diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,25968 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 3.0,
+  "eval_steps": 500,
+  "global_step": 3705,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0008097165991902834,
+      "grad_norm": 8.212680311060101,
+      "learning_rate": 4.0000000000000003e-07,
+      "loss": 1.8584,
+      "step": 1
+    },
+    {
+      "epoch": 0.0016194331983805667,
+      "grad_norm": 8.289202083488915,
+      "learning_rate": 8.000000000000001e-07,
+      "loss": 1.8781,
+      "step": 2
+    },
+    {
+      "epoch": 0.0024291497975708503,
+      "grad_norm": 8.121793660526253,
+      "learning_rate": 1.2000000000000002e-06,
+      "loss": 1.8632,
+      "step": 3
+    },
+    {
+      "epoch": 0.0032388663967611335,
+      "grad_norm": 8.282365329091148,
+      "learning_rate": 1.6000000000000001e-06,
+      "loss": 1.8428,
+      "step": 4
+    },
+    {
+      "epoch": 0.004048582995951417,
+      "grad_norm": 6.205624862594086,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 1.8148,
+      "step": 5
+    },
+    {
+      "epoch": 0.004858299595141701,
+      "grad_norm": 3.874015809560423,
+      "learning_rate": 2.4000000000000003e-06,
+      "loss": 1.8443,
+      "step": 6
+    },
+    {
+      "epoch": 0.005668016194331984,
+      "grad_norm": 3.295685239888731,
+      "learning_rate": 2.8000000000000003e-06,
+      "loss": 1.7624,
+      "step": 7
+    },
+    {
+      "epoch": 0.006477732793522267,
+      "grad_norm": 3.35229547435271,
+      "learning_rate": 3.2000000000000003e-06,
+      "loss": 1.7892,
+      "step": 8
+    },
+    {
+      "epoch": 0.0072874493927125505,
+      "grad_norm": 3.6152531340095924,
+      "learning_rate": 3.6000000000000003e-06,
+      "loss": 1.7692,
+      "step": 9
+    },
+    {
+      "epoch": 0.008097165991902834,
+      "grad_norm": 2.839020140388393,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 1.7546,
+      "step": 10
+    },
+    {
+      "epoch": 0.008906882591093117,
+      "grad_norm": 3.280470763841352,
+      "learning_rate": 4.4e-06,
+      "loss": 1.781,
+      "step": 11
+    },
+    {
+      "epoch": 0.009716599190283401,
+      "grad_norm": 2.919143442971804,
+      "learning_rate": 4.800000000000001e-06,
+      "loss": 1.7255,
+      "step": 12
+    },
+    {
+      "epoch": 0.010526315789473684,
+      "grad_norm": 2.1780641385791957,
+      "learning_rate": 5.2e-06,
+      "loss": 1.7069,
+      "step": 13
+    },
+    {
+      "epoch": 0.011336032388663968,
+      "grad_norm": 2.0864543101995356,
+      "learning_rate": 5.600000000000001e-06,
+      "loss": 1.7801,
+      "step": 14
+    },
+    {
+      "epoch": 0.012145748987854251,
+      "grad_norm": 2.433989036218932,
+      "learning_rate": 6e-06,
+      "loss": 1.6886,
+      "step": 15
+    },
+    {
+      "epoch": 0.012955465587044534,
+      "grad_norm": 2.108153096043187,
+      "learning_rate": 6.4000000000000006e-06,
+      "loss": 1.7845,
+      "step": 16
+    },
+    {
+      "epoch": 0.013765182186234818,
+      "grad_norm": 2.147481330273472,
+      "learning_rate": 6.800000000000001e-06,
+      "loss": 1.7888,
+      "step": 17
+    },
+    {
+      "epoch": 0.014574898785425101,
+      "grad_norm": 2.0230155835526658,
+      "learning_rate": 7.2000000000000005e-06,
+      "loss": 1.7484,
+      "step": 18
+    },
+    {
+      "epoch": 0.015384615384615385,
+      "grad_norm": 1.8351338768344623,
+      "learning_rate": 7.600000000000001e-06,
+      "loss": 1.6815,
+      "step": 19
+    },
+    {
+      "epoch": 0.016194331983805668,
+      "grad_norm": 1.8385443541885478,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 1.7259,
+      "step": 20
+    },
+    {
+      "epoch": 0.01700404858299595,
+      "grad_norm": 1.9450404722821018,
+      "learning_rate": 8.400000000000001e-06,
+      "loss": 1.7513,
+      "step": 21
+    },
+    {
+      "epoch": 0.017813765182186234,
+      "grad_norm": 1.7329165869002578,
+      "learning_rate": 8.8e-06,
+      "loss": 1.6934,
+      "step": 22
+    },
+    {
+      "epoch": 0.01862348178137652,
+      "grad_norm": 1.8043465898806423,
+      "learning_rate": 9.200000000000002e-06,
+      "loss": 1.7411,
+      "step": 23
+    },
+    {
+      "epoch": 0.019433198380566803,
+      "grad_norm": 1.6575139155258072,
+      "learning_rate": 9.600000000000001e-06,
+      "loss": 1.7141,
+      "step": 24
+    },
+    {
+      "epoch": 0.020242914979757085,
+      "grad_norm": 1.6149477634364793,
+      "learning_rate": 1e-05,
+      "loss": 1.6359,
+      "step": 25
+    },
+    {
+      "epoch": 0.021052631578947368,
+      "grad_norm": 1.720501765932264,
+      "learning_rate": 1.04e-05,
+      "loss": 1.6983,
+      "step": 26
+    },
+    {
+      "epoch": 0.02186234817813765,
+      "grad_norm": 1.669970332575193,
+      "learning_rate": 1.0800000000000002e-05,
+      "loss": 1.7,
+      "step": 27
+    },
+    {
+      "epoch": 0.022672064777327937,
+      "grad_norm": 1.588644479806073,
+      "learning_rate": 1.1200000000000001e-05,
+      "loss": 1.6018,
+      "step": 28
+    },
+    {
+      "epoch": 0.02348178137651822,
+      "grad_norm": 1.6915129269465283,
+      "learning_rate": 1.16e-05,
+      "loss": 1.6732,
+      "step": 29
+    },
+    {
+      "epoch": 0.024291497975708502,
+      "grad_norm": 1.67842726423677,
+      "learning_rate": 1.2e-05,
+      "loss": 1.7291,
+      "step": 30
+    },
+    {
+      "epoch": 0.025101214574898785,
+      "grad_norm": 1.662602613195812,
+      "learning_rate": 1.2400000000000002e-05,
+      "loss": 1.714,
+      "step": 31
+    },
+    {
+      "epoch": 0.025910931174089068,
+      "grad_norm": 1.5692125530449308,
+      "learning_rate": 1.2800000000000001e-05,
+      "loss": 1.7681,
+      "step": 32
+    },
+    {
+      "epoch": 0.026720647773279354,
+      "grad_norm": 1.5593466931369984,
+      "learning_rate": 1.3200000000000002e-05,
+      "loss": 1.6625,
+      "step": 33
+    },
+    {
+      "epoch": 0.027530364372469637,
+      "grad_norm": 1.5314296376489476,
+      "learning_rate": 1.3600000000000002e-05,
+      "loss": 1.6711,
+      "step": 34
+    },
+    {
+      "epoch": 0.02834008097165992,
+      "grad_norm": 1.7793373457448158,
+      "learning_rate": 1.4e-05,
+      "loss": 1.7056,
+      "step": 35
+    },
+    {
+      "epoch": 0.029149797570850202,
+      "grad_norm": 1.5178537061767556,
+      "learning_rate": 1.4400000000000001e-05,
+      "loss": 1.7037,
+      "step": 36
+    },
+    {
+      "epoch": 0.029959514170040485,
+      "grad_norm": 1.5459895717695264,
+      "learning_rate": 1.48e-05,
+      "loss": 1.7683,
+      "step": 37
+    },
+    {
+      "epoch": 0.03076923076923077,
+      "grad_norm": 1.5735552822755368,
+      "learning_rate": 1.5200000000000002e-05,
+      "loss": 1.6872,
+      "step": 38
+    },
+    {
+      "epoch": 0.031578947368421054,
+      "grad_norm": 1.4358615071215153,
+      "learning_rate": 1.5600000000000003e-05,
+      "loss": 1.7402,
+      "step": 39
+    },
+    {
+      "epoch": 0.032388663967611336,
+      "grad_norm": 1.712555492023088,
+      "learning_rate": 1.6000000000000003e-05,
+      "loss": 1.7413,
+      "step": 40
+    },
+    {
+      "epoch": 0.03319838056680162,
+      "grad_norm": 1.527931327004627,
+      "learning_rate": 1.64e-05,
+      "loss": 1.7361,
+      "step": 41
+    },
+    {
+      "epoch": 0.0340080971659919,
+      "grad_norm": 1.5654078241724048,
+      "learning_rate": 1.6800000000000002e-05,
+      "loss": 1.7174,
+      "step": 42
+    },
+    {
+      "epoch": 0.034817813765182185,
+      "grad_norm": 1.5967020304138295,
+      "learning_rate": 1.72e-05,
+      "loss": 1.6863,
+      "step": 43
+    },
+    {
+      "epoch": 0.03562753036437247,
+      "grad_norm": 1.5449242343523315,
+      "learning_rate": 1.76e-05,
+      "loss": 1.7186,
+      "step": 44
+    },
+    {
+      "epoch": 0.03643724696356275,
+      "grad_norm": 1.5664894285095186,
+      "learning_rate": 1.8e-05,
+      "loss": 1.677,
+      "step": 45
+    },
+    {
+      "epoch": 0.03724696356275304,
+      "grad_norm": 1.6465253425137338,
+      "learning_rate": 1.8400000000000003e-05,
+      "loss": 1.767,
+      "step": 46
+    },
+    {
+      "epoch": 0.03805668016194332,
+      "grad_norm": 1.6352491164142435,
+      "learning_rate": 1.88e-05,
+      "loss": 1.5998,
+      "step": 47
+    },
+    {
+      "epoch": 0.038866396761133605,
+      "grad_norm": 1.6363309288994599,
+      "learning_rate": 1.9200000000000003e-05,
+      "loss": 1.6266,
+      "step": 48
+    },
+    {
+      "epoch": 0.03967611336032389,
+      "grad_norm": 1.8173439281602377,
+      "learning_rate": 1.9600000000000002e-05,
+      "loss": 1.6592,
+      "step": 49
+    },
+    {
+      "epoch": 0.04048582995951417,
+      "grad_norm": 1.5272192265553266,
+      "learning_rate": 2e-05,
+      "loss": 1.738,
+      "step": 50
+    },
+    {
+      "epoch": 0.04129554655870445,
+      "grad_norm": 1.8649531326109419,
+      "learning_rate": 1.9999996306016426e-05,
+      "loss": 1.7247,
+      "step": 51
+    },
+    {
+      "epoch": 0.042105263157894736,
+      "grad_norm": 1.7074872326048571,
+      "learning_rate": 1.9999985224068418e-05,
+      "loss": 1.7084,
+      "step": 52
+    },
+    {
+      "epoch": 0.04291497975708502,
+      "grad_norm": 1.4977194700952365,
+      "learning_rate": 1.9999966754164176e-05,
+      "loss": 1.7055,
+      "step": 53
+    },
+    {
+      "epoch": 0.0437246963562753,
+      "grad_norm": 1.6230626841795457,
+      "learning_rate": 1.9999940896317337e-05,
+      "loss": 1.5649,
+      "step": 54
+    },
+    {
+      "epoch": 0.044534412955465584,
+      "grad_norm": 1.6916647362719663,
+      "learning_rate": 1.9999907650547006e-05,
+      "loss": 1.7438,
+      "step": 55
+    },
+    {
+      "epoch": 0.045344129554655874,
+      "grad_norm": 1.4840562908731014,
+      "learning_rate": 1.999986701687775e-05,
+      "loss": 1.6202,
+      "step": 56
+    },
+    {
+      "epoch": 0.046153846153846156,
+      "grad_norm": 2.1064626020990596,
+      "learning_rate": 1.9999818995339587e-05,
+      "loss": 1.6874,
+      "step": 57
+    },
+    {
+      "epoch": 0.04696356275303644,
+      "grad_norm": 1.514821876037186,
+      "learning_rate": 1.999976358596799e-05,
+      "loss": 1.6145,
+      "step": 58
+    },
+    {
+      "epoch": 0.04777327935222672,
+      "grad_norm": 1.5837774362892976,
+      "learning_rate": 1.9999700788803902e-05,
+      "loss": 1.7424,
+      "step": 59
+    },
+    {
+      "epoch": 0.048582995951417005,
+      "grad_norm": 1.5015921648476362,
+      "learning_rate": 1.999963060389371e-05,
+      "loss": 1.7074,
+      "step": 60
+    },
+    {
+      "epoch": 0.04939271255060729,
+      "grad_norm": 1.5081987493072453,
+      "learning_rate": 1.9999553031289277e-05,
+      "loss": 1.7671,
+      "step": 61
+    },
+    {
+      "epoch": 0.05020242914979757,
+      "grad_norm": 1.449828681342794,
+      "learning_rate": 1.9999468071047904e-05,
+      "loss": 1.7608,
+      "step": 62
+    },
+    {
+      "epoch": 0.05101214574898785,
+      "grad_norm": 1.601994782437138,
+      "learning_rate": 1.9999375723232362e-05,
+      "loss": 1.7219,
+      "step": 63
+    },
+    {
+      "epoch": 0.051821862348178135,
+      "grad_norm": 1.5204297943669711,
+      "learning_rate": 1.999927598791088e-05,
+      "loss": 1.756,
+      "step": 64
+    },
+    {
+      "epoch": 0.05263157894736842,
+      "grad_norm": 1.4755593142051295,
+      "learning_rate": 1.9999168865157137e-05,
+      "loss": 1.6837,
+      "step": 65
+    },
+    {
+      "epoch": 0.05344129554655871,
+      "grad_norm": 1.5345109179851053,
+      "learning_rate": 1.999905435505028e-05,
+      "loss": 1.7347,
+      "step": 66
+    },
+    {
+      "epoch": 0.05425101214574899,
+      "grad_norm": 1.6001626392593242,
+      "learning_rate": 1.9998932457674904e-05,
+      "loss": 1.654,
+      "step": 67
+    },
+    {
+      "epoch": 0.05506072874493927,
+      "grad_norm": 1.4701409508033054,
+      "learning_rate": 1.999880317312107e-05,
+      "loss": 1.7161,
+      "step": 68
+    },
+    {
+      "epoch": 0.055870445344129556,
+      "grad_norm": 1.3692751510613166,
+      "learning_rate": 1.999866650148429e-05,
+      "loss": 1.7001,
+      "step": 69
+    },
+    {
+      "epoch": 0.05668016194331984,
+      "grad_norm": 1.48694331218701,
+      "learning_rate": 1.999852244286554e-05,
+      "loss": 1.812,
+      "step": 70
+    },
+    {
+      "epoch": 0.05748987854251012,
+      "grad_norm": 1.3924742554067722,
+      "learning_rate": 1.999837099737125e-05,
+      "loss": 1.6882,
+      "step": 71
+    },
+    {
+      "epoch": 0.058299595141700404,
+      "grad_norm": 1.4558556277711194,
+      "learning_rate": 1.9998212165113305e-05,
+      "loss": 1.7295,
+      "step": 72
+    },
+    {
+      "epoch": 0.05910931174089069,
+      "grad_norm": 1.4086032761044172,
+      "learning_rate": 1.999804594620905e-05,
+      "loss": 1.6551,
+      "step": 73
+    },
+    {
+      "epoch": 0.05991902834008097,
+      "grad_norm": 1.4939708802308547,
+      "learning_rate": 1.999787234078129e-05,
+      "loss": 1.7266,
+      "step": 74
+    },
+    {
+      "epoch": 0.06072874493927125,
+      "grad_norm": 1.4235299645544777,
+      "learning_rate": 1.9997691348958278e-05,
+      "loss": 1.7154,
+      "step": 75
+    },
+    {
+      "epoch": 0.06153846153846154,
+      "grad_norm": 1.644992651490282,
+      "learning_rate": 1.9997502970873736e-05,
+      "loss": 1.7414,
+      "step": 76
+    },
+    {
+      "epoch": 0.062348178137651825,
+      "grad_norm": 1.4731685460504425,
+      "learning_rate": 1.9997307206666835e-05,
+      "loss": 1.6953,
+      "step": 77
+    },
+    {
+      "epoch": 0.06315789473684211,
+      "grad_norm": 1.4684965944812223,
+      "learning_rate": 1.9997104056482206e-05,
+      "loss": 1.7127,
+      "step": 78
+    },
+    {
+      "epoch": 0.06396761133603239,
+      "grad_norm": 1.5057059153504087,
+      "learning_rate": 1.9996893520469934e-05,
+      "loss": 1.7393,
+      "step": 79
+    },
+    {
+      "epoch": 0.06477732793522267,
+      "grad_norm": 1.3158005918665274,
+      "learning_rate": 1.999667559878556e-05,
+      "loss": 1.6802,
+      "step": 80
+    },
+    {
+      "epoch": 0.06558704453441296,
+      "grad_norm": 1.405150913782594,
+      "learning_rate": 1.9996450291590093e-05,
+      "loss": 1.7208,
+      "step": 81
+    },
+    {
+      "epoch": 0.06639676113360324,
+      "grad_norm": 1.540226649005545,
+      "learning_rate": 1.9996217599049978e-05,
+      "loss": 1.7939,
+      "step": 82
+    },
+    {
+      "epoch": 0.06720647773279352,
+      "grad_norm": 1.4885304031726005,
+      "learning_rate": 1.9995977521337134e-05,
+      "loss": 1.6682,
+      "step": 83
+    },
+    {
+      "epoch": 0.0680161943319838,
+      "grad_norm": 1.3901851985498637,
+      "learning_rate": 1.9995730058628928e-05,
+      "loss": 1.667,
+      "step": 84
+    },
+    {
+      "epoch": 0.06882591093117409,
+      "grad_norm": 1.4725073250334886,
+      "learning_rate": 1.9995475211108183e-05,
+      "loss": 1.7547,
+      "step": 85
+    },
+    {
+      "epoch": 0.06963562753036437,
+      "grad_norm": 1.452556987551732,
+      "learning_rate": 1.9995212978963185e-05,
+      "loss": 1.7608,
+      "step": 86
+    },
+    {
+      "epoch": 0.07044534412955465,
+      "grad_norm": 1.4996449206080487,
+      "learning_rate": 1.9994943362387666e-05,
+      "loss": 1.6664,
+      "step": 87
+    },
+    {
+      "epoch": 0.07125506072874493,
+      "grad_norm": 1.4199683810583805,
+      "learning_rate": 1.9994666361580815e-05,
+      "loss": 1.6367,
+      "step": 88
+    },
+    {
+      "epoch": 0.07206477732793522,
+      "grad_norm": 1.571126361723556,
+      "learning_rate": 1.999438197674729e-05,
+      "loss": 1.7765,
+      "step": 89
+    },
+    {
+      "epoch": 0.0728744939271255,
+      "grad_norm": 1.3926335473604161,
+      "learning_rate": 1.9994090208097176e-05,
+      "loss": 1.722,
+      "step": 90
+    },
+    {
+      "epoch": 0.07368421052631578,
+      "grad_norm": 1.4552870672529465,
+      "learning_rate": 1.9993791055846048e-05,
+      "loss": 1.6651,
+      "step": 91
+    },
+    {
+      "epoch": 0.07449392712550608,
+      "grad_norm": 1.379931558535098,
+      "learning_rate": 1.999348452021491e-05,
+      "loss": 1.7096,
+      "step": 92
+    },
+    {
+      "epoch": 0.07530364372469636,
+      "grad_norm": 1.4198581520574205,
+      "learning_rate": 1.9993170601430233e-05,
+      "loss": 1.6601,
+      "step": 93
+    },
+    {
+      "epoch": 0.07611336032388664,
+      "grad_norm": 1.5078914507853327,
+      "learning_rate": 1.9992849299723933e-05,
+      "loss": 1.8005,
+      "step": 94
+    },
+    {
+      "epoch": 0.07692307692307693,
+      "grad_norm": 1.5076216489476741,
+      "learning_rate": 1.9992520615333393e-05,
+      "loss": 1.682,
+      "step": 95
+    },
+    {
+      "epoch": 0.07773279352226721,
+      "grad_norm": 1.3937658114601108,
+      "learning_rate": 1.9992184548501444e-05,
+      "loss": 1.6415,
+      "step": 96
+    },
+    {
+      "epoch": 0.07854251012145749,
+      "grad_norm": 1.5041710747062258,
+      "learning_rate": 1.9991841099476365e-05,
+      "loss": 1.6654,
+      "step": 97
+    },
+    {
+      "epoch": 0.07935222672064778,
+      "grad_norm": 1.4006737880954128,
+      "learning_rate": 1.9991490268511903e-05,
+      "loss": 1.6976,
+      "step": 98
+    },
+    {
+      "epoch": 0.08016194331983806,
+      "grad_norm": 1.5631879172620895,
+      "learning_rate": 1.9991132055867244e-05,
+      "loss": 1.669,
+      "step": 99
+    },
+    {
+      "epoch": 0.08097165991902834,
+      "grad_norm": 1.394392254430093,
+      "learning_rate": 1.9990766461807037e-05,
+      "loss": 1.6968,
+      "step": 100
+    },
+    {
+      "epoch": 0.08178137651821862,
+      "grad_norm": 1.472525619978714,
+      "learning_rate": 1.9990393486601385e-05,
+      "loss": 1.6773,
+      "step": 101
+    },
+    {
+      "epoch": 0.0825910931174089,
+      "grad_norm": 1.378206403293524,
+      "learning_rate": 1.9990013130525835e-05,
+      "loss": 1.6821,
+      "step": 102
+    },
+    {
+      "epoch": 0.08340080971659919,
+      "grad_norm": 1.328026656261711,
+      "learning_rate": 1.9989625393861397e-05,
+      "loss": 1.6103,
+      "step": 103
+    },
+    {
+      "epoch": 0.08421052631578947,
+      "grad_norm": 1.4064722782717616,
+      "learning_rate": 1.9989230276894525e-05,
+      "loss": 1.7324,
+      "step": 104
+    },
+    {
+      "epoch": 0.08502024291497975,
+      "grad_norm": 1.4813376774614593,
+      "learning_rate": 1.9988827779917138e-05,
+      "loss": 1.7272,
+      "step": 105
+    },
+    {
+      "epoch": 0.08582995951417004,
+      "grad_norm": 1.3687982499187121,
+      "learning_rate": 1.998841790322659e-05,
+      "loss": 1.673,
+      "step": 106
+    },
+    {
+      "epoch": 0.08663967611336032,
+      "grad_norm": 1.422876362819873,
+      "learning_rate": 1.9988000647125703e-05,
+      "loss": 1.7597,
+      "step": 107
+    },
+    {
+      "epoch": 0.0874493927125506,
+      "grad_norm": 1.308219863689626,
+      "learning_rate": 1.9987576011922743e-05,
+      "loss": 1.7697,
+      "step": 108
+    },
+    {
+      "epoch": 0.08825910931174089,
+      "grad_norm": 1.4183268142705985,
+      "learning_rate": 1.9987143997931428e-05,
+      "loss": 1.6809,
+      "step": 109
+    },
+    {
+      "epoch": 0.08906882591093117,
+      "grad_norm": 1.4357441361156031,
+      "learning_rate": 1.9986704605470932e-05,
+      "loss": 1.6358,
+      "step": 110
+    },
+    {
+      "epoch": 0.08987854251012145,
+      "grad_norm": 1.3529581082840894,
+      "learning_rate": 1.998625783486587e-05,
+      "loss": 1.6681,
+      "step": 111
+    },
+    {
+      "epoch": 0.09068825910931175,
+      "grad_norm": 1.408818420235949,
+      "learning_rate": 1.998580368644632e-05,
+      "loss": 1.6938,
+      "step": 112
+    },
+    {
+      "epoch": 0.09149797570850203,
+      "grad_norm": 1.4451180541868947,
+      "learning_rate": 1.998534216054781e-05,
+      "loss": 1.7739,
+      "step": 113
+    },
+    {
+      "epoch": 0.09230769230769231,
+      "grad_norm": 1.2941303668841717,
+      "learning_rate": 1.9984873257511296e-05,
+      "loss": 1.6704,
+      "step": 114
+    },
+    {
+      "epoch": 0.0931174089068826,
+      "grad_norm": 1.440123130144802,
+      "learning_rate": 1.9984396977683223e-05,
+      "loss": 1.7456,
+      "step": 115
+    },
+    {
+      "epoch": 0.09392712550607288,
+      "grad_norm": 1.3845008076241843,
+      "learning_rate": 1.998391332141545e-05,
+      "loss": 1.6968,
+      "step": 116
+    },
+    {
+      "epoch": 0.09473684210526316,
+      "grad_norm": 1.5220029955302263,
+      "learning_rate": 1.998342228906531e-05,
+      "loss": 1.8585,
+      "step": 117
+    },
+    {
+      "epoch": 0.09554655870445344,
+      "grad_norm": 1.350510831118027,
+      "learning_rate": 1.998292388099557e-05,
+      "loss": 1.7009,
+      "step": 118
+    },
+    {
+      "epoch": 0.09635627530364373,
+      "grad_norm": 1.410947289102555,
+      "learning_rate": 1.9982418097574458e-05,
+      "loss": 1.7082,
+      "step": 119
+    },
+    {
+      "epoch": 0.09716599190283401,
+      "grad_norm": 1.4674312785560688,
+      "learning_rate": 1.998190493917564e-05,
+      "loss": 1.6933,
+      "step": 120
+    },
+    {
+      "epoch": 0.09797570850202429,
+      "grad_norm": 1.3586528211235367,
+      "learning_rate": 1.9981384406178235e-05,
+      "loss": 1.697,
+      "step": 121
+    },
+    {
+      "epoch": 0.09878542510121457,
+      "grad_norm": 1.323874170111494,
+      "learning_rate": 1.998085649896682e-05,
+      "loss": 1.6996,
+      "step": 122
+    },
+    {
+      "epoch": 0.09959514170040486,
+      "grad_norm": 1.3857202766997219,
+      "learning_rate": 1.99803212179314e-05,
+      "loss": 1.6158,
+      "step": 123
+    },
+    {
+      "epoch": 0.10040485829959514,
+      "grad_norm": 1.3135040066156267,
+      "learning_rate": 1.9979778563467446e-05,
+      "loss": 1.6917,
+      "step": 124
+    },
+    {
+      "epoch": 0.10121457489878542,
+      "grad_norm": 1.450855282133572,
+      "learning_rate": 1.9979228535975866e-05,
+      "loss": 1.6773,
+      "step": 125
+    },
+    {
+      "epoch": 0.1020242914979757,
+      "grad_norm": 1.347188642125951,
+      "learning_rate": 1.997867113586302e-05,
+      "loss": 1.7144,
+      "step": 126
+    },
+    {
+      "epoch": 0.10283400809716599,
+      "grad_norm": 1.345981107132026,
+      "learning_rate": 1.997810636354071e-05,
+      "loss": 1.6041,
+      "step": 127
+    },
+    {
+      "epoch": 0.10364372469635627,
+      "grad_norm": 1.2972173884632567,
+      "learning_rate": 1.9977534219426195e-05,
+      "loss": 1.6855,
+      "step": 128
+    },
+    {
+      "epoch": 0.10445344129554655,
+      "grad_norm": 1.275527287436048,
+      "learning_rate": 1.997695470394217e-05,
+      "loss": 1.696,
+      "step": 129
+    },
+    {
+      "epoch": 0.10526315789473684,
+      "grad_norm": 1.3215678390799637,
+      "learning_rate": 1.9976367817516773e-05,
+      "loss": 1.6935,
+      "step": 130
+    },
+    {
+      "epoch": 0.10607287449392712,
+      "grad_norm": 1.3828255635499427,
+      "learning_rate": 1.99757735605836e-05,
+      "loss": 1.7256,
+      "step": 131
+    },
+    {
+      "epoch": 0.10688259109311742,
+      "grad_norm": 1.448631635568024,
+      "learning_rate": 1.997517193358169e-05,
+      "loss": 1.6388,
+      "step": 132
+    },
+    {
+      "epoch": 0.1076923076923077,
+      "grad_norm": 1.3338631760173656,
+      "learning_rate": 1.9974562936955513e-05,
+      "loss": 1.6965,
+      "step": 133
+    },
+    {
+      "epoch": 0.10850202429149798,
+      "grad_norm": 1.3424092875489613,
+      "learning_rate": 1.9973946571155e-05,
+      "loss": 1.709,
+      "step": 134
+    },
+    {
+      "epoch": 0.10931174089068826,
+      "grad_norm": 1.3898696103299275,
+      "learning_rate": 1.9973322836635517e-05,
+      "loss": 1.6822,
+      "step": 135
+    },
+    {
+      "epoch": 0.11012145748987855,
+      "grad_norm": 1.4706810809601643,
+      "learning_rate": 1.997269173385788e-05,
+      "loss": 1.7058,
+      "step": 136
+    },
+    {
+      "epoch": 0.11093117408906883,
+      "grad_norm": 1.4026119282799607,
+      "learning_rate": 1.9972053263288346e-05,
+      "loss": 1.6531,
+      "step": 137
+    },
+    {
+      "epoch": 0.11174089068825911,
+      "grad_norm": 1.3288506201522425,
+      "learning_rate": 1.9971407425398614e-05,
+      "loss": 1.6532,
+      "step": 138
+    },
+    {
+      "epoch": 0.1125506072874494,
+      "grad_norm": 1.5239912432928768,
+      "learning_rate": 1.9970754220665824e-05,
+      "loss": 1.7604,
+      "step": 139
+    },
+    {
+      "epoch": 0.11336032388663968,
+      "grad_norm": 1.444106448065426,
+      "learning_rate": 1.9970093649572567e-05,
+      "loss": 1.7596,
+      "step": 140
+    },
+    {
+      "epoch": 0.11417004048582996,
+      "grad_norm": 1.3200905343039002,
+      "learning_rate": 1.9969425712606864e-05,
+      "loss": 1.6867,
+      "step": 141
+    },
+    {
+      "epoch": 0.11497975708502024,
+      "grad_norm": 1.3681666300591042,
+      "learning_rate": 1.996875041026219e-05,
+      "loss": 1.6352,
+      "step": 142
+    },
+    {
+      "epoch": 0.11578947368421053,
+      "grad_norm": 1.4975233179603091,
+      "learning_rate": 1.9968067743037453e-05,
+      "loss": 1.679,
+      "step": 143
+    },
+    {
+      "epoch": 0.11659919028340081,
+      "grad_norm": 1.3172078248631487,
+      "learning_rate": 1.9967377711437008e-05,
+      "loss": 1.7421,
+      "step": 144
+    },
+    {
+      "epoch": 0.11740890688259109,
+      "grad_norm": 1.260492754026676,
+      "learning_rate": 1.9966680315970647e-05,
+      "loss": 1.5965,
+      "step": 145
+    },
+    {
+      "epoch": 0.11821862348178137,
+      "grad_norm": 1.3786263323616315,
+      "learning_rate": 1.9965975557153604e-05,
+      "loss": 1.7367,
+      "step": 146
+    },
+    {
+      "epoch": 0.11902834008097166,
+      "grad_norm": 1.3104589610244388,
+      "learning_rate": 1.996526343550655e-05,
+      "loss": 1.6468,
+      "step": 147
+    },
+    {
+      "epoch": 0.11983805668016194,
+      "grad_norm": 1.3212338618017334,
+      "learning_rate": 1.99645439515556e-05,
+      "loss": 1.6932,
+      "step": 148
+    },
+    {
+      "epoch": 0.12064777327935222,
+      "grad_norm": 1.2823071957240442,
+      "learning_rate": 1.9963817105832305e-05,
+      "loss": 1.6809,
+      "step": 149
+    },
+    {
+      "epoch": 0.1214574898785425,
+      "grad_norm": 1.2331269340221573,
+      "learning_rate": 1.996308289887366e-05,
+      "loss": 1.7086,
+      "step": 150
+    },
+    {
+      "epoch": 0.12226720647773279,
+      "grad_norm": 1.3183668422073427,
+      "learning_rate": 1.9962341331222092e-05,
+      "loss": 1.6732,
+      "step": 151
+    },
+    {
+      "epoch": 0.12307692307692308,
+      "grad_norm": 1.3211093232564248,
+      "learning_rate": 1.996159240342547e-05,
+      "loss": 1.6752,
+      "step": 152
+    },
+    {
+      "epoch": 0.12388663967611337,
+      "grad_norm": 1.3807756953246901,
+      "learning_rate": 1.9960836116037095e-05,
+      "loss": 1.6819,
+      "step": 153
+    },
+    {
+      "epoch": 0.12469635627530365,
+      "grad_norm": 1.229226565824642,
+      "learning_rate": 1.9960072469615716e-05,
+      "loss": 1.7211,
+      "step": 154
+    },
+    {
+      "epoch": 0.12550607287449392,
+      "grad_norm": 1.37757201668452,
+      "learning_rate": 1.9959301464725507e-05,
+      "loss": 1.7252,
+      "step": 155
+    },
+    {
+      "epoch": 0.12631578947368421,
+      "grad_norm": 1.3625152047582556,
+      "learning_rate": 1.9958523101936083e-05,
+      "loss": 1.7572,
+      "step": 156
+    },
+    {
+      "epoch": 0.12712550607287448,
+      "grad_norm": 1.3700626216344285,
+      "learning_rate": 1.9957737381822505e-05,
+      "loss": 1.7384,
+      "step": 157
+    },
+    {
+      "epoch": 0.12793522267206478,
+      "grad_norm": 1.3265891040200295,
+      "learning_rate": 1.9956944304965257e-05,
+      "loss": 1.7058,
+      "step": 158
+    },
+    {
+      "epoch": 0.12874493927125505,
+      "grad_norm": 1.477875528364369,
+      "learning_rate": 1.9956143871950252e-05,
+      "loss": 1.6846,
+      "step": 159
+    },
+    {
+      "epoch": 0.12955465587044535,
+      "grad_norm": 1.3079566287214603,
+      "learning_rate": 1.995533608336886e-05,
+      "loss": 1.7252,
+      "step": 160
+    },
+    {
+      "epoch": 0.13036437246963561,
+      "grad_norm": 1.422258383650131,
+      "learning_rate": 1.9954520939817863e-05,
+      "loss": 1.7018,
+      "step": 161
+    },
+    {
+      "epoch": 0.1311740890688259,
+      "grad_norm": 1.4264781196637615,
+      "learning_rate": 1.9953698441899494e-05,
+      "loss": 1.6585,
+      "step": 162
+    },
+    {
+      "epoch": 0.1319838056680162,
+      "grad_norm": 1.2963346637584106,
+      "learning_rate": 1.9952868590221403e-05,
+      "loss": 1.6369,
+      "step": 163
+    },
+    {
+      "epoch": 0.13279352226720648,
+      "grad_norm": 1.5083900065401323,
+      "learning_rate": 1.9952031385396694e-05,
+      "loss": 1.7224,
+      "step": 164
+    },
+    {
+      "epoch": 0.13360323886639677,
+      "grad_norm": 1.5179214854876595,
+      "learning_rate": 1.995118682804388e-05,
+      "loss": 1.7588,
+      "step": 165
+    },
+    {
+      "epoch": 0.13441295546558704,
+      "grad_norm": 1.3017296176425739,
+      "learning_rate": 1.995033491878692e-05,
+      "loss": 1.7246,
+      "step": 166
+    },
+    {
+      "epoch": 0.13522267206477734,
+      "grad_norm": 1.505962687706292,
+      "learning_rate": 1.9949475658255207e-05,
+      "loss": 1.7567,
+      "step": 167
+    },
+    {
+      "epoch": 0.1360323886639676,
+      "grad_norm": 1.2555297272542156,
+      "learning_rate": 1.994860904708355e-05,
+      "loss": 1.7062,
+      "step": 168
+    },
+    {
+      "epoch": 0.1368421052631579,
+      "grad_norm": 1.485635141654141,
+      "learning_rate": 1.994773508591221e-05,
+      "loss": 1.6812,
+      "step": 169
+    },
+    {
+      "epoch": 0.13765182186234817,
+      "grad_norm": 1.2908226955038942,
+      "learning_rate": 1.9946853775386857e-05,
+      "loss": 1.6867,
+      "step": 170
+    },
+    {
+      "epoch": 0.13846153846153847,
+      "grad_norm": 1.32792940285529,
+      "learning_rate": 1.9945965116158605e-05,
+      "loss": 1.6529,
+      "step": 171
+    },
+    {
+      "epoch": 0.13927125506072874,
+      "grad_norm": 1.2871011365969764,
+      "learning_rate": 1.9945069108883993e-05,
+      "loss": 1.6857,
+      "step": 172
+    },
+    {
+      "epoch": 0.14008097165991903,
+      "grad_norm": 1.3266369892431742,
+      "learning_rate": 1.994416575422499e-05,
+      "loss": 1.7121,
+      "step": 173
+    },
+    {
+      "epoch": 0.1408906882591093,
+      "grad_norm": 1.3444583702433448,
+      "learning_rate": 1.9943255052848984e-05,
+      "loss": 1.6784,
+      "step": 174
+    },
+    {
+      "epoch": 0.1417004048582996,
+      "grad_norm": 1.2393516324343374,
+      "learning_rate": 1.9942337005428805e-05,
+      "loss": 1.6686,
+      "step": 175
+    },
+    {
+      "epoch": 0.14251012145748987,
+      "grad_norm": 1.3248524170639142,
+      "learning_rate": 1.99414116126427e-05,
+      "loss": 1.6597,
+      "step": 176
+    },
+    {
+      "epoch": 0.14331983805668017,
+      "grad_norm": 1.1663228528592244,
+      "learning_rate": 1.9940478875174346e-05,
+      "loss": 1.5819,
+      "step": 177
+    },
+    {
+      "epoch": 0.14412955465587043,
+      "grad_norm": 1.3787198785492285,
+      "learning_rate": 1.9939538793712852e-05,
+      "loss": 1.6482,
+      "step": 178
+    },
+    {
+      "epoch": 0.14493927125506073,
+      "grad_norm": 1.350980964072927,
+      "learning_rate": 1.993859136895274e-05,
+      "loss": 1.6578,
+      "step": 179
+    },
+    {
+      "epoch": 0.145748987854251,
+      "grad_norm": 1.4172264638957366,
+      "learning_rate": 1.9937636601593965e-05,
+      "loss": 1.7364,
+      "step": 180
+    },
+    {
+      "epoch": 0.1465587044534413,
+      "grad_norm": 1.2164996262071472,
+      "learning_rate": 1.9936674492341913e-05,
+      "loss": 1.6566,
+      "step": 181
+    },
+    {
+      "epoch": 0.14736842105263157,
+      "grad_norm": 1.163342097978053,
+      "learning_rate": 1.9935705041907375e-05,
+      "loss": 1.6742,
+      "step": 182
+    },
+    {
+      "epoch": 0.14817813765182186,
+      "grad_norm": 1.3251945549485633,
+      "learning_rate": 1.9934728251006593e-05,
+      "loss": 1.6875,
+      "step": 183
+    },
+    {
+      "epoch": 0.14898785425101216,
+      "grad_norm": 1.2641308251973702,
+      "learning_rate": 1.9933744120361202e-05,
+      "loss": 1.7407,
+      "step": 184
+    },
+    {
+      "epoch": 0.14979757085020243,
+      "grad_norm": 1.3486708746298524,
+      "learning_rate": 1.9932752650698285e-05,
+      "loss": 1.6667,
+      "step": 185
+    },
+    {
+      "epoch": 0.15060728744939272,
+      "grad_norm": 1.2926218776022609,
+      "learning_rate": 1.993175384275033e-05,
+      "loss": 1.6742,
+      "step": 186
+    },
+    {
+      "epoch": 0.151417004048583,
+      "grad_norm": 1.367440454792052,
+      "learning_rate": 1.9930747697255263e-05,
+      "loss": 1.6944,
+      "step": 187
+    },
+    {
+      "epoch": 0.1522267206477733,
+      "grad_norm": 1.3527036774330383,
+      "learning_rate": 1.992973421495641e-05,
+      "loss": 1.626,
+      "step": 188
+    },
+    {
+      "epoch": 0.15303643724696356,
+      "grad_norm": 1.287788530805172,
+      "learning_rate": 1.992871339660253e-05,
+      "loss": 1.6921,
+      "step": 189
+    },
+    {
+      "epoch": 0.15384615384615385,
+      "grad_norm": 1.4084861938787652,
+      "learning_rate": 1.9927685242947804e-05,
+      "loss": 1.7108,
+      "step": 190
+    },
+    {
+      "epoch": 0.15465587044534412,
+      "grad_norm": 1.279148085366124,
+      "learning_rate": 1.9926649754751825e-05,
+      "loss": 1.6675,
+      "step": 191
+    },
+    {
+      "epoch": 0.15546558704453442,
+      "grad_norm": 1.2664930272396757,
+      "learning_rate": 1.9925606932779615e-05,
+      "loss": 1.7115,
+      "step": 192
+    },
+    {
+      "epoch": 0.1562753036437247,
+      "grad_norm": 1.321724868931457,
+      "learning_rate": 1.99245567778016e-05,
+      "loss": 1.6596,
+      "step": 193
+    },
+    {
+      "epoch": 0.15708502024291499,
+      "grad_norm": 1.2948713460179608,
+      "learning_rate": 1.9923499290593637e-05,
+      "loss": 1.6664,
+      "step": 194
+    },
+    {
+      "epoch": 0.15789473684210525,
+      "grad_norm": 1.3818189742460931,
+      "learning_rate": 1.9922434471936987e-05,
+      "loss": 1.7305,
+      "step": 195
+    },
+    {
+      "epoch": 0.15870445344129555,
+      "grad_norm": 1.2479572194235047,
+      "learning_rate": 1.9921362322618337e-05,
+      "loss": 1.7553,
+      "step": 196
+    },
+    {
+      "epoch": 0.15951417004048582,
+      "grad_norm": 1.2898516448229447,
+      "learning_rate": 1.9920282843429795e-05,
+      "loss": 1.65,
+      "step": 197
+    },
+    {
+      "epoch": 0.16032388663967612,
+      "grad_norm": 1.344427118443871,
+      "learning_rate": 1.9919196035168865e-05,
+      "loss": 1.6366,
+      "step": 198
+    },
+    {
+      "epoch": 0.16113360323886639,
+      "grad_norm": 1.2531134681362286,
+      "learning_rate": 1.9918101898638488e-05,
+      "loss": 1.6608,
+      "step": 199
+    },
+    {
+      "epoch": 0.16194331983805668,
+      "grad_norm": 1.5273799001568773,
+      "learning_rate": 1.9917000434647e-05,
+      "loss": 1.6969,
+      "step": 200
+    },
+    {
+      "epoch": 0.16275303643724695,
+      "grad_norm": 1.4281468350234128,
+      "learning_rate": 1.9915891644008164e-05,
+      "loss": 1.6933,
+      "step": 201
+    },
+    {
+      "epoch": 0.16356275303643725,
+      "grad_norm": 1.2479553523055322,
+      "learning_rate": 1.991477552754115e-05,
+      "loss": 1.633,
+      "step": 202
+    },
+    {
+      "epoch": 0.16437246963562754,
+      "grad_norm": 1.3889081568820743,
+      "learning_rate": 1.9913652086070535e-05,
+      "loss": 1.6847,
+      "step": 203
+    },
+    {
+      "epoch": 0.1651821862348178,
+      "grad_norm": 1.3301717360886383,
+      "learning_rate": 1.9912521320426327e-05,
+      "loss": 1.6712,
+      "step": 204
+    },
+    {
+      "epoch": 0.1659919028340081,
+      "grad_norm": 1.2916336401520587,
+      "learning_rate": 1.991138323144392e-05,
+      "loss": 1.6672,
+      "step": 205
+    },
+    {
+      "epoch": 0.16680161943319838,
+      "grad_norm": 1.2694035888961783,
+      "learning_rate": 1.9910237819964135e-05,
+      "loss": 1.6759,
+      "step": 206
+    },
+    {
+      "epoch": 0.16761133603238867,
+      "grad_norm": 1.2658690812785272,
+      "learning_rate": 1.9909085086833198e-05,
+      "loss": 1.6743,
+      "step": 207
+    },
+    {
+      "epoch": 0.16842105263157894,
+      "grad_norm": 1.2672804576639893,
+      "learning_rate": 1.9907925032902745e-05,
+      "loss": 1.664,
+      "step": 208
+    },
+    {
+      "epoch": 0.16923076923076924,
+      "grad_norm": 1.1845691219627643,
+      "learning_rate": 1.9906757659029817e-05,
+      "loss": 1.6352,
+      "step": 209
+    },
+    {
+      "epoch": 0.1700404858299595,
+      "grad_norm": 1.260533932342834,
+      "learning_rate": 1.990558296607687e-05,
+      "loss": 1.6717,
+      "step": 210
+    },
+    {
+      "epoch": 0.1708502024291498,
+      "grad_norm": 1.1724140852250677,
+      "learning_rate": 1.9904400954911763e-05,
+      "loss": 1.5936,
+      "step": 211
+    },
+    {
+      "epoch": 0.17165991902834007,
+      "grad_norm": 1.3376574942903026,
+      "learning_rate": 1.990321162640776e-05,
+      "loss": 1.6479,
+      "step": 212
+    },
+    {
+      "epoch": 0.17246963562753037,
+      "grad_norm": 1.3302629754630635,
+      "learning_rate": 1.9902014981443532e-05,
+      "loss": 1.6688,
+      "step": 213
+    },
+    {
+      "epoch": 0.17327935222672064,
+      "grad_norm": 1.2504083776545458,
+      "learning_rate": 1.9900811020903158e-05,
+      "loss": 1.6221,
+      "step": 214
+    },
+    {
+      "epoch": 0.17408906882591094,
+      "grad_norm": 1.279410572569658,
+      "learning_rate": 1.9899599745676123e-05,
+      "loss": 1.7135,
+      "step": 215
+    },
+    {
+      "epoch": 0.1748987854251012,
+      "grad_norm": 1.3819762266443132,
+      "learning_rate": 1.989838115665731e-05,
+      "loss": 1.6995,
+      "step": 216
+    },
+    {
+      "epoch": 0.1757085020242915,
+      "grad_norm": 1.260595817978411,
+      "learning_rate": 1.9897155254747006e-05,
+      "loss": 1.6651,
+      "step": 217
+    },
+    {
+      "epoch": 0.17651821862348177,
+      "grad_norm": 1.3746335152221298,
+      "learning_rate": 1.989592204085091e-05,
+      "loss": 1.732,
+      "step": 218
+    },
+    {
+      "epoch": 0.17732793522267207,
+      "grad_norm": 1.2946738994826934,
+      "learning_rate": 1.9894681515880106e-05,
+      "loss": 1.6498,
+      "step": 219
+    },
+    {
+      "epoch": 0.17813765182186234,
+      "grad_norm": 1.3214716293210784,
+      "learning_rate": 1.9893433680751105e-05,
+      "loss": 1.6552,
+      "step": 220
+    },
+    {
+      "epoch": 0.17894736842105263,
+      "grad_norm": 1.267390260422078,
+      "learning_rate": 1.9892178536385788e-05,
+      "loss": 1.6601,
+      "step": 221
+    },
+    {
+      "epoch": 0.1797570850202429,
+      "grad_norm": 1.2661450912214733,
+      "learning_rate": 1.9890916083711463e-05,
+      "loss": 1.6884,
+      "step": 222
+    },
+    {
+      "epoch": 0.1805668016194332,
+      "grad_norm": 1.2915140670465655,
+      "learning_rate": 1.9889646323660816e-05,
+      "loss": 1.7442,
+      "step": 223
+    },
+    {
+      "epoch": 0.1813765182186235,
+      "grad_norm": 1.2082943407898998,
+      "learning_rate": 1.9888369257171952e-05,
+      "loss": 1.6767,
+      "step": 224
+    },
+    {
+      "epoch": 0.18218623481781376,
+      "grad_norm": 1.3731016770046895,
+      "learning_rate": 1.9887084885188354e-05,
+      "loss": 1.7334,
+      "step": 225
+    },
+    {
+      "epoch": 0.18299595141700406,
+      "grad_norm": 1.2555298583209036,
+      "learning_rate": 1.988579320865892e-05,
+      "loss": 1.7203,
+      "step": 226
+    },
+    {
+      "epoch": 0.18380566801619433,
+      "grad_norm": 1.263140649035035,
+      "learning_rate": 1.988449422853793e-05,
+      "loss": 1.6895,
+      "step": 227
+    },
+    {
+      "epoch": 0.18461538461538463,
+      "grad_norm": 1.2749775919052824,
+      "learning_rate": 1.9883187945785067e-05,
+      "loss": 1.6948,
+      "step": 228
+    },
+    {
+      "epoch": 0.1854251012145749,
+      "grad_norm": 1.2429895536427575,
+      "learning_rate": 1.9881874361365413e-05,
+      "loss": 1.6361,
+      "step": 229
+    },
+    {
+      "epoch": 0.1862348178137652,
+      "grad_norm": 1.2416951269903427,
+      "learning_rate": 1.9880553476249437e-05,
+      "loss": 1.6758,
+      "step": 230
+    },
+    {
+      "epoch": 0.18704453441295546,
+      "grad_norm": 1.239555283382809,
+      "learning_rate": 1.9879225291413e-05,
+      "loss": 1.6989,
+      "step": 231
+    },
+    {
+      "epoch": 0.18785425101214576,
+      "grad_norm": 1.2919283930709724,
+      "learning_rate": 1.9877889807837373e-05,
+      "loss": 1.6731,
+      "step": 232
+    },
+    {
+      "epoch": 0.18866396761133603,
+      "grad_norm": 1.2233652648633166,
+      "learning_rate": 1.9876547026509194e-05,
+      "loss": 1.6967,
+      "step": 233
+    },
+    {
+      "epoch": 0.18947368421052632,
+      "grad_norm": 1.2745651498433337,
+      "learning_rate": 1.987519694842051e-05,
+      "loss": 1.7336,
+      "step": 234
+    },
+    {
+      "epoch": 0.1902834008097166,
+      "grad_norm": 1.2732698824293152,
+      "learning_rate": 1.9873839574568756e-05,
+      "loss": 1.6757,
+      "step": 235
+    },
+    {
+      "epoch": 0.1910931174089069,
+      "grad_norm": 1.3370656954094273,
+      "learning_rate": 1.9872474905956752e-05,
+      "loss": 1.7574,
+      "step": 236
+    },
+    {
+      "epoch": 0.19190283400809716,
+      "grad_norm": 1.337679917073954,
+      "learning_rate": 1.9871102943592717e-05,
+      "loss": 1.6757,
+      "step": 237
+    },
+    {
+      "epoch": 0.19271255060728745,
+      "grad_norm": 1.3246803706512758,
+      "learning_rate": 1.9869723688490247e-05,
+      "loss": 1.7,
+      "step": 238
+    },
+    {
+      "epoch": 0.19352226720647772,
+      "grad_norm": 1.2000006625597532,
+      "learning_rate": 1.9868337141668333e-05,
+      "loss": 1.6744,
+      "step": 239
+    },
+    {
+      "epoch": 0.19433198380566802,
+      "grad_norm": 1.3328073640320248,
+      "learning_rate": 1.9866943304151346e-05,
+      "loss": 1.6301,
+      "step": 240
+    },
+    {
+      "epoch": 0.1951417004048583,
+      "grad_norm": 1.2595984845371397,
+      "learning_rate": 1.9865542176969055e-05,
+      "loss": 1.7314,
+      "step": 241
+    },
+    {
+      "epoch": 0.19595141700404858,
+      "grad_norm": 1.2697481850011105,
+      "learning_rate": 1.986413376115661e-05,
+      "loss": 1.7403,
+      "step": 242
+    },
+    {
+      "epoch": 0.19676113360323888,
+      "grad_norm": 1.184815618587701,
+      "learning_rate": 1.9862718057754536e-05,
+      "loss": 1.6913,
+      "step": 243
+    },
+    {
+      "epoch": 0.19757085020242915,
+      "grad_norm": 1.2575508256727617,
+      "learning_rate": 1.9861295067808754e-05,
+      "loss": 1.6972,
+      "step": 244
+    },
+    {
+      "epoch": 0.19838056680161945,
+      "grad_norm": 1.2037684360733698,
+      "learning_rate": 1.9859864792370565e-05,
+      "loss": 1.6808,
+      "step": 245
+    },
+    {
+      "epoch": 0.19919028340080971,
+      "grad_norm": 1.1984658119885334,
+      "learning_rate": 1.985842723249665e-05,
+      "loss": 1.6486,
+      "step": 246
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.3302394542035112,
+      "learning_rate": 1.985698238924908e-05,
+      "loss": 1.6491,
+      "step": 247
+    },
+    {
+      "epoch": 0.20080971659919028,
+      "grad_norm": 1.2819039031405692,
+      "learning_rate": 1.9855530263695287e-05,
+      "loss": 1.7145,
+      "step": 248
+    },
+    {
+      "epoch": 0.20161943319838058,
+      "grad_norm": 1.2645396518124346,
+      "learning_rate": 1.9854070856908113e-05,
+      "loss": 1.6766,
+      "step": 249
+    },
+    {
+      "epoch": 0.20242914979757085,
+      "grad_norm": 1.3551588648492328,
+      "learning_rate": 1.985260416996575e-05,
+      "loss": 1.7532,
+      "step": 250
+    },
+    {
+      "epoch": 0.20323886639676114,
+      "grad_norm": 1.1872824323555884,
+      "learning_rate": 1.9851130203951787e-05,
+      "loss": 1.722,
+      "step": 251
+    },
+    {
+      "epoch": 0.2040485829959514,
+      "grad_norm": 1.3233391136396748,
+      "learning_rate": 1.9849648959955187e-05,
+      "loss": 1.6066,
+      "step": 252
+    },
+    {
+      "epoch": 0.2048582995951417,
+      "grad_norm": 1.3232293456672852,
+      "learning_rate": 1.9848160439070284e-05,
+      "loss": 1.6556,
+      "step": 253
+    },
+    {
+      "epoch": 0.20566801619433198,
+      "grad_norm": 1.3252440210019765,
+      "learning_rate": 1.9846664642396793e-05,
+      "loss": 1.6771,
+      "step": 254
+    },
+    {
+      "epoch": 0.20647773279352227,
+      "grad_norm": 1.2091779306727122,
+      "learning_rate": 1.9845161571039805e-05,
+      "loss": 1.7038,
+      "step": 255
+    },
+    {
+      "epoch": 0.20728744939271254,
+      "grad_norm": 1.2887939358404155,
+      "learning_rate": 1.9843651226109784e-05,
+      "loss": 1.6948,
+      "step": 256
+    },
+    {
+      "epoch": 0.20809716599190284,
+      "grad_norm": 1.2583822543738294,
+      "learning_rate": 1.984213360872257e-05,
+      "loss": 1.6888,
+      "step": 257
+    },
+    {
+      "epoch": 0.2089068825910931,
+      "grad_norm": 1.3279683247483347,
+      "learning_rate": 1.9840608719999367e-05,
+      "loss": 1.7059,
+      "step": 258
+    },
+    {
+      "epoch": 0.2097165991902834,
+      "grad_norm": 1.270081125900351,
+      "learning_rate": 1.9839076561066766e-05,
+      "loss": 1.7175,
+      "step": 259
+    },
+    {
+      "epoch": 0.21052631578947367,
+      "grad_norm": 1.254273959928416,
+      "learning_rate": 1.983753713305672e-05,
+      "loss": 1.6763,
+      "step": 260
+    },
+    {
+      "epoch": 0.21133603238866397,
+      "grad_norm": 1.3521388079612242,
+      "learning_rate": 1.9835990437106542e-05,
+      "loss": 1.6948,
+      "step": 261
+    },
+    {
+      "epoch": 0.21214574898785424,
+      "grad_norm": 1.267536006413824,
+      "learning_rate": 1.983443647435894e-05,
+      "loss": 1.675,
+      "step": 262
+    },
+    {
+      "epoch": 0.21295546558704453,
+      "grad_norm": 1.3388930511150345,
+      "learning_rate": 1.9832875245961972e-05,
+      "loss": 1.7291,
+      "step": 263
+    },
+    {
+      "epoch": 0.21376518218623483,
+      "grad_norm": 1.2427093269086842,
+      "learning_rate": 1.9831306753069066e-05,
+      "loss": 1.6094,
+      "step": 264
+    },
+    {
+      "epoch": 0.2145748987854251,
+      "grad_norm": 1.349674351751306,
+      "learning_rate": 1.982973099683902e-05,
+      "loss": 1.6964,
+      "step": 265
+    },
+    {
+      "epoch": 0.2153846153846154,
+      "grad_norm": 1.365432576876562,
+      "learning_rate": 1.9828147978436e-05,
+      "loss": 1.7176,
+      "step": 266
+    },
+    {
+      "epoch": 0.21619433198380567,
+      "grad_norm": 1.2307745643880328,
+      "learning_rate": 1.982655769902953e-05,
+      "loss": 1.6596,
+      "step": 267
+    },
+    {
+      "epoch": 0.21700404858299596,
+      "grad_norm": 1.2552754011065252,
+      "learning_rate": 1.9824960159794512e-05,
+      "loss": 1.7251,
+      "step": 268
+    },
+    {
+      "epoch": 0.21781376518218623,
+      "grad_norm": 1.2821452869506051,
+      "learning_rate": 1.9823355361911192e-05,
+      "loss": 1.6452,
+      "step": 269
+    },
+    {
+      "epoch": 0.21862348178137653,
+      "grad_norm": 1.2062731542559535,
+      "learning_rate": 1.98217433065652e-05,
+      "loss": 1.6814,
+      "step": 270
+    },
+    {
+      "epoch": 0.2194331983805668,
+      "grad_norm": 1.2262168648135,
+      "learning_rate": 1.9820123994947505e-05,
+      "loss": 1.6816,
+      "step": 271
+    },
+    {
+      "epoch": 0.2202429149797571,
+      "grad_norm": 1.2938841697813679,
+      "learning_rate": 1.981849742825446e-05,
+      "loss": 1.7483,
+      "step": 272
+    },
+    {
+      "epoch": 0.22105263157894736,
+      "grad_norm": 1.2645687786911326,
+      "learning_rate": 1.981686360768776e-05,
+      "loss": 1.6805,
+      "step": 273
+    },
+    {
+      "epoch": 0.22186234817813766,
+      "grad_norm": 1.222316415160104,
+      "learning_rate": 1.9815222534454472e-05,
+      "loss": 1.7249,
+      "step": 274
+    },
+    {
+      "epoch": 0.22267206477732793,
+      "grad_norm": 1.2204096535680333,
+      "learning_rate": 1.9813574209767013e-05,
+      "loss": 1.6393,
+      "step": 275
+    },
+    {
+      "epoch": 0.22348178137651822,
+      "grad_norm": 1.2272417519081043,
+      "learning_rate": 1.981191863484316e-05,
+      "loss": 1.665,
+      "step": 276
+    },
+    {
+      "epoch": 0.2242914979757085,
+      "grad_norm": 1.2688048841631627,
+      "learning_rate": 1.9810255810906046e-05,
+      "loss": 1.6725,
+      "step": 277
+    },
+    {
+      "epoch": 0.2251012145748988,
+      "grad_norm": 1.1858272375472187,
+      "learning_rate": 1.9808585739184156e-05,
+      "loss": 1.6861,
+      "step": 278
+    },
+    {
+      "epoch": 0.22591093117408906,
+      "grad_norm": 1.2436292029105176,
+      "learning_rate": 1.980690842091134e-05,
+      "loss": 1.679,
+      "step": 279
+    },
+    {
+      "epoch": 0.22672064777327935,
+      "grad_norm": 1.2348131442739951,
+      "learning_rate": 1.9805223857326794e-05,
+      "loss": 1.6451,
+      "step": 280
+    },
+    {
+      "epoch": 0.22753036437246962,
+      "grad_norm": 1.195784828107809,
+      "learning_rate": 1.9803532049675062e-05,
+      "loss": 1.637,
+      "step": 281
+    },
+    {
+      "epoch": 0.22834008097165992,
+      "grad_norm": 1.2391790592489964,
+      "learning_rate": 1.9801832999206057e-05,
+      "loss": 1.7539,
+      "step": 282
+    },
+    {
+      "epoch": 0.2291497975708502,
+      "grad_norm": 1.1984543549350761,
+      "learning_rate": 1.980012670717502e-05,
+      "loss": 1.6307,
+      "step": 283
+    },
+    {
+      "epoch": 0.22995951417004049,
+      "grad_norm": 1.2082020794861243,
+      "learning_rate": 1.9798413174842565e-05,
+      "loss": 1.7061,
+      "step": 284
+    },
+    {
+      "epoch": 0.23076923076923078,
+      "grad_norm": 1.17049178703938,
+      "learning_rate": 1.9796692403474632e-05,
+      "loss": 1.6614,
+      "step": 285
+    },
+    {
+      "epoch": 0.23157894736842105,
+      "grad_norm": 1.1776506077280637,
+      "learning_rate": 1.9794964394342532e-05,
+      "loss": 1.7012,
+      "step": 286
+    },
+    {
+      "epoch": 0.23238866396761135,
+      "grad_norm": 1.250032564625677,
+      "learning_rate": 1.9793229148722907e-05,
+      "loss": 1.7472,
+      "step": 287
+    },
+    {
+      "epoch": 0.23319838056680162,
+      "grad_norm": 1.1923743701902663,
+      "learning_rate": 1.979148666789775e-05,
+      "loss": 1.6169,
+      "step": 288
+    },
+    {
+      "epoch": 0.2340080971659919,
+      "grad_norm": 1.2445641617846566,
+      "learning_rate": 1.9789736953154405e-05,
+      "loss": 1.6697,
+      "step": 289
+    },
+    {
+      "epoch": 0.23481781376518218,
+      "grad_norm": 1.2338781631842157,
+      "learning_rate": 1.9787980005785553e-05,
+      "loss": 1.6691,
+      "step": 290
+    },
+    {
+      "epoch": 0.23562753036437248,
+      "grad_norm": 1.263470721107049,
+      "learning_rate": 1.9786215827089216e-05,
+      "loss": 1.6716,
+      "step": 291
+    },
+    {
+      "epoch": 0.23643724696356275,
+      "grad_norm": 1.2561973670317472,
+      "learning_rate": 1.978444441836877e-05,
+      "loss": 1.6759,
+      "step": 292
+    },
+    {
+      "epoch": 0.23724696356275304,
+      "grad_norm": 1.2208658433077548,
+      "learning_rate": 1.9782665780932926e-05,
+      "loss": 1.6881,
+      "step": 293
+    },
+    {
+      "epoch": 0.2380566801619433,
+      "grad_norm": 1.2129567864078932,
+      "learning_rate": 1.9780879916095733e-05,
+      "loss": 1.6417,
+      "step": 294
+    },
+    {
+      "epoch": 0.2388663967611336,
+      "grad_norm": 1.260630886604336,
+      "learning_rate": 1.977908682517658e-05,
+      "loss": 1.6602,
+      "step": 295
+    },
+    {
+      "epoch": 0.23967611336032388,
+      "grad_norm": 1.2506139069671103,
+      "learning_rate": 1.97772865095002e-05,
+      "loss": 1.7444,
+      "step": 296
+    },
+    {
+      "epoch": 0.24048582995951417,
+      "grad_norm": 1.360572780404123,
+      "learning_rate": 1.9775478970396663e-05,
+      "loss": 1.6975,
+      "step": 297
+    },
+    {
+      "epoch": 0.24129554655870444,
+      "grad_norm": 1.2420340903090243,
+      "learning_rate": 1.9773664209201368e-05,
+      "loss": 1.6733,
+      "step": 298
+    },
+    {
+      "epoch": 0.24210526315789474,
+      "grad_norm": 1.2691192435537564,
+      "learning_rate": 1.977184222725505e-05,
+      "loss": 1.6496,
+      "step": 299
+    },
+    {
+      "epoch": 0.242914979757085,
+      "grad_norm": 1.251420231930437,
+      "learning_rate": 1.9770013025903797e-05,
+      "loss": 1.6639,
+      "step": 300
+    },
+    {
+      "epoch": 0.2437246963562753,
+      "grad_norm": 1.3007823671114107,
+      "learning_rate": 1.9768176606499005e-05,
+      "loss": 1.7034,
+      "step": 301
+    },
+    {
+      "epoch": 0.24453441295546557,
+      "grad_norm": 1.2809858772390412,
+      "learning_rate": 1.976633297039742e-05,
+      "loss": 1.6903,
+      "step": 302
+    },
+    {
+      "epoch": 0.24534412955465587,
+      "grad_norm": 1.2280345470671303,
+      "learning_rate": 1.976448211896111e-05,
+      "loss": 1.6723,
+      "step": 303
+    },
+    {
+      "epoch": 0.24615384615384617,
+      "grad_norm": 1.1497220827533912,
+      "learning_rate": 1.9762624053557485e-05,
+      "loss": 1.6747,
+      "step": 304
+    },
+    {
+      "epoch": 0.24696356275303644,
+      "grad_norm": 1.318495032765947,
+      "learning_rate": 1.9760758775559275e-05,
+      "loss": 1.6121,
+      "step": 305
+    },
+    {
+      "epoch": 0.24777327935222673,
+      "grad_norm": 1.2399534094992168,
+      "learning_rate": 1.9758886286344536e-05,
+      "loss": 1.7047,
+      "step": 306
+    },
+    {
+      "epoch": 0.248582995951417,
+      "grad_norm": 1.2398070884711676,
+      "learning_rate": 1.9757006587296664e-05,
+      "loss": 1.6521,
+      "step": 307
+    },
+    {
+      "epoch": 0.2493927125506073,
+      "grad_norm": 1.148807377709599,
+      "learning_rate": 1.975511967980437e-05,
+      "loss": 1.7039,
+      "step": 308
+    },
+    {
+      "epoch": 0.25020242914979757,
+      "grad_norm": 1.213998992883127,
+      "learning_rate": 1.9753225565261695e-05,
+      "loss": 1.6879,
+      "step": 309
+    },
+    {
+      "epoch": 0.25101214574898784,
+      "grad_norm": 1.2397569747054522,
+      "learning_rate": 1.9751324245068008e-05,
+      "loss": 1.7233,
+      "step": 310
+    },
+    {
+      "epoch": 0.25182186234817816,
+      "grad_norm": 1.1848797839629908,
+      "learning_rate": 1.9749415720627993e-05,
+      "loss": 1.6697,
+      "step": 311
+    },
+    {
+      "epoch": 0.25263157894736843,
+      "grad_norm": 1.334789053117042,
+      "learning_rate": 1.974749999335167e-05,
+      "loss": 1.6449,
+      "step": 312
+    },
+    {
+      "epoch": 0.2534412955465587,
+      "grad_norm": 1.1684958236354195,
+      "learning_rate": 1.974557706465436e-05,
+      "loss": 1.6163,
+      "step": 313
+    },
+    {
+      "epoch": 0.25425101214574897,
+      "grad_norm": 1.1517446837404286,
+      "learning_rate": 1.9743646935956727e-05,
+      "loss": 1.6145,
+      "step": 314
+    },
+    {
+      "epoch": 0.2550607287449393,
+      "grad_norm": 1.3651988245433082,
+      "learning_rate": 1.974170960868474e-05,
+      "loss": 1.7181,
+      "step": 315
+    },
+    {
+      "epoch": 0.25587044534412956,
+      "grad_norm": 1.2278060656467518,
+      "learning_rate": 1.973976508426969e-05,
+      "loss": 1.7003,
+      "step": 316
+    },
+    {
+      "epoch": 0.25668016194331983,
+      "grad_norm": 1.298927878086763,
+      "learning_rate": 1.9737813364148187e-05,
+      "loss": 1.7066,
+      "step": 317
+    },
+    {
+      "epoch": 0.2574898785425101,
+      "grad_norm": 1.280731725281388,
+      "learning_rate": 1.973585444976215e-05,
+      "loss": 1.6962,
+      "step": 318
+    },
+    {
+      "epoch": 0.2582995951417004,
+      "grad_norm": 1.3481441118932556,
+      "learning_rate": 1.973388834255882e-05,
+      "loss": 1.6744,
+      "step": 319
+    },
+    {
+      "epoch": 0.2591093117408907,
+      "grad_norm": 1.2693166429730327,
+      "learning_rate": 1.973191504399076e-05,
+      "loss": 1.6762,
+      "step": 320
+    },
+    {
+      "epoch": 0.25991902834008096,
+      "grad_norm": 1.2340493742688543,
+      "learning_rate": 1.9729934555515823e-05,
+      "loss": 1.6857,
+      "step": 321
+    },
+    {
+      "epoch": 0.26072874493927123,
+      "grad_norm": 1.2452283567023132,
+      "learning_rate": 1.9727946878597193e-05,
+      "loss": 1.6444,
+      "step": 322
+    },
+    {
+      "epoch": 0.26153846153846155,
+      "grad_norm": 1.3029700463193483,
+      "learning_rate": 1.9725952014703366e-05,
+      "loss": 1.7188,
+      "step": 323
+    },
+    {
+      "epoch": 0.2623481781376518,
+      "grad_norm": 1.179155915390884,
+      "learning_rate": 1.9723949965308132e-05,
+      "loss": 1.6801,
+      "step": 324
+    },
+    {
+      "epoch": 0.2631578947368421,
+      "grad_norm": 1.286301672885254,
+      "learning_rate": 1.97219407318906e-05,
+      "loss": 1.6617,
+      "step": 325
+    },
+    {
+      "epoch": 0.2639676113360324,
+      "grad_norm": 1.2990996064383904,
+      "learning_rate": 1.9719924315935185e-05,
+      "loss": 1.7269,
+      "step": 326
+    },
+    {
+      "epoch": 0.2647773279352227,
+      "grad_norm": 1.2224528081088977,
+      "learning_rate": 1.971790071893161e-05,
+      "loss": 1.7372,
+      "step": 327
+    },
+    {
+      "epoch": 0.26558704453441295,
+      "grad_norm": 1.2367757457178075,
+      "learning_rate": 1.9715869942374902e-05,
+      "loss": 1.6821,
+      "step": 328
+    },
+    {
+      "epoch": 0.2663967611336032,
+      "grad_norm": 1.2373862157826907,
+      "learning_rate": 1.9713831987765394e-05,
+      "loss": 1.7366,
+      "step": 329
+    },
+    {
+      "epoch": 0.26720647773279355,
+      "grad_norm": 1.2623515990334666,
+      "learning_rate": 1.9711786856608714e-05,
+      "loss": 1.6406,
+      "step": 330
+    },
+    {
+      "epoch": 0.2680161943319838,
+      "grad_norm": 1.1587501394940165,
+      "learning_rate": 1.9709734550415804e-05,
+      "loss": 1.6572,
+      "step": 331
+    },
+    {
+      "epoch": 0.2688259109311741,
+      "grad_norm": 1.3564850748615598,
+      "learning_rate": 1.97076750707029e-05,
+      "loss": 1.6981,
+      "step": 332
+    },
+    {
+      "epoch": 0.26963562753036435,
+      "grad_norm": 1.1910339514470332,
+      "learning_rate": 1.9705608418991534e-05,
+      "loss": 1.712,
+      "step": 333
+    },
+    {
+      "epoch": 0.2704453441295547,
+      "grad_norm": 1.2174211006655304,
+      "learning_rate": 1.9703534596808547e-05,
+      "loss": 1.6783,
+      "step": 334
+    },
+    {
+      "epoch": 0.27125506072874495,
+      "grad_norm": 1.1830375775538684,
+      "learning_rate": 1.970145360568607e-05,
+      "loss": 1.7121,
+      "step": 335
+    },
+    {
+      "epoch": 0.2720647773279352,
+      "grad_norm": 1.1619637628427306,
+      "learning_rate": 1.9699365447161535e-05,
+      "loss": 1.6468,
+      "step": 336
+    },
+    {
+      "epoch": 0.2728744939271255,
+      "grad_norm": 1.2028809053995226,
+      "learning_rate": 1.969727012277766e-05,
+      "loss": 1.6848,
+      "step": 337
+    },
+    {
+      "epoch": 0.2736842105263158,
+      "grad_norm": 1.1621842682222943,
+      "learning_rate": 1.969516763408247e-05,
+      "loss": 1.644,
+      "step": 338
+    },
+    {
+      "epoch": 0.2744939271255061,
+      "grad_norm": 1.2484226953452637,
+      "learning_rate": 1.9693057982629277e-05,
+      "loss": 1.6179,
+      "step": 339
+    },
+    {
+      "epoch": 0.27530364372469635,
+      "grad_norm": 1.3119636180148664,
+      "learning_rate": 1.969094116997668e-05,
+      "loss": 1.7032,
+      "step": 340
+    },
+    {
+      "epoch": 0.2761133603238866,
+      "grad_norm": 1.2508340881483084,
+      "learning_rate": 1.9688817197688576e-05,
+      "loss": 1.6472,
+      "step": 341
+    },
+    {
+      "epoch": 0.27692307692307694,
+      "grad_norm": 1.202434258153614,
+      "learning_rate": 1.968668606733415e-05,
+      "loss": 1.6996,
+      "step": 342
+    },
+    {
+      "epoch": 0.2777327935222672,
+      "grad_norm": 1.1862376977274527,
+      "learning_rate": 1.9684547780487873e-05,
+      "loss": 1.6536,
+      "step": 343
+    },
+    {
+      "epoch": 0.2785425101214575,
+      "grad_norm": 1.3042945008375981,
+      "learning_rate": 1.9682402338729504e-05,
+      "loss": 1.714,
+      "step": 344
+    },
+    {
+      "epoch": 0.2793522267206478,
+      "grad_norm": 1.2321323526734465,
+      "learning_rate": 1.968024974364408e-05,
+      "loss": 1.6788,
+      "step": 345
+    },
+    {
+      "epoch": 0.28016194331983807,
+      "grad_norm": 1.176640860678081,
+      "learning_rate": 1.967808999682195e-05,
+      "loss": 1.6697,
+      "step": 346
+    },
+    {
+      "epoch": 0.28097165991902834,
+      "grad_norm": 1.2519779435098668,
+      "learning_rate": 1.9675923099858712e-05,
+      "loss": 1.663,
+      "step": 347
+    },
+    {
+      "epoch": 0.2817813765182186,
+      "grad_norm": 1.2594751793446184,
+      "learning_rate": 1.9673749054355268e-05,
+      "loss": 1.7454,
+      "step": 348
+    },
+    {
+      "epoch": 0.28259109311740893,
+      "grad_norm": 1.1910600819228896,
+      "learning_rate": 1.9671567861917796e-05,
+      "loss": 1.6278,
+      "step": 349
+    },
+    {
+      "epoch": 0.2834008097165992,
+      "grad_norm": 1.1545397374773405,
+      "learning_rate": 1.9669379524157755e-05,
+      "loss": 1.6888,
+      "step": 350
+    },
+    {
+      "epoch": 0.28421052631578947,
+      "grad_norm": 1.1774954626925083,
+      "learning_rate": 1.9667184042691877e-05,
+      "loss": 1.62,
+      "step": 351
+    },
+    {
+      "epoch": 0.28502024291497974,
+      "grad_norm": 1.1669061183444,
+      "learning_rate": 1.966498141914218e-05,
+      "loss": 1.6305,
+      "step": 352
+    },
+    {
+      "epoch": 0.28582995951417006,
+      "grad_norm": 1.2826242009462285,
+      "learning_rate": 1.9662771655135954e-05,
+      "loss": 1.7025,
+      "step": 353
+    },
+    {
+      "epoch": 0.28663967611336033,
+      "grad_norm": 1.215510453464448,
+      "learning_rate": 1.9660554752305763e-05,
+      "loss": 1.7183,
+      "step": 354
+    },
+    {
+      "epoch": 0.2874493927125506,
+      "grad_norm": 1.1500016494292455,
+      "learning_rate": 1.9658330712289456e-05,
+      "loss": 1.6474,
+      "step": 355
+    },
+    {
+      "epoch": 0.28825910931174087,
+      "grad_norm": 1.1771753941978587,
+      "learning_rate": 1.965609953673014e-05,
+      "loss": 1.6303,
+      "step": 356
+    },
+    {
+      "epoch": 0.2890688259109312,
+      "grad_norm": 1.1888955521350377,
+      "learning_rate": 1.9653861227276197e-05,
+      "loss": 1.7237,
+      "step": 357
+    },
+    {
+      "epoch": 0.28987854251012146,
+      "grad_norm": 1.3602337615622975,
+      "learning_rate": 1.9651615785581287e-05,
+      "loss": 1.6749,
+      "step": 358
+    },
+    {
+      "epoch": 0.29068825910931173,
+      "grad_norm": 1.1994666759895225,
+      "learning_rate": 1.9649363213304337e-05,
+      "loss": 1.6335,
+      "step": 359
+    },
+    {
+      "epoch": 0.291497975708502,
+      "grad_norm": 1.1952494705657468,
+      "learning_rate": 1.9647103512109535e-05,
+      "loss": 1.6583,
+      "step": 360
+    },
+    {
+      "epoch": 0.2923076923076923,
+      "grad_norm": 1.2493452825647193,
+      "learning_rate": 1.9644836683666347e-05,
+      "loss": 1.7336,
+      "step": 361
+    },
+    {
+      "epoch": 0.2931174089068826,
+      "grad_norm": 1.2043236715266186,
+      "learning_rate": 1.9642562729649492e-05,
+      "loss": 1.6457,
+      "step": 362
+    },
+    {
+      "epoch": 0.29392712550607286,
+      "grad_norm": 1.3258481801373758,
+      "learning_rate": 1.964028165173896e-05,
+      "loss": 1.7147,
+      "step": 363
+    },
+    {
+      "epoch": 0.29473684210526313,
+      "grad_norm": 1.3308234537433237,
+      "learning_rate": 1.963799345162001e-05,
+      "loss": 1.7239,
+      "step": 364
+    },
+    {
+      "epoch": 0.29554655870445345,
+      "grad_norm": 1.2105650981201128,
+      "learning_rate": 1.9635698130983153e-05,
+      "loss": 1.69,
+      "step": 365
+    },
+    {
+      "epoch": 0.2963562753036437,
+      "grad_norm": 1.4724018359938562,
+      "learning_rate": 1.9633395691524163e-05,
+      "loss": 1.7904,
+      "step": 366
+    },
+    {
+      "epoch": 0.297165991902834,
+      "grad_norm": 1.167011876559427,
+      "learning_rate": 1.9631086134944076e-05,
+      "loss": 1.6719,
+      "step": 367
+    },
+    {
+      "epoch": 0.2979757085020243,
+      "grad_norm": 1.2115645319681754,
+      "learning_rate": 1.9628769462949187e-05,
+      "loss": 1.7076,
+      "step": 368
+    },
+    {
+      "epoch": 0.2987854251012146,
+      "grad_norm": 1.2930285867259568,
+      "learning_rate": 1.9626445677251043e-05,
+      "loss": 1.7555,
+      "step": 369
+    },
+    {
+      "epoch": 0.29959514170040485,
+      "grad_norm": 1.1530085275414237,
+      "learning_rate": 1.962411477956645e-05,
+      "loss": 1.6965,
+      "step": 370
+    },
+    {
+      "epoch": 0.3004048582995951,
+      "grad_norm": 1.4097103876759387,
+      "learning_rate": 1.9621776771617464e-05,
+      "loss": 1.6768,
+      "step": 371
+    },
+    {
+      "epoch": 0.30121457489878545,
+      "grad_norm": 1.168938484091032,
+      "learning_rate": 1.9619431655131404e-05,
+      "loss": 1.6209,
+      "step": 372
+    },
+    {
+      "epoch": 0.3020242914979757,
+      "grad_norm": 1.2057005406635852,
+      "learning_rate": 1.961707943184083e-05,
+      "loss": 1.6893,
+      "step": 373
+    },
+    {
+      "epoch": 0.302834008097166,
+      "grad_norm": 1.2784392150421353,
+      "learning_rate": 1.9614720103483562e-05,
+      "loss": 1.7162,
+      "step": 374
+    },
+    {
+      "epoch": 0.30364372469635625,
+      "grad_norm": 1.2120518452791202,
+      "learning_rate": 1.9612353671802658e-05,
+      "loss": 1.5952,
+      "step": 375
+    },
+    {
+      "epoch": 0.3044534412955466,
+      "grad_norm": 1.2873149200801106,
+      "learning_rate": 1.960998013854643e-05,
+      "loss": 1.7035,
+      "step": 376
+    },
+    {
+      "epoch": 0.30526315789473685,
+      "grad_norm": 1.3006849812067949,
+      "learning_rate": 1.960759950546844e-05,
+      "loss": 1.6379,
+      "step": 377
+    },
+    {
+      "epoch": 0.3060728744939271,
+      "grad_norm": 1.2990265392717135,
+      "learning_rate": 1.960521177432749e-05,
+      "loss": 1.7101,
+      "step": 378
+    },
+    {
+      "epoch": 0.3068825910931174,
+      "grad_norm": 1.3082477762819935,
+      "learning_rate": 1.9602816946887634e-05,
+      "loss": 1.691,
+      "step": 379
+    },
+    {
+      "epoch": 0.3076923076923077,
+      "grad_norm": 1.2057202532696198,
+      "learning_rate": 1.960041502491815e-05,
+      "loss": 1.7229,
+      "step": 380
+    },
+    {
+      "epoch": 0.308502024291498,
+      "grad_norm": 1.155390745921477,
+      "learning_rate": 1.959800601019358e-05,
+      "loss": 1.649,
+      "step": 381
+    },
+    {
+      "epoch": 0.30931174089068825,
+      "grad_norm": 1.2547425138455335,
+      "learning_rate": 1.9595589904493696e-05,
+      "loss": 1.7094,
+      "step": 382
+    },
+    {
+      "epoch": 0.3101214574898785,
+      "grad_norm": 1.2268479082620285,
+      "learning_rate": 1.9593166709603503e-05,
+      "loss": 1.6612,
+      "step": 383
+    },
+    {
+      "epoch": 0.31093117408906884,
+      "grad_norm": 1.247022945456135,
+      "learning_rate": 1.9590736427313255e-05,
+      "loss": 1.7201,
+      "step": 384
+    },
+    {
+      "epoch": 0.3117408906882591,
+      "grad_norm": 1.1780738130918533,
+      "learning_rate": 1.9588299059418434e-05,
+      "loss": 1.707,
+      "step": 385
+    },
+    {
+      "epoch": 0.3125506072874494,
+      "grad_norm": 1.3110540287499943,
+      "learning_rate": 1.958585460771976e-05,
+      "loss": 1.5952,
+      "step": 386
+    },
+    {
+      "epoch": 0.3133603238866397,
+      "grad_norm": 1.2356287734723737,
+      "learning_rate": 1.9583403074023183e-05,
+      "loss": 1.7514,
+      "step": 387
+    },
+    {
+      "epoch": 0.31417004048582997,
+      "grad_norm": 1.251297643939624,
+      "learning_rate": 1.9580944460139896e-05,
+      "loss": 1.6577,
+      "step": 388
+    },
+    {
+      "epoch": 0.31497975708502024,
+      "grad_norm": 1.283479017126295,
+      "learning_rate": 1.9578478767886303e-05,
+      "loss": 1.6909,
+      "step": 389
+    },
+    {
+      "epoch": 0.3157894736842105,
+      "grad_norm": 1.1892005294340031,
+      "learning_rate": 1.957600599908406e-05,
+      "loss": 1.6956,
+      "step": 390
+    },
+    {
+      "epoch": 0.31659919028340083,
+      "grad_norm": 1.299357039138039,
+      "learning_rate": 1.957352615556004e-05,
+      "loss": 1.6994,
+      "step": 391
+    },
+    {
+      "epoch": 0.3174089068825911,
+      "grad_norm": 1.2192402309654093,
+      "learning_rate": 1.9571039239146332e-05,
+      "loss": 1.7085,
+      "step": 392
+    },
+    {
+      "epoch": 0.31821862348178137,
+      "grad_norm": 1.2089295961333866,
+      "learning_rate": 1.9568545251680272e-05,
+      "loss": 1.7312,
+      "step": 393
+    },
+    {
+      "epoch": 0.31902834008097164,
+      "grad_norm": 1.2002238729382657,
+      "learning_rate": 1.956604419500441e-05,
+      "loss": 1.6919,
+      "step": 394
+    },
+    {
+      "epoch": 0.31983805668016196,
+      "grad_norm": 1.1372502072400479,
+      "learning_rate": 1.9563536070966513e-05,
+      "loss": 1.6842,
+      "step": 395
+    },
+    {
+      "epoch": 0.32064777327935223,
+      "grad_norm": 1.1848402112810035,
+      "learning_rate": 1.956102088141958e-05,
+      "loss": 1.6854,
+      "step": 396
+    },
+    {
+      "epoch": 0.3214574898785425,
+      "grad_norm": 1.212172944607681,
+      "learning_rate": 1.9558498628221816e-05,
+      "loss": 1.6693,
+      "step": 397
+    },
+    {
+      "epoch": 0.32226720647773277,
+      "grad_norm": 1.2323473874443758,
+      "learning_rate": 1.9555969313236666e-05,
+      "loss": 1.6916,
+      "step": 398
+    },
+    {
+      "epoch": 0.3230769230769231,
+      "grad_norm": 1.197869807167727,
+      "learning_rate": 1.955343293833277e-05,
+      "loss": 1.6777,
+      "step": 399
+    },
+    {
+      "epoch": 0.32388663967611336,
+      "grad_norm": 1.2182554044770237,
+      "learning_rate": 1.9550889505383996e-05,
+      "loss": 1.7348,
+      "step": 400
+    },
+    {
+      "epoch": 0.32469635627530363,
+      "grad_norm": 1.2204060971165689,
+      "learning_rate": 1.954833901626943e-05,
+      "loss": 1.6656,
+      "step": 401
+    },
+    {
+      "epoch": 0.3255060728744939,
+      "grad_norm": 1.2003350489961402,
+      "learning_rate": 1.9545781472873354e-05,
+      "loss": 1.6162,
+      "step": 402
+    },
+    {
+      "epoch": 0.3263157894736842,
+      "grad_norm": 1.3712370071917266,
+      "learning_rate": 1.954321687708528e-05,
+      "loss": 1.6689,
+      "step": 403
+    },
+    {
+      "epoch": 0.3271255060728745,
+      "grad_norm": 1.2066042919530333,
+      "learning_rate": 1.954064523079992e-05,
+      "loss": 1.6092,
+      "step": 404
+    },
+    {
+      "epoch": 0.32793522267206476,
+      "grad_norm": 1.2696649824839026,
+      "learning_rate": 1.9538066535917196e-05,
+      "loss": 1.6442,
+      "step": 405
+    },
+    {
+      "epoch": 0.3287449392712551,
+      "grad_norm": 1.470140395099795,
+      "learning_rate": 1.9535480794342248e-05,
+      "loss": 1.6058,
+      "step": 406
+    },
+    {
+      "epoch": 0.32955465587044536,
+      "grad_norm": 1.3119894938694066,
+      "learning_rate": 1.9532888007985408e-05,
+      "loss": 1.7058,
+      "step": 407
+    },
+    {
+      "epoch": 0.3303643724696356,
+      "grad_norm": 1.4968426504458228,
+      "learning_rate": 1.9530288178762213e-05,
+      "loss": 1.6929,
+      "step": 408
+    },
+    {
+      "epoch": 0.3311740890688259,
+      "grad_norm": 1.2601976626437872,
+      "learning_rate": 1.9527681308593412e-05,
+      "loss": 1.6336,
+      "step": 409
+    },
+    {
+      "epoch": 0.3319838056680162,
+      "grad_norm": 1.3452338661981322,
+      "learning_rate": 1.952506739940496e-05,
+      "loss": 1.7445,
+      "step": 410
+    },
+    {
+      "epoch": 0.3327935222672065,
+      "grad_norm": 1.2717619212491589,
+      "learning_rate": 1.9522446453127994e-05,
+      "loss": 1.6666,
+      "step": 411
+    },
+    {
+      "epoch": 0.33360323886639676,
+      "grad_norm": 1.1984329741126858,
+      "learning_rate": 1.951981847169886e-05,
+      "loss": 1.7448,
+      "step": 412
+    },
+    {
+      "epoch": 0.334412955465587,
+      "grad_norm": 1.1961449074373034,
+      "learning_rate": 1.951718345705911e-05,
+      "loss": 1.667,
+      "step": 413
+    },
+    {
+      "epoch": 0.33522267206477735,
+      "grad_norm": 1.330214100403057,
+      "learning_rate": 1.9514541411155478e-05,
+      "loss": 1.7108,
+      "step": 414
+    },
+    {
+      "epoch": 0.3360323886639676,
+      "grad_norm": 1.183842085273524,
+      "learning_rate": 1.9511892335939904e-05,
+      "loss": 1.621,
+      "step": 415
+    },
+    {
+      "epoch": 0.3368421052631579,
+      "grad_norm": 1.2664210311376265,
+      "learning_rate": 1.950923623336951e-05,
+      "loss": 1.6512,
+      "step": 416
+    },
+    {
+      "epoch": 0.33765182186234816,
+      "grad_norm": 1.1261494079421026,
+      "learning_rate": 1.9506573105406623e-05,
+      "loss": 1.6123,
+      "step": 417
+    },
+    {
+      "epoch": 0.3384615384615385,
+      "grad_norm": 1.27913282791293,
+      "learning_rate": 1.9503902954018748e-05,
+      "loss": 1.7083,
+      "step": 418
+    },
+    {
+      "epoch": 0.33927125506072875,
+      "grad_norm": 1.1019506289059786,
+      "learning_rate": 1.9501225781178586e-05,
+      "loss": 1.6501,
+      "step": 419
+    },
+    {
+      "epoch": 0.340080971659919,
+      "grad_norm": 1.1163063251074246,
+      "learning_rate": 1.9498541588864022e-05,
+      "loss": 1.626,
+      "step": 420
+    },
+    {
+      "epoch": 0.3408906882591093,
+      "grad_norm": 1.1208221900644928,
+      "learning_rate": 1.9495850379058127e-05,
+      "loss": 1.6486,
+      "step": 421
+    },
+    {
+      "epoch": 0.3417004048582996,
+      "grad_norm": 1.1252985647383595,
+      "learning_rate": 1.9493152153749162e-05,
+      "loss": 1.6098,
+      "step": 422
+    },
+    {
+      "epoch": 0.3425101214574899,
+      "grad_norm": 1.1658129970131546,
+      "learning_rate": 1.9490446914930564e-05,
+      "loss": 1.6067,
+      "step": 423
+    },
+    {
+      "epoch": 0.34331983805668015,
+      "grad_norm": 1.2465190601078746,
+      "learning_rate": 1.9487734664600956e-05,
+      "loss": 1.7689,
+      "step": 424
+    },
+    {
+      "epoch": 0.3441295546558704,
+      "grad_norm": 1.2032865897994165,
+      "learning_rate": 1.948501540476414e-05,
+      "loss": 1.6348,
+      "step": 425
+    },
+    {
+      "epoch": 0.34493927125506074,
+      "grad_norm": 1.3211565552357492,
+      "learning_rate": 1.9482289137429098e-05,
+      "loss": 1.6349,
+      "step": 426
+    },
+    {
+      "epoch": 0.345748987854251,
+      "grad_norm": 1.1307592944834632,
+      "learning_rate": 1.9479555864609985e-05,
+      "loss": 1.593,
+      "step": 427
+    },
+    {
+      "epoch": 0.3465587044534413,
+      "grad_norm": 1.2787848323635016,
+      "learning_rate": 1.947681558832613e-05,
+      "loss": 1.7022,
+      "step": 428
+    },
+    {
+      "epoch": 0.3473684210526316,
+      "grad_norm": 1.2091547875890887,
+      "learning_rate": 1.9474068310602048e-05,
+      "loss": 1.6566,
+      "step": 429
+    },
+    {
+      "epoch": 0.3481781376518219,
+      "grad_norm": 1.2483509469560066,
+      "learning_rate": 1.9471314033467413e-05,
+      "loss": 1.6993,
+      "step": 430
+    },
+    {
+      "epoch": 0.34898785425101214,
+      "grad_norm": 1.2498538878012477,
+      "learning_rate": 1.9468552758957076e-05,
+      "loss": 1.7252,
+      "step": 431
+    },
+    {
+      "epoch": 0.3497975708502024,
+      "grad_norm": 1.1989025930704638,
+      "learning_rate": 1.9465784489111063e-05,
+      "loss": 1.6999,
+      "step": 432
+    },
+    {
+      "epoch": 0.35060728744939273,
+      "grad_norm": 1.1762119281728465,
+      "learning_rate": 1.9463009225974558e-05,
+      "loss": 1.6501,
+      "step": 433
+    },
+    {
+      "epoch": 0.351417004048583,
+      "grad_norm": 1.2950834978750903,
+      "learning_rate": 1.9460226971597916e-05,
+      "loss": 1.7399,
+      "step": 434
+    },
+    {
+      "epoch": 0.3522267206477733,
+      "grad_norm": 1.2455256194553699,
+      "learning_rate": 1.945743772803666e-05,
+      "loss": 1.6708,
+      "step": 435
+    },
+    {
+      "epoch": 0.35303643724696354,
+      "grad_norm": 1.2202828687384126,
+      "learning_rate": 1.945464149735147e-05,
+      "loss": 1.6896,
+      "step": 436
+    },
+    {
+      "epoch": 0.35384615384615387,
+      "grad_norm": 1.158420520477845,
+      "learning_rate": 1.94518382816082e-05,
+      "loss": 1.6619,
+      "step": 437
+    },
+    {
+      "epoch": 0.35465587044534413,
+      "grad_norm": 1.1779822358008685,
+      "learning_rate": 1.9449028082877843e-05,
+      "loss": 1.6571,
+      "step": 438
+    },
+    {
+      "epoch": 0.3554655870445344,
+      "grad_norm": 1.1534682512913468,
+      "learning_rate": 1.944621090323658e-05,
+      "loss": 1.5662,
+      "step": 439
+    },
+    {
+      "epoch": 0.3562753036437247,
+      "grad_norm": 1.231944980721245,
+      "learning_rate": 1.9443386744765726e-05,
+      "loss": 1.6072,
+      "step": 440
+    },
+    {
+      "epoch": 0.357085020242915,
+      "grad_norm": 1.2309431429104132,
+      "learning_rate": 1.9440555609551758e-05,
+      "loss": 1.6478,
+      "step": 441
+    },
+    {
+      "epoch": 0.35789473684210527,
+      "grad_norm": 1.1256347239923377,
+      "learning_rate": 1.9437717499686313e-05,
+      "loss": 1.5872,
+      "step": 442
+    },
+    {
+      "epoch": 0.35870445344129553,
+      "grad_norm": 1.177945395683949,
+      "learning_rate": 1.9434872417266176e-05,
+      "loss": 1.6847,
+      "step": 443
+    },
+    {
+      "epoch": 0.3595141700404858,
+      "grad_norm": 1.1817750781529393,
+      "learning_rate": 1.943202036439329e-05,
+      "loss": 1.6555,
+      "step": 444
+    },
+    {
+      "epoch": 0.3603238866396761,
+      "grad_norm": 1.155486180808622,
+      "learning_rate": 1.942916134317473e-05,
+      "loss": 1.703,
+      "step": 445
+    },
+    {
+      "epoch": 0.3611336032388664,
+      "grad_norm": 1.2249672033301868,
+      "learning_rate": 1.9426295355722745e-05,
+      "loss": 1.6686,
+      "step": 446
+    },
+    {
+      "epoch": 0.36194331983805667,
+      "grad_norm": 1.1939431449591948,
+      "learning_rate": 1.9423422404154708e-05,
+      "loss": 1.6403,
+      "step": 447
+    },
+    {
+      "epoch": 0.362753036437247,
+      "grad_norm": 1.283040329656735,
+      "learning_rate": 1.942054249059315e-05,
+      "loss": 1.7033,
+      "step": 448
+    },
+    {
+      "epoch": 0.36356275303643726,
+      "grad_norm": 1.2302822837989298,
+      "learning_rate": 1.941765561716574e-05,
+      "loss": 1.6309,
+      "step": 449
+    },
+    {
+      "epoch": 0.3643724696356275,
+      "grad_norm": 1.258830672551768,
+      "learning_rate": 1.9414761786005293e-05,
+      "loss": 1.6807,
+      "step": 450
+    },
+    {
+      "epoch": 0.3651821862348178,
+      "grad_norm": 1.2346970982772958,
+      "learning_rate": 1.9411860999249762e-05,
+      "loss": 1.6063,
+      "step": 451
+    },
+    {
+      "epoch": 0.3659919028340081,
+      "grad_norm": 1.323015686111149,
+      "learning_rate": 1.9408953259042236e-05,
+      "loss": 1.6917,
+      "step": 452
+    },
+    {
+      "epoch": 0.3668016194331984,
+      "grad_norm": 1.1863762011370744,
+      "learning_rate": 1.9406038567530944e-05,
+      "loss": 1.6365,
+      "step": 453
+    },
+    {
+      "epoch": 0.36761133603238866,
+      "grad_norm": 1.2001660652965502,
+      "learning_rate": 1.9403116926869253e-05,
+      "loss": 1.6383,
+      "step": 454
+    },
+    {
+      "epoch": 0.3684210526315789,
+      "grad_norm": 1.2089353111723722,
+      "learning_rate": 1.9400188339215657e-05,
+      "loss": 1.6778,
+      "step": 455
+    },
+    {
+      "epoch": 0.36923076923076925,
+      "grad_norm": 1.2546231222228346,
+      "learning_rate": 1.9397252806733793e-05,
+      "loss": 1.6405,
+      "step": 456
+    },
+    {
+      "epoch": 0.3700404858299595,
+      "grad_norm": 1.3359592739488402,
+      "learning_rate": 1.939431033159242e-05,
+      "loss": 1.643,
+      "step": 457
+    },
+    {
+      "epoch": 0.3708502024291498,
+      "grad_norm": 1.1715630102315027,
+      "learning_rate": 1.9391360915965426e-05,
+      "loss": 1.5935,
+      "step": 458
+    },
+    {
+      "epoch": 0.37165991902834006,
+      "grad_norm": 1.1385520627989953,
+      "learning_rate": 1.9388404562031836e-05,
+      "loss": 1.6979,
+      "step": 459
+    },
+    {
+      "epoch": 0.3724696356275304,
+      "grad_norm": 1.2399884358282447,
+      "learning_rate": 1.9385441271975786e-05,
+      "loss": 1.6434,
+      "step": 460
+    },
+    {
+      "epoch": 0.37327935222672065,
+      "grad_norm": 1.163385165068389,
+      "learning_rate": 1.9382471047986555e-05,
+      "loss": 1.6838,
+      "step": 461
+    },
+    {
+      "epoch": 0.3740890688259109,
+      "grad_norm": 1.1225008795816183,
+      "learning_rate": 1.9379493892258527e-05,
+      "loss": 1.6541,
+      "step": 462
+    },
+    {
+      "epoch": 0.3748987854251012,
+      "grad_norm": 1.2531286384618345,
+      "learning_rate": 1.937650980699122e-05,
+      "loss": 1.654,
+      "step": 463
+    },
+    {
+      "epoch": 0.3757085020242915,
+      "grad_norm": 1.1876213186476707,
+      "learning_rate": 1.9373518794389263e-05,
+      "loss": 1.6901,
+      "step": 464
+    },
+    {
+      "epoch": 0.3765182186234818,
+      "grad_norm": 1.112836379139156,
+      "learning_rate": 1.9370520856662406e-05,
+      "loss": 1.6399,
+      "step": 465
+    },
+    {
+      "epoch": 0.37732793522267205,
+      "grad_norm": 1.3185471861887414,
+      "learning_rate": 1.9367515996025516e-05,
+      "loss": 1.6746,
+      "step": 466
+    },
+    {
+      "epoch": 0.3781376518218624,
+      "grad_norm": 1.3149218591633627,
+      "learning_rate": 1.9364504214698578e-05,
+      "loss": 1.6721,
+      "step": 467
+    },
+    {
+      "epoch": 0.37894736842105264,
+      "grad_norm": 1.1367079723459552,
+      "learning_rate": 1.936148551490668e-05,
+      "loss": 1.7142,
+      "step": 468
+    },
+    {
+      "epoch": 0.3797570850202429,
+      "grad_norm": 1.1910905835785592,
+      "learning_rate": 1.935845989888003e-05,
+      "loss": 1.6307,
+      "step": 469
+    },
+    {
+      "epoch": 0.3805668016194332,
+      "grad_norm": 1.1936772698797782,
+      "learning_rate": 1.9355427368853946e-05,
+      "loss": 1.6976,
+      "step": 470
+    },
+    {
+      "epoch": 0.3813765182186235,
+      "grad_norm": 1.2030550020345074,
+      "learning_rate": 1.935238792706885e-05,
+      "loss": 1.7098,
+      "step": 471
+    },
+    {
+      "epoch": 0.3821862348178138,
+      "grad_norm": 1.171000490443613,
+      "learning_rate": 1.934934157577027e-05,
+      "loss": 1.7119,
+      "step": 472
+    },
+    {
+      "epoch": 0.38299595141700404,
+      "grad_norm": 1.164404989871964,
+      "learning_rate": 1.934628831720884e-05,
+      "loss": 1.6364,
+      "step": 473
+    },
+    {
+      "epoch": 0.3838056680161943,
+      "grad_norm": 1.1421132414106066,
+      "learning_rate": 1.9343228153640296e-05,
+      "loss": 1.6652,
+      "step": 474
+    },
+    {
+      "epoch": 0.38461538461538464,
+      "grad_norm": 1.1666775642165048,
+      "learning_rate": 1.9340161087325483e-05,
+      "loss": 1.6711,
+      "step": 475
+    },
+    {
+      "epoch": 0.3854251012145749,
+      "grad_norm": 1.1725292925597457,
+      "learning_rate": 1.9337087120530335e-05,
+      "loss": 1.5852,
+      "step": 476
+    },
+    {
+      "epoch": 0.3862348178137652,
+      "grad_norm": 1.2217452550533756,
+      "learning_rate": 1.9334006255525884e-05,
+      "loss": 1.6768,
+      "step": 477
+    },
+    {
+      "epoch": 0.38704453441295544,
+      "grad_norm": 1.1442586319126382,
+      "learning_rate": 1.9330918494588275e-05,
+      "loss": 1.6791,
+      "step": 478
+    },
+    {
+      "epoch": 0.38785425101214577,
+      "grad_norm": 1.0946129383309182,
+      "learning_rate": 1.9327823839998726e-05,
+      "loss": 1.7151,
+      "step": 479
+    },
+    {
+      "epoch": 0.38866396761133604,
+      "grad_norm": 1.2204161246543435,
+      "learning_rate": 1.932472229404356e-05,
+      "loss": 1.6237,
+      "step": 480
+    },
+    {
+      "epoch": 0.3894736842105263,
+      "grad_norm": 1.1880149763428396,
+      "learning_rate": 1.932161385901419e-05,
+      "loss": 1.6761,
+      "step": 481
+    },
+    {
+      "epoch": 0.3902834008097166,
+      "grad_norm": 1.2592668241512581,
+      "learning_rate": 1.931849853720712e-05,
+      "loss": 1.5898,
+      "step": 482
+    },
+    {
+      "epoch": 0.3910931174089069,
+      "grad_norm": 1.209479528339513,
+      "learning_rate": 1.931537633092393e-05,
+      "loss": 1.6279,
+      "step": 483
+    },
+    {
+      "epoch": 0.39190283400809717,
+      "grad_norm": 1.2513199284132253,
+      "learning_rate": 1.9312247242471306e-05,
+      "loss": 1.6343,
+      "step": 484
+    },
+    {
+      "epoch": 0.39271255060728744,
+      "grad_norm": 1.1989743043638685,
+      "learning_rate": 1.9309111274161005e-05,
+      "loss": 1.6881,
+      "step": 485
+    },
+    {
+      "epoch": 0.39352226720647776,
+      "grad_norm": 1.2273758702155306,
+      "learning_rate": 1.930596842830987e-05,
+      "loss": 1.6703,
+      "step": 486
+    },
+    {
+      "epoch": 0.39433198380566803,
+      "grad_norm": 1.1911609997920505,
+      "learning_rate": 1.9302818707239822e-05,
+      "loss": 1.6359,
+      "step": 487
+    },
+    {
+      "epoch": 0.3951417004048583,
+      "grad_norm": 1.1555821594774291,
+      "learning_rate": 1.9299662113277867e-05,
+      "loss": 1.7005,
+      "step": 488
+    },
+    {
+      "epoch": 0.39595141700404857,
+      "grad_norm": 1.161714222746895,
+      "learning_rate": 1.929649864875609e-05,
+      "loss": 1.6259,
+      "step": 489
+    },
+    {
+      "epoch": 0.3967611336032389,
+      "grad_norm": 1.2024264839115575,
+      "learning_rate": 1.9293328316011645e-05,
+      "loss": 1.6454,
+      "step": 490
+    },
+    {
+      "epoch": 0.39757085020242916,
+      "grad_norm": 1.182376794250563,
+      "learning_rate": 1.929015111738676e-05,
+      "loss": 1.6839,
+      "step": 491
+    },
+    {
+      "epoch": 0.39838056680161943,
+      "grad_norm": 1.2463525022449193,
+      "learning_rate": 1.9286967055228744e-05,
+      "loss": 1.7041,
+      "step": 492
+    },
+    {
+      "epoch": 0.3991902834008097,
+      "grad_norm": 1.1765792609257826,
+      "learning_rate": 1.928377613188997e-05,
+      "loss": 1.6963,
+      "step": 493
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.1464191471865726,
+      "learning_rate": 1.9280578349727882e-05,
+      "loss": 1.7587,
+      "step": 494
+    },
+    {
+      "epoch": 0.4008097165991903,
+      "grad_norm": 1.2250198503430287,
+      "learning_rate": 1.927737371110499e-05,
+      "loss": 1.6011,
+      "step": 495
+    },
+    {
+      "epoch": 0.40161943319838056,
+      "grad_norm": 1.2273800649829825,
+      "learning_rate": 1.927416221838887e-05,
+      "loss": 1.7219,
+      "step": 496
+    },
+    {
+      "epoch": 0.40242914979757083,
+      "grad_norm": 1.1823664342882607,
+      "learning_rate": 1.9270943873952162e-05,
+      "loss": 1.6612,
+      "step": 497
+    },
+    {
+      "epoch": 0.40323886639676115,
+      "grad_norm": 1.2129258567045582,
+      "learning_rate": 1.9267718680172574e-05,
+      "loss": 1.6969,
+      "step": 498
+    },
+    {
+      "epoch": 0.4040485829959514,
+      "grad_norm": 1.142011991460994,
+      "learning_rate": 1.926448663943286e-05,
+      "loss": 1.6331,
+      "step": 499
+    },
+    {
+      "epoch": 0.4048582995951417,
+      "grad_norm": 1.1900750233989383,
+      "learning_rate": 1.9261247754120846e-05,
+      "loss": 1.5953,
+      "step": 500
+    },
+    {
+      "epoch": 0.40566801619433196,
+      "grad_norm": 1.1876038257951016,
+      "learning_rate": 1.925800202662941e-05,
+      "loss": 1.635,
+      "step": 501
+    },
+    {
+      "epoch": 0.4064777327935223,
+      "grad_norm": 1.1483754178488699,
+      "learning_rate": 1.9254749459356482e-05,
+      "loss": 1.672,
+      "step": 502
+    },
+    {
+      "epoch": 0.40728744939271255,
+      "grad_norm": 1.2261880502908553,
+      "learning_rate": 1.9251490054705053e-05,
+      "loss": 1.7095,
+      "step": 503
+    },
+    {
+      "epoch": 0.4080971659919028,
+      "grad_norm": 1.2012813998498333,
+      "learning_rate": 1.9248223815083155e-05,
+      "loss": 1.6417,
+      "step": 504
+    },
+    {
+      "epoch": 0.4089068825910931,
+      "grad_norm": 1.1128246203440006,
+      "learning_rate": 1.924495074290388e-05,
+      "loss": 1.6963,
+      "step": 505
+    },
+    {
+      "epoch": 0.4097165991902834,
+      "grad_norm": 1.1860636204765815,
+      "learning_rate": 1.9241670840585357e-05,
+      "loss": 1.6741,
+      "step": 506
+    },
+    {
+      "epoch": 0.4105263157894737,
+      "grad_norm": 1.3472485519941122,
+      "learning_rate": 1.923838411055077e-05,
+      "loss": 1.6403,
+      "step": 507
+    },
+    {
+      "epoch": 0.41133603238866395,
+      "grad_norm": 1.243377675921519,
+      "learning_rate": 1.923509055522835e-05,
+      "loss": 1.6858,
+      "step": 508
+    },
+    {
+      "epoch": 0.4121457489878543,
+      "grad_norm": 1.1780699401889794,
+      "learning_rate": 1.9231790177051354e-05,
+      "loss": 1.7021,
+      "step": 509
+    },
+    {
+      "epoch": 0.41295546558704455,
+      "grad_norm": 1.1235409565713237,
+      "learning_rate": 1.92284829784581e-05,
+      "loss": 1.8083,
+      "step": 510
+    },
+    {
+      "epoch": 0.4137651821862348,
+      "grad_norm": 1.2751087877042944,
+      "learning_rate": 1.922516896189193e-05,
+      "loss": 1.6194,
+      "step": 511
+    },
+    {
+      "epoch": 0.4145748987854251,
+      "grad_norm": 1.1876545871499165,
+      "learning_rate": 1.922184812980123e-05,
+      "loss": 1.6526,
+      "step": 512
+    },
+    {
+      "epoch": 0.4153846153846154,
+      "grad_norm": 1.1881650289467287,
+      "learning_rate": 1.921852048463942e-05,
+      "loss": 1.6574,
+      "step": 513
+    },
+    {
+      "epoch": 0.4161943319838057,
+      "grad_norm": 1.2137821843512242,
+      "learning_rate": 1.9215186028864955e-05,
+      "loss": 1.631,
+      "step": 514
+    },
+    {
+      "epoch": 0.41700404858299595,
+      "grad_norm": 1.2205249821982338,
+      "learning_rate": 1.9211844764941318e-05,
+      "loss": 1.6862,
+      "step": 515
+    },
+    {
+      "epoch": 0.4178137651821862,
+      "grad_norm": 1.229223049157285,
+      "learning_rate": 1.920849669533702e-05,
+      "loss": 1.7166,
+      "step": 516
+    },
+    {
+      "epoch": 0.41862348178137654,
+      "grad_norm": 1.1248482042561723,
+      "learning_rate": 1.920514182252561e-05,
+      "loss": 1.7045,
+      "step": 517
+    },
+    {
+      "epoch": 0.4194331983805668,
+      "grad_norm": 1.1551125743150399,
+      "learning_rate": 1.9201780148985657e-05,
+      "loss": 1.6242,
+      "step": 518
+    },
+    {
+      "epoch": 0.4202429149797571,
+      "grad_norm": 1.1669084355179586,
+      "learning_rate": 1.9198411677200753e-05,
+      "loss": 1.708,
+      "step": 519
+    },
+    {
+      "epoch": 0.42105263157894735,
+      "grad_norm": 1.1195972354108037,
+      "learning_rate": 1.919503640965951e-05,
+      "loss": 1.6543,
+      "step": 520
+    },
+    {
+      "epoch": 0.42186234817813767,
+      "grad_norm": 1.1430422701282208,
+      "learning_rate": 1.919165434885557e-05,
+      "loss": 1.6929,
+      "step": 521
+    },
+    {
+      "epoch": 0.42267206477732794,
+      "grad_norm": 1.248012426300318,
+      "learning_rate": 1.9188265497287587e-05,
+      "loss": 1.6886,
+      "step": 522
+    },
+    {
+      "epoch": 0.4234817813765182,
+      "grad_norm": 1.2892887919197453,
+      "learning_rate": 1.9184869857459233e-05,
+      "loss": 1.6493,
+      "step": 523
+    },
+    {
+      "epoch": 0.4242914979757085,
+      "grad_norm": 1.1122675853550497,
+      "learning_rate": 1.918146743187919e-05,
+      "loss": 1.6726,
+      "step": 524
+    },
+    {
+      "epoch": 0.4251012145748988,
+      "grad_norm": 1.2687690810982857,
+      "learning_rate": 1.917805822306117e-05,
+      "loss": 1.6835,
+      "step": 525
+    },
+    {
+      "epoch": 0.42591093117408907,
+      "grad_norm": 1.248733843761085,
+      "learning_rate": 1.9174642233523876e-05,
+      "loss": 1.6022,
+      "step": 526
+    },
+    {
+      "epoch": 0.42672064777327934,
+      "grad_norm": 1.1610405823160372,
+      "learning_rate": 1.9171219465791037e-05,
+      "loss": 1.6776,
+      "step": 527
+    },
+    {
+      "epoch": 0.42753036437246966,
+      "grad_norm": 1.1844102471692486,
+      "learning_rate": 1.9167789922391374e-05,
+      "loss": 1.7224,
+      "step": 528
+    },
+    {
+      "epoch": 0.42834008097165993,
+      "grad_norm": 1.1045406051602051,
+      "learning_rate": 1.916435360585863e-05,
+      "loss": 1.6201,
+      "step": 529
+    },
+    {
+      "epoch": 0.4291497975708502,
+      "grad_norm": 1.1718884314754825,
+      "learning_rate": 1.916091051873154e-05,
+      "loss": 1.6654,
+      "step": 530
+    },
+    {
+      "epoch": 0.42995951417004047,
+      "grad_norm": 1.2149881721900548,
+      "learning_rate": 1.915746066355385e-05,
+      "loss": 1.6386,
+      "step": 531
+    },
+    {
+      "epoch": 0.4307692307692308,
+      "grad_norm": 1.122281282890927,
+      "learning_rate": 1.9154004042874295e-05,
+      "loss": 1.6699,
+      "step": 532
+    },
+    {
+      "epoch": 0.43157894736842106,
+      "grad_norm": 1.1414789158751815,
+      "learning_rate": 1.915054065924662e-05,
+      "loss": 1.7295,
+      "step": 533
+    },
+    {
+      "epoch": 0.43238866396761133,
+      "grad_norm": 1.1773843495524834,
+      "learning_rate": 1.914707051522956e-05,
+      "loss": 1.6519,
+      "step": 534
+    },
+    {
+      "epoch": 0.4331983805668016,
+      "grad_norm": 1.1370478468226877,
+      "learning_rate": 1.9143593613386845e-05,
+      "loss": 1.663,
+      "step": 535
+    },
+    {
+      "epoch": 0.4340080971659919,
+      "grad_norm": 1.2049613179791026,
+      "learning_rate": 1.9140109956287202e-05,
+      "loss": 1.6453,
+      "step": 536
+    },
+    {
+      "epoch": 0.4348178137651822,
+      "grad_norm": 1.1634868725319234,
+      "learning_rate": 1.9136619546504344e-05,
+      "loss": 1.626,
+      "step": 537
+    },
+    {
+      "epoch": 0.43562753036437246,
+      "grad_norm": 1.151511606446221,
+      "learning_rate": 1.9133122386616972e-05,
+      "loss": 1.676,
+      "step": 538
+    },
+    {
+      "epoch": 0.43643724696356273,
+      "grad_norm": 1.1352020906409594,
+      "learning_rate": 1.9129618479208775e-05,
+      "loss": 1.665,
+      "step": 539
+    },
+    {
+      "epoch": 0.43724696356275305,
+      "grad_norm": 1.0886967891950499,
+      "learning_rate": 1.9126107826868436e-05,
+      "loss": 1.6408,
+      "step": 540
+    },
+    {
+      "epoch": 0.4380566801619433,
+      "grad_norm": 1.209588887486826,
+      "learning_rate": 1.91225904321896e-05,
+      "loss": 1.7039,
+      "step": 541
+    },
+    {
+      "epoch": 0.4388663967611336,
+      "grad_norm": 1.140155316990926,
+      "learning_rate": 1.9119066297770924e-05,
+      "loss": 1.6966,
+      "step": 542
+    },
+    {
+      "epoch": 0.43967611336032386,
+      "grad_norm": 1.1285673550031636,
+      "learning_rate": 1.9115535426216018e-05,
+      "loss": 1.6644,
+      "step": 543
+    },
+    {
+      "epoch": 0.4404858299595142,
+      "grad_norm": 1.179973190293816,
+      "learning_rate": 1.9111997820133472e-05,
+      "loss": 1.7061,
+      "step": 544
+    },
+    {
+      "epoch": 0.44129554655870445,
+      "grad_norm": 1.308360305063932,
+      "learning_rate": 1.9108453482136866e-05,
+      "loss": 1.7163,
+      "step": 545
+    },
+    {
+      "epoch": 0.4421052631578947,
+      "grad_norm": 1.1712581078253976,
+      "learning_rate": 1.9104902414844746e-05,
+      "loss": 1.644,
+      "step": 546
+    },
+    {
+      "epoch": 0.44291497975708505,
+      "grad_norm": 1.1720636280682097,
+      "learning_rate": 1.9101344620880625e-05,
+      "loss": 1.7228,
+      "step": 547
+    },
+    {
+      "epoch": 0.4437246963562753,
+      "grad_norm": 1.2371895291825066,
+      "learning_rate": 1.909778010287299e-05,
+      "loss": 1.6121,
+      "step": 548
+    },
+    {
+      "epoch": 0.4445344129554656,
+      "grad_norm": 1.0466963526468558,
+      "learning_rate": 1.9094208863455296e-05,
+      "loss": 1.6056,
+      "step": 549
+    },
+    {
+      "epoch": 0.44534412955465585,
+      "grad_norm": 1.1539410237456056,
+      "learning_rate": 1.9090630905265963e-05,
+      "loss": 1.7385,
+      "step": 550
+    },
+    {
+      "epoch": 0.4461538461538462,
+      "grad_norm": 1.1690769087937798,
+      "learning_rate": 1.9087046230948373e-05,
+      "loss": 1.7135,
+      "step": 551
+    },
+    {
+      "epoch": 0.44696356275303645,
+      "grad_norm": 1.2143571533897817,
+      "learning_rate": 1.9083454843150875e-05,
+      "loss": 1.6558,
+      "step": 552
+    },
+    {
+      "epoch": 0.4477732793522267,
+      "grad_norm": 1.1034768392638734,
+      "learning_rate": 1.9079856744526775e-05,
+      "loss": 1.643,
+      "step": 553
+    },
+    {
+      "epoch": 0.448582995951417,
+      "grad_norm": 1.150409314573767,
+      "learning_rate": 1.9076251937734328e-05,
+      "loss": 1.6522,
+      "step": 554
+    },
+    {
+      "epoch": 0.4493927125506073,
+      "grad_norm": 1.2891397853585957,
+      "learning_rate": 1.9072640425436762e-05,
+      "loss": 1.6858,
+      "step": 555
+    },
+    {
+      "epoch": 0.4502024291497976,
+      "grad_norm": 1.2158215396752632,
+      "learning_rate": 1.906902221030225e-05,
+      "loss": 1.6065,
+      "step": 556
+    },
+    {
+      "epoch": 0.45101214574898785,
+      "grad_norm": 1.1200407349127612,
+      "learning_rate": 1.9065397295003917e-05,
+      "loss": 1.651,
+      "step": 557
+    },
+    {
+      "epoch": 0.4518218623481781,
+      "grad_norm": 1.1631363825604073,
+      "learning_rate": 1.9061765682219833e-05,
+      "loss": 1.6214,
+      "step": 558
+    },
+    {
+      "epoch": 0.45263157894736844,
+      "grad_norm": 1.1276692541556212,
+      "learning_rate": 1.9058127374633027e-05,
+      "loss": 1.6493,
+      "step": 559
+    },
+    {
+      "epoch": 0.4534412955465587,
+      "grad_norm": 1.1960342249167806,
+      "learning_rate": 1.905448237493147e-05,
+      "loss": 1.6425,
+      "step": 560
+    },
+    {
+      "epoch": 0.454251012145749,
+      "grad_norm": 1.1128837104785292,
+      "learning_rate": 1.905083068580807e-05,
+      "loss": 1.6333,
+      "step": 561
+    },
+    {
+      "epoch": 0.45506072874493925,
+      "grad_norm": 1.2256171630437762,
+      "learning_rate": 1.9047172309960685e-05,
+      "loss": 1.6754,
+      "step": 562
+    },
+    {
+      "epoch": 0.45587044534412957,
+      "grad_norm": 1.2397790580338823,
+      "learning_rate": 1.9043507250092113e-05,
+      "loss": 1.6666,
+      "step": 563
+    },
+    {
+      "epoch": 0.45668016194331984,
+      "grad_norm": 1.1082657174113733,
+      "learning_rate": 1.9039835508910086e-05,
+      "loss": 1.6955,
+      "step": 564
+    },
+    {
+      "epoch": 0.4574898785425101,
+      "grad_norm": 1.1906286541004671,
+      "learning_rate": 1.9036157089127278e-05,
+      "loss": 1.6531,
+      "step": 565
+    },
+    {
+      "epoch": 0.4582995951417004,
+      "grad_norm": 1.1424432978234285,
+      "learning_rate": 1.903247199346129e-05,
+      "loss": 1.7165,
+      "step": 566
+    },
+    {
+      "epoch": 0.4591093117408907,
+      "grad_norm": 1.209971661625088,
+      "learning_rate": 1.902878022463466e-05,
+      "loss": 1.7129,
+      "step": 567
+    },
+    {
+      "epoch": 0.45991902834008097,
+      "grad_norm": 1.1858172748323061,
+      "learning_rate": 1.9025081785374854e-05,
+      "loss": 1.6542,
+      "step": 568
+    },
+    {
+      "epoch": 0.46072874493927124,
+      "grad_norm": 1.133164376536185,
+      "learning_rate": 1.9021376678414266e-05,
+      "loss": 1.6183,
+      "step": 569
+    },
+    {
+      "epoch": 0.46153846153846156,
+      "grad_norm": 1.0890524812398936,
+      "learning_rate": 1.901766490649022e-05,
+      "loss": 1.6258,
+      "step": 570
+    },
+    {
+      "epoch": 0.46234817813765183,
+      "grad_norm": 1.2014659731558368,
+      "learning_rate": 1.901394647234496e-05,
+      "loss": 1.7053,
+      "step": 571
+    },
+    {
+      "epoch": 0.4631578947368421,
+      "grad_norm": 1.1482794624137653,
+      "learning_rate": 1.901022137872565e-05,
+      "loss": 1.6733,
+      "step": 572
+    },
+    {
+      "epoch": 0.46396761133603237,
+      "grad_norm": 1.1400965356949173,
+      "learning_rate": 1.9006489628384374e-05,
+      "loss": 1.6498,
+      "step": 573
+    },
+    {
+      "epoch": 0.4647773279352227,
+      "grad_norm": 1.1411852291997495,
+      "learning_rate": 1.9002751224078148e-05,
+      "loss": 1.6479,
+      "step": 574
+    },
+    {
+      "epoch": 0.46558704453441296,
+      "grad_norm": 1.1952888204227503,
+      "learning_rate": 1.8999006168568883e-05,
+      "loss": 1.6397,
+      "step": 575
+    },
+    {
+      "epoch": 0.46639676113360323,
+      "grad_norm": 1.1088918517492075,
+      "learning_rate": 1.899525446462342e-05,
+      "loss": 1.6055,
+      "step": 576
+    },
+    {
+      "epoch": 0.4672064777327935,
+      "grad_norm": 1.1036176426671134,
+      "learning_rate": 1.89914961150135e-05,
+      "loss": 1.5637,
+      "step": 577
+    },
+    {
+      "epoch": 0.4680161943319838,
+      "grad_norm": 1.2620277434980276,
+      "learning_rate": 1.8987731122515783e-05,
+      "loss": 1.7563,
+      "step": 578
+    },
+    {
+      "epoch": 0.4688259109311741,
+      "grad_norm": 1.2114730386349555,
+      "learning_rate": 1.8983959489911833e-05,
+      "loss": 1.6029,
+      "step": 579
+    },
+    {
+      "epoch": 0.46963562753036436,
+      "grad_norm": 1.1429824486613362,
+      "learning_rate": 1.8980181219988117e-05,
+      "loss": 1.6325,
+      "step": 580
+    },
+    {
+      "epoch": 0.47044534412955463,
+      "grad_norm": 1.2235645463897764,
+      "learning_rate": 1.897639631553601e-05,
+      "loss": 1.6529,
+      "step": 581
+    },
+    {
+      "epoch": 0.47125506072874496,
+      "grad_norm": 1.2114311834584224,
+      "learning_rate": 1.897260477935179e-05,
+      "loss": 1.7054,
+      "step": 582
+    },
+    {
+      "epoch": 0.4720647773279352,
+      "grad_norm": 1.166387438730835,
+      "learning_rate": 1.8968806614236625e-05,
+      "loss": 1.5569,
+      "step": 583
+    },
+    {
+      "epoch": 0.4728744939271255,
+      "grad_norm": 1.195523192072768,
+      "learning_rate": 1.8965001822996597e-05,
+      "loss": 1.6743,
+      "step": 584
+    },
+    {
+      "epoch": 0.47368421052631576,
+      "grad_norm": 1.1792702417522374,
+      "learning_rate": 1.8961190408442662e-05,
+      "loss": 1.5886,
+      "step": 585
+    },
+    {
+      "epoch": 0.4744939271255061,
+      "grad_norm": 1.2088013728568938,
+      "learning_rate": 1.8957372373390686e-05,
+      "loss": 1.7337,
+      "step": 586
+    },
+    {
+      "epoch": 0.47530364372469636,
+      "grad_norm": 1.1827861831578297,
+      "learning_rate": 1.895354772066142e-05,
+      "loss": 1.6138,
+      "step": 587
+    },
+    {
+      "epoch": 0.4761133603238866,
+      "grad_norm": 1.2447777229069645,
+      "learning_rate": 1.8949716453080508e-05,
+      "loss": 1.6402,
+      "step": 588
+    },
+    {
+      "epoch": 0.47692307692307695,
+      "grad_norm": 1.2650084378850481,
+      "learning_rate": 1.894587857347847e-05,
+      "loss": 1.7021,
+      "step": 589
+    },
+    {
+      "epoch": 0.4777327935222672,
+      "grad_norm": 1.1035611492696404,
+      "learning_rate": 1.8942034084690727e-05,
+      "loss": 1.666,
+      "step": 590
+    },
+    {
+      "epoch": 0.4785425101214575,
+      "grad_norm": 1.2449301610510557,
+      "learning_rate": 1.893818298955757e-05,
+      "loss": 1.6586,
+      "step": 591
+    },
+    {
+      "epoch": 0.47935222672064776,
+      "grad_norm": 1.233085689754001,
+      "learning_rate": 1.8934325290924177e-05,
+      "loss": 1.6828,
+      "step": 592
+    },
+    {
+      "epoch": 0.4801619433198381,
+      "grad_norm": 1.1810873725015931,
+      "learning_rate": 1.8930460991640606e-05,
+      "loss": 1.6581,
+      "step": 593
+    },
+    {
+      "epoch": 0.48097165991902835,
+      "grad_norm": 1.2288793304782808,
+      "learning_rate": 1.8926590094561784e-05,
+      "loss": 1.691,
+      "step": 594
+    },
+    {
+      "epoch": 0.4817813765182186,
+      "grad_norm": 1.2201380059232416,
+      "learning_rate": 1.8922712602547516e-05,
+      "loss": 1.6666,
+      "step": 595
+    },
+    {
+      "epoch": 0.4825910931174089,
+      "grad_norm": 1.145832667110585,
+      "learning_rate": 1.891882851846249e-05,
+      "loss": 1.678,
+      "step": 596
+    },
+    {
+      "epoch": 0.4834008097165992,
+      "grad_norm": 1.1830711402174343,
+      "learning_rate": 1.891493784517624e-05,
+      "loss": 1.6358,
+      "step": 597
+    },
+    {
+      "epoch": 0.4842105263157895,
+      "grad_norm": 1.2028619373981595,
+      "learning_rate": 1.8911040585563196e-05,
+      "loss": 1.7163,
+      "step": 598
+    },
+    {
+      "epoch": 0.48502024291497975,
+      "grad_norm": 1.1632991822145227,
+      "learning_rate": 1.8907136742502633e-05,
+      "loss": 1.7096,
+      "step": 599
+    },
+    {
+      "epoch": 0.48582995951417,
+      "grad_norm": 1.177024703746382,
+      "learning_rate": 1.89032263188787e-05,
+      "loss": 1.6444,
+      "step": 600
+    },
+    {
+      "epoch": 0.48663967611336034,
+      "grad_norm": 1.2097396232078657,
+      "learning_rate": 1.8899309317580403e-05,
+      "loss": 1.7032,
+      "step": 601
+    },
+    {
+      "epoch": 0.4874493927125506,
+      "grad_norm": 1.0312500629058605,
+      "learning_rate": 1.8895385741501608e-05,
+      "loss": 1.6393,
+      "step": 602
+    },
+    {
+      "epoch": 0.4882591093117409,
+      "grad_norm": 1.1817251113637202,
+      "learning_rate": 1.889145559354105e-05,
+      "loss": 1.6357,
+      "step": 603
+    },
+    {
+      "epoch": 0.48906882591093115,
+      "grad_norm": 1.192720416754525,
+      "learning_rate": 1.88875188766023e-05,
+      "loss": 1.6244,
+      "step": 604
+    },
+    {
+      "epoch": 0.4898785425101215,
+      "grad_norm": 1.1432510606242128,
+      "learning_rate": 1.8883575593593793e-05,
+      "loss": 1.6741,
+      "step": 605
+    },
+    {
+      "epoch": 0.49068825910931174,
+      "grad_norm": 1.1569463862557519,
+      "learning_rate": 1.8879625747428815e-05,
+      "loss": 1.6761,
+      "step": 606
+    },
+    {
+      "epoch": 0.491497975708502,
+      "grad_norm": 1.187603448727291,
+      "learning_rate": 1.8875669341025498e-05,
+      "loss": 1.728,
+      "step": 607
+    },
+    {
+      "epoch": 0.49230769230769234,
+      "grad_norm": 1.2130166999284056,
+      "learning_rate": 1.8871706377306826e-05,
+      "loss": 1.6865,
+      "step": 608
+    },
+    {
+      "epoch": 0.4931174089068826,
+      "grad_norm": 1.1670591496027172,
+      "learning_rate": 1.886773685920062e-05,
+      "loss": 1.6744,
+      "step": 609
+    },
+    {
+      "epoch": 0.4939271255060729,
+      "grad_norm": 1.2163771157473018,
+      "learning_rate": 1.8863760789639548e-05,
+      "loss": 1.7203,
+      "step": 610
+    },
+    {
+      "epoch": 0.49473684210526314,
+      "grad_norm": 1.229371456590416,
+      "learning_rate": 1.8859778171561118e-05,
+      "loss": 1.7059,
+      "step": 611
+    },
+    {
+      "epoch": 0.49554655870445347,
+      "grad_norm": 1.192832189327257,
+      "learning_rate": 1.8855789007907672e-05,
+      "loss": 1.6862,
+      "step": 612
+    },
+    {
+      "epoch": 0.49635627530364373,
+      "grad_norm": 1.2220933781088503,
+      "learning_rate": 1.885179330162639e-05,
+      "loss": 1.6475,
+      "step": 613
+    },
+    {
+      "epoch": 0.497165991902834,
+      "grad_norm": 1.1263582806182404,
+      "learning_rate": 1.8847791055669297e-05,
+      "loss": 1.5831,
+      "step": 614
+    },
+    {
+      "epoch": 0.4979757085020243,
+      "grad_norm": 1.2054405676608149,
+      "learning_rate": 1.8843782272993225e-05,
+      "loss": 1.6485,
+      "step": 615
+    },
+    {
+      "epoch": 0.4987854251012146,
+      "grad_norm": 1.1513573612013865,
+      "learning_rate": 1.883976695655986e-05,
+      "loss": 1.6391,
+      "step": 616
+    },
+    {
+      "epoch": 0.49959514170040487,
+      "grad_norm": 1.1215561830277645,
+      "learning_rate": 1.88357451093357e-05,
+      "loss": 1.689,
+      "step": 617
+    },
+    {
+      "epoch": 0.5004048582995951,
+      "grad_norm": 1.2642509822675319,
+      "learning_rate": 1.8831716734292074e-05,
+      "loss": 1.6564,
+      "step": 618
+    },
+    {
+      "epoch": 0.5012145748987854,
+      "grad_norm": 1.1935335114428585,
+      "learning_rate": 1.882768183440513e-05,
+      "loss": 1.6118,
+      "step": 619
+    },
+    {
+      "epoch": 0.5020242914979757,
+      "grad_norm": 1.2850491601729415,
+      "learning_rate": 1.8823640412655844e-05,
+      "loss": 1.652,
+      "step": 620
+    },
+    {
+      "epoch": 0.5028340080971659,
+      "grad_norm": 1.202299368491898,
+      "learning_rate": 1.881959247203e-05,
+      "loss": 1.6723,
+      "step": 621
+    },
+    {
+      "epoch": 0.5036437246963563,
+      "grad_norm": 1.0583336420886984,
+      "learning_rate": 1.8815538015518203e-05,
+      "loss": 1.6403,
+      "step": 622
+    },
+    {
+      "epoch": 0.5044534412955466,
+      "grad_norm": 1.24127140876251,
+      "learning_rate": 1.8811477046115877e-05,
+      "loss": 1.692,
+      "step": 623
+    },
+    {
+      "epoch": 0.5052631578947369,
+      "grad_norm": 1.174302573947688,
+      "learning_rate": 1.880740956682325e-05,
+      "loss": 1.7306,
+      "step": 624
+    },
+    {
+      "epoch": 0.5060728744939271,
+      "grad_norm": 1.1958655509103453,
+      "learning_rate": 1.880333558064536e-05,
+      "loss": 1.7106,
+      "step": 625
+    },
+    {
+      "epoch": 0.5068825910931174,
+      "grad_norm": 1.230409287496346,
+      "learning_rate": 1.8799255090592056e-05,
+      "loss": 1.6266,
+      "step": 626
+    },
+    {
+      "epoch": 0.5076923076923077,
+      "grad_norm": 1.1988542561396345,
+      "learning_rate": 1.8795168099677992e-05,
+      "loss": 1.6516,
+      "step": 627
+    },
+    {
+      "epoch": 0.5085020242914979,
+      "grad_norm": 1.1887473097347059,
+      "learning_rate": 1.8791074610922624e-05,
+      "loss": 1.7295,
+      "step": 628
+    },
+    {
+      "epoch": 0.5093117408906883,
+      "grad_norm": 1.3303290476414642,
+      "learning_rate": 1.8786974627350206e-05,
+      "loss": 1.6802,
+      "step": 629
+    },
+    {
+      "epoch": 0.5101214574898786,
+      "grad_norm": 1.1121862887308984,
+      "learning_rate": 1.878286815198979e-05,
+      "loss": 1.6759,
+      "step": 630
+    },
+    {
+      "epoch": 0.5109311740890689,
+      "grad_norm": 1.1527683977442384,
+      "learning_rate": 1.8778755187875236e-05,
+      "loss": 1.675,
+      "step": 631
+    },
+    {
+      "epoch": 0.5117408906882591,
+      "grad_norm": 1.1283416836021556,
+      "learning_rate": 1.877463573804518e-05,
+      "loss": 1.7042,
+      "step": 632
+    },
+    {
+      "epoch": 0.5125506072874494,
+      "grad_norm": 1.111511552571126,
+      "learning_rate": 1.877050980554306e-05,
+      "loss": 1.6402,
+      "step": 633
+    },
+    {
+      "epoch": 0.5133603238866397,
+      "grad_norm": 1.1686794783887113,
+      "learning_rate": 1.8766377393417104e-05,
+      "loss": 1.6266,
+      "step": 634
+    },
+    {
+      "epoch": 0.5141700404858299,
+      "grad_norm": 1.0774322060832857,
+      "learning_rate": 1.876223850472032e-05,
+      "loss": 1.6109,
+      "step": 635
+    },
+    {
+      "epoch": 0.5149797570850202,
+      "grad_norm": 1.1204612114124226,
+      "learning_rate": 1.875809314251051e-05,
+      "loss": 1.5896,
+      "step": 636
+    },
+    {
+      "epoch": 0.5157894736842106,
+      "grad_norm": 1.1411930639576617,
+      "learning_rate": 1.8753941309850248e-05,
+      "loss": 1.6531,
+      "step": 637
+    },
+    {
+      "epoch": 0.5165991902834008,
+      "grad_norm": 1.246386803807688,
+      "learning_rate": 1.8749783009806898e-05,
+      "loss": 1.7025,
+      "step": 638
+    },
+    {
+      "epoch": 0.5174089068825911,
+      "grad_norm": 1.176836096422221,
+      "learning_rate": 1.8745618245452596e-05,
+      "loss": 1.6579,
+      "step": 639
+    },
+    {
+      "epoch": 0.5182186234817814,
+      "grad_norm": 1.10956833415483,
+      "learning_rate": 1.8741447019864263e-05,
+      "loss": 1.6245,
+      "step": 640
+    },
+    {
+      "epoch": 0.5190283400809717,
+      "grad_norm": 1.1188976219487057,
+      "learning_rate": 1.873726933612358e-05,
+      "loss": 1.6296,
+      "step": 641
+    },
+    {
+      "epoch": 0.5198380566801619,
+      "grad_norm": 1.2300857210684373,
+      "learning_rate": 1.873308519731701e-05,
+      "loss": 1.6823,
+      "step": 642
+    },
+    {
+      "epoch": 0.5206477732793522,
+      "grad_norm": 1.1437764748313324,
+      "learning_rate": 1.872889460653578e-05,
+      "loss": 1.6442,
+      "step": 643
+    },
+    {
+      "epoch": 0.5214574898785425,
+      "grad_norm": 1.1135134095150818,
+      "learning_rate": 1.872469756687588e-05,
+      "loss": 1.5917,
+      "step": 644
+    },
+    {
+      "epoch": 0.5222672064777328,
+      "grad_norm": 1.1837475043251595,
+      "learning_rate": 1.872049408143808e-05,
+      "loss": 1.7025,
+      "step": 645
+    },
+    {
+      "epoch": 0.5230769230769231,
+      "grad_norm": 1.1575973920612108,
+      "learning_rate": 1.8716284153327887e-05,
+      "loss": 1.6407,
+      "step": 646
+    },
+    {
+      "epoch": 0.5238866396761134,
+      "grad_norm": 1.1699024098280943,
+      "learning_rate": 1.871206778565559e-05,
+      "loss": 1.6838,
+      "step": 647
+    },
+    {
+      "epoch": 0.5246963562753036,
+      "grad_norm": 1.1006692187902312,
+      "learning_rate": 1.870784498153623e-05,
+      "loss": 1.6798,
+      "step": 648
+    },
+    {
+      "epoch": 0.5255060728744939,
+      "grad_norm": 1.1232448009893263,
+      "learning_rate": 1.87036157440896e-05,
+      "loss": 1.6862,
+      "step": 649
+    },
+    {
+      "epoch": 0.5263157894736842,
+      "grad_norm": 1.1938575522143917,
+      "learning_rate": 1.8699380076440242e-05,
+      "loss": 1.7397,
+      "step": 650
+    },
+    {
+      "epoch": 0.5271255060728745,
+      "grad_norm": 1.1943224317319683,
+      "learning_rate": 1.8695137981717452e-05,
+      "loss": 1.6201,
+      "step": 651
+    },
+    {
+      "epoch": 0.5279352226720648,
+      "grad_norm": 1.1217940510846476,
+      "learning_rate": 1.8690889463055285e-05,
+      "loss": 1.637,
+      "step": 652
+    },
+    {
+      "epoch": 0.5287449392712551,
+      "grad_norm": 1.1039437922379434,
+      "learning_rate": 1.8686634523592523e-05,
+      "loss": 1.6459,
+      "step": 653
+    },
+    {
+      "epoch": 0.5295546558704454,
+      "grad_norm": 1.1566442258336185,
+      "learning_rate": 1.868237316647271e-05,
+      "loss": 1.669,
+      "step": 654
+    },
+    {
+      "epoch": 0.5303643724696356,
+      "grad_norm": 1.2402960513081014,
+      "learning_rate": 1.8678105394844114e-05,
+      "loss": 1.6826,
+      "step": 655
+    },
+    {
+      "epoch": 0.5311740890688259,
+      "grad_norm": 1.1427219527507249,
+      "learning_rate": 1.8673831211859758e-05,
+      "loss": 1.6865,
+      "step": 656
+    },
+    {
+      "epoch": 0.5319838056680162,
+      "grad_norm": 1.1397055186220963,
+      "learning_rate": 1.866955062067739e-05,
+      "loss": 1.6982,
+      "step": 657
+    },
+    {
+      "epoch": 0.5327935222672064,
+      "grad_norm": 1.1723541895025018,
+      "learning_rate": 1.8665263624459497e-05,
+      "loss": 1.6186,
+      "step": 658
+    },
+    {
+      "epoch": 0.5336032388663967,
+      "grad_norm": 1.0732319910985701,
+      "learning_rate": 1.86609702263733e-05,
+      "loss": 1.7125,
+      "step": 659
+    },
+    {
+      "epoch": 0.5344129554655871,
+      "grad_norm": 1.102482299703252,
+      "learning_rate": 1.8656670429590745e-05,
+      "loss": 1.7209,
+      "step": 660
+    },
+    {
+      "epoch": 0.5352226720647774,
+      "grad_norm": 1.1656759203730942,
+      "learning_rate": 1.8652364237288507e-05,
+      "loss": 1.6379,
+      "step": 661
+    },
+    {
+      "epoch": 0.5360323886639676,
+      "grad_norm": 1.1796882306648648,
+      "learning_rate": 1.864805165264799e-05,
+      "loss": 1.6733,
+      "step": 662
+    },
+    {
+      "epoch": 0.5368421052631579,
+      "grad_norm": 1.0979399574684228,
+      "learning_rate": 1.8643732678855314e-05,
+      "loss": 1.6439,
+      "step": 663
+    },
+    {
+      "epoch": 0.5376518218623482,
+      "grad_norm": 1.1656259629485042,
+      "learning_rate": 1.8639407319101325e-05,
+      "loss": 1.7003,
+      "step": 664
+    },
+    {
+      "epoch": 0.5384615384615384,
+      "grad_norm": 1.109949494450682,
+      "learning_rate": 1.8635075576581587e-05,
+      "loss": 1.6913,
+      "step": 665
+    },
+    {
+      "epoch": 0.5392712550607287,
+      "grad_norm": 1.1384719929801121,
+      "learning_rate": 1.8630737454496374e-05,
+      "loss": 1.6767,
+      "step": 666
+    },
+    {
+      "epoch": 0.540080971659919,
+      "grad_norm": 1.1146510327537758,
+      "learning_rate": 1.8626392956050675e-05,
+      "loss": 1.6193,
+      "step": 667
+    },
+    {
+      "epoch": 0.5408906882591094,
+      "grad_norm": 1.1533020616397756,
+      "learning_rate": 1.862204208445419e-05,
+      "loss": 1.6373,
+      "step": 668
+    },
+    {
+      "epoch": 0.5417004048582996,
+      "grad_norm": 1.1409514813984398,
+      "learning_rate": 1.8617684842921337e-05,
+      "loss": 1.5814,
+      "step": 669
+    },
+    {
+      "epoch": 0.5425101214574899,
+      "grad_norm": 1.1892690565319142,
+      "learning_rate": 1.861332123467122e-05,
+      "loss": 1.71,
+      "step": 670
+    },
+    {
+      "epoch": 0.5433198380566802,
+      "grad_norm": 1.0770444731078834,
+      "learning_rate": 1.8608951262927667e-05,
+      "loss": 1.6821,
+      "step": 671
+    },
+    {
+      "epoch": 0.5441295546558704,
+      "grad_norm": 1.2201808284178386,
+      "learning_rate": 1.8604574930919198e-05,
+      "loss": 1.608,
+      "step": 672
+    },
+    {
+      "epoch": 0.5449392712550607,
+      "grad_norm": 1.1156767804046464,
+      "learning_rate": 1.860019224187903e-05,
+      "loss": 1.6091,
+      "step": 673
+    },
+    {
+      "epoch": 0.545748987854251,
+      "grad_norm": 1.1312661674531492,
+      "learning_rate": 1.8595803199045083e-05,
+      "loss": 1.6964,
+      "step": 674
+    },
+    {
+      "epoch": 0.5465587044534413,
+      "grad_norm": 1.2293348497150374,
+      "learning_rate": 1.859140780565996e-05,
+      "loss": 1.6872,
+      "step": 675
+    },
+    {
+      "epoch": 0.5473684210526316,
+      "grad_norm": 1.062664670461368,
+      "learning_rate": 1.858700606497097e-05,
+      "loss": 1.6528,
+      "step": 676
+    },
+    {
+      "epoch": 0.5481781376518219,
+      "grad_norm": 1.1400817362492008,
+      "learning_rate": 1.85825979802301e-05,
+      "loss": 1.6562,
+      "step": 677
+    },
+    {
+      "epoch": 0.5489878542510122,
+      "grad_norm": 1.0787086213793413,
+      "learning_rate": 1.8578183554694035e-05,
+      "loss": 1.7133,
+      "step": 678
+    },
+    {
+      "epoch": 0.5497975708502024,
+      "grad_norm": 1.0952800726386747,
+      "learning_rate": 1.8573762791624132e-05,
+      "loss": 1.6316,
+      "step": 679
+    },
+    {
+      "epoch": 0.5506072874493927,
+      "grad_norm": 1.140497018650844,
+      "learning_rate": 1.856933569428644e-05,
+      "loss": 1.5434,
+      "step": 680
+    },
+    {
+      "epoch": 0.551417004048583,
+      "grad_norm": 1.155661350966275,
+      "learning_rate": 1.856490226595168e-05,
+      "loss": 1.7302,
+      "step": 681
+    },
+    {
+      "epoch": 0.5522267206477732,
+      "grad_norm": 1.162518892367102,
+      "learning_rate": 1.856046250989526e-05,
+      "loss": 1.7335,
+      "step": 682
+    },
+    {
+      "epoch": 0.5530364372469636,
+      "grad_norm": 1.1661792099551183,
+      "learning_rate": 1.8556016429397248e-05,
+      "loss": 1.6408,
+      "step": 683
+    },
+    {
+      "epoch": 0.5538461538461539,
+      "grad_norm": 1.129401750350817,
+      "learning_rate": 1.8551564027742404e-05,
+      "loss": 1.6326,
+      "step": 684
+    },
+    {
+      "epoch": 0.5546558704453441,
+      "grad_norm": 1.2694837443603195,
+      "learning_rate": 1.8547105308220142e-05,
+      "loss": 1.6501,
+      "step": 685
+    },
+    {
+      "epoch": 0.5554655870445344,
+      "grad_norm": 1.0791347009798566,
+      "learning_rate": 1.854264027412455e-05,
+      "loss": 1.6494,
+      "step": 686
+    },
+    {
+      "epoch": 0.5562753036437247,
+      "grad_norm": 1.2420182578602168,
+      "learning_rate": 1.853816892875438e-05,
+      "loss": 1.6413,
+      "step": 687
+    },
+    {
+      "epoch": 0.557085020242915,
+      "grad_norm": 1.2039522453979503,
+      "learning_rate": 1.853369127541305e-05,
+      "loss": 1.639,
+      "step": 688
+    },
+    {
+      "epoch": 0.5578947368421052,
+      "grad_norm": 1.136925790043882,
+      "learning_rate": 1.8529207317408634e-05,
+      "loss": 1.6843,
+      "step": 689
+    },
+    {
+      "epoch": 0.5587044534412956,
+      "grad_norm": 1.2928638734791758,
+      "learning_rate": 1.852471705805387e-05,
+      "loss": 1.5799,
+      "step": 690
+    },
+    {
+      "epoch": 0.5595141700404859,
+      "grad_norm": 1.0426663613551523,
+      "learning_rate": 1.8520220500666133e-05,
+      "loss": 1.6005,
+      "step": 691
+    },
+    {
+      "epoch": 0.5603238866396761,
+      "grad_norm": 1.3130919141243858,
+      "learning_rate": 1.8515717648567476e-05,
+      "loss": 1.7198,
+      "step": 692
+    },
+    {
+      "epoch": 0.5611336032388664,
+      "grad_norm": 1.0955024598419019,
+      "learning_rate": 1.8511208505084593e-05,
+      "loss": 1.6197,
+      "step": 693
+    },
+    {
+      "epoch": 0.5619433198380567,
+      "grad_norm": 1.0883596134382716,
+      "learning_rate": 1.850669307354882e-05,
+      "loss": 1.6603,
+      "step": 694
+    },
+    {
+      "epoch": 0.562753036437247,
+      "grad_norm": 1.218741988618394,
+      "learning_rate": 1.8502171357296144e-05,
+      "loss": 1.6684,
+      "step": 695
+    },
+    {
+      "epoch": 0.5635627530364372,
+      "grad_norm": 1.1391537110795147,
+      "learning_rate": 1.8497643359667193e-05,
+      "loss": 1.6579,
+      "step": 696
+    },
+    {
+      "epoch": 0.5643724696356275,
+      "grad_norm": 1.1387711931236417,
+      "learning_rate": 1.8493109084007236e-05,
+      "loss": 1.6465,
+      "step": 697
+    },
+    {
+      "epoch": 0.5651821862348179,
+      "grad_norm": 1.1254009149677715,
+      "learning_rate": 1.8488568533666183e-05,
+      "loss": 1.6042,
+      "step": 698
+    },
+    {
+      "epoch": 0.5659919028340081,
+      "grad_norm": 1.1257268411329382,
+      "learning_rate": 1.848402171199858e-05,
+      "loss": 1.6514,
+      "step": 699
+    },
+    {
+      "epoch": 0.5668016194331984,
+      "grad_norm": 1.0927876479165208,
+      "learning_rate": 1.84794686223636e-05,
+      "loss": 1.6822,
+      "step": 700
+    },
+    {
+      "epoch": 0.5676113360323887,
+      "grad_norm": 1.090171472139705,
+      "learning_rate": 1.8474909268125053e-05,
+      "loss": 1.7033,
+      "step": 701
+    },
+    {
+      "epoch": 0.5684210526315789,
+      "grad_norm": 1.088693023569449,
+      "learning_rate": 1.8470343652651374e-05,
+      "loss": 1.615,
+      "step": 702
+    },
+    {
+      "epoch": 0.5692307692307692,
+      "grad_norm": 1.3432139320476364,
+      "learning_rate": 1.846577177931562e-05,
+      "loss": 1.6546,
+      "step": 703
+    },
+    {
+      "epoch": 0.5700404858299595,
+      "grad_norm": 1.0705730809837788,
+      "learning_rate": 1.8461193651495482e-05,
+      "loss": 1.6723,
+      "step": 704
+    },
+    {
+      "epoch": 0.5708502024291497,
+      "grad_norm": 1.121041236995467,
+      "learning_rate": 1.8456609272573268e-05,
+      "loss": 1.5757,
+      "step": 705
+    },
+    {
+      "epoch": 0.5716599190283401,
+      "grad_norm": 1.2097795973645553,
+      "learning_rate": 1.8452018645935895e-05,
+      "loss": 1.6559,
+      "step": 706
+    },
+    {
+      "epoch": 0.5724696356275304,
+      "grad_norm": 1.1312684155340222,
+      "learning_rate": 1.844742177497491e-05,
+      "loss": 1.6322,
+      "step": 707
+    },
+    {
+      "epoch": 0.5732793522267207,
+      "grad_norm": 1.126110922221143,
+      "learning_rate": 1.8442818663086456e-05,
+      "loss": 1.6482,
+      "step": 708
+    },
+    {
+      "epoch": 0.5740890688259109,
+      "grad_norm": 1.1010830873261865,
+      "learning_rate": 1.8438209313671307e-05,
+      "loss": 1.6829,
+      "step": 709
+    },
+    {
+      "epoch": 0.5748987854251012,
+      "grad_norm": 1.2743855652131832,
+      "learning_rate": 1.8433593730134835e-05,
+      "loss": 1.669,
+      "step": 710
+    },
+    {
+      "epoch": 0.5757085020242915,
+      "grad_norm": 1.1448636929742997,
+      "learning_rate": 1.842897191588701e-05,
+      "loss": 1.692,
+      "step": 711
+    },
+    {
+      "epoch": 0.5765182186234817,
+      "grad_norm": 1.1106904762427132,
+      "learning_rate": 1.842434387434242e-05,
+      "loss": 1.6477,
+      "step": 712
+    },
+    {
+      "epoch": 0.5773279352226721,
+      "grad_norm": 1.2682672769652985,
+      "learning_rate": 1.8419709608920243e-05,
+      "loss": 1.6194,
+      "step": 713
+    },
+    {
+      "epoch": 0.5781376518218624,
+      "grad_norm": 1.2002428737470912,
+      "learning_rate": 1.8415069123044263e-05,
+      "loss": 1.6244,
+      "step": 714
+    },
+    {
+      "epoch": 0.5789473684210527,
+      "grad_norm": 1.105036420479548,
+      "learning_rate": 1.841042242014285e-05,
+      "loss": 1.6496,
+      "step": 715
+    },
+    {
+      "epoch": 0.5797570850202429,
+      "grad_norm": 1.1794855208133117,
+      "learning_rate": 1.840576950364898e-05,
+      "loss": 1.6911,
+      "step": 716
+    },
+    {
+      "epoch": 0.5805668016194332,
+      "grad_norm": 1.1055378048289626,
+      "learning_rate": 1.8401110377000206e-05,
+      "loss": 1.6052,
+      "step": 717
+    },
+    {
+      "epoch": 0.5813765182186235,
+      "grad_norm": 1.2287534777350115,
+      "learning_rate": 1.839644504363868e-05,
+      "loss": 1.6905,
+      "step": 718
+    },
+    {
+      "epoch": 0.5821862348178137,
+      "grad_norm": 1.1569748526557413,
+      "learning_rate": 1.839177350701113e-05,
+      "loss": 1.6255,
+      "step": 719
+    },
+    {
+      "epoch": 0.582995951417004,
+      "grad_norm": 1.1818168990284688,
+      "learning_rate": 1.838709577056888e-05,
+      "loss": 1.5795,
+      "step": 720
+    },
+    {
+      "epoch": 0.5838056680161944,
+      "grad_norm": 1.1555576007146946,
+      "learning_rate": 1.838241183776782e-05,
+      "loss": 1.6674,
+      "step": 721
+    },
+    {
+      "epoch": 0.5846153846153846,
+      "grad_norm": 1.1401390602412884,
+      "learning_rate": 1.8377721712068424e-05,
+      "loss": 1.675,
+      "step": 722
+    },
+    {
+      "epoch": 0.5854251012145749,
+      "grad_norm": 1.162723139199304,
+      "learning_rate": 1.8373025396935743e-05,
+      "loss": 1.5534,
+      "step": 723
+    },
+    {
+      "epoch": 0.5862348178137652,
+      "grad_norm": 1.1799321705036754,
+      "learning_rate": 1.8368322895839397e-05,
+      "loss": 1.639,
+      "step": 724
+    },
+    {
+      "epoch": 0.5870445344129555,
+      "grad_norm": 1.2240975875375573,
+      "learning_rate": 1.8363614212253585e-05,
+      "loss": 1.597,
+      "step": 725
+    },
+    {
+      "epoch": 0.5878542510121457,
+      "grad_norm": 1.1698625227265869,
+      "learning_rate": 1.8358899349657063e-05,
+      "loss": 1.7323,
+      "step": 726
+    },
+    {
+      "epoch": 0.588663967611336,
+      "grad_norm": 1.1339957776380405,
+      "learning_rate": 1.8354178311533152e-05,
+      "loss": 1.6183,
+      "step": 727
+    },
+    {
+      "epoch": 0.5894736842105263,
+      "grad_norm": 1.2392896642339328,
+      "learning_rate": 1.8349451101369742e-05,
+      "loss": 1.6272,
+      "step": 728
+    },
+    {
+      "epoch": 0.5902834008097166,
+      "grad_norm": 1.1522226881000637,
+      "learning_rate": 1.8344717722659285e-05,
+      "loss": 1.6365,
+      "step": 729
+    },
+    {
+      "epoch": 0.5910931174089069,
+      "grad_norm": 1.2425239844288822,
+      "learning_rate": 1.833997817889878e-05,
+      "loss": 1.5836,
+      "step": 730
+    },
+    {
+      "epoch": 0.5919028340080972,
+      "grad_norm": 1.1120981246598145,
+      "learning_rate": 1.833523247358979e-05,
+      "loss": 1.6734,
+      "step": 731
+    },
+    {
+      "epoch": 0.5927125506072874,
+      "grad_norm": 1.1298710482715226,
+      "learning_rate": 1.8330480610238424e-05,
+      "loss": 1.607,
+      "step": 732
+    },
+    {
+      "epoch": 0.5935222672064777,
+      "grad_norm": 1.1377574517661373,
+      "learning_rate": 1.8325722592355344e-05,
+      "loss": 1.6454,
+      "step": 733
+    },
+    {
+      "epoch": 0.594331983805668,
+      "grad_norm": 1.1538583104420033,
+      "learning_rate": 1.8320958423455756e-05,
+      "loss": 1.6921,
+      "step": 734
+    },
+    {
+      "epoch": 0.5951417004048583,
+      "grad_norm": 1.0793087009627134,
+      "learning_rate": 1.8316188107059418e-05,
+      "loss": 1.6726,
+      "step": 735
+    },
+    {
+      "epoch": 0.5959514170040486,
+      "grad_norm": 1.1201444715810196,
+      "learning_rate": 1.8311411646690616e-05,
+      "loss": 1.6588,
+      "step": 736
+    },
+    {
+      "epoch": 0.5967611336032389,
+      "grad_norm": 1.096654674640351,
+      "learning_rate": 1.8306629045878192e-05,
+      "loss": 1.6684,
+      "step": 737
+    },
+    {
+      "epoch": 0.5975708502024292,
+      "grad_norm": 1.1079216672953291,
+      "learning_rate": 1.8301840308155507e-05,
+      "loss": 1.6618,
+      "step": 738
+    },
+    {
+      "epoch": 0.5983805668016194,
+      "grad_norm": 1.0879802564948844,
+      "learning_rate": 1.8297045437060474e-05,
+      "loss": 1.7383,
+      "step": 739
+    },
+    {
+      "epoch": 0.5991902834008097,
+      "grad_norm": 1.1424198121843356,
+      "learning_rate": 1.8292244436135517e-05,
+      "loss": 1.6709,
+      "step": 740
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.1430964238759362,
+      "learning_rate": 1.828743730892761e-05,
+      "loss": 1.64,
+      "step": 741
+    },
+    {
+      "epoch": 0.6008097165991902,
+      "grad_norm": 1.0895066483885778,
+      "learning_rate": 1.8282624058988237e-05,
+      "loss": 1.6584,
+      "step": 742
+    },
+    {
+      "epoch": 0.6016194331983805,
+      "grad_norm": 1.1291234703082627,
+      "learning_rate": 1.827780468987341e-05,
+      "loss": 1.6452,
+      "step": 743
+    },
+    {
+      "epoch": 0.6024291497975709,
+      "grad_norm": 1.115557218075291,
+      "learning_rate": 1.8272979205143674e-05,
+      "loss": 1.6076,
+      "step": 744
+    },
+    {
+      "epoch": 0.6032388663967612,
+      "grad_norm": 1.1171405856483771,
+      "learning_rate": 1.8268147608364068e-05,
+      "loss": 1.6046,
+      "step": 745
+    },
+    {
+      "epoch": 0.6040485829959514,
+      "grad_norm": 1.0965261272427589,
+      "learning_rate": 1.8263309903104163e-05,
+      "loss": 1.6013,
+      "step": 746
+    },
+    {
+      "epoch": 0.6048582995951417,
+      "grad_norm": 1.2185450067830268,
+      "learning_rate": 1.8258466092938042e-05,
+      "loss": 1.6519,
+      "step": 747
+    },
+    {
+      "epoch": 0.605668016194332,
+      "grad_norm": 1.1702812963364408,
+      "learning_rate": 1.82536161814443e-05,
+      "loss": 1.6953,
+      "step": 748
+    },
+    {
+      "epoch": 0.6064777327935222,
+      "grad_norm": 1.0686726410287588,
+      "learning_rate": 1.8248760172206024e-05,
+      "loss": 1.6262,
+      "step": 749
+    },
+    {
+      "epoch": 0.6072874493927125,
+      "grad_norm": 1.136164025439449,
+      "learning_rate": 1.8243898068810833e-05,
+      "loss": 1.6661,
+      "step": 750
+    },
+    {
+      "epoch": 0.6080971659919029,
+      "grad_norm": 1.1954715351994403,
+      "learning_rate": 1.8239029874850823e-05,
+      "loss": 1.674,
+      "step": 751
+    },
+    {
+      "epoch": 0.6089068825910932,
+      "grad_norm": 1.1184558352613179,
+      "learning_rate": 1.82341555939226e-05,
+      "loss": 1.721,
+      "step": 752
+    },
+    {
+      "epoch": 0.6097165991902834,
+      "grad_norm": 1.129119070175351,
+      "learning_rate": 1.822927522962727e-05,
+      "loss": 1.6457,
+      "step": 753
+    },
+    {
+      "epoch": 0.6105263157894737,
+      "grad_norm": 1.1410321623438338,
+      "learning_rate": 1.822438878557043e-05,
+      "loss": 1.6529,
+      "step": 754
+    },
+    {
+      "epoch": 0.611336032388664,
+      "grad_norm": 1.1589440461468792,
+      "learning_rate": 1.8219496265362164e-05,
+      "loss": 1.7046,
+      "step": 755
+    },
+    {
+      "epoch": 0.6121457489878542,
+      "grad_norm": 1.115213970991795,
+      "learning_rate": 1.8214597672617054e-05,
+      "loss": 1.7141,
+      "step": 756
+    },
+    {
+      "epoch": 0.6129554655870445,
+      "grad_norm": 1.1739020879997546,
+      "learning_rate": 1.8209693010954166e-05,
+      "loss": 1.6513,
+      "step": 757
+    },
+    {
+      "epoch": 0.6137651821862348,
+      "grad_norm": 1.1420116503362077,
+      "learning_rate": 1.820478228399704e-05,
+      "loss": 1.6475,
+      "step": 758
+    },
+    {
+      "epoch": 0.6145748987854251,
+      "grad_norm": 1.2176203227373104,
+      "learning_rate": 1.819986549537372e-05,
+      "loss": 1.6648,
+      "step": 759
+    },
+    {
+      "epoch": 0.6153846153846154,
+      "grad_norm": 1.1052645276051432,
+      "learning_rate": 1.81949426487167e-05,
+      "loss": 1.6193,
+      "step": 760
+    },
+    {
+      "epoch": 0.6161943319838057,
+      "grad_norm": 1.099724825808027,
+      "learning_rate": 1.819001374766296e-05,
+      "loss": 1.5659,
+      "step": 761
+    },
+    {
+      "epoch": 0.617004048582996,
+      "grad_norm": 1.255382813893447,
+      "learning_rate": 1.818507879585397e-05,
+      "loss": 1.6868,
+      "step": 762
+    },
+    {
+      "epoch": 0.6178137651821862,
+      "grad_norm": 1.1337123092988888,
+      "learning_rate": 1.8180137796935648e-05,
+      "loss": 1.7117,
+      "step": 763
+    },
+    {
+      "epoch": 0.6186234817813765,
+      "grad_norm": 1.0916001764698793,
+      "learning_rate": 1.8175190754558384e-05,
+      "loss": 1.6079,
+      "step": 764
+    },
+    {
+      "epoch": 0.6194331983805668,
+      "grad_norm": 1.1462547249605433,
+      "learning_rate": 1.8170237672377046e-05,
+      "loss": 1.6266,
+      "step": 765
+    },
+    {
+      "epoch": 0.620242914979757,
+      "grad_norm": 1.184448821537451,
+      "learning_rate": 1.8165278554050946e-05,
+      "loss": 1.7218,
+      "step": 766
+    },
+    {
+      "epoch": 0.6210526315789474,
+      "grad_norm": 1.1988569585117117,
+      "learning_rate": 1.8160313403243874e-05,
+      "loss": 1.7002,
+      "step": 767
+    },
+    {
+      "epoch": 0.6218623481781377,
+      "grad_norm": 1.1270467979508891,
+      "learning_rate": 1.8155342223624054e-05,
+      "loss": 1.6241,
+      "step": 768
+    },
+    {
+      "epoch": 0.622672064777328,
+      "grad_norm": 1.0627366879538067,
+      "learning_rate": 1.8150365018864192e-05,
+      "loss": 1.63,
+      "step": 769
+    },
+    {
+      "epoch": 0.6234817813765182,
+      "grad_norm": 1.1499690148470678,
+      "learning_rate": 1.814538179264142e-05,
+      "loss": 1.6664,
+      "step": 770
+    },
+    {
+      "epoch": 0.6242914979757085,
+      "grad_norm": 1.1381510719594987,
+      "learning_rate": 1.8140392548637333e-05,
+      "loss": 1.6597,
+      "step": 771
+    },
+    {
+      "epoch": 0.6251012145748988,
+      "grad_norm": 1.1404467641041274,
+      "learning_rate": 1.8135397290537967e-05,
+      "loss": 1.6967,
+      "step": 772
+    },
+    {
+      "epoch": 0.625910931174089,
+      "grad_norm": 1.1388300535406857,
+      "learning_rate": 1.81303960220338e-05,
+      "loss": 1.607,
+      "step": 773
+    },
+    {
+      "epoch": 0.6267206477732794,
+      "grad_norm": 1.1302382932024566,
+      "learning_rate": 1.812538874681976e-05,
+      "loss": 1.6584,
+      "step": 774
+    },
+    {
+      "epoch": 0.6275303643724697,
+      "grad_norm": 1.1014453045454282,
+      "learning_rate": 1.81203754685952e-05,
+      "loss": 1.6767,
+      "step": 775
+    },
+    {
+      "epoch": 0.6283400809716599,
+      "grad_norm": 1.1824063773454823,
+      "learning_rate": 1.8115356191063913e-05,
+      "loss": 1.6373,
+      "step": 776
+    },
+    {
+      "epoch": 0.6291497975708502,
+      "grad_norm": 1.072773503139197,
+      "learning_rate": 1.811033091793413e-05,
+      "loss": 1.6094,
+      "step": 777
+    },
+    {
+      "epoch": 0.6299595141700405,
+      "grad_norm": 1.1293001784836003,
+      "learning_rate": 1.8105299652918496e-05,
+      "loss": 1.6591,
+      "step": 778
+    },
+    {
+      "epoch": 0.6307692307692307,
+      "grad_norm": 1.2465499878402888,
+      "learning_rate": 1.8100262399734102e-05,
+      "loss": 1.7228,
+      "step": 779
+    },
+    {
+      "epoch": 0.631578947368421,
+      "grad_norm": 1.0649028197252695,
+      "learning_rate": 1.8095219162102453e-05,
+      "loss": 1.6456,
+      "step": 780
+    },
+    {
+      "epoch": 0.6323886639676113,
+      "grad_norm": 1.2144051445060644,
+      "learning_rate": 1.8090169943749477e-05,
+      "loss": 1.6515,
+      "step": 781
+    },
+    {
+      "epoch": 0.6331983805668017,
+      "grad_norm": 1.0813125393674639,
+      "learning_rate": 1.8085114748405514e-05,
+      "loss": 1.6634,
+      "step": 782
+    },
+    {
+      "epoch": 0.6340080971659919,
+      "grad_norm": 1.0907352727369097,
+      "learning_rate": 1.8080053579805333e-05,
+      "loss": 1.647,
+      "step": 783
+    },
+    {
+      "epoch": 0.6348178137651822,
+      "grad_norm": 1.1795860026440994,
+      "learning_rate": 1.8074986441688102e-05,
+      "loss": 1.6494,
+      "step": 784
+    },
+    {
+      "epoch": 0.6356275303643725,
+      "grad_norm": 1.0999290606147423,
+      "learning_rate": 1.8069913337797414e-05,
+      "loss": 1.6061,
+      "step": 785
+    },
+    {
+      "epoch": 0.6364372469635627,
+      "grad_norm": 1.1271587687692461,
+      "learning_rate": 1.8064834271881252e-05,
+      "loss": 1.6512,
+      "step": 786
+    },
+    {
+      "epoch": 0.637246963562753,
+      "grad_norm": 1.2475268374395798,
+      "learning_rate": 1.805974924769202e-05,
+      "loss": 1.5999,
+      "step": 787
+    },
+    {
+      "epoch": 0.6380566801619433,
+      "grad_norm": 1.1666492616855777,
+      "learning_rate": 1.8054658268986517e-05,
+      "loss": 1.6134,
+      "step": 788
+    },
+    {
+      "epoch": 0.6388663967611335,
+      "grad_norm": 1.190647434820481,
+      "learning_rate": 1.8049561339525938e-05,
+      "loss": 1.6325,
+      "step": 789
+    },
+    {
+      "epoch": 0.6396761133603239,
+      "grad_norm": 1.1567789620189808,
+      "learning_rate": 1.804445846307588e-05,
+      "loss": 1.6126,
+      "step": 790
+    },
+    {
+      "epoch": 0.6404858299595142,
+      "grad_norm": 1.1205352247753744,
+      "learning_rate": 1.803934964340633e-05,
+      "loss": 1.6759,
+      "step": 791
+    },
+    {
+      "epoch": 0.6412955465587045,
+      "grad_norm": 1.1623098699690382,
+      "learning_rate": 1.803423488429167e-05,
+      "loss": 1.6557,
+      "step": 792
+    },
+    {
+      "epoch": 0.6421052631578947,
+      "grad_norm": 1.0452603483361436,
+      "learning_rate": 1.8029114189510664e-05,
+      "loss": 1.637,
+      "step": 793
+    },
+    {
+      "epoch": 0.642914979757085,
+      "grad_norm": 1.1457994199931247,
+      "learning_rate": 1.8023987562846468e-05,
+      "loss": 1.7003,
+      "step": 794
+    },
+    {
+      "epoch": 0.6437246963562753,
+      "grad_norm": 1.1078125949520747,
+      "learning_rate": 1.801885500808661e-05,
+      "loss": 1.6411,
+      "step": 795
+    },
+    {
+      "epoch": 0.6445344129554655,
+      "grad_norm": 1.063287107792573,
+      "learning_rate": 1.8013716529023013e-05,
+      "loss": 1.5944,
+      "step": 796
+    },
+    {
+      "epoch": 0.6453441295546559,
+      "grad_norm": 1.1463230389504844,
+      "learning_rate": 1.8008572129451963e-05,
+      "loss": 1.6437,
+      "step": 797
+    },
+    {
+      "epoch": 0.6461538461538462,
+      "grad_norm": 1.0745077662710292,
+      "learning_rate": 1.800342181317413e-05,
+      "loss": 1.6141,
+      "step": 798
+    },
+    {
+      "epoch": 0.6469635627530365,
+      "grad_norm": 1.1207965270353415,
+      "learning_rate": 1.7998265583994544e-05,
+      "loss": 1.6043,
+      "step": 799
+    },
+    {
+      "epoch": 0.6477732793522267,
+      "grad_norm": 1.0890890467994665,
+      "learning_rate": 1.7993103445722615e-05,
+      "loss": 1.6425,
+      "step": 800
+    },
+    {
+      "epoch": 0.648582995951417,
+      "grad_norm": 1.13852357372144,
+      "learning_rate": 1.7987935402172114e-05,
+      "loss": 1.6762,
+      "step": 801
+    },
+    {
+      "epoch": 0.6493927125506073,
+      "grad_norm": 1.1427359144052136,
+      "learning_rate": 1.7982761457161175e-05,
+      "loss": 1.6973,
+      "step": 802
+    },
+    {
+      "epoch": 0.6502024291497975,
+      "grad_norm": 1.0845396194326615,
+      "learning_rate": 1.7977581614512286e-05,
+      "loss": 1.6781,
+      "step": 803
+    },
+    {
+      "epoch": 0.6510121457489878,
+      "grad_norm": 1.1491199289525131,
+      "learning_rate": 1.7972395878052304e-05,
+      "loss": 1.6786,
+      "step": 804
+    },
+    {
+      "epoch": 0.6518218623481782,
+      "grad_norm": 1.0901262157214715,
+      "learning_rate": 1.7967204251612432e-05,
+      "loss": 1.6037,
+      "step": 805
+    },
+    {
+      "epoch": 0.6526315789473685,
+      "grad_norm": 1.1102939312877738,
+      "learning_rate": 1.796200673902823e-05,
+      "loss": 1.6314,
+      "step": 806
+    },
+    {
+      "epoch": 0.6534412955465587,
+      "grad_norm": 1.1409639624890162,
+      "learning_rate": 1.7956803344139592e-05,
+      "loss": 1.7136,
+      "step": 807
+    },
+    {
+      "epoch": 0.654251012145749,
+      "grad_norm": 1.0991234850498368,
+      "learning_rate": 1.795159407079078e-05,
+      "loss": 1.6653,
+      "step": 808
+    },
+    {
+      "epoch": 0.6550607287449393,
+      "grad_norm": 1.1172599612597327,
+      "learning_rate": 1.7946378922830386e-05,
+      "loss": 1.625,
+      "step": 809
+    },
+    {
+      "epoch": 0.6558704453441295,
+      "grad_norm": 1.0552333200507726,
+      "learning_rate": 1.7941157904111346e-05,
+      "loss": 1.7037,
+      "step": 810
+    },
+    {
+      "epoch": 0.6566801619433198,
+      "grad_norm": 1.080338568915127,
+      "learning_rate": 1.7935931018490923e-05,
+      "loss": 1.6834,
+      "step": 811
+    },
+    {
+      "epoch": 0.6574898785425102,
+      "grad_norm": 1.086653104326212,
+      "learning_rate": 1.7930698269830733e-05,
+      "loss": 1.6364,
+      "step": 812
+    },
+    {
+      "epoch": 0.6582995951417004,
+      "grad_norm": 1.0712198372881643,
+      "learning_rate": 1.7925459661996707e-05,
+      "loss": 1.5958,
+      "step": 813
+    },
+    {
+      "epoch": 0.6591093117408907,
+      "grad_norm": 1.1256703101949748,
+      "learning_rate": 1.7920215198859114e-05,
+      "loss": 1.6405,
+      "step": 814
+    },
+    {
+      "epoch": 0.659919028340081,
+      "grad_norm": 1.1116563555135301,
+      "learning_rate": 1.7914964884292543e-05,
+      "loss": 1.6689,
+      "step": 815
+    },
+    {
+      "epoch": 0.6607287449392713,
+      "grad_norm": 1.12107713293399,
+      "learning_rate": 1.7909708722175914e-05,
+      "loss": 1.6928,
+      "step": 816
+    },
+    {
+      "epoch": 0.6615384615384615,
+      "grad_norm": 1.1619550044989124,
+      "learning_rate": 1.7904446716392457e-05,
+      "loss": 1.6034,
+      "step": 817
+    },
+    {
+      "epoch": 0.6623481781376518,
+      "grad_norm": 1.1005819448367389,
+      "learning_rate": 1.789917887082973e-05,
+      "loss": 1.6888,
+      "step": 818
+    },
+    {
+      "epoch": 0.6631578947368421,
+      "grad_norm": 1.1827182789963149,
+      "learning_rate": 1.7893905189379594e-05,
+      "loss": 1.6213,
+      "step": 819
+    },
+    {
+      "epoch": 0.6639676113360324,
+      "grad_norm": 1.1422059406503742,
+      "learning_rate": 1.7888625675938237e-05,
+      "loss": 1.6266,
+      "step": 820
+    },
+    {
+      "epoch": 0.6647773279352227,
+      "grad_norm": 1.0470507697677474,
+      "learning_rate": 1.7883340334406136e-05,
+      "loss": 1.6561,
+      "step": 821
+    },
+    {
+      "epoch": 0.665587044534413,
+      "grad_norm": 1.1216106844368927,
+      "learning_rate": 1.7878049168688087e-05,
+      "loss": 1.6705,
+      "step": 822
+    },
+    {
+      "epoch": 0.6663967611336032,
+      "grad_norm": 1.2138659468039459,
+      "learning_rate": 1.787275218269319e-05,
+      "loss": 1.6336,
+      "step": 823
+    },
+    {
+      "epoch": 0.6672064777327935,
+      "grad_norm": 1.0986476320949485,
+      "learning_rate": 1.7867449380334834e-05,
+      "loss": 1.7078,
+      "step": 824
+    },
+    {
+      "epoch": 0.6680161943319838,
+      "grad_norm": 1.288830051080927,
+      "learning_rate": 1.7862140765530718e-05,
+      "loss": 1.6254,
+      "step": 825
+    },
+    {
+      "epoch": 0.668825910931174,
+      "grad_norm": 1.1447438193383035,
+      "learning_rate": 1.7856826342202828e-05,
+      "loss": 1.6201,
+      "step": 826
+    },
+    {
+      "epoch": 0.6696356275303643,
+      "grad_norm": 1.1649226403537143,
+      "learning_rate": 1.785150611427744e-05,
+      "loss": 1.6864,
+      "step": 827
+    },
+    {
+      "epoch": 0.6704453441295547,
+      "grad_norm": 1.118218283333164,
+      "learning_rate": 1.7846180085685122e-05,
+      "loss": 1.6007,
+      "step": 828
+    },
+    {
+      "epoch": 0.671255060728745,
+      "grad_norm": 1.1014855871694855,
+      "learning_rate": 1.7840848260360728e-05,
+      "loss": 1.6499,
+      "step": 829
+    },
+    {
+      "epoch": 0.6720647773279352,
+      "grad_norm": 1.2218105006616033,
+      "learning_rate": 1.783551064224339e-05,
+      "loss": 1.6338,
+      "step": 830
+    },
+    {
+      "epoch": 0.6728744939271255,
+      "grad_norm": 1.1476489793760862,
+      "learning_rate": 1.7830167235276524e-05,
+      "loss": 1.6073,
+      "step": 831
+    },
+    {
+      "epoch": 0.6736842105263158,
+      "grad_norm": 1.1653286803987006,
+      "learning_rate": 1.7824818043407828e-05,
+      "loss": 1.5989,
+      "step": 832
+    },
+    {
+      "epoch": 0.674493927125506,
+      "grad_norm": 1.1159344507824178,
+      "learning_rate": 1.7819463070589256e-05,
+      "loss": 1.6211,
+      "step": 833
+    },
+    {
+      "epoch": 0.6753036437246963,
+      "grad_norm": 1.0909665741148316,
+      "learning_rate": 1.781410232077705e-05,
+      "loss": 1.6052,
+      "step": 834
+    },
+    {
+      "epoch": 0.6761133603238867,
+      "grad_norm": 1.1405758039320455,
+      "learning_rate": 1.7808735797931715e-05,
+      "loss": 1.7158,
+      "step": 835
+    },
+    {
+      "epoch": 0.676923076923077,
+      "grad_norm": 1.1662611008186603,
+      "learning_rate": 1.780336350601802e-05,
+      "loss": 1.714,
+      "step": 836
+    },
+    {
+      "epoch": 0.6777327935222672,
+      "grad_norm": 1.1318426366131789,
+      "learning_rate": 1.7797985449004996e-05,
+      "loss": 1.6362,
+      "step": 837
+    },
+    {
+      "epoch": 0.6785425101214575,
+      "grad_norm": 1.1122001699040855,
+      "learning_rate": 1.7792601630865937e-05,
+      "loss": 1.6337,
+      "step": 838
+    },
+    {
+      "epoch": 0.6793522267206478,
+      "grad_norm": 1.1309770551503653,
+      "learning_rate": 1.7787212055578383e-05,
+      "loss": 1.5989,
+      "step": 839
+    },
+    {
+      "epoch": 0.680161943319838,
+      "grad_norm": 1.1673264771022611,
+      "learning_rate": 1.7781816727124138e-05,
+      "loss": 1.6649,
+      "step": 840
+    },
+    {
+      "epoch": 0.6809716599190283,
+      "grad_norm": 1.1003607212760034,
+      "learning_rate": 1.7776415649489257e-05,
+      "loss": 1.6183,
+      "step": 841
+    },
+    {
+      "epoch": 0.6817813765182186,
+      "grad_norm": 1.1092869043197529,
+      "learning_rate": 1.7771008826664036e-05,
+      "loss": 1.6678,
+      "step": 842
+    },
+    {
+      "epoch": 0.682591093117409,
+      "grad_norm": 1.2211528749408376,
+      "learning_rate": 1.7765596262643013e-05,
+      "loss": 1.6584,
+      "step": 843
+    },
+    {
+      "epoch": 0.6834008097165992,
+      "grad_norm": 1.1758598116755494,
+      "learning_rate": 1.776017796142498e-05,
+      "loss": 1.6415,
+      "step": 844
+    },
+    {
+      "epoch": 0.6842105263157895,
+      "grad_norm": 1.1767326941123255,
+      "learning_rate": 1.7754753927012955e-05,
+      "loss": 1.6265,
+      "step": 845
+    },
+    {
+      "epoch": 0.6850202429149798,
+      "grad_norm": 1.1747867351240824,
+      "learning_rate": 1.77493241634142e-05,
+      "loss": 1.625,
+      "step": 846
+    },
+    {
+      "epoch": 0.68582995951417,
+      "grad_norm": 1.1345683268627573,
+      "learning_rate": 1.7743888674640203e-05,
+      "loss": 1.6418,
+      "step": 847
+    },
+    {
+      "epoch": 0.6866396761133603,
+      "grad_norm": 1.1283603891809075,
+      "learning_rate": 1.773844746470669e-05,
+      "loss": 1.6412,
+      "step": 848
+    },
+    {
+      "epoch": 0.6874493927125506,
+      "grad_norm": 1.1378377091313905,
+      "learning_rate": 1.7733000537633605e-05,
+      "loss": 1.6298,
+      "step": 849
+    },
+    {
+      "epoch": 0.6882591093117408,
+      "grad_norm": 1.1253702667997967,
+      "learning_rate": 1.7727547897445117e-05,
+      "loss": 1.6472,
+      "step": 850
+    },
+    {
+      "epoch": 0.6890688259109312,
+      "grad_norm": 1.1529285701924186,
+      "learning_rate": 1.772208954816963e-05,
+      "loss": 1.6063,
+      "step": 851
+    },
+    {
+      "epoch": 0.6898785425101215,
+      "grad_norm": 1.2088268908844393,
+      "learning_rate": 1.771662549383974e-05,
+      "loss": 1.6893,
+      "step": 852
+    },
+    {
+      "epoch": 0.6906882591093118,
+      "grad_norm": 1.1008137117384047,
+      "learning_rate": 1.7711155738492286e-05,
+      "loss": 1.7038,
+      "step": 853
+    },
+    {
+      "epoch": 0.691497975708502,
+      "grad_norm": 1.1203578405697336,
+      "learning_rate": 1.7705680286168297e-05,
+      "loss": 1.691,
+      "step": 854
+    },
+    {
+      "epoch": 0.6923076923076923,
+      "grad_norm": 1.0904874884732818,
+      "learning_rate": 1.770019914091302e-05,
+      "loss": 1.66,
+      "step": 855
+    },
+    {
+      "epoch": 0.6931174089068826,
+      "grad_norm": 1.2342232334641767,
+      "learning_rate": 1.769471230677591e-05,
+      "loss": 1.6029,
+      "step": 856
+    },
+    {
+      "epoch": 0.6939271255060728,
+      "grad_norm": 1.1892221297007457,
+      "learning_rate": 1.7689219787810618e-05,
+      "loss": 1.6115,
+      "step": 857
+    },
+    {
+      "epoch": 0.6947368421052632,
+      "grad_norm": 1.1436275460030911,
+      "learning_rate": 1.7683721588075005e-05,
+      "loss": 1.6779,
+      "step": 858
+    },
+    {
+      "epoch": 0.6955465587044535,
+      "grad_norm": 1.0531746133706077,
+      "learning_rate": 1.7678217711631115e-05,
+      "loss": 1.6254,
+      "step": 859
+    },
+    {
+      "epoch": 0.6963562753036437,
+      "grad_norm": 1.204345691882841,
+      "learning_rate": 1.76727081625452e-05,
+      "loss": 1.7005,
+      "step": 860
+    },
+    {
+      "epoch": 0.697165991902834,
+      "grad_norm": 1.1113015688625698,
+      "learning_rate": 1.7667192944887696e-05,
+      "loss": 1.6242,
+      "step": 861
+    },
+    {
+      "epoch": 0.6979757085020243,
+      "grad_norm": 1.0642180179681062,
+      "learning_rate": 1.7661672062733226e-05,
+      "loss": 1.6057,
+      "step": 862
+    },
+    {
+      "epoch": 0.6987854251012146,
+      "grad_norm": 1.1074568311327786,
+      "learning_rate": 1.76561455201606e-05,
+      "loss": 1.6569,
+      "step": 863
+    },
+    {
+      "epoch": 0.6995951417004048,
+      "grad_norm": 1.1475467339069338,
+      "learning_rate": 1.765061332125281e-05,
+      "loss": 1.6668,
+      "step": 864
+    },
+    {
+      "epoch": 0.7004048582995951,
+      "grad_norm": 1.0613765631942633,
+      "learning_rate": 1.7645075470097024e-05,
+      "loss": 1.6471,
+      "step": 865
+    },
+    {
+      "epoch": 0.7012145748987855,
+      "grad_norm": 1.1262125184431948,
+      "learning_rate": 1.7639531970784594e-05,
+      "loss": 1.632,
+      "step": 866
+    },
+    {
+      "epoch": 0.7020242914979757,
+      "grad_norm": 1.053405645515529,
+      "learning_rate": 1.763398282741103e-05,
+      "loss": 1.6628,
+      "step": 867
+    },
+    {
+      "epoch": 0.702834008097166,
+      "grad_norm": 1.053485822528232,
+      "learning_rate": 1.762842804407603e-05,
+      "loss": 1.6268,
+      "step": 868
+    },
+    {
+      "epoch": 0.7036437246963563,
+      "grad_norm": 1.1067274819740494,
+      "learning_rate": 1.7622867624883446e-05,
+      "loss": 1.6268,
+      "step": 869
+    },
+    {
+      "epoch": 0.7044534412955465,
+      "grad_norm": 1.1365300210950073,
+      "learning_rate": 1.7617301573941296e-05,
+      "loss": 1.6751,
+      "step": 870
+    },
+    {
+      "epoch": 0.7052631578947368,
+      "grad_norm": 1.0944652038336933,
+      "learning_rate": 1.7611729895361763e-05,
+      "loss": 1.7041,
+      "step": 871
+    },
+    {
+      "epoch": 0.7060728744939271,
+      "grad_norm": 1.0536399181541465,
+      "learning_rate": 1.760615259326118e-05,
+      "loss": 1.6283,
+      "step": 872
+    },
+    {
+      "epoch": 0.7068825910931175,
+      "grad_norm": 1.0489723553006804,
+      "learning_rate": 1.760056967176005e-05,
+      "loss": 1.7136,
+      "step": 873
+    },
+    {
+      "epoch": 0.7076923076923077,
+      "grad_norm": 1.0719751555919763,
+      "learning_rate": 1.7594981134983003e-05,
+      "loss": 1.6001,
+      "step": 874
+    },
+    {
+      "epoch": 0.708502024291498,
+      "grad_norm": 1.1499881882739111,
+      "learning_rate": 1.758938698705884e-05,
+      "loss": 1.677,
+      "step": 875
+    },
+    {
+      "epoch": 0.7093117408906883,
+      "grad_norm": 1.101761750936134,
+      "learning_rate": 1.75837872321205e-05,
+      "loss": 1.5747,
+      "step": 876
+    },
+    {
+      "epoch": 0.7101214574898785,
+      "grad_norm": 1.077980130931622,
+      "learning_rate": 1.757818187430506e-05,
+      "loss": 1.6827,
+      "step": 877
+    },
+    {
+      "epoch": 0.7109311740890688,
+      "grad_norm": 1.0897713278469439,
+      "learning_rate": 1.757257091775374e-05,
+      "loss": 1.659,
+      "step": 878
+    },
+    {
+      "epoch": 0.7117408906882591,
+      "grad_norm": 1.4838670869876838,
+      "learning_rate": 1.7566954366611896e-05,
+      "loss": 1.6481,
+      "step": 879
+    },
+    {
+      "epoch": 0.7125506072874493,
+      "grad_norm": 1.1522319490945014,
+      "learning_rate": 1.7561332225029022e-05,
+      "loss": 1.6426,
+      "step": 880
+    },
+    {
+      "epoch": 0.7133603238866397,
+      "grad_norm": 1.1601026341967655,
+      "learning_rate": 1.7555704497158734e-05,
+      "loss": 1.6299,
+      "step": 881
+    },
+    {
+      "epoch": 0.71417004048583,
+      "grad_norm": 1.0844245975035776,
+      "learning_rate": 1.755007118715878e-05,
+      "loss": 1.6366,
+      "step": 882
+    },
+    {
+      "epoch": 0.7149797570850203,
+      "grad_norm": 1.106288732347795,
+      "learning_rate": 1.754443229919103e-05,
+      "loss": 1.6275,
+      "step": 883
+    },
+    {
+      "epoch": 0.7157894736842105,
+      "grad_norm": 1.0464400021747562,
+      "learning_rate": 1.7538787837421475e-05,
+      "loss": 1.5273,
+      "step": 884
+    },
+    {
+      "epoch": 0.7165991902834008,
+      "grad_norm": 1.1156141109394004,
+      "learning_rate": 1.7533137806020226e-05,
+      "loss": 1.6415,
+      "step": 885
+    },
+    {
+      "epoch": 0.7174089068825911,
+      "grad_norm": 1.1617480668291316,
+      "learning_rate": 1.752748220916151e-05,
+      "loss": 1.6285,
+      "step": 886
+    },
+    {
+      "epoch": 0.7182186234817813,
+      "grad_norm": 1.0663413051063326,
+      "learning_rate": 1.752182105102366e-05,
+      "loss": 1.6715,
+      "step": 887
+    },
+    {
+      "epoch": 0.7190283400809716,
+      "grad_norm": 1.1351571282717112,
+      "learning_rate": 1.7516154335789123e-05,
+      "loss": 1.676,
+      "step": 888
+    },
+    {
+      "epoch": 0.719838056680162,
+      "grad_norm": 1.0745374695294985,
+      "learning_rate": 1.751048206764445e-05,
+      "loss": 1.6808,
+      "step": 889
+    },
+    {
+      "epoch": 0.7206477732793523,
+      "grad_norm": 1.1821923099732665,
+      "learning_rate": 1.7504804250780292e-05,
+      "loss": 1.6717,
+      "step": 890
+    },
+    {
+      "epoch": 0.7214574898785425,
+      "grad_norm": 1.1136129753607058,
+      "learning_rate": 1.7499120889391403e-05,
+      "loss": 1.6181,
+      "step": 891
+    },
+    {
+      "epoch": 0.7222672064777328,
+      "grad_norm": 1.114536036094605,
+      "learning_rate": 1.7493431987676628e-05,
+      "loss": 1.6641,
+      "step": 892
+    },
+    {
+      "epoch": 0.7230769230769231,
+      "grad_norm": 1.2106559019950824,
+      "learning_rate": 1.7487737549838915e-05,
+      "loss": 1.673,
+      "step": 893
+    },
+    {
+      "epoch": 0.7238866396761133,
+      "grad_norm": 1.1049671120265863,
+      "learning_rate": 1.748203758008529e-05,
+      "loss": 1.6829,
+      "step": 894
+    },
+    {
+      "epoch": 0.7246963562753036,
+      "grad_norm": 1.0330699329285005,
+      "learning_rate": 1.747633208262688e-05,
+      "loss": 1.5981,
+      "step": 895
+    },
+    {
+      "epoch": 0.725506072874494,
+      "grad_norm": 1.122795995838303,
+      "learning_rate": 1.747062106167888e-05,
+      "loss": 1.6874,
+      "step": 896
+    },
+    {
+      "epoch": 0.7263157894736842,
+      "grad_norm": 1.0411327718132521,
+      "learning_rate": 1.7464904521460574e-05,
+      "loss": 1.6655,
+      "step": 897
+    },
+    {
+      "epoch": 0.7271255060728745,
+      "grad_norm": 1.094203881932362,
+      "learning_rate": 1.7459182466195328e-05,
+      "loss": 1.6611,
+      "step": 898
+    },
+    {
+      "epoch": 0.7279352226720648,
+      "grad_norm": 1.1197343663542865,
+      "learning_rate": 1.7453454900110575e-05,
+      "loss": 1.6484,
+      "step": 899
+    },
+    {
+      "epoch": 0.728744939271255,
+      "grad_norm": 1.0891684716999215,
+      "learning_rate": 1.744772182743782e-05,
+      "loss": 1.6823,
+      "step": 900
+    },
+    {
+      "epoch": 0.7295546558704453,
+      "grad_norm": 1.0117179095825084,
+      "learning_rate": 1.744198325241264e-05,
+      "loss": 1.5874,
+      "step": 901
+    },
+    {
+      "epoch": 0.7303643724696356,
+      "grad_norm": 1.11669367321381,
+      "learning_rate": 1.743623917927468e-05,
+      "loss": 1.6327,
+      "step": 902
+    },
+    {
+      "epoch": 0.7311740890688259,
+      "grad_norm": 1.0477262743830713,
+      "learning_rate": 1.7430489612267634e-05,
+      "loss": 1.5939,
+      "step": 903
+    },
+    {
+      "epoch": 0.7319838056680162,
+      "grad_norm": 1.1755335589263856,
+      "learning_rate": 1.742473455563927e-05,
+      "loss": 1.6671,
+      "step": 904
+    },
+    {
+      "epoch": 0.7327935222672065,
+      "grad_norm": 1.1271564518170447,
+      "learning_rate": 1.74189740136414e-05,
+      "loss": 1.6179,
+      "step": 905
+    },
+    {
+      "epoch": 0.7336032388663968,
+      "grad_norm": 1.1024780353509462,
+      "learning_rate": 1.7413207990529897e-05,
+      "loss": 1.6315,
+      "step": 906
+    },
+    {
+      "epoch": 0.734412955465587,
+      "grad_norm": 1.057413980088522,
+      "learning_rate": 1.7407436490564675e-05,
+      "loss": 1.6019,
+      "step": 907
+    },
+    {
+      "epoch": 0.7352226720647773,
+      "grad_norm": 1.0903114614208873,
+      "learning_rate": 1.740165951800971e-05,
+      "loss": 1.6435,
+      "step": 908
+    },
+    {
+      "epoch": 0.7360323886639676,
+      "grad_norm": 1.1465351066882732,
+      "learning_rate": 1.7395877077132996e-05,
+      "loss": 1.6659,
+      "step": 909
+    },
+    {
+      "epoch": 0.7368421052631579,
+      "grad_norm": 1.0879106203040465,
+      "learning_rate": 1.7390089172206594e-05,
+      "loss": 1.6035,
+      "step": 910
+    },
+    {
+      "epoch": 0.7376518218623481,
+      "grad_norm": 1.0251909256160818,
+      "learning_rate": 1.738429580750658e-05,
+      "loss": 1.6751,
+      "step": 911
+    },
+    {
+      "epoch": 0.7384615384615385,
+      "grad_norm": 1.101454173734882,
+      "learning_rate": 1.7378496987313078e-05,
+      "loss": 1.6924,
+      "step": 912
+    },
+    {
+      "epoch": 0.7392712550607288,
+      "grad_norm": 1.0632875331068412,
+      "learning_rate": 1.7372692715910236e-05,
+      "loss": 1.7052,
+      "step": 913
+    },
+    {
+      "epoch": 0.740080971659919,
+      "grad_norm": 1.2019849547754717,
+      "learning_rate": 1.736688299758623e-05,
+      "loss": 1.6453,
+      "step": 914
+    },
+    {
+      "epoch": 0.7408906882591093,
+      "grad_norm": 1.0424618834849857,
+      "learning_rate": 1.736106783663326e-05,
+      "loss": 1.6095,
+      "step": 915
+    },
+    {
+      "epoch": 0.7417004048582996,
+      "grad_norm": 1.119522754130525,
+      "learning_rate": 1.735524723734755e-05,
+      "loss": 1.5714,
+      "step": 916
+    },
+    {
+      "epoch": 0.7425101214574898,
+      "grad_norm": 1.1126763401822644,
+      "learning_rate": 1.7349421204029343e-05,
+      "loss": 1.5884,
+      "step": 917
+    },
+    {
+      "epoch": 0.7433198380566801,
+      "grad_norm": 1.1436206847704529,
+      "learning_rate": 1.734358974098288e-05,
+      "loss": 1.6716,
+      "step": 918
+    },
+    {
+      "epoch": 0.7441295546558705,
+      "grad_norm": 1.1033761896656453,
+      "learning_rate": 1.7337752852516443e-05,
+      "loss": 1.6642,
+      "step": 919
+    },
+    {
+      "epoch": 0.7449392712550608,
+      "grad_norm": 1.0771058808303946,
+      "learning_rate": 1.7331910542942298e-05,
+      "loss": 1.6655,
+      "step": 920
+    },
+    {
+      "epoch": 0.745748987854251,
+      "grad_norm": 1.1540799878521129,
+      "learning_rate": 1.732606281657672e-05,
+      "loss": 1.6533,
+      "step": 921
+    },
+    {
+      "epoch": 0.7465587044534413,
+      "grad_norm": 1.1632383389035768,
+      "learning_rate": 1.732020967774e-05,
+      "loss": 1.6251,
+      "step": 922
+    },
+    {
+      "epoch": 0.7473684210526316,
+      "grad_norm": 1.0554497626152348,
+      "learning_rate": 1.7314351130756412e-05,
+      "loss": 1.6913,
+      "step": 923
+    },
+    {
+      "epoch": 0.7481781376518218,
+      "grad_norm": 1.0558743109746518,
+      "learning_rate": 1.7308487179954233e-05,
+      "loss": 1.6377,
+      "step": 924
+    },
+    {
+      "epoch": 0.7489878542510121,
+      "grad_norm": 1.0485799147334454,
+      "learning_rate": 1.7302617829665725e-05,
+      "loss": 1.6837,
+      "step": 925
+    },
+    {
+      "epoch": 0.7497975708502024,
+      "grad_norm": 1.1396931481843333,
+      "learning_rate": 1.729674308422715e-05,
+      "loss": 1.6475,
+      "step": 926
+    },
+    {
+      "epoch": 0.7506072874493928,
+      "grad_norm": 1.089831914564661,
+      "learning_rate": 1.7290862947978753e-05,
+      "loss": 1.6612,
+      "step": 927
+    },
+    {
+      "epoch": 0.751417004048583,
+      "grad_norm": 1.214076933636336,
+      "learning_rate": 1.7284977425264755e-05,
+      "loss": 1.6558,
+      "step": 928
+    },
+    {
+      "epoch": 0.7522267206477733,
+      "grad_norm": 1.0860938684718011,
+      "learning_rate": 1.727908652043336e-05,
+      "loss": 1.6885,
+      "step": 929
+    },
+    {
+      "epoch": 0.7530364372469636,
+      "grad_norm": 1.101065862505211,
+      "learning_rate": 1.7273190237836757e-05,
+      "loss": 1.6578,
+      "step": 930
+    },
+    {
+      "epoch": 0.7538461538461538,
+      "grad_norm": 1.1342711520778153,
+      "learning_rate": 1.726728858183109e-05,
+      "loss": 1.5637,
+      "step": 931
+    },
+    {
+      "epoch": 0.7546558704453441,
+      "grad_norm": 1.0939363758915137,
+      "learning_rate": 1.726138155677649e-05,
+      "loss": 1.6303,
+      "step": 932
+    },
+    {
+      "epoch": 0.7554655870445344,
+      "grad_norm": 1.0310679045269706,
+      "learning_rate": 1.725546916703704e-05,
+      "loss": 1.5678,
+      "step": 933
+    },
+    {
+      "epoch": 0.7562753036437248,
+      "grad_norm": 1.1728811689558527,
+      "learning_rate": 1.7249551416980806e-05,
+      "loss": 1.6611,
+      "step": 934
+    },
+    {
+      "epoch": 0.757085020242915,
+      "grad_norm": 1.1888012056779975,
+      "learning_rate": 1.7243628310979793e-05,
+      "loss": 1.6931,
+      "step": 935
+    },
+    {
+      "epoch": 0.7578947368421053,
+      "grad_norm": 1.1056546162441991,
+      "learning_rate": 1.7237699853409974e-05,
+      "loss": 1.6788,
+      "step": 936
+    },
+    {
+      "epoch": 0.7587044534412956,
+      "grad_norm": 1.1020429695381275,
+      "learning_rate": 1.7231766048651272e-05,
+      "loss": 1.6557,
+      "step": 937
+    },
+    {
+      "epoch": 0.7595141700404858,
+      "grad_norm": 1.174830307710626,
+      "learning_rate": 1.722582690108757e-05,
+      "loss": 1.5964,
+      "step": 938
+    },
+    {
+      "epoch": 0.7603238866396761,
+      "grad_norm": 1.1111495814988965,
+      "learning_rate": 1.7219882415106685e-05,
+      "loss": 1.6547,
+      "step": 939
+    },
+    {
+      "epoch": 0.7611336032388664,
+      "grad_norm": 1.1511962350490403,
+      "learning_rate": 1.7213932595100384e-05,
+      "loss": 1.6447,
+      "step": 940
+    },
+    {
+      "epoch": 0.7619433198380566,
+      "grad_norm": 1.1124431654124347,
+      "learning_rate": 1.7207977445464374e-05,
+      "loss": 1.6756,
+      "step": 941
+    },
+    {
+      "epoch": 0.762753036437247,
+      "grad_norm": 1.094632192703889,
+      "learning_rate": 1.7202016970598303e-05,
+      "loss": 1.6216,
+      "step": 942
+    },
+    {
+      "epoch": 0.7635627530364373,
+      "grad_norm": 1.0848286346166132,
+      "learning_rate": 1.7196051174905746e-05,
+      "loss": 1.6162,
+      "step": 943
+    },
+    {
+      "epoch": 0.7643724696356275,
+      "grad_norm": 1.117835402570124,
+      "learning_rate": 1.719008006279422e-05,
+      "loss": 1.6988,
+      "step": 944
+    },
+    {
+      "epoch": 0.7651821862348178,
+      "grad_norm": 1.0695545232404842,
+      "learning_rate": 1.7184103638675157e-05,
+      "loss": 1.6417,
+      "step": 945
+    },
+    {
+      "epoch": 0.7659919028340081,
+      "grad_norm": 1.2187648019898212,
+      "learning_rate": 1.7178121906963925e-05,
+      "loss": 1.5581,
+      "step": 946
+    },
+    {
+      "epoch": 0.7668016194331984,
+      "grad_norm": 1.1428081923285947,
+      "learning_rate": 1.71721348720798e-05,
+      "loss": 1.5958,
+      "step": 947
+    },
+    {
+      "epoch": 0.7676113360323886,
+      "grad_norm": 1.1258009592099063,
+      "learning_rate": 1.7166142538445986e-05,
+      "loss": 1.7054,
+      "step": 948
+    },
+    {
+      "epoch": 0.7684210526315789,
+      "grad_norm": 1.122110087454184,
+      "learning_rate": 1.7160144910489602e-05,
+      "loss": 1.6479,
+      "step": 949
+    },
+    {
+      "epoch": 0.7692307692307693,
+      "grad_norm": 1.2197448941368936,
+      "learning_rate": 1.715414199264168e-05,
+      "loss": 1.6362,
+      "step": 950
+    },
+    {
+      "epoch": 0.7700404858299595,
+      "grad_norm": 1.081986455490488,
+      "learning_rate": 1.7148133789337145e-05,
+      "loss": 1.6376,
+      "step": 951
+    },
+    {
+      "epoch": 0.7708502024291498,
+      "grad_norm": 1.0866654156187887,
+      "learning_rate": 1.7142120305014848e-05,
+      "loss": 1.6275,
+      "step": 952
+    },
+    {
+      "epoch": 0.7716599190283401,
+      "grad_norm": 1.1078330177065072,
+      "learning_rate": 1.7136101544117526e-05,
+      "loss": 1.62,
+      "step": 953
+    },
+    {
+      "epoch": 0.7724696356275303,
+      "grad_norm": 1.0817145396361252,
+      "learning_rate": 1.7130077511091817e-05,
+      "loss": 1.6552,
+      "step": 954
+    },
+    {
+      "epoch": 0.7732793522267206,
+      "grad_norm": 1.105540137190186,
+      "learning_rate": 1.7124048210388268e-05,
+      "loss": 1.6459,
+      "step": 955
+    },
+    {
+      "epoch": 0.7740890688259109,
+      "grad_norm": 1.0847839201754785,
+      "learning_rate": 1.7118013646461295e-05,
+      "loss": 1.6759,
+      "step": 956
+    },
+    {
+      "epoch": 0.7748987854251013,
+      "grad_norm": 1.084427169214924,
+      "learning_rate": 1.711197382376922e-05,
+      "loss": 1.6833,
+      "step": 957
+    },
+    {
+      "epoch": 0.7757085020242915,
+      "grad_norm": 1.131195586447811,
+      "learning_rate": 1.710592874677424e-05,
+      "loss": 1.6091,
+      "step": 958
+    },
+    {
+      "epoch": 0.7765182186234818,
+      "grad_norm": 1.0521712965264904,
+      "learning_rate": 1.7099878419942444e-05,
+      "loss": 1.6319,
+      "step": 959
+    },
+    {
+      "epoch": 0.7773279352226721,
+      "grad_norm": 1.0638585192349859,
+      "learning_rate": 1.709382284774379e-05,
+      "loss": 1.6039,
+      "step": 960
+    },
+    {
+      "epoch": 0.7781376518218623,
+      "grad_norm": 1.1423083079533596,
+      "learning_rate": 1.7087762034652113e-05,
+      "loss": 1.672,
+      "step": 961
+    },
+    {
+      "epoch": 0.7789473684210526,
+      "grad_norm": 1.1028125449584352,
+      "learning_rate": 1.7081695985145124e-05,
+      "loss": 1.6914,
+      "step": 962
+    },
+    {
+      "epoch": 0.7797570850202429,
+      "grad_norm": 1.1268022996799139,
+      "learning_rate": 1.7075624703704404e-05,
+      "loss": 1.7132,
+      "step": 963
+    },
+    {
+      "epoch": 0.7805668016194331,
+      "grad_norm": 1.1209367289547245,
+      "learning_rate": 1.7069548194815387e-05,
+      "loss": 1.6357,
+      "step": 964
+    },
+    {
+      "epoch": 0.7813765182186235,
+      "grad_norm": 1.1887708228854563,
+      "learning_rate": 1.706346646296739e-05,
+      "loss": 1.5665,
+      "step": 965
+    },
+    {
+      "epoch": 0.7821862348178138,
+      "grad_norm": 1.118443433275976,
+      "learning_rate": 1.7057379512653565e-05,
+      "loss": 1.6763,
+      "step": 966
+    },
+    {
+      "epoch": 0.7829959514170041,
+      "grad_norm": 1.0454329186401545,
+      "learning_rate": 1.7051287348370934e-05,
+      "loss": 1.5866,
+      "step": 967
+    },
+    {
+      "epoch": 0.7838056680161943,
+      "grad_norm": 1.076636884210294,
+      "learning_rate": 1.704518997462037e-05,
+      "loss": 1.6171,
+      "step": 968
+    },
+    {
+      "epoch": 0.7846153846153846,
+      "grad_norm": 1.02837143519001,
+      "learning_rate": 1.7039087395906593e-05,
+      "loss": 1.6411,
+      "step": 969
+    },
+    {
+      "epoch": 0.7854251012145749,
+      "grad_norm": 1.0924312535153355,
+      "learning_rate": 1.7032979616738167e-05,
+      "loss": 1.6169,
+      "step": 970
+    },
+    {
+      "epoch": 0.7862348178137651,
+      "grad_norm": 1.0621344949493872,
+      "learning_rate": 1.7026866641627503e-05,
+      "loss": 1.6543,
+      "step": 971
+    },
+    {
+      "epoch": 0.7870445344129555,
+      "grad_norm": 1.0840593933228118,
+      "learning_rate": 1.7020748475090835e-05,
+      "loss": 1.6604,
+      "step": 972
+    },
+    {
+      "epoch": 0.7878542510121458,
+      "grad_norm": 1.0324309043206548,
+      "learning_rate": 1.701462512164826e-05,
+      "loss": 1.6547,
+      "step": 973
+    },
+    {
+      "epoch": 0.7886639676113361,
+      "grad_norm": 1.1347770119446707,
+      "learning_rate": 1.700849658582368e-05,
+      "loss": 1.6716,
+      "step": 974
+    },
+    {
+      "epoch": 0.7894736842105263,
+      "grad_norm": 1.1772883414899646,
+      "learning_rate": 1.7002362872144843e-05,
+      "loss": 1.6968,
+      "step": 975
+    },
+    {
+      "epoch": 0.7902834008097166,
+      "grad_norm": 1.1298734658918796,
+      "learning_rate": 1.6996223985143314e-05,
+      "loss": 1.7227,
+      "step": 976
+    },
+    {
+      "epoch": 0.7910931174089069,
+      "grad_norm": 1.0893915856547574,
+      "learning_rate": 1.6990079929354485e-05,
+      "loss": 1.6051,
+      "step": 977
+    },
+    {
+      "epoch": 0.7919028340080971,
+      "grad_norm": 1.0564375360450058,
+      "learning_rate": 1.698393070931756e-05,
+      "loss": 1.6565,
+      "step": 978
+    },
+    {
+      "epoch": 0.7927125506072874,
+      "grad_norm": 1.0465087498034826,
+      "learning_rate": 1.6977776329575566e-05,
+      "loss": 1.6737,
+      "step": 979
+    },
+    {
+      "epoch": 0.7935222672064778,
+      "grad_norm": 1.0938693523776055,
+      "learning_rate": 1.697161679467534e-05,
+      "loss": 1.6149,
+      "step": 980
+    },
+    {
+      "epoch": 0.794331983805668,
+      "grad_norm": 1.1328659849792375,
+      "learning_rate": 1.696545210916752e-05,
+      "loss": 1.6203,
+      "step": 981
+    },
+    {
+      "epoch": 0.7951417004048583,
+      "grad_norm": 1.1231511367270903,
+      "learning_rate": 1.695928227760656e-05,
+      "loss": 1.7056,
+      "step": 982
+    },
+    {
+      "epoch": 0.7959514170040486,
+      "grad_norm": 1.1090374560264693,
+      "learning_rate": 1.6953107304550714e-05,
+      "loss": 1.5986,
+      "step": 983
+    },
+    {
+      "epoch": 0.7967611336032389,
+      "grad_norm": 1.0770482845738483,
+      "learning_rate": 1.694692719456202e-05,
+      "loss": 1.6314,
+      "step": 984
+    },
+    {
+      "epoch": 0.7975708502024291,
+      "grad_norm": 1.0762090702083147,
+      "learning_rate": 1.6940741952206342e-05,
+      "loss": 1.6346,
+      "step": 985
+    },
+    {
+      "epoch": 0.7983805668016194,
+      "grad_norm": 1.0669244133657225,
+      "learning_rate": 1.69345515820533e-05,
+      "loss": 1.6763,
+      "step": 986
+    },
+    {
+      "epoch": 0.7991902834008097,
+      "grad_norm": 1.062879048024205,
+      "learning_rate": 1.6928356088676325e-05,
+      "loss": 1.7014,
+      "step": 987
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0947855744497017,
+      "learning_rate": 1.6922155476652625e-05,
+      "loss": 1.631,
+      "step": 988
+    },
+    {
+      "epoch": 0.8008097165991903,
+      "grad_norm": 1.034452979238415,
+      "learning_rate": 1.6915949750563202e-05,
+      "loss": 1.6141,
+      "step": 989
+    },
+    {
+      "epoch": 0.8016194331983806,
+      "grad_norm": 1.0485411735651209,
+      "learning_rate": 1.6909738914992812e-05,
+      "loss": 1.6518,
+      "step": 990
+    },
+    {
+      "epoch": 0.8024291497975709,
+      "grad_norm": 1.055581466504577,
+      "learning_rate": 1.6903522974530005e-05,
+      "loss": 1.613,
+      "step": 991
+    },
+    {
+      "epoch": 0.8032388663967611,
+      "grad_norm": 1.0560876471593497,
+      "learning_rate": 1.6897301933767103e-05,
+      "loss": 1.7046,
+      "step": 992
+    },
+    {
+      "epoch": 0.8040485829959514,
+      "grad_norm": 1.1003223798755561,
+      "learning_rate": 1.6891075797300184e-05,
+      "loss": 1.6625,
+      "step": 993
+    },
+    {
+      "epoch": 0.8048582995951417,
+      "grad_norm": 1.1021700834191077,
+      "learning_rate": 1.6884844569729097e-05,
+      "loss": 1.6678,
+      "step": 994
+    },
+    {
+      "epoch": 0.805668016194332,
+      "grad_norm": 1.1166959526421196,
+      "learning_rate": 1.6878608255657457e-05,
+      "loss": 1.6381,
+      "step": 995
+    },
+    {
+      "epoch": 0.8064777327935223,
+      "grad_norm": 1.065512317839742,
+      "learning_rate": 1.687236685969263e-05,
+      "loss": 1.6296,
+      "step": 996
+    },
+    {
+      "epoch": 0.8072874493927126,
+      "grad_norm": 1.1156855620907848,
+      "learning_rate": 1.6866120386445737e-05,
+      "loss": 1.6858,
+      "step": 997
+    },
+    {
+      "epoch": 0.8080971659919028,
+      "grad_norm": 1.0953706109586931,
+      "learning_rate": 1.6859868840531654e-05,
+      "loss": 1.7635,
+      "step": 998
+    },
+    {
+      "epoch": 0.8089068825910931,
+      "grad_norm": 1.0233507628776497,
+      "learning_rate": 1.6853612226569005e-05,
+      "loss": 1.6315,
+      "step": 999
+    },
+    {
+      "epoch": 0.8097165991902834,
+      "grad_norm": 1.1751028012727203,
+      "learning_rate": 1.6847350549180148e-05,
+      "loss": 1.6798,
+      "step": 1000
+    },
+    {
+      "epoch": 0.8105263157894737,
+      "grad_norm": 1.128106775328797,
+      "learning_rate": 1.68410838129912e-05,
+      "loss": 1.667,
+      "step": 1001
+    },
+    {
+      "epoch": 0.8113360323886639,
+      "grad_norm": 1.0812082492824788,
+      "learning_rate": 1.6834812022632e-05,
+      "loss": 1.6104,
+      "step": 1002
+    },
+    {
+      "epoch": 0.8121457489878543,
+      "grad_norm": 1.1490248467998732,
+      "learning_rate": 1.682853518273612e-05,
+      "loss": 1.6542,
+      "step": 1003
+    },
+    {
+      "epoch": 0.8129554655870446,
+      "grad_norm": 1.0715868026423407,
+      "learning_rate": 1.6822253297940876e-05,
+      "loss": 1.6435,
+      "step": 1004
+    },
+    {
+      "epoch": 0.8137651821862348,
+      "grad_norm": 1.117058675593047,
+      "learning_rate": 1.6815966372887305e-05,
+      "loss": 1.6082,
+      "step": 1005
+    },
+    {
+      "epoch": 0.8145748987854251,
+      "grad_norm": 1.1438973084718087,
+      "learning_rate": 1.6809674412220166e-05,
+      "loss": 1.6409,
+      "step": 1006
+    },
+    {
+      "epoch": 0.8153846153846154,
+      "grad_norm": 1.1167195347301986,
+      "learning_rate": 1.6803377420587935e-05,
+      "loss": 1.6139,
+      "step": 1007
+    },
+    {
+      "epoch": 0.8161943319838056,
+      "grad_norm": 1.050529006455886,
+      "learning_rate": 1.679707540264281e-05,
+      "loss": 1.6202,
+      "step": 1008
+    },
+    {
+      "epoch": 0.8170040485829959,
+      "grad_norm": 1.0456426778952734,
+      "learning_rate": 1.6790768363040704e-05,
+      "loss": 1.6511,
+      "step": 1009
+    },
+    {
+      "epoch": 0.8178137651821862,
+      "grad_norm": 1.0744209810818393,
+      "learning_rate": 1.6784456306441234e-05,
+      "loss": 1.6343,
+      "step": 1010
+    },
+    {
+      "epoch": 0.8186234817813766,
+      "grad_norm": 1.0454475493984707,
+      "learning_rate": 1.6778139237507728e-05,
+      "loss": 1.6411,
+      "step": 1011
+    },
+    {
+      "epoch": 0.8194331983805668,
+      "grad_norm": 1.1450988763255063,
+      "learning_rate": 1.6771817160907214e-05,
+      "loss": 1.6547,
+      "step": 1012
+    },
+    {
+      "epoch": 0.8202429149797571,
+      "grad_norm": 1.0820294221210558,
+      "learning_rate": 1.6765490081310426e-05,
+      "loss": 1.7049,
+      "step": 1013
+    },
+    {
+      "epoch": 0.8210526315789474,
+      "grad_norm": 1.0795056916204937,
+      "learning_rate": 1.6759158003391783e-05,
+      "loss": 1.6503,
+      "step": 1014
+    },
+    {
+      "epoch": 0.8218623481781376,
+      "grad_norm": 1.0754101028602392,
+      "learning_rate": 1.675282093182941e-05,
+      "loss": 1.6015,
+      "step": 1015
+    },
+    {
+      "epoch": 0.8226720647773279,
+      "grad_norm": 1.0902544138244956,
+      "learning_rate": 1.674647887130511e-05,
+      "loss": 1.6625,
+      "step": 1016
+    },
+    {
+      "epoch": 0.8234817813765182,
+      "grad_norm": 1.0728952819859472,
+      "learning_rate": 1.674013182650438e-05,
+      "loss": 1.6318,
+      "step": 1017
+    },
+    {
+      "epoch": 0.8242914979757086,
+      "grad_norm": 1.087191133019499,
+      "learning_rate": 1.673377980211639e-05,
+      "loss": 1.607,
+      "step": 1018
+    },
+    {
+      "epoch": 0.8251012145748988,
+      "grad_norm": 1.092471584016294,
+      "learning_rate": 1.6727422802834e-05,
+      "loss": 1.6151,
+      "step": 1019
+    },
+    {
+      "epoch": 0.8259109311740891,
+      "grad_norm": 1.0942633386482077,
+      "learning_rate": 1.672106083335374e-05,
+      "loss": 1.6538,
+      "step": 1020
+    },
+    {
+      "epoch": 0.8267206477732794,
+      "grad_norm": 1.0611257052505665,
+      "learning_rate": 1.671469389837581e-05,
+      "loss": 1.6813,
+      "step": 1021
+    },
+    {
+      "epoch": 0.8275303643724696,
+      "grad_norm": 1.1243190686891869,
+      "learning_rate": 1.6708322002604085e-05,
+      "loss": 1.6294,
+      "step": 1022
+    },
+    {
+      "epoch": 0.8283400809716599,
+      "grad_norm": 1.1202657006614518,
+      "learning_rate": 1.6701945150746094e-05,
+      "loss": 1.6243,
+      "step": 1023
+    },
+    {
+      "epoch": 0.8291497975708502,
+      "grad_norm": 1.1727083505566944,
+      "learning_rate": 1.6695563347513036e-05,
+      "loss": 1.631,
+      "step": 1024
+    },
+    {
+      "epoch": 0.8299595141700404,
+      "grad_norm": 1.0959511539538467,
+      "learning_rate": 1.6689176597619773e-05,
+      "loss": 1.7506,
+      "step": 1025
+    },
+    {
+      "epoch": 0.8307692307692308,
+      "grad_norm": 1.1456195238430023,
+      "learning_rate": 1.6682784905784808e-05,
+      "loss": 1.6222,
+      "step": 1026
+    },
+    {
+      "epoch": 0.8315789473684211,
+      "grad_norm": 1.1142846493104133,
+      "learning_rate": 1.6676388276730305e-05,
+      "loss": 1.5693,
+      "step": 1027
+    },
+    {
+      "epoch": 0.8323886639676114,
+      "grad_norm": 1.054064976471123,
+      "learning_rate": 1.6669986715182064e-05,
+      "loss": 1.677,
+      "step": 1028
+    },
+    {
+      "epoch": 0.8331983805668016,
+      "grad_norm": 1.026287560156129,
+      "learning_rate": 1.6663580225869554e-05,
+      "loss": 1.711,
+      "step": 1029
+    },
+    {
+      "epoch": 0.8340080971659919,
+      "grad_norm": 1.1059699325557255,
+      "learning_rate": 1.6657168813525855e-05,
+      "loss": 1.6314,
+      "step": 1030
+    },
+    {
+      "epoch": 0.8348178137651822,
+      "grad_norm": 1.0663812567827384,
+      "learning_rate": 1.6650752482887698e-05,
+      "loss": 1.6292,
+      "step": 1031
+    },
+    {
+      "epoch": 0.8356275303643724,
+      "grad_norm": 1.1308684576404386,
+      "learning_rate": 1.6644331238695454e-05,
+      "loss": 1.6018,
+      "step": 1032
+    },
+    {
+      "epoch": 0.8364372469635628,
+      "grad_norm": 1.0658872985205445,
+      "learning_rate": 1.6637905085693113e-05,
+      "loss": 1.603,
+      "step": 1033
+    },
+    {
+      "epoch": 0.8372469635627531,
+      "grad_norm": 1.0560438364461333,
+      "learning_rate": 1.6631474028628298e-05,
+      "loss": 1.6701,
+      "step": 1034
+    },
+    {
+      "epoch": 0.8380566801619433,
+      "grad_norm": 1.097675248377845,
+      "learning_rate": 1.662503807225225e-05,
+      "loss": 1.6673,
+      "step": 1035
+    },
+    {
+      "epoch": 0.8388663967611336,
+      "grad_norm": 1.1463142445281125,
+      "learning_rate": 1.6618597221319835e-05,
+      "loss": 1.6401,
+      "step": 1036
+    },
+    {
+      "epoch": 0.8396761133603239,
+      "grad_norm": 1.0991588788894604,
+      "learning_rate": 1.6612151480589526e-05,
+      "loss": 1.6822,
+      "step": 1037
+    },
+    {
+      "epoch": 0.8404858299595142,
+      "grad_norm": 1.1432757857765516,
+      "learning_rate": 1.6605700854823427e-05,
+      "loss": 1.6552,
+      "step": 1038
+    },
+    {
+      "epoch": 0.8412955465587044,
+      "grad_norm": 1.0996521956343472,
+      "learning_rate": 1.659924534878723e-05,
+      "loss": 1.6155,
+      "step": 1039
+    },
+    {
+      "epoch": 0.8421052631578947,
+      "grad_norm": 1.2092130985611158,
+      "learning_rate": 1.659278496725024e-05,
+      "loss": 1.6697,
+      "step": 1040
+    },
+    {
+      "epoch": 0.8429149797570851,
+      "grad_norm": 1.1859643866916718,
+      "learning_rate": 1.6586319714985372e-05,
+      "loss": 1.6514,
+      "step": 1041
+    },
+    {
+      "epoch": 0.8437246963562753,
+      "grad_norm": 1.096493975856786,
+      "learning_rate": 1.6579849596769132e-05,
+      "loss": 1.6892,
+      "step": 1042
+    },
+    {
+      "epoch": 0.8445344129554656,
+      "grad_norm": 1.041583638770386,
+      "learning_rate": 1.6573374617381622e-05,
+      "loss": 1.5834,
+      "step": 1043
+    },
+    {
+      "epoch": 0.8453441295546559,
+      "grad_norm": 1.256429469555766,
+      "learning_rate": 1.656689478160653e-05,
+      "loss": 1.5953,
+      "step": 1044
+    },
+    {
+      "epoch": 0.8461538461538461,
+      "grad_norm": 1.1156384269735415,
+      "learning_rate": 1.6560410094231144e-05,
+      "loss": 1.608,
+      "step": 1045
+    },
+    {
+      "epoch": 0.8469635627530364,
+      "grad_norm": 1.2175373188796819,
+      "learning_rate": 1.655392056004633e-05,
+      "loss": 1.7043,
+      "step": 1046
+    },
+    {
+      "epoch": 0.8477732793522267,
+      "grad_norm": 1.1837512398752925,
+      "learning_rate": 1.6547426183846527e-05,
+      "loss": 1.6364,
+      "step": 1047
+    },
+    {
+      "epoch": 0.848582995951417,
+      "grad_norm": 1.0731112756132954,
+      "learning_rate": 1.6540926970429768e-05,
+      "loss": 1.6826,
+      "step": 1048
+    },
+    {
+      "epoch": 0.8493927125506073,
+      "grad_norm": 1.2861445088088037,
+      "learning_rate": 1.6534422924597647e-05,
+      "loss": 1.671,
+      "step": 1049
+    },
+    {
+      "epoch": 0.8502024291497976,
+      "grad_norm": 1.2667796024188396,
+      "learning_rate": 1.6527914051155328e-05,
+      "loss": 1.6239,
+      "step": 1050
+    },
+    {
+      "epoch": 0.8510121457489879,
+      "grad_norm": 1.0621089700252984,
+      "learning_rate": 1.652140035491155e-05,
+      "loss": 1.6278,
+      "step": 1051
+    },
+    {
+      "epoch": 0.8518218623481781,
+      "grad_norm": 1.1959034368355665,
+      "learning_rate": 1.6514881840678606e-05,
+      "loss": 1.6383,
+      "step": 1052
+    },
+    {
+      "epoch": 0.8526315789473684,
+      "grad_norm": 1.1679490573250864,
+      "learning_rate": 1.650835851327236e-05,
+      "loss": 1.6685,
+      "step": 1053
+    },
+    {
+      "epoch": 0.8534412955465587,
+      "grad_norm": 1.0724625429268337,
+      "learning_rate": 1.6501830377512214e-05,
+      "loss": 1.6101,
+      "step": 1054
+    },
+    {
+      "epoch": 0.854251012145749,
+      "grad_norm": 1.0992880410878538,
+      "learning_rate": 1.6495297438221145e-05,
+      "loss": 1.6482,
+      "step": 1055
+    },
+    {
+      "epoch": 0.8550607287449393,
+      "grad_norm": 1.2082881920964228,
+      "learning_rate": 1.6488759700225663e-05,
+      "loss": 1.6947,
+      "step": 1056
+    },
+    {
+      "epoch": 0.8558704453441296,
+      "grad_norm": 1.064085800343604,
+      "learning_rate": 1.6482217168355824e-05,
+      "loss": 1.7083,
+      "step": 1057
+    },
+    {
+      "epoch": 0.8566801619433199,
+      "grad_norm": 1.2300421861334183,
+      "learning_rate": 1.647566984744523e-05,
+      "loss": 1.6707,
+      "step": 1058
+    },
+    {
+      "epoch": 0.8574898785425101,
+      "grad_norm": 1.168895681590581,
+      "learning_rate": 1.6469117742331023e-05,
+      "loss": 1.5947,
+      "step": 1059
+    },
+    {
+      "epoch": 0.8582995951417004,
+      "grad_norm": 1.0860055758849936,
+      "learning_rate": 1.6462560857853876e-05,
+      "loss": 1.6144,
+      "step": 1060
+    },
+    {
+      "epoch": 0.8591093117408907,
+      "grad_norm": 1.2059795818651675,
+      "learning_rate": 1.645599919885799e-05,
+      "loss": 1.7368,
+      "step": 1061
+    },
+    {
+      "epoch": 0.8599190283400809,
+      "grad_norm": 1.1705528700907784,
+      "learning_rate": 1.6449432770191104e-05,
+      "loss": 1.7067,
+      "step": 1062
+    },
+    {
+      "epoch": 0.8607287449392712,
+      "grad_norm": 1.0305975110654955,
+      "learning_rate": 1.6442861576704467e-05,
+      "loss": 1.6808,
+      "step": 1063
+    },
+    {
+      "epoch": 0.8615384615384616,
+      "grad_norm": 1.0672600681647995,
+      "learning_rate": 1.6436285623252863e-05,
+      "loss": 1.6516,
+      "step": 1064
+    },
+    {
+      "epoch": 0.8623481781376519,
+      "grad_norm": 1.0921123850603622,
+      "learning_rate": 1.6429704914694573e-05,
+      "loss": 1.6578,
+      "step": 1065
+    },
+    {
+      "epoch": 0.8631578947368421,
+      "grad_norm": 1.1420424627061119,
+      "learning_rate": 1.6423119455891412e-05,
+      "loss": 1.6166,
+      "step": 1066
+    },
+    {
+      "epoch": 0.8639676113360324,
+      "grad_norm": 1.1974027318652682,
+      "learning_rate": 1.6416529251708695e-05,
+      "loss": 1.5715,
+      "step": 1067
+    },
+    {
+      "epoch": 0.8647773279352227,
+      "grad_norm": 1.0470830216699876,
+      "learning_rate": 1.6409934307015237e-05,
+      "loss": 1.6083,
+      "step": 1068
+    },
+    {
+      "epoch": 0.8655870445344129,
+      "grad_norm": 1.070487613342487,
+      "learning_rate": 1.6403334626683373e-05,
+      "loss": 1.596,
+      "step": 1069
+    },
+    {
+      "epoch": 0.8663967611336032,
+      "grad_norm": 1.1529368703348062,
+      "learning_rate": 1.6396730215588913e-05,
+      "loss": 1.7128,
+      "step": 1070
+    },
+    {
+      "epoch": 0.8672064777327935,
+      "grad_norm": 1.1573095803039568,
+      "learning_rate": 1.639012107861118e-05,
+      "loss": 1.5791,
+      "step": 1071
+    },
+    {
+      "epoch": 0.8680161943319838,
+      "grad_norm": 1.0795488410207628,
+      "learning_rate": 1.6383507220632983e-05,
+      "loss": 1.6544,
+      "step": 1072
+    },
+    {
+      "epoch": 0.8688259109311741,
+      "grad_norm": 1.037145472528182,
+      "learning_rate": 1.6376888646540617e-05,
+      "loss": 1.6401,
+      "step": 1073
+    },
+    {
+      "epoch": 0.8696356275303644,
+      "grad_norm": 1.060717046272381,
+      "learning_rate": 1.6370265361223864e-05,
+      "loss": 1.6801,
+      "step": 1074
+    },
+    {
+      "epoch": 0.8704453441295547,
+      "grad_norm": 1.0964370572935223,
+      "learning_rate": 1.6363637369575984e-05,
+      "loss": 1.6636,
+      "step": 1075
+    },
+    {
+      "epoch": 0.8712550607287449,
+      "grad_norm": 1.0310388617035904,
+      "learning_rate": 1.6357004676493716e-05,
+      "loss": 1.5856,
+      "step": 1076
+    },
+    {
+      "epoch": 0.8720647773279352,
+      "grad_norm": 1.0168478194178103,
+      "learning_rate": 1.635036728687727e-05,
+      "loss": 1.5626,
+      "step": 1077
+    },
+    {
+      "epoch": 0.8728744939271255,
+      "grad_norm": 1.1227525434781638,
+      "learning_rate": 1.6343725205630335e-05,
+      "loss": 1.6922,
+      "step": 1078
+    },
+    {
+      "epoch": 0.8736842105263158,
+      "grad_norm": 1.0084316594907183,
+      "learning_rate": 1.633707843766005e-05,
+      "loss": 1.6777,
+      "step": 1079
+    },
+    {
+      "epoch": 0.8744939271255061,
+      "grad_norm": 1.0733928985684549,
+      "learning_rate": 1.633042698787703e-05,
+      "loss": 1.6883,
+      "step": 1080
+    },
+    {
+      "epoch": 0.8753036437246964,
+      "grad_norm": 1.0420921585152885,
+      "learning_rate": 1.632377086119534e-05,
+      "loss": 1.5657,
+      "step": 1081
+    },
+    {
+      "epoch": 0.8761133603238866,
+      "grad_norm": 1.0479710213227003,
+      "learning_rate": 1.631711006253251e-05,
+      "loss": 1.7318,
+      "step": 1082
+    },
+    {
+      "epoch": 0.8769230769230769,
+      "grad_norm": 1.0675821904258518,
+      "learning_rate": 1.6310444596809514e-05,
+      "loss": 1.6281,
+      "step": 1083
+    },
+    {
+      "epoch": 0.8777327935222672,
+      "grad_norm": 1.0092848547062299,
+      "learning_rate": 1.6303774468950776e-05,
+      "loss": 1.5721,
+      "step": 1084
+    },
+    {
+      "epoch": 0.8785425101214575,
+      "grad_norm": 1.1037917612800165,
+      "learning_rate": 1.6297099683884163e-05,
+      "loss": 1.588,
+      "step": 1085
+    },
+    {
+      "epoch": 0.8793522267206477,
+      "grad_norm": 1.081509721545105,
+      "learning_rate": 1.629042024654099e-05,
+      "loss": 1.6159,
+      "step": 1086
+    },
+    {
+      "epoch": 0.8801619433198381,
+      "grad_norm": 1.0836597364116385,
+      "learning_rate": 1.6283736161855995e-05,
+      "loss": 1.6192,
+      "step": 1087
+    },
+    {
+      "epoch": 0.8809716599190284,
+      "grad_norm": 1.0587543748723958,
+      "learning_rate": 1.6277047434767364e-05,
+      "loss": 1.5687,
+      "step": 1088
+    },
+    {
+      "epoch": 0.8817813765182186,
+      "grad_norm": 1.0747135995302248,
+      "learning_rate": 1.6270354070216704e-05,
+      "loss": 1.6513,
+      "step": 1089
+    },
+    {
+      "epoch": 0.8825910931174089,
+      "grad_norm": 1.0869005500984592,
+      "learning_rate": 1.626365607314905e-05,
+      "loss": 1.6559,
+      "step": 1090
+    },
+    {
+      "epoch": 0.8834008097165992,
+      "grad_norm": 1.0595343275908542,
+      "learning_rate": 1.625695344851286e-05,
+      "loss": 1.6598,
+      "step": 1091
+    },
+    {
+      "epoch": 0.8842105263157894,
+      "grad_norm": 1.1107620614268396,
+      "learning_rate": 1.6250246201260017e-05,
+      "loss": 1.6897,
+      "step": 1092
+    },
+    {
+      "epoch": 0.8850202429149797,
+      "grad_norm": 1.1907108746856276,
+      "learning_rate": 1.624353433634581e-05,
+      "loss": 1.6426,
+      "step": 1093
+    },
+    {
+      "epoch": 0.8858299595141701,
+      "grad_norm": 1.046609129170575,
+      "learning_rate": 1.6236817858728937e-05,
+      "loss": 1.5677,
+      "step": 1094
+    },
+    {
+      "epoch": 0.8866396761133604,
+      "grad_norm": 1.0744260551582498,
+      "learning_rate": 1.6230096773371514e-05,
+      "loss": 1.6268,
+      "step": 1095
+    },
+    {
+      "epoch": 0.8874493927125506,
+      "grad_norm": 1.1340503009391705,
+      "learning_rate": 1.622337108523906e-05,
+      "loss": 1.6032,
+      "step": 1096
+    },
+    {
+      "epoch": 0.8882591093117409,
+      "grad_norm": 1.07216583644219,
+      "learning_rate": 1.621664079930049e-05,
+      "loss": 1.66,
+      "step": 1097
+    },
+    {
+      "epoch": 0.8890688259109312,
+      "grad_norm": 1.137011474293611,
+      "learning_rate": 1.620990592052811e-05,
+      "loss": 1.6506,
+      "step": 1098
+    },
+    {
+      "epoch": 0.8898785425101214,
+      "grad_norm": 1.0925701713040545,
+      "learning_rate": 1.620316645389764e-05,
+      "loss": 1.6441,
+      "step": 1099
+    },
+    {
+      "epoch": 0.8906882591093117,
+      "grad_norm": 0.9809230441134872,
+      "learning_rate": 1.619642240438816e-05,
+      "loss": 1.6116,
+      "step": 1100
+    },
+    {
+      "epoch": 0.891497975708502,
+      "grad_norm": 1.0227226563848153,
+      "learning_rate": 1.618967377698216e-05,
+      "loss": 1.6869,
+      "step": 1101
+    },
+    {
+      "epoch": 0.8923076923076924,
+      "grad_norm": 1.0936896377565115,
+      "learning_rate": 1.6182920576665508e-05,
+      "loss": 1.6799,
+      "step": 1102
+    },
+    {
+      "epoch": 0.8931174089068826,
+      "grad_norm": 1.064709684635027,
+      "learning_rate": 1.6176162808427437e-05,
+      "loss": 1.6547,
+      "step": 1103
+    },
+    {
+      "epoch": 0.8939271255060729,
+      "grad_norm": 1.089323924603956,
+      "learning_rate": 1.6169400477260566e-05,
+      "loss": 1.6514,
+      "step": 1104
+    },
+    {
+      "epoch": 0.8947368421052632,
+      "grad_norm": 1.051251367154035,
+      "learning_rate": 1.616263358816089e-05,
+      "loss": 1.5567,
+      "step": 1105
+    },
+    {
+      "epoch": 0.8955465587044534,
+      "grad_norm": 1.0955130855844533,
+      "learning_rate": 1.6155862146127757e-05,
+      "loss": 1.6892,
+      "step": 1106
+    },
+    {
+      "epoch": 0.8963562753036437,
+      "grad_norm": 1.2427921629069014,
+      "learning_rate": 1.6149086156163893e-05,
+      "loss": 1.6672,
+      "step": 1107
+    },
+    {
+      "epoch": 0.897165991902834,
+      "grad_norm": 1.0531149381114697,
+      "learning_rate": 1.6142305623275367e-05,
+      "loss": 1.6756,
+      "step": 1108
+    },
+    {
+      "epoch": 0.8979757085020242,
+      "grad_norm": 1.1185947127125788,
+      "learning_rate": 1.6135520552471625e-05,
+      "loss": 1.6263,
+      "step": 1109
+    },
+    {
+      "epoch": 0.8987854251012146,
+      "grad_norm": 1.0114886200304354,
+      "learning_rate": 1.612873094876545e-05,
+      "loss": 1.6887,
+      "step": 1110
+    },
+    {
+      "epoch": 0.8995951417004049,
+      "grad_norm": 1.1380754110775286,
+      "learning_rate": 1.612193681717298e-05,
+      "loss": 1.6664,
+      "step": 1111
+    },
+    {
+      "epoch": 0.9004048582995952,
+      "grad_norm": 1.0844337360811653,
+      "learning_rate": 1.611513816271369e-05,
+      "loss": 1.6717,
+      "step": 1112
+    },
+    {
+      "epoch": 0.9012145748987854,
+      "grad_norm": 1.0684169391575367,
+      "learning_rate": 1.6108334990410413e-05,
+      "loss": 1.6056,
+      "step": 1113
+    },
+    {
+      "epoch": 0.9020242914979757,
+      "grad_norm": 1.0828647321774048,
+      "learning_rate": 1.610152730528931e-05,
+      "loss": 1.6916,
+      "step": 1114
+    },
+    {
+      "epoch": 0.902834008097166,
+      "grad_norm": 1.0539257127891122,
+      "learning_rate": 1.6094715112379874e-05,
+      "loss": 1.6259,
+      "step": 1115
+    },
+    {
+      "epoch": 0.9036437246963562,
+      "grad_norm": 1.0495197445873474,
+      "learning_rate": 1.6087898416714928e-05,
+      "loss": 1.6449,
+      "step": 1116
+    },
+    {
+      "epoch": 0.9044534412955466,
+      "grad_norm": 1.037832176782658,
+      "learning_rate": 1.608107722333063e-05,
+      "loss": 1.5811,
+      "step": 1117
+    },
+    {
+      "epoch": 0.9052631578947369,
+      "grad_norm": 1.1204020070782086,
+      "learning_rate": 1.607425153726645e-05,
+      "loss": 1.6671,
+      "step": 1118
+    },
+    {
+      "epoch": 0.9060728744939271,
+      "grad_norm": 1.0947786620111954,
+      "learning_rate": 1.6067421363565185e-05,
+      "loss": 1.6438,
+      "step": 1119
+    },
+    {
+      "epoch": 0.9068825910931174,
+      "grad_norm": 1.1703591478489133,
+      "learning_rate": 1.6060586707272943e-05,
+      "loss": 1.5142,
+      "step": 1120
+    },
+    {
+      "epoch": 0.9076923076923077,
+      "grad_norm": 1.1101060613426852,
+      "learning_rate": 1.6053747573439147e-05,
+      "loss": 1.6466,
+      "step": 1121
+    },
+    {
+      "epoch": 0.908502024291498,
+      "grad_norm": 0.9844797162954405,
+      "learning_rate": 1.6046903967116532e-05,
+      "loss": 1.5892,
+      "step": 1122
+    },
+    {
+      "epoch": 0.9093117408906882,
+      "grad_norm": 1.2066011149235545,
+      "learning_rate": 1.604005589336112e-05,
+      "loss": 1.6379,
+      "step": 1123
+    },
+    {
+      "epoch": 0.9101214574898785,
+      "grad_norm": 1.12551790929863,
+      "learning_rate": 1.6033203357232255e-05,
+      "loss": 1.5908,
+      "step": 1124
+    },
+    {
+      "epoch": 0.9109311740890689,
+      "grad_norm": 1.0539443046535852,
+      "learning_rate": 1.6026346363792565e-05,
+      "loss": 1.6407,
+      "step": 1125
+    },
+    {
+      "epoch": 0.9117408906882591,
+      "grad_norm": 1.2718959974915909,
+      "learning_rate": 1.6019484918107977e-05,
+      "loss": 1.6657,
+      "step": 1126
+    },
+    {
+      "epoch": 0.9125506072874494,
+      "grad_norm": 1.1172864236142332,
+      "learning_rate": 1.60126190252477e-05,
+      "loss": 1.5977,
+      "step": 1127
+    },
+    {
+      "epoch": 0.9133603238866397,
+      "grad_norm": 1.029629174783446,
+      "learning_rate": 1.600574869028423e-05,
+      "loss": 1.5827,
+      "step": 1128
+    },
+    {
+      "epoch": 0.91417004048583,
+      "grad_norm": 1.1276448324458075,
+      "learning_rate": 1.599887391829336e-05,
+      "loss": 1.5556,
+      "step": 1129
+    },
+    {
+      "epoch": 0.9149797570850202,
+      "grad_norm": 1.234475202377491,
+      "learning_rate": 1.599199471435414e-05,
+      "loss": 1.6962,
+      "step": 1130
+    },
+    {
+      "epoch": 0.9157894736842105,
+      "grad_norm": 1.0658020028121022,
+      "learning_rate": 1.5985111083548905e-05,
+      "loss": 1.6202,
+      "step": 1131
+    },
+    {
+      "epoch": 0.9165991902834008,
+      "grad_norm": 1.1895352614746753,
+      "learning_rate": 1.5978223030963257e-05,
+      "loss": 1.5915,
+      "step": 1132
+    },
+    {
+      "epoch": 0.9174089068825911,
+      "grad_norm": 1.1294963341176811,
+      "learning_rate": 1.5971330561686073e-05,
+      "loss": 1.6618,
+      "step": 1133
+    },
+    {
+      "epoch": 0.9182186234817814,
+      "grad_norm": 1.0383726295620532,
+      "learning_rate": 1.596443368080948e-05,
+      "loss": 1.7049,
+      "step": 1134
+    },
+    {
+      "epoch": 0.9190283400809717,
+      "grad_norm": 1.0765024500512697,
+      "learning_rate": 1.5957532393428872e-05,
+      "loss": 1.6578,
+      "step": 1135
+    },
+    {
+      "epoch": 0.9198380566801619,
+      "grad_norm": 1.0435257860758407,
+      "learning_rate": 1.5950626704642898e-05,
+      "loss": 1.5491,
+      "step": 1136
+    },
+    {
+      "epoch": 0.9206477732793522,
+      "grad_norm": 1.040007314475756,
+      "learning_rate": 1.594371661955346e-05,
+      "loss": 1.5721,
+      "step": 1137
+    },
+    {
+      "epoch": 0.9214574898785425,
+      "grad_norm": 1.1078111485151643,
+      "learning_rate": 1.5936802143265708e-05,
+      "loss": 1.6908,
+      "step": 1138
+    },
+    {
+      "epoch": 0.9222672064777327,
+      "grad_norm": 1.1320470811777017,
+      "learning_rate": 1.592988328088803e-05,
+      "loss": 1.6745,
+      "step": 1139
+    },
+    {
+      "epoch": 0.9230769230769231,
+      "grad_norm": 1.016464328891757,
+      "learning_rate": 1.5922960037532057e-05,
+      "loss": 1.5978,
+      "step": 1140
+    },
+    {
+      "epoch": 0.9238866396761134,
+      "grad_norm": 1.0843626078556605,
+      "learning_rate": 1.5916032418312665e-05,
+      "loss": 1.5708,
+      "step": 1141
+    },
+    {
+      "epoch": 0.9246963562753037,
+      "grad_norm": 1.1378816169740498,
+      "learning_rate": 1.5909100428347953e-05,
+      "loss": 1.7212,
+      "step": 1142
+    },
+    {
+      "epoch": 0.9255060728744939,
+      "grad_norm": 1.0606143686664404,
+      "learning_rate": 1.590216407275925e-05,
+      "loss": 1.5639,
+      "step": 1143
+    },
+    {
+      "epoch": 0.9263157894736842,
+      "grad_norm": 1.3367149432598577,
+      "learning_rate": 1.5895223356671116e-05,
+      "loss": 1.6906,
+      "step": 1144
+    },
+    {
+      "epoch": 0.9271255060728745,
+      "grad_norm": 1.0416656067812,
+      "learning_rate": 1.588827828521133e-05,
+      "loss": 1.5926,
+      "step": 1145
+    },
+    {
+      "epoch": 0.9279352226720647,
+      "grad_norm": 1.0644446557633547,
+      "learning_rate": 1.5881328863510885e-05,
+      "loss": 1.6572,
+      "step": 1146
+    },
+    {
+      "epoch": 0.928744939271255,
+      "grad_norm": 1.0307821440286953,
+      "learning_rate": 1.5874375096703993e-05,
+      "loss": 1.6722,
+      "step": 1147
+    },
+    {
+      "epoch": 0.9295546558704454,
+      "grad_norm": 1.0777429781962335,
+      "learning_rate": 1.5867416989928077e-05,
+      "loss": 1.6206,
+      "step": 1148
+    },
+    {
+      "epoch": 0.9303643724696357,
+      "grad_norm": 1.0968718914039448,
+      "learning_rate": 1.5860454548323755e-05,
+      "loss": 1.6033,
+      "step": 1149
+    },
+    {
+      "epoch": 0.9311740890688259,
+      "grad_norm": 1.094113134525346,
+      "learning_rate": 1.585348777703486e-05,
+      "loss": 1.6282,
+      "step": 1150
+    },
+    {
+      "epoch": 0.9319838056680162,
+      "grad_norm": 1.110942326892508,
+      "learning_rate": 1.5846516681208425e-05,
+      "loss": 1.6293,
+      "step": 1151
+    },
+    {
+      "epoch": 0.9327935222672065,
+      "grad_norm": 1.1089357151037313,
+      "learning_rate": 1.5839541265994663e-05,
+      "loss": 1.7625,
+      "step": 1152
+    },
+    {
+      "epoch": 0.9336032388663967,
+      "grad_norm": 1.0813117854976675,
+      "learning_rate": 1.5832561536546998e-05,
+      "loss": 1.6626,
+      "step": 1153
+    },
+    {
+      "epoch": 0.934412955465587,
+      "grad_norm": 1.0395260288355859,
+      "learning_rate": 1.5825577498022027e-05,
+      "loss": 1.6422,
+      "step": 1154
+    },
+    {
+      "epoch": 0.9352226720647774,
+      "grad_norm": 1.0434934564308764,
+      "learning_rate": 1.581858915557953e-05,
+      "loss": 1.6427,
+      "step": 1155
+    },
+    {
+      "epoch": 0.9360323886639677,
+      "grad_norm": 1.1131287757276462,
+      "learning_rate": 1.5811596514382474e-05,
+      "loss": 1.6684,
+      "step": 1156
+    },
+    {
+      "epoch": 0.9368421052631579,
+      "grad_norm": 1.0535686300577007,
+      "learning_rate": 1.5804599579597007e-05,
+      "loss": 1.5772,
+      "step": 1157
+    },
+    {
+      "epoch": 0.9376518218623482,
+      "grad_norm": 1.0265557890806118,
+      "learning_rate": 1.5797598356392433e-05,
+      "loss": 1.6554,
+      "step": 1158
+    },
+    {
+      "epoch": 0.9384615384615385,
+      "grad_norm": 1.040297515122633,
+      "learning_rate": 1.5790592849941234e-05,
+      "loss": 1.6032,
+      "step": 1159
+    },
+    {
+      "epoch": 0.9392712550607287,
+      "grad_norm": 1.0584403749937723,
+      "learning_rate": 1.5783583065419054e-05,
+      "loss": 1.6032,
+      "step": 1160
+    },
+    {
+      "epoch": 0.940080971659919,
+      "grad_norm": 1.1762205201783962,
+      "learning_rate": 1.5776569008004705e-05,
+      "loss": 1.6195,
+      "step": 1161
+    },
+    {
+      "epoch": 0.9408906882591093,
+      "grad_norm": 1.116940246849939,
+      "learning_rate": 1.5769550682880143e-05,
+      "loss": 1.6004,
+      "step": 1162
+    },
+    {
+      "epoch": 0.9417004048582996,
+      "grad_norm": 1.0651667577950013,
+      "learning_rate": 1.5762528095230488e-05,
+      "loss": 1.584,
+      "step": 1163
+    },
+    {
+      "epoch": 0.9425101214574899,
+      "grad_norm": 1.1032179981000514,
+      "learning_rate": 1.5755501250244e-05,
+      "loss": 1.6358,
+      "step": 1164
+    },
+    {
+      "epoch": 0.9433198380566802,
+      "grad_norm": 1.1087189553464278,
+      "learning_rate": 1.5748470153112093e-05,
+      "loss": 1.5852,
+      "step": 1165
+    },
+    {
+      "epoch": 0.9441295546558705,
+      "grad_norm": 1.0912761006902574,
+      "learning_rate": 1.574143480902932e-05,
+      "loss": 1.6718,
+      "step": 1166
+    },
+    {
+      "epoch": 0.9449392712550607,
+      "grad_norm": 1.0720080749650522,
+      "learning_rate": 1.5734395223193367e-05,
+      "loss": 1.637,
+      "step": 1167
+    },
+    {
+      "epoch": 0.945748987854251,
+      "grad_norm": 1.1016199611874298,
+      "learning_rate": 1.5727351400805054e-05,
+      "loss": 1.7111,
+      "step": 1168
+    },
+    {
+      "epoch": 0.9465587044534413,
+      "grad_norm": 1.0389321277980663,
+      "learning_rate": 1.572030334706834e-05,
+      "loss": 1.6109,
+      "step": 1169
+    },
+    {
+      "epoch": 0.9473684210526315,
+      "grad_norm": 1.1153836189672202,
+      "learning_rate": 1.57132510671903e-05,
+      "loss": 1.6806,
+      "step": 1170
+    },
+    {
+      "epoch": 0.9481781376518219,
+      "grad_norm": 1.0852648007480221,
+      "learning_rate": 1.5706194566381136e-05,
+      "loss": 1.5954,
+      "step": 1171
+    },
+    {
+      "epoch": 0.9489878542510122,
+      "grad_norm": 1.1024355588691714,
+      "learning_rate": 1.5699133849854164e-05,
+      "loss": 1.7109,
+      "step": 1172
+    },
+    {
+      "epoch": 0.9497975708502024,
+      "grad_norm": 1.1103223087215968,
+      "learning_rate": 1.5692068922825826e-05,
+      "loss": 1.6344,
+      "step": 1173
+    },
+    {
+      "epoch": 0.9506072874493927,
+      "grad_norm": 1.0273649828440345,
+      "learning_rate": 1.5684999790515664e-05,
+      "loss": 1.6134,
+      "step": 1174
+    },
+    {
+      "epoch": 0.951417004048583,
+      "grad_norm": 1.025710171640773,
+      "learning_rate": 1.5677926458146327e-05,
+      "loss": 1.6368,
+      "step": 1175
+    },
+    {
+      "epoch": 0.9522267206477733,
+      "grad_norm": 1.080001734059546,
+      "learning_rate": 1.567084893094357e-05,
+      "loss": 1.6528,
+      "step": 1176
+    },
+    {
+      "epoch": 0.9530364372469635,
+      "grad_norm": 1.0606923317026444,
+      "learning_rate": 1.566376721413625e-05,
+      "loss": 1.5999,
+      "step": 1177
+    },
+    {
+      "epoch": 0.9538461538461539,
+      "grad_norm": 1.0103172442950479,
+      "learning_rate": 1.5656681312956316e-05,
+      "loss": 1.5915,
+      "step": 1178
+    },
+    {
+      "epoch": 0.9546558704453442,
+      "grad_norm": 1.1057352054726133,
+      "learning_rate": 1.5649591232638804e-05,
+      "loss": 1.6497,
+      "step": 1179
+    },
+    {
+      "epoch": 0.9554655870445344,
+      "grad_norm": 1.1047227327452962,
+      "learning_rate": 1.5642496978421842e-05,
+      "loss": 1.6919,
+      "step": 1180
+    },
+    {
+      "epoch": 0.9562753036437247,
+      "grad_norm": 1.0283713285578344,
+      "learning_rate": 1.563539855554665e-05,
+      "loss": 1.6248,
+      "step": 1181
+    },
+    {
+      "epoch": 0.957085020242915,
+      "grad_norm": 1.0214669851609182,
+      "learning_rate": 1.5628295969257515e-05,
+      "loss": 1.623,
+      "step": 1182
+    },
+    {
+      "epoch": 0.9578947368421052,
+      "grad_norm": 1.0696244337426537,
+      "learning_rate": 1.5621189224801797e-05,
+      "loss": 1.6771,
+      "step": 1183
+    },
+    {
+      "epoch": 0.9587044534412955,
+      "grad_norm": 1.0663459977571883,
+      "learning_rate": 1.5614078327429947e-05,
+      "loss": 1.7245,
+      "step": 1184
+    },
+    {
+      "epoch": 0.9595141700404858,
+      "grad_norm": 1.100787082491128,
+      "learning_rate": 1.560696328239547e-05,
+      "loss": 1.6435,
+      "step": 1185
+    },
+    {
+      "epoch": 0.9603238866396762,
+      "grad_norm": 1.0861397245205235,
+      "learning_rate": 1.559984409495493e-05,
+      "loss": 1.5699,
+      "step": 1186
+    },
+    {
+      "epoch": 0.9611336032388664,
+      "grad_norm": 1.0927150205723406,
+      "learning_rate": 1.5592720770367967e-05,
+      "loss": 1.6555,
+      "step": 1187
+    },
+    {
+      "epoch": 0.9619433198380567,
+      "grad_norm": 1.0739743395413222,
+      "learning_rate": 1.5585593313897267e-05,
+      "loss": 1.5801,
+      "step": 1188
+    },
+    {
+      "epoch": 0.962753036437247,
+      "grad_norm": 1.058740021769692,
+      "learning_rate": 1.5578461730808575e-05,
+      "loss": 1.6698,
+      "step": 1189
+    },
+    {
+      "epoch": 0.9635627530364372,
+      "grad_norm": 1.0525423850505513,
+      "learning_rate": 1.5571326026370676e-05,
+      "loss": 1.5322,
+      "step": 1190
+    },
+    {
+      "epoch": 0.9643724696356275,
+      "grad_norm": 1.075841311496856,
+      "learning_rate": 1.5564186205855407e-05,
+      "loss": 1.5897,
+      "step": 1191
+    },
+    {
+      "epoch": 0.9651821862348178,
+      "grad_norm": 1.1185012350730061,
+      "learning_rate": 1.5557042274537644e-05,
+      "loss": 1.632,
+      "step": 1192
+    },
+    {
+      "epoch": 0.965991902834008,
+      "grad_norm": 1.0818043247678821,
+      "learning_rate": 1.5549894237695302e-05,
+      "loss": 1.6112,
+      "step": 1193
+    },
+    {
+      "epoch": 0.9668016194331984,
+      "grad_norm": 1.0762571974215163,
+      "learning_rate": 1.5542742100609324e-05,
+      "loss": 1.5975,
+      "step": 1194
+    },
+    {
+      "epoch": 0.9676113360323887,
+      "grad_norm": 1.1294759095742624,
+      "learning_rate": 1.5535585868563688e-05,
+      "loss": 1.628,
+      "step": 1195
+    },
+    {
+      "epoch": 0.968421052631579,
+      "grad_norm": 1.0550884938111755,
+      "learning_rate": 1.552842554684539e-05,
+      "loss": 1.6819,
+      "step": 1196
+    },
+    {
+      "epoch": 0.9692307692307692,
+      "grad_norm": 1.1137617349520605,
+      "learning_rate": 1.5521261140744458e-05,
+      "loss": 1.5694,
+      "step": 1197
+    },
+    {
+      "epoch": 0.9700404858299595,
+      "grad_norm": 1.122314201391633,
+      "learning_rate": 1.551409265555393e-05,
+      "loss": 1.7246,
+      "step": 1198
+    },
+    {
+      "epoch": 0.9708502024291498,
+      "grad_norm": 1.0252937720177295,
+      "learning_rate": 1.5506920096569857e-05,
+      "loss": 1.5807,
+      "step": 1199
+    },
+    {
+      "epoch": 0.97165991902834,
+      "grad_norm": 1.066566322393118,
+      "learning_rate": 1.5499743469091303e-05,
+      "loss": 1.5806,
+      "step": 1200
+    },
+    {
+      "epoch": 0.9724696356275304,
+      "grad_norm": 1.1862181447108153,
+      "learning_rate": 1.5492562778420342e-05,
+      "loss": 1.6677,
+      "step": 1201
+    },
+    {
+      "epoch": 0.9732793522267207,
+      "grad_norm": 1.1600571833816153,
+      "learning_rate": 1.5485378029862034e-05,
+      "loss": 1.6181,
+      "step": 1202
+    },
+    {
+      "epoch": 0.974089068825911,
+      "grad_norm": 1.1209706305697864,
+      "learning_rate": 1.547818922872446e-05,
+      "loss": 1.595,
+      "step": 1203
+    },
+    {
+      "epoch": 0.9748987854251012,
+      "grad_norm": 1.2007659516216234,
+      "learning_rate": 1.547099638031867e-05,
+      "loss": 1.5757,
+      "step": 1204
+    },
+    {
+      "epoch": 0.9757085020242915,
+      "grad_norm": 1.087592209286536,
+      "learning_rate": 1.5463799489958727e-05,
+      "loss": 1.6574,
+      "step": 1205
+    },
+    {
+      "epoch": 0.9765182186234818,
+      "grad_norm": 1.0599754950787568,
+      "learning_rate": 1.5456598562961666e-05,
+      "loss": 1.6501,
+      "step": 1206
+    },
+    {
+      "epoch": 0.977327935222672,
+      "grad_norm": 1.2383281394995351,
+      "learning_rate": 1.544939360464751e-05,
+      "loss": 1.6212,
+      "step": 1207
+    },
+    {
+      "epoch": 0.9781376518218623,
+      "grad_norm": 1.1245330174866524,
+      "learning_rate": 1.5442184620339252e-05,
+      "loss": 1.6466,
+      "step": 1208
+    },
+    {
+      "epoch": 0.9789473684210527,
+      "grad_norm": 0.9799617502733805,
+      "learning_rate": 1.5434971615362875e-05,
+      "loss": 1.6056,
+      "step": 1209
+    },
+    {
+      "epoch": 0.979757085020243,
+      "grad_norm": 1.142502619460788,
+      "learning_rate": 1.542775459504732e-05,
+      "loss": 1.6629,
+      "step": 1210
+    },
+    {
+      "epoch": 0.9805668016194332,
+      "grad_norm": 1.0942507084406352,
+      "learning_rate": 1.5420533564724495e-05,
+      "loss": 1.6764,
+      "step": 1211
+    },
+    {
+      "epoch": 0.9813765182186235,
+      "grad_norm": 1.161749663929768,
+      "learning_rate": 1.5413308529729274e-05,
+      "loss": 1.6983,
+      "step": 1212
+    },
+    {
+      "epoch": 0.9821862348178138,
+      "grad_norm": 1.0441007142843908,
+      "learning_rate": 1.5406079495399495e-05,
+      "loss": 1.5531,
+      "step": 1213
+    },
+    {
+      "epoch": 0.982995951417004,
+      "grad_norm": 1.0129014720851601,
+      "learning_rate": 1.5398846467075937e-05,
+      "loss": 1.5144,
+      "step": 1214
+    },
+    {
+      "epoch": 0.9838056680161943,
+      "grad_norm": 1.0916694795654285,
+      "learning_rate": 1.5391609450102346e-05,
+      "loss": 1.5713,
+      "step": 1215
+    },
+    {
+      "epoch": 0.9846153846153847,
+      "grad_norm": 1.0119872612879202,
+      "learning_rate": 1.5384368449825395e-05,
+      "loss": 1.6537,
+      "step": 1216
+    },
+    {
+      "epoch": 0.9854251012145749,
+      "grad_norm": 1.089804356576127,
+      "learning_rate": 1.5377123471594723e-05,
+      "loss": 1.6905,
+      "step": 1217
+    },
+    {
+      "epoch": 0.9862348178137652,
+      "grad_norm": 1.018296330143969,
+      "learning_rate": 1.536987452076289e-05,
+      "loss": 1.5738,
+      "step": 1218
+    },
+    {
+      "epoch": 0.9870445344129555,
+      "grad_norm": 1.1463126951348903,
+      "learning_rate": 1.5362621602685394e-05,
+      "loss": 1.5971,
+      "step": 1219
+    },
+    {
+      "epoch": 0.9878542510121457,
+      "grad_norm": 1.0831047769241162,
+      "learning_rate": 1.5355364722720674e-05,
+      "loss": 1.6642,
+      "step": 1220
+    },
+    {
+      "epoch": 0.988663967611336,
+      "grad_norm": 1.0229539605576596,
+      "learning_rate": 1.5348103886230086e-05,
+      "loss": 1.6169,
+      "step": 1221
+    },
+    {
+      "epoch": 0.9894736842105263,
+      "grad_norm": 1.0706268272715531,
+      "learning_rate": 1.5340839098577912e-05,
+      "loss": 1.6369,
+      "step": 1222
+    },
+    {
+      "epoch": 0.9902834008097166,
+      "grad_norm": 1.076282820889626,
+      "learning_rate": 1.5333570365131353e-05,
+      "loss": 1.6125,
+      "step": 1223
+    },
+    {
+      "epoch": 0.9910931174089069,
+      "grad_norm": 1.0521363851747962,
+      "learning_rate": 1.5326297691260526e-05,
+      "loss": 1.6126,
+      "step": 1224
+    },
+    {
+      "epoch": 0.9919028340080972,
+      "grad_norm": 1.028881455867217,
+      "learning_rate": 1.5319021082338458e-05,
+      "loss": 1.6541,
+      "step": 1225
+    },
+    {
+      "epoch": 0.9927125506072875,
+      "grad_norm": 1.04183651822074,
+      "learning_rate": 1.5311740543741088e-05,
+      "loss": 1.6238,
+      "step": 1226
+    },
+    {
+      "epoch": 0.9935222672064777,
+      "grad_norm": 1.0858918280353071,
+      "learning_rate": 1.5304456080847247e-05,
+      "loss": 1.6627,
+      "step": 1227
+    },
+    {
+      "epoch": 0.994331983805668,
+      "grad_norm": 1.0035165826353316,
+      "learning_rate": 1.5297167699038673e-05,
+      "loss": 1.6204,
+      "step": 1228
+    },
+    {
+      "epoch": 0.9951417004048583,
+      "grad_norm": 1.0816188271259373,
+      "learning_rate": 1.5289875403700005e-05,
+      "loss": 1.6276,
+      "step": 1229
+    },
+    {
+      "epoch": 0.9959514170040485,
+      "grad_norm": 1.0599193319828009,
+      "learning_rate": 1.5282579200218762e-05,
+      "loss": 1.6398,
+      "step": 1230
+    },
+    {
+      "epoch": 0.9967611336032388,
+      "grad_norm": 1.0919178042931876,
+      "learning_rate": 1.5275279093985355e-05,
+      "loss": 1.5611,
+      "step": 1231
+    },
+    {
+      "epoch": 0.9975708502024292,
+      "grad_norm": 1.1107755668554948,
+      "learning_rate": 1.5267975090393078e-05,
+      "loss": 1.715,
+      "step": 1232
+    },
+    {
+      "epoch": 0.9983805668016195,
+      "grad_norm": 0.9966294564153475,
+      "learning_rate": 1.526066719483811e-05,
+      "loss": 1.6605,
+      "step": 1233
+    },
+    {
+      "epoch": 0.9991902834008097,
+      "grad_norm": 1.1569636814062747,
+      "learning_rate": 1.5253355412719498e-05,
+      "loss": 1.6844,
+      "step": 1234
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 1.0838838624298,
+      "learning_rate": 1.5246039749439159e-05,
+      "loss": 1.6023,
+      "step": 1235
+    },
+    {
+      "epoch": 1.0008097165991903,
+      "grad_norm": 1.3033229459063758,
+      "learning_rate": 1.5238720210401881e-05,
+      "loss": 1.4524,
+      "step": 1236
+    },
+    {
+      "epoch": 1.0016194331983805,
+      "grad_norm": 1.1817666802596205,
+      "learning_rate": 1.5231396801015321e-05,
+      "loss": 1.4273,
+      "step": 1237
+    },
+    {
+      "epoch": 1.0024291497975708,
+      "grad_norm": 1.1157969184839365,
+      "learning_rate": 1.5224069526689981e-05,
+      "loss": 1.4376,
+      "step": 1238
+    },
+    {
+      "epoch": 1.003238866396761,
+      "grad_norm": 1.1746908857400815,
+      "learning_rate": 1.5216738392839241e-05,
+      "loss": 1.3419,
+      "step": 1239
+    },
+    {
+      "epoch": 1.0040485829959513,
+      "grad_norm": 1.5854462006834873,
+      "learning_rate": 1.5209403404879305e-05,
+      "loss": 1.3791,
+      "step": 1240
+    },
+    {
+      "epoch": 1.0048582995951416,
+      "grad_norm": 1.4607487661303684,
+      "learning_rate": 1.5202064568229242e-05,
+      "loss": 1.3514,
+      "step": 1241
+    },
+    {
+      "epoch": 1.0056680161943319,
+      "grad_norm": 1.4180363035835093,
+      "learning_rate": 1.5194721888310966e-05,
+      "loss": 1.4033,
+      "step": 1242
+    },
+    {
+      "epoch": 1.0064777327935224,
+      "grad_norm": 1.3518288634394915,
+      "learning_rate": 1.5187375370549218e-05,
+      "loss": 1.3413,
+      "step": 1243
+    },
+    {
+      "epoch": 1.0072874493927126,
+      "grad_norm": 1.3776243915536928,
+      "learning_rate": 1.5180025020371585e-05,
+      "loss": 1.4354,
+      "step": 1244
+    },
+    {
+      "epoch": 1.008097165991903,
+      "grad_norm": 1.263117229706304,
+      "learning_rate": 1.5172670843208477e-05,
+      "loss": 1.3634,
+      "step": 1245
+    },
+    {
+      "epoch": 1.0089068825910932,
+      "grad_norm": 1.2727019839645528,
+      "learning_rate": 1.5165312844493146e-05,
+      "loss": 1.3325,
+      "step": 1246
+    },
+    {
+      "epoch": 1.0097165991902834,
+      "grad_norm": 1.2423022959844323,
+      "learning_rate": 1.5157951029661644e-05,
+      "loss": 1.3249,
+      "step": 1247
+    },
+    {
+      "epoch": 1.0105263157894737,
+      "grad_norm": 1.3569865807346784,
+      "learning_rate": 1.5150585404152864e-05,
+      "loss": 1.388,
+      "step": 1248
+    },
+    {
+      "epoch": 1.011336032388664,
+      "grad_norm": 1.3839955804790385,
+      "learning_rate": 1.5143215973408505e-05,
+      "loss": 1.3716,
+      "step": 1249
+    },
+    {
+      "epoch": 1.0121457489878543,
+      "grad_norm": 1.281605620473294,
+      "learning_rate": 1.5135842742873077e-05,
+      "loss": 1.3419,
+      "step": 1250
+    },
+    {
+      "epoch": 1.0129554655870445,
+      "grad_norm": 1.240875523498585,
+      "learning_rate": 1.5128465717993898e-05,
+      "loss": 1.4467,
+      "step": 1251
+    },
+    {
+      "epoch": 1.0137651821862348,
+      "grad_norm": 1.2013839184847355,
+      "learning_rate": 1.5121084904221093e-05,
+      "loss": 1.3489,
+      "step": 1252
+    },
+    {
+      "epoch": 1.014574898785425,
+      "grad_norm": 1.224004173475795,
+      "learning_rate": 1.5113700307007575e-05,
+      "loss": 1.3784,
+      "step": 1253
+    },
+    {
+      "epoch": 1.0153846153846153,
+      "grad_norm": 1.2009133574758046,
+      "learning_rate": 1.510631193180907e-05,
+      "loss": 1.3149,
+      "step": 1254
+    },
+    {
+      "epoch": 1.0161943319838056,
+      "grad_norm": 1.2303064249057094,
+      "learning_rate": 1.5098919784084083e-05,
+      "loss": 1.3512,
+      "step": 1255
+    },
+    {
+      "epoch": 1.0170040485829959,
+      "grad_norm": 1.2885247341623098,
+      "learning_rate": 1.50915238692939e-05,
+      "loss": 1.3995,
+      "step": 1256
+    },
+    {
+      "epoch": 1.0178137651821861,
+      "grad_norm": 1.2509468790226062,
+      "learning_rate": 1.5084124192902612e-05,
+      "loss": 1.3568,
+      "step": 1257
+    },
+    {
+      "epoch": 1.0186234817813766,
+      "grad_norm": 1.2329823773251085,
+      "learning_rate": 1.5076720760377064e-05,
+      "loss": 1.3742,
+      "step": 1258
+    },
+    {
+      "epoch": 1.019433198380567,
+      "grad_norm": 1.2514982887751502,
+      "learning_rate": 1.5069313577186892e-05,
+      "loss": 1.4078,
+      "step": 1259
+    },
+    {
+      "epoch": 1.0202429149797572,
+      "grad_norm": 1.1878670599291559,
+      "learning_rate": 1.5061902648804503e-05,
+      "loss": 1.3712,
+      "step": 1260
+    },
+    {
+      "epoch": 1.0210526315789474,
+      "grad_norm": 1.1702105583771456,
+      "learning_rate": 1.5054487980705059e-05,
+      "loss": 1.3456,
+      "step": 1261
+    },
+    {
+      "epoch": 1.0218623481781377,
+      "grad_norm": 1.2390980605393822,
+      "learning_rate": 1.5047069578366497e-05,
+      "loss": 1.3626,
+      "step": 1262
+    },
+    {
+      "epoch": 1.022672064777328,
+      "grad_norm": 1.1903294920255179,
+      "learning_rate": 1.5039647447269508e-05,
+      "loss": 1.3562,
+      "step": 1263
+    },
+    {
+      "epoch": 1.0234817813765182,
+      "grad_norm": 1.210946578195798,
+      "learning_rate": 1.5032221592897536e-05,
+      "loss": 1.3543,
+      "step": 1264
+    },
+    {
+      "epoch": 1.0242914979757085,
+      "grad_norm": 1.1617580200681779,
+      "learning_rate": 1.502479202073678e-05,
+      "loss": 1.3388,
+      "step": 1265
+    },
+    {
+      "epoch": 1.0251012145748988,
+      "grad_norm": 1.2440798916020672,
+      "learning_rate": 1.5017358736276183e-05,
+      "loss": 1.378,
+      "step": 1266
+    },
+    {
+      "epoch": 1.025910931174089,
+      "grad_norm": 1.2628930798261047,
+      "learning_rate": 1.500992174500743e-05,
+      "loss": 1.3923,
+      "step": 1267
+    },
+    {
+      "epoch": 1.0267206477732793,
+      "grad_norm": 1.2359583176315028,
+      "learning_rate": 1.5002481052424945e-05,
+      "loss": 1.3872,
+      "step": 1268
+    },
+    {
+      "epoch": 1.0275303643724696,
+      "grad_norm": 1.1963174169460538,
+      "learning_rate": 1.499503666402589e-05,
+      "loss": 1.3951,
+      "step": 1269
+    },
+    {
+      "epoch": 1.0283400809716599,
+      "grad_norm": 1.124590149714284,
+      "learning_rate": 1.4987588585310154e-05,
+      "loss": 1.361,
+      "step": 1270
+    },
+    {
+      "epoch": 1.0291497975708501,
+      "grad_norm": 1.1617067643230052,
+      "learning_rate": 1.4980136821780348e-05,
+      "loss": 1.3614,
+      "step": 1271
+    },
+    {
+      "epoch": 1.0299595141700404,
+      "grad_norm": 1.6170111738264419,
+      "learning_rate": 1.497268137894182e-05,
+      "loss": 1.4219,
+      "step": 1272
+    },
+    {
+      "epoch": 1.0307692307692307,
+      "grad_norm": 1.2275764161254092,
+      "learning_rate": 1.4965222262302621e-05,
+      "loss": 1.3835,
+      "step": 1273
+    },
+    {
+      "epoch": 1.0315789473684212,
+      "grad_norm": 1.1996462946756752,
+      "learning_rate": 1.4957759477373519e-05,
+      "loss": 1.2693,
+      "step": 1274
+    },
+    {
+      "epoch": 1.0323886639676114,
+      "grad_norm": 1.2125778356953747,
+      "learning_rate": 1.4950293029668004e-05,
+      "loss": 1.3383,
+      "step": 1275
+    },
+    {
+      "epoch": 1.0331983805668017,
+      "grad_norm": 1.2012198190789183,
+      "learning_rate": 1.4942822924702252e-05,
+      "loss": 1.374,
+      "step": 1276
+    },
+    {
+      "epoch": 1.034008097165992,
+      "grad_norm": 1.2130852863124042,
+      "learning_rate": 1.4935349167995161e-05,
+      "loss": 1.4402,
+      "step": 1277
+    },
+    {
+      "epoch": 1.0348178137651822,
+      "grad_norm": 1.1740546729540866,
+      "learning_rate": 1.4927871765068314e-05,
+      "loss": 1.3831,
+      "step": 1278
+    },
+    {
+      "epoch": 1.0356275303643725,
+      "grad_norm": 1.1917159960823642,
+      "learning_rate": 1.4920390721445993e-05,
+      "loss": 1.3729,
+      "step": 1279
+    },
+    {
+      "epoch": 1.0364372469635628,
+      "grad_norm": 1.1733824111242548,
+      "learning_rate": 1.4912906042655164e-05,
+      "loss": 1.3949,
+      "step": 1280
+    },
+    {
+      "epoch": 1.037246963562753,
+      "grad_norm": 1.1752871025521956,
+      "learning_rate": 1.4905417734225488e-05,
+      "loss": 1.3553,
+      "step": 1281
+    },
+    {
+      "epoch": 1.0380566801619433,
+      "grad_norm": 1.1715814660303838,
+      "learning_rate": 1.4897925801689304e-05,
+      "loss": 1.3838,
+      "step": 1282
+    },
+    {
+      "epoch": 1.0388663967611336,
+      "grad_norm": 1.2415703860859453,
+      "learning_rate": 1.4890430250581622e-05,
+      "loss": 1.3841,
+      "step": 1283
+    },
+    {
+      "epoch": 1.0396761133603238,
+      "grad_norm": 1.1825484273450233,
+      "learning_rate": 1.4882931086440133e-05,
+      "loss": 1.3573,
+      "step": 1284
+    },
+    {
+      "epoch": 1.040485829959514,
+      "grad_norm": 1.1504373958278675,
+      "learning_rate": 1.4875428314805195e-05,
+      "loss": 1.3256,
+      "step": 1285
+    },
+    {
+      "epoch": 1.0412955465587044,
+      "grad_norm": 1.1436853666808242,
+      "learning_rate": 1.4867921941219834e-05,
+      "loss": 1.3932,
+      "step": 1286
+    },
+    {
+      "epoch": 1.0421052631578946,
+      "grad_norm": 1.162386050137052,
+      "learning_rate": 1.4860411971229728e-05,
+      "loss": 1.3845,
+      "step": 1287
+    },
+    {
+      "epoch": 1.042914979757085,
+      "grad_norm": 1.2092585781453715,
+      "learning_rate": 1.485289841038322e-05,
+      "loss": 1.3369,
+      "step": 1288
+    },
+    {
+      "epoch": 1.0437246963562754,
+      "grad_norm": 1.145037292414635,
+      "learning_rate": 1.484538126423131e-05,
+      "loss": 1.4054,
+      "step": 1289
+    },
+    {
+      "epoch": 1.0445344129554657,
+      "grad_norm": 1.1686385666583774,
+      "learning_rate": 1.483786053832763e-05,
+      "loss": 1.3353,
+      "step": 1290
+    },
+    {
+      "epoch": 1.045344129554656,
+      "grad_norm": 1.2275739238178558,
+      "learning_rate": 1.483033623822848e-05,
+      "loss": 1.4264,
+      "step": 1291
+    },
+    {
+      "epoch": 1.0461538461538462,
+      "grad_norm": 1.1623896668490261,
+      "learning_rate": 1.4822808369492778e-05,
+      "loss": 1.4345,
+      "step": 1292
+    },
+    {
+      "epoch": 1.0469635627530365,
+      "grad_norm": 1.2296387845097625,
+      "learning_rate": 1.4815276937682094e-05,
+      "loss": 1.3207,
+      "step": 1293
+    },
+    {
+      "epoch": 1.0477732793522267,
+      "grad_norm": 1.176201622341318,
+      "learning_rate": 1.4807741948360625e-05,
+      "loss": 1.3585,
+      "step": 1294
+    },
+    {
+      "epoch": 1.048582995951417,
+      "grad_norm": 1.1610794617830476,
+      "learning_rate": 1.4800203407095194e-05,
+      "loss": 1.3479,
+      "step": 1295
+    },
+    {
+      "epoch": 1.0493927125506073,
+      "grad_norm": 1.1776599630056936,
+      "learning_rate": 1.4792661319455252e-05,
+      "loss": 1.4027,
+      "step": 1296
+    },
+    {
+      "epoch": 1.0502024291497976,
+      "grad_norm": 1.2338228585850806,
+      "learning_rate": 1.4785115691012866e-05,
+      "loss": 1.3332,
+      "step": 1297
+    },
+    {
+      "epoch": 1.0510121457489878,
+      "grad_norm": 1.1880604320942654,
+      "learning_rate": 1.4777566527342729e-05,
+      "loss": 1.3304,
+      "step": 1298
+    },
+    {
+      "epoch": 1.051821862348178,
+      "grad_norm": 1.1817231459155564,
+      "learning_rate": 1.4770013834022128e-05,
+      "loss": 1.3714,
+      "step": 1299
+    },
+    {
+      "epoch": 1.0526315789473684,
+      "grad_norm": 1.185176582407899,
+      "learning_rate": 1.4762457616630972e-05,
+      "loss": 1.2879,
+      "step": 1300
+    },
+    {
+      "epoch": 1.0534412955465586,
+      "grad_norm": 1.1828727222185689,
+      "learning_rate": 1.4754897880751776e-05,
+      "loss": 1.2923,
+      "step": 1301
+    },
+    {
+      "epoch": 1.054251012145749,
+      "grad_norm": 1.1744325557146087,
+      "learning_rate": 1.474733463196964e-05,
+      "loss": 1.3548,
+      "step": 1302
+    },
+    {
+      "epoch": 1.0550607287449392,
+      "grad_norm": 1.160850787157178,
+      "learning_rate": 1.4739767875872271e-05,
+      "loss": 1.4059,
+      "step": 1303
+    },
+    {
+      "epoch": 1.0558704453441297,
+      "grad_norm": 1.176977157825984,
+      "learning_rate": 1.473219761804996e-05,
+      "loss": 1.3351,
+      "step": 1304
+    },
+    {
+      "epoch": 1.05668016194332,
+      "grad_norm": 1.2503348243091958,
+      "learning_rate": 1.4724623864095595e-05,
+      "loss": 1.4432,
+      "step": 1305
+    },
+    {
+      "epoch": 1.0574898785425102,
+      "grad_norm": 1.1872676938264057,
+      "learning_rate": 1.4717046619604636e-05,
+      "loss": 1.4101,
+      "step": 1306
+    },
+    {
+      "epoch": 1.0582995951417005,
+      "grad_norm": 1.168594441423703,
+      "learning_rate": 1.4709465890175125e-05,
+      "loss": 1.3947,
+      "step": 1307
+    },
+    {
+      "epoch": 1.0591093117408907,
+      "grad_norm": 1.1705898387765592,
+      "learning_rate": 1.4701881681407684e-05,
+      "loss": 1.3358,
+      "step": 1308
+    },
+    {
+      "epoch": 1.059919028340081,
+      "grad_norm": 1.1603679171263597,
+      "learning_rate": 1.46942939989055e-05,
+      "loss": 1.3784,
+      "step": 1309
+    },
+    {
+      "epoch": 1.0607287449392713,
+      "grad_norm": 1.164400476925873,
+      "learning_rate": 1.4686702848274328e-05,
+      "loss": 1.3856,
+      "step": 1310
+    },
+    {
+      "epoch": 1.0615384615384615,
+      "grad_norm": 1.1554651486479175,
+      "learning_rate": 1.4679108235122482e-05,
+      "loss": 1.3397,
+      "step": 1311
+    },
+    {
+      "epoch": 1.0623481781376518,
+      "grad_norm": 1.2155869311427083,
+      "learning_rate": 1.467151016506084e-05,
+      "loss": 1.3729,
+      "step": 1312
+    },
+    {
+      "epoch": 1.063157894736842,
+      "grad_norm": 1.2133615435457628,
+      "learning_rate": 1.4663908643702836e-05,
+      "loss": 1.3122,
+      "step": 1313
+    },
+    {
+      "epoch": 1.0639676113360323,
+      "grad_norm": 1.1634564916763344,
+      "learning_rate": 1.465630367666444e-05,
+      "loss": 1.3354,
+      "step": 1314
+    },
+    {
+      "epoch": 1.0647773279352226,
+      "grad_norm": 1.2001955918988352,
+      "learning_rate": 1.4648695269564182e-05,
+      "loss": 1.4082,
+      "step": 1315
+    },
+    {
+      "epoch": 1.0655870445344129,
+      "grad_norm": 1.1379663684306365,
+      "learning_rate": 1.4641083428023124e-05,
+      "loss": 1.3493,
+      "step": 1316
+    },
+    {
+      "epoch": 1.0663967611336032,
+      "grad_norm": 1.1718455198343034,
+      "learning_rate": 1.4633468157664879e-05,
+      "loss": 1.3505,
+      "step": 1317
+    },
+    {
+      "epoch": 1.0672064777327934,
+      "grad_norm": 1.1646759648158063,
+      "learning_rate": 1.4625849464115571e-05,
+      "loss": 1.3558,
+      "step": 1318
+    },
+    {
+      "epoch": 1.068016194331984,
+      "grad_norm": 1.162310000496439,
+      "learning_rate": 1.4618227353003878e-05,
+      "loss": 1.3958,
+      "step": 1319
+    },
+    {
+      "epoch": 1.0688259109311742,
+      "grad_norm": 1.1488822317358849,
+      "learning_rate": 1.461060182996098e-05,
+      "loss": 1.4,
+      "step": 1320
+    },
+    {
+      "epoch": 1.0696356275303645,
+      "grad_norm": 1.2033656269010369,
+      "learning_rate": 1.4602972900620596e-05,
+      "loss": 1.3656,
+      "step": 1321
+    },
+    {
+      "epoch": 1.0704453441295547,
+      "grad_norm": 1.2654857663774528,
+      "learning_rate": 1.459534057061895e-05,
+      "loss": 1.314,
+      "step": 1322
+    },
+    {
+      "epoch": 1.071255060728745,
+      "grad_norm": 1.1409457494007944,
+      "learning_rate": 1.4587704845594784e-05,
+      "loss": 1.3914,
+      "step": 1323
+    },
+    {
+      "epoch": 1.0720647773279353,
+      "grad_norm": 1.1625130696761021,
+      "learning_rate": 1.4580065731189344e-05,
+      "loss": 1.3809,
+      "step": 1324
+    },
+    {
+      "epoch": 1.0728744939271255,
+      "grad_norm": 1.1754811815518604,
+      "learning_rate": 1.4572423233046386e-05,
+      "loss": 1.357,
+      "step": 1325
+    },
+    {
+      "epoch": 1.0736842105263158,
+      "grad_norm": 1.274907802092002,
+      "learning_rate": 1.456477735681216e-05,
+      "loss": 1.3543,
+      "step": 1326
+    },
+    {
+      "epoch": 1.074493927125506,
+      "grad_norm": 1.194810175856374,
+      "learning_rate": 1.455712810813542e-05,
+      "loss": 1.3274,
+      "step": 1327
+    },
+    {
+      "epoch": 1.0753036437246963,
+      "grad_norm": 1.1841749394103487,
+      "learning_rate": 1.4549475492667395e-05,
+      "loss": 1.4214,
+      "step": 1328
+    },
+    {
+      "epoch": 1.0761133603238866,
+      "grad_norm": 1.1616739871602735,
+      "learning_rate": 1.4541819516061824e-05,
+      "loss": 1.3143,
+      "step": 1329
+    },
+    {
+      "epoch": 1.0769230769230769,
+      "grad_norm": 1.235456801171255,
+      "learning_rate": 1.4534160183974908e-05,
+      "loss": 1.3862,
+      "step": 1330
+    },
+    {
+      "epoch": 1.0777327935222671,
+      "grad_norm": 1.185526004922521,
+      "learning_rate": 1.4526497502065343e-05,
+      "loss": 1.3827,
+      "step": 1331
+    },
+    {
+      "epoch": 1.0785425101214574,
+      "grad_norm": 1.181146915990783,
+      "learning_rate": 1.4518831475994287e-05,
+      "loss": 1.3566,
+      "step": 1332
+    },
+    {
+      "epoch": 1.0793522267206477,
+      "grad_norm": 1.2073400109649837,
+      "learning_rate": 1.4511162111425377e-05,
+      "loss": 1.3772,
+      "step": 1333
+    },
+    {
+      "epoch": 1.0801619433198382,
+      "grad_norm": 1.2061215113680948,
+      "learning_rate": 1.450348941402472e-05,
+      "loss": 1.4011,
+      "step": 1334
+    },
+    {
+      "epoch": 1.0809716599190284,
+      "grad_norm": 1.2058666314849154,
+      "learning_rate": 1.4495813389460875e-05,
+      "loss": 1.3617,
+      "step": 1335
+    },
+    {
+      "epoch": 1.0817813765182187,
+      "grad_norm": 1.1903692062674975,
+      "learning_rate": 1.448813404340486e-05,
+      "loss": 1.2916,
+      "step": 1336
+    },
+    {
+      "epoch": 1.082591093117409,
+      "grad_norm": 1.2340367866655837,
+      "learning_rate": 1.4480451381530159e-05,
+      "loss": 1.2935,
+      "step": 1337
+    },
+    {
+      "epoch": 1.0834008097165992,
+      "grad_norm": 1.2659821523943215,
+      "learning_rate": 1.447276540951269e-05,
+      "loss": 1.362,
+      "step": 1338
+    },
+    {
+      "epoch": 1.0842105263157895,
+      "grad_norm": 1.2228755854040374,
+      "learning_rate": 1.4465076133030828e-05,
+      "loss": 1.3781,
+      "step": 1339
+    },
+    {
+      "epoch": 1.0850202429149798,
+      "grad_norm": 1.1570938076905548,
+      "learning_rate": 1.4457383557765385e-05,
+      "loss": 1.3808,
+      "step": 1340
+    },
+    {
+      "epoch": 1.08582995951417,
+      "grad_norm": 1.1637617032486594,
+      "learning_rate": 1.4449687689399607e-05,
+      "loss": 1.3515,
+      "step": 1341
+    },
+    {
+      "epoch": 1.0866396761133603,
+      "grad_norm": 1.1678076483156412,
+      "learning_rate": 1.4441988533619182e-05,
+      "loss": 1.3391,
+      "step": 1342
+    },
+    {
+      "epoch": 1.0874493927125506,
+      "grad_norm": 1.2160441877476182,
+      "learning_rate": 1.4434286096112215e-05,
+      "loss": 1.3475,
+      "step": 1343
+    },
+    {
+      "epoch": 1.0882591093117409,
+      "grad_norm": 1.235249585783021,
+      "learning_rate": 1.4426580382569241e-05,
+      "loss": 1.2836,
+      "step": 1344
+    },
+    {
+      "epoch": 1.0890688259109311,
+      "grad_norm": 1.2382294650559509,
+      "learning_rate": 1.4418871398683227e-05,
+      "loss": 1.392,
+      "step": 1345
+    },
+    {
+      "epoch": 1.0898785425101214,
+      "grad_norm": 1.173232332418364,
+      "learning_rate": 1.4411159150149532e-05,
+      "loss": 1.3071,
+      "step": 1346
+    },
+    {
+      "epoch": 1.0906882591093117,
+      "grad_norm": 1.1510674680418143,
+      "learning_rate": 1.4403443642665946e-05,
+      "loss": 1.3228,
+      "step": 1347
+    },
+    {
+      "epoch": 1.091497975708502,
+      "grad_norm": 1.1961923726867185,
+      "learning_rate": 1.439572488193266e-05,
+      "loss": 1.3194,
+      "step": 1348
+    },
+    {
+      "epoch": 1.0923076923076924,
+      "grad_norm": 1.1902807444112073,
+      "learning_rate": 1.438800287365227e-05,
+      "loss": 1.3624,
+      "step": 1349
+    },
+    {
+      "epoch": 1.0931174089068827,
+      "grad_norm": 1.1834064174607142,
+      "learning_rate": 1.4380277623529766e-05,
+      "loss": 1.3173,
+      "step": 1350
+    },
+    {
+      "epoch": 1.093927125506073,
+      "grad_norm": 1.2136583995550188,
+      "learning_rate": 1.437254913727254e-05,
+      "loss": 1.3443,
+      "step": 1351
+    },
+    {
+      "epoch": 1.0947368421052632,
+      "grad_norm": 1.154981299062566,
+      "learning_rate": 1.4364817420590373e-05,
+      "loss": 1.3872,
+      "step": 1352
+    },
+    {
+      "epoch": 1.0955465587044535,
+      "grad_norm": 1.2102553779995304,
+      "learning_rate": 1.4357082479195435e-05,
+      "loss": 1.2916,
+      "step": 1353
+    },
+    {
+      "epoch": 1.0963562753036438,
+      "grad_norm": 1.188609622826024,
+      "learning_rate": 1.434934431880227e-05,
+      "loss": 1.3302,
+      "step": 1354
+    },
+    {
+      "epoch": 1.097165991902834,
+      "grad_norm": 1.200714765859928,
+      "learning_rate": 1.4341602945127806e-05,
+      "loss": 1.2963,
+      "step": 1355
+    },
+    {
+      "epoch": 1.0979757085020243,
+      "grad_norm": 1.1787138315863135,
+      "learning_rate": 1.4333858363891346e-05,
+      "loss": 1.3517,
+      "step": 1356
+    },
+    {
+      "epoch": 1.0987854251012146,
+      "grad_norm": 1.2170954780680936,
+      "learning_rate": 1.4326110580814563e-05,
+      "loss": 1.3765,
+      "step": 1357
+    },
+    {
+      "epoch": 1.0995951417004048,
+      "grad_norm": 1.250816449134447,
+      "learning_rate": 1.431835960162149e-05,
+      "loss": 1.3175,
+      "step": 1358
+    },
+    {
+      "epoch": 1.1004048582995951,
+      "grad_norm": 1.177624560111963,
+      "learning_rate": 1.4310605432038527e-05,
+      "loss": 1.3935,
+      "step": 1359
+    },
+    {
+      "epoch": 1.1012145748987854,
+      "grad_norm": 1.2197836972826388,
+      "learning_rate": 1.4302848077794427e-05,
+      "loss": 1.4803,
+      "step": 1360
+    },
+    {
+      "epoch": 1.1020242914979756,
+      "grad_norm": 1.2127800632997665,
+      "learning_rate": 1.42950875446203e-05,
+      "loss": 1.3772,
+      "step": 1361
+    },
+    {
+      "epoch": 1.102834008097166,
+      "grad_norm": 1.23399617147964,
+      "learning_rate": 1.4287323838249603e-05,
+      "loss": 1.3533,
+      "step": 1362
+    },
+    {
+      "epoch": 1.1036437246963562,
+      "grad_norm": 1.2240074250465314,
+      "learning_rate": 1.4279556964418135e-05,
+      "loss": 1.3569,
+      "step": 1363
+    },
+    {
+      "epoch": 1.1044534412955465,
+      "grad_norm": 1.2564824376072379,
+      "learning_rate": 1.4271786928864037e-05,
+      "loss": 1.4501,
+      "step": 1364
+    },
+    {
+      "epoch": 1.1052631578947367,
+      "grad_norm": 1.1991759525151162,
+      "learning_rate": 1.426401373732779e-05,
+      "loss": 1.3907,
+      "step": 1365
+    },
+    {
+      "epoch": 1.1060728744939272,
+      "grad_norm": 1.2378861157976855,
+      "learning_rate": 1.4256237395552195e-05,
+      "loss": 1.3469,
+      "step": 1366
+    },
+    {
+      "epoch": 1.1068825910931175,
+      "grad_norm": 1.2292213801115794,
+      "learning_rate": 1.4248457909282391e-05,
+      "loss": 1.3883,
+      "step": 1367
+    },
+    {
+      "epoch": 1.1076923076923078,
+      "grad_norm": 1.1914894912948184,
+      "learning_rate": 1.4240675284265838e-05,
+      "loss": 1.3029,
+      "step": 1368
+    },
+    {
+      "epoch": 1.108502024291498,
+      "grad_norm": 1.180343522184691,
+      "learning_rate": 1.4232889526252316e-05,
+      "loss": 1.3656,
+      "step": 1369
+    },
+    {
+      "epoch": 1.1093117408906883,
+      "grad_norm": 1.1905529250584874,
+      "learning_rate": 1.422510064099391e-05,
+      "loss": 1.3206,
+      "step": 1370
+    },
+    {
+      "epoch": 1.1101214574898786,
+      "grad_norm": 1.211461020000152,
+      "learning_rate": 1.421730863424503e-05,
+      "loss": 1.2979,
+      "step": 1371
+    },
+    {
+      "epoch": 1.1109311740890688,
+      "grad_norm": 1.2001565370308576,
+      "learning_rate": 1.4209513511762381e-05,
+      "loss": 1.3776,
+      "step": 1372
+    },
+    {
+      "epoch": 1.111740890688259,
+      "grad_norm": 1.2955199788028005,
+      "learning_rate": 1.420171527930498e-05,
+      "loss": 1.3443,
+      "step": 1373
+    },
+    {
+      "epoch": 1.1125506072874494,
+      "grad_norm": 1.1599118475957895,
+      "learning_rate": 1.4193913942634122e-05,
+      "loss": 1.3644,
+      "step": 1374
+    },
+    {
+      "epoch": 1.1133603238866396,
+      "grad_norm": 1.1585810167212327,
+      "learning_rate": 1.4186109507513425e-05,
+      "loss": 1.3979,
+      "step": 1375
+    },
+    {
+      "epoch": 1.11417004048583,
+      "grad_norm": 1.2720040494303266,
+      "learning_rate": 1.417830197970877e-05,
+      "loss": 1.3597,
+      "step": 1376
+    },
+    {
+      "epoch": 1.1149797570850202,
+      "grad_norm": 1.2098574998597365,
+      "learning_rate": 1.4170491364988336e-05,
+      "loss": 1.4478,
+      "step": 1377
+    },
+    {
+      "epoch": 1.1157894736842104,
+      "grad_norm": 1.2313276250589487,
+      "learning_rate": 1.416267766912258e-05,
+      "loss": 1.3725,
+      "step": 1378
+    },
+    {
+      "epoch": 1.1165991902834007,
+      "grad_norm": 1.1949141861834391,
+      "learning_rate": 1.4154860897884234e-05,
+      "loss": 1.394,
+      "step": 1379
+    },
+    {
+      "epoch": 1.117408906882591,
+      "grad_norm": 1.1798433594200126,
+      "learning_rate": 1.4147041057048303e-05,
+      "loss": 1.3759,
+      "step": 1380
+    },
+    {
+      "epoch": 1.1182186234817815,
+      "grad_norm": 1.2126045567567152,
+      "learning_rate": 1.4139218152392058e-05,
+      "loss": 1.4342,
+      "step": 1381
+    },
+    {
+      "epoch": 1.1190283400809717,
+      "grad_norm": 1.1831352962334098,
+      "learning_rate": 1.4131392189695037e-05,
+      "loss": 1.3377,
+      "step": 1382
+    },
+    {
+      "epoch": 1.119838056680162,
+      "grad_norm": 1.1965883843750686,
+      "learning_rate": 1.4123563174739036e-05,
+      "loss": 1.3369,
+      "step": 1383
+    },
+    {
+      "epoch": 1.1206477732793523,
+      "grad_norm": 1.1612564728192138,
+      "learning_rate": 1.4115731113308106e-05,
+      "loss": 1.3275,
+      "step": 1384
+    },
+    {
+      "epoch": 1.1214574898785425,
+      "grad_norm": 1.2213439759905513,
+      "learning_rate": 1.4107896011188546e-05,
+      "loss": 1.3599,
+      "step": 1385
+    },
+    {
+      "epoch": 1.1222672064777328,
+      "grad_norm": 1.2321562126222285,
+      "learning_rate": 1.4100057874168906e-05,
+      "loss": 1.4032,
+      "step": 1386
+    },
+    {
+      "epoch": 1.123076923076923,
+      "grad_norm": 1.2058709080617371,
+      "learning_rate": 1.4092216708039974e-05,
+      "loss": 1.3196,
+      "step": 1387
+    },
+    {
+      "epoch": 1.1238866396761134,
+      "grad_norm": 1.1731302542020694,
+      "learning_rate": 1.4084372518594777e-05,
+      "loss": 1.372,
+      "step": 1388
+    },
+    {
+      "epoch": 1.1246963562753036,
+      "grad_norm": 1.2473568748894555,
+      "learning_rate": 1.4076525311628581e-05,
+      "loss": 1.3703,
+      "step": 1389
+    },
+    {
+      "epoch": 1.125506072874494,
+      "grad_norm": 1.1445398846114365,
+      "learning_rate": 1.4068675092938872e-05,
+      "loss": 1.3289,
+      "step": 1390
+    },
+    {
+      "epoch": 1.1263157894736842,
+      "grad_norm": 1.178653116534255,
+      "learning_rate": 1.406082186832537e-05,
+      "loss": 1.3859,
+      "step": 1391
+    },
+    {
+      "epoch": 1.1271255060728744,
+      "grad_norm": 1.2540075296621336,
+      "learning_rate": 1.4052965643590006e-05,
+      "loss": 1.3663,
+      "step": 1392
+    },
+    {
+      "epoch": 1.1279352226720647,
+      "grad_norm": 1.278208376958459,
+      "learning_rate": 1.4045106424536938e-05,
+      "loss": 1.4342,
+      "step": 1393
+    },
+    {
+      "epoch": 1.128744939271255,
+      "grad_norm": 1.2422771469402274,
+      "learning_rate": 1.403724421697253e-05,
+      "loss": 1.3634,
+      "step": 1394
+    },
+    {
+      "epoch": 1.1295546558704452,
+      "grad_norm": 1.220480716235926,
+      "learning_rate": 1.4029379026705352e-05,
+      "loss": 1.3913,
+      "step": 1395
+    },
+    {
+      "epoch": 1.1303643724696357,
+      "grad_norm": 1.2008169986282988,
+      "learning_rate": 1.4021510859546184e-05,
+      "loss": 1.3316,
+      "step": 1396
+    },
+    {
+      "epoch": 1.131174089068826,
+      "grad_norm": 1.206940001904269,
+      "learning_rate": 1.4013639721308004e-05,
+      "loss": 1.3366,
+      "step": 1397
+    },
+    {
+      "epoch": 1.1319838056680163,
+      "grad_norm": 1.1864838007376735,
+      "learning_rate": 1.4005765617805977e-05,
+      "loss": 1.3345,
+      "step": 1398
+    },
+    {
+      "epoch": 1.1327935222672065,
+      "grad_norm": 1.1806069165225561,
+      "learning_rate": 1.3997888554857468e-05,
+      "loss": 1.3515,
+      "step": 1399
+    },
+    {
+      "epoch": 1.1336032388663968,
+      "grad_norm": 1.244689296436616,
+      "learning_rate": 1.3990008538282027e-05,
+      "loss": 1.4887,
+      "step": 1400
+    },
+    {
+      "epoch": 1.134412955465587,
+      "grad_norm": 1.2092159259655217,
+      "learning_rate": 1.3982125573901384e-05,
+      "loss": 1.3932,
+      "step": 1401
+    },
+    {
+      "epoch": 1.1352226720647773,
+      "grad_norm": 1.201131078854986,
+      "learning_rate": 1.3974239667539445e-05,
+      "loss": 1.3621,
+      "step": 1402
+    },
+    {
+      "epoch": 1.1360323886639676,
+      "grad_norm": 1.2951035372189361,
+      "learning_rate": 1.396635082502229e-05,
+      "loss": 1.3733,
+      "step": 1403
+    },
+    {
+      "epoch": 1.1368421052631579,
+      "grad_norm": 1.2124240532342052,
+      "learning_rate": 1.3958459052178175e-05,
+      "loss": 1.3841,
+      "step": 1404
+    },
+    {
+      "epoch": 1.1376518218623481,
+      "grad_norm": 1.2078175074517925,
+      "learning_rate": 1.3950564354837512e-05,
+      "loss": 1.3062,
+      "step": 1405
+    },
+    {
+      "epoch": 1.1384615384615384,
+      "grad_norm": 1.2235438631380933,
+      "learning_rate": 1.3942666738832879e-05,
+      "loss": 1.3788,
+      "step": 1406
+    },
+    {
+      "epoch": 1.1392712550607287,
+      "grad_norm": 1.2535687484192757,
+      "learning_rate": 1.3934766209999012e-05,
+      "loss": 1.32,
+      "step": 1407
+    },
+    {
+      "epoch": 1.140080971659919,
+      "grad_norm": 1.1792494735598216,
+      "learning_rate": 1.3926862774172789e-05,
+      "loss": 1.3501,
+      "step": 1408
+    },
+    {
+      "epoch": 1.1408906882591092,
+      "grad_norm": 1.2550576055364107,
+      "learning_rate": 1.391895643719325e-05,
+      "loss": 1.3685,
+      "step": 1409
+    },
+    {
+      "epoch": 1.1417004048582995,
+      "grad_norm": 1.187775539751402,
+      "learning_rate": 1.391104720490156e-05,
+      "loss": 1.3756,
+      "step": 1410
+    },
+    {
+      "epoch": 1.14251012145749,
+      "grad_norm": 1.1716789792191453,
+      "learning_rate": 1.3903135083141046e-05,
+      "loss": 1.3657,
+      "step": 1411
+    },
+    {
+      "epoch": 1.1433198380566802,
+      "grad_norm": 1.1878226281741322,
+      "learning_rate": 1.389522007775715e-05,
+      "loss": 1.3513,
+      "step": 1412
+    },
+    {
+      "epoch": 1.1441295546558705,
+      "grad_norm": 1.172121127796374,
+      "learning_rate": 1.3887302194597455e-05,
+      "loss": 1.3747,
+      "step": 1413
+    },
+    {
+      "epoch": 1.1449392712550608,
+      "grad_norm": 1.1977041128005224,
+      "learning_rate": 1.3879381439511664e-05,
+      "loss": 1.3213,
+      "step": 1414
+    },
+    {
+      "epoch": 1.145748987854251,
+      "grad_norm": 1.1951668935892397,
+      "learning_rate": 1.387145781835161e-05,
+      "loss": 1.3693,
+      "step": 1415
+    },
+    {
+      "epoch": 1.1465587044534413,
+      "grad_norm": 1.2425273987483105,
+      "learning_rate": 1.3863531336971231e-05,
+      "loss": 1.3388,
+      "step": 1416
+    },
+    {
+      "epoch": 1.1473684210526316,
+      "grad_norm": 1.2738201413338597,
+      "learning_rate": 1.3855602001226596e-05,
+      "loss": 1.4744,
+      "step": 1417
+    },
+    {
+      "epoch": 1.1481781376518219,
+      "grad_norm": 1.2067856613703325,
+      "learning_rate": 1.384766981697586e-05,
+      "loss": 1.3698,
+      "step": 1418
+    },
+    {
+      "epoch": 1.1489878542510121,
+      "grad_norm": 1.2309122591312711,
+      "learning_rate": 1.3839734790079304e-05,
+      "loss": 1.4008,
+      "step": 1419
+    },
+    {
+      "epoch": 1.1497975708502024,
+      "grad_norm": 1.1924268380366414,
+      "learning_rate": 1.3831796926399295e-05,
+      "loss": 1.4003,
+      "step": 1420
+    },
+    {
+      "epoch": 1.1506072874493927,
+      "grad_norm": 1.252471702177511,
+      "learning_rate": 1.3823856231800301e-05,
+      "loss": 1.3868,
+      "step": 1421
+    },
+    {
+      "epoch": 1.151417004048583,
+      "grad_norm": 1.209467941700512,
+      "learning_rate": 1.3815912712148885e-05,
+      "loss": 1.3534,
+      "step": 1422
+    },
+    {
+      "epoch": 1.1522267206477732,
+      "grad_norm": 1.1971438481581644,
+      "learning_rate": 1.3807966373313689e-05,
+      "loss": 1.355,
+      "step": 1423
+    },
+    {
+      "epoch": 1.1530364372469635,
+      "grad_norm": 1.2066657341311795,
+      "learning_rate": 1.380001722116544e-05,
+      "loss": 1.3531,
+      "step": 1424
+    },
+    {
+      "epoch": 1.1538461538461537,
+      "grad_norm": 1.1713117606005494,
+      "learning_rate": 1.3792065261576953e-05,
+      "loss": 1.3873,
+      "step": 1425
+    },
+    {
+      "epoch": 1.1546558704453442,
+      "grad_norm": 1.2248765090877878,
+      "learning_rate": 1.3784110500423104e-05,
+      "loss": 1.3402,
+      "step": 1426
+    },
+    {
+      "epoch": 1.1554655870445345,
+      "grad_norm": 1.2378827680661408,
+      "learning_rate": 1.3776152943580846e-05,
+      "loss": 1.2761,
+      "step": 1427
+    },
+    {
+      "epoch": 1.1562753036437248,
+      "grad_norm": 1.1711500146006906,
+      "learning_rate": 1.3768192596929195e-05,
+      "loss": 1.377,
+      "step": 1428
+    },
+    {
+      "epoch": 1.157085020242915,
+      "grad_norm": 1.2549196705935586,
+      "learning_rate": 1.376022946634923e-05,
+      "loss": 1.4122,
+      "step": 1429
+    },
+    {
+      "epoch": 1.1578947368421053,
+      "grad_norm": 1.221604782084902,
+      "learning_rate": 1.3752263557724088e-05,
+      "loss": 1.4329,
+      "step": 1430
+    },
+    {
+      "epoch": 1.1587044534412956,
+      "grad_norm": 1.244326599131787,
+      "learning_rate": 1.3744294876938953e-05,
+      "loss": 1.4161,
+      "step": 1431
+    },
+    {
+      "epoch": 1.1595141700404858,
+      "grad_norm": 1.1776858381109803,
+      "learning_rate": 1.3736323429881056e-05,
+      "loss": 1.411,
+      "step": 1432
+    },
+    {
+      "epoch": 1.1603238866396761,
+      "grad_norm": 1.2341697465268895,
+      "learning_rate": 1.3728349222439682e-05,
+      "loss": 1.3794,
+      "step": 1433
+    },
+    {
+      "epoch": 1.1611336032388664,
+      "grad_norm": 1.2025715980532212,
+      "learning_rate": 1.3720372260506152e-05,
+      "loss": 1.4168,
+      "step": 1434
+    },
+    {
+      "epoch": 1.1619433198380567,
+      "grad_norm": 1.2316761185815908,
+      "learning_rate": 1.3712392549973814e-05,
+      "loss": 1.4173,
+      "step": 1435
+    },
+    {
+      "epoch": 1.162753036437247,
+      "grad_norm": 1.1692638075647215,
+      "learning_rate": 1.370441009673805e-05,
+      "loss": 1.3268,
+      "step": 1436
+    },
+    {
+      "epoch": 1.1635627530364372,
+      "grad_norm": 1.2081335987676685,
+      "learning_rate": 1.3696424906696275e-05,
+      "loss": 1.3603,
+      "step": 1437
+    },
+    {
+      "epoch": 1.1643724696356275,
+      "grad_norm": 1.2286436380651753,
+      "learning_rate": 1.3688436985747922e-05,
+      "loss": 1.398,
+      "step": 1438
+    },
+    {
+      "epoch": 1.1651821862348177,
+      "grad_norm": 1.2708424368976516,
+      "learning_rate": 1.3680446339794436e-05,
+      "loss": 1.36,
+      "step": 1439
+    },
+    {
+      "epoch": 1.165991902834008,
+      "grad_norm": 1.2883548798199778,
+      "learning_rate": 1.3672452974739278e-05,
+      "loss": 1.4225,
+      "step": 1440
+    },
+    {
+      "epoch": 1.1668016194331985,
+      "grad_norm": 1.2173917634376092,
+      "learning_rate": 1.366445689648793e-05,
+      "loss": 1.315,
+      "step": 1441
+    },
+    {
+      "epoch": 1.1676113360323888,
+      "grad_norm": 1.242957617527675,
+      "learning_rate": 1.3656458110947864e-05,
+      "loss": 1.3515,
+      "step": 1442
+    },
+    {
+      "epoch": 1.168421052631579,
+      "grad_norm": 1.2488134934449235,
+      "learning_rate": 1.364845662402855e-05,
+      "loss": 1.3045,
+      "step": 1443
+    },
+    {
+      "epoch": 1.1692307692307693,
+      "grad_norm": 1.2204335700413085,
+      "learning_rate": 1.3640452441641466e-05,
+      "loss": 1.396,
+      "step": 1444
+    },
+    {
+      "epoch": 1.1700404858299596,
+      "grad_norm": 1.28192971889865,
+      "learning_rate": 1.3632445569700078e-05,
+      "loss": 1.3832,
+      "step": 1445
+    },
+    {
+      "epoch": 1.1708502024291498,
+      "grad_norm": 1.244268004826121,
+      "learning_rate": 1.362443601411983e-05,
+      "loss": 1.3672,
+      "step": 1446
+    },
+    {
+      "epoch": 1.17165991902834,
+      "grad_norm": 1.155976566029991,
+      "learning_rate": 1.361642378081816e-05,
+      "loss": 1.2883,
+      "step": 1447
+    },
+    {
+      "epoch": 1.1724696356275304,
+      "grad_norm": 1.2611118512015467,
+      "learning_rate": 1.3608408875714478e-05,
+      "loss": 1.3181,
+      "step": 1448
+    },
+    {
+      "epoch": 1.1732793522267206,
+      "grad_norm": 1.1853603357638687,
+      "learning_rate": 1.3600391304730174e-05,
+      "loss": 1.3431,
+      "step": 1449
+    },
+    {
+      "epoch": 1.174089068825911,
+      "grad_norm": 1.1956700418684363,
+      "learning_rate": 1.3592371073788595e-05,
+      "loss": 1.4223,
+      "step": 1450
+    },
+    {
+      "epoch": 1.1748987854251012,
+      "grad_norm": 1.2850373149613217,
+      "learning_rate": 1.3584348188815066e-05,
+      "loss": 1.4498,
+      "step": 1451
+    },
+    {
+      "epoch": 1.1757085020242914,
+      "grad_norm": 1.193949378852388,
+      "learning_rate": 1.357632265573687e-05,
+      "loss": 1.3437,
+      "step": 1452
+    },
+    {
+      "epoch": 1.1765182186234817,
+      "grad_norm": 1.1874525953583595,
+      "learning_rate": 1.356829448048324e-05,
+      "loss": 1.3978,
+      "step": 1453
+    },
+    {
+      "epoch": 1.177327935222672,
+      "grad_norm": 1.194001231503012,
+      "learning_rate": 1.3560263668985366e-05,
+      "loss": 1.2924,
+      "step": 1454
+    },
+    {
+      "epoch": 1.1781376518218623,
+      "grad_norm": 1.1791770992975843,
+      "learning_rate": 1.355223022717639e-05,
+      "loss": 1.3411,
+      "step": 1455
+    },
+    {
+      "epoch": 1.1789473684210527,
+      "grad_norm": 1.2243559430427213,
+      "learning_rate": 1.3544194160991388e-05,
+      "loss": 1.4058,
+      "step": 1456
+    },
+    {
+      "epoch": 1.1797570850202428,
+      "grad_norm": 1.1894939791550536,
+      "learning_rate": 1.353615547636738e-05,
+      "loss": 1.3892,
+      "step": 1457
+    },
+    {
+      "epoch": 1.1805668016194333,
+      "grad_norm": 1.2776607461433331,
+      "learning_rate": 1.3528114179243322e-05,
+      "loss": 1.3352,
+      "step": 1458
+    },
+    {
+      "epoch": 1.1813765182186236,
+      "grad_norm": 1.201017466707103,
+      "learning_rate": 1.3520070275560093e-05,
+      "loss": 1.3605,
+      "step": 1459
+    },
+    {
+      "epoch": 1.1821862348178138,
+      "grad_norm": 1.2221052297543349,
+      "learning_rate": 1.3512023771260507e-05,
+      "loss": 1.3831,
+      "step": 1460
+    },
+    {
+      "epoch": 1.182995951417004,
+      "grad_norm": 1.2535433317816922,
+      "learning_rate": 1.3503974672289295e-05,
+      "loss": 1.4064,
+      "step": 1461
+    },
+    {
+      "epoch": 1.1838056680161944,
+      "grad_norm": 1.25001419317799,
+      "learning_rate": 1.3495922984593101e-05,
+      "loss": 1.354,
+      "step": 1462
+    },
+    {
+      "epoch": 1.1846153846153846,
+      "grad_norm": 1.2089559561561227,
+      "learning_rate": 1.3487868714120494e-05,
+      "loss": 1.3245,
+      "step": 1463
+    },
+    {
+      "epoch": 1.185425101214575,
+      "grad_norm": 1.1679267338188017,
+      "learning_rate": 1.347981186682193e-05,
+      "loss": 1.4321,
+      "step": 1464
+    },
+    {
+      "epoch": 1.1862348178137652,
+      "grad_norm": 1.2411376067872044,
+      "learning_rate": 1.347175244864979e-05,
+      "loss": 1.4119,
+      "step": 1465
+    },
+    {
+      "epoch": 1.1870445344129554,
+      "grad_norm": 1.2595186312439597,
+      "learning_rate": 1.3463690465558346e-05,
+      "loss": 1.3369,
+      "step": 1466
+    },
+    {
+      "epoch": 1.1878542510121457,
+      "grad_norm": 1.1909117279605281,
+      "learning_rate": 1.3455625923503762e-05,
+      "loss": 1.3863,
+      "step": 1467
+    },
+    {
+      "epoch": 1.188663967611336,
+      "grad_norm": 1.2054369811230554,
+      "learning_rate": 1.344755882844409e-05,
+      "loss": 1.3767,
+      "step": 1468
+    },
+    {
+      "epoch": 1.1894736842105262,
+      "grad_norm": 1.2323426762141974,
+      "learning_rate": 1.3439489186339283e-05,
+      "loss": 1.377,
+      "step": 1469
+    },
+    {
+      "epoch": 1.1902834008097165,
+      "grad_norm": 1.2454039577692348,
+      "learning_rate": 1.3431417003151162e-05,
+      "loss": 1.3446,
+      "step": 1470
+    },
+    {
+      "epoch": 1.191093117408907,
+      "grad_norm": 1.2587301942793938,
+      "learning_rate": 1.3423342284843428e-05,
+      "loss": 1.3762,
+      "step": 1471
+    },
+    {
+      "epoch": 1.191902834008097,
+      "grad_norm": 1.2722119186788876,
+      "learning_rate": 1.3415265037381657e-05,
+      "loss": 1.3654,
+      "step": 1472
+    },
+    {
+      "epoch": 1.1927125506072875,
+      "grad_norm": 1.2805355921406205,
+      "learning_rate": 1.3407185266733294e-05,
+      "loss": 1.3376,
+      "step": 1473
+    },
+    {
+      "epoch": 1.1935222672064778,
+      "grad_norm": 1.2081785962123999,
+      "learning_rate": 1.3399102978867648e-05,
+      "loss": 1.4041,
+      "step": 1474
+    },
+    {
+      "epoch": 1.194331983805668,
+      "grad_norm": 1.1925409611242321,
+      "learning_rate": 1.3391018179755886e-05,
+      "loss": 1.3684,
+      "step": 1475
+    },
+    {
+      "epoch": 1.1951417004048583,
+      "grad_norm": 1.227526446454291,
+      "learning_rate": 1.3382930875371028e-05,
+      "loss": 1.4175,
+      "step": 1476
+    },
+    {
+      "epoch": 1.1959514170040486,
+      "grad_norm": 1.1817699551420382,
+      "learning_rate": 1.3374841071687949e-05,
+      "loss": 1.3924,
+      "step": 1477
+    },
+    {
+      "epoch": 1.1967611336032389,
+      "grad_norm": 1.1756054088231052,
+      "learning_rate": 1.3366748774683376e-05,
+      "loss": 1.357,
+      "step": 1478
+    },
+    {
+      "epoch": 1.1975708502024291,
+      "grad_norm": 1.222659746027629,
+      "learning_rate": 1.3358653990335863e-05,
+      "loss": 1.3636,
+      "step": 1479
+    },
+    {
+      "epoch": 1.1983805668016194,
+      "grad_norm": 1.2400190458109341,
+      "learning_rate": 1.3350556724625809e-05,
+      "loss": 1.3603,
+      "step": 1480
+    },
+    {
+      "epoch": 1.1991902834008097,
+      "grad_norm": 1.2088314629672083,
+      "learning_rate": 1.3342456983535457e-05,
+      "loss": 1.416,
+      "step": 1481
+    },
+    {
+      "epoch": 1.2,
+      "grad_norm": 1.2142465353436538,
+      "learning_rate": 1.3334354773048863e-05,
+      "loss": 1.3035,
+      "step": 1482
+    },
+    {
+      "epoch": 1.2008097165991902,
+      "grad_norm": 1.1986889230221165,
+      "learning_rate": 1.3326250099151911e-05,
+      "loss": 1.3957,
+      "step": 1483
+    },
+    {
+      "epoch": 1.2016194331983805,
+      "grad_norm": 1.2413091007676673,
+      "learning_rate": 1.331814296783231e-05,
+      "loss": 1.3504,
+      "step": 1484
+    },
+    {
+      "epoch": 1.2024291497975708,
+      "grad_norm": 1.2256638183999424,
+      "learning_rate": 1.3310033385079589e-05,
+      "loss": 1.3678,
+      "step": 1485
+    },
+    {
+      "epoch": 1.2032388663967613,
+      "grad_norm": 1.2400693672871794,
+      "learning_rate": 1.330192135688507e-05,
+      "loss": 1.4066,
+      "step": 1486
+    },
+    {
+      "epoch": 1.2040485829959513,
+      "grad_norm": 1.2362345111970168,
+      "learning_rate": 1.3293806889241898e-05,
+      "loss": 1.3888,
+      "step": 1487
+    },
+    {
+      "epoch": 1.2048582995951418,
+      "grad_norm": 1.2587009847578705,
+      "learning_rate": 1.3285689988145011e-05,
+      "loss": 1.4359,
+      "step": 1488
+    },
+    {
+      "epoch": 1.205668016194332,
+      "grad_norm": 1.1914663304562276,
+      "learning_rate": 1.3277570659591159e-05,
+      "loss": 1.3391,
+      "step": 1489
+    },
+    {
+      "epoch": 1.2064777327935223,
+      "grad_norm": 1.1834896782723838,
+      "learning_rate": 1.3269448909578866e-05,
+      "loss": 1.4029,
+      "step": 1490
+    },
+    {
+      "epoch": 1.2072874493927126,
+      "grad_norm": 1.1813018458720312,
+      "learning_rate": 1.3261324744108454e-05,
+      "loss": 1.415,
+      "step": 1491
+    },
+    {
+      "epoch": 1.2080971659919029,
+      "grad_norm": 1.1818375702199966,
+      "learning_rate": 1.3253198169182033e-05,
+      "loss": 1.3473,
+      "step": 1492
+    },
+    {
+      "epoch": 1.2089068825910931,
+      "grad_norm": 1.194541380193379,
+      "learning_rate": 1.3245069190803495e-05,
+      "loss": 1.3753,
+      "step": 1493
+    },
+    {
+      "epoch": 1.2097165991902834,
+      "grad_norm": 1.2134975986163172,
+      "learning_rate": 1.3236937814978493e-05,
+      "loss": 1.3514,
+      "step": 1494
+    },
+    {
+      "epoch": 1.2105263157894737,
+      "grad_norm": 1.1847950487403296,
+      "learning_rate": 1.3228804047714462e-05,
+      "loss": 1.3309,
+      "step": 1495
+    },
+    {
+      "epoch": 1.211336032388664,
+      "grad_norm": 1.1484211925319983,
+      "learning_rate": 1.322066789502061e-05,
+      "loss": 1.3631,
+      "step": 1496
+    },
+    {
+      "epoch": 1.2121457489878542,
+      "grad_norm": 1.2388971531409982,
+      "learning_rate": 1.3212529362907894e-05,
+      "loss": 1.3734,
+      "step": 1497
+    },
+    {
+      "epoch": 1.2129554655870445,
+      "grad_norm": 1.247111008797553,
+      "learning_rate": 1.3204388457389032e-05,
+      "loss": 1.3013,
+      "step": 1498
+    },
+    {
+      "epoch": 1.2137651821862347,
+      "grad_norm": 1.2555192712705792,
+      "learning_rate": 1.319624518447851e-05,
+      "loss": 1.3507,
+      "step": 1499
+    },
+    {
+      "epoch": 1.214574898785425,
+      "grad_norm": 1.3054090163651753,
+      "learning_rate": 1.3188099550192537e-05,
+      "loss": 1.3748,
+      "step": 1500
+    },
+    {
+      "epoch": 1.2153846153846155,
+      "grad_norm": 1.250883089302353,
+      "learning_rate": 1.317995156054909e-05,
+      "loss": 1.3597,
+      "step": 1501
+    },
+    {
+      "epoch": 1.2161943319838056,
+      "grad_norm": 1.2376049269488396,
+      "learning_rate": 1.3171801221567872e-05,
+      "loss": 1.3149,
+      "step": 1502
+    },
+    {
+      "epoch": 1.217004048582996,
+      "grad_norm": 1.1862482520157116,
+      "learning_rate": 1.3163648539270333e-05,
+      "loss": 1.4034,
+      "step": 1503
+    },
+    {
+      "epoch": 1.2178137651821863,
+      "grad_norm": 1.26463939374948,
+      "learning_rate": 1.315549351967964e-05,
+      "loss": 1.3365,
+      "step": 1504
+    },
+    {
+      "epoch": 1.2186234817813766,
+      "grad_norm": 1.2437748606370247,
+      "learning_rate": 1.31473361688207e-05,
+      "loss": 1.4052,
+      "step": 1505
+    },
+    {
+      "epoch": 1.2194331983805669,
+      "grad_norm": 1.2684663340272135,
+      "learning_rate": 1.3139176492720137e-05,
+      "loss": 1.3346,
+      "step": 1506
+    },
+    {
+      "epoch": 1.2202429149797571,
+      "grad_norm": 1.2311472588728176,
+      "learning_rate": 1.3131014497406288e-05,
+      "loss": 1.3399,
+      "step": 1507
+    },
+    {
+      "epoch": 1.2210526315789474,
+      "grad_norm": 1.262810193750128,
+      "learning_rate": 1.3122850188909216e-05,
+      "loss": 1.3896,
+      "step": 1508
+    },
+    {
+      "epoch": 1.2218623481781377,
+      "grad_norm": 1.227527399226647,
+      "learning_rate": 1.3114683573260677e-05,
+      "loss": 1.408,
+      "step": 1509
+    },
+    {
+      "epoch": 1.222672064777328,
+      "grad_norm": 1.2375397432956372,
+      "learning_rate": 1.3106514656494147e-05,
+      "loss": 1.3952,
+      "step": 1510
+    },
+    {
+      "epoch": 1.2234817813765182,
+      "grad_norm": 1.1550935938629225,
+      "learning_rate": 1.3098343444644793e-05,
+      "loss": 1.3888,
+      "step": 1511
+    },
+    {
+      "epoch": 1.2242914979757085,
+      "grad_norm": 1.1942656359224868,
+      "learning_rate": 1.3090169943749475e-05,
+      "loss": 1.3832,
+      "step": 1512
+    },
+    {
+      "epoch": 1.2251012145748987,
+      "grad_norm": 1.1727975184000596,
+      "learning_rate": 1.3081994159846753e-05,
+      "loss": 1.3812,
+      "step": 1513
+    },
+    {
+      "epoch": 1.225910931174089,
+      "grad_norm": 1.197178362301058,
+      "learning_rate": 1.3073816098976871e-05,
+      "loss": 1.3731,
+      "step": 1514
+    },
+    {
+      "epoch": 1.2267206477732793,
+      "grad_norm": 1.2142805284091027,
+      "learning_rate": 1.3065635767181748e-05,
+      "loss": 1.4057,
+      "step": 1515
+    },
+    {
+      "epoch": 1.2275303643724695,
+      "grad_norm": 1.1711215174781089,
+      "learning_rate": 1.3057453170504988e-05,
+      "loss": 1.4099,
+      "step": 1516
+    },
+    {
+      "epoch": 1.2283400809716598,
+      "grad_norm": 1.183153242000738,
+      "learning_rate": 1.304926831499187e-05,
+      "loss": 1.374,
+      "step": 1517
+    },
+    {
+      "epoch": 1.2291497975708503,
+      "grad_norm": 1.2202202744850468,
+      "learning_rate": 1.3041081206689335e-05,
+      "loss": 1.3325,
+      "step": 1518
+    },
+    {
+      "epoch": 1.2299595141700406,
+      "grad_norm": 1.1858098990909935,
+      "learning_rate": 1.3032891851645994e-05,
+      "loss": 1.3086,
+      "step": 1519
+    },
+    {
+      "epoch": 1.2307692307692308,
+      "grad_norm": 1.230959944898959,
+      "learning_rate": 1.302470025591211e-05,
+      "loss": 1.3373,
+      "step": 1520
+    },
+    {
+      "epoch": 1.231578947368421,
+      "grad_norm": 1.1451597949607148,
+      "learning_rate": 1.3016506425539615e-05,
+      "loss": 1.3721,
+      "step": 1521
+    },
+    {
+      "epoch": 1.2323886639676114,
+      "grad_norm": 1.3441292904286972,
+      "learning_rate": 1.3008310366582081e-05,
+      "loss": 1.4799,
+      "step": 1522
+    },
+    {
+      "epoch": 1.2331983805668016,
+      "grad_norm": 1.2341770886622436,
+      "learning_rate": 1.300011208509473e-05,
+      "loss": 1.3638,
+      "step": 1523
+    },
+    {
+      "epoch": 1.234008097165992,
+      "grad_norm": 1.1915183190289644,
+      "learning_rate": 1.2991911587134416e-05,
+      "loss": 1.4123,
+      "step": 1524
+    },
+    {
+      "epoch": 1.2348178137651822,
+      "grad_norm": 1.2162901503483015,
+      "learning_rate": 1.2983708878759655e-05,
+      "loss": 1.4574,
+      "step": 1525
+    },
+    {
+      "epoch": 1.2356275303643725,
+      "grad_norm": 1.2621342134046012,
+      "learning_rate": 1.2975503966030574e-05,
+      "loss": 1.3648,
+      "step": 1526
+    },
+    {
+      "epoch": 1.2364372469635627,
+      "grad_norm": 1.1896912894693648,
+      "learning_rate": 1.2967296855008932e-05,
+      "loss": 1.3118,
+      "step": 1527
+    },
+    {
+      "epoch": 1.237246963562753,
+      "grad_norm": 1.2023777230870145,
+      "learning_rate": 1.2959087551758121e-05,
+      "loss": 1.4103,
+      "step": 1528
+    },
+    {
+      "epoch": 1.2380566801619433,
+      "grad_norm": 1.228550678791859,
+      "learning_rate": 1.2950876062343147e-05,
+      "loss": 1.3102,
+      "step": 1529
+    },
+    {
+      "epoch": 1.2388663967611335,
+      "grad_norm": 1.2650925867678326,
+      "learning_rate": 1.2942662392830632e-05,
+      "loss": 1.388,
+      "step": 1530
+    },
+    {
+      "epoch": 1.2396761133603238,
+      "grad_norm": 1.1719459344716467,
+      "learning_rate": 1.2934446549288801e-05,
+      "loss": 1.3515,
+      "step": 1531
+    },
+    {
+      "epoch": 1.240485829959514,
+      "grad_norm": 1.2442167711303942,
+      "learning_rate": 1.2926228537787498e-05,
+      "loss": 1.3685,
+      "step": 1532
+    },
+    {
+      "epoch": 1.2412955465587046,
+      "grad_norm": 1.192558403731443,
+      "learning_rate": 1.2918008364398164e-05,
+      "loss": 1.3777,
+      "step": 1533
+    },
+    {
+      "epoch": 1.2421052631578948,
+      "grad_norm": 1.187107614370798,
+      "learning_rate": 1.2909786035193836e-05,
+      "loss": 1.444,
+      "step": 1534
+    },
+    {
+      "epoch": 1.242914979757085,
+      "grad_norm": 1.2441337176267466,
+      "learning_rate": 1.290156155624914e-05,
+      "loss": 1.3631,
+      "step": 1535
+    },
+    {
+      "epoch": 1.2437246963562754,
+      "grad_norm": 1.174168892326793,
+      "learning_rate": 1.2893334933640296e-05,
+      "loss": 1.3321,
+      "step": 1536
+    },
+    {
+      "epoch": 1.2445344129554656,
+      "grad_norm": 1.208004365868136,
+      "learning_rate": 1.2885106173445108e-05,
+      "loss": 1.363,
+      "step": 1537
+    },
+    {
+      "epoch": 1.245344129554656,
+      "grad_norm": 1.1950930591538556,
+      "learning_rate": 1.287687528174295e-05,
+      "loss": 1.3179,
+      "step": 1538
+    },
+    {
+      "epoch": 1.2461538461538462,
+      "grad_norm": 1.1790688919955996,
+      "learning_rate": 1.2868642264614787e-05,
+      "loss": 1.4061,
+      "step": 1539
+    },
+    {
+      "epoch": 1.2469635627530364,
+      "grad_norm": 1.1987689806967812,
+      "learning_rate": 1.286040712814314e-05,
+      "loss": 1.3569,
+      "step": 1540
+    },
+    {
+      "epoch": 1.2477732793522267,
+      "grad_norm": 1.1817370751262748,
+      "learning_rate": 1.2852169878412102e-05,
+      "loss": 1.3369,
+      "step": 1541
+    },
+    {
+      "epoch": 1.248582995951417,
+      "grad_norm": 1.2282449878735011,
+      "learning_rate": 1.2843930521507324e-05,
+      "loss": 1.3283,
+      "step": 1542
+    },
+    {
+      "epoch": 1.2493927125506072,
+      "grad_norm": 1.26597563474877,
+      "learning_rate": 1.2835689063516019e-05,
+      "loss": 1.4247,
+      "step": 1543
+    },
+    {
+      "epoch": 1.2502024291497975,
+      "grad_norm": 1.2159532514071452,
+      "learning_rate": 1.2827445510526945e-05,
+      "loss": 1.3501,
+      "step": 1544
+    },
+    {
+      "epoch": 1.2510121457489878,
+      "grad_norm": 1.2015354671386715,
+      "learning_rate": 1.2819199868630419e-05,
+      "loss": 1.4067,
+      "step": 1545
+    },
+    {
+      "epoch": 1.2518218623481783,
+      "grad_norm": 1.1488073343742868,
+      "learning_rate": 1.2810952143918284e-05,
+      "loss": 1.3881,
+      "step": 1546
+    },
+    {
+      "epoch": 1.2526315789473683,
+      "grad_norm": 1.1469365945891747,
+      "learning_rate": 1.2802702342483941e-05,
+      "loss": 1.2904,
+      "step": 1547
+    },
+    {
+      "epoch": 1.2534412955465588,
+      "grad_norm": 1.2210656858397426,
+      "learning_rate": 1.279445047042231e-05,
+      "loss": 1.3727,
+      "step": 1548
+    },
+    {
+      "epoch": 1.2542510121457489,
+      "grad_norm": 1.199892214891525,
+      "learning_rate": 1.278619653382985e-05,
+      "loss": 1.4002,
+      "step": 1549
+    },
+    {
+      "epoch": 1.2550607287449393,
+      "grad_norm": 1.233459919301486,
+      "learning_rate": 1.2777940538804545e-05,
+      "loss": 1.4107,
+      "step": 1550
+    },
+    {
+      "epoch": 1.2558704453441296,
+      "grad_norm": 1.147765149659166,
+      "learning_rate": 1.2769682491445892e-05,
+      "loss": 1.3451,
+      "step": 1551
+    },
+    {
+      "epoch": 1.2566801619433199,
+      "grad_norm": 1.1348340061961697,
+      "learning_rate": 1.276142239785491e-05,
+      "loss": 1.2995,
+      "step": 1552
+    },
+    {
+      "epoch": 1.2574898785425102,
+      "grad_norm": 1.2149990877087464,
+      "learning_rate": 1.275316026413413e-05,
+      "loss": 1.2791,
+      "step": 1553
+    },
+    {
+      "epoch": 1.2582995951417004,
+      "grad_norm": 1.1996055435289459,
+      "learning_rate": 1.274489609638759e-05,
+      "loss": 1.3854,
+      "step": 1554
+    },
+    {
+      "epoch": 1.2591093117408907,
+      "grad_norm": 1.2048053640102194,
+      "learning_rate": 1.2736629900720832e-05,
+      "loss": 1.3213,
+      "step": 1555
+    },
+    {
+      "epoch": 1.259919028340081,
+      "grad_norm": 1.1862492357917775,
+      "learning_rate": 1.2728361683240889e-05,
+      "loss": 1.4001,
+      "step": 1556
+    },
+    {
+      "epoch": 1.2607287449392712,
+      "grad_norm": 1.246276511809421,
+      "learning_rate": 1.2720091450056293e-05,
+      "loss": 1.3218,
+      "step": 1557
+    },
+    {
+      "epoch": 1.2615384615384615,
+      "grad_norm": 1.223220081058916,
+      "learning_rate": 1.2711819207277071e-05,
+      "loss": 1.356,
+      "step": 1558
+    },
+    {
+      "epoch": 1.2623481781376518,
+      "grad_norm": 1.2502562370531392,
+      "learning_rate": 1.2703544961014727e-05,
+      "loss": 1.3222,
+      "step": 1559
+    },
+    {
+      "epoch": 1.263157894736842,
+      "grad_norm": 1.224324008996058,
+      "learning_rate": 1.2695268717382242e-05,
+      "loss": 1.3478,
+      "step": 1560
+    },
+    {
+      "epoch": 1.2639676113360325,
+      "grad_norm": 1.2105633319160907,
+      "learning_rate": 1.268699048249408e-05,
+      "loss": 1.4391,
+      "step": 1561
+    },
+    {
+      "epoch": 1.2647773279352226,
+      "grad_norm": 1.2837901936738587,
+      "learning_rate": 1.2678710262466178e-05,
+      "loss": 1.4009,
+      "step": 1562
+    },
+    {
+      "epoch": 1.265587044534413,
+      "grad_norm": 1.2548245764229349,
+      "learning_rate": 1.2670428063415932e-05,
+      "loss": 1.4045,
+      "step": 1563
+    },
+    {
+      "epoch": 1.266396761133603,
+      "grad_norm": 1.1561209573769804,
+      "learning_rate": 1.26621438914622e-05,
+      "loss": 1.3793,
+      "step": 1564
+    },
+    {
+      "epoch": 1.2672064777327936,
+      "grad_norm": 1.1691572154518026,
+      "learning_rate": 1.2653857752725305e-05,
+      "loss": 1.3663,
+      "step": 1565
+    },
+    {
+      "epoch": 1.2680161943319839,
+      "grad_norm": 1.2306512954942797,
+      "learning_rate": 1.2645569653327024e-05,
+      "loss": 1.3454,
+      "step": 1566
+    },
+    {
+      "epoch": 1.2688259109311741,
+      "grad_norm": 1.1802175610787244,
+      "learning_rate": 1.2637279599390569e-05,
+      "loss": 1.369,
+      "step": 1567
+    },
+    {
+      "epoch": 1.2696356275303644,
+      "grad_norm": 1.1959566114951496,
+      "learning_rate": 1.2628987597040605e-05,
+      "loss": 1.3326,
+      "step": 1568
+    },
+    {
+      "epoch": 1.2704453441295547,
+      "grad_norm": 1.2258150541786852,
+      "learning_rate": 1.2620693652403241e-05,
+      "loss": 1.4098,
+      "step": 1569
+    },
+    {
+      "epoch": 1.271255060728745,
+      "grad_norm": 1.2413950001716263,
+      "learning_rate": 1.2612397771606015e-05,
+      "loss": 1.328,
+      "step": 1570
+    },
+    {
+      "epoch": 1.2720647773279352,
+      "grad_norm": 1.1898669800873198,
+      "learning_rate": 1.2604099960777896e-05,
+      "loss": 1.3843,
+      "step": 1571
+    },
+    {
+      "epoch": 1.2728744939271255,
+      "grad_norm": 1.2359955799975981,
+      "learning_rate": 1.2595800226049277e-05,
+      "loss": 1.4085,
+      "step": 1572
+    },
+    {
+      "epoch": 1.2736842105263158,
+      "grad_norm": 1.2178561670112904,
+      "learning_rate": 1.258749857355198e-05,
+      "loss": 1.2919,
+      "step": 1573
+    },
+    {
+      "epoch": 1.274493927125506,
+      "grad_norm": 1.2042766779705247,
+      "learning_rate": 1.2579195009419234e-05,
+      "loss": 1.3142,
+      "step": 1574
+    },
+    {
+      "epoch": 1.2753036437246963,
+      "grad_norm": 1.2110686014257106,
+      "learning_rate": 1.2570889539785683e-05,
+      "loss": 1.3382,
+      "step": 1575
+    },
+    {
+      "epoch": 1.2761133603238866,
+      "grad_norm": 1.2052344706519602,
+      "learning_rate": 1.2562582170787385e-05,
+      "loss": 1.4298,
+      "step": 1576
+    },
+    {
+      "epoch": 1.2769230769230768,
+      "grad_norm": 1.243022267671595,
+      "learning_rate": 1.2554272908561798e-05,
+      "loss": 1.3593,
+      "step": 1577
+    },
+    {
+      "epoch": 1.2777327935222673,
+      "grad_norm": 1.2033161377378727,
+      "learning_rate": 1.2545961759247775e-05,
+      "loss": 1.3927,
+      "step": 1578
+    },
+    {
+      "epoch": 1.2785425101214574,
+      "grad_norm": 1.2330632054756852,
+      "learning_rate": 1.2537648728985565e-05,
+      "loss": 1.4024,
+      "step": 1579
+    },
+    {
+      "epoch": 1.2793522267206479,
+      "grad_norm": 1.201478870274986,
+      "learning_rate": 1.2529333823916807e-05,
+      "loss": 1.4012,
+      "step": 1580
+    },
+    {
+      "epoch": 1.2801619433198381,
+      "grad_norm": 1.1664964260293886,
+      "learning_rate": 1.2521017050184531e-05,
+      "loss": 1.3771,
+      "step": 1581
+    },
+    {
+      "epoch": 1.2809716599190284,
+      "grad_norm": 1.180036690598496,
+      "learning_rate": 1.251269841393314e-05,
+      "loss": 1.3307,
+      "step": 1582
+    },
+    {
+      "epoch": 1.2817813765182187,
+      "grad_norm": 1.2026900475737514,
+      "learning_rate": 1.2504377921308408e-05,
+      "loss": 1.3842,
+      "step": 1583
+    },
+    {
+      "epoch": 1.282591093117409,
+      "grad_norm": 1.2152832958337092,
+      "learning_rate": 1.2496055578457496e-05,
+      "loss": 1.323,
+      "step": 1584
+    },
+    {
+      "epoch": 1.2834008097165992,
+      "grad_norm": 1.1654873586192795,
+      "learning_rate": 1.2487731391528919e-05,
+      "loss": 1.3791,
+      "step": 1585
+    },
+    {
+      "epoch": 1.2842105263157895,
+      "grad_norm": 1.2189615995410232,
+      "learning_rate": 1.2479405366672562e-05,
+      "loss": 1.4297,
+      "step": 1586
+    },
+    {
+      "epoch": 1.2850202429149797,
+      "grad_norm": 1.2231303739404338,
+      "learning_rate": 1.2471077510039665e-05,
+      "loss": 1.4837,
+      "step": 1587
+    },
+    {
+      "epoch": 1.28582995951417,
+      "grad_norm": 1.1937220120755236,
+      "learning_rate": 1.2462747827782818e-05,
+      "loss": 1.3918,
+      "step": 1588
+    },
+    {
+      "epoch": 1.2866396761133603,
+      "grad_norm": 1.263374882520897,
+      "learning_rate": 1.2454416326055964e-05,
+      "loss": 1.4037,
+      "step": 1589
+    },
+    {
+      "epoch": 1.2874493927125505,
+      "grad_norm": 1.2138622497968994,
+      "learning_rate": 1.2446083011014389e-05,
+      "loss": 1.3822,
+      "step": 1590
+    },
+    {
+      "epoch": 1.2882591093117408,
+      "grad_norm": 1.1753589986143762,
+      "learning_rate": 1.2437747888814722e-05,
+      "loss": 1.3099,
+      "step": 1591
+    },
+    {
+      "epoch": 1.289068825910931,
+      "grad_norm": 1.195846367618967,
+      "learning_rate": 1.242941096561492e-05,
+      "loss": 1.3619,
+      "step": 1592
+    },
+    {
+      "epoch": 1.2898785425101216,
+      "grad_norm": 1.1675847607196994,
+      "learning_rate": 1.2421072247574277e-05,
+      "loss": 1.3881,
+      "step": 1593
+    },
+    {
+      "epoch": 1.2906882591093116,
+      "grad_norm": 1.2109119056603415,
+      "learning_rate": 1.2412731740853405e-05,
+      "loss": 1.3412,
+      "step": 1594
+    },
+    {
+      "epoch": 1.291497975708502,
+      "grad_norm": 1.1465692161369532,
+      "learning_rate": 1.2404389451614253e-05,
+      "loss": 1.3789,
+      "step": 1595
+    },
+    {
+      "epoch": 1.2923076923076924,
+      "grad_norm": 1.1968403107276433,
+      "learning_rate": 1.2396045386020066e-05,
+      "loss": 1.3803,
+      "step": 1596
+    },
+    {
+      "epoch": 1.2931174089068826,
+      "grad_norm": 1.3220159374383795,
+      "learning_rate": 1.2387699550235419e-05,
+      "loss": 1.3292,
+      "step": 1597
+    },
+    {
+      "epoch": 1.293927125506073,
+      "grad_norm": 1.1851391120160597,
+      "learning_rate": 1.2379351950426188e-05,
+      "loss": 1.388,
+      "step": 1598
+    },
+    {
+      "epoch": 1.2947368421052632,
+      "grad_norm": 1.216804570437914,
+      "learning_rate": 1.2371002592759553e-05,
+      "loss": 1.3748,
+      "step": 1599
+    },
+    {
+      "epoch": 1.2955465587044535,
+      "grad_norm": 1.3015331306057583,
+      "learning_rate": 1.2362651483403985e-05,
+      "loss": 1.3344,
+      "step": 1600
+    },
+    {
+      "epoch": 1.2963562753036437,
+      "grad_norm": 1.1871910644240773,
+      "learning_rate": 1.2354298628529263e-05,
+      "loss": 1.3712,
+      "step": 1601
+    },
+    {
+      "epoch": 1.297165991902834,
+      "grad_norm": 1.2506226744247222,
+      "learning_rate": 1.2345944034306447e-05,
+      "loss": 1.3352,
+      "step": 1602
+    },
+    {
+      "epoch": 1.2979757085020243,
+      "grad_norm": 1.222415145371624,
+      "learning_rate": 1.2337587706907885e-05,
+      "loss": 1.4298,
+      "step": 1603
+    },
+    {
+      "epoch": 1.2987854251012145,
+      "grad_norm": 1.2046074904547404,
+      "learning_rate": 1.2329229652507199e-05,
+      "loss": 1.3016,
+      "step": 1604
+    },
+    {
+      "epoch": 1.2995951417004048,
+      "grad_norm": 1.1789481943544442,
+      "learning_rate": 1.2320869877279297e-05,
+      "loss": 1.4301,
+      "step": 1605
+    },
+    {
+      "epoch": 1.300404858299595,
+      "grad_norm": 1.2312716488423778,
+      "learning_rate": 1.2312508387400356e-05,
+      "loss": 1.3756,
+      "step": 1606
+    },
+    {
+      "epoch": 1.3012145748987853,
+      "grad_norm": 1.1876391454771218,
+      "learning_rate": 1.230414518904781e-05,
+      "loss": 1.347,
+      "step": 1607
+    },
+    {
+      "epoch": 1.3020242914979758,
+      "grad_norm": 1.1818075707743736,
+      "learning_rate": 1.2295780288400365e-05,
+      "loss": 1.2704,
+      "step": 1608
+    },
+    {
+      "epoch": 1.3028340080971659,
+      "grad_norm": 1.230394929800309,
+      "learning_rate": 1.2287413691637986e-05,
+      "loss": 1.4056,
+      "step": 1609
+    },
+    {
+      "epoch": 1.3036437246963564,
+      "grad_norm": 1.2093149673925736,
+      "learning_rate": 1.2279045404941883e-05,
+      "loss": 1.355,
+      "step": 1610
+    },
+    {
+      "epoch": 1.3044534412955466,
+      "grad_norm": 1.2368989128597225,
+      "learning_rate": 1.2270675434494523e-05,
+      "loss": 1.3272,
+      "step": 1611
+    },
+    {
+      "epoch": 1.305263157894737,
+      "grad_norm": 1.2457418382739227,
+      "learning_rate": 1.2262303786479603e-05,
+      "loss": 1.3158,
+      "step": 1612
+    },
+    {
+      "epoch": 1.3060728744939272,
+      "grad_norm": 1.193000113626177,
+      "learning_rate": 1.2253930467082082e-05,
+      "loss": 1.3006,
+      "step": 1613
+    },
+    {
+      "epoch": 1.3068825910931174,
+      "grad_norm": 1.177523460512185,
+      "learning_rate": 1.2245555482488134e-05,
+      "loss": 1.4537,
+      "step": 1614
+    },
+    {
+      "epoch": 1.3076923076923077,
+      "grad_norm": 1.20594416334784,
+      "learning_rate": 1.2237178838885168e-05,
+      "loss": 1.3691,
+      "step": 1615
+    },
+    {
+      "epoch": 1.308502024291498,
+      "grad_norm": 1.2486014445170293,
+      "learning_rate": 1.2228800542461828e-05,
+      "loss": 1.333,
+      "step": 1616
+    },
+    {
+      "epoch": 1.3093117408906882,
+      "grad_norm": 1.183684994706002,
+      "learning_rate": 1.2220420599407965e-05,
+      "loss": 1.3391,
+      "step": 1617
+    },
+    {
+      "epoch": 1.3101214574898785,
+      "grad_norm": 1.2014854800059243,
+      "learning_rate": 1.2212039015914656e-05,
+      "loss": 1.3952,
+      "step": 1618
+    },
+    {
+      "epoch": 1.3109311740890688,
+      "grad_norm": 1.2525560938912015,
+      "learning_rate": 1.2203655798174188e-05,
+      "loss": 1.3096,
+      "step": 1619
+    },
+    {
+      "epoch": 1.311740890688259,
+      "grad_norm": 1.1545428943491032,
+      "learning_rate": 1.2195270952380052e-05,
+      "loss": 1.3919,
+      "step": 1620
+    },
+    {
+      "epoch": 1.3125506072874493,
+      "grad_norm": 1.1981415320424176,
+      "learning_rate": 1.2186884484726948e-05,
+      "loss": 1.3424,
+      "step": 1621
+    },
+    {
+      "epoch": 1.3133603238866396,
+      "grad_norm": 1.2551393649132647,
+      "learning_rate": 1.2178496401410772e-05,
+      "loss": 1.3471,
+      "step": 1622
+    },
+    {
+      "epoch": 1.31417004048583,
+      "grad_norm": 1.2509875632018495,
+      "learning_rate": 1.2170106708628604e-05,
+      "loss": 1.3454,
+      "step": 1623
+    },
+    {
+      "epoch": 1.3149797570850201,
+      "grad_norm": 1.2300196893904036,
+      "learning_rate": 1.2161715412578729e-05,
+      "loss": 1.3805,
+      "step": 1624
+    },
+    {
+      "epoch": 1.3157894736842106,
+      "grad_norm": 1.1907433595568389,
+      "learning_rate": 1.215332251946061e-05,
+      "loss": 1.3729,
+      "step": 1625
+    },
+    {
+      "epoch": 1.3165991902834009,
+      "grad_norm": 1.16306796314768,
+      "learning_rate": 1.2144928035474886e-05,
+      "loss": 1.314,
+      "step": 1626
+    },
+    {
+      "epoch": 1.3174089068825912,
+      "grad_norm": 1.1442311854910465,
+      "learning_rate": 1.213653196682337e-05,
+      "loss": 1.3982,
+      "step": 1627
+    },
+    {
+      "epoch": 1.3182186234817814,
+      "grad_norm": 1.2255785520913092,
+      "learning_rate": 1.2128134319709057e-05,
+      "loss": 1.4047,
+      "step": 1628
+    },
+    {
+      "epoch": 1.3190283400809717,
+      "grad_norm": 1.144767078265005,
+      "learning_rate": 1.21197351003361e-05,
+      "loss": 1.3996,
+      "step": 1629
+    },
+    {
+      "epoch": 1.319838056680162,
+      "grad_norm": 1.2322890116383336,
+      "learning_rate": 1.2111334314909811e-05,
+      "loss": 1.3791,
+      "step": 1630
+    },
+    {
+      "epoch": 1.3206477732793522,
+      "grad_norm": 1.2107662604876777,
+      "learning_rate": 1.2102931969636664e-05,
+      "loss": 1.3275,
+      "step": 1631
+    },
+    {
+      "epoch": 1.3214574898785425,
+      "grad_norm": 1.2271628627713616,
+      "learning_rate": 1.2094528070724286e-05,
+      "loss": 1.3068,
+      "step": 1632
+    },
+    {
+      "epoch": 1.3222672064777328,
+      "grad_norm": 1.240148364284859,
+      "learning_rate": 1.2086122624381446e-05,
+      "loss": 1.3663,
+      "step": 1633
+    },
+    {
+      "epoch": 1.323076923076923,
+      "grad_norm": 1.2055440099566974,
+      "learning_rate": 1.2077715636818066e-05,
+      "loss": 1.4334,
+      "step": 1634
+    },
+    {
+      "epoch": 1.3238866396761133,
+      "grad_norm": 1.230867923850948,
+      "learning_rate": 1.2069307114245197e-05,
+      "loss": 1.3325,
+      "step": 1635
+    },
+    {
+      "epoch": 1.3246963562753036,
+      "grad_norm": 1.1621119132494113,
+      "learning_rate": 1.2060897062875027e-05,
+      "loss": 1.3705,
+      "step": 1636
+    },
+    {
+      "epoch": 1.3255060728744938,
+      "grad_norm": 1.2483246064419058,
+      "learning_rate": 1.2052485488920877e-05,
+      "loss": 1.3473,
+      "step": 1637
+    },
+    {
+      "epoch": 1.3263157894736843,
+      "grad_norm": 1.2207617901689776,
+      "learning_rate": 1.2044072398597188e-05,
+      "loss": 1.3736,
+      "step": 1638
+    },
+    {
+      "epoch": 1.3271255060728744,
+      "grad_norm": 1.181566130734583,
+      "learning_rate": 1.2035657798119527e-05,
+      "loss": 1.3351,
+      "step": 1639
+    },
+    {
+      "epoch": 1.3279352226720649,
+      "grad_norm": 1.2130403891894281,
+      "learning_rate": 1.2027241693704567e-05,
+      "loss": 1.4527,
+      "step": 1640
+    },
+    {
+      "epoch": 1.3287449392712551,
+      "grad_norm": 1.2379457536405214,
+      "learning_rate": 1.2018824091570103e-05,
+      "loss": 1.4002,
+      "step": 1641
+    },
+    {
+      "epoch": 1.3295546558704454,
+      "grad_norm": 1.2287910940298021,
+      "learning_rate": 1.2010404997935032e-05,
+      "loss": 1.4004,
+      "step": 1642
+    },
+    {
+      "epoch": 1.3303643724696357,
+      "grad_norm": 1.1412833841977206,
+      "learning_rate": 1.2001984419019353e-05,
+      "loss": 1.3346,
+      "step": 1643
+    },
+    {
+      "epoch": 1.331174089068826,
+      "grad_norm": 1.15781602550099,
+      "learning_rate": 1.1993562361044157e-05,
+      "loss": 1.4022,
+      "step": 1644
+    },
+    {
+      "epoch": 1.3319838056680162,
+      "grad_norm": 1.1570840461641474,
+      "learning_rate": 1.1985138830231638e-05,
+      "loss": 1.3471,
+      "step": 1645
+    },
+    {
+      "epoch": 1.3327935222672065,
+      "grad_norm": 1.1957634256697691,
+      "learning_rate": 1.1976713832805071e-05,
+      "loss": 1.3155,
+      "step": 1646
+    },
+    {
+      "epoch": 1.3336032388663968,
+      "grad_norm": 1.1870263239595638,
+      "learning_rate": 1.1968287374988819e-05,
+      "loss": 1.3684,
+      "step": 1647
+    },
+    {
+      "epoch": 1.334412955465587,
+      "grad_norm": 1.2264800114026502,
+      "learning_rate": 1.1959859463008316e-05,
+      "loss": 1.3983,
+      "step": 1648
+    },
+    {
+      "epoch": 1.3352226720647773,
+      "grad_norm": 1.2388487173488427,
+      "learning_rate": 1.1951430103090079e-05,
+      "loss": 1.3473,
+      "step": 1649
+    },
+    {
+      "epoch": 1.3360323886639676,
+      "grad_norm": 1.2542679619160093,
+      "learning_rate": 1.1942999301461694e-05,
+      "loss": 1.2761,
+      "step": 1650
+    },
+    {
+      "epoch": 1.3368421052631578,
+      "grad_norm": 1.1812219242563955,
+      "learning_rate": 1.1934567064351802e-05,
+      "loss": 1.3625,
+      "step": 1651
+    },
+    {
+      "epoch": 1.337651821862348,
+      "grad_norm": 1.2130955096327123,
+      "learning_rate": 1.192613339799012e-05,
+      "loss": 1.3876,
+      "step": 1652
+    },
+    {
+      "epoch": 1.3384615384615386,
+      "grad_norm": 1.261762862681965,
+      "learning_rate": 1.1917698308607409e-05,
+      "loss": 1.3848,
+      "step": 1653
+    },
+    {
+      "epoch": 1.3392712550607286,
+      "grad_norm": 1.187112353472305,
+      "learning_rate": 1.1909261802435485e-05,
+      "loss": 1.3847,
+      "step": 1654
+    },
+    {
+      "epoch": 1.3400809716599191,
+      "grad_norm": 1.1853951778926974,
+      "learning_rate": 1.1900823885707216e-05,
+      "loss": 1.3522,
+      "step": 1655
+    },
+    {
+      "epoch": 1.3408906882591092,
+      "grad_norm": 1.2146857185059894,
+      "learning_rate": 1.1892384564656499e-05,
+      "loss": 1.3787,
+      "step": 1656
+    },
+    {
+      "epoch": 1.3417004048582997,
+      "grad_norm": 1.3043886121989412,
+      "learning_rate": 1.1883943845518282e-05,
+      "loss": 1.3538,
+      "step": 1657
+    },
+    {
+      "epoch": 1.34251012145749,
+      "grad_norm": 1.184170924694241,
+      "learning_rate": 1.187550173452854e-05,
+      "loss": 1.3652,
+      "step": 1658
+    },
+    {
+      "epoch": 1.3433198380566802,
+      "grad_norm": 1.2197338560870996,
+      "learning_rate": 1.1867058237924276e-05,
+      "loss": 1.3972,
+      "step": 1659
+    },
+    {
+      "epoch": 1.3441295546558705,
+      "grad_norm": 1.158381117646045,
+      "learning_rate": 1.1858613361943518e-05,
+      "loss": 1.3444,
+      "step": 1660
+    },
+    {
+      "epoch": 1.3449392712550607,
+      "grad_norm": 1.1488130103833536,
+      "learning_rate": 1.1850167112825316e-05,
+      "loss": 1.3224,
+      "step": 1661
+    },
+    {
+      "epoch": 1.345748987854251,
+      "grad_norm": 1.189395323195293,
+      "learning_rate": 1.1841719496809725e-05,
+      "loss": 1.3928,
+      "step": 1662
+    },
+    {
+      "epoch": 1.3465587044534413,
+      "grad_norm": 1.2004459789438373,
+      "learning_rate": 1.1833270520137819e-05,
+      "loss": 1.3517,
+      "step": 1663
+    },
+    {
+      "epoch": 1.3473684210526315,
+      "grad_norm": 1.204495169303041,
+      "learning_rate": 1.182482018905167e-05,
+      "loss": 1.4155,
+      "step": 1664
+    },
+    {
+      "epoch": 1.3481781376518218,
+      "grad_norm": 1.2633604245127454,
+      "learning_rate": 1.1816368509794365e-05,
+      "loss": 1.4195,
+      "step": 1665
+    },
+    {
+      "epoch": 1.348987854251012,
+      "grad_norm": 1.180591301251816,
+      "learning_rate": 1.1807915488609968e-05,
+      "loss": 1.3682,
+      "step": 1666
+    },
+    {
+      "epoch": 1.3497975708502024,
+      "grad_norm": 1.1375804044964366,
+      "learning_rate": 1.1799461131743548e-05,
+      "loss": 1.3662,
+      "step": 1667
+    },
+    {
+      "epoch": 1.3506072874493928,
+      "grad_norm": 1.1650602887537418,
+      "learning_rate": 1.179100544544115e-05,
+      "loss": 1.3532,
+      "step": 1668
+    },
+    {
+      "epoch": 1.351417004048583,
+      "grad_norm": 1.1656597543760445,
+      "learning_rate": 1.1782548435949814e-05,
+      "loss": 1.3439,
+      "step": 1669
+    },
+    {
+      "epoch": 1.3522267206477734,
+      "grad_norm": 1.1984145456570525,
+      "learning_rate": 1.177409010951755e-05,
+      "loss": 1.3584,
+      "step": 1670
+    },
+    {
+      "epoch": 1.3530364372469634,
+      "grad_norm": 1.2125180037933434,
+      "learning_rate": 1.1765630472393338e-05,
+      "loss": 1.3553,
+      "step": 1671
+    },
+    {
+      "epoch": 1.353846153846154,
+      "grad_norm": 1.278487148224879,
+      "learning_rate": 1.1757169530827129e-05,
+      "loss": 1.3817,
+      "step": 1672
+    },
+    {
+      "epoch": 1.3546558704453442,
+      "grad_norm": 1.2618389577279896,
+      "learning_rate": 1.1748707291069846e-05,
+      "loss": 1.4483,
+      "step": 1673
+    },
+    {
+      "epoch": 1.3554655870445345,
+      "grad_norm": 1.2332866269192326,
+      "learning_rate": 1.1740243759373358e-05,
+      "loss": 1.3525,
+      "step": 1674
+    },
+    {
+      "epoch": 1.3562753036437247,
+      "grad_norm": 1.2162494031574367,
+      "learning_rate": 1.1731778941990497e-05,
+      "loss": 1.3197,
+      "step": 1675
+    },
+    {
+      "epoch": 1.357085020242915,
+      "grad_norm": 1.182031395601063,
+      "learning_rate": 1.1723312845175041e-05,
+      "loss": 1.3705,
+      "step": 1676
+    },
+    {
+      "epoch": 1.3578947368421053,
+      "grad_norm": 1.188685842481188,
+      "learning_rate": 1.1714845475181716e-05,
+      "loss": 1.3931,
+      "step": 1677
+    },
+    {
+      "epoch": 1.3587044534412955,
+      "grad_norm": 1.216181329532205,
+      "learning_rate": 1.1706376838266185e-05,
+      "loss": 1.3821,
+      "step": 1678
+    },
+    {
+      "epoch": 1.3595141700404858,
+      "grad_norm": 1.2693089283894199,
+      "learning_rate": 1.169790694068505e-05,
+      "loss": 1.3803,
+      "step": 1679
+    },
+    {
+      "epoch": 1.360323886639676,
+      "grad_norm": 1.214964553877776,
+      "learning_rate": 1.1689435788695844e-05,
+      "loss": 1.3528,
+      "step": 1680
+    },
+    {
+      "epoch": 1.3611336032388663,
+      "grad_norm": 1.2317976939891357,
+      "learning_rate": 1.1680963388557028e-05,
+      "loss": 1.4125,
+      "step": 1681
+    },
+    {
+      "epoch": 1.3619433198380566,
+      "grad_norm": 1.2844113480291337,
+      "learning_rate": 1.1672489746527979e-05,
+      "loss": 1.3941,
+      "step": 1682
+    },
+    {
+      "epoch": 1.362753036437247,
+      "grad_norm": 1.2279453781778626,
+      "learning_rate": 1.1664014868869e-05,
+      "loss": 1.3184,
+      "step": 1683
+    },
+    {
+      "epoch": 1.3635627530364371,
+      "grad_norm": 1.2537576929676622,
+      "learning_rate": 1.16555387618413e-05,
+      "loss": 1.3569,
+      "step": 1684
+    },
+    {
+      "epoch": 1.3643724696356276,
+      "grad_norm": 1.2833642997336299,
+      "learning_rate": 1.1647061431707e-05,
+      "loss": 1.4056,
+      "step": 1685
+    },
+    {
+      "epoch": 1.3651821862348177,
+      "grad_norm": 1.235092352736005,
+      "learning_rate": 1.1638582884729127e-05,
+      "loss": 1.4097,
+      "step": 1686
+    },
+    {
+      "epoch": 1.3659919028340082,
+      "grad_norm": 1.187634661590241,
+      "learning_rate": 1.16301031271716e-05,
+      "loss": 1.4244,
+      "step": 1687
+    },
+    {
+      "epoch": 1.3668016194331984,
+      "grad_norm": 1.19737827161909,
+      "learning_rate": 1.1621622165299233e-05,
+      "loss": 1.3609,
+      "step": 1688
+    },
+    {
+      "epoch": 1.3676113360323887,
+      "grad_norm": 1.2414142941245716,
+      "learning_rate": 1.161314000537774e-05,
+      "loss": 1.4491,
+      "step": 1689
+    },
+    {
+      "epoch": 1.368421052631579,
+      "grad_norm": 1.231861969240117,
+      "learning_rate": 1.1604656653673707e-05,
+      "loss": 1.3283,
+      "step": 1690
+    },
+    {
+      "epoch": 1.3692307692307693,
+      "grad_norm": 1.1830853058793631,
+      "learning_rate": 1.1596172116454609e-05,
+      "loss": 1.3827,
+      "step": 1691
+    },
+    {
+      "epoch": 1.3700404858299595,
+      "grad_norm": 1.2656452551880795,
+      "learning_rate": 1.1587686399988793e-05,
+      "loss": 1.419,
+      "step": 1692
+    },
+    {
+      "epoch": 1.3708502024291498,
+      "grad_norm": 1.2452981014868072,
+      "learning_rate": 1.157919951054548e-05,
+      "loss": 1.3345,
+      "step": 1693
+    },
+    {
+      "epoch": 1.37165991902834,
+      "grad_norm": 1.2262829471723262,
+      "learning_rate": 1.1570711454394759e-05,
+      "loss": 1.3604,
+      "step": 1694
+    },
+    {
+      "epoch": 1.3724696356275303,
+      "grad_norm": 1.1921863334032212,
+      "learning_rate": 1.156222223780757e-05,
+      "loss": 1.3358,
+      "step": 1695
+    },
+    {
+      "epoch": 1.3732793522267206,
+      "grad_norm": 1.2818863761177053,
+      "learning_rate": 1.1553731867055724e-05,
+      "loss": 1.3675,
+      "step": 1696
+    },
+    {
+      "epoch": 1.3740890688259109,
+      "grad_norm": 1.2743486168767448,
+      "learning_rate": 1.1545240348411877e-05,
+      "loss": 1.4062,
+      "step": 1697
+    },
+    {
+      "epoch": 1.3748987854251011,
+      "grad_norm": 1.2406034474747183,
+      "learning_rate": 1.1536747688149537e-05,
+      "loss": 1.4102,
+      "step": 1698
+    },
+    {
+      "epoch": 1.3757085020242914,
+      "grad_norm": 1.2104446295299616,
+      "learning_rate": 1.1528253892543053e-05,
+      "loss": 1.44,
+      "step": 1699
+    },
+    {
+      "epoch": 1.376518218623482,
+      "grad_norm": 1.176123338179055,
+      "learning_rate": 1.1519758967867608e-05,
+      "loss": 1.4084,
+      "step": 1700
+    },
+    {
+      "epoch": 1.377327935222672,
+      "grad_norm": 1.2508166417186997,
+      "learning_rate": 1.1511262920399233e-05,
+      "loss": 1.3585,
+      "step": 1701
+    },
+    {
+      "epoch": 1.3781376518218624,
+      "grad_norm": 1.1964180815673577,
+      "learning_rate": 1.1502765756414776e-05,
+      "loss": 1.3996,
+      "step": 1702
+    },
+    {
+      "epoch": 1.3789473684210527,
+      "grad_norm": 1.2382693764014767,
+      "learning_rate": 1.1494267482191912e-05,
+      "loss": 1.2927,
+      "step": 1703
+    },
+    {
+      "epoch": 1.379757085020243,
+      "grad_norm": 1.346059359155808,
+      "learning_rate": 1.1485768104009141e-05,
+      "loss": 1.3396,
+      "step": 1704
+    },
+    {
+      "epoch": 1.3805668016194332,
+      "grad_norm": 1.3195598291408253,
+      "learning_rate": 1.1477267628145777e-05,
+      "loss": 1.3325,
+      "step": 1705
+    },
+    {
+      "epoch": 1.3813765182186235,
+      "grad_norm": 1.278714888738768,
+      "learning_rate": 1.146876606088194e-05,
+      "loss": 1.3388,
+      "step": 1706
+    },
+    {
+      "epoch": 1.3821862348178138,
+      "grad_norm": 1.2060319063156253,
+      "learning_rate": 1.1460263408498557e-05,
+      "loss": 1.337,
+      "step": 1707
+    },
+    {
+      "epoch": 1.382995951417004,
+      "grad_norm": 1.3011345426618277,
+      "learning_rate": 1.1451759677277367e-05,
+      "loss": 1.4089,
+      "step": 1708
+    },
+    {
+      "epoch": 1.3838056680161943,
+      "grad_norm": 1.238105150141394,
+      "learning_rate": 1.1443254873500897e-05,
+      "loss": 1.3463,
+      "step": 1709
+    },
+    {
+      "epoch": 1.3846153846153846,
+      "grad_norm": 1.2043600689977974,
+      "learning_rate": 1.1434749003452467e-05,
+      "loss": 1.4289,
+      "step": 1710
+    },
+    {
+      "epoch": 1.3854251012145749,
+      "grad_norm": 1.252328542955408,
+      "learning_rate": 1.1426242073416183e-05,
+      "loss": 1.3714,
+      "step": 1711
+    },
+    {
+      "epoch": 1.3862348178137651,
+      "grad_norm": 1.302041753695186,
+      "learning_rate": 1.1417734089676939e-05,
+      "loss": 1.363,
+      "step": 1712
+    },
+    {
+      "epoch": 1.3870445344129554,
+      "grad_norm": 1.2218471138487301,
+      "learning_rate": 1.140922505852041e-05,
+      "loss": 1.3402,
+      "step": 1713
+    },
+    {
+      "epoch": 1.3878542510121457,
+      "grad_norm": 1.2397585007328973,
+      "learning_rate": 1.1400714986233035e-05,
+      "loss": 1.3859,
+      "step": 1714
+    },
+    {
+      "epoch": 1.3886639676113361,
+      "grad_norm": 1.2723411222710095,
+      "learning_rate": 1.1392203879102027e-05,
+      "loss": 1.4312,
+      "step": 1715
+    },
+    {
+      "epoch": 1.3894736842105262,
+      "grad_norm": 1.2616308378503096,
+      "learning_rate": 1.1383691743415364e-05,
+      "loss": 1.3441,
+      "step": 1716
+    },
+    {
+      "epoch": 1.3902834008097167,
+      "grad_norm": 1.2096630244554334,
+      "learning_rate": 1.1375178585461788e-05,
+      "loss": 1.3332,
+      "step": 1717
+    },
+    {
+      "epoch": 1.391093117408907,
+      "grad_norm": 1.159693781497365,
+      "learning_rate": 1.136666441153079e-05,
+      "loss": 1.3252,
+      "step": 1718
+    },
+    {
+      "epoch": 1.3919028340080972,
+      "grad_norm": 1.1854318858942796,
+      "learning_rate": 1.1358149227912613e-05,
+      "loss": 1.3676,
+      "step": 1719
+    },
+    {
+      "epoch": 1.3927125506072875,
+      "grad_norm": 1.2233170433120137,
+      "learning_rate": 1.1349633040898246e-05,
+      "loss": 1.399,
+      "step": 1720
+    },
+    {
+      "epoch": 1.3935222672064778,
+      "grad_norm": 1.2879065439442032,
+      "learning_rate": 1.1341115856779423e-05,
+      "loss": 1.4255,
+      "step": 1721
+    },
+    {
+      "epoch": 1.394331983805668,
+      "grad_norm": 1.2707729718124967,
+      "learning_rate": 1.133259768184861e-05,
+      "loss": 1.3265,
+      "step": 1722
+    },
+    {
+      "epoch": 1.3951417004048583,
+      "grad_norm": 1.236173964659415,
+      "learning_rate": 1.1324078522399005e-05,
+      "loss": 1.4351,
+      "step": 1723
+    },
+    {
+      "epoch": 1.3959514170040486,
+      "grad_norm": 1.2153770369914687,
+      "learning_rate": 1.1315558384724537e-05,
+      "loss": 1.4008,
+      "step": 1724
+    },
+    {
+      "epoch": 1.3967611336032388,
+      "grad_norm": 1.1899505676690036,
+      "learning_rate": 1.1307037275119854e-05,
+      "loss": 1.3839,
+      "step": 1725
+    },
+    {
+      "epoch": 1.397570850202429,
+      "grad_norm": 1.1928149560420538,
+      "learning_rate": 1.1298515199880327e-05,
+      "loss": 1.3546,
+      "step": 1726
+    },
+    {
+      "epoch": 1.3983805668016194,
+      "grad_norm": 1.2238669730900604,
+      "learning_rate": 1.1289992165302036e-05,
+      "loss": 1.2984,
+      "step": 1727
+    },
+    {
+      "epoch": 1.3991902834008096,
+      "grad_norm": 1.1887889650994987,
+      "learning_rate": 1.1281468177681767e-05,
+      "loss": 1.3361,
+      "step": 1728
+    },
+    {
+      "epoch": 1.4,
+      "grad_norm": 1.2214260231171459,
+      "learning_rate": 1.1272943243317017e-05,
+      "loss": 1.3368,
+      "step": 1729
+    },
+    {
+      "epoch": 1.4008097165991904,
+      "grad_norm": 1.206685962166214,
+      "learning_rate": 1.1264417368505981e-05,
+      "loss": 1.377,
+      "step": 1730
+    },
+    {
+      "epoch": 1.4016194331983804,
+      "grad_norm": 1.2113952130760868,
+      "learning_rate": 1.1255890559547549e-05,
+      "loss": 1.3545,
+      "step": 1731
+    },
+    {
+      "epoch": 1.402429149797571,
+      "grad_norm": 1.1657293085603149,
+      "learning_rate": 1.1247362822741292e-05,
+      "loss": 1.3407,
+      "step": 1732
+    },
+    {
+      "epoch": 1.4032388663967612,
+      "grad_norm": 1.1487734412817017,
+      "learning_rate": 1.123883416438748e-05,
+      "loss": 1.2937,
+      "step": 1733
+    },
+    {
+      "epoch": 1.4040485829959515,
+      "grad_norm": 1.1513182139430822,
+      "learning_rate": 1.1230304590787059e-05,
+      "loss": 1.4015,
+      "step": 1734
+    },
+    {
+      "epoch": 1.4048582995951417,
+      "grad_norm": 1.2294748713620087,
+      "learning_rate": 1.1221774108241646e-05,
+      "loss": 1.4282,
+      "step": 1735
+    },
+    {
+      "epoch": 1.405668016194332,
+      "grad_norm": 1.2580084797725928,
+      "learning_rate": 1.121324272305353e-05,
+      "loss": 1.4106,
+      "step": 1736
+    },
+    {
+      "epoch": 1.4064777327935223,
+      "grad_norm": 1.2246093554948188,
+      "learning_rate": 1.1204710441525677e-05,
+      "loss": 1.3874,
+      "step": 1737
+    },
+    {
+      "epoch": 1.4072874493927126,
+      "grad_norm": 1.2289401792154757,
+      "learning_rate": 1.119617726996171e-05,
+      "loss": 1.4159,
+      "step": 1738
+    },
+    {
+      "epoch": 1.4080971659919028,
+      "grad_norm": 1.1896362142853578,
+      "learning_rate": 1.1187643214665905e-05,
+      "loss": 1.3545,
+      "step": 1739
+    },
+    {
+      "epoch": 1.408906882591093,
+      "grad_norm": 1.1629116832244053,
+      "learning_rate": 1.117910828194319e-05,
+      "loss": 1.3602,
+      "step": 1740
+    },
+    {
+      "epoch": 1.4097165991902834,
+      "grad_norm": 1.2609850112703145,
+      "learning_rate": 1.117057247809915e-05,
+      "loss": 1.4164,
+      "step": 1741
+    },
+    {
+      "epoch": 1.4105263157894736,
+      "grad_norm": 1.220342239587353,
+      "learning_rate": 1.1162035809440005e-05,
+      "loss": 1.3738,
+      "step": 1742
+    },
+    {
+      "epoch": 1.411336032388664,
+      "grad_norm": 1.168299823561213,
+      "learning_rate": 1.1153498282272626e-05,
+      "loss": 1.3956,
+      "step": 1743
+    },
+    {
+      "epoch": 1.4121457489878542,
+      "grad_norm": 1.1755175754663574,
+      "learning_rate": 1.11449599029045e-05,
+      "loss": 1.3535,
+      "step": 1744
+    },
+    {
+      "epoch": 1.4129554655870447,
+      "grad_norm": 1.2143297219911169,
+      "learning_rate": 1.1136420677643763e-05,
+      "loss": 1.3979,
+      "step": 1745
+    },
+    {
+      "epoch": 1.4137651821862347,
+      "grad_norm": 1.1767586889180381,
+      "learning_rate": 1.1127880612799158e-05,
+      "loss": 1.4074,
+      "step": 1746
+    },
+    {
+      "epoch": 1.4145748987854252,
+      "grad_norm": 1.1817264441831037,
+      "learning_rate": 1.1119339714680062e-05,
+      "loss": 1.3971,
+      "step": 1747
+    },
+    {
+      "epoch": 1.4153846153846155,
+      "grad_norm": 1.1851169547567846,
+      "learning_rate": 1.111079798959646e-05,
+      "loss": 1.3668,
+      "step": 1748
+    },
+    {
+      "epoch": 1.4161943319838057,
+      "grad_norm": 1.2004222820473625,
+      "learning_rate": 1.1102255443858953e-05,
+      "loss": 1.3868,
+      "step": 1749
+    },
+    {
+      "epoch": 1.417004048582996,
+      "grad_norm": 1.1884889082407253,
+      "learning_rate": 1.1093712083778748e-05,
+      "loss": 1.3551,
+      "step": 1750
+    },
+    {
+      "epoch": 1.4178137651821863,
+      "grad_norm": 1.1839821452630235,
+      "learning_rate": 1.108516791566764e-05,
+      "loss": 1.3438,
+      "step": 1751
+    },
+    {
+      "epoch": 1.4186234817813765,
+      "grad_norm": 1.2051936032476567,
+      "learning_rate": 1.1076622945838045e-05,
+      "loss": 1.3842,
+      "step": 1752
+    },
+    {
+      "epoch": 1.4194331983805668,
+      "grad_norm": 1.2344360426305427,
+      "learning_rate": 1.1068077180602953e-05,
+      "loss": 1.3924,
+      "step": 1753
+    },
+    {
+      "epoch": 1.420242914979757,
+      "grad_norm": 1.2188954123830977,
+      "learning_rate": 1.1059530626275948e-05,
+      "loss": 1.4137,
+      "step": 1754
+    },
+    {
+      "epoch": 1.4210526315789473,
+      "grad_norm": 1.2359084462091137,
+      "learning_rate": 1.1050983289171195e-05,
+      "loss": 1.3869,
+      "step": 1755
+    },
+    {
+      "epoch": 1.4218623481781376,
+      "grad_norm": 1.2213164143326687,
+      "learning_rate": 1.1042435175603439e-05,
+      "loss": 1.3499,
+      "step": 1756
+    },
+    {
+      "epoch": 1.4226720647773279,
+      "grad_norm": 1.2425021636630276,
+      "learning_rate": 1.1033886291888004e-05,
+      "loss": 1.3418,
+      "step": 1757
+    },
+    {
+      "epoch": 1.4234817813765182,
+      "grad_norm": 1.188641014563366,
+      "learning_rate": 1.102533664434077e-05,
+      "loss": 1.3177,
+      "step": 1758
+    },
+    {
+      "epoch": 1.4242914979757084,
+      "grad_norm": 1.1958395164768627,
+      "learning_rate": 1.1016786239278188e-05,
+      "loss": 1.3896,
+      "step": 1759
+    },
+    {
+      "epoch": 1.425101214574899,
+      "grad_norm": 1.1862097808375378,
+      "learning_rate": 1.1008235083017272e-05,
+      "loss": 1.3035,
+      "step": 1760
+    },
+    {
+      "epoch": 1.425910931174089,
+      "grad_norm": 1.2172514085079498,
+      "learning_rate": 1.0999683181875591e-05,
+      "loss": 1.3385,
+      "step": 1761
+    },
+    {
+      "epoch": 1.4267206477732794,
+      "grad_norm": 1.263795641893233,
+      "learning_rate": 1.0991130542171255e-05,
+      "loss": 1.3504,
+      "step": 1762
+    },
+    {
+      "epoch": 1.4275303643724697,
+      "grad_norm": 1.218807375034323,
+      "learning_rate": 1.0982577170222934e-05,
+      "loss": 1.3728,
+      "step": 1763
+    },
+    {
+      "epoch": 1.42834008097166,
+      "grad_norm": 1.2785005424913352,
+      "learning_rate": 1.0974023072349824e-05,
+      "loss": 1.3292,
+      "step": 1764
+    },
+    {
+      "epoch": 1.4291497975708503,
+      "grad_norm": 1.1955787893390333,
+      "learning_rate": 1.096546825487167e-05,
+      "loss": 1.4012,
+      "step": 1765
+    },
+    {
+      "epoch": 1.4299595141700405,
+      "grad_norm": 1.2472396886736037,
+      "learning_rate": 1.0956912724108737e-05,
+      "loss": 1.3221,
+      "step": 1766
+    },
+    {
+      "epoch": 1.4307692307692308,
+      "grad_norm": 1.2548879430356208,
+      "learning_rate": 1.0948356486381829e-05,
+      "loss": 1.3776,
+      "step": 1767
+    },
+    {
+      "epoch": 1.431578947368421,
+      "grad_norm": 1.242576432624912,
+      "learning_rate": 1.0939799548012262e-05,
+      "loss": 1.4025,
+      "step": 1768
+    },
+    {
+      "epoch": 1.4323886639676113,
+      "grad_norm": 1.2574563622805863,
+      "learning_rate": 1.0931241915321877e-05,
+      "loss": 1.412,
+      "step": 1769
+    },
+    {
+      "epoch": 1.4331983805668016,
+      "grad_norm": 1.1995040638978254,
+      "learning_rate": 1.092268359463302e-05,
+      "loss": 1.3246,
+      "step": 1770
+    },
+    {
+      "epoch": 1.4340080971659919,
+      "grad_norm": 1.2739773809734911,
+      "learning_rate": 1.0914124592268557e-05,
+      "loss": 1.3698,
+      "step": 1771
+    },
+    {
+      "epoch": 1.4348178137651821,
+      "grad_norm": 1.2199023602628671,
+      "learning_rate": 1.0905564914551847e-05,
+      "loss": 1.434,
+      "step": 1772
+    },
+    {
+      "epoch": 1.4356275303643724,
+      "grad_norm": 1.2464548777989193,
+      "learning_rate": 1.0897004567806754e-05,
+      "loss": 1.4201,
+      "step": 1773
+    },
+    {
+      "epoch": 1.4364372469635627,
+      "grad_norm": 1.214101570752936,
+      "learning_rate": 1.088844355835763e-05,
+      "loss": 1.3015,
+      "step": 1774
+    },
+    {
+      "epoch": 1.4372469635627532,
+      "grad_norm": 1.163110661848742,
+      "learning_rate": 1.0879881892529325e-05,
+      "loss": 1.3638,
+      "step": 1775
+    },
+    {
+      "epoch": 1.4380566801619432,
+      "grad_norm": 1.1936498971792766,
+      "learning_rate": 1.0871319576647166e-05,
+      "loss": 1.3742,
+      "step": 1776
+    },
+    {
+      "epoch": 1.4388663967611337,
+      "grad_norm": 1.2451768719667269,
+      "learning_rate": 1.0862756617036965e-05,
+      "loss": 1.4094,
+      "step": 1777
+    },
+    {
+      "epoch": 1.4396761133603238,
+      "grad_norm": 1.1478943380908433,
+      "learning_rate": 1.085419302002501e-05,
+      "loss": 1.3577,
+      "step": 1778
+    },
+    {
+      "epoch": 1.4404858299595142,
+      "grad_norm": 1.2145161220203882,
+      "learning_rate": 1.0845628791938058e-05,
+      "loss": 1.3611,
+      "step": 1779
+    },
+    {
+      "epoch": 1.4412955465587045,
+      "grad_norm": 1.23124963332477,
+      "learning_rate": 1.0837063939103332e-05,
+      "loss": 1.3967,
+      "step": 1780
+    },
+    {
+      "epoch": 1.4421052631578948,
+      "grad_norm": 1.1833240982834867,
+      "learning_rate": 1.0828498467848515e-05,
+      "loss": 1.381,
+      "step": 1781
+    },
+    {
+      "epoch": 1.442914979757085,
+      "grad_norm": 1.2492813677148087,
+      "learning_rate": 1.0819932384501755e-05,
+      "loss": 1.3991,
+      "step": 1782
+    },
+    {
+      "epoch": 1.4437246963562753,
+      "grad_norm": 1.340019879016165,
+      "learning_rate": 1.081136569539164e-05,
+      "loss": 1.3585,
+      "step": 1783
+    },
+    {
+      "epoch": 1.4445344129554656,
+      "grad_norm": 1.2206768859147779,
+      "learning_rate": 1.0802798406847213e-05,
+      "loss": 1.4541,
+      "step": 1784
+    },
+    {
+      "epoch": 1.4453441295546559,
+      "grad_norm": 1.1832342780679208,
+      "learning_rate": 1.0794230525197959e-05,
+      "loss": 1.3752,
+      "step": 1785
+    },
+    {
+      "epoch": 1.4461538461538461,
+      "grad_norm": 1.196053859850085,
+      "learning_rate": 1.0785662056773805e-05,
+      "loss": 1.327,
+      "step": 1786
+    },
+    {
+      "epoch": 1.4469635627530364,
+      "grad_norm": 1.211397922712786,
+      "learning_rate": 1.0777093007905102e-05,
+      "loss": 1.4449,
+      "step": 1787
+    },
+    {
+      "epoch": 1.4477732793522267,
+      "grad_norm": 1.2836918819568495,
+      "learning_rate": 1.0768523384922635e-05,
+      "loss": 1.342,
+      "step": 1788
+    },
+    {
+      "epoch": 1.448582995951417,
+      "grad_norm": 1.2272591940165982,
+      "learning_rate": 1.0759953194157617e-05,
+      "loss": 1.4064,
+      "step": 1789
+    },
+    {
+      "epoch": 1.4493927125506074,
+      "grad_norm": 1.2062079148051859,
+      "learning_rate": 1.0751382441941677e-05,
+      "loss": 1.3203,
+      "step": 1790
+    },
+    {
+      "epoch": 1.4502024291497975,
+      "grad_norm": 1.116389048563294,
+      "learning_rate": 1.0742811134606856e-05,
+      "loss": 1.3733,
+      "step": 1791
+    },
+    {
+      "epoch": 1.451012145748988,
+      "grad_norm": 1.25258756424204,
+      "learning_rate": 1.0734239278485608e-05,
+      "loss": 1.3817,
+      "step": 1792
+    },
+    {
+      "epoch": 1.451821862348178,
+      "grad_norm": 1.208392946627262,
+      "learning_rate": 1.0725666879910792e-05,
+      "loss": 1.3108,
+      "step": 1793
+    },
+    {
+      "epoch": 1.4526315789473685,
+      "grad_norm": 1.1948986892653393,
+      "learning_rate": 1.071709394521567e-05,
+      "loss": 1.4184,
+      "step": 1794
+    },
+    {
+      "epoch": 1.4534412955465588,
+      "grad_norm": 1.1408273455202025,
+      "learning_rate": 1.0708520480733895e-05,
+      "loss": 1.4017,
+      "step": 1795
+    },
+    {
+      "epoch": 1.454251012145749,
+      "grad_norm": 1.1858291893292587,
+      "learning_rate": 1.0699946492799515e-05,
+      "loss": 1.3898,
+      "step": 1796
+    },
+    {
+      "epoch": 1.4550607287449393,
+      "grad_norm": 1.1442203734732623,
+      "learning_rate": 1.0691371987746968e-05,
+      "loss": 1.3862,
+      "step": 1797
+    },
+    {
+      "epoch": 1.4558704453441296,
+      "grad_norm": 1.1796945509151606,
+      "learning_rate": 1.0682796971911067e-05,
+      "loss": 1.3721,
+      "step": 1798
+    },
+    {
+      "epoch": 1.4566801619433198,
+      "grad_norm": 1.2518379890147606,
+      "learning_rate": 1.0674221451627003e-05,
+      "loss": 1.4382,
+      "step": 1799
+    },
+    {
+      "epoch": 1.45748987854251,
+      "grad_norm": 1.1544556302617863,
+      "learning_rate": 1.0665645433230345e-05,
+      "loss": 1.3804,
+      "step": 1800
+    },
+    {
+      "epoch": 1.4582995951417004,
+      "grad_norm": 1.2433303275893108,
+      "learning_rate": 1.065706892305703e-05,
+      "loss": 1.4092,
+      "step": 1801
+    },
+    {
+      "epoch": 1.4591093117408906,
+      "grad_norm": 1.277806150716417,
+      "learning_rate": 1.0648491927443352e-05,
+      "loss": 1.3671,
+      "step": 1802
+    },
+    {
+      "epoch": 1.459919028340081,
+      "grad_norm": 1.2764604355710378,
+      "learning_rate": 1.0639914452725966e-05,
+      "loss": 1.3823,
+      "step": 1803
+    },
+    {
+      "epoch": 1.4607287449392712,
+      "grad_norm": 1.1879055573384147,
+      "learning_rate": 1.0631336505241885e-05,
+      "loss": 1.3828,
+      "step": 1804
+    },
+    {
+      "epoch": 1.4615384615384617,
+      "grad_norm": 1.2161612489980715,
+      "learning_rate": 1.0622758091328469e-05,
+      "loss": 1.3319,
+      "step": 1805
+    },
+    {
+      "epoch": 1.4623481781376517,
+      "grad_norm": 1.2261075654968276,
+      "learning_rate": 1.0614179217323418e-05,
+      "loss": 1.3419,
+      "step": 1806
+    },
+    {
+      "epoch": 1.4631578947368422,
+      "grad_norm": 1.2117166223283236,
+      "learning_rate": 1.0605599889564782e-05,
+      "loss": 1.3463,
+      "step": 1807
+    },
+    {
+      "epoch": 1.4639676113360323,
+      "grad_norm": 1.2031717859192184,
+      "learning_rate": 1.0597020114390932e-05,
+      "loss": 1.3635,
+      "step": 1808
+    },
+    {
+      "epoch": 1.4647773279352228,
+      "grad_norm": 1.2031388198613926,
+      "learning_rate": 1.0588439898140586e-05,
+      "loss": 1.3451,
+      "step": 1809
+    },
+    {
+      "epoch": 1.465587044534413,
+      "grad_norm": 1.3474438239487712,
+      "learning_rate": 1.0579859247152774e-05,
+      "loss": 1.4206,
+      "step": 1810
+    },
+    {
+      "epoch": 1.4663967611336033,
+      "grad_norm": 1.2444786342836838,
+      "learning_rate": 1.0571278167766857e-05,
+      "loss": 1.3636,
+      "step": 1811
+    },
+    {
+      "epoch": 1.4672064777327936,
+      "grad_norm": 1.219013237010411,
+      "learning_rate": 1.0562696666322502e-05,
+      "loss": 1.3193,
+      "step": 1812
+    },
+    {
+      "epoch": 1.4680161943319838,
+      "grad_norm": 1.1690703137014555,
+      "learning_rate": 1.05541147491597e-05,
+      "loss": 1.315,
+      "step": 1813
+    },
+    {
+      "epoch": 1.468825910931174,
+      "grad_norm": 1.2063145766276413,
+      "learning_rate": 1.0545532422618742e-05,
+      "loss": 1.3397,
+      "step": 1814
+    },
+    {
+      "epoch": 1.4696356275303644,
+      "grad_norm": 1.2003457207424921,
+      "learning_rate": 1.0536949693040224e-05,
+      "loss": 1.3279,
+      "step": 1815
+    },
+    {
+      "epoch": 1.4704453441295546,
+      "grad_norm": 1.2020614644434036,
+      "learning_rate": 1.0528366566765032e-05,
+      "loss": 1.408,
+      "step": 1816
+    },
+    {
+      "epoch": 1.471255060728745,
+      "grad_norm": 1.281360311469663,
+      "learning_rate": 1.0519783050134358e-05,
+      "loss": 1.3833,
+      "step": 1817
+    },
+    {
+      "epoch": 1.4720647773279352,
+      "grad_norm": 1.2169999342183757,
+      "learning_rate": 1.0511199149489673e-05,
+      "loss": 1.3173,
+      "step": 1818
+    },
+    {
+      "epoch": 1.4728744939271254,
+      "grad_norm": 1.1845396830671333,
+      "learning_rate": 1.0502614871172736e-05,
+      "loss": 1.3344,
+      "step": 1819
+    },
+    {
+      "epoch": 1.4736842105263157,
+      "grad_norm": 1.2867321451496658,
+      "learning_rate": 1.0494030221525582e-05,
+      "loss": 1.4286,
+      "step": 1820
+    },
+    {
+      "epoch": 1.474493927125506,
+      "grad_norm": 1.2064628750455095,
+      "learning_rate": 1.0485445206890522e-05,
+      "loss": 1.3356,
+      "step": 1821
+    },
+    {
+      "epoch": 1.4753036437246965,
+      "grad_norm": 1.228960560736752,
+      "learning_rate": 1.0476859833610142e-05,
+      "loss": 1.4522,
+      "step": 1822
+    },
+    {
+      "epoch": 1.4761133603238865,
+      "grad_norm": 1.2058195619871348,
+      "learning_rate": 1.046827410802728e-05,
+      "loss": 1.3681,
+      "step": 1823
+    },
+    {
+      "epoch": 1.476923076923077,
+      "grad_norm": 1.1852192696221284,
+      "learning_rate": 1.0459688036485044e-05,
+      "loss": 1.3427,
+      "step": 1824
+    },
+    {
+      "epoch": 1.4777327935222673,
+      "grad_norm": 1.20736083607171,
+      "learning_rate": 1.0451101625326798e-05,
+      "loss": 1.4147,
+      "step": 1825
+    },
+    {
+      "epoch": 1.4785425101214575,
+      "grad_norm": 1.2516509317308389,
+      "learning_rate": 1.0442514880896156e-05,
+      "loss": 1.3319,
+      "step": 1826
+    },
+    {
+      "epoch": 1.4793522267206478,
+      "grad_norm": 1.2212025089204959,
+      "learning_rate": 1.043392780953697e-05,
+      "loss": 1.34,
+      "step": 1827
+    },
+    {
+      "epoch": 1.480161943319838,
+      "grad_norm": 1.241429196352147,
+      "learning_rate": 1.0425340417593341e-05,
+      "loss": 1.3743,
+      "step": 1828
+    },
+    {
+      "epoch": 1.4809716599190283,
+      "grad_norm": 1.18843326826917,
+      "learning_rate": 1.0416752711409612e-05,
+      "loss": 1.3653,
+      "step": 1829
+    },
+    {
+      "epoch": 1.4817813765182186,
+      "grad_norm": 1.218215122004799,
+      "learning_rate": 1.0408164697330348e-05,
+      "loss": 1.3632,
+      "step": 1830
+    },
+    {
+      "epoch": 1.4825910931174089,
+      "grad_norm": 1.203207018787504,
+      "learning_rate": 1.0399576381700346e-05,
+      "loss": 1.3731,
+      "step": 1831
+    },
+    {
+      "epoch": 1.4834008097165992,
+      "grad_norm": 1.2359647155470024,
+      "learning_rate": 1.0390987770864623e-05,
+      "loss": 1.363,
+      "step": 1832
+    },
+    {
+      "epoch": 1.4842105263157894,
+      "grad_norm": 1.211501248267098,
+      "learning_rate": 1.0382398871168421e-05,
+      "loss": 1.3222,
+      "step": 1833
+    },
+    {
+      "epoch": 1.4850202429149797,
+      "grad_norm": 1.1936705020590501,
+      "learning_rate": 1.0373809688957192e-05,
+      "loss": 1.3877,
+      "step": 1834
+    },
+    {
+      "epoch": 1.48582995951417,
+      "grad_norm": 1.1937889310278138,
+      "learning_rate": 1.0365220230576592e-05,
+      "loss": 1.4284,
+      "step": 1835
+    },
+    {
+      "epoch": 1.4866396761133602,
+      "grad_norm": 1.1808597278513449,
+      "learning_rate": 1.035663050237248e-05,
+      "loss": 1.3622,
+      "step": 1836
+    },
+    {
+      "epoch": 1.4874493927125507,
+      "grad_norm": 1.1900999195170403,
+      "learning_rate": 1.0348040510690929e-05,
+      "loss": 1.4129,
+      "step": 1837
+    },
+    {
+      "epoch": 1.4882591093117408,
+      "grad_norm": 1.2368390047722708,
+      "learning_rate": 1.033945026187819e-05,
+      "loss": 1.3577,
+      "step": 1838
+    },
+    {
+      "epoch": 1.4890688259109313,
+      "grad_norm": 1.250086709756194,
+      "learning_rate": 1.0330859762280712e-05,
+      "loss": 1.3997,
+      "step": 1839
+    },
+    {
+      "epoch": 1.4898785425101215,
+      "grad_norm": 1.2472030369660945,
+      "learning_rate": 1.0322269018245128e-05,
+      "loss": 1.3501,
+      "step": 1840
+    },
+    {
+      "epoch": 1.4906882591093118,
+      "grad_norm": 1.2452205777512158,
+      "learning_rate": 1.0313678036118253e-05,
+      "loss": 1.3399,
+      "step": 1841
+    },
+    {
+      "epoch": 1.491497975708502,
+      "grad_norm": 1.2034361473861883,
+      "learning_rate": 1.0305086822247077e-05,
+      "loss": 1.3746,
+      "step": 1842
+    },
+    {
+      "epoch": 1.4923076923076923,
+      "grad_norm": 1.2952780810330986,
+      "learning_rate": 1.0296495382978756e-05,
+      "loss": 1.3704,
+      "step": 1843
+    },
+    {
+      "epoch": 1.4931174089068826,
+      "grad_norm": 1.2614416983159662,
+      "learning_rate": 1.0287903724660617e-05,
+      "loss": 1.3932,
+      "step": 1844
+    },
+    {
+      "epoch": 1.4939271255060729,
+      "grad_norm": 1.207803764074404,
+      "learning_rate": 1.0279311853640157e-05,
+      "loss": 1.3694,
+      "step": 1845
+    },
+    {
+      "epoch": 1.4947368421052631,
+      "grad_norm": 1.181956708022318,
+      "learning_rate": 1.0270719776265017e-05,
+      "loss": 1.3917,
+      "step": 1846
+    },
+    {
+      "epoch": 1.4955465587044534,
+      "grad_norm": 1.1444875887339165,
+      "learning_rate": 1.0262127498882992e-05,
+      "loss": 1.3259,
+      "step": 1847
+    },
+    {
+      "epoch": 1.4963562753036437,
+      "grad_norm": 1.2022791631767018,
+      "learning_rate": 1.0253535027842032e-05,
+      "loss": 1.3427,
+      "step": 1848
+    },
+    {
+      "epoch": 1.497165991902834,
+      "grad_norm": 1.2070972768865362,
+      "learning_rate": 1.024494236949023e-05,
+      "loss": 1.3986,
+      "step": 1849
+    },
+    {
+      "epoch": 1.4979757085020242,
+      "grad_norm": 1.2147753961877357,
+      "learning_rate": 1.0236349530175807e-05,
+      "loss": 1.3675,
+      "step": 1850
+    },
+    {
+      "epoch": 1.4987854251012145,
+      "grad_norm": 1.1591467128751487,
+      "learning_rate": 1.0227756516247127e-05,
+      "loss": 1.3841,
+      "step": 1851
+    },
+    {
+      "epoch": 1.499595141700405,
+      "grad_norm": 1.2069756938729441,
+      "learning_rate": 1.0219163334052682e-05,
+      "loss": 1.3365,
+      "step": 1852
+    },
+    {
+      "epoch": 1.500404858299595,
+      "grad_norm": 1.2440005880081049,
+      "learning_rate": 1.0210569989941085e-05,
+      "loss": 1.357,
+      "step": 1853
+    },
+    {
+      "epoch": 1.5012145748987855,
+      "grad_norm": 1.2000119522275203,
+      "learning_rate": 1.020197649026107e-05,
+      "loss": 1.3938,
+      "step": 1854
+    },
+    {
+      "epoch": 1.5020242914979756,
+      "grad_norm": 1.187811020715873,
+      "learning_rate": 1.019338284136149e-05,
+      "loss": 1.3288,
+      "step": 1855
+    },
+    {
+      "epoch": 1.502834008097166,
+      "grad_norm": 1.1684805914260927,
+      "learning_rate": 1.01847890495913e-05,
+      "loss": 1.3791,
+      "step": 1856
+    },
+    {
+      "epoch": 1.5036437246963563,
+      "grad_norm": 1.2362993183012971,
+      "learning_rate": 1.0176195121299567e-05,
+      "loss": 1.394,
+      "step": 1857
+    },
+    {
+      "epoch": 1.5044534412955466,
+      "grad_norm": 1.2284232654390486,
+      "learning_rate": 1.0167601062835459e-05,
+      "loss": 1.3728,
+      "step": 1858
+    },
+    {
+      "epoch": 1.5052631578947369,
+      "grad_norm": 1.2440453874825037,
+      "learning_rate": 1.0159006880548237e-05,
+      "loss": 1.3519,
+      "step": 1859
+    },
+    {
+      "epoch": 1.5060728744939271,
+      "grad_norm": 1.14460238709053,
+      "learning_rate": 1.015041258078725e-05,
+      "loss": 1.365,
+      "step": 1860
+    },
+    {
+      "epoch": 1.5068825910931174,
+      "grad_norm": 1.1391435364818712,
+      "learning_rate": 1.0141818169901945e-05,
+      "loss": 1.3191,
+      "step": 1861
+    },
+    {
+      "epoch": 1.5076923076923077,
+      "grad_norm": 1.1847183299604456,
+      "learning_rate": 1.013322365424184e-05,
+      "loss": 1.4322,
+      "step": 1862
+    },
+    {
+      "epoch": 1.508502024291498,
+      "grad_norm": 1.2242479719103367,
+      "learning_rate": 1.012462904015654e-05,
+      "loss": 1.3728,
+      "step": 1863
+    },
+    {
+      "epoch": 1.5093117408906882,
+      "grad_norm": 1.2141368241612376,
+      "learning_rate": 1.011603433399571e-05,
+      "loss": 1.4276,
+      "step": 1864
+    },
+    {
+      "epoch": 1.5101214574898787,
+      "grad_norm": 1.197406422654741,
+      "learning_rate": 1.0107439542109097e-05,
+      "loss": 1.387,
+      "step": 1865
+    },
+    {
+      "epoch": 1.5109311740890687,
+      "grad_norm": 1.141149626390884,
+      "learning_rate": 1.0098844670846504e-05,
+      "loss": 1.3297,
+      "step": 1866
+    },
+    {
+      "epoch": 1.5117408906882592,
+      "grad_norm": 1.2001491455611746,
+      "learning_rate": 1.0090249726557795e-05,
+      "loss": 1.2897,
+      "step": 1867
+    },
+    {
+      "epoch": 1.5125506072874493,
+      "grad_norm": 1.1667785806675015,
+      "learning_rate": 1.0081654715592881e-05,
+      "loss": 1.3459,
+      "step": 1868
+    },
+    {
+      "epoch": 1.5133603238866398,
+      "grad_norm": 1.1839167724233615,
+      "learning_rate": 1.007305964430173e-05,
+      "loss": 1.3219,
+      "step": 1869
+    },
+    {
+      "epoch": 1.5141700404858298,
+      "grad_norm": 1.2144434763318714,
+      "learning_rate": 1.0064464519034358e-05,
+      "loss": 1.3816,
+      "step": 1870
+    },
+    {
+      "epoch": 1.5149797570850203,
+      "grad_norm": 1.2135857961398115,
+      "learning_rate": 1.005586934614081e-05,
+      "loss": 1.4097,
+      "step": 1871
+    },
+    {
+      "epoch": 1.5157894736842106,
+      "grad_norm": 1.2095532800796946,
+      "learning_rate": 1.004727413197117e-05,
+      "loss": 1.332,
+      "step": 1872
+    },
+    {
+      "epoch": 1.5165991902834008,
+      "grad_norm": 1.2181508610257055,
+      "learning_rate": 1.0038678882875557e-05,
+      "loss": 1.3892,
+      "step": 1873
+    },
+    {
+      "epoch": 1.5174089068825911,
+      "grad_norm": 1.2026433653521167,
+      "learning_rate": 1.0030083605204115e-05,
+      "loss": 1.4083,
+      "step": 1874
+    },
+    {
+      "epoch": 1.5182186234817814,
+      "grad_norm": 1.1773517841295498,
+      "learning_rate": 1.0021488305307003e-05,
+      "loss": 1.3584,
+      "step": 1875
+    },
+    {
+      "epoch": 1.5190283400809717,
+      "grad_norm": 1.231433388400886,
+      "learning_rate": 1.00128929895344e-05,
+      "loss": 1.374,
+      "step": 1876
+    },
+    {
+      "epoch": 1.519838056680162,
+      "grad_norm": 1.2332502916196273,
+      "learning_rate": 1.0004297664236502e-05,
+      "loss": 1.4351,
+      "step": 1877
+    },
+    {
+      "epoch": 1.5206477732793522,
+      "grad_norm": 1.213292815409757,
+      "learning_rate": 9.9957023357635e-06,
+      "loss": 1.2996,
+      "step": 1878
+    },
+    {
+      "epoch": 1.5214574898785425,
+      "grad_norm": 1.2346861234405477,
+      "learning_rate": 9.9871070104656e-06,
+      "loss": 1.3471,
+      "step": 1879
+    },
+    {
+      "epoch": 1.522267206477733,
+      "grad_norm": 1.254827265085309,
+      "learning_rate": 9.978511694692999e-06,
+      "loss": 1.3533,
+      "step": 1880
+    },
+    {
+      "epoch": 1.523076923076923,
+      "grad_norm": 1.1917367114901438,
+      "learning_rate": 9.969916394795888e-06,
+      "loss": 1.3383,
+      "step": 1881
+    },
+    {
+      "epoch": 1.5238866396761135,
+      "grad_norm": 1.180916076508306,
+      "learning_rate": 9.961321117124444e-06,
+      "loss": 1.3722,
+      "step": 1882
+    },
+    {
+      "epoch": 1.5246963562753035,
+      "grad_norm": 1.2010789539898061,
+      "learning_rate": 9.952725868028831e-06,
+      "loss": 1.342,
+      "step": 1883
+    },
+    {
+      "epoch": 1.525506072874494,
+      "grad_norm": 1.2029836828056097,
+      "learning_rate": 9.944130653859195e-06,
+      "loss": 1.3714,
+      "step": 1884
+    },
+    {
+      "epoch": 1.526315789473684,
+      "grad_norm": 1.230235446023344,
+      "learning_rate": 9.935535480965647e-06,
+      "loss": 1.419,
+      "step": 1885
+    },
+    {
+      "epoch": 1.5271255060728746,
+      "grad_norm": 1.2304103134503401,
+      "learning_rate": 9.92694035569827e-06,
+      "loss": 1.3459,
+      "step": 1886
+    },
+    {
+      "epoch": 1.5279352226720648,
+      "grad_norm": 1.2003919942780323,
+      "learning_rate": 9.918345284407122e-06,
+      "loss": 1.4122,
+      "step": 1887
+    },
+    {
+      "epoch": 1.528744939271255,
+      "grad_norm": 1.2639140447939796,
+      "learning_rate": 9.909750273442208e-06,
+      "loss": 1.3368,
+      "step": 1888
+    },
+    {
+      "epoch": 1.5295546558704454,
+      "grad_norm": 1.2269119630065992,
+      "learning_rate": 9.901155329153498e-06,
+      "loss": 1.3554,
+      "step": 1889
+    },
+    {
+      "epoch": 1.5303643724696356,
+      "grad_norm": 1.2657559146957076,
+      "learning_rate": 9.892560457890907e-06,
+      "loss": 1.4207,
+      "step": 1890
+    },
+    {
+      "epoch": 1.531174089068826,
+      "grad_norm": 1.1993508806620081,
+      "learning_rate": 9.883965666004293e-06,
+      "loss": 1.3638,
+      "step": 1891
+    },
+    {
+      "epoch": 1.5319838056680162,
+      "grad_norm": 1.219522089298675,
+      "learning_rate": 9.875370959843465e-06,
+      "loss": 1.3768,
+      "step": 1892
+    },
+    {
+      "epoch": 1.5327935222672064,
+      "grad_norm": 1.2178201278785572,
+      "learning_rate": 9.866776345758166e-06,
+      "loss": 1.408,
+      "step": 1893
+    },
+    {
+      "epoch": 1.5336032388663967,
+      "grad_norm": 1.1451898153111888,
+      "learning_rate": 9.858181830098058e-06,
+      "loss": 1.3561,
+      "step": 1894
+    },
+    {
+      "epoch": 1.5344129554655872,
+      "grad_norm": 1.2469841070465375,
+      "learning_rate": 9.849587419212751e-06,
+      "loss": 1.4224,
+      "step": 1895
+    },
+    {
+      "epoch": 1.5352226720647772,
+      "grad_norm": 1.1845691507885716,
+      "learning_rate": 9.840993119451768e-06,
+      "loss": 1.4053,
+      "step": 1896
+    },
+    {
+      "epoch": 1.5360323886639677,
+      "grad_norm": 1.1330718380072349,
+      "learning_rate": 9.832398937164545e-06,
+      "loss": 1.3581,
+      "step": 1897
+    },
+    {
+      "epoch": 1.5368421052631578,
+      "grad_norm": 1.1691776979372561,
+      "learning_rate": 9.823804878700434e-06,
+      "loss": 1.3208,
+      "step": 1898
+    },
+    {
+      "epoch": 1.5376518218623483,
+      "grad_norm": 1.1868100114985074,
+      "learning_rate": 9.815210950408703e-06,
+      "loss": 1.351,
+      "step": 1899
+    },
+    {
+      "epoch": 1.5384615384615383,
+      "grad_norm": 1.1980676569924105,
+      "learning_rate": 9.806617158638515e-06,
+      "loss": 1.3409,
+      "step": 1900
+    },
+    {
+      "epoch": 1.5392712550607288,
+      "grad_norm": 1.2286369763296952,
+      "learning_rate": 9.798023509738932e-06,
+      "loss": 1.3736,
+      "step": 1901
+    },
+    {
+      "epoch": 1.5400809716599189,
+      "grad_norm": 1.230286054787407,
+      "learning_rate": 9.789430010058918e-06,
+      "loss": 1.3594,
+      "step": 1902
+    },
+    {
+      "epoch": 1.5408906882591094,
+      "grad_norm": 1.1833188959041652,
+      "learning_rate": 9.78083666594732e-06,
+      "loss": 1.3609,
+      "step": 1903
+    },
+    {
+      "epoch": 1.5417004048582996,
+      "grad_norm": 1.225456208787112,
+      "learning_rate": 9.772243483752876e-06,
+      "loss": 1.3461,
+      "step": 1904
+    },
+    {
+      "epoch": 1.54251012145749,
+      "grad_norm": 1.1594495050829507,
+      "learning_rate": 9.763650469824198e-06,
+      "loss": 1.3852,
+      "step": 1905
+    },
+    {
+      "epoch": 1.5433198380566802,
+      "grad_norm": 1.2470244297864663,
+      "learning_rate": 9.755057630509774e-06,
+      "loss": 1.3146,
+      "step": 1906
+    },
+    {
+      "epoch": 1.5441295546558704,
+      "grad_norm": 1.2036173468340592,
+      "learning_rate": 9.746464972157971e-06,
+      "loss": 1.3576,
+      "step": 1907
+    },
+    {
+      "epoch": 1.5449392712550607,
+      "grad_norm": 1.235082778436893,
+      "learning_rate": 9.737872501117013e-06,
+      "loss": 1.3347,
+      "step": 1908
+    },
+    {
+      "epoch": 1.545748987854251,
+      "grad_norm": 1.247100018489477,
+      "learning_rate": 9.729280223734988e-06,
+      "loss": 1.4114,
+      "step": 1909
+    },
+    {
+      "epoch": 1.5465587044534415,
+      "grad_norm": 1.1457263494567647,
+      "learning_rate": 9.720688146359843e-06,
+      "loss": 1.386,
+      "step": 1910
+    },
+    {
+      "epoch": 1.5473684210526315,
+      "grad_norm": 1.1729296197232764,
+      "learning_rate": 9.712096275339381e-06,
+      "loss": 1.3698,
+      "step": 1911
+    },
+    {
+      "epoch": 1.548178137651822,
+      "grad_norm": 1.1426028010449,
+      "learning_rate": 9.703504617021247e-06,
+      "loss": 1.286,
+      "step": 1912
+    },
+    {
+      "epoch": 1.548987854251012,
+      "grad_norm": 1.2004758472296264,
+      "learning_rate": 9.694913177752927e-06,
+      "loss": 1.4253,
+      "step": 1913
+    },
+    {
+      "epoch": 1.5497975708502025,
+      "grad_norm": 1.2117453405073084,
+      "learning_rate": 9.68632196388175e-06,
+      "loss": 1.2984,
+      "step": 1914
+    },
+    {
+      "epoch": 1.5506072874493926,
+      "grad_norm": 1.2033634956141823,
+      "learning_rate": 9.677730981754875e-06,
+      "loss": 1.2999,
+      "step": 1915
+    },
+    {
+      "epoch": 1.551417004048583,
+      "grad_norm": 1.2564126564820501,
+      "learning_rate": 9.669140237719292e-06,
+      "loss": 1.3905,
+      "step": 1916
+    },
+    {
+      "epoch": 1.5522267206477731,
+      "grad_norm": 1.2062084382081446,
+      "learning_rate": 9.660549738121814e-06,
+      "loss": 1.3715,
+      "step": 1917
+    },
+    {
+      "epoch": 1.5530364372469636,
+      "grad_norm": 1.2829351534612063,
+      "learning_rate": 9.651959489309073e-06,
+      "loss": 1.3494,
+      "step": 1918
+    },
+    {
+      "epoch": 1.5538461538461539,
+      "grad_norm": 1.2383174742670793,
+      "learning_rate": 9.643369497627521e-06,
+      "loss": 1.4225,
+      "step": 1919
+    },
+    {
+      "epoch": 1.5546558704453441,
+      "grad_norm": 1.222803913479318,
+      "learning_rate": 9.634779769423412e-06,
+      "loss": 1.3943,
+      "step": 1920
+    },
+    {
+      "epoch": 1.5554655870445344,
+      "grad_norm": 1.193502143064031,
+      "learning_rate": 9.62619031104281e-06,
+      "loss": 1.3762,
+      "step": 1921
+    },
+    {
+      "epoch": 1.5562753036437247,
+      "grad_norm": 1.1907724889410738,
+      "learning_rate": 9.61760112883158e-06,
+      "loss": 1.3944,
+      "step": 1922
+    },
+    {
+      "epoch": 1.557085020242915,
+      "grad_norm": 1.198345852396561,
+      "learning_rate": 9.609012229135379e-06,
+      "loss": 1.3854,
+      "step": 1923
+    },
+    {
+      "epoch": 1.5578947368421052,
+      "grad_norm": 1.2182523045236915,
+      "learning_rate": 9.600423618299659e-06,
+      "loss": 1.367,
+      "step": 1924
+    },
+    {
+      "epoch": 1.5587044534412957,
+      "grad_norm": 1.1805065924488671,
+      "learning_rate": 9.591835302669657e-06,
+      "loss": 1.3546,
+      "step": 1925
+    },
+    {
+      "epoch": 1.5595141700404858,
+      "grad_norm": 1.2249464401299464,
+      "learning_rate": 9.58324728859039e-06,
+      "loss": 1.3609,
+      "step": 1926
+    },
+    {
+      "epoch": 1.5603238866396762,
+      "grad_norm": 1.2243122967938407,
+      "learning_rate": 9.57465958240666e-06,
+      "loss": 1.3999,
+      "step": 1927
+    },
+    {
+      "epoch": 1.5611336032388663,
+      "grad_norm": 1.2274491912534922,
+      "learning_rate": 9.566072190463032e-06,
+      "loss": 1.306,
+      "step": 1928
+    },
+    {
+      "epoch": 1.5619433198380568,
+      "grad_norm": 1.2488797733856527,
+      "learning_rate": 9.557485119103849e-06,
+      "loss": 1.3739,
+      "step": 1929
+    },
+    {
+      "epoch": 1.5627530364372468,
+      "grad_norm": 1.2239459163300153,
+      "learning_rate": 9.548898374673205e-06,
+      "loss": 1.3517,
+      "step": 1930
+    },
+    {
+      "epoch": 1.5635627530364373,
+      "grad_norm": 1.2178846415184048,
+      "learning_rate": 9.540311963514957e-06,
+      "loss": 1.3731,
+      "step": 1931
+    },
+    {
+      "epoch": 1.5643724696356274,
+      "grad_norm": 1.265558847290855,
+      "learning_rate": 9.531725891972725e-06,
+      "loss": 1.4236,
+      "step": 1932
+    },
+    {
+      "epoch": 1.5651821862348179,
+      "grad_norm": 1.1681836474926013,
+      "learning_rate": 9.523140166389864e-06,
+      "loss": 1.4025,
+      "step": 1933
+    },
+    {
+      "epoch": 1.5659919028340081,
+      "grad_norm": 1.2040234728629644,
+      "learning_rate": 9.514554793109477e-06,
+      "loss": 1.4339,
+      "step": 1934
+    },
+    {
+      "epoch": 1.5668016194331984,
+      "grad_norm": 1.265710944730449,
+      "learning_rate": 9.505969778474418e-06,
+      "loss": 1.3561,
+      "step": 1935
+    },
+    {
+      "epoch": 1.5676113360323887,
+      "grad_norm": 1.1846379371827758,
+      "learning_rate": 9.497385128827266e-06,
+      "loss": 1.3652,
+      "step": 1936
+    },
+    {
+      "epoch": 1.568421052631579,
+      "grad_norm": 1.174180232851037,
+      "learning_rate": 9.48880085051033e-06,
+      "loss": 1.2807,
+      "step": 1937
+    },
+    {
+      "epoch": 1.5692307692307692,
+      "grad_norm": 1.242422752714555,
+      "learning_rate": 9.480216949865644e-06,
+      "loss": 1.375,
+      "step": 1938
+    },
+    {
+      "epoch": 1.5700404858299595,
+      "grad_norm": 1.239409495392598,
+      "learning_rate": 9.471633433234972e-06,
+      "loss": 1.3435,
+      "step": 1939
+    },
+    {
+      "epoch": 1.5708502024291497,
+      "grad_norm": 1.2564868657700816,
+      "learning_rate": 9.463050306959782e-06,
+      "loss": 1.3503,
+      "step": 1940
+    },
+    {
+      "epoch": 1.57165991902834,
+      "grad_norm": 1.2398521277680479,
+      "learning_rate": 9.454467577381263e-06,
+      "loss": 1.3266,
+      "step": 1941
+    },
+    {
+      "epoch": 1.5724696356275305,
+      "grad_norm": 1.2135937983968192,
+      "learning_rate": 9.445885250840301e-06,
+      "loss": 1.3269,
+      "step": 1942
+    },
+    {
+      "epoch": 1.5732793522267206,
+      "grad_norm": 1.1895520607669423,
+      "learning_rate": 9.4373033336775e-06,
+      "loss": 1.3121,
+      "step": 1943
+    },
+    {
+      "epoch": 1.574089068825911,
+      "grad_norm": 1.1824970891733284,
+      "learning_rate": 9.428721832233148e-06,
+      "loss": 1.3569,
+      "step": 1944
+    },
+    {
+      "epoch": 1.574898785425101,
+      "grad_norm": 1.2023256894687824,
+      "learning_rate": 9.42014075284723e-06,
+      "loss": 1.4145,
+      "step": 1945
+    },
+    {
+      "epoch": 1.5757085020242916,
+      "grad_norm": 1.18100975780351,
+      "learning_rate": 9.411560101859417e-06,
+      "loss": 1.3616,
+      "step": 1946
+    },
+    {
+      "epoch": 1.5765182186234816,
+      "grad_norm": 1.2357755167800077,
+      "learning_rate": 9.402979885609071e-06,
+      "loss": 1.3371,
+      "step": 1947
+    },
+    {
+      "epoch": 1.5773279352226721,
+      "grad_norm": 1.2279972155267704,
+      "learning_rate": 9.394400110435225e-06,
+      "loss": 1.3994,
+      "step": 1948
+    },
+    {
+      "epoch": 1.5781376518218624,
+      "grad_norm": 1.1914709750868986,
+      "learning_rate": 9.385820782676584e-06,
+      "loss": 1.3488,
+      "step": 1949
+    },
+    {
+      "epoch": 1.5789473684210527,
+      "grad_norm": 1.1994067475194277,
+      "learning_rate": 9.377241908671533e-06,
+      "loss": 1.3631,
+      "step": 1950
+    },
+    {
+      "epoch": 1.579757085020243,
+      "grad_norm": 1.1718520198213915,
+      "learning_rate": 9.368663494758115e-06,
+      "loss": 1.3229,
+      "step": 1951
+    },
+    {
+      "epoch": 1.5805668016194332,
+      "grad_norm": 1.2111814984055784,
+      "learning_rate": 9.360085547274036e-06,
+      "loss": 1.3721,
+      "step": 1952
+    },
+    {
+      "epoch": 1.5813765182186235,
+      "grad_norm": 1.2166854804513723,
+      "learning_rate": 9.351508072556651e-06,
+      "loss": 1.3382,
+      "step": 1953
+    },
+    {
+      "epoch": 1.5821862348178137,
+      "grad_norm": 1.2328123213604463,
+      "learning_rate": 9.342931076942973e-06,
+      "loss": 1.4074,
+      "step": 1954
+    },
+    {
+      "epoch": 1.582995951417004,
+      "grad_norm": 1.2394612442811164,
+      "learning_rate": 9.334354566769658e-06,
+      "loss": 1.3018,
+      "step": 1955
+    },
+    {
+      "epoch": 1.5838056680161943,
+      "grad_norm": 1.2112567483732097,
+      "learning_rate": 9.325778548373e-06,
+      "loss": 1.2831,
+      "step": 1956
+    },
+    {
+      "epoch": 1.5846153846153848,
+      "grad_norm": 1.1733648170105058,
+      "learning_rate": 9.317203028088938e-06,
+      "loss": 1.3761,
+      "step": 1957
+    },
+    {
+      "epoch": 1.5854251012145748,
+      "grad_norm": 1.2203734468272454,
+      "learning_rate": 9.308628012253032e-06,
+      "loss": 1.3634,
+      "step": 1958
+    },
+    {
+      "epoch": 1.5862348178137653,
+      "grad_norm": 1.189447117782888,
+      "learning_rate": 9.300053507200487e-06,
+      "loss": 1.4065,
+      "step": 1959
+    },
+    {
+      "epoch": 1.5870445344129553,
+      "grad_norm": 1.2206492902377073,
+      "learning_rate": 9.291479519266108e-06,
+      "loss": 1.3927,
+      "step": 1960
+    },
+    {
+      "epoch": 1.5878542510121458,
+      "grad_norm": 1.1783321735346974,
+      "learning_rate": 9.282906054784333e-06,
+      "loss": 1.3714,
+      "step": 1961
+    },
+    {
+      "epoch": 1.5886639676113359,
+      "grad_norm": 1.165535947374695,
+      "learning_rate": 9.274333120089211e-06,
+      "loss": 1.3888,
+      "step": 1962
+    },
+    {
+      "epoch": 1.5894736842105264,
+      "grad_norm": 1.2497678418850922,
+      "learning_rate": 9.265760721514397e-06,
+      "loss": 1.3666,
+      "step": 1963
+    },
+    {
+      "epoch": 1.5902834008097166,
+      "grad_norm": 1.2311086453584528,
+      "learning_rate": 9.257188865393148e-06,
+      "loss": 1.411,
+      "step": 1964
+    },
+    {
+      "epoch": 1.591093117408907,
+      "grad_norm": 1.1853289463831915,
+      "learning_rate": 9.248617558058328e-06,
+      "loss": 1.3884,
+      "step": 1965
+    },
+    {
+      "epoch": 1.5919028340080972,
+      "grad_norm": 1.2149617440803817,
+      "learning_rate": 9.240046805842383e-06,
+      "loss": 1.3175,
+      "step": 1966
+    },
+    {
+      "epoch": 1.5927125506072874,
+      "grad_norm": 1.2303091959416708,
+      "learning_rate": 9.231476615077366e-06,
+      "loss": 1.3443,
+      "step": 1967
+    },
+    {
+      "epoch": 1.5935222672064777,
+      "grad_norm": 1.2403372778652388,
+      "learning_rate": 9.2229069920949e-06,
+      "loss": 1.3279,
+      "step": 1968
+    },
+    {
+      "epoch": 1.594331983805668,
+      "grad_norm": 1.2729706893834039,
+      "learning_rate": 9.214337943226199e-06,
+      "loss": 1.4091,
+      "step": 1969
+    },
+    {
+      "epoch": 1.5951417004048583,
+      "grad_norm": 1.1994129479779139,
+      "learning_rate": 9.205769474802045e-06,
+      "loss": 1.3433,
+      "step": 1970
+    },
+    {
+      "epoch": 1.5959514170040485,
+      "grad_norm": 1.2304449200153609,
+      "learning_rate": 9.19720159315279e-06,
+      "loss": 1.2943,
+      "step": 1971
+    },
+    {
+      "epoch": 1.596761133603239,
+      "grad_norm": 1.2186847444701223,
+      "learning_rate": 9.188634304608366e-06,
+      "loss": 1.3831,
+      "step": 1972
+    },
+    {
+      "epoch": 1.597570850202429,
+      "grad_norm": 1.2840247109912253,
+      "learning_rate": 9.180067615498251e-06,
+      "loss": 1.3342,
+      "step": 1973
+    },
+    {
+      "epoch": 1.5983805668016196,
+      "grad_norm": 1.183188023738025,
+      "learning_rate": 9.171501532151486e-06,
+      "loss": 1.3873,
+      "step": 1974
+    },
+    {
+      "epoch": 1.5991902834008096,
+      "grad_norm": 1.1776750182121347,
+      "learning_rate": 9.162936060896672e-06,
+      "loss": 1.3365,
+      "step": 1975
+    },
+    {
+      "epoch": 1.6,
+      "grad_norm": 1.328668412553802,
+      "learning_rate": 9.154371208061943e-06,
+      "loss": 1.3847,
+      "step": 1976
+    },
+    {
+      "epoch": 1.6008097165991901,
+      "grad_norm": 1.2168461558396328,
+      "learning_rate": 9.145806979974991e-06,
+      "loss": 1.3555,
+      "step": 1977
+    },
+    {
+      "epoch": 1.6016194331983806,
+      "grad_norm": 1.1622150592681906,
+      "learning_rate": 9.137243382963039e-06,
+      "loss": 1.3405,
+      "step": 1978
+    },
+    {
+      "epoch": 1.602429149797571,
+      "grad_norm": 1.2436189606158277,
+      "learning_rate": 9.128680423352839e-06,
+      "loss": 1.3591,
+      "step": 1979
+    },
+    {
+      "epoch": 1.6032388663967612,
+      "grad_norm": 1.2610631910478973,
+      "learning_rate": 9.12011810747068e-06,
+      "loss": 1.4099,
+      "step": 1980
+    },
+    {
+      "epoch": 1.6040485829959514,
+      "grad_norm": 1.2026899264299413,
+      "learning_rate": 9.111556441642375e-06,
+      "loss": 1.3482,
+      "step": 1981
+    },
+    {
+      "epoch": 1.6048582995951417,
+      "grad_norm": 1.2364975494939607,
+      "learning_rate": 9.10299543219325e-06,
+      "loss": 1.3467,
+      "step": 1982
+    },
+    {
+      "epoch": 1.605668016194332,
+      "grad_norm": 1.2106651427399964,
+      "learning_rate": 9.094435085448153e-06,
+      "loss": 1.3102,
+      "step": 1983
+    },
+    {
+      "epoch": 1.6064777327935222,
+      "grad_norm": 1.1715140130298451,
+      "learning_rate": 9.085875407731444e-06,
+      "loss": 1.3365,
+      "step": 1984
+    },
+    {
+      "epoch": 1.6072874493927125,
+      "grad_norm": 1.1983281783955513,
+      "learning_rate": 9.07731640536698e-06,
+      "loss": 1.4136,
+      "step": 1985
+    },
+    {
+      "epoch": 1.6080971659919028,
+      "grad_norm": 1.1502136649316144,
+      "learning_rate": 9.068758084678126e-06,
+      "loss": 1.3614,
+      "step": 1986
+    },
+    {
+      "epoch": 1.6089068825910933,
+      "grad_norm": 1.213144325311362,
+      "learning_rate": 9.060200451987741e-06,
+      "loss": 1.4013,
+      "step": 1987
+    },
+    {
+      "epoch": 1.6097165991902833,
+      "grad_norm": 1.2047368862136312,
+      "learning_rate": 9.051643513618176e-06,
+      "loss": 1.3126,
+      "step": 1988
+    },
+    {
+      "epoch": 1.6105263157894738,
+      "grad_norm": 1.2612831728284535,
+      "learning_rate": 9.043087275891266e-06,
+      "loss": 1.3652,
+      "step": 1989
+    },
+    {
+      "epoch": 1.6113360323886639,
+      "grad_norm": 1.1804234187245097,
+      "learning_rate": 9.034531745128334e-06,
+      "loss": 1.3749,
+      "step": 1990
+    },
+    {
+      "epoch": 1.6121457489878543,
+      "grad_norm": 1.1934129239753541,
+      "learning_rate": 9.025976927650176e-06,
+      "loss": 1.3223,
+      "step": 1991
+    },
+    {
+      "epoch": 1.6129554655870444,
+      "grad_norm": 1.2290844683292121,
+      "learning_rate": 9.017422829777068e-06,
+      "loss": 1.3693,
+      "step": 1992
+    },
+    {
+      "epoch": 1.6137651821862349,
+      "grad_norm": 1.234955363945963,
+      "learning_rate": 9.008869457828748e-06,
+      "loss": 1.3732,
+      "step": 1993
+    },
+    {
+      "epoch": 1.6145748987854251,
+      "grad_norm": 1.2343436283938842,
+      "learning_rate": 9.000316818124412e-06,
+      "loss": 1.3267,
+      "step": 1994
+    },
+    {
+      "epoch": 1.6153846153846154,
+      "grad_norm": 1.1943281327624646,
+      "learning_rate": 8.991764916982731e-06,
+      "loss": 1.2832,
+      "step": 1995
+    },
+    {
+      "epoch": 1.6161943319838057,
+      "grad_norm": 1.241814934842723,
+      "learning_rate": 8.98321376072182e-06,
+      "loss": 1.4049,
+      "step": 1996
+    },
+    {
+      "epoch": 1.617004048582996,
+      "grad_norm": 1.2281313938106257,
+      "learning_rate": 8.974663355659237e-06,
+      "loss": 1.3636,
+      "step": 1997
+    },
+    {
+      "epoch": 1.6178137651821862,
+      "grad_norm": 1.2025757332697389,
+      "learning_rate": 8.966113708111998e-06,
+      "loss": 1.2603,
+      "step": 1998
+    },
+    {
+      "epoch": 1.6186234817813765,
+      "grad_norm": 1.209319926631317,
+      "learning_rate": 8.957564824396561e-06,
+      "loss": 1.3905,
+      "step": 1999
+    },
+    {
+      "epoch": 1.6194331983805668,
+      "grad_norm": 1.2048829491205542,
+      "learning_rate": 8.949016710828808e-06,
+      "loss": 1.3835,
+      "step": 2000
+    },
+    {
+      "epoch": 1.620242914979757,
+      "grad_norm": 1.1837603866283808,
+      "learning_rate": 8.940469373724054e-06,
+      "loss": 1.3579,
+      "step": 2001
+    },
+    {
+      "epoch": 1.6210526315789475,
+      "grad_norm": 7.115639597248133,
+      "learning_rate": 8.93192281939705e-06,
+      "loss": 1.328,
+      "step": 2002
+    },
+    {
+      "epoch": 1.6218623481781376,
+      "grad_norm": 1.225986024046339,
+      "learning_rate": 8.923377054161959e-06,
+      "loss": 1.3977,
+      "step": 2003
+    },
+    {
+      "epoch": 1.622672064777328,
+      "grad_norm": 1.173818300527572,
+      "learning_rate": 8.914832084332363e-06,
+      "loss": 1.4135,
+      "step": 2004
+    },
+    {
+      "epoch": 1.623481781376518,
+      "grad_norm": 1.2141976317329137,
+      "learning_rate": 8.906287916221259e-06,
+      "loss": 1.3668,
+      "step": 2005
+    },
+    {
+      "epoch": 1.6242914979757086,
+      "grad_norm": 1.1459187530159702,
+      "learning_rate": 8.897744556141047e-06,
+      "loss": 1.3697,
+      "step": 2006
+    },
+    {
+      "epoch": 1.6251012145748986,
+      "grad_norm": 1.1655094826381835,
+      "learning_rate": 8.88920201040354e-06,
+      "loss": 1.3506,
+      "step": 2007
+    },
+    {
+      "epoch": 1.6259109311740891,
+      "grad_norm": 1.2237829104408862,
+      "learning_rate": 8.880660285319941e-06,
+      "loss": 1.4369,
+      "step": 2008
+    },
+    {
+      "epoch": 1.6267206477732794,
+      "grad_norm": 1.2400129820732735,
+      "learning_rate": 8.872119387200844e-06,
+      "loss": 1.3921,
+      "step": 2009
+    },
+    {
+      "epoch": 1.6275303643724697,
+      "grad_norm": 1.2312502428364493,
+      "learning_rate": 8.863579322356242e-06,
+      "loss": 1.2692,
+      "step": 2010
+    },
+    {
+      "epoch": 1.62834008097166,
+      "grad_norm": 1.2294849309798999,
+      "learning_rate": 8.855040097095504e-06,
+      "loss": 1.3047,
+      "step": 2011
+    },
+    {
+      "epoch": 1.6291497975708502,
+      "grad_norm": 1.224426204175276,
+      "learning_rate": 8.846501717727378e-06,
+      "loss": 1.4028,
+      "step": 2012
+    },
+    {
+      "epoch": 1.6299595141700405,
+      "grad_norm": 1.2489783782355426,
+      "learning_rate": 8.837964190559998e-06,
+      "loss": 1.384,
+      "step": 2013
+    },
+    {
+      "epoch": 1.6307692307692307,
+      "grad_norm": 1.2161261861703765,
+      "learning_rate": 8.829427521900852e-06,
+      "loss": 1.4003,
+      "step": 2014
+    },
+    {
+      "epoch": 1.631578947368421,
+      "grad_norm": 1.1628457277298792,
+      "learning_rate": 8.820891718056815e-06,
+      "loss": 1.3563,
+      "step": 2015
+    },
+    {
+      "epoch": 1.6323886639676113,
+      "grad_norm": 1.1999497209012282,
+      "learning_rate": 8.8123567853341e-06,
+      "loss": 1.3742,
+      "step": 2016
+    },
+    {
+      "epoch": 1.6331983805668018,
+      "grad_norm": 1.1700076641674673,
+      "learning_rate": 8.803822730038292e-06,
+      "loss": 1.3734,
+      "step": 2017
+    },
+    {
+      "epoch": 1.6340080971659918,
+      "grad_norm": 1.184177246603208,
+      "learning_rate": 8.795289558474325e-06,
+      "loss": 1.3199,
+      "step": 2018
+    },
+    {
+      "epoch": 1.6348178137651823,
+      "grad_norm": 1.207218490736053,
+      "learning_rate": 8.786757276946473e-06,
+      "loss": 1.4515,
+      "step": 2019
+    },
+    {
+      "epoch": 1.6356275303643724,
+      "grad_norm": 1.1821552126386663,
+      "learning_rate": 8.77822589175836e-06,
+      "loss": 1.383,
+      "step": 2020
+    },
+    {
+      "epoch": 1.6364372469635629,
+      "grad_norm": 1.2274446623556885,
+      "learning_rate": 8.769695409212946e-06,
+      "loss": 1.3733,
+      "step": 2021
+    },
+    {
+      "epoch": 1.637246963562753,
+      "grad_norm": 1.2480667270847892,
+      "learning_rate": 8.76116583561252e-06,
+      "loss": 1.4254,
+      "step": 2022
+    },
+    {
+      "epoch": 1.6380566801619434,
+      "grad_norm": 1.2019082482183845,
+      "learning_rate": 8.752637177258708e-06,
+      "loss": 1.3923,
+      "step": 2023
+    },
+    {
+      "epoch": 1.6388663967611334,
+      "grad_norm": 1.1636761907843682,
+      "learning_rate": 8.744109440452455e-06,
+      "loss": 1.3485,
+      "step": 2024
+    },
+    {
+      "epoch": 1.639676113360324,
+      "grad_norm": 1.2715956411428166,
+      "learning_rate": 8.73558263149402e-06,
+      "loss": 1.2978,
+      "step": 2025
+    },
+    {
+      "epoch": 1.6404858299595142,
+      "grad_norm": 1.2120052316388328,
+      "learning_rate": 8.727056756682985e-06,
+      "loss": 1.4335,
+      "step": 2026
+    },
+    {
+      "epoch": 1.6412955465587045,
+      "grad_norm": 1.2616855428891294,
+      "learning_rate": 8.718531822318236e-06,
+      "loss": 1.3146,
+      "step": 2027
+    },
+    {
+      "epoch": 1.6421052631578947,
+      "grad_norm": 1.2332469667194608,
+      "learning_rate": 8.71000783469797e-06,
+      "loss": 1.4207,
+      "step": 2028
+    },
+    {
+      "epoch": 1.642914979757085,
+      "grad_norm": 1.184361570240323,
+      "learning_rate": 8.701484800119678e-06,
+      "loss": 1.4211,
+      "step": 2029
+    },
+    {
+      "epoch": 1.6437246963562753,
+      "grad_norm": 1.1901404032712164,
+      "learning_rate": 8.692962724880148e-06,
+      "loss": 1.2767,
+      "step": 2030
+    },
+    {
+      "epoch": 1.6445344129554655,
+      "grad_norm": 1.1932059161816782,
+      "learning_rate": 8.684441615275465e-06,
+      "loss": 1.3932,
+      "step": 2031
+    },
+    {
+      "epoch": 1.645344129554656,
+      "grad_norm": 1.1891787691549471,
+      "learning_rate": 8.675921477600996e-06,
+      "loss": 1.3344,
+      "step": 2032
+    },
+    {
+      "epoch": 1.646153846153846,
+      "grad_norm": 1.2379204694615744,
+      "learning_rate": 8.667402318151394e-06,
+      "loss": 1.3101,
+      "step": 2033
+    },
+    {
+      "epoch": 1.6469635627530366,
+      "grad_norm": 1.1672160458877656,
+      "learning_rate": 8.65888414322058e-06,
+      "loss": 1.3827,
+      "step": 2034
+    },
+    {
+      "epoch": 1.6477732793522266,
+      "grad_norm": 1.2157475649535694,
+      "learning_rate": 8.650366959101757e-06,
+      "loss": 1.3832,
+      "step": 2035
+    },
+    {
+      "epoch": 1.648582995951417,
+      "grad_norm": 1.1812552928529416,
+      "learning_rate": 8.641850772087392e-06,
+      "loss": 1.3611,
+      "step": 2036
+    },
+    {
+      "epoch": 1.6493927125506072,
+      "grad_norm": 1.2769305561908675,
+      "learning_rate": 8.633335588469215e-06,
+      "loss": 1.3163,
+      "step": 2037
+    },
+    {
+      "epoch": 1.6502024291497976,
+      "grad_norm": 1.1910880418512533,
+      "learning_rate": 8.62482141453821e-06,
+      "loss": 1.3636,
+      "step": 2038
+    },
+    {
+      "epoch": 1.6510121457489877,
+      "grad_norm": 1.2422148934633142,
+      "learning_rate": 8.616308256584636e-06,
+      "loss": 1.3399,
+      "step": 2039
+    },
+    {
+      "epoch": 1.6518218623481782,
+      "grad_norm": 1.2039779168904876,
+      "learning_rate": 8.607796120897978e-06,
+      "loss": 1.3927,
+      "step": 2040
+    },
+    {
+      "epoch": 1.6526315789473685,
+      "grad_norm": 1.2033223841568303,
+      "learning_rate": 8.599285013766969e-06,
+      "loss": 1.3786,
+      "step": 2041
+    },
+    {
+      "epoch": 1.6534412955465587,
+      "grad_norm": 1.2064528480180141,
+      "learning_rate": 8.590774941479594e-06,
+      "loss": 1.3453,
+      "step": 2042
+    },
+    {
+      "epoch": 1.654251012145749,
+      "grad_norm": 1.2144333343052642,
+      "learning_rate": 8.582265910323063e-06,
+      "loss": 1.3132,
+      "step": 2043
+    },
+    {
+      "epoch": 1.6550607287449393,
+      "grad_norm": 1.2356297509544136,
+      "learning_rate": 8.57375792658382e-06,
+      "loss": 1.384,
+      "step": 2044
+    },
+    {
+      "epoch": 1.6558704453441295,
+      "grad_norm": 1.2113533133382017,
+      "learning_rate": 8.565250996547538e-06,
+      "loss": 1.4371,
+      "step": 2045
+    },
+    {
+      "epoch": 1.6566801619433198,
+      "grad_norm": 1.19928327777792,
+      "learning_rate": 8.556745126499104e-06,
+      "loss": 1.3767,
+      "step": 2046
+    },
+    {
+      "epoch": 1.6574898785425103,
+      "grad_norm": 1.2044105767224726,
+      "learning_rate": 8.548240322722634e-06,
+      "loss": 1.3739,
+      "step": 2047
+    },
+    {
+      "epoch": 1.6582995951417003,
+      "grad_norm": 1.2310691515165295,
+      "learning_rate": 8.539736591501444e-06,
+      "loss": 1.4491,
+      "step": 2048
+    },
+    {
+      "epoch": 1.6591093117408908,
+      "grad_norm": 1.1954706292971082,
+      "learning_rate": 8.531233939118064e-06,
+      "loss": 1.3675,
+      "step": 2049
+    },
+    {
+      "epoch": 1.6599190283400809,
+      "grad_norm": 1.1942959686441643,
+      "learning_rate": 8.522732371854228e-06,
+      "loss": 1.5045,
+      "step": 2050
+    },
+    {
+      "epoch": 1.6607287449392714,
+      "grad_norm": 1.1935221793122746,
+      "learning_rate": 8.514231895990862e-06,
+      "loss": 1.3651,
+      "step": 2051
+    },
+    {
+      "epoch": 1.6615384615384614,
+      "grad_norm": 1.1971712548588236,
+      "learning_rate": 8.50573251780809e-06,
+      "loss": 1.3726,
+      "step": 2052
+    },
+    {
+      "epoch": 1.662348178137652,
+      "grad_norm": 1.189506111043996,
+      "learning_rate": 8.497234243585229e-06,
+      "loss": 1.4117,
+      "step": 2053
+    },
+    {
+      "epoch": 1.663157894736842,
+      "grad_norm": 1.1995587258339253,
+      "learning_rate": 8.488737079600767e-06,
+      "loss": 1.3298,
+      "step": 2054
+    },
+    {
+      "epoch": 1.6639676113360324,
+      "grad_norm": 1.1893431869171311,
+      "learning_rate": 8.480241032132394e-06,
+      "loss": 1.368,
+      "step": 2055
+    },
+    {
+      "epoch": 1.6647773279352227,
+      "grad_norm": 1.2451970649626896,
+      "learning_rate": 8.47174610745695e-06,
+      "loss": 1.3723,
+      "step": 2056
+    },
+    {
+      "epoch": 1.665587044534413,
+      "grad_norm": 1.19193248474202,
+      "learning_rate": 8.463252311850466e-06,
+      "loss": 1.3371,
+      "step": 2057
+    },
+    {
+      "epoch": 1.6663967611336032,
+      "grad_norm": 1.2609189929314095,
+      "learning_rate": 8.454759651588127e-06,
+      "loss": 1.4365,
+      "step": 2058
+    },
+    {
+      "epoch": 1.6672064777327935,
+      "grad_norm": 1.214748523646701,
+      "learning_rate": 8.446268132944279e-06,
+      "loss": 1.3793,
+      "step": 2059
+    },
+    {
+      "epoch": 1.6680161943319838,
+      "grad_norm": 1.2916028366953496,
+      "learning_rate": 8.437777762192434e-06,
+      "loss": 1.4646,
+      "step": 2060
+    },
+    {
+      "epoch": 1.668825910931174,
+      "grad_norm": 1.217799194319695,
+      "learning_rate": 8.429288545605248e-06,
+      "loss": 1.374,
+      "step": 2061
+    },
+    {
+      "epoch": 1.6696356275303643,
+      "grad_norm": 1.1681254207450982,
+      "learning_rate": 8.42080048945452e-06,
+      "loss": 1.3769,
+      "step": 2062
+    },
+    {
+      "epoch": 1.6704453441295546,
+      "grad_norm": 1.2364124010512334,
+      "learning_rate": 8.412313600011209e-06,
+      "loss": 1.3553,
+      "step": 2063
+    },
+    {
+      "epoch": 1.671255060728745,
+      "grad_norm": 1.237450931928774,
+      "learning_rate": 8.403827883545393e-06,
+      "loss": 1.4733,
+      "step": 2064
+    },
+    {
+      "epoch": 1.6720647773279351,
+      "grad_norm": 1.1760825158488026,
+      "learning_rate": 8.395343346326295e-06,
+      "loss": 1.3914,
+      "step": 2065
+    },
+    {
+      "epoch": 1.6728744939271256,
+      "grad_norm": 1.2383834233283117,
+      "learning_rate": 8.386859994622266e-06,
+      "loss": 1.3251,
+      "step": 2066
+    },
+    {
+      "epoch": 1.6736842105263157,
+      "grad_norm": 1.2192262734981085,
+      "learning_rate": 8.378377834700769e-06,
+      "loss": 1.3688,
+      "step": 2067
+    },
+    {
+      "epoch": 1.6744939271255062,
+      "grad_norm": 1.1753730133070337,
+      "learning_rate": 8.369896872828406e-06,
+      "loss": 1.376,
+      "step": 2068
+    },
+    {
+      "epoch": 1.6753036437246962,
+      "grad_norm": 1.1758350186139863,
+      "learning_rate": 8.361417115270878e-06,
+      "loss": 1.3457,
+      "step": 2069
+    },
+    {
+      "epoch": 1.6761133603238867,
+      "grad_norm": 1.2192998551029626,
+      "learning_rate": 8.352938568293e-06,
+      "loss": 1.3951,
+      "step": 2070
+    },
+    {
+      "epoch": 1.676923076923077,
+      "grad_norm": 1.2527451825200207,
+      "learning_rate": 8.3444612381587e-06,
+      "loss": 1.3647,
+      "step": 2071
+    },
+    {
+      "epoch": 1.6777327935222672,
+      "grad_norm": 1.2447953070318596,
+      "learning_rate": 8.335985131131002e-06,
+      "loss": 1.4072,
+      "step": 2072
+    },
+    {
+      "epoch": 1.6785425101214575,
+      "grad_norm": 1.2245743293643092,
+      "learning_rate": 8.327510253472023e-06,
+      "loss": 1.3841,
+      "step": 2073
+    },
+    {
+      "epoch": 1.6793522267206478,
+      "grad_norm": 1.210563123337093,
+      "learning_rate": 8.319036611442974e-06,
+      "loss": 1.3674,
+      "step": 2074
+    },
+    {
+      "epoch": 1.680161943319838,
+      "grad_norm": 1.215284281361853,
+      "learning_rate": 8.310564211304159e-06,
+      "loss": 1.3919,
+      "step": 2075
+    },
+    {
+      "epoch": 1.6809716599190283,
+      "grad_norm": 1.1606361526638542,
+      "learning_rate": 8.302093059314955e-06,
+      "loss": 1.4128,
+      "step": 2076
+    },
+    {
+      "epoch": 1.6817813765182186,
+      "grad_norm": 1.2280121763912009,
+      "learning_rate": 8.293623161733819e-06,
+      "loss": 1.3144,
+      "step": 2077
+    },
+    {
+      "epoch": 1.6825910931174088,
+      "grad_norm": 1.1672110474120607,
+      "learning_rate": 8.285154524818288e-06,
+      "loss": 1.3403,
+      "step": 2078
+    },
+    {
+      "epoch": 1.6834008097165993,
+      "grad_norm": 1.1959606638614337,
+      "learning_rate": 8.27668715482496e-06,
+      "loss": 1.4199,
+      "step": 2079
+    },
+    {
+      "epoch": 1.6842105263157894,
+      "grad_norm": 1.178527090126989,
+      "learning_rate": 8.268221058009506e-06,
+      "loss": 1.331,
+      "step": 2080
+    },
+    {
+      "epoch": 1.6850202429149799,
+      "grad_norm": 1.212039723837128,
+      "learning_rate": 8.259756240626646e-06,
+      "loss": 1.3688,
+      "step": 2081
+    },
+    {
+      "epoch": 1.68582995951417,
+      "grad_norm": 1.2117727419701412,
+      "learning_rate": 8.251292708930156e-06,
+      "loss": 1.4021,
+      "step": 2082
+    },
+    {
+      "epoch": 1.6866396761133604,
+      "grad_norm": 1.1328543998014728,
+      "learning_rate": 8.242830469172873e-06,
+      "loss": 1.2837,
+      "step": 2083
+    },
+    {
+      "epoch": 1.6874493927125505,
+      "grad_norm": 1.174797996844683,
+      "learning_rate": 8.234369527606667e-06,
+      "loss": 1.3522,
+      "step": 2084
+    },
+    {
+      "epoch": 1.688259109311741,
+      "grad_norm": 1.187195276450799,
+      "learning_rate": 8.225909890482456e-06,
+      "loss": 1.3685,
+      "step": 2085
+    },
+    {
+      "epoch": 1.6890688259109312,
+      "grad_norm": 1.2247680029203445,
+      "learning_rate": 8.217451564050185e-06,
+      "loss": 1.3567,
+      "step": 2086
+    },
+    {
+      "epoch": 1.6898785425101215,
+      "grad_norm": 1.1886609297985533,
+      "learning_rate": 8.20899455455885e-06,
+      "loss": 1.336,
+      "step": 2087
+    },
+    {
+      "epoch": 1.6906882591093118,
+      "grad_norm": 1.209073662488914,
+      "learning_rate": 8.200538868256455e-06,
+      "loss": 1.3241,
+      "step": 2088
+    },
+    {
+      "epoch": 1.691497975708502,
+      "grad_norm": 1.2521452085782394,
+      "learning_rate": 8.192084511390033e-06,
+      "loss": 1.3865,
+      "step": 2089
+    },
+    {
+      "epoch": 1.6923076923076923,
+      "grad_norm": 1.1899332902396975,
+      "learning_rate": 8.183631490205636e-06,
+      "loss": 1.3589,
+      "step": 2090
+    },
+    {
+      "epoch": 1.6931174089068826,
+      "grad_norm": 1.1695363453646483,
+      "learning_rate": 8.17517981094833e-06,
+      "loss": 1.3299,
+      "step": 2091
+    },
+    {
+      "epoch": 1.6939271255060728,
+      "grad_norm": 1.1873531255692886,
+      "learning_rate": 8.166729479862185e-06,
+      "loss": 1.41,
+      "step": 2092
+    },
+    {
+      "epoch": 1.694736842105263,
+      "grad_norm": 1.177164040688767,
+      "learning_rate": 8.15828050319028e-06,
+      "loss": 1.376,
+      "step": 2093
+    },
+    {
+      "epoch": 1.6955465587044536,
+      "grad_norm": 1.2131772752466448,
+      "learning_rate": 8.149832887174686e-06,
+      "loss": 1.3092,
+      "step": 2094
+    },
+    {
+      "epoch": 1.6963562753036436,
+      "grad_norm": 1.1863052171248314,
+      "learning_rate": 8.141386638056482e-06,
+      "loss": 1.3064,
+      "step": 2095
+    },
+    {
+      "epoch": 1.6971659919028341,
+      "grad_norm": 1.2087525026441543,
+      "learning_rate": 8.132941762075726e-06,
+      "loss": 1.3384,
+      "step": 2096
+    },
+    {
+      "epoch": 1.6979757085020242,
+      "grad_norm": 1.1909825383271502,
+      "learning_rate": 8.124498265471462e-06,
+      "loss": 1.3282,
+      "step": 2097
+    },
+    {
+      "epoch": 1.6987854251012147,
+      "grad_norm": 1.2073382417531069,
+      "learning_rate": 8.116056154481721e-06,
+      "loss": 1.4204,
+      "step": 2098
+    },
+    {
+      "epoch": 1.6995951417004047,
+      "grad_norm": 1.2336753723660032,
+      "learning_rate": 8.107615435343506e-06,
+      "loss": 1.4096,
+      "step": 2099
+    },
+    {
+      "epoch": 1.7004048582995952,
+      "grad_norm": 1.2244965473022886,
+      "learning_rate": 8.09917611429279e-06,
+      "loss": 1.3345,
+      "step": 2100
+    },
+    {
+      "epoch": 1.7012145748987855,
+      "grad_norm": 1.2365934665855882,
+      "learning_rate": 8.090738197564519e-06,
+      "loss": 1.3549,
+      "step": 2101
+    },
+    {
+      "epoch": 1.7020242914979757,
+      "grad_norm": 1.2403759805495438,
+      "learning_rate": 8.082301691392593e-06,
+      "loss": 1.4056,
+      "step": 2102
+    },
+    {
+      "epoch": 1.702834008097166,
+      "grad_norm": 1.23974335524669,
+      "learning_rate": 8.073866602009883e-06,
+      "loss": 1.314,
+      "step": 2103
+    },
+    {
+      "epoch": 1.7036437246963563,
+      "grad_norm": 1.2444897063130003,
+      "learning_rate": 8.0654329356482e-06,
+      "loss": 1.3807,
+      "step": 2104
+    },
+    {
+      "epoch": 1.7044534412955465,
+      "grad_norm": 1.1949535882055378,
+      "learning_rate": 8.057000698538311e-06,
+      "loss": 1.3949,
+      "step": 2105
+    },
+    {
+      "epoch": 1.7052631578947368,
+      "grad_norm": 1.2104242593278358,
+      "learning_rate": 8.048569896909925e-06,
+      "loss": 1.2957,
+      "step": 2106
+    },
+    {
+      "epoch": 1.706072874493927,
+      "grad_norm": 1.242882880007763,
+      "learning_rate": 8.040140536991688e-06,
+      "loss": 1.3838,
+      "step": 2107
+    },
+    {
+      "epoch": 1.7068825910931174,
+      "grad_norm": 1.2414139288923256,
+      "learning_rate": 8.031712625011186e-06,
+      "loss": 1.3637,
+      "step": 2108
+    },
+    {
+      "epoch": 1.7076923076923078,
+      "grad_norm": 1.1934131947311941,
+      "learning_rate": 8.023286167194934e-06,
+      "loss": 1.345,
+      "step": 2109
+    },
+    {
+      "epoch": 1.708502024291498,
+      "grad_norm": 1.2339934690903747,
+      "learning_rate": 8.014861169768362e-06,
+      "loss": 1.43,
+      "step": 2110
+    },
+    {
+      "epoch": 1.7093117408906884,
+      "grad_norm": 1.172849536919364,
+      "learning_rate": 8.006437638955846e-06,
+      "loss": 1.4032,
+      "step": 2111
+    },
+    {
+      "epoch": 1.7101214574898784,
+      "grad_norm": 1.1971353964773175,
+      "learning_rate": 7.99801558098065e-06,
+      "loss": 1.3609,
+      "step": 2112
+    },
+    {
+      "epoch": 1.710931174089069,
+      "grad_norm": 1.210793522412267,
+      "learning_rate": 7.98959500206497e-06,
+      "loss": 1.3799,
+      "step": 2113
+    },
+    {
+      "epoch": 1.711740890688259,
+      "grad_norm": 1.1992770532754535,
+      "learning_rate": 7.9811759084299e-06,
+      "loss": 1.3861,
+      "step": 2114
+    },
+    {
+      "epoch": 1.7125506072874495,
+      "grad_norm": 1.250496757926102,
+      "learning_rate": 7.972758306295436e-06,
+      "loss": 1.3585,
+      "step": 2115
+    },
+    {
+      "epoch": 1.7133603238866397,
+      "grad_norm": 1.2143039563042084,
+      "learning_rate": 7.964342201880478e-06,
+      "loss": 1.3448,
+      "step": 2116
+    },
+    {
+      "epoch": 1.71417004048583,
+      "grad_norm": 1.2503268692965592,
+      "learning_rate": 7.955927601402817e-06,
+      "loss": 1.342,
+      "step": 2117
+    },
+    {
+      "epoch": 1.7149797570850203,
+      "grad_norm": 1.272542261752361,
+      "learning_rate": 7.947514511079126e-06,
+      "loss": 1.4221,
+      "step": 2118
+    },
+    {
+      "epoch": 1.7157894736842105,
+      "grad_norm": 1.2723130848911237,
+      "learning_rate": 7.939102937124975e-06,
+      "loss": 1.4166,
+      "step": 2119
+    },
+    {
+      "epoch": 1.7165991902834008,
+      "grad_norm": 1.2332492013002891,
+      "learning_rate": 7.930692885754806e-06,
+      "loss": 1.3435,
+      "step": 2120
+    },
+    {
+      "epoch": 1.717408906882591,
+      "grad_norm": 1.2106787991881476,
+      "learning_rate": 7.922284363181937e-06,
+      "loss": 1.3128,
+      "step": 2121
+    },
+    {
+      "epoch": 1.7182186234817813,
+      "grad_norm": 1.2133263111046222,
+      "learning_rate": 7.913877375618555e-06,
+      "loss": 1.3606,
+      "step": 2122
+    },
+    {
+      "epoch": 1.7190283400809716,
+      "grad_norm": 1.2393512093973866,
+      "learning_rate": 7.90547192927572e-06,
+      "loss": 1.3813,
+      "step": 2123
+    },
+    {
+      "epoch": 1.719838056680162,
+      "grad_norm": 1.18596615707014,
+      "learning_rate": 7.897068030363341e-06,
+      "loss": 1.3718,
+      "step": 2124
+    },
+    {
+      "epoch": 1.7206477732793521,
+      "grad_norm": 1.1774499718505693,
+      "learning_rate": 7.888665685090194e-06,
+      "loss": 1.3439,
+      "step": 2125
+    },
+    {
+      "epoch": 1.7214574898785426,
+      "grad_norm": 1.196454598660986,
+      "learning_rate": 7.880264899663901e-06,
+      "loss": 1.3454,
+      "step": 2126
+    },
+    {
+      "epoch": 1.7222672064777327,
+      "grad_norm": 1.247233467944127,
+      "learning_rate": 7.871865680290943e-06,
+      "loss": 1.3977,
+      "step": 2127
+    },
+    {
+      "epoch": 1.7230769230769232,
+      "grad_norm": 1.189652568647767,
+      "learning_rate": 7.863468033176632e-06,
+      "loss": 1.3581,
+      "step": 2128
+    },
+    {
+      "epoch": 1.7238866396761132,
+      "grad_norm": 1.1938170625000661,
+      "learning_rate": 7.855071964525115e-06,
+      "loss": 1.3333,
+      "step": 2129
+    },
+    {
+      "epoch": 1.7246963562753037,
+      "grad_norm": 1.1673683171310625,
+      "learning_rate": 7.846677480539392e-06,
+      "loss": 1.3601,
+      "step": 2130
+    },
+    {
+      "epoch": 1.725506072874494,
+      "grad_norm": 1.1775636728244776,
+      "learning_rate": 7.838284587421273e-06,
+      "loss": 1.3889,
+      "step": 2131
+    },
+    {
+      "epoch": 1.7263157894736842,
+      "grad_norm": 1.2347561246730585,
+      "learning_rate": 7.829893291371399e-06,
+      "loss": 1.3683,
+      "step": 2132
+    },
+    {
+      "epoch": 1.7271255060728745,
+      "grad_norm": 1.1960868396656386,
+      "learning_rate": 7.821503598589234e-06,
+      "loss": 1.364,
+      "step": 2133
+    },
+    {
+      "epoch": 1.7279352226720648,
+      "grad_norm": 1.1576414391616703,
+      "learning_rate": 7.813115515273052e-06,
+      "loss": 1.3586,
+      "step": 2134
+    },
+    {
+      "epoch": 1.728744939271255,
+      "grad_norm": 1.2728426715326342,
+      "learning_rate": 7.80472904761995e-06,
+      "loss": 1.3417,
+      "step": 2135
+    },
+    {
+      "epoch": 1.7295546558704453,
+      "grad_norm": 1.196854159805804,
+      "learning_rate": 7.796344201825816e-06,
+      "loss": 1.3683,
+      "step": 2136
+    },
+    {
+      "epoch": 1.7303643724696356,
+      "grad_norm": 1.2227814698116768,
+      "learning_rate": 7.787960984085346e-06,
+      "loss": 1.3132,
+      "step": 2137
+    },
+    {
+      "epoch": 1.7311740890688259,
+      "grad_norm": 1.1904404210188404,
+      "learning_rate": 7.779579400592039e-06,
+      "loss": 1.2808,
+      "step": 2138
+    },
+    {
+      "epoch": 1.7319838056680164,
+      "grad_norm": 1.2071293931299591,
+      "learning_rate": 7.771199457538177e-06,
+      "loss": 1.3415,
+      "step": 2139
+    },
+    {
+      "epoch": 1.7327935222672064,
+      "grad_norm": 1.2589835991554157,
+      "learning_rate": 7.762821161114834e-06,
+      "loss": 1.3671,
+      "step": 2140
+    },
+    {
+      "epoch": 1.733603238866397,
+      "grad_norm": 1.2502098184841257,
+      "learning_rate": 7.754444517511869e-06,
+      "loss": 1.3771,
+      "step": 2141
+    },
+    {
+      "epoch": 1.734412955465587,
+      "grad_norm": 1.194748364690635,
+      "learning_rate": 7.746069532917918e-06,
+      "loss": 1.3148,
+      "step": 2142
+    },
+    {
+      "epoch": 1.7352226720647774,
+      "grad_norm": 1.2615364812188445,
+      "learning_rate": 7.737696213520397e-06,
+      "loss": 1.3587,
+      "step": 2143
+    },
+    {
+      "epoch": 1.7360323886639675,
+      "grad_norm": 1.1791684259343356,
+      "learning_rate": 7.72932456550548e-06,
+      "loss": 1.335,
+      "step": 2144
+    },
+    {
+      "epoch": 1.736842105263158,
+      "grad_norm": 1.2804937427129839,
+      "learning_rate": 7.720954595058118e-06,
+      "loss": 1.3563,
+      "step": 2145
+    },
+    {
+      "epoch": 1.737651821862348,
+      "grad_norm": 1.2678895788470363,
+      "learning_rate": 7.712586308362017e-06,
+      "loss": 1.3297,
+      "step": 2146
+    },
+    {
+      "epoch": 1.7384615384615385,
+      "grad_norm": 1.1864809688777394,
+      "learning_rate": 7.704219711599637e-06,
+      "loss": 1.3111,
+      "step": 2147
+    },
+    {
+      "epoch": 1.7392712550607288,
+      "grad_norm": 1.1913833583807754,
+      "learning_rate": 7.695854810952194e-06,
+      "loss": 1.3823,
+      "step": 2148
+    },
+    {
+      "epoch": 1.740080971659919,
+      "grad_norm": 1.2454501834374865,
+      "learning_rate": 7.687491612599651e-06,
+      "loss": 1.3384,
+      "step": 2149
+    },
+    {
+      "epoch": 1.7408906882591093,
+      "grad_norm": 1.2311094048424676,
+      "learning_rate": 7.679130122720704e-06,
+      "loss": 1.4009,
+      "step": 2150
+    },
+    {
+      "epoch": 1.7417004048582996,
+      "grad_norm": 1.2146362073312729,
+      "learning_rate": 7.670770347492804e-06,
+      "loss": 1.3641,
+      "step": 2151
+    },
+    {
+      "epoch": 1.7425101214574898,
+      "grad_norm": 1.214516287695011,
+      "learning_rate": 7.662412293092118e-06,
+      "loss": 1.3468,
+      "step": 2152
+    },
+    {
+      "epoch": 1.7433198380566801,
+      "grad_norm": 1.1764628400623898,
+      "learning_rate": 7.654055965693556e-06,
+      "loss": 1.3273,
+      "step": 2153
+    },
+    {
+      "epoch": 1.7441295546558706,
+      "grad_norm": 1.1411724141436934,
+      "learning_rate": 7.64570137147074e-06,
+      "loss": 1.3301,
+      "step": 2154
+    },
+    {
+      "epoch": 1.7449392712550607,
+      "grad_norm": 1.2122093274665457,
+      "learning_rate": 7.637348516596016e-06,
+      "loss": 1.3525,
+      "step": 2155
+    },
+    {
+      "epoch": 1.7457489878542511,
+      "grad_norm": 1.2399842029700836,
+      "learning_rate": 7.628997407240453e-06,
+      "loss": 1.3807,
+      "step": 2156
+    },
+    {
+      "epoch": 1.7465587044534412,
+      "grad_norm": 1.256148734557012,
+      "learning_rate": 7.620648049573815e-06,
+      "loss": 1.3519,
+      "step": 2157
+    },
+    {
+      "epoch": 1.7473684210526317,
+      "grad_norm": 1.2286990308441597,
+      "learning_rate": 7.61230044976458e-06,
+      "loss": 1.3506,
+      "step": 2158
+    },
+    {
+      "epoch": 1.7481781376518217,
+      "grad_norm": 1.2260430970223977,
+      "learning_rate": 7.603954613979933e-06,
+      "loss": 1.2832,
+      "step": 2159
+    },
+    {
+      "epoch": 1.7489878542510122,
+      "grad_norm": 1.1988962090230735,
+      "learning_rate": 7.59561054838575e-06,
+      "loss": 1.307,
+      "step": 2160
+    },
+    {
+      "epoch": 1.7497975708502023,
+      "grad_norm": 1.1876276282662488,
+      "learning_rate": 7.587268259146596e-06,
+      "loss": 1.3742,
+      "step": 2161
+    },
+    {
+      "epoch": 1.7506072874493928,
+      "grad_norm": 1.327874554832877,
+      "learning_rate": 7.578927752425727e-06,
+      "loss": 1.3885,
+      "step": 2162
+    },
+    {
+      "epoch": 1.751417004048583,
+      "grad_norm": 1.2617574575652093,
+      "learning_rate": 7.570589034385083e-06,
+      "loss": 1.4301,
+      "step": 2163
+    },
+    {
+      "epoch": 1.7522267206477733,
+      "grad_norm": 1.1822211358693704,
+      "learning_rate": 7.562252111185282e-06,
+      "loss": 1.3268,
+      "step": 2164
+    },
+    {
+      "epoch": 1.7530364372469636,
+      "grad_norm": 1.2264998801603384,
+      "learning_rate": 7.5539169889856135e-06,
+      "loss": 1.3632,
+      "step": 2165
+    },
+    {
+      "epoch": 1.7538461538461538,
+      "grad_norm": 1.2308240591640838,
+      "learning_rate": 7.545583673944038e-06,
+      "loss": 1.3703,
+      "step": 2166
+    },
+    {
+      "epoch": 1.754655870445344,
+      "grad_norm": 1.2034026004405005,
+      "learning_rate": 7.537252172217185e-06,
+      "loss": 1.4068,
+      "step": 2167
+    },
+    {
+      "epoch": 1.7554655870445344,
+      "grad_norm": 1.2716169869380658,
+      "learning_rate": 7.528922489960339e-06,
+      "loss": 1.3546,
+      "step": 2168
+    },
+    {
+      "epoch": 1.7562753036437249,
+      "grad_norm": 1.1963463371260916,
+      "learning_rate": 7.52059463332744e-06,
+      "loss": 1.3827,
+      "step": 2169
+    },
+    {
+      "epoch": 1.757085020242915,
+      "grad_norm": 1.2214699387119217,
+      "learning_rate": 7.512268608471083e-06,
+      "loss": 1.3597,
+      "step": 2170
+    },
+    {
+      "epoch": 1.7578947368421054,
+      "grad_norm": 1.1980661545941786,
+      "learning_rate": 7.503944421542508e-06,
+      "loss": 1.4244,
+      "step": 2171
+    },
+    {
+      "epoch": 1.7587044534412954,
+      "grad_norm": 1.2346433817261153,
+      "learning_rate": 7.495622078691597e-06,
+      "loss": 1.3887,
+      "step": 2172
+    },
+    {
+      "epoch": 1.759514170040486,
+      "grad_norm": 1.2316791370226217,
+      "learning_rate": 7.487301586066866e-06,
+      "loss": 1.4128,
+      "step": 2173
+    },
+    {
+      "epoch": 1.760323886639676,
+      "grad_norm": 1.247303111408137,
+      "learning_rate": 7.47898294981547e-06,
+      "loss": 1.3464,
+      "step": 2174
+    },
+    {
+      "epoch": 1.7611336032388665,
+      "grad_norm": 1.1858468665595168,
+      "learning_rate": 7.470666176083193e-06,
+      "loss": 1.3297,
+      "step": 2175
+    },
+    {
+      "epoch": 1.7619433198380565,
+      "grad_norm": 1.17298641444367,
+      "learning_rate": 7.462351271014438e-06,
+      "loss": 1.3335,
+      "step": 2176
+    },
+    {
+      "epoch": 1.762753036437247,
+      "grad_norm": 1.2071311078649347,
+      "learning_rate": 7.454038240752228e-06,
+      "loss": 1.3957,
+      "step": 2177
+    },
+    {
+      "epoch": 1.7635627530364373,
+      "grad_norm": 1.2383361119848715,
+      "learning_rate": 7.4457270914382056e-06,
+      "loss": 1.3592,
+      "step": 2178
+    },
+    {
+      "epoch": 1.7643724696356275,
+      "grad_norm": 1.2321882683113468,
+      "learning_rate": 7.437417829212618e-06,
+      "loss": 1.3527,
+      "step": 2179
+    },
+    {
+      "epoch": 1.7651821862348178,
+      "grad_norm": 1.1470511569497224,
+      "learning_rate": 7.42911046021432e-06,
+      "loss": 1.3536,
+      "step": 2180
+    },
+    {
+      "epoch": 1.765991902834008,
+      "grad_norm": 1.1704186219839523,
+      "learning_rate": 7.420804990580772e-06,
+      "loss": 1.3738,
+      "step": 2181
+    },
+    {
+      "epoch": 1.7668016194331984,
+      "grad_norm": 1.155250015703002,
+      "learning_rate": 7.4125014264480225e-06,
+      "loss": 1.3488,
+      "step": 2182
+    },
+    {
+      "epoch": 1.7676113360323886,
+      "grad_norm": 1.2361131817316744,
+      "learning_rate": 7.404199773950724e-06,
+      "loss": 1.3283,
+      "step": 2183
+    },
+    {
+      "epoch": 1.768421052631579,
+      "grad_norm": 1.1572149457442449,
+      "learning_rate": 7.395900039222108e-06,
+      "loss": 1.4089,
+      "step": 2184
+    },
+    {
+      "epoch": 1.7692307692307692,
+      "grad_norm": 1.2024459906835356,
+      "learning_rate": 7.387602228393987e-06,
+      "loss": 1.3428,
+      "step": 2185
+    },
+    {
+      "epoch": 1.7700404858299597,
+      "grad_norm": 1.2444788254439085,
+      "learning_rate": 7.379306347596762e-06,
+      "loss": 1.3889,
+      "step": 2186
+    },
+    {
+      "epoch": 1.7708502024291497,
+      "grad_norm": 1.2199415758978813,
+      "learning_rate": 7.371012402959399e-06,
+      "loss": 1.3438,
+      "step": 2187
+    },
+    {
+      "epoch": 1.7716599190283402,
+      "grad_norm": 1.147217425822318,
+      "learning_rate": 7.362720400609437e-06,
+      "loss": 1.4056,
+      "step": 2188
+    },
+    {
+      "epoch": 1.7724696356275302,
+      "grad_norm": 1.2527673146015181,
+      "learning_rate": 7.354430346672983e-06,
+      "loss": 1.3813,
+      "step": 2189
+    },
+    {
+      "epoch": 1.7732793522267207,
+      "grad_norm": 1.261309500485428,
+      "learning_rate": 7.346142247274695e-06,
+      "loss": 1.4284,
+      "step": 2190
+    },
+    {
+      "epoch": 1.7740890688259108,
+      "grad_norm": 1.2320074330261714,
+      "learning_rate": 7.337856108537802e-06,
+      "loss": 1.4159,
+      "step": 2191
+    },
+    {
+      "epoch": 1.7748987854251013,
+      "grad_norm": 1.247234002362416,
+      "learning_rate": 7.329571936584072e-06,
+      "loss": 1.4277,
+      "step": 2192
+    },
+    {
+      "epoch": 1.7757085020242915,
+      "grad_norm": 1.2396500425068102,
+      "learning_rate": 7.321289737533826e-06,
+      "loss": 1.2784,
+      "step": 2193
+    },
+    {
+      "epoch": 1.7765182186234818,
+      "grad_norm": 1.221512450585911,
+      "learning_rate": 7.313009517505923e-06,
+      "loss": 1.4277,
+      "step": 2194
+    },
+    {
+      "epoch": 1.777327935222672,
+      "grad_norm": 1.2207042254565934,
+      "learning_rate": 7.304731282617762e-06,
+      "loss": 1.3541,
+      "step": 2195
+    },
+    {
+      "epoch": 1.7781376518218623,
+      "grad_norm": 1.179448361791702,
+      "learning_rate": 7.29645503898528e-06,
+      "loss": 1.3417,
+      "step": 2196
+    },
+    {
+      "epoch": 1.7789473684210526,
+      "grad_norm": 1.210105556711578,
+      "learning_rate": 7.288180792722934e-06,
+      "loss": 1.3518,
+      "step": 2197
+    },
+    {
+      "epoch": 1.7797570850202429,
+      "grad_norm": 1.2517556094115736,
+      "learning_rate": 7.279908549943708e-06,
+      "loss": 1.3949,
+      "step": 2198
+    },
+    {
+      "epoch": 1.7805668016194331,
+      "grad_norm": 1.213422808224917,
+      "learning_rate": 7.271638316759116e-06,
+      "loss": 1.2904,
+      "step": 2199
+    },
+    {
+      "epoch": 1.7813765182186234,
+      "grad_norm": 1.21215761739376,
+      "learning_rate": 7.263370099279173e-06,
+      "loss": 1.282,
+      "step": 2200
+    },
+    {
+      "epoch": 1.782186234817814,
+      "grad_norm": 1.2357528688986366,
+      "learning_rate": 7.255103903612413e-06,
+      "loss": 1.3415,
+      "step": 2201
+    },
+    {
+      "epoch": 1.782995951417004,
+      "grad_norm": 1.2102242154201936,
+      "learning_rate": 7.246839735865874e-06,
+      "loss": 1.3856,
+      "step": 2202
+    },
+    {
+      "epoch": 1.7838056680161944,
+      "grad_norm": 1.2066789803427276,
+      "learning_rate": 7.238577602145094e-06,
+      "loss": 1.3408,
+      "step": 2203
+    },
+    {
+      "epoch": 1.7846153846153845,
+      "grad_norm": 1.2002015685758425,
+      "learning_rate": 7.230317508554113e-06,
+      "loss": 1.3005,
+      "step": 2204
+    },
+    {
+      "epoch": 1.785425101214575,
+      "grad_norm": 1.2371780286395402,
+      "learning_rate": 7.2220594611954606e-06,
+      "loss": 1.3464,
+      "step": 2205
+    },
+    {
+      "epoch": 1.786234817813765,
+      "grad_norm": 1.2122162279977085,
+      "learning_rate": 7.21380346617015e-06,
+      "loss": 1.4097,
+      "step": 2206
+    },
+    {
+      "epoch": 1.7870445344129555,
+      "grad_norm": 1.1890616495957214,
+      "learning_rate": 7.20554952957769e-06,
+      "loss": 1.3381,
+      "step": 2207
+    },
+    {
+      "epoch": 1.7878542510121458,
+      "grad_norm": 1.2240916448009678,
+      "learning_rate": 7.197297657516062e-06,
+      "loss": 1.2961,
+      "step": 2208
+    },
+    {
+      "epoch": 1.788663967611336,
+      "grad_norm": 1.2291351852794627,
+      "learning_rate": 7.189047856081719e-06,
+      "loss": 1.3057,
+      "step": 2209
+    },
+    {
+      "epoch": 1.7894736842105263,
+      "grad_norm": 1.2303308564438198,
+      "learning_rate": 7.1808001313695855e-06,
+      "loss": 1.434,
+      "step": 2210
+    },
+    {
+      "epoch": 1.7902834008097166,
+      "grad_norm": 1.2068008557972785,
+      "learning_rate": 7.172554489473057e-06,
+      "loss": 1.371,
+      "step": 2211
+    },
+    {
+      "epoch": 1.7910931174089069,
+      "grad_norm": 1.236399538012193,
+      "learning_rate": 7.164310936483986e-06,
+      "loss": 1.3062,
+      "step": 2212
+    },
+    {
+      "epoch": 1.7919028340080971,
+      "grad_norm": 1.1998903560290575,
+      "learning_rate": 7.156069478492679e-06,
+      "loss": 1.3007,
+      "step": 2213
+    },
+    {
+      "epoch": 1.7927125506072874,
+      "grad_norm": 1.2525924525988954,
+      "learning_rate": 7.1478301215878975e-06,
+      "loss": 1.4231,
+      "step": 2214
+    },
+    {
+      "epoch": 1.7935222672064777,
+      "grad_norm": 1.2758950545332266,
+      "learning_rate": 7.1395928718568605e-06,
+      "loss": 1.359,
+      "step": 2215
+    },
+    {
+      "epoch": 1.7943319838056682,
+      "grad_norm": 1.2001272107207346,
+      "learning_rate": 7.131357735385213e-06,
+      "loss": 1.3625,
+      "step": 2216
+    },
+    {
+      "epoch": 1.7951417004048582,
+      "grad_norm": 1.2200265505607486,
+      "learning_rate": 7.123124718257052e-06,
+      "loss": 1.3997,
+      "step": 2217
+    },
+    {
+      "epoch": 1.7959514170040487,
+      "grad_norm": 1.2159603840435709,
+      "learning_rate": 7.114893826554896e-06,
+      "loss": 1.3696,
+      "step": 2218
+    },
+    {
+      "epoch": 1.7967611336032387,
+      "grad_norm": 1.2133557545824332,
+      "learning_rate": 7.106665066359708e-06,
+      "loss": 1.4013,
+      "step": 2219
+    },
+    {
+      "epoch": 1.7975708502024292,
+      "grad_norm": 1.2428462413099133,
+      "learning_rate": 7.098438443750865e-06,
+      "loss": 1.4124,
+      "step": 2220
+    },
+    {
+      "epoch": 1.7983805668016193,
+      "grad_norm": 1.1962533108457247,
+      "learning_rate": 7.0902139648061676e-06,
+      "loss": 1.3797,
+      "step": 2221
+    },
+    {
+      "epoch": 1.7991902834008098,
+      "grad_norm": 1.1949448871594797,
+      "learning_rate": 7.081991635601835e-06,
+      "loss": 1.3777,
+      "step": 2222
+    },
+    {
+      "epoch": 1.8,
+      "grad_norm": 1.1739839756203938,
+      "learning_rate": 7.073771462212502e-06,
+      "loss": 1.3431,
+      "step": 2223
+    },
+    {
+      "epoch": 1.8008097165991903,
+      "grad_norm": 1.2246364614759249,
+      "learning_rate": 7.065553450711202e-06,
+      "loss": 1.3858,
+      "step": 2224
+    },
+    {
+      "epoch": 1.8016194331983806,
+      "grad_norm": 1.2541194418394814,
+      "learning_rate": 7.057337607169373e-06,
+      "loss": 1.4208,
+      "step": 2225
+    },
+    {
+      "epoch": 1.8024291497975709,
+      "grad_norm": 1.1625873389342016,
+      "learning_rate": 7.049123937656855e-06,
+      "loss": 1.2865,
+      "step": 2226
+    },
+    {
+      "epoch": 1.8032388663967611,
+      "grad_norm": 1.1789523515967617,
+      "learning_rate": 7.040912448241881e-06,
+      "loss": 1.3901,
+      "step": 2227
+    },
+    {
+      "epoch": 1.8040485829959514,
+      "grad_norm": 1.1706741638509202,
+      "learning_rate": 7.032703144991071e-06,
+      "loss": 1.3657,
+      "step": 2228
+    },
+    {
+      "epoch": 1.8048582995951417,
+      "grad_norm": 1.1921078298881833,
+      "learning_rate": 7.024496033969432e-06,
+      "loss": 1.3611,
+      "step": 2229
+    },
+    {
+      "epoch": 1.805668016194332,
+      "grad_norm": 1.17097174113593,
+      "learning_rate": 7.016291121240346e-06,
+      "loss": 1.3633,
+      "step": 2230
+    },
+    {
+      "epoch": 1.8064777327935224,
+      "grad_norm": 1.1852035894468138,
+      "learning_rate": 7.0080884128655844e-06,
+      "loss": 1.3807,
+      "step": 2231
+    },
+    {
+      "epoch": 1.8072874493927125,
+      "grad_norm": 1.2725267927716697,
+      "learning_rate": 6.999887914905275e-06,
+      "loss": 1.4361,
+      "step": 2232
+    },
+    {
+      "epoch": 1.808097165991903,
+      "grad_norm": 1.220459422861511,
+      "learning_rate": 6.991689633417922e-06,
+      "loss": 1.3571,
+      "step": 2233
+    },
+    {
+      "epoch": 1.808906882591093,
+      "grad_norm": 1.1788740570003207,
+      "learning_rate": 6.983493574460387e-06,
+      "loss": 1.3274,
+      "step": 2234
+    },
+    {
+      "epoch": 1.8097165991902835,
+      "grad_norm": 1.3416750880261363,
+      "learning_rate": 6.975299744087891e-06,
+      "loss": 1.3581,
+      "step": 2235
+    },
+    {
+      "epoch": 1.8105263157894735,
+      "grad_norm": 1.225592957552821,
+      "learning_rate": 6.967108148354012e-06,
+      "loss": 1.4076,
+      "step": 2236
+    },
+    {
+      "epoch": 1.811336032388664,
+      "grad_norm": 1.2485859196199363,
+      "learning_rate": 6.958918793310669e-06,
+      "loss": 1.434,
+      "step": 2237
+    },
+    {
+      "epoch": 1.8121457489878543,
+      "grad_norm": 1.179048945702912,
+      "learning_rate": 6.950731685008132e-06,
+      "loss": 1.3373,
+      "step": 2238
+    },
+    {
+      "epoch": 1.8129554655870446,
+      "grad_norm": 1.208305356954193,
+      "learning_rate": 6.942546829495014e-06,
+      "loss": 1.3677,
+      "step": 2239
+    },
+    {
+      "epoch": 1.8137651821862348,
+      "grad_norm": 1.179294023466567,
+      "learning_rate": 6.934364232818254e-06,
+      "loss": 1.3397,
+      "step": 2240
+    },
+    {
+      "epoch": 1.814574898785425,
+      "grad_norm": 1.238375865893255,
+      "learning_rate": 6.926183901023134e-06,
+      "loss": 1.3259,
+      "step": 2241
+    },
+    {
+      "epoch": 1.8153846153846154,
+      "grad_norm": 1.2004335795147096,
+      "learning_rate": 6.91800584015325e-06,
+      "loss": 1.3741,
+      "step": 2242
+    },
+    {
+      "epoch": 1.8161943319838056,
+      "grad_norm": 1.2101607717425258,
+      "learning_rate": 6.909830056250527e-06,
+      "loss": 1.3455,
+      "step": 2243
+    },
+    {
+      "epoch": 1.817004048582996,
+      "grad_norm": 1.203030571859603,
+      "learning_rate": 6.901656555355212e-06,
+      "loss": 1.3705,
+      "step": 2244
+    },
+    {
+      "epoch": 1.8178137651821862,
+      "grad_norm": 1.276448418257517,
+      "learning_rate": 6.8934853435058566e-06,
+      "loss": 1.3609,
+      "step": 2245
+    },
+    {
+      "epoch": 1.8186234817813767,
+      "grad_norm": 1.1499994230561823,
+      "learning_rate": 6.8853164267393234e-06,
+      "loss": 1.3902,
+      "step": 2246
+    },
+    {
+      "epoch": 1.8194331983805667,
+      "grad_norm": 1.1943909447731087,
+      "learning_rate": 6.877149811090785e-06,
+      "loss": 1.2936,
+      "step": 2247
+    },
+    {
+      "epoch": 1.8202429149797572,
+      "grad_norm": 1.2276088320117058,
+      "learning_rate": 6.8689855025937124e-06,
+      "loss": 1.3459,
+      "step": 2248
+    },
+    {
+      "epoch": 1.8210526315789473,
+      "grad_norm": 1.2131982314479368,
+      "learning_rate": 6.860823507279868e-06,
+      "loss": 1.3071,
+      "step": 2249
+    },
+    {
+      "epoch": 1.8218623481781377,
+      "grad_norm": 1.2662064967702196,
+      "learning_rate": 6.852663831179303e-06,
+      "loss": 1.3959,
+      "step": 2250
+    },
+    {
+      "epoch": 1.8226720647773278,
+      "grad_norm": 1.28182916895796,
+      "learning_rate": 6.844506480320363e-06,
+      "loss": 1.3637,
+      "step": 2251
+    },
+    {
+      "epoch": 1.8234817813765183,
+      "grad_norm": 1.1875870250303597,
+      "learning_rate": 6.836351460729673e-06,
+      "loss": 1.354,
+      "step": 2252
+    },
+    {
+      "epoch": 1.8242914979757086,
+      "grad_norm": 1.2135144065873424,
+      "learning_rate": 6.828198778432131e-06,
+      "loss": 1.3731,
+      "step": 2253
+    },
+    {
+      "epoch": 1.8251012145748988,
+      "grad_norm": 1.190208517610891,
+      "learning_rate": 6.820048439450913e-06,
+      "loss": 1.3976,
+      "step": 2254
+    },
+    {
+      "epoch": 1.825910931174089,
+      "grad_norm": 1.1613715093942938,
+      "learning_rate": 6.811900449807465e-06,
+      "loss": 1.3521,
+      "step": 2255
+    },
+    {
+      "epoch": 1.8267206477732794,
+      "grad_norm": 1.1600854701348962,
+      "learning_rate": 6.803754815521495e-06,
+      "loss": 1.3905,
+      "step": 2256
+    },
+    {
+      "epoch": 1.8275303643724696,
+      "grad_norm": 1.1683331040020393,
+      "learning_rate": 6.7956115426109695e-06,
+      "loss": 1.3801,
+      "step": 2257
+    },
+    {
+      "epoch": 1.82834008097166,
+      "grad_norm": 1.1228286175250644,
+      "learning_rate": 6.78747063709211e-06,
+      "loss": 1.3634,
+      "step": 2258
+    },
+    {
+      "epoch": 1.8291497975708502,
+      "grad_norm": 1.1651839094767948,
+      "learning_rate": 6.779332104979394e-06,
+      "loss": 1.3268,
+      "step": 2259
+    },
+    {
+      "epoch": 1.8299595141700404,
+      "grad_norm": 1.1999740716361973,
+      "learning_rate": 6.771195952285541e-06,
+      "loss": 1.3391,
+      "step": 2260
+    },
+    {
+      "epoch": 1.830769230769231,
+      "grad_norm": 1.1698993424325916,
+      "learning_rate": 6.763062185021511e-06,
+      "loss": 1.3434,
+      "step": 2261
+    },
+    {
+      "epoch": 1.831578947368421,
+      "grad_norm": 1.197859926690675,
+      "learning_rate": 6.754930809196507e-06,
+      "loss": 1.3116,
+      "step": 2262
+    },
+    {
+      "epoch": 1.8323886639676115,
+      "grad_norm": 1.2075936874086974,
+      "learning_rate": 6.746801830817966e-06,
+      "loss": 1.3516,
+      "step": 2263
+    },
+    {
+      "epoch": 1.8331983805668015,
+      "grad_norm": 1.2280800864883632,
+      "learning_rate": 6.738675255891548e-06,
+      "loss": 1.3995,
+      "step": 2264
+    },
+    {
+      "epoch": 1.834008097165992,
+      "grad_norm": 1.186966643343012,
+      "learning_rate": 6.730551090421137e-06,
+      "loss": 1.3301,
+      "step": 2265
+    },
+    {
+      "epoch": 1.834817813765182,
+      "grad_norm": 1.1752482043139705,
+      "learning_rate": 6.7224293404088445e-06,
+      "loss": 1.288,
+      "step": 2266
+    },
+    {
+      "epoch": 1.8356275303643725,
+      "grad_norm": 1.1737943908874262,
+      "learning_rate": 6.714310011854989e-06,
+      "loss": 1.379,
+      "step": 2267
+    },
+    {
+      "epoch": 1.8364372469635628,
+      "grad_norm": 1.1881877312022122,
+      "learning_rate": 6.7061931107581055e-06,
+      "loss": 1.3928,
+      "step": 2268
+    },
+    {
+      "epoch": 1.837246963562753,
+      "grad_norm": 1.2419072015459927,
+      "learning_rate": 6.698078643114935e-06,
+      "loss": 1.3845,
+      "step": 2269
+    },
+    {
+      "epoch": 1.8380566801619433,
+      "grad_norm": 1.1961721466530393,
+      "learning_rate": 6.689966614920414e-06,
+      "loss": 1.3745,
+      "step": 2270
+    },
+    {
+      "epoch": 1.8388663967611336,
+      "grad_norm": 1.151446689287607,
+      "learning_rate": 6.681857032167689e-06,
+      "loss": 1.3118,
+      "step": 2271
+    },
+    {
+      "epoch": 1.8396761133603239,
+      "grad_norm": 1.196438525357998,
+      "learning_rate": 6.673749900848092e-06,
+      "loss": 1.3183,
+      "step": 2272
+    },
+    {
+      "epoch": 1.8404858299595142,
+      "grad_norm": 1.2018986018410842,
+      "learning_rate": 6.665645226951141e-06,
+      "loss": 1.4007,
+      "step": 2273
+    },
+    {
+      "epoch": 1.8412955465587044,
+      "grad_norm": 1.175635606943612,
+      "learning_rate": 6.657543016464546e-06,
+      "loss": 1.39,
+      "step": 2274
+    },
+    {
+      "epoch": 1.8421052631578947,
+      "grad_norm": 1.2177369319411393,
+      "learning_rate": 6.6494432753741935e-06,
+      "loss": 1.3989,
+      "step": 2275
+    },
+    {
+      "epoch": 1.8429149797570852,
+      "grad_norm": 1.15435997929181,
+      "learning_rate": 6.641346009664142e-06,
+      "loss": 1.3644,
+      "step": 2276
+    },
+    {
+      "epoch": 1.8437246963562752,
+      "grad_norm": 1.1812983300442905,
+      "learning_rate": 6.63325122531663e-06,
+      "loss": 1.3,
+      "step": 2277
+    },
+    {
+      "epoch": 1.8445344129554657,
+      "grad_norm": 1.1786903009596423,
+      "learning_rate": 6.62515892831205e-06,
+      "loss": 1.3513,
+      "step": 2278
+    },
+    {
+      "epoch": 1.8453441295546558,
+      "grad_norm": 1.1533542073236585,
+      "learning_rate": 6.6170691246289744e-06,
+      "loss": 1.3916,
+      "step": 2279
+    },
+    {
+      "epoch": 1.8461538461538463,
+      "grad_norm": 1.122771919065958,
+      "learning_rate": 6.608981820244116e-06,
+      "loss": 1.2917,
+      "step": 2280
+    },
+    {
+      "epoch": 1.8469635627530363,
+      "grad_norm": 1.1565154319571267,
+      "learning_rate": 6.600897021132353e-06,
+      "loss": 1.3448,
+      "step": 2281
+    },
+    {
+      "epoch": 1.8477732793522268,
+      "grad_norm": 1.1994419287429507,
+      "learning_rate": 6.592814733266708e-06,
+      "loss": 1.3973,
+      "step": 2282
+    },
+    {
+      "epoch": 1.8485829959514168,
+      "grad_norm": 1.2145511142569878,
+      "learning_rate": 6.5847349626183444e-06,
+      "loss": 1.2976,
+      "step": 2283
+    },
+    {
+      "epoch": 1.8493927125506073,
+      "grad_norm": 1.2470922439448142,
+      "learning_rate": 6.576657715156576e-06,
+      "loss": 1.4211,
+      "step": 2284
+    },
+    {
+      "epoch": 1.8502024291497976,
+      "grad_norm": 1.1849987198146783,
+      "learning_rate": 6.568582996848844e-06,
+      "loss": 1.4217,
+      "step": 2285
+    },
+    {
+      "epoch": 1.8510121457489879,
+      "grad_norm": 1.2246339182732113,
+      "learning_rate": 6.560510813660719e-06,
+      "loss": 1.347,
+      "step": 2286
+    },
+    {
+      "epoch": 1.8518218623481781,
+      "grad_norm": 1.2511212401796088,
+      "learning_rate": 6.5524411715559125e-06,
+      "loss": 1.4248,
+      "step": 2287
+    },
+    {
+      "epoch": 1.8526315789473684,
+      "grad_norm": 1.2304999920366795,
+      "learning_rate": 6.544374076496243e-06,
+      "loss": 1.3445,
+      "step": 2288
+    },
+    {
+      "epoch": 1.8534412955465587,
+      "grad_norm": 1.2349915243981289,
+      "learning_rate": 6.536309534441658e-06,
+      "loss": 1.3398,
+      "step": 2289
+    },
+    {
+      "epoch": 1.854251012145749,
+      "grad_norm": 1.1884282705027234,
+      "learning_rate": 6.528247551350213e-06,
+      "loss": 1.2971,
+      "step": 2290
+    },
+    {
+      "epoch": 1.8550607287449394,
+      "grad_norm": 1.181801214740644,
+      "learning_rate": 6.5201881331780725e-06,
+      "loss": 1.3792,
+      "step": 2291
+    },
+    {
+      "epoch": 1.8558704453441295,
+      "grad_norm": 1.2135112397667827,
+      "learning_rate": 6.512131285879513e-06,
+      "loss": 1.3898,
+      "step": 2292
+    },
+    {
+      "epoch": 1.85668016194332,
+      "grad_norm": 1.2180427276548653,
+      "learning_rate": 6.504077015406902e-06,
+      "loss": 1.3809,
+      "step": 2293
+    },
+    {
+      "epoch": 1.85748987854251,
+      "grad_norm": 1.1561446887739786,
+      "learning_rate": 6.496025327710707e-06,
+      "loss": 1.3113,
+      "step": 2294
+    },
+    {
+      "epoch": 1.8582995951417005,
+      "grad_norm": 1.2247735536559143,
+      "learning_rate": 6.487976228739493e-06,
+      "loss": 1.3531,
+      "step": 2295
+    },
+    {
+      "epoch": 1.8591093117408906,
+      "grad_norm": 1.2152586819122253,
+      "learning_rate": 6.4799297244399085e-06,
+      "loss": 1.3964,
+      "step": 2296
+    },
+    {
+      "epoch": 1.859919028340081,
+      "grad_norm": 1.2454451424496096,
+      "learning_rate": 6.471885820756683e-06,
+      "loss": 1.3904,
+      "step": 2297
+    },
+    {
+      "epoch": 1.860728744939271,
+      "grad_norm": 1.1655322922036064,
+      "learning_rate": 6.463844523632622e-06,
+      "loss": 1.3364,
+      "step": 2298
+    },
+    {
+      "epoch": 1.8615384615384616,
+      "grad_norm": 1.2996377865771735,
+      "learning_rate": 6.455805839008615e-06,
+      "loss": 1.4075,
+      "step": 2299
+    },
+    {
+      "epoch": 1.8623481781376519,
+      "grad_norm": 1.1964728670717037,
+      "learning_rate": 6.4477697728236146e-06,
+      "loss": 1.3517,
+      "step": 2300
+    },
+    {
+      "epoch": 1.8631578947368421,
+      "grad_norm": 1.16799096604399,
+      "learning_rate": 6.439736331014637e-06,
+      "loss": 1.3809,
+      "step": 2301
+    },
+    {
+      "epoch": 1.8639676113360324,
+      "grad_norm": 1.2252233979220486,
+      "learning_rate": 6.431705519516763e-06,
+      "loss": 1.3675,
+      "step": 2302
+    },
+    {
+      "epoch": 1.8647773279352227,
+      "grad_norm": 1.2104500576735182,
+      "learning_rate": 6.4236773442631325e-06,
+      "loss": 1.3805,
+      "step": 2303
+    },
+    {
+      "epoch": 1.865587044534413,
+      "grad_norm": 1.3193178213753143,
+      "learning_rate": 6.415651811184935e-06,
+      "loss": 1.3997,
+      "step": 2304
+    },
+    {
+      "epoch": 1.8663967611336032,
+      "grad_norm": 1.2018055746906207,
+      "learning_rate": 6.407628926211409e-06,
+      "loss": 1.3964,
+      "step": 2305
+    },
+    {
+      "epoch": 1.8672064777327935,
+      "grad_norm": 1.174760468858027,
+      "learning_rate": 6.39960869526983e-06,
+      "loss": 1.3464,
+      "step": 2306
+    },
+    {
+      "epoch": 1.8680161943319837,
+      "grad_norm": 1.1810773457338066,
+      "learning_rate": 6.391591124285524e-06,
+      "loss": 1.2858,
+      "step": 2307
+    },
+    {
+      "epoch": 1.8688259109311742,
+      "grad_norm": 1.2262009022694145,
+      "learning_rate": 6.383576219181844e-06,
+      "loss": 1.3559,
+      "step": 2308
+    },
+    {
+      "epoch": 1.8696356275303643,
+      "grad_norm": 1.2773594939431656,
+      "learning_rate": 6.375563985880174e-06,
+      "loss": 1.3705,
+      "step": 2309
+    },
+    {
+      "epoch": 1.8704453441295548,
+      "grad_norm": 1.2208108751520061,
+      "learning_rate": 6.367554430299924e-06,
+      "loss": 1.3333,
+      "step": 2310
+    },
+    {
+      "epoch": 1.8712550607287448,
+      "grad_norm": 1.4160184181629654,
+      "learning_rate": 6.3595475583585344e-06,
+      "loss": 1.3816,
+      "step": 2311
+    },
+    {
+      "epoch": 1.8720647773279353,
+      "grad_norm": 1.2094735305859114,
+      "learning_rate": 6.351543375971453e-06,
+      "loss": 1.3275,
+      "step": 2312
+    },
+    {
+      "epoch": 1.8728744939271254,
+      "grad_norm": 1.2855800361668372,
+      "learning_rate": 6.34354188905214e-06,
+      "loss": 1.4212,
+      "step": 2313
+    },
+    {
+      "epoch": 1.8736842105263158,
+      "grad_norm": 1.1929901078467973,
+      "learning_rate": 6.335543103512072e-06,
+      "loss": 1.368,
+      "step": 2314
+    },
+    {
+      "epoch": 1.874493927125506,
+      "grad_norm": 1.1975185842016003,
+      "learning_rate": 6.327547025260723e-06,
+      "loss": 1.3372,
+      "step": 2315
+    },
+    {
+      "epoch": 1.8753036437246964,
+      "grad_norm": 1.2685945017090825,
+      "learning_rate": 6.319553660205569e-06,
+      "loss": 1.3461,
+      "step": 2316
+    },
+    {
+      "epoch": 1.8761133603238866,
+      "grad_norm": 1.1962632720706978,
+      "learning_rate": 6.3115630142520835e-06,
+      "loss": 1.404,
+      "step": 2317
+    },
+    {
+      "epoch": 1.876923076923077,
+      "grad_norm": 1.198552614864501,
+      "learning_rate": 6.303575093303725e-06,
+      "loss": 1.4111,
+      "step": 2318
+    },
+    {
+      "epoch": 1.8777327935222672,
+      "grad_norm": 1.1891573003084546,
+      "learning_rate": 6.2955899032619515e-06,
+      "loss": 1.3048,
+      "step": 2319
+    },
+    {
+      "epoch": 1.8785425101214575,
+      "grad_norm": 1.1495642468155007,
+      "learning_rate": 6.287607450026189e-06,
+      "loss": 1.4159,
+      "step": 2320
+    },
+    {
+      "epoch": 1.8793522267206477,
+      "grad_norm": 1.1707115053399164,
+      "learning_rate": 6.27962773949385e-06,
+      "loss": 1.4223,
+      "step": 2321
+    },
+    {
+      "epoch": 1.880161943319838,
+      "grad_norm": 1.2148777682926812,
+      "learning_rate": 6.271650777560318e-06,
+      "loss": 1.3416,
+      "step": 2322
+    },
+    {
+      "epoch": 1.8809716599190285,
+      "grad_norm": 1.222151610467891,
+      "learning_rate": 6.263676570118948e-06,
+      "loss": 1.3497,
+      "step": 2323
+    },
+    {
+      "epoch": 1.8817813765182185,
+      "grad_norm": 1.1961853962501905,
+      "learning_rate": 6.2557051230610534e-06,
+      "loss": 1.3338,
+      "step": 2324
+    },
+    {
+      "epoch": 1.882591093117409,
+      "grad_norm": 1.1926637799351172,
+      "learning_rate": 6.247736442275918e-06,
+      "loss": 1.4051,
+      "step": 2325
+    },
+    {
+      "epoch": 1.883400809716599,
+      "grad_norm": 1.1918108744163147,
+      "learning_rate": 6.239770533650771e-06,
+      "loss": 1.398,
+      "step": 2326
+    },
+    {
+      "epoch": 1.8842105263157896,
+      "grad_norm": 1.234775751606103,
+      "learning_rate": 6.231807403070806e-06,
+      "loss": 1.3671,
+      "step": 2327
+    },
+    {
+      "epoch": 1.8850202429149796,
+      "grad_norm": 1.2099918714991094,
+      "learning_rate": 6.223847056419154e-06,
+      "loss": 1.3623,
+      "step": 2328
+    },
+    {
+      "epoch": 1.88582995951417,
+      "grad_norm": 1.2291790207100255,
+      "learning_rate": 6.215889499576898e-06,
+      "loss": 1.3359,
+      "step": 2329
+    },
+    {
+      "epoch": 1.8866396761133604,
+      "grad_norm": 1.2504678829644937,
+      "learning_rate": 6.2079347384230505e-06,
+      "loss": 1.3474,
+      "step": 2330
+    },
+    {
+      "epoch": 1.8874493927125506,
+      "grad_norm": 1.1571650257183987,
+      "learning_rate": 6.199982778834561e-06,
+      "loss": 1.3704,
+      "step": 2331
+    },
+    {
+      "epoch": 1.888259109311741,
+      "grad_norm": 1.192980883967285,
+      "learning_rate": 6.192033626686316e-06,
+      "loss": 1.3893,
+      "step": 2332
+    },
+    {
+      "epoch": 1.8890688259109312,
+      "grad_norm": 1.2365704933508046,
+      "learning_rate": 6.1840872878511215e-06,
+      "loss": 1.3739,
+      "step": 2333
+    },
+    {
+      "epoch": 1.8898785425101214,
+      "grad_norm": 1.2430427056052022,
+      "learning_rate": 6.1761437681997e-06,
+      "loss": 1.3773,
+      "step": 2334
+    },
+    {
+      "epoch": 1.8906882591093117,
+      "grad_norm": 1.2140264228246762,
+      "learning_rate": 6.168203073600706e-06,
+      "loss": 1.3265,
+      "step": 2335
+    },
+    {
+      "epoch": 1.891497975708502,
+      "grad_norm": 1.1949388259542468,
+      "learning_rate": 6.160265209920698e-06,
+      "loss": 1.4013,
+      "step": 2336
+    },
+    {
+      "epoch": 1.8923076923076922,
+      "grad_norm": 1.2825641807955082,
+      "learning_rate": 6.152330183024142e-06,
+      "loss": 1.3582,
+      "step": 2337
+    },
+    {
+      "epoch": 1.8931174089068827,
+      "grad_norm": 1.1900913534174808,
+      "learning_rate": 6.1443979987734086e-06,
+      "loss": 1.3772,
+      "step": 2338
+    },
+    {
+      "epoch": 1.8939271255060728,
+      "grad_norm": 1.2005697884400928,
+      "learning_rate": 6.1364686630287694e-06,
+      "loss": 1.3625,
+      "step": 2339
+    },
+    {
+      "epoch": 1.8947368421052633,
+      "grad_norm": 1.2630270064813116,
+      "learning_rate": 6.128542181648395e-06,
+      "loss": 1.3342,
+      "step": 2340
+    },
+    {
+      "epoch": 1.8955465587044533,
+      "grad_norm": 1.2002800976397625,
+      "learning_rate": 6.120618560488341e-06,
+      "loss": 1.3267,
+      "step": 2341
+    },
+    {
+      "epoch": 1.8963562753036438,
+      "grad_norm": 1.215442464121004,
+      "learning_rate": 6.112697805402548e-06,
+      "loss": 1.3688,
+      "step": 2342
+    },
+    {
+      "epoch": 1.8971659919028339,
+      "grad_norm": 1.2288864318659574,
+      "learning_rate": 6.104779922242851e-06,
+      "loss": 1.3636,
+      "step": 2343
+    },
+    {
+      "epoch": 1.8979757085020244,
+      "grad_norm": 1.194317008155087,
+      "learning_rate": 6.096864916858957e-06,
+      "loss": 1.32,
+      "step": 2344
+    },
+    {
+      "epoch": 1.8987854251012146,
+      "grad_norm": 1.2124723409709166,
+      "learning_rate": 6.088952795098442e-06,
+      "loss": 1.3127,
+      "step": 2345
+    },
+    {
+      "epoch": 1.8995951417004049,
+      "grad_norm": 1.188500019539976,
+      "learning_rate": 6.081043562806754e-06,
+      "loss": 1.388,
+      "step": 2346
+    },
+    {
+      "epoch": 1.9004048582995952,
+      "grad_norm": 1.1552719513756389,
+      "learning_rate": 6.073137225827213e-06,
+      "loss": 1.3228,
+      "step": 2347
+    },
+    {
+      "epoch": 1.9012145748987854,
+      "grad_norm": 1.2237471760776777,
+      "learning_rate": 6.065233790000993e-06,
+      "loss": 1.3489,
+      "step": 2348
+    },
+    {
+      "epoch": 1.9020242914979757,
+      "grad_norm": 1.1792800865204056,
+      "learning_rate": 6.057333261167122e-06,
+      "loss": 1.3479,
+      "step": 2349
+    },
+    {
+      "epoch": 1.902834008097166,
+      "grad_norm": 1.199559285857203,
+      "learning_rate": 6.049435645162487e-06,
+      "loss": 1.3519,
+      "step": 2350
+    },
+    {
+      "epoch": 1.9036437246963562,
+      "grad_norm": 1.181957404617024,
+      "learning_rate": 6.041540947821827e-06,
+      "loss": 1.3371,
+      "step": 2351
+    },
+    {
+      "epoch": 1.9044534412955465,
+      "grad_norm": 1.2505276548541542,
+      "learning_rate": 6.0336491749777115e-06,
+      "loss": 1.4061,
+      "step": 2352
+    },
+    {
+      "epoch": 1.905263157894737,
+      "grad_norm": 1.195960955129131,
+      "learning_rate": 6.025760332460558e-06,
+      "loss": 1.3586,
+      "step": 2353
+    },
+    {
+      "epoch": 1.906072874493927,
+      "grad_norm": 1.1707714015115926,
+      "learning_rate": 6.01787442609862e-06,
+      "loss": 1.3412,
+      "step": 2354
+    },
+    {
+      "epoch": 1.9068825910931175,
+      "grad_norm": 1.140910772316072,
+      "learning_rate": 6.009991461717977e-06,
+      "loss": 1.3314,
+      "step": 2355
+    },
+    {
+      "epoch": 1.9076923076923076,
+      "grad_norm": 1.23349460378788,
+      "learning_rate": 6.002111445142533e-06,
+      "loss": 1.3255,
+      "step": 2356
+    },
+    {
+      "epoch": 1.908502024291498,
+      "grad_norm": 1.2462129342380746,
+      "learning_rate": 5.994234382194026e-06,
+      "loss": 1.3989,
+      "step": 2357
+    },
+    {
+      "epoch": 1.9093117408906881,
+      "grad_norm": 1.2688670795842787,
+      "learning_rate": 5.986360278691998e-06,
+      "loss": 1.3718,
+      "step": 2358
+    },
+    {
+      "epoch": 1.9101214574898786,
+      "grad_norm": 1.2095702791808156,
+      "learning_rate": 5.978489140453817e-06,
+      "loss": 1.3534,
+      "step": 2359
+    },
+    {
+      "epoch": 1.9109311740890689,
+      "grad_norm": 1.1982546045786988,
+      "learning_rate": 5.9706209732946495e-06,
+      "loss": 1.3671,
+      "step": 2360
+    },
+    {
+      "epoch": 1.9117408906882591,
+      "grad_norm": 1.1822667381374636,
+      "learning_rate": 5.962755783027473e-06,
+      "loss": 1.2912,
+      "step": 2361
+    },
+    {
+      "epoch": 1.9125506072874494,
+      "grad_norm": 1.242784856625613,
+      "learning_rate": 5.954893575463064e-06,
+      "loss": 1.369,
+      "step": 2362
+    },
+    {
+      "epoch": 1.9133603238866397,
+      "grad_norm": 1.2344835123108033,
+      "learning_rate": 5.9470343564099975e-06,
+      "loss": 1.3867,
+      "step": 2363
+    },
+    {
+      "epoch": 1.91417004048583,
+      "grad_norm": 1.2420930503800902,
+      "learning_rate": 5.939178131674633e-06,
+      "loss": 1.3344,
+      "step": 2364
+    },
+    {
+      "epoch": 1.9149797570850202,
+      "grad_norm": 1.2002007408254076,
+      "learning_rate": 5.931324907061131e-06,
+      "loss": 1.4089,
+      "step": 2365
+    },
+    {
+      "epoch": 1.9157894736842105,
+      "grad_norm": 1.1431280050371697,
+      "learning_rate": 5.92347468837142e-06,
+      "loss": 1.3774,
+      "step": 2366
+    },
+    {
+      "epoch": 1.9165991902834008,
+      "grad_norm": 1.2176788699652055,
+      "learning_rate": 5.915627481405224e-06,
+      "loss": 1.3993,
+      "step": 2367
+    },
+    {
+      "epoch": 1.9174089068825912,
+      "grad_norm": 1.1570899200524465,
+      "learning_rate": 5.907783291960027e-06,
+      "loss": 1.2658,
+      "step": 2368
+    },
+    {
+      "epoch": 1.9182186234817813,
+      "grad_norm": 1.2498879079115792,
+      "learning_rate": 5.899942125831097e-06,
+      "loss": 1.286,
+      "step": 2369
+    },
+    {
+      "epoch": 1.9190283400809718,
+      "grad_norm": 1.2146345067146647,
+      "learning_rate": 5.892103988811457e-06,
+      "loss": 1.4007,
+      "step": 2370
+    },
+    {
+      "epoch": 1.9198380566801618,
+      "grad_norm": 1.2400136674029751,
+      "learning_rate": 5.884268886691898e-06,
+      "loss": 1.3853,
+      "step": 2371
+    },
+    {
+      "epoch": 1.9206477732793523,
+      "grad_norm": 1.2266043645344005,
+      "learning_rate": 5.876436825260967e-06,
+      "loss": 1.3606,
+      "step": 2372
+    },
+    {
+      "epoch": 1.9214574898785424,
+      "grad_norm": 1.1901114955712122,
+      "learning_rate": 5.868607810304967e-06,
+      "loss": 1.4107,
+      "step": 2373
+    },
+    {
+      "epoch": 1.9222672064777329,
+      "grad_norm": 1.2389492781589038,
+      "learning_rate": 5.860781847607943e-06,
+      "loss": 1.3883,
+      "step": 2374
+    },
+    {
+      "epoch": 1.9230769230769231,
+      "grad_norm": 1.223439005294772,
+      "learning_rate": 5.852958942951701e-06,
+      "loss": 1.3636,
+      "step": 2375
+    },
+    {
+      "epoch": 1.9238866396761134,
+      "grad_norm": 1.1896039830081746,
+      "learning_rate": 5.845139102115769e-06,
+      "loss": 1.3259,
+      "step": 2376
+    },
+    {
+      "epoch": 1.9246963562753037,
+      "grad_norm": 1.2049235269890495,
+      "learning_rate": 5.837322330877421e-06,
+      "loss": 1.3567,
+      "step": 2377
+    },
+    {
+      "epoch": 1.925506072874494,
+      "grad_norm": 1.2157467237161108,
+      "learning_rate": 5.829508635011667e-06,
+      "loss": 1.3588,
+      "step": 2378
+    },
+    {
+      "epoch": 1.9263157894736842,
+      "grad_norm": 1.248188375178168,
+      "learning_rate": 5.821698020291234e-06,
+      "loss": 1.3572,
+      "step": 2379
+    },
+    {
+      "epoch": 1.9271255060728745,
+      "grad_norm": 1.212346730632553,
+      "learning_rate": 5.8138904924865766e-06,
+      "loss": 1.3043,
+      "step": 2380
+    },
+    {
+      "epoch": 1.9279352226720647,
+      "grad_norm": 1.1824421069183566,
+      "learning_rate": 5.806086057365878e-06,
+      "loss": 1.3943,
+      "step": 2381
+    },
+    {
+      "epoch": 1.928744939271255,
+      "grad_norm": 1.1763429489984647,
+      "learning_rate": 5.798284720695022e-06,
+      "loss": 1.365,
+      "step": 2382
+    },
+    {
+      "epoch": 1.9295546558704455,
+      "grad_norm": 1.204932841462832,
+      "learning_rate": 5.790486488237619e-06,
+      "loss": 1.3377,
+      "step": 2383
+    },
+    {
+      "epoch": 1.9303643724696355,
+      "grad_norm": 1.2045578783510957,
+      "learning_rate": 5.782691365754971e-06,
+      "loss": 1.4084,
+      "step": 2384
+    },
+    {
+      "epoch": 1.931174089068826,
+      "grad_norm": 1.2608711906516323,
+      "learning_rate": 5.774899359006092e-06,
+      "loss": 1.2661,
+      "step": 2385
+    },
+    {
+      "epoch": 1.931983805668016,
+      "grad_norm": 1.1990033949590717,
+      "learning_rate": 5.76711047374769e-06,
+      "loss": 1.3528,
+      "step": 2386
+    },
+    {
+      "epoch": 1.9327935222672066,
+      "grad_norm": 1.172948011576386,
+      "learning_rate": 5.759324715734166e-06,
+      "loss": 1.4189,
+      "step": 2387
+    },
+    {
+      "epoch": 1.9336032388663966,
+      "grad_norm": 1.241926186312199,
+      "learning_rate": 5.7515420907176105e-06,
+      "loss": 1.4028,
+      "step": 2388
+    },
+    {
+      "epoch": 1.9344129554655871,
+      "grad_norm": 1.1964066986025503,
+      "learning_rate": 5.743762604447809e-06,
+      "loss": 1.3317,
+      "step": 2389
+    },
+    {
+      "epoch": 1.9352226720647774,
+      "grad_norm": 1.1945494693033618,
+      "learning_rate": 5.735986262672211e-06,
+      "loss": 1.341,
+      "step": 2390
+    },
+    {
+      "epoch": 1.9360323886639677,
+      "grad_norm": 1.1755268375572105,
+      "learning_rate": 5.728213071135962e-06,
+      "loss": 1.3911,
+      "step": 2391
+    },
+    {
+      "epoch": 1.936842105263158,
+      "grad_norm": 1.1804944891484381,
+      "learning_rate": 5.720443035581867e-06,
+      "loss": 1.3697,
+      "step": 2392
+    },
+    {
+      "epoch": 1.9376518218623482,
+      "grad_norm": 1.2046317934373794,
+      "learning_rate": 5.712676161750399e-06,
+      "loss": 1.4039,
+      "step": 2393
+    },
+    {
+      "epoch": 1.9384615384615385,
+      "grad_norm": 1.1637358644282172,
+      "learning_rate": 5.704912455379703e-06,
+      "loss": 1.363,
+      "step": 2394
+    },
+    {
+      "epoch": 1.9392712550607287,
+      "grad_norm": 1.1974299223091525,
+      "learning_rate": 5.697151922205575e-06,
+      "loss": 1.3971,
+      "step": 2395
+    },
+    {
+      "epoch": 1.940080971659919,
+      "grad_norm": 1.157981186836544,
+      "learning_rate": 5.689394567961477e-06,
+      "loss": 1.2967,
+      "step": 2396
+    },
+    {
+      "epoch": 1.9408906882591093,
+      "grad_norm": 1.1670968402428008,
+      "learning_rate": 5.681640398378514e-06,
+      "loss": 1.3849,
+      "step": 2397
+    },
+    {
+      "epoch": 1.9417004048582998,
+      "grad_norm": 1.1292919479938222,
+      "learning_rate": 5.673889419185439e-06,
+      "loss": 1.4253,
+      "step": 2398
+    },
+    {
+      "epoch": 1.9425101214574898,
+      "grad_norm": 1.213702805085859,
+      "learning_rate": 5.666141636108655e-06,
+      "loss": 1.3198,
+      "step": 2399
+    },
+    {
+      "epoch": 1.9433198380566803,
+      "grad_norm": 1.157847450719407,
+      "learning_rate": 5.658397054872197e-06,
+      "loss": 1.3175,
+      "step": 2400
+    },
+    {
+      "epoch": 1.9441295546558703,
+      "grad_norm": 1.2106036490901373,
+      "learning_rate": 5.650655681197734e-06,
+      "loss": 1.4105,
+      "step": 2401
+    },
+    {
+      "epoch": 1.9449392712550608,
+      "grad_norm": 1.219262613527522,
+      "learning_rate": 5.642917520804569e-06,
+      "loss": 1.3505,
+      "step": 2402
+    },
+    {
+      "epoch": 1.9457489878542509,
+      "grad_norm": 1.1742978362533214,
+      "learning_rate": 5.635182579409626e-06,
+      "loss": 1.299,
+      "step": 2403
+    },
+    {
+      "epoch": 1.9465587044534414,
+      "grad_norm": 1.2293963021779448,
+      "learning_rate": 5.627450862727461e-06,
+      "loss": 1.3391,
+      "step": 2404
+    },
+    {
+      "epoch": 1.9473684210526314,
+      "grad_norm": 1.2034045616590177,
+      "learning_rate": 5.619722376470238e-06,
+      "loss": 1.3877,
+      "step": 2405
+    },
+    {
+      "epoch": 1.948178137651822,
+      "grad_norm": 1.2272183799034553,
+      "learning_rate": 5.611997126347732e-06,
+      "loss": 1.3988,
+      "step": 2406
+    },
+    {
+      "epoch": 1.9489878542510122,
+      "grad_norm": 1.2695110730414088,
+      "learning_rate": 5.604275118067341e-06,
+      "loss": 1.3653,
+      "step": 2407
+    },
+    {
+      "epoch": 1.9497975708502024,
+      "grad_norm": 1.248909293580144,
+      "learning_rate": 5.596556357334056e-06,
+      "loss": 1.3562,
+      "step": 2408
+    },
+    {
+      "epoch": 1.9506072874493927,
+      "grad_norm": 1.2131088122427793,
+      "learning_rate": 5.588840849850472e-06,
+      "loss": 1.3928,
+      "step": 2409
+    },
+    {
+      "epoch": 1.951417004048583,
+      "grad_norm": 1.1900481644176077,
+      "learning_rate": 5.581128601316774e-06,
+      "loss": 1.3777,
+      "step": 2410
+    },
+    {
+      "epoch": 1.9522267206477733,
+      "grad_norm": 1.2099498350109,
+      "learning_rate": 5.573419617430758e-06,
+      "loss": 1.3499,
+      "step": 2411
+    },
+    {
+      "epoch": 1.9530364372469635,
+      "grad_norm": 1.217148269565451,
+      "learning_rate": 5.565713903887788e-06,
+      "loss": 1.2877,
+      "step": 2412
+    },
+    {
+      "epoch": 1.953846153846154,
+      "grad_norm": 1.2490893741382743,
+      "learning_rate": 5.558011466380824e-06,
+      "loss": 1.3768,
+      "step": 2413
+    },
+    {
+      "epoch": 1.954655870445344,
+      "grad_norm": 1.240461136852775,
+      "learning_rate": 5.550312310600394e-06,
+      "loss": 1.4229,
+      "step": 2414
+    },
+    {
+      "epoch": 1.9554655870445345,
+      "grad_norm": 1.2295726427309241,
+      "learning_rate": 5.542616442234618e-06,
+      "loss": 1.3347,
+      "step": 2415
+    },
+    {
+      "epoch": 1.9562753036437246,
+      "grad_norm": 1.2008815138181963,
+      "learning_rate": 5.534923866969175e-06,
+      "loss": 1.4176,
+      "step": 2416
+    },
+    {
+      "epoch": 1.957085020242915,
+      "grad_norm": 1.1676560157055105,
+      "learning_rate": 5.527234590487314e-06,
+      "loss": 1.3314,
+      "step": 2417
+    },
+    {
+      "epoch": 1.9578947368421051,
+      "grad_norm": 1.1701886277830924,
+      "learning_rate": 5.5195486184698435e-06,
+      "loss": 1.363,
+      "step": 2418
+    },
+    {
+      "epoch": 1.9587044534412956,
+      "grad_norm": 1.1778230913774466,
+      "learning_rate": 5.511865956595142e-06,
+      "loss": 1.3683,
+      "step": 2419
+    },
+    {
+      "epoch": 1.9595141700404857,
+      "grad_norm": 1.173338300441953,
+      "learning_rate": 5.504186610539131e-06,
+      "loss": 1.3393,
+      "step": 2420
+    },
+    {
+      "epoch": 1.9603238866396762,
+      "grad_norm": 1.1767463019339799,
+      "learning_rate": 5.496510585975285e-06,
+      "loss": 1.3235,
+      "step": 2421
+    },
+    {
+      "epoch": 1.9611336032388664,
+      "grad_norm": 1.2644216036741416,
+      "learning_rate": 5.488837888574623e-06,
+      "loss": 1.3373,
+      "step": 2422
+    },
+    {
+      "epoch": 1.9619433198380567,
+      "grad_norm": 1.2103895583574469,
+      "learning_rate": 5.4811685240057165e-06,
+      "loss": 1.3983,
+      "step": 2423
+    },
+    {
+      "epoch": 1.962753036437247,
+      "grad_norm": 1.1846356504436384,
+      "learning_rate": 5.473502497934663e-06,
+      "loss": 1.3294,
+      "step": 2424
+    },
+    {
+      "epoch": 1.9635627530364372,
+      "grad_norm": 1.2235189988456545,
+      "learning_rate": 5.465839816025093e-06,
+      "loss": 1.3276,
+      "step": 2425
+    },
+    {
+      "epoch": 1.9643724696356275,
+      "grad_norm": 1.2314816730763305,
+      "learning_rate": 5.458180483938179e-06,
+      "loss": 1.4498,
+      "step": 2426
+    },
+    {
+      "epoch": 1.9651821862348178,
+      "grad_norm": 1.212596357115965,
+      "learning_rate": 5.450524507332606e-06,
+      "loss": 1.3656,
+      "step": 2427
+    },
+    {
+      "epoch": 1.965991902834008,
+      "grad_norm": 1.1639800229825885,
+      "learning_rate": 5.442871891864585e-06,
+      "loss": 1.3776,
+      "step": 2428
+    },
+    {
+      "epoch": 1.9668016194331983,
+      "grad_norm": 1.1367690723379495,
+      "learning_rate": 5.435222643187843e-06,
+      "loss": 1.3242,
+      "step": 2429
+    },
+    {
+      "epoch": 1.9676113360323888,
+      "grad_norm": 1.2165886607110268,
+      "learning_rate": 5.427576766953615e-06,
+      "loss": 1.3591,
+      "step": 2430
+    },
+    {
+      "epoch": 1.9684210526315788,
+      "grad_norm": 1.1790081897002724,
+      "learning_rate": 5.419934268810659e-06,
+      "loss": 1.3848,
+      "step": 2431
+    },
+    {
+      "epoch": 1.9692307692307693,
+      "grad_norm": 1.2315181532137929,
+      "learning_rate": 5.412295154405217e-06,
+      "loss": 1.3431,
+      "step": 2432
+    },
+    {
+      "epoch": 1.9700404858299594,
+      "grad_norm": 1.2728808844442372,
+      "learning_rate": 5.4046594293810515e-06,
+      "loss": 1.3639,
+      "step": 2433
+    },
+    {
+      "epoch": 1.9708502024291499,
+      "grad_norm": 1.232337828546579,
+      "learning_rate": 5.397027099379406e-06,
+      "loss": 1.3697,
+      "step": 2434
+    },
+    {
+      "epoch": 1.97165991902834,
+      "grad_norm": 1.243179350091489,
+      "learning_rate": 5.3893981700390215e-06,
+      "loss": 1.3252,
+      "step": 2435
+    },
+    {
+      "epoch": 1.9724696356275304,
+      "grad_norm": 1.2371857762137026,
+      "learning_rate": 5.381772646996128e-06,
+      "loss": 1.3827,
+      "step": 2436
+    },
+    {
+      "epoch": 1.9732793522267207,
+      "grad_norm": 1.288958547364644,
+      "learning_rate": 5.374150535884433e-06,
+      "loss": 1.3265,
+      "step": 2437
+    },
+    {
+      "epoch": 1.974089068825911,
+      "grad_norm": 1.2475048960238355,
+      "learning_rate": 5.3665318423351255e-06,
+      "loss": 1.2822,
+      "step": 2438
+    },
+    {
+      "epoch": 1.9748987854251012,
+      "grad_norm": 1.1986060107719656,
+      "learning_rate": 5.358916571976878e-06,
+      "loss": 1.3558,
+      "step": 2439
+    },
+    {
+      "epoch": 1.9757085020242915,
+      "grad_norm": 1.1809389390768783,
+      "learning_rate": 5.35130473043582e-06,
+      "loss": 1.3428,
+      "step": 2440
+    },
+    {
+      "epoch": 1.9765182186234818,
+      "grad_norm": 1.1854316658221693,
+      "learning_rate": 5.343696323335564e-06,
+      "loss": 1.4093,
+      "step": 2441
+    },
+    {
+      "epoch": 1.977327935222672,
+      "grad_norm": 1.211156343447039,
+      "learning_rate": 5.336091356297168e-06,
+      "loss": 1.2688,
+      "step": 2442
+    },
+    {
+      "epoch": 1.9781376518218623,
+      "grad_norm": 1.2978073337950151,
+      "learning_rate": 5.328489834939162e-06,
+      "loss": 1.3924,
+      "step": 2443
+    },
+    {
+      "epoch": 1.9789473684210526,
+      "grad_norm": 1.1739628299787876,
+      "learning_rate": 5.320891764877522e-06,
+      "loss": 1.371,
+      "step": 2444
+    },
+    {
+      "epoch": 1.979757085020243,
+      "grad_norm": 1.198921849736746,
+      "learning_rate": 5.313297151725679e-06,
+      "loss": 1.3149,
+      "step": 2445
+    },
+    {
+      "epoch": 1.980566801619433,
+      "grad_norm": 1.2012656881257282,
+      "learning_rate": 5.305706001094504e-06,
+      "loss": 1.3979,
+      "step": 2446
+    },
+    {
+      "epoch": 1.9813765182186236,
+      "grad_norm": 1.1827392393082905,
+      "learning_rate": 5.298118318592316e-06,
+      "loss": 1.3565,
+      "step": 2447
+    },
+    {
+      "epoch": 1.9821862348178136,
+      "grad_norm": 1.2628168108725613,
+      "learning_rate": 5.290534109824875e-06,
+      "loss": 1.3705,
+      "step": 2448
+    },
+    {
+      "epoch": 1.9829959514170041,
+      "grad_norm": 1.248756806447712,
+      "learning_rate": 5.282953380395366e-06,
+      "loss": 1.2367,
+      "step": 2449
+    },
+    {
+      "epoch": 1.9838056680161942,
+      "grad_norm": 1.2171206230007487,
+      "learning_rate": 5.275376135904408e-06,
+      "loss": 1.3403,
+      "step": 2450
+    },
+    {
+      "epoch": 1.9846153846153847,
+      "grad_norm": 1.289234216018497,
+      "learning_rate": 5.267802381950042e-06,
+      "loss": 1.3819,
+      "step": 2451
+    },
+    {
+      "epoch": 1.985425101214575,
+      "grad_norm": 1.2241063424118177,
+      "learning_rate": 5.260232124127734e-06,
+      "loss": 1.3171,
+      "step": 2452
+    },
+    {
+      "epoch": 1.9862348178137652,
+      "grad_norm": 1.2256066411153002,
+      "learning_rate": 5.252665368030362e-06,
+      "loss": 1.4032,
+      "step": 2453
+    },
+    {
+      "epoch": 1.9870445344129555,
+      "grad_norm": 1.2295354974621036,
+      "learning_rate": 5.245102119248227e-06,
+      "loss": 1.3562,
+      "step": 2454
+    },
+    {
+      "epoch": 1.9878542510121457,
+      "grad_norm": 1.207357620526114,
+      "learning_rate": 5.2375423833690255e-06,
+      "loss": 1.3962,
+      "step": 2455
+    },
+    {
+      "epoch": 1.988663967611336,
+      "grad_norm": 1.2123767685021807,
+      "learning_rate": 5.229986165977874e-06,
+      "loss": 1.3524,
+      "step": 2456
+    },
+    {
+      "epoch": 1.9894736842105263,
+      "grad_norm": 1.1985170410291037,
+      "learning_rate": 5.222433472657276e-06,
+      "loss": 1.342,
+      "step": 2457
+    },
+    {
+      "epoch": 1.9902834008097166,
+      "grad_norm": 1.2549278916904512,
+      "learning_rate": 5.214884308987136e-06,
+      "loss": 1.2894,
+      "step": 2458
+    },
+    {
+      "epoch": 1.9910931174089068,
+      "grad_norm": 1.2721385349454946,
+      "learning_rate": 5.207338680544754e-06,
+      "loss": 1.3614,
+      "step": 2459
+    },
+    {
+      "epoch": 1.9919028340080973,
+      "grad_norm": 1.1730908013315933,
+      "learning_rate": 5.1997965929048125e-06,
+      "loss": 1.3449,
+      "step": 2460
+    },
+    {
+      "epoch": 1.9927125506072874,
+      "grad_norm": 1.237418271403445,
+      "learning_rate": 5.192258051639378e-06,
+      "loss": 1.3128,
+      "step": 2461
+    },
+    {
+      "epoch": 1.9935222672064778,
+      "grad_norm": 1.195751862323348,
+      "learning_rate": 5.184723062317905e-06,
+      "loss": 1.3297,
+      "step": 2462
+    },
+    {
+      "epoch": 1.994331983805668,
+      "grad_norm": 1.1842709296523235,
+      "learning_rate": 5.177191630507221e-06,
+      "loss": 1.3214,
+      "step": 2463
+    },
+    {
+      "epoch": 1.9951417004048584,
+      "grad_norm": 1.1803392844860812,
+      "learning_rate": 5.169663761771522e-06,
+      "loss": 1.415,
+      "step": 2464
+    },
+    {
+      "epoch": 1.9959514170040484,
+      "grad_norm": 1.1998108340136762,
+      "learning_rate": 5.1621394616723705e-06,
+      "loss": 1.3449,
+      "step": 2465
+    },
+    {
+      "epoch": 1.996761133603239,
+      "grad_norm": 1.2193208223683405,
+      "learning_rate": 5.154618735768695e-06,
+      "loss": 1.2948,
+      "step": 2466
+    },
+    {
+      "epoch": 1.9975708502024292,
+      "grad_norm": 1.1939406883857941,
+      "learning_rate": 5.147101589616783e-06,
+      "loss": 1.3627,
+      "step": 2467
+    },
+    {
+      "epoch": 1.9983805668016195,
+      "grad_norm": 1.209031327432166,
+      "learning_rate": 5.139588028770275e-06,
+      "loss": 1.3994,
+      "step": 2468
+    },
+    {
+      "epoch": 1.9991902834008097,
+      "grad_norm": 1.169763485845793,
+      "learning_rate": 5.13207805878017e-06,
+      "loss": 1.3228,
+      "step": 2469
+    },
+    {
+      "epoch": 2.0,
+      "grad_norm": 1.532603626298016,
+      "learning_rate": 5.124571685194804e-06,
+      "loss": 1.2018,
+      "step": 2470
+    },
+    {
+      "epoch": 2.0008097165991905,
+      "grad_norm": 2.854383023719633,
+      "learning_rate": 5.1170689135598675e-06,
+      "loss": 1.0829,
+      "step": 2471
+    },
+    {
+      "epoch": 2.0016194331983805,
+      "grad_norm": 3.050414009521738,
+      "learning_rate": 5.10956974941838e-06,
+      "loss": 1.0063,
+      "step": 2472
+    },
+    {
+      "epoch": 2.002429149797571,
+      "grad_norm": 2.4422306774820672,
+      "learning_rate": 5.102074198310701e-06,
+      "loss": 1.0396,
+      "step": 2473
+    },
+    {
+      "epoch": 2.003238866396761,
+      "grad_norm": 2.0451498041031044,
+      "learning_rate": 5.094582265774515e-06,
+      "loss": 0.9888,
+      "step": 2474
+    },
+    {
+      "epoch": 2.0040485829959516,
+      "grad_norm": 4.046585838573169,
+      "learning_rate": 5.087093957344841e-06,
+      "loss": 1.0822,
+      "step": 2475
+    },
+    {
+      "epoch": 2.0048582995951416,
+      "grad_norm": 4.926775146816468,
+      "learning_rate": 5.079609278554011e-06,
+      "loss": 1.0585,
+      "step": 2476
+    },
+    {
+      "epoch": 2.005668016194332,
+      "grad_norm": 3.3365273107950393,
+      "learning_rate": 5.07212823493169e-06,
+      "loss": 0.9847,
+      "step": 2477
+    },
+    {
+      "epoch": 2.006477732793522,
+      "grad_norm": 2.24516980725426,
+      "learning_rate": 5.064650832004839e-06,
+      "loss": 1.0298,
+      "step": 2478
+    },
+    {
+      "epoch": 2.0072874493927126,
+      "grad_norm": 1.7417896088574096,
+      "learning_rate": 5.057177075297748e-06,
+      "loss": 0.9663,
+      "step": 2479
+    },
+    {
+      "epoch": 2.0080971659919027,
+      "grad_norm": 2.031730299689621,
+      "learning_rate": 5.049706970332e-06,
+      "loss": 0.9755,
+      "step": 2480
+    },
+    {
+      "epoch": 2.008906882591093,
+      "grad_norm": 1.9976555875126931,
+      "learning_rate": 5.0422405226264825e-06,
+      "loss": 0.9815,
+      "step": 2481
+    },
+    {
+      "epoch": 2.0097165991902832,
+      "grad_norm": 1.9153896673542294,
+      "learning_rate": 5.034777737697384e-06,
+      "loss": 0.9825,
+      "step": 2482
+    },
+    {
+      "epoch": 2.0105263157894737,
+      "grad_norm": 1.6395893752150272,
+      "learning_rate": 5.027318621058182e-06,
+      "loss": 1.0408,
+      "step": 2483
+    },
+    {
+      "epoch": 2.0113360323886638,
+      "grad_norm": 1.6423936187066637,
+      "learning_rate": 5.019863178219653e-06,
+      "loss": 0.9731,
+      "step": 2484
+    },
+    {
+      "epoch": 2.0121457489878543,
+      "grad_norm": 1.7371817317654348,
+      "learning_rate": 5.0124114146898505e-06,
+      "loss": 0.9819,
+      "step": 2485
+    },
+    {
+      "epoch": 2.0129554655870447,
+      "grad_norm": 1.7846901106310733,
+      "learning_rate": 5.004963335974112e-06,
+      "loss": 1.0394,
+      "step": 2486
+    },
+    {
+      "epoch": 2.013765182186235,
+      "grad_norm": 1.7793293478267915,
+      "learning_rate": 4.997518947575058e-06,
+      "loss": 1.0592,
+      "step": 2487
+    },
+    {
+      "epoch": 2.0145748987854253,
+      "grad_norm": 1.6437092415729435,
+      "learning_rate": 4.990078254992574e-06,
+      "loss": 0.999,
+      "step": 2488
+    },
+    {
+      "epoch": 2.0153846153846153,
+      "grad_norm": 1.6534586866603884,
+      "learning_rate": 4.982641263723822e-06,
+      "loss": 1.0597,
+      "step": 2489
+    },
+    {
+      "epoch": 2.016194331983806,
+      "grad_norm": 1.5683644605257934,
+      "learning_rate": 4.9752079792632244e-06,
+      "loss": 1.0008,
+      "step": 2490
+    },
+    {
+      "epoch": 2.017004048582996,
+      "grad_norm": 1.6067812383281972,
+      "learning_rate": 4.967778407102466e-06,
+      "loss": 0.9898,
+      "step": 2491
+    },
+    {
+      "epoch": 2.0178137651821864,
+      "grad_norm": 1.6310900165582578,
+      "learning_rate": 4.960352552730495e-06,
+      "loss": 0.978,
+      "step": 2492
+    },
+    {
+      "epoch": 2.0186234817813764,
+      "grad_norm": 1.6529602670760182,
+      "learning_rate": 4.952930421633506e-06,
+      "loss": 0.9264,
+      "step": 2493
+    },
+    {
+      "epoch": 2.019433198380567,
+      "grad_norm": 1.7217510219463308,
+      "learning_rate": 4.945512019294941e-06,
+      "loss": 0.9976,
+      "step": 2494
+    },
+    {
+      "epoch": 2.020242914979757,
+      "grad_norm": 1.7026003567072832,
+      "learning_rate": 4.938097351195499e-06,
+      "loss": 0.9611,
+      "step": 2495
+    },
+    {
+      "epoch": 2.0210526315789474,
+      "grad_norm": 1.6872762681564812,
+      "learning_rate": 4.9306864228131094e-06,
+      "loss": 0.9497,
+      "step": 2496
+    },
+    {
+      "epoch": 2.0218623481781375,
+      "grad_norm": 1.6295705608523692,
+      "learning_rate": 4.92327923962294e-06,
+      "loss": 1.054,
+      "step": 2497
+    },
+    {
+      "epoch": 2.022672064777328,
+      "grad_norm": 1.6156871771493067,
+      "learning_rate": 4.91587580709739e-06,
+      "loss": 1.0382,
+      "step": 2498
+    },
+    {
+      "epoch": 2.023481781376518,
+      "grad_norm": 1.667645461991409,
+      "learning_rate": 4.9084761307061e-06,
+      "loss": 1.0005,
+      "step": 2499
+    },
+    {
+      "epoch": 2.0242914979757085,
+      "grad_norm": 1.6226089763668954,
+      "learning_rate": 4.9010802159159224e-06,
+      "loss": 0.9241,
+      "step": 2500
+    },
+    {
+      "epoch": 2.025101214574899,
+      "grad_norm": 1.6532518189213758,
+      "learning_rate": 4.893688068190933e-06,
+      "loss": 1.0058,
+      "step": 2501
+    },
+    {
+      "epoch": 2.025910931174089,
+      "grad_norm": 1.5598422496487256,
+      "learning_rate": 4.886299692992425e-06,
+      "loss": 0.9634,
+      "step": 2502
+    },
+    {
+      "epoch": 2.0267206477732795,
+      "grad_norm": 1.6760718021088303,
+      "learning_rate": 4.878915095778911e-06,
+      "loss": 0.9279,
+      "step": 2503
+    },
+    {
+      "epoch": 2.0275303643724696,
+      "grad_norm": 1.6857184074433873,
+      "learning_rate": 4.871534282006105e-06,
+      "loss": 1.033,
+      "step": 2504
+    },
+    {
+      "epoch": 2.02834008097166,
+      "grad_norm": 1.6164129806079546,
+      "learning_rate": 4.864157257126928e-06,
+      "loss": 1.0426,
+      "step": 2505
+    },
+    {
+      "epoch": 2.02914979757085,
+      "grad_norm": 1.6723447210528488,
+      "learning_rate": 4.856784026591497e-06,
+      "loss": 0.9716,
+      "step": 2506
+    },
+    {
+      "epoch": 2.0299595141700406,
+      "grad_norm": 1.574208483958483,
+      "learning_rate": 4.849414595847138e-06,
+      "loss": 1.0813,
+      "step": 2507
+    },
+    {
+      "epoch": 2.0307692307692307,
+      "grad_norm": 1.5924598069963187,
+      "learning_rate": 4.84204897033836e-06,
+      "loss": 1.0232,
+      "step": 2508
+    },
+    {
+      "epoch": 2.031578947368421,
+      "grad_norm": 1.6310804523408304,
+      "learning_rate": 4.834687155506861e-06,
+      "loss": 0.9985,
+      "step": 2509
+    },
+    {
+      "epoch": 2.032388663967611,
+      "grad_norm": 1.6173868253735737,
+      "learning_rate": 4.8273291567915225e-06,
+      "loss": 0.9707,
+      "step": 2510
+    },
+    {
+      "epoch": 2.0331983805668017,
+      "grad_norm": 1.665833483770265,
+      "learning_rate": 4.8199749796284175e-06,
+      "loss": 0.9886,
+      "step": 2511
+    },
+    {
+      "epoch": 2.0340080971659917,
+      "grad_norm": 1.5330111731342393,
+      "learning_rate": 4.812624629450785e-06,
+      "loss": 1.0049,
+      "step": 2512
+    },
+    {
+      "epoch": 2.0348178137651822,
+      "grad_norm": 1.6868658667526049,
+      "learning_rate": 4.805278111689035e-06,
+      "loss": 0.9885,
+      "step": 2513
+    },
+    {
+      "epoch": 2.0356275303643723,
+      "grad_norm": 1.7897259905366785,
+      "learning_rate": 4.797935431770758e-06,
+      "loss": 0.9826,
+      "step": 2514
+    },
+    {
+      "epoch": 2.0364372469635628,
+      "grad_norm": 1.7488839728709102,
+      "learning_rate": 4.790596595120699e-06,
+      "loss": 1.0691,
+      "step": 2515
+    },
+    {
+      "epoch": 2.0372469635627533,
+      "grad_norm": 1.6698774320169687,
+      "learning_rate": 4.783261607160764e-06,
+      "loss": 1.0223,
+      "step": 2516
+    },
+    {
+      "epoch": 2.0380566801619433,
+      "grad_norm": 1.6944023574692242,
+      "learning_rate": 4.775930473310021e-06,
+      "loss": 0.9541,
+      "step": 2517
+    },
+    {
+      "epoch": 2.038866396761134,
+      "grad_norm": 1.6030259296495724,
+      "learning_rate": 4.768603198984683e-06,
+      "loss": 0.9359,
+      "step": 2518
+    },
+    {
+      "epoch": 2.039676113360324,
+      "grad_norm": 1.7509997817466008,
+      "learning_rate": 4.761279789598122e-06,
+      "loss": 0.9543,
+      "step": 2519
+    },
+    {
+      "epoch": 2.0404858299595143,
+      "grad_norm": 1.6053230945872223,
+      "learning_rate": 4.753960250560843e-06,
+      "loss": 1.0359,
+      "step": 2520
+    },
+    {
+      "epoch": 2.0412955465587044,
+      "grad_norm": 1.6470550257288128,
+      "learning_rate": 4.746644587280505e-06,
+      "loss": 1.0407,
+      "step": 2521
+    },
+    {
+      "epoch": 2.042105263157895,
+      "grad_norm": 1.5960268921634935,
+      "learning_rate": 4.739332805161892e-06,
+      "loss": 1.0435,
+      "step": 2522
+    },
+    {
+      "epoch": 2.042914979757085,
+      "grad_norm": 1.6787492821611882,
+      "learning_rate": 4.732024909606923e-06,
+      "loss": 0.918,
+      "step": 2523
+    },
+    {
+      "epoch": 2.0437246963562754,
+      "grad_norm": 1.686860082454776,
+      "learning_rate": 4.7247209060146495e-06,
+      "loss": 1.0009,
+      "step": 2524
+    },
+    {
+      "epoch": 2.0445344129554655,
+      "grad_norm": 1.8404325405304682,
+      "learning_rate": 4.7174207997812436e-06,
+      "loss": 0.9994,
+      "step": 2525
+    },
+    {
+      "epoch": 2.045344129554656,
+      "grad_norm": 1.6172176781453194,
+      "learning_rate": 4.710124596299998e-06,
+      "loss": 1.0434,
+      "step": 2526
+    },
+    {
+      "epoch": 2.046153846153846,
+      "grad_norm": 1.603579157288264,
+      "learning_rate": 4.70283230096133e-06,
+      "loss": 0.9923,
+      "step": 2527
+    },
+    {
+      "epoch": 2.0469635627530365,
+      "grad_norm": 1.6474458591059657,
+      "learning_rate": 4.6955439191527556e-06,
+      "loss": 0.9799,
+      "step": 2528
+    },
+    {
+      "epoch": 2.0477732793522265,
+      "grad_norm": 1.6208921660937512,
+      "learning_rate": 4.688259456258916e-06,
+      "loss": 1.0971,
+      "step": 2529
+    },
+    {
+      "epoch": 2.048582995951417,
+      "grad_norm": 1.6056850691607139,
+      "learning_rate": 4.680978917661544e-06,
+      "loss": 0.9683,
+      "step": 2530
+    },
+    {
+      "epoch": 2.049392712550607,
+      "grad_norm": 1.6643700134128743,
+      "learning_rate": 4.673702308739478e-06,
+      "loss": 1.0623,
+      "step": 2531
+    },
+    {
+      "epoch": 2.0502024291497976,
+      "grad_norm": 1.6630243276082293,
+      "learning_rate": 4.666429634868651e-06,
+      "loss": 0.9393,
+      "step": 2532
+    },
+    {
+      "epoch": 2.051012145748988,
+      "grad_norm": 1.6267769934323475,
+      "learning_rate": 4.659160901422094e-06,
+      "loss": 1.0042,
+      "step": 2533
+    },
+    {
+      "epoch": 2.051821862348178,
+      "grad_norm": 1.798215794043288,
+      "learning_rate": 4.651896113769917e-06,
+      "loss": 1.0247,
+      "step": 2534
+    },
+    {
+      "epoch": 2.0526315789473686,
+      "grad_norm": 1.5817141672309059,
+      "learning_rate": 4.6446352772793256e-06,
+      "loss": 1.045,
+      "step": 2535
+    },
+    {
+      "epoch": 2.0534412955465586,
+      "grad_norm": 1.589804333120599,
+      "learning_rate": 4.637378397314607e-06,
+      "loss": 1.0288,
+      "step": 2536
+    },
+    {
+      "epoch": 2.054251012145749,
+      "grad_norm": 1.6164730177481579,
+      "learning_rate": 4.630125479237114e-06,
+      "loss": 0.9591,
+      "step": 2537
+    },
+    {
+      "epoch": 2.055060728744939,
+      "grad_norm": 1.5869372349513844,
+      "learning_rate": 4.622876528405281e-06,
+      "loss": 1.0036,
+      "step": 2538
+    },
+    {
+      "epoch": 2.0558704453441297,
+      "grad_norm": 1.6778929953471509,
+      "learning_rate": 4.615631550174609e-06,
+      "loss": 1.0513,
+      "step": 2539
+    },
+    {
+      "epoch": 2.0566801619433197,
+      "grad_norm": 1.5623260111430552,
+      "learning_rate": 4.608390549897661e-06,
+      "loss": 1.0336,
+      "step": 2540
+    },
+    {
+      "epoch": 2.05748987854251,
+      "grad_norm": 1.6803766333569634,
+      "learning_rate": 4.601153532924064e-06,
+      "loss": 0.9903,
+      "step": 2541
+    },
+    {
+      "epoch": 2.0582995951417002,
+      "grad_norm": 1.7893026888075112,
+      "learning_rate": 4.593920504600508e-06,
+      "loss": 0.9702,
+      "step": 2542
+    },
+    {
+      "epoch": 2.0591093117408907,
+      "grad_norm": 1.7193621454208152,
+      "learning_rate": 4.586691470270725e-06,
+      "loss": 1.0157,
+      "step": 2543
+    },
+    {
+      "epoch": 2.059919028340081,
+      "grad_norm": 1.660520678676988,
+      "learning_rate": 4.579466435275506e-06,
+      "loss": 0.9825,
+      "step": 2544
+    },
+    {
+      "epoch": 2.0607287449392713,
+      "grad_norm": 1.654111044168718,
+      "learning_rate": 4.5722454049526825e-06,
+      "loss": 0.9855,
+      "step": 2545
+    },
+    {
+      "epoch": 2.0615384615384613,
+      "grad_norm": 1.6536593350319584,
+      "learning_rate": 4.565028384637127e-06,
+      "loss": 0.9827,
+      "step": 2546
+    },
+    {
+      "epoch": 2.062348178137652,
+      "grad_norm": 1.7206634958914158,
+      "learning_rate": 4.557815379660749e-06,
+      "loss": 0.9644,
+      "step": 2547
+    },
+    {
+      "epoch": 2.0631578947368423,
+      "grad_norm": 1.6167239693996955,
+      "learning_rate": 4.550606395352496e-06,
+      "loss": 1.0163,
+      "step": 2548
+    },
+    {
+      "epoch": 2.0639676113360323,
+      "grad_norm": 1.6089251662105508,
+      "learning_rate": 4.543401437038335e-06,
+      "loss": 0.9805,
+      "step": 2549
+    },
+    {
+      "epoch": 2.064777327935223,
+      "grad_norm": 1.6783012771976722,
+      "learning_rate": 4.536200510041271e-06,
+      "loss": 1.0095,
+      "step": 2550
+    },
+    {
+      "epoch": 2.065587044534413,
+      "grad_norm": 1.7505966956320587,
+      "learning_rate": 4.5290036196813294e-06,
+      "loss": 0.9746,
+      "step": 2551
+    },
+    {
+      "epoch": 2.0663967611336034,
+      "grad_norm": 1.8266001199253628,
+      "learning_rate": 4.521810771275543e-06,
+      "loss": 1.0251,
+      "step": 2552
+    },
+    {
+      "epoch": 2.0672064777327934,
+      "grad_norm": 1.6164027260208844,
+      "learning_rate": 4.514621970137967e-06,
+      "loss": 1.0155,
+      "step": 2553
+    },
+    {
+      "epoch": 2.068016194331984,
+      "grad_norm": 1.7050380636821716,
+      "learning_rate": 4.507437221579662e-06,
+      "loss": 0.9616,
+      "step": 2554
+    },
+    {
+      "epoch": 2.068825910931174,
+      "grad_norm": 1.7946850395276819,
+      "learning_rate": 4.5002565309087e-06,
+      "loss": 0.9502,
+      "step": 2555
+    },
+    {
+      "epoch": 2.0696356275303645,
+      "grad_norm": 1.585579416581573,
+      "learning_rate": 4.493079903430144e-06,
+      "loss": 0.998,
+      "step": 2556
+    },
+    {
+      "epoch": 2.0704453441295545,
+      "grad_norm": 1.6347584013839378,
+      "learning_rate": 4.485907344446073e-06,
+      "loss": 1.006,
+      "step": 2557
+    },
+    {
+      "epoch": 2.071255060728745,
+      "grad_norm": 1.7338467691904824,
+      "learning_rate": 4.478738859255542e-06,
+      "loss": 1.0128,
+      "step": 2558
+    },
+    {
+      "epoch": 2.072064777327935,
+      "grad_norm": 1.8039959794063494,
+      "learning_rate": 4.4715744531546115e-06,
+      "loss": 0.9555,
+      "step": 2559
+    },
+    {
+      "epoch": 2.0728744939271255,
+      "grad_norm": 1.7310213842710942,
+      "learning_rate": 4.4644141314363165e-06,
+      "loss": 1.0433,
+      "step": 2560
+    },
+    {
+      "epoch": 2.0736842105263156,
+      "grad_norm": 1.5892088009452698,
+      "learning_rate": 4.45725789939068e-06,
+      "loss": 0.9779,
+      "step": 2561
+    },
+    {
+      "epoch": 2.074493927125506,
+      "grad_norm": 1.5328100810210468,
+      "learning_rate": 4.450105762304703e-06,
+      "loss": 1.057,
+      "step": 2562
+    },
+    {
+      "epoch": 2.0753036437246966,
+      "grad_norm": 1.5920760677864,
+      "learning_rate": 4.44295772546236e-06,
+      "loss": 1.0849,
+      "step": 2563
+    },
+    {
+      "epoch": 2.0761133603238866,
+      "grad_norm": 1.660029789485431,
+      "learning_rate": 4.435813794144596e-06,
+      "loss": 1.0217,
+      "step": 2564
+    },
+    {
+      "epoch": 2.076923076923077,
+      "grad_norm": 1.665382546622789,
+      "learning_rate": 4.4286739736293285e-06,
+      "loss": 0.9562,
+      "step": 2565
+    },
+    {
+      "epoch": 2.077732793522267,
+      "grad_norm": 1.6552078961930563,
+      "learning_rate": 4.421538269191427e-06,
+      "loss": 1.031,
+      "step": 2566
+    },
+    {
+      "epoch": 2.0785425101214576,
+      "grad_norm": 1.6606164514071944,
+      "learning_rate": 4.414406686102734e-06,
+      "loss": 0.9781,
+      "step": 2567
+    },
+    {
+      "epoch": 2.0793522267206477,
+      "grad_norm": 1.6124589826824705,
+      "learning_rate": 4.407279229632034e-06,
+      "loss": 1.0374,
+      "step": 2568
+    },
+    {
+      "epoch": 2.080161943319838,
+      "grad_norm": 1.7193015635747628,
+      "learning_rate": 4.400155905045073e-06,
+      "loss": 1.0336,
+      "step": 2569
+    },
+    {
+      "epoch": 2.080971659919028,
+      "grad_norm": 1.7315044687603365,
+      "learning_rate": 4.393036717604536e-06,
+      "loss": 0.9621,
+      "step": 2570
+    },
+    {
+      "epoch": 2.0817813765182187,
+      "grad_norm": 1.7098596418635459,
+      "learning_rate": 4.385921672570054e-06,
+      "loss": 0.9662,
+      "step": 2571
+    },
+    {
+      "epoch": 2.0825910931174088,
+      "grad_norm": 1.607788160578827,
+      "learning_rate": 4.378810775198203e-06,
+      "loss": 1.0082,
+      "step": 2572
+    },
+    {
+      "epoch": 2.0834008097165992,
+      "grad_norm": 1.653291357704816,
+      "learning_rate": 4.371704030742491e-06,
+      "loss": 1.0356,
+      "step": 2573
+    },
+    {
+      "epoch": 2.0842105263157893,
+      "grad_norm": 1.6881096442374148,
+      "learning_rate": 4.36460144445335e-06,
+      "loss": 1.0342,
+      "step": 2574
+    },
+    {
+      "epoch": 2.08502024291498,
+      "grad_norm": 1.7125542227995731,
+      "learning_rate": 4.357503021578158e-06,
+      "loss": 0.9266,
+      "step": 2575
+    },
+    {
+      "epoch": 2.08582995951417,
+      "grad_norm": 1.670269245177905,
+      "learning_rate": 4.3504087673612e-06,
+      "loss": 1.001,
+      "step": 2576
+    },
+    {
+      "epoch": 2.0866396761133603,
+      "grad_norm": 1.5784965452737714,
+      "learning_rate": 4.343318687043691e-06,
+      "loss": 1.0012,
+      "step": 2577
+    },
+    {
+      "epoch": 2.087449392712551,
+      "grad_norm": 1.6681995748372629,
+      "learning_rate": 4.336232785863756e-06,
+      "loss": 1.0095,
+      "step": 2578
+    },
+    {
+      "epoch": 2.088259109311741,
+      "grad_norm": 1.748694757073851,
+      "learning_rate": 4.329151069056432e-06,
+      "loss": 0.9591,
+      "step": 2579
+    },
+    {
+      "epoch": 2.0890688259109313,
+      "grad_norm": 1.7719387348638524,
+      "learning_rate": 4.322073541853677e-06,
+      "loss": 0.9696,
+      "step": 2580
+    },
+    {
+      "epoch": 2.0898785425101214,
+      "grad_norm": 1.64114377111076,
+      "learning_rate": 4.3150002094843415e-06,
+      "loss": 0.9863,
+      "step": 2581
+    },
+    {
+      "epoch": 2.090688259109312,
+      "grad_norm": 1.64738596574169,
+      "learning_rate": 4.307931077174175e-06,
+      "loss": 1.0393,
+      "step": 2582
+    },
+    {
+      "epoch": 2.091497975708502,
+      "grad_norm": 1.691860211239468,
+      "learning_rate": 4.300866150145837e-06,
+      "loss": 0.9525,
+      "step": 2583
+    },
+    {
+      "epoch": 2.0923076923076924,
+      "grad_norm": 1.576591959382383,
+      "learning_rate": 4.293805433618869e-06,
+      "loss": 1.0705,
+      "step": 2584
+    },
+    {
+      "epoch": 2.0931174089068825,
+      "grad_norm": 1.5738372098947238,
+      "learning_rate": 4.286748932809707e-06,
+      "loss": 1.0264,
+      "step": 2585
+    },
+    {
+      "epoch": 2.093927125506073,
+      "grad_norm": 1.6490322231201804,
+      "learning_rate": 4.279696652931663e-06,
+      "loss": 0.9917,
+      "step": 2586
+    },
+    {
+      "epoch": 2.094736842105263,
+      "grad_norm": 1.8188564774664804,
+      "learning_rate": 4.272648599194948e-06,
+      "loss": 0.9347,
+      "step": 2587
+    },
+    {
+      "epoch": 2.0955465587044535,
+      "grad_norm": 1.8581675482172395,
+      "learning_rate": 4.265604776806638e-06,
+      "loss": 0.9164,
+      "step": 2588
+    },
+    {
+      "epoch": 2.0963562753036435,
+      "grad_norm": 1.8577584612627356,
+      "learning_rate": 4.258565190970684e-06,
+      "loss": 1.0061,
+      "step": 2589
+    },
+    {
+      "epoch": 2.097165991902834,
+      "grad_norm": 1.743492463005408,
+      "learning_rate": 4.2515298468879064e-06,
+      "loss": 1.0067,
+      "step": 2590
+    },
+    {
+      "epoch": 2.097975708502024,
+      "grad_norm": 1.646875858354436,
+      "learning_rate": 4.244498749756e-06,
+      "loss": 1.0357,
+      "step": 2591
+    },
+    {
+      "epoch": 2.0987854251012146,
+      "grad_norm": 1.7035172799966736,
+      "learning_rate": 4.237471904769514e-06,
+      "loss": 1.0574,
+      "step": 2592
+    },
+    {
+      "epoch": 2.099595141700405,
+      "grad_norm": 1.729516687109549,
+      "learning_rate": 4.2304493171198605e-06,
+      "loss": 1.0129,
+      "step": 2593
+    },
+    {
+      "epoch": 2.100404858299595,
+      "grad_norm": 1.8017575479884655,
+      "learning_rate": 4.223430991995296e-06,
+      "loss": 1.0019,
+      "step": 2594
+    },
+    {
+      "epoch": 2.1012145748987856,
+      "grad_norm": 1.6594942547323506,
+      "learning_rate": 4.216416934580947e-06,
+      "loss": 0.9952,
+      "step": 2595
+    },
+    {
+      "epoch": 2.1020242914979756,
+      "grad_norm": 1.7239899018632698,
+      "learning_rate": 4.2094071500587695e-06,
+      "loss": 0.9851,
+      "step": 2596
+    },
+    {
+      "epoch": 2.102834008097166,
+      "grad_norm": 1.703904052950475,
+      "learning_rate": 4.202401643607572e-06,
+      "loss": 1.0284,
+      "step": 2597
+    },
+    {
+      "epoch": 2.103643724696356,
+      "grad_norm": 1.8593640278653734,
+      "learning_rate": 4.1954004204029945e-06,
+      "loss": 0.9578,
+      "step": 2598
+    },
+    {
+      "epoch": 2.1044534412955467,
+      "grad_norm": 1.856774903804991,
+      "learning_rate": 4.188403485617526e-06,
+      "loss": 1.0438,
+      "step": 2599
+    },
+    {
+      "epoch": 2.1052631578947367,
+      "grad_norm": 1.78449027364225,
+      "learning_rate": 4.181410844420473e-06,
+      "loss": 0.9355,
+      "step": 2600
+    },
+    {
+      "epoch": 2.106072874493927,
+      "grad_norm": 1.6445451971076204,
+      "learning_rate": 4.174422501977976e-06,
+      "loss": 1.0028,
+      "step": 2601
+    },
+    {
+      "epoch": 2.1068825910931173,
+      "grad_norm": 1.6462257848218957,
+      "learning_rate": 4.167438463453003e-06,
+      "loss": 0.9803,
+      "step": 2602
+    },
+    {
+      "epoch": 2.1076923076923078,
+      "grad_norm": 1.6907368674533254,
+      "learning_rate": 4.160458734005337e-06,
+      "loss": 1.0294,
+      "step": 2603
+    },
+    {
+      "epoch": 2.108502024291498,
+      "grad_norm": 1.6636161911260379,
+      "learning_rate": 4.153483318791579e-06,
+      "loss": 0.9813,
+      "step": 2604
+    },
+    {
+      "epoch": 2.1093117408906883,
+      "grad_norm": 1.7183152187153377,
+      "learning_rate": 4.146512222965144e-06,
+      "loss": 0.9823,
+      "step": 2605
+    },
+    {
+      "epoch": 2.1101214574898783,
+      "grad_norm": 1.6814396856451228,
+      "learning_rate": 4.139545451676248e-06,
+      "loss": 1.032,
+      "step": 2606
+    },
+    {
+      "epoch": 2.110931174089069,
+      "grad_norm": 1.7645866557280991,
+      "learning_rate": 4.1325830100719275e-06,
+      "loss": 1.0356,
+      "step": 2607
+    },
+    {
+      "epoch": 2.1117408906882593,
+      "grad_norm": 1.6961679753232464,
+      "learning_rate": 4.125624903296009e-06,
+      "loss": 1.0811,
+      "step": 2608
+    },
+    {
+      "epoch": 2.1125506072874494,
+      "grad_norm": 1.6525910699956303,
+      "learning_rate": 4.118671136489115e-06,
+      "loss": 0.9451,
+      "step": 2609
+    },
+    {
+      "epoch": 2.11336032388664,
+      "grad_norm": 1.6360702685596986,
+      "learning_rate": 4.111721714788671e-06,
+      "loss": 0.9375,
+      "step": 2610
+    },
+    {
+      "epoch": 2.11417004048583,
+      "grad_norm": 1.6179640067775025,
+      "learning_rate": 4.104776643328886e-06,
+      "loss": 1.0224,
+      "step": 2611
+    },
+    {
+      "epoch": 2.1149797570850204,
+      "grad_norm": 1.6525238056683458,
+      "learning_rate": 4.097835927240753e-06,
+      "loss": 1.0351,
+      "step": 2612
+    },
+    {
+      "epoch": 2.1157894736842104,
+      "grad_norm": 1.585412724307839,
+      "learning_rate": 4.090899571652053e-06,
+      "loss": 0.9931,
+      "step": 2613
+    },
+    {
+      "epoch": 2.116599190283401,
+      "grad_norm": 1.646625538543497,
+      "learning_rate": 4.083967581687338e-06,
+      "loss": 0.9385,
+      "step": 2614
+    },
+    {
+      "epoch": 2.117408906882591,
+      "grad_norm": 1.6984647506858865,
+      "learning_rate": 4.077039962467946e-06,
+      "loss": 0.9629,
+      "step": 2615
+    },
+    {
+      "epoch": 2.1182186234817815,
+      "grad_norm": 1.6719455263043292,
+      "learning_rate": 4.070116719111973e-06,
+      "loss": 1.0239,
+      "step": 2616
+    },
+    {
+      "epoch": 2.1190283400809715,
+      "grad_norm": 1.7112777230654765,
+      "learning_rate": 4.063197856734295e-06,
+      "loss": 0.979,
+      "step": 2617
+    },
+    {
+      "epoch": 2.119838056680162,
+      "grad_norm": 1.6987541653351368,
+      "learning_rate": 4.056283380446542e-06,
+      "loss": 1.0153,
+      "step": 2618
+    },
+    {
+      "epoch": 2.120647773279352,
+      "grad_norm": 1.6871431309146279,
+      "learning_rate": 4.049373295357105e-06,
+      "loss": 1.0376,
+      "step": 2619
+    },
+    {
+      "epoch": 2.1214574898785425,
+      "grad_norm": 1.6903927451402816,
+      "learning_rate": 4.042467606571134e-06,
+      "loss": 0.9581,
+      "step": 2620
+    },
+    {
+      "epoch": 2.1222672064777326,
+      "grad_norm": 1.6637711833229312,
+      "learning_rate": 4.0355663191905285e-06,
+      "loss": 1.0201,
+      "step": 2621
+    },
+    {
+      "epoch": 2.123076923076923,
+      "grad_norm": 1.7111155440518553,
+      "learning_rate": 4.028669438313933e-06,
+      "loss": 0.9114,
+      "step": 2622
+    },
+    {
+      "epoch": 2.1238866396761136,
+      "grad_norm": 1.7277429546609078,
+      "learning_rate": 4.0217769690367426e-06,
+      "loss": 0.9072,
+      "step": 2623
+    },
+    {
+      "epoch": 2.1246963562753036,
+      "grad_norm": 1.6772060985996178,
+      "learning_rate": 4.014888916451097e-06,
+      "loss": 0.9913,
+      "step": 2624
+    },
+    {
+      "epoch": 2.125506072874494,
+      "grad_norm": 1.6078641618211225,
+      "learning_rate": 4.008005285645863e-06,
+      "loss": 1.069,
+      "step": 2625
+    },
+    {
+      "epoch": 2.126315789473684,
+      "grad_norm": 1.620520443252991,
+      "learning_rate": 4.001126081706643e-06,
+      "loss": 0.9733,
+      "step": 2626
+    },
+    {
+      "epoch": 2.1271255060728746,
+      "grad_norm": 1.6078988933775202,
+      "learning_rate": 3.994251309715772e-06,
+      "loss": 1.0605,
+      "step": 2627
+    },
+    {
+      "epoch": 2.1279352226720647,
+      "grad_norm": 1.6029687010632365,
+      "learning_rate": 3.9873809747523075e-06,
+      "loss": 0.9842,
+      "step": 2628
+    },
+    {
+      "epoch": 2.128744939271255,
+      "grad_norm": 1.705951335796002,
+      "learning_rate": 3.98051508189203e-06,
+      "loss": 1.0009,
+      "step": 2629
+    },
+    {
+      "epoch": 2.1295546558704452,
+      "grad_norm": 1.6635774008101378,
+      "learning_rate": 3.973653636207437e-06,
+      "loss": 0.8982,
+      "step": 2630
+    },
+    {
+      "epoch": 2.1303643724696357,
+      "grad_norm": 1.7082989524349448,
+      "learning_rate": 3.966796642767745e-06,
+      "loss": 1.0141,
+      "step": 2631
+    },
+    {
+      "epoch": 2.1311740890688258,
+      "grad_norm": 1.6653197593597078,
+      "learning_rate": 3.959944106638881e-06,
+      "loss": 0.9809,
+      "step": 2632
+    },
+    {
+      "epoch": 2.1319838056680163,
+      "grad_norm": 1.6678594956056483,
+      "learning_rate": 3.953096032883473e-06,
+      "loss": 0.98,
+      "step": 2633
+    },
+    {
+      "epoch": 2.1327935222672063,
+      "grad_norm": 1.7005284089411195,
+      "learning_rate": 3.946252426560855e-06,
+      "loss": 0.9979,
+      "step": 2634
+    },
+    {
+      "epoch": 2.133603238866397,
+      "grad_norm": 1.6151071121961287,
+      "learning_rate": 3.939413292727061e-06,
+      "loss": 0.9968,
+      "step": 2635
+    },
+    {
+      "epoch": 2.134412955465587,
+      "grad_norm": 1.6806724677037892,
+      "learning_rate": 3.932578636434822e-06,
+      "loss": 0.9318,
+      "step": 2636
+    },
+    {
+      "epoch": 2.1352226720647773,
+      "grad_norm": 1.7311866714728976,
+      "learning_rate": 3.9257484627335545e-06,
+      "loss": 0.965,
+      "step": 2637
+    },
+    {
+      "epoch": 2.136032388663968,
+      "grad_norm": 1.7711121211961232,
+      "learning_rate": 3.9189227766693715e-06,
+      "loss": 0.9824,
+      "step": 2638
+    },
+    {
+      "epoch": 2.136842105263158,
+      "grad_norm": 1.684667995414289,
+      "learning_rate": 3.912101583285072e-06,
+      "loss": 0.915,
+      "step": 2639
+    },
+    {
+      "epoch": 2.1376518218623484,
+      "grad_norm": 1.7191802163538186,
+      "learning_rate": 3.9052848876201285e-06,
+      "loss": 0.9592,
+      "step": 2640
+    },
+    {
+      "epoch": 2.1384615384615384,
+      "grad_norm": 1.6407834559216017,
+      "learning_rate": 3.898472694710692e-06,
+      "loss": 0.9821,
+      "step": 2641
+    },
+    {
+      "epoch": 2.139271255060729,
+      "grad_norm": 1.6128402655438978,
+      "learning_rate": 3.891665009589588e-06,
+      "loss": 0.9976,
+      "step": 2642
+    },
+    {
+      "epoch": 2.140080971659919,
+      "grad_norm": 1.7113632799708938,
+      "learning_rate": 3.884861837286314e-06,
+      "loss": 0.9884,
+      "step": 2643
+    },
+    {
+      "epoch": 2.1408906882591094,
+      "grad_norm": 1.6447541227519287,
+      "learning_rate": 3.878063182827025e-06,
+      "loss": 0.9051,
+      "step": 2644
+    },
+    {
+      "epoch": 2.1417004048582995,
+      "grad_norm": 1.700018398274763,
+      "learning_rate": 3.8712690512345555e-06,
+      "loss": 1.0079,
+      "step": 2645
+    },
+    {
+      "epoch": 2.14251012145749,
+      "grad_norm": 1.7278282489587906,
+      "learning_rate": 3.8644794475283754e-06,
+      "loss": 0.9488,
+      "step": 2646
+    },
+    {
+      "epoch": 2.14331983805668,
+      "grad_norm": 1.7030788584383703,
+      "learning_rate": 3.857694376724634e-06,
+      "loss": 0.9766,
+      "step": 2647
+    },
+    {
+      "epoch": 2.1441295546558705,
+      "grad_norm": 1.6975892598366673,
+      "learning_rate": 3.850913843836111e-06,
+      "loss": 0.9645,
+      "step": 2648
+    },
+    {
+      "epoch": 2.1449392712550606,
+      "grad_norm": 1.6362040151463373,
+      "learning_rate": 3.844137853872245e-06,
+      "loss": 0.9809,
+      "step": 2649
+    },
+    {
+      "epoch": 2.145748987854251,
+      "grad_norm": 1.6598578036651748,
+      "learning_rate": 3.837366411839114e-06,
+      "loss": 0.9822,
+      "step": 2650
+    },
+    {
+      "epoch": 2.146558704453441,
+      "grad_norm": 1.6357704564069009,
+      "learning_rate": 3.830599522739437e-06,
+      "loss": 1.0175,
+      "step": 2651
+    },
+    {
+      "epoch": 2.1473684210526316,
+      "grad_norm": 1.6232730145470107,
+      "learning_rate": 3.823837191572567e-06,
+      "loss": 0.9417,
+      "step": 2652
+    },
+    {
+      "epoch": 2.148178137651822,
+      "grad_norm": 1.6210092712146753,
+      "learning_rate": 3.817079423334497e-06,
+      "loss": 0.9656,
+      "step": 2653
+    },
+    {
+      "epoch": 2.148987854251012,
+      "grad_norm": 1.688514119577297,
+      "learning_rate": 3.8103262230178395e-06,
+      "loss": 1.0027,
+      "step": 2654
+    },
+    {
+      "epoch": 2.1497975708502026,
+      "grad_norm": 1.6972006744474526,
+      "learning_rate": 3.8035775956118416e-06,
+      "loss": 1.0177,
+      "step": 2655
+    },
+    {
+      "epoch": 2.1506072874493927,
+      "grad_norm": 1.6862311012034847,
+      "learning_rate": 3.7968335461023654e-06,
+      "loss": 0.9403,
+      "step": 2656
+    },
+    {
+      "epoch": 2.151417004048583,
+      "grad_norm": 1.6576918628517623,
+      "learning_rate": 3.790094079471891e-06,
+      "loss": 0.9654,
+      "step": 2657
+    },
+    {
+      "epoch": 2.152226720647773,
+      "grad_norm": 1.705286165779682,
+      "learning_rate": 3.7833592006995144e-06,
+      "loss": 1.0502,
+      "step": 2658
+    },
+    {
+      "epoch": 2.1530364372469637,
+      "grad_norm": 1.6153316475940291,
+      "learning_rate": 3.77662891476094e-06,
+      "loss": 0.9806,
+      "step": 2659
+    },
+    {
+      "epoch": 2.1538461538461537,
+      "grad_norm": 1.6880698498431286,
+      "learning_rate": 3.7699032266284863e-06,
+      "loss": 1.0253,
+      "step": 2660
+    },
+    {
+      "epoch": 2.1546558704453442,
+      "grad_norm": 1.7777836789038124,
+      "learning_rate": 3.7631821412710668e-06,
+      "loss": 0.9982,
+      "step": 2661
+    },
+    {
+      "epoch": 2.1554655870445343,
+      "grad_norm": 1.6474905913845634,
+      "learning_rate": 3.7564656636541928e-06,
+      "loss": 0.9736,
+      "step": 2662
+    },
+    {
+      "epoch": 2.1562753036437248,
+      "grad_norm": 1.64817525670547,
+      "learning_rate": 3.7497537987399836e-06,
+      "loss": 0.9996,
+      "step": 2663
+    },
+    {
+      "epoch": 2.157085020242915,
+      "grad_norm": 1.6328494859671836,
+      "learning_rate": 3.7430465514871405e-06,
+      "loss": 0.9529,
+      "step": 2664
+    },
+    {
+      "epoch": 2.1578947368421053,
+      "grad_norm": 1.5460907995988389,
+      "learning_rate": 3.736343926850954e-06,
+      "loss": 0.9776,
+      "step": 2665
+    },
+    {
+      "epoch": 2.1587044534412954,
+      "grad_norm": 1.7999028275226707,
+      "learning_rate": 3.729645929783302e-06,
+      "loss": 0.9444,
+      "step": 2666
+    },
+    {
+      "epoch": 2.159514170040486,
+      "grad_norm": 1.6674054885247187,
+      "learning_rate": 3.7229525652326392e-06,
+      "loss": 0.9698,
+      "step": 2667
+    },
+    {
+      "epoch": 2.1603238866396763,
+      "grad_norm": 1.7459838166181108,
+      "learning_rate": 3.7162638381440077e-06,
+      "loss": 0.9791,
+      "step": 2668
+    },
+    {
+      "epoch": 2.1611336032388664,
+      "grad_norm": 1.7171429916272887,
+      "learning_rate": 3.709579753459015e-06,
+      "loss": 1.0177,
+      "step": 2669
+    },
+    {
+      "epoch": 2.161943319838057,
+      "grad_norm": 1.729852115207468,
+      "learning_rate": 3.702900316115836e-06,
+      "loss": 1.0042,
+      "step": 2670
+    },
+    {
+      "epoch": 2.162753036437247,
+      "grad_norm": 1.7165892661582856,
+      "learning_rate": 3.6962255310492256e-06,
+      "loss": 0.9719,
+      "step": 2671
+    },
+    {
+      "epoch": 2.1635627530364374,
+      "grad_norm": 1.628850570144987,
+      "learning_rate": 3.689555403190488e-06,
+      "loss": 1.0131,
+      "step": 2672
+    },
+    {
+      "epoch": 2.1643724696356275,
+      "grad_norm": 1.6296844076079475,
+      "learning_rate": 3.6828899374674933e-06,
+      "loss": 0.9733,
+      "step": 2673
+    },
+    {
+      "epoch": 2.165182186234818,
+      "grad_norm": 1.623822465699174,
+      "learning_rate": 3.67622913880466e-06,
+      "loss": 1.0049,
+      "step": 2674
+    },
+    {
+      "epoch": 2.165991902834008,
+      "grad_norm": 1.6705667827629207,
+      "learning_rate": 3.6695730121229734e-06,
+      "loss": 1.0427,
+      "step": 2675
+    },
+    {
+      "epoch": 2.1668016194331985,
+      "grad_norm": 1.6901112337582092,
+      "learning_rate": 3.6629215623399526e-06,
+      "loss": 0.9016,
+      "step": 2676
+    },
+    {
+      "epoch": 2.1676113360323885,
+      "grad_norm": 1.635822547002648,
+      "learning_rate": 3.6562747943696696e-06,
+      "loss": 1.0312,
+      "step": 2677
+    },
+    {
+      "epoch": 2.168421052631579,
+      "grad_norm": 1.773469367884702,
+      "learning_rate": 3.6496327131227284e-06,
+      "loss": 0.9799,
+      "step": 2678
+    },
+    {
+      "epoch": 2.169230769230769,
+      "grad_norm": 1.703119798535184,
+      "learning_rate": 3.6429953235062853e-06,
+      "loss": 0.9642,
+      "step": 2679
+    },
+    {
+      "epoch": 2.1700404858299596,
+      "grad_norm": 1.5950037755310869,
+      "learning_rate": 3.6363626304240185e-06,
+      "loss": 0.9954,
+      "step": 2680
+    },
+    {
+      "epoch": 2.1708502024291496,
+      "grad_norm": 1.6651131112700766,
+      "learning_rate": 3.629734638776139e-06,
+      "loss": 0.9598,
+      "step": 2681
+    },
+    {
+      "epoch": 2.17165991902834,
+      "grad_norm": 1.674961128025693,
+      "learning_rate": 3.6231113534593833e-06,
+      "loss": 0.9485,
+      "step": 2682
+    },
+    {
+      "epoch": 2.1724696356275306,
+      "grad_norm": 1.6927204922354684,
+      "learning_rate": 3.616492779367018e-06,
+      "loss": 1.0114,
+      "step": 2683
+    },
+    {
+      "epoch": 2.1732793522267206,
+      "grad_norm": 1.6141482144444916,
+      "learning_rate": 3.609878921388822e-06,
+      "loss": 1.0204,
+      "step": 2684
+    },
+    {
+      "epoch": 2.174089068825911,
+      "grad_norm": 1.5938306093069654,
+      "learning_rate": 3.6032697844110896e-06,
+      "loss": 1.0281,
+      "step": 2685
+    },
+    {
+      "epoch": 2.174898785425101,
+      "grad_norm": 1.674185209940073,
+      "learning_rate": 3.596665373316629e-06,
+      "loss": 1.0156,
+      "step": 2686
+    },
+    {
+      "epoch": 2.1757085020242917,
+      "grad_norm": 1.652753408219103,
+      "learning_rate": 3.590065692984762e-06,
+      "loss": 0.9686,
+      "step": 2687
+    },
+    {
+      "epoch": 2.1765182186234817,
+      "grad_norm": 1.6554790863907958,
+      "learning_rate": 3.583470748291309e-06,
+      "loss": 0.961,
+      "step": 2688
+    },
+    {
+      "epoch": 2.177327935222672,
+      "grad_norm": 1.7595295916415807,
+      "learning_rate": 3.5768805441085885e-06,
+      "loss": 1.0052,
+      "step": 2689
+    },
+    {
+      "epoch": 2.1781376518218623,
+      "grad_norm": 1.7129389096310346,
+      "learning_rate": 3.5702950853054284e-06,
+      "loss": 1.0033,
+      "step": 2690
+    },
+    {
+      "epoch": 2.1789473684210527,
+      "grad_norm": 1.6958273168912117,
+      "learning_rate": 3.5637143767471427e-06,
+      "loss": 1.0284,
+      "step": 2691
+    },
+    {
+      "epoch": 2.179757085020243,
+      "grad_norm": 1.6851614860716266,
+      "learning_rate": 3.5571384232955365e-06,
+      "loss": 0.971,
+      "step": 2692
+    },
+    {
+      "epoch": 2.1805668016194333,
+      "grad_norm": 1.6326875705295927,
+      "learning_rate": 3.550567229808901e-06,
+      "loss": 1.002,
+      "step": 2693
+    },
+    {
+      "epoch": 2.1813765182186233,
+      "grad_norm": 1.6508788198860427,
+      "learning_rate": 3.5440008011420103e-06,
+      "loss": 1.0164,
+      "step": 2694
+    },
+    {
+      "epoch": 2.182186234817814,
+      "grad_norm": 1.6307505301931884,
+      "learning_rate": 3.5374391421461273e-06,
+      "loss": 0.9336,
+      "step": 2695
+    },
+    {
+      "epoch": 2.182995951417004,
+      "grad_norm": 1.6240785364108545,
+      "learning_rate": 3.5308822576689805e-06,
+      "loss": 1.0049,
+      "step": 2696
+    },
+    {
+      "epoch": 2.1838056680161944,
+      "grad_norm": 1.6543399441514246,
+      "learning_rate": 3.5243301525547714e-06,
+      "loss": 1.0455,
+      "step": 2697
+    },
+    {
+      "epoch": 2.184615384615385,
+      "grad_norm": 1.650315122489001,
+      "learning_rate": 3.5177828316441797e-06,
+      "loss": 0.9574,
+      "step": 2698
+    },
+    {
+      "epoch": 2.185425101214575,
+      "grad_norm": 1.6675806777098334,
+      "learning_rate": 3.511240299774341e-06,
+      "loss": 1.005,
+      "step": 2699
+    },
+    {
+      "epoch": 2.1862348178137654,
+      "grad_norm": 1.6634779623745304,
+      "learning_rate": 3.5047025617788578e-06,
+      "loss": 1.0074,
+      "step": 2700
+    },
+    {
+      "epoch": 2.1870445344129554,
+      "grad_norm": 1.6256164088565501,
+      "learning_rate": 3.4981696224877893e-06,
+      "loss": 0.9875,
+      "step": 2701
+    },
+    {
+      "epoch": 2.187854251012146,
+      "grad_norm": 1.7291877052522024,
+      "learning_rate": 3.491641486727645e-06,
+      "loss": 0.9844,
+      "step": 2702
+    },
+    {
+      "epoch": 2.188663967611336,
+      "grad_norm": 1.7129557045779524,
+      "learning_rate": 3.4851181593213967e-06,
+      "loss": 0.9625,
+      "step": 2703
+    },
+    {
+      "epoch": 2.1894736842105265,
+      "grad_norm": 1.7620888900482394,
+      "learning_rate": 3.478599645088453e-06,
+      "loss": 0.9872,
+      "step": 2704
+    },
+    {
+      "epoch": 2.1902834008097165,
+      "grad_norm": 1.6649502085884391,
+      "learning_rate": 3.4720859488446744e-06,
+      "loss": 0.9264,
+      "step": 2705
+    },
+    {
+      "epoch": 2.191093117408907,
+      "grad_norm": 1.6084607566937745,
+      "learning_rate": 3.4655770754023574e-06,
+      "loss": 0.9531,
+      "step": 2706
+    },
+    {
+      "epoch": 2.191902834008097,
+      "grad_norm": 1.6974794965679043,
+      "learning_rate": 3.4590730295702356e-06,
+      "loss": 0.9283,
+      "step": 2707
+    },
+    {
+      "epoch": 2.1927125506072875,
+      "grad_norm": 1.687339468024081,
+      "learning_rate": 3.452573816153476e-06,
+      "loss": 0.9698,
+      "step": 2708
+    },
+    {
+      "epoch": 2.1935222672064776,
+      "grad_norm": 1.7543324832553966,
+      "learning_rate": 3.446079439953677e-06,
+      "loss": 0.9411,
+      "step": 2709
+    },
+    {
+      "epoch": 2.194331983805668,
+      "grad_norm": 1.711648121979403,
+      "learning_rate": 3.4395899057688575e-06,
+      "loss": 0.9817,
+      "step": 2710
+    },
+    {
+      "epoch": 2.195141700404858,
+      "grad_norm": 1.664633436704711,
+      "learning_rate": 3.4331052183934687e-06,
+      "loss": 0.9834,
+      "step": 2711
+    },
+    {
+      "epoch": 2.1959514170040486,
+      "grad_norm": 1.655215067217286,
+      "learning_rate": 3.4266253826183805e-06,
+      "loss": 0.9665,
+      "step": 2712
+    },
+    {
+      "epoch": 2.196761133603239,
+      "grad_norm": 1.7006734272766477,
+      "learning_rate": 3.4201504032308695e-06,
+      "loss": 0.9777,
+      "step": 2713
+    },
+    {
+      "epoch": 2.197570850202429,
+      "grad_norm": 1.714785930728774,
+      "learning_rate": 3.41368028501463e-06,
+      "loss": 0.9656,
+      "step": 2714
+    },
+    {
+      "epoch": 2.1983805668016196,
+      "grad_norm": 1.6909113867716612,
+      "learning_rate": 3.407215032749763e-06,
+      "loss": 1.0659,
+      "step": 2715
+    },
+    {
+      "epoch": 2.1991902834008097,
+      "grad_norm": 1.68682986907009,
+      "learning_rate": 3.4007546512127764e-06,
+      "loss": 0.9103,
+      "step": 2716
+    },
+    {
+      "epoch": 2.2,
+      "grad_norm": 1.6137573322333,
+      "learning_rate": 3.3942991451765793e-06,
+      "loss": 0.9957,
+      "step": 2717
+    },
+    {
+      "epoch": 2.2008097165991902,
+      "grad_norm": 1.6129056026950326,
+      "learning_rate": 3.387848519410475e-06,
+      "loss": 0.9935,
+      "step": 2718
+    },
+    {
+      "epoch": 2.2016194331983807,
+      "grad_norm": 1.694139432275297,
+      "learning_rate": 3.3814027786801675e-06,
+      "loss": 0.9418,
+      "step": 2719
+    },
+    {
+      "epoch": 2.2024291497975708,
+      "grad_norm": 1.7634838421878205,
+      "learning_rate": 3.374961927747751e-06,
+      "loss": 0.9193,
+      "step": 2720
+    },
+    {
+      "epoch": 2.2032388663967613,
+      "grad_norm": 1.6980171565314421,
+      "learning_rate": 3.3685259713717034e-06,
+      "loss": 0.9727,
+      "step": 2721
+    },
+    {
+      "epoch": 2.2040485829959513,
+      "grad_norm": 1.722930414761989,
+      "learning_rate": 3.362094914306888e-06,
+      "loss": 0.93,
+      "step": 2722
+    },
+    {
+      "epoch": 2.204858299595142,
+      "grad_norm": 1.6124571834590236,
+      "learning_rate": 3.355668761304548e-06,
+      "loss": 1.1032,
+      "step": 2723
+    },
+    {
+      "epoch": 2.205668016194332,
+      "grad_norm": 1.6394333882696688,
+      "learning_rate": 3.349247517112305e-06,
+      "loss": 1.0452,
+      "step": 2724
+    },
+    {
+      "epoch": 2.2064777327935223,
+      "grad_norm": 1.6345795057358297,
+      "learning_rate": 3.342831186474149e-06,
+      "loss": 0.9504,
+      "step": 2725
+    },
+    {
+      "epoch": 2.2072874493927124,
+      "grad_norm": 1.6475621511846046,
+      "learning_rate": 3.336419774130447e-06,
+      "loss": 1.0154,
+      "step": 2726
+    },
+    {
+      "epoch": 2.208097165991903,
+      "grad_norm": 1.6238349459249526,
+      "learning_rate": 3.3300132848179346e-06,
+      "loss": 0.9662,
+      "step": 2727
+    },
+    {
+      "epoch": 2.208906882591093,
+      "grad_norm": 1.6615756750809165,
+      "learning_rate": 3.3236117232696984e-06,
+      "loss": 1.0466,
+      "step": 2728
+    },
+    {
+      "epoch": 2.2097165991902834,
+      "grad_norm": 1.7120719200934522,
+      "learning_rate": 3.3172150942151947e-06,
+      "loss": 0.9345,
+      "step": 2729
+    },
+    {
+      "epoch": 2.2105263157894735,
+      "grad_norm": 1.6804653865746528,
+      "learning_rate": 3.31082340238023e-06,
+      "loss": 0.9928,
+      "step": 2730
+    },
+    {
+      "epoch": 2.211336032388664,
+      "grad_norm": 1.7147429622035564,
+      "learning_rate": 3.3044366524869652e-06,
+      "loss": 0.9555,
+      "step": 2731
+    },
+    {
+      "epoch": 2.2121457489878544,
+      "grad_norm": 1.664455141304305,
+      "learning_rate": 3.2980548492539064e-06,
+      "loss": 1.0162,
+      "step": 2732
+    },
+    {
+      "epoch": 2.2129554655870445,
+      "grad_norm": 1.7280945010756663,
+      "learning_rate": 3.291677997395918e-06,
+      "loss": 0.9628,
+      "step": 2733
+    },
+    {
+      "epoch": 2.213765182186235,
+      "grad_norm": 1.673275550114484,
+      "learning_rate": 3.2853061016241884e-06,
+      "loss": 1.0242,
+      "step": 2734
+    },
+    {
+      "epoch": 2.214574898785425,
+      "grad_norm": 1.7272385557412295,
+      "learning_rate": 3.27893916664626e-06,
+      "loss": 0.9813,
+      "step": 2735
+    },
+    {
+      "epoch": 2.2153846153846155,
+      "grad_norm": 1.6681828389749171,
+      "learning_rate": 3.2725771971660002e-06,
+      "loss": 0.9479,
+      "step": 2736
+    },
+    {
+      "epoch": 2.2161943319838056,
+      "grad_norm": 1.6481760708593136,
+      "learning_rate": 3.266220197883613e-06,
+      "loss": 0.9863,
+      "step": 2737
+    },
+    {
+      "epoch": 2.217004048582996,
+      "grad_norm": 1.6193831329668813,
+      "learning_rate": 3.259868173495626e-06,
+      "loss": 0.972,
+      "step": 2738
+    },
+    {
+      "epoch": 2.217813765182186,
+      "grad_norm": 1.6439379984114413,
+      "learning_rate": 3.2535211286948955e-06,
+      "loss": 0.9832,
+      "step": 2739
+    },
+    {
+      "epoch": 2.2186234817813766,
+      "grad_norm": 1.6572224667024693,
+      "learning_rate": 3.2471790681705928e-06,
+      "loss": 1.036,
+      "step": 2740
+    },
+    {
+      "epoch": 2.2194331983805666,
+      "grad_norm": 1.6609689760873279,
+      "learning_rate": 3.2408419966082195e-06,
+      "loss": 1.0051,
+      "step": 2741
+    },
+    {
+      "epoch": 2.220242914979757,
+      "grad_norm": 1.6875131566811306,
+      "learning_rate": 3.2345099186895758e-06,
+      "loss": 0.9738,
+      "step": 2742
+    },
+    {
+      "epoch": 2.221052631578947,
+      "grad_norm": 1.7178275452285305,
+      "learning_rate": 3.2281828390927873e-06,
+      "loss": 1.0007,
+      "step": 2743
+    },
+    {
+      "epoch": 2.2218623481781377,
+      "grad_norm": 1.7172578500910411,
+      "learning_rate": 3.221860762492275e-06,
+      "loss": 1.0334,
+      "step": 2744
+    },
+    {
+      "epoch": 2.2226720647773277,
+      "grad_norm": 1.6450445517301915,
+      "learning_rate": 3.215543693558769e-06,
+      "loss": 1.0215,
+      "step": 2745
+    },
+    {
+      "epoch": 2.223481781376518,
+      "grad_norm": 1.6443211869924201,
+      "learning_rate": 3.2092316369593e-06,
+      "loss": 0.9702,
+      "step": 2746
+    },
+    {
+      "epoch": 2.2242914979757087,
+      "grad_norm": 1.6072557369507485,
+      "learning_rate": 3.20292459735719e-06,
+      "loss": 0.9568,
+      "step": 2747
+    },
+    {
+      "epoch": 2.2251012145748987,
+      "grad_norm": 1.5890263149144106,
+      "learning_rate": 3.1966225794120666e-06,
+      "loss": 0.9662,
+      "step": 2748
+    },
+    {
+      "epoch": 2.2259109311740892,
+      "grad_norm": 1.6626050407972168,
+      "learning_rate": 3.1903255877798365e-06,
+      "loss": 0.9184,
+      "step": 2749
+    },
+    {
+      "epoch": 2.2267206477732793,
+      "grad_norm": 1.6429805630498902,
+      "learning_rate": 3.1840336271126935e-06,
+      "loss": 0.9815,
+      "step": 2750
+    },
+    {
+      "epoch": 2.2275303643724698,
+      "grad_norm": 1.7143845132000823,
+      "learning_rate": 3.1777467020591236e-06,
+      "loss": 1.0032,
+      "step": 2751
+    },
+    {
+      "epoch": 2.22834008097166,
+      "grad_norm": 1.7160771696928598,
+      "learning_rate": 3.1714648172638827e-06,
+      "loss": 0.9347,
+      "step": 2752
+    },
+    {
+      "epoch": 2.2291497975708503,
+      "grad_norm": 1.737109438949758,
+      "learning_rate": 3.165187977368007e-06,
+      "loss": 0.9485,
+      "step": 2753
+    },
+    {
+      "epoch": 2.2299595141700403,
+      "grad_norm": 1.6534753583589024,
+      "learning_rate": 3.158916187008806e-06,
+      "loss": 0.9403,
+      "step": 2754
+    },
+    {
+      "epoch": 2.230769230769231,
+      "grad_norm": 1.6881735304925691,
+      "learning_rate": 3.152649450819852e-06,
+      "loss": 0.989,
+      "step": 2755
+    },
+    {
+      "epoch": 2.231578947368421,
+      "grad_norm": 1.683248270203457,
+      "learning_rate": 3.146387773431e-06,
+      "loss": 1.0457,
+      "step": 2756
+    },
+    {
+      "epoch": 2.2323886639676114,
+      "grad_norm": 1.716360447509341,
+      "learning_rate": 3.1401311594683494e-06,
+      "loss": 1.011,
+      "step": 2757
+    },
+    {
+      "epoch": 2.2331983805668014,
+      "grad_norm": 1.726117425194692,
+      "learning_rate": 3.1338796135542647e-06,
+      "loss": 0.9894,
+      "step": 2758
+    },
+    {
+      "epoch": 2.234008097165992,
+      "grad_norm": 1.722136573207579,
+      "learning_rate": 3.1276331403073733e-06,
+      "loss": 0.9064,
+      "step": 2759
+    },
+    {
+      "epoch": 2.234817813765182,
+      "grad_norm": 1.6672196996268878,
+      "learning_rate": 3.1213917443425456e-06,
+      "loss": 0.9917,
+      "step": 2760
+    },
+    {
+      "epoch": 2.2356275303643725,
+      "grad_norm": 1.7657193112711613,
+      "learning_rate": 3.1151554302709063e-06,
+      "loss": 1.0076,
+      "step": 2761
+    },
+    {
+      "epoch": 2.236437246963563,
+      "grad_norm": 1.6902194564599813,
+      "learning_rate": 3.108924202699819e-06,
+      "loss": 1.0375,
+      "step": 2762
+    },
+    {
+      "epoch": 2.237246963562753,
+      "grad_norm": 1.7990661956680112,
+      "learning_rate": 3.1026980662328997e-06,
+      "loss": 0.9798,
+      "step": 2763
+    },
+    {
+      "epoch": 2.2380566801619435,
+      "grad_norm": 1.6588109292539122,
+      "learning_rate": 3.096477025469996e-06,
+      "loss": 0.9786,
+      "step": 2764
+    },
+    {
+      "epoch": 2.2388663967611335,
+      "grad_norm": 1.7473866019174604,
+      "learning_rate": 3.0902610850071922e-06,
+      "loss": 1.001,
+      "step": 2765
+    },
+    {
+      "epoch": 2.239676113360324,
+      "grad_norm": 1.5963776881440452,
+      "learning_rate": 3.084050249436802e-06,
+      "loss": 0.9453,
+      "step": 2766
+    },
+    {
+      "epoch": 2.240485829959514,
+      "grad_norm": 1.6511854722362378,
+      "learning_rate": 3.077844523347374e-06,
+      "loss": 0.9677,
+      "step": 2767
+    },
+    {
+      "epoch": 2.2412955465587046,
+      "grad_norm": 1.8031304104053283,
+      "learning_rate": 3.0716439113236785e-06,
+      "loss": 0.9372,
+      "step": 2768
+    },
+    {
+      "epoch": 2.2421052631578946,
+      "grad_norm": 1.6716997949617207,
+      "learning_rate": 3.0654484179467047e-06,
+      "loss": 0.9719,
+      "step": 2769
+    },
+    {
+      "epoch": 2.242914979757085,
+      "grad_norm": 1.623759924284454,
+      "learning_rate": 3.0592580477936606e-06,
+      "loss": 0.9986,
+      "step": 2770
+    },
+    {
+      "epoch": 2.243724696356275,
+      "grad_norm": 1.750132587533433,
+      "learning_rate": 3.0530728054379787e-06,
+      "loss": 0.9968,
+      "step": 2771
+    },
+    {
+      "epoch": 2.2445344129554656,
+      "grad_norm": 1.6683052360621997,
+      "learning_rate": 3.0468926954492907e-06,
+      "loss": 0.9192,
+      "step": 2772
+    },
+    {
+      "epoch": 2.2453441295546557,
+      "grad_norm": 1.7517498796441024,
+      "learning_rate": 3.0407177223934426e-06,
+      "loss": 0.9623,
+      "step": 2773
+    },
+    {
+      "epoch": 2.246153846153846,
+      "grad_norm": 1.7300413517444366,
+      "learning_rate": 3.034547890832481e-06,
+      "loss": 1.0161,
+      "step": 2774
+    },
+    {
+      "epoch": 2.246963562753036,
+      "grad_norm": 1.6389215305609568,
+      "learning_rate": 3.0283832053246644e-06,
+      "loss": 0.9079,
+      "step": 2775
+    },
+    {
+      "epoch": 2.2477732793522267,
+      "grad_norm": 1.6760927536723689,
+      "learning_rate": 3.022223670424437e-06,
+      "loss": 0.9628,
+      "step": 2776
+    },
+    {
+      "epoch": 2.248582995951417,
+      "grad_norm": 1.6857597576034187,
+      "learning_rate": 3.016069290682441e-06,
+      "loss": 0.9542,
+      "step": 2777
+    },
+    {
+      "epoch": 2.2493927125506072,
+      "grad_norm": 1.6715075659799083,
+      "learning_rate": 3.009920070645518e-06,
+      "loss": 1.048,
+      "step": 2778
+    },
+    {
+      "epoch": 2.2502024291497977,
+      "grad_norm": 1.6998697843304018,
+      "learning_rate": 3.0037760148566874e-06,
+      "loss": 0.9779,
+      "step": 2779
+    },
+    {
+      "epoch": 2.251012145748988,
+      "grad_norm": 1.7822427598135133,
+      "learning_rate": 2.99763712785516e-06,
+      "loss": 0.8956,
+      "step": 2780
+    },
+    {
+      "epoch": 2.2518218623481783,
+      "grad_norm": 1.6755771799162749,
+      "learning_rate": 2.9915034141763234e-06,
+      "loss": 0.9552,
+      "step": 2781
+    },
+    {
+      "epoch": 2.2526315789473683,
+      "grad_norm": 1.7432840821094,
+      "learning_rate": 2.9853748783517435e-06,
+      "loss": 1.0165,
+      "step": 2782
+    },
+    {
+      "epoch": 2.253441295546559,
+      "grad_norm": 1.7142235267700365,
+      "learning_rate": 2.9792515249091657e-06,
+      "loss": 1.0134,
+      "step": 2783
+    },
+    {
+      "epoch": 2.254251012145749,
+      "grad_norm": 1.7197547849869146,
+      "learning_rate": 2.973133358372504e-06,
+      "loss": 1.0114,
+      "step": 2784
+    },
+    {
+      "epoch": 2.2550607287449393,
+      "grad_norm": 1.6535591134631202,
+      "learning_rate": 2.967020383261834e-06,
+      "loss": 1.0338,
+      "step": 2785
+    },
+    {
+      "epoch": 2.2558704453441294,
+      "grad_norm": 1.5835662400274373,
+      "learning_rate": 2.960912604093409e-06,
+      "loss": 1.0032,
+      "step": 2786
+    },
+    {
+      "epoch": 2.25668016194332,
+      "grad_norm": 1.5736860921183824,
+      "learning_rate": 2.954810025379633e-06,
+      "loss": 1.0048,
+      "step": 2787
+    },
+    {
+      "epoch": 2.25748987854251,
+      "grad_norm": 1.7772332085899527,
+      "learning_rate": 2.948712651629071e-06,
+      "loss": 0.9409,
+      "step": 2788
+    },
+    {
+      "epoch": 2.2582995951417004,
+      "grad_norm": 1.8153302390569306,
+      "learning_rate": 2.9426204873464414e-06,
+      "loss": 0.9485,
+      "step": 2789
+    },
+    {
+      "epoch": 2.2591093117408905,
+      "grad_norm": 1.6435844735478726,
+      "learning_rate": 2.9365335370326143e-06,
+      "loss": 0.9743,
+      "step": 2790
+    },
+    {
+      "epoch": 2.259919028340081,
+      "grad_norm": 1.7393090327566456,
+      "learning_rate": 2.9304518051846143e-06,
+      "loss": 0.9736,
+      "step": 2791
+    },
+    {
+      "epoch": 2.2607287449392715,
+      "grad_norm": 1.6924799124970187,
+      "learning_rate": 2.924375296295597e-06,
+      "loss": 0.9779,
+      "step": 2792
+    },
+    {
+      "epoch": 2.2615384615384615,
+      "grad_norm": 1.6747668468696226,
+      "learning_rate": 2.9183040148548757e-06,
+      "loss": 0.9792,
+      "step": 2793
+    },
+    {
+      "epoch": 2.262348178137652,
+      "grad_norm": 1.8608821257094024,
+      "learning_rate": 2.9122379653478894e-06,
+      "loss": 0.9717,
+      "step": 2794
+    },
+    {
+      "epoch": 2.263157894736842,
+      "grad_norm": 1.6857679680681223,
+      "learning_rate": 2.9061771522562143e-06,
+      "loss": 0.9856,
+      "step": 2795
+    },
+    {
+      "epoch": 2.2639676113360325,
+      "grad_norm": 1.5974903308608277,
+      "learning_rate": 2.90012158005756e-06,
+      "loss": 0.9913,
+      "step": 2796
+    },
+    {
+      "epoch": 2.2647773279352226,
+      "grad_norm": 1.699823981139351,
+      "learning_rate": 2.8940712532257633e-06,
+      "loss": 0.9562,
+      "step": 2797
+    },
+    {
+      "epoch": 2.265587044534413,
+      "grad_norm": 1.6553947780834652,
+      "learning_rate": 2.8880261762307837e-06,
+      "loss": 0.9666,
+      "step": 2798
+    },
+    {
+      "epoch": 2.266396761133603,
+      "grad_norm": 1.6975659495235669,
+      "learning_rate": 2.8819863535387083e-06,
+      "loss": 0.983,
+      "step": 2799
+    },
+    {
+      "epoch": 2.2672064777327936,
+      "grad_norm": 1.6417408606225983,
+      "learning_rate": 2.875951789611734e-06,
+      "loss": 0.9544,
+      "step": 2800
+    },
+    {
+      "epoch": 2.2680161943319836,
+      "grad_norm": 1.6970530026128983,
+      "learning_rate": 2.8699224889081825e-06,
+      "loss": 1.0598,
+      "step": 2801
+    },
+    {
+      "epoch": 2.268825910931174,
+      "grad_norm": 1.6632904185152642,
+      "learning_rate": 2.8638984558824777e-06,
+      "loss": 1.0024,
+      "step": 2802
+    },
+    {
+      "epoch": 2.269635627530364,
+      "grad_norm": 1.7530525316894194,
+      "learning_rate": 2.857879694985156e-06,
+      "loss": 1.0393,
+      "step": 2803
+    },
+    {
+      "epoch": 2.2704453441295547,
+      "grad_norm": 1.6495544362471581,
+      "learning_rate": 2.851866210662858e-06,
+      "loss": 0.9628,
+      "step": 2804
+    },
+    {
+      "epoch": 2.2712550607287447,
+      "grad_norm": 1.636310387000526,
+      "learning_rate": 2.8458580073583262e-06,
+      "loss": 0.9938,
+      "step": 2805
+    },
+    {
+      "epoch": 2.272064777327935,
+      "grad_norm": 1.6690369023149698,
+      "learning_rate": 2.839855089510398e-06,
+      "loss": 0.9759,
+      "step": 2806
+    },
+    {
+      "epoch": 2.2728744939271257,
+      "grad_norm": 1.7396582090782353,
+      "learning_rate": 2.8338574615540136e-06,
+      "loss": 1.0541,
+      "step": 2807
+    },
+    {
+      "epoch": 2.2736842105263158,
+      "grad_norm": 1.6449963646693437,
+      "learning_rate": 2.827865127920203e-06,
+      "loss": 0.9446,
+      "step": 2808
+    },
+    {
+      "epoch": 2.2744939271255062,
+      "grad_norm": 1.71078655363176,
+      "learning_rate": 2.821878093036079e-06,
+      "loss": 0.9266,
+      "step": 2809
+    },
+    {
+      "epoch": 2.2753036437246963,
+      "grad_norm": 1.722249474010595,
+      "learning_rate": 2.8158963613248437e-06,
+      "loss": 0.9671,
+      "step": 2810
+    },
+    {
+      "epoch": 2.276113360323887,
+      "grad_norm": 1.6729015082048833,
+      "learning_rate": 2.8099199372057818e-06,
+      "loss": 0.995,
+      "step": 2811
+    },
+    {
+      "epoch": 2.276923076923077,
+      "grad_norm": 1.7356714743398183,
+      "learning_rate": 2.803948825094255e-06,
+      "loss": 1.0026,
+      "step": 2812
+    },
+    {
+      "epoch": 2.2777327935222673,
+      "grad_norm": 1.7087100915274263,
+      "learning_rate": 2.7979830294016985e-06,
+      "loss": 0.9608,
+      "step": 2813
+    },
+    {
+      "epoch": 2.2785425101214574,
+      "grad_norm": 1.6622618230635267,
+      "learning_rate": 2.792022554535625e-06,
+      "loss": 1.0405,
+      "step": 2814
+    },
+    {
+      "epoch": 2.279352226720648,
+      "grad_norm": 1.6282728230738879,
+      "learning_rate": 2.7860674048996174e-06,
+      "loss": 1.0188,
+      "step": 2815
+    },
+    {
+      "epoch": 2.280161943319838,
+      "grad_norm": 1.6465016952207205,
+      "learning_rate": 2.780117584893317e-06,
+      "loss": 1.0251,
+      "step": 2816
+    },
+    {
+      "epoch": 2.2809716599190284,
+      "grad_norm": 1.6700201012055247,
+      "learning_rate": 2.774173098912433e-06,
+      "loss": 0.9897,
+      "step": 2817
+    },
+    {
+      "epoch": 2.2817813765182184,
+      "grad_norm": 1.654997034971534,
+      "learning_rate": 2.76823395134873e-06,
+      "loss": 0.9931,
+      "step": 2818
+    },
+    {
+      "epoch": 2.282591093117409,
+      "grad_norm": 1.6463795592129717,
+      "learning_rate": 2.7623001465900323e-06,
+      "loss": 0.9646,
+      "step": 2819
+    },
+    {
+      "epoch": 2.283400809716599,
+      "grad_norm": 1.689096129045378,
+      "learning_rate": 2.756371689020214e-06,
+      "loss": 1.037,
+      "step": 2820
+    },
+    {
+      "epoch": 2.2842105263157895,
+      "grad_norm": 1.739455812749162,
+      "learning_rate": 2.7504485830191985e-06,
+      "loss": 1.0553,
+      "step": 2821
+    },
+    {
+      "epoch": 2.28502024291498,
+      "grad_norm": 1.737347509141222,
+      "learning_rate": 2.7445308329629593e-06,
+      "loss": 0.9492,
+      "step": 2822
+    },
+    {
+      "epoch": 2.28582995951417,
+      "grad_norm": 1.7541768615419884,
+      "learning_rate": 2.738618443223513e-06,
+      "loss": 0.954,
+      "step": 2823
+    },
+    {
+      "epoch": 2.2866396761133605,
+      "grad_norm": 1.680772296035219,
+      "learning_rate": 2.7327114181689117e-06,
+      "loss": 1.0036,
+      "step": 2824
+    },
+    {
+      "epoch": 2.2874493927125505,
+      "grad_norm": 1.7198767387432525,
+      "learning_rate": 2.7268097621632473e-06,
+      "loss": 1.0243,
+      "step": 2825
+    },
+    {
+      "epoch": 2.288259109311741,
+      "grad_norm": 1.7541843187045827,
+      "learning_rate": 2.7209134795666404e-06,
+      "loss": 0.9636,
+      "step": 2826
+    },
+    {
+      "epoch": 2.289068825910931,
+      "grad_norm": 1.7969143994647754,
+      "learning_rate": 2.715022574735249e-06,
+      "loss": 0.956,
+      "step": 2827
+    },
+    {
+      "epoch": 2.2898785425101216,
+      "grad_norm": 1.7235098932288484,
+      "learning_rate": 2.709137052021248e-06,
+      "loss": 0.9696,
+      "step": 2828
+    },
+    {
+      "epoch": 2.2906882591093116,
+      "grad_norm": 1.7147800208396093,
+      "learning_rate": 2.7032569157728503e-06,
+      "loss": 0.9945,
+      "step": 2829
+    },
+    {
+      "epoch": 2.291497975708502,
+      "grad_norm": 1.6634199908382443,
+      "learning_rate": 2.697382170334275e-06,
+      "loss": 0.9715,
+      "step": 2830
+    },
+    {
+      "epoch": 2.292307692307692,
+      "grad_norm": 1.7025316188789756,
+      "learning_rate": 2.6915128200457706e-06,
+      "loss": 0.9584,
+      "step": 2831
+    },
+    {
+      "epoch": 2.2931174089068826,
+      "grad_norm": 1.7135149326400867,
+      "learning_rate": 2.68564886924359e-06,
+      "loss": 0.9675,
+      "step": 2832
+    },
+    {
+      "epoch": 2.2939271255060727,
+      "grad_norm": 1.697368241080622,
+      "learning_rate": 2.679790322260002e-06,
+      "loss": 1.0273,
+      "step": 2833
+    },
+    {
+      "epoch": 2.294736842105263,
+      "grad_norm": 1.6304843444402075,
+      "learning_rate": 2.673937183423282e-06,
+      "loss": 0.998,
+      "step": 2834
+    },
+    {
+      "epoch": 2.2955465587044532,
+      "grad_norm": 1.6737945308046214,
+      "learning_rate": 2.6680894570577042e-06,
+      "loss": 0.9654,
+      "step": 2835
+    },
+    {
+      "epoch": 2.2963562753036437,
+      "grad_norm": 1.6851080909667857,
+      "learning_rate": 2.6622471474835585e-06,
+      "loss": 0.9605,
+      "step": 2836
+    },
+    {
+      "epoch": 2.297165991902834,
+      "grad_norm": 1.7289635381651745,
+      "learning_rate": 2.6564102590171204e-06,
+      "loss": 1.0554,
+      "step": 2837
+    },
+    {
+      "epoch": 2.2979757085020243,
+      "grad_norm": 1.6276718372129562,
+      "learning_rate": 2.6505787959706607e-06,
+      "loss": 0.9341,
+      "step": 2838
+    },
+    {
+      "epoch": 2.2987854251012148,
+      "grad_norm": 1.775958697906154,
+      "learning_rate": 2.6447527626524504e-06,
+      "loss": 0.929,
+      "step": 2839
+    },
+    {
+      "epoch": 2.299595141700405,
+      "grad_norm": 1.6271115167068193,
+      "learning_rate": 2.638932163366742e-06,
+      "loss": 1.0366,
+      "step": 2840
+    },
+    {
+      "epoch": 2.3004048582995953,
+      "grad_norm": 1.7404302828297658,
+      "learning_rate": 2.633117002413774e-06,
+      "loss": 1.0187,
+      "step": 2841
+    },
+    {
+      "epoch": 2.3012145748987853,
+      "grad_norm": 1.7172201475593398,
+      "learning_rate": 2.6273072840897685e-06,
+      "loss": 0.9908,
+      "step": 2842
+    },
+    {
+      "epoch": 2.302024291497976,
+      "grad_norm": 1.7540383203717043,
+      "learning_rate": 2.6215030126869235e-06,
+      "loss": 0.9614,
+      "step": 2843
+    },
+    {
+      "epoch": 2.302834008097166,
+      "grad_norm": 1.7302429828447257,
+      "learning_rate": 2.6157041924934223e-06,
+      "loss": 0.9773,
+      "step": 2844
+    },
+    {
+      "epoch": 2.3036437246963564,
+      "grad_norm": 1.621473177336414,
+      "learning_rate": 2.6099108277934105e-06,
+      "loss": 0.9527,
+      "step": 2845
+    },
+    {
+      "epoch": 2.3044534412955464,
+      "grad_norm": 1.6861502912702049,
+      "learning_rate": 2.604122922867004e-06,
+      "loss": 1.0025,
+      "step": 2846
+    },
+    {
+      "epoch": 2.305263157894737,
+      "grad_norm": 1.6470296585152888,
+      "learning_rate": 2.5983404819902937e-06,
+      "loss": 1.0029,
+      "step": 2847
+    },
+    {
+      "epoch": 2.306072874493927,
+      "grad_norm": 1.732391716550845,
+      "learning_rate": 2.592563509435325e-06,
+      "loss": 0.9638,
+      "step": 2848
+    },
+    {
+      "epoch": 2.3068825910931174,
+      "grad_norm": 1.6830485680896312,
+      "learning_rate": 2.586792009470107e-06,
+      "loss": 1.038,
+      "step": 2849
+    },
+    {
+      "epoch": 2.3076923076923075,
+      "grad_norm": 1.7578887787386592,
+      "learning_rate": 2.581025986358602e-06,
+      "loss": 0.9445,
+      "step": 2850
+    },
+    {
+      "epoch": 2.308502024291498,
+      "grad_norm": 1.7346747000095792,
+      "learning_rate": 2.575265444360733e-06,
+      "loss": 0.9888,
+      "step": 2851
+    },
+    {
+      "epoch": 2.3093117408906885,
+      "grad_norm": 1.7156101092464429,
+      "learning_rate": 2.5695103877323678e-06,
+      "loss": 1.0043,
+      "step": 2852
+    },
+    {
+      "epoch": 2.3101214574898785,
+      "grad_norm": 1.6144048959910942,
+      "learning_rate": 2.563760820725325e-06,
+      "loss": 1.0411,
+      "step": 2853
+    },
+    {
+      "epoch": 2.310931174089069,
+      "grad_norm": 1.6919970110753482,
+      "learning_rate": 2.5580167475873595e-06,
+      "loss": 0.9792,
+      "step": 2854
+    },
+    {
+      "epoch": 2.311740890688259,
+      "grad_norm": 1.763114954790134,
+      "learning_rate": 2.5522781725621814e-06,
+      "loss": 0.9677,
+      "step": 2855
+    },
+    {
+      "epoch": 2.3125506072874495,
+      "grad_norm": 1.7245366432759541,
+      "learning_rate": 2.5465450998894294e-06,
+      "loss": 0.9699,
+      "step": 2856
+    },
+    {
+      "epoch": 2.3133603238866396,
+      "grad_norm": 1.7077901989875472,
+      "learning_rate": 2.540817533804676e-06,
+      "loss": 1.0007,
+      "step": 2857
+    },
+    {
+      "epoch": 2.31417004048583,
+      "grad_norm": 1.6847209459419308,
+      "learning_rate": 2.535095478539428e-06,
+      "loss": 1.0254,
+      "step": 2858
+    },
+    {
+      "epoch": 2.31497975708502,
+      "grad_norm": 1.706175440021964,
+      "learning_rate": 2.529378938321124e-06,
+      "loss": 0.9536,
+      "step": 2859
+    },
+    {
+      "epoch": 2.3157894736842106,
+      "grad_norm": 1.7691390032285985,
+      "learning_rate": 2.523667917373125e-06,
+      "loss": 1.0103,
+      "step": 2860
+    },
+    {
+      "epoch": 2.3165991902834007,
+      "grad_norm": 1.8077440425603108,
+      "learning_rate": 2.517962419914712e-06,
+      "loss": 0.9207,
+      "step": 2861
+    },
+    {
+      "epoch": 2.317408906882591,
+      "grad_norm": 1.6663947281505385,
+      "learning_rate": 2.512262450161087e-06,
+      "loss": 0.9743,
+      "step": 2862
+    },
+    {
+      "epoch": 2.318218623481781,
+      "grad_norm": 1.6909579843290505,
+      "learning_rate": 2.5065680123233737e-06,
+      "loss": 0.9382,
+      "step": 2863
+    },
+    {
+      "epoch": 2.3190283400809717,
+      "grad_norm": 1.6847267233784076,
+      "learning_rate": 2.5008791106086015e-06,
+      "loss": 0.9839,
+      "step": 2864
+    },
+    {
+      "epoch": 2.3198380566801617,
+      "grad_norm": 1.7283759037557056,
+      "learning_rate": 2.4951957492197097e-06,
+      "loss": 0.9233,
+      "step": 2865
+    },
+    {
+      "epoch": 2.3206477732793522,
+      "grad_norm": 1.6108950115636136,
+      "learning_rate": 2.4895179323555517e-06,
+      "loss": 0.9222,
+      "step": 2866
+    },
+    {
+      "epoch": 2.3214574898785427,
+      "grad_norm": 1.759239040012402,
+      "learning_rate": 2.483845664210879e-06,
+      "loss": 1.0203,
+      "step": 2867
+    },
+    {
+      "epoch": 2.3222672064777328,
+      "grad_norm": 1.6569900658909889,
+      "learning_rate": 2.478178948976342e-06,
+      "loss": 1.0225,
+      "step": 2868
+    },
+    {
+      "epoch": 2.3230769230769233,
+      "grad_norm": 1.6165160420682663,
+      "learning_rate": 2.4725177908384936e-06,
+      "loss": 0.9963,
+      "step": 2869
+    },
+    {
+      "epoch": 2.3238866396761133,
+      "grad_norm": 1.6910680194151018,
+      "learning_rate": 2.4668621939797745e-06,
+      "loss": 0.9845,
+      "step": 2870
+    },
+    {
+      "epoch": 2.324696356275304,
+      "grad_norm": 1.7564925975810206,
+      "learning_rate": 2.461212162578527e-06,
+      "loss": 1.0053,
+      "step": 2871
+    },
+    {
+      "epoch": 2.325506072874494,
+      "grad_norm": 1.6736697804350984,
+      "learning_rate": 2.455567700808974e-06,
+      "loss": 1.0172,
+      "step": 2872
+    },
+    {
+      "epoch": 2.3263157894736843,
+      "grad_norm": 1.6235333292980543,
+      "learning_rate": 2.4499288128412214e-06,
+      "loss": 1.0632,
+      "step": 2873
+    },
+    {
+      "epoch": 2.3271255060728744,
+      "grad_norm": 1.7363325081342043,
+      "learning_rate": 2.4442955028412672e-06,
+      "loss": 0.9349,
+      "step": 2874
+    },
+    {
+      "epoch": 2.327935222672065,
+      "grad_norm": 1.6683416186196407,
+      "learning_rate": 2.438667774970981e-06,
+      "loss": 0.9533,
+      "step": 2875
+    },
+    {
+      "epoch": 2.328744939271255,
+      "grad_norm": 1.7103163002011812,
+      "learning_rate": 2.433045633388106e-06,
+      "loss": 0.9585,
+      "step": 2876
+    },
+    {
+      "epoch": 2.3295546558704454,
+      "grad_norm": 1.6372160886662246,
+      "learning_rate": 2.4274290822462656e-06,
+      "loss": 0.9704,
+      "step": 2877
+    },
+    {
+      "epoch": 2.3303643724696355,
+      "grad_norm": 1.7503426443559542,
+      "learning_rate": 2.4218181256949434e-06,
+      "loss": 1.0724,
+      "step": 2878
+    },
+    {
+      "epoch": 2.331174089068826,
+      "grad_norm": 1.6224949379470495,
+      "learning_rate": 2.4162127678795045e-06,
+      "loss": 0.9353,
+      "step": 2879
+    },
+    {
+      "epoch": 2.331983805668016,
+      "grad_norm": 1.7179059296387311,
+      "learning_rate": 2.4106130129411608e-06,
+      "loss": 0.9951,
+      "step": 2880
+    },
+    {
+      "epoch": 2.3327935222672065,
+      "grad_norm": 1.6820449772558062,
+      "learning_rate": 2.405018865016999e-06,
+      "loss": 1.0007,
+      "step": 2881
+    },
+    {
+      "epoch": 2.333603238866397,
+      "grad_norm": 1.640951194648763,
+      "learning_rate": 2.3994303282399544e-06,
+      "loss": 0.9456,
+      "step": 2882
+    },
+    {
+      "epoch": 2.334412955465587,
+      "grad_norm": 1.6725791781090846,
+      "learning_rate": 2.3938474067388208e-06,
+      "loss": 0.9674,
+      "step": 2883
+    },
+    {
+      "epoch": 2.3352226720647775,
+      "grad_norm": 1.8442889599518857,
+      "learning_rate": 2.388270104638242e-06,
+      "loss": 0.9674,
+      "step": 2884
+    },
+    {
+      "epoch": 2.3360323886639676,
+      "grad_norm": 1.676925947550545,
+      "learning_rate": 2.3826984260587084e-06,
+      "loss": 0.9523,
+      "step": 2885
+    },
+    {
+      "epoch": 2.336842105263158,
+      "grad_norm": 1.717243161461183,
+      "learning_rate": 2.3771323751165563e-06,
+      "loss": 0.9958,
+      "step": 2886
+    },
+    {
+      "epoch": 2.337651821862348,
+      "grad_norm": 1.7024062048223407,
+      "learning_rate": 2.3715719559239727e-06,
+      "loss": 1.0246,
+      "step": 2887
+    },
+    {
+      "epoch": 2.3384615384615386,
+      "grad_norm": 1.6526087755299945,
+      "learning_rate": 2.3660171725889703e-06,
+      "loss": 0.9659,
+      "step": 2888
+    },
+    {
+      "epoch": 2.3392712550607286,
+      "grad_norm": 1.6804957380379042,
+      "learning_rate": 2.360468029215409e-06,
+      "loss": 1.0254,
+      "step": 2889
+    },
+    {
+      "epoch": 2.340080971659919,
+      "grad_norm": 1.6575687386773823,
+      "learning_rate": 2.354924529902978e-06,
+      "loss": 1.0599,
+      "step": 2890
+    },
+    {
+      "epoch": 2.340890688259109,
+      "grad_norm": 1.6411725270506965,
+      "learning_rate": 2.349386678747194e-06,
+      "loss": 1.0325,
+      "step": 2891
+    },
+    {
+      "epoch": 2.3417004048582997,
+      "grad_norm": 1.6593350291881466,
+      "learning_rate": 2.343854479839405e-06,
+      "loss": 0.9796,
+      "step": 2892
+    },
+    {
+      "epoch": 2.3425101214574897,
+      "grad_norm": 1.6411951141419265,
+      "learning_rate": 2.3383279372667787e-06,
+      "loss": 0.9762,
+      "step": 2893
+    },
+    {
+      "epoch": 2.34331983805668,
+      "grad_norm": 1.6675022296883175,
+      "learning_rate": 2.332807055112306e-06,
+      "loss": 0.9739,
+      "step": 2894
+    },
+    {
+      "epoch": 2.3441295546558703,
+      "grad_norm": 1.6845312790330145,
+      "learning_rate": 2.327291837454799e-06,
+      "loss": 0.9776,
+      "step": 2895
+    },
+    {
+      "epoch": 2.3449392712550607,
+      "grad_norm": 1.6631177985983452,
+      "learning_rate": 2.3217822883688855e-06,
+      "loss": 0.9856,
+      "step": 2896
+    },
+    {
+      "epoch": 2.3457489878542512,
+      "grad_norm": 1.7002279127037514,
+      "learning_rate": 2.316278411924998e-06,
+      "loss": 1.004,
+      "step": 2897
+    },
+    {
+      "epoch": 2.3465587044534413,
+      "grad_norm": 1.7044705523610206,
+      "learning_rate": 2.310780212189384e-06,
+      "loss": 0.9071,
+      "step": 2898
+    },
+    {
+      "epoch": 2.3473684210526318,
+      "grad_norm": 1.7588763360318254,
+      "learning_rate": 2.3052876932240943e-06,
+      "loss": 0.939,
+      "step": 2899
+    },
+    {
+      "epoch": 2.348178137651822,
+      "grad_norm": 1.6865532061555057,
+      "learning_rate": 2.2998008590869838e-06,
+      "loss": 0.9706,
+      "step": 2900
+    },
+    {
+      "epoch": 2.3489878542510123,
+      "grad_norm": 1.8099761327981618,
+      "learning_rate": 2.294319713831705e-06,
+      "loss": 0.9509,
+      "step": 2901
+    },
+    {
+      "epoch": 2.3497975708502024,
+      "grad_norm": 1.655965409511294,
+      "learning_rate": 2.2888442615077145e-06,
+      "loss": 0.9411,
+      "step": 2902
+    },
+    {
+      "epoch": 2.350607287449393,
+      "grad_norm": 1.661395531775376,
+      "learning_rate": 2.2833745061602587e-06,
+      "loss": 1.011,
+      "step": 2903
+    },
+    {
+      "epoch": 2.351417004048583,
+      "grad_norm": 1.6324577056900818,
+      "learning_rate": 2.277910451830373e-06,
+      "loss": 1.0316,
+      "step": 2904
+    },
+    {
+      "epoch": 2.3522267206477734,
+      "grad_norm": 1.6836043670245235,
+      "learning_rate": 2.2724521025548828e-06,
+      "loss": 1.0146,
+      "step": 2905
+    },
+    {
+      "epoch": 2.3530364372469634,
+      "grad_norm": 1.7435287774310333,
+      "learning_rate": 2.2669994623664006e-06,
+      "loss": 0.9744,
+      "step": 2906
+    },
+    {
+      "epoch": 2.353846153846154,
+      "grad_norm": 1.7490120988400564,
+      "learning_rate": 2.2615525352933156e-06,
+      "loss": 0.9675,
+      "step": 2907
+    },
+    {
+      "epoch": 2.354655870445344,
+      "grad_norm": 1.6560540854955874,
+      "learning_rate": 2.256111325359801e-06,
+      "loss": 0.9834,
+      "step": 2908
+    },
+    {
+      "epoch": 2.3554655870445345,
+      "grad_norm": 1.6937685129731164,
+      "learning_rate": 2.250675836585803e-06,
+      "loss": 1.0068,
+      "step": 2909
+    },
+    {
+      "epoch": 2.3562753036437245,
+      "grad_norm": 1.6622436941899015,
+      "learning_rate": 2.245246072987045e-06,
+      "loss": 0.9918,
+      "step": 2910
+    },
+    {
+      "epoch": 2.357085020242915,
+      "grad_norm": 1.7854484056765068,
+      "learning_rate": 2.2398220385750213e-06,
+      "loss": 1.0529,
+      "step": 2911
+    },
+    {
+      "epoch": 2.3578947368421055,
+      "grad_norm": 1.6881598706045204,
+      "learning_rate": 2.234403737356987e-06,
+      "loss": 0.9479,
+      "step": 2912
+    },
+    {
+      "epoch": 2.3587044534412955,
+      "grad_norm": 1.7582794873521799,
+      "learning_rate": 2.228991173335967e-06,
+      "loss": 1.0117,
+      "step": 2913
+    },
+    {
+      "epoch": 2.3595141700404856,
+      "grad_norm": 1.6302584416633792,
+      "learning_rate": 2.2235843505107447e-06,
+      "loss": 0.9808,
+      "step": 2914
+    },
+    {
+      "epoch": 2.360323886639676,
+      "grad_norm": 1.741990356965201,
+      "learning_rate": 2.2181832728758635e-06,
+      "loss": 0.9415,
+      "step": 2915
+    },
+    {
+      "epoch": 2.3611336032388666,
+      "grad_norm": 1.7084803772752033,
+      "learning_rate": 2.2127879444216184e-06,
+      "loss": 1.0223,
+      "step": 2916
+    },
+    {
+      "epoch": 2.3619433198380566,
+      "grad_norm": 1.680568814995532,
+      "learning_rate": 2.2073983691340673e-06,
+      "loss": 0.9924,
+      "step": 2917
+    },
+    {
+      "epoch": 2.362753036437247,
+      "grad_norm": 1.6884873524658965,
+      "learning_rate": 2.202014550995003e-06,
+      "loss": 0.946,
+      "step": 2918
+    },
+    {
+      "epoch": 2.363562753036437,
+      "grad_norm": 1.6973802116265153,
+      "learning_rate": 2.1966364939819797e-06,
+      "loss": 0.9663,
+      "step": 2919
+    },
+    {
+      "epoch": 2.3643724696356276,
+      "grad_norm": 1.6489385262541483,
+      "learning_rate": 2.191264202068286e-06,
+      "loss": 0.985,
+      "step": 2920
+    },
+    {
+      "epoch": 2.3651821862348177,
+      "grad_norm": 1.6792446355838644,
+      "learning_rate": 2.185897679222951e-06,
+      "loss": 0.9868,
+      "step": 2921
+    },
+    {
+      "epoch": 2.365991902834008,
+      "grad_norm": 1.6804030489521116,
+      "learning_rate": 2.180536929410748e-06,
+      "loss": 1.0269,
+      "step": 2922
+    },
+    {
+      "epoch": 2.3668016194331982,
+      "grad_norm": 1.6589791003331682,
+      "learning_rate": 2.1751819565921774e-06,
+      "loss": 0.9795,
+      "step": 2923
+    },
+    {
+      "epoch": 2.3676113360323887,
+      "grad_norm": 1.6922330016889355,
+      "learning_rate": 2.169832764723475e-06,
+      "loss": 0.9302,
+      "step": 2924
+    },
+    {
+      "epoch": 2.3684210526315788,
+      "grad_norm": 1.617132099924979,
+      "learning_rate": 2.1644893577566118e-06,
+      "loss": 0.9114,
+      "step": 2925
+    },
+    {
+      "epoch": 2.3692307692307693,
+      "grad_norm": 1.683358875672709,
+      "learning_rate": 2.1591517396392735e-06,
+      "loss": 1.0196,
+      "step": 2926
+    },
+    {
+      "epoch": 2.3700404858299597,
+      "grad_norm": 1.662270804025337,
+      "learning_rate": 2.15381991431488e-06,
+      "loss": 0.9346,
+      "step": 2927
+    },
+    {
+      "epoch": 2.37085020242915,
+      "grad_norm": 1.5560767626917962,
+      "learning_rate": 2.1484938857225636e-06,
+      "loss": 1.0323,
+      "step": 2928
+    },
+    {
+      "epoch": 2.37165991902834,
+      "grad_norm": 1.6168560582415574,
+      "learning_rate": 2.1431736577971763e-06,
+      "loss": 1.0005,
+      "step": 2929
+    },
+    {
+      "epoch": 2.3724696356275303,
+      "grad_norm": 1.6890995844263526,
+      "learning_rate": 2.137859234469286e-06,
+      "loss": 1.0447,
+      "step": 2930
+    },
+    {
+      "epoch": 2.373279352226721,
+      "grad_norm": 1.6394854020625869,
+      "learning_rate": 2.132550619665168e-06,
+      "loss": 0.9983,
+      "step": 2931
+    },
+    {
+      "epoch": 2.374089068825911,
+      "grad_norm": 1.6603872171093927,
+      "learning_rate": 2.1272478173068147e-06,
+      "loss": 0.9744,
+      "step": 2932
+    },
+    {
+      "epoch": 2.3748987854251014,
+      "grad_norm": 1.6252945957266938,
+      "learning_rate": 2.1219508313119163e-06,
+      "loss": 0.947,
+      "step": 2933
+    },
+    {
+      "epoch": 2.3757085020242914,
+      "grad_norm": 1.7572254760843087,
+      "learning_rate": 2.1166596655938676e-06,
+      "loss": 0.9481,
+      "step": 2934
+    },
+    {
+      "epoch": 2.376518218623482,
+      "grad_norm": 1.680859166288277,
+      "learning_rate": 2.1113743240617668e-06,
+      "loss": 0.9892,
+      "step": 2935
+    },
+    {
+      "epoch": 2.377327935222672,
+      "grad_norm": 1.7086071225889468,
+      "learning_rate": 2.1060948106204072e-06,
+      "loss": 1.0586,
+      "step": 2936
+    },
+    {
+      "epoch": 2.3781376518218624,
+      "grad_norm": 1.6757697718883062,
+      "learning_rate": 2.100821129170274e-06,
+      "loss": 0.9855,
+      "step": 2937
+    },
+    {
+      "epoch": 2.3789473684210525,
+      "grad_norm": 1.6316446681345318,
+      "learning_rate": 2.0955532836075445e-06,
+      "loss": 1.0192,
+      "step": 2938
+    },
+    {
+      "epoch": 2.379757085020243,
+      "grad_norm": 1.6928291171366985,
+      "learning_rate": 2.090291277824089e-06,
+      "loss": 0.974,
+      "step": 2939
+    },
+    {
+      "epoch": 2.380566801619433,
+      "grad_norm": 1.6550872597895547,
+      "learning_rate": 2.08503511570746e-06,
+      "loss": 1.069,
+      "step": 2940
+    },
+    {
+      "epoch": 2.3813765182186235,
+      "grad_norm": 1.6475664661450247,
+      "learning_rate": 2.0797848011408906e-06,
+      "loss": 0.9001,
+      "step": 2941
+    },
+    {
+      "epoch": 2.382186234817814,
+      "grad_norm": 1.6992711495500126,
+      "learning_rate": 2.0745403380032947e-06,
+      "loss": 0.9789,
+      "step": 2942
+    },
+    {
+      "epoch": 2.382995951417004,
+      "grad_norm": 1.7621872944691108,
+      "learning_rate": 2.0693017301692698e-06,
+      "loss": 1.0104,
+      "step": 2943
+    },
+    {
+      "epoch": 2.383805668016194,
+      "grad_norm": 1.6841266297477848,
+      "learning_rate": 2.0640689815090777e-06,
+      "loss": 0.9808,
+      "step": 2944
+    },
+    {
+      "epoch": 2.3846153846153846,
+      "grad_norm": 1.7074153273587775,
+      "learning_rate": 2.058842095888658e-06,
+      "loss": 0.9962,
+      "step": 2945
+    },
+    {
+      "epoch": 2.385425101214575,
+      "grad_norm": 1.6367366387306832,
+      "learning_rate": 2.053621077169613e-06,
+      "loss": 0.9874,
+      "step": 2946
+    },
+    {
+      "epoch": 2.386234817813765,
+      "grad_norm": 1.6204049507354517,
+      "learning_rate": 2.0484059292092196e-06,
+      "loss": 1.047,
+      "step": 2947
+    },
+    {
+      "epoch": 2.3870445344129556,
+      "grad_norm": 1.6516699596226871,
+      "learning_rate": 2.0431966558604097e-06,
+      "loss": 1.0036,
+      "step": 2948
+    },
+    {
+      "epoch": 2.3878542510121457,
+      "grad_norm": 1.630232092481421,
+      "learning_rate": 2.0379932609717767e-06,
+      "loss": 1.016,
+      "step": 2949
+    },
+    {
+      "epoch": 2.388663967611336,
+      "grad_norm": 1.6118486578131621,
+      "learning_rate": 2.0327957483875693e-06,
+      "loss": 1.0195,
+      "step": 2950
+    },
+    {
+      "epoch": 2.389473684210526,
+      "grad_norm": 1.679620212591038,
+      "learning_rate": 2.0276041219476985e-06,
+      "loss": 1.0092,
+      "step": 2951
+    },
+    {
+      "epoch": 2.3902834008097167,
+      "grad_norm": 1.6192037143814413,
+      "learning_rate": 2.0224183854877165e-06,
+      "loss": 1.0289,
+      "step": 2952
+    },
+    {
+      "epoch": 2.3910931174089067,
+      "grad_norm": 1.6913936221204569,
+      "learning_rate": 2.0172385428388288e-06,
+      "loss": 0.9022,
+      "step": 2953
+    },
+    {
+      "epoch": 2.3919028340080972,
+      "grad_norm": 1.712872458206929,
+      "learning_rate": 2.0120645978278887e-06,
+      "loss": 0.9932,
+      "step": 2954
+    },
+    {
+      "epoch": 2.3927125506072873,
+      "grad_norm": 1.6358216619621548,
+      "learning_rate": 2.006896554277388e-06,
+      "loss": 0.9879,
+      "step": 2955
+    },
+    {
+      "epoch": 2.3935222672064778,
+      "grad_norm": 1.7962624248609893,
+      "learning_rate": 2.0017344160054597e-06,
+      "loss": 0.9944,
+      "step": 2956
+    },
+    {
+      "epoch": 2.3943319838056683,
+      "grad_norm": 1.7097327349314781,
+      "learning_rate": 1.996578186825876e-06,
+      "loss": 0.9837,
+      "step": 2957
+    },
+    {
+      "epoch": 2.3951417004048583,
+      "grad_norm": 1.7984710742922152,
+      "learning_rate": 1.991427870548038e-06,
+      "loss": 1.0043,
+      "step": 2958
+    },
+    {
+      "epoch": 2.3959514170040483,
+      "grad_norm": 1.7633179302331299,
+      "learning_rate": 1.9862834709769897e-06,
+      "loss": 0.9943,
+      "step": 2959
+    },
+    {
+      "epoch": 2.396761133603239,
+      "grad_norm": 1.6081821375494785,
+      "learning_rate": 1.981144991913392e-06,
+      "loss": 1.0207,
+      "step": 2960
+    },
+    {
+      "epoch": 2.3975708502024293,
+      "grad_norm": 1.6574008938350164,
+      "learning_rate": 1.976012437153535e-06,
+      "loss": 1.0171,
+      "step": 2961
+    },
+    {
+      "epoch": 2.3983805668016194,
+      "grad_norm": 1.6854338035743857,
+      "learning_rate": 1.970885810489337e-06,
+      "loss": 0.9726,
+      "step": 2962
+    },
+    {
+      "epoch": 2.39919028340081,
+      "grad_norm": 1.6012012823118715,
+      "learning_rate": 1.9657651157083322e-06,
+      "loss": 0.9743,
+      "step": 2963
+    },
+    {
+      "epoch": 2.4,
+      "grad_norm": 3.279317777475881,
+      "learning_rate": 1.960650356593672e-06,
+      "loss": 1.0349,
+      "step": 2964
+    },
+    {
+      "epoch": 2.4008097165991904,
+      "grad_norm": 1.6559506921615998,
+      "learning_rate": 1.9555415369241228e-06,
+      "loss": 1.0237,
+      "step": 2965
+    },
+    {
+      "epoch": 2.4016194331983804,
+      "grad_norm": 1.696563890993013,
+      "learning_rate": 1.9504386604740632e-06,
+      "loss": 1.034,
+      "step": 2966
+    },
+    {
+      "epoch": 2.402429149797571,
+      "grad_norm": 1.6229320190285514,
+      "learning_rate": 1.9453417310134857e-06,
+      "loss": 0.9817,
+      "step": 2967
+    },
+    {
+      "epoch": 2.403238866396761,
+      "grad_norm": 1.717120363215466,
+      "learning_rate": 1.9402507523079794e-06,
+      "loss": 0.9337,
+      "step": 2968
+    },
+    {
+      "epoch": 2.4040485829959515,
+      "grad_norm": 1.8037959085157829,
+      "learning_rate": 1.9351657281187484e-06,
+      "loss": 0.9729,
+      "step": 2969
+    },
+    {
+      "epoch": 2.4048582995951415,
+      "grad_norm": 1.6684409182222344,
+      "learning_rate": 1.930086662202589e-06,
+      "loss": 1.0042,
+      "step": 2970
+    },
+    {
+      "epoch": 2.405668016194332,
+      "grad_norm": 1.698807530710582,
+      "learning_rate": 1.9250135583119e-06,
+      "loss": 0.9531,
+      "step": 2971
+    },
+    {
+      "epoch": 2.4064777327935225,
+      "grad_norm": 1.6713412841686315,
+      "learning_rate": 1.9199464201946717e-06,
+      "loss": 0.9905,
+      "step": 2972
+    },
+    {
+      "epoch": 2.4072874493927126,
+      "grad_norm": 1.6587820441376417,
+      "learning_rate": 1.9148852515944893e-06,
+      "loss": 0.9788,
+      "step": 2973
+    },
+    {
+      "epoch": 2.4080971659919026,
+      "grad_norm": 1.6459864544000329,
+      "learning_rate": 1.9098300562505266e-06,
+      "loss": 0.9819,
+      "step": 2974
+    },
+    {
+      "epoch": 2.408906882591093,
+      "grad_norm": 1.6811137602745856,
+      "learning_rate": 1.9047808378975485e-06,
+      "loss": 1.0037,
+      "step": 2975
+    },
+    {
+      "epoch": 2.4097165991902836,
+      "grad_norm": 1.6452093088766784,
+      "learning_rate": 1.8997376002658974e-06,
+      "loss": 0.9404,
+      "step": 2976
+    },
+    {
+      "epoch": 2.4105263157894736,
+      "grad_norm": 1.7009168710382707,
+      "learning_rate": 1.894700347081505e-06,
+      "loss": 0.9519,
+      "step": 2977
+    },
+    {
+      "epoch": 2.411336032388664,
+      "grad_norm": 1.6634076517052085,
+      "learning_rate": 1.8896690820658758e-06,
+      "loss": 1.0487,
+      "step": 2978
+    },
+    {
+      "epoch": 2.412145748987854,
+      "grad_norm": 1.6737820388111349,
+      "learning_rate": 1.8846438089360896e-06,
+      "loss": 1.0075,
+      "step": 2979
+    },
+    {
+      "epoch": 2.4129554655870447,
+      "grad_norm": 1.7249507058482862,
+      "learning_rate": 1.8796245314048046e-06,
+      "loss": 1.0098,
+      "step": 2980
+    },
+    {
+      "epoch": 2.4137651821862347,
+      "grad_norm": 1.7251783475658493,
+      "learning_rate": 1.874611253180244e-06,
+      "loss": 0.9775,
+      "step": 2981
+    },
+    {
+      "epoch": 2.414574898785425,
+      "grad_norm": 1.6136959740020433,
+      "learning_rate": 1.8696039779662012e-06,
+      "loss": 0.9962,
+      "step": 2982
+    },
+    {
+      "epoch": 2.4153846153846152,
+      "grad_norm": 1.6425732401790396,
+      "learning_rate": 1.8646027094620345e-06,
+      "loss": 0.9816,
+      "step": 2983
+    },
+    {
+      "epoch": 2.4161943319838057,
+      "grad_norm": 1.8229989217932228,
+      "learning_rate": 1.8596074513626694e-06,
+      "loss": 0.9151,
+      "step": 2984
+    },
+    {
+      "epoch": 2.417004048582996,
+      "grad_norm": 1.7061703390123246,
+      "learning_rate": 1.8546182073585828e-06,
+      "loss": 1.0206,
+      "step": 2985
+    },
+    {
+      "epoch": 2.4178137651821863,
+      "grad_norm": 1.6590284413749856,
+      "learning_rate": 1.8496349811358116e-06,
+      "loss": 1.0051,
+      "step": 2986
+    },
+    {
+      "epoch": 2.4186234817813768,
+      "grad_norm": 1.7189358586751335,
+      "learning_rate": 1.8446577763759478e-06,
+      "loss": 0.8961,
+      "step": 2987
+    },
+    {
+      "epoch": 2.419433198380567,
+      "grad_norm": 1.7082228107534516,
+      "learning_rate": 1.8396865967561317e-06,
+      "loss": 1.0398,
+      "step": 2988
+    },
+    {
+      "epoch": 2.420242914979757,
+      "grad_norm": 1.6578774454136742,
+      "learning_rate": 1.8347214459490548e-06,
+      "loss": 0.9466,
+      "step": 2989
+    },
+    {
+      "epoch": 2.4210526315789473,
+      "grad_norm": 1.7186094993741328,
+      "learning_rate": 1.829762327622958e-06,
+      "loss": 1.0007,
+      "step": 2990
+    },
+    {
+      "epoch": 2.421862348178138,
+      "grad_norm": 1.7016828071558703,
+      "learning_rate": 1.8248092454416166e-06,
+      "loss": 1.0301,
+      "step": 2991
+    },
+    {
+      "epoch": 2.422672064777328,
+      "grad_norm": 1.7081136775369437,
+      "learning_rate": 1.8198622030643564e-06,
+      "loss": 0.9879,
+      "step": 2992
+    },
+    {
+      "epoch": 2.4234817813765184,
+      "grad_norm": 1.7205639728799287,
+      "learning_rate": 1.814921204146033e-06,
+      "loss": 0.9609,
+      "step": 2993
+    },
+    {
+      "epoch": 2.4242914979757084,
+      "grad_norm": 1.6612183377575551,
+      "learning_rate": 1.8099862523370415e-06,
+      "loss": 0.9652,
+      "step": 2994
+    },
+    {
+      "epoch": 2.425101214574899,
+      "grad_norm": 1.674361206817772,
+      "learning_rate": 1.805057351283307e-06,
+      "loss": 0.9703,
+      "step": 2995
+    },
+    {
+      "epoch": 2.425910931174089,
+      "grad_norm": 1.5856102512600692,
+      "learning_rate": 1.8001345046262864e-06,
+      "loss": 0.9733,
+      "step": 2996
+    },
+    {
+      "epoch": 2.4267206477732794,
+      "grad_norm": 1.63966324613164,
+      "learning_rate": 1.7952177160029594e-06,
+      "loss": 1.0133,
+      "step": 2997
+    },
+    {
+      "epoch": 2.4275303643724695,
+      "grad_norm": 1.6634576125787222,
+      "learning_rate": 1.7903069890458347e-06,
+      "loss": 0.9702,
+      "step": 2998
+    },
+    {
+      "epoch": 2.42834008097166,
+      "grad_norm": 1.7282172413441768,
+      "learning_rate": 1.7854023273829467e-06,
+      "loss": 0.8979,
+      "step": 2999
+    },
+    {
+      "epoch": 2.42914979757085,
+      "grad_norm": 1.6699414779639086,
+      "learning_rate": 1.7805037346378384e-06,
+      "loss": 0.9921,
+      "step": 3000
+    },
+    {
+      "epoch": 2.4299595141700405,
+      "grad_norm": 1.6478445275579376,
+      "learning_rate": 1.7756112144295745e-06,
+      "loss": 1.0098,
+      "step": 3001
+    },
+    {
+      "epoch": 2.430769230769231,
+      "grad_norm": 1.7177993329143486,
+      "learning_rate": 1.7707247703727325e-06,
+      "loss": 0.9902,
+      "step": 3002
+    },
+    {
+      "epoch": 2.431578947368421,
+      "grad_norm": 1.6717118318062545,
+      "learning_rate": 1.7658444060774028e-06,
+      "loss": 0.9317,
+      "step": 3003
+    },
+    {
+      "epoch": 2.432388663967611,
+      "grad_norm": 1.656893954893965,
+      "learning_rate": 1.7609701251491796e-06,
+      "loss": 1.0093,
+      "step": 3004
+    },
+    {
+      "epoch": 2.4331983805668016,
+      "grad_norm": 1.672458797771242,
+      "learning_rate": 1.756101931189169e-06,
+      "loss": 0.994,
+      "step": 3005
+    },
+    {
+      "epoch": 2.434008097165992,
+      "grad_norm": 1.7979249229953589,
+      "learning_rate": 1.7512398277939735e-06,
+      "loss": 0.9501,
+      "step": 3006
+    },
+    {
+      "epoch": 2.434817813765182,
+      "grad_norm": 1.7675347987379522,
+      "learning_rate": 1.7463838185557024e-06,
+      "loss": 0.9573,
+      "step": 3007
+    },
+    {
+      "epoch": 2.4356275303643726,
+      "grad_norm": 1.6829178005541199,
+      "learning_rate": 1.7415339070619586e-06,
+      "loss": 0.9926,
+      "step": 3008
+    },
+    {
+      "epoch": 2.4364372469635627,
+      "grad_norm": 1.7579588916573732,
+      "learning_rate": 1.73669009689584e-06,
+      "loss": 0.9126,
+      "step": 3009
+    },
+    {
+      "epoch": 2.437246963562753,
+      "grad_norm": 1.739729410392981,
+      "learning_rate": 1.7318523916359376e-06,
+      "loss": 0.9609,
+      "step": 3010
+    },
+    {
+      "epoch": 2.438056680161943,
+      "grad_norm": 1.6681245792489996,
+      "learning_rate": 1.7270207948563323e-06,
+      "loss": 1.0533,
+      "step": 3011
+    },
+    {
+      "epoch": 2.4388663967611337,
+      "grad_norm": 1.696281337597073,
+      "learning_rate": 1.7221953101265888e-06,
+      "loss": 0.9251,
+      "step": 3012
+    },
+    {
+      "epoch": 2.4396761133603238,
+      "grad_norm": 1.6189610965490386,
+      "learning_rate": 1.7173759410117663e-06,
+      "loss": 0.9526,
+      "step": 3013
+    },
+    {
+      "epoch": 2.4404858299595142,
+      "grad_norm": 1.724367632228813,
+      "learning_rate": 1.7125626910723915e-06,
+      "loss": 0.9634,
+      "step": 3014
+    },
+    {
+      "epoch": 2.4412955465587043,
+      "grad_norm": 1.6773880195931092,
+      "learning_rate": 1.7077555638644838e-06,
+      "loss": 0.9881,
+      "step": 3015
+    },
+    {
+      "epoch": 2.442105263157895,
+      "grad_norm": 1.7197370339372542,
+      "learning_rate": 1.7029545629395306e-06,
+      "loss": 0.9735,
+      "step": 3016
+    },
+    {
+      "epoch": 2.4429149797570853,
+      "grad_norm": 1.7700548528489848,
+      "learning_rate": 1.6981596918444953e-06,
+      "loss": 1.0035,
+      "step": 3017
+    },
+    {
+      "epoch": 2.4437246963562753,
+      "grad_norm": 1.7118799526231616,
+      "learning_rate": 1.6933709541218125e-06,
+      "loss": 1.0057,
+      "step": 3018
+    },
+    {
+      "epoch": 2.4445344129554654,
+      "grad_norm": 1.6696288935408505,
+      "learning_rate": 1.6885883533093839e-06,
+      "loss": 1.0043,
+      "step": 3019
+    },
+    {
+      "epoch": 2.445344129554656,
+      "grad_norm": 1.6491620840254673,
+      "learning_rate": 1.6838118929405856e-06,
+      "loss": 0.9407,
+      "step": 3020
+    },
+    {
+      "epoch": 2.4461538461538463,
+      "grad_norm": 1.659428974002118,
+      "learning_rate": 1.6790415765442458e-06,
+      "loss": 1.0047,
+      "step": 3021
+    },
+    {
+      "epoch": 2.4469635627530364,
+      "grad_norm": 1.6354681687186046,
+      "learning_rate": 1.6742774076446578e-06,
+      "loss": 0.986,
+      "step": 3022
+    },
+    {
+      "epoch": 2.447773279352227,
+      "grad_norm": 1.6775886642207607,
+      "learning_rate": 1.6695193897615781e-06,
+      "loss": 0.9518,
+      "step": 3023
+    },
+    {
+      "epoch": 2.448582995951417,
+      "grad_norm": 1.6648950365114745,
+      "learning_rate": 1.6647675264102126e-06,
+      "loss": 0.9382,
+      "step": 3024
+    },
+    {
+      "epoch": 2.4493927125506074,
+      "grad_norm": 1.646166538722141,
+      "learning_rate": 1.660021821101222e-06,
+      "loss": 0.998,
+      "step": 3025
+    },
+    {
+      "epoch": 2.4502024291497975,
+      "grad_norm": 1.6268027966670777,
+      "learning_rate": 1.6552822773407163e-06,
+      "loss": 1.0009,
+      "step": 3026
+    },
+    {
+      "epoch": 2.451012145748988,
+      "grad_norm": 1.6955695087440212,
+      "learning_rate": 1.6505488986302586e-06,
+      "loss": 1.0315,
+      "step": 3027
+    },
+    {
+      "epoch": 2.451821862348178,
+      "grad_norm": 1.6851857353319628,
+      "learning_rate": 1.645821688466851e-06,
+      "loss": 0.9873,
+      "step": 3028
+    },
+    {
+      "epoch": 2.4526315789473685,
+      "grad_norm": 1.6305706879900275,
+      "learning_rate": 1.6411006503429428e-06,
+      "loss": 0.912,
+      "step": 3029
+    },
+    {
+      "epoch": 2.4534412955465585,
+      "grad_norm": 1.6588245540634141,
+      "learning_rate": 1.6363857877464161e-06,
+      "loss": 1.0123,
+      "step": 3030
+    },
+    {
+      "epoch": 2.454251012145749,
+      "grad_norm": 1.6911002514254574,
+      "learning_rate": 1.6316771041606027e-06,
+      "loss": 0.9738,
+      "step": 3031
+    },
+    {
+      "epoch": 2.455060728744939,
+      "grad_norm": 1.7042305673840112,
+      "learning_rate": 1.6269746030642607e-06,
+      "loss": 0.9485,
+      "step": 3032
+    },
+    {
+      "epoch": 2.4558704453441296,
+      "grad_norm": 1.620560078412615,
+      "learning_rate": 1.6222782879315802e-06,
+      "loss": 0.9963,
+      "step": 3033
+    },
+    {
+      "epoch": 2.4566801619433196,
+      "grad_norm": 1.6956591304380106,
+      "learning_rate": 1.6175881622321832e-06,
+      "loss": 0.9385,
+      "step": 3034
+    },
+    {
+      "epoch": 2.45748987854251,
+      "grad_norm": 1.710112971818022,
+      "learning_rate": 1.6129042294311227e-06,
+      "loss": 0.9377,
+      "step": 3035
+    },
+    {
+      "epoch": 2.4582995951417006,
+      "grad_norm": 1.7103868534898998,
+      "learning_rate": 1.6082264929888702e-06,
+      "loss": 0.997,
+      "step": 3036
+    },
+    {
+      "epoch": 2.4591093117408906,
+      "grad_norm": 1.5708740432232011,
+      "learning_rate": 1.603554956361324e-06,
+      "loss": 0.9684,
+      "step": 3037
+    },
+    {
+      "epoch": 2.459919028340081,
+      "grad_norm": 1.7023113258556464,
+      "learning_rate": 1.5988896229997952e-06,
+      "loss": 0.9485,
+      "step": 3038
+    },
+    {
+      "epoch": 2.460728744939271,
+      "grad_norm": 1.709049003849265,
+      "learning_rate": 1.5942304963510236e-06,
+      "loss": 0.9504,
+      "step": 3039
+    },
+    {
+      "epoch": 2.4615384615384617,
+      "grad_norm": 1.7137018057495825,
+      "learning_rate": 1.5895775798571523e-06,
+      "loss": 0.962,
+      "step": 3040
+    },
+    {
+      "epoch": 2.4623481781376517,
+      "grad_norm": 1.6578218057308711,
+      "learning_rate": 1.5849308769557393e-06,
+      "loss": 0.9717,
+      "step": 3041
+    },
+    {
+      "epoch": 2.463157894736842,
+      "grad_norm": 1.7381084856427103,
+      "learning_rate": 1.5802903910797584e-06,
+      "loss": 0.9704,
+      "step": 3042
+    },
+    {
+      "epoch": 2.4639676113360323,
+      "grad_norm": 1.7855052345593487,
+      "learning_rate": 1.575656125657583e-06,
+      "loss": 0.9614,
+      "step": 3043
+    },
+    {
+      "epoch": 2.4647773279352228,
+      "grad_norm": 1.6487843421840669,
+      "learning_rate": 1.5710280841129932e-06,
+      "loss": 1.0039,
+      "step": 3044
+    },
+    {
+      "epoch": 2.465587044534413,
+      "grad_norm": 1.6882419309228567,
+      "learning_rate": 1.5664062698651706e-06,
+      "loss": 0.9755,
+      "step": 3045
+    },
+    {
+      "epoch": 2.4663967611336033,
+      "grad_norm": 1.6447631511348773,
+      "learning_rate": 1.5617906863286936e-06,
+      "loss": 1.026,
+      "step": 3046
+    },
+    {
+      "epoch": 2.4672064777327933,
+      "grad_norm": 1.7529100807012996,
+      "learning_rate": 1.5571813369135457e-06,
+      "loss": 0.9185,
+      "step": 3047
+    },
+    {
+      "epoch": 2.468016194331984,
+      "grad_norm": 1.7744701731060566,
+      "learning_rate": 1.5525782250250953e-06,
+      "loss": 0.9455,
+      "step": 3048
+    },
+    {
+      "epoch": 2.468825910931174,
+      "grad_norm": 1.7939853067544942,
+      "learning_rate": 1.5479813540641064e-06,
+      "loss": 1.0429,
+      "step": 3049
+    },
+    {
+      "epoch": 2.4696356275303644,
+      "grad_norm": 1.8049689554151258,
+      "learning_rate": 1.5433907274267357e-06,
+      "loss": 0.9599,
+      "step": 3050
+    },
+    {
+      "epoch": 2.470445344129555,
+      "grad_norm": 1.7257816506826475,
+      "learning_rate": 1.538806348504519e-06,
+      "loss": 1.0007,
+      "step": 3051
+    },
+    {
+      "epoch": 2.471255060728745,
+      "grad_norm": 1.6882725120061068,
+      "learning_rate": 1.534228220684384e-06,
+      "loss": 0.9825,
+      "step": 3052
+    },
+    {
+      "epoch": 2.4720647773279354,
+      "grad_norm": 1.5881294637631733,
+      "learning_rate": 1.5296563473486325e-06,
+      "loss": 0.9745,
+      "step": 3053
+    },
+    {
+      "epoch": 2.4728744939271254,
+      "grad_norm": 1.730530663215343,
+      "learning_rate": 1.525090731874951e-06,
+      "loss": 0.9413,
+      "step": 3054
+    },
+    {
+      "epoch": 2.473684210526316,
+      "grad_norm": 1.669751163835487,
+      "learning_rate": 1.5205313776364028e-06,
+      "loss": 0.9574,
+      "step": 3055
+    },
+    {
+      "epoch": 2.474493927125506,
+      "grad_norm": 1.6752509320040634,
+      "learning_rate": 1.5159782880014207e-06,
+      "loss": 1.0109,
+      "step": 3056
+    },
+    {
+      "epoch": 2.4753036437246965,
+      "grad_norm": 1.6347253102331993,
+      "learning_rate": 1.511431466333817e-06,
+      "loss": 0.9337,
+      "step": 3057
+    },
+    {
+      "epoch": 2.4761133603238865,
+      "grad_norm": 1.6962317193949117,
+      "learning_rate": 1.506890915992766e-06,
+      "loss": 1.0413,
+      "step": 3058
+    },
+    {
+      "epoch": 2.476923076923077,
+      "grad_norm": 1.6430785203789686,
+      "learning_rate": 1.5023566403328105e-06,
+      "loss": 0.9546,
+      "step": 3059
+    },
+    {
+      "epoch": 2.477732793522267,
+      "grad_norm": 1.6664785121693346,
+      "learning_rate": 1.4978286427038602e-06,
+      "loss": 0.9703,
+      "step": 3060
+    },
+    {
+      "epoch": 2.4785425101214575,
+      "grad_norm": 1.614002773301002,
+      "learning_rate": 1.4933069264511834e-06,
+      "loss": 1.029,
+      "step": 3061
+    },
+    {
+      "epoch": 2.4793522267206476,
+      "grad_norm": 1.6894913124402384,
+      "learning_rate": 1.488791494915408e-06,
+      "loss": 0.8944,
+      "step": 3062
+    },
+    {
+      "epoch": 2.480161943319838,
+      "grad_norm": 1.5775586891733708,
+      "learning_rate": 1.4842823514325244e-06,
+      "loss": 0.9362,
+      "step": 3063
+    },
+    {
+      "epoch": 2.480971659919028,
+      "grad_norm": 1.645201425602457,
+      "learning_rate": 1.4797794993338676e-06,
+      "loss": 0.9915,
+      "step": 3064
+    },
+    {
+      "epoch": 2.4817813765182186,
+      "grad_norm": 1.6765332328829647,
+      "learning_rate": 1.4752829419461357e-06,
+      "loss": 1.0843,
+      "step": 3065
+    },
+    {
+      "epoch": 2.482591093117409,
+      "grad_norm": 1.718267617032678,
+      "learning_rate": 1.4707926825913676e-06,
+      "loss": 0.9818,
+      "step": 3066
+    },
+    {
+      "epoch": 2.483400809716599,
+      "grad_norm": 1.6812421248362865,
+      "learning_rate": 1.4663087245869512e-06,
+      "loss": 0.9961,
+      "step": 3067
+    },
+    {
+      "epoch": 2.4842105263157896,
+      "grad_norm": 1.6802970912002102,
+      "learning_rate": 1.4618310712456218e-06,
+      "loss": 0.9768,
+      "step": 3068
+    },
+    {
+      "epoch": 2.4850202429149797,
+      "grad_norm": 1.7391114674842005,
+      "learning_rate": 1.4573597258754546e-06,
+      "loss": 0.9987,
+      "step": 3069
+    },
+    {
+      "epoch": 2.48582995951417,
+      "grad_norm": 1.694339682778352,
+      "learning_rate": 1.4528946917798603e-06,
+      "loss": 0.9906,
+      "step": 3070
+    },
+    {
+      "epoch": 2.4866396761133602,
+      "grad_norm": 1.7363838335149249,
+      "learning_rate": 1.448435972257597e-06,
+      "loss": 0.9866,
+      "step": 3071
+    },
+    {
+      "epoch": 2.4874493927125507,
+      "grad_norm": 1.6870306102440977,
+      "learning_rate": 1.4439835706027526e-06,
+      "loss": 0.9262,
+      "step": 3072
+    },
+    {
+      "epoch": 2.4882591093117408,
+      "grad_norm": 1.6692336890689885,
+      "learning_rate": 1.4395374901047443e-06,
+      "loss": 0.9166,
+      "step": 3073
+    },
+    {
+      "epoch": 2.4890688259109313,
+      "grad_norm": 1.6886739622081277,
+      "learning_rate": 1.4350977340483218e-06,
+      "loss": 1.0644,
+      "step": 3074
+    },
+    {
+      "epoch": 2.4898785425101213,
+      "grad_norm": 1.6364093438608744,
+      "learning_rate": 1.4306643057135638e-06,
+      "loss": 1.0401,
+      "step": 3075
+    },
+    {
+      "epoch": 2.490688259109312,
+      "grad_norm": 1.6954035597211397,
+      "learning_rate": 1.4262372083758714e-06,
+      "loss": 0.981,
+      "step": 3076
+    },
+    {
+      "epoch": 2.491497975708502,
+      "grad_norm": 1.6508448957202095,
+      "learning_rate": 1.4218164453059669e-06,
+      "loss": 1.0243,
+      "step": 3077
+    },
+    {
+      "epoch": 2.4923076923076923,
+      "grad_norm": 1.7298928655955168,
+      "learning_rate": 1.4174020197699e-06,
+      "loss": 0.9714,
+      "step": 3078
+    },
+    {
+      "epoch": 2.4931174089068824,
+      "grad_norm": 1.668595707670543,
+      "learning_rate": 1.4129939350290312e-06,
+      "loss": 1.01,
+      "step": 3079
+    },
+    {
+      "epoch": 2.493927125506073,
+      "grad_norm": 1.7117470225490021,
+      "learning_rate": 1.4085921943400416e-06,
+      "loss": 0.961,
+      "step": 3080
+    },
+    {
+      "epoch": 2.4947368421052634,
+      "grad_norm": 1.6420783440308226,
+      "learning_rate": 1.404196800954921e-06,
+      "loss": 0.954,
+      "step": 3081
+    },
+    {
+      "epoch": 2.4955465587044534,
+      "grad_norm": 1.6312366446037767,
+      "learning_rate": 1.3998077581209712e-06,
+      "loss": 0.9967,
+      "step": 3082
+    },
+    {
+      "epoch": 2.496356275303644,
+      "grad_norm": 1.7256120345016341,
+      "learning_rate": 1.3954250690808036e-06,
+      "loss": 0.9527,
+      "step": 3083
+    },
+    {
+      "epoch": 2.497165991902834,
+      "grad_norm": 1.6410665457677764,
+      "learning_rate": 1.3910487370723347e-06,
+      "loss": 1.0166,
+      "step": 3084
+    },
+    {
+      "epoch": 2.4979757085020244,
+      "grad_norm": 1.683999810079675,
+      "learning_rate": 1.3866787653287804e-06,
+      "loss": 0.8987,
+      "step": 3085
+    },
+    {
+      "epoch": 2.4987854251012145,
+      "grad_norm": 1.7049093242978388,
+      "learning_rate": 1.3823151570786653e-06,
+      "loss": 0.9651,
+      "step": 3086
+    },
+    {
+      "epoch": 2.499595141700405,
+      "grad_norm": 1.7012673477415141,
+      "learning_rate": 1.3779579155458089e-06,
+      "loss": 0.9147,
+      "step": 3087
+    },
+    {
+      "epoch": 2.500404858299595,
+      "grad_norm": 1.7058323956163846,
+      "learning_rate": 1.3736070439493277e-06,
+      "loss": 0.9477,
+      "step": 3088
+    },
+    {
+      "epoch": 2.5012145748987855,
+      "grad_norm": 1.7742651035823065,
+      "learning_rate": 1.369262545503629e-06,
+      "loss": 0.9419,
+      "step": 3089
+    },
+    {
+      "epoch": 2.5020242914979756,
+      "grad_norm": 1.5760497645036358,
+      "learning_rate": 1.3649244234184157e-06,
+      "loss": 0.9739,
+      "step": 3090
+    },
+    {
+      "epoch": 2.502834008097166,
+      "grad_norm": 1.637791167494247,
+      "learning_rate": 1.3605926808986758e-06,
+      "loss": 0.9618,
+      "step": 3091
+    },
+    {
+      "epoch": 2.5036437246963565,
+      "grad_norm": 1.6641954657060896,
+      "learning_rate": 1.3562673211446863e-06,
+      "loss": 0.992,
+      "step": 3092
+    },
+    {
+      "epoch": 2.5044534412955466,
+      "grad_norm": 1.687575151546589,
+      "learning_rate": 1.3519483473520124e-06,
+      "loss": 0.946,
+      "step": 3093
+    },
+    {
+      "epoch": 2.5052631578947366,
+      "grad_norm": 1.6264510448768228,
+      "learning_rate": 1.3476357627114945e-06,
+      "loss": 1.0184,
+      "step": 3094
+    },
+    {
+      "epoch": 2.506072874493927,
+      "grad_norm": 1.630753363552384,
+      "learning_rate": 1.3433295704092586e-06,
+      "loss": 0.9742,
+      "step": 3095
+    },
+    {
+      "epoch": 2.5068825910931176,
+      "grad_norm": 1.7544663565446108,
+      "learning_rate": 1.3390297736267033e-06,
+      "loss": 0.9339,
+      "step": 3096
+    },
+    {
+      "epoch": 2.5076923076923077,
+      "grad_norm": 1.6250326610562258,
+      "learning_rate": 1.3347363755405064e-06,
+      "loss": 0.9355,
+      "step": 3097
+    },
+    {
+      "epoch": 2.5085020242914977,
+      "grad_norm": 1.7027349777509002,
+      "learning_rate": 1.3304493793226135e-06,
+      "loss": 0.9723,
+      "step": 3098
+    },
+    {
+      "epoch": 2.509311740890688,
+      "grad_norm": 1.8050244837715836,
+      "learning_rate": 1.3261687881402464e-06,
+      "loss": 0.99,
+      "step": 3099
+    },
+    {
+      "epoch": 2.5101214574898787,
+      "grad_norm": 1.7168877357885066,
+      "learning_rate": 1.3218946051558867e-06,
+      "loss": 1.0173,
+      "step": 3100
+    },
+    {
+      "epoch": 2.5109311740890687,
+      "grad_norm": 1.7435696786750583,
+      "learning_rate": 1.3176268335272934e-06,
+      "loss": 1.0018,
+      "step": 3101
+    },
+    {
+      "epoch": 2.5117408906882592,
+      "grad_norm": 1.5938621762485412,
+      "learning_rate": 1.3133654764074765e-06,
+      "loss": 1.111,
+      "step": 3102
+    },
+    {
+      "epoch": 2.5125506072874493,
+      "grad_norm": 1.6757330684573215,
+      "learning_rate": 1.3091105369447166e-06,
+      "loss": 0.9999,
+      "step": 3103
+    },
+    {
+      "epoch": 2.5133603238866398,
+      "grad_norm": 1.7130683949058882,
+      "learning_rate": 1.3048620182825478e-06,
+      "loss": 1.0611,
+      "step": 3104
+    },
+    {
+      "epoch": 2.51417004048583,
+      "grad_norm": 1.7070694868968392,
+      "learning_rate": 1.3006199235597628e-06,
+      "loss": 0.9069,
+      "step": 3105
+    },
+    {
+      "epoch": 2.5149797570850203,
+      "grad_norm": 1.6521586885628545,
+      "learning_rate": 1.2963842559104045e-06,
+      "loss": 1.0559,
+      "step": 3106
+    },
+    {
+      "epoch": 2.515789473684211,
+      "grad_norm": 1.6872800067468905,
+      "learning_rate": 1.29215501846377e-06,
+      "loss": 1.0301,
+      "step": 3107
+    },
+    {
+      "epoch": 2.516599190283401,
+      "grad_norm": 1.6209272379948023,
+      "learning_rate": 1.2879322143444095e-06,
+      "loss": 0.9728,
+      "step": 3108
+    },
+    {
+      "epoch": 2.517408906882591,
+      "grad_norm": 1.6880406144910838,
+      "learning_rate": 1.2837158466721155e-06,
+      "loss": 0.982,
+      "step": 3109
+    },
+    {
+      "epoch": 2.5182186234817814,
+      "grad_norm": 1.7156607604203853,
+      "learning_rate": 1.279505918561923e-06,
+      "loss": 1.0017,
+      "step": 3110
+    },
+    {
+      "epoch": 2.519028340080972,
+      "grad_norm": 1.6981687300306323,
+      "learning_rate": 1.2753024331241193e-06,
+      "loss": 0.9594,
+      "step": 3111
+    },
+    {
+      "epoch": 2.519838056680162,
+      "grad_norm": 1.7067120833158638,
+      "learning_rate": 1.2711053934642225e-06,
+      "loss": 0.9487,
+      "step": 3112
+    },
+    {
+      "epoch": 2.520647773279352,
+      "grad_norm": 1.689789662984842,
+      "learning_rate": 1.2669148026829915e-06,
+      "loss": 1.0139,
+      "step": 3113
+    },
+    {
+      "epoch": 2.5214574898785425,
+      "grad_norm": 1.6470500284421414,
+      "learning_rate": 1.2627306638764213e-06,
+      "loss": 1.0077,
+      "step": 3114
+    },
+    {
+      "epoch": 2.522267206477733,
+      "grad_norm": 1.5826496661345488,
+      "learning_rate": 1.2585529801357377e-06,
+      "loss": 1.0494,
+      "step": 3115
+    },
+    {
+      "epoch": 2.523076923076923,
+      "grad_norm": 1.742074564535538,
+      "learning_rate": 1.2543817545474036e-06,
+      "loss": 1.0169,
+      "step": 3116
+    },
+    {
+      "epoch": 2.5238866396761135,
+      "grad_norm": 1.7660057671815605,
+      "learning_rate": 1.250216990193105e-06,
+      "loss": 0.9641,
+      "step": 3117
+    },
+    {
+      "epoch": 2.5246963562753035,
+      "grad_norm": 1.652494097329047,
+      "learning_rate": 1.246058690149755e-06,
+      "loss": 0.9172,
+      "step": 3118
+    },
+    {
+      "epoch": 2.525506072874494,
+      "grad_norm": 1.6461968449457183,
+      "learning_rate": 1.2419068574894943e-06,
+      "loss": 1.0005,
+      "step": 3119
+    },
+    {
+      "epoch": 2.526315789473684,
+      "grad_norm": 1.6624721041701427,
+      "learning_rate": 1.2377614952796825e-06,
+      "loss": 0.9542,
+      "step": 3120
+    },
+    {
+      "epoch": 2.5271255060728746,
+      "grad_norm": 1.7097282118454058,
+      "learning_rate": 1.2336226065828993e-06,
+      "loss": 0.9636,
+      "step": 3121
+    },
+    {
+      "epoch": 2.527935222672065,
+      "grad_norm": 1.7315637749359838,
+      "learning_rate": 1.2294901944569394e-06,
+      "loss": 0.9522,
+      "step": 3122
+    },
+    {
+      "epoch": 2.528744939271255,
+      "grad_norm": 1.68876777571607,
+      "learning_rate": 1.22536426195482e-06,
+      "loss": 0.9978,
+      "step": 3123
+    },
+    {
+      "epoch": 2.529554655870445,
+      "grad_norm": 1.7501789467762574,
+      "learning_rate": 1.2212448121247643e-06,
+      "loss": 0.9859,
+      "step": 3124
+    },
+    {
+      "epoch": 2.5303643724696356,
+      "grad_norm": 1.6915559058052043,
+      "learning_rate": 1.217131848010209e-06,
+      "loss": 0.9187,
+      "step": 3125
+    },
+    {
+      "epoch": 2.531174089068826,
+      "grad_norm": 1.7489759930129678,
+      "learning_rate": 1.2130253726497954e-06,
+      "loss": 0.9737,
+      "step": 3126
+    },
+    {
+      "epoch": 2.531983805668016,
+      "grad_norm": 1.6670553247236088,
+      "learning_rate": 1.2089253890773789e-06,
+      "loss": 0.8828,
+      "step": 3127
+    },
+    {
+      "epoch": 2.532793522267206,
+      "grad_norm": 1.6651695693818536,
+      "learning_rate": 1.204831900322011e-06,
+      "loss": 0.9528,
+      "step": 3128
+    },
+    {
+      "epoch": 2.5336032388663967,
+      "grad_norm": 1.6250021225860003,
+      "learning_rate": 1.200744909407946e-06,
+      "loss": 0.9981,
+      "step": 3129
+    },
+    {
+      "epoch": 2.534412955465587,
+      "grad_norm": 1.7704071649167492,
+      "learning_rate": 1.196664419354644e-06,
+      "loss": 0.9586,
+      "step": 3130
+    },
+    {
+      "epoch": 2.5352226720647772,
+      "grad_norm": 1.732509606852866,
+      "learning_rate": 1.1925904331767545e-06,
+      "loss": 0.9765,
+      "step": 3131
+    },
+    {
+      "epoch": 2.5360323886639677,
+      "grad_norm": 1.7830253671280303,
+      "learning_rate": 1.1885229538841259e-06,
+      "loss": 0.9934,
+      "step": 3132
+    },
+    {
+      "epoch": 2.536842105263158,
+      "grad_norm": 1.7228909788826947,
+      "learning_rate": 1.1844619844817995e-06,
+      "loss": 0.9926,
+      "step": 3133
+    },
+    {
+      "epoch": 2.5376518218623483,
+      "grad_norm": 1.7301843770006853,
+      "learning_rate": 1.1804075279700023e-06,
+      "loss": 0.9108,
+      "step": 3134
+    },
+    {
+      "epoch": 2.5384615384615383,
+      "grad_norm": 1.7297103310761568,
+      "learning_rate": 1.176359587344158e-06,
+      "loss": 0.9898,
+      "step": 3135
+    },
+    {
+      "epoch": 2.539271255060729,
+      "grad_norm": 1.6692620409074408,
+      "learning_rate": 1.17231816559487e-06,
+      "loss": 0.9766,
+      "step": 3136
+    },
+    {
+      "epoch": 2.540080971659919,
+      "grad_norm": 1.7138975481366907,
+      "learning_rate": 1.168283265707927e-06,
+      "loss": 1.0342,
+      "step": 3137
+    },
+    {
+      "epoch": 2.5408906882591094,
+      "grad_norm": 1.8684377760837239,
+      "learning_rate": 1.1642548906643003e-06,
+      "loss": 0.917,
+      "step": 3138
+    },
+    {
+      "epoch": 2.5417004048582994,
+      "grad_norm": 1.6682784606850616,
+      "learning_rate": 1.160233043440141e-06,
+      "loss": 0.9585,
+      "step": 3139
+    },
+    {
+      "epoch": 2.54251012145749,
+      "grad_norm": 1.646760129502982,
+      "learning_rate": 1.1562177270067766e-06,
+      "loss": 0.9977,
+      "step": 3140
+    },
+    {
+      "epoch": 2.5433198380566804,
+      "grad_norm": 1.7761301320650948,
+      "learning_rate": 1.1522089443307083e-06,
+      "loss": 1.0222,
+      "step": 3141
+    },
+    {
+      "epoch": 2.5441295546558704,
+      "grad_norm": 1.713346287093232,
+      "learning_rate": 1.1482066983736095e-06,
+      "loss": 1.0214,
+      "step": 3142
+    },
+    {
+      "epoch": 2.5449392712550605,
+      "grad_norm": 1.739109874170232,
+      "learning_rate": 1.1442109920923317e-06,
+      "loss": 0.9595,
+      "step": 3143
+    },
+    {
+      "epoch": 2.545748987854251,
+      "grad_norm": 1.797935688363354,
+      "learning_rate": 1.1402218284388845e-06,
+      "loss": 0.9891,
+      "step": 3144
+    },
+    {
+      "epoch": 2.5465587044534415,
+      "grad_norm": 1.7311448213719867,
+      "learning_rate": 1.1362392103604536e-06,
+      "loss": 0.9481,
+      "step": 3145
+    },
+    {
+      "epoch": 2.5473684210526315,
+      "grad_norm": 1.6344396564565618,
+      "learning_rate": 1.132263140799381e-06,
+      "loss": 0.9162,
+      "step": 3146
+    },
+    {
+      "epoch": 2.548178137651822,
+      "grad_norm": 1.7616052113712068,
+      "learning_rate": 1.1282936226931762e-06,
+      "loss": 0.947,
+      "step": 3147
+    },
+    {
+      "epoch": 2.548987854251012,
+      "grad_norm": 1.671278873181062,
+      "learning_rate": 1.124330658974503e-06,
+      "loss": 0.9724,
+      "step": 3148
+    },
+    {
+      "epoch": 2.5497975708502025,
+      "grad_norm": 1.624337775639022,
+      "learning_rate": 1.120374252571188e-06,
+      "loss": 0.9633,
+      "step": 3149
+    },
+    {
+      "epoch": 2.5506072874493926,
+      "grad_norm": 1.7342053818211673,
+      "learning_rate": 1.1164244064062101e-06,
+      "loss": 0.9351,
+      "step": 3150
+    },
+    {
+      "epoch": 2.551417004048583,
+      "grad_norm": 1.7210916677441777,
+      "learning_rate": 1.112481123397704e-06,
+      "loss": 0.936,
+      "step": 3151
+    },
+    {
+      "epoch": 2.552226720647773,
+      "grad_norm": 1.7346383956750289,
+      "learning_rate": 1.1085444064589523e-06,
+      "loss": 0.9564,
+      "step": 3152
+    },
+    {
+      "epoch": 2.5530364372469636,
+      "grad_norm": 1.7122232488976015,
+      "learning_rate": 1.104614258498392e-06,
+      "loss": 1.0299,
+      "step": 3153
+    },
+    {
+      "epoch": 2.5538461538461537,
+      "grad_norm": 1.6894970983661606,
+      "learning_rate": 1.1006906824196006e-06,
+      "loss": 1.052,
+      "step": 3154
+    },
+    {
+      "epoch": 2.554655870445344,
+      "grad_norm": 1.7405016755106155,
+      "learning_rate": 1.0967736811213048e-06,
+      "loss": 0.9761,
+      "step": 3155
+    },
+    {
+      "epoch": 2.5554655870445346,
+      "grad_norm": 1.6635809223324118,
+      "learning_rate": 1.0928632574973718e-06,
+      "loss": 0.9735,
+      "step": 3156
+    },
+    {
+      "epoch": 2.5562753036437247,
+      "grad_norm": 1.5865674939483103,
+      "learning_rate": 1.0889594144368088e-06,
+      "loss": 0.9703,
+      "step": 3157
+    },
+    {
+      "epoch": 2.5570850202429147,
+      "grad_norm": 1.7886384841861689,
+      "learning_rate": 1.0850621548237604e-06,
+      "loss": 0.953,
+      "step": 3158
+    },
+    {
+      "epoch": 2.557894736842105,
+      "grad_norm": 1.65444440912227,
+      "learning_rate": 1.081171481537513e-06,
+      "loss": 1.0416,
+      "step": 3159
+    },
+    {
+      "epoch": 2.5587044534412957,
+      "grad_norm": 1.6010494418210732,
+      "learning_rate": 1.0772873974524833e-06,
+      "loss": 0.978,
+      "step": 3160
+    },
+    {
+      "epoch": 2.5595141700404858,
+      "grad_norm": 1.753931513386066,
+      "learning_rate": 1.0734099054382186e-06,
+      "loss": 1.0272,
+      "step": 3161
+    },
+    {
+      "epoch": 2.5603238866396762,
+      "grad_norm": 1.7002625406870047,
+      "learning_rate": 1.069539008359397e-06,
+      "loss": 0.9895,
+      "step": 3162
+    },
+    {
+      "epoch": 2.5611336032388663,
+      "grad_norm": 1.696885279537435,
+      "learning_rate": 1.0656747090758246e-06,
+      "loss": 1.042,
+      "step": 3163
+    },
+    {
+      "epoch": 2.561943319838057,
+      "grad_norm": 1.6725123732842846,
+      "learning_rate": 1.061817010442433e-06,
+      "loss": 0.9873,
+      "step": 3164
+    },
+    {
+      "epoch": 2.562753036437247,
+      "grad_norm": 1.744422716922843,
+      "learning_rate": 1.0579659153092759e-06,
+      "loss": 0.9915,
+      "step": 3165
+    },
+    {
+      "epoch": 2.5635627530364373,
+      "grad_norm": 1.6982287591882945,
+      "learning_rate": 1.0541214265215328e-06,
+      "loss": 0.993,
+      "step": 3166
+    },
+    {
+      "epoch": 2.5643724696356274,
+      "grad_norm": 1.6580163858315626,
+      "learning_rate": 1.0502835469194961e-06,
+      "loss": 1.0379,
+      "step": 3167
+    },
+    {
+      "epoch": 2.565182186234818,
+      "grad_norm": 1.6503684640252827,
+      "learning_rate": 1.0464522793385822e-06,
+      "loss": 0.9683,
+      "step": 3168
+    },
+    {
+      "epoch": 2.565991902834008,
+      "grad_norm": 1.6600515587726554,
+      "learning_rate": 1.0426276266093172e-06,
+      "loss": 1.0482,
+      "step": 3169
+    },
+    {
+      "epoch": 2.5668016194331984,
+      "grad_norm": 1.7091294278437665,
+      "learning_rate": 1.0388095915573427e-06,
+      "loss": 1.0417,
+      "step": 3170
+    },
+    {
+      "epoch": 2.567611336032389,
+      "grad_norm": 1.6555794090881855,
+      "learning_rate": 1.034998177003409e-06,
+      "loss": 0.9302,
+      "step": 3171
+    },
+    {
+      "epoch": 2.568421052631579,
+      "grad_norm": 1.7961476476441376,
+      "learning_rate": 1.0311933857633772e-06,
+      "loss": 0.967,
+      "step": 3172
+    },
+    {
+      "epoch": 2.569230769230769,
+      "grad_norm": 1.6778244264366644,
+      "learning_rate": 1.027395220648213e-06,
+      "loss": 1.0396,
+      "step": 3173
+    },
+    {
+      "epoch": 2.5700404858299595,
+      "grad_norm": 1.6859023669000333,
+      "learning_rate": 1.0236036844639897e-06,
+      "loss": 0.9793,
+      "step": 3174
+    },
+    {
+      "epoch": 2.57085020242915,
+      "grad_norm": 1.7334612254526203,
+      "learning_rate": 1.0198187800118842e-06,
+      "loss": 0.9165,
+      "step": 3175
+    },
+    {
+      "epoch": 2.57165991902834,
+      "grad_norm": 1.726216684645843,
+      "learning_rate": 1.0160405100881699e-06,
+      "loss": 1.0163,
+      "step": 3176
+    },
+    {
+      "epoch": 2.5724696356275305,
+      "grad_norm": 1.6800195699709957,
+      "learning_rate": 1.0122688774842194e-06,
+      "loss": 0.9615,
+      "step": 3177
+    },
+    {
+      "epoch": 2.5732793522267206,
+      "grad_norm": 1.7496662315340423,
+      "learning_rate": 1.0085038849865025e-06,
+      "loss": 0.9433,
+      "step": 3178
+    },
+    {
+      "epoch": 2.574089068825911,
+      "grad_norm": 1.669757711256767,
+      "learning_rate": 1.0047455353765845e-06,
+      "loss": 0.998,
+      "step": 3179
+    },
+    {
+      "epoch": 2.574898785425101,
+      "grad_norm": 1.7649644449020268,
+      "learning_rate": 1.0009938314311186e-06,
+      "loss": 0.8943,
+      "step": 3180
+    },
+    {
+      "epoch": 2.5757085020242916,
+      "grad_norm": 1.665726033058726,
+      "learning_rate": 9.972487759218551e-07,
+      "loss": 0.9636,
+      "step": 3181
+    },
+    {
+      "epoch": 2.5765182186234816,
+      "grad_norm": 1.6714613239053895,
+      "learning_rate": 9.935103716156258e-07,
+      "loss": 0.9908,
+      "step": 3182
+    },
+    {
+      "epoch": 2.577327935222672,
+      "grad_norm": 1.780013775907417,
+      "learning_rate": 9.897786212743543e-07,
+      "loss": 0.9168,
+      "step": 3183
+    },
+    {
+      "epoch": 2.578137651821862,
+      "grad_norm": 1.6637866101217857,
+      "learning_rate": 9.860535276550443e-07,
+      "loss": 0.9549,
+      "step": 3184
+    },
+    {
+      "epoch": 2.5789473684210527,
+      "grad_norm": 1.6430785762195612,
+      "learning_rate": 9.82335093509782e-07,
+      "loss": 0.9702,
+      "step": 3185
+    },
+    {
+      "epoch": 2.579757085020243,
+      "grad_norm": 1.7132270223349593,
+      "learning_rate": 9.786233215857354e-07,
+      "loss": 0.9692,
+      "step": 3186
+    },
+    {
+      "epoch": 2.580566801619433,
+      "grad_norm": 1.7060451332011317,
+      "learning_rate": 9.74918214625149e-07,
+      "loss": 0.9941,
+      "step": 3187
+    },
+    {
+      "epoch": 2.5813765182186232,
+      "grad_norm": 1.6254697719748301,
+      "learning_rate": 9.712197753653418e-07,
+      "loss": 1.0413,
+      "step": 3188
+    },
+    {
+      "epoch": 2.5821862348178137,
+      "grad_norm": 1.6890779172726746,
+      "learning_rate": 9.675280065387117e-07,
+      "loss": 0.9546,
+      "step": 3189
+    },
+    {
+      "epoch": 2.582995951417004,
+      "grad_norm": 1.6665698649055485,
+      "learning_rate": 9.638429108727232e-07,
+      "loss": 0.9145,
+      "step": 3190
+    },
+    {
+      "epoch": 2.5838056680161943,
+      "grad_norm": 1.7455050077057428,
+      "learning_rate": 9.601644910899144e-07,
+      "loss": 0.9475,
+      "step": 3191
+    },
+    {
+      "epoch": 2.5846153846153848,
+      "grad_norm": 1.707769647747537,
+      "learning_rate": 9.56492749907889e-07,
+      "loss": 0.9925,
+      "step": 3192
+    },
+    {
+      "epoch": 2.585425101214575,
+      "grad_norm": 1.7626431911020435,
+      "learning_rate": 9.528276900393185e-07,
+      "loss": 0.9197,
+      "step": 3193
+    },
+    {
+      "epoch": 2.5862348178137653,
+      "grad_norm": 1.6723942662699407,
+      "learning_rate": 9.491693141919345e-07,
+      "loss": 0.9899,
+      "step": 3194
+    },
+    {
+      "epoch": 2.5870445344129553,
+      "grad_norm": 1.66701720323838,
+      "learning_rate": 9.455176250685338e-07,
+      "loss": 1.0288,
+      "step": 3195
+    },
+    {
+      "epoch": 2.587854251012146,
+      "grad_norm": 1.6539257270825336,
+      "learning_rate": 9.418726253669741e-07,
+      "loss": 1.0232,
+      "step": 3196
+    },
+    {
+      "epoch": 2.588663967611336,
+      "grad_norm": 1.5465095186573317,
+      "learning_rate": 9.38234317780169e-07,
+      "loss": 0.9658,
+      "step": 3197
+    },
+    {
+      "epoch": 2.5894736842105264,
+      "grad_norm": 1.7029852892332802,
+      "learning_rate": 9.346027049960849e-07,
+      "loss": 0.9513,
+      "step": 3198
+    },
+    {
+      "epoch": 2.5902834008097164,
+      "grad_norm": 1.831209877056402,
+      "learning_rate": 9.309777896977501e-07,
+      "loss": 0.9345,
+      "step": 3199
+    },
+    {
+      "epoch": 2.591093117408907,
+      "grad_norm": 1.6872660132013921,
+      "learning_rate": 9.27359574563238e-07,
+      "loss": 0.9709,
+      "step": 3200
+    },
+    {
+      "epoch": 2.5919028340080974,
+      "grad_norm": 1.6893163114944878,
+      "learning_rate": 9.237480622656736e-07,
+      "loss": 0.9725,
+      "step": 3201
+    },
+    {
+      "epoch": 2.5927125506072874,
+      "grad_norm": 1.6058163315797132,
+      "learning_rate": 9.201432554732304e-07,
+      "loss": 1.0278,
+      "step": 3202
+    },
+    {
+      "epoch": 2.5935222672064775,
+      "grad_norm": 1.692466394304998,
+      "learning_rate": 9.165451568491257e-07,
+      "loss": 1.0333,
+      "step": 3203
+    },
+    {
+      "epoch": 2.594331983805668,
+      "grad_norm": 1.640592263798185,
+      "learning_rate": 9.129537690516277e-07,
+      "loss": 0.9587,
+      "step": 3204
+    },
+    {
+      "epoch": 2.5951417004048585,
+      "grad_norm": 1.6407337256128884,
+      "learning_rate": 9.093690947340406e-07,
+      "loss": 0.937,
+      "step": 3205
+    },
+    {
+      "epoch": 2.5959514170040485,
+      "grad_norm": 1.738587051999838,
+      "learning_rate": 9.057911365447058e-07,
+      "loss": 0.9966,
+      "step": 3206
+    },
+    {
+      "epoch": 2.596761133603239,
+      "grad_norm": 1.6250090533624972,
+      "learning_rate": 9.022198971270124e-07,
+      "loss": 0.9348,
+      "step": 3207
+    },
+    {
+      "epoch": 2.597570850202429,
+      "grad_norm": 1.643745528654048,
+      "learning_rate": 8.986553791193775e-07,
+      "loss": 1.0099,
+      "step": 3208
+    },
+    {
+      "epoch": 2.5983805668016196,
+      "grad_norm": 1.6883666457612738,
+      "learning_rate": 8.950975851552568e-07,
+      "loss": 0.9812,
+      "step": 3209
+    },
+    {
+      "epoch": 2.5991902834008096,
+      "grad_norm": 1.7178288521933882,
+      "learning_rate": 8.915465178631344e-07,
+      "loss": 0.9968,
+      "step": 3210
+    },
+    {
+      "epoch": 2.6,
+      "grad_norm": 1.6723879917548437,
+      "learning_rate": 8.880021798665295e-07,
+      "loss": 0.9755,
+      "step": 3211
+    },
+    {
+      "epoch": 2.60080971659919,
+      "grad_norm": 1.705852934667388,
+      "learning_rate": 8.844645737839874e-07,
+      "loss": 0.9901,
+      "step": 3212
+    },
+    {
+      "epoch": 2.6016194331983806,
+      "grad_norm": 1.7322533252061503,
+      "learning_rate": 8.809337022290787e-07,
+      "loss": 0.938,
+      "step": 3213
+    },
+    {
+      "epoch": 2.6024291497975707,
+      "grad_norm": 1.6855599969482573,
+      "learning_rate": 8.774095678103978e-07,
+      "loss": 0.9581,
+      "step": 3214
+    },
+    {
+      "epoch": 2.603238866396761,
+      "grad_norm": 1.6274201556331367,
+      "learning_rate": 8.738921731315686e-07,
+      "loss": 1.0798,
+      "step": 3215
+    },
+    {
+      "epoch": 2.6040485829959517,
+      "grad_norm": 1.61403911744669,
+      "learning_rate": 8.70381520791227e-07,
+      "loss": 1.0021,
+      "step": 3216
+    },
+    {
+      "epoch": 2.6048582995951417,
+      "grad_norm": 1.678255519779955,
+      "learning_rate": 8.668776133830315e-07,
+      "loss": 0.9717,
+      "step": 3217
+    },
+    {
+      "epoch": 2.6056680161943317,
+      "grad_norm": 1.627162580752701,
+      "learning_rate": 8.633804534956591e-07,
+      "loss": 0.8836,
+      "step": 3218
+    },
+    {
+      "epoch": 2.6064777327935222,
+      "grad_norm": 1.6220265843378787,
+      "learning_rate": 8.598900437127999e-07,
+      "loss": 0.9926,
+      "step": 3219
+    },
+    {
+      "epoch": 2.6072874493927127,
+      "grad_norm": 1.637954735309685,
+      "learning_rate": 8.564063866131567e-07,
+      "loss": 1.0443,
+      "step": 3220
+    },
+    {
+      "epoch": 2.6080971659919028,
+      "grad_norm": 1.6447001385429936,
+      "learning_rate": 8.529294847704428e-07,
+      "loss": 0.8968,
+      "step": 3221
+    },
+    {
+      "epoch": 2.6089068825910933,
+      "grad_norm": 1.732595510365215,
+      "learning_rate": 8.494593407533814e-07,
+      "loss": 0.9386,
+      "step": 3222
+    },
+    {
+      "epoch": 2.6097165991902833,
+      "grad_norm": 1.6660686105327704,
+      "learning_rate": 8.459959571257071e-07,
+      "loss": 1.0091,
+      "step": 3223
+    },
+    {
+      "epoch": 2.610526315789474,
+      "grad_norm": 1.7013721994565274,
+      "learning_rate": 8.425393364461542e-07,
+      "loss": 0.9747,
+      "step": 3224
+    },
+    {
+      "epoch": 2.611336032388664,
+      "grad_norm": 1.7819007792713526,
+      "learning_rate": 8.390894812684602e-07,
+      "loss": 0.9579,
+      "step": 3225
+    },
+    {
+      "epoch": 2.6121457489878543,
+      "grad_norm": 1.5960953357473817,
+      "learning_rate": 8.356463941413717e-07,
+      "loss": 1.0558,
+      "step": 3226
+    },
+    {
+      "epoch": 2.6129554655870444,
+      "grad_norm": 1.7717830181251943,
+      "learning_rate": 8.322100776086272e-07,
+      "loss": 1.0088,
+      "step": 3227
+    },
+    {
+      "epoch": 2.613765182186235,
+      "grad_norm": 1.7439967991496321,
+      "learning_rate": 8.287805342089672e-07,
+      "loss": 0.9053,
+      "step": 3228
+    },
+    {
+      "epoch": 2.614574898785425,
+      "grad_norm": 1.670184838895179,
+      "learning_rate": 8.253577664761259e-07,
+      "loss": 1.0389,
+      "step": 3229
+    },
+    {
+      "epoch": 2.6153846153846154,
+      "grad_norm": 1.7399555221858654,
+      "learning_rate": 8.219417769388316e-07,
+      "loss": 0.9939,
+      "step": 3230
+    },
+    {
+      "epoch": 2.616194331983806,
+      "grad_norm": 1.6749694700524644,
+      "learning_rate": 8.1853256812081e-07,
+      "loss": 1.0025,
+      "step": 3231
+    },
+    {
+      "epoch": 2.617004048582996,
+      "grad_norm": 1.7379585443686947,
+      "learning_rate": 8.151301425407699e-07,
+      "loss": 0.9861,
+      "step": 3232
+    },
+    {
+      "epoch": 2.617813765182186,
+      "grad_norm": 1.6065227038635674,
+      "learning_rate": 8.117345027124146e-07,
+      "loss": 1.0168,
+      "step": 3233
+    },
+    {
+      "epoch": 2.6186234817813765,
+      "grad_norm": 1.7012911860488888,
+      "learning_rate": 8.083456511444309e-07,
+      "loss": 1.0043,
+      "step": 3234
+    },
+    {
+      "epoch": 2.619433198380567,
+      "grad_norm": 1.716691520394183,
+      "learning_rate": 8.049635903404907e-07,
+      "loss": 0.9689,
+      "step": 3235
+    },
+    {
+      "epoch": 2.620242914979757,
+      "grad_norm": 1.7278598183805447,
+      "learning_rate": 8.015883227992505e-07,
+      "loss": 1.027,
+      "step": 3236
+    },
+    {
+      "epoch": 2.6210526315789475,
+      "grad_norm": 1.7367934858252325,
+      "learning_rate": 7.982198510143457e-07,
+      "loss": 0.9283,
+      "step": 3237
+    },
+    {
+      "epoch": 2.6218623481781376,
+      "grad_norm": 1.6429152171351453,
+      "learning_rate": 7.948581774743902e-07,
+      "loss": 0.9303,
+      "step": 3238
+    },
+    {
+      "epoch": 2.622672064777328,
+      "grad_norm": 1.7790796815176042,
+      "learning_rate": 7.915033046629817e-07,
+      "loss": 0.9561,
+      "step": 3239
+    },
+    {
+      "epoch": 2.623481781376518,
+      "grad_norm": 1.686258952709826,
+      "learning_rate": 7.881552350586863e-07,
+      "loss": 0.8947,
+      "step": 3240
+    },
+    {
+      "epoch": 2.6242914979757086,
+      "grad_norm": 1.758970152066431,
+      "learning_rate": 7.848139711350489e-07,
+      "loss": 1.048,
+      "step": 3241
+    },
+    {
+      "epoch": 2.6251012145748986,
+      "grad_norm": 1.6233206462939287,
+      "learning_rate": 7.814795153605825e-07,
+      "loss": 1.0282,
+      "step": 3242
+    },
+    {
+      "epoch": 2.625910931174089,
+      "grad_norm": 1.5853805998689159,
+      "learning_rate": 7.781518701987734e-07,
+      "loss": 1.0069,
+      "step": 3243
+    },
+    {
+      "epoch": 2.626720647773279,
+      "grad_norm": 1.849575708030945,
+      "learning_rate": 7.748310381080749e-07,
+      "loss": 0.904,
+      "step": 3244
+    },
+    {
+      "epoch": 2.6275303643724697,
+      "grad_norm": 1.655980362780037,
+      "learning_rate": 7.715170215419043e-07,
+      "loss": 0.988,
+      "step": 3245
+    },
+    {
+      "epoch": 2.62834008097166,
+      "grad_norm": 1.6899812614311005,
+      "learning_rate": 7.682098229486478e-07,
+      "loss": 0.9808,
+      "step": 3246
+    },
+    {
+      "epoch": 2.62914979757085,
+      "grad_norm": 1.6541647003198807,
+      "learning_rate": 7.649094447716532e-07,
+      "loss": 0.961,
+      "step": 3247
+    },
+    {
+      "epoch": 2.6299595141700403,
+      "grad_norm": 1.6996182835373292,
+      "learning_rate": 7.616158894492298e-07,
+      "loss": 0.9375,
+      "step": 3248
+    },
+    {
+      "epoch": 2.6307692307692307,
+      "grad_norm": 1.644339997790146,
+      "learning_rate": 7.583291594146458e-07,
+      "loss": 1.0002,
+      "step": 3249
+    },
+    {
+      "epoch": 2.6315789473684212,
+      "grad_norm": 1.7094644080217443,
+      "learning_rate": 7.550492570961243e-07,
+      "loss": 0.9905,
+      "step": 3250
+    },
+    {
+      "epoch": 2.6323886639676113,
+      "grad_norm": 1.724822229310986,
+      "learning_rate": 7.517761849168481e-07,
+      "loss": 0.9991,
+      "step": 3251
+    },
+    {
+      "epoch": 2.6331983805668018,
+      "grad_norm": 1.6081123719870767,
+      "learning_rate": 7.485099452949507e-07,
+      "loss": 0.9177,
+      "step": 3252
+    },
+    {
+      "epoch": 2.634008097165992,
+      "grad_norm": 1.7226834341746038,
+      "learning_rate": 7.452505406435184e-07,
+      "loss": 0.9244,
+      "step": 3253
+    },
+    {
+      "epoch": 2.6348178137651823,
+      "grad_norm": 1.6628814186493102,
+      "learning_rate": 7.419979733705929e-07,
+      "loss": 1.0378,
+      "step": 3254
+    },
+    {
+      "epoch": 2.6356275303643724,
+      "grad_norm": 1.738063003473798,
+      "learning_rate": 7.387522458791552e-07,
+      "loss": 0.9352,
+      "step": 3255
+    },
+    {
+      "epoch": 2.636437246963563,
+      "grad_norm": 1.680315939042461,
+      "learning_rate": 7.355133605671417e-07,
+      "loss": 1.0976,
+      "step": 3256
+    },
+    {
+      "epoch": 2.637246963562753,
+      "grad_norm": 1.665843160940704,
+      "learning_rate": 7.322813198274303e-07,
+      "loss": 0.9783,
+      "step": 3257
+    },
+    {
+      "epoch": 2.6380566801619434,
+      "grad_norm": 1.702934773134636,
+      "learning_rate": 7.290561260478401e-07,
+      "loss": 1.012,
+      "step": 3258
+    },
+    {
+      "epoch": 2.6388663967611334,
+      "grad_norm": 1.6427499028473984,
+      "learning_rate": 7.258377816111339e-07,
+      "loss": 1.0654,
+      "step": 3259
+    },
+    {
+      "epoch": 2.639676113360324,
+      "grad_norm": 1.681106134788699,
+      "learning_rate": 7.226262888950153e-07,
+      "loss": 0.9816,
+      "step": 3260
+    },
+    {
+      "epoch": 2.6404858299595144,
+      "grad_norm": 1.6644509739713758,
+      "learning_rate": 7.194216502721219e-07,
+      "loss": 1.0321,
+      "step": 3261
+    },
+    {
+      "epoch": 2.6412955465587045,
+      "grad_norm": 1.6556000089922558,
+      "learning_rate": 7.16223868110032e-07,
+      "loss": 0.9842,
+      "step": 3262
+    },
+    {
+      "epoch": 2.6421052631578945,
+      "grad_norm": 1.6874491415426638,
+      "learning_rate": 7.130329447712581e-07,
+      "loss": 0.9057,
+      "step": 3263
+    },
+    {
+      "epoch": 2.642914979757085,
+      "grad_norm": 1.7091112739486392,
+      "learning_rate": 7.098488826132422e-07,
+      "loss": 1.008,
+      "step": 3264
+    },
+    {
+      "epoch": 2.6437246963562755,
+      "grad_norm": 1.6178389429525835,
+      "learning_rate": 7.066716839883592e-07,
+      "loss": 0.9924,
+      "step": 3265
+    },
+    {
+      "epoch": 2.6445344129554655,
+      "grad_norm": 1.6485423266523553,
+      "learning_rate": 7.035013512439126e-07,
+      "loss": 1.054,
+      "step": 3266
+    },
+    {
+      "epoch": 2.645344129554656,
+      "grad_norm": 1.661813731422273,
+      "learning_rate": 7.003378867221344e-07,
+      "loss": 0.9126,
+      "step": 3267
+    },
+    {
+      "epoch": 2.646153846153846,
+      "grad_norm": 1.6864743875444674,
+      "learning_rate": 6.971812927601806e-07,
+      "loss": 0.9689,
+      "step": 3268
+    },
+    {
+      "epoch": 2.6469635627530366,
+      "grad_norm": 1.7764344486102688,
+      "learning_rate": 6.940315716901347e-07,
+      "loss": 0.9694,
+      "step": 3269
+    },
+    {
+      "epoch": 2.6477732793522266,
+      "grad_norm": 1.7426314403267598,
+      "learning_rate": 6.908887258389974e-07,
+      "loss": 1.0287,
+      "step": 3270
+    },
+    {
+      "epoch": 2.648582995951417,
+      "grad_norm": 1.6967622835020153,
+      "learning_rate": 6.877527575286958e-07,
+      "loss": 0.8432,
+      "step": 3271
+    },
+    {
+      "epoch": 2.649392712550607,
+      "grad_norm": 1.659074402231809,
+      "learning_rate": 6.846236690760721e-07,
+      "loss": 0.931,
+      "step": 3272
+    },
+    {
+      "epoch": 2.6502024291497976,
+      "grad_norm": 1.6674943656562866,
+      "learning_rate": 6.815014627928862e-07,
+      "loss": 0.9656,
+      "step": 3273
+    },
+    {
+      "epoch": 2.6510121457489877,
+      "grad_norm": 1.710449217756733,
+      "learning_rate": 6.783861409858128e-07,
+      "loss": 0.9369,
+      "step": 3274
+    },
+    {
+      "epoch": 2.651821862348178,
+      "grad_norm": 1.7486026332376843,
+      "learning_rate": 6.752777059564431e-07,
+      "loss": 0.9058,
+      "step": 3275
+    },
+    {
+      "epoch": 2.6526315789473687,
+      "grad_norm": 1.7290326137066063,
+      "learning_rate": 6.721761600012766e-07,
+      "loss": 0.9418,
+      "step": 3276
+    },
+    {
+      "epoch": 2.6534412955465587,
+      "grad_norm": 1.6974512106097555,
+      "learning_rate": 6.690815054117283e-07,
+      "loss": 0.9965,
+      "step": 3277
+    },
+    {
+      "epoch": 2.6542510121457488,
+      "grad_norm": 1.570906783647935,
+      "learning_rate": 6.659937444741149e-07,
+      "loss": 0.9696,
+      "step": 3278
+    },
+    {
+      "epoch": 2.6550607287449393,
+      "grad_norm": 1.680749541696145,
+      "learning_rate": 6.629128794696694e-07,
+      "loss": 0.8977,
+      "step": 3279
+    },
+    {
+      "epoch": 2.6558704453441297,
+      "grad_norm": 1.6928204758699041,
+      "learning_rate": 6.598389126745209e-07,
+      "loss": 1.0138,
+      "step": 3280
+    },
+    {
+      "epoch": 2.65668016194332,
+      "grad_norm": 1.5916031288683266,
+      "learning_rate": 6.567718463597061e-07,
+      "loss": 1.0033,
+      "step": 3281
+    },
+    {
+      "epoch": 2.6574898785425103,
+      "grad_norm": 1.7010292729133476,
+      "learning_rate": 6.537116827911649e-07,
+      "loss": 0.9784,
+      "step": 3282
+    },
+    {
+      "epoch": 2.6582995951417003,
+      "grad_norm": 1.7137596107033524,
+      "learning_rate": 6.506584242297332e-07,
+      "loss": 0.9155,
+      "step": 3283
+    },
+    {
+      "epoch": 2.659109311740891,
+      "grad_norm": 1.744228981789455,
+      "learning_rate": 6.476120729311531e-07,
+      "loss": 0.9438,
+      "step": 3284
+    },
+    {
+      "epoch": 2.659919028340081,
+      "grad_norm": 1.6938026901844712,
+      "learning_rate": 6.445726311460553e-07,
+      "loss": 1.0437,
+      "step": 3285
+    },
+    {
+      "epoch": 2.6607287449392714,
+      "grad_norm": 1.7064562907469736,
+      "learning_rate": 6.415401011199707e-07,
+      "loss": 0.9533,
+      "step": 3286
+    },
+    {
+      "epoch": 2.6615384615384614,
+      "grad_norm": 1.653652412179681,
+      "learning_rate": 6.385144850933222e-07,
+      "loss": 0.9629,
+      "step": 3287
+    },
+    {
+      "epoch": 2.662348178137652,
+      "grad_norm": 1.7289904638695122,
+      "learning_rate": 6.354957853014254e-07,
+      "loss": 1.017,
+      "step": 3288
+    },
+    {
+      "epoch": 2.663157894736842,
+      "grad_norm": 1.7152833008239363,
+      "learning_rate": 6.324840039744862e-07,
+      "loss": 0.9085,
+      "step": 3289
+    },
+    {
+      "epoch": 2.6639676113360324,
+      "grad_norm": 1.7289503260777432,
+      "learning_rate": 6.29479143337598e-07,
+      "loss": 0.8867,
+      "step": 3290
+    },
+    {
+      "epoch": 2.664777327935223,
+      "grad_norm": 1.6648250831416591,
+      "learning_rate": 6.264812056107406e-07,
+      "loss": 0.9589,
+      "step": 3291
+    },
+    {
+      "epoch": 2.665587044534413,
+      "grad_norm": 1.7548308959687973,
+      "learning_rate": 6.234901930087822e-07,
+      "loss": 0.9672,
+      "step": 3292
+    },
+    {
+      "epoch": 2.666396761133603,
+      "grad_norm": 1.6680110599487112,
+      "learning_rate": 6.205061077414743e-07,
+      "loss": 0.9872,
+      "step": 3293
+    },
+    {
+      "epoch": 2.6672064777327935,
+      "grad_norm": 1.7292774474924442,
+      "learning_rate": 6.175289520134464e-07,
+      "loss": 0.9107,
+      "step": 3294
+    },
+    {
+      "epoch": 2.668016194331984,
+      "grad_norm": 1.7431371411748449,
+      "learning_rate": 6.145587280242138e-07,
+      "loss": 1.009,
+      "step": 3295
+    },
+    {
+      "epoch": 2.668825910931174,
+      "grad_norm": 1.787321743542622,
+      "learning_rate": 6.115954379681666e-07,
+      "loss": 0.9275,
+      "step": 3296
+    },
+    {
+      "epoch": 2.669635627530364,
+      "grad_norm": 1.6827076549424453,
+      "learning_rate": 6.086390840345758e-07,
+      "loss": 0.9042,
+      "step": 3297
+    },
+    {
+      "epoch": 2.6704453441295546,
+      "grad_norm": 1.6938249525960054,
+      "learning_rate": 6.05689668407582e-07,
+      "loss": 1.0097,
+      "step": 3298
+    },
+    {
+      "epoch": 2.671255060728745,
+      "grad_norm": 1.745931288902666,
+      "learning_rate": 6.027471932662087e-07,
+      "loss": 1.028,
+      "step": 3299
+    },
+    {
+      "epoch": 2.672064777327935,
+      "grad_norm": 1.6780881861216097,
+      "learning_rate": 5.99811660784344e-07,
+      "loss": 1.0207,
+      "step": 3300
+    },
+    {
+      "epoch": 2.6728744939271256,
+      "grad_norm": 1.666413076144872,
+      "learning_rate": 5.968830731307507e-07,
+      "loss": 0.9445,
+      "step": 3301
+    },
+    {
+      "epoch": 2.6736842105263157,
+      "grad_norm": 1.6645120361386307,
+      "learning_rate": 5.93961432469058e-07,
+      "loss": 0.9988,
+      "step": 3302
+    },
+    {
+      "epoch": 2.674493927125506,
+      "grad_norm": 1.6392363775063152,
+      "learning_rate": 5.910467409577669e-07,
+      "loss": 0.8894,
+      "step": 3303
+    },
+    {
+      "epoch": 2.675303643724696,
+      "grad_norm": 1.6878734747415562,
+      "learning_rate": 5.881390007502397e-07,
+      "loss": 0.9665,
+      "step": 3304
+    },
+    {
+      "epoch": 2.6761133603238867,
+      "grad_norm": 1.7768958238131116,
+      "learning_rate": 5.852382139947077e-07,
+      "loss": 0.9143,
+      "step": 3305
+    },
+    {
+      "epoch": 2.676923076923077,
+      "grad_norm": 1.5994762220145167,
+      "learning_rate": 5.82344382834259e-07,
+      "loss": 1.0549,
+      "step": 3306
+    },
+    {
+      "epoch": 2.6777327935222672,
+      "grad_norm": 1.7374717323879252,
+      "learning_rate": 5.7945750940685e-07,
+      "loss": 1.0182,
+      "step": 3307
+    },
+    {
+      "epoch": 2.6785425101214573,
+      "grad_norm": 1.650625300217036,
+      "learning_rate": 5.765775958452935e-07,
+      "loss": 1.0116,
+      "step": 3308
+    },
+    {
+      "epoch": 2.6793522267206478,
+      "grad_norm": 1.7121361852833241,
+      "learning_rate": 5.737046442772576e-07,
+      "loss": 1.0502,
+      "step": 3309
+    },
+    {
+      "epoch": 2.6801619433198383,
+      "grad_norm": 1.6502806678346016,
+      "learning_rate": 5.708386568252688e-07,
+      "loss": 0.9503,
+      "step": 3310
+    },
+    {
+      "epoch": 2.6809716599190283,
+      "grad_norm": 1.686124777598768,
+      "learning_rate": 5.679796356067135e-07,
+      "loss": 0.9896,
+      "step": 3311
+    },
+    {
+      "epoch": 2.6817813765182184,
+      "grad_norm": 1.6024788294961974,
+      "learning_rate": 5.651275827338242e-07,
+      "loss": 1.0105,
+      "step": 3312
+    },
+    {
+      "epoch": 2.682591093117409,
+      "grad_norm": 1.7206476831624593,
+      "learning_rate": 5.622825003136878e-07,
+      "loss": 0.977,
+      "step": 3313
+    },
+    {
+      "epoch": 2.6834008097165993,
+      "grad_norm": 1.6455934468913418,
+      "learning_rate": 5.594443904482439e-07,
+      "loss": 0.9792,
+      "step": 3314
+    },
+    {
+      "epoch": 2.6842105263157894,
+      "grad_norm": 1.7190031808580448,
+      "learning_rate": 5.566132552342784e-07,
+      "loss": 1.0892,
+      "step": 3315
+    },
+    {
+      "epoch": 2.68502024291498,
+      "grad_norm": 1.6275442282090031,
+      "learning_rate": 5.53789096763423e-07,
+      "loss": 0.9998,
+      "step": 3316
+    },
+    {
+      "epoch": 2.68582995951417,
+      "grad_norm": 1.6723953815880728,
+      "learning_rate": 5.509719171221583e-07,
+      "loss": 0.9851,
+      "step": 3317
+    },
+    {
+      "epoch": 2.6866396761133604,
+      "grad_norm": 1.5877924488779105,
+      "learning_rate": 5.481617183918053e-07,
+      "loss": 0.9935,
+      "step": 3318
+    },
+    {
+      "epoch": 2.6874493927125505,
+      "grad_norm": 1.725319551891715,
+      "learning_rate": 5.45358502648532e-07,
+      "loss": 0.9045,
+      "step": 3319
+    },
+    {
+      "epoch": 2.688259109311741,
+      "grad_norm": 1.6281167669786418,
+      "learning_rate": 5.425622719633428e-07,
+      "loss": 0.997,
+      "step": 3320
+    },
+    {
+      "epoch": 2.6890688259109314,
+      "grad_norm": 1.6635114803247797,
+      "learning_rate": 5.397730284020863e-07,
+      "loss": 1.04,
+      "step": 3321
+    },
+    {
+      "epoch": 2.6898785425101215,
+      "grad_norm": 1.6162289487244115,
+      "learning_rate": 5.369907740254454e-07,
+      "loss": 0.9811,
+      "step": 3322
+    },
+    {
+      "epoch": 2.6906882591093115,
+      "grad_norm": 1.7327330282080413,
+      "learning_rate": 5.342155108889391e-07,
+      "loss": 0.9866,
+      "step": 3323
+    },
+    {
+      "epoch": 2.691497975708502,
+      "grad_norm": 1.673161064229558,
+      "learning_rate": 5.31447241042925e-07,
+      "loss": 0.9979,
+      "step": 3324
+    },
+    {
+      "epoch": 2.6923076923076925,
+      "grad_norm": 1.5832573352945025,
+      "learning_rate": 5.286859665325905e-07,
+      "loss": 0.9559,
+      "step": 3325
+    },
+    {
+      "epoch": 2.6931174089068826,
+      "grad_norm": 1.6724549773177813,
+      "learning_rate": 5.259316893979549e-07,
+      "loss": 0.9106,
+      "step": 3326
+    },
+    {
+      "epoch": 2.6939271255060726,
+      "grad_norm": 1.6314795157115118,
+      "learning_rate": 5.231844116738716e-07,
+      "loss": 0.972,
+      "step": 3327
+    },
+    {
+      "epoch": 2.694736842105263,
+      "grad_norm": 1.6811218193411817,
+      "learning_rate": 5.204441353900169e-07,
+      "loss": 0.9769,
+      "step": 3328
+    },
+    {
+      "epoch": 2.6955465587044536,
+      "grad_norm": 1.728655218184744,
+      "learning_rate": 5.177108625709026e-07,
+      "loss": 1.0373,
+      "step": 3329
+    },
+    {
+      "epoch": 2.6963562753036436,
+      "grad_norm": 1.652166979844383,
+      "learning_rate": 5.149845952358589e-07,
+      "loss": 1.0337,
+      "step": 3330
+    },
+    {
+      "epoch": 2.697165991902834,
+      "grad_norm": 1.668823613835129,
+      "learning_rate": 5.122653353990437e-07,
+      "loss": 0.9601,
+      "step": 3331
+    },
+    {
+      "epoch": 2.697975708502024,
+      "grad_norm": 1.619860141890713,
+      "learning_rate": 5.095530850694375e-07,
+      "loss": 1.0214,
+      "step": 3332
+    },
+    {
+      "epoch": 2.6987854251012147,
+      "grad_norm": 1.7472110705893544,
+      "learning_rate": 5.068478462508409e-07,
+      "loss": 0.9791,
+      "step": 3333
+    },
+    {
+      "epoch": 2.6995951417004047,
+      "grad_norm": 1.6679499678127792,
+      "learning_rate": 5.04149620941875e-07,
+      "loss": 0.9824,
+      "step": 3334
+    },
+    {
+      "epoch": 2.700404858299595,
+      "grad_norm": 1.6663417046865165,
+      "learning_rate": 5.014584111359811e-07,
+      "loss": 0.9539,
+      "step": 3335
+    },
+    {
+      "epoch": 2.7012145748987857,
+      "grad_norm": 1.672587987397243,
+      "learning_rate": 4.987742188214162e-07,
+      "loss": 0.9183,
+      "step": 3336
+    },
+    {
+      "epoch": 2.7020242914979757,
+      "grad_norm": 1.7188596938545961,
+      "learning_rate": 4.960970459812542e-07,
+      "loss": 0.9772,
+      "step": 3337
+    },
+    {
+      "epoch": 2.702834008097166,
+      "grad_norm": 1.6817958789804741,
+      "learning_rate": 4.934268945933784e-07,
+      "loss": 0.9823,
+      "step": 3338
+    },
+    {
+      "epoch": 2.7036437246963563,
+      "grad_norm": 1.6819764284995349,
+      "learning_rate": 4.907637666304898e-07,
+      "loss": 0.9582,
+      "step": 3339
+    },
+    {
+      "epoch": 2.7044534412955468,
+      "grad_norm": 1.6643127350965405,
+      "learning_rate": 4.881076640600979e-07,
+      "loss": 0.9677,
+      "step": 3340
+    },
+    {
+      "epoch": 2.705263157894737,
+      "grad_norm": 1.777174635148692,
+      "learning_rate": 4.854585888445218e-07,
+      "loss": 0.9946,
+      "step": 3341
+    },
+    {
+      "epoch": 2.706072874493927,
+      "grad_norm": 1.6692254557052726,
+      "learning_rate": 4.828165429408926e-07,
+      "loss": 1.0203,
+      "step": 3342
+    },
+    {
+      "epoch": 2.7068825910931174,
+      "grad_norm": 1.7053185479226727,
+      "learning_rate": 4.801815283011413e-07,
+      "loss": 0.9442,
+      "step": 3343
+    },
+    {
+      "epoch": 2.707692307692308,
+      "grad_norm": 1.700154883033202,
+      "learning_rate": 4.775535468720105e-07,
+      "loss": 0.9464,
+      "step": 3344
+    },
+    {
+      "epoch": 2.708502024291498,
+      "grad_norm": 1.7405575127592974,
+      "learning_rate": 4.7493260059504497e-07,
+      "loss": 0.9651,
+      "step": 3345
+    },
+    {
+      "epoch": 2.7093117408906884,
+      "grad_norm": 1.6499300506429417,
+      "learning_rate": 4.7231869140658804e-07,
+      "loss": 0.9186,
+      "step": 3346
+    },
+    {
+      "epoch": 2.7101214574898784,
+      "grad_norm": 1.684334696306486,
+      "learning_rate": 4.6971182123779045e-07,
+      "loss": 0.8985,
+      "step": 3347
+    },
+    {
+      "epoch": 2.710931174089069,
+      "grad_norm": 1.6675095321437483,
+      "learning_rate": 4.6711199201459833e-07,
+      "loss": 1.0096,
+      "step": 3348
+    },
+    {
+      "epoch": 2.711740890688259,
+      "grad_norm": 1.7918664382749507,
+      "learning_rate": 4.645192056577541e-07,
+      "loss": 1.0911,
+      "step": 3349
+    },
+    {
+      "epoch": 2.7125506072874495,
+      "grad_norm": 1.7239165785172468,
+      "learning_rate": 4.6193346408280216e-07,
+      "loss": 1.0048,
+      "step": 3350
+    },
+    {
+      "epoch": 2.71336032388664,
+      "grad_norm": 1.6745541046010592,
+      "learning_rate": 4.5935476920008213e-07,
+      "loss": 0.9741,
+      "step": 3351
+    },
+    {
+      "epoch": 2.71417004048583,
+      "grad_norm": 1.770541331198052,
+      "learning_rate": 4.5678312291472347e-07,
+      "loss": 0.9229,
+      "step": 3352
+    },
+    {
+      "epoch": 2.71497975708502,
+      "grad_norm": 1.7454452776343088,
+      "learning_rate": 4.542185271266486e-07,
+      "loss": 0.9203,
+      "step": 3353
+    },
+    {
+      "epoch": 2.7157894736842105,
+      "grad_norm": 1.6546940699373445,
+      "learning_rate": 4.516609837305741e-07,
+      "loss": 0.9511,
+      "step": 3354
+    },
+    {
+      "epoch": 2.716599190283401,
+      "grad_norm": 1.673530449759611,
+      "learning_rate": 4.491104946160052e-07,
+      "loss": 0.9926,
+      "step": 3355
+    },
+    {
+      "epoch": 2.717408906882591,
+      "grad_norm": 1.6784803821715035,
+      "learning_rate": 4.465670616672313e-07,
+      "loss": 0.951,
+      "step": 3356
+    },
+    {
+      "epoch": 2.718218623481781,
+      "grad_norm": 1.7286011969705144,
+      "learning_rate": 4.440306867633359e-07,
+      "loss": 0.9863,
+      "step": 3357
+    },
+    {
+      "epoch": 2.7190283400809716,
+      "grad_norm": 1.759405070641066,
+      "learning_rate": 4.4150137177818243e-07,
+      "loss": 1.0233,
+      "step": 3358
+    },
+    {
+      "epoch": 2.719838056680162,
+      "grad_norm": 1.7032342660835975,
+      "learning_rate": 4.389791185804237e-07,
+      "loss": 0.9514,
+      "step": 3359
+    },
+    {
+      "epoch": 2.720647773279352,
+      "grad_norm": 1.7760045827311535,
+      "learning_rate": 4.3646392903348823e-07,
+      "loss": 0.9509,
+      "step": 3360
+    },
+    {
+      "epoch": 2.7214574898785426,
+      "grad_norm": 1.703842102243129,
+      "learning_rate": 4.3395580499559276e-07,
+      "loss": 0.9985,
+      "step": 3361
+    },
+    {
+      "epoch": 2.7222672064777327,
+      "grad_norm": 1.7058090054545807,
+      "learning_rate": 4.3145474831972845e-07,
+      "loss": 0.994,
+      "step": 3362
+    },
+    {
+      "epoch": 2.723076923076923,
+      "grad_norm": 1.704324426436363,
+      "learning_rate": 4.2896076085367056e-07,
+      "loss": 0.9599,
+      "step": 3363
+    },
+    {
+      "epoch": 2.723886639676113,
+      "grad_norm": 1.6980076718179375,
+      "learning_rate": 4.264738444399652e-07,
+      "loss": 0.9828,
+      "step": 3364
+    },
+    {
+      "epoch": 2.7246963562753037,
+      "grad_norm": 1.6087858125711858,
+      "learning_rate": 4.2399400091594154e-07,
+      "loss": 1.0022,
+      "step": 3365
+    },
+    {
+      "epoch": 2.725506072874494,
+      "grad_norm": 1.6883922800088609,
+      "learning_rate": 4.2152123211369633e-07,
+      "loss": 0.9971,
+      "step": 3366
+    },
+    {
+      "epoch": 2.7263157894736842,
+      "grad_norm": 1.7249610106754352,
+      "learning_rate": 4.1905553986010707e-07,
+      "loss": 0.9756,
+      "step": 3367
+    },
+    {
+      "epoch": 2.7271255060728743,
+      "grad_norm": 1.650676985334564,
+      "learning_rate": 4.165969259768177e-07,
+      "loss": 0.9695,
+      "step": 3368
+    },
+    {
+      "epoch": 2.727935222672065,
+      "grad_norm": 1.7286467505133185,
+      "learning_rate": 4.1414539228024297e-07,
+      "loss": 1.0002,
+      "step": 3369
+    },
+    {
+      "epoch": 2.7287449392712553,
+      "grad_norm": 1.7055124440922338,
+      "learning_rate": 4.117009405815686e-07,
+      "loss": 0.9834,
+      "step": 3370
+    },
+    {
+      "epoch": 2.7295546558704453,
+      "grad_norm": 1.7056058182983371,
+      "learning_rate": 4.0926357268674667e-07,
+      "loss": 0.9676,
+      "step": 3371
+    },
+    {
+      "epoch": 2.7303643724696354,
+      "grad_norm": 1.7942169407381743,
+      "learning_rate": 4.068332903964978e-07,
+      "loss": 1.0179,
+      "step": 3372
+    },
+    {
+      "epoch": 2.731174089068826,
+      "grad_norm": 1.7569046659689378,
+      "learning_rate": 4.0441009550630683e-07,
+      "loss": 0.9497,
+      "step": 3373
+    },
+    {
+      "epoch": 2.7319838056680164,
+      "grad_norm": 1.7915957106720466,
+      "learning_rate": 4.0199398980641955e-07,
+      "loss": 1.0083,
+      "step": 3374
+    },
+    {
+      "epoch": 2.7327935222672064,
+      "grad_norm": 1.6604774269182196,
+      "learning_rate": 3.9958497508185036e-07,
+      "loss": 0.9262,
+      "step": 3375
+    },
+    {
+      "epoch": 2.733603238866397,
+      "grad_norm": 1.7981493121093652,
+      "learning_rate": 3.9718305311236996e-07,
+      "loss": 0.9811,
+      "step": 3376
+    },
+    {
+      "epoch": 2.734412955465587,
+      "grad_norm": 1.7140059177794198,
+      "learning_rate": 3.9478822567251e-07,
+      "loss": 1.02,
+      "step": 3377
+    },
+    {
+      "epoch": 2.7352226720647774,
+      "grad_norm": 1.7037447623006154,
+      "learning_rate": 3.924004945315618e-07,
+      "loss": 1.0029,
+      "step": 3378
+    },
+    {
+      "epoch": 2.7360323886639675,
+      "grad_norm": 1.6598577339137044,
+      "learning_rate": 3.900198614535711e-07,
+      "loss": 0.9981,
+      "step": 3379
+    },
+    {
+      "epoch": 2.736842105263158,
+      "grad_norm": 1.643004118226858,
+      "learning_rate": 3.8764632819734526e-07,
+      "loss": 0.9708,
+      "step": 3380
+    },
+    {
+      "epoch": 2.737651821862348,
+      "grad_norm": 1.6338597557404355,
+      "learning_rate": 3.852798965164406e-07,
+      "loss": 1.0315,
+      "step": 3381
+    },
+    {
+      "epoch": 2.7384615384615385,
+      "grad_norm": 1.6815159903047672,
+      "learning_rate": 3.8292056815916965e-07,
+      "loss": 0.9838,
+      "step": 3382
+    },
+    {
+      "epoch": 2.7392712550607285,
+      "grad_norm": 1.6880628778391817,
+      "learning_rate": 3.805683448685971e-07,
+      "loss": 0.9882,
+      "step": 3383
+    },
+    {
+      "epoch": 2.740080971659919,
+      "grad_norm": 1.7337464917612249,
+      "learning_rate": 3.782232283825371e-07,
+      "loss": 0.9449,
+      "step": 3384
+    },
+    {
+      "epoch": 2.7408906882591095,
+      "grad_norm": 1.618990796172125,
+      "learning_rate": 3.758852204335539e-07,
+      "loss": 0.9886,
+      "step": 3385
+    },
+    {
+      "epoch": 2.7417004048582996,
+      "grad_norm": 1.7016005187057022,
+      "learning_rate": 3.735543227489591e-07,
+      "loss": 1.032,
+      "step": 3386
+    },
+    {
+      "epoch": 2.7425101214574896,
+      "grad_norm": 1.6521289220479534,
+      "learning_rate": 3.712305370508151e-07,
+      "loss": 0.9023,
+      "step": 3387
+    },
+    {
+      "epoch": 2.74331983805668,
+      "grad_norm": 1.6818920788449179,
+      "learning_rate": 3.6891386505592543e-07,
+      "loss": 0.9534,
+      "step": 3388
+    },
+    {
+      "epoch": 2.7441295546558706,
+      "grad_norm": 1.6371230429182253,
+      "learning_rate": 3.6660430847583973e-07,
+      "loss": 1.0219,
+      "step": 3389
+    },
+    {
+      "epoch": 2.7449392712550607,
+      "grad_norm": 1.6513777881762906,
+      "learning_rate": 3.643018690168487e-07,
+      "loss": 1.0237,
+      "step": 3390
+    },
+    {
+      "epoch": 2.745748987854251,
+      "grad_norm": 1.702844558572882,
+      "learning_rate": 3.620065483799917e-07,
+      "loss": 1.0022,
+      "step": 3391
+    },
+    {
+      "epoch": 2.746558704453441,
+      "grad_norm": 1.6684488841014413,
+      "learning_rate": 3.5971834826104114e-07,
+      "loss": 1.0313,
+      "step": 3392
+    },
+    {
+      "epoch": 2.7473684210526317,
+      "grad_norm": 1.678173521295049,
+      "learning_rate": 3.5743727035051245e-07,
+      "loss": 0.9862,
+      "step": 3393
+    },
+    {
+      "epoch": 2.7481781376518217,
+      "grad_norm": 1.6963824088714325,
+      "learning_rate": 3.551633163336565e-07,
+      "loss": 0.9448,
+      "step": 3394
+    },
+    {
+      "epoch": 2.748987854251012,
+      "grad_norm": 1.650614217397481,
+      "learning_rate": 3.5289648789046616e-07,
+      "loss": 0.9745,
+      "step": 3395
+    },
+    {
+      "epoch": 2.7497975708502023,
+      "grad_norm": 1.6241513485984185,
+      "learning_rate": 3.5063678669566616e-07,
+      "loss": 0.9782,
+      "step": 3396
+    },
+    {
+      "epoch": 2.7506072874493928,
+      "grad_norm": 1.7366532492678954,
+      "learning_rate": 3.4838421441871553e-07,
+      "loss": 0.9545,
+      "step": 3397
+    },
+    {
+      "epoch": 2.751417004048583,
+      "grad_norm": 1.643649947081736,
+      "learning_rate": 3.4613877272380526e-07,
+      "loss": 0.9422,
+      "step": 3398
+    },
+    {
+      "epoch": 2.7522267206477733,
+      "grad_norm": 1.7397550528137697,
+      "learning_rate": 3.4390046326986506e-07,
+      "loss": 0.9788,
+      "step": 3399
+    },
+    {
+      "epoch": 2.753036437246964,
+      "grad_norm": 1.7012047353401265,
+      "learning_rate": 3.4166928771054653e-07,
+      "loss": 1.0243,
+      "step": 3400
+    },
+    {
+      "epoch": 2.753846153846154,
+      "grad_norm": 1.6224734532349014,
+      "learning_rate": 3.394452476942367e-07,
+      "loss": 1.0144,
+      "step": 3401
+    },
+    {
+      "epoch": 2.754655870445344,
+      "grad_norm": 1.6480456374209134,
+      "learning_rate": 3.37228344864049e-07,
+      "loss": 0.9859,
+      "step": 3402
+    },
+    {
+      "epoch": 2.7554655870445344,
+      "grad_norm": 1.6460293515314257,
+      "learning_rate": 3.350185808578232e-07,
+      "loss": 1.0034,
+      "step": 3403
+    },
+    {
+      "epoch": 2.756275303643725,
+      "grad_norm": 1.757720445735907,
+      "learning_rate": 3.328159573081258e-07,
+      "loss": 1.0231,
+      "step": 3404
+    },
+    {
+      "epoch": 2.757085020242915,
+      "grad_norm": 1.7467262631474252,
+      "learning_rate": 3.3062047584224934e-07,
+      "loss": 1.0036,
+      "step": 3405
+    },
+    {
+      "epoch": 2.7578947368421054,
+      "grad_norm": 1.665170966625211,
+      "learning_rate": 3.284321380822053e-07,
+      "loss": 1.0402,
+      "step": 3406
+    },
+    {
+      "epoch": 2.7587044534412954,
+      "grad_norm": 1.6956185677978453,
+      "learning_rate": 3.262509456447327e-07,
+      "loss": 0.998,
+      "step": 3407
+    },
+    {
+      "epoch": 2.759514170040486,
+      "grad_norm": 1.7519136605866905,
+      "learning_rate": 3.240769001412891e-07,
+      "loss": 0.9528,
+      "step": 3408
+    },
+    {
+      "epoch": 2.760323886639676,
+      "grad_norm": 1.7155229720938194,
+      "learning_rate": 3.21910003178052e-07,
+      "loss": 0.9239,
+      "step": 3409
+    },
+    {
+      "epoch": 2.7611336032388665,
+      "grad_norm": 1.6753569936900223,
+      "learning_rate": 3.197502563559185e-07,
+      "loss": 1.0247,
+      "step": 3410
+    },
+    {
+      "epoch": 2.7619433198380565,
+      "grad_norm": 1.6291004521416133,
+      "learning_rate": 3.1759766127050116e-07,
+      "loss": 1.0405,
+      "step": 3411
+    },
+    {
+      "epoch": 2.762753036437247,
+      "grad_norm": 1.7193595883451167,
+      "learning_rate": 3.1545221951213125e-07,
+      "loss": 0.9007,
+      "step": 3412
+    },
+    {
+      "epoch": 2.763562753036437,
+      "grad_norm": 1.7346002461970236,
+      "learning_rate": 3.13313932665853e-07,
+      "loss": 0.922,
+      "step": 3413
+    },
+    {
+      "epoch": 2.7643724696356275,
+      "grad_norm": 1.6724531692395965,
+      "learning_rate": 3.1118280231142496e-07,
+      "loss": 0.9998,
+      "step": 3414
+    },
+    {
+      "epoch": 2.765182186234818,
+      "grad_norm": 1.7916657747046403,
+      "learning_rate": 3.0905883002332213e-07,
+      "loss": 0.9282,
+      "step": 3415
+    },
+    {
+      "epoch": 2.765991902834008,
+      "grad_norm": 1.6206762833247699,
+      "learning_rate": 3.069420173707249e-07,
+      "loss": 1.0229,
+      "step": 3416
+    },
+    {
+      "epoch": 2.766801619433198,
+      "grad_norm": 1.7079187810682614,
+      "learning_rate": 3.048323659175301e-07,
+      "loss": 0.9421,
+      "step": 3417
+    },
+    {
+      "epoch": 2.7676113360323886,
+      "grad_norm": 1.6489425364151793,
+      "learning_rate": 3.027298772223419e-07,
+      "loss": 1.0534,
+      "step": 3418
+    },
+    {
+      "epoch": 2.768421052631579,
+      "grad_norm": 1.6128333988404466,
+      "learning_rate": 3.006345528384691e-07,
+      "loss": 0.9396,
+      "step": 3419
+    },
+    {
+      "epoch": 2.769230769230769,
+      "grad_norm": 1.6925085208629478,
+      "learning_rate": 2.985463943139322e-07,
+      "loss": 0.9452,
+      "step": 3420
+    },
+    {
+      "epoch": 2.7700404858299597,
+      "grad_norm": 1.728805527613981,
+      "learning_rate": 2.96465403191456e-07,
+      "loss": 0.9629,
+      "step": 3421
+    },
+    {
+      "epoch": 2.7708502024291497,
+      "grad_norm": 1.6026531892520557,
+      "learning_rate": 2.943915810084685e-07,
+      "loss": 0.9992,
+      "step": 3422
+    },
+    {
+      "epoch": 2.77165991902834,
+      "grad_norm": 1.7066029483605916,
+      "learning_rate": 2.923249292971042e-07,
+      "loss": 1.0369,
+      "step": 3423
+    },
+    {
+      "epoch": 2.7724696356275302,
+      "grad_norm": 1.6385263272251447,
+      "learning_rate": 2.9026544958419833e-07,
+      "loss": 1.0205,
+      "step": 3424
+    },
+    {
+      "epoch": 2.7732793522267207,
+      "grad_norm": 1.6324065649477524,
+      "learning_rate": 2.882131433912883e-07,
+      "loss": 0.9805,
+      "step": 3425
+    },
+    {
+      "epoch": 2.7740890688259108,
+      "grad_norm": 1.6647087033829902,
+      "learning_rate": 2.8616801223461e-07,
+      "loss": 1.0149,
+      "step": 3426
+    },
+    {
+      "epoch": 2.7748987854251013,
+      "grad_norm": 1.7242997046116755,
+      "learning_rate": 2.841300576250994e-07,
+      "loss": 0.9302,
+      "step": 3427
+    },
+    {
+      "epoch": 2.7757085020242913,
+      "grad_norm": 1.6351854391159,
+      "learning_rate": 2.8209928106839204e-07,
+      "loss": 0.9689,
+      "step": 3428
+    },
+    {
+      "epoch": 2.776518218623482,
+      "grad_norm": 1.6547619822957484,
+      "learning_rate": 2.800756840648178e-07,
+      "loss": 1.041,
+      "step": 3429
+    },
+    {
+      "epoch": 2.7773279352226723,
+      "grad_norm": 1.6907136500019693,
+      "learning_rate": 2.7805926810940297e-07,
+      "loss": 1.0214,
+      "step": 3430
+    },
+    {
+      "epoch": 2.7781376518218623,
+      "grad_norm": 1.6858591321540695,
+      "learning_rate": 2.7605003469187044e-07,
+      "loss": 1.0106,
+      "step": 3431
+    },
+    {
+      "epoch": 2.7789473684210524,
+      "grad_norm": 1.761735692393002,
+      "learning_rate": 2.74047985296636e-07,
+      "loss": 1.0111,
+      "step": 3432
+    },
+    {
+      "epoch": 2.779757085020243,
+      "grad_norm": 1.6840626486171097,
+      "learning_rate": 2.720531214028055e-07,
+      "loss": 0.9443,
+      "step": 3433
+    },
+    {
+      "epoch": 2.7805668016194334,
+      "grad_norm": 1.7212890453015592,
+      "learning_rate": 2.700654444841777e-07,
+      "loss": 0.9862,
+      "step": 3434
+    },
+    {
+      "epoch": 2.7813765182186234,
+      "grad_norm": 1.6376530049169284,
+      "learning_rate": 2.6808495600924355e-07,
+      "loss": 0.9878,
+      "step": 3435
+    },
+    {
+      "epoch": 2.782186234817814,
+      "grad_norm": 1.6855350369784343,
+      "learning_rate": 2.661116574411793e-07,
+      "loss": 0.9569,
+      "step": 3436
+    },
+    {
+      "epoch": 2.782995951417004,
+      "grad_norm": 1.774069519379109,
+      "learning_rate": 2.6414555023785204e-07,
+      "loss": 0.9306,
+      "step": 3437
+    },
+    {
+      "epoch": 2.7838056680161944,
+      "grad_norm": 1.592262691990277,
+      "learning_rate": 2.6218663585181547e-07,
+      "loss": 1.0135,
+      "step": 3438
+    },
+    {
+      "epoch": 2.7846153846153845,
+      "grad_norm": 1.7023463434984625,
+      "learning_rate": 2.602349157303108e-07,
+      "loss": 0.9985,
+      "step": 3439
+    },
+    {
+      "epoch": 2.785425101214575,
+      "grad_norm": 1.7075079740626764,
+      "learning_rate": 2.582903913152612e-07,
+      "loss": 0.9859,
+      "step": 3440
+    },
+    {
+      "epoch": 2.786234817813765,
+      "grad_norm": 1.6394777254757238,
+      "learning_rate": 2.563530640432732e-07,
+      "loss": 0.9818,
+      "step": 3441
+    },
+    {
+      "epoch": 2.7870445344129555,
+      "grad_norm": 1.6722333217908798,
+      "learning_rate": 2.5442293534564067e-07,
+      "loss": 0.9843,
+      "step": 3442
+    },
+    {
+      "epoch": 2.7878542510121456,
+      "grad_norm": 1.64564423365469,
+      "learning_rate": 2.525000066483352e-07,
+      "loss": 0.9853,
+      "step": 3443
+    },
+    {
+      "epoch": 2.788663967611336,
+      "grad_norm": 1.5885870007756162,
+      "learning_rate": 2.5058427937200816e-07,
+      "loss": 0.9864,
+      "step": 3444
+    },
+    {
+      "epoch": 2.7894736842105265,
+      "grad_norm": 1.6567088157067262,
+      "learning_rate": 2.4867575493199515e-07,
+      "loss": 0.9909,
+      "step": 3445
+    },
+    {
+      "epoch": 2.7902834008097166,
+      "grad_norm": 1.7241493918176931,
+      "learning_rate": 2.467744347383072e-07,
+      "loss": 0.9471,
+      "step": 3446
+    },
+    {
+      "epoch": 2.7910931174089066,
+      "grad_norm": 1.5953145152003532,
+      "learning_rate": 2.44880320195634e-07,
+      "loss": 1.0425,
+      "step": 3447
+    },
+    {
+      "epoch": 2.791902834008097,
+      "grad_norm": 1.704422723526784,
+      "learning_rate": 2.4299341270333955e-07,
+      "loss": 0.9508,
+      "step": 3448
+    },
+    {
+      "epoch": 2.7927125506072876,
+      "grad_norm": 1.6615202773394242,
+      "learning_rate": 2.4111371365546643e-07,
+      "loss": 0.9808,
+      "step": 3449
+    },
+    {
+      "epoch": 2.7935222672064777,
+      "grad_norm": 1.6622420527909882,
+      "learning_rate": 2.392412244407294e-07,
+      "loss": 0.9952,
+      "step": 3450
+    },
+    {
+      "epoch": 2.794331983805668,
+      "grad_norm": 1.7028019856102616,
+      "learning_rate": 2.373759464425174e-07,
+      "loss": 1.0009,
+      "step": 3451
+    },
+    {
+      "epoch": 2.795141700404858,
+      "grad_norm": 1.6318324035656953,
+      "learning_rate": 2.3551788103889027e-07,
+      "loss": 0.9645,
+      "step": 3452
+    },
+    {
+      "epoch": 2.7959514170040487,
+      "grad_norm": 1.6683299909476288,
+      "learning_rate": 2.3366702960258336e-07,
+      "loss": 1.0226,
+      "step": 3453
+    },
+    {
+      "epoch": 2.7967611336032387,
+      "grad_norm": 1.7972866276194772,
+      "learning_rate": 2.3182339350099724e-07,
+      "loss": 0.9662,
+      "step": 3454
+    },
+    {
+      "epoch": 2.7975708502024292,
+      "grad_norm": 1.7250223932824822,
+      "learning_rate": 2.2998697409620573e-07,
+      "loss": 1.0532,
+      "step": 3455
+    },
+    {
+      "epoch": 2.7983805668016193,
+      "grad_norm": 1.6934825177542452,
+      "learning_rate": 2.2815777274495022e-07,
+      "loss": 1.0051,
+      "step": 3456
+    },
+    {
+      "epoch": 2.7991902834008098,
+      "grad_norm": 1.7033740545061948,
+      "learning_rate": 2.2633579079863632e-07,
+      "loss": 0.997,
+      "step": 3457
+    },
+    {
+      "epoch": 2.8,
+      "grad_norm": 1.7292274929656826,
+      "learning_rate": 2.2452102960334064e-07,
+      "loss": 0.9835,
+      "step": 3458
+    },
+    {
+      "epoch": 2.8008097165991903,
+      "grad_norm": 1.6514262457771633,
+      "learning_rate": 2.2271349049979962e-07,
+      "loss": 0.9233,
+      "step": 3459
+    },
+    {
+      "epoch": 2.801619433198381,
+      "grad_norm": 1.7415649309506156,
+      "learning_rate": 2.2091317482342056e-07,
+      "loss": 0.9879,
+      "step": 3460
+    },
+    {
+      "epoch": 2.802429149797571,
+      "grad_norm": 1.6947139567148872,
+      "learning_rate": 2.1912008390426953e-07,
+      "loss": 0.9346,
+      "step": 3461
+    },
+    {
+      "epoch": 2.803238866396761,
+      "grad_norm": 1.760629650452238,
+      "learning_rate": 2.1733421906707464e-07,
+      "loss": 1.0005,
+      "step": 3462
+    },
+    {
+      "epoch": 2.8040485829959514,
+      "grad_norm": 1.6060526892387195,
+      "learning_rate": 2.1555558163122935e-07,
+      "loss": 1.075,
+      "step": 3463
+    },
+    {
+      "epoch": 2.804858299595142,
+      "grad_norm": 1.6926722698944767,
+      "learning_rate": 2.137841729107848e-07,
+      "loss": 0.8934,
+      "step": 3464
+    },
+    {
+      "epoch": 2.805668016194332,
+      "grad_norm": 1.6918582251550736,
+      "learning_rate": 2.1201999421445074e-07,
+      "loss": 0.9724,
+      "step": 3465
+    },
+    {
+      "epoch": 2.8064777327935224,
+      "grad_norm": 1.6163610863078754,
+      "learning_rate": 2.1026304684559685e-07,
+      "loss": 1.0206,
+      "step": 3466
+    },
+    {
+      "epoch": 2.8072874493927125,
+      "grad_norm": 1.6797477934507816,
+      "learning_rate": 2.0851333210225032e-07,
+      "loss": 0.9478,
+      "step": 3467
+    },
+    {
+      "epoch": 2.808097165991903,
+      "grad_norm": 1.6836631767653443,
+      "learning_rate": 2.0677085127709495e-07,
+      "loss": 1.0166,
+      "step": 3468
+    },
+    {
+      "epoch": 2.808906882591093,
+      "grad_norm": 1.7134147411086924,
+      "learning_rate": 2.0503560565747092e-07,
+      "loss": 0.9374,
+      "step": 3469
+    },
+    {
+      "epoch": 2.8097165991902835,
+      "grad_norm": 1.7189098388876702,
+      "learning_rate": 2.0330759652536835e-07,
+      "loss": 0.9788,
+      "step": 3470
+    },
+    {
+      "epoch": 2.8105263157894735,
+      "grad_norm": 1.6505141468009865,
+      "learning_rate": 2.0158682515743933e-07,
+      "loss": 0.9734,
+      "step": 3471
+    },
+    {
+      "epoch": 2.811336032388664,
+      "grad_norm": 1.6506249870105763,
+      "learning_rate": 1.9987329282498024e-07,
+      "loss": 0.9544,
+      "step": 3472
+    },
+    {
+      "epoch": 2.812145748987854,
+      "grad_norm": 1.7316946412284435,
+      "learning_rate": 1.9816700079394625e-07,
+      "loss": 1.0232,
+      "step": 3473
+    },
+    {
+      "epoch": 2.8129554655870446,
+      "grad_norm": 1.6533763703707396,
+      "learning_rate": 1.964679503249367e-07,
+      "loss": 0.9491,
+      "step": 3474
+    },
+    {
+      "epoch": 2.813765182186235,
+      "grad_norm": 1.76162742728525,
+      "learning_rate": 1.9477614267320867e-07,
+      "loss": 1.0347,
+      "step": 3475
+    },
+    {
+      "epoch": 2.814574898785425,
+      "grad_norm": 1.6828209980534072,
+      "learning_rate": 1.9309157908866116e-07,
+      "loss": 0.9581,
+      "step": 3476
+    },
+    {
+      "epoch": 2.815384615384615,
+      "grad_norm": 1.674090593565143,
+      "learning_rate": 1.9141426081584537e-07,
+      "loss": 0.9604,
+      "step": 3477
+    },
+    {
+      "epoch": 2.8161943319838056,
+      "grad_norm": 1.725348808312188,
+      "learning_rate": 1.8974418909395774e-07,
+      "loss": 1.0223,
+      "step": 3478
+    },
+    {
+      "epoch": 2.817004048582996,
+      "grad_norm": 1.6949407572974873,
+      "learning_rate": 1.880813651568425e-07,
+      "loss": 1.0619,
+      "step": 3479
+    },
+    {
+      "epoch": 2.817813765182186,
+      "grad_norm": 1.662538589931312,
+      "learning_rate": 1.8642579023298913e-07,
+      "loss": 1.0112,
+      "step": 3480
+    },
+    {
+      "epoch": 2.8186234817813767,
+      "grad_norm": 1.7612232629543667,
+      "learning_rate": 1.8477746554552922e-07,
+      "loss": 0.9445,
+      "step": 3481
+    },
+    {
+      "epoch": 2.8194331983805667,
+      "grad_norm": 1.6705151929988862,
+      "learning_rate": 1.831363923122409e-07,
+      "loss": 0.9485,
+      "step": 3482
+    },
+    {
+      "epoch": 2.820242914979757,
+      "grad_norm": 1.6573033708453293,
+      "learning_rate": 1.815025717455432e-07,
+      "loss": 0.9873,
+      "step": 3483
+    },
+    {
+      "epoch": 2.8210526315789473,
+      "grad_norm": 1.6767470480305524,
+      "learning_rate": 1.7987600505249726e-07,
+      "loss": 0.9749,
+      "step": 3484
+    },
+    {
+      "epoch": 2.8218623481781377,
+      "grad_norm": 1.7183296390160356,
+      "learning_rate": 1.7825669343480624e-07,
+      "loss": 0.9317,
+      "step": 3485
+    },
+    {
+      "epoch": 2.822672064777328,
+      "grad_norm": 1.6844074820085622,
+      "learning_rate": 1.7664463808880983e-07,
+      "loss": 0.9173,
+      "step": 3486
+    },
+    {
+      "epoch": 2.8234817813765183,
+      "grad_norm": 1.6126455021944093,
+      "learning_rate": 1.7503984020549203e-07,
+      "loss": 0.9177,
+      "step": 3487
+    },
+    {
+      "epoch": 2.8242914979757083,
+      "grad_norm": 1.674650005216152,
+      "learning_rate": 1.7344230097047111e-07,
+      "loss": 0.9383,
+      "step": 3488
+    },
+    {
+      "epoch": 2.825101214574899,
+      "grad_norm": 1.7018972026288255,
+      "learning_rate": 1.7185202156400294e-07,
+      "loss": 0.9821,
+      "step": 3489
+    },
+    {
+      "epoch": 2.8259109311740893,
+      "grad_norm": 1.6593676997323308,
+      "learning_rate": 1.7026900316098217e-07,
+      "loss": 0.9953,
+      "step": 3490
+    },
+    {
+      "epoch": 2.8267206477732794,
+      "grad_norm": 1.6763973087070592,
+      "learning_rate": 1.6869324693093768e-07,
+      "loss": 0.97,
+      "step": 3491
+    },
+    {
+      "epoch": 2.8275303643724694,
+      "grad_norm": 1.6756347235552076,
+      "learning_rate": 1.6712475403803164e-07,
+      "loss": 1.0831,
+      "step": 3492
+    },
+    {
+      "epoch": 2.82834008097166,
+      "grad_norm": 1.6961902060855358,
+      "learning_rate": 1.655635256410615e-07,
+      "loss": 1.0187,
+      "step": 3493
+    },
+    {
+      "epoch": 2.8291497975708504,
+      "grad_norm": 1.7106926602042642,
+      "learning_rate": 1.6400956289345903e-07,
+      "loss": 0.9455,
+      "step": 3494
+    },
+    {
+      "epoch": 2.8299595141700404,
+      "grad_norm": 1.6339795546096871,
+      "learning_rate": 1.6246286694328594e-07,
+      "loss": 1.0447,
+      "step": 3495
+    },
+    {
+      "epoch": 2.830769230769231,
+      "grad_norm": 1.6545843031250616,
+      "learning_rate": 1.6092343893323593e-07,
+      "loss": 0.9805,
+      "step": 3496
+    },
+    {
+      "epoch": 2.831578947368421,
+      "grad_norm": 1.6809638570842909,
+      "learning_rate": 1.5939128000063364e-07,
+      "loss": 0.9301,
+      "step": 3497
+    },
+    {
+      "epoch": 2.8323886639676115,
+      "grad_norm": 1.699507326131043,
+      "learning_rate": 1.5786639127743363e-07,
+      "loss": 0.9757,
+      "step": 3498
+    },
+    {
+      "epoch": 2.8331983805668015,
+      "grad_norm": 1.7076118973222212,
+      "learning_rate": 1.5634877389021695e-07,
+      "loss": 0.9538,
+      "step": 3499
+    },
+    {
+      "epoch": 2.834008097165992,
+      "grad_norm": 1.6599897623636726,
+      "learning_rate": 1.5483842896019675e-07,
+      "loss": 1.0356,
+      "step": 3500
+    },
+    {
+      "epoch": 2.834817813765182,
+      "grad_norm": 1.71624063763336,
+      "learning_rate": 1.5333535760320929e-07,
+      "loss": 0.96,
+      "step": 3501
+    },
+    {
+      "epoch": 2.8356275303643725,
+      "grad_norm": 1.7240409170070614,
+      "learning_rate": 1.518395609297185e-07,
+      "loss": 0.9491,
+      "step": 3502
+    },
+    {
+      "epoch": 2.8364372469635626,
+      "grad_norm": 1.6299987076331925,
+      "learning_rate": 1.5035104004481604e-07,
+      "loss": 1.0197,
+      "step": 3503
+    },
+    {
+      "epoch": 2.837246963562753,
+      "grad_norm": 1.5841940758881194,
+      "learning_rate": 1.4886979604821328e-07,
+      "loss": 0.9758,
+      "step": 3504
+    },
+    {
+      "epoch": 2.8380566801619436,
+      "grad_norm": 1.6462756907053675,
+      "learning_rate": 1.473958300342504e-07,
+      "loss": 0.975,
+      "step": 3505
+    },
+    {
+      "epoch": 2.8388663967611336,
+      "grad_norm": 1.6923662140676001,
+      "learning_rate": 1.4592914309188965e-07,
+      "loss": 0.9598,
+      "step": 3506
+    },
+    {
+      "epoch": 2.8396761133603237,
+      "grad_norm": 1.71758711743743,
+      "learning_rate": 1.4446973630471207e-07,
+      "loss": 0.9187,
+      "step": 3507
+    },
+    {
+      "epoch": 2.840485829959514,
+      "grad_norm": 1.7314496796265455,
+      "learning_rate": 1.4301761075092402e-07,
+      "loss": 0.9947,
+      "step": 3508
+    },
+    {
+      "epoch": 2.8412955465587046,
+      "grad_norm": 1.6658482977303863,
+      "learning_rate": 1.415727675033507e-07,
+      "loss": 0.9994,
+      "step": 3509
+    },
+    {
+      "epoch": 2.8421052631578947,
+      "grad_norm": 1.652120289375282,
+      "learning_rate": 1.401352076294371e-07,
+      "loss": 1.0019,
+      "step": 3510
+    },
+    {
+      "epoch": 2.842914979757085,
+      "grad_norm": 1.7489773597175606,
+      "learning_rate": 1.3870493219124814e-07,
+      "loss": 1.0033,
+      "step": 3511
+    },
+    {
+      "epoch": 2.8437246963562752,
+      "grad_norm": 1.703873872469804,
+      "learning_rate": 1.3728194224546742e-07,
+      "loss": 0.9716,
+      "step": 3512
+    },
+    {
+      "epoch": 2.8445344129554657,
+      "grad_norm": 1.7380413078833634,
+      "learning_rate": 1.35866238843394e-07,
+      "loss": 0.9694,
+      "step": 3513
+    },
+    {
+      "epoch": 2.8453441295546558,
+      "grad_norm": 1.7289000733110216,
+      "learning_rate": 1.3445782303094568e-07,
+      "loss": 0.9484,
+      "step": 3514
+    },
+    {
+      "epoch": 2.8461538461538463,
+      "grad_norm": 1.748270696926121,
+      "learning_rate": 1.3305669584865565e-07,
+      "loss": 0.9799,
+      "step": 3515
+    },
+    {
+      "epoch": 2.8469635627530363,
+      "grad_norm": 1.6413747236212781,
+      "learning_rate": 1.3166285833167147e-07,
+      "loss": 1.0151,
+      "step": 3516
+    },
+    {
+      "epoch": 2.847773279352227,
+      "grad_norm": 1.6667240650346242,
+      "learning_rate": 1.3027631150975606e-07,
+      "loss": 0.9758,
+      "step": 3517
+    },
+    {
+      "epoch": 2.848582995951417,
+      "grad_norm": 1.6872730298067318,
+      "learning_rate": 1.2889705640728445e-07,
+      "loss": 0.9228,
+      "step": 3518
+    },
+    {
+      "epoch": 2.8493927125506073,
+      "grad_norm": 1.6295638968820234,
+      "learning_rate": 1.275250940432471e-07,
+      "loss": 0.979,
+      "step": 3519
+    },
+    {
+      "epoch": 2.850202429149798,
+      "grad_norm": 1.6588363373893749,
+      "learning_rate": 1.261604254312454e-07,
+      "loss": 1.0192,
+      "step": 3520
+    },
+    {
+      "epoch": 2.851012145748988,
+      "grad_norm": 1.6683421475461793,
+      "learning_rate": 1.248030515794907e-07,
+      "loss": 0.9543,
+      "step": 3521
+    },
+    {
+      "epoch": 2.851821862348178,
+      "grad_norm": 1.6524719881626193,
+      "learning_rate": 1.2345297349080852e-07,
+      "loss": 0.9376,
+      "step": 3522
+    },
+    {
+      "epoch": 2.8526315789473684,
+      "grad_norm": 1.698816182274864,
+      "learning_rate": 1.221101921626311e-07,
+      "loss": 0.969,
+      "step": 3523
+    },
+    {
+      "epoch": 2.853441295546559,
+      "grad_norm": 1.6990536494185664,
+      "learning_rate": 1.2077470858699925e-07,
+      "loss": 0.9081,
+      "step": 3524
+    },
+    {
+      "epoch": 2.854251012145749,
+      "grad_norm": 1.6689902626693056,
+      "learning_rate": 1.1944652375056597e-07,
+      "loss": 0.9413,
+      "step": 3525
+    },
+    {
+      "epoch": 2.8550607287449394,
+      "grad_norm": 1.7541568188456584,
+      "learning_rate": 1.1812563863458859e-07,
+      "loss": 0.9368,
+      "step": 3526
+    },
+    {
+      "epoch": 2.8558704453441295,
+      "grad_norm": 1.6900489394163216,
+      "learning_rate": 1.1681205421493425e-07,
+      "loss": 1.0227,
+      "step": 3527
+    },
+    {
+      "epoch": 2.85668016194332,
+      "grad_norm": 1.6593250577589962,
+      "learning_rate": 1.1550577146207331e-07,
+      "loss": 0.9822,
+      "step": 3528
+    },
+    {
+      "epoch": 2.85748987854251,
+      "grad_norm": 1.706412070441469,
+      "learning_rate": 1.1420679134108382e-07,
+      "loss": 0.9934,
+      "step": 3529
+    },
+    {
+      "epoch": 2.8582995951417005,
+      "grad_norm": 1.6579862881360155,
+      "learning_rate": 1.1291511481164807e-07,
+      "loss": 0.9466,
+      "step": 3530
+    },
+    {
+      "epoch": 2.8591093117408906,
+      "grad_norm": 1.7170922289715347,
+      "learning_rate": 1.1163074282805165e-07,
+      "loss": 0.9103,
+      "step": 3531
+    },
+    {
+      "epoch": 2.859919028340081,
+      "grad_norm": 1.7114518472316398,
+      "learning_rate": 1.1035367633918436e-07,
+      "loss": 0.9184,
+      "step": 3532
+    },
+    {
+      "epoch": 2.860728744939271,
+      "grad_norm": 1.731323195336544,
+      "learning_rate": 1.0908391628854042e-07,
+      "loss": 0.9779,
+      "step": 3533
+    },
+    {
+      "epoch": 2.8615384615384616,
+      "grad_norm": 1.7111499506843564,
+      "learning_rate": 1.0782146361421275e-07,
+      "loss": 1.0177,
+      "step": 3534
+    },
+    {
+      "epoch": 2.862348178137652,
+      "grad_norm": 1.6337915289731522,
+      "learning_rate": 1.0656631924889749e-07,
+      "loss": 0.974,
+      "step": 3535
+    },
+    {
+      "epoch": 2.863157894736842,
+      "grad_norm": 1.6468027663173628,
+      "learning_rate": 1.0531848411989287e-07,
+      "loss": 1.0301,
+      "step": 3536
+    },
+    {
+      "epoch": 2.863967611336032,
+      "grad_norm": 1.704350995042486,
+      "learning_rate": 1.0407795914909258e-07,
+      "loss": 1.0112,
+      "step": 3537
+    },
+    {
+      "epoch": 2.8647773279352227,
+      "grad_norm": 1.6680112193827918,
+      "learning_rate": 1.0284474525299459e-07,
+      "loss": 1.0021,
+      "step": 3538
+    },
+    {
+      "epoch": 2.865587044534413,
+      "grad_norm": 1.630227489137973,
+      "learning_rate": 1.0161884334269234e-07,
+      "loss": 0.9871,
+      "step": 3539
+    },
+    {
+      "epoch": 2.866396761133603,
+      "grad_norm": 1.7269095533127516,
+      "learning_rate": 1.0040025432387801e-07,
+      "loss": 0.9789,
+      "step": 3540
+    },
+    {
+      "epoch": 2.8672064777327932,
+      "grad_norm": 1.645077368110772,
+      "learning_rate": 9.918897909684144e-08,
+      "loss": 0.9977,
+      "step": 3541
+    },
+    {
+      "epoch": 2.8680161943319837,
+      "grad_norm": 1.6531908723812594,
+      "learning_rate": 9.798501855646792e-08,
+      "loss": 1.0315,
+      "step": 3542
+    },
+    {
+      "epoch": 2.8688259109311742,
+      "grad_norm": 1.6773021644682622,
+      "learning_rate": 9.678837359224148e-08,
+      "loss": 0.9943,
+      "step": 3543
+    },
+    {
+      "epoch": 2.8696356275303643,
+      "grad_norm": 1.6110461386436772,
+      "learning_rate": 9.559904508823825e-08,
+      "loss": 0.9716,
+      "step": 3544
+    },
+    {
+      "epoch": 2.8704453441295548,
+      "grad_norm": 1.547369809807511,
+      "learning_rate": 9.441703392313095e-08,
+      "loss": 0.9595,
+      "step": 3545
+    },
+    {
+      "epoch": 2.871255060728745,
+      "grad_norm": 1.6684471743402367,
+      "learning_rate": 9.324234097018436e-08,
+      "loss": 0.9031,
+      "step": 3546
+    },
+    {
+      "epoch": 2.8720647773279353,
+      "grad_norm": 1.6770696827512974,
+      "learning_rate": 9.20749670972576e-08,
+      "loss": 0.9372,
+      "step": 3547
+    },
+    {
+      "epoch": 2.8728744939271254,
+      "grad_norm": 1.7263050739430634,
+      "learning_rate": 9.091491316680411e-08,
+      "loss": 0.997,
+      "step": 3548
+    },
+    {
+      "epoch": 2.873684210526316,
+      "grad_norm": 1.6963472723142492,
+      "learning_rate": 8.976218003586722e-08,
+      "loss": 1.0112,
+      "step": 3549
+    },
+    {
+      "epoch": 2.8744939271255063,
+      "grad_norm": 1.7236666081560383,
+      "learning_rate": 8.861676855608237e-08,
+      "loss": 0.9723,
+      "step": 3550
+    },
+    {
+      "epoch": 2.8753036437246964,
+      "grad_norm": 1.6959061896936787,
+      "learning_rate": 8.747867957367595e-08,
+      "loss": 0.9236,
+      "step": 3551
+    },
+    {
+      "epoch": 2.8761133603238864,
+      "grad_norm": 1.7019359454703777,
+      "learning_rate": 8.634791392946429e-08,
+      "loss": 0.9874,
+      "step": 3552
+    },
+    {
+      "epoch": 2.876923076923077,
+      "grad_norm": 1.6598850043781408,
+      "learning_rate": 8.522447245885356e-08,
+      "loss": 0.9448,
+      "step": 3553
+    },
+    {
+      "epoch": 2.8777327935222674,
+      "grad_norm": 1.657178469902909,
+      "learning_rate": 8.410835599183875e-08,
+      "loss": 1.0182,
+      "step": 3554
+    },
+    {
+      "epoch": 2.8785425101214575,
+      "grad_norm": 1.7487204603764595,
+      "learning_rate": 8.299956535300135e-08,
+      "loss": 0.9883,
+      "step": 3555
+    },
+    {
+      "epoch": 2.8793522267206475,
+      "grad_norm": 1.7130380837811467,
+      "learning_rate": 8.189810136151388e-08,
+      "loss": 0.9205,
+      "step": 3556
+    },
+    {
+      "epoch": 2.880161943319838,
+      "grad_norm": 1.7416692432911214,
+      "learning_rate": 8.08039648311354e-08,
+      "loss": 1.0081,
+      "step": 3557
+    },
+    {
+      "epoch": 2.8809716599190285,
+      "grad_norm": 1.7940488637799301,
+      "learning_rate": 7.971715657020706e-08,
+      "loss": 1.0193,
+      "step": 3558
+    },
+    {
+      "epoch": 2.8817813765182185,
+      "grad_norm": 1.7643531171720135,
+      "learning_rate": 7.863767738166217e-08,
+      "loss": 0.9777,
+      "step": 3559
+    },
+    {
+      "epoch": 2.882591093117409,
+      "grad_norm": 1.7549243652435764,
+      "learning_rate": 7.756552806301498e-08,
+      "loss": 0.9255,
+      "step": 3560
+    },
+    {
+      "epoch": 2.883400809716599,
+      "grad_norm": 1.6037364745875238,
+      "learning_rate": 7.650070940636634e-08,
+      "loss": 1.0126,
+      "step": 3561
+    },
+    {
+      "epoch": 2.8842105263157896,
+      "grad_norm": 1.6938156063789074,
+      "learning_rate": 7.54432221984014e-08,
+      "loss": 1.001,
+      "step": 3562
+    },
+    {
+      "epoch": 2.8850202429149796,
+      "grad_norm": 1.640394353127388,
+      "learning_rate": 7.43930672203863e-08,
+      "loss": 1.0106,
+      "step": 3563
+    },
+    {
+      "epoch": 2.88582995951417,
+      "grad_norm": 1.682827146515058,
+      "learning_rate": 7.335024524817492e-08,
+      "loss": 0.9672,
+      "step": 3564
+    },
+    {
+      "epoch": 2.8866396761133606,
+      "grad_norm": 1.64220479848791,
+      "learning_rate": 7.23147570521987e-08,
+      "loss": 1.0053,
+      "step": 3565
+    },
+    {
+      "epoch": 2.8874493927125506,
+      "grad_norm": 1.6800144926826894,
+      "learning_rate": 7.128660339747239e-08,
+      "loss": 1.0534,
+      "step": 3566
+    },
+    {
+      "epoch": 2.8882591093117407,
+      "grad_norm": 1.6706178385744694,
+      "learning_rate": 7.026578504359394e-08,
+      "loss": 0.9858,
+      "step": 3567
+    },
+    {
+      "epoch": 2.889068825910931,
+      "grad_norm": 1.6725170356062997,
+      "learning_rate": 6.925230274474003e-08,
+      "loss": 0.9447,
+      "step": 3568
+    },
+    {
+      "epoch": 2.8898785425101217,
+      "grad_norm": 1.7484788389494295,
+      "learning_rate": 6.824615724966843e-08,
+      "loss": 0.982,
+      "step": 3569
+    },
+    {
+      "epoch": 2.8906882591093117,
+      "grad_norm": 1.7412794412751307,
+      "learning_rate": 6.724734930171561e-08,
+      "loss": 0.9185,
+      "step": 3570
+    },
+    {
+      "epoch": 2.8914979757085018,
+      "grad_norm": 1.7112532122718904,
+      "learning_rate": 6.625587963879909e-08,
+      "loss": 0.884,
+      "step": 3571
+    },
+    {
+      "epoch": 2.8923076923076922,
+      "grad_norm": 1.669951486754946,
+      "learning_rate": 6.527174899341071e-08,
+      "loss": 1.0064,
+      "step": 3572
+    },
+    {
+      "epoch": 2.8931174089068827,
+      "grad_norm": 1.780604450503182,
+      "learning_rate": 6.429495809262554e-08,
+      "loss": 0.9273,
+      "step": 3573
+    },
+    {
+      "epoch": 2.893927125506073,
+      "grad_norm": 1.6302873462180967,
+      "learning_rate": 6.332550765809075e-08,
+      "loss": 0.9506,
+      "step": 3574
+    },
+    {
+      "epoch": 2.8947368421052633,
+      "grad_norm": 1.7710784892224958,
+      "learning_rate": 6.236339840603677e-08,
+      "loss": 0.9671,
+      "step": 3575
+    },
+    {
+      "epoch": 2.8955465587044533,
+      "grad_norm": 1.7228628615096542,
+      "learning_rate": 6.140863104726391e-08,
+      "loss": 0.9461,
+      "step": 3576
+    },
+    {
+      "epoch": 2.896356275303644,
+      "grad_norm": 1.7105420499553026,
+      "learning_rate": 6.046120628715124e-08,
+      "loss": 0.9828,
+      "step": 3577
+    },
+    {
+      "epoch": 2.897165991902834,
+      "grad_norm": 1.6456141765558494,
+      "learning_rate": 5.952112482565442e-08,
+      "loss": 0.9371,
+      "step": 3578
+    },
+    {
+      "epoch": 2.8979757085020244,
+      "grad_norm": 1.629221588967243,
+      "learning_rate": 5.858838735730232e-08,
+      "loss": 0.9767,
+      "step": 3579
+    },
+    {
+      "epoch": 2.898785425101215,
+      "grad_norm": 1.7148433089723767,
+      "learning_rate": 5.766299457119817e-08,
+      "loss": 0.9423,
+      "step": 3580
+    },
+    {
+      "epoch": 2.899595141700405,
+      "grad_norm": 1.6770711797146378,
+      "learning_rate": 5.674494715101841e-08,
+      "loss": 0.9575,
+      "step": 3581
+    },
+    {
+      "epoch": 2.900404858299595,
+      "grad_norm": 1.7228669637843592,
+      "learning_rate": 5.583424577501273e-08,
+      "loss": 0.982,
+      "step": 3582
+    },
+    {
+      "epoch": 2.9012145748987854,
+      "grad_norm": 1.6410483294411622,
+      "learning_rate": 5.4930891116007355e-08,
+      "loss": 0.9662,
+      "step": 3583
+    },
+    {
+      "epoch": 2.902024291497976,
+      "grad_norm": 1.695947041652728,
+      "learning_rate": 5.40348838413951e-08,
+      "loss": 0.9336,
+      "step": 3584
+    },
+    {
+      "epoch": 2.902834008097166,
+      "grad_norm": 1.7092686814192066,
+      "learning_rate": 5.3146224613144225e-08,
+      "loss": 1.0737,
+      "step": 3585
+    },
+    {
+      "epoch": 2.903643724696356,
+      "grad_norm": 1.7080949259326936,
+      "learning_rate": 5.2264914087792885e-08,
+      "loss": 0.9843,
+      "step": 3586
+    },
+    {
+      "epoch": 2.9044534412955465,
+      "grad_norm": 1.6572515892039745,
+      "learning_rate": 5.139095291645024e-08,
+      "loss": 1.0179,
+      "step": 3587
+    },
+    {
+      "epoch": 2.905263157894737,
+      "grad_norm": 1.6690887807850592,
+      "learning_rate": 5.052434174479759e-08,
+      "loss": 1.0066,
+      "step": 3588
+    },
+    {
+      "epoch": 2.906072874493927,
+      "grad_norm": 1.6532884012275588,
+      "learning_rate": 4.966508121308167e-08,
+      "loss": 0.9931,
+      "step": 3589
+    },
+    {
+      "epoch": 2.9068825910931175,
+      "grad_norm": 1.7133322607154833,
+      "learning_rate": 4.8813171956123565e-08,
+      "loss": 0.9854,
+      "step": 3590
+    },
+    {
+      "epoch": 2.9076923076923076,
+      "grad_norm": 1.6567745485844279,
+      "learning_rate": 4.796861460330982e-08,
+      "loss": 1.0568,
+      "step": 3591
+    },
+    {
+      "epoch": 2.908502024291498,
+      "grad_norm": 1.7343588764303524,
+      "learning_rate": 4.713140977859687e-08,
+      "loss": 0.9529,
+      "step": 3592
+    },
+    {
+      "epoch": 2.909311740890688,
+      "grad_norm": 1.6208499139806953,
+      "learning_rate": 4.630155810050885e-08,
+      "loss": 1.0186,
+      "step": 3593
+    },
+    {
+      "epoch": 2.9101214574898786,
+      "grad_norm": 1.6493435535547711,
+      "learning_rate": 4.547906018213866e-08,
+      "loss": 1.0102,
+      "step": 3594
+    },
+    {
+      "epoch": 2.910931174089069,
+      "grad_norm": 1.6692153459998045,
+      "learning_rate": 4.4663916631143554e-08,
+      "loss": 0.9606,
+      "step": 3595
+    },
+    {
+      "epoch": 2.911740890688259,
+      "grad_norm": 1.6721684961284737,
+      "learning_rate": 4.3856128049749594e-08,
+      "loss": 0.9796,
+      "step": 3596
+    },
+    {
+      "epoch": 2.912550607287449,
+      "grad_norm": 1.7315280769395167,
+      "learning_rate": 4.3055695034747155e-08,
+      "loss": 0.9748,
+      "step": 3597
+    },
+    {
+      "epoch": 2.9133603238866397,
+      "grad_norm": 1.6378741930140095,
+      "learning_rate": 4.226261817749544e-08,
+      "loss": 0.9593,
+      "step": 3598
+    },
+    {
+      "epoch": 2.91417004048583,
+      "grad_norm": 1.7322062631852746,
+      "learning_rate": 4.147689806391575e-08,
+      "loss": 1.0119,
+      "step": 3599
+    },
+    {
+      "epoch": 2.91497975708502,
+      "grad_norm": 1.7138380672008324,
+      "learning_rate": 4.069853527449596e-08,
+      "loss": 0.988,
+      "step": 3600
+    },
+    {
+      "epoch": 2.9157894736842103,
+      "grad_norm": 1.7209134121305623,
+      "learning_rate": 3.9927530384288314e-08,
+      "loss": 0.9688,
+      "step": 3601
+    },
+    {
+      "epoch": 2.9165991902834008,
+      "grad_norm": 1.7412002078631348,
+      "learning_rate": 3.916388396290716e-08,
+      "loss": 0.9903,
+      "step": 3602
+    },
+    {
+      "epoch": 2.9174089068825912,
+      "grad_norm": 1.7072796682597313,
+      "learning_rate": 3.840759657453452e-08,
+      "loss": 0.9591,
+      "step": 3603
+    },
+    {
+      "epoch": 2.9182186234817813,
+      "grad_norm": 1.6648056164804599,
+      "learning_rate": 3.7658668777910135e-08,
+      "loss": 0.9767,
+      "step": 3604
+    },
+    {
+      "epoch": 2.919028340080972,
+      "grad_norm": 1.7237893699815425,
+      "learning_rate": 3.691710112634139e-08,
+      "loss": 1.0162,
+      "step": 3605
+    },
+    {
+      "epoch": 2.919838056680162,
+      "grad_norm": 1.7961044154386043,
+      "learning_rate": 3.61828941676956e-08,
+      "loss": 1.0186,
+      "step": 3606
+    },
+    {
+      "epoch": 2.9206477732793523,
+      "grad_norm": 1.6497363776121279,
+      "learning_rate": 3.54560484444022e-08,
+      "loss": 0.909,
+      "step": 3607
+    },
+    {
+      "epoch": 2.9214574898785424,
+      "grad_norm": 1.747468367170849,
+      "learning_rate": 3.473656449345275e-08,
+      "loss": 0.89,
+      "step": 3608
+    },
+    {
+      "epoch": 2.922267206477733,
+      "grad_norm": 1.762439292957918,
+      "learning_rate": 3.402444284639872e-08,
+      "loss": 0.9111,
+      "step": 3609
+    },
+    {
+      "epoch": 2.9230769230769234,
+      "grad_norm": 1.7149546863394658,
+      "learning_rate": 3.3319684029354815e-08,
+      "loss": 0.9661,
+      "step": 3610
+    },
+    {
+      "epoch": 2.9238866396761134,
+      "grad_norm": 1.6630865806186383,
+      "learning_rate": 3.262228856299343e-08,
+      "loss": 0.9029,
+      "step": 3611
+    },
+    {
+      "epoch": 2.9246963562753034,
+      "grad_norm": 1.7363529145465648,
+      "learning_rate": 3.193225696254798e-08,
+      "loss": 0.9848,
+      "step": 3612
+    },
+    {
+      "epoch": 2.925506072874494,
+      "grad_norm": 1.6797985549185608,
+      "learning_rate": 3.124958973781178e-08,
+      "loss": 1.0399,
+      "step": 3613
+    },
+    {
+      "epoch": 2.9263157894736844,
+      "grad_norm": 1.6469873014958698,
+      "learning_rate": 3.057428739313695e-08,
+      "loss": 0.9589,
+      "step": 3614
+    },
+    {
+      "epoch": 2.9271255060728745,
+      "grad_norm": 1.7250247951219637,
+      "learning_rate": 2.9906350427435505e-08,
+      "loss": 0.937,
+      "step": 3615
+    },
+    {
+      "epoch": 2.9279352226720645,
+      "grad_norm": 1.6796115243111611,
+      "learning_rate": 2.924577933417716e-08,
+      "loss": 0.9864,
+      "step": 3616
+    },
+    {
+      "epoch": 2.928744939271255,
+      "grad_norm": 1.6550805457651048,
+      "learning_rate": 2.85925746013882e-08,
+      "loss": 1.0157,
+      "step": 3617
+    },
+    {
+      "epoch": 2.9295546558704455,
+      "grad_norm": 1.7491759234899906,
+      "learning_rate": 2.7946736711654822e-08,
+      "loss": 1.004,
+      "step": 3618
+    },
+    {
+      "epoch": 2.9303643724696355,
+      "grad_norm": 1.64099300892061,
+      "learning_rate": 2.7308266142119788e-08,
+      "loss": 1.0059,
+      "step": 3619
+    },
+    {
+      "epoch": 2.931174089068826,
+      "grad_norm": 1.6299619725247847,
+      "learning_rate": 2.667716336448356e-08,
+      "loss": 0.9822,
+      "step": 3620
+    },
+    {
+      "epoch": 2.931983805668016,
+      "grad_norm": 1.6619338478776728,
+      "learning_rate": 2.605342884500206e-08,
+      "loss": 0.9879,
+      "step": 3621
+    },
+    {
+      "epoch": 2.9327935222672066,
+      "grad_norm": 1.6869289651046775,
+      "learning_rate": 2.5437063044488895e-08,
+      "loss": 0.9443,
+      "step": 3622
+    },
+    {
+      "epoch": 2.9336032388663966,
+      "grad_norm": 1.650516004800676,
+      "learning_rate": 2.4828066418314256e-08,
+      "loss": 0.9792,
+      "step": 3623
+    },
+    {
+      "epoch": 2.934412955465587,
+      "grad_norm": 1.714896614796574,
+      "learning_rate": 2.422643941640046e-08,
+      "loss": 0.9162,
+      "step": 3624
+    },
+    {
+      "epoch": 2.9352226720647776,
+      "grad_norm": 1.614786417424853,
+      "learning_rate": 2.3632182483228628e-08,
+      "loss": 0.9934,
+      "step": 3625
+    },
+    {
+      "epoch": 2.9360323886639677,
+      "grad_norm": 1.7199405245576622,
+      "learning_rate": 2.3045296057834232e-08,
+      "loss": 0.9467,
+      "step": 3626
+    },
+    {
+      "epoch": 2.9368421052631577,
+      "grad_norm": 1.689626777761019,
+      "learning_rate": 2.2465780573807105e-08,
+      "loss": 0.9896,
+      "step": 3627
+    },
+    {
+      "epoch": 2.937651821862348,
+      "grad_norm": 1.618965258362145,
+      "learning_rate": 2.1893636459289213e-08,
+      "loss": 0.9987,
+      "step": 3628
+    },
+    {
+      "epoch": 2.9384615384615387,
+      "grad_norm": 1.6953456161595155,
+      "learning_rate": 2.132886413698243e-08,
+      "loss": 0.9242,
+      "step": 3629
+    },
+    {
+      "epoch": 2.9392712550607287,
+      "grad_norm": 1.640775504856468,
+      "learning_rate": 2.077146402413521e-08,
+      "loss": 0.9815,
+      "step": 3630
+    },
+    {
+      "epoch": 2.9400809716599188,
+      "grad_norm": 1.630716994540923,
+      "learning_rate": 2.0221436532555928e-08,
+      "loss": 0.9879,
+      "step": 3631
+    },
+    {
+      "epoch": 2.9408906882591093,
+      "grad_norm": 1.6746857067688603,
+      "learning_rate": 1.9678782068600633e-08,
+      "loss": 0.9805,
+      "step": 3632
+    },
+    {
+      "epoch": 2.9417004048582998,
+      "grad_norm": 1.692709178383313,
+      "learning_rate": 1.9143501033181965e-08,
+      "loss": 0.9926,
+      "step": 3633
+    },
+    {
+      "epoch": 2.94251012145749,
+      "grad_norm": 1.722482470681197,
+      "learning_rate": 1.8615593821763587e-08,
+      "loss": 1.0066,
+      "step": 3634
+    },
+    {
+      "epoch": 2.9433198380566803,
+      "grad_norm": 1.6955658283586104,
+      "learning_rate": 1.80950608243613e-08,
+      "loss": 1.0168,
+      "step": 3635
+    },
+    {
+      "epoch": 2.9441295546558703,
+      "grad_norm": 1.6593146288029987,
+      "learning_rate": 1.758190242554303e-08,
+      "loss": 0.9471,
+      "step": 3636
+    },
+    {
+      "epoch": 2.944939271255061,
+      "grad_norm": 1.6437181186269008,
+      "learning_rate": 1.7076119004429958e-08,
+      "loss": 1.0134,
+      "step": 3637
+    },
+    {
+      "epoch": 2.945748987854251,
+      "grad_norm": 1.7191306396154686,
+      "learning_rate": 1.657771093469096e-08,
+      "loss": 0.8814,
+      "step": 3638
+    },
+    {
+      "epoch": 2.9465587044534414,
+      "grad_norm": 1.6761533860830595,
+      "learning_rate": 1.6086678584550374e-08,
+      "loss": 0.9793,
+      "step": 3639
+    },
+    {
+      "epoch": 2.9473684210526314,
+      "grad_norm": 1.704443447654609,
+      "learning_rate": 1.5603022316780235e-08,
+      "loss": 0.9372,
+      "step": 3640
+    },
+    {
+      "epoch": 2.948178137651822,
+      "grad_norm": 1.7083386239114013,
+      "learning_rate": 1.5126742488703604e-08,
+      "loss": 0.9712,
+      "step": 3641
+    },
+    {
+      "epoch": 2.948987854251012,
+      "grad_norm": 1.6835922657350433,
+      "learning_rate": 1.4657839452195677e-08,
+      "loss": 0.9932,
+      "step": 3642
+    },
+    {
+      "epoch": 2.9497975708502024,
+      "grad_norm": 1.7376719503586473,
+      "learning_rate": 1.4196313553680453e-08,
+      "loss": 0.8903,
+      "step": 3643
+    },
+    {
+      "epoch": 2.950607287449393,
+      "grad_norm": 1.7489663276342433,
+      "learning_rate": 1.3742165134130736e-08,
+      "loss": 0.9544,
+      "step": 3644
+    },
+    {
+      "epoch": 2.951417004048583,
+      "grad_norm": 1.6975365462225342,
+      "learning_rate": 1.329539452907036e-08,
+      "loss": 0.9856,
+      "step": 3645
+    },
+    {
+      "epoch": 2.952226720647773,
+      "grad_norm": 1.7039651291858493,
+      "learning_rate": 1.285600206857196e-08,
+      "loss": 0.9986,
+      "step": 3646
+    },
+    {
+      "epoch": 2.9530364372469635,
+      "grad_norm": 1.6540752577120743,
+      "learning_rate": 1.2423988077258087e-08,
+      "loss": 1.0064,
+      "step": 3647
+    },
+    {
+      "epoch": 2.953846153846154,
+      "grad_norm": 1.563653194661081,
+      "learning_rate": 1.1999352874297876e-08,
+      "loss": 1.0462,
+      "step": 3648
+    },
+    {
+      "epoch": 2.954655870445344,
+      "grad_norm": 1.6611119980537208,
+      "learning_rate": 1.1582096773410379e-08,
+      "loss": 1.0157,
+      "step": 3649
+    },
+    {
+      "epoch": 2.9554655870445345,
+      "grad_norm": 1.6209215418673872,
+      "learning_rate": 1.117222008286456e-08,
+      "loss": 1.0152,
+      "step": 3650
+    },
+    {
+      "epoch": 2.9562753036437246,
+      "grad_norm": 1.651092000152933,
+      "learning_rate": 1.0769723105474861e-08,
+      "loss": 0.9698,
+      "step": 3651
+    },
+    {
+      "epoch": 2.957085020242915,
+      "grad_norm": 1.6427078147132284,
+      "learning_rate": 1.0374606138605636e-08,
+      "loss": 0.9562,
+      "step": 3652
+    },
+    {
+      "epoch": 2.957894736842105,
+      "grad_norm": 1.6736772750275095,
+      "learning_rate": 9.986869474166716e-09,
+      "loss": 0.9024,
+      "step": 3653
+    },
+    {
+      "epoch": 2.9587044534412956,
+      "grad_norm": 1.744834463851239,
+      "learning_rate": 9.606513398617846e-09,
+      "loss": 0.8867,
+      "step": 3654
+    },
+    {
+      "epoch": 2.9595141700404857,
+      "grad_norm": 1.58327630324171,
+      "learning_rate": 9.233538192963132e-09,
+      "loss": 1.0089,
+      "step": 3655
+    },
+    {
+      "epoch": 2.960323886639676,
+      "grad_norm": 1.6438964342520213,
+      "learning_rate": 8.867944132757711e-09,
+      "loss": 0.9828,
+      "step": 3656
+    },
+    {
+      "epoch": 2.961133603238866,
+      "grad_norm": 1.7224600116058253,
+      "learning_rate": 8.50973148809997e-09,
+      "loss": 0.9382,
+      "step": 3657
+    },
+    {
+      "epoch": 2.9619433198380567,
+      "grad_norm": 1.6459642766819444,
+      "learning_rate": 8.158900523635993e-09,
+      "loss": 0.9835,
+      "step": 3658
+    },
+    {
+      "epoch": 2.962753036437247,
+      "grad_norm": 1.6777112954645979,
+      "learning_rate": 7.815451498559557e-09,
+      "loss": 0.9928,
+      "step": 3659
+    },
+    {
+      "epoch": 2.9635627530364372,
+      "grad_norm": 1.7128316731395987,
+      "learning_rate": 7.479384666608802e-09,
+      "loss": 0.9999,
+      "step": 3660
+    },
+    {
+      "epoch": 2.9643724696356273,
+      "grad_norm": 1.7331941430919588,
+      "learning_rate": 7.150700276068457e-09,
+      "loss": 0.9244,
+      "step": 3661
+    },
+    {
+      "epoch": 2.9651821862348178,
+      "grad_norm": 1.6425820505456277,
+      "learning_rate": 6.82939856977094e-09,
+      "loss": 1.0678,
+      "step": 3662
+    },
+    {
+      "epoch": 2.9659919028340083,
+      "grad_norm": 1.6923771797910705,
+      "learning_rate": 6.515479785091927e-09,
+      "loss": 0.9698,
+      "step": 3663
+    },
+    {
+      "epoch": 2.9668016194331983,
+      "grad_norm": 1.6765814114028854,
+      "learning_rate": 6.208944153953678e-09,
+      "loss": 0.9733,
+      "step": 3664
+    },
+    {
+      "epoch": 2.967611336032389,
+      "grad_norm": 1.631450848919302,
+      "learning_rate": 5.909791902823925e-09,
+      "loss": 0.9317,
+      "step": 3665
+    },
+    {
+      "epoch": 2.968421052631579,
+      "grad_norm": 1.6990001884177788,
+      "learning_rate": 5.618023252714766e-09,
+      "loss": 0.9603,
+      "step": 3666
+    },
+    {
+      "epoch": 2.9692307692307693,
+      "grad_norm": 1.6821450995668512,
+      "learning_rate": 5.333638419184883e-09,
+      "loss": 0.9634,
+      "step": 3667
+    },
+    {
+      "epoch": 2.9700404858299594,
+      "grad_norm": 1.6746580159744173,
+      "learning_rate": 5.056637612336212e-09,
+      "loss": 0.8863,
+      "step": 3668
+    },
+    {
+      "epoch": 2.97085020242915,
+      "grad_norm": 1.6902742756024816,
+      "learning_rate": 4.787021036816164e-09,
+      "loss": 0.9718,
+      "step": 3669
+    },
+    {
+      "epoch": 2.97165991902834,
+      "grad_norm": 1.7087829075461096,
+      "learning_rate": 4.524788891816512e-09,
+      "loss": 0.9482,
+      "step": 3670
+    },
+    {
+      "epoch": 2.9724696356275304,
+      "grad_norm": 1.7218111890670291,
+      "learning_rate": 4.269941371073394e-09,
+      "loss": 0.9897,
+      "step": 3671
+    },
+    {
+      "epoch": 2.9732793522267205,
+      "grad_norm": 1.6710781319892993,
+      "learning_rate": 4.022478662867313e-09,
+      "loss": 0.9778,
+      "step": 3672
+    },
+    {
+      "epoch": 2.974089068825911,
+      "grad_norm": 1.6227351235688,
+      "learning_rate": 3.782400950023135e-09,
+      "loss": 0.9521,
+      "step": 3673
+    },
+    {
+      "epoch": 2.9748987854251014,
+      "grad_norm": 1.6542823286693975,
+      "learning_rate": 3.5497084099100907e-09,
+      "loss": 0.9691,
+      "step": 3674
+    },
+    {
+      "epoch": 2.9757085020242915,
+      "grad_norm": 1.6462032546016088,
+      "learning_rate": 3.3244012144395545e-09,
+      "loss": 0.9743,
+      "step": 3675
+    },
+    {
+      "epoch": 2.9765182186234815,
+      "grad_norm": 1.6774459355985822,
+      "learning_rate": 3.1064795300683735e-09,
+      "loss": 1.0424,
+      "step": 3676
+    },
+    {
+      "epoch": 2.977327935222672,
+      "grad_norm": 1.6990599030124527,
+      "learning_rate": 2.8959435177955407e-09,
+      "loss": 0.9464,
+      "step": 3677
+    },
+    {
+      "epoch": 2.9781376518218625,
+      "grad_norm": 1.6902993214670898,
+      "learning_rate": 2.692793333165522e-09,
+      "loss": 1.0075,
+      "step": 3678
+    },
+    {
+      "epoch": 2.9789473684210526,
+      "grad_norm": 1.623677495815907,
+      "learning_rate": 2.4970291262649272e-09,
+      "loss": 0.9633,
+      "step": 3679
+    },
+    {
+      "epoch": 2.979757085020243,
+      "grad_norm": 1.7265662360747522,
+      "learning_rate": 2.3086510417225093e-09,
+      "loss": 1.0392,
+      "step": 3680
+    },
+    {
+      "epoch": 2.980566801619433,
+      "grad_norm": 1.6751044099068313,
+      "learning_rate": 2.1276592187124966e-09,
+      "loss": 0.9776,
+      "step": 3681
+    },
+    {
+      "epoch": 2.9813765182186236,
+      "grad_norm": 1.6716882137403286,
+      "learning_rate": 1.95405379095126e-09,
+      "loss": 0.9679,
+      "step": 3682
+    },
+    {
+      "epoch": 2.9821862348178136,
+      "grad_norm": 1.7059793709856448,
+      "learning_rate": 1.7878348866962047e-09,
+      "loss": 0.9525,
+      "step": 3683
+    },
+    {
+      "epoch": 2.982995951417004,
+      "grad_norm": 1.6552987765137663,
+      "learning_rate": 1.6290026287513194e-09,
+      "loss": 1.0131,
+      "step": 3684
+    },
+    {
+      "epoch": 2.983805668016194,
+      "grad_norm": 1.6707647453986292,
+      "learning_rate": 1.4775571344605167e-09,
+      "loss": 0.9977,
+      "step": 3685
+    },
+    {
+      "epoch": 2.9846153846153847,
+      "grad_norm": 1.6942525860442892,
+      "learning_rate": 1.3334985157109625e-09,
+      "loss": 0.9585,
+      "step": 3686
+    },
+    {
+      "epoch": 2.9854251012145747,
+      "grad_norm": 1.6490557158214814,
+      "learning_rate": 1.1968268789330773e-09,
+      "loss": 0.9482,
+      "step": 3687
+    },
+    {
+      "epoch": 2.986234817813765,
+      "grad_norm": 1.6249913718833546,
+      "learning_rate": 1.0675423250994244e-09,
+      "loss": 0.9683,
+      "step": 3688
+    },
+    {
+      "epoch": 2.9870445344129557,
+      "grad_norm": 1.6778588925327733,
+      "learning_rate": 9.456449497247112e-10,
+      "loss": 0.9532,
+      "step": 3689
+    },
+    {
+      "epoch": 2.9878542510121457,
+      "grad_norm": 1.6962491724057556,
+      "learning_rate": 8.311348428657884e-10,
+      "loss": 0.9811,
+      "step": 3690
+    },
+    {
+      "epoch": 2.988663967611336,
+      "grad_norm": 1.6716685263614628,
+      "learning_rate": 7.240120891238711e-10,
+      "loss": 0.9267,
+      "step": 3691
+    },
+    {
+      "epoch": 2.9894736842105263,
+      "grad_norm": 1.7528045485583463,
+      "learning_rate": 6.242767676400974e-10,
+      "loss": 0.9343,
+      "step": 3692
+    },
+    {
+      "epoch": 2.9902834008097168,
+      "grad_norm": 1.7066732550694934,
+      "learning_rate": 5.31928952098859e-10,
+      "loss": 0.956,
+      "step": 3693
+    },
+    {
+      "epoch": 2.991093117408907,
+      "grad_norm": 1.6307704468251862,
+      "learning_rate": 4.469687107255816e-10,
+      "loss": 0.984,
+      "step": 3694
+    },
+    {
+      "epoch": 2.9919028340080973,
+      "grad_norm": 1.7099555082281752,
+      "learning_rate": 3.6939610628894396e-10,
+      "loss": 0.9458,
+      "step": 3695
+    },
+    {
+      "epoch": 2.9927125506072874,
+      "grad_norm": 1.6724491743800372,
+      "learning_rate": 2.9921119609976903e-10,
+      "loss": 0.958,
+      "step": 3696
+    },
+    {
+      "epoch": 2.993522267206478,
+      "grad_norm": 1.7205791609019359,
+      "learning_rate": 2.364140320110231e-10,
+      "loss": 0.8848,
+      "step": 3697
+    },
+    {
+      "epoch": 2.994331983805668,
+      "grad_norm": 1.7073529607261773,
+      "learning_rate": 1.8100466041559573e-10,
+      "loss": 0.9302,
+      "step": 3698
+    },
+    {
+      "epoch": 2.9951417004048584,
+      "grad_norm": 1.6422918196631109,
+      "learning_rate": 1.3298312225074051e-10,
+      "loss": 0.9791,
+      "step": 3699
+    },
+    {
+      "epoch": 2.9959514170040484,
+      "grad_norm": 1.7056433896621659,
+      "learning_rate": 9.234945299363418e-11,
+      "loss": 0.9875,
+      "step": 3700
+    },
+    {
+      "epoch": 2.996761133603239,
+      "grad_norm": 1.6180923945102257,
+      "learning_rate": 5.910368266470734e-11,
+      "loss": 0.9499,
+      "step": 3701
+    },
+    {
+      "epoch": 2.997570850202429,
+      "grad_norm": 1.6943866994325132,
+      "learning_rate": 3.324583582653418e-11,
+      "loss": 1.0206,
+      "step": 3702
+    },
+    {
+      "epoch": 2.9983805668016195,
+      "grad_norm": 1.5972662423702781,
+      "learning_rate": 1.477593158272228e-11,
+      "loss": 1.0339,
+      "step": 3703
+    },
+    {
+      "epoch": 2.99919028340081,
+      "grad_norm": 1.6264801708756609,
+      "learning_rate": 3.693983577912619e-12,
+      "loss": 0.9317,
+      "step": 3704
+    },
+    {
+      "epoch": 3.0,
+      "grad_norm": 1.6339826687405365,
+      "learning_rate": 0.0,
+      "loss": 0.9082,
+      "step": 3705
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 3705,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 3,
+  "save_steps": 618,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 4.96481039548416e+16,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}