoch": 23.054313099041533,
+ "grad_norm": 0.08642578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0383,
+ "step": 7216
+ },
+ {
+ "epoch": 23.057507987220447,
+ "grad_norm": 0.09619140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0465,
+ "step": 7217
+ },
+ {
+ "epoch": 23.06070287539936,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0342,
+ "step": 7218
+ },
+ {
+ "epoch": 23.063897763578275,
+ "grad_norm": 0.1083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0434,
+ "step": 7219
+ },
+ {
+ "epoch": 23.06709265175719,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0531,
+ "step": 7220
+ },
+ {
+ "epoch": 23.070287539936103,
+ "grad_norm": 0.061767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.036,
+ "step": 7221
+ },
+ {
+ "epoch": 23.073482428115017,
+ "grad_norm": 0.11083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0386,
+ "step": 7222
+ },
+ {
+ "epoch": 23.076677316293928,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0357,
+ "step": 7223
+ },
+ {
+ "epoch": 23.079872204472842,
+ "grad_norm": 0.087890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0541,
+ "step": 7224
+ },
+ {
+ "epoch": 23.083067092651756,
+ "grad_norm": 0.08544921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0269,
+ "step": 7225
+ },
+ {
+ "epoch": 23.08626198083067,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0397,
+ "step": 7226
+ },
+ {
+ "epoch": 23.089456869009584,
+ "grad_norm": 0.0947265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0392,
+ "step": 7227
+ },
+ {
+ "epoch": 23.0926517571885,
+ "grad_norm": 0.056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.043,
+ "step": 7228
+ },
+ {
+ "epoch": 23.095846645367413,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0344,
+ "step": 7229
+ },
+ {
+ "epoch": 23.099041533546327,
+ "grad_norm": 0.0537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.043,
+ "step": 7230
+ },
+ {
+ "epoch": 23.10223642172524,
+ "grad_norm": 0.051513671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0363,
+ "step": 7231
+ },
+ {
+ "epoch": 23.105431309904155,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0462,
+ "step": 7232
+ },
+ {
+ "epoch": 23.108626198083066,
+ "grad_norm": 0.056884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0382,
+ "step": 7233
+ },
+ {
+ "epoch": 23.11182108626198,
+ "grad_norm": 0.06103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0387,
+ "step": 7234
+ },
+ {
+ "epoch": 23.115015974440894,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0447,
+ "step": 7235
+ },
+ {
+ "epoch": 23.118210862619808,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0563,
+ "step": 7236
+ },
+ {
+ "epoch": 23.121405750798722,
+ "grad_norm": 0.11376953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0448,
+ "step": 7237
+ },
+ {
+ "epoch": 23.124600638977636,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0466,
+ "step": 7238
+ },
+ {
+ "epoch": 23.12779552715655,
+ "grad_norm": 0.115234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0419,
+ "step": 7239
+ },
+ {
+ "epoch": 23.130990415335464,
+ "grad_norm": 0.0859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0226,
+ "step": 7240
+ },
+ {
+ "epoch": 23.13418530351438,
+ "grad_norm": 0.1494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0387,
+ "step": 7241
+ },
+ {
+ "epoch": 23.13738019169329,
+ "grad_norm": 0.0888671875,
+ "learning_rate": 0.0005,
+ "loss": 1.05,
+ "step": 7242
+ },
+ {
+ "epoch": 23.140575079872203,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0415,
+ "step": 7243
+ },
+ {
+ "epoch": 23.143769968051117,
+ "grad_norm": 0.091796875,
+ "learning_rate": 0.0005,
+ "loss": 1.047,
+ "step": 7244
+ },
+ {
+ "epoch": 23.14696485623003,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0335,
+ "step": 7245
+ },
+ {
+ "epoch": 23.150159744408946,
+ "grad_norm": 0.10498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0463,
+ "step": 7246
+ },
+ {
+ "epoch": 23.15335463258786,
+ "grad_norm": 0.08154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.036,
+ "step": 7247
+ },
+ {
+ "epoch": 23.156549520766774,
+ "grad_norm": 0.12353515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0381,
+ "step": 7248
+ },
+ {
+ "epoch": 23.159744408945688,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0465,
+ "step": 7249
+ },
+ {
+ "epoch": 23.162939297124602,
+ "grad_norm": 0.07861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0544,
+ "step": 7250
+ },
+ {
+ "epoch": 23.166134185303516,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0426,
+ "step": 7251
+ },
+ {
+ "epoch": 23.169329073482427,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0358,
+ "step": 7252
+ },
+ {
+ "epoch": 23.17252396166134,
+ "grad_norm": 0.146484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0541,
+ "step": 7253
+ },
+ {
+ "epoch": 23.175718849840255,
+ "grad_norm": 0.062255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0506,
+ "step": 7254
+ },
+ {
+ "epoch": 23.17891373801917,
+ "grad_norm": 0.10791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0397,
+ "step": 7255
+ },
+ {
+ "epoch": 23.182108626198083,
+ "grad_norm": 0.083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0397,
+ "step": 7256
+ },
+ {
+ "epoch": 23.185303514376997,
+ "grad_norm": 0.10791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0457,
+ "step": 7257
+ },
+ {
+ "epoch": 23.18849840255591,
+ "grad_norm": 0.0859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0444,
+ "step": 7258
+ },
+ {
+ "epoch": 23.191693290734825,
+ "grad_norm": 0.0869140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0508,
+ "step": 7259
+ },
+ {
+ "epoch": 23.19488817891374,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0409,
+ "step": 7260
+ },
+ {
+ "epoch": 23.19808306709265,
+ "grad_norm": 0.054931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0378,
+ "step": 7261
+ },
+ {
+ "epoch": 23.201277955271564,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0337,
+ "step": 7262
+ },
+ {
+ "epoch": 23.20447284345048,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0343,
+ "step": 7263
+ },
+ {
+ "epoch": 23.207667731629392,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0522,
+ "step": 7264
+ },
+ {
+ "epoch": 23.210862619808307,
+ "grad_norm": 0.056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0511,
+ "step": 7265
+ },
+ {
+ "epoch": 23.21405750798722,
+ "grad_norm": 0.055419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0358,
+ "step": 7266
+ },
+ {
+ "epoch": 23.217252396166135,
+ "grad_norm": 0.05517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0476,
+ "step": 7267
+ },
+ {
+ "epoch": 23.22044728434505,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0487,
+ "step": 7268
+ },
+ {
+ "epoch": 23.223642172523963,
+ "grad_norm": 0.048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0312,
+ "step": 7269
+ },
+ {
+ "epoch": 23.226837060702877,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0517,
+ "step": 7270
+ },
+ {
+ "epoch": 23.230031948881788,
+ "grad_norm": 0.0537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0446,
+ "step": 7271
+ },
+ {
+ "epoch": 23.233226837060702,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0272,
+ "step": 7272
+ },
+ {
+ "epoch": 23.236421725239616,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.038,
+ "step": 7273
+ },
+ {
+ "epoch": 23.23961661341853,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0379,
+ "step": 7274
+ },
+ {
+ "epoch": 23.242811501597444,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0367,
+ "step": 7275
+ },
+ {
+ "epoch": 23.24600638977636,
+ "grad_norm": 0.09912109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0464,
+ "step": 7276
+ },
+ {
+ "epoch": 23.249201277955272,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0306,
+ "step": 7277
+ },
+ {
+ "epoch": 23.252396166134186,
+ "grad_norm": 0.11962890625,
+ "learning_rate": 0.0005,
+ "loss": 1.037,
+ "step": 7278
+ },
+ {
+ "epoch": 23.2555910543131,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0264,
+ "step": 7279
+ },
+ {
+ "epoch": 23.25878594249201,
+ "grad_norm": 0.1025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0528,
+ "step": 7280
+ },
+ {
+ "epoch": 23.261980830670925,
+ "grad_norm": 0.05517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0477,
+ "step": 7281
+ },
+ {
+ "epoch": 23.26517571884984,
+ "grad_norm": 0.0888671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0528,
+ "step": 7282
+ },
+ {
+ "epoch": 23.268370607028753,
+ "grad_norm": 0.059814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0502,
+ "step": 7283
+ },
+ {
+ "epoch": 23.271565495207668,
+ "grad_norm": 0.1005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0453,
+ "step": 7284
+ },
+ {
+ "epoch": 23.27476038338658,
+ "grad_norm": 0.0537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0596,
+ "step": 7285
+ },
+ {
+ "epoch": 23.277955271565496,
+ "grad_norm": 0.0810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0507,
+ "step": 7286
+ },
+ {
+ "epoch": 23.28115015974441,
+ "grad_norm": 0.0615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0318,
+ "step": 7287
+ },
+ {
+ "epoch": 23.284345047923324,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0464,
+ "step": 7288
+ },
+ {
+ "epoch": 23.287539936102238,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0423,
+ "step": 7289
+ },
+ {
+ "epoch": 23.29073482428115,
+ "grad_norm": 0.05810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0508,
+ "step": 7290
+ },
+ {
+ "epoch": 23.293929712460063,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.047,
+ "step": 7291
+ },
+ {
+ "epoch": 23.297124600638977,
+ "grad_norm": 0.052734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0369,
+ "step": 7292
+ },
+ {
+ "epoch": 23.30031948881789,
+ "grad_norm": 0.0556640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0281,
+ "step": 7293
+ },
+ {
+ "epoch": 23.303514376996805,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0392,
+ "step": 7294
+ },
+ {
+ "epoch": 23.30670926517572,
+ "grad_norm": 0.08935546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0419,
+ "step": 7295
+ },
+ {
+ "epoch": 23.309904153354633,
+ "grad_norm": 0.0615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0443,
+ "step": 7296
+ },
+ {
+ "epoch": 23.313099041533548,
+ "grad_norm": 0.1005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0425,
+ "step": 7297
+ },
+ {
+ "epoch": 23.31629392971246,
+ "grad_norm": 0.051513671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0423,
+ "step": 7298
+ },
+ {
+ "epoch": 23.319488817891372,
+ "grad_norm": 0.08740234375,
+ "learning_rate": 0.0005,
+ "loss": 1.049,
+ "step": 7299
+ },
+ {
+ "epoch": 23.322683706070286,
+ "grad_norm": 0.05322265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0474,
+ "step": 7300
+ },
+ {
+ "epoch": 23.3258785942492,
+ "grad_norm": 0.052978515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0502,
+ "step": 7301
+ },
+ {
+ "epoch": 23.329073482428115,
+ "grad_norm": 0.058837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0345,
+ "step": 7302
+ },
+ {
+ "epoch": 23.33226837060703,
+ "grad_norm": 0.0595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.047,
+ "step": 7303
+ },
+ {
+ "epoch": 23.335463258785943,
+ "grad_norm": 0.05126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0466,
+ "step": 7304
+ },
+ {
+ "epoch": 23.338658146964857,
+ "grad_norm": 0.05419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0474,
+ "step": 7305
+ },
+ {
+ "epoch": 23.34185303514377,
+ "grad_norm": 0.056884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0412,
+ "step": 7306
+ },
+ {
+ "epoch": 23.345047923322685,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0543,
+ "step": 7307
+ },
+ {
+ "epoch": 23.3482428115016,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0394,
+ "step": 7308
+ },
+ {
+ "epoch": 23.35143769968051,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0429,
+ "step": 7309
+ },
+ {
+ "epoch": 23.354632587859424,
+ "grad_norm": 0.057861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.03,
+ "step": 7310
+ },
+ {
+ "epoch": 23.357827476038338,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0538,
+ "step": 7311
+ },
+ {
+ "epoch": 23.361022364217252,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0484,
+ "step": 7312
+ },
+ {
+ "epoch": 23.364217252396166,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0411,
+ "step": 7313
+ },
+ {
+ "epoch": 23.36741214057508,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0523,
+ "step": 7314
+ },
+ {
+ "epoch": 23.370607028753994,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0445,
+ "step": 7315
+ },
+ {
+ "epoch": 23.37380191693291,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0318,
+ "step": 7316
+ },
+ {
+ "epoch": 23.376996805111823,
+ "grad_norm": 0.11572265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0299,
+ "step": 7317
+ },
+ {
+ "epoch": 23.380191693290733,
+ "grad_norm": 0.0830078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0427,
+ "step": 7318
+ },
+ {
+ "epoch": 23.383386581469647,
+ "grad_norm": 0.0986328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0437,
+ "step": 7319
+ },
+ {
+ "epoch": 23.38658146964856,
+ "grad_norm": 0.05859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0408,
+ "step": 7320
+ },
+ {
+ "epoch": 23.389776357827476,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0428,
+ "step": 7321
+ },
+ {
+ "epoch": 23.39297124600639,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0471,
+ "step": 7322
+ },
+ {
+ "epoch": 23.396166134185304,
+ "grad_norm": 0.056396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0333,
+ "step": 7323
+ },
+ {
+ "epoch": 23.399361022364218,
+ "grad_norm": 0.0810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0478,
+ "step": 7324
+ },
+ {
+ "epoch": 23.402555910543132,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0305,
+ "step": 7325
+ },
+ {
+ "epoch": 23.405750798722046,
+ "grad_norm": 0.06201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0549,
+ "step": 7326
+ },
+ {
+ "epoch": 23.408945686900957,
+ "grad_norm": 0.10009765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0472,
+ "step": 7327
+ },
+ {
+ "epoch": 23.41214057507987,
+ "grad_norm": 0.054443359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0416,
+ "step": 7328
+ },
+ {
+ "epoch": 23.415335463258785,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0381,
+ "step": 7329
+ },
+ {
+ "epoch": 23.4185303514377,
+ "grad_norm": 0.050537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0385,
+ "step": 7330
+ },
+ {
+ "epoch": 23.421725239616613,
+ "grad_norm": 0.059326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0424,
+ "step": 7331
+ },
+ {
+ "epoch": 23.424920127795527,
+ "grad_norm": 0.05810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0521,
+ "step": 7332
+ },
+ {
+ "epoch": 23.42811501597444,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0488,
+ "step": 7333
+ },
+ {
+ "epoch": 23.431309904153355,
+ "grad_norm": 0.056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0461,
+ "step": 7334
+ },
+ {
+ "epoch": 23.43450479233227,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0379,
+ "step": 7335
+ },
+ {
+ "epoch": 23.437699680511184,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0484,
+ "step": 7336
+ },
+ {
+ "epoch": 23.440894568690094,
+ "grad_norm": 0.08740234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0323,
+ "step": 7337
+ },
+ {
+ "epoch": 23.44408945686901,
+ "grad_norm": 0.080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0513,
+ "step": 7338
+ },
+ {
+ "epoch": 23.447284345047922,
+ "grad_norm": 0.056884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0532,
+ "step": 7339
+ },
+ {
+ "epoch": 23.450479233226837,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.04,
+ "step": 7340
+ },
+ {
+ "epoch": 23.45367412140575,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.035,
+ "step": 7341
+ },
+ {
+ "epoch": 23.456869009584665,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0478,
+ "step": 7342
+ },
+ {
+ "epoch": 23.46006389776358,
+ "grad_norm": 0.0810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0399,
+ "step": 7343
+ },
+ {
+ "epoch": 23.463258785942493,
+ "grad_norm": 0.051025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0346,
+ "step": 7344
+ },
+ {
+ "epoch": 23.466453674121407,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0476,
+ "step": 7345
+ },
+ {
+ "epoch": 23.46964856230032,
+ "grad_norm": 0.062255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0378,
+ "step": 7346
+ },
+ {
+ "epoch": 23.472843450479232,
+ "grad_norm": 0.053466796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0422,
+ "step": 7347
+ },
+ {
+ "epoch": 23.476038338658146,
+ "grad_norm": 0.06103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.042,
+ "step": 7348
+ },
+ {
+ "epoch": 23.47923322683706,
+ "grad_norm": 0.05029296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0458,
+ "step": 7349
+ },
+ {
+ "epoch": 23.482428115015974,
+ "grad_norm": 0.05029296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0461,
+ "step": 7350
+ },
+ {
+ "epoch": 23.48562300319489,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0489,
+ "step": 7351
+ },
+ {
+ "epoch": 23.488817891373802,
+ "grad_norm": 0.049560546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0415,
+ "step": 7352
+ },
+ {
+ "epoch": 23.492012779552716,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0338,
+ "step": 7353
+ },
+ {
+ "epoch": 23.49520766773163,
+ "grad_norm": 0.095703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0482,
+ "step": 7354
+ },
+ {
+ "epoch": 23.498402555910545,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0325,
+ "step": 7355
+ },
+ {
+ "epoch": 23.501597444089455,
+ "grad_norm": 0.126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0426,
+ "step": 7356
+ },
+ {
+ "epoch": 23.50479233226837,
+ "grad_norm": 0.055419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0516,
+ "step": 7357
+ },
+ {
+ "epoch": 23.507987220447284,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0416,
+ "step": 7358
+ },
+ {
+ "epoch": 23.511182108626198,
+ "grad_norm": 0.10009765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0353,
+ "step": 7359
+ },
+ {
+ "epoch": 23.51437699680511,
+ "grad_norm": 0.0595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0551,
+ "step": 7360
+ },
+ {
+ "epoch": 23.517571884984026,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0475,
+ "step": 7361
+ },
+ {
+ "epoch": 23.52076677316294,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0471,
+ "step": 7362
+ },
+ {
+ "epoch": 23.523961661341854,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0431,
+ "step": 7363
+ },
+ {
+ "epoch": 23.527156549520768,
+ "grad_norm": 0.0986328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0473,
+ "step": 7364
+ },
+ {
+ "epoch": 23.53035143769968,
+ "grad_norm": 0.09375,
+ "learning_rate": 0.0005,
+ "loss": 1.0437,
+ "step": 7365
+ },
+ {
+ "epoch": 23.533546325878593,
+ "grad_norm": 0.09375,
+ "learning_rate": 0.0005,
+ "loss": 1.0506,
+ "step": 7366
+ },
+ {
+ "epoch": 23.536741214057507,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0342,
+ "step": 7367
+ },
+ {
+ "epoch": 23.53993610223642,
+ "grad_norm": 0.10302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0465,
+ "step": 7368
+ },
+ {
+ "epoch": 23.543130990415335,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0354,
+ "step": 7369
+ },
+ {
+ "epoch": 23.54632587859425,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0513,
+ "step": 7370
+ },
+ {
+ "epoch": 23.549520766773163,
+ "grad_norm": 0.05810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0405,
+ "step": 7371
+ },
+ {
+ "epoch": 23.552715654952078,
+ "grad_norm": 0.050048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0553,
+ "step": 7372
+ },
+ {
+ "epoch": 23.55591054313099,
+ "grad_norm": 0.056884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0417,
+ "step": 7373
+ },
+ {
+ "epoch": 23.559105431309906,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0558,
+ "step": 7374
+ },
+ {
+ "epoch": 23.562300319488816,
+ "grad_norm": 0.049072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.036,
+ "step": 7375
+ },
+ {
+ "epoch": 23.56549520766773,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0484,
+ "step": 7376
+ },
+ {
+ "epoch": 23.568690095846645,
+ "grad_norm": 0.052490234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0393,
+ "step": 7377
+ },
+ {
+ "epoch": 23.57188498402556,
+ "grad_norm": 0.0498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0302,
+ "step": 7378
+ },
+ {
+ "epoch": 23.575079872204473,
+ "grad_norm": 0.057373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0427,
+ "step": 7379
+ },
+ {
+ "epoch": 23.578274760383387,
+ "grad_norm": 0.0517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0383,
+ "step": 7380
+ },
+ {
+ "epoch": 23.5814696485623,
+ "grad_norm": 0.056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0436,
+ "step": 7381
+ },
+ {
+ "epoch": 23.584664536741215,
+ "grad_norm": 0.059326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0344,
+ "step": 7382
+ },
+ {
+ "epoch": 23.58785942492013,
+ "grad_norm": 0.10595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0418,
+ "step": 7383
+ },
+ {
+ "epoch": 23.591054313099043,
+ "grad_norm": 0.05859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0407,
+ "step": 7384
+ },
+ {
+ "epoch": 23.594249201277954,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.047,
+ "step": 7385
+ },
+ {
+ "epoch": 23.597444089456868,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0306,
+ "step": 7386
+ },
+ {
+ "epoch": 23.600638977635782,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0448,
+ "step": 7387
+ },
+ {
+ "epoch": 23.603833865814696,
+ "grad_norm": 0.053466796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0471,
+ "step": 7388
+ },
+ {
+ "epoch": 23.60702875399361,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0367,
+ "step": 7389
+ },
+ {
+ "epoch": 23.610223642172524,
+ "grad_norm": 0.05859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0419,
+ "step": 7390
+ },
+ {
+ "epoch": 23.61341853035144,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0405,
+ "step": 7391
+ },
+ {
+ "epoch": 23.616613418530353,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0392,
+ "step": 7392
+ },
+ {
+ "epoch": 23.619808306709267,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0339,
+ "step": 7393
+ },
+ {
+ "epoch": 23.623003194888177,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0491,
+ "step": 7394
+ },
+ {
+ "epoch": 23.62619808306709,
+ "grad_norm": 0.09912109375,
+ "learning_rate": 0.0005,
+ "loss": 1.042,
+ "step": 7395
+ },
+ {
+ "epoch": 23.629392971246006,
+ "grad_norm": 0.146484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0465,
+ "step": 7396
+ },
+ {
+ "epoch": 23.63258785942492,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.039,
+ "step": 7397
+ },
+ {
+ "epoch": 23.635782747603834,
+ "grad_norm": 0.095703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0376,
+ "step": 7398
+ },
+ {
+ "epoch": 23.638977635782748,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0269,
+ "step": 7399
+ },
+ {
+ "epoch": 23.642172523961662,
+ "grad_norm": 0.1201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0524,
+ "step": 7400
+ },
+ {
+ "epoch": 23.645367412140576,
+ "grad_norm": 0.158203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0437,
+ "step": 7401
+ },
+ {
+ "epoch": 23.64856230031949,
+ "grad_norm": 0.054931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0405,
+ "step": 7402
+ },
+ {
+ "epoch": 23.6517571884984,
+ "grad_norm": 0.09521484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0389,
+ "step": 7403
+ },
+ {
+ "epoch": 23.654952076677315,
+ "grad_norm": 0.05029296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0464,
+ "step": 7404
+ },
+ {
+ "epoch": 23.65814696485623,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0463,
+ "step": 7405
+ },
+ {
+ "epoch": 23.661341853035143,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0519,
+ "step": 7406
+ },
+ {
+ "epoch": 23.664536741214057,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0424,
+ "step": 7407
+ },
+ {
+ "epoch": 23.66773162939297,
+ "grad_norm": 0.0615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0462,
+ "step": 7408
+ },
+ {
+ "epoch": 23.670926517571885,
+ "grad_norm": 0.049560546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0272,
+ "step": 7409
+ },
+ {
+ "epoch": 23.6741214057508,
+ "grad_norm": 0.091796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0446,
+ "step": 7410
+ },
+ {
+ "epoch": 23.677316293929714,
+ "grad_norm": 0.06103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0418,
+ "step": 7411
+ },
+ {
+ "epoch": 23.680511182108628,
+ "grad_norm": 0.051513671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0424,
+ "step": 7412
+ },
+ {
+ "epoch": 23.68370607028754,
+ "grad_norm": 0.07861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0471,
+ "step": 7413
+ },
+ {
+ "epoch": 23.686900958466452,
+ "grad_norm": 0.0537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0375,
+ "step": 7414
+ },
+ {
+ "epoch": 23.690095846645367,
+ "grad_norm": 0.05224609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0557,
+ "step": 7415
+ },
+ {
+ "epoch": 23.69329073482428,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0388,
+ "step": 7416
+ },
+ {
+ "epoch": 23.696485623003195,
+ "grad_norm": 0.05517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0442,
+ "step": 7417
+ },
+ {
+ "epoch": 23.69968051118211,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0337,
+ "step": 7418
+ },
+ {
+ "epoch": 23.702875399361023,
+ "grad_norm": 0.055419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0419,
+ "step": 7419
+ },
+ {
+ "epoch": 23.706070287539937,
+ "grad_norm": 0.05029296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0449,
+ "step": 7420
+ },
+ {
+ "epoch": 23.70926517571885,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0483,
+ "step": 7421
+ },
+ {
+ "epoch": 23.712460063897765,
+ "grad_norm": 0.05517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0479,
+ "step": 7422
+ },
+ {
+ "epoch": 23.715654952076676,
+ "grad_norm": 0.052490234375,
+ "learning_rate": 0.0005,
+ "loss": 1.04,
+ "step": 7423
+ },
+ {
+ "epoch": 23.71884984025559,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0381,
+ "step": 7424
+ },
+ {
+ "epoch": 23.722044728434504,
+ "grad_norm": 0.05126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0475,
+ "step": 7425
+ },
+ {
+ "epoch": 23.72523961661342,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0517,
+ "step": 7426
+ },
+ {
+ "epoch": 23.728434504792332,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0395,
+ "step": 7427
+ },
+ {
+ "epoch": 23.731629392971247,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0339,
+ "step": 7428
+ },
+ {
+ "epoch": 23.73482428115016,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0429,
+ "step": 7429
+ },
+ {
+ "epoch": 23.738019169329075,
+ "grad_norm": 0.05224609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0445,
+ "step": 7430
+ },
+ {
+ "epoch": 23.74121405750799,
+ "grad_norm": 0.0966796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0417,
+ "step": 7431
+ },
+ {
+ "epoch": 23.7444089456869,
+ "grad_norm": 0.056884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0361,
+ "step": 7432
+ },
+ {
+ "epoch": 23.747603833865814,
+ "grad_norm": 0.10400390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0417,
+ "step": 7433
+ },
+ {
+ "epoch": 23.750798722044728,
+ "grad_norm": 0.05126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0278,
+ "step": 7434
+ },
+ {
+ "epoch": 23.75399361022364,
+ "grad_norm": 0.062255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0493,
+ "step": 7435
+ },
+ {
+ "epoch": 23.757188498402556,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.038,
+ "step": 7436
+ },
+ {
+ "epoch": 23.76038338658147,
+ "grad_norm": 0.08740234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0513,
+ "step": 7437
+ },
+ {
+ "epoch": 23.763578274760384,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0364,
+ "step": 7438
+ },
+ {
+ "epoch": 23.766773162939298,
+ "grad_norm": 0.0966796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0484,
+ "step": 7439
+ },
+ {
+ "epoch": 23.769968051118212,
+ "grad_norm": 0.125,
+ "learning_rate": 0.0005,
+ "loss": 1.0355,
+ "step": 7440
+ },
+ {
+ "epoch": 23.773162939297123,
+ "grad_norm": 0.1552734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0381,
+ "step": 7441
+ },
+ {
+ "epoch": 23.776357827476037,
+ "grad_norm": 0.10302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0417,
+ "step": 7442
+ },
+ {
+ "epoch": 23.77955271565495,
+ "grad_norm": 0.1533203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0391,
+ "step": 7443
+ },
+ {
+ "epoch": 23.782747603833865,
+ "grad_norm": 0.0986328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0434,
+ "step": 7444
+ },
+ {
+ "epoch": 23.78594249201278,
+ "grad_norm": 0.10546875,
+ "learning_rate": 0.0005,
+ "loss": 1.05,
+ "step": 7445
+ },
+ {
+ "epoch": 23.789137380191693,
+ "grad_norm": 0.1708984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0451,
+ "step": 7446
+ },
+ {
+ "epoch": 23.792332268370608,
+ "grad_norm": 0.09130859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0322,
+ "step": 7447
+ },
+ {
+ "epoch": 23.79552715654952,
+ "grad_norm": 0.115234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0421,
+ "step": 7448
+ },
+ {
+ "epoch": 23.798722044728436,
+ "grad_norm": 0.09228515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0554,
+ "step": 7449
+ },
+ {
+ "epoch": 23.80191693290735,
+ "grad_norm": 0.11669921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0534,
+ "step": 7450
+ },
+ {
+ "epoch": 23.80511182108626,
+ "grad_norm": 0.1494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0335,
+ "step": 7451
+ },
+ {
+ "epoch": 23.808306709265175,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.045,
+ "step": 7452
+ },
+ {
+ "epoch": 23.81150159744409,
+ "grad_norm": 0.1279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0352,
+ "step": 7453
+ },
+ {
+ "epoch": 23.814696485623003,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0346,
+ "step": 7454
+ },
+ {
+ "epoch": 23.817891373801917,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0382,
+ "step": 7455
+ },
+ {
+ "epoch": 23.82108626198083,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0363,
+ "step": 7456
+ },
+ {
+ "epoch": 23.824281150159745,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0419,
+ "step": 7457
+ },
+ {
+ "epoch": 23.82747603833866,
+ "grad_norm": 0.08544921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0568,
+ "step": 7458
+ },
+ {
+ "epoch": 23.830670926517573,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0462,
+ "step": 7459
+ },
+ {
+ "epoch": 23.833865814696484,
+ "grad_norm": 0.08203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0421,
+ "step": 7460
+ },
+ {
+ "epoch": 23.837060702875398,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0463,
+ "step": 7461
+ },
+ {
+ "epoch": 23.840255591054312,
+ "grad_norm": 0.057373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.049,
+ "step": 7462
+ },
+ {
+ "epoch": 23.843450479233226,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.042,
+ "step": 7463
+ },
+ {
+ "epoch": 23.84664536741214,
+ "grad_norm": 0.055908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0371,
+ "step": 7464
+ },
+ {
+ "epoch": 23.849840255591054,
+ "grad_norm": 0.111328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0306,
+ "step": 7465
+ },
+ {
+ "epoch": 23.85303514376997,
+ "grad_norm": 0.0595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0432,
+ "step": 7466
+ },
+ {
+ "epoch": 23.856230031948883,
+ "grad_norm": 0.11181640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0464,
+ "step": 7467
+ },
+ {
+ "epoch": 23.859424920127797,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0461,
+ "step": 7468
+ },
+ {
+ "epoch": 23.86261980830671,
+ "grad_norm": 0.056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0523,
+ "step": 7469
+ },
+ {
+ "epoch": 23.86581469648562,
+ "grad_norm": 0.05810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0386,
+ "step": 7470
+ },
+ {
+ "epoch": 23.869009584664536,
+ "grad_norm": 0.056396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0364,
+ "step": 7471
+ },
+ {
+ "epoch": 23.87220447284345,
+ "grad_norm": 0.05419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0507,
+ "step": 7472
+ },
+ {
+ "epoch": 23.875399361022364,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0328,
+ "step": 7473
+ },
+ {
+ "epoch": 23.878594249201278,
+ "grad_norm": 0.058837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0407,
+ "step": 7474
+ },
+ {
+ "epoch": 23.881789137380192,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0512,
+ "step": 7475
+ },
+ {
+ "epoch": 23.884984025559106,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0483,
+ "step": 7476
+ },
+ {
+ "epoch": 23.88817891373802,
+ "grad_norm": 0.05615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0456,
+ "step": 7477
+ },
+ {
+ "epoch": 23.891373801916934,
+ "grad_norm": 0.05615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0428,
+ "step": 7478
+ },
+ {
+ "epoch": 23.894568690095845,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0466,
+ "step": 7479
+ },
+ {
+ "epoch": 23.89776357827476,
+ "grad_norm": 0.050537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0417,
+ "step": 7480
+ },
+ {
+ "epoch": 23.900958466453673,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0354,
+ "step": 7481
+ },
+ {
+ "epoch": 23.904153354632587,
+ "grad_norm": 0.09423828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0459,
+ "step": 7482
+ },
+ {
+ "epoch": 23.9073482428115,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0405,
+ "step": 7483
+ },
+ {
+ "epoch": 23.910543130990416,
+ "grad_norm": 0.0947265625,
+ "learning_rate": 0.0005,
+ "loss": 1.042,
+ "step": 7484
+ },
+ {
+ "epoch": 23.91373801916933,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0403,
+ "step": 7485
+ },
+ {
+ "epoch": 23.916932907348244,
+ "grad_norm": 0.05859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0395,
+ "step": 7486
+ },
+ {
+ "epoch": 23.920127795527158,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0475,
+ "step": 7487
+ },
+ {
+ "epoch": 23.923322683706072,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0477,
+ "step": 7488
+ },
+ {
+ "epoch": 23.926517571884983,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0445,
+ "step": 7489
+ },
+ {
+ "epoch": 23.929712460063897,
+ "grad_norm": 0.060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0395,
+ "step": 7490
+ },
+ {
+ "epoch": 23.93290734824281,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0462,
+ "step": 7491
+ },
+ {
+ "epoch": 23.936102236421725,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.047,
+ "step": 7492
+ },
+ {
+ "epoch": 23.93929712460064,
+ "grad_norm": 0.12109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0411,
+ "step": 7493
+ },
+ {
+ "epoch": 23.942492012779553,
+ "grad_norm": 0.05810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0379,
+ "step": 7494
+ },
+ {
+ "epoch": 23.945686900958467,
+ "grad_norm": 0.099609375,
+ "learning_rate": 0.0005,
+ "loss": 1.066,
+ "step": 7495
+ },
+ {
+ "epoch": 23.94888178913738,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0456,
+ "step": 7496
+ },
+ {
+ "epoch": 23.952076677316295,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0543,
+ "step": 7497
+ },
+ {
+ "epoch": 23.955271565495206,
+ "grad_norm": 0.08642578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0437,
+ "step": 7498
+ },
+ {
+ "epoch": 23.95846645367412,
+ "grad_norm": 0.109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0457,
+ "step": 7499
+ },
+ {
+ "epoch": 23.961661341853034,
+ "grad_norm": 0.08642578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0458,
+ "step": 7500
+ },
+ {
+ "epoch": 23.96485623003195,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0494,
+ "step": 7501
+ },
+ {
+ "epoch": 23.968051118210862,
+ "grad_norm": 0.1513671875,
+ "learning_rate": 0.0005,
+ "loss": 1.05,
+ "step": 7502
+ },
+ {
+ "epoch": 23.971246006389777,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0568,
+ "step": 7503
+ },
+ {
+ "epoch": 23.97444089456869,
+ "grad_norm": 0.08447265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0359,
+ "step": 7504
+ },
+ {
+ "epoch": 23.977635782747605,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0419,
+ "step": 7505
+ },
+ {
+ "epoch": 23.98083067092652,
+ "grad_norm": 0.154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0481,
+ "step": 7506
+ },
+ {
+ "epoch": 23.984025559105433,
+ "grad_norm": 0.0546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0367,
+ "step": 7507
+ },
+ {
+ "epoch": 23.987220447284344,
+ "grad_norm": 0.05859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0425,
+ "step": 7508
+ },
+ {
+ "epoch": 23.990415335463258,
+ "grad_norm": 0.09326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.042,
+ "step": 7509
+ },
+ {
+ "epoch": 23.99361022364217,
+ "grad_norm": 0.059814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0352,
+ "step": 7510
+ },
+ {
+ "epoch": 23.996805111821086,
+ "grad_norm": 0.060791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0405,
+ "step": 7511
+ },
+ {
+ "epoch": 24.0,
+ "grad_norm": 0.060302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0512,
+ "step": 7512
+ },
+ {
+ "epoch": 24.003194888178914,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0482,
+ "step": 7513
+ },
+ {
+ "epoch": 24.00638977635783,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0389,
+ "step": 7514
+ },
+ {
+ "epoch": 24.009584664536742,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0491,
+ "step": 7515
+ },
+ {
+ "epoch": 24.012779552715656,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0366,
+ "step": 7516
+ },
+ {
+ "epoch": 24.015974440894567,
+ "grad_norm": 0.0810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0455,
+ "step": 7517
+ },
+ {
+ "epoch": 24.01916932907348,
+ "grad_norm": 0.057861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0521,
+ "step": 7518
+ },
+ {
+ "epoch": 24.022364217252395,
+ "grad_norm": 0.08447265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0424,
+ "step": 7519
+ },
+ {
+ "epoch": 24.02555910543131,
+ "grad_norm": 0.05517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0375,
+ "step": 7520
+ },
+ {
+ "epoch": 24.028753993610223,
+ "grad_norm": 0.08837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0418,
+ "step": 7521
+ },
+ {
+ "epoch": 24.031948881789138,
+ "grad_norm": 0.05615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0446,
+ "step": 7522
+ },
+ {
+ "epoch": 24.03514376996805,
+ "grad_norm": 0.0849609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0479,
+ "step": 7523
+ },
+ {
+ "epoch": 24.038338658146966,
+ "grad_norm": 0.05712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0425,
+ "step": 7524
+ },
+ {
+ "epoch": 24.04153354632588,
+ "grad_norm": 0.1005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0347,
+ "step": 7525
+ },
+ {
+ "epoch": 24.044728434504794,
+ "grad_norm": 0.10107421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0513,
+ "step": 7526
+ },
+ {
+ "epoch": 24.047923322683705,
+ "grad_norm": 0.05908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0395,
+ "step": 7527
+ },
+ {
+ "epoch": 24.05111821086262,
+ "grad_norm": 0.09912109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0324,
+ "step": 7528
+ },
+ {
+ "epoch": 24.054313099041533,
+ "grad_norm": 0.051513671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0335,
+ "step": 7529
+ },
+ {
+ "epoch": 24.057507987220447,
+ "grad_norm": 0.060791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0381,
+ "step": 7530
+ },
+ {
+ "epoch": 24.06070287539936,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0547,
+ "step": 7531
+ },
+ {
+ "epoch": 24.063897763578275,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0508,
+ "step": 7532
+ },
+ {
+ "epoch": 24.06709265175719,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.043,
+ "step": 7533
+ },
+ {
+ "epoch": 24.070287539936103,
+ "grad_norm": 0.087890625,
+ "learning_rate": 0.0005,
+ "loss": 1.039,
+ "step": 7534
+ },
+ {
+ "epoch": 24.073482428115017,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.047,
+ "step": 7535
+ },
+ {
+ "epoch": 24.076677316293928,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.043,
+ "step": 7536
+ },
+ {
+ "epoch": 24.079872204472842,
+ "grad_norm": 0.10791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0447,
+ "step": 7537
+ },
+ {
+ "epoch": 24.083067092651756,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0397,
+ "step": 7538
+ },
+ {
+ "epoch": 24.08626198083067,
+ "grad_norm": 0.1298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0408,
+ "step": 7539
+ },
+ {
+ "epoch": 24.089456869009584,
+ "grad_norm": 0.08642578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0446,
+ "step": 7540
+ },
+ {
+ "epoch": 24.0926517571885,
+ "grad_norm": 0.1279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0399,
+ "step": 7541
+ },
+ {
+ "epoch": 24.095846645367413,
+ "grad_norm": 0.080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0305,
+ "step": 7542
+ },
+ {
+ "epoch": 24.099041533546327,
+ "grad_norm": 0.11572265625,
+ "learning_rate": 0.0005,
+ "loss": 1.039,
+ "step": 7543
+ },
+ {
+ "epoch": 24.10223642172524,
+ "grad_norm": 0.07861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0481,
+ "step": 7544
+ },
+ {
+ "epoch": 24.105431309904155,
+ "grad_norm": 0.08837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0409,
+ "step": 7545
+ },
+ {
+ "epoch": 24.108626198083066,
+ "grad_norm": 0.08935546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0238,
+ "step": 7546
+ },
+ {
+ "epoch": 24.11182108626198,
+ "grad_norm": 0.08447265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0365,
+ "step": 7547
+ },
+ {
+ "epoch": 24.115015974440894,
+ "grad_norm": 0.08447265625,
+ "learning_rate": 0.0005,
+ "loss": 1.042,
+ "step": 7548
+ },
+ {
+ "epoch": 24.118210862619808,
+ "grad_norm": 0.09375,
+ "learning_rate": 0.0005,
+ "loss": 1.0308,
+ "step": 7549
+ },
+ {
+ "epoch": 24.121405750798722,
+ "grad_norm": 0.123046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0425,
+ "step": 7550
+ },
+ {
+ "epoch": 24.124600638977636,
+ "grad_norm": 0.08935546875,
+ "learning_rate": 0.0005,
+ "loss": 1.045,
+ "step": 7551
+ },
+ {
+ "epoch": 24.12779552715655,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0375,
+ "step": 7552
+ },
+ {
+ "epoch": 24.130990415335464,
+ "grad_norm": 0.0595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0434,
+ "step": 7553
+ },
+ {
+ "epoch": 24.13418530351438,
+ "grad_norm": 0.0849609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0455,
+ "step": 7554
+ },
+ {
+ "epoch": 24.13738019169329,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0444,
+ "step": 7555
+ },
+ {
+ "epoch": 24.140575079872203,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0384,
+ "step": 7556
+ },
+ {
+ "epoch": 24.143769968051117,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0451,
+ "step": 7557
+ },
+ {
+ "epoch": 24.14696485623003,
+ "grad_norm": 0.0595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0479,
+ "step": 7558
+ },
+ {
+ "epoch": 24.150159744408946,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0491,
+ "step": 7559
+ },
+ {
+ "epoch": 24.15335463258786,
+ "grad_norm": 0.05908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0434,
+ "step": 7560
+ },
+ {
+ "epoch": 24.156549520766774,
+ "grad_norm": 0.053466796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0443,
+ "step": 7561
+ },
+ {
+ "epoch": 24.159744408945688,
+ "grad_norm": 0.0517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0366,
+ "step": 7562
+ },
+ {
+ "epoch": 24.162939297124602,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0439,
+ "step": 7563
+ },
+ {
+ "epoch": 24.166134185303516,
+ "grad_norm": 0.052490234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0403,
+ "step": 7564
+ },
+ {
+ "epoch": 24.169329073482427,
+ "grad_norm": 0.0546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0505,
+ "step": 7565
+ },
+ {
+ "epoch": 24.17252396166134,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0417,
+ "step": 7566
+ },
+ {
+ "epoch": 24.175718849840255,
+ "grad_norm": 0.0556640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0362,
+ "step": 7567
+ },
+ {
+ "epoch": 24.17891373801917,
+ "grad_norm": 0.06005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0344,
+ "step": 7568
+ },
+ {
+ "epoch": 24.182108626198083,
+ "grad_norm": 0.049560546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0372,
+ "step": 7569
+ },
+ {
+ "epoch": 24.185303514376997,
+ "grad_norm": 0.05517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0472,
+ "step": 7570
+ },
+ {
+ "epoch": 24.18849840255591,
+ "grad_norm": 0.05029296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0369,
+ "step": 7571
+ },
+ {
+ "epoch": 24.191693290734825,
+ "grad_norm": 0.05224609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0469,
+ "step": 7572
+ },
+ {
+ "epoch": 24.19488817891374,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0355,
+ "step": 7573
+ },
+ {
+ "epoch": 24.19808306709265,
+ "grad_norm": 0.04541015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0387,
+ "step": 7574
+ },
+ {
+ "epoch": 24.201277955271564,
+ "grad_norm": 0.05126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0505,
+ "step": 7575
+ },
+ {
+ "epoch": 24.20447284345048,
+ "grad_norm": 0.0498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0407,
+ "step": 7576
+ },
+ {
+ "epoch": 24.207667731629392,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0484,
+ "step": 7577
+ },
+ {
+ "epoch": 24.210862619808307,
+ "grad_norm": 0.052734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0413,
+ "step": 7578
+ },
+ {
+ "epoch": 24.21405750798722,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0528,
+ "step": 7579
+ },
+ {
+ "epoch": 24.217252396166135,
+ "grad_norm": 0.0546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0443,
+ "step": 7580
+ },
+ {
+ "epoch": 24.22044728434505,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0504,
+ "step": 7581
+ },
+ {
+ "epoch": 24.223642172523963,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0322,
+ "step": 7582
+ },
+ {
+ "epoch": 24.226837060702877,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.051,
+ "step": 7583
+ },
+ {
+ "epoch": 24.230031948881788,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0439,
+ "step": 7584
+ },
+ {
+ "epoch": 24.233226837060702,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0324,
+ "step": 7585
+ },
+ {
+ "epoch": 24.236421725239616,
+ "grad_norm": 0.107421875,
+ "learning_rate": 0.0005,
+ "loss": 1.054,
+ "step": 7586
+ },
+ {
+ "epoch": 24.23961661341853,
+ "grad_norm": 0.1005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0406,
+ "step": 7587
+ },
+ {
+ "epoch": 24.242811501597444,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0348,
+ "step": 7588
+ },
+ {
+ "epoch": 24.24600638977636,
+ "grad_norm": 0.11328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0298,
+ "step": 7589
+ },
+ {
+ "epoch": 24.249201277955272,
+ "grad_norm": 0.060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0513,
+ "step": 7590
+ },
+ {
+ "epoch": 24.252396166134186,
+ "grad_norm": 0.0986328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0383,
+ "step": 7591
+ },
+ {
+ "epoch": 24.2555910543131,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0493,
+ "step": 7592
+ },
+ {
+ "epoch": 24.25878594249201,
+ "grad_norm": 0.140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0368,
+ "step": 7593
+ },
+ {
+ "epoch": 24.261980830670925,
+ "grad_norm": 0.051513671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0473,
+ "step": 7594
+ },
+ {
+ "epoch": 24.26517571884984,
+ "grad_norm": 0.1064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0341,
+ "step": 7595
+ },
+ {
+ "epoch": 24.268370607028753,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0362,
+ "step": 7596
+ },
+ {
+ "epoch": 24.271565495207668,
+ "grad_norm": 0.1259765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0426,
+ "step": 7597
+ },
+ {
+ "epoch": 24.27476038338658,
+ "grad_norm": 0.11962890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0492,
+ "step": 7598
+ },
+ {
+ "epoch": 24.277955271565496,
+ "grad_norm": 0.1064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0397,
+ "step": 7599
+ },
+ {
+ "epoch": 24.28115015974441,
+ "grad_norm": 0.1982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0479,
+ "step": 7600
+ },
+ {
+ "epoch": 24.284345047923324,
+ "grad_norm": 0.048583984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0458,
+ "step": 7601
+ },
+ {
+ "epoch": 24.287539936102238,
+ "grad_norm": 0.16015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0546,
+ "step": 7602
+ },
+ {
+ "epoch": 24.29073482428115,
+ "grad_norm": 0.054931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0376,
+ "step": 7603
+ },
+ {
+ "epoch": 24.293929712460063,
+ "grad_norm": 0.1435546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0409,
+ "step": 7604
+ },
+ {
+ "epoch": 24.297124600638977,
+ "grad_norm": 0.08203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0498,
+ "step": 7605
+ },
+ {
+ "epoch": 24.30031948881789,
+ "grad_norm": 0.125,
+ "learning_rate": 0.0005,
+ "loss": 1.0389,
+ "step": 7606
+ },
+ {
+ "epoch": 24.303514376996805,
+ "grad_norm": 0.1123046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0526,
+ "step": 7607
+ },
+ {
+ "epoch": 24.30670926517572,
+ "grad_norm": 0.1083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0372,
+ "step": 7608
+ },
+ {
+ "epoch": 24.309904153354633,
+ "grad_norm": 0.1943359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0495,
+ "step": 7609
+ },
+ {
+ "epoch": 24.313099041533548,
+ "grad_norm": 0.05712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.031,
+ "step": 7610
+ },
+ {
+ "epoch": 24.31629392971246,
+ "grad_norm": 0.1484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0467,
+ "step": 7611
+ },
+ {
+ "epoch": 24.319488817891372,
+ "grad_norm": 0.09912109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0429,
+ "step": 7612
+ },
+ {
+ "epoch": 24.322683706070286,
+ "grad_norm": 0.12158203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0319,
+ "step": 7613
+ },
+ {
+ "epoch": 24.3258785942492,
+ "grad_norm": 0.2021484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0304,
+ "step": 7614
+ },
+ {
+ "epoch": 24.329073482428115,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0384,
+ "step": 7615
+ },
+ {
+ "epoch": 24.33226837060703,
+ "grad_norm": 0.271484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0423,
+ "step": 7616
+ },
+ {
+ "epoch": 24.335463258785943,
+ "grad_norm": 0.2236328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0414,
+ "step": 7617
+ },
+ {
+ "epoch": 24.338658146964857,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0453,
+ "step": 7618
+ },
+ {
+ "epoch": 24.34185303514377,
+ "grad_norm": 0.12353515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0323,
+ "step": 7619
+ },
+ {
+ "epoch": 24.345047923322685,
+ "grad_norm": 0.1083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0345,
+ "step": 7620
+ },
+ {
+ "epoch": 24.3482428115016,
+ "grad_norm": 0.09033203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0509,
+ "step": 7621
+ },
+ {
+ "epoch": 24.35143769968051,
+ "grad_norm": 0.14453125,
+ "learning_rate": 0.0005,
+ "loss": 1.051,
+ "step": 7622
+ },
+ {
+ "epoch": 24.354632587859424,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0434,
+ "step": 7623
+ },
+ {
+ "epoch": 24.357827476038338,
+ "grad_norm": 0.150390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0385,
+ "step": 7624
+ },
+ {
+ "epoch": 24.361022364217252,
+ "grad_norm": 0.1005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.047,
+ "step": 7625
+ },
+ {
+ "epoch": 24.364217252396166,
+ "grad_norm": 0.2490234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0401,
+ "step": 7626
+ },
+ {
+ "epoch": 24.36741214057508,
+ "grad_norm": 0.142578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0386,
+ "step": 7627
+ },
+ {
+ "epoch": 24.370607028753994,
+ "grad_norm": 0.0986328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0439,
+ "step": 7628
+ },
+ {
+ "epoch": 24.37380191693291,
+ "grad_norm": 0.1474609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0509,
+ "step": 7629
+ },
+ {
+ "epoch": 24.376996805111823,
+ "grad_norm": 0.1533203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0337,
+ "step": 7630
+ },
+ {
+ "epoch": 24.380191693290733,
+ "grad_norm": 0.1962890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0513,
+ "step": 7631
+ },
+ {
+ "epoch": 24.383386581469647,
+ "grad_norm": 0.189453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0468,
+ "step": 7632
+ },
+ {
+ "epoch": 24.38658146964856,
+ "grad_norm": 0.1767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.047,
+ "step": 7633
+ },
+ {
+ "epoch": 24.389776357827476,
+ "grad_norm": 0.1796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0415,
+ "step": 7634
+ },
+ {
+ "epoch": 24.39297124600639,
+ "grad_norm": 0.1923828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0431,
+ "step": 7635
+ },
+ {
+ "epoch": 24.396166134185304,
+ "grad_norm": 0.171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0474,
+ "step": 7636
+ },
+ {
+ "epoch": 24.399361022364218,
+ "grad_norm": 0.228515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0316,
+ "step": 7637
+ },
+ {
+ "epoch": 24.402555910543132,
+ "grad_norm": 0.1943359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0439,
+ "step": 7638
+ },
+ {
+ "epoch": 24.405750798722046,
+ "grad_norm": 0.306640625,
+ "learning_rate": 0.0005,
+ "loss": 1.046,
+ "step": 7639
+ },
+ {
+ "epoch": 24.408945686900957,
+ "grad_norm": 0.189453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0428,
+ "step": 7640
+ },
+ {
+ "epoch": 24.41214057507987,
+ "grad_norm": 0.42578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0419,
+ "step": 7641
+ },
+ {
+ "epoch": 24.415335463258785,
+ "grad_norm": 0.15234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0525,
+ "step": 7642
+ },
+ {
+ "epoch": 24.4185303514377,
+ "grad_norm": 0.287109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0361,
+ "step": 7643
+ },
+ {
+ "epoch": 24.421725239616613,
+ "grad_norm": 0.1845703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0537,
+ "step": 7644
+ },
+ {
+ "epoch": 24.424920127795527,
+ "grad_norm": 0.396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0387,
+ "step": 7645
+ },
+ {
+ "epoch": 24.42811501597444,
+ "grad_norm": 0.16015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0448,
+ "step": 7646
+ },
+ {
+ "epoch": 24.431309904153355,
+ "grad_norm": 0.26171875,
+ "learning_rate": 0.0005,
+ "loss": 1.036,
+ "step": 7647
+ },
+ {
+ "epoch": 24.43450479233227,
+ "grad_norm": 0.2021484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0369,
+ "step": 7648
+ },
+ {
+ "epoch": 24.437699680511184,
+ "grad_norm": 0.2890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0449,
+ "step": 7649
+ },
+ {
+ "epoch": 24.440894568690094,
+ "grad_norm": 0.1982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0435,
+ "step": 7650
+ },
+ {
+ "epoch": 24.44408945686901,
+ "grad_norm": 0.298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0372,
+ "step": 7651
+ },
+ {
+ "epoch": 24.447284345047922,
+ "grad_norm": 0.1787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0492,
+ "step": 7652
+ },
+ {
+ "epoch": 24.450479233226837,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0473,
+ "step": 7653
+ },
+ {
+ "epoch": 24.45367412140575,
+ "grad_norm": 0.1318359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0447,
+ "step": 7654
+ },
+ {
+ "epoch": 24.456869009584665,
+ "grad_norm": 0.1123046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0471,
+ "step": 7655
+ },
+ {
+ "epoch": 24.46006389776358,
+ "grad_norm": 0.11181640625,
+ "learning_rate": 0.0005,
+ "loss": 1.032,
+ "step": 7656
+ },
+ {
+ "epoch": 24.463258785942493,
+ "grad_norm": 0.1689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0465,
+ "step": 7657
+ },
+ {
+ "epoch": 24.466453674121407,
+ "grad_norm": 0.1015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0505,
+ "step": 7658
+ },
+ {
+ "epoch": 24.46964856230032,
+ "grad_norm": 0.2041015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0499,
+ "step": 7659
+ },
+ {
+ "epoch": 24.472843450479232,
+ "grad_norm": 0.1064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.031,
+ "step": 7660
+ },
+ {
+ "epoch": 24.476038338658146,
+ "grad_norm": 0.1181640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0336,
+ "step": 7661
+ },
+ {
+ "epoch": 24.47923322683706,
+ "grad_norm": 0.12109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0569,
+ "step": 7662
+ },
+ {
+ "epoch": 24.482428115015974,
+ "grad_norm": 0.0966796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0512,
+ "step": 7663
+ },
+ {
+ "epoch": 24.48562300319489,
+ "grad_norm": 0.11572265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0433,
+ "step": 7664
+ },
+ {
+ "epoch": 24.488817891373802,
+ "grad_norm": 0.162109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0435,
+ "step": 7665
+ },
+ {
+ "epoch": 24.492012779552716,
+ "grad_norm": 0.12451171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0457,
+ "step": 7666
+ },
+ {
+ "epoch": 24.49520766773163,
+ "grad_norm": 0.2177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.042,
+ "step": 7667
+ },
+ {
+ "epoch": 24.498402555910545,
+ "grad_norm": 0.07861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0501,
+ "step": 7668
+ },
+ {
+ "epoch": 24.501597444089455,
+ "grad_norm": 0.08837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0415,
+ "step": 7669
+ },
+ {
+ "epoch": 24.50479233226837,
+ "grad_norm": 0.083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0472,
+ "step": 7670
+ },
+ {
+ "epoch": 24.507987220447284,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0409,
+ "step": 7671
+ },
+ {
+ "epoch": 24.511182108626198,
+ "grad_norm": 0.046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0467,
+ "step": 7672
+ },
+ {
+ "epoch": 24.51437699680511,
+ "grad_norm": 0.05810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0426,
+ "step": 7673
+ },
+ {
+ "epoch": 24.517571884984026,
+ "grad_norm": 0.054931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0344,
+ "step": 7674
+ },
+ {
+ "epoch": 24.52076677316294,
+ "grad_norm": 0.06103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0545,
+ "step": 7675
+ },
+ {
+ "epoch": 24.523961661341854,
+ "grad_norm": 0.0517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0531,
+ "step": 7676
+ },
+ {
+ "epoch": 24.527156549520768,
+ "grad_norm": 0.052734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0467,
+ "step": 7677
+ },
+ {
+ "epoch": 24.53035143769968,
+ "grad_norm": 0.060302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0431,
+ "step": 7678
+ },
+ {
+ "epoch": 24.533546325878593,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0343,
+ "step": 7679
+ },
+ {
+ "epoch": 24.536741214057507,
+ "grad_norm": 0.048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.049,
+ "step": 7680
+ },
+ {
+ "epoch": 24.53993610223642,
+ "grad_norm": 0.05908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.044,
+ "step": 7681
+ },
+ {
+ "epoch": 24.543130990415335,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0378,
+ "step": 7682
+ },
+ {
+ "epoch": 24.54632587859425,
+ "grad_norm": 0.08935546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0458,
+ "step": 7683
+ },
+ {
+ "epoch": 24.549520766773163,
+ "grad_norm": 0.09375,
+ "learning_rate": 0.0005,
+ "loss": 1.0232,
+ "step": 7684
+ },
+ {
+ "epoch": 24.552715654952078,
+ "grad_norm": 0.1171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0432,
+ "step": 7685
+ },
+ {
+ "epoch": 24.55591054313099,
+ "grad_norm": 0.049560546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0396,
+ "step": 7686
+ },
+ {
+ "epoch": 24.559105431309906,
+ "grad_norm": 0.0595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0384,
+ "step": 7687
+ },
+ {
+ "epoch": 24.562300319488816,
+ "grad_norm": 0.062255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0392,
+ "step": 7688
+ },
+ {
+ "epoch": 24.56549520766773,
+ "grad_norm": 0.0615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.045,
+ "step": 7689
+ },
+ {
+ "epoch": 24.568690095846645,
+ "grad_norm": 0.08544921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0444,
+ "step": 7690
+ },
+ {
+ "epoch": 24.57188498402556,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0504,
+ "step": 7691
+ },
+ {
+ "epoch": 24.575079872204473,
+ "grad_norm": 0.0458984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0417,
+ "step": 7692
+ },
+ {
+ "epoch": 24.578274760383387,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0444,
+ "step": 7693
+ },
+ {
+ "epoch": 24.5814696485623,
+ "grad_norm": 0.08349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0542,
+ "step": 7694
+ },
+ {
+ "epoch": 24.584664536741215,
+ "grad_norm": 0.08251953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0425,
+ "step": 7695
+ },
+ {
+ "epoch": 24.58785942492013,
+ "grad_norm": 0.08544921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0492,
+ "step": 7696
+ },
+ {
+ "epoch": 24.591054313099043,
+ "grad_norm": 0.0595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0413,
+ "step": 7697
+ },
+ {
+ "epoch": 24.594249201277954,
+ "grad_norm": 0.0546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0476,
+ "step": 7698
+ },
+ {
+ "epoch": 24.597444089456868,
+ "grad_norm": 0.08349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0437,
+ "step": 7699
+ },
+ {
+ "epoch": 24.600638977635782,
+ "grad_norm": 0.054443359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0478,
+ "step": 7700
+ },
+ {
+ "epoch": 24.603833865814696,
+ "grad_norm": 0.06201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0526,
+ "step": 7701
+ },
+ {
+ "epoch": 24.60702875399361,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0417,
+ "step": 7702
+ },
+ {
+ "epoch": 24.610223642172524,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.033,
+ "step": 7703
+ },
+ {
+ "epoch": 24.61341853035144,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0394,
+ "step": 7704
+ },
+ {
+ "epoch": 24.616613418530353,
+ "grad_norm": 0.07861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0372,
+ "step": 7705
+ },
+ {
+ "epoch": 24.619808306709267,
+ "grad_norm": 0.05908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0448,
+ "step": 7706
+ },
+ {
+ "epoch": 24.623003194888177,
+ "grad_norm": 0.1123046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0408,
+ "step": 7707
+ },
+ {
+ "epoch": 24.62619808306709,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0334,
+ "step": 7708
+ },
+ {
+ "epoch": 24.629392971246006,
+ "grad_norm": 0.09814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0359,
+ "step": 7709
+ },
+ {
+ "epoch": 24.63258785942492,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0343,
+ "step": 7710
+ },
+ {
+ "epoch": 24.635782747603834,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0512,
+ "step": 7711
+ },
+ {
+ "epoch": 24.638977635782748,
+ "grad_norm": 0.051513671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0407,
+ "step": 7712
+ },
+ {
+ "epoch": 24.642172523961662,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0451,
+ "step": 7713
+ },
+ {
+ "epoch": 24.645367412140576,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0462,
+ "step": 7714
+ },
+ {
+ "epoch": 24.64856230031949,
+ "grad_norm": 0.05908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0421,
+ "step": 7715
+ },
+ {
+ "epoch": 24.6517571884984,
+ "grad_norm": 0.0478515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0441,
+ "step": 7716
+ },
+ {
+ "epoch": 24.654952076677315,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0388,
+ "step": 7717
+ },
+ {
+ "epoch": 24.65814696485623,
+ "grad_norm": 0.0556640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0376,
+ "step": 7718
+ },
+ {
+ "epoch": 24.661341853035143,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.041,
+ "step": 7719
+ },
+ {
+ "epoch": 24.664536741214057,
+ "grad_norm": 0.051025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0548,
+ "step": 7720
+ },
+ {
+ "epoch": 24.66773162939297,
+ "grad_norm": 0.1103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0533,
+ "step": 7721
+ },
+ {
+ "epoch": 24.670926517571885,
+ "grad_norm": 0.047119140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0458,
+ "step": 7722
+ },
+ {
+ "epoch": 24.6741214057508,
+ "grad_norm": 0.12890625,
+ "learning_rate": 0.0005,
+ "loss": 1.041,
+ "step": 7723
+ },
+ {
+ "epoch": 24.677316293929714,
+ "grad_norm": 0.0556640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0452,
+ "step": 7724
+ },
+ {
+ "epoch": 24.680511182108628,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0379,
+ "step": 7725
+ },
+ {
+ "epoch": 24.68370607028754,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0437,
+ "step": 7726
+ },
+ {
+ "epoch": 24.686900958466452,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0443,
+ "step": 7727
+ },
+ {
+ "epoch": 24.690095846645367,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0563,
+ "step": 7728
+ },
+ {
+ "epoch": 24.69329073482428,
+ "grad_norm": 0.12255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0556,
+ "step": 7729
+ },
+ {
+ "epoch": 24.696485623003195,
+ "grad_norm": 0.0537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0327,
+ "step": 7730
+ },
+ {
+ "epoch": 24.69968051118211,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0469,
+ "step": 7731
+ },
+ {
+ "epoch": 24.702875399361023,
+ "grad_norm": 0.05859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0288,
+ "step": 7732
+ },
+ {
+ "epoch": 24.706070287539937,
+ "grad_norm": 0.08544921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0408,
+ "step": 7733
+ },
+ {
+ "epoch": 24.70926517571885,
+ "grad_norm": 0.051513671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0409,
+ "step": 7734
+ },
+ {
+ "epoch": 24.712460063897765,
+ "grad_norm": 0.09130859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0434,
+ "step": 7735
+ },
+ {
+ "epoch": 24.715654952076676,
+ "grad_norm": 0.059814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.041,
+ "step": 7736
+ },
+ {
+ "epoch": 24.71884984025559,
+ "grad_norm": 0.09326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0434,
+ "step": 7737
+ },
+ {
+ "epoch": 24.722044728434504,
+ "grad_norm": 0.05078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0424,
+ "step": 7738
+ },
+ {
+ "epoch": 24.72523961661342,
+ "grad_norm": 0.0947265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0519,
+ "step": 7739
+ },
+ {
+ "epoch": 24.728434504792332,
+ "grad_norm": 0.058349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0306,
+ "step": 7740
+ },
+ {
+ "epoch": 24.731629392971247,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0285,
+ "step": 7741
+ },
+ {
+ "epoch": 24.73482428115016,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0377,
+ "step": 7742
+ },
+ {
+ "epoch": 24.738019169329075,
+ "grad_norm": 0.06201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0433,
+ "step": 7743
+ },
+ {
+ "epoch": 24.74121405750799,
+ "grad_norm": 0.0986328125,
+ "learning_rate": 0.0005,
+ "loss": 1.051,
+ "step": 7744
+ },
+ {
+ "epoch": 24.7444089456869,
+ "grad_norm": 0.054443359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0496,
+ "step": 7745
+ },
+ {
+ "epoch": 24.747603833865814,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0402,
+ "step": 7746
+ },
+ {
+ "epoch": 24.750798722044728,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0475,
+ "step": 7747
+ },
+ {
+ "epoch": 24.75399361022364,
+ "grad_norm": 0.091796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0552,
+ "step": 7748
+ },
+ {
+ "epoch": 24.757188498402556,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0387,
+ "step": 7749
+ },
+ {
+ "epoch": 24.76038338658147,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0463,
+ "step": 7750
+ },
+ {
+ "epoch": 24.763578274760384,
+ "grad_norm": 0.060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0479,
+ "step": 7751
+ },
+ {
+ "epoch": 24.766773162939298,
+ "grad_norm": 0.05419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0342,
+ "step": 7752
+ },
+ {
+ "epoch": 24.769968051118212,
+ "grad_norm": 0.080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.042,
+ "step": 7753
+ },
+ {
+ "epoch": 24.773162939297123,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0426,
+ "step": 7754
+ },
+ {
+ "epoch": 24.776357827476037,
+ "grad_norm": 0.05859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0467,
+ "step": 7755
+ },
+ {
+ "epoch": 24.77955271565495,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0335,
+ "step": 7756
+ },
+ {
+ "epoch": 24.782747603833865,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0423,
+ "step": 7757
+ },
+ {
+ "epoch": 24.78594249201278,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0454,
+ "step": 7758
+ },
+ {
+ "epoch": 24.789137380191693,
+ "grad_norm": 0.055908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0496,
+ "step": 7759
+ },
+ {
+ "epoch": 24.792332268370608,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0447,
+ "step": 7760
+ },
+ {
+ "epoch": 24.79552715654952,
+ "grad_norm": 0.057861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0435,
+ "step": 7761
+ },
+ {
+ "epoch": 24.798722044728436,
+ "grad_norm": 0.055419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0422,
+ "step": 7762
+ },
+ {
+ "epoch": 24.80191693290735,
+ "grad_norm": 0.054443359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0384,
+ "step": 7763
+ },
+ {
+ "epoch": 24.80511182108626,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0422,
+ "step": 7764
+ },
+ {
+ "epoch": 24.808306709265175,
+ "grad_norm": 0.09375,
+ "learning_rate": 0.0005,
+ "loss": 1.0481,
+ "step": 7765
+ },
+ {
+ "epoch": 24.81150159744409,
+ "grad_norm": 0.047607421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0468,
+ "step": 7766
+ },
+ {
+ "epoch": 24.814696485623003,
+ "grad_norm": 0.05810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.042,
+ "step": 7767
+ },
+ {
+ "epoch": 24.817891373801917,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0589,
+ "step": 7768
+ },
+ {
+ "epoch": 24.82108626198083,
+ "grad_norm": 0.060791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0429,
+ "step": 7769
+ },
+ {
+ "epoch": 24.824281150159745,
+ "grad_norm": 0.0849609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0351,
+ "step": 7770
+ },
+ {
+ "epoch": 24.82747603833866,
+ "grad_norm": 0.10791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0373,
+ "step": 7771
+ },
+ {
+ "epoch": 24.830670926517573,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0418,
+ "step": 7772
+ },
+ {
+ "epoch": 24.833865814696484,
+ "grad_norm": 0.053466796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0429,
+ "step": 7773
+ },
+ {
+ "epoch": 24.837060702875398,
+ "grad_norm": 0.052978515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0403,
+ "step": 7774
+ },
+ {
+ "epoch": 24.840255591054312,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0295,
+ "step": 7775
+ },
+ {
+ "epoch": 24.843450479233226,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0434,
+ "step": 7776
+ },
+ {
+ "epoch": 24.84664536741214,
+ "grad_norm": 0.0576171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0309,
+ "step": 7777
+ },
+ {
+ "epoch": 24.849840255591054,
+ "grad_norm": 0.061767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0403,
+ "step": 7778
+ },
+ {
+ "epoch": 24.85303514376997,
+ "grad_norm": 0.062255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.032,
+ "step": 7779
+ },
+ {
+ "epoch": 24.856230031948883,
+ "grad_norm": 0.052734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0429,
+ "step": 7780
+ },
+ {
+ "epoch": 24.859424920127797,
+ "grad_norm": 0.060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0293,
+ "step": 7781
+ },
+ {
+ "epoch": 24.86261980830671,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0584,
+ "step": 7782
+ },
+ {
+ "epoch": 24.86581469648562,
+ "grad_norm": 0.051513671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0325,
+ "step": 7783
+ },
+ {
+ "epoch": 24.869009584664536,
+ "grad_norm": 0.0908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0409,
+ "step": 7784
+ },
+ {
+ "epoch": 24.87220447284345,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0521,
+ "step": 7785
+ },
+ {
+ "epoch": 24.875399361022364,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0344,
+ "step": 7786
+ },
+ {
+ "epoch": 24.878594249201278,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0423,
+ "step": 7787
+ },
+ {
+ "epoch": 24.881789137380192,
+ "grad_norm": 0.0927734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0402,
+ "step": 7788
+ },
+ {
+ "epoch": 24.884984025559106,
+ "grad_norm": 0.0595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.037,
+ "step": 7789
+ },
+ {
+ "epoch": 24.88817891373802,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0357,
+ "step": 7790
+ },
+ {
+ "epoch": 24.891373801916934,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0434,
+ "step": 7791
+ },
+ {
+ "epoch": 24.894568690095845,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0461,
+ "step": 7792
+ },
+ {
+ "epoch": 24.89776357827476,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0445,
+ "step": 7793
+ },
+ {
+ "epoch": 24.900958466453673,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0338,
+ "step": 7794
+ },
+ {
+ "epoch": 24.904153354632587,
+ "grad_norm": 0.059326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0368,
+ "step": 7795
+ },
+ {
+ "epoch": 24.9073482428115,
+ "grad_norm": 0.05615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0308,
+ "step": 7796
+ },
+ {
+ "epoch": 24.910543130990416,
+ "grad_norm": 0.0830078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0436,
+ "step": 7797
+ },
+ {
+ "epoch": 24.91373801916933,
+ "grad_norm": 0.054931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0444,
+ "step": 7798
+ },
+ {
+ "epoch": 24.916932907348244,
+ "grad_norm": 0.053955078125,
+ "learning_rate": 0.0005,
+ "loss": 1.042,
+ "step": 7799
+ },
+ {
+ "epoch": 24.920127795527158,
+ "grad_norm": 0.059326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0547,
+ "step": 7800
+ },
+ {
+ "epoch": 24.923322683706072,
+ "grad_norm": 0.046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0479,
+ "step": 7801
+ },
+ {
+ "epoch": 24.926517571884983,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0363,
+ "step": 7802
+ },
+ {
+ "epoch": 24.929712460063897,
+ "grad_norm": 0.0595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0378,
+ "step": 7803
+ },
+ {
+ "epoch": 24.93290734824281,
+ "grad_norm": 0.053466796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0429,
+ "step": 7804
+ },
+ {
+ "epoch": 24.936102236421725,
+ "grad_norm": 0.08056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0384,
+ "step": 7805
+ },
+ {
+ "epoch": 24.93929712460064,
+ "grad_norm": 0.05517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0394,
+ "step": 7806
+ },
+ {
+ "epoch": 24.942492012779553,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0451,
+ "step": 7807
+ },
+ {
+ "epoch": 24.945686900958467,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0382,
+ "step": 7808
+ },
+ {
+ "epoch": 24.94888178913738,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0422,
+ "step": 7809
+ },
+ {
+ "epoch": 24.952076677316295,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0507,
+ "step": 7810
+ },
+ {
+ "epoch": 24.955271565495206,
+ "grad_norm": 0.10791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0385,
+ "step": 7811
+ },
+ {
+ "epoch": 24.95846645367412,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0379,
+ "step": 7812
+ },
+ {
+ "epoch": 24.961661341853034,
+ "grad_norm": 0.12890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0433,
+ "step": 7813
+ },
+ {
+ "epoch": 24.96485623003195,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0415,
+ "step": 7814
+ },
+ {
+ "epoch": 24.968051118210862,
+ "grad_norm": 0.0830078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0326,
+ "step": 7815
+ },
+ {
+ "epoch": 24.971246006389777,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0386,
+ "step": 7816
+ },
+ {
+ "epoch": 24.97444089456869,
+ "grad_norm": 0.08251953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0375,
+ "step": 7817
+ },
+ {
+ "epoch": 24.977635782747605,
+ "grad_norm": 0.060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0488,
+ "step": 7818
+ },
+ {
+ "epoch": 24.98083067092652,
+ "grad_norm": 0.0810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0507,
+ "step": 7819
+ },
+ {
+ "epoch": 24.984025559105433,
+ "grad_norm": 0.0927734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0492,
+ "step": 7820
+ },
+ {
+ "epoch": 24.987220447284344,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0495,
+ "step": 7821
+ },
+ {
+ "epoch": 24.990415335463258,
+ "grad_norm": 0.0927734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0434,
+ "step": 7822
+ },
+ {
+ "epoch": 24.99361022364217,
+ "grad_norm": 0.052490234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0545,
+ "step": 7823
+ },
+ {
+ "epoch": 24.996805111821086,
+ "grad_norm": 0.10986328125,
+ "learning_rate": 0.0005,
+ "loss": 1.043,
+ "step": 7824
+ },
+ {
+ "epoch": 25.0,
+ "grad_norm": 0.08544921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0469,
+ "step": 7825
+ },
+ {
+ "epoch": 25.003194888178914,
+ "grad_norm": 0.15234375,
+ "learning_rate": 0.0005,
+ "loss": 1.037,
+ "step": 7826
+ },
+ {
+ "epoch": 25.00638977635783,
+ "grad_norm": 0.059814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0385,
+ "step": 7827
+ },
+ {
+ "epoch": 25.009584664536742,
+ "grad_norm": 0.083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.028,
+ "step": 7828
+ },
+ {
+ "epoch": 25.012779552715656,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.044,
+ "step": 7829
+ },
+ {
+ "epoch": 25.015974440894567,
+ "grad_norm": 0.080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0481,
+ "step": 7830
+ },
+ {
+ "epoch": 25.01916932907348,
+ "grad_norm": 0.062255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0541,
+ "step": 7831
+ },
+ {
+ "epoch": 25.022364217252395,
+ "grad_norm": 0.09765625,
+ "learning_rate": 0.0005,
+ "loss": 1.05,
+ "step": 7832
+ },
+ {
+ "epoch": 25.02555910543131,
+ "grad_norm": 0.046142578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0264,
+ "step": 7833
+ },
+ {
+ "epoch": 25.028753993610223,
+ "grad_norm": 0.10791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.045,
+ "step": 7834
+ },
+ {
+ "epoch": 25.031948881789138,
+ "grad_norm": 0.058837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0424,
+ "step": 7835
+ },
+ {
+ "epoch": 25.03514376996805,
+ "grad_norm": 0.0908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0536,
+ "step": 7836
+ },
+ {
+ "epoch": 25.038338658146966,
+ "grad_norm": 0.055908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0417,
+ "step": 7837
+ },
+ {
+ "epoch": 25.04153354632588,
+ "grad_norm": 0.059814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0529,
+ "step": 7838
+ },
+ {
+ "epoch": 25.044728434504794,
+ "grad_norm": 0.0576171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0475,
+ "step": 7839
+ },
+ {
+ "epoch": 25.047923322683705,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0448,
+ "step": 7840
+ },
+ {
+ "epoch": 25.05111821086262,
+ "grad_norm": 0.060302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0405,
+ "step": 7841
+ },
+ {
+ "epoch": 25.054313099041533,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0288,
+ "step": 7842
+ },
+ {
+ "epoch": 25.057507987220447,
+ "grad_norm": 0.05712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0419,
+ "step": 7843
+ },
+ {
+ "epoch": 25.06070287539936,
+ "grad_norm": 0.052490234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0384,
+ "step": 7844
+ },
+ {
+ "epoch": 25.063897763578275,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0457,
+ "step": 7845
+ },
+ {
+ "epoch": 25.06709265175719,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.037,
+ "step": 7846
+ },
+ {
+ "epoch": 25.070287539936103,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0313,
+ "step": 7847
+ },
+ {
+ "epoch": 25.073482428115017,
+ "grad_norm": 0.052734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0426,
+ "step": 7848
+ },
+ {
+ "epoch": 25.076677316293928,
+ "grad_norm": 0.049072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0529,
+ "step": 7849
+ },
+ {
+ "epoch": 25.079872204472842,
+ "grad_norm": 0.060302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0392,
+ "step": 7850
+ },
+ {
+ "epoch": 25.083067092651756,
+ "grad_norm": 0.04833984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0446,
+ "step": 7851
+ },
+ {
+ "epoch": 25.08626198083067,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0357,
+ "step": 7852
+ },
+ {
+ "epoch": 25.089456869009584,
+ "grad_norm": 0.0615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0368,
+ "step": 7853
+ },
+ {
+ "epoch": 25.0926517571885,
+ "grad_norm": 0.060302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0382,
+ "step": 7854
+ },
+ {
+ "epoch": 25.095846645367413,
+ "grad_norm": 0.05712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0345,
+ "step": 7855
+ },
+ {
+ "epoch": 25.099041533546327,
+ "grad_norm": 0.052001953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0397,
+ "step": 7856
+ },
+ {
+ "epoch": 25.10223642172524,
+ "grad_norm": 0.06005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0533,
+ "step": 7857
+ },
+ {
+ "epoch": 25.105431309904155,
+ "grad_norm": 0.0517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0415,
+ "step": 7858
+ },
+ {
+ "epoch": 25.108626198083066,
+ "grad_norm": 0.051513671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0636,
+ "step": 7859
+ },
+ {
+ "epoch": 25.11182108626198,
+ "grad_norm": 0.0546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0352,
+ "step": 7860
+ },
+ {
+ "epoch": 25.115015974440894,
+ "grad_norm": 0.052490234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0522,
+ "step": 7861
+ },
+ {
+ "epoch": 25.118210862619808,
+ "grad_norm": 0.055908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0456,
+ "step": 7862
+ },
+ {
+ "epoch": 25.121405750798722,
+ "grad_norm": 0.052490234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0502,
+ "step": 7863
+ },
+ {
+ "epoch": 25.124600638977636,
+ "grad_norm": 0.056884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0414,
+ "step": 7864
+ },
+ {
+ "epoch": 25.12779552715655,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0375,
+ "step": 7865
+ },
+ {
+ "epoch": 25.130990415335464,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0371,
+ "step": 7866
+ },
+ {
+ "epoch": 25.13418530351438,
+ "grad_norm": 0.09228515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0428,
+ "step": 7867
+ },
+ {
+ "epoch": 25.13738019169329,
+ "grad_norm": 0.061279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0508,
+ "step": 7868
+ },
+ {
+ "epoch": 25.140575079872203,
+ "grad_norm": 0.05517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0464,
+ "step": 7869
+ },
+ {
+ "epoch": 25.143769968051117,
+ "grad_norm": 0.0615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0432,
+ "step": 7870
+ },
+ {
+ "epoch": 25.14696485623003,
+ "grad_norm": 0.057373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0377,
+ "step": 7871
+ },
+ {
+ "epoch": 25.150159744408946,
+ "grad_norm": 0.08740234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0376,
+ "step": 7872
+ },
+ {
+ "epoch": 25.15335463258786,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0418,
+ "step": 7873
+ },
+ {
+ "epoch": 25.156549520766774,
+ "grad_norm": 0.0615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0396,
+ "step": 7874
+ },
+ {
+ "epoch": 25.159744408945688,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0369,
+ "step": 7875
+ },
+ {
+ "epoch": 25.162939297124602,
+ "grad_norm": 0.057861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0464,
+ "step": 7876
+ },
+ {
+ "epoch": 25.166134185303516,
+ "grad_norm": 0.062255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0507,
+ "step": 7877
+ },
+ {
+ "epoch": 25.169329073482427,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0441,
+ "step": 7878
+ },
+ {
+ "epoch": 25.17252396166134,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0371,
+ "step": 7879
+ },
+ {
+ "epoch": 25.175718849840255,
+ "grad_norm": 0.07861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0371,
+ "step": 7880
+ },
+ {
+ "epoch": 25.17891373801917,
+ "grad_norm": 0.08251953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0396,
+ "step": 7881
+ },
+ {
+ "epoch": 25.182108626198083,
+ "grad_norm": 0.09912109375,
+ "learning_rate": 0.0005,
+ "loss": 1.037,
+ "step": 7882
+ },
+ {
+ "epoch": 25.185303514376997,
+ "grad_norm": 0.0546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0333,
+ "step": 7883
+ },
+ {
+ "epoch": 25.18849840255591,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0425,
+ "step": 7884
+ },
+ {
+ "epoch": 25.191693290734825,
+ "grad_norm": 0.08740234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0398,
+ "step": 7885
+ },
+ {
+ "epoch": 25.19488817891374,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0446,
+ "step": 7886
+ },
+ {
+ "epoch": 25.19808306709265,
+ "grad_norm": 0.10888671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0387,
+ "step": 7887
+ },
+ {
+ "epoch": 25.201277955271564,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0476,
+ "step": 7888
+ },
+ {
+ "epoch": 25.20447284345048,
+ "grad_norm": 0.10791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0431,
+ "step": 7889
+ },
+ {
+ "epoch": 25.207667731629392,
+ "grad_norm": 0.08056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0449,
+ "step": 7890
+ },
+ {
+ "epoch": 25.210862619808307,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0456,
+ "step": 7891
+ },
+ {
+ "epoch": 25.21405750798722,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0497,
+ "step": 7892
+ },
+ {
+ "epoch": 25.217252396166135,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.042,
+ "step": 7893
+ },
+ {
+ "epoch": 25.22044728434505,
+ "grad_norm": 0.053955078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0478,
+ "step": 7894
+ },
+ {
+ "epoch": 25.223642172523963,
+ "grad_norm": 0.055419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0482,
+ "step": 7895
+ },
+ {
+ "epoch": 25.226837060702877,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0518,
+ "step": 7896
+ },
+ {
+ "epoch": 25.230031948881788,
+ "grad_norm": 0.0908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0402,
+ "step": 7897
+ },
+ {
+ "epoch": 25.233226837060702,
+ "grad_norm": 0.146484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0366,
+ "step": 7898
+ },
+ {
+ "epoch": 25.236421725239616,
+ "grad_norm": 0.08154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0313,
+ "step": 7899
+ },
+ {
+ "epoch": 25.23961661341853,
+ "grad_norm": 0.1240234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0473,
+ "step": 7900
+ },
+ {
+ "epoch": 25.242811501597444,
+ "grad_norm": 0.09228515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0405,
+ "step": 7901
+ },
+ {
+ "epoch": 25.24600638977636,
+ "grad_norm": 0.0859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0306,
+ "step": 7902
+ },
+ {
+ "epoch": 25.249201277955272,
+ "grad_norm": 0.099609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0514,
+ "step": 7903
+ },
+ {
+ "epoch": 25.252396166134186,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0582,
+ "step": 7904
+ },
+ {
+ "epoch": 25.2555910543131,
+ "grad_norm": 0.0830078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0364,
+ "step": 7905
+ },
+ {
+ "epoch": 25.25878594249201,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0358,
+ "step": 7906
+ },
+ {
+ "epoch": 25.261980830670925,
+ "grad_norm": 0.1650390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0335,
+ "step": 7907
+ },
+ {
+ "epoch": 25.26517571884984,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0374,
+ "step": 7908
+ },
+ {
+ "epoch": 25.268370607028753,
+ "grad_norm": 0.11962890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0519,
+ "step": 7909
+ },
+ {
+ "epoch": 25.271565495207668,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0441,
+ "step": 7910
+ },
+ {
+ "epoch": 25.27476038338658,
+ "grad_norm": 0.08251953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0448,
+ "step": 7911
+ },
+ {
+ "epoch": 25.277955271565496,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0338,
+ "step": 7912
+ },
+ {
+ "epoch": 25.28115015974441,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.037,
+ "step": 7913
+ },
+ {
+ "epoch": 25.284345047923324,
+ "grad_norm": 0.08984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0446,
+ "step": 7914
+ },
+ {
+ "epoch": 25.287539936102238,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0416,
+ "step": 7915
+ },
+ {
+ "epoch": 25.29073482428115,
+ "grad_norm": 0.12451171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0523,
+ "step": 7916
+ },
+ {
+ "epoch": 25.293929712460063,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0373,
+ "step": 7917
+ },
+ {
+ "epoch": 25.297124600638977,
+ "grad_norm": 0.1474609375,
+ "learning_rate": 0.0005,
+ "loss": 1.044,
+ "step": 7918
+ },
+ {
+ "epoch": 25.30031948881789,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0453,
+ "step": 7919
+ },
+ {
+ "epoch": 25.303514376996805,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.034,
+ "step": 7920
+ },
+ {
+ "epoch": 25.30670926517572,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0393,
+ "step": 7921
+ },
+ {
+ "epoch": 25.309904153354633,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.043,
+ "step": 7922
+ },
+ {
+ "epoch": 25.313099041533548,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0512,
+ "step": 7923
+ },
+ {
+ "epoch": 25.31629392971246,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0449,
+ "step": 7924
+ },
+ {
+ "epoch": 25.319488817891372,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0398,
+ "step": 7925
+ },
+ {
+ "epoch": 25.322683706070286,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0408,
+ "step": 7926
+ },
+ {
+ "epoch": 25.3258785942492,
+ "grad_norm": 0.048095703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0305,
+ "step": 7927
+ },
+ {
+ "epoch": 25.329073482428115,
+ "grad_norm": 0.05224609375,
+ "learning_rate": 0.0005,
+ "loss": 1.031,
+ "step": 7928
+ },
+ {
+ "epoch": 25.33226837060703,
+ "grad_norm": 0.05615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0517,
+ "step": 7929
+ },
+ {
+ "epoch": 25.335463258785943,
+ "grad_norm": 0.05615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0427,
+ "step": 7930
+ },
+ {
+ "epoch": 25.338658146964857,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0317,
+ "step": 7931
+ },
+ {
+ "epoch": 25.34185303514377,
+ "grad_norm": 0.05029296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0425,
+ "step": 7932
+ },
+ {
+ "epoch": 25.345047923322685,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0511,
+ "step": 7933
+ },
+ {
+ "epoch": 25.3482428115016,
+ "grad_norm": 0.051513671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0349,
+ "step": 7934
+ },
+ {
+ "epoch": 25.35143769968051,
+ "grad_norm": 0.0869140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0503,
+ "step": 7935
+ },
+ {
+ "epoch": 25.354632587859424,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0535,
+ "step": 7936
+ },
+ {
+ "epoch": 25.357827476038338,
+ "grad_norm": 0.05078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0423,
+ "step": 7937
+ },
+ {
+ "epoch": 25.361022364217252,
+ "grad_norm": 0.059814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0355,
+ "step": 7938
+ },
+ {
+ "epoch": 25.364217252396166,
+ "grad_norm": 0.0556640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0366,
+ "step": 7939
+ },
+ {
+ "epoch": 25.36741214057508,
+ "grad_norm": 0.09228515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0474,
+ "step": 7940
+ },
+ {
+ "epoch": 25.370607028753994,
+ "grad_norm": 0.0869140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0421,
+ "step": 7941
+ },
+ {
+ "epoch": 25.37380191693291,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0424,
+ "step": 7942
+ },
+ {
+ "epoch": 25.376996805111823,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0418,
+ "step": 7943
+ },
+ {
+ "epoch": 25.380191693290733,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0448,
+ "step": 7944
+ },
+ {
+ "epoch": 25.383386581469647,
+ "grad_norm": 0.060302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0478,
+ "step": 7945
+ },
+ {
+ "epoch": 25.38658146964856,
+ "grad_norm": 0.061279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0467,
+ "step": 7946
+ },
+ {
+ "epoch": 25.389776357827476,
+ "grad_norm": 0.08544921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0397,
+ "step": 7947
+ },
+ {
+ "epoch": 25.39297124600639,
+ "grad_norm": 0.09619140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0492,
+ "step": 7948
+ },
+ {
+ "epoch": 25.396166134185304,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0443,
+ "step": 7949
+ },
+ {
+ "epoch": 25.399361022364218,
+ "grad_norm": 0.08837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0432,
+ "step": 7950
+ },
+ {
+ "epoch": 25.402555910543132,
+ "grad_norm": 0.061279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0392,
+ "step": 7951
+ },
+ {
+ "epoch": 25.405750798722046,
+ "grad_norm": 0.057373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.027,
+ "step": 7952
+ },
+ {
+ "epoch": 25.408945686900957,
+ "grad_norm": 0.054443359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0471,
+ "step": 7953
+ },
+ {
+ "epoch": 25.41214057507987,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0458,
+ "step": 7954
+ },
+ {
+ "epoch": 25.415335463258785,
+ "grad_norm": 0.0537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0426,
+ "step": 7955
+ },
+ {
+ "epoch": 25.4185303514377,
+ "grad_norm": 0.099609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0325,
+ "step": 7956
+ },
+ {
+ "epoch": 25.421725239616613,
+ "grad_norm": 0.05615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0314,
+ "step": 7957
+ },
+ {
+ "epoch": 25.424920127795527,
+ "grad_norm": 0.130859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0418,
+ "step": 7958
+ },
+ {
+ "epoch": 25.42811501597444,
+ "grad_norm": 0.053955078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0448,
+ "step": 7959
+ },
+ {
+ "epoch": 25.431309904153355,
+ "grad_norm": 0.1630859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0399,
+ "step": 7960
+ },
+ {
+ "epoch": 25.43450479233227,
+ "grad_norm": 0.1162109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0405,
+ "step": 7961
+ },
+ {
+ "epoch": 25.437699680511184,
+ "grad_norm": 0.09033203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0453,
+ "step": 7962
+ },
+ {
+ "epoch": 25.440894568690094,
+ "grad_norm": 0.10693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0323,
+ "step": 7963
+ },
+ {
+ "epoch": 25.44408945686901,
+ "grad_norm": 0.05224609375,
+ "learning_rate": 0.0005,
+ "loss": 1.048,
+ "step": 7964
+ },
+ {
+ "epoch": 25.447284345047922,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0408,
+ "step": 7965
+ },
+ {
+ "epoch": 25.450479233226837,
+ "grad_norm": 0.05419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0221,
+ "step": 7966
+ },
+ {
+ "epoch": 25.45367412140575,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0389,
+ "step": 7967
+ },
+ {
+ "epoch": 25.456869009584665,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0469,
+ "step": 7968
+ },
+ {
+ "epoch": 25.46006389776358,
+ "grad_norm": 0.130859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0409,
+ "step": 7969
+ },
+ {
+ "epoch": 25.463258785942493,
+ "grad_norm": 0.056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0364,
+ "step": 7970
+ },
+ {
+ "epoch": 25.466453674121407,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.044,
+ "step": 7971
+ },
+ {
+ "epoch": 25.46964856230032,
+ "grad_norm": 0.10302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0405,
+ "step": 7972
+ },
+ {
+ "epoch": 25.472843450479232,
+ "grad_norm": 0.12451171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0555,
+ "step": 7973
+ },
+ {
+ "epoch": 25.476038338658146,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0423,
+ "step": 7974
+ },
+ {
+ "epoch": 25.47923322683706,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0395,
+ "step": 7975
+ },
+ {
+ "epoch": 25.482428115015974,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0442,
+ "step": 7976
+ },
+ {
+ "epoch": 25.48562300319489,
+ "grad_norm": 0.057861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0476,
+ "step": 7977
+ },
+ {
+ "epoch": 25.488817891373802,
+ "grad_norm": 0.060302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0482,
+ "step": 7978
+ },
+ {
+ "epoch": 25.492012779552716,
+ "grad_norm": 0.0595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0477,
+ "step": 7979
+ },
+ {
+ "epoch": 25.49520766773163,
+ "grad_norm": 0.05517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0474,
+ "step": 7980
+ },
+ {
+ "epoch": 25.498402555910545,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0492,
+ "step": 7981
+ },
+ {
+ "epoch": 25.501597444089455,
+ "grad_norm": 0.0576171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0336,
+ "step": 7982
+ },
+ {
+ "epoch": 25.50479233226837,
+ "grad_norm": 0.054443359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0412,
+ "step": 7983
+ },
+ {
+ "epoch": 25.507987220447284,
+ "grad_norm": 0.103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0478,
+ "step": 7984
+ },
+ {
+ "epoch": 25.511182108626198,
+ "grad_norm": 0.043701171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0363,
+ "step": 7985
+ },
+ {
+ "epoch": 25.51437699680511,
+ "grad_norm": 0.1279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0349,
+ "step": 7986
+ },
+ {
+ "epoch": 25.517571884984026,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0334,
+ "step": 7987
+ },
+ {
+ "epoch": 25.52076677316294,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0301,
+ "step": 7988
+ },
+ {
+ "epoch": 25.523961661341854,
+ "grad_norm": 0.062255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0422,
+ "step": 7989
+ },
+ {
+ "epoch": 25.527156549520768,
+ "grad_norm": 0.05419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0418,
+ "step": 7990
+ },
+ {
+ "epoch": 25.53035143769968,
+ "grad_norm": 0.051513671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0458,
+ "step": 7991
+ },
+ {
+ "epoch": 25.533546325878593,
+ "grad_norm": 0.08349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0405,
+ "step": 7992
+ },
+ {
+ "epoch": 25.536741214057507,
+ "grad_norm": 0.059326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0433,
+ "step": 7993
+ },
+ {
+ "epoch": 25.53993610223642,
+ "grad_norm": 0.08203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0199,
+ "step": 7994
+ },
+ {
+ "epoch": 25.543130990415335,
+ "grad_norm": 0.1826171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0353,
+ "step": 7995
+ },
+ {
+ "epoch": 25.54632587859425,
+ "grad_norm": 0.05615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0511,
+ "step": 7996
+ },
+ {
+ "epoch": 25.549520766773163,
+ "grad_norm": 0.08349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0468,
+ "step": 7997
+ },
+ {
+ "epoch": 25.552715654952078,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0455,
+ "step": 7998
+ },
+ {
+ "epoch": 25.55591054313099,
+ "grad_norm": 0.054443359375,
+ "learning_rate": 0.0005,
+ "loss": 1.038,
+ "step": 7999
+ },
+ {
+ "epoch": 25.559105431309906,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0368,
+ "step": 8000
+ },
+ {
+ "epoch": 25.562300319488816,
+ "grad_norm": 0.051025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0397,
+ "step": 8001
+ },
+ {
+ "epoch": 25.56549520766773,
+ "grad_norm": 0.0830078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0324,
+ "step": 8002
+ },
+ {
+ "epoch": 25.568690095846645,
+ "grad_norm": 0.08251953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0317,
+ "step": 8003
+ },
+ {
+ "epoch": 25.57188498402556,
+ "grad_norm": 0.111328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0553,
+ "step": 8004
+ },
+ {
+ "epoch": 25.575079872204473,
+ "grad_norm": 0.1123046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0399,
+ "step": 8005
+ },
+ {
+ "epoch": 25.578274760383387,
+ "grad_norm": 0.1474609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0469,
+ "step": 8006
+ },
+ {
+ "epoch": 25.5814696485623,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0371,
+ "step": 8007
+ },
+ {
+ "epoch": 25.584664536741215,
+ "grad_norm": 0.10791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0485,
+ "step": 8008
+ },
+ {
+ "epoch": 25.58785942492013,
+ "grad_norm": 0.08203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0559,
+ "step": 8009
+ },
+ {
+ "epoch": 25.591054313099043,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0499,
+ "step": 8010
+ },
+ {
+ "epoch": 25.594249201277954,
+ "grad_norm": 0.1123046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0329,
+ "step": 8011
+ },
+ {
+ "epoch": 25.597444089456868,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0434,
+ "step": 8012
+ },
+ {
+ "epoch": 25.600638977635782,
+ "grad_norm": 0.09521484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0328,
+ "step": 8013
+ },
+ {
+ "epoch": 25.603833865814696,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0304,
+ "step": 8014
+ },
+ {
+ "epoch": 25.60702875399361,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0499,
+ "step": 8015
+ },
+ {
+ "epoch": 25.610223642172524,
+ "grad_norm": 0.08154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0502,
+ "step": 8016
+ },
+ {
+ "epoch": 25.61341853035144,
+ "grad_norm": 0.0830078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0398,
+ "step": 8017
+ },
+ {
+ "epoch": 25.616613418530353,
+ "grad_norm": 0.10205078125,
+ "learning_rate": 0.0005,
+ "loss": 1.033,
+ "step": 8018
+ },
+ {
+ "epoch": 25.619808306709267,
+ "grad_norm": 0.1455078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0347,
+ "step": 8019
+ },
+ {
+ "epoch": 25.623003194888177,
+ "grad_norm": 0.052490234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0352,
+ "step": 8020
+ },
+ {
+ "epoch": 25.62619808306709,
+ "grad_norm": 0.115234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0292,
+ "step": 8021
+ },
+ {
+ "epoch": 25.629392971246006,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.035,
+ "step": 8022
+ },
+ {
+ "epoch": 25.63258785942492,
+ "grad_norm": 0.12158203125,
+ "learning_rate": 0.0005,
+ "loss": 1.052,
+ "step": 8023
+ },
+ {
+ "epoch": 25.635782747603834,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0364,
+ "step": 8024
+ },
+ {
+ "epoch": 25.638977635782748,
+ "grad_norm": 0.1318359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0378,
+ "step": 8025
+ },
+ {
+ "epoch": 25.642172523961662,
+ "grad_norm": 0.1611328125,
+ "learning_rate": 0.0005,
+ "loss": 1.04,
+ "step": 8026
+ },
+ {
+ "epoch": 25.645367412140576,
+ "grad_norm": 0.06005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0258,
+ "step": 8027
+ },
+ {
+ "epoch": 25.64856230031949,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0457,
+ "step": 8028
+ },
+ {
+ "epoch": 25.6517571884984,
+ "grad_norm": 0.0869140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0397,
+ "step": 8029
+ },
+ {
+ "epoch": 25.654952076677315,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0435,
+ "step": 8030
+ },
+ {
+ "epoch": 25.65814696485623,
+ "grad_norm": 0.0556640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0277,
+ "step": 8031
+ },
+ {
+ "epoch": 25.661341853035143,
+ "grad_norm": 0.0595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0334,
+ "step": 8032
+ },
+ {
+ "epoch": 25.664536741214057,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0405,
+ "step": 8033
+ },
+ {
+ "epoch": 25.66773162939297,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0483,
+ "step": 8034
+ },
+ {
+ "epoch": 25.670926517571885,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0449,
+ "step": 8035
+ },
+ {
+ "epoch": 25.6741214057508,
+ "grad_norm": 0.056884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.034,
+ "step": 8036
+ },
+ {
+ "epoch": 25.677316293929714,
+ "grad_norm": 0.11181640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0346,
+ "step": 8037
+ },
+ {
+ "epoch": 25.680511182108628,
+ "grad_norm": 0.1728515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0242,
+ "step": 8038
+ },
+ {
+ "epoch": 25.68370607028754,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0502,
+ "step": 8039
+ },
+ {
+ "epoch": 25.686900958466452,
+ "grad_norm": 0.1279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0379,
+ "step": 8040
+ },
+ {
+ "epoch": 25.690095846645367,
+ "grad_norm": 0.05517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0457,
+ "step": 8041
+ },
+ {
+ "epoch": 25.69329073482428,
+ "grad_norm": 0.10693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0354,
+ "step": 8042
+ },
+ {
+ "epoch": 25.696485623003195,
+ "grad_norm": 0.046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0385,
+ "step": 8043
+ },
+ {
+ "epoch": 25.69968051118211,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0373,
+ "step": 8044
+ },
+ {
+ "epoch": 25.702875399361023,
+ "grad_norm": 0.061279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0433,
+ "step": 8045
+ },
+ {
+ "epoch": 25.706070287539937,
+ "grad_norm": 0.05908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0363,
+ "step": 8046
+ },
+ {
+ "epoch": 25.70926517571885,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0382,
+ "step": 8047
+ },
+ {
+ "epoch": 25.712460063897765,
+ "grad_norm": 0.052001953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0417,
+ "step": 8048
+ },
+ {
+ "epoch": 25.715654952076676,
+ "grad_norm": 0.06005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0594,
+ "step": 8049
+ },
+ {
+ "epoch": 25.71884984025559,
+ "grad_norm": 0.058349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0445,
+ "step": 8050
+ },
+ {
+ "epoch": 25.722044728434504,
+ "grad_norm": 0.08642578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0385,
+ "step": 8051
+ },
+ {
+ "epoch": 25.72523961661342,
+ "grad_norm": 0.0576171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0402,
+ "step": 8052
+ },
+ {
+ "epoch": 25.728434504792332,
+ "grad_norm": 0.08203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0387,
+ "step": 8053
+ },
+ {
+ "epoch": 25.731629392971247,
+ "grad_norm": 0.0478515625,
+ "learning_rate": 0.0005,
+ "loss": 1.03,
+ "step": 8054
+ },
+ {
+ "epoch": 25.73482428115016,
+ "grad_norm": 0.060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.028,
+ "step": 8055
+ },
+ {
+ "epoch": 25.738019169329075,
+ "grad_norm": 0.054931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0343,
+ "step": 8056
+ },
+ {
+ "epoch": 25.74121405750799,
+ "grad_norm": 0.05615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0447,
+ "step": 8057
+ },
+ {
+ "epoch": 25.7444089456869,
+ "grad_norm": 0.051513671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0421,
+ "step": 8058
+ },
+ {
+ "epoch": 25.747603833865814,
+ "grad_norm": 0.05078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0296,
+ "step": 8059
+ },
+ {
+ "epoch": 25.750798722044728,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0326,
+ "step": 8060
+ },
+ {
+ "epoch": 25.75399361022364,
+ "grad_norm": 0.052978515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0393,
+ "step": 8061
+ },
+ {
+ "epoch": 25.757188498402556,
+ "grad_norm": 0.08251953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0507,
+ "step": 8062
+ },
+ {
+ "epoch": 25.76038338658147,
+ "grad_norm": 0.051025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0383,
+ "step": 8063
+ },
+ {
+ "epoch": 25.763578274760384,
+ "grad_norm": 0.060791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0454,
+ "step": 8064
+ },
+ {
+ "epoch": 25.766773162939298,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0361,
+ "step": 8065
+ },
+ {
+ "epoch": 25.769968051118212,
+ "grad_norm": 0.051025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0482,
+ "step": 8066
+ },
+ {
+ "epoch": 25.773162939297123,
+ "grad_norm": 0.09716796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0448,
+ "step": 8067
+ },
+ {
+ "epoch": 25.776357827476037,
+ "grad_norm": 0.05322265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0469,
+ "step": 8068
+ },
+ {
+ "epoch": 25.77955271565495,
+ "grad_norm": 0.06201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0483,
+ "step": 8069
+ },
+ {
+ "epoch": 25.782747603833865,
+ "grad_norm": 0.06005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0443,
+ "step": 8070
+ },
+ {
+ "epoch": 25.78594249201278,
+ "grad_norm": 0.057373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0352,
+ "step": 8071
+ },
+ {
+ "epoch": 25.789137380191693,
+ "grad_norm": 0.060302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0466,
+ "step": 8072
+ },
+ {
+ "epoch": 25.792332268370608,
+ "grad_norm": 0.052734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0346,
+ "step": 8073
+ },
+ {
+ "epoch": 25.79552715654952,
+ "grad_norm": 0.051025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0301,
+ "step": 8074
+ },
+ {
+ "epoch": 25.798722044728436,
+ "grad_norm": 0.05517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0312,
+ "step": 8075
+ },
+ {
+ "epoch": 25.80191693290735,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.042,
+ "step": 8076
+ },
+ {
+ "epoch": 25.80511182108626,
+ "grad_norm": 0.048583984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0411,
+ "step": 8077
+ },
+ {
+ "epoch": 25.808306709265175,
+ "grad_norm": 0.0556640625,
+ "learning_rate": 0.0005,
+ "loss": 1.043,
+ "step": 8078
+ },
+ {
+ "epoch": 25.81150159744409,
+ "grad_norm": 0.057373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0377,
+ "step": 8079
+ },
+ {
+ "epoch": 25.814696485623003,
+ "grad_norm": 0.060302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0295,
+ "step": 8080
+ },
+ {
+ "epoch": 25.817891373801917,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.049,
+ "step": 8081
+ },
+ {
+ "epoch": 25.82108626198083,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0362,
+ "step": 8082
+ },
+ {
+ "epoch": 25.824281150159745,
+ "grad_norm": 0.0556640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0418,
+ "step": 8083
+ },
+ {
+ "epoch": 25.82747603833866,
+ "grad_norm": 0.056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0523,
+ "step": 8084
+ },
+ {
+ "epoch": 25.830670926517573,
+ "grad_norm": 0.06201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0374,
+ "step": 8085
+ },
+ {
+ "epoch": 25.833865814696484,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0598,
+ "step": 8086
+ },
+ {
+ "epoch": 25.837060702875398,
+ "grad_norm": 0.1064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0231,
+ "step": 8087
+ },
+ {
+ "epoch": 25.840255591054312,
+ "grad_norm": 0.05859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0429,
+ "step": 8088
+ },
+ {
+ "epoch": 25.843450479233226,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0318,
+ "step": 8089
+ },
+ {
+ "epoch": 25.84664536741214,
+ "grad_norm": 0.060791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0505,
+ "step": 8090
+ },
+ {
+ "epoch": 25.849840255591054,
+ "grad_norm": 0.060302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0366,
+ "step": 8091
+ },
+ {
+ "epoch": 25.85303514376997,
+ "grad_norm": 0.049072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0401,
+ "step": 8092
+ },
+ {
+ "epoch": 25.856230031948883,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0332,
+ "step": 8093
+ },
+ {
+ "epoch": 25.859424920127797,
+ "grad_norm": 0.056396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0518,
+ "step": 8094
+ },
+ {
+ "epoch": 25.86261980830671,
+ "grad_norm": 0.08837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0292,
+ "step": 8095
+ },
+ {
+ "epoch": 25.86581469648562,
+ "grad_norm": 0.056396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0394,
+ "step": 8096
+ },
+ {
+ "epoch": 25.869009584664536,
+ "grad_norm": 0.0546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0448,
+ "step": 8097
+ },
+ {
+ "epoch": 25.87220447284345,
+ "grad_norm": 0.08447265625,
+ "learning_rate": 0.0005,
+ "loss": 1.042,
+ "step": 8098
+ },
+ {
+ "epoch": 25.875399361022364,
+ "grad_norm": 0.060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0259,
+ "step": 8099
+ },
+ {
+ "epoch": 25.878594249201278,
+ "grad_norm": 0.057373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0348,
+ "step": 8100
+ },
+ {
+ "epoch": 25.881789137380192,
+ "grad_norm": 0.057861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.042,
+ "step": 8101
+ },
+ {
+ "epoch": 25.884984025559106,
+ "grad_norm": 0.060791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0386,
+ "step": 8102
+ },
+ {
+ "epoch": 25.88817891373802,
+ "grad_norm": 0.0546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0452,
+ "step": 8103
+ },
+ {
+ "epoch": 25.891373801916934,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0356,
+ "step": 8104
+ },
+ {
+ "epoch": 25.894568690095845,
+ "grad_norm": 0.050048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.041,
+ "step": 8105
+ },
+ {
+ "epoch": 25.89776357827476,
+ "grad_norm": 0.051025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0505,
+ "step": 8106
+ },
+ {
+ "epoch": 25.900958466453673,
+ "grad_norm": 0.0498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0364,
+ "step": 8107
+ },
+ {
+ "epoch": 25.904153354632587,
+ "grad_norm": 0.0458984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0404,
+ "step": 8108
+ },
+ {
+ "epoch": 25.9073482428115,
+ "grad_norm": 0.055908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0325,
+ "step": 8109
+ },
+ {
+ "epoch": 25.910543130990416,
+ "grad_norm": 0.058349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0333,
+ "step": 8110
+ },
+ {
+ "epoch": 25.91373801916933,
+ "grad_norm": 0.05078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0431,
+ "step": 8111
+ },
+ {
+ "epoch": 25.916932907348244,
+ "grad_norm": 0.056884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0346,
+ "step": 8112
+ },
+ {
+ "epoch": 25.920127795527158,
+ "grad_norm": 0.06005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0352,
+ "step": 8113
+ },
+ {
+ "epoch": 25.923322683706072,
+ "grad_norm": 0.05419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0388,
+ "step": 8114
+ },
+ {
+ "epoch": 25.926517571884983,
+ "grad_norm": 0.087890625,
+ "learning_rate": 0.0005,
+ "loss": 1.05,
+ "step": 8115
+ },
+ {
+ "epoch": 25.929712460063897,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0512,
+ "step": 8116
+ },
+ {
+ "epoch": 25.93290734824281,
+ "grad_norm": 0.09375,
+ "learning_rate": 0.0005,
+ "loss": 1.0347,
+ "step": 8117
+ },
+ {
+ "epoch": 25.936102236421725,
+ "grad_norm": 0.0810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0451,
+ "step": 8118
+ },
+ {
+ "epoch": 25.93929712460064,
+ "grad_norm": 0.056396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0391,
+ "step": 8119
+ },
+ {
+ "epoch": 25.942492012779553,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0361,
+ "step": 8120
+ },
+ {
+ "epoch": 25.945686900958467,
+ "grad_norm": 0.05810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.037,
+ "step": 8121
+ },
+ {
+ "epoch": 25.94888178913738,
+ "grad_norm": 0.051025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.041,
+ "step": 8122
+ },
+ {
+ "epoch": 25.952076677316295,
+ "grad_norm": 0.048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0429,
+ "step": 8123
+ },
+ {
+ "epoch": 25.955271565495206,
+ "grad_norm": 0.0595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0325,
+ "step": 8124
+ },
+ {
+ "epoch": 25.95846645367412,
+ "grad_norm": 0.058349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0413,
+ "step": 8125
+ },
+ {
+ "epoch": 25.961661341853034,
+ "grad_norm": 0.0556640625,
+ "learning_rate": 0.0005,
+ "loss": 1.048,
+ "step": 8126
+ },
+ {
+ "epoch": 25.96485623003195,
+ "grad_norm": 0.050048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0378,
+ "step": 8127
+ },
+ {
+ "epoch": 25.968051118210862,
+ "grad_norm": 0.0556640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0273,
+ "step": 8128
+ },
+ {
+ "epoch": 25.971246006389777,
+ "grad_norm": 0.05126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0425,
+ "step": 8129
+ },
+ {
+ "epoch": 25.97444089456869,
+ "grad_norm": 0.0537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0495,
+ "step": 8130
+ },
+ {
+ "epoch": 25.977635782747605,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.033,
+ "step": 8131
+ },
+ {
+ "epoch": 25.98083067092652,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0312,
+ "step": 8132
+ },
+ {
+ "epoch": 25.984025559105433,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0384,
+ "step": 8133
+ },
+ {
+ "epoch": 25.987220447284344,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0393,
+ "step": 8134
+ },
+ {
+ "epoch": 25.990415335463258,
+ "grad_norm": 0.05029296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0379,
+ "step": 8135
+ },
+ {
+ "epoch": 25.99361022364217,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0491,
+ "step": 8136
+ },
+ {
+ "epoch": 25.996805111821086,
+ "grad_norm": 0.049072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0438,
+ "step": 8137
+ },
+ {
+ "epoch": 26.0,
+ "grad_norm": 0.05810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0388,
+ "step": 8138
+ },
+ {
+ "epoch": 26.003194888178914,
+ "grad_norm": 0.057861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0348,
+ "step": 8139
+ },
+ {
+ "epoch": 26.00638977635783,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0428,
+ "step": 8140
+ },
+ {
+ "epoch": 26.009584664536742,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0454,
+ "step": 8141
+ },
+ {
+ "epoch": 26.012779552715656,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0387,
+ "step": 8142
+ },
+ {
+ "epoch": 26.015974440894567,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0309,
+ "step": 8143
+ },
+ {
+ "epoch": 26.01916932907348,
+ "grad_norm": 0.0888671875,
+ "learning_rate": 0.0005,
+ "loss": 1.045,
+ "step": 8144
+ },
+ {
+ "epoch": 26.022364217252395,
+ "grad_norm": 0.056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0386,
+ "step": 8145
+ },
+ {
+ "epoch": 26.02555910543131,
+ "grad_norm": 0.060302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0401,
+ "step": 8146
+ },
+ {
+ "epoch": 26.028753993610223,
+ "grad_norm": 0.050048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0397,
+ "step": 8147
+ },
+ {
+ "epoch": 26.031948881789138,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0434,
+ "step": 8148
+ },
+ {
+ "epoch": 26.03514376996805,
+ "grad_norm": 0.0615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0439,
+ "step": 8149
+ },
+ {
+ "epoch": 26.038338658146966,
+ "grad_norm": 0.045166015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0404,
+ "step": 8150
+ },
+ {
+ "epoch": 26.04153354632588,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0506,
+ "step": 8151
+ },
+ {
+ "epoch": 26.044728434504794,
+ "grad_norm": 0.05126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0476,
+ "step": 8152
+ },
+ {
+ "epoch": 26.047923322683705,
+ "grad_norm": 0.06201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0369,
+ "step": 8153
+ },
+ {
+ "epoch": 26.05111821086262,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0423,
+ "step": 8154
+ },
+ {
+ "epoch": 26.054313099041533,
+ "grad_norm": 0.05126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0628,
+ "step": 8155
+ },
+ {
+ "epoch": 26.057507987220447,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0466,
+ "step": 8156
+ },
+ {
+ "epoch": 26.06070287539936,
+ "grad_norm": 0.0908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0345,
+ "step": 8157
+ },
+ {
+ "epoch": 26.063897763578275,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0465,
+ "step": 8158
+ },
+ {
+ "epoch": 26.06709265175719,
+ "grad_norm": 0.05224609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0477,
+ "step": 8159
+ },
+ {
+ "epoch": 26.070287539936103,
+ "grad_norm": 0.05615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0484,
+ "step": 8160
+ },
+ {
+ "epoch": 26.073482428115017,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0408,
+ "step": 8161
+ },
+ {
+ "epoch": 26.076677316293928,
+ "grad_norm": 0.061767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.035,
+ "step": 8162
+ },
+ {
+ "epoch": 26.079872204472842,
+ "grad_norm": 0.058837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0398,
+ "step": 8163
+ },
+ {
+ "epoch": 26.083067092651756,
+ "grad_norm": 0.08447265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0514,
+ "step": 8164
+ },
+ {
+ "epoch": 26.08626198083067,
+ "grad_norm": 0.0595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0437,
+ "step": 8165
+ },
+ {
+ "epoch": 26.089456869009584,
+ "grad_norm": 0.123046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0343,
+ "step": 8166
+ },
+ {
+ "epoch": 26.0926517571885,
+ "grad_norm": 0.050537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0389,
+ "step": 8167
+ },
+ {
+ "epoch": 26.095846645367413,
+ "grad_norm": 0.061767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0326,
+ "step": 8168
+ },
+ {
+ "epoch": 26.099041533546327,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0399,
+ "step": 8169
+ },
+ {
+ "epoch": 26.10223642172524,
+ "grad_norm": 0.05859375,
+ "learning_rate": 0.0005,
+ "loss": 1.039,
+ "step": 8170
+ },
+ {
+ "epoch": 26.105431309904155,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0269,
+ "step": 8171
+ },
+ {
+ "epoch": 26.108626198083066,
+ "grad_norm": 0.053466796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0535,
+ "step": 8172
+ },
+ {
+ "epoch": 26.11182108626198,
+ "grad_norm": 0.056396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0403,
+ "step": 8173
+ },
+ {
+ "epoch": 26.115015974440894,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0448,
+ "step": 8174
+ },
+ {
+ "epoch": 26.118210862619808,
+ "grad_norm": 0.055908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0415,
+ "step": 8175
+ },
+ {
+ "epoch": 26.121405750798722,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0264,
+ "step": 8176
+ },
+ {
+ "epoch": 26.124600638977636,
+ "grad_norm": 0.05908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.047,
+ "step": 8177
+ },
+ {
+ "epoch": 26.12779552715655,
+ "grad_norm": 0.0537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.046,
+ "step": 8178
+ },
+ {
+ "epoch": 26.130990415335464,
+ "grad_norm": 0.051025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0371,
+ "step": 8179
+ },
+ {
+ "epoch": 26.13418530351438,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0373,
+ "step": 8180
+ },
+ {
+ "epoch": 26.13738019169329,
+ "grad_norm": 0.056884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0457,
+ "step": 8181
+ },
+ {
+ "epoch": 26.140575079872203,
+ "grad_norm": 0.05322265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0327,
+ "step": 8182
+ },
+ {
+ "epoch": 26.143769968051117,
+ "grad_norm": 0.08447265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0338,
+ "step": 8183
+ },
+ {
+ "epoch": 26.14696485623003,
+ "grad_norm": 0.048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0326,
+ "step": 8184
+ },
+ {
+ "epoch": 26.150159744408946,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0482,
+ "step": 8185
+ },
+ {
+ "epoch": 26.15335463258786,
+ "grad_norm": 0.107421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0386,
+ "step": 8186
+ },
+ {
+ "epoch": 26.156549520766774,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0381,
+ "step": 8187
+ },
+ {
+ "epoch": 26.159744408945688,
+ "grad_norm": 0.1357421875,
+ "learning_rate": 0.0005,
+ "loss": 1.038,
+ "step": 8188
+ },
+ {
+ "epoch": 26.162939297124602,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0343,
+ "step": 8189
+ },
+ {
+ "epoch": 26.166134185303516,
+ "grad_norm": 0.1259765625,
+ "learning_rate": 0.0005,
+ "loss": 1.032,
+ "step": 8190
+ },
+ {
+ "epoch": 26.169329073482427,
+ "grad_norm": 0.08642578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0363,
+ "step": 8191
+ },
+ {
+ "epoch": 26.17252396166134,
+ "grad_norm": 0.0947265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0366,
+ "step": 8192
+ },
+ {
+ "epoch": 26.175718849840255,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.05,
+ "step": 8193
+ },
+ {
+ "epoch": 26.17891373801917,
+ "grad_norm": 0.091796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0402,
+ "step": 8194
+ },
+ {
+ "epoch": 26.182108626198083,
+ "grad_norm": 0.08251953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0415,
+ "step": 8195
+ },
+ {
+ "epoch": 26.185303514376997,
+ "grad_norm": 0.1083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0347,
+ "step": 8196
+ },
+ {
+ "epoch": 26.18849840255591,
+ "grad_norm": 0.050048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0365,
+ "step": 8197
+ },
+ {
+ "epoch": 26.191693290734825,
+ "grad_norm": 0.09033203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0453,
+ "step": 8198
+ },
+ {
+ "epoch": 26.19488817891374,
+ "grad_norm": 0.05517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0446,
+ "step": 8199
+ },
+ {
+ "epoch": 26.19808306709265,
+ "grad_norm": 0.057373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0339,
+ "step": 8200
+ },
+ {
+ "epoch": 26.201277955271564,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0466,
+ "step": 8201
+ },
+ {
+ "epoch": 26.20447284345048,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0436,
+ "step": 8202
+ },
+ {
+ "epoch": 26.207667731629392,
+ "grad_norm": 0.05859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0282,
+ "step": 8203
+ },
+ {
+ "epoch": 26.210862619808307,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0431,
+ "step": 8204
+ },
+ {
+ "epoch": 26.21405750798722,
+ "grad_norm": 0.053955078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0492,
+ "step": 8205
+ },
+ {
+ "epoch": 26.217252396166135,
+ "grad_norm": 0.06201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0465,
+ "step": 8206
+ },
+ {
+ "epoch": 26.22044728434505,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0374,
+ "step": 8207
+ },
+ {
+ "epoch": 26.223642172523963,
+ "grad_norm": 0.060302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0301,
+ "step": 8208
+ },
+ {
+ "epoch": 26.226837060702877,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0332,
+ "step": 8209
+ },
+ {
+ "epoch": 26.230031948881788,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0315,
+ "step": 8210
+ },
+ {
+ "epoch": 26.233226837060702,
+ "grad_norm": 0.057861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0491,
+ "step": 8211
+ },
+ {
+ "epoch": 26.236421725239616,
+ "grad_norm": 0.0615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0461,
+ "step": 8212
+ },
+ {
+ "epoch": 26.23961661341853,
+ "grad_norm": 0.055908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0305,
+ "step": 8213
+ },
+ {
+ "epoch": 26.242811501597444,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0423,
+ "step": 8214
+ },
+ {
+ "epoch": 26.24600638977636,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0448,
+ "step": 8215
+ },
+ {
+ "epoch": 26.249201277955272,
+ "grad_norm": 0.0810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0371,
+ "step": 8216
+ },
+ {
+ "epoch": 26.252396166134186,
+ "grad_norm": 0.05419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0352,
+ "step": 8217
+ },
+ {
+ "epoch": 26.2555910543131,
+ "grad_norm": 0.050537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0422,
+ "step": 8218
+ },
+ {
+ "epoch": 26.25878594249201,
+ "grad_norm": 0.053466796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0387,
+ "step": 8219
+ },
+ {
+ "epoch": 26.261980830670925,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0324,
+ "step": 8220
+ },
+ {
+ "epoch": 26.26517571884984,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0237,
+ "step": 8221
+ },
+ {
+ "epoch": 26.268370607028753,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0382,
+ "step": 8222
+ },
+ {
+ "epoch": 26.271565495207668,
+ "grad_norm": 0.08349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0443,
+ "step": 8223
+ },
+ {
+ "epoch": 26.27476038338658,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0441,
+ "step": 8224
+ },
+ {
+ "epoch": 26.277955271565496,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0373,
+ "step": 8225
+ },
+ {
+ "epoch": 26.28115015974441,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.041,
+ "step": 8226
+ },
+ {
+ "epoch": 26.284345047923324,
+ "grad_norm": 0.0810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.052,
+ "step": 8227
+ },
+ {
+ "epoch": 26.287539936102238,
+ "grad_norm": 0.1015625,
+ "learning_rate": 0.0005,
+ "loss": 1.041,
+ "step": 8228
+ },
+ {
+ "epoch": 26.29073482428115,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0432,
+ "step": 8229
+ },
+ {
+ "epoch": 26.293929712460063,
+ "grad_norm": 0.11572265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0406,
+ "step": 8230
+ },
+ {
+ "epoch": 26.297124600638977,
+ "grad_norm": 0.05322265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0305,
+ "step": 8231
+ },
+ {
+ "epoch": 26.30031948881789,
+ "grad_norm": 0.09912109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0559,
+ "step": 8232
+ },
+ {
+ "epoch": 26.303514376996805,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0305,
+ "step": 8233
+ },
+ {
+ "epoch": 26.30670926517572,
+ "grad_norm": 0.10009765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0295,
+ "step": 8234
+ },
+ {
+ "epoch": 26.309904153354633,
+ "grad_norm": 0.08984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0567,
+ "step": 8235
+ },
+ {
+ "epoch": 26.313099041533548,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.037,
+ "step": 8236
+ },
+ {
+ "epoch": 26.31629392971246,
+ "grad_norm": 0.08837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0439,
+ "step": 8237
+ },
+ {
+ "epoch": 26.319488817891372,
+ "grad_norm": 0.053466796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0446,
+ "step": 8238
+ },
+ {
+ "epoch": 26.322683706070286,
+ "grad_norm": 0.11328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0424,
+ "step": 8239
+ },
+ {
+ "epoch": 26.3258785942492,
+ "grad_norm": 0.0517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0534,
+ "step": 8240
+ },
+ {
+ "epoch": 26.329073482428115,
+ "grad_norm": 0.1142578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0377,
+ "step": 8241
+ },
+ {
+ "epoch": 26.33226837060703,
+ "grad_norm": 0.109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0338,
+ "step": 8242
+ },
+ {
+ "epoch": 26.335463258785943,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0457,
+ "step": 8243
+ },
+ {
+ "epoch": 26.338658146964857,
+ "grad_norm": 0.05517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0339,
+ "step": 8244
+ },
+ {
+ "epoch": 26.34185303514377,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0363,
+ "step": 8245
+ },
+ {
+ "epoch": 26.345047923322685,
+ "grad_norm": 0.061767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0399,
+ "step": 8246
+ },
+ {
+ "epoch": 26.3482428115016,
+ "grad_norm": 0.04833984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0424,
+ "step": 8247
+ },
+ {
+ "epoch": 26.35143769968051,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.04,
+ "step": 8248
+ },
+ {
+ "epoch": 26.354632587859424,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0291,
+ "step": 8249
+ },
+ {
+ "epoch": 26.357827476038338,
+ "grad_norm": 0.123046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0431,
+ "step": 8250
+ },
+ {
+ "epoch": 26.361022364217252,
+ "grad_norm": 0.052490234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0406,
+ "step": 8251
+ },
+ {
+ "epoch": 26.364217252396166,
+ "grad_norm": 0.12109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0277,
+ "step": 8252
+ },
+ {
+ "epoch": 26.36741214057508,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0439,
+ "step": 8253
+ },
+ {
+ "epoch": 26.370607028753994,
+ "grad_norm": 0.056884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0373,
+ "step": 8254
+ },
+ {
+ "epoch": 26.37380191693291,
+ "grad_norm": 0.0537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.042,
+ "step": 8255
+ },
+ {
+ "epoch": 26.376996805111823,
+ "grad_norm": 0.049560546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0505,
+ "step": 8256
+ },
+ {
+ "epoch": 26.380191693290733,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0467,
+ "step": 8257
+ },
+ {
+ "epoch": 26.383386581469647,
+ "grad_norm": 0.05712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0347,
+ "step": 8258
+ },
+ {
+ "epoch": 26.38658146964856,
+ "grad_norm": 0.0546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0437,
+ "step": 8259
+ },
+ {
+ "epoch": 26.389776357827476,
+ "grad_norm": 0.0556640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0427,
+ "step": 8260
+ },
+ {
+ "epoch": 26.39297124600639,
+ "grad_norm": 0.047119140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0411,
+ "step": 8261
+ },
+ {
+ "epoch": 26.396166134185304,
+ "grad_norm": 0.051025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0308,
+ "step": 8262
+ },
+ {
+ "epoch": 26.399361022364218,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0347,
+ "step": 8263
+ },
+ {
+ "epoch": 26.402555910543132,
+ "grad_norm": 0.11376953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0467,
+ "step": 8264
+ },
+ {
+ "epoch": 26.405750798722046,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0362,
+ "step": 8265
+ },
+ {
+ "epoch": 26.408945686900957,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0382,
+ "step": 8266
+ },
+ {
+ "epoch": 26.41214057507987,
+ "grad_norm": 0.130859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0406,
+ "step": 8267
+ },
+ {
+ "epoch": 26.415335463258785,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0429,
+ "step": 8268
+ },
+ {
+ "epoch": 26.4185303514377,
+ "grad_norm": 0.09912109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0306,
+ "step": 8269
+ },
+ {
+ "epoch": 26.421725239616613,
+ "grad_norm": 0.059814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0397,
+ "step": 8270
+ },
+ {
+ "epoch": 26.424920127795527,
+ "grad_norm": 0.095703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0469,
+ "step": 8271
+ },
+ {
+ "epoch": 26.42811501597444,
+ "grad_norm": 0.09228515625,
+ "learning_rate": 0.0005,
+ "loss": 1.038,
+ "step": 8272
+ },
+ {
+ "epoch": 26.431309904153355,
+ "grad_norm": 0.1103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0365,
+ "step": 8273
+ },
+ {
+ "epoch": 26.43450479233227,
+ "grad_norm": 0.142578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0499,
+ "step": 8274
+ },
+ {
+ "epoch": 26.437699680511184,
+ "grad_norm": 0.1435546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0344,
+ "step": 8275
+ },
+ {
+ "epoch": 26.440894568690094,
+ "grad_norm": 0.2353515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0324,
+ "step": 8276
+ },
+ {
+ "epoch": 26.44408945686901,
+ "grad_norm": 0.1298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0421,
+ "step": 8277
+ },
+ {
+ "epoch": 26.447284345047922,
+ "grad_norm": 0.08984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0453,
+ "step": 8278
+ },
+ {
+ "epoch": 26.450479233226837,
+ "grad_norm": 0.1982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0375,
+ "step": 8279
+ },
+ {
+ "epoch": 26.45367412140575,
+ "grad_norm": 0.1689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0376,
+ "step": 8280
+ },
+ {
+ "epoch": 26.456869009584665,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0388,
+ "step": 8281
+ },
+ {
+ "epoch": 26.46006389776358,
+ "grad_norm": 0.12353515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0347,
+ "step": 8282
+ },
+ {
+ "epoch": 26.463258785942493,
+ "grad_norm": 0.08935546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0428,
+ "step": 8283
+ },
+ {
+ "epoch": 26.466453674121407,
+ "grad_norm": 0.1875,
+ "learning_rate": 0.0005,
+ "loss": 1.0398,
+ "step": 8284
+ },
+ {
+ "epoch": 26.46964856230032,
+ "grad_norm": 0.05908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.039,
+ "step": 8285
+ },
+ {
+ "epoch": 26.472843450479232,
+ "grad_norm": 0.11376953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0323,
+ "step": 8286
+ },
+ {
+ "epoch": 26.476038338658146,
+ "grad_norm": 0.059814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0394,
+ "step": 8287
+ },
+ {
+ "epoch": 26.47923322683706,
+ "grad_norm": 0.10009765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0439,
+ "step": 8288
+ },
+ {
+ "epoch": 26.482428115015974,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0377,
+ "step": 8289
+ },
+ {
+ "epoch": 26.48562300319489,
+ "grad_norm": 0.08203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0522,
+ "step": 8290
+ },
+ {
+ "epoch": 26.488817891373802,
+ "grad_norm": 0.09423828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0401,
+ "step": 8291
+ },
+ {
+ "epoch": 26.492012779552716,
+ "grad_norm": 0.053955078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0425,
+ "step": 8292
+ },
+ {
+ "epoch": 26.49520766773163,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0301,
+ "step": 8293
+ },
+ {
+ "epoch": 26.498402555910545,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0403,
+ "step": 8294
+ },
+ {
+ "epoch": 26.501597444089455,
+ "grad_norm": 0.115234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0412,
+ "step": 8295
+ },
+ {
+ "epoch": 26.50479233226837,
+ "grad_norm": 0.061767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0474,
+ "step": 8296
+ },
+ {
+ "epoch": 26.507987220447284,
+ "grad_norm": 0.146484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0414,
+ "step": 8297
+ },
+ {
+ "epoch": 26.511182108626198,
+ "grad_norm": 0.1904296875,
+ "learning_rate": 0.0005,
+ "loss": 1.037,
+ "step": 8298
+ },
+ {
+ "epoch": 26.51437699680511,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0375,
+ "step": 8299
+ },
+ {
+ "epoch": 26.517571884984026,
+ "grad_norm": 0.1982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0345,
+ "step": 8300
+ },
+ {
+ "epoch": 26.52076677316294,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0562,
+ "step": 8301
+ },
+ {
+ "epoch": 26.523961661341854,
+ "grad_norm": 0.1494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0497,
+ "step": 8302
+ },
+ {
+ "epoch": 26.527156549520768,
+ "grad_norm": 0.07861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.053,
+ "step": 8303
+ },
+ {
+ "epoch": 26.53035143769968,
+ "grad_norm": 0.091796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0412,
+ "step": 8304
+ },
+ {
+ "epoch": 26.533546325878593,
+ "grad_norm": 0.08837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0486,
+ "step": 8305
+ },
+ {
+ "epoch": 26.536741214057507,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0344,
+ "step": 8306
+ },
+ {
+ "epoch": 26.53993610223642,
+ "grad_norm": 0.0478515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0423,
+ "step": 8307
+ },
+ {
+ "epoch": 26.543130990415335,
+ "grad_norm": 0.0546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0515,
+ "step": 8308
+ },
+ {
+ "epoch": 26.54632587859425,
+ "grad_norm": 0.08349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0459,
+ "step": 8309
+ },
+ {
+ "epoch": 26.549520766773163,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0362,
+ "step": 8310
+ },
+ {
+ "epoch": 26.552715654952078,
+ "grad_norm": 0.126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0418,
+ "step": 8311
+ },
+ {
+ "epoch": 26.55591054313099,
+ "grad_norm": 0.060791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0476,
+ "step": 8312
+ },
+ {
+ "epoch": 26.559105431309906,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0408,
+ "step": 8313
+ },
+ {
+ "epoch": 26.562300319488816,
+ "grad_norm": 0.056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0367,
+ "step": 8314
+ },
+ {
+ "epoch": 26.56549520766773,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0454,
+ "step": 8315
+ },
+ {
+ "epoch": 26.568690095846645,
+ "grad_norm": 0.0615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0439,
+ "step": 8316
+ },
+ {
+ "epoch": 26.57188498402556,
+ "grad_norm": 0.05908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.046,
+ "step": 8317
+ },
+ {
+ "epoch": 26.575079872204473,
+ "grad_norm": 0.09521484375,
+ "learning_rate": 0.0005,
+ "loss": 1.044,
+ "step": 8318
+ },
+ {
+ "epoch": 26.578274760383387,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0281,
+ "step": 8319
+ },
+ {
+ "epoch": 26.5814696485623,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0408,
+ "step": 8320
+ },
+ {
+ "epoch": 26.584664536741215,
+ "grad_norm": 0.0830078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0351,
+ "step": 8321
+ },
+ {
+ "epoch": 26.58785942492013,
+ "grad_norm": 0.05126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0435,
+ "step": 8322
+ },
+ {
+ "epoch": 26.591054313099043,
+ "grad_norm": 0.09423828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0386,
+ "step": 8323
+ },
+ {
+ "epoch": 26.594249201277954,
+ "grad_norm": 0.061767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0388,
+ "step": 8324
+ },
+ {
+ "epoch": 26.597444089456868,
+ "grad_norm": 0.05419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0379,
+ "step": 8325
+ },
+ {
+ "epoch": 26.600638977635782,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0325,
+ "step": 8326
+ },
+ {
+ "epoch": 26.603833865814696,
+ "grad_norm": 0.0810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0265,
+ "step": 8327
+ },
+ {
+ "epoch": 26.60702875399361,
+ "grad_norm": 0.056884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0376,
+ "step": 8328
+ },
+ {
+ "epoch": 26.610223642172524,
+ "grad_norm": 0.053955078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0495,
+ "step": 8329
+ },
+ {
+ "epoch": 26.61341853035144,
+ "grad_norm": 0.051025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0451,
+ "step": 8330
+ },
+ {
+ "epoch": 26.616613418530353,
+ "grad_norm": 0.052490234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0486,
+ "step": 8331
+ },
+ {
+ "epoch": 26.619808306709267,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0382,
+ "step": 8332
+ },
+ {
+ "epoch": 26.623003194888177,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0438,
+ "step": 8333
+ },
+ {
+ "epoch": 26.62619808306709,
+ "grad_norm": 0.059814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0321,
+ "step": 8334
+ },
+ {
+ "epoch": 26.629392971246006,
+ "grad_norm": 0.054931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0344,
+ "step": 8335
+ },
+ {
+ "epoch": 26.63258785942492,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.039,
+ "step": 8336
+ },
+ {
+ "epoch": 26.635782747603834,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0432,
+ "step": 8337
+ },
+ {
+ "epoch": 26.638977635782748,
+ "grad_norm": 0.05810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.041,
+ "step": 8338
+ },
+ {
+ "epoch": 26.642172523961662,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0489,
+ "step": 8339
+ },
+ {
+ "epoch": 26.645367412140576,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0361,
+ "step": 8340
+ },
+ {
+ "epoch": 26.64856230031949,
+ "grad_norm": 0.054931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0342,
+ "step": 8341
+ },
+ {
+ "epoch": 26.6517571884984,
+ "grad_norm": 0.08447265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0509,
+ "step": 8342
+ },
+ {
+ "epoch": 26.654952076677315,
+ "grad_norm": 0.0556640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0477,
+ "step": 8343
+ },
+ {
+ "epoch": 26.65814696485623,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0366,
+ "step": 8344
+ },
+ {
+ "epoch": 26.661341853035143,
+ "grad_norm": 0.061279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0505,
+ "step": 8345
+ },
+ {
+ "epoch": 26.664536741214057,
+ "grad_norm": 0.057861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0422,
+ "step": 8346
+ },
+ {
+ "epoch": 26.66773162939297,
+ "grad_norm": 0.05029296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0351,
+ "step": 8347
+ },
+ {
+ "epoch": 26.670926517571885,
+ "grad_norm": 0.080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0406,
+ "step": 8348
+ },
+ {
+ "epoch": 26.6741214057508,
+ "grad_norm": 0.0556640625,
+ "learning_rate": 0.0005,
+ "loss": 1.032,
+ "step": 8349
+ },
+ {
+ "epoch": 26.677316293929714,
+ "grad_norm": 0.11474609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0491,
+ "step": 8350
+ },
+ {
+ "epoch": 26.680511182108628,
+ "grad_norm": 0.09814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0377,
+ "step": 8351
+ },
+ {
+ "epoch": 26.68370607028754,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0371,
+ "step": 8352
+ },
+ {
+ "epoch": 26.686900958466452,
+ "grad_norm": 0.10009765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0481,
+ "step": 8353
+ },
+ {
+ "epoch": 26.690095846645367,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0453,
+ "step": 8354
+ },
+ {
+ "epoch": 26.69329073482428,
+ "grad_norm": 0.1201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0364,
+ "step": 8355
+ },
+ {
+ "epoch": 26.696485623003195,
+ "grad_norm": 0.052001953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0542,
+ "step": 8356
+ },
+ {
+ "epoch": 26.69968051118211,
+ "grad_norm": 0.126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0358,
+ "step": 8357
+ },
+ {
+ "epoch": 26.702875399361023,
+ "grad_norm": 0.051513671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0457,
+ "step": 8358
+ },
+ {
+ "epoch": 26.706070287539937,
+ "grad_norm": 0.12890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0278,
+ "step": 8359
+ },
+ {
+ "epoch": 26.70926517571885,
+ "grad_norm": 0.053955078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0435,
+ "step": 8360
+ },
+ {
+ "epoch": 26.712460063897765,
+ "grad_norm": 0.09619140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0525,
+ "step": 8361
+ },
+ {
+ "epoch": 26.715654952076676,
+ "grad_norm": 0.04833984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0441,
+ "step": 8362
+ },
+ {
+ "epoch": 26.71884984025559,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0378,
+ "step": 8363
+ },
+ {
+ "epoch": 26.722044728434504,
+ "grad_norm": 0.059814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0259,
+ "step": 8364
+ },
+ {
+ "epoch": 26.72523961661342,
+ "grad_norm": 0.10693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0346,
+ "step": 8365
+ },
+ {
+ "epoch": 26.728434504792332,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0333,
+ "step": 8366
+ },
+ {
+ "epoch": 26.731629392971247,
+ "grad_norm": 0.1025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.038,
+ "step": 8367
+ },
+ {
+ "epoch": 26.73482428115016,
+ "grad_norm": 0.0849609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0398,
+ "step": 8368
+ },
+ {
+ "epoch": 26.738019169329075,
+ "grad_norm": 0.1640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0266,
+ "step": 8369
+ },
+ {
+ "epoch": 26.74121405750799,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0449,
+ "step": 8370
+ },
+ {
+ "epoch": 26.7444089456869,
+ "grad_norm": 0.236328125,
+ "learning_rate": 0.0005,
+ "loss": 1.041,
+ "step": 8371
+ },
+ {
+ "epoch": 26.747603833865814,
+ "grad_norm": 0.1474609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0388,
+ "step": 8372
+ },
+ {
+ "epoch": 26.750798722044728,
+ "grad_norm": 0.15234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0545,
+ "step": 8373
+ },
+ {
+ "epoch": 26.75399361022364,
+ "grad_norm": 0.330078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0343,
+ "step": 8374
+ },
+ {
+ "epoch": 26.757188498402556,
+ "grad_norm": 0.259765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0411,
+ "step": 8375
+ },
+ {
+ "epoch": 26.76038338658147,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0357,
+ "step": 8376
+ },
+ {
+ "epoch": 26.763578274760384,
+ "grad_norm": 0.345703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0382,
+ "step": 8377
+ },
+ {
+ "epoch": 26.766773162939298,
+ "grad_norm": 0.39453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0407,
+ "step": 8378
+ },
+ {
+ "epoch": 26.769968051118212,
+ "grad_norm": 0.2119140625,
+ "learning_rate": 0.0005,
+ "loss": 1.034,
+ "step": 8379
+ },
+ {
+ "epoch": 26.773162939297123,
+ "grad_norm": 0.1923828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0255,
+ "step": 8380
+ },
+ {
+ "epoch": 26.776357827476037,
+ "grad_norm": 0.447265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0311,
+ "step": 8381
+ },
+ {
+ "epoch": 26.77955271565495,
+ "grad_norm": 0.412109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0428,
+ "step": 8382
+ },
+ {
+ "epoch": 26.782747603833865,
+ "grad_norm": 0.087890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0423,
+ "step": 8383
+ },
+ {
+ "epoch": 26.78594249201278,
+ "grad_norm": 0.28515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0421,
+ "step": 8384
+ },
+ {
+ "epoch": 26.789137380191693,
+ "grad_norm": 0.208984375,
+ "learning_rate": 0.0005,
+ "loss": 1.035,
+ "step": 8385
+ },
+ {
+ "epoch": 26.792332268370608,
+ "grad_norm": 0.111328125,
+ "learning_rate": 0.0005,
+ "loss": 1.05,
+ "step": 8386
+ },
+ {
+ "epoch": 26.79552715654952,
+ "grad_norm": 0.30859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0543,
+ "step": 8387
+ },
+ {
+ "epoch": 26.798722044728436,
+ "grad_norm": 0.2734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0192,
+ "step": 8388
+ },
+ {
+ "epoch": 26.80191693290735,
+ "grad_norm": 0.1494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0392,
+ "step": 8389
+ },
+ {
+ "epoch": 26.80511182108626,
+ "grad_norm": 0.53515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0427,
+ "step": 8390
+ },
+ {
+ "epoch": 26.808306709265175,
+ "grad_norm": 0.72265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0508,
+ "step": 8391
+ },
+ {
+ "epoch": 26.81150159744409,
+ "grad_norm": 0.921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0441,
+ "step": 8392
+ },
+ {
+ "epoch": 26.814696485623003,
+ "grad_norm": 1.6171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0719,
+ "step": 8393
+ },
+ {
+ "epoch": 26.817891373801917,
+ "grad_norm": 0.609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0386,
+ "step": 8394
+ },
+ {
+ "epoch": 26.82108626198083,
+ "grad_norm": 0.298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0388,
+ "step": 8395
+ },
+ {
+ "epoch": 26.824281150159745,
+ "grad_norm": 1.0859375,
+ "learning_rate": 0.0005,
+ "loss": 1.047,
+ "step": 8396
+ },
+ {
+ "epoch": 26.82747603833866,
+ "grad_norm": 1.375,
+ "learning_rate": 0.0005,
+ "loss": 1.0445,
+ "step": 8397
+ },
+ {
+ "epoch": 26.830670926517573,
+ "grad_norm": 0.1552734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0481,
+ "step": 8398
+ },
+ {
+ "epoch": 26.833865814696484,
+ "grad_norm": 1.5234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0562,
+ "step": 8399
+ },
+ {
+ "epoch": 26.837060702875398,
+ "grad_norm": 0.328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0292,
+ "step": 8400
+ },
+ {
+ "epoch": 26.840255591054312,
+ "grad_norm": 0.66015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0556,
+ "step": 8401
+ },
+ {
+ "epoch": 26.843450479233226,
+ "grad_norm": 1.0234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0579,
+ "step": 8402
+ },
+ {
+ "epoch": 26.84664536741214,
+ "grad_norm": 0.7578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0533,
+ "step": 8403
+ },
+ {
+ "epoch": 26.849840255591054,
+ "grad_norm": 0.1640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0458,
+ "step": 8404
+ },
+ {
+ "epoch": 26.85303514376997,
+ "grad_norm": 0.86328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0352,
+ "step": 8405
+ },
+ {
+ "epoch": 26.856230031948883,
+ "grad_norm": 0.76953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0378,
+ "step": 8406
+ },
+ {
+ "epoch": 26.859424920127797,
+ "grad_norm": 0.185546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0442,
+ "step": 8407
+ },
+ {
+ "epoch": 26.86261980830671,
+ "grad_norm": 0.73046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0391,
+ "step": 8408
+ },
+ {
+ "epoch": 26.86581469648562,
+ "grad_norm": 0.3671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0327,
+ "step": 8409
+ },
+ {
+ "epoch": 26.869009584664536,
+ "grad_norm": 0.369140625,
+ "learning_rate": 0.0005,
+ "loss": 1.035,
+ "step": 8410
+ },
+ {
+ "epoch": 26.87220447284345,
+ "grad_norm": 0.46875,
+ "learning_rate": 0.0005,
+ "loss": 1.0307,
+ "step": 8411
+ },
+ {
+ "epoch": 26.875399361022364,
+ "grad_norm": 0.2421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0524,
+ "step": 8412
+ },
+ {
+ "epoch": 26.878594249201278,
+ "grad_norm": 0.61328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0458,
+ "step": 8413
+ },
+ {
+ "epoch": 26.881789137380192,
+ "grad_norm": 0.119140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0409,
+ "step": 8414
+ },
+ {
+ "epoch": 26.884984025559106,
+ "grad_norm": 0.4296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0386,
+ "step": 8415
+ },
+ {
+ "epoch": 26.88817891373802,
+ "grad_norm": 0.1533203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0446,
+ "step": 8416
+ },
+ {
+ "epoch": 26.891373801916934,
+ "grad_norm": 0.302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0432,
+ "step": 8417
+ },
+ {
+ "epoch": 26.894568690095845,
+ "grad_norm": 0.1484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0459,
+ "step": 8418
+ },
+ {
+ "epoch": 26.89776357827476,
+ "grad_norm": 0.1767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0501,
+ "step": 8419
+ },
+ {
+ "epoch": 26.900958466453673,
+ "grad_norm": 0.1279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0329,
+ "step": 8420
+ },
+ {
+ "epoch": 26.904153354632587,
+ "grad_norm": 0.1484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0421,
+ "step": 8421
+ },
+ {
+ "epoch": 26.9073482428115,
+ "grad_norm": 0.10302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0499,
+ "step": 8422
+ },
+ {
+ "epoch": 26.910543130990416,
+ "grad_norm": 0.1298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0351,
+ "step": 8423
+ },
+ {
+ "epoch": 26.91373801916933,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.041,
+ "step": 8424
+ },
+ {
+ "epoch": 26.916932907348244,
+ "grad_norm": 0.10009765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0408,
+ "step": 8425
+ },
+ {
+ "epoch": 26.920127795527158,
+ "grad_norm": 0.0908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0505,
+ "step": 8426
+ },
+ {
+ "epoch": 26.923322683706072,
+ "grad_norm": 0.11328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0448,
+ "step": 8427
+ },
+ {
+ "epoch": 26.926517571884983,
+ "grad_norm": 0.09814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0435,
+ "step": 8428
+ },
+ {
+ "epoch": 26.929712460063897,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0347,
+ "step": 8429
+ },
+ {
+ "epoch": 26.93290734824281,
+ "grad_norm": 0.1533203125,
+ "learning_rate": 0.0005,
+ "loss": 1.036,
+ "step": 8430
+ },
+ {
+ "epoch": 26.936102236421725,
+ "grad_norm": 0.1318359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0457,
+ "step": 8431
+ },
+ {
+ "epoch": 26.93929712460064,
+ "grad_norm": 0.1552734375,
+ "learning_rate": 0.0005,
+ "loss": 1.038,
+ "step": 8432
+ },
+ {
+ "epoch": 26.942492012779553,
+ "grad_norm": 0.1064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0369,
+ "step": 8433
+ },
+ {
+ "epoch": 26.945686900958467,
+ "grad_norm": 0.1640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0464,
+ "step": 8434
+ },
+ {
+ "epoch": 26.94888178913738,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0506,
+ "step": 8435
+ },
+ {
+ "epoch": 26.952076677316295,
+ "grad_norm": 0.140625,
+ "learning_rate": 0.0005,
+ "loss": 1.043,
+ "step": 8436
+ },
+ {
+ "epoch": 26.955271565495206,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0333,
+ "step": 8437
+ },
+ {
+ "epoch": 26.95846645367412,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0349,
+ "step": 8438
+ },
+ {
+ "epoch": 26.961661341853034,
+ "grad_norm": 0.115234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0453,
+ "step": 8439
+ },
+ {
+ "epoch": 26.96485623003195,
+ "grad_norm": 0.058349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0603,
+ "step": 8440
+ },
+ {
+ "epoch": 26.968051118210862,
+ "grad_norm": 0.087890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0361,
+ "step": 8441
+ },
+ {
+ "epoch": 26.971246006389777,
+ "grad_norm": 0.08056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0274,
+ "step": 8442
+ },
+ {
+ "epoch": 26.97444089456869,
+ "grad_norm": 0.05517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0476,
+ "step": 8443
+ },
+ {
+ "epoch": 26.977635782747605,
+ "grad_norm": 0.095703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0396,
+ "step": 8444
+ },
+ {
+ "epoch": 26.98083067092652,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0331,
+ "step": 8445
+ },
+ {
+ "epoch": 26.984025559105433,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0575,
+ "step": 8446
+ },
+ {
+ "epoch": 26.987220447284344,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0423,
+ "step": 8447
+ },
+ {
+ "epoch": 26.990415335463258,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0565,
+ "step": 8448
+ },
+ {
+ "epoch": 26.99361022364217,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0481,
+ "step": 8449
+ },
+ {
+ "epoch": 26.996805111821086,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.03,
+ "step": 8450
+ },
+ {
+ "epoch": 27.0,
+ "grad_norm": 0.083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0451,
+ "step": 8451
+ },
+ {
+ "epoch": 27.003194888178914,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.039,
+ "step": 8452
+ },
+ {
+ "epoch": 27.00638977635783,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0386,
+ "step": 8453
+ },
+ {
+ "epoch": 27.009584664536742,
+ "grad_norm": 0.06201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0346,
+ "step": 8454
+ },
+ {
+ "epoch": 27.012779552715656,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0408,
+ "step": 8455
+ },
+ {
+ "epoch": 27.015974440894567,
+ "grad_norm": 0.049560546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0392,
+ "step": 8456
+ },
+ {
+ "epoch": 27.01916932907348,
+ "grad_norm": 0.058349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0383,
+ "step": 8457
+ },
+ {
+ "epoch": 27.022364217252395,
+ "grad_norm": 0.060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0452,
+ "step": 8458
+ },
+ {
+ "epoch": 27.02555910543131,
+ "grad_norm": 0.052001953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0388,
+ "step": 8459
+ },
+ {
+ "epoch": 27.028753993610223,
+ "grad_norm": 0.054931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0425,
+ "step": 8460
+ },
+ {
+ "epoch": 27.031948881789138,
+ "grad_norm": 0.059814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0332,
+ "step": 8461
+ },
+ {
+ "epoch": 27.03514376996805,
+ "grad_norm": 0.051025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.044,
+ "step": 8462
+ },
+ {
+ "epoch": 27.038338658146966,
+ "grad_norm": 0.0576171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0425,
+ "step": 8463
+ },
+ {
+ "epoch": 27.04153354632588,
+ "grad_norm": 0.0830078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0406,
+ "step": 8464
+ },
+ {
+ "epoch": 27.044728434504794,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.05,
+ "step": 8465
+ },
+ {
+ "epoch": 27.047923322683705,
+ "grad_norm": 0.06005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.032,
+ "step": 8466
+ },
+ {
+ "epoch": 27.05111821086262,
+ "grad_norm": 0.056884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0393,
+ "step": 8467
+ },
+ {
+ "epoch": 27.054313099041533,
+ "grad_norm": 0.048583984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0264,
+ "step": 8468
+ },
+ {
+ "epoch": 27.057507987220447,
+ "grad_norm": 0.05322265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0414,
+ "step": 8469
+ },
+ {
+ "epoch": 27.06070287539936,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0423,
+ "step": 8470
+ },
+ {
+ "epoch": 27.063897763578275,
+ "grad_norm": 0.052734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0396,
+ "step": 8471
+ },
+ {
+ "epoch": 27.06709265175719,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0578,
+ "step": 8472
+ },
+ {
+ "epoch": 27.070287539936103,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0324,
+ "step": 8473
+ },
+ {
+ "epoch": 27.073482428115017,
+ "grad_norm": 0.05908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0411,
+ "step": 8474
+ },
+ {
+ "epoch": 27.076677316293928,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0379,
+ "step": 8475
+ },
+ {
+ "epoch": 27.079872204472842,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0369,
+ "step": 8476
+ },
+ {
+ "epoch": 27.083067092651756,
+ "grad_norm": 0.05126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0453,
+ "step": 8477
+ },
+ {
+ "epoch": 27.08626198083067,
+ "grad_norm": 0.0537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0491,
+ "step": 8478
+ },
+ {
+ "epoch": 27.089456869009584,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0399,
+ "step": 8479
+ },
+ {
+ "epoch": 27.0926517571885,
+ "grad_norm": 0.05126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0574,
+ "step": 8480
+ },
+ {
+ "epoch": 27.095846645367413,
+ "grad_norm": 0.04931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0492,
+ "step": 8481
+ },
+ {
+ "epoch": 27.099041533546327,
+ "grad_norm": 0.047119140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0305,
+ "step": 8482
+ },
+ {
+ "epoch": 27.10223642172524,
+ "grad_norm": 0.05126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0376,
+ "step": 8483
+ },
+ {
+ "epoch": 27.105431309904155,
+ "grad_norm": 0.05029296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0369,
+ "step": 8484
+ },
+ {
+ "epoch": 27.108626198083066,
+ "grad_norm": 0.0546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0401,
+ "step": 8485
+ },
+ {
+ "epoch": 27.11182108626198,
+ "grad_norm": 0.056396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0383,
+ "step": 8486
+ },
+ {
+ "epoch": 27.115015974440894,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0346,
+ "step": 8487
+ },
+ {
+ "epoch": 27.118210862619808,
+ "grad_norm": 0.0498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0235,
+ "step": 8488
+ },
+ {
+ "epoch": 27.121405750798722,
+ "grad_norm": 0.05322265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0508,
+ "step": 8489
+ },
+ {
+ "epoch": 27.124600638977636,
+ "grad_norm": 0.05029296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0324,
+ "step": 8490
+ },
+ {
+ "epoch": 27.12779552715655,
+ "grad_norm": 0.05810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.036,
+ "step": 8491
+ },
+ {
+ "epoch": 27.130990415335464,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0392,
+ "step": 8492
+ },
+ {
+ "epoch": 27.13418530351438,
+ "grad_norm": 0.058349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0455,
+ "step": 8493
+ },
+ {
+ "epoch": 27.13738019169329,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0398,
+ "step": 8494
+ },
+ {
+ "epoch": 27.140575079872203,
+ "grad_norm": 0.06201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.037,
+ "step": 8495
+ },
+ {
+ "epoch": 27.143769968051117,
+ "grad_norm": 0.05810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0452,
+ "step": 8496
+ },
+ {
+ "epoch": 27.14696485623003,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0373,
+ "step": 8497
+ },
+ {
+ "epoch": 27.150159744408946,
+ "grad_norm": 0.052978515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0458,
+ "step": 8498
+ },
+ {
+ "epoch": 27.15335463258786,
+ "grad_norm": 0.053466796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0415,
+ "step": 8499
+ },
+ {
+ "epoch": 27.156549520766774,
+ "grad_norm": 0.046630859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0456,
+ "step": 8500
+ },
+ {
+ "epoch": 27.159744408945688,
+ "grad_norm": 0.05615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.049,
+ "step": 8501
+ },
+ {
+ "epoch": 27.162939297124602,
+ "grad_norm": 0.050537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.042,
+ "step": 8502
+ },
+ {
+ "epoch": 27.166134185303516,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0358,
+ "step": 8503
+ },
+ {
+ "epoch": 27.169329073482427,
+ "grad_norm": 0.049560546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0349,
+ "step": 8504
+ },
+ {
+ "epoch": 27.17252396166134,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0385,
+ "step": 8505
+ },
+ {
+ "epoch": 27.175718849840255,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0364,
+ "step": 8506
+ },
+ {
+ "epoch": 27.17891373801917,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0371,
+ "step": 8507
+ },
+ {
+ "epoch": 27.182108626198083,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0453,
+ "step": 8508
+ },
+ {
+ "epoch": 27.185303514376997,
+ "grad_norm": 0.11767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0423,
+ "step": 8509
+ },
+ {
+ "epoch": 27.18849840255591,
+ "grad_norm": 0.060791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0405,
+ "step": 8510
+ },
+ {
+ "epoch": 27.191693290734825,
+ "grad_norm": 0.059326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0335,
+ "step": 8511
+ },
+ {
+ "epoch": 27.19488817891374,
+ "grad_norm": 0.095703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0394,
+ "step": 8512
+ },
+ {
+ "epoch": 27.19808306709265,
+ "grad_norm": 0.05859375,
+ "learning_rate": 0.0005,
+ "loss": 1.038,
+ "step": 8513
+ },
+ {
+ "epoch": 27.201277955271564,
+ "grad_norm": 0.05712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0511,
+ "step": 8514
+ },
+ {
+ "epoch": 27.20447284345048,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0349,
+ "step": 8515
+ },
+ {
+ "epoch": 27.207667731629392,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0429,
+ "step": 8516
+ },
+ {
+ "epoch": 27.210862619808307,
+ "grad_norm": 0.08935546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0482,
+ "step": 8517
+ },
+ {
+ "epoch": 27.21405750798722,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0395,
+ "step": 8518
+ },
+ {
+ "epoch": 27.217252396166135,
+ "grad_norm": 0.08203125,
+ "learning_rate": 0.0005,
+ "loss": 1.045,
+ "step": 8519
+ },
+ {
+ "epoch": 27.22044728434505,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0276,
+ "step": 8520
+ },
+ {
+ "epoch": 27.223642172523963,
+ "grad_norm": 0.057861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0315,
+ "step": 8521
+ },
+ {
+ "epoch": 27.226837060702877,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0332,
+ "step": 8522
+ },
+ {
+ "epoch": 27.230031948881788,
+ "grad_norm": 0.052978515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0444,
+ "step": 8523
+ },
+ {
+ "epoch": 27.233226837060702,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0464,
+ "step": 8524
+ },
+ {
+ "epoch": 27.236421725239616,
+ "grad_norm": 0.050537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0404,
+ "step": 8525
+ },
+ {
+ "epoch": 27.23961661341853,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0446,
+ "step": 8526
+ },
+ {
+ "epoch": 27.242811501597444,
+ "grad_norm": 0.044677734375,
+ "learning_rate": 0.0005,
+ "loss": 1.03,
+ "step": 8527
+ },
+ {
+ "epoch": 27.24600638977636,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0376,
+ "step": 8528
+ },
+ {
+ "epoch": 27.249201277955272,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0381,
+ "step": 8529
+ },
+ {
+ "epoch": 27.252396166134186,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0518,
+ "step": 8530
+ },
+ {
+ "epoch": 27.2555910543131,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0354,
+ "step": 8531
+ },
+ {
+ "epoch": 27.25878594249201,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0482,
+ "step": 8532
+ },
+ {
+ "epoch": 27.261980830670925,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0415,
+ "step": 8533
+ },
+ {
+ "epoch": 27.26517571884984,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0416,
+ "step": 8534
+ },
+ {
+ "epoch": 27.268370607028753,
+ "grad_norm": 0.059326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0241,
+ "step": 8535
+ },
+ {
+ "epoch": 27.271565495207668,
+ "grad_norm": 0.05517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.042,
+ "step": 8536
+ },
+ {
+ "epoch": 27.27476038338658,
+ "grad_norm": 0.058349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0495,
+ "step": 8537
+ },
+ {
+ "epoch": 27.277955271565496,
+ "grad_norm": 0.050537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0495,
+ "step": 8538
+ },
+ {
+ "epoch": 27.28115015974441,
+ "grad_norm": 0.08154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0267,
+ "step": 8539
+ },
+ {
+ "epoch": 27.284345047923324,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0414,
+ "step": 8540
+ },
+ {
+ "epoch": 27.287539936102238,
+ "grad_norm": 0.05859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0431,
+ "step": 8541
+ },
+ {
+ "epoch": 27.29073482428115,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0497,
+ "step": 8542
+ },
+ {
+ "epoch": 27.293929712460063,
+ "grad_norm": 0.057861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0352,
+ "step": 8543
+ },
+ {
+ "epoch": 27.297124600638977,
+ "grad_norm": 0.054443359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0367,
+ "step": 8544
+ },
+ {
+ "epoch": 27.30031948881789,
+ "grad_norm": 0.055908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.041,
+ "step": 8545
+ },
+ {
+ "epoch": 27.303514376996805,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0326,
+ "step": 8546
+ },
+ {
+ "epoch": 27.30670926517572,
+ "grad_norm": 0.060791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0417,
+ "step": 8547
+ },
+ {
+ "epoch": 27.309904153354633,
+ "grad_norm": 0.08349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0467,
+ "step": 8548
+ },
+ {
+ "epoch": 27.313099041533548,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0336,
+ "step": 8549
+ },
+ {
+ "epoch": 27.31629392971246,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0374,
+ "step": 8550
+ },
+ {
+ "epoch": 27.319488817891372,
+ "grad_norm": 0.1005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0455,
+ "step": 8551
+ },
+ {
+ "epoch": 27.322683706070286,
+ "grad_norm": 0.06201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0394,
+ "step": 8552
+ },
+ {
+ "epoch": 27.3258785942492,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0446,
+ "step": 8553
+ },
+ {
+ "epoch": 27.329073482428115,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0362,
+ "step": 8554
+ },
+ {
+ "epoch": 27.33226837060703,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0455,
+ "step": 8555
+ },
+ {
+ "epoch": 27.335463258785943,
+ "grad_norm": 0.05712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0428,
+ "step": 8556
+ },
+ {
+ "epoch": 27.338658146964857,
+ "grad_norm": 0.05810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0285,
+ "step": 8557
+ },
+ {
+ "epoch": 27.34185303514377,
+ "grad_norm": 0.056396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0358,
+ "step": 8558
+ },
+ {
+ "epoch": 27.345047923322685,
+ "grad_norm": 0.051025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0421,
+ "step": 8559
+ },
+ {
+ "epoch": 27.3482428115016,
+ "grad_norm": 0.05078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0363,
+ "step": 8560
+ },
+ {
+ "epoch": 27.35143769968051,
+ "grad_norm": 0.052734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0243,
+ "step": 8561
+ },
+ {
+ "epoch": 27.354632587859424,
+ "grad_norm": 0.04541015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0429,
+ "step": 8562
+ },
+ {
+ "epoch": 27.357827476038338,
+ "grad_norm": 0.05908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0286,
+ "step": 8563
+ },
+ {
+ "epoch": 27.361022364217252,
+ "grad_norm": 0.059814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0363,
+ "step": 8564
+ },
+ {
+ "epoch": 27.364217252396166,
+ "grad_norm": 0.052001953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0254,
+ "step": 8565
+ },
+ {
+ "epoch": 27.36741214057508,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0391,
+ "step": 8566
+ },
+ {
+ "epoch": 27.370607028753994,
+ "grad_norm": 0.05078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0415,
+ "step": 8567
+ },
+ {
+ "epoch": 27.37380191693291,
+ "grad_norm": 0.056396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0428,
+ "step": 8568
+ },
+ {
+ "epoch": 27.376996805111823,
+ "grad_norm": 0.049072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0375,
+ "step": 8569
+ },
+ {
+ "epoch": 27.380191693290733,
+ "grad_norm": 0.04833984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0345,
+ "step": 8570
+ },
+ {
+ "epoch": 27.383386581469647,
+ "grad_norm": 0.056884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0389,
+ "step": 8571
+ },
+ {
+ "epoch": 27.38658146964856,
+ "grad_norm": 0.08056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0322,
+ "step": 8572
+ },
+ {
+ "epoch": 27.389776357827476,
+ "grad_norm": 0.08203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0322,
+ "step": 8573
+ },
+ {
+ "epoch": 27.39297124600639,
+ "grad_norm": 0.119140625,
+ "learning_rate": 0.0005,
+ "loss": 1.038,
+ "step": 8574
+ },
+ {
+ "epoch": 27.396166134185304,
+ "grad_norm": 0.091796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0337,
+ "step": 8575
+ },
+ {
+ "epoch": 27.399361022364218,
+ "grad_norm": 0.126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0376,
+ "step": 8576
+ },
+ {
+ "epoch": 27.402555910543132,
+ "grad_norm": 0.08349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.038,
+ "step": 8577
+ },
+ {
+ "epoch": 27.405750798722046,
+ "grad_norm": 0.10693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0383,
+ "step": 8578
+ },
+ {
+ "epoch": 27.408945686900957,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.032,
+ "step": 8579
+ },
+ {
+ "epoch": 27.41214057507987,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0352,
+ "step": 8580
+ },
+ {
+ "epoch": 27.415335463258785,
+ "grad_norm": 0.055908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0467,
+ "step": 8581
+ },
+ {
+ "epoch": 27.4185303514377,
+ "grad_norm": 0.0546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0351,
+ "step": 8582
+ },
+ {
+ "epoch": 27.421725239616613,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0303,
+ "step": 8583
+ },
+ {
+ "epoch": 27.424920127795527,
+ "grad_norm": 0.05126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0518,
+ "step": 8584
+ },
+ {
+ "epoch": 27.42811501597444,
+ "grad_norm": 0.0615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0403,
+ "step": 8585
+ },
+ {
+ "epoch": 27.431309904153355,
+ "grad_norm": 0.053466796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0366,
+ "step": 8586
+ },
+ {
+ "epoch": 27.43450479233227,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0362,
+ "step": 8587
+ },
+ {
+ "epoch": 27.437699680511184,
+ "grad_norm": 0.05322265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0355,
+ "step": 8588
+ },
+ {
+ "epoch": 27.440894568690094,
+ "grad_norm": 0.056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0263,
+ "step": 8589
+ },
+ {
+ "epoch": 27.44408945686901,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.046,
+ "step": 8590
+ },
+ {
+ "epoch": 27.447284345047922,
+ "grad_norm": 0.051025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0458,
+ "step": 8591
+ },
+ {
+ "epoch": 27.450479233226837,
+ "grad_norm": 0.061279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0281,
+ "step": 8592
+ },
+ {
+ "epoch": 27.45367412140575,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.043,
+ "step": 8593
+ },
+ {
+ "epoch": 27.456869009584665,
+ "grad_norm": 0.047119140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0379,
+ "step": 8594
+ },
+ {
+ "epoch": 27.46006389776358,
+ "grad_norm": 0.059326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0233,
+ "step": 8595
+ },
+ {
+ "epoch": 27.463258785942493,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0384,
+ "step": 8596
+ },
+ {
+ "epoch": 27.466453674121407,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0394,
+ "step": 8597
+ },
+ {
+ "epoch": 27.46964856230032,
+ "grad_norm": 0.10009765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0266,
+ "step": 8598
+ },
+ {
+ "epoch": 27.472843450479232,
+ "grad_norm": 0.05224609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0315,
+ "step": 8599
+ },
+ {
+ "epoch": 27.476038338658146,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0332,
+ "step": 8600
+ },
+ {
+ "epoch": 27.47923322683706,
+ "grad_norm": 0.09033203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0482,
+ "step": 8601
+ },
+ {
+ "epoch": 27.482428115015974,
+ "grad_norm": 0.08544921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0387,
+ "step": 8602
+ },
+ {
+ "epoch": 27.48562300319489,
+ "grad_norm": 0.08544921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0444,
+ "step": 8603
+ },
+ {
+ "epoch": 27.488817891373802,
+ "grad_norm": 0.05078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0361,
+ "step": 8604
+ },
+ {
+ "epoch": 27.492012779552716,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0339,
+ "step": 8605
+ },
+ {
+ "epoch": 27.49520766773163,
+ "grad_norm": 0.0517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0333,
+ "step": 8606
+ },
+ {
+ "epoch": 27.498402555910545,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0388,
+ "step": 8607
+ },
+ {
+ "epoch": 27.501597444089455,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0264,
+ "step": 8608
+ },
+ {
+ "epoch": 27.50479233226837,
+ "grad_norm": 0.125,
+ "learning_rate": 0.0005,
+ "loss": 1.0353,
+ "step": 8609
+ },
+ {
+ "epoch": 27.507987220447284,
+ "grad_norm": 0.05810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0494,
+ "step": 8610
+ },
+ {
+ "epoch": 27.511182108626198,
+ "grad_norm": 0.060302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0371,
+ "step": 8611
+ },
+ {
+ "epoch": 27.51437699680511,
+ "grad_norm": 0.09423828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0452,
+ "step": 8612
+ },
+ {
+ "epoch": 27.517571884984026,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.036,
+ "step": 8613
+ },
+ {
+ "epoch": 27.52076677316294,
+ "grad_norm": 0.050537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0486,
+ "step": 8614
+ },
+ {
+ "epoch": 27.523961661341854,
+ "grad_norm": 0.0498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0315,
+ "step": 8615
+ },
+ {
+ "epoch": 27.527156549520768,
+ "grad_norm": 0.047119140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0399,
+ "step": 8616
+ },
+ {
+ "epoch": 27.53035143769968,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0489,
+ "step": 8617
+ },
+ {
+ "epoch": 27.533546325878593,
+ "grad_norm": 0.052490234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0469,
+ "step": 8618
+ },
+ {
+ "epoch": 27.536741214057507,
+ "grad_norm": 0.0546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0413,
+ "step": 8619
+ },
+ {
+ "epoch": 27.53993610223642,
+ "grad_norm": 0.09423828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0295,
+ "step": 8620
+ },
+ {
+ "epoch": 27.543130990415335,
+ "grad_norm": 0.057861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0447,
+ "step": 8621
+ },
+ {
+ "epoch": 27.54632587859425,
+ "grad_norm": 0.1005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0367,
+ "step": 8622
+ },
+ {
+ "epoch": 27.549520766773163,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.033,
+ "step": 8623
+ },
+ {
+ "epoch": 27.552715654952078,
+ "grad_norm": 0.051025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0359,
+ "step": 8624
+ },
+ {
+ "epoch": 27.55591054313099,
+ "grad_norm": 0.05908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0287,
+ "step": 8625
+ },
+ {
+ "epoch": 27.559105431309906,
+ "grad_norm": 0.062255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0417,
+ "step": 8626
+ },
+ {
+ "epoch": 27.562300319488816,
+ "grad_norm": 0.051513671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0356,
+ "step": 8627
+ },
+ {
+ "epoch": 27.56549520766773,
+ "grad_norm": 0.052978515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0385,
+ "step": 8628
+ },
+ {
+ "epoch": 27.568690095846645,
+ "grad_norm": 0.08203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0335,
+ "step": 8629
+ },
+ {
+ "epoch": 27.57188498402556,
+ "grad_norm": 0.050537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0263,
+ "step": 8630
+ },
+ {
+ "epoch": 27.575079872204473,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0449,
+ "step": 8631
+ },
+ {
+ "epoch": 27.578274760383387,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.027,
+ "step": 8632
+ },
+ {
+ "epoch": 27.5814696485623,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0374,
+ "step": 8633
+ },
+ {
+ "epoch": 27.584664536741215,
+ "grad_norm": 0.054443359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0364,
+ "step": 8634
+ },
+ {
+ "epoch": 27.58785942492013,
+ "grad_norm": 0.119140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0359,
+ "step": 8635
+ },
+ {
+ "epoch": 27.591054313099043,
+ "grad_norm": 0.05810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0382,
+ "step": 8636
+ },
+ {
+ "epoch": 27.594249201277954,
+ "grad_norm": 0.08251953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0401,
+ "step": 8637
+ },
+ {
+ "epoch": 27.597444089456868,
+ "grad_norm": 0.08349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0402,
+ "step": 8638
+ },
+ {
+ "epoch": 27.600638977635782,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0226,
+ "step": 8639
+ },
+ {
+ "epoch": 27.603833865814696,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0413,
+ "step": 8640
+ },
+ {
+ "epoch": 27.60702875399361,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0284,
+ "step": 8641
+ },
+ {
+ "epoch": 27.610223642172524,
+ "grad_norm": 0.060302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0375,
+ "step": 8642
+ },
+ {
+ "epoch": 27.61341853035144,
+ "grad_norm": 0.056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0414,
+ "step": 8643
+ },
+ {
+ "epoch": 27.616613418530353,
+ "grad_norm": 0.05126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0179,
+ "step": 8644
+ },
+ {
+ "epoch": 27.619808306709267,
+ "grad_norm": 0.08251953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0455,
+ "step": 8645
+ },
+ {
+ "epoch": 27.623003194888177,
+ "grad_norm": 0.05419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0499,
+ "step": 8646
+ },
+ {
+ "epoch": 27.62619808306709,
+ "grad_norm": 0.060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0438,
+ "step": 8647
+ },
+ {
+ "epoch": 27.629392971246006,
+ "grad_norm": 0.060791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0443,
+ "step": 8648
+ },
+ {
+ "epoch": 27.63258785942492,
+ "grad_norm": 0.0888671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0268,
+ "step": 8649
+ },
+ {
+ "epoch": 27.635782747603834,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0462,
+ "step": 8650
+ },
+ {
+ "epoch": 27.638977635782748,
+ "grad_norm": 0.048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.042,
+ "step": 8651
+ },
+ {
+ "epoch": 27.642172523961662,
+ "grad_norm": 0.08349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0362,
+ "step": 8652
+ },
+ {
+ "epoch": 27.645367412140576,
+ "grad_norm": 0.058349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.046,
+ "step": 8653
+ },
+ {
+ "epoch": 27.64856230031949,
+ "grad_norm": 0.0810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0511,
+ "step": 8654
+ },
+ {
+ "epoch": 27.6517571884984,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0215,
+ "step": 8655
+ },
+ {
+ "epoch": 27.654952076677315,
+ "grad_norm": 0.11474609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0454,
+ "step": 8656
+ },
+ {
+ "epoch": 27.65814696485623,
+ "grad_norm": 0.05810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0459,
+ "step": 8657
+ },
+ {
+ "epoch": 27.661341853035143,
+ "grad_norm": 0.08642578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0338,
+ "step": 8658
+ },
+ {
+ "epoch": 27.664536741214057,
+ "grad_norm": 0.0517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0366,
+ "step": 8659
+ },
+ {
+ "epoch": 27.66773162939297,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0442,
+ "step": 8660
+ },
+ {
+ "epoch": 27.670926517571885,
+ "grad_norm": 0.05224609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0497,
+ "step": 8661
+ },
+ {
+ "epoch": 27.6741214057508,
+ "grad_norm": 0.053466796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0481,
+ "step": 8662
+ },
+ {
+ "epoch": 27.677316293929714,
+ "grad_norm": 0.051025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0384,
+ "step": 8663
+ },
+ {
+ "epoch": 27.680511182108628,
+ "grad_norm": 0.052490234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0293,
+ "step": 8664
+ },
+ {
+ "epoch": 27.68370607028754,
+ "grad_norm": 0.052490234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0435,
+ "step": 8665
+ },
+ {
+ "epoch": 27.686900958466452,
+ "grad_norm": 0.053466796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0379,
+ "step": 8666
+ },
+ {
+ "epoch": 27.690095846645367,
+ "grad_norm": 0.0517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0397,
+ "step": 8667
+ },
+ {
+ "epoch": 27.69329073482428,
+ "grad_norm": 0.052734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0489,
+ "step": 8668
+ },
+ {
+ "epoch": 27.696485623003195,
+ "grad_norm": 0.05517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0366,
+ "step": 8669
+ },
+ {
+ "epoch": 27.69968051118211,
+ "grad_norm": 0.05859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0546,
+ "step": 8670
+ },
+ {
+ "epoch": 27.702875399361023,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0405,
+ "step": 8671
+ },
+ {
+ "epoch": 27.706070287539937,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0349,
+ "step": 8672
+ },
+ {
+ "epoch": 27.70926517571885,
+ "grad_norm": 0.060791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0451,
+ "step": 8673
+ },
+ {
+ "epoch": 27.712460063897765,
+ "grad_norm": 0.049072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0311,
+ "step": 8674
+ },
+ {
+ "epoch": 27.715654952076676,
+ "grad_norm": 0.05419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0418,
+ "step": 8675
+ },
+ {
+ "epoch": 27.71884984025559,
+ "grad_norm": 0.054443359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0357,
+ "step": 8676
+ },
+ {
+ "epoch": 27.722044728434504,
+ "grad_norm": 0.05810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0475,
+ "step": 8677
+ },
+ {
+ "epoch": 27.72523961661342,
+ "grad_norm": 0.056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0338,
+ "step": 8678
+ },
+ {
+ "epoch": 27.728434504792332,
+ "grad_norm": 0.05810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0412,
+ "step": 8679
+ },
+ {
+ "epoch": 27.731629392971247,
+ "grad_norm": 0.048583984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0369,
+ "step": 8680
+ },
+ {
+ "epoch": 27.73482428115016,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0472,
+ "step": 8681
+ },
+ {
+ "epoch": 27.738019169329075,
+ "grad_norm": 0.054931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0358,
+ "step": 8682
+ },
+ {
+ "epoch": 27.74121405750799,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.038,
+ "step": 8683
+ },
+ {
+ "epoch": 27.7444089456869,
+ "grad_norm": 0.053466796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0347,
+ "step": 8684
+ },
+ {
+ "epoch": 27.747603833865814,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0433,
+ "step": 8685
+ },
+ {
+ "epoch": 27.750798722044728,
+ "grad_norm": 0.04931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0326,
+ "step": 8686
+ },
+ {
+ "epoch": 27.75399361022364,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0372,
+ "step": 8687
+ },
+ {
+ "epoch": 27.757188498402556,
+ "grad_norm": 0.058837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0341,
+ "step": 8688
+ },
+ {
+ "epoch": 27.76038338658147,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0353,
+ "step": 8689
+ },
+ {
+ "epoch": 27.763578274760384,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0367,
+ "step": 8690
+ },
+ {
+ "epoch": 27.766773162939298,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0413,
+ "step": 8691
+ },
+ {
+ "epoch": 27.769968051118212,
+ "grad_norm": 0.05908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0406,
+ "step": 8692
+ },
+ {
+ "epoch": 27.773162939297123,
+ "grad_norm": 0.0498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0376,
+ "step": 8693
+ },
+ {
+ "epoch": 27.776357827476037,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0453,
+ "step": 8694
+ },
+ {
+ "epoch": 27.77955271565495,
+ "grad_norm": 0.05224609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0402,
+ "step": 8695
+ },
+ {
+ "epoch": 27.782747603833865,
+ "grad_norm": 0.059814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0475,
+ "step": 8696
+ },
+ {
+ "epoch": 27.78594249201278,
+ "grad_norm": 0.04833984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0435,
+ "step": 8697
+ },
+ {
+ "epoch": 27.789137380191693,
+ "grad_norm": 0.05029296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0312,
+ "step": 8698
+ },
+ {
+ "epoch": 27.792332268370608,
+ "grad_norm": 0.06005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0398,
+ "step": 8699
+ },
+ {
+ "epoch": 27.79552715654952,
+ "grad_norm": 0.08349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0341,
+ "step": 8700
+ },
+ {
+ "epoch": 27.798722044728436,
+ "grad_norm": 0.059814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0559,
+ "step": 8701
+ },
+ {
+ "epoch": 27.80191693290735,
+ "grad_norm": 0.080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0484,
+ "step": 8702
+ },
+ {
+ "epoch": 27.80511182108626,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0378,
+ "step": 8703
+ },
+ {
+ "epoch": 27.808306709265175,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0476,
+ "step": 8704
+ },
+ {
+ "epoch": 27.81150159744409,
+ "grad_norm": 0.056396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0371,
+ "step": 8705
+ },
+ {
+ "epoch": 27.814696485623003,
+ "grad_norm": 0.048095703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0329,
+ "step": 8706
+ },
+ {
+ "epoch": 27.817891373801917,
+ "grad_norm": 0.05126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0386,
+ "step": 8707
+ },
+ {
+ "epoch": 27.82108626198083,
+ "grad_norm": 0.0537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0474,
+ "step": 8708
+ },
+ {
+ "epoch": 27.824281150159745,
+ "grad_norm": 0.052001953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0403,
+ "step": 8709
+ },
+ {
+ "epoch": 27.82747603833866,
+ "grad_norm": 0.044677734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0387,
+ "step": 8710
+ },
+ {
+ "epoch": 27.830670926517573,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0347,
+ "step": 8711
+ },
+ {
+ "epoch": 27.833865814696484,
+ "grad_norm": 0.050537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0326,
+ "step": 8712
+ },
+ {
+ "epoch": 27.837060702875398,
+ "grad_norm": 0.0888671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0356,
+ "step": 8713
+ },
+ {
+ "epoch": 27.840255591054312,
+ "grad_norm": 0.046142578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0326,
+ "step": 8714
+ },
+ {
+ "epoch": 27.843450479233226,
+ "grad_norm": 0.08447265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0365,
+ "step": 8715
+ },
+ {
+ "epoch": 27.84664536741214,
+ "grad_norm": 0.05615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0463,
+ "step": 8716
+ },
+ {
+ "epoch": 27.849840255591054,
+ "grad_norm": 0.04541015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0333,
+ "step": 8717
+ },
+ {
+ "epoch": 27.85303514376997,
+ "grad_norm": 0.062255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0419,
+ "step": 8718
+ },
+ {
+ "epoch": 27.856230031948883,
+ "grad_norm": 0.05810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0422,
+ "step": 8719
+ },
+ {
+ "epoch": 27.859424920127797,
+ "grad_norm": 0.052734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0441,
+ "step": 8720
+ },
+ {
+ "epoch": 27.86261980830671,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0422,
+ "step": 8721
+ },
+ {
+ "epoch": 27.86581469648562,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0415,
+ "step": 8722
+ },
+ {
+ "epoch": 27.869009584664536,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0441,
+ "step": 8723
+ },
+ {
+ "epoch": 27.87220447284345,
+ "grad_norm": 0.05859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0409,
+ "step": 8724
+ },
+ {
+ "epoch": 27.875399361022364,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0328,
+ "step": 8725
+ },
+ {
+ "epoch": 27.878594249201278,
+ "grad_norm": 0.05859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0351,
+ "step": 8726
+ },
+ {
+ "epoch": 27.881789137380192,
+ "grad_norm": 0.08642578125,
+ "learning_rate": 0.0005,
+ "loss": 1.033,
+ "step": 8727
+ },
+ {
+ "epoch": 27.884984025559106,
+ "grad_norm": 0.1259765625,
+ "learning_rate": 0.0005,
+ "loss": 1.045,
+ "step": 8728
+ },
+ {
+ "epoch": 27.88817891373802,
+ "grad_norm": 0.189453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0503,
+ "step": 8729
+ },
+ {
+ "epoch": 27.891373801916934,
+ "grad_norm": 0.09033203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0411,
+ "step": 8730
+ },
+ {
+ "epoch": 27.894568690095845,
+ "grad_norm": 0.1689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0502,
+ "step": 8731
+ },
+ {
+ "epoch": 27.89776357827476,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0398,
+ "step": 8732
+ },
+ {
+ "epoch": 27.900958466453673,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0437,
+ "step": 8733
+ },
+ {
+ "epoch": 27.904153354632587,
+ "grad_norm": 0.054931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0374,
+ "step": 8734
+ },
+ {
+ "epoch": 27.9073482428115,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0334,
+ "step": 8735
+ },
+ {
+ "epoch": 27.910543130990416,
+ "grad_norm": 0.10205078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0394,
+ "step": 8736
+ },
+ {
+ "epoch": 27.91373801916933,
+ "grad_norm": 0.08203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0387,
+ "step": 8737
+ },
+ {
+ "epoch": 27.916932907348244,
+ "grad_norm": 0.054443359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0494,
+ "step": 8738
+ },
+ {
+ "epoch": 27.920127795527158,
+ "grad_norm": 0.1953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0318,
+ "step": 8739
+ },
+ {
+ "epoch": 27.923322683706072,
+ "grad_norm": 0.1533203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0303,
+ "step": 8740
+ },
+ {
+ "epoch": 27.926517571884983,
+ "grad_norm": 0.09521484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0317,
+ "step": 8741
+ },
+ {
+ "epoch": 27.929712460063897,
+ "grad_norm": 0.09912109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0323,
+ "step": 8742
+ },
+ {
+ "epoch": 27.93290734824281,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0384,
+ "step": 8743
+ },
+ {
+ "epoch": 27.936102236421725,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0361,
+ "step": 8744
+ },
+ {
+ "epoch": 27.93929712460064,
+ "grad_norm": 0.05712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0265,
+ "step": 8745
+ },
+ {
+ "epoch": 27.942492012779553,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0323,
+ "step": 8746
+ },
+ {
+ "epoch": 27.945686900958467,
+ "grad_norm": 0.08203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0481,
+ "step": 8747
+ },
+ {
+ "epoch": 27.94888178913738,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0447,
+ "step": 8748
+ },
+ {
+ "epoch": 27.952076677316295,
+ "grad_norm": 0.08251953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0364,
+ "step": 8749
+ },
+ {
+ "epoch": 27.955271565495206,
+ "grad_norm": 0.08935546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0457,
+ "step": 8750
+ },
+ {
+ "epoch": 27.95846645367412,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.041,
+ "step": 8751
+ },
+ {
+ "epoch": 27.961661341853034,
+ "grad_norm": 0.047607421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0241,
+ "step": 8752
+ },
+ {
+ "epoch": 27.96485623003195,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0479,
+ "step": 8753
+ },
+ {
+ "epoch": 27.968051118210862,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0208,
+ "step": 8754
+ },
+ {
+ "epoch": 27.971246006389777,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0344,
+ "step": 8755
+ },
+ {
+ "epoch": 27.97444089456869,
+ "grad_norm": 0.05615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0326,
+ "step": 8756
+ },
+ {
+ "epoch": 27.977635782747605,
+ "grad_norm": 0.087890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0368,
+ "step": 8757
+ },
+ {
+ "epoch": 27.98083067092652,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0374,
+ "step": 8758
+ },
+ {
+ "epoch": 27.984025559105433,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0412,
+ "step": 8759
+ },
+ {
+ "epoch": 27.987220447284344,
+ "grad_norm": 0.1279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0402,
+ "step": 8760
+ },
+ {
+ "epoch": 27.990415335463258,
+ "grad_norm": 0.0888671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0423,
+ "step": 8761
+ },
+ {
+ "epoch": 27.99361022364217,
+ "grad_norm": 0.11767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0344,
+ "step": 8762
+ },
+ {
+ "epoch": 27.996805111821086,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0444,
+ "step": 8763
+ },
+ {
+ "epoch": 28.0,
+ "grad_norm": 0.08837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0451,
+ "step": 8764
+ },
+ {
+ "epoch": 28.003194888178914,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0334,
+ "step": 8765
+ },
+ {
+ "epoch": 28.00638977635783,
+ "grad_norm": 0.107421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0372,
+ "step": 8766
+ },
+ {
+ "epoch": 28.009584664536742,
+ "grad_norm": 0.05322265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0455,
+ "step": 8767
+ },
+ {
+ "epoch": 28.012779552715656,
+ "grad_norm": 0.10302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0427,
+ "step": 8768
+ },
+ {
+ "epoch": 28.015974440894567,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0441,
+ "step": 8769
+ },
+ {
+ "epoch": 28.01916932907348,
+ "grad_norm": 0.07861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0438,
+ "step": 8770
+ },
+ {
+ "epoch": 28.022364217252395,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0484,
+ "step": 8771
+ },
+ {
+ "epoch": 28.02555910543131,
+ "grad_norm": 0.1015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0333,
+ "step": 8772
+ },
+ {
+ "epoch": 28.028753993610223,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0365,
+ "step": 8773
+ },
+ {
+ "epoch": 28.031948881789138,
+ "grad_norm": 0.0849609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0364,
+ "step": 8774
+ },
+ {
+ "epoch": 28.03514376996805,
+ "grad_norm": 0.0556640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0477,
+ "step": 8775
+ },
+ {
+ "epoch": 28.038338658146966,
+ "grad_norm": 0.1015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0328,
+ "step": 8776
+ },
+ {
+ "epoch": 28.04153354632588,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0421,
+ "step": 8777
+ },
+ {
+ "epoch": 28.044728434504794,
+ "grad_norm": 0.10009765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0469,
+ "step": 8778
+ },
+ {
+ "epoch": 28.047923322683705,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.022,
+ "step": 8779
+ },
+ {
+ "epoch": 28.05111821086262,
+ "grad_norm": 0.087890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0479,
+ "step": 8780
+ },
+ {
+ "epoch": 28.054313099041533,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0368,
+ "step": 8781
+ },
+ {
+ "epoch": 28.057507987220447,
+ "grad_norm": 0.1142578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0275,
+ "step": 8782
+ },
+ {
+ "epoch": 28.06070287539936,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0352,
+ "step": 8783
+ },
+ {
+ "epoch": 28.063897763578275,
+ "grad_norm": 0.0615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0416,
+ "step": 8784
+ },
+ {
+ "epoch": 28.06709265175719,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.038,
+ "step": 8785
+ },
+ {
+ "epoch": 28.070287539936103,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0359,
+ "step": 8786
+ },
+ {
+ "epoch": 28.073482428115017,
+ "grad_norm": 0.05224609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0432,
+ "step": 8787
+ },
+ {
+ "epoch": 28.076677316293928,
+ "grad_norm": 0.058837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0197,
+ "step": 8788
+ },
+ {
+ "epoch": 28.079872204472842,
+ "grad_norm": 0.049560546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0348,
+ "step": 8789
+ },
+ {
+ "epoch": 28.083067092651756,
+ "grad_norm": 0.05908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0379,
+ "step": 8790
+ },
+ {
+ "epoch": 28.08626198083067,
+ "grad_norm": 0.0556640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0333,
+ "step": 8791
+ },
+ {
+ "epoch": 28.089456869009584,
+ "grad_norm": 0.054443359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0349,
+ "step": 8792
+ },
+ {
+ "epoch": 28.0926517571885,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0482,
+ "step": 8793
+ },
+ {
+ "epoch": 28.095846645367413,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0325,
+ "step": 8794
+ },
+ {
+ "epoch": 28.099041533546327,
+ "grad_norm": 0.058349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0287,
+ "step": 8795
+ },
+ {
+ "epoch": 28.10223642172524,
+ "grad_norm": 0.052001953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0427,
+ "step": 8796
+ },
+ {
+ "epoch": 28.105431309904155,
+ "grad_norm": 0.087890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0467,
+ "step": 8797
+ },
+ {
+ "epoch": 28.108626198083066,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0317,
+ "step": 8798
+ },
+ {
+ "epoch": 28.11182108626198,
+ "grad_norm": 0.11181640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0363,
+ "step": 8799
+ },
+ {
+ "epoch": 28.115015974440894,
+ "grad_norm": 0.10205078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0347,
+ "step": 8800
+ },
+ {
+ "epoch": 28.118210862619808,
+ "grad_norm": 0.10546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0306,
+ "step": 8801
+ },
+ {
+ "epoch": 28.121405750798722,
+ "grad_norm": 0.08203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0449,
+ "step": 8802
+ },
+ {
+ "epoch": 28.124600638977636,
+ "grad_norm": 0.0546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0488,
+ "step": 8803
+ },
+ {
+ "epoch": 28.12779552715655,
+ "grad_norm": 0.0439453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0343,
+ "step": 8804
+ },
+ {
+ "epoch": 28.130990415335464,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0381,
+ "step": 8805
+ },
+ {
+ "epoch": 28.13418530351438,
+ "grad_norm": 0.055419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0426,
+ "step": 8806
+ },
+ {
+ "epoch": 28.13738019169329,
+ "grad_norm": 0.048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0386,
+ "step": 8807
+ },
+ {
+ "epoch": 28.140575079872203,
+ "grad_norm": 0.0546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0445,
+ "step": 8808
+ },
+ {
+ "epoch": 28.143769968051117,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0309,
+ "step": 8809
+ },
+ {
+ "epoch": 28.14696485623003,
+ "grad_norm": 0.05859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0314,
+ "step": 8810
+ },
+ {
+ "epoch": 28.150159744408946,
+ "grad_norm": 0.057861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0385,
+ "step": 8811
+ },
+ {
+ "epoch": 28.15335463258786,
+ "grad_norm": 0.09375,
+ "learning_rate": 0.0005,
+ "loss": 1.0447,
+ "step": 8812
+ },
+ {
+ "epoch": 28.156549520766774,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0336,
+ "step": 8813
+ },
+ {
+ "epoch": 28.159744408945688,
+ "grad_norm": 0.055419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0387,
+ "step": 8814
+ },
+ {
+ "epoch": 28.162939297124602,
+ "grad_norm": 0.060302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0331,
+ "step": 8815
+ },
+ {
+ "epoch": 28.166134185303516,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0336,
+ "step": 8816
+ },
+ {
+ "epoch": 28.169329073482427,
+ "grad_norm": 0.099609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0333,
+ "step": 8817
+ },
+ {
+ "epoch": 28.17252396166134,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0334,
+ "step": 8818
+ },
+ {
+ "epoch": 28.175718849840255,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0455,
+ "step": 8819
+ },
+ {
+ "epoch": 28.17891373801917,
+ "grad_norm": 0.08740234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0456,
+ "step": 8820
+ },
+ {
+ "epoch": 28.182108626198083,
+ "grad_norm": 0.1162109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0337,
+ "step": 8821
+ },
+ {
+ "epoch": 28.185303514376997,
+ "grad_norm": 0.10107421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0386,
+ "step": 8822
+ },
+ {
+ "epoch": 28.18849840255591,
+ "grad_norm": 0.09130859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0328,
+ "step": 8823
+ },
+ {
+ "epoch": 28.191693290734825,
+ "grad_norm": 0.09130859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0391,
+ "step": 8824
+ },
+ {
+ "epoch": 28.19488817891374,
+ "grad_norm": 0.05322265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0457,
+ "step": 8825
+ },
+ {
+ "epoch": 28.19808306709265,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0502,
+ "step": 8826
+ },
+ {
+ "epoch": 28.201277955271564,
+ "grad_norm": 0.051025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0521,
+ "step": 8827
+ },
+ {
+ "epoch": 28.20447284345048,
+ "grad_norm": 0.051025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0361,
+ "step": 8828
+ },
+ {
+ "epoch": 28.207667731629392,
+ "grad_norm": 0.06005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0379,
+ "step": 8829
+ },
+ {
+ "epoch": 28.210862619808307,
+ "grad_norm": 0.0556640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0585,
+ "step": 8830
+ },
+ {
+ "epoch": 28.21405750798722,
+ "grad_norm": 0.0517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0534,
+ "step": 8831
+ },
+ {
+ "epoch": 28.217252396166135,
+ "grad_norm": 0.052734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0436,
+ "step": 8832
+ },
+ {
+ "epoch": 28.22044728434505,
+ "grad_norm": 0.057373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0273,
+ "step": 8833
+ },
+ {
+ "epoch": 28.223642172523963,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0482,
+ "step": 8834
+ },
+ {
+ "epoch": 28.226837060702877,
+ "grad_norm": 0.05517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0368,
+ "step": 8835
+ },
+ {
+ "epoch": 28.230031948881788,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0416,
+ "step": 8836
+ },
+ {
+ "epoch": 28.233226837060702,
+ "grad_norm": 0.0615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0417,
+ "step": 8837
+ },
+ {
+ "epoch": 28.236421725239616,
+ "grad_norm": 0.059814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0396,
+ "step": 8838
+ },
+ {
+ "epoch": 28.23961661341853,
+ "grad_norm": 0.10205078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0291,
+ "step": 8839
+ },
+ {
+ "epoch": 28.242811501597444,
+ "grad_norm": 0.052001953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0302,
+ "step": 8840
+ },
+ {
+ "epoch": 28.24600638977636,
+ "grad_norm": 0.1396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0389,
+ "step": 8841
+ },
+ {
+ "epoch": 28.249201277955272,
+ "grad_norm": 0.08642578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0279,
+ "step": 8842
+ },
+ {
+ "epoch": 28.252396166134186,
+ "grad_norm": 0.1845703125,
+ "learning_rate": 0.0005,
+ "loss": 1.038,
+ "step": 8843
+ },
+ {
+ "epoch": 28.2555910543131,
+ "grad_norm": 0.09375,
+ "learning_rate": 0.0005,
+ "loss": 1.0488,
+ "step": 8844
+ },
+ {
+ "epoch": 28.25878594249201,
+ "grad_norm": 0.12255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0418,
+ "step": 8845
+ },
+ {
+ "epoch": 28.261980830670925,
+ "grad_norm": 0.1015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0406,
+ "step": 8846
+ },
+ {
+ "epoch": 28.26517571884984,
+ "grad_norm": 0.10205078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0563,
+ "step": 8847
+ },
+ {
+ "epoch": 28.268370607028753,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0497,
+ "step": 8848
+ },
+ {
+ "epoch": 28.271565495207668,
+ "grad_norm": 0.0849609375,
+ "learning_rate": 0.0005,
+ "loss": 1.044,
+ "step": 8849
+ },
+ {
+ "epoch": 28.27476038338658,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0496,
+ "step": 8850
+ },
+ {
+ "epoch": 28.277955271565496,
+ "grad_norm": 0.083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0439,
+ "step": 8851
+ },
+ {
+ "epoch": 28.28115015974441,
+ "grad_norm": 0.08203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0296,
+ "step": 8852
+ },
+ {
+ "epoch": 28.284345047923324,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0408,
+ "step": 8853
+ },
+ {
+ "epoch": 28.287539936102238,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0449,
+ "step": 8854
+ },
+ {
+ "epoch": 28.29073482428115,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0374,
+ "step": 8855
+ },
+ {
+ "epoch": 28.293929712460063,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0261,
+ "step": 8856
+ },
+ {
+ "epoch": 28.297124600638977,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0412,
+ "step": 8857
+ },
+ {
+ "epoch": 28.30031948881789,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0476,
+ "step": 8858
+ },
+ {
+ "epoch": 28.303514376996805,
+ "grad_norm": 0.060791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0398,
+ "step": 8859
+ },
+ {
+ "epoch": 28.30670926517572,
+ "grad_norm": 0.053466796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0423,
+ "step": 8860
+ },
+ {
+ "epoch": 28.309904153354633,
+ "grad_norm": 0.050537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0332,
+ "step": 8861
+ },
+ {
+ "epoch": 28.313099041533548,
+ "grad_norm": 0.05029296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0374,
+ "step": 8862
+ },
+ {
+ "epoch": 28.31629392971246,
+ "grad_norm": 0.058837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.04,
+ "step": 8863
+ },
+ {
+ "epoch": 28.319488817891372,
+ "grad_norm": 0.052734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0452,
+ "step": 8864
+ },
+ {
+ "epoch": 28.322683706070286,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0396,
+ "step": 8865
+ },
+ {
+ "epoch": 28.3258785942492,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0448,
+ "step": 8866
+ },
+ {
+ "epoch": 28.329073482428115,
+ "grad_norm": 0.060302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0415,
+ "step": 8867
+ },
+ {
+ "epoch": 28.33226837060703,
+ "grad_norm": 0.1103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0424,
+ "step": 8868
+ },
+ {
+ "epoch": 28.335463258785943,
+ "grad_norm": 0.08935546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0373,
+ "step": 8869
+ },
+ {
+ "epoch": 28.338658146964857,
+ "grad_norm": 0.10302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0361,
+ "step": 8870
+ },
+ {
+ "epoch": 28.34185303514377,
+ "grad_norm": 0.083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0407,
+ "step": 8871
+ },
+ {
+ "epoch": 28.345047923322685,
+ "grad_norm": 0.055419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0419,
+ "step": 8872
+ },
+ {
+ "epoch": 28.3482428115016,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0477,
+ "step": 8873
+ },
+ {
+ "epoch": 28.35143769968051,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.044,
+ "step": 8874
+ },
+ {
+ "epoch": 28.354632587859424,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0549,
+ "step": 8875
+ },
+ {
+ "epoch": 28.357827476038338,
+ "grad_norm": 0.0595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0432,
+ "step": 8876
+ },
+ {
+ "epoch": 28.361022364217252,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0405,
+ "step": 8877
+ },
+ {
+ "epoch": 28.364217252396166,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0327,
+ "step": 8878
+ },
+ {
+ "epoch": 28.36741214057508,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0345,
+ "step": 8879
+ },
+ {
+ "epoch": 28.370607028753994,
+ "grad_norm": 0.08203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0441,
+ "step": 8880
+ },
+ {
+ "epoch": 28.37380191693291,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0504,
+ "step": 8881
+ },
+ {
+ "epoch": 28.376996805111823,
+ "grad_norm": 0.08203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0416,
+ "step": 8882
+ },
+ {
+ "epoch": 28.380191693290733,
+ "grad_norm": 0.055908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0351,
+ "step": 8883
+ },
+ {
+ "epoch": 28.383386581469647,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0485,
+ "step": 8884
+ },
+ {
+ "epoch": 28.38658146964856,
+ "grad_norm": 0.058837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.043,
+ "step": 8885
+ },
+ {
+ "epoch": 28.389776357827476,
+ "grad_norm": 0.0556640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0569,
+ "step": 8886
+ },
+ {
+ "epoch": 28.39297124600639,
+ "grad_norm": 0.055419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0293,
+ "step": 8887
+ },
+ {
+ "epoch": 28.396166134185304,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0402,
+ "step": 8888
+ },
+ {
+ "epoch": 28.399361022364218,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0341,
+ "step": 8889
+ },
+ {
+ "epoch": 28.402555910543132,
+ "grad_norm": 0.049560546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0386,
+ "step": 8890
+ },
+ {
+ "epoch": 28.405750798722046,
+ "grad_norm": 0.05322265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0386,
+ "step": 8891
+ },
+ {
+ "epoch": 28.408945686900957,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0356,
+ "step": 8892
+ },
+ {
+ "epoch": 28.41214057507987,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0384,
+ "step": 8893
+ },
+ {
+ "epoch": 28.415335463258785,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0409,
+ "step": 8894
+ },
+ {
+ "epoch": 28.4185303514377,
+ "grad_norm": 0.07861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0493,
+ "step": 8895
+ },
+ {
+ "epoch": 28.421725239616613,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0522,
+ "step": 8896
+ },
+ {
+ "epoch": 28.424920127795527,
+ "grad_norm": 0.1142578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0326,
+ "step": 8897
+ },
+ {
+ "epoch": 28.42811501597444,
+ "grad_norm": 0.052001953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0258,
+ "step": 8898
+ },
+ {
+ "epoch": 28.431309904153355,
+ "grad_norm": 0.05908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0392,
+ "step": 8899
+ },
+ {
+ "epoch": 28.43450479233227,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0425,
+ "step": 8900
+ },
+ {
+ "epoch": 28.437699680511184,
+ "grad_norm": 0.07861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0469,
+ "step": 8901
+ },
+ {
+ "epoch": 28.440894568690094,
+ "grad_norm": 0.099609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0449,
+ "step": 8902
+ },
+ {
+ "epoch": 28.44408945686901,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0448,
+ "step": 8903
+ },
+ {
+ "epoch": 28.447284345047922,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0319,
+ "step": 8904
+ },
+ {
+ "epoch": 28.450479233226837,
+ "grad_norm": 0.1044921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0408,
+ "step": 8905
+ },
+ {
+ "epoch": 28.45367412140575,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0309,
+ "step": 8906
+ },
+ {
+ "epoch": 28.456869009584665,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0402,
+ "step": 8907
+ },
+ {
+ "epoch": 28.46006389776358,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0379,
+ "step": 8908
+ },
+ {
+ "epoch": 28.463258785942493,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0348,
+ "step": 8909
+ },
+ {
+ "epoch": 28.466453674121407,
+ "grad_norm": 0.050048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0392,
+ "step": 8910
+ },
+ {
+ "epoch": 28.46964856230032,
+ "grad_norm": 0.05810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.029,
+ "step": 8911
+ },
+ {
+ "epoch": 28.472843450479232,
+ "grad_norm": 0.0546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0463,
+ "step": 8912
+ },
+ {
+ "epoch": 28.476038338658146,
+ "grad_norm": 0.060302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0347,
+ "step": 8913
+ },
+ {
+ "epoch": 28.47923322683706,
+ "grad_norm": 0.061279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0539,
+ "step": 8914
+ },
+ {
+ "epoch": 28.482428115015974,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0262,
+ "step": 8915
+ },
+ {
+ "epoch": 28.48562300319489,
+ "grad_norm": 0.057373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0352,
+ "step": 8916
+ },
+ {
+ "epoch": 28.488817891373802,
+ "grad_norm": 0.062255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.038,
+ "step": 8917
+ },
+ {
+ "epoch": 28.492012779552716,
+ "grad_norm": 0.05126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0301,
+ "step": 8918
+ },
+ {
+ "epoch": 28.49520766773163,
+ "grad_norm": 0.08154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0424,
+ "step": 8919
+ },
+ {
+ "epoch": 28.498402555910545,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0363,
+ "step": 8920
+ },
+ {
+ "epoch": 28.501597444089455,
+ "grad_norm": 0.091796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0376,
+ "step": 8921
+ },
+ {
+ "epoch": 28.50479233226837,
+ "grad_norm": 0.057861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0427,
+ "step": 8922
+ },
+ {
+ "epoch": 28.507987220447284,
+ "grad_norm": 0.0986328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0506,
+ "step": 8923
+ },
+ {
+ "epoch": 28.511182108626198,
+ "grad_norm": 0.0556640625,
+ "learning_rate": 0.0005,
+ "loss": 1.033,
+ "step": 8924
+ },
+ {
+ "epoch": 28.51437699680511,
+ "grad_norm": 0.05224609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0433,
+ "step": 8925
+ },
+ {
+ "epoch": 28.517571884984026,
+ "grad_norm": 0.048095703125,
+ "learning_rate": 0.0005,
+ "loss": 1.031,
+ "step": 8926
+ },
+ {
+ "epoch": 28.52076677316294,
+ "grad_norm": 0.059814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0501,
+ "step": 8927
+ },
+ {
+ "epoch": 28.523961661341854,
+ "grad_norm": 0.05078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0231,
+ "step": 8928
+ },
+ {
+ "epoch": 28.527156549520768,
+ "grad_norm": 0.08984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0273,
+ "step": 8929
+ },
+ {
+ "epoch": 28.53035143769968,
+ "grad_norm": 0.055908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0484,
+ "step": 8930
+ },
+ {
+ "epoch": 28.533546325878593,
+ "grad_norm": 0.0966796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0455,
+ "step": 8931
+ },
+ {
+ "epoch": 28.536741214057507,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0369,
+ "step": 8932
+ },
+ {
+ "epoch": 28.53993610223642,
+ "grad_norm": 0.08056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0407,
+ "step": 8933
+ },
+ {
+ "epoch": 28.543130990415335,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.035,
+ "step": 8934
+ },
+ {
+ "epoch": 28.54632587859425,
+ "grad_norm": 0.055419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0379,
+ "step": 8935
+ },
+ {
+ "epoch": 28.549520766773163,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0396,
+ "step": 8936
+ },
+ {
+ "epoch": 28.552715654952078,
+ "grad_norm": 0.060302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0498,
+ "step": 8937
+ },
+ {
+ "epoch": 28.55591054313099,
+ "grad_norm": 0.06201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.033,
+ "step": 8938
+ },
+ {
+ "epoch": 28.559105431309906,
+ "grad_norm": 0.062255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0362,
+ "step": 8939
+ },
+ {
+ "epoch": 28.562300319488816,
+ "grad_norm": 0.0859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0416,
+ "step": 8940
+ },
+ {
+ "epoch": 28.56549520766773,
+ "grad_norm": 0.061279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0408,
+ "step": 8941
+ },
+ {
+ "epoch": 28.568690095846645,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0524,
+ "step": 8942
+ },
+ {
+ "epoch": 28.57188498402556,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0236,
+ "step": 8943
+ },
+ {
+ "epoch": 28.575079872204473,
+ "grad_norm": 0.061279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0236,
+ "step": 8944
+ },
+ {
+ "epoch": 28.578274760383387,
+ "grad_norm": 0.09423828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0438,
+ "step": 8945
+ },
+ {
+ "epoch": 28.5814696485623,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0257,
+ "step": 8946
+ },
+ {
+ "epoch": 28.584664536741215,
+ "grad_norm": 0.12060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0392,
+ "step": 8947
+ },
+ {
+ "epoch": 28.58785942492013,
+ "grad_norm": 0.10693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0299,
+ "step": 8948
+ },
+ {
+ "epoch": 28.591054313099043,
+ "grad_norm": 0.1533203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0383,
+ "step": 8949
+ },
+ {
+ "epoch": 28.594249201277954,
+ "grad_norm": 0.08203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0442,
+ "step": 8950
+ },
+ {
+ "epoch": 28.597444089456868,
+ "grad_norm": 0.095703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0417,
+ "step": 8951
+ },
+ {
+ "epoch": 28.600638977635782,
+ "grad_norm": 0.0810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0325,
+ "step": 8952
+ },
+ {
+ "epoch": 28.603833865814696,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0372,
+ "step": 8953
+ },
+ {
+ "epoch": 28.60702875399361,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0308,
+ "step": 8954
+ },
+ {
+ "epoch": 28.610223642172524,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0542,
+ "step": 8955
+ },
+ {
+ "epoch": 28.61341853035144,
+ "grad_norm": 0.08935546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0368,
+ "step": 8956
+ },
+ {
+ "epoch": 28.616613418530353,
+ "grad_norm": 0.0927734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0298,
+ "step": 8957
+ },
+ {
+ "epoch": 28.619808306709267,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0407,
+ "step": 8958
+ },
+ {
+ "epoch": 28.623003194888177,
+ "grad_norm": 0.08642578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0323,
+ "step": 8959
+ },
+ {
+ "epoch": 28.62619808306709,
+ "grad_norm": 0.054931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0446,
+ "step": 8960
+ },
+ {
+ "epoch": 28.629392971246006,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0394,
+ "step": 8961
+ },
+ {
+ "epoch": 28.63258785942492,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0537,
+ "step": 8962
+ },
+ {
+ "epoch": 28.635782747603834,
+ "grad_norm": 0.058837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0387,
+ "step": 8963
+ },
+ {
+ "epoch": 28.638977635782748,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0392,
+ "step": 8964
+ },
+ {
+ "epoch": 28.642172523961662,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0313,
+ "step": 8965
+ },
+ {
+ "epoch": 28.645367412140576,
+ "grad_norm": 0.052001953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0328,
+ "step": 8966
+ },
+ {
+ "epoch": 28.64856230031949,
+ "grad_norm": 0.06005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0566,
+ "step": 8967
+ },
+ {
+ "epoch": 28.6517571884984,
+ "grad_norm": 0.060302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0455,
+ "step": 8968
+ },
+ {
+ "epoch": 28.654952076677315,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0366,
+ "step": 8969
+ },
+ {
+ "epoch": 28.65814696485623,
+ "grad_norm": 0.0966796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0386,
+ "step": 8970
+ },
+ {
+ "epoch": 28.661341853035143,
+ "grad_norm": 0.062255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0374,
+ "step": 8971
+ },
+ {
+ "epoch": 28.664536741214057,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0378,
+ "step": 8972
+ },
+ {
+ "epoch": 28.66773162939297,
+ "grad_norm": 0.091796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0406,
+ "step": 8973
+ },
+ {
+ "epoch": 28.670926517571885,
+ "grad_norm": 0.05078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0373,
+ "step": 8974
+ },
+ {
+ "epoch": 28.6741214057508,
+ "grad_norm": 0.10107421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0394,
+ "step": 8975
+ },
+ {
+ "epoch": 28.677316293929714,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0462,
+ "step": 8976
+ },
+ {
+ "epoch": 28.680511182108628,
+ "grad_norm": 0.058837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0356,
+ "step": 8977
+ },
+ {
+ "epoch": 28.68370607028754,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0441,
+ "step": 8978
+ },
+ {
+ "epoch": 28.686900958466452,
+ "grad_norm": 0.059326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0361,
+ "step": 8979
+ },
+ {
+ "epoch": 28.690095846645367,
+ "grad_norm": 0.05322265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0377,
+ "step": 8980
+ },
+ {
+ "epoch": 28.69329073482428,
+ "grad_norm": 0.05712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0391,
+ "step": 8981
+ },
+ {
+ "epoch": 28.696485623003195,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0329,
+ "step": 8982
+ },
+ {
+ "epoch": 28.69968051118211,
+ "grad_norm": 0.05224609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0528,
+ "step": 8983
+ },
+ {
+ "epoch": 28.702875399361023,
+ "grad_norm": 0.08837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.03,
+ "step": 8984
+ },
+ {
+ "epoch": 28.706070287539937,
+ "grad_norm": 0.050048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0443,
+ "step": 8985
+ },
+ {
+ "epoch": 28.70926517571885,
+ "grad_norm": 0.0927734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0364,
+ "step": 8986
+ },
+ {
+ "epoch": 28.712460063897765,
+ "grad_norm": 0.05517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0426,
+ "step": 8987
+ },
+ {
+ "epoch": 28.715654952076676,
+ "grad_norm": 0.09130859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0408,
+ "step": 8988
+ },
+ {
+ "epoch": 28.71884984025559,
+ "grad_norm": 0.05712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0376,
+ "step": 8989
+ },
+ {
+ "epoch": 28.722044728434504,
+ "grad_norm": 0.0595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0359,
+ "step": 8990
+ },
+ {
+ "epoch": 28.72523961661342,
+ "grad_norm": 0.059814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0344,
+ "step": 8991
+ },
+ {
+ "epoch": 28.728434504792332,
+ "grad_norm": 0.06103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0384,
+ "step": 8992
+ },
+ {
+ "epoch": 28.731629392971247,
+ "grad_norm": 0.05419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0504,
+ "step": 8993
+ },
+ {
+ "epoch": 28.73482428115016,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0461,
+ "step": 8994
+ },
+ {
+ "epoch": 28.738019169329075,
+ "grad_norm": 0.091796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0439,
+ "step": 8995
+ },
+ {
+ "epoch": 28.74121405750799,
+ "grad_norm": 0.059326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0505,
+ "step": 8996
+ },
+ {
+ "epoch": 28.7444089456869,
+ "grad_norm": 0.1220703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0381,
+ "step": 8997
+ },
+ {
+ "epoch": 28.747603833865814,
+ "grad_norm": 0.05029296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0325,
+ "step": 8998
+ },
+ {
+ "epoch": 28.750798722044728,
+ "grad_norm": 0.1357421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0464,
+ "step": 8999
+ },
+ {
+ "epoch": 28.75399361022364,
+ "grad_norm": 0.05859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0282,
+ "step": 9000
+ },
+ {
+ "epoch": 28.757188498402556,
+ "grad_norm": 0.1240234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0362,
+ "step": 9001
+ },
+ {
+ "epoch": 28.76038338658147,
+ "grad_norm": 0.228515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0396,
+ "step": 9002
+ },
+ {
+ "epoch": 28.763578274760384,
+ "grad_norm": 0.08935546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0404,
+ "step": 9003
+ },
+ {
+ "epoch": 28.766773162939298,
+ "grad_norm": 0.10888671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0449,
+ "step": 9004
+ },
+ {
+ "epoch": 28.769968051118212,
+ "grad_norm": 0.11962890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0362,
+ "step": 9005
+ },
+ {
+ "epoch": 28.773162939297123,
+ "grad_norm": 0.09326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0392,
+ "step": 9006
+ },
+ {
+ "epoch": 28.776357827476037,
+ "grad_norm": 0.12890625,
+ "learning_rate": 0.0005,
+ "loss": 1.039,
+ "step": 9007
+ },
+ {
+ "epoch": 28.77955271565495,
+ "grad_norm": 0.08740234375,
+ "learning_rate": 0.0005,
+ "loss": 1.042,
+ "step": 9008
+ },
+ {
+ "epoch": 28.782747603833865,
+ "grad_norm": 0.1494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0419,
+ "step": 9009
+ },
+ {
+ "epoch": 28.78594249201278,
+ "grad_norm": 0.0888671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0395,
+ "step": 9010
+ },
+ {
+ "epoch": 28.789137380191693,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0472,
+ "step": 9011
+ },
+ {
+ "epoch": 28.792332268370608,
+ "grad_norm": 0.125,
+ "learning_rate": 0.0005,
+ "loss": 1.0348,
+ "step": 9012
+ },
+ {
+ "epoch": 28.79552715654952,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.041,
+ "step": 9013
+ },
+ {
+ "epoch": 28.798722044728436,
+ "grad_norm": 0.08154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0486,
+ "step": 9014
+ },
+ {
+ "epoch": 28.80191693290735,
+ "grad_norm": 0.056396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0312,
+ "step": 9015
+ },
+ {
+ "epoch": 28.80511182108626,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0492,
+ "step": 9016
+ },
+ {
+ "epoch": 28.808306709265175,
+ "grad_norm": 0.060791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0369,
+ "step": 9017
+ },
+ {
+ "epoch": 28.81150159744409,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0372,
+ "step": 9018
+ },
+ {
+ "epoch": 28.814696485623003,
+ "grad_norm": 0.08154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0364,
+ "step": 9019
+ },
+ {
+ "epoch": 28.817891373801917,
+ "grad_norm": 0.0546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0289,
+ "step": 9020
+ },
+ {
+ "epoch": 28.82108626198083,
+ "grad_norm": 0.060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0464,
+ "step": 9021
+ },
+ {
+ "epoch": 28.824281150159745,
+ "grad_norm": 0.054931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0449,
+ "step": 9022
+ },
+ {
+ "epoch": 28.82747603833866,
+ "grad_norm": 0.05859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0397,
+ "step": 9023
+ },
+ {
+ "epoch": 28.830670926517573,
+ "grad_norm": 0.052490234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0292,
+ "step": 9024
+ },
+ {
+ "epoch": 28.833865814696484,
+ "grad_norm": 0.059814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0362,
+ "step": 9025
+ },
+ {
+ "epoch": 28.837060702875398,
+ "grad_norm": 0.083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0268,
+ "step": 9026
+ },
+ {
+ "epoch": 28.840255591054312,
+ "grad_norm": 0.060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0446,
+ "step": 9027
+ },
+ {
+ "epoch": 28.843450479233226,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0332,
+ "step": 9028
+ },
+ {
+ "epoch": 28.84664536741214,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0451,
+ "step": 9029
+ },
+ {
+ "epoch": 28.849840255591054,
+ "grad_norm": 0.08447265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0274,
+ "step": 9030
+ },
+ {
+ "epoch": 28.85303514376997,
+ "grad_norm": 0.057373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0376,
+ "step": 9031
+ },
+ {
+ "epoch": 28.856230031948883,
+ "grad_norm": 0.061767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.037,
+ "step": 9032
+ },
+ {
+ "epoch": 28.859424920127797,
+ "grad_norm": 0.061767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0393,
+ "step": 9033
+ },
+ {
+ "epoch": 28.86261980830671,
+ "grad_norm": 0.0546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0386,
+ "step": 9034
+ },
+ {
+ "epoch": 28.86581469648562,
+ "grad_norm": 0.05224609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0423,
+ "step": 9035
+ },
+ {
+ "epoch": 28.869009584664536,
+ "grad_norm": 0.053955078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0312,
+ "step": 9036
+ },
+ {
+ "epoch": 28.87220447284345,
+ "grad_norm": 0.052490234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0478,
+ "step": 9037
+ },
+ {
+ "epoch": 28.875399361022364,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0467,
+ "step": 9038
+ },
+ {
+ "epoch": 28.878594249201278,
+ "grad_norm": 0.051025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0263,
+ "step": 9039
+ },
+ {
+ "epoch": 28.881789137380192,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0347,
+ "step": 9040
+ },
+ {
+ "epoch": 28.884984025559106,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.04,
+ "step": 9041
+ },
+ {
+ "epoch": 28.88817891373802,
+ "grad_norm": 0.09033203125,
+ "learning_rate": 0.0005,
+ "loss": 1.035,
+ "step": 9042
+ },
+ {
+ "epoch": 28.891373801916934,
+ "grad_norm": 0.0556640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0277,
+ "step": 9043
+ },
+ {
+ "epoch": 28.894568690095845,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0362,
+ "step": 9044
+ },
+ {
+ "epoch": 28.89776357827476,
+ "grad_norm": 0.06005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0342,
+ "step": 9045
+ },
+ {
+ "epoch": 28.900958466453673,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0425,
+ "step": 9046
+ },
+ {
+ "epoch": 28.904153354632587,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0375,
+ "step": 9047
+ },
+ {
+ "epoch": 28.9073482428115,
+ "grad_norm": 0.08056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0415,
+ "step": 9048
+ },
+ {
+ "epoch": 28.910543130990416,
+ "grad_norm": 0.095703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0475,
+ "step": 9049
+ },
+ {
+ "epoch": 28.91373801916933,
+ "grad_norm": 0.09033203125,
+ "learning_rate": 0.0005,
+ "loss": 1.039,
+ "step": 9050
+ },
+ {
+ "epoch": 28.916932907348244,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0333,
+ "step": 9051
+ },
+ {
+ "epoch": 28.920127795527158,
+ "grad_norm": 0.099609375,
+ "learning_rate": 0.0005,
+ "loss": 1.028,
+ "step": 9052
+ },
+ {
+ "epoch": 28.923322683706072,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0331,
+ "step": 9053
+ },
+ {
+ "epoch": 28.926517571884983,
+ "grad_norm": 0.09521484375,
+ "learning_rate": 0.0005,
+ "loss": 1.028,
+ "step": 9054
+ },
+ {
+ "epoch": 28.929712460063897,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0289,
+ "step": 9055
+ },
+ {
+ "epoch": 28.93290734824281,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.04,
+ "step": 9056
+ },
+ {
+ "epoch": 28.936102236421725,
+ "grad_norm": 0.060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0444,
+ "step": 9057
+ },
+ {
+ "epoch": 28.93929712460064,
+ "grad_norm": 0.07861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0371,
+ "step": 9058
+ },
+ {
+ "epoch": 28.942492012779553,
+ "grad_norm": 0.061279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0401,
+ "step": 9059
+ },
+ {
+ "epoch": 28.945686900958467,
+ "grad_norm": 0.07861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.043,
+ "step": 9060
+ },
+ {
+ "epoch": 28.94888178913738,
+ "grad_norm": 0.0595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0409,
+ "step": 9061
+ },
+ {
+ "epoch": 28.952076677316295,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0338,
+ "step": 9062
+ },
+ {
+ "epoch": 28.955271565495206,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0433,
+ "step": 9063
+ },
+ {
+ "epoch": 28.95846645367412,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0347,
+ "step": 9064
+ },
+ {
+ "epoch": 28.961661341853034,
+ "grad_norm": 0.0556640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0455,
+ "step": 9065
+ },
+ {
+ "epoch": 28.96485623003195,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0343,
+ "step": 9066
+ },
+ {
+ "epoch": 28.968051118210862,
+ "grad_norm": 0.0546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0485,
+ "step": 9067
+ },
+ {
+ "epoch": 28.971246006389777,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0288,
+ "step": 9068
+ },
+ {
+ "epoch": 28.97444089456869,
+ "grad_norm": 0.06005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0364,
+ "step": 9069
+ },
+ {
+ "epoch": 28.977635782747605,
+ "grad_norm": 0.09228515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0389,
+ "step": 9070
+ },
+ {
+ "epoch": 28.98083067092652,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0288,
+ "step": 9071
+ },
+ {
+ "epoch": 28.984025559105433,
+ "grad_norm": 0.09033203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0327,
+ "step": 9072
+ },
+ {
+ "epoch": 28.987220447284344,
+ "grad_norm": 0.08642578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0352,
+ "step": 9073
+ },
+ {
+ "epoch": 28.990415335463258,
+ "grad_norm": 0.13671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0448,
+ "step": 9074
+ },
+ {
+ "epoch": 28.99361022364217,
+ "grad_norm": 0.055908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0332,
+ "step": 9075
+ },
+ {
+ "epoch": 28.996805111821086,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0322,
+ "step": 9076
+ },
+ {
+ "epoch": 29.0,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0366,
+ "step": 9077
+ },
+ {
+ "epoch": 29.003194888178914,
+ "grad_norm": 0.0615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0379,
+ "step": 9078
+ },
+ {
+ "epoch": 29.00638977635783,
+ "grad_norm": 0.05712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0436,
+ "step": 9079
+ },
+ {
+ "epoch": 29.009584664536742,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.044,
+ "step": 9080
+ },
+ {
+ "epoch": 29.012779552715656,
+ "grad_norm": 0.0498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0426,
+ "step": 9081
+ },
+ {
+ "epoch": 29.015974440894567,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0399,
+ "step": 9082
+ },
+ {
+ "epoch": 29.01916932907348,
+ "grad_norm": 0.06103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0348,
+ "step": 9083
+ },
+ {
+ "epoch": 29.022364217252395,
+ "grad_norm": 0.05908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0319,
+ "step": 9084
+ },
+ {
+ "epoch": 29.02555910543131,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0275,
+ "step": 9085
+ },
+ {
+ "epoch": 29.028753993610223,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0312,
+ "step": 9086
+ },
+ {
+ "epoch": 29.031948881789138,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0381,
+ "step": 9087
+ },
+ {
+ "epoch": 29.03514376996805,
+ "grad_norm": 0.0966796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0412,
+ "step": 9088
+ },
+ {
+ "epoch": 29.038338658146966,
+ "grad_norm": 0.059814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0463,
+ "step": 9089
+ },
+ {
+ "epoch": 29.04153354632588,
+ "grad_norm": 0.0869140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0394,
+ "step": 9090
+ },
+ {
+ "epoch": 29.044728434504794,
+ "grad_norm": 0.08349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.034,
+ "step": 9091
+ },
+ {
+ "epoch": 29.047923322683705,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0521,
+ "step": 9092
+ },
+ {
+ "epoch": 29.05111821086262,
+ "grad_norm": 0.08837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0416,
+ "step": 9093
+ },
+ {
+ "epoch": 29.054313099041533,
+ "grad_norm": 0.057373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0379,
+ "step": 9094
+ },
+ {
+ "epoch": 29.057507987220447,
+ "grad_norm": 0.08056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0341,
+ "step": 9095
+ },
+ {
+ "epoch": 29.06070287539936,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0449,
+ "step": 9096
+ },
+ {
+ "epoch": 29.063897763578275,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0311,
+ "step": 9097
+ },
+ {
+ "epoch": 29.06709265175719,
+ "grad_norm": 0.08837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0489,
+ "step": 9098
+ },
+ {
+ "epoch": 29.070287539936103,
+ "grad_norm": 0.0830078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0361,
+ "step": 9099
+ },
+ {
+ "epoch": 29.073482428115017,
+ "grad_norm": 0.060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0445,
+ "step": 9100
+ },
+ {
+ "epoch": 29.076677316293928,
+ "grad_norm": 0.08349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0315,
+ "step": 9101
+ },
+ {
+ "epoch": 29.079872204472842,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0312,
+ "step": 9102
+ },
+ {
+ "epoch": 29.083067092651756,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0442,
+ "step": 9103
+ },
+ {
+ "epoch": 29.08626198083067,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0379,
+ "step": 9104
+ },
+ {
+ "epoch": 29.089456869009584,
+ "grad_norm": 0.09130859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0355,
+ "step": 9105
+ },
+ {
+ "epoch": 29.0926517571885,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0402,
+ "step": 9106
+ },
+ {
+ "epoch": 29.095846645367413,
+ "grad_norm": 0.10205078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0345,
+ "step": 9107
+ },
+ {
+ "epoch": 29.099041533546327,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.041,
+ "step": 9108
+ },
+ {
+ "epoch": 29.10223642172524,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0429,
+ "step": 9109
+ },
+ {
+ "epoch": 29.105431309904155,
+ "grad_norm": 0.146484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0463,
+ "step": 9110
+ },
+ {
+ "epoch": 29.108626198083066,
+ "grad_norm": 0.060302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0484,
+ "step": 9111
+ },
+ {
+ "epoch": 29.11182108626198,
+ "grad_norm": 0.16015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0465,
+ "step": 9112
+ },
+ {
+ "epoch": 29.115015974440894,
+ "grad_norm": 0.055419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0437,
+ "step": 9113
+ },
+ {
+ "epoch": 29.118210862619808,
+ "grad_norm": 0.134765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0359,
+ "step": 9114
+ },
+ {
+ "epoch": 29.121405750798722,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0269,
+ "step": 9115
+ },
+ {
+ "epoch": 29.124600638977636,
+ "grad_norm": 0.12890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0349,
+ "step": 9116
+ },
+ {
+ "epoch": 29.12779552715655,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0412,
+ "step": 9117
+ },
+ {
+ "epoch": 29.130990415335464,
+ "grad_norm": 0.107421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0341,
+ "step": 9118
+ },
+ {
+ "epoch": 29.13418530351438,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0418,
+ "step": 9119
+ },
+ {
+ "epoch": 29.13738019169329,
+ "grad_norm": 0.130859375,
+ "learning_rate": 0.0005,
+ "loss": 1.04,
+ "step": 9120
+ },
+ {
+ "epoch": 29.140575079872203,
+ "grad_norm": 0.0546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0415,
+ "step": 9121
+ },
+ {
+ "epoch": 29.143769968051117,
+ "grad_norm": 0.12890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0396,
+ "step": 9122
+ },
+ {
+ "epoch": 29.14696485623003,
+ "grad_norm": 0.048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0475,
+ "step": 9123
+ },
+ {
+ "epoch": 29.150159744408946,
+ "grad_norm": 0.1572265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0421,
+ "step": 9124
+ },
+ {
+ "epoch": 29.15335463258786,
+ "grad_norm": 0.07861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0381,
+ "step": 9125
+ },
+ {
+ "epoch": 29.156549520766774,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0501,
+ "step": 9126
+ },
+ {
+ "epoch": 29.159744408945688,
+ "grad_norm": 0.0537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0268,
+ "step": 9127
+ },
+ {
+ "epoch": 29.162939297124602,
+ "grad_norm": 0.05615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0315,
+ "step": 9128
+ },
+ {
+ "epoch": 29.166134185303516,
+ "grad_norm": 0.0498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0411,
+ "step": 9129
+ },
+ {
+ "epoch": 29.169329073482427,
+ "grad_norm": 0.046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0515,
+ "step": 9130
+ },
+ {
+ "epoch": 29.17252396166134,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0286,
+ "step": 9131
+ },
+ {
+ "epoch": 29.175718849840255,
+ "grad_norm": 0.05126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0452,
+ "step": 9132
+ },
+ {
+ "epoch": 29.17891373801917,
+ "grad_norm": 0.062255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0365,
+ "step": 9133
+ },
+ {
+ "epoch": 29.182108626198083,
+ "grad_norm": 0.0546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0449,
+ "step": 9134
+ },
+ {
+ "epoch": 29.185303514376997,
+ "grad_norm": 0.056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0351,
+ "step": 9135
+ },
+ {
+ "epoch": 29.18849840255591,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0358,
+ "step": 9136
+ },
+ {
+ "epoch": 29.191693290734825,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0374,
+ "step": 9137
+ },
+ {
+ "epoch": 29.19488817891374,
+ "grad_norm": 0.046142578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0308,
+ "step": 9138
+ },
+ {
+ "epoch": 29.19808306709265,
+ "grad_norm": 0.052490234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0347,
+ "step": 9139
+ },
+ {
+ "epoch": 29.201277955271564,
+ "grad_norm": 0.045654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0243,
+ "step": 9140
+ },
+ {
+ "epoch": 29.20447284345048,
+ "grad_norm": 0.0478515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0479,
+ "step": 9141
+ },
+ {
+ "epoch": 29.207667731629392,
+ "grad_norm": 0.051513671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0357,
+ "step": 9142
+ },
+ {
+ "epoch": 29.210862619808307,
+ "grad_norm": 0.049560546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0342,
+ "step": 9143
+ },
+ {
+ "epoch": 29.21405750798722,
+ "grad_norm": 0.049072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0477,
+ "step": 9144
+ },
+ {
+ "epoch": 29.217252396166135,
+ "grad_norm": 0.058349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0316,
+ "step": 9145
+ },
+ {
+ "epoch": 29.22044728434505,
+ "grad_norm": 0.049560546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0433,
+ "step": 9146
+ },
+ {
+ "epoch": 29.223642172523963,
+ "grad_norm": 0.06005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.042,
+ "step": 9147
+ },
+ {
+ "epoch": 29.226837060702877,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0406,
+ "step": 9148
+ },
+ {
+ "epoch": 29.230031948881788,
+ "grad_norm": 0.08251953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0388,
+ "step": 9149
+ },
+ {
+ "epoch": 29.233226837060702,
+ "grad_norm": 0.05712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0403,
+ "step": 9150
+ },
+ {
+ "epoch": 29.236421725239616,
+ "grad_norm": 0.1279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0336,
+ "step": 9151
+ },
+ {
+ "epoch": 29.23961661341853,
+ "grad_norm": 0.05615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.037,
+ "step": 9152
+ },
+ {
+ "epoch": 29.242811501597444,
+ "grad_norm": 0.1318359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0382,
+ "step": 9153
+ },
+ {
+ "epoch": 29.24600638977636,
+ "grad_norm": 0.061767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0392,
+ "step": 9154
+ },
+ {
+ "epoch": 29.249201277955272,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0343,
+ "step": 9155
+ },
+ {
+ "epoch": 29.252396166134186,
+ "grad_norm": 0.058837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0465,
+ "step": 9156
+ },
+ {
+ "epoch": 29.2555910543131,
+ "grad_norm": 0.061279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.046,
+ "step": 9157
+ },
+ {
+ "epoch": 29.25878594249201,
+ "grad_norm": 0.0556640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0299,
+ "step": 9158
+ },
+ {
+ "epoch": 29.261980830670925,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0396,
+ "step": 9159
+ },
+ {
+ "epoch": 29.26517571884984,
+ "grad_norm": 0.057373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0327,
+ "step": 9160
+ },
+ {
+ "epoch": 29.268370607028753,
+ "grad_norm": 0.052978515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0201,
+ "step": 9161
+ },
+ {
+ "epoch": 29.271565495207668,
+ "grad_norm": 0.05615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0396,
+ "step": 9162
+ },
+ {
+ "epoch": 29.27476038338658,
+ "grad_norm": 0.052978515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0388,
+ "step": 9163
+ },
+ {
+ "epoch": 29.277955271565496,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0527,
+ "step": 9164
+ },
+ {
+ "epoch": 29.28115015974441,
+ "grad_norm": 0.05615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0363,
+ "step": 9165
+ },
+ {
+ "epoch": 29.284345047923324,
+ "grad_norm": 0.05078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0378,
+ "step": 9166
+ },
+ {
+ "epoch": 29.287539936102238,
+ "grad_norm": 0.06201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0355,
+ "step": 9167
+ },
+ {
+ "epoch": 29.29073482428115,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0461,
+ "step": 9168
+ },
+ {
+ "epoch": 29.293929712460063,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0408,
+ "step": 9169
+ },
+ {
+ "epoch": 29.297124600638977,
+ "grad_norm": 0.05908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0413,
+ "step": 9170
+ },
+ {
+ "epoch": 29.30031948881789,
+ "grad_norm": 0.0615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0405,
+ "step": 9171
+ },
+ {
+ "epoch": 29.303514376996805,
+ "grad_norm": 0.0595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0373,
+ "step": 9172
+ },
+ {
+ "epoch": 29.30670926517572,
+ "grad_norm": 0.054931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0371,
+ "step": 9173
+ },
+ {
+ "epoch": 29.309904153354633,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0401,
+ "step": 9174
+ },
+ {
+ "epoch": 29.313099041533548,
+ "grad_norm": 0.05859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0366,
+ "step": 9175
+ },
+ {
+ "epoch": 29.31629392971246,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0494,
+ "step": 9176
+ },
+ {
+ "epoch": 29.319488817891372,
+ "grad_norm": 0.0537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0277,
+ "step": 9177
+ },
+ {
+ "epoch": 29.322683706070286,
+ "grad_norm": 0.08203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0325,
+ "step": 9178
+ },
+ {
+ "epoch": 29.3258785942492,
+ "grad_norm": 0.060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0378,
+ "step": 9179
+ },
+ {
+ "epoch": 29.329073482428115,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0419,
+ "step": 9180
+ },
+ {
+ "epoch": 29.33226837060703,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0515,
+ "step": 9181
+ },
+ {
+ "epoch": 29.335463258785943,
+ "grad_norm": 0.09228515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0339,
+ "step": 9182
+ },
+ {
+ "epoch": 29.338658146964857,
+ "grad_norm": 0.053466796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0443,
+ "step": 9183
+ },
+ {
+ "epoch": 29.34185303514377,
+ "grad_norm": 0.057373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0356,
+ "step": 9184
+ },
+ {
+ "epoch": 29.345047923322685,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0428,
+ "step": 9185
+ },
+ {
+ "epoch": 29.3482428115016,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0358,
+ "step": 9186
+ },
+ {
+ "epoch": 29.35143769968051,
+ "grad_norm": 0.056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0358,
+ "step": 9187
+ },
+ {
+ "epoch": 29.354632587859424,
+ "grad_norm": 0.052490234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0259,
+ "step": 9188
+ },
+ {
+ "epoch": 29.357827476038338,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0374,
+ "step": 9189
+ },
+ {
+ "epoch": 29.361022364217252,
+ "grad_norm": 0.055419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.052,
+ "step": 9190
+ },
+ {
+ "epoch": 29.364217252396166,
+ "grad_norm": 0.0537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0288,
+ "step": 9191
+ },
+ {
+ "epoch": 29.36741214057508,
+ "grad_norm": 0.054443359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0401,
+ "step": 9192
+ },
+ {
+ "epoch": 29.370607028753994,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.046,
+ "step": 9193
+ },
+ {
+ "epoch": 29.37380191693291,
+ "grad_norm": 0.058349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0466,
+ "step": 9194
+ },
+ {
+ "epoch": 29.376996805111823,
+ "grad_norm": 0.10546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0496,
+ "step": 9195
+ },
+ {
+ "epoch": 29.380191693290733,
+ "grad_norm": 0.04931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0417,
+ "step": 9196
+ },
+ {
+ "epoch": 29.383386581469647,
+ "grad_norm": 0.0517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0337,
+ "step": 9197
+ },
+ {
+ "epoch": 29.38658146964856,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0313,
+ "step": 9198
+ },
+ {
+ "epoch": 29.389776357827476,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.045,
+ "step": 9199
+ },
+ {
+ "epoch": 29.39297124600639,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0393,
+ "step": 9200
+ },
+ {
+ "epoch": 29.396166134185304,
+ "grad_norm": 0.054931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0401,
+ "step": 9201
+ },
+ {
+ "epoch": 29.399361022364218,
+ "grad_norm": 0.060302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0261,
+ "step": 9202
+ },
+ {
+ "epoch": 29.402555910543132,
+ "grad_norm": 0.047607421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0439,
+ "step": 9203
+ },
+ {
+ "epoch": 29.405750798722046,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0349,
+ "step": 9204
+ },
+ {
+ "epoch": 29.408945686900957,
+ "grad_norm": 0.05078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0491,
+ "step": 9205
+ },
+ {
+ "epoch": 29.41214057507987,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.039,
+ "step": 9206
+ },
+ {
+ "epoch": 29.415335463258785,
+ "grad_norm": 0.052978515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0268,
+ "step": 9207
+ },
+ {
+ "epoch": 29.4185303514377,
+ "grad_norm": 0.080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0421,
+ "step": 9208
+ },
+ {
+ "epoch": 29.421725239616613,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0303,
+ "step": 9209
+ },
+ {
+ "epoch": 29.424920127795527,
+ "grad_norm": 0.0927734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0312,
+ "step": 9210
+ },
+ {
+ "epoch": 29.42811501597444,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0359,
+ "step": 9211
+ },
+ {
+ "epoch": 29.431309904153355,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0331,
+ "step": 9212
+ },
+ {
+ "epoch": 29.43450479233227,
+ "grad_norm": 0.050048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0431,
+ "step": 9213
+ },
+ {
+ "epoch": 29.437699680511184,
+ "grad_norm": 0.08740234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0385,
+ "step": 9214
+ },
+ {
+ "epoch": 29.440894568690094,
+ "grad_norm": 0.08935546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0304,
+ "step": 9215
+ },
+ {
+ "epoch": 29.44408945686901,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0429,
+ "step": 9216
+ },
+ {
+ "epoch": 29.447284345047922,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0456,
+ "step": 9217
+ },
+ {
+ "epoch": 29.450479233226837,
+ "grad_norm": 0.0986328125,
+ "learning_rate": 0.0005,
+ "loss": 1.038,
+ "step": 9218
+ },
+ {
+ "epoch": 29.45367412140575,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0262,
+ "step": 9219
+ },
+ {
+ "epoch": 29.456869009584665,
+ "grad_norm": 0.08203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0284,
+ "step": 9220
+ },
+ {
+ "epoch": 29.46006389776358,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0355,
+ "step": 9221
+ },
+ {
+ "epoch": 29.463258785942493,
+ "grad_norm": 0.130859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0486,
+ "step": 9222
+ },
+ {
+ "epoch": 29.466453674121407,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0439,
+ "step": 9223
+ },
+ {
+ "epoch": 29.46964856230032,
+ "grad_norm": 0.095703125,
+ "learning_rate": 0.0005,
+ "loss": 1.04,
+ "step": 9224
+ },
+ {
+ "epoch": 29.472843450479232,
+ "grad_norm": 0.080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0264,
+ "step": 9225
+ },
+ {
+ "epoch": 29.476038338658146,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0476,
+ "step": 9226
+ },
+ {
+ "epoch": 29.47923322683706,
+ "grad_norm": 0.05810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0421,
+ "step": 9227
+ },
+ {
+ "epoch": 29.482428115015974,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0552,
+ "step": 9228
+ },
+ {
+ "epoch": 29.48562300319489,
+ "grad_norm": 0.0556640625,
+ "learning_rate": 0.0005,
+ "loss": 1.025,
+ "step": 9229
+ },
+ {
+ "epoch": 29.488817891373802,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0324,
+ "step": 9230
+ },
+ {
+ "epoch": 29.492012779552716,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0381,
+ "step": 9231
+ },
+ {
+ "epoch": 29.49520766773163,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0377,
+ "step": 9232
+ },
+ {
+ "epoch": 29.498402555910545,
+ "grad_norm": 0.05517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0308,
+ "step": 9233
+ },
+ {
+ "epoch": 29.501597444089455,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0424,
+ "step": 9234
+ },
+ {
+ "epoch": 29.50479233226837,
+ "grad_norm": 0.049560546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0405,
+ "step": 9235
+ },
+ {
+ "epoch": 29.507987220447284,
+ "grad_norm": 0.091796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0419,
+ "step": 9236
+ },
+ {
+ "epoch": 29.511182108626198,
+ "grad_norm": 0.0830078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0278,
+ "step": 9237
+ },
+ {
+ "epoch": 29.51437699680511,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0356,
+ "step": 9238
+ },
+ {
+ "epoch": 29.517571884984026,
+ "grad_norm": 0.048583984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0422,
+ "step": 9239
+ },
+ {
+ "epoch": 29.52076677316294,
+ "grad_norm": 0.1474609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0395,
+ "step": 9240
+ },
+ {
+ "epoch": 29.523961661341854,
+ "grad_norm": 0.0517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0426,
+ "step": 9241
+ },
+ {
+ "epoch": 29.527156549520768,
+ "grad_norm": 0.10546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0354,
+ "step": 9242
+ },
+ {
+ "epoch": 29.53035143769968,
+ "grad_norm": 0.1025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0325,
+ "step": 9243
+ },
+ {
+ "epoch": 29.533546325878593,
+ "grad_norm": 0.1044921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0328,
+ "step": 9244
+ },
+ {
+ "epoch": 29.536741214057507,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0404,
+ "step": 9245
+ },
+ {
+ "epoch": 29.53993610223642,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0437,
+ "step": 9246
+ },
+ {
+ "epoch": 29.543130990415335,
+ "grad_norm": 0.08935546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0401,
+ "step": 9247
+ },
+ {
+ "epoch": 29.54632587859425,
+ "grad_norm": 0.0849609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0431,
+ "step": 9248
+ },
+ {
+ "epoch": 29.549520766773163,
+ "grad_norm": 0.06005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0395,
+ "step": 9249
+ },
+ {
+ "epoch": 29.552715654952078,
+ "grad_norm": 0.05029296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0401,
+ "step": 9250
+ },
+ {
+ "epoch": 29.55591054313099,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0319,
+ "step": 9251
+ },
+ {
+ "epoch": 29.559105431309906,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0413,
+ "step": 9252
+ },
+ {
+ "epoch": 29.562300319488816,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0421,
+ "step": 9253
+ },
+ {
+ "epoch": 29.56549520766773,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0557,
+ "step": 9254
+ },
+ {
+ "epoch": 29.568690095846645,
+ "grad_norm": 0.08056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0411,
+ "step": 9255
+ },
+ {
+ "epoch": 29.57188498402556,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0338,
+ "step": 9256
+ },
+ {
+ "epoch": 29.575079872204473,
+ "grad_norm": 0.10400390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0466,
+ "step": 9257
+ },
+ {
+ "epoch": 29.578274760383387,
+ "grad_norm": 0.12158203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0364,
+ "step": 9258
+ },
+ {
+ "epoch": 29.5814696485623,
+ "grad_norm": 0.22265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0421,
+ "step": 9259
+ },
+ {
+ "epoch": 29.584664536741215,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0339,
+ "step": 9260
+ },
+ {
+ "epoch": 29.58785942492013,
+ "grad_norm": 0.130859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0261,
+ "step": 9261
+ },
+ {
+ "epoch": 29.591054313099043,
+ "grad_norm": 0.057373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0343,
+ "step": 9262
+ },
+ {
+ "epoch": 29.594249201277954,
+ "grad_norm": 0.0810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0416,
+ "step": 9263
+ },
+ {
+ "epoch": 29.597444089456868,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0251,
+ "step": 9264
+ },
+ {
+ "epoch": 29.600638977635782,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0428,
+ "step": 9265
+ },
+ {
+ "epoch": 29.603833865814696,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0365,
+ "step": 9266
+ },
+ {
+ "epoch": 29.60702875399361,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0366,
+ "step": 9267
+ },
+ {
+ "epoch": 29.610223642172524,
+ "grad_norm": 0.056884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0345,
+ "step": 9268
+ },
+ {
+ "epoch": 29.61341853035144,
+ "grad_norm": 0.08447265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0448,
+ "step": 9269
+ },
+ {
+ "epoch": 29.616613418530353,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0347,
+ "step": 9270
+ },
+ {
+ "epoch": 29.619808306709267,
+ "grad_norm": 0.10302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.042,
+ "step": 9271
+ },
+ {
+ "epoch": 29.623003194888177,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0424,
+ "step": 9272
+ },
+ {
+ "epoch": 29.62619808306709,
+ "grad_norm": 0.08203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0322,
+ "step": 9273
+ },
+ {
+ "epoch": 29.629392971246006,
+ "grad_norm": 0.051513671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0381,
+ "step": 9274
+ },
+ {
+ "epoch": 29.63258785942492,
+ "grad_norm": 0.091796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0416,
+ "step": 9275
+ },
+ {
+ "epoch": 29.635782747603834,
+ "grad_norm": 0.050048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0527,
+ "step": 9276
+ },
+ {
+ "epoch": 29.638977635782748,
+ "grad_norm": 0.08740234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0427,
+ "step": 9277
+ },
+ {
+ "epoch": 29.642172523961662,
+ "grad_norm": 0.046142578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0423,
+ "step": 9278
+ },
+ {
+ "epoch": 29.645367412140576,
+ "grad_norm": 0.08935546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0352,
+ "step": 9279
+ },
+ {
+ "epoch": 29.64856230031949,
+ "grad_norm": 0.052734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0354,
+ "step": 9280
+ },
+ {
+ "epoch": 29.6517571884984,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0375,
+ "step": 9281
+ },
+ {
+ "epoch": 29.654952076677315,
+ "grad_norm": 0.046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0332,
+ "step": 9282
+ },
+ {
+ "epoch": 29.65814696485623,
+ "grad_norm": 0.05224609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0312,
+ "step": 9283
+ },
+ {
+ "epoch": 29.661341853035143,
+ "grad_norm": 0.05029296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0358,
+ "step": 9284
+ },
+ {
+ "epoch": 29.664536741214057,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0283,
+ "step": 9285
+ },
+ {
+ "epoch": 29.66773162939297,
+ "grad_norm": 0.051513671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0389,
+ "step": 9286
+ },
+ {
+ "epoch": 29.670926517571885,
+ "grad_norm": 0.052490234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0315,
+ "step": 9287
+ },
+ {
+ "epoch": 29.6741214057508,
+ "grad_norm": 0.056396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.037,
+ "step": 9288
+ },
+ {
+ "epoch": 29.677316293929714,
+ "grad_norm": 0.058349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0331,
+ "step": 9289
+ },
+ {
+ "epoch": 29.680511182108628,
+ "grad_norm": 0.04443359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0357,
+ "step": 9290
+ },
+ {
+ "epoch": 29.68370607028754,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0409,
+ "step": 9291
+ },
+ {
+ "epoch": 29.686900958466452,
+ "grad_norm": 0.060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0319,
+ "step": 9292
+ },
+ {
+ "epoch": 29.690095846645367,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0435,
+ "step": 9293
+ },
+ {
+ "epoch": 29.69329073482428,
+ "grad_norm": 0.0556640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0404,
+ "step": 9294
+ },
+ {
+ "epoch": 29.696485623003195,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0444,
+ "step": 9295
+ },
+ {
+ "epoch": 29.69968051118211,
+ "grad_norm": 0.05322265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0437,
+ "step": 9296
+ },
+ {
+ "epoch": 29.702875399361023,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0466,
+ "step": 9297
+ },
+ {
+ "epoch": 29.706070287539937,
+ "grad_norm": 0.053466796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0323,
+ "step": 9298
+ },
+ {
+ "epoch": 29.70926517571885,
+ "grad_norm": 0.060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0312,
+ "step": 9299
+ },
+ {
+ "epoch": 29.712460063897765,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0447,
+ "step": 9300
+ },
+ {
+ "epoch": 29.715654952076676,
+ "grad_norm": 0.061767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.045,
+ "step": 9301
+ },
+ {
+ "epoch": 29.71884984025559,
+ "grad_norm": 0.08837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0307,
+ "step": 9302
+ },
+ {
+ "epoch": 29.722044728434504,
+ "grad_norm": 0.0556640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0299,
+ "step": 9303
+ },
+ {
+ "epoch": 29.72523961661342,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0267,
+ "step": 9304
+ },
+ {
+ "epoch": 29.728434504792332,
+ "grad_norm": 0.09423828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0203,
+ "step": 9305
+ },
+ {
+ "epoch": 29.731629392971247,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0246,
+ "step": 9306
+ },
+ {
+ "epoch": 29.73482428115016,
+ "grad_norm": 0.05615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0379,
+ "step": 9307
+ },
+ {
+ "epoch": 29.738019169329075,
+ "grad_norm": 0.0927734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0382,
+ "step": 9308
+ },
+ {
+ "epoch": 29.74121405750799,
+ "grad_norm": 0.058349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.03,
+ "step": 9309
+ },
+ {
+ "epoch": 29.7444089456869,
+ "grad_norm": 0.083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0414,
+ "step": 9310
+ },
+ {
+ "epoch": 29.747603833865814,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0155,
+ "step": 9311
+ },
+ {
+ "epoch": 29.750798722044728,
+ "grad_norm": 0.0615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.038,
+ "step": 9312
+ },
+ {
+ "epoch": 29.75399361022364,
+ "grad_norm": 0.08251953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0427,
+ "step": 9313
+ },
+ {
+ "epoch": 29.757188498402556,
+ "grad_norm": 0.05078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0483,
+ "step": 9314
+ },
+ {
+ "epoch": 29.76038338658147,
+ "grad_norm": 0.05908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0495,
+ "step": 9315
+ },
+ {
+ "epoch": 29.763578274760384,
+ "grad_norm": 0.0537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0293,
+ "step": 9316
+ },
+ {
+ "epoch": 29.766773162939298,
+ "grad_norm": 0.0810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0364,
+ "step": 9317
+ },
+ {
+ "epoch": 29.769968051118212,
+ "grad_norm": 0.050537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0409,
+ "step": 9318
+ },
+ {
+ "epoch": 29.773162939297123,
+ "grad_norm": 0.08984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0248,
+ "step": 9319
+ },
+ {
+ "epoch": 29.776357827476037,
+ "grad_norm": 0.08935546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0529,
+ "step": 9320
+ },
+ {
+ "epoch": 29.77955271565495,
+ "grad_norm": 0.109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0363,
+ "step": 9321
+ },
+ {
+ "epoch": 29.782747603833865,
+ "grad_norm": 0.08642578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0478,
+ "step": 9322
+ },
+ {
+ "epoch": 29.78594249201278,
+ "grad_norm": 0.0888671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0326,
+ "step": 9323
+ },
+ {
+ "epoch": 29.789137380191693,
+ "grad_norm": 0.046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0409,
+ "step": 9324
+ },
+ {
+ "epoch": 29.792332268370608,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.034,
+ "step": 9325
+ },
+ {
+ "epoch": 29.79552715654952,
+ "grad_norm": 0.046630859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0344,
+ "step": 9326
+ },
+ {
+ "epoch": 29.798722044728436,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0339,
+ "step": 9327
+ },
+ {
+ "epoch": 29.80191693290735,
+ "grad_norm": 0.059326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.033,
+ "step": 9328
+ },
+ {
+ "epoch": 29.80511182108626,
+ "grad_norm": 0.051025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0376,
+ "step": 9329
+ },
+ {
+ "epoch": 29.808306709265175,
+ "grad_norm": 0.0810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0369,
+ "step": 9330
+ },
+ {
+ "epoch": 29.81150159744409,
+ "grad_norm": 0.04931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0387,
+ "step": 9331
+ },
+ {
+ "epoch": 29.814696485623003,
+ "grad_norm": 0.053466796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0444,
+ "step": 9332
+ },
+ {
+ "epoch": 29.817891373801917,
+ "grad_norm": 0.04736328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0454,
+ "step": 9333
+ },
+ {
+ "epoch": 29.82108626198083,
+ "grad_norm": 0.05712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0351,
+ "step": 9334
+ },
+ {
+ "epoch": 29.824281150159745,
+ "grad_norm": 0.057373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0349,
+ "step": 9335
+ },
+ {
+ "epoch": 29.82747603833866,
+ "grad_norm": 0.046142578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0338,
+ "step": 9336
+ },
+ {
+ "epoch": 29.830670926517573,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0258,
+ "step": 9337
+ },
+ {
+ "epoch": 29.833865814696484,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0341,
+ "step": 9338
+ },
+ {
+ "epoch": 29.837060702875398,
+ "grad_norm": 0.06201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.041,
+ "step": 9339
+ },
+ {
+ "epoch": 29.840255591054312,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0479,
+ "step": 9340
+ },
+ {
+ "epoch": 29.843450479233226,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0267,
+ "step": 9341
+ },
+ {
+ "epoch": 29.84664536741214,
+ "grad_norm": 0.057861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0428,
+ "step": 9342
+ },
+ {
+ "epoch": 29.849840255591054,
+ "grad_norm": 0.08935546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0449,
+ "step": 9343
+ },
+ {
+ "epoch": 29.85303514376997,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0412,
+ "step": 9344
+ },
+ {
+ "epoch": 29.856230031948883,
+ "grad_norm": 0.059814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0298,
+ "step": 9345
+ },
+ {
+ "epoch": 29.859424920127797,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0386,
+ "step": 9346
+ },
+ {
+ "epoch": 29.86261980830671,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0332,
+ "step": 9347
+ },
+ {
+ "epoch": 29.86581469648562,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0308,
+ "step": 9348
+ },
+ {
+ "epoch": 29.869009584664536,
+ "grad_norm": 0.0595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0362,
+ "step": 9349
+ },
+ {
+ "epoch": 29.87220447284345,
+ "grad_norm": 0.0947265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0351,
+ "step": 9350
+ },
+ {
+ "epoch": 29.875399361022364,
+ "grad_norm": 0.0556640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0351,
+ "step": 9351
+ },
+ {
+ "epoch": 29.878594249201278,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0353,
+ "step": 9352
+ },
+ {
+ "epoch": 29.881789137380192,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0384,
+ "step": 9353
+ },
+ {
+ "epoch": 29.884984025559106,
+ "grad_norm": 0.059326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0395,
+ "step": 9354
+ },
+ {
+ "epoch": 29.88817891373802,
+ "grad_norm": 0.062255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0426,
+ "step": 9355
+ },
+ {
+ "epoch": 29.891373801916934,
+ "grad_norm": 0.05419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0482,
+ "step": 9356
+ },
+ {
+ "epoch": 29.894568690095845,
+ "grad_norm": 0.09716796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0283,
+ "step": 9357
+ },
+ {
+ "epoch": 29.89776357827476,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0325,
+ "step": 9358
+ },
+ {
+ "epoch": 29.900958466453673,
+ "grad_norm": 0.0615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0271,
+ "step": 9359
+ },
+ {
+ "epoch": 29.904153354632587,
+ "grad_norm": 0.051025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0355,
+ "step": 9360
+ },
+ {
+ "epoch": 29.9073482428115,
+ "grad_norm": 0.054443359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0293,
+ "step": 9361
+ },
+ {
+ "epoch": 29.910543130990416,
+ "grad_norm": 0.059326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.038,
+ "step": 9362
+ },
+ {
+ "epoch": 29.91373801916933,
+ "grad_norm": 0.0546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0493,
+ "step": 9363
+ },
+ {
+ "epoch": 29.916932907348244,
+ "grad_norm": 0.060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0463,
+ "step": 9364
+ },
+ {
+ "epoch": 29.920127795527158,
+ "grad_norm": 0.1015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0332,
+ "step": 9365
+ },
+ {
+ "epoch": 29.923322683706072,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.053,
+ "step": 9366
+ },
+ {
+ "epoch": 29.926517571884983,
+ "grad_norm": 0.11083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0302,
+ "step": 9367
+ },
+ {
+ "epoch": 29.929712460063897,
+ "grad_norm": 0.0595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0344,
+ "step": 9368
+ },
+ {
+ "epoch": 29.93290734824281,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.046,
+ "step": 9369
+ },
+ {
+ "epoch": 29.936102236421725,
+ "grad_norm": 0.080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0455,
+ "step": 9370
+ },
+ {
+ "epoch": 29.93929712460064,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0423,
+ "step": 9371
+ },
+ {
+ "epoch": 29.942492012779553,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.049,
+ "step": 9372
+ },
+ {
+ "epoch": 29.945686900958467,
+ "grad_norm": 0.052490234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0354,
+ "step": 9373
+ },
+ {
+ "epoch": 29.94888178913738,
+ "grad_norm": 0.057373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0327,
+ "step": 9374
+ },
+ {
+ "epoch": 29.952076677316295,
+ "grad_norm": 0.057861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0378,
+ "step": 9375
+ },
+ {
+ "epoch": 29.955271565495206,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.033,
+ "step": 9376
+ },
+ {
+ "epoch": 29.95846645367412,
+ "grad_norm": 0.051025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0354,
+ "step": 9377
+ },
+ {
+ "epoch": 29.961661341853034,
+ "grad_norm": 0.053955078125,
+ "learning_rate": 0.0005,
+ "loss": 1.033,
+ "step": 9378
+ },
+ {
+ "epoch": 29.96485623003195,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0407,
+ "step": 9379
+ },
+ {
+ "epoch": 29.968051118210862,
+ "grad_norm": 0.08837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0366,
+ "step": 9380
+ },
+ {
+ "epoch": 29.971246006389777,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0304,
+ "step": 9381
+ },
+ {
+ "epoch": 29.97444089456869,
+ "grad_norm": 0.09814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0576,
+ "step": 9382
+ },
+ {
+ "epoch": 29.977635782747605,
+ "grad_norm": 0.049072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0349,
+ "step": 9383
+ },
+ {
+ "epoch": 29.98083067092652,
+ "grad_norm": 0.087890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0481,
+ "step": 9384
+ },
+ {
+ "epoch": 29.984025559105433,
+ "grad_norm": 0.0927734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0344,
+ "step": 9385
+ },
+ {
+ "epoch": 29.987220447284344,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0359,
+ "step": 9386
+ },
+ {
+ "epoch": 29.990415335463258,
+ "grad_norm": 0.09765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0448,
+ "step": 9387
+ },
+ {
+ "epoch": 29.99361022364217,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0464,
+ "step": 9388
+ },
+ {
+ "epoch": 29.996805111821086,
+ "grad_norm": 0.054443359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0522,
+ "step": 9389
+ },
+ {
+ "epoch": 30.0,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0545,
+ "step": 9390
+ },
+ {
+ "epoch": 30.003194888178914,
+ "grad_norm": 0.048583984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0338,
+ "step": 9391
+ },
+ {
+ "epoch": 30.00638977635783,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0432,
+ "step": 9392
+ },
+ {
+ "epoch": 30.009584664536742,
+ "grad_norm": 0.062255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0283,
+ "step": 9393
+ },
+ {
+ "epoch": 30.012779552715656,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0375,
+ "step": 9394
+ },
+ {
+ "epoch": 30.015974440894567,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0248,
+ "step": 9395
+ },
+ {
+ "epoch": 30.01916932907348,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0302,
+ "step": 9396
+ },
+ {
+ "epoch": 30.022364217252395,
+ "grad_norm": 0.10107421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0438,
+ "step": 9397
+ },
+ {
+ "epoch": 30.02555910543131,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0307,
+ "step": 9398
+ },
+ {
+ "epoch": 30.028753993610223,
+ "grad_norm": 0.09619140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0485,
+ "step": 9399
+ },
+ {
+ "epoch": 30.031948881789138,
+ "grad_norm": 0.057861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0393,
+ "step": 9400
+ },
+ {
+ "epoch": 30.03514376996805,
+ "grad_norm": 0.061767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0492,
+ "step": 9401
+ },
+ {
+ "epoch": 30.038338658146966,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0192,
+ "step": 9402
+ },
+ {
+ "epoch": 30.04153354632588,
+ "grad_norm": 0.061767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0453,
+ "step": 9403
+ },
+ {
+ "epoch": 30.044728434504794,
+ "grad_norm": 0.08154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0369,
+ "step": 9404
+ },
+ {
+ "epoch": 30.047923322683705,
+ "grad_norm": 0.05029296875,
+ "learning_rate": 0.0005,
+ "loss": 1.035,
+ "step": 9405
+ },
+ {
+ "epoch": 30.05111821086262,
+ "grad_norm": 0.10693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0422,
+ "step": 9406
+ },
+ {
+ "epoch": 30.054313099041533,
+ "grad_norm": 0.05712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0366,
+ "step": 9407
+ },
+ {
+ "epoch": 30.057507987220447,
+ "grad_norm": 0.058837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0292,
+ "step": 9408
+ },
+ {
+ "epoch": 30.06070287539936,
+ "grad_norm": 0.058349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0331,
+ "step": 9409
+ },
+ {
+ "epoch": 30.063897763578275,
+ "grad_norm": 0.05859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0401,
+ "step": 9410
+ },
+ {
+ "epoch": 30.06709265175719,
+ "grad_norm": 0.058837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0465,
+ "step": 9411
+ },
+ {
+ "epoch": 30.070287539936103,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0258,
+ "step": 9412
+ },
+ {
+ "epoch": 30.073482428115017,
+ "grad_norm": 0.053955078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0373,
+ "step": 9413
+ },
+ {
+ "epoch": 30.076677316293928,
+ "grad_norm": 0.05712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0434,
+ "step": 9414
+ },
+ {
+ "epoch": 30.079872204472842,
+ "grad_norm": 0.052490234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0385,
+ "step": 9415
+ },
+ {
+ "epoch": 30.083067092651756,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0383,
+ "step": 9416
+ },
+ {
+ "epoch": 30.08626198083067,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0426,
+ "step": 9417
+ },
+ {
+ "epoch": 30.089456869009584,
+ "grad_norm": 0.059814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0521,
+ "step": 9418
+ },
+ {
+ "epoch": 30.0926517571885,
+ "grad_norm": 0.062255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.053,
+ "step": 9419
+ },
+ {
+ "epoch": 30.095846645367413,
+ "grad_norm": 0.0595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0353,
+ "step": 9420
+ },
+ {
+ "epoch": 30.099041533546327,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0449,
+ "step": 9421
+ },
+ {
+ "epoch": 30.10223642172524,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0435,
+ "step": 9422
+ },
+ {
+ "epoch": 30.105431309904155,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0427,
+ "step": 9423
+ },
+ {
+ "epoch": 30.108626198083066,
+ "grad_norm": 0.0478515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0455,
+ "step": 9424
+ },
+ {
+ "epoch": 30.11182108626198,
+ "grad_norm": 0.052490234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0373,
+ "step": 9425
+ },
+ {
+ "epoch": 30.115015974440894,
+ "grad_norm": 0.060791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0359,
+ "step": 9426
+ },
+ {
+ "epoch": 30.118210862619808,
+ "grad_norm": 0.061279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0406,
+ "step": 9427
+ },
+ {
+ "epoch": 30.121405750798722,
+ "grad_norm": 0.06103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0492,
+ "step": 9428
+ },
+ {
+ "epoch": 30.124600638977636,
+ "grad_norm": 0.0927734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0394,
+ "step": 9429
+ },
+ {
+ "epoch": 30.12779552715655,
+ "grad_norm": 0.045166015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0439,
+ "step": 9430
+ },
+ {
+ "epoch": 30.130990415335464,
+ "grad_norm": 0.05859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0392,
+ "step": 9431
+ },
+ {
+ "epoch": 30.13418530351438,
+ "grad_norm": 0.05078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0342,
+ "step": 9432
+ },
+ {
+ "epoch": 30.13738019169329,
+ "grad_norm": 0.08984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0321,
+ "step": 9433
+ },
+ {
+ "epoch": 30.140575079872203,
+ "grad_norm": 0.051513671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0357,
+ "step": 9434
+ },
+ {
+ "epoch": 30.143769968051117,
+ "grad_norm": 0.0869140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0365,
+ "step": 9435
+ },
+ {
+ "epoch": 30.14696485623003,
+ "grad_norm": 0.046630859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0552,
+ "step": 9436
+ },
+ {
+ "epoch": 30.150159744408946,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0398,
+ "step": 9437
+ },
+ {
+ "epoch": 30.15335463258786,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0294,
+ "step": 9438
+ },
+ {
+ "epoch": 30.156549520766774,
+ "grad_norm": 0.0576171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0446,
+ "step": 9439
+ },
+ {
+ "epoch": 30.159744408945688,
+ "grad_norm": 0.044677734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0475,
+ "step": 9440
+ },
+ {
+ "epoch": 30.162939297124602,
+ "grad_norm": 0.054931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0335,
+ "step": 9441
+ },
+ {
+ "epoch": 30.166134185303516,
+ "grad_norm": 0.055419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0238,
+ "step": 9442
+ },
+ {
+ "epoch": 30.169329073482427,
+ "grad_norm": 0.08984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0384,
+ "step": 9443
+ },
+ {
+ "epoch": 30.17252396166134,
+ "grad_norm": 0.05078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0467,
+ "step": 9444
+ },
+ {
+ "epoch": 30.175718849840255,
+ "grad_norm": 0.10205078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0377,
+ "step": 9445
+ },
+ {
+ "epoch": 30.17891373801917,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0382,
+ "step": 9446
+ },
+ {
+ "epoch": 30.182108626198083,
+ "grad_norm": 0.09716796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0461,
+ "step": 9447
+ },
+ {
+ "epoch": 30.185303514376997,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.03,
+ "step": 9448
+ },
+ {
+ "epoch": 30.18849840255591,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0341,
+ "step": 9449
+ },
+ {
+ "epoch": 30.191693290734825,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0377,
+ "step": 9450
+ },
+ {
+ "epoch": 30.19488817891374,
+ "grad_norm": 0.08837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0343,
+ "step": 9451
+ },
+ {
+ "epoch": 30.19808306709265,
+ "grad_norm": 0.0546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0364,
+ "step": 9452
+ },
+ {
+ "epoch": 30.201277955271564,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0275,
+ "step": 9453
+ },
+ {
+ "epoch": 30.20447284345048,
+ "grad_norm": 0.05810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0352,
+ "step": 9454
+ },
+ {
+ "epoch": 30.207667731629392,
+ "grad_norm": 0.048583984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0495,
+ "step": 9455
+ },
+ {
+ "epoch": 30.210862619808307,
+ "grad_norm": 0.06103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0366,
+ "step": 9456
+ },
+ {
+ "epoch": 30.21405750798722,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0388,
+ "step": 9457
+ },
+ {
+ "epoch": 30.217252396166135,
+ "grad_norm": 0.0576171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0455,
+ "step": 9458
+ },
+ {
+ "epoch": 30.22044728434505,
+ "grad_norm": 0.056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.034,
+ "step": 9459
+ },
+ {
+ "epoch": 30.223642172523963,
+ "grad_norm": 0.051513671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0291,
+ "step": 9460
+ },
+ {
+ "epoch": 30.226837060702877,
+ "grad_norm": 0.058837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0569,
+ "step": 9461
+ },
+ {
+ "epoch": 30.230031948881788,
+ "grad_norm": 0.08544921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0244,
+ "step": 9462
+ },
+ {
+ "epoch": 30.233226837060702,
+ "grad_norm": 0.052001953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0491,
+ "step": 9463
+ },
+ {
+ "epoch": 30.236421725239616,
+ "grad_norm": 0.052978515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0333,
+ "step": 9464
+ },
+ {
+ "epoch": 30.23961661341853,
+ "grad_norm": 0.0830078125,
+ "learning_rate": 0.0005,
+ "loss": 1.044,
+ "step": 9465
+ },
+ {
+ "epoch": 30.242811501597444,
+ "grad_norm": 0.0615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0439,
+ "step": 9466
+ },
+ {
+ "epoch": 30.24600638977636,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0344,
+ "step": 9467
+ },
+ {
+ "epoch": 30.249201277955272,
+ "grad_norm": 0.05126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.04,
+ "step": 9468
+ },
+ {
+ "epoch": 30.252396166134186,
+ "grad_norm": 0.061767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0311,
+ "step": 9469
+ },
+ {
+ "epoch": 30.2555910543131,
+ "grad_norm": 0.057373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0358,
+ "step": 9470
+ },
+ {
+ "epoch": 30.25878594249201,
+ "grad_norm": 0.054931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0397,
+ "step": 9471
+ },
+ {
+ "epoch": 30.261980830670925,
+ "grad_norm": 0.055908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.039,
+ "step": 9472
+ },
+ {
+ "epoch": 30.26517571884984,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0237,
+ "step": 9473
+ },
+ {
+ "epoch": 30.268370607028753,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0331,
+ "step": 9474
+ },
+ {
+ "epoch": 30.271565495207668,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0343,
+ "step": 9475
+ },
+ {
+ "epoch": 30.27476038338658,
+ "grad_norm": 0.057861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0345,
+ "step": 9476
+ },
+ {
+ "epoch": 30.277955271565496,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0414,
+ "step": 9477
+ },
+ {
+ "epoch": 30.28115015974441,
+ "grad_norm": 0.0869140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0522,
+ "step": 9478
+ },
+ {
+ "epoch": 30.284345047923324,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0399,
+ "step": 9479
+ },
+ {
+ "epoch": 30.287539936102238,
+ "grad_norm": 0.05322265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0305,
+ "step": 9480
+ },
+ {
+ "epoch": 30.29073482428115,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0387,
+ "step": 9481
+ },
+ {
+ "epoch": 30.293929712460063,
+ "grad_norm": 0.058349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.031,
+ "step": 9482
+ },
+ {
+ "epoch": 30.297124600638977,
+ "grad_norm": 0.09228515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0349,
+ "step": 9483
+ },
+ {
+ "epoch": 30.30031948881789,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0264,
+ "step": 9484
+ },
+ {
+ "epoch": 30.303514376996805,
+ "grad_norm": 0.0888671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0376,
+ "step": 9485
+ },
+ {
+ "epoch": 30.30670926517572,
+ "grad_norm": 0.0615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0448,
+ "step": 9486
+ },
+ {
+ "epoch": 30.309904153354633,
+ "grad_norm": 0.09765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0387,
+ "step": 9487
+ },
+ {
+ "epoch": 30.313099041533548,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0338,
+ "step": 9488
+ },
+ {
+ "epoch": 30.31629392971246,
+ "grad_norm": 0.050048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0394,
+ "step": 9489
+ },
+ {
+ "epoch": 30.319488817891372,
+ "grad_norm": 0.10595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0392,
+ "step": 9490
+ },
+ {
+ "epoch": 30.322683706070286,
+ "grad_norm": 0.08154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0339,
+ "step": 9491
+ },
+ {
+ "epoch": 30.3258785942492,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0353,
+ "step": 9492
+ },
+ {
+ "epoch": 30.329073482428115,
+ "grad_norm": 0.09130859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0313,
+ "step": 9493
+ },
+ {
+ "epoch": 30.33226837060703,
+ "grad_norm": 0.0986328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0361,
+ "step": 9494
+ },
+ {
+ "epoch": 30.335463258785943,
+ "grad_norm": 0.0947265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0447,
+ "step": 9495
+ },
+ {
+ "epoch": 30.338658146964857,
+ "grad_norm": 0.08203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0514,
+ "step": 9496
+ },
+ {
+ "epoch": 30.34185303514377,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0279,
+ "step": 9497
+ },
+ {
+ "epoch": 30.345047923322685,
+ "grad_norm": 0.0810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0423,
+ "step": 9498
+ },
+ {
+ "epoch": 30.3482428115016,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0447,
+ "step": 9499
+ },
+ {
+ "epoch": 30.35143769968051,
+ "grad_norm": 0.09912109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0393,
+ "step": 9500
+ },
+ {
+ "epoch": 30.354632587859424,
+ "grad_norm": 0.07861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0254,
+ "step": 9501
+ },
+ {
+ "epoch": 30.357827476038338,
+ "grad_norm": 0.109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0361,
+ "step": 9502
+ },
+ {
+ "epoch": 30.361022364217252,
+ "grad_norm": 0.060791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0525,
+ "step": 9503
+ },
+ {
+ "epoch": 30.364217252396166,
+ "grad_norm": 0.0927734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0303,
+ "step": 9504
+ },
+ {
+ "epoch": 30.36741214057508,
+ "grad_norm": 0.05078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0312,
+ "step": 9505
+ },
+ {
+ "epoch": 30.370607028753994,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.048,
+ "step": 9506
+ },
+ {
+ "epoch": 30.37380191693291,
+ "grad_norm": 0.049560546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0407,
+ "step": 9507
+ },
+ {
+ "epoch": 30.376996805111823,
+ "grad_norm": 0.058837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0413,
+ "step": 9508
+ },
+ {
+ "epoch": 30.380191693290733,
+ "grad_norm": 0.05419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0332,
+ "step": 9509
+ },
+ {
+ "epoch": 30.383386581469647,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0315,
+ "step": 9510
+ },
+ {
+ "epoch": 30.38658146964856,
+ "grad_norm": 0.054443359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0159,
+ "step": 9511
+ },
+ {
+ "epoch": 30.389776357827476,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0345,
+ "step": 9512
+ },
+ {
+ "epoch": 30.39297124600639,
+ "grad_norm": 0.062255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0488,
+ "step": 9513
+ },
+ {
+ "epoch": 30.396166134185304,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0366,
+ "step": 9514
+ },
+ {
+ "epoch": 30.399361022364218,
+ "grad_norm": 0.062255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0328,
+ "step": 9515
+ },
+ {
+ "epoch": 30.402555910543132,
+ "grad_norm": 0.0546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0436,
+ "step": 9516
+ },
+ {
+ "epoch": 30.405750798722046,
+ "grad_norm": 0.05224609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0337,
+ "step": 9517
+ },
+ {
+ "epoch": 30.408945686900957,
+ "grad_norm": 0.0498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0333,
+ "step": 9518
+ },
+ {
+ "epoch": 30.41214057507987,
+ "grad_norm": 0.059326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0386,
+ "step": 9519
+ },
+ {
+ "epoch": 30.415335463258785,
+ "grad_norm": 0.05224609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0351,
+ "step": 9520
+ },
+ {
+ "epoch": 30.4185303514377,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0369,
+ "step": 9521
+ },
+ {
+ "epoch": 30.421725239616613,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0393,
+ "step": 9522
+ },
+ {
+ "epoch": 30.424920127795527,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0372,
+ "step": 9523
+ },
+ {
+ "epoch": 30.42811501597444,
+ "grad_norm": 0.0546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0405,
+ "step": 9524
+ },
+ {
+ "epoch": 30.431309904153355,
+ "grad_norm": 0.052978515625,
+ "learning_rate": 0.0005,
+ "loss": 1.035,
+ "step": 9525
+ },
+ {
+ "epoch": 30.43450479233227,
+ "grad_norm": 0.058349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0343,
+ "step": 9526
+ },
+ {
+ "epoch": 30.437699680511184,
+ "grad_norm": 0.057861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0336,
+ "step": 9527
+ },
+ {
+ "epoch": 30.440894568690094,
+ "grad_norm": 0.06103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0421,
+ "step": 9528
+ },
+ {
+ "epoch": 30.44408945686901,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0341,
+ "step": 9529
+ },
+ {
+ "epoch": 30.447284345047922,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.027,
+ "step": 9530
+ },
+ {
+ "epoch": 30.450479233226837,
+ "grad_norm": 0.1015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0345,
+ "step": 9531
+ },
+ {
+ "epoch": 30.45367412140575,
+ "grad_norm": 0.08251953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0312,
+ "step": 9532
+ },
+ {
+ "epoch": 30.456869009584665,
+ "grad_norm": 0.08544921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0369,
+ "step": 9533
+ },
+ {
+ "epoch": 30.46006389776358,
+ "grad_norm": 0.07861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0402,
+ "step": 9534
+ },
+ {
+ "epoch": 30.463258785942493,
+ "grad_norm": 0.048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.038,
+ "step": 9535
+ },
+ {
+ "epoch": 30.466453674121407,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0415,
+ "step": 9536
+ },
+ {
+ "epoch": 30.46964856230032,
+ "grad_norm": 0.05615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0397,
+ "step": 9537
+ },
+ {
+ "epoch": 30.472843450479232,
+ "grad_norm": 0.06103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.041,
+ "step": 9538
+ },
+ {
+ "epoch": 30.476038338658146,
+ "grad_norm": 0.0908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0318,
+ "step": 9539
+ },
+ {
+ "epoch": 30.47923322683706,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0374,
+ "step": 9540
+ },
+ {
+ "epoch": 30.482428115015974,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0437,
+ "step": 9541
+ },
+ {
+ "epoch": 30.48562300319489,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0283,
+ "step": 9542
+ },
+ {
+ "epoch": 30.488817891373802,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0452,
+ "step": 9543
+ },
+ {
+ "epoch": 30.492012779552716,
+ "grad_norm": 0.05908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0368,
+ "step": 9544
+ },
+ {
+ "epoch": 30.49520766773163,
+ "grad_norm": 0.05859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0361,
+ "step": 9545
+ },
+ {
+ "epoch": 30.498402555910545,
+ "grad_norm": 0.0546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0243,
+ "step": 9546
+ },
+ {
+ "epoch": 30.501597444089455,
+ "grad_norm": 0.08349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0324,
+ "step": 9547
+ },
+ {
+ "epoch": 30.50479233226837,
+ "grad_norm": 0.0546875,
+ "learning_rate": 0.0005,
+ "loss": 1.047,
+ "step": 9548
+ },
+ {
+ "epoch": 30.507987220447284,
+ "grad_norm": 0.055908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0354,
+ "step": 9549
+ },
+ {
+ "epoch": 30.511182108626198,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0422,
+ "step": 9550
+ },
+ {
+ "epoch": 30.51437699680511,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0419,
+ "step": 9551
+ },
+ {
+ "epoch": 30.517571884984026,
+ "grad_norm": 0.0849609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0479,
+ "step": 9552
+ },
+ {
+ "epoch": 30.52076677316294,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0363,
+ "step": 9553
+ },
+ {
+ "epoch": 30.523961661341854,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.035,
+ "step": 9554
+ },
+ {
+ "epoch": 30.527156549520768,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0328,
+ "step": 9555
+ },
+ {
+ "epoch": 30.53035143769968,
+ "grad_norm": 0.10400390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0479,
+ "step": 9556
+ },
+ {
+ "epoch": 30.533546325878593,
+ "grad_norm": 0.10009765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0334,
+ "step": 9557
+ },
+ {
+ "epoch": 30.536741214057507,
+ "grad_norm": 0.1689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0418,
+ "step": 9558
+ },
+ {
+ "epoch": 30.53993610223642,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0349,
+ "step": 9559
+ },
+ {
+ "epoch": 30.543130990415335,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0352,
+ "step": 9560
+ },
+ {
+ "epoch": 30.54632587859425,
+ "grad_norm": 0.0537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0455,
+ "step": 9561
+ },
+ {
+ "epoch": 30.549520766773163,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0414,
+ "step": 9562
+ },
+ {
+ "epoch": 30.552715654952078,
+ "grad_norm": 0.0556640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0222,
+ "step": 9563
+ },
+ {
+ "epoch": 30.55591054313099,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0374,
+ "step": 9564
+ },
+ {
+ "epoch": 30.559105431309906,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0374,
+ "step": 9565
+ },
+ {
+ "epoch": 30.562300319488816,
+ "grad_norm": 0.08837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0419,
+ "step": 9566
+ },
+ {
+ "epoch": 30.56549520766773,
+ "grad_norm": 0.10205078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0307,
+ "step": 9567
+ },
+ {
+ "epoch": 30.568690095846645,
+ "grad_norm": 0.1416015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0323,
+ "step": 9568
+ },
+ {
+ "epoch": 30.57188498402556,
+ "grad_norm": 0.06005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0337,
+ "step": 9569
+ },
+ {
+ "epoch": 30.575079872204473,
+ "grad_norm": 0.099609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0428,
+ "step": 9570
+ },
+ {
+ "epoch": 30.578274760383387,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0443,
+ "step": 9571
+ },
+ {
+ "epoch": 30.5814696485623,
+ "grad_norm": 0.06201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0389,
+ "step": 9572
+ },
+ {
+ "epoch": 30.584664536741215,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0442,
+ "step": 9573
+ },
+ {
+ "epoch": 30.58785942492013,
+ "grad_norm": 0.09423828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0403,
+ "step": 9574
+ },
+ {
+ "epoch": 30.591054313099043,
+ "grad_norm": 0.053466796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0303,
+ "step": 9575
+ },
+ {
+ "epoch": 30.594249201277954,
+ "grad_norm": 0.05029296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0424,
+ "step": 9576
+ },
+ {
+ "epoch": 30.597444089456868,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0415,
+ "step": 9577
+ },
+ {
+ "epoch": 30.600638977635782,
+ "grad_norm": 0.050537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0392,
+ "step": 9578
+ },
+ {
+ "epoch": 30.603833865814696,
+ "grad_norm": 0.05078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0474,
+ "step": 9579
+ },
+ {
+ "epoch": 30.60702875399361,
+ "grad_norm": 0.058837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0421,
+ "step": 9580
+ },
+ {
+ "epoch": 30.610223642172524,
+ "grad_norm": 0.052001953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0316,
+ "step": 9581
+ },
+ {
+ "epoch": 30.61341853035144,
+ "grad_norm": 0.06005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0344,
+ "step": 9582
+ },
+ {
+ "epoch": 30.616613418530353,
+ "grad_norm": 0.058349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0406,
+ "step": 9583
+ },
+ {
+ "epoch": 30.619808306709267,
+ "grad_norm": 0.0810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0372,
+ "step": 9584
+ },
+ {
+ "epoch": 30.623003194888177,
+ "grad_norm": 0.08447265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0383,
+ "step": 9585
+ },
+ {
+ "epoch": 30.62619808306709,
+ "grad_norm": 0.051025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0331,
+ "step": 9586
+ },
+ {
+ "epoch": 30.629392971246006,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0382,
+ "step": 9587
+ },
+ {
+ "epoch": 30.63258785942492,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0266,
+ "step": 9588
+ },
+ {
+ "epoch": 30.635782747603834,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0415,
+ "step": 9589
+ },
+ {
+ "epoch": 30.638977635782748,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0336,
+ "step": 9590
+ },
+ {
+ "epoch": 30.642172523961662,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0366,
+ "step": 9591
+ },
+ {
+ "epoch": 30.645367412140576,
+ "grad_norm": 0.054443359375,
+ "learning_rate": 0.0005,
+ "loss": 1.034,
+ "step": 9592
+ },
+ {
+ "epoch": 30.64856230031949,
+ "grad_norm": 0.05126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0428,
+ "step": 9593
+ },
+ {
+ "epoch": 30.6517571884984,
+ "grad_norm": 0.0537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0367,
+ "step": 9594
+ },
+ {
+ "epoch": 30.654952076677315,
+ "grad_norm": 0.0537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0321,
+ "step": 9595
+ },
+ {
+ "epoch": 30.65814696485623,
+ "grad_norm": 0.051513671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0404,
+ "step": 9596
+ },
+ {
+ "epoch": 30.661341853035143,
+ "grad_norm": 0.056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0344,
+ "step": 9597
+ },
+ {
+ "epoch": 30.664536741214057,
+ "grad_norm": 0.061767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0345,
+ "step": 9598
+ },
+ {
+ "epoch": 30.66773162939297,
+ "grad_norm": 0.08544921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0492,
+ "step": 9599
+ },
+ {
+ "epoch": 30.670926517571885,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0296,
+ "step": 9600
+ },
+ {
+ "epoch": 30.6741214057508,
+ "grad_norm": 0.0859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0409,
+ "step": 9601
+ },
+ {
+ "epoch": 30.677316293929714,
+ "grad_norm": 0.061767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0322,
+ "step": 9602
+ },
+ {
+ "epoch": 30.680511182108628,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0352,
+ "step": 9603
+ },
+ {
+ "epoch": 30.68370607028754,
+ "grad_norm": 0.0947265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0383,
+ "step": 9604
+ },
+ {
+ "epoch": 30.686900958466452,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0311,
+ "step": 9605
+ },
+ {
+ "epoch": 30.690095846645367,
+ "grad_norm": 0.1337890625,
+ "learning_rate": 0.0005,
+ "loss": 1.036,
+ "step": 9606
+ },
+ {
+ "epoch": 30.69329073482428,
+ "grad_norm": 0.08251953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0444,
+ "step": 9607
+ },
+ {
+ "epoch": 30.696485623003195,
+ "grad_norm": 0.10791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0316,
+ "step": 9608
+ },
+ {
+ "epoch": 30.69968051118211,
+ "grad_norm": 0.0810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0296,
+ "step": 9609
+ },
+ {
+ "epoch": 30.702875399361023,
+ "grad_norm": 0.07861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0358,
+ "step": 9610
+ },
+ {
+ "epoch": 30.706070287539937,
+ "grad_norm": 0.058349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0372,
+ "step": 9611
+ },
+ {
+ "epoch": 30.70926517571885,
+ "grad_norm": 0.04931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0486,
+ "step": 9612
+ },
+ {
+ "epoch": 30.712460063897765,
+ "grad_norm": 0.05615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0412,
+ "step": 9613
+ },
+ {
+ "epoch": 30.715654952076676,
+ "grad_norm": 0.05615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0301,
+ "step": 9614
+ },
+ {
+ "epoch": 30.71884984025559,
+ "grad_norm": 0.050048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0238,
+ "step": 9615
+ },
+ {
+ "epoch": 30.722044728434504,
+ "grad_norm": 0.05078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0506,
+ "step": 9616
+ },
+ {
+ "epoch": 30.72523961661342,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0367,
+ "step": 9617
+ },
+ {
+ "epoch": 30.728434504792332,
+ "grad_norm": 0.056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0345,
+ "step": 9618
+ },
+ {
+ "epoch": 30.731629392971247,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0396,
+ "step": 9619
+ },
+ {
+ "epoch": 30.73482428115016,
+ "grad_norm": 0.061767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0352,
+ "step": 9620
+ },
+ {
+ "epoch": 30.738019169329075,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0524,
+ "step": 9621
+ },
+ {
+ "epoch": 30.74121405750799,
+ "grad_norm": 0.1591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.03,
+ "step": 9622
+ },
+ {
+ "epoch": 30.7444089456869,
+ "grad_norm": 0.10986328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0429,
+ "step": 9623
+ },
+ {
+ "epoch": 30.747603833865814,
+ "grad_norm": 0.1162109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0386,
+ "step": 9624
+ },
+ {
+ "epoch": 30.750798722044728,
+ "grad_norm": 0.1005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0438,
+ "step": 9625
+ },
+ {
+ "epoch": 30.75399361022364,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0434,
+ "step": 9626
+ },
+ {
+ "epoch": 30.757188498402556,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0292,
+ "step": 9627
+ },
+ {
+ "epoch": 30.76038338658147,
+ "grad_norm": 0.058349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.049,
+ "step": 9628
+ },
+ {
+ "epoch": 30.763578274760384,
+ "grad_norm": 0.0966796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0443,
+ "step": 9629
+ },
+ {
+ "epoch": 30.766773162939298,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0416,
+ "step": 9630
+ },
+ {
+ "epoch": 30.769968051118212,
+ "grad_norm": 0.09912109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0441,
+ "step": 9631
+ },
+ {
+ "epoch": 30.773162939297123,
+ "grad_norm": 0.060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0313,
+ "step": 9632
+ },
+ {
+ "epoch": 30.776357827476037,
+ "grad_norm": 0.08984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0261,
+ "step": 9633
+ },
+ {
+ "epoch": 30.77955271565495,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0438,
+ "step": 9634
+ },
+ {
+ "epoch": 30.782747603833865,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0387,
+ "step": 9635
+ },
+ {
+ "epoch": 30.78594249201278,
+ "grad_norm": 0.10791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0381,
+ "step": 9636
+ },
+ {
+ "epoch": 30.789137380191693,
+ "grad_norm": 0.1962890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0241,
+ "step": 9637
+ },
+ {
+ "epoch": 30.792332268370608,
+ "grad_norm": 0.058837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0292,
+ "step": 9638
+ },
+ {
+ "epoch": 30.79552715654952,
+ "grad_norm": 0.10888671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0377,
+ "step": 9639
+ },
+ {
+ "epoch": 30.798722044728436,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0398,
+ "step": 9640
+ },
+ {
+ "epoch": 30.80191693290735,
+ "grad_norm": 0.1064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0353,
+ "step": 9641
+ },
+ {
+ "epoch": 30.80511182108626,
+ "grad_norm": 0.08544921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0385,
+ "step": 9642
+ },
+ {
+ "epoch": 30.808306709265175,
+ "grad_norm": 0.0595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0322,
+ "step": 9643
+ },
+ {
+ "epoch": 30.81150159744409,
+ "grad_norm": 0.08154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0377,
+ "step": 9644
+ },
+ {
+ "epoch": 30.814696485623003,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0331,
+ "step": 9645
+ },
+ {
+ "epoch": 30.817891373801917,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0304,
+ "step": 9646
+ },
+ {
+ "epoch": 30.82108626198083,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0415,
+ "step": 9647
+ },
+ {
+ "epoch": 30.824281150159745,
+ "grad_norm": 0.0966796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0299,
+ "step": 9648
+ },
+ {
+ "epoch": 30.82747603833866,
+ "grad_norm": 0.0859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0457,
+ "step": 9649
+ },
+ {
+ "epoch": 30.830670926517573,
+ "grad_norm": 0.083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0399,
+ "step": 9650
+ },
+ {
+ "epoch": 30.833865814696484,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0354,
+ "step": 9651
+ },
+ {
+ "epoch": 30.837060702875398,
+ "grad_norm": 0.09033203125,
+ "learning_rate": 0.0005,
+ "loss": 1.046,
+ "step": 9652
+ },
+ {
+ "epoch": 30.840255591054312,
+ "grad_norm": 0.09326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.037,
+ "step": 9653
+ },
+ {
+ "epoch": 30.843450479233226,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0427,
+ "step": 9654
+ },
+ {
+ "epoch": 30.84664536741214,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.044,
+ "step": 9655
+ },
+ {
+ "epoch": 30.849840255591054,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0445,
+ "step": 9656
+ },
+ {
+ "epoch": 30.85303514376997,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0391,
+ "step": 9657
+ },
+ {
+ "epoch": 30.856230031948883,
+ "grad_norm": 0.0576171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0428,
+ "step": 9658
+ },
+ {
+ "epoch": 30.859424920127797,
+ "grad_norm": 0.054931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0429,
+ "step": 9659
+ },
+ {
+ "epoch": 30.86261980830671,
+ "grad_norm": 0.05712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.036,
+ "step": 9660
+ },
+ {
+ "epoch": 30.86581469648562,
+ "grad_norm": 0.080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0351,
+ "step": 9661
+ },
+ {
+ "epoch": 30.869009584664536,
+ "grad_norm": 0.057861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0426,
+ "step": 9662
+ },
+ {
+ "epoch": 30.87220447284345,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0345,
+ "step": 9663
+ },
+ {
+ "epoch": 30.875399361022364,
+ "grad_norm": 0.0556640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0317,
+ "step": 9664
+ },
+ {
+ "epoch": 30.878594249201278,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0386,
+ "step": 9665
+ },
+ {
+ "epoch": 30.881789137380192,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0463,
+ "step": 9666
+ },
+ {
+ "epoch": 30.884984025559106,
+ "grad_norm": 0.0869140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0365,
+ "step": 9667
+ },
+ {
+ "epoch": 30.88817891373802,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.021,
+ "step": 9668
+ },
+ {
+ "epoch": 30.891373801916934,
+ "grad_norm": 0.052978515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0289,
+ "step": 9669
+ },
+ {
+ "epoch": 30.894568690095845,
+ "grad_norm": 0.083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0359,
+ "step": 9670
+ },
+ {
+ "epoch": 30.89776357827476,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0271,
+ "step": 9671
+ },
+ {
+ "epoch": 30.900958466453673,
+ "grad_norm": 0.126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0479,
+ "step": 9672
+ },
+ {
+ "epoch": 30.904153354632587,
+ "grad_norm": 0.057373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0382,
+ "step": 9673
+ },
+ {
+ "epoch": 30.9073482428115,
+ "grad_norm": 0.10107421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0305,
+ "step": 9674
+ },
+ {
+ "epoch": 30.910543130990416,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.036,
+ "step": 9675
+ },
+ {
+ "epoch": 30.91373801916933,
+ "grad_norm": 0.09765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0411,
+ "step": 9676
+ },
+ {
+ "epoch": 30.916932907348244,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0403,
+ "step": 9677
+ },
+ {
+ "epoch": 30.920127795527158,
+ "grad_norm": 0.0927734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0385,
+ "step": 9678
+ },
+ {
+ "epoch": 30.923322683706072,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0298,
+ "step": 9679
+ },
+ {
+ "epoch": 30.926517571884983,
+ "grad_norm": 0.10107421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0438,
+ "step": 9680
+ },
+ {
+ "epoch": 30.929712460063897,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0406,
+ "step": 9681
+ },
+ {
+ "epoch": 30.93290734824281,
+ "grad_norm": 0.123046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0385,
+ "step": 9682
+ },
+ {
+ "epoch": 30.936102236421725,
+ "grad_norm": 0.087890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0371,
+ "step": 9683
+ },
+ {
+ "epoch": 30.93929712460064,
+ "grad_norm": 0.11376953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0351,
+ "step": 9684
+ },
+ {
+ "epoch": 30.942492012779553,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0417,
+ "step": 9685
+ },
+ {
+ "epoch": 30.945686900958467,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0527,
+ "step": 9686
+ },
+ {
+ "epoch": 30.94888178913738,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0438,
+ "step": 9687
+ },
+ {
+ "epoch": 30.952076677316295,
+ "grad_norm": 0.10107421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0358,
+ "step": 9688
+ },
+ {
+ "epoch": 30.955271565495206,
+ "grad_norm": 0.056396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0358,
+ "step": 9689
+ },
+ {
+ "epoch": 30.95846645367412,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0388,
+ "step": 9690
+ },
+ {
+ "epoch": 30.961661341853034,
+ "grad_norm": 0.056884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0334,
+ "step": 9691
+ },
+ {
+ "epoch": 30.96485623003195,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0198,
+ "step": 9692
+ },
+ {
+ "epoch": 30.968051118210862,
+ "grad_norm": 0.059814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0326,
+ "step": 9693
+ },
+ {
+ "epoch": 30.971246006389777,
+ "grad_norm": 0.09814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0444,
+ "step": 9694
+ },
+ {
+ "epoch": 30.97444089456869,
+ "grad_norm": 0.05517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0254,
+ "step": 9695
+ },
+ {
+ "epoch": 30.977635782747605,
+ "grad_norm": 0.06005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0422,
+ "step": 9696
+ },
+ {
+ "epoch": 30.98083067092652,
+ "grad_norm": 0.052490234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0313,
+ "step": 9697
+ },
+ {
+ "epoch": 30.984025559105433,
+ "grad_norm": 0.0810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0325,
+ "step": 9698
+ },
+ {
+ "epoch": 30.987220447284344,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0501,
+ "step": 9699
+ },
+ {
+ "epoch": 30.990415335463258,
+ "grad_norm": 0.09130859375,
+ "learning_rate": 0.0005,
+ "loss": 1.038,
+ "step": 9700
+ },
+ {
+ "epoch": 30.99361022364217,
+ "grad_norm": 0.0546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0294,
+ "step": 9701
+ },
+ {
+ "epoch": 30.996805111821086,
+ "grad_norm": 0.05712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0312,
+ "step": 9702
+ },
+ {
+ "epoch": 31.0,
+ "grad_norm": 0.05615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0263,
+ "step": 9703
+ },
+ {
+ "epoch": 31.003194888178914,
+ "grad_norm": 0.05322265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0277,
+ "step": 9704
+ },
+ {
+ "epoch": 31.00638977635783,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0385,
+ "step": 9705
+ },
+ {
+ "epoch": 31.009584664536742,
+ "grad_norm": 0.049560546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0498,
+ "step": 9706
+ },
+ {
+ "epoch": 31.012779552715656,
+ "grad_norm": 0.055419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0245,
+ "step": 9707
+ },
+ {
+ "epoch": 31.015974440894567,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0236,
+ "step": 9708
+ },
+ {
+ "epoch": 31.01916932907348,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0306,
+ "step": 9709
+ },
+ {
+ "epoch": 31.022364217252395,
+ "grad_norm": 0.05078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0209,
+ "step": 9710
+ },
+ {
+ "epoch": 31.02555910543131,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0491,
+ "step": 9711
+ },
+ {
+ "epoch": 31.028753993610223,
+ "grad_norm": 0.0576171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0411,
+ "step": 9712
+ },
+ {
+ "epoch": 31.031948881789138,
+ "grad_norm": 0.059326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0399,
+ "step": 9713
+ },
+ {
+ "epoch": 31.03514376996805,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0296,
+ "step": 9714
+ },
+ {
+ "epoch": 31.038338658146966,
+ "grad_norm": 0.06201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0459,
+ "step": 9715
+ },
+ {
+ "epoch": 31.04153354632588,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0325,
+ "step": 9716
+ },
+ {
+ "epoch": 31.044728434504794,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0357,
+ "step": 9717
+ },
+ {
+ "epoch": 31.047923322683705,
+ "grad_norm": 0.0576171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0406,
+ "step": 9718
+ },
+ {
+ "epoch": 31.05111821086262,
+ "grad_norm": 0.05810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0366,
+ "step": 9719
+ },
+ {
+ "epoch": 31.054313099041533,
+ "grad_norm": 0.05322265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0281,
+ "step": 9720
+ },
+ {
+ "epoch": 31.057507987220447,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0304,
+ "step": 9721
+ },
+ {
+ "epoch": 31.06070287539936,
+ "grad_norm": 0.05810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.044,
+ "step": 9722
+ },
+ {
+ "epoch": 31.063897763578275,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0407,
+ "step": 9723
+ },
+ {
+ "epoch": 31.06709265175719,
+ "grad_norm": 0.05224609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0464,
+ "step": 9724
+ },
+ {
+ "epoch": 31.070287539936103,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0433,
+ "step": 9725
+ },
+ {
+ "epoch": 31.073482428115017,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0245,
+ "step": 9726
+ },
+ {
+ "epoch": 31.076677316293928,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0354,
+ "step": 9727
+ },
+ {
+ "epoch": 31.079872204472842,
+ "grad_norm": 0.08642578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0449,
+ "step": 9728
+ },
+ {
+ "epoch": 31.083067092651756,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0415,
+ "step": 9729
+ },
+ {
+ "epoch": 31.08626198083067,
+ "grad_norm": 0.051025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0488,
+ "step": 9730
+ },
+ {
+ "epoch": 31.089456869009584,
+ "grad_norm": 0.06005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0385,
+ "step": 9731
+ },
+ {
+ "epoch": 31.0926517571885,
+ "grad_norm": 0.05517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0267,
+ "step": 9732
+ },
+ {
+ "epoch": 31.095846645367413,
+ "grad_norm": 0.052734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0371,
+ "step": 9733
+ },
+ {
+ "epoch": 31.099041533546327,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0325,
+ "step": 9734
+ },
+ {
+ "epoch": 31.10223642172524,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0323,
+ "step": 9735
+ },
+ {
+ "epoch": 31.105431309904155,
+ "grad_norm": 0.06103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0436,
+ "step": 9736
+ },
+ {
+ "epoch": 31.108626198083066,
+ "grad_norm": 0.05517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0282,
+ "step": 9737
+ },
+ {
+ "epoch": 31.11182108626198,
+ "grad_norm": 0.05908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.042,
+ "step": 9738
+ },
+ {
+ "epoch": 31.115015974440894,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0392,
+ "step": 9739
+ },
+ {
+ "epoch": 31.118210862619808,
+ "grad_norm": 0.06103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.046,
+ "step": 9740
+ },
+ {
+ "epoch": 31.121405750798722,
+ "grad_norm": 0.053955078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0412,
+ "step": 9741
+ },
+ {
+ "epoch": 31.124600638977636,
+ "grad_norm": 0.052001953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0464,
+ "step": 9742
+ },
+ {
+ "epoch": 31.12779552715655,
+ "grad_norm": 0.05322265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0313,
+ "step": 9743
+ },
+ {
+ "epoch": 31.130990415335464,
+ "grad_norm": 0.054443359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0339,
+ "step": 9744
+ },
+ {
+ "epoch": 31.13418530351438,
+ "grad_norm": 0.0517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0246,
+ "step": 9745
+ },
+ {
+ "epoch": 31.13738019169329,
+ "grad_norm": 0.054931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0353,
+ "step": 9746
+ },
+ {
+ "epoch": 31.140575079872203,
+ "grad_norm": 0.05078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0449,
+ "step": 9747
+ },
+ {
+ "epoch": 31.143769968051117,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0254,
+ "step": 9748
+ },
+ {
+ "epoch": 31.14696485623003,
+ "grad_norm": 0.045166015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0272,
+ "step": 9749
+ },
+ {
+ "epoch": 31.150159744408946,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0355,
+ "step": 9750
+ },
+ {
+ "epoch": 31.15335463258786,
+ "grad_norm": 0.10205078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0447,
+ "step": 9751
+ },
+ {
+ "epoch": 31.156549520766774,
+ "grad_norm": 0.053466796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0315,
+ "step": 9752
+ },
+ {
+ "epoch": 31.159744408945688,
+ "grad_norm": 0.0576171875,
+ "learning_rate": 0.0005,
+ "loss": 1.038,
+ "step": 9753
+ },
+ {
+ "epoch": 31.162939297124602,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0429,
+ "step": 9754
+ },
+ {
+ "epoch": 31.166134185303516,
+ "grad_norm": 0.05712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.034,
+ "step": 9755
+ },
+ {
+ "epoch": 31.169329073482427,
+ "grad_norm": 0.1337890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0351,
+ "step": 9756
+ },
+ {
+ "epoch": 31.17252396166134,
+ "grad_norm": 0.15234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0442,
+ "step": 9757
+ },
+ {
+ "epoch": 31.175718849840255,
+ "grad_norm": 0.12109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0486,
+ "step": 9758
+ },
+ {
+ "epoch": 31.17891373801917,
+ "grad_norm": 0.263671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0496,
+ "step": 9759
+ },
+ {
+ "epoch": 31.182108626198083,
+ "grad_norm": 0.130859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0437,
+ "step": 9760
+ },
+ {
+ "epoch": 31.185303514376997,
+ "grad_norm": 0.1552734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0307,
+ "step": 9761
+ },
+ {
+ "epoch": 31.18849840255591,
+ "grad_norm": 0.2431640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0439,
+ "step": 9762
+ },
+ {
+ "epoch": 31.191693290734825,
+ "grad_norm": 0.05712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.045,
+ "step": 9763
+ },
+ {
+ "epoch": 31.19488817891374,
+ "grad_norm": 0.177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0282,
+ "step": 9764
+ },
+ {
+ "epoch": 31.19808306709265,
+ "grad_norm": 0.05029296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0496,
+ "step": 9765
+ },
+ {
+ "epoch": 31.201277955271564,
+ "grad_norm": 0.1162109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0385,
+ "step": 9766
+ },
+ {
+ "epoch": 31.20447284345048,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0361,
+ "step": 9767
+ },
+ {
+ "epoch": 31.207667731629392,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0467,
+ "step": 9768
+ },
+ {
+ "epoch": 31.210862619808307,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0251,
+ "step": 9769
+ },
+ {
+ "epoch": 31.21405750798722,
+ "grad_norm": 0.091796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0424,
+ "step": 9770
+ },
+ {
+ "epoch": 31.217252396166135,
+ "grad_norm": 0.080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0543,
+ "step": 9771
+ },
+ {
+ "epoch": 31.22044728434505,
+ "grad_norm": 0.08984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0407,
+ "step": 9772
+ },
+ {
+ "epoch": 31.223642172523963,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0405,
+ "step": 9773
+ },
+ {
+ "epoch": 31.226837060702877,
+ "grad_norm": 0.119140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0301,
+ "step": 9774
+ },
+ {
+ "epoch": 31.230031948881788,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0344,
+ "step": 9775
+ },
+ {
+ "epoch": 31.233226837060702,
+ "grad_norm": 0.10986328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0404,
+ "step": 9776
+ },
+ {
+ "epoch": 31.236421725239616,
+ "grad_norm": 0.07861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0301,
+ "step": 9777
+ },
+ {
+ "epoch": 31.23961661341853,
+ "grad_norm": 0.1142578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0443,
+ "step": 9778
+ },
+ {
+ "epoch": 31.242811501597444,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0392,
+ "step": 9779
+ },
+ {
+ "epoch": 31.24600638977636,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0413,
+ "step": 9780
+ },
+ {
+ "epoch": 31.249201277955272,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0348,
+ "step": 9781
+ },
+ {
+ "epoch": 31.252396166134186,
+ "grad_norm": 0.08984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0358,
+ "step": 9782
+ },
+ {
+ "epoch": 31.2555910543131,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0204,
+ "step": 9783
+ },
+ {
+ "epoch": 31.25878594249201,
+ "grad_norm": 0.08203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0419,
+ "step": 9784
+ },
+ {
+ "epoch": 31.261980830670925,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0414,
+ "step": 9785
+ },
+ {
+ "epoch": 31.26517571884984,
+ "grad_norm": 0.130859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0257,
+ "step": 9786
+ },
+ {
+ "epoch": 31.268370607028753,
+ "grad_norm": 0.1494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0379,
+ "step": 9787
+ },
+ {
+ "epoch": 31.271565495207668,
+ "grad_norm": 0.1533203125,
+ "learning_rate": 0.0005,
+ "loss": 1.034,
+ "step": 9788
+ },
+ {
+ "epoch": 31.27476038338658,
+ "grad_norm": 0.26953125,
+ "learning_rate": 0.0005,
+ "loss": 1.041,
+ "step": 9789
+ },
+ {
+ "epoch": 31.277955271565496,
+ "grad_norm": 0.10498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0566,
+ "step": 9790
+ },
+ {
+ "epoch": 31.28115015974441,
+ "grad_norm": 0.10595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0263,
+ "step": 9791
+ },
+ {
+ "epoch": 31.284345047923324,
+ "grad_norm": 0.10009765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0228,
+ "step": 9792
+ },
+ {
+ "epoch": 31.287539936102238,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0298,
+ "step": 9793
+ },
+ {
+ "epoch": 31.29073482428115,
+ "grad_norm": 0.1083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0325,
+ "step": 9794
+ },
+ {
+ "epoch": 31.293929712460063,
+ "grad_norm": 0.0908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0354,
+ "step": 9795
+ },
+ {
+ "epoch": 31.297124600638977,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0419,
+ "step": 9796
+ },
+ {
+ "epoch": 31.30031948881789,
+ "grad_norm": 0.08935546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0337,
+ "step": 9797
+ },
+ {
+ "epoch": 31.303514376996805,
+ "grad_norm": 0.1123046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0394,
+ "step": 9798
+ },
+ {
+ "epoch": 31.30670926517572,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0308,
+ "step": 9799
+ },
+ {
+ "epoch": 31.309904153354633,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0483,
+ "step": 9800
+ },
+ {
+ "epoch": 31.313099041533548,
+ "grad_norm": 0.10400390625,
+ "learning_rate": 0.0005,
+ "loss": 1.04,
+ "step": 9801
+ },
+ {
+ "epoch": 31.31629392971246,
+ "grad_norm": 0.08447265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0371,
+ "step": 9802
+ },
+ {
+ "epoch": 31.319488817891372,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0356,
+ "step": 9803
+ },
+ {
+ "epoch": 31.322683706070286,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0379,
+ "step": 9804
+ },
+ {
+ "epoch": 31.3258785942492,
+ "grad_norm": 0.06005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0283,
+ "step": 9805
+ },
+ {
+ "epoch": 31.329073482428115,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0342,
+ "step": 9806
+ },
+ {
+ "epoch": 31.33226837060703,
+ "grad_norm": 0.0595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0357,
+ "step": 9807
+ },
+ {
+ "epoch": 31.335463258785943,
+ "grad_norm": 0.0576171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0354,
+ "step": 9808
+ },
+ {
+ "epoch": 31.338658146964857,
+ "grad_norm": 0.050048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0399,
+ "step": 9809
+ },
+ {
+ "epoch": 31.34185303514377,
+ "grad_norm": 0.055908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0395,
+ "step": 9810
+ },
+ {
+ "epoch": 31.345047923322685,
+ "grad_norm": 0.05029296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0291,
+ "step": 9811
+ },
+ {
+ "epoch": 31.3482428115016,
+ "grad_norm": 0.056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0406,
+ "step": 9812
+ },
+ {
+ "epoch": 31.35143769968051,
+ "grad_norm": 0.05224609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0416,
+ "step": 9813
+ },
+ {
+ "epoch": 31.354632587859424,
+ "grad_norm": 0.056884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0321,
+ "step": 9814
+ },
+ {
+ "epoch": 31.357827476038338,
+ "grad_norm": 0.04541015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0378,
+ "step": 9815
+ },
+ {
+ "epoch": 31.361022364217252,
+ "grad_norm": 0.059326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.03,
+ "step": 9816
+ },
+ {
+ "epoch": 31.364217252396166,
+ "grad_norm": 0.048095703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0269,
+ "step": 9817
+ },
+ {
+ "epoch": 31.36741214057508,
+ "grad_norm": 0.08056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0391,
+ "step": 9818
+ },
+ {
+ "epoch": 31.370607028753994,
+ "grad_norm": 0.052001953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0432,
+ "step": 9819
+ },
+ {
+ "epoch": 31.37380191693291,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0419,
+ "step": 9820
+ },
+ {
+ "epoch": 31.376996805111823,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0452,
+ "step": 9821
+ },
+ {
+ "epoch": 31.380191693290733,
+ "grad_norm": 0.05908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0405,
+ "step": 9822
+ },
+ {
+ "epoch": 31.383386581469647,
+ "grad_norm": 0.08349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0397,
+ "step": 9823
+ },
+ {
+ "epoch": 31.38658146964856,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0363,
+ "step": 9824
+ },
+ {
+ "epoch": 31.389776357827476,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0433,
+ "step": 9825
+ },
+ {
+ "epoch": 31.39297124600639,
+ "grad_norm": 0.057373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0426,
+ "step": 9826
+ },
+ {
+ "epoch": 31.396166134185304,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0388,
+ "step": 9827
+ },
+ {
+ "epoch": 31.399361022364218,
+ "grad_norm": 0.048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0394,
+ "step": 9828
+ },
+ {
+ "epoch": 31.402555910543132,
+ "grad_norm": 0.060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0348,
+ "step": 9829
+ },
+ {
+ "epoch": 31.405750798722046,
+ "grad_norm": 0.048095703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0394,
+ "step": 9830
+ },
+ {
+ "epoch": 31.408945686900957,
+ "grad_norm": 0.056884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0478,
+ "step": 9831
+ },
+ {
+ "epoch": 31.41214057507987,
+ "grad_norm": 0.054931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0326,
+ "step": 9832
+ },
+ {
+ "epoch": 31.415335463258785,
+ "grad_norm": 0.05029296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0364,
+ "step": 9833
+ },
+ {
+ "epoch": 31.4185303514377,
+ "grad_norm": 0.05712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0324,
+ "step": 9834
+ },
+ {
+ "epoch": 31.421725239616613,
+ "grad_norm": 0.05859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0415,
+ "step": 9835
+ },
+ {
+ "epoch": 31.424920127795527,
+ "grad_norm": 0.0595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0429,
+ "step": 9836
+ },
+ {
+ "epoch": 31.42811501597444,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0427,
+ "step": 9837
+ },
+ {
+ "epoch": 31.431309904153355,
+ "grad_norm": 0.049072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0477,
+ "step": 9838
+ },
+ {
+ "epoch": 31.43450479233227,
+ "grad_norm": 0.05126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.034,
+ "step": 9839
+ },
+ {
+ "epoch": 31.437699680511184,
+ "grad_norm": 0.059326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0422,
+ "step": 9840
+ },
+ {
+ "epoch": 31.440894568690094,
+ "grad_norm": 0.0537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0376,
+ "step": 9841
+ },
+ {
+ "epoch": 31.44408945686901,
+ "grad_norm": 0.059814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0403,
+ "step": 9842
+ },
+ {
+ "epoch": 31.447284345047922,
+ "grad_norm": 0.058349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0324,
+ "step": 9843
+ },
+ {
+ "epoch": 31.450479233226837,
+ "grad_norm": 0.050537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0369,
+ "step": 9844
+ },
+ {
+ "epoch": 31.45367412140575,
+ "grad_norm": 0.060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.044,
+ "step": 9845
+ },
+ {
+ "epoch": 31.456869009584665,
+ "grad_norm": 0.06103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0465,
+ "step": 9846
+ },
+ {
+ "epoch": 31.46006389776358,
+ "grad_norm": 0.059326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0356,
+ "step": 9847
+ },
+ {
+ "epoch": 31.463258785942493,
+ "grad_norm": 0.08056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0399,
+ "step": 9848
+ },
+ {
+ "epoch": 31.466453674121407,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0408,
+ "step": 9849
+ },
+ {
+ "epoch": 31.46964856230032,
+ "grad_norm": 0.0595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0414,
+ "step": 9850
+ },
+ {
+ "epoch": 31.472843450479232,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0296,
+ "step": 9851
+ },
+ {
+ "epoch": 31.476038338658146,
+ "grad_norm": 0.1005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0528,
+ "step": 9852
+ },
+ {
+ "epoch": 31.47923322683706,
+ "grad_norm": 0.0849609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0323,
+ "step": 9853
+ },
+ {
+ "epoch": 31.482428115015974,
+ "grad_norm": 0.08251953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0329,
+ "step": 9854
+ },
+ {
+ "epoch": 31.48562300319489,
+ "grad_norm": 0.0908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0406,
+ "step": 9855
+ },
+ {
+ "epoch": 31.488817891373802,
+ "grad_norm": 0.056396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0293,
+ "step": 9856
+ },
+ {
+ "epoch": 31.492012779552716,
+ "grad_norm": 0.05859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0342,
+ "step": 9857
+ },
+ {
+ "epoch": 31.49520766773163,
+ "grad_norm": 0.05908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0307,
+ "step": 9858
+ },
+ {
+ "epoch": 31.498402555910545,
+ "grad_norm": 0.08642578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0364,
+ "step": 9859
+ },
+ {
+ "epoch": 31.501597444089455,
+ "grad_norm": 0.06201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0373,
+ "step": 9860
+ },
+ {
+ "epoch": 31.50479233226837,
+ "grad_norm": 0.08349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0318,
+ "step": 9861
+ },
+ {
+ "epoch": 31.507987220447284,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0486,
+ "step": 9862
+ },
+ {
+ "epoch": 31.511182108626198,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0439,
+ "step": 9863
+ },
+ {
+ "epoch": 31.51437699680511,
+ "grad_norm": 0.061279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0494,
+ "step": 9864
+ },
+ {
+ "epoch": 31.517571884984026,
+ "grad_norm": 0.059814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0415,
+ "step": 9865
+ },
+ {
+ "epoch": 31.52076677316294,
+ "grad_norm": 0.08203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0353,
+ "step": 9866
+ },
+ {
+ "epoch": 31.523961661341854,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0387,
+ "step": 9867
+ },
+ {
+ "epoch": 31.527156549520768,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0368,
+ "step": 9868
+ },
+ {
+ "epoch": 31.53035143769968,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0419,
+ "step": 9869
+ },
+ {
+ "epoch": 31.533546325878593,
+ "grad_norm": 0.056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0408,
+ "step": 9870
+ },
+ {
+ "epoch": 31.536741214057507,
+ "grad_norm": 0.0478515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0407,
+ "step": 9871
+ },
+ {
+ "epoch": 31.53993610223642,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0303,
+ "step": 9872
+ },
+ {
+ "epoch": 31.543130990415335,
+ "grad_norm": 0.054931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0597,
+ "step": 9873
+ },
+ {
+ "epoch": 31.54632587859425,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.038,
+ "step": 9874
+ },
+ {
+ "epoch": 31.549520766773163,
+ "grad_norm": 0.04931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0328,
+ "step": 9875
+ },
+ {
+ "epoch": 31.552715654952078,
+ "grad_norm": 0.0458984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0341,
+ "step": 9876
+ },
+ {
+ "epoch": 31.55591054313099,
+ "grad_norm": 0.052734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0352,
+ "step": 9877
+ },
+ {
+ "epoch": 31.559105431309906,
+ "grad_norm": 0.048583984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0353,
+ "step": 9878
+ },
+ {
+ "epoch": 31.562300319488816,
+ "grad_norm": 0.05126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0469,
+ "step": 9879
+ },
+ {
+ "epoch": 31.56549520766773,
+ "grad_norm": 0.050048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0373,
+ "step": 9880
+ },
+ {
+ "epoch": 31.568690095846645,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0337,
+ "step": 9881
+ },
+ {
+ "epoch": 31.57188498402556,
+ "grad_norm": 0.05224609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0282,
+ "step": 9882
+ },
+ {
+ "epoch": 31.575079872204473,
+ "grad_norm": 0.051025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0436,
+ "step": 9883
+ },
+ {
+ "epoch": 31.578274760383387,
+ "grad_norm": 0.056396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0347,
+ "step": 9884
+ },
+ {
+ "epoch": 31.5814696485623,
+ "grad_norm": 0.05859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0406,
+ "step": 9885
+ },
+ {
+ "epoch": 31.584664536741215,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0322,
+ "step": 9886
+ },
+ {
+ "epoch": 31.58785942492013,
+ "grad_norm": 0.09619140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0363,
+ "step": 9887
+ },
+ {
+ "epoch": 31.591054313099043,
+ "grad_norm": 0.09326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0454,
+ "step": 9888
+ },
+ {
+ "epoch": 31.594249201277954,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0379,
+ "step": 9889
+ },
+ {
+ "epoch": 31.597444089456868,
+ "grad_norm": 0.05859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0356,
+ "step": 9890
+ },
+ {
+ "epoch": 31.600638977635782,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0413,
+ "step": 9891
+ },
+ {
+ "epoch": 31.603833865814696,
+ "grad_norm": 0.061767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0357,
+ "step": 9892
+ },
+ {
+ "epoch": 31.60702875399361,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0255,
+ "step": 9893
+ },
+ {
+ "epoch": 31.610223642172524,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0377,
+ "step": 9894
+ },
+ {
+ "epoch": 31.61341853035144,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0418,
+ "step": 9895
+ },
+ {
+ "epoch": 31.616613418530353,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.04,
+ "step": 9896
+ },
+ {
+ "epoch": 31.619808306709267,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0307,
+ "step": 9897
+ },
+ {
+ "epoch": 31.623003194888177,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0331,
+ "step": 9898
+ },
+ {
+ "epoch": 31.62619808306709,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0352,
+ "step": 9899
+ },
+ {
+ "epoch": 31.629392971246006,
+ "grad_norm": 0.05712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0372,
+ "step": 9900
+ },
+ {
+ "epoch": 31.63258785942492,
+ "grad_norm": 0.08349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0478,
+ "step": 9901
+ },
+ {
+ "epoch": 31.635782747603834,
+ "grad_norm": 0.06201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0316,
+ "step": 9902
+ },
+ {
+ "epoch": 31.638977635782748,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0412,
+ "step": 9903
+ },
+ {
+ "epoch": 31.642172523961662,
+ "grad_norm": 0.05517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0311,
+ "step": 9904
+ },
+ {
+ "epoch": 31.645367412140576,
+ "grad_norm": 0.103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0506,
+ "step": 9905
+ },
+ {
+ "epoch": 31.64856230031949,
+ "grad_norm": 0.0927734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0354,
+ "step": 9906
+ },
+ {
+ "epoch": 31.6517571884984,
+ "grad_norm": 0.09033203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0373,
+ "step": 9907
+ },
+ {
+ "epoch": 31.654952076677315,
+ "grad_norm": 0.08740234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0421,
+ "step": 9908
+ },
+ {
+ "epoch": 31.65814696485623,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0325,
+ "step": 9909
+ },
+ {
+ "epoch": 31.661341853035143,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0296,
+ "step": 9910
+ },
+ {
+ "epoch": 31.664536741214057,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.041,
+ "step": 9911
+ },
+ {
+ "epoch": 31.66773162939297,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0486,
+ "step": 9912
+ },
+ {
+ "epoch": 31.670926517571885,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0316,
+ "step": 9913
+ },
+ {
+ "epoch": 31.6741214057508,
+ "grad_norm": 0.05517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0281,
+ "step": 9914
+ },
+ {
+ "epoch": 31.677316293929714,
+ "grad_norm": 0.061279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0369,
+ "step": 9915
+ },
+ {
+ "epoch": 31.680511182108628,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0459,
+ "step": 9916
+ },
+ {
+ "epoch": 31.68370607028754,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0438,
+ "step": 9917
+ },
+ {
+ "epoch": 31.686900958466452,
+ "grad_norm": 0.059814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.03,
+ "step": 9918
+ },
+ {
+ "epoch": 31.690095846645367,
+ "grad_norm": 0.09130859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0232,
+ "step": 9919
+ },
+ {
+ "epoch": 31.69329073482428,
+ "grad_norm": 0.05712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0502,
+ "step": 9920
+ },
+ {
+ "epoch": 31.696485623003195,
+ "grad_norm": 0.09521484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0259,
+ "step": 9921
+ },
+ {
+ "epoch": 31.69968051118211,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0415,
+ "step": 9922
+ },
+ {
+ "epoch": 31.702875399361023,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0409,
+ "step": 9923
+ },
+ {
+ "epoch": 31.706070287539937,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0443,
+ "step": 9924
+ },
+ {
+ "epoch": 31.70926517571885,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0405,
+ "step": 9925
+ },
+ {
+ "epoch": 31.712460063897765,
+ "grad_norm": 0.06005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0432,
+ "step": 9926
+ },
+ {
+ "epoch": 31.715654952076676,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0374,
+ "step": 9927
+ },
+ {
+ "epoch": 31.71884984025559,
+ "grad_norm": 0.08203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0458,
+ "step": 9928
+ },
+ {
+ "epoch": 31.722044728434504,
+ "grad_norm": 0.05517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.034,
+ "step": 9929
+ },
+ {
+ "epoch": 31.72523961661342,
+ "grad_norm": 0.08203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0485,
+ "step": 9930
+ },
+ {
+ "epoch": 31.728434504792332,
+ "grad_norm": 0.061767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0324,
+ "step": 9931
+ },
+ {
+ "epoch": 31.731629392971247,
+ "grad_norm": 0.05517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0376,
+ "step": 9932
+ },
+ {
+ "epoch": 31.73482428115016,
+ "grad_norm": 0.05419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0338,
+ "step": 9933
+ },
+ {
+ "epoch": 31.738019169329075,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0364,
+ "step": 9934
+ },
+ {
+ "epoch": 31.74121405750799,
+ "grad_norm": 0.050537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0372,
+ "step": 9935
+ },
+ {
+ "epoch": 31.7444089456869,
+ "grad_norm": 0.052001953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0494,
+ "step": 9936
+ },
+ {
+ "epoch": 31.747603833865814,
+ "grad_norm": 0.057373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0437,
+ "step": 9937
+ },
+ {
+ "epoch": 31.750798722044728,
+ "grad_norm": 0.052978515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0383,
+ "step": 9938
+ },
+ {
+ "epoch": 31.75399361022364,
+ "grad_norm": 0.05029296875,
+ "learning_rate": 0.0005,
+ "loss": 1.036,
+ "step": 9939
+ },
+ {
+ "epoch": 31.757188498402556,
+ "grad_norm": 0.0517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.045,
+ "step": 9940
+ },
+ {
+ "epoch": 31.76038338658147,
+ "grad_norm": 0.05029296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0335,
+ "step": 9941
+ },
+ {
+ "epoch": 31.763578274760384,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0121,
+ "step": 9942
+ },
+ {
+ "epoch": 31.766773162939298,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0201,
+ "step": 9943
+ },
+ {
+ "epoch": 31.769968051118212,
+ "grad_norm": 0.052001953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0307,
+ "step": 9944
+ },
+ {
+ "epoch": 31.773162939297123,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0368,
+ "step": 9945
+ },
+ {
+ "epoch": 31.776357827476037,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0374,
+ "step": 9946
+ },
+ {
+ "epoch": 31.77955271565495,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0347,
+ "step": 9947
+ },
+ {
+ "epoch": 31.782747603833865,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0354,
+ "step": 9948
+ },
+ {
+ "epoch": 31.78594249201278,
+ "grad_norm": 0.059326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0256,
+ "step": 9949
+ },
+ {
+ "epoch": 31.789137380191693,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0409,
+ "step": 9950
+ },
+ {
+ "epoch": 31.792332268370608,
+ "grad_norm": 0.044189453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0322,
+ "step": 9951
+ },
+ {
+ "epoch": 31.79552715654952,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0488,
+ "step": 9952
+ },
+ {
+ "epoch": 31.798722044728436,
+ "grad_norm": 0.049560546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0381,
+ "step": 9953
+ },
+ {
+ "epoch": 31.80191693290735,
+ "grad_norm": 0.052978515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0444,
+ "step": 9954
+ },
+ {
+ "epoch": 31.80511182108626,
+ "grad_norm": 0.0458984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0401,
+ "step": 9955
+ },
+ {
+ "epoch": 31.808306709265175,
+ "grad_norm": 0.058837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0397,
+ "step": 9956
+ },
+ {
+ "epoch": 31.81150159744409,
+ "grad_norm": 0.052978515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0237,
+ "step": 9957
+ },
+ {
+ "epoch": 31.814696485623003,
+ "grad_norm": 0.05712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0365,
+ "step": 9958
+ },
+ {
+ "epoch": 31.817891373801917,
+ "grad_norm": 0.0537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0422,
+ "step": 9959
+ },
+ {
+ "epoch": 31.82108626198083,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0368,
+ "step": 9960
+ },
+ {
+ "epoch": 31.824281150159745,
+ "grad_norm": 0.05517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0429,
+ "step": 9961
+ },
+ {
+ "epoch": 31.82747603833866,
+ "grad_norm": 0.05615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0377,
+ "step": 9962
+ },
+ {
+ "epoch": 31.830670926517573,
+ "grad_norm": 0.0537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0379,
+ "step": 9963
+ },
+ {
+ "epoch": 31.833865814696484,
+ "grad_norm": 0.0498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0516,
+ "step": 9964
+ },
+ {
+ "epoch": 31.837060702875398,
+ "grad_norm": 0.05517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0195,
+ "step": 9965
+ },
+ {
+ "epoch": 31.840255591054312,
+ "grad_norm": 0.047119140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0403,
+ "step": 9966
+ },
+ {
+ "epoch": 31.843450479233226,
+ "grad_norm": 0.0478515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0326,
+ "step": 9967
+ },
+ {
+ "epoch": 31.84664536741214,
+ "grad_norm": 0.05517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0358,
+ "step": 9968
+ },
+ {
+ "epoch": 31.849840255591054,
+ "grad_norm": 0.046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0377,
+ "step": 9969
+ },
+ {
+ "epoch": 31.85303514376997,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0241,
+ "step": 9970
+ },
+ {
+ "epoch": 31.856230031948883,
+ "grad_norm": 0.059814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0312,
+ "step": 9971
+ },
+ {
+ "epoch": 31.859424920127797,
+ "grad_norm": 0.045166015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0317,
+ "step": 9972
+ },
+ {
+ "epoch": 31.86261980830671,
+ "grad_norm": 0.056396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0394,
+ "step": 9973
+ },
+ {
+ "epoch": 31.86581469648562,
+ "grad_norm": 0.0595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.04,
+ "step": 9974
+ },
+ {
+ "epoch": 31.869009584664536,
+ "grad_norm": 0.05029296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0385,
+ "step": 9975
+ },
+ {
+ "epoch": 31.87220447284345,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0395,
+ "step": 9976
+ },
+ {
+ "epoch": 31.875399361022364,
+ "grad_norm": 0.048095703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0384,
+ "step": 9977
+ },
+ {
+ "epoch": 31.878594249201278,
+ "grad_norm": 0.12060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0359,
+ "step": 9978
+ },
+ {
+ "epoch": 31.881789137380192,
+ "grad_norm": 0.05615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0441,
+ "step": 9979
+ },
+ {
+ "epoch": 31.884984025559106,
+ "grad_norm": 0.09423828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0433,
+ "step": 9980
+ },
+ {
+ "epoch": 31.88817891373802,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0261,
+ "step": 9981
+ },
+ {
+ "epoch": 31.891373801916934,
+ "grad_norm": 0.1044921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0317,
+ "step": 9982
+ },
+ {
+ "epoch": 31.894568690095845,
+ "grad_norm": 0.10595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0381,
+ "step": 9983
+ },
+ {
+ "epoch": 31.89776357827476,
+ "grad_norm": 0.05322265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0295,
+ "step": 9984
+ },
+ {
+ "epoch": 31.900958466453673,
+ "grad_norm": 0.158203125,
+ "learning_rate": 0.0005,
+ "loss": 1.048,
+ "step": 9985
+ },
+ {
+ "epoch": 31.904153354632587,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.035,
+ "step": 9986
+ },
+ {
+ "epoch": 31.9073482428115,
+ "grad_norm": 0.15234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0285,
+ "step": 9987
+ },
+ {
+ "epoch": 31.910543130990416,
+ "grad_norm": 0.083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0418,
+ "step": 9988
+ },
+ {
+ "epoch": 31.91373801916933,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0359,
+ "step": 9989
+ },
+ {
+ "epoch": 31.916932907348244,
+ "grad_norm": 0.1318359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0323,
+ "step": 9990
+ },
+ {
+ "epoch": 31.920127795527158,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0379,
+ "step": 9991
+ },
+ {
+ "epoch": 31.923322683706072,
+ "grad_norm": 0.1708984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0487,
+ "step": 9992
+ },
+ {
+ "epoch": 31.926517571884983,
+ "grad_norm": 0.1044921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0304,
+ "step": 9993
+ },
+ {
+ "epoch": 31.929712460063897,
+ "grad_norm": 0.126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0343,
+ "step": 9994
+ },
+ {
+ "epoch": 31.93290734824281,
+ "grad_norm": 0.1435546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0345,
+ "step": 9995
+ },
+ {
+ "epoch": 31.936102236421725,
+ "grad_norm": 0.1142578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0356,
+ "step": 9996
+ },
+ {
+ "epoch": 31.93929712460064,
+ "grad_norm": 0.056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0355,
+ "step": 9997
+ },
+ {
+ "epoch": 31.942492012779553,
+ "grad_norm": 0.09375,
+ "learning_rate": 0.0005,
+ "loss": 1.0516,
+ "step": 9998
+ },
+ {
+ "epoch": 31.945686900958467,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0401,
+ "step": 9999
+ },
+ {
+ "epoch": 31.94888178913738,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0234,
+ "step": 10000
+ },
+ {
+ "epoch": 31.952076677316295,
+ "grad_norm": 0.0615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0397,
+ "step": 10001
+ },
+ {
+ "epoch": 31.955271565495206,
+ "grad_norm": 0.0859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0323,
+ "step": 10002
+ },
+ {
+ "epoch": 31.95846645367412,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0331,
+ "step": 10003
+ },
+ {
+ "epoch": 31.961661341853034,
+ "grad_norm": 0.06103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0374,
+ "step": 10004
+ },
+ {
+ "epoch": 31.96485623003195,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0195,
+ "step": 10005
+ },
+ {
+ "epoch": 31.968051118210862,
+ "grad_norm": 0.13671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0364,
+ "step": 10006
+ },
+ {
+ "epoch": 31.971246006389777,
+ "grad_norm": 0.095703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0423,
+ "step": 10007
+ },
+ {
+ "epoch": 31.97444089456869,
+ "grad_norm": 0.1083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0373,
+ "step": 10008
+ },
+ {
+ "epoch": 31.977635782747605,
+ "grad_norm": 0.103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0418,
+ "step": 10009
+ },
+ {
+ "epoch": 31.98083067092652,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0324,
+ "step": 10010
+ },
+ {
+ "epoch": 31.984025559105433,
+ "grad_norm": 0.08251953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0354,
+ "step": 10011
+ },
+ {
+ "epoch": 31.987220447284344,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0351,
+ "step": 10012
+ },
+ {
+ "epoch": 31.990415335463258,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0432,
+ "step": 10013
+ },
+ {
+ "epoch": 31.99361022364217,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0332,
+ "step": 10014
+ },
+ {
+ "epoch": 31.996805111821086,
+ "grad_norm": 0.125,
+ "learning_rate": 0.0005,
+ "loss": 1.0377,
+ "step": 10015
+ },
+ {
+ "epoch": 32.0,
+ "grad_norm": 0.054443359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0451,
+ "step": 10016
+ },
+ {
+ "epoch": 32.00319488817891,
+ "grad_norm": 0.11767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0373,
+ "step": 10017
+ },
+ {
+ "epoch": 32.00638977635783,
+ "grad_norm": 0.05517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.035,
+ "step": 10018
+ },
+ {
+ "epoch": 32.00958466453674,
+ "grad_norm": 0.09619140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0439,
+ "step": 10019
+ },
+ {
+ "epoch": 32.01277955271566,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.047,
+ "step": 10020
+ },
+ {
+ "epoch": 32.01597444089457,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0318,
+ "step": 10021
+ },
+ {
+ "epoch": 32.019169329073485,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0394,
+ "step": 10022
+ },
+ {
+ "epoch": 32.022364217252395,
+ "grad_norm": 0.053955078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0264,
+ "step": 10023
+ },
+ {
+ "epoch": 32.02555910543131,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.035,
+ "step": 10024
+ },
+ {
+ "epoch": 32.02875399361022,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0221,
+ "step": 10025
+ },
+ {
+ "epoch": 32.031948881789134,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0386,
+ "step": 10026
+ },
+ {
+ "epoch": 32.03514376996805,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0377,
+ "step": 10027
+ },
+ {
+ "epoch": 32.03833865814696,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.035,
+ "step": 10028
+ },
+ {
+ "epoch": 32.04153354632588,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0381,
+ "step": 10029
+ },
+ {
+ "epoch": 32.04472843450479,
+ "grad_norm": 0.10400390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0514,
+ "step": 10030
+ },
+ {
+ "epoch": 32.04792332268371,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.033,
+ "step": 10031
+ },
+ {
+ "epoch": 32.05111821086262,
+ "grad_norm": 0.06201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0356,
+ "step": 10032
+ },
+ {
+ "epoch": 32.054313099041536,
+ "grad_norm": 0.0947265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0313,
+ "step": 10033
+ },
+ {
+ "epoch": 32.05750798722045,
+ "grad_norm": 0.058349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0547,
+ "step": 10034
+ },
+ {
+ "epoch": 32.06070287539936,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.045,
+ "step": 10035
+ },
+ {
+ "epoch": 32.063897763578275,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0314,
+ "step": 10036
+ },
+ {
+ "epoch": 32.067092651757186,
+ "grad_norm": 0.054931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0319,
+ "step": 10037
+ },
+ {
+ "epoch": 32.0702875399361,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0391,
+ "step": 10038
+ },
+ {
+ "epoch": 32.073482428115014,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.041,
+ "step": 10039
+ },
+ {
+ "epoch": 32.07667731629393,
+ "grad_norm": 0.05810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0281,
+ "step": 10040
+ },
+ {
+ "epoch": 32.07987220447284,
+ "grad_norm": 0.1025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0436,
+ "step": 10041
+ },
+ {
+ "epoch": 32.08306709265176,
+ "grad_norm": 0.0498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0249,
+ "step": 10042
+ },
+ {
+ "epoch": 32.08626198083067,
+ "grad_norm": 0.05810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0344,
+ "step": 10043
+ },
+ {
+ "epoch": 32.08945686900959,
+ "grad_norm": 0.05712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0523,
+ "step": 10044
+ },
+ {
+ "epoch": 32.0926517571885,
+ "grad_norm": 0.053466796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0341,
+ "step": 10045
+ },
+ {
+ "epoch": 32.09584664536741,
+ "grad_norm": 0.0810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0389,
+ "step": 10046
+ },
+ {
+ "epoch": 32.09904153354633,
+ "grad_norm": 0.051513671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0362,
+ "step": 10047
+ },
+ {
+ "epoch": 32.10223642172524,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0301,
+ "step": 10048
+ },
+ {
+ "epoch": 32.105431309904155,
+ "grad_norm": 0.058837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0432,
+ "step": 10049
+ },
+ {
+ "epoch": 32.108626198083066,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0355,
+ "step": 10050
+ },
+ {
+ "epoch": 32.11182108626198,
+ "grad_norm": 0.060791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0427,
+ "step": 10051
+ },
+ {
+ "epoch": 32.115015974440894,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0362,
+ "step": 10052
+ },
+ {
+ "epoch": 32.11821086261981,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0333,
+ "step": 10053
+ },
+ {
+ "epoch": 32.12140575079872,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0268,
+ "step": 10054
+ },
+ {
+ "epoch": 32.12460063897763,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0301,
+ "step": 10055
+ },
+ {
+ "epoch": 32.12779552715655,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0317,
+ "step": 10056
+ },
+ {
+ "epoch": 32.13099041533546,
+ "grad_norm": 0.0830078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0351,
+ "step": 10057
+ },
+ {
+ "epoch": 32.13418530351438,
+ "grad_norm": 0.050537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0364,
+ "step": 10058
+ },
+ {
+ "epoch": 32.13738019169329,
+ "grad_norm": 0.099609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0533,
+ "step": 10059
+ },
+ {
+ "epoch": 32.14057507987221,
+ "grad_norm": 0.052734375,
+ "learning_rate": 0.0005,
+ "loss": 1.036,
+ "step": 10060
+ },
+ {
+ "epoch": 32.14376996805112,
+ "grad_norm": 0.06005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0474,
+ "step": 10061
+ },
+ {
+ "epoch": 32.146964856230035,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0446,
+ "step": 10062
+ },
+ {
+ "epoch": 32.150159744408946,
+ "grad_norm": 0.09228515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0268,
+ "step": 10063
+ },
+ {
+ "epoch": 32.153354632587856,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0445,
+ "step": 10064
+ },
+ {
+ "epoch": 32.156549520766774,
+ "grad_norm": 0.0986328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0388,
+ "step": 10065
+ },
+ {
+ "epoch": 32.159744408945684,
+ "grad_norm": 0.060302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0495,
+ "step": 10066
+ },
+ {
+ "epoch": 32.1629392971246,
+ "grad_norm": 0.080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.041,
+ "step": 10067
+ },
+ {
+ "epoch": 32.16613418530351,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0377,
+ "step": 10068
+ },
+ {
+ "epoch": 32.16932907348243,
+ "grad_norm": 0.059814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.037,
+ "step": 10069
+ },
+ {
+ "epoch": 32.17252396166134,
+ "grad_norm": 0.080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0286,
+ "step": 10070
+ },
+ {
+ "epoch": 32.17571884984026,
+ "grad_norm": 0.056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0148,
+ "step": 10071
+ },
+ {
+ "epoch": 32.17891373801917,
+ "grad_norm": 0.058837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0415,
+ "step": 10072
+ },
+ {
+ "epoch": 32.18210862619808,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0391,
+ "step": 10073
+ },
+ {
+ "epoch": 32.185303514377,
+ "grad_norm": 0.062255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.032,
+ "step": 10074
+ },
+ {
+ "epoch": 32.18849840255591,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0391,
+ "step": 10075
+ },
+ {
+ "epoch": 32.191693290734825,
+ "grad_norm": 0.05419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0372,
+ "step": 10076
+ },
+ {
+ "epoch": 32.194888178913736,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0389,
+ "step": 10077
+ },
+ {
+ "epoch": 32.198083067092654,
+ "grad_norm": 0.053466796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0371,
+ "step": 10078
+ },
+ {
+ "epoch": 32.201277955271564,
+ "grad_norm": 0.05859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0312,
+ "step": 10079
+ },
+ {
+ "epoch": 32.20447284345048,
+ "grad_norm": 0.046630859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0482,
+ "step": 10080
+ },
+ {
+ "epoch": 32.20766773162939,
+ "grad_norm": 0.04833984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0387,
+ "step": 10081
+ },
+ {
+ "epoch": 32.21086261980831,
+ "grad_norm": 0.0478515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0311,
+ "step": 10082
+ },
+ {
+ "epoch": 32.21405750798722,
+ "grad_norm": 0.05029296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0391,
+ "step": 10083
+ },
+ {
+ "epoch": 32.21725239616613,
+ "grad_norm": 0.050048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0303,
+ "step": 10084
+ },
+ {
+ "epoch": 32.22044728434505,
+ "grad_norm": 0.053955078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0503,
+ "step": 10085
+ },
+ {
+ "epoch": 32.22364217252396,
+ "grad_norm": 0.0576171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0373,
+ "step": 10086
+ },
+ {
+ "epoch": 32.22683706070288,
+ "grad_norm": 0.04736328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0279,
+ "step": 10087
+ },
+ {
+ "epoch": 32.23003194888179,
+ "grad_norm": 0.056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.039,
+ "step": 10088
+ },
+ {
+ "epoch": 32.233226837060705,
+ "grad_norm": 0.058349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0423,
+ "step": 10089
+ },
+ {
+ "epoch": 32.236421725239616,
+ "grad_norm": 0.05810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.036,
+ "step": 10090
+ },
+ {
+ "epoch": 32.239616613418534,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0344,
+ "step": 10091
+ },
+ {
+ "epoch": 32.242811501597444,
+ "grad_norm": 0.04931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0408,
+ "step": 10092
+ },
+ {
+ "epoch": 32.246006389776355,
+ "grad_norm": 0.051025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0316,
+ "step": 10093
+ },
+ {
+ "epoch": 32.24920127795527,
+ "grad_norm": 0.0458984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0266,
+ "step": 10094
+ },
+ {
+ "epoch": 32.25239616613418,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.037,
+ "step": 10095
+ },
+ {
+ "epoch": 32.2555910543131,
+ "grad_norm": 0.047119140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0405,
+ "step": 10096
+ },
+ {
+ "epoch": 32.25878594249201,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.042,
+ "step": 10097
+ },
+ {
+ "epoch": 32.26198083067093,
+ "grad_norm": 0.05029296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0318,
+ "step": 10098
+ },
+ {
+ "epoch": 32.26517571884984,
+ "grad_norm": 0.043701171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0334,
+ "step": 10099
+ },
+ {
+ "epoch": 32.26837060702876,
+ "grad_norm": 0.05810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0463,
+ "step": 10100
+ },
+ {
+ "epoch": 32.27156549520767,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0331,
+ "step": 10101
+ },
+ {
+ "epoch": 32.27476038338658,
+ "grad_norm": 0.062255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0295,
+ "step": 10102
+ },
+ {
+ "epoch": 32.277955271565496,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0256,
+ "step": 10103
+ },
+ {
+ "epoch": 32.281150159744406,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0342,
+ "step": 10104
+ },
+ {
+ "epoch": 32.284345047923324,
+ "grad_norm": 0.062255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0278,
+ "step": 10105
+ },
+ {
+ "epoch": 32.287539936102235,
+ "grad_norm": 0.11962890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0401,
+ "step": 10106
+ },
+ {
+ "epoch": 32.29073482428115,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0386,
+ "step": 10107
+ },
+ {
+ "epoch": 32.29392971246006,
+ "grad_norm": 0.166015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0368,
+ "step": 10108
+ },
+ {
+ "epoch": 32.29712460063898,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0385,
+ "step": 10109
+ },
+ {
+ "epoch": 32.30031948881789,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.036,
+ "step": 10110
+ },
+ {
+ "epoch": 32.3035143769968,
+ "grad_norm": 0.055908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0387,
+ "step": 10111
+ },
+ {
+ "epoch": 32.30670926517572,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0333,
+ "step": 10112
+ },
+ {
+ "epoch": 32.30990415335463,
+ "grad_norm": 0.05322265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0397,
+ "step": 10113
+ },
+ {
+ "epoch": 32.31309904153355,
+ "grad_norm": 0.0537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0289,
+ "step": 10114
+ },
+ {
+ "epoch": 32.31629392971246,
+ "grad_norm": 0.059326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0373,
+ "step": 10115
+ },
+ {
+ "epoch": 32.319488817891376,
+ "grad_norm": 0.05126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0381,
+ "step": 10116
+ },
+ {
+ "epoch": 32.322683706070286,
+ "grad_norm": 0.05810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0351,
+ "step": 10117
+ },
+ {
+ "epoch": 32.325878594249204,
+ "grad_norm": 0.057373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0257,
+ "step": 10118
+ },
+ {
+ "epoch": 32.329073482428115,
+ "grad_norm": 0.05908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0401,
+ "step": 10119
+ },
+ {
+ "epoch": 32.33226837060703,
+ "grad_norm": 0.056396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0453,
+ "step": 10120
+ },
+ {
+ "epoch": 32.33546325878594,
+ "grad_norm": 0.09521484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0277,
+ "step": 10121
+ },
+ {
+ "epoch": 32.33865814696485,
+ "grad_norm": 0.054931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0307,
+ "step": 10122
+ },
+ {
+ "epoch": 32.34185303514377,
+ "grad_norm": 0.0576171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0474,
+ "step": 10123
+ },
+ {
+ "epoch": 32.34504792332268,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0321,
+ "step": 10124
+ },
+ {
+ "epoch": 32.3482428115016,
+ "grad_norm": 0.060302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0369,
+ "step": 10125
+ },
+ {
+ "epoch": 32.35143769968051,
+ "grad_norm": 0.049560546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0297,
+ "step": 10126
+ },
+ {
+ "epoch": 32.35463258785943,
+ "grad_norm": 0.0517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0308,
+ "step": 10127
+ },
+ {
+ "epoch": 32.35782747603834,
+ "grad_norm": 0.0517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0408,
+ "step": 10128
+ },
+ {
+ "epoch": 32.361022364217256,
+ "grad_norm": 0.046142578125,
+ "learning_rate": 0.0005,
+ "loss": 1.026,
+ "step": 10129
+ },
+ {
+ "epoch": 32.364217252396166,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0272,
+ "step": 10130
+ },
+ {
+ "epoch": 32.36741214057508,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0357,
+ "step": 10131
+ },
+ {
+ "epoch": 32.370607028753994,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0379,
+ "step": 10132
+ },
+ {
+ "epoch": 32.373801916932905,
+ "grad_norm": 0.10302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0247,
+ "step": 10133
+ },
+ {
+ "epoch": 32.37699680511182,
+ "grad_norm": 0.054931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0277,
+ "step": 10134
+ },
+ {
+ "epoch": 32.38019169329073,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0403,
+ "step": 10135
+ },
+ {
+ "epoch": 32.38338658146965,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0346,
+ "step": 10136
+ },
+ {
+ "epoch": 32.38658146964856,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0435,
+ "step": 10137
+ },
+ {
+ "epoch": 32.38977635782748,
+ "grad_norm": 0.0517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0409,
+ "step": 10138
+ },
+ {
+ "epoch": 32.39297124600639,
+ "grad_norm": 0.05029296875,
+ "learning_rate": 0.0005,
+ "loss": 1.043,
+ "step": 10139
+ },
+ {
+ "epoch": 32.3961661341853,
+ "grad_norm": 0.054443359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0365,
+ "step": 10140
+ },
+ {
+ "epoch": 32.39936102236422,
+ "grad_norm": 0.049560546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0279,
+ "step": 10141
+ },
+ {
+ "epoch": 32.40255591054313,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0472,
+ "step": 10142
+ },
+ {
+ "epoch": 32.405750798722046,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0341,
+ "step": 10143
+ },
+ {
+ "epoch": 32.40894568690096,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0263,
+ "step": 10144
+ },
+ {
+ "epoch": 32.412140575079874,
+ "grad_norm": 0.0869140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0352,
+ "step": 10145
+ },
+ {
+ "epoch": 32.415335463258785,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0404,
+ "step": 10146
+ },
+ {
+ "epoch": 32.4185303514377,
+ "grad_norm": 0.0556640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0353,
+ "step": 10147
+ },
+ {
+ "epoch": 32.42172523961661,
+ "grad_norm": 0.0556640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0389,
+ "step": 10148
+ },
+ {
+ "epoch": 32.424920127795524,
+ "grad_norm": 0.09912109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0331,
+ "step": 10149
+ },
+ {
+ "epoch": 32.42811501597444,
+ "grad_norm": 0.04931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0383,
+ "step": 10150
+ },
+ {
+ "epoch": 32.43130990415335,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0434,
+ "step": 10151
+ },
+ {
+ "epoch": 32.43450479233227,
+ "grad_norm": 0.09716796875,
+ "learning_rate": 0.0005,
+ "loss": 1.043,
+ "step": 10152
+ },
+ {
+ "epoch": 32.43769968051118,
+ "grad_norm": 0.06005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0415,
+ "step": 10153
+ },
+ {
+ "epoch": 32.4408945686901,
+ "grad_norm": 0.08740234375,
+ "learning_rate": 0.0005,
+ "loss": 1.025,
+ "step": 10154
+ },
+ {
+ "epoch": 32.44408945686901,
+ "grad_norm": 0.123046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0352,
+ "step": 10155
+ },
+ {
+ "epoch": 32.447284345047926,
+ "grad_norm": 0.1279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0329,
+ "step": 10156
+ },
+ {
+ "epoch": 32.45047923322684,
+ "grad_norm": 0.091796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0415,
+ "step": 10157
+ },
+ {
+ "epoch": 32.453674121405754,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0227,
+ "step": 10158
+ },
+ {
+ "epoch": 32.456869009584665,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0339,
+ "step": 10159
+ },
+ {
+ "epoch": 32.460063897763575,
+ "grad_norm": 0.0888671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0395,
+ "step": 10160
+ },
+ {
+ "epoch": 32.46325878594249,
+ "grad_norm": 0.0595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0347,
+ "step": 10161
+ },
+ {
+ "epoch": 32.466453674121404,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0511,
+ "step": 10162
+ },
+ {
+ "epoch": 32.46964856230032,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0333,
+ "step": 10163
+ },
+ {
+ "epoch": 32.47284345047923,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0315,
+ "step": 10164
+ },
+ {
+ "epoch": 32.47603833865815,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0273,
+ "step": 10165
+ },
+ {
+ "epoch": 32.47923322683706,
+ "grad_norm": 0.0927734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0304,
+ "step": 10166
+ },
+ {
+ "epoch": 32.48242811501598,
+ "grad_norm": 0.0810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0405,
+ "step": 10167
+ },
+ {
+ "epoch": 32.48562300319489,
+ "grad_norm": 0.10693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0354,
+ "step": 10168
+ },
+ {
+ "epoch": 32.4888178913738,
+ "grad_norm": 0.08349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0401,
+ "step": 10169
+ },
+ {
+ "epoch": 32.49201277955272,
+ "grad_norm": 0.048583984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0382,
+ "step": 10170
+ },
+ {
+ "epoch": 32.49520766773163,
+ "grad_norm": 0.054931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.047,
+ "step": 10171
+ },
+ {
+ "epoch": 32.498402555910545,
+ "grad_norm": 0.051513671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0318,
+ "step": 10172
+ },
+ {
+ "epoch": 32.501597444089455,
+ "grad_norm": 0.060302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.025,
+ "step": 10173
+ },
+ {
+ "epoch": 32.50479233226837,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0337,
+ "step": 10174
+ },
+ {
+ "epoch": 32.50798722044728,
+ "grad_norm": 0.060302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0327,
+ "step": 10175
+ },
+ {
+ "epoch": 32.5111821086262,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0253,
+ "step": 10176
+ },
+ {
+ "epoch": 32.51437699680511,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0292,
+ "step": 10177
+ },
+ {
+ "epoch": 32.51757188498402,
+ "grad_norm": 0.058837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0477,
+ "step": 10178
+ },
+ {
+ "epoch": 32.52076677316294,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.025,
+ "step": 10179
+ },
+ {
+ "epoch": 32.52396166134185,
+ "grad_norm": 0.0830078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0245,
+ "step": 10180
+ },
+ {
+ "epoch": 32.52715654952077,
+ "grad_norm": 0.158203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0323,
+ "step": 10181
+ },
+ {
+ "epoch": 32.53035143769968,
+ "grad_norm": 0.05322265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0426,
+ "step": 10182
+ },
+ {
+ "epoch": 32.533546325878596,
+ "grad_norm": 0.09912109375,
+ "learning_rate": 0.0005,
+ "loss": 1.037,
+ "step": 10183
+ },
+ {
+ "epoch": 32.53674121405751,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0391,
+ "step": 10184
+ },
+ {
+ "epoch": 32.539936102236425,
+ "grad_norm": 0.09814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0286,
+ "step": 10185
+ },
+ {
+ "epoch": 32.543130990415335,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0266,
+ "step": 10186
+ },
+ {
+ "epoch": 32.546325878594246,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0507,
+ "step": 10187
+ },
+ {
+ "epoch": 32.54952076677316,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0421,
+ "step": 10188
+ },
+ {
+ "epoch": 32.552715654952074,
+ "grad_norm": 0.0615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0317,
+ "step": 10189
+ },
+ {
+ "epoch": 32.55591054313099,
+ "grad_norm": 0.053955078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0318,
+ "step": 10190
+ },
+ {
+ "epoch": 32.5591054313099,
+ "grad_norm": 0.058349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0365,
+ "step": 10191
+ },
+ {
+ "epoch": 32.56230031948882,
+ "grad_norm": 0.0576171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0296,
+ "step": 10192
+ },
+ {
+ "epoch": 32.56549520766773,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0397,
+ "step": 10193
+ },
+ {
+ "epoch": 32.56869009584665,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0375,
+ "step": 10194
+ },
+ {
+ "epoch": 32.57188498402556,
+ "grad_norm": 0.049560546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0585,
+ "step": 10195
+ },
+ {
+ "epoch": 32.575079872204476,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0326,
+ "step": 10196
+ },
+ {
+ "epoch": 32.57827476038339,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0508,
+ "step": 10197
+ },
+ {
+ "epoch": 32.5814696485623,
+ "grad_norm": 0.05126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0382,
+ "step": 10198
+ },
+ {
+ "epoch": 32.584664536741215,
+ "grad_norm": 0.1259765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0428,
+ "step": 10199
+ },
+ {
+ "epoch": 32.587859424920126,
+ "grad_norm": 0.0537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.031,
+ "step": 10200
+ },
+ {
+ "epoch": 32.59105431309904,
+ "grad_norm": 0.08251953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0447,
+ "step": 10201
+ },
+ {
+ "epoch": 32.594249201277954,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0433,
+ "step": 10202
+ },
+ {
+ "epoch": 32.59744408945687,
+ "grad_norm": 0.055908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0301,
+ "step": 10203
+ },
+ {
+ "epoch": 32.60063897763578,
+ "grad_norm": 0.058837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0324,
+ "step": 10204
+ },
+ {
+ "epoch": 32.6038338658147,
+ "grad_norm": 0.05322265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0395,
+ "step": 10205
+ },
+ {
+ "epoch": 32.60702875399361,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0334,
+ "step": 10206
+ },
+ {
+ "epoch": 32.61022364217252,
+ "grad_norm": 0.058837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0379,
+ "step": 10207
+ },
+ {
+ "epoch": 32.61341853035144,
+ "grad_norm": 0.06103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0349,
+ "step": 10208
+ },
+ {
+ "epoch": 32.61661341853035,
+ "grad_norm": 0.054443359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0392,
+ "step": 10209
+ },
+ {
+ "epoch": 32.61980830670927,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0346,
+ "step": 10210
+ },
+ {
+ "epoch": 32.62300319488818,
+ "grad_norm": 0.052001953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0356,
+ "step": 10211
+ },
+ {
+ "epoch": 32.626198083067095,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0373,
+ "step": 10212
+ },
+ {
+ "epoch": 32.629392971246006,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0329,
+ "step": 10213
+ },
+ {
+ "epoch": 32.63258785942492,
+ "grad_norm": 0.053466796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0408,
+ "step": 10214
+ },
+ {
+ "epoch": 32.635782747603834,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0324,
+ "step": 10215
+ },
+ {
+ "epoch": 32.638977635782744,
+ "grad_norm": 0.05517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0327,
+ "step": 10216
+ },
+ {
+ "epoch": 32.64217252396166,
+ "grad_norm": 0.04638671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0438,
+ "step": 10217
+ },
+ {
+ "epoch": 32.64536741214057,
+ "grad_norm": 0.052734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0341,
+ "step": 10218
+ },
+ {
+ "epoch": 32.64856230031949,
+ "grad_norm": 0.054931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0396,
+ "step": 10219
+ },
+ {
+ "epoch": 32.6517571884984,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0464,
+ "step": 10220
+ },
+ {
+ "epoch": 32.65495207667732,
+ "grad_norm": 0.0458984375,
+ "learning_rate": 0.0005,
+ "loss": 1.043,
+ "step": 10221
+ },
+ {
+ "epoch": 32.65814696485623,
+ "grad_norm": 0.12109375,
+ "learning_rate": 0.0005,
+ "loss": 1.034,
+ "step": 10222
+ },
+ {
+ "epoch": 32.66134185303515,
+ "grad_norm": 0.046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0336,
+ "step": 10223
+ },
+ {
+ "epoch": 32.66453674121406,
+ "grad_norm": 0.095703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0331,
+ "step": 10224
+ },
+ {
+ "epoch": 32.66773162939297,
+ "grad_norm": 0.043701171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0324,
+ "step": 10225
+ },
+ {
+ "epoch": 32.670926517571885,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0458,
+ "step": 10226
+ },
+ {
+ "epoch": 32.674121405750796,
+ "grad_norm": 0.046142578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0319,
+ "step": 10227
+ },
+ {
+ "epoch": 32.677316293929714,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0443,
+ "step": 10228
+ },
+ {
+ "epoch": 32.680511182108624,
+ "grad_norm": 0.0537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0357,
+ "step": 10229
+ },
+ {
+ "epoch": 32.68370607028754,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0459,
+ "step": 10230
+ },
+ {
+ "epoch": 32.68690095846645,
+ "grad_norm": 0.056396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0299,
+ "step": 10231
+ },
+ {
+ "epoch": 32.69009584664537,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0339,
+ "step": 10232
+ },
+ {
+ "epoch": 32.69329073482428,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0466,
+ "step": 10233
+ },
+ {
+ "epoch": 32.6964856230032,
+ "grad_norm": 0.1259765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0464,
+ "step": 10234
+ },
+ {
+ "epoch": 32.69968051118211,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0294,
+ "step": 10235
+ },
+ {
+ "epoch": 32.70287539936102,
+ "grad_norm": 0.10986328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0305,
+ "step": 10236
+ },
+ {
+ "epoch": 32.70607028753994,
+ "grad_norm": 0.060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0445,
+ "step": 10237
+ },
+ {
+ "epoch": 32.70926517571885,
+ "grad_norm": 0.1279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0461,
+ "step": 10238
+ },
+ {
+ "epoch": 32.712460063897765,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0332,
+ "step": 10239
+ },
+ {
+ "epoch": 32.715654952076676,
+ "grad_norm": 0.11767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0369,
+ "step": 10240
+ },
+ {
+ "epoch": 32.718849840255594,
+ "grad_norm": 0.080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0374,
+ "step": 10241
+ },
+ {
+ "epoch": 32.722044728434504,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0339,
+ "step": 10242
+ },
+ {
+ "epoch": 32.72523961661342,
+ "grad_norm": 0.095703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0372,
+ "step": 10243
+ },
+ {
+ "epoch": 32.72843450479233,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0357,
+ "step": 10244
+ },
+ {
+ "epoch": 32.73162939297124,
+ "grad_norm": 0.056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0316,
+ "step": 10245
+ },
+ {
+ "epoch": 32.73482428115016,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0312,
+ "step": 10246
+ },
+ {
+ "epoch": 32.73801916932907,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0413,
+ "step": 10247
+ },
+ {
+ "epoch": 32.74121405750799,
+ "grad_norm": 0.1181640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0477,
+ "step": 10248
+ },
+ {
+ "epoch": 32.7444089456869,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.037,
+ "step": 10249
+ },
+ {
+ "epoch": 32.74760383386582,
+ "grad_norm": 0.126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0341,
+ "step": 10250
+ },
+ {
+ "epoch": 32.75079872204473,
+ "grad_norm": 0.0615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0263,
+ "step": 10251
+ },
+ {
+ "epoch": 32.753993610223645,
+ "grad_norm": 0.1396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0256,
+ "step": 10252
+ },
+ {
+ "epoch": 32.757188498402556,
+ "grad_norm": 0.0615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0374,
+ "step": 10253
+ },
+ {
+ "epoch": 32.760383386581466,
+ "grad_norm": 0.125,
+ "learning_rate": 0.0005,
+ "loss": 1.0315,
+ "step": 10254
+ },
+ {
+ "epoch": 32.763578274760384,
+ "grad_norm": 0.1435546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0527,
+ "step": 10255
+ },
+ {
+ "epoch": 32.766773162939295,
+ "grad_norm": 0.20703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0316,
+ "step": 10256
+ },
+ {
+ "epoch": 32.76996805111821,
+ "grad_norm": 0.091796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0426,
+ "step": 10257
+ },
+ {
+ "epoch": 32.77316293929712,
+ "grad_norm": 0.1650390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0362,
+ "step": 10258
+ },
+ {
+ "epoch": 32.77635782747604,
+ "grad_norm": 0.1142578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0286,
+ "step": 10259
+ },
+ {
+ "epoch": 32.77955271565495,
+ "grad_norm": 0.1123046875,
+ "learning_rate": 0.0005,
+ "loss": 1.042,
+ "step": 10260
+ },
+ {
+ "epoch": 32.78274760383387,
+ "grad_norm": 0.1328125,
+ "learning_rate": 0.0005,
+ "loss": 1.046,
+ "step": 10261
+ },
+ {
+ "epoch": 32.78594249201278,
+ "grad_norm": 0.06201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0378,
+ "step": 10262
+ },
+ {
+ "epoch": 32.78913738019169,
+ "grad_norm": 0.1875,
+ "learning_rate": 0.0005,
+ "loss": 1.0394,
+ "step": 10263
+ },
+ {
+ "epoch": 32.79233226837061,
+ "grad_norm": 0.091796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0385,
+ "step": 10264
+ },
+ {
+ "epoch": 32.79552715654952,
+ "grad_norm": 0.12451171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0254,
+ "step": 10265
+ },
+ {
+ "epoch": 32.798722044728436,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.034,
+ "step": 10266
+ },
+ {
+ "epoch": 32.801916932907346,
+ "grad_norm": 0.10498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0482,
+ "step": 10267
+ },
+ {
+ "epoch": 32.805111821086264,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0351,
+ "step": 10268
+ },
+ {
+ "epoch": 32.808306709265175,
+ "grad_norm": 0.095703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0408,
+ "step": 10269
+ },
+ {
+ "epoch": 32.81150159744409,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0313,
+ "step": 10270
+ },
+ {
+ "epoch": 32.814696485623,
+ "grad_norm": 0.058349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0315,
+ "step": 10271
+ },
+ {
+ "epoch": 32.81789137380191,
+ "grad_norm": 0.055419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.036,
+ "step": 10272
+ },
+ {
+ "epoch": 32.82108626198083,
+ "grad_norm": 0.062255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0425,
+ "step": 10273
+ },
+ {
+ "epoch": 32.82428115015974,
+ "grad_norm": 0.05322265625,
+ "learning_rate": 0.0005,
+ "loss": 1.041,
+ "step": 10274
+ },
+ {
+ "epoch": 32.82747603833866,
+ "grad_norm": 0.05517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0451,
+ "step": 10275
+ },
+ {
+ "epoch": 32.83067092651757,
+ "grad_norm": 0.061279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.043,
+ "step": 10276
+ },
+ {
+ "epoch": 32.83386581469649,
+ "grad_norm": 0.054931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0198,
+ "step": 10277
+ },
+ {
+ "epoch": 32.8370607028754,
+ "grad_norm": 0.062255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0486,
+ "step": 10278
+ },
+ {
+ "epoch": 32.840255591054316,
+ "grad_norm": 0.08349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.035,
+ "step": 10279
+ },
+ {
+ "epoch": 32.843450479233226,
+ "grad_norm": 0.09423828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0382,
+ "step": 10280
+ },
+ {
+ "epoch": 32.846645367412144,
+ "grad_norm": 0.060791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0279,
+ "step": 10281
+ },
+ {
+ "epoch": 32.849840255591054,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0384,
+ "step": 10282
+ },
+ {
+ "epoch": 32.853035143769965,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0293,
+ "step": 10283
+ },
+ {
+ "epoch": 32.85623003194888,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0421,
+ "step": 10284
+ },
+ {
+ "epoch": 32.85942492012779,
+ "grad_norm": 0.1044921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0358,
+ "step": 10285
+ },
+ {
+ "epoch": 32.86261980830671,
+ "grad_norm": 0.0576171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0343,
+ "step": 10286
+ },
+ {
+ "epoch": 32.86581469648562,
+ "grad_norm": 0.11279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.033,
+ "step": 10287
+ },
+ {
+ "epoch": 32.86900958466454,
+ "grad_norm": 0.0888671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0401,
+ "step": 10288
+ },
+ {
+ "epoch": 32.87220447284345,
+ "grad_norm": 0.09423828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0303,
+ "step": 10289
+ },
+ {
+ "epoch": 32.87539936102237,
+ "grad_norm": 0.08154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0401,
+ "step": 10290
+ },
+ {
+ "epoch": 32.87859424920128,
+ "grad_norm": 0.10888671875,
+ "learning_rate": 0.0005,
+ "loss": 1.043,
+ "step": 10291
+ },
+ {
+ "epoch": 32.88178913738019,
+ "grad_norm": 0.1171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0442,
+ "step": 10292
+ },
+ {
+ "epoch": 32.884984025559106,
+ "grad_norm": 0.08935546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0431,
+ "step": 10293
+ },
+ {
+ "epoch": 32.88817891373802,
+ "grad_norm": 0.1455078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0396,
+ "step": 10294
+ },
+ {
+ "epoch": 32.891373801916934,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0411,
+ "step": 10295
+ },
+ {
+ "epoch": 32.894568690095845,
+ "grad_norm": 0.158203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0531,
+ "step": 10296
+ },
+ {
+ "epoch": 32.89776357827476,
+ "grad_norm": 0.052734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0271,
+ "step": 10297
+ },
+ {
+ "epoch": 32.90095846645367,
+ "grad_norm": 0.10302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0323,
+ "step": 10298
+ },
+ {
+ "epoch": 32.90415335463259,
+ "grad_norm": 0.0517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0359,
+ "step": 10299
+ },
+ {
+ "epoch": 32.9073482428115,
+ "grad_norm": 0.115234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0399,
+ "step": 10300
+ },
+ {
+ "epoch": 32.91054313099041,
+ "grad_norm": 0.046142578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0362,
+ "step": 10301
+ },
+ {
+ "epoch": 32.91373801916933,
+ "grad_norm": 0.11865234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0388,
+ "step": 10302
+ },
+ {
+ "epoch": 32.91693290734824,
+ "grad_norm": 0.04931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0272,
+ "step": 10303
+ },
+ {
+ "epoch": 32.92012779552716,
+ "grad_norm": 0.11865234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0553,
+ "step": 10304
+ },
+ {
+ "epoch": 32.92332268370607,
+ "grad_norm": 0.053955078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0454,
+ "step": 10305
+ },
+ {
+ "epoch": 32.926517571884986,
+ "grad_norm": 0.11181640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0449,
+ "step": 10306
+ },
+ {
+ "epoch": 32.9297124600639,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0436,
+ "step": 10307
+ },
+ {
+ "epoch": 32.932907348242814,
+ "grad_norm": 0.1015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0314,
+ "step": 10308
+ },
+ {
+ "epoch": 32.936102236421725,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0326,
+ "step": 10309
+ },
+ {
+ "epoch": 32.93929712460064,
+ "grad_norm": 0.0849609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0254,
+ "step": 10310
+ },
+ {
+ "epoch": 32.94249201277955,
+ "grad_norm": 0.08203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0359,
+ "step": 10311
+ },
+ {
+ "epoch": 32.945686900958464,
+ "grad_norm": 0.1494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0335,
+ "step": 10312
+ },
+ {
+ "epoch": 32.94888178913738,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0325,
+ "step": 10313
+ },
+ {
+ "epoch": 32.95207667731629,
+ "grad_norm": 0.1875,
+ "learning_rate": 0.0005,
+ "loss": 1.0373,
+ "step": 10314
+ },
+ {
+ "epoch": 32.95527156549521,
+ "grad_norm": 0.1064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.042,
+ "step": 10315
+ },
+ {
+ "epoch": 32.95846645367412,
+ "grad_norm": 0.3046875,
+ "learning_rate": 0.0005,
+ "loss": 1.037,
+ "step": 10316
+ },
+ {
+ "epoch": 32.96166134185304,
+ "grad_norm": 0.12158203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0303,
+ "step": 10317
+ },
+ {
+ "epoch": 32.96485623003195,
+ "grad_norm": 0.1689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0372,
+ "step": 10318
+ },
+ {
+ "epoch": 32.968051118210866,
+ "grad_norm": 0.09716796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0216,
+ "step": 10319
+ },
+ {
+ "epoch": 32.97124600638978,
+ "grad_norm": 0.1513671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0298,
+ "step": 10320
+ },
+ {
+ "epoch": 32.97444089456869,
+ "grad_norm": 0.1416015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0409,
+ "step": 10321
+ },
+ {
+ "epoch": 32.977635782747605,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0384,
+ "step": 10322
+ },
+ {
+ "epoch": 32.980830670926515,
+ "grad_norm": 0.09765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0437,
+ "step": 10323
+ },
+ {
+ "epoch": 32.98402555910543,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0407,
+ "step": 10324
+ },
+ {
+ "epoch": 32.98722044728434,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0386,
+ "step": 10325
+ },
+ {
+ "epoch": 32.99041533546326,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0408,
+ "step": 10326
+ },
+ {
+ "epoch": 32.99361022364217,
+ "grad_norm": 0.10205078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0477,
+ "step": 10327
+ },
+ {
+ "epoch": 32.99680511182109,
+ "grad_norm": 0.058349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0297,
+ "step": 10328
+ },
+ {
+ "epoch": 33.0,
+ "grad_norm": 0.1162109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0436,
+ "step": 10329
+ },
+ {
+ "epoch": 33.00319488817891,
+ "grad_norm": 0.052978515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0411,
+ "step": 10330
+ },
+ {
+ "epoch": 33.00638977635783,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0465,
+ "step": 10331
+ },
+ {
+ "epoch": 33.00958466453674,
+ "grad_norm": 0.06201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0302,
+ "step": 10332
+ },
+ {
+ "epoch": 33.01277955271566,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0393,
+ "step": 10333
+ },
+ {
+ "epoch": 33.01597444089457,
+ "grad_norm": 0.051025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0181,
+ "step": 10334
+ },
+ {
+ "epoch": 33.019169329073485,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0423,
+ "step": 10335
+ },
+ {
+ "epoch": 33.022364217252395,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0346,
+ "step": 10336
+ },
+ {
+ "epoch": 33.02555910543131,
+ "grad_norm": 0.053955078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0354,
+ "step": 10337
+ },
+ {
+ "epoch": 33.02875399361022,
+ "grad_norm": 0.08984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0363,
+ "step": 10338
+ },
+ {
+ "epoch": 33.031948881789134,
+ "grad_norm": 0.05419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0377,
+ "step": 10339
+ },
+ {
+ "epoch": 33.03514376996805,
+ "grad_norm": 0.060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0368,
+ "step": 10340
+ },
+ {
+ "epoch": 33.03833865814696,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0439,
+ "step": 10341
+ },
+ {
+ "epoch": 33.04153354632588,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0313,
+ "step": 10342
+ },
+ {
+ "epoch": 33.04472843450479,
+ "grad_norm": 0.05810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0362,
+ "step": 10343
+ },
+ {
+ "epoch": 33.04792332268371,
+ "grad_norm": 0.08251953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0371,
+ "step": 10344
+ },
+ {
+ "epoch": 33.05111821086262,
+ "grad_norm": 0.05419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.035,
+ "step": 10345
+ },
+ {
+ "epoch": 33.054313099041536,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0359,
+ "step": 10346
+ },
+ {
+ "epoch": 33.05750798722045,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0403,
+ "step": 10347
+ },
+ {
+ "epoch": 33.06070287539936,
+ "grad_norm": 0.061279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0414,
+ "step": 10348
+ },
+ {
+ "epoch": 33.063897763578275,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0302,
+ "step": 10349
+ },
+ {
+ "epoch": 33.067092651757186,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0361,
+ "step": 10350
+ },
+ {
+ "epoch": 33.0702875399361,
+ "grad_norm": 0.12451171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0359,
+ "step": 10351
+ },
+ {
+ "epoch": 33.073482428115014,
+ "grad_norm": 0.061767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0284,
+ "step": 10352
+ },
+ {
+ "epoch": 33.07667731629393,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0358,
+ "step": 10353
+ },
+ {
+ "epoch": 33.07987220447284,
+ "grad_norm": 0.06201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0424,
+ "step": 10354
+ },
+ {
+ "epoch": 33.08306709265176,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0418,
+ "step": 10355
+ },
+ {
+ "epoch": 33.08626198083067,
+ "grad_norm": 0.0517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0304,
+ "step": 10356
+ },
+ {
+ "epoch": 33.08945686900959,
+ "grad_norm": 0.080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0364,
+ "step": 10357
+ },
+ {
+ "epoch": 33.0926517571885,
+ "grad_norm": 0.050537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.032,
+ "step": 10358
+ },
+ {
+ "epoch": 33.09584664536741,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.044,
+ "step": 10359
+ },
+ {
+ "epoch": 33.09904153354633,
+ "grad_norm": 0.058837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0244,
+ "step": 10360
+ },
+ {
+ "epoch": 33.10223642172524,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0415,
+ "step": 10361
+ },
+ {
+ "epoch": 33.105431309904155,
+ "grad_norm": 0.057373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.045,
+ "step": 10362
+ },
+ {
+ "epoch": 33.108626198083066,
+ "grad_norm": 0.05810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0385,
+ "step": 10363
+ },
+ {
+ "epoch": 33.11182108626198,
+ "grad_norm": 0.0546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0474,
+ "step": 10364
+ },
+ {
+ "epoch": 33.115015974440894,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0378,
+ "step": 10365
+ },
+ {
+ "epoch": 33.11821086261981,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.041,
+ "step": 10366
+ },
+ {
+ "epoch": 33.12140575079872,
+ "grad_norm": 0.058837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0399,
+ "step": 10367
+ },
+ {
+ "epoch": 33.12460063897763,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.038,
+ "step": 10368
+ },
+ {
+ "epoch": 33.12779552715655,
+ "grad_norm": 0.054931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0381,
+ "step": 10369
+ },
+ {
+ "epoch": 33.13099041533546,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0361,
+ "step": 10370
+ },
+ {
+ "epoch": 33.13418530351438,
+ "grad_norm": 0.048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0474,
+ "step": 10371
+ },
+ {
+ "epoch": 33.13738019169329,
+ "grad_norm": 0.058837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0297,
+ "step": 10372
+ },
+ {
+ "epoch": 33.14057507987221,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0409,
+ "step": 10373
+ },
+ {
+ "epoch": 33.14376996805112,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0319,
+ "step": 10374
+ },
+ {
+ "epoch": 33.146964856230035,
+ "grad_norm": 0.08544921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0389,
+ "step": 10375
+ },
+ {
+ "epoch": 33.150159744408946,
+ "grad_norm": 0.09521484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0311,
+ "step": 10376
+ },
+ {
+ "epoch": 33.153354632587856,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0323,
+ "step": 10377
+ },
+ {
+ "epoch": 33.156549520766774,
+ "grad_norm": 0.1103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0355,
+ "step": 10378
+ },
+ {
+ "epoch": 33.159744408945684,
+ "grad_norm": 0.048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0212,
+ "step": 10379
+ },
+ {
+ "epoch": 33.1629392971246,
+ "grad_norm": 0.05615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0362,
+ "step": 10380
+ },
+ {
+ "epoch": 33.16613418530351,
+ "grad_norm": 0.06201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0386,
+ "step": 10381
+ },
+ {
+ "epoch": 33.16932907348243,
+ "grad_norm": 0.0556640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0283,
+ "step": 10382
+ },
+ {
+ "epoch": 33.17252396166134,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0413,
+ "step": 10383
+ },
+ {
+ "epoch": 33.17571884984026,
+ "grad_norm": 0.052001953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0327,
+ "step": 10384
+ },
+ {
+ "epoch": 33.17891373801917,
+ "grad_norm": 0.09375,
+ "learning_rate": 0.0005,
+ "loss": 1.0327,
+ "step": 10385
+ },
+ {
+ "epoch": 33.18210862619808,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0275,
+ "step": 10386
+ },
+ {
+ "epoch": 33.185303514377,
+ "grad_norm": 0.080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0213,
+ "step": 10387
+ },
+ {
+ "epoch": 33.18849840255591,
+ "grad_norm": 0.0888671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0344,
+ "step": 10388
+ },
+ {
+ "epoch": 33.191693290734825,
+ "grad_norm": 0.09716796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0365,
+ "step": 10389
+ },
+ {
+ "epoch": 33.194888178913736,
+ "grad_norm": 0.0537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.035,
+ "step": 10390
+ },
+ {
+ "epoch": 33.198083067092654,
+ "grad_norm": 0.0966796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0426,
+ "step": 10391
+ },
+ {
+ "epoch": 33.201277955271564,
+ "grad_norm": 0.060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0412,
+ "step": 10392
+ },
+ {
+ "epoch": 33.20447284345048,
+ "grad_norm": 0.08740234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0349,
+ "step": 10393
+ },
+ {
+ "epoch": 33.20766773162939,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0481,
+ "step": 10394
+ },
+ {
+ "epoch": 33.21086261980831,
+ "grad_norm": 0.08740234375,
+ "learning_rate": 0.0005,
+ "loss": 1.037,
+ "step": 10395
+ },
+ {
+ "epoch": 33.21405750798722,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0356,
+ "step": 10396
+ },
+ {
+ "epoch": 33.21725239616613,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0288,
+ "step": 10397
+ },
+ {
+ "epoch": 33.22044728434505,
+ "grad_norm": 0.11328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0273,
+ "step": 10398
+ },
+ {
+ "epoch": 33.22364217252396,
+ "grad_norm": 0.10791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0492,
+ "step": 10399
+ },
+ {
+ "epoch": 33.22683706070288,
+ "grad_norm": 0.115234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0341,
+ "step": 10400
+ },
+ {
+ "epoch": 33.23003194888179,
+ "grad_norm": 0.1357421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0114,
+ "step": 10401
+ },
+ {
+ "epoch": 33.233226837060705,
+ "grad_norm": 0.0888671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0428,
+ "step": 10402
+ },
+ {
+ "epoch": 33.236421725239616,
+ "grad_norm": 0.10693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0323,
+ "step": 10403
+ },
+ {
+ "epoch": 33.239616613418534,
+ "grad_norm": 0.10595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0292,
+ "step": 10404
+ },
+ {
+ "epoch": 33.242811501597444,
+ "grad_norm": 0.11572265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0265,
+ "step": 10405
+ },
+ {
+ "epoch": 33.246006389776355,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0473,
+ "step": 10406
+ },
+ {
+ "epoch": 33.24920127795527,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0443,
+ "step": 10407
+ },
+ {
+ "epoch": 33.25239616613418,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0292,
+ "step": 10408
+ },
+ {
+ "epoch": 33.2555910543131,
+ "grad_norm": 0.05712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0349,
+ "step": 10409
+ },
+ {
+ "epoch": 33.25878594249201,
+ "grad_norm": 0.05712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0335,
+ "step": 10410
+ },
+ {
+ "epoch": 33.26198083067093,
+ "grad_norm": 0.051025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0255,
+ "step": 10411
+ },
+ {
+ "epoch": 33.26517571884984,
+ "grad_norm": 0.05908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0463,
+ "step": 10412
+ },
+ {
+ "epoch": 33.26837060702876,
+ "grad_norm": 0.05078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0465,
+ "step": 10413
+ },
+ {
+ "epoch": 33.27156549520767,
+ "grad_norm": 0.095703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0312,
+ "step": 10414
+ },
+ {
+ "epoch": 33.27476038338658,
+ "grad_norm": 0.054443359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0267,
+ "step": 10415
+ },
+ {
+ "epoch": 33.277955271565496,
+ "grad_norm": 0.123046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0321,
+ "step": 10416
+ },
+ {
+ "epoch": 33.281150159744406,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0438,
+ "step": 10417
+ },
+ {
+ "epoch": 33.284345047923324,
+ "grad_norm": 0.09716796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0368,
+ "step": 10418
+ },
+ {
+ "epoch": 33.287539936102235,
+ "grad_norm": 0.059814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0307,
+ "step": 10419
+ },
+ {
+ "epoch": 33.29073482428115,
+ "grad_norm": 0.1044921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0557,
+ "step": 10420
+ },
+ {
+ "epoch": 33.29392971246006,
+ "grad_norm": 0.080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0362,
+ "step": 10421
+ },
+ {
+ "epoch": 33.29712460063898,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0361,
+ "step": 10422
+ },
+ {
+ "epoch": 33.30031948881789,
+ "grad_norm": 0.08935546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0358,
+ "step": 10423
+ },
+ {
+ "epoch": 33.3035143769968,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0286,
+ "step": 10424
+ },
+ {
+ "epoch": 33.30670926517572,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0378,
+ "step": 10425
+ },
+ {
+ "epoch": 33.30990415335463,
+ "grad_norm": 0.061279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0554,
+ "step": 10426
+ },
+ {
+ "epoch": 33.31309904153355,
+ "grad_norm": 0.08349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0309,
+ "step": 10427
+ },
+ {
+ "epoch": 33.31629392971246,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0388,
+ "step": 10428
+ },
+ {
+ "epoch": 33.319488817891376,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0373,
+ "step": 10429
+ },
+ {
+ "epoch": 33.322683706070286,
+ "grad_norm": 0.05224609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0394,
+ "step": 10430
+ },
+ {
+ "epoch": 33.325878594249204,
+ "grad_norm": 0.0615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0448,
+ "step": 10431
+ },
+ {
+ "epoch": 33.329073482428115,
+ "grad_norm": 0.051513671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0424,
+ "step": 10432
+ },
+ {
+ "epoch": 33.33226837060703,
+ "grad_norm": 0.04833984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0366,
+ "step": 10433
+ },
+ {
+ "epoch": 33.33546325878594,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0349,
+ "step": 10434
+ },
+ {
+ "epoch": 33.33865814696485,
+ "grad_norm": 0.052490234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0394,
+ "step": 10435
+ },
+ {
+ "epoch": 33.34185303514377,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0321,
+ "step": 10436
+ },
+ {
+ "epoch": 33.34504792332268,
+ "grad_norm": 0.062255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0524,
+ "step": 10437
+ },
+ {
+ "epoch": 33.3482428115016,
+ "grad_norm": 0.058349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0386,
+ "step": 10438
+ },
+ {
+ "epoch": 33.35143769968051,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0307,
+ "step": 10439
+ },
+ {
+ "epoch": 33.35463258785943,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0359,
+ "step": 10440
+ },
+ {
+ "epoch": 33.35782747603834,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0293,
+ "step": 10441
+ },
+ {
+ "epoch": 33.361022364217256,
+ "grad_norm": 0.080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0258,
+ "step": 10442
+ },
+ {
+ "epoch": 33.364217252396166,
+ "grad_norm": 0.10791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0419,
+ "step": 10443
+ },
+ {
+ "epoch": 33.36741214057508,
+ "grad_norm": 0.1328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0245,
+ "step": 10444
+ },
+ {
+ "epoch": 33.370607028753994,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0278,
+ "step": 10445
+ },
+ {
+ "epoch": 33.373801916932905,
+ "grad_norm": 0.1396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0376,
+ "step": 10446
+ },
+ {
+ "epoch": 33.37699680511182,
+ "grad_norm": 0.055908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0378,
+ "step": 10447
+ },
+ {
+ "epoch": 33.38019169329073,
+ "grad_norm": 0.11474609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0329,
+ "step": 10448
+ },
+ {
+ "epoch": 33.38338658146965,
+ "grad_norm": 0.05078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0297,
+ "step": 10449
+ },
+ {
+ "epoch": 33.38658146964856,
+ "grad_norm": 0.0869140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0287,
+ "step": 10450
+ },
+ {
+ "epoch": 33.38977635782748,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0454,
+ "step": 10451
+ },
+ {
+ "epoch": 33.39297124600639,
+ "grad_norm": 0.1240234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0344,
+ "step": 10452
+ },
+ {
+ "epoch": 33.3961661341853,
+ "grad_norm": 0.048095703125,
+ "learning_rate": 0.0005,
+ "loss": 1.036,
+ "step": 10453
+ },
+ {
+ "epoch": 33.39936102236422,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0263,
+ "step": 10454
+ },
+ {
+ "epoch": 33.40255591054313,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0397,
+ "step": 10455
+ },
+ {
+ "epoch": 33.405750798722046,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0466,
+ "step": 10456
+ },
+ {
+ "epoch": 33.40894568690096,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0365,
+ "step": 10457
+ },
+ {
+ "epoch": 33.412140575079874,
+ "grad_norm": 0.08203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0308,
+ "step": 10458
+ },
+ {
+ "epoch": 33.415335463258785,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0401,
+ "step": 10459
+ },
+ {
+ "epoch": 33.4185303514377,
+ "grad_norm": 0.09814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0327,
+ "step": 10460
+ },
+ {
+ "epoch": 33.42172523961661,
+ "grad_norm": 0.0830078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0359,
+ "step": 10461
+ },
+ {
+ "epoch": 33.424920127795524,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0375,
+ "step": 10462
+ },
+ {
+ "epoch": 33.42811501597444,
+ "grad_norm": 0.08203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0432,
+ "step": 10463
+ },
+ {
+ "epoch": 33.43130990415335,
+ "grad_norm": 0.056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0327,
+ "step": 10464
+ },
+ {
+ "epoch": 33.43450479233227,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0422,
+ "step": 10465
+ },
+ {
+ "epoch": 33.43769968051118,
+ "grad_norm": 0.049560546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0381,
+ "step": 10466
+ },
+ {
+ "epoch": 33.4408945686901,
+ "grad_norm": 0.05322265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0341,
+ "step": 10467
+ },
+ {
+ "epoch": 33.44408945686901,
+ "grad_norm": 0.07861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0388,
+ "step": 10468
+ },
+ {
+ "epoch": 33.447284345047926,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0292,
+ "step": 10469
+ },
+ {
+ "epoch": 33.45047923322684,
+ "grad_norm": 0.0615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0421,
+ "step": 10470
+ },
+ {
+ "epoch": 33.453674121405754,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0452,
+ "step": 10471
+ },
+ {
+ "epoch": 33.456869009584665,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0301,
+ "step": 10472
+ },
+ {
+ "epoch": 33.460063897763575,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0438,
+ "step": 10473
+ },
+ {
+ "epoch": 33.46325878594249,
+ "grad_norm": 0.0546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0385,
+ "step": 10474
+ },
+ {
+ "epoch": 33.466453674121404,
+ "grad_norm": 0.05126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0271,
+ "step": 10475
+ },
+ {
+ "epoch": 33.46964856230032,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0327,
+ "step": 10476
+ },
+ {
+ "epoch": 33.47284345047923,
+ "grad_norm": 0.053955078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0219,
+ "step": 10477
+ },
+ {
+ "epoch": 33.47603833865815,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.028,
+ "step": 10478
+ },
+ {
+ "epoch": 33.47923322683706,
+ "grad_norm": 0.0859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0356,
+ "step": 10479
+ },
+ {
+ "epoch": 33.48242811501598,
+ "grad_norm": 0.08740234375,
+ "learning_rate": 0.0005,
+ "loss": 1.028,
+ "step": 10480
+ },
+ {
+ "epoch": 33.48562300319489,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0404,
+ "step": 10481
+ },
+ {
+ "epoch": 33.4888178913738,
+ "grad_norm": 0.07861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0236,
+ "step": 10482
+ },
+ {
+ "epoch": 33.49201277955272,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0341,
+ "step": 10483
+ },
+ {
+ "epoch": 33.49520766773163,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0319,
+ "step": 10484
+ },
+ {
+ "epoch": 33.498402555910545,
+ "grad_norm": 0.060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0213,
+ "step": 10485
+ },
+ {
+ "epoch": 33.501597444089455,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.036,
+ "step": 10486
+ },
+ {
+ "epoch": 33.50479233226837,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0396,
+ "step": 10487
+ },
+ {
+ "epoch": 33.50798722044728,
+ "grad_norm": 0.0810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0373,
+ "step": 10488
+ },
+ {
+ "epoch": 33.5111821086262,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0416,
+ "step": 10489
+ },
+ {
+ "epoch": 33.51437699680511,
+ "grad_norm": 0.06103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0243,
+ "step": 10490
+ },
+ {
+ "epoch": 33.51757188498402,
+ "grad_norm": 0.06103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0405,
+ "step": 10491
+ },
+ {
+ "epoch": 33.52076677316294,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0322,
+ "step": 10492
+ },
+ {
+ "epoch": 33.52396166134185,
+ "grad_norm": 0.04736328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0305,
+ "step": 10493
+ },
+ {
+ "epoch": 33.52715654952077,
+ "grad_norm": 0.091796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0405,
+ "step": 10494
+ },
+ {
+ "epoch": 33.53035143769968,
+ "grad_norm": 0.0595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.046,
+ "step": 10495
+ },
+ {
+ "epoch": 33.533546325878596,
+ "grad_norm": 0.053466796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0331,
+ "step": 10496
+ },
+ {
+ "epoch": 33.53674121405751,
+ "grad_norm": 0.08349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0244,
+ "step": 10497
+ },
+ {
+ "epoch": 33.539936102236425,
+ "grad_norm": 0.05712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0311,
+ "step": 10498
+ },
+ {
+ "epoch": 33.543130990415335,
+ "grad_norm": 0.05078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0413,
+ "step": 10499
+ },
+ {
+ "epoch": 33.546325878594246,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0368,
+ "step": 10500
+ },
+ {
+ "epoch": 33.54952076677316,
+ "grad_norm": 0.05322265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0491,
+ "step": 10501
+ },
+ {
+ "epoch": 33.552715654952074,
+ "grad_norm": 0.060791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0326,
+ "step": 10502
+ },
+ {
+ "epoch": 33.55591054313099,
+ "grad_norm": 0.0537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.035,
+ "step": 10503
+ },
+ {
+ "epoch": 33.5591054313099,
+ "grad_norm": 0.056396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0384,
+ "step": 10504
+ },
+ {
+ "epoch": 33.56230031948882,
+ "grad_norm": 0.087890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0295,
+ "step": 10505
+ },
+ {
+ "epoch": 33.56549520766773,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0313,
+ "step": 10506
+ },
+ {
+ "epoch": 33.56869009584665,
+ "grad_norm": 0.09716796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0396,
+ "step": 10507
+ },
+ {
+ "epoch": 33.57188498402556,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0329,
+ "step": 10508
+ },
+ {
+ "epoch": 33.575079872204476,
+ "grad_norm": 0.08447265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0315,
+ "step": 10509
+ },
+ {
+ "epoch": 33.57827476038339,
+ "grad_norm": 0.06201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0305,
+ "step": 10510
+ },
+ {
+ "epoch": 33.5814696485623,
+ "grad_norm": 0.0859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0461,
+ "step": 10511
+ },
+ {
+ "epoch": 33.584664536741215,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0458,
+ "step": 10512
+ },
+ {
+ "epoch": 33.587859424920126,
+ "grad_norm": 0.060302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0358,
+ "step": 10513
+ },
+ {
+ "epoch": 33.59105431309904,
+ "grad_norm": 0.05419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0246,
+ "step": 10514
+ },
+ {
+ "epoch": 33.594249201277954,
+ "grad_norm": 0.051513671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0225,
+ "step": 10515
+ },
+ {
+ "epoch": 33.59744408945687,
+ "grad_norm": 0.056884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0305,
+ "step": 10516
+ },
+ {
+ "epoch": 33.60063897763578,
+ "grad_norm": 0.08154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0326,
+ "step": 10517
+ },
+ {
+ "epoch": 33.6038338658147,
+ "grad_norm": 0.044189453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0285,
+ "step": 10518
+ },
+ {
+ "epoch": 33.60702875399361,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0318,
+ "step": 10519
+ },
+ {
+ "epoch": 33.61022364217252,
+ "grad_norm": 0.056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.037,
+ "step": 10520
+ },
+ {
+ "epoch": 33.61341853035144,
+ "grad_norm": 0.055419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0466,
+ "step": 10521
+ },
+ {
+ "epoch": 33.61661341853035,
+ "grad_norm": 0.05126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0362,
+ "step": 10522
+ },
+ {
+ "epoch": 33.61980830670927,
+ "grad_norm": 0.05810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0314,
+ "step": 10523
+ },
+ {
+ "epoch": 33.62300319488818,
+ "grad_norm": 0.056396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0316,
+ "step": 10524
+ },
+ {
+ "epoch": 33.626198083067095,
+ "grad_norm": 0.05224609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0397,
+ "step": 10525
+ },
+ {
+ "epoch": 33.629392971246006,
+ "grad_norm": 0.052978515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0391,
+ "step": 10526
+ },
+ {
+ "epoch": 33.63258785942492,
+ "grad_norm": 0.059326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0342,
+ "step": 10527
+ },
+ {
+ "epoch": 33.635782747603834,
+ "grad_norm": 0.05078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0287,
+ "step": 10528
+ },
+ {
+ "epoch": 33.638977635782744,
+ "grad_norm": 0.0498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0375,
+ "step": 10529
+ },
+ {
+ "epoch": 33.64217252396166,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0327,
+ "step": 10530
+ },
+ {
+ "epoch": 33.64536741214057,
+ "grad_norm": 0.052978515625,
+ "learning_rate": 0.0005,
+ "loss": 1.018,
+ "step": 10531
+ },
+ {
+ "epoch": 33.64856230031949,
+ "grad_norm": 0.083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.026,
+ "step": 10532
+ },
+ {
+ "epoch": 33.6517571884984,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.04,
+ "step": 10533
+ },
+ {
+ "epoch": 33.65495207667732,
+ "grad_norm": 0.04833984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0463,
+ "step": 10534
+ },
+ {
+ "epoch": 33.65814696485623,
+ "grad_norm": 0.091796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0428,
+ "step": 10535
+ },
+ {
+ "epoch": 33.66134185303515,
+ "grad_norm": 0.062255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0411,
+ "step": 10536
+ },
+ {
+ "epoch": 33.66453674121406,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0306,
+ "step": 10537
+ },
+ {
+ "epoch": 33.66773162939297,
+ "grad_norm": 0.057861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0376,
+ "step": 10538
+ },
+ {
+ "epoch": 33.670926517571885,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0312,
+ "step": 10539
+ },
+ {
+ "epoch": 33.674121405750796,
+ "grad_norm": 0.05615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0393,
+ "step": 10540
+ },
+ {
+ "epoch": 33.677316293929714,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0257,
+ "step": 10541
+ },
+ {
+ "epoch": 33.680511182108624,
+ "grad_norm": 0.05810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0405,
+ "step": 10542
+ },
+ {
+ "epoch": 33.68370607028754,
+ "grad_norm": 0.0546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0439,
+ "step": 10543
+ },
+ {
+ "epoch": 33.68690095846645,
+ "grad_norm": 0.1044921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0436,
+ "step": 10544
+ },
+ {
+ "epoch": 33.69009584664537,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0505,
+ "step": 10545
+ },
+ {
+ "epoch": 33.69329073482428,
+ "grad_norm": 0.059326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0306,
+ "step": 10546
+ },
+ {
+ "epoch": 33.6964856230032,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0372,
+ "step": 10547
+ },
+ {
+ "epoch": 33.69968051118211,
+ "grad_norm": 0.09130859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0373,
+ "step": 10548
+ },
+ {
+ "epoch": 33.70287539936102,
+ "grad_norm": 0.0595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.041,
+ "step": 10549
+ },
+ {
+ "epoch": 33.70607028753994,
+ "grad_norm": 0.083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0419,
+ "step": 10550
+ },
+ {
+ "epoch": 33.70926517571885,
+ "grad_norm": 0.05908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0308,
+ "step": 10551
+ },
+ {
+ "epoch": 33.712460063897765,
+ "grad_norm": 0.052490234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0391,
+ "step": 10552
+ },
+ {
+ "epoch": 33.715654952076676,
+ "grad_norm": 0.052490234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0406,
+ "step": 10553
+ },
+ {
+ "epoch": 33.718849840255594,
+ "grad_norm": 0.050048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.041,
+ "step": 10554
+ },
+ {
+ "epoch": 33.722044728434504,
+ "grad_norm": 0.0615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0344,
+ "step": 10555
+ },
+ {
+ "epoch": 33.72523961661342,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0422,
+ "step": 10556
+ },
+ {
+ "epoch": 33.72843450479233,
+ "grad_norm": 0.06201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0409,
+ "step": 10557
+ },
+ {
+ "epoch": 33.73162939297124,
+ "grad_norm": 0.045166015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0327,
+ "step": 10558
+ },
+ {
+ "epoch": 33.73482428115016,
+ "grad_norm": 0.08056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.034,
+ "step": 10559
+ },
+ {
+ "epoch": 33.73801916932907,
+ "grad_norm": 0.0478515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0271,
+ "step": 10560
+ },
+ {
+ "epoch": 33.74121405750799,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0433,
+ "step": 10561
+ },
+ {
+ "epoch": 33.7444089456869,
+ "grad_norm": 0.1318359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0251,
+ "step": 10562
+ },
+ {
+ "epoch": 33.74760383386582,
+ "grad_norm": 0.05126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0358,
+ "step": 10563
+ },
+ {
+ "epoch": 33.75079872204473,
+ "grad_norm": 0.11767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0364,
+ "step": 10564
+ },
+ {
+ "epoch": 33.753993610223645,
+ "grad_norm": 0.045166015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0357,
+ "step": 10565
+ },
+ {
+ "epoch": 33.757188498402556,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0438,
+ "step": 10566
+ },
+ {
+ "epoch": 33.760383386581466,
+ "grad_norm": 0.04345703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0341,
+ "step": 10567
+ },
+ {
+ "epoch": 33.763578274760384,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0378,
+ "step": 10568
+ },
+ {
+ "epoch": 33.766773162939295,
+ "grad_norm": 0.0498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0339,
+ "step": 10569
+ },
+ {
+ "epoch": 33.76996805111821,
+ "grad_norm": 0.08837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.047,
+ "step": 10570
+ },
+ {
+ "epoch": 33.77316293929712,
+ "grad_norm": 0.0810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0184,
+ "step": 10571
+ },
+ {
+ "epoch": 33.77635782747604,
+ "grad_norm": 0.056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0396,
+ "step": 10572
+ },
+ {
+ "epoch": 33.77955271565495,
+ "grad_norm": 0.056884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.032,
+ "step": 10573
+ },
+ {
+ "epoch": 33.78274760383387,
+ "grad_norm": 0.055908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0273,
+ "step": 10574
+ },
+ {
+ "epoch": 33.78594249201278,
+ "grad_norm": 0.058349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.044,
+ "step": 10575
+ },
+ {
+ "epoch": 33.78913738019169,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0307,
+ "step": 10576
+ },
+ {
+ "epoch": 33.79233226837061,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0253,
+ "step": 10577
+ },
+ {
+ "epoch": 33.79552715654952,
+ "grad_norm": 0.05615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0459,
+ "step": 10578
+ },
+ {
+ "epoch": 33.798722044728436,
+ "grad_norm": 0.046142578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0292,
+ "step": 10579
+ },
+ {
+ "epoch": 33.801916932907346,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0327,
+ "step": 10580
+ },
+ {
+ "epoch": 33.805111821086264,
+ "grad_norm": 0.0517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0403,
+ "step": 10581
+ },
+ {
+ "epoch": 33.808306709265175,
+ "grad_norm": 0.0595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0378,
+ "step": 10582
+ },
+ {
+ "epoch": 33.81150159744409,
+ "grad_norm": 0.080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0317,
+ "step": 10583
+ },
+ {
+ "epoch": 33.814696485623,
+ "grad_norm": 0.0810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0301,
+ "step": 10584
+ },
+ {
+ "epoch": 33.81789137380191,
+ "grad_norm": 0.052490234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0299,
+ "step": 10585
+ },
+ {
+ "epoch": 33.82108626198083,
+ "grad_norm": 0.058349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0322,
+ "step": 10586
+ },
+ {
+ "epoch": 33.82428115015974,
+ "grad_norm": 0.060302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0484,
+ "step": 10587
+ },
+ {
+ "epoch": 33.82747603833866,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0233,
+ "step": 10588
+ },
+ {
+ "epoch": 33.83067092651757,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.038,
+ "step": 10589
+ },
+ {
+ "epoch": 33.83386581469649,
+ "grad_norm": 0.057373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0334,
+ "step": 10590
+ },
+ {
+ "epoch": 33.8370607028754,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0396,
+ "step": 10591
+ },
+ {
+ "epoch": 33.840255591054316,
+ "grad_norm": 0.0498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0414,
+ "step": 10592
+ },
+ {
+ "epoch": 33.843450479233226,
+ "grad_norm": 0.05322265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0249,
+ "step": 10593
+ },
+ {
+ "epoch": 33.846645367412144,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0272,
+ "step": 10594
+ },
+ {
+ "epoch": 33.849840255591054,
+ "grad_norm": 0.0439453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0353,
+ "step": 10595
+ },
+ {
+ "epoch": 33.853035143769965,
+ "grad_norm": 0.062255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.05,
+ "step": 10596
+ },
+ {
+ "epoch": 33.85623003194888,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0344,
+ "step": 10597
+ },
+ {
+ "epoch": 33.85942492012779,
+ "grad_norm": 0.07861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0314,
+ "step": 10598
+ },
+ {
+ "epoch": 33.86261980830671,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.028,
+ "step": 10599
+ },
+ {
+ "epoch": 33.86581469648562,
+ "grad_norm": 0.087890625,
+ "learning_rate": 0.0005,
+ "loss": 1.033,
+ "step": 10600
+ },
+ {
+ "epoch": 33.86900958466454,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0346,
+ "step": 10601
+ },
+ {
+ "epoch": 33.87220447284345,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0377,
+ "step": 10602
+ },
+ {
+ "epoch": 33.87539936102237,
+ "grad_norm": 0.0537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0398,
+ "step": 10603
+ },
+ {
+ "epoch": 33.87859424920128,
+ "grad_norm": 0.060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0353,
+ "step": 10604
+ },
+ {
+ "epoch": 33.88178913738019,
+ "grad_norm": 0.04833984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0457,
+ "step": 10605
+ },
+ {
+ "epoch": 33.884984025559106,
+ "grad_norm": 0.055908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0216,
+ "step": 10606
+ },
+ {
+ "epoch": 33.88817891373802,
+ "grad_norm": 0.06103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0206,
+ "step": 10607
+ },
+ {
+ "epoch": 33.891373801916934,
+ "grad_norm": 0.061279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.038,
+ "step": 10608
+ },
+ {
+ "epoch": 33.894568690095845,
+ "grad_norm": 0.05078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0402,
+ "step": 10609
+ },
+ {
+ "epoch": 33.89776357827476,
+ "grad_norm": 0.053955078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0368,
+ "step": 10610
+ },
+ {
+ "epoch": 33.90095846645367,
+ "grad_norm": 0.0595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0264,
+ "step": 10611
+ },
+ {
+ "epoch": 33.90415335463259,
+ "grad_norm": 0.0546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0473,
+ "step": 10612
+ },
+ {
+ "epoch": 33.9073482428115,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0459,
+ "step": 10613
+ },
+ {
+ "epoch": 33.91054313099041,
+ "grad_norm": 0.060791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0325,
+ "step": 10614
+ },
+ {
+ "epoch": 33.91373801916933,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0421,
+ "step": 10615
+ },
+ {
+ "epoch": 33.91693290734824,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0326,
+ "step": 10616
+ },
+ {
+ "epoch": 33.92012779552716,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0372,
+ "step": 10617
+ },
+ {
+ "epoch": 33.92332268370607,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0434,
+ "step": 10618
+ },
+ {
+ "epoch": 33.926517571884986,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0416,
+ "step": 10619
+ },
+ {
+ "epoch": 33.9297124600639,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0403,
+ "step": 10620
+ },
+ {
+ "epoch": 33.932907348242814,
+ "grad_norm": 0.0986328125,
+ "learning_rate": 0.0005,
+ "loss": 1.036,
+ "step": 10621
+ },
+ {
+ "epoch": 33.936102236421725,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0415,
+ "step": 10622
+ },
+ {
+ "epoch": 33.93929712460064,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0331,
+ "step": 10623
+ },
+ {
+ "epoch": 33.94249201277955,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0345,
+ "step": 10624
+ },
+ {
+ "epoch": 33.945686900958464,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0548,
+ "step": 10625
+ },
+ {
+ "epoch": 33.94888178913738,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0273,
+ "step": 10626
+ },
+ {
+ "epoch": 33.95207667731629,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0269,
+ "step": 10627
+ },
+ {
+ "epoch": 33.95527156549521,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0419,
+ "step": 10628
+ },
+ {
+ "epoch": 33.95846645367412,
+ "grad_norm": 0.08056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.032,
+ "step": 10629
+ },
+ {
+ "epoch": 33.96166134185304,
+ "grad_norm": 0.0537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0304,
+ "step": 10630
+ },
+ {
+ "epoch": 33.96485623003195,
+ "grad_norm": 0.09033203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0437,
+ "step": 10631
+ },
+ {
+ "epoch": 33.968051118210866,
+ "grad_norm": 0.053466796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0373,
+ "step": 10632
+ },
+ {
+ "epoch": 33.97124600638978,
+ "grad_norm": 0.08056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0338,
+ "step": 10633
+ },
+ {
+ "epoch": 33.97444089456869,
+ "grad_norm": 0.050537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0338,
+ "step": 10634
+ },
+ {
+ "epoch": 33.977635782747605,
+ "grad_norm": 0.046142578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0349,
+ "step": 10635
+ },
+ {
+ "epoch": 33.980830670926515,
+ "grad_norm": 0.05078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0308,
+ "step": 10636
+ },
+ {
+ "epoch": 33.98402555910543,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0256,
+ "step": 10637
+ },
+ {
+ "epoch": 33.98722044728434,
+ "grad_norm": 0.055908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0239,
+ "step": 10638
+ },
+ {
+ "epoch": 33.99041533546326,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0396,
+ "step": 10639
+ },
+ {
+ "epoch": 33.99361022364217,
+ "grad_norm": 0.058837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0365,
+ "step": 10640
+ },
+ {
+ "epoch": 33.99680511182109,
+ "grad_norm": 0.0927734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0395,
+ "step": 10641
+ },
+ {
+ "epoch": 34.0,
+ "grad_norm": 0.1318359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0385,
+ "step": 10642
+ },
+ {
+ "epoch": 34.00319488817891,
+ "grad_norm": 0.1083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0536,
+ "step": 10643
+ },
+ {
+ "epoch": 34.00638977635783,
+ "grad_norm": 0.1337890625,
+ "learning_rate": 0.0005,
+ "loss": 1.03,
+ "step": 10644
+ },
+ {
+ "epoch": 34.00958466453674,
+ "grad_norm": 0.07861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0352,
+ "step": 10645
+ },
+ {
+ "epoch": 34.01277955271566,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0318,
+ "step": 10646
+ },
+ {
+ "epoch": 34.01597444089457,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0407,
+ "step": 10647
+ },
+ {
+ "epoch": 34.019169329073485,
+ "grad_norm": 0.08740234375,
+ "learning_rate": 0.0005,
+ "loss": 1.03,
+ "step": 10648
+ },
+ {
+ "epoch": 34.022364217252395,
+ "grad_norm": 0.1357421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0397,
+ "step": 10649
+ },
+ {
+ "epoch": 34.02555910543131,
+ "grad_norm": 0.1376953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0415,
+ "step": 10650
+ },
+ {
+ "epoch": 34.02875399361022,
+ "grad_norm": 0.169921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0344,
+ "step": 10651
+ },
+ {
+ "epoch": 34.031948881789134,
+ "grad_norm": 0.095703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0344,
+ "step": 10652
+ },
+ {
+ "epoch": 34.03514376996805,
+ "grad_norm": 0.12255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0338,
+ "step": 10653
+ },
+ {
+ "epoch": 34.03833865814696,
+ "grad_norm": 0.087890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0235,
+ "step": 10654
+ },
+ {
+ "epoch": 34.04153354632588,
+ "grad_norm": 0.0888671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0196,
+ "step": 10655
+ },
+ {
+ "epoch": 34.04472843450479,
+ "grad_norm": 0.0986328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0183,
+ "step": 10656
+ },
+ {
+ "epoch": 34.04792332268371,
+ "grad_norm": 0.083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0468,
+ "step": 10657
+ },
+ {
+ "epoch": 34.05111821086262,
+ "grad_norm": 0.05908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0326,
+ "step": 10658
+ },
+ {
+ "epoch": 34.054313099041536,
+ "grad_norm": 0.099609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0417,
+ "step": 10659
+ },
+ {
+ "epoch": 34.05750798722045,
+ "grad_norm": 0.08154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0404,
+ "step": 10660
+ },
+ {
+ "epoch": 34.06070287539936,
+ "grad_norm": 0.125,
+ "learning_rate": 0.0005,
+ "loss": 1.0339,
+ "step": 10661
+ },
+ {
+ "epoch": 34.063897763578275,
+ "grad_norm": 0.1904296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0363,
+ "step": 10662
+ },
+ {
+ "epoch": 34.067092651757186,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0388,
+ "step": 10663
+ },
+ {
+ "epoch": 34.0702875399361,
+ "grad_norm": 0.224609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0442,
+ "step": 10664
+ },
+ {
+ "epoch": 34.073482428115014,
+ "grad_norm": 0.10205078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0443,
+ "step": 10665
+ },
+ {
+ "epoch": 34.07667731629393,
+ "grad_norm": 0.142578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0386,
+ "step": 10666
+ },
+ {
+ "epoch": 34.07987220447284,
+ "grad_norm": 0.142578125,
+ "learning_rate": 0.0005,
+ "loss": 1.043,
+ "step": 10667
+ },
+ {
+ "epoch": 34.08306709265176,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0362,
+ "step": 10668
+ },
+ {
+ "epoch": 34.08626198083067,
+ "grad_norm": 0.12890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0413,
+ "step": 10669
+ },
+ {
+ "epoch": 34.08945686900959,
+ "grad_norm": 0.04541015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0219,
+ "step": 10670
+ },
+ {
+ "epoch": 34.0926517571885,
+ "grad_norm": 0.1240234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0207,
+ "step": 10671
+ },
+ {
+ "epoch": 34.09584664536741,
+ "grad_norm": 0.043701171875,
+ "learning_rate": 0.0005,
+ "loss": 1.034,
+ "step": 10672
+ },
+ {
+ "epoch": 34.09904153354633,
+ "grad_norm": 0.1416015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0317,
+ "step": 10673
+ },
+ {
+ "epoch": 34.10223642172524,
+ "grad_norm": 0.046630859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0274,
+ "step": 10674
+ },
+ {
+ "epoch": 34.105431309904155,
+ "grad_norm": 0.10693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0423,
+ "step": 10675
+ },
+ {
+ "epoch": 34.108626198083066,
+ "grad_norm": 0.103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0356,
+ "step": 10676
+ },
+ {
+ "epoch": 34.11182108626198,
+ "grad_norm": 0.12060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0392,
+ "step": 10677
+ },
+ {
+ "epoch": 34.115015974440894,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0426,
+ "step": 10678
+ },
+ {
+ "epoch": 34.11821086261981,
+ "grad_norm": 0.0966796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0297,
+ "step": 10679
+ },
+ {
+ "epoch": 34.12140575079872,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0341,
+ "step": 10680
+ },
+ {
+ "epoch": 34.12460063897763,
+ "grad_norm": 0.1005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0324,
+ "step": 10681
+ },
+ {
+ "epoch": 34.12779552715655,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0343,
+ "step": 10682
+ },
+ {
+ "epoch": 34.13099041533546,
+ "grad_norm": 0.123046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0363,
+ "step": 10683
+ },
+ {
+ "epoch": 34.13418530351438,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0243,
+ "step": 10684
+ },
+ {
+ "epoch": 34.13738019169329,
+ "grad_norm": 0.054931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0289,
+ "step": 10685
+ },
+ {
+ "epoch": 34.14057507987221,
+ "grad_norm": 0.05224609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0381,
+ "step": 10686
+ },
+ {
+ "epoch": 34.14376996805112,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0281,
+ "step": 10687
+ },
+ {
+ "epoch": 34.146964856230035,
+ "grad_norm": 0.0517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0397,
+ "step": 10688
+ },
+ {
+ "epoch": 34.150159744408946,
+ "grad_norm": 0.09423828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0434,
+ "step": 10689
+ },
+ {
+ "epoch": 34.153354632587856,
+ "grad_norm": 0.08154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0427,
+ "step": 10690
+ },
+ {
+ "epoch": 34.156549520766774,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0224,
+ "step": 10691
+ },
+ {
+ "epoch": 34.159744408945684,
+ "grad_norm": 0.047119140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0268,
+ "step": 10692
+ },
+ {
+ "epoch": 34.1629392971246,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0476,
+ "step": 10693
+ },
+ {
+ "epoch": 34.16613418530351,
+ "grad_norm": 0.08447265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0317,
+ "step": 10694
+ },
+ {
+ "epoch": 34.16932907348243,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0443,
+ "step": 10695
+ },
+ {
+ "epoch": 34.17252396166134,
+ "grad_norm": 0.07861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0345,
+ "step": 10696
+ },
+ {
+ "epoch": 34.17571884984026,
+ "grad_norm": 0.057861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.029,
+ "step": 10697
+ },
+ {
+ "epoch": 34.17891373801917,
+ "grad_norm": 0.0517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.03,
+ "step": 10698
+ },
+ {
+ "epoch": 34.18210862619808,
+ "grad_norm": 0.058837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0415,
+ "step": 10699
+ },
+ {
+ "epoch": 34.185303514377,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0385,
+ "step": 10700
+ },
+ {
+ "epoch": 34.18849840255591,
+ "grad_norm": 0.0869140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0332,
+ "step": 10701
+ },
+ {
+ "epoch": 34.191693290734825,
+ "grad_norm": 0.1181640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0287,
+ "step": 10702
+ },
+ {
+ "epoch": 34.194888178913736,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0272,
+ "step": 10703
+ },
+ {
+ "epoch": 34.198083067092654,
+ "grad_norm": 0.1357421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0347,
+ "step": 10704
+ },
+ {
+ "epoch": 34.201277955271564,
+ "grad_norm": 0.049072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.041,
+ "step": 10705
+ },
+ {
+ "epoch": 34.20447284345048,
+ "grad_norm": 0.1162109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0471,
+ "step": 10706
+ },
+ {
+ "epoch": 34.20766773162939,
+ "grad_norm": 0.059814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0314,
+ "step": 10707
+ },
+ {
+ "epoch": 34.21086261980831,
+ "grad_norm": 0.1396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.032,
+ "step": 10708
+ },
+ {
+ "epoch": 34.21405750798722,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0334,
+ "step": 10709
+ },
+ {
+ "epoch": 34.21725239616613,
+ "grad_norm": 0.16796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0404,
+ "step": 10710
+ },
+ {
+ "epoch": 34.22044728434505,
+ "grad_norm": 0.0546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0357,
+ "step": 10711
+ },
+ {
+ "epoch": 34.22364217252396,
+ "grad_norm": 0.134765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0469,
+ "step": 10712
+ },
+ {
+ "epoch": 34.22683706070288,
+ "grad_norm": 0.0927734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0444,
+ "step": 10713
+ },
+ {
+ "epoch": 34.23003194888179,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0284,
+ "step": 10714
+ },
+ {
+ "epoch": 34.233226837060705,
+ "grad_norm": 0.056396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0399,
+ "step": 10715
+ },
+ {
+ "epoch": 34.236421725239616,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0439,
+ "step": 10716
+ },
+ {
+ "epoch": 34.239616613418534,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0344,
+ "step": 10717
+ },
+ {
+ "epoch": 34.242811501597444,
+ "grad_norm": 0.05517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0349,
+ "step": 10718
+ },
+ {
+ "epoch": 34.246006389776355,
+ "grad_norm": 0.08984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0404,
+ "step": 10719
+ },
+ {
+ "epoch": 34.24920127795527,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0406,
+ "step": 10720
+ },
+ {
+ "epoch": 34.25239616613418,
+ "grad_norm": 0.1201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0327,
+ "step": 10721
+ },
+ {
+ "epoch": 34.2555910543131,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0267,
+ "step": 10722
+ },
+ {
+ "epoch": 34.25878594249201,
+ "grad_norm": 0.10595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0304,
+ "step": 10723
+ },
+ {
+ "epoch": 34.26198083067093,
+ "grad_norm": 0.048583984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0365,
+ "step": 10724
+ },
+ {
+ "epoch": 34.26517571884984,
+ "grad_norm": 0.08984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0389,
+ "step": 10725
+ },
+ {
+ "epoch": 34.26837060702876,
+ "grad_norm": 0.0498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0378,
+ "step": 10726
+ },
+ {
+ "epoch": 34.27156549520767,
+ "grad_norm": 0.0849609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0164,
+ "step": 10727
+ },
+ {
+ "epoch": 34.27476038338658,
+ "grad_norm": 0.05322265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0397,
+ "step": 10728
+ },
+ {
+ "epoch": 34.277955271565496,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0448,
+ "step": 10729
+ },
+ {
+ "epoch": 34.281150159744406,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0475,
+ "step": 10730
+ },
+ {
+ "epoch": 34.284345047923324,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0326,
+ "step": 10731
+ },
+ {
+ "epoch": 34.287539936102235,
+ "grad_norm": 0.09765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0342,
+ "step": 10732
+ },
+ {
+ "epoch": 34.29073482428115,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0404,
+ "step": 10733
+ },
+ {
+ "epoch": 34.29392971246006,
+ "grad_norm": 0.059814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.026,
+ "step": 10734
+ },
+ {
+ "epoch": 34.29712460063898,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.041,
+ "step": 10735
+ },
+ {
+ "epoch": 34.30031948881789,
+ "grad_norm": 0.056396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0316,
+ "step": 10736
+ },
+ {
+ "epoch": 34.3035143769968,
+ "grad_norm": 0.05615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0253,
+ "step": 10737
+ },
+ {
+ "epoch": 34.30670926517572,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.034,
+ "step": 10738
+ },
+ {
+ "epoch": 34.30990415335463,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.035,
+ "step": 10739
+ },
+ {
+ "epoch": 34.31309904153355,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0406,
+ "step": 10740
+ },
+ {
+ "epoch": 34.31629392971246,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0285,
+ "step": 10741
+ },
+ {
+ "epoch": 34.319488817891376,
+ "grad_norm": 0.0615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0398,
+ "step": 10742
+ },
+ {
+ "epoch": 34.322683706070286,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0367,
+ "step": 10743
+ },
+ {
+ "epoch": 34.325878594249204,
+ "grad_norm": 0.061279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0294,
+ "step": 10744
+ },
+ {
+ "epoch": 34.329073482428115,
+ "grad_norm": 0.062255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0394,
+ "step": 10745
+ },
+ {
+ "epoch": 34.33226837060703,
+ "grad_norm": 0.087890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0408,
+ "step": 10746
+ },
+ {
+ "epoch": 34.33546325878594,
+ "grad_norm": 0.09814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0342,
+ "step": 10747
+ },
+ {
+ "epoch": 34.33865814696485,
+ "grad_norm": 0.09033203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0277,
+ "step": 10748
+ },
+ {
+ "epoch": 34.34185303514377,
+ "grad_norm": 0.052978515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0311,
+ "step": 10749
+ },
+ {
+ "epoch": 34.34504792332268,
+ "grad_norm": 0.12158203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0371,
+ "step": 10750
+ },
+ {
+ "epoch": 34.3482428115016,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0386,
+ "step": 10751
+ },
+ {
+ "epoch": 34.35143769968051,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0407,
+ "step": 10752
+ },
+ {
+ "epoch": 34.35463258785943,
+ "grad_norm": 0.091796875,
+ "learning_rate": 0.0005,
+ "loss": 1.031,
+ "step": 10753
+ },
+ {
+ "epoch": 34.35782747603834,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0317,
+ "step": 10754
+ },
+ {
+ "epoch": 34.361022364217256,
+ "grad_norm": 0.1171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0458,
+ "step": 10755
+ },
+ {
+ "epoch": 34.364217252396166,
+ "grad_norm": 0.099609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0221,
+ "step": 10756
+ },
+ {
+ "epoch": 34.36741214057508,
+ "grad_norm": 0.08154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0305,
+ "step": 10757
+ },
+ {
+ "epoch": 34.370607028753994,
+ "grad_norm": 0.10595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0307,
+ "step": 10758
+ },
+ {
+ "epoch": 34.373801916932905,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0349,
+ "step": 10759
+ },
+ {
+ "epoch": 34.37699680511182,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0343,
+ "step": 10760
+ },
+ {
+ "epoch": 34.38019169329073,
+ "grad_norm": 0.11572265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0227,
+ "step": 10761
+ },
+ {
+ "epoch": 34.38338658146965,
+ "grad_norm": 0.05419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0359,
+ "step": 10762
+ },
+ {
+ "epoch": 34.38658146964856,
+ "grad_norm": 0.1357421875,
+ "learning_rate": 0.0005,
+ "loss": 1.034,
+ "step": 10763
+ },
+ {
+ "epoch": 34.38977635782748,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0399,
+ "step": 10764
+ },
+ {
+ "epoch": 34.39297124600639,
+ "grad_norm": 0.10302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0393,
+ "step": 10765
+ },
+ {
+ "epoch": 34.3961661341853,
+ "grad_norm": 0.10693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0305,
+ "step": 10766
+ },
+ {
+ "epoch": 34.39936102236422,
+ "grad_norm": 0.09423828125,
+ "learning_rate": 0.0005,
+ "loss": 1.028,
+ "step": 10767
+ },
+ {
+ "epoch": 34.40255591054313,
+ "grad_norm": 0.12109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0134,
+ "step": 10768
+ },
+ {
+ "epoch": 34.405750798722046,
+ "grad_norm": 0.140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0448,
+ "step": 10769
+ },
+ {
+ "epoch": 34.40894568690096,
+ "grad_norm": 0.1845703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0238,
+ "step": 10770
+ },
+ {
+ "epoch": 34.412140575079874,
+ "grad_norm": 0.140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0295,
+ "step": 10771
+ },
+ {
+ "epoch": 34.415335463258785,
+ "grad_norm": 0.30078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0312,
+ "step": 10772
+ },
+ {
+ "epoch": 34.4185303514377,
+ "grad_norm": 0.146484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0328,
+ "step": 10773
+ },
+ {
+ "epoch": 34.42172523961661,
+ "grad_norm": 0.1240234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0349,
+ "step": 10774
+ },
+ {
+ "epoch": 34.424920127795524,
+ "grad_norm": 0.21484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0323,
+ "step": 10775
+ },
+ {
+ "epoch": 34.42811501597444,
+ "grad_norm": 0.055908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0433,
+ "step": 10776
+ },
+ {
+ "epoch": 34.43130990415335,
+ "grad_norm": 0.2138671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0352,
+ "step": 10777
+ },
+ {
+ "epoch": 34.43450479233227,
+ "grad_norm": 0.154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0325,
+ "step": 10778
+ },
+ {
+ "epoch": 34.43769968051118,
+ "grad_norm": 0.109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0511,
+ "step": 10779
+ },
+ {
+ "epoch": 34.4408945686901,
+ "grad_norm": 0.12158203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0228,
+ "step": 10780
+ },
+ {
+ "epoch": 34.44408945686901,
+ "grad_norm": 0.11962890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0363,
+ "step": 10781
+ },
+ {
+ "epoch": 34.447284345047926,
+ "grad_norm": 0.15234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0293,
+ "step": 10782
+ },
+ {
+ "epoch": 34.45047923322684,
+ "grad_norm": 0.115234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0396,
+ "step": 10783
+ },
+ {
+ "epoch": 34.453674121405754,
+ "grad_norm": 0.224609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0313,
+ "step": 10784
+ },
+ {
+ "epoch": 34.456869009584665,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0415,
+ "step": 10785
+ },
+ {
+ "epoch": 34.460063897763575,
+ "grad_norm": 0.2275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0342,
+ "step": 10786
+ },
+ {
+ "epoch": 34.46325878594249,
+ "grad_norm": 0.054443359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0301,
+ "step": 10787
+ },
+ {
+ "epoch": 34.466453674121404,
+ "grad_norm": 0.169921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0375,
+ "step": 10788
+ },
+ {
+ "epoch": 34.46964856230032,
+ "grad_norm": 0.0966796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0357,
+ "step": 10789
+ },
+ {
+ "epoch": 34.47284345047923,
+ "grad_norm": 0.099609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0386,
+ "step": 10790
+ },
+ {
+ "epoch": 34.47603833865815,
+ "grad_norm": 0.08154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0369,
+ "step": 10791
+ },
+ {
+ "epoch": 34.47923322683706,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0412,
+ "step": 10792
+ },
+ {
+ "epoch": 34.48242811501598,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0485,
+ "step": 10793
+ },
+ {
+ "epoch": 34.48562300319489,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0256,
+ "step": 10794
+ },
+ {
+ "epoch": 34.4888178913738,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0335,
+ "step": 10795
+ },
+ {
+ "epoch": 34.49201277955272,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0383,
+ "step": 10796
+ },
+ {
+ "epoch": 34.49520766773163,
+ "grad_norm": 0.056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0524,
+ "step": 10797
+ },
+ {
+ "epoch": 34.498402555910545,
+ "grad_norm": 0.09619140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0328,
+ "step": 10798
+ },
+ {
+ "epoch": 34.501597444089455,
+ "grad_norm": 0.05908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0397,
+ "step": 10799
+ },
+ {
+ "epoch": 34.50479233226837,
+ "grad_norm": 0.08984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0398,
+ "step": 10800
+ },
+ {
+ "epoch": 34.50798722044728,
+ "grad_norm": 0.08056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0329,
+ "step": 10801
+ },
+ {
+ "epoch": 34.5111821086262,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.039,
+ "step": 10802
+ },
+ {
+ "epoch": 34.51437699680511,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0406,
+ "step": 10803
+ },
+ {
+ "epoch": 34.51757188498402,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.034,
+ "step": 10804
+ },
+ {
+ "epoch": 34.52076677316294,
+ "grad_norm": 0.0537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0402,
+ "step": 10805
+ },
+ {
+ "epoch": 34.52396166134185,
+ "grad_norm": 0.08349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0313,
+ "step": 10806
+ },
+ {
+ "epoch": 34.52715654952077,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0336,
+ "step": 10807
+ },
+ {
+ "epoch": 34.53035143769968,
+ "grad_norm": 0.062255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0417,
+ "step": 10808
+ },
+ {
+ "epoch": 34.533546325878596,
+ "grad_norm": 0.058349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0387,
+ "step": 10809
+ },
+ {
+ "epoch": 34.53674121405751,
+ "grad_norm": 0.050537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0336,
+ "step": 10810
+ },
+ {
+ "epoch": 34.539936102236425,
+ "grad_norm": 0.058837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0377,
+ "step": 10811
+ },
+ {
+ "epoch": 34.543130990415335,
+ "grad_norm": 0.05419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0389,
+ "step": 10812
+ },
+ {
+ "epoch": 34.546325878594246,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.037,
+ "step": 10813
+ },
+ {
+ "epoch": 34.54952076677316,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0385,
+ "step": 10814
+ },
+ {
+ "epoch": 34.552715654952074,
+ "grad_norm": 0.051025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0411,
+ "step": 10815
+ },
+ {
+ "epoch": 34.55591054313099,
+ "grad_norm": 0.049072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0306,
+ "step": 10816
+ },
+ {
+ "epoch": 34.5591054313099,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0386,
+ "step": 10817
+ },
+ {
+ "epoch": 34.56230031948882,
+ "grad_norm": 0.05419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0352,
+ "step": 10818
+ },
+ {
+ "epoch": 34.56549520766773,
+ "grad_norm": 0.048095703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0446,
+ "step": 10819
+ },
+ {
+ "epoch": 34.56869009584665,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0315,
+ "step": 10820
+ },
+ {
+ "epoch": 34.57188498402556,
+ "grad_norm": 0.055419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0325,
+ "step": 10821
+ },
+ {
+ "epoch": 34.575079872204476,
+ "grad_norm": 0.07861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0406,
+ "step": 10822
+ },
+ {
+ "epoch": 34.57827476038339,
+ "grad_norm": 0.055419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0417,
+ "step": 10823
+ },
+ {
+ "epoch": 34.5814696485623,
+ "grad_norm": 0.08154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0309,
+ "step": 10824
+ },
+ {
+ "epoch": 34.584664536741215,
+ "grad_norm": 0.046142578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0332,
+ "step": 10825
+ },
+ {
+ "epoch": 34.587859424920126,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0366,
+ "step": 10826
+ },
+ {
+ "epoch": 34.59105431309904,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0249,
+ "step": 10827
+ },
+ {
+ "epoch": 34.594249201277954,
+ "grad_norm": 0.0927734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0308,
+ "step": 10828
+ },
+ {
+ "epoch": 34.59744408945687,
+ "grad_norm": 0.041748046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0345,
+ "step": 10829
+ },
+ {
+ "epoch": 34.60063897763578,
+ "grad_norm": 0.058837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0461,
+ "step": 10830
+ },
+ {
+ "epoch": 34.6038338658147,
+ "grad_norm": 0.045166015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0384,
+ "step": 10831
+ },
+ {
+ "epoch": 34.60702875399361,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0281,
+ "step": 10832
+ },
+ {
+ "epoch": 34.61022364217252,
+ "grad_norm": 0.046142578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0343,
+ "step": 10833
+ },
+ {
+ "epoch": 34.61341853035144,
+ "grad_norm": 0.1123046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0278,
+ "step": 10834
+ },
+ {
+ "epoch": 34.61661341853035,
+ "grad_norm": 0.04541015625,
+ "learning_rate": 0.0005,
+ "loss": 1.041,
+ "step": 10835
+ },
+ {
+ "epoch": 34.61980830670927,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0331,
+ "step": 10836
+ },
+ {
+ "epoch": 34.62300319488818,
+ "grad_norm": 0.07861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0244,
+ "step": 10837
+ },
+ {
+ "epoch": 34.626198083067095,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0357,
+ "step": 10838
+ },
+ {
+ "epoch": 34.629392971246006,
+ "grad_norm": 0.049560546875,
+ "learning_rate": 0.0005,
+ "loss": 1.038,
+ "step": 10839
+ },
+ {
+ "epoch": 34.63258785942492,
+ "grad_norm": 0.050537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0273,
+ "step": 10840
+ },
+ {
+ "epoch": 34.635782747603834,
+ "grad_norm": 0.048095703125,
+ "learning_rate": 0.0005,
+ "loss": 1.047,
+ "step": 10841
+ },
+ {
+ "epoch": 34.638977635782744,
+ "grad_norm": 0.061767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0327,
+ "step": 10842
+ },
+ {
+ "epoch": 34.64217252396166,
+ "grad_norm": 0.055419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0442,
+ "step": 10843
+ },
+ {
+ "epoch": 34.64536741214057,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0378,
+ "step": 10844
+ },
+ {
+ "epoch": 34.64856230031949,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0419,
+ "step": 10845
+ },
+ {
+ "epoch": 34.6517571884984,
+ "grad_norm": 0.06103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0335,
+ "step": 10846
+ },
+ {
+ "epoch": 34.65495207667732,
+ "grad_norm": 0.080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0343,
+ "step": 10847
+ },
+ {
+ "epoch": 34.65814696485623,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0261,
+ "step": 10848
+ },
+ {
+ "epoch": 34.66134185303515,
+ "grad_norm": 0.09033203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0341,
+ "step": 10849
+ },
+ {
+ "epoch": 34.66453674121406,
+ "grad_norm": 0.07861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0339,
+ "step": 10850
+ },
+ {
+ "epoch": 34.66773162939297,
+ "grad_norm": 0.05322265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0286,
+ "step": 10851
+ },
+ {
+ "epoch": 34.670926517571885,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0317,
+ "step": 10852
+ },
+ {
+ "epoch": 34.674121405750796,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0467,
+ "step": 10853
+ },
+ {
+ "epoch": 34.677316293929714,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0464,
+ "step": 10854
+ },
+ {
+ "epoch": 34.680511182108624,
+ "grad_norm": 0.051025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0349,
+ "step": 10855
+ },
+ {
+ "epoch": 34.68370607028754,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0327,
+ "step": 10856
+ },
+ {
+ "epoch": 34.68690095846645,
+ "grad_norm": 0.046875,
+ "learning_rate": 0.0005,
+ "loss": 1.033,
+ "step": 10857
+ },
+ {
+ "epoch": 34.69009584664537,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0391,
+ "step": 10858
+ },
+ {
+ "epoch": 34.69329073482428,
+ "grad_norm": 0.049072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0329,
+ "step": 10859
+ },
+ {
+ "epoch": 34.6964856230032,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0352,
+ "step": 10860
+ },
+ {
+ "epoch": 34.69968051118211,
+ "grad_norm": 0.046142578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0428,
+ "step": 10861
+ },
+ {
+ "epoch": 34.70287539936102,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0322,
+ "step": 10862
+ },
+ {
+ "epoch": 34.70607028753994,
+ "grad_norm": 0.054443359375,
+ "learning_rate": 0.0005,
+ "loss": 1.03,
+ "step": 10863
+ },
+ {
+ "epoch": 34.70926517571885,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0358,
+ "step": 10864
+ },
+ {
+ "epoch": 34.712460063897765,
+ "grad_norm": 0.054931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0376,
+ "step": 10865
+ },
+ {
+ "epoch": 34.715654952076676,
+ "grad_norm": 0.0927734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0399,
+ "step": 10866
+ },
+ {
+ "epoch": 34.718849840255594,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0443,
+ "step": 10867
+ },
+ {
+ "epoch": 34.722044728434504,
+ "grad_norm": 0.150390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0377,
+ "step": 10868
+ },
+ {
+ "epoch": 34.72523961661342,
+ "grad_norm": 0.08740234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0288,
+ "step": 10869
+ },
+ {
+ "epoch": 34.72843450479233,
+ "grad_norm": 0.1728515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0322,
+ "step": 10870
+ },
+ {
+ "epoch": 34.73162939297124,
+ "grad_norm": 0.0517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0322,
+ "step": 10871
+ },
+ {
+ "epoch": 34.73482428115016,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0459,
+ "step": 10872
+ },
+ {
+ "epoch": 34.73801916932907,
+ "grad_norm": 0.047119140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0316,
+ "step": 10873
+ },
+ {
+ "epoch": 34.74121405750799,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0358,
+ "step": 10874
+ },
+ {
+ "epoch": 34.7444089456869,
+ "grad_norm": 0.051025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0297,
+ "step": 10875
+ },
+ {
+ "epoch": 34.74760383386582,
+ "grad_norm": 0.057861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0323,
+ "step": 10876
+ },
+ {
+ "epoch": 34.75079872204473,
+ "grad_norm": 0.0478515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0361,
+ "step": 10877
+ },
+ {
+ "epoch": 34.753993610223645,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0375,
+ "step": 10878
+ },
+ {
+ "epoch": 34.757188498402556,
+ "grad_norm": 0.060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0461,
+ "step": 10879
+ },
+ {
+ "epoch": 34.760383386581466,
+ "grad_norm": 0.0498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.025,
+ "step": 10880
+ },
+ {
+ "epoch": 34.763578274760384,
+ "grad_norm": 0.080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0206,
+ "step": 10881
+ },
+ {
+ "epoch": 34.766773162939295,
+ "grad_norm": 0.057861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0398,
+ "step": 10882
+ },
+ {
+ "epoch": 34.76996805111821,
+ "grad_norm": 0.06005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0327,
+ "step": 10883
+ },
+ {
+ "epoch": 34.77316293929712,
+ "grad_norm": 0.060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0485,
+ "step": 10884
+ },
+ {
+ "epoch": 34.77635782747604,
+ "grad_norm": 0.0546875,
+ "learning_rate": 0.0005,
+ "loss": 1.033,
+ "step": 10885
+ },
+ {
+ "epoch": 34.77955271565495,
+ "grad_norm": 0.107421875,
+ "learning_rate": 0.0005,
+ "loss": 1.022,
+ "step": 10886
+ },
+ {
+ "epoch": 34.78274760383387,
+ "grad_norm": 0.0615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0329,
+ "step": 10887
+ },
+ {
+ "epoch": 34.78594249201278,
+ "grad_norm": 0.08740234375,
+ "learning_rate": 0.0005,
+ "loss": 1.027,
+ "step": 10888
+ },
+ {
+ "epoch": 34.78913738019169,
+ "grad_norm": 0.049072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0345,
+ "step": 10889
+ },
+ {
+ "epoch": 34.79233226837061,
+ "grad_norm": 0.0810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0406,
+ "step": 10890
+ },
+ {
+ "epoch": 34.79552715654952,
+ "grad_norm": 0.09130859375,
+ "learning_rate": 0.0005,
+ "loss": 1.024,
+ "step": 10891
+ },
+ {
+ "epoch": 34.798722044728436,
+ "grad_norm": 0.1337890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0313,
+ "step": 10892
+ },
+ {
+ "epoch": 34.801916932907346,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0296,
+ "step": 10893
+ },
+ {
+ "epoch": 34.805111821086264,
+ "grad_norm": 0.1611328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0347,
+ "step": 10894
+ },
+ {
+ "epoch": 34.808306709265175,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.048,
+ "step": 10895
+ },
+ {
+ "epoch": 34.81150159744409,
+ "grad_norm": 0.059326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.028,
+ "step": 10896
+ },
+ {
+ "epoch": 34.814696485623,
+ "grad_norm": 0.0908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0318,
+ "step": 10897
+ },
+ {
+ "epoch": 34.81789137380191,
+ "grad_norm": 0.09228515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0452,
+ "step": 10898
+ },
+ {
+ "epoch": 34.82108626198083,
+ "grad_norm": 0.058837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0408,
+ "step": 10899
+ },
+ {
+ "epoch": 34.82428115015974,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.044,
+ "step": 10900
+ },
+ {
+ "epoch": 34.82747603833866,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0462,
+ "step": 10901
+ },
+ {
+ "epoch": 34.83067092651757,
+ "grad_norm": 0.05419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0317,
+ "step": 10902
+ },
+ {
+ "epoch": 34.83386581469649,
+ "grad_norm": 0.057373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0281,
+ "step": 10903
+ },
+ {
+ "epoch": 34.8370607028754,
+ "grad_norm": 0.05712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0407,
+ "step": 10904
+ },
+ {
+ "epoch": 34.840255591054316,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0435,
+ "step": 10905
+ },
+ {
+ "epoch": 34.843450479233226,
+ "grad_norm": 0.0869140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0298,
+ "step": 10906
+ },
+ {
+ "epoch": 34.846645367412144,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0295,
+ "step": 10907
+ },
+ {
+ "epoch": 34.849840255591054,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0329,
+ "step": 10908
+ },
+ {
+ "epoch": 34.853035143769965,
+ "grad_norm": 0.0517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0348,
+ "step": 10909
+ },
+ {
+ "epoch": 34.85623003194888,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0381,
+ "step": 10910
+ },
+ {
+ "epoch": 34.85942492012779,
+ "grad_norm": 0.05078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0313,
+ "step": 10911
+ },
+ {
+ "epoch": 34.86261980830671,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0274,
+ "step": 10912
+ },
+ {
+ "epoch": 34.86581469648562,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0461,
+ "step": 10913
+ },
+ {
+ "epoch": 34.86900958466454,
+ "grad_norm": 0.10302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0184,
+ "step": 10914
+ },
+ {
+ "epoch": 34.87220447284345,
+ "grad_norm": 0.0830078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0279,
+ "step": 10915
+ },
+ {
+ "epoch": 34.87539936102237,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0363,
+ "step": 10916
+ },
+ {
+ "epoch": 34.87859424920128,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0115,
+ "step": 10917
+ },
+ {
+ "epoch": 34.88178913738019,
+ "grad_norm": 0.09423828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0432,
+ "step": 10918
+ },
+ {
+ "epoch": 34.884984025559106,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.032,
+ "step": 10919
+ },
+ {
+ "epoch": 34.88817891373802,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0355,
+ "step": 10920
+ },
+ {
+ "epoch": 34.891373801916934,
+ "grad_norm": 0.09912109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0441,
+ "step": 10921
+ },
+ {
+ "epoch": 34.894568690095845,
+ "grad_norm": 0.1328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0323,
+ "step": 10922
+ },
+ {
+ "epoch": 34.89776357827476,
+ "grad_norm": 0.08447265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0305,
+ "step": 10923
+ },
+ {
+ "epoch": 34.90095846645367,
+ "grad_norm": 0.150390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0367,
+ "step": 10924
+ },
+ {
+ "epoch": 34.90415335463259,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0322,
+ "step": 10925
+ },
+ {
+ "epoch": 34.9073482428115,
+ "grad_norm": 0.1455078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0326,
+ "step": 10926
+ },
+ {
+ "epoch": 34.91054313099041,
+ "grad_norm": 0.08203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0354,
+ "step": 10927
+ },
+ {
+ "epoch": 34.91373801916933,
+ "grad_norm": 0.142578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0388,
+ "step": 10928
+ },
+ {
+ "epoch": 34.91693290734824,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0253,
+ "step": 10929
+ },
+ {
+ "epoch": 34.92012779552716,
+ "grad_norm": 0.111328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0337,
+ "step": 10930
+ },
+ {
+ "epoch": 34.92332268370607,
+ "grad_norm": 0.08984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0291,
+ "step": 10931
+ },
+ {
+ "epoch": 34.926517571884986,
+ "grad_norm": 0.08447265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0424,
+ "step": 10932
+ },
+ {
+ "epoch": 34.9297124600639,
+ "grad_norm": 0.10009765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0506,
+ "step": 10933
+ },
+ {
+ "epoch": 34.932907348242814,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0365,
+ "step": 10934
+ },
+ {
+ "epoch": 34.936102236421725,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0325,
+ "step": 10935
+ },
+ {
+ "epoch": 34.93929712460064,
+ "grad_norm": 0.1025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0419,
+ "step": 10936
+ },
+ {
+ "epoch": 34.94249201277955,
+ "grad_norm": 0.10400390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0288,
+ "step": 10937
+ },
+ {
+ "epoch": 34.945686900958464,
+ "grad_norm": 0.1015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0281,
+ "step": 10938
+ },
+ {
+ "epoch": 34.94888178913738,
+ "grad_norm": 0.15234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0333,
+ "step": 10939
+ },
+ {
+ "epoch": 34.95207667731629,
+ "grad_norm": 0.1083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.033,
+ "step": 10940
+ },
+ {
+ "epoch": 34.95527156549521,
+ "grad_norm": 0.19921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0362,
+ "step": 10941
+ },
+ {
+ "epoch": 34.95846645367412,
+ "grad_norm": 0.060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0341,
+ "step": 10942
+ },
+ {
+ "epoch": 34.96166134185304,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0321,
+ "step": 10943
+ },
+ {
+ "epoch": 34.96485623003195,
+ "grad_norm": 0.1064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0472,
+ "step": 10944
+ },
+ {
+ "epoch": 34.968051118210866,
+ "grad_norm": 0.08447265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0407,
+ "step": 10945
+ },
+ {
+ "epoch": 34.97124600638978,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0336,
+ "step": 10946
+ },
+ {
+ "epoch": 34.97444089456869,
+ "grad_norm": 0.0859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0226,
+ "step": 10947
+ },
+ {
+ "epoch": 34.977635782747605,
+ "grad_norm": 0.05615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0354,
+ "step": 10948
+ },
+ {
+ "epoch": 34.980830670926515,
+ "grad_norm": 0.057861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0368,
+ "step": 10949
+ },
+ {
+ "epoch": 34.98402555910543,
+ "grad_norm": 0.0546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0259,
+ "step": 10950
+ },
+ {
+ "epoch": 34.98722044728434,
+ "grad_norm": 0.0615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0339,
+ "step": 10951
+ },
+ {
+ "epoch": 34.99041533546326,
+ "grad_norm": 0.05419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0322,
+ "step": 10952
+ },
+ {
+ "epoch": 34.99361022364217,
+ "grad_norm": 0.055908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0349,
+ "step": 10953
+ },
+ {
+ "epoch": 34.99680511182109,
+ "grad_norm": 0.05419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.028,
+ "step": 10954
+ },
+ {
+ "epoch": 35.0,
+ "grad_norm": 0.047119140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0317,
+ "step": 10955
+ },
+ {
+ "epoch": 35.00319488817891,
+ "grad_norm": 0.059326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0393,
+ "step": 10956
+ },
+ {
+ "epoch": 35.00638977635783,
+ "grad_norm": 0.0537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0411,
+ "step": 10957
+ },
+ {
+ "epoch": 35.00958466453674,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0423,
+ "step": 10958
+ },
+ {
+ "epoch": 35.01277955271566,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0335,
+ "step": 10959
+ },
+ {
+ "epoch": 35.01597444089457,
+ "grad_norm": 0.055908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0252,
+ "step": 10960
+ },
+ {
+ "epoch": 35.019169329073485,
+ "grad_norm": 0.049072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0253,
+ "step": 10961
+ },
+ {
+ "epoch": 35.022364217252395,
+ "grad_norm": 0.056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0258,
+ "step": 10962
+ },
+ {
+ "epoch": 35.02555910543131,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0302,
+ "step": 10963
+ },
+ {
+ "epoch": 35.02875399361022,
+ "grad_norm": 0.051025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0369,
+ "step": 10964
+ },
+ {
+ "epoch": 35.031948881789134,
+ "grad_norm": 0.1015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0328,
+ "step": 10965
+ },
+ {
+ "epoch": 35.03514376996805,
+ "grad_norm": 0.046142578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0274,
+ "step": 10966
+ },
+ {
+ "epoch": 35.03833865814696,
+ "grad_norm": 0.061767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0347,
+ "step": 10967
+ },
+ {
+ "epoch": 35.04153354632588,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0282,
+ "step": 10968
+ },
+ {
+ "epoch": 35.04472843450479,
+ "grad_norm": 0.0478515625,
+ "learning_rate": 0.0005,
+ "loss": 1.03,
+ "step": 10969
+ },
+ {
+ "epoch": 35.04792332268371,
+ "grad_norm": 0.0537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0327,
+ "step": 10970
+ },
+ {
+ "epoch": 35.05111821086262,
+ "grad_norm": 0.04931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0454,
+ "step": 10971
+ },
+ {
+ "epoch": 35.054313099041536,
+ "grad_norm": 0.053466796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0346,
+ "step": 10972
+ },
+ {
+ "epoch": 35.05750798722045,
+ "grad_norm": 0.05712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0413,
+ "step": 10973
+ },
+ {
+ "epoch": 35.06070287539936,
+ "grad_norm": 0.05615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.031,
+ "step": 10974
+ },
+ {
+ "epoch": 35.063897763578275,
+ "grad_norm": 0.050048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0435,
+ "step": 10975
+ },
+ {
+ "epoch": 35.067092651757186,
+ "grad_norm": 0.059326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.023,
+ "step": 10976
+ },
+ {
+ "epoch": 35.0702875399361,
+ "grad_norm": 0.048583984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0288,
+ "step": 10977
+ },
+ {
+ "epoch": 35.073482428115014,
+ "grad_norm": 0.057861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0437,
+ "step": 10978
+ },
+ {
+ "epoch": 35.07667731629393,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0311,
+ "step": 10979
+ },
+ {
+ "epoch": 35.07987220447284,
+ "grad_norm": 0.10546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0248,
+ "step": 10980
+ },
+ {
+ "epoch": 35.08306709265176,
+ "grad_norm": 0.125,
+ "learning_rate": 0.0005,
+ "loss": 1.0409,
+ "step": 10981
+ },
+ {
+ "epoch": 35.08626198083067,
+ "grad_norm": 0.046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0341,
+ "step": 10982
+ },
+ {
+ "epoch": 35.08945686900959,
+ "grad_norm": 0.0615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0345,
+ "step": 10983
+ },
+ {
+ "epoch": 35.0926517571885,
+ "grad_norm": 0.04736328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0377,
+ "step": 10984
+ },
+ {
+ "epoch": 35.09584664536741,
+ "grad_norm": 0.057373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0337,
+ "step": 10985
+ },
+ {
+ "epoch": 35.09904153354633,
+ "grad_norm": 0.0869140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0327,
+ "step": 10986
+ },
+ {
+ "epoch": 35.10223642172524,
+ "grad_norm": 0.095703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0252,
+ "step": 10987
+ },
+ {
+ "epoch": 35.105431309904155,
+ "grad_norm": 0.1572265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0333,
+ "step": 10988
+ },
+ {
+ "epoch": 35.108626198083066,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.041,
+ "step": 10989
+ },
+ {
+ "epoch": 35.11182108626198,
+ "grad_norm": 0.06103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0256,
+ "step": 10990
+ },
+ {
+ "epoch": 35.115015974440894,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0213,
+ "step": 10991
+ },
+ {
+ "epoch": 35.11821086261981,
+ "grad_norm": 0.060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0324,
+ "step": 10992
+ },
+ {
+ "epoch": 35.12140575079872,
+ "grad_norm": 0.055908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0364,
+ "step": 10993
+ },
+ {
+ "epoch": 35.12460063897763,
+ "grad_norm": 0.050048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0396,
+ "step": 10994
+ },
+ {
+ "epoch": 35.12779552715655,
+ "grad_norm": 0.053955078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0337,
+ "step": 10995
+ },
+ {
+ "epoch": 35.13099041533546,
+ "grad_norm": 0.05517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.028,
+ "step": 10996
+ },
+ {
+ "epoch": 35.13418530351438,
+ "grad_norm": 0.060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0352,
+ "step": 10997
+ },
+ {
+ "epoch": 35.13738019169329,
+ "grad_norm": 0.060302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0406,
+ "step": 10998
+ },
+ {
+ "epoch": 35.14057507987221,
+ "grad_norm": 0.057373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0375,
+ "step": 10999
+ },
+ {
+ "epoch": 35.14376996805112,
+ "grad_norm": 0.04638671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0411,
+ "step": 11000
+ },
+ {
+ "epoch": 35.146964856230035,
+ "grad_norm": 0.08251953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0397,
+ "step": 11001
+ },
+ {
+ "epoch": 35.150159744408946,
+ "grad_norm": 0.05126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0296,
+ "step": 11002
+ },
+ {
+ "epoch": 35.153354632587856,
+ "grad_norm": 0.05078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0264,
+ "step": 11003
+ },
+ {
+ "epoch": 35.156549520766774,
+ "grad_norm": 0.061767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0421,
+ "step": 11004
+ },
+ {
+ "epoch": 35.159744408945684,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0308,
+ "step": 11005
+ },
+ {
+ "epoch": 35.1629392971246,
+ "grad_norm": 0.051513671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0334,
+ "step": 11006
+ },
+ {
+ "epoch": 35.16613418530351,
+ "grad_norm": 0.06201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0324,
+ "step": 11007
+ },
+ {
+ "epoch": 35.16932907348243,
+ "grad_norm": 0.058349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0347,
+ "step": 11008
+ },
+ {
+ "epoch": 35.17252396166134,
+ "grad_norm": 0.083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0298,
+ "step": 11009
+ },
+ {
+ "epoch": 35.17571884984026,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0407,
+ "step": 11010
+ },
+ {
+ "epoch": 35.17891373801917,
+ "grad_norm": 0.1455078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0343,
+ "step": 11011
+ },
+ {
+ "epoch": 35.18210862619808,
+ "grad_norm": 0.1435546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0297,
+ "step": 11012
+ },
+ {
+ "epoch": 35.185303514377,
+ "grad_norm": 0.11669921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0379,
+ "step": 11013
+ },
+ {
+ "epoch": 35.18849840255591,
+ "grad_norm": 0.1328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0457,
+ "step": 11014
+ },
+ {
+ "epoch": 35.191693290734825,
+ "grad_norm": 0.05859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0374,
+ "step": 11015
+ },
+ {
+ "epoch": 35.194888178913736,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0382,
+ "step": 11016
+ },
+ {
+ "epoch": 35.198083067092654,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0363,
+ "step": 11017
+ },
+ {
+ "epoch": 35.201277955271564,
+ "grad_norm": 0.08740234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0332,
+ "step": 11018
+ },
+ {
+ "epoch": 35.20447284345048,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0359,
+ "step": 11019
+ },
+ {
+ "epoch": 35.20766773162939,
+ "grad_norm": 0.1279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0348,
+ "step": 11020
+ },
+ {
+ "epoch": 35.21086261980831,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.033,
+ "step": 11021
+ },
+ {
+ "epoch": 35.21405750798722,
+ "grad_norm": 0.142578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0331,
+ "step": 11022
+ },
+ {
+ "epoch": 35.21725239616613,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.033,
+ "step": 11023
+ },
+ {
+ "epoch": 35.22044728434505,
+ "grad_norm": 0.1982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0332,
+ "step": 11024
+ },
+ {
+ "epoch": 35.22364217252396,
+ "grad_norm": 0.08349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0388,
+ "step": 11025
+ },
+ {
+ "epoch": 35.22683706070288,
+ "grad_norm": 0.1318359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0413,
+ "step": 11026
+ },
+ {
+ "epoch": 35.23003194888179,
+ "grad_norm": 0.1240234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0347,
+ "step": 11027
+ },
+ {
+ "epoch": 35.233226837060705,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0322,
+ "step": 11028
+ },
+ {
+ "epoch": 35.236421725239616,
+ "grad_norm": 0.11865234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0409,
+ "step": 11029
+ },
+ {
+ "epoch": 35.239616613418534,
+ "grad_norm": 0.05224609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0374,
+ "step": 11030
+ },
+ {
+ "epoch": 35.242811501597444,
+ "grad_norm": 0.1279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.042,
+ "step": 11031
+ },
+ {
+ "epoch": 35.246006389776355,
+ "grad_norm": 0.060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0293,
+ "step": 11032
+ },
+ {
+ "epoch": 35.24920127795527,
+ "grad_norm": 0.12158203125,
+ "learning_rate": 0.0005,
+ "loss": 1.032,
+ "step": 11033
+ },
+ {
+ "epoch": 35.25239616613418,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0398,
+ "step": 11034
+ },
+ {
+ "epoch": 35.2555910543131,
+ "grad_norm": 0.083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.032,
+ "step": 11035
+ },
+ {
+ "epoch": 35.25878594249201,
+ "grad_norm": 0.04736328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0356,
+ "step": 11036
+ },
+ {
+ "epoch": 35.26198083067093,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0265,
+ "step": 11037
+ },
+ {
+ "epoch": 35.26517571884984,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0361,
+ "step": 11038
+ },
+ {
+ "epoch": 35.26837060702876,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0192,
+ "step": 11039
+ },
+ {
+ "epoch": 35.27156549520767,
+ "grad_norm": 0.05078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0274,
+ "step": 11040
+ },
+ {
+ "epoch": 35.27476038338658,
+ "grad_norm": 0.05908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0429,
+ "step": 11041
+ },
+ {
+ "epoch": 35.277955271565496,
+ "grad_norm": 0.052001953125,
+ "learning_rate": 0.0005,
+ "loss": 1.029,
+ "step": 11042
+ },
+ {
+ "epoch": 35.281150159744406,
+ "grad_norm": 0.06103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.036,
+ "step": 11043
+ },
+ {
+ "epoch": 35.284345047923324,
+ "grad_norm": 0.0498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0386,
+ "step": 11044
+ },
+ {
+ "epoch": 35.287539936102235,
+ "grad_norm": 0.056884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0313,
+ "step": 11045
+ },
+ {
+ "epoch": 35.29073482428115,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.032,
+ "step": 11046
+ },
+ {
+ "epoch": 35.29392971246006,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0391,
+ "step": 11047
+ },
+ {
+ "epoch": 35.29712460063898,
+ "grad_norm": 0.0869140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0421,
+ "step": 11048
+ },
+ {
+ "epoch": 35.30031948881789,
+ "grad_norm": 0.058349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0328,
+ "step": 11049
+ },
+ {
+ "epoch": 35.3035143769968,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0261,
+ "step": 11050
+ },
+ {
+ "epoch": 35.30670926517572,
+ "grad_norm": 0.056396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0361,
+ "step": 11051
+ },
+ {
+ "epoch": 35.30990415335463,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0421,
+ "step": 11052
+ },
+ {
+ "epoch": 35.31309904153355,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0339,
+ "step": 11053
+ },
+ {
+ "epoch": 35.31629392971246,
+ "grad_norm": 0.08203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0366,
+ "step": 11054
+ },
+ {
+ "epoch": 35.319488817891376,
+ "grad_norm": 0.10107421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0426,
+ "step": 11055
+ },
+ {
+ "epoch": 35.322683706070286,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0367,
+ "step": 11056
+ },
+ {
+ "epoch": 35.325878594249204,
+ "grad_norm": 0.140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0339,
+ "step": 11057
+ },
+ {
+ "epoch": 35.329073482428115,
+ "grad_norm": 0.09375,
+ "learning_rate": 0.0005,
+ "loss": 1.0442,
+ "step": 11058
+ },
+ {
+ "epoch": 35.33226837060703,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0487,
+ "step": 11059
+ },
+ {
+ "epoch": 35.33546325878594,
+ "grad_norm": 0.09521484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0309,
+ "step": 11060
+ },
+ {
+ "epoch": 35.33865814696485,
+ "grad_norm": 0.0869140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0389,
+ "step": 11061
+ },
+ {
+ "epoch": 35.34185303514377,
+ "grad_norm": 0.09912109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0314,
+ "step": 11062
+ },
+ {
+ "epoch": 35.34504792332268,
+ "grad_norm": 0.08935546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0358,
+ "step": 11063
+ },
+ {
+ "epoch": 35.3482428115016,
+ "grad_norm": 0.126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.04,
+ "step": 11064
+ },
+ {
+ "epoch": 35.35143769968051,
+ "grad_norm": 0.0576171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0412,
+ "step": 11065
+ },
+ {
+ "epoch": 35.35463258785943,
+ "grad_norm": 0.11962890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0406,
+ "step": 11066
+ },
+ {
+ "epoch": 35.35782747603834,
+ "grad_norm": 0.050537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.041,
+ "step": 11067
+ },
+ {
+ "epoch": 35.361022364217256,
+ "grad_norm": 0.16015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0411,
+ "step": 11068
+ },
+ {
+ "epoch": 35.364217252396166,
+ "grad_norm": 0.04931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.031,
+ "step": 11069
+ },
+ {
+ "epoch": 35.36741214057508,
+ "grad_norm": 0.140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0329,
+ "step": 11070
+ },
+ {
+ "epoch": 35.370607028753994,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0441,
+ "step": 11071
+ },
+ {
+ "epoch": 35.373801916932905,
+ "grad_norm": 0.1708984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0342,
+ "step": 11072
+ },
+ {
+ "epoch": 35.37699680511182,
+ "grad_norm": 0.058837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0307,
+ "step": 11073
+ },
+ {
+ "epoch": 35.38019169329073,
+ "grad_norm": 0.1005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0222,
+ "step": 11074
+ },
+ {
+ "epoch": 35.38338658146965,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0314,
+ "step": 11075
+ },
+ {
+ "epoch": 35.38658146964856,
+ "grad_norm": 0.0810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0442,
+ "step": 11076
+ },
+ {
+ "epoch": 35.38977635782748,
+ "grad_norm": 0.0498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0386,
+ "step": 11077
+ },
+ {
+ "epoch": 35.39297124600639,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0388,
+ "step": 11078
+ },
+ {
+ "epoch": 35.3961661341853,
+ "grad_norm": 0.06005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0366,
+ "step": 11079
+ },
+ {
+ "epoch": 35.39936102236422,
+ "grad_norm": 0.053466796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0415,
+ "step": 11080
+ },
+ {
+ "epoch": 35.40255591054313,
+ "grad_norm": 0.04931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0427,
+ "step": 11081
+ },
+ {
+ "epoch": 35.405750798722046,
+ "grad_norm": 0.07861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0241,
+ "step": 11082
+ },
+ {
+ "epoch": 35.40894568690096,
+ "grad_norm": 0.04931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0465,
+ "step": 11083
+ },
+ {
+ "epoch": 35.412140575079874,
+ "grad_norm": 0.0546875,
+ "learning_rate": 0.0005,
+ "loss": 1.029,
+ "step": 11084
+ },
+ {
+ "epoch": 35.415335463258785,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0307,
+ "step": 11085
+ },
+ {
+ "epoch": 35.4185303514377,
+ "grad_norm": 0.0498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0403,
+ "step": 11086
+ },
+ {
+ "epoch": 35.42172523961661,
+ "grad_norm": 0.058349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0317,
+ "step": 11087
+ },
+ {
+ "epoch": 35.424920127795524,
+ "grad_norm": 0.045166015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0349,
+ "step": 11088
+ },
+ {
+ "epoch": 35.42811501597444,
+ "grad_norm": 0.055908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0309,
+ "step": 11089
+ },
+ {
+ "epoch": 35.43130990415335,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0338,
+ "step": 11090
+ },
+ {
+ "epoch": 35.43450479233227,
+ "grad_norm": 0.0869140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0336,
+ "step": 11091
+ },
+ {
+ "epoch": 35.43769968051118,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0354,
+ "step": 11092
+ },
+ {
+ "epoch": 35.4408945686901,
+ "grad_norm": 0.055419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0329,
+ "step": 11093
+ },
+ {
+ "epoch": 35.44408945686901,
+ "grad_norm": 0.058349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0393,
+ "step": 11094
+ },
+ {
+ "epoch": 35.447284345047926,
+ "grad_norm": 0.0556640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0361,
+ "step": 11095
+ },
+ {
+ "epoch": 35.45047923322684,
+ "grad_norm": 0.04833984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0375,
+ "step": 11096
+ },
+ {
+ "epoch": 35.453674121405754,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0271,
+ "step": 11097
+ },
+ {
+ "epoch": 35.456869009584665,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0309,
+ "step": 11098
+ },
+ {
+ "epoch": 35.460063897763575,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.031,
+ "step": 11099
+ },
+ {
+ "epoch": 35.46325878594249,
+ "grad_norm": 0.052734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0362,
+ "step": 11100
+ },
+ {
+ "epoch": 35.466453674121404,
+ "grad_norm": 0.0458984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0389,
+ "step": 11101
+ },
+ {
+ "epoch": 35.46964856230032,
+ "grad_norm": 0.04833984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0231,
+ "step": 11102
+ },
+ {
+ "epoch": 35.47284345047923,
+ "grad_norm": 0.05859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0271,
+ "step": 11103
+ },
+ {
+ "epoch": 35.47603833865815,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.041,
+ "step": 11104
+ },
+ {
+ "epoch": 35.47923322683706,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0391,
+ "step": 11105
+ },
+ {
+ "epoch": 35.48242811501598,
+ "grad_norm": 0.0537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0373,
+ "step": 11106
+ },
+ {
+ "epoch": 35.48562300319489,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.037,
+ "step": 11107
+ },
+ {
+ "epoch": 35.4888178913738,
+ "grad_norm": 0.053466796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0323,
+ "step": 11108
+ },
+ {
+ "epoch": 35.49201277955272,
+ "grad_norm": 0.1044921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0402,
+ "step": 11109
+ },
+ {
+ "epoch": 35.49520766773163,
+ "grad_norm": 0.0830078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0332,
+ "step": 11110
+ },
+ {
+ "epoch": 35.498402555910545,
+ "grad_norm": 0.056884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0341,
+ "step": 11111
+ },
+ {
+ "epoch": 35.501597444089455,
+ "grad_norm": 0.050537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0266,
+ "step": 11112
+ },
+ {
+ "epoch": 35.50479233226837,
+ "grad_norm": 0.057861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0211,
+ "step": 11113
+ },
+ {
+ "epoch": 35.50798722044728,
+ "grad_norm": 0.055908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0496,
+ "step": 11114
+ },
+ {
+ "epoch": 35.5111821086262,
+ "grad_norm": 0.060791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0275,
+ "step": 11115
+ },
+ {
+ "epoch": 35.51437699680511,
+ "grad_norm": 0.0498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0386,
+ "step": 11116
+ },
+ {
+ "epoch": 35.51757188498402,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0386,
+ "step": 11117
+ },
+ {
+ "epoch": 35.52076677316294,
+ "grad_norm": 0.062255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0354,
+ "step": 11118
+ },
+ {
+ "epoch": 35.52396166134185,
+ "grad_norm": 0.1298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.038,
+ "step": 11119
+ },
+ {
+ "epoch": 35.52715654952077,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0441,
+ "step": 11120
+ },
+ {
+ "epoch": 35.53035143769968,
+ "grad_norm": 0.0830078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0268,
+ "step": 11121
+ },
+ {
+ "epoch": 35.533546325878596,
+ "grad_norm": 0.054931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0299,
+ "step": 11122
+ },
+ {
+ "epoch": 35.53674121405751,
+ "grad_norm": 0.1044921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0324,
+ "step": 11123
+ },
+ {
+ "epoch": 35.539936102236425,
+ "grad_norm": 0.10107421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0534,
+ "step": 11124
+ },
+ {
+ "epoch": 35.543130990415335,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0352,
+ "step": 11125
+ },
+ {
+ "epoch": 35.546325878594246,
+ "grad_norm": 0.0849609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0328,
+ "step": 11126
+ },
+ {
+ "epoch": 35.54952076677316,
+ "grad_norm": 0.08740234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0404,
+ "step": 11127
+ },
+ {
+ "epoch": 35.552715654952074,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0288,
+ "step": 11128
+ },
+ {
+ "epoch": 35.55591054313099,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.034,
+ "step": 11129
+ },
+ {
+ "epoch": 35.5591054313099,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0372,
+ "step": 11130
+ },
+ {
+ "epoch": 35.56230031948882,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0391,
+ "step": 11131
+ },
+ {
+ "epoch": 35.56549520766773,
+ "grad_norm": 0.05859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0361,
+ "step": 11132
+ },
+ {
+ "epoch": 35.56869009584665,
+ "grad_norm": 0.058837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0261,
+ "step": 11133
+ },
+ {
+ "epoch": 35.57188498402556,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0398,
+ "step": 11134
+ },
+ {
+ "epoch": 35.575079872204476,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0347,
+ "step": 11135
+ },
+ {
+ "epoch": 35.57827476038339,
+ "grad_norm": 0.0576171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0203,
+ "step": 11136
+ },
+ {
+ "epoch": 35.5814696485623,
+ "grad_norm": 0.055419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0291,
+ "step": 11137
+ },
+ {
+ "epoch": 35.584664536741215,
+ "grad_norm": 0.05517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.038,
+ "step": 11138
+ },
+ {
+ "epoch": 35.587859424920126,
+ "grad_norm": 0.0517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.042,
+ "step": 11139
+ },
+ {
+ "epoch": 35.59105431309904,
+ "grad_norm": 0.054443359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0275,
+ "step": 11140
+ },
+ {
+ "epoch": 35.594249201277954,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0323,
+ "step": 11141
+ },
+ {
+ "epoch": 35.59744408945687,
+ "grad_norm": 0.04833984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0186,
+ "step": 11142
+ },
+ {
+ "epoch": 35.60063897763578,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0321,
+ "step": 11143
+ },
+ {
+ "epoch": 35.6038338658147,
+ "grad_norm": 0.055908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0366,
+ "step": 11144
+ },
+ {
+ "epoch": 35.60702875399361,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0403,
+ "step": 11145
+ },
+ {
+ "epoch": 35.61022364217252,
+ "grad_norm": 0.0595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0336,
+ "step": 11146
+ },
+ {
+ "epoch": 35.61341853035144,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0351,
+ "step": 11147
+ },
+ {
+ "epoch": 35.61661341853035,
+ "grad_norm": 0.056884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0447,
+ "step": 11148
+ },
+ {
+ "epoch": 35.61980830670927,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0357,
+ "step": 11149
+ },
+ {
+ "epoch": 35.62300319488818,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0417,
+ "step": 11150
+ },
+ {
+ "epoch": 35.626198083067095,
+ "grad_norm": 0.05078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0359,
+ "step": 11151
+ },
+ {
+ "epoch": 35.629392971246006,
+ "grad_norm": 0.08154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0398,
+ "step": 11152
+ },
+ {
+ "epoch": 35.63258785942492,
+ "grad_norm": 0.052001953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0312,
+ "step": 11153
+ },
+ {
+ "epoch": 35.635782747603834,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.037,
+ "step": 11154
+ },
+ {
+ "epoch": 35.638977635782744,
+ "grad_norm": 0.060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0295,
+ "step": 11155
+ },
+ {
+ "epoch": 35.64217252396166,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0349,
+ "step": 11156
+ },
+ {
+ "epoch": 35.64536741214057,
+ "grad_norm": 0.0810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0248,
+ "step": 11157
+ },
+ {
+ "epoch": 35.64856230031949,
+ "grad_norm": 0.057861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0429,
+ "step": 11158
+ },
+ {
+ "epoch": 35.6517571884984,
+ "grad_norm": 0.08935546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0332,
+ "step": 11159
+ },
+ {
+ "epoch": 35.65495207667732,
+ "grad_norm": 0.0595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0271,
+ "step": 11160
+ },
+ {
+ "epoch": 35.65814696485623,
+ "grad_norm": 0.04638671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0381,
+ "step": 11161
+ },
+ {
+ "epoch": 35.66134185303515,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0349,
+ "step": 11162
+ },
+ {
+ "epoch": 35.66453674121406,
+ "grad_norm": 0.05126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0436,
+ "step": 11163
+ },
+ {
+ "epoch": 35.66773162939297,
+ "grad_norm": 0.052001953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0245,
+ "step": 11164
+ },
+ {
+ "epoch": 35.670926517571885,
+ "grad_norm": 0.05419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0243,
+ "step": 11165
+ },
+ {
+ "epoch": 35.674121405750796,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0433,
+ "step": 11166
+ },
+ {
+ "epoch": 35.677316293929714,
+ "grad_norm": 0.053955078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0325,
+ "step": 11167
+ },
+ {
+ "epoch": 35.680511182108624,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0302,
+ "step": 11168
+ },
+ {
+ "epoch": 35.68370607028754,
+ "grad_norm": 0.05908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0349,
+ "step": 11169
+ },
+ {
+ "epoch": 35.68690095846645,
+ "grad_norm": 0.06201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0301,
+ "step": 11170
+ },
+ {
+ "epoch": 35.69009584664537,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0315,
+ "step": 11171
+ },
+ {
+ "epoch": 35.69329073482428,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0416,
+ "step": 11172
+ },
+ {
+ "epoch": 35.6964856230032,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0302,
+ "step": 11173
+ },
+ {
+ "epoch": 35.69968051118211,
+ "grad_norm": 0.05078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0404,
+ "step": 11174
+ },
+ {
+ "epoch": 35.70287539936102,
+ "grad_norm": 0.1005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0325,
+ "step": 11175
+ },
+ {
+ "epoch": 35.70607028753994,
+ "grad_norm": 0.055419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0214,
+ "step": 11176
+ },
+ {
+ "epoch": 35.70926517571885,
+ "grad_norm": 0.10205078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0381,
+ "step": 11177
+ },
+ {
+ "epoch": 35.712460063897765,
+ "grad_norm": 0.0546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0407,
+ "step": 11178
+ },
+ {
+ "epoch": 35.715654952076676,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0405,
+ "step": 11179
+ },
+ {
+ "epoch": 35.718849840255594,
+ "grad_norm": 0.057373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0532,
+ "step": 11180
+ },
+ {
+ "epoch": 35.722044728434504,
+ "grad_norm": 0.052001953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0445,
+ "step": 11181
+ },
+ {
+ "epoch": 35.72523961661342,
+ "grad_norm": 0.0576171875,
+ "learning_rate": 0.0005,
+ "loss": 1.036,
+ "step": 11182
+ },
+ {
+ "epoch": 35.72843450479233,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0301,
+ "step": 11183
+ },
+ {
+ "epoch": 35.73162939297124,
+ "grad_norm": 0.10595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0438,
+ "step": 11184
+ },
+ {
+ "epoch": 35.73482428115016,
+ "grad_norm": 0.055908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0367,
+ "step": 11185
+ },
+ {
+ "epoch": 35.73801916932907,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0283,
+ "step": 11186
+ },
+ {
+ "epoch": 35.74121405750799,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0388,
+ "step": 11187
+ },
+ {
+ "epoch": 35.7444089456869,
+ "grad_norm": 0.05712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0345,
+ "step": 11188
+ },
+ {
+ "epoch": 35.74760383386582,
+ "grad_norm": 0.06103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0255,
+ "step": 11189
+ },
+ {
+ "epoch": 35.75079872204473,
+ "grad_norm": 0.056396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0316,
+ "step": 11190
+ },
+ {
+ "epoch": 35.753993610223645,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0268,
+ "step": 11191
+ },
+ {
+ "epoch": 35.757188498402556,
+ "grad_norm": 0.048095703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0289,
+ "step": 11192
+ },
+ {
+ "epoch": 35.760383386581466,
+ "grad_norm": 0.08056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0263,
+ "step": 11193
+ },
+ {
+ "epoch": 35.763578274760384,
+ "grad_norm": 0.05322265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0456,
+ "step": 11194
+ },
+ {
+ "epoch": 35.766773162939295,
+ "grad_norm": 0.08251953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0317,
+ "step": 11195
+ },
+ {
+ "epoch": 35.76996805111821,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0353,
+ "step": 11196
+ },
+ {
+ "epoch": 35.77316293929712,
+ "grad_norm": 0.06103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0323,
+ "step": 11197
+ },
+ {
+ "epoch": 35.77635782747604,
+ "grad_norm": 0.05908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0354,
+ "step": 11198
+ },
+ {
+ "epoch": 35.77955271565495,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0328,
+ "step": 11199
+ },
+ {
+ "epoch": 35.78274760383387,
+ "grad_norm": 0.10498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0397,
+ "step": 11200
+ },
+ {
+ "epoch": 35.78594249201278,
+ "grad_norm": 0.224609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0264,
+ "step": 11201
+ },
+ {
+ "epoch": 35.78913738019169,
+ "grad_norm": 0.04541015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0377,
+ "step": 11202
+ },
+ {
+ "epoch": 35.79233226837061,
+ "grad_norm": 0.1318359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0334,
+ "step": 11203
+ },
+ {
+ "epoch": 35.79552715654952,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0351,
+ "step": 11204
+ },
+ {
+ "epoch": 35.798722044728436,
+ "grad_norm": 0.080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0405,
+ "step": 11205
+ },
+ {
+ "epoch": 35.801916932907346,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0263,
+ "step": 11206
+ },
+ {
+ "epoch": 35.805111821086264,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0224,
+ "step": 11207
+ },
+ {
+ "epoch": 35.808306709265175,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0364,
+ "step": 11208
+ },
+ {
+ "epoch": 35.81150159744409,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0409,
+ "step": 11209
+ },
+ {
+ "epoch": 35.814696485623,
+ "grad_norm": 0.047607421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0211,
+ "step": 11210
+ },
+ {
+ "epoch": 35.81789137380191,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0315,
+ "step": 11211
+ },
+ {
+ "epoch": 35.82108626198083,
+ "grad_norm": 0.08935546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0317,
+ "step": 11212
+ },
+ {
+ "epoch": 35.82428115015974,
+ "grad_norm": 0.052978515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0205,
+ "step": 11213
+ },
+ {
+ "epoch": 35.82747603833866,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0389,
+ "step": 11214
+ },
+ {
+ "epoch": 35.83067092651757,
+ "grad_norm": 0.052490234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0328,
+ "step": 11215
+ },
+ {
+ "epoch": 35.83386581469649,
+ "grad_norm": 0.08154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.029,
+ "step": 11216
+ },
+ {
+ "epoch": 35.8370607028754,
+ "grad_norm": 0.062255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0342,
+ "step": 11217
+ },
+ {
+ "epoch": 35.840255591054316,
+ "grad_norm": 0.12060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0213,
+ "step": 11218
+ },
+ {
+ "epoch": 35.843450479233226,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0418,
+ "step": 11219
+ },
+ {
+ "epoch": 35.846645367412144,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0357,
+ "step": 11220
+ },
+ {
+ "epoch": 35.849840255591054,
+ "grad_norm": 0.08740234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0381,
+ "step": 11221
+ },
+ {
+ "epoch": 35.853035143769965,
+ "grad_norm": 0.05810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0366,
+ "step": 11222
+ },
+ {
+ "epoch": 35.85623003194888,
+ "grad_norm": 0.10888671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0407,
+ "step": 11223
+ },
+ {
+ "epoch": 35.85942492012779,
+ "grad_norm": 0.052001953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0323,
+ "step": 11224
+ },
+ {
+ "epoch": 35.86261980830671,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.048,
+ "step": 11225
+ },
+ {
+ "epoch": 35.86581469648562,
+ "grad_norm": 0.05908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.038,
+ "step": 11226
+ },
+ {
+ "epoch": 35.86900958466454,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.043,
+ "step": 11227
+ },
+ {
+ "epoch": 35.87220447284345,
+ "grad_norm": 0.10107421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0383,
+ "step": 11228
+ },
+ {
+ "epoch": 35.87539936102237,
+ "grad_norm": 0.0498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0411,
+ "step": 11229
+ },
+ {
+ "epoch": 35.87859424920128,
+ "grad_norm": 0.059326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0366,
+ "step": 11230
+ },
+ {
+ "epoch": 35.88178913738019,
+ "grad_norm": 0.056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.045,
+ "step": 11231
+ },
+ {
+ "epoch": 35.884984025559106,
+ "grad_norm": 0.0556640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0317,
+ "step": 11232
+ },
+ {
+ "epoch": 35.88817891373802,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0355,
+ "step": 11233
+ },
+ {
+ "epoch": 35.891373801916934,
+ "grad_norm": 0.0595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0389,
+ "step": 11234
+ },
+ {
+ "epoch": 35.894568690095845,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.035,
+ "step": 11235
+ },
+ {
+ "epoch": 35.89776357827476,
+ "grad_norm": 0.053466796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0442,
+ "step": 11236
+ },
+ {
+ "epoch": 35.90095846645367,
+ "grad_norm": 0.05712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0414,
+ "step": 11237
+ },
+ {
+ "epoch": 35.90415335463259,
+ "grad_norm": 0.09130859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0376,
+ "step": 11238
+ },
+ {
+ "epoch": 35.9073482428115,
+ "grad_norm": 0.049072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0339,
+ "step": 11239
+ },
+ {
+ "epoch": 35.91054313099041,
+ "grad_norm": 0.054443359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0361,
+ "step": 11240
+ },
+ {
+ "epoch": 35.91373801916933,
+ "grad_norm": 0.05908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0406,
+ "step": 11241
+ },
+ {
+ "epoch": 35.91693290734824,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.03,
+ "step": 11242
+ },
+ {
+ "epoch": 35.92012779552716,
+ "grad_norm": 0.0615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0432,
+ "step": 11243
+ },
+ {
+ "epoch": 35.92332268370607,
+ "grad_norm": 0.0810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0361,
+ "step": 11244
+ },
+ {
+ "epoch": 35.926517571884986,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0393,
+ "step": 11245
+ },
+ {
+ "epoch": 35.9297124600639,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0387,
+ "step": 11246
+ },
+ {
+ "epoch": 35.932907348242814,
+ "grad_norm": 0.1044921875,
+ "learning_rate": 0.0005,
+ "loss": 1.049,
+ "step": 11247
+ },
+ {
+ "epoch": 35.936102236421725,
+ "grad_norm": 0.103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0299,
+ "step": 11248
+ },
+ {
+ "epoch": 35.93929712460064,
+ "grad_norm": 0.10498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0234,
+ "step": 11249
+ },
+ {
+ "epoch": 35.94249201277955,
+ "grad_norm": 0.1240234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0393,
+ "step": 11250
+ },
+ {
+ "epoch": 35.945686900958464,
+ "grad_norm": 0.1044921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0397,
+ "step": 11251
+ },
+ {
+ "epoch": 35.94888178913738,
+ "grad_norm": 0.0830078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0319,
+ "step": 11252
+ },
+ {
+ "epoch": 35.95207667731629,
+ "grad_norm": 0.1044921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0377,
+ "step": 11253
+ },
+ {
+ "epoch": 35.95527156549521,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0354,
+ "step": 11254
+ },
+ {
+ "epoch": 35.95846645367412,
+ "grad_norm": 0.11572265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0339,
+ "step": 11255
+ },
+ {
+ "epoch": 35.96166134185304,
+ "grad_norm": 0.05859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0303,
+ "step": 11256
+ },
+ {
+ "epoch": 35.96485623003195,
+ "grad_norm": 0.1259765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0305,
+ "step": 11257
+ },
+ {
+ "epoch": 35.968051118210866,
+ "grad_norm": 0.0537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.035,
+ "step": 11258
+ },
+ {
+ "epoch": 35.97124600638978,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0528,
+ "step": 11259
+ },
+ {
+ "epoch": 35.97444089456869,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0323,
+ "step": 11260
+ },
+ {
+ "epoch": 35.977635782747605,
+ "grad_norm": 0.08154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0304,
+ "step": 11261
+ },
+ {
+ "epoch": 35.980830670926515,
+ "grad_norm": 0.055419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0383,
+ "step": 11262
+ },
+ {
+ "epoch": 35.98402555910543,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0313,
+ "step": 11263
+ },
+ {
+ "epoch": 35.98722044728434,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0287,
+ "step": 11264
+ },
+ {
+ "epoch": 35.99041533546326,
+ "grad_norm": 0.062255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0392,
+ "step": 11265
+ },
+ {
+ "epoch": 35.99361022364217,
+ "grad_norm": 0.053466796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0461,
+ "step": 11266
+ },
+ {
+ "epoch": 35.99680511182109,
+ "grad_norm": 0.060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0243,
+ "step": 11267
+ },
+ {
+ "epoch": 36.0,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0374,
+ "step": 11268
+ },
+ {
+ "epoch": 36.00319488817891,
+ "grad_norm": 0.0947265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0301,
+ "step": 11269
+ },
+ {
+ "epoch": 36.00638977635783,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0336,
+ "step": 11270
+ },
+ {
+ "epoch": 36.00958466453674,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0337,
+ "step": 11271
+ },
+ {
+ "epoch": 36.01277955271566,
+ "grad_norm": 0.091796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0335,
+ "step": 11272
+ },
+ {
+ "epoch": 36.01597444089457,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0275,
+ "step": 11273
+ },
+ {
+ "epoch": 36.019169329073485,
+ "grad_norm": 0.0498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0418,
+ "step": 11274
+ },
+ {
+ "epoch": 36.022364217252395,
+ "grad_norm": 0.05126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0412,
+ "step": 11275
+ },
+ {
+ "epoch": 36.02555910543131,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0364,
+ "step": 11276
+ },
+ {
+ "epoch": 36.02875399361022,
+ "grad_norm": 0.054443359375,
+ "learning_rate": 0.0005,
+ "loss": 1.036,
+ "step": 11277
+ },
+ {
+ "epoch": 36.031948881789134,
+ "grad_norm": 0.0576171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0323,
+ "step": 11278
+ },
+ {
+ "epoch": 36.03514376996805,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0301,
+ "step": 11279
+ },
+ {
+ "epoch": 36.03833865814696,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.044,
+ "step": 11280
+ },
+ {
+ "epoch": 36.04153354632588,
+ "grad_norm": 0.0908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0407,
+ "step": 11281
+ },
+ {
+ "epoch": 36.04472843450479,
+ "grad_norm": 0.051025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0296,
+ "step": 11282
+ },
+ {
+ "epoch": 36.04792332268371,
+ "grad_norm": 0.09228515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0337,
+ "step": 11283
+ },
+ {
+ "epoch": 36.05111821086262,
+ "grad_norm": 0.05859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0444,
+ "step": 11284
+ },
+ {
+ "epoch": 36.054313099041536,
+ "grad_norm": 0.08447265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0299,
+ "step": 11285
+ },
+ {
+ "epoch": 36.05750798722045,
+ "grad_norm": 0.04541015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0314,
+ "step": 11286
+ },
+ {
+ "epoch": 36.06070287539936,
+ "grad_norm": 0.0859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0245,
+ "step": 11287
+ },
+ {
+ "epoch": 36.063897763578275,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0342,
+ "step": 11288
+ },
+ {
+ "epoch": 36.067092651757186,
+ "grad_norm": 0.09033203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0215,
+ "step": 11289
+ },
+ {
+ "epoch": 36.0702875399361,
+ "grad_norm": 0.05224609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0152,
+ "step": 11290
+ },
+ {
+ "epoch": 36.073482428115014,
+ "grad_norm": 0.08251953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0353,
+ "step": 11291
+ },
+ {
+ "epoch": 36.07667731629393,
+ "grad_norm": 0.057373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0424,
+ "step": 11292
+ },
+ {
+ "epoch": 36.07987220447284,
+ "grad_norm": 0.0595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0289,
+ "step": 11293
+ },
+ {
+ "epoch": 36.08306709265176,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0322,
+ "step": 11294
+ },
+ {
+ "epoch": 36.08626198083067,
+ "grad_norm": 0.048583984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0143,
+ "step": 11295
+ },
+ {
+ "epoch": 36.08945686900959,
+ "grad_norm": 0.09716796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0354,
+ "step": 11296
+ },
+ {
+ "epoch": 36.0926517571885,
+ "grad_norm": 0.099609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0435,
+ "step": 11297
+ },
+ {
+ "epoch": 36.09584664536741,
+ "grad_norm": 0.1513671875,
+ "learning_rate": 0.0005,
+ "loss": 1.04,
+ "step": 11298
+ },
+ {
+ "epoch": 36.09904153354633,
+ "grad_norm": 0.208984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0308,
+ "step": 11299
+ },
+ {
+ "epoch": 36.10223642172524,
+ "grad_norm": 0.1455078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0388,
+ "step": 11300
+ },
+ {
+ "epoch": 36.105431309904155,
+ "grad_norm": 0.3515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0393,
+ "step": 11301
+ },
+ {
+ "epoch": 36.108626198083066,
+ "grad_norm": 0.2109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0283,
+ "step": 11302
+ },
+ {
+ "epoch": 36.11182108626198,
+ "grad_norm": 0.1416015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0324,
+ "step": 11303
+ },
+ {
+ "epoch": 36.115015974440894,
+ "grad_norm": 0.275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0419,
+ "step": 11304
+ },
+ {
+ "epoch": 36.11821086261981,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.036,
+ "step": 11305
+ },
+ {
+ "epoch": 36.12140575079872,
+ "grad_norm": 0.1904296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0274,
+ "step": 11306
+ },
+ {
+ "epoch": 36.12460063897763,
+ "grad_norm": 0.11083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0366,
+ "step": 11307
+ },
+ {
+ "epoch": 36.12779552715655,
+ "grad_norm": 0.16796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0457,
+ "step": 11308
+ },
+ {
+ "epoch": 36.13099041533546,
+ "grad_norm": 0.1884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0271,
+ "step": 11309
+ },
+ {
+ "epoch": 36.13418530351438,
+ "grad_norm": 0.16015625,
+ "learning_rate": 0.0005,
+ "loss": 1.016,
+ "step": 11310
+ },
+ {
+ "epoch": 36.13738019169329,
+ "grad_norm": 0.333984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0292,
+ "step": 11311
+ },
+ {
+ "epoch": 36.14057507987221,
+ "grad_norm": 0.12158203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0386,
+ "step": 11312
+ },
+ {
+ "epoch": 36.14376996805112,
+ "grad_norm": 0.166015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0376,
+ "step": 11313
+ },
+ {
+ "epoch": 36.146964856230035,
+ "grad_norm": 0.1513671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0324,
+ "step": 11314
+ },
+ {
+ "epoch": 36.150159744408946,
+ "grad_norm": 0.11376953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0286,
+ "step": 11315
+ },
+ {
+ "epoch": 36.153354632587856,
+ "grad_norm": 0.2080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0354,
+ "step": 11316
+ },
+ {
+ "epoch": 36.156549520766774,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0356,
+ "step": 11317
+ },
+ {
+ "epoch": 36.159744408945684,
+ "grad_norm": 0.1416015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0372,
+ "step": 11318
+ },
+ {
+ "epoch": 36.1629392971246,
+ "grad_norm": 0.06005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.037,
+ "step": 11319
+ },
+ {
+ "epoch": 36.16613418530351,
+ "grad_norm": 0.10400390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0364,
+ "step": 11320
+ },
+ {
+ "epoch": 36.16932907348243,
+ "grad_norm": 0.04833984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0298,
+ "step": 11321
+ },
+ {
+ "epoch": 36.17252396166134,
+ "grad_norm": 0.10791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0445,
+ "step": 11322
+ },
+ {
+ "epoch": 36.17571884984026,
+ "grad_norm": 0.048583984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0383,
+ "step": 11323
+ },
+ {
+ "epoch": 36.17891373801917,
+ "grad_norm": 0.057373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0393,
+ "step": 11324
+ },
+ {
+ "epoch": 36.18210862619808,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0527,
+ "step": 11325
+ },
+ {
+ "epoch": 36.185303514377,
+ "grad_norm": 0.048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0356,
+ "step": 11326
+ },
+ {
+ "epoch": 36.18849840255591,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0362,
+ "step": 11327
+ },
+ {
+ "epoch": 36.191693290734825,
+ "grad_norm": 0.054931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0392,
+ "step": 11328
+ },
+ {
+ "epoch": 36.194888178913736,
+ "grad_norm": 0.08251953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0245,
+ "step": 11329
+ },
+ {
+ "epoch": 36.198083067092654,
+ "grad_norm": 0.0947265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0455,
+ "step": 11330
+ },
+ {
+ "epoch": 36.201277955271564,
+ "grad_norm": 0.04638671875,
+ "learning_rate": 0.0005,
+ "loss": 1.036,
+ "step": 11331
+ },
+ {
+ "epoch": 36.20447284345048,
+ "grad_norm": 0.046630859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0226,
+ "step": 11332
+ },
+ {
+ "epoch": 36.20766773162939,
+ "grad_norm": 0.052734375,
+ "learning_rate": 0.0005,
+ "loss": 1.032,
+ "step": 11333
+ },
+ {
+ "epoch": 36.21086261980831,
+ "grad_norm": 0.049072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0426,
+ "step": 11334
+ },
+ {
+ "epoch": 36.21405750798722,
+ "grad_norm": 0.062255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0388,
+ "step": 11335
+ },
+ {
+ "epoch": 36.21725239616613,
+ "grad_norm": 0.04541015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0263,
+ "step": 11336
+ },
+ {
+ "epoch": 36.22044728434505,
+ "grad_norm": 0.056884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0371,
+ "step": 11337
+ },
+ {
+ "epoch": 36.22364217252396,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0358,
+ "step": 11338
+ },
+ {
+ "epoch": 36.22683706070288,
+ "grad_norm": 0.0556640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0391,
+ "step": 11339
+ },
+ {
+ "epoch": 36.23003194888179,
+ "grad_norm": 0.08837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0288,
+ "step": 11340
+ },
+ {
+ "epoch": 36.233226837060705,
+ "grad_norm": 0.046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0293,
+ "step": 11341
+ },
+ {
+ "epoch": 36.236421725239616,
+ "grad_norm": 0.087890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0433,
+ "step": 11342
+ },
+ {
+ "epoch": 36.239616613418534,
+ "grad_norm": 0.04833984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0246,
+ "step": 11343
+ },
+ {
+ "epoch": 36.242811501597444,
+ "grad_norm": 0.051513671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0398,
+ "step": 11344
+ },
+ {
+ "epoch": 36.246006389776355,
+ "grad_norm": 0.08642578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0304,
+ "step": 11345
+ },
+ {
+ "epoch": 36.24920127795527,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0321,
+ "step": 11346
+ },
+ {
+ "epoch": 36.25239616613418,
+ "grad_norm": 0.10498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0321,
+ "step": 11347
+ },
+ {
+ "epoch": 36.2555910543131,
+ "grad_norm": 0.08740234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0345,
+ "step": 11348
+ },
+ {
+ "epoch": 36.25878594249201,
+ "grad_norm": 0.10498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0407,
+ "step": 11349
+ },
+ {
+ "epoch": 36.26198083067093,
+ "grad_norm": 0.10791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0289,
+ "step": 11350
+ },
+ {
+ "epoch": 36.26517571884984,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0481,
+ "step": 11351
+ },
+ {
+ "epoch": 36.26837060702876,
+ "grad_norm": 0.11767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0297,
+ "step": 11352
+ },
+ {
+ "epoch": 36.27156549520767,
+ "grad_norm": 0.1259765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0411,
+ "step": 11353
+ },
+ {
+ "epoch": 36.27476038338658,
+ "grad_norm": 0.1337890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0383,
+ "step": 11354
+ },
+ {
+ "epoch": 36.277955271565496,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0266,
+ "step": 11355
+ },
+ {
+ "epoch": 36.281150159744406,
+ "grad_norm": 0.125,
+ "learning_rate": 0.0005,
+ "loss": 1.0274,
+ "step": 11356
+ },
+ {
+ "epoch": 36.284345047923324,
+ "grad_norm": 0.060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0308,
+ "step": 11357
+ },
+ {
+ "epoch": 36.287539936102235,
+ "grad_norm": 0.1845703125,
+ "learning_rate": 0.0005,
+ "loss": 1.035,
+ "step": 11358
+ },
+ {
+ "epoch": 36.29073482428115,
+ "grad_norm": 0.1025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0272,
+ "step": 11359
+ },
+ {
+ "epoch": 36.29392971246006,
+ "grad_norm": 0.13671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0359,
+ "step": 11360
+ },
+ {
+ "epoch": 36.29712460063898,
+ "grad_norm": 0.09423828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0335,
+ "step": 11361
+ },
+ {
+ "epoch": 36.30031948881789,
+ "grad_norm": 0.1220703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0398,
+ "step": 11362
+ },
+ {
+ "epoch": 36.3035143769968,
+ "grad_norm": 0.10205078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0199,
+ "step": 11363
+ },
+ {
+ "epoch": 36.30670926517572,
+ "grad_norm": 0.126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0371,
+ "step": 11364
+ },
+ {
+ "epoch": 36.30990415335463,
+ "grad_norm": 0.154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.028,
+ "step": 11365
+ },
+ {
+ "epoch": 36.31309904153355,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0261,
+ "step": 11366
+ },
+ {
+ "epoch": 36.31629392971246,
+ "grad_norm": 0.1552734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0457,
+ "step": 11367
+ },
+ {
+ "epoch": 36.319488817891376,
+ "grad_norm": 0.060302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.038,
+ "step": 11368
+ },
+ {
+ "epoch": 36.322683706070286,
+ "grad_norm": 0.1259765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0316,
+ "step": 11369
+ },
+ {
+ "epoch": 36.325878594249204,
+ "grad_norm": 0.07861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.033,
+ "step": 11370
+ },
+ {
+ "epoch": 36.329073482428115,
+ "grad_norm": 0.19140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0335,
+ "step": 11371
+ },
+ {
+ "epoch": 36.33226837060703,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0258,
+ "step": 11372
+ },
+ {
+ "epoch": 36.33546325878594,
+ "grad_norm": 0.1376953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0365,
+ "step": 11373
+ },
+ {
+ "epoch": 36.33865814696485,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0391,
+ "step": 11374
+ },
+ {
+ "epoch": 36.34185303514377,
+ "grad_norm": 0.11572265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0401,
+ "step": 11375
+ },
+ {
+ "epoch": 36.34504792332268,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0382,
+ "step": 11376
+ },
+ {
+ "epoch": 36.3482428115016,
+ "grad_norm": 0.10009765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0293,
+ "step": 11377
+ },
+ {
+ "epoch": 36.35143769968051,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0324,
+ "step": 11378
+ },
+ {
+ "epoch": 36.35463258785943,
+ "grad_norm": 0.12109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0354,
+ "step": 11379
+ },
+ {
+ "epoch": 36.35782747603834,
+ "grad_norm": 0.057373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0329,
+ "step": 11380
+ },
+ {
+ "epoch": 36.361022364217256,
+ "grad_norm": 0.1162109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0412,
+ "step": 11381
+ },
+ {
+ "epoch": 36.364217252396166,
+ "grad_norm": 0.052001953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0376,
+ "step": 11382
+ },
+ {
+ "epoch": 36.36741214057508,
+ "grad_norm": 0.09619140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0385,
+ "step": 11383
+ },
+ {
+ "epoch": 36.370607028753994,
+ "grad_norm": 0.052734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0191,
+ "step": 11384
+ },
+ {
+ "epoch": 36.373801916932905,
+ "grad_norm": 0.0810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.029,
+ "step": 11385
+ },
+ {
+ "epoch": 36.37699680511182,
+ "grad_norm": 0.058349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0408,
+ "step": 11386
+ },
+ {
+ "epoch": 36.38019169329073,
+ "grad_norm": 0.08251953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0433,
+ "step": 11387
+ },
+ {
+ "epoch": 36.38338658146965,
+ "grad_norm": 0.05517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0439,
+ "step": 11388
+ },
+ {
+ "epoch": 36.38658146964856,
+ "grad_norm": 0.06005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0311,
+ "step": 11389
+ },
+ {
+ "epoch": 36.38977635782748,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0357,
+ "step": 11390
+ },
+ {
+ "epoch": 36.39297124600639,
+ "grad_norm": 0.056884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0372,
+ "step": 11391
+ },
+ {
+ "epoch": 36.3961661341853,
+ "grad_norm": 0.07861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.033,
+ "step": 11392
+ },
+ {
+ "epoch": 36.39936102236422,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0326,
+ "step": 11393
+ },
+ {
+ "epoch": 36.40255591054313,
+ "grad_norm": 0.09814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0326,
+ "step": 11394
+ },
+ {
+ "epoch": 36.405750798722046,
+ "grad_norm": 0.05029296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0407,
+ "step": 11395
+ },
+ {
+ "epoch": 36.40894568690096,
+ "grad_norm": 0.09619140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0306,
+ "step": 11396
+ },
+ {
+ "epoch": 36.412140575079874,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0371,
+ "step": 11397
+ },
+ {
+ "epoch": 36.415335463258785,
+ "grad_norm": 0.05029296875,
+ "learning_rate": 0.0005,
+ "loss": 1.041,
+ "step": 11398
+ },
+ {
+ "epoch": 36.4185303514377,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0476,
+ "step": 11399
+ },
+ {
+ "epoch": 36.42172523961661,
+ "grad_norm": 0.0498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0312,
+ "step": 11400
+ },
+ {
+ "epoch": 36.424920127795524,
+ "grad_norm": 0.053466796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0312,
+ "step": 11401
+ },
+ {
+ "epoch": 36.42811501597444,
+ "grad_norm": 0.052734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0489,
+ "step": 11402
+ },
+ {
+ "epoch": 36.43130990415335,
+ "grad_norm": 0.049072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0355,
+ "step": 11403
+ },
+ {
+ "epoch": 36.43450479233227,
+ "grad_norm": 0.057373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0303,
+ "step": 11404
+ },
+ {
+ "epoch": 36.43769968051118,
+ "grad_norm": 0.051513671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0424,
+ "step": 11405
+ },
+ {
+ "epoch": 36.4408945686901,
+ "grad_norm": 0.055908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0351,
+ "step": 11406
+ },
+ {
+ "epoch": 36.44408945686901,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0466,
+ "step": 11407
+ },
+ {
+ "epoch": 36.447284345047926,
+ "grad_norm": 0.050048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0366,
+ "step": 11408
+ },
+ {
+ "epoch": 36.45047923322684,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0229,
+ "step": 11409
+ },
+ {
+ "epoch": 36.453674121405754,
+ "grad_norm": 0.058349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0378,
+ "step": 11410
+ },
+ {
+ "epoch": 36.456869009584665,
+ "grad_norm": 0.048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0208,
+ "step": 11411
+ },
+ {
+ "epoch": 36.460063897763575,
+ "grad_norm": 0.05908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0356,
+ "step": 11412
+ },
+ {
+ "epoch": 36.46325878594249,
+ "grad_norm": 0.05224609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0375,
+ "step": 11413
+ },
+ {
+ "epoch": 36.466453674121404,
+ "grad_norm": 0.050048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0299,
+ "step": 11414
+ },
+ {
+ "epoch": 36.46964856230032,
+ "grad_norm": 0.05615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0445,
+ "step": 11415
+ },
+ {
+ "epoch": 36.47284345047923,
+ "grad_norm": 0.05712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0397,
+ "step": 11416
+ },
+ {
+ "epoch": 36.47603833865815,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0333,
+ "step": 11417
+ },
+ {
+ "epoch": 36.47923322683706,
+ "grad_norm": 0.047119140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0267,
+ "step": 11418
+ },
+ {
+ "epoch": 36.48242811501598,
+ "grad_norm": 0.048583984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0265,
+ "step": 11419
+ },
+ {
+ "epoch": 36.48562300319489,
+ "grad_norm": 0.0576171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0309,
+ "step": 11420
+ },
+ {
+ "epoch": 36.4888178913738,
+ "grad_norm": 0.04931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0409,
+ "step": 11421
+ },
+ {
+ "epoch": 36.49201277955272,
+ "grad_norm": 0.052490234375,
+ "learning_rate": 0.0005,
+ "loss": 1.037,
+ "step": 11422
+ },
+ {
+ "epoch": 36.49520766773163,
+ "grad_norm": 0.049560546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0332,
+ "step": 11423
+ },
+ {
+ "epoch": 36.498402555910545,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0374,
+ "step": 11424
+ },
+ {
+ "epoch": 36.501597444089455,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0299,
+ "step": 11425
+ },
+ {
+ "epoch": 36.50479233226837,
+ "grad_norm": 0.05322265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0415,
+ "step": 11426
+ },
+ {
+ "epoch": 36.50798722044728,
+ "grad_norm": 0.05029296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0239,
+ "step": 11427
+ },
+ {
+ "epoch": 36.5111821086262,
+ "grad_norm": 0.052978515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0301,
+ "step": 11428
+ },
+ {
+ "epoch": 36.51437699680511,
+ "grad_norm": 0.053955078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0248,
+ "step": 11429
+ },
+ {
+ "epoch": 36.51757188498402,
+ "grad_norm": 0.05224609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0508,
+ "step": 11430
+ },
+ {
+ "epoch": 36.52076677316294,
+ "grad_norm": 0.044189453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0304,
+ "step": 11431
+ },
+ {
+ "epoch": 36.52396166134185,
+ "grad_norm": 0.05029296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0384,
+ "step": 11432
+ },
+ {
+ "epoch": 36.52715654952077,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0275,
+ "step": 11433
+ },
+ {
+ "epoch": 36.53035143769968,
+ "grad_norm": 0.0556640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0199,
+ "step": 11434
+ },
+ {
+ "epoch": 36.533546325878596,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0435,
+ "step": 11435
+ },
+ {
+ "epoch": 36.53674121405751,
+ "grad_norm": 0.04541015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0252,
+ "step": 11436
+ },
+ {
+ "epoch": 36.539936102236425,
+ "grad_norm": 0.052734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0252,
+ "step": 11437
+ },
+ {
+ "epoch": 36.543130990415335,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0305,
+ "step": 11438
+ },
+ {
+ "epoch": 36.546325878594246,
+ "grad_norm": 0.052490234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0333,
+ "step": 11439
+ },
+ {
+ "epoch": 36.54952076677316,
+ "grad_norm": 0.062255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.025,
+ "step": 11440
+ },
+ {
+ "epoch": 36.552715654952074,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.041,
+ "step": 11441
+ },
+ {
+ "epoch": 36.55591054313099,
+ "grad_norm": 0.056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0342,
+ "step": 11442
+ },
+ {
+ "epoch": 36.5591054313099,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0297,
+ "step": 11443
+ },
+ {
+ "epoch": 36.56230031948882,
+ "grad_norm": 0.04931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0415,
+ "step": 11444
+ },
+ {
+ "epoch": 36.56549520766773,
+ "grad_norm": 0.052734375,
+ "learning_rate": 0.0005,
+ "loss": 1.032,
+ "step": 11445
+ },
+ {
+ "epoch": 36.56869009584665,
+ "grad_norm": 0.058349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.032,
+ "step": 11446
+ },
+ {
+ "epoch": 36.57188498402556,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0483,
+ "step": 11447
+ },
+ {
+ "epoch": 36.575079872204476,
+ "grad_norm": 0.049072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0363,
+ "step": 11448
+ },
+ {
+ "epoch": 36.57827476038339,
+ "grad_norm": 0.10009765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0404,
+ "step": 11449
+ },
+ {
+ "epoch": 36.5814696485623,
+ "grad_norm": 0.0556640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0367,
+ "step": 11450
+ },
+ {
+ "epoch": 36.584664536741215,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0213,
+ "step": 11451
+ },
+ {
+ "epoch": 36.587859424920126,
+ "grad_norm": 0.0849609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0292,
+ "step": 11452
+ },
+ {
+ "epoch": 36.59105431309904,
+ "grad_norm": 0.09423828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0195,
+ "step": 11453
+ },
+ {
+ "epoch": 36.594249201277954,
+ "grad_norm": 0.056884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0333,
+ "step": 11454
+ },
+ {
+ "epoch": 36.59744408945687,
+ "grad_norm": 0.09130859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0452,
+ "step": 11455
+ },
+ {
+ "epoch": 36.60063897763578,
+ "grad_norm": 0.051025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0394,
+ "step": 11456
+ },
+ {
+ "epoch": 36.6038338658147,
+ "grad_norm": 0.0859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0336,
+ "step": 11457
+ },
+ {
+ "epoch": 36.60702875399361,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0121,
+ "step": 11458
+ },
+ {
+ "epoch": 36.61022364217252,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0403,
+ "step": 11459
+ },
+ {
+ "epoch": 36.61341853035144,
+ "grad_norm": 0.0966796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0347,
+ "step": 11460
+ },
+ {
+ "epoch": 36.61661341853035,
+ "grad_norm": 0.11865234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0414,
+ "step": 11461
+ },
+ {
+ "epoch": 36.61980830670927,
+ "grad_norm": 0.103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0238,
+ "step": 11462
+ },
+ {
+ "epoch": 36.62300319488818,
+ "grad_norm": 0.12890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0322,
+ "step": 11463
+ },
+ {
+ "epoch": 36.626198083067095,
+ "grad_norm": 0.109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0358,
+ "step": 11464
+ },
+ {
+ "epoch": 36.629392971246006,
+ "grad_norm": 0.10546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0328,
+ "step": 11465
+ },
+ {
+ "epoch": 36.63258785942492,
+ "grad_norm": 0.134765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0518,
+ "step": 11466
+ },
+ {
+ "epoch": 36.635782747603834,
+ "grad_norm": 0.0859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0224,
+ "step": 11467
+ },
+ {
+ "epoch": 36.638977635782744,
+ "grad_norm": 0.154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0265,
+ "step": 11468
+ },
+ {
+ "epoch": 36.64217252396166,
+ "grad_norm": 0.05029296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0403,
+ "step": 11469
+ },
+ {
+ "epoch": 36.64536741214057,
+ "grad_norm": 0.09814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0318,
+ "step": 11470
+ },
+ {
+ "epoch": 36.64856230031949,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0218,
+ "step": 11471
+ },
+ {
+ "epoch": 36.6517571884984,
+ "grad_norm": 0.107421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0298,
+ "step": 11472
+ },
+ {
+ "epoch": 36.65495207667732,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0196,
+ "step": 11473
+ },
+ {
+ "epoch": 36.65814696485623,
+ "grad_norm": 0.1357421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0352,
+ "step": 11474
+ },
+ {
+ "epoch": 36.66134185303515,
+ "grad_norm": 0.05419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0361,
+ "step": 11475
+ },
+ {
+ "epoch": 36.66453674121406,
+ "grad_norm": 0.1337890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0277,
+ "step": 11476
+ },
+ {
+ "epoch": 36.66773162939297,
+ "grad_norm": 0.05908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0229,
+ "step": 11477
+ },
+ {
+ "epoch": 36.670926517571885,
+ "grad_norm": 0.10693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0362,
+ "step": 11478
+ },
+ {
+ "epoch": 36.674121405750796,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0278,
+ "step": 11479
+ },
+ {
+ "epoch": 36.677316293929714,
+ "grad_norm": 0.08837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0463,
+ "step": 11480
+ },
+ {
+ "epoch": 36.680511182108624,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0338,
+ "step": 11481
+ },
+ {
+ "epoch": 36.68370607028754,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.033,
+ "step": 11482
+ },
+ {
+ "epoch": 36.68690095846645,
+ "grad_norm": 0.05029296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0293,
+ "step": 11483
+ },
+ {
+ "epoch": 36.69009584664537,
+ "grad_norm": 0.05712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0506,
+ "step": 11484
+ },
+ {
+ "epoch": 36.69329073482428,
+ "grad_norm": 0.062255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0367,
+ "step": 11485
+ },
+ {
+ "epoch": 36.6964856230032,
+ "grad_norm": 0.053955078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0305,
+ "step": 11486
+ },
+ {
+ "epoch": 36.69968051118211,
+ "grad_norm": 0.057373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0323,
+ "step": 11487
+ },
+ {
+ "epoch": 36.70287539936102,
+ "grad_norm": 0.05126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.027,
+ "step": 11488
+ },
+ {
+ "epoch": 36.70607028753994,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0242,
+ "step": 11489
+ },
+ {
+ "epoch": 36.70926517571885,
+ "grad_norm": 0.05078125,
+ "learning_rate": 0.0005,
+ "loss": 1.044,
+ "step": 11490
+ },
+ {
+ "epoch": 36.712460063897765,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0265,
+ "step": 11491
+ },
+ {
+ "epoch": 36.715654952076676,
+ "grad_norm": 0.0546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0412,
+ "step": 11492
+ },
+ {
+ "epoch": 36.718849840255594,
+ "grad_norm": 0.111328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0321,
+ "step": 11493
+ },
+ {
+ "epoch": 36.722044728434504,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.027,
+ "step": 11494
+ },
+ {
+ "epoch": 36.72523961661342,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0423,
+ "step": 11495
+ },
+ {
+ "epoch": 36.72843450479233,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0292,
+ "step": 11496
+ },
+ {
+ "epoch": 36.73162939297124,
+ "grad_norm": 0.0537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0284,
+ "step": 11497
+ },
+ {
+ "epoch": 36.73482428115016,
+ "grad_norm": 0.056884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.029,
+ "step": 11498
+ },
+ {
+ "epoch": 36.73801916932907,
+ "grad_norm": 0.0537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0267,
+ "step": 11499
+ },
+ {
+ "epoch": 36.74121405750799,
+ "grad_norm": 0.06005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0377,
+ "step": 11500
+ },
+ {
+ "epoch": 36.7444089456869,
+ "grad_norm": 0.0458984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0362,
+ "step": 11501
+ },
+ {
+ "epoch": 36.74760383386582,
+ "grad_norm": 0.0556640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0429,
+ "step": 11502
+ },
+ {
+ "epoch": 36.75079872204473,
+ "grad_norm": 0.058837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0414,
+ "step": 11503
+ },
+ {
+ "epoch": 36.753993610223645,
+ "grad_norm": 0.054931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0436,
+ "step": 11504
+ },
+ {
+ "epoch": 36.757188498402556,
+ "grad_norm": 0.05859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0394,
+ "step": 11505
+ },
+ {
+ "epoch": 36.760383386581466,
+ "grad_norm": 0.055908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0392,
+ "step": 11506
+ },
+ {
+ "epoch": 36.763578274760384,
+ "grad_norm": 0.0556640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0306,
+ "step": 11507
+ },
+ {
+ "epoch": 36.766773162939295,
+ "grad_norm": 0.060791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0341,
+ "step": 11508
+ },
+ {
+ "epoch": 36.76996805111821,
+ "grad_norm": 0.0576171875,
+ "learning_rate": 0.0005,
+ "loss": 1.037,
+ "step": 11509
+ },
+ {
+ "epoch": 36.77316293929712,
+ "grad_norm": 0.05322265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0452,
+ "step": 11510
+ },
+ {
+ "epoch": 36.77635782747604,
+ "grad_norm": 0.07861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0341,
+ "step": 11511
+ },
+ {
+ "epoch": 36.77955271565495,
+ "grad_norm": 0.05126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0352,
+ "step": 11512
+ },
+ {
+ "epoch": 36.78274760383387,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0313,
+ "step": 11513
+ },
+ {
+ "epoch": 36.78594249201278,
+ "grad_norm": 0.055908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0358,
+ "step": 11514
+ },
+ {
+ "epoch": 36.78913738019169,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0235,
+ "step": 11515
+ },
+ {
+ "epoch": 36.79233226837061,
+ "grad_norm": 0.08154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0461,
+ "step": 11516
+ },
+ {
+ "epoch": 36.79552715654952,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0342,
+ "step": 11517
+ },
+ {
+ "epoch": 36.798722044728436,
+ "grad_norm": 0.08740234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0311,
+ "step": 11518
+ },
+ {
+ "epoch": 36.801916932907346,
+ "grad_norm": 0.054443359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0307,
+ "step": 11519
+ },
+ {
+ "epoch": 36.805111821086264,
+ "grad_norm": 0.06005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0279,
+ "step": 11520
+ },
+ {
+ "epoch": 36.808306709265175,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0252,
+ "step": 11521
+ },
+ {
+ "epoch": 36.81150159744409,
+ "grad_norm": 0.061279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0245,
+ "step": 11522
+ },
+ {
+ "epoch": 36.814696485623,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0247,
+ "step": 11523
+ },
+ {
+ "epoch": 36.81789137380191,
+ "grad_norm": 0.095703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0243,
+ "step": 11524
+ },
+ {
+ "epoch": 36.82108626198083,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0435,
+ "step": 11525
+ },
+ {
+ "epoch": 36.82428115015974,
+ "grad_norm": 0.099609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0408,
+ "step": 11526
+ },
+ {
+ "epoch": 36.82747603833866,
+ "grad_norm": 0.05517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.034,
+ "step": 11527
+ },
+ {
+ "epoch": 36.83067092651757,
+ "grad_norm": 0.115234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0287,
+ "step": 11528
+ },
+ {
+ "epoch": 36.83386581469649,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0317,
+ "step": 11529
+ },
+ {
+ "epoch": 36.8370607028754,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0427,
+ "step": 11530
+ },
+ {
+ "epoch": 36.840255591054316,
+ "grad_norm": 0.1103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0361,
+ "step": 11531
+ },
+ {
+ "epoch": 36.843450479233226,
+ "grad_norm": 0.061767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0288,
+ "step": 11532
+ },
+ {
+ "epoch": 36.846645367412144,
+ "grad_norm": 0.083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0409,
+ "step": 11533
+ },
+ {
+ "epoch": 36.849840255591054,
+ "grad_norm": 0.0810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0425,
+ "step": 11534
+ },
+ {
+ "epoch": 36.853035143769965,
+ "grad_norm": 0.12451171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0177,
+ "step": 11535
+ },
+ {
+ "epoch": 36.85623003194888,
+ "grad_norm": 0.197265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0387,
+ "step": 11536
+ },
+ {
+ "epoch": 36.85942492012779,
+ "grad_norm": 0.09521484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0311,
+ "step": 11537
+ },
+ {
+ "epoch": 36.86261980830671,
+ "grad_norm": 0.1416015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0282,
+ "step": 11538
+ },
+ {
+ "epoch": 36.86581469648562,
+ "grad_norm": 0.1376953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0318,
+ "step": 11539
+ },
+ {
+ "epoch": 36.86900958466454,
+ "grad_norm": 0.1396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0138,
+ "step": 11540
+ },
+ {
+ "epoch": 36.87220447284345,
+ "grad_norm": 0.13671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0406,
+ "step": 11541
+ },
+ {
+ "epoch": 36.87539936102237,
+ "grad_norm": 0.087890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0263,
+ "step": 11542
+ },
+ {
+ "epoch": 36.87859424920128,
+ "grad_norm": 0.1455078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0277,
+ "step": 11543
+ },
+ {
+ "epoch": 36.88178913738019,
+ "grad_norm": 0.1240234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0401,
+ "step": 11544
+ },
+ {
+ "epoch": 36.884984025559106,
+ "grad_norm": 0.126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0384,
+ "step": 11545
+ },
+ {
+ "epoch": 36.88817891373802,
+ "grad_norm": 0.1220703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0234,
+ "step": 11546
+ },
+ {
+ "epoch": 36.891373801916934,
+ "grad_norm": 0.1328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0377,
+ "step": 11547
+ },
+ {
+ "epoch": 36.894568690095845,
+ "grad_norm": 0.16015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0382,
+ "step": 11548
+ },
+ {
+ "epoch": 36.89776357827476,
+ "grad_norm": 0.08837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.039,
+ "step": 11549
+ },
+ {
+ "epoch": 36.90095846645367,
+ "grad_norm": 0.2041015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0415,
+ "step": 11550
+ },
+ {
+ "epoch": 36.90415335463259,
+ "grad_norm": 0.061767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0408,
+ "step": 11551
+ },
+ {
+ "epoch": 36.9073482428115,
+ "grad_norm": 0.158203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0343,
+ "step": 11552
+ },
+ {
+ "epoch": 36.91054313099041,
+ "grad_norm": 0.103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0384,
+ "step": 11553
+ },
+ {
+ "epoch": 36.91373801916933,
+ "grad_norm": 0.12890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0349,
+ "step": 11554
+ },
+ {
+ "epoch": 36.91693290734824,
+ "grad_norm": 0.1376953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0295,
+ "step": 11555
+ },
+ {
+ "epoch": 36.92012779552716,
+ "grad_norm": 0.1357421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0257,
+ "step": 11556
+ },
+ {
+ "epoch": 36.92332268370607,
+ "grad_norm": 0.2080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0409,
+ "step": 11557
+ },
+ {
+ "epoch": 36.926517571884986,
+ "grad_norm": 0.1005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0361,
+ "step": 11558
+ },
+ {
+ "epoch": 36.9297124600639,
+ "grad_norm": 0.2001953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0269,
+ "step": 11559
+ },
+ {
+ "epoch": 36.932907348242814,
+ "grad_norm": 0.099609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0316,
+ "step": 11560
+ },
+ {
+ "epoch": 36.936102236421725,
+ "grad_norm": 0.0966796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0305,
+ "step": 11561
+ },
+ {
+ "epoch": 36.93929712460064,
+ "grad_norm": 0.09423828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0357,
+ "step": 11562
+ },
+ {
+ "epoch": 36.94249201277955,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0412,
+ "step": 11563
+ },
+ {
+ "epoch": 36.945686900958464,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0269,
+ "step": 11564
+ },
+ {
+ "epoch": 36.94888178913738,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0273,
+ "step": 11565
+ },
+ {
+ "epoch": 36.95207667731629,
+ "grad_norm": 0.0908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0372,
+ "step": 11566
+ },
+ {
+ "epoch": 36.95527156549521,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0426,
+ "step": 11567
+ },
+ {
+ "epoch": 36.95846645367412,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0343,
+ "step": 11568
+ },
+ {
+ "epoch": 36.96166134185304,
+ "grad_norm": 0.0986328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0384,
+ "step": 11569
+ },
+ {
+ "epoch": 36.96485623003195,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0451,
+ "step": 11570
+ },
+ {
+ "epoch": 36.968051118210866,
+ "grad_norm": 0.05810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0302,
+ "step": 11571
+ },
+ {
+ "epoch": 36.97124600638978,
+ "grad_norm": 0.055419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0434,
+ "step": 11572
+ },
+ {
+ "epoch": 36.97444089456869,
+ "grad_norm": 0.049560546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0314,
+ "step": 11573
+ },
+ {
+ "epoch": 36.977635782747605,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0341,
+ "step": 11574
+ },
+ {
+ "epoch": 36.980830670926515,
+ "grad_norm": 0.05126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0472,
+ "step": 11575
+ },
+ {
+ "epoch": 36.98402555910543,
+ "grad_norm": 0.05224609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0354,
+ "step": 11576
+ },
+ {
+ "epoch": 36.98722044728434,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0378,
+ "step": 11577
+ },
+ {
+ "epoch": 36.99041533546326,
+ "grad_norm": 0.062255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0312,
+ "step": 11578
+ },
+ {
+ "epoch": 36.99361022364217,
+ "grad_norm": 0.09326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0369,
+ "step": 11579
+ },
+ {
+ "epoch": 36.99680511182109,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0464,
+ "step": 11580
+ },
+ {
+ "epoch": 37.0,
+ "grad_norm": 0.080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0263,
+ "step": 11581
+ },
+ {
+ "epoch": 37.00319488817891,
+ "grad_norm": 0.058349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0202,
+ "step": 11582
+ },
+ {
+ "epoch": 37.00638977635783,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0232,
+ "step": 11583
+ },
+ {
+ "epoch": 37.00958466453674,
+ "grad_norm": 0.05419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.025,
+ "step": 11584
+ },
+ {
+ "epoch": 37.01277955271566,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0417,
+ "step": 11585
+ },
+ {
+ "epoch": 37.01597444089457,
+ "grad_norm": 0.05517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0341,
+ "step": 11586
+ },
+ {
+ "epoch": 37.019169329073485,
+ "grad_norm": 0.111328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0419,
+ "step": 11587
+ },
+ {
+ "epoch": 37.022364217252395,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0258,
+ "step": 11588
+ },
+ {
+ "epoch": 37.02555910543131,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0395,
+ "step": 11589
+ },
+ {
+ "epoch": 37.02875399361022,
+ "grad_norm": 0.056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0472,
+ "step": 11590
+ },
+ {
+ "epoch": 37.031948881789134,
+ "grad_norm": 0.056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0256,
+ "step": 11591
+ },
+ {
+ "epoch": 37.03514376996805,
+ "grad_norm": 0.05224609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0222,
+ "step": 11592
+ },
+ {
+ "epoch": 37.03833865814696,
+ "grad_norm": 0.0556640625,
+ "learning_rate": 0.0005,
+ "loss": 1.029,
+ "step": 11593
+ },
+ {
+ "epoch": 37.04153354632588,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0383,
+ "step": 11594
+ },
+ {
+ "epoch": 37.04472843450479,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.032,
+ "step": 11595
+ },
+ {
+ "epoch": 37.04792332268371,
+ "grad_norm": 0.053466796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0318,
+ "step": 11596
+ },
+ {
+ "epoch": 37.05111821086262,
+ "grad_norm": 0.05078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0324,
+ "step": 11597
+ },
+ {
+ "epoch": 37.054313099041536,
+ "grad_norm": 0.0849609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0264,
+ "step": 11598
+ },
+ {
+ "epoch": 37.05750798722045,
+ "grad_norm": 0.05224609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0357,
+ "step": 11599
+ },
+ {
+ "epoch": 37.06070287539936,
+ "grad_norm": 0.05908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0304,
+ "step": 11600
+ },
+ {
+ "epoch": 37.063897763578275,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0343,
+ "step": 11601
+ },
+ {
+ "epoch": 37.067092651757186,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0428,
+ "step": 11602
+ },
+ {
+ "epoch": 37.0702875399361,
+ "grad_norm": 0.061767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0436,
+ "step": 11603
+ },
+ {
+ "epoch": 37.073482428115014,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0372,
+ "step": 11604
+ },
+ {
+ "epoch": 37.07667731629393,
+ "grad_norm": 0.047607421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0356,
+ "step": 11605
+ },
+ {
+ "epoch": 37.07987220447284,
+ "grad_norm": 0.07861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.031,
+ "step": 11606
+ },
+ {
+ "epoch": 37.08306709265176,
+ "grad_norm": 0.08740234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0353,
+ "step": 11607
+ },
+ {
+ "epoch": 37.08626198083067,
+ "grad_norm": 0.09033203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0251,
+ "step": 11608
+ },
+ {
+ "epoch": 37.08945686900959,
+ "grad_norm": 0.150390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0345,
+ "step": 11609
+ },
+ {
+ "epoch": 37.0926517571885,
+ "grad_norm": 0.11962890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0359,
+ "step": 11610
+ },
+ {
+ "epoch": 37.09584664536741,
+ "grad_norm": 0.1513671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0294,
+ "step": 11611
+ },
+ {
+ "epoch": 37.09904153354633,
+ "grad_norm": 0.11328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0323,
+ "step": 11612
+ },
+ {
+ "epoch": 37.10223642172524,
+ "grad_norm": 0.1181640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0382,
+ "step": 11613
+ },
+ {
+ "epoch": 37.105431309904155,
+ "grad_norm": 0.095703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0497,
+ "step": 11614
+ },
+ {
+ "epoch": 37.108626198083066,
+ "grad_norm": 0.1259765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0418,
+ "step": 11615
+ },
+ {
+ "epoch": 37.11182108626198,
+ "grad_norm": 0.05517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0402,
+ "step": 11616
+ },
+ {
+ "epoch": 37.115015974440894,
+ "grad_norm": 0.16015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0324,
+ "step": 11617
+ },
+ {
+ "epoch": 37.11821086261981,
+ "grad_norm": 0.0966796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0253,
+ "step": 11618
+ },
+ {
+ "epoch": 37.12140575079872,
+ "grad_norm": 0.1240234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0353,
+ "step": 11619
+ },
+ {
+ "epoch": 37.12460063897763,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0364,
+ "step": 11620
+ },
+ {
+ "epoch": 37.12779552715655,
+ "grad_norm": 0.14453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0299,
+ "step": 11621
+ },
+ {
+ "epoch": 37.13099041533546,
+ "grad_norm": 0.10888671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0423,
+ "step": 11622
+ },
+ {
+ "epoch": 37.13418530351438,
+ "grad_norm": 0.0966796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0263,
+ "step": 11623
+ },
+ {
+ "epoch": 37.13738019169329,
+ "grad_norm": 0.0849609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0467,
+ "step": 11624
+ },
+ {
+ "epoch": 37.14057507987221,
+ "grad_norm": 0.1025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0483,
+ "step": 11625
+ },
+ {
+ "epoch": 37.14376996805112,
+ "grad_norm": 0.059326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0334,
+ "step": 11626
+ },
+ {
+ "epoch": 37.146964856230035,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0304,
+ "step": 11627
+ },
+ {
+ "epoch": 37.150159744408946,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0337,
+ "step": 11628
+ },
+ {
+ "epoch": 37.153354632587856,
+ "grad_norm": 0.060302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0288,
+ "step": 11629
+ },
+ {
+ "epoch": 37.156549520766774,
+ "grad_norm": 0.08544921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0307,
+ "step": 11630
+ },
+ {
+ "epoch": 37.159744408945684,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0339,
+ "step": 11631
+ },
+ {
+ "epoch": 37.1629392971246,
+ "grad_norm": 0.05615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0528,
+ "step": 11632
+ },
+ {
+ "epoch": 37.16613418530351,
+ "grad_norm": 0.050048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0489,
+ "step": 11633
+ },
+ {
+ "epoch": 37.16932907348243,
+ "grad_norm": 0.05810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0305,
+ "step": 11634
+ },
+ {
+ "epoch": 37.17252396166134,
+ "grad_norm": 0.057861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0398,
+ "step": 11635
+ },
+ {
+ "epoch": 37.17571884984026,
+ "grad_norm": 0.05615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0331,
+ "step": 11636
+ },
+ {
+ "epoch": 37.17891373801917,
+ "grad_norm": 0.05615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0415,
+ "step": 11637
+ },
+ {
+ "epoch": 37.18210862619808,
+ "grad_norm": 0.053466796875,
+ "learning_rate": 0.0005,
+ "loss": 1.033,
+ "step": 11638
+ },
+ {
+ "epoch": 37.185303514377,
+ "grad_norm": 0.055419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0357,
+ "step": 11639
+ },
+ {
+ "epoch": 37.18849840255591,
+ "grad_norm": 0.061279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0482,
+ "step": 11640
+ },
+ {
+ "epoch": 37.191693290734825,
+ "grad_norm": 0.052734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0313,
+ "step": 11641
+ },
+ {
+ "epoch": 37.194888178913736,
+ "grad_norm": 0.05322265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0313,
+ "step": 11642
+ },
+ {
+ "epoch": 37.198083067092654,
+ "grad_norm": 0.0927734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0307,
+ "step": 11643
+ },
+ {
+ "epoch": 37.201277955271564,
+ "grad_norm": 0.09326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0348,
+ "step": 11644
+ },
+ {
+ "epoch": 37.20447284345048,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0347,
+ "step": 11645
+ },
+ {
+ "epoch": 37.20766773162939,
+ "grad_norm": 0.05078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0324,
+ "step": 11646
+ },
+ {
+ "epoch": 37.21086261980831,
+ "grad_norm": 0.09228515625,
+ "learning_rate": 0.0005,
+ "loss": 1.036,
+ "step": 11647
+ },
+ {
+ "epoch": 37.21405750798722,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0257,
+ "step": 11648
+ },
+ {
+ "epoch": 37.21725239616613,
+ "grad_norm": 0.09375,
+ "learning_rate": 0.0005,
+ "loss": 1.0221,
+ "step": 11649
+ },
+ {
+ "epoch": 37.22044728434505,
+ "grad_norm": 0.10595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0276,
+ "step": 11650
+ },
+ {
+ "epoch": 37.22364217252396,
+ "grad_norm": 0.1396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0292,
+ "step": 11651
+ },
+ {
+ "epoch": 37.22683706070288,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0392,
+ "step": 11652
+ },
+ {
+ "epoch": 37.23003194888179,
+ "grad_norm": 0.047607421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0328,
+ "step": 11653
+ },
+ {
+ "epoch": 37.233226837060705,
+ "grad_norm": 0.10400390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0303,
+ "step": 11654
+ },
+ {
+ "epoch": 37.236421725239616,
+ "grad_norm": 0.053955078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0154,
+ "step": 11655
+ },
+ {
+ "epoch": 37.239616613418534,
+ "grad_norm": 0.0869140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0308,
+ "step": 11656
+ },
+ {
+ "epoch": 37.242811501597444,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0335,
+ "step": 11657
+ },
+ {
+ "epoch": 37.246006389776355,
+ "grad_norm": 0.04833984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0423,
+ "step": 11658
+ },
+ {
+ "epoch": 37.24920127795527,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0359,
+ "step": 11659
+ },
+ {
+ "epoch": 37.25239616613418,
+ "grad_norm": 0.052734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0357,
+ "step": 11660
+ },
+ {
+ "epoch": 37.2555910543131,
+ "grad_norm": 0.048583984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0516,
+ "step": 11661
+ },
+ {
+ "epoch": 37.25878594249201,
+ "grad_norm": 0.057373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0254,
+ "step": 11662
+ },
+ {
+ "epoch": 37.26198083067093,
+ "grad_norm": 0.08251953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0377,
+ "step": 11663
+ },
+ {
+ "epoch": 37.26517571884984,
+ "grad_norm": 0.048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0248,
+ "step": 11664
+ },
+ {
+ "epoch": 37.26837060702876,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0246,
+ "step": 11665
+ },
+ {
+ "epoch": 37.27156549520767,
+ "grad_norm": 0.05419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0348,
+ "step": 11666
+ },
+ {
+ "epoch": 37.27476038338658,
+ "grad_norm": 0.05322265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0275,
+ "step": 11667
+ },
+ {
+ "epoch": 37.277955271565496,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0441,
+ "step": 11668
+ },
+ {
+ "epoch": 37.281150159744406,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0353,
+ "step": 11669
+ },
+ {
+ "epoch": 37.284345047923324,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0272,
+ "step": 11670
+ },
+ {
+ "epoch": 37.287539936102235,
+ "grad_norm": 0.051513671875,
+ "learning_rate": 0.0005,
+ "loss": 1.038,
+ "step": 11671
+ },
+ {
+ "epoch": 37.29073482428115,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0352,
+ "step": 11672
+ },
+ {
+ "epoch": 37.29392971246006,
+ "grad_norm": 0.053955078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0257,
+ "step": 11673
+ },
+ {
+ "epoch": 37.29712460063898,
+ "grad_norm": 0.05322265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0302,
+ "step": 11674
+ },
+ {
+ "epoch": 37.30031948881789,
+ "grad_norm": 0.048095703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0471,
+ "step": 11675
+ },
+ {
+ "epoch": 37.3035143769968,
+ "grad_norm": 0.0576171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0342,
+ "step": 11676
+ },
+ {
+ "epoch": 37.30670926517572,
+ "grad_norm": 0.0458984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0334,
+ "step": 11677
+ },
+ {
+ "epoch": 37.30990415335463,
+ "grad_norm": 0.08154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0302,
+ "step": 11678
+ },
+ {
+ "epoch": 37.31309904153355,
+ "grad_norm": 0.0498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0363,
+ "step": 11679
+ },
+ {
+ "epoch": 37.31629392971246,
+ "grad_norm": 0.040771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0317,
+ "step": 11680
+ },
+ {
+ "epoch": 37.319488817891376,
+ "grad_norm": 0.056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0307,
+ "step": 11681
+ },
+ {
+ "epoch": 37.322683706070286,
+ "grad_norm": 0.046630859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0412,
+ "step": 11682
+ },
+ {
+ "epoch": 37.325878594249204,
+ "grad_norm": 0.048095703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0259,
+ "step": 11683
+ },
+ {
+ "epoch": 37.329073482428115,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0394,
+ "step": 11684
+ },
+ {
+ "epoch": 37.33226837060703,
+ "grad_norm": 0.0517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0296,
+ "step": 11685
+ },
+ {
+ "epoch": 37.33546325878594,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0311,
+ "step": 11686
+ },
+ {
+ "epoch": 37.33865814696485,
+ "grad_norm": 0.06103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0156,
+ "step": 11687
+ },
+ {
+ "epoch": 37.34185303514377,
+ "grad_norm": 0.06005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.036,
+ "step": 11688
+ },
+ {
+ "epoch": 37.34504792332268,
+ "grad_norm": 0.0966796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0337,
+ "step": 11689
+ },
+ {
+ "epoch": 37.3482428115016,
+ "grad_norm": 0.05419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0379,
+ "step": 11690
+ },
+ {
+ "epoch": 37.35143769968051,
+ "grad_norm": 0.05419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0372,
+ "step": 11691
+ },
+ {
+ "epoch": 37.35463258785943,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0335,
+ "step": 11692
+ },
+ {
+ "epoch": 37.35782747603834,
+ "grad_norm": 0.047119140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0315,
+ "step": 11693
+ },
+ {
+ "epoch": 37.361022364217256,
+ "grad_norm": 0.054931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0356,
+ "step": 11694
+ },
+ {
+ "epoch": 37.364217252396166,
+ "grad_norm": 0.060791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0237,
+ "step": 11695
+ },
+ {
+ "epoch": 37.36741214057508,
+ "grad_norm": 0.05810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0393,
+ "step": 11696
+ },
+ {
+ "epoch": 37.370607028753994,
+ "grad_norm": 0.052001953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0334,
+ "step": 11697
+ },
+ {
+ "epoch": 37.373801916932905,
+ "grad_norm": 0.052490234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0334,
+ "step": 11698
+ },
+ {
+ "epoch": 37.37699680511182,
+ "grad_norm": 0.05517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0438,
+ "step": 11699
+ },
+ {
+ "epoch": 37.38019169329073,
+ "grad_norm": 0.051025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0263,
+ "step": 11700
+ },
+ {
+ "epoch": 37.38338658146965,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0414,
+ "step": 11701
+ },
+ {
+ "epoch": 37.38658146964856,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.043,
+ "step": 11702
+ },
+ {
+ "epoch": 37.38977635782748,
+ "grad_norm": 0.048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0347,
+ "step": 11703
+ },
+ {
+ "epoch": 37.39297124600639,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0314,
+ "step": 11704
+ },
+ {
+ "epoch": 37.3961661341853,
+ "grad_norm": 0.1025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0364,
+ "step": 11705
+ },
+ {
+ "epoch": 37.39936102236422,
+ "grad_norm": 0.048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0223,
+ "step": 11706
+ },
+ {
+ "epoch": 37.40255591054313,
+ "grad_norm": 0.060302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0359,
+ "step": 11707
+ },
+ {
+ "epoch": 37.405750798722046,
+ "grad_norm": 0.052734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0198,
+ "step": 11708
+ },
+ {
+ "epoch": 37.40894568690096,
+ "grad_norm": 0.061279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0371,
+ "step": 11709
+ },
+ {
+ "epoch": 37.412140575079874,
+ "grad_norm": 0.056884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0431,
+ "step": 11710
+ },
+ {
+ "epoch": 37.415335463258785,
+ "grad_norm": 0.053955078125,
+ "learning_rate": 0.0005,
+ "loss": 1.031,
+ "step": 11711
+ },
+ {
+ "epoch": 37.4185303514377,
+ "grad_norm": 0.059814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.029,
+ "step": 11712
+ },
+ {
+ "epoch": 37.42172523961661,
+ "grad_norm": 0.046142578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0332,
+ "step": 11713
+ },
+ {
+ "epoch": 37.424920127795524,
+ "grad_norm": 0.051513671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0394,
+ "step": 11714
+ },
+ {
+ "epoch": 37.42811501597444,
+ "grad_norm": 0.0537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0336,
+ "step": 11715
+ },
+ {
+ "epoch": 37.43130990415335,
+ "grad_norm": 0.04931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0353,
+ "step": 11716
+ },
+ {
+ "epoch": 37.43450479233227,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0308,
+ "step": 11717
+ },
+ {
+ "epoch": 37.43769968051118,
+ "grad_norm": 0.05712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0438,
+ "step": 11718
+ },
+ {
+ "epoch": 37.4408945686901,
+ "grad_norm": 0.057373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0362,
+ "step": 11719
+ },
+ {
+ "epoch": 37.44408945686901,
+ "grad_norm": 0.0859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0394,
+ "step": 11720
+ },
+ {
+ "epoch": 37.447284345047926,
+ "grad_norm": 0.051513671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0347,
+ "step": 11721
+ },
+ {
+ "epoch": 37.45047923322684,
+ "grad_norm": 0.095703125,
+ "learning_rate": 0.0005,
+ "loss": 1.032,
+ "step": 11722
+ },
+ {
+ "epoch": 37.453674121405754,
+ "grad_norm": 0.058837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0433,
+ "step": 11723
+ },
+ {
+ "epoch": 37.456869009584665,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0342,
+ "step": 11724
+ },
+ {
+ "epoch": 37.460063897763575,
+ "grad_norm": 0.050048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0327,
+ "step": 11725
+ },
+ {
+ "epoch": 37.46325878594249,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0452,
+ "step": 11726
+ },
+ {
+ "epoch": 37.466453674121404,
+ "grad_norm": 0.04443359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0299,
+ "step": 11727
+ },
+ {
+ "epoch": 37.46964856230032,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0197,
+ "step": 11728
+ },
+ {
+ "epoch": 37.47284345047923,
+ "grad_norm": 0.055419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0354,
+ "step": 11729
+ },
+ {
+ "epoch": 37.47603833865815,
+ "grad_norm": 0.0546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0275,
+ "step": 11730
+ },
+ {
+ "epoch": 37.47923322683706,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0321,
+ "step": 11731
+ },
+ {
+ "epoch": 37.48242811501598,
+ "grad_norm": 0.0927734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0406,
+ "step": 11732
+ },
+ {
+ "epoch": 37.48562300319489,
+ "grad_norm": 0.09619140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0388,
+ "step": 11733
+ },
+ {
+ "epoch": 37.4888178913738,
+ "grad_norm": 0.055419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0258,
+ "step": 11734
+ },
+ {
+ "epoch": 37.49201277955272,
+ "grad_norm": 0.0986328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0327,
+ "step": 11735
+ },
+ {
+ "epoch": 37.49520766773163,
+ "grad_norm": 0.08056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.032,
+ "step": 11736
+ },
+ {
+ "epoch": 37.498402555910545,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0411,
+ "step": 11737
+ },
+ {
+ "epoch": 37.501597444089455,
+ "grad_norm": 0.10595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0309,
+ "step": 11738
+ },
+ {
+ "epoch": 37.50479233226837,
+ "grad_norm": 0.061767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0328,
+ "step": 11739
+ },
+ {
+ "epoch": 37.50798722044728,
+ "grad_norm": 0.056396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0353,
+ "step": 11740
+ },
+ {
+ "epoch": 37.5111821086262,
+ "grad_norm": 0.08837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0345,
+ "step": 11741
+ },
+ {
+ "epoch": 37.51437699680511,
+ "grad_norm": 0.0859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0515,
+ "step": 11742
+ },
+ {
+ "epoch": 37.51757188498402,
+ "grad_norm": 0.06005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0306,
+ "step": 11743
+ },
+ {
+ "epoch": 37.52076677316294,
+ "grad_norm": 0.10205078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0385,
+ "step": 11744
+ },
+ {
+ "epoch": 37.52396166134185,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0332,
+ "step": 11745
+ },
+ {
+ "epoch": 37.52715654952077,
+ "grad_norm": 0.08056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0328,
+ "step": 11746
+ },
+ {
+ "epoch": 37.53035143769968,
+ "grad_norm": 0.0849609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0331,
+ "step": 11747
+ },
+ {
+ "epoch": 37.533546325878596,
+ "grad_norm": 0.08349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0298,
+ "step": 11748
+ },
+ {
+ "epoch": 37.53674121405751,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0395,
+ "step": 11749
+ },
+ {
+ "epoch": 37.539936102236425,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0401,
+ "step": 11750
+ },
+ {
+ "epoch": 37.543130990415335,
+ "grad_norm": 0.0595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0369,
+ "step": 11751
+ },
+ {
+ "epoch": 37.546325878594246,
+ "grad_norm": 0.05224609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0335,
+ "step": 11752
+ },
+ {
+ "epoch": 37.54952076677316,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.041,
+ "step": 11753
+ },
+ {
+ "epoch": 37.552715654952074,
+ "grad_norm": 0.05810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0382,
+ "step": 11754
+ },
+ {
+ "epoch": 37.55591054313099,
+ "grad_norm": 0.062255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0326,
+ "step": 11755
+ },
+ {
+ "epoch": 37.5591054313099,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.039,
+ "step": 11756
+ },
+ {
+ "epoch": 37.56230031948882,
+ "grad_norm": 0.08056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0292,
+ "step": 11757
+ },
+ {
+ "epoch": 37.56549520766773,
+ "grad_norm": 0.130859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0412,
+ "step": 11758
+ },
+ {
+ "epoch": 37.56869009584665,
+ "grad_norm": 0.09716796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0311,
+ "step": 11759
+ },
+ {
+ "epoch": 37.57188498402556,
+ "grad_norm": 0.173828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0317,
+ "step": 11760
+ },
+ {
+ "epoch": 37.575079872204476,
+ "grad_norm": 0.08349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0377,
+ "step": 11761
+ },
+ {
+ "epoch": 37.57827476038339,
+ "grad_norm": 0.154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0344,
+ "step": 11762
+ },
+ {
+ "epoch": 37.5814696485623,
+ "grad_norm": 0.2001953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0376,
+ "step": 11763
+ },
+ {
+ "epoch": 37.584664536741215,
+ "grad_norm": 0.140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0309,
+ "step": 11764
+ },
+ {
+ "epoch": 37.587859424920126,
+ "grad_norm": 0.302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0355,
+ "step": 11765
+ },
+ {
+ "epoch": 37.59105431309904,
+ "grad_norm": 0.10888671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0409,
+ "step": 11766
+ },
+ {
+ "epoch": 37.594249201277954,
+ "grad_norm": 0.216796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0338,
+ "step": 11767
+ },
+ {
+ "epoch": 37.59744408945687,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0322,
+ "step": 11768
+ },
+ {
+ "epoch": 37.60063897763578,
+ "grad_norm": 0.134765625,
+ "learning_rate": 0.0005,
+ "loss": 1.034,
+ "step": 11769
+ },
+ {
+ "epoch": 37.6038338658147,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0299,
+ "step": 11770
+ },
+ {
+ "epoch": 37.60702875399361,
+ "grad_norm": 0.1455078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0284,
+ "step": 11771
+ },
+ {
+ "epoch": 37.61022364217252,
+ "grad_norm": 0.07861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0353,
+ "step": 11772
+ },
+ {
+ "epoch": 37.61341853035144,
+ "grad_norm": 0.09375,
+ "learning_rate": 0.0005,
+ "loss": 1.0423,
+ "step": 11773
+ },
+ {
+ "epoch": 37.61661341853035,
+ "grad_norm": 0.10400390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0409,
+ "step": 11774
+ },
+ {
+ "epoch": 37.61980830670927,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.046,
+ "step": 11775
+ },
+ {
+ "epoch": 37.62300319488818,
+ "grad_norm": 0.10107421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0321,
+ "step": 11776
+ },
+ {
+ "epoch": 37.626198083067095,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0315,
+ "step": 11777
+ },
+ {
+ "epoch": 37.629392971246006,
+ "grad_norm": 0.0498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0405,
+ "step": 11778
+ },
+ {
+ "epoch": 37.63258785942492,
+ "grad_norm": 0.08349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0287,
+ "step": 11779
+ },
+ {
+ "epoch": 37.635782747603834,
+ "grad_norm": 0.1025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0395,
+ "step": 11780
+ },
+ {
+ "epoch": 37.638977635782744,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0438,
+ "step": 11781
+ },
+ {
+ "epoch": 37.64217252396166,
+ "grad_norm": 0.08251953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0362,
+ "step": 11782
+ },
+ {
+ "epoch": 37.64536741214057,
+ "grad_norm": 0.125,
+ "learning_rate": 0.0005,
+ "loss": 1.0362,
+ "step": 11783
+ },
+ {
+ "epoch": 37.64856230031949,
+ "grad_norm": 0.08349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.045,
+ "step": 11784
+ },
+ {
+ "epoch": 37.6517571884984,
+ "grad_norm": 0.08935546875,
+ "learning_rate": 0.0005,
+ "loss": 1.032,
+ "step": 11785
+ },
+ {
+ "epoch": 37.65495207667732,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0352,
+ "step": 11786
+ },
+ {
+ "epoch": 37.65814696485623,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0248,
+ "step": 11787
+ },
+ {
+ "epoch": 37.66134185303515,
+ "grad_norm": 0.125,
+ "learning_rate": 0.0005,
+ "loss": 1.0384,
+ "step": 11788
+ },
+ {
+ "epoch": 37.66453674121406,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0423,
+ "step": 11789
+ },
+ {
+ "epoch": 37.66773162939297,
+ "grad_norm": 0.1181640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0196,
+ "step": 11790
+ },
+ {
+ "epoch": 37.670926517571885,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0452,
+ "step": 11791
+ },
+ {
+ "epoch": 37.674121405750796,
+ "grad_norm": 0.05810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0405,
+ "step": 11792
+ },
+ {
+ "epoch": 37.677316293929714,
+ "grad_norm": 0.1240234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0181,
+ "step": 11793
+ },
+ {
+ "epoch": 37.680511182108624,
+ "grad_norm": 0.054443359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0317,
+ "step": 11794
+ },
+ {
+ "epoch": 37.68370607028754,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0438,
+ "step": 11795
+ },
+ {
+ "epoch": 37.68690095846645,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0372,
+ "step": 11796
+ },
+ {
+ "epoch": 37.69009584664537,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0364,
+ "step": 11797
+ },
+ {
+ "epoch": 37.69329073482428,
+ "grad_norm": 0.051025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0302,
+ "step": 11798
+ },
+ {
+ "epoch": 37.6964856230032,
+ "grad_norm": 0.052001953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0377,
+ "step": 11799
+ },
+ {
+ "epoch": 37.69968051118211,
+ "grad_norm": 0.0546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0275,
+ "step": 11800
+ },
+ {
+ "epoch": 37.70287539936102,
+ "grad_norm": 0.06201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0335,
+ "step": 11801
+ },
+ {
+ "epoch": 37.70607028753994,
+ "grad_norm": 0.056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0393,
+ "step": 11802
+ },
+ {
+ "epoch": 37.70926517571885,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0376,
+ "step": 11803
+ },
+ {
+ "epoch": 37.712460063897765,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0394,
+ "step": 11804
+ },
+ {
+ "epoch": 37.715654952076676,
+ "grad_norm": 0.05029296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0409,
+ "step": 11805
+ },
+ {
+ "epoch": 37.718849840255594,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0377,
+ "step": 11806
+ },
+ {
+ "epoch": 37.722044728434504,
+ "grad_norm": 0.05419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0406,
+ "step": 11807
+ },
+ {
+ "epoch": 37.72523961661342,
+ "grad_norm": 0.059326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0329,
+ "step": 11808
+ },
+ {
+ "epoch": 37.72843450479233,
+ "grad_norm": 0.048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0363,
+ "step": 11809
+ },
+ {
+ "epoch": 37.73162939297124,
+ "grad_norm": 0.08984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0398,
+ "step": 11810
+ },
+ {
+ "epoch": 37.73482428115016,
+ "grad_norm": 0.057373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0265,
+ "step": 11811
+ },
+ {
+ "epoch": 37.73801916932907,
+ "grad_norm": 0.0810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0387,
+ "step": 11812
+ },
+ {
+ "epoch": 37.74121405750799,
+ "grad_norm": 0.09033203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0358,
+ "step": 11813
+ },
+ {
+ "epoch": 37.7444089456869,
+ "grad_norm": 0.053466796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0337,
+ "step": 11814
+ },
+ {
+ "epoch": 37.74760383386582,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0241,
+ "step": 11815
+ },
+ {
+ "epoch": 37.75079872204473,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.031,
+ "step": 11816
+ },
+ {
+ "epoch": 37.753993610223645,
+ "grad_norm": 0.10546875,
+ "learning_rate": 0.0005,
+ "loss": 1.031,
+ "step": 11817
+ },
+ {
+ "epoch": 37.757188498402556,
+ "grad_norm": 0.0859375,
+ "learning_rate": 0.0005,
+ "loss": 1.026,
+ "step": 11818
+ },
+ {
+ "epoch": 37.760383386581466,
+ "grad_norm": 0.1005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0302,
+ "step": 11819
+ },
+ {
+ "epoch": 37.763578274760384,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0306,
+ "step": 11820
+ },
+ {
+ "epoch": 37.766773162939295,
+ "grad_norm": 0.10302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0393,
+ "step": 11821
+ },
+ {
+ "epoch": 37.76996805111821,
+ "grad_norm": 0.056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0427,
+ "step": 11822
+ },
+ {
+ "epoch": 37.77316293929712,
+ "grad_norm": 0.1123046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0364,
+ "step": 11823
+ },
+ {
+ "epoch": 37.77635782747604,
+ "grad_norm": 0.080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0438,
+ "step": 11824
+ },
+ {
+ "epoch": 37.77955271565495,
+ "grad_norm": 0.1005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0378,
+ "step": 11825
+ },
+ {
+ "epoch": 37.78274760383387,
+ "grad_norm": 0.11181640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0299,
+ "step": 11826
+ },
+ {
+ "epoch": 37.78594249201278,
+ "grad_norm": 0.0986328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0294,
+ "step": 11827
+ },
+ {
+ "epoch": 37.78913738019169,
+ "grad_norm": 0.11767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0303,
+ "step": 11828
+ },
+ {
+ "epoch": 37.79233226837061,
+ "grad_norm": 0.107421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0221,
+ "step": 11829
+ },
+ {
+ "epoch": 37.79552715654952,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0323,
+ "step": 11830
+ },
+ {
+ "epoch": 37.798722044728436,
+ "grad_norm": 0.08544921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0327,
+ "step": 11831
+ },
+ {
+ "epoch": 37.801916932907346,
+ "grad_norm": 0.055419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0447,
+ "step": 11832
+ },
+ {
+ "epoch": 37.805111821086264,
+ "grad_norm": 0.087890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0289,
+ "step": 11833
+ },
+ {
+ "epoch": 37.808306709265175,
+ "grad_norm": 0.05615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0359,
+ "step": 11834
+ },
+ {
+ "epoch": 37.81150159744409,
+ "grad_norm": 0.058349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0404,
+ "step": 11835
+ },
+ {
+ "epoch": 37.814696485623,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0418,
+ "step": 11836
+ },
+ {
+ "epoch": 37.81789137380191,
+ "grad_norm": 0.05810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0423,
+ "step": 11837
+ },
+ {
+ "epoch": 37.82108626198083,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0301,
+ "step": 11838
+ },
+ {
+ "epoch": 37.82428115015974,
+ "grad_norm": 0.060302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0486,
+ "step": 11839
+ },
+ {
+ "epoch": 37.82747603833866,
+ "grad_norm": 0.08984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0343,
+ "step": 11840
+ },
+ {
+ "epoch": 37.83067092651757,
+ "grad_norm": 0.05712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0323,
+ "step": 11841
+ },
+ {
+ "epoch": 37.83386581469649,
+ "grad_norm": 0.0830078125,
+ "learning_rate": 0.0005,
+ "loss": 1.026,
+ "step": 11842
+ },
+ {
+ "epoch": 37.8370607028754,
+ "grad_norm": 0.044677734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0411,
+ "step": 11843
+ },
+ {
+ "epoch": 37.840255591054316,
+ "grad_norm": 0.059326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0318,
+ "step": 11844
+ },
+ {
+ "epoch": 37.843450479233226,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0402,
+ "step": 11845
+ },
+ {
+ "epoch": 37.846645367412144,
+ "grad_norm": 0.0595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0292,
+ "step": 11846
+ },
+ {
+ "epoch": 37.849840255591054,
+ "grad_norm": 0.060791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0343,
+ "step": 11847
+ },
+ {
+ "epoch": 37.853035143769965,
+ "grad_norm": 0.09228515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0244,
+ "step": 11848
+ },
+ {
+ "epoch": 37.85623003194888,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0259,
+ "step": 11849
+ },
+ {
+ "epoch": 37.85942492012779,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0384,
+ "step": 11850
+ },
+ {
+ "epoch": 37.86261980830671,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0326,
+ "step": 11851
+ },
+ {
+ "epoch": 37.86581469648562,
+ "grad_norm": 0.046630859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0359,
+ "step": 11852
+ },
+ {
+ "epoch": 37.86900958466454,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0307,
+ "step": 11853
+ },
+ {
+ "epoch": 37.87220447284345,
+ "grad_norm": 0.055908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0357,
+ "step": 11854
+ },
+ {
+ "epoch": 37.87539936102237,
+ "grad_norm": 0.05224609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0409,
+ "step": 11855
+ },
+ {
+ "epoch": 37.87859424920128,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0265,
+ "step": 11856
+ },
+ {
+ "epoch": 37.88178913738019,
+ "grad_norm": 0.053955078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0207,
+ "step": 11857
+ },
+ {
+ "epoch": 37.884984025559106,
+ "grad_norm": 0.0498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0399,
+ "step": 11858
+ },
+ {
+ "epoch": 37.88817891373802,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0311,
+ "step": 11859
+ },
+ {
+ "epoch": 37.891373801916934,
+ "grad_norm": 0.060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0295,
+ "step": 11860
+ },
+ {
+ "epoch": 37.894568690095845,
+ "grad_norm": 0.08837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0198,
+ "step": 11861
+ },
+ {
+ "epoch": 37.89776357827476,
+ "grad_norm": 0.05908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0404,
+ "step": 11862
+ },
+ {
+ "epoch": 37.90095846645367,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.027,
+ "step": 11863
+ },
+ {
+ "epoch": 37.90415335463259,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0422,
+ "step": 11864
+ },
+ {
+ "epoch": 37.9073482428115,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0145,
+ "step": 11865
+ },
+ {
+ "epoch": 37.91054313099041,
+ "grad_norm": 0.060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0358,
+ "step": 11866
+ },
+ {
+ "epoch": 37.91373801916933,
+ "grad_norm": 0.138671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0277,
+ "step": 11867
+ },
+ {
+ "epoch": 37.91693290734824,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0306,
+ "step": 11868
+ },
+ {
+ "epoch": 37.92012779552716,
+ "grad_norm": 0.09912109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0318,
+ "step": 11869
+ },
+ {
+ "epoch": 37.92332268370607,
+ "grad_norm": 0.050048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0421,
+ "step": 11870
+ },
+ {
+ "epoch": 37.926517571884986,
+ "grad_norm": 0.09814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0293,
+ "step": 11871
+ },
+ {
+ "epoch": 37.9297124600639,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0346,
+ "step": 11872
+ },
+ {
+ "epoch": 37.932907348242814,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0327,
+ "step": 11873
+ },
+ {
+ "epoch": 37.936102236421725,
+ "grad_norm": 0.055908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0281,
+ "step": 11874
+ },
+ {
+ "epoch": 37.93929712460064,
+ "grad_norm": 0.0947265625,
+ "learning_rate": 0.0005,
+ "loss": 1.035,
+ "step": 11875
+ },
+ {
+ "epoch": 37.94249201277955,
+ "grad_norm": 0.056396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0476,
+ "step": 11876
+ },
+ {
+ "epoch": 37.945686900958464,
+ "grad_norm": 0.09033203125,
+ "learning_rate": 0.0005,
+ "loss": 1.035,
+ "step": 11877
+ },
+ {
+ "epoch": 37.94888178913738,
+ "grad_norm": 0.04638671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0361,
+ "step": 11878
+ },
+ {
+ "epoch": 37.95207667731629,
+ "grad_norm": 0.09619140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0348,
+ "step": 11879
+ },
+ {
+ "epoch": 37.95527156549521,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0283,
+ "step": 11880
+ },
+ {
+ "epoch": 37.95846645367412,
+ "grad_norm": 0.1103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0413,
+ "step": 11881
+ },
+ {
+ "epoch": 37.96166134185304,
+ "grad_norm": 0.05078125,
+ "learning_rate": 0.0005,
+ "loss": 1.04,
+ "step": 11882
+ },
+ {
+ "epoch": 37.96485623003195,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0382,
+ "step": 11883
+ },
+ {
+ "epoch": 37.968051118210866,
+ "grad_norm": 0.057861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0321,
+ "step": 11884
+ },
+ {
+ "epoch": 37.97124600638978,
+ "grad_norm": 0.1103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0219,
+ "step": 11885
+ },
+ {
+ "epoch": 37.97444089456869,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0344,
+ "step": 11886
+ },
+ {
+ "epoch": 37.977635782747605,
+ "grad_norm": 0.10693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0334,
+ "step": 11887
+ },
+ {
+ "epoch": 37.980830670926515,
+ "grad_norm": 0.091796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0331,
+ "step": 11888
+ },
+ {
+ "epoch": 37.98402555910543,
+ "grad_norm": 0.1171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0403,
+ "step": 11889
+ },
+ {
+ "epoch": 37.98722044728434,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0375,
+ "step": 11890
+ },
+ {
+ "epoch": 37.99041533546326,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0352,
+ "step": 11891
+ },
+ {
+ "epoch": 37.99361022364217,
+ "grad_norm": 0.08642578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0392,
+ "step": 11892
+ },
+ {
+ "epoch": 37.99680511182109,
+ "grad_norm": 0.1328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0408,
+ "step": 11893
+ },
+ {
+ "epoch": 38.0,
+ "grad_norm": 0.10205078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0384,
+ "step": 11894
+ },
+ {
+ "epoch": 38.00319488817891,
+ "grad_norm": 0.154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.009,
+ "step": 11895
+ },
+ {
+ "epoch": 38.00638977635783,
+ "grad_norm": 0.052001953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0377,
+ "step": 11896
+ },
+ {
+ "epoch": 38.00958466453674,
+ "grad_norm": 0.10888671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0207,
+ "step": 11897
+ },
+ {
+ "epoch": 38.01277955271566,
+ "grad_norm": 0.059814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.018,
+ "step": 11898
+ },
+ {
+ "epoch": 38.01597444089457,
+ "grad_norm": 0.09765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0278,
+ "step": 11899
+ },
+ {
+ "epoch": 38.019169329073485,
+ "grad_norm": 0.058349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0299,
+ "step": 11900
+ },
+ {
+ "epoch": 38.022364217252395,
+ "grad_norm": 0.103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0322,
+ "step": 11901
+ },
+ {
+ "epoch": 38.02555910543131,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0302,
+ "step": 11902
+ },
+ {
+ "epoch": 38.02875399361022,
+ "grad_norm": 0.11865234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0317,
+ "step": 11903
+ },
+ {
+ "epoch": 38.031948881789134,
+ "grad_norm": 0.061767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0325,
+ "step": 11904
+ },
+ {
+ "epoch": 38.03514376996805,
+ "grad_norm": 0.109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0337,
+ "step": 11905
+ },
+ {
+ "epoch": 38.03833865814696,
+ "grad_norm": 0.046142578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0332,
+ "step": 11906
+ },
+ {
+ "epoch": 38.04153354632588,
+ "grad_norm": 0.12255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0328,
+ "step": 11907
+ },
+ {
+ "epoch": 38.04472843450479,
+ "grad_norm": 0.054443359375,
+ "learning_rate": 0.0005,
+ "loss": 1.022,
+ "step": 11908
+ },
+ {
+ "epoch": 38.04792332268371,
+ "grad_norm": 0.11181640625,
+ "learning_rate": 0.0005,
+ "loss": 1.037,
+ "step": 11909
+ },
+ {
+ "epoch": 38.05111821086262,
+ "grad_norm": 0.0537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0219,
+ "step": 11910
+ },
+ {
+ "epoch": 38.054313099041536,
+ "grad_norm": 0.1025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.03,
+ "step": 11911
+ },
+ {
+ "epoch": 38.05750798722045,
+ "grad_norm": 0.0537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0379,
+ "step": 11912
+ },
+ {
+ "epoch": 38.06070287539936,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0377,
+ "step": 11913
+ },
+ {
+ "epoch": 38.063897763578275,
+ "grad_norm": 0.056884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.02,
+ "step": 11914
+ },
+ {
+ "epoch": 38.067092651757186,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0343,
+ "step": 11915
+ },
+ {
+ "epoch": 38.0702875399361,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0378,
+ "step": 11916
+ },
+ {
+ "epoch": 38.073482428115014,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.034,
+ "step": 11917
+ },
+ {
+ "epoch": 38.07667731629393,
+ "grad_norm": 0.08154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0445,
+ "step": 11918
+ },
+ {
+ "epoch": 38.07987220447284,
+ "grad_norm": 0.08056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0355,
+ "step": 11919
+ },
+ {
+ "epoch": 38.08306709265176,
+ "grad_norm": 0.09130859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0256,
+ "step": 11920
+ },
+ {
+ "epoch": 38.08626198083067,
+ "grad_norm": 0.11181640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0417,
+ "step": 11921
+ },
+ {
+ "epoch": 38.08945686900959,
+ "grad_norm": 0.052001953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0331,
+ "step": 11922
+ },
+ {
+ "epoch": 38.0926517571885,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0384,
+ "step": 11923
+ },
+ {
+ "epoch": 38.09584664536741,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0302,
+ "step": 11924
+ },
+ {
+ "epoch": 38.09904153354633,
+ "grad_norm": 0.051513671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0306,
+ "step": 11925
+ },
+ {
+ "epoch": 38.10223642172524,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0482,
+ "step": 11926
+ },
+ {
+ "epoch": 38.105431309904155,
+ "grad_norm": 0.057861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0351,
+ "step": 11927
+ },
+ {
+ "epoch": 38.108626198083066,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0353,
+ "step": 11928
+ },
+ {
+ "epoch": 38.11182108626198,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0302,
+ "step": 11929
+ },
+ {
+ "epoch": 38.115015974440894,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0208,
+ "step": 11930
+ },
+ {
+ "epoch": 38.11821086261981,
+ "grad_norm": 0.1259765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0346,
+ "step": 11931
+ },
+ {
+ "epoch": 38.12140575079872,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0323,
+ "step": 11932
+ },
+ {
+ "epoch": 38.12460063897763,
+ "grad_norm": 0.11279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0338,
+ "step": 11933
+ },
+ {
+ "epoch": 38.12779552715655,
+ "grad_norm": 0.08154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0313,
+ "step": 11934
+ },
+ {
+ "epoch": 38.13099041533546,
+ "grad_norm": 0.1044921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0351,
+ "step": 11935
+ },
+ {
+ "epoch": 38.13418530351438,
+ "grad_norm": 0.10009765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0274,
+ "step": 11936
+ },
+ {
+ "epoch": 38.13738019169329,
+ "grad_norm": 0.1435546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0331,
+ "step": 11937
+ },
+ {
+ "epoch": 38.14057507987221,
+ "grad_norm": 0.1259765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0343,
+ "step": 11938
+ },
+ {
+ "epoch": 38.14376996805112,
+ "grad_norm": 0.2275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.031,
+ "step": 11939
+ },
+ {
+ "epoch": 38.146964856230035,
+ "grad_norm": 0.04833984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0334,
+ "step": 11940
+ },
+ {
+ "epoch": 38.150159744408946,
+ "grad_norm": 0.1103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0272,
+ "step": 11941
+ },
+ {
+ "epoch": 38.153354632587856,
+ "grad_norm": 0.061279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0328,
+ "step": 11942
+ },
+ {
+ "epoch": 38.156549520766774,
+ "grad_norm": 0.1279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0434,
+ "step": 11943
+ },
+ {
+ "epoch": 38.159744408945684,
+ "grad_norm": 0.0546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0301,
+ "step": 11944
+ },
+ {
+ "epoch": 38.1629392971246,
+ "grad_norm": 0.12890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0323,
+ "step": 11945
+ },
+ {
+ "epoch": 38.16613418530351,
+ "grad_norm": 0.053955078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0427,
+ "step": 11946
+ },
+ {
+ "epoch": 38.16932907348243,
+ "grad_norm": 0.09521484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0382,
+ "step": 11947
+ },
+ {
+ "epoch": 38.17252396166134,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0226,
+ "step": 11948
+ },
+ {
+ "epoch": 38.17571884984026,
+ "grad_norm": 0.1337890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0329,
+ "step": 11949
+ },
+ {
+ "epoch": 38.17891373801917,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.025,
+ "step": 11950
+ },
+ {
+ "epoch": 38.18210862619808,
+ "grad_norm": 0.1298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0339,
+ "step": 11951
+ },
+ {
+ "epoch": 38.185303514377,
+ "grad_norm": 0.056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0341,
+ "step": 11952
+ },
+ {
+ "epoch": 38.18849840255591,
+ "grad_norm": 0.091796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0348,
+ "step": 11953
+ },
+ {
+ "epoch": 38.191693290734825,
+ "grad_norm": 0.052978515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0331,
+ "step": 11954
+ },
+ {
+ "epoch": 38.194888178913736,
+ "grad_norm": 0.05810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0339,
+ "step": 11955
+ },
+ {
+ "epoch": 38.198083067092654,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0388,
+ "step": 11956
+ },
+ {
+ "epoch": 38.201277955271564,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0359,
+ "step": 11957
+ },
+ {
+ "epoch": 38.20447284345048,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0318,
+ "step": 11958
+ },
+ {
+ "epoch": 38.20766773162939,
+ "grad_norm": 0.1650390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0429,
+ "step": 11959
+ },
+ {
+ "epoch": 38.21086261980831,
+ "grad_norm": 0.10205078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0253,
+ "step": 11960
+ },
+ {
+ "epoch": 38.21405750798722,
+ "grad_norm": 0.12158203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0221,
+ "step": 11961
+ },
+ {
+ "epoch": 38.21725239616613,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0375,
+ "step": 11962
+ },
+ {
+ "epoch": 38.22044728434505,
+ "grad_norm": 0.0830078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0353,
+ "step": 11963
+ },
+ {
+ "epoch": 38.22364217252396,
+ "grad_norm": 0.05615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0247,
+ "step": 11964
+ },
+ {
+ "epoch": 38.22683706070288,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0477,
+ "step": 11965
+ },
+ {
+ "epoch": 38.23003194888179,
+ "grad_norm": 0.083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0309,
+ "step": 11966
+ },
+ {
+ "epoch": 38.233226837060705,
+ "grad_norm": 0.0869140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0234,
+ "step": 11967
+ },
+ {
+ "epoch": 38.236421725239616,
+ "grad_norm": 0.059326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0416,
+ "step": 11968
+ },
+ {
+ "epoch": 38.239616613418534,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0312,
+ "step": 11969
+ },
+ {
+ "epoch": 38.242811501597444,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0454,
+ "step": 11970
+ },
+ {
+ "epoch": 38.246006389776355,
+ "grad_norm": 0.06201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0478,
+ "step": 11971
+ },
+ {
+ "epoch": 38.24920127795527,
+ "grad_norm": 0.0810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0344,
+ "step": 11972
+ },
+ {
+ "epoch": 38.25239616613418,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0271,
+ "step": 11973
+ },
+ {
+ "epoch": 38.2555910543131,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0239,
+ "step": 11974
+ },
+ {
+ "epoch": 38.25878594249201,
+ "grad_norm": 0.1162109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0323,
+ "step": 11975
+ },
+ {
+ "epoch": 38.26198083067093,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0251,
+ "step": 11976
+ },
+ {
+ "epoch": 38.26517571884984,
+ "grad_norm": 0.08544921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0349,
+ "step": 11977
+ },
+ {
+ "epoch": 38.26837060702876,
+ "grad_norm": 0.0595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.026,
+ "step": 11978
+ },
+ {
+ "epoch": 38.27156549520767,
+ "grad_norm": 0.05224609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0182,
+ "step": 11979
+ },
+ {
+ "epoch": 38.27476038338658,
+ "grad_norm": 0.056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.032,
+ "step": 11980
+ },
+ {
+ "epoch": 38.277955271565496,
+ "grad_norm": 0.053955078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0483,
+ "step": 11981
+ },
+ {
+ "epoch": 38.281150159744406,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0356,
+ "step": 11982
+ },
+ {
+ "epoch": 38.284345047923324,
+ "grad_norm": 0.046142578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0345,
+ "step": 11983
+ },
+ {
+ "epoch": 38.287539936102235,
+ "grad_norm": 0.058349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0246,
+ "step": 11984
+ },
+ {
+ "epoch": 38.29073482428115,
+ "grad_norm": 0.0517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0312,
+ "step": 11985
+ },
+ {
+ "epoch": 38.29392971246006,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0299,
+ "step": 11986
+ },
+ {
+ "epoch": 38.29712460063898,
+ "grad_norm": 0.1123046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0213,
+ "step": 11987
+ },
+ {
+ "epoch": 38.30031948881789,
+ "grad_norm": 0.11083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0405,
+ "step": 11988
+ },
+ {
+ "epoch": 38.3035143769968,
+ "grad_norm": 0.0986328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0292,
+ "step": 11989
+ },
+ {
+ "epoch": 38.30670926517572,
+ "grad_norm": 0.09130859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0319,
+ "step": 11990
+ },
+ {
+ "epoch": 38.30990415335463,
+ "grad_norm": 0.06103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0248,
+ "step": 11991
+ },
+ {
+ "epoch": 38.31309904153355,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0389,
+ "step": 11992
+ },
+ {
+ "epoch": 38.31629392971246,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.037,
+ "step": 11993
+ },
+ {
+ "epoch": 38.319488817891376,
+ "grad_norm": 0.0810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0475,
+ "step": 11994
+ },
+ {
+ "epoch": 38.322683706070286,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0276,
+ "step": 11995
+ },
+ {
+ "epoch": 38.325878594249204,
+ "grad_norm": 0.056396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0353,
+ "step": 11996
+ },
+ {
+ "epoch": 38.329073482428115,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0314,
+ "step": 11997
+ },
+ {
+ "epoch": 38.33226837060703,
+ "grad_norm": 0.046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0289,
+ "step": 11998
+ },
+ {
+ "epoch": 38.33546325878594,
+ "grad_norm": 0.05322265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0433,
+ "step": 11999
+ },
+ {
+ "epoch": 38.33865814696485,
+ "grad_norm": 0.07861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0434,
+ "step": 12000
+ },
+ {
+ "epoch": 38.34185303514377,
+ "grad_norm": 0.056884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0375,
+ "step": 12001
+ },
+ {
+ "epoch": 38.34504792332268,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.034,
+ "step": 12002
+ },
+ {
+ "epoch": 38.3482428115016,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0375,
+ "step": 12003
+ },
+ {
+ "epoch": 38.35143769968051,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0314,
+ "step": 12004
+ },
+ {
+ "epoch": 38.35463258785943,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0373,
+ "step": 12005
+ },
+ {
+ "epoch": 38.35782747603834,
+ "grad_norm": 0.11865234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0344,
+ "step": 12006
+ },
+ {
+ "epoch": 38.361022364217256,
+ "grad_norm": 0.056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0309,
+ "step": 12007
+ },
+ {
+ "epoch": 38.364217252396166,
+ "grad_norm": 0.10302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0365,
+ "step": 12008
+ },
+ {
+ "epoch": 38.36741214057508,
+ "grad_norm": 0.050537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0268,
+ "step": 12009
+ },
+ {
+ "epoch": 38.370607028753994,
+ "grad_norm": 0.083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0471,
+ "step": 12010
+ },
+ {
+ "epoch": 38.373801916932905,
+ "grad_norm": 0.055419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0239,
+ "step": 12011
+ },
+ {
+ "epoch": 38.37699680511182,
+ "grad_norm": 0.138671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0289,
+ "step": 12012
+ },
+ {
+ "epoch": 38.38019169329073,
+ "grad_norm": 0.1337890625,
+ "learning_rate": 0.0005,
+ "loss": 1.024,
+ "step": 12013
+ },
+ {
+ "epoch": 38.38338658146965,
+ "grad_norm": 0.091796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0348,
+ "step": 12014
+ },
+ {
+ "epoch": 38.38658146964856,
+ "grad_norm": 0.12890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0345,
+ "step": 12015
+ },
+ {
+ "epoch": 38.38977635782748,
+ "grad_norm": 0.08740234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0316,
+ "step": 12016
+ },
+ {
+ "epoch": 38.39297124600639,
+ "grad_norm": 0.08984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0368,
+ "step": 12017
+ },
+ {
+ "epoch": 38.3961661341853,
+ "grad_norm": 0.09033203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0455,
+ "step": 12018
+ },
+ {
+ "epoch": 38.39936102236422,
+ "grad_norm": 0.10302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0318,
+ "step": 12019
+ },
+ {
+ "epoch": 38.40255591054313,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0277,
+ "step": 12020
+ },
+ {
+ "epoch": 38.405750798722046,
+ "grad_norm": 0.09228515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0356,
+ "step": 12021
+ },
+ {
+ "epoch": 38.40894568690096,
+ "grad_norm": 0.0830078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0423,
+ "step": 12022
+ },
+ {
+ "epoch": 38.412140575079874,
+ "grad_norm": 0.0908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0426,
+ "step": 12023
+ },
+ {
+ "epoch": 38.415335463258785,
+ "grad_norm": 0.060302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0294,
+ "step": 12024
+ },
+ {
+ "epoch": 38.4185303514377,
+ "grad_norm": 0.08642578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0374,
+ "step": 12025
+ },
+ {
+ "epoch": 38.42172523961661,
+ "grad_norm": 0.060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0378,
+ "step": 12026
+ },
+ {
+ "epoch": 38.424920127795524,
+ "grad_norm": 0.057861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.027,
+ "step": 12027
+ },
+ {
+ "epoch": 38.42811501597444,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.033,
+ "step": 12028
+ },
+ {
+ "epoch": 38.43130990415335,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0362,
+ "step": 12029
+ },
+ {
+ "epoch": 38.43450479233227,
+ "grad_norm": 0.0517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0301,
+ "step": 12030
+ },
+ {
+ "epoch": 38.43769968051118,
+ "grad_norm": 0.055419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0372,
+ "step": 12031
+ },
+ {
+ "epoch": 38.4408945686901,
+ "grad_norm": 0.08740234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0418,
+ "step": 12032
+ },
+ {
+ "epoch": 38.44408945686901,
+ "grad_norm": 0.051025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0294,
+ "step": 12033
+ },
+ {
+ "epoch": 38.447284345047926,
+ "grad_norm": 0.080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0354,
+ "step": 12034
+ },
+ {
+ "epoch": 38.45047923322684,
+ "grad_norm": 0.05615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.035,
+ "step": 12035
+ },
+ {
+ "epoch": 38.453674121405754,
+ "grad_norm": 0.06103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0302,
+ "step": 12036
+ },
+ {
+ "epoch": 38.456869009584665,
+ "grad_norm": 0.050048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0109,
+ "step": 12037
+ },
+ {
+ "epoch": 38.460063897763575,
+ "grad_norm": 0.0966796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0275,
+ "step": 12038
+ },
+ {
+ "epoch": 38.46325878594249,
+ "grad_norm": 0.095703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0353,
+ "step": 12039
+ },
+ {
+ "epoch": 38.466453674121404,
+ "grad_norm": 0.0478515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0314,
+ "step": 12040
+ },
+ {
+ "epoch": 38.46964856230032,
+ "grad_norm": 0.059814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0415,
+ "step": 12041
+ },
+ {
+ "epoch": 38.47284345047923,
+ "grad_norm": 0.056884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0282,
+ "step": 12042
+ },
+ {
+ "epoch": 38.47603833865815,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0234,
+ "step": 12043
+ },
+ {
+ "epoch": 38.47923322683706,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0292,
+ "step": 12044
+ },
+ {
+ "epoch": 38.48242811501598,
+ "grad_norm": 0.0947265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0277,
+ "step": 12045
+ },
+ {
+ "epoch": 38.48562300319489,
+ "grad_norm": 0.0908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0376,
+ "step": 12046
+ },
+ {
+ "epoch": 38.4888178913738,
+ "grad_norm": 0.08935546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0271,
+ "step": 12047
+ },
+ {
+ "epoch": 38.49201277955272,
+ "grad_norm": 0.051513671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0389,
+ "step": 12048
+ },
+ {
+ "epoch": 38.49520766773163,
+ "grad_norm": 0.1279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0412,
+ "step": 12049
+ },
+ {
+ "epoch": 38.498402555910545,
+ "grad_norm": 0.0556640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0438,
+ "step": 12050
+ },
+ {
+ "epoch": 38.501597444089455,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0366,
+ "step": 12051
+ },
+ {
+ "epoch": 38.50479233226837,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0462,
+ "step": 12052
+ },
+ {
+ "epoch": 38.50798722044728,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.035,
+ "step": 12053
+ },
+ {
+ "epoch": 38.5111821086262,
+ "grad_norm": 0.061279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0318,
+ "step": 12054
+ },
+ {
+ "epoch": 38.51437699680511,
+ "grad_norm": 0.056396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0254,
+ "step": 12055
+ },
+ {
+ "epoch": 38.51757188498402,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0347,
+ "step": 12056
+ },
+ {
+ "epoch": 38.52076677316294,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0302,
+ "step": 12057
+ },
+ {
+ "epoch": 38.52396166134185,
+ "grad_norm": 0.099609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0445,
+ "step": 12058
+ },
+ {
+ "epoch": 38.52715654952077,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.027,
+ "step": 12059
+ },
+ {
+ "epoch": 38.53035143769968,
+ "grad_norm": 0.08837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0431,
+ "step": 12060
+ },
+ {
+ "epoch": 38.533546325878596,
+ "grad_norm": 0.059326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0439,
+ "step": 12061
+ },
+ {
+ "epoch": 38.53674121405751,
+ "grad_norm": 0.0595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0438,
+ "step": 12062
+ },
+ {
+ "epoch": 38.539936102236425,
+ "grad_norm": 0.05078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0311,
+ "step": 12063
+ },
+ {
+ "epoch": 38.543130990415335,
+ "grad_norm": 0.05322265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0334,
+ "step": 12064
+ },
+ {
+ "epoch": 38.546325878594246,
+ "grad_norm": 0.056884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0397,
+ "step": 12065
+ },
+ {
+ "epoch": 38.54952076677316,
+ "grad_norm": 0.054443359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0233,
+ "step": 12066
+ },
+ {
+ "epoch": 38.552715654952074,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.039,
+ "step": 12067
+ },
+ {
+ "epoch": 38.55591054313099,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.027,
+ "step": 12068
+ },
+ {
+ "epoch": 38.5591054313099,
+ "grad_norm": 0.060791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0413,
+ "step": 12069
+ },
+ {
+ "epoch": 38.56230031948882,
+ "grad_norm": 0.052490234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0514,
+ "step": 12070
+ },
+ {
+ "epoch": 38.56549520766773,
+ "grad_norm": 0.052978515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0334,
+ "step": 12071
+ },
+ {
+ "epoch": 38.56869009584665,
+ "grad_norm": 0.0517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0274,
+ "step": 12072
+ },
+ {
+ "epoch": 38.57188498402556,
+ "grad_norm": 0.05419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0208,
+ "step": 12073
+ },
+ {
+ "epoch": 38.575079872204476,
+ "grad_norm": 0.08056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0389,
+ "step": 12074
+ },
+ {
+ "epoch": 38.57827476038339,
+ "grad_norm": 0.047607421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0279,
+ "step": 12075
+ },
+ {
+ "epoch": 38.5814696485623,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0284,
+ "step": 12076
+ },
+ {
+ "epoch": 38.584664536741215,
+ "grad_norm": 0.1181640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0401,
+ "step": 12077
+ },
+ {
+ "epoch": 38.587859424920126,
+ "grad_norm": 0.051025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0379,
+ "step": 12078
+ },
+ {
+ "epoch": 38.59105431309904,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.026,
+ "step": 12079
+ },
+ {
+ "epoch": 38.594249201277954,
+ "grad_norm": 0.047607421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0345,
+ "step": 12080
+ },
+ {
+ "epoch": 38.59744408945687,
+ "grad_norm": 0.08740234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0383,
+ "step": 12081
+ },
+ {
+ "epoch": 38.60063897763578,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0364,
+ "step": 12082
+ },
+ {
+ "epoch": 38.6038338658147,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0335,
+ "step": 12083
+ },
+ {
+ "epoch": 38.60702875399361,
+ "grad_norm": 0.09912109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0399,
+ "step": 12084
+ },
+ {
+ "epoch": 38.61022364217252,
+ "grad_norm": 0.10107421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0394,
+ "step": 12085
+ },
+ {
+ "epoch": 38.61341853035144,
+ "grad_norm": 0.0478515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0389,
+ "step": 12086
+ },
+ {
+ "epoch": 38.61661341853035,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0401,
+ "step": 12087
+ },
+ {
+ "epoch": 38.61980830670927,
+ "grad_norm": 0.055908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0304,
+ "step": 12088
+ },
+ {
+ "epoch": 38.62300319488818,
+ "grad_norm": 0.05810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0343,
+ "step": 12089
+ },
+ {
+ "epoch": 38.626198083067095,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0309,
+ "step": 12090
+ },
+ {
+ "epoch": 38.629392971246006,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0388,
+ "step": 12091
+ },
+ {
+ "epoch": 38.63258785942492,
+ "grad_norm": 0.08544921875,
+ "learning_rate": 0.0005,
+ "loss": 1.017,
+ "step": 12092
+ },
+ {
+ "epoch": 38.635782747603834,
+ "grad_norm": 0.09765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0246,
+ "step": 12093
+ },
+ {
+ "epoch": 38.638977635782744,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0189,
+ "step": 12094
+ },
+ {
+ "epoch": 38.64217252396166,
+ "grad_norm": 0.1259765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0319,
+ "step": 12095
+ },
+ {
+ "epoch": 38.64536741214057,
+ "grad_norm": 0.056396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0255,
+ "step": 12096
+ },
+ {
+ "epoch": 38.64856230031949,
+ "grad_norm": 0.1298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0351,
+ "step": 12097
+ },
+ {
+ "epoch": 38.6517571884984,
+ "grad_norm": 0.1015625,
+ "learning_rate": 0.0005,
+ "loss": 1.035,
+ "step": 12098
+ },
+ {
+ "epoch": 38.65495207667732,
+ "grad_norm": 0.2080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0348,
+ "step": 12099
+ },
+ {
+ "epoch": 38.65814696485623,
+ "grad_norm": 0.08154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0308,
+ "step": 12100
+ },
+ {
+ "epoch": 38.66134185303515,
+ "grad_norm": 0.09716796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0362,
+ "step": 12101
+ },
+ {
+ "epoch": 38.66453674121406,
+ "grad_norm": 0.095703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0391,
+ "step": 12102
+ },
+ {
+ "epoch": 38.66773162939297,
+ "grad_norm": 0.1015625,
+ "learning_rate": 0.0005,
+ "loss": 1.033,
+ "step": 12103
+ },
+ {
+ "epoch": 38.670926517571885,
+ "grad_norm": 0.0810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0188,
+ "step": 12104
+ },
+ {
+ "epoch": 38.674121405750796,
+ "grad_norm": 0.08642578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0474,
+ "step": 12105
+ },
+ {
+ "epoch": 38.677316293929714,
+ "grad_norm": 0.0595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0354,
+ "step": 12106
+ },
+ {
+ "epoch": 38.680511182108624,
+ "grad_norm": 0.08056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.035,
+ "step": 12107
+ },
+ {
+ "epoch": 38.68370607028754,
+ "grad_norm": 0.0556640625,
+ "learning_rate": 0.0005,
+ "loss": 1.032,
+ "step": 12108
+ },
+ {
+ "epoch": 38.68690095846645,
+ "grad_norm": 0.0869140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0486,
+ "step": 12109
+ },
+ {
+ "epoch": 38.69009584664537,
+ "grad_norm": 0.08740234375,
+ "learning_rate": 0.0005,
+ "loss": 1.019,
+ "step": 12110
+ },
+ {
+ "epoch": 38.69329073482428,
+ "grad_norm": 0.10595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0388,
+ "step": 12111
+ },
+ {
+ "epoch": 38.6964856230032,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0375,
+ "step": 12112
+ },
+ {
+ "epoch": 38.69968051118211,
+ "grad_norm": 0.1552734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0346,
+ "step": 12113
+ },
+ {
+ "epoch": 38.70287539936102,
+ "grad_norm": 0.0830078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0432,
+ "step": 12114
+ },
+ {
+ "epoch": 38.70607028753994,
+ "grad_norm": 0.1025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0372,
+ "step": 12115
+ },
+ {
+ "epoch": 38.70926517571885,
+ "grad_norm": 0.08642578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0292,
+ "step": 12116
+ },
+ {
+ "epoch": 38.712460063897765,
+ "grad_norm": 0.0947265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0251,
+ "step": 12117
+ },
+ {
+ "epoch": 38.715654952076676,
+ "grad_norm": 0.08349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.038,
+ "step": 12118
+ },
+ {
+ "epoch": 38.718849840255594,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0348,
+ "step": 12119
+ },
+ {
+ "epoch": 38.722044728434504,
+ "grad_norm": 0.1083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0333,
+ "step": 12120
+ },
+ {
+ "epoch": 38.72523961661342,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0383,
+ "step": 12121
+ },
+ {
+ "epoch": 38.72843450479233,
+ "grad_norm": 0.1806640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0264,
+ "step": 12122
+ },
+ {
+ "epoch": 38.73162939297124,
+ "grad_norm": 0.053466796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0225,
+ "step": 12123
+ },
+ {
+ "epoch": 38.73482428115016,
+ "grad_norm": 0.212890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0426,
+ "step": 12124
+ },
+ {
+ "epoch": 38.73801916932907,
+ "grad_norm": 0.1630859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0369,
+ "step": 12125
+ },
+ {
+ "epoch": 38.74121405750799,
+ "grad_norm": 0.1103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0431,
+ "step": 12126
+ },
+ {
+ "epoch": 38.7444089456869,
+ "grad_norm": 0.275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0447,
+ "step": 12127
+ },
+ {
+ "epoch": 38.74760383386582,
+ "grad_norm": 0.11572265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0251,
+ "step": 12128
+ },
+ {
+ "epoch": 38.75079872204473,
+ "grad_norm": 0.171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0396,
+ "step": 12129
+ },
+ {
+ "epoch": 38.753993610223645,
+ "grad_norm": 0.224609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0361,
+ "step": 12130
+ },
+ {
+ "epoch": 38.757188498402556,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0306,
+ "step": 12131
+ },
+ {
+ "epoch": 38.760383386581466,
+ "grad_norm": 0.2431640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0414,
+ "step": 12132
+ },
+ {
+ "epoch": 38.763578274760384,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0285,
+ "step": 12133
+ },
+ {
+ "epoch": 38.766773162939295,
+ "grad_norm": 0.1630859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0247,
+ "step": 12134
+ },
+ {
+ "epoch": 38.76996805111821,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0465,
+ "step": 12135
+ },
+ {
+ "epoch": 38.77316293929712,
+ "grad_norm": 0.2109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0221,
+ "step": 12136
+ },
+ {
+ "epoch": 38.77635782747604,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0324,
+ "step": 12137
+ },
+ {
+ "epoch": 38.77955271565495,
+ "grad_norm": 0.1845703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0311,
+ "step": 12138
+ },
+ {
+ "epoch": 38.78274760383387,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.038,
+ "step": 12139
+ },
+ {
+ "epoch": 38.78594249201278,
+ "grad_norm": 0.10791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.035,
+ "step": 12140
+ },
+ {
+ "epoch": 38.78913738019169,
+ "grad_norm": 0.08447265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0397,
+ "step": 12141
+ },
+ {
+ "epoch": 38.79233226837061,
+ "grad_norm": 0.091796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0384,
+ "step": 12142
+ },
+ {
+ "epoch": 38.79552715654952,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0335,
+ "step": 12143
+ },
+ {
+ "epoch": 38.798722044728436,
+ "grad_norm": 0.09130859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0239,
+ "step": 12144
+ },
+ {
+ "epoch": 38.801916932907346,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0294,
+ "step": 12145
+ },
+ {
+ "epoch": 38.805111821086264,
+ "grad_norm": 0.0859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0361,
+ "step": 12146
+ },
+ {
+ "epoch": 38.808306709265175,
+ "grad_norm": 0.08984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0405,
+ "step": 12147
+ },
+ {
+ "epoch": 38.81150159744409,
+ "grad_norm": 0.103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.034,
+ "step": 12148
+ },
+ {
+ "epoch": 38.814696485623,
+ "grad_norm": 0.1162109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0324,
+ "step": 12149
+ },
+ {
+ "epoch": 38.81789137380191,
+ "grad_norm": 0.09375,
+ "learning_rate": 0.0005,
+ "loss": 1.0477,
+ "step": 12150
+ },
+ {
+ "epoch": 38.82108626198083,
+ "grad_norm": 0.1474609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0334,
+ "step": 12151
+ },
+ {
+ "epoch": 38.82428115015974,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.022,
+ "step": 12152
+ },
+ {
+ "epoch": 38.82747603833866,
+ "grad_norm": 0.0830078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0349,
+ "step": 12153
+ },
+ {
+ "epoch": 38.83067092651757,
+ "grad_norm": 0.0986328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0293,
+ "step": 12154
+ },
+ {
+ "epoch": 38.83386581469649,
+ "grad_norm": 0.0849609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0491,
+ "step": 12155
+ },
+ {
+ "epoch": 38.8370607028754,
+ "grad_norm": 0.052734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0304,
+ "step": 12156
+ },
+ {
+ "epoch": 38.840255591054316,
+ "grad_norm": 0.05908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0352,
+ "step": 12157
+ },
+ {
+ "epoch": 38.843450479233226,
+ "grad_norm": 0.05322265625,
+ "learning_rate": 0.0005,
+ "loss": 1.02,
+ "step": 12158
+ },
+ {
+ "epoch": 38.846645367412144,
+ "grad_norm": 0.04833984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0311,
+ "step": 12159
+ },
+ {
+ "epoch": 38.849840255591054,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0298,
+ "step": 12160
+ },
+ {
+ "epoch": 38.853035143769965,
+ "grad_norm": 0.045654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0331,
+ "step": 12161
+ },
+ {
+ "epoch": 38.85623003194888,
+ "grad_norm": 0.049072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0358,
+ "step": 12162
+ },
+ {
+ "epoch": 38.85942492012779,
+ "grad_norm": 0.04931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0375,
+ "step": 12163
+ },
+ {
+ "epoch": 38.86261980830671,
+ "grad_norm": 0.054443359375,
+ "learning_rate": 0.0005,
+ "loss": 1.05,
+ "step": 12164
+ },
+ {
+ "epoch": 38.86581469648562,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0362,
+ "step": 12165
+ },
+ {
+ "epoch": 38.86900958466454,
+ "grad_norm": 0.0478515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0336,
+ "step": 12166
+ },
+ {
+ "epoch": 38.87220447284345,
+ "grad_norm": 0.059326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.033,
+ "step": 12167
+ },
+ {
+ "epoch": 38.87539936102237,
+ "grad_norm": 0.052001953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0368,
+ "step": 12168
+ },
+ {
+ "epoch": 38.87859424920128,
+ "grad_norm": 0.050048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0381,
+ "step": 12169
+ },
+ {
+ "epoch": 38.88178913738019,
+ "grad_norm": 0.0517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.046,
+ "step": 12170
+ },
+ {
+ "epoch": 38.884984025559106,
+ "grad_norm": 0.09814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0349,
+ "step": 12171
+ },
+ {
+ "epoch": 38.88817891373802,
+ "grad_norm": 0.05322265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0397,
+ "step": 12172
+ },
+ {
+ "epoch": 38.891373801916934,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0322,
+ "step": 12173
+ },
+ {
+ "epoch": 38.894568690095845,
+ "grad_norm": 0.08203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0437,
+ "step": 12174
+ },
+ {
+ "epoch": 38.89776357827476,
+ "grad_norm": 0.059814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0247,
+ "step": 12175
+ },
+ {
+ "epoch": 38.90095846645367,
+ "grad_norm": 0.08740234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0273,
+ "step": 12176
+ },
+ {
+ "epoch": 38.90415335463259,
+ "grad_norm": 0.062255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0433,
+ "step": 12177
+ },
+ {
+ "epoch": 38.9073482428115,
+ "grad_norm": 0.0859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0307,
+ "step": 12178
+ },
+ {
+ "epoch": 38.91054313099041,
+ "grad_norm": 0.05126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0323,
+ "step": 12179
+ },
+ {
+ "epoch": 38.91373801916933,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0193,
+ "step": 12180
+ },
+ {
+ "epoch": 38.91693290734824,
+ "grad_norm": 0.051513671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0367,
+ "step": 12181
+ },
+ {
+ "epoch": 38.92012779552716,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0259,
+ "step": 12182
+ },
+ {
+ "epoch": 38.92332268370607,
+ "grad_norm": 0.0478515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0412,
+ "step": 12183
+ },
+ {
+ "epoch": 38.926517571884986,
+ "grad_norm": 0.0849609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0405,
+ "step": 12184
+ },
+ {
+ "epoch": 38.9297124600639,
+ "grad_norm": 0.057861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0274,
+ "step": 12185
+ },
+ {
+ "epoch": 38.932907348242814,
+ "grad_norm": 0.053955078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0261,
+ "step": 12186
+ },
+ {
+ "epoch": 38.936102236421725,
+ "grad_norm": 0.053955078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0356,
+ "step": 12187
+ },
+ {
+ "epoch": 38.93929712460064,
+ "grad_norm": 0.051513671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0478,
+ "step": 12188
+ },
+ {
+ "epoch": 38.94249201277955,
+ "grad_norm": 0.04833984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0351,
+ "step": 12189
+ },
+ {
+ "epoch": 38.945686900958464,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0206,
+ "step": 12190
+ },
+ {
+ "epoch": 38.94888178913738,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0359,
+ "step": 12191
+ },
+ {
+ "epoch": 38.95207667731629,
+ "grad_norm": 0.0537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.033,
+ "step": 12192
+ },
+ {
+ "epoch": 38.95527156549521,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0367,
+ "step": 12193
+ },
+ {
+ "epoch": 38.95846645367412,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0283,
+ "step": 12194
+ },
+ {
+ "epoch": 38.96166134185304,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0399,
+ "step": 12195
+ },
+ {
+ "epoch": 38.96485623003195,
+ "grad_norm": 0.052490234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0286,
+ "step": 12196
+ },
+ {
+ "epoch": 38.968051118210866,
+ "grad_norm": 0.0478515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0375,
+ "step": 12197
+ },
+ {
+ "epoch": 38.97124600638978,
+ "grad_norm": 0.051513671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0301,
+ "step": 12198
+ },
+ {
+ "epoch": 38.97444089456869,
+ "grad_norm": 0.0517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0388,
+ "step": 12199
+ },
+ {
+ "epoch": 38.977635782747605,
+ "grad_norm": 0.056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0395,
+ "step": 12200
+ },
+ {
+ "epoch": 38.980830670926515,
+ "grad_norm": 0.05810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0326,
+ "step": 12201
+ },
+ {
+ "epoch": 38.98402555910543,
+ "grad_norm": 0.05615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0285,
+ "step": 12202
+ },
+ {
+ "epoch": 38.98722044728434,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0301,
+ "step": 12203
+ },
+ {
+ "epoch": 38.99041533546326,
+ "grad_norm": 0.05908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0379,
+ "step": 12204
+ },
+ {
+ "epoch": 38.99361022364217,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0258,
+ "step": 12205
+ },
+ {
+ "epoch": 38.99680511182109,
+ "grad_norm": 0.0595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0359,
+ "step": 12206
+ },
+ {
+ "epoch": 39.0,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.022,
+ "step": 12207
+ },
+ {
+ "epoch": 39.00319488817891,
+ "grad_norm": 0.051025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.03,
+ "step": 12208
+ },
+ {
+ "epoch": 39.00638977635783,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0403,
+ "step": 12209
+ },
+ {
+ "epoch": 39.00958466453674,
+ "grad_norm": 0.052978515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0334,
+ "step": 12210
+ },
+ {
+ "epoch": 39.01277955271566,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0225,
+ "step": 12211
+ },
+ {
+ "epoch": 39.01597444089457,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0405,
+ "step": 12212
+ },
+ {
+ "epoch": 39.019169329073485,
+ "grad_norm": 0.10107421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0322,
+ "step": 12213
+ },
+ {
+ "epoch": 39.022364217252395,
+ "grad_norm": 0.08349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0367,
+ "step": 12214
+ },
+ {
+ "epoch": 39.02555910543131,
+ "grad_norm": 0.142578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0216,
+ "step": 12215
+ },
+ {
+ "epoch": 39.02875399361022,
+ "grad_norm": 0.05419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0349,
+ "step": 12216
+ },
+ {
+ "epoch": 39.031948881789134,
+ "grad_norm": 0.1474609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0384,
+ "step": 12217
+ },
+ {
+ "epoch": 39.03514376996805,
+ "grad_norm": 0.048095703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0398,
+ "step": 12218
+ },
+ {
+ "epoch": 39.03833865814696,
+ "grad_norm": 0.1591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0312,
+ "step": 12219
+ },
+ {
+ "epoch": 39.04153354632588,
+ "grad_norm": 0.08837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0371,
+ "step": 12220
+ },
+ {
+ "epoch": 39.04472843450479,
+ "grad_norm": 0.12890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0328,
+ "step": 12221
+ },
+ {
+ "epoch": 39.04792332268371,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0351,
+ "step": 12222
+ },
+ {
+ "epoch": 39.05111821086262,
+ "grad_norm": 0.1494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0381,
+ "step": 12223
+ },
+ {
+ "epoch": 39.054313099041536,
+ "grad_norm": 0.142578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0336,
+ "step": 12224
+ },
+ {
+ "epoch": 39.05750798722045,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.032,
+ "step": 12225
+ },
+ {
+ "epoch": 39.06070287539936,
+ "grad_norm": 0.09228515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0294,
+ "step": 12226
+ },
+ {
+ "epoch": 39.063897763578275,
+ "grad_norm": 0.08544921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0256,
+ "step": 12227
+ },
+ {
+ "epoch": 39.067092651757186,
+ "grad_norm": 0.08740234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0437,
+ "step": 12228
+ },
+ {
+ "epoch": 39.0702875399361,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0372,
+ "step": 12229
+ },
+ {
+ "epoch": 39.073482428115014,
+ "grad_norm": 0.08203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0274,
+ "step": 12230
+ },
+ {
+ "epoch": 39.07667731629393,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0439,
+ "step": 12231
+ },
+ {
+ "epoch": 39.07987220447284,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0355,
+ "step": 12232
+ },
+ {
+ "epoch": 39.08306709265176,
+ "grad_norm": 0.08251953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0309,
+ "step": 12233
+ },
+ {
+ "epoch": 39.08626198083067,
+ "grad_norm": 0.0546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0271,
+ "step": 12234
+ },
+ {
+ "epoch": 39.08945686900959,
+ "grad_norm": 0.10498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0231,
+ "step": 12235
+ },
+ {
+ "epoch": 39.0926517571885,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0394,
+ "step": 12236
+ },
+ {
+ "epoch": 39.09584664536741,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0213,
+ "step": 12237
+ },
+ {
+ "epoch": 39.09904153354633,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0323,
+ "step": 12238
+ },
+ {
+ "epoch": 39.10223642172524,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0353,
+ "step": 12239
+ },
+ {
+ "epoch": 39.105431309904155,
+ "grad_norm": 0.048583984375,
+ "learning_rate": 0.0005,
+ "loss": 1.023,
+ "step": 12240
+ },
+ {
+ "epoch": 39.108626198083066,
+ "grad_norm": 0.08154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0305,
+ "step": 12241
+ },
+ {
+ "epoch": 39.11182108626198,
+ "grad_norm": 0.05322265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0358,
+ "step": 12242
+ },
+ {
+ "epoch": 39.115015974440894,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.032,
+ "step": 12243
+ },
+ {
+ "epoch": 39.11821086261981,
+ "grad_norm": 0.0615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0363,
+ "step": 12244
+ },
+ {
+ "epoch": 39.12140575079872,
+ "grad_norm": 0.0595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0343,
+ "step": 12245
+ },
+ {
+ "epoch": 39.12460063897763,
+ "grad_norm": 0.0576171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0385,
+ "step": 12246
+ },
+ {
+ "epoch": 39.12779552715655,
+ "grad_norm": 0.059326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0319,
+ "step": 12247
+ },
+ {
+ "epoch": 39.13099041533546,
+ "grad_norm": 0.0615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0309,
+ "step": 12248
+ },
+ {
+ "epoch": 39.13418530351438,
+ "grad_norm": 0.050537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0349,
+ "step": 12249
+ },
+ {
+ "epoch": 39.13738019169329,
+ "grad_norm": 0.083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.025,
+ "step": 12250
+ },
+ {
+ "epoch": 39.14057507987221,
+ "grad_norm": 0.04736328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0346,
+ "step": 12251
+ },
+ {
+ "epoch": 39.14376996805112,
+ "grad_norm": 0.09423828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0399,
+ "step": 12252
+ },
+ {
+ "epoch": 39.146964856230035,
+ "grad_norm": 0.05859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0301,
+ "step": 12253
+ },
+ {
+ "epoch": 39.150159744408946,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0382,
+ "step": 12254
+ },
+ {
+ "epoch": 39.153354632587856,
+ "grad_norm": 0.0517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0266,
+ "step": 12255
+ },
+ {
+ "epoch": 39.156549520766774,
+ "grad_norm": 0.0498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0346,
+ "step": 12256
+ },
+ {
+ "epoch": 39.159744408945684,
+ "grad_norm": 0.04931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0391,
+ "step": 12257
+ },
+ {
+ "epoch": 39.1629392971246,
+ "grad_norm": 0.05029296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0439,
+ "step": 12258
+ },
+ {
+ "epoch": 39.16613418530351,
+ "grad_norm": 0.057373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0449,
+ "step": 12259
+ },
+ {
+ "epoch": 39.16932907348243,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0362,
+ "step": 12260
+ },
+ {
+ "epoch": 39.17252396166134,
+ "grad_norm": 0.0498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0304,
+ "step": 12261
+ },
+ {
+ "epoch": 39.17571884984026,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0329,
+ "step": 12262
+ },
+ {
+ "epoch": 39.17891373801917,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0306,
+ "step": 12263
+ },
+ {
+ "epoch": 39.18210862619808,
+ "grad_norm": 0.05859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0321,
+ "step": 12264
+ },
+ {
+ "epoch": 39.185303514377,
+ "grad_norm": 0.054443359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0317,
+ "step": 12265
+ },
+ {
+ "epoch": 39.18849840255591,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0378,
+ "step": 12266
+ },
+ {
+ "epoch": 39.191693290734825,
+ "grad_norm": 0.0927734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0316,
+ "step": 12267
+ },
+ {
+ "epoch": 39.194888178913736,
+ "grad_norm": 0.047119140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0335,
+ "step": 12268
+ },
+ {
+ "epoch": 39.198083067092654,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0408,
+ "step": 12269
+ },
+ {
+ "epoch": 39.201277955271564,
+ "grad_norm": 0.054443359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0347,
+ "step": 12270
+ },
+ {
+ "epoch": 39.20447284345048,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0229,
+ "step": 12271
+ },
+ {
+ "epoch": 39.20766773162939,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.031,
+ "step": 12272
+ },
+ {
+ "epoch": 39.21086261980831,
+ "grad_norm": 0.047119140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0341,
+ "step": 12273
+ },
+ {
+ "epoch": 39.21405750798722,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0314,
+ "step": 12274
+ },
+ {
+ "epoch": 39.21725239616613,
+ "grad_norm": 0.04736328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0341,
+ "step": 12275
+ },
+ {
+ "epoch": 39.22044728434505,
+ "grad_norm": 0.0947265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0245,
+ "step": 12276
+ },
+ {
+ "epoch": 39.22364217252396,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0276,
+ "step": 12277
+ },
+ {
+ "epoch": 39.22683706070288,
+ "grad_norm": 0.09033203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0342,
+ "step": 12278
+ },
+ {
+ "epoch": 39.23003194888179,
+ "grad_norm": 0.0537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0157,
+ "step": 12279
+ },
+ {
+ "epoch": 39.233226837060705,
+ "grad_norm": 0.0849609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0275,
+ "step": 12280
+ },
+ {
+ "epoch": 39.236421725239616,
+ "grad_norm": 0.061767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0318,
+ "step": 12281
+ },
+ {
+ "epoch": 39.239616613418534,
+ "grad_norm": 0.0517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0398,
+ "step": 12282
+ },
+ {
+ "epoch": 39.242811501597444,
+ "grad_norm": 0.057373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.03,
+ "step": 12283
+ },
+ {
+ "epoch": 39.246006389776355,
+ "grad_norm": 0.05517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.043,
+ "step": 12284
+ },
+ {
+ "epoch": 39.24920127795527,
+ "grad_norm": 0.048095703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0295,
+ "step": 12285
+ },
+ {
+ "epoch": 39.25239616613418,
+ "grad_norm": 0.055419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0396,
+ "step": 12286
+ },
+ {
+ "epoch": 39.2555910543131,
+ "grad_norm": 0.04736328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0355,
+ "step": 12287
+ },
+ {
+ "epoch": 39.25878594249201,
+ "grad_norm": 0.052001953125,
+ "learning_rate": 0.0005,
+ "loss": 1.036,
+ "step": 12288
+ },
+ {
+ "epoch": 39.26198083067093,
+ "grad_norm": 0.058837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0332,
+ "step": 12289
+ },
+ {
+ "epoch": 39.26517571884984,
+ "grad_norm": 0.05517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0449,
+ "step": 12290
+ },
+ {
+ "epoch": 39.26837060702876,
+ "grad_norm": 0.052490234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0287,
+ "step": 12291
+ },
+ {
+ "epoch": 39.27156549520767,
+ "grad_norm": 0.049072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.042,
+ "step": 12292
+ },
+ {
+ "epoch": 39.27476038338658,
+ "grad_norm": 0.050048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0264,
+ "step": 12293
+ },
+ {
+ "epoch": 39.277955271565496,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0216,
+ "step": 12294
+ },
+ {
+ "epoch": 39.281150159744406,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0379,
+ "step": 12295
+ },
+ {
+ "epoch": 39.284345047923324,
+ "grad_norm": 0.058349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0353,
+ "step": 12296
+ },
+ {
+ "epoch": 39.287539936102235,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0244,
+ "step": 12297
+ },
+ {
+ "epoch": 39.29073482428115,
+ "grad_norm": 0.05126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0416,
+ "step": 12298
+ },
+ {
+ "epoch": 39.29392971246006,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0342,
+ "step": 12299
+ },
+ {
+ "epoch": 39.29712460063898,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.028,
+ "step": 12300
+ },
+ {
+ "epoch": 39.30031948881789,
+ "grad_norm": 0.061279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0409,
+ "step": 12301
+ },
+ {
+ "epoch": 39.3035143769968,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0448,
+ "step": 12302
+ },
+ {
+ "epoch": 39.30670926517572,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0303,
+ "step": 12303
+ },
+ {
+ "epoch": 39.30990415335463,
+ "grad_norm": 0.0498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0196,
+ "step": 12304
+ },
+ {
+ "epoch": 39.31309904153355,
+ "grad_norm": 0.05517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.03,
+ "step": 12305
+ },
+ {
+ "epoch": 39.31629392971246,
+ "grad_norm": 0.0615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0364,
+ "step": 12306
+ },
+ {
+ "epoch": 39.319488817891376,
+ "grad_norm": 0.0576171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0285,
+ "step": 12307
+ },
+ {
+ "epoch": 39.322683706070286,
+ "grad_norm": 0.0615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0281,
+ "step": 12308
+ },
+ {
+ "epoch": 39.325878594249204,
+ "grad_norm": 0.09326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0387,
+ "step": 12309
+ },
+ {
+ "epoch": 39.329073482428115,
+ "grad_norm": 0.0830078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0232,
+ "step": 12310
+ },
+ {
+ "epoch": 39.33226837060703,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0312,
+ "step": 12311
+ },
+ {
+ "epoch": 39.33546325878594,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0293,
+ "step": 12312
+ },
+ {
+ "epoch": 39.33865814696485,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0442,
+ "step": 12313
+ },
+ {
+ "epoch": 39.34185303514377,
+ "grad_norm": 0.052734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0384,
+ "step": 12314
+ },
+ {
+ "epoch": 39.34504792332268,
+ "grad_norm": 0.048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0378,
+ "step": 12315
+ },
+ {
+ "epoch": 39.3482428115016,
+ "grad_norm": 0.051025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.034,
+ "step": 12316
+ },
+ {
+ "epoch": 39.35143769968051,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0286,
+ "step": 12317
+ },
+ {
+ "epoch": 39.35463258785943,
+ "grad_norm": 0.05419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.031,
+ "step": 12318
+ },
+ {
+ "epoch": 39.35782747603834,
+ "grad_norm": 0.06005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0312,
+ "step": 12319
+ },
+ {
+ "epoch": 39.361022364217256,
+ "grad_norm": 0.05322265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0256,
+ "step": 12320
+ },
+ {
+ "epoch": 39.364217252396166,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0378,
+ "step": 12321
+ },
+ {
+ "epoch": 39.36741214057508,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.038,
+ "step": 12322
+ },
+ {
+ "epoch": 39.370607028753994,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0439,
+ "step": 12323
+ },
+ {
+ "epoch": 39.373801916932905,
+ "grad_norm": 0.09033203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0278,
+ "step": 12324
+ },
+ {
+ "epoch": 39.37699680511182,
+ "grad_norm": 0.11572265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0339,
+ "step": 12325
+ },
+ {
+ "epoch": 39.38019169329073,
+ "grad_norm": 0.0927734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0384,
+ "step": 12326
+ },
+ {
+ "epoch": 39.38338658146965,
+ "grad_norm": 0.060791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0452,
+ "step": 12327
+ },
+ {
+ "epoch": 39.38658146964856,
+ "grad_norm": 0.1240234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0275,
+ "step": 12328
+ },
+ {
+ "epoch": 39.38977635782748,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0309,
+ "step": 12329
+ },
+ {
+ "epoch": 39.39297124600639,
+ "grad_norm": 0.0966796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0426,
+ "step": 12330
+ },
+ {
+ "epoch": 39.3961661341853,
+ "grad_norm": 0.11181640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0332,
+ "step": 12331
+ },
+ {
+ "epoch": 39.39936102236422,
+ "grad_norm": 0.0595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0377,
+ "step": 12332
+ },
+ {
+ "epoch": 39.40255591054313,
+ "grad_norm": 0.1494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0276,
+ "step": 12333
+ },
+ {
+ "epoch": 39.405750798722046,
+ "grad_norm": 0.053466796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0365,
+ "step": 12334
+ },
+ {
+ "epoch": 39.40894568690096,
+ "grad_norm": 0.115234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0277,
+ "step": 12335
+ },
+ {
+ "epoch": 39.412140575079874,
+ "grad_norm": 0.0908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0359,
+ "step": 12336
+ },
+ {
+ "epoch": 39.415335463258785,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0278,
+ "step": 12337
+ },
+ {
+ "epoch": 39.4185303514377,
+ "grad_norm": 0.08154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0297,
+ "step": 12338
+ },
+ {
+ "epoch": 39.42172523961661,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0387,
+ "step": 12339
+ },
+ {
+ "epoch": 39.424920127795524,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0227,
+ "step": 12340
+ },
+ {
+ "epoch": 39.42811501597444,
+ "grad_norm": 0.0859375,
+ "learning_rate": 0.0005,
+ "loss": 1.035,
+ "step": 12341
+ },
+ {
+ "epoch": 39.43130990415335,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0317,
+ "step": 12342
+ },
+ {
+ "epoch": 39.43450479233227,
+ "grad_norm": 0.09814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0389,
+ "step": 12343
+ },
+ {
+ "epoch": 39.43769968051118,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.038,
+ "step": 12344
+ },
+ {
+ "epoch": 39.4408945686901,
+ "grad_norm": 0.056396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0366,
+ "step": 12345
+ },
+ {
+ "epoch": 39.44408945686901,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0459,
+ "step": 12346
+ },
+ {
+ "epoch": 39.447284345047926,
+ "grad_norm": 0.05615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0297,
+ "step": 12347
+ },
+ {
+ "epoch": 39.45047923322684,
+ "grad_norm": 0.06201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0353,
+ "step": 12348
+ },
+ {
+ "epoch": 39.453674121405754,
+ "grad_norm": 0.0908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0308,
+ "step": 12349
+ },
+ {
+ "epoch": 39.456869009584665,
+ "grad_norm": 0.04833984375,
+ "learning_rate": 0.0005,
+ "loss": 1.046,
+ "step": 12350
+ },
+ {
+ "epoch": 39.460063897763575,
+ "grad_norm": 0.0595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0319,
+ "step": 12351
+ },
+ {
+ "epoch": 39.46325878594249,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0337,
+ "step": 12352
+ },
+ {
+ "epoch": 39.466453674121404,
+ "grad_norm": 0.061279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0309,
+ "step": 12353
+ },
+ {
+ "epoch": 39.46964856230032,
+ "grad_norm": 0.0537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0299,
+ "step": 12354
+ },
+ {
+ "epoch": 39.47284345047923,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0275,
+ "step": 12355
+ },
+ {
+ "epoch": 39.47603833865815,
+ "grad_norm": 0.09716796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0368,
+ "step": 12356
+ },
+ {
+ "epoch": 39.47923322683706,
+ "grad_norm": 0.091796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0262,
+ "step": 12357
+ },
+ {
+ "epoch": 39.48242811501598,
+ "grad_norm": 0.10498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0338,
+ "step": 12358
+ },
+ {
+ "epoch": 39.48562300319489,
+ "grad_norm": 0.1083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0342,
+ "step": 12359
+ },
+ {
+ "epoch": 39.4888178913738,
+ "grad_norm": 0.1005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0412,
+ "step": 12360
+ },
+ {
+ "epoch": 39.49201277955272,
+ "grad_norm": 0.1015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0302,
+ "step": 12361
+ },
+ {
+ "epoch": 39.49520766773163,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0216,
+ "step": 12362
+ },
+ {
+ "epoch": 39.498402555910545,
+ "grad_norm": 0.1640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0366,
+ "step": 12363
+ },
+ {
+ "epoch": 39.501597444089455,
+ "grad_norm": 0.12353515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0387,
+ "step": 12364
+ },
+ {
+ "epoch": 39.50479233226837,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0204,
+ "step": 12365
+ },
+ {
+ "epoch": 39.50798722044728,
+ "grad_norm": 0.1259765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0296,
+ "step": 12366
+ },
+ {
+ "epoch": 39.5111821086262,
+ "grad_norm": 0.10107421875,
+ "learning_rate": 0.0005,
+ "loss": 1.041,
+ "step": 12367
+ },
+ {
+ "epoch": 39.51437699680511,
+ "grad_norm": 0.1259765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0386,
+ "step": 12368
+ },
+ {
+ "epoch": 39.51757188498402,
+ "grad_norm": 0.11181640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0396,
+ "step": 12369
+ },
+ {
+ "epoch": 39.52076677316294,
+ "grad_norm": 0.11865234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0288,
+ "step": 12370
+ },
+ {
+ "epoch": 39.52396166134185,
+ "grad_norm": 0.16015625,
+ "learning_rate": 0.0005,
+ "loss": 1.03,
+ "step": 12371
+ },
+ {
+ "epoch": 39.52715654952077,
+ "grad_norm": 0.0947265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0241,
+ "step": 12372
+ },
+ {
+ "epoch": 39.53035143769968,
+ "grad_norm": 0.1923828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0313,
+ "step": 12373
+ },
+ {
+ "epoch": 39.533546325878596,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0317,
+ "step": 12374
+ },
+ {
+ "epoch": 39.53674121405751,
+ "grad_norm": 0.1298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0374,
+ "step": 12375
+ },
+ {
+ "epoch": 39.539936102236425,
+ "grad_norm": 0.1142578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0315,
+ "step": 12376
+ },
+ {
+ "epoch": 39.543130990415335,
+ "grad_norm": 0.09228515625,
+ "learning_rate": 0.0005,
+ "loss": 1.028,
+ "step": 12377
+ },
+ {
+ "epoch": 39.546325878594246,
+ "grad_norm": 0.1552734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0226,
+ "step": 12378
+ },
+ {
+ "epoch": 39.54952076677316,
+ "grad_norm": 0.08251953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0376,
+ "step": 12379
+ },
+ {
+ "epoch": 39.552715654952074,
+ "grad_norm": 0.154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0303,
+ "step": 12380
+ },
+ {
+ "epoch": 39.55591054313099,
+ "grad_norm": 0.080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0375,
+ "step": 12381
+ },
+ {
+ "epoch": 39.5591054313099,
+ "grad_norm": 0.1689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0378,
+ "step": 12382
+ },
+ {
+ "epoch": 39.56230031948882,
+ "grad_norm": 0.06005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0254,
+ "step": 12383
+ },
+ {
+ "epoch": 39.56549520766773,
+ "grad_norm": 0.181640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0336,
+ "step": 12384
+ },
+ {
+ "epoch": 39.56869009584665,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0239,
+ "step": 12385
+ },
+ {
+ "epoch": 39.57188498402556,
+ "grad_norm": 0.1533203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0359,
+ "step": 12386
+ },
+ {
+ "epoch": 39.575079872204476,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0357,
+ "step": 12387
+ },
+ {
+ "epoch": 39.57827476038339,
+ "grad_norm": 0.1884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0256,
+ "step": 12388
+ },
+ {
+ "epoch": 39.5814696485623,
+ "grad_norm": 0.057373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0357,
+ "step": 12389
+ },
+ {
+ "epoch": 39.584664536741215,
+ "grad_norm": 0.1474609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0367,
+ "step": 12390
+ },
+ {
+ "epoch": 39.587859424920126,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0343,
+ "step": 12391
+ },
+ {
+ "epoch": 39.59105431309904,
+ "grad_norm": 0.138671875,
+ "learning_rate": 0.0005,
+ "loss": 1.036,
+ "step": 12392
+ },
+ {
+ "epoch": 39.594249201277954,
+ "grad_norm": 0.060791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0323,
+ "step": 12393
+ },
+ {
+ "epoch": 39.59744408945687,
+ "grad_norm": 0.08984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0357,
+ "step": 12394
+ },
+ {
+ "epoch": 39.60063897763578,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0269,
+ "step": 12395
+ },
+ {
+ "epoch": 39.6038338658147,
+ "grad_norm": 0.08837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0215,
+ "step": 12396
+ },
+ {
+ "epoch": 39.60702875399361,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0412,
+ "step": 12397
+ },
+ {
+ "epoch": 39.61022364217252,
+ "grad_norm": 0.11376953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0348,
+ "step": 12398
+ },
+ {
+ "epoch": 39.61341853035144,
+ "grad_norm": 0.06201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.043,
+ "step": 12399
+ },
+ {
+ "epoch": 39.61661341853035,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0269,
+ "step": 12400
+ },
+ {
+ "epoch": 39.61980830670927,
+ "grad_norm": 0.061279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.029,
+ "step": 12401
+ },
+ {
+ "epoch": 39.62300319488818,
+ "grad_norm": 0.08251953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0345,
+ "step": 12402
+ },
+ {
+ "epoch": 39.626198083067095,
+ "grad_norm": 0.0830078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0355,
+ "step": 12403
+ },
+ {
+ "epoch": 39.629392971246006,
+ "grad_norm": 0.0615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0361,
+ "step": 12404
+ },
+ {
+ "epoch": 39.63258785942492,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0192,
+ "step": 12405
+ },
+ {
+ "epoch": 39.635782747603834,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0222,
+ "step": 12406
+ },
+ {
+ "epoch": 39.638977635782744,
+ "grad_norm": 0.05126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.043,
+ "step": 12407
+ },
+ {
+ "epoch": 39.64217252396166,
+ "grad_norm": 0.1357421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0288,
+ "step": 12408
+ },
+ {
+ "epoch": 39.64536741214057,
+ "grad_norm": 0.0810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0382,
+ "step": 12409
+ },
+ {
+ "epoch": 39.64856230031949,
+ "grad_norm": 0.09912109375,
+ "learning_rate": 0.0005,
+ "loss": 1.022,
+ "step": 12410
+ },
+ {
+ "epoch": 39.6517571884984,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0432,
+ "step": 12411
+ },
+ {
+ "epoch": 39.65495207667732,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0318,
+ "step": 12412
+ },
+ {
+ "epoch": 39.65814696485623,
+ "grad_norm": 0.1162109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0411,
+ "step": 12413
+ },
+ {
+ "epoch": 39.66134185303515,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0257,
+ "step": 12414
+ },
+ {
+ "epoch": 39.66453674121406,
+ "grad_norm": 0.10986328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0394,
+ "step": 12415
+ },
+ {
+ "epoch": 39.66773162939297,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0369,
+ "step": 12416
+ },
+ {
+ "epoch": 39.670926517571885,
+ "grad_norm": 0.08447265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0322,
+ "step": 12417
+ },
+ {
+ "epoch": 39.674121405750796,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0271,
+ "step": 12418
+ },
+ {
+ "epoch": 39.677316293929714,
+ "grad_norm": 0.05126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0208,
+ "step": 12419
+ },
+ {
+ "epoch": 39.680511182108624,
+ "grad_norm": 0.048583984375,
+ "learning_rate": 0.0005,
+ "loss": 1.038,
+ "step": 12420
+ },
+ {
+ "epoch": 39.68370607028754,
+ "grad_norm": 0.05322265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0295,
+ "step": 12421
+ },
+ {
+ "epoch": 39.68690095846645,
+ "grad_norm": 0.058837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0263,
+ "step": 12422
+ },
+ {
+ "epoch": 39.69009584664537,
+ "grad_norm": 0.0517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0369,
+ "step": 12423
+ },
+ {
+ "epoch": 39.69329073482428,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.031,
+ "step": 12424
+ },
+ {
+ "epoch": 39.6964856230032,
+ "grad_norm": 0.06201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0292,
+ "step": 12425
+ },
+ {
+ "epoch": 39.69968051118211,
+ "grad_norm": 0.1533203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0294,
+ "step": 12426
+ },
+ {
+ "epoch": 39.70287539936102,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0293,
+ "step": 12427
+ },
+ {
+ "epoch": 39.70607028753994,
+ "grad_norm": 0.134765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0303,
+ "step": 12428
+ },
+ {
+ "epoch": 39.70926517571885,
+ "grad_norm": 0.05419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.043,
+ "step": 12429
+ },
+ {
+ "epoch": 39.712460063897765,
+ "grad_norm": 0.1103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0358,
+ "step": 12430
+ },
+ {
+ "epoch": 39.715654952076676,
+ "grad_norm": 0.052734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0284,
+ "step": 12431
+ },
+ {
+ "epoch": 39.718849840255594,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0284,
+ "step": 12432
+ },
+ {
+ "epoch": 39.722044728434504,
+ "grad_norm": 0.05419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0432,
+ "step": 12433
+ },
+ {
+ "epoch": 39.72523961661342,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0337,
+ "step": 12434
+ },
+ {
+ "epoch": 39.72843450479233,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.037,
+ "step": 12435
+ },
+ {
+ "epoch": 39.73162939297124,
+ "grad_norm": 0.0556640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0342,
+ "step": 12436
+ },
+ {
+ "epoch": 39.73482428115016,
+ "grad_norm": 0.08251953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0392,
+ "step": 12437
+ },
+ {
+ "epoch": 39.73801916932907,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.047,
+ "step": 12438
+ },
+ {
+ "epoch": 39.74121405750799,
+ "grad_norm": 0.09814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.043,
+ "step": 12439
+ },
+ {
+ "epoch": 39.7444089456869,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0385,
+ "step": 12440
+ },
+ {
+ "epoch": 39.74760383386582,
+ "grad_norm": 0.0556640625,
+ "learning_rate": 0.0005,
+ "loss": 1.029,
+ "step": 12441
+ },
+ {
+ "epoch": 39.75079872204473,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0563,
+ "step": 12442
+ },
+ {
+ "epoch": 39.753993610223645,
+ "grad_norm": 0.0537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0368,
+ "step": 12443
+ },
+ {
+ "epoch": 39.757188498402556,
+ "grad_norm": 0.08349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0404,
+ "step": 12444
+ },
+ {
+ "epoch": 39.760383386581466,
+ "grad_norm": 0.09716796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0172,
+ "step": 12445
+ },
+ {
+ "epoch": 39.763578274760384,
+ "grad_norm": 0.09375,
+ "learning_rate": 0.0005,
+ "loss": 1.0309,
+ "step": 12446
+ },
+ {
+ "epoch": 39.766773162939295,
+ "grad_norm": 0.1337890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0262,
+ "step": 12447
+ },
+ {
+ "epoch": 39.76996805111821,
+ "grad_norm": 0.091796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0371,
+ "step": 12448
+ },
+ {
+ "epoch": 39.77316293929712,
+ "grad_norm": 0.18359375,
+ "learning_rate": 0.0005,
+ "loss": 1.033,
+ "step": 12449
+ },
+ {
+ "epoch": 39.77635782747604,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0309,
+ "step": 12450
+ },
+ {
+ "epoch": 39.77955271565495,
+ "grad_norm": 0.1259765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0245,
+ "step": 12451
+ },
+ {
+ "epoch": 39.78274760383387,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0271,
+ "step": 12452
+ },
+ {
+ "epoch": 39.78594249201278,
+ "grad_norm": 0.107421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0396,
+ "step": 12453
+ },
+ {
+ "epoch": 39.78913738019169,
+ "grad_norm": 0.058837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0321,
+ "step": 12454
+ },
+ {
+ "epoch": 39.79233226837061,
+ "grad_norm": 0.0986328125,
+ "learning_rate": 0.0005,
+ "loss": 1.037,
+ "step": 12455
+ },
+ {
+ "epoch": 39.79552715654952,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0332,
+ "step": 12456
+ },
+ {
+ "epoch": 39.798722044728436,
+ "grad_norm": 0.12890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0429,
+ "step": 12457
+ },
+ {
+ "epoch": 39.801916932907346,
+ "grad_norm": 0.12109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0381,
+ "step": 12458
+ },
+ {
+ "epoch": 39.805111821086264,
+ "grad_norm": 0.08447265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0427,
+ "step": 12459
+ },
+ {
+ "epoch": 39.808306709265175,
+ "grad_norm": 0.099609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0397,
+ "step": 12460
+ },
+ {
+ "epoch": 39.81150159744409,
+ "grad_norm": 0.08984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0277,
+ "step": 12461
+ },
+ {
+ "epoch": 39.814696485623,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0345,
+ "step": 12462
+ },
+ {
+ "epoch": 39.81789137380191,
+ "grad_norm": 0.0849609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0377,
+ "step": 12463
+ },
+ {
+ "epoch": 39.82108626198083,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0289,
+ "step": 12464
+ },
+ {
+ "epoch": 39.82428115015974,
+ "grad_norm": 0.08056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.034,
+ "step": 12465
+ },
+ {
+ "epoch": 39.82747603833866,
+ "grad_norm": 0.09375,
+ "learning_rate": 0.0005,
+ "loss": 1.0419,
+ "step": 12466
+ },
+ {
+ "epoch": 39.83067092651757,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0322,
+ "step": 12467
+ },
+ {
+ "epoch": 39.83386581469649,
+ "grad_norm": 0.1123046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0271,
+ "step": 12468
+ },
+ {
+ "epoch": 39.8370607028754,
+ "grad_norm": 0.083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.04,
+ "step": 12469
+ },
+ {
+ "epoch": 39.840255591054316,
+ "grad_norm": 0.09521484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0244,
+ "step": 12470
+ },
+ {
+ "epoch": 39.843450479233226,
+ "grad_norm": 0.0966796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0427,
+ "step": 12471
+ },
+ {
+ "epoch": 39.846645367412144,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0396,
+ "step": 12472
+ },
+ {
+ "epoch": 39.849840255591054,
+ "grad_norm": 0.08203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0411,
+ "step": 12473
+ },
+ {
+ "epoch": 39.853035143769965,
+ "grad_norm": 0.09130859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0358,
+ "step": 12474
+ },
+ {
+ "epoch": 39.85623003194888,
+ "grad_norm": 0.059326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0393,
+ "step": 12475
+ },
+ {
+ "epoch": 39.85942492012779,
+ "grad_norm": 0.09375,
+ "learning_rate": 0.0005,
+ "loss": 1.0327,
+ "step": 12476
+ },
+ {
+ "epoch": 39.86261980830671,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0263,
+ "step": 12477
+ },
+ {
+ "epoch": 39.86581469648562,
+ "grad_norm": 0.1494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0289,
+ "step": 12478
+ },
+ {
+ "epoch": 39.86900958466454,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0387,
+ "step": 12479
+ },
+ {
+ "epoch": 39.87220447284345,
+ "grad_norm": 0.138671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0227,
+ "step": 12480
+ },
+ {
+ "epoch": 39.87539936102237,
+ "grad_norm": 0.052734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0315,
+ "step": 12481
+ },
+ {
+ "epoch": 39.87859424920128,
+ "grad_norm": 0.1416015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0282,
+ "step": 12482
+ },
+ {
+ "epoch": 39.88178913738019,
+ "grad_norm": 0.050537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0381,
+ "step": 12483
+ },
+ {
+ "epoch": 39.884984025559106,
+ "grad_norm": 0.1337890625,
+ "learning_rate": 0.0005,
+ "loss": 1.042,
+ "step": 12484
+ },
+ {
+ "epoch": 39.88817891373802,
+ "grad_norm": 0.04296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0321,
+ "step": 12485
+ },
+ {
+ "epoch": 39.891373801916934,
+ "grad_norm": 0.103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0407,
+ "step": 12486
+ },
+ {
+ "epoch": 39.894568690095845,
+ "grad_norm": 0.05078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0285,
+ "step": 12487
+ },
+ {
+ "epoch": 39.89776357827476,
+ "grad_norm": 0.09716796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0239,
+ "step": 12488
+ },
+ {
+ "epoch": 39.90095846645367,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0387,
+ "step": 12489
+ },
+ {
+ "epoch": 39.90415335463259,
+ "grad_norm": 0.111328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0289,
+ "step": 12490
+ },
+ {
+ "epoch": 39.9073482428115,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0257,
+ "step": 12491
+ },
+ {
+ "epoch": 39.91054313099041,
+ "grad_norm": 0.09765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0469,
+ "step": 12492
+ },
+ {
+ "epoch": 39.91373801916933,
+ "grad_norm": 0.0498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0192,
+ "step": 12493
+ },
+ {
+ "epoch": 39.91693290734824,
+ "grad_norm": 0.08740234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0443,
+ "step": 12494
+ },
+ {
+ "epoch": 39.92012779552716,
+ "grad_norm": 0.050048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0356,
+ "step": 12495
+ },
+ {
+ "epoch": 39.92332268370607,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0302,
+ "step": 12496
+ },
+ {
+ "epoch": 39.926517571884986,
+ "grad_norm": 0.10107421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0331,
+ "step": 12497
+ },
+ {
+ "epoch": 39.9297124600639,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0322,
+ "step": 12498
+ },
+ {
+ "epoch": 39.932907348242814,
+ "grad_norm": 0.083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0371,
+ "step": 12499
+ },
+ {
+ "epoch": 39.936102236421725,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0323,
+ "step": 12500
+ },
+ {
+ "epoch": 39.93929712460064,
+ "grad_norm": 0.059814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0317,
+ "step": 12501
+ },
+ {
+ "epoch": 39.94249201277955,
+ "grad_norm": 0.054931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0317,
+ "step": 12502
+ },
+ {
+ "epoch": 39.945686900958464,
+ "grad_norm": 0.05517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0329,
+ "step": 12503
+ },
+ {
+ "epoch": 39.94888178913738,
+ "grad_norm": 0.052001953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0296,
+ "step": 12504
+ },
+ {
+ "epoch": 39.95207667731629,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.031,
+ "step": 12505
+ },
+ {
+ "epoch": 39.95527156549521,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0146,
+ "step": 12506
+ },
+ {
+ "epoch": 39.95846645367412,
+ "grad_norm": 0.0556640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0232,
+ "step": 12507
+ },
+ {
+ "epoch": 39.96166134185304,
+ "grad_norm": 0.045166015625,
+ "learning_rate": 0.0005,
+ "loss": 1.035,
+ "step": 12508
+ },
+ {
+ "epoch": 39.96485623003195,
+ "grad_norm": 0.05908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0323,
+ "step": 12509
+ },
+ {
+ "epoch": 39.968051118210866,
+ "grad_norm": 0.05224609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0254,
+ "step": 12510
+ },
+ {
+ "epoch": 39.97124600638978,
+ "grad_norm": 0.048095703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0361,
+ "step": 12511
+ },
+ {
+ "epoch": 39.97444089456869,
+ "grad_norm": 0.06005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0295,
+ "step": 12512
+ },
+ {
+ "epoch": 39.977635782747605,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0325,
+ "step": 12513
+ },
+ {
+ "epoch": 39.980830670926515,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0319,
+ "step": 12514
+ },
+ {
+ "epoch": 39.98402555910543,
+ "grad_norm": 0.058837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0362,
+ "step": 12515
+ },
+ {
+ "epoch": 39.98722044728434,
+ "grad_norm": 0.056884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0299,
+ "step": 12516
+ },
+ {
+ "epoch": 39.99041533546326,
+ "grad_norm": 0.0478515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0365,
+ "step": 12517
+ },
+ {
+ "epoch": 39.99361022364217,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.041,
+ "step": 12518
+ },
+ {
+ "epoch": 39.99680511182109,
+ "grad_norm": 0.095703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0263,
+ "step": 12519
+ },
+ {
+ "epoch": 40.0,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0262,
+ "step": 12520
+ },
+ {
+ "epoch": 40.00319488817891,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0276,
+ "step": 12521
+ },
+ {
+ "epoch": 40.00638977635783,
+ "grad_norm": 0.09619140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0264,
+ "step": 12522
+ },
+ {
+ "epoch": 40.00958466453674,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0316,
+ "step": 12523
+ },
+ {
+ "epoch": 40.01277955271566,
+ "grad_norm": 0.058837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0368,
+ "step": 12524
+ },
+ {
+ "epoch": 40.01597444089457,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.04,
+ "step": 12525
+ },
+ {
+ "epoch": 40.019169329073485,
+ "grad_norm": 0.10400390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0363,
+ "step": 12526
+ },
+ {
+ "epoch": 40.022364217252395,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0275,
+ "step": 12527
+ },
+ {
+ "epoch": 40.02555910543131,
+ "grad_norm": 0.08740234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0318,
+ "step": 12528
+ },
+ {
+ "epoch": 40.02875399361022,
+ "grad_norm": 0.091796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0232,
+ "step": 12529
+ },
+ {
+ "epoch": 40.031948881789134,
+ "grad_norm": 0.059814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0221,
+ "step": 12530
+ },
+ {
+ "epoch": 40.03514376996805,
+ "grad_norm": 0.048095703125,
+ "learning_rate": 0.0005,
+ "loss": 1.022,
+ "step": 12531
+ },
+ {
+ "epoch": 40.03833865814696,
+ "grad_norm": 0.06005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0277,
+ "step": 12532
+ },
+ {
+ "epoch": 40.04153354632588,
+ "grad_norm": 0.05224609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0408,
+ "step": 12533
+ },
+ {
+ "epoch": 40.04472843450479,
+ "grad_norm": 0.05322265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0253,
+ "step": 12534
+ },
+ {
+ "epoch": 40.04792332268371,
+ "grad_norm": 0.05029296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0327,
+ "step": 12535
+ },
+ {
+ "epoch": 40.05111821086262,
+ "grad_norm": 0.049072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0315,
+ "step": 12536
+ },
+ {
+ "epoch": 40.054313099041536,
+ "grad_norm": 0.061279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0269,
+ "step": 12537
+ },
+ {
+ "epoch": 40.05750798722045,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0274,
+ "step": 12538
+ },
+ {
+ "epoch": 40.06070287539936,
+ "grad_norm": 0.05712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0248,
+ "step": 12539
+ },
+ {
+ "epoch": 40.063897763578275,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0281,
+ "step": 12540
+ },
+ {
+ "epoch": 40.067092651757186,
+ "grad_norm": 0.052978515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0248,
+ "step": 12541
+ },
+ {
+ "epoch": 40.0702875399361,
+ "grad_norm": 0.06005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0294,
+ "step": 12542
+ },
+ {
+ "epoch": 40.073482428115014,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0279,
+ "step": 12543
+ },
+ {
+ "epoch": 40.07667731629393,
+ "grad_norm": 0.052001953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0447,
+ "step": 12544
+ },
+ {
+ "epoch": 40.07987220447284,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0286,
+ "step": 12545
+ },
+ {
+ "epoch": 40.08306709265176,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0272,
+ "step": 12546
+ },
+ {
+ "epoch": 40.08626198083067,
+ "grad_norm": 0.05615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0325,
+ "step": 12547
+ },
+ {
+ "epoch": 40.08945686900959,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0256,
+ "step": 12548
+ },
+ {
+ "epoch": 40.0926517571885,
+ "grad_norm": 0.0576171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0242,
+ "step": 12549
+ },
+ {
+ "epoch": 40.09584664536741,
+ "grad_norm": 0.05322265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0352,
+ "step": 12550
+ },
+ {
+ "epoch": 40.09904153354633,
+ "grad_norm": 0.044677734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0336,
+ "step": 12551
+ },
+ {
+ "epoch": 40.10223642172524,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0296,
+ "step": 12552
+ },
+ {
+ "epoch": 40.105431309904155,
+ "grad_norm": 0.059326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0405,
+ "step": 12553
+ },
+ {
+ "epoch": 40.108626198083066,
+ "grad_norm": 0.047607421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0352,
+ "step": 12554
+ },
+ {
+ "epoch": 40.11182108626198,
+ "grad_norm": 0.061279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0341,
+ "step": 12555
+ },
+ {
+ "epoch": 40.115015974440894,
+ "grad_norm": 0.043701171875,
+ "learning_rate": 0.0005,
+ "loss": 1.027,
+ "step": 12556
+ },
+ {
+ "epoch": 40.11821086261981,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0449,
+ "step": 12557
+ },
+ {
+ "epoch": 40.12140575079872,
+ "grad_norm": 0.048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0326,
+ "step": 12558
+ },
+ {
+ "epoch": 40.12460063897763,
+ "grad_norm": 0.060302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0378,
+ "step": 12559
+ },
+ {
+ "epoch": 40.12779552715655,
+ "grad_norm": 0.048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0368,
+ "step": 12560
+ },
+ {
+ "epoch": 40.13099041533546,
+ "grad_norm": 0.0478515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0383,
+ "step": 12561
+ },
+ {
+ "epoch": 40.13418530351438,
+ "grad_norm": 0.046142578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0349,
+ "step": 12562
+ },
+ {
+ "epoch": 40.13738019169329,
+ "grad_norm": 0.048095703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0301,
+ "step": 12563
+ },
+ {
+ "epoch": 40.14057507987221,
+ "grad_norm": 0.044677734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0267,
+ "step": 12564
+ },
+ {
+ "epoch": 40.14376996805112,
+ "grad_norm": 0.052734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0338,
+ "step": 12565
+ },
+ {
+ "epoch": 40.146964856230035,
+ "grad_norm": 0.047607421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0421,
+ "step": 12566
+ },
+ {
+ "epoch": 40.150159744408946,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0315,
+ "step": 12567
+ },
+ {
+ "epoch": 40.153354632587856,
+ "grad_norm": 0.047119140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0364,
+ "step": 12568
+ },
+ {
+ "epoch": 40.156549520766774,
+ "grad_norm": 0.05078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0366,
+ "step": 12569
+ },
+ {
+ "epoch": 40.159744408945684,
+ "grad_norm": 0.0556640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0153,
+ "step": 12570
+ },
+ {
+ "epoch": 40.1629392971246,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0331,
+ "step": 12571
+ },
+ {
+ "epoch": 40.16613418530351,
+ "grad_norm": 0.055419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0426,
+ "step": 12572
+ },
+ {
+ "epoch": 40.16932907348243,
+ "grad_norm": 0.0546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0273,
+ "step": 12573
+ },
+ {
+ "epoch": 40.17252396166134,
+ "grad_norm": 0.052734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0384,
+ "step": 12574
+ },
+ {
+ "epoch": 40.17571884984026,
+ "grad_norm": 0.05078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0307,
+ "step": 12575
+ },
+ {
+ "epoch": 40.17891373801917,
+ "grad_norm": 0.053955078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0415,
+ "step": 12576
+ },
+ {
+ "epoch": 40.18210862619808,
+ "grad_norm": 0.08203125,
+ "learning_rate": 0.0005,
+ "loss": 1.037,
+ "step": 12577
+ },
+ {
+ "epoch": 40.185303514377,
+ "grad_norm": 0.06103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0316,
+ "step": 12578
+ },
+ {
+ "epoch": 40.18849840255591,
+ "grad_norm": 0.046142578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0241,
+ "step": 12579
+ },
+ {
+ "epoch": 40.191693290734825,
+ "grad_norm": 0.04833984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0243,
+ "step": 12580
+ },
+ {
+ "epoch": 40.194888178913736,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0298,
+ "step": 12581
+ },
+ {
+ "epoch": 40.198083067092654,
+ "grad_norm": 0.050537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0289,
+ "step": 12582
+ },
+ {
+ "epoch": 40.201277955271564,
+ "grad_norm": 0.05517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0356,
+ "step": 12583
+ },
+ {
+ "epoch": 40.20447284345048,
+ "grad_norm": 0.053466796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0352,
+ "step": 12584
+ },
+ {
+ "epoch": 40.20766773162939,
+ "grad_norm": 0.056884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.019,
+ "step": 12585
+ },
+ {
+ "epoch": 40.21086261980831,
+ "grad_norm": 0.05419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.031,
+ "step": 12586
+ },
+ {
+ "epoch": 40.21405750798722,
+ "grad_norm": 0.05029296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0257,
+ "step": 12587
+ },
+ {
+ "epoch": 40.21725239616613,
+ "grad_norm": 0.05615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0279,
+ "step": 12588
+ },
+ {
+ "epoch": 40.22044728434505,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0306,
+ "step": 12589
+ },
+ {
+ "epoch": 40.22364217252396,
+ "grad_norm": 0.052490234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0317,
+ "step": 12590
+ },
+ {
+ "epoch": 40.22683706070288,
+ "grad_norm": 0.08349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.035,
+ "step": 12591
+ },
+ {
+ "epoch": 40.23003194888179,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.034,
+ "step": 12592
+ },
+ {
+ "epoch": 40.233226837060705,
+ "grad_norm": 0.046630859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0265,
+ "step": 12593
+ },
+ {
+ "epoch": 40.236421725239616,
+ "grad_norm": 0.053466796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0232,
+ "step": 12594
+ },
+ {
+ "epoch": 40.239616613418534,
+ "grad_norm": 0.050537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0345,
+ "step": 12595
+ },
+ {
+ "epoch": 40.242811501597444,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0355,
+ "step": 12596
+ },
+ {
+ "epoch": 40.246006389776355,
+ "grad_norm": 0.060791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.028,
+ "step": 12597
+ },
+ {
+ "epoch": 40.24920127795527,
+ "grad_norm": 0.0927734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0301,
+ "step": 12598
+ },
+ {
+ "epoch": 40.25239616613418,
+ "grad_norm": 0.04931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0294,
+ "step": 12599
+ },
+ {
+ "epoch": 40.2555910543131,
+ "grad_norm": 0.0908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0384,
+ "step": 12600
+ },
+ {
+ "epoch": 40.25878594249201,
+ "grad_norm": 0.05126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0198,
+ "step": 12601
+ },
+ {
+ "epoch": 40.26198083067093,
+ "grad_norm": 0.05908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0169,
+ "step": 12602
+ },
+ {
+ "epoch": 40.26517571884984,
+ "grad_norm": 0.10205078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0307,
+ "step": 12603
+ },
+ {
+ "epoch": 40.26837060702876,
+ "grad_norm": 0.09228515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0282,
+ "step": 12604
+ },
+ {
+ "epoch": 40.27156549520767,
+ "grad_norm": 0.10546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0366,
+ "step": 12605
+ },
+ {
+ "epoch": 40.27476038338658,
+ "grad_norm": 0.109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0364,
+ "step": 12606
+ },
+ {
+ "epoch": 40.277955271565496,
+ "grad_norm": 0.060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0308,
+ "step": 12607
+ },
+ {
+ "epoch": 40.281150159744406,
+ "grad_norm": 0.08447265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0452,
+ "step": 12608
+ },
+ {
+ "epoch": 40.284345047923324,
+ "grad_norm": 0.08642578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0336,
+ "step": 12609
+ },
+ {
+ "epoch": 40.287539936102235,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0274,
+ "step": 12610
+ },
+ {
+ "epoch": 40.29073482428115,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0309,
+ "step": 12611
+ },
+ {
+ "epoch": 40.29392971246006,
+ "grad_norm": 0.08544921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0287,
+ "step": 12612
+ },
+ {
+ "epoch": 40.29712460063898,
+ "grad_norm": 0.0595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0427,
+ "step": 12613
+ },
+ {
+ "epoch": 40.30031948881789,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0384,
+ "step": 12614
+ },
+ {
+ "epoch": 40.3035143769968,
+ "grad_norm": 0.054443359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0302,
+ "step": 12615
+ },
+ {
+ "epoch": 40.30670926517572,
+ "grad_norm": 0.095703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0431,
+ "step": 12616
+ },
+ {
+ "epoch": 40.30990415335463,
+ "grad_norm": 0.047119140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0389,
+ "step": 12617
+ },
+ {
+ "epoch": 40.31309904153355,
+ "grad_norm": 0.0595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0302,
+ "step": 12618
+ },
+ {
+ "epoch": 40.31629392971246,
+ "grad_norm": 0.04638671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0482,
+ "step": 12619
+ },
+ {
+ "epoch": 40.319488817891376,
+ "grad_norm": 0.05517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0361,
+ "step": 12620
+ },
+ {
+ "epoch": 40.322683706070286,
+ "grad_norm": 0.05078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0366,
+ "step": 12621
+ },
+ {
+ "epoch": 40.325878594249204,
+ "grad_norm": 0.04541015625,
+ "learning_rate": 0.0005,
+ "loss": 1.029,
+ "step": 12622
+ },
+ {
+ "epoch": 40.329073482428115,
+ "grad_norm": 0.048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0246,
+ "step": 12623
+ },
+ {
+ "epoch": 40.33226837060703,
+ "grad_norm": 0.04736328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0241,
+ "step": 12624
+ },
+ {
+ "epoch": 40.33546325878594,
+ "grad_norm": 0.05615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.03,
+ "step": 12625
+ },
+ {
+ "epoch": 40.33865814696485,
+ "grad_norm": 0.0458984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0158,
+ "step": 12626
+ },
+ {
+ "epoch": 40.34185303514377,
+ "grad_norm": 0.05810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0278,
+ "step": 12627
+ },
+ {
+ "epoch": 40.34504792332268,
+ "grad_norm": 0.048095703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0434,
+ "step": 12628
+ },
+ {
+ "epoch": 40.3482428115016,
+ "grad_norm": 0.05126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.033,
+ "step": 12629
+ },
+ {
+ "epoch": 40.35143769968051,
+ "grad_norm": 0.0498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0234,
+ "step": 12630
+ },
+ {
+ "epoch": 40.35463258785943,
+ "grad_norm": 0.059326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.027,
+ "step": 12631
+ },
+ {
+ "epoch": 40.35782747603834,
+ "grad_norm": 0.04931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0254,
+ "step": 12632
+ },
+ {
+ "epoch": 40.361022364217256,
+ "grad_norm": 0.0498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0367,
+ "step": 12633
+ },
+ {
+ "epoch": 40.364217252396166,
+ "grad_norm": 0.056396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.042,
+ "step": 12634
+ },
+ {
+ "epoch": 40.36741214057508,
+ "grad_norm": 0.0478515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0396,
+ "step": 12635
+ },
+ {
+ "epoch": 40.370607028753994,
+ "grad_norm": 0.052978515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0242,
+ "step": 12636
+ },
+ {
+ "epoch": 40.373801916932905,
+ "grad_norm": 0.054931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0312,
+ "step": 12637
+ },
+ {
+ "epoch": 40.37699680511182,
+ "grad_norm": 0.05810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0357,
+ "step": 12638
+ },
+ {
+ "epoch": 40.38019169329073,
+ "grad_norm": 0.049072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0368,
+ "step": 12639
+ },
+ {
+ "epoch": 40.38338658146965,
+ "grad_norm": 0.0869140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0289,
+ "step": 12640
+ },
+ {
+ "epoch": 40.38658146964856,
+ "grad_norm": 0.0849609375,
+ "learning_rate": 0.0005,
+ "loss": 1.034,
+ "step": 12641
+ },
+ {
+ "epoch": 40.38977635782748,
+ "grad_norm": 0.059814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0333,
+ "step": 12642
+ },
+ {
+ "epoch": 40.39297124600639,
+ "grad_norm": 0.09228515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0305,
+ "step": 12643
+ },
+ {
+ "epoch": 40.3961661341853,
+ "grad_norm": 0.049072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0372,
+ "step": 12644
+ },
+ {
+ "epoch": 40.39936102236422,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0325,
+ "step": 12645
+ },
+ {
+ "epoch": 40.40255591054313,
+ "grad_norm": 0.058349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.044,
+ "step": 12646
+ },
+ {
+ "epoch": 40.405750798722046,
+ "grad_norm": 0.09130859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0118,
+ "step": 12647
+ },
+ {
+ "epoch": 40.40894568690096,
+ "grad_norm": 0.08447265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0285,
+ "step": 12648
+ },
+ {
+ "epoch": 40.412140575079874,
+ "grad_norm": 0.1259765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0253,
+ "step": 12649
+ },
+ {
+ "epoch": 40.415335463258785,
+ "grad_norm": 0.05810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0223,
+ "step": 12650
+ },
+ {
+ "epoch": 40.4185303514377,
+ "grad_norm": 0.123046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0366,
+ "step": 12651
+ },
+ {
+ "epoch": 40.42172523961661,
+ "grad_norm": 0.041015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0212,
+ "step": 12652
+ },
+ {
+ "epoch": 40.424920127795524,
+ "grad_norm": 0.09326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0221,
+ "step": 12653
+ },
+ {
+ "epoch": 40.42811501597444,
+ "grad_norm": 0.051025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0384,
+ "step": 12654
+ },
+ {
+ "epoch": 40.43130990415335,
+ "grad_norm": 0.10107421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0438,
+ "step": 12655
+ },
+ {
+ "epoch": 40.43450479233227,
+ "grad_norm": 0.058837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0373,
+ "step": 12656
+ },
+ {
+ "epoch": 40.43769968051118,
+ "grad_norm": 0.1240234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0295,
+ "step": 12657
+ },
+ {
+ "epoch": 40.4408945686901,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0418,
+ "step": 12658
+ },
+ {
+ "epoch": 40.44408945686901,
+ "grad_norm": 0.10107421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0485,
+ "step": 12659
+ },
+ {
+ "epoch": 40.447284345047926,
+ "grad_norm": 0.0859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0314,
+ "step": 12660
+ },
+ {
+ "epoch": 40.45047923322684,
+ "grad_norm": 0.1025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0353,
+ "step": 12661
+ },
+ {
+ "epoch": 40.453674121405754,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0323,
+ "step": 12662
+ },
+ {
+ "epoch": 40.456869009584665,
+ "grad_norm": 0.0859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0359,
+ "step": 12663
+ },
+ {
+ "epoch": 40.460063897763575,
+ "grad_norm": 0.052001953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0336,
+ "step": 12664
+ },
+ {
+ "epoch": 40.46325878594249,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0323,
+ "step": 12665
+ },
+ {
+ "epoch": 40.466453674121404,
+ "grad_norm": 0.05810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0258,
+ "step": 12666
+ },
+ {
+ "epoch": 40.46964856230032,
+ "grad_norm": 0.055419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0495,
+ "step": 12667
+ },
+ {
+ "epoch": 40.47284345047923,
+ "grad_norm": 0.09033203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0195,
+ "step": 12668
+ },
+ {
+ "epoch": 40.47603833865815,
+ "grad_norm": 0.053466796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0292,
+ "step": 12669
+ },
+ {
+ "epoch": 40.47923322683706,
+ "grad_norm": 0.091796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0181,
+ "step": 12670
+ },
+ {
+ "epoch": 40.48242811501598,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0399,
+ "step": 12671
+ },
+ {
+ "epoch": 40.48562300319489,
+ "grad_norm": 0.08642578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0219,
+ "step": 12672
+ },
+ {
+ "epoch": 40.4888178913738,
+ "grad_norm": 0.10107421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0417,
+ "step": 12673
+ },
+ {
+ "epoch": 40.49201277955272,
+ "grad_norm": 0.12890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0347,
+ "step": 12674
+ },
+ {
+ "epoch": 40.49520766773163,
+ "grad_norm": 0.061279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0305,
+ "step": 12675
+ },
+ {
+ "epoch": 40.498402555910545,
+ "grad_norm": 0.0927734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0192,
+ "step": 12676
+ },
+ {
+ "epoch": 40.501597444089455,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0297,
+ "step": 12677
+ },
+ {
+ "epoch": 40.50479233226837,
+ "grad_norm": 0.08544921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0429,
+ "step": 12678
+ },
+ {
+ "epoch": 40.50798722044728,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0324,
+ "step": 12679
+ },
+ {
+ "epoch": 40.5111821086262,
+ "grad_norm": 0.11669921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0226,
+ "step": 12680
+ },
+ {
+ "epoch": 40.51437699680511,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0433,
+ "step": 12681
+ },
+ {
+ "epoch": 40.51757188498402,
+ "grad_norm": 0.0947265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0323,
+ "step": 12682
+ },
+ {
+ "epoch": 40.52076677316294,
+ "grad_norm": 0.0546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0381,
+ "step": 12683
+ },
+ {
+ "epoch": 40.52396166134185,
+ "grad_norm": 0.0810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0273,
+ "step": 12684
+ },
+ {
+ "epoch": 40.52715654952077,
+ "grad_norm": 0.0859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0374,
+ "step": 12685
+ },
+ {
+ "epoch": 40.53035143769968,
+ "grad_norm": 0.0849609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0358,
+ "step": 12686
+ },
+ {
+ "epoch": 40.533546325878596,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0305,
+ "step": 12687
+ },
+ {
+ "epoch": 40.53674121405751,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0359,
+ "step": 12688
+ },
+ {
+ "epoch": 40.539936102236425,
+ "grad_norm": 0.054931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.027,
+ "step": 12689
+ },
+ {
+ "epoch": 40.543130990415335,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0359,
+ "step": 12690
+ },
+ {
+ "epoch": 40.546325878594246,
+ "grad_norm": 0.05224609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0268,
+ "step": 12691
+ },
+ {
+ "epoch": 40.54952076677316,
+ "grad_norm": 0.04833984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0273,
+ "step": 12692
+ },
+ {
+ "epoch": 40.552715654952074,
+ "grad_norm": 0.0537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0272,
+ "step": 12693
+ },
+ {
+ "epoch": 40.55591054313099,
+ "grad_norm": 0.05517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0367,
+ "step": 12694
+ },
+ {
+ "epoch": 40.5591054313099,
+ "grad_norm": 0.0517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0336,
+ "step": 12695
+ },
+ {
+ "epoch": 40.56230031948882,
+ "grad_norm": 0.056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0219,
+ "step": 12696
+ },
+ {
+ "epoch": 40.56549520766773,
+ "grad_norm": 0.053466796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0325,
+ "step": 12697
+ },
+ {
+ "epoch": 40.56869009584665,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0306,
+ "step": 12698
+ },
+ {
+ "epoch": 40.57188498402556,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0202,
+ "step": 12699
+ },
+ {
+ "epoch": 40.575079872204476,
+ "grad_norm": 0.05224609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0271,
+ "step": 12700
+ },
+ {
+ "epoch": 40.57827476038339,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0357,
+ "step": 12701
+ },
+ {
+ "epoch": 40.5814696485623,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0392,
+ "step": 12702
+ },
+ {
+ "epoch": 40.584664536741215,
+ "grad_norm": 0.1123046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0306,
+ "step": 12703
+ },
+ {
+ "epoch": 40.587859424920126,
+ "grad_norm": 0.08544921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0328,
+ "step": 12704
+ },
+ {
+ "epoch": 40.59105431309904,
+ "grad_norm": 0.058837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0387,
+ "step": 12705
+ },
+ {
+ "epoch": 40.594249201277954,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0311,
+ "step": 12706
+ },
+ {
+ "epoch": 40.59744408945687,
+ "grad_norm": 0.05712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0274,
+ "step": 12707
+ },
+ {
+ "epoch": 40.60063897763578,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.029,
+ "step": 12708
+ },
+ {
+ "epoch": 40.6038338658147,
+ "grad_norm": 0.044921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0353,
+ "step": 12709
+ },
+ {
+ "epoch": 40.60702875399361,
+ "grad_norm": 0.107421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0355,
+ "step": 12710
+ },
+ {
+ "epoch": 40.61022364217252,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0449,
+ "step": 12711
+ },
+ {
+ "epoch": 40.61341853035144,
+ "grad_norm": 0.057861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0219,
+ "step": 12712
+ },
+ {
+ "epoch": 40.61661341853035,
+ "grad_norm": 0.05517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0342,
+ "step": 12713
+ },
+ {
+ "epoch": 40.61980830670927,
+ "grad_norm": 0.05517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0333,
+ "step": 12714
+ },
+ {
+ "epoch": 40.62300319488818,
+ "grad_norm": 0.049560546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0275,
+ "step": 12715
+ },
+ {
+ "epoch": 40.626198083067095,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0411,
+ "step": 12716
+ },
+ {
+ "epoch": 40.629392971246006,
+ "grad_norm": 0.060791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0467,
+ "step": 12717
+ },
+ {
+ "epoch": 40.63258785942492,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.024,
+ "step": 12718
+ },
+ {
+ "epoch": 40.635782747603834,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0283,
+ "step": 12719
+ },
+ {
+ "epoch": 40.638977635782744,
+ "grad_norm": 0.0830078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0298,
+ "step": 12720
+ },
+ {
+ "epoch": 40.64217252396166,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0253,
+ "step": 12721
+ },
+ {
+ "epoch": 40.64536741214057,
+ "grad_norm": 0.1396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0391,
+ "step": 12722
+ },
+ {
+ "epoch": 40.64856230031949,
+ "grad_norm": 0.08154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0226,
+ "step": 12723
+ },
+ {
+ "epoch": 40.6517571884984,
+ "grad_norm": 0.058837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0357,
+ "step": 12724
+ },
+ {
+ "epoch": 40.65495207667732,
+ "grad_norm": 0.1220703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0308,
+ "step": 12725
+ },
+ {
+ "epoch": 40.65814696485623,
+ "grad_norm": 0.06201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0233,
+ "step": 12726
+ },
+ {
+ "epoch": 40.66134185303515,
+ "grad_norm": 0.058837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0268,
+ "step": 12727
+ },
+ {
+ "epoch": 40.66453674121406,
+ "grad_norm": 0.09619140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0335,
+ "step": 12728
+ },
+ {
+ "epoch": 40.66773162939297,
+ "grad_norm": 0.061767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0393,
+ "step": 12729
+ },
+ {
+ "epoch": 40.670926517571885,
+ "grad_norm": 0.0966796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0297,
+ "step": 12730
+ },
+ {
+ "epoch": 40.674121405750796,
+ "grad_norm": 0.09130859375,
+ "learning_rate": 0.0005,
+ "loss": 1.035,
+ "step": 12731
+ },
+ {
+ "epoch": 40.677316293929714,
+ "grad_norm": 0.099609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0296,
+ "step": 12732
+ },
+ {
+ "epoch": 40.680511182108624,
+ "grad_norm": 0.10791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0411,
+ "step": 12733
+ },
+ {
+ "epoch": 40.68370607028754,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.04,
+ "step": 12734
+ },
+ {
+ "epoch": 40.68690095846645,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0206,
+ "step": 12735
+ },
+ {
+ "epoch": 40.69009584664537,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.038,
+ "step": 12736
+ },
+ {
+ "epoch": 40.69329073482428,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.038,
+ "step": 12737
+ },
+ {
+ "epoch": 40.6964856230032,
+ "grad_norm": 0.061767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.02,
+ "step": 12738
+ },
+ {
+ "epoch": 40.69968051118211,
+ "grad_norm": 0.053466796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0353,
+ "step": 12739
+ },
+ {
+ "epoch": 40.70287539936102,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0348,
+ "step": 12740
+ },
+ {
+ "epoch": 40.70607028753994,
+ "grad_norm": 0.0556640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0218,
+ "step": 12741
+ },
+ {
+ "epoch": 40.70926517571885,
+ "grad_norm": 0.0478515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0376,
+ "step": 12742
+ },
+ {
+ "epoch": 40.712460063897765,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0311,
+ "step": 12743
+ },
+ {
+ "epoch": 40.715654952076676,
+ "grad_norm": 0.056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0325,
+ "step": 12744
+ },
+ {
+ "epoch": 40.718849840255594,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0279,
+ "step": 12745
+ },
+ {
+ "epoch": 40.722044728434504,
+ "grad_norm": 0.06103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0245,
+ "step": 12746
+ },
+ {
+ "epoch": 40.72523961661342,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0325,
+ "step": 12747
+ },
+ {
+ "epoch": 40.72843450479233,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0322,
+ "step": 12748
+ },
+ {
+ "epoch": 40.73162939297124,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0318,
+ "step": 12749
+ },
+ {
+ "epoch": 40.73482428115016,
+ "grad_norm": 0.056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0268,
+ "step": 12750
+ },
+ {
+ "epoch": 40.73801916932907,
+ "grad_norm": 0.060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.025,
+ "step": 12751
+ },
+ {
+ "epoch": 40.74121405750799,
+ "grad_norm": 0.08935546875,
+ "learning_rate": 0.0005,
+ "loss": 1.035,
+ "step": 12752
+ },
+ {
+ "epoch": 40.7444089456869,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0327,
+ "step": 12753
+ },
+ {
+ "epoch": 40.74760383386582,
+ "grad_norm": 0.08544921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0326,
+ "step": 12754
+ },
+ {
+ "epoch": 40.75079872204473,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.032,
+ "step": 12755
+ },
+ {
+ "epoch": 40.753993610223645,
+ "grad_norm": 0.051025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0412,
+ "step": 12756
+ },
+ {
+ "epoch": 40.757188498402556,
+ "grad_norm": 0.059814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0345,
+ "step": 12757
+ },
+ {
+ "epoch": 40.760383386581466,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0288,
+ "step": 12758
+ },
+ {
+ "epoch": 40.763578274760384,
+ "grad_norm": 0.0869140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0393,
+ "step": 12759
+ },
+ {
+ "epoch": 40.766773162939295,
+ "grad_norm": 0.0615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0377,
+ "step": 12760
+ },
+ {
+ "epoch": 40.76996805111821,
+ "grad_norm": 0.11083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0315,
+ "step": 12761
+ },
+ {
+ "epoch": 40.77316293929712,
+ "grad_norm": 0.05517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0257,
+ "step": 12762
+ },
+ {
+ "epoch": 40.77635782747604,
+ "grad_norm": 0.09326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0434,
+ "step": 12763
+ },
+ {
+ "epoch": 40.77955271565495,
+ "grad_norm": 0.05126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0418,
+ "step": 12764
+ },
+ {
+ "epoch": 40.78274760383387,
+ "grad_norm": 0.0478515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0359,
+ "step": 12765
+ },
+ {
+ "epoch": 40.78594249201278,
+ "grad_norm": 0.058837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0368,
+ "step": 12766
+ },
+ {
+ "epoch": 40.78913738019169,
+ "grad_norm": 0.047119140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0385,
+ "step": 12767
+ },
+ {
+ "epoch": 40.79233226837061,
+ "grad_norm": 0.060791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0314,
+ "step": 12768
+ },
+ {
+ "epoch": 40.79552715654952,
+ "grad_norm": 0.043212890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0345,
+ "step": 12769
+ },
+ {
+ "epoch": 40.798722044728436,
+ "grad_norm": 0.061767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0352,
+ "step": 12770
+ },
+ {
+ "epoch": 40.801916932907346,
+ "grad_norm": 0.05419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0287,
+ "step": 12771
+ },
+ {
+ "epoch": 40.805111821086264,
+ "grad_norm": 0.0869140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0373,
+ "step": 12772
+ },
+ {
+ "epoch": 40.808306709265175,
+ "grad_norm": 0.047119140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0354,
+ "step": 12773
+ },
+ {
+ "epoch": 40.81150159744409,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0353,
+ "step": 12774
+ },
+ {
+ "epoch": 40.814696485623,
+ "grad_norm": 0.050537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0263,
+ "step": 12775
+ },
+ {
+ "epoch": 40.81789137380191,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0406,
+ "step": 12776
+ },
+ {
+ "epoch": 40.82108626198083,
+ "grad_norm": 0.0498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0296,
+ "step": 12777
+ },
+ {
+ "epoch": 40.82428115015974,
+ "grad_norm": 0.050048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0271,
+ "step": 12778
+ },
+ {
+ "epoch": 40.82747603833866,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0293,
+ "step": 12779
+ },
+ {
+ "epoch": 40.83067092651757,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0243,
+ "step": 12780
+ },
+ {
+ "epoch": 40.83386581469649,
+ "grad_norm": 0.05615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0347,
+ "step": 12781
+ },
+ {
+ "epoch": 40.8370607028754,
+ "grad_norm": 0.049072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.036,
+ "step": 12782
+ },
+ {
+ "epoch": 40.840255591054316,
+ "grad_norm": 0.052490234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0299,
+ "step": 12783
+ },
+ {
+ "epoch": 40.843450479233226,
+ "grad_norm": 0.05908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0353,
+ "step": 12784
+ },
+ {
+ "epoch": 40.846645367412144,
+ "grad_norm": 0.046630859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0286,
+ "step": 12785
+ },
+ {
+ "epoch": 40.849840255591054,
+ "grad_norm": 0.052490234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0267,
+ "step": 12786
+ },
+ {
+ "epoch": 40.853035143769965,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.051,
+ "step": 12787
+ },
+ {
+ "epoch": 40.85623003194888,
+ "grad_norm": 0.05224609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0368,
+ "step": 12788
+ },
+ {
+ "epoch": 40.85942492012779,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0408,
+ "step": 12789
+ },
+ {
+ "epoch": 40.86261980830671,
+ "grad_norm": 0.06201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0435,
+ "step": 12790
+ },
+ {
+ "epoch": 40.86581469648562,
+ "grad_norm": 0.050537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0292,
+ "step": 12791
+ },
+ {
+ "epoch": 40.86900958466454,
+ "grad_norm": 0.04638671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0453,
+ "step": 12792
+ },
+ {
+ "epoch": 40.87220447284345,
+ "grad_norm": 0.04736328125,
+ "learning_rate": 0.0005,
+ "loss": 1.04,
+ "step": 12793
+ },
+ {
+ "epoch": 40.87539936102237,
+ "grad_norm": 0.052001953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0361,
+ "step": 12794
+ },
+ {
+ "epoch": 40.87859424920128,
+ "grad_norm": 0.05224609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0405,
+ "step": 12795
+ },
+ {
+ "epoch": 40.88178913738019,
+ "grad_norm": 0.05029296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0299,
+ "step": 12796
+ },
+ {
+ "epoch": 40.884984025559106,
+ "grad_norm": 0.0517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0336,
+ "step": 12797
+ },
+ {
+ "epoch": 40.88817891373802,
+ "grad_norm": 0.044189453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0253,
+ "step": 12798
+ },
+ {
+ "epoch": 40.891373801916934,
+ "grad_norm": 0.059326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0319,
+ "step": 12799
+ },
+ {
+ "epoch": 40.894568690095845,
+ "grad_norm": 0.05322265625,
+ "learning_rate": 0.0005,
+ "loss": 1.023,
+ "step": 12800
+ },
+ {
+ "epoch": 40.89776357827476,
+ "grad_norm": 0.057373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0431,
+ "step": 12801
+ },
+ {
+ "epoch": 40.90095846645367,
+ "grad_norm": 0.08154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.04,
+ "step": 12802
+ },
+ {
+ "epoch": 40.90415335463259,
+ "grad_norm": 0.05712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0277,
+ "step": 12803
+ },
+ {
+ "epoch": 40.9073482428115,
+ "grad_norm": 0.0927734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0348,
+ "step": 12804
+ },
+ {
+ "epoch": 40.91054313099041,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0306,
+ "step": 12805
+ },
+ {
+ "epoch": 40.91373801916933,
+ "grad_norm": 0.049560546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0275,
+ "step": 12806
+ },
+ {
+ "epoch": 40.91693290734824,
+ "grad_norm": 0.049072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0203,
+ "step": 12807
+ },
+ {
+ "epoch": 40.92012779552716,
+ "grad_norm": 0.061279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.033,
+ "step": 12808
+ },
+ {
+ "epoch": 40.92332268370607,
+ "grad_norm": 0.060791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0304,
+ "step": 12809
+ },
+ {
+ "epoch": 40.926517571884986,
+ "grad_norm": 0.049560546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0374,
+ "step": 12810
+ },
+ {
+ "epoch": 40.9297124600639,
+ "grad_norm": 0.051513671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0313,
+ "step": 12811
+ },
+ {
+ "epoch": 40.932907348242814,
+ "grad_norm": 0.087890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0379,
+ "step": 12812
+ },
+ {
+ "epoch": 40.936102236421725,
+ "grad_norm": 0.0517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0247,
+ "step": 12813
+ },
+ {
+ "epoch": 40.93929712460064,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0344,
+ "step": 12814
+ },
+ {
+ "epoch": 40.94249201277955,
+ "grad_norm": 0.045166015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0292,
+ "step": 12815
+ },
+ {
+ "epoch": 40.945686900958464,
+ "grad_norm": 0.06201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0263,
+ "step": 12816
+ },
+ {
+ "epoch": 40.94888178913738,
+ "grad_norm": 0.057373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0163,
+ "step": 12817
+ },
+ {
+ "epoch": 40.95207667731629,
+ "grad_norm": 0.103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0385,
+ "step": 12818
+ },
+ {
+ "epoch": 40.95527156549521,
+ "grad_norm": 0.053955078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0294,
+ "step": 12819
+ },
+ {
+ "epoch": 40.95846645367412,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.036,
+ "step": 12820
+ },
+ {
+ "epoch": 40.96166134185304,
+ "grad_norm": 0.056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0403,
+ "step": 12821
+ },
+ {
+ "epoch": 40.96485623003195,
+ "grad_norm": 0.061279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.034,
+ "step": 12822
+ },
+ {
+ "epoch": 40.968051118210866,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0237,
+ "step": 12823
+ },
+ {
+ "epoch": 40.97124600638978,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0306,
+ "step": 12824
+ },
+ {
+ "epoch": 40.97444089456869,
+ "grad_norm": 0.0986328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0322,
+ "step": 12825
+ },
+ {
+ "epoch": 40.977635782747605,
+ "grad_norm": 0.052001953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0385,
+ "step": 12826
+ },
+ {
+ "epoch": 40.980830670926515,
+ "grad_norm": 0.05224609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0372,
+ "step": 12827
+ },
+ {
+ "epoch": 40.98402555910543,
+ "grad_norm": 0.045166015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0441,
+ "step": 12828
+ },
+ {
+ "epoch": 40.98722044728434,
+ "grad_norm": 0.05322265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0337,
+ "step": 12829
+ },
+ {
+ "epoch": 40.99041533546326,
+ "grad_norm": 0.061767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0353,
+ "step": 12830
+ },
+ {
+ "epoch": 40.99361022364217,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0304,
+ "step": 12831
+ },
+ {
+ "epoch": 40.99680511182109,
+ "grad_norm": 0.0830078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0234,
+ "step": 12832
+ },
+ {
+ "epoch": 41.0,
+ "grad_norm": 0.06005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0245,
+ "step": 12833
+ },
+ {
+ "epoch": 41.00319488817891,
+ "grad_norm": 0.0966796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0424,
+ "step": 12834
+ },
+ {
+ "epoch": 41.00638977635783,
+ "grad_norm": 0.055419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0369,
+ "step": 12835
+ },
+ {
+ "epoch": 41.00958466453674,
+ "grad_norm": 0.09033203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0198,
+ "step": 12836
+ },
+ {
+ "epoch": 41.01277955271566,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0233,
+ "step": 12837
+ },
+ {
+ "epoch": 41.01597444089457,
+ "grad_norm": 0.08251953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0352,
+ "step": 12838
+ },
+ {
+ "epoch": 41.019169329073485,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0289,
+ "step": 12839
+ },
+ {
+ "epoch": 41.022364217252395,
+ "grad_norm": 0.05126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.029,
+ "step": 12840
+ },
+ {
+ "epoch": 41.02555910543131,
+ "grad_norm": 0.0546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0346,
+ "step": 12841
+ },
+ {
+ "epoch": 41.02875399361022,
+ "grad_norm": 0.045166015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0332,
+ "step": 12842
+ },
+ {
+ "epoch": 41.031948881789134,
+ "grad_norm": 0.0537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0338,
+ "step": 12843
+ },
+ {
+ "epoch": 41.03514376996805,
+ "grad_norm": 0.048095703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0249,
+ "step": 12844
+ },
+ {
+ "epoch": 41.03833865814696,
+ "grad_norm": 0.045654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0306,
+ "step": 12845
+ },
+ {
+ "epoch": 41.04153354632588,
+ "grad_norm": 0.06201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0281,
+ "step": 12846
+ },
+ {
+ "epoch": 41.04472843450479,
+ "grad_norm": 0.046875,
+ "learning_rate": 0.0005,
+ "loss": 1.029,
+ "step": 12847
+ },
+ {
+ "epoch": 41.04792332268371,
+ "grad_norm": 0.05615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.03,
+ "step": 12848
+ },
+ {
+ "epoch": 41.05111821086262,
+ "grad_norm": 0.08056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0374,
+ "step": 12849
+ },
+ {
+ "epoch": 41.054313099041536,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0305,
+ "step": 12850
+ },
+ {
+ "epoch": 41.05750798722045,
+ "grad_norm": 0.138671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0283,
+ "step": 12851
+ },
+ {
+ "epoch": 41.06070287539936,
+ "grad_norm": 0.09228515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0381,
+ "step": 12852
+ },
+ {
+ "epoch": 41.063897763578275,
+ "grad_norm": 0.06005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0295,
+ "step": 12853
+ },
+ {
+ "epoch": 41.067092651757186,
+ "grad_norm": 0.1845703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0251,
+ "step": 12854
+ },
+ {
+ "epoch": 41.0702875399361,
+ "grad_norm": 0.08056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0329,
+ "step": 12855
+ },
+ {
+ "epoch": 41.073482428115014,
+ "grad_norm": 0.16015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0314,
+ "step": 12856
+ },
+ {
+ "epoch": 41.07667731629393,
+ "grad_norm": 0.1337890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0357,
+ "step": 12857
+ },
+ {
+ "epoch": 41.07987220447284,
+ "grad_norm": 0.10888671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0359,
+ "step": 12858
+ },
+ {
+ "epoch": 41.08306709265176,
+ "grad_norm": 0.1572265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0292,
+ "step": 12859
+ },
+ {
+ "epoch": 41.08626198083067,
+ "grad_norm": 0.08740234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0353,
+ "step": 12860
+ },
+ {
+ "epoch": 41.08945686900959,
+ "grad_norm": 0.08837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0347,
+ "step": 12861
+ },
+ {
+ "epoch": 41.0926517571885,
+ "grad_norm": 0.09716796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0366,
+ "step": 12862
+ },
+ {
+ "epoch": 41.09584664536741,
+ "grad_norm": 0.087890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0376,
+ "step": 12863
+ },
+ {
+ "epoch": 41.09904153354633,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0293,
+ "step": 12864
+ },
+ {
+ "epoch": 41.10223642172524,
+ "grad_norm": 0.10986328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0322,
+ "step": 12865
+ },
+ {
+ "epoch": 41.105431309904155,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0222,
+ "step": 12866
+ },
+ {
+ "epoch": 41.108626198083066,
+ "grad_norm": 0.12109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0364,
+ "step": 12867
+ },
+ {
+ "epoch": 41.11182108626198,
+ "grad_norm": 0.0810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0282,
+ "step": 12868
+ },
+ {
+ "epoch": 41.115015974440894,
+ "grad_norm": 0.08251953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0249,
+ "step": 12869
+ },
+ {
+ "epoch": 41.11821086261981,
+ "grad_norm": 0.06201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0447,
+ "step": 12870
+ },
+ {
+ "epoch": 41.12140575079872,
+ "grad_norm": 0.08154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.033,
+ "step": 12871
+ },
+ {
+ "epoch": 41.12460063897763,
+ "grad_norm": 0.0830078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0272,
+ "step": 12872
+ },
+ {
+ "epoch": 41.12779552715655,
+ "grad_norm": 0.0537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0342,
+ "step": 12873
+ },
+ {
+ "epoch": 41.13099041533546,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0266,
+ "step": 12874
+ },
+ {
+ "epoch": 41.13418530351438,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0223,
+ "step": 12875
+ },
+ {
+ "epoch": 41.13738019169329,
+ "grad_norm": 0.0498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.037,
+ "step": 12876
+ },
+ {
+ "epoch": 41.14057507987221,
+ "grad_norm": 0.061767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0359,
+ "step": 12877
+ },
+ {
+ "epoch": 41.14376996805112,
+ "grad_norm": 0.05517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0311,
+ "step": 12878
+ },
+ {
+ "epoch": 41.146964856230035,
+ "grad_norm": 0.08447265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0379,
+ "step": 12879
+ },
+ {
+ "epoch": 41.150159744408946,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0344,
+ "step": 12880
+ },
+ {
+ "epoch": 41.153354632587856,
+ "grad_norm": 0.091796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0281,
+ "step": 12881
+ },
+ {
+ "epoch": 41.156549520766774,
+ "grad_norm": 0.0576171875,
+ "learning_rate": 0.0005,
+ "loss": 1.043,
+ "step": 12882
+ },
+ {
+ "epoch": 41.159744408945684,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0305,
+ "step": 12883
+ },
+ {
+ "epoch": 41.1629392971246,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0313,
+ "step": 12884
+ },
+ {
+ "epoch": 41.16613418530351,
+ "grad_norm": 0.0849609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0362,
+ "step": 12885
+ },
+ {
+ "epoch": 41.16932907348243,
+ "grad_norm": 0.05859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0241,
+ "step": 12886
+ },
+ {
+ "epoch": 41.17252396166134,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0237,
+ "step": 12887
+ },
+ {
+ "epoch": 41.17571884984026,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0381,
+ "step": 12888
+ },
+ {
+ "epoch": 41.17891373801917,
+ "grad_norm": 0.11376953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0352,
+ "step": 12889
+ },
+ {
+ "epoch": 41.18210862619808,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0368,
+ "step": 12890
+ },
+ {
+ "epoch": 41.185303514377,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0428,
+ "step": 12891
+ },
+ {
+ "epoch": 41.18849840255591,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0164,
+ "step": 12892
+ },
+ {
+ "epoch": 41.191693290734825,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0377,
+ "step": 12893
+ },
+ {
+ "epoch": 41.194888178913736,
+ "grad_norm": 0.042236328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0292,
+ "step": 12894
+ },
+ {
+ "epoch": 41.198083067092654,
+ "grad_norm": 0.056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0383,
+ "step": 12895
+ },
+ {
+ "epoch": 41.201277955271564,
+ "grad_norm": 0.05419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0289,
+ "step": 12896
+ },
+ {
+ "epoch": 41.20447284345048,
+ "grad_norm": 0.05908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.021,
+ "step": 12897
+ },
+ {
+ "epoch": 41.20766773162939,
+ "grad_norm": 0.060302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0363,
+ "step": 12898
+ },
+ {
+ "epoch": 41.21086261980831,
+ "grad_norm": 0.052490234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0212,
+ "step": 12899
+ },
+ {
+ "epoch": 41.21405750798722,
+ "grad_norm": 0.060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0321,
+ "step": 12900
+ },
+ {
+ "epoch": 41.21725239616613,
+ "grad_norm": 0.052490234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0335,
+ "step": 12901
+ },
+ {
+ "epoch": 41.22044728434505,
+ "grad_norm": 0.08154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0302,
+ "step": 12902
+ },
+ {
+ "epoch": 41.22364217252396,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0394,
+ "step": 12903
+ },
+ {
+ "epoch": 41.22683706070288,
+ "grad_norm": 0.04931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.035,
+ "step": 12904
+ },
+ {
+ "epoch": 41.23003194888179,
+ "grad_norm": 0.056396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0255,
+ "step": 12905
+ },
+ {
+ "epoch": 41.233226837060705,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0437,
+ "step": 12906
+ },
+ {
+ "epoch": 41.236421725239616,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0356,
+ "step": 12907
+ },
+ {
+ "epoch": 41.239616613418534,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0229,
+ "step": 12908
+ },
+ {
+ "epoch": 41.242811501597444,
+ "grad_norm": 0.08349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0341,
+ "step": 12909
+ },
+ {
+ "epoch": 41.246006389776355,
+ "grad_norm": 0.054443359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0262,
+ "step": 12910
+ },
+ {
+ "epoch": 41.24920127795527,
+ "grad_norm": 0.05615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0368,
+ "step": 12911
+ },
+ {
+ "epoch": 41.25239616613418,
+ "grad_norm": 0.048583984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0407,
+ "step": 12912
+ },
+ {
+ "epoch": 41.2555910543131,
+ "grad_norm": 0.057373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0269,
+ "step": 12913
+ },
+ {
+ "epoch": 41.25878594249201,
+ "grad_norm": 0.0546875,
+ "learning_rate": 0.0005,
+ "loss": 1.032,
+ "step": 12914
+ },
+ {
+ "epoch": 41.26198083067093,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0328,
+ "step": 12915
+ },
+ {
+ "epoch": 41.26517571884984,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0183,
+ "step": 12916
+ },
+ {
+ "epoch": 41.26837060702876,
+ "grad_norm": 0.05029296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0259,
+ "step": 12917
+ },
+ {
+ "epoch": 41.27156549520767,
+ "grad_norm": 0.051513671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0352,
+ "step": 12918
+ },
+ {
+ "epoch": 41.27476038338658,
+ "grad_norm": 0.057373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0271,
+ "step": 12919
+ },
+ {
+ "epoch": 41.277955271565496,
+ "grad_norm": 0.0595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0343,
+ "step": 12920
+ },
+ {
+ "epoch": 41.281150159744406,
+ "grad_norm": 0.05615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.029,
+ "step": 12921
+ },
+ {
+ "epoch": 41.284345047923324,
+ "grad_norm": 0.052001953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0315,
+ "step": 12922
+ },
+ {
+ "epoch": 41.287539936102235,
+ "grad_norm": 0.05517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0381,
+ "step": 12923
+ },
+ {
+ "epoch": 41.29073482428115,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0326,
+ "step": 12924
+ },
+ {
+ "epoch": 41.29392971246006,
+ "grad_norm": 0.06005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0357,
+ "step": 12925
+ },
+ {
+ "epoch": 41.29712460063898,
+ "grad_norm": 0.051513671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0233,
+ "step": 12926
+ },
+ {
+ "epoch": 41.30031948881789,
+ "grad_norm": 0.055908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0335,
+ "step": 12927
+ },
+ {
+ "epoch": 41.3035143769968,
+ "grad_norm": 0.0556640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0411,
+ "step": 12928
+ },
+ {
+ "epoch": 41.30670926517572,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0326,
+ "step": 12929
+ },
+ {
+ "epoch": 41.30990415335463,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0279,
+ "step": 12930
+ },
+ {
+ "epoch": 41.31309904153355,
+ "grad_norm": 0.091796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0258,
+ "step": 12931
+ },
+ {
+ "epoch": 41.31629392971246,
+ "grad_norm": 0.04736328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0308,
+ "step": 12932
+ },
+ {
+ "epoch": 41.319488817891376,
+ "grad_norm": 0.050537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0276,
+ "step": 12933
+ },
+ {
+ "epoch": 41.322683706070286,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0221,
+ "step": 12934
+ },
+ {
+ "epoch": 41.325878594249204,
+ "grad_norm": 0.046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0322,
+ "step": 12935
+ },
+ {
+ "epoch": 41.329073482428115,
+ "grad_norm": 0.0947265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0268,
+ "step": 12936
+ },
+ {
+ "epoch": 41.33226837060703,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0231,
+ "step": 12937
+ },
+ {
+ "epoch": 41.33546325878594,
+ "grad_norm": 0.0869140625,
+ "learning_rate": 0.0005,
+ "loss": 1.026,
+ "step": 12938
+ },
+ {
+ "epoch": 41.33865814696485,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0316,
+ "step": 12939
+ },
+ {
+ "epoch": 41.34185303514377,
+ "grad_norm": 0.05078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0279,
+ "step": 12940
+ },
+ {
+ "epoch": 41.34504792332268,
+ "grad_norm": 0.0947265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0392,
+ "step": 12941
+ },
+ {
+ "epoch": 41.3482428115016,
+ "grad_norm": 0.05224609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0349,
+ "step": 12942
+ },
+ {
+ "epoch": 41.35143769968051,
+ "grad_norm": 0.08056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0323,
+ "step": 12943
+ },
+ {
+ "epoch": 41.35463258785943,
+ "grad_norm": 0.052978515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0391,
+ "step": 12944
+ },
+ {
+ "epoch": 41.35782747603834,
+ "grad_norm": 0.1005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0261,
+ "step": 12945
+ },
+ {
+ "epoch": 41.361022364217256,
+ "grad_norm": 0.123046875,
+ "learning_rate": 0.0005,
+ "loss": 1.024,
+ "step": 12946
+ },
+ {
+ "epoch": 41.364217252396166,
+ "grad_norm": 0.248046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0241,
+ "step": 12947
+ },
+ {
+ "epoch": 41.36741214057508,
+ "grad_norm": 0.10595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0266,
+ "step": 12948
+ },
+ {
+ "epoch": 41.370607028753994,
+ "grad_norm": 0.326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0357,
+ "step": 12949
+ },
+ {
+ "epoch": 41.373801916932905,
+ "grad_norm": 0.3046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0408,
+ "step": 12950
+ },
+ {
+ "epoch": 41.37699680511182,
+ "grad_norm": 0.21484375,
+ "learning_rate": 0.0005,
+ "loss": 1.034,
+ "step": 12951
+ },
+ {
+ "epoch": 41.38019169329073,
+ "grad_norm": 0.08544921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0393,
+ "step": 12952
+ },
+ {
+ "epoch": 41.38338658146965,
+ "grad_norm": 0.283203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0381,
+ "step": 12953
+ },
+ {
+ "epoch": 41.38658146964856,
+ "grad_norm": 0.263671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0343,
+ "step": 12954
+ },
+ {
+ "epoch": 41.38977635782748,
+ "grad_norm": 0.09228515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0312,
+ "step": 12955
+ },
+ {
+ "epoch": 41.39297124600639,
+ "grad_norm": 0.3203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0317,
+ "step": 12956
+ },
+ {
+ "epoch": 41.3961661341853,
+ "grad_norm": 0.1083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0373,
+ "step": 12957
+ },
+ {
+ "epoch": 41.39936102236422,
+ "grad_norm": 0.1826171875,
+ "learning_rate": 0.0005,
+ "loss": 1.033,
+ "step": 12958
+ },
+ {
+ "epoch": 41.40255591054313,
+ "grad_norm": 0.130859375,
+ "learning_rate": 0.0005,
+ "loss": 1.025,
+ "step": 12959
+ },
+ {
+ "epoch": 41.405750798722046,
+ "grad_norm": 0.11181640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0275,
+ "step": 12960
+ },
+ {
+ "epoch": 41.40894568690096,
+ "grad_norm": 0.162109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0268,
+ "step": 12961
+ },
+ {
+ "epoch": 41.412140575079874,
+ "grad_norm": 0.09765625,
+ "learning_rate": 0.0005,
+ "loss": 1.037,
+ "step": 12962
+ },
+ {
+ "epoch": 41.415335463258785,
+ "grad_norm": 0.255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0263,
+ "step": 12963
+ },
+ {
+ "epoch": 41.4185303514377,
+ "grad_norm": 0.10498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0281,
+ "step": 12964
+ },
+ {
+ "epoch": 41.42172523961661,
+ "grad_norm": 0.16015625,
+ "learning_rate": 0.0005,
+ "loss": 1.029,
+ "step": 12965
+ },
+ {
+ "epoch": 41.424920127795524,
+ "grad_norm": 0.08740234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0305,
+ "step": 12966
+ },
+ {
+ "epoch": 41.42811501597444,
+ "grad_norm": 0.15234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0196,
+ "step": 12967
+ },
+ {
+ "epoch": 41.43130990415335,
+ "grad_norm": 0.052001953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0377,
+ "step": 12968
+ },
+ {
+ "epoch": 41.43450479233227,
+ "grad_norm": 0.142578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0305,
+ "step": 12969
+ },
+ {
+ "epoch": 41.43769968051118,
+ "grad_norm": 0.0947265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0343,
+ "step": 12970
+ },
+ {
+ "epoch": 41.4408945686901,
+ "grad_norm": 0.1484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0267,
+ "step": 12971
+ },
+ {
+ "epoch": 41.44408945686901,
+ "grad_norm": 0.10302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0367,
+ "step": 12972
+ },
+ {
+ "epoch": 41.447284345047926,
+ "grad_norm": 0.1611328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0297,
+ "step": 12973
+ },
+ {
+ "epoch": 41.45047923322684,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0289,
+ "step": 12974
+ },
+ {
+ "epoch": 41.453674121405754,
+ "grad_norm": 0.15625,
+ "learning_rate": 0.0005,
+ "loss": 1.0355,
+ "step": 12975
+ },
+ {
+ "epoch": 41.456869009584665,
+ "grad_norm": 0.05810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0307,
+ "step": 12976
+ },
+ {
+ "epoch": 41.460063897763575,
+ "grad_norm": 0.1484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0375,
+ "step": 12977
+ },
+ {
+ "epoch": 41.46325878594249,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0418,
+ "step": 12978
+ },
+ {
+ "epoch": 41.466453674121404,
+ "grad_norm": 0.1171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0362,
+ "step": 12979
+ },
+ {
+ "epoch": 41.46964856230032,
+ "grad_norm": 0.09912109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0211,
+ "step": 12980
+ },
+ {
+ "epoch": 41.47284345047923,
+ "grad_norm": 0.10302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0384,
+ "step": 12981
+ },
+ {
+ "epoch": 41.47603833865815,
+ "grad_norm": 0.16796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0381,
+ "step": 12982
+ },
+ {
+ "epoch": 41.47923322683706,
+ "grad_norm": 0.099609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0299,
+ "step": 12983
+ },
+ {
+ "epoch": 41.48242811501598,
+ "grad_norm": 0.26171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0332,
+ "step": 12984
+ },
+ {
+ "epoch": 41.48562300319489,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0296,
+ "step": 12985
+ },
+ {
+ "epoch": 41.4888178913738,
+ "grad_norm": 0.232421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0356,
+ "step": 12986
+ },
+ {
+ "epoch": 41.49201277955272,
+ "grad_norm": 0.1806640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0212,
+ "step": 12987
+ },
+ {
+ "epoch": 41.49520766773163,
+ "grad_norm": 0.09375,
+ "learning_rate": 0.0005,
+ "loss": 1.0321,
+ "step": 12988
+ },
+ {
+ "epoch": 41.498402555910545,
+ "grad_norm": 0.240234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0262,
+ "step": 12989
+ },
+ {
+ "epoch": 41.501597444089455,
+ "grad_norm": 0.051025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0399,
+ "step": 12990
+ },
+ {
+ "epoch": 41.50479233226837,
+ "grad_norm": 0.259765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0259,
+ "step": 12991
+ },
+ {
+ "epoch": 41.50798722044728,
+ "grad_norm": 0.1513671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0366,
+ "step": 12992
+ },
+ {
+ "epoch": 41.5111821086262,
+ "grad_norm": 0.1826171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0357,
+ "step": 12993
+ },
+ {
+ "epoch": 41.51437699680511,
+ "grad_norm": 0.1953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0282,
+ "step": 12994
+ },
+ {
+ "epoch": 41.51757188498402,
+ "grad_norm": 0.1513671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0267,
+ "step": 12995
+ },
+ {
+ "epoch": 41.52076677316294,
+ "grad_norm": 0.255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0379,
+ "step": 12996
+ },
+ {
+ "epoch": 41.52396166134185,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0464,
+ "step": 12997
+ },
+ {
+ "epoch": 41.52715654952077,
+ "grad_norm": 0.154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0331,
+ "step": 12998
+ },
+ {
+ "epoch": 41.53035143769968,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0296,
+ "step": 12999
+ },
+ {
+ "epoch": 41.533546325878596,
+ "grad_norm": 0.158203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0324,
+ "step": 13000
+ },
+ {
+ "epoch": 41.53674121405751,
+ "grad_norm": 0.0888671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0323,
+ "step": 13001
+ },
+ {
+ "epoch": 41.539936102236425,
+ "grad_norm": 0.12451171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0439,
+ "step": 13002
+ },
+ {
+ "epoch": 41.543130990415335,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0347,
+ "step": 13003
+ },
+ {
+ "epoch": 41.546325878594246,
+ "grad_norm": 0.0966796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0528,
+ "step": 13004
+ },
+ {
+ "epoch": 41.54952076677316,
+ "grad_norm": 0.058837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0357,
+ "step": 13005
+ },
+ {
+ "epoch": 41.552715654952074,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.028,
+ "step": 13006
+ },
+ {
+ "epoch": 41.55591054313099,
+ "grad_norm": 0.045654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0403,
+ "step": 13007
+ },
+ {
+ "epoch": 41.5591054313099,
+ "grad_norm": 0.087890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0373,
+ "step": 13008
+ },
+ {
+ "epoch": 41.56230031948882,
+ "grad_norm": 0.052734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0445,
+ "step": 13009
+ },
+ {
+ "epoch": 41.56549520766773,
+ "grad_norm": 0.11376953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0303,
+ "step": 13010
+ },
+ {
+ "epoch": 41.56869009584665,
+ "grad_norm": 0.05908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0394,
+ "step": 13011
+ },
+ {
+ "epoch": 41.57188498402556,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0235,
+ "step": 13012
+ },
+ {
+ "epoch": 41.575079872204476,
+ "grad_norm": 0.05029296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0273,
+ "step": 13013
+ },
+ {
+ "epoch": 41.57827476038339,
+ "grad_norm": 0.052001953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0346,
+ "step": 13014
+ },
+ {
+ "epoch": 41.5814696485623,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0276,
+ "step": 13015
+ },
+ {
+ "epoch": 41.584664536741215,
+ "grad_norm": 0.045654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0327,
+ "step": 13016
+ },
+ {
+ "epoch": 41.587859424920126,
+ "grad_norm": 0.04638671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0283,
+ "step": 13017
+ },
+ {
+ "epoch": 41.59105431309904,
+ "grad_norm": 0.04638671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0341,
+ "step": 13018
+ },
+ {
+ "epoch": 41.594249201277954,
+ "grad_norm": 0.04931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0288,
+ "step": 13019
+ },
+ {
+ "epoch": 41.59744408945687,
+ "grad_norm": 0.08740234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0334,
+ "step": 13020
+ },
+ {
+ "epoch": 41.60063897763578,
+ "grad_norm": 0.0498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0307,
+ "step": 13021
+ },
+ {
+ "epoch": 41.6038338658147,
+ "grad_norm": 0.057373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.027,
+ "step": 13022
+ },
+ {
+ "epoch": 41.60702875399361,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0414,
+ "step": 13023
+ },
+ {
+ "epoch": 41.61022364217252,
+ "grad_norm": 0.08056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0252,
+ "step": 13024
+ },
+ {
+ "epoch": 41.61341853035144,
+ "grad_norm": 0.0859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0206,
+ "step": 13025
+ },
+ {
+ "epoch": 41.61661341853035,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0266,
+ "step": 13026
+ },
+ {
+ "epoch": 41.61980830670927,
+ "grad_norm": 0.059326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0316,
+ "step": 13027
+ },
+ {
+ "epoch": 41.62300319488818,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0294,
+ "step": 13028
+ },
+ {
+ "epoch": 41.626198083067095,
+ "grad_norm": 0.05712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0286,
+ "step": 13029
+ },
+ {
+ "epoch": 41.629392971246006,
+ "grad_norm": 0.052978515625,
+ "learning_rate": 0.0005,
+ "loss": 1.039,
+ "step": 13030
+ },
+ {
+ "epoch": 41.63258785942492,
+ "grad_norm": 0.04833984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0277,
+ "step": 13031
+ },
+ {
+ "epoch": 41.635782747603834,
+ "grad_norm": 0.052978515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0295,
+ "step": 13032
+ },
+ {
+ "epoch": 41.638977635782744,
+ "grad_norm": 0.0537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0304,
+ "step": 13033
+ },
+ {
+ "epoch": 41.64217252396166,
+ "grad_norm": 0.058837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.039,
+ "step": 13034
+ },
+ {
+ "epoch": 41.64536741214057,
+ "grad_norm": 0.054931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0178,
+ "step": 13035
+ },
+ {
+ "epoch": 41.64856230031949,
+ "grad_norm": 0.044677734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0195,
+ "step": 13036
+ },
+ {
+ "epoch": 41.6517571884984,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0357,
+ "step": 13037
+ },
+ {
+ "epoch": 41.65495207667732,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0305,
+ "step": 13038
+ },
+ {
+ "epoch": 41.65814696485623,
+ "grad_norm": 0.046142578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0363,
+ "step": 13039
+ },
+ {
+ "epoch": 41.66134185303515,
+ "grad_norm": 0.044677734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0298,
+ "step": 13040
+ },
+ {
+ "epoch": 41.66453674121406,
+ "grad_norm": 0.061767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0287,
+ "step": 13041
+ },
+ {
+ "epoch": 41.66773162939297,
+ "grad_norm": 0.0546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0298,
+ "step": 13042
+ },
+ {
+ "epoch": 41.670926517571885,
+ "grad_norm": 0.046875,
+ "learning_rate": 0.0005,
+ "loss": 1.027,
+ "step": 13043
+ },
+ {
+ "epoch": 41.674121405750796,
+ "grad_norm": 0.054931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0298,
+ "step": 13044
+ },
+ {
+ "epoch": 41.677316293929714,
+ "grad_norm": 0.042724609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0295,
+ "step": 13045
+ },
+ {
+ "epoch": 41.680511182108624,
+ "grad_norm": 0.04833984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0213,
+ "step": 13046
+ },
+ {
+ "epoch": 41.68370607028754,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0261,
+ "step": 13047
+ },
+ {
+ "epoch": 41.68690095846645,
+ "grad_norm": 0.044921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0335,
+ "step": 13048
+ },
+ {
+ "epoch": 41.69009584664537,
+ "grad_norm": 0.0439453125,
+ "learning_rate": 0.0005,
+ "loss": 1.032,
+ "step": 13049
+ },
+ {
+ "epoch": 41.69329073482428,
+ "grad_norm": 0.04296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0378,
+ "step": 13050
+ },
+ {
+ "epoch": 41.6964856230032,
+ "grad_norm": 0.051513671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0442,
+ "step": 13051
+ },
+ {
+ "epoch": 41.69968051118211,
+ "grad_norm": 0.0556640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0339,
+ "step": 13052
+ },
+ {
+ "epoch": 41.70287539936102,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0249,
+ "step": 13053
+ },
+ {
+ "epoch": 41.70607028753994,
+ "grad_norm": 0.061279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0296,
+ "step": 13054
+ },
+ {
+ "epoch": 41.70926517571885,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0234,
+ "step": 13055
+ },
+ {
+ "epoch": 41.712460063897765,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0309,
+ "step": 13056
+ },
+ {
+ "epoch": 41.715654952076676,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0367,
+ "step": 13057
+ },
+ {
+ "epoch": 41.718849840255594,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0303,
+ "step": 13058
+ },
+ {
+ "epoch": 41.722044728434504,
+ "grad_norm": 0.10009765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0423,
+ "step": 13059
+ },
+ {
+ "epoch": 41.72523961661342,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0269,
+ "step": 13060
+ },
+ {
+ "epoch": 41.72843450479233,
+ "grad_norm": 0.0966796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0271,
+ "step": 13061
+ },
+ {
+ "epoch": 41.73162939297124,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0288,
+ "step": 13062
+ },
+ {
+ "epoch": 41.73482428115016,
+ "grad_norm": 0.056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0367,
+ "step": 13063
+ },
+ {
+ "epoch": 41.73801916932907,
+ "grad_norm": 0.1474609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0385,
+ "step": 13064
+ },
+ {
+ "epoch": 41.74121405750799,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0285,
+ "step": 13065
+ },
+ {
+ "epoch": 41.7444089456869,
+ "grad_norm": 0.1494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0294,
+ "step": 13066
+ },
+ {
+ "epoch": 41.74760383386582,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0317,
+ "step": 13067
+ },
+ {
+ "epoch": 41.75079872204473,
+ "grad_norm": 0.11328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0338,
+ "step": 13068
+ },
+ {
+ "epoch": 41.753993610223645,
+ "grad_norm": 0.05810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0243,
+ "step": 13069
+ },
+ {
+ "epoch": 41.757188498402556,
+ "grad_norm": 0.10888671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0295,
+ "step": 13070
+ },
+ {
+ "epoch": 41.760383386581466,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0349,
+ "step": 13071
+ },
+ {
+ "epoch": 41.763578274760384,
+ "grad_norm": 0.09130859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0252,
+ "step": 13072
+ },
+ {
+ "epoch": 41.766773162939295,
+ "grad_norm": 0.050048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0398,
+ "step": 13073
+ },
+ {
+ "epoch": 41.76996805111821,
+ "grad_norm": 0.07861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0346,
+ "step": 13074
+ },
+ {
+ "epoch": 41.77316293929712,
+ "grad_norm": 0.052734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0266,
+ "step": 13075
+ },
+ {
+ "epoch": 41.77635782747604,
+ "grad_norm": 0.10888671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0406,
+ "step": 13076
+ },
+ {
+ "epoch": 41.77955271565495,
+ "grad_norm": 0.09814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0414,
+ "step": 13077
+ },
+ {
+ "epoch": 41.78274760383387,
+ "grad_norm": 0.10205078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0247,
+ "step": 13078
+ },
+ {
+ "epoch": 41.78594249201278,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0327,
+ "step": 13079
+ },
+ {
+ "epoch": 41.78913738019169,
+ "grad_norm": 0.09228515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0213,
+ "step": 13080
+ },
+ {
+ "epoch": 41.79233226837061,
+ "grad_norm": 0.050537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0354,
+ "step": 13081
+ },
+ {
+ "epoch": 41.79552715654952,
+ "grad_norm": 0.1142578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0363,
+ "step": 13082
+ },
+ {
+ "epoch": 41.798722044728436,
+ "grad_norm": 0.048095703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0339,
+ "step": 13083
+ },
+ {
+ "epoch": 41.801916932907346,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0297,
+ "step": 13084
+ },
+ {
+ "epoch": 41.805111821086264,
+ "grad_norm": 0.099609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0191,
+ "step": 13085
+ },
+ {
+ "epoch": 41.808306709265175,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0297,
+ "step": 13086
+ },
+ {
+ "epoch": 41.81150159744409,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0301,
+ "step": 13087
+ },
+ {
+ "epoch": 41.814696485623,
+ "grad_norm": 0.0556640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0445,
+ "step": 13088
+ },
+ {
+ "epoch": 41.81789137380191,
+ "grad_norm": 0.05029296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0335,
+ "step": 13089
+ },
+ {
+ "epoch": 41.82108626198083,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0331,
+ "step": 13090
+ },
+ {
+ "epoch": 41.82428115015974,
+ "grad_norm": 0.051513671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0233,
+ "step": 13091
+ },
+ {
+ "epoch": 41.82747603833866,
+ "grad_norm": 0.060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0336,
+ "step": 13092
+ },
+ {
+ "epoch": 41.83067092651757,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0389,
+ "step": 13093
+ },
+ {
+ "epoch": 41.83386581469649,
+ "grad_norm": 0.06005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0326,
+ "step": 13094
+ },
+ {
+ "epoch": 41.8370607028754,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0384,
+ "step": 13095
+ },
+ {
+ "epoch": 41.840255591054316,
+ "grad_norm": 0.1337890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0337,
+ "step": 13096
+ },
+ {
+ "epoch": 41.843450479233226,
+ "grad_norm": 0.1015625,
+ "learning_rate": 0.0005,
+ "loss": 1.04,
+ "step": 13097
+ },
+ {
+ "epoch": 41.846645367412144,
+ "grad_norm": 0.1474609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0356,
+ "step": 13098
+ },
+ {
+ "epoch": 41.849840255591054,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0285,
+ "step": 13099
+ },
+ {
+ "epoch": 41.853035143769965,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0413,
+ "step": 13100
+ },
+ {
+ "epoch": 41.85623003194888,
+ "grad_norm": 0.08349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0283,
+ "step": 13101
+ },
+ {
+ "epoch": 41.85942492012779,
+ "grad_norm": 0.08056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0383,
+ "step": 13102
+ },
+ {
+ "epoch": 41.86261980830671,
+ "grad_norm": 0.058837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0355,
+ "step": 13103
+ },
+ {
+ "epoch": 41.86581469648562,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0255,
+ "step": 13104
+ },
+ {
+ "epoch": 41.86900958466454,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0244,
+ "step": 13105
+ },
+ {
+ "epoch": 41.87220447284345,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0293,
+ "step": 13106
+ },
+ {
+ "epoch": 41.87539936102237,
+ "grad_norm": 0.08935546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0336,
+ "step": 13107
+ },
+ {
+ "epoch": 41.87859424920128,
+ "grad_norm": 0.08056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0387,
+ "step": 13108
+ },
+ {
+ "epoch": 41.88178913738019,
+ "grad_norm": 0.09130859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0401,
+ "step": 13109
+ },
+ {
+ "epoch": 41.884984025559106,
+ "grad_norm": 0.09814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0212,
+ "step": 13110
+ },
+ {
+ "epoch": 41.88817891373802,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0315,
+ "step": 13111
+ },
+ {
+ "epoch": 41.891373801916934,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.038,
+ "step": 13112
+ },
+ {
+ "epoch": 41.894568690095845,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0236,
+ "step": 13113
+ },
+ {
+ "epoch": 41.89776357827476,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0316,
+ "step": 13114
+ },
+ {
+ "epoch": 41.90095846645367,
+ "grad_norm": 0.08154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0286,
+ "step": 13115
+ },
+ {
+ "epoch": 41.90415335463259,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0269,
+ "step": 13116
+ },
+ {
+ "epoch": 41.9073482428115,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0241,
+ "step": 13117
+ },
+ {
+ "epoch": 41.91054313099041,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0262,
+ "step": 13118
+ },
+ {
+ "epoch": 41.91373801916933,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0268,
+ "step": 13119
+ },
+ {
+ "epoch": 41.91693290734824,
+ "grad_norm": 0.08349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0326,
+ "step": 13120
+ },
+ {
+ "epoch": 41.92012779552716,
+ "grad_norm": 0.046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0341,
+ "step": 13121
+ },
+ {
+ "epoch": 41.92332268370607,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.023,
+ "step": 13122
+ },
+ {
+ "epoch": 41.926517571884986,
+ "grad_norm": 0.055419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.033,
+ "step": 13123
+ },
+ {
+ "epoch": 41.9297124600639,
+ "grad_norm": 0.043701171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0321,
+ "step": 13124
+ },
+ {
+ "epoch": 41.932907348242814,
+ "grad_norm": 0.06103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0377,
+ "step": 13125
+ },
+ {
+ "epoch": 41.936102236421725,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0203,
+ "step": 13126
+ },
+ {
+ "epoch": 41.93929712460064,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0319,
+ "step": 13127
+ },
+ {
+ "epoch": 41.94249201277955,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0226,
+ "step": 13128
+ },
+ {
+ "epoch": 41.945686900958464,
+ "grad_norm": 0.05615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0278,
+ "step": 13129
+ },
+ {
+ "epoch": 41.94888178913738,
+ "grad_norm": 0.058837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0314,
+ "step": 13130
+ },
+ {
+ "epoch": 41.95207667731629,
+ "grad_norm": 0.056396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0382,
+ "step": 13131
+ },
+ {
+ "epoch": 41.95527156549521,
+ "grad_norm": 0.056396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.038,
+ "step": 13132
+ },
+ {
+ "epoch": 41.95846645367412,
+ "grad_norm": 0.07861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.032,
+ "step": 13133
+ },
+ {
+ "epoch": 41.96166134185304,
+ "grad_norm": 0.125,
+ "learning_rate": 0.0005,
+ "loss": 1.0307,
+ "step": 13134
+ },
+ {
+ "epoch": 41.96485623003195,
+ "grad_norm": 0.05712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0415,
+ "step": 13135
+ },
+ {
+ "epoch": 41.968051118210866,
+ "grad_norm": 0.1376953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0359,
+ "step": 13136
+ },
+ {
+ "epoch": 41.97124600638978,
+ "grad_norm": 0.05078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0329,
+ "step": 13137
+ },
+ {
+ "epoch": 41.97444089456869,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0287,
+ "step": 13138
+ },
+ {
+ "epoch": 41.977635782747605,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0399,
+ "step": 13139
+ },
+ {
+ "epoch": 41.980830670926515,
+ "grad_norm": 0.051513671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0323,
+ "step": 13140
+ },
+ {
+ "epoch": 41.98402555910543,
+ "grad_norm": 0.0478515625,
+ "learning_rate": 0.0005,
+ "loss": 1.04,
+ "step": 13141
+ },
+ {
+ "epoch": 41.98722044728434,
+ "grad_norm": 0.055908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0163,
+ "step": 13142
+ },
+ {
+ "epoch": 41.99041533546326,
+ "grad_norm": 0.04541015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0275,
+ "step": 13143
+ },
+ {
+ "epoch": 41.99361022364217,
+ "grad_norm": 0.045166015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0311,
+ "step": 13144
+ },
+ {
+ "epoch": 41.99680511182109,
+ "grad_norm": 0.054443359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0332,
+ "step": 13145
+ },
+ {
+ "epoch": 42.0,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0211,
+ "step": 13146
+ },
+ {
+ "epoch": 42.00319488817891,
+ "grad_norm": 0.04541015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0329,
+ "step": 13147
+ },
+ {
+ "epoch": 42.00638977635783,
+ "grad_norm": 0.05029296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0292,
+ "step": 13148
+ },
+ {
+ "epoch": 42.00958466453674,
+ "grad_norm": 0.0478515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0396,
+ "step": 13149
+ },
+ {
+ "epoch": 42.01277955271566,
+ "grad_norm": 0.0556640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0317,
+ "step": 13150
+ },
+ {
+ "epoch": 42.01597444089457,
+ "grad_norm": 0.0595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0306,
+ "step": 13151
+ },
+ {
+ "epoch": 42.019169329073485,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0192,
+ "step": 13152
+ },
+ {
+ "epoch": 42.022364217252395,
+ "grad_norm": 0.087890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0295,
+ "step": 13153
+ },
+ {
+ "epoch": 42.02555910543131,
+ "grad_norm": 0.080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.023,
+ "step": 13154
+ },
+ {
+ "epoch": 42.02875399361022,
+ "grad_norm": 0.1689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0237,
+ "step": 13155
+ },
+ {
+ "epoch": 42.031948881789134,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0291,
+ "step": 13156
+ },
+ {
+ "epoch": 42.03514376996805,
+ "grad_norm": 0.126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0298,
+ "step": 13157
+ },
+ {
+ "epoch": 42.03833865814696,
+ "grad_norm": 0.0595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0231,
+ "step": 13158
+ },
+ {
+ "epoch": 42.04153354632588,
+ "grad_norm": 0.09423828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0359,
+ "step": 13159
+ },
+ {
+ "epoch": 42.04472843450479,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0297,
+ "step": 13160
+ },
+ {
+ "epoch": 42.04792332268371,
+ "grad_norm": 0.080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0315,
+ "step": 13161
+ },
+ {
+ "epoch": 42.05111821086262,
+ "grad_norm": 0.05859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0277,
+ "step": 13162
+ },
+ {
+ "epoch": 42.054313099041536,
+ "grad_norm": 0.0810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.023,
+ "step": 13163
+ },
+ {
+ "epoch": 42.05750798722045,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0365,
+ "step": 13164
+ },
+ {
+ "epoch": 42.06070287539936,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0306,
+ "step": 13165
+ },
+ {
+ "epoch": 42.063897763578275,
+ "grad_norm": 0.05859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0322,
+ "step": 13166
+ },
+ {
+ "epoch": 42.067092651757186,
+ "grad_norm": 0.083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0222,
+ "step": 13167
+ },
+ {
+ "epoch": 42.0702875399361,
+ "grad_norm": 0.08447265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0277,
+ "step": 13168
+ },
+ {
+ "epoch": 42.073482428115014,
+ "grad_norm": 0.10400390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0279,
+ "step": 13169
+ },
+ {
+ "epoch": 42.07667731629393,
+ "grad_norm": 0.11865234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0428,
+ "step": 13170
+ },
+ {
+ "epoch": 42.07987220447284,
+ "grad_norm": 0.10498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0254,
+ "step": 13171
+ },
+ {
+ "epoch": 42.08306709265176,
+ "grad_norm": 0.11669921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0312,
+ "step": 13172
+ },
+ {
+ "epoch": 42.08626198083067,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0384,
+ "step": 13173
+ },
+ {
+ "epoch": 42.08945686900959,
+ "grad_norm": 0.11572265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0255,
+ "step": 13174
+ },
+ {
+ "epoch": 42.0926517571885,
+ "grad_norm": 0.051025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0368,
+ "step": 13175
+ },
+ {
+ "epoch": 42.09584664536741,
+ "grad_norm": 0.11474609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0298,
+ "step": 13176
+ },
+ {
+ "epoch": 42.09904153354633,
+ "grad_norm": 0.049560546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0388,
+ "step": 13177
+ },
+ {
+ "epoch": 42.10223642172524,
+ "grad_norm": 0.08984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0326,
+ "step": 13178
+ },
+ {
+ "epoch": 42.105431309904155,
+ "grad_norm": 0.048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0236,
+ "step": 13179
+ },
+ {
+ "epoch": 42.108626198083066,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0355,
+ "step": 13180
+ },
+ {
+ "epoch": 42.11182108626198,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0332,
+ "step": 13181
+ },
+ {
+ "epoch": 42.115015974440894,
+ "grad_norm": 0.11572265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0212,
+ "step": 13182
+ },
+ {
+ "epoch": 42.11821086261981,
+ "grad_norm": 0.047119140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0287,
+ "step": 13183
+ },
+ {
+ "epoch": 42.12140575079872,
+ "grad_norm": 0.11279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0438,
+ "step": 13184
+ },
+ {
+ "epoch": 42.12460063897763,
+ "grad_norm": 0.05517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0365,
+ "step": 13185
+ },
+ {
+ "epoch": 42.12779552715655,
+ "grad_norm": 0.095703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0361,
+ "step": 13186
+ },
+ {
+ "epoch": 42.13099041533546,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0321,
+ "step": 13187
+ },
+ {
+ "epoch": 42.13418530351438,
+ "grad_norm": 0.162109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0289,
+ "step": 13188
+ },
+ {
+ "epoch": 42.13738019169329,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0257,
+ "step": 13189
+ },
+ {
+ "epoch": 42.14057507987221,
+ "grad_norm": 0.22265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0435,
+ "step": 13190
+ },
+ {
+ "epoch": 42.14376996805112,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0191,
+ "step": 13191
+ },
+ {
+ "epoch": 42.146964856230035,
+ "grad_norm": 0.1357421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0388,
+ "step": 13192
+ },
+ {
+ "epoch": 42.150159744408946,
+ "grad_norm": 0.05712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0353,
+ "step": 13193
+ },
+ {
+ "epoch": 42.153354632587856,
+ "grad_norm": 0.107421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0413,
+ "step": 13194
+ },
+ {
+ "epoch": 42.156549520766774,
+ "grad_norm": 0.057373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0261,
+ "step": 13195
+ },
+ {
+ "epoch": 42.159744408945684,
+ "grad_norm": 0.09423828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0278,
+ "step": 13196
+ },
+ {
+ "epoch": 42.1629392971246,
+ "grad_norm": 0.048583984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0258,
+ "step": 13197
+ },
+ {
+ "epoch": 42.16613418530351,
+ "grad_norm": 0.10546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0406,
+ "step": 13198
+ },
+ {
+ "epoch": 42.16932907348243,
+ "grad_norm": 0.08837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0306,
+ "step": 13199
+ },
+ {
+ "epoch": 42.17252396166134,
+ "grad_norm": 0.08349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0212,
+ "step": 13200
+ },
+ {
+ "epoch": 42.17571884984026,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0341,
+ "step": 13201
+ },
+ {
+ "epoch": 42.17891373801917,
+ "grad_norm": 0.08740234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0133,
+ "step": 13202
+ },
+ {
+ "epoch": 42.18210862619808,
+ "grad_norm": 0.048583984375,
+ "learning_rate": 0.0005,
+ "loss": 1.034,
+ "step": 13203
+ },
+ {
+ "epoch": 42.185303514377,
+ "grad_norm": 0.080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0292,
+ "step": 13204
+ },
+ {
+ "epoch": 42.18849840255591,
+ "grad_norm": 0.05419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0289,
+ "step": 13205
+ },
+ {
+ "epoch": 42.191693290734825,
+ "grad_norm": 0.050048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0319,
+ "step": 13206
+ },
+ {
+ "epoch": 42.194888178913736,
+ "grad_norm": 0.10400390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0343,
+ "step": 13207
+ },
+ {
+ "epoch": 42.198083067092654,
+ "grad_norm": 0.051025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0281,
+ "step": 13208
+ },
+ {
+ "epoch": 42.201277955271564,
+ "grad_norm": 0.10546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0356,
+ "step": 13209
+ },
+ {
+ "epoch": 42.20447284345048,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0295,
+ "step": 13210
+ },
+ {
+ "epoch": 42.20766773162939,
+ "grad_norm": 0.10546875,
+ "learning_rate": 0.0005,
+ "loss": 1.032,
+ "step": 13211
+ },
+ {
+ "epoch": 42.21086261980831,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0353,
+ "step": 13212
+ },
+ {
+ "epoch": 42.21405750798722,
+ "grad_norm": 0.056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.041,
+ "step": 13213
+ },
+ {
+ "epoch": 42.21725239616613,
+ "grad_norm": 0.06103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0429,
+ "step": 13214
+ },
+ {
+ "epoch": 42.22044728434505,
+ "grad_norm": 0.08154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.038,
+ "step": 13215
+ },
+ {
+ "epoch": 42.22364217252396,
+ "grad_norm": 0.06103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.039,
+ "step": 13216
+ },
+ {
+ "epoch": 42.22683706070288,
+ "grad_norm": 0.056884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0324,
+ "step": 13217
+ },
+ {
+ "epoch": 42.23003194888179,
+ "grad_norm": 0.044677734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0317,
+ "step": 13218
+ },
+ {
+ "epoch": 42.233226837060705,
+ "grad_norm": 0.059814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.034,
+ "step": 13219
+ },
+ {
+ "epoch": 42.236421725239616,
+ "grad_norm": 0.0537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.033,
+ "step": 13220
+ },
+ {
+ "epoch": 42.239616613418534,
+ "grad_norm": 0.05517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0308,
+ "step": 13221
+ },
+ {
+ "epoch": 42.242811501597444,
+ "grad_norm": 0.0556640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0272,
+ "step": 13222
+ },
+ {
+ "epoch": 42.246006389776355,
+ "grad_norm": 0.0556640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0282,
+ "step": 13223
+ },
+ {
+ "epoch": 42.24920127795527,
+ "grad_norm": 0.05078125,
+ "learning_rate": 0.0005,
+ "loss": 1.029,
+ "step": 13224
+ },
+ {
+ "epoch": 42.25239616613418,
+ "grad_norm": 0.045166015625,
+ "learning_rate": 0.0005,
+ "loss": 1.03,
+ "step": 13225
+ },
+ {
+ "epoch": 42.2555910543131,
+ "grad_norm": 0.0595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0481,
+ "step": 13226
+ },
+ {
+ "epoch": 42.25878594249201,
+ "grad_norm": 0.04638671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0275,
+ "step": 13227
+ },
+ {
+ "epoch": 42.26198083067093,
+ "grad_norm": 0.056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0294,
+ "step": 13228
+ },
+ {
+ "epoch": 42.26517571884984,
+ "grad_norm": 0.0546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0327,
+ "step": 13229
+ },
+ {
+ "epoch": 42.26837060702876,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0374,
+ "step": 13230
+ },
+ {
+ "epoch": 42.27156549520767,
+ "grad_norm": 0.056396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0378,
+ "step": 13231
+ },
+ {
+ "epoch": 42.27476038338658,
+ "grad_norm": 0.051025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0268,
+ "step": 13232
+ },
+ {
+ "epoch": 42.277955271565496,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0315,
+ "step": 13233
+ },
+ {
+ "epoch": 42.281150159744406,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0238,
+ "step": 13234
+ },
+ {
+ "epoch": 42.284345047923324,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0376,
+ "step": 13235
+ },
+ {
+ "epoch": 42.287539936102235,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0284,
+ "step": 13236
+ },
+ {
+ "epoch": 42.29073482428115,
+ "grad_norm": 0.05615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0382,
+ "step": 13237
+ },
+ {
+ "epoch": 42.29392971246006,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0246,
+ "step": 13238
+ },
+ {
+ "epoch": 42.29712460063898,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0334,
+ "step": 13239
+ },
+ {
+ "epoch": 42.30031948881789,
+ "grad_norm": 0.058837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0434,
+ "step": 13240
+ },
+ {
+ "epoch": 42.3035143769968,
+ "grad_norm": 0.0859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0253,
+ "step": 13241
+ },
+ {
+ "epoch": 42.30670926517572,
+ "grad_norm": 0.05029296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0262,
+ "step": 13242
+ },
+ {
+ "epoch": 42.30990415335463,
+ "grad_norm": 0.08740234375,
+ "learning_rate": 0.0005,
+ "loss": 1.02,
+ "step": 13243
+ },
+ {
+ "epoch": 42.31309904153355,
+ "grad_norm": 0.05029296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0298,
+ "step": 13244
+ },
+ {
+ "epoch": 42.31629392971246,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0294,
+ "step": 13245
+ },
+ {
+ "epoch": 42.319488817891376,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.035,
+ "step": 13246
+ },
+ {
+ "epoch": 42.322683706070286,
+ "grad_norm": 0.11376953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0256,
+ "step": 13247
+ },
+ {
+ "epoch": 42.325878594249204,
+ "grad_norm": 0.05908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0292,
+ "step": 13248
+ },
+ {
+ "epoch": 42.329073482428115,
+ "grad_norm": 0.09375,
+ "learning_rate": 0.0005,
+ "loss": 1.0301,
+ "step": 13249
+ },
+ {
+ "epoch": 42.33226837060703,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0353,
+ "step": 13250
+ },
+ {
+ "epoch": 42.33546325878594,
+ "grad_norm": 0.08740234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0365,
+ "step": 13251
+ },
+ {
+ "epoch": 42.33865814696485,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0293,
+ "step": 13252
+ },
+ {
+ "epoch": 42.34185303514377,
+ "grad_norm": 0.0927734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0309,
+ "step": 13253
+ },
+ {
+ "epoch": 42.34504792332268,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0255,
+ "step": 13254
+ },
+ {
+ "epoch": 42.3482428115016,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.039,
+ "step": 13255
+ },
+ {
+ "epoch": 42.35143769968051,
+ "grad_norm": 0.09326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0431,
+ "step": 13256
+ },
+ {
+ "epoch": 42.35463258785943,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0273,
+ "step": 13257
+ },
+ {
+ "epoch": 42.35782747603834,
+ "grad_norm": 0.05517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0401,
+ "step": 13258
+ },
+ {
+ "epoch": 42.361022364217256,
+ "grad_norm": 0.07861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0337,
+ "step": 13259
+ },
+ {
+ "epoch": 42.364217252396166,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0321,
+ "step": 13260
+ },
+ {
+ "epoch": 42.36741214057508,
+ "grad_norm": 0.10986328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0324,
+ "step": 13261
+ },
+ {
+ "epoch": 42.370607028753994,
+ "grad_norm": 0.04638671875,
+ "learning_rate": 0.0005,
+ "loss": 1.022,
+ "step": 13262
+ },
+ {
+ "epoch": 42.373801916932905,
+ "grad_norm": 0.05712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0332,
+ "step": 13263
+ },
+ {
+ "epoch": 42.37699680511182,
+ "grad_norm": 0.0537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0285,
+ "step": 13264
+ },
+ {
+ "epoch": 42.38019169329073,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0306,
+ "step": 13265
+ },
+ {
+ "epoch": 42.38338658146965,
+ "grad_norm": 0.05029296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0454,
+ "step": 13266
+ },
+ {
+ "epoch": 42.38658146964856,
+ "grad_norm": 0.060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0314,
+ "step": 13267
+ },
+ {
+ "epoch": 42.38977635782748,
+ "grad_norm": 0.054931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0381,
+ "step": 13268
+ },
+ {
+ "epoch": 42.39297124600639,
+ "grad_norm": 0.06201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0277,
+ "step": 13269
+ },
+ {
+ "epoch": 42.3961661341853,
+ "grad_norm": 0.0966796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0312,
+ "step": 13270
+ },
+ {
+ "epoch": 42.39936102236422,
+ "grad_norm": 0.055419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0341,
+ "step": 13271
+ },
+ {
+ "epoch": 42.40255591054313,
+ "grad_norm": 0.08740234375,
+ "learning_rate": 0.0005,
+ "loss": 1.035,
+ "step": 13272
+ },
+ {
+ "epoch": 42.405750798722046,
+ "grad_norm": 0.054443359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0374,
+ "step": 13273
+ },
+ {
+ "epoch": 42.40894568690096,
+ "grad_norm": 0.09228515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0298,
+ "step": 13274
+ },
+ {
+ "epoch": 42.412140575079874,
+ "grad_norm": 0.0595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0339,
+ "step": 13275
+ },
+ {
+ "epoch": 42.415335463258785,
+ "grad_norm": 0.0849609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0259,
+ "step": 13276
+ },
+ {
+ "epoch": 42.4185303514377,
+ "grad_norm": 0.1923828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0264,
+ "step": 13277
+ },
+ {
+ "epoch": 42.42172523961661,
+ "grad_norm": 0.3125,
+ "learning_rate": 0.0005,
+ "loss": 1.0369,
+ "step": 13278
+ },
+ {
+ "epoch": 42.424920127795524,
+ "grad_norm": 0.09521484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0326,
+ "step": 13279
+ },
+ {
+ "epoch": 42.42811501597444,
+ "grad_norm": 0.173828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0327,
+ "step": 13280
+ },
+ {
+ "epoch": 42.43130990415335,
+ "grad_norm": 0.205078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0389,
+ "step": 13281
+ },
+ {
+ "epoch": 42.43450479233227,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0384,
+ "step": 13282
+ },
+ {
+ "epoch": 42.43769968051118,
+ "grad_norm": 0.2353515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0348,
+ "step": 13283
+ },
+ {
+ "epoch": 42.4408945686901,
+ "grad_norm": 0.27734375,
+ "learning_rate": 0.0005,
+ "loss": 1.03,
+ "step": 13284
+ },
+ {
+ "epoch": 42.44408945686901,
+ "grad_norm": 0.059814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0286,
+ "step": 13285
+ },
+ {
+ "epoch": 42.447284345047926,
+ "grad_norm": 0.25390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0278,
+ "step": 13286
+ },
+ {
+ "epoch": 42.45047923322684,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0418,
+ "step": 13287
+ },
+ {
+ "epoch": 42.453674121405754,
+ "grad_norm": 0.1640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0244,
+ "step": 13288
+ },
+ {
+ "epoch": 42.456869009584665,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0401,
+ "step": 13289
+ },
+ {
+ "epoch": 42.460063897763575,
+ "grad_norm": 0.0830078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0229,
+ "step": 13290
+ },
+ {
+ "epoch": 42.46325878594249,
+ "grad_norm": 0.08984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0217,
+ "step": 13291
+ },
+ {
+ "epoch": 42.466453674121404,
+ "grad_norm": 0.080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0366,
+ "step": 13292
+ },
+ {
+ "epoch": 42.46964856230032,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0292,
+ "step": 13293
+ },
+ {
+ "epoch": 42.47284345047923,
+ "grad_norm": 0.08203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0302,
+ "step": 13294
+ },
+ {
+ "epoch": 42.47603833865815,
+ "grad_norm": 0.11474609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0313,
+ "step": 13295
+ },
+ {
+ "epoch": 42.47923322683706,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0429,
+ "step": 13296
+ },
+ {
+ "epoch": 42.48242811501598,
+ "grad_norm": 0.052978515625,
+ "learning_rate": 0.0005,
+ "loss": 1.033,
+ "step": 13297
+ },
+ {
+ "epoch": 42.48562300319489,
+ "grad_norm": 0.10302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0358,
+ "step": 13298
+ },
+ {
+ "epoch": 42.4888178913738,
+ "grad_norm": 0.05322265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0396,
+ "step": 13299
+ },
+ {
+ "epoch": 42.49201277955272,
+ "grad_norm": 0.09326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.035,
+ "step": 13300
+ },
+ {
+ "epoch": 42.49520766773163,
+ "grad_norm": 0.080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0331,
+ "step": 13301
+ },
+ {
+ "epoch": 42.498402555910545,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0258,
+ "step": 13302
+ },
+ {
+ "epoch": 42.501597444089455,
+ "grad_norm": 0.05615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0351,
+ "step": 13303
+ },
+ {
+ "epoch": 42.50479233226837,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0286,
+ "step": 13304
+ },
+ {
+ "epoch": 42.50798722044728,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0283,
+ "step": 13305
+ },
+ {
+ "epoch": 42.5111821086262,
+ "grad_norm": 0.08349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0314,
+ "step": 13306
+ },
+ {
+ "epoch": 42.51437699680511,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0327,
+ "step": 13307
+ },
+ {
+ "epoch": 42.51757188498402,
+ "grad_norm": 0.0888671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0324,
+ "step": 13308
+ },
+ {
+ "epoch": 42.52076677316294,
+ "grad_norm": 0.06005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.03,
+ "step": 13309
+ },
+ {
+ "epoch": 42.52396166134185,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.036,
+ "step": 13310
+ },
+ {
+ "epoch": 42.52715654952077,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0217,
+ "step": 13311
+ },
+ {
+ "epoch": 42.53035143769968,
+ "grad_norm": 0.095703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0354,
+ "step": 13312
+ },
+ {
+ "epoch": 42.533546325878596,
+ "grad_norm": 0.057373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0185,
+ "step": 13313
+ },
+ {
+ "epoch": 42.53674121405751,
+ "grad_norm": 0.08837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0415,
+ "step": 13314
+ },
+ {
+ "epoch": 42.539936102236425,
+ "grad_norm": 0.05810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0345,
+ "step": 13315
+ },
+ {
+ "epoch": 42.543130990415335,
+ "grad_norm": 0.08544921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0333,
+ "step": 13316
+ },
+ {
+ "epoch": 42.546325878594246,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0329,
+ "step": 13317
+ },
+ {
+ "epoch": 42.54952076677316,
+ "grad_norm": 0.12451171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0347,
+ "step": 13318
+ },
+ {
+ "epoch": 42.552715654952074,
+ "grad_norm": 0.1396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0282,
+ "step": 13319
+ },
+ {
+ "epoch": 42.55591054313099,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0311,
+ "step": 13320
+ },
+ {
+ "epoch": 42.5591054313099,
+ "grad_norm": 0.046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0276,
+ "step": 13321
+ },
+ {
+ "epoch": 42.56230031948882,
+ "grad_norm": 0.09423828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0378,
+ "step": 13322
+ },
+ {
+ "epoch": 42.56549520766773,
+ "grad_norm": 0.058349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0295,
+ "step": 13323
+ },
+ {
+ "epoch": 42.56869009584665,
+ "grad_norm": 0.04541015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0385,
+ "step": 13324
+ },
+ {
+ "epoch": 42.57188498402556,
+ "grad_norm": 0.052978515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0275,
+ "step": 13325
+ },
+ {
+ "epoch": 42.575079872204476,
+ "grad_norm": 0.058837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0329,
+ "step": 13326
+ },
+ {
+ "epoch": 42.57827476038339,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.031,
+ "step": 13327
+ },
+ {
+ "epoch": 42.5814696485623,
+ "grad_norm": 0.046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0304,
+ "step": 13328
+ },
+ {
+ "epoch": 42.584664536741215,
+ "grad_norm": 0.0986328125,
+ "learning_rate": 0.0005,
+ "loss": 1.029,
+ "step": 13329
+ },
+ {
+ "epoch": 42.587859424920126,
+ "grad_norm": 0.048095703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0466,
+ "step": 13330
+ },
+ {
+ "epoch": 42.59105431309904,
+ "grad_norm": 0.055419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.031,
+ "step": 13331
+ },
+ {
+ "epoch": 42.594249201277954,
+ "grad_norm": 0.08984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0275,
+ "step": 13332
+ },
+ {
+ "epoch": 42.59744408945687,
+ "grad_norm": 0.05615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0339,
+ "step": 13333
+ },
+ {
+ "epoch": 42.60063897763578,
+ "grad_norm": 0.053466796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0326,
+ "step": 13334
+ },
+ {
+ "epoch": 42.6038338658147,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0167,
+ "step": 13335
+ },
+ {
+ "epoch": 42.60702875399361,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0353,
+ "step": 13336
+ },
+ {
+ "epoch": 42.61022364217252,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0147,
+ "step": 13337
+ },
+ {
+ "epoch": 42.61341853035144,
+ "grad_norm": 0.0908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0299,
+ "step": 13338
+ },
+ {
+ "epoch": 42.61661341853035,
+ "grad_norm": 0.087890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0269,
+ "step": 13339
+ },
+ {
+ "epoch": 42.61980830670927,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0355,
+ "step": 13340
+ },
+ {
+ "epoch": 42.62300319488818,
+ "grad_norm": 0.12353515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0417,
+ "step": 13341
+ },
+ {
+ "epoch": 42.626198083067095,
+ "grad_norm": 0.060791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0368,
+ "step": 13342
+ },
+ {
+ "epoch": 42.629392971246006,
+ "grad_norm": 0.1142578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0393,
+ "step": 13343
+ },
+ {
+ "epoch": 42.63258785942492,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0387,
+ "step": 13344
+ },
+ {
+ "epoch": 42.635782747603834,
+ "grad_norm": 0.08056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.029,
+ "step": 13345
+ },
+ {
+ "epoch": 42.638977635782744,
+ "grad_norm": 0.1005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0311,
+ "step": 13346
+ },
+ {
+ "epoch": 42.64217252396166,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0317,
+ "step": 13347
+ },
+ {
+ "epoch": 42.64536741214057,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0248,
+ "step": 13348
+ },
+ {
+ "epoch": 42.64856230031949,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0313,
+ "step": 13349
+ },
+ {
+ "epoch": 42.6517571884984,
+ "grad_norm": 0.059814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0204,
+ "step": 13350
+ },
+ {
+ "epoch": 42.65495207667732,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0345,
+ "step": 13351
+ },
+ {
+ "epoch": 42.65814696485623,
+ "grad_norm": 0.1015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0449,
+ "step": 13352
+ },
+ {
+ "epoch": 42.66134185303515,
+ "grad_norm": 0.058349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.037,
+ "step": 13353
+ },
+ {
+ "epoch": 42.66453674121406,
+ "grad_norm": 0.09375,
+ "learning_rate": 0.0005,
+ "loss": 1.0319,
+ "step": 13354
+ },
+ {
+ "epoch": 42.66773162939297,
+ "grad_norm": 0.06103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0363,
+ "step": 13355
+ },
+ {
+ "epoch": 42.670926517571885,
+ "grad_norm": 0.056884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0243,
+ "step": 13356
+ },
+ {
+ "epoch": 42.674121405750796,
+ "grad_norm": 0.056396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0341,
+ "step": 13357
+ },
+ {
+ "epoch": 42.677316293929714,
+ "grad_norm": 0.05419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.031,
+ "step": 13358
+ },
+ {
+ "epoch": 42.680511182108624,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0318,
+ "step": 13359
+ },
+ {
+ "epoch": 42.68370607028754,
+ "grad_norm": 0.0810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0328,
+ "step": 13360
+ },
+ {
+ "epoch": 42.68690095846645,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0251,
+ "step": 13361
+ },
+ {
+ "epoch": 42.69009584664537,
+ "grad_norm": 0.1376953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0322,
+ "step": 13362
+ },
+ {
+ "epoch": 42.69329073482428,
+ "grad_norm": 0.053466796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0258,
+ "step": 13363
+ },
+ {
+ "epoch": 42.6964856230032,
+ "grad_norm": 0.1259765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0206,
+ "step": 13364
+ },
+ {
+ "epoch": 42.69968051118211,
+ "grad_norm": 0.04541015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0278,
+ "step": 13365
+ },
+ {
+ "epoch": 42.70287539936102,
+ "grad_norm": 0.083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0157,
+ "step": 13366
+ },
+ {
+ "epoch": 42.70607028753994,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0278,
+ "step": 13367
+ },
+ {
+ "epoch": 42.70926517571885,
+ "grad_norm": 0.10693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0218,
+ "step": 13368
+ },
+ {
+ "epoch": 42.712460063897765,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0337,
+ "step": 13369
+ },
+ {
+ "epoch": 42.715654952076676,
+ "grad_norm": 0.1416015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0359,
+ "step": 13370
+ },
+ {
+ "epoch": 42.718849840255594,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0372,
+ "step": 13371
+ },
+ {
+ "epoch": 42.722044728434504,
+ "grad_norm": 0.1455078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0419,
+ "step": 13372
+ },
+ {
+ "epoch": 42.72523961661342,
+ "grad_norm": 0.060302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0289,
+ "step": 13373
+ },
+ {
+ "epoch": 42.72843450479233,
+ "grad_norm": 0.1337890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0291,
+ "step": 13374
+ },
+ {
+ "epoch": 42.73162939297124,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0381,
+ "step": 13375
+ },
+ {
+ "epoch": 42.73482428115016,
+ "grad_norm": 0.1259765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0345,
+ "step": 13376
+ },
+ {
+ "epoch": 42.73801916932907,
+ "grad_norm": 0.08349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0265,
+ "step": 13377
+ },
+ {
+ "epoch": 42.74121405750799,
+ "grad_norm": 0.0888671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0309,
+ "step": 13378
+ },
+ {
+ "epoch": 42.7444089456869,
+ "grad_norm": 0.0498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0298,
+ "step": 13379
+ },
+ {
+ "epoch": 42.74760383386582,
+ "grad_norm": 0.08935546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0228,
+ "step": 13380
+ },
+ {
+ "epoch": 42.75079872204473,
+ "grad_norm": 0.05126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0284,
+ "step": 13381
+ },
+ {
+ "epoch": 42.753993610223645,
+ "grad_norm": 0.049072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0363,
+ "step": 13382
+ },
+ {
+ "epoch": 42.757188498402556,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0337,
+ "step": 13383
+ },
+ {
+ "epoch": 42.760383386581466,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.038,
+ "step": 13384
+ },
+ {
+ "epoch": 42.763578274760384,
+ "grad_norm": 0.052734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0192,
+ "step": 13385
+ },
+ {
+ "epoch": 42.766773162939295,
+ "grad_norm": 0.1005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0379,
+ "step": 13386
+ },
+ {
+ "epoch": 42.76996805111821,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0378,
+ "step": 13387
+ },
+ {
+ "epoch": 42.77316293929712,
+ "grad_norm": 0.11669921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0331,
+ "step": 13388
+ },
+ {
+ "epoch": 42.77635782747604,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0458,
+ "step": 13389
+ },
+ {
+ "epoch": 42.77955271565495,
+ "grad_norm": 0.1162109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0252,
+ "step": 13390
+ },
+ {
+ "epoch": 42.78274760383387,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0421,
+ "step": 13391
+ },
+ {
+ "epoch": 42.78594249201278,
+ "grad_norm": 0.125,
+ "learning_rate": 0.0005,
+ "loss": 1.0293,
+ "step": 13392
+ },
+ {
+ "epoch": 42.78913738019169,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0429,
+ "step": 13393
+ },
+ {
+ "epoch": 42.79233226837061,
+ "grad_norm": 0.166015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0261,
+ "step": 13394
+ },
+ {
+ "epoch": 42.79552715654952,
+ "grad_norm": 0.134765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0332,
+ "step": 13395
+ },
+ {
+ "epoch": 42.798722044728436,
+ "grad_norm": 0.181640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0459,
+ "step": 13396
+ },
+ {
+ "epoch": 42.801916932907346,
+ "grad_norm": 0.220703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0314,
+ "step": 13397
+ },
+ {
+ "epoch": 42.805111821086264,
+ "grad_norm": 0.080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0348,
+ "step": 13398
+ },
+ {
+ "epoch": 42.808306709265175,
+ "grad_norm": 0.16796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0335,
+ "step": 13399
+ },
+ {
+ "epoch": 42.81150159744409,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.034,
+ "step": 13400
+ },
+ {
+ "epoch": 42.814696485623,
+ "grad_norm": 0.10986328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0313,
+ "step": 13401
+ },
+ {
+ "epoch": 42.81789137380191,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0228,
+ "step": 13402
+ },
+ {
+ "epoch": 42.82108626198083,
+ "grad_norm": 0.09326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0286,
+ "step": 13403
+ },
+ {
+ "epoch": 42.82428115015974,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0369,
+ "step": 13404
+ },
+ {
+ "epoch": 42.82747603833866,
+ "grad_norm": 0.08447265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0325,
+ "step": 13405
+ },
+ {
+ "epoch": 42.83067092651757,
+ "grad_norm": 0.058837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0239,
+ "step": 13406
+ },
+ {
+ "epoch": 42.83386581469649,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0251,
+ "step": 13407
+ },
+ {
+ "epoch": 42.8370607028754,
+ "grad_norm": 0.0810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0411,
+ "step": 13408
+ },
+ {
+ "epoch": 42.840255591054316,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0326,
+ "step": 13409
+ },
+ {
+ "epoch": 42.843450479233226,
+ "grad_norm": 0.051025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0259,
+ "step": 13410
+ },
+ {
+ "epoch": 42.846645367412144,
+ "grad_norm": 0.05126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0215,
+ "step": 13411
+ },
+ {
+ "epoch": 42.849840255591054,
+ "grad_norm": 0.05078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0274,
+ "step": 13412
+ },
+ {
+ "epoch": 42.853035143769965,
+ "grad_norm": 0.059326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0301,
+ "step": 13413
+ },
+ {
+ "epoch": 42.85623003194888,
+ "grad_norm": 0.0478515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0197,
+ "step": 13414
+ },
+ {
+ "epoch": 42.85942492012779,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0356,
+ "step": 13415
+ },
+ {
+ "epoch": 42.86261980830671,
+ "grad_norm": 0.1005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0214,
+ "step": 13416
+ },
+ {
+ "epoch": 42.86581469648562,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0297,
+ "step": 13417
+ },
+ {
+ "epoch": 42.86900958466454,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0418,
+ "step": 13418
+ },
+ {
+ "epoch": 42.87220447284345,
+ "grad_norm": 0.046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0324,
+ "step": 13419
+ },
+ {
+ "epoch": 42.87539936102237,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0351,
+ "step": 13420
+ },
+ {
+ "epoch": 42.87859424920128,
+ "grad_norm": 0.050048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0318,
+ "step": 13421
+ },
+ {
+ "epoch": 42.88178913738019,
+ "grad_norm": 0.043701171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0364,
+ "step": 13422
+ },
+ {
+ "epoch": 42.884984025559106,
+ "grad_norm": 0.05126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.032,
+ "step": 13423
+ },
+ {
+ "epoch": 42.88817891373802,
+ "grad_norm": 0.049560546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0292,
+ "step": 13424
+ },
+ {
+ "epoch": 42.891373801916934,
+ "grad_norm": 0.05224609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0322,
+ "step": 13425
+ },
+ {
+ "epoch": 42.894568690095845,
+ "grad_norm": 0.05029296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0325,
+ "step": 13426
+ },
+ {
+ "epoch": 42.89776357827476,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.032,
+ "step": 13427
+ },
+ {
+ "epoch": 42.90095846645367,
+ "grad_norm": 0.04833984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0331,
+ "step": 13428
+ },
+ {
+ "epoch": 42.90415335463259,
+ "grad_norm": 0.08349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.039,
+ "step": 13429
+ },
+ {
+ "epoch": 42.9073482428115,
+ "grad_norm": 0.062255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0321,
+ "step": 13430
+ },
+ {
+ "epoch": 42.91054313099041,
+ "grad_norm": 0.059814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0396,
+ "step": 13431
+ },
+ {
+ "epoch": 42.91373801916933,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.022,
+ "step": 13432
+ },
+ {
+ "epoch": 42.91693290734824,
+ "grad_norm": 0.048095703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0312,
+ "step": 13433
+ },
+ {
+ "epoch": 42.92012779552716,
+ "grad_norm": 0.0556640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0391,
+ "step": 13434
+ },
+ {
+ "epoch": 42.92332268370607,
+ "grad_norm": 0.054443359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0308,
+ "step": 13435
+ },
+ {
+ "epoch": 42.926517571884986,
+ "grad_norm": 0.04736328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0384,
+ "step": 13436
+ },
+ {
+ "epoch": 42.9297124600639,
+ "grad_norm": 0.054443359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0267,
+ "step": 13437
+ },
+ {
+ "epoch": 42.932907348242814,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0288,
+ "step": 13438
+ },
+ {
+ "epoch": 42.936102236421725,
+ "grad_norm": 0.048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0293,
+ "step": 13439
+ },
+ {
+ "epoch": 42.93929712460064,
+ "grad_norm": 0.1005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0195,
+ "step": 13440
+ },
+ {
+ "epoch": 42.94249201277955,
+ "grad_norm": 0.056396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0377,
+ "step": 13441
+ },
+ {
+ "epoch": 42.945686900958464,
+ "grad_norm": 0.11376953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0335,
+ "step": 13442
+ },
+ {
+ "epoch": 42.94888178913738,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0199,
+ "step": 13443
+ },
+ {
+ "epoch": 42.95207667731629,
+ "grad_norm": 0.0986328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0392,
+ "step": 13444
+ },
+ {
+ "epoch": 42.95527156549521,
+ "grad_norm": 0.06005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0406,
+ "step": 13445
+ },
+ {
+ "epoch": 42.95846645367412,
+ "grad_norm": 0.05859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0361,
+ "step": 13446
+ },
+ {
+ "epoch": 42.96166134185304,
+ "grad_norm": 0.058837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0334,
+ "step": 13447
+ },
+ {
+ "epoch": 42.96485623003195,
+ "grad_norm": 0.056884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0356,
+ "step": 13448
+ },
+ {
+ "epoch": 42.968051118210866,
+ "grad_norm": 0.048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0262,
+ "step": 13449
+ },
+ {
+ "epoch": 42.97124600638978,
+ "grad_norm": 0.048583984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0335,
+ "step": 13450
+ },
+ {
+ "epoch": 42.97444089456869,
+ "grad_norm": 0.0595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0315,
+ "step": 13451
+ },
+ {
+ "epoch": 42.977635782747605,
+ "grad_norm": 0.05908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0359,
+ "step": 13452
+ },
+ {
+ "epoch": 42.980830670926515,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0324,
+ "step": 13453
+ },
+ {
+ "epoch": 42.98402555910543,
+ "grad_norm": 0.045654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0245,
+ "step": 13454
+ },
+ {
+ "epoch": 42.98722044728434,
+ "grad_norm": 0.05419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0337,
+ "step": 13455
+ },
+ {
+ "epoch": 42.99041533546326,
+ "grad_norm": 0.05908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0304,
+ "step": 13456
+ },
+ {
+ "epoch": 42.99361022364217,
+ "grad_norm": 0.053466796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0381,
+ "step": 13457
+ },
+ {
+ "epoch": 42.99680511182109,
+ "grad_norm": 0.051513671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0261,
+ "step": 13458
+ },
+ {
+ "epoch": 43.0,
+ "grad_norm": 0.08056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.028,
+ "step": 13459
+ },
+ {
+ "epoch": 43.00319488817891,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0293,
+ "step": 13460
+ },
+ {
+ "epoch": 43.00638977635783,
+ "grad_norm": 0.0537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0269,
+ "step": 13461
+ },
+ {
+ "epoch": 43.00958466453674,
+ "grad_norm": 0.056396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0317,
+ "step": 13462
+ },
+ {
+ "epoch": 43.01277955271566,
+ "grad_norm": 0.04736328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0361,
+ "step": 13463
+ },
+ {
+ "epoch": 43.01597444089457,
+ "grad_norm": 0.055908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.036,
+ "step": 13464
+ },
+ {
+ "epoch": 43.019169329073485,
+ "grad_norm": 0.057373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0263,
+ "step": 13465
+ },
+ {
+ "epoch": 43.022364217252395,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0328,
+ "step": 13466
+ },
+ {
+ "epoch": 43.02555910543131,
+ "grad_norm": 0.051025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0213,
+ "step": 13467
+ },
+ {
+ "epoch": 43.02875399361022,
+ "grad_norm": 0.053955078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0284,
+ "step": 13468
+ },
+ {
+ "epoch": 43.031948881789134,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.036,
+ "step": 13469
+ },
+ {
+ "epoch": 43.03514376996805,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0407,
+ "step": 13470
+ },
+ {
+ "epoch": 43.03833865814696,
+ "grad_norm": 0.0556640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0262,
+ "step": 13471
+ },
+ {
+ "epoch": 43.04153354632588,
+ "grad_norm": 0.05224609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0166,
+ "step": 13472
+ },
+ {
+ "epoch": 43.04472843450479,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0334,
+ "step": 13473
+ },
+ {
+ "epoch": 43.04792332268371,
+ "grad_norm": 0.060302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0388,
+ "step": 13474
+ },
+ {
+ "epoch": 43.05111821086262,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0229,
+ "step": 13475
+ },
+ {
+ "epoch": 43.054313099041536,
+ "grad_norm": 0.051513671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0272,
+ "step": 13476
+ },
+ {
+ "epoch": 43.05750798722045,
+ "grad_norm": 0.1044921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0291,
+ "step": 13477
+ },
+ {
+ "epoch": 43.06070287539936,
+ "grad_norm": 0.06103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.036,
+ "step": 13478
+ },
+ {
+ "epoch": 43.063897763578275,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0259,
+ "step": 13479
+ },
+ {
+ "epoch": 43.067092651757186,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.038,
+ "step": 13480
+ },
+ {
+ "epoch": 43.0702875399361,
+ "grad_norm": 0.04736328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0355,
+ "step": 13481
+ },
+ {
+ "epoch": 43.073482428115014,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0253,
+ "step": 13482
+ },
+ {
+ "epoch": 43.07667731629393,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0198,
+ "step": 13483
+ },
+ {
+ "epoch": 43.07987220447284,
+ "grad_norm": 0.052001953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0409,
+ "step": 13484
+ },
+ {
+ "epoch": 43.08306709265176,
+ "grad_norm": 0.062255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0299,
+ "step": 13485
+ },
+ {
+ "epoch": 43.08626198083067,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0348,
+ "step": 13486
+ },
+ {
+ "epoch": 43.08945686900959,
+ "grad_norm": 0.05615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0345,
+ "step": 13487
+ },
+ {
+ "epoch": 43.0926517571885,
+ "grad_norm": 0.05419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0359,
+ "step": 13488
+ },
+ {
+ "epoch": 43.09584664536741,
+ "grad_norm": 0.05517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0412,
+ "step": 13489
+ },
+ {
+ "epoch": 43.09904153354633,
+ "grad_norm": 0.048583984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0397,
+ "step": 13490
+ },
+ {
+ "epoch": 43.10223642172524,
+ "grad_norm": 0.048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0197,
+ "step": 13491
+ },
+ {
+ "epoch": 43.105431309904155,
+ "grad_norm": 0.05419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0304,
+ "step": 13492
+ },
+ {
+ "epoch": 43.108626198083066,
+ "grad_norm": 0.049072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0363,
+ "step": 13493
+ },
+ {
+ "epoch": 43.11182108626198,
+ "grad_norm": 0.0576171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0313,
+ "step": 13494
+ },
+ {
+ "epoch": 43.115015974440894,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0369,
+ "step": 13495
+ },
+ {
+ "epoch": 43.11821086261981,
+ "grad_norm": 0.05615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0348,
+ "step": 13496
+ },
+ {
+ "epoch": 43.12140575079872,
+ "grad_norm": 0.046142578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0402,
+ "step": 13497
+ },
+ {
+ "epoch": 43.12460063897763,
+ "grad_norm": 0.09521484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0312,
+ "step": 13498
+ },
+ {
+ "epoch": 43.12779552715655,
+ "grad_norm": 0.04296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0392,
+ "step": 13499
+ },
+ {
+ "epoch": 43.13099041533546,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0327,
+ "step": 13500
+ },
+ {
+ "epoch": 43.13418530351438,
+ "grad_norm": 0.05078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0297,
+ "step": 13501
+ },
+ {
+ "epoch": 43.13738019169329,
+ "grad_norm": 0.058837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0322,
+ "step": 13502
+ },
+ {
+ "epoch": 43.14057507987221,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0424,
+ "step": 13503
+ },
+ {
+ "epoch": 43.14376996805112,
+ "grad_norm": 0.05322265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0334,
+ "step": 13504
+ },
+ {
+ "epoch": 43.146964856230035,
+ "grad_norm": 0.046630859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0287,
+ "step": 13505
+ },
+ {
+ "epoch": 43.150159744408946,
+ "grad_norm": 0.057861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0367,
+ "step": 13506
+ },
+ {
+ "epoch": 43.153354632587856,
+ "grad_norm": 0.04736328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0249,
+ "step": 13507
+ },
+ {
+ "epoch": 43.156549520766774,
+ "grad_norm": 0.057373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0377,
+ "step": 13508
+ },
+ {
+ "epoch": 43.159744408945684,
+ "grad_norm": 0.056396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0286,
+ "step": 13509
+ },
+ {
+ "epoch": 43.1629392971246,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0289,
+ "step": 13510
+ },
+ {
+ "epoch": 43.16613418530351,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0279,
+ "step": 13511
+ },
+ {
+ "epoch": 43.16932907348243,
+ "grad_norm": 0.046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0269,
+ "step": 13512
+ },
+ {
+ "epoch": 43.17252396166134,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0268,
+ "step": 13513
+ },
+ {
+ "epoch": 43.17571884984026,
+ "grad_norm": 0.051513671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0358,
+ "step": 13514
+ },
+ {
+ "epoch": 43.17891373801917,
+ "grad_norm": 0.050537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.026,
+ "step": 13515
+ },
+ {
+ "epoch": 43.18210862619808,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0329,
+ "step": 13516
+ },
+ {
+ "epoch": 43.185303514377,
+ "grad_norm": 0.052001953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0366,
+ "step": 13517
+ },
+ {
+ "epoch": 43.18849840255591,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0375,
+ "step": 13518
+ },
+ {
+ "epoch": 43.191693290734825,
+ "grad_norm": 0.054931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0341,
+ "step": 13519
+ },
+ {
+ "epoch": 43.194888178913736,
+ "grad_norm": 0.07861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0361,
+ "step": 13520
+ },
+ {
+ "epoch": 43.198083067092654,
+ "grad_norm": 0.044921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0439,
+ "step": 13521
+ },
+ {
+ "epoch": 43.201277955271564,
+ "grad_norm": 0.10205078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0404,
+ "step": 13522
+ },
+ {
+ "epoch": 43.20447284345048,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0289,
+ "step": 13523
+ },
+ {
+ "epoch": 43.20766773162939,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0247,
+ "step": 13524
+ },
+ {
+ "epoch": 43.21086261980831,
+ "grad_norm": 0.08056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0301,
+ "step": 13525
+ },
+ {
+ "epoch": 43.21405750798722,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0408,
+ "step": 13526
+ },
+ {
+ "epoch": 43.21725239616613,
+ "grad_norm": 0.051025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0275,
+ "step": 13527
+ },
+ {
+ "epoch": 43.22044728434505,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0369,
+ "step": 13528
+ },
+ {
+ "epoch": 43.22364217252396,
+ "grad_norm": 0.05859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0337,
+ "step": 13529
+ },
+ {
+ "epoch": 43.22683706070288,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0221,
+ "step": 13530
+ },
+ {
+ "epoch": 43.23003194888179,
+ "grad_norm": 0.048095703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0322,
+ "step": 13531
+ },
+ {
+ "epoch": 43.233226837060705,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0298,
+ "step": 13532
+ },
+ {
+ "epoch": 43.236421725239616,
+ "grad_norm": 0.059326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0318,
+ "step": 13533
+ },
+ {
+ "epoch": 43.239616613418534,
+ "grad_norm": 0.10302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0408,
+ "step": 13534
+ },
+ {
+ "epoch": 43.242811501597444,
+ "grad_norm": 0.047607421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0308,
+ "step": 13535
+ },
+ {
+ "epoch": 43.246006389776355,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.025,
+ "step": 13536
+ },
+ {
+ "epoch": 43.24920127795527,
+ "grad_norm": 0.061279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0444,
+ "step": 13537
+ },
+ {
+ "epoch": 43.25239616613418,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0289,
+ "step": 13538
+ },
+ {
+ "epoch": 43.2555910543131,
+ "grad_norm": 0.049560546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0343,
+ "step": 13539
+ },
+ {
+ "epoch": 43.25878594249201,
+ "grad_norm": 0.05908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0153,
+ "step": 13540
+ },
+ {
+ "epoch": 43.26198083067093,
+ "grad_norm": 0.051513671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0295,
+ "step": 13541
+ },
+ {
+ "epoch": 43.26517571884984,
+ "grad_norm": 0.05712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0342,
+ "step": 13542
+ },
+ {
+ "epoch": 43.26837060702876,
+ "grad_norm": 0.062255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0336,
+ "step": 13543
+ },
+ {
+ "epoch": 43.27156549520767,
+ "grad_norm": 0.059814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0283,
+ "step": 13544
+ },
+ {
+ "epoch": 43.27476038338658,
+ "grad_norm": 0.052001953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0295,
+ "step": 13545
+ },
+ {
+ "epoch": 43.277955271565496,
+ "grad_norm": 0.09765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0324,
+ "step": 13546
+ },
+ {
+ "epoch": 43.281150159744406,
+ "grad_norm": 0.04833984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0291,
+ "step": 13547
+ },
+ {
+ "epoch": 43.284345047923324,
+ "grad_norm": 0.095703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0379,
+ "step": 13548
+ },
+ {
+ "epoch": 43.287539936102235,
+ "grad_norm": 0.05419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0237,
+ "step": 13549
+ },
+ {
+ "epoch": 43.29073482428115,
+ "grad_norm": 0.10546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0231,
+ "step": 13550
+ },
+ {
+ "epoch": 43.29392971246006,
+ "grad_norm": 0.08740234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0414,
+ "step": 13551
+ },
+ {
+ "epoch": 43.29712460063898,
+ "grad_norm": 0.08154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0395,
+ "step": 13552
+ },
+ {
+ "epoch": 43.30031948881789,
+ "grad_norm": 0.09521484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0215,
+ "step": 13553
+ },
+ {
+ "epoch": 43.3035143769968,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0086,
+ "step": 13554
+ },
+ {
+ "epoch": 43.30670926517572,
+ "grad_norm": 0.11474609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0331,
+ "step": 13555
+ },
+ {
+ "epoch": 43.30990415335463,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.032,
+ "step": 13556
+ },
+ {
+ "epoch": 43.31309904153355,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0449,
+ "step": 13557
+ },
+ {
+ "epoch": 43.31629392971246,
+ "grad_norm": 0.080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0347,
+ "step": 13558
+ },
+ {
+ "epoch": 43.319488817891376,
+ "grad_norm": 0.1259765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0271,
+ "step": 13559
+ },
+ {
+ "epoch": 43.322683706070286,
+ "grad_norm": 0.1337890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0351,
+ "step": 13560
+ },
+ {
+ "epoch": 43.325878594249204,
+ "grad_norm": 0.051025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0297,
+ "step": 13561
+ },
+ {
+ "epoch": 43.329073482428115,
+ "grad_norm": 0.08251953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0279,
+ "step": 13562
+ },
+ {
+ "epoch": 43.33226837060703,
+ "grad_norm": 0.059814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0281,
+ "step": 13563
+ },
+ {
+ "epoch": 43.33546325878594,
+ "grad_norm": 0.10888671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0336,
+ "step": 13564
+ },
+ {
+ "epoch": 43.33865814696485,
+ "grad_norm": 0.058349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0303,
+ "step": 13565
+ },
+ {
+ "epoch": 43.34185303514377,
+ "grad_norm": 0.10009765625,
+ "learning_rate": 0.0005,
+ "loss": 1.038,
+ "step": 13566
+ },
+ {
+ "epoch": 43.34504792332268,
+ "grad_norm": 0.051513671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0233,
+ "step": 13567
+ },
+ {
+ "epoch": 43.3482428115016,
+ "grad_norm": 0.11572265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0211,
+ "step": 13568
+ },
+ {
+ "epoch": 43.35143769968051,
+ "grad_norm": 0.048583984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0253,
+ "step": 13569
+ },
+ {
+ "epoch": 43.35463258785943,
+ "grad_norm": 0.1328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0294,
+ "step": 13570
+ },
+ {
+ "epoch": 43.35782747603834,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0365,
+ "step": 13571
+ },
+ {
+ "epoch": 43.361022364217256,
+ "grad_norm": 0.10107421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0274,
+ "step": 13572
+ },
+ {
+ "epoch": 43.364217252396166,
+ "grad_norm": 0.08154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0322,
+ "step": 13573
+ },
+ {
+ "epoch": 43.36741214057508,
+ "grad_norm": 0.09619140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0278,
+ "step": 13574
+ },
+ {
+ "epoch": 43.370607028753994,
+ "grad_norm": 0.08447265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0318,
+ "step": 13575
+ },
+ {
+ "epoch": 43.373801916932905,
+ "grad_norm": 0.09375,
+ "learning_rate": 0.0005,
+ "loss": 1.0206,
+ "step": 13576
+ },
+ {
+ "epoch": 43.37699680511182,
+ "grad_norm": 0.126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.027,
+ "step": 13577
+ },
+ {
+ "epoch": 43.38019169329073,
+ "grad_norm": 0.052490234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0215,
+ "step": 13578
+ },
+ {
+ "epoch": 43.38338658146965,
+ "grad_norm": 0.1171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0359,
+ "step": 13579
+ },
+ {
+ "epoch": 43.38658146964856,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0424,
+ "step": 13580
+ },
+ {
+ "epoch": 43.38977635782748,
+ "grad_norm": 0.126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0307,
+ "step": 13581
+ },
+ {
+ "epoch": 43.39297124600639,
+ "grad_norm": 0.062255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0297,
+ "step": 13582
+ },
+ {
+ "epoch": 43.3961661341853,
+ "grad_norm": 0.09521484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0262,
+ "step": 13583
+ },
+ {
+ "epoch": 43.39936102236422,
+ "grad_norm": 0.05126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0267,
+ "step": 13584
+ },
+ {
+ "epoch": 43.40255591054313,
+ "grad_norm": 0.1103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0239,
+ "step": 13585
+ },
+ {
+ "epoch": 43.405750798722046,
+ "grad_norm": 0.08154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0295,
+ "step": 13586
+ },
+ {
+ "epoch": 43.40894568690096,
+ "grad_norm": 0.1328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0364,
+ "step": 13587
+ },
+ {
+ "epoch": 43.412140575079874,
+ "grad_norm": 0.1435546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0251,
+ "step": 13588
+ },
+ {
+ "epoch": 43.415335463258785,
+ "grad_norm": 0.1201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0281,
+ "step": 13589
+ },
+ {
+ "epoch": 43.4185303514377,
+ "grad_norm": 0.1689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0275,
+ "step": 13590
+ },
+ {
+ "epoch": 43.42172523961661,
+ "grad_norm": 0.1123046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0238,
+ "step": 13591
+ },
+ {
+ "epoch": 43.424920127795524,
+ "grad_norm": 0.10302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0329,
+ "step": 13592
+ },
+ {
+ "epoch": 43.42811501597444,
+ "grad_norm": 0.11083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0374,
+ "step": 13593
+ },
+ {
+ "epoch": 43.43130990415335,
+ "grad_norm": 0.10791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0246,
+ "step": 13594
+ },
+ {
+ "epoch": 43.43450479233227,
+ "grad_norm": 0.1318359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0245,
+ "step": 13595
+ },
+ {
+ "epoch": 43.43769968051118,
+ "grad_norm": 0.1396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.025,
+ "step": 13596
+ },
+ {
+ "epoch": 43.4408945686901,
+ "grad_norm": 0.1298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0135,
+ "step": 13597
+ },
+ {
+ "epoch": 43.44408945686901,
+ "grad_norm": 0.126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0221,
+ "step": 13598
+ },
+ {
+ "epoch": 43.447284345047926,
+ "grad_norm": 0.10205078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0352,
+ "step": 13599
+ },
+ {
+ "epoch": 43.45047923322684,
+ "grad_norm": 0.0927734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0353,
+ "step": 13600
+ },
+ {
+ "epoch": 43.453674121405754,
+ "grad_norm": 0.1171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0257,
+ "step": 13601
+ },
+ {
+ "epoch": 43.456869009584665,
+ "grad_norm": 0.1455078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0301,
+ "step": 13602
+ },
+ {
+ "epoch": 43.460063897763575,
+ "grad_norm": 0.11083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0314,
+ "step": 13603
+ },
+ {
+ "epoch": 43.46325878594249,
+ "grad_norm": 0.1103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0399,
+ "step": 13604
+ },
+ {
+ "epoch": 43.466453674121404,
+ "grad_norm": 0.08447265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0245,
+ "step": 13605
+ },
+ {
+ "epoch": 43.46964856230032,
+ "grad_norm": 0.0615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.029,
+ "step": 13606
+ },
+ {
+ "epoch": 43.47284345047923,
+ "grad_norm": 0.053955078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0244,
+ "step": 13607
+ },
+ {
+ "epoch": 43.47603833865815,
+ "grad_norm": 0.08251953125,
+ "learning_rate": 0.0005,
+ "loss": 1.037,
+ "step": 13608
+ },
+ {
+ "epoch": 43.47923322683706,
+ "grad_norm": 0.0576171875,
+ "learning_rate": 0.0005,
+ "loss": 1.023,
+ "step": 13609
+ },
+ {
+ "epoch": 43.48242811501598,
+ "grad_norm": 0.057861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0334,
+ "step": 13610
+ },
+ {
+ "epoch": 43.48562300319489,
+ "grad_norm": 0.09423828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0462,
+ "step": 13611
+ },
+ {
+ "epoch": 43.4888178913738,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0238,
+ "step": 13612
+ },
+ {
+ "epoch": 43.49201277955272,
+ "grad_norm": 0.08740234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0284,
+ "step": 13613
+ },
+ {
+ "epoch": 43.49520766773163,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0301,
+ "step": 13614
+ },
+ {
+ "epoch": 43.498402555910545,
+ "grad_norm": 0.062255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0305,
+ "step": 13615
+ },
+ {
+ "epoch": 43.501597444089455,
+ "grad_norm": 0.05712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.039,
+ "step": 13616
+ },
+ {
+ "epoch": 43.50479233226837,
+ "grad_norm": 0.049560546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0281,
+ "step": 13617
+ },
+ {
+ "epoch": 43.50798722044728,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0248,
+ "step": 13618
+ },
+ {
+ "epoch": 43.5111821086262,
+ "grad_norm": 0.04931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0244,
+ "step": 13619
+ },
+ {
+ "epoch": 43.51437699680511,
+ "grad_norm": 0.06005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0375,
+ "step": 13620
+ },
+ {
+ "epoch": 43.51757188498402,
+ "grad_norm": 0.045654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.031,
+ "step": 13621
+ },
+ {
+ "epoch": 43.52076677316294,
+ "grad_norm": 0.046142578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0331,
+ "step": 13622
+ },
+ {
+ "epoch": 43.52396166134185,
+ "grad_norm": 0.0556640625,
+ "learning_rate": 0.0005,
+ "loss": 1.02,
+ "step": 13623
+ },
+ {
+ "epoch": 43.52715654952077,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0283,
+ "step": 13624
+ },
+ {
+ "epoch": 43.53035143769968,
+ "grad_norm": 0.052001953125,
+ "learning_rate": 0.0005,
+ "loss": 1.024,
+ "step": 13625
+ },
+ {
+ "epoch": 43.533546325878596,
+ "grad_norm": 0.058349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0322,
+ "step": 13626
+ },
+ {
+ "epoch": 43.53674121405751,
+ "grad_norm": 0.0576171875,
+ "learning_rate": 0.0005,
+ "loss": 1.033,
+ "step": 13627
+ },
+ {
+ "epoch": 43.539936102236425,
+ "grad_norm": 0.0546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0259,
+ "step": 13628
+ },
+ {
+ "epoch": 43.543130990415335,
+ "grad_norm": 0.054931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0275,
+ "step": 13629
+ },
+ {
+ "epoch": 43.546325878594246,
+ "grad_norm": 0.0859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0285,
+ "step": 13630
+ },
+ {
+ "epoch": 43.54952076677316,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0307,
+ "step": 13631
+ },
+ {
+ "epoch": 43.552715654952074,
+ "grad_norm": 0.1162109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0271,
+ "step": 13632
+ },
+ {
+ "epoch": 43.55591054313099,
+ "grad_norm": 0.057861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0298,
+ "step": 13633
+ },
+ {
+ "epoch": 43.5591054313099,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0281,
+ "step": 13634
+ },
+ {
+ "epoch": 43.56230031948882,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0308,
+ "step": 13635
+ },
+ {
+ "epoch": 43.56549520766773,
+ "grad_norm": 0.08056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0321,
+ "step": 13636
+ },
+ {
+ "epoch": 43.56869009584665,
+ "grad_norm": 0.051513671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0343,
+ "step": 13637
+ },
+ {
+ "epoch": 43.57188498402556,
+ "grad_norm": 0.10791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0196,
+ "step": 13638
+ },
+ {
+ "epoch": 43.575079872204476,
+ "grad_norm": 0.05908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0374,
+ "step": 13639
+ },
+ {
+ "epoch": 43.57827476038339,
+ "grad_norm": 0.103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0404,
+ "step": 13640
+ },
+ {
+ "epoch": 43.5814696485623,
+ "grad_norm": 0.0986328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0305,
+ "step": 13641
+ },
+ {
+ "epoch": 43.584664536741215,
+ "grad_norm": 0.059326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0203,
+ "step": 13642
+ },
+ {
+ "epoch": 43.587859424920126,
+ "grad_norm": 0.11376953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0377,
+ "step": 13643
+ },
+ {
+ "epoch": 43.59105431309904,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0278,
+ "step": 13644
+ },
+ {
+ "epoch": 43.594249201277954,
+ "grad_norm": 0.09326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0343,
+ "step": 13645
+ },
+ {
+ "epoch": 43.59744408945687,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0357,
+ "step": 13646
+ },
+ {
+ "epoch": 43.60063897763578,
+ "grad_norm": 0.0546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0337,
+ "step": 13647
+ },
+ {
+ "epoch": 43.6038338658147,
+ "grad_norm": 0.10400390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0218,
+ "step": 13648
+ },
+ {
+ "epoch": 43.60702875399361,
+ "grad_norm": 0.054443359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0371,
+ "step": 13649
+ },
+ {
+ "epoch": 43.61022364217252,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0201,
+ "step": 13650
+ },
+ {
+ "epoch": 43.61341853035144,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0376,
+ "step": 13651
+ },
+ {
+ "epoch": 43.61661341853035,
+ "grad_norm": 0.0888671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0295,
+ "step": 13652
+ },
+ {
+ "epoch": 43.61980830670927,
+ "grad_norm": 0.0537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0244,
+ "step": 13653
+ },
+ {
+ "epoch": 43.62300319488818,
+ "grad_norm": 0.1357421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0283,
+ "step": 13654
+ },
+ {
+ "epoch": 43.626198083067095,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0347,
+ "step": 13655
+ },
+ {
+ "epoch": 43.629392971246006,
+ "grad_norm": 0.10205078125,
+ "learning_rate": 0.0005,
+ "loss": 1.019,
+ "step": 13656
+ },
+ {
+ "epoch": 43.63258785942492,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0346,
+ "step": 13657
+ },
+ {
+ "epoch": 43.635782747603834,
+ "grad_norm": 0.052734375,
+ "learning_rate": 0.0005,
+ "loss": 1.032,
+ "step": 13658
+ },
+ {
+ "epoch": 43.638977635782744,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0412,
+ "step": 13659
+ },
+ {
+ "epoch": 43.64217252396166,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0199,
+ "step": 13660
+ },
+ {
+ "epoch": 43.64536741214057,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0233,
+ "step": 13661
+ },
+ {
+ "epoch": 43.64856230031949,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0352,
+ "step": 13662
+ },
+ {
+ "epoch": 43.6517571884984,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0327,
+ "step": 13663
+ },
+ {
+ "epoch": 43.65495207667732,
+ "grad_norm": 0.045654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0428,
+ "step": 13664
+ },
+ {
+ "epoch": 43.65814696485623,
+ "grad_norm": 0.05810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0272,
+ "step": 13665
+ },
+ {
+ "epoch": 43.66134185303515,
+ "grad_norm": 0.091796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0342,
+ "step": 13666
+ },
+ {
+ "epoch": 43.66453674121406,
+ "grad_norm": 0.05224609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0282,
+ "step": 13667
+ },
+ {
+ "epoch": 43.66773162939297,
+ "grad_norm": 0.10498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0167,
+ "step": 13668
+ },
+ {
+ "epoch": 43.670926517571885,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0205,
+ "step": 13669
+ },
+ {
+ "epoch": 43.674121405750796,
+ "grad_norm": 0.10107421875,
+ "learning_rate": 0.0005,
+ "loss": 1.026,
+ "step": 13670
+ },
+ {
+ "epoch": 43.677316293929714,
+ "grad_norm": 0.0830078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0425,
+ "step": 13671
+ },
+ {
+ "epoch": 43.680511182108624,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0282,
+ "step": 13672
+ },
+ {
+ "epoch": 43.68370607028754,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0268,
+ "step": 13673
+ },
+ {
+ "epoch": 43.68690095846645,
+ "grad_norm": 0.08447265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0367,
+ "step": 13674
+ },
+ {
+ "epoch": 43.69009584664537,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0277,
+ "step": 13675
+ },
+ {
+ "epoch": 43.69329073482428,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0366,
+ "step": 13676
+ },
+ {
+ "epoch": 43.6964856230032,
+ "grad_norm": 0.048583984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0396,
+ "step": 13677
+ },
+ {
+ "epoch": 43.69968051118211,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0233,
+ "step": 13678
+ },
+ {
+ "epoch": 43.70287539936102,
+ "grad_norm": 0.049560546875,
+ "learning_rate": 0.0005,
+ "loss": 1.018,
+ "step": 13679
+ },
+ {
+ "epoch": 43.70607028753994,
+ "grad_norm": 0.050537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0319,
+ "step": 13680
+ },
+ {
+ "epoch": 43.70926517571885,
+ "grad_norm": 0.05517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.033,
+ "step": 13681
+ },
+ {
+ "epoch": 43.712460063897765,
+ "grad_norm": 0.0537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0345,
+ "step": 13682
+ },
+ {
+ "epoch": 43.715654952076676,
+ "grad_norm": 0.047119140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0326,
+ "step": 13683
+ },
+ {
+ "epoch": 43.718849840255594,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0256,
+ "step": 13684
+ },
+ {
+ "epoch": 43.722044728434504,
+ "grad_norm": 0.05419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0305,
+ "step": 13685
+ },
+ {
+ "epoch": 43.72523961661342,
+ "grad_norm": 0.047607421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0425,
+ "step": 13686
+ },
+ {
+ "epoch": 43.72843450479233,
+ "grad_norm": 0.05810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0242,
+ "step": 13687
+ },
+ {
+ "epoch": 43.73162939297124,
+ "grad_norm": 0.048583984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0296,
+ "step": 13688
+ },
+ {
+ "epoch": 43.73482428115016,
+ "grad_norm": 0.05419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0282,
+ "step": 13689
+ },
+ {
+ "epoch": 43.73801916932907,
+ "grad_norm": 0.107421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0222,
+ "step": 13690
+ },
+ {
+ "epoch": 43.74121405750799,
+ "grad_norm": 0.080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0355,
+ "step": 13691
+ },
+ {
+ "epoch": 43.7444089456869,
+ "grad_norm": 0.09228515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0373,
+ "step": 13692
+ },
+ {
+ "epoch": 43.74760383386582,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0221,
+ "step": 13693
+ },
+ {
+ "epoch": 43.75079872204473,
+ "grad_norm": 0.0849609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0301,
+ "step": 13694
+ },
+ {
+ "epoch": 43.753993610223645,
+ "grad_norm": 0.09423828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0286,
+ "step": 13695
+ },
+ {
+ "epoch": 43.757188498402556,
+ "grad_norm": 0.05615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0296,
+ "step": 13696
+ },
+ {
+ "epoch": 43.760383386581466,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0227,
+ "step": 13697
+ },
+ {
+ "epoch": 43.763578274760384,
+ "grad_norm": 0.046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0226,
+ "step": 13698
+ },
+ {
+ "epoch": 43.766773162939295,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0315,
+ "step": 13699
+ },
+ {
+ "epoch": 43.76996805111821,
+ "grad_norm": 0.05322265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0204,
+ "step": 13700
+ },
+ {
+ "epoch": 43.77316293929712,
+ "grad_norm": 0.053466796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0284,
+ "step": 13701
+ },
+ {
+ "epoch": 43.77635782747604,
+ "grad_norm": 0.0517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0428,
+ "step": 13702
+ },
+ {
+ "epoch": 43.77955271565495,
+ "grad_norm": 0.061279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0395,
+ "step": 13703
+ },
+ {
+ "epoch": 43.78274760383387,
+ "grad_norm": 0.059326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0319,
+ "step": 13704
+ },
+ {
+ "epoch": 43.78594249201278,
+ "grad_norm": 0.083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.025,
+ "step": 13705
+ },
+ {
+ "epoch": 43.78913738019169,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0344,
+ "step": 13706
+ },
+ {
+ "epoch": 43.79233226837061,
+ "grad_norm": 0.09326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0242,
+ "step": 13707
+ },
+ {
+ "epoch": 43.79552715654952,
+ "grad_norm": 0.051025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0336,
+ "step": 13708
+ },
+ {
+ "epoch": 43.798722044728436,
+ "grad_norm": 0.10107421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0063,
+ "step": 13709
+ },
+ {
+ "epoch": 43.801916932907346,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0341,
+ "step": 13710
+ },
+ {
+ "epoch": 43.805111821086264,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0444,
+ "step": 13711
+ },
+ {
+ "epoch": 43.808306709265175,
+ "grad_norm": 0.052490234375,
+ "learning_rate": 0.0005,
+ "loss": 1.036,
+ "step": 13712
+ },
+ {
+ "epoch": 43.81150159744409,
+ "grad_norm": 0.054443359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0374,
+ "step": 13713
+ },
+ {
+ "epoch": 43.814696485623,
+ "grad_norm": 0.061279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0263,
+ "step": 13714
+ },
+ {
+ "epoch": 43.81789137380191,
+ "grad_norm": 0.0830078125,
+ "learning_rate": 0.0005,
+ "loss": 1.03,
+ "step": 13715
+ },
+ {
+ "epoch": 43.82108626198083,
+ "grad_norm": 0.050537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0303,
+ "step": 13716
+ },
+ {
+ "epoch": 43.82428115015974,
+ "grad_norm": 0.09423828125,
+ "learning_rate": 0.0005,
+ "loss": 1.025,
+ "step": 13717
+ },
+ {
+ "epoch": 43.82747603833866,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0413,
+ "step": 13718
+ },
+ {
+ "epoch": 43.83067092651757,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0275,
+ "step": 13719
+ },
+ {
+ "epoch": 43.83386581469649,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0205,
+ "step": 13720
+ },
+ {
+ "epoch": 43.8370607028754,
+ "grad_norm": 0.061279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0213,
+ "step": 13721
+ },
+ {
+ "epoch": 43.840255591054316,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0154,
+ "step": 13722
+ },
+ {
+ "epoch": 43.843450479233226,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0343,
+ "step": 13723
+ },
+ {
+ "epoch": 43.846645367412144,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0317,
+ "step": 13724
+ },
+ {
+ "epoch": 43.849840255591054,
+ "grad_norm": 0.0830078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0375,
+ "step": 13725
+ },
+ {
+ "epoch": 43.853035143769965,
+ "grad_norm": 0.061279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0374,
+ "step": 13726
+ },
+ {
+ "epoch": 43.85623003194888,
+ "grad_norm": 0.1396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.029,
+ "step": 13727
+ },
+ {
+ "epoch": 43.85942492012779,
+ "grad_norm": 0.051513671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0307,
+ "step": 13728
+ },
+ {
+ "epoch": 43.86261980830671,
+ "grad_norm": 0.091796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0301,
+ "step": 13729
+ },
+ {
+ "epoch": 43.86581469648562,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0323,
+ "step": 13730
+ },
+ {
+ "epoch": 43.86900958466454,
+ "grad_norm": 0.061767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0259,
+ "step": 13731
+ },
+ {
+ "epoch": 43.87220447284345,
+ "grad_norm": 0.10009765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0194,
+ "step": 13732
+ },
+ {
+ "epoch": 43.87539936102237,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0239,
+ "step": 13733
+ },
+ {
+ "epoch": 43.87859424920128,
+ "grad_norm": 0.08447265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0283,
+ "step": 13734
+ },
+ {
+ "epoch": 43.88178913738019,
+ "grad_norm": 0.0810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0284,
+ "step": 13735
+ },
+ {
+ "epoch": 43.884984025559106,
+ "grad_norm": 0.05810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0324,
+ "step": 13736
+ },
+ {
+ "epoch": 43.88817891373802,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0251,
+ "step": 13737
+ },
+ {
+ "epoch": 43.891373801916934,
+ "grad_norm": 0.054443359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0429,
+ "step": 13738
+ },
+ {
+ "epoch": 43.894568690095845,
+ "grad_norm": 0.058837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0173,
+ "step": 13739
+ },
+ {
+ "epoch": 43.89776357827476,
+ "grad_norm": 0.053955078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0173,
+ "step": 13740
+ },
+ {
+ "epoch": 43.90095846645367,
+ "grad_norm": 0.05419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0349,
+ "step": 13741
+ },
+ {
+ "epoch": 43.90415335463259,
+ "grad_norm": 0.08251953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0353,
+ "step": 13742
+ },
+ {
+ "epoch": 43.9073482428115,
+ "grad_norm": 0.05517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0362,
+ "step": 13743
+ },
+ {
+ "epoch": 43.91054313099041,
+ "grad_norm": 0.049072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0314,
+ "step": 13744
+ },
+ {
+ "epoch": 43.91373801916933,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0316,
+ "step": 13745
+ },
+ {
+ "epoch": 43.91693290734824,
+ "grad_norm": 0.0458984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0379,
+ "step": 13746
+ },
+ {
+ "epoch": 43.92012779552716,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0345,
+ "step": 13747
+ },
+ {
+ "epoch": 43.92332268370607,
+ "grad_norm": 0.05615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0156,
+ "step": 13748
+ },
+ {
+ "epoch": 43.926517571884986,
+ "grad_norm": 0.059814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0335,
+ "step": 13749
+ },
+ {
+ "epoch": 43.9297124600639,
+ "grad_norm": 0.061279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0355,
+ "step": 13750
+ },
+ {
+ "epoch": 43.932907348242814,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0271,
+ "step": 13751
+ },
+ {
+ "epoch": 43.936102236421725,
+ "grad_norm": 0.11767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0299,
+ "step": 13752
+ },
+ {
+ "epoch": 43.93929712460064,
+ "grad_norm": 0.08837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0434,
+ "step": 13753
+ },
+ {
+ "epoch": 43.94249201277955,
+ "grad_norm": 0.08837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0211,
+ "step": 13754
+ },
+ {
+ "epoch": 43.945686900958464,
+ "grad_norm": 0.0869140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0285,
+ "step": 13755
+ },
+ {
+ "epoch": 43.94888178913738,
+ "grad_norm": 0.08056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0368,
+ "step": 13756
+ },
+ {
+ "epoch": 43.95207667731629,
+ "grad_norm": 0.11279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0257,
+ "step": 13757
+ },
+ {
+ "epoch": 43.95527156549521,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.03,
+ "step": 13758
+ },
+ {
+ "epoch": 43.95846645367412,
+ "grad_norm": 0.058837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0269,
+ "step": 13759
+ },
+ {
+ "epoch": 43.96166134185304,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.038,
+ "step": 13760
+ },
+ {
+ "epoch": 43.96485623003195,
+ "grad_norm": 0.0966796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0299,
+ "step": 13761
+ },
+ {
+ "epoch": 43.968051118210866,
+ "grad_norm": 0.061767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0255,
+ "step": 13762
+ },
+ {
+ "epoch": 43.97124600638978,
+ "grad_norm": 0.09521484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0346,
+ "step": 13763
+ },
+ {
+ "epoch": 43.97444089456869,
+ "grad_norm": 0.053466796875,
+ "learning_rate": 0.0005,
+ "loss": 1.029,
+ "step": 13764
+ },
+ {
+ "epoch": 43.977635782747605,
+ "grad_norm": 0.05517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.026,
+ "step": 13765
+ },
+ {
+ "epoch": 43.980830670926515,
+ "grad_norm": 0.083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0332,
+ "step": 13766
+ },
+ {
+ "epoch": 43.98402555910543,
+ "grad_norm": 0.053955078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0288,
+ "step": 13767
+ },
+ {
+ "epoch": 43.98722044728434,
+ "grad_norm": 0.050048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0312,
+ "step": 13768
+ },
+ {
+ "epoch": 43.99041533546326,
+ "grad_norm": 0.05712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0286,
+ "step": 13769
+ },
+ {
+ "epoch": 43.99361022364217,
+ "grad_norm": 0.059814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0171,
+ "step": 13770
+ },
+ {
+ "epoch": 43.99680511182109,
+ "grad_norm": 0.04833984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0277,
+ "step": 13771
+ },
+ {
+ "epoch": 44.0,
+ "grad_norm": 0.055908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0269,
+ "step": 13772
+ },
+ {
+ "epoch": 44.00319488817891,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0284,
+ "step": 13773
+ },
+ {
+ "epoch": 44.00638977635783,
+ "grad_norm": 0.11328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0305,
+ "step": 13774
+ },
+ {
+ "epoch": 44.00958466453674,
+ "grad_norm": 0.054931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0364,
+ "step": 13775
+ },
+ {
+ "epoch": 44.01277955271566,
+ "grad_norm": 0.048095703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0316,
+ "step": 13776
+ },
+ {
+ "epoch": 44.01597444089457,
+ "grad_norm": 0.04443359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0232,
+ "step": 13777
+ },
+ {
+ "epoch": 44.019169329073485,
+ "grad_norm": 0.06201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0324,
+ "step": 13778
+ },
+ {
+ "epoch": 44.022364217252395,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0136,
+ "step": 13779
+ },
+ {
+ "epoch": 44.02555910543131,
+ "grad_norm": 0.05126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0269,
+ "step": 13780
+ },
+ {
+ "epoch": 44.02875399361022,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0184,
+ "step": 13781
+ },
+ {
+ "epoch": 44.031948881789134,
+ "grad_norm": 0.05615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0312,
+ "step": 13782
+ },
+ {
+ "epoch": 44.03514376996805,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0302,
+ "step": 13783
+ },
+ {
+ "epoch": 44.03833865814696,
+ "grad_norm": 0.060791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0275,
+ "step": 13784
+ },
+ {
+ "epoch": 44.04153354632588,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.032,
+ "step": 13785
+ },
+ {
+ "epoch": 44.04472843450479,
+ "grad_norm": 0.04931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0206,
+ "step": 13786
+ },
+ {
+ "epoch": 44.04792332268371,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0212,
+ "step": 13787
+ },
+ {
+ "epoch": 44.05111821086262,
+ "grad_norm": 0.057373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0192,
+ "step": 13788
+ },
+ {
+ "epoch": 44.054313099041536,
+ "grad_norm": 0.053466796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0179,
+ "step": 13789
+ },
+ {
+ "epoch": 44.05750798722045,
+ "grad_norm": 0.0859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0304,
+ "step": 13790
+ },
+ {
+ "epoch": 44.06070287539936,
+ "grad_norm": 0.05078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0246,
+ "step": 13791
+ },
+ {
+ "epoch": 44.063897763578275,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0347,
+ "step": 13792
+ },
+ {
+ "epoch": 44.067092651757186,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.024,
+ "step": 13793
+ },
+ {
+ "epoch": 44.0702875399361,
+ "grad_norm": 0.054931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0214,
+ "step": 13794
+ },
+ {
+ "epoch": 44.073482428115014,
+ "grad_norm": 0.05224609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0282,
+ "step": 13795
+ },
+ {
+ "epoch": 44.07667731629393,
+ "grad_norm": 0.0927734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0296,
+ "step": 13796
+ },
+ {
+ "epoch": 44.07987220447284,
+ "grad_norm": 0.08056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0264,
+ "step": 13797
+ },
+ {
+ "epoch": 44.08306709265176,
+ "grad_norm": 0.054931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0306,
+ "step": 13798
+ },
+ {
+ "epoch": 44.08626198083067,
+ "grad_norm": 0.1201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0168,
+ "step": 13799
+ },
+ {
+ "epoch": 44.08945686900959,
+ "grad_norm": 0.0546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0359,
+ "step": 13800
+ },
+ {
+ "epoch": 44.0926517571885,
+ "grad_norm": 0.080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0344,
+ "step": 13801
+ },
+ {
+ "epoch": 44.09584664536741,
+ "grad_norm": 0.1279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0217,
+ "step": 13802
+ },
+ {
+ "epoch": 44.09904153354633,
+ "grad_norm": 0.0537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0368,
+ "step": 13803
+ },
+ {
+ "epoch": 44.10223642172524,
+ "grad_norm": 0.109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0287,
+ "step": 13804
+ },
+ {
+ "epoch": 44.105431309904155,
+ "grad_norm": 0.134765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0355,
+ "step": 13805
+ },
+ {
+ "epoch": 44.108626198083066,
+ "grad_norm": 0.21875,
+ "learning_rate": 0.0005,
+ "loss": 1.043,
+ "step": 13806
+ },
+ {
+ "epoch": 44.11182108626198,
+ "grad_norm": 0.1044921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0399,
+ "step": 13807
+ },
+ {
+ "epoch": 44.115015974440894,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0231,
+ "step": 13808
+ },
+ {
+ "epoch": 44.11821086261981,
+ "grad_norm": 0.150390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0361,
+ "step": 13809
+ },
+ {
+ "epoch": 44.12140575079872,
+ "grad_norm": 0.0576171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0229,
+ "step": 13810
+ },
+ {
+ "epoch": 44.12460063897763,
+ "grad_norm": 0.08740234375,
+ "learning_rate": 0.0005,
+ "loss": 1.038,
+ "step": 13811
+ },
+ {
+ "epoch": 44.12779552715655,
+ "grad_norm": 0.1376953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0241,
+ "step": 13812
+ },
+ {
+ "epoch": 44.13099041533546,
+ "grad_norm": 0.1142578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0324,
+ "step": 13813
+ },
+ {
+ "epoch": 44.13418530351438,
+ "grad_norm": 0.10986328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0349,
+ "step": 13814
+ },
+ {
+ "epoch": 44.13738019169329,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0217,
+ "step": 13815
+ },
+ {
+ "epoch": 44.14057507987221,
+ "grad_norm": 0.09912109375,
+ "learning_rate": 0.0005,
+ "loss": 1.027,
+ "step": 13816
+ },
+ {
+ "epoch": 44.14376996805112,
+ "grad_norm": 0.0927734375,
+ "learning_rate": 0.0005,
+ "loss": 1.029,
+ "step": 13817
+ },
+ {
+ "epoch": 44.146964856230035,
+ "grad_norm": 0.0869140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0376,
+ "step": 13818
+ },
+ {
+ "epoch": 44.150159744408946,
+ "grad_norm": 0.08544921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0243,
+ "step": 13819
+ },
+ {
+ "epoch": 44.153354632587856,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0367,
+ "step": 13820
+ },
+ {
+ "epoch": 44.156549520766774,
+ "grad_norm": 0.054443359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0194,
+ "step": 13821
+ },
+ {
+ "epoch": 44.159744408945684,
+ "grad_norm": 0.056396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0219,
+ "step": 13822
+ },
+ {
+ "epoch": 44.1629392971246,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0337,
+ "step": 13823
+ },
+ {
+ "epoch": 44.16613418530351,
+ "grad_norm": 0.059326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0287,
+ "step": 13824
+ },
+ {
+ "epoch": 44.16932907348243,
+ "grad_norm": 0.11572265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0299,
+ "step": 13825
+ },
+ {
+ "epoch": 44.17252396166134,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0355,
+ "step": 13826
+ },
+ {
+ "epoch": 44.17571884984026,
+ "grad_norm": 0.09375,
+ "learning_rate": 0.0005,
+ "loss": 1.0268,
+ "step": 13827
+ },
+ {
+ "epoch": 44.17891373801917,
+ "grad_norm": 0.1044921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0348,
+ "step": 13828
+ },
+ {
+ "epoch": 44.18210862619808,
+ "grad_norm": 0.08056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0242,
+ "step": 13829
+ },
+ {
+ "epoch": 44.185303514377,
+ "grad_norm": 0.08056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0359,
+ "step": 13830
+ },
+ {
+ "epoch": 44.18849840255591,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0314,
+ "step": 13831
+ },
+ {
+ "epoch": 44.191693290734825,
+ "grad_norm": 0.056396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0393,
+ "step": 13832
+ },
+ {
+ "epoch": 44.194888178913736,
+ "grad_norm": 0.052978515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0459,
+ "step": 13833
+ },
+ {
+ "epoch": 44.198083067092654,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0338,
+ "step": 13834
+ },
+ {
+ "epoch": 44.201277955271564,
+ "grad_norm": 0.123046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0327,
+ "step": 13835
+ },
+ {
+ "epoch": 44.20447284345048,
+ "grad_norm": 0.0517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0325,
+ "step": 13836
+ },
+ {
+ "epoch": 44.20766773162939,
+ "grad_norm": 0.10107421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0262,
+ "step": 13837
+ },
+ {
+ "epoch": 44.21086261980831,
+ "grad_norm": 0.052001953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0329,
+ "step": 13838
+ },
+ {
+ "epoch": 44.21405750798722,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0281,
+ "step": 13839
+ },
+ {
+ "epoch": 44.21725239616613,
+ "grad_norm": 0.05224609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0215,
+ "step": 13840
+ },
+ {
+ "epoch": 44.22044728434505,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0347,
+ "step": 13841
+ },
+ {
+ "epoch": 44.22364217252396,
+ "grad_norm": 0.049072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.027,
+ "step": 13842
+ },
+ {
+ "epoch": 44.22683706070288,
+ "grad_norm": 0.08935546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0317,
+ "step": 13843
+ },
+ {
+ "epoch": 44.23003194888179,
+ "grad_norm": 0.047607421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0288,
+ "step": 13844
+ },
+ {
+ "epoch": 44.233226837060705,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0265,
+ "step": 13845
+ },
+ {
+ "epoch": 44.236421725239616,
+ "grad_norm": 0.06005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0332,
+ "step": 13846
+ },
+ {
+ "epoch": 44.239616613418534,
+ "grad_norm": 0.0927734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0328,
+ "step": 13847
+ },
+ {
+ "epoch": 44.242811501597444,
+ "grad_norm": 0.052978515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0288,
+ "step": 13848
+ },
+ {
+ "epoch": 44.246006389776355,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0468,
+ "step": 13849
+ },
+ {
+ "epoch": 44.24920127795527,
+ "grad_norm": 0.04931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0305,
+ "step": 13850
+ },
+ {
+ "epoch": 44.25239616613418,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0263,
+ "step": 13851
+ },
+ {
+ "epoch": 44.2555910543131,
+ "grad_norm": 0.05126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0364,
+ "step": 13852
+ },
+ {
+ "epoch": 44.25878594249201,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0379,
+ "step": 13853
+ },
+ {
+ "epoch": 44.26198083067093,
+ "grad_norm": 0.04833984375,
+ "learning_rate": 0.0005,
+ "loss": 1.028,
+ "step": 13854
+ },
+ {
+ "epoch": 44.26517571884984,
+ "grad_norm": 0.051513671875,
+ "learning_rate": 0.0005,
+ "loss": 1.036,
+ "step": 13855
+ },
+ {
+ "epoch": 44.26837060702876,
+ "grad_norm": 0.0576171875,
+ "learning_rate": 0.0005,
+ "loss": 1.019,
+ "step": 13856
+ },
+ {
+ "epoch": 44.27156549520767,
+ "grad_norm": 0.045166015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0309,
+ "step": 13857
+ },
+ {
+ "epoch": 44.27476038338658,
+ "grad_norm": 0.08740234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0247,
+ "step": 13858
+ },
+ {
+ "epoch": 44.277955271565496,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0308,
+ "step": 13859
+ },
+ {
+ "epoch": 44.281150159744406,
+ "grad_norm": 0.08203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0258,
+ "step": 13860
+ },
+ {
+ "epoch": 44.284345047923324,
+ "grad_norm": 0.056396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.04,
+ "step": 13861
+ },
+ {
+ "epoch": 44.287539936102235,
+ "grad_norm": 0.05419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0254,
+ "step": 13862
+ },
+ {
+ "epoch": 44.29073482428115,
+ "grad_norm": 0.048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0417,
+ "step": 13863
+ },
+ {
+ "epoch": 44.29392971246006,
+ "grad_norm": 0.0595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0199,
+ "step": 13864
+ },
+ {
+ "epoch": 44.29712460063898,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0291,
+ "step": 13865
+ },
+ {
+ "epoch": 44.30031948881789,
+ "grad_norm": 0.04931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0368,
+ "step": 13866
+ },
+ {
+ "epoch": 44.3035143769968,
+ "grad_norm": 0.05126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0345,
+ "step": 13867
+ },
+ {
+ "epoch": 44.30670926517572,
+ "grad_norm": 0.0517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0335,
+ "step": 13868
+ },
+ {
+ "epoch": 44.30990415335463,
+ "grad_norm": 0.053955078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0436,
+ "step": 13869
+ },
+ {
+ "epoch": 44.31309904153355,
+ "grad_norm": 0.08837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0217,
+ "step": 13870
+ },
+ {
+ "epoch": 44.31629392971246,
+ "grad_norm": 0.05615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0357,
+ "step": 13871
+ },
+ {
+ "epoch": 44.319488817891376,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0374,
+ "step": 13872
+ },
+ {
+ "epoch": 44.322683706070286,
+ "grad_norm": 0.0615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0271,
+ "step": 13873
+ },
+ {
+ "epoch": 44.325878594249204,
+ "grad_norm": 0.055908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0251,
+ "step": 13874
+ },
+ {
+ "epoch": 44.329073482428115,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0317,
+ "step": 13875
+ },
+ {
+ "epoch": 44.33226837060703,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0293,
+ "step": 13876
+ },
+ {
+ "epoch": 44.33546325878594,
+ "grad_norm": 0.091796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0319,
+ "step": 13877
+ },
+ {
+ "epoch": 44.33865814696485,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0391,
+ "step": 13878
+ },
+ {
+ "epoch": 44.34185303514377,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0244,
+ "step": 13879
+ },
+ {
+ "epoch": 44.34504792332268,
+ "grad_norm": 0.09326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0375,
+ "step": 13880
+ },
+ {
+ "epoch": 44.3482428115016,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0251,
+ "step": 13881
+ },
+ {
+ "epoch": 44.35143769968051,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0167,
+ "step": 13882
+ },
+ {
+ "epoch": 44.35463258785943,
+ "grad_norm": 0.062255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0277,
+ "step": 13883
+ },
+ {
+ "epoch": 44.35782747603834,
+ "grad_norm": 0.059814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0275,
+ "step": 13884
+ },
+ {
+ "epoch": 44.361022364217256,
+ "grad_norm": 0.06201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0399,
+ "step": 13885
+ },
+ {
+ "epoch": 44.364217252396166,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0288,
+ "step": 13886
+ },
+ {
+ "epoch": 44.36741214057508,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0273,
+ "step": 13887
+ },
+ {
+ "epoch": 44.370607028753994,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0286,
+ "step": 13888
+ },
+ {
+ "epoch": 44.373801916932905,
+ "grad_norm": 0.09619140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0373,
+ "step": 13889
+ },
+ {
+ "epoch": 44.37699680511182,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0233,
+ "step": 13890
+ },
+ {
+ "epoch": 44.38019169329073,
+ "grad_norm": 0.058837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0422,
+ "step": 13891
+ },
+ {
+ "epoch": 44.38338658146965,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0232,
+ "step": 13892
+ },
+ {
+ "epoch": 44.38658146964856,
+ "grad_norm": 0.052490234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0335,
+ "step": 13893
+ },
+ {
+ "epoch": 44.38977635782748,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0333,
+ "step": 13894
+ },
+ {
+ "epoch": 44.39297124600639,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0279,
+ "step": 13895
+ },
+ {
+ "epoch": 44.3961661341853,
+ "grad_norm": 0.09326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0293,
+ "step": 13896
+ },
+ {
+ "epoch": 44.39936102236422,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0288,
+ "step": 13897
+ },
+ {
+ "epoch": 44.40255591054313,
+ "grad_norm": 0.1591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0287,
+ "step": 13898
+ },
+ {
+ "epoch": 44.405750798722046,
+ "grad_norm": 0.09375,
+ "learning_rate": 0.0005,
+ "loss": 1.0301,
+ "step": 13899
+ },
+ {
+ "epoch": 44.40894568690096,
+ "grad_norm": 0.2119140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0247,
+ "step": 13900
+ },
+ {
+ "epoch": 44.412140575079874,
+ "grad_norm": 0.091796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0191,
+ "step": 13901
+ },
+ {
+ "epoch": 44.415335463258785,
+ "grad_norm": 0.1796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0404,
+ "step": 13902
+ },
+ {
+ "epoch": 44.4185303514377,
+ "grad_norm": 0.123046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0328,
+ "step": 13903
+ },
+ {
+ "epoch": 44.42172523961661,
+ "grad_norm": 0.08251953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0271,
+ "step": 13904
+ },
+ {
+ "epoch": 44.424920127795524,
+ "grad_norm": 0.1416015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0277,
+ "step": 13905
+ },
+ {
+ "epoch": 44.42811501597444,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0339,
+ "step": 13906
+ },
+ {
+ "epoch": 44.43130990415335,
+ "grad_norm": 0.154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0223,
+ "step": 13907
+ },
+ {
+ "epoch": 44.43450479233227,
+ "grad_norm": 0.08251953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0157,
+ "step": 13908
+ },
+ {
+ "epoch": 44.43769968051118,
+ "grad_norm": 0.130859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0326,
+ "step": 13909
+ },
+ {
+ "epoch": 44.4408945686901,
+ "grad_norm": 0.091796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0296,
+ "step": 13910
+ },
+ {
+ "epoch": 44.44408945686901,
+ "grad_norm": 0.111328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0275,
+ "step": 13911
+ },
+ {
+ "epoch": 44.447284345047926,
+ "grad_norm": 0.11181640625,
+ "learning_rate": 0.0005,
+ "loss": 1.023,
+ "step": 13912
+ },
+ {
+ "epoch": 44.45047923322684,
+ "grad_norm": 0.0908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0367,
+ "step": 13913
+ },
+ {
+ "epoch": 44.453674121405754,
+ "grad_norm": 0.1201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0381,
+ "step": 13914
+ },
+ {
+ "epoch": 44.456869009584665,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0186,
+ "step": 13915
+ },
+ {
+ "epoch": 44.460063897763575,
+ "grad_norm": 0.111328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0319,
+ "step": 13916
+ },
+ {
+ "epoch": 44.46325878594249,
+ "grad_norm": 0.057861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0199,
+ "step": 13917
+ },
+ {
+ "epoch": 44.466453674121404,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0271,
+ "step": 13918
+ },
+ {
+ "epoch": 44.46964856230032,
+ "grad_norm": 0.09619140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0404,
+ "step": 13919
+ },
+ {
+ "epoch": 44.47284345047923,
+ "grad_norm": 0.1435546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0343,
+ "step": 13920
+ },
+ {
+ "epoch": 44.47603833865815,
+ "grad_norm": 0.09375,
+ "learning_rate": 0.0005,
+ "loss": 1.0253,
+ "step": 13921
+ },
+ {
+ "epoch": 44.47923322683706,
+ "grad_norm": 0.208984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0316,
+ "step": 13922
+ },
+ {
+ "epoch": 44.48242811501598,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0215,
+ "step": 13923
+ },
+ {
+ "epoch": 44.48562300319489,
+ "grad_norm": 0.10693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0266,
+ "step": 13924
+ },
+ {
+ "epoch": 44.4888178913738,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0293,
+ "step": 13925
+ },
+ {
+ "epoch": 44.49201277955272,
+ "grad_norm": 0.09375,
+ "learning_rate": 0.0005,
+ "loss": 1.0329,
+ "step": 13926
+ },
+ {
+ "epoch": 44.49520766773163,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0253,
+ "step": 13927
+ },
+ {
+ "epoch": 44.498402555910545,
+ "grad_norm": 0.08203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0366,
+ "step": 13928
+ },
+ {
+ "epoch": 44.501597444089455,
+ "grad_norm": 0.111328125,
+ "learning_rate": 0.0005,
+ "loss": 1.023,
+ "step": 13929
+ },
+ {
+ "epoch": 44.50479233226837,
+ "grad_norm": 0.08251953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0285,
+ "step": 13930
+ },
+ {
+ "epoch": 44.50798722044728,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0352,
+ "step": 13931
+ },
+ {
+ "epoch": 44.5111821086262,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0316,
+ "step": 13932
+ },
+ {
+ "epoch": 44.51437699680511,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0265,
+ "step": 13933
+ },
+ {
+ "epoch": 44.51757188498402,
+ "grad_norm": 0.058837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0266,
+ "step": 13934
+ },
+ {
+ "epoch": 44.52076677316294,
+ "grad_norm": 0.08544921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0274,
+ "step": 13935
+ },
+ {
+ "epoch": 44.52396166134185,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0295,
+ "step": 13936
+ },
+ {
+ "epoch": 44.52715654952077,
+ "grad_norm": 0.1171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0213,
+ "step": 13937
+ },
+ {
+ "epoch": 44.53035143769968,
+ "grad_norm": 0.055908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0355,
+ "step": 13938
+ },
+ {
+ "epoch": 44.533546325878596,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0265,
+ "step": 13939
+ },
+ {
+ "epoch": 44.53674121405751,
+ "grad_norm": 0.08349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0329,
+ "step": 13940
+ },
+ {
+ "epoch": 44.539936102236425,
+ "grad_norm": 0.09765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0318,
+ "step": 13941
+ },
+ {
+ "epoch": 44.543130990415335,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0361,
+ "step": 13942
+ },
+ {
+ "epoch": 44.546325878594246,
+ "grad_norm": 0.10986328125,
+ "learning_rate": 0.0005,
+ "loss": 1.028,
+ "step": 13943
+ },
+ {
+ "epoch": 44.54952076677316,
+ "grad_norm": 0.056396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0301,
+ "step": 13944
+ },
+ {
+ "epoch": 44.552715654952074,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0345,
+ "step": 13945
+ },
+ {
+ "epoch": 44.55591054313099,
+ "grad_norm": 0.1064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0248,
+ "step": 13946
+ },
+ {
+ "epoch": 44.5591054313099,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0284,
+ "step": 13947
+ },
+ {
+ "epoch": 44.56230031948882,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0315,
+ "step": 13948
+ },
+ {
+ "epoch": 44.56549520766773,
+ "grad_norm": 0.06103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0168,
+ "step": 13949
+ },
+ {
+ "epoch": 44.56869009584665,
+ "grad_norm": 0.050048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0311,
+ "step": 13950
+ },
+ {
+ "epoch": 44.57188498402556,
+ "grad_norm": 0.060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0367,
+ "step": 13951
+ },
+ {
+ "epoch": 44.575079872204476,
+ "grad_norm": 0.09326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0237,
+ "step": 13952
+ },
+ {
+ "epoch": 44.57827476038339,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0292,
+ "step": 13953
+ },
+ {
+ "epoch": 44.5814696485623,
+ "grad_norm": 0.1435546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0379,
+ "step": 13954
+ },
+ {
+ "epoch": 44.584664536741215,
+ "grad_norm": 0.08056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0315,
+ "step": 13955
+ },
+ {
+ "epoch": 44.587859424920126,
+ "grad_norm": 0.1669921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0394,
+ "step": 13956
+ },
+ {
+ "epoch": 44.59105431309904,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0297,
+ "step": 13957
+ },
+ {
+ "epoch": 44.594249201277954,
+ "grad_norm": 0.057373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.022,
+ "step": 13958
+ },
+ {
+ "epoch": 44.59744408945687,
+ "grad_norm": 0.173828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0258,
+ "step": 13959
+ },
+ {
+ "epoch": 44.60063897763578,
+ "grad_norm": 0.09716796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0266,
+ "step": 13960
+ },
+ {
+ "epoch": 44.6038338658147,
+ "grad_norm": 0.0546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0401,
+ "step": 13961
+ },
+ {
+ "epoch": 44.60702875399361,
+ "grad_norm": 0.099609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0317,
+ "step": 13962
+ },
+ {
+ "epoch": 44.61022364217252,
+ "grad_norm": 0.1044921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0326,
+ "step": 13963
+ },
+ {
+ "epoch": 44.61341853035144,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0377,
+ "step": 13964
+ },
+ {
+ "epoch": 44.61661341853035,
+ "grad_norm": 0.10888671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0313,
+ "step": 13965
+ },
+ {
+ "epoch": 44.61980830670927,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0154,
+ "step": 13966
+ },
+ {
+ "epoch": 44.62300319488818,
+ "grad_norm": 0.059326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0278,
+ "step": 13967
+ },
+ {
+ "epoch": 44.626198083067095,
+ "grad_norm": 0.1357421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0195,
+ "step": 13968
+ },
+ {
+ "epoch": 44.629392971246006,
+ "grad_norm": 0.0537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0268,
+ "step": 13969
+ },
+ {
+ "epoch": 44.63258785942492,
+ "grad_norm": 0.10498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0318,
+ "step": 13970
+ },
+ {
+ "epoch": 44.635782747603834,
+ "grad_norm": 0.049560546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0277,
+ "step": 13971
+ },
+ {
+ "epoch": 44.638977635782744,
+ "grad_norm": 0.056884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0166,
+ "step": 13972
+ },
+ {
+ "epoch": 44.64217252396166,
+ "grad_norm": 0.05029296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0329,
+ "step": 13973
+ },
+ {
+ "epoch": 44.64536741214057,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.028,
+ "step": 13974
+ },
+ {
+ "epoch": 44.64856230031949,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.029,
+ "step": 13975
+ },
+ {
+ "epoch": 44.6517571884984,
+ "grad_norm": 0.0810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0276,
+ "step": 13976
+ },
+ {
+ "epoch": 44.65495207667732,
+ "grad_norm": 0.056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0311,
+ "step": 13977
+ },
+ {
+ "epoch": 44.65814696485623,
+ "grad_norm": 0.062255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0191,
+ "step": 13978
+ },
+ {
+ "epoch": 44.66134185303515,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0263,
+ "step": 13979
+ },
+ {
+ "epoch": 44.66453674121406,
+ "grad_norm": 0.05322265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0269,
+ "step": 13980
+ },
+ {
+ "epoch": 44.66773162939297,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0174,
+ "step": 13981
+ },
+ {
+ "epoch": 44.670926517571885,
+ "grad_norm": 0.08154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0368,
+ "step": 13982
+ },
+ {
+ "epoch": 44.674121405750796,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0247,
+ "step": 13983
+ },
+ {
+ "epoch": 44.677316293929714,
+ "grad_norm": 0.08251953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0218,
+ "step": 13984
+ },
+ {
+ "epoch": 44.680511182108624,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0345,
+ "step": 13985
+ },
+ {
+ "epoch": 44.68370607028754,
+ "grad_norm": 0.10400390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0308,
+ "step": 13986
+ },
+ {
+ "epoch": 44.68690095846645,
+ "grad_norm": 0.04931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0329,
+ "step": 13987
+ },
+ {
+ "epoch": 44.69009584664537,
+ "grad_norm": 0.083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0245,
+ "step": 13988
+ },
+ {
+ "epoch": 44.69329073482428,
+ "grad_norm": 0.0478515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0257,
+ "step": 13989
+ },
+ {
+ "epoch": 44.6964856230032,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0267,
+ "step": 13990
+ },
+ {
+ "epoch": 44.69968051118211,
+ "grad_norm": 0.052001953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0326,
+ "step": 13991
+ },
+ {
+ "epoch": 44.70287539936102,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.025,
+ "step": 13992
+ },
+ {
+ "epoch": 44.70607028753994,
+ "grad_norm": 0.0576171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0266,
+ "step": 13993
+ },
+ {
+ "epoch": 44.70926517571885,
+ "grad_norm": 0.059814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0392,
+ "step": 13994
+ },
+ {
+ "epoch": 44.712460063897765,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.027,
+ "step": 13995
+ },
+ {
+ "epoch": 44.715654952076676,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0376,
+ "step": 13996
+ },
+ {
+ "epoch": 44.718849840255594,
+ "grad_norm": 0.050537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0212,
+ "step": 13997
+ },
+ {
+ "epoch": 44.722044728434504,
+ "grad_norm": 0.048583984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0297,
+ "step": 13998
+ },
+ {
+ "epoch": 44.72523961661342,
+ "grad_norm": 0.04541015625,
+ "learning_rate": 0.0005,
+ "loss": 1.038,
+ "step": 13999
+ },
+ {
+ "epoch": 44.72843450479233,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0299,
+ "step": 14000
+ },
+ {
+ "epoch": 44.73162939297124,
+ "grad_norm": 0.052978515625,
+ "learning_rate": 0.0005,
+ "loss": 1.025,
+ "step": 14001
+ },
+ {
+ "epoch": 44.73482428115016,
+ "grad_norm": 0.060791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0344,
+ "step": 14002
+ },
+ {
+ "epoch": 44.73801916932907,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0306,
+ "step": 14003
+ },
+ {
+ "epoch": 44.74121405750799,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.032,
+ "step": 14004
+ },
+ {
+ "epoch": 44.7444089456869,
+ "grad_norm": 0.1474609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0346,
+ "step": 14005
+ },
+ {
+ "epoch": 44.74760383386582,
+ "grad_norm": 0.1064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.026,
+ "step": 14006
+ },
+ {
+ "epoch": 44.75079872204473,
+ "grad_norm": 0.158203125,
+ "learning_rate": 0.0005,
+ "loss": 1.03,
+ "step": 14007
+ },
+ {
+ "epoch": 44.753993610223645,
+ "grad_norm": 0.1083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0294,
+ "step": 14008
+ },
+ {
+ "epoch": 44.757188498402556,
+ "grad_norm": 0.169921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0284,
+ "step": 14009
+ },
+ {
+ "epoch": 44.760383386581466,
+ "grad_norm": 0.08984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0282,
+ "step": 14010
+ },
+ {
+ "epoch": 44.763578274760384,
+ "grad_norm": 0.08056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0327,
+ "step": 14011
+ },
+ {
+ "epoch": 44.766773162939295,
+ "grad_norm": 0.0810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.031,
+ "step": 14012
+ },
+ {
+ "epoch": 44.76996805111821,
+ "grad_norm": 0.09814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0317,
+ "step": 14013
+ },
+ {
+ "epoch": 44.77316293929712,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0248,
+ "step": 14014
+ },
+ {
+ "epoch": 44.77635782747604,
+ "grad_norm": 0.1005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0204,
+ "step": 14015
+ },
+ {
+ "epoch": 44.77955271565495,
+ "grad_norm": 0.16015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0251,
+ "step": 14016
+ },
+ {
+ "epoch": 44.78274760383387,
+ "grad_norm": 0.11181640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0202,
+ "step": 14017
+ },
+ {
+ "epoch": 44.78594249201278,
+ "grad_norm": 0.08740234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0286,
+ "step": 14018
+ },
+ {
+ "epoch": 44.78913738019169,
+ "grad_norm": 0.08837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0256,
+ "step": 14019
+ },
+ {
+ "epoch": 44.79233226837061,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.026,
+ "step": 14020
+ },
+ {
+ "epoch": 44.79552715654952,
+ "grad_norm": 0.134765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0302,
+ "step": 14021
+ },
+ {
+ "epoch": 44.798722044728436,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0404,
+ "step": 14022
+ },
+ {
+ "epoch": 44.801916932907346,
+ "grad_norm": 0.09912109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0287,
+ "step": 14023
+ },
+ {
+ "epoch": 44.805111821086264,
+ "grad_norm": 0.103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0403,
+ "step": 14024
+ },
+ {
+ "epoch": 44.808306709265175,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0311,
+ "step": 14025
+ },
+ {
+ "epoch": 44.81150159744409,
+ "grad_norm": 0.09619140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0279,
+ "step": 14026
+ },
+ {
+ "epoch": 44.814696485623,
+ "grad_norm": 0.1220703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0216,
+ "step": 14027
+ },
+ {
+ "epoch": 44.81789137380191,
+ "grad_norm": 0.130859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0448,
+ "step": 14028
+ },
+ {
+ "epoch": 44.82108626198083,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0272,
+ "step": 14029
+ },
+ {
+ "epoch": 44.82428115015974,
+ "grad_norm": 0.09375,
+ "learning_rate": 0.0005,
+ "loss": 1.0331,
+ "step": 14030
+ },
+ {
+ "epoch": 44.82747603833866,
+ "grad_norm": 0.05908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.025,
+ "step": 14031
+ },
+ {
+ "epoch": 44.83067092651757,
+ "grad_norm": 0.09228515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0276,
+ "step": 14032
+ },
+ {
+ "epoch": 44.83386581469649,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0291,
+ "step": 14033
+ },
+ {
+ "epoch": 44.8370607028754,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0352,
+ "step": 14034
+ },
+ {
+ "epoch": 44.840255591054316,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0374,
+ "step": 14035
+ },
+ {
+ "epoch": 44.843450479233226,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0305,
+ "step": 14036
+ },
+ {
+ "epoch": 44.846645367412144,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0259,
+ "step": 14037
+ },
+ {
+ "epoch": 44.849840255591054,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0352,
+ "step": 14038
+ },
+ {
+ "epoch": 44.853035143769965,
+ "grad_norm": 0.05322265625,
+ "learning_rate": 0.0005,
+ "loss": 1.021,
+ "step": 14039
+ },
+ {
+ "epoch": 44.85623003194888,
+ "grad_norm": 0.091796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0278,
+ "step": 14040
+ },
+ {
+ "epoch": 44.85942492012779,
+ "grad_norm": 0.05859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0212,
+ "step": 14041
+ },
+ {
+ "epoch": 44.86261980830671,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0331,
+ "step": 14042
+ },
+ {
+ "epoch": 44.86581469648562,
+ "grad_norm": 0.099609375,
+ "learning_rate": 0.0005,
+ "loss": 1.03,
+ "step": 14043
+ },
+ {
+ "epoch": 44.86900958466454,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0211,
+ "step": 14044
+ },
+ {
+ "epoch": 44.87220447284345,
+ "grad_norm": 0.1318359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0394,
+ "step": 14045
+ },
+ {
+ "epoch": 44.87539936102237,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0346,
+ "step": 14046
+ },
+ {
+ "epoch": 44.87859424920128,
+ "grad_norm": 0.1259765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0357,
+ "step": 14047
+ },
+ {
+ "epoch": 44.88178913738019,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0181,
+ "step": 14048
+ },
+ {
+ "epoch": 44.884984025559106,
+ "grad_norm": 0.109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0254,
+ "step": 14049
+ },
+ {
+ "epoch": 44.88817891373802,
+ "grad_norm": 0.06005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.023,
+ "step": 14050
+ },
+ {
+ "epoch": 44.891373801916934,
+ "grad_norm": 0.0927734375,
+ "learning_rate": 0.0005,
+ "loss": 1.035,
+ "step": 14051
+ },
+ {
+ "epoch": 44.894568690095845,
+ "grad_norm": 0.050048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.049,
+ "step": 14052
+ },
+ {
+ "epoch": 44.89776357827476,
+ "grad_norm": 0.08349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0223,
+ "step": 14053
+ },
+ {
+ "epoch": 44.90095846645367,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0307,
+ "step": 14054
+ },
+ {
+ "epoch": 44.90415335463259,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0267,
+ "step": 14055
+ },
+ {
+ "epoch": 44.9073482428115,
+ "grad_norm": 0.04931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0204,
+ "step": 14056
+ },
+ {
+ "epoch": 44.91054313099041,
+ "grad_norm": 0.05615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0203,
+ "step": 14057
+ },
+ {
+ "epoch": 44.91373801916933,
+ "grad_norm": 0.052978515625,
+ "learning_rate": 0.0005,
+ "loss": 1.034,
+ "step": 14058
+ },
+ {
+ "epoch": 44.91693290734824,
+ "grad_norm": 0.055419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0334,
+ "step": 14059
+ },
+ {
+ "epoch": 44.92012779552716,
+ "grad_norm": 0.060302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0324,
+ "step": 14060
+ },
+ {
+ "epoch": 44.92332268370607,
+ "grad_norm": 0.050537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0353,
+ "step": 14061
+ },
+ {
+ "epoch": 44.926517571884986,
+ "grad_norm": 0.08056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0358,
+ "step": 14062
+ },
+ {
+ "epoch": 44.9297124600639,
+ "grad_norm": 0.048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0321,
+ "step": 14063
+ },
+ {
+ "epoch": 44.932907348242814,
+ "grad_norm": 0.0595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0209,
+ "step": 14064
+ },
+ {
+ "epoch": 44.936102236421725,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.039,
+ "step": 14065
+ },
+ {
+ "epoch": 44.93929712460064,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0298,
+ "step": 14066
+ },
+ {
+ "epoch": 44.94249201277955,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0362,
+ "step": 14067
+ },
+ {
+ "epoch": 44.945686900958464,
+ "grad_norm": 0.05322265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0237,
+ "step": 14068
+ },
+ {
+ "epoch": 44.94888178913738,
+ "grad_norm": 0.091796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0376,
+ "step": 14069
+ },
+ {
+ "epoch": 44.95207667731629,
+ "grad_norm": 0.1064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0317,
+ "step": 14070
+ },
+ {
+ "epoch": 44.95527156549521,
+ "grad_norm": 0.0927734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0396,
+ "step": 14071
+ },
+ {
+ "epoch": 44.95846645367412,
+ "grad_norm": 0.1748046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0332,
+ "step": 14072
+ },
+ {
+ "epoch": 44.96166134185304,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0347,
+ "step": 14073
+ },
+ {
+ "epoch": 44.96485623003195,
+ "grad_norm": 0.158203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0267,
+ "step": 14074
+ },
+ {
+ "epoch": 44.968051118210866,
+ "grad_norm": 0.09716796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0306,
+ "step": 14075
+ },
+ {
+ "epoch": 44.97124600638978,
+ "grad_norm": 0.232421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0208,
+ "step": 14076
+ },
+ {
+ "epoch": 44.97444089456869,
+ "grad_norm": 0.0908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0321,
+ "step": 14077
+ },
+ {
+ "epoch": 44.977635782747605,
+ "grad_norm": 0.1337890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0179,
+ "step": 14078
+ },
+ {
+ "epoch": 44.980830670926515,
+ "grad_norm": 0.0986328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0246,
+ "step": 14079
+ },
+ {
+ "epoch": 44.98402555910543,
+ "grad_norm": 0.0810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0297,
+ "step": 14080
+ },
+ {
+ "epoch": 44.98722044728434,
+ "grad_norm": 0.08935546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0224,
+ "step": 14081
+ },
+ {
+ "epoch": 44.99041533546326,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0366,
+ "step": 14082
+ },
+ {
+ "epoch": 44.99361022364217,
+ "grad_norm": 0.07861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0269,
+ "step": 14083
+ },
+ {
+ "epoch": 44.99680511182109,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0383,
+ "step": 14084
+ },
+ {
+ "epoch": 45.0,
+ "grad_norm": 0.1220703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0321,
+ "step": 14085
+ },
+ {
+ "epoch": 45.00319488817891,
+ "grad_norm": 0.04638671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0274,
+ "step": 14086
+ },
+ {
+ "epoch": 45.00638977635783,
+ "grad_norm": 0.138671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0322,
+ "step": 14087
+ },
+ {
+ "epoch": 45.00958466453674,
+ "grad_norm": 0.045654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.037,
+ "step": 14088
+ },
+ {
+ "epoch": 45.01277955271566,
+ "grad_norm": 0.0947265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0201,
+ "step": 14089
+ },
+ {
+ "epoch": 45.01597444089457,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0358,
+ "step": 14090
+ },
+ {
+ "epoch": 45.019169329073485,
+ "grad_norm": 0.10546875,
+ "learning_rate": 0.0005,
+ "loss": 1.042,
+ "step": 14091
+ },
+ {
+ "epoch": 45.022364217252395,
+ "grad_norm": 0.059326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0457,
+ "step": 14092
+ },
+ {
+ "epoch": 45.02555910543131,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0305,
+ "step": 14093
+ },
+ {
+ "epoch": 45.02875399361022,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0264,
+ "step": 14094
+ },
+ {
+ "epoch": 45.031948881789134,
+ "grad_norm": 0.05224609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0311,
+ "step": 14095
+ },
+ {
+ "epoch": 45.03514376996805,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0284,
+ "step": 14096
+ },
+ {
+ "epoch": 45.03833865814696,
+ "grad_norm": 0.05615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0236,
+ "step": 14097
+ },
+ {
+ "epoch": 45.04153354632588,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0291,
+ "step": 14098
+ },
+ {
+ "epoch": 45.04472843450479,
+ "grad_norm": 0.05810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0276,
+ "step": 14099
+ },
+ {
+ "epoch": 45.04792332268371,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0223,
+ "step": 14100
+ },
+ {
+ "epoch": 45.05111821086262,
+ "grad_norm": 0.056396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0323,
+ "step": 14101
+ },
+ {
+ "epoch": 45.054313099041536,
+ "grad_norm": 0.0498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.017,
+ "step": 14102
+ },
+ {
+ "epoch": 45.05750798722045,
+ "grad_norm": 0.0498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0198,
+ "step": 14103
+ },
+ {
+ "epoch": 45.06070287539936,
+ "grad_norm": 0.055419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0263,
+ "step": 14104
+ },
+ {
+ "epoch": 45.063897763578275,
+ "grad_norm": 0.045166015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0345,
+ "step": 14105
+ },
+ {
+ "epoch": 45.067092651757186,
+ "grad_norm": 0.083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0188,
+ "step": 14106
+ },
+ {
+ "epoch": 45.0702875399361,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0355,
+ "step": 14107
+ },
+ {
+ "epoch": 45.073482428115014,
+ "grad_norm": 0.1513671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0385,
+ "step": 14108
+ },
+ {
+ "epoch": 45.07667731629393,
+ "grad_norm": 0.05859375,
+ "learning_rate": 0.0005,
+ "loss": 1.024,
+ "step": 14109
+ },
+ {
+ "epoch": 45.07987220447284,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0274,
+ "step": 14110
+ },
+ {
+ "epoch": 45.08306709265176,
+ "grad_norm": 0.05517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0247,
+ "step": 14111
+ },
+ {
+ "epoch": 45.08626198083067,
+ "grad_norm": 0.060302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0216,
+ "step": 14112
+ },
+ {
+ "epoch": 45.08945686900959,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0271,
+ "step": 14113
+ },
+ {
+ "epoch": 45.0926517571885,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0309,
+ "step": 14114
+ },
+ {
+ "epoch": 45.09584664536741,
+ "grad_norm": 0.0859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0278,
+ "step": 14115
+ },
+ {
+ "epoch": 45.09904153354633,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0371,
+ "step": 14116
+ },
+ {
+ "epoch": 45.10223642172524,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0367,
+ "step": 14117
+ },
+ {
+ "epoch": 45.105431309904155,
+ "grad_norm": 0.08251953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0305,
+ "step": 14118
+ },
+ {
+ "epoch": 45.108626198083066,
+ "grad_norm": 0.09423828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0371,
+ "step": 14119
+ },
+ {
+ "epoch": 45.11182108626198,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0196,
+ "step": 14120
+ },
+ {
+ "epoch": 45.115015974440894,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0171,
+ "step": 14121
+ },
+ {
+ "epoch": 45.11821086261981,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.03,
+ "step": 14122
+ },
+ {
+ "epoch": 45.12140575079872,
+ "grad_norm": 0.0810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0267,
+ "step": 14123
+ },
+ {
+ "epoch": 45.12460063897763,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0311,
+ "step": 14124
+ },
+ {
+ "epoch": 45.12779552715655,
+ "grad_norm": 0.083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0297,
+ "step": 14125
+ },
+ {
+ "epoch": 45.13099041533546,
+ "grad_norm": 0.052978515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0325,
+ "step": 14126
+ },
+ {
+ "epoch": 45.13418530351438,
+ "grad_norm": 0.08740234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0284,
+ "step": 14127
+ },
+ {
+ "epoch": 45.13738019169329,
+ "grad_norm": 0.046630859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0235,
+ "step": 14128
+ },
+ {
+ "epoch": 45.14057507987221,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0214,
+ "step": 14129
+ },
+ {
+ "epoch": 45.14376996805112,
+ "grad_norm": 0.10595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0292,
+ "step": 14130
+ },
+ {
+ "epoch": 45.146964856230035,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0381,
+ "step": 14131
+ },
+ {
+ "epoch": 45.150159744408946,
+ "grad_norm": 0.1044921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0331,
+ "step": 14132
+ },
+ {
+ "epoch": 45.153354632587856,
+ "grad_norm": 0.05322265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0254,
+ "step": 14133
+ },
+ {
+ "epoch": 45.156549520766774,
+ "grad_norm": 0.051513671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0278,
+ "step": 14134
+ },
+ {
+ "epoch": 45.159744408945684,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0322,
+ "step": 14135
+ },
+ {
+ "epoch": 45.1629392971246,
+ "grad_norm": 0.053466796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0257,
+ "step": 14136
+ },
+ {
+ "epoch": 45.16613418530351,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0226,
+ "step": 14137
+ },
+ {
+ "epoch": 45.16932907348243,
+ "grad_norm": 0.056884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0357,
+ "step": 14138
+ },
+ {
+ "epoch": 45.17252396166134,
+ "grad_norm": 0.0537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0347,
+ "step": 14139
+ },
+ {
+ "epoch": 45.17571884984026,
+ "grad_norm": 0.05712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0272,
+ "step": 14140
+ },
+ {
+ "epoch": 45.17891373801917,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0307,
+ "step": 14141
+ },
+ {
+ "epoch": 45.18210862619808,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0309,
+ "step": 14142
+ },
+ {
+ "epoch": 45.185303514377,
+ "grad_norm": 0.052734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0251,
+ "step": 14143
+ },
+ {
+ "epoch": 45.18849840255591,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0268,
+ "step": 14144
+ },
+ {
+ "epoch": 45.191693290734825,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0281,
+ "step": 14145
+ },
+ {
+ "epoch": 45.194888178913736,
+ "grad_norm": 0.047607421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0258,
+ "step": 14146
+ },
+ {
+ "epoch": 45.198083067092654,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0277,
+ "step": 14147
+ },
+ {
+ "epoch": 45.201277955271564,
+ "grad_norm": 0.0869140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0284,
+ "step": 14148
+ },
+ {
+ "epoch": 45.20447284345048,
+ "grad_norm": 0.059814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0192,
+ "step": 14149
+ },
+ {
+ "epoch": 45.20766773162939,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0354,
+ "step": 14150
+ },
+ {
+ "epoch": 45.21086261980831,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0392,
+ "step": 14151
+ },
+ {
+ "epoch": 45.21405750798722,
+ "grad_norm": 0.0576171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0272,
+ "step": 14152
+ },
+ {
+ "epoch": 45.21725239616613,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.024,
+ "step": 14153
+ },
+ {
+ "epoch": 45.22044728434505,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0327,
+ "step": 14154
+ },
+ {
+ "epoch": 45.22364217252396,
+ "grad_norm": 0.053955078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0278,
+ "step": 14155
+ },
+ {
+ "epoch": 45.22683706070288,
+ "grad_norm": 0.08154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0273,
+ "step": 14156
+ },
+ {
+ "epoch": 45.23003194888179,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0351,
+ "step": 14157
+ },
+ {
+ "epoch": 45.233226837060705,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0241,
+ "step": 14158
+ },
+ {
+ "epoch": 45.236421725239616,
+ "grad_norm": 0.055419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0312,
+ "step": 14159
+ },
+ {
+ "epoch": 45.239616613418534,
+ "grad_norm": 0.052978515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0278,
+ "step": 14160
+ },
+ {
+ "epoch": 45.242811501597444,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0318,
+ "step": 14161
+ },
+ {
+ "epoch": 45.246006389776355,
+ "grad_norm": 0.05810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0262,
+ "step": 14162
+ },
+ {
+ "epoch": 45.24920127795527,
+ "grad_norm": 0.050537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0434,
+ "step": 14163
+ },
+ {
+ "epoch": 45.25239616613418,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0218,
+ "step": 14164
+ },
+ {
+ "epoch": 45.2555910543131,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0343,
+ "step": 14165
+ },
+ {
+ "epoch": 45.25878594249201,
+ "grad_norm": 0.1123046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0244,
+ "step": 14166
+ },
+ {
+ "epoch": 45.26198083067093,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.022,
+ "step": 14167
+ },
+ {
+ "epoch": 45.26517571884984,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0284,
+ "step": 14168
+ },
+ {
+ "epoch": 45.26837060702876,
+ "grad_norm": 0.1162109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0175,
+ "step": 14169
+ },
+ {
+ "epoch": 45.27156549520767,
+ "grad_norm": 0.09912109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0289,
+ "step": 14170
+ },
+ {
+ "epoch": 45.27476038338658,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0312,
+ "step": 14171
+ },
+ {
+ "epoch": 45.277955271565496,
+ "grad_norm": 0.099609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0286,
+ "step": 14172
+ },
+ {
+ "epoch": 45.281150159744406,
+ "grad_norm": 0.055419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0353,
+ "step": 14173
+ },
+ {
+ "epoch": 45.284345047923324,
+ "grad_norm": 0.09375,
+ "learning_rate": 0.0005,
+ "loss": 1.0163,
+ "step": 14174
+ },
+ {
+ "epoch": 45.287539936102235,
+ "grad_norm": 0.056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0337,
+ "step": 14175
+ },
+ {
+ "epoch": 45.29073482428115,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0102,
+ "step": 14176
+ },
+ {
+ "epoch": 45.29392971246006,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0168,
+ "step": 14177
+ },
+ {
+ "epoch": 45.29712460063898,
+ "grad_norm": 0.050537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0312,
+ "step": 14178
+ },
+ {
+ "epoch": 45.30031948881789,
+ "grad_norm": 0.0595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0321,
+ "step": 14179
+ },
+ {
+ "epoch": 45.3035143769968,
+ "grad_norm": 0.056884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0332,
+ "step": 14180
+ },
+ {
+ "epoch": 45.30670926517572,
+ "grad_norm": 0.0546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0314,
+ "step": 14181
+ },
+ {
+ "epoch": 45.30990415335463,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0446,
+ "step": 14182
+ },
+ {
+ "epoch": 45.31309904153355,
+ "grad_norm": 0.0556640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0258,
+ "step": 14183
+ },
+ {
+ "epoch": 45.31629392971246,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.026,
+ "step": 14184
+ },
+ {
+ "epoch": 45.319488817891376,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.043,
+ "step": 14185
+ },
+ {
+ "epoch": 45.322683706070286,
+ "grad_norm": 0.051025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0359,
+ "step": 14186
+ },
+ {
+ "epoch": 45.325878594249204,
+ "grad_norm": 0.1396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0279,
+ "step": 14187
+ },
+ {
+ "epoch": 45.329073482428115,
+ "grad_norm": 0.083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0349,
+ "step": 14188
+ },
+ {
+ "epoch": 45.33226837060703,
+ "grad_norm": 0.10009765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0407,
+ "step": 14189
+ },
+ {
+ "epoch": 45.33546325878594,
+ "grad_norm": 0.09716796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0223,
+ "step": 14190
+ },
+ {
+ "epoch": 45.33865814696485,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0273,
+ "step": 14191
+ },
+ {
+ "epoch": 45.34185303514377,
+ "grad_norm": 0.08203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0264,
+ "step": 14192
+ },
+ {
+ "epoch": 45.34504792332268,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0385,
+ "step": 14193
+ },
+ {
+ "epoch": 45.3482428115016,
+ "grad_norm": 0.05419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0405,
+ "step": 14194
+ },
+ {
+ "epoch": 45.35143769968051,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0211,
+ "step": 14195
+ },
+ {
+ "epoch": 45.35463258785943,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.028,
+ "step": 14196
+ },
+ {
+ "epoch": 45.35782747603834,
+ "grad_norm": 0.05419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0255,
+ "step": 14197
+ },
+ {
+ "epoch": 45.361022364217256,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0381,
+ "step": 14198
+ },
+ {
+ "epoch": 45.364217252396166,
+ "grad_norm": 0.0830078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0246,
+ "step": 14199
+ },
+ {
+ "epoch": 45.36741214057508,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0134,
+ "step": 14200
+ },
+ {
+ "epoch": 45.370607028753994,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0299,
+ "step": 14201
+ },
+ {
+ "epoch": 45.373801916932905,
+ "grad_norm": 0.0537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0325,
+ "step": 14202
+ },
+ {
+ "epoch": 45.37699680511182,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0172,
+ "step": 14203
+ },
+ {
+ "epoch": 45.38019169329073,
+ "grad_norm": 0.05517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0338,
+ "step": 14204
+ },
+ {
+ "epoch": 45.38338658146965,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0384,
+ "step": 14205
+ },
+ {
+ "epoch": 45.38658146964856,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0308,
+ "step": 14206
+ },
+ {
+ "epoch": 45.38977635782748,
+ "grad_norm": 0.054443359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0245,
+ "step": 14207
+ },
+ {
+ "epoch": 45.39297124600639,
+ "grad_norm": 0.0859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0291,
+ "step": 14208
+ },
+ {
+ "epoch": 45.3961661341853,
+ "grad_norm": 0.05078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0205,
+ "step": 14209
+ },
+ {
+ "epoch": 45.39936102236422,
+ "grad_norm": 0.05615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0198,
+ "step": 14210
+ },
+ {
+ "epoch": 45.40255591054313,
+ "grad_norm": 0.0517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0332,
+ "step": 14211
+ },
+ {
+ "epoch": 45.405750798722046,
+ "grad_norm": 0.051025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.028,
+ "step": 14212
+ },
+ {
+ "epoch": 45.40894568690096,
+ "grad_norm": 0.04736328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0341,
+ "step": 14213
+ },
+ {
+ "epoch": 45.412140575079874,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0285,
+ "step": 14214
+ },
+ {
+ "epoch": 45.415335463258785,
+ "grad_norm": 0.04833984375,
+ "learning_rate": 0.0005,
+ "loss": 1.03,
+ "step": 14215
+ },
+ {
+ "epoch": 45.4185303514377,
+ "grad_norm": 0.050537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0108,
+ "step": 14216
+ },
+ {
+ "epoch": 45.42172523961661,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0354,
+ "step": 14217
+ },
+ {
+ "epoch": 45.424920127795524,
+ "grad_norm": 0.049560546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0366,
+ "step": 14218
+ },
+ {
+ "epoch": 45.42811501597444,
+ "grad_norm": 0.05126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0274,
+ "step": 14219
+ },
+ {
+ "epoch": 45.43130990415335,
+ "grad_norm": 0.052490234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0071,
+ "step": 14220
+ },
+ {
+ "epoch": 45.43450479233227,
+ "grad_norm": 0.051513671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0232,
+ "step": 14221
+ },
+ {
+ "epoch": 45.43769968051118,
+ "grad_norm": 0.048583984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0324,
+ "step": 14222
+ },
+ {
+ "epoch": 45.4408945686901,
+ "grad_norm": 0.053955078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0372,
+ "step": 14223
+ },
+ {
+ "epoch": 45.44408945686901,
+ "grad_norm": 0.06201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.028,
+ "step": 14224
+ },
+ {
+ "epoch": 45.447284345047926,
+ "grad_norm": 0.0556640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0286,
+ "step": 14225
+ },
+ {
+ "epoch": 45.45047923322684,
+ "grad_norm": 0.050048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0211,
+ "step": 14226
+ },
+ {
+ "epoch": 45.453674121405754,
+ "grad_norm": 0.051025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0441,
+ "step": 14227
+ },
+ {
+ "epoch": 45.456869009584665,
+ "grad_norm": 0.0517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0283,
+ "step": 14228
+ },
+ {
+ "epoch": 45.460063897763575,
+ "grad_norm": 0.09033203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0311,
+ "step": 14229
+ },
+ {
+ "epoch": 45.46325878594249,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0253,
+ "step": 14230
+ },
+ {
+ "epoch": 45.466453674121404,
+ "grad_norm": 0.0859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0276,
+ "step": 14231
+ },
+ {
+ "epoch": 45.46964856230032,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0252,
+ "step": 14232
+ },
+ {
+ "epoch": 45.47284345047923,
+ "grad_norm": 0.0595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0282,
+ "step": 14233
+ },
+ {
+ "epoch": 45.47603833865815,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0283,
+ "step": 14234
+ },
+ {
+ "epoch": 45.47923322683706,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0393,
+ "step": 14235
+ },
+ {
+ "epoch": 45.48242811501598,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0354,
+ "step": 14236
+ },
+ {
+ "epoch": 45.48562300319489,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0298,
+ "step": 14237
+ },
+ {
+ "epoch": 45.4888178913738,
+ "grad_norm": 0.04931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0298,
+ "step": 14238
+ },
+ {
+ "epoch": 45.49201277955272,
+ "grad_norm": 0.0869140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0359,
+ "step": 14239
+ },
+ {
+ "epoch": 45.49520766773163,
+ "grad_norm": 0.08642578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0372,
+ "step": 14240
+ },
+ {
+ "epoch": 45.498402555910545,
+ "grad_norm": 0.09228515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0194,
+ "step": 14241
+ },
+ {
+ "epoch": 45.501597444089455,
+ "grad_norm": 0.054443359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0311,
+ "step": 14242
+ },
+ {
+ "epoch": 45.50479233226837,
+ "grad_norm": 0.0927734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0299,
+ "step": 14243
+ },
+ {
+ "epoch": 45.50798722044728,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0273,
+ "step": 14244
+ },
+ {
+ "epoch": 45.5111821086262,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0439,
+ "step": 14245
+ },
+ {
+ "epoch": 45.51437699680511,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0187,
+ "step": 14246
+ },
+ {
+ "epoch": 45.51757188498402,
+ "grad_norm": 0.08251953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0361,
+ "step": 14247
+ },
+ {
+ "epoch": 45.52076677316294,
+ "grad_norm": 0.0888671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0271,
+ "step": 14248
+ },
+ {
+ "epoch": 45.52396166134185,
+ "grad_norm": 0.1025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.042,
+ "step": 14249
+ },
+ {
+ "epoch": 45.52715654952077,
+ "grad_norm": 0.1376953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0237,
+ "step": 14250
+ },
+ {
+ "epoch": 45.53035143769968,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0342,
+ "step": 14251
+ },
+ {
+ "epoch": 45.533546325878596,
+ "grad_norm": 0.08349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0266,
+ "step": 14252
+ },
+ {
+ "epoch": 45.53674121405751,
+ "grad_norm": 0.052001953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0367,
+ "step": 14253
+ },
+ {
+ "epoch": 45.539936102236425,
+ "grad_norm": 0.09375,
+ "learning_rate": 0.0005,
+ "loss": 1.0296,
+ "step": 14254
+ },
+ {
+ "epoch": 45.543130990415335,
+ "grad_norm": 0.052978515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0388,
+ "step": 14255
+ },
+ {
+ "epoch": 45.546325878594246,
+ "grad_norm": 0.0830078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0348,
+ "step": 14256
+ },
+ {
+ "epoch": 45.54952076677316,
+ "grad_norm": 0.046142578125,
+ "learning_rate": 0.0005,
+ "loss": 1.038,
+ "step": 14257
+ },
+ {
+ "epoch": 45.552715654952074,
+ "grad_norm": 0.103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0248,
+ "step": 14258
+ },
+ {
+ "epoch": 45.55591054313099,
+ "grad_norm": 0.091796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0193,
+ "step": 14259
+ },
+ {
+ "epoch": 45.5591054313099,
+ "grad_norm": 0.1298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0365,
+ "step": 14260
+ },
+ {
+ "epoch": 45.56230031948882,
+ "grad_norm": 0.1201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0314,
+ "step": 14261
+ },
+ {
+ "epoch": 45.56549520766773,
+ "grad_norm": 0.11376953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0319,
+ "step": 14262
+ },
+ {
+ "epoch": 45.56869009584665,
+ "grad_norm": 0.142578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0306,
+ "step": 14263
+ },
+ {
+ "epoch": 45.57188498402556,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0368,
+ "step": 14264
+ },
+ {
+ "epoch": 45.575079872204476,
+ "grad_norm": 0.0966796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0289,
+ "step": 14265
+ },
+ {
+ "epoch": 45.57827476038339,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0325,
+ "step": 14266
+ },
+ {
+ "epoch": 45.5814696485623,
+ "grad_norm": 0.130859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0353,
+ "step": 14267
+ },
+ {
+ "epoch": 45.584664536741215,
+ "grad_norm": 0.05322265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0391,
+ "step": 14268
+ },
+ {
+ "epoch": 45.587859424920126,
+ "grad_norm": 0.13671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0289,
+ "step": 14269
+ },
+ {
+ "epoch": 45.59105431309904,
+ "grad_norm": 0.055419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0253,
+ "step": 14270
+ },
+ {
+ "epoch": 45.594249201277954,
+ "grad_norm": 0.12060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0322,
+ "step": 14271
+ },
+ {
+ "epoch": 45.59744408945687,
+ "grad_norm": 0.0810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0344,
+ "step": 14272
+ },
+ {
+ "epoch": 45.60063897763578,
+ "grad_norm": 0.1826171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0278,
+ "step": 14273
+ },
+ {
+ "epoch": 45.6038338658147,
+ "grad_norm": 0.046630859375,
+ "learning_rate": 0.0005,
+ "loss": 1.035,
+ "step": 14274
+ },
+ {
+ "epoch": 45.60702875399361,
+ "grad_norm": 0.17578125,
+ "learning_rate": 0.0005,
+ "loss": 1.037,
+ "step": 14275
+ },
+ {
+ "epoch": 45.61022364217252,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0277,
+ "step": 14276
+ },
+ {
+ "epoch": 45.61341853035144,
+ "grad_norm": 0.10009765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0178,
+ "step": 14277
+ },
+ {
+ "epoch": 45.61661341853035,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0269,
+ "step": 14278
+ },
+ {
+ "epoch": 45.61980830670927,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0233,
+ "step": 14279
+ },
+ {
+ "epoch": 45.62300319488818,
+ "grad_norm": 0.0439453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0325,
+ "step": 14280
+ },
+ {
+ "epoch": 45.626198083067095,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0353,
+ "step": 14281
+ },
+ {
+ "epoch": 45.629392971246006,
+ "grad_norm": 0.0556640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0379,
+ "step": 14282
+ },
+ {
+ "epoch": 45.63258785942492,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0249,
+ "step": 14283
+ },
+ {
+ "epoch": 45.635782747603834,
+ "grad_norm": 0.05859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0175,
+ "step": 14284
+ },
+ {
+ "epoch": 45.638977635782744,
+ "grad_norm": 0.05078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0369,
+ "step": 14285
+ },
+ {
+ "epoch": 45.64217252396166,
+ "grad_norm": 0.0595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0239,
+ "step": 14286
+ },
+ {
+ "epoch": 45.64536741214057,
+ "grad_norm": 0.053955078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0385,
+ "step": 14287
+ },
+ {
+ "epoch": 45.64856230031949,
+ "grad_norm": 0.0517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0298,
+ "step": 14288
+ },
+ {
+ "epoch": 45.6517571884984,
+ "grad_norm": 0.050537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0413,
+ "step": 14289
+ },
+ {
+ "epoch": 45.65495207667732,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.029,
+ "step": 14290
+ },
+ {
+ "epoch": 45.65814696485623,
+ "grad_norm": 0.046630859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0427,
+ "step": 14291
+ },
+ {
+ "epoch": 45.66134185303515,
+ "grad_norm": 0.053466796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0424,
+ "step": 14292
+ },
+ {
+ "epoch": 45.66453674121406,
+ "grad_norm": 0.0595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0307,
+ "step": 14293
+ },
+ {
+ "epoch": 45.66773162939297,
+ "grad_norm": 0.056396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0273,
+ "step": 14294
+ },
+ {
+ "epoch": 45.670926517571885,
+ "grad_norm": 0.08203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0194,
+ "step": 14295
+ },
+ {
+ "epoch": 45.674121405750796,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0303,
+ "step": 14296
+ },
+ {
+ "epoch": 45.677316293929714,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0348,
+ "step": 14297
+ },
+ {
+ "epoch": 45.680511182108624,
+ "grad_norm": 0.04638671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0358,
+ "step": 14298
+ },
+ {
+ "epoch": 45.68370607028754,
+ "grad_norm": 0.087890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0456,
+ "step": 14299
+ },
+ {
+ "epoch": 45.68690095846645,
+ "grad_norm": 0.060302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0367,
+ "step": 14300
+ },
+ {
+ "epoch": 45.69009584664537,
+ "grad_norm": 0.06201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0332,
+ "step": 14301
+ },
+ {
+ "epoch": 45.69329073482428,
+ "grad_norm": 0.1015625,
+ "learning_rate": 0.0005,
+ "loss": 1.025,
+ "step": 14302
+ },
+ {
+ "epoch": 45.6964856230032,
+ "grad_norm": 0.1064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0412,
+ "step": 14303
+ },
+ {
+ "epoch": 45.69968051118211,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.02,
+ "step": 14304
+ },
+ {
+ "epoch": 45.70287539936102,
+ "grad_norm": 0.08203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0422,
+ "step": 14305
+ },
+ {
+ "epoch": 45.70607028753994,
+ "grad_norm": 0.09814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0424,
+ "step": 14306
+ },
+ {
+ "epoch": 45.70926517571885,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0248,
+ "step": 14307
+ },
+ {
+ "epoch": 45.712460063897765,
+ "grad_norm": 0.09619140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0404,
+ "step": 14308
+ },
+ {
+ "epoch": 45.715654952076676,
+ "grad_norm": 0.05908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0339,
+ "step": 14309
+ },
+ {
+ "epoch": 45.718849840255594,
+ "grad_norm": 0.11181640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0379,
+ "step": 14310
+ },
+ {
+ "epoch": 45.722044728434504,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0279,
+ "step": 14311
+ },
+ {
+ "epoch": 45.72523961661342,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0224,
+ "step": 14312
+ },
+ {
+ "epoch": 45.72843450479233,
+ "grad_norm": 0.05419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0126,
+ "step": 14313
+ },
+ {
+ "epoch": 45.73162939297124,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0447,
+ "step": 14314
+ },
+ {
+ "epoch": 45.73482428115016,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0292,
+ "step": 14315
+ },
+ {
+ "epoch": 45.73801916932907,
+ "grad_norm": 0.10791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0302,
+ "step": 14316
+ },
+ {
+ "epoch": 45.74121405750799,
+ "grad_norm": 0.0830078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0271,
+ "step": 14317
+ },
+ {
+ "epoch": 45.7444089456869,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0307,
+ "step": 14318
+ },
+ {
+ "epoch": 45.74760383386582,
+ "grad_norm": 0.12353515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0278,
+ "step": 14319
+ },
+ {
+ "epoch": 45.75079872204473,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0316,
+ "step": 14320
+ },
+ {
+ "epoch": 45.753993610223645,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0339,
+ "step": 14321
+ },
+ {
+ "epoch": 45.757188498402556,
+ "grad_norm": 0.08251953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0366,
+ "step": 14322
+ },
+ {
+ "epoch": 45.760383386581466,
+ "grad_norm": 0.061279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0287,
+ "step": 14323
+ },
+ {
+ "epoch": 45.763578274760384,
+ "grad_norm": 0.0576171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0322,
+ "step": 14324
+ },
+ {
+ "epoch": 45.766773162939295,
+ "grad_norm": 0.0498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0222,
+ "step": 14325
+ },
+ {
+ "epoch": 45.76996805111821,
+ "grad_norm": 0.059326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0241,
+ "step": 14326
+ },
+ {
+ "epoch": 45.77316293929712,
+ "grad_norm": 0.080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.04,
+ "step": 14327
+ },
+ {
+ "epoch": 45.77635782747604,
+ "grad_norm": 0.10205078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0279,
+ "step": 14328
+ },
+ {
+ "epoch": 45.77955271565495,
+ "grad_norm": 0.05810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0222,
+ "step": 14329
+ },
+ {
+ "epoch": 45.78274760383387,
+ "grad_norm": 0.08056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0372,
+ "step": 14330
+ },
+ {
+ "epoch": 45.78594249201278,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0224,
+ "step": 14331
+ },
+ {
+ "epoch": 45.78913738019169,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0198,
+ "step": 14332
+ },
+ {
+ "epoch": 45.79233226837061,
+ "grad_norm": 0.05615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0248,
+ "step": 14333
+ },
+ {
+ "epoch": 45.79552715654952,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0345,
+ "step": 14334
+ },
+ {
+ "epoch": 45.798722044728436,
+ "grad_norm": 0.05322265625,
+ "learning_rate": 0.0005,
+ "loss": 1.032,
+ "step": 14335
+ },
+ {
+ "epoch": 45.801916932907346,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0269,
+ "step": 14336
+ },
+ {
+ "epoch": 45.805111821086264,
+ "grad_norm": 0.05712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0296,
+ "step": 14337
+ },
+ {
+ "epoch": 45.808306709265175,
+ "grad_norm": 0.051025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.019,
+ "step": 14338
+ },
+ {
+ "epoch": 45.81150159744409,
+ "grad_norm": 0.062255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0347,
+ "step": 14339
+ },
+ {
+ "epoch": 45.814696485623,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0264,
+ "step": 14340
+ },
+ {
+ "epoch": 45.81789137380191,
+ "grad_norm": 0.05419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0287,
+ "step": 14341
+ },
+ {
+ "epoch": 45.82108626198083,
+ "grad_norm": 0.057861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0256,
+ "step": 14342
+ },
+ {
+ "epoch": 45.82428115015974,
+ "grad_norm": 0.056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.025,
+ "step": 14343
+ },
+ {
+ "epoch": 45.82747603833866,
+ "grad_norm": 0.055419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0301,
+ "step": 14344
+ },
+ {
+ "epoch": 45.83067092651757,
+ "grad_norm": 0.057861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0355,
+ "step": 14345
+ },
+ {
+ "epoch": 45.83386581469649,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0213,
+ "step": 14346
+ },
+ {
+ "epoch": 45.8370607028754,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0271,
+ "step": 14347
+ },
+ {
+ "epoch": 45.840255591054316,
+ "grad_norm": 0.04931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.04,
+ "step": 14348
+ },
+ {
+ "epoch": 45.843450479233226,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0365,
+ "step": 14349
+ },
+ {
+ "epoch": 45.846645367412144,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0373,
+ "step": 14350
+ },
+ {
+ "epoch": 45.849840255591054,
+ "grad_norm": 0.049560546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0308,
+ "step": 14351
+ },
+ {
+ "epoch": 45.853035143769965,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0316,
+ "step": 14352
+ },
+ {
+ "epoch": 45.85623003194888,
+ "grad_norm": 0.080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0436,
+ "step": 14353
+ },
+ {
+ "epoch": 45.85942492012779,
+ "grad_norm": 0.056884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0295,
+ "step": 14354
+ },
+ {
+ "epoch": 45.86261980830671,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.03,
+ "step": 14355
+ },
+ {
+ "epoch": 45.86581469648562,
+ "grad_norm": 0.052734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0282,
+ "step": 14356
+ },
+ {
+ "epoch": 45.86900958466454,
+ "grad_norm": 0.05908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0277,
+ "step": 14357
+ },
+ {
+ "epoch": 45.87220447284345,
+ "grad_norm": 0.0556640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0362,
+ "step": 14358
+ },
+ {
+ "epoch": 45.87539936102237,
+ "grad_norm": 0.05810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0352,
+ "step": 14359
+ },
+ {
+ "epoch": 45.87859424920128,
+ "grad_norm": 0.05322265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0352,
+ "step": 14360
+ },
+ {
+ "epoch": 45.88178913738019,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0325,
+ "step": 14361
+ },
+ {
+ "epoch": 45.884984025559106,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0264,
+ "step": 14362
+ },
+ {
+ "epoch": 45.88817891373802,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0352,
+ "step": 14363
+ },
+ {
+ "epoch": 45.891373801916934,
+ "grad_norm": 0.050537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0192,
+ "step": 14364
+ },
+ {
+ "epoch": 45.894568690095845,
+ "grad_norm": 0.059814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0253,
+ "step": 14365
+ },
+ {
+ "epoch": 45.89776357827476,
+ "grad_norm": 0.048583984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0279,
+ "step": 14366
+ },
+ {
+ "epoch": 45.90095846645367,
+ "grad_norm": 0.0595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0353,
+ "step": 14367
+ },
+ {
+ "epoch": 45.90415335463259,
+ "grad_norm": 0.056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0352,
+ "step": 14368
+ },
+ {
+ "epoch": 45.9073482428115,
+ "grad_norm": 0.049072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0385,
+ "step": 14369
+ },
+ {
+ "epoch": 45.91054313099041,
+ "grad_norm": 0.047607421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0439,
+ "step": 14370
+ },
+ {
+ "epoch": 45.91373801916933,
+ "grad_norm": 0.056396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0298,
+ "step": 14371
+ },
+ {
+ "epoch": 45.91693290734824,
+ "grad_norm": 0.055908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0277,
+ "step": 14372
+ },
+ {
+ "epoch": 45.92012779552716,
+ "grad_norm": 0.053466796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0262,
+ "step": 14373
+ },
+ {
+ "epoch": 45.92332268370607,
+ "grad_norm": 0.08642578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0409,
+ "step": 14374
+ },
+ {
+ "epoch": 45.926517571884986,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0329,
+ "step": 14375
+ },
+ {
+ "epoch": 45.9297124600639,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0311,
+ "step": 14376
+ },
+ {
+ "epoch": 45.932907348242814,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0353,
+ "step": 14377
+ },
+ {
+ "epoch": 45.936102236421725,
+ "grad_norm": 0.057861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.031,
+ "step": 14378
+ },
+ {
+ "epoch": 45.93929712460064,
+ "grad_norm": 0.05224609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0335,
+ "step": 14379
+ },
+ {
+ "epoch": 45.94249201277955,
+ "grad_norm": 0.050048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0239,
+ "step": 14380
+ },
+ {
+ "epoch": 45.945686900958464,
+ "grad_norm": 0.057861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0209,
+ "step": 14381
+ },
+ {
+ "epoch": 45.94888178913738,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0267,
+ "step": 14382
+ },
+ {
+ "epoch": 45.95207667731629,
+ "grad_norm": 0.051025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0357,
+ "step": 14383
+ },
+ {
+ "epoch": 45.95527156549521,
+ "grad_norm": 0.07861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0277,
+ "step": 14384
+ },
+ {
+ "epoch": 45.95846645367412,
+ "grad_norm": 0.05419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0337,
+ "step": 14385
+ },
+ {
+ "epoch": 45.96166134185304,
+ "grad_norm": 0.11181640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0253,
+ "step": 14386
+ },
+ {
+ "epoch": 45.96485623003195,
+ "grad_norm": 0.080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0188,
+ "step": 14387
+ },
+ {
+ "epoch": 45.968051118210866,
+ "grad_norm": 0.12890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0304,
+ "step": 14388
+ },
+ {
+ "epoch": 45.97124600638978,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0248,
+ "step": 14389
+ },
+ {
+ "epoch": 45.97444089456869,
+ "grad_norm": 0.11279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0333,
+ "step": 14390
+ },
+ {
+ "epoch": 45.977635782747605,
+ "grad_norm": 0.050048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0172,
+ "step": 14391
+ },
+ {
+ "epoch": 45.980830670926515,
+ "grad_norm": 0.1025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0344,
+ "step": 14392
+ },
+ {
+ "epoch": 45.98402555910543,
+ "grad_norm": 0.051025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0268,
+ "step": 14393
+ },
+ {
+ "epoch": 45.98722044728434,
+ "grad_norm": 0.060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0305,
+ "step": 14394
+ },
+ {
+ "epoch": 45.99041533546326,
+ "grad_norm": 0.052978515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0447,
+ "step": 14395
+ },
+ {
+ "epoch": 45.99361022364217,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0387,
+ "step": 14396
+ },
+ {
+ "epoch": 45.99680511182109,
+ "grad_norm": 0.047607421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0283,
+ "step": 14397
+ },
+ {
+ "epoch": 46.0,
+ "grad_norm": 0.058349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0355,
+ "step": 14398
+ },
+ {
+ "epoch": 46.00319488817891,
+ "grad_norm": 0.05029296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0566,
+ "step": 14399
+ },
+ {
+ "epoch": 46.00638977635783,
+ "grad_norm": 0.056884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0377,
+ "step": 14400
+ },
+ {
+ "epoch": 46.00958466453674,
+ "grad_norm": 0.05322265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0287,
+ "step": 14401
+ },
+ {
+ "epoch": 46.01277955271566,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0284,
+ "step": 14402
+ },
+ {
+ "epoch": 46.01597444089457,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0174,
+ "step": 14403
+ },
+ {
+ "epoch": 46.019169329073485,
+ "grad_norm": 0.055908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0336,
+ "step": 14404
+ },
+ {
+ "epoch": 46.022364217252395,
+ "grad_norm": 0.05712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0295,
+ "step": 14405
+ },
+ {
+ "epoch": 46.02555910543131,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0307,
+ "step": 14406
+ },
+ {
+ "epoch": 46.02875399361022,
+ "grad_norm": 0.05615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0371,
+ "step": 14407
+ },
+ {
+ "epoch": 46.031948881789134,
+ "grad_norm": 0.0830078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0357,
+ "step": 14408
+ },
+ {
+ "epoch": 46.03514376996805,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.026,
+ "step": 14409
+ },
+ {
+ "epoch": 46.03833865814696,
+ "grad_norm": 0.0546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0273,
+ "step": 14410
+ },
+ {
+ "epoch": 46.04153354632588,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0237,
+ "step": 14411
+ },
+ {
+ "epoch": 46.04472843450479,
+ "grad_norm": 0.10791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0294,
+ "step": 14412
+ },
+ {
+ "epoch": 46.04792332268371,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0158,
+ "step": 14413
+ },
+ {
+ "epoch": 46.05111821086262,
+ "grad_norm": 0.059814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.036,
+ "step": 14414
+ },
+ {
+ "epoch": 46.054313099041536,
+ "grad_norm": 0.0615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0186,
+ "step": 14415
+ },
+ {
+ "epoch": 46.05750798722045,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0278,
+ "step": 14416
+ },
+ {
+ "epoch": 46.06070287539936,
+ "grad_norm": 0.09423828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0223,
+ "step": 14417
+ },
+ {
+ "epoch": 46.063897763578275,
+ "grad_norm": 0.10498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0444,
+ "step": 14418
+ },
+ {
+ "epoch": 46.067092651757186,
+ "grad_norm": 0.0869140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0351,
+ "step": 14419
+ },
+ {
+ "epoch": 46.0702875399361,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0286,
+ "step": 14420
+ },
+ {
+ "epoch": 46.073482428115014,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0308,
+ "step": 14421
+ },
+ {
+ "epoch": 46.07667731629393,
+ "grad_norm": 0.055419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0126,
+ "step": 14422
+ },
+ {
+ "epoch": 46.07987220447284,
+ "grad_norm": 0.044921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0222,
+ "step": 14423
+ },
+ {
+ "epoch": 46.08306709265176,
+ "grad_norm": 0.07861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0305,
+ "step": 14424
+ },
+ {
+ "epoch": 46.08626198083067,
+ "grad_norm": 0.0537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0144,
+ "step": 14425
+ },
+ {
+ "epoch": 46.08945686900959,
+ "grad_norm": 0.08447265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0248,
+ "step": 14426
+ },
+ {
+ "epoch": 46.0926517571885,
+ "grad_norm": 0.046142578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0195,
+ "step": 14427
+ },
+ {
+ "epoch": 46.09584664536741,
+ "grad_norm": 0.091796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0305,
+ "step": 14428
+ },
+ {
+ "epoch": 46.09904153354633,
+ "grad_norm": 0.049072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.033,
+ "step": 14429
+ },
+ {
+ "epoch": 46.10223642172524,
+ "grad_norm": 0.052001953125,
+ "learning_rate": 0.0005,
+ "loss": 1.029,
+ "step": 14430
+ },
+ {
+ "epoch": 46.105431309904155,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0213,
+ "step": 14431
+ },
+ {
+ "epoch": 46.108626198083066,
+ "grad_norm": 0.052001953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0222,
+ "step": 14432
+ },
+ {
+ "epoch": 46.11182108626198,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0297,
+ "step": 14433
+ },
+ {
+ "epoch": 46.115015974440894,
+ "grad_norm": 0.053466796875,
+ "learning_rate": 0.0005,
+ "loss": 1.028,
+ "step": 14434
+ },
+ {
+ "epoch": 46.11821086261981,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0217,
+ "step": 14435
+ },
+ {
+ "epoch": 46.12140575079872,
+ "grad_norm": 0.0537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0256,
+ "step": 14436
+ },
+ {
+ "epoch": 46.12460063897763,
+ "grad_norm": 0.0810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0233,
+ "step": 14437
+ },
+ {
+ "epoch": 46.12779552715655,
+ "grad_norm": 0.05126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0279,
+ "step": 14438
+ },
+ {
+ "epoch": 46.13099041533546,
+ "grad_norm": 0.09814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0379,
+ "step": 14439
+ },
+ {
+ "epoch": 46.13418530351438,
+ "grad_norm": 0.04833984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0249,
+ "step": 14440
+ },
+ {
+ "epoch": 46.13738019169329,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0278,
+ "step": 14441
+ },
+ {
+ "epoch": 46.14057507987221,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0187,
+ "step": 14442
+ },
+ {
+ "epoch": 46.14376996805112,
+ "grad_norm": 0.10107421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0301,
+ "step": 14443
+ },
+ {
+ "epoch": 46.146964856230035,
+ "grad_norm": 0.05126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0246,
+ "step": 14444
+ },
+ {
+ "epoch": 46.150159744408946,
+ "grad_norm": 0.08447265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0182,
+ "step": 14445
+ },
+ {
+ "epoch": 46.153354632587856,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0262,
+ "step": 14446
+ },
+ {
+ "epoch": 46.156549520766774,
+ "grad_norm": 0.1162109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0263,
+ "step": 14447
+ },
+ {
+ "epoch": 46.159744408945684,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0345,
+ "step": 14448
+ },
+ {
+ "epoch": 46.1629392971246,
+ "grad_norm": 0.16015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0355,
+ "step": 14449
+ },
+ {
+ "epoch": 46.16613418530351,
+ "grad_norm": 0.056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0359,
+ "step": 14450
+ },
+ {
+ "epoch": 46.16932907348243,
+ "grad_norm": 0.177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0281,
+ "step": 14451
+ },
+ {
+ "epoch": 46.17252396166134,
+ "grad_norm": 0.05224609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0314,
+ "step": 14452
+ },
+ {
+ "epoch": 46.17571884984026,
+ "grad_norm": 0.142578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0272,
+ "step": 14453
+ },
+ {
+ "epoch": 46.17891373801917,
+ "grad_norm": 0.054931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0365,
+ "step": 14454
+ },
+ {
+ "epoch": 46.18210862619808,
+ "grad_norm": 0.10205078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0363,
+ "step": 14455
+ },
+ {
+ "epoch": 46.185303514377,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0219,
+ "step": 14456
+ },
+ {
+ "epoch": 46.18849840255591,
+ "grad_norm": 0.09375,
+ "learning_rate": 0.0005,
+ "loss": 1.0322,
+ "step": 14457
+ },
+ {
+ "epoch": 46.191693290734825,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0209,
+ "step": 14458
+ },
+ {
+ "epoch": 46.194888178913736,
+ "grad_norm": 0.10595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0317,
+ "step": 14459
+ },
+ {
+ "epoch": 46.198083067092654,
+ "grad_norm": 0.049560546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0176,
+ "step": 14460
+ },
+ {
+ "epoch": 46.201277955271564,
+ "grad_norm": 0.05322265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0227,
+ "step": 14461
+ },
+ {
+ "epoch": 46.20447284345048,
+ "grad_norm": 0.045166015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0271,
+ "step": 14462
+ },
+ {
+ "epoch": 46.20766773162939,
+ "grad_norm": 0.051513671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0237,
+ "step": 14463
+ },
+ {
+ "epoch": 46.21086261980831,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0182,
+ "step": 14464
+ },
+ {
+ "epoch": 46.21405750798722,
+ "grad_norm": 0.045654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0288,
+ "step": 14465
+ },
+ {
+ "epoch": 46.21725239616613,
+ "grad_norm": 0.06201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0368,
+ "step": 14466
+ },
+ {
+ "epoch": 46.22044728434505,
+ "grad_norm": 0.1337890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0327,
+ "step": 14467
+ },
+ {
+ "epoch": 46.22364217252396,
+ "grad_norm": 0.06005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0222,
+ "step": 14468
+ },
+ {
+ "epoch": 46.22683706070288,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0284,
+ "step": 14469
+ },
+ {
+ "epoch": 46.23003194888179,
+ "grad_norm": 0.05029296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0281,
+ "step": 14470
+ },
+ {
+ "epoch": 46.233226837060705,
+ "grad_norm": 0.05810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.039,
+ "step": 14471
+ },
+ {
+ "epoch": 46.236421725239616,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0294,
+ "step": 14472
+ },
+ {
+ "epoch": 46.239616613418534,
+ "grad_norm": 0.051025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0266,
+ "step": 14473
+ },
+ {
+ "epoch": 46.242811501597444,
+ "grad_norm": 0.0546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0263,
+ "step": 14474
+ },
+ {
+ "epoch": 46.246006389776355,
+ "grad_norm": 0.0498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0311,
+ "step": 14475
+ },
+ {
+ "epoch": 46.24920127795527,
+ "grad_norm": 0.053466796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0269,
+ "step": 14476
+ },
+ {
+ "epoch": 46.25239616613418,
+ "grad_norm": 0.048583984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0243,
+ "step": 14477
+ },
+ {
+ "epoch": 46.2555910543131,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0259,
+ "step": 14478
+ },
+ {
+ "epoch": 46.25878594249201,
+ "grad_norm": 0.05810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0336,
+ "step": 14479
+ },
+ {
+ "epoch": 46.26198083067093,
+ "grad_norm": 0.080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0407,
+ "step": 14480
+ },
+ {
+ "epoch": 46.26517571884984,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0248,
+ "step": 14481
+ },
+ {
+ "epoch": 46.26837060702876,
+ "grad_norm": 0.10595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0233,
+ "step": 14482
+ },
+ {
+ "epoch": 46.27156549520767,
+ "grad_norm": 0.0830078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0217,
+ "step": 14483
+ },
+ {
+ "epoch": 46.27476038338658,
+ "grad_norm": 0.13671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0364,
+ "step": 14484
+ },
+ {
+ "epoch": 46.277955271565496,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.037,
+ "step": 14485
+ },
+ {
+ "epoch": 46.281150159744406,
+ "grad_norm": 0.1611328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0247,
+ "step": 14486
+ },
+ {
+ "epoch": 46.284345047923324,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.021,
+ "step": 14487
+ },
+ {
+ "epoch": 46.287539936102235,
+ "grad_norm": 0.17578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0203,
+ "step": 14488
+ },
+ {
+ "epoch": 46.29073482428115,
+ "grad_norm": 0.055419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.027,
+ "step": 14489
+ },
+ {
+ "epoch": 46.29392971246006,
+ "grad_norm": 0.142578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0324,
+ "step": 14490
+ },
+ {
+ "epoch": 46.29712460063898,
+ "grad_norm": 0.048095703125,
+ "learning_rate": 0.0005,
+ "loss": 1.033,
+ "step": 14491
+ },
+ {
+ "epoch": 46.30031948881789,
+ "grad_norm": 0.1318359375,
+ "learning_rate": 0.0005,
+ "loss": 1.038,
+ "step": 14492
+ },
+ {
+ "epoch": 46.3035143769968,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0322,
+ "step": 14493
+ },
+ {
+ "epoch": 46.30670926517572,
+ "grad_norm": 0.057861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0346,
+ "step": 14494
+ },
+ {
+ "epoch": 46.30990415335463,
+ "grad_norm": 0.054931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0188,
+ "step": 14495
+ },
+ {
+ "epoch": 46.31309904153355,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0394,
+ "step": 14496
+ },
+ {
+ "epoch": 46.31629392971246,
+ "grad_norm": 0.0576171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0369,
+ "step": 14497
+ },
+ {
+ "epoch": 46.319488817891376,
+ "grad_norm": 0.05126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0337,
+ "step": 14498
+ },
+ {
+ "epoch": 46.322683706070286,
+ "grad_norm": 0.05078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0265,
+ "step": 14499
+ },
+ {
+ "epoch": 46.325878594249204,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.023,
+ "step": 14500
+ },
+ {
+ "epoch": 46.329073482428115,
+ "grad_norm": 0.05419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0177,
+ "step": 14501
+ },
+ {
+ "epoch": 46.33226837060703,
+ "grad_norm": 0.05322265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0239,
+ "step": 14502
+ },
+ {
+ "epoch": 46.33546325878594,
+ "grad_norm": 0.05908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0228,
+ "step": 14503
+ },
+ {
+ "epoch": 46.33865814696485,
+ "grad_norm": 0.0537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0304,
+ "step": 14504
+ },
+ {
+ "epoch": 46.34185303514377,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0321,
+ "step": 14505
+ },
+ {
+ "epoch": 46.34504792332268,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0332,
+ "step": 14506
+ },
+ {
+ "epoch": 46.3482428115016,
+ "grad_norm": 0.05517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0281,
+ "step": 14507
+ },
+ {
+ "epoch": 46.35143769968051,
+ "grad_norm": 0.056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0306,
+ "step": 14508
+ },
+ {
+ "epoch": 46.35463258785943,
+ "grad_norm": 0.05078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0205,
+ "step": 14509
+ },
+ {
+ "epoch": 46.35782747603834,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0354,
+ "step": 14510
+ },
+ {
+ "epoch": 46.361022364217256,
+ "grad_norm": 0.046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0262,
+ "step": 14511
+ },
+ {
+ "epoch": 46.364217252396166,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0346,
+ "step": 14512
+ },
+ {
+ "epoch": 46.36741214057508,
+ "grad_norm": 0.05322265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0357,
+ "step": 14513
+ },
+ {
+ "epoch": 46.370607028753994,
+ "grad_norm": 0.053955078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0385,
+ "step": 14514
+ },
+ {
+ "epoch": 46.373801916932905,
+ "grad_norm": 0.0888671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0267,
+ "step": 14515
+ },
+ {
+ "epoch": 46.37699680511182,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0404,
+ "step": 14516
+ },
+ {
+ "epoch": 46.38019169329073,
+ "grad_norm": 0.09423828125,
+ "learning_rate": 0.0005,
+ "loss": 1.026,
+ "step": 14517
+ },
+ {
+ "epoch": 46.38338658146965,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0312,
+ "step": 14518
+ },
+ {
+ "epoch": 46.38658146964856,
+ "grad_norm": 0.10400390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0337,
+ "step": 14519
+ },
+ {
+ "epoch": 46.38977635782748,
+ "grad_norm": 0.05224609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0189,
+ "step": 14520
+ },
+ {
+ "epoch": 46.39297124600639,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0325,
+ "step": 14521
+ },
+ {
+ "epoch": 46.3961661341853,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0357,
+ "step": 14522
+ },
+ {
+ "epoch": 46.39936102236422,
+ "grad_norm": 0.1123046875,
+ "learning_rate": 0.0005,
+ "loss": 1.034,
+ "step": 14523
+ },
+ {
+ "epoch": 46.40255591054313,
+ "grad_norm": 0.1005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0336,
+ "step": 14524
+ },
+ {
+ "epoch": 46.405750798722046,
+ "grad_norm": 0.060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0267,
+ "step": 14525
+ },
+ {
+ "epoch": 46.40894568690096,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0359,
+ "step": 14526
+ },
+ {
+ "epoch": 46.412140575079874,
+ "grad_norm": 0.107421875,
+ "learning_rate": 0.0005,
+ "loss": 1.026,
+ "step": 14527
+ },
+ {
+ "epoch": 46.415335463258785,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.027,
+ "step": 14528
+ },
+ {
+ "epoch": 46.4185303514377,
+ "grad_norm": 0.1259765625,
+ "learning_rate": 0.0005,
+ "loss": 1.035,
+ "step": 14529
+ },
+ {
+ "epoch": 46.42172523961661,
+ "grad_norm": 0.08935546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0174,
+ "step": 14530
+ },
+ {
+ "epoch": 46.424920127795524,
+ "grad_norm": 0.146484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0206,
+ "step": 14531
+ },
+ {
+ "epoch": 46.42811501597444,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0093,
+ "step": 14532
+ },
+ {
+ "epoch": 46.43130990415335,
+ "grad_norm": 0.09033203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0216,
+ "step": 14533
+ },
+ {
+ "epoch": 46.43450479233227,
+ "grad_norm": 0.058349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0254,
+ "step": 14534
+ },
+ {
+ "epoch": 46.43769968051118,
+ "grad_norm": 0.099609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0283,
+ "step": 14535
+ },
+ {
+ "epoch": 46.4408945686901,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0333,
+ "step": 14536
+ },
+ {
+ "epoch": 46.44408945686901,
+ "grad_norm": 0.0849609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0297,
+ "step": 14537
+ },
+ {
+ "epoch": 46.447284345047926,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0262,
+ "step": 14538
+ },
+ {
+ "epoch": 46.45047923322684,
+ "grad_norm": 0.109375,
+ "learning_rate": 0.0005,
+ "loss": 1.034,
+ "step": 14539
+ },
+ {
+ "epoch": 46.453674121405754,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0341,
+ "step": 14540
+ },
+ {
+ "epoch": 46.456869009584665,
+ "grad_norm": 0.09423828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0244,
+ "step": 14541
+ },
+ {
+ "epoch": 46.460063897763575,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0301,
+ "step": 14542
+ },
+ {
+ "epoch": 46.46325878594249,
+ "grad_norm": 0.1357421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0268,
+ "step": 14543
+ },
+ {
+ "epoch": 46.466453674121404,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0407,
+ "step": 14544
+ },
+ {
+ "epoch": 46.46964856230032,
+ "grad_norm": 0.15625,
+ "learning_rate": 0.0005,
+ "loss": 1.0161,
+ "step": 14545
+ },
+ {
+ "epoch": 46.47284345047923,
+ "grad_norm": 0.1865234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0315,
+ "step": 14546
+ },
+ {
+ "epoch": 46.47603833865815,
+ "grad_norm": 0.0966796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0337,
+ "step": 14547
+ },
+ {
+ "epoch": 46.47923322683706,
+ "grad_norm": 0.1533203125,
+ "learning_rate": 0.0005,
+ "loss": 1.031,
+ "step": 14548
+ },
+ {
+ "epoch": 46.48242811501598,
+ "grad_norm": 0.11181640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0264,
+ "step": 14549
+ },
+ {
+ "epoch": 46.48562300319489,
+ "grad_norm": 0.08544921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0256,
+ "step": 14550
+ },
+ {
+ "epoch": 46.4888178913738,
+ "grad_norm": 0.1708984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0411,
+ "step": 14551
+ },
+ {
+ "epoch": 46.49201277955272,
+ "grad_norm": 0.1748046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0279,
+ "step": 14552
+ },
+ {
+ "epoch": 46.49520766773163,
+ "grad_norm": 0.31640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0352,
+ "step": 14553
+ },
+ {
+ "epoch": 46.498402555910545,
+ "grad_norm": 0.2060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0387,
+ "step": 14554
+ },
+ {
+ "epoch": 46.501597444089455,
+ "grad_norm": 0.18359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0336,
+ "step": 14555
+ },
+ {
+ "epoch": 46.50479233226837,
+ "grad_norm": 0.1787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0227,
+ "step": 14556
+ },
+ {
+ "epoch": 46.50798722044728,
+ "grad_norm": 0.169921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0325,
+ "step": 14557
+ },
+ {
+ "epoch": 46.5111821086262,
+ "grad_norm": 0.216796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0295,
+ "step": 14558
+ },
+ {
+ "epoch": 46.51437699680511,
+ "grad_norm": 0.150390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0257,
+ "step": 14559
+ },
+ {
+ "epoch": 46.51757188498402,
+ "grad_norm": 0.306640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0353,
+ "step": 14560
+ },
+ {
+ "epoch": 46.52076677316294,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0409,
+ "step": 14561
+ },
+ {
+ "epoch": 46.52396166134185,
+ "grad_norm": 0.1982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.027,
+ "step": 14562
+ },
+ {
+ "epoch": 46.52715654952077,
+ "grad_norm": 0.0947265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0336,
+ "step": 14563
+ },
+ {
+ "epoch": 46.53035143769968,
+ "grad_norm": 0.2177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0464,
+ "step": 14564
+ },
+ {
+ "epoch": 46.533546325878596,
+ "grad_norm": 0.1376953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0362,
+ "step": 14565
+ },
+ {
+ "epoch": 46.53674121405751,
+ "grad_norm": 0.10400390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0341,
+ "step": 14566
+ },
+ {
+ "epoch": 46.539936102236425,
+ "grad_norm": 0.11279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.031,
+ "step": 14567
+ },
+ {
+ "epoch": 46.543130990415335,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0273,
+ "step": 14568
+ },
+ {
+ "epoch": 46.546325878594246,
+ "grad_norm": 0.1064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0257,
+ "step": 14569
+ },
+ {
+ "epoch": 46.54952076677316,
+ "grad_norm": 0.09765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0204,
+ "step": 14570
+ },
+ {
+ "epoch": 46.552715654952074,
+ "grad_norm": 0.1953125,
+ "learning_rate": 0.0005,
+ "loss": 1.026,
+ "step": 14571
+ },
+ {
+ "epoch": 46.55591054313099,
+ "grad_norm": 0.053955078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0335,
+ "step": 14572
+ },
+ {
+ "epoch": 46.5591054313099,
+ "grad_norm": 0.1611328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0323,
+ "step": 14573
+ },
+ {
+ "epoch": 46.56230031948882,
+ "grad_norm": 0.0888671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0382,
+ "step": 14574
+ },
+ {
+ "epoch": 46.56549520766773,
+ "grad_norm": 0.267578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0348,
+ "step": 14575
+ },
+ {
+ "epoch": 46.56869009584665,
+ "grad_norm": 0.10205078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0304,
+ "step": 14576
+ },
+ {
+ "epoch": 46.57188498402556,
+ "grad_norm": 0.185546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0311,
+ "step": 14577
+ },
+ {
+ "epoch": 46.575079872204476,
+ "grad_norm": 0.181640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0285,
+ "step": 14578
+ },
+ {
+ "epoch": 46.57827476038339,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0348,
+ "step": 14579
+ },
+ {
+ "epoch": 46.5814696485623,
+ "grad_norm": 0.1953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0223,
+ "step": 14580
+ },
+ {
+ "epoch": 46.584664536741215,
+ "grad_norm": 0.1103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0186,
+ "step": 14581
+ },
+ {
+ "epoch": 46.587859424920126,
+ "grad_norm": 0.201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0286,
+ "step": 14582
+ },
+ {
+ "epoch": 46.59105431309904,
+ "grad_norm": 0.2294921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0359,
+ "step": 14583
+ },
+ {
+ "epoch": 46.594249201277954,
+ "grad_norm": 0.1328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0338,
+ "step": 14584
+ },
+ {
+ "epoch": 46.59744408945687,
+ "grad_norm": 0.33203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0326,
+ "step": 14585
+ },
+ {
+ "epoch": 46.60063897763578,
+ "grad_norm": 0.09765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0279,
+ "step": 14586
+ },
+ {
+ "epoch": 46.6038338658147,
+ "grad_norm": 0.3046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0291,
+ "step": 14587
+ },
+ {
+ "epoch": 46.60702875399361,
+ "grad_norm": 0.185546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0313,
+ "step": 14588
+ },
+ {
+ "epoch": 46.61022364217252,
+ "grad_norm": 0.1884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0369,
+ "step": 14589
+ },
+ {
+ "epoch": 46.61341853035144,
+ "grad_norm": 0.166015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0306,
+ "step": 14590
+ },
+ {
+ "epoch": 46.61661341853035,
+ "grad_norm": 0.130859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0318,
+ "step": 14591
+ },
+ {
+ "epoch": 46.61980830670927,
+ "grad_norm": 0.14453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0429,
+ "step": 14592
+ },
+ {
+ "epoch": 46.62300319488818,
+ "grad_norm": 0.2421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0411,
+ "step": 14593
+ },
+ {
+ "epoch": 46.626198083067095,
+ "grad_norm": 0.279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0269,
+ "step": 14594
+ },
+ {
+ "epoch": 46.629392971246006,
+ "grad_norm": 0.1357421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0337,
+ "step": 14595
+ },
+ {
+ "epoch": 46.63258785942492,
+ "grad_norm": 0.458984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0397,
+ "step": 14596
+ },
+ {
+ "epoch": 46.635782747603834,
+ "grad_norm": 0.404296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0342,
+ "step": 14597
+ },
+ {
+ "epoch": 46.638977635782744,
+ "grad_norm": 0.25390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0278,
+ "step": 14598
+ },
+ {
+ "epoch": 46.64217252396166,
+ "grad_norm": 0.15625,
+ "learning_rate": 0.0005,
+ "loss": 1.0224,
+ "step": 14599
+ },
+ {
+ "epoch": 46.64536741214057,
+ "grad_norm": 0.375,
+ "learning_rate": 0.0005,
+ "loss": 1.0321,
+ "step": 14600
+ },
+ {
+ "epoch": 46.64856230031949,
+ "grad_norm": 0.357421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0302,
+ "step": 14601
+ },
+ {
+ "epoch": 46.6517571884984,
+ "grad_norm": 0.08740234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0292,
+ "step": 14602
+ },
+ {
+ "epoch": 46.65495207667732,
+ "grad_norm": 0.248046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0294,
+ "step": 14603
+ },
+ {
+ "epoch": 46.65814696485623,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.021,
+ "step": 14604
+ },
+ {
+ "epoch": 46.66134185303515,
+ "grad_norm": 0.2041015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0333,
+ "step": 14605
+ },
+ {
+ "epoch": 46.66453674121406,
+ "grad_norm": 0.15234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0188,
+ "step": 14606
+ },
+ {
+ "epoch": 46.66773162939297,
+ "grad_norm": 0.166015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0344,
+ "step": 14607
+ },
+ {
+ "epoch": 46.670926517571885,
+ "grad_norm": 0.208984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0361,
+ "step": 14608
+ },
+ {
+ "epoch": 46.674121405750796,
+ "grad_norm": 0.10107421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0263,
+ "step": 14609
+ },
+ {
+ "epoch": 46.677316293929714,
+ "grad_norm": 0.23046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0467,
+ "step": 14610
+ },
+ {
+ "epoch": 46.680511182108624,
+ "grad_norm": 0.08837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.034,
+ "step": 14611
+ },
+ {
+ "epoch": 46.68370607028754,
+ "grad_norm": 0.21484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0358,
+ "step": 14612
+ },
+ {
+ "epoch": 46.68690095846645,
+ "grad_norm": 0.1318359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0319,
+ "step": 14613
+ },
+ {
+ "epoch": 46.69009584664537,
+ "grad_norm": 0.1787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0292,
+ "step": 14614
+ },
+ {
+ "epoch": 46.69329073482428,
+ "grad_norm": 0.1357421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0271,
+ "step": 14615
+ },
+ {
+ "epoch": 46.6964856230032,
+ "grad_norm": 0.15625,
+ "learning_rate": 0.0005,
+ "loss": 1.0328,
+ "step": 14616
+ },
+ {
+ "epoch": 46.69968051118211,
+ "grad_norm": 0.11865234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0258,
+ "step": 14617
+ },
+ {
+ "epoch": 46.70287539936102,
+ "grad_norm": 0.12451171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0357,
+ "step": 14618
+ },
+ {
+ "epoch": 46.70607028753994,
+ "grad_norm": 0.08984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0335,
+ "step": 14619
+ },
+ {
+ "epoch": 46.70926517571885,
+ "grad_norm": 0.1083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0294,
+ "step": 14620
+ },
+ {
+ "epoch": 46.712460063897765,
+ "grad_norm": 0.09814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0385,
+ "step": 14621
+ },
+ {
+ "epoch": 46.715654952076676,
+ "grad_norm": 0.107421875,
+ "learning_rate": 0.0005,
+ "loss": 1.011,
+ "step": 14622
+ },
+ {
+ "epoch": 46.718849840255594,
+ "grad_norm": 0.1865234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0257,
+ "step": 14623
+ },
+ {
+ "epoch": 46.722044728434504,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0353,
+ "step": 14624
+ },
+ {
+ "epoch": 46.72523961661342,
+ "grad_norm": 0.21875,
+ "learning_rate": 0.0005,
+ "loss": 1.0346,
+ "step": 14625
+ },
+ {
+ "epoch": 46.72843450479233,
+ "grad_norm": 0.11572265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0308,
+ "step": 14626
+ },
+ {
+ "epoch": 46.73162939297124,
+ "grad_norm": 0.1416015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0364,
+ "step": 14627
+ },
+ {
+ "epoch": 46.73482428115016,
+ "grad_norm": 0.055419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.022,
+ "step": 14628
+ },
+ {
+ "epoch": 46.73801916932907,
+ "grad_norm": 0.1552734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0172,
+ "step": 14629
+ },
+ {
+ "epoch": 46.74121405750799,
+ "grad_norm": 0.10400390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0285,
+ "step": 14630
+ },
+ {
+ "epoch": 46.7444089456869,
+ "grad_norm": 0.1201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0356,
+ "step": 14631
+ },
+ {
+ "epoch": 46.74760383386582,
+ "grad_norm": 0.12353515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0297,
+ "step": 14632
+ },
+ {
+ "epoch": 46.75079872204473,
+ "grad_norm": 0.10107421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0205,
+ "step": 14633
+ },
+ {
+ "epoch": 46.753993610223645,
+ "grad_norm": 0.11279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0219,
+ "step": 14634
+ },
+ {
+ "epoch": 46.757188498402556,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0207,
+ "step": 14635
+ },
+ {
+ "epoch": 46.760383386581466,
+ "grad_norm": 0.1025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0267,
+ "step": 14636
+ },
+ {
+ "epoch": 46.763578274760384,
+ "grad_norm": 0.08349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0226,
+ "step": 14637
+ },
+ {
+ "epoch": 46.766773162939295,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0376,
+ "step": 14638
+ },
+ {
+ "epoch": 46.76996805111821,
+ "grad_norm": 0.061279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0367,
+ "step": 14639
+ },
+ {
+ "epoch": 46.77316293929712,
+ "grad_norm": 0.0927734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0291,
+ "step": 14640
+ },
+ {
+ "epoch": 46.77635782747604,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0175,
+ "step": 14641
+ },
+ {
+ "epoch": 46.77955271565495,
+ "grad_norm": 0.1005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0256,
+ "step": 14642
+ },
+ {
+ "epoch": 46.78274760383387,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0331,
+ "step": 14643
+ },
+ {
+ "epoch": 46.78594249201278,
+ "grad_norm": 0.08642578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0295,
+ "step": 14644
+ },
+ {
+ "epoch": 46.78913738019169,
+ "grad_norm": 0.08544921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0349,
+ "step": 14645
+ },
+ {
+ "epoch": 46.79233226837061,
+ "grad_norm": 0.1064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0344,
+ "step": 14646
+ },
+ {
+ "epoch": 46.79552715654952,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0333,
+ "step": 14647
+ },
+ {
+ "epoch": 46.798722044728436,
+ "grad_norm": 0.1396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0318,
+ "step": 14648
+ },
+ {
+ "epoch": 46.801916932907346,
+ "grad_norm": 0.05908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0219,
+ "step": 14649
+ },
+ {
+ "epoch": 46.805111821086264,
+ "grad_norm": 0.1728515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0361,
+ "step": 14650
+ },
+ {
+ "epoch": 46.808306709265175,
+ "grad_norm": 0.0810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0356,
+ "step": 14651
+ },
+ {
+ "epoch": 46.81150159744409,
+ "grad_norm": 0.123046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0296,
+ "step": 14652
+ },
+ {
+ "epoch": 46.814696485623,
+ "grad_norm": 0.0859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0222,
+ "step": 14653
+ },
+ {
+ "epoch": 46.81789137380191,
+ "grad_norm": 0.08984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0288,
+ "step": 14654
+ },
+ {
+ "epoch": 46.82108626198083,
+ "grad_norm": 0.0546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0293,
+ "step": 14655
+ },
+ {
+ "epoch": 46.82428115015974,
+ "grad_norm": 0.08984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0468,
+ "step": 14656
+ },
+ {
+ "epoch": 46.82747603833866,
+ "grad_norm": 0.08447265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0213,
+ "step": 14657
+ },
+ {
+ "epoch": 46.83067092651757,
+ "grad_norm": 0.0927734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0222,
+ "step": 14658
+ },
+ {
+ "epoch": 46.83386581469649,
+ "grad_norm": 0.142578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0321,
+ "step": 14659
+ },
+ {
+ "epoch": 46.8370607028754,
+ "grad_norm": 0.10107421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0252,
+ "step": 14660
+ },
+ {
+ "epoch": 46.840255591054316,
+ "grad_norm": 0.1748046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0235,
+ "step": 14661
+ },
+ {
+ "epoch": 46.843450479233226,
+ "grad_norm": 0.10546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0173,
+ "step": 14662
+ },
+ {
+ "epoch": 46.846645367412144,
+ "grad_norm": 0.1640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0288,
+ "step": 14663
+ },
+ {
+ "epoch": 46.849840255591054,
+ "grad_norm": 0.0947265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0377,
+ "step": 14664
+ },
+ {
+ "epoch": 46.853035143769965,
+ "grad_norm": 0.1689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.018,
+ "step": 14665
+ },
+ {
+ "epoch": 46.85623003194888,
+ "grad_norm": 0.052490234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0338,
+ "step": 14666
+ },
+ {
+ "epoch": 46.85942492012779,
+ "grad_norm": 0.1591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0202,
+ "step": 14667
+ },
+ {
+ "epoch": 46.86261980830671,
+ "grad_norm": 0.0478515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0237,
+ "step": 14668
+ },
+ {
+ "epoch": 46.86581469648562,
+ "grad_norm": 0.15625,
+ "learning_rate": 0.0005,
+ "loss": 1.0243,
+ "step": 14669
+ },
+ {
+ "epoch": 46.86900958466454,
+ "grad_norm": 0.05712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0308,
+ "step": 14670
+ },
+ {
+ "epoch": 46.87220447284345,
+ "grad_norm": 0.1494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0425,
+ "step": 14671
+ },
+ {
+ "epoch": 46.87539936102237,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0362,
+ "step": 14672
+ },
+ {
+ "epoch": 46.87859424920128,
+ "grad_norm": 0.130859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0362,
+ "step": 14673
+ },
+ {
+ "epoch": 46.88178913738019,
+ "grad_norm": 0.0517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0403,
+ "step": 14674
+ },
+ {
+ "epoch": 46.884984025559106,
+ "grad_norm": 0.08056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0331,
+ "step": 14675
+ },
+ {
+ "epoch": 46.88817891373802,
+ "grad_norm": 0.0576171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0343,
+ "step": 14676
+ },
+ {
+ "epoch": 46.891373801916934,
+ "grad_norm": 0.0859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0379,
+ "step": 14677
+ },
+ {
+ "epoch": 46.894568690095845,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0295,
+ "step": 14678
+ },
+ {
+ "epoch": 46.89776357827476,
+ "grad_norm": 0.1025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0386,
+ "step": 14679
+ },
+ {
+ "epoch": 46.90095846645367,
+ "grad_norm": 0.056396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0256,
+ "step": 14680
+ },
+ {
+ "epoch": 46.90415335463259,
+ "grad_norm": 0.10107421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0348,
+ "step": 14681
+ },
+ {
+ "epoch": 46.9073482428115,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0382,
+ "step": 14682
+ },
+ {
+ "epoch": 46.91054313099041,
+ "grad_norm": 0.11376953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0347,
+ "step": 14683
+ },
+ {
+ "epoch": 46.91373801916933,
+ "grad_norm": 0.052490234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0262,
+ "step": 14684
+ },
+ {
+ "epoch": 46.91693290734824,
+ "grad_norm": 0.11181640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0283,
+ "step": 14685
+ },
+ {
+ "epoch": 46.92012779552716,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0341,
+ "step": 14686
+ },
+ {
+ "epoch": 46.92332268370607,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.022,
+ "step": 14687
+ },
+ {
+ "epoch": 46.926517571884986,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0285,
+ "step": 14688
+ },
+ {
+ "epoch": 46.9297124600639,
+ "grad_norm": 0.056884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0371,
+ "step": 14689
+ },
+ {
+ "epoch": 46.932907348242814,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0285,
+ "step": 14690
+ },
+ {
+ "epoch": 46.936102236421725,
+ "grad_norm": 0.06005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0291,
+ "step": 14691
+ },
+ {
+ "epoch": 46.93929712460064,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.024,
+ "step": 14692
+ },
+ {
+ "epoch": 46.94249201277955,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0359,
+ "step": 14693
+ },
+ {
+ "epoch": 46.945686900958464,
+ "grad_norm": 0.11865234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0376,
+ "step": 14694
+ },
+ {
+ "epoch": 46.94888178913738,
+ "grad_norm": 0.11083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0141,
+ "step": 14695
+ },
+ {
+ "epoch": 46.95207667731629,
+ "grad_norm": 0.08251953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0222,
+ "step": 14696
+ },
+ {
+ "epoch": 46.95527156549521,
+ "grad_norm": 0.09521484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0146,
+ "step": 14697
+ },
+ {
+ "epoch": 46.95846645367412,
+ "grad_norm": 0.059326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0294,
+ "step": 14698
+ },
+ {
+ "epoch": 46.96166134185304,
+ "grad_norm": 0.0576171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0251,
+ "step": 14699
+ },
+ {
+ "epoch": 46.96485623003195,
+ "grad_norm": 0.0908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0213,
+ "step": 14700
+ },
+ {
+ "epoch": 46.968051118210866,
+ "grad_norm": 0.11962890625,
+ "learning_rate": 0.0005,
+ "loss": 1.009,
+ "step": 14701
+ },
+ {
+ "epoch": 46.97124600638978,
+ "grad_norm": 0.08056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0389,
+ "step": 14702
+ },
+ {
+ "epoch": 46.97444089456869,
+ "grad_norm": 0.1376953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0176,
+ "step": 14703
+ },
+ {
+ "epoch": 46.977635782747605,
+ "grad_norm": 0.1162109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0295,
+ "step": 14704
+ },
+ {
+ "epoch": 46.980830670926515,
+ "grad_norm": 0.2109375,
+ "learning_rate": 0.0005,
+ "loss": 1.034,
+ "step": 14705
+ },
+ {
+ "epoch": 46.98402555910543,
+ "grad_norm": 0.0869140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0362,
+ "step": 14706
+ },
+ {
+ "epoch": 46.98722044728434,
+ "grad_norm": 0.2060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0296,
+ "step": 14707
+ },
+ {
+ "epoch": 46.99041533546326,
+ "grad_norm": 0.1005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0387,
+ "step": 14708
+ },
+ {
+ "epoch": 46.99361022364217,
+ "grad_norm": 0.1669921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0292,
+ "step": 14709
+ },
+ {
+ "epoch": 46.99680511182109,
+ "grad_norm": 0.1435546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0289,
+ "step": 14710
+ },
+ {
+ "epoch": 47.0,
+ "grad_norm": 0.1142578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0374,
+ "step": 14711
+ },
+ {
+ "epoch": 47.00319488817891,
+ "grad_norm": 0.1630859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0267,
+ "step": 14712
+ },
+ {
+ "epoch": 47.00638977635783,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0203,
+ "step": 14713
+ },
+ {
+ "epoch": 47.00958466453674,
+ "grad_norm": 0.2041015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0252,
+ "step": 14714
+ },
+ {
+ "epoch": 47.01277955271566,
+ "grad_norm": 0.1171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0299,
+ "step": 14715
+ },
+ {
+ "epoch": 47.01597444089457,
+ "grad_norm": 0.1416015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0302,
+ "step": 14716
+ },
+ {
+ "epoch": 47.019169329073485,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0332,
+ "step": 14717
+ },
+ {
+ "epoch": 47.022364217252395,
+ "grad_norm": 0.10546875,
+ "learning_rate": 0.0005,
+ "loss": 1.03,
+ "step": 14718
+ },
+ {
+ "epoch": 47.02555910543131,
+ "grad_norm": 0.048583984375,
+ "learning_rate": 0.0005,
+ "loss": 1.02,
+ "step": 14719
+ },
+ {
+ "epoch": 47.02875399361022,
+ "grad_norm": 0.1328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0385,
+ "step": 14720
+ },
+ {
+ "epoch": 47.031948881789134,
+ "grad_norm": 0.0517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0287,
+ "step": 14721
+ },
+ {
+ "epoch": 47.03514376996805,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0322,
+ "step": 14722
+ },
+ {
+ "epoch": 47.03833865814696,
+ "grad_norm": 0.052001953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0437,
+ "step": 14723
+ },
+ {
+ "epoch": 47.04153354632588,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0355,
+ "step": 14724
+ },
+ {
+ "epoch": 47.04472843450479,
+ "grad_norm": 0.0546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0315,
+ "step": 14725
+ },
+ {
+ "epoch": 47.04792332268371,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0228,
+ "step": 14726
+ },
+ {
+ "epoch": 47.05111821086262,
+ "grad_norm": 0.052490234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0294,
+ "step": 14727
+ },
+ {
+ "epoch": 47.054313099041536,
+ "grad_norm": 0.087890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0269,
+ "step": 14728
+ },
+ {
+ "epoch": 47.05750798722045,
+ "grad_norm": 0.049072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0365,
+ "step": 14729
+ },
+ {
+ "epoch": 47.06070287539936,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0243,
+ "step": 14730
+ },
+ {
+ "epoch": 47.063897763578275,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0279,
+ "step": 14731
+ },
+ {
+ "epoch": 47.067092651757186,
+ "grad_norm": 0.0810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.024,
+ "step": 14732
+ },
+ {
+ "epoch": 47.0702875399361,
+ "grad_norm": 0.052978515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0223,
+ "step": 14733
+ },
+ {
+ "epoch": 47.073482428115014,
+ "grad_norm": 0.10546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0287,
+ "step": 14734
+ },
+ {
+ "epoch": 47.07667731629393,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0271,
+ "step": 14735
+ },
+ {
+ "epoch": 47.07987220447284,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0345,
+ "step": 14736
+ },
+ {
+ "epoch": 47.08306709265176,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0312,
+ "step": 14737
+ },
+ {
+ "epoch": 47.08626198083067,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0375,
+ "step": 14738
+ },
+ {
+ "epoch": 47.08945686900959,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0152,
+ "step": 14739
+ },
+ {
+ "epoch": 47.0926517571885,
+ "grad_norm": 0.051513671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0364,
+ "step": 14740
+ },
+ {
+ "epoch": 47.09584664536741,
+ "grad_norm": 0.059814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.035,
+ "step": 14741
+ },
+ {
+ "epoch": 47.09904153354633,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0265,
+ "step": 14742
+ },
+ {
+ "epoch": 47.10223642172524,
+ "grad_norm": 0.046630859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0309,
+ "step": 14743
+ },
+ {
+ "epoch": 47.105431309904155,
+ "grad_norm": 0.049072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0373,
+ "step": 14744
+ },
+ {
+ "epoch": 47.108626198083066,
+ "grad_norm": 0.053955078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0276,
+ "step": 14745
+ },
+ {
+ "epoch": 47.11182108626198,
+ "grad_norm": 0.050537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0244,
+ "step": 14746
+ },
+ {
+ "epoch": 47.115015974440894,
+ "grad_norm": 0.05126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0426,
+ "step": 14747
+ },
+ {
+ "epoch": 47.11821086261981,
+ "grad_norm": 0.04248046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0201,
+ "step": 14748
+ },
+ {
+ "epoch": 47.12140575079872,
+ "grad_norm": 0.046142578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0302,
+ "step": 14749
+ },
+ {
+ "epoch": 47.12460063897763,
+ "grad_norm": 0.046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0342,
+ "step": 14750
+ },
+ {
+ "epoch": 47.12779552715655,
+ "grad_norm": 0.05029296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0277,
+ "step": 14751
+ },
+ {
+ "epoch": 47.13099041533546,
+ "grad_norm": 0.053466796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0361,
+ "step": 14752
+ },
+ {
+ "epoch": 47.13418530351438,
+ "grad_norm": 0.0576171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0276,
+ "step": 14753
+ },
+ {
+ "epoch": 47.13738019169329,
+ "grad_norm": 0.046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0333,
+ "step": 14754
+ },
+ {
+ "epoch": 47.14057507987221,
+ "grad_norm": 0.051025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.034,
+ "step": 14755
+ },
+ {
+ "epoch": 47.14376996805112,
+ "grad_norm": 0.048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0398,
+ "step": 14756
+ },
+ {
+ "epoch": 47.146964856230035,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0169,
+ "step": 14757
+ },
+ {
+ "epoch": 47.150159744408946,
+ "grad_norm": 0.061279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0302,
+ "step": 14758
+ },
+ {
+ "epoch": 47.153354632587856,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0334,
+ "step": 14759
+ },
+ {
+ "epoch": 47.156549520766774,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0237,
+ "step": 14760
+ },
+ {
+ "epoch": 47.159744408945684,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0317,
+ "step": 14761
+ },
+ {
+ "epoch": 47.1629392971246,
+ "grad_norm": 0.0810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0249,
+ "step": 14762
+ },
+ {
+ "epoch": 47.16613418530351,
+ "grad_norm": 0.046630859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0144,
+ "step": 14763
+ },
+ {
+ "epoch": 47.16932907348243,
+ "grad_norm": 0.060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0276,
+ "step": 14764
+ },
+ {
+ "epoch": 47.17252396166134,
+ "grad_norm": 0.0546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0216,
+ "step": 14765
+ },
+ {
+ "epoch": 47.17571884984026,
+ "grad_norm": 0.04443359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0264,
+ "step": 14766
+ },
+ {
+ "epoch": 47.17891373801917,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0327,
+ "step": 14767
+ },
+ {
+ "epoch": 47.18210862619808,
+ "grad_norm": 0.053955078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0197,
+ "step": 14768
+ },
+ {
+ "epoch": 47.185303514377,
+ "grad_norm": 0.0537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0332,
+ "step": 14769
+ },
+ {
+ "epoch": 47.18849840255591,
+ "grad_norm": 0.0615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0265,
+ "step": 14770
+ },
+ {
+ "epoch": 47.191693290734825,
+ "grad_norm": 0.08935546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0179,
+ "step": 14771
+ },
+ {
+ "epoch": 47.194888178913736,
+ "grad_norm": 0.08203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0254,
+ "step": 14772
+ },
+ {
+ "epoch": 47.198083067092654,
+ "grad_norm": 0.0830078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0196,
+ "step": 14773
+ },
+ {
+ "epoch": 47.201277955271564,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0292,
+ "step": 14774
+ },
+ {
+ "epoch": 47.20447284345048,
+ "grad_norm": 0.062255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0223,
+ "step": 14775
+ },
+ {
+ "epoch": 47.20766773162939,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0223,
+ "step": 14776
+ },
+ {
+ "epoch": 47.21086261980831,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0264,
+ "step": 14777
+ },
+ {
+ "epoch": 47.21405750798722,
+ "grad_norm": 0.0830078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0277,
+ "step": 14778
+ },
+ {
+ "epoch": 47.21725239616613,
+ "grad_norm": 0.05078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0448,
+ "step": 14779
+ },
+ {
+ "epoch": 47.22044728434505,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0305,
+ "step": 14780
+ },
+ {
+ "epoch": 47.22364217252396,
+ "grad_norm": 0.045166015625,
+ "learning_rate": 0.0005,
+ "loss": 1.027,
+ "step": 14781
+ },
+ {
+ "epoch": 47.22683706070288,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0238,
+ "step": 14782
+ },
+ {
+ "epoch": 47.23003194888179,
+ "grad_norm": 0.06201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0158,
+ "step": 14783
+ },
+ {
+ "epoch": 47.233226837060705,
+ "grad_norm": 0.08447265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0337,
+ "step": 14784
+ },
+ {
+ "epoch": 47.236421725239616,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0176,
+ "step": 14785
+ },
+ {
+ "epoch": 47.239616613418534,
+ "grad_norm": 0.1123046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0335,
+ "step": 14786
+ },
+ {
+ "epoch": 47.242811501597444,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0312,
+ "step": 14787
+ },
+ {
+ "epoch": 47.246006389776355,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0312,
+ "step": 14788
+ },
+ {
+ "epoch": 47.24920127795527,
+ "grad_norm": 0.056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0309,
+ "step": 14789
+ },
+ {
+ "epoch": 47.25239616613418,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0339,
+ "step": 14790
+ },
+ {
+ "epoch": 47.2555910543131,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0291,
+ "step": 14791
+ },
+ {
+ "epoch": 47.25878594249201,
+ "grad_norm": 0.052734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0367,
+ "step": 14792
+ },
+ {
+ "epoch": 47.26198083067093,
+ "grad_norm": 0.07861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0182,
+ "step": 14793
+ },
+ {
+ "epoch": 47.26517571884984,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.019,
+ "step": 14794
+ },
+ {
+ "epoch": 47.26837060702876,
+ "grad_norm": 0.0556640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0231,
+ "step": 14795
+ },
+ {
+ "epoch": 47.27156549520767,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0311,
+ "step": 14796
+ },
+ {
+ "epoch": 47.27476038338658,
+ "grad_norm": 0.06005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0265,
+ "step": 14797
+ },
+ {
+ "epoch": 47.277955271565496,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0253,
+ "step": 14798
+ },
+ {
+ "epoch": 47.281150159744406,
+ "grad_norm": 0.09765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0394,
+ "step": 14799
+ },
+ {
+ "epoch": 47.284345047923324,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0291,
+ "step": 14800
+ },
+ {
+ "epoch": 47.287539936102235,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.025,
+ "step": 14801
+ },
+ {
+ "epoch": 47.29073482428115,
+ "grad_norm": 0.060302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0273,
+ "step": 14802
+ },
+ {
+ "epoch": 47.29392971246006,
+ "grad_norm": 0.048095703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0287,
+ "step": 14803
+ },
+ {
+ "epoch": 47.29712460063898,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.033,
+ "step": 14804
+ },
+ {
+ "epoch": 47.30031948881789,
+ "grad_norm": 0.05078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0361,
+ "step": 14805
+ },
+ {
+ "epoch": 47.3035143769968,
+ "grad_norm": 0.060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0448,
+ "step": 14806
+ },
+ {
+ "epoch": 47.30670926517572,
+ "grad_norm": 0.095703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0261,
+ "step": 14807
+ },
+ {
+ "epoch": 47.30990415335463,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0316,
+ "step": 14808
+ },
+ {
+ "epoch": 47.31309904153355,
+ "grad_norm": 0.08642578125,
+ "learning_rate": 0.0005,
+ "loss": 1.045,
+ "step": 14809
+ },
+ {
+ "epoch": 47.31629392971246,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0212,
+ "step": 14810
+ },
+ {
+ "epoch": 47.319488817891376,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0302,
+ "step": 14811
+ },
+ {
+ "epoch": 47.322683706070286,
+ "grad_norm": 0.08984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0169,
+ "step": 14812
+ },
+ {
+ "epoch": 47.325878594249204,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0326,
+ "step": 14813
+ },
+ {
+ "epoch": 47.329073482428115,
+ "grad_norm": 0.04931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0391,
+ "step": 14814
+ },
+ {
+ "epoch": 47.33226837060703,
+ "grad_norm": 0.0576171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0237,
+ "step": 14815
+ },
+ {
+ "epoch": 47.33546325878594,
+ "grad_norm": 0.087890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0346,
+ "step": 14816
+ },
+ {
+ "epoch": 47.33865814696485,
+ "grad_norm": 0.05224609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0236,
+ "step": 14817
+ },
+ {
+ "epoch": 47.34185303514377,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0219,
+ "step": 14818
+ },
+ {
+ "epoch": 47.34504792332268,
+ "grad_norm": 0.09326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0409,
+ "step": 14819
+ },
+ {
+ "epoch": 47.3482428115016,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0307,
+ "step": 14820
+ },
+ {
+ "epoch": 47.35143769968051,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0354,
+ "step": 14821
+ },
+ {
+ "epoch": 47.35463258785943,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0385,
+ "step": 14822
+ },
+ {
+ "epoch": 47.35782747603834,
+ "grad_norm": 0.0908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0307,
+ "step": 14823
+ },
+ {
+ "epoch": 47.361022364217256,
+ "grad_norm": 0.0537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0331,
+ "step": 14824
+ },
+ {
+ "epoch": 47.364217252396166,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0314,
+ "step": 14825
+ },
+ {
+ "epoch": 47.36741214057508,
+ "grad_norm": 0.048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0238,
+ "step": 14826
+ },
+ {
+ "epoch": 47.370607028753994,
+ "grad_norm": 0.061279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0335,
+ "step": 14827
+ },
+ {
+ "epoch": 47.373801916932905,
+ "grad_norm": 0.06005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.012,
+ "step": 14828
+ },
+ {
+ "epoch": 47.37699680511182,
+ "grad_norm": 0.05224609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0242,
+ "step": 14829
+ },
+ {
+ "epoch": 47.38019169329073,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0214,
+ "step": 14830
+ },
+ {
+ "epoch": 47.38338658146965,
+ "grad_norm": 0.057373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0218,
+ "step": 14831
+ },
+ {
+ "epoch": 47.38658146964856,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0409,
+ "step": 14832
+ },
+ {
+ "epoch": 47.38977635782748,
+ "grad_norm": 0.05517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0409,
+ "step": 14833
+ },
+ {
+ "epoch": 47.39297124600639,
+ "grad_norm": 0.055908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0185,
+ "step": 14834
+ },
+ {
+ "epoch": 47.3961661341853,
+ "grad_norm": 0.07861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0305,
+ "step": 14835
+ },
+ {
+ "epoch": 47.39936102236422,
+ "grad_norm": 0.09619140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0422,
+ "step": 14836
+ },
+ {
+ "epoch": 47.40255591054313,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0264,
+ "step": 14837
+ },
+ {
+ "epoch": 47.405750798722046,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0294,
+ "step": 14838
+ },
+ {
+ "epoch": 47.40894568690096,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0343,
+ "step": 14839
+ },
+ {
+ "epoch": 47.412140575079874,
+ "grad_norm": 0.054931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0337,
+ "step": 14840
+ },
+ {
+ "epoch": 47.415335463258785,
+ "grad_norm": 0.0517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0229,
+ "step": 14841
+ },
+ {
+ "epoch": 47.4185303514377,
+ "grad_norm": 0.06005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0255,
+ "step": 14842
+ },
+ {
+ "epoch": 47.42172523961661,
+ "grad_norm": 0.05078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0291,
+ "step": 14843
+ },
+ {
+ "epoch": 47.424920127795524,
+ "grad_norm": 0.0517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0254,
+ "step": 14844
+ },
+ {
+ "epoch": 47.42811501597444,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0174,
+ "step": 14845
+ },
+ {
+ "epoch": 47.43130990415335,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.029,
+ "step": 14846
+ },
+ {
+ "epoch": 47.43450479233227,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0271,
+ "step": 14847
+ },
+ {
+ "epoch": 47.43769968051118,
+ "grad_norm": 0.047119140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0396,
+ "step": 14848
+ },
+ {
+ "epoch": 47.4408945686901,
+ "grad_norm": 0.080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0308,
+ "step": 14849
+ },
+ {
+ "epoch": 47.44408945686901,
+ "grad_norm": 0.0498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0401,
+ "step": 14850
+ },
+ {
+ "epoch": 47.447284345047926,
+ "grad_norm": 0.049560546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0191,
+ "step": 14851
+ },
+ {
+ "epoch": 47.45047923322684,
+ "grad_norm": 0.056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.031,
+ "step": 14852
+ },
+ {
+ "epoch": 47.453674121405754,
+ "grad_norm": 0.054931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0364,
+ "step": 14853
+ },
+ {
+ "epoch": 47.456869009584665,
+ "grad_norm": 0.09765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0344,
+ "step": 14854
+ },
+ {
+ "epoch": 47.460063897763575,
+ "grad_norm": 0.062255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0324,
+ "step": 14855
+ },
+ {
+ "epoch": 47.46325878594249,
+ "grad_norm": 0.09326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0179,
+ "step": 14856
+ },
+ {
+ "epoch": 47.466453674121404,
+ "grad_norm": 0.056396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0215,
+ "step": 14857
+ },
+ {
+ "epoch": 47.46964856230032,
+ "grad_norm": 0.08349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.033,
+ "step": 14858
+ },
+ {
+ "epoch": 47.47284345047923,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0366,
+ "step": 14859
+ },
+ {
+ "epoch": 47.47603833865815,
+ "grad_norm": 0.052490234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0181,
+ "step": 14860
+ },
+ {
+ "epoch": 47.47923322683706,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0309,
+ "step": 14861
+ },
+ {
+ "epoch": 47.48242811501598,
+ "grad_norm": 0.06201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0223,
+ "step": 14862
+ },
+ {
+ "epoch": 47.48562300319489,
+ "grad_norm": 0.047119140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0362,
+ "step": 14863
+ },
+ {
+ "epoch": 47.4888178913738,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0232,
+ "step": 14864
+ },
+ {
+ "epoch": 47.49201277955272,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0308,
+ "step": 14865
+ },
+ {
+ "epoch": 47.49520766773163,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0437,
+ "step": 14866
+ },
+ {
+ "epoch": 47.498402555910545,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0309,
+ "step": 14867
+ },
+ {
+ "epoch": 47.501597444089455,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0266,
+ "step": 14868
+ },
+ {
+ "epoch": 47.50479233226837,
+ "grad_norm": 0.05810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0287,
+ "step": 14869
+ },
+ {
+ "epoch": 47.50798722044728,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0313,
+ "step": 14870
+ },
+ {
+ "epoch": 47.5111821086262,
+ "grad_norm": 0.06005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.035,
+ "step": 14871
+ },
+ {
+ "epoch": 47.51437699680511,
+ "grad_norm": 0.08154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0332,
+ "step": 14872
+ },
+ {
+ "epoch": 47.51757188498402,
+ "grad_norm": 0.062255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0248,
+ "step": 14873
+ },
+ {
+ "epoch": 47.52076677316294,
+ "grad_norm": 0.08642578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0109,
+ "step": 14874
+ },
+ {
+ "epoch": 47.52396166134185,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0341,
+ "step": 14875
+ },
+ {
+ "epoch": 47.52715654952077,
+ "grad_norm": 0.0869140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0274,
+ "step": 14876
+ },
+ {
+ "epoch": 47.53035143769968,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0361,
+ "step": 14877
+ },
+ {
+ "epoch": 47.533546325878596,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0169,
+ "step": 14878
+ },
+ {
+ "epoch": 47.53674121405751,
+ "grad_norm": 0.052734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0177,
+ "step": 14879
+ },
+ {
+ "epoch": 47.539936102236425,
+ "grad_norm": 0.06103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0297,
+ "step": 14880
+ },
+ {
+ "epoch": 47.543130990415335,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0315,
+ "step": 14881
+ },
+ {
+ "epoch": 47.546325878594246,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.034,
+ "step": 14882
+ },
+ {
+ "epoch": 47.54952076677316,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0297,
+ "step": 14883
+ },
+ {
+ "epoch": 47.552715654952074,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0237,
+ "step": 14884
+ },
+ {
+ "epoch": 47.55591054313099,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0269,
+ "step": 14885
+ },
+ {
+ "epoch": 47.5591054313099,
+ "grad_norm": 0.10302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0304,
+ "step": 14886
+ },
+ {
+ "epoch": 47.56230031948882,
+ "grad_norm": 0.0830078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0208,
+ "step": 14887
+ },
+ {
+ "epoch": 47.56549520766773,
+ "grad_norm": 0.09619140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0288,
+ "step": 14888
+ },
+ {
+ "epoch": 47.56869009584665,
+ "grad_norm": 0.08154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.026,
+ "step": 14889
+ },
+ {
+ "epoch": 47.57188498402556,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0405,
+ "step": 14890
+ },
+ {
+ "epoch": 47.575079872204476,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0367,
+ "step": 14891
+ },
+ {
+ "epoch": 47.57827476038339,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0228,
+ "step": 14892
+ },
+ {
+ "epoch": 47.5814696485623,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0375,
+ "step": 14893
+ },
+ {
+ "epoch": 47.584664536741215,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0365,
+ "step": 14894
+ },
+ {
+ "epoch": 47.587859424920126,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0479,
+ "step": 14895
+ },
+ {
+ "epoch": 47.59105431309904,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0205,
+ "step": 14896
+ },
+ {
+ "epoch": 47.594249201277954,
+ "grad_norm": 0.05859375,
+ "learning_rate": 0.0005,
+ "loss": 1.033,
+ "step": 14897
+ },
+ {
+ "epoch": 47.59744408945687,
+ "grad_norm": 0.0810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0235,
+ "step": 14898
+ },
+ {
+ "epoch": 47.60063897763578,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0279,
+ "step": 14899
+ },
+ {
+ "epoch": 47.6038338658147,
+ "grad_norm": 0.09912109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0329,
+ "step": 14900
+ },
+ {
+ "epoch": 47.60702875399361,
+ "grad_norm": 0.10107421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0219,
+ "step": 14901
+ },
+ {
+ "epoch": 47.61022364217252,
+ "grad_norm": 0.14453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0291,
+ "step": 14902
+ },
+ {
+ "epoch": 47.61341853035144,
+ "grad_norm": 0.057373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0434,
+ "step": 14903
+ },
+ {
+ "epoch": 47.61661341853035,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0212,
+ "step": 14904
+ },
+ {
+ "epoch": 47.61980830670927,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0269,
+ "step": 14905
+ },
+ {
+ "epoch": 47.62300319488818,
+ "grad_norm": 0.09423828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0269,
+ "step": 14906
+ },
+ {
+ "epoch": 47.626198083067095,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0306,
+ "step": 14907
+ },
+ {
+ "epoch": 47.629392971246006,
+ "grad_norm": 0.140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0327,
+ "step": 14908
+ },
+ {
+ "epoch": 47.63258785942492,
+ "grad_norm": 0.05029296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0408,
+ "step": 14909
+ },
+ {
+ "epoch": 47.635782747603834,
+ "grad_norm": 0.1396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0197,
+ "step": 14910
+ },
+ {
+ "epoch": 47.638977635782744,
+ "grad_norm": 0.08056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0302,
+ "step": 14911
+ },
+ {
+ "epoch": 47.64217252396166,
+ "grad_norm": 0.142578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0341,
+ "step": 14912
+ },
+ {
+ "epoch": 47.64536741214057,
+ "grad_norm": 0.09765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0379,
+ "step": 14913
+ },
+ {
+ "epoch": 47.64856230031949,
+ "grad_norm": 0.154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0292,
+ "step": 14914
+ },
+ {
+ "epoch": 47.6517571884984,
+ "grad_norm": 0.10107421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0329,
+ "step": 14915
+ },
+ {
+ "epoch": 47.65495207667732,
+ "grad_norm": 0.11083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.046,
+ "step": 14916
+ },
+ {
+ "epoch": 47.65814696485623,
+ "grad_norm": 0.1259765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0367,
+ "step": 14917
+ },
+ {
+ "epoch": 47.66134185303515,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0316,
+ "step": 14918
+ },
+ {
+ "epoch": 47.66453674121406,
+ "grad_norm": 0.1259765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0272,
+ "step": 14919
+ },
+ {
+ "epoch": 47.66773162939297,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0378,
+ "step": 14920
+ },
+ {
+ "epoch": 47.670926517571885,
+ "grad_norm": 0.21484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0314,
+ "step": 14921
+ },
+ {
+ "epoch": 47.674121405750796,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0318,
+ "step": 14922
+ },
+ {
+ "epoch": 47.677316293929714,
+ "grad_norm": 0.1767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0389,
+ "step": 14923
+ },
+ {
+ "epoch": 47.680511182108624,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0449,
+ "step": 14924
+ },
+ {
+ "epoch": 47.68370607028754,
+ "grad_norm": 0.1533203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0501,
+ "step": 14925
+ },
+ {
+ "epoch": 47.68690095846645,
+ "grad_norm": 0.08740234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0277,
+ "step": 14926
+ },
+ {
+ "epoch": 47.69009584664537,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0298,
+ "step": 14927
+ },
+ {
+ "epoch": 47.69329073482428,
+ "grad_norm": 0.054931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0277,
+ "step": 14928
+ },
+ {
+ "epoch": 47.6964856230032,
+ "grad_norm": 0.123046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0261,
+ "step": 14929
+ },
+ {
+ "epoch": 47.69968051118211,
+ "grad_norm": 0.09033203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0328,
+ "step": 14930
+ },
+ {
+ "epoch": 47.70287539936102,
+ "grad_norm": 0.0986328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0292,
+ "step": 14931
+ },
+ {
+ "epoch": 47.70607028753994,
+ "grad_norm": 0.07861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0292,
+ "step": 14932
+ },
+ {
+ "epoch": 47.70926517571885,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0323,
+ "step": 14933
+ },
+ {
+ "epoch": 47.712460063897765,
+ "grad_norm": 0.057861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0191,
+ "step": 14934
+ },
+ {
+ "epoch": 47.715654952076676,
+ "grad_norm": 0.0908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0275,
+ "step": 14935
+ },
+ {
+ "epoch": 47.718849840255594,
+ "grad_norm": 0.06005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0217,
+ "step": 14936
+ },
+ {
+ "epoch": 47.722044728434504,
+ "grad_norm": 0.1328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0253,
+ "step": 14937
+ },
+ {
+ "epoch": 47.72523961661342,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0292,
+ "step": 14938
+ },
+ {
+ "epoch": 47.72843450479233,
+ "grad_norm": 0.08837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0287,
+ "step": 14939
+ },
+ {
+ "epoch": 47.73162939297124,
+ "grad_norm": 0.054443359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0166,
+ "step": 14940
+ },
+ {
+ "epoch": 47.73482428115016,
+ "grad_norm": 0.08056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0263,
+ "step": 14941
+ },
+ {
+ "epoch": 47.73801916932907,
+ "grad_norm": 0.057861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0252,
+ "step": 14942
+ },
+ {
+ "epoch": 47.74121405750799,
+ "grad_norm": 0.052001953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0275,
+ "step": 14943
+ },
+ {
+ "epoch": 47.7444089456869,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.033,
+ "step": 14944
+ },
+ {
+ "epoch": 47.74760383386582,
+ "grad_norm": 0.0498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0349,
+ "step": 14945
+ },
+ {
+ "epoch": 47.75079872204473,
+ "grad_norm": 0.08056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0321,
+ "step": 14946
+ },
+ {
+ "epoch": 47.753993610223645,
+ "grad_norm": 0.046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0278,
+ "step": 14947
+ },
+ {
+ "epoch": 47.757188498402556,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0302,
+ "step": 14948
+ },
+ {
+ "epoch": 47.760383386581466,
+ "grad_norm": 0.052001953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0386,
+ "step": 14949
+ },
+ {
+ "epoch": 47.763578274760384,
+ "grad_norm": 0.062255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0241,
+ "step": 14950
+ },
+ {
+ "epoch": 47.766773162939295,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0267,
+ "step": 14951
+ },
+ {
+ "epoch": 47.76996805111821,
+ "grad_norm": 0.055419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0276,
+ "step": 14952
+ },
+ {
+ "epoch": 47.77316293929712,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0241,
+ "step": 14953
+ },
+ {
+ "epoch": 47.77635782747604,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0258,
+ "step": 14954
+ },
+ {
+ "epoch": 47.77955271565495,
+ "grad_norm": 0.0498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0207,
+ "step": 14955
+ },
+ {
+ "epoch": 47.78274760383387,
+ "grad_norm": 0.048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.024,
+ "step": 14956
+ },
+ {
+ "epoch": 47.78594249201278,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0205,
+ "step": 14957
+ },
+ {
+ "epoch": 47.78913738019169,
+ "grad_norm": 0.060791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0212,
+ "step": 14958
+ },
+ {
+ "epoch": 47.79233226837061,
+ "grad_norm": 0.10498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0259,
+ "step": 14959
+ },
+ {
+ "epoch": 47.79552715654952,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0273,
+ "step": 14960
+ },
+ {
+ "epoch": 47.798722044728436,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0414,
+ "step": 14961
+ },
+ {
+ "epoch": 47.801916932907346,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0306,
+ "step": 14962
+ },
+ {
+ "epoch": 47.805111821086264,
+ "grad_norm": 0.08056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0337,
+ "step": 14963
+ },
+ {
+ "epoch": 47.808306709265175,
+ "grad_norm": 0.1083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0241,
+ "step": 14964
+ },
+ {
+ "epoch": 47.81150159744409,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0289,
+ "step": 14965
+ },
+ {
+ "epoch": 47.814696485623,
+ "grad_norm": 0.08447265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0287,
+ "step": 14966
+ },
+ {
+ "epoch": 47.81789137380191,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0317,
+ "step": 14967
+ },
+ {
+ "epoch": 47.82108626198083,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.028,
+ "step": 14968
+ },
+ {
+ "epoch": 47.82428115015974,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0421,
+ "step": 14969
+ },
+ {
+ "epoch": 47.82747603833866,
+ "grad_norm": 0.08154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0338,
+ "step": 14970
+ },
+ {
+ "epoch": 47.83067092651757,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0224,
+ "step": 14971
+ },
+ {
+ "epoch": 47.83386581469649,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0367,
+ "step": 14972
+ },
+ {
+ "epoch": 47.8370607028754,
+ "grad_norm": 0.0439453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0325,
+ "step": 14973
+ },
+ {
+ "epoch": 47.840255591054316,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0297,
+ "step": 14974
+ },
+ {
+ "epoch": 47.843450479233226,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0322,
+ "step": 14975
+ },
+ {
+ "epoch": 47.846645367412144,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0309,
+ "step": 14976
+ },
+ {
+ "epoch": 47.849840255591054,
+ "grad_norm": 0.0869140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0292,
+ "step": 14977
+ },
+ {
+ "epoch": 47.853035143769965,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0316,
+ "step": 14978
+ },
+ {
+ "epoch": 47.85623003194888,
+ "grad_norm": 0.08154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0338,
+ "step": 14979
+ },
+ {
+ "epoch": 47.85942492012779,
+ "grad_norm": 0.08447265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0209,
+ "step": 14980
+ },
+ {
+ "epoch": 47.86261980830671,
+ "grad_norm": 0.05859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0196,
+ "step": 14981
+ },
+ {
+ "epoch": 47.86581469648562,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0264,
+ "step": 14982
+ },
+ {
+ "epoch": 47.86900958466454,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0257,
+ "step": 14983
+ },
+ {
+ "epoch": 47.87220447284345,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0214,
+ "step": 14984
+ },
+ {
+ "epoch": 47.87539936102237,
+ "grad_norm": 0.0546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0138,
+ "step": 14985
+ },
+ {
+ "epoch": 47.87859424920128,
+ "grad_norm": 0.05517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0312,
+ "step": 14986
+ },
+ {
+ "epoch": 47.88178913738019,
+ "grad_norm": 0.049560546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0267,
+ "step": 14987
+ },
+ {
+ "epoch": 47.884984025559106,
+ "grad_norm": 0.055419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0358,
+ "step": 14988
+ },
+ {
+ "epoch": 47.88817891373802,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0306,
+ "step": 14989
+ },
+ {
+ "epoch": 47.891373801916934,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0263,
+ "step": 14990
+ },
+ {
+ "epoch": 47.894568690095845,
+ "grad_norm": 0.0966796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0206,
+ "step": 14991
+ },
+ {
+ "epoch": 47.89776357827476,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0198,
+ "step": 14992
+ },
+ {
+ "epoch": 47.90095846645367,
+ "grad_norm": 0.1005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0287,
+ "step": 14993
+ },
+ {
+ "epoch": 47.90415335463259,
+ "grad_norm": 0.05419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0336,
+ "step": 14994
+ },
+ {
+ "epoch": 47.9073482428115,
+ "grad_norm": 0.08642578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0275,
+ "step": 14995
+ },
+ {
+ "epoch": 47.91054313099041,
+ "grad_norm": 0.05859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0275,
+ "step": 14996
+ },
+ {
+ "epoch": 47.91373801916933,
+ "grad_norm": 0.091796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0225,
+ "step": 14997
+ },
+ {
+ "epoch": 47.91693290734824,
+ "grad_norm": 0.09716796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0435,
+ "step": 14998
+ },
+ {
+ "epoch": 47.92012779552716,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0286,
+ "step": 14999
+ },
+ {
+ "epoch": 47.92332268370607,
+ "grad_norm": 0.1337890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0364,
+ "step": 15000
+ },
+ {
+ "epoch": 47.926517571884986,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0286,
+ "step": 15001
+ },
+ {
+ "epoch": 47.9297124600639,
+ "grad_norm": 0.193359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0238,
+ "step": 15002
+ },
+ {
+ "epoch": 47.932907348242814,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0196,
+ "step": 15003
+ },
+ {
+ "epoch": 47.936102236421725,
+ "grad_norm": 0.169921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0301,
+ "step": 15004
+ },
+ {
+ "epoch": 47.93929712460064,
+ "grad_norm": 0.09521484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0228,
+ "step": 15005
+ },
+ {
+ "epoch": 47.94249201277955,
+ "grad_norm": 0.1328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0255,
+ "step": 15006
+ },
+ {
+ "epoch": 47.945686900958464,
+ "grad_norm": 0.08251953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0245,
+ "step": 15007
+ },
+ {
+ "epoch": 47.94888178913738,
+ "grad_norm": 0.126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0283,
+ "step": 15008
+ },
+ {
+ "epoch": 47.95207667731629,
+ "grad_norm": 0.061279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0314,
+ "step": 15009
+ },
+ {
+ "epoch": 47.95527156549521,
+ "grad_norm": 0.119140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0336,
+ "step": 15010
+ },
+ {
+ "epoch": 47.95846645367412,
+ "grad_norm": 0.06005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0299,
+ "step": 15011
+ },
+ {
+ "epoch": 47.96166134185304,
+ "grad_norm": 0.1171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0359,
+ "step": 15012
+ },
+ {
+ "epoch": 47.96485623003195,
+ "grad_norm": 0.0859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0275,
+ "step": 15013
+ },
+ {
+ "epoch": 47.968051118210866,
+ "grad_norm": 0.1318359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0325,
+ "step": 15014
+ },
+ {
+ "epoch": 47.97124600638978,
+ "grad_norm": 0.08251953125,
+ "learning_rate": 0.0005,
+ "loss": 1.026,
+ "step": 15015
+ },
+ {
+ "epoch": 47.97444089456869,
+ "grad_norm": 0.09912109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0198,
+ "step": 15016
+ },
+ {
+ "epoch": 47.977635782747605,
+ "grad_norm": 0.10107421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0234,
+ "step": 15017
+ },
+ {
+ "epoch": 47.980830670926515,
+ "grad_norm": 0.14453125,
+ "learning_rate": 0.0005,
+ "loss": 1.03,
+ "step": 15018
+ },
+ {
+ "epoch": 47.98402555910543,
+ "grad_norm": 0.04736328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0229,
+ "step": 15019
+ },
+ {
+ "epoch": 47.98722044728434,
+ "grad_norm": 0.1494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0222,
+ "step": 15020
+ },
+ {
+ "epoch": 47.99041533546326,
+ "grad_norm": 0.08935546875,
+ "learning_rate": 0.0005,
+ "loss": 1.017,
+ "step": 15021
+ },
+ {
+ "epoch": 47.99361022364217,
+ "grad_norm": 0.11962890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0299,
+ "step": 15022
+ },
+ {
+ "epoch": 47.99680511182109,
+ "grad_norm": 0.07861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.027,
+ "step": 15023
+ },
+ {
+ "epoch": 48.0,
+ "grad_norm": 0.11767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0336,
+ "step": 15024
+ },
+ {
+ "epoch": 48.00319488817891,
+ "grad_norm": 0.0927734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0263,
+ "step": 15025
+ },
+ {
+ "epoch": 48.00638977635783,
+ "grad_norm": 0.10595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0324,
+ "step": 15026
+ },
+ {
+ "epoch": 48.00958466453674,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0236,
+ "step": 15027
+ },
+ {
+ "epoch": 48.01277955271566,
+ "grad_norm": 0.1337890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0251,
+ "step": 15028
+ },
+ {
+ "epoch": 48.01597444089457,
+ "grad_norm": 0.056884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0242,
+ "step": 15029
+ },
+ {
+ "epoch": 48.019169329073485,
+ "grad_norm": 0.11279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0282,
+ "step": 15030
+ },
+ {
+ "epoch": 48.022364217252395,
+ "grad_norm": 0.05517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0278,
+ "step": 15031
+ },
+ {
+ "epoch": 48.02555910543131,
+ "grad_norm": 0.103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0278,
+ "step": 15032
+ },
+ {
+ "epoch": 48.02875399361022,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0276,
+ "step": 15033
+ },
+ {
+ "epoch": 48.031948881789134,
+ "grad_norm": 0.08251953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0233,
+ "step": 15034
+ },
+ {
+ "epoch": 48.03514376996805,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0354,
+ "step": 15035
+ },
+ {
+ "epoch": 48.03833865814696,
+ "grad_norm": 0.0947265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0313,
+ "step": 15036
+ },
+ {
+ "epoch": 48.04153354632588,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0211,
+ "step": 15037
+ },
+ {
+ "epoch": 48.04472843450479,
+ "grad_norm": 0.09033203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0279,
+ "step": 15038
+ },
+ {
+ "epoch": 48.04792332268371,
+ "grad_norm": 0.052734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0316,
+ "step": 15039
+ },
+ {
+ "epoch": 48.05111821086262,
+ "grad_norm": 0.08251953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0318,
+ "step": 15040
+ },
+ {
+ "epoch": 48.054313099041536,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0359,
+ "step": 15041
+ },
+ {
+ "epoch": 48.05750798722045,
+ "grad_norm": 0.11376953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0246,
+ "step": 15042
+ },
+ {
+ "epoch": 48.06070287539936,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0265,
+ "step": 15043
+ },
+ {
+ "epoch": 48.063897763578275,
+ "grad_norm": 0.083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0297,
+ "step": 15044
+ },
+ {
+ "epoch": 48.067092651757186,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0247,
+ "step": 15045
+ },
+ {
+ "epoch": 48.0702875399361,
+ "grad_norm": 0.0556640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0212,
+ "step": 15046
+ },
+ {
+ "epoch": 48.073482428115014,
+ "grad_norm": 0.05615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0264,
+ "step": 15047
+ },
+ {
+ "epoch": 48.07667731629393,
+ "grad_norm": 0.051025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0205,
+ "step": 15048
+ },
+ {
+ "epoch": 48.07987220447284,
+ "grad_norm": 0.050537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0341,
+ "step": 15049
+ },
+ {
+ "epoch": 48.08306709265176,
+ "grad_norm": 0.058349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0186,
+ "step": 15050
+ },
+ {
+ "epoch": 48.08626198083067,
+ "grad_norm": 0.04638671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0328,
+ "step": 15051
+ },
+ {
+ "epoch": 48.08945686900959,
+ "grad_norm": 0.0595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0323,
+ "step": 15052
+ },
+ {
+ "epoch": 48.0926517571885,
+ "grad_norm": 0.051025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0356,
+ "step": 15053
+ },
+ {
+ "epoch": 48.09584664536741,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0267,
+ "step": 15054
+ },
+ {
+ "epoch": 48.09904153354633,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0251,
+ "step": 15055
+ },
+ {
+ "epoch": 48.10223642172524,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0248,
+ "step": 15056
+ },
+ {
+ "epoch": 48.105431309904155,
+ "grad_norm": 0.05029296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0414,
+ "step": 15057
+ },
+ {
+ "epoch": 48.108626198083066,
+ "grad_norm": 0.053466796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0277,
+ "step": 15058
+ },
+ {
+ "epoch": 48.11182108626198,
+ "grad_norm": 0.0478515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0378,
+ "step": 15059
+ },
+ {
+ "epoch": 48.115015974440894,
+ "grad_norm": 0.045166015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0349,
+ "step": 15060
+ },
+ {
+ "epoch": 48.11821086261981,
+ "grad_norm": 0.060302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0239,
+ "step": 15061
+ },
+ {
+ "epoch": 48.12140575079872,
+ "grad_norm": 0.044189453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0295,
+ "step": 15062
+ },
+ {
+ "epoch": 48.12460063897763,
+ "grad_norm": 0.09765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0157,
+ "step": 15063
+ },
+ {
+ "epoch": 48.12779552715655,
+ "grad_norm": 0.048095703125,
+ "learning_rate": 0.0005,
+ "loss": 1.03,
+ "step": 15064
+ },
+ {
+ "epoch": 48.13099041533546,
+ "grad_norm": 0.10546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0277,
+ "step": 15065
+ },
+ {
+ "epoch": 48.13418530351438,
+ "grad_norm": 0.062255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0253,
+ "step": 15066
+ },
+ {
+ "epoch": 48.13738019169329,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0253,
+ "step": 15067
+ },
+ {
+ "epoch": 48.14057507987221,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0282,
+ "step": 15068
+ },
+ {
+ "epoch": 48.14376996805112,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0287,
+ "step": 15069
+ },
+ {
+ "epoch": 48.146964856230035,
+ "grad_norm": 0.04736328125,
+ "learning_rate": 0.0005,
+ "loss": 1.025,
+ "step": 15070
+ },
+ {
+ "epoch": 48.150159744408946,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0321,
+ "step": 15071
+ },
+ {
+ "epoch": 48.153354632587856,
+ "grad_norm": 0.056396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0265,
+ "step": 15072
+ },
+ {
+ "epoch": 48.156549520766774,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.036,
+ "step": 15073
+ },
+ {
+ "epoch": 48.159744408945684,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0417,
+ "step": 15074
+ },
+ {
+ "epoch": 48.1629392971246,
+ "grad_norm": 0.08544921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0332,
+ "step": 15075
+ },
+ {
+ "epoch": 48.16613418530351,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.026,
+ "step": 15076
+ },
+ {
+ "epoch": 48.16932907348243,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0126,
+ "step": 15077
+ },
+ {
+ "epoch": 48.17252396166134,
+ "grad_norm": 0.080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0158,
+ "step": 15078
+ },
+ {
+ "epoch": 48.17571884984026,
+ "grad_norm": 0.0830078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0262,
+ "step": 15079
+ },
+ {
+ "epoch": 48.17891373801917,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0366,
+ "step": 15080
+ },
+ {
+ "epoch": 48.18210862619808,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0222,
+ "step": 15081
+ },
+ {
+ "epoch": 48.185303514377,
+ "grad_norm": 0.095703125,
+ "learning_rate": 0.0005,
+ "loss": 1.026,
+ "step": 15082
+ },
+ {
+ "epoch": 48.18849840255591,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0212,
+ "step": 15083
+ },
+ {
+ "epoch": 48.191693290734825,
+ "grad_norm": 0.08056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0286,
+ "step": 15084
+ },
+ {
+ "epoch": 48.194888178913736,
+ "grad_norm": 0.044189453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0294,
+ "step": 15085
+ },
+ {
+ "epoch": 48.198083067092654,
+ "grad_norm": 0.0517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0335,
+ "step": 15086
+ },
+ {
+ "epoch": 48.201277955271564,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.031,
+ "step": 15087
+ },
+ {
+ "epoch": 48.20447284345048,
+ "grad_norm": 0.0830078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0337,
+ "step": 15088
+ },
+ {
+ "epoch": 48.20766773162939,
+ "grad_norm": 0.103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0204,
+ "step": 15089
+ },
+ {
+ "epoch": 48.21086261980831,
+ "grad_norm": 0.126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0251,
+ "step": 15090
+ },
+ {
+ "epoch": 48.21405750798722,
+ "grad_norm": 0.0869140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0372,
+ "step": 15091
+ },
+ {
+ "epoch": 48.21725239616613,
+ "grad_norm": 0.1513671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0259,
+ "step": 15092
+ },
+ {
+ "epoch": 48.22044728434505,
+ "grad_norm": 0.059814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0243,
+ "step": 15093
+ },
+ {
+ "epoch": 48.22364217252396,
+ "grad_norm": 0.1416015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0261,
+ "step": 15094
+ },
+ {
+ "epoch": 48.22683706070288,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0143,
+ "step": 15095
+ },
+ {
+ "epoch": 48.23003194888179,
+ "grad_norm": 0.12451171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0295,
+ "step": 15096
+ },
+ {
+ "epoch": 48.233226837060705,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0335,
+ "step": 15097
+ },
+ {
+ "epoch": 48.236421725239616,
+ "grad_norm": 0.10693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0195,
+ "step": 15098
+ },
+ {
+ "epoch": 48.239616613418534,
+ "grad_norm": 0.09033203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0211,
+ "step": 15099
+ },
+ {
+ "epoch": 48.242811501597444,
+ "grad_norm": 0.12353515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0298,
+ "step": 15100
+ },
+ {
+ "epoch": 48.246006389776355,
+ "grad_norm": 0.099609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0311,
+ "step": 15101
+ },
+ {
+ "epoch": 48.24920127795527,
+ "grad_norm": 0.171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0272,
+ "step": 15102
+ },
+ {
+ "epoch": 48.25239616613418,
+ "grad_norm": 0.0830078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0478,
+ "step": 15103
+ },
+ {
+ "epoch": 48.2555910543131,
+ "grad_norm": 0.244140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0394,
+ "step": 15104
+ },
+ {
+ "epoch": 48.25878594249201,
+ "grad_norm": 0.091796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0255,
+ "step": 15105
+ },
+ {
+ "epoch": 48.26198083067093,
+ "grad_norm": 0.126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0209,
+ "step": 15106
+ },
+ {
+ "epoch": 48.26517571884984,
+ "grad_norm": 0.08642578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0271,
+ "step": 15107
+ },
+ {
+ "epoch": 48.26837060702876,
+ "grad_norm": 0.125,
+ "learning_rate": 0.0005,
+ "loss": 1.0277,
+ "step": 15108
+ },
+ {
+ "epoch": 48.27156549520767,
+ "grad_norm": 0.158203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0243,
+ "step": 15109
+ },
+ {
+ "epoch": 48.27476038338658,
+ "grad_norm": 0.0849609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0271,
+ "step": 15110
+ },
+ {
+ "epoch": 48.277955271565496,
+ "grad_norm": 0.154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0366,
+ "step": 15111
+ },
+ {
+ "epoch": 48.281150159744406,
+ "grad_norm": 0.08203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0318,
+ "step": 15112
+ },
+ {
+ "epoch": 48.284345047923324,
+ "grad_norm": 0.1572265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0214,
+ "step": 15113
+ },
+ {
+ "epoch": 48.287539936102235,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0187,
+ "step": 15114
+ },
+ {
+ "epoch": 48.29073482428115,
+ "grad_norm": 0.1328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0364,
+ "step": 15115
+ },
+ {
+ "epoch": 48.29392971246006,
+ "grad_norm": 0.0986328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0233,
+ "step": 15116
+ },
+ {
+ "epoch": 48.29712460063898,
+ "grad_norm": 0.17578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0361,
+ "step": 15117
+ },
+ {
+ "epoch": 48.30031948881789,
+ "grad_norm": 0.04736328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0298,
+ "step": 15118
+ },
+ {
+ "epoch": 48.3035143769968,
+ "grad_norm": 0.1943359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0276,
+ "step": 15119
+ },
+ {
+ "epoch": 48.30670926517572,
+ "grad_norm": 0.09814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0239,
+ "step": 15120
+ },
+ {
+ "epoch": 48.30990415335463,
+ "grad_norm": 0.0986328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0275,
+ "step": 15121
+ },
+ {
+ "epoch": 48.31309904153355,
+ "grad_norm": 0.061767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0176,
+ "step": 15122
+ },
+ {
+ "epoch": 48.31629392971246,
+ "grad_norm": 0.09326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.031,
+ "step": 15123
+ },
+ {
+ "epoch": 48.319488817891376,
+ "grad_norm": 0.050048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0195,
+ "step": 15124
+ },
+ {
+ "epoch": 48.322683706070286,
+ "grad_norm": 0.09814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0201,
+ "step": 15125
+ },
+ {
+ "epoch": 48.325878594249204,
+ "grad_norm": 0.05517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0274,
+ "step": 15126
+ },
+ {
+ "epoch": 48.329073482428115,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.014,
+ "step": 15127
+ },
+ {
+ "epoch": 48.33226837060703,
+ "grad_norm": 0.051513671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0227,
+ "step": 15128
+ },
+ {
+ "epoch": 48.33546325878594,
+ "grad_norm": 0.056396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0248,
+ "step": 15129
+ },
+ {
+ "epoch": 48.33865814696485,
+ "grad_norm": 0.05615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0341,
+ "step": 15130
+ },
+ {
+ "epoch": 48.34185303514377,
+ "grad_norm": 0.0478515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0244,
+ "step": 15131
+ },
+ {
+ "epoch": 48.34504792332268,
+ "grad_norm": 0.0537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0207,
+ "step": 15132
+ },
+ {
+ "epoch": 48.3482428115016,
+ "grad_norm": 0.053466796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0347,
+ "step": 15133
+ },
+ {
+ "epoch": 48.35143769968051,
+ "grad_norm": 0.05712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0227,
+ "step": 15134
+ },
+ {
+ "epoch": 48.35463258785943,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0313,
+ "step": 15135
+ },
+ {
+ "epoch": 48.35782747603834,
+ "grad_norm": 0.051025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0243,
+ "step": 15136
+ },
+ {
+ "epoch": 48.361022364217256,
+ "grad_norm": 0.052978515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0319,
+ "step": 15137
+ },
+ {
+ "epoch": 48.364217252396166,
+ "grad_norm": 0.0576171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0303,
+ "step": 15138
+ },
+ {
+ "epoch": 48.36741214057508,
+ "grad_norm": 0.057373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0302,
+ "step": 15139
+ },
+ {
+ "epoch": 48.370607028753994,
+ "grad_norm": 0.047119140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0353,
+ "step": 15140
+ },
+ {
+ "epoch": 48.373801916932905,
+ "grad_norm": 0.0927734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0205,
+ "step": 15141
+ },
+ {
+ "epoch": 48.37699680511182,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0224,
+ "step": 15142
+ },
+ {
+ "epoch": 48.38019169329073,
+ "grad_norm": 0.060791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0308,
+ "step": 15143
+ },
+ {
+ "epoch": 48.38338658146965,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0324,
+ "step": 15144
+ },
+ {
+ "epoch": 48.38658146964856,
+ "grad_norm": 0.048095703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0387,
+ "step": 15145
+ },
+ {
+ "epoch": 48.38977635782748,
+ "grad_norm": 0.052978515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0312,
+ "step": 15146
+ },
+ {
+ "epoch": 48.39297124600639,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0426,
+ "step": 15147
+ },
+ {
+ "epoch": 48.3961661341853,
+ "grad_norm": 0.0478515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0305,
+ "step": 15148
+ },
+ {
+ "epoch": 48.39936102236422,
+ "grad_norm": 0.08154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0245,
+ "step": 15149
+ },
+ {
+ "epoch": 48.40255591054313,
+ "grad_norm": 0.052978515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0391,
+ "step": 15150
+ },
+ {
+ "epoch": 48.405750798722046,
+ "grad_norm": 0.091796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0267,
+ "step": 15151
+ },
+ {
+ "epoch": 48.40894568690096,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0323,
+ "step": 15152
+ },
+ {
+ "epoch": 48.412140575079874,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0234,
+ "step": 15153
+ },
+ {
+ "epoch": 48.415335463258785,
+ "grad_norm": 0.046142578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0274,
+ "step": 15154
+ },
+ {
+ "epoch": 48.4185303514377,
+ "grad_norm": 0.0947265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0286,
+ "step": 15155
+ },
+ {
+ "epoch": 48.42172523961661,
+ "grad_norm": 0.0927734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0262,
+ "step": 15156
+ },
+ {
+ "epoch": 48.424920127795524,
+ "grad_norm": 0.09814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0309,
+ "step": 15157
+ },
+ {
+ "epoch": 48.42811501597444,
+ "grad_norm": 0.09619140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0189,
+ "step": 15158
+ },
+ {
+ "epoch": 48.43130990415335,
+ "grad_norm": 0.091796875,
+ "learning_rate": 0.0005,
+ "loss": 1.025,
+ "step": 15159
+ },
+ {
+ "epoch": 48.43450479233227,
+ "grad_norm": 0.056884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0366,
+ "step": 15160
+ },
+ {
+ "epoch": 48.43769968051118,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.038,
+ "step": 15161
+ },
+ {
+ "epoch": 48.4408945686901,
+ "grad_norm": 0.054931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0425,
+ "step": 15162
+ },
+ {
+ "epoch": 48.44408945686901,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0274,
+ "step": 15163
+ },
+ {
+ "epoch": 48.447284345047926,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0285,
+ "step": 15164
+ },
+ {
+ "epoch": 48.45047923322684,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0195,
+ "step": 15165
+ },
+ {
+ "epoch": 48.453674121405754,
+ "grad_norm": 0.11962890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0313,
+ "step": 15166
+ },
+ {
+ "epoch": 48.456869009584665,
+ "grad_norm": 0.1201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0405,
+ "step": 15167
+ },
+ {
+ "epoch": 48.460063897763575,
+ "grad_norm": 0.0908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0325,
+ "step": 15168
+ },
+ {
+ "epoch": 48.46325878594249,
+ "grad_norm": 0.1220703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0367,
+ "step": 15169
+ },
+ {
+ "epoch": 48.466453674121404,
+ "grad_norm": 0.1181640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0183,
+ "step": 15170
+ },
+ {
+ "epoch": 48.46964856230032,
+ "grad_norm": 0.1533203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0263,
+ "step": 15171
+ },
+ {
+ "epoch": 48.47284345047923,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0244,
+ "step": 15172
+ },
+ {
+ "epoch": 48.47603833865815,
+ "grad_norm": 0.09619140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0271,
+ "step": 15173
+ },
+ {
+ "epoch": 48.47923322683706,
+ "grad_norm": 0.057861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0218,
+ "step": 15174
+ },
+ {
+ "epoch": 48.48242811501598,
+ "grad_norm": 0.080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.031,
+ "step": 15175
+ },
+ {
+ "epoch": 48.48562300319489,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0305,
+ "step": 15176
+ },
+ {
+ "epoch": 48.4888178913738,
+ "grad_norm": 0.0947265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0187,
+ "step": 15177
+ },
+ {
+ "epoch": 48.49201277955272,
+ "grad_norm": 0.09375,
+ "learning_rate": 0.0005,
+ "loss": 1.0246,
+ "step": 15178
+ },
+ {
+ "epoch": 48.49520766773163,
+ "grad_norm": 0.060302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0214,
+ "step": 15179
+ },
+ {
+ "epoch": 48.498402555910545,
+ "grad_norm": 0.04736328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0282,
+ "step": 15180
+ },
+ {
+ "epoch": 48.501597444089455,
+ "grad_norm": 0.050537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0324,
+ "step": 15181
+ },
+ {
+ "epoch": 48.50479233226837,
+ "grad_norm": 0.10302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0325,
+ "step": 15182
+ },
+ {
+ "epoch": 48.50798722044728,
+ "grad_norm": 0.05810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0316,
+ "step": 15183
+ },
+ {
+ "epoch": 48.5111821086262,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0314,
+ "step": 15184
+ },
+ {
+ "epoch": 48.51437699680511,
+ "grad_norm": 0.056884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0258,
+ "step": 15185
+ },
+ {
+ "epoch": 48.51757188498402,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.027,
+ "step": 15186
+ },
+ {
+ "epoch": 48.52076677316294,
+ "grad_norm": 0.05224609375,
+ "learning_rate": 0.0005,
+ "loss": 1.026,
+ "step": 15187
+ },
+ {
+ "epoch": 48.52396166134185,
+ "grad_norm": 0.05712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0332,
+ "step": 15188
+ },
+ {
+ "epoch": 48.52715654952077,
+ "grad_norm": 0.062255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0348,
+ "step": 15189
+ },
+ {
+ "epoch": 48.53035143769968,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0348,
+ "step": 15190
+ },
+ {
+ "epoch": 48.533546325878596,
+ "grad_norm": 0.091796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0348,
+ "step": 15191
+ },
+ {
+ "epoch": 48.53674121405751,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.028,
+ "step": 15192
+ },
+ {
+ "epoch": 48.539936102236425,
+ "grad_norm": 0.09326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0219,
+ "step": 15193
+ },
+ {
+ "epoch": 48.543130990415335,
+ "grad_norm": 0.055419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0362,
+ "step": 15194
+ },
+ {
+ "epoch": 48.546325878594246,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0275,
+ "step": 15195
+ },
+ {
+ "epoch": 48.54952076677316,
+ "grad_norm": 0.054931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0277,
+ "step": 15196
+ },
+ {
+ "epoch": 48.552715654952074,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0254,
+ "step": 15197
+ },
+ {
+ "epoch": 48.55591054313099,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0268,
+ "step": 15198
+ },
+ {
+ "epoch": 48.5591054313099,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0306,
+ "step": 15199
+ },
+ {
+ "epoch": 48.56230031948882,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0168,
+ "step": 15200
+ },
+ {
+ "epoch": 48.56549520766773,
+ "grad_norm": 0.08056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0211,
+ "step": 15201
+ },
+ {
+ "epoch": 48.56869009584665,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0338,
+ "step": 15202
+ },
+ {
+ "epoch": 48.57188498402556,
+ "grad_norm": 0.08251953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0343,
+ "step": 15203
+ },
+ {
+ "epoch": 48.575079872204476,
+ "grad_norm": 0.08642578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0408,
+ "step": 15204
+ },
+ {
+ "epoch": 48.57827476038339,
+ "grad_norm": 0.1650390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0325,
+ "step": 15205
+ },
+ {
+ "epoch": 48.5814696485623,
+ "grad_norm": 0.049560546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0304,
+ "step": 15206
+ },
+ {
+ "epoch": 48.584664536741215,
+ "grad_norm": 0.1796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0264,
+ "step": 15207
+ },
+ {
+ "epoch": 48.587859424920126,
+ "grad_norm": 0.1005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.031,
+ "step": 15208
+ },
+ {
+ "epoch": 48.59105431309904,
+ "grad_norm": 0.24609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0241,
+ "step": 15209
+ },
+ {
+ "epoch": 48.594249201277954,
+ "grad_norm": 0.11181640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0322,
+ "step": 15210
+ },
+ {
+ "epoch": 48.59744408945687,
+ "grad_norm": 0.224609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0403,
+ "step": 15211
+ },
+ {
+ "epoch": 48.60063897763578,
+ "grad_norm": 0.1923828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0197,
+ "step": 15212
+ },
+ {
+ "epoch": 48.6038338658147,
+ "grad_norm": 0.07861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.037,
+ "step": 15213
+ },
+ {
+ "epoch": 48.60702875399361,
+ "grad_norm": 0.228515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0322,
+ "step": 15214
+ },
+ {
+ "epoch": 48.61022364217252,
+ "grad_norm": 0.083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0285,
+ "step": 15215
+ },
+ {
+ "epoch": 48.61341853035144,
+ "grad_norm": 0.193359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0417,
+ "step": 15216
+ },
+ {
+ "epoch": 48.61661341853035,
+ "grad_norm": 0.208984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0356,
+ "step": 15217
+ },
+ {
+ "epoch": 48.61980830670927,
+ "grad_norm": 0.1201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0238,
+ "step": 15218
+ },
+ {
+ "epoch": 48.62300319488818,
+ "grad_norm": 0.26953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0345,
+ "step": 15219
+ },
+ {
+ "epoch": 48.626198083067095,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0312,
+ "step": 15220
+ },
+ {
+ "epoch": 48.629392971246006,
+ "grad_norm": 0.1845703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0218,
+ "step": 15221
+ },
+ {
+ "epoch": 48.63258785942492,
+ "grad_norm": 0.1044921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0365,
+ "step": 15222
+ },
+ {
+ "epoch": 48.635782747603834,
+ "grad_norm": 0.1923828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0266,
+ "step": 15223
+ },
+ {
+ "epoch": 48.638977635782744,
+ "grad_norm": 0.1015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0341,
+ "step": 15224
+ },
+ {
+ "epoch": 48.64217252396166,
+ "grad_norm": 0.134765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0249,
+ "step": 15225
+ },
+ {
+ "epoch": 48.64536741214057,
+ "grad_norm": 0.1240234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0429,
+ "step": 15226
+ },
+ {
+ "epoch": 48.64856230031949,
+ "grad_norm": 0.0986328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0255,
+ "step": 15227
+ },
+ {
+ "epoch": 48.6517571884984,
+ "grad_norm": 0.1083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0163,
+ "step": 15228
+ },
+ {
+ "epoch": 48.65495207667732,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0223,
+ "step": 15229
+ },
+ {
+ "epoch": 48.65814696485623,
+ "grad_norm": 0.1064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0309,
+ "step": 15230
+ },
+ {
+ "epoch": 48.66134185303515,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0309,
+ "step": 15231
+ },
+ {
+ "epoch": 48.66453674121406,
+ "grad_norm": 0.115234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0259,
+ "step": 15232
+ },
+ {
+ "epoch": 48.66773162939297,
+ "grad_norm": 0.08349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0245,
+ "step": 15233
+ },
+ {
+ "epoch": 48.670926517571885,
+ "grad_norm": 0.107421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0391,
+ "step": 15234
+ },
+ {
+ "epoch": 48.674121405750796,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0247,
+ "step": 15235
+ },
+ {
+ "epoch": 48.677316293929714,
+ "grad_norm": 0.08447265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0296,
+ "step": 15236
+ },
+ {
+ "epoch": 48.680511182108624,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0342,
+ "step": 15237
+ },
+ {
+ "epoch": 48.68370607028754,
+ "grad_norm": 0.06103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0301,
+ "step": 15238
+ },
+ {
+ "epoch": 48.68690095846645,
+ "grad_norm": 0.09326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0493,
+ "step": 15239
+ },
+ {
+ "epoch": 48.69009584664537,
+ "grad_norm": 0.0615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0299,
+ "step": 15240
+ },
+ {
+ "epoch": 48.69329073482428,
+ "grad_norm": 0.1064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0278,
+ "step": 15241
+ },
+ {
+ "epoch": 48.6964856230032,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.027,
+ "step": 15242
+ },
+ {
+ "epoch": 48.69968051118211,
+ "grad_norm": 0.1416015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0292,
+ "step": 15243
+ },
+ {
+ "epoch": 48.70287539936102,
+ "grad_norm": 0.1025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0267,
+ "step": 15244
+ },
+ {
+ "epoch": 48.70607028753994,
+ "grad_norm": 0.11865234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0384,
+ "step": 15245
+ },
+ {
+ "epoch": 48.70926517571885,
+ "grad_norm": 0.1650390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0304,
+ "step": 15246
+ },
+ {
+ "epoch": 48.712460063897765,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0278,
+ "step": 15247
+ },
+ {
+ "epoch": 48.715654952076676,
+ "grad_norm": 0.138671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0371,
+ "step": 15248
+ },
+ {
+ "epoch": 48.718849840255594,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0289,
+ "step": 15249
+ },
+ {
+ "epoch": 48.722044728434504,
+ "grad_norm": 0.08544921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0241,
+ "step": 15250
+ },
+ {
+ "epoch": 48.72523961661342,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0235,
+ "step": 15251
+ },
+ {
+ "epoch": 48.72843450479233,
+ "grad_norm": 0.08203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0235,
+ "step": 15252
+ },
+ {
+ "epoch": 48.73162939297124,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0228,
+ "step": 15253
+ },
+ {
+ "epoch": 48.73482428115016,
+ "grad_norm": 0.062255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0211,
+ "step": 15254
+ },
+ {
+ "epoch": 48.73801916932907,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0335,
+ "step": 15255
+ },
+ {
+ "epoch": 48.74121405750799,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0227,
+ "step": 15256
+ },
+ {
+ "epoch": 48.7444089456869,
+ "grad_norm": 0.08056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0223,
+ "step": 15257
+ },
+ {
+ "epoch": 48.74760383386582,
+ "grad_norm": 0.0517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0306,
+ "step": 15258
+ },
+ {
+ "epoch": 48.75079872204473,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0194,
+ "step": 15259
+ },
+ {
+ "epoch": 48.753993610223645,
+ "grad_norm": 0.061767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.034,
+ "step": 15260
+ },
+ {
+ "epoch": 48.757188498402556,
+ "grad_norm": 0.057373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0303,
+ "step": 15261
+ },
+ {
+ "epoch": 48.760383386581466,
+ "grad_norm": 0.056396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0404,
+ "step": 15262
+ },
+ {
+ "epoch": 48.763578274760384,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0279,
+ "step": 15263
+ },
+ {
+ "epoch": 48.766773162939295,
+ "grad_norm": 0.05126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0262,
+ "step": 15264
+ },
+ {
+ "epoch": 48.76996805111821,
+ "grad_norm": 0.054931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0237,
+ "step": 15265
+ },
+ {
+ "epoch": 48.77316293929712,
+ "grad_norm": 0.09423828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0284,
+ "step": 15266
+ },
+ {
+ "epoch": 48.77635782747604,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0409,
+ "step": 15267
+ },
+ {
+ "epoch": 48.77955271565495,
+ "grad_norm": 0.062255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0162,
+ "step": 15268
+ },
+ {
+ "epoch": 48.78274760383387,
+ "grad_norm": 0.0556640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0291,
+ "step": 15269
+ },
+ {
+ "epoch": 48.78594249201278,
+ "grad_norm": 0.060791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0285,
+ "step": 15270
+ },
+ {
+ "epoch": 48.78913738019169,
+ "grad_norm": 0.056396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0348,
+ "step": 15271
+ },
+ {
+ "epoch": 48.79233226837061,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0304,
+ "step": 15272
+ },
+ {
+ "epoch": 48.79552715654952,
+ "grad_norm": 0.051513671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0352,
+ "step": 15273
+ },
+ {
+ "epoch": 48.798722044728436,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0273,
+ "step": 15274
+ },
+ {
+ "epoch": 48.801916932907346,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.03,
+ "step": 15275
+ },
+ {
+ "epoch": 48.805111821086264,
+ "grad_norm": 0.053466796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0186,
+ "step": 15276
+ },
+ {
+ "epoch": 48.808306709265175,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0318,
+ "step": 15277
+ },
+ {
+ "epoch": 48.81150159744409,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0192,
+ "step": 15278
+ },
+ {
+ "epoch": 48.814696485623,
+ "grad_norm": 0.052001953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0256,
+ "step": 15279
+ },
+ {
+ "epoch": 48.81789137380191,
+ "grad_norm": 0.056396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0287,
+ "step": 15280
+ },
+ {
+ "epoch": 48.82108626198083,
+ "grad_norm": 0.08349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0294,
+ "step": 15281
+ },
+ {
+ "epoch": 48.82428115015974,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0321,
+ "step": 15282
+ },
+ {
+ "epoch": 48.82747603833866,
+ "grad_norm": 0.062255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0309,
+ "step": 15283
+ },
+ {
+ "epoch": 48.83067092651757,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0317,
+ "step": 15284
+ },
+ {
+ "epoch": 48.83386581469649,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0302,
+ "step": 15285
+ },
+ {
+ "epoch": 48.8370607028754,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0281,
+ "step": 15286
+ },
+ {
+ "epoch": 48.840255591054316,
+ "grad_norm": 0.08056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0381,
+ "step": 15287
+ },
+ {
+ "epoch": 48.843450479233226,
+ "grad_norm": 0.050537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.035,
+ "step": 15288
+ },
+ {
+ "epoch": 48.846645367412144,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0335,
+ "step": 15289
+ },
+ {
+ "epoch": 48.849840255591054,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0385,
+ "step": 15290
+ },
+ {
+ "epoch": 48.853035143769965,
+ "grad_norm": 0.0576171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0317,
+ "step": 15291
+ },
+ {
+ "epoch": 48.85623003194888,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0281,
+ "step": 15292
+ },
+ {
+ "epoch": 48.85942492012779,
+ "grad_norm": 0.058837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0227,
+ "step": 15293
+ },
+ {
+ "epoch": 48.86261980830671,
+ "grad_norm": 0.050048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0273,
+ "step": 15294
+ },
+ {
+ "epoch": 48.86581469648562,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0223,
+ "step": 15295
+ },
+ {
+ "epoch": 48.86900958466454,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.046,
+ "step": 15296
+ },
+ {
+ "epoch": 48.87220447284345,
+ "grad_norm": 0.0517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0156,
+ "step": 15297
+ },
+ {
+ "epoch": 48.87539936102237,
+ "grad_norm": 0.10205078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0166,
+ "step": 15298
+ },
+ {
+ "epoch": 48.87859424920128,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0247,
+ "step": 15299
+ },
+ {
+ "epoch": 48.88178913738019,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0275,
+ "step": 15300
+ },
+ {
+ "epoch": 48.884984025559106,
+ "grad_norm": 0.051513671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0198,
+ "step": 15301
+ },
+ {
+ "epoch": 48.88817891373802,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0291,
+ "step": 15302
+ },
+ {
+ "epoch": 48.891373801916934,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0308,
+ "step": 15303
+ },
+ {
+ "epoch": 48.894568690095845,
+ "grad_norm": 0.09228515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0309,
+ "step": 15304
+ },
+ {
+ "epoch": 48.89776357827476,
+ "grad_norm": 0.0830078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0305,
+ "step": 15305
+ },
+ {
+ "epoch": 48.90095846645367,
+ "grad_norm": 0.15234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0355,
+ "step": 15306
+ },
+ {
+ "epoch": 48.90415335463259,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0251,
+ "step": 15307
+ },
+ {
+ "epoch": 48.9073482428115,
+ "grad_norm": 0.119140625,
+ "learning_rate": 0.0005,
+ "loss": 1.03,
+ "step": 15308
+ },
+ {
+ "epoch": 48.91054313099041,
+ "grad_norm": 0.08447265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0394,
+ "step": 15309
+ },
+ {
+ "epoch": 48.91373801916933,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0304,
+ "step": 15310
+ },
+ {
+ "epoch": 48.91693290734824,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0187,
+ "step": 15311
+ },
+ {
+ "epoch": 48.92012779552716,
+ "grad_norm": 0.08447265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0326,
+ "step": 15312
+ },
+ {
+ "epoch": 48.92332268370607,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0334,
+ "step": 15313
+ },
+ {
+ "epoch": 48.926517571884986,
+ "grad_norm": 0.087890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0143,
+ "step": 15314
+ },
+ {
+ "epoch": 48.9297124600639,
+ "grad_norm": 0.08642578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0329,
+ "step": 15315
+ },
+ {
+ "epoch": 48.932907348242814,
+ "grad_norm": 0.08544921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0289,
+ "step": 15316
+ },
+ {
+ "epoch": 48.936102236421725,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0215,
+ "step": 15317
+ },
+ {
+ "epoch": 48.93929712460064,
+ "grad_norm": 0.095703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0309,
+ "step": 15318
+ },
+ {
+ "epoch": 48.94249201277955,
+ "grad_norm": 0.0810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0302,
+ "step": 15319
+ },
+ {
+ "epoch": 48.945686900958464,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0362,
+ "step": 15320
+ },
+ {
+ "epoch": 48.94888178913738,
+ "grad_norm": 0.10986328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0338,
+ "step": 15321
+ },
+ {
+ "epoch": 48.95207667731629,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0264,
+ "step": 15322
+ },
+ {
+ "epoch": 48.95527156549521,
+ "grad_norm": 0.1015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0275,
+ "step": 15323
+ },
+ {
+ "epoch": 48.95846645367412,
+ "grad_norm": 0.09423828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0276,
+ "step": 15324
+ },
+ {
+ "epoch": 48.96166134185304,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0325,
+ "step": 15325
+ },
+ {
+ "epoch": 48.96485623003195,
+ "grad_norm": 0.0810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0212,
+ "step": 15326
+ },
+ {
+ "epoch": 48.968051118210866,
+ "grad_norm": 0.0615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0218,
+ "step": 15327
+ },
+ {
+ "epoch": 48.97124600638978,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0332,
+ "step": 15328
+ },
+ {
+ "epoch": 48.97444089456869,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0328,
+ "step": 15329
+ },
+ {
+ "epoch": 48.977635782747605,
+ "grad_norm": 0.0927734375,
+ "learning_rate": 0.0005,
+ "loss": 1.03,
+ "step": 15330
+ },
+ {
+ "epoch": 48.980830670926515,
+ "grad_norm": 0.056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0292,
+ "step": 15331
+ },
+ {
+ "epoch": 48.98402555910543,
+ "grad_norm": 0.057861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0313,
+ "step": 15332
+ },
+ {
+ "epoch": 48.98722044728434,
+ "grad_norm": 0.08056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0213,
+ "step": 15333
+ },
+ {
+ "epoch": 48.99041533546326,
+ "grad_norm": 0.06103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.026,
+ "step": 15334
+ },
+ {
+ "epoch": 48.99361022364217,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0258,
+ "step": 15335
+ },
+ {
+ "epoch": 48.99680511182109,
+ "grad_norm": 0.05908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.029,
+ "step": 15336
+ },
+ {
+ "epoch": 49.0,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0315,
+ "step": 15337
+ },
+ {
+ "epoch": 49.00319488817891,
+ "grad_norm": 0.05908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0305,
+ "step": 15338
+ },
+ {
+ "epoch": 49.00638977635783,
+ "grad_norm": 0.058837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0155,
+ "step": 15339
+ },
+ {
+ "epoch": 49.00958466453674,
+ "grad_norm": 0.080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0235,
+ "step": 15340
+ },
+ {
+ "epoch": 49.01277955271566,
+ "grad_norm": 0.05517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0271,
+ "step": 15341
+ },
+ {
+ "epoch": 49.01597444089457,
+ "grad_norm": 0.053955078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0162,
+ "step": 15342
+ },
+ {
+ "epoch": 49.019169329073485,
+ "grad_norm": 0.04541015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0279,
+ "step": 15343
+ },
+ {
+ "epoch": 49.022364217252395,
+ "grad_norm": 0.05517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0228,
+ "step": 15344
+ },
+ {
+ "epoch": 49.02555910543131,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0252,
+ "step": 15345
+ },
+ {
+ "epoch": 49.02875399361022,
+ "grad_norm": 0.083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0229,
+ "step": 15346
+ },
+ {
+ "epoch": 49.031948881789134,
+ "grad_norm": 0.053466796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0275,
+ "step": 15347
+ },
+ {
+ "epoch": 49.03514376996805,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0265,
+ "step": 15348
+ },
+ {
+ "epoch": 49.03833865814696,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0216,
+ "step": 15349
+ },
+ {
+ "epoch": 49.04153354632588,
+ "grad_norm": 0.080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0292,
+ "step": 15350
+ },
+ {
+ "epoch": 49.04472843450479,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0193,
+ "step": 15351
+ },
+ {
+ "epoch": 49.04792332268371,
+ "grad_norm": 0.09521484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0264,
+ "step": 15352
+ },
+ {
+ "epoch": 49.05111821086262,
+ "grad_norm": 0.08251953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0294,
+ "step": 15353
+ },
+ {
+ "epoch": 49.054313099041536,
+ "grad_norm": 0.091796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0361,
+ "step": 15354
+ },
+ {
+ "epoch": 49.05750798722045,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0338,
+ "step": 15355
+ },
+ {
+ "epoch": 49.06070287539936,
+ "grad_norm": 0.080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0249,
+ "step": 15356
+ },
+ {
+ "epoch": 49.063897763578275,
+ "grad_norm": 0.0859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0372,
+ "step": 15357
+ },
+ {
+ "epoch": 49.067092651757186,
+ "grad_norm": 0.11181640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0304,
+ "step": 15358
+ },
+ {
+ "epoch": 49.0702875399361,
+ "grad_norm": 0.083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0227,
+ "step": 15359
+ },
+ {
+ "epoch": 49.073482428115014,
+ "grad_norm": 0.1640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0363,
+ "step": 15360
+ },
+ {
+ "epoch": 49.07667731629393,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.033,
+ "step": 15361
+ },
+ {
+ "epoch": 49.07987220447284,
+ "grad_norm": 0.1953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0226,
+ "step": 15362
+ },
+ {
+ "epoch": 49.08306709265176,
+ "grad_norm": 0.04736328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0262,
+ "step": 15363
+ },
+ {
+ "epoch": 49.08626198083067,
+ "grad_norm": 0.10400390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0335,
+ "step": 15364
+ },
+ {
+ "epoch": 49.08945686900959,
+ "grad_norm": 0.0546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0329,
+ "step": 15365
+ },
+ {
+ "epoch": 49.0926517571885,
+ "grad_norm": 0.095703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0328,
+ "step": 15366
+ },
+ {
+ "epoch": 49.09584664536741,
+ "grad_norm": 0.05712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0288,
+ "step": 15367
+ },
+ {
+ "epoch": 49.09904153354633,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0123,
+ "step": 15368
+ },
+ {
+ "epoch": 49.10223642172524,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0237,
+ "step": 15369
+ },
+ {
+ "epoch": 49.105431309904155,
+ "grad_norm": 0.0576171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0345,
+ "step": 15370
+ },
+ {
+ "epoch": 49.108626198083066,
+ "grad_norm": 0.051025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0249,
+ "step": 15371
+ },
+ {
+ "epoch": 49.11182108626198,
+ "grad_norm": 0.05322265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0286,
+ "step": 15372
+ },
+ {
+ "epoch": 49.115015974440894,
+ "grad_norm": 0.060791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0361,
+ "step": 15373
+ },
+ {
+ "epoch": 49.11821086261981,
+ "grad_norm": 0.051513671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0246,
+ "step": 15374
+ },
+ {
+ "epoch": 49.12140575079872,
+ "grad_norm": 0.054443359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0339,
+ "step": 15375
+ },
+ {
+ "epoch": 49.12460063897763,
+ "grad_norm": 0.0498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0335,
+ "step": 15376
+ },
+ {
+ "epoch": 49.12779552715655,
+ "grad_norm": 0.045166015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0187,
+ "step": 15377
+ },
+ {
+ "epoch": 49.13099041533546,
+ "grad_norm": 0.04833984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0354,
+ "step": 15378
+ },
+ {
+ "epoch": 49.13418530351438,
+ "grad_norm": 0.056884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0338,
+ "step": 15379
+ },
+ {
+ "epoch": 49.13738019169329,
+ "grad_norm": 0.057861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.033,
+ "step": 15380
+ },
+ {
+ "epoch": 49.14057507987221,
+ "grad_norm": 0.049072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0234,
+ "step": 15381
+ },
+ {
+ "epoch": 49.14376996805112,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0248,
+ "step": 15382
+ },
+ {
+ "epoch": 49.146964856230035,
+ "grad_norm": 0.050048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0318,
+ "step": 15383
+ },
+ {
+ "epoch": 49.150159744408946,
+ "grad_norm": 0.05126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0467,
+ "step": 15384
+ },
+ {
+ "epoch": 49.153354632587856,
+ "grad_norm": 0.0546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0297,
+ "step": 15385
+ },
+ {
+ "epoch": 49.156549520766774,
+ "grad_norm": 0.080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0299,
+ "step": 15386
+ },
+ {
+ "epoch": 49.159744408945684,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0263,
+ "step": 15387
+ },
+ {
+ "epoch": 49.1629392971246,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.017,
+ "step": 15388
+ },
+ {
+ "epoch": 49.16613418530351,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0289,
+ "step": 15389
+ },
+ {
+ "epoch": 49.16932907348243,
+ "grad_norm": 0.051025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0278,
+ "step": 15390
+ },
+ {
+ "epoch": 49.17252396166134,
+ "grad_norm": 0.0830078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0287,
+ "step": 15391
+ },
+ {
+ "epoch": 49.17571884984026,
+ "grad_norm": 0.059814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0321,
+ "step": 15392
+ },
+ {
+ "epoch": 49.17891373801917,
+ "grad_norm": 0.053466796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0308,
+ "step": 15393
+ },
+ {
+ "epoch": 49.18210862619808,
+ "grad_norm": 0.05078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0203,
+ "step": 15394
+ },
+ {
+ "epoch": 49.185303514377,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0364,
+ "step": 15395
+ },
+ {
+ "epoch": 49.18849840255591,
+ "grad_norm": 0.050537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0292,
+ "step": 15396
+ },
+ {
+ "epoch": 49.191693290734825,
+ "grad_norm": 0.056396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0248,
+ "step": 15397
+ },
+ {
+ "epoch": 49.194888178913736,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0363,
+ "step": 15398
+ },
+ {
+ "epoch": 49.198083067092654,
+ "grad_norm": 0.05029296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0218,
+ "step": 15399
+ },
+ {
+ "epoch": 49.201277955271564,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.022,
+ "step": 15400
+ },
+ {
+ "epoch": 49.20447284345048,
+ "grad_norm": 0.051025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0287,
+ "step": 15401
+ },
+ {
+ "epoch": 49.20766773162939,
+ "grad_norm": 0.05224609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0205,
+ "step": 15402
+ },
+ {
+ "epoch": 49.21086261980831,
+ "grad_norm": 0.045166015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0247,
+ "step": 15403
+ },
+ {
+ "epoch": 49.21405750798722,
+ "grad_norm": 0.060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0389,
+ "step": 15404
+ },
+ {
+ "epoch": 49.21725239616613,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0242,
+ "step": 15405
+ },
+ {
+ "epoch": 49.22044728434505,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0225,
+ "step": 15406
+ },
+ {
+ "epoch": 49.22364217252396,
+ "grad_norm": 0.0478515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0321,
+ "step": 15407
+ },
+ {
+ "epoch": 49.22683706070288,
+ "grad_norm": 0.083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0351,
+ "step": 15408
+ },
+ {
+ "epoch": 49.23003194888179,
+ "grad_norm": 0.04736328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0272,
+ "step": 15409
+ },
+ {
+ "epoch": 49.233226837060705,
+ "grad_norm": 0.1015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0381,
+ "step": 15410
+ },
+ {
+ "epoch": 49.236421725239616,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0212,
+ "step": 15411
+ },
+ {
+ "epoch": 49.239616613418534,
+ "grad_norm": 0.048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0321,
+ "step": 15412
+ },
+ {
+ "epoch": 49.242811501597444,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0351,
+ "step": 15413
+ },
+ {
+ "epoch": 49.246006389776355,
+ "grad_norm": 0.060302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0346,
+ "step": 15414
+ },
+ {
+ "epoch": 49.24920127795527,
+ "grad_norm": 0.04736328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0245,
+ "step": 15415
+ },
+ {
+ "epoch": 49.25239616613418,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0255,
+ "step": 15416
+ },
+ {
+ "epoch": 49.2555910543131,
+ "grad_norm": 0.08251953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0348,
+ "step": 15417
+ },
+ {
+ "epoch": 49.25878594249201,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0331,
+ "step": 15418
+ },
+ {
+ "epoch": 49.26198083067093,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0301,
+ "step": 15419
+ },
+ {
+ "epoch": 49.26517571884984,
+ "grad_norm": 0.060791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0311,
+ "step": 15420
+ },
+ {
+ "epoch": 49.26837060702876,
+ "grad_norm": 0.04931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0252,
+ "step": 15421
+ },
+ {
+ "epoch": 49.27156549520767,
+ "grad_norm": 0.080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0184,
+ "step": 15422
+ },
+ {
+ "epoch": 49.27476038338658,
+ "grad_norm": 0.0537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.02,
+ "step": 15423
+ },
+ {
+ "epoch": 49.277955271565496,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0305,
+ "step": 15424
+ },
+ {
+ "epoch": 49.281150159744406,
+ "grad_norm": 0.04931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0287,
+ "step": 15425
+ },
+ {
+ "epoch": 49.284345047923324,
+ "grad_norm": 0.08349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.033,
+ "step": 15426
+ },
+ {
+ "epoch": 49.287539936102235,
+ "grad_norm": 0.052978515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0284,
+ "step": 15427
+ },
+ {
+ "epoch": 49.29073482428115,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0217,
+ "step": 15428
+ },
+ {
+ "epoch": 49.29392971246006,
+ "grad_norm": 0.059814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0239,
+ "step": 15429
+ },
+ {
+ "epoch": 49.29712460063898,
+ "grad_norm": 0.059326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0305,
+ "step": 15430
+ },
+ {
+ "epoch": 49.30031948881789,
+ "grad_norm": 0.05908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0235,
+ "step": 15431
+ },
+ {
+ "epoch": 49.3035143769968,
+ "grad_norm": 0.05517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0202,
+ "step": 15432
+ },
+ {
+ "epoch": 49.30670926517572,
+ "grad_norm": 0.048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0126,
+ "step": 15433
+ },
+ {
+ "epoch": 49.30990415335463,
+ "grad_norm": 0.10888671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0325,
+ "step": 15434
+ },
+ {
+ "epoch": 49.31309904153355,
+ "grad_norm": 0.061767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.027,
+ "step": 15435
+ },
+ {
+ "epoch": 49.31629392971246,
+ "grad_norm": 0.10546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0263,
+ "step": 15436
+ },
+ {
+ "epoch": 49.319488817891376,
+ "grad_norm": 0.08203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0335,
+ "step": 15437
+ },
+ {
+ "epoch": 49.322683706070286,
+ "grad_norm": 0.057861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.021,
+ "step": 15438
+ },
+ {
+ "epoch": 49.325878594249204,
+ "grad_norm": 0.12255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0267,
+ "step": 15439
+ },
+ {
+ "epoch": 49.329073482428115,
+ "grad_norm": 0.0556640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0198,
+ "step": 15440
+ },
+ {
+ "epoch": 49.33226837060703,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0378,
+ "step": 15441
+ },
+ {
+ "epoch": 49.33546325878594,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0402,
+ "step": 15442
+ },
+ {
+ "epoch": 49.33865814696485,
+ "grad_norm": 0.05126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0364,
+ "step": 15443
+ },
+ {
+ "epoch": 49.34185303514377,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0223,
+ "step": 15444
+ },
+ {
+ "epoch": 49.34504792332268,
+ "grad_norm": 0.0615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0295,
+ "step": 15445
+ },
+ {
+ "epoch": 49.3482428115016,
+ "grad_norm": 0.05224609375,
+ "learning_rate": 0.0005,
+ "loss": 1.025,
+ "step": 15446
+ },
+ {
+ "epoch": 49.35143769968051,
+ "grad_norm": 0.058349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0272,
+ "step": 15447
+ },
+ {
+ "epoch": 49.35463258785943,
+ "grad_norm": 0.052978515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0353,
+ "step": 15448
+ },
+ {
+ "epoch": 49.35782747603834,
+ "grad_norm": 0.053955078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0289,
+ "step": 15449
+ },
+ {
+ "epoch": 49.361022364217256,
+ "grad_norm": 0.056884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0142,
+ "step": 15450
+ },
+ {
+ "epoch": 49.364217252396166,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0244,
+ "step": 15451
+ },
+ {
+ "epoch": 49.36741214057508,
+ "grad_norm": 0.08349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0299,
+ "step": 15452
+ },
+ {
+ "epoch": 49.370607028753994,
+ "grad_norm": 0.0498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0276,
+ "step": 15453
+ },
+ {
+ "epoch": 49.373801916932905,
+ "grad_norm": 0.052734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0273,
+ "step": 15454
+ },
+ {
+ "epoch": 49.37699680511182,
+ "grad_norm": 0.052734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0293,
+ "step": 15455
+ },
+ {
+ "epoch": 49.38019169329073,
+ "grad_norm": 0.04931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0366,
+ "step": 15456
+ },
+ {
+ "epoch": 49.38338658146965,
+ "grad_norm": 0.052001953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0328,
+ "step": 15457
+ },
+ {
+ "epoch": 49.38658146964856,
+ "grad_norm": 0.0537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0279,
+ "step": 15458
+ },
+ {
+ "epoch": 49.38977635782748,
+ "grad_norm": 0.09326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0319,
+ "step": 15459
+ },
+ {
+ "epoch": 49.39297124600639,
+ "grad_norm": 0.052001953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0233,
+ "step": 15460
+ },
+ {
+ "epoch": 49.3961661341853,
+ "grad_norm": 0.11181640625,
+ "learning_rate": 0.0005,
+ "loss": 1.032,
+ "step": 15461
+ },
+ {
+ "epoch": 49.39936102236422,
+ "grad_norm": 0.10546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0309,
+ "step": 15462
+ },
+ {
+ "epoch": 49.40255591054313,
+ "grad_norm": 0.1259765625,
+ "learning_rate": 0.0005,
+ "loss": 1.019,
+ "step": 15463
+ },
+ {
+ "epoch": 49.405750798722046,
+ "grad_norm": 0.166015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0223,
+ "step": 15464
+ },
+ {
+ "epoch": 49.40894568690096,
+ "grad_norm": 0.142578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0225,
+ "step": 15465
+ },
+ {
+ "epoch": 49.412140575079874,
+ "grad_norm": 0.1181640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0313,
+ "step": 15466
+ },
+ {
+ "epoch": 49.415335463258785,
+ "grad_norm": 0.08349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0183,
+ "step": 15467
+ },
+ {
+ "epoch": 49.4185303514377,
+ "grad_norm": 0.2197265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0303,
+ "step": 15468
+ },
+ {
+ "epoch": 49.42172523961661,
+ "grad_norm": 0.10546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0313,
+ "step": 15469
+ },
+ {
+ "epoch": 49.424920127795524,
+ "grad_norm": 0.1865234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0263,
+ "step": 15470
+ },
+ {
+ "epoch": 49.42811501597444,
+ "grad_norm": 0.1708984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0174,
+ "step": 15471
+ },
+ {
+ "epoch": 49.43130990415335,
+ "grad_norm": 0.2412109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0247,
+ "step": 15472
+ },
+ {
+ "epoch": 49.43450479233227,
+ "grad_norm": 0.173828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0227,
+ "step": 15473
+ },
+ {
+ "epoch": 49.43769968051118,
+ "grad_norm": 0.2294921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0451,
+ "step": 15474
+ },
+ {
+ "epoch": 49.4408945686901,
+ "grad_norm": 0.1943359375,
+ "learning_rate": 0.0005,
+ "loss": 1.027,
+ "step": 15475
+ },
+ {
+ "epoch": 49.44408945686901,
+ "grad_norm": 0.2421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0262,
+ "step": 15476
+ },
+ {
+ "epoch": 49.447284345047926,
+ "grad_norm": 0.1611328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0324,
+ "step": 15477
+ },
+ {
+ "epoch": 49.45047923322684,
+ "grad_norm": 0.17578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0271,
+ "step": 15478
+ },
+ {
+ "epoch": 49.453674121405754,
+ "grad_norm": 0.14453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0372,
+ "step": 15479
+ },
+ {
+ "epoch": 49.456869009584665,
+ "grad_norm": 0.080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.036,
+ "step": 15480
+ },
+ {
+ "epoch": 49.460063897763575,
+ "grad_norm": 0.140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0298,
+ "step": 15481
+ },
+ {
+ "epoch": 49.46325878594249,
+ "grad_norm": 0.0849609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0257,
+ "step": 15482
+ },
+ {
+ "epoch": 49.466453674121404,
+ "grad_norm": 0.142578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0417,
+ "step": 15483
+ },
+ {
+ "epoch": 49.46964856230032,
+ "grad_norm": 0.11376953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0219,
+ "step": 15484
+ },
+ {
+ "epoch": 49.47284345047923,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0276,
+ "step": 15485
+ },
+ {
+ "epoch": 49.47603833865815,
+ "grad_norm": 0.08642578125,
+ "learning_rate": 0.0005,
+ "loss": 1.03,
+ "step": 15486
+ },
+ {
+ "epoch": 49.47923322683706,
+ "grad_norm": 0.1201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0145,
+ "step": 15487
+ },
+ {
+ "epoch": 49.48242811501598,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0362,
+ "step": 15488
+ },
+ {
+ "epoch": 49.48562300319489,
+ "grad_norm": 0.11328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0346,
+ "step": 15489
+ },
+ {
+ "epoch": 49.4888178913738,
+ "grad_norm": 0.0908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0259,
+ "step": 15490
+ },
+ {
+ "epoch": 49.49201277955272,
+ "grad_norm": 0.1298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0309,
+ "step": 15491
+ },
+ {
+ "epoch": 49.49520766773163,
+ "grad_norm": 0.10693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0344,
+ "step": 15492
+ },
+ {
+ "epoch": 49.498402555910545,
+ "grad_norm": 0.08349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0398,
+ "step": 15493
+ },
+ {
+ "epoch": 49.501597444089455,
+ "grad_norm": 0.154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0255,
+ "step": 15494
+ },
+ {
+ "epoch": 49.50479233226837,
+ "grad_norm": 0.11474609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0268,
+ "step": 15495
+ },
+ {
+ "epoch": 49.50798722044728,
+ "grad_norm": 0.17578125,
+ "learning_rate": 0.0005,
+ "loss": 1.028,
+ "step": 15496
+ },
+ {
+ "epoch": 49.5111821086262,
+ "grad_norm": 0.158203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0235,
+ "step": 15497
+ },
+ {
+ "epoch": 49.51437699680511,
+ "grad_norm": 0.2041015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0253,
+ "step": 15498
+ },
+ {
+ "epoch": 49.51757188498402,
+ "grad_norm": 0.1162109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0344,
+ "step": 15499
+ },
+ {
+ "epoch": 49.52076677316294,
+ "grad_norm": 0.0888671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0362,
+ "step": 15500
+ },
+ {
+ "epoch": 49.52396166134185,
+ "grad_norm": 0.1318359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0317,
+ "step": 15501
+ },
+ {
+ "epoch": 49.52715654952077,
+ "grad_norm": 0.1328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0215,
+ "step": 15502
+ },
+ {
+ "epoch": 49.53035143769968,
+ "grad_norm": 0.1376953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0262,
+ "step": 15503
+ },
+ {
+ "epoch": 49.533546325878596,
+ "grad_norm": 0.1328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0341,
+ "step": 15504
+ },
+ {
+ "epoch": 49.53674121405751,
+ "grad_norm": 0.1552734375,
+ "learning_rate": 0.0005,
+ "loss": 1.025,
+ "step": 15505
+ },
+ {
+ "epoch": 49.539936102236425,
+ "grad_norm": 0.162109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0182,
+ "step": 15506
+ },
+ {
+ "epoch": 49.543130990415335,
+ "grad_norm": 0.1357421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0305,
+ "step": 15507
+ },
+ {
+ "epoch": 49.546325878594246,
+ "grad_norm": 0.103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0275,
+ "step": 15508
+ },
+ {
+ "epoch": 49.54952076677316,
+ "grad_norm": 0.12890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0239,
+ "step": 15509
+ },
+ {
+ "epoch": 49.552715654952074,
+ "grad_norm": 0.08447265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0316,
+ "step": 15510
+ },
+ {
+ "epoch": 49.55591054313099,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0365,
+ "step": 15511
+ },
+ {
+ "epoch": 49.5591054313099,
+ "grad_norm": 0.09130859375,
+ "learning_rate": 0.0005,
+ "loss": 1.019,
+ "step": 15512
+ },
+ {
+ "epoch": 49.56230031948882,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0289,
+ "step": 15513
+ },
+ {
+ "epoch": 49.56549520766773,
+ "grad_norm": 0.1064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0361,
+ "step": 15514
+ },
+ {
+ "epoch": 49.56869009584665,
+ "grad_norm": 0.119140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0195,
+ "step": 15515
+ },
+ {
+ "epoch": 49.57188498402556,
+ "grad_norm": 0.1474609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0241,
+ "step": 15516
+ },
+ {
+ "epoch": 49.575079872204476,
+ "grad_norm": 0.1572265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0191,
+ "step": 15517
+ },
+ {
+ "epoch": 49.57827476038339,
+ "grad_norm": 0.11376953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0279,
+ "step": 15518
+ },
+ {
+ "epoch": 49.5814696485623,
+ "grad_norm": 0.08447265625,
+ "learning_rate": 0.0005,
+ "loss": 1.028,
+ "step": 15519
+ },
+ {
+ "epoch": 49.584664536741215,
+ "grad_norm": 0.1171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0376,
+ "step": 15520
+ },
+ {
+ "epoch": 49.587859424920126,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0278,
+ "step": 15521
+ },
+ {
+ "epoch": 49.59105431309904,
+ "grad_norm": 0.062255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0216,
+ "step": 15522
+ },
+ {
+ "epoch": 49.594249201277954,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0311,
+ "step": 15523
+ },
+ {
+ "epoch": 49.59744408945687,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0436,
+ "step": 15524
+ },
+ {
+ "epoch": 49.60063897763578,
+ "grad_norm": 0.09521484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0348,
+ "step": 15525
+ },
+ {
+ "epoch": 49.6038338658147,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0193,
+ "step": 15526
+ },
+ {
+ "epoch": 49.60702875399361,
+ "grad_norm": 0.10595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0271,
+ "step": 15527
+ },
+ {
+ "epoch": 49.61022364217252,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0272,
+ "step": 15528
+ },
+ {
+ "epoch": 49.61341853035144,
+ "grad_norm": 0.109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0258,
+ "step": 15529
+ },
+ {
+ "epoch": 49.61661341853035,
+ "grad_norm": 0.10205078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0337,
+ "step": 15530
+ },
+ {
+ "epoch": 49.61980830670927,
+ "grad_norm": 0.1904296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0248,
+ "step": 15531
+ },
+ {
+ "epoch": 49.62300319488818,
+ "grad_norm": 0.053466796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0323,
+ "step": 15532
+ },
+ {
+ "epoch": 49.626198083067095,
+ "grad_norm": 0.0966796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0202,
+ "step": 15533
+ },
+ {
+ "epoch": 49.629392971246006,
+ "grad_norm": 0.0517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0322,
+ "step": 15534
+ },
+ {
+ "epoch": 49.63258785942492,
+ "grad_norm": 0.087890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0394,
+ "step": 15535
+ },
+ {
+ "epoch": 49.635782747603834,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0269,
+ "step": 15536
+ },
+ {
+ "epoch": 49.638977635782744,
+ "grad_norm": 0.1298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0183,
+ "step": 15537
+ },
+ {
+ "epoch": 49.64217252396166,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0295,
+ "step": 15538
+ },
+ {
+ "epoch": 49.64536741214057,
+ "grad_norm": 0.1767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0277,
+ "step": 15539
+ },
+ {
+ "epoch": 49.64856230031949,
+ "grad_norm": 0.05712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0258,
+ "step": 15540
+ },
+ {
+ "epoch": 49.6517571884984,
+ "grad_norm": 0.1484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0358,
+ "step": 15541
+ },
+ {
+ "epoch": 49.65495207667732,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0246,
+ "step": 15542
+ },
+ {
+ "epoch": 49.65814696485623,
+ "grad_norm": 0.15234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0241,
+ "step": 15543
+ },
+ {
+ "epoch": 49.66134185303515,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0482,
+ "step": 15544
+ },
+ {
+ "epoch": 49.66453674121406,
+ "grad_norm": 0.095703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0252,
+ "step": 15545
+ },
+ {
+ "epoch": 49.66773162939297,
+ "grad_norm": 0.0498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0363,
+ "step": 15546
+ },
+ {
+ "epoch": 49.670926517571885,
+ "grad_norm": 0.10595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0254,
+ "step": 15547
+ },
+ {
+ "epoch": 49.674121405750796,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0281,
+ "step": 15548
+ },
+ {
+ "epoch": 49.677316293929714,
+ "grad_norm": 0.11669921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0288,
+ "step": 15549
+ },
+ {
+ "epoch": 49.680511182108624,
+ "grad_norm": 0.08642578125,
+ "learning_rate": 0.0005,
+ "loss": 1.024,
+ "step": 15550
+ },
+ {
+ "epoch": 49.68370607028754,
+ "grad_norm": 0.107421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0212,
+ "step": 15551
+ },
+ {
+ "epoch": 49.68690095846645,
+ "grad_norm": 0.052734375,
+ "learning_rate": 0.0005,
+ "loss": 1.019,
+ "step": 15552
+ },
+ {
+ "epoch": 49.69009584664537,
+ "grad_norm": 0.11572265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0281,
+ "step": 15553
+ },
+ {
+ "epoch": 49.69329073482428,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0136,
+ "step": 15554
+ },
+ {
+ "epoch": 49.6964856230032,
+ "grad_norm": 0.142578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0292,
+ "step": 15555
+ },
+ {
+ "epoch": 49.69968051118211,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0319,
+ "step": 15556
+ },
+ {
+ "epoch": 49.70287539936102,
+ "grad_norm": 0.1337890625,
+ "learning_rate": 0.0005,
+ "loss": 1.029,
+ "step": 15557
+ },
+ {
+ "epoch": 49.70607028753994,
+ "grad_norm": 0.10107421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0337,
+ "step": 15558
+ },
+ {
+ "epoch": 49.70926517571885,
+ "grad_norm": 0.1337890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0254,
+ "step": 15559
+ },
+ {
+ "epoch": 49.712460063897765,
+ "grad_norm": 0.10498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0216,
+ "step": 15560
+ },
+ {
+ "epoch": 49.715654952076676,
+ "grad_norm": 0.12890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0301,
+ "step": 15561
+ },
+ {
+ "epoch": 49.718849840255594,
+ "grad_norm": 0.11669921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0206,
+ "step": 15562
+ },
+ {
+ "epoch": 49.722044728434504,
+ "grad_norm": 0.09765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0237,
+ "step": 15563
+ },
+ {
+ "epoch": 49.72523961661342,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0247,
+ "step": 15564
+ },
+ {
+ "epoch": 49.72843450479233,
+ "grad_norm": 0.10400390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0328,
+ "step": 15565
+ },
+ {
+ "epoch": 49.73162939297124,
+ "grad_norm": 0.0595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0366,
+ "step": 15566
+ },
+ {
+ "epoch": 49.73482428115016,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0343,
+ "step": 15567
+ },
+ {
+ "epoch": 49.73801916932907,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0267,
+ "step": 15568
+ },
+ {
+ "epoch": 49.74121405750799,
+ "grad_norm": 0.053466796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0292,
+ "step": 15569
+ },
+ {
+ "epoch": 49.7444089456869,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0284,
+ "step": 15570
+ },
+ {
+ "epoch": 49.74760383386582,
+ "grad_norm": 0.05419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0356,
+ "step": 15571
+ },
+ {
+ "epoch": 49.75079872204473,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0331,
+ "step": 15572
+ },
+ {
+ "epoch": 49.753993610223645,
+ "grad_norm": 0.051513671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0243,
+ "step": 15573
+ },
+ {
+ "epoch": 49.757188498402556,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0302,
+ "step": 15574
+ },
+ {
+ "epoch": 49.760383386581466,
+ "grad_norm": 0.051513671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0263,
+ "step": 15575
+ },
+ {
+ "epoch": 49.763578274760384,
+ "grad_norm": 0.08056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0408,
+ "step": 15576
+ },
+ {
+ "epoch": 49.766773162939295,
+ "grad_norm": 0.0498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0256,
+ "step": 15577
+ },
+ {
+ "epoch": 49.76996805111821,
+ "grad_norm": 0.09814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0352,
+ "step": 15578
+ },
+ {
+ "epoch": 49.77316293929712,
+ "grad_norm": 0.060302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0317,
+ "step": 15579
+ },
+ {
+ "epoch": 49.77635782747604,
+ "grad_norm": 0.09375,
+ "learning_rate": 0.0005,
+ "loss": 1.0218,
+ "step": 15580
+ },
+ {
+ "epoch": 49.77955271565495,
+ "grad_norm": 0.053955078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0249,
+ "step": 15581
+ },
+ {
+ "epoch": 49.78274760383387,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0168,
+ "step": 15582
+ },
+ {
+ "epoch": 49.78594249201278,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.033,
+ "step": 15583
+ },
+ {
+ "epoch": 49.78913738019169,
+ "grad_norm": 0.05419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0321,
+ "step": 15584
+ },
+ {
+ "epoch": 49.79233226837061,
+ "grad_norm": 0.1015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0342,
+ "step": 15585
+ },
+ {
+ "epoch": 49.79552715654952,
+ "grad_norm": 0.05078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0267,
+ "step": 15586
+ },
+ {
+ "epoch": 49.798722044728436,
+ "grad_norm": 0.08447265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0234,
+ "step": 15587
+ },
+ {
+ "epoch": 49.801916932907346,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0337,
+ "step": 15588
+ },
+ {
+ "epoch": 49.805111821086264,
+ "grad_norm": 0.12060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0299,
+ "step": 15589
+ },
+ {
+ "epoch": 49.808306709265175,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0272,
+ "step": 15590
+ },
+ {
+ "epoch": 49.81150159744409,
+ "grad_norm": 0.11572265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0338,
+ "step": 15591
+ },
+ {
+ "epoch": 49.814696485623,
+ "grad_norm": 0.05224609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0255,
+ "step": 15592
+ },
+ {
+ "epoch": 49.81789137380191,
+ "grad_norm": 0.08837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0284,
+ "step": 15593
+ },
+ {
+ "epoch": 49.82108626198083,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0194,
+ "step": 15594
+ },
+ {
+ "epoch": 49.82428115015974,
+ "grad_norm": 0.0947265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0288,
+ "step": 15595
+ },
+ {
+ "epoch": 49.82747603833866,
+ "grad_norm": 0.058837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0158,
+ "step": 15596
+ },
+ {
+ "epoch": 49.83067092651757,
+ "grad_norm": 0.09814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0203,
+ "step": 15597
+ },
+ {
+ "epoch": 49.83386581469649,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0273,
+ "step": 15598
+ },
+ {
+ "epoch": 49.8370607028754,
+ "grad_norm": 0.10498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0251,
+ "step": 15599
+ },
+ {
+ "epoch": 49.840255591054316,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0309,
+ "step": 15600
+ },
+ {
+ "epoch": 49.843450479233226,
+ "grad_norm": 0.1416015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0386,
+ "step": 15601
+ },
+ {
+ "epoch": 49.846645367412144,
+ "grad_norm": 0.050048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0429,
+ "step": 15602
+ },
+ {
+ "epoch": 49.849840255591054,
+ "grad_norm": 0.12109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0316,
+ "step": 15603
+ },
+ {
+ "epoch": 49.853035143769965,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0381,
+ "step": 15604
+ },
+ {
+ "epoch": 49.85623003194888,
+ "grad_norm": 0.169921875,
+ "learning_rate": 0.0005,
+ "loss": 1.029,
+ "step": 15605
+ },
+ {
+ "epoch": 49.85942492012779,
+ "grad_norm": 0.053466796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0273,
+ "step": 15606
+ },
+ {
+ "epoch": 49.86261980830671,
+ "grad_norm": 0.18359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0266,
+ "step": 15607
+ },
+ {
+ "epoch": 49.86581469648562,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0352,
+ "step": 15608
+ },
+ {
+ "epoch": 49.86900958466454,
+ "grad_norm": 0.154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0227,
+ "step": 15609
+ },
+ {
+ "epoch": 49.87220447284345,
+ "grad_norm": 0.05029296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0275,
+ "step": 15610
+ },
+ {
+ "epoch": 49.87539936102237,
+ "grad_norm": 0.1162109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0302,
+ "step": 15611
+ },
+ {
+ "epoch": 49.87859424920128,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0226,
+ "step": 15612
+ },
+ {
+ "epoch": 49.88178913738019,
+ "grad_norm": 0.08740234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0283,
+ "step": 15613
+ },
+ {
+ "epoch": 49.884984025559106,
+ "grad_norm": 0.059814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0358,
+ "step": 15614
+ },
+ {
+ "epoch": 49.88817891373802,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0234,
+ "step": 15615
+ },
+ {
+ "epoch": 49.891373801916934,
+ "grad_norm": 0.0849609375,
+ "learning_rate": 0.0005,
+ "loss": 1.018,
+ "step": 15616
+ },
+ {
+ "epoch": 49.894568690095845,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0307,
+ "step": 15617
+ },
+ {
+ "epoch": 49.89776357827476,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.029,
+ "step": 15618
+ },
+ {
+ "epoch": 49.90095846645367,
+ "grad_norm": 0.0498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0277,
+ "step": 15619
+ },
+ {
+ "epoch": 49.90415335463259,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0245,
+ "step": 15620
+ },
+ {
+ "epoch": 49.9073482428115,
+ "grad_norm": 0.054443359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0308,
+ "step": 15621
+ },
+ {
+ "epoch": 49.91054313099041,
+ "grad_norm": 0.055908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0344,
+ "step": 15622
+ },
+ {
+ "epoch": 49.91373801916933,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0287,
+ "step": 15623
+ },
+ {
+ "epoch": 49.91693290734824,
+ "grad_norm": 0.054443359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0296,
+ "step": 15624
+ },
+ {
+ "epoch": 49.92012779552716,
+ "grad_norm": 0.048095703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0386,
+ "step": 15625
+ },
+ {
+ "epoch": 49.92332268370607,
+ "grad_norm": 0.061767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0352,
+ "step": 15626
+ },
+ {
+ "epoch": 49.926517571884986,
+ "grad_norm": 0.051025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0171,
+ "step": 15627
+ },
+ {
+ "epoch": 49.9297124600639,
+ "grad_norm": 0.0458984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0271,
+ "step": 15628
+ },
+ {
+ "epoch": 49.932907348242814,
+ "grad_norm": 0.044921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0325,
+ "step": 15629
+ },
+ {
+ "epoch": 49.936102236421725,
+ "grad_norm": 0.048583984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0247,
+ "step": 15630
+ },
+ {
+ "epoch": 49.93929712460064,
+ "grad_norm": 0.054443359375,
+ "learning_rate": 0.0005,
+ "loss": 1.027,
+ "step": 15631
+ },
+ {
+ "epoch": 49.94249201277955,
+ "grad_norm": 0.048583984375,
+ "learning_rate": 0.0005,
+ "loss": 1.032,
+ "step": 15632
+ },
+ {
+ "epoch": 49.945686900958464,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.037,
+ "step": 15633
+ },
+ {
+ "epoch": 49.94888178913738,
+ "grad_norm": 0.1328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0276,
+ "step": 15634
+ },
+ {
+ "epoch": 49.95207667731629,
+ "grad_norm": 0.09423828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0302,
+ "step": 15635
+ },
+ {
+ "epoch": 49.95527156549521,
+ "grad_norm": 0.07861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0228,
+ "step": 15636
+ },
+ {
+ "epoch": 49.95846645367412,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0267,
+ "step": 15637
+ },
+ {
+ "epoch": 49.96166134185304,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.024,
+ "step": 15638
+ },
+ {
+ "epoch": 49.96485623003195,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0269,
+ "step": 15639
+ },
+ {
+ "epoch": 49.968051118210866,
+ "grad_norm": 0.048095703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0256,
+ "step": 15640
+ },
+ {
+ "epoch": 49.97124600638978,
+ "grad_norm": 0.0595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0307,
+ "step": 15641
+ },
+ {
+ "epoch": 49.97444089456869,
+ "grad_norm": 0.05322265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0201,
+ "step": 15642
+ },
+ {
+ "epoch": 49.977635782747605,
+ "grad_norm": 0.050537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0395,
+ "step": 15643
+ },
+ {
+ "epoch": 49.980830670926515,
+ "grad_norm": 0.04833984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0295,
+ "step": 15644
+ },
+ {
+ "epoch": 49.98402555910543,
+ "grad_norm": 0.0498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0268,
+ "step": 15645
+ },
+ {
+ "epoch": 49.98722044728434,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0305,
+ "step": 15646
+ },
+ {
+ "epoch": 49.99041533546326,
+ "grad_norm": 0.05810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0383,
+ "step": 15647
+ },
+ {
+ "epoch": 49.99361022364217,
+ "grad_norm": 0.055908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0279,
+ "step": 15648
+ },
+ {
+ "epoch": 49.99680511182109,
+ "grad_norm": 0.045166015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0311,
+ "step": 15649
+ },
+ {
+ "epoch": 50.0,
+ "grad_norm": 0.051513671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0293,
+ "step": 15650
+ },
+ {
+ "epoch": 50.00319488817891,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0294,
+ "step": 15651
+ },
+ {
+ "epoch": 50.00638977635783,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0101,
+ "step": 15652
+ },
+ {
+ "epoch": 50.00958466453674,
+ "grad_norm": 0.053466796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0326,
+ "step": 15653
+ },
+ {
+ "epoch": 50.01277955271566,
+ "grad_norm": 0.0537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0217,
+ "step": 15654
+ },
+ {
+ "epoch": 50.01597444089457,
+ "grad_norm": 0.049072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.025,
+ "step": 15655
+ },
+ {
+ "epoch": 50.019169329073485,
+ "grad_norm": 0.04736328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0321,
+ "step": 15656
+ },
+ {
+ "epoch": 50.022364217252395,
+ "grad_norm": 0.056396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0248,
+ "step": 15657
+ },
+ {
+ "epoch": 50.02555910543131,
+ "grad_norm": 0.05126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0239,
+ "step": 15658
+ },
+ {
+ "epoch": 50.02875399361022,
+ "grad_norm": 0.05078125,
+ "learning_rate": 0.0005,
+ "loss": 1.017,
+ "step": 15659
+ },
+ {
+ "epoch": 50.031948881789134,
+ "grad_norm": 0.048583984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0328,
+ "step": 15660
+ },
+ {
+ "epoch": 50.03514376996805,
+ "grad_norm": 0.054443359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0253,
+ "step": 15661
+ },
+ {
+ "epoch": 50.03833865814696,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0285,
+ "step": 15662
+ },
+ {
+ "epoch": 50.04153354632588,
+ "grad_norm": 0.04931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0315,
+ "step": 15663
+ },
+ {
+ "epoch": 50.04472843450479,
+ "grad_norm": 0.06201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0289,
+ "step": 15664
+ },
+ {
+ "epoch": 50.04792332268371,
+ "grad_norm": 0.08203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0269,
+ "step": 15665
+ },
+ {
+ "epoch": 50.05111821086262,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0333,
+ "step": 15666
+ },
+ {
+ "epoch": 50.054313099041536,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0245,
+ "step": 15667
+ },
+ {
+ "epoch": 50.05750798722045,
+ "grad_norm": 0.08056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0277,
+ "step": 15668
+ },
+ {
+ "epoch": 50.06070287539936,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0244,
+ "step": 15669
+ },
+ {
+ "epoch": 50.063897763578275,
+ "grad_norm": 0.0546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0161,
+ "step": 15670
+ },
+ {
+ "epoch": 50.067092651757186,
+ "grad_norm": 0.056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0257,
+ "step": 15671
+ },
+ {
+ "epoch": 50.0702875399361,
+ "grad_norm": 0.050048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0209,
+ "step": 15672
+ },
+ {
+ "epoch": 50.073482428115014,
+ "grad_norm": 0.044921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0415,
+ "step": 15673
+ },
+ {
+ "epoch": 50.07667731629393,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0282,
+ "step": 15674
+ },
+ {
+ "epoch": 50.07987220447284,
+ "grad_norm": 0.05517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.029,
+ "step": 15675
+ },
+ {
+ "epoch": 50.08306709265176,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0309,
+ "step": 15676
+ },
+ {
+ "epoch": 50.08626198083067,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0175,
+ "step": 15677
+ },
+ {
+ "epoch": 50.08945686900959,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0342,
+ "step": 15678
+ },
+ {
+ "epoch": 50.0926517571885,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0218,
+ "step": 15679
+ },
+ {
+ "epoch": 50.09584664536741,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0363,
+ "step": 15680
+ },
+ {
+ "epoch": 50.09904153354633,
+ "grad_norm": 0.059326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0274,
+ "step": 15681
+ },
+ {
+ "epoch": 50.10223642172524,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0358,
+ "step": 15682
+ },
+ {
+ "epoch": 50.105431309904155,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0243,
+ "step": 15683
+ },
+ {
+ "epoch": 50.108626198083066,
+ "grad_norm": 0.05078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0315,
+ "step": 15684
+ },
+ {
+ "epoch": 50.11182108626198,
+ "grad_norm": 0.048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0242,
+ "step": 15685
+ },
+ {
+ "epoch": 50.115015974440894,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0299,
+ "step": 15686
+ },
+ {
+ "epoch": 50.11821086261981,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0458,
+ "step": 15687
+ },
+ {
+ "epoch": 50.12140575079872,
+ "grad_norm": 0.083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0314,
+ "step": 15688
+ },
+ {
+ "epoch": 50.12460063897763,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0294,
+ "step": 15689
+ },
+ {
+ "epoch": 50.12779552715655,
+ "grad_norm": 0.05419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.027,
+ "step": 15690
+ },
+ {
+ "epoch": 50.13099041533546,
+ "grad_norm": 0.08251953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0482,
+ "step": 15691
+ },
+ {
+ "epoch": 50.13418530351438,
+ "grad_norm": 0.052978515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0182,
+ "step": 15692
+ },
+ {
+ "epoch": 50.13738019169329,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0351,
+ "step": 15693
+ },
+ {
+ "epoch": 50.14057507987221,
+ "grad_norm": 0.05517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0174,
+ "step": 15694
+ },
+ {
+ "epoch": 50.14376996805112,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0348,
+ "step": 15695
+ },
+ {
+ "epoch": 50.146964856230035,
+ "grad_norm": 0.055908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0356,
+ "step": 15696
+ },
+ {
+ "epoch": 50.150159744408946,
+ "grad_norm": 0.05322265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0219,
+ "step": 15697
+ },
+ {
+ "epoch": 50.153354632587856,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0246,
+ "step": 15698
+ },
+ {
+ "epoch": 50.156549520766774,
+ "grad_norm": 0.06005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0289,
+ "step": 15699
+ },
+ {
+ "epoch": 50.159744408945684,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0296,
+ "step": 15700
+ },
+ {
+ "epoch": 50.1629392971246,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.024,
+ "step": 15701
+ },
+ {
+ "epoch": 50.16613418530351,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0296,
+ "step": 15702
+ },
+ {
+ "epoch": 50.16932907348243,
+ "grad_norm": 0.050537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0303,
+ "step": 15703
+ },
+ {
+ "epoch": 50.17252396166134,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0261,
+ "step": 15704
+ },
+ {
+ "epoch": 50.17571884984026,
+ "grad_norm": 0.056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0258,
+ "step": 15705
+ },
+ {
+ "epoch": 50.17891373801917,
+ "grad_norm": 0.04833984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0335,
+ "step": 15706
+ },
+ {
+ "epoch": 50.18210862619808,
+ "grad_norm": 0.04931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0315,
+ "step": 15707
+ },
+ {
+ "epoch": 50.185303514377,
+ "grad_norm": 0.055419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0241,
+ "step": 15708
+ },
+ {
+ "epoch": 50.18849840255591,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0301,
+ "step": 15709
+ },
+ {
+ "epoch": 50.191693290734825,
+ "grad_norm": 0.054443359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0173,
+ "step": 15710
+ },
+ {
+ "epoch": 50.194888178913736,
+ "grad_norm": 0.08251953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0367,
+ "step": 15711
+ },
+ {
+ "epoch": 50.198083067092654,
+ "grad_norm": 0.05517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0254,
+ "step": 15712
+ },
+ {
+ "epoch": 50.201277955271564,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0291,
+ "step": 15713
+ },
+ {
+ "epoch": 50.20447284345048,
+ "grad_norm": 0.054931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0273,
+ "step": 15714
+ },
+ {
+ "epoch": 50.20766773162939,
+ "grad_norm": 0.054931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0221,
+ "step": 15715
+ },
+ {
+ "epoch": 50.21086261980831,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0385,
+ "step": 15716
+ },
+ {
+ "epoch": 50.21405750798722,
+ "grad_norm": 0.051513671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0351,
+ "step": 15717
+ },
+ {
+ "epoch": 50.21725239616613,
+ "grad_norm": 0.09375,
+ "learning_rate": 0.0005,
+ "loss": 1.0236,
+ "step": 15718
+ },
+ {
+ "epoch": 50.22044728434505,
+ "grad_norm": 0.04931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0337,
+ "step": 15719
+ },
+ {
+ "epoch": 50.22364217252396,
+ "grad_norm": 0.048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0258,
+ "step": 15720
+ },
+ {
+ "epoch": 50.22683706070288,
+ "grad_norm": 0.1044921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0321,
+ "step": 15721
+ },
+ {
+ "epoch": 50.23003194888179,
+ "grad_norm": 0.047607421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0301,
+ "step": 15722
+ },
+ {
+ "epoch": 50.233226837060705,
+ "grad_norm": 0.05908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0438,
+ "step": 15723
+ },
+ {
+ "epoch": 50.236421725239616,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0213,
+ "step": 15724
+ },
+ {
+ "epoch": 50.239616613418534,
+ "grad_norm": 0.05908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0218,
+ "step": 15725
+ },
+ {
+ "epoch": 50.242811501597444,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0264,
+ "step": 15726
+ },
+ {
+ "epoch": 50.246006389776355,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0169,
+ "step": 15727
+ },
+ {
+ "epoch": 50.24920127795527,
+ "grad_norm": 0.046630859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0108,
+ "step": 15728
+ },
+ {
+ "epoch": 50.25239616613418,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.04,
+ "step": 15729
+ },
+ {
+ "epoch": 50.2555910543131,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0295,
+ "step": 15730
+ },
+ {
+ "epoch": 50.25878594249201,
+ "grad_norm": 0.051513671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0305,
+ "step": 15731
+ },
+ {
+ "epoch": 50.26198083067093,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0417,
+ "step": 15732
+ },
+ {
+ "epoch": 50.26517571884984,
+ "grad_norm": 0.08349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0197,
+ "step": 15733
+ },
+ {
+ "epoch": 50.26837060702876,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0214,
+ "step": 15734
+ },
+ {
+ "epoch": 50.27156549520767,
+ "grad_norm": 0.099609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0371,
+ "step": 15735
+ },
+ {
+ "epoch": 50.27476038338658,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0361,
+ "step": 15736
+ },
+ {
+ "epoch": 50.277955271565496,
+ "grad_norm": 0.0986328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0288,
+ "step": 15737
+ },
+ {
+ "epoch": 50.281150159744406,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0352,
+ "step": 15738
+ },
+ {
+ "epoch": 50.284345047923324,
+ "grad_norm": 0.10595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0174,
+ "step": 15739
+ },
+ {
+ "epoch": 50.287539936102235,
+ "grad_norm": 0.060791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0315,
+ "step": 15740
+ },
+ {
+ "epoch": 50.29073482428115,
+ "grad_norm": 0.08740234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0128,
+ "step": 15741
+ },
+ {
+ "epoch": 50.29392971246006,
+ "grad_norm": 0.0546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0386,
+ "step": 15742
+ },
+ {
+ "epoch": 50.29712460063898,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0229,
+ "step": 15743
+ },
+ {
+ "epoch": 50.30031948881789,
+ "grad_norm": 0.08740234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0218,
+ "step": 15744
+ },
+ {
+ "epoch": 50.3035143769968,
+ "grad_norm": 0.08642578125,
+ "learning_rate": 0.0005,
+ "loss": 1.031,
+ "step": 15745
+ },
+ {
+ "epoch": 50.30670926517572,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0345,
+ "step": 15746
+ },
+ {
+ "epoch": 50.30990415335463,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0264,
+ "step": 15747
+ },
+ {
+ "epoch": 50.31309904153355,
+ "grad_norm": 0.05712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0202,
+ "step": 15748
+ },
+ {
+ "epoch": 50.31629392971246,
+ "grad_norm": 0.0517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0195,
+ "step": 15749
+ },
+ {
+ "epoch": 50.319488817891376,
+ "grad_norm": 0.04296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0252,
+ "step": 15750
+ },
+ {
+ "epoch": 50.322683706070286,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0237,
+ "step": 15751
+ },
+ {
+ "epoch": 50.325878594249204,
+ "grad_norm": 0.0556640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0373,
+ "step": 15752
+ },
+ {
+ "epoch": 50.329073482428115,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0216,
+ "step": 15753
+ },
+ {
+ "epoch": 50.33226837060703,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0271,
+ "step": 15754
+ },
+ {
+ "epoch": 50.33546325878594,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0258,
+ "step": 15755
+ },
+ {
+ "epoch": 50.33865814696485,
+ "grad_norm": 0.061279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0294,
+ "step": 15756
+ },
+ {
+ "epoch": 50.34185303514377,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0247,
+ "step": 15757
+ },
+ {
+ "epoch": 50.34504792332268,
+ "grad_norm": 0.10498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.019,
+ "step": 15758
+ },
+ {
+ "epoch": 50.3482428115016,
+ "grad_norm": 0.138671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0236,
+ "step": 15759
+ },
+ {
+ "epoch": 50.35143769968051,
+ "grad_norm": 0.173828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0259,
+ "step": 15760
+ },
+ {
+ "epoch": 50.35463258785943,
+ "grad_norm": 0.12890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0265,
+ "step": 15761
+ },
+ {
+ "epoch": 50.35782747603834,
+ "grad_norm": 0.197265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0309,
+ "step": 15762
+ },
+ {
+ "epoch": 50.361022364217256,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0256,
+ "step": 15763
+ },
+ {
+ "epoch": 50.364217252396166,
+ "grad_norm": 0.10546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0355,
+ "step": 15764
+ },
+ {
+ "epoch": 50.36741214057508,
+ "grad_norm": 0.0517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0352,
+ "step": 15765
+ },
+ {
+ "epoch": 50.370607028753994,
+ "grad_norm": 0.103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0205,
+ "step": 15766
+ },
+ {
+ "epoch": 50.373801916932905,
+ "grad_norm": 0.052978515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0221,
+ "step": 15767
+ },
+ {
+ "epoch": 50.37699680511182,
+ "grad_norm": 0.10400390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0274,
+ "step": 15768
+ },
+ {
+ "epoch": 50.38019169329073,
+ "grad_norm": 0.054931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0314,
+ "step": 15769
+ },
+ {
+ "epoch": 50.38338658146965,
+ "grad_norm": 0.10205078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0215,
+ "step": 15770
+ },
+ {
+ "epoch": 50.38658146964856,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0236,
+ "step": 15771
+ },
+ {
+ "epoch": 50.38977635782748,
+ "grad_norm": 0.123046875,
+ "learning_rate": 0.0005,
+ "loss": 1.027,
+ "step": 15772
+ },
+ {
+ "epoch": 50.39297124600639,
+ "grad_norm": 0.095703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0358,
+ "step": 15773
+ },
+ {
+ "epoch": 50.3961661341853,
+ "grad_norm": 0.10302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0271,
+ "step": 15774
+ },
+ {
+ "epoch": 50.39936102236422,
+ "grad_norm": 0.08447265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0267,
+ "step": 15775
+ },
+ {
+ "epoch": 50.40255591054313,
+ "grad_norm": 0.091796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0237,
+ "step": 15776
+ },
+ {
+ "epoch": 50.405750798722046,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0152,
+ "step": 15777
+ },
+ {
+ "epoch": 50.40894568690096,
+ "grad_norm": 0.1025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0377,
+ "step": 15778
+ },
+ {
+ "epoch": 50.412140575079874,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0337,
+ "step": 15779
+ },
+ {
+ "epoch": 50.415335463258785,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0316,
+ "step": 15780
+ },
+ {
+ "epoch": 50.4185303514377,
+ "grad_norm": 0.059814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0283,
+ "step": 15781
+ },
+ {
+ "epoch": 50.42172523961661,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0261,
+ "step": 15782
+ },
+ {
+ "epoch": 50.424920127795524,
+ "grad_norm": 0.052734375,
+ "learning_rate": 0.0005,
+ "loss": 1.03,
+ "step": 15783
+ },
+ {
+ "epoch": 50.42811501597444,
+ "grad_norm": 0.057373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0319,
+ "step": 15784
+ },
+ {
+ "epoch": 50.43130990415335,
+ "grad_norm": 0.0576171875,
+ "learning_rate": 0.0005,
+ "loss": 1.03,
+ "step": 15785
+ },
+ {
+ "epoch": 50.43450479233227,
+ "grad_norm": 0.05615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0278,
+ "step": 15786
+ },
+ {
+ "epoch": 50.43769968051118,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0262,
+ "step": 15787
+ },
+ {
+ "epoch": 50.4408945686901,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0235,
+ "step": 15788
+ },
+ {
+ "epoch": 50.44408945686901,
+ "grad_norm": 0.060791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0265,
+ "step": 15789
+ },
+ {
+ "epoch": 50.447284345047926,
+ "grad_norm": 0.056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0383,
+ "step": 15790
+ },
+ {
+ "epoch": 50.45047923322684,
+ "grad_norm": 0.047119140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0359,
+ "step": 15791
+ },
+ {
+ "epoch": 50.453674121405754,
+ "grad_norm": 0.052978515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0284,
+ "step": 15792
+ },
+ {
+ "epoch": 50.456869009584665,
+ "grad_norm": 0.061767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0253,
+ "step": 15793
+ },
+ {
+ "epoch": 50.460063897763575,
+ "grad_norm": 0.0478515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0214,
+ "step": 15794
+ },
+ {
+ "epoch": 50.46325878594249,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0269,
+ "step": 15795
+ },
+ {
+ "epoch": 50.466453674121404,
+ "grad_norm": 0.054931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0194,
+ "step": 15796
+ },
+ {
+ "epoch": 50.46964856230032,
+ "grad_norm": 0.062255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0434,
+ "step": 15797
+ },
+ {
+ "epoch": 50.47284345047923,
+ "grad_norm": 0.0556640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0299,
+ "step": 15798
+ },
+ {
+ "epoch": 50.47603833865815,
+ "grad_norm": 0.05078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0342,
+ "step": 15799
+ },
+ {
+ "epoch": 50.47923322683706,
+ "grad_norm": 0.04150390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0204,
+ "step": 15800
+ },
+ {
+ "epoch": 50.48242811501598,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.032,
+ "step": 15801
+ },
+ {
+ "epoch": 50.48562300319489,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0273,
+ "step": 15802
+ },
+ {
+ "epoch": 50.4888178913738,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0246,
+ "step": 15803
+ },
+ {
+ "epoch": 50.49201277955272,
+ "grad_norm": 0.05224609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0348,
+ "step": 15804
+ },
+ {
+ "epoch": 50.49520766773163,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0193,
+ "step": 15805
+ },
+ {
+ "epoch": 50.498402555910545,
+ "grad_norm": 0.05517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.019,
+ "step": 15806
+ },
+ {
+ "epoch": 50.501597444089455,
+ "grad_norm": 0.05712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0245,
+ "step": 15807
+ },
+ {
+ "epoch": 50.50479233226837,
+ "grad_norm": 0.052001953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0307,
+ "step": 15808
+ },
+ {
+ "epoch": 50.50798722044728,
+ "grad_norm": 0.056884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0253,
+ "step": 15809
+ },
+ {
+ "epoch": 50.5111821086262,
+ "grad_norm": 0.060791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0193,
+ "step": 15810
+ },
+ {
+ "epoch": 50.51437699680511,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0302,
+ "step": 15811
+ },
+ {
+ "epoch": 50.51757188498402,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0251,
+ "step": 15812
+ },
+ {
+ "epoch": 50.52076677316294,
+ "grad_norm": 0.0478515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0348,
+ "step": 15813
+ },
+ {
+ "epoch": 50.52396166134185,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0289,
+ "step": 15814
+ },
+ {
+ "epoch": 50.52715654952077,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0302,
+ "step": 15815
+ },
+ {
+ "epoch": 50.53035143769968,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0285,
+ "step": 15816
+ },
+ {
+ "epoch": 50.533546325878596,
+ "grad_norm": 0.059814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0347,
+ "step": 15817
+ },
+ {
+ "epoch": 50.53674121405751,
+ "grad_norm": 0.099609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0239,
+ "step": 15818
+ },
+ {
+ "epoch": 50.539936102236425,
+ "grad_norm": 0.049072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.021,
+ "step": 15819
+ },
+ {
+ "epoch": 50.543130990415335,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0217,
+ "step": 15820
+ },
+ {
+ "epoch": 50.546325878594246,
+ "grad_norm": 0.08056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0264,
+ "step": 15821
+ },
+ {
+ "epoch": 50.54952076677316,
+ "grad_norm": 0.08154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0354,
+ "step": 15822
+ },
+ {
+ "epoch": 50.552715654952074,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0282,
+ "step": 15823
+ },
+ {
+ "epoch": 50.55591054313099,
+ "grad_norm": 0.08203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0171,
+ "step": 15824
+ },
+ {
+ "epoch": 50.5591054313099,
+ "grad_norm": 0.08642578125,
+ "learning_rate": 0.0005,
+ "loss": 1.03,
+ "step": 15825
+ },
+ {
+ "epoch": 50.56230031948882,
+ "grad_norm": 0.1220703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0347,
+ "step": 15826
+ },
+ {
+ "epoch": 50.56549520766773,
+ "grad_norm": 0.05419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0291,
+ "step": 15827
+ },
+ {
+ "epoch": 50.56869009584665,
+ "grad_norm": 0.053955078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0357,
+ "step": 15828
+ },
+ {
+ "epoch": 50.57188498402556,
+ "grad_norm": 0.05322265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0213,
+ "step": 15829
+ },
+ {
+ "epoch": 50.575079872204476,
+ "grad_norm": 0.0478515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0316,
+ "step": 15830
+ },
+ {
+ "epoch": 50.57827476038339,
+ "grad_norm": 0.057373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0282,
+ "step": 15831
+ },
+ {
+ "epoch": 50.5814696485623,
+ "grad_norm": 0.05859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0247,
+ "step": 15832
+ },
+ {
+ "epoch": 50.584664536741215,
+ "grad_norm": 0.05126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0257,
+ "step": 15833
+ },
+ {
+ "epoch": 50.587859424920126,
+ "grad_norm": 0.048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0281,
+ "step": 15834
+ },
+ {
+ "epoch": 50.59105431309904,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.016,
+ "step": 15835
+ },
+ {
+ "epoch": 50.594249201277954,
+ "grad_norm": 0.08984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0239,
+ "step": 15836
+ },
+ {
+ "epoch": 50.59744408945687,
+ "grad_norm": 0.0517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0337,
+ "step": 15837
+ },
+ {
+ "epoch": 50.60063897763578,
+ "grad_norm": 0.08642578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0275,
+ "step": 15838
+ },
+ {
+ "epoch": 50.6038338658147,
+ "grad_norm": 0.051513671875,
+ "learning_rate": 0.0005,
+ "loss": 1.034,
+ "step": 15839
+ },
+ {
+ "epoch": 50.60702875399361,
+ "grad_norm": 0.0556640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0319,
+ "step": 15840
+ },
+ {
+ "epoch": 50.61022364217252,
+ "grad_norm": 0.05322265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0272,
+ "step": 15841
+ },
+ {
+ "epoch": 50.61341853035144,
+ "grad_norm": 0.04345703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0278,
+ "step": 15842
+ },
+ {
+ "epoch": 50.61661341853035,
+ "grad_norm": 0.05615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0266,
+ "step": 15843
+ },
+ {
+ "epoch": 50.61980830670927,
+ "grad_norm": 0.055908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.024,
+ "step": 15844
+ },
+ {
+ "epoch": 50.62300319488818,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0221,
+ "step": 15845
+ },
+ {
+ "epoch": 50.626198083067095,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0239,
+ "step": 15846
+ },
+ {
+ "epoch": 50.629392971246006,
+ "grad_norm": 0.0556640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0371,
+ "step": 15847
+ },
+ {
+ "epoch": 50.63258785942492,
+ "grad_norm": 0.0546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0243,
+ "step": 15848
+ },
+ {
+ "epoch": 50.635782747603834,
+ "grad_norm": 0.053466796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0178,
+ "step": 15849
+ },
+ {
+ "epoch": 50.638977635782744,
+ "grad_norm": 0.049560546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0323,
+ "step": 15850
+ },
+ {
+ "epoch": 50.64217252396166,
+ "grad_norm": 0.09423828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0187,
+ "step": 15851
+ },
+ {
+ "epoch": 50.64536741214057,
+ "grad_norm": 0.08203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0196,
+ "step": 15852
+ },
+ {
+ "epoch": 50.64856230031949,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0353,
+ "step": 15853
+ },
+ {
+ "epoch": 50.6517571884984,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0196,
+ "step": 15854
+ },
+ {
+ "epoch": 50.65495207667732,
+ "grad_norm": 0.060302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.025,
+ "step": 15855
+ },
+ {
+ "epoch": 50.65814696485623,
+ "grad_norm": 0.051513671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0277,
+ "step": 15856
+ },
+ {
+ "epoch": 50.66134185303515,
+ "grad_norm": 0.050537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0297,
+ "step": 15857
+ },
+ {
+ "epoch": 50.66453674121406,
+ "grad_norm": 0.046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0385,
+ "step": 15858
+ },
+ {
+ "epoch": 50.66773162939297,
+ "grad_norm": 0.06005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0341,
+ "step": 15859
+ },
+ {
+ "epoch": 50.670926517571885,
+ "grad_norm": 0.057861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.021,
+ "step": 15860
+ },
+ {
+ "epoch": 50.674121405750796,
+ "grad_norm": 0.0966796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0444,
+ "step": 15861
+ },
+ {
+ "epoch": 50.677316293929714,
+ "grad_norm": 0.042724609375,
+ "learning_rate": 0.0005,
+ "loss": 1.018,
+ "step": 15862
+ },
+ {
+ "epoch": 50.680511182108624,
+ "grad_norm": 0.12255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0368,
+ "step": 15863
+ },
+ {
+ "epoch": 50.68370607028754,
+ "grad_norm": 0.044677734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0224,
+ "step": 15864
+ },
+ {
+ "epoch": 50.68690095846645,
+ "grad_norm": 0.12060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0304,
+ "step": 15865
+ },
+ {
+ "epoch": 50.69009584664537,
+ "grad_norm": 0.05078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0224,
+ "step": 15866
+ },
+ {
+ "epoch": 50.69329073482428,
+ "grad_norm": 0.126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0237,
+ "step": 15867
+ },
+ {
+ "epoch": 50.6964856230032,
+ "grad_norm": 0.047607421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0239,
+ "step": 15868
+ },
+ {
+ "epoch": 50.69968051118211,
+ "grad_norm": 0.1171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0257,
+ "step": 15869
+ },
+ {
+ "epoch": 50.70287539936102,
+ "grad_norm": 0.055908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0266,
+ "step": 15870
+ },
+ {
+ "epoch": 50.70607028753994,
+ "grad_norm": 0.107421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0348,
+ "step": 15871
+ },
+ {
+ "epoch": 50.70926517571885,
+ "grad_norm": 0.051025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0345,
+ "step": 15872
+ },
+ {
+ "epoch": 50.712460063897765,
+ "grad_norm": 0.1005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0234,
+ "step": 15873
+ },
+ {
+ "epoch": 50.715654952076676,
+ "grad_norm": 0.060302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0251,
+ "step": 15874
+ },
+ {
+ "epoch": 50.718849840255594,
+ "grad_norm": 0.11572265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0319,
+ "step": 15875
+ },
+ {
+ "epoch": 50.722044728434504,
+ "grad_norm": 0.049560546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0347,
+ "step": 15876
+ },
+ {
+ "epoch": 50.72523961661342,
+ "grad_norm": 0.12890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0312,
+ "step": 15877
+ },
+ {
+ "epoch": 50.72843450479233,
+ "grad_norm": 0.09423828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0271,
+ "step": 15878
+ },
+ {
+ "epoch": 50.73162939297124,
+ "grad_norm": 0.125,
+ "learning_rate": 0.0005,
+ "loss": 1.0086,
+ "step": 15879
+ },
+ {
+ "epoch": 50.73482428115016,
+ "grad_norm": 0.061767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0312,
+ "step": 15880
+ },
+ {
+ "epoch": 50.73801916932907,
+ "grad_norm": 0.1513671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0313,
+ "step": 15881
+ },
+ {
+ "epoch": 50.74121405750799,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0285,
+ "step": 15882
+ },
+ {
+ "epoch": 50.7444089456869,
+ "grad_norm": 0.059814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0313,
+ "step": 15883
+ },
+ {
+ "epoch": 50.74760383386582,
+ "grad_norm": 0.083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0372,
+ "step": 15884
+ },
+ {
+ "epoch": 50.75079872204473,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0206,
+ "step": 15885
+ },
+ {
+ "epoch": 50.753993610223645,
+ "grad_norm": 0.1328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0272,
+ "step": 15886
+ },
+ {
+ "epoch": 50.757188498402556,
+ "grad_norm": 0.123046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0216,
+ "step": 15887
+ },
+ {
+ "epoch": 50.760383386581466,
+ "grad_norm": 0.103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0199,
+ "step": 15888
+ },
+ {
+ "epoch": 50.763578274760384,
+ "grad_norm": 0.10791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0195,
+ "step": 15889
+ },
+ {
+ "epoch": 50.766773162939295,
+ "grad_norm": 0.1328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0317,
+ "step": 15890
+ },
+ {
+ "epoch": 50.76996805111821,
+ "grad_norm": 0.126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0347,
+ "step": 15891
+ },
+ {
+ "epoch": 50.77316293929712,
+ "grad_norm": 0.09326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0342,
+ "step": 15892
+ },
+ {
+ "epoch": 50.77635782747604,
+ "grad_norm": 0.1298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0272,
+ "step": 15893
+ },
+ {
+ "epoch": 50.77955271565495,
+ "grad_norm": 0.054443359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0333,
+ "step": 15894
+ },
+ {
+ "epoch": 50.78274760383387,
+ "grad_norm": 0.10595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0285,
+ "step": 15895
+ },
+ {
+ "epoch": 50.78594249201278,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0307,
+ "step": 15896
+ },
+ {
+ "epoch": 50.78913738019169,
+ "grad_norm": 0.0859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0248,
+ "step": 15897
+ },
+ {
+ "epoch": 50.79233226837061,
+ "grad_norm": 0.0888671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0252,
+ "step": 15898
+ },
+ {
+ "epoch": 50.79552715654952,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0432,
+ "step": 15899
+ },
+ {
+ "epoch": 50.798722044728436,
+ "grad_norm": 0.08349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.046,
+ "step": 15900
+ },
+ {
+ "epoch": 50.801916932907346,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0351,
+ "step": 15901
+ },
+ {
+ "epoch": 50.805111821086264,
+ "grad_norm": 0.0849609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0213,
+ "step": 15902
+ },
+ {
+ "epoch": 50.808306709265175,
+ "grad_norm": 0.0849609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0248,
+ "step": 15903
+ },
+ {
+ "epoch": 50.81150159744409,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0367,
+ "step": 15904
+ },
+ {
+ "epoch": 50.814696485623,
+ "grad_norm": 0.095703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0304,
+ "step": 15905
+ },
+ {
+ "epoch": 50.81789137380191,
+ "grad_norm": 0.061279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.042,
+ "step": 15906
+ },
+ {
+ "epoch": 50.82108626198083,
+ "grad_norm": 0.049560546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0271,
+ "step": 15907
+ },
+ {
+ "epoch": 50.82428115015974,
+ "grad_norm": 0.061279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0312,
+ "step": 15908
+ },
+ {
+ "epoch": 50.82747603833866,
+ "grad_norm": 0.044677734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0205,
+ "step": 15909
+ },
+ {
+ "epoch": 50.83067092651757,
+ "grad_norm": 0.057861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0324,
+ "step": 15910
+ },
+ {
+ "epoch": 50.83386581469649,
+ "grad_norm": 0.057861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0236,
+ "step": 15911
+ },
+ {
+ "epoch": 50.8370607028754,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0291,
+ "step": 15912
+ },
+ {
+ "epoch": 50.840255591054316,
+ "grad_norm": 0.0546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0334,
+ "step": 15913
+ },
+ {
+ "epoch": 50.843450479233226,
+ "grad_norm": 0.059814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0311,
+ "step": 15914
+ },
+ {
+ "epoch": 50.846645367412144,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0221,
+ "step": 15915
+ },
+ {
+ "epoch": 50.849840255591054,
+ "grad_norm": 0.060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0364,
+ "step": 15916
+ },
+ {
+ "epoch": 50.853035143769965,
+ "grad_norm": 0.04931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0381,
+ "step": 15917
+ },
+ {
+ "epoch": 50.85623003194888,
+ "grad_norm": 0.0517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0332,
+ "step": 15918
+ },
+ {
+ "epoch": 50.85942492012779,
+ "grad_norm": 0.0556640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0378,
+ "step": 15919
+ },
+ {
+ "epoch": 50.86261980830671,
+ "grad_norm": 0.057861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.042,
+ "step": 15920
+ },
+ {
+ "epoch": 50.86581469648562,
+ "grad_norm": 0.058349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0417,
+ "step": 15921
+ },
+ {
+ "epoch": 50.86900958466454,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0266,
+ "step": 15922
+ },
+ {
+ "epoch": 50.87220447284345,
+ "grad_norm": 0.043212890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0223,
+ "step": 15923
+ },
+ {
+ "epoch": 50.87539936102237,
+ "grad_norm": 0.056884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0287,
+ "step": 15924
+ },
+ {
+ "epoch": 50.87859424920128,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0281,
+ "step": 15925
+ },
+ {
+ "epoch": 50.88178913738019,
+ "grad_norm": 0.050537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0305,
+ "step": 15926
+ },
+ {
+ "epoch": 50.884984025559106,
+ "grad_norm": 0.051025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0292,
+ "step": 15927
+ },
+ {
+ "epoch": 50.88817891373802,
+ "grad_norm": 0.046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0244,
+ "step": 15928
+ },
+ {
+ "epoch": 50.891373801916934,
+ "grad_norm": 0.04541015625,
+ "learning_rate": 0.0005,
+ "loss": 1.021,
+ "step": 15929
+ },
+ {
+ "epoch": 50.894568690095845,
+ "grad_norm": 0.044677734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0071,
+ "step": 15930
+ },
+ {
+ "epoch": 50.89776357827476,
+ "grad_norm": 0.052734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0325,
+ "step": 15931
+ },
+ {
+ "epoch": 50.90095846645367,
+ "grad_norm": 0.051025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.027,
+ "step": 15932
+ },
+ {
+ "epoch": 50.90415335463259,
+ "grad_norm": 0.05224609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0239,
+ "step": 15933
+ },
+ {
+ "epoch": 50.9073482428115,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0252,
+ "step": 15934
+ },
+ {
+ "epoch": 50.91054313099041,
+ "grad_norm": 0.05126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0357,
+ "step": 15935
+ },
+ {
+ "epoch": 50.91373801916933,
+ "grad_norm": 0.05419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0365,
+ "step": 15936
+ },
+ {
+ "epoch": 50.91693290734824,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0321,
+ "step": 15937
+ },
+ {
+ "epoch": 50.92012779552716,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.019,
+ "step": 15938
+ },
+ {
+ "epoch": 50.92332268370607,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0162,
+ "step": 15939
+ },
+ {
+ "epoch": 50.926517571884986,
+ "grad_norm": 0.061279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0318,
+ "step": 15940
+ },
+ {
+ "epoch": 50.9297124600639,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0362,
+ "step": 15941
+ },
+ {
+ "epoch": 50.932907348242814,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0254,
+ "step": 15942
+ },
+ {
+ "epoch": 50.936102236421725,
+ "grad_norm": 0.0908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0234,
+ "step": 15943
+ },
+ {
+ "epoch": 50.93929712460064,
+ "grad_norm": 0.05322265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0233,
+ "step": 15944
+ },
+ {
+ "epoch": 50.94249201277955,
+ "grad_norm": 0.059326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0351,
+ "step": 15945
+ },
+ {
+ "epoch": 50.945686900958464,
+ "grad_norm": 0.054931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0425,
+ "step": 15946
+ },
+ {
+ "epoch": 50.94888178913738,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0309,
+ "step": 15947
+ },
+ {
+ "epoch": 50.95207667731629,
+ "grad_norm": 0.05078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0317,
+ "step": 15948
+ },
+ {
+ "epoch": 50.95527156549521,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0388,
+ "step": 15949
+ },
+ {
+ "epoch": 50.95846645367412,
+ "grad_norm": 0.060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.026,
+ "step": 15950
+ },
+ {
+ "epoch": 50.96166134185304,
+ "grad_norm": 0.11083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.014,
+ "step": 15951
+ },
+ {
+ "epoch": 50.96485623003195,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0325,
+ "step": 15952
+ },
+ {
+ "epoch": 50.968051118210866,
+ "grad_norm": 0.0859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0255,
+ "step": 15953
+ },
+ {
+ "epoch": 50.97124600638978,
+ "grad_norm": 0.05419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0161,
+ "step": 15954
+ },
+ {
+ "epoch": 50.97444089456869,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0321,
+ "step": 15955
+ },
+ {
+ "epoch": 50.977635782747605,
+ "grad_norm": 0.059814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0328,
+ "step": 15956
+ },
+ {
+ "epoch": 50.980830670926515,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0256,
+ "step": 15957
+ },
+ {
+ "epoch": 50.98402555910543,
+ "grad_norm": 0.056884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0219,
+ "step": 15958
+ },
+ {
+ "epoch": 50.98722044728434,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0294,
+ "step": 15959
+ },
+ {
+ "epoch": 50.99041533546326,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0202,
+ "step": 15960
+ },
+ {
+ "epoch": 50.99361022364217,
+ "grad_norm": 0.07861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0408,
+ "step": 15961
+ },
+ {
+ "epoch": 50.99680511182109,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0389,
+ "step": 15962
+ },
+ {
+ "epoch": 51.0,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0308,
+ "step": 15963
+ },
+ {
+ "epoch": 51.00319488817891,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0302,
+ "step": 15964
+ },
+ {
+ "epoch": 51.00638977635783,
+ "grad_norm": 0.09228515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0168,
+ "step": 15965
+ },
+ {
+ "epoch": 51.00958466453674,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0242,
+ "step": 15966
+ },
+ {
+ "epoch": 51.01277955271566,
+ "grad_norm": 0.05712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0222,
+ "step": 15967
+ },
+ {
+ "epoch": 51.01597444089457,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0287,
+ "step": 15968
+ },
+ {
+ "epoch": 51.019169329073485,
+ "grad_norm": 0.057373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.024,
+ "step": 15969
+ },
+ {
+ "epoch": 51.022364217252395,
+ "grad_norm": 0.052978515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0224,
+ "step": 15970
+ },
+ {
+ "epoch": 51.02555910543131,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0284,
+ "step": 15971
+ },
+ {
+ "epoch": 51.02875399361022,
+ "grad_norm": 0.0517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0347,
+ "step": 15972
+ },
+ {
+ "epoch": 51.031948881789134,
+ "grad_norm": 0.06201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0249,
+ "step": 15973
+ },
+ {
+ "epoch": 51.03514376996805,
+ "grad_norm": 0.060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0264,
+ "step": 15974
+ },
+ {
+ "epoch": 51.03833865814696,
+ "grad_norm": 0.1416015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0284,
+ "step": 15975
+ },
+ {
+ "epoch": 51.04153354632588,
+ "grad_norm": 0.05712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0292,
+ "step": 15976
+ },
+ {
+ "epoch": 51.04472843450479,
+ "grad_norm": 0.053466796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0238,
+ "step": 15977
+ },
+ {
+ "epoch": 51.04792332268371,
+ "grad_norm": 0.1005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0067,
+ "step": 15978
+ },
+ {
+ "epoch": 51.05111821086262,
+ "grad_norm": 0.058837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0308,
+ "step": 15979
+ },
+ {
+ "epoch": 51.054313099041536,
+ "grad_norm": 0.0966796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0242,
+ "step": 15980
+ },
+ {
+ "epoch": 51.05750798722045,
+ "grad_norm": 0.0595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0164,
+ "step": 15981
+ },
+ {
+ "epoch": 51.06070287539936,
+ "grad_norm": 0.052001953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0064,
+ "step": 15982
+ },
+ {
+ "epoch": 51.063897763578275,
+ "grad_norm": 0.05712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0265,
+ "step": 15983
+ },
+ {
+ "epoch": 51.067092651757186,
+ "grad_norm": 0.0517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0239,
+ "step": 15984
+ },
+ {
+ "epoch": 51.0702875399361,
+ "grad_norm": 0.0595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0209,
+ "step": 15985
+ },
+ {
+ "epoch": 51.073482428115014,
+ "grad_norm": 0.047119140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0277,
+ "step": 15986
+ },
+ {
+ "epoch": 51.07667731629393,
+ "grad_norm": 0.047119140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0296,
+ "step": 15987
+ },
+ {
+ "epoch": 51.07987220447284,
+ "grad_norm": 0.048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0265,
+ "step": 15988
+ },
+ {
+ "epoch": 51.08306709265176,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0272,
+ "step": 15989
+ },
+ {
+ "epoch": 51.08626198083067,
+ "grad_norm": 0.057373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0357,
+ "step": 15990
+ },
+ {
+ "epoch": 51.08945686900959,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0264,
+ "step": 15991
+ },
+ {
+ "epoch": 51.0926517571885,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0158,
+ "step": 15992
+ },
+ {
+ "epoch": 51.09584664536741,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0321,
+ "step": 15993
+ },
+ {
+ "epoch": 51.09904153354633,
+ "grad_norm": 0.0849609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0226,
+ "step": 15994
+ },
+ {
+ "epoch": 51.10223642172524,
+ "grad_norm": 0.09228515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0331,
+ "step": 15995
+ },
+ {
+ "epoch": 51.105431309904155,
+ "grad_norm": 0.05224609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0318,
+ "step": 15996
+ },
+ {
+ "epoch": 51.108626198083066,
+ "grad_norm": 0.08642578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0256,
+ "step": 15997
+ },
+ {
+ "epoch": 51.11182108626198,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0393,
+ "step": 15998
+ },
+ {
+ "epoch": 51.115015974440894,
+ "grad_norm": 0.055908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0225,
+ "step": 15999
+ },
+ {
+ "epoch": 51.11821086261981,
+ "grad_norm": 0.11572265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0297,
+ "step": 16000
+ },
+ {
+ "epoch": 51.12140575079872,
+ "grad_norm": 0.10595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.017,
+ "step": 16001
+ },
+ {
+ "epoch": 51.12460063897763,
+ "grad_norm": 0.15625,
+ "learning_rate": 0.0005,
+ "loss": 1.0306,
+ "step": 16002
+ },
+ {
+ "epoch": 51.12779552715655,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.032,
+ "step": 16003
+ },
+ {
+ "epoch": 51.13099041533546,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.023,
+ "step": 16004
+ },
+ {
+ "epoch": 51.13418530351438,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.044,
+ "step": 16005
+ },
+ {
+ "epoch": 51.13738019169329,
+ "grad_norm": 0.09521484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0304,
+ "step": 16006
+ },
+ {
+ "epoch": 51.14057507987221,
+ "grad_norm": 0.05078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0347,
+ "step": 16007
+ },
+ {
+ "epoch": 51.14376996805112,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0393,
+ "step": 16008
+ },
+ {
+ "epoch": 51.146964856230035,
+ "grad_norm": 0.060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0378,
+ "step": 16009
+ },
+ {
+ "epoch": 51.150159744408946,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0328,
+ "step": 16010
+ },
+ {
+ "epoch": 51.153354632587856,
+ "grad_norm": 0.05322265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0195,
+ "step": 16011
+ },
+ {
+ "epoch": 51.156549520766774,
+ "grad_norm": 0.05908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.028,
+ "step": 16012
+ },
+ {
+ "epoch": 51.159744408945684,
+ "grad_norm": 0.08740234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0269,
+ "step": 16013
+ },
+ {
+ "epoch": 51.1629392971246,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.03,
+ "step": 16014
+ },
+ {
+ "epoch": 51.16613418530351,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0382,
+ "step": 16015
+ },
+ {
+ "epoch": 51.16932907348243,
+ "grad_norm": 0.05126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0211,
+ "step": 16016
+ },
+ {
+ "epoch": 51.17252396166134,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0166,
+ "step": 16017
+ },
+ {
+ "epoch": 51.17571884984026,
+ "grad_norm": 0.044189453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0293,
+ "step": 16018
+ },
+ {
+ "epoch": 51.17891373801917,
+ "grad_norm": 0.05810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0339,
+ "step": 16019
+ },
+ {
+ "epoch": 51.18210862619808,
+ "grad_norm": 0.09130859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0206,
+ "step": 16020
+ },
+ {
+ "epoch": 51.185303514377,
+ "grad_norm": 0.05029296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0239,
+ "step": 16021
+ },
+ {
+ "epoch": 51.18849840255591,
+ "grad_norm": 0.0869140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0335,
+ "step": 16022
+ },
+ {
+ "epoch": 51.191693290734825,
+ "grad_norm": 0.058837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.028,
+ "step": 16023
+ },
+ {
+ "epoch": 51.194888178913736,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0344,
+ "step": 16024
+ },
+ {
+ "epoch": 51.198083067092654,
+ "grad_norm": 0.0537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0219,
+ "step": 16025
+ },
+ {
+ "epoch": 51.201277955271564,
+ "grad_norm": 0.056396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0344,
+ "step": 16026
+ },
+ {
+ "epoch": 51.20447284345048,
+ "grad_norm": 0.060302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0339,
+ "step": 16027
+ },
+ {
+ "epoch": 51.20766773162939,
+ "grad_norm": 0.0595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0268,
+ "step": 16028
+ },
+ {
+ "epoch": 51.21086261980831,
+ "grad_norm": 0.04931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0194,
+ "step": 16029
+ },
+ {
+ "epoch": 51.21405750798722,
+ "grad_norm": 0.09814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0219,
+ "step": 16030
+ },
+ {
+ "epoch": 51.21725239616613,
+ "grad_norm": 0.047119140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0261,
+ "step": 16031
+ },
+ {
+ "epoch": 51.22044728434505,
+ "grad_norm": 0.08203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0392,
+ "step": 16032
+ },
+ {
+ "epoch": 51.22364217252396,
+ "grad_norm": 0.07861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0263,
+ "step": 16033
+ },
+ {
+ "epoch": 51.22683706070288,
+ "grad_norm": 0.06201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0414,
+ "step": 16034
+ },
+ {
+ "epoch": 51.23003194888179,
+ "grad_norm": 0.056396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0339,
+ "step": 16035
+ },
+ {
+ "epoch": 51.233226837060705,
+ "grad_norm": 0.0517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0201,
+ "step": 16036
+ },
+ {
+ "epoch": 51.236421725239616,
+ "grad_norm": 0.047119140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0197,
+ "step": 16037
+ },
+ {
+ "epoch": 51.239616613418534,
+ "grad_norm": 0.11083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0296,
+ "step": 16038
+ },
+ {
+ "epoch": 51.242811501597444,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.03,
+ "step": 16039
+ },
+ {
+ "epoch": 51.246006389776355,
+ "grad_norm": 0.05810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.033,
+ "step": 16040
+ },
+ {
+ "epoch": 51.24920127795527,
+ "grad_norm": 0.05419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0303,
+ "step": 16041
+ },
+ {
+ "epoch": 51.25239616613418,
+ "grad_norm": 0.052001953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0298,
+ "step": 16042
+ },
+ {
+ "epoch": 51.2555910543131,
+ "grad_norm": 0.051513671875,
+ "learning_rate": 0.0005,
+ "loss": 1.031,
+ "step": 16043
+ },
+ {
+ "epoch": 51.25878594249201,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0143,
+ "step": 16044
+ },
+ {
+ "epoch": 51.26198083067093,
+ "grad_norm": 0.056396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0484,
+ "step": 16045
+ },
+ {
+ "epoch": 51.26517571884984,
+ "grad_norm": 0.052978515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0152,
+ "step": 16046
+ },
+ {
+ "epoch": 51.26837060702876,
+ "grad_norm": 0.09619140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0253,
+ "step": 16047
+ },
+ {
+ "epoch": 51.27156549520767,
+ "grad_norm": 0.049560546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0257,
+ "step": 16048
+ },
+ {
+ "epoch": 51.27476038338658,
+ "grad_norm": 0.08056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0218,
+ "step": 16049
+ },
+ {
+ "epoch": 51.277955271565496,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0369,
+ "step": 16050
+ },
+ {
+ "epoch": 51.281150159744406,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0265,
+ "step": 16051
+ },
+ {
+ "epoch": 51.284345047923324,
+ "grad_norm": 0.08837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0285,
+ "step": 16052
+ },
+ {
+ "epoch": 51.287539936102235,
+ "grad_norm": 0.0849609375,
+ "learning_rate": 0.0005,
+ "loss": 1.029,
+ "step": 16053
+ },
+ {
+ "epoch": 51.29073482428115,
+ "grad_norm": 0.080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0322,
+ "step": 16054
+ },
+ {
+ "epoch": 51.29392971246006,
+ "grad_norm": 0.09619140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0273,
+ "step": 16055
+ },
+ {
+ "epoch": 51.29712460063898,
+ "grad_norm": 0.0849609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0283,
+ "step": 16056
+ },
+ {
+ "epoch": 51.30031948881789,
+ "grad_norm": 0.1044921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0337,
+ "step": 16057
+ },
+ {
+ "epoch": 51.3035143769968,
+ "grad_norm": 0.09326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0282,
+ "step": 16058
+ },
+ {
+ "epoch": 51.30670926517572,
+ "grad_norm": 0.10009765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0321,
+ "step": 16059
+ },
+ {
+ "epoch": 51.30990415335463,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0333,
+ "step": 16060
+ },
+ {
+ "epoch": 51.31309904153355,
+ "grad_norm": 0.09716796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0258,
+ "step": 16061
+ },
+ {
+ "epoch": 51.31629392971246,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0296,
+ "step": 16062
+ },
+ {
+ "epoch": 51.319488817891376,
+ "grad_norm": 0.10009765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0332,
+ "step": 16063
+ },
+ {
+ "epoch": 51.322683706070286,
+ "grad_norm": 0.056884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0175,
+ "step": 16064
+ },
+ {
+ "epoch": 51.325878594249204,
+ "grad_norm": 0.087890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0279,
+ "step": 16065
+ },
+ {
+ "epoch": 51.329073482428115,
+ "grad_norm": 0.05224609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0303,
+ "step": 16066
+ },
+ {
+ "epoch": 51.33226837060703,
+ "grad_norm": 0.11962890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0383,
+ "step": 16067
+ },
+ {
+ "epoch": 51.33546325878594,
+ "grad_norm": 0.1259765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0307,
+ "step": 16068
+ },
+ {
+ "epoch": 51.33865814696485,
+ "grad_norm": 0.28125,
+ "learning_rate": 0.0005,
+ "loss": 1.0272,
+ "step": 16069
+ },
+ {
+ "epoch": 51.34185303514377,
+ "grad_norm": 0.1162109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0283,
+ "step": 16070
+ },
+ {
+ "epoch": 51.34504792332268,
+ "grad_norm": 0.154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0305,
+ "step": 16071
+ },
+ {
+ "epoch": 51.3482428115016,
+ "grad_norm": 0.1220703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0378,
+ "step": 16072
+ },
+ {
+ "epoch": 51.35143769968051,
+ "grad_norm": 0.08447265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0285,
+ "step": 16073
+ },
+ {
+ "epoch": 51.35463258785943,
+ "grad_norm": 0.0986328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0199,
+ "step": 16074
+ },
+ {
+ "epoch": 51.35782747603834,
+ "grad_norm": 0.0830078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0249,
+ "step": 16075
+ },
+ {
+ "epoch": 51.361022364217256,
+ "grad_norm": 0.08935546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0294,
+ "step": 16076
+ },
+ {
+ "epoch": 51.364217252396166,
+ "grad_norm": 0.125,
+ "learning_rate": 0.0005,
+ "loss": 1.0205,
+ "step": 16077
+ },
+ {
+ "epoch": 51.36741214057508,
+ "grad_norm": 0.10302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0258,
+ "step": 16078
+ },
+ {
+ "epoch": 51.370607028753994,
+ "grad_norm": 0.142578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0254,
+ "step": 16079
+ },
+ {
+ "epoch": 51.373801916932905,
+ "grad_norm": 0.095703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0362,
+ "step": 16080
+ },
+ {
+ "epoch": 51.37699680511182,
+ "grad_norm": 0.1171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0277,
+ "step": 16081
+ },
+ {
+ "epoch": 51.38019169329073,
+ "grad_norm": 0.1015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0222,
+ "step": 16082
+ },
+ {
+ "epoch": 51.38338658146965,
+ "grad_norm": 0.0888671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0237,
+ "step": 16083
+ },
+ {
+ "epoch": 51.38658146964856,
+ "grad_norm": 0.08154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.022,
+ "step": 16084
+ },
+ {
+ "epoch": 51.38977635782748,
+ "grad_norm": 0.08203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0134,
+ "step": 16085
+ },
+ {
+ "epoch": 51.39297124600639,
+ "grad_norm": 0.0615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0273,
+ "step": 16086
+ },
+ {
+ "epoch": 51.3961661341853,
+ "grad_norm": 0.08447265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0215,
+ "step": 16087
+ },
+ {
+ "epoch": 51.39936102236422,
+ "grad_norm": 0.059326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0364,
+ "step": 16088
+ },
+ {
+ "epoch": 51.40255591054313,
+ "grad_norm": 0.08056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0306,
+ "step": 16089
+ },
+ {
+ "epoch": 51.405750798722046,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0246,
+ "step": 16090
+ },
+ {
+ "epoch": 51.40894568690096,
+ "grad_norm": 0.060791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.026,
+ "step": 16091
+ },
+ {
+ "epoch": 51.412140575079874,
+ "grad_norm": 0.10498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0276,
+ "step": 16092
+ },
+ {
+ "epoch": 51.415335463258785,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0384,
+ "step": 16093
+ },
+ {
+ "epoch": 51.4185303514377,
+ "grad_norm": 0.0615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0367,
+ "step": 16094
+ },
+ {
+ "epoch": 51.42172523961661,
+ "grad_norm": 0.056884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0321,
+ "step": 16095
+ },
+ {
+ "epoch": 51.424920127795524,
+ "grad_norm": 0.059326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0298,
+ "step": 16096
+ },
+ {
+ "epoch": 51.42811501597444,
+ "grad_norm": 0.05810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.021,
+ "step": 16097
+ },
+ {
+ "epoch": 51.43130990415335,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0352,
+ "step": 16098
+ },
+ {
+ "epoch": 51.43450479233227,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0297,
+ "step": 16099
+ },
+ {
+ "epoch": 51.43769968051118,
+ "grad_norm": 0.05126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0306,
+ "step": 16100
+ },
+ {
+ "epoch": 51.4408945686901,
+ "grad_norm": 0.05126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0295,
+ "step": 16101
+ },
+ {
+ "epoch": 51.44408945686901,
+ "grad_norm": 0.05712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.026,
+ "step": 16102
+ },
+ {
+ "epoch": 51.447284345047926,
+ "grad_norm": 0.054931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0224,
+ "step": 16103
+ },
+ {
+ "epoch": 51.45047923322684,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.027,
+ "step": 16104
+ },
+ {
+ "epoch": 51.453674121405754,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0265,
+ "step": 16105
+ },
+ {
+ "epoch": 51.456869009584665,
+ "grad_norm": 0.08154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0206,
+ "step": 16106
+ },
+ {
+ "epoch": 51.460063897763575,
+ "grad_norm": 0.06201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0258,
+ "step": 16107
+ },
+ {
+ "epoch": 51.46325878594249,
+ "grad_norm": 0.080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0173,
+ "step": 16108
+ },
+ {
+ "epoch": 51.466453674121404,
+ "grad_norm": 0.10009765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0182,
+ "step": 16109
+ },
+ {
+ "epoch": 51.46964856230032,
+ "grad_norm": 0.0556640625,
+ "learning_rate": 0.0005,
+ "loss": 1.035,
+ "step": 16110
+ },
+ {
+ "epoch": 51.47284345047923,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.044,
+ "step": 16111
+ },
+ {
+ "epoch": 51.47603833865815,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0286,
+ "step": 16112
+ },
+ {
+ "epoch": 51.47923322683706,
+ "grad_norm": 0.11328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0285,
+ "step": 16113
+ },
+ {
+ "epoch": 51.48242811501598,
+ "grad_norm": 0.0810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0256,
+ "step": 16114
+ },
+ {
+ "epoch": 51.48562300319489,
+ "grad_norm": 0.0810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.026,
+ "step": 16115
+ },
+ {
+ "epoch": 51.4888178913738,
+ "grad_norm": 0.07861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0218,
+ "step": 16116
+ },
+ {
+ "epoch": 51.49201277955272,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0352,
+ "step": 16117
+ },
+ {
+ "epoch": 51.49520766773163,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0231,
+ "step": 16118
+ },
+ {
+ "epoch": 51.498402555910545,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0283,
+ "step": 16119
+ },
+ {
+ "epoch": 51.501597444089455,
+ "grad_norm": 0.08740234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0293,
+ "step": 16120
+ },
+ {
+ "epoch": 51.50479233226837,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0325,
+ "step": 16121
+ },
+ {
+ "epoch": 51.50798722044728,
+ "grad_norm": 0.052978515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0353,
+ "step": 16122
+ },
+ {
+ "epoch": 51.5111821086262,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0226,
+ "step": 16123
+ },
+ {
+ "epoch": 51.51437699680511,
+ "grad_norm": 0.052490234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0155,
+ "step": 16124
+ },
+ {
+ "epoch": 51.51757188498402,
+ "grad_norm": 0.056396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0275,
+ "step": 16125
+ },
+ {
+ "epoch": 51.52076677316294,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0373,
+ "step": 16126
+ },
+ {
+ "epoch": 51.52396166134185,
+ "grad_norm": 0.054443359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0237,
+ "step": 16127
+ },
+ {
+ "epoch": 51.52715654952077,
+ "grad_norm": 0.10205078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0333,
+ "step": 16128
+ },
+ {
+ "epoch": 51.53035143769968,
+ "grad_norm": 0.0947265625,
+ "learning_rate": 0.0005,
+ "loss": 1.034,
+ "step": 16129
+ },
+ {
+ "epoch": 51.533546325878596,
+ "grad_norm": 0.0576171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0441,
+ "step": 16130
+ },
+ {
+ "epoch": 51.53674121405751,
+ "grad_norm": 0.06005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0345,
+ "step": 16131
+ },
+ {
+ "epoch": 51.539936102236425,
+ "grad_norm": 0.046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0172,
+ "step": 16132
+ },
+ {
+ "epoch": 51.543130990415335,
+ "grad_norm": 0.052490234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0265,
+ "step": 16133
+ },
+ {
+ "epoch": 51.546325878594246,
+ "grad_norm": 0.051025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0304,
+ "step": 16134
+ },
+ {
+ "epoch": 51.54952076677316,
+ "grad_norm": 0.05712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0217,
+ "step": 16135
+ },
+ {
+ "epoch": 51.552715654952074,
+ "grad_norm": 0.05126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0357,
+ "step": 16136
+ },
+ {
+ "epoch": 51.55591054313099,
+ "grad_norm": 0.04931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0249,
+ "step": 16137
+ },
+ {
+ "epoch": 51.5591054313099,
+ "grad_norm": 0.0615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0243,
+ "step": 16138
+ },
+ {
+ "epoch": 51.56230031948882,
+ "grad_norm": 0.056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0155,
+ "step": 16139
+ },
+ {
+ "epoch": 51.56549520766773,
+ "grad_norm": 0.056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0256,
+ "step": 16140
+ },
+ {
+ "epoch": 51.56869009584665,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0194,
+ "step": 16141
+ },
+ {
+ "epoch": 51.57188498402556,
+ "grad_norm": 0.052978515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0241,
+ "step": 16142
+ },
+ {
+ "epoch": 51.575079872204476,
+ "grad_norm": 0.1298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0157,
+ "step": 16143
+ },
+ {
+ "epoch": 51.57827476038339,
+ "grad_norm": 0.045654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0314,
+ "step": 16144
+ },
+ {
+ "epoch": 51.5814696485623,
+ "grad_norm": 0.1181640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0232,
+ "step": 16145
+ },
+ {
+ "epoch": 51.584664536741215,
+ "grad_norm": 0.05712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0403,
+ "step": 16146
+ },
+ {
+ "epoch": 51.587859424920126,
+ "grad_norm": 0.115234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0354,
+ "step": 16147
+ },
+ {
+ "epoch": 51.59105431309904,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0297,
+ "step": 16148
+ },
+ {
+ "epoch": 51.594249201277954,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0199,
+ "step": 16149
+ },
+ {
+ "epoch": 51.59744408945687,
+ "grad_norm": 0.0546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0339,
+ "step": 16150
+ },
+ {
+ "epoch": 51.60063897763578,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0152,
+ "step": 16151
+ },
+ {
+ "epoch": 51.6038338658147,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0309,
+ "step": 16152
+ },
+ {
+ "epoch": 51.60702875399361,
+ "grad_norm": 0.107421875,
+ "learning_rate": 0.0005,
+ "loss": 1.012,
+ "step": 16153
+ },
+ {
+ "epoch": 51.61022364217252,
+ "grad_norm": 0.060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0286,
+ "step": 16154
+ },
+ {
+ "epoch": 51.61341853035144,
+ "grad_norm": 0.095703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0343,
+ "step": 16155
+ },
+ {
+ "epoch": 51.61661341853035,
+ "grad_norm": 0.052001953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0319,
+ "step": 16156
+ },
+ {
+ "epoch": 51.61980830670927,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0268,
+ "step": 16157
+ },
+ {
+ "epoch": 51.62300319488818,
+ "grad_norm": 0.05322265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0288,
+ "step": 16158
+ },
+ {
+ "epoch": 51.626198083067095,
+ "grad_norm": 0.09375,
+ "learning_rate": 0.0005,
+ "loss": 1.0278,
+ "step": 16159
+ },
+ {
+ "epoch": 51.629392971246006,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0278,
+ "step": 16160
+ },
+ {
+ "epoch": 51.63258785942492,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0273,
+ "step": 16161
+ },
+ {
+ "epoch": 51.635782747603834,
+ "grad_norm": 0.052001953125,
+ "learning_rate": 0.0005,
+ "loss": 1.027,
+ "step": 16162
+ },
+ {
+ "epoch": 51.638977635782744,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0351,
+ "step": 16163
+ },
+ {
+ "epoch": 51.64217252396166,
+ "grad_norm": 0.054931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0286,
+ "step": 16164
+ },
+ {
+ "epoch": 51.64536741214057,
+ "grad_norm": 0.0546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0178,
+ "step": 16165
+ },
+ {
+ "epoch": 51.64856230031949,
+ "grad_norm": 0.046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0255,
+ "step": 16166
+ },
+ {
+ "epoch": 51.6517571884984,
+ "grad_norm": 0.054931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.031,
+ "step": 16167
+ },
+ {
+ "epoch": 51.65495207667732,
+ "grad_norm": 0.045166015625,
+ "learning_rate": 0.0005,
+ "loss": 1.027,
+ "step": 16168
+ },
+ {
+ "epoch": 51.65814696485623,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0297,
+ "step": 16169
+ },
+ {
+ "epoch": 51.66134185303515,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0291,
+ "step": 16170
+ },
+ {
+ "epoch": 51.66453674121406,
+ "grad_norm": 0.05908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0292,
+ "step": 16171
+ },
+ {
+ "epoch": 51.66773162939297,
+ "grad_norm": 0.04248046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0232,
+ "step": 16172
+ },
+ {
+ "epoch": 51.670926517571885,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0253,
+ "step": 16173
+ },
+ {
+ "epoch": 51.674121405750796,
+ "grad_norm": 0.061767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.022,
+ "step": 16174
+ },
+ {
+ "epoch": 51.677316293929714,
+ "grad_norm": 0.08203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0354,
+ "step": 16175
+ },
+ {
+ "epoch": 51.680511182108624,
+ "grad_norm": 0.05029296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0253,
+ "step": 16176
+ },
+ {
+ "epoch": 51.68370607028754,
+ "grad_norm": 0.048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0216,
+ "step": 16177
+ },
+ {
+ "epoch": 51.68690095846645,
+ "grad_norm": 0.057373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0208,
+ "step": 16178
+ },
+ {
+ "epoch": 51.69009584664537,
+ "grad_norm": 0.056884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0332,
+ "step": 16179
+ },
+ {
+ "epoch": 51.69329073482428,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0295,
+ "step": 16180
+ },
+ {
+ "epoch": 51.6964856230032,
+ "grad_norm": 0.049560546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0331,
+ "step": 16181
+ },
+ {
+ "epoch": 51.69968051118211,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0384,
+ "step": 16182
+ },
+ {
+ "epoch": 51.70287539936102,
+ "grad_norm": 0.0478515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0335,
+ "step": 16183
+ },
+ {
+ "epoch": 51.70607028753994,
+ "grad_norm": 0.05810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0231,
+ "step": 16184
+ },
+ {
+ "epoch": 51.70926517571885,
+ "grad_norm": 0.08203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0342,
+ "step": 16185
+ },
+ {
+ "epoch": 51.712460063897765,
+ "grad_norm": 0.061767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.025,
+ "step": 16186
+ },
+ {
+ "epoch": 51.715654952076676,
+ "grad_norm": 0.09130859375,
+ "learning_rate": 0.0005,
+ "loss": 1.018,
+ "step": 16187
+ },
+ {
+ "epoch": 51.718849840255594,
+ "grad_norm": 0.0869140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0231,
+ "step": 16188
+ },
+ {
+ "epoch": 51.722044728434504,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0208,
+ "step": 16189
+ },
+ {
+ "epoch": 51.72523961661342,
+ "grad_norm": 0.0869140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0284,
+ "step": 16190
+ },
+ {
+ "epoch": 51.72843450479233,
+ "grad_norm": 0.050537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0184,
+ "step": 16191
+ },
+ {
+ "epoch": 51.73162939297124,
+ "grad_norm": 0.052978515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0233,
+ "step": 16192
+ },
+ {
+ "epoch": 51.73482428115016,
+ "grad_norm": 0.06005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0342,
+ "step": 16193
+ },
+ {
+ "epoch": 51.73801916932907,
+ "grad_norm": 0.051025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0251,
+ "step": 16194
+ },
+ {
+ "epoch": 51.74121405750799,
+ "grad_norm": 0.0517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0337,
+ "step": 16195
+ },
+ {
+ "epoch": 51.7444089456869,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0248,
+ "step": 16196
+ },
+ {
+ "epoch": 51.74760383386582,
+ "grad_norm": 0.060302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0286,
+ "step": 16197
+ },
+ {
+ "epoch": 51.75079872204473,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0218,
+ "step": 16198
+ },
+ {
+ "epoch": 51.753993610223645,
+ "grad_norm": 0.0595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0323,
+ "step": 16199
+ },
+ {
+ "epoch": 51.757188498402556,
+ "grad_norm": 0.059814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0225,
+ "step": 16200
+ },
+ {
+ "epoch": 51.760383386581466,
+ "grad_norm": 0.08935546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0266,
+ "step": 16201
+ },
+ {
+ "epoch": 51.763578274760384,
+ "grad_norm": 0.055908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0275,
+ "step": 16202
+ },
+ {
+ "epoch": 51.766773162939295,
+ "grad_norm": 0.058837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.024,
+ "step": 16203
+ },
+ {
+ "epoch": 51.76996805111821,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0234,
+ "step": 16204
+ },
+ {
+ "epoch": 51.77316293929712,
+ "grad_norm": 0.05029296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0306,
+ "step": 16205
+ },
+ {
+ "epoch": 51.77635782747604,
+ "grad_norm": 0.0546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0287,
+ "step": 16206
+ },
+ {
+ "epoch": 51.77955271565495,
+ "grad_norm": 0.046142578125,
+ "learning_rate": 0.0005,
+ "loss": 1.026,
+ "step": 16207
+ },
+ {
+ "epoch": 51.78274760383387,
+ "grad_norm": 0.050537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0274,
+ "step": 16208
+ },
+ {
+ "epoch": 51.78594249201278,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0288,
+ "step": 16209
+ },
+ {
+ "epoch": 51.78913738019169,
+ "grad_norm": 0.09619140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0345,
+ "step": 16210
+ },
+ {
+ "epoch": 51.79233226837061,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0269,
+ "step": 16211
+ },
+ {
+ "epoch": 51.79552715654952,
+ "grad_norm": 0.09033203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0246,
+ "step": 16212
+ },
+ {
+ "epoch": 51.798722044728436,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0383,
+ "step": 16213
+ },
+ {
+ "epoch": 51.801916932907346,
+ "grad_norm": 0.09033203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0208,
+ "step": 16214
+ },
+ {
+ "epoch": 51.805111821086264,
+ "grad_norm": 0.055908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0327,
+ "step": 16215
+ },
+ {
+ "epoch": 51.808306709265175,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0277,
+ "step": 16216
+ },
+ {
+ "epoch": 51.81150159744409,
+ "grad_norm": 0.10888671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0251,
+ "step": 16217
+ },
+ {
+ "epoch": 51.814696485623,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0341,
+ "step": 16218
+ },
+ {
+ "epoch": 51.81789137380191,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0296,
+ "step": 16219
+ },
+ {
+ "epoch": 51.82108626198083,
+ "grad_norm": 0.056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.029,
+ "step": 16220
+ },
+ {
+ "epoch": 51.82428115015974,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0246,
+ "step": 16221
+ },
+ {
+ "epoch": 51.82747603833866,
+ "grad_norm": 0.0517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0341,
+ "step": 16222
+ },
+ {
+ "epoch": 51.83067092651757,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0294,
+ "step": 16223
+ },
+ {
+ "epoch": 51.83386581469649,
+ "grad_norm": 0.047607421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0362,
+ "step": 16224
+ },
+ {
+ "epoch": 51.8370607028754,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0263,
+ "step": 16225
+ },
+ {
+ "epoch": 51.840255591054316,
+ "grad_norm": 0.059326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0229,
+ "step": 16226
+ },
+ {
+ "epoch": 51.843450479233226,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0253,
+ "step": 16227
+ },
+ {
+ "epoch": 51.846645367412144,
+ "grad_norm": 0.05712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0266,
+ "step": 16228
+ },
+ {
+ "epoch": 51.849840255591054,
+ "grad_norm": 0.05029296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0177,
+ "step": 16229
+ },
+ {
+ "epoch": 51.853035143769965,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0263,
+ "step": 16230
+ },
+ {
+ "epoch": 51.85623003194888,
+ "grad_norm": 0.058349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0345,
+ "step": 16231
+ },
+ {
+ "epoch": 51.85942492012779,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0248,
+ "step": 16232
+ },
+ {
+ "epoch": 51.86261980830671,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0333,
+ "step": 16233
+ },
+ {
+ "epoch": 51.86581469648562,
+ "grad_norm": 0.052001953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0348,
+ "step": 16234
+ },
+ {
+ "epoch": 51.86900958466454,
+ "grad_norm": 0.0595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0283,
+ "step": 16235
+ },
+ {
+ "epoch": 51.87220447284345,
+ "grad_norm": 0.051025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0289,
+ "step": 16236
+ },
+ {
+ "epoch": 51.87539936102237,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.033,
+ "step": 16237
+ },
+ {
+ "epoch": 51.87859424920128,
+ "grad_norm": 0.056396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0272,
+ "step": 16238
+ },
+ {
+ "epoch": 51.88178913738019,
+ "grad_norm": 0.057373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0359,
+ "step": 16239
+ },
+ {
+ "epoch": 51.884984025559106,
+ "grad_norm": 0.046630859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0303,
+ "step": 16240
+ },
+ {
+ "epoch": 51.88817891373802,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0257,
+ "step": 16241
+ },
+ {
+ "epoch": 51.891373801916934,
+ "grad_norm": 0.061279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0255,
+ "step": 16242
+ },
+ {
+ "epoch": 51.894568690095845,
+ "grad_norm": 0.049072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.038,
+ "step": 16243
+ },
+ {
+ "epoch": 51.89776357827476,
+ "grad_norm": 0.055908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0176,
+ "step": 16244
+ },
+ {
+ "epoch": 51.90095846645367,
+ "grad_norm": 0.08154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0317,
+ "step": 16245
+ },
+ {
+ "epoch": 51.90415335463259,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0274,
+ "step": 16246
+ },
+ {
+ "epoch": 51.9073482428115,
+ "grad_norm": 0.05712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0311,
+ "step": 16247
+ },
+ {
+ "epoch": 51.91054313099041,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0276,
+ "step": 16248
+ },
+ {
+ "epoch": 51.91373801916933,
+ "grad_norm": 0.056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0348,
+ "step": 16249
+ },
+ {
+ "epoch": 51.91693290734824,
+ "grad_norm": 0.061279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.034,
+ "step": 16250
+ },
+ {
+ "epoch": 51.92012779552716,
+ "grad_norm": 0.0517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0169,
+ "step": 16251
+ },
+ {
+ "epoch": 51.92332268370607,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0267,
+ "step": 16252
+ },
+ {
+ "epoch": 51.926517571884986,
+ "grad_norm": 0.053466796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0139,
+ "step": 16253
+ },
+ {
+ "epoch": 51.9297124600639,
+ "grad_norm": 0.052001953125,
+ "learning_rate": 0.0005,
+ "loss": 1.008,
+ "step": 16254
+ },
+ {
+ "epoch": 51.932907348242814,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0228,
+ "step": 16255
+ },
+ {
+ "epoch": 51.936102236421725,
+ "grad_norm": 0.046630859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0223,
+ "step": 16256
+ },
+ {
+ "epoch": 51.93929712460064,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.022,
+ "step": 16257
+ },
+ {
+ "epoch": 51.94249201277955,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0289,
+ "step": 16258
+ },
+ {
+ "epoch": 51.945686900958464,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.027,
+ "step": 16259
+ },
+ {
+ "epoch": 51.94888178913738,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0257,
+ "step": 16260
+ },
+ {
+ "epoch": 51.95207667731629,
+ "grad_norm": 0.0576171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0298,
+ "step": 16261
+ },
+ {
+ "epoch": 51.95527156549521,
+ "grad_norm": 0.050048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0322,
+ "step": 16262
+ },
+ {
+ "epoch": 51.95846645367412,
+ "grad_norm": 0.0458984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0197,
+ "step": 16263
+ },
+ {
+ "epoch": 51.96166134185304,
+ "grad_norm": 0.050537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0326,
+ "step": 16264
+ },
+ {
+ "epoch": 51.96485623003195,
+ "grad_norm": 0.1015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0333,
+ "step": 16265
+ },
+ {
+ "epoch": 51.968051118210866,
+ "grad_norm": 0.046630859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0142,
+ "step": 16266
+ },
+ {
+ "epoch": 51.97124600638978,
+ "grad_norm": 0.0615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0379,
+ "step": 16267
+ },
+ {
+ "epoch": 51.97444089456869,
+ "grad_norm": 0.057373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0246,
+ "step": 16268
+ },
+ {
+ "epoch": 51.977635782747605,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0259,
+ "step": 16269
+ },
+ {
+ "epoch": 51.980830670926515,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0132,
+ "step": 16270
+ },
+ {
+ "epoch": 51.98402555910543,
+ "grad_norm": 0.08984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0318,
+ "step": 16271
+ },
+ {
+ "epoch": 51.98722044728434,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0314,
+ "step": 16272
+ },
+ {
+ "epoch": 51.99041533546326,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0308,
+ "step": 16273
+ },
+ {
+ "epoch": 51.99361022364217,
+ "grad_norm": 0.11181640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0394,
+ "step": 16274
+ },
+ {
+ "epoch": 51.99680511182109,
+ "grad_norm": 0.1396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.023,
+ "step": 16275
+ },
+ {
+ "epoch": 52.0,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0289,
+ "step": 16276
+ },
+ {
+ "epoch": 52.00319488817891,
+ "grad_norm": 0.1328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0138,
+ "step": 16277
+ },
+ {
+ "epoch": 52.00638977635783,
+ "grad_norm": 0.061279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0272,
+ "step": 16278
+ },
+ {
+ "epoch": 52.00958466453674,
+ "grad_norm": 0.09326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0293,
+ "step": 16279
+ },
+ {
+ "epoch": 52.01277955271566,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0315,
+ "step": 16280
+ },
+ {
+ "epoch": 52.01597444089457,
+ "grad_norm": 0.0576171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0431,
+ "step": 16281
+ },
+ {
+ "epoch": 52.019169329073485,
+ "grad_norm": 0.0927734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0237,
+ "step": 16282
+ },
+ {
+ "epoch": 52.022364217252395,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0372,
+ "step": 16283
+ },
+ {
+ "epoch": 52.02555910543131,
+ "grad_norm": 0.087890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0272,
+ "step": 16284
+ },
+ {
+ "epoch": 52.02875399361022,
+ "grad_norm": 0.103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0219,
+ "step": 16285
+ },
+ {
+ "epoch": 52.031948881789134,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0327,
+ "step": 16286
+ },
+ {
+ "epoch": 52.03514376996805,
+ "grad_norm": 0.12158203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0221,
+ "step": 16287
+ },
+ {
+ "epoch": 52.03833865814696,
+ "grad_norm": 0.08154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0359,
+ "step": 16288
+ },
+ {
+ "epoch": 52.04153354632588,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0281,
+ "step": 16289
+ },
+ {
+ "epoch": 52.04472843450479,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0335,
+ "step": 16290
+ },
+ {
+ "epoch": 52.04792332268371,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0247,
+ "step": 16291
+ },
+ {
+ "epoch": 52.05111821086262,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0201,
+ "step": 16292
+ },
+ {
+ "epoch": 52.054313099041536,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0246,
+ "step": 16293
+ },
+ {
+ "epoch": 52.05750798722045,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0275,
+ "step": 16294
+ },
+ {
+ "epoch": 52.06070287539936,
+ "grad_norm": 0.061767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.03,
+ "step": 16295
+ },
+ {
+ "epoch": 52.063897763578275,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0372,
+ "step": 16296
+ },
+ {
+ "epoch": 52.067092651757186,
+ "grad_norm": 0.05615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0348,
+ "step": 16297
+ },
+ {
+ "epoch": 52.0702875399361,
+ "grad_norm": 0.053955078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0357,
+ "step": 16298
+ },
+ {
+ "epoch": 52.073482428115014,
+ "grad_norm": 0.08349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0266,
+ "step": 16299
+ },
+ {
+ "epoch": 52.07667731629393,
+ "grad_norm": 0.0859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0276,
+ "step": 16300
+ },
+ {
+ "epoch": 52.07987220447284,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0235,
+ "step": 16301
+ },
+ {
+ "epoch": 52.08306709265176,
+ "grad_norm": 0.058349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0277,
+ "step": 16302
+ },
+ {
+ "epoch": 52.08626198083067,
+ "grad_norm": 0.09375,
+ "learning_rate": 0.0005,
+ "loss": 1.0282,
+ "step": 16303
+ },
+ {
+ "epoch": 52.08945686900959,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0296,
+ "step": 16304
+ },
+ {
+ "epoch": 52.0926517571885,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0152,
+ "step": 16305
+ },
+ {
+ "epoch": 52.09584664536741,
+ "grad_norm": 0.056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.028,
+ "step": 16306
+ },
+ {
+ "epoch": 52.09904153354633,
+ "grad_norm": 0.150390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0173,
+ "step": 16307
+ },
+ {
+ "epoch": 52.10223642172524,
+ "grad_norm": 0.0859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0215,
+ "step": 16308
+ },
+ {
+ "epoch": 52.105431309904155,
+ "grad_norm": 0.08154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0262,
+ "step": 16309
+ },
+ {
+ "epoch": 52.108626198083066,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0167,
+ "step": 16310
+ },
+ {
+ "epoch": 52.11182108626198,
+ "grad_norm": 0.111328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0302,
+ "step": 16311
+ },
+ {
+ "epoch": 52.115015974440894,
+ "grad_norm": 0.08984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0271,
+ "step": 16312
+ },
+ {
+ "epoch": 52.11821086261981,
+ "grad_norm": 0.059814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0177,
+ "step": 16313
+ },
+ {
+ "epoch": 52.12140575079872,
+ "grad_norm": 0.0830078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0315,
+ "step": 16314
+ },
+ {
+ "epoch": 52.12460063897763,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0254,
+ "step": 16315
+ },
+ {
+ "epoch": 52.12779552715655,
+ "grad_norm": 0.050048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0228,
+ "step": 16316
+ },
+ {
+ "epoch": 52.13099041533546,
+ "grad_norm": 0.05908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0251,
+ "step": 16317
+ },
+ {
+ "epoch": 52.13418530351438,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0313,
+ "step": 16318
+ },
+ {
+ "epoch": 52.13738019169329,
+ "grad_norm": 0.04638671875,
+ "learning_rate": 0.0005,
+ "loss": 1.019,
+ "step": 16319
+ },
+ {
+ "epoch": 52.14057507987221,
+ "grad_norm": 0.055419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0316,
+ "step": 16320
+ },
+ {
+ "epoch": 52.14376996805112,
+ "grad_norm": 0.056396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0191,
+ "step": 16321
+ },
+ {
+ "epoch": 52.146964856230035,
+ "grad_norm": 0.056884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0281,
+ "step": 16322
+ },
+ {
+ "epoch": 52.150159744408946,
+ "grad_norm": 0.051025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0208,
+ "step": 16323
+ },
+ {
+ "epoch": 52.153354632587856,
+ "grad_norm": 0.06201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0268,
+ "step": 16324
+ },
+ {
+ "epoch": 52.156549520766774,
+ "grad_norm": 0.04638671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0328,
+ "step": 16325
+ },
+ {
+ "epoch": 52.159744408945684,
+ "grad_norm": 0.05224609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0406,
+ "step": 16326
+ },
+ {
+ "epoch": 52.1629392971246,
+ "grad_norm": 0.06103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0386,
+ "step": 16327
+ },
+ {
+ "epoch": 52.16613418530351,
+ "grad_norm": 0.046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0251,
+ "step": 16328
+ },
+ {
+ "epoch": 52.16932907348243,
+ "grad_norm": 0.053466796875,
+ "learning_rate": 0.0005,
+ "loss": 1.032,
+ "step": 16329
+ },
+ {
+ "epoch": 52.17252396166134,
+ "grad_norm": 0.06201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0176,
+ "step": 16330
+ },
+ {
+ "epoch": 52.17571884984026,
+ "grad_norm": 0.057861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0187,
+ "step": 16331
+ },
+ {
+ "epoch": 52.17891373801917,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0328,
+ "step": 16332
+ },
+ {
+ "epoch": 52.18210862619808,
+ "grad_norm": 0.056396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0293,
+ "step": 16333
+ },
+ {
+ "epoch": 52.185303514377,
+ "grad_norm": 0.04736328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0239,
+ "step": 16334
+ },
+ {
+ "epoch": 52.18849840255591,
+ "grad_norm": 0.0498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0351,
+ "step": 16335
+ },
+ {
+ "epoch": 52.191693290734825,
+ "grad_norm": 0.04345703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0308,
+ "step": 16336
+ },
+ {
+ "epoch": 52.194888178913736,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0125,
+ "step": 16337
+ },
+ {
+ "epoch": 52.198083067092654,
+ "grad_norm": 0.044921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0375,
+ "step": 16338
+ },
+ {
+ "epoch": 52.201277955271564,
+ "grad_norm": 0.0439453125,
+ "learning_rate": 0.0005,
+ "loss": 1.031,
+ "step": 16339
+ },
+ {
+ "epoch": 52.20447284345048,
+ "grad_norm": 0.08349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0321,
+ "step": 16340
+ },
+ {
+ "epoch": 52.20766773162939,
+ "grad_norm": 0.052978515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0312,
+ "step": 16341
+ },
+ {
+ "epoch": 52.21086261980831,
+ "grad_norm": 0.048095703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0363,
+ "step": 16342
+ },
+ {
+ "epoch": 52.21405750798722,
+ "grad_norm": 0.080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0228,
+ "step": 16343
+ },
+ {
+ "epoch": 52.21725239616613,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0301,
+ "step": 16344
+ },
+ {
+ "epoch": 52.22044728434505,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0229,
+ "step": 16345
+ },
+ {
+ "epoch": 52.22364217252396,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0206,
+ "step": 16346
+ },
+ {
+ "epoch": 52.22683706070288,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0206,
+ "step": 16347
+ },
+ {
+ "epoch": 52.23003194888179,
+ "grad_norm": 0.054443359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0243,
+ "step": 16348
+ },
+ {
+ "epoch": 52.233226837060705,
+ "grad_norm": 0.062255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0256,
+ "step": 16349
+ },
+ {
+ "epoch": 52.236421725239616,
+ "grad_norm": 0.083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0299,
+ "step": 16350
+ },
+ {
+ "epoch": 52.239616613418534,
+ "grad_norm": 0.126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0256,
+ "step": 16351
+ },
+ {
+ "epoch": 52.242811501597444,
+ "grad_norm": 0.0810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0291,
+ "step": 16352
+ },
+ {
+ "epoch": 52.246006389776355,
+ "grad_norm": 0.1650390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0281,
+ "step": 16353
+ },
+ {
+ "epoch": 52.24920127795527,
+ "grad_norm": 0.0576171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0285,
+ "step": 16354
+ },
+ {
+ "epoch": 52.25239616613418,
+ "grad_norm": 0.1689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0351,
+ "step": 16355
+ },
+ {
+ "epoch": 52.2555910543131,
+ "grad_norm": 0.053466796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0195,
+ "step": 16356
+ },
+ {
+ "epoch": 52.25878594249201,
+ "grad_norm": 0.1337890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0258,
+ "step": 16357
+ },
+ {
+ "epoch": 52.26198083067093,
+ "grad_norm": 0.05859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0331,
+ "step": 16358
+ },
+ {
+ "epoch": 52.26517571884984,
+ "grad_norm": 0.0986328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0375,
+ "step": 16359
+ },
+ {
+ "epoch": 52.26837060702876,
+ "grad_norm": 0.0517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0163,
+ "step": 16360
+ },
+ {
+ "epoch": 52.27156549520767,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0124,
+ "step": 16361
+ },
+ {
+ "epoch": 52.27476038338658,
+ "grad_norm": 0.08935546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0409,
+ "step": 16362
+ },
+ {
+ "epoch": 52.277955271565496,
+ "grad_norm": 0.15625,
+ "learning_rate": 0.0005,
+ "loss": 1.0358,
+ "step": 16363
+ },
+ {
+ "epoch": 52.281150159744406,
+ "grad_norm": 0.051513671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0154,
+ "step": 16364
+ },
+ {
+ "epoch": 52.284345047923324,
+ "grad_norm": 0.162109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0236,
+ "step": 16365
+ },
+ {
+ "epoch": 52.287539936102235,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0254,
+ "step": 16366
+ },
+ {
+ "epoch": 52.29073482428115,
+ "grad_norm": 0.267578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0296,
+ "step": 16367
+ },
+ {
+ "epoch": 52.29392971246006,
+ "grad_norm": 0.09423828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0132,
+ "step": 16368
+ },
+ {
+ "epoch": 52.29712460063898,
+ "grad_norm": 0.12890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0281,
+ "step": 16369
+ },
+ {
+ "epoch": 52.30031948881789,
+ "grad_norm": 0.099609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0341,
+ "step": 16370
+ },
+ {
+ "epoch": 52.3035143769968,
+ "grad_norm": 0.095703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0301,
+ "step": 16371
+ },
+ {
+ "epoch": 52.30670926517572,
+ "grad_norm": 0.11669921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0194,
+ "step": 16372
+ },
+ {
+ "epoch": 52.30990415335463,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0247,
+ "step": 16373
+ },
+ {
+ "epoch": 52.31309904153355,
+ "grad_norm": 0.080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0249,
+ "step": 16374
+ },
+ {
+ "epoch": 52.31629392971246,
+ "grad_norm": 0.091796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0292,
+ "step": 16375
+ },
+ {
+ "epoch": 52.319488817891376,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0268,
+ "step": 16376
+ },
+ {
+ "epoch": 52.322683706070286,
+ "grad_norm": 0.1298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0245,
+ "step": 16377
+ },
+ {
+ "epoch": 52.325878594249204,
+ "grad_norm": 0.1591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0421,
+ "step": 16378
+ },
+ {
+ "epoch": 52.329073482428115,
+ "grad_norm": 0.08837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0371,
+ "step": 16379
+ },
+ {
+ "epoch": 52.33226837060703,
+ "grad_norm": 0.1845703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0208,
+ "step": 16380
+ },
+ {
+ "epoch": 52.33546325878594,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.027,
+ "step": 16381
+ },
+ {
+ "epoch": 52.33865814696485,
+ "grad_norm": 0.263671875,
+ "learning_rate": 0.0005,
+ "loss": 1.027,
+ "step": 16382
+ },
+ {
+ "epoch": 52.34185303514377,
+ "grad_norm": 0.12890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0317,
+ "step": 16383
+ },
+ {
+ "epoch": 52.34504792332268,
+ "grad_norm": 0.1416015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0281,
+ "step": 16384
+ },
+ {
+ "epoch": 52.3482428115016,
+ "grad_norm": 0.11474609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0255,
+ "step": 16385
+ },
+ {
+ "epoch": 52.35143769968051,
+ "grad_norm": 0.10498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0213,
+ "step": 16386
+ },
+ {
+ "epoch": 52.35463258785943,
+ "grad_norm": 0.11279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0336,
+ "step": 16387
+ },
+ {
+ "epoch": 52.35782747603834,
+ "grad_norm": 0.09326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0256,
+ "step": 16388
+ },
+ {
+ "epoch": 52.361022364217256,
+ "grad_norm": 0.1845703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0378,
+ "step": 16389
+ },
+ {
+ "epoch": 52.364217252396166,
+ "grad_norm": 0.1376953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0219,
+ "step": 16390
+ },
+ {
+ "epoch": 52.36741214057508,
+ "grad_norm": 0.146484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0291,
+ "step": 16391
+ },
+ {
+ "epoch": 52.370607028753994,
+ "grad_norm": 0.11279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.026,
+ "step": 16392
+ },
+ {
+ "epoch": 52.373801916932905,
+ "grad_norm": 0.1376953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0311,
+ "step": 16393
+ },
+ {
+ "epoch": 52.37699680511182,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0287,
+ "step": 16394
+ },
+ {
+ "epoch": 52.38019169329073,
+ "grad_norm": 0.10498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0284,
+ "step": 16395
+ },
+ {
+ "epoch": 52.38338658146965,
+ "grad_norm": 0.09423828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0272,
+ "step": 16396
+ },
+ {
+ "epoch": 52.38658146964856,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0347,
+ "step": 16397
+ },
+ {
+ "epoch": 52.38977635782748,
+ "grad_norm": 0.103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0363,
+ "step": 16398
+ },
+ {
+ "epoch": 52.39297124600639,
+ "grad_norm": 0.04541015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0241,
+ "step": 16399
+ },
+ {
+ "epoch": 52.3961661341853,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0164,
+ "step": 16400
+ },
+ {
+ "epoch": 52.39936102236422,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0261,
+ "step": 16401
+ },
+ {
+ "epoch": 52.40255591054313,
+ "grad_norm": 0.054931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0322,
+ "step": 16402
+ },
+ {
+ "epoch": 52.405750798722046,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0273,
+ "step": 16403
+ },
+ {
+ "epoch": 52.40894568690096,
+ "grad_norm": 0.0498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0321,
+ "step": 16404
+ },
+ {
+ "epoch": 52.412140575079874,
+ "grad_norm": 0.08251953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0335,
+ "step": 16405
+ },
+ {
+ "epoch": 52.415335463258785,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0255,
+ "step": 16406
+ },
+ {
+ "epoch": 52.4185303514377,
+ "grad_norm": 0.09033203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0116,
+ "step": 16407
+ },
+ {
+ "epoch": 52.42172523961661,
+ "grad_norm": 0.05859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0262,
+ "step": 16408
+ },
+ {
+ "epoch": 52.424920127795524,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0198,
+ "step": 16409
+ },
+ {
+ "epoch": 52.42811501597444,
+ "grad_norm": 0.04736328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0231,
+ "step": 16410
+ },
+ {
+ "epoch": 52.43130990415335,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0289,
+ "step": 16411
+ },
+ {
+ "epoch": 52.43450479233227,
+ "grad_norm": 0.059326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0158,
+ "step": 16412
+ },
+ {
+ "epoch": 52.43769968051118,
+ "grad_norm": 0.054443359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0185,
+ "step": 16413
+ },
+ {
+ "epoch": 52.4408945686901,
+ "grad_norm": 0.045166015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0296,
+ "step": 16414
+ },
+ {
+ "epoch": 52.44408945686901,
+ "grad_norm": 0.053466796875,
+ "learning_rate": 0.0005,
+ "loss": 1.02,
+ "step": 16415
+ },
+ {
+ "epoch": 52.447284345047926,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0206,
+ "step": 16416
+ },
+ {
+ "epoch": 52.45047923322684,
+ "grad_norm": 0.05859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0446,
+ "step": 16417
+ },
+ {
+ "epoch": 52.453674121405754,
+ "grad_norm": 0.103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0263,
+ "step": 16418
+ },
+ {
+ "epoch": 52.456869009584665,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0286,
+ "step": 16419
+ },
+ {
+ "epoch": 52.460063897763575,
+ "grad_norm": 0.099609375,
+ "learning_rate": 0.0005,
+ "loss": 1.032,
+ "step": 16420
+ },
+ {
+ "epoch": 52.46325878594249,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0359,
+ "step": 16421
+ },
+ {
+ "epoch": 52.466453674121404,
+ "grad_norm": 0.111328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0384,
+ "step": 16422
+ },
+ {
+ "epoch": 52.46964856230032,
+ "grad_norm": 0.056884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0198,
+ "step": 16423
+ },
+ {
+ "epoch": 52.47284345047923,
+ "grad_norm": 0.1162109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0411,
+ "step": 16424
+ },
+ {
+ "epoch": 52.47603833865815,
+ "grad_norm": 0.080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0314,
+ "step": 16425
+ },
+ {
+ "epoch": 52.47923322683706,
+ "grad_norm": 0.087890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0246,
+ "step": 16426
+ },
+ {
+ "epoch": 52.48242811501598,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0404,
+ "step": 16427
+ },
+ {
+ "epoch": 52.48562300319489,
+ "grad_norm": 0.1328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0272,
+ "step": 16428
+ },
+ {
+ "epoch": 52.4888178913738,
+ "grad_norm": 0.1103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0241,
+ "step": 16429
+ },
+ {
+ "epoch": 52.49201277955272,
+ "grad_norm": 0.1435546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0283,
+ "step": 16430
+ },
+ {
+ "epoch": 52.49520766773163,
+ "grad_norm": 0.162109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0223,
+ "step": 16431
+ },
+ {
+ "epoch": 52.498402555910545,
+ "grad_norm": 0.1572265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0109,
+ "step": 16432
+ },
+ {
+ "epoch": 52.501597444089455,
+ "grad_norm": 0.232421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0336,
+ "step": 16433
+ },
+ {
+ "epoch": 52.50479233226837,
+ "grad_norm": 0.12451171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0271,
+ "step": 16434
+ },
+ {
+ "epoch": 52.50798722044728,
+ "grad_norm": 0.1708984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0162,
+ "step": 16435
+ },
+ {
+ "epoch": 52.5111821086262,
+ "grad_norm": 0.1455078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0308,
+ "step": 16436
+ },
+ {
+ "epoch": 52.51437699680511,
+ "grad_norm": 0.123046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0265,
+ "step": 16437
+ },
+ {
+ "epoch": 52.51757188498402,
+ "grad_norm": 0.16015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0219,
+ "step": 16438
+ },
+ {
+ "epoch": 52.52076677316294,
+ "grad_norm": 0.08544921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0266,
+ "step": 16439
+ },
+ {
+ "epoch": 52.52396166134185,
+ "grad_norm": 0.2265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0241,
+ "step": 16440
+ },
+ {
+ "epoch": 52.52715654952077,
+ "grad_norm": 0.154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0233,
+ "step": 16441
+ },
+ {
+ "epoch": 52.53035143769968,
+ "grad_norm": 0.1982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0389,
+ "step": 16442
+ },
+ {
+ "epoch": 52.533546325878596,
+ "grad_norm": 0.212890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0204,
+ "step": 16443
+ },
+ {
+ "epoch": 52.53674121405751,
+ "grad_norm": 0.181640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0295,
+ "step": 16444
+ },
+ {
+ "epoch": 52.539936102236425,
+ "grad_norm": 0.328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0236,
+ "step": 16445
+ },
+ {
+ "epoch": 52.543130990415335,
+ "grad_norm": 0.10986328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0284,
+ "step": 16446
+ },
+ {
+ "epoch": 52.546325878594246,
+ "grad_norm": 0.220703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0321,
+ "step": 16447
+ },
+ {
+ "epoch": 52.54952076677316,
+ "grad_norm": 0.181640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0276,
+ "step": 16448
+ },
+ {
+ "epoch": 52.552715654952074,
+ "grad_norm": 0.1318359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0364,
+ "step": 16449
+ },
+ {
+ "epoch": 52.55591054313099,
+ "grad_norm": 0.34375,
+ "learning_rate": 0.0005,
+ "loss": 1.0386,
+ "step": 16450
+ },
+ {
+ "epoch": 52.5591054313099,
+ "grad_norm": 0.3671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0423,
+ "step": 16451
+ },
+ {
+ "epoch": 52.56230031948882,
+ "grad_norm": 0.126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0196,
+ "step": 16452
+ },
+ {
+ "epoch": 52.56549520766773,
+ "grad_norm": 0.244140625,
+ "learning_rate": 0.0005,
+ "loss": 1.029,
+ "step": 16453
+ },
+ {
+ "epoch": 52.56869009584665,
+ "grad_norm": 0.2158203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0329,
+ "step": 16454
+ },
+ {
+ "epoch": 52.57188498402556,
+ "grad_norm": 0.16015625,
+ "learning_rate": 0.0005,
+ "loss": 1.03,
+ "step": 16455
+ },
+ {
+ "epoch": 52.575079872204476,
+ "grad_norm": 0.30078125,
+ "learning_rate": 0.0005,
+ "loss": 1.024,
+ "step": 16456
+ },
+ {
+ "epoch": 52.57827476038339,
+ "grad_norm": 0.11279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0276,
+ "step": 16457
+ },
+ {
+ "epoch": 52.5814696485623,
+ "grad_norm": 0.134765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0283,
+ "step": 16458
+ },
+ {
+ "epoch": 52.584664536741215,
+ "grad_norm": 0.1279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.044,
+ "step": 16459
+ },
+ {
+ "epoch": 52.587859424920126,
+ "grad_norm": 0.11376953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0387,
+ "step": 16460
+ },
+ {
+ "epoch": 52.59105431309904,
+ "grad_norm": 0.1962890625,
+ "learning_rate": 0.0005,
+ "loss": 1.021,
+ "step": 16461
+ },
+ {
+ "epoch": 52.594249201277954,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0237,
+ "step": 16462
+ },
+ {
+ "epoch": 52.59744408945687,
+ "grad_norm": 0.259765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0315,
+ "step": 16463
+ },
+ {
+ "epoch": 52.60063897763578,
+ "grad_norm": 0.1552734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0276,
+ "step": 16464
+ },
+ {
+ "epoch": 52.6038338658147,
+ "grad_norm": 0.1845703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0258,
+ "step": 16465
+ },
+ {
+ "epoch": 52.60702875399361,
+ "grad_norm": 0.2138671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0256,
+ "step": 16466
+ },
+ {
+ "epoch": 52.61022364217252,
+ "grad_norm": 0.107421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0377,
+ "step": 16467
+ },
+ {
+ "epoch": 52.61341853035144,
+ "grad_norm": 0.171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0202,
+ "step": 16468
+ },
+ {
+ "epoch": 52.61661341853035,
+ "grad_norm": 0.10009765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0284,
+ "step": 16469
+ },
+ {
+ "epoch": 52.61980830670927,
+ "grad_norm": 0.162109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0125,
+ "step": 16470
+ },
+ {
+ "epoch": 52.62300319488818,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.032,
+ "step": 16471
+ },
+ {
+ "epoch": 52.626198083067095,
+ "grad_norm": 0.1611328125,
+ "learning_rate": 0.0005,
+ "loss": 1.034,
+ "step": 16472
+ },
+ {
+ "epoch": 52.629392971246006,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0187,
+ "step": 16473
+ },
+ {
+ "epoch": 52.63258785942492,
+ "grad_norm": 0.1630859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0291,
+ "step": 16474
+ },
+ {
+ "epoch": 52.635782747603834,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.025,
+ "step": 16475
+ },
+ {
+ "epoch": 52.638977635782744,
+ "grad_norm": 0.173828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0294,
+ "step": 16476
+ },
+ {
+ "epoch": 52.64217252396166,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0265,
+ "step": 16477
+ },
+ {
+ "epoch": 52.64536741214057,
+ "grad_norm": 0.2177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0314,
+ "step": 16478
+ },
+ {
+ "epoch": 52.64856230031949,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0341,
+ "step": 16479
+ },
+ {
+ "epoch": 52.6517571884984,
+ "grad_norm": 0.12890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0283,
+ "step": 16480
+ },
+ {
+ "epoch": 52.65495207667732,
+ "grad_norm": 0.055419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0087,
+ "step": 16481
+ },
+ {
+ "epoch": 52.65814696485623,
+ "grad_norm": 0.13671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0267,
+ "step": 16482
+ },
+ {
+ "epoch": 52.66134185303515,
+ "grad_norm": 0.08154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.023,
+ "step": 16483
+ },
+ {
+ "epoch": 52.66453674121406,
+ "grad_norm": 0.134765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0148,
+ "step": 16484
+ },
+ {
+ "epoch": 52.66773162939297,
+ "grad_norm": 0.0830078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0152,
+ "step": 16485
+ },
+ {
+ "epoch": 52.670926517571885,
+ "grad_norm": 0.12890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0303,
+ "step": 16486
+ },
+ {
+ "epoch": 52.674121405750796,
+ "grad_norm": 0.08984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0256,
+ "step": 16487
+ },
+ {
+ "epoch": 52.677316293929714,
+ "grad_norm": 0.1513671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0312,
+ "step": 16488
+ },
+ {
+ "epoch": 52.680511182108624,
+ "grad_norm": 0.054931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0171,
+ "step": 16489
+ },
+ {
+ "epoch": 52.68370607028754,
+ "grad_norm": 0.12890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0245,
+ "step": 16490
+ },
+ {
+ "epoch": 52.68690095846645,
+ "grad_norm": 0.0517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0306,
+ "step": 16491
+ },
+ {
+ "epoch": 52.69009584664537,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0266,
+ "step": 16492
+ },
+ {
+ "epoch": 52.69329073482428,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0249,
+ "step": 16493
+ },
+ {
+ "epoch": 52.6964856230032,
+ "grad_norm": 0.0849609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0221,
+ "step": 16494
+ },
+ {
+ "epoch": 52.69968051118211,
+ "grad_norm": 0.09033203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0338,
+ "step": 16495
+ },
+ {
+ "epoch": 52.70287539936102,
+ "grad_norm": 0.10205078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0272,
+ "step": 16496
+ },
+ {
+ "epoch": 52.70607028753994,
+ "grad_norm": 0.060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.026,
+ "step": 16497
+ },
+ {
+ "epoch": 52.70926517571885,
+ "grad_norm": 0.0595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0262,
+ "step": 16498
+ },
+ {
+ "epoch": 52.712460063897765,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0285,
+ "step": 16499
+ },
+ {
+ "epoch": 52.715654952076676,
+ "grad_norm": 0.046142578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0256,
+ "step": 16500
+ },
+ {
+ "epoch": 52.718849840255594,
+ "grad_norm": 0.083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0264,
+ "step": 16501
+ },
+ {
+ "epoch": 52.722044728434504,
+ "grad_norm": 0.043212890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0266,
+ "step": 16502
+ },
+ {
+ "epoch": 52.72523961661342,
+ "grad_norm": 0.08154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0316,
+ "step": 16503
+ },
+ {
+ "epoch": 52.72843450479233,
+ "grad_norm": 0.049560546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0306,
+ "step": 16504
+ },
+ {
+ "epoch": 52.73162939297124,
+ "grad_norm": 0.060302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.023,
+ "step": 16505
+ },
+ {
+ "epoch": 52.73482428115016,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0272,
+ "step": 16506
+ },
+ {
+ "epoch": 52.73801916932907,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0324,
+ "step": 16507
+ },
+ {
+ "epoch": 52.74121405750799,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0151,
+ "step": 16508
+ },
+ {
+ "epoch": 52.7444089456869,
+ "grad_norm": 0.0927734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0229,
+ "step": 16509
+ },
+ {
+ "epoch": 52.74760383386582,
+ "grad_norm": 0.057861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0262,
+ "step": 16510
+ },
+ {
+ "epoch": 52.75079872204473,
+ "grad_norm": 0.060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0317,
+ "step": 16511
+ },
+ {
+ "epoch": 52.753993610223645,
+ "grad_norm": 0.050537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0275,
+ "step": 16512
+ },
+ {
+ "epoch": 52.757188498402556,
+ "grad_norm": 0.0517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.029,
+ "step": 16513
+ },
+ {
+ "epoch": 52.760383386581466,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0391,
+ "step": 16514
+ },
+ {
+ "epoch": 52.763578274760384,
+ "grad_norm": 0.08056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0427,
+ "step": 16515
+ },
+ {
+ "epoch": 52.766773162939295,
+ "grad_norm": 0.056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0183,
+ "step": 16516
+ },
+ {
+ "epoch": 52.76996805111821,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0416,
+ "step": 16517
+ },
+ {
+ "epoch": 52.77316293929712,
+ "grad_norm": 0.044677734375,
+ "learning_rate": 0.0005,
+ "loss": 1.026,
+ "step": 16518
+ },
+ {
+ "epoch": 52.77635782747604,
+ "grad_norm": 0.0537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0268,
+ "step": 16519
+ },
+ {
+ "epoch": 52.77955271565495,
+ "grad_norm": 0.051513671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0199,
+ "step": 16520
+ },
+ {
+ "epoch": 52.78274760383387,
+ "grad_norm": 0.058349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0253,
+ "step": 16521
+ },
+ {
+ "epoch": 52.78594249201278,
+ "grad_norm": 0.06201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0294,
+ "step": 16522
+ },
+ {
+ "epoch": 52.78913738019169,
+ "grad_norm": 0.080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0411,
+ "step": 16523
+ },
+ {
+ "epoch": 52.79233226837061,
+ "grad_norm": 0.05224609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0259,
+ "step": 16524
+ },
+ {
+ "epoch": 52.79552715654952,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0225,
+ "step": 16525
+ },
+ {
+ "epoch": 52.798722044728436,
+ "grad_norm": 0.049072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0107,
+ "step": 16526
+ },
+ {
+ "epoch": 52.801916932907346,
+ "grad_norm": 0.055419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0163,
+ "step": 16527
+ },
+ {
+ "epoch": 52.805111821086264,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0187,
+ "step": 16528
+ },
+ {
+ "epoch": 52.808306709265175,
+ "grad_norm": 0.0546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0209,
+ "step": 16529
+ },
+ {
+ "epoch": 52.81150159744409,
+ "grad_norm": 0.055419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0357,
+ "step": 16530
+ },
+ {
+ "epoch": 52.814696485623,
+ "grad_norm": 0.054443359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0334,
+ "step": 16531
+ },
+ {
+ "epoch": 52.81789137380191,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0088,
+ "step": 16532
+ },
+ {
+ "epoch": 52.82108626198083,
+ "grad_norm": 0.055419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.032,
+ "step": 16533
+ },
+ {
+ "epoch": 52.82428115015974,
+ "grad_norm": 0.05224609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0254,
+ "step": 16534
+ },
+ {
+ "epoch": 52.82747603833866,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0348,
+ "step": 16535
+ },
+ {
+ "epoch": 52.83067092651757,
+ "grad_norm": 0.06201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0283,
+ "step": 16536
+ },
+ {
+ "epoch": 52.83386581469649,
+ "grad_norm": 0.06201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0269,
+ "step": 16537
+ },
+ {
+ "epoch": 52.8370607028754,
+ "grad_norm": 0.052978515625,
+ "learning_rate": 0.0005,
+ "loss": 1.026,
+ "step": 16538
+ },
+ {
+ "epoch": 52.840255591054316,
+ "grad_norm": 0.0478515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0236,
+ "step": 16539
+ },
+ {
+ "epoch": 52.843450479233226,
+ "grad_norm": 0.053955078125,
+ "learning_rate": 0.0005,
+ "loss": 1.024,
+ "step": 16540
+ },
+ {
+ "epoch": 52.846645367412144,
+ "grad_norm": 0.05078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0313,
+ "step": 16541
+ },
+ {
+ "epoch": 52.849840255591054,
+ "grad_norm": 0.054931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0299,
+ "step": 16542
+ },
+ {
+ "epoch": 52.853035143769965,
+ "grad_norm": 0.09814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0272,
+ "step": 16543
+ },
+ {
+ "epoch": 52.85623003194888,
+ "grad_norm": 0.056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0121,
+ "step": 16544
+ },
+ {
+ "epoch": 52.85942492012779,
+ "grad_norm": 0.1005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0292,
+ "step": 16545
+ },
+ {
+ "epoch": 52.86261980830671,
+ "grad_norm": 0.11083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0281,
+ "step": 16546
+ },
+ {
+ "epoch": 52.86581469648562,
+ "grad_norm": 0.1123046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0291,
+ "step": 16547
+ },
+ {
+ "epoch": 52.86900958466454,
+ "grad_norm": 0.08203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0354,
+ "step": 16548
+ },
+ {
+ "epoch": 52.87220447284345,
+ "grad_norm": 0.07861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0204,
+ "step": 16549
+ },
+ {
+ "epoch": 52.87539936102237,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0288,
+ "step": 16550
+ },
+ {
+ "epoch": 52.87859424920128,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0276,
+ "step": 16551
+ },
+ {
+ "epoch": 52.88178913738019,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0176,
+ "step": 16552
+ },
+ {
+ "epoch": 52.884984025559106,
+ "grad_norm": 0.05615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0417,
+ "step": 16553
+ },
+ {
+ "epoch": 52.88817891373802,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0329,
+ "step": 16554
+ },
+ {
+ "epoch": 52.891373801916934,
+ "grad_norm": 0.048583984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0313,
+ "step": 16555
+ },
+ {
+ "epoch": 52.894568690095845,
+ "grad_norm": 0.050537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0201,
+ "step": 16556
+ },
+ {
+ "epoch": 52.89776357827476,
+ "grad_norm": 0.048583984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0162,
+ "step": 16557
+ },
+ {
+ "epoch": 52.90095846645367,
+ "grad_norm": 0.05224609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0318,
+ "step": 16558
+ },
+ {
+ "epoch": 52.90415335463259,
+ "grad_norm": 0.05908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0242,
+ "step": 16559
+ },
+ {
+ "epoch": 52.9073482428115,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0228,
+ "step": 16560
+ },
+ {
+ "epoch": 52.91054313099041,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0313,
+ "step": 16561
+ },
+ {
+ "epoch": 52.91373801916933,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0216,
+ "step": 16562
+ },
+ {
+ "epoch": 52.91693290734824,
+ "grad_norm": 0.056396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0252,
+ "step": 16563
+ },
+ {
+ "epoch": 52.92012779552716,
+ "grad_norm": 0.052978515625,
+ "learning_rate": 0.0005,
+ "loss": 1.029,
+ "step": 16564
+ },
+ {
+ "epoch": 52.92332268370607,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0213,
+ "step": 16565
+ },
+ {
+ "epoch": 52.926517571884986,
+ "grad_norm": 0.05712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0368,
+ "step": 16566
+ },
+ {
+ "epoch": 52.9297124600639,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.028,
+ "step": 16567
+ },
+ {
+ "epoch": 52.932907348242814,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0228,
+ "step": 16568
+ },
+ {
+ "epoch": 52.936102236421725,
+ "grad_norm": 0.0947265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0195,
+ "step": 16569
+ },
+ {
+ "epoch": 52.93929712460064,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0343,
+ "step": 16570
+ },
+ {
+ "epoch": 52.94249201277955,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0398,
+ "step": 16571
+ },
+ {
+ "epoch": 52.945686900958464,
+ "grad_norm": 0.1162109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0288,
+ "step": 16572
+ },
+ {
+ "epoch": 52.94888178913738,
+ "grad_norm": 0.055908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0367,
+ "step": 16573
+ },
+ {
+ "epoch": 52.95207667731629,
+ "grad_norm": 0.10986328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0172,
+ "step": 16574
+ },
+ {
+ "epoch": 52.95527156549521,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0184,
+ "step": 16575
+ },
+ {
+ "epoch": 52.95846645367412,
+ "grad_norm": 0.1005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0266,
+ "step": 16576
+ },
+ {
+ "epoch": 52.96166134185304,
+ "grad_norm": 0.0927734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0237,
+ "step": 16577
+ },
+ {
+ "epoch": 52.96485623003195,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0255,
+ "step": 16578
+ },
+ {
+ "epoch": 52.968051118210866,
+ "grad_norm": 0.10205078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0129,
+ "step": 16579
+ },
+ {
+ "epoch": 52.97124600638978,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0399,
+ "step": 16580
+ },
+ {
+ "epoch": 52.97444089456869,
+ "grad_norm": 0.11865234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0316,
+ "step": 16581
+ },
+ {
+ "epoch": 52.977635782747605,
+ "grad_norm": 0.052734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0232,
+ "step": 16582
+ },
+ {
+ "epoch": 52.980830670926515,
+ "grad_norm": 0.1103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.027,
+ "step": 16583
+ },
+ {
+ "epoch": 52.98402555910543,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0169,
+ "step": 16584
+ },
+ {
+ "epoch": 52.98722044728434,
+ "grad_norm": 0.09912109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0244,
+ "step": 16585
+ },
+ {
+ "epoch": 52.99041533546326,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0245,
+ "step": 16586
+ },
+ {
+ "epoch": 52.99361022364217,
+ "grad_norm": 0.07861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0357,
+ "step": 16587
+ },
+ {
+ "epoch": 52.99680511182109,
+ "grad_norm": 0.051025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0267,
+ "step": 16588
+ },
+ {
+ "epoch": 53.0,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0239,
+ "step": 16589
+ },
+ {
+ "epoch": 53.00319488817891,
+ "grad_norm": 0.057373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0271,
+ "step": 16590
+ },
+ {
+ "epoch": 53.00638977635783,
+ "grad_norm": 0.06005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0186,
+ "step": 16591
+ },
+ {
+ "epoch": 53.00958466453674,
+ "grad_norm": 0.05126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0329,
+ "step": 16592
+ },
+ {
+ "epoch": 53.01277955271566,
+ "grad_norm": 0.048095703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0275,
+ "step": 16593
+ },
+ {
+ "epoch": 53.01597444089457,
+ "grad_norm": 0.0615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0206,
+ "step": 16594
+ },
+ {
+ "epoch": 53.019169329073485,
+ "grad_norm": 0.04736328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0315,
+ "step": 16595
+ },
+ {
+ "epoch": 53.022364217252395,
+ "grad_norm": 0.054443359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0291,
+ "step": 16596
+ },
+ {
+ "epoch": 53.02555910543131,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0216,
+ "step": 16597
+ },
+ {
+ "epoch": 53.02875399361022,
+ "grad_norm": 0.055908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0278,
+ "step": 16598
+ },
+ {
+ "epoch": 53.031948881789134,
+ "grad_norm": 0.050537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0187,
+ "step": 16599
+ },
+ {
+ "epoch": 53.03514376996805,
+ "grad_norm": 0.053955078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0393,
+ "step": 16600
+ },
+ {
+ "epoch": 53.03833865814696,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.022,
+ "step": 16601
+ },
+ {
+ "epoch": 53.04153354632588,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0227,
+ "step": 16602
+ },
+ {
+ "epoch": 53.04472843450479,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0225,
+ "step": 16603
+ },
+ {
+ "epoch": 53.04792332268371,
+ "grad_norm": 0.10595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0353,
+ "step": 16604
+ },
+ {
+ "epoch": 53.05111821086262,
+ "grad_norm": 0.062255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0316,
+ "step": 16605
+ },
+ {
+ "epoch": 53.054313099041536,
+ "grad_norm": 0.048095703125,
+ "learning_rate": 0.0005,
+ "loss": 1.032,
+ "step": 16606
+ },
+ {
+ "epoch": 53.05750798722045,
+ "grad_norm": 0.061767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0399,
+ "step": 16607
+ },
+ {
+ "epoch": 53.06070287539936,
+ "grad_norm": 0.060791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0329,
+ "step": 16608
+ },
+ {
+ "epoch": 53.063897763578275,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0184,
+ "step": 16609
+ },
+ {
+ "epoch": 53.067092651757186,
+ "grad_norm": 0.07861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0212,
+ "step": 16610
+ },
+ {
+ "epoch": 53.0702875399361,
+ "grad_norm": 0.0859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0154,
+ "step": 16611
+ },
+ {
+ "epoch": 53.073482428115014,
+ "grad_norm": 0.0478515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0222,
+ "step": 16612
+ },
+ {
+ "epoch": 53.07667731629393,
+ "grad_norm": 0.09814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0285,
+ "step": 16613
+ },
+ {
+ "epoch": 53.07987220447284,
+ "grad_norm": 0.048583984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0353,
+ "step": 16614
+ },
+ {
+ "epoch": 53.08306709265176,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0249,
+ "step": 16615
+ },
+ {
+ "epoch": 53.08626198083067,
+ "grad_norm": 0.048095703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0315,
+ "step": 16616
+ },
+ {
+ "epoch": 53.08945686900959,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0325,
+ "step": 16617
+ },
+ {
+ "epoch": 53.0926517571885,
+ "grad_norm": 0.0458984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0188,
+ "step": 16618
+ },
+ {
+ "epoch": 53.09584664536741,
+ "grad_norm": 0.05078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0349,
+ "step": 16619
+ },
+ {
+ "epoch": 53.09904153354633,
+ "grad_norm": 0.047119140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0208,
+ "step": 16620
+ },
+ {
+ "epoch": 53.10223642172524,
+ "grad_norm": 0.06103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0258,
+ "step": 16621
+ },
+ {
+ "epoch": 53.105431309904155,
+ "grad_norm": 0.058837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0263,
+ "step": 16622
+ },
+ {
+ "epoch": 53.108626198083066,
+ "grad_norm": 0.059326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0262,
+ "step": 16623
+ },
+ {
+ "epoch": 53.11182108626198,
+ "grad_norm": 0.06103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0362,
+ "step": 16624
+ },
+ {
+ "epoch": 53.115015974440894,
+ "grad_norm": 0.05615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0248,
+ "step": 16625
+ },
+ {
+ "epoch": 53.11821086261981,
+ "grad_norm": 0.08447265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0282,
+ "step": 16626
+ },
+ {
+ "epoch": 53.12140575079872,
+ "grad_norm": 0.0478515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0329,
+ "step": 16627
+ },
+ {
+ "epoch": 53.12460063897763,
+ "grad_norm": 0.05615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0249,
+ "step": 16628
+ },
+ {
+ "epoch": 53.12779552715655,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0236,
+ "step": 16629
+ },
+ {
+ "epoch": 53.13099041533546,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0314,
+ "step": 16630
+ },
+ {
+ "epoch": 53.13418530351438,
+ "grad_norm": 0.05712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0246,
+ "step": 16631
+ },
+ {
+ "epoch": 53.13738019169329,
+ "grad_norm": 0.0869140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0282,
+ "step": 16632
+ },
+ {
+ "epoch": 53.14057507987221,
+ "grad_norm": 0.10546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0183,
+ "step": 16633
+ },
+ {
+ "epoch": 53.14376996805112,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.019,
+ "step": 16634
+ },
+ {
+ "epoch": 53.146964856230035,
+ "grad_norm": 0.09326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0292,
+ "step": 16635
+ },
+ {
+ "epoch": 53.150159744408946,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0249,
+ "step": 16636
+ },
+ {
+ "epoch": 53.153354632587856,
+ "grad_norm": 0.05078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0183,
+ "step": 16637
+ },
+ {
+ "epoch": 53.156549520766774,
+ "grad_norm": 0.058837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0182,
+ "step": 16638
+ },
+ {
+ "epoch": 53.159744408945684,
+ "grad_norm": 0.05810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0327,
+ "step": 16639
+ },
+ {
+ "epoch": 53.1629392971246,
+ "grad_norm": 0.05224609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0192,
+ "step": 16640
+ },
+ {
+ "epoch": 53.16613418530351,
+ "grad_norm": 0.061767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0277,
+ "step": 16641
+ },
+ {
+ "epoch": 53.16932907348243,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0331,
+ "step": 16642
+ },
+ {
+ "epoch": 53.17252396166134,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0229,
+ "step": 16643
+ },
+ {
+ "epoch": 53.17571884984026,
+ "grad_norm": 0.07861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0285,
+ "step": 16644
+ },
+ {
+ "epoch": 53.17891373801917,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0382,
+ "step": 16645
+ },
+ {
+ "epoch": 53.18210862619808,
+ "grad_norm": 0.10498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0194,
+ "step": 16646
+ },
+ {
+ "epoch": 53.185303514377,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0348,
+ "step": 16647
+ },
+ {
+ "epoch": 53.18849840255591,
+ "grad_norm": 0.09716796875,
+ "learning_rate": 0.0005,
+ "loss": 1.028,
+ "step": 16648
+ },
+ {
+ "epoch": 53.191693290734825,
+ "grad_norm": 0.057861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0281,
+ "step": 16649
+ },
+ {
+ "epoch": 53.194888178913736,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0254,
+ "step": 16650
+ },
+ {
+ "epoch": 53.198083067092654,
+ "grad_norm": 0.05322265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0346,
+ "step": 16651
+ },
+ {
+ "epoch": 53.201277955271564,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0368,
+ "step": 16652
+ },
+ {
+ "epoch": 53.20447284345048,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0333,
+ "step": 16653
+ },
+ {
+ "epoch": 53.20766773162939,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0276,
+ "step": 16654
+ },
+ {
+ "epoch": 53.21086261980831,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0132,
+ "step": 16655
+ },
+ {
+ "epoch": 53.21405750798722,
+ "grad_norm": 0.059814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0164,
+ "step": 16656
+ },
+ {
+ "epoch": 53.21725239616613,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0261,
+ "step": 16657
+ },
+ {
+ "epoch": 53.22044728434505,
+ "grad_norm": 0.051513671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0151,
+ "step": 16658
+ },
+ {
+ "epoch": 53.22364217252396,
+ "grad_norm": 0.0546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0353,
+ "step": 16659
+ },
+ {
+ "epoch": 53.22683706070288,
+ "grad_norm": 0.053466796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0247,
+ "step": 16660
+ },
+ {
+ "epoch": 53.23003194888179,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0298,
+ "step": 16661
+ },
+ {
+ "epoch": 53.233226837060705,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0348,
+ "step": 16662
+ },
+ {
+ "epoch": 53.236421725239616,
+ "grad_norm": 0.060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0419,
+ "step": 16663
+ },
+ {
+ "epoch": 53.239616613418534,
+ "grad_norm": 0.08056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0206,
+ "step": 16664
+ },
+ {
+ "epoch": 53.242811501597444,
+ "grad_norm": 0.056884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0288,
+ "step": 16665
+ },
+ {
+ "epoch": 53.246006389776355,
+ "grad_norm": 0.062255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0417,
+ "step": 16666
+ },
+ {
+ "epoch": 53.24920127795527,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0304,
+ "step": 16667
+ },
+ {
+ "epoch": 53.25239616613418,
+ "grad_norm": 0.0615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0427,
+ "step": 16668
+ },
+ {
+ "epoch": 53.2555910543131,
+ "grad_norm": 0.087890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0253,
+ "step": 16669
+ },
+ {
+ "epoch": 53.25878594249201,
+ "grad_norm": 0.049072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0309,
+ "step": 16670
+ },
+ {
+ "epoch": 53.26198083067093,
+ "grad_norm": 0.07861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0307,
+ "step": 16671
+ },
+ {
+ "epoch": 53.26517571884984,
+ "grad_norm": 0.046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0345,
+ "step": 16672
+ },
+ {
+ "epoch": 53.26837060702876,
+ "grad_norm": 0.056396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0177,
+ "step": 16673
+ },
+ {
+ "epoch": 53.27156549520767,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0209,
+ "step": 16674
+ },
+ {
+ "epoch": 53.27476038338658,
+ "grad_norm": 0.0546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0268,
+ "step": 16675
+ },
+ {
+ "epoch": 53.277955271565496,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0284,
+ "step": 16676
+ },
+ {
+ "epoch": 53.281150159744406,
+ "grad_norm": 0.053466796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0288,
+ "step": 16677
+ },
+ {
+ "epoch": 53.284345047923324,
+ "grad_norm": 0.057861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0284,
+ "step": 16678
+ },
+ {
+ "epoch": 53.287539936102235,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0389,
+ "step": 16679
+ },
+ {
+ "epoch": 53.29073482428115,
+ "grad_norm": 0.09130859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0293,
+ "step": 16680
+ },
+ {
+ "epoch": 53.29392971246006,
+ "grad_norm": 0.045654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0288,
+ "step": 16681
+ },
+ {
+ "epoch": 53.29712460063898,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.023,
+ "step": 16682
+ },
+ {
+ "epoch": 53.30031948881789,
+ "grad_norm": 0.061767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0428,
+ "step": 16683
+ },
+ {
+ "epoch": 53.3035143769968,
+ "grad_norm": 0.12109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0324,
+ "step": 16684
+ },
+ {
+ "epoch": 53.30670926517572,
+ "grad_norm": 0.060302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0232,
+ "step": 16685
+ },
+ {
+ "epoch": 53.30990415335463,
+ "grad_norm": 0.123046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0237,
+ "step": 16686
+ },
+ {
+ "epoch": 53.31309904153355,
+ "grad_norm": 0.09716796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0126,
+ "step": 16687
+ },
+ {
+ "epoch": 53.31629392971246,
+ "grad_norm": 0.103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0344,
+ "step": 16688
+ },
+ {
+ "epoch": 53.319488817891376,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0219,
+ "step": 16689
+ },
+ {
+ "epoch": 53.322683706070286,
+ "grad_norm": 0.061279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0427,
+ "step": 16690
+ },
+ {
+ "epoch": 53.325878594249204,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0287,
+ "step": 16691
+ },
+ {
+ "epoch": 53.329073482428115,
+ "grad_norm": 0.05517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.042,
+ "step": 16692
+ },
+ {
+ "epoch": 53.33226837060703,
+ "grad_norm": 0.05029296875,
+ "learning_rate": 0.0005,
+ "loss": 1.029,
+ "step": 16693
+ },
+ {
+ "epoch": 53.33546325878594,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0212,
+ "step": 16694
+ },
+ {
+ "epoch": 53.33865814696485,
+ "grad_norm": 0.058349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.029,
+ "step": 16695
+ },
+ {
+ "epoch": 53.34185303514377,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0317,
+ "step": 16696
+ },
+ {
+ "epoch": 53.34504792332268,
+ "grad_norm": 0.056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0298,
+ "step": 16697
+ },
+ {
+ "epoch": 53.3482428115016,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0294,
+ "step": 16698
+ },
+ {
+ "epoch": 53.35143769968051,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0318,
+ "step": 16699
+ },
+ {
+ "epoch": 53.35463258785943,
+ "grad_norm": 0.09130859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0257,
+ "step": 16700
+ },
+ {
+ "epoch": 53.35782747603834,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0313,
+ "step": 16701
+ },
+ {
+ "epoch": 53.361022364217256,
+ "grad_norm": 0.0498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.027,
+ "step": 16702
+ },
+ {
+ "epoch": 53.364217252396166,
+ "grad_norm": 0.126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0196,
+ "step": 16703
+ },
+ {
+ "epoch": 53.36741214057508,
+ "grad_norm": 0.050537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0245,
+ "step": 16704
+ },
+ {
+ "epoch": 53.370607028753994,
+ "grad_norm": 0.080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0217,
+ "step": 16705
+ },
+ {
+ "epoch": 53.373801916932905,
+ "grad_norm": 0.052978515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0226,
+ "step": 16706
+ },
+ {
+ "epoch": 53.37699680511182,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0185,
+ "step": 16707
+ },
+ {
+ "epoch": 53.38019169329073,
+ "grad_norm": 0.05419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0247,
+ "step": 16708
+ },
+ {
+ "epoch": 53.38338658146965,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0363,
+ "step": 16709
+ },
+ {
+ "epoch": 53.38658146964856,
+ "grad_norm": 0.051513671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0369,
+ "step": 16710
+ },
+ {
+ "epoch": 53.38977635782748,
+ "grad_norm": 0.058349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0154,
+ "step": 16711
+ },
+ {
+ "epoch": 53.39297124600639,
+ "grad_norm": 0.05078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0232,
+ "step": 16712
+ },
+ {
+ "epoch": 53.3961661341853,
+ "grad_norm": 0.056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0345,
+ "step": 16713
+ },
+ {
+ "epoch": 53.39936102236422,
+ "grad_norm": 0.056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0186,
+ "step": 16714
+ },
+ {
+ "epoch": 53.40255591054313,
+ "grad_norm": 0.04345703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0174,
+ "step": 16715
+ },
+ {
+ "epoch": 53.405750798722046,
+ "grad_norm": 0.055908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0248,
+ "step": 16716
+ },
+ {
+ "epoch": 53.40894568690096,
+ "grad_norm": 0.056884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0204,
+ "step": 16717
+ },
+ {
+ "epoch": 53.412140575079874,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0187,
+ "step": 16718
+ },
+ {
+ "epoch": 53.415335463258785,
+ "grad_norm": 0.048095703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0288,
+ "step": 16719
+ },
+ {
+ "epoch": 53.4185303514377,
+ "grad_norm": 0.10107421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0303,
+ "step": 16720
+ },
+ {
+ "epoch": 53.42172523961661,
+ "grad_norm": 0.080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0184,
+ "step": 16721
+ },
+ {
+ "epoch": 53.424920127795524,
+ "grad_norm": 0.1005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0259,
+ "step": 16722
+ },
+ {
+ "epoch": 53.42811501597444,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0184,
+ "step": 16723
+ },
+ {
+ "epoch": 53.43130990415335,
+ "grad_norm": 0.083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0271,
+ "step": 16724
+ },
+ {
+ "epoch": 53.43450479233227,
+ "grad_norm": 0.048583984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0321,
+ "step": 16725
+ },
+ {
+ "epoch": 53.43769968051118,
+ "grad_norm": 0.08056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0308,
+ "step": 16726
+ },
+ {
+ "epoch": 53.4408945686901,
+ "grad_norm": 0.08447265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0225,
+ "step": 16727
+ },
+ {
+ "epoch": 53.44408945686901,
+ "grad_norm": 0.1591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0238,
+ "step": 16728
+ },
+ {
+ "epoch": 53.447284345047926,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.03,
+ "step": 16729
+ },
+ {
+ "epoch": 53.45047923322684,
+ "grad_norm": 0.1376953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0361,
+ "step": 16730
+ },
+ {
+ "epoch": 53.453674121405754,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0284,
+ "step": 16731
+ },
+ {
+ "epoch": 53.456869009584665,
+ "grad_norm": 0.09619140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0465,
+ "step": 16732
+ },
+ {
+ "epoch": 53.460063897763575,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0125,
+ "step": 16733
+ },
+ {
+ "epoch": 53.46325878594249,
+ "grad_norm": 0.08837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0207,
+ "step": 16734
+ },
+ {
+ "epoch": 53.466453674121404,
+ "grad_norm": 0.095703125,
+ "learning_rate": 0.0005,
+ "loss": 1.021,
+ "step": 16735
+ },
+ {
+ "epoch": 53.46964856230032,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0228,
+ "step": 16736
+ },
+ {
+ "epoch": 53.47284345047923,
+ "grad_norm": 0.10302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0214,
+ "step": 16737
+ },
+ {
+ "epoch": 53.47603833865815,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0254,
+ "step": 16738
+ },
+ {
+ "epoch": 53.47923322683706,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0264,
+ "step": 16739
+ },
+ {
+ "epoch": 53.48242811501598,
+ "grad_norm": 0.060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0243,
+ "step": 16740
+ },
+ {
+ "epoch": 53.48562300319489,
+ "grad_norm": 0.060302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0232,
+ "step": 16741
+ },
+ {
+ "epoch": 53.4888178913738,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0248,
+ "step": 16742
+ },
+ {
+ "epoch": 53.49201277955272,
+ "grad_norm": 0.051025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.03,
+ "step": 16743
+ },
+ {
+ "epoch": 53.49520766773163,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0247,
+ "step": 16744
+ },
+ {
+ "epoch": 53.498402555910545,
+ "grad_norm": 0.054443359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0207,
+ "step": 16745
+ },
+ {
+ "epoch": 53.501597444089455,
+ "grad_norm": 0.08056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0211,
+ "step": 16746
+ },
+ {
+ "epoch": 53.50479233226837,
+ "grad_norm": 0.0498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0281,
+ "step": 16747
+ },
+ {
+ "epoch": 53.50798722044728,
+ "grad_norm": 0.08935546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0254,
+ "step": 16748
+ },
+ {
+ "epoch": 53.5111821086262,
+ "grad_norm": 0.05126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0012,
+ "step": 16749
+ },
+ {
+ "epoch": 53.51437699680511,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0217,
+ "step": 16750
+ },
+ {
+ "epoch": 53.51757188498402,
+ "grad_norm": 0.053466796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0272,
+ "step": 16751
+ },
+ {
+ "epoch": 53.52076677316294,
+ "grad_norm": 0.057861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0252,
+ "step": 16752
+ },
+ {
+ "epoch": 53.52396166134185,
+ "grad_norm": 0.058349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.021,
+ "step": 16753
+ },
+ {
+ "epoch": 53.52715654952077,
+ "grad_norm": 0.055908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0223,
+ "step": 16754
+ },
+ {
+ "epoch": 53.53035143769968,
+ "grad_norm": 0.0546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0287,
+ "step": 16755
+ },
+ {
+ "epoch": 53.533546325878596,
+ "grad_norm": 0.061279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0297,
+ "step": 16756
+ },
+ {
+ "epoch": 53.53674121405751,
+ "grad_norm": 0.062255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0249,
+ "step": 16757
+ },
+ {
+ "epoch": 53.539936102236425,
+ "grad_norm": 0.05126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0263,
+ "step": 16758
+ },
+ {
+ "epoch": 53.543130990415335,
+ "grad_norm": 0.051513671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0351,
+ "step": 16759
+ },
+ {
+ "epoch": 53.546325878594246,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0269,
+ "step": 16760
+ },
+ {
+ "epoch": 53.54952076677316,
+ "grad_norm": 0.07861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0284,
+ "step": 16761
+ },
+ {
+ "epoch": 53.552715654952074,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0202,
+ "step": 16762
+ },
+ {
+ "epoch": 53.55591054313099,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0313,
+ "step": 16763
+ },
+ {
+ "epoch": 53.5591054313099,
+ "grad_norm": 0.054443359375,
+ "learning_rate": 0.0005,
+ "loss": 1.028,
+ "step": 16764
+ },
+ {
+ "epoch": 53.56230031948882,
+ "grad_norm": 0.048583984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0158,
+ "step": 16765
+ },
+ {
+ "epoch": 53.56549520766773,
+ "grad_norm": 0.048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0289,
+ "step": 16766
+ },
+ {
+ "epoch": 53.56869009584665,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0228,
+ "step": 16767
+ },
+ {
+ "epoch": 53.57188498402556,
+ "grad_norm": 0.052978515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0296,
+ "step": 16768
+ },
+ {
+ "epoch": 53.575079872204476,
+ "grad_norm": 0.09423828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0282,
+ "step": 16769
+ },
+ {
+ "epoch": 53.57827476038339,
+ "grad_norm": 0.04833984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0309,
+ "step": 16770
+ },
+ {
+ "epoch": 53.5814696485623,
+ "grad_norm": 0.058837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0245,
+ "step": 16771
+ },
+ {
+ "epoch": 53.584664536741215,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.032,
+ "step": 16772
+ },
+ {
+ "epoch": 53.587859424920126,
+ "grad_norm": 0.0458984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0351,
+ "step": 16773
+ },
+ {
+ "epoch": 53.59105431309904,
+ "grad_norm": 0.059814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0246,
+ "step": 16774
+ },
+ {
+ "epoch": 53.594249201277954,
+ "grad_norm": 0.048583984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0306,
+ "step": 16775
+ },
+ {
+ "epoch": 53.59744408945687,
+ "grad_norm": 0.05908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0302,
+ "step": 16776
+ },
+ {
+ "epoch": 53.60063897763578,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0264,
+ "step": 16777
+ },
+ {
+ "epoch": 53.6038338658147,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0265,
+ "step": 16778
+ },
+ {
+ "epoch": 53.60702875399361,
+ "grad_norm": 0.044189453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0306,
+ "step": 16779
+ },
+ {
+ "epoch": 53.61022364217252,
+ "grad_norm": 0.0498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0178,
+ "step": 16780
+ },
+ {
+ "epoch": 53.61341853035144,
+ "grad_norm": 0.051025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0444,
+ "step": 16781
+ },
+ {
+ "epoch": 53.61661341853035,
+ "grad_norm": 0.045654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.034,
+ "step": 16782
+ },
+ {
+ "epoch": 53.61980830670927,
+ "grad_norm": 0.0576171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0323,
+ "step": 16783
+ },
+ {
+ "epoch": 53.62300319488818,
+ "grad_norm": 0.050537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0305,
+ "step": 16784
+ },
+ {
+ "epoch": 53.626198083067095,
+ "grad_norm": 0.053955078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0248,
+ "step": 16785
+ },
+ {
+ "epoch": 53.629392971246006,
+ "grad_norm": 0.056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0395,
+ "step": 16786
+ },
+ {
+ "epoch": 53.63258785942492,
+ "grad_norm": 0.051025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.032,
+ "step": 16787
+ },
+ {
+ "epoch": 53.635782747603834,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0337,
+ "step": 16788
+ },
+ {
+ "epoch": 53.638977635782744,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0103,
+ "step": 16789
+ },
+ {
+ "epoch": 53.64217252396166,
+ "grad_norm": 0.052978515625,
+ "learning_rate": 0.0005,
+ "loss": 1.028,
+ "step": 16790
+ },
+ {
+ "epoch": 53.64536741214057,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0241,
+ "step": 16791
+ },
+ {
+ "epoch": 53.64856230031949,
+ "grad_norm": 0.07861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0216,
+ "step": 16792
+ },
+ {
+ "epoch": 53.6517571884984,
+ "grad_norm": 0.09326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0235,
+ "step": 16793
+ },
+ {
+ "epoch": 53.65495207667732,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0353,
+ "step": 16794
+ },
+ {
+ "epoch": 53.65814696485623,
+ "grad_norm": 0.056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0218,
+ "step": 16795
+ },
+ {
+ "epoch": 53.66134185303515,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0295,
+ "step": 16796
+ },
+ {
+ "epoch": 53.66453674121406,
+ "grad_norm": 0.08154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0241,
+ "step": 16797
+ },
+ {
+ "epoch": 53.66773162939297,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.043,
+ "step": 16798
+ },
+ {
+ "epoch": 53.670926517571885,
+ "grad_norm": 0.0810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0312,
+ "step": 16799
+ },
+ {
+ "epoch": 53.674121405750796,
+ "grad_norm": 0.087890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0231,
+ "step": 16800
+ },
+ {
+ "epoch": 53.677316293929714,
+ "grad_norm": 0.158203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0283,
+ "step": 16801
+ },
+ {
+ "epoch": 53.680511182108624,
+ "grad_norm": 0.08154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0334,
+ "step": 16802
+ },
+ {
+ "epoch": 53.68370607028754,
+ "grad_norm": 0.10986328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0236,
+ "step": 16803
+ },
+ {
+ "epoch": 53.68690095846645,
+ "grad_norm": 0.052734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0357,
+ "step": 16804
+ },
+ {
+ "epoch": 53.69009584664537,
+ "grad_norm": 0.0810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0346,
+ "step": 16805
+ },
+ {
+ "epoch": 53.69329073482428,
+ "grad_norm": 0.0517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0245,
+ "step": 16806
+ },
+ {
+ "epoch": 53.6964856230032,
+ "grad_norm": 0.083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0316,
+ "step": 16807
+ },
+ {
+ "epoch": 53.69968051118211,
+ "grad_norm": 0.054443359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0143,
+ "step": 16808
+ },
+ {
+ "epoch": 53.70287539936102,
+ "grad_norm": 0.09228515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0133,
+ "step": 16809
+ },
+ {
+ "epoch": 53.70607028753994,
+ "grad_norm": 0.09375,
+ "learning_rate": 0.0005,
+ "loss": 1.0222,
+ "step": 16810
+ },
+ {
+ "epoch": 53.70926517571885,
+ "grad_norm": 0.083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0313,
+ "step": 16811
+ },
+ {
+ "epoch": 53.712460063897765,
+ "grad_norm": 0.0908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0264,
+ "step": 16812
+ },
+ {
+ "epoch": 53.715654952076676,
+ "grad_norm": 0.08056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0221,
+ "step": 16813
+ },
+ {
+ "epoch": 53.718849840255594,
+ "grad_norm": 0.10791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0344,
+ "step": 16814
+ },
+ {
+ "epoch": 53.722044728434504,
+ "grad_norm": 0.047607421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0295,
+ "step": 16815
+ },
+ {
+ "epoch": 53.72523961661342,
+ "grad_norm": 0.111328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0211,
+ "step": 16816
+ },
+ {
+ "epoch": 53.72843450479233,
+ "grad_norm": 0.06103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0283,
+ "step": 16817
+ },
+ {
+ "epoch": 53.73162939297124,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0281,
+ "step": 16818
+ },
+ {
+ "epoch": 53.73482428115016,
+ "grad_norm": 0.06201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0145,
+ "step": 16819
+ },
+ {
+ "epoch": 53.73801916932907,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0297,
+ "step": 16820
+ },
+ {
+ "epoch": 53.74121405750799,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0245,
+ "step": 16821
+ },
+ {
+ "epoch": 53.7444089456869,
+ "grad_norm": 0.10546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0239,
+ "step": 16822
+ },
+ {
+ "epoch": 53.74760383386582,
+ "grad_norm": 0.056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0204,
+ "step": 16823
+ },
+ {
+ "epoch": 53.75079872204473,
+ "grad_norm": 0.087890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0223,
+ "step": 16824
+ },
+ {
+ "epoch": 53.753993610223645,
+ "grad_norm": 0.060302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0256,
+ "step": 16825
+ },
+ {
+ "epoch": 53.757188498402556,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.025,
+ "step": 16826
+ },
+ {
+ "epoch": 53.760383386581466,
+ "grad_norm": 0.056396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0285,
+ "step": 16827
+ },
+ {
+ "epoch": 53.763578274760384,
+ "grad_norm": 0.04736328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0327,
+ "step": 16828
+ },
+ {
+ "epoch": 53.766773162939295,
+ "grad_norm": 0.0927734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0297,
+ "step": 16829
+ },
+ {
+ "epoch": 53.76996805111821,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0173,
+ "step": 16830
+ },
+ {
+ "epoch": 53.77316293929712,
+ "grad_norm": 0.0869140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0327,
+ "step": 16831
+ },
+ {
+ "epoch": 53.77635782747604,
+ "grad_norm": 0.0927734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0297,
+ "step": 16832
+ },
+ {
+ "epoch": 53.77955271565495,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0363,
+ "step": 16833
+ },
+ {
+ "epoch": 53.78274760383387,
+ "grad_norm": 0.051513671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0356,
+ "step": 16834
+ },
+ {
+ "epoch": 53.78594249201278,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.034,
+ "step": 16835
+ },
+ {
+ "epoch": 53.78913738019169,
+ "grad_norm": 0.056884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0287,
+ "step": 16836
+ },
+ {
+ "epoch": 53.79233226837061,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0329,
+ "step": 16837
+ },
+ {
+ "epoch": 53.79552715654952,
+ "grad_norm": 0.04052734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0183,
+ "step": 16838
+ },
+ {
+ "epoch": 53.798722044728436,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.016,
+ "step": 16839
+ },
+ {
+ "epoch": 53.801916932907346,
+ "grad_norm": 0.04296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0298,
+ "step": 16840
+ },
+ {
+ "epoch": 53.805111821086264,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0268,
+ "step": 16841
+ },
+ {
+ "epoch": 53.808306709265175,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0256,
+ "step": 16842
+ },
+ {
+ "epoch": 53.81150159744409,
+ "grad_norm": 0.05517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0159,
+ "step": 16843
+ },
+ {
+ "epoch": 53.814696485623,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0273,
+ "step": 16844
+ },
+ {
+ "epoch": 53.81789137380191,
+ "grad_norm": 0.0517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0231,
+ "step": 16845
+ },
+ {
+ "epoch": 53.82108626198083,
+ "grad_norm": 0.056396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0288,
+ "step": 16846
+ },
+ {
+ "epoch": 53.82428115015974,
+ "grad_norm": 0.0576171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0346,
+ "step": 16847
+ },
+ {
+ "epoch": 53.82747603833866,
+ "grad_norm": 0.06201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0264,
+ "step": 16848
+ },
+ {
+ "epoch": 53.83067092651757,
+ "grad_norm": 0.0869140625,
+ "learning_rate": 0.0005,
+ "loss": 1.026,
+ "step": 16849
+ },
+ {
+ "epoch": 53.83386581469649,
+ "grad_norm": 0.08056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0403,
+ "step": 16850
+ },
+ {
+ "epoch": 53.8370607028754,
+ "grad_norm": 0.12451171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0191,
+ "step": 16851
+ },
+ {
+ "epoch": 53.840255591054316,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0363,
+ "step": 16852
+ },
+ {
+ "epoch": 53.843450479233226,
+ "grad_norm": 0.0546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0142,
+ "step": 16853
+ },
+ {
+ "epoch": 53.846645367412144,
+ "grad_norm": 0.052978515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0337,
+ "step": 16854
+ },
+ {
+ "epoch": 53.849840255591054,
+ "grad_norm": 0.09033203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0232,
+ "step": 16855
+ },
+ {
+ "epoch": 53.853035143769965,
+ "grad_norm": 0.115234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0281,
+ "step": 16856
+ },
+ {
+ "epoch": 53.85623003194888,
+ "grad_norm": 0.0517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0229,
+ "step": 16857
+ },
+ {
+ "epoch": 53.85942492012779,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0271,
+ "step": 16858
+ },
+ {
+ "epoch": 53.86261980830671,
+ "grad_norm": 0.056884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0252,
+ "step": 16859
+ },
+ {
+ "epoch": 53.86581469648562,
+ "grad_norm": 0.05615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0261,
+ "step": 16860
+ },
+ {
+ "epoch": 53.86900958466454,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0297,
+ "step": 16861
+ },
+ {
+ "epoch": 53.87220447284345,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0144,
+ "step": 16862
+ },
+ {
+ "epoch": 53.87539936102237,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0337,
+ "step": 16863
+ },
+ {
+ "epoch": 53.87859424920128,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0248,
+ "step": 16864
+ },
+ {
+ "epoch": 53.88178913738019,
+ "grad_norm": 0.04638671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0264,
+ "step": 16865
+ },
+ {
+ "epoch": 53.884984025559106,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0398,
+ "step": 16866
+ },
+ {
+ "epoch": 53.88817891373802,
+ "grad_norm": 0.05126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0208,
+ "step": 16867
+ },
+ {
+ "epoch": 53.891373801916934,
+ "grad_norm": 0.05712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0259,
+ "step": 16868
+ },
+ {
+ "epoch": 53.894568690095845,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0335,
+ "step": 16869
+ },
+ {
+ "epoch": 53.89776357827476,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0232,
+ "step": 16870
+ },
+ {
+ "epoch": 53.90095846645367,
+ "grad_norm": 0.049072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.017,
+ "step": 16871
+ },
+ {
+ "epoch": 53.90415335463259,
+ "grad_norm": 0.055908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0228,
+ "step": 16872
+ },
+ {
+ "epoch": 53.9073482428115,
+ "grad_norm": 0.0498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0283,
+ "step": 16873
+ },
+ {
+ "epoch": 53.91054313099041,
+ "grad_norm": 0.056884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0208,
+ "step": 16874
+ },
+ {
+ "epoch": 53.91373801916933,
+ "grad_norm": 0.052978515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0284,
+ "step": 16875
+ },
+ {
+ "epoch": 53.91693290734824,
+ "grad_norm": 0.04296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0247,
+ "step": 16876
+ },
+ {
+ "epoch": 53.92012779552716,
+ "grad_norm": 0.08251953125,
+ "learning_rate": 0.0005,
+ "loss": 1.027,
+ "step": 16877
+ },
+ {
+ "epoch": 53.92332268370607,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0281,
+ "step": 16878
+ },
+ {
+ "epoch": 53.926517571884986,
+ "grad_norm": 0.058837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.036,
+ "step": 16879
+ },
+ {
+ "epoch": 53.9297124600639,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0194,
+ "step": 16880
+ },
+ {
+ "epoch": 53.932907348242814,
+ "grad_norm": 0.0537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0321,
+ "step": 16881
+ },
+ {
+ "epoch": 53.936102236421725,
+ "grad_norm": 0.07861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0222,
+ "step": 16882
+ },
+ {
+ "epoch": 53.93929712460064,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.025,
+ "step": 16883
+ },
+ {
+ "epoch": 53.94249201277955,
+ "grad_norm": 0.057373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0269,
+ "step": 16884
+ },
+ {
+ "epoch": 53.945686900958464,
+ "grad_norm": 0.058837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0197,
+ "step": 16885
+ },
+ {
+ "epoch": 53.94888178913738,
+ "grad_norm": 0.060791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0246,
+ "step": 16886
+ },
+ {
+ "epoch": 53.95207667731629,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0255,
+ "step": 16887
+ },
+ {
+ "epoch": 53.95527156549521,
+ "grad_norm": 0.058837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0278,
+ "step": 16888
+ },
+ {
+ "epoch": 53.95846645367412,
+ "grad_norm": 0.0986328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0269,
+ "step": 16889
+ },
+ {
+ "epoch": 53.96166134185304,
+ "grad_norm": 0.053466796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0286,
+ "step": 16890
+ },
+ {
+ "epoch": 53.96485623003195,
+ "grad_norm": 0.109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0345,
+ "step": 16891
+ },
+ {
+ "epoch": 53.968051118210866,
+ "grad_norm": 0.0517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0191,
+ "step": 16892
+ },
+ {
+ "epoch": 53.97124600638978,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0318,
+ "step": 16893
+ },
+ {
+ "epoch": 53.97444089456869,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0241,
+ "step": 16894
+ },
+ {
+ "epoch": 53.977635782747605,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0288,
+ "step": 16895
+ },
+ {
+ "epoch": 53.980830670926515,
+ "grad_norm": 0.052001953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0337,
+ "step": 16896
+ },
+ {
+ "epoch": 53.98402555910543,
+ "grad_norm": 0.062255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0241,
+ "step": 16897
+ },
+ {
+ "epoch": 53.98722044728434,
+ "grad_norm": 0.052490234375,
+ "learning_rate": 0.0005,
+ "loss": 1.013,
+ "step": 16898
+ },
+ {
+ "epoch": 53.99041533546326,
+ "grad_norm": 0.051513671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0347,
+ "step": 16899
+ },
+ {
+ "epoch": 53.99361022364217,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0276,
+ "step": 16900
+ },
+ {
+ "epoch": 53.99680511182109,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0296,
+ "step": 16901
+ },
+ {
+ "epoch": 54.0,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0276,
+ "step": 16902
+ },
+ {
+ "epoch": 54.00319488817891,
+ "grad_norm": 0.08447265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0236,
+ "step": 16903
+ },
+ {
+ "epoch": 54.00638977635783,
+ "grad_norm": 0.09326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0226,
+ "step": 16904
+ },
+ {
+ "epoch": 54.00958466453674,
+ "grad_norm": 0.062255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0262,
+ "step": 16905
+ },
+ {
+ "epoch": 54.01277955271566,
+ "grad_norm": 0.0830078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0204,
+ "step": 16906
+ },
+ {
+ "epoch": 54.01597444089457,
+ "grad_norm": 0.058349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0302,
+ "step": 16907
+ },
+ {
+ "epoch": 54.019169329073485,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.024,
+ "step": 16908
+ },
+ {
+ "epoch": 54.022364217252395,
+ "grad_norm": 0.08203125,
+ "learning_rate": 0.0005,
+ "loss": 1.041,
+ "step": 16909
+ },
+ {
+ "epoch": 54.02555910543131,
+ "grad_norm": 0.1279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0365,
+ "step": 16910
+ },
+ {
+ "epoch": 54.02875399361022,
+ "grad_norm": 0.0927734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0373,
+ "step": 16911
+ },
+ {
+ "epoch": 54.031948881789134,
+ "grad_norm": 0.1611328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0521,
+ "step": 16912
+ },
+ {
+ "epoch": 54.03514376996805,
+ "grad_norm": 0.0888671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0266,
+ "step": 16913
+ },
+ {
+ "epoch": 54.03833865814696,
+ "grad_norm": 0.1572265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0166,
+ "step": 16914
+ },
+ {
+ "epoch": 54.04153354632588,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.032,
+ "step": 16915
+ },
+ {
+ "epoch": 54.04472843450479,
+ "grad_norm": 0.1474609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0172,
+ "step": 16916
+ },
+ {
+ "epoch": 54.04792332268371,
+ "grad_norm": 0.052734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0305,
+ "step": 16917
+ },
+ {
+ "epoch": 54.05111821086262,
+ "grad_norm": 0.12890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0329,
+ "step": 16918
+ },
+ {
+ "epoch": 54.054313099041536,
+ "grad_norm": 0.05810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0238,
+ "step": 16919
+ },
+ {
+ "epoch": 54.05750798722045,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0252,
+ "step": 16920
+ },
+ {
+ "epoch": 54.06070287539936,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0221,
+ "step": 16921
+ },
+ {
+ "epoch": 54.063897763578275,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0332,
+ "step": 16922
+ },
+ {
+ "epoch": 54.067092651757186,
+ "grad_norm": 0.06201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0285,
+ "step": 16923
+ },
+ {
+ "epoch": 54.0702875399361,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0172,
+ "step": 16924
+ },
+ {
+ "epoch": 54.073482428115014,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0246,
+ "step": 16925
+ },
+ {
+ "epoch": 54.07667731629393,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0187,
+ "step": 16926
+ },
+ {
+ "epoch": 54.07987220447284,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0371,
+ "step": 16927
+ },
+ {
+ "epoch": 54.08306709265176,
+ "grad_norm": 0.11572265625,
+ "learning_rate": 0.0005,
+ "loss": 1.028,
+ "step": 16928
+ },
+ {
+ "epoch": 54.08626198083067,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0211,
+ "step": 16929
+ },
+ {
+ "epoch": 54.08945686900959,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0156,
+ "step": 16930
+ },
+ {
+ "epoch": 54.0926517571885,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.029,
+ "step": 16931
+ },
+ {
+ "epoch": 54.09584664536741,
+ "grad_norm": 0.061767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0253,
+ "step": 16932
+ },
+ {
+ "epoch": 54.09904153354633,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0277,
+ "step": 16933
+ },
+ {
+ "epoch": 54.10223642172524,
+ "grad_norm": 0.09521484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0256,
+ "step": 16934
+ },
+ {
+ "epoch": 54.105431309904155,
+ "grad_norm": 0.060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0364,
+ "step": 16935
+ },
+ {
+ "epoch": 54.108626198083066,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0337,
+ "step": 16936
+ },
+ {
+ "epoch": 54.11182108626198,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0261,
+ "step": 16937
+ },
+ {
+ "epoch": 54.115015974440894,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0237,
+ "step": 16938
+ },
+ {
+ "epoch": 54.11821086261981,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0317,
+ "step": 16939
+ },
+ {
+ "epoch": 54.12140575079872,
+ "grad_norm": 0.05517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0239,
+ "step": 16940
+ },
+ {
+ "epoch": 54.12460063897763,
+ "grad_norm": 0.047119140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0378,
+ "step": 16941
+ },
+ {
+ "epoch": 54.12779552715655,
+ "grad_norm": 0.06103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0161,
+ "step": 16942
+ },
+ {
+ "epoch": 54.13099041533546,
+ "grad_norm": 0.061279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0256,
+ "step": 16943
+ },
+ {
+ "epoch": 54.13418530351438,
+ "grad_norm": 0.041748046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0248,
+ "step": 16944
+ },
+ {
+ "epoch": 54.13738019169329,
+ "grad_norm": 0.046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0196,
+ "step": 16945
+ },
+ {
+ "epoch": 54.14057507987221,
+ "grad_norm": 0.046630859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0313,
+ "step": 16946
+ },
+ {
+ "epoch": 54.14376996805112,
+ "grad_norm": 0.051513671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0183,
+ "step": 16947
+ },
+ {
+ "epoch": 54.146964856230035,
+ "grad_norm": 0.052978515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0228,
+ "step": 16948
+ },
+ {
+ "epoch": 54.150159744408946,
+ "grad_norm": 0.052978515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0292,
+ "step": 16949
+ },
+ {
+ "epoch": 54.153354632587856,
+ "grad_norm": 0.0546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0321,
+ "step": 16950
+ },
+ {
+ "epoch": 54.156549520766774,
+ "grad_norm": 0.056884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0307,
+ "step": 16951
+ },
+ {
+ "epoch": 54.159744408945684,
+ "grad_norm": 0.056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0228,
+ "step": 16952
+ },
+ {
+ "epoch": 54.1629392971246,
+ "grad_norm": 0.0546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0197,
+ "step": 16953
+ },
+ {
+ "epoch": 54.16613418530351,
+ "grad_norm": 0.0498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0325,
+ "step": 16954
+ },
+ {
+ "epoch": 54.16932907348243,
+ "grad_norm": 0.05712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.016,
+ "step": 16955
+ },
+ {
+ "epoch": 54.17252396166134,
+ "grad_norm": 0.045166015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0268,
+ "step": 16956
+ },
+ {
+ "epoch": 54.17571884984026,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0341,
+ "step": 16957
+ },
+ {
+ "epoch": 54.17891373801917,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0226,
+ "step": 16958
+ },
+ {
+ "epoch": 54.18210862619808,
+ "grad_norm": 0.050048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0272,
+ "step": 16959
+ },
+ {
+ "epoch": 54.185303514377,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0389,
+ "step": 16960
+ },
+ {
+ "epoch": 54.18849840255591,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0293,
+ "step": 16961
+ },
+ {
+ "epoch": 54.191693290734825,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0316,
+ "step": 16962
+ },
+ {
+ "epoch": 54.194888178913736,
+ "grad_norm": 0.05322265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0287,
+ "step": 16963
+ },
+ {
+ "epoch": 54.198083067092654,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0363,
+ "step": 16964
+ },
+ {
+ "epoch": 54.201277955271564,
+ "grad_norm": 0.060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0206,
+ "step": 16965
+ },
+ {
+ "epoch": 54.20447284345048,
+ "grad_norm": 0.05126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0261,
+ "step": 16966
+ },
+ {
+ "epoch": 54.20766773162939,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0282,
+ "step": 16967
+ },
+ {
+ "epoch": 54.21086261980831,
+ "grad_norm": 0.055419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0278,
+ "step": 16968
+ },
+ {
+ "epoch": 54.21405750798722,
+ "grad_norm": 0.049072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0292,
+ "step": 16969
+ },
+ {
+ "epoch": 54.21725239616613,
+ "grad_norm": 0.08056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0267,
+ "step": 16970
+ },
+ {
+ "epoch": 54.22044728434505,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0257,
+ "step": 16971
+ },
+ {
+ "epoch": 54.22364217252396,
+ "grad_norm": 0.10498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0229,
+ "step": 16972
+ },
+ {
+ "epoch": 54.22683706070288,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.025,
+ "step": 16973
+ },
+ {
+ "epoch": 54.23003194888179,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0364,
+ "step": 16974
+ },
+ {
+ "epoch": 54.233226837060705,
+ "grad_norm": 0.1005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0175,
+ "step": 16975
+ },
+ {
+ "epoch": 54.236421725239616,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0264,
+ "step": 16976
+ },
+ {
+ "epoch": 54.239616613418534,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.036,
+ "step": 16977
+ },
+ {
+ "epoch": 54.242811501597444,
+ "grad_norm": 0.0810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0266,
+ "step": 16978
+ },
+ {
+ "epoch": 54.246006389776355,
+ "grad_norm": 0.04931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0339,
+ "step": 16979
+ },
+ {
+ "epoch": 54.24920127795527,
+ "grad_norm": 0.08740234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0234,
+ "step": 16980
+ },
+ {
+ "epoch": 54.25239616613418,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0326,
+ "step": 16981
+ },
+ {
+ "epoch": 54.2555910543131,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0217,
+ "step": 16982
+ },
+ {
+ "epoch": 54.25878594249201,
+ "grad_norm": 0.05517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0275,
+ "step": 16983
+ },
+ {
+ "epoch": 54.26198083067093,
+ "grad_norm": 0.05615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0281,
+ "step": 16984
+ },
+ {
+ "epoch": 54.26517571884984,
+ "grad_norm": 0.0517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0268,
+ "step": 16985
+ },
+ {
+ "epoch": 54.26837060702876,
+ "grad_norm": 0.057373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0296,
+ "step": 16986
+ },
+ {
+ "epoch": 54.27156549520767,
+ "grad_norm": 0.058349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0289,
+ "step": 16987
+ },
+ {
+ "epoch": 54.27476038338658,
+ "grad_norm": 0.049072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.03,
+ "step": 16988
+ },
+ {
+ "epoch": 54.277955271565496,
+ "grad_norm": 0.048583984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0325,
+ "step": 16989
+ },
+ {
+ "epoch": 54.281150159744406,
+ "grad_norm": 0.0576171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0168,
+ "step": 16990
+ },
+ {
+ "epoch": 54.284345047923324,
+ "grad_norm": 0.045654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0199,
+ "step": 16991
+ },
+ {
+ "epoch": 54.287539936102235,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0306,
+ "step": 16992
+ },
+ {
+ "epoch": 54.29073482428115,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0226,
+ "step": 16993
+ },
+ {
+ "epoch": 54.29392971246006,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0282,
+ "step": 16994
+ },
+ {
+ "epoch": 54.29712460063898,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0265,
+ "step": 16995
+ },
+ {
+ "epoch": 54.30031948881789,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0241,
+ "step": 16996
+ },
+ {
+ "epoch": 54.3035143769968,
+ "grad_norm": 0.0888671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0293,
+ "step": 16997
+ },
+ {
+ "epoch": 54.30670926517572,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0166,
+ "step": 16998
+ },
+ {
+ "epoch": 54.30990415335463,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0358,
+ "step": 16999
+ },
+ {
+ "epoch": 54.31309904153355,
+ "grad_norm": 0.1328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0327,
+ "step": 17000
+ },
+ {
+ "epoch": 54.31629392971246,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.032,
+ "step": 17001
+ },
+ {
+ "epoch": 54.319488817891376,
+ "grad_norm": 0.130859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0244,
+ "step": 17002
+ },
+ {
+ "epoch": 54.322683706070286,
+ "grad_norm": 0.0869140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0242,
+ "step": 17003
+ },
+ {
+ "epoch": 54.325878594249204,
+ "grad_norm": 0.1552734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0136,
+ "step": 17004
+ },
+ {
+ "epoch": 54.329073482428115,
+ "grad_norm": 0.119140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0146,
+ "step": 17005
+ },
+ {
+ "epoch": 54.33226837060703,
+ "grad_norm": 0.11279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0293,
+ "step": 17006
+ },
+ {
+ "epoch": 54.33546325878594,
+ "grad_norm": 0.26171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0312,
+ "step": 17007
+ },
+ {
+ "epoch": 54.33865814696485,
+ "grad_norm": 0.203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0221,
+ "step": 17008
+ },
+ {
+ "epoch": 54.34185303514377,
+ "grad_norm": 0.142578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0313,
+ "step": 17009
+ },
+ {
+ "epoch": 54.34504792332268,
+ "grad_norm": 0.31640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0328,
+ "step": 17010
+ },
+ {
+ "epoch": 54.3482428115016,
+ "grad_norm": 0.125,
+ "learning_rate": 0.0005,
+ "loss": 1.028,
+ "step": 17011
+ },
+ {
+ "epoch": 54.35143769968051,
+ "grad_norm": 0.12890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0152,
+ "step": 17012
+ },
+ {
+ "epoch": 54.35463258785943,
+ "grad_norm": 0.138671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0283,
+ "step": 17013
+ },
+ {
+ "epoch": 54.35782747603834,
+ "grad_norm": 0.0888671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0262,
+ "step": 17014
+ },
+ {
+ "epoch": 54.361022364217256,
+ "grad_norm": 0.2060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0289,
+ "step": 17015
+ },
+ {
+ "epoch": 54.364217252396166,
+ "grad_norm": 0.08056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0271,
+ "step": 17016
+ },
+ {
+ "epoch": 54.36741214057508,
+ "grad_norm": 0.1630859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0341,
+ "step": 17017
+ },
+ {
+ "epoch": 54.370607028753994,
+ "grad_norm": 0.126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0251,
+ "step": 17018
+ },
+ {
+ "epoch": 54.373801916932905,
+ "grad_norm": 0.12451171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0236,
+ "step": 17019
+ },
+ {
+ "epoch": 54.37699680511182,
+ "grad_norm": 0.1484375,
+ "learning_rate": 0.0005,
+ "loss": 1.04,
+ "step": 17020
+ },
+ {
+ "epoch": 54.38019169329073,
+ "grad_norm": 0.10205078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0231,
+ "step": 17021
+ },
+ {
+ "epoch": 54.38338658146965,
+ "grad_norm": 0.1787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0273,
+ "step": 17022
+ },
+ {
+ "epoch": 54.38658146964856,
+ "grad_norm": 0.09765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0211,
+ "step": 17023
+ },
+ {
+ "epoch": 54.38977635782748,
+ "grad_norm": 0.224609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0308,
+ "step": 17024
+ },
+ {
+ "epoch": 54.39297124600639,
+ "grad_norm": 0.08740234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0312,
+ "step": 17025
+ },
+ {
+ "epoch": 54.3961661341853,
+ "grad_norm": 0.1279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0291,
+ "step": 17026
+ },
+ {
+ "epoch": 54.39936102236422,
+ "grad_norm": 0.07861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0306,
+ "step": 17027
+ },
+ {
+ "epoch": 54.40255591054313,
+ "grad_norm": 0.1279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0263,
+ "step": 17028
+ },
+ {
+ "epoch": 54.405750798722046,
+ "grad_norm": 0.111328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0339,
+ "step": 17029
+ },
+ {
+ "epoch": 54.40894568690096,
+ "grad_norm": 0.11083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0217,
+ "step": 17030
+ },
+ {
+ "epoch": 54.412140575079874,
+ "grad_norm": 0.1875,
+ "learning_rate": 0.0005,
+ "loss": 1.025,
+ "step": 17031
+ },
+ {
+ "epoch": 54.415335463258785,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0329,
+ "step": 17032
+ },
+ {
+ "epoch": 54.4185303514377,
+ "grad_norm": 0.240234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0374,
+ "step": 17033
+ },
+ {
+ "epoch": 54.42172523961661,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0239,
+ "step": 17034
+ },
+ {
+ "epoch": 54.424920127795524,
+ "grad_norm": 0.1484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0238,
+ "step": 17035
+ },
+ {
+ "epoch": 54.42811501597444,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0322,
+ "step": 17036
+ },
+ {
+ "epoch": 54.43130990415335,
+ "grad_norm": 0.0908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0301,
+ "step": 17037
+ },
+ {
+ "epoch": 54.43450479233227,
+ "grad_norm": 0.05224609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0305,
+ "step": 17038
+ },
+ {
+ "epoch": 54.43769968051118,
+ "grad_norm": 0.08642578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0316,
+ "step": 17039
+ },
+ {
+ "epoch": 54.4408945686901,
+ "grad_norm": 0.0908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0336,
+ "step": 17040
+ },
+ {
+ "epoch": 54.44408945686901,
+ "grad_norm": 0.048095703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0236,
+ "step": 17041
+ },
+ {
+ "epoch": 54.447284345047926,
+ "grad_norm": 0.05126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0301,
+ "step": 17042
+ },
+ {
+ "epoch": 54.45047923322684,
+ "grad_norm": 0.05419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0161,
+ "step": 17043
+ },
+ {
+ "epoch": 54.453674121405754,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0276,
+ "step": 17044
+ },
+ {
+ "epoch": 54.456869009584665,
+ "grad_norm": 0.052490234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0214,
+ "step": 17045
+ },
+ {
+ "epoch": 54.460063897763575,
+ "grad_norm": 0.0576171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0351,
+ "step": 17046
+ },
+ {
+ "epoch": 54.46325878594249,
+ "grad_norm": 0.05615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0415,
+ "step": 17047
+ },
+ {
+ "epoch": 54.466453674121404,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0191,
+ "step": 17048
+ },
+ {
+ "epoch": 54.46964856230032,
+ "grad_norm": 0.05126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0303,
+ "step": 17049
+ },
+ {
+ "epoch": 54.47284345047923,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0248,
+ "step": 17050
+ },
+ {
+ "epoch": 54.47603833865815,
+ "grad_norm": 0.047607421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0262,
+ "step": 17051
+ },
+ {
+ "epoch": 54.47923322683706,
+ "grad_norm": 0.04541015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0291,
+ "step": 17052
+ },
+ {
+ "epoch": 54.48242811501598,
+ "grad_norm": 0.047119140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0357,
+ "step": 17053
+ },
+ {
+ "epoch": 54.48562300319489,
+ "grad_norm": 0.058837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0293,
+ "step": 17054
+ },
+ {
+ "epoch": 54.4888178913738,
+ "grad_norm": 0.0576171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0298,
+ "step": 17055
+ },
+ {
+ "epoch": 54.49201277955272,
+ "grad_norm": 0.05908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0279,
+ "step": 17056
+ },
+ {
+ "epoch": 54.49520766773163,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0276,
+ "step": 17057
+ },
+ {
+ "epoch": 54.498402555910545,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0244,
+ "step": 17058
+ },
+ {
+ "epoch": 54.501597444089455,
+ "grad_norm": 0.1015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0149,
+ "step": 17059
+ },
+ {
+ "epoch": 54.50479233226837,
+ "grad_norm": 0.05078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0227,
+ "step": 17060
+ },
+ {
+ "epoch": 54.50798722044728,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0252,
+ "step": 17061
+ },
+ {
+ "epoch": 54.5111821086262,
+ "grad_norm": 0.080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0234,
+ "step": 17062
+ },
+ {
+ "epoch": 54.51437699680511,
+ "grad_norm": 0.10302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0296,
+ "step": 17063
+ },
+ {
+ "epoch": 54.51757188498402,
+ "grad_norm": 0.0869140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0185,
+ "step": 17064
+ },
+ {
+ "epoch": 54.52076677316294,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0219,
+ "step": 17065
+ },
+ {
+ "epoch": 54.52396166134185,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0324,
+ "step": 17066
+ },
+ {
+ "epoch": 54.52715654952077,
+ "grad_norm": 0.083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0212,
+ "step": 17067
+ },
+ {
+ "epoch": 54.53035143769968,
+ "grad_norm": 0.060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.024,
+ "step": 17068
+ },
+ {
+ "epoch": 54.533546325878596,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0339,
+ "step": 17069
+ },
+ {
+ "epoch": 54.53674121405751,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0265,
+ "step": 17070
+ },
+ {
+ "epoch": 54.539936102236425,
+ "grad_norm": 0.05615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0234,
+ "step": 17071
+ },
+ {
+ "epoch": 54.543130990415335,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0384,
+ "step": 17072
+ },
+ {
+ "epoch": 54.546325878594246,
+ "grad_norm": 0.052001953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0191,
+ "step": 17073
+ },
+ {
+ "epoch": 54.54952076677316,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0325,
+ "step": 17074
+ },
+ {
+ "epoch": 54.552715654952074,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0263,
+ "step": 17075
+ },
+ {
+ "epoch": 54.55591054313099,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0283,
+ "step": 17076
+ },
+ {
+ "epoch": 54.5591054313099,
+ "grad_norm": 0.057373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0285,
+ "step": 17077
+ },
+ {
+ "epoch": 54.56230031948882,
+ "grad_norm": 0.05517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0355,
+ "step": 17078
+ },
+ {
+ "epoch": 54.56549520766773,
+ "grad_norm": 0.059326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0308,
+ "step": 17079
+ },
+ {
+ "epoch": 54.56869009584665,
+ "grad_norm": 0.056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0158,
+ "step": 17080
+ },
+ {
+ "epoch": 54.57188498402556,
+ "grad_norm": 0.05126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0265,
+ "step": 17081
+ },
+ {
+ "epoch": 54.575079872204476,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0238,
+ "step": 17082
+ },
+ {
+ "epoch": 54.57827476038339,
+ "grad_norm": 0.053955078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0411,
+ "step": 17083
+ },
+ {
+ "epoch": 54.5814696485623,
+ "grad_norm": 0.0927734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0316,
+ "step": 17084
+ },
+ {
+ "epoch": 54.584664536741215,
+ "grad_norm": 0.0478515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0327,
+ "step": 17085
+ },
+ {
+ "epoch": 54.587859424920126,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0255,
+ "step": 17086
+ },
+ {
+ "epoch": 54.59105431309904,
+ "grad_norm": 0.052734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0219,
+ "step": 17087
+ },
+ {
+ "epoch": 54.594249201277954,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0291,
+ "step": 17088
+ },
+ {
+ "epoch": 54.59744408945687,
+ "grad_norm": 0.058349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0312,
+ "step": 17089
+ },
+ {
+ "epoch": 54.60063897763578,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0403,
+ "step": 17090
+ },
+ {
+ "epoch": 54.6038338658147,
+ "grad_norm": 0.055419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0206,
+ "step": 17091
+ },
+ {
+ "epoch": 54.60702875399361,
+ "grad_norm": 0.057373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0248,
+ "step": 17092
+ },
+ {
+ "epoch": 54.61022364217252,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0275,
+ "step": 17093
+ },
+ {
+ "epoch": 54.61341853035144,
+ "grad_norm": 0.044921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0309,
+ "step": 17094
+ },
+ {
+ "epoch": 54.61661341853035,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0241,
+ "step": 17095
+ },
+ {
+ "epoch": 54.61980830670927,
+ "grad_norm": 0.0439453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0293,
+ "step": 17096
+ },
+ {
+ "epoch": 54.62300319488818,
+ "grad_norm": 0.1025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0243,
+ "step": 17097
+ },
+ {
+ "epoch": 54.626198083067095,
+ "grad_norm": 0.05615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.023,
+ "step": 17098
+ },
+ {
+ "epoch": 54.629392971246006,
+ "grad_norm": 0.056884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0133,
+ "step": 17099
+ },
+ {
+ "epoch": 54.63258785942492,
+ "grad_norm": 0.09423828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0251,
+ "step": 17100
+ },
+ {
+ "epoch": 54.635782747603834,
+ "grad_norm": 0.056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.03,
+ "step": 17101
+ },
+ {
+ "epoch": 54.638977635782744,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0262,
+ "step": 17102
+ },
+ {
+ "epoch": 54.64217252396166,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0263,
+ "step": 17103
+ },
+ {
+ "epoch": 54.64536741214057,
+ "grad_norm": 0.057861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0283,
+ "step": 17104
+ },
+ {
+ "epoch": 54.64856230031949,
+ "grad_norm": 0.047119140625,
+ "learning_rate": 0.0005,
+ "loss": 1.035,
+ "step": 17105
+ },
+ {
+ "epoch": 54.6517571884984,
+ "grad_norm": 0.04736328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0396,
+ "step": 17106
+ },
+ {
+ "epoch": 54.65495207667732,
+ "grad_norm": 0.051025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0357,
+ "step": 17107
+ },
+ {
+ "epoch": 54.65814696485623,
+ "grad_norm": 0.058349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0159,
+ "step": 17108
+ },
+ {
+ "epoch": 54.66134185303515,
+ "grad_norm": 0.059326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0362,
+ "step": 17109
+ },
+ {
+ "epoch": 54.66453674121406,
+ "grad_norm": 0.059326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0283,
+ "step": 17110
+ },
+ {
+ "epoch": 54.66773162939297,
+ "grad_norm": 0.05078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0323,
+ "step": 17111
+ },
+ {
+ "epoch": 54.670926517571885,
+ "grad_norm": 0.043212890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0249,
+ "step": 17112
+ },
+ {
+ "epoch": 54.674121405750796,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0308,
+ "step": 17113
+ },
+ {
+ "epoch": 54.677316293929714,
+ "grad_norm": 0.05615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0344,
+ "step": 17114
+ },
+ {
+ "epoch": 54.680511182108624,
+ "grad_norm": 0.057373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0253,
+ "step": 17115
+ },
+ {
+ "epoch": 54.68370607028754,
+ "grad_norm": 0.0517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0189,
+ "step": 17116
+ },
+ {
+ "epoch": 54.68690095846645,
+ "grad_norm": 0.0517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0344,
+ "step": 17117
+ },
+ {
+ "epoch": 54.69009584664537,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0223,
+ "step": 17118
+ },
+ {
+ "epoch": 54.69329073482428,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0224,
+ "step": 17119
+ },
+ {
+ "epoch": 54.6964856230032,
+ "grad_norm": 0.10498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0219,
+ "step": 17120
+ },
+ {
+ "epoch": 54.69968051118211,
+ "grad_norm": 0.055419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0225,
+ "step": 17121
+ },
+ {
+ "epoch": 54.70287539936102,
+ "grad_norm": 0.0517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0252,
+ "step": 17122
+ },
+ {
+ "epoch": 54.70607028753994,
+ "grad_norm": 0.0595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0308,
+ "step": 17123
+ },
+ {
+ "epoch": 54.70926517571885,
+ "grad_norm": 0.10107421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0273,
+ "step": 17124
+ },
+ {
+ "epoch": 54.712460063897765,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0268,
+ "step": 17125
+ },
+ {
+ "epoch": 54.715654952076676,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0201,
+ "step": 17126
+ },
+ {
+ "epoch": 54.718849840255594,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.019,
+ "step": 17127
+ },
+ {
+ "epoch": 54.722044728434504,
+ "grad_norm": 0.057861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0228,
+ "step": 17128
+ },
+ {
+ "epoch": 54.72523961661342,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0059,
+ "step": 17129
+ },
+ {
+ "epoch": 54.72843450479233,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0224,
+ "step": 17130
+ },
+ {
+ "epoch": 54.73162939297124,
+ "grad_norm": 0.0576171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0125,
+ "step": 17131
+ },
+ {
+ "epoch": 54.73482428115016,
+ "grad_norm": 0.10498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0201,
+ "step": 17132
+ },
+ {
+ "epoch": 54.73801916932907,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0341,
+ "step": 17133
+ },
+ {
+ "epoch": 54.74121405750799,
+ "grad_norm": 0.08056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0269,
+ "step": 17134
+ },
+ {
+ "epoch": 54.7444089456869,
+ "grad_norm": 0.056884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0195,
+ "step": 17135
+ },
+ {
+ "epoch": 54.74760383386582,
+ "grad_norm": 0.08740234375,
+ "learning_rate": 0.0005,
+ "loss": 1.032,
+ "step": 17136
+ },
+ {
+ "epoch": 54.75079872204473,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.023,
+ "step": 17137
+ },
+ {
+ "epoch": 54.753993610223645,
+ "grad_norm": 0.09423828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0245,
+ "step": 17138
+ },
+ {
+ "epoch": 54.757188498402556,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0207,
+ "step": 17139
+ },
+ {
+ "epoch": 54.760383386581466,
+ "grad_norm": 0.130859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0222,
+ "step": 17140
+ },
+ {
+ "epoch": 54.763578274760384,
+ "grad_norm": 0.08740234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0282,
+ "step": 17141
+ },
+ {
+ "epoch": 54.766773162939295,
+ "grad_norm": 0.15625,
+ "learning_rate": 0.0005,
+ "loss": 1.0278,
+ "step": 17142
+ },
+ {
+ "epoch": 54.76996805111821,
+ "grad_norm": 0.09423828125,
+ "learning_rate": 0.0005,
+ "loss": 1.019,
+ "step": 17143
+ },
+ {
+ "epoch": 54.77316293929712,
+ "grad_norm": 0.25,
+ "learning_rate": 0.0005,
+ "loss": 1.0391,
+ "step": 17144
+ },
+ {
+ "epoch": 54.77635782747604,
+ "grad_norm": 0.1162109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0154,
+ "step": 17145
+ },
+ {
+ "epoch": 54.77955271565495,
+ "grad_norm": 0.126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0108,
+ "step": 17146
+ },
+ {
+ "epoch": 54.78274760383387,
+ "grad_norm": 0.1005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0237,
+ "step": 17147
+ },
+ {
+ "epoch": 54.78594249201278,
+ "grad_norm": 0.09814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0336,
+ "step": 17148
+ },
+ {
+ "epoch": 54.78913738019169,
+ "grad_norm": 0.1123046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0303,
+ "step": 17149
+ },
+ {
+ "epoch": 54.79233226837061,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.025,
+ "step": 17150
+ },
+ {
+ "epoch": 54.79552715654952,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0188,
+ "step": 17151
+ },
+ {
+ "epoch": 54.798722044728436,
+ "grad_norm": 0.1103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0369,
+ "step": 17152
+ },
+ {
+ "epoch": 54.801916932907346,
+ "grad_norm": 0.09619140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0158,
+ "step": 17153
+ },
+ {
+ "epoch": 54.805111821086264,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0173,
+ "step": 17154
+ },
+ {
+ "epoch": 54.808306709265175,
+ "grad_norm": 0.107421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0289,
+ "step": 17155
+ },
+ {
+ "epoch": 54.81150159744409,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0265,
+ "step": 17156
+ },
+ {
+ "epoch": 54.814696485623,
+ "grad_norm": 0.1337890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0384,
+ "step": 17157
+ },
+ {
+ "epoch": 54.81789137380191,
+ "grad_norm": 0.0576171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0343,
+ "step": 17158
+ },
+ {
+ "epoch": 54.82108626198083,
+ "grad_norm": 0.0869140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0291,
+ "step": 17159
+ },
+ {
+ "epoch": 54.82428115015974,
+ "grad_norm": 0.09814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0302,
+ "step": 17160
+ },
+ {
+ "epoch": 54.82747603833866,
+ "grad_norm": 0.10693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0247,
+ "step": 17161
+ },
+ {
+ "epoch": 54.83067092651757,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0294,
+ "step": 17162
+ },
+ {
+ "epoch": 54.83386581469649,
+ "grad_norm": 0.1142578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0256,
+ "step": 17163
+ },
+ {
+ "epoch": 54.8370607028754,
+ "grad_norm": 0.057373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0261,
+ "step": 17164
+ },
+ {
+ "epoch": 54.840255591054316,
+ "grad_norm": 0.1015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0167,
+ "step": 17165
+ },
+ {
+ "epoch": 54.843450479233226,
+ "grad_norm": 0.05322265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0412,
+ "step": 17166
+ },
+ {
+ "epoch": 54.846645367412144,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0209,
+ "step": 17167
+ },
+ {
+ "epoch": 54.849840255591054,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0322,
+ "step": 17168
+ },
+ {
+ "epoch": 54.853035143769965,
+ "grad_norm": 0.08984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0289,
+ "step": 17169
+ },
+ {
+ "epoch": 54.85623003194888,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0261,
+ "step": 17170
+ },
+ {
+ "epoch": 54.85942492012779,
+ "grad_norm": 0.07861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0162,
+ "step": 17171
+ },
+ {
+ "epoch": 54.86261980830671,
+ "grad_norm": 0.1416015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0181,
+ "step": 17172
+ },
+ {
+ "epoch": 54.86581469648562,
+ "grad_norm": 0.11572265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0315,
+ "step": 17173
+ },
+ {
+ "epoch": 54.86900958466454,
+ "grad_norm": 0.15625,
+ "learning_rate": 0.0005,
+ "loss": 1.0397,
+ "step": 17174
+ },
+ {
+ "epoch": 54.87220447284345,
+ "grad_norm": 0.16796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0364,
+ "step": 17175
+ },
+ {
+ "epoch": 54.87539936102237,
+ "grad_norm": 0.10400390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0198,
+ "step": 17176
+ },
+ {
+ "epoch": 54.87859424920128,
+ "grad_norm": 0.12451171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0185,
+ "step": 17177
+ },
+ {
+ "epoch": 54.88178913738019,
+ "grad_norm": 0.1640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0265,
+ "step": 17178
+ },
+ {
+ "epoch": 54.884984025559106,
+ "grad_norm": 0.16015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0247,
+ "step": 17179
+ },
+ {
+ "epoch": 54.88817891373802,
+ "grad_norm": 0.189453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0303,
+ "step": 17180
+ },
+ {
+ "epoch": 54.891373801916934,
+ "grad_norm": 0.2578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0323,
+ "step": 17181
+ },
+ {
+ "epoch": 54.894568690095845,
+ "grad_norm": 0.1083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0208,
+ "step": 17182
+ },
+ {
+ "epoch": 54.89776357827476,
+ "grad_norm": 0.20703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0237,
+ "step": 17183
+ },
+ {
+ "epoch": 54.90095846645367,
+ "grad_norm": 0.150390625,
+ "learning_rate": 0.0005,
+ "loss": 1.017,
+ "step": 17184
+ },
+ {
+ "epoch": 54.90415335463259,
+ "grad_norm": 0.11962890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0259,
+ "step": 17185
+ },
+ {
+ "epoch": 54.9073482428115,
+ "grad_norm": 0.1513671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0303,
+ "step": 17186
+ },
+ {
+ "epoch": 54.91054313099041,
+ "grad_norm": 0.09716796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0265,
+ "step": 17187
+ },
+ {
+ "epoch": 54.91373801916933,
+ "grad_norm": 0.1806640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0188,
+ "step": 17188
+ },
+ {
+ "epoch": 54.91693290734824,
+ "grad_norm": 0.08642578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0289,
+ "step": 17189
+ },
+ {
+ "epoch": 54.92012779552716,
+ "grad_norm": 0.1953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0231,
+ "step": 17190
+ },
+ {
+ "epoch": 54.92332268370607,
+ "grad_norm": 0.1318359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0324,
+ "step": 17191
+ },
+ {
+ "epoch": 54.926517571884986,
+ "grad_norm": 0.2177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0367,
+ "step": 17192
+ },
+ {
+ "epoch": 54.9297124600639,
+ "grad_norm": 0.361328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0331,
+ "step": 17193
+ },
+ {
+ "epoch": 54.932907348242814,
+ "grad_norm": 0.060302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0302,
+ "step": 17194
+ },
+ {
+ "epoch": 54.936102236421725,
+ "grad_norm": 0.27734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0296,
+ "step": 17195
+ },
+ {
+ "epoch": 54.93929712460064,
+ "grad_norm": 0.1748046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0144,
+ "step": 17196
+ },
+ {
+ "epoch": 54.94249201277955,
+ "grad_norm": 0.1279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0269,
+ "step": 17197
+ },
+ {
+ "epoch": 54.945686900958464,
+ "grad_norm": 0.2490234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0178,
+ "step": 17198
+ },
+ {
+ "epoch": 54.94888178913738,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0349,
+ "step": 17199
+ },
+ {
+ "epoch": 54.95207667731629,
+ "grad_norm": 0.2412109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0196,
+ "step": 17200
+ },
+ {
+ "epoch": 54.95527156549521,
+ "grad_norm": 0.0888671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0187,
+ "step": 17201
+ },
+ {
+ "epoch": 54.95846645367412,
+ "grad_norm": 0.21484375,
+ "learning_rate": 0.0005,
+ "loss": 1.029,
+ "step": 17202
+ },
+ {
+ "epoch": 54.96166134185304,
+ "grad_norm": 0.1435546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0294,
+ "step": 17203
+ },
+ {
+ "epoch": 54.96485623003195,
+ "grad_norm": 0.119140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0314,
+ "step": 17204
+ },
+ {
+ "epoch": 54.968051118210866,
+ "grad_norm": 0.1357421875,
+ "learning_rate": 0.0005,
+ "loss": 1.032,
+ "step": 17205
+ },
+ {
+ "epoch": 54.97124600638978,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.029,
+ "step": 17206
+ },
+ {
+ "epoch": 54.97444089456869,
+ "grad_norm": 0.130859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0302,
+ "step": 17207
+ },
+ {
+ "epoch": 54.977635782747605,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0267,
+ "step": 17208
+ },
+ {
+ "epoch": 54.980830670926515,
+ "grad_norm": 0.12158203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0347,
+ "step": 17209
+ },
+ {
+ "epoch": 54.98402555910543,
+ "grad_norm": 0.060791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0339,
+ "step": 17210
+ },
+ {
+ "epoch": 54.98722044728434,
+ "grad_norm": 0.10205078125,
+ "learning_rate": 0.0005,
+ "loss": 1.017,
+ "step": 17211
+ },
+ {
+ "epoch": 54.99041533546326,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0237,
+ "step": 17212
+ },
+ {
+ "epoch": 54.99361022364217,
+ "grad_norm": 0.1064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0191,
+ "step": 17213
+ },
+ {
+ "epoch": 54.99680511182109,
+ "grad_norm": 0.09716796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0353,
+ "step": 17214
+ },
+ {
+ "epoch": 55.0,
+ "grad_norm": 0.1513671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0333,
+ "step": 17215
+ },
+ {
+ "epoch": 55.00319488817891,
+ "grad_norm": 0.08544921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0201,
+ "step": 17216
+ },
+ {
+ "epoch": 55.00638977635783,
+ "grad_norm": 0.1845703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0351,
+ "step": 17217
+ },
+ {
+ "epoch": 55.00958466453674,
+ "grad_norm": 0.05224609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0343,
+ "step": 17218
+ },
+ {
+ "epoch": 55.01277955271566,
+ "grad_norm": 0.1552734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0283,
+ "step": 17219
+ },
+ {
+ "epoch": 55.01597444089457,
+ "grad_norm": 0.052734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0177,
+ "step": 17220
+ },
+ {
+ "epoch": 55.019169329073485,
+ "grad_norm": 0.10986328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0343,
+ "step": 17221
+ },
+ {
+ "epoch": 55.022364217252395,
+ "grad_norm": 0.056884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.032,
+ "step": 17222
+ },
+ {
+ "epoch": 55.02555910543131,
+ "grad_norm": 0.1005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0238,
+ "step": 17223
+ },
+ {
+ "epoch": 55.02875399361022,
+ "grad_norm": 0.05810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0223,
+ "step": 17224
+ },
+ {
+ "epoch": 55.031948881789134,
+ "grad_norm": 0.0927734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0362,
+ "step": 17225
+ },
+ {
+ "epoch": 55.03514376996805,
+ "grad_norm": 0.0537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0241,
+ "step": 17226
+ },
+ {
+ "epoch": 55.03833865814696,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0316,
+ "step": 17227
+ },
+ {
+ "epoch": 55.04153354632588,
+ "grad_norm": 0.0517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0208,
+ "step": 17228
+ },
+ {
+ "epoch": 55.04472843450479,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0228,
+ "step": 17229
+ },
+ {
+ "epoch": 55.04792332268371,
+ "grad_norm": 0.05517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0213,
+ "step": 17230
+ },
+ {
+ "epoch": 55.05111821086262,
+ "grad_norm": 0.0888671875,
+ "learning_rate": 0.0005,
+ "loss": 1.038,
+ "step": 17231
+ },
+ {
+ "epoch": 55.054313099041536,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0132,
+ "step": 17232
+ },
+ {
+ "epoch": 55.05750798722045,
+ "grad_norm": 0.05419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0215,
+ "step": 17233
+ },
+ {
+ "epoch": 55.06070287539936,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0308,
+ "step": 17234
+ },
+ {
+ "epoch": 55.063897763578275,
+ "grad_norm": 0.0576171875,
+ "learning_rate": 0.0005,
+ "loss": 1.018,
+ "step": 17235
+ },
+ {
+ "epoch": 55.067092651757186,
+ "grad_norm": 0.057373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0248,
+ "step": 17236
+ },
+ {
+ "epoch": 55.0702875399361,
+ "grad_norm": 0.048095703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0369,
+ "step": 17237
+ },
+ {
+ "epoch": 55.073482428115014,
+ "grad_norm": 0.060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0225,
+ "step": 17238
+ },
+ {
+ "epoch": 55.07667731629393,
+ "grad_norm": 0.06103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0193,
+ "step": 17239
+ },
+ {
+ "epoch": 55.07987220447284,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0277,
+ "step": 17240
+ },
+ {
+ "epoch": 55.08306709265176,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0259,
+ "step": 17241
+ },
+ {
+ "epoch": 55.08626198083067,
+ "grad_norm": 0.11865234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0352,
+ "step": 17242
+ },
+ {
+ "epoch": 55.08945686900959,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0229,
+ "step": 17243
+ },
+ {
+ "epoch": 55.0926517571885,
+ "grad_norm": 0.10791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0239,
+ "step": 17244
+ },
+ {
+ "epoch": 55.09584664536741,
+ "grad_norm": 0.0546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0179,
+ "step": 17245
+ },
+ {
+ "epoch": 55.09904153354633,
+ "grad_norm": 0.08056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0219,
+ "step": 17246
+ },
+ {
+ "epoch": 55.10223642172524,
+ "grad_norm": 0.060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0267,
+ "step": 17247
+ },
+ {
+ "epoch": 55.105431309904155,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0252,
+ "step": 17248
+ },
+ {
+ "epoch": 55.108626198083066,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0114,
+ "step": 17249
+ },
+ {
+ "epoch": 55.11182108626198,
+ "grad_norm": 0.053466796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0249,
+ "step": 17250
+ },
+ {
+ "epoch": 55.115015974440894,
+ "grad_norm": 0.051513671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0135,
+ "step": 17251
+ },
+ {
+ "epoch": 55.11821086261981,
+ "grad_norm": 0.054443359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0248,
+ "step": 17252
+ },
+ {
+ "epoch": 55.12140575079872,
+ "grad_norm": 0.053955078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0334,
+ "step": 17253
+ },
+ {
+ "epoch": 55.12460063897763,
+ "grad_norm": 0.048583984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0132,
+ "step": 17254
+ },
+ {
+ "epoch": 55.12779552715655,
+ "grad_norm": 0.1259765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0222,
+ "step": 17255
+ },
+ {
+ "epoch": 55.13099041533546,
+ "grad_norm": 0.057373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0265,
+ "step": 17256
+ },
+ {
+ "epoch": 55.13418530351438,
+ "grad_norm": 0.1318359375,
+ "learning_rate": 0.0005,
+ "loss": 1.024,
+ "step": 17257
+ },
+ {
+ "epoch": 55.13738019169329,
+ "grad_norm": 0.10400390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0129,
+ "step": 17258
+ },
+ {
+ "epoch": 55.14057507987221,
+ "grad_norm": 0.1689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0358,
+ "step": 17259
+ },
+ {
+ "epoch": 55.14376996805112,
+ "grad_norm": 0.12255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0263,
+ "step": 17260
+ },
+ {
+ "epoch": 55.146964856230035,
+ "grad_norm": 0.19140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0221,
+ "step": 17261
+ },
+ {
+ "epoch": 55.150159744408946,
+ "grad_norm": 0.1025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0289,
+ "step": 17262
+ },
+ {
+ "epoch": 55.153354632587856,
+ "grad_norm": 0.173828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0241,
+ "step": 17263
+ },
+ {
+ "epoch": 55.156549520766774,
+ "grad_norm": 0.095703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0278,
+ "step": 17264
+ },
+ {
+ "epoch": 55.159744408945684,
+ "grad_norm": 0.09521484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0269,
+ "step": 17265
+ },
+ {
+ "epoch": 55.1629392971246,
+ "grad_norm": 0.08251953125,
+ "learning_rate": 0.0005,
+ "loss": 1.024,
+ "step": 17266
+ },
+ {
+ "epoch": 55.16613418530351,
+ "grad_norm": 0.0810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0263,
+ "step": 17267
+ },
+ {
+ "epoch": 55.16932907348243,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0201,
+ "step": 17268
+ },
+ {
+ "epoch": 55.17252396166134,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0199,
+ "step": 17269
+ },
+ {
+ "epoch": 55.17571884984026,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0114,
+ "step": 17270
+ },
+ {
+ "epoch": 55.17891373801917,
+ "grad_norm": 0.0576171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0335,
+ "step": 17271
+ },
+ {
+ "epoch": 55.18210862619808,
+ "grad_norm": 0.058349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0295,
+ "step": 17272
+ },
+ {
+ "epoch": 55.185303514377,
+ "grad_norm": 0.0986328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0264,
+ "step": 17273
+ },
+ {
+ "epoch": 55.18849840255591,
+ "grad_norm": 0.046630859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0182,
+ "step": 17274
+ },
+ {
+ "epoch": 55.191693290734825,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0227,
+ "step": 17275
+ },
+ {
+ "epoch": 55.194888178913736,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0162,
+ "step": 17276
+ },
+ {
+ "epoch": 55.198083067092654,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0316,
+ "step": 17277
+ },
+ {
+ "epoch": 55.201277955271564,
+ "grad_norm": 0.05859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0338,
+ "step": 17278
+ },
+ {
+ "epoch": 55.20447284345048,
+ "grad_norm": 0.049560546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0316,
+ "step": 17279
+ },
+ {
+ "epoch": 55.20766773162939,
+ "grad_norm": 0.0517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0343,
+ "step": 17280
+ },
+ {
+ "epoch": 55.21086261980831,
+ "grad_norm": 0.055419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.027,
+ "step": 17281
+ },
+ {
+ "epoch": 55.21405750798722,
+ "grad_norm": 0.056884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0331,
+ "step": 17282
+ },
+ {
+ "epoch": 55.21725239616613,
+ "grad_norm": 0.052490234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0301,
+ "step": 17283
+ },
+ {
+ "epoch": 55.22044728434505,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0242,
+ "step": 17284
+ },
+ {
+ "epoch": 55.22364217252396,
+ "grad_norm": 0.09521484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0175,
+ "step": 17285
+ },
+ {
+ "epoch": 55.22683706070288,
+ "grad_norm": 0.05419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.026,
+ "step": 17286
+ },
+ {
+ "epoch": 55.23003194888179,
+ "grad_norm": 0.119140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0269,
+ "step": 17287
+ },
+ {
+ "epoch": 55.233226837060705,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0245,
+ "step": 17288
+ },
+ {
+ "epoch": 55.236421725239616,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0255,
+ "step": 17289
+ },
+ {
+ "epoch": 55.239616613418534,
+ "grad_norm": 0.09423828125,
+ "learning_rate": 0.0005,
+ "loss": 1.026,
+ "step": 17290
+ },
+ {
+ "epoch": 55.242811501597444,
+ "grad_norm": 0.060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0233,
+ "step": 17291
+ },
+ {
+ "epoch": 55.246006389776355,
+ "grad_norm": 0.08154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0187,
+ "step": 17292
+ },
+ {
+ "epoch": 55.24920127795527,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0302,
+ "step": 17293
+ },
+ {
+ "epoch": 55.25239616613418,
+ "grad_norm": 0.058349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0229,
+ "step": 17294
+ },
+ {
+ "epoch": 55.2555910543131,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0328,
+ "step": 17295
+ },
+ {
+ "epoch": 55.25878594249201,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0375,
+ "step": 17296
+ },
+ {
+ "epoch": 55.26198083067093,
+ "grad_norm": 0.11669921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0325,
+ "step": 17297
+ },
+ {
+ "epoch": 55.26517571884984,
+ "grad_norm": 0.08935546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0316,
+ "step": 17298
+ },
+ {
+ "epoch": 55.26837060702876,
+ "grad_norm": 0.1767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.034,
+ "step": 17299
+ },
+ {
+ "epoch": 55.27156549520767,
+ "grad_norm": 0.056396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0226,
+ "step": 17300
+ },
+ {
+ "epoch": 55.27476038338658,
+ "grad_norm": 0.12451171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0371,
+ "step": 17301
+ },
+ {
+ "epoch": 55.277955271565496,
+ "grad_norm": 0.09033203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0201,
+ "step": 17302
+ },
+ {
+ "epoch": 55.281150159744406,
+ "grad_norm": 0.1455078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0275,
+ "step": 17303
+ },
+ {
+ "epoch": 55.284345047923324,
+ "grad_norm": 0.08984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0171,
+ "step": 17304
+ },
+ {
+ "epoch": 55.287539936102235,
+ "grad_norm": 0.1484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0308,
+ "step": 17305
+ },
+ {
+ "epoch": 55.29073482428115,
+ "grad_norm": 0.1689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0296,
+ "step": 17306
+ },
+ {
+ "epoch": 55.29392971246006,
+ "grad_norm": 0.08203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0173,
+ "step": 17307
+ },
+ {
+ "epoch": 55.29712460063898,
+ "grad_norm": 0.1884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0152,
+ "step": 17308
+ },
+ {
+ "epoch": 55.30031948881789,
+ "grad_norm": 0.080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0352,
+ "step": 17309
+ },
+ {
+ "epoch": 55.3035143769968,
+ "grad_norm": 0.1494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0333,
+ "step": 17310
+ },
+ {
+ "epoch": 55.30670926517572,
+ "grad_norm": 0.080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0359,
+ "step": 17311
+ },
+ {
+ "epoch": 55.30990415335463,
+ "grad_norm": 0.1650390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0299,
+ "step": 17312
+ },
+ {
+ "epoch": 55.31309904153355,
+ "grad_norm": 0.1298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0328,
+ "step": 17313
+ },
+ {
+ "epoch": 55.31629392971246,
+ "grad_norm": 0.18359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0264,
+ "step": 17314
+ },
+ {
+ "epoch": 55.319488817891376,
+ "grad_norm": 0.1025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0179,
+ "step": 17315
+ },
+ {
+ "epoch": 55.322683706070286,
+ "grad_norm": 0.1748046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0327,
+ "step": 17316
+ },
+ {
+ "epoch": 55.325878594249204,
+ "grad_norm": 0.1875,
+ "learning_rate": 0.0005,
+ "loss": 1.0333,
+ "step": 17317
+ },
+ {
+ "epoch": 55.329073482428115,
+ "grad_norm": 0.10205078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0371,
+ "step": 17318
+ },
+ {
+ "epoch": 55.33226837060703,
+ "grad_norm": 0.26171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0308,
+ "step": 17319
+ },
+ {
+ "epoch": 55.33546325878594,
+ "grad_norm": 0.2294921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0287,
+ "step": 17320
+ },
+ {
+ "epoch": 55.33865814696485,
+ "grad_norm": 0.1806640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0253,
+ "step": 17321
+ },
+ {
+ "epoch": 55.34185303514377,
+ "grad_norm": 0.458984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0298,
+ "step": 17322
+ },
+ {
+ "epoch": 55.34504792332268,
+ "grad_norm": 0.3125,
+ "learning_rate": 0.0005,
+ "loss": 1.0078,
+ "step": 17323
+ },
+ {
+ "epoch": 55.3482428115016,
+ "grad_norm": 0.212890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0136,
+ "step": 17324
+ },
+ {
+ "epoch": 55.35143769968051,
+ "grad_norm": 0.64453125,
+ "learning_rate": 0.0005,
+ "loss": 1.036,
+ "step": 17325
+ },
+ {
+ "epoch": 55.35463258785943,
+ "grad_norm": 1.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0352,
+ "step": 17326
+ },
+ {
+ "epoch": 55.35782747603834,
+ "grad_norm": 1.40625,
+ "learning_rate": 0.0005,
+ "loss": 1.0476,
+ "step": 17327
+ },
+ {
+ "epoch": 55.361022364217256,
+ "grad_norm": 0.275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.034,
+ "step": 17328
+ },
+ {
+ "epoch": 55.364217252396166,
+ "grad_norm": 1.390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0351,
+ "step": 17329
+ },
+ {
+ "epoch": 55.36741214057508,
+ "grad_norm": 1.2578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0331,
+ "step": 17330
+ },
+ {
+ "epoch": 55.370607028753994,
+ "grad_norm": 0.33203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0397,
+ "step": 17331
+ },
+ {
+ "epoch": 55.373801916932905,
+ "grad_norm": 1.1015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0365,
+ "step": 17332
+ },
+ {
+ "epoch": 55.37699680511182,
+ "grad_norm": 0.875,
+ "learning_rate": 0.0005,
+ "loss": 1.0316,
+ "step": 17333
+ },
+ {
+ "epoch": 55.38019169329073,
+ "grad_norm": 0.2158203125,
+ "learning_rate": 0.0005,
+ "loss": 1.026,
+ "step": 17334
+ },
+ {
+ "epoch": 55.38338658146965,
+ "grad_norm": 1.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0332,
+ "step": 17335
+ },
+ {
+ "epoch": 55.38658146964856,
+ "grad_norm": 0.80078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0198,
+ "step": 17336
+ },
+ {
+ "epoch": 55.38977635782748,
+ "grad_norm": 0.4375,
+ "learning_rate": 0.0005,
+ "loss": 1.0454,
+ "step": 17337
+ },
+ {
+ "epoch": 55.39297124600639,
+ "grad_norm": 1.6796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0516,
+ "step": 17338
+ },
+ {
+ "epoch": 55.3961661341853,
+ "grad_norm": 0.45703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0347,
+ "step": 17339
+ },
+ {
+ "epoch": 55.39936102236422,
+ "grad_norm": 1.28125,
+ "learning_rate": 0.0005,
+ "loss": 1.0421,
+ "step": 17340
+ },
+ {
+ "epoch": 55.40255591054313,
+ "grad_norm": 0.69921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0426,
+ "step": 17341
+ },
+ {
+ "epoch": 55.405750798722046,
+ "grad_norm": 1.5625,
+ "learning_rate": 0.0005,
+ "loss": 1.0609,
+ "step": 17342
+ },
+ {
+ "epoch": 55.40894568690096,
+ "grad_norm": 0.828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0376,
+ "step": 17343
+ },
+ {
+ "epoch": 55.412140575079874,
+ "grad_norm": 1.65625,
+ "learning_rate": 0.0005,
+ "loss": 1.0674,
+ "step": 17344
+ },
+ {
+ "epoch": 55.415335463258785,
+ "grad_norm": 1.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0475,
+ "step": 17345
+ },
+ {
+ "epoch": 55.4185303514377,
+ "grad_norm": 1.3125,
+ "learning_rate": 0.0005,
+ "loss": 1.0533,
+ "step": 17346
+ },
+ {
+ "epoch": 55.42172523961661,
+ "grad_norm": 1.140625,
+ "learning_rate": 0.0005,
+ "loss": 1.048,
+ "step": 17347
+ },
+ {
+ "epoch": 55.424920127795524,
+ "grad_norm": 1.2109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0547,
+ "step": 17348
+ },
+ {
+ "epoch": 55.42811501597444,
+ "grad_norm": 0.91015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0476,
+ "step": 17349
+ },
+ {
+ "epoch": 55.43130990415335,
+ "grad_norm": 1.421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0629,
+ "step": 17350
+ },
+ {
+ "epoch": 55.43450479233227,
+ "grad_norm": 1.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0636,
+ "step": 17351
+ },
+ {
+ "epoch": 55.43769968051118,
+ "grad_norm": 1.2734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0707,
+ "step": 17352
+ },
+ {
+ "epoch": 55.4408945686901,
+ "grad_norm": 1.1328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0722,
+ "step": 17353
+ },
+ {
+ "epoch": 55.44408945686901,
+ "grad_norm": 0.66796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0507,
+ "step": 17354
+ },
+ {
+ "epoch": 55.447284345047926,
+ "grad_norm": 0.890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0511,
+ "step": 17355
+ },
+ {
+ "epoch": 55.45047923322684,
+ "grad_norm": 0.4375,
+ "learning_rate": 0.0005,
+ "loss": 1.0394,
+ "step": 17356
+ },
+ {
+ "epoch": 55.453674121405754,
+ "grad_norm": 0.65234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0543,
+ "step": 17357
+ },
+ {
+ "epoch": 55.456869009584665,
+ "grad_norm": 0.26171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0392,
+ "step": 17358
+ },
+ {
+ "epoch": 55.460063897763575,
+ "grad_norm": 0.52734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0517,
+ "step": 17359
+ },
+ {
+ "epoch": 55.46325878594249,
+ "grad_norm": 0.228515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0325,
+ "step": 17360
+ },
+ {
+ "epoch": 55.466453674121404,
+ "grad_norm": 0.357421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0336,
+ "step": 17361
+ },
+ {
+ "epoch": 55.46964856230032,
+ "grad_norm": 0.2373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0237,
+ "step": 17362
+ },
+ {
+ "epoch": 55.47284345047923,
+ "grad_norm": 0.2890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0405,
+ "step": 17363
+ },
+ {
+ "epoch": 55.47603833865815,
+ "grad_norm": 0.236328125,
+ "learning_rate": 0.0005,
+ "loss": 1.026,
+ "step": 17364
+ },
+ {
+ "epoch": 55.47923322683706,
+ "grad_norm": 0.162109375,
+ "learning_rate": 0.0005,
+ "loss": 1.028,
+ "step": 17365
+ },
+ {
+ "epoch": 55.48242811501598,
+ "grad_norm": 0.255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0265,
+ "step": 17366
+ },
+ {
+ "epoch": 55.48562300319489,
+ "grad_norm": 0.1142578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0315,
+ "step": 17367
+ },
+ {
+ "epoch": 55.4888178913738,
+ "grad_norm": 0.1796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0238,
+ "step": 17368
+ },
+ {
+ "epoch": 55.49201277955272,
+ "grad_norm": 0.18359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0362,
+ "step": 17369
+ },
+ {
+ "epoch": 55.49520766773163,
+ "grad_norm": 0.095703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0304,
+ "step": 17370
+ },
+ {
+ "epoch": 55.498402555910545,
+ "grad_norm": 0.166015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0244,
+ "step": 17371
+ },
+ {
+ "epoch": 55.501597444089455,
+ "grad_norm": 0.1650390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0227,
+ "step": 17372
+ },
+ {
+ "epoch": 55.50479233226837,
+ "grad_norm": 0.087890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0322,
+ "step": 17373
+ },
+ {
+ "epoch": 55.50798722044728,
+ "grad_norm": 0.173828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0293,
+ "step": 17374
+ },
+ {
+ "epoch": 55.5111821086262,
+ "grad_norm": 0.09765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0323,
+ "step": 17375
+ },
+ {
+ "epoch": 55.51437699680511,
+ "grad_norm": 0.109375,
+ "learning_rate": 0.0005,
+ "loss": 1.024,
+ "step": 17376
+ },
+ {
+ "epoch": 55.51757188498402,
+ "grad_norm": 0.11328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0373,
+ "step": 17377
+ },
+ {
+ "epoch": 55.52076677316294,
+ "grad_norm": 0.0908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0265,
+ "step": 17378
+ },
+ {
+ "epoch": 55.52396166134185,
+ "grad_norm": 0.08056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0208,
+ "step": 17379
+ },
+ {
+ "epoch": 55.52715654952077,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0349,
+ "step": 17380
+ },
+ {
+ "epoch": 55.53035143769968,
+ "grad_norm": 0.08642578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0232,
+ "step": 17381
+ },
+ {
+ "epoch": 55.533546325878596,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0318,
+ "step": 17382
+ },
+ {
+ "epoch": 55.53674121405751,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0304,
+ "step": 17383
+ },
+ {
+ "epoch": 55.539936102236425,
+ "grad_norm": 0.08056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0366,
+ "step": 17384
+ },
+ {
+ "epoch": 55.543130990415335,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0279,
+ "step": 17385
+ },
+ {
+ "epoch": 55.546325878594246,
+ "grad_norm": 0.05615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.029,
+ "step": 17386
+ },
+ {
+ "epoch": 55.54952076677316,
+ "grad_norm": 0.09375,
+ "learning_rate": 0.0005,
+ "loss": 1.0198,
+ "step": 17387
+ },
+ {
+ "epoch": 55.552715654952074,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.02,
+ "step": 17388
+ },
+ {
+ "epoch": 55.55591054313099,
+ "grad_norm": 0.08056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0206,
+ "step": 17389
+ },
+ {
+ "epoch": 55.5591054313099,
+ "grad_norm": 0.0830078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0314,
+ "step": 17390
+ },
+ {
+ "epoch": 55.56230031948882,
+ "grad_norm": 0.05712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.036,
+ "step": 17391
+ },
+ {
+ "epoch": 55.56549520766773,
+ "grad_norm": 0.08447265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0359,
+ "step": 17392
+ },
+ {
+ "epoch": 55.56869009584665,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0224,
+ "step": 17393
+ },
+ {
+ "epoch": 55.57188498402556,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0191,
+ "step": 17394
+ },
+ {
+ "epoch": 55.575079872204476,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0327,
+ "step": 17395
+ },
+ {
+ "epoch": 55.57827476038339,
+ "grad_norm": 0.05615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0229,
+ "step": 17396
+ },
+ {
+ "epoch": 55.5814696485623,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0161,
+ "step": 17397
+ },
+ {
+ "epoch": 55.584664536741215,
+ "grad_norm": 0.045654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0212,
+ "step": 17398
+ },
+ {
+ "epoch": 55.587859424920126,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0256,
+ "step": 17399
+ },
+ {
+ "epoch": 55.59105431309904,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0411,
+ "step": 17400
+ },
+ {
+ "epoch": 55.594249201277954,
+ "grad_norm": 0.046630859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0289,
+ "step": 17401
+ },
+ {
+ "epoch": 55.59744408945687,
+ "grad_norm": 0.054931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0421,
+ "step": 17402
+ },
+ {
+ "epoch": 55.60063897763578,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0235,
+ "step": 17403
+ },
+ {
+ "epoch": 55.6038338658147,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0176,
+ "step": 17404
+ },
+ {
+ "epoch": 55.60702875399361,
+ "grad_norm": 0.04345703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0241,
+ "step": 17405
+ },
+ {
+ "epoch": 55.61022364217252,
+ "grad_norm": 0.0498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0448,
+ "step": 17406
+ },
+ {
+ "epoch": 55.61341853035144,
+ "grad_norm": 0.059326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0267,
+ "step": 17407
+ },
+ {
+ "epoch": 55.61661341853035,
+ "grad_norm": 0.04931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0203,
+ "step": 17408
+ },
+ {
+ "epoch": 55.61980830670927,
+ "grad_norm": 0.10302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.039,
+ "step": 17409
+ },
+ {
+ "epoch": 55.62300319488818,
+ "grad_norm": 0.08349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0317,
+ "step": 17410
+ },
+ {
+ "epoch": 55.626198083067095,
+ "grad_norm": 0.08203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0381,
+ "step": 17411
+ },
+ {
+ "epoch": 55.629392971246006,
+ "grad_norm": 0.10400390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0112,
+ "step": 17412
+ },
+ {
+ "epoch": 55.63258785942492,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0292,
+ "step": 17413
+ },
+ {
+ "epoch": 55.635782747603834,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.028,
+ "step": 17414
+ },
+ {
+ "epoch": 55.638977635782744,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0128,
+ "step": 17415
+ },
+ {
+ "epoch": 55.64217252396166,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0313,
+ "step": 17416
+ },
+ {
+ "epoch": 55.64536741214057,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0227,
+ "step": 17417
+ },
+ {
+ "epoch": 55.64856230031949,
+ "grad_norm": 0.08251953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0343,
+ "step": 17418
+ },
+ {
+ "epoch": 55.6517571884984,
+ "grad_norm": 0.11279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0324,
+ "step": 17419
+ },
+ {
+ "epoch": 55.65495207667732,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0109,
+ "step": 17420
+ },
+ {
+ "epoch": 55.65814696485623,
+ "grad_norm": 0.11865234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0111,
+ "step": 17421
+ },
+ {
+ "epoch": 55.66134185303515,
+ "grad_norm": 0.07861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0324,
+ "step": 17422
+ },
+ {
+ "epoch": 55.66453674121406,
+ "grad_norm": 0.142578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0244,
+ "step": 17423
+ },
+ {
+ "epoch": 55.66773162939297,
+ "grad_norm": 0.08349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0151,
+ "step": 17424
+ },
+ {
+ "epoch": 55.670926517571885,
+ "grad_norm": 0.130859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0262,
+ "step": 17425
+ },
+ {
+ "epoch": 55.674121405750796,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0375,
+ "step": 17426
+ },
+ {
+ "epoch": 55.677316293929714,
+ "grad_norm": 0.1123046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0381,
+ "step": 17427
+ },
+ {
+ "epoch": 55.680511182108624,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0217,
+ "step": 17428
+ },
+ {
+ "epoch": 55.68370607028754,
+ "grad_norm": 0.10595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.018,
+ "step": 17429
+ },
+ {
+ "epoch": 55.68690095846645,
+ "grad_norm": 0.08837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0316,
+ "step": 17430
+ },
+ {
+ "epoch": 55.69009584664537,
+ "grad_norm": 0.1494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0345,
+ "step": 17431
+ },
+ {
+ "epoch": 55.69329073482428,
+ "grad_norm": 0.09814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0183,
+ "step": 17432
+ },
+ {
+ "epoch": 55.6964856230032,
+ "grad_norm": 0.1923828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0339,
+ "step": 17433
+ },
+ {
+ "epoch": 55.69968051118211,
+ "grad_norm": 0.08935546875,
+ "learning_rate": 0.0005,
+ "loss": 1.02,
+ "step": 17434
+ },
+ {
+ "epoch": 55.70287539936102,
+ "grad_norm": 0.208984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0193,
+ "step": 17435
+ },
+ {
+ "epoch": 55.70607028753994,
+ "grad_norm": 0.08154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0205,
+ "step": 17436
+ },
+ {
+ "epoch": 55.70926517571885,
+ "grad_norm": 0.208984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0314,
+ "step": 17437
+ },
+ {
+ "epoch": 55.712460063897765,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0332,
+ "step": 17438
+ },
+ {
+ "epoch": 55.715654952076676,
+ "grad_norm": 0.1708984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0299,
+ "step": 17439
+ },
+ {
+ "epoch": 55.718849840255594,
+ "grad_norm": 0.08154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0289,
+ "step": 17440
+ },
+ {
+ "epoch": 55.722044728434504,
+ "grad_norm": 0.134765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0266,
+ "step": 17441
+ },
+ {
+ "epoch": 55.72523961661342,
+ "grad_norm": 0.11181640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0213,
+ "step": 17442
+ },
+ {
+ "epoch": 55.72843450479233,
+ "grad_norm": 0.169921875,
+ "learning_rate": 0.0005,
+ "loss": 1.022,
+ "step": 17443
+ },
+ {
+ "epoch": 55.73162939297124,
+ "grad_norm": 0.115234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0213,
+ "step": 17444
+ },
+ {
+ "epoch": 55.73482428115016,
+ "grad_norm": 0.1767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0281,
+ "step": 17445
+ },
+ {
+ "epoch": 55.73801916932907,
+ "grad_norm": 0.115234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0209,
+ "step": 17446
+ },
+ {
+ "epoch": 55.74121405750799,
+ "grad_norm": 0.14453125,
+ "learning_rate": 0.0005,
+ "loss": 1.032,
+ "step": 17447
+ },
+ {
+ "epoch": 55.7444089456869,
+ "grad_norm": 0.1328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0165,
+ "step": 17448
+ },
+ {
+ "epoch": 55.74760383386582,
+ "grad_norm": 0.10888671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0221,
+ "step": 17449
+ },
+ {
+ "epoch": 55.75079872204473,
+ "grad_norm": 0.1298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0359,
+ "step": 17450
+ },
+ {
+ "epoch": 55.753993610223645,
+ "grad_norm": 0.052978515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0316,
+ "step": 17451
+ },
+ {
+ "epoch": 55.757188498402556,
+ "grad_norm": 0.1015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0325,
+ "step": 17452
+ },
+ {
+ "epoch": 55.760383386581466,
+ "grad_norm": 0.053466796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0295,
+ "step": 17453
+ },
+ {
+ "epoch": 55.763578274760384,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0246,
+ "step": 17454
+ },
+ {
+ "epoch": 55.766773162939295,
+ "grad_norm": 0.08251953125,
+ "learning_rate": 0.0005,
+ "loss": 1.028,
+ "step": 17455
+ },
+ {
+ "epoch": 55.76996805111821,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0357,
+ "step": 17456
+ },
+ {
+ "epoch": 55.77316293929712,
+ "grad_norm": 0.0888671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0302,
+ "step": 17457
+ },
+ {
+ "epoch": 55.77635782747604,
+ "grad_norm": 0.0517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0216,
+ "step": 17458
+ },
+ {
+ "epoch": 55.77955271565495,
+ "grad_norm": 0.10009765625,
+ "learning_rate": 0.0005,
+ "loss": 1.032,
+ "step": 17459
+ },
+ {
+ "epoch": 55.78274760383387,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0231,
+ "step": 17460
+ },
+ {
+ "epoch": 55.78594249201278,
+ "grad_norm": 0.047119140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0233,
+ "step": 17461
+ },
+ {
+ "epoch": 55.78913738019169,
+ "grad_norm": 0.08984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0278,
+ "step": 17462
+ },
+ {
+ "epoch": 55.79233226837061,
+ "grad_norm": 0.052490234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0253,
+ "step": 17463
+ },
+ {
+ "epoch": 55.79552715654952,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0258,
+ "step": 17464
+ },
+ {
+ "epoch": 55.798722044728436,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0356,
+ "step": 17465
+ },
+ {
+ "epoch": 55.801916932907346,
+ "grad_norm": 0.080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0179,
+ "step": 17466
+ },
+ {
+ "epoch": 55.805111821086264,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0158,
+ "step": 17467
+ },
+ {
+ "epoch": 55.808306709265175,
+ "grad_norm": 0.046142578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0387,
+ "step": 17468
+ },
+ {
+ "epoch": 55.81150159744409,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0141,
+ "step": 17469
+ },
+ {
+ "epoch": 55.814696485623,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0395,
+ "step": 17470
+ },
+ {
+ "epoch": 55.81789137380191,
+ "grad_norm": 0.046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0183,
+ "step": 17471
+ },
+ {
+ "epoch": 55.82108626198083,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0249,
+ "step": 17472
+ },
+ {
+ "epoch": 55.82428115015974,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0275,
+ "step": 17473
+ },
+ {
+ "epoch": 55.82747603833866,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.015,
+ "step": 17474
+ },
+ {
+ "epoch": 55.83067092651757,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0261,
+ "step": 17475
+ },
+ {
+ "epoch": 55.83386581469649,
+ "grad_norm": 0.0576171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0334,
+ "step": 17476
+ },
+ {
+ "epoch": 55.8370607028754,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0248,
+ "step": 17477
+ },
+ {
+ "epoch": 55.840255591054316,
+ "grad_norm": 0.059814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0255,
+ "step": 17478
+ },
+ {
+ "epoch": 55.843450479233226,
+ "grad_norm": 0.08154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.022,
+ "step": 17479
+ },
+ {
+ "epoch": 55.846645367412144,
+ "grad_norm": 0.046630859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0178,
+ "step": 17480
+ },
+ {
+ "epoch": 55.849840255591054,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0315,
+ "step": 17481
+ },
+ {
+ "epoch": 55.853035143769965,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0303,
+ "step": 17482
+ },
+ {
+ "epoch": 55.85623003194888,
+ "grad_norm": 0.055419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0272,
+ "step": 17483
+ },
+ {
+ "epoch": 55.85942492012779,
+ "grad_norm": 0.061767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0321,
+ "step": 17484
+ },
+ {
+ "epoch": 55.86261980830671,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0307,
+ "step": 17485
+ },
+ {
+ "epoch": 55.86581469648562,
+ "grad_norm": 0.09765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0128,
+ "step": 17486
+ },
+ {
+ "epoch": 55.86900958466454,
+ "grad_norm": 0.0830078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0213,
+ "step": 17487
+ },
+ {
+ "epoch": 55.87220447284345,
+ "grad_norm": 0.1748046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0269,
+ "step": 17488
+ },
+ {
+ "epoch": 55.87539936102237,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0394,
+ "step": 17489
+ },
+ {
+ "epoch": 55.87859424920128,
+ "grad_norm": 0.166015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0361,
+ "step": 17490
+ },
+ {
+ "epoch": 55.88178913738019,
+ "grad_norm": 0.09326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0223,
+ "step": 17491
+ },
+ {
+ "epoch": 55.884984025559106,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0271,
+ "step": 17492
+ },
+ {
+ "epoch": 55.88817891373802,
+ "grad_norm": 0.1416015625,
+ "learning_rate": 0.0005,
+ "loss": 1.029,
+ "step": 17493
+ },
+ {
+ "epoch": 55.891373801916934,
+ "grad_norm": 0.11669921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0137,
+ "step": 17494
+ },
+ {
+ "epoch": 55.894568690095845,
+ "grad_norm": 0.1591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0278,
+ "step": 17495
+ },
+ {
+ "epoch": 55.89776357827476,
+ "grad_norm": 0.1669921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0261,
+ "step": 17496
+ },
+ {
+ "epoch": 55.90095846645367,
+ "grad_norm": 0.10693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0363,
+ "step": 17497
+ },
+ {
+ "epoch": 55.90415335463259,
+ "grad_norm": 0.11669921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0148,
+ "step": 17498
+ },
+ {
+ "epoch": 55.9073482428115,
+ "grad_norm": 0.12890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0361,
+ "step": 17499
+ },
+ {
+ "epoch": 55.91054313099041,
+ "grad_norm": 0.10693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0247,
+ "step": 17500
+ },
+ {
+ "epoch": 55.91373801916933,
+ "grad_norm": 0.15234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0209,
+ "step": 17501
+ },
+ {
+ "epoch": 55.91693290734824,
+ "grad_norm": 0.05419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0318,
+ "step": 17502
+ },
+ {
+ "epoch": 55.92012779552716,
+ "grad_norm": 0.140625,
+ "learning_rate": 0.0005,
+ "loss": 1.03,
+ "step": 17503
+ },
+ {
+ "epoch": 55.92332268370607,
+ "grad_norm": 0.1572265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0191,
+ "step": 17504
+ },
+ {
+ "epoch": 55.926517571884986,
+ "grad_norm": 0.09912109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0378,
+ "step": 17505
+ },
+ {
+ "epoch": 55.9297124600639,
+ "grad_norm": 0.142578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0186,
+ "step": 17506
+ },
+ {
+ "epoch": 55.932907348242814,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0209,
+ "step": 17507
+ },
+ {
+ "epoch": 55.936102236421725,
+ "grad_norm": 0.059814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0215,
+ "step": 17508
+ },
+ {
+ "epoch": 55.93929712460064,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0209,
+ "step": 17509
+ },
+ {
+ "epoch": 55.94249201277955,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.028,
+ "step": 17510
+ },
+ {
+ "epoch": 55.945686900958464,
+ "grad_norm": 0.056884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0224,
+ "step": 17511
+ },
+ {
+ "epoch": 55.94888178913738,
+ "grad_norm": 0.046142578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0279,
+ "step": 17512
+ },
+ {
+ "epoch": 55.95207667731629,
+ "grad_norm": 0.08056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0293,
+ "step": 17513
+ },
+ {
+ "epoch": 55.95527156549521,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0176,
+ "step": 17514
+ },
+ {
+ "epoch": 55.95846645367412,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0309,
+ "step": 17515
+ },
+ {
+ "epoch": 55.96166134185304,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0275,
+ "step": 17516
+ },
+ {
+ "epoch": 55.96485623003195,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0285,
+ "step": 17517
+ },
+ {
+ "epoch": 55.968051118210866,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.031,
+ "step": 17518
+ },
+ {
+ "epoch": 55.97124600638978,
+ "grad_norm": 0.048583984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0233,
+ "step": 17519
+ },
+ {
+ "epoch": 55.97444089456869,
+ "grad_norm": 0.08740234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0173,
+ "step": 17520
+ },
+ {
+ "epoch": 55.977635782747605,
+ "grad_norm": 0.080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0301,
+ "step": 17521
+ },
+ {
+ "epoch": 55.980830670926515,
+ "grad_norm": 0.055419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0249,
+ "step": 17522
+ },
+ {
+ "epoch": 55.98402555910543,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0296,
+ "step": 17523
+ },
+ {
+ "epoch": 55.98722044728434,
+ "grad_norm": 0.047119140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0287,
+ "step": 17524
+ },
+ {
+ "epoch": 55.99041533546326,
+ "grad_norm": 0.052734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0167,
+ "step": 17525
+ },
+ {
+ "epoch": 55.99361022364217,
+ "grad_norm": 0.052490234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0324,
+ "step": 17526
+ },
+ {
+ "epoch": 55.99680511182109,
+ "grad_norm": 0.05712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0281,
+ "step": 17527
+ },
+ {
+ "epoch": 56.0,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0225,
+ "step": 17528
+ },
+ {
+ "epoch": 56.00319488817891,
+ "grad_norm": 0.056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0171,
+ "step": 17529
+ },
+ {
+ "epoch": 56.00638977635783,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0278,
+ "step": 17530
+ },
+ {
+ "epoch": 56.00958466453674,
+ "grad_norm": 0.09033203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0165,
+ "step": 17531
+ },
+ {
+ "epoch": 56.01277955271566,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0293,
+ "step": 17532
+ },
+ {
+ "epoch": 56.01597444089457,
+ "grad_norm": 0.12890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0264,
+ "step": 17533
+ },
+ {
+ "epoch": 56.019169329073485,
+ "grad_norm": 0.052734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0279,
+ "step": 17534
+ },
+ {
+ "epoch": 56.022364217252395,
+ "grad_norm": 0.05615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0318,
+ "step": 17535
+ },
+ {
+ "epoch": 56.02555910543131,
+ "grad_norm": 0.060302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0318,
+ "step": 17536
+ },
+ {
+ "epoch": 56.02875399361022,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0354,
+ "step": 17537
+ },
+ {
+ "epoch": 56.031948881789134,
+ "grad_norm": 0.051513671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0324,
+ "step": 17538
+ },
+ {
+ "epoch": 56.03514376996805,
+ "grad_norm": 0.046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0164,
+ "step": 17539
+ },
+ {
+ "epoch": 56.03833865814696,
+ "grad_norm": 0.056884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.029,
+ "step": 17540
+ },
+ {
+ "epoch": 56.04153354632588,
+ "grad_norm": 0.050048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0126,
+ "step": 17541
+ },
+ {
+ "epoch": 56.04472843450479,
+ "grad_norm": 0.05615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0275,
+ "step": 17542
+ },
+ {
+ "epoch": 56.04792332268371,
+ "grad_norm": 0.051025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.028,
+ "step": 17543
+ },
+ {
+ "epoch": 56.05111821086262,
+ "grad_norm": 0.056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0279,
+ "step": 17544
+ },
+ {
+ "epoch": 56.054313099041536,
+ "grad_norm": 0.04931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0194,
+ "step": 17545
+ },
+ {
+ "epoch": 56.05750798722045,
+ "grad_norm": 0.054443359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0261,
+ "step": 17546
+ },
+ {
+ "epoch": 56.06070287539936,
+ "grad_norm": 0.049072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0279,
+ "step": 17547
+ },
+ {
+ "epoch": 56.063897763578275,
+ "grad_norm": 0.043701171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0333,
+ "step": 17548
+ },
+ {
+ "epoch": 56.067092651757186,
+ "grad_norm": 0.052001953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0172,
+ "step": 17549
+ },
+ {
+ "epoch": 56.0702875399361,
+ "grad_norm": 0.047119140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0173,
+ "step": 17550
+ },
+ {
+ "epoch": 56.073482428115014,
+ "grad_norm": 0.0458984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0176,
+ "step": 17551
+ },
+ {
+ "epoch": 56.07667731629393,
+ "grad_norm": 0.0546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0286,
+ "step": 17552
+ },
+ {
+ "epoch": 56.07987220447284,
+ "grad_norm": 0.050048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0117,
+ "step": 17553
+ },
+ {
+ "epoch": 56.08306709265176,
+ "grad_norm": 0.043212890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0192,
+ "step": 17554
+ },
+ {
+ "epoch": 56.08626198083067,
+ "grad_norm": 0.051025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0451,
+ "step": 17555
+ },
+ {
+ "epoch": 56.08945686900959,
+ "grad_norm": 0.044921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0178,
+ "step": 17556
+ },
+ {
+ "epoch": 56.0926517571885,
+ "grad_norm": 0.05859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0096,
+ "step": 17557
+ },
+ {
+ "epoch": 56.09584664536741,
+ "grad_norm": 0.04541015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0218,
+ "step": 17558
+ },
+ {
+ "epoch": 56.09904153354633,
+ "grad_norm": 0.049072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0238,
+ "step": 17559
+ },
+ {
+ "epoch": 56.10223642172524,
+ "grad_norm": 0.05517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0114,
+ "step": 17560
+ },
+ {
+ "epoch": 56.105431309904155,
+ "grad_norm": 0.051513671875,
+ "learning_rate": 0.0005,
+ "loss": 1.02,
+ "step": 17561
+ },
+ {
+ "epoch": 56.108626198083066,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0299,
+ "step": 17562
+ },
+ {
+ "epoch": 56.11182108626198,
+ "grad_norm": 0.048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0266,
+ "step": 17563
+ },
+ {
+ "epoch": 56.115015974440894,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0308,
+ "step": 17564
+ },
+ {
+ "epoch": 56.11821086261981,
+ "grad_norm": 0.05419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.02,
+ "step": 17565
+ },
+ {
+ "epoch": 56.12140575079872,
+ "grad_norm": 0.05419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.019,
+ "step": 17566
+ },
+ {
+ "epoch": 56.12460063897763,
+ "grad_norm": 0.060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0269,
+ "step": 17567
+ },
+ {
+ "epoch": 56.12779552715655,
+ "grad_norm": 0.048583984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0241,
+ "step": 17568
+ },
+ {
+ "epoch": 56.13099041533546,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0289,
+ "step": 17569
+ },
+ {
+ "epoch": 56.13418530351438,
+ "grad_norm": 0.0517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0298,
+ "step": 17570
+ },
+ {
+ "epoch": 56.13738019169329,
+ "grad_norm": 0.055908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0249,
+ "step": 17571
+ },
+ {
+ "epoch": 56.14057507987221,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0271,
+ "step": 17572
+ },
+ {
+ "epoch": 56.14376996805112,
+ "grad_norm": 0.056884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0143,
+ "step": 17573
+ },
+ {
+ "epoch": 56.146964856230035,
+ "grad_norm": 0.08203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0243,
+ "step": 17574
+ },
+ {
+ "epoch": 56.150159744408946,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.019,
+ "step": 17575
+ },
+ {
+ "epoch": 56.153354632587856,
+ "grad_norm": 0.06005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0259,
+ "step": 17576
+ },
+ {
+ "epoch": 56.156549520766774,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0282,
+ "step": 17577
+ },
+ {
+ "epoch": 56.159744408945684,
+ "grad_norm": 0.0595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0289,
+ "step": 17578
+ },
+ {
+ "epoch": 56.1629392971246,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0123,
+ "step": 17579
+ },
+ {
+ "epoch": 56.16613418530351,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0349,
+ "step": 17580
+ },
+ {
+ "epoch": 56.16932907348243,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0317,
+ "step": 17581
+ },
+ {
+ "epoch": 56.17252396166134,
+ "grad_norm": 0.08935546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0242,
+ "step": 17582
+ },
+ {
+ "epoch": 56.17571884984026,
+ "grad_norm": 0.061767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0213,
+ "step": 17583
+ },
+ {
+ "epoch": 56.17891373801917,
+ "grad_norm": 0.08642578125,
+ "learning_rate": 0.0005,
+ "loss": 1.017,
+ "step": 17584
+ },
+ {
+ "epoch": 56.18210862619808,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.025,
+ "step": 17585
+ },
+ {
+ "epoch": 56.185303514377,
+ "grad_norm": 0.0546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0341,
+ "step": 17586
+ },
+ {
+ "epoch": 56.18849840255591,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0354,
+ "step": 17587
+ },
+ {
+ "epoch": 56.191693290734825,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0248,
+ "step": 17588
+ },
+ {
+ "epoch": 56.194888178913736,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0158,
+ "step": 17589
+ },
+ {
+ "epoch": 56.198083067092654,
+ "grad_norm": 0.08349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0195,
+ "step": 17590
+ },
+ {
+ "epoch": 56.201277955271564,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0334,
+ "step": 17591
+ },
+ {
+ "epoch": 56.20447284345048,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0312,
+ "step": 17592
+ },
+ {
+ "epoch": 56.20766773162939,
+ "grad_norm": 0.0498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0251,
+ "step": 17593
+ },
+ {
+ "epoch": 56.21086261980831,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0293,
+ "step": 17594
+ },
+ {
+ "epoch": 56.21405750798722,
+ "grad_norm": 0.08837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0208,
+ "step": 17595
+ },
+ {
+ "epoch": 56.21725239616613,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0377,
+ "step": 17596
+ },
+ {
+ "epoch": 56.22044728434505,
+ "grad_norm": 0.0458984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0132,
+ "step": 17597
+ },
+ {
+ "epoch": 56.22364217252396,
+ "grad_norm": 0.059814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0191,
+ "step": 17598
+ },
+ {
+ "epoch": 56.22683706070288,
+ "grad_norm": 0.05908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0299,
+ "step": 17599
+ },
+ {
+ "epoch": 56.23003194888179,
+ "grad_norm": 0.06201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0158,
+ "step": 17600
+ },
+ {
+ "epoch": 56.233226837060705,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0269,
+ "step": 17601
+ },
+ {
+ "epoch": 56.236421725239616,
+ "grad_norm": 0.0439453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0351,
+ "step": 17602
+ },
+ {
+ "epoch": 56.239616613418534,
+ "grad_norm": 0.08203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0251,
+ "step": 17603
+ },
+ {
+ "epoch": 56.242811501597444,
+ "grad_norm": 0.07861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0226,
+ "step": 17604
+ },
+ {
+ "epoch": 56.246006389776355,
+ "grad_norm": 0.05029296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0298,
+ "step": 17605
+ },
+ {
+ "epoch": 56.24920127795527,
+ "grad_norm": 0.057861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0215,
+ "step": 17606
+ },
+ {
+ "epoch": 56.25239616613418,
+ "grad_norm": 0.05908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0215,
+ "step": 17607
+ },
+ {
+ "epoch": 56.2555910543131,
+ "grad_norm": 0.05078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0224,
+ "step": 17608
+ },
+ {
+ "epoch": 56.25878594249201,
+ "grad_norm": 0.050048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0257,
+ "step": 17609
+ },
+ {
+ "epoch": 56.26198083067093,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0197,
+ "step": 17610
+ },
+ {
+ "epoch": 56.26517571884984,
+ "grad_norm": 0.05810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0172,
+ "step": 17611
+ },
+ {
+ "epoch": 56.26837060702876,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0275,
+ "step": 17612
+ },
+ {
+ "epoch": 56.27156549520767,
+ "grad_norm": 0.044677734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0321,
+ "step": 17613
+ },
+ {
+ "epoch": 56.27476038338658,
+ "grad_norm": 0.043212890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0173,
+ "step": 17614
+ },
+ {
+ "epoch": 56.277955271565496,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0404,
+ "step": 17615
+ },
+ {
+ "epoch": 56.281150159744406,
+ "grad_norm": 0.052978515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0338,
+ "step": 17616
+ },
+ {
+ "epoch": 56.284345047923324,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0323,
+ "step": 17617
+ },
+ {
+ "epoch": 56.287539936102235,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0267,
+ "step": 17618
+ },
+ {
+ "epoch": 56.29073482428115,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0343,
+ "step": 17619
+ },
+ {
+ "epoch": 56.29392971246006,
+ "grad_norm": 0.059326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.026,
+ "step": 17620
+ },
+ {
+ "epoch": 56.29712460063898,
+ "grad_norm": 0.05224609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0266,
+ "step": 17621
+ },
+ {
+ "epoch": 56.30031948881789,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0285,
+ "step": 17622
+ },
+ {
+ "epoch": 56.3035143769968,
+ "grad_norm": 0.047607421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0328,
+ "step": 17623
+ },
+ {
+ "epoch": 56.30670926517572,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0218,
+ "step": 17624
+ },
+ {
+ "epoch": 56.30990415335463,
+ "grad_norm": 0.05029296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0292,
+ "step": 17625
+ },
+ {
+ "epoch": 56.31309904153355,
+ "grad_norm": 0.05078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0249,
+ "step": 17626
+ },
+ {
+ "epoch": 56.31629392971246,
+ "grad_norm": 0.052001953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0261,
+ "step": 17627
+ },
+ {
+ "epoch": 56.319488817891376,
+ "grad_norm": 0.060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0326,
+ "step": 17628
+ },
+ {
+ "epoch": 56.322683706070286,
+ "grad_norm": 0.0498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.025,
+ "step": 17629
+ },
+ {
+ "epoch": 56.325878594249204,
+ "grad_norm": 0.0478515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0204,
+ "step": 17630
+ },
+ {
+ "epoch": 56.329073482428115,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0215,
+ "step": 17631
+ },
+ {
+ "epoch": 56.33226837060703,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0227,
+ "step": 17632
+ },
+ {
+ "epoch": 56.33546325878594,
+ "grad_norm": 0.05859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0276,
+ "step": 17633
+ },
+ {
+ "epoch": 56.33865814696485,
+ "grad_norm": 0.0859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0246,
+ "step": 17634
+ },
+ {
+ "epoch": 56.34185303514377,
+ "grad_norm": 0.051513671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0179,
+ "step": 17635
+ },
+ {
+ "epoch": 56.34504792332268,
+ "grad_norm": 0.057373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0342,
+ "step": 17636
+ },
+ {
+ "epoch": 56.3482428115016,
+ "grad_norm": 0.061767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.032,
+ "step": 17637
+ },
+ {
+ "epoch": 56.35143769968051,
+ "grad_norm": 0.0458984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0221,
+ "step": 17638
+ },
+ {
+ "epoch": 56.35463258785943,
+ "grad_norm": 0.060302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0237,
+ "step": 17639
+ },
+ {
+ "epoch": 56.35782747603834,
+ "grad_norm": 0.051025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0221,
+ "step": 17640
+ },
+ {
+ "epoch": 56.361022364217256,
+ "grad_norm": 0.059814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0239,
+ "step": 17641
+ },
+ {
+ "epoch": 56.364217252396166,
+ "grad_norm": 0.048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0179,
+ "step": 17642
+ },
+ {
+ "epoch": 56.36741214057508,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0193,
+ "step": 17643
+ },
+ {
+ "epoch": 56.370607028753994,
+ "grad_norm": 0.056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0178,
+ "step": 17644
+ },
+ {
+ "epoch": 56.373801916932905,
+ "grad_norm": 0.05126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.029,
+ "step": 17645
+ },
+ {
+ "epoch": 56.37699680511182,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0249,
+ "step": 17646
+ },
+ {
+ "epoch": 56.38019169329073,
+ "grad_norm": 0.050048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0271,
+ "step": 17647
+ },
+ {
+ "epoch": 56.38338658146965,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0241,
+ "step": 17648
+ },
+ {
+ "epoch": 56.38658146964856,
+ "grad_norm": 0.059326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0197,
+ "step": 17649
+ },
+ {
+ "epoch": 56.38977635782748,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0348,
+ "step": 17650
+ },
+ {
+ "epoch": 56.39297124600639,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0265,
+ "step": 17651
+ },
+ {
+ "epoch": 56.3961661341853,
+ "grad_norm": 0.054443359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0301,
+ "step": 17652
+ },
+ {
+ "epoch": 56.39936102236422,
+ "grad_norm": 0.044677734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0254,
+ "step": 17653
+ },
+ {
+ "epoch": 56.40255591054313,
+ "grad_norm": 0.1201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0281,
+ "step": 17654
+ },
+ {
+ "epoch": 56.405750798722046,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0251,
+ "step": 17655
+ },
+ {
+ "epoch": 56.40894568690096,
+ "grad_norm": 0.0576171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0329,
+ "step": 17656
+ },
+ {
+ "epoch": 56.412140575079874,
+ "grad_norm": 0.11572265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0305,
+ "step": 17657
+ },
+ {
+ "epoch": 56.415335463258785,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.025,
+ "step": 17658
+ },
+ {
+ "epoch": 56.4185303514377,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0251,
+ "step": 17659
+ },
+ {
+ "epoch": 56.42172523961661,
+ "grad_norm": 0.1083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.033,
+ "step": 17660
+ },
+ {
+ "epoch": 56.424920127795524,
+ "grad_norm": 0.051513671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0328,
+ "step": 17661
+ },
+ {
+ "epoch": 56.42811501597444,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0337,
+ "step": 17662
+ },
+ {
+ "epoch": 56.43130990415335,
+ "grad_norm": 0.09521484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0349,
+ "step": 17663
+ },
+ {
+ "epoch": 56.43450479233227,
+ "grad_norm": 0.05810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0301,
+ "step": 17664
+ },
+ {
+ "epoch": 56.43769968051118,
+ "grad_norm": 0.0908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.012,
+ "step": 17665
+ },
+ {
+ "epoch": 56.4408945686901,
+ "grad_norm": 0.1298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0264,
+ "step": 17666
+ },
+ {
+ "epoch": 56.44408945686901,
+ "grad_norm": 0.10791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0261,
+ "step": 17667
+ },
+ {
+ "epoch": 56.447284345047926,
+ "grad_norm": 0.08447265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0315,
+ "step": 17668
+ },
+ {
+ "epoch": 56.45047923322684,
+ "grad_norm": 0.1201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0282,
+ "step": 17669
+ },
+ {
+ "epoch": 56.453674121405754,
+ "grad_norm": 0.05859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0368,
+ "step": 17670
+ },
+ {
+ "epoch": 56.456869009584665,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0214,
+ "step": 17671
+ },
+ {
+ "epoch": 56.460063897763575,
+ "grad_norm": 0.1279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0168,
+ "step": 17672
+ },
+ {
+ "epoch": 56.46325878594249,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0173,
+ "step": 17673
+ },
+ {
+ "epoch": 56.466453674121404,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0142,
+ "step": 17674
+ },
+ {
+ "epoch": 56.46964856230032,
+ "grad_norm": 0.056884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0187,
+ "step": 17675
+ },
+ {
+ "epoch": 56.47284345047923,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0256,
+ "step": 17676
+ },
+ {
+ "epoch": 56.47603833865815,
+ "grad_norm": 0.055419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0211,
+ "step": 17677
+ },
+ {
+ "epoch": 56.47923322683706,
+ "grad_norm": 0.047607421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0233,
+ "step": 17678
+ },
+ {
+ "epoch": 56.48242811501598,
+ "grad_norm": 0.05322265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0308,
+ "step": 17679
+ },
+ {
+ "epoch": 56.48562300319489,
+ "grad_norm": 0.0478515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0367,
+ "step": 17680
+ },
+ {
+ "epoch": 56.4888178913738,
+ "grad_norm": 0.046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0286,
+ "step": 17681
+ },
+ {
+ "epoch": 56.49201277955272,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0129,
+ "step": 17682
+ },
+ {
+ "epoch": 56.49520766773163,
+ "grad_norm": 0.046142578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0203,
+ "step": 17683
+ },
+ {
+ "epoch": 56.498402555910545,
+ "grad_norm": 0.08837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0276,
+ "step": 17684
+ },
+ {
+ "epoch": 56.501597444089455,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.03,
+ "step": 17685
+ },
+ {
+ "epoch": 56.50479233226837,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0163,
+ "step": 17686
+ },
+ {
+ "epoch": 56.50798722044728,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0211,
+ "step": 17687
+ },
+ {
+ "epoch": 56.5111821086262,
+ "grad_norm": 0.044189453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0194,
+ "step": 17688
+ },
+ {
+ "epoch": 56.51437699680511,
+ "grad_norm": 0.047119140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0178,
+ "step": 17689
+ },
+ {
+ "epoch": 56.51757188498402,
+ "grad_norm": 0.050537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0449,
+ "step": 17690
+ },
+ {
+ "epoch": 56.52076677316294,
+ "grad_norm": 0.05908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0211,
+ "step": 17691
+ },
+ {
+ "epoch": 56.52396166134185,
+ "grad_norm": 0.0537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0245,
+ "step": 17692
+ },
+ {
+ "epoch": 56.52715654952077,
+ "grad_norm": 0.057861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0231,
+ "step": 17693
+ },
+ {
+ "epoch": 56.53035143769968,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0213,
+ "step": 17694
+ },
+ {
+ "epoch": 56.533546325878596,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0314,
+ "step": 17695
+ },
+ {
+ "epoch": 56.53674121405751,
+ "grad_norm": 0.0556640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0222,
+ "step": 17696
+ },
+ {
+ "epoch": 56.539936102236425,
+ "grad_norm": 0.04833984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0193,
+ "step": 17697
+ },
+ {
+ "epoch": 56.543130990415335,
+ "grad_norm": 0.08447265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0177,
+ "step": 17698
+ },
+ {
+ "epoch": 56.546325878594246,
+ "grad_norm": 0.0498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0238,
+ "step": 17699
+ },
+ {
+ "epoch": 56.54952076677316,
+ "grad_norm": 0.060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0374,
+ "step": 17700
+ },
+ {
+ "epoch": 56.552715654952074,
+ "grad_norm": 0.0849609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0228,
+ "step": 17701
+ },
+ {
+ "epoch": 56.55591054313099,
+ "grad_norm": 0.08447265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0342,
+ "step": 17702
+ },
+ {
+ "epoch": 56.5591054313099,
+ "grad_norm": 0.08349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0362,
+ "step": 17703
+ },
+ {
+ "epoch": 56.56230031948882,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0278,
+ "step": 17704
+ },
+ {
+ "epoch": 56.56549520766773,
+ "grad_norm": 0.045166015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0291,
+ "step": 17705
+ },
+ {
+ "epoch": 56.56869009584665,
+ "grad_norm": 0.056396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.025,
+ "step": 17706
+ },
+ {
+ "epoch": 56.57188498402556,
+ "grad_norm": 0.054931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0261,
+ "step": 17707
+ },
+ {
+ "epoch": 56.575079872204476,
+ "grad_norm": 0.0595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0212,
+ "step": 17708
+ },
+ {
+ "epoch": 56.57827476038339,
+ "grad_norm": 0.057861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0113,
+ "step": 17709
+ },
+ {
+ "epoch": 56.5814696485623,
+ "grad_norm": 0.06005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0284,
+ "step": 17710
+ },
+ {
+ "epoch": 56.584664536741215,
+ "grad_norm": 0.05615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.02,
+ "step": 17711
+ },
+ {
+ "epoch": 56.587859424920126,
+ "grad_norm": 0.05029296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0175,
+ "step": 17712
+ },
+ {
+ "epoch": 56.59105431309904,
+ "grad_norm": 0.0517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0175,
+ "step": 17713
+ },
+ {
+ "epoch": 56.594249201277954,
+ "grad_norm": 0.0576171875,
+ "learning_rate": 0.0005,
+ "loss": 1.018,
+ "step": 17714
+ },
+ {
+ "epoch": 56.59744408945687,
+ "grad_norm": 0.057861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0293,
+ "step": 17715
+ },
+ {
+ "epoch": 56.60063897763578,
+ "grad_norm": 0.05810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0313,
+ "step": 17716
+ },
+ {
+ "epoch": 56.6038338658147,
+ "grad_norm": 0.05029296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0345,
+ "step": 17717
+ },
+ {
+ "epoch": 56.60702875399361,
+ "grad_norm": 0.050537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0224,
+ "step": 17718
+ },
+ {
+ "epoch": 56.61022364217252,
+ "grad_norm": 0.056396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0262,
+ "step": 17719
+ },
+ {
+ "epoch": 56.61341853035144,
+ "grad_norm": 0.052978515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0327,
+ "step": 17720
+ },
+ {
+ "epoch": 56.61661341853035,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0269,
+ "step": 17721
+ },
+ {
+ "epoch": 56.61980830670927,
+ "grad_norm": 0.056396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0324,
+ "step": 17722
+ },
+ {
+ "epoch": 56.62300319488818,
+ "grad_norm": 0.05908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0183,
+ "step": 17723
+ },
+ {
+ "epoch": 56.626198083067095,
+ "grad_norm": 0.056884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0201,
+ "step": 17724
+ },
+ {
+ "epoch": 56.629392971246006,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0312,
+ "step": 17725
+ },
+ {
+ "epoch": 56.63258785942492,
+ "grad_norm": 0.0546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0232,
+ "step": 17726
+ },
+ {
+ "epoch": 56.635782747603834,
+ "grad_norm": 0.05615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0169,
+ "step": 17727
+ },
+ {
+ "epoch": 56.638977635782744,
+ "grad_norm": 0.052978515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0224,
+ "step": 17728
+ },
+ {
+ "epoch": 56.64217252396166,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0354,
+ "step": 17729
+ },
+ {
+ "epoch": 56.64536741214057,
+ "grad_norm": 0.059326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0093,
+ "step": 17730
+ },
+ {
+ "epoch": 56.64856230031949,
+ "grad_norm": 0.05810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0361,
+ "step": 17731
+ },
+ {
+ "epoch": 56.6517571884984,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.028,
+ "step": 17732
+ },
+ {
+ "epoch": 56.65495207667732,
+ "grad_norm": 0.059326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0294,
+ "step": 17733
+ },
+ {
+ "epoch": 56.65814696485623,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0198,
+ "step": 17734
+ },
+ {
+ "epoch": 56.66134185303515,
+ "grad_norm": 0.05712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0368,
+ "step": 17735
+ },
+ {
+ "epoch": 56.66453674121406,
+ "grad_norm": 0.056396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0268,
+ "step": 17736
+ },
+ {
+ "epoch": 56.66773162939297,
+ "grad_norm": 0.08837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0211,
+ "step": 17737
+ },
+ {
+ "epoch": 56.670926517571885,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0279,
+ "step": 17738
+ },
+ {
+ "epoch": 56.674121405750796,
+ "grad_norm": 0.11962890625,
+ "learning_rate": 0.0005,
+ "loss": 1.019,
+ "step": 17739
+ },
+ {
+ "epoch": 56.677316293929714,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0207,
+ "step": 17740
+ },
+ {
+ "epoch": 56.680511182108624,
+ "grad_norm": 0.0546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0301,
+ "step": 17741
+ },
+ {
+ "epoch": 56.68370607028754,
+ "grad_norm": 0.08544921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0251,
+ "step": 17742
+ },
+ {
+ "epoch": 56.68690095846645,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0389,
+ "step": 17743
+ },
+ {
+ "epoch": 56.69009584664537,
+ "grad_norm": 0.05615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0283,
+ "step": 17744
+ },
+ {
+ "epoch": 56.69329073482428,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0248,
+ "step": 17745
+ },
+ {
+ "epoch": 56.6964856230032,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0196,
+ "step": 17746
+ },
+ {
+ "epoch": 56.69968051118211,
+ "grad_norm": 0.09130859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0356,
+ "step": 17747
+ },
+ {
+ "epoch": 56.70287539936102,
+ "grad_norm": 0.062255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0222,
+ "step": 17748
+ },
+ {
+ "epoch": 56.70607028753994,
+ "grad_norm": 0.05859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0303,
+ "step": 17749
+ },
+ {
+ "epoch": 56.70926517571885,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0245,
+ "step": 17750
+ },
+ {
+ "epoch": 56.712460063897765,
+ "grad_norm": 0.056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0225,
+ "step": 17751
+ },
+ {
+ "epoch": 56.715654952076676,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0187,
+ "step": 17752
+ },
+ {
+ "epoch": 56.718849840255594,
+ "grad_norm": 0.057861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0245,
+ "step": 17753
+ },
+ {
+ "epoch": 56.722044728434504,
+ "grad_norm": 0.08203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0279,
+ "step": 17754
+ },
+ {
+ "epoch": 56.72523961661342,
+ "grad_norm": 0.05322265625,
+ "learning_rate": 0.0005,
+ "loss": 1.019,
+ "step": 17755
+ },
+ {
+ "epoch": 56.72843450479233,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0141,
+ "step": 17756
+ },
+ {
+ "epoch": 56.73162939297124,
+ "grad_norm": 0.055908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0242,
+ "step": 17757
+ },
+ {
+ "epoch": 56.73482428115016,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0349,
+ "step": 17758
+ },
+ {
+ "epoch": 56.73801916932907,
+ "grad_norm": 0.052490234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0278,
+ "step": 17759
+ },
+ {
+ "epoch": 56.74121405750799,
+ "grad_norm": 0.052734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0213,
+ "step": 17760
+ },
+ {
+ "epoch": 56.7444089456869,
+ "grad_norm": 0.046630859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0223,
+ "step": 17761
+ },
+ {
+ "epoch": 56.74760383386582,
+ "grad_norm": 0.049072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0275,
+ "step": 17762
+ },
+ {
+ "epoch": 56.75079872204473,
+ "grad_norm": 0.0595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0297,
+ "step": 17763
+ },
+ {
+ "epoch": 56.753993610223645,
+ "grad_norm": 0.04736328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0238,
+ "step": 17764
+ },
+ {
+ "epoch": 56.757188498402556,
+ "grad_norm": 0.046142578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0255,
+ "step": 17765
+ },
+ {
+ "epoch": 56.760383386581466,
+ "grad_norm": 0.052490234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0224,
+ "step": 17766
+ },
+ {
+ "epoch": 56.763578274760384,
+ "grad_norm": 0.052734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0257,
+ "step": 17767
+ },
+ {
+ "epoch": 56.766773162939295,
+ "grad_norm": 0.058837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0234,
+ "step": 17768
+ },
+ {
+ "epoch": 56.76996805111821,
+ "grad_norm": 0.07861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.016,
+ "step": 17769
+ },
+ {
+ "epoch": 56.77316293929712,
+ "grad_norm": 0.054443359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0144,
+ "step": 17770
+ },
+ {
+ "epoch": 56.77635782747604,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0163,
+ "step": 17771
+ },
+ {
+ "epoch": 56.77955271565495,
+ "grad_norm": 0.046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0386,
+ "step": 17772
+ },
+ {
+ "epoch": 56.78274760383387,
+ "grad_norm": 0.04931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0235,
+ "step": 17773
+ },
+ {
+ "epoch": 56.78594249201278,
+ "grad_norm": 0.054443359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0161,
+ "step": 17774
+ },
+ {
+ "epoch": 56.78913738019169,
+ "grad_norm": 0.056884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0286,
+ "step": 17775
+ },
+ {
+ "epoch": 56.79233226837061,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0169,
+ "step": 17776
+ },
+ {
+ "epoch": 56.79552715654952,
+ "grad_norm": 0.052978515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0261,
+ "step": 17777
+ },
+ {
+ "epoch": 56.798722044728436,
+ "grad_norm": 0.0458984375,
+ "learning_rate": 0.0005,
+ "loss": 1.03,
+ "step": 17778
+ },
+ {
+ "epoch": 56.801916932907346,
+ "grad_norm": 0.05078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0293,
+ "step": 17779
+ },
+ {
+ "epoch": 56.805111821086264,
+ "grad_norm": 0.052490234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0267,
+ "step": 17780
+ },
+ {
+ "epoch": 56.808306709265175,
+ "grad_norm": 0.05126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0289,
+ "step": 17781
+ },
+ {
+ "epoch": 56.81150159744409,
+ "grad_norm": 0.056396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0311,
+ "step": 17782
+ },
+ {
+ "epoch": 56.814696485623,
+ "grad_norm": 0.056396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0233,
+ "step": 17783
+ },
+ {
+ "epoch": 56.81789137380191,
+ "grad_norm": 0.0537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0115,
+ "step": 17784
+ },
+ {
+ "epoch": 56.82108626198083,
+ "grad_norm": 0.054931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0293,
+ "step": 17785
+ },
+ {
+ "epoch": 56.82428115015974,
+ "grad_norm": 0.0869140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0312,
+ "step": 17786
+ },
+ {
+ "epoch": 56.82747603833866,
+ "grad_norm": 0.0556640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0256,
+ "step": 17787
+ },
+ {
+ "epoch": 56.83067092651757,
+ "grad_norm": 0.1083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0221,
+ "step": 17788
+ },
+ {
+ "epoch": 56.83386581469649,
+ "grad_norm": 0.099609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0296,
+ "step": 17789
+ },
+ {
+ "epoch": 56.8370607028754,
+ "grad_norm": 0.054443359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0271,
+ "step": 17790
+ },
+ {
+ "epoch": 56.840255591054316,
+ "grad_norm": 0.1025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0275,
+ "step": 17791
+ },
+ {
+ "epoch": 56.843450479233226,
+ "grad_norm": 0.08447265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0233,
+ "step": 17792
+ },
+ {
+ "epoch": 56.846645367412144,
+ "grad_norm": 0.08154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0294,
+ "step": 17793
+ },
+ {
+ "epoch": 56.849840255591054,
+ "grad_norm": 0.22265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0208,
+ "step": 17794
+ },
+ {
+ "epoch": 56.853035143769965,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0312,
+ "step": 17795
+ },
+ {
+ "epoch": 56.85623003194888,
+ "grad_norm": 0.09375,
+ "learning_rate": 0.0005,
+ "loss": 1.0298,
+ "step": 17796
+ },
+ {
+ "epoch": 56.85942492012779,
+ "grad_norm": 0.0859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0263,
+ "step": 17797
+ },
+ {
+ "epoch": 56.86261980830671,
+ "grad_norm": 0.08642578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0238,
+ "step": 17798
+ },
+ {
+ "epoch": 56.86581469648562,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0356,
+ "step": 17799
+ },
+ {
+ "epoch": 56.86900958466454,
+ "grad_norm": 0.10888671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0258,
+ "step": 17800
+ },
+ {
+ "epoch": 56.87220447284345,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0216,
+ "step": 17801
+ },
+ {
+ "epoch": 56.87539936102237,
+ "grad_norm": 0.08154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0321,
+ "step": 17802
+ },
+ {
+ "epoch": 56.87859424920128,
+ "grad_norm": 0.059814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0333,
+ "step": 17803
+ },
+ {
+ "epoch": 56.88178913738019,
+ "grad_norm": 0.04736328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0168,
+ "step": 17804
+ },
+ {
+ "epoch": 56.884984025559106,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0244,
+ "step": 17805
+ },
+ {
+ "epoch": 56.88817891373802,
+ "grad_norm": 0.056884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0296,
+ "step": 17806
+ },
+ {
+ "epoch": 56.891373801916934,
+ "grad_norm": 0.08056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0286,
+ "step": 17807
+ },
+ {
+ "epoch": 56.894568690095845,
+ "grad_norm": 0.048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0229,
+ "step": 17808
+ },
+ {
+ "epoch": 56.89776357827476,
+ "grad_norm": 0.05419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0218,
+ "step": 17809
+ },
+ {
+ "epoch": 56.90095846645367,
+ "grad_norm": 0.052490234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0306,
+ "step": 17810
+ },
+ {
+ "epoch": 56.90415335463259,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0184,
+ "step": 17811
+ },
+ {
+ "epoch": 56.9073482428115,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.016,
+ "step": 17812
+ },
+ {
+ "epoch": 56.91054313099041,
+ "grad_norm": 0.058837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0364,
+ "step": 17813
+ },
+ {
+ "epoch": 56.91373801916933,
+ "grad_norm": 0.058837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0228,
+ "step": 17814
+ },
+ {
+ "epoch": 56.91693290734824,
+ "grad_norm": 0.05078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0216,
+ "step": 17815
+ },
+ {
+ "epoch": 56.92012779552716,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0262,
+ "step": 17816
+ },
+ {
+ "epoch": 56.92332268370607,
+ "grad_norm": 0.048095703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0147,
+ "step": 17817
+ },
+ {
+ "epoch": 56.926517571884986,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0224,
+ "step": 17818
+ },
+ {
+ "epoch": 56.9297124600639,
+ "grad_norm": 0.048095703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0264,
+ "step": 17819
+ },
+ {
+ "epoch": 56.932907348242814,
+ "grad_norm": 0.046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0215,
+ "step": 17820
+ },
+ {
+ "epoch": 56.936102236421725,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0182,
+ "step": 17821
+ },
+ {
+ "epoch": 56.93929712460064,
+ "grad_norm": 0.04736328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0237,
+ "step": 17822
+ },
+ {
+ "epoch": 56.94249201277955,
+ "grad_norm": 0.0537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0307,
+ "step": 17823
+ },
+ {
+ "epoch": 56.945686900958464,
+ "grad_norm": 0.047607421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0242,
+ "step": 17824
+ },
+ {
+ "epoch": 56.94888178913738,
+ "grad_norm": 0.045654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0208,
+ "step": 17825
+ },
+ {
+ "epoch": 56.95207667731629,
+ "grad_norm": 0.045166015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0235,
+ "step": 17826
+ },
+ {
+ "epoch": 56.95527156549521,
+ "grad_norm": 0.050537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0244,
+ "step": 17827
+ },
+ {
+ "epoch": 56.95846645367412,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0262,
+ "step": 17828
+ },
+ {
+ "epoch": 56.96166134185304,
+ "grad_norm": 0.04248046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0319,
+ "step": 17829
+ },
+ {
+ "epoch": 56.96485623003195,
+ "grad_norm": 0.047607421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0289,
+ "step": 17830
+ },
+ {
+ "epoch": 56.968051118210866,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0197,
+ "step": 17831
+ },
+ {
+ "epoch": 56.97124600638978,
+ "grad_norm": 0.04833984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0321,
+ "step": 17832
+ },
+ {
+ "epoch": 56.97444089456869,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0232,
+ "step": 17833
+ },
+ {
+ "epoch": 56.977635782747605,
+ "grad_norm": 0.05712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0358,
+ "step": 17834
+ },
+ {
+ "epoch": 56.980830670926515,
+ "grad_norm": 0.0517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0241,
+ "step": 17835
+ },
+ {
+ "epoch": 56.98402555910543,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0244,
+ "step": 17836
+ },
+ {
+ "epoch": 56.98722044728434,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0265,
+ "step": 17837
+ },
+ {
+ "epoch": 56.99041533546326,
+ "grad_norm": 0.08642578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0252,
+ "step": 17838
+ },
+ {
+ "epoch": 56.99361022364217,
+ "grad_norm": 0.0869140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0337,
+ "step": 17839
+ },
+ {
+ "epoch": 56.99680511182109,
+ "grad_norm": 0.0595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0312,
+ "step": 17840
+ },
+ {
+ "epoch": 57.0,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0216,
+ "step": 17841
+ },
+ {
+ "epoch": 57.00319488817891,
+ "grad_norm": 0.053466796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0373,
+ "step": 17842
+ },
+ {
+ "epoch": 57.00638977635783,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0253,
+ "step": 17843
+ },
+ {
+ "epoch": 57.00958466453674,
+ "grad_norm": 0.049560546875,
+ "learning_rate": 0.0005,
+ "loss": 1.028,
+ "step": 17844
+ },
+ {
+ "epoch": 57.01277955271566,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.028,
+ "step": 17845
+ },
+ {
+ "epoch": 57.01597444089457,
+ "grad_norm": 0.05859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0238,
+ "step": 17846
+ },
+ {
+ "epoch": 57.019169329073485,
+ "grad_norm": 0.05224609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0225,
+ "step": 17847
+ },
+ {
+ "epoch": 57.022364217252395,
+ "grad_norm": 0.11669921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0137,
+ "step": 17848
+ },
+ {
+ "epoch": 57.02555910543131,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0279,
+ "step": 17849
+ },
+ {
+ "epoch": 57.02875399361022,
+ "grad_norm": 0.091796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0219,
+ "step": 17850
+ },
+ {
+ "epoch": 57.031948881789134,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0328,
+ "step": 17851
+ },
+ {
+ "epoch": 57.03514376996805,
+ "grad_norm": 0.06201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0264,
+ "step": 17852
+ },
+ {
+ "epoch": 57.03833865814696,
+ "grad_norm": 0.0888671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0314,
+ "step": 17853
+ },
+ {
+ "epoch": 57.04153354632588,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0213,
+ "step": 17854
+ },
+ {
+ "epoch": 57.04472843450479,
+ "grad_norm": 0.08154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.013,
+ "step": 17855
+ },
+ {
+ "epoch": 57.04792332268371,
+ "grad_norm": 0.050537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0373,
+ "step": 17856
+ },
+ {
+ "epoch": 57.05111821086262,
+ "grad_norm": 0.0947265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0222,
+ "step": 17857
+ },
+ {
+ "epoch": 57.054313099041536,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0325,
+ "step": 17858
+ },
+ {
+ "epoch": 57.05750798722045,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0297,
+ "step": 17859
+ },
+ {
+ "epoch": 57.06070287539936,
+ "grad_norm": 0.1103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0275,
+ "step": 17860
+ },
+ {
+ "epoch": 57.063897763578275,
+ "grad_norm": 0.11083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0288,
+ "step": 17861
+ },
+ {
+ "epoch": 57.067092651757186,
+ "grad_norm": 0.08154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0267,
+ "step": 17862
+ },
+ {
+ "epoch": 57.0702875399361,
+ "grad_norm": 0.0859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0154,
+ "step": 17863
+ },
+ {
+ "epoch": 57.073482428115014,
+ "grad_norm": 0.059814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0229,
+ "step": 17864
+ },
+ {
+ "epoch": 57.07667731629393,
+ "grad_norm": 0.0595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0299,
+ "step": 17865
+ },
+ {
+ "epoch": 57.07987220447284,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0307,
+ "step": 17866
+ },
+ {
+ "epoch": 57.08306709265176,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0275,
+ "step": 17867
+ },
+ {
+ "epoch": 57.08626198083067,
+ "grad_norm": 0.0517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0218,
+ "step": 17868
+ },
+ {
+ "epoch": 57.08945686900959,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0264,
+ "step": 17869
+ },
+ {
+ "epoch": 57.0926517571885,
+ "grad_norm": 0.0537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0163,
+ "step": 17870
+ },
+ {
+ "epoch": 57.09584664536741,
+ "grad_norm": 0.061767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0213,
+ "step": 17871
+ },
+ {
+ "epoch": 57.09904153354633,
+ "grad_norm": 0.057373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.009,
+ "step": 17872
+ },
+ {
+ "epoch": 57.10223642172524,
+ "grad_norm": 0.083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0101,
+ "step": 17873
+ },
+ {
+ "epoch": 57.105431309904155,
+ "grad_norm": 0.04638671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0221,
+ "step": 17874
+ },
+ {
+ "epoch": 57.108626198083066,
+ "grad_norm": 0.095703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0132,
+ "step": 17875
+ },
+ {
+ "epoch": 57.11182108626198,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0239,
+ "step": 17876
+ },
+ {
+ "epoch": 57.115015974440894,
+ "grad_norm": 0.10888671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0159,
+ "step": 17877
+ },
+ {
+ "epoch": 57.11821086261981,
+ "grad_norm": 0.1044921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0327,
+ "step": 17878
+ },
+ {
+ "epoch": 57.12140575079872,
+ "grad_norm": 0.126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0201,
+ "step": 17879
+ },
+ {
+ "epoch": 57.12460063897763,
+ "grad_norm": 0.126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0298,
+ "step": 17880
+ },
+ {
+ "epoch": 57.12779552715655,
+ "grad_norm": 0.1533203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0199,
+ "step": 17881
+ },
+ {
+ "epoch": 57.13099041533546,
+ "grad_norm": 0.11767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0261,
+ "step": 17882
+ },
+ {
+ "epoch": 57.13418530351438,
+ "grad_norm": 0.1337890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0276,
+ "step": 17883
+ },
+ {
+ "epoch": 57.13738019169329,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0317,
+ "step": 17884
+ },
+ {
+ "epoch": 57.14057507987221,
+ "grad_norm": 0.1025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.041,
+ "step": 17885
+ },
+ {
+ "epoch": 57.14376996805112,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0312,
+ "step": 17886
+ },
+ {
+ "epoch": 57.146964856230035,
+ "grad_norm": 0.11328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0154,
+ "step": 17887
+ },
+ {
+ "epoch": 57.150159744408946,
+ "grad_norm": 0.08251953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0228,
+ "step": 17888
+ },
+ {
+ "epoch": 57.153354632587856,
+ "grad_norm": 0.1220703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0182,
+ "step": 17889
+ },
+ {
+ "epoch": 57.156549520766774,
+ "grad_norm": 0.0908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0228,
+ "step": 17890
+ },
+ {
+ "epoch": 57.159744408945684,
+ "grad_norm": 0.08740234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0259,
+ "step": 17891
+ },
+ {
+ "epoch": 57.1629392971246,
+ "grad_norm": 0.091796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0211,
+ "step": 17892
+ },
+ {
+ "epoch": 57.16613418530351,
+ "grad_norm": 0.05517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0418,
+ "step": 17893
+ },
+ {
+ "epoch": 57.16932907348243,
+ "grad_norm": 0.0615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0314,
+ "step": 17894
+ },
+ {
+ "epoch": 57.17252396166134,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0223,
+ "step": 17895
+ },
+ {
+ "epoch": 57.17571884984026,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.025,
+ "step": 17896
+ },
+ {
+ "epoch": 57.17891373801917,
+ "grad_norm": 0.09375,
+ "learning_rate": 0.0005,
+ "loss": 1.0228,
+ "step": 17897
+ },
+ {
+ "epoch": 57.18210862619808,
+ "grad_norm": 0.05419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0187,
+ "step": 17898
+ },
+ {
+ "epoch": 57.185303514377,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0297,
+ "step": 17899
+ },
+ {
+ "epoch": 57.18849840255591,
+ "grad_norm": 0.055419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0194,
+ "step": 17900
+ },
+ {
+ "epoch": 57.191693290734825,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0252,
+ "step": 17901
+ },
+ {
+ "epoch": 57.194888178913736,
+ "grad_norm": 0.056396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0249,
+ "step": 17902
+ },
+ {
+ "epoch": 57.198083067092654,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0251,
+ "step": 17903
+ },
+ {
+ "epoch": 57.201277955271564,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0231,
+ "step": 17904
+ },
+ {
+ "epoch": 57.20447284345048,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0279,
+ "step": 17905
+ },
+ {
+ "epoch": 57.20766773162939,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0208,
+ "step": 17906
+ },
+ {
+ "epoch": 57.21086261980831,
+ "grad_norm": 0.052490234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0204,
+ "step": 17907
+ },
+ {
+ "epoch": 57.21405750798722,
+ "grad_norm": 0.050537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0307,
+ "step": 17908
+ },
+ {
+ "epoch": 57.21725239616613,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0389,
+ "step": 17909
+ },
+ {
+ "epoch": 57.22044728434505,
+ "grad_norm": 0.046630859375,
+ "learning_rate": 0.0005,
+ "loss": 1.025,
+ "step": 17910
+ },
+ {
+ "epoch": 57.22364217252396,
+ "grad_norm": 0.052978515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0193,
+ "step": 17911
+ },
+ {
+ "epoch": 57.22683706070288,
+ "grad_norm": 0.06201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0267,
+ "step": 17912
+ },
+ {
+ "epoch": 57.23003194888179,
+ "grad_norm": 0.056396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0303,
+ "step": 17913
+ },
+ {
+ "epoch": 57.233226837060705,
+ "grad_norm": 0.0576171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0277,
+ "step": 17914
+ },
+ {
+ "epoch": 57.236421725239616,
+ "grad_norm": 0.0478515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0197,
+ "step": 17915
+ },
+ {
+ "epoch": 57.239616613418534,
+ "grad_norm": 0.05322265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0301,
+ "step": 17916
+ },
+ {
+ "epoch": 57.242811501597444,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0313,
+ "step": 17917
+ },
+ {
+ "epoch": 57.246006389776355,
+ "grad_norm": 0.051513671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0257,
+ "step": 17918
+ },
+ {
+ "epoch": 57.24920127795527,
+ "grad_norm": 0.04833984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0281,
+ "step": 17919
+ },
+ {
+ "epoch": 57.25239616613418,
+ "grad_norm": 0.04931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0216,
+ "step": 17920
+ },
+ {
+ "epoch": 57.2555910543131,
+ "grad_norm": 0.0458984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0283,
+ "step": 17921
+ },
+ {
+ "epoch": 57.25878594249201,
+ "grad_norm": 0.095703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0143,
+ "step": 17922
+ },
+ {
+ "epoch": 57.26198083067093,
+ "grad_norm": 0.050537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0326,
+ "step": 17923
+ },
+ {
+ "epoch": 57.26517571884984,
+ "grad_norm": 0.047607421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0175,
+ "step": 17924
+ },
+ {
+ "epoch": 57.26837060702876,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0384,
+ "step": 17925
+ },
+ {
+ "epoch": 57.27156549520767,
+ "grad_norm": 0.04541015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0127,
+ "step": 17926
+ },
+ {
+ "epoch": 57.27476038338658,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0473,
+ "step": 17927
+ },
+ {
+ "epoch": 57.277955271565496,
+ "grad_norm": 0.052490234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0233,
+ "step": 17928
+ },
+ {
+ "epoch": 57.281150159744406,
+ "grad_norm": 0.07861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0196,
+ "step": 17929
+ },
+ {
+ "epoch": 57.284345047923324,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0239,
+ "step": 17930
+ },
+ {
+ "epoch": 57.287539936102235,
+ "grad_norm": 0.05419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0204,
+ "step": 17931
+ },
+ {
+ "epoch": 57.29073482428115,
+ "grad_norm": 0.05224609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0142,
+ "step": 17932
+ },
+ {
+ "epoch": 57.29392971246006,
+ "grad_norm": 0.09228515625,
+ "learning_rate": 0.0005,
+ "loss": 1.033,
+ "step": 17933
+ },
+ {
+ "epoch": 57.29712460063898,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0252,
+ "step": 17934
+ },
+ {
+ "epoch": 57.30031948881789,
+ "grad_norm": 0.04931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0242,
+ "step": 17935
+ },
+ {
+ "epoch": 57.3035143769968,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0333,
+ "step": 17936
+ },
+ {
+ "epoch": 57.30670926517572,
+ "grad_norm": 0.048095703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0243,
+ "step": 17937
+ },
+ {
+ "epoch": 57.30990415335463,
+ "grad_norm": 0.048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0198,
+ "step": 17938
+ },
+ {
+ "epoch": 57.31309904153355,
+ "grad_norm": 0.048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.019,
+ "step": 17939
+ },
+ {
+ "epoch": 57.31629392971246,
+ "grad_norm": 0.045654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0263,
+ "step": 17940
+ },
+ {
+ "epoch": 57.319488817891376,
+ "grad_norm": 0.0478515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0288,
+ "step": 17941
+ },
+ {
+ "epoch": 57.322683706070286,
+ "grad_norm": 0.0498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0264,
+ "step": 17942
+ },
+ {
+ "epoch": 57.325878594249204,
+ "grad_norm": 0.05126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0252,
+ "step": 17943
+ },
+ {
+ "epoch": 57.329073482428115,
+ "grad_norm": 0.04541015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0257,
+ "step": 17944
+ },
+ {
+ "epoch": 57.33226837060703,
+ "grad_norm": 0.07861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0377,
+ "step": 17945
+ },
+ {
+ "epoch": 57.33546325878594,
+ "grad_norm": 0.052978515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0293,
+ "step": 17946
+ },
+ {
+ "epoch": 57.33865814696485,
+ "grad_norm": 0.047119140625,
+ "learning_rate": 0.0005,
+ "loss": 1.035,
+ "step": 17947
+ },
+ {
+ "epoch": 57.34185303514377,
+ "grad_norm": 0.09033203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0254,
+ "step": 17948
+ },
+ {
+ "epoch": 57.34504792332268,
+ "grad_norm": 0.046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0253,
+ "step": 17949
+ },
+ {
+ "epoch": 57.3482428115016,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0253,
+ "step": 17950
+ },
+ {
+ "epoch": 57.35143769968051,
+ "grad_norm": 0.05810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0207,
+ "step": 17951
+ },
+ {
+ "epoch": 57.35463258785943,
+ "grad_norm": 0.057373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.024,
+ "step": 17952
+ },
+ {
+ "epoch": 57.35782747603834,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0205,
+ "step": 17953
+ },
+ {
+ "epoch": 57.361022364217256,
+ "grad_norm": 0.06201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0249,
+ "step": 17954
+ },
+ {
+ "epoch": 57.364217252396166,
+ "grad_norm": 0.052734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0262,
+ "step": 17955
+ },
+ {
+ "epoch": 57.36741214057508,
+ "grad_norm": 0.0849609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0212,
+ "step": 17956
+ },
+ {
+ "epoch": 57.370607028753994,
+ "grad_norm": 0.050048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0251,
+ "step": 17957
+ },
+ {
+ "epoch": 57.373801916932905,
+ "grad_norm": 0.056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0309,
+ "step": 17958
+ },
+ {
+ "epoch": 57.37699680511182,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0277,
+ "step": 17959
+ },
+ {
+ "epoch": 57.38019169329073,
+ "grad_norm": 0.054931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0162,
+ "step": 17960
+ },
+ {
+ "epoch": 57.38338658146965,
+ "grad_norm": 0.0556640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0225,
+ "step": 17961
+ },
+ {
+ "epoch": 57.38658146964856,
+ "grad_norm": 0.04638671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0317,
+ "step": 17962
+ },
+ {
+ "epoch": 57.38977635782748,
+ "grad_norm": 0.050048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.016,
+ "step": 17963
+ },
+ {
+ "epoch": 57.39297124600639,
+ "grad_norm": 0.05224609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0236,
+ "step": 17964
+ },
+ {
+ "epoch": 57.3961661341853,
+ "grad_norm": 0.0537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0163,
+ "step": 17965
+ },
+ {
+ "epoch": 57.39936102236422,
+ "grad_norm": 0.047119140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0118,
+ "step": 17966
+ },
+ {
+ "epoch": 57.40255591054313,
+ "grad_norm": 0.06005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0272,
+ "step": 17967
+ },
+ {
+ "epoch": 57.405750798722046,
+ "grad_norm": 0.0517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0223,
+ "step": 17968
+ },
+ {
+ "epoch": 57.40894568690096,
+ "grad_norm": 0.055419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0302,
+ "step": 17969
+ },
+ {
+ "epoch": 57.412140575079874,
+ "grad_norm": 0.052490234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0225,
+ "step": 17970
+ },
+ {
+ "epoch": 57.415335463258785,
+ "grad_norm": 0.05078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0281,
+ "step": 17971
+ },
+ {
+ "epoch": 57.4185303514377,
+ "grad_norm": 0.06201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0281,
+ "step": 17972
+ },
+ {
+ "epoch": 57.42172523961661,
+ "grad_norm": 0.04736328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0124,
+ "step": 17973
+ },
+ {
+ "epoch": 57.424920127795524,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0161,
+ "step": 17974
+ },
+ {
+ "epoch": 57.42811501597444,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0253,
+ "step": 17975
+ },
+ {
+ "epoch": 57.43130990415335,
+ "grad_norm": 0.0517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.022,
+ "step": 17976
+ },
+ {
+ "epoch": 57.43450479233227,
+ "grad_norm": 0.0537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0324,
+ "step": 17977
+ },
+ {
+ "epoch": 57.43769968051118,
+ "grad_norm": 0.0546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0296,
+ "step": 17978
+ },
+ {
+ "epoch": 57.4408945686901,
+ "grad_norm": 0.055908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.017,
+ "step": 17979
+ },
+ {
+ "epoch": 57.44408945686901,
+ "grad_norm": 0.06201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0125,
+ "step": 17980
+ },
+ {
+ "epoch": 57.447284345047926,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.034,
+ "step": 17981
+ },
+ {
+ "epoch": 57.45047923322684,
+ "grad_norm": 0.07861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.022,
+ "step": 17982
+ },
+ {
+ "epoch": 57.453674121405754,
+ "grad_norm": 0.048583984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0188,
+ "step": 17983
+ },
+ {
+ "epoch": 57.456869009584665,
+ "grad_norm": 0.058837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0224,
+ "step": 17984
+ },
+ {
+ "epoch": 57.460063897763575,
+ "grad_norm": 0.049072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0254,
+ "step": 17985
+ },
+ {
+ "epoch": 57.46325878594249,
+ "grad_norm": 0.04833984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0251,
+ "step": 17986
+ },
+ {
+ "epoch": 57.466453674121404,
+ "grad_norm": 0.05126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0176,
+ "step": 17987
+ },
+ {
+ "epoch": 57.46964856230032,
+ "grad_norm": 0.057861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0337,
+ "step": 17988
+ },
+ {
+ "epoch": 57.47284345047923,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0226,
+ "step": 17989
+ },
+ {
+ "epoch": 57.47603833865815,
+ "grad_norm": 0.048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0277,
+ "step": 17990
+ },
+ {
+ "epoch": 57.47923322683706,
+ "grad_norm": 0.04931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0302,
+ "step": 17991
+ },
+ {
+ "epoch": 57.48242811501598,
+ "grad_norm": 0.062255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0341,
+ "step": 17992
+ },
+ {
+ "epoch": 57.48562300319489,
+ "grad_norm": 0.05810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0378,
+ "step": 17993
+ },
+ {
+ "epoch": 57.4888178913738,
+ "grad_norm": 0.058349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0297,
+ "step": 17994
+ },
+ {
+ "epoch": 57.49201277955272,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0231,
+ "step": 17995
+ },
+ {
+ "epoch": 57.49520766773163,
+ "grad_norm": 0.054931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0277,
+ "step": 17996
+ },
+ {
+ "epoch": 57.498402555910545,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.034,
+ "step": 17997
+ },
+ {
+ "epoch": 57.501597444089455,
+ "grad_norm": 0.056884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0219,
+ "step": 17998
+ },
+ {
+ "epoch": 57.50479233226837,
+ "grad_norm": 0.08544921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0196,
+ "step": 17999
+ },
+ {
+ "epoch": 57.50798722044728,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0164,
+ "step": 18000
+ },
+ {
+ "epoch": 57.5111821086262,
+ "grad_norm": 0.0576171875,
+ "learning_rate": 0.0005,
+ "loss": 1.022,
+ "step": 18001
+ },
+ {
+ "epoch": 57.51437699680511,
+ "grad_norm": 0.1083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0213,
+ "step": 18002
+ },
+ {
+ "epoch": 57.51757188498402,
+ "grad_norm": 0.046142578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0193,
+ "step": 18003
+ },
+ {
+ "epoch": 57.52076677316294,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0118,
+ "step": 18004
+ },
+ {
+ "epoch": 57.52396166134185,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0272,
+ "step": 18005
+ },
+ {
+ "epoch": 57.52715654952077,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0367,
+ "step": 18006
+ },
+ {
+ "epoch": 57.53035143769968,
+ "grad_norm": 0.09912109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0165,
+ "step": 18007
+ },
+ {
+ "epoch": 57.533546325878596,
+ "grad_norm": 0.056884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0189,
+ "step": 18008
+ },
+ {
+ "epoch": 57.53674121405751,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0225,
+ "step": 18009
+ },
+ {
+ "epoch": 57.539936102236425,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.013,
+ "step": 18010
+ },
+ {
+ "epoch": 57.543130990415335,
+ "grad_norm": 0.1162109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0205,
+ "step": 18011
+ },
+ {
+ "epoch": 57.546325878594246,
+ "grad_norm": 0.0595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0248,
+ "step": 18012
+ },
+ {
+ "epoch": 57.54952076677316,
+ "grad_norm": 0.09130859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0251,
+ "step": 18013
+ },
+ {
+ "epoch": 57.552715654952074,
+ "grad_norm": 0.09228515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0236,
+ "step": 18014
+ },
+ {
+ "epoch": 57.55591054313099,
+ "grad_norm": 0.08984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0295,
+ "step": 18015
+ },
+ {
+ "epoch": 57.5591054313099,
+ "grad_norm": 0.0869140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0363,
+ "step": 18016
+ },
+ {
+ "epoch": 57.56230031948882,
+ "grad_norm": 0.049560546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0339,
+ "step": 18017
+ },
+ {
+ "epoch": 57.56549520766773,
+ "grad_norm": 0.08984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0298,
+ "step": 18018
+ },
+ {
+ "epoch": 57.56869009584665,
+ "grad_norm": 0.056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0231,
+ "step": 18019
+ },
+ {
+ "epoch": 57.57188498402556,
+ "grad_norm": 0.103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0244,
+ "step": 18020
+ },
+ {
+ "epoch": 57.575079872204476,
+ "grad_norm": 0.052978515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0284,
+ "step": 18021
+ },
+ {
+ "epoch": 57.57827476038339,
+ "grad_norm": 0.083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0399,
+ "step": 18022
+ },
+ {
+ "epoch": 57.5814696485623,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0282,
+ "step": 18023
+ },
+ {
+ "epoch": 57.584664536741215,
+ "grad_norm": 0.057373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0184,
+ "step": 18024
+ },
+ {
+ "epoch": 57.587859424920126,
+ "grad_norm": 0.083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.025,
+ "step": 18025
+ },
+ {
+ "epoch": 57.59105431309904,
+ "grad_norm": 0.052001953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0369,
+ "step": 18026
+ },
+ {
+ "epoch": 57.594249201277954,
+ "grad_norm": 0.099609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0289,
+ "step": 18027
+ },
+ {
+ "epoch": 57.59744408945687,
+ "grad_norm": 0.059326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0171,
+ "step": 18028
+ },
+ {
+ "epoch": 57.60063897763578,
+ "grad_norm": 0.0859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0463,
+ "step": 18029
+ },
+ {
+ "epoch": 57.6038338658147,
+ "grad_norm": 0.057373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0233,
+ "step": 18030
+ },
+ {
+ "epoch": 57.60702875399361,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0219,
+ "step": 18031
+ },
+ {
+ "epoch": 57.61022364217252,
+ "grad_norm": 0.09423828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0278,
+ "step": 18032
+ },
+ {
+ "epoch": 57.61341853035144,
+ "grad_norm": 0.050537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0271,
+ "step": 18033
+ },
+ {
+ "epoch": 57.61661341853035,
+ "grad_norm": 0.0849609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0287,
+ "step": 18034
+ },
+ {
+ "epoch": 57.61980830670927,
+ "grad_norm": 0.044921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0258,
+ "step": 18035
+ },
+ {
+ "epoch": 57.62300319488818,
+ "grad_norm": 0.08935546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0315,
+ "step": 18036
+ },
+ {
+ "epoch": 57.626198083067095,
+ "grad_norm": 0.0966796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0286,
+ "step": 18037
+ },
+ {
+ "epoch": 57.629392971246006,
+ "grad_norm": 0.044921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0369,
+ "step": 18038
+ },
+ {
+ "epoch": 57.63258785942492,
+ "grad_norm": 0.0947265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0271,
+ "step": 18039
+ },
+ {
+ "epoch": 57.635782747603834,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0311,
+ "step": 18040
+ },
+ {
+ "epoch": 57.638977635782744,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0163,
+ "step": 18041
+ },
+ {
+ "epoch": 57.64217252396166,
+ "grad_norm": 0.1328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0259,
+ "step": 18042
+ },
+ {
+ "epoch": 57.64536741214057,
+ "grad_norm": 0.087890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0189,
+ "step": 18043
+ },
+ {
+ "epoch": 57.64856230031949,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0343,
+ "step": 18044
+ },
+ {
+ "epoch": 57.6517571884984,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0243,
+ "step": 18045
+ },
+ {
+ "epoch": 57.65495207667732,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0326,
+ "step": 18046
+ },
+ {
+ "epoch": 57.65814696485623,
+ "grad_norm": 0.08740234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0341,
+ "step": 18047
+ },
+ {
+ "epoch": 57.66134185303515,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0393,
+ "step": 18048
+ },
+ {
+ "epoch": 57.66453674121406,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0251,
+ "step": 18049
+ },
+ {
+ "epoch": 57.66773162939297,
+ "grad_norm": 0.09423828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0274,
+ "step": 18050
+ },
+ {
+ "epoch": 57.670926517571885,
+ "grad_norm": 0.04931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0315,
+ "step": 18051
+ },
+ {
+ "epoch": 57.674121405750796,
+ "grad_norm": 0.09912109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0316,
+ "step": 18052
+ },
+ {
+ "epoch": 57.677316293929714,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0181,
+ "step": 18053
+ },
+ {
+ "epoch": 57.680511182108624,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0296,
+ "step": 18054
+ },
+ {
+ "epoch": 57.68370607028754,
+ "grad_norm": 0.0908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0117,
+ "step": 18055
+ },
+ {
+ "epoch": 57.68690095846645,
+ "grad_norm": 0.052734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0315,
+ "step": 18056
+ },
+ {
+ "epoch": 57.69009584664537,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.025,
+ "step": 18057
+ },
+ {
+ "epoch": 57.69329073482428,
+ "grad_norm": 0.0498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.032,
+ "step": 18058
+ },
+ {
+ "epoch": 57.6964856230032,
+ "grad_norm": 0.054443359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0294,
+ "step": 18059
+ },
+ {
+ "epoch": 57.69968051118211,
+ "grad_norm": 0.057861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0349,
+ "step": 18060
+ },
+ {
+ "epoch": 57.70287539936102,
+ "grad_norm": 0.0498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0268,
+ "step": 18061
+ },
+ {
+ "epoch": 57.70607028753994,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.031,
+ "step": 18062
+ },
+ {
+ "epoch": 57.70926517571885,
+ "grad_norm": 0.052734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0234,
+ "step": 18063
+ },
+ {
+ "epoch": 57.712460063897765,
+ "grad_norm": 0.05126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0246,
+ "step": 18064
+ },
+ {
+ "epoch": 57.715654952076676,
+ "grad_norm": 0.04931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.025,
+ "step": 18065
+ },
+ {
+ "epoch": 57.718849840255594,
+ "grad_norm": 0.0439453125,
+ "learning_rate": 0.0005,
+ "loss": 1.026,
+ "step": 18066
+ },
+ {
+ "epoch": 57.722044728434504,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0262,
+ "step": 18067
+ },
+ {
+ "epoch": 57.72523961661342,
+ "grad_norm": 0.04833984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0209,
+ "step": 18068
+ },
+ {
+ "epoch": 57.72843450479233,
+ "grad_norm": 0.08544921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0187,
+ "step": 18069
+ },
+ {
+ "epoch": 57.73162939297124,
+ "grad_norm": 0.061767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0264,
+ "step": 18070
+ },
+ {
+ "epoch": 57.73482428115016,
+ "grad_norm": 0.04443359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0265,
+ "step": 18071
+ },
+ {
+ "epoch": 57.73801916932907,
+ "grad_norm": 0.08935546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0257,
+ "step": 18072
+ },
+ {
+ "epoch": 57.74121405750799,
+ "grad_norm": 0.04736328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0267,
+ "step": 18073
+ },
+ {
+ "epoch": 57.7444089456869,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0184,
+ "step": 18074
+ },
+ {
+ "epoch": 57.74760383386582,
+ "grad_norm": 0.04638671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0206,
+ "step": 18075
+ },
+ {
+ "epoch": 57.75079872204473,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0181,
+ "step": 18076
+ },
+ {
+ "epoch": 57.753993610223645,
+ "grad_norm": 0.050048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.03,
+ "step": 18077
+ },
+ {
+ "epoch": 57.757188498402556,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0278,
+ "step": 18078
+ },
+ {
+ "epoch": 57.760383386581466,
+ "grad_norm": 0.05029296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0246,
+ "step": 18079
+ },
+ {
+ "epoch": 57.763578274760384,
+ "grad_norm": 0.04736328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0257,
+ "step": 18080
+ },
+ {
+ "epoch": 57.766773162939295,
+ "grad_norm": 0.04736328125,
+ "learning_rate": 0.0005,
+ "loss": 1.031,
+ "step": 18081
+ },
+ {
+ "epoch": 57.76996805111821,
+ "grad_norm": 0.052001953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0141,
+ "step": 18082
+ },
+ {
+ "epoch": 57.77316293929712,
+ "grad_norm": 0.05615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0225,
+ "step": 18083
+ },
+ {
+ "epoch": 57.77635782747604,
+ "grad_norm": 0.058837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0267,
+ "step": 18084
+ },
+ {
+ "epoch": 57.77955271565495,
+ "grad_norm": 0.0478515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0193,
+ "step": 18085
+ },
+ {
+ "epoch": 57.78274760383387,
+ "grad_norm": 0.05419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0183,
+ "step": 18086
+ },
+ {
+ "epoch": 57.78594249201278,
+ "grad_norm": 0.046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0249,
+ "step": 18087
+ },
+ {
+ "epoch": 57.78913738019169,
+ "grad_norm": 0.050537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0249,
+ "step": 18088
+ },
+ {
+ "epoch": 57.79233226837061,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0169,
+ "step": 18089
+ },
+ {
+ "epoch": 57.79552715654952,
+ "grad_norm": 0.046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0279,
+ "step": 18090
+ },
+ {
+ "epoch": 57.798722044728436,
+ "grad_norm": 0.054931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.025,
+ "step": 18091
+ },
+ {
+ "epoch": 57.801916932907346,
+ "grad_norm": 0.061767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0335,
+ "step": 18092
+ },
+ {
+ "epoch": 57.805111821086264,
+ "grad_norm": 0.044921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0411,
+ "step": 18093
+ },
+ {
+ "epoch": 57.808306709265175,
+ "grad_norm": 0.0478515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0236,
+ "step": 18094
+ },
+ {
+ "epoch": 57.81150159744409,
+ "grad_norm": 0.0517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.026,
+ "step": 18095
+ },
+ {
+ "epoch": 57.814696485623,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0234,
+ "step": 18096
+ },
+ {
+ "epoch": 57.81789137380191,
+ "grad_norm": 0.0556640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0338,
+ "step": 18097
+ },
+ {
+ "epoch": 57.82108626198083,
+ "grad_norm": 0.0517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0199,
+ "step": 18098
+ },
+ {
+ "epoch": 57.82428115015974,
+ "grad_norm": 0.051025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0308,
+ "step": 18099
+ },
+ {
+ "epoch": 57.82747603833866,
+ "grad_norm": 0.05712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.014,
+ "step": 18100
+ },
+ {
+ "epoch": 57.83067092651757,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0244,
+ "step": 18101
+ },
+ {
+ "epoch": 57.83386581469649,
+ "grad_norm": 0.048583984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0271,
+ "step": 18102
+ },
+ {
+ "epoch": 57.8370607028754,
+ "grad_norm": 0.047607421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0358,
+ "step": 18103
+ },
+ {
+ "epoch": 57.840255591054316,
+ "grad_norm": 0.060791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0268,
+ "step": 18104
+ },
+ {
+ "epoch": 57.843450479233226,
+ "grad_norm": 0.08447265625,
+ "learning_rate": 0.0005,
+ "loss": 1.03,
+ "step": 18105
+ },
+ {
+ "epoch": 57.846645367412144,
+ "grad_norm": 0.059814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0221,
+ "step": 18106
+ },
+ {
+ "epoch": 57.849840255591054,
+ "grad_norm": 0.0810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0273,
+ "step": 18107
+ },
+ {
+ "epoch": 57.853035143769965,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0237,
+ "step": 18108
+ },
+ {
+ "epoch": 57.85623003194888,
+ "grad_norm": 0.055419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0287,
+ "step": 18109
+ },
+ {
+ "epoch": 57.85942492012779,
+ "grad_norm": 0.07861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0255,
+ "step": 18110
+ },
+ {
+ "epoch": 57.86261980830671,
+ "grad_norm": 0.0556640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0375,
+ "step": 18111
+ },
+ {
+ "epoch": 57.86581469648562,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0301,
+ "step": 18112
+ },
+ {
+ "epoch": 57.86900958466454,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0256,
+ "step": 18113
+ },
+ {
+ "epoch": 57.87220447284345,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0282,
+ "step": 18114
+ },
+ {
+ "epoch": 57.87539936102237,
+ "grad_norm": 0.095703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0149,
+ "step": 18115
+ },
+ {
+ "epoch": 57.87859424920128,
+ "grad_norm": 0.0859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0386,
+ "step": 18116
+ },
+ {
+ "epoch": 57.88178913738019,
+ "grad_norm": 0.11181640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0166,
+ "step": 18117
+ },
+ {
+ "epoch": 57.884984025559106,
+ "grad_norm": 0.06005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0261,
+ "step": 18118
+ },
+ {
+ "epoch": 57.88817891373802,
+ "grad_norm": 0.11962890625,
+ "learning_rate": 0.0005,
+ "loss": 1.021,
+ "step": 18119
+ },
+ {
+ "epoch": 57.891373801916934,
+ "grad_norm": 0.0849609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0269,
+ "step": 18120
+ },
+ {
+ "epoch": 57.894568690095845,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0106,
+ "step": 18121
+ },
+ {
+ "epoch": 57.89776357827476,
+ "grad_norm": 0.058349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0197,
+ "step": 18122
+ },
+ {
+ "epoch": 57.90095846645367,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0166,
+ "step": 18123
+ },
+ {
+ "epoch": 57.90415335463259,
+ "grad_norm": 0.05859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0253,
+ "step": 18124
+ },
+ {
+ "epoch": 57.9073482428115,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0326,
+ "step": 18125
+ },
+ {
+ "epoch": 57.91054313099041,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.012,
+ "step": 18126
+ },
+ {
+ "epoch": 57.91373801916933,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0299,
+ "step": 18127
+ },
+ {
+ "epoch": 57.91693290734824,
+ "grad_norm": 0.0869140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0106,
+ "step": 18128
+ },
+ {
+ "epoch": 57.92012779552716,
+ "grad_norm": 0.0439453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0282,
+ "step": 18129
+ },
+ {
+ "epoch": 57.92332268370607,
+ "grad_norm": 0.05615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.035,
+ "step": 18130
+ },
+ {
+ "epoch": 57.926517571884986,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0294,
+ "step": 18131
+ },
+ {
+ "epoch": 57.9297124600639,
+ "grad_norm": 0.046630859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0306,
+ "step": 18132
+ },
+ {
+ "epoch": 57.932907348242814,
+ "grad_norm": 0.05322265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0271,
+ "step": 18133
+ },
+ {
+ "epoch": 57.936102236421725,
+ "grad_norm": 0.0458984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0338,
+ "step": 18134
+ },
+ {
+ "epoch": 57.93929712460064,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0085,
+ "step": 18135
+ },
+ {
+ "epoch": 57.94249201277955,
+ "grad_norm": 0.048583984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0178,
+ "step": 18136
+ },
+ {
+ "epoch": 57.945686900958464,
+ "grad_norm": 0.0595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0268,
+ "step": 18137
+ },
+ {
+ "epoch": 57.94888178913738,
+ "grad_norm": 0.043212890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0275,
+ "step": 18138
+ },
+ {
+ "epoch": 57.95207667731629,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0163,
+ "step": 18139
+ },
+ {
+ "epoch": 57.95527156549521,
+ "grad_norm": 0.05224609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0224,
+ "step": 18140
+ },
+ {
+ "epoch": 57.95846645367412,
+ "grad_norm": 0.04638671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0283,
+ "step": 18141
+ },
+ {
+ "epoch": 57.96166134185304,
+ "grad_norm": 0.04931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0295,
+ "step": 18142
+ },
+ {
+ "epoch": 57.96485623003195,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0322,
+ "step": 18143
+ },
+ {
+ "epoch": 57.968051118210866,
+ "grad_norm": 0.0537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0327,
+ "step": 18144
+ },
+ {
+ "epoch": 57.97124600638978,
+ "grad_norm": 0.044189453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0107,
+ "step": 18145
+ },
+ {
+ "epoch": 57.97444089456869,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0263,
+ "step": 18146
+ },
+ {
+ "epoch": 57.977635782747605,
+ "grad_norm": 0.0546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0325,
+ "step": 18147
+ },
+ {
+ "epoch": 57.980830670926515,
+ "grad_norm": 0.0517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0362,
+ "step": 18148
+ },
+ {
+ "epoch": 57.98402555910543,
+ "grad_norm": 0.0810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0409,
+ "step": 18149
+ },
+ {
+ "epoch": 57.98722044728434,
+ "grad_norm": 0.06103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0229,
+ "step": 18150
+ },
+ {
+ "epoch": 57.99041533546326,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0338,
+ "step": 18151
+ },
+ {
+ "epoch": 57.99361022364217,
+ "grad_norm": 0.047607421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0262,
+ "step": 18152
+ },
+ {
+ "epoch": 57.99680511182109,
+ "grad_norm": 0.05078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0192,
+ "step": 18153
+ },
+ {
+ "epoch": 58.0,
+ "grad_norm": 0.05029296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0217,
+ "step": 18154
+ },
+ {
+ "epoch": 58.00319488817891,
+ "grad_norm": 0.052734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0224,
+ "step": 18155
+ },
+ {
+ "epoch": 58.00638977635783,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0202,
+ "step": 18156
+ },
+ {
+ "epoch": 58.00958466453674,
+ "grad_norm": 0.059326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0091,
+ "step": 18157
+ },
+ {
+ "epoch": 58.01277955271566,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0168,
+ "step": 18158
+ },
+ {
+ "epoch": 58.01597444089457,
+ "grad_norm": 0.095703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0248,
+ "step": 18159
+ },
+ {
+ "epoch": 58.019169329073485,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0303,
+ "step": 18160
+ },
+ {
+ "epoch": 58.022364217252395,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0239,
+ "step": 18161
+ },
+ {
+ "epoch": 58.02555910543131,
+ "grad_norm": 0.06005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0287,
+ "step": 18162
+ },
+ {
+ "epoch": 58.02875399361022,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0214,
+ "step": 18163
+ },
+ {
+ "epoch": 58.031948881789134,
+ "grad_norm": 0.08251953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0179,
+ "step": 18164
+ },
+ {
+ "epoch": 58.03514376996805,
+ "grad_norm": 0.051025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0269,
+ "step": 18165
+ },
+ {
+ "epoch": 58.03833865814696,
+ "grad_norm": 0.0537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0293,
+ "step": 18166
+ },
+ {
+ "epoch": 58.04153354632588,
+ "grad_norm": 0.052001953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0225,
+ "step": 18167
+ },
+ {
+ "epoch": 58.04472843450479,
+ "grad_norm": 0.05078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0206,
+ "step": 18168
+ },
+ {
+ "epoch": 58.04792332268371,
+ "grad_norm": 0.05322265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0239,
+ "step": 18169
+ },
+ {
+ "epoch": 58.05111821086262,
+ "grad_norm": 0.049072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0173,
+ "step": 18170
+ },
+ {
+ "epoch": 58.054313099041536,
+ "grad_norm": 0.050537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0216,
+ "step": 18171
+ },
+ {
+ "epoch": 58.05750798722045,
+ "grad_norm": 0.052001953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0231,
+ "step": 18172
+ },
+ {
+ "epoch": 58.06070287539936,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0233,
+ "step": 18173
+ },
+ {
+ "epoch": 58.063897763578275,
+ "grad_norm": 0.052734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0164,
+ "step": 18174
+ },
+ {
+ "epoch": 58.067092651757186,
+ "grad_norm": 0.05712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0224,
+ "step": 18175
+ },
+ {
+ "epoch": 58.0702875399361,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0239,
+ "step": 18176
+ },
+ {
+ "epoch": 58.073482428115014,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0209,
+ "step": 18177
+ },
+ {
+ "epoch": 58.07667731629393,
+ "grad_norm": 0.0556640625,
+ "learning_rate": 0.0005,
+ "loss": 1.033,
+ "step": 18178
+ },
+ {
+ "epoch": 58.07987220447284,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0137,
+ "step": 18179
+ },
+ {
+ "epoch": 58.08306709265176,
+ "grad_norm": 0.06201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0257,
+ "step": 18180
+ },
+ {
+ "epoch": 58.08626198083067,
+ "grad_norm": 0.053466796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0363,
+ "step": 18181
+ },
+ {
+ "epoch": 58.08945686900959,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0264,
+ "step": 18182
+ },
+ {
+ "epoch": 58.0926517571885,
+ "grad_norm": 0.061279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0274,
+ "step": 18183
+ },
+ {
+ "epoch": 58.09584664536741,
+ "grad_norm": 0.0537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0162,
+ "step": 18184
+ },
+ {
+ "epoch": 58.09904153354633,
+ "grad_norm": 0.061279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0202,
+ "step": 18185
+ },
+ {
+ "epoch": 58.10223642172524,
+ "grad_norm": 0.045654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0178,
+ "step": 18186
+ },
+ {
+ "epoch": 58.105431309904155,
+ "grad_norm": 0.05126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0224,
+ "step": 18187
+ },
+ {
+ "epoch": 58.108626198083066,
+ "grad_norm": 0.08447265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0279,
+ "step": 18188
+ },
+ {
+ "epoch": 58.11182108626198,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0119,
+ "step": 18189
+ },
+ {
+ "epoch": 58.115015974440894,
+ "grad_norm": 0.06103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0323,
+ "step": 18190
+ },
+ {
+ "epoch": 58.11821086261981,
+ "grad_norm": 0.048095703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0118,
+ "step": 18191
+ },
+ {
+ "epoch": 58.12140575079872,
+ "grad_norm": 0.061279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0327,
+ "step": 18192
+ },
+ {
+ "epoch": 58.12460063897763,
+ "grad_norm": 0.055419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0288,
+ "step": 18193
+ },
+ {
+ "epoch": 58.12779552715655,
+ "grad_norm": 0.04541015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0155,
+ "step": 18194
+ },
+ {
+ "epoch": 58.13099041533546,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0247,
+ "step": 18195
+ },
+ {
+ "epoch": 58.13418530351438,
+ "grad_norm": 0.057861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0268,
+ "step": 18196
+ },
+ {
+ "epoch": 58.13738019169329,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0196,
+ "step": 18197
+ },
+ {
+ "epoch": 58.14057507987221,
+ "grad_norm": 0.08447265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0228,
+ "step": 18198
+ },
+ {
+ "epoch": 58.14376996805112,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0191,
+ "step": 18199
+ },
+ {
+ "epoch": 58.146964856230035,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0127,
+ "step": 18200
+ },
+ {
+ "epoch": 58.150159744408946,
+ "grad_norm": 0.06005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0404,
+ "step": 18201
+ },
+ {
+ "epoch": 58.153354632587856,
+ "grad_norm": 0.053466796875,
+ "learning_rate": 0.0005,
+ "loss": 1.028,
+ "step": 18202
+ },
+ {
+ "epoch": 58.156549520766774,
+ "grad_norm": 0.059326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0202,
+ "step": 18203
+ },
+ {
+ "epoch": 58.159744408945684,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0222,
+ "step": 18204
+ },
+ {
+ "epoch": 58.1629392971246,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0213,
+ "step": 18205
+ },
+ {
+ "epoch": 58.16613418530351,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0237,
+ "step": 18206
+ },
+ {
+ "epoch": 58.16932907348243,
+ "grad_norm": 0.055419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0293,
+ "step": 18207
+ },
+ {
+ "epoch": 58.17252396166134,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0266,
+ "step": 18208
+ },
+ {
+ "epoch": 58.17571884984026,
+ "grad_norm": 0.05419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0296,
+ "step": 18209
+ },
+ {
+ "epoch": 58.17891373801917,
+ "grad_norm": 0.0869140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0257,
+ "step": 18210
+ },
+ {
+ "epoch": 58.18210862619808,
+ "grad_norm": 0.0595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0244,
+ "step": 18211
+ },
+ {
+ "epoch": 58.185303514377,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0344,
+ "step": 18212
+ },
+ {
+ "epoch": 58.18849840255591,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0232,
+ "step": 18213
+ },
+ {
+ "epoch": 58.191693290734825,
+ "grad_norm": 0.05126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0254,
+ "step": 18214
+ },
+ {
+ "epoch": 58.194888178913736,
+ "grad_norm": 0.07861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0278,
+ "step": 18215
+ },
+ {
+ "epoch": 58.198083067092654,
+ "grad_norm": 0.0546875,
+ "learning_rate": 0.0005,
+ "loss": 1.019,
+ "step": 18216
+ },
+ {
+ "epoch": 58.201277955271564,
+ "grad_norm": 0.08447265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0288,
+ "step": 18217
+ },
+ {
+ "epoch": 58.20447284345048,
+ "grad_norm": 0.05712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0218,
+ "step": 18218
+ },
+ {
+ "epoch": 58.20766773162939,
+ "grad_norm": 0.054443359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0228,
+ "step": 18219
+ },
+ {
+ "epoch": 58.21086261980831,
+ "grad_norm": 0.05810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.03,
+ "step": 18220
+ },
+ {
+ "epoch": 58.21405750798722,
+ "grad_norm": 0.060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.031,
+ "step": 18221
+ },
+ {
+ "epoch": 58.21725239616613,
+ "grad_norm": 0.060302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0312,
+ "step": 18222
+ },
+ {
+ "epoch": 58.22044728434505,
+ "grad_norm": 0.052001953125,
+ "learning_rate": 0.0005,
+ "loss": 1.032,
+ "step": 18223
+ },
+ {
+ "epoch": 58.22364217252396,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0236,
+ "step": 18224
+ },
+ {
+ "epoch": 58.22683706070288,
+ "grad_norm": 0.06103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0231,
+ "step": 18225
+ },
+ {
+ "epoch": 58.23003194888179,
+ "grad_norm": 0.04736328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0369,
+ "step": 18226
+ },
+ {
+ "epoch": 58.233226837060705,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0347,
+ "step": 18227
+ },
+ {
+ "epoch": 58.236421725239616,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0384,
+ "step": 18228
+ },
+ {
+ "epoch": 58.239616613418534,
+ "grad_norm": 0.052001953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0314,
+ "step": 18229
+ },
+ {
+ "epoch": 58.242811501597444,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0152,
+ "step": 18230
+ },
+ {
+ "epoch": 58.246006389776355,
+ "grad_norm": 0.0546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0243,
+ "step": 18231
+ },
+ {
+ "epoch": 58.24920127795527,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0213,
+ "step": 18232
+ },
+ {
+ "epoch": 58.25239616613418,
+ "grad_norm": 0.0478515625,
+ "learning_rate": 0.0005,
+ "loss": 1.031,
+ "step": 18233
+ },
+ {
+ "epoch": 58.2555910543131,
+ "grad_norm": 0.08740234375,
+ "learning_rate": 0.0005,
+ "loss": 1.032,
+ "step": 18234
+ },
+ {
+ "epoch": 58.25878594249201,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.013,
+ "step": 18235
+ },
+ {
+ "epoch": 58.26198083067093,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0218,
+ "step": 18236
+ },
+ {
+ "epoch": 58.26517571884984,
+ "grad_norm": 0.048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0345,
+ "step": 18237
+ },
+ {
+ "epoch": 58.26837060702876,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0268,
+ "step": 18238
+ },
+ {
+ "epoch": 58.27156549520767,
+ "grad_norm": 0.052001953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0236,
+ "step": 18239
+ },
+ {
+ "epoch": 58.27476038338658,
+ "grad_norm": 0.047607421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0312,
+ "step": 18240
+ },
+ {
+ "epoch": 58.277955271565496,
+ "grad_norm": 0.05517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.022,
+ "step": 18241
+ },
+ {
+ "epoch": 58.281150159744406,
+ "grad_norm": 0.052490234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0204,
+ "step": 18242
+ },
+ {
+ "epoch": 58.284345047923324,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0253,
+ "step": 18243
+ },
+ {
+ "epoch": 58.287539936102235,
+ "grad_norm": 0.04931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0245,
+ "step": 18244
+ },
+ {
+ "epoch": 58.29073482428115,
+ "grad_norm": 0.0556640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0393,
+ "step": 18245
+ },
+ {
+ "epoch": 58.29392971246006,
+ "grad_norm": 0.052001953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0266,
+ "step": 18246
+ },
+ {
+ "epoch": 58.29712460063898,
+ "grad_norm": 0.05029296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0247,
+ "step": 18247
+ },
+ {
+ "epoch": 58.30031948881789,
+ "grad_norm": 0.0849609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0305,
+ "step": 18248
+ },
+ {
+ "epoch": 58.3035143769968,
+ "grad_norm": 0.048583984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0256,
+ "step": 18249
+ },
+ {
+ "epoch": 58.30670926517572,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0279,
+ "step": 18250
+ },
+ {
+ "epoch": 58.30990415335463,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0315,
+ "step": 18251
+ },
+ {
+ "epoch": 58.31309904153355,
+ "grad_norm": 0.044189453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0301,
+ "step": 18252
+ },
+ {
+ "epoch": 58.31629392971246,
+ "grad_norm": 0.0869140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0167,
+ "step": 18253
+ },
+ {
+ "epoch": 58.319488817891376,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0466,
+ "step": 18254
+ },
+ {
+ "epoch": 58.322683706070286,
+ "grad_norm": 0.126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0257,
+ "step": 18255
+ },
+ {
+ "epoch": 58.325878594249204,
+ "grad_norm": 0.08447265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0233,
+ "step": 18256
+ },
+ {
+ "epoch": 58.329073482428115,
+ "grad_norm": 0.08544921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0233,
+ "step": 18257
+ },
+ {
+ "epoch": 58.33226837060703,
+ "grad_norm": 0.060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.028,
+ "step": 18258
+ },
+ {
+ "epoch": 58.33546325878594,
+ "grad_norm": 0.099609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0295,
+ "step": 18259
+ },
+ {
+ "epoch": 58.33865814696485,
+ "grad_norm": 0.051513671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0337,
+ "step": 18260
+ },
+ {
+ "epoch": 58.34185303514377,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0332,
+ "step": 18261
+ },
+ {
+ "epoch": 58.34504792332268,
+ "grad_norm": 0.052490234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0129,
+ "step": 18262
+ },
+ {
+ "epoch": 58.3482428115016,
+ "grad_norm": 0.056396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0201,
+ "step": 18263
+ },
+ {
+ "epoch": 58.35143769968051,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0217,
+ "step": 18264
+ },
+ {
+ "epoch": 58.35463258785943,
+ "grad_norm": 0.060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0317,
+ "step": 18265
+ },
+ {
+ "epoch": 58.35782747603834,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0258,
+ "step": 18266
+ },
+ {
+ "epoch": 58.361022364217256,
+ "grad_norm": 0.045654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0206,
+ "step": 18267
+ },
+ {
+ "epoch": 58.364217252396166,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0241,
+ "step": 18268
+ },
+ {
+ "epoch": 58.36741214057508,
+ "grad_norm": 0.05224609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0311,
+ "step": 18269
+ },
+ {
+ "epoch": 58.370607028753994,
+ "grad_norm": 0.05078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0233,
+ "step": 18270
+ },
+ {
+ "epoch": 58.373801916932905,
+ "grad_norm": 0.05859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0165,
+ "step": 18271
+ },
+ {
+ "epoch": 58.37699680511182,
+ "grad_norm": 0.05810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0202,
+ "step": 18272
+ },
+ {
+ "epoch": 58.38019169329073,
+ "grad_norm": 0.0576171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0318,
+ "step": 18273
+ },
+ {
+ "epoch": 58.38338658146965,
+ "grad_norm": 0.052490234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0179,
+ "step": 18274
+ },
+ {
+ "epoch": 58.38658146964856,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.026,
+ "step": 18275
+ },
+ {
+ "epoch": 58.38977635782748,
+ "grad_norm": 0.07861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0225,
+ "step": 18276
+ },
+ {
+ "epoch": 58.39297124600639,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0154,
+ "step": 18277
+ },
+ {
+ "epoch": 58.3961661341853,
+ "grad_norm": 0.09326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0224,
+ "step": 18278
+ },
+ {
+ "epoch": 58.39936102236422,
+ "grad_norm": 0.044189453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0367,
+ "step": 18279
+ },
+ {
+ "epoch": 58.40255591054313,
+ "grad_norm": 0.095703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0249,
+ "step": 18280
+ },
+ {
+ "epoch": 58.405750798722046,
+ "grad_norm": 0.04736328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0291,
+ "step": 18281
+ },
+ {
+ "epoch": 58.40894568690096,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.024,
+ "step": 18282
+ },
+ {
+ "epoch": 58.412140575079874,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0309,
+ "step": 18283
+ },
+ {
+ "epoch": 58.415335463258785,
+ "grad_norm": 0.057861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0301,
+ "step": 18284
+ },
+ {
+ "epoch": 58.4185303514377,
+ "grad_norm": 0.05908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0191,
+ "step": 18285
+ },
+ {
+ "epoch": 58.42172523961661,
+ "grad_norm": 0.04541015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0187,
+ "step": 18286
+ },
+ {
+ "epoch": 58.424920127795524,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0176,
+ "step": 18287
+ },
+ {
+ "epoch": 58.42811501597444,
+ "grad_norm": 0.052734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0263,
+ "step": 18288
+ },
+ {
+ "epoch": 58.43130990415335,
+ "grad_norm": 0.054931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0276,
+ "step": 18289
+ },
+ {
+ "epoch": 58.43450479233227,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.026,
+ "step": 18290
+ },
+ {
+ "epoch": 58.43769968051118,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0191,
+ "step": 18291
+ },
+ {
+ "epoch": 58.4408945686901,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.025,
+ "step": 18292
+ },
+ {
+ "epoch": 58.44408945686901,
+ "grad_norm": 0.04541015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0157,
+ "step": 18293
+ },
+ {
+ "epoch": 58.447284345047926,
+ "grad_norm": 0.059326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0285,
+ "step": 18294
+ },
+ {
+ "epoch": 58.45047923322684,
+ "grad_norm": 0.045166015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0266,
+ "step": 18295
+ },
+ {
+ "epoch": 58.453674121405754,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0235,
+ "step": 18296
+ },
+ {
+ "epoch": 58.456869009584665,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0244,
+ "step": 18297
+ },
+ {
+ "epoch": 58.460063897763575,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0212,
+ "step": 18298
+ },
+ {
+ "epoch": 58.46325878594249,
+ "grad_norm": 0.0615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0235,
+ "step": 18299
+ },
+ {
+ "epoch": 58.466453674121404,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0242,
+ "step": 18300
+ },
+ {
+ "epoch": 58.46964856230032,
+ "grad_norm": 0.09326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0231,
+ "step": 18301
+ },
+ {
+ "epoch": 58.47284345047923,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0201,
+ "step": 18302
+ },
+ {
+ "epoch": 58.47603833865815,
+ "grad_norm": 0.09765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0413,
+ "step": 18303
+ },
+ {
+ "epoch": 58.47923322683706,
+ "grad_norm": 0.146484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0256,
+ "step": 18304
+ },
+ {
+ "epoch": 58.48242811501598,
+ "grad_norm": 0.056884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0242,
+ "step": 18305
+ },
+ {
+ "epoch": 58.48562300319489,
+ "grad_norm": 0.1220703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0325,
+ "step": 18306
+ },
+ {
+ "epoch": 58.4888178913738,
+ "grad_norm": 0.1611328125,
+ "learning_rate": 0.0005,
+ "loss": 1.026,
+ "step": 18307
+ },
+ {
+ "epoch": 58.49201277955272,
+ "grad_norm": 0.08349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0214,
+ "step": 18308
+ },
+ {
+ "epoch": 58.49520766773163,
+ "grad_norm": 0.1875,
+ "learning_rate": 0.0005,
+ "loss": 1.0228,
+ "step": 18309
+ },
+ {
+ "epoch": 58.498402555910545,
+ "grad_norm": 0.091796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0227,
+ "step": 18310
+ },
+ {
+ "epoch": 58.501597444089455,
+ "grad_norm": 0.1689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0285,
+ "step": 18311
+ },
+ {
+ "epoch": 58.50479233226837,
+ "grad_norm": 0.1455078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0151,
+ "step": 18312
+ },
+ {
+ "epoch": 58.50798722044728,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0289,
+ "step": 18313
+ },
+ {
+ "epoch": 58.5111821086262,
+ "grad_norm": 0.1630859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0248,
+ "step": 18314
+ },
+ {
+ "epoch": 58.51437699680511,
+ "grad_norm": 0.10888671875,
+ "learning_rate": 0.0005,
+ "loss": 1.021,
+ "step": 18315
+ },
+ {
+ "epoch": 58.51757188498402,
+ "grad_norm": 0.1611328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0177,
+ "step": 18316
+ },
+ {
+ "epoch": 58.52076677316294,
+ "grad_norm": 0.1572265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0236,
+ "step": 18317
+ },
+ {
+ "epoch": 58.52396166134185,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0205,
+ "step": 18318
+ },
+ {
+ "epoch": 58.52715654952077,
+ "grad_norm": 0.162109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0212,
+ "step": 18319
+ },
+ {
+ "epoch": 58.53035143769968,
+ "grad_norm": 0.10107421875,
+ "learning_rate": 0.0005,
+ "loss": 1.026,
+ "step": 18320
+ },
+ {
+ "epoch": 58.533546325878596,
+ "grad_norm": 0.1376953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0257,
+ "step": 18321
+ },
+ {
+ "epoch": 58.53674121405751,
+ "grad_norm": 0.1259765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0243,
+ "step": 18322
+ },
+ {
+ "epoch": 58.539936102236425,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0312,
+ "step": 18323
+ },
+ {
+ "epoch": 58.543130990415335,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.036,
+ "step": 18324
+ },
+ {
+ "epoch": 58.546325878594246,
+ "grad_norm": 0.0908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0316,
+ "step": 18325
+ },
+ {
+ "epoch": 58.54952076677316,
+ "grad_norm": 0.06103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0129,
+ "step": 18326
+ },
+ {
+ "epoch": 58.552715654952074,
+ "grad_norm": 0.0908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0149,
+ "step": 18327
+ },
+ {
+ "epoch": 58.55591054313099,
+ "grad_norm": 0.103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0263,
+ "step": 18328
+ },
+ {
+ "epoch": 58.5591054313099,
+ "grad_norm": 0.08447265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0149,
+ "step": 18329
+ },
+ {
+ "epoch": 58.56230031948882,
+ "grad_norm": 0.1484375,
+ "learning_rate": 0.0005,
+ "loss": 1.027,
+ "step": 18330
+ },
+ {
+ "epoch": 58.56549520766773,
+ "grad_norm": 0.1416015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0285,
+ "step": 18331
+ },
+ {
+ "epoch": 58.56869009584665,
+ "grad_norm": 0.1669921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0215,
+ "step": 18332
+ },
+ {
+ "epoch": 58.57188498402556,
+ "grad_norm": 0.123046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0186,
+ "step": 18333
+ },
+ {
+ "epoch": 58.575079872204476,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0243,
+ "step": 18334
+ },
+ {
+ "epoch": 58.57827476038339,
+ "grad_norm": 0.16796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0288,
+ "step": 18335
+ },
+ {
+ "epoch": 58.5814696485623,
+ "grad_norm": 0.0859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0258,
+ "step": 18336
+ },
+ {
+ "epoch": 58.584664536741215,
+ "grad_norm": 0.08349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0268,
+ "step": 18337
+ },
+ {
+ "epoch": 58.587859424920126,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0221,
+ "step": 18338
+ },
+ {
+ "epoch": 58.59105431309904,
+ "grad_norm": 0.09619140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0336,
+ "step": 18339
+ },
+ {
+ "epoch": 58.594249201277954,
+ "grad_norm": 0.053955078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0292,
+ "step": 18340
+ },
+ {
+ "epoch": 58.59744408945687,
+ "grad_norm": 0.08740234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0205,
+ "step": 18341
+ },
+ {
+ "epoch": 58.60063897763578,
+ "grad_norm": 0.057373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0243,
+ "step": 18342
+ },
+ {
+ "epoch": 58.6038338658147,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0282,
+ "step": 18343
+ },
+ {
+ "epoch": 58.60702875399361,
+ "grad_norm": 0.09423828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0232,
+ "step": 18344
+ },
+ {
+ "epoch": 58.61022364217252,
+ "grad_norm": 0.055419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0244,
+ "step": 18345
+ },
+ {
+ "epoch": 58.61341853035144,
+ "grad_norm": 0.060302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0238,
+ "step": 18346
+ },
+ {
+ "epoch": 58.61661341853035,
+ "grad_norm": 0.0810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0365,
+ "step": 18347
+ },
+ {
+ "epoch": 58.61980830670927,
+ "grad_norm": 0.05712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0236,
+ "step": 18348
+ },
+ {
+ "epoch": 58.62300319488818,
+ "grad_norm": 0.0869140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0177,
+ "step": 18349
+ },
+ {
+ "epoch": 58.626198083067095,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0202,
+ "step": 18350
+ },
+ {
+ "epoch": 58.629392971246006,
+ "grad_norm": 0.06005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.029,
+ "step": 18351
+ },
+ {
+ "epoch": 58.63258785942492,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.022,
+ "step": 18352
+ },
+ {
+ "epoch": 58.635782747603834,
+ "grad_norm": 0.08642578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0246,
+ "step": 18353
+ },
+ {
+ "epoch": 58.638977635782744,
+ "grad_norm": 0.054931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0291,
+ "step": 18354
+ },
+ {
+ "epoch": 58.64217252396166,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0319,
+ "step": 18355
+ },
+ {
+ "epoch": 58.64536741214057,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0205,
+ "step": 18356
+ },
+ {
+ "epoch": 58.64856230031949,
+ "grad_norm": 0.046142578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0233,
+ "step": 18357
+ },
+ {
+ "epoch": 58.6517571884984,
+ "grad_norm": 0.0517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0247,
+ "step": 18358
+ },
+ {
+ "epoch": 58.65495207667732,
+ "grad_norm": 0.0810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0259,
+ "step": 18359
+ },
+ {
+ "epoch": 58.65814696485623,
+ "grad_norm": 0.052734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0208,
+ "step": 18360
+ },
+ {
+ "epoch": 58.66134185303515,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0167,
+ "step": 18361
+ },
+ {
+ "epoch": 58.66453674121406,
+ "grad_norm": 0.048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0322,
+ "step": 18362
+ },
+ {
+ "epoch": 58.66773162939297,
+ "grad_norm": 0.054931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0173,
+ "step": 18363
+ },
+ {
+ "epoch": 58.670926517571885,
+ "grad_norm": 0.062255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0217,
+ "step": 18364
+ },
+ {
+ "epoch": 58.674121405750796,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0161,
+ "step": 18365
+ },
+ {
+ "epoch": 58.677316293929714,
+ "grad_norm": 0.08349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0132,
+ "step": 18366
+ },
+ {
+ "epoch": 58.680511182108624,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0183,
+ "step": 18367
+ },
+ {
+ "epoch": 58.68370607028754,
+ "grad_norm": 0.09033203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0222,
+ "step": 18368
+ },
+ {
+ "epoch": 58.68690095846645,
+ "grad_norm": 0.049560546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0114,
+ "step": 18369
+ },
+ {
+ "epoch": 58.69009584664537,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0194,
+ "step": 18370
+ },
+ {
+ "epoch": 58.69329073482428,
+ "grad_norm": 0.06201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0238,
+ "step": 18371
+ },
+ {
+ "epoch": 58.6964856230032,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0269,
+ "step": 18372
+ },
+ {
+ "epoch": 58.69968051118211,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.03,
+ "step": 18373
+ },
+ {
+ "epoch": 58.70287539936102,
+ "grad_norm": 0.052734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0251,
+ "step": 18374
+ },
+ {
+ "epoch": 58.70607028753994,
+ "grad_norm": 0.06005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0193,
+ "step": 18375
+ },
+ {
+ "epoch": 58.70926517571885,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0378,
+ "step": 18376
+ },
+ {
+ "epoch": 58.712460063897765,
+ "grad_norm": 0.051025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.021,
+ "step": 18377
+ },
+ {
+ "epoch": 58.715654952076676,
+ "grad_norm": 0.061767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0234,
+ "step": 18378
+ },
+ {
+ "epoch": 58.718849840255594,
+ "grad_norm": 0.047607421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0197,
+ "step": 18379
+ },
+ {
+ "epoch": 58.722044728434504,
+ "grad_norm": 0.04833984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0227,
+ "step": 18380
+ },
+ {
+ "epoch": 58.72523961661342,
+ "grad_norm": 0.045654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0173,
+ "step": 18381
+ },
+ {
+ "epoch": 58.72843450479233,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0125,
+ "step": 18382
+ },
+ {
+ "epoch": 58.73162939297124,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0211,
+ "step": 18383
+ },
+ {
+ "epoch": 58.73482428115016,
+ "grad_norm": 0.055908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0288,
+ "step": 18384
+ },
+ {
+ "epoch": 58.73801916932907,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0259,
+ "step": 18385
+ },
+ {
+ "epoch": 58.74121405750799,
+ "grad_norm": 0.08447265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0333,
+ "step": 18386
+ },
+ {
+ "epoch": 58.7444089456869,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0251,
+ "step": 18387
+ },
+ {
+ "epoch": 58.74760383386582,
+ "grad_norm": 0.058837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0296,
+ "step": 18388
+ },
+ {
+ "epoch": 58.75079872204473,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0166,
+ "step": 18389
+ },
+ {
+ "epoch": 58.753993610223645,
+ "grad_norm": 0.061767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0295,
+ "step": 18390
+ },
+ {
+ "epoch": 58.757188498402556,
+ "grad_norm": 0.050537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0147,
+ "step": 18391
+ },
+ {
+ "epoch": 58.760383386581466,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0299,
+ "step": 18392
+ },
+ {
+ "epoch": 58.763578274760384,
+ "grad_norm": 0.052734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0187,
+ "step": 18393
+ },
+ {
+ "epoch": 58.766773162939295,
+ "grad_norm": 0.05908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0263,
+ "step": 18394
+ },
+ {
+ "epoch": 58.76996805111821,
+ "grad_norm": 0.059814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0203,
+ "step": 18395
+ },
+ {
+ "epoch": 58.77316293929712,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0163,
+ "step": 18396
+ },
+ {
+ "epoch": 58.77635782747604,
+ "grad_norm": 0.05615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0348,
+ "step": 18397
+ },
+ {
+ "epoch": 58.77955271565495,
+ "grad_norm": 0.0595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0281,
+ "step": 18398
+ },
+ {
+ "epoch": 58.78274760383387,
+ "grad_norm": 0.05029296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0181,
+ "step": 18399
+ },
+ {
+ "epoch": 58.78594249201278,
+ "grad_norm": 0.055419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0224,
+ "step": 18400
+ },
+ {
+ "epoch": 58.78913738019169,
+ "grad_norm": 0.0595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0223,
+ "step": 18401
+ },
+ {
+ "epoch": 58.79233226837061,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0328,
+ "step": 18402
+ },
+ {
+ "epoch": 58.79552715654952,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0284,
+ "step": 18403
+ },
+ {
+ "epoch": 58.798722044728436,
+ "grad_norm": 0.060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0264,
+ "step": 18404
+ },
+ {
+ "epoch": 58.801916932907346,
+ "grad_norm": 0.048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0261,
+ "step": 18405
+ },
+ {
+ "epoch": 58.805111821086264,
+ "grad_norm": 0.054931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0106,
+ "step": 18406
+ },
+ {
+ "epoch": 58.808306709265175,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.029,
+ "step": 18407
+ },
+ {
+ "epoch": 58.81150159744409,
+ "grad_norm": 0.04931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0286,
+ "step": 18408
+ },
+ {
+ "epoch": 58.814696485623,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0299,
+ "step": 18409
+ },
+ {
+ "epoch": 58.81789137380191,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0261,
+ "step": 18410
+ },
+ {
+ "epoch": 58.82108626198083,
+ "grad_norm": 0.05712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0283,
+ "step": 18411
+ },
+ {
+ "epoch": 58.82428115015974,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0268,
+ "step": 18412
+ },
+ {
+ "epoch": 58.82747603833866,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0318,
+ "step": 18413
+ },
+ {
+ "epoch": 58.83067092651757,
+ "grad_norm": 0.05078125,
+ "learning_rate": 0.0005,
+ "loss": 1.041,
+ "step": 18414
+ },
+ {
+ "epoch": 58.83386581469649,
+ "grad_norm": 0.10400390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0276,
+ "step": 18415
+ },
+ {
+ "epoch": 58.8370607028754,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0188,
+ "step": 18416
+ },
+ {
+ "epoch": 58.840255591054316,
+ "grad_norm": 0.0576171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0291,
+ "step": 18417
+ },
+ {
+ "epoch": 58.843450479233226,
+ "grad_norm": 0.1298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.032,
+ "step": 18418
+ },
+ {
+ "epoch": 58.846645367412144,
+ "grad_norm": 0.08544921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0231,
+ "step": 18419
+ },
+ {
+ "epoch": 58.849840255591054,
+ "grad_norm": 0.0576171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0267,
+ "step": 18420
+ },
+ {
+ "epoch": 58.853035143769965,
+ "grad_norm": 0.1455078125,
+ "learning_rate": 0.0005,
+ "loss": 1.016,
+ "step": 18421
+ },
+ {
+ "epoch": 58.85623003194888,
+ "grad_norm": 0.11181640625,
+ "learning_rate": 0.0005,
+ "loss": 1.021,
+ "step": 18422
+ },
+ {
+ "epoch": 58.85942492012779,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0319,
+ "step": 18423
+ },
+ {
+ "epoch": 58.86261980830671,
+ "grad_norm": 0.1396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.033,
+ "step": 18424
+ },
+ {
+ "epoch": 58.86581469648562,
+ "grad_norm": 0.09228515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0279,
+ "step": 18425
+ },
+ {
+ "epoch": 58.86900958466454,
+ "grad_norm": 0.12451171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0246,
+ "step": 18426
+ },
+ {
+ "epoch": 58.87220447284345,
+ "grad_norm": 0.154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.024,
+ "step": 18427
+ },
+ {
+ "epoch": 58.87539936102237,
+ "grad_norm": 0.09814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0234,
+ "step": 18428
+ },
+ {
+ "epoch": 58.87859424920128,
+ "grad_norm": 0.1435546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0141,
+ "step": 18429
+ },
+ {
+ "epoch": 58.88178913738019,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0223,
+ "step": 18430
+ },
+ {
+ "epoch": 58.884984025559106,
+ "grad_norm": 0.0927734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0314,
+ "step": 18431
+ },
+ {
+ "epoch": 58.88817891373802,
+ "grad_norm": 0.142578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0097,
+ "step": 18432
+ },
+ {
+ "epoch": 58.891373801916934,
+ "grad_norm": 0.0439453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0286,
+ "step": 18433
+ },
+ {
+ "epoch": 58.894568690095845,
+ "grad_norm": 0.10107421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0206,
+ "step": 18434
+ },
+ {
+ "epoch": 58.89776357827476,
+ "grad_norm": 0.08349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0375,
+ "step": 18435
+ },
+ {
+ "epoch": 58.90095846645367,
+ "grad_norm": 0.08544921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0393,
+ "step": 18436
+ },
+ {
+ "epoch": 58.90415335463259,
+ "grad_norm": 0.11572265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0297,
+ "step": 18437
+ },
+ {
+ "epoch": 58.9073482428115,
+ "grad_norm": 0.08935546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0344,
+ "step": 18438
+ },
+ {
+ "epoch": 58.91054313099041,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0318,
+ "step": 18439
+ },
+ {
+ "epoch": 58.91373801916933,
+ "grad_norm": 0.1123046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0212,
+ "step": 18440
+ },
+ {
+ "epoch": 58.91693290734824,
+ "grad_norm": 0.056396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0265,
+ "step": 18441
+ },
+ {
+ "epoch": 58.92012779552716,
+ "grad_norm": 0.05517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0294,
+ "step": 18442
+ },
+ {
+ "epoch": 58.92332268370607,
+ "grad_norm": 0.059814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0259,
+ "step": 18443
+ },
+ {
+ "epoch": 58.926517571884986,
+ "grad_norm": 0.052734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0222,
+ "step": 18444
+ },
+ {
+ "epoch": 58.9297124600639,
+ "grad_norm": 0.0498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0272,
+ "step": 18445
+ },
+ {
+ "epoch": 58.932907348242814,
+ "grad_norm": 0.0537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0185,
+ "step": 18446
+ },
+ {
+ "epoch": 58.936102236421725,
+ "grad_norm": 0.05810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0266,
+ "step": 18447
+ },
+ {
+ "epoch": 58.93929712460064,
+ "grad_norm": 0.055419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0423,
+ "step": 18448
+ },
+ {
+ "epoch": 58.94249201277955,
+ "grad_norm": 0.06103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0235,
+ "step": 18449
+ },
+ {
+ "epoch": 58.945686900958464,
+ "grad_norm": 0.052978515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0144,
+ "step": 18450
+ },
+ {
+ "epoch": 58.94888178913738,
+ "grad_norm": 0.0458984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0286,
+ "step": 18451
+ },
+ {
+ "epoch": 58.95207667731629,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0285,
+ "step": 18452
+ },
+ {
+ "epoch": 58.95527156549521,
+ "grad_norm": 0.048583984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0259,
+ "step": 18453
+ },
+ {
+ "epoch": 58.95846645367412,
+ "grad_norm": 0.04736328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0303,
+ "step": 18454
+ },
+ {
+ "epoch": 58.96166134185304,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0254,
+ "step": 18455
+ },
+ {
+ "epoch": 58.96485623003195,
+ "grad_norm": 0.051025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.02,
+ "step": 18456
+ },
+ {
+ "epoch": 58.968051118210866,
+ "grad_norm": 0.052734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0203,
+ "step": 18457
+ },
+ {
+ "epoch": 58.97124600638978,
+ "grad_norm": 0.050048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0255,
+ "step": 18458
+ },
+ {
+ "epoch": 58.97444089456869,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0205,
+ "step": 18459
+ },
+ {
+ "epoch": 58.977635782747605,
+ "grad_norm": 0.046630859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0259,
+ "step": 18460
+ },
+ {
+ "epoch": 58.980830670926515,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0214,
+ "step": 18461
+ },
+ {
+ "epoch": 58.98402555910543,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0213,
+ "step": 18462
+ },
+ {
+ "epoch": 58.98722044728434,
+ "grad_norm": 0.05517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.026,
+ "step": 18463
+ },
+ {
+ "epoch": 58.99041533546326,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0308,
+ "step": 18464
+ },
+ {
+ "epoch": 58.99361022364217,
+ "grad_norm": 0.050537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0266,
+ "step": 18465
+ },
+ {
+ "epoch": 58.99680511182109,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0268,
+ "step": 18466
+ },
+ {
+ "epoch": 59.0,
+ "grad_norm": 0.1015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0211,
+ "step": 18467
+ },
+ {
+ "epoch": 59.00319488817891,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0317,
+ "step": 18468
+ },
+ {
+ "epoch": 59.00638977635783,
+ "grad_norm": 0.1181640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0237,
+ "step": 18469
+ },
+ {
+ "epoch": 59.00958466453674,
+ "grad_norm": 0.1005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0207,
+ "step": 18470
+ },
+ {
+ "epoch": 59.01277955271566,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0421,
+ "step": 18471
+ },
+ {
+ "epoch": 59.01597444089457,
+ "grad_norm": 0.09814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0261,
+ "step": 18472
+ },
+ {
+ "epoch": 59.019169329073485,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.026,
+ "step": 18473
+ },
+ {
+ "epoch": 59.022364217252395,
+ "grad_norm": 0.05908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0249,
+ "step": 18474
+ },
+ {
+ "epoch": 59.02555910543131,
+ "grad_norm": 0.1298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0203,
+ "step": 18475
+ },
+ {
+ "epoch": 59.02875399361022,
+ "grad_norm": 0.08740234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0278,
+ "step": 18476
+ },
+ {
+ "epoch": 59.031948881789134,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0224,
+ "step": 18477
+ },
+ {
+ "epoch": 59.03514376996805,
+ "grad_norm": 0.10693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0365,
+ "step": 18478
+ },
+ {
+ "epoch": 59.03833865814696,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0283,
+ "step": 18479
+ },
+ {
+ "epoch": 59.04153354632588,
+ "grad_norm": 0.052001953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0217,
+ "step": 18480
+ },
+ {
+ "epoch": 59.04472843450479,
+ "grad_norm": 0.09521484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0236,
+ "step": 18481
+ },
+ {
+ "epoch": 59.04792332268371,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0237,
+ "step": 18482
+ },
+ {
+ "epoch": 59.05111821086262,
+ "grad_norm": 0.052734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0239,
+ "step": 18483
+ },
+ {
+ "epoch": 59.054313099041536,
+ "grad_norm": 0.056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0292,
+ "step": 18484
+ },
+ {
+ "epoch": 59.05750798722045,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0197,
+ "step": 18485
+ },
+ {
+ "epoch": 59.06070287539936,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0228,
+ "step": 18486
+ },
+ {
+ "epoch": 59.063897763578275,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0216,
+ "step": 18487
+ },
+ {
+ "epoch": 59.067092651757186,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0224,
+ "step": 18488
+ },
+ {
+ "epoch": 59.0702875399361,
+ "grad_norm": 0.048583984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0308,
+ "step": 18489
+ },
+ {
+ "epoch": 59.073482428115014,
+ "grad_norm": 0.054443359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0244,
+ "step": 18490
+ },
+ {
+ "epoch": 59.07667731629393,
+ "grad_norm": 0.048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0206,
+ "step": 18491
+ },
+ {
+ "epoch": 59.07987220447284,
+ "grad_norm": 0.05078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0166,
+ "step": 18492
+ },
+ {
+ "epoch": 59.08306709265176,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0333,
+ "step": 18493
+ },
+ {
+ "epoch": 59.08626198083067,
+ "grad_norm": 0.057861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0307,
+ "step": 18494
+ },
+ {
+ "epoch": 59.08945686900959,
+ "grad_norm": 0.04248046875,
+ "learning_rate": 0.0005,
+ "loss": 1.026,
+ "step": 18495
+ },
+ {
+ "epoch": 59.0926517571885,
+ "grad_norm": 0.059326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0235,
+ "step": 18496
+ },
+ {
+ "epoch": 59.09584664536741,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.027,
+ "step": 18497
+ },
+ {
+ "epoch": 59.09904153354633,
+ "grad_norm": 0.0517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0188,
+ "step": 18498
+ },
+ {
+ "epoch": 59.10223642172524,
+ "grad_norm": 0.056884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0243,
+ "step": 18499
+ },
+ {
+ "epoch": 59.105431309904155,
+ "grad_norm": 0.051513671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0234,
+ "step": 18500
+ },
+ {
+ "epoch": 59.108626198083066,
+ "grad_norm": 0.05810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0235,
+ "step": 18501
+ },
+ {
+ "epoch": 59.11182108626198,
+ "grad_norm": 0.04638671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0247,
+ "step": 18502
+ },
+ {
+ "epoch": 59.115015974440894,
+ "grad_norm": 0.055419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0262,
+ "step": 18503
+ },
+ {
+ "epoch": 59.11821086261981,
+ "grad_norm": 0.06005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0165,
+ "step": 18504
+ },
+ {
+ "epoch": 59.12140575079872,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0207,
+ "step": 18505
+ },
+ {
+ "epoch": 59.12460063897763,
+ "grad_norm": 0.054931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0311,
+ "step": 18506
+ },
+ {
+ "epoch": 59.12779552715655,
+ "grad_norm": 0.050048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0305,
+ "step": 18507
+ },
+ {
+ "epoch": 59.13099041533546,
+ "grad_norm": 0.054931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0288,
+ "step": 18508
+ },
+ {
+ "epoch": 59.13418530351438,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.028,
+ "step": 18509
+ },
+ {
+ "epoch": 59.13738019169329,
+ "grad_norm": 0.047607421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0217,
+ "step": 18510
+ },
+ {
+ "epoch": 59.14057507987221,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0247,
+ "step": 18511
+ },
+ {
+ "epoch": 59.14376996805112,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0315,
+ "step": 18512
+ },
+ {
+ "epoch": 59.146964856230035,
+ "grad_norm": 0.053466796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0197,
+ "step": 18513
+ },
+ {
+ "epoch": 59.150159744408946,
+ "grad_norm": 0.08642578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0246,
+ "step": 18514
+ },
+ {
+ "epoch": 59.153354632587856,
+ "grad_norm": 0.04150390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0259,
+ "step": 18515
+ },
+ {
+ "epoch": 59.156549520766774,
+ "grad_norm": 0.049560546875,
+ "learning_rate": 0.0005,
+ "loss": 1.024,
+ "step": 18516
+ },
+ {
+ "epoch": 59.159744408945684,
+ "grad_norm": 0.049072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0292,
+ "step": 18517
+ },
+ {
+ "epoch": 59.1629392971246,
+ "grad_norm": 0.042236328125,
+ "learning_rate": 0.0005,
+ "loss": 1.029,
+ "step": 18518
+ },
+ {
+ "epoch": 59.16613418530351,
+ "grad_norm": 0.04833984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0394,
+ "step": 18519
+ },
+ {
+ "epoch": 59.16932907348243,
+ "grad_norm": 0.0458984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0305,
+ "step": 18520
+ },
+ {
+ "epoch": 59.17252396166134,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.017,
+ "step": 18521
+ },
+ {
+ "epoch": 59.17571884984026,
+ "grad_norm": 0.048583984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0272,
+ "step": 18522
+ },
+ {
+ "epoch": 59.17891373801917,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0348,
+ "step": 18523
+ },
+ {
+ "epoch": 59.18210862619808,
+ "grad_norm": 0.09423828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0207,
+ "step": 18524
+ },
+ {
+ "epoch": 59.185303514377,
+ "grad_norm": 0.06201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0267,
+ "step": 18525
+ },
+ {
+ "epoch": 59.18849840255591,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0165,
+ "step": 18526
+ },
+ {
+ "epoch": 59.191693290734825,
+ "grad_norm": 0.04296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0185,
+ "step": 18527
+ },
+ {
+ "epoch": 59.194888178913736,
+ "grad_norm": 0.05029296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0371,
+ "step": 18528
+ },
+ {
+ "epoch": 59.198083067092654,
+ "grad_norm": 0.061279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.022,
+ "step": 18529
+ },
+ {
+ "epoch": 59.201277955271564,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0116,
+ "step": 18530
+ },
+ {
+ "epoch": 59.20447284345048,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0247,
+ "step": 18531
+ },
+ {
+ "epoch": 59.20766773162939,
+ "grad_norm": 0.058837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0236,
+ "step": 18532
+ },
+ {
+ "epoch": 59.21086261980831,
+ "grad_norm": 0.126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0081,
+ "step": 18533
+ },
+ {
+ "epoch": 59.21405750798722,
+ "grad_norm": 0.05615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0222,
+ "step": 18534
+ },
+ {
+ "epoch": 59.21725239616613,
+ "grad_norm": 0.0859375,
+ "learning_rate": 0.0005,
+ "loss": 1.03,
+ "step": 18535
+ },
+ {
+ "epoch": 59.22044728434505,
+ "grad_norm": 0.053955078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0276,
+ "step": 18536
+ },
+ {
+ "epoch": 59.22364217252396,
+ "grad_norm": 0.050537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0223,
+ "step": 18537
+ },
+ {
+ "epoch": 59.22683706070288,
+ "grad_norm": 0.060791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0132,
+ "step": 18538
+ },
+ {
+ "epoch": 59.23003194888179,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0266,
+ "step": 18539
+ },
+ {
+ "epoch": 59.233226837060705,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0284,
+ "step": 18540
+ },
+ {
+ "epoch": 59.236421725239616,
+ "grad_norm": 0.050048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0167,
+ "step": 18541
+ },
+ {
+ "epoch": 59.239616613418534,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0172,
+ "step": 18542
+ },
+ {
+ "epoch": 59.242811501597444,
+ "grad_norm": 0.0576171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0253,
+ "step": 18543
+ },
+ {
+ "epoch": 59.246006389776355,
+ "grad_norm": 0.08642578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0161,
+ "step": 18544
+ },
+ {
+ "epoch": 59.24920127795527,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0279,
+ "step": 18545
+ },
+ {
+ "epoch": 59.25239616613418,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0314,
+ "step": 18546
+ },
+ {
+ "epoch": 59.2555910543131,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0197,
+ "step": 18547
+ },
+ {
+ "epoch": 59.25878594249201,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0238,
+ "step": 18548
+ },
+ {
+ "epoch": 59.26198083067093,
+ "grad_norm": 0.08837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0174,
+ "step": 18549
+ },
+ {
+ "epoch": 59.26517571884984,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0237,
+ "step": 18550
+ },
+ {
+ "epoch": 59.26837060702876,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.017,
+ "step": 18551
+ },
+ {
+ "epoch": 59.27156549520767,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0263,
+ "step": 18552
+ },
+ {
+ "epoch": 59.27476038338658,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0266,
+ "step": 18553
+ },
+ {
+ "epoch": 59.277955271565496,
+ "grad_norm": 0.056396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.014,
+ "step": 18554
+ },
+ {
+ "epoch": 59.281150159744406,
+ "grad_norm": 0.050048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0296,
+ "step": 18555
+ },
+ {
+ "epoch": 59.284345047923324,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0248,
+ "step": 18556
+ },
+ {
+ "epoch": 59.287539936102235,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0281,
+ "step": 18557
+ },
+ {
+ "epoch": 59.29073482428115,
+ "grad_norm": 0.11669921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0268,
+ "step": 18558
+ },
+ {
+ "epoch": 59.29392971246006,
+ "grad_norm": 0.048583984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0255,
+ "step": 18559
+ },
+ {
+ "epoch": 59.29712460063898,
+ "grad_norm": 0.05908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0217,
+ "step": 18560
+ },
+ {
+ "epoch": 59.30031948881789,
+ "grad_norm": 0.055419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0308,
+ "step": 18561
+ },
+ {
+ "epoch": 59.3035143769968,
+ "grad_norm": 0.08447265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0303,
+ "step": 18562
+ },
+ {
+ "epoch": 59.30670926517572,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0258,
+ "step": 18563
+ },
+ {
+ "epoch": 59.30990415335463,
+ "grad_norm": 0.060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0252,
+ "step": 18564
+ },
+ {
+ "epoch": 59.31309904153355,
+ "grad_norm": 0.060791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0308,
+ "step": 18565
+ },
+ {
+ "epoch": 59.31629392971246,
+ "grad_norm": 0.06201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0324,
+ "step": 18566
+ },
+ {
+ "epoch": 59.319488817891376,
+ "grad_norm": 0.04541015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0254,
+ "step": 18567
+ },
+ {
+ "epoch": 59.322683706070286,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0201,
+ "step": 18568
+ },
+ {
+ "epoch": 59.325878594249204,
+ "grad_norm": 0.04931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0174,
+ "step": 18569
+ },
+ {
+ "epoch": 59.329073482428115,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0168,
+ "step": 18570
+ },
+ {
+ "epoch": 59.33226837060703,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0193,
+ "step": 18571
+ },
+ {
+ "epoch": 59.33546325878594,
+ "grad_norm": 0.08056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0341,
+ "step": 18572
+ },
+ {
+ "epoch": 59.33865814696485,
+ "grad_norm": 0.048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0294,
+ "step": 18573
+ },
+ {
+ "epoch": 59.34185303514377,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0199,
+ "step": 18574
+ },
+ {
+ "epoch": 59.34504792332268,
+ "grad_norm": 0.0498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0316,
+ "step": 18575
+ },
+ {
+ "epoch": 59.3482428115016,
+ "grad_norm": 0.0556640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0264,
+ "step": 18576
+ },
+ {
+ "epoch": 59.35143769968051,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0288,
+ "step": 18577
+ },
+ {
+ "epoch": 59.35463258785943,
+ "grad_norm": 0.0498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0304,
+ "step": 18578
+ },
+ {
+ "epoch": 59.35782747603834,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0143,
+ "step": 18579
+ },
+ {
+ "epoch": 59.361022364217256,
+ "grad_norm": 0.051025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0262,
+ "step": 18580
+ },
+ {
+ "epoch": 59.364217252396166,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0399,
+ "step": 18581
+ },
+ {
+ "epoch": 59.36741214057508,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0288,
+ "step": 18582
+ },
+ {
+ "epoch": 59.370607028753994,
+ "grad_norm": 0.062255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0258,
+ "step": 18583
+ },
+ {
+ "epoch": 59.373801916932905,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0064,
+ "step": 18584
+ },
+ {
+ "epoch": 59.37699680511182,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0338,
+ "step": 18585
+ },
+ {
+ "epoch": 59.38019169329073,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0178,
+ "step": 18586
+ },
+ {
+ "epoch": 59.38338658146965,
+ "grad_norm": 0.0810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0169,
+ "step": 18587
+ },
+ {
+ "epoch": 59.38658146964856,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0101,
+ "step": 18588
+ },
+ {
+ "epoch": 59.38977635782748,
+ "grad_norm": 0.08056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0304,
+ "step": 18589
+ },
+ {
+ "epoch": 59.39297124600639,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.024,
+ "step": 18590
+ },
+ {
+ "epoch": 59.3961661341853,
+ "grad_norm": 0.050048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0144,
+ "step": 18591
+ },
+ {
+ "epoch": 59.39936102236422,
+ "grad_norm": 0.058349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0228,
+ "step": 18592
+ },
+ {
+ "epoch": 59.40255591054313,
+ "grad_norm": 0.056396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0175,
+ "step": 18593
+ },
+ {
+ "epoch": 59.405750798722046,
+ "grad_norm": 0.052734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0301,
+ "step": 18594
+ },
+ {
+ "epoch": 59.40894568690096,
+ "grad_norm": 0.052734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0241,
+ "step": 18595
+ },
+ {
+ "epoch": 59.412140575079874,
+ "grad_norm": 0.08935546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0202,
+ "step": 18596
+ },
+ {
+ "epoch": 59.415335463258785,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0381,
+ "step": 18597
+ },
+ {
+ "epoch": 59.4185303514377,
+ "grad_norm": 0.07861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0391,
+ "step": 18598
+ },
+ {
+ "epoch": 59.42172523961661,
+ "grad_norm": 0.08203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0272,
+ "step": 18599
+ },
+ {
+ "epoch": 59.424920127795524,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0239,
+ "step": 18600
+ },
+ {
+ "epoch": 59.42811501597444,
+ "grad_norm": 0.056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0246,
+ "step": 18601
+ },
+ {
+ "epoch": 59.43130990415335,
+ "grad_norm": 0.0849609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0182,
+ "step": 18602
+ },
+ {
+ "epoch": 59.43450479233227,
+ "grad_norm": 0.049560546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0297,
+ "step": 18603
+ },
+ {
+ "epoch": 59.43769968051118,
+ "grad_norm": 0.056396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0276,
+ "step": 18604
+ },
+ {
+ "epoch": 59.4408945686901,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0278,
+ "step": 18605
+ },
+ {
+ "epoch": 59.44408945686901,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0236,
+ "step": 18606
+ },
+ {
+ "epoch": 59.447284345047926,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0321,
+ "step": 18607
+ },
+ {
+ "epoch": 59.45047923322684,
+ "grad_norm": 0.0556640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0255,
+ "step": 18608
+ },
+ {
+ "epoch": 59.453674121405754,
+ "grad_norm": 0.0849609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0151,
+ "step": 18609
+ },
+ {
+ "epoch": 59.456869009584665,
+ "grad_norm": 0.048583984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0306,
+ "step": 18610
+ },
+ {
+ "epoch": 59.460063897763575,
+ "grad_norm": 0.08447265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0307,
+ "step": 18611
+ },
+ {
+ "epoch": 59.46325878594249,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0406,
+ "step": 18612
+ },
+ {
+ "epoch": 59.466453674121404,
+ "grad_norm": 0.05029296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0304,
+ "step": 18613
+ },
+ {
+ "epoch": 59.46964856230032,
+ "grad_norm": 0.050537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0322,
+ "step": 18614
+ },
+ {
+ "epoch": 59.47284345047923,
+ "grad_norm": 0.049072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0275,
+ "step": 18615
+ },
+ {
+ "epoch": 59.47603833865815,
+ "grad_norm": 0.059814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0223,
+ "step": 18616
+ },
+ {
+ "epoch": 59.47923322683706,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0259,
+ "step": 18617
+ },
+ {
+ "epoch": 59.48242811501598,
+ "grad_norm": 0.045166015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0256,
+ "step": 18618
+ },
+ {
+ "epoch": 59.48562300319489,
+ "grad_norm": 0.054443359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0184,
+ "step": 18619
+ },
+ {
+ "epoch": 59.4888178913738,
+ "grad_norm": 0.044921875,
+ "learning_rate": 0.0005,
+ "loss": 1.021,
+ "step": 18620
+ },
+ {
+ "epoch": 59.49201277955272,
+ "grad_norm": 0.05224609375,
+ "learning_rate": 0.0005,
+ "loss": 1.021,
+ "step": 18621
+ },
+ {
+ "epoch": 59.49520766773163,
+ "grad_norm": 0.046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0298,
+ "step": 18622
+ },
+ {
+ "epoch": 59.498402555910545,
+ "grad_norm": 0.0546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0266,
+ "step": 18623
+ },
+ {
+ "epoch": 59.501597444089455,
+ "grad_norm": 0.056396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0101,
+ "step": 18624
+ },
+ {
+ "epoch": 59.50479233226837,
+ "grad_norm": 0.052734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0231,
+ "step": 18625
+ },
+ {
+ "epoch": 59.50798722044728,
+ "grad_norm": 0.0888671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0212,
+ "step": 18626
+ },
+ {
+ "epoch": 59.5111821086262,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0317,
+ "step": 18627
+ },
+ {
+ "epoch": 59.51437699680511,
+ "grad_norm": 0.09326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0194,
+ "step": 18628
+ },
+ {
+ "epoch": 59.51757188498402,
+ "grad_norm": 0.05859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0238,
+ "step": 18629
+ },
+ {
+ "epoch": 59.52076677316294,
+ "grad_norm": 0.0498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0143,
+ "step": 18630
+ },
+ {
+ "epoch": 59.52396166134185,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0237,
+ "step": 18631
+ },
+ {
+ "epoch": 59.52715654952077,
+ "grad_norm": 0.045654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0335,
+ "step": 18632
+ },
+ {
+ "epoch": 59.53035143769968,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0179,
+ "step": 18633
+ },
+ {
+ "epoch": 59.533546325878596,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0161,
+ "step": 18634
+ },
+ {
+ "epoch": 59.53674121405751,
+ "grad_norm": 0.048583984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0297,
+ "step": 18635
+ },
+ {
+ "epoch": 59.539936102236425,
+ "grad_norm": 0.061767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0252,
+ "step": 18636
+ },
+ {
+ "epoch": 59.543130990415335,
+ "grad_norm": 0.059814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0343,
+ "step": 18637
+ },
+ {
+ "epoch": 59.546325878594246,
+ "grad_norm": 0.045166015625,
+ "learning_rate": 0.0005,
+ "loss": 1.029,
+ "step": 18638
+ },
+ {
+ "epoch": 59.54952076677316,
+ "grad_norm": 0.049560546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0308,
+ "step": 18639
+ },
+ {
+ "epoch": 59.552715654952074,
+ "grad_norm": 0.054443359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0258,
+ "step": 18640
+ },
+ {
+ "epoch": 59.55591054313099,
+ "grad_norm": 0.050048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0203,
+ "step": 18641
+ },
+ {
+ "epoch": 59.5591054313099,
+ "grad_norm": 0.0498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0244,
+ "step": 18642
+ },
+ {
+ "epoch": 59.56230031948882,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0208,
+ "step": 18643
+ },
+ {
+ "epoch": 59.56549520766773,
+ "grad_norm": 0.060791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.029,
+ "step": 18644
+ },
+ {
+ "epoch": 59.56869009584665,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.027,
+ "step": 18645
+ },
+ {
+ "epoch": 59.57188498402556,
+ "grad_norm": 0.0478515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0274,
+ "step": 18646
+ },
+ {
+ "epoch": 59.575079872204476,
+ "grad_norm": 0.051513671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0293,
+ "step": 18647
+ },
+ {
+ "epoch": 59.57827476038339,
+ "grad_norm": 0.04443359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0213,
+ "step": 18648
+ },
+ {
+ "epoch": 59.5814696485623,
+ "grad_norm": 0.05908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0261,
+ "step": 18649
+ },
+ {
+ "epoch": 59.584664536741215,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0291,
+ "step": 18650
+ },
+ {
+ "epoch": 59.587859424920126,
+ "grad_norm": 0.10498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0268,
+ "step": 18651
+ },
+ {
+ "epoch": 59.59105431309904,
+ "grad_norm": 0.04833984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0346,
+ "step": 18652
+ },
+ {
+ "epoch": 59.594249201277954,
+ "grad_norm": 0.056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0275,
+ "step": 18653
+ },
+ {
+ "epoch": 59.59744408945687,
+ "grad_norm": 0.09130859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0272,
+ "step": 18654
+ },
+ {
+ "epoch": 59.60063897763578,
+ "grad_norm": 0.061279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.017,
+ "step": 18655
+ },
+ {
+ "epoch": 59.6038338658147,
+ "grad_norm": 0.050048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0213,
+ "step": 18656
+ },
+ {
+ "epoch": 59.60702875399361,
+ "grad_norm": 0.08251953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0222,
+ "step": 18657
+ },
+ {
+ "epoch": 59.61022364217252,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0288,
+ "step": 18658
+ },
+ {
+ "epoch": 59.61341853035144,
+ "grad_norm": 0.05078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0263,
+ "step": 18659
+ },
+ {
+ "epoch": 59.61661341853035,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0339,
+ "step": 18660
+ },
+ {
+ "epoch": 59.61980830670927,
+ "grad_norm": 0.057861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0253,
+ "step": 18661
+ },
+ {
+ "epoch": 59.62300319488818,
+ "grad_norm": 0.0517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0137,
+ "step": 18662
+ },
+ {
+ "epoch": 59.626198083067095,
+ "grad_norm": 0.0556640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0296,
+ "step": 18663
+ },
+ {
+ "epoch": 59.629392971246006,
+ "grad_norm": 0.053466796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0265,
+ "step": 18664
+ },
+ {
+ "epoch": 59.63258785942492,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0367,
+ "step": 18665
+ },
+ {
+ "epoch": 59.635782747603834,
+ "grad_norm": 0.053466796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0258,
+ "step": 18666
+ },
+ {
+ "epoch": 59.638977635782744,
+ "grad_norm": 0.05078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0226,
+ "step": 18667
+ },
+ {
+ "epoch": 59.64217252396166,
+ "grad_norm": 0.05078125,
+ "learning_rate": 0.0005,
+ "loss": 1.031,
+ "step": 18668
+ },
+ {
+ "epoch": 59.64536741214057,
+ "grad_norm": 0.059326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0147,
+ "step": 18669
+ },
+ {
+ "epoch": 59.64856230031949,
+ "grad_norm": 0.055908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0269,
+ "step": 18670
+ },
+ {
+ "epoch": 59.6517571884984,
+ "grad_norm": 0.059326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0182,
+ "step": 18671
+ },
+ {
+ "epoch": 59.65495207667732,
+ "grad_norm": 0.05224609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0308,
+ "step": 18672
+ },
+ {
+ "epoch": 59.65814696485623,
+ "grad_norm": 0.08984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0249,
+ "step": 18673
+ },
+ {
+ "epoch": 59.66134185303515,
+ "grad_norm": 0.060791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.02,
+ "step": 18674
+ },
+ {
+ "epoch": 59.66453674121406,
+ "grad_norm": 0.0556640625,
+ "learning_rate": 0.0005,
+ "loss": 1.025,
+ "step": 18675
+ },
+ {
+ "epoch": 59.66773162939297,
+ "grad_norm": 0.060302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0174,
+ "step": 18676
+ },
+ {
+ "epoch": 59.670926517571885,
+ "grad_norm": 0.049560546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0169,
+ "step": 18677
+ },
+ {
+ "epoch": 59.674121405750796,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.033,
+ "step": 18678
+ },
+ {
+ "epoch": 59.677316293929714,
+ "grad_norm": 0.051513671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0303,
+ "step": 18679
+ },
+ {
+ "epoch": 59.680511182108624,
+ "grad_norm": 0.05859375,
+ "learning_rate": 0.0005,
+ "loss": 1.017,
+ "step": 18680
+ },
+ {
+ "epoch": 59.68370607028754,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0221,
+ "step": 18681
+ },
+ {
+ "epoch": 59.68690095846645,
+ "grad_norm": 0.080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0186,
+ "step": 18682
+ },
+ {
+ "epoch": 59.69009584664537,
+ "grad_norm": 0.06005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0199,
+ "step": 18683
+ },
+ {
+ "epoch": 59.69329073482428,
+ "grad_norm": 0.0869140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0144,
+ "step": 18684
+ },
+ {
+ "epoch": 59.6964856230032,
+ "grad_norm": 0.0615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.02,
+ "step": 18685
+ },
+ {
+ "epoch": 59.69968051118211,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0166,
+ "step": 18686
+ },
+ {
+ "epoch": 59.70287539936102,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0379,
+ "step": 18687
+ },
+ {
+ "epoch": 59.70607028753994,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0193,
+ "step": 18688
+ },
+ {
+ "epoch": 59.70926517571885,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0197,
+ "step": 18689
+ },
+ {
+ "epoch": 59.712460063897765,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.026,
+ "step": 18690
+ },
+ {
+ "epoch": 59.715654952076676,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0236,
+ "step": 18691
+ },
+ {
+ "epoch": 59.718849840255594,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0163,
+ "step": 18692
+ },
+ {
+ "epoch": 59.722044728434504,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0203,
+ "step": 18693
+ },
+ {
+ "epoch": 59.72523961661342,
+ "grad_norm": 0.0546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0213,
+ "step": 18694
+ },
+ {
+ "epoch": 59.72843450479233,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0185,
+ "step": 18695
+ },
+ {
+ "epoch": 59.73162939297124,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0234,
+ "step": 18696
+ },
+ {
+ "epoch": 59.73482428115016,
+ "grad_norm": 0.083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0309,
+ "step": 18697
+ },
+ {
+ "epoch": 59.73801916932907,
+ "grad_norm": 0.056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0349,
+ "step": 18698
+ },
+ {
+ "epoch": 59.74121405750799,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0216,
+ "step": 18699
+ },
+ {
+ "epoch": 59.7444089456869,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0222,
+ "step": 18700
+ },
+ {
+ "epoch": 59.74760383386582,
+ "grad_norm": 0.0546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0334,
+ "step": 18701
+ },
+ {
+ "epoch": 59.75079872204473,
+ "grad_norm": 0.055419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0293,
+ "step": 18702
+ },
+ {
+ "epoch": 59.753993610223645,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0241,
+ "step": 18703
+ },
+ {
+ "epoch": 59.757188498402556,
+ "grad_norm": 0.056396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0311,
+ "step": 18704
+ },
+ {
+ "epoch": 59.760383386581466,
+ "grad_norm": 0.05126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0291,
+ "step": 18705
+ },
+ {
+ "epoch": 59.763578274760384,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0355,
+ "step": 18706
+ },
+ {
+ "epoch": 59.766773162939295,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.019,
+ "step": 18707
+ },
+ {
+ "epoch": 59.76996805111821,
+ "grad_norm": 0.05908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0219,
+ "step": 18708
+ },
+ {
+ "epoch": 59.77316293929712,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0309,
+ "step": 18709
+ },
+ {
+ "epoch": 59.77635782747604,
+ "grad_norm": 0.0947265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0201,
+ "step": 18710
+ },
+ {
+ "epoch": 59.77955271565495,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0337,
+ "step": 18711
+ },
+ {
+ "epoch": 59.78274760383387,
+ "grad_norm": 0.11376953125,
+ "learning_rate": 0.0005,
+ "loss": 1.02,
+ "step": 18712
+ },
+ {
+ "epoch": 59.78594249201278,
+ "grad_norm": 0.05419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0294,
+ "step": 18713
+ },
+ {
+ "epoch": 59.78913738019169,
+ "grad_norm": 0.08251953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0206,
+ "step": 18714
+ },
+ {
+ "epoch": 59.79233226837061,
+ "grad_norm": 0.083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0285,
+ "step": 18715
+ },
+ {
+ "epoch": 59.79552715654952,
+ "grad_norm": 0.06201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.037,
+ "step": 18716
+ },
+ {
+ "epoch": 59.798722044728436,
+ "grad_norm": 0.054443359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0243,
+ "step": 18717
+ },
+ {
+ "epoch": 59.801916932907346,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0291,
+ "step": 18718
+ },
+ {
+ "epoch": 59.805111821086264,
+ "grad_norm": 0.09619140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0245,
+ "step": 18719
+ },
+ {
+ "epoch": 59.808306709265175,
+ "grad_norm": 0.05517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0186,
+ "step": 18720
+ },
+ {
+ "epoch": 59.81150159744409,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0273,
+ "step": 18721
+ },
+ {
+ "epoch": 59.814696485623,
+ "grad_norm": 0.0869140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0258,
+ "step": 18722
+ },
+ {
+ "epoch": 59.81789137380191,
+ "grad_norm": 0.06005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0339,
+ "step": 18723
+ },
+ {
+ "epoch": 59.82108626198083,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0279,
+ "step": 18724
+ },
+ {
+ "epoch": 59.82428115015974,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0363,
+ "step": 18725
+ },
+ {
+ "epoch": 59.82747603833866,
+ "grad_norm": 0.0546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0169,
+ "step": 18726
+ },
+ {
+ "epoch": 59.83067092651757,
+ "grad_norm": 0.04736328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0233,
+ "step": 18727
+ },
+ {
+ "epoch": 59.83386581469649,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0383,
+ "step": 18728
+ },
+ {
+ "epoch": 59.8370607028754,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0267,
+ "step": 18729
+ },
+ {
+ "epoch": 59.840255591054316,
+ "grad_norm": 0.057373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0239,
+ "step": 18730
+ },
+ {
+ "epoch": 59.843450479233226,
+ "grad_norm": 0.0556640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0326,
+ "step": 18731
+ },
+ {
+ "epoch": 59.846645367412144,
+ "grad_norm": 0.0556640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0246,
+ "step": 18732
+ },
+ {
+ "epoch": 59.849840255591054,
+ "grad_norm": 0.048583984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0322,
+ "step": 18733
+ },
+ {
+ "epoch": 59.853035143769965,
+ "grad_norm": 0.044189453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0305,
+ "step": 18734
+ },
+ {
+ "epoch": 59.85623003194888,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.026,
+ "step": 18735
+ },
+ {
+ "epoch": 59.85942492012779,
+ "grad_norm": 0.0458984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0179,
+ "step": 18736
+ },
+ {
+ "epoch": 59.86261980830671,
+ "grad_norm": 0.060791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0271,
+ "step": 18737
+ },
+ {
+ "epoch": 59.86581469648562,
+ "grad_norm": 0.0478515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0247,
+ "step": 18738
+ },
+ {
+ "epoch": 59.86900958466454,
+ "grad_norm": 0.047119140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0211,
+ "step": 18739
+ },
+ {
+ "epoch": 59.87220447284345,
+ "grad_norm": 0.048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0308,
+ "step": 18740
+ },
+ {
+ "epoch": 59.87539936102237,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0213,
+ "step": 18741
+ },
+ {
+ "epoch": 59.87859424920128,
+ "grad_norm": 0.054931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0356,
+ "step": 18742
+ },
+ {
+ "epoch": 59.88178913738019,
+ "grad_norm": 0.046630859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0228,
+ "step": 18743
+ },
+ {
+ "epoch": 59.884984025559106,
+ "grad_norm": 0.058349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0233,
+ "step": 18744
+ },
+ {
+ "epoch": 59.88817891373802,
+ "grad_norm": 0.045654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0166,
+ "step": 18745
+ },
+ {
+ "epoch": 59.891373801916934,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0234,
+ "step": 18746
+ },
+ {
+ "epoch": 59.894568690095845,
+ "grad_norm": 0.05517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0245,
+ "step": 18747
+ },
+ {
+ "epoch": 59.89776357827476,
+ "grad_norm": 0.050048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0285,
+ "step": 18748
+ },
+ {
+ "epoch": 59.90095846645367,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0188,
+ "step": 18749
+ },
+ {
+ "epoch": 59.90415335463259,
+ "grad_norm": 0.0556640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0294,
+ "step": 18750
+ },
+ {
+ "epoch": 59.9073482428115,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0252,
+ "step": 18751
+ },
+ {
+ "epoch": 59.91054313099041,
+ "grad_norm": 0.0927734375,
+ "learning_rate": 0.0005,
+ "loss": 1.032,
+ "step": 18752
+ },
+ {
+ "epoch": 59.91373801916933,
+ "grad_norm": 0.083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0216,
+ "step": 18753
+ },
+ {
+ "epoch": 59.91693290734824,
+ "grad_norm": 0.055908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0299,
+ "step": 18754
+ },
+ {
+ "epoch": 59.92012779552716,
+ "grad_norm": 0.0830078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0307,
+ "step": 18755
+ },
+ {
+ "epoch": 59.92332268370607,
+ "grad_norm": 0.08544921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0256,
+ "step": 18756
+ },
+ {
+ "epoch": 59.926517571884986,
+ "grad_norm": 0.059814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0295,
+ "step": 18757
+ },
+ {
+ "epoch": 59.9297124600639,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0341,
+ "step": 18758
+ },
+ {
+ "epoch": 59.932907348242814,
+ "grad_norm": 0.057373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0301,
+ "step": 18759
+ },
+ {
+ "epoch": 59.936102236421725,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0323,
+ "step": 18760
+ },
+ {
+ "epoch": 59.93929712460064,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0287,
+ "step": 18761
+ },
+ {
+ "epoch": 59.94249201277955,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0212,
+ "step": 18762
+ },
+ {
+ "epoch": 59.945686900958464,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0193,
+ "step": 18763
+ },
+ {
+ "epoch": 59.94888178913738,
+ "grad_norm": 0.051025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0177,
+ "step": 18764
+ },
+ {
+ "epoch": 59.95207667731629,
+ "grad_norm": 0.056396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0135,
+ "step": 18765
+ },
+ {
+ "epoch": 59.95527156549521,
+ "grad_norm": 0.05224609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0162,
+ "step": 18766
+ },
+ {
+ "epoch": 59.95846645367412,
+ "grad_norm": 0.08154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0208,
+ "step": 18767
+ },
+ {
+ "epoch": 59.96166134185304,
+ "grad_norm": 0.05419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0213,
+ "step": 18768
+ },
+ {
+ "epoch": 59.96485623003195,
+ "grad_norm": 0.09521484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0107,
+ "step": 18769
+ },
+ {
+ "epoch": 59.968051118210866,
+ "grad_norm": 0.05908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.025,
+ "step": 18770
+ },
+ {
+ "epoch": 59.97124600638978,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0158,
+ "step": 18771
+ },
+ {
+ "epoch": 59.97444089456869,
+ "grad_norm": 0.087890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0101,
+ "step": 18772
+ },
+ {
+ "epoch": 59.977635782747605,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0247,
+ "step": 18773
+ },
+ {
+ "epoch": 59.980830670926515,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0218,
+ "step": 18774
+ },
+ {
+ "epoch": 59.98402555910543,
+ "grad_norm": 0.061279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0293,
+ "step": 18775
+ },
+ {
+ "epoch": 59.98722044728434,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0176,
+ "step": 18776
+ },
+ {
+ "epoch": 59.99041533546326,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0372,
+ "step": 18777
+ },
+ {
+ "epoch": 59.99361022364217,
+ "grad_norm": 0.061279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.021,
+ "step": 18778
+ },
+ {
+ "epoch": 59.99680511182109,
+ "grad_norm": 0.05126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0204,
+ "step": 18779
+ },
+ {
+ "epoch": 60.0,
+ "grad_norm": 0.057861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0274,
+ "step": 18780
+ },
+ {
+ "epoch": 60.00319488817891,
+ "grad_norm": 0.0478515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0305,
+ "step": 18781
+ },
+ {
+ "epoch": 60.00638977635783,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0174,
+ "step": 18782
+ },
+ {
+ "epoch": 60.00958466453674,
+ "grad_norm": 0.0458984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0244,
+ "step": 18783
+ },
+ {
+ "epoch": 60.01277955271566,
+ "grad_norm": 0.052001953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0252,
+ "step": 18784
+ },
+ {
+ "epoch": 60.01597444089457,
+ "grad_norm": 0.05419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0253,
+ "step": 18785
+ },
+ {
+ "epoch": 60.019169329073485,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0249,
+ "step": 18786
+ },
+ {
+ "epoch": 60.022364217252395,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.02,
+ "step": 18787
+ },
+ {
+ "epoch": 60.02555910543131,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0338,
+ "step": 18788
+ },
+ {
+ "epoch": 60.02875399361022,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0157,
+ "step": 18789
+ },
+ {
+ "epoch": 60.031948881789134,
+ "grad_norm": 0.0537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0248,
+ "step": 18790
+ },
+ {
+ "epoch": 60.03514376996805,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0196,
+ "step": 18791
+ },
+ {
+ "epoch": 60.03833865814696,
+ "grad_norm": 0.052978515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0149,
+ "step": 18792
+ },
+ {
+ "epoch": 60.04153354632588,
+ "grad_norm": 0.060791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0243,
+ "step": 18793
+ },
+ {
+ "epoch": 60.04472843450479,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0281,
+ "step": 18794
+ },
+ {
+ "epoch": 60.04792332268371,
+ "grad_norm": 0.05126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0213,
+ "step": 18795
+ },
+ {
+ "epoch": 60.05111821086262,
+ "grad_norm": 0.08056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0251,
+ "step": 18796
+ },
+ {
+ "epoch": 60.054313099041536,
+ "grad_norm": 0.056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0264,
+ "step": 18797
+ },
+ {
+ "epoch": 60.05750798722045,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0164,
+ "step": 18798
+ },
+ {
+ "epoch": 60.06070287539936,
+ "grad_norm": 0.049072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0318,
+ "step": 18799
+ },
+ {
+ "epoch": 60.063897763578275,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0204,
+ "step": 18800
+ },
+ {
+ "epoch": 60.067092651757186,
+ "grad_norm": 0.056396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0231,
+ "step": 18801
+ },
+ {
+ "epoch": 60.0702875399361,
+ "grad_norm": 0.10009765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0308,
+ "step": 18802
+ },
+ {
+ "epoch": 60.073482428115014,
+ "grad_norm": 0.059814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0114,
+ "step": 18803
+ },
+ {
+ "epoch": 60.07667731629393,
+ "grad_norm": 0.11962890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0138,
+ "step": 18804
+ },
+ {
+ "epoch": 60.07987220447284,
+ "grad_norm": 0.054443359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0376,
+ "step": 18805
+ },
+ {
+ "epoch": 60.08306709265176,
+ "grad_norm": 0.1171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0158,
+ "step": 18806
+ },
+ {
+ "epoch": 60.08626198083067,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0226,
+ "step": 18807
+ },
+ {
+ "epoch": 60.08945686900959,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0357,
+ "step": 18808
+ },
+ {
+ "epoch": 60.0926517571885,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0291,
+ "step": 18809
+ },
+ {
+ "epoch": 60.09584664536741,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0318,
+ "step": 18810
+ },
+ {
+ "epoch": 60.09904153354633,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0226,
+ "step": 18811
+ },
+ {
+ "epoch": 60.10223642172524,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0141,
+ "step": 18812
+ },
+ {
+ "epoch": 60.105431309904155,
+ "grad_norm": 0.060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0318,
+ "step": 18813
+ },
+ {
+ "epoch": 60.108626198083066,
+ "grad_norm": 0.05908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.024,
+ "step": 18814
+ },
+ {
+ "epoch": 60.11182108626198,
+ "grad_norm": 0.058837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0202,
+ "step": 18815
+ },
+ {
+ "epoch": 60.115015974440894,
+ "grad_norm": 0.046630859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0319,
+ "step": 18816
+ },
+ {
+ "epoch": 60.11821086261981,
+ "grad_norm": 0.052490234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0247,
+ "step": 18817
+ },
+ {
+ "epoch": 60.12140575079872,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0208,
+ "step": 18818
+ },
+ {
+ "epoch": 60.12460063897763,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0308,
+ "step": 18819
+ },
+ {
+ "epoch": 60.12779552715655,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0222,
+ "step": 18820
+ },
+ {
+ "epoch": 60.13099041533546,
+ "grad_norm": 0.05810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0239,
+ "step": 18821
+ },
+ {
+ "epoch": 60.13418530351438,
+ "grad_norm": 0.060302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0251,
+ "step": 18822
+ },
+ {
+ "epoch": 60.13738019169329,
+ "grad_norm": 0.057373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0169,
+ "step": 18823
+ },
+ {
+ "epoch": 60.14057507987221,
+ "grad_norm": 0.05224609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0264,
+ "step": 18824
+ },
+ {
+ "epoch": 60.14376996805112,
+ "grad_norm": 0.042236328125,
+ "learning_rate": 0.0005,
+ "loss": 0.998,
+ "step": 18825
+ },
+ {
+ "epoch": 60.146964856230035,
+ "grad_norm": 0.046142578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0182,
+ "step": 18826
+ },
+ {
+ "epoch": 60.150159744408946,
+ "grad_norm": 0.05078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0188,
+ "step": 18827
+ },
+ {
+ "epoch": 60.153354632587856,
+ "grad_norm": 0.05419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0235,
+ "step": 18828
+ },
+ {
+ "epoch": 60.156549520766774,
+ "grad_norm": 0.0478515625,
+ "learning_rate": 0.0005,
+ "loss": 1.018,
+ "step": 18829
+ },
+ {
+ "epoch": 60.159744408945684,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0299,
+ "step": 18830
+ },
+ {
+ "epoch": 60.1629392971246,
+ "grad_norm": 0.046630859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0283,
+ "step": 18831
+ },
+ {
+ "epoch": 60.16613418530351,
+ "grad_norm": 0.04931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0231,
+ "step": 18832
+ },
+ {
+ "epoch": 60.16932907348243,
+ "grad_norm": 0.058349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0136,
+ "step": 18833
+ },
+ {
+ "epoch": 60.17252396166134,
+ "grad_norm": 0.051513671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0167,
+ "step": 18834
+ },
+ {
+ "epoch": 60.17571884984026,
+ "grad_norm": 0.04736328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0139,
+ "step": 18835
+ },
+ {
+ "epoch": 60.17891373801917,
+ "grad_norm": 0.056396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0193,
+ "step": 18836
+ },
+ {
+ "epoch": 60.18210862619808,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0322,
+ "step": 18837
+ },
+ {
+ "epoch": 60.185303514377,
+ "grad_norm": 0.047607421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0292,
+ "step": 18838
+ },
+ {
+ "epoch": 60.18849840255591,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0263,
+ "step": 18839
+ },
+ {
+ "epoch": 60.191693290734825,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0241,
+ "step": 18840
+ },
+ {
+ "epoch": 60.194888178913736,
+ "grad_norm": 0.059326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0304,
+ "step": 18841
+ },
+ {
+ "epoch": 60.198083067092654,
+ "grad_norm": 0.05859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0191,
+ "step": 18842
+ },
+ {
+ "epoch": 60.201277955271564,
+ "grad_norm": 0.047607421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0196,
+ "step": 18843
+ },
+ {
+ "epoch": 60.20447284345048,
+ "grad_norm": 0.04833984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0313,
+ "step": 18844
+ },
+ {
+ "epoch": 60.20766773162939,
+ "grad_norm": 0.08203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0226,
+ "step": 18845
+ },
+ {
+ "epoch": 60.21086261980831,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0257,
+ "step": 18846
+ },
+ {
+ "epoch": 60.21405750798722,
+ "grad_norm": 0.08056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0291,
+ "step": 18847
+ },
+ {
+ "epoch": 60.21725239616613,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0266,
+ "step": 18848
+ },
+ {
+ "epoch": 60.22044728434505,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.022,
+ "step": 18849
+ },
+ {
+ "epoch": 60.22364217252396,
+ "grad_norm": 0.080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0268,
+ "step": 18850
+ },
+ {
+ "epoch": 60.22683706070288,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0201,
+ "step": 18851
+ },
+ {
+ "epoch": 60.23003194888179,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0287,
+ "step": 18852
+ },
+ {
+ "epoch": 60.233226837060705,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.021,
+ "step": 18853
+ },
+ {
+ "epoch": 60.236421725239616,
+ "grad_norm": 0.06103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0394,
+ "step": 18854
+ },
+ {
+ "epoch": 60.239616613418534,
+ "grad_norm": 0.08984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0325,
+ "step": 18855
+ },
+ {
+ "epoch": 60.242811501597444,
+ "grad_norm": 0.08544921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0304,
+ "step": 18856
+ },
+ {
+ "epoch": 60.246006389776355,
+ "grad_norm": 0.095703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0204,
+ "step": 18857
+ },
+ {
+ "epoch": 60.24920127795527,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0312,
+ "step": 18858
+ },
+ {
+ "epoch": 60.25239616613418,
+ "grad_norm": 0.080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0259,
+ "step": 18859
+ },
+ {
+ "epoch": 60.2555910543131,
+ "grad_norm": 0.107421875,
+ "learning_rate": 0.0005,
+ "loss": 1.024,
+ "step": 18860
+ },
+ {
+ "epoch": 60.25878594249201,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0325,
+ "step": 18861
+ },
+ {
+ "epoch": 60.26198083067093,
+ "grad_norm": 0.1005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0203,
+ "step": 18862
+ },
+ {
+ "epoch": 60.26517571884984,
+ "grad_norm": 0.05859375,
+ "learning_rate": 0.0005,
+ "loss": 1.021,
+ "step": 18863
+ },
+ {
+ "epoch": 60.26837060702876,
+ "grad_norm": 0.1318359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0224,
+ "step": 18864
+ },
+ {
+ "epoch": 60.27156549520767,
+ "grad_norm": 0.050048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0151,
+ "step": 18865
+ },
+ {
+ "epoch": 60.27476038338658,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0356,
+ "step": 18866
+ },
+ {
+ "epoch": 60.277955271565496,
+ "grad_norm": 0.12451171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0175,
+ "step": 18867
+ },
+ {
+ "epoch": 60.281150159744406,
+ "grad_norm": 0.10205078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0248,
+ "step": 18868
+ },
+ {
+ "epoch": 60.284345047923324,
+ "grad_norm": 0.06201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0284,
+ "step": 18869
+ },
+ {
+ "epoch": 60.287539936102235,
+ "grad_norm": 0.0478515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0243,
+ "step": 18870
+ },
+ {
+ "epoch": 60.29073482428115,
+ "grad_norm": 0.09716796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0246,
+ "step": 18871
+ },
+ {
+ "epoch": 60.29392971246006,
+ "grad_norm": 0.058349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0182,
+ "step": 18872
+ },
+ {
+ "epoch": 60.29712460063898,
+ "grad_norm": 0.0869140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0213,
+ "step": 18873
+ },
+ {
+ "epoch": 60.30031948881789,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0249,
+ "step": 18874
+ },
+ {
+ "epoch": 60.3035143769968,
+ "grad_norm": 0.095703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0233,
+ "step": 18875
+ },
+ {
+ "epoch": 60.30670926517572,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0317,
+ "step": 18876
+ },
+ {
+ "epoch": 60.30990415335463,
+ "grad_norm": 0.11181640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0152,
+ "step": 18877
+ },
+ {
+ "epoch": 60.31309904153355,
+ "grad_norm": 0.08740234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0257,
+ "step": 18878
+ },
+ {
+ "epoch": 60.31629392971246,
+ "grad_norm": 0.10595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0228,
+ "step": 18879
+ },
+ {
+ "epoch": 60.319488817891376,
+ "grad_norm": 0.09130859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0202,
+ "step": 18880
+ },
+ {
+ "epoch": 60.322683706070286,
+ "grad_norm": 0.0830078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0215,
+ "step": 18881
+ },
+ {
+ "epoch": 60.325878594249204,
+ "grad_norm": 0.08154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0351,
+ "step": 18882
+ },
+ {
+ "epoch": 60.329073482428115,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0288,
+ "step": 18883
+ },
+ {
+ "epoch": 60.33226837060703,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0174,
+ "step": 18884
+ },
+ {
+ "epoch": 60.33546325878594,
+ "grad_norm": 0.09326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0193,
+ "step": 18885
+ },
+ {
+ "epoch": 60.33865814696485,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.014,
+ "step": 18886
+ },
+ {
+ "epoch": 60.34185303514377,
+ "grad_norm": 0.08056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0327,
+ "step": 18887
+ },
+ {
+ "epoch": 60.34504792332268,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0271,
+ "step": 18888
+ },
+ {
+ "epoch": 60.3482428115016,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0303,
+ "step": 18889
+ },
+ {
+ "epoch": 60.35143769968051,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0254,
+ "step": 18890
+ },
+ {
+ "epoch": 60.35463258785943,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0323,
+ "step": 18891
+ },
+ {
+ "epoch": 60.35782747603834,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0283,
+ "step": 18892
+ },
+ {
+ "epoch": 60.361022364217256,
+ "grad_norm": 0.05908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0221,
+ "step": 18893
+ },
+ {
+ "epoch": 60.364217252396166,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0206,
+ "step": 18894
+ },
+ {
+ "epoch": 60.36741214057508,
+ "grad_norm": 0.053466796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0145,
+ "step": 18895
+ },
+ {
+ "epoch": 60.370607028753994,
+ "grad_norm": 0.05908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0312,
+ "step": 18896
+ },
+ {
+ "epoch": 60.373801916932905,
+ "grad_norm": 0.057373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0217,
+ "step": 18897
+ },
+ {
+ "epoch": 60.37699680511182,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0296,
+ "step": 18898
+ },
+ {
+ "epoch": 60.38019169329073,
+ "grad_norm": 0.06201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0203,
+ "step": 18899
+ },
+ {
+ "epoch": 60.38338658146965,
+ "grad_norm": 0.049560546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0289,
+ "step": 18900
+ },
+ {
+ "epoch": 60.38658146964856,
+ "grad_norm": 0.048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0251,
+ "step": 18901
+ },
+ {
+ "epoch": 60.38977635782748,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0258,
+ "step": 18902
+ },
+ {
+ "epoch": 60.39297124600639,
+ "grad_norm": 0.11279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0084,
+ "step": 18903
+ },
+ {
+ "epoch": 60.3961661341853,
+ "grad_norm": 0.1552734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0284,
+ "step": 18904
+ },
+ {
+ "epoch": 60.39936102236422,
+ "grad_norm": 0.1162109375,
+ "learning_rate": 0.0005,
+ "loss": 1.025,
+ "step": 18905
+ },
+ {
+ "epoch": 60.40255591054313,
+ "grad_norm": 0.142578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0233,
+ "step": 18906
+ },
+ {
+ "epoch": 60.405750798722046,
+ "grad_norm": 0.126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0277,
+ "step": 18907
+ },
+ {
+ "epoch": 60.40894568690096,
+ "grad_norm": 0.1875,
+ "learning_rate": 0.0005,
+ "loss": 1.0219,
+ "step": 18908
+ },
+ {
+ "epoch": 60.412140575079874,
+ "grad_norm": 0.09765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0378,
+ "step": 18909
+ },
+ {
+ "epoch": 60.415335463258785,
+ "grad_norm": 0.2275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0325,
+ "step": 18910
+ },
+ {
+ "epoch": 60.4185303514377,
+ "grad_norm": 0.054443359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0252,
+ "step": 18911
+ },
+ {
+ "epoch": 60.42172523961661,
+ "grad_norm": 0.1572265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0262,
+ "step": 18912
+ },
+ {
+ "epoch": 60.424920127795524,
+ "grad_norm": 0.10009765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0164,
+ "step": 18913
+ },
+ {
+ "epoch": 60.42811501597444,
+ "grad_norm": 0.1181640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0243,
+ "step": 18914
+ },
+ {
+ "epoch": 60.43130990415335,
+ "grad_norm": 0.1396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0291,
+ "step": 18915
+ },
+ {
+ "epoch": 60.43450479233227,
+ "grad_norm": 0.091796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0401,
+ "step": 18916
+ },
+ {
+ "epoch": 60.43769968051118,
+ "grad_norm": 0.08447265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0354,
+ "step": 18917
+ },
+ {
+ "epoch": 60.4408945686901,
+ "grad_norm": 0.09375,
+ "learning_rate": 0.0005,
+ "loss": 1.0286,
+ "step": 18918
+ },
+ {
+ "epoch": 60.44408945686901,
+ "grad_norm": 0.05419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0277,
+ "step": 18919
+ },
+ {
+ "epoch": 60.447284345047926,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.029,
+ "step": 18920
+ },
+ {
+ "epoch": 60.45047923322684,
+ "grad_norm": 0.052001953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0201,
+ "step": 18921
+ },
+ {
+ "epoch": 60.453674121405754,
+ "grad_norm": 0.08203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0163,
+ "step": 18922
+ },
+ {
+ "epoch": 60.456869009584665,
+ "grad_norm": 0.050048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0157,
+ "step": 18923
+ },
+ {
+ "epoch": 60.460063897763575,
+ "grad_norm": 0.059814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0358,
+ "step": 18924
+ },
+ {
+ "epoch": 60.46325878594249,
+ "grad_norm": 0.057861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0239,
+ "step": 18925
+ },
+ {
+ "epoch": 60.466453674121404,
+ "grad_norm": 0.05615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0271,
+ "step": 18926
+ },
+ {
+ "epoch": 60.46964856230032,
+ "grad_norm": 0.051025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.019,
+ "step": 18927
+ },
+ {
+ "epoch": 60.47284345047923,
+ "grad_norm": 0.052978515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0137,
+ "step": 18928
+ },
+ {
+ "epoch": 60.47603833865815,
+ "grad_norm": 0.05078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0281,
+ "step": 18929
+ },
+ {
+ "epoch": 60.47923322683706,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0229,
+ "step": 18930
+ },
+ {
+ "epoch": 60.48242811501598,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0267,
+ "step": 18931
+ },
+ {
+ "epoch": 60.48562300319489,
+ "grad_norm": 0.044677734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0214,
+ "step": 18932
+ },
+ {
+ "epoch": 60.4888178913738,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0294,
+ "step": 18933
+ },
+ {
+ "epoch": 60.49201277955272,
+ "grad_norm": 0.048095703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0244,
+ "step": 18934
+ },
+ {
+ "epoch": 60.49520766773163,
+ "grad_norm": 0.052978515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0317,
+ "step": 18935
+ },
+ {
+ "epoch": 60.498402555910545,
+ "grad_norm": 0.060302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0325,
+ "step": 18936
+ },
+ {
+ "epoch": 60.501597444089455,
+ "grad_norm": 0.0576171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0319,
+ "step": 18937
+ },
+ {
+ "epoch": 60.50479233226837,
+ "grad_norm": 0.05810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0347,
+ "step": 18938
+ },
+ {
+ "epoch": 60.50798722044728,
+ "grad_norm": 0.053466796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0329,
+ "step": 18939
+ },
+ {
+ "epoch": 60.5111821086262,
+ "grad_norm": 0.0830078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0279,
+ "step": 18940
+ },
+ {
+ "epoch": 60.51437699680511,
+ "grad_norm": 0.055908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0312,
+ "step": 18941
+ },
+ {
+ "epoch": 60.51757188498402,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0231,
+ "step": 18942
+ },
+ {
+ "epoch": 60.52076677316294,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.022,
+ "step": 18943
+ },
+ {
+ "epoch": 60.52396166134185,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0191,
+ "step": 18944
+ },
+ {
+ "epoch": 60.52715654952077,
+ "grad_norm": 0.059814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0181,
+ "step": 18945
+ },
+ {
+ "epoch": 60.53035143769968,
+ "grad_norm": 0.058349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0219,
+ "step": 18946
+ },
+ {
+ "epoch": 60.533546325878596,
+ "grad_norm": 0.062255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0299,
+ "step": 18947
+ },
+ {
+ "epoch": 60.53674121405751,
+ "grad_norm": 0.05517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0117,
+ "step": 18948
+ },
+ {
+ "epoch": 60.539936102236425,
+ "grad_norm": 0.049560546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0125,
+ "step": 18949
+ },
+ {
+ "epoch": 60.543130990415335,
+ "grad_norm": 0.0537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0237,
+ "step": 18950
+ },
+ {
+ "epoch": 60.546325878594246,
+ "grad_norm": 0.08544921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0201,
+ "step": 18951
+ },
+ {
+ "epoch": 60.54952076677316,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.023,
+ "step": 18952
+ },
+ {
+ "epoch": 60.552715654952074,
+ "grad_norm": 0.060302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0239,
+ "step": 18953
+ },
+ {
+ "epoch": 60.55591054313099,
+ "grad_norm": 0.083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0175,
+ "step": 18954
+ },
+ {
+ "epoch": 60.5591054313099,
+ "grad_norm": 0.049560546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0123,
+ "step": 18955
+ },
+ {
+ "epoch": 60.56230031948882,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0248,
+ "step": 18956
+ },
+ {
+ "epoch": 60.56549520766773,
+ "grad_norm": 0.057373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0228,
+ "step": 18957
+ },
+ {
+ "epoch": 60.56869009584665,
+ "grad_norm": 0.06103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0222,
+ "step": 18958
+ },
+ {
+ "epoch": 60.57188498402556,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0159,
+ "step": 18959
+ },
+ {
+ "epoch": 60.575079872204476,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0384,
+ "step": 18960
+ },
+ {
+ "epoch": 60.57827476038339,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0265,
+ "step": 18961
+ },
+ {
+ "epoch": 60.5814696485623,
+ "grad_norm": 0.055908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0191,
+ "step": 18962
+ },
+ {
+ "epoch": 60.584664536741215,
+ "grad_norm": 0.11572265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0221,
+ "step": 18963
+ },
+ {
+ "epoch": 60.587859424920126,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0344,
+ "step": 18964
+ },
+ {
+ "epoch": 60.59105431309904,
+ "grad_norm": 0.060302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.018,
+ "step": 18965
+ },
+ {
+ "epoch": 60.594249201277954,
+ "grad_norm": 0.0576171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0246,
+ "step": 18966
+ },
+ {
+ "epoch": 60.59744408945687,
+ "grad_norm": 0.05126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0227,
+ "step": 18967
+ },
+ {
+ "epoch": 60.60063897763578,
+ "grad_norm": 0.062255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0354,
+ "step": 18968
+ },
+ {
+ "epoch": 60.6038338658147,
+ "grad_norm": 0.050537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0288,
+ "step": 18969
+ },
+ {
+ "epoch": 60.60702875399361,
+ "grad_norm": 0.05859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0139,
+ "step": 18970
+ },
+ {
+ "epoch": 60.61022364217252,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0376,
+ "step": 18971
+ },
+ {
+ "epoch": 60.61341853035144,
+ "grad_norm": 0.0517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0275,
+ "step": 18972
+ },
+ {
+ "epoch": 60.61661341853035,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0171,
+ "step": 18973
+ },
+ {
+ "epoch": 60.61980830670927,
+ "grad_norm": 0.05859375,
+ "learning_rate": 0.0005,
+ "loss": 1.019,
+ "step": 18974
+ },
+ {
+ "epoch": 60.62300319488818,
+ "grad_norm": 0.055908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0233,
+ "step": 18975
+ },
+ {
+ "epoch": 60.626198083067095,
+ "grad_norm": 0.05224609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0269,
+ "step": 18976
+ },
+ {
+ "epoch": 60.629392971246006,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0252,
+ "step": 18977
+ },
+ {
+ "epoch": 60.63258785942492,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0313,
+ "step": 18978
+ },
+ {
+ "epoch": 60.635782747603834,
+ "grad_norm": 0.056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0147,
+ "step": 18979
+ },
+ {
+ "epoch": 60.638977635782744,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0213,
+ "step": 18980
+ },
+ {
+ "epoch": 60.64217252396166,
+ "grad_norm": 0.053466796875,
+ "learning_rate": 0.0005,
+ "loss": 1.026,
+ "step": 18981
+ },
+ {
+ "epoch": 60.64536741214057,
+ "grad_norm": 0.09033203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0092,
+ "step": 18982
+ },
+ {
+ "epoch": 60.64856230031949,
+ "grad_norm": 0.05615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0284,
+ "step": 18983
+ },
+ {
+ "epoch": 60.6517571884984,
+ "grad_norm": 0.061767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0272,
+ "step": 18984
+ },
+ {
+ "epoch": 60.65495207667732,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0319,
+ "step": 18985
+ },
+ {
+ "epoch": 60.65814696485623,
+ "grad_norm": 0.09130859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0231,
+ "step": 18986
+ },
+ {
+ "epoch": 60.66134185303515,
+ "grad_norm": 0.09619140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0151,
+ "step": 18987
+ },
+ {
+ "epoch": 60.66453674121406,
+ "grad_norm": 0.11572265625,
+ "learning_rate": 0.0005,
+ "loss": 1.032,
+ "step": 18988
+ },
+ {
+ "epoch": 60.66773162939297,
+ "grad_norm": 0.09130859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0267,
+ "step": 18989
+ },
+ {
+ "epoch": 60.670926517571885,
+ "grad_norm": 0.05517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0249,
+ "step": 18990
+ },
+ {
+ "epoch": 60.674121405750796,
+ "grad_norm": 0.09130859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0344,
+ "step": 18991
+ },
+ {
+ "epoch": 60.677316293929714,
+ "grad_norm": 0.09228515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0327,
+ "step": 18992
+ },
+ {
+ "epoch": 60.680511182108624,
+ "grad_norm": 0.08837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0277,
+ "step": 18993
+ },
+ {
+ "epoch": 60.68370607028754,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0222,
+ "step": 18994
+ },
+ {
+ "epoch": 60.68690095846645,
+ "grad_norm": 0.08642578125,
+ "learning_rate": 0.0005,
+ "loss": 1.031,
+ "step": 18995
+ },
+ {
+ "epoch": 60.69009584664537,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0311,
+ "step": 18996
+ },
+ {
+ "epoch": 60.69329073482428,
+ "grad_norm": 0.08154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0268,
+ "step": 18997
+ },
+ {
+ "epoch": 60.6964856230032,
+ "grad_norm": 0.08447265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0212,
+ "step": 18998
+ },
+ {
+ "epoch": 60.69968051118211,
+ "grad_norm": 0.1201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0145,
+ "step": 18999
+ },
+ {
+ "epoch": 60.70287539936102,
+ "grad_norm": 0.1220703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0236,
+ "step": 19000
+ },
+ {
+ "epoch": 60.70607028753994,
+ "grad_norm": 0.12109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0246,
+ "step": 19001
+ },
+ {
+ "epoch": 60.70926517571885,
+ "grad_norm": 0.1396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0322,
+ "step": 19002
+ },
+ {
+ "epoch": 60.712460063897765,
+ "grad_norm": 0.146484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0168,
+ "step": 19003
+ },
+ {
+ "epoch": 60.715654952076676,
+ "grad_norm": 0.1416015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0159,
+ "step": 19004
+ },
+ {
+ "epoch": 60.718849840255594,
+ "grad_norm": 0.158203125,
+ "learning_rate": 0.0005,
+ "loss": 1.027,
+ "step": 19005
+ },
+ {
+ "epoch": 60.722044728434504,
+ "grad_norm": 0.1572265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0194,
+ "step": 19006
+ },
+ {
+ "epoch": 60.72523961661342,
+ "grad_norm": 0.1611328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0278,
+ "step": 19007
+ },
+ {
+ "epoch": 60.72843450479233,
+ "grad_norm": 0.1396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0199,
+ "step": 19008
+ },
+ {
+ "epoch": 60.73162939297124,
+ "grad_norm": 0.1259765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0172,
+ "step": 19009
+ },
+ {
+ "epoch": 60.73482428115016,
+ "grad_norm": 0.0849609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0282,
+ "step": 19010
+ },
+ {
+ "epoch": 60.73801916932907,
+ "grad_norm": 0.10791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0166,
+ "step": 19011
+ },
+ {
+ "epoch": 60.74121405750799,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.031,
+ "step": 19012
+ },
+ {
+ "epoch": 60.7444089456869,
+ "grad_norm": 0.10791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0198,
+ "step": 19013
+ },
+ {
+ "epoch": 60.74760383386582,
+ "grad_norm": 0.1015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0259,
+ "step": 19014
+ },
+ {
+ "epoch": 60.75079872204473,
+ "grad_norm": 0.130859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0244,
+ "step": 19015
+ },
+ {
+ "epoch": 60.753993610223645,
+ "grad_norm": 0.09423828125,
+ "learning_rate": 0.0005,
+ "loss": 1.022,
+ "step": 19016
+ },
+ {
+ "epoch": 60.757188498402556,
+ "grad_norm": 0.134765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0249,
+ "step": 19017
+ },
+ {
+ "epoch": 60.760383386581466,
+ "grad_norm": 0.09521484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0314,
+ "step": 19018
+ },
+ {
+ "epoch": 60.763578274760384,
+ "grad_norm": 0.1328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0213,
+ "step": 19019
+ },
+ {
+ "epoch": 60.766773162939295,
+ "grad_norm": 0.109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0317,
+ "step": 19020
+ },
+ {
+ "epoch": 60.76996805111821,
+ "grad_norm": 0.10791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0327,
+ "step": 19021
+ },
+ {
+ "epoch": 60.77316293929712,
+ "grad_norm": 0.0888671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0292,
+ "step": 19022
+ },
+ {
+ "epoch": 60.77635782747604,
+ "grad_norm": 0.0498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0212,
+ "step": 19023
+ },
+ {
+ "epoch": 60.77955271565495,
+ "grad_norm": 0.0859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0233,
+ "step": 19024
+ },
+ {
+ "epoch": 60.78274760383387,
+ "grad_norm": 0.049072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0268,
+ "step": 19025
+ },
+ {
+ "epoch": 60.78594249201278,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.024,
+ "step": 19026
+ },
+ {
+ "epoch": 60.78913738019169,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0211,
+ "step": 19027
+ },
+ {
+ "epoch": 60.79233226837061,
+ "grad_norm": 0.052001953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0174,
+ "step": 19028
+ },
+ {
+ "epoch": 60.79552715654952,
+ "grad_norm": 0.1064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0193,
+ "step": 19029
+ },
+ {
+ "epoch": 60.798722044728436,
+ "grad_norm": 0.048583984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0175,
+ "step": 19030
+ },
+ {
+ "epoch": 60.801916932907346,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0211,
+ "step": 19031
+ },
+ {
+ "epoch": 60.805111821086264,
+ "grad_norm": 0.05712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0304,
+ "step": 19032
+ },
+ {
+ "epoch": 60.808306709265175,
+ "grad_norm": 0.052001953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0089,
+ "step": 19033
+ },
+ {
+ "epoch": 60.81150159744409,
+ "grad_norm": 0.05712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.019,
+ "step": 19034
+ },
+ {
+ "epoch": 60.814696485623,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0412,
+ "step": 19035
+ },
+ {
+ "epoch": 60.81789137380191,
+ "grad_norm": 0.0537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0238,
+ "step": 19036
+ },
+ {
+ "epoch": 60.82108626198083,
+ "grad_norm": 0.060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0238,
+ "step": 19037
+ },
+ {
+ "epoch": 60.82428115015974,
+ "grad_norm": 0.05517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.021,
+ "step": 19038
+ },
+ {
+ "epoch": 60.82747603833866,
+ "grad_norm": 0.08056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0259,
+ "step": 19039
+ },
+ {
+ "epoch": 60.83067092651757,
+ "grad_norm": 0.05224609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0272,
+ "step": 19040
+ },
+ {
+ "epoch": 60.83386581469649,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0303,
+ "step": 19041
+ },
+ {
+ "epoch": 60.8370607028754,
+ "grad_norm": 0.055908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0308,
+ "step": 19042
+ },
+ {
+ "epoch": 60.840255591054316,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0149,
+ "step": 19043
+ },
+ {
+ "epoch": 60.843450479233226,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0215,
+ "step": 19044
+ },
+ {
+ "epoch": 60.846645367412144,
+ "grad_norm": 0.04638671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0239,
+ "step": 19045
+ },
+ {
+ "epoch": 60.849840255591054,
+ "grad_norm": 0.09130859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0203,
+ "step": 19046
+ },
+ {
+ "epoch": 60.853035143769965,
+ "grad_norm": 0.059326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.028,
+ "step": 19047
+ },
+ {
+ "epoch": 60.85623003194888,
+ "grad_norm": 0.0947265625,
+ "learning_rate": 0.0005,
+ "loss": 1.027,
+ "step": 19048
+ },
+ {
+ "epoch": 60.85942492012779,
+ "grad_norm": 0.0576171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0245,
+ "step": 19049
+ },
+ {
+ "epoch": 60.86261980830671,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.022,
+ "step": 19050
+ },
+ {
+ "epoch": 60.86581469648562,
+ "grad_norm": 0.0615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0332,
+ "step": 19051
+ },
+ {
+ "epoch": 60.86900958466454,
+ "grad_norm": 0.0966796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0193,
+ "step": 19052
+ },
+ {
+ "epoch": 60.87220447284345,
+ "grad_norm": 0.061279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0256,
+ "step": 19053
+ },
+ {
+ "epoch": 60.87539936102237,
+ "grad_norm": 0.051513671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0217,
+ "step": 19054
+ },
+ {
+ "epoch": 60.87859424920128,
+ "grad_norm": 0.05908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0384,
+ "step": 19055
+ },
+ {
+ "epoch": 60.88178913738019,
+ "grad_norm": 0.087890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0233,
+ "step": 19056
+ },
+ {
+ "epoch": 60.884984025559106,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0256,
+ "step": 19057
+ },
+ {
+ "epoch": 60.88817891373802,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0289,
+ "step": 19058
+ },
+ {
+ "epoch": 60.891373801916934,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0356,
+ "step": 19059
+ },
+ {
+ "epoch": 60.894568690095845,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0153,
+ "step": 19060
+ },
+ {
+ "epoch": 60.89776357827476,
+ "grad_norm": 0.056884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0289,
+ "step": 19061
+ },
+ {
+ "epoch": 60.90095846645367,
+ "grad_norm": 0.126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0318,
+ "step": 19062
+ },
+ {
+ "epoch": 60.90415335463259,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0209,
+ "step": 19063
+ },
+ {
+ "epoch": 60.9073482428115,
+ "grad_norm": 0.057373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.015,
+ "step": 19064
+ },
+ {
+ "epoch": 60.91054313099041,
+ "grad_norm": 0.0888671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0252,
+ "step": 19065
+ },
+ {
+ "epoch": 60.91373801916933,
+ "grad_norm": 0.11474609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0278,
+ "step": 19066
+ },
+ {
+ "epoch": 60.91693290734824,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0295,
+ "step": 19067
+ },
+ {
+ "epoch": 60.92012779552716,
+ "grad_norm": 0.10888671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0265,
+ "step": 19068
+ },
+ {
+ "epoch": 60.92332268370607,
+ "grad_norm": 0.10400390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0274,
+ "step": 19069
+ },
+ {
+ "epoch": 60.926517571884986,
+ "grad_norm": 0.08203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0257,
+ "step": 19070
+ },
+ {
+ "epoch": 60.9297124600639,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0213,
+ "step": 19071
+ },
+ {
+ "epoch": 60.932907348242814,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0308,
+ "step": 19072
+ },
+ {
+ "epoch": 60.936102236421725,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0323,
+ "step": 19073
+ },
+ {
+ "epoch": 60.93929712460064,
+ "grad_norm": 0.054443359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0074,
+ "step": 19074
+ },
+ {
+ "epoch": 60.94249201277955,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0245,
+ "step": 19075
+ },
+ {
+ "epoch": 60.945686900958464,
+ "grad_norm": 0.0537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0206,
+ "step": 19076
+ },
+ {
+ "epoch": 60.94888178913738,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0268,
+ "step": 19077
+ },
+ {
+ "epoch": 60.95207667731629,
+ "grad_norm": 0.06005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0275,
+ "step": 19078
+ },
+ {
+ "epoch": 60.95527156549521,
+ "grad_norm": 0.060302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0216,
+ "step": 19079
+ },
+ {
+ "epoch": 60.95846645367412,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0178,
+ "step": 19080
+ },
+ {
+ "epoch": 60.96166134185304,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0205,
+ "step": 19081
+ },
+ {
+ "epoch": 60.96485623003195,
+ "grad_norm": 0.058349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0141,
+ "step": 19082
+ },
+ {
+ "epoch": 60.968051118210866,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0214,
+ "step": 19083
+ },
+ {
+ "epoch": 60.97124600638978,
+ "grad_norm": 0.0517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0309,
+ "step": 19084
+ },
+ {
+ "epoch": 60.97444089456869,
+ "grad_norm": 0.05810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0263,
+ "step": 19085
+ },
+ {
+ "epoch": 60.977635782747605,
+ "grad_norm": 0.0556640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0279,
+ "step": 19086
+ },
+ {
+ "epoch": 60.980830670926515,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0199,
+ "step": 19087
+ },
+ {
+ "epoch": 60.98402555910543,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0171,
+ "step": 19088
+ },
+ {
+ "epoch": 60.98722044728434,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0254,
+ "step": 19089
+ },
+ {
+ "epoch": 60.99041533546326,
+ "grad_norm": 0.10107421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0266,
+ "step": 19090
+ },
+ {
+ "epoch": 60.99361022364217,
+ "grad_norm": 0.08837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0237,
+ "step": 19091
+ },
+ {
+ "epoch": 60.99680511182109,
+ "grad_norm": 0.0810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0213,
+ "step": 19092
+ },
+ {
+ "epoch": 61.0,
+ "grad_norm": 0.10302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0207,
+ "step": 19093
+ },
+ {
+ "epoch": 61.00319488817891,
+ "grad_norm": 0.060791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0328,
+ "step": 19094
+ },
+ {
+ "epoch": 61.00638977635783,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0136,
+ "step": 19095
+ },
+ {
+ "epoch": 61.00958466453674,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0243,
+ "step": 19096
+ },
+ {
+ "epoch": 61.01277955271566,
+ "grad_norm": 0.0517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0324,
+ "step": 19097
+ },
+ {
+ "epoch": 61.01597444089457,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0262,
+ "step": 19098
+ },
+ {
+ "epoch": 61.019169329073485,
+ "grad_norm": 0.0458984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0261,
+ "step": 19099
+ },
+ {
+ "epoch": 61.022364217252395,
+ "grad_norm": 0.04833984375,
+ "learning_rate": 0.0005,
+ "loss": 1.035,
+ "step": 19100
+ },
+ {
+ "epoch": 61.02555910543131,
+ "grad_norm": 0.05126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0223,
+ "step": 19101
+ },
+ {
+ "epoch": 61.02875399361022,
+ "grad_norm": 0.0517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0186,
+ "step": 19102
+ },
+ {
+ "epoch": 61.031948881789134,
+ "grad_norm": 0.054443359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0245,
+ "step": 19103
+ },
+ {
+ "epoch": 61.03514376996805,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0251,
+ "step": 19104
+ },
+ {
+ "epoch": 61.03833865814696,
+ "grad_norm": 0.054443359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0217,
+ "step": 19105
+ },
+ {
+ "epoch": 61.04153354632588,
+ "grad_norm": 0.060791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0163,
+ "step": 19106
+ },
+ {
+ "epoch": 61.04472843450479,
+ "grad_norm": 0.058349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0145,
+ "step": 19107
+ },
+ {
+ "epoch": 61.04792332268371,
+ "grad_norm": 0.0546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0293,
+ "step": 19108
+ },
+ {
+ "epoch": 61.05111821086262,
+ "grad_norm": 0.0478515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0192,
+ "step": 19109
+ },
+ {
+ "epoch": 61.054313099041536,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0233,
+ "step": 19110
+ },
+ {
+ "epoch": 61.05750798722045,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0298,
+ "step": 19111
+ },
+ {
+ "epoch": 61.06070287539936,
+ "grad_norm": 0.060302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0272,
+ "step": 19112
+ },
+ {
+ "epoch": 61.063897763578275,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0219,
+ "step": 19113
+ },
+ {
+ "epoch": 61.067092651757186,
+ "grad_norm": 0.0537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0242,
+ "step": 19114
+ },
+ {
+ "epoch": 61.0702875399361,
+ "grad_norm": 0.1201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0132,
+ "step": 19115
+ },
+ {
+ "epoch": 61.073482428115014,
+ "grad_norm": 0.08203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0285,
+ "step": 19116
+ },
+ {
+ "epoch": 61.07667731629393,
+ "grad_norm": 0.07861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0189,
+ "step": 19117
+ },
+ {
+ "epoch": 61.07987220447284,
+ "grad_norm": 0.09716796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0215,
+ "step": 19118
+ },
+ {
+ "epoch": 61.08306709265176,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0198,
+ "step": 19119
+ },
+ {
+ "epoch": 61.08626198083067,
+ "grad_norm": 0.10693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0295,
+ "step": 19120
+ },
+ {
+ "epoch": 61.08945686900959,
+ "grad_norm": 0.12109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0314,
+ "step": 19121
+ },
+ {
+ "epoch": 61.0926517571885,
+ "grad_norm": 0.11669921875,
+ "learning_rate": 0.0005,
+ "loss": 1.019,
+ "step": 19122
+ },
+ {
+ "epoch": 61.09584664536741,
+ "grad_norm": 0.1435546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0322,
+ "step": 19123
+ },
+ {
+ "epoch": 61.09904153354633,
+ "grad_norm": 0.09814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0193,
+ "step": 19124
+ },
+ {
+ "epoch": 61.10223642172524,
+ "grad_norm": 0.1259765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0227,
+ "step": 19125
+ },
+ {
+ "epoch": 61.105431309904155,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.031,
+ "step": 19126
+ },
+ {
+ "epoch": 61.108626198083066,
+ "grad_norm": 0.05419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0188,
+ "step": 19127
+ },
+ {
+ "epoch": 61.11182108626198,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0247,
+ "step": 19128
+ },
+ {
+ "epoch": 61.115015974440894,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0327,
+ "step": 19129
+ },
+ {
+ "epoch": 61.11821086261981,
+ "grad_norm": 0.083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0266,
+ "step": 19130
+ },
+ {
+ "epoch": 61.12140575079872,
+ "grad_norm": 0.058349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0346,
+ "step": 19131
+ },
+ {
+ "epoch": 61.12460063897763,
+ "grad_norm": 0.059326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0295,
+ "step": 19132
+ },
+ {
+ "epoch": 61.12779552715655,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0291,
+ "step": 19133
+ },
+ {
+ "epoch": 61.13099041533546,
+ "grad_norm": 0.05859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0252,
+ "step": 19134
+ },
+ {
+ "epoch": 61.13418530351438,
+ "grad_norm": 0.058349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0203,
+ "step": 19135
+ },
+ {
+ "epoch": 61.13738019169329,
+ "grad_norm": 0.058349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0238,
+ "step": 19136
+ },
+ {
+ "epoch": 61.14057507987221,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0324,
+ "step": 19137
+ },
+ {
+ "epoch": 61.14376996805112,
+ "grad_norm": 0.0537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0331,
+ "step": 19138
+ },
+ {
+ "epoch": 61.146964856230035,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0167,
+ "step": 19139
+ },
+ {
+ "epoch": 61.150159744408946,
+ "grad_norm": 0.047607421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0282,
+ "step": 19140
+ },
+ {
+ "epoch": 61.153354632587856,
+ "grad_norm": 0.047607421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0246,
+ "step": 19141
+ },
+ {
+ "epoch": 61.156549520766774,
+ "grad_norm": 0.0556640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0214,
+ "step": 19142
+ },
+ {
+ "epoch": 61.159744408945684,
+ "grad_norm": 0.05712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0249,
+ "step": 19143
+ },
+ {
+ "epoch": 61.1629392971246,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0135,
+ "step": 19144
+ },
+ {
+ "epoch": 61.16613418530351,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0306,
+ "step": 19145
+ },
+ {
+ "epoch": 61.16932907348243,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0237,
+ "step": 19146
+ },
+ {
+ "epoch": 61.17252396166134,
+ "grad_norm": 0.0888671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0195,
+ "step": 19147
+ },
+ {
+ "epoch": 61.17571884984026,
+ "grad_norm": 0.048583984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0189,
+ "step": 19148
+ },
+ {
+ "epoch": 61.17891373801917,
+ "grad_norm": 0.05322265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0241,
+ "step": 19149
+ },
+ {
+ "epoch": 61.18210862619808,
+ "grad_norm": 0.053466796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0295,
+ "step": 19150
+ },
+ {
+ "epoch": 61.185303514377,
+ "grad_norm": 0.05126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0312,
+ "step": 19151
+ },
+ {
+ "epoch": 61.18849840255591,
+ "grad_norm": 0.08642578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0229,
+ "step": 19152
+ },
+ {
+ "epoch": 61.191693290734825,
+ "grad_norm": 0.060791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0263,
+ "step": 19153
+ },
+ {
+ "epoch": 61.194888178913736,
+ "grad_norm": 0.05615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0164,
+ "step": 19154
+ },
+ {
+ "epoch": 61.198083067092654,
+ "grad_norm": 0.1044921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0194,
+ "step": 19155
+ },
+ {
+ "epoch": 61.201277955271564,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0165,
+ "step": 19156
+ },
+ {
+ "epoch": 61.20447284345048,
+ "grad_norm": 0.0546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0325,
+ "step": 19157
+ },
+ {
+ "epoch": 61.20766773162939,
+ "grad_norm": 0.06103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0143,
+ "step": 19158
+ },
+ {
+ "epoch": 61.21086261980831,
+ "grad_norm": 0.048095703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0234,
+ "step": 19159
+ },
+ {
+ "epoch": 61.21405750798722,
+ "grad_norm": 0.05712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0275,
+ "step": 19160
+ },
+ {
+ "epoch": 61.21725239616613,
+ "grad_norm": 0.06005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0275,
+ "step": 19161
+ },
+ {
+ "epoch": 61.22044728434505,
+ "grad_norm": 0.059326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0292,
+ "step": 19162
+ },
+ {
+ "epoch": 61.22364217252396,
+ "grad_norm": 0.04638671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0216,
+ "step": 19163
+ },
+ {
+ "epoch": 61.22683706070288,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0185,
+ "step": 19164
+ },
+ {
+ "epoch": 61.23003194888179,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0266,
+ "step": 19165
+ },
+ {
+ "epoch": 61.233226837060705,
+ "grad_norm": 0.048095703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0172,
+ "step": 19166
+ },
+ {
+ "epoch": 61.236421725239616,
+ "grad_norm": 0.08251953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0115,
+ "step": 19167
+ },
+ {
+ "epoch": 61.239616613418534,
+ "grad_norm": 0.049072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0245,
+ "step": 19168
+ },
+ {
+ "epoch": 61.242811501597444,
+ "grad_norm": 0.054931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0137,
+ "step": 19169
+ },
+ {
+ "epoch": 61.246006389776355,
+ "grad_norm": 0.04345703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0183,
+ "step": 19170
+ },
+ {
+ "epoch": 61.24920127795527,
+ "grad_norm": 0.04541015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0271,
+ "step": 19171
+ },
+ {
+ "epoch": 61.25239616613418,
+ "grad_norm": 0.044921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0228,
+ "step": 19172
+ },
+ {
+ "epoch": 61.2555910543131,
+ "grad_norm": 0.058837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0293,
+ "step": 19173
+ },
+ {
+ "epoch": 61.25878594249201,
+ "grad_norm": 0.044189453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0298,
+ "step": 19174
+ },
+ {
+ "epoch": 61.26198083067093,
+ "grad_norm": 0.08056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0111,
+ "step": 19175
+ },
+ {
+ "epoch": 61.26517571884984,
+ "grad_norm": 0.05419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.013,
+ "step": 19176
+ },
+ {
+ "epoch": 61.26837060702876,
+ "grad_norm": 0.05224609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0249,
+ "step": 19177
+ },
+ {
+ "epoch": 61.27156549520767,
+ "grad_norm": 0.1162109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0219,
+ "step": 19178
+ },
+ {
+ "epoch": 61.27476038338658,
+ "grad_norm": 0.0546875,
+ "learning_rate": 0.0005,
+ "loss": 1.043,
+ "step": 19179
+ },
+ {
+ "epoch": 61.277955271565496,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0111,
+ "step": 19180
+ },
+ {
+ "epoch": 61.281150159744406,
+ "grad_norm": 0.058837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.028,
+ "step": 19181
+ },
+ {
+ "epoch": 61.284345047923324,
+ "grad_norm": 0.05615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0133,
+ "step": 19182
+ },
+ {
+ "epoch": 61.287539936102235,
+ "grad_norm": 0.06005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0402,
+ "step": 19183
+ },
+ {
+ "epoch": 61.29073482428115,
+ "grad_norm": 0.0908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.021,
+ "step": 19184
+ },
+ {
+ "epoch": 61.29392971246006,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0265,
+ "step": 19185
+ },
+ {
+ "epoch": 61.29712460063898,
+ "grad_norm": 0.046630859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0361,
+ "step": 19186
+ },
+ {
+ "epoch": 61.30031948881789,
+ "grad_norm": 0.09521484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0168,
+ "step": 19187
+ },
+ {
+ "epoch": 61.3035143769968,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0242,
+ "step": 19188
+ },
+ {
+ "epoch": 61.30670926517572,
+ "grad_norm": 0.087890625,
+ "learning_rate": 0.0005,
+ "loss": 1.024,
+ "step": 19189
+ },
+ {
+ "epoch": 61.30990415335463,
+ "grad_norm": 0.07861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.034,
+ "step": 19190
+ },
+ {
+ "epoch": 61.31309904153355,
+ "grad_norm": 0.1025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0179,
+ "step": 19191
+ },
+ {
+ "epoch": 61.31629392971246,
+ "grad_norm": 0.05859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0278,
+ "step": 19192
+ },
+ {
+ "epoch": 61.319488817891376,
+ "grad_norm": 0.10986328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0193,
+ "step": 19193
+ },
+ {
+ "epoch": 61.322683706070286,
+ "grad_norm": 0.09814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0151,
+ "step": 19194
+ },
+ {
+ "epoch": 61.325878594249204,
+ "grad_norm": 0.138671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0287,
+ "step": 19195
+ },
+ {
+ "epoch": 61.329073482428115,
+ "grad_norm": 0.130859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0194,
+ "step": 19196
+ },
+ {
+ "epoch": 61.33226837060703,
+ "grad_norm": 0.0966796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0342,
+ "step": 19197
+ },
+ {
+ "epoch": 61.33546325878594,
+ "grad_norm": 0.09521484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0233,
+ "step": 19198
+ },
+ {
+ "epoch": 61.33865814696485,
+ "grad_norm": 0.12255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0234,
+ "step": 19199
+ },
+ {
+ "epoch": 61.34185303514377,
+ "grad_norm": 0.1328125,
+ "learning_rate": 0.0005,
+ "loss": 1.031,
+ "step": 19200
+ },
+ {
+ "epoch": 61.34504792332268,
+ "grad_norm": 0.11181640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0296,
+ "step": 19201
+ },
+ {
+ "epoch": 61.3482428115016,
+ "grad_norm": 0.0595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0279,
+ "step": 19202
+ },
+ {
+ "epoch": 61.35143769968051,
+ "grad_norm": 0.115234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0278,
+ "step": 19203
+ },
+ {
+ "epoch": 61.35463258785943,
+ "grad_norm": 0.11376953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0259,
+ "step": 19204
+ },
+ {
+ "epoch": 61.35782747603834,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0203,
+ "step": 19205
+ },
+ {
+ "epoch": 61.361022364217256,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0219,
+ "step": 19206
+ },
+ {
+ "epoch": 61.364217252396166,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0257,
+ "step": 19207
+ },
+ {
+ "epoch": 61.36741214057508,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0305,
+ "step": 19208
+ },
+ {
+ "epoch": 61.370607028753994,
+ "grad_norm": 0.08837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0245,
+ "step": 19209
+ },
+ {
+ "epoch": 61.373801916932905,
+ "grad_norm": 0.10986328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0258,
+ "step": 19210
+ },
+ {
+ "epoch": 61.37699680511182,
+ "grad_norm": 0.1171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0267,
+ "step": 19211
+ },
+ {
+ "epoch": 61.38019169329073,
+ "grad_norm": 0.060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0261,
+ "step": 19212
+ },
+ {
+ "epoch": 61.38338658146965,
+ "grad_norm": 0.08056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0215,
+ "step": 19213
+ },
+ {
+ "epoch": 61.38658146964856,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0239,
+ "step": 19214
+ },
+ {
+ "epoch": 61.38977635782748,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0235,
+ "step": 19215
+ },
+ {
+ "epoch": 61.39297124600639,
+ "grad_norm": 0.06103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.018,
+ "step": 19216
+ },
+ {
+ "epoch": 61.3961661341853,
+ "grad_norm": 0.051025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0294,
+ "step": 19217
+ },
+ {
+ "epoch": 61.39936102236422,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0403,
+ "step": 19218
+ },
+ {
+ "epoch": 61.40255591054313,
+ "grad_norm": 0.046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0279,
+ "step": 19219
+ },
+ {
+ "epoch": 61.405750798722046,
+ "grad_norm": 0.055908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.034,
+ "step": 19220
+ },
+ {
+ "epoch": 61.40894568690096,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0254,
+ "step": 19221
+ },
+ {
+ "epoch": 61.412140575079874,
+ "grad_norm": 0.0888671875,
+ "learning_rate": 0.0005,
+ "loss": 1.035,
+ "step": 19222
+ },
+ {
+ "epoch": 61.415335463258785,
+ "grad_norm": 0.06005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0254,
+ "step": 19223
+ },
+ {
+ "epoch": 61.4185303514377,
+ "grad_norm": 0.04541015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0165,
+ "step": 19224
+ },
+ {
+ "epoch": 61.42172523961661,
+ "grad_norm": 0.053955078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0202,
+ "step": 19225
+ },
+ {
+ "epoch": 61.424920127795524,
+ "grad_norm": 0.0458984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0252,
+ "step": 19226
+ },
+ {
+ "epoch": 61.42811501597444,
+ "grad_norm": 0.042236328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0159,
+ "step": 19227
+ },
+ {
+ "epoch": 61.43130990415335,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0256,
+ "step": 19228
+ },
+ {
+ "epoch": 61.43450479233227,
+ "grad_norm": 0.049560546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0173,
+ "step": 19229
+ },
+ {
+ "epoch": 61.43769968051118,
+ "grad_norm": 0.0576171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0155,
+ "step": 19230
+ },
+ {
+ "epoch": 61.4408945686901,
+ "grad_norm": 0.061767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0131,
+ "step": 19231
+ },
+ {
+ "epoch": 61.44408945686901,
+ "grad_norm": 0.049072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0302,
+ "step": 19232
+ },
+ {
+ "epoch": 61.447284345047926,
+ "grad_norm": 0.08447265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0238,
+ "step": 19233
+ },
+ {
+ "epoch": 61.45047923322684,
+ "grad_norm": 0.049072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0262,
+ "step": 19234
+ },
+ {
+ "epoch": 61.453674121405754,
+ "grad_norm": 0.058349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0211,
+ "step": 19235
+ },
+ {
+ "epoch": 61.456869009584665,
+ "grad_norm": 0.058837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.031,
+ "step": 19236
+ },
+ {
+ "epoch": 61.460063897763575,
+ "grad_norm": 0.0537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0171,
+ "step": 19237
+ },
+ {
+ "epoch": 61.46325878594249,
+ "grad_norm": 0.056884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0219,
+ "step": 19238
+ },
+ {
+ "epoch": 61.466453674121404,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.009,
+ "step": 19239
+ },
+ {
+ "epoch": 61.46964856230032,
+ "grad_norm": 0.05517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0319,
+ "step": 19240
+ },
+ {
+ "epoch": 61.47284345047923,
+ "grad_norm": 0.044189453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0264,
+ "step": 19241
+ },
+ {
+ "epoch": 61.47603833865815,
+ "grad_norm": 0.080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0195,
+ "step": 19242
+ },
+ {
+ "epoch": 61.47923322683706,
+ "grad_norm": 0.05712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0282,
+ "step": 19243
+ },
+ {
+ "epoch": 61.48242811501598,
+ "grad_norm": 0.08203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0219,
+ "step": 19244
+ },
+ {
+ "epoch": 61.48562300319489,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0288,
+ "step": 19245
+ },
+ {
+ "epoch": 61.4888178913738,
+ "grad_norm": 0.046630859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0315,
+ "step": 19246
+ },
+ {
+ "epoch": 61.49201277955272,
+ "grad_norm": 0.08349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0138,
+ "step": 19247
+ },
+ {
+ "epoch": 61.49520766773163,
+ "grad_norm": 0.056396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0348,
+ "step": 19248
+ },
+ {
+ "epoch": 61.498402555910545,
+ "grad_norm": 0.051513671875,
+ "learning_rate": 0.0005,
+ "loss": 1.022,
+ "step": 19249
+ },
+ {
+ "epoch": 61.501597444089455,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0134,
+ "step": 19250
+ },
+ {
+ "epoch": 61.50479233226837,
+ "grad_norm": 0.056396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0209,
+ "step": 19251
+ },
+ {
+ "epoch": 61.50798722044728,
+ "grad_norm": 0.06201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0274,
+ "step": 19252
+ },
+ {
+ "epoch": 61.5111821086262,
+ "grad_norm": 0.044189453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0199,
+ "step": 19253
+ },
+ {
+ "epoch": 61.51437699680511,
+ "grad_norm": 0.095703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0212,
+ "step": 19254
+ },
+ {
+ "epoch": 61.51757188498402,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0258,
+ "step": 19255
+ },
+ {
+ "epoch": 61.52076677316294,
+ "grad_norm": 0.095703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0373,
+ "step": 19256
+ },
+ {
+ "epoch": 61.52396166134185,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0184,
+ "step": 19257
+ },
+ {
+ "epoch": 61.52715654952077,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0169,
+ "step": 19258
+ },
+ {
+ "epoch": 61.53035143769968,
+ "grad_norm": 0.0615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0447,
+ "step": 19259
+ },
+ {
+ "epoch": 61.533546325878596,
+ "grad_norm": 0.07861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0211,
+ "step": 19260
+ },
+ {
+ "epoch": 61.53674121405751,
+ "grad_norm": 0.055908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0257,
+ "step": 19261
+ },
+ {
+ "epoch": 61.539936102236425,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0195,
+ "step": 19262
+ },
+ {
+ "epoch": 61.543130990415335,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0157,
+ "step": 19263
+ },
+ {
+ "epoch": 61.546325878594246,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.023,
+ "step": 19264
+ },
+ {
+ "epoch": 61.54952076677316,
+ "grad_norm": 0.06103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0229,
+ "step": 19265
+ },
+ {
+ "epoch": 61.552715654952074,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0249,
+ "step": 19266
+ },
+ {
+ "epoch": 61.55591054313099,
+ "grad_norm": 0.044921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0241,
+ "step": 19267
+ },
+ {
+ "epoch": 61.5591054313099,
+ "grad_norm": 0.056884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0266,
+ "step": 19268
+ },
+ {
+ "epoch": 61.56230031948882,
+ "grad_norm": 0.050048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0214,
+ "step": 19269
+ },
+ {
+ "epoch": 61.56549520766773,
+ "grad_norm": 0.0576171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0247,
+ "step": 19270
+ },
+ {
+ "epoch": 61.56869009584665,
+ "grad_norm": 0.047607421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0144,
+ "step": 19271
+ },
+ {
+ "epoch": 61.57188498402556,
+ "grad_norm": 0.05126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0172,
+ "step": 19272
+ },
+ {
+ "epoch": 61.575079872204476,
+ "grad_norm": 0.04931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0342,
+ "step": 19273
+ },
+ {
+ "epoch": 61.57827476038339,
+ "grad_norm": 0.0556640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0185,
+ "step": 19274
+ },
+ {
+ "epoch": 61.5814696485623,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0284,
+ "step": 19275
+ },
+ {
+ "epoch": 61.584664536741215,
+ "grad_norm": 0.04736328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0287,
+ "step": 19276
+ },
+ {
+ "epoch": 61.587859424920126,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0256,
+ "step": 19277
+ },
+ {
+ "epoch": 61.59105431309904,
+ "grad_norm": 0.05078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0236,
+ "step": 19278
+ },
+ {
+ "epoch": 61.594249201277954,
+ "grad_norm": 0.10791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0228,
+ "step": 19279
+ },
+ {
+ "epoch": 61.59744408945687,
+ "grad_norm": 0.046142578125,
+ "learning_rate": 0.0005,
+ "loss": 1.026,
+ "step": 19280
+ },
+ {
+ "epoch": 61.60063897763578,
+ "grad_norm": 0.11279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0259,
+ "step": 19281
+ },
+ {
+ "epoch": 61.6038338658147,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0217,
+ "step": 19282
+ },
+ {
+ "epoch": 61.60702875399361,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0144,
+ "step": 19283
+ },
+ {
+ "epoch": 61.61022364217252,
+ "grad_norm": 0.0810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0256,
+ "step": 19284
+ },
+ {
+ "epoch": 61.61341853035144,
+ "grad_norm": 0.083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0296,
+ "step": 19285
+ },
+ {
+ "epoch": 61.61661341853035,
+ "grad_norm": 0.054931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0331,
+ "step": 19286
+ },
+ {
+ "epoch": 61.61980830670927,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0204,
+ "step": 19287
+ },
+ {
+ "epoch": 61.62300319488818,
+ "grad_norm": 0.10107421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0177,
+ "step": 19288
+ },
+ {
+ "epoch": 61.626198083067095,
+ "grad_norm": 0.0869140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0241,
+ "step": 19289
+ },
+ {
+ "epoch": 61.629392971246006,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0223,
+ "step": 19290
+ },
+ {
+ "epoch": 61.63258785942492,
+ "grad_norm": 0.05810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0293,
+ "step": 19291
+ },
+ {
+ "epoch": 61.635782747603834,
+ "grad_norm": 0.08740234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0246,
+ "step": 19292
+ },
+ {
+ "epoch": 61.638977635782744,
+ "grad_norm": 0.08544921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0153,
+ "step": 19293
+ },
+ {
+ "epoch": 61.64217252396166,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0179,
+ "step": 19294
+ },
+ {
+ "epoch": 61.64536741214057,
+ "grad_norm": 0.0517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0283,
+ "step": 19295
+ },
+ {
+ "epoch": 61.64856230031949,
+ "grad_norm": 0.08349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0218,
+ "step": 19296
+ },
+ {
+ "epoch": 61.6517571884984,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0158,
+ "step": 19297
+ },
+ {
+ "epoch": 61.65495207667732,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0285,
+ "step": 19298
+ },
+ {
+ "epoch": 61.65814696485623,
+ "grad_norm": 0.055419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0223,
+ "step": 19299
+ },
+ {
+ "epoch": 61.66134185303515,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0271,
+ "step": 19300
+ },
+ {
+ "epoch": 61.66453674121406,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.015,
+ "step": 19301
+ },
+ {
+ "epoch": 61.66773162939297,
+ "grad_norm": 0.05078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0201,
+ "step": 19302
+ },
+ {
+ "epoch": 61.670926517571885,
+ "grad_norm": 0.046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0336,
+ "step": 19303
+ },
+ {
+ "epoch": 61.674121405750796,
+ "grad_norm": 0.05908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.022,
+ "step": 19304
+ },
+ {
+ "epoch": 61.677316293929714,
+ "grad_norm": 0.061279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0306,
+ "step": 19305
+ },
+ {
+ "epoch": 61.680511182108624,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0207,
+ "step": 19306
+ },
+ {
+ "epoch": 61.68370607028754,
+ "grad_norm": 0.04736328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0306,
+ "step": 19307
+ },
+ {
+ "epoch": 61.68690095846645,
+ "grad_norm": 0.05419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.028,
+ "step": 19308
+ },
+ {
+ "epoch": 61.69009584664537,
+ "grad_norm": 0.060302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0282,
+ "step": 19309
+ },
+ {
+ "epoch": 61.69329073482428,
+ "grad_norm": 0.09423828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0261,
+ "step": 19310
+ },
+ {
+ "epoch": 61.6964856230032,
+ "grad_norm": 0.0869140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0267,
+ "step": 19311
+ },
+ {
+ "epoch": 61.69968051118211,
+ "grad_norm": 0.0546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0342,
+ "step": 19312
+ },
+ {
+ "epoch": 61.70287539936102,
+ "grad_norm": 0.051025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0247,
+ "step": 19313
+ },
+ {
+ "epoch": 61.70607028753994,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0251,
+ "step": 19314
+ },
+ {
+ "epoch": 61.70926517571885,
+ "grad_norm": 0.05712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0432,
+ "step": 19315
+ },
+ {
+ "epoch": 61.712460063897765,
+ "grad_norm": 0.0576171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0211,
+ "step": 19316
+ },
+ {
+ "epoch": 61.715654952076676,
+ "grad_norm": 0.057861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0234,
+ "step": 19317
+ },
+ {
+ "epoch": 61.718849840255594,
+ "grad_norm": 0.057861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0165,
+ "step": 19318
+ },
+ {
+ "epoch": 61.722044728434504,
+ "grad_norm": 0.05322265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0273,
+ "step": 19319
+ },
+ {
+ "epoch": 61.72523961661342,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0187,
+ "step": 19320
+ },
+ {
+ "epoch": 61.72843450479233,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0246,
+ "step": 19321
+ },
+ {
+ "epoch": 61.73162939297124,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0135,
+ "step": 19322
+ },
+ {
+ "epoch": 61.73482428115016,
+ "grad_norm": 0.0595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0366,
+ "step": 19323
+ },
+ {
+ "epoch": 61.73801916932907,
+ "grad_norm": 0.048095703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0202,
+ "step": 19324
+ },
+ {
+ "epoch": 61.74121405750799,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0217,
+ "step": 19325
+ },
+ {
+ "epoch": 61.7444089456869,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0245,
+ "step": 19326
+ },
+ {
+ "epoch": 61.74760383386582,
+ "grad_norm": 0.049072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0298,
+ "step": 19327
+ },
+ {
+ "epoch": 61.75079872204473,
+ "grad_norm": 0.1044921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0187,
+ "step": 19328
+ },
+ {
+ "epoch": 61.753993610223645,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0272,
+ "step": 19329
+ },
+ {
+ "epoch": 61.757188498402556,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0271,
+ "step": 19330
+ },
+ {
+ "epoch": 61.760383386581466,
+ "grad_norm": 0.0537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0228,
+ "step": 19331
+ },
+ {
+ "epoch": 61.763578274760384,
+ "grad_norm": 0.1103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.032,
+ "step": 19332
+ },
+ {
+ "epoch": 61.766773162939295,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0208,
+ "step": 19333
+ },
+ {
+ "epoch": 61.76996805111821,
+ "grad_norm": 0.10302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0319,
+ "step": 19334
+ },
+ {
+ "epoch": 61.77316293929712,
+ "grad_norm": 0.08056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0229,
+ "step": 19335
+ },
+ {
+ "epoch": 61.77635782747604,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0188,
+ "step": 19336
+ },
+ {
+ "epoch": 61.77955271565495,
+ "grad_norm": 0.08203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0206,
+ "step": 19337
+ },
+ {
+ "epoch": 61.78274760383387,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0312,
+ "step": 19338
+ },
+ {
+ "epoch": 61.78594249201278,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0187,
+ "step": 19339
+ },
+ {
+ "epoch": 61.78913738019169,
+ "grad_norm": 0.061279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0204,
+ "step": 19340
+ },
+ {
+ "epoch": 61.79233226837061,
+ "grad_norm": 0.08251953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0254,
+ "step": 19341
+ },
+ {
+ "epoch": 61.79552715654952,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0255,
+ "step": 19342
+ },
+ {
+ "epoch": 61.798722044728436,
+ "grad_norm": 0.123046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0115,
+ "step": 19343
+ },
+ {
+ "epoch": 61.801916932907346,
+ "grad_norm": 0.0849609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0207,
+ "step": 19344
+ },
+ {
+ "epoch": 61.805111821086264,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0208,
+ "step": 19345
+ },
+ {
+ "epoch": 61.808306709265175,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0214,
+ "step": 19346
+ },
+ {
+ "epoch": 61.81150159744409,
+ "grad_norm": 0.051025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0244,
+ "step": 19347
+ },
+ {
+ "epoch": 61.814696485623,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0244,
+ "step": 19348
+ },
+ {
+ "epoch": 61.81789137380191,
+ "grad_norm": 0.0517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0319,
+ "step": 19349
+ },
+ {
+ "epoch": 61.82108626198083,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.023,
+ "step": 19350
+ },
+ {
+ "epoch": 61.82428115015974,
+ "grad_norm": 0.048583984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0258,
+ "step": 19351
+ },
+ {
+ "epoch": 61.82747603833866,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0164,
+ "step": 19352
+ },
+ {
+ "epoch": 61.83067092651757,
+ "grad_norm": 0.052490234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0318,
+ "step": 19353
+ },
+ {
+ "epoch": 61.83386581469649,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0272,
+ "step": 19354
+ },
+ {
+ "epoch": 61.8370607028754,
+ "grad_norm": 0.05517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0103,
+ "step": 19355
+ },
+ {
+ "epoch": 61.840255591054316,
+ "grad_norm": 0.054931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0192,
+ "step": 19356
+ },
+ {
+ "epoch": 61.843450479233226,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0166,
+ "step": 19357
+ },
+ {
+ "epoch": 61.846645367412144,
+ "grad_norm": 0.0478515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0122,
+ "step": 19358
+ },
+ {
+ "epoch": 61.849840255591054,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0304,
+ "step": 19359
+ },
+ {
+ "epoch": 61.853035143769965,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0233,
+ "step": 19360
+ },
+ {
+ "epoch": 61.85623003194888,
+ "grad_norm": 0.060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0335,
+ "step": 19361
+ },
+ {
+ "epoch": 61.85942492012779,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0234,
+ "step": 19362
+ },
+ {
+ "epoch": 61.86261980830671,
+ "grad_norm": 0.054931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0259,
+ "step": 19363
+ },
+ {
+ "epoch": 61.86581469648562,
+ "grad_norm": 0.052734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0189,
+ "step": 19364
+ },
+ {
+ "epoch": 61.86900958466454,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0244,
+ "step": 19365
+ },
+ {
+ "epoch": 61.87220447284345,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0269,
+ "step": 19366
+ },
+ {
+ "epoch": 61.87539936102237,
+ "grad_norm": 0.0908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0138,
+ "step": 19367
+ },
+ {
+ "epoch": 61.87859424920128,
+ "grad_norm": 0.099609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0313,
+ "step": 19368
+ },
+ {
+ "epoch": 61.88178913738019,
+ "grad_norm": 0.115234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0196,
+ "step": 19369
+ },
+ {
+ "epoch": 61.884984025559106,
+ "grad_norm": 0.09423828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0301,
+ "step": 19370
+ },
+ {
+ "epoch": 61.88817891373802,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0326,
+ "step": 19371
+ },
+ {
+ "epoch": 61.891373801916934,
+ "grad_norm": 0.0849609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0264,
+ "step": 19372
+ },
+ {
+ "epoch": 61.894568690095845,
+ "grad_norm": 0.049560546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0237,
+ "step": 19373
+ },
+ {
+ "epoch": 61.89776357827476,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0152,
+ "step": 19374
+ },
+ {
+ "epoch": 61.90095846645367,
+ "grad_norm": 0.056396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0221,
+ "step": 19375
+ },
+ {
+ "epoch": 61.90415335463259,
+ "grad_norm": 0.05908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0286,
+ "step": 19376
+ },
+ {
+ "epoch": 61.9073482428115,
+ "grad_norm": 0.052490234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0231,
+ "step": 19377
+ },
+ {
+ "epoch": 61.91054313099041,
+ "grad_norm": 0.05419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0336,
+ "step": 19378
+ },
+ {
+ "epoch": 61.91373801916933,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0216,
+ "step": 19379
+ },
+ {
+ "epoch": 61.91693290734824,
+ "grad_norm": 0.05419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0308,
+ "step": 19380
+ },
+ {
+ "epoch": 61.92012779552716,
+ "grad_norm": 0.05712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0279,
+ "step": 19381
+ },
+ {
+ "epoch": 61.92332268370607,
+ "grad_norm": 0.0478515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0167,
+ "step": 19382
+ },
+ {
+ "epoch": 61.926517571884986,
+ "grad_norm": 0.053955078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0215,
+ "step": 19383
+ },
+ {
+ "epoch": 61.9297124600639,
+ "grad_norm": 0.0556640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0373,
+ "step": 19384
+ },
+ {
+ "epoch": 61.932907348242814,
+ "grad_norm": 0.057861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0164,
+ "step": 19385
+ },
+ {
+ "epoch": 61.936102236421725,
+ "grad_norm": 0.054931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0287,
+ "step": 19386
+ },
+ {
+ "epoch": 61.93929712460064,
+ "grad_norm": 0.059814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0257,
+ "step": 19387
+ },
+ {
+ "epoch": 61.94249201277955,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0273,
+ "step": 19388
+ },
+ {
+ "epoch": 61.945686900958464,
+ "grad_norm": 0.062255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0361,
+ "step": 19389
+ },
+ {
+ "epoch": 61.94888178913738,
+ "grad_norm": 0.0830078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0245,
+ "step": 19390
+ },
+ {
+ "epoch": 61.95207667731629,
+ "grad_norm": 0.08984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0273,
+ "step": 19391
+ },
+ {
+ "epoch": 61.95527156549521,
+ "grad_norm": 0.083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.034,
+ "step": 19392
+ },
+ {
+ "epoch": 61.95846645367412,
+ "grad_norm": 0.0615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0195,
+ "step": 19393
+ },
+ {
+ "epoch": 61.96166134185304,
+ "grad_norm": 0.056396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0215,
+ "step": 19394
+ },
+ {
+ "epoch": 61.96485623003195,
+ "grad_norm": 0.10546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0213,
+ "step": 19395
+ },
+ {
+ "epoch": 61.968051118210866,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.041,
+ "step": 19396
+ },
+ {
+ "epoch": 61.97124600638978,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0222,
+ "step": 19397
+ },
+ {
+ "epoch": 61.97444089456869,
+ "grad_norm": 0.0810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0254,
+ "step": 19398
+ },
+ {
+ "epoch": 61.977635782747605,
+ "grad_norm": 0.048583984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0221,
+ "step": 19399
+ },
+ {
+ "epoch": 61.980830670926515,
+ "grad_norm": 0.056396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0207,
+ "step": 19400
+ },
+ {
+ "epoch": 61.98402555910543,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0268,
+ "step": 19401
+ },
+ {
+ "epoch": 61.98722044728434,
+ "grad_norm": 0.057373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0276,
+ "step": 19402
+ },
+ {
+ "epoch": 61.99041533546326,
+ "grad_norm": 0.04541015625,
+ "learning_rate": 0.0005,
+ "loss": 1.021,
+ "step": 19403
+ },
+ {
+ "epoch": 61.99361022364217,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0174,
+ "step": 19404
+ },
+ {
+ "epoch": 61.99680511182109,
+ "grad_norm": 0.056884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0168,
+ "step": 19405
+ },
+ {
+ "epoch": 62.0,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0397,
+ "step": 19406
+ },
+ {
+ "epoch": 62.00319488817891,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0263,
+ "step": 19407
+ },
+ {
+ "epoch": 62.00638977635783,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.014,
+ "step": 19408
+ },
+ {
+ "epoch": 62.00958466453674,
+ "grad_norm": 0.10986328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0117,
+ "step": 19409
+ },
+ {
+ "epoch": 62.01277955271566,
+ "grad_norm": 0.08447265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0289,
+ "step": 19410
+ },
+ {
+ "epoch": 62.01597444089457,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0272,
+ "step": 19411
+ },
+ {
+ "epoch": 62.019169329073485,
+ "grad_norm": 0.1162109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0235,
+ "step": 19412
+ },
+ {
+ "epoch": 62.022364217252395,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0396,
+ "step": 19413
+ },
+ {
+ "epoch": 62.02555910543131,
+ "grad_norm": 0.051513671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0257,
+ "step": 19414
+ },
+ {
+ "epoch": 62.02875399361022,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0253,
+ "step": 19415
+ },
+ {
+ "epoch": 62.031948881789134,
+ "grad_norm": 0.056396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0216,
+ "step": 19416
+ },
+ {
+ "epoch": 62.03514376996805,
+ "grad_norm": 0.049560546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0241,
+ "step": 19417
+ },
+ {
+ "epoch": 62.03833865814696,
+ "grad_norm": 0.0810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0269,
+ "step": 19418
+ },
+ {
+ "epoch": 62.04153354632588,
+ "grad_norm": 0.062255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0287,
+ "step": 19419
+ },
+ {
+ "epoch": 62.04472843450479,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0192,
+ "step": 19420
+ },
+ {
+ "epoch": 62.04792332268371,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0181,
+ "step": 19421
+ },
+ {
+ "epoch": 62.05111821086262,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0256,
+ "step": 19422
+ },
+ {
+ "epoch": 62.054313099041536,
+ "grad_norm": 0.09912109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0217,
+ "step": 19423
+ },
+ {
+ "epoch": 62.05750798722045,
+ "grad_norm": 0.050537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.021,
+ "step": 19424
+ },
+ {
+ "epoch": 62.06070287539936,
+ "grad_norm": 0.087890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0317,
+ "step": 19425
+ },
+ {
+ "epoch": 62.063897763578275,
+ "grad_norm": 0.08251953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0361,
+ "step": 19426
+ },
+ {
+ "epoch": 62.067092651757186,
+ "grad_norm": 0.054931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0296,
+ "step": 19427
+ },
+ {
+ "epoch": 62.0702875399361,
+ "grad_norm": 0.0986328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0334,
+ "step": 19428
+ },
+ {
+ "epoch": 62.073482428115014,
+ "grad_norm": 0.054931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0173,
+ "step": 19429
+ },
+ {
+ "epoch": 62.07667731629393,
+ "grad_norm": 0.05517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0232,
+ "step": 19430
+ },
+ {
+ "epoch": 62.07987220447284,
+ "grad_norm": 0.062255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0257,
+ "step": 19431
+ },
+ {
+ "epoch": 62.08306709265176,
+ "grad_norm": 0.053466796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0197,
+ "step": 19432
+ },
+ {
+ "epoch": 62.08626198083067,
+ "grad_norm": 0.05126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0322,
+ "step": 19433
+ },
+ {
+ "epoch": 62.08945686900959,
+ "grad_norm": 0.048095703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0181,
+ "step": 19434
+ },
+ {
+ "epoch": 62.0926517571885,
+ "grad_norm": 0.04931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0231,
+ "step": 19435
+ },
+ {
+ "epoch": 62.09584664536741,
+ "grad_norm": 0.07861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0284,
+ "step": 19436
+ },
+ {
+ "epoch": 62.09904153354633,
+ "grad_norm": 0.048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0173,
+ "step": 19437
+ },
+ {
+ "epoch": 62.10223642172524,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0309,
+ "step": 19438
+ },
+ {
+ "epoch": 62.105431309904155,
+ "grad_norm": 0.059326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0164,
+ "step": 19439
+ },
+ {
+ "epoch": 62.108626198083066,
+ "grad_norm": 0.052490234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0095,
+ "step": 19440
+ },
+ {
+ "epoch": 62.11182108626198,
+ "grad_norm": 0.062255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0212,
+ "step": 19441
+ },
+ {
+ "epoch": 62.115015974440894,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.029,
+ "step": 19442
+ },
+ {
+ "epoch": 62.11821086261981,
+ "grad_norm": 0.05908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0209,
+ "step": 19443
+ },
+ {
+ "epoch": 62.12140575079872,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0141,
+ "step": 19444
+ },
+ {
+ "epoch": 62.12460063897763,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0184,
+ "step": 19445
+ },
+ {
+ "epoch": 62.12779552715655,
+ "grad_norm": 0.10205078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0198,
+ "step": 19446
+ },
+ {
+ "epoch": 62.13099041533546,
+ "grad_norm": 0.057373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0294,
+ "step": 19447
+ },
+ {
+ "epoch": 62.13418530351438,
+ "grad_norm": 0.057861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0315,
+ "step": 19448
+ },
+ {
+ "epoch": 62.13738019169329,
+ "grad_norm": 0.056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0218,
+ "step": 19449
+ },
+ {
+ "epoch": 62.14057507987221,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0217,
+ "step": 19450
+ },
+ {
+ "epoch": 62.14376996805112,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0143,
+ "step": 19451
+ },
+ {
+ "epoch": 62.146964856230035,
+ "grad_norm": 0.08984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0247,
+ "step": 19452
+ },
+ {
+ "epoch": 62.150159744408946,
+ "grad_norm": 0.08349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0352,
+ "step": 19453
+ },
+ {
+ "epoch": 62.153354632587856,
+ "grad_norm": 0.061767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.023,
+ "step": 19454
+ },
+ {
+ "epoch": 62.156549520766774,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0225,
+ "step": 19455
+ },
+ {
+ "epoch": 62.159744408945684,
+ "grad_norm": 0.060791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0216,
+ "step": 19456
+ },
+ {
+ "epoch": 62.1629392971246,
+ "grad_norm": 0.045166015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0193,
+ "step": 19457
+ },
+ {
+ "epoch": 62.16613418530351,
+ "grad_norm": 0.06201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0219,
+ "step": 19458
+ },
+ {
+ "epoch": 62.16932907348243,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0121,
+ "step": 19459
+ },
+ {
+ "epoch": 62.17252396166134,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0245,
+ "step": 19460
+ },
+ {
+ "epoch": 62.17571884984026,
+ "grad_norm": 0.048095703125,
+ "learning_rate": 0.0005,
+ "loss": 1.032,
+ "step": 19461
+ },
+ {
+ "epoch": 62.17891373801917,
+ "grad_norm": 0.060791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0383,
+ "step": 19462
+ },
+ {
+ "epoch": 62.18210862619808,
+ "grad_norm": 0.0537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.027,
+ "step": 19463
+ },
+ {
+ "epoch": 62.185303514377,
+ "grad_norm": 0.08837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.022,
+ "step": 19464
+ },
+ {
+ "epoch": 62.18849840255591,
+ "grad_norm": 0.052978515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0207,
+ "step": 19465
+ },
+ {
+ "epoch": 62.191693290734825,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0237,
+ "step": 19466
+ },
+ {
+ "epoch": 62.194888178913736,
+ "grad_norm": 0.060791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0209,
+ "step": 19467
+ },
+ {
+ "epoch": 62.198083067092654,
+ "grad_norm": 0.0849609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0228,
+ "step": 19468
+ },
+ {
+ "epoch": 62.201277955271564,
+ "grad_norm": 0.052734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0167,
+ "step": 19469
+ },
+ {
+ "epoch": 62.20447284345048,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0275,
+ "step": 19470
+ },
+ {
+ "epoch": 62.20766773162939,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0283,
+ "step": 19471
+ },
+ {
+ "epoch": 62.21086261980831,
+ "grad_norm": 0.09375,
+ "learning_rate": 0.0005,
+ "loss": 1.0243,
+ "step": 19472
+ },
+ {
+ "epoch": 62.21405750798722,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0209,
+ "step": 19473
+ },
+ {
+ "epoch": 62.21725239616613,
+ "grad_norm": 0.091796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0245,
+ "step": 19474
+ },
+ {
+ "epoch": 62.22044728434505,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0334,
+ "step": 19475
+ },
+ {
+ "epoch": 62.22364217252396,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.016,
+ "step": 19476
+ },
+ {
+ "epoch": 62.22683706070288,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0354,
+ "step": 19477
+ },
+ {
+ "epoch": 62.23003194888179,
+ "grad_norm": 0.095703125,
+ "learning_rate": 0.0005,
+ "loss": 1.026,
+ "step": 19478
+ },
+ {
+ "epoch": 62.233226837060705,
+ "grad_norm": 0.054443359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0258,
+ "step": 19479
+ },
+ {
+ "epoch": 62.236421725239616,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0299,
+ "step": 19480
+ },
+ {
+ "epoch": 62.239616613418534,
+ "grad_norm": 0.08642578125,
+ "learning_rate": 0.0005,
+ "loss": 1.021,
+ "step": 19481
+ },
+ {
+ "epoch": 62.242811501597444,
+ "grad_norm": 0.058837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0218,
+ "step": 19482
+ },
+ {
+ "epoch": 62.246006389776355,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0264,
+ "step": 19483
+ },
+ {
+ "epoch": 62.24920127795527,
+ "grad_norm": 0.060302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0324,
+ "step": 19484
+ },
+ {
+ "epoch": 62.25239616613418,
+ "grad_norm": 0.05712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0281,
+ "step": 19485
+ },
+ {
+ "epoch": 62.2555910543131,
+ "grad_norm": 0.052001953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0269,
+ "step": 19486
+ },
+ {
+ "epoch": 62.25878594249201,
+ "grad_norm": 0.053466796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0272,
+ "step": 19487
+ },
+ {
+ "epoch": 62.26198083067093,
+ "grad_norm": 0.046142578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0218,
+ "step": 19488
+ },
+ {
+ "epoch": 62.26517571884984,
+ "grad_norm": 0.0537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0282,
+ "step": 19489
+ },
+ {
+ "epoch": 62.26837060702876,
+ "grad_norm": 0.04443359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0199,
+ "step": 19490
+ },
+ {
+ "epoch": 62.27156549520767,
+ "grad_norm": 0.050048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0254,
+ "step": 19491
+ },
+ {
+ "epoch": 62.27476038338658,
+ "grad_norm": 0.04736328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0209,
+ "step": 19492
+ },
+ {
+ "epoch": 62.277955271565496,
+ "grad_norm": 0.05859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0108,
+ "step": 19493
+ },
+ {
+ "epoch": 62.281150159744406,
+ "grad_norm": 0.058349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0167,
+ "step": 19494
+ },
+ {
+ "epoch": 62.284345047923324,
+ "grad_norm": 0.062255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0284,
+ "step": 19495
+ },
+ {
+ "epoch": 62.287539936102235,
+ "grad_norm": 0.052001953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0146,
+ "step": 19496
+ },
+ {
+ "epoch": 62.29073482428115,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0302,
+ "step": 19497
+ },
+ {
+ "epoch": 62.29392971246006,
+ "grad_norm": 0.050537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0264,
+ "step": 19498
+ },
+ {
+ "epoch": 62.29712460063898,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0293,
+ "step": 19499
+ },
+ {
+ "epoch": 62.30031948881789,
+ "grad_norm": 0.04736328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0242,
+ "step": 19500
+ },
+ {
+ "epoch": 62.3035143769968,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.013,
+ "step": 19501
+ },
+ {
+ "epoch": 62.30670926517572,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0079,
+ "step": 19502
+ },
+ {
+ "epoch": 62.30990415335463,
+ "grad_norm": 0.053466796875,
+ "learning_rate": 0.0005,
+ "loss": 1.021,
+ "step": 19503
+ },
+ {
+ "epoch": 62.31309904153355,
+ "grad_norm": 0.06103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0332,
+ "step": 19504
+ },
+ {
+ "epoch": 62.31629392971246,
+ "grad_norm": 0.052001953125,
+ "learning_rate": 0.0005,
+ "loss": 1.018,
+ "step": 19505
+ },
+ {
+ "epoch": 62.319488817891376,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0249,
+ "step": 19506
+ },
+ {
+ "epoch": 62.322683706070286,
+ "grad_norm": 0.060302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0182,
+ "step": 19507
+ },
+ {
+ "epoch": 62.325878594249204,
+ "grad_norm": 0.05078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0242,
+ "step": 19508
+ },
+ {
+ "epoch": 62.329073482428115,
+ "grad_norm": 0.08349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0354,
+ "step": 19509
+ },
+ {
+ "epoch": 62.33226837060703,
+ "grad_norm": 0.04833984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0226,
+ "step": 19510
+ },
+ {
+ "epoch": 62.33546325878594,
+ "grad_norm": 0.05322265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0244,
+ "step": 19511
+ },
+ {
+ "epoch": 62.33865814696485,
+ "grad_norm": 0.050048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.027,
+ "step": 19512
+ },
+ {
+ "epoch": 62.34185303514377,
+ "grad_norm": 0.0458984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0249,
+ "step": 19513
+ },
+ {
+ "epoch": 62.34504792332268,
+ "grad_norm": 0.0615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0215,
+ "step": 19514
+ },
+ {
+ "epoch": 62.3482428115016,
+ "grad_norm": 0.0595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0244,
+ "step": 19515
+ },
+ {
+ "epoch": 62.35143769968051,
+ "grad_norm": 0.054931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0193,
+ "step": 19516
+ },
+ {
+ "epoch": 62.35463258785943,
+ "grad_norm": 0.051513671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0184,
+ "step": 19517
+ },
+ {
+ "epoch": 62.35782747603834,
+ "grad_norm": 0.05029296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0228,
+ "step": 19518
+ },
+ {
+ "epoch": 62.361022364217256,
+ "grad_norm": 0.057861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0217,
+ "step": 19519
+ },
+ {
+ "epoch": 62.364217252396166,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0191,
+ "step": 19520
+ },
+ {
+ "epoch": 62.36741214057508,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0228,
+ "step": 19521
+ },
+ {
+ "epoch": 62.370607028753994,
+ "grad_norm": 0.09814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0157,
+ "step": 19522
+ },
+ {
+ "epoch": 62.373801916932905,
+ "grad_norm": 0.06201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0226,
+ "step": 19523
+ },
+ {
+ "epoch": 62.37699680511182,
+ "grad_norm": 0.05029296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0223,
+ "step": 19524
+ },
+ {
+ "epoch": 62.38019169329073,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0378,
+ "step": 19525
+ },
+ {
+ "epoch": 62.38338658146965,
+ "grad_norm": 0.0478515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0221,
+ "step": 19526
+ },
+ {
+ "epoch": 62.38658146964856,
+ "grad_norm": 0.0947265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0291,
+ "step": 19527
+ },
+ {
+ "epoch": 62.38977635782748,
+ "grad_norm": 0.0556640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0238,
+ "step": 19528
+ },
+ {
+ "epoch": 62.39297124600639,
+ "grad_norm": 0.0830078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0304,
+ "step": 19529
+ },
+ {
+ "epoch": 62.3961661341853,
+ "grad_norm": 0.0517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0201,
+ "step": 19530
+ },
+ {
+ "epoch": 62.39936102236422,
+ "grad_norm": 0.06005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0122,
+ "step": 19531
+ },
+ {
+ "epoch": 62.40255591054313,
+ "grad_norm": 0.05126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0327,
+ "step": 19532
+ },
+ {
+ "epoch": 62.405750798722046,
+ "grad_norm": 0.052490234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0216,
+ "step": 19533
+ },
+ {
+ "epoch": 62.40894568690096,
+ "grad_norm": 0.05224609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0315,
+ "step": 19534
+ },
+ {
+ "epoch": 62.412140575079874,
+ "grad_norm": 0.055908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0231,
+ "step": 19535
+ },
+ {
+ "epoch": 62.415335463258785,
+ "grad_norm": 0.0595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0236,
+ "step": 19536
+ },
+ {
+ "epoch": 62.4185303514377,
+ "grad_norm": 0.04541015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0282,
+ "step": 19537
+ },
+ {
+ "epoch": 62.42172523961661,
+ "grad_norm": 0.09814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0301,
+ "step": 19538
+ },
+ {
+ "epoch": 62.424920127795524,
+ "grad_norm": 0.044677734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0205,
+ "step": 19539
+ },
+ {
+ "epoch": 62.42811501597444,
+ "grad_norm": 0.05224609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0194,
+ "step": 19540
+ },
+ {
+ "epoch": 62.43130990415335,
+ "grad_norm": 0.059814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0221,
+ "step": 19541
+ },
+ {
+ "epoch": 62.43450479233227,
+ "grad_norm": 0.05419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0239,
+ "step": 19542
+ },
+ {
+ "epoch": 62.43769968051118,
+ "grad_norm": 0.056396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0178,
+ "step": 19543
+ },
+ {
+ "epoch": 62.4408945686901,
+ "grad_norm": 0.05908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0335,
+ "step": 19544
+ },
+ {
+ "epoch": 62.44408945686901,
+ "grad_norm": 0.054443359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0161,
+ "step": 19545
+ },
+ {
+ "epoch": 62.447284345047926,
+ "grad_norm": 0.0576171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0222,
+ "step": 19546
+ },
+ {
+ "epoch": 62.45047923322684,
+ "grad_norm": 0.044677734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0253,
+ "step": 19547
+ },
+ {
+ "epoch": 62.453674121405754,
+ "grad_norm": 0.052490234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0174,
+ "step": 19548
+ },
+ {
+ "epoch": 62.456869009584665,
+ "grad_norm": 0.05029296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0255,
+ "step": 19549
+ },
+ {
+ "epoch": 62.460063897763575,
+ "grad_norm": 0.052001953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0302,
+ "step": 19550
+ },
+ {
+ "epoch": 62.46325878594249,
+ "grad_norm": 0.080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0295,
+ "step": 19551
+ },
+ {
+ "epoch": 62.466453674121404,
+ "grad_norm": 0.09228515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0219,
+ "step": 19552
+ },
+ {
+ "epoch": 62.46964856230032,
+ "grad_norm": 0.0947265625,
+ "learning_rate": 0.0005,
+ "loss": 1.025,
+ "step": 19553
+ },
+ {
+ "epoch": 62.47284345047923,
+ "grad_norm": 0.052001953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0399,
+ "step": 19554
+ },
+ {
+ "epoch": 62.47603833865815,
+ "grad_norm": 0.059814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0391,
+ "step": 19555
+ },
+ {
+ "epoch": 62.47923322683706,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0311,
+ "step": 19556
+ },
+ {
+ "epoch": 62.48242811501598,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0271,
+ "step": 19557
+ },
+ {
+ "epoch": 62.48562300319489,
+ "grad_norm": 0.052734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0271,
+ "step": 19558
+ },
+ {
+ "epoch": 62.4888178913738,
+ "grad_norm": 0.0859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0272,
+ "step": 19559
+ },
+ {
+ "epoch": 62.49201277955272,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0197,
+ "step": 19560
+ },
+ {
+ "epoch": 62.49520766773163,
+ "grad_norm": 0.055419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0254,
+ "step": 19561
+ },
+ {
+ "epoch": 62.498402555910545,
+ "grad_norm": 0.0830078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0124,
+ "step": 19562
+ },
+ {
+ "epoch": 62.501597444089455,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0211,
+ "step": 19563
+ },
+ {
+ "epoch": 62.50479233226837,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0187,
+ "step": 19564
+ },
+ {
+ "epoch": 62.50798722044728,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0241,
+ "step": 19565
+ },
+ {
+ "epoch": 62.5111821086262,
+ "grad_norm": 0.1259765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0295,
+ "step": 19566
+ },
+ {
+ "epoch": 62.51437699680511,
+ "grad_norm": 0.059814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0237,
+ "step": 19567
+ },
+ {
+ "epoch": 62.51757188498402,
+ "grad_norm": 0.08447265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0295,
+ "step": 19568
+ },
+ {
+ "epoch": 62.52076677316294,
+ "grad_norm": 0.0966796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0198,
+ "step": 19569
+ },
+ {
+ "epoch": 62.52396166134185,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0131,
+ "step": 19570
+ },
+ {
+ "epoch": 62.52715654952077,
+ "grad_norm": 0.0859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0161,
+ "step": 19571
+ },
+ {
+ "epoch": 62.53035143769968,
+ "grad_norm": 0.119140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0288,
+ "step": 19572
+ },
+ {
+ "epoch": 62.533546325878596,
+ "grad_norm": 0.052978515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0205,
+ "step": 19573
+ },
+ {
+ "epoch": 62.53674121405751,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0302,
+ "step": 19574
+ },
+ {
+ "epoch": 62.539936102236425,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0145,
+ "step": 19575
+ },
+ {
+ "epoch": 62.543130990415335,
+ "grad_norm": 0.06103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0242,
+ "step": 19576
+ },
+ {
+ "epoch": 62.546325878594246,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0258,
+ "step": 19577
+ },
+ {
+ "epoch": 62.54952076677316,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0198,
+ "step": 19578
+ },
+ {
+ "epoch": 62.552715654952074,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0237,
+ "step": 19579
+ },
+ {
+ "epoch": 62.55591054313099,
+ "grad_norm": 0.062255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0173,
+ "step": 19580
+ },
+ {
+ "epoch": 62.5591054313099,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.017,
+ "step": 19581
+ },
+ {
+ "epoch": 62.56230031948882,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0263,
+ "step": 19582
+ },
+ {
+ "epoch": 62.56549520766773,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0193,
+ "step": 19583
+ },
+ {
+ "epoch": 62.56869009584665,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0215,
+ "step": 19584
+ },
+ {
+ "epoch": 62.57188498402556,
+ "grad_norm": 0.0986328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0299,
+ "step": 19585
+ },
+ {
+ "epoch": 62.575079872204476,
+ "grad_norm": 0.05419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0252,
+ "step": 19586
+ },
+ {
+ "epoch": 62.57827476038339,
+ "grad_norm": 0.1484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0292,
+ "step": 19587
+ },
+ {
+ "epoch": 62.5814696485623,
+ "grad_norm": 0.054931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0193,
+ "step": 19588
+ },
+ {
+ "epoch": 62.584664536741215,
+ "grad_norm": 0.138671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0291,
+ "step": 19589
+ },
+ {
+ "epoch": 62.587859424920126,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0363,
+ "step": 19590
+ },
+ {
+ "epoch": 62.59105431309904,
+ "grad_norm": 0.0810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0143,
+ "step": 19591
+ },
+ {
+ "epoch": 62.594249201277954,
+ "grad_norm": 0.060791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0244,
+ "step": 19592
+ },
+ {
+ "epoch": 62.59744408945687,
+ "grad_norm": 0.0888671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0301,
+ "step": 19593
+ },
+ {
+ "epoch": 62.60063897763578,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0145,
+ "step": 19594
+ },
+ {
+ "epoch": 62.6038338658147,
+ "grad_norm": 0.095703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0239,
+ "step": 19595
+ },
+ {
+ "epoch": 62.60702875399361,
+ "grad_norm": 0.0908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0259,
+ "step": 19596
+ },
+ {
+ "epoch": 62.61022364217252,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.013,
+ "step": 19597
+ },
+ {
+ "epoch": 62.61341853035144,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.032,
+ "step": 19598
+ },
+ {
+ "epoch": 62.61661341853035,
+ "grad_norm": 0.09130859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0263,
+ "step": 19599
+ },
+ {
+ "epoch": 62.61980830670927,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0218,
+ "step": 19600
+ },
+ {
+ "epoch": 62.62300319488818,
+ "grad_norm": 0.0546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0147,
+ "step": 19601
+ },
+ {
+ "epoch": 62.626198083067095,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.038,
+ "step": 19602
+ },
+ {
+ "epoch": 62.629392971246006,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0208,
+ "step": 19603
+ },
+ {
+ "epoch": 62.63258785942492,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0349,
+ "step": 19604
+ },
+ {
+ "epoch": 62.635782747603834,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0238,
+ "step": 19605
+ },
+ {
+ "epoch": 62.638977635782744,
+ "grad_norm": 0.052978515625,
+ "learning_rate": 0.0005,
+ "loss": 1.026,
+ "step": 19606
+ },
+ {
+ "epoch": 62.64217252396166,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0197,
+ "step": 19607
+ },
+ {
+ "epoch": 62.64536741214057,
+ "grad_norm": 0.046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0164,
+ "step": 19608
+ },
+ {
+ "epoch": 62.64856230031949,
+ "grad_norm": 0.056884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0183,
+ "step": 19609
+ },
+ {
+ "epoch": 62.6517571884984,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0213,
+ "step": 19610
+ },
+ {
+ "epoch": 62.65495207667732,
+ "grad_norm": 0.0556640625,
+ "learning_rate": 0.0005,
+ "loss": 1.014,
+ "step": 19611
+ },
+ {
+ "epoch": 62.65814696485623,
+ "grad_norm": 0.057373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0184,
+ "step": 19612
+ },
+ {
+ "epoch": 62.66134185303515,
+ "grad_norm": 0.047119140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0257,
+ "step": 19613
+ },
+ {
+ "epoch": 62.66453674121406,
+ "grad_norm": 0.08154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0308,
+ "step": 19614
+ },
+ {
+ "epoch": 62.66773162939297,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0242,
+ "step": 19615
+ },
+ {
+ "epoch": 62.670926517571885,
+ "grad_norm": 0.10791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0268,
+ "step": 19616
+ },
+ {
+ "epoch": 62.674121405750796,
+ "grad_norm": 0.0986328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0267,
+ "step": 19617
+ },
+ {
+ "epoch": 62.677316293929714,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0274,
+ "step": 19618
+ },
+ {
+ "epoch": 62.680511182108624,
+ "grad_norm": 0.1357421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0304,
+ "step": 19619
+ },
+ {
+ "epoch": 62.68370607028754,
+ "grad_norm": 0.061767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0127,
+ "step": 19620
+ },
+ {
+ "epoch": 62.68690095846645,
+ "grad_norm": 0.10107421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0135,
+ "step": 19621
+ },
+ {
+ "epoch": 62.69009584664537,
+ "grad_norm": 0.048095703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0239,
+ "step": 19622
+ },
+ {
+ "epoch": 62.69329073482428,
+ "grad_norm": 0.055419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0291,
+ "step": 19623
+ },
+ {
+ "epoch": 62.6964856230032,
+ "grad_norm": 0.043701171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0254,
+ "step": 19624
+ },
+ {
+ "epoch": 62.69968051118211,
+ "grad_norm": 0.047607421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0136,
+ "step": 19625
+ },
+ {
+ "epoch": 62.70287539936102,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0221,
+ "step": 19626
+ },
+ {
+ "epoch": 62.70607028753994,
+ "grad_norm": 0.05419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0125,
+ "step": 19627
+ },
+ {
+ "epoch": 62.70926517571885,
+ "grad_norm": 0.05126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.024,
+ "step": 19628
+ },
+ {
+ "epoch": 62.712460063897765,
+ "grad_norm": 0.05322265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0304,
+ "step": 19629
+ },
+ {
+ "epoch": 62.715654952076676,
+ "grad_norm": 0.046630859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0109,
+ "step": 19630
+ },
+ {
+ "epoch": 62.718849840255594,
+ "grad_norm": 0.052978515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0351,
+ "step": 19631
+ },
+ {
+ "epoch": 62.722044728434504,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0199,
+ "step": 19632
+ },
+ {
+ "epoch": 62.72523961661342,
+ "grad_norm": 0.052734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0266,
+ "step": 19633
+ },
+ {
+ "epoch": 62.72843450479233,
+ "grad_norm": 0.0576171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0285,
+ "step": 19634
+ },
+ {
+ "epoch": 62.73162939297124,
+ "grad_norm": 0.0556640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0081,
+ "step": 19635
+ },
+ {
+ "epoch": 62.73482428115016,
+ "grad_norm": 0.04443359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0196,
+ "step": 19636
+ },
+ {
+ "epoch": 62.73801916932907,
+ "grad_norm": 0.0556640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0355,
+ "step": 19637
+ },
+ {
+ "epoch": 62.74121405750799,
+ "grad_norm": 0.04541015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0327,
+ "step": 19638
+ },
+ {
+ "epoch": 62.7444089456869,
+ "grad_norm": 0.05615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0213,
+ "step": 19639
+ },
+ {
+ "epoch": 62.74760383386582,
+ "grad_norm": 0.053955078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0303,
+ "step": 19640
+ },
+ {
+ "epoch": 62.75079872204473,
+ "grad_norm": 0.060791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0129,
+ "step": 19641
+ },
+ {
+ "epoch": 62.753993610223645,
+ "grad_norm": 0.054931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0238,
+ "step": 19642
+ },
+ {
+ "epoch": 62.757188498402556,
+ "grad_norm": 0.056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0144,
+ "step": 19643
+ },
+ {
+ "epoch": 62.760383386581466,
+ "grad_norm": 0.0439453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0138,
+ "step": 19644
+ },
+ {
+ "epoch": 62.763578274760384,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.01,
+ "step": 19645
+ },
+ {
+ "epoch": 62.766773162939295,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0218,
+ "step": 19646
+ },
+ {
+ "epoch": 62.76996805111821,
+ "grad_norm": 0.04931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0209,
+ "step": 19647
+ },
+ {
+ "epoch": 62.77316293929712,
+ "grad_norm": 0.053466796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0104,
+ "step": 19648
+ },
+ {
+ "epoch": 62.77635782747604,
+ "grad_norm": 0.04638671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0261,
+ "step": 19649
+ },
+ {
+ "epoch": 62.77955271565495,
+ "grad_norm": 0.051025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0181,
+ "step": 19650
+ },
+ {
+ "epoch": 62.78274760383387,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.017,
+ "step": 19651
+ },
+ {
+ "epoch": 62.78594249201278,
+ "grad_norm": 0.046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0322,
+ "step": 19652
+ },
+ {
+ "epoch": 62.78913738019169,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0339,
+ "step": 19653
+ },
+ {
+ "epoch": 62.79233226837061,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0229,
+ "step": 19654
+ },
+ {
+ "epoch": 62.79552715654952,
+ "grad_norm": 0.055908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0269,
+ "step": 19655
+ },
+ {
+ "epoch": 62.798722044728436,
+ "grad_norm": 0.0576171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0231,
+ "step": 19656
+ },
+ {
+ "epoch": 62.801916932907346,
+ "grad_norm": 0.134765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0225,
+ "step": 19657
+ },
+ {
+ "epoch": 62.805111821086264,
+ "grad_norm": 0.052490234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0237,
+ "step": 19658
+ },
+ {
+ "epoch": 62.808306709265175,
+ "grad_norm": 0.1005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0227,
+ "step": 19659
+ },
+ {
+ "epoch": 62.81150159744409,
+ "grad_norm": 0.09130859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0271,
+ "step": 19660
+ },
+ {
+ "epoch": 62.814696485623,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0236,
+ "step": 19661
+ },
+ {
+ "epoch": 62.81789137380191,
+ "grad_norm": 0.080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0218,
+ "step": 19662
+ },
+ {
+ "epoch": 62.82108626198083,
+ "grad_norm": 0.09130859375,
+ "learning_rate": 0.0005,
+ "loss": 1.037,
+ "step": 19663
+ },
+ {
+ "epoch": 62.82428115015974,
+ "grad_norm": 0.07861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.033,
+ "step": 19664
+ },
+ {
+ "epoch": 62.82747603833866,
+ "grad_norm": 0.107421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0258,
+ "step": 19665
+ },
+ {
+ "epoch": 62.83067092651757,
+ "grad_norm": 0.06103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0329,
+ "step": 19666
+ },
+ {
+ "epoch": 62.83386581469649,
+ "grad_norm": 0.1259765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0229,
+ "step": 19667
+ },
+ {
+ "epoch": 62.8370607028754,
+ "grad_norm": 0.09228515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0367,
+ "step": 19668
+ },
+ {
+ "epoch": 62.840255591054316,
+ "grad_norm": 0.1005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0263,
+ "step": 19669
+ },
+ {
+ "epoch": 62.843450479233226,
+ "grad_norm": 0.08544921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0202,
+ "step": 19670
+ },
+ {
+ "epoch": 62.846645367412144,
+ "grad_norm": 0.059814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.02,
+ "step": 19671
+ },
+ {
+ "epoch": 62.849840255591054,
+ "grad_norm": 0.083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0169,
+ "step": 19672
+ },
+ {
+ "epoch": 62.853035143769965,
+ "grad_norm": 0.058349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0262,
+ "step": 19673
+ },
+ {
+ "epoch": 62.85623003194888,
+ "grad_norm": 0.1044921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0083,
+ "step": 19674
+ },
+ {
+ "epoch": 62.85942492012779,
+ "grad_norm": 0.053466796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0125,
+ "step": 19675
+ },
+ {
+ "epoch": 62.86261980830671,
+ "grad_norm": 0.091796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0241,
+ "step": 19676
+ },
+ {
+ "epoch": 62.86581469648562,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.016,
+ "step": 19677
+ },
+ {
+ "epoch": 62.86900958466454,
+ "grad_norm": 0.0888671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0269,
+ "step": 19678
+ },
+ {
+ "epoch": 62.87220447284345,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0086,
+ "step": 19679
+ },
+ {
+ "epoch": 62.87539936102237,
+ "grad_norm": 0.1103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0193,
+ "step": 19680
+ },
+ {
+ "epoch": 62.87859424920128,
+ "grad_norm": 0.1005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0279,
+ "step": 19681
+ },
+ {
+ "epoch": 62.88178913738019,
+ "grad_norm": 0.10986328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0148,
+ "step": 19682
+ },
+ {
+ "epoch": 62.884984025559106,
+ "grad_norm": 0.0986328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0276,
+ "step": 19683
+ },
+ {
+ "epoch": 62.88817891373802,
+ "grad_norm": 0.04931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.026,
+ "step": 19684
+ },
+ {
+ "epoch": 62.891373801916934,
+ "grad_norm": 0.0810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0243,
+ "step": 19685
+ },
+ {
+ "epoch": 62.894568690095845,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0283,
+ "step": 19686
+ },
+ {
+ "epoch": 62.89776357827476,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0228,
+ "step": 19687
+ },
+ {
+ "epoch": 62.90095846645367,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0244,
+ "step": 19688
+ },
+ {
+ "epoch": 62.90415335463259,
+ "grad_norm": 0.08642578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0262,
+ "step": 19689
+ },
+ {
+ "epoch": 62.9073482428115,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0317,
+ "step": 19690
+ },
+ {
+ "epoch": 62.91054313099041,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0271,
+ "step": 19691
+ },
+ {
+ "epoch": 62.91373801916933,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0256,
+ "step": 19692
+ },
+ {
+ "epoch": 62.91693290734824,
+ "grad_norm": 0.1513671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0292,
+ "step": 19693
+ },
+ {
+ "epoch": 62.92012779552716,
+ "grad_norm": 0.150390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0233,
+ "step": 19694
+ },
+ {
+ "epoch": 62.92332268370607,
+ "grad_norm": 0.1181640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0225,
+ "step": 19695
+ },
+ {
+ "epoch": 62.926517571884986,
+ "grad_norm": 0.10302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0226,
+ "step": 19696
+ },
+ {
+ "epoch": 62.9297124600639,
+ "grad_norm": 0.10498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0215,
+ "step": 19697
+ },
+ {
+ "epoch": 62.932907348242814,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0268,
+ "step": 19698
+ },
+ {
+ "epoch": 62.936102236421725,
+ "grad_norm": 0.058837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0278,
+ "step": 19699
+ },
+ {
+ "epoch": 62.93929712460064,
+ "grad_norm": 0.1748046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0158,
+ "step": 19700
+ },
+ {
+ "epoch": 62.94249201277955,
+ "grad_norm": 0.05712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0165,
+ "step": 19701
+ },
+ {
+ "epoch": 62.945686900958464,
+ "grad_norm": 0.1484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0154,
+ "step": 19702
+ },
+ {
+ "epoch": 62.94888178913738,
+ "grad_norm": 0.11083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0231,
+ "step": 19703
+ },
+ {
+ "epoch": 62.95207667731629,
+ "grad_norm": 0.1298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0244,
+ "step": 19704
+ },
+ {
+ "epoch": 62.95527156549521,
+ "grad_norm": 0.123046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0267,
+ "step": 19705
+ },
+ {
+ "epoch": 62.95846645367412,
+ "grad_norm": 0.1162109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0133,
+ "step": 19706
+ },
+ {
+ "epoch": 62.96166134185304,
+ "grad_norm": 0.158203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0177,
+ "step": 19707
+ },
+ {
+ "epoch": 62.96485623003195,
+ "grad_norm": 0.1640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0251,
+ "step": 19708
+ },
+ {
+ "epoch": 62.968051118210866,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0353,
+ "step": 19709
+ },
+ {
+ "epoch": 62.97124600638978,
+ "grad_norm": 0.13671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0214,
+ "step": 19710
+ },
+ {
+ "epoch": 62.97444089456869,
+ "grad_norm": 0.09228515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0297,
+ "step": 19711
+ },
+ {
+ "epoch": 62.977635782747605,
+ "grad_norm": 0.09912109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0255,
+ "step": 19712
+ },
+ {
+ "epoch": 62.980830670926515,
+ "grad_norm": 0.123046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0235,
+ "step": 19713
+ },
+ {
+ "epoch": 62.98402555910543,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0183,
+ "step": 19714
+ },
+ {
+ "epoch": 62.98722044728434,
+ "grad_norm": 0.1015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0127,
+ "step": 19715
+ },
+ {
+ "epoch": 62.99041533546326,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0192,
+ "step": 19716
+ },
+ {
+ "epoch": 62.99361022364217,
+ "grad_norm": 0.056396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0281,
+ "step": 19717
+ },
+ {
+ "epoch": 62.99680511182109,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.027,
+ "step": 19718
+ },
+ {
+ "epoch": 63.0,
+ "grad_norm": 0.055908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.026,
+ "step": 19719
+ },
+ {
+ "epoch": 63.00319488817891,
+ "grad_norm": 0.052001953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0263,
+ "step": 19720
+ },
+ {
+ "epoch": 63.00638977635783,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0322,
+ "step": 19721
+ },
+ {
+ "epoch": 63.00958466453674,
+ "grad_norm": 0.052978515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0191,
+ "step": 19722
+ },
+ {
+ "epoch": 63.01277955271566,
+ "grad_norm": 0.0556640625,
+ "learning_rate": 0.0005,
+ "loss": 1.026,
+ "step": 19723
+ },
+ {
+ "epoch": 63.01597444089457,
+ "grad_norm": 0.080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0152,
+ "step": 19724
+ },
+ {
+ "epoch": 63.019169329073485,
+ "grad_norm": 0.0478515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0332,
+ "step": 19725
+ },
+ {
+ "epoch": 63.022364217252395,
+ "grad_norm": 0.0556640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0218,
+ "step": 19726
+ },
+ {
+ "epoch": 63.02555910543131,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0301,
+ "step": 19727
+ },
+ {
+ "epoch": 63.02875399361022,
+ "grad_norm": 0.05859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0214,
+ "step": 19728
+ },
+ {
+ "epoch": 63.031948881789134,
+ "grad_norm": 0.058837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0251,
+ "step": 19729
+ },
+ {
+ "epoch": 63.03514376996805,
+ "grad_norm": 0.08349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0215,
+ "step": 19730
+ },
+ {
+ "epoch": 63.03833865814696,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0216,
+ "step": 19731
+ },
+ {
+ "epoch": 63.04153354632588,
+ "grad_norm": 0.054443359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0217,
+ "step": 19732
+ },
+ {
+ "epoch": 63.04472843450479,
+ "grad_norm": 0.057861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0238,
+ "step": 19733
+ },
+ {
+ "epoch": 63.04792332268371,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0161,
+ "step": 19734
+ },
+ {
+ "epoch": 63.05111821086262,
+ "grad_norm": 0.049560546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0267,
+ "step": 19735
+ },
+ {
+ "epoch": 63.054313099041536,
+ "grad_norm": 0.05078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0394,
+ "step": 19736
+ },
+ {
+ "epoch": 63.05750798722045,
+ "grad_norm": 0.05126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.014,
+ "step": 19737
+ },
+ {
+ "epoch": 63.06070287539936,
+ "grad_norm": 0.0576171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0315,
+ "step": 19738
+ },
+ {
+ "epoch": 63.063897763578275,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0344,
+ "step": 19739
+ },
+ {
+ "epoch": 63.067092651757186,
+ "grad_norm": 0.045654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0297,
+ "step": 19740
+ },
+ {
+ "epoch": 63.0702875399361,
+ "grad_norm": 0.0537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0153,
+ "step": 19741
+ },
+ {
+ "epoch": 63.073482428115014,
+ "grad_norm": 0.0546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0369,
+ "step": 19742
+ },
+ {
+ "epoch": 63.07667731629393,
+ "grad_norm": 0.042724609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0191,
+ "step": 19743
+ },
+ {
+ "epoch": 63.07987220447284,
+ "grad_norm": 0.0517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0177,
+ "step": 19744
+ },
+ {
+ "epoch": 63.08306709265176,
+ "grad_norm": 0.04736328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0154,
+ "step": 19745
+ },
+ {
+ "epoch": 63.08626198083067,
+ "grad_norm": 0.060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.016,
+ "step": 19746
+ },
+ {
+ "epoch": 63.08945686900959,
+ "grad_norm": 0.0869140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0321,
+ "step": 19747
+ },
+ {
+ "epoch": 63.0926517571885,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0247,
+ "step": 19748
+ },
+ {
+ "epoch": 63.09584664536741,
+ "grad_norm": 0.09375,
+ "learning_rate": 0.0005,
+ "loss": 1.0339,
+ "step": 19749
+ },
+ {
+ "epoch": 63.09904153354633,
+ "grad_norm": 0.058349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0195,
+ "step": 19750
+ },
+ {
+ "epoch": 63.10223642172524,
+ "grad_norm": 0.061279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0169,
+ "step": 19751
+ },
+ {
+ "epoch": 63.105431309904155,
+ "grad_norm": 0.08837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0265,
+ "step": 19752
+ },
+ {
+ "epoch": 63.108626198083066,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0169,
+ "step": 19753
+ },
+ {
+ "epoch": 63.11182108626198,
+ "grad_norm": 0.1103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0178,
+ "step": 19754
+ },
+ {
+ "epoch": 63.115015974440894,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0184,
+ "step": 19755
+ },
+ {
+ "epoch": 63.11821086261981,
+ "grad_norm": 0.09130859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0238,
+ "step": 19756
+ },
+ {
+ "epoch": 63.12140575079872,
+ "grad_norm": 0.060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0267,
+ "step": 19757
+ },
+ {
+ "epoch": 63.12460063897763,
+ "grad_norm": 0.0927734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0099,
+ "step": 19758
+ },
+ {
+ "epoch": 63.12779552715655,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0206,
+ "step": 19759
+ },
+ {
+ "epoch": 63.13099041533546,
+ "grad_norm": 0.0849609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0228,
+ "step": 19760
+ },
+ {
+ "epoch": 63.13418530351438,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0248,
+ "step": 19761
+ },
+ {
+ "epoch": 63.13738019169329,
+ "grad_norm": 0.1337890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0132,
+ "step": 19762
+ },
+ {
+ "epoch": 63.14057507987221,
+ "grad_norm": 0.08203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0152,
+ "step": 19763
+ },
+ {
+ "epoch": 63.14376996805112,
+ "grad_norm": 0.171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0217,
+ "step": 19764
+ },
+ {
+ "epoch": 63.146964856230035,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.014,
+ "step": 19765
+ },
+ {
+ "epoch": 63.150159744408946,
+ "grad_norm": 0.134765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0277,
+ "step": 19766
+ },
+ {
+ "epoch": 63.153354632587856,
+ "grad_norm": 0.11376953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0165,
+ "step": 19767
+ },
+ {
+ "epoch": 63.156549520766774,
+ "grad_norm": 0.1357421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0212,
+ "step": 19768
+ },
+ {
+ "epoch": 63.159744408945684,
+ "grad_norm": 0.11767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0339,
+ "step": 19769
+ },
+ {
+ "epoch": 63.1629392971246,
+ "grad_norm": 0.1484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0223,
+ "step": 19770
+ },
+ {
+ "epoch": 63.16613418530351,
+ "grad_norm": 0.08642578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0285,
+ "step": 19771
+ },
+ {
+ "epoch": 63.16932907348243,
+ "grad_norm": 0.13671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0173,
+ "step": 19772
+ },
+ {
+ "epoch": 63.17252396166134,
+ "grad_norm": 0.07861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0298,
+ "step": 19773
+ },
+ {
+ "epoch": 63.17571884984026,
+ "grad_norm": 0.1259765625,
+ "learning_rate": 0.0005,
+ "loss": 1.027,
+ "step": 19774
+ },
+ {
+ "epoch": 63.17891373801917,
+ "grad_norm": 0.08154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0208,
+ "step": 19775
+ },
+ {
+ "epoch": 63.18210862619808,
+ "grad_norm": 0.11279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0149,
+ "step": 19776
+ },
+ {
+ "epoch": 63.185303514377,
+ "grad_norm": 0.1005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0275,
+ "step": 19777
+ },
+ {
+ "epoch": 63.18849840255591,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0186,
+ "step": 19778
+ },
+ {
+ "epoch": 63.191693290734825,
+ "grad_norm": 0.125,
+ "learning_rate": 0.0005,
+ "loss": 1.0263,
+ "step": 19779
+ },
+ {
+ "epoch": 63.194888178913736,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0185,
+ "step": 19780
+ },
+ {
+ "epoch": 63.198083067092654,
+ "grad_norm": 0.10986328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0222,
+ "step": 19781
+ },
+ {
+ "epoch": 63.201277955271564,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0341,
+ "step": 19782
+ },
+ {
+ "epoch": 63.20447284345048,
+ "grad_norm": 0.177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0312,
+ "step": 19783
+ },
+ {
+ "epoch": 63.20766773162939,
+ "grad_norm": 0.0908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.01,
+ "step": 19784
+ },
+ {
+ "epoch": 63.21086261980831,
+ "grad_norm": 0.2080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.027,
+ "step": 19785
+ },
+ {
+ "epoch": 63.21405750798722,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0228,
+ "step": 19786
+ },
+ {
+ "epoch": 63.21725239616613,
+ "grad_norm": 0.169921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0272,
+ "step": 19787
+ },
+ {
+ "epoch": 63.22044728434505,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0219,
+ "step": 19788
+ },
+ {
+ "epoch": 63.22364217252396,
+ "grad_norm": 0.140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0138,
+ "step": 19789
+ },
+ {
+ "epoch": 63.22683706070288,
+ "grad_norm": 0.08251953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0254,
+ "step": 19790
+ },
+ {
+ "epoch": 63.23003194888179,
+ "grad_norm": 0.1640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0253,
+ "step": 19791
+ },
+ {
+ "epoch": 63.233226837060705,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0243,
+ "step": 19792
+ },
+ {
+ "epoch": 63.236421725239616,
+ "grad_norm": 0.1552734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0201,
+ "step": 19793
+ },
+ {
+ "epoch": 63.239616613418534,
+ "grad_norm": 0.1220703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0363,
+ "step": 19794
+ },
+ {
+ "epoch": 63.242811501597444,
+ "grad_norm": 0.2216796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0172,
+ "step": 19795
+ },
+ {
+ "epoch": 63.246006389776355,
+ "grad_norm": 0.1435546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0213,
+ "step": 19796
+ },
+ {
+ "epoch": 63.24920127795527,
+ "grad_norm": 0.359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0237,
+ "step": 19797
+ },
+ {
+ "epoch": 63.25239616613418,
+ "grad_norm": 0.047607421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0321,
+ "step": 19798
+ },
+ {
+ "epoch": 63.2555910543131,
+ "grad_norm": 0.345703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0258,
+ "step": 19799
+ },
+ {
+ "epoch": 63.25878594249201,
+ "grad_norm": 0.1142578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0308,
+ "step": 19800
+ },
+ {
+ "epoch": 63.26198083067093,
+ "grad_norm": 0.2236328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0236,
+ "step": 19801
+ },
+ {
+ "epoch": 63.26517571884984,
+ "grad_norm": 0.140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0154,
+ "step": 19802
+ },
+ {
+ "epoch": 63.26837060702876,
+ "grad_norm": 0.1923828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0288,
+ "step": 19803
+ },
+ {
+ "epoch": 63.27156549520767,
+ "grad_norm": 0.2177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.026,
+ "step": 19804
+ },
+ {
+ "epoch": 63.27476038338658,
+ "grad_norm": 0.150390625,
+ "learning_rate": 0.0005,
+ "loss": 1.03,
+ "step": 19805
+ },
+ {
+ "epoch": 63.277955271565496,
+ "grad_norm": 0.2392578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0117,
+ "step": 19806
+ },
+ {
+ "epoch": 63.281150159744406,
+ "grad_norm": 0.1357421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0264,
+ "step": 19807
+ },
+ {
+ "epoch": 63.284345047923324,
+ "grad_norm": 0.203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0219,
+ "step": 19808
+ },
+ {
+ "epoch": 63.287539936102235,
+ "grad_norm": 0.1474609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0131,
+ "step": 19809
+ },
+ {
+ "epoch": 63.29073482428115,
+ "grad_norm": 0.169921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0164,
+ "step": 19810
+ },
+ {
+ "epoch": 63.29392971246006,
+ "grad_norm": 0.1962890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0195,
+ "step": 19811
+ },
+ {
+ "epoch": 63.29712460063898,
+ "grad_norm": 0.24609375,
+ "learning_rate": 0.0005,
+ "loss": 1.019,
+ "step": 19812
+ },
+ {
+ "epoch": 63.30031948881789,
+ "grad_norm": 0.1953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0279,
+ "step": 19813
+ },
+ {
+ "epoch": 63.3035143769968,
+ "grad_norm": 0.41015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0241,
+ "step": 19814
+ },
+ {
+ "epoch": 63.30670926517572,
+ "grad_norm": 0.08544921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0279,
+ "step": 19815
+ },
+ {
+ "epoch": 63.30990415335463,
+ "grad_norm": 0.42578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0211,
+ "step": 19816
+ },
+ {
+ "epoch": 63.31309904153355,
+ "grad_norm": 0.0556640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0232,
+ "step": 19817
+ },
+ {
+ "epoch": 63.31629392971246,
+ "grad_norm": 0.31640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0176,
+ "step": 19818
+ },
+ {
+ "epoch": 63.319488817891376,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0306,
+ "step": 19819
+ },
+ {
+ "epoch": 63.322683706070286,
+ "grad_norm": 0.251953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0216,
+ "step": 19820
+ },
+ {
+ "epoch": 63.325878594249204,
+ "grad_norm": 0.08642578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0209,
+ "step": 19821
+ },
+ {
+ "epoch": 63.329073482428115,
+ "grad_norm": 0.2216796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0249,
+ "step": 19822
+ },
+ {
+ "epoch": 63.33226837060703,
+ "grad_norm": 0.11279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0361,
+ "step": 19823
+ },
+ {
+ "epoch": 63.33546325878594,
+ "grad_norm": 0.2392578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0201,
+ "step": 19824
+ },
+ {
+ "epoch": 63.33865814696485,
+ "grad_norm": 0.1552734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0296,
+ "step": 19825
+ },
+ {
+ "epoch": 63.34185303514377,
+ "grad_norm": 0.421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0294,
+ "step": 19826
+ },
+ {
+ "epoch": 63.34504792332268,
+ "grad_norm": 0.1201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0237,
+ "step": 19827
+ },
+ {
+ "epoch": 63.3482428115016,
+ "grad_norm": 0.3203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0219,
+ "step": 19828
+ },
+ {
+ "epoch": 63.35143769968051,
+ "grad_norm": 0.2138671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0182,
+ "step": 19829
+ },
+ {
+ "epoch": 63.35463258785943,
+ "grad_norm": 0.2177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0256,
+ "step": 19830
+ },
+ {
+ "epoch": 63.35782747603834,
+ "grad_norm": 0.169921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0217,
+ "step": 19831
+ },
+ {
+ "epoch": 63.361022364217256,
+ "grad_norm": 0.2294921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0262,
+ "step": 19832
+ },
+ {
+ "epoch": 63.364217252396166,
+ "grad_norm": 0.1376953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0208,
+ "step": 19833
+ },
+ {
+ "epoch": 63.36741214057508,
+ "grad_norm": 0.2001953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0242,
+ "step": 19834
+ },
+ {
+ "epoch": 63.370607028753994,
+ "grad_norm": 0.0498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0331,
+ "step": 19835
+ },
+ {
+ "epoch": 63.373801916932905,
+ "grad_norm": 0.193359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0124,
+ "step": 19836
+ },
+ {
+ "epoch": 63.37699680511182,
+ "grad_norm": 0.12158203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0177,
+ "step": 19837
+ },
+ {
+ "epoch": 63.38019169329073,
+ "grad_norm": 0.1484375,
+ "learning_rate": 0.0005,
+ "loss": 1.029,
+ "step": 19838
+ },
+ {
+ "epoch": 63.38338658146965,
+ "grad_norm": 0.1572265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0285,
+ "step": 19839
+ },
+ {
+ "epoch": 63.38658146964856,
+ "grad_norm": 0.0849609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0126,
+ "step": 19840
+ },
+ {
+ "epoch": 63.38977635782748,
+ "grad_norm": 0.146484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0155,
+ "step": 19841
+ },
+ {
+ "epoch": 63.39297124600639,
+ "grad_norm": 0.12451171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0239,
+ "step": 19842
+ },
+ {
+ "epoch": 63.3961661341853,
+ "grad_norm": 0.1103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0232,
+ "step": 19843
+ },
+ {
+ "epoch": 63.39936102236422,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0266,
+ "step": 19844
+ },
+ {
+ "epoch": 63.40255591054313,
+ "grad_norm": 0.08837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0238,
+ "step": 19845
+ },
+ {
+ "epoch": 63.405750798722046,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0379,
+ "step": 19846
+ },
+ {
+ "epoch": 63.40894568690096,
+ "grad_norm": 0.060791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0287,
+ "step": 19847
+ },
+ {
+ "epoch": 63.412140575079874,
+ "grad_norm": 0.0888671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0219,
+ "step": 19848
+ },
+ {
+ "epoch": 63.415335463258785,
+ "grad_norm": 0.04931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.02,
+ "step": 19849
+ },
+ {
+ "epoch": 63.4185303514377,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0134,
+ "step": 19850
+ },
+ {
+ "epoch": 63.42172523961661,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0258,
+ "step": 19851
+ },
+ {
+ "epoch": 63.424920127795524,
+ "grad_norm": 0.0810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.021,
+ "step": 19852
+ },
+ {
+ "epoch": 63.42811501597444,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0346,
+ "step": 19853
+ },
+ {
+ "epoch": 63.43130990415335,
+ "grad_norm": 0.08935546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0175,
+ "step": 19854
+ },
+ {
+ "epoch": 63.43450479233227,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0185,
+ "step": 19855
+ },
+ {
+ "epoch": 63.43769968051118,
+ "grad_norm": 0.060302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0164,
+ "step": 19856
+ },
+ {
+ "epoch": 63.4408945686901,
+ "grad_norm": 0.0869140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0267,
+ "step": 19857
+ },
+ {
+ "epoch": 63.44408945686901,
+ "grad_norm": 0.046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0267,
+ "step": 19858
+ },
+ {
+ "epoch": 63.447284345047926,
+ "grad_norm": 0.05712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0391,
+ "step": 19859
+ },
+ {
+ "epoch": 63.45047923322684,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0218,
+ "step": 19860
+ },
+ {
+ "epoch": 63.453674121405754,
+ "grad_norm": 0.0576171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0319,
+ "step": 19861
+ },
+ {
+ "epoch": 63.456869009584665,
+ "grad_norm": 0.049560546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0263,
+ "step": 19862
+ },
+ {
+ "epoch": 63.460063897763575,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.029,
+ "step": 19863
+ },
+ {
+ "epoch": 63.46325878594249,
+ "grad_norm": 0.0859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0256,
+ "step": 19864
+ },
+ {
+ "epoch": 63.466453674121404,
+ "grad_norm": 0.062255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0324,
+ "step": 19865
+ },
+ {
+ "epoch": 63.46964856230032,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0255,
+ "step": 19866
+ },
+ {
+ "epoch": 63.47284345047923,
+ "grad_norm": 0.048583984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0257,
+ "step": 19867
+ },
+ {
+ "epoch": 63.47603833865815,
+ "grad_norm": 0.06005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0211,
+ "step": 19868
+ },
+ {
+ "epoch": 63.47923322683706,
+ "grad_norm": 0.0546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0287,
+ "step": 19869
+ },
+ {
+ "epoch": 63.48242811501598,
+ "grad_norm": 0.049560546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0308,
+ "step": 19870
+ },
+ {
+ "epoch": 63.48562300319489,
+ "grad_norm": 0.0546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0251,
+ "step": 19871
+ },
+ {
+ "epoch": 63.4888178913738,
+ "grad_norm": 0.049560546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0079,
+ "step": 19872
+ },
+ {
+ "epoch": 63.49201277955272,
+ "grad_norm": 0.046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0236,
+ "step": 19873
+ },
+ {
+ "epoch": 63.49520766773163,
+ "grad_norm": 0.08984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0142,
+ "step": 19874
+ },
+ {
+ "epoch": 63.498402555910545,
+ "grad_norm": 0.04296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0296,
+ "step": 19875
+ },
+ {
+ "epoch": 63.501597444089455,
+ "grad_norm": 0.058837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0229,
+ "step": 19876
+ },
+ {
+ "epoch": 63.50479233226837,
+ "grad_norm": 0.08984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0278,
+ "step": 19877
+ },
+ {
+ "epoch": 63.50798722044728,
+ "grad_norm": 0.050048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0269,
+ "step": 19878
+ },
+ {
+ "epoch": 63.5111821086262,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0333,
+ "step": 19879
+ },
+ {
+ "epoch": 63.51437699680511,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0161,
+ "step": 19880
+ },
+ {
+ "epoch": 63.51757188498402,
+ "grad_norm": 0.055908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0262,
+ "step": 19881
+ },
+ {
+ "epoch": 63.52076677316294,
+ "grad_norm": 0.0556640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0315,
+ "step": 19882
+ },
+ {
+ "epoch": 63.52396166134185,
+ "grad_norm": 0.05224609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0338,
+ "step": 19883
+ },
+ {
+ "epoch": 63.52715654952077,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0339,
+ "step": 19884
+ },
+ {
+ "epoch": 63.53035143769968,
+ "grad_norm": 0.05810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0189,
+ "step": 19885
+ },
+ {
+ "epoch": 63.533546325878596,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0175,
+ "step": 19886
+ },
+ {
+ "epoch": 63.53674121405751,
+ "grad_norm": 0.06201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0258,
+ "step": 19887
+ },
+ {
+ "epoch": 63.539936102236425,
+ "grad_norm": 0.057373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0355,
+ "step": 19888
+ },
+ {
+ "epoch": 63.543130990415335,
+ "grad_norm": 0.061279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0211,
+ "step": 19889
+ },
+ {
+ "epoch": 63.546325878594246,
+ "grad_norm": 0.08642578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0238,
+ "step": 19890
+ },
+ {
+ "epoch": 63.54952076677316,
+ "grad_norm": 0.08642578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0353,
+ "step": 19891
+ },
+ {
+ "epoch": 63.552715654952074,
+ "grad_norm": 0.05322265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0304,
+ "step": 19892
+ },
+ {
+ "epoch": 63.55591054313099,
+ "grad_norm": 0.05712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.021,
+ "step": 19893
+ },
+ {
+ "epoch": 63.5591054313099,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0314,
+ "step": 19894
+ },
+ {
+ "epoch": 63.56230031948882,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0258,
+ "step": 19895
+ },
+ {
+ "epoch": 63.56549520766773,
+ "grad_norm": 0.05908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0266,
+ "step": 19896
+ },
+ {
+ "epoch": 63.56869009584665,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0298,
+ "step": 19897
+ },
+ {
+ "epoch": 63.57188498402556,
+ "grad_norm": 0.05859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0323,
+ "step": 19898
+ },
+ {
+ "epoch": 63.575079872204476,
+ "grad_norm": 0.0537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0221,
+ "step": 19899
+ },
+ {
+ "epoch": 63.57827476038339,
+ "grad_norm": 0.058349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.012,
+ "step": 19900
+ },
+ {
+ "epoch": 63.5814696485623,
+ "grad_norm": 0.0576171875,
+ "learning_rate": 0.0005,
+ "loss": 1.026,
+ "step": 19901
+ },
+ {
+ "epoch": 63.584664536741215,
+ "grad_norm": 0.057373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0191,
+ "step": 19902
+ },
+ {
+ "epoch": 63.587859424920126,
+ "grad_norm": 0.080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.022,
+ "step": 19903
+ },
+ {
+ "epoch": 63.59105431309904,
+ "grad_norm": 0.08203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0234,
+ "step": 19904
+ },
+ {
+ "epoch": 63.594249201277954,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0218,
+ "step": 19905
+ },
+ {
+ "epoch": 63.59744408945687,
+ "grad_norm": 0.048095703125,
+ "learning_rate": 0.0005,
+ "loss": 1.031,
+ "step": 19906
+ },
+ {
+ "epoch": 63.60063897763578,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0234,
+ "step": 19907
+ },
+ {
+ "epoch": 63.6038338658147,
+ "grad_norm": 0.0615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0144,
+ "step": 19908
+ },
+ {
+ "epoch": 63.60702875399361,
+ "grad_norm": 0.0595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.02,
+ "step": 19909
+ },
+ {
+ "epoch": 63.61022364217252,
+ "grad_norm": 0.058837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0187,
+ "step": 19910
+ },
+ {
+ "epoch": 63.61341853035144,
+ "grad_norm": 0.062255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0234,
+ "step": 19911
+ },
+ {
+ "epoch": 63.61661341853035,
+ "grad_norm": 0.057373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0282,
+ "step": 19912
+ },
+ {
+ "epoch": 63.61980830670927,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0272,
+ "step": 19913
+ },
+ {
+ "epoch": 63.62300319488818,
+ "grad_norm": 0.0517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0221,
+ "step": 19914
+ },
+ {
+ "epoch": 63.626198083067095,
+ "grad_norm": 0.09033203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0239,
+ "step": 19915
+ },
+ {
+ "epoch": 63.629392971246006,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0308,
+ "step": 19916
+ },
+ {
+ "epoch": 63.63258785942492,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0281,
+ "step": 19917
+ },
+ {
+ "epoch": 63.635782747603834,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0288,
+ "step": 19918
+ },
+ {
+ "epoch": 63.638977635782744,
+ "grad_norm": 0.0517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0303,
+ "step": 19919
+ },
+ {
+ "epoch": 63.64217252396166,
+ "grad_norm": 0.05712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.02,
+ "step": 19920
+ },
+ {
+ "epoch": 63.64536741214057,
+ "grad_norm": 0.050048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.019,
+ "step": 19921
+ },
+ {
+ "epoch": 63.64856230031949,
+ "grad_norm": 0.04931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0231,
+ "step": 19922
+ },
+ {
+ "epoch": 63.6517571884984,
+ "grad_norm": 0.049560546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0252,
+ "step": 19923
+ },
+ {
+ "epoch": 63.65495207667732,
+ "grad_norm": 0.050537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0163,
+ "step": 19924
+ },
+ {
+ "epoch": 63.65814696485623,
+ "grad_norm": 0.05224609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0317,
+ "step": 19925
+ },
+ {
+ "epoch": 63.66134185303515,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0331,
+ "step": 19926
+ },
+ {
+ "epoch": 63.66453674121406,
+ "grad_norm": 0.055908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0217,
+ "step": 19927
+ },
+ {
+ "epoch": 63.66773162939297,
+ "grad_norm": 0.044921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0156,
+ "step": 19928
+ },
+ {
+ "epoch": 63.670926517571885,
+ "grad_norm": 0.047607421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0167,
+ "step": 19929
+ },
+ {
+ "epoch": 63.674121405750796,
+ "grad_norm": 0.044677734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0234,
+ "step": 19930
+ },
+ {
+ "epoch": 63.677316293929714,
+ "grad_norm": 0.050048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0232,
+ "step": 19931
+ },
+ {
+ "epoch": 63.680511182108624,
+ "grad_norm": 0.056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0275,
+ "step": 19932
+ },
+ {
+ "epoch": 63.68370607028754,
+ "grad_norm": 0.05126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0262,
+ "step": 19933
+ },
+ {
+ "epoch": 63.68690095846645,
+ "grad_norm": 0.049560546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0138,
+ "step": 19934
+ },
+ {
+ "epoch": 63.69009584664537,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0267,
+ "step": 19935
+ },
+ {
+ "epoch": 63.69329073482428,
+ "grad_norm": 0.052978515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0233,
+ "step": 19936
+ },
+ {
+ "epoch": 63.6964856230032,
+ "grad_norm": 0.047607421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0239,
+ "step": 19937
+ },
+ {
+ "epoch": 63.69968051118211,
+ "grad_norm": 0.0615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0258,
+ "step": 19938
+ },
+ {
+ "epoch": 63.70287539936102,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0151,
+ "step": 19939
+ },
+ {
+ "epoch": 63.70607028753994,
+ "grad_norm": 0.05322265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0149,
+ "step": 19940
+ },
+ {
+ "epoch": 63.70926517571885,
+ "grad_norm": 0.0849609375,
+ "learning_rate": 0.0005,
+ "loss": 1.029,
+ "step": 19941
+ },
+ {
+ "epoch": 63.712460063897765,
+ "grad_norm": 0.09619140625,
+ "learning_rate": 0.0005,
+ "loss": 1.03,
+ "step": 19942
+ },
+ {
+ "epoch": 63.715654952076676,
+ "grad_norm": 0.06005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0361,
+ "step": 19943
+ },
+ {
+ "epoch": 63.718849840255594,
+ "grad_norm": 0.08056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0247,
+ "step": 19944
+ },
+ {
+ "epoch": 63.722044728434504,
+ "grad_norm": 0.0888671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0196,
+ "step": 19945
+ },
+ {
+ "epoch": 63.72523961661342,
+ "grad_norm": 0.0546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0239,
+ "step": 19946
+ },
+ {
+ "epoch": 63.72843450479233,
+ "grad_norm": 0.10302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0283,
+ "step": 19947
+ },
+ {
+ "epoch": 63.73162939297124,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0353,
+ "step": 19948
+ },
+ {
+ "epoch": 63.73482428115016,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.01,
+ "step": 19949
+ },
+ {
+ "epoch": 63.73801916932907,
+ "grad_norm": 0.05419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0185,
+ "step": 19950
+ },
+ {
+ "epoch": 63.74121405750799,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0183,
+ "step": 19951
+ },
+ {
+ "epoch": 63.7444089456869,
+ "grad_norm": 0.06201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0233,
+ "step": 19952
+ },
+ {
+ "epoch": 63.74760383386582,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0183,
+ "step": 19953
+ },
+ {
+ "epoch": 63.75079872204473,
+ "grad_norm": 0.052734375,
+ "learning_rate": 0.0005,
+ "loss": 1.023,
+ "step": 19954
+ },
+ {
+ "epoch": 63.753993610223645,
+ "grad_norm": 0.05517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0224,
+ "step": 19955
+ },
+ {
+ "epoch": 63.757188498402556,
+ "grad_norm": 0.057373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0288,
+ "step": 19956
+ },
+ {
+ "epoch": 63.760383386581466,
+ "grad_norm": 0.05859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0148,
+ "step": 19957
+ },
+ {
+ "epoch": 63.763578274760384,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0158,
+ "step": 19958
+ },
+ {
+ "epoch": 63.766773162939295,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0136,
+ "step": 19959
+ },
+ {
+ "epoch": 63.76996805111821,
+ "grad_norm": 0.05029296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0312,
+ "step": 19960
+ },
+ {
+ "epoch": 63.77316293929712,
+ "grad_norm": 0.05078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0245,
+ "step": 19961
+ },
+ {
+ "epoch": 63.77635782747604,
+ "grad_norm": 0.057373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0213,
+ "step": 19962
+ },
+ {
+ "epoch": 63.77955271565495,
+ "grad_norm": 0.05859375,
+ "learning_rate": 0.0005,
+ "loss": 1.025,
+ "step": 19963
+ },
+ {
+ "epoch": 63.78274760383387,
+ "grad_norm": 0.048095703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0148,
+ "step": 19964
+ },
+ {
+ "epoch": 63.78594249201278,
+ "grad_norm": 0.0517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0202,
+ "step": 19965
+ },
+ {
+ "epoch": 63.78913738019169,
+ "grad_norm": 0.053466796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0198,
+ "step": 19966
+ },
+ {
+ "epoch": 63.79233226837061,
+ "grad_norm": 0.0537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0177,
+ "step": 19967
+ },
+ {
+ "epoch": 63.79552715654952,
+ "grad_norm": 0.0537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0335,
+ "step": 19968
+ },
+ {
+ "epoch": 63.798722044728436,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0159,
+ "step": 19969
+ },
+ {
+ "epoch": 63.801916932907346,
+ "grad_norm": 0.059814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0241,
+ "step": 19970
+ },
+ {
+ "epoch": 63.805111821086264,
+ "grad_norm": 0.046142578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0183,
+ "step": 19971
+ },
+ {
+ "epoch": 63.808306709265175,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0295,
+ "step": 19972
+ },
+ {
+ "epoch": 63.81150159744409,
+ "grad_norm": 0.04833984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0381,
+ "step": 19973
+ },
+ {
+ "epoch": 63.814696485623,
+ "grad_norm": 0.04345703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0271,
+ "step": 19974
+ },
+ {
+ "epoch": 63.81789137380191,
+ "grad_norm": 0.05029296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0242,
+ "step": 19975
+ },
+ {
+ "epoch": 63.82108626198083,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0281,
+ "step": 19976
+ },
+ {
+ "epoch": 63.82428115015974,
+ "grad_norm": 0.056884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0196,
+ "step": 19977
+ },
+ {
+ "epoch": 63.82747603833866,
+ "grad_norm": 0.055419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0291,
+ "step": 19978
+ },
+ {
+ "epoch": 63.83067092651757,
+ "grad_norm": 0.0537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0232,
+ "step": 19979
+ },
+ {
+ "epoch": 63.83386581469649,
+ "grad_norm": 0.049560546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0171,
+ "step": 19980
+ },
+ {
+ "epoch": 63.8370607028754,
+ "grad_norm": 0.05810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0165,
+ "step": 19981
+ },
+ {
+ "epoch": 63.840255591054316,
+ "grad_norm": 0.0517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0287,
+ "step": 19982
+ },
+ {
+ "epoch": 63.843450479233226,
+ "grad_norm": 0.05029296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0258,
+ "step": 19983
+ },
+ {
+ "epoch": 63.846645367412144,
+ "grad_norm": 0.056396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.023,
+ "step": 19984
+ },
+ {
+ "epoch": 63.849840255591054,
+ "grad_norm": 0.060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0189,
+ "step": 19985
+ },
+ {
+ "epoch": 63.853035143769965,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0257,
+ "step": 19986
+ },
+ {
+ "epoch": 63.85623003194888,
+ "grad_norm": 0.06201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.016,
+ "step": 19987
+ },
+ {
+ "epoch": 63.85942492012779,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0236,
+ "step": 19988
+ },
+ {
+ "epoch": 63.86261980830671,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0324,
+ "step": 19989
+ },
+ {
+ "epoch": 63.86581469648562,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0194,
+ "step": 19990
+ },
+ {
+ "epoch": 63.86900958466454,
+ "grad_norm": 0.044189453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0188,
+ "step": 19991
+ },
+ {
+ "epoch": 63.87220447284345,
+ "grad_norm": 0.0537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0301,
+ "step": 19992
+ },
+ {
+ "epoch": 63.87539936102237,
+ "grad_norm": 0.0439453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0245,
+ "step": 19993
+ },
+ {
+ "epoch": 63.87859424920128,
+ "grad_norm": 0.059814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0221,
+ "step": 19994
+ },
+ {
+ "epoch": 63.88178913738019,
+ "grad_norm": 0.046630859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0131,
+ "step": 19995
+ },
+ {
+ "epoch": 63.884984025559106,
+ "grad_norm": 0.051025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0227,
+ "step": 19996
+ },
+ {
+ "epoch": 63.88817891373802,
+ "grad_norm": 0.0498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0301,
+ "step": 19997
+ },
+ {
+ "epoch": 63.891373801916934,
+ "grad_norm": 0.062255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0357,
+ "step": 19998
+ },
+ {
+ "epoch": 63.894568690095845,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0262,
+ "step": 19999
+ },
+ {
+ "epoch": 63.89776357827476,
+ "grad_norm": 0.0771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0114,
+ "step": 20000
+ },
+ {
+ "epoch": 63.90095846645367,
+ "grad_norm": 0.048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0296,
+ "step": 20001
+ },
+ {
+ "epoch": 63.90415335463259,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.023,
+ "step": 20002
+ },
+ {
+ "epoch": 63.9073482428115,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0253,
+ "step": 20003
+ },
+ {
+ "epoch": 63.91054313099041,
+ "grad_norm": 0.057373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0271,
+ "step": 20004
+ },
+ {
+ "epoch": 63.91373801916933,
+ "grad_norm": 0.05712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0269,
+ "step": 20005
+ },
+ {
+ "epoch": 63.91693290734824,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0201,
+ "step": 20006
+ },
+ {
+ "epoch": 63.92012779552716,
+ "grad_norm": 0.06005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0254,
+ "step": 20007
+ },
+ {
+ "epoch": 63.92332268370607,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0216,
+ "step": 20008
+ },
+ {
+ "epoch": 63.926517571884986,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0282,
+ "step": 20009
+ },
+ {
+ "epoch": 63.9297124600639,
+ "grad_norm": 0.05517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0256,
+ "step": 20010
+ },
+ {
+ "epoch": 63.932907348242814,
+ "grad_norm": 0.0556640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0261,
+ "step": 20011
+ },
+ {
+ "epoch": 63.936102236421725,
+ "grad_norm": 0.057373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0205,
+ "step": 20012
+ },
+ {
+ "epoch": 63.93929712460064,
+ "grad_norm": 0.0498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0169,
+ "step": 20013
+ },
+ {
+ "epoch": 63.94249201277955,
+ "grad_norm": 0.05322265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0285,
+ "step": 20014
+ },
+ {
+ "epoch": 63.945686900958464,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0288,
+ "step": 20015
+ },
+ {
+ "epoch": 63.94888178913738,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0219,
+ "step": 20016
+ },
+ {
+ "epoch": 63.95207667731629,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0215,
+ "step": 20017
+ },
+ {
+ "epoch": 63.95527156549521,
+ "grad_norm": 0.1103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0221,
+ "step": 20018
+ },
+ {
+ "epoch": 63.95846645367412,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0272,
+ "step": 20019
+ },
+ {
+ "epoch": 63.96166134185304,
+ "grad_norm": 0.09619140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0296,
+ "step": 20020
+ },
+ {
+ "epoch": 63.96485623003195,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0316,
+ "step": 20021
+ },
+ {
+ "epoch": 63.968051118210866,
+ "grad_norm": 0.052978515625,
+ "learning_rate": 0.0005,
+ "loss": 1.023,
+ "step": 20022
+ },
+ {
+ "epoch": 63.97124600638978,
+ "grad_norm": 0.1103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0195,
+ "step": 20023
+ },
+ {
+ "epoch": 63.97444089456869,
+ "grad_norm": 0.09765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0192,
+ "step": 20024
+ },
+ {
+ "epoch": 63.977635782747605,
+ "grad_norm": 0.14453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0201,
+ "step": 20025
+ },
+ {
+ "epoch": 63.980830670926515,
+ "grad_norm": 0.10888671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0226,
+ "step": 20026
+ },
+ {
+ "epoch": 63.98402555910543,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0217,
+ "step": 20027
+ },
+ {
+ "epoch": 63.98722044728434,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0139,
+ "step": 20028
+ },
+ {
+ "epoch": 63.99041533546326,
+ "grad_norm": 0.05322265625,
+ "learning_rate": 0.0005,
+ "loss": 1.035,
+ "step": 20029
+ },
+ {
+ "epoch": 63.99361022364217,
+ "grad_norm": 0.05712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0242,
+ "step": 20030
+ },
+ {
+ "epoch": 63.99680511182109,
+ "grad_norm": 0.054443359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0175,
+ "step": 20031
+ },
+ {
+ "epoch": 64.0,
+ "grad_norm": 0.055908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0302,
+ "step": 20032
+ },
+ {
+ "epoch": 64.00319488817891,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0215,
+ "step": 20033
+ },
+ {
+ "epoch": 64.00638977635782,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0184,
+ "step": 20034
+ },
+ {
+ "epoch": 64.00958466453675,
+ "grad_norm": 0.050537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0283,
+ "step": 20035
+ },
+ {
+ "epoch": 64.01277955271566,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0156,
+ "step": 20036
+ },
+ {
+ "epoch": 64.01597444089457,
+ "grad_norm": 0.08447265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0203,
+ "step": 20037
+ },
+ {
+ "epoch": 64.01916932907348,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0286,
+ "step": 20038
+ },
+ {
+ "epoch": 64.0223642172524,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0291,
+ "step": 20039
+ },
+ {
+ "epoch": 64.02555910543131,
+ "grad_norm": 0.048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0285,
+ "step": 20040
+ },
+ {
+ "epoch": 64.02875399361022,
+ "grad_norm": 0.05615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0225,
+ "step": 20041
+ },
+ {
+ "epoch": 64.03194888178913,
+ "grad_norm": 0.057373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0236,
+ "step": 20042
+ },
+ {
+ "epoch": 64.03514376996804,
+ "grad_norm": 0.053955078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0247,
+ "step": 20043
+ },
+ {
+ "epoch": 64.03833865814697,
+ "grad_norm": 0.052001953125,
+ "learning_rate": 0.0005,
+ "loss": 1.023,
+ "step": 20044
+ },
+ {
+ "epoch": 64.04153354632588,
+ "grad_norm": 0.05419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0264,
+ "step": 20045
+ },
+ {
+ "epoch": 64.04472843450479,
+ "grad_norm": 0.045654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0278,
+ "step": 20046
+ },
+ {
+ "epoch": 64.0479233226837,
+ "grad_norm": 0.05029296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0264,
+ "step": 20047
+ },
+ {
+ "epoch": 64.05111821086263,
+ "grad_norm": 0.083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0313,
+ "step": 20048
+ },
+ {
+ "epoch": 64.05431309904154,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0189,
+ "step": 20049
+ },
+ {
+ "epoch": 64.05750798722045,
+ "grad_norm": 0.05419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.017,
+ "step": 20050
+ },
+ {
+ "epoch": 64.06070287539936,
+ "grad_norm": 0.1171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0193,
+ "step": 20051
+ },
+ {
+ "epoch": 64.06389776357827,
+ "grad_norm": 0.05615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.02,
+ "step": 20052
+ },
+ {
+ "epoch": 64.06709265175719,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0221,
+ "step": 20053
+ },
+ {
+ "epoch": 64.0702875399361,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0363,
+ "step": 20054
+ },
+ {
+ "epoch": 64.07348242811501,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0128,
+ "step": 20055
+ },
+ {
+ "epoch": 64.07667731629392,
+ "grad_norm": 0.058349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0211,
+ "step": 20056
+ },
+ {
+ "epoch": 64.07987220447285,
+ "grad_norm": 0.130859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0344,
+ "step": 20057
+ },
+ {
+ "epoch": 64.08306709265176,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0176,
+ "step": 20058
+ },
+ {
+ "epoch": 64.08626198083067,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0305,
+ "step": 20059
+ },
+ {
+ "epoch": 64.08945686900958,
+ "grad_norm": 0.11279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0218,
+ "step": 20060
+ },
+ {
+ "epoch": 64.09265175718849,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0234,
+ "step": 20061
+ },
+ {
+ "epoch": 64.09584664536742,
+ "grad_norm": 0.05517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0098,
+ "step": 20062
+ },
+ {
+ "epoch": 64.09904153354633,
+ "grad_norm": 0.08642578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0243,
+ "step": 20063
+ },
+ {
+ "epoch": 64.10223642172524,
+ "grad_norm": 0.1044921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0284,
+ "step": 20064
+ },
+ {
+ "epoch": 64.10543130990415,
+ "grad_norm": 0.0595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0331,
+ "step": 20065
+ },
+ {
+ "epoch": 64.10862619808307,
+ "grad_norm": 0.052734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0243,
+ "step": 20066
+ },
+ {
+ "epoch": 64.11182108626198,
+ "grad_norm": 0.04931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.022,
+ "step": 20067
+ },
+ {
+ "epoch": 64.1150159744409,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0276,
+ "step": 20068
+ },
+ {
+ "epoch": 64.1182108626198,
+ "grad_norm": 0.0517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0235,
+ "step": 20069
+ },
+ {
+ "epoch": 64.12140575079871,
+ "grad_norm": 0.08935546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0221,
+ "step": 20070
+ },
+ {
+ "epoch": 64.12460063897764,
+ "grad_norm": 0.05078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0308,
+ "step": 20071
+ },
+ {
+ "epoch": 64.12779552715655,
+ "grad_norm": 0.11181640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0163,
+ "step": 20072
+ },
+ {
+ "epoch": 64.13099041533546,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0363,
+ "step": 20073
+ },
+ {
+ "epoch": 64.13418530351437,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0296,
+ "step": 20074
+ },
+ {
+ "epoch": 64.1373801916933,
+ "grad_norm": 0.087890625,
+ "learning_rate": 0.0005,
+ "loss": 1.014,
+ "step": 20075
+ },
+ {
+ "epoch": 64.1405750798722,
+ "grad_norm": 0.053466796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0239,
+ "step": 20076
+ },
+ {
+ "epoch": 64.14376996805112,
+ "grad_norm": 0.10400390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0367,
+ "step": 20077
+ },
+ {
+ "epoch": 64.14696485623003,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0187,
+ "step": 20078
+ },
+ {
+ "epoch": 64.15015974440895,
+ "grad_norm": 0.05029296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0241,
+ "step": 20079
+ },
+ {
+ "epoch": 64.15335463258786,
+ "grad_norm": 0.08203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0233,
+ "step": 20080
+ },
+ {
+ "epoch": 64.15654952076677,
+ "grad_norm": 0.0546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0114,
+ "step": 20081
+ },
+ {
+ "epoch": 64.15974440894568,
+ "grad_norm": 0.0927734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0272,
+ "step": 20082
+ },
+ {
+ "epoch": 64.1629392971246,
+ "grad_norm": 0.044921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0143,
+ "step": 20083
+ },
+ {
+ "epoch": 64.16613418530352,
+ "grad_norm": 0.08154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0135,
+ "step": 20084
+ },
+ {
+ "epoch": 64.16932907348243,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.027,
+ "step": 20085
+ },
+ {
+ "epoch": 64.17252396166134,
+ "grad_norm": 0.087890625,
+ "learning_rate": 0.0005,
+ "loss": 1.024,
+ "step": 20086
+ },
+ {
+ "epoch": 64.17571884984025,
+ "grad_norm": 0.08642578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0188,
+ "step": 20087
+ },
+ {
+ "epoch": 64.17891373801918,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0269,
+ "step": 20088
+ },
+ {
+ "epoch": 64.18210862619809,
+ "grad_norm": 0.10009765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0225,
+ "step": 20089
+ },
+ {
+ "epoch": 64.185303514377,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0166,
+ "step": 20090
+ },
+ {
+ "epoch": 64.18849840255591,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.012,
+ "step": 20091
+ },
+ {
+ "epoch": 64.19169329073482,
+ "grad_norm": 0.083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0176,
+ "step": 20092
+ },
+ {
+ "epoch": 64.19488817891374,
+ "grad_norm": 0.10205078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0106,
+ "step": 20093
+ },
+ {
+ "epoch": 64.19808306709265,
+ "grad_norm": 0.10107421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0186,
+ "step": 20094
+ },
+ {
+ "epoch": 64.20127795527156,
+ "grad_norm": 0.05615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0236,
+ "step": 20095
+ },
+ {
+ "epoch": 64.20447284345047,
+ "grad_norm": 0.142578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0392,
+ "step": 20096
+ },
+ {
+ "epoch": 64.2076677316294,
+ "grad_norm": 0.060302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0326,
+ "step": 20097
+ },
+ {
+ "epoch": 64.21086261980831,
+ "grad_norm": 0.10498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0228,
+ "step": 20098
+ },
+ {
+ "epoch": 64.21405750798722,
+ "grad_norm": 0.10302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.022,
+ "step": 20099
+ },
+ {
+ "epoch": 64.21725239616613,
+ "grad_norm": 0.09130859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0228,
+ "step": 20100
+ },
+ {
+ "epoch": 64.22044728434504,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0147,
+ "step": 20101
+ },
+ {
+ "epoch": 64.22364217252397,
+ "grad_norm": 0.045166015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0122,
+ "step": 20102
+ },
+ {
+ "epoch": 64.22683706070288,
+ "grad_norm": 0.09521484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0145,
+ "step": 20103
+ },
+ {
+ "epoch": 64.23003194888179,
+ "grad_norm": 0.0478515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0273,
+ "step": 20104
+ },
+ {
+ "epoch": 64.2332268370607,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0237,
+ "step": 20105
+ },
+ {
+ "epoch": 64.23642172523962,
+ "grad_norm": 0.05419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0176,
+ "step": 20106
+ },
+ {
+ "epoch": 64.23961661341853,
+ "grad_norm": 0.045166015625,
+ "learning_rate": 0.0005,
+ "loss": 1.029,
+ "step": 20107
+ },
+ {
+ "epoch": 64.24281150159744,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0379,
+ "step": 20108
+ },
+ {
+ "epoch": 64.24600638977635,
+ "grad_norm": 0.057861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0177,
+ "step": 20109
+ },
+ {
+ "epoch": 64.24920127795527,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0274,
+ "step": 20110
+ },
+ {
+ "epoch": 64.25239616613419,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0185,
+ "step": 20111
+ },
+ {
+ "epoch": 64.2555910543131,
+ "grad_norm": 0.055419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0144,
+ "step": 20112
+ },
+ {
+ "epoch": 64.25878594249201,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0233,
+ "step": 20113
+ },
+ {
+ "epoch": 64.26198083067092,
+ "grad_norm": 0.08447265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0174,
+ "step": 20114
+ },
+ {
+ "epoch": 64.26517571884985,
+ "grad_norm": 0.060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0276,
+ "step": 20115
+ },
+ {
+ "epoch": 64.26837060702876,
+ "grad_norm": 0.06103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0263,
+ "step": 20116
+ },
+ {
+ "epoch": 64.27156549520767,
+ "grad_norm": 0.057861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0182,
+ "step": 20117
+ },
+ {
+ "epoch": 64.27476038338658,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.015,
+ "step": 20118
+ },
+ {
+ "epoch": 64.27795527156549,
+ "grad_norm": 0.056396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0246,
+ "step": 20119
+ },
+ {
+ "epoch": 64.28115015974441,
+ "grad_norm": 0.0830078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0177,
+ "step": 20120
+ },
+ {
+ "epoch": 64.28434504792332,
+ "grad_norm": 0.0458984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0082,
+ "step": 20121
+ },
+ {
+ "epoch": 64.28753993610223,
+ "grad_norm": 0.043212890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0226,
+ "step": 20122
+ },
+ {
+ "epoch": 64.29073482428115,
+ "grad_norm": 0.06005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0121,
+ "step": 20123
+ },
+ {
+ "epoch": 64.29392971246007,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0208,
+ "step": 20124
+ },
+ {
+ "epoch": 64.29712460063898,
+ "grad_norm": 0.047119140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0369,
+ "step": 20125
+ },
+ {
+ "epoch": 64.30031948881789,
+ "grad_norm": 0.09130859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0202,
+ "step": 20126
+ },
+ {
+ "epoch": 64.3035143769968,
+ "grad_norm": 0.09765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0246,
+ "step": 20127
+ },
+ {
+ "epoch": 64.30670926517571,
+ "grad_norm": 0.060791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.014,
+ "step": 20128
+ },
+ {
+ "epoch": 64.30990415335464,
+ "grad_norm": 0.08203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0177,
+ "step": 20129
+ },
+ {
+ "epoch": 64.31309904153355,
+ "grad_norm": 0.057861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0286,
+ "step": 20130
+ },
+ {
+ "epoch": 64.31629392971246,
+ "grad_norm": 0.050048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0201,
+ "step": 20131
+ },
+ {
+ "epoch": 64.31948881789137,
+ "grad_norm": 0.09521484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0318,
+ "step": 20132
+ },
+ {
+ "epoch": 64.3226837060703,
+ "grad_norm": 0.056396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0311,
+ "step": 20133
+ },
+ {
+ "epoch": 64.3258785942492,
+ "grad_norm": 0.0458984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0171,
+ "step": 20134
+ },
+ {
+ "epoch": 64.32907348242811,
+ "grad_norm": 0.061767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0215,
+ "step": 20135
+ },
+ {
+ "epoch": 64.33226837060703,
+ "grad_norm": 0.0849609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0182,
+ "step": 20136
+ },
+ {
+ "epoch": 64.33546325878594,
+ "grad_norm": 0.052001953125,
+ "learning_rate": 0.0005,
+ "loss": 1.01,
+ "step": 20137
+ },
+ {
+ "epoch": 64.33865814696486,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0259,
+ "step": 20138
+ },
+ {
+ "epoch": 64.34185303514377,
+ "grad_norm": 0.052001953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0189,
+ "step": 20139
+ },
+ {
+ "epoch": 64.34504792332268,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0169,
+ "step": 20140
+ },
+ {
+ "epoch": 64.34824281150159,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0127,
+ "step": 20141
+ },
+ {
+ "epoch": 64.35143769968052,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0283,
+ "step": 20142
+ },
+ {
+ "epoch": 64.35463258785943,
+ "grad_norm": 0.0498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0266,
+ "step": 20143
+ },
+ {
+ "epoch": 64.35782747603834,
+ "grad_norm": 0.08544921875,
+ "learning_rate": 0.0005,
+ "loss": 1.014,
+ "step": 20144
+ },
+ {
+ "epoch": 64.36102236421725,
+ "grad_norm": 0.05712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.035,
+ "step": 20145
+ },
+ {
+ "epoch": 64.36421725239616,
+ "grad_norm": 0.05224609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0288,
+ "step": 20146
+ },
+ {
+ "epoch": 64.36741214057508,
+ "grad_norm": 0.052978515625,
+ "learning_rate": 0.0005,
+ "loss": 1.022,
+ "step": 20147
+ },
+ {
+ "epoch": 64.370607028754,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0199,
+ "step": 20148
+ },
+ {
+ "epoch": 64.3738019169329,
+ "grad_norm": 0.062255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0319,
+ "step": 20149
+ },
+ {
+ "epoch": 64.37699680511182,
+ "grad_norm": 0.08154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0262,
+ "step": 20150
+ },
+ {
+ "epoch": 64.38019169329074,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0351,
+ "step": 20151
+ },
+ {
+ "epoch": 64.38338658146965,
+ "grad_norm": 0.05615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0213,
+ "step": 20152
+ },
+ {
+ "epoch": 64.38658146964856,
+ "grad_norm": 0.06103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0227,
+ "step": 20153
+ },
+ {
+ "epoch": 64.38977635782747,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0218,
+ "step": 20154
+ },
+ {
+ "epoch": 64.3929712460064,
+ "grad_norm": 0.052001953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0284,
+ "step": 20155
+ },
+ {
+ "epoch": 64.39616613418531,
+ "grad_norm": 0.095703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0187,
+ "step": 20156
+ },
+ {
+ "epoch": 64.39936102236422,
+ "grad_norm": 0.0576171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0145,
+ "step": 20157
+ },
+ {
+ "epoch": 64.40255591054313,
+ "grad_norm": 0.060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0295,
+ "step": 20158
+ },
+ {
+ "epoch": 64.40575079872204,
+ "grad_norm": 0.055419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0184,
+ "step": 20159
+ },
+ {
+ "epoch": 64.40894568690096,
+ "grad_norm": 0.055419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0319,
+ "step": 20160
+ },
+ {
+ "epoch": 64.41214057507987,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0294,
+ "step": 20161
+ },
+ {
+ "epoch": 64.41533546325878,
+ "grad_norm": 0.08740234375,
+ "learning_rate": 0.0005,
+ "loss": 1.025,
+ "step": 20162
+ },
+ {
+ "epoch": 64.4185303514377,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.021,
+ "step": 20163
+ },
+ {
+ "epoch": 64.42172523961662,
+ "grad_norm": 0.1044921875,
+ "learning_rate": 0.0005,
+ "loss": 1.022,
+ "step": 20164
+ },
+ {
+ "epoch": 64.42492012779553,
+ "grad_norm": 0.06005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0159,
+ "step": 20165
+ },
+ {
+ "epoch": 64.42811501597444,
+ "grad_norm": 0.053955078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0286,
+ "step": 20166
+ },
+ {
+ "epoch": 64.43130990415335,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0252,
+ "step": 20167
+ },
+ {
+ "epoch": 64.43450479233226,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0215,
+ "step": 20168
+ },
+ {
+ "epoch": 64.43769968051119,
+ "grad_norm": 0.059814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0201,
+ "step": 20169
+ },
+ {
+ "epoch": 64.4408945686901,
+ "grad_norm": 0.09423828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0213,
+ "step": 20170
+ },
+ {
+ "epoch": 64.44408945686901,
+ "grad_norm": 0.08349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0106,
+ "step": 20171
+ },
+ {
+ "epoch": 64.44728434504792,
+ "grad_norm": 0.05322265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0062,
+ "step": 20172
+ },
+ {
+ "epoch": 64.45047923322684,
+ "grad_norm": 0.059326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0282,
+ "step": 20173
+ },
+ {
+ "epoch": 64.45367412140575,
+ "grad_norm": 0.107421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0266,
+ "step": 20174
+ },
+ {
+ "epoch": 64.45686900958466,
+ "grad_norm": 0.099609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0138,
+ "step": 20175
+ },
+ {
+ "epoch": 64.46006389776358,
+ "grad_norm": 0.126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0205,
+ "step": 20176
+ },
+ {
+ "epoch": 64.46325878594249,
+ "grad_norm": 0.1162109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0322,
+ "step": 20177
+ },
+ {
+ "epoch": 64.46645367412141,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0101,
+ "step": 20178
+ },
+ {
+ "epoch": 64.46964856230032,
+ "grad_norm": 0.10791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0166,
+ "step": 20179
+ },
+ {
+ "epoch": 64.47284345047923,
+ "grad_norm": 0.09912109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0256,
+ "step": 20180
+ },
+ {
+ "epoch": 64.47603833865814,
+ "grad_norm": 0.1064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0163,
+ "step": 20181
+ },
+ {
+ "epoch": 64.47923322683707,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0214,
+ "step": 20182
+ },
+ {
+ "epoch": 64.48242811501598,
+ "grad_norm": 0.09765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0139,
+ "step": 20183
+ },
+ {
+ "epoch": 64.48562300319489,
+ "grad_norm": 0.0615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0263,
+ "step": 20184
+ },
+ {
+ "epoch": 64.4888178913738,
+ "grad_norm": 0.1240234375,
+ "learning_rate": 0.0005,
+ "loss": 1.022,
+ "step": 20185
+ },
+ {
+ "epoch": 64.49201277955271,
+ "grad_norm": 0.12890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0214,
+ "step": 20186
+ },
+ {
+ "epoch": 64.49520766773163,
+ "grad_norm": 0.09912109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0232,
+ "step": 20187
+ },
+ {
+ "epoch": 64.49840255591054,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0338,
+ "step": 20188
+ },
+ {
+ "epoch": 64.50159744408946,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0263,
+ "step": 20189
+ },
+ {
+ "epoch": 64.50479233226837,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0304,
+ "step": 20190
+ },
+ {
+ "epoch": 64.50798722044729,
+ "grad_norm": 0.10498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0223,
+ "step": 20191
+ },
+ {
+ "epoch": 64.5111821086262,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0206,
+ "step": 20192
+ },
+ {
+ "epoch": 64.51437699680511,
+ "grad_norm": 0.055908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0262,
+ "step": 20193
+ },
+ {
+ "epoch": 64.51757188498402,
+ "grad_norm": 0.095703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0288,
+ "step": 20194
+ },
+ {
+ "epoch": 64.52076677316293,
+ "grad_norm": 0.0810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.024,
+ "step": 20195
+ },
+ {
+ "epoch": 64.52396166134186,
+ "grad_norm": 0.059326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.03,
+ "step": 20196
+ },
+ {
+ "epoch": 64.52715654952077,
+ "grad_norm": 0.095703125,
+ "learning_rate": 0.0005,
+ "loss": 1.025,
+ "step": 20197
+ },
+ {
+ "epoch": 64.53035143769968,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0192,
+ "step": 20198
+ },
+ {
+ "epoch": 64.53354632587859,
+ "grad_norm": 0.0615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0326,
+ "step": 20199
+ },
+ {
+ "epoch": 64.53674121405751,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0284,
+ "step": 20200
+ },
+ {
+ "epoch": 64.53993610223642,
+ "grad_norm": 0.0947265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0228,
+ "step": 20201
+ },
+ {
+ "epoch": 64.54313099041534,
+ "grad_norm": 0.061279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0298,
+ "step": 20202
+ },
+ {
+ "epoch": 64.54632587859425,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0176,
+ "step": 20203
+ },
+ {
+ "epoch": 64.54952076677316,
+ "grad_norm": 0.0537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0286,
+ "step": 20204
+ },
+ {
+ "epoch": 64.55271565495208,
+ "grad_norm": 0.06201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0219,
+ "step": 20205
+ },
+ {
+ "epoch": 64.55591054313099,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0242,
+ "step": 20206
+ },
+ {
+ "epoch": 64.5591054313099,
+ "grad_norm": 0.0576171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0262,
+ "step": 20207
+ },
+ {
+ "epoch": 64.56230031948881,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0295,
+ "step": 20208
+ },
+ {
+ "epoch": 64.56549520766774,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0213,
+ "step": 20209
+ },
+ {
+ "epoch": 64.56869009584665,
+ "grad_norm": 0.057861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0249,
+ "step": 20210
+ },
+ {
+ "epoch": 64.57188498402556,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0295,
+ "step": 20211
+ },
+ {
+ "epoch": 64.57507987220447,
+ "grad_norm": 0.08642578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0218,
+ "step": 20212
+ },
+ {
+ "epoch": 64.57827476038338,
+ "grad_norm": 0.06103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0304,
+ "step": 20213
+ },
+ {
+ "epoch": 64.5814696485623,
+ "grad_norm": 0.055908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0186,
+ "step": 20214
+ },
+ {
+ "epoch": 64.58466453674122,
+ "grad_norm": 0.056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0212,
+ "step": 20215
+ },
+ {
+ "epoch": 64.58785942492013,
+ "grad_norm": 0.045166015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0213,
+ "step": 20216
+ },
+ {
+ "epoch": 64.59105431309904,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0346,
+ "step": 20217
+ },
+ {
+ "epoch": 64.59424920127796,
+ "grad_norm": 0.050048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.019,
+ "step": 20218
+ },
+ {
+ "epoch": 64.59744408945687,
+ "grad_norm": 0.056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0147,
+ "step": 20219
+ },
+ {
+ "epoch": 64.60063897763578,
+ "grad_norm": 0.061767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0268,
+ "step": 20220
+ },
+ {
+ "epoch": 64.60383386581469,
+ "grad_norm": 0.047119140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0324,
+ "step": 20221
+ },
+ {
+ "epoch": 64.6070287539936,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0211,
+ "step": 20222
+ },
+ {
+ "epoch": 64.61022364217253,
+ "grad_norm": 0.0615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0273,
+ "step": 20223
+ },
+ {
+ "epoch": 64.61341853035144,
+ "grad_norm": 0.060791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0158,
+ "step": 20224
+ },
+ {
+ "epoch": 64.61661341853035,
+ "grad_norm": 0.05517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0119,
+ "step": 20225
+ },
+ {
+ "epoch": 64.61980830670926,
+ "grad_norm": 0.051025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0232,
+ "step": 20226
+ },
+ {
+ "epoch": 64.62300319488818,
+ "grad_norm": 0.080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0111,
+ "step": 20227
+ },
+ {
+ "epoch": 64.6261980830671,
+ "grad_norm": 0.0498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0287,
+ "step": 20228
+ },
+ {
+ "epoch": 64.629392971246,
+ "grad_norm": 0.04638671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0224,
+ "step": 20229
+ },
+ {
+ "epoch": 64.63258785942492,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0163,
+ "step": 20230
+ },
+ {
+ "epoch": 64.63578274760384,
+ "grad_norm": 0.050537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0332,
+ "step": 20231
+ },
+ {
+ "epoch": 64.63897763578275,
+ "grad_norm": 0.0458984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0168,
+ "step": 20232
+ },
+ {
+ "epoch": 64.64217252396166,
+ "grad_norm": 0.05126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0259,
+ "step": 20233
+ },
+ {
+ "epoch": 64.64536741214057,
+ "grad_norm": 0.046630859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0207,
+ "step": 20234
+ },
+ {
+ "epoch": 64.64856230031948,
+ "grad_norm": 0.0556640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0214,
+ "step": 20235
+ },
+ {
+ "epoch": 64.65175718849841,
+ "grad_norm": 0.045166015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0228,
+ "step": 20236
+ },
+ {
+ "epoch": 64.65495207667732,
+ "grad_norm": 0.04541015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0254,
+ "step": 20237
+ },
+ {
+ "epoch": 64.65814696485623,
+ "grad_norm": 0.050048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0171,
+ "step": 20238
+ },
+ {
+ "epoch": 64.66134185303514,
+ "grad_norm": 0.062255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0244,
+ "step": 20239
+ },
+ {
+ "epoch": 64.66453674121406,
+ "grad_norm": 0.047607421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0204,
+ "step": 20240
+ },
+ {
+ "epoch": 64.66773162939297,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0231,
+ "step": 20241
+ },
+ {
+ "epoch": 64.67092651757189,
+ "grad_norm": 0.054931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0364,
+ "step": 20242
+ },
+ {
+ "epoch": 64.6741214057508,
+ "grad_norm": 0.0546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0288,
+ "step": 20243
+ },
+ {
+ "epoch": 64.6773162939297,
+ "grad_norm": 0.046630859375,
+ "learning_rate": 0.0005,
+ "loss": 1.015,
+ "step": 20244
+ },
+ {
+ "epoch": 64.68051118210863,
+ "grad_norm": 0.0595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0286,
+ "step": 20245
+ },
+ {
+ "epoch": 64.68370607028754,
+ "grad_norm": 0.056396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0345,
+ "step": 20246
+ },
+ {
+ "epoch": 64.68690095846645,
+ "grad_norm": 0.059814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0235,
+ "step": 20247
+ },
+ {
+ "epoch": 64.69009584664536,
+ "grad_norm": 0.051513671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0208,
+ "step": 20248
+ },
+ {
+ "epoch": 64.69329073482429,
+ "grad_norm": 0.05322265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0222,
+ "step": 20249
+ },
+ {
+ "epoch": 64.6964856230032,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0266,
+ "step": 20250
+ },
+ {
+ "epoch": 64.69968051118211,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0237,
+ "step": 20251
+ },
+ {
+ "epoch": 64.70287539936102,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0264,
+ "step": 20252
+ },
+ {
+ "epoch": 64.70607028753993,
+ "grad_norm": 0.0478515625,
+ "learning_rate": 0.0005,
+ "loss": 1.032,
+ "step": 20253
+ },
+ {
+ "epoch": 64.70926517571885,
+ "grad_norm": 0.05224609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0303,
+ "step": 20254
+ },
+ {
+ "epoch": 64.71246006389777,
+ "grad_norm": 0.056396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0377,
+ "step": 20255
+ },
+ {
+ "epoch": 64.71565495207668,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0314,
+ "step": 20256
+ },
+ {
+ "epoch": 64.71884984025559,
+ "grad_norm": 0.055419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0098,
+ "step": 20257
+ },
+ {
+ "epoch": 64.72204472843451,
+ "grad_norm": 0.08056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0234,
+ "step": 20258
+ },
+ {
+ "epoch": 64.72523961661342,
+ "grad_norm": 0.05224609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0287,
+ "step": 20259
+ },
+ {
+ "epoch": 64.72843450479233,
+ "grad_norm": 0.046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0223,
+ "step": 20260
+ },
+ {
+ "epoch": 64.73162939297124,
+ "grad_norm": 0.06005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0242,
+ "step": 20261
+ },
+ {
+ "epoch": 64.73482428115015,
+ "grad_norm": 0.051025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0219,
+ "step": 20262
+ },
+ {
+ "epoch": 64.73801916932908,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0264,
+ "step": 20263
+ },
+ {
+ "epoch": 64.74121405750799,
+ "grad_norm": 0.08056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0254,
+ "step": 20264
+ },
+ {
+ "epoch": 64.7444089456869,
+ "grad_norm": 0.09521484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0088,
+ "step": 20265
+ },
+ {
+ "epoch": 64.74760383386581,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0126,
+ "step": 20266
+ },
+ {
+ "epoch": 64.75079872204473,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0201,
+ "step": 20267
+ },
+ {
+ "epoch": 64.75399361022365,
+ "grad_norm": 0.052490234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0316,
+ "step": 20268
+ },
+ {
+ "epoch": 64.75718849840256,
+ "grad_norm": 0.05224609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0216,
+ "step": 20269
+ },
+ {
+ "epoch": 64.76038338658147,
+ "grad_norm": 0.05224609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0283,
+ "step": 20270
+ },
+ {
+ "epoch": 64.76357827476038,
+ "grad_norm": 0.0478515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0293,
+ "step": 20271
+ },
+ {
+ "epoch": 64.7667731629393,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0271,
+ "step": 20272
+ },
+ {
+ "epoch": 64.76996805111821,
+ "grad_norm": 0.057861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0236,
+ "step": 20273
+ },
+ {
+ "epoch": 64.77316293929712,
+ "grad_norm": 0.04833984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0201,
+ "step": 20274
+ },
+ {
+ "epoch": 64.77635782747603,
+ "grad_norm": 0.0458984375,
+ "learning_rate": 0.0005,
+ "loss": 1.03,
+ "step": 20275
+ },
+ {
+ "epoch": 64.77955271565496,
+ "grad_norm": 0.044921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0211,
+ "step": 20276
+ },
+ {
+ "epoch": 64.78274760383387,
+ "grad_norm": 0.053955078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0219,
+ "step": 20277
+ },
+ {
+ "epoch": 64.78594249201278,
+ "grad_norm": 0.047607421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0232,
+ "step": 20278
+ },
+ {
+ "epoch": 64.78913738019169,
+ "grad_norm": 0.0458984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0212,
+ "step": 20279
+ },
+ {
+ "epoch": 64.7923322683706,
+ "grad_norm": 0.046630859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0227,
+ "step": 20280
+ },
+ {
+ "epoch": 64.79552715654953,
+ "grad_norm": 0.0615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0202,
+ "step": 20281
+ },
+ {
+ "epoch": 64.79872204472844,
+ "grad_norm": 0.060302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0221,
+ "step": 20282
+ },
+ {
+ "epoch": 64.80191693290735,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0181,
+ "step": 20283
+ },
+ {
+ "epoch": 64.80511182108626,
+ "grad_norm": 0.0595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.025,
+ "step": 20284
+ },
+ {
+ "epoch": 64.80830670926518,
+ "grad_norm": 0.08544921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0239,
+ "step": 20285
+ },
+ {
+ "epoch": 64.81150159744409,
+ "grad_norm": 0.095703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0171,
+ "step": 20286
+ },
+ {
+ "epoch": 64.814696485623,
+ "grad_norm": 0.04638671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0318,
+ "step": 20287
+ },
+ {
+ "epoch": 64.81789137380191,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0441,
+ "step": 20288
+ },
+ {
+ "epoch": 64.82108626198082,
+ "grad_norm": 0.083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0173,
+ "step": 20289
+ },
+ {
+ "epoch": 64.82428115015975,
+ "grad_norm": 0.0537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.02,
+ "step": 20290
+ },
+ {
+ "epoch": 64.82747603833866,
+ "grad_norm": 0.05126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0326,
+ "step": 20291
+ },
+ {
+ "epoch": 64.83067092651757,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0189,
+ "step": 20292
+ },
+ {
+ "epoch": 64.83386581469648,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0169,
+ "step": 20293
+ },
+ {
+ "epoch": 64.8370607028754,
+ "grad_norm": 0.059326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.028,
+ "step": 20294
+ },
+ {
+ "epoch": 64.84025559105432,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0315,
+ "step": 20295
+ },
+ {
+ "epoch": 64.84345047923323,
+ "grad_norm": 0.049072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0193,
+ "step": 20296
+ },
+ {
+ "epoch": 64.84664536741214,
+ "grad_norm": 0.050048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0213,
+ "step": 20297
+ },
+ {
+ "epoch": 64.84984025559105,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0259,
+ "step": 20298
+ },
+ {
+ "epoch": 64.85303514376997,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0199,
+ "step": 20299
+ },
+ {
+ "epoch": 64.85623003194888,
+ "grad_norm": 0.0517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0309,
+ "step": 20300
+ },
+ {
+ "epoch": 64.8594249201278,
+ "grad_norm": 0.057861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0223,
+ "step": 20301
+ },
+ {
+ "epoch": 64.8626198083067,
+ "grad_norm": 0.08349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0301,
+ "step": 20302
+ },
+ {
+ "epoch": 64.86581469648563,
+ "grad_norm": 0.049072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0246,
+ "step": 20303
+ },
+ {
+ "epoch": 64.86900958466454,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0212,
+ "step": 20304
+ },
+ {
+ "epoch": 64.87220447284345,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0167,
+ "step": 20305
+ },
+ {
+ "epoch": 64.87539936102236,
+ "grad_norm": 0.060302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0294,
+ "step": 20306
+ },
+ {
+ "epoch": 64.87859424920129,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0235,
+ "step": 20307
+ },
+ {
+ "epoch": 64.8817891373802,
+ "grad_norm": 0.050537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0259,
+ "step": 20308
+ },
+ {
+ "epoch": 64.8849840255591,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0266,
+ "step": 20309
+ },
+ {
+ "epoch": 64.88817891373802,
+ "grad_norm": 0.05810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0178,
+ "step": 20310
+ },
+ {
+ "epoch": 64.89137380191693,
+ "grad_norm": 0.07568359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0248,
+ "step": 20311
+ },
+ {
+ "epoch": 64.89456869009585,
+ "grad_norm": 0.040771484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0158,
+ "step": 20312
+ },
+ {
+ "epoch": 64.89776357827476,
+ "grad_norm": 0.08056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0228,
+ "step": 20313
+ },
+ {
+ "epoch": 64.90095846645367,
+ "grad_norm": 0.04541015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0187,
+ "step": 20314
+ },
+ {
+ "epoch": 64.90415335463258,
+ "grad_norm": 0.09130859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0179,
+ "step": 20315
+ },
+ {
+ "epoch": 64.90734824281151,
+ "grad_norm": 0.049560546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0172,
+ "step": 20316
+ },
+ {
+ "epoch": 64.91054313099042,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0166,
+ "step": 20317
+ },
+ {
+ "epoch": 64.91373801916933,
+ "grad_norm": 0.059326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0234,
+ "step": 20318
+ },
+ {
+ "epoch": 64.91693290734824,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0321,
+ "step": 20319
+ },
+ {
+ "epoch": 64.92012779552715,
+ "grad_norm": 0.048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0272,
+ "step": 20320
+ },
+ {
+ "epoch": 64.92332268370608,
+ "grad_norm": 0.061279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0257,
+ "step": 20321
+ },
+ {
+ "epoch": 64.92651757188499,
+ "grad_norm": 0.053466796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0188,
+ "step": 20322
+ },
+ {
+ "epoch": 64.9297124600639,
+ "grad_norm": 0.053955078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0248,
+ "step": 20323
+ },
+ {
+ "epoch": 64.93290734824281,
+ "grad_norm": 0.1005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0178,
+ "step": 20324
+ },
+ {
+ "epoch": 64.93610223642173,
+ "grad_norm": 0.1025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0318,
+ "step": 20325
+ },
+ {
+ "epoch": 64.93929712460064,
+ "grad_norm": 0.123046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0134,
+ "step": 20326
+ },
+ {
+ "epoch": 64.94249201277955,
+ "grad_norm": 0.08837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.019,
+ "step": 20327
+ },
+ {
+ "epoch": 64.94568690095846,
+ "grad_norm": 0.11572265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0322,
+ "step": 20328
+ },
+ {
+ "epoch": 64.94888178913737,
+ "grad_norm": 0.05126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.028,
+ "step": 20329
+ },
+ {
+ "epoch": 64.9520766773163,
+ "grad_norm": 0.103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0271,
+ "step": 20330
+ },
+ {
+ "epoch": 64.95527156549521,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.024,
+ "step": 20331
+ },
+ {
+ "epoch": 64.95846645367412,
+ "grad_norm": 0.057861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0174,
+ "step": 20332
+ },
+ {
+ "epoch": 64.96166134185303,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0291,
+ "step": 20333
+ },
+ {
+ "epoch": 64.96485623003196,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0233,
+ "step": 20334
+ },
+ {
+ "epoch": 64.96805111821087,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0244,
+ "step": 20335
+ },
+ {
+ "epoch": 64.97124600638978,
+ "grad_norm": 0.10302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0226,
+ "step": 20336
+ },
+ {
+ "epoch": 64.97444089456869,
+ "grad_norm": 0.053466796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0281,
+ "step": 20337
+ },
+ {
+ "epoch": 64.9776357827476,
+ "grad_norm": 0.10205078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0281,
+ "step": 20338
+ },
+ {
+ "epoch": 64.98083067092652,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.032,
+ "step": 20339
+ },
+ {
+ "epoch": 64.98402555910543,
+ "grad_norm": 0.05322265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0173,
+ "step": 20340
+ },
+ {
+ "epoch": 64.98722044728434,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0249,
+ "step": 20341
+ },
+ {
+ "epoch": 64.99041533546325,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0183,
+ "step": 20342
+ },
+ {
+ "epoch": 64.99361022364218,
+ "grad_norm": 0.052978515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0231,
+ "step": 20343
+ },
+ {
+ "epoch": 64.99680511182109,
+ "grad_norm": 0.0810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0303,
+ "step": 20344
+ },
+ {
+ "epoch": 65.0,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0353,
+ "step": 20345
+ },
+ {
+ "epoch": 65.00319488817891,
+ "grad_norm": 0.08740234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0229,
+ "step": 20346
+ },
+ {
+ "epoch": 65.00638977635782,
+ "grad_norm": 0.052001953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0157,
+ "step": 20347
+ },
+ {
+ "epoch": 65.00958466453675,
+ "grad_norm": 0.056884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0191,
+ "step": 20348
+ },
+ {
+ "epoch": 65.01277955271566,
+ "grad_norm": 0.055419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0265,
+ "step": 20349
+ },
+ {
+ "epoch": 65.01597444089457,
+ "grad_norm": 0.061279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0209,
+ "step": 20350
+ },
+ {
+ "epoch": 65.01916932907348,
+ "grad_norm": 0.0498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.014,
+ "step": 20351
+ },
+ {
+ "epoch": 65.0223642172524,
+ "grad_norm": 0.061279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0203,
+ "step": 20352
+ },
+ {
+ "epoch": 65.02555910543131,
+ "grad_norm": 0.04736328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0234,
+ "step": 20353
+ },
+ {
+ "epoch": 65.02875399361022,
+ "grad_norm": 0.0556640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0184,
+ "step": 20354
+ },
+ {
+ "epoch": 65.03194888178913,
+ "grad_norm": 0.050048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0161,
+ "step": 20355
+ },
+ {
+ "epoch": 65.03514376996804,
+ "grad_norm": 0.045654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0212,
+ "step": 20356
+ },
+ {
+ "epoch": 65.03833865814697,
+ "grad_norm": 0.049072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0161,
+ "step": 20357
+ },
+ {
+ "epoch": 65.04153354632588,
+ "grad_norm": 0.060302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0275,
+ "step": 20358
+ },
+ {
+ "epoch": 65.04472843450479,
+ "grad_norm": 0.0498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0316,
+ "step": 20359
+ },
+ {
+ "epoch": 65.0479233226837,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0244,
+ "step": 20360
+ },
+ {
+ "epoch": 65.05111821086263,
+ "grad_norm": 0.053955078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0245,
+ "step": 20361
+ },
+ {
+ "epoch": 65.05431309904154,
+ "grad_norm": 0.052734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0188,
+ "step": 20362
+ },
+ {
+ "epoch": 65.05750798722045,
+ "grad_norm": 0.059814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0241,
+ "step": 20363
+ },
+ {
+ "epoch": 65.06070287539936,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0214,
+ "step": 20364
+ },
+ {
+ "epoch": 65.06389776357827,
+ "grad_norm": 0.053955078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0213,
+ "step": 20365
+ },
+ {
+ "epoch": 65.06709265175719,
+ "grad_norm": 0.050048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0192,
+ "step": 20366
+ },
+ {
+ "epoch": 65.0702875399361,
+ "grad_norm": 0.056396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0227,
+ "step": 20367
+ },
+ {
+ "epoch": 65.07348242811501,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0174,
+ "step": 20368
+ },
+ {
+ "epoch": 65.07667731629392,
+ "grad_norm": 0.058837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.033,
+ "step": 20369
+ },
+ {
+ "epoch": 65.07987220447285,
+ "grad_norm": 0.060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0253,
+ "step": 20370
+ },
+ {
+ "epoch": 65.08306709265176,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0232,
+ "step": 20371
+ },
+ {
+ "epoch": 65.08626198083067,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0189,
+ "step": 20372
+ },
+ {
+ "epoch": 65.08945686900958,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.03,
+ "step": 20373
+ },
+ {
+ "epoch": 65.09265175718849,
+ "grad_norm": 0.08544921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0236,
+ "step": 20374
+ },
+ {
+ "epoch": 65.09584664536742,
+ "grad_norm": 0.05810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0165,
+ "step": 20375
+ },
+ {
+ "epoch": 65.09904153354633,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0208,
+ "step": 20376
+ },
+ {
+ "epoch": 65.10223642172524,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0278,
+ "step": 20377
+ },
+ {
+ "epoch": 65.10543130990415,
+ "grad_norm": 0.05126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0284,
+ "step": 20378
+ },
+ {
+ "epoch": 65.10862619808307,
+ "grad_norm": 0.052734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0214,
+ "step": 20379
+ },
+ {
+ "epoch": 65.11182108626198,
+ "grad_norm": 0.05712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0243,
+ "step": 20380
+ },
+ {
+ "epoch": 65.1150159744409,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0257,
+ "step": 20381
+ },
+ {
+ "epoch": 65.1182108626198,
+ "grad_norm": 0.056884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0178,
+ "step": 20382
+ },
+ {
+ "epoch": 65.12140575079871,
+ "grad_norm": 0.053955078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0181,
+ "step": 20383
+ },
+ {
+ "epoch": 65.12460063897764,
+ "grad_norm": 0.051025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.026,
+ "step": 20384
+ },
+ {
+ "epoch": 65.12779552715655,
+ "grad_norm": 0.0498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0274,
+ "step": 20385
+ },
+ {
+ "epoch": 65.13099041533546,
+ "grad_norm": 0.050537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0145,
+ "step": 20386
+ },
+ {
+ "epoch": 65.13418530351437,
+ "grad_norm": 0.047119140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0318,
+ "step": 20387
+ },
+ {
+ "epoch": 65.1373801916933,
+ "grad_norm": 0.0576171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0237,
+ "step": 20388
+ },
+ {
+ "epoch": 65.1405750798722,
+ "grad_norm": 0.059814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0109,
+ "step": 20389
+ },
+ {
+ "epoch": 65.14376996805112,
+ "grad_norm": 0.060791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0266,
+ "step": 20390
+ },
+ {
+ "epoch": 65.14696485623003,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0286,
+ "step": 20391
+ },
+ {
+ "epoch": 65.15015974440895,
+ "grad_norm": 0.054931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0263,
+ "step": 20392
+ },
+ {
+ "epoch": 65.15335463258786,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0088,
+ "step": 20393
+ },
+ {
+ "epoch": 65.15654952076677,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0204,
+ "step": 20394
+ },
+ {
+ "epoch": 65.15974440894568,
+ "grad_norm": 0.061767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.023,
+ "step": 20395
+ },
+ {
+ "epoch": 65.1629392971246,
+ "grad_norm": 0.048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0255,
+ "step": 20396
+ },
+ {
+ "epoch": 65.16613418530352,
+ "grad_norm": 0.061279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0265,
+ "step": 20397
+ },
+ {
+ "epoch": 65.16932907348243,
+ "grad_norm": 0.051025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0285,
+ "step": 20398
+ },
+ {
+ "epoch": 65.17252396166134,
+ "grad_norm": 0.053955078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0111,
+ "step": 20399
+ },
+ {
+ "epoch": 65.17571884984025,
+ "grad_norm": 0.049560546875,
+ "learning_rate": 0.0005,
+ "loss": 1.016,
+ "step": 20400
+ },
+ {
+ "epoch": 65.17891373801918,
+ "grad_norm": 0.04638671875,
+ "learning_rate": 0.0005,
+ "loss": 1.011,
+ "step": 20401
+ },
+ {
+ "epoch": 65.18210862619809,
+ "grad_norm": 0.060791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0225,
+ "step": 20402
+ },
+ {
+ "epoch": 65.185303514377,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0296,
+ "step": 20403
+ },
+ {
+ "epoch": 65.18849840255591,
+ "grad_norm": 0.0458984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0273,
+ "step": 20404
+ },
+ {
+ "epoch": 65.19169329073482,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0208,
+ "step": 20405
+ },
+ {
+ "epoch": 65.19488817891374,
+ "grad_norm": 0.051513671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0218,
+ "step": 20406
+ },
+ {
+ "epoch": 65.19808306709265,
+ "grad_norm": 0.05859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0333,
+ "step": 20407
+ },
+ {
+ "epoch": 65.20127795527156,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0207,
+ "step": 20408
+ },
+ {
+ "epoch": 65.20447284345047,
+ "grad_norm": 0.08251953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0218,
+ "step": 20409
+ },
+ {
+ "epoch": 65.2076677316294,
+ "grad_norm": 0.0849609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0204,
+ "step": 20410
+ },
+ {
+ "epoch": 65.21086261980831,
+ "grad_norm": 0.15625,
+ "learning_rate": 0.0005,
+ "loss": 1.011,
+ "step": 20411
+ },
+ {
+ "epoch": 65.21405750798722,
+ "grad_norm": 0.1123046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0293,
+ "step": 20412
+ },
+ {
+ "epoch": 65.21725239616613,
+ "grad_norm": 0.08837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0276,
+ "step": 20413
+ },
+ {
+ "epoch": 65.22044728434504,
+ "grad_norm": 0.09228515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0229,
+ "step": 20414
+ },
+ {
+ "epoch": 65.22364217252397,
+ "grad_norm": 0.08203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0272,
+ "step": 20415
+ },
+ {
+ "epoch": 65.22683706070288,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0123,
+ "step": 20416
+ },
+ {
+ "epoch": 65.23003194888179,
+ "grad_norm": 0.11572265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0144,
+ "step": 20417
+ },
+ {
+ "epoch": 65.2332268370607,
+ "grad_norm": 0.06103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0285,
+ "step": 20418
+ },
+ {
+ "epoch": 65.23642172523962,
+ "grad_norm": 0.10302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0299,
+ "step": 20419
+ },
+ {
+ "epoch": 65.23961661341853,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0294,
+ "step": 20420
+ },
+ {
+ "epoch": 65.24281150159744,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0272,
+ "step": 20421
+ },
+ {
+ "epoch": 65.24600638977635,
+ "grad_norm": 0.05908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0229,
+ "step": 20422
+ },
+ {
+ "epoch": 65.24920127795527,
+ "grad_norm": 0.05712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0179,
+ "step": 20423
+ },
+ {
+ "epoch": 65.25239616613419,
+ "grad_norm": 0.050048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0282,
+ "step": 20424
+ },
+ {
+ "epoch": 65.2555910543131,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0247,
+ "step": 20425
+ },
+ {
+ "epoch": 65.25878594249201,
+ "grad_norm": 0.052001953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0366,
+ "step": 20426
+ },
+ {
+ "epoch": 65.26198083067092,
+ "grad_norm": 0.08984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0171,
+ "step": 20427
+ },
+ {
+ "epoch": 65.26517571884985,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0213,
+ "step": 20428
+ },
+ {
+ "epoch": 65.26837060702876,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0096,
+ "step": 20429
+ },
+ {
+ "epoch": 65.27156549520767,
+ "grad_norm": 0.0576171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0249,
+ "step": 20430
+ },
+ {
+ "epoch": 65.27476038338658,
+ "grad_norm": 0.083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0267,
+ "step": 20431
+ },
+ {
+ "epoch": 65.27795527156549,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0283,
+ "step": 20432
+ },
+ {
+ "epoch": 65.28115015974441,
+ "grad_norm": 0.053466796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0253,
+ "step": 20433
+ },
+ {
+ "epoch": 65.28434504792332,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0288,
+ "step": 20434
+ },
+ {
+ "epoch": 65.28753993610223,
+ "grad_norm": 0.049560546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0151,
+ "step": 20435
+ },
+ {
+ "epoch": 65.29073482428115,
+ "grad_norm": 0.05517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0186,
+ "step": 20436
+ },
+ {
+ "epoch": 65.29392971246007,
+ "grad_norm": 0.0546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0245,
+ "step": 20437
+ },
+ {
+ "epoch": 65.29712460063898,
+ "grad_norm": 0.0439453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0178,
+ "step": 20438
+ },
+ {
+ "epoch": 65.30031948881789,
+ "grad_norm": 0.046630859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0299,
+ "step": 20439
+ },
+ {
+ "epoch": 65.3035143769968,
+ "grad_norm": 0.058837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0266,
+ "step": 20440
+ },
+ {
+ "epoch": 65.30670926517571,
+ "grad_norm": 0.047119140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0225,
+ "step": 20441
+ },
+ {
+ "epoch": 65.30990415335464,
+ "grad_norm": 0.0498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0341,
+ "step": 20442
+ },
+ {
+ "epoch": 65.31309904153355,
+ "grad_norm": 0.047607421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0224,
+ "step": 20443
+ },
+ {
+ "epoch": 65.31629392971246,
+ "grad_norm": 0.049560546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0191,
+ "step": 20444
+ },
+ {
+ "epoch": 65.31948881789137,
+ "grad_norm": 0.05712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.015,
+ "step": 20445
+ },
+ {
+ "epoch": 65.3226837060703,
+ "grad_norm": 0.0478515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0249,
+ "step": 20446
+ },
+ {
+ "epoch": 65.3258785942492,
+ "grad_norm": 0.04833984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0143,
+ "step": 20447
+ },
+ {
+ "epoch": 65.32907348242811,
+ "grad_norm": 0.043701171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0232,
+ "step": 20448
+ },
+ {
+ "epoch": 65.33226837060703,
+ "grad_norm": 0.05126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0136,
+ "step": 20449
+ },
+ {
+ "epoch": 65.33546325878594,
+ "grad_norm": 0.046630859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0077,
+ "step": 20450
+ },
+ {
+ "epoch": 65.33865814696486,
+ "grad_norm": 0.04736328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0275,
+ "step": 20451
+ },
+ {
+ "epoch": 65.34185303514377,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0207,
+ "step": 20452
+ },
+ {
+ "epoch": 65.34504792332268,
+ "grad_norm": 0.052978515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0234,
+ "step": 20453
+ },
+ {
+ "epoch": 65.34824281150159,
+ "grad_norm": 0.050537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.02,
+ "step": 20454
+ },
+ {
+ "epoch": 65.35143769968052,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0259,
+ "step": 20455
+ },
+ {
+ "epoch": 65.35463258785943,
+ "grad_norm": 0.05859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0301,
+ "step": 20456
+ },
+ {
+ "epoch": 65.35782747603834,
+ "grad_norm": 0.08984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0294,
+ "step": 20457
+ },
+ {
+ "epoch": 65.36102236421725,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0227,
+ "step": 20458
+ },
+ {
+ "epoch": 65.36421725239616,
+ "grad_norm": 0.11767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0314,
+ "step": 20459
+ },
+ {
+ "epoch": 65.36741214057508,
+ "grad_norm": 0.08056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.023,
+ "step": 20460
+ },
+ {
+ "epoch": 65.370607028754,
+ "grad_norm": 0.09375,
+ "learning_rate": 0.0005,
+ "loss": 1.0161,
+ "step": 20461
+ },
+ {
+ "epoch": 65.3738019169329,
+ "grad_norm": 0.0966796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0222,
+ "step": 20462
+ },
+ {
+ "epoch": 65.37699680511182,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0221,
+ "step": 20463
+ },
+ {
+ "epoch": 65.38019169329074,
+ "grad_norm": 0.1201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0213,
+ "step": 20464
+ },
+ {
+ "epoch": 65.38338658146965,
+ "grad_norm": 0.05712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0193,
+ "step": 20465
+ },
+ {
+ "epoch": 65.38658146964856,
+ "grad_norm": 0.1357421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0265,
+ "step": 20466
+ },
+ {
+ "epoch": 65.38977635782747,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0225,
+ "step": 20467
+ },
+ {
+ "epoch": 65.3929712460064,
+ "grad_norm": 0.126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0175,
+ "step": 20468
+ },
+ {
+ "epoch": 65.39616613418531,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.021,
+ "step": 20469
+ },
+ {
+ "epoch": 65.39936102236422,
+ "grad_norm": 0.08935546875,
+ "learning_rate": 0.0005,
+ "loss": 1.028,
+ "step": 20470
+ },
+ {
+ "epoch": 65.40255591054313,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0238,
+ "step": 20471
+ },
+ {
+ "epoch": 65.40575079872204,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0164,
+ "step": 20472
+ },
+ {
+ "epoch": 65.40894568690096,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0257,
+ "step": 20473
+ },
+ {
+ "epoch": 65.41214057507987,
+ "grad_norm": 0.054443359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0209,
+ "step": 20474
+ },
+ {
+ "epoch": 65.41533546325878,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0201,
+ "step": 20475
+ },
+ {
+ "epoch": 65.4185303514377,
+ "grad_norm": 0.05517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0242,
+ "step": 20476
+ },
+ {
+ "epoch": 65.42172523961662,
+ "grad_norm": 0.060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0206,
+ "step": 20477
+ },
+ {
+ "epoch": 65.42492012779553,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0335,
+ "step": 20478
+ },
+ {
+ "epoch": 65.42811501597444,
+ "grad_norm": 0.0830078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0282,
+ "step": 20479
+ },
+ {
+ "epoch": 65.43130990415335,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0269,
+ "step": 20480
+ },
+ {
+ "epoch": 65.43450479233226,
+ "grad_norm": 0.0947265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0163,
+ "step": 20481
+ },
+ {
+ "epoch": 65.43769968051119,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.018,
+ "step": 20482
+ },
+ {
+ "epoch": 65.4408945686901,
+ "grad_norm": 0.05322265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0167,
+ "step": 20483
+ },
+ {
+ "epoch": 65.44408945686901,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0288,
+ "step": 20484
+ },
+ {
+ "epoch": 65.44728434504792,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0278,
+ "step": 20485
+ },
+ {
+ "epoch": 65.45047923322684,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0227,
+ "step": 20486
+ },
+ {
+ "epoch": 65.45367412140575,
+ "grad_norm": 0.0546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0229,
+ "step": 20487
+ },
+ {
+ "epoch": 65.45686900958466,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0138,
+ "step": 20488
+ },
+ {
+ "epoch": 65.46006389776358,
+ "grad_norm": 0.050048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0299,
+ "step": 20489
+ },
+ {
+ "epoch": 65.46325878594249,
+ "grad_norm": 0.04931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0261,
+ "step": 20490
+ },
+ {
+ "epoch": 65.46645367412141,
+ "grad_norm": 0.11328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0299,
+ "step": 20491
+ },
+ {
+ "epoch": 65.46964856230032,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0293,
+ "step": 20492
+ },
+ {
+ "epoch": 65.47284345047923,
+ "grad_norm": 0.0966796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0311,
+ "step": 20493
+ },
+ {
+ "epoch": 65.47603833865814,
+ "grad_norm": 0.11865234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0246,
+ "step": 20494
+ },
+ {
+ "epoch": 65.47923322683707,
+ "grad_norm": 0.06201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0355,
+ "step": 20495
+ },
+ {
+ "epoch": 65.48242811501598,
+ "grad_norm": 0.08740234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0179,
+ "step": 20496
+ },
+ {
+ "epoch": 65.48562300319489,
+ "grad_norm": 0.1328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0228,
+ "step": 20497
+ },
+ {
+ "epoch": 65.4888178913738,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0147,
+ "step": 20498
+ },
+ {
+ "epoch": 65.49201277955271,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0226,
+ "step": 20499
+ },
+ {
+ "epoch": 65.49520766773163,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.023,
+ "step": 20500
+ },
+ {
+ "epoch": 65.49840255591054,
+ "grad_norm": 0.054443359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0274,
+ "step": 20501
+ },
+ {
+ "epoch": 65.50159744408946,
+ "grad_norm": 0.0556640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0073,
+ "step": 20502
+ },
+ {
+ "epoch": 65.50479233226837,
+ "grad_norm": 0.05126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0177,
+ "step": 20503
+ },
+ {
+ "epoch": 65.50798722044729,
+ "grad_norm": 0.049072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0199,
+ "step": 20504
+ },
+ {
+ "epoch": 65.5111821086262,
+ "grad_norm": 0.048583984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0148,
+ "step": 20505
+ },
+ {
+ "epoch": 65.51437699680511,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0221,
+ "step": 20506
+ },
+ {
+ "epoch": 65.51757188498402,
+ "grad_norm": 0.054931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0254,
+ "step": 20507
+ },
+ {
+ "epoch": 65.52076677316293,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0269,
+ "step": 20508
+ },
+ {
+ "epoch": 65.52396166134186,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0191,
+ "step": 20509
+ },
+ {
+ "epoch": 65.52715654952077,
+ "grad_norm": 0.048095703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0205,
+ "step": 20510
+ },
+ {
+ "epoch": 65.53035143769968,
+ "grad_norm": 0.058349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0193,
+ "step": 20511
+ },
+ {
+ "epoch": 65.53354632587859,
+ "grad_norm": 0.05029296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0172,
+ "step": 20512
+ },
+ {
+ "epoch": 65.53674121405751,
+ "grad_norm": 0.04931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0262,
+ "step": 20513
+ },
+ {
+ "epoch": 65.53993610223642,
+ "grad_norm": 0.0537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0291,
+ "step": 20514
+ },
+ {
+ "epoch": 65.54313099041534,
+ "grad_norm": 0.04931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0153,
+ "step": 20515
+ },
+ {
+ "epoch": 65.54632587859425,
+ "grad_norm": 0.049072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0144,
+ "step": 20516
+ },
+ {
+ "epoch": 65.54952076677316,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0253,
+ "step": 20517
+ },
+ {
+ "epoch": 65.55271565495208,
+ "grad_norm": 0.05224609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0254,
+ "step": 20518
+ },
+ {
+ "epoch": 65.55591054313099,
+ "grad_norm": 0.05517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0167,
+ "step": 20519
+ },
+ {
+ "epoch": 65.5591054313099,
+ "grad_norm": 0.055908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.012,
+ "step": 20520
+ },
+ {
+ "epoch": 65.56230031948881,
+ "grad_norm": 0.051513671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0253,
+ "step": 20521
+ },
+ {
+ "epoch": 65.56549520766774,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0268,
+ "step": 20522
+ },
+ {
+ "epoch": 65.56869009584665,
+ "grad_norm": 0.054931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0204,
+ "step": 20523
+ },
+ {
+ "epoch": 65.57188498402556,
+ "grad_norm": 0.08984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0201,
+ "step": 20524
+ },
+ {
+ "epoch": 65.57507987220447,
+ "grad_norm": 0.0478515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0214,
+ "step": 20525
+ },
+ {
+ "epoch": 65.57827476038338,
+ "grad_norm": 0.08544921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0174,
+ "step": 20526
+ },
+ {
+ "epoch": 65.5814696485623,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0156,
+ "step": 20527
+ },
+ {
+ "epoch": 65.58466453674122,
+ "grad_norm": 0.0927734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0301,
+ "step": 20528
+ },
+ {
+ "epoch": 65.58785942492013,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0188,
+ "step": 20529
+ },
+ {
+ "epoch": 65.59105431309904,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0273,
+ "step": 20530
+ },
+ {
+ "epoch": 65.59424920127796,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0225,
+ "step": 20531
+ },
+ {
+ "epoch": 65.59744408945687,
+ "grad_norm": 0.062255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0262,
+ "step": 20532
+ },
+ {
+ "epoch": 65.60063897763578,
+ "grad_norm": 0.080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0265,
+ "step": 20533
+ },
+ {
+ "epoch": 65.60383386581469,
+ "grad_norm": 0.04833984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0162,
+ "step": 20534
+ },
+ {
+ "epoch": 65.6070287539936,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0165,
+ "step": 20535
+ },
+ {
+ "epoch": 65.61022364217253,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0206,
+ "step": 20536
+ },
+ {
+ "epoch": 65.61341853035144,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0293,
+ "step": 20537
+ },
+ {
+ "epoch": 65.61661341853035,
+ "grad_norm": 0.080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0343,
+ "step": 20538
+ },
+ {
+ "epoch": 65.61980830670926,
+ "grad_norm": 0.054443359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0092,
+ "step": 20539
+ },
+ {
+ "epoch": 65.62300319488818,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0115,
+ "step": 20540
+ },
+ {
+ "epoch": 65.6261980830671,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0274,
+ "step": 20541
+ },
+ {
+ "epoch": 65.629392971246,
+ "grad_norm": 0.052978515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0214,
+ "step": 20542
+ },
+ {
+ "epoch": 65.63258785942492,
+ "grad_norm": 0.08935546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0303,
+ "step": 20543
+ },
+ {
+ "epoch": 65.63578274760384,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0124,
+ "step": 20544
+ },
+ {
+ "epoch": 65.63897763578275,
+ "grad_norm": 0.10693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0339,
+ "step": 20545
+ },
+ {
+ "epoch": 65.64217252396166,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0185,
+ "step": 20546
+ },
+ {
+ "epoch": 65.64536741214057,
+ "grad_norm": 0.1162109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0336,
+ "step": 20547
+ },
+ {
+ "epoch": 65.64856230031948,
+ "grad_norm": 0.09375,
+ "learning_rate": 0.0005,
+ "loss": 1.0139,
+ "step": 20548
+ },
+ {
+ "epoch": 65.65175718849841,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0145,
+ "step": 20549
+ },
+ {
+ "epoch": 65.65495207667732,
+ "grad_norm": 0.055908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0189,
+ "step": 20550
+ },
+ {
+ "epoch": 65.65814696485623,
+ "grad_norm": 0.07958984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0238,
+ "step": 20551
+ },
+ {
+ "epoch": 65.66134185303514,
+ "grad_norm": 0.0986328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0285,
+ "step": 20552
+ },
+ {
+ "epoch": 65.66453674121406,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0105,
+ "step": 20553
+ },
+ {
+ "epoch": 65.66773162939297,
+ "grad_norm": 0.08251953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0258,
+ "step": 20554
+ },
+ {
+ "epoch": 65.67092651757189,
+ "grad_norm": 0.12353515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0253,
+ "step": 20555
+ },
+ {
+ "epoch": 65.6741214057508,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0148,
+ "step": 20556
+ },
+ {
+ "epoch": 65.6773162939297,
+ "grad_norm": 0.11181640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0183,
+ "step": 20557
+ },
+ {
+ "epoch": 65.68051118210863,
+ "grad_norm": 0.09716796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0275,
+ "step": 20558
+ },
+ {
+ "epoch": 65.68370607028754,
+ "grad_norm": 0.050537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0333,
+ "step": 20559
+ },
+ {
+ "epoch": 65.68690095846645,
+ "grad_norm": 0.1240234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0255,
+ "step": 20560
+ },
+ {
+ "epoch": 65.69009584664536,
+ "grad_norm": 0.095703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0205,
+ "step": 20561
+ },
+ {
+ "epoch": 65.69329073482429,
+ "grad_norm": 0.1318359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0281,
+ "step": 20562
+ },
+ {
+ "epoch": 65.6964856230032,
+ "grad_norm": 0.09033203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0314,
+ "step": 20563
+ },
+ {
+ "epoch": 65.69968051118211,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0327,
+ "step": 20564
+ },
+ {
+ "epoch": 65.70287539936102,
+ "grad_norm": 0.060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0224,
+ "step": 20565
+ },
+ {
+ "epoch": 65.70607028753993,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.022,
+ "step": 20566
+ },
+ {
+ "epoch": 65.70926517571885,
+ "grad_norm": 0.057861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0385,
+ "step": 20567
+ },
+ {
+ "epoch": 65.71246006389777,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0226,
+ "step": 20568
+ },
+ {
+ "epoch": 65.71565495207668,
+ "grad_norm": 0.0537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0248,
+ "step": 20569
+ },
+ {
+ "epoch": 65.71884984025559,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0202,
+ "step": 20570
+ },
+ {
+ "epoch": 65.72204472843451,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0214,
+ "step": 20571
+ },
+ {
+ "epoch": 65.72523961661342,
+ "grad_norm": 0.0595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0268,
+ "step": 20572
+ },
+ {
+ "epoch": 65.72843450479233,
+ "grad_norm": 0.060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.026,
+ "step": 20573
+ },
+ {
+ "epoch": 65.73162939297124,
+ "grad_norm": 0.05810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0181,
+ "step": 20574
+ },
+ {
+ "epoch": 65.73482428115015,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.028,
+ "step": 20575
+ },
+ {
+ "epoch": 65.73801916932908,
+ "grad_norm": 0.1015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0273,
+ "step": 20576
+ },
+ {
+ "epoch": 65.74121405750799,
+ "grad_norm": 0.049560546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0271,
+ "step": 20577
+ },
+ {
+ "epoch": 65.7444089456869,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0397,
+ "step": 20578
+ },
+ {
+ "epoch": 65.74760383386581,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0265,
+ "step": 20579
+ },
+ {
+ "epoch": 65.75079872204473,
+ "grad_norm": 0.060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0281,
+ "step": 20580
+ },
+ {
+ "epoch": 65.75399361022365,
+ "grad_norm": 0.080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0287,
+ "step": 20581
+ },
+ {
+ "epoch": 65.75718849840256,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0282,
+ "step": 20582
+ },
+ {
+ "epoch": 65.76038338658147,
+ "grad_norm": 0.053955078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0294,
+ "step": 20583
+ },
+ {
+ "epoch": 65.76357827476038,
+ "grad_norm": 0.099609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0203,
+ "step": 20584
+ },
+ {
+ "epoch": 65.7667731629393,
+ "grad_norm": 0.047607421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0204,
+ "step": 20585
+ },
+ {
+ "epoch": 65.76996805111821,
+ "grad_norm": 0.08984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0268,
+ "step": 20586
+ },
+ {
+ "epoch": 65.77316293929712,
+ "grad_norm": 0.0888671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0144,
+ "step": 20587
+ },
+ {
+ "epoch": 65.77635782747603,
+ "grad_norm": 0.05615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0312,
+ "step": 20588
+ },
+ {
+ "epoch": 65.77955271565496,
+ "grad_norm": 0.1123046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0176,
+ "step": 20589
+ },
+ {
+ "epoch": 65.78274760383387,
+ "grad_norm": 0.08447265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0358,
+ "step": 20590
+ },
+ {
+ "epoch": 65.78594249201278,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0205,
+ "step": 20591
+ },
+ {
+ "epoch": 65.78913738019169,
+ "grad_norm": 0.119140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0209,
+ "step": 20592
+ },
+ {
+ "epoch": 65.7923322683706,
+ "grad_norm": 0.1142578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0185,
+ "step": 20593
+ },
+ {
+ "epoch": 65.79552715654953,
+ "grad_norm": 0.05078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0098,
+ "step": 20594
+ },
+ {
+ "epoch": 65.79872204472844,
+ "grad_norm": 0.15625,
+ "learning_rate": 0.0005,
+ "loss": 1.0224,
+ "step": 20595
+ },
+ {
+ "epoch": 65.80191693290735,
+ "grad_norm": 0.1494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.024,
+ "step": 20596
+ },
+ {
+ "epoch": 65.80511182108626,
+ "grad_norm": 0.1845703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0258,
+ "step": 20597
+ },
+ {
+ "epoch": 65.80830670926518,
+ "grad_norm": 0.1875,
+ "learning_rate": 0.0005,
+ "loss": 1.0185,
+ "step": 20598
+ },
+ {
+ "epoch": 65.81150159744409,
+ "grad_norm": 0.1611328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0202,
+ "step": 20599
+ },
+ {
+ "epoch": 65.814696485623,
+ "grad_norm": 0.134765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0338,
+ "step": 20600
+ },
+ {
+ "epoch": 65.81789137380191,
+ "grad_norm": 0.08544921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0254,
+ "step": 20601
+ },
+ {
+ "epoch": 65.82108626198082,
+ "grad_norm": 0.12109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0206,
+ "step": 20602
+ },
+ {
+ "epoch": 65.82428115015975,
+ "grad_norm": 0.0869140625,
+ "learning_rate": 0.0005,
+ "loss": 1.033,
+ "step": 20603
+ },
+ {
+ "epoch": 65.82747603833866,
+ "grad_norm": 0.048583984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0147,
+ "step": 20604
+ },
+ {
+ "epoch": 65.83067092651757,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.017,
+ "step": 20605
+ },
+ {
+ "epoch": 65.83386581469648,
+ "grad_norm": 0.06103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0236,
+ "step": 20606
+ },
+ {
+ "epoch": 65.8370607028754,
+ "grad_norm": 0.048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0193,
+ "step": 20607
+ },
+ {
+ "epoch": 65.84025559105432,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0287,
+ "step": 20608
+ },
+ {
+ "epoch": 65.84345047923323,
+ "grad_norm": 0.0537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0174,
+ "step": 20609
+ },
+ {
+ "epoch": 65.84664536741214,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0213,
+ "step": 20610
+ },
+ {
+ "epoch": 65.84984025559105,
+ "grad_norm": 0.058837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0204,
+ "step": 20611
+ },
+ {
+ "epoch": 65.85303514376997,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0117,
+ "step": 20612
+ },
+ {
+ "epoch": 65.85623003194888,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0171,
+ "step": 20613
+ },
+ {
+ "epoch": 65.8594249201278,
+ "grad_norm": 0.056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0282,
+ "step": 20614
+ },
+ {
+ "epoch": 65.8626198083067,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0274,
+ "step": 20615
+ },
+ {
+ "epoch": 65.86581469648563,
+ "grad_norm": 0.107421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0231,
+ "step": 20616
+ },
+ {
+ "epoch": 65.86900958466454,
+ "grad_norm": 0.061767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0232,
+ "step": 20617
+ },
+ {
+ "epoch": 65.87220447284345,
+ "grad_norm": 0.10009765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0205,
+ "step": 20618
+ },
+ {
+ "epoch": 65.87539936102236,
+ "grad_norm": 0.052490234375,
+ "learning_rate": 0.0005,
+ "loss": 1.035,
+ "step": 20619
+ },
+ {
+ "epoch": 65.87859424920129,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0278,
+ "step": 20620
+ },
+ {
+ "epoch": 65.8817891373802,
+ "grad_norm": 0.1171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0112,
+ "step": 20621
+ },
+ {
+ "epoch": 65.8849840255591,
+ "grad_norm": 0.11181640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0277,
+ "step": 20622
+ },
+ {
+ "epoch": 65.88817891373802,
+ "grad_norm": 0.115234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0249,
+ "step": 20623
+ },
+ {
+ "epoch": 65.89137380191693,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0231,
+ "step": 20624
+ },
+ {
+ "epoch": 65.89456869009585,
+ "grad_norm": 0.13671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0313,
+ "step": 20625
+ },
+ {
+ "epoch": 65.89776357827476,
+ "grad_norm": 0.05908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0209,
+ "step": 20626
+ },
+ {
+ "epoch": 65.90095846645367,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0124,
+ "step": 20627
+ },
+ {
+ "epoch": 65.90415335463258,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0358,
+ "step": 20628
+ },
+ {
+ "epoch": 65.90734824281151,
+ "grad_norm": 0.057861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0296,
+ "step": 20629
+ },
+ {
+ "epoch": 65.91054313099042,
+ "grad_norm": 0.060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0262,
+ "step": 20630
+ },
+ {
+ "epoch": 65.91373801916933,
+ "grad_norm": 0.06298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0407,
+ "step": 20631
+ },
+ {
+ "epoch": 65.91693290734824,
+ "grad_norm": 0.0546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0088,
+ "step": 20632
+ },
+ {
+ "epoch": 65.92012779552715,
+ "grad_norm": 0.052978515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0312,
+ "step": 20633
+ },
+ {
+ "epoch": 65.92332268370608,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0269,
+ "step": 20634
+ },
+ {
+ "epoch": 65.92651757188499,
+ "grad_norm": 0.053466796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0164,
+ "step": 20635
+ },
+ {
+ "epoch": 65.9297124600639,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0241,
+ "step": 20636
+ },
+ {
+ "epoch": 65.93290734824281,
+ "grad_norm": 0.056884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0204,
+ "step": 20637
+ },
+ {
+ "epoch": 65.93610223642173,
+ "grad_norm": 0.0498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0248,
+ "step": 20638
+ },
+ {
+ "epoch": 65.93929712460064,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0284,
+ "step": 20639
+ },
+ {
+ "epoch": 65.94249201277955,
+ "grad_norm": 0.056884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0157,
+ "step": 20640
+ },
+ {
+ "epoch": 65.94568690095846,
+ "grad_norm": 0.058349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0277,
+ "step": 20641
+ },
+ {
+ "epoch": 65.94888178913737,
+ "grad_norm": 0.056884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0268,
+ "step": 20642
+ },
+ {
+ "epoch": 65.9520766773163,
+ "grad_norm": 0.047119140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0303,
+ "step": 20643
+ },
+ {
+ "epoch": 65.95527156549521,
+ "grad_norm": 0.04736328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0232,
+ "step": 20644
+ },
+ {
+ "epoch": 65.95846645367412,
+ "grad_norm": 0.060791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0227,
+ "step": 20645
+ },
+ {
+ "epoch": 65.96166134185303,
+ "grad_norm": 0.0478515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0209,
+ "step": 20646
+ },
+ {
+ "epoch": 65.96485623003196,
+ "grad_norm": 0.045166015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0224,
+ "step": 20647
+ },
+ {
+ "epoch": 65.96805111821087,
+ "grad_norm": 0.04931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0201,
+ "step": 20648
+ },
+ {
+ "epoch": 65.97124600638978,
+ "grad_norm": 0.0576171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0217,
+ "step": 20649
+ },
+ {
+ "epoch": 65.97444089456869,
+ "grad_norm": 0.047607421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0212,
+ "step": 20650
+ },
+ {
+ "epoch": 65.9776357827476,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0275,
+ "step": 20651
+ },
+ {
+ "epoch": 65.98083067092652,
+ "grad_norm": 0.05029296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0238,
+ "step": 20652
+ },
+ {
+ "epoch": 65.98402555910543,
+ "grad_norm": 0.043212890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0271,
+ "step": 20653
+ },
+ {
+ "epoch": 65.98722044728434,
+ "grad_norm": 0.0576171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0185,
+ "step": 20654
+ },
+ {
+ "epoch": 65.99041533546325,
+ "grad_norm": 0.046630859375,
+ "learning_rate": 0.0005,
+ "loss": 1.024,
+ "step": 20655
+ },
+ {
+ "epoch": 65.99361022364218,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0242,
+ "step": 20656
+ },
+ {
+ "epoch": 65.99680511182109,
+ "grad_norm": 0.059814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0286,
+ "step": 20657
+ },
+ {
+ "epoch": 66.0,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0173,
+ "step": 20658
+ },
+ {
+ "epoch": 66.00319488817891,
+ "grad_norm": 0.05078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0332,
+ "step": 20659
+ },
+ {
+ "epoch": 66.00638977635782,
+ "grad_norm": 0.055419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.021,
+ "step": 20660
+ },
+ {
+ "epoch": 66.00958466453675,
+ "grad_norm": 0.045166015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0066,
+ "step": 20661
+ },
+ {
+ "epoch": 66.01277955271566,
+ "grad_norm": 0.0546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0244,
+ "step": 20662
+ },
+ {
+ "epoch": 66.01597444089457,
+ "grad_norm": 0.056884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.023,
+ "step": 20663
+ },
+ {
+ "epoch": 66.01916932907348,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.025,
+ "step": 20664
+ },
+ {
+ "epoch": 66.0223642172524,
+ "grad_norm": 0.046630859375,
+ "learning_rate": 0.0005,
+ "loss": 1.018,
+ "step": 20665
+ },
+ {
+ "epoch": 66.02555910543131,
+ "grad_norm": 0.052978515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0196,
+ "step": 20666
+ },
+ {
+ "epoch": 66.02875399361022,
+ "grad_norm": 0.0498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0311,
+ "step": 20667
+ },
+ {
+ "epoch": 66.03194888178913,
+ "grad_norm": 0.04736328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0219,
+ "step": 20668
+ },
+ {
+ "epoch": 66.03514376996804,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0175,
+ "step": 20669
+ },
+ {
+ "epoch": 66.03833865814697,
+ "grad_norm": 0.0595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0345,
+ "step": 20670
+ },
+ {
+ "epoch": 66.04153354632588,
+ "grad_norm": 0.0908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0236,
+ "step": 20671
+ },
+ {
+ "epoch": 66.04472843450479,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0212,
+ "step": 20672
+ },
+ {
+ "epoch": 66.0479233226837,
+ "grad_norm": 0.0859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0305,
+ "step": 20673
+ },
+ {
+ "epoch": 66.05111821086263,
+ "grad_norm": 0.12109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0177,
+ "step": 20674
+ },
+ {
+ "epoch": 66.05431309904154,
+ "grad_norm": 0.052490234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0237,
+ "step": 20675
+ },
+ {
+ "epoch": 66.05750798722045,
+ "grad_norm": 0.0810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0202,
+ "step": 20676
+ },
+ {
+ "epoch": 66.06070287539936,
+ "grad_norm": 0.09814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0271,
+ "step": 20677
+ },
+ {
+ "epoch": 66.06389776357827,
+ "grad_norm": 0.055908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0255,
+ "step": 20678
+ },
+ {
+ "epoch": 66.06709265175719,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0185,
+ "step": 20679
+ },
+ {
+ "epoch": 66.0702875399361,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0292,
+ "step": 20680
+ },
+ {
+ "epoch": 66.07348242811501,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0338,
+ "step": 20681
+ },
+ {
+ "epoch": 66.07667731629392,
+ "grad_norm": 0.072265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0211,
+ "step": 20682
+ },
+ {
+ "epoch": 66.07987220447285,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0246,
+ "step": 20683
+ },
+ {
+ "epoch": 66.08306709265176,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0291,
+ "step": 20684
+ },
+ {
+ "epoch": 66.08626198083067,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0234,
+ "step": 20685
+ },
+ {
+ "epoch": 66.08945686900958,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0104,
+ "step": 20686
+ },
+ {
+ "epoch": 66.09265175718849,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0278,
+ "step": 20687
+ },
+ {
+ "epoch": 66.09584664536742,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0193,
+ "step": 20688
+ },
+ {
+ "epoch": 66.09904153354633,
+ "grad_norm": 0.061767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0149,
+ "step": 20689
+ },
+ {
+ "epoch": 66.10223642172524,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0196,
+ "step": 20690
+ },
+ {
+ "epoch": 66.10543130990415,
+ "grad_norm": 0.08349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0193,
+ "step": 20691
+ },
+ {
+ "epoch": 66.10862619808307,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0253,
+ "step": 20692
+ },
+ {
+ "epoch": 66.11182108626198,
+ "grad_norm": 0.051025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0281,
+ "step": 20693
+ },
+ {
+ "epoch": 66.1150159744409,
+ "grad_norm": 0.06591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0254,
+ "step": 20694
+ },
+ {
+ "epoch": 66.1182108626198,
+ "grad_norm": 0.060791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0306,
+ "step": 20695
+ },
+ {
+ "epoch": 66.12140575079871,
+ "grad_norm": 0.0517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0275,
+ "step": 20696
+ },
+ {
+ "epoch": 66.12460063897764,
+ "grad_norm": 0.0556640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0254,
+ "step": 20697
+ },
+ {
+ "epoch": 66.12779552715655,
+ "grad_norm": 0.04736328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0149,
+ "step": 20698
+ },
+ {
+ "epoch": 66.13099041533546,
+ "grad_norm": 0.0830078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0223,
+ "step": 20699
+ },
+ {
+ "epoch": 66.13418530351437,
+ "grad_norm": 0.051025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0163,
+ "step": 20700
+ },
+ {
+ "epoch": 66.1373801916933,
+ "grad_norm": 0.0458984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0307,
+ "step": 20701
+ },
+ {
+ "epoch": 66.1405750798722,
+ "grad_norm": 0.08447265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0203,
+ "step": 20702
+ },
+ {
+ "epoch": 66.14376996805112,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0197,
+ "step": 20703
+ },
+ {
+ "epoch": 66.14696485623003,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0203,
+ "step": 20704
+ },
+ {
+ "epoch": 66.15015974440895,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0219,
+ "step": 20705
+ },
+ {
+ "epoch": 66.15335463258786,
+ "grad_norm": 0.04638671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0268,
+ "step": 20706
+ },
+ {
+ "epoch": 66.15654952076677,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0258,
+ "step": 20707
+ },
+ {
+ "epoch": 66.15974440894568,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0116,
+ "step": 20708
+ },
+ {
+ "epoch": 66.1629392971246,
+ "grad_norm": 0.060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0138,
+ "step": 20709
+ },
+ {
+ "epoch": 66.16613418530352,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0195,
+ "step": 20710
+ },
+ {
+ "epoch": 66.16932907348243,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0227,
+ "step": 20711
+ },
+ {
+ "epoch": 66.17252396166134,
+ "grad_norm": 0.09912109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0201,
+ "step": 20712
+ },
+ {
+ "epoch": 66.17571884984025,
+ "grad_norm": 0.04638671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0225,
+ "step": 20713
+ },
+ {
+ "epoch": 66.17891373801918,
+ "grad_norm": 0.0546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0169,
+ "step": 20714
+ },
+ {
+ "epoch": 66.18210862619809,
+ "grad_norm": 0.0615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0238,
+ "step": 20715
+ },
+ {
+ "epoch": 66.185303514377,
+ "grad_norm": 0.061279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.028,
+ "step": 20716
+ },
+ {
+ "epoch": 66.18849840255591,
+ "grad_norm": 0.061279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0085,
+ "step": 20717
+ },
+ {
+ "epoch": 66.19169329073482,
+ "grad_norm": 0.055908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.022,
+ "step": 20718
+ },
+ {
+ "epoch": 66.19488817891374,
+ "grad_norm": 0.09228515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0211,
+ "step": 20719
+ },
+ {
+ "epoch": 66.19808306709265,
+ "grad_norm": 0.1240234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0129,
+ "step": 20720
+ },
+ {
+ "epoch": 66.20127795527156,
+ "grad_norm": 0.0908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0171,
+ "step": 20721
+ },
+ {
+ "epoch": 66.20447284345047,
+ "grad_norm": 0.05859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0168,
+ "step": 20722
+ },
+ {
+ "epoch": 66.2076677316294,
+ "grad_norm": 0.11962890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0328,
+ "step": 20723
+ },
+ {
+ "epoch": 66.21086261980831,
+ "grad_norm": 0.04833984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0272,
+ "step": 20724
+ },
+ {
+ "epoch": 66.21405750798722,
+ "grad_norm": 0.0888671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0185,
+ "step": 20725
+ },
+ {
+ "epoch": 66.21725239616613,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0253,
+ "step": 20726
+ },
+ {
+ "epoch": 66.22044728434504,
+ "grad_norm": 0.08447265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0169,
+ "step": 20727
+ },
+ {
+ "epoch": 66.22364217252397,
+ "grad_norm": 0.09912109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0287,
+ "step": 20728
+ },
+ {
+ "epoch": 66.22683706070288,
+ "grad_norm": 0.078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0336,
+ "step": 20729
+ },
+ {
+ "epoch": 66.23003194888179,
+ "grad_norm": 0.08251953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0178,
+ "step": 20730
+ },
+ {
+ "epoch": 66.2332268370607,
+ "grad_norm": 0.052001953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0272,
+ "step": 20731
+ },
+ {
+ "epoch": 66.23642172523962,
+ "grad_norm": 0.07421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0265,
+ "step": 20732
+ },
+ {
+ "epoch": 66.23961661341853,
+ "grad_norm": 0.051025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0318,
+ "step": 20733
+ },
+ {
+ "epoch": 66.24281150159744,
+ "grad_norm": 0.103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0127,
+ "step": 20734
+ },
+ {
+ "epoch": 66.24600638977635,
+ "grad_norm": 0.061767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0217,
+ "step": 20735
+ },
+ {
+ "epoch": 66.24920127795527,
+ "grad_norm": 0.11865234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0126,
+ "step": 20736
+ },
+ {
+ "epoch": 66.25239616613419,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0379,
+ "step": 20737
+ },
+ {
+ "epoch": 66.2555910543131,
+ "grad_norm": 0.0849609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0205,
+ "step": 20738
+ },
+ {
+ "epoch": 66.25878594249201,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.019,
+ "step": 20739
+ },
+ {
+ "epoch": 66.26198083067092,
+ "grad_norm": 0.06396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0231,
+ "step": 20740
+ },
+ {
+ "epoch": 66.26517571884985,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0212,
+ "step": 20741
+ },
+ {
+ "epoch": 66.26837060702876,
+ "grad_norm": 0.05908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0124,
+ "step": 20742
+ },
+ {
+ "epoch": 66.27156549520767,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0178,
+ "step": 20743
+ },
+ {
+ "epoch": 66.27476038338658,
+ "grad_norm": 0.060302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0098,
+ "step": 20744
+ },
+ {
+ "epoch": 66.27795527156549,
+ "grad_norm": 0.0615234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0231,
+ "step": 20745
+ },
+ {
+ "epoch": 66.28115015974441,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0209,
+ "step": 20746
+ },
+ {
+ "epoch": 66.28434504792332,
+ "grad_norm": 0.056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0314,
+ "step": 20747
+ },
+ {
+ "epoch": 66.28753993610223,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0214,
+ "step": 20748
+ },
+ {
+ "epoch": 66.29073482428115,
+ "grad_norm": 0.045166015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0282,
+ "step": 20749
+ },
+ {
+ "epoch": 66.29392971246007,
+ "grad_norm": 0.1025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0272,
+ "step": 20750
+ },
+ {
+ "epoch": 66.29712460063898,
+ "grad_norm": 0.05224609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0228,
+ "step": 20751
+ },
+ {
+ "epoch": 66.30031948881789,
+ "grad_norm": 0.06494140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0261,
+ "step": 20752
+ },
+ {
+ "epoch": 66.3035143769968,
+ "grad_norm": 0.052734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0223,
+ "step": 20753
+ },
+ {
+ "epoch": 66.30670926517571,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0232,
+ "step": 20754
+ },
+ {
+ "epoch": 66.30990415335464,
+ "grad_norm": 0.05322265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0243,
+ "step": 20755
+ },
+ {
+ "epoch": 66.31309904153355,
+ "grad_norm": 0.06103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.028,
+ "step": 20756
+ },
+ {
+ "epoch": 66.31629392971246,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0248,
+ "step": 20757
+ },
+ {
+ "epoch": 66.31948881789137,
+ "grad_norm": 0.0537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0231,
+ "step": 20758
+ },
+ {
+ "epoch": 66.3226837060703,
+ "grad_norm": 0.045166015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0139,
+ "step": 20759
+ },
+ {
+ "epoch": 66.3258785942492,
+ "grad_norm": 0.046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0373,
+ "step": 20760
+ },
+ {
+ "epoch": 66.32907348242811,
+ "grad_norm": 0.052734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0122,
+ "step": 20761
+ },
+ {
+ "epoch": 66.33226837060703,
+ "grad_norm": 0.052490234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0229,
+ "step": 20762
+ },
+ {
+ "epoch": 66.33546325878594,
+ "grad_norm": 0.056396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0353,
+ "step": 20763
+ },
+ {
+ "epoch": 66.33865814696486,
+ "grad_norm": 0.05029296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0081,
+ "step": 20764
+ },
+ {
+ "epoch": 66.34185303514377,
+ "grad_norm": 0.0517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0158,
+ "step": 20765
+ },
+ {
+ "epoch": 66.34504792332268,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.017,
+ "step": 20766
+ },
+ {
+ "epoch": 66.34824281150159,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0202,
+ "step": 20767
+ },
+ {
+ "epoch": 66.35143769968052,
+ "grad_norm": 0.060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0184,
+ "step": 20768
+ },
+ {
+ "epoch": 66.35463258785943,
+ "grad_norm": 0.057861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0256,
+ "step": 20769
+ },
+ {
+ "epoch": 66.35782747603834,
+ "grad_norm": 0.052001953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0259,
+ "step": 20770
+ },
+ {
+ "epoch": 66.36102236421725,
+ "grad_norm": 0.052490234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0287,
+ "step": 20771
+ },
+ {
+ "epoch": 66.36421725239616,
+ "grad_norm": 0.050537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0279,
+ "step": 20772
+ },
+ {
+ "epoch": 66.36741214057508,
+ "grad_norm": 0.052490234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0171,
+ "step": 20773
+ },
+ {
+ "epoch": 66.370607028754,
+ "grad_norm": 0.05859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0171,
+ "step": 20774
+ },
+ {
+ "epoch": 66.3738019169329,
+ "grad_norm": 0.060791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0174,
+ "step": 20775
+ },
+ {
+ "epoch": 66.37699680511182,
+ "grad_norm": 0.052978515625,
+ "learning_rate": 0.0005,
+ "loss": 1.023,
+ "step": 20776
+ },
+ {
+ "epoch": 66.38019169329074,
+ "grad_norm": 0.059326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0356,
+ "step": 20777
+ },
+ {
+ "epoch": 66.38338658146965,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0169,
+ "step": 20778
+ },
+ {
+ "epoch": 66.38658146964856,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.028,
+ "step": 20779
+ },
+ {
+ "epoch": 66.38977635782747,
+ "grad_norm": 0.138671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0165,
+ "step": 20780
+ },
+ {
+ "epoch": 66.3929712460064,
+ "grad_norm": 0.1435546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0207,
+ "step": 20781
+ },
+ {
+ "epoch": 66.39616613418531,
+ "grad_norm": 0.1591796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0264,
+ "step": 20782
+ },
+ {
+ "epoch": 66.39936102236422,
+ "grad_norm": 0.1298828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0272,
+ "step": 20783
+ },
+ {
+ "epoch": 66.40255591054313,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0168,
+ "step": 20784
+ },
+ {
+ "epoch": 66.40575079872204,
+ "grad_norm": 0.1201171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0105,
+ "step": 20785
+ },
+ {
+ "epoch": 66.40894568690096,
+ "grad_norm": 0.08447265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0176,
+ "step": 20786
+ },
+ {
+ "epoch": 66.41214057507987,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0145,
+ "step": 20787
+ },
+ {
+ "epoch": 66.41533546325878,
+ "grad_norm": 0.1513671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0203,
+ "step": 20788
+ },
+ {
+ "epoch": 66.4185303514377,
+ "grad_norm": 0.0859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0185,
+ "step": 20789
+ },
+ {
+ "epoch": 66.42172523961662,
+ "grad_norm": 0.09326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0082,
+ "step": 20790
+ },
+ {
+ "epoch": 66.42492012779553,
+ "grad_norm": 0.0888671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0205,
+ "step": 20791
+ },
+ {
+ "epoch": 66.42811501597444,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0266,
+ "step": 20792
+ },
+ {
+ "epoch": 66.43130990415335,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0244,
+ "step": 20793
+ },
+ {
+ "epoch": 66.43450479233226,
+ "grad_norm": 0.09033203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0252,
+ "step": 20794
+ },
+ {
+ "epoch": 66.43769968051119,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0249,
+ "step": 20795
+ },
+ {
+ "epoch": 66.4408945686901,
+ "grad_norm": 0.08154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0276,
+ "step": 20796
+ },
+ {
+ "epoch": 66.44408945686901,
+ "grad_norm": 0.0859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0282,
+ "step": 20797
+ },
+ {
+ "epoch": 66.44728434504792,
+ "grad_norm": 0.0556640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0253,
+ "step": 20798
+ },
+ {
+ "epoch": 66.45047923322684,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0243,
+ "step": 20799
+ },
+ {
+ "epoch": 66.45367412140575,
+ "grad_norm": 0.056884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.02,
+ "step": 20800
+ },
+ {
+ "epoch": 66.45686900958466,
+ "grad_norm": 0.054443359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0256,
+ "step": 20801
+ },
+ {
+ "epoch": 66.46006389776358,
+ "grad_norm": 0.0654296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0346,
+ "step": 20802
+ },
+ {
+ "epoch": 66.46325878594249,
+ "grad_norm": 0.0478515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0224,
+ "step": 20803
+ },
+ {
+ "epoch": 66.46645367412141,
+ "grad_norm": 0.052978515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0124,
+ "step": 20804
+ },
+ {
+ "epoch": 66.46964856230032,
+ "grad_norm": 0.05517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0265,
+ "step": 20805
+ },
+ {
+ "epoch": 66.47284345047923,
+ "grad_norm": 0.048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0141,
+ "step": 20806
+ },
+ {
+ "epoch": 66.47603833865814,
+ "grad_norm": 0.061279296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0167,
+ "step": 20807
+ },
+ {
+ "epoch": 66.47923322683707,
+ "grad_norm": 0.054443359375,
+ "learning_rate": 0.0005,
+ "loss": 1.025,
+ "step": 20808
+ },
+ {
+ "epoch": 66.48242811501598,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0088,
+ "step": 20809
+ },
+ {
+ "epoch": 66.48562300319489,
+ "grad_norm": 0.04638671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0241,
+ "step": 20810
+ },
+ {
+ "epoch": 66.4888178913738,
+ "grad_norm": 0.061767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0218,
+ "step": 20811
+ },
+ {
+ "epoch": 66.49201277955271,
+ "grad_norm": 0.049560546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0237,
+ "step": 20812
+ },
+ {
+ "epoch": 66.49520766773163,
+ "grad_norm": 0.0546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0253,
+ "step": 20813
+ },
+ {
+ "epoch": 66.49840255591054,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0205,
+ "step": 20814
+ },
+ {
+ "epoch": 66.50159744408946,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0226,
+ "step": 20815
+ },
+ {
+ "epoch": 66.50479233226837,
+ "grad_norm": 0.08056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0271,
+ "step": 20816
+ },
+ {
+ "epoch": 66.50798722044729,
+ "grad_norm": 0.05517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0349,
+ "step": 20817
+ },
+ {
+ "epoch": 66.5111821086262,
+ "grad_norm": 0.099609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0278,
+ "step": 20818
+ },
+ {
+ "epoch": 66.51437699680511,
+ "grad_norm": 0.0810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0275,
+ "step": 20819
+ },
+ {
+ "epoch": 66.51757188498402,
+ "grad_norm": 0.189453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0324,
+ "step": 20820
+ },
+ {
+ "epoch": 66.52076677316293,
+ "grad_norm": 0.05859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0232,
+ "step": 20821
+ },
+ {
+ "epoch": 66.52396166134186,
+ "grad_norm": 0.1533203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0271,
+ "step": 20822
+ },
+ {
+ "epoch": 66.52715654952077,
+ "grad_norm": 0.10986328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0266,
+ "step": 20823
+ },
+ {
+ "epoch": 66.53035143769968,
+ "grad_norm": 0.1318359375,
+ "learning_rate": 0.0005,
+ "loss": 1.01,
+ "step": 20824
+ },
+ {
+ "epoch": 66.53354632587859,
+ "grad_norm": 0.12255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.023,
+ "step": 20825
+ },
+ {
+ "epoch": 66.53674121405751,
+ "grad_norm": 0.1484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0356,
+ "step": 20826
+ },
+ {
+ "epoch": 66.53993610223642,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0261,
+ "step": 20827
+ },
+ {
+ "epoch": 66.54313099041534,
+ "grad_norm": 0.1005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0174,
+ "step": 20828
+ },
+ {
+ "epoch": 66.54632587859425,
+ "grad_norm": 0.064453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0161,
+ "step": 20829
+ },
+ {
+ "epoch": 66.54952076677316,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0211,
+ "step": 20830
+ },
+ {
+ "epoch": 66.55271565495208,
+ "grad_norm": 0.1103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0315,
+ "step": 20831
+ },
+ {
+ "epoch": 66.55591054313099,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0267,
+ "step": 20832
+ },
+ {
+ "epoch": 66.5591054313099,
+ "grad_norm": 0.087890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0162,
+ "step": 20833
+ },
+ {
+ "epoch": 66.56230031948881,
+ "grad_norm": 0.0537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0217,
+ "step": 20834
+ },
+ {
+ "epoch": 66.56549520766774,
+ "grad_norm": 0.06005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0378,
+ "step": 20835
+ },
+ {
+ "epoch": 66.56869009584665,
+ "grad_norm": 0.055908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0127,
+ "step": 20836
+ },
+ {
+ "epoch": 66.57188498402556,
+ "grad_norm": 0.055908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0202,
+ "step": 20837
+ },
+ {
+ "epoch": 66.57507987220447,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0186,
+ "step": 20838
+ },
+ {
+ "epoch": 66.57827476038338,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0187,
+ "step": 20839
+ },
+ {
+ "epoch": 66.5814696485623,
+ "grad_norm": 0.046630859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0323,
+ "step": 20840
+ },
+ {
+ "epoch": 66.58466453674122,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0221,
+ "step": 20841
+ },
+ {
+ "epoch": 66.58785942492013,
+ "grad_norm": 0.0576171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0167,
+ "step": 20842
+ },
+ {
+ "epoch": 66.59105431309904,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0154,
+ "step": 20843
+ },
+ {
+ "epoch": 66.59424920127796,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.017,
+ "step": 20844
+ },
+ {
+ "epoch": 66.59744408945687,
+ "grad_norm": 0.05322265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0174,
+ "step": 20845
+ },
+ {
+ "epoch": 66.60063897763578,
+ "grad_norm": 0.0927734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0213,
+ "step": 20846
+ },
+ {
+ "epoch": 66.60383386581469,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0282,
+ "step": 20847
+ },
+ {
+ "epoch": 66.6070287539936,
+ "grad_norm": 0.0869140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0362,
+ "step": 20848
+ },
+ {
+ "epoch": 66.61022364217253,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0247,
+ "step": 20849
+ },
+ {
+ "epoch": 66.61341853035144,
+ "grad_norm": 0.056396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0314,
+ "step": 20850
+ },
+ {
+ "epoch": 66.61661341853035,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0202,
+ "step": 20851
+ },
+ {
+ "epoch": 66.61980830670926,
+ "grad_norm": 0.05224609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0159,
+ "step": 20852
+ },
+ {
+ "epoch": 66.62300319488818,
+ "grad_norm": 0.08544921875,
+ "learning_rate": 0.0005,
+ "loss": 1.024,
+ "step": 20853
+ },
+ {
+ "epoch": 66.6261980830671,
+ "grad_norm": 0.059326171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0179,
+ "step": 20854
+ },
+ {
+ "epoch": 66.629392971246,
+ "grad_norm": 0.05029296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0325,
+ "step": 20855
+ },
+ {
+ "epoch": 66.63258785942492,
+ "grad_norm": 0.061767578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0264,
+ "step": 20856
+ },
+ {
+ "epoch": 66.63578274760384,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0174,
+ "step": 20857
+ },
+ {
+ "epoch": 66.63897763578275,
+ "grad_norm": 0.05517578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0132,
+ "step": 20858
+ },
+ {
+ "epoch": 66.64217252396166,
+ "grad_norm": 0.060791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.013,
+ "step": 20859
+ },
+ {
+ "epoch": 66.64536741214057,
+ "grad_norm": 0.0537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0156,
+ "step": 20860
+ },
+ {
+ "epoch": 66.64856230031948,
+ "grad_norm": 0.052978515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0181,
+ "step": 20861
+ },
+ {
+ "epoch": 66.65175718849841,
+ "grad_norm": 0.046142578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0168,
+ "step": 20862
+ },
+ {
+ "epoch": 66.65495207667732,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0257,
+ "step": 20863
+ },
+ {
+ "epoch": 66.65814696485623,
+ "grad_norm": 0.046142578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0201,
+ "step": 20864
+ },
+ {
+ "epoch": 66.66134185303514,
+ "grad_norm": 0.05126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0305,
+ "step": 20865
+ },
+ {
+ "epoch": 66.66453674121406,
+ "grad_norm": 0.0498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0261,
+ "step": 20866
+ },
+ {
+ "epoch": 66.66773162939297,
+ "grad_norm": 0.04443359375,
+ "learning_rate": 0.0005,
+ "loss": 1.019,
+ "step": 20867
+ },
+ {
+ "epoch": 66.67092651757189,
+ "grad_norm": 0.058837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0144,
+ "step": 20868
+ },
+ {
+ "epoch": 66.6741214057508,
+ "grad_norm": 0.08447265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0279,
+ "step": 20869
+ },
+ {
+ "epoch": 66.6773162939297,
+ "grad_norm": 0.04541015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0314,
+ "step": 20870
+ },
+ {
+ "epoch": 66.68051118210863,
+ "grad_norm": 0.11328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0159,
+ "step": 20871
+ },
+ {
+ "epoch": 66.68370607028754,
+ "grad_norm": 0.06640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0207,
+ "step": 20872
+ },
+ {
+ "epoch": 66.68690095846645,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0161,
+ "step": 20873
+ },
+ {
+ "epoch": 66.69009584664536,
+ "grad_norm": 0.052978515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0237,
+ "step": 20874
+ },
+ {
+ "epoch": 66.69329073482429,
+ "grad_norm": 0.053955078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0193,
+ "step": 20875
+ },
+ {
+ "epoch": 66.6964856230032,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0235,
+ "step": 20876
+ },
+ {
+ "epoch": 66.69968051118211,
+ "grad_norm": 0.076171875,
+ "learning_rate": 0.0005,
+ "loss": 1.025,
+ "step": 20877
+ },
+ {
+ "epoch": 66.70287539936102,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0252,
+ "step": 20878
+ },
+ {
+ "epoch": 66.70607028753993,
+ "grad_norm": 0.04345703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0247,
+ "step": 20879
+ },
+ {
+ "epoch": 66.70926517571885,
+ "grad_norm": 0.10498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0215,
+ "step": 20880
+ },
+ {
+ "epoch": 66.71246006389777,
+ "grad_norm": 0.059814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0256,
+ "step": 20881
+ },
+ {
+ "epoch": 66.71565495207668,
+ "grad_norm": 0.0498046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0222,
+ "step": 20882
+ },
+ {
+ "epoch": 66.71884984025559,
+ "grad_norm": 0.062255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0213,
+ "step": 20883
+ },
+ {
+ "epoch": 66.72204472843451,
+ "grad_norm": 0.08349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0179,
+ "step": 20884
+ },
+ {
+ "epoch": 66.72523961661342,
+ "grad_norm": 0.04296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0302,
+ "step": 20885
+ },
+ {
+ "epoch": 66.72843450479233,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0256,
+ "step": 20886
+ },
+ {
+ "epoch": 66.73162939297124,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0103,
+ "step": 20887
+ },
+ {
+ "epoch": 66.73482428115015,
+ "grad_norm": 0.053955078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0389,
+ "step": 20888
+ },
+ {
+ "epoch": 66.73801916932908,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0231,
+ "step": 20889
+ },
+ {
+ "epoch": 66.74121405750799,
+ "grad_norm": 0.047119140625,
+ "learning_rate": 0.0005,
+ "loss": 1.019,
+ "step": 20890
+ },
+ {
+ "epoch": 66.7444089456869,
+ "grad_norm": 0.060546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0227,
+ "step": 20891
+ },
+ {
+ "epoch": 66.74760383386581,
+ "grad_norm": 0.0458984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0229,
+ "step": 20892
+ },
+ {
+ "epoch": 66.75079872204473,
+ "grad_norm": 0.04736328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0133,
+ "step": 20893
+ },
+ {
+ "epoch": 66.75399361022365,
+ "grad_norm": 0.05224609375,
+ "learning_rate": 0.0005,
+ "loss": 1.027,
+ "step": 20894
+ },
+ {
+ "epoch": 66.75718849840256,
+ "grad_norm": 0.05126953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0162,
+ "step": 20895
+ },
+ {
+ "epoch": 66.76038338658147,
+ "grad_norm": 0.0703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0198,
+ "step": 20896
+ },
+ {
+ "epoch": 66.76357827476038,
+ "grad_norm": 0.0791015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0174,
+ "step": 20897
+ },
+ {
+ "epoch": 66.7667731629393,
+ "grad_norm": 0.07763671875,
+ "learning_rate": 0.0005,
+ "loss": 1.026,
+ "step": 20898
+ },
+ {
+ "epoch": 66.76996805111821,
+ "grad_norm": 0.06689453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0208,
+ "step": 20899
+ },
+ {
+ "epoch": 66.77316293929712,
+ "grad_norm": 0.08349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0208,
+ "step": 20900
+ },
+ {
+ "epoch": 66.77635782747603,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0178,
+ "step": 20901
+ },
+ {
+ "epoch": 66.77955271565496,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0275,
+ "step": 20902
+ },
+ {
+ "epoch": 66.78274760383387,
+ "grad_norm": 0.09814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0261,
+ "step": 20903
+ },
+ {
+ "epoch": 66.78594249201278,
+ "grad_norm": 0.051025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0238,
+ "step": 20904
+ },
+ {
+ "epoch": 66.78913738019169,
+ "grad_norm": 0.10302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.025,
+ "step": 20905
+ },
+ {
+ "epoch": 66.7923322683706,
+ "grad_norm": 0.08203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0359,
+ "step": 20906
+ },
+ {
+ "epoch": 66.79552715654953,
+ "grad_norm": 0.050048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0182,
+ "step": 20907
+ },
+ {
+ "epoch": 66.79872204472844,
+ "grad_norm": 0.1337890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0286,
+ "step": 20908
+ },
+ {
+ "epoch": 66.80191693290735,
+ "grad_norm": 0.07373046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0276,
+ "step": 20909
+ },
+ {
+ "epoch": 66.80511182108626,
+ "grad_norm": 0.12255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0238,
+ "step": 20910
+ },
+ {
+ "epoch": 66.80830670926518,
+ "grad_norm": 0.1416015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0289,
+ "step": 20911
+ },
+ {
+ "epoch": 66.81150159744409,
+ "grad_norm": 0.1474609375,
+ "learning_rate": 0.0005,
+ "loss": 1.019,
+ "step": 20912
+ },
+ {
+ "epoch": 66.814696485623,
+ "grad_norm": 0.125,
+ "learning_rate": 0.0005,
+ "loss": 1.0187,
+ "step": 20913
+ },
+ {
+ "epoch": 66.81789137380191,
+ "grad_norm": 0.146484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0267,
+ "step": 20914
+ },
+ {
+ "epoch": 66.82108626198082,
+ "grad_norm": 0.11962890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0327,
+ "step": 20915
+ },
+ {
+ "epoch": 66.82428115015975,
+ "grad_norm": 0.1533203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0231,
+ "step": 20916
+ },
+ {
+ "epoch": 66.82747603833866,
+ "grad_norm": 0.103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0171,
+ "step": 20917
+ },
+ {
+ "epoch": 66.83067092651757,
+ "grad_norm": 0.11962890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0317,
+ "step": 20918
+ },
+ {
+ "epoch": 66.83386581469648,
+ "grad_norm": 0.08837890625,
+ "learning_rate": 0.0005,
+ "loss": 1.0188,
+ "step": 20919
+ },
+ {
+ "epoch": 66.8370607028754,
+ "grad_norm": 0.08154296875,
+ "learning_rate": 0.0005,
+ "loss": 1.0119,
+ "step": 20920
+ },
+ {
+ "epoch": 66.84025559105432,
+ "grad_norm": 0.1025390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0312,
+ "step": 20921
+ },
+ {
+ "epoch": 66.84345047923323,
+ "grad_norm": 0.0712890625,
+ "learning_rate": 0.0005,
+ "loss": 1.031,
+ "step": 20922
+ },
+ {
+ "epoch": 66.84664536741214,
+ "grad_norm": 0.10009765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0298,
+ "step": 20923
+ },
+ {
+ "epoch": 66.84984025559105,
+ "grad_norm": 0.07666015625,
+ "learning_rate": 0.0005,
+ "loss": 1.021,
+ "step": 20924
+ },
+ {
+ "epoch": 66.85303514376997,
+ "grad_norm": 0.10693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0395,
+ "step": 20925
+ },
+ {
+ "epoch": 66.85623003194888,
+ "grad_norm": 0.10546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0251,
+ "step": 20926
+ },
+ {
+ "epoch": 66.8594249201278,
+ "grad_norm": 0.123046875,
+ "learning_rate": 0.0005,
+ "loss": 1.0215,
+ "step": 20927
+ },
+ {
+ "epoch": 66.8626198083067,
+ "grad_norm": 0.06005859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0342,
+ "step": 20928
+ },
+ {
+ "epoch": 66.86581469648563,
+ "grad_norm": 0.1171875,
+ "learning_rate": 0.0005,
+ "loss": 1.0168,
+ "step": 20929
+ },
+ {
+ "epoch": 66.86900958466454,
+ "grad_norm": 0.07861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0198,
+ "step": 20930
+ },
+ {
+ "epoch": 66.87220447284345,
+ "grad_norm": 0.0830078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0272,
+ "step": 20931
+ },
+ {
+ "epoch": 66.87539936102236,
+ "grad_norm": 0.0751953125,
+ "learning_rate": 0.0005,
+ "loss": 1.0304,
+ "step": 20932
+ },
+ {
+ "epoch": 66.87859424920129,
+ "grad_norm": 0.0732421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0197,
+ "step": 20933
+ },
+ {
+ "epoch": 66.8817891373802,
+ "grad_norm": 0.054443359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0239,
+ "step": 20934
+ },
+ {
+ "epoch": 66.8849840255591,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0191,
+ "step": 20935
+ },
+ {
+ "epoch": 66.88817891373802,
+ "grad_norm": 0.062255859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0198,
+ "step": 20936
+ },
+ {
+ "epoch": 66.89137380191693,
+ "grad_norm": 0.058349609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0284,
+ "step": 20937
+ },
+ {
+ "epoch": 66.89456869009585,
+ "grad_norm": 0.059814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0344,
+ "step": 20938
+ },
+ {
+ "epoch": 66.89776357827476,
+ "grad_norm": 0.060302734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0241,
+ "step": 20939
+ },
+ {
+ "epoch": 66.90095846645367,
+ "grad_norm": 0.0830078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0155,
+ "step": 20940
+ },
+ {
+ "epoch": 66.90415335463258,
+ "grad_norm": 0.0625,
+ "learning_rate": 0.0005,
+ "loss": 1.0326,
+ "step": 20941
+ },
+ {
+ "epoch": 66.90734824281151,
+ "grad_norm": 0.06787109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0203,
+ "step": 20942
+ },
+ {
+ "epoch": 66.91054313099042,
+ "grad_norm": 0.05224609375,
+ "learning_rate": 0.0005,
+ "loss": 1.0209,
+ "step": 20943
+ },
+ {
+ "epoch": 66.91373801916933,
+ "grad_norm": 0.06103515625,
+ "learning_rate": 0.0005,
+ "loss": 1.02,
+ "step": 20944
+ },
+ {
+ "epoch": 66.91693290734824,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0195,
+ "step": 20945
+ },
+ {
+ "epoch": 66.92012779552715,
+ "grad_norm": 0.068359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0245,
+ "step": 20946
+ },
+ {
+ "epoch": 66.92332268370608,
+ "grad_norm": 0.0810546875,
+ "learning_rate": 0.0005,
+ "loss": 1.0177,
+ "step": 20947
+ },
+ {
+ "epoch": 66.92651757188499,
+ "grad_norm": 0.083984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0223,
+ "step": 20948
+ },
+ {
+ "epoch": 66.9297124600639,
+ "grad_norm": 0.10595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0162,
+ "step": 20949
+ },
+ {
+ "epoch": 66.93290734824281,
+ "grad_norm": 0.07470703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0146,
+ "step": 20950
+ },
+ {
+ "epoch": 66.93610223642173,
+ "grad_norm": 0.0947265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0248,
+ "step": 20951
+ },
+ {
+ "epoch": 66.93929712460064,
+ "grad_norm": 0.057861328125,
+ "learning_rate": 0.0005,
+ "loss": 1.0303,
+ "step": 20952
+ },
+ {
+ "epoch": 66.94249201277955,
+ "grad_norm": 0.0888671875,
+ "learning_rate": 0.0005,
+ "loss": 1.0103,
+ "step": 20953
+ },
+ {
+ "epoch": 66.94568690095846,
+ "grad_norm": 0.05908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0156,
+ "step": 20954
+ },
+ {
+ "epoch": 66.94888178913737,
+ "grad_norm": 0.059814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.0225,
+ "step": 20955
+ },
+ {
+ "epoch": 66.9520766773163,
+ "grad_norm": 0.07275390625,
+ "learning_rate": 0.0005,
+ "loss": 1.0092,
+ "step": 20956
+ },
+ {
+ "epoch": 66.95527156549521,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0191,
+ "step": 20957
+ },
+ {
+ "epoch": 66.95846645367412,
+ "grad_norm": 0.053955078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0297,
+ "step": 20958
+ },
+ {
+ "epoch": 66.96166134185303,
+ "grad_norm": 0.07177734375,
+ "learning_rate": 0.0005,
+ "loss": 1.0198,
+ "step": 20959
+ },
+ {
+ "epoch": 66.96485623003196,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0315,
+ "step": 20960
+ },
+ {
+ "epoch": 66.96805111821087,
+ "grad_norm": 0.055419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0251,
+ "step": 20961
+ },
+ {
+ "epoch": 66.97124600638978,
+ "grad_norm": 0.056396484375,
+ "learning_rate": 0.0005,
+ "loss": 1.0253,
+ "step": 20962
+ },
+ {
+ "epoch": 66.97444089456869,
+ "grad_norm": 0.054931640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0228,
+ "step": 20963
+ },
+ {
+ "epoch": 66.9776357827476,
+ "grad_norm": 0.05419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0164,
+ "step": 20964
+ },
+ {
+ "epoch": 66.98083067092652,
+ "grad_norm": 0.052490234375,
+ "learning_rate": 0.0005,
+ "loss": 1.0237,
+ "step": 20965
+ },
+ {
+ "epoch": 66.98402555910543,
+ "grad_norm": 0.05419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0321,
+ "step": 20966
+ },
+ {
+ "epoch": 66.98722044728434,
+ "grad_norm": 0.0537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0313,
+ "step": 20967
+ },
+ {
+ "epoch": 66.99041533546325,
+ "grad_norm": 0.055419921875,
+ "learning_rate": 0.0005,
+ "loss": 1.0207,
+ "step": 20968
+ },
+ {
+ "epoch": 66.99361022364218,
+ "grad_norm": 0.04541015625,
+ "learning_rate": 0.0005,
+ "loss": 1.0347,
+ "step": 20969
+ },
+ {
+ "epoch": 66.99680511182109,
+ "grad_norm": 0.053466796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0262,
+ "step": 20970
+ },
+ {
+ "epoch": 67.0,
+ "grad_norm": 0.04833984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0268,
+ "step": 20971
+ },
+ {
+ "epoch": 67.00319488817891,
+ "grad_norm": 0.050537109375,
+ "learning_rate": 0.0005,
+ "loss": 1.0157,
+ "step": 20972
+ },
+ {
+ "epoch": 67.00638977635782,
+ "grad_norm": 0.047119140625,
+ "learning_rate": 0.0005,
+ "loss": 1.0301,
+ "step": 20973
+ },
+ {
+ "epoch": 67.00958466453675,
+ "grad_norm": 0.07080078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0159,
+ "step": 20974
+ },
+ {
+ "epoch": 67.01277955271566,
+ "grad_norm": 0.0439453125,
+ "learning_rate": 0.0005,
+ "loss": 1.021,
+ "step": 20975
+ },
+ {
+ "epoch": 67.01597444089457,
+ "grad_norm": 0.054443359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0207,
+ "step": 20976
+ },
+ {
+ "epoch": 67.01916932907348,
+ "grad_norm": 0.056640625,
+ "learning_rate": 0.0005,
+ "loss": 1.0175,
+ "step": 20977
+ },
+ {
+ "epoch": 67.0223642172524,
+ "grad_norm": 0.06982421875,
+ "learning_rate": 0.0005,
+ "loss": 1.0241,
+ "step": 20978
+ },
+ {
+ "epoch": 67.02555910543131,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0282,
+ "step": 20979
+ },
+ {
+ "epoch": 67.02875399361022,
+ "grad_norm": 0.06884765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0229,
+ "step": 20980
+ },
+ {
+ "epoch": 67.03194888178913,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0212,
+ "step": 20981
+ },
+ {
+ "epoch": 67.03514376996804,
+ "grad_norm": 0.048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0199,
+ "step": 20982
+ },
+ {
+ "epoch": 67.03833865814697,
+ "grad_norm": 0.053955078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0294,
+ "step": 20983
+ },
+ {
+ "epoch": 67.04153354632588,
+ "grad_norm": 0.055908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0331,
+ "step": 20984
+ },
+ {
+ "epoch": 67.04472843450479,
+ "grad_norm": 0.0458984375,
+ "learning_rate": 0.0005,
+ "loss": 1.0209,
+ "step": 20985
+ },
+ {
+ "epoch": 67.0479233226837,
+ "grad_norm": 0.0673828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0107,
+ "step": 20986
+ },
+ {
+ "epoch": 67.05111821086263,
+ "grad_norm": 0.053466796875,
+ "learning_rate": 0.0005,
+ "loss": 1.0307,
+ "step": 20987
+ },
+ {
+ "epoch": 67.05431309904154,
+ "grad_norm": 0.0634765625,
+ "learning_rate": 0.0005,
+ "loss": 1.0155,
+ "step": 20988
+ },
+ {
+ "epoch": 67.05750798722045,
+ "grad_norm": 0.052978515625,
+ "learning_rate": 0.0005,
+ "loss": 1.0229,
+ "step": 20989
+ },
+ {
+ "epoch": 67.06070287539936,
+ "grad_norm": 0.059814453125,
+ "learning_rate": 0.0005,
+ "loss": 1.005,
+ "step": 20990
+ },
+ {
+ "epoch": 67.06389776357827,
+ "grad_norm": 0.05908203125,
+ "learning_rate": 0.0005,
+ "loss": 1.0119,
+ "step": 20991
+ },
+ {
+ "epoch": 67.06709265175719,
+ "grad_norm": 0.05859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0249,
+ "step": 20992
+ },
+ {
+ "epoch": 67.0702875399361,
+ "grad_norm": 0.04736328125,
+ "learning_rate": 0.0005,
+ "loss": 1.021,
+ "step": 20993
+ },
+ {
+ "epoch": 67.07348242811501,
+ "grad_norm": 0.050048828125,
+ "learning_rate": 0.0005,
+ "loss": 1.0237,
+ "step": 20994
+ },
+ {
+ "epoch": 67.07667731629392,
+ "grad_norm": 0.05078125,
+ "learning_rate": 0.0005,
+ "loss": 1.0231,
+ "step": 20995
+ },
+ {
+ "epoch": 67.07987220447285,
+ "grad_norm": 0.0693359375,
+ "learning_rate": 0.0005,
+ "loss": 1.0265,
+ "step": 20996
+ },
+ {
+ "epoch": 67.08306709265176,
+ "grad_norm": 0.046142578125,
+ "learning_rate": 0.0005,
+ "loss": 1.0177,
+ "step": 20997
+ },
+ {
+ "epoch": 67.08626198083067,
+ "grad_norm": 0.046630859375,
+ "learning_rate": 0.0005,
+ "loss": 1.0239,
+ "step": 20998
+ },
+ {
+ "epoch": 67.08945686900958,
+ "grad_norm": 0.0595703125,
+ "learning_rate": 0.0005,
+ "loss": 1.0164,
+ "step": 20999
+ },
+ {
+ "epoch": 67.09265175718849,
+ "grad_norm": 0.05322265625,
+ "learning_rate": 0.0005,
+ "loss": 1.0194,
+ "step": 21000
+ }
+ ],
+ "logging_steps": 1.0,
+ "max_steps": 187800,
+ "num_input_tokens_seen": 0,
+ "num_train_epochs": 600,
+ "save_steps": 1000,
+ "stateful_callbacks": {
+ "TrainerControl": {
+ "args": {
+ "should_epoch_stop": false,
+ "should_evaluate": false,
+ "should_log": false,
+ "should_save": true,
+ "should_training_stop": false
+ },
+ "attributes": {}
+ }
+ },
+ "total_flos": 3.0273567086990066e+19,
+ "train_batch_size": 256,
+ "trial_name": null,
+ "trial_params": null
+}