diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,20860 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 1.0,
+  "eval_steps": 500,
+  "global_step": 2974,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0,
+      "grad_norm": 9.113551904483701,
+      "learning_rate": 1.1111111111111112e-07,
+      "loss": 1.7379,
+      "step": 1
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.666641640137465,
+      "learning_rate": 2.2222222222222224e-07,
+      "loss": 1.5163,
+      "step": 2
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 9.285391292243393,
+      "learning_rate": 3.3333333333333335e-07,
+      "loss": 1.7072,
+      "step": 3
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.503071677683805,
+      "learning_rate": 4.444444444444445e-07,
+      "loss": 1.3861,
+      "step": 4
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.789306421021227,
+      "learning_rate": 5.555555555555555e-07,
+      "loss": 1.5914,
+      "step": 5
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.217252543871634,
+      "learning_rate": 6.666666666666667e-07,
+      "loss": 1.6276,
+      "step": 6
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.784070522914973,
+      "learning_rate": 7.777777777777779e-07,
+      "loss": 1.5337,
+      "step": 7
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.444320674640562,
+      "learning_rate": 8.88888888888889e-07,
+      "loss": 1.464,
+      "step": 8
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 8.72678728491744,
+      "learning_rate": 1.0000000000000002e-06,
+      "loss": 1.588,
+      "step": 9
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 7.776067172623341,
+      "learning_rate": 1.111111111111111e-06,
+      "loss": 1.6185,
+      "step": 10
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.99786134094402,
+      "learning_rate": 1.2222222222222223e-06,
+      "loss": 1.3517,
+      "step": 11
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 5.346807250807995,
+      "learning_rate": 1.3333333333333334e-06,
+      "loss": 1.3359,
+      "step": 12
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 6.74017885339056,
+      "learning_rate": 1.4444444444444445e-06,
+      "loss": 1.6038,
+      "step": 13
+    },
+    {
+      "epoch": 0.0,
+      "grad_norm": 4.492812874131746,
+      "learning_rate": 1.5555555555555558e-06,
+      "loss": 1.4196,
+      "step": 14
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.7044093902138435,
+      "learning_rate": 1.6666666666666667e-06,
+      "loss": 1.438,
+      "step": 15
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 4.724535172792871,
+      "learning_rate": 1.777777777777778e-06,
+      "loss": 1.4499,
+      "step": 16
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 3.5606546819174683,
+      "learning_rate": 1.888888888888889e-06,
+      "loss": 1.3924,
+      "step": 17
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 2.85723565391118,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 1.391,
+      "step": 18
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5875373341251489,
+      "learning_rate": 2.1111111111111114e-06,
+      "loss": 1.1959,
+      "step": 19
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.369388948458357,
+      "learning_rate": 2.222222222222222e-06,
+      "loss": 1.2046,
+      "step": 20
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.2029638585524494,
+      "learning_rate": 2.3333333333333336e-06,
+      "loss": 1.287,
+      "step": 21
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.25437144945714,
+      "learning_rate": 2.4444444444444447e-06,
+      "loss": 1.0921,
+      "step": 22
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4364661380803248,
+      "learning_rate": 2.5555555555555557e-06,
+      "loss": 1.303,
+      "step": 23
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.2108231155796993,
+      "learning_rate": 2.666666666666667e-06,
+      "loss": 1.2288,
+      "step": 24
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5149902856324549,
+      "learning_rate": 2.7777777777777783e-06,
+      "loss": 1.3917,
+      "step": 25
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.052353129355817,
+      "learning_rate": 2.888888888888889e-06,
+      "loss": 1.1873,
+      "step": 26
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5439273334633483,
+      "learning_rate": 3e-06,
+      "loss": 1.3348,
+      "step": 27
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.8704969628863402,
+      "learning_rate": 3.1111111111111116e-06,
+      "loss": 1.2203,
+      "step": 28
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.5917427545842355,
+      "learning_rate": 3.2222222222222227e-06,
+      "loss": 1.2433,
+      "step": 29
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7857106733969363,
+      "learning_rate": 3.3333333333333333e-06,
+      "loss": 1.1287,
+      "step": 30
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.81561386237172,
+      "learning_rate": 3.444444444444445e-06,
+      "loss": 1.17,
+      "step": 31
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.7300899606590168,
+      "learning_rate": 3.555555555555556e-06,
+      "loss": 1.2502,
+      "step": 32
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.4710539247056669,
+      "learning_rate": 3.6666666666666666e-06,
+      "loss": 1.0919,
+      "step": 33
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0878691460684309,
+      "learning_rate": 3.777777777777778e-06,
+      "loss": 1.339,
+      "step": 34
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.1952000724786025,
+      "learning_rate": 3.88888888888889e-06,
+      "loss": 1.0695,
+      "step": 35
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.0124415853522692,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 1.0406,
+      "step": 36
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9023595709906321,
+      "learning_rate": 4.111111111111111e-06,
+      "loss": 1.3004,
+      "step": 37
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.3920082618633798,
+      "learning_rate": 4.222222222222223e-06,
+      "loss": 0.9488,
+      "step": 38
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8711560431335008,
+      "learning_rate": 4.333333333333334e-06,
+      "loss": 1.2819,
+      "step": 39
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 1.2634072392645208,
+      "learning_rate": 4.444444444444444e-06,
+      "loss": 1.2314,
+      "step": 40
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.8206010270905869,
+      "learning_rate": 4.555555555555556e-06,
+      "loss": 1.1718,
+      "step": 41
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.762796445859709,
+      "learning_rate": 4.666666666666667e-06,
+      "loss": 1.2114,
+      "step": 42
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.7840882720584422,
+      "learning_rate": 4.777777777777778e-06,
+      "loss": 1.0674,
+      "step": 43
+    },
+    {
+      "epoch": 0.01,
+      "grad_norm": 0.9500797358743636,
+      "learning_rate": 4.888888888888889e-06,
+      "loss": 1.4348,
+      "step": 44
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9332005451631037,
+      "learning_rate": 5e-06,
+      "loss": 1.1185,
+      "step": 45
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7135349849193702,
+      "learning_rate": 5.1111111111111115e-06,
+      "loss": 1.1611,
+      "step": 46
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.0059610528798153,
+      "learning_rate": 5.2222222222222226e-06,
+      "loss": 1.1526,
+      "step": 47
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.3109916549414267,
+      "learning_rate": 5.333333333333334e-06,
+      "loss": 0.8651,
+      "step": 48
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8647854394148043,
+      "learning_rate": 5.444444444444445e-06,
+      "loss": 1.304,
+      "step": 49
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8478898741206349,
+      "learning_rate": 5.555555555555557e-06,
+      "loss": 1.0122,
+      "step": 50
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7924732869600487,
+      "learning_rate": 5.666666666666667e-06,
+      "loss": 1.247,
+      "step": 51
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.9915722230625104,
+      "learning_rate": 5.777777777777778e-06,
+      "loss": 1.2123,
+      "step": 52
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.1196528881796783,
+      "learning_rate": 5.88888888888889e-06,
+      "loss": 1.1736,
+      "step": 53
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7213922987957131,
+      "learning_rate": 6e-06,
+      "loss": 1.1073,
+      "step": 54
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7144716559787306,
+      "learning_rate": 6.111111111111112e-06,
+      "loss": 1.0763,
+      "step": 55
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7399560428741311,
+      "learning_rate": 6.222222222222223e-06,
+      "loss": 1.1409,
+      "step": 56
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.9177012804944056,
+      "learning_rate": 6.333333333333333e-06,
+      "loss": 0.9933,
+      "step": 57
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6258101423489458,
+      "learning_rate": 6.444444444444445e-06,
+      "loss": 1.308,
+      "step": 58
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.774905437562704,
+      "learning_rate": 6.555555555555556e-06,
+      "loss": 0.9512,
+      "step": 59
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6604359493411501,
+      "learning_rate": 6.666666666666667e-06,
+      "loss": 1.2613,
+      "step": 60
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6797518417321902,
+      "learning_rate": 6.777777777777779e-06,
+      "loss": 0.9448,
+      "step": 61
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.692670578876648,
+      "learning_rate": 6.88888888888889e-06,
+      "loss": 0.9521,
+      "step": 62
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6736929700592789,
+      "learning_rate": 7e-06,
+      "loss": 1.2557,
+      "step": 63
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.760750336196496,
+      "learning_rate": 7.111111111111112e-06,
+      "loss": 1.1524,
+      "step": 64
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 1.95208474356618,
+      "learning_rate": 7.222222222222223e-06,
+      "loss": 0.9749,
+      "step": 65
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6330691780212638,
+      "learning_rate": 7.333333333333333e-06,
+      "loss": 1.0214,
+      "step": 66
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7027721615587439,
+      "learning_rate": 7.444444444444445e-06,
+      "loss": 0.9667,
+      "step": 67
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6112230323368782,
+      "learning_rate": 7.555555555555556e-06,
+      "loss": 1.1759,
+      "step": 68
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7114571836326415,
+      "learning_rate": 7.666666666666667e-06,
+      "loss": 1.0877,
+      "step": 69
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6367657076980286,
+      "learning_rate": 7.77777777777778e-06,
+      "loss": 1.1572,
+      "step": 70
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.7169446592018366,
+      "learning_rate": 7.88888888888889e-06,
+      "loss": 1.0678,
+      "step": 71
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.6153872376312953,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 1.1506,
+      "step": 72
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 0.8051829594292012,
+      "learning_rate": 8.111111111111112e-06,
+      "loss": 1.038,
+      "step": 73
+    },
+    {
+      "epoch": 0.02,
+      "grad_norm": 8.421602678146645,
+      "learning_rate": 8.222222222222222e-06,
+      "loss": 1.0459,
+      "step": 74
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6568728706881283,
+      "learning_rate": 8.333333333333334e-06,
+      "loss": 1.0514,
+      "step": 75
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6933390745206598,
+      "learning_rate": 8.444444444444446e-06,
+      "loss": 1.0316,
+      "step": 76
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6319894221508225,
+      "learning_rate": 8.555555555555556e-06,
+      "loss": 1.0707,
+      "step": 77
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6484460927430371,
+      "learning_rate": 8.666666666666668e-06,
+      "loss": 0.8955,
+      "step": 78
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6412412597686028,
+      "learning_rate": 8.777777777777778e-06,
+      "loss": 1.0412,
+      "step": 79
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6249848568842076,
+      "learning_rate": 8.888888888888888e-06,
+      "loss": 1.1485,
+      "step": 80
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5779448469526501,
+      "learning_rate": 9e-06,
+      "loss": 1.1263,
+      "step": 81
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7928794387514632,
+      "learning_rate": 9.111111111111112e-06,
+      "loss": 1.0573,
+      "step": 82
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 11.090589133661634,
+      "learning_rate": 9.222222222222224e-06,
+      "loss": 1.1394,
+      "step": 83
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7656421345857701,
+      "learning_rate": 9.333333333333334e-06,
+      "loss": 0.9838,
+      "step": 84
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6349694970578184,
+      "learning_rate": 9.444444444444445e-06,
+      "loss": 1.1032,
+      "step": 85
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7495166582854168,
+      "learning_rate": 9.555555555555556e-06,
+      "loss": 1.0593,
+      "step": 86
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.624738091847216,
+      "learning_rate": 9.666666666666667e-06,
+      "loss": 1.1374,
+      "step": 87
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6467217611363693,
+      "learning_rate": 9.777777777777779e-06,
+      "loss": 0.9366,
+      "step": 88
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6830216508444934,
+      "learning_rate": 9.88888888888889e-06,
+      "loss": 1.1443,
+      "step": 89
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6096308883326953,
+      "learning_rate": 1e-05,
+      "loss": 0.9363,
+      "step": 90
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 5.03815160174929,
+      "learning_rate": 9.999997033466898e-06,
+      "loss": 0.9331,
+      "step": 91
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6157052909342736,
+      "learning_rate": 9.999988133871106e-06,
+      "loss": 1.1873,
+      "step": 92
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6296814693811345,
+      "learning_rate": 9.999973301223189e-06,
+      "loss": 0.9486,
+      "step": 93
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7307285772972432,
+      "learning_rate": 9.999952535540743e-06,
+      "loss": 1.1042,
+      "step": 94
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5323668548074265,
+      "learning_rate": 9.999925836848413e-06,
+      "loss": 1.0378,
+      "step": 95
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6945477093732292,
+      "learning_rate": 9.99989320517788e-06,
+      "loss": 1.1623,
+      "step": 96
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6433895620887395,
+      "learning_rate": 9.999854640567861e-06,
+      "loss": 1.1039,
+      "step": 97
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5842976951577084,
+      "learning_rate": 9.999810143064122e-06,
+      "loss": 0.9365,
+      "step": 98
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.642410912594165,
+      "learning_rate": 9.999759712719461e-06,
+      "loss": 1.1889,
+      "step": 99
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.6834415034045189,
+      "learning_rate": 9.999703349593721e-06,
+      "loss": 0.9125,
+      "step": 100
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 3.499954753879034,
+      "learning_rate": 9.999641053753784e-06,
+      "loss": 1.0228,
+      "step": 101
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.8239984118546664,
+      "learning_rate": 9.999572825273569e-06,
+      "loss": 1.0466,
+      "step": 102
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.5965062178022119,
+      "learning_rate": 9.999498664234038e-06,
+      "loss": 1.0925,
+      "step": 103
+    },
+    {
+      "epoch": 0.03,
+      "grad_norm": 0.7595482038631068,
+      "learning_rate": 9.999418570723189e-06,
+      "loss": 1.1829,
+      "step": 104
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5370583886870598,
+      "learning_rate": 9.999332544836066e-06,
+      "loss": 0.9787,
+      "step": 105
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5265866797541217,
+      "learning_rate": 9.999240586674749e-06,
+      "loss": 1.1078,
+      "step": 106
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6417343161827469,
+      "learning_rate": 9.999142696348351e-06,
+      "loss": 0.8688,
+      "step": 107
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7485007581766425,
+      "learning_rate": 9.999038873973034e-06,
+      "loss": 1.1221,
+      "step": 108
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9701130652280301,
+      "learning_rate": 9.998929119671996e-06,
+      "loss": 1.0153,
+      "step": 109
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.445211686406718,
+      "learning_rate": 9.99881343357547e-06,
+      "loss": 0.8183,
+      "step": 110
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7920490401420517,
+      "learning_rate": 9.998691815820732e-06,
+      "loss": 1.1654,
+      "step": 111
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.100134095404715,
+      "learning_rate": 9.998564266552096e-06,
+      "loss": 0.9582,
+      "step": 112
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5769508495123952,
+      "learning_rate": 9.998430785920912e-06,
+      "loss": 1.1256,
+      "step": 113
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6689066897708291,
+      "learning_rate": 9.99829137408557e-06,
+      "loss": 0.9715,
+      "step": 114
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.7485365617328634,
+      "learning_rate": 9.9981460312115e-06,
+      "loss": 1.046,
+      "step": 115
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.692302941222489,
+      "learning_rate": 9.997994757471165e-06,
+      "loss": 0.8989,
+      "step": 116
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6204328292162147,
+      "learning_rate": 9.997837553044072e-06,
+      "loss": 0.8562,
+      "step": 117
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.099975666530968,
+      "learning_rate": 9.997674418116759e-06,
+      "loss": 0.956,
+      "step": 118
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.4841383943094081,
+      "learning_rate": 9.997505352882803e-06,
+      "loss": 0.9996,
+      "step": 119
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5784174104321163,
+      "learning_rate": 9.997330357542823e-06,
+      "loss": 1.0922,
+      "step": 120
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.9529906523568871,
+      "learning_rate": 9.997149432304468e-06,
+      "loss": 1.006,
+      "step": 121
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5456364543354274,
+      "learning_rate": 9.996962577382428e-06,
+      "loss": 1.0299,
+      "step": 122
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6103682494289949,
+      "learning_rate": 9.996769792998424e-06,
+      "loss": 0.9805,
+      "step": 123
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5956823055219069,
+      "learning_rate": 9.996571079381222e-06,
+      "loss": 0.9807,
+      "step": 124
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5868444289881356,
+      "learning_rate": 9.996366436766612e-06,
+      "loss": 1.1717,
+      "step": 125
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 1.647808421916753,
+      "learning_rate": 9.99615586539743e-06,
+      "loss": 0.9551,
+      "step": 126
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8549885276167155,
+      "learning_rate": 9.995939365523542e-06,
+      "loss": 0.7909,
+      "step": 127
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6736538373369514,
+      "learning_rate": 9.99571693740185e-06,
+      "loss": 0.9968,
+      "step": 128
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5323860910681362,
+      "learning_rate": 9.995488581296287e-06,
+      "loss": 0.9381,
+      "step": 129
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.5799258734927842,
+      "learning_rate": 9.995254297477827e-06,
+      "loss": 1.052,
+      "step": 130
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.40555266054332817,
+      "learning_rate": 9.995014086224473e-06,
+      "loss": 1.0247,
+      "step": 131
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.8313722623249578,
+      "learning_rate": 9.994767947821261e-06,
+      "loss": 1.114,
+      "step": 132
+    },
+    {
+      "epoch": 0.04,
+      "grad_norm": 0.6159881086165419,
+      "learning_rate": 9.994515882560267e-06,
+      "loss": 1.0206,
+      "step": 133
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5765420827669706,
+      "learning_rate": 9.99425789074059e-06,
+      "loss": 1.0255,
+      "step": 134
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6072108148377163,
+      "learning_rate": 9.99399397266837e-06,
+      "loss": 0.9779,
+      "step": 135
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5575071285419874,
+      "learning_rate": 9.993724128656772e-06,
+      "loss": 0.7238,
+      "step": 136
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5484619900480536,
+      "learning_rate": 9.993448359026e-06,
+      "loss": 1.0957,
+      "step": 137
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6476859960298905,
+      "learning_rate": 9.993166664103283e-06,
+      "loss": 0.9129,
+      "step": 138
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5584717784698555,
+      "learning_rate": 9.992879044222887e-06,
+      "loss": 1.0078,
+      "step": 139
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8891251962670573,
+      "learning_rate": 9.992585499726104e-06,
+      "loss": 0.9896,
+      "step": 140
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.874523517171636,
+      "learning_rate": 9.992286030961257e-06,
+      "loss": 0.8692,
+      "step": 141
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6804767621255267,
+      "learning_rate": 9.9919806382837e-06,
+      "loss": 1.1029,
+      "step": 142
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.47872384695467546,
+      "learning_rate": 9.991669322055818e-06,
+      "loss": 1.0301,
+      "step": 143
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5706255497301249,
+      "learning_rate": 9.99135208264702e-06,
+      "loss": 1.1003,
+      "step": 144
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.584096053153678,
+      "learning_rate": 9.991028920433747e-06,
+      "loss": 0.7893,
+      "step": 145
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5686176684625571,
+      "learning_rate": 9.99069983579947e-06,
+      "loss": 0.9837,
+      "step": 146
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6766074180282442,
+      "learning_rate": 9.990364829134682e-06,
+      "loss": 1.0003,
+      "step": 147
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 1.5717121842062827,
+      "learning_rate": 9.990023900836909e-06,
+      "loss": 0.9145,
+      "step": 148
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7575418977089604,
+      "learning_rate": 9.9896770513107e-06,
+      "loss": 0.9694,
+      "step": 149
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5931531244198177,
+      "learning_rate": 9.98932428096763e-06,
+      "loss": 0.9884,
+      "step": 150
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.5639076282633193,
+      "learning_rate": 9.988965590226303e-06,
+      "loss": 1.0644,
+      "step": 151
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6369351706982362,
+      "learning_rate": 9.988600979512345e-06,
+      "loss": 0.9616,
+      "step": 152
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7739799531591574,
+      "learning_rate": 9.988230449258409e-06,
+      "loss": 1.0453,
+      "step": 153
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7880235118597533,
+      "learning_rate": 9.987853999904169e-06,
+      "loss": 0.7048,
+      "step": 154
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.8614599286997509,
+      "learning_rate": 9.987471631896326e-06,
+      "loss": 1.0319,
+      "step": 155
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.437366450131843,
+      "learning_rate": 9.987083345688604e-06,
+      "loss": 1.0908,
+      "step": 156
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6169345528702436,
+      "learning_rate": 9.986689141741748e-06,
+      "loss": 0.9286,
+      "step": 157
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.610257586302272,
+      "learning_rate": 9.986289020523525e-06,
+      "loss": 1.0694,
+      "step": 158
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.6073126333489027,
+      "learning_rate": 9.985882982508723e-06,
+      "loss": 0.9538,
+      "step": 159
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7612774737827293,
+      "learning_rate": 9.985471028179155e-06,
+      "loss": 0.9208,
+      "step": 160
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.630742110233408,
+      "learning_rate": 9.985053158023647e-06,
+      "loss": 1.0563,
+      "step": 161
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.7204672321591241,
+      "learning_rate": 9.984629372538055e-06,
+      "loss": 0.8701,
+      "step": 162
+    },
+    {
+      "epoch": 0.05,
+      "grad_norm": 0.9447008580031212,
+      "learning_rate": 9.984199672225246e-06,
+      "loss": 0.8548,
+      "step": 163
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8357579937088263,
+      "learning_rate": 9.983764057595105e-06,
+      "loss": 0.8893,
+      "step": 164
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7989934702212893,
+      "learning_rate": 9.983322529164543e-06,
+      "loss": 0.9271,
+      "step": 165
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5802826976856544,
+      "learning_rate": 9.98287508745748e-06,
+      "loss": 1.0502,
+      "step": 166
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6403437591683995,
+      "learning_rate": 9.982421733004857e-06,
+      "loss": 1.0072,
+      "step": 167
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5708405057411571,
+      "learning_rate": 9.981962466344633e-06,
+      "loss": 1.1796,
+      "step": 168
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.4204252182417166,
+      "learning_rate": 9.981497288021778e-06,
+      "loss": 0.9959,
+      "step": 169
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6953918741411399,
+      "learning_rate": 9.981026198588276e-06,
+      "loss": 1.0982,
+      "step": 170
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6168950518222326,
+      "learning_rate": 9.980549198603132e-06,
+      "loss": 0.6971,
+      "step": 171
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.634482656988875,
+      "learning_rate": 9.98006628863236e-06,
+      "loss": 0.9567,
+      "step": 172
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5543531233947134,
+      "learning_rate": 9.979577469248984e-06,
+      "loss": 0.8863,
+      "step": 173
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6960202180885926,
+      "learning_rate": 9.979082741033047e-06,
+      "loss": 0.908,
+      "step": 174
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7184794126810341,
+      "learning_rate": 9.9785821045716e-06,
+      "loss": 1.1404,
+      "step": 175
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6302022139154456,
+      "learning_rate": 9.978075560458703e-06,
+      "loss": 0.8822,
+      "step": 176
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.8606038251163217,
+      "learning_rate": 9.977563109295429e-06,
+      "loss": 1.0643,
+      "step": 177
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6476131274883036,
+      "learning_rate": 9.977044751689858e-06,
+      "loss": 0.8189,
+      "step": 178
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6285108549011541,
+      "learning_rate": 9.976520488257082e-06,
+      "loss": 1.1198,
+      "step": 179
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5591779433195079,
+      "learning_rate": 9.975990319619199e-06,
+      "loss": 0.817,
+      "step": 180
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6442548490291593,
+      "learning_rate": 9.975454246405312e-06,
+      "loss": 0.8783,
+      "step": 181
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6141326200646599,
+      "learning_rate": 9.974912269251534e-06,
+      "loss": 1.1487,
+      "step": 182
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.47077141150771157,
+      "learning_rate": 9.974364388800983e-06,
+      "loss": 1.0232,
+      "step": 183
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6323904771012486,
+      "learning_rate": 9.97381060570378e-06,
+      "loss": 0.9804,
+      "step": 184
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.82445384759489,
+      "learning_rate": 9.97325092061705e-06,
+      "loss": 0.8852,
+      "step": 185
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 1.2743959744980293,
+      "learning_rate": 9.972685334204926e-06,
+      "loss": 0.8875,
+      "step": 186
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5658881360125698,
+      "learning_rate": 9.972113847138538e-06,
+      "loss": 1.0069,
+      "step": 187
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7795452924194631,
+      "learning_rate": 9.971536460096021e-06,
+      "loss": 0.9696,
+      "step": 188
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.7859839289734134,
+      "learning_rate": 9.97095317376251e-06,
+      "loss": 1.1285,
+      "step": 189
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5845612785208404,
+      "learning_rate": 9.970363988830138e-06,
+      "loss": 0.6905,
+      "step": 190
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6357386327190812,
+      "learning_rate": 9.969768905998045e-06,
+      "loss": 1.0197,
+      "step": 191
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.6241579399228288,
+      "learning_rate": 9.96916792597236e-06,
+      "loss": 1.1463,
+      "step": 192
+    },
+    {
+      "epoch": 0.06,
+      "grad_norm": 0.5891823096457172,
+      "learning_rate": 9.968561049466214e-06,
+      "loss": 0.885,
+      "step": 193
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.46321994895259255,
+      "learning_rate": 9.967948277199736e-06,
+      "loss": 1.0243,
+      "step": 194
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7126845167755501,
+      "learning_rate": 9.96732960990005e-06,
+      "loss": 0.9449,
+      "step": 195
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6705124362756845,
+      "learning_rate": 9.966705048301272e-06,
+      "loss": 1.0781,
+      "step": 196
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8260634886292914,
+      "learning_rate": 9.96607459314452e-06,
+      "loss": 0.8725,
+      "step": 197
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.8141165464738289,
+      "learning_rate": 9.965438245177895e-06,
+      "loss": 0.6877,
+      "step": 198
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6840059728119295,
+      "learning_rate": 9.964796005156498e-06,
+      "loss": 1.0856,
+      "step": 199
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6901750536413432,
+      "learning_rate": 9.964147873842423e-06,
+      "loss": 0.9326,
+      "step": 200
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5686332076647092,
+      "learning_rate": 9.963493852004746e-06,
+      "loss": 1.1657,
+      "step": 201
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.64072838732007,
+      "learning_rate": 9.96283394041954e-06,
+      "loss": 0.9592,
+      "step": 202
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5334824795620187,
+      "learning_rate": 9.962168139869866e-06,
+      "loss": 1.014,
+      "step": 203
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6276182567877997,
+      "learning_rate": 9.96149645114577e-06,
+      "loss": 1.0496,
+      "step": 204
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5075251883061104,
+      "learning_rate": 9.960818875044288e-06,
+      "loss": 0.9987,
+      "step": 205
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7058557513797454,
+      "learning_rate": 9.960135412369441e-06,
+      "loss": 0.7625,
+      "step": 206
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6504983850136618,
+      "learning_rate": 9.959446063932232e-06,
+      "loss": 0.9193,
+      "step": 207
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.586393913452422,
+      "learning_rate": 9.958750830550654e-06,
+      "loss": 1.0117,
+      "step": 208
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6426568788931324,
+      "learning_rate": 9.95804971304968e-06,
+      "loss": 0.838,
+      "step": 209
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 5.48423034879515,
+      "learning_rate": 9.957342712261263e-06,
+      "loss": 0.9163,
+      "step": 210
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7636535530469188,
+      "learning_rate": 9.956629829024342e-06,
+      "loss": 1.0426,
+      "step": 211
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6177587768244475,
+      "learning_rate": 9.955911064184834e-06,
+      "loss": 1.0468,
+      "step": 212
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5262523822758027,
+      "learning_rate": 9.95518641859563e-06,
+      "loss": 0.9376,
+      "step": 213
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5208897427506387,
+      "learning_rate": 9.95445589311661e-06,
+      "loss": 0.9548,
+      "step": 214
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7155103485896604,
+      "learning_rate": 9.953719488614625e-06,
+      "loss": 1.0825,
+      "step": 215
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4455122797905697,
+      "learning_rate": 9.952977205963496e-06,
+      "loss": 0.7011,
+      "step": 216
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5848429203378096,
+      "learning_rate": 9.95222904604403e-06,
+      "loss": 0.8599,
+      "step": 217
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.4041169771051739,
+      "learning_rate": 9.951475009744006e-06,
+      "loss": 1.0574,
+      "step": 218
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.7614026229997751,
+      "learning_rate": 9.950715097958168e-06,
+      "loss": 0.8336,
+      "step": 219
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5654051082417987,
+      "learning_rate": 9.94994931158824e-06,
+      "loss": 1.094,
+      "step": 220
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.6486590997163361,
+      "learning_rate": 9.949177651542917e-06,
+      "loss": 0.8587,
+      "step": 221
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.598608123431916,
+      "learning_rate": 9.948400118737856e-06,
+      "loss": 0.8353,
+      "step": 222
+    },
+    {
+      "epoch": 0.07,
+      "grad_norm": 0.5631267008314914,
+      "learning_rate": 9.94761671409569e-06,
+      "loss": 1.0439,
+      "step": 223
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4729999948514551,
+      "learning_rate": 9.946827438546016e-06,
+      "loss": 0.6776,
+      "step": 224
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5534236056119295,
+      "learning_rate": 9.946032293025402e-06,
+      "loss": 1.0315,
+      "step": 225
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.746002862658215,
+      "learning_rate": 9.945231278477374e-06,
+      "loss": 1.0057,
+      "step": 226
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.7302208874584205,
+      "learning_rate": 9.944424395852432e-06,
+      "loss": 1.068,
+      "step": 227
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6096260752322689,
+      "learning_rate": 9.943611646108029e-06,
+      "loss": 0.8818,
+      "step": 228
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6756327343367196,
+      "learning_rate": 9.942793030208585e-06,
+      "loss": 0.9147,
+      "step": 229
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5895164295946276,
+      "learning_rate": 9.941968549125481e-06,
+      "loss": 1.0979,
+      "step": 230
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4330388884712142,
+      "learning_rate": 9.94113820383706e-06,
+      "loss": 0.9485,
+      "step": 231
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5491403301333749,
+      "learning_rate": 9.940301995328616e-06,
+      "loss": 1.0452,
+      "step": 232
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6194374062445085,
+      "learning_rate": 9.939459924592407e-06,
+      "loss": 0.8666,
+      "step": 233
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.4425573243315801,
+      "learning_rate": 9.938611992627647e-06,
+      "loss": 0.7291,
+      "step": 234
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6643984264963485,
+      "learning_rate": 9.937758200440501e-06,
+      "loss": 1.0098,
+      "step": 235
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6153555458642626,
+      "learning_rate": 9.936898549044093e-06,
+      "loss": 0.8222,
+      "step": 236
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6015925292473556,
+      "learning_rate": 9.936033039458494e-06,
+      "loss": 1.0327,
+      "step": 237
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6101280322780526,
+      "learning_rate": 9.935161672710728e-06,
+      "loss": 0.9977,
+      "step": 238
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 1.3249438551908315,
+      "learning_rate": 9.934284449834775e-06,
+      "loss": 1.0414,
+      "step": 239
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6088213415472031,
+      "learning_rate": 9.933401371871555e-06,
+      "loss": 0.921,
+      "step": 240
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.618591559944766,
+      "learning_rate": 9.932512439868942e-06,
+      "loss": 0.957,
+      "step": 241
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5246265685349965,
+      "learning_rate": 9.931617654881753e-06,
+      "loss": 0.9706,
+      "step": 242
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.41372151657961087,
+      "learning_rate": 9.930717017971753e-06,
+      "loss": 0.85,
+      "step": 243
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6547040771141608,
+      "learning_rate": 9.929810530207651e-06,
+      "loss": 1.1364,
+      "step": 244
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6518749936935838,
+      "learning_rate": 9.928898192665095e-06,
+      "loss": 0.8715,
+      "step": 245
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.664605861280631,
+      "learning_rate": 9.927980006426677e-06,
+      "loss": 1.059,
+      "step": 246
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6040893846717752,
+      "learning_rate": 9.92705597258193e-06,
+      "loss": 0.8337,
+      "step": 247
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5579921591977253,
+      "learning_rate": 9.926126092227323e-06,
+      "loss": 0.8546,
+      "step": 248
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.6799065782655582,
+      "learning_rate": 9.925190366466269e-06,
+      "loss": 1.117,
+      "step": 249
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5870303484039056,
+      "learning_rate": 9.924248796409107e-06,
+      "loss": 1.0132,
+      "step": 250
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.46845036582509797,
+      "learning_rate": 9.923301383173119e-06,
+      "loss": 0.7543,
+      "step": 251
+    },
+    {
+      "epoch": 0.08,
+      "grad_norm": 0.5932892192881746,
+      "learning_rate": 9.922348127882518e-06,
+      "loss": 0.9116,
+      "step": 252
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6499403879776278,
+      "learning_rate": 9.921389031668449e-06,
+      "loss": 0.9054,
+      "step": 253
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.544623242609157,
+      "learning_rate": 9.920424095668988e-06,
+      "loss": 0.9941,
+      "step": 254
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5292309719167626,
+      "learning_rate": 9.919453321029141e-06,
+      "loss": 1.0092,
+      "step": 255
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5365557107557589,
+      "learning_rate": 9.918476708900843e-06,
+      "loss": 0.9896,
+      "step": 256
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 1.6578813837173556,
+      "learning_rate": 9.917494260442956e-06,
+      "loss": 0.9066,
+      "step": 257
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5473555079833043,
+      "learning_rate": 9.916505976821262e-06,
+      "loss": 0.9258,
+      "step": 258
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.9298154295536868,
+      "learning_rate": 9.915511859208475e-06,
+      "loss": 0.9331,
+      "step": 259
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.42585252813547586,
+      "learning_rate": 9.914511908784227e-06,
+      "loss": 0.7927,
+      "step": 260
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6627469825884029,
+      "learning_rate": 9.913506126735074e-06,
+      "loss": 0.9692,
+      "step": 261
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5369978606595845,
+      "learning_rate": 9.912494514254487e-06,
+      "loss": 0.8895,
+      "step": 262
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.589709578377587,
+      "learning_rate": 9.911477072542861e-06,
+      "loss": 1.0417,
+      "step": 263
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6761237584770096,
+      "learning_rate": 9.910453802807506e-06,
+      "loss": 0.8079,
+      "step": 264
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6108107055902207,
+      "learning_rate": 9.909424706262647e-06,
+      "loss": 0.9398,
+      "step": 265
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6129901789320793,
+      "learning_rate": 9.908389784129424e-06,
+      "loss": 1.0206,
+      "step": 266
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.4008590213387604,
+      "learning_rate": 9.907349037635888e-06,
+      "loss": 1.0479,
+      "step": 267
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7510281994703654,
+      "learning_rate": 9.906302468017002e-06,
+      "loss": 0.937,
+      "step": 268
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5708093192553663,
+      "learning_rate": 9.905250076514642e-06,
+      "loss": 0.6741,
+      "step": 269
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6693250548920765,
+      "learning_rate": 9.904191864377588e-06,
+      "loss": 1.1373,
+      "step": 270
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5065523337717195,
+      "learning_rate": 9.90312783286153e-06,
+      "loss": 0.8307,
+      "step": 271
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5334718815986009,
+      "learning_rate": 9.902057983229059e-06,
+      "loss": 0.9693,
+      "step": 272
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6022358340880763,
+      "learning_rate": 9.900982316749676e-06,
+      "loss": 1.0309,
+      "step": 273
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6217767008089107,
+      "learning_rate": 9.899900834699778e-06,
+      "loss": 0.7995,
+      "step": 274
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5936050034636983,
+      "learning_rate": 9.898813538362669e-06,
+      "loss": 0.9195,
+      "step": 275
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5523693710578743,
+      "learning_rate": 9.897720429028547e-06,
+      "loss": 0.9549,
+      "step": 276
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7158500029317744,
+      "learning_rate": 9.89662150799451e-06,
+      "loss": 0.7869,
+      "step": 277
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6832570396628003,
+      "learning_rate": 9.895516776564555e-06,
+      "loss": 1.0084,
+      "step": 278
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.6515647079845315,
+      "learning_rate": 9.894406236049569e-06,
+      "loss": 0.8382,
+      "step": 279
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.5535959952208508,
+      "learning_rate": 9.893289887767331e-06,
+      "loss": 1.054,
+      "step": 280
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.8076133260950338,
+      "learning_rate": 9.89216773304252e-06,
+      "loss": 0.9692,
+      "step": 281
+    },
+    {
+      "epoch": 0.09,
+      "grad_norm": 0.7504710571257389,
+      "learning_rate": 9.891039773206698e-06,
+      "loss": 1.0334,
+      "step": 282
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5305888004295811,
+      "learning_rate": 9.889906009598316e-06,
+      "loss": 1.0521,
+      "step": 283
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6512227952393306,
+      "learning_rate": 9.888766443562712e-06,
+      "loss": 0.8424,
+      "step": 284
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5621896782432038,
+      "learning_rate": 9.887621076452114e-06,
+      "loss": 1.1036,
+      "step": 285
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5307511306775973,
+      "learning_rate": 9.886469909625624e-06,
+      "loss": 0.7791,
+      "step": 286
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7851180792872695,
+      "learning_rate": 9.885312944449239e-06,
+      "loss": 0.696,
+      "step": 287
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7566428505027745,
+      "learning_rate": 9.884150182295821e-06,
+      "loss": 1.0432,
+      "step": 288
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6649746448937598,
+      "learning_rate": 9.882981624545127e-06,
+      "loss": 1.0452,
+      "step": 289
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.691263588897666,
+      "learning_rate": 9.881807272583776e-06,
+      "loss": 0.938,
+      "step": 290
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5621726051198118,
+      "learning_rate": 9.880627127805276e-06,
+      "loss": 0.8807,
+      "step": 291
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4629085515209141,
+      "learning_rate": 9.879441191609997e-06,
+      "loss": 0.9777,
+      "step": 292
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.624384895152709,
+      "learning_rate": 9.87824946540519e-06,
+      "loss": 0.8675,
+      "step": 293
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.633491524703973,
+      "learning_rate": 9.877051950604972e-06,
+      "loss": 1.0329,
+      "step": 294
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5129047680584472,
+      "learning_rate": 9.87584864863033e-06,
+      "loss": 0.9564,
+      "step": 295
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6150243755796482,
+      "learning_rate": 9.874639560909118e-06,
+      "loss": 0.7576,
+      "step": 296
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6314666966183794,
+      "learning_rate": 9.873424688876057e-06,
+      "loss": 0.8953,
+      "step": 297
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.9380478852994985,
+      "learning_rate": 9.872204033972727e-06,
+      "loss": 0.8634,
+      "step": 298
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5536948045638821,
+      "learning_rate": 9.870977597647575e-06,
+      "loss": 1.0545,
+      "step": 299
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6488262899345278,
+      "learning_rate": 9.869745381355906e-06,
+      "loss": 0.8589,
+      "step": 300
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6937225371856561,
+      "learning_rate": 9.868507386559887e-06,
+      "loss": 0.9617,
+      "step": 301
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6276220538437542,
+      "learning_rate": 9.867263614728535e-06,
+      "loss": 0.8177,
+      "step": 302
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7186435162122448,
+      "learning_rate": 9.866014067337729e-06,
+      "loss": 0.9895,
+      "step": 303
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.528289497257951,
+      "learning_rate": 9.864758745870196e-06,
+      "loss": 0.7654,
+      "step": 304
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.4601392213986357,
+      "learning_rate": 9.863497651815519e-06,
+      "loss": 0.9513,
+      "step": 305
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5558134268320157,
+      "learning_rate": 9.862230786670129e-06,
+      "loss": 1.0056,
+      "step": 306
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.7319111951319074,
+      "learning_rate": 9.860958151937303e-06,
+      "loss": 1.0011,
+      "step": 307
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5102229795624809,
+      "learning_rate": 9.85967974912717e-06,
+      "loss": 0.9856,
+      "step": 308
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.6575739595648712,
+      "learning_rate": 9.858395579756695e-06,
+      "loss": 0.8368,
+      "step": 309
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.550573456772846,
+      "learning_rate": 9.857105645349694e-06,
+      "loss": 0.9541,
+      "step": 310
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.5821467085215216,
+      "learning_rate": 9.855809947436819e-06,
+      "loss": 1.1707,
+      "step": 311
+    },
+    {
+      "epoch": 0.1,
+      "grad_norm": 0.8541175321930523,
+      "learning_rate": 9.854508487555562e-06,
+      "loss": 0.8029,
+      "step": 312
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4004861126656386,
+      "learning_rate": 9.853201267250252e-06,
+      "loss": 0.7004,
+      "step": 313
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6152291348067506,
+      "learning_rate": 9.851888288072053e-06,
+      "loss": 0.9646,
+      "step": 314
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6317341620565768,
+      "learning_rate": 9.850569551578969e-06,
+      "loss": 1.0362,
+      "step": 315
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.37637632811692995,
+      "learning_rate": 9.849245059335824e-06,
+      "loss": 0.9997,
+      "step": 316
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7955932456137332,
+      "learning_rate": 9.847914812914282e-06,
+      "loss": 0.8746,
+      "step": 317
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6120753615438862,
+      "learning_rate": 9.846578813892827e-06,
+      "loss": 0.9569,
+      "step": 318
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5771857162309362,
+      "learning_rate": 9.845237063856777e-06,
+      "loss": 0.9394,
+      "step": 319
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5631465602703184,
+      "learning_rate": 9.843889564398269e-06,
+      "loss": 0.9542,
+      "step": 320
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5361718542976157,
+      "learning_rate": 9.842536317116262e-06,
+      "loss": 0.9229,
+      "step": 321
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4470364612299326,
+      "learning_rate": 9.841177323616539e-06,
+      "loss": 0.7099,
+      "step": 322
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5722642575743017,
+      "learning_rate": 9.8398125855117e-06,
+      "loss": 1.0582,
+      "step": 323
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6059920041882674,
+      "learning_rate": 9.838442104421162e-06,
+      "loss": 0.8893,
+      "step": 324
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5938009284438333,
+      "learning_rate": 9.837065881971153e-06,
+      "loss": 0.9635,
+      "step": 325
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.648468909841806,
+      "learning_rate": 9.835683919794719e-06,
+      "loss": 0.8643,
+      "step": 326
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7128481437128603,
+      "learning_rate": 9.834296219531713e-06,
+      "loss": 0.9449,
+      "step": 327
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.7112302378381029,
+      "learning_rate": 9.832902782828801e-06,
+      "loss": 0.9966,
+      "step": 328
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4198256811680223,
+      "learning_rate": 9.831503611339452e-06,
+      "loss": 0.9908,
+      "step": 329
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5045826714370206,
+      "learning_rate": 9.83009870672394e-06,
+      "loss": 1.0909,
+      "step": 330
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.47493368425847504,
+      "learning_rate": 9.828688070649346e-06,
+      "loss": 0.6265,
+      "step": 331
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5545531556007843,
+      "learning_rate": 9.82727170478955e-06,
+      "loss": 0.9667,
+      "step": 332
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6665375114032382,
+      "learning_rate": 9.825849610825227e-06,
+      "loss": 1.0029,
+      "step": 333
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6642137291206877,
+      "learning_rate": 9.824421790443855e-06,
+      "loss": 1.0647,
+      "step": 334
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5138508661573852,
+      "learning_rate": 9.822988245339701e-06,
+      "loss": 0.7621,
+      "step": 335
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.6243396367964964,
+      "learning_rate": 9.821548977213834e-06,
+      "loss": 0.9179,
+      "step": 336
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5593116332050566,
+      "learning_rate": 9.820103987774108e-06,
+      "loss": 0.9629,
+      "step": 337
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.565586203253813,
+      "learning_rate": 9.818653278735163e-06,
+      "loss": 0.8503,
+      "step": 338
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.4886869663592565,
+      "learning_rate": 9.81719685181843e-06,
+      "loss": 0.7447,
+      "step": 339
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5404216820963684,
+      "learning_rate": 9.815734708752126e-06,
+      "loss": 0.8731,
+      "step": 340
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.46497043434146657,
+      "learning_rate": 9.81426685127125e-06,
+      "loss": 1.0199,
+      "step": 341
+    },
+    {
+      "epoch": 0.11,
+      "grad_norm": 0.5774286165593494,
+      "learning_rate": 9.81279328111758e-06,
+      "loss": 1.0741,
+      "step": 342
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5944851295560838,
+      "learning_rate": 9.811314000039674e-06,
+      "loss": 0.8914,
+      "step": 343
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5745536097030528,
+      "learning_rate": 9.809829009792868e-06,
+      "loss": 1.0425,
+      "step": 344
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5278274709947967,
+      "learning_rate": 9.808338312139267e-06,
+      "loss": 0.8939,
+      "step": 345
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7217137282610596,
+      "learning_rate": 9.806841908847758e-06,
+      "loss": 1.0693,
+      "step": 346
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6421946313320681,
+      "learning_rate": 9.805339801693988e-06,
+      "loss": 0.9419,
+      "step": 347
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6616603010461214,
+      "learning_rate": 9.80383199246038e-06,
+      "loss": 0.8855,
+      "step": 348
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5260594266261932,
+      "learning_rate": 9.802318482936121e-06,
+      "loss": 0.7367,
+      "step": 349
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7053273492181624,
+      "learning_rate": 9.800799274917159e-06,
+      "loss": 0.8791,
+      "step": 350
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6191435305171418,
+      "learning_rate": 9.799274370206207e-06,
+      "loss": 0.9934,
+      "step": 351
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5513150809500457,
+      "learning_rate": 9.79774377061274e-06,
+      "loss": 0.8463,
+      "step": 352
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5591414637264921,
+      "learning_rate": 9.796207477952984e-06,
+      "loss": 0.9553,
+      "step": 353
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5614099714047394,
+      "learning_rate": 9.794665494049926e-06,
+      "loss": 1.1124,
+      "step": 354
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.4638812582233279,
+      "learning_rate": 9.793117820733304e-06,
+      "loss": 0.9635,
+      "step": 355
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.8904672606944889,
+      "learning_rate": 9.791564459839609e-06,
+      "loss": 1.0246,
+      "step": 356
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.49537762529097945,
+      "learning_rate": 9.790005413212075e-06,
+      "loss": 0.6915,
+      "step": 357
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.545642896176596,
+      "learning_rate": 9.788440682700695e-06,
+      "loss": 0.9864,
+      "step": 358
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5503872609807979,
+      "learning_rate": 9.786870270162193e-06,
+      "loss": 0.7184,
+      "step": 359
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5945956687343261,
+      "learning_rate": 9.785294177460043e-06,
+      "loss": 0.917,
+      "step": 360
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5580851317115412,
+      "learning_rate": 9.783712406464459e-06,
+      "loss": 1.0678,
+      "step": 361
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6738806708104627,
+      "learning_rate": 9.782124959052388e-06,
+      "loss": 0.8696,
+      "step": 362
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6207451336008124,
+      "learning_rate": 9.780531837107519e-06,
+      "loss": 0.9775,
+      "step": 363
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5432962135466998,
+      "learning_rate": 9.77893304252027e-06,
+      "loss": 0.9806,
+      "step": 364
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5795661111897684,
+      "learning_rate": 9.777328577187795e-06,
+      "loss": 0.9285,
+      "step": 365
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.7985093517406228,
+      "learning_rate": 9.775718443013969e-06,
+      "loss": 0.7749,
+      "step": 366
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5553393452792643,
+      "learning_rate": 9.774102641909402e-06,
+      "loss": 0.9227,
+      "step": 367
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.650560902492215,
+      "learning_rate": 9.772481175791422e-06,
+      "loss": 1.0343,
+      "step": 368
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.47634610650418613,
+      "learning_rate": 9.770854046584085e-06,
+      "loss": 0.9699,
+      "step": 369
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.6118120644112486,
+      "learning_rate": 9.769221256218165e-06,
+      "loss": 0.9426,
+      "step": 370
+    },
+    {
+      "epoch": 0.12,
+      "grad_norm": 0.5459082942793801,
+      "learning_rate": 9.767582806631148e-06,
+      "loss": 0.7746,
+      "step": 371
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5860984209497028,
+      "learning_rate": 9.765938699767245e-06,
+      "loss": 0.7991,
+      "step": 372
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5913316216472443,
+      "learning_rate": 9.76428893757737e-06,
+      "loss": 0.9674,
+      "step": 373
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6150654715152247,
+      "learning_rate": 9.762633522019159e-06,
+      "loss": 0.9825,
+      "step": 374
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5102049772529004,
+      "learning_rate": 9.760972455056943e-06,
+      "loss": 0.9658,
+      "step": 375
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8494326344734157,
+      "learning_rate": 9.75930573866177e-06,
+      "loss": 0.625,
+      "step": 376
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5736977218822077,
+      "learning_rate": 9.75763337481139e-06,
+      "loss": 0.9926,
+      "step": 377
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6445652449134818,
+      "learning_rate": 9.755955365490246e-06,
+      "loss": 1.0623,
+      "step": 378
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6769674311749971,
+      "learning_rate": 9.754271712689491e-06,
+      "loss": 0.8396,
+      "step": 379
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.46009072671586665,
+      "learning_rate": 9.752582418406969e-06,
+      "loss": 1.0228,
+      "step": 380
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5738690765918417,
+      "learning_rate": 9.750887484647215e-06,
+      "loss": 0.8378,
+      "step": 381
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5487114890939861,
+      "learning_rate": 9.749186913421465e-06,
+      "loss": 0.9939,
+      "step": 382
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8328785904457929,
+      "learning_rate": 9.747480706747637e-06,
+      "loss": 0.6375,
+      "step": 383
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5851011670587644,
+      "learning_rate": 9.745768866650339e-06,
+      "loss": 0.9306,
+      "step": 384
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5967704211042556,
+      "learning_rate": 9.744051395160863e-06,
+      "loss": 0.9782,
+      "step": 385
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.8163861138879221,
+      "learning_rate": 9.742328294317181e-06,
+      "loss": 0.8787,
+      "step": 386
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5890038140571896,
+      "learning_rate": 9.74059956616395e-06,
+      "loss": 1.1483,
+      "step": 387
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.60078850405636,
+      "learning_rate": 9.738865212752505e-06,
+      "loss": 0.9929,
+      "step": 388
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6567908680228934,
+      "learning_rate": 9.737125236140845e-06,
+      "loss": 0.8785,
+      "step": 389
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7155816676576612,
+      "learning_rate": 9.735379638393654e-06,
+      "loss": 1.0028,
+      "step": 390
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5099173826957298,
+      "learning_rate": 9.73362842158228e-06,
+      "loss": 0.9629,
+      "step": 391
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5263090903811279,
+      "learning_rate": 9.73187158778474e-06,
+      "loss": 0.6756,
+      "step": 392
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 1.0978566857429124,
+      "learning_rate": 9.73010913908572e-06,
+      "loss": 0.8761,
+      "step": 393
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.6183421239177268,
+      "learning_rate": 9.72834107757656e-06,
+      "loss": 1.0078,
+      "step": 394
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5546891927462532,
+      "learning_rate": 9.726567405355268e-06,
+      "loss": 0.7478,
+      "step": 395
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.549963959624232,
+      "learning_rate": 9.724788124526504e-06,
+      "loss": 1.0426,
+      "step": 396
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5384401602849483,
+      "learning_rate": 9.723003237201588e-06,
+      "loss": 0.8755,
+      "step": 397
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.7223408597967157,
+      "learning_rate": 9.721212745498493e-06,
+      "loss": 0.9742,
+      "step": 398
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5701105056091407,
+      "learning_rate": 9.719416651541839e-06,
+      "loss": 0.9182,
+      "step": 399
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 2.72901586796143,
+      "learning_rate": 9.717614957462892e-06,
+      "loss": 0.9496,
+      "step": 400
+    },
+    {
+      "epoch": 0.13,
+      "grad_norm": 0.5235443907607203,
+      "learning_rate": 9.71580766539957e-06,
+      "loss": 1.0009,
+      "step": 401
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.49893386673043194,
+      "learning_rate": 9.713994777496427e-06,
+      "loss": 0.6215,
+      "step": 402
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6494703882594394,
+      "learning_rate": 9.712176295904661e-06,
+      "loss": 0.9599,
+      "step": 403
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5961598395950723,
+      "learning_rate": 9.710352222782104e-06,
+      "loss": 0.9295,
+      "step": 404
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6826712834484538,
+      "learning_rate": 9.70852256029323e-06,
+      "loss": 0.8143,
+      "step": 405
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.4460785212974887,
+      "learning_rate": 9.706687310609137e-06,
+      "loss": 1.0335,
+      "step": 406
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6880498578333829,
+      "learning_rate": 9.704846475907558e-06,
+      "loss": 0.8039,
+      "step": 407
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5801094270042266,
+      "learning_rate": 9.70300005837285e-06,
+      "loss": 1.0487,
+      "step": 408
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5447757857792805,
+      "learning_rate": 9.701148060196001e-06,
+      "loss": 0.9332,
+      "step": 409
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5651838920997194,
+      "learning_rate": 9.699290483574611e-06,
+      "loss": 0.6241,
+      "step": 410
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.509330653221104,
+      "learning_rate": 9.697427330712908e-06,
+      "loss": 1.0826,
+      "step": 411
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5473710987688465,
+      "learning_rate": 9.695558603821735e-06,
+      "loss": 0.9143,
+      "step": 412
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 1.040425450086061,
+      "learning_rate": 9.693684305118545e-06,
+      "loss": 1.0376,
+      "step": 413
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6199244100356727,
+      "learning_rate": 9.691804436827409e-06,
+      "loss": 0.8645,
+      "step": 414
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.758085370317742,
+      "learning_rate": 9.689919001178998e-06,
+      "loss": 0.8816,
+      "step": 415
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5529447378488773,
+      "learning_rate": 9.6880280004106e-06,
+      "loss": 1.0433,
+      "step": 416
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.48522837038972855,
+      "learning_rate": 9.686131436766104e-06,
+      "loss": 0.9357,
+      "step": 417
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6061981182888246,
+      "learning_rate": 9.68422931249599e-06,
+      "loss": 0.9832,
+      "step": 418
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.4825548999741968,
+      "learning_rate": 9.682321629857348e-06,
+      "loss": 0.807,
+      "step": 419
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5160556923600843,
+      "learning_rate": 9.68040839111386e-06,
+      "loss": 0.8088,
+      "step": 420
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5925293271597755,
+      "learning_rate": 9.678489598535799e-06,
+      "loss": 0.8966,
+      "step": 421
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6060682591811569,
+      "learning_rate": 9.67656525440003e-06,
+      "loss": 0.7686,
+      "step": 422
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5575940925697245,
+      "learning_rate": 9.674635360990005e-06,
+      "loss": 1.0744,
+      "step": 423
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.64525213859389,
+      "learning_rate": 9.672699920595763e-06,
+      "loss": 0.8914,
+      "step": 424
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6284233470261462,
+      "learning_rate": 9.670758935513922e-06,
+      "loss": 0.9495,
+      "step": 425
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5988351309047422,
+      "learning_rate": 9.66881240804768e-06,
+      "loss": 0.9851,
+      "step": 426
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6658063448298998,
+      "learning_rate": 9.666860340506809e-06,
+      "loss": 0.9953,
+      "step": 427
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.5809729442275439,
+      "learning_rate": 9.664902735207664e-06,
+      "loss": 0.8141,
+      "step": 428
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.48915883899184043,
+      "learning_rate": 9.662939594473164e-06,
+      "loss": 0.8331,
+      "step": 429
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6728754771926357,
+      "learning_rate": 9.660970920632798e-06,
+      "loss": 1.0833,
+      "step": 430
+    },
+    {
+      "epoch": 0.14,
+      "grad_norm": 0.6149040732868122,
+      "learning_rate": 9.658996716022618e-06,
+      "loss": 0.9186,
+      "step": 431
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.521750685700819,
+      "learning_rate": 9.657016982985242e-06,
+      "loss": 0.9554,
+      "step": 432
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.947554296727018,
+      "learning_rate": 9.655031723869848e-06,
+      "loss": 0.8281,
+      "step": 433
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 1.0467818627513927,
+      "learning_rate": 9.653040941032173e-06,
+      "loss": 0.8908,
+      "step": 434
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.8010429471838942,
+      "learning_rate": 9.651044636834501e-06,
+      "loss": 1.071,
+      "step": 435
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.437401316351953,
+      "learning_rate": 9.64904281364568e-06,
+      "loss": 0.7954,
+      "step": 436
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6279312290820683,
+      "learning_rate": 9.647035473841093e-06,
+      "loss": 0.9634,
+      "step": 437
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6569948798780001,
+      "learning_rate": 9.64502261980268e-06,
+      "loss": 0.9663,
+      "step": 438
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5526856701620505,
+      "learning_rate": 9.643004253918919e-06,
+      "loss": 1.0032,
+      "step": 439
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5968223971628108,
+      "learning_rate": 9.64098037858483e-06,
+      "loss": 0.8361,
+      "step": 440
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.38323346546039927,
+      "learning_rate": 9.638950996201972e-06,
+      "loss": 0.9619,
+      "step": 441
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5679663788406791,
+      "learning_rate": 9.636916109178433e-06,
+      "loss": 1.0666,
+      "step": 442
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7253280144000139,
+      "learning_rate": 9.63487571992884e-06,
+      "loss": 0.7714,
+      "step": 443
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5670252480177316,
+      "learning_rate": 9.632829830874345e-06,
+      "loss": 0.8835,
+      "step": 444
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6092329870070661,
+      "learning_rate": 9.630778444442628e-06,
+      "loss": 0.8842,
+      "step": 445
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.4325144553317635,
+      "learning_rate": 9.628721563067888e-06,
+      "loss": 0.8031,
+      "step": 446
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6564597491306782,
+      "learning_rate": 9.626659189190852e-06,
+      "loss": 0.8459,
+      "step": 447
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5808602529354175,
+      "learning_rate": 9.62459132525876e-06,
+      "loss": 0.8992,
+      "step": 448
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5562653507401123,
+      "learning_rate": 9.62251797372536e-06,
+      "loss": 0.9343,
+      "step": 449
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.7786554741093257,
+      "learning_rate": 9.620439137050927e-06,
+      "loss": 0.8615,
+      "step": 450
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6533362239923904,
+      "learning_rate": 9.618354817702231e-06,
+      "loss": 0.9244,
+      "step": 451
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5113879722176844,
+      "learning_rate": 9.616265018152555e-06,
+      "loss": 0.949,
+      "step": 452
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.3707901422385307,
+      "learning_rate": 9.614169740881683e-06,
+      "loss": 1.0125,
+      "step": 453
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.43303734545050104,
+      "learning_rate": 9.612068988375898e-06,
+      "loss": 0.6633,
+      "step": 454
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6703783594578631,
+      "learning_rate": 9.609962763127978e-06,
+      "loss": 0.932,
+      "step": 455
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5467986292480502,
+      "learning_rate": 9.607851067637205e-06,
+      "loss": 0.9298,
+      "step": 456
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.472445451586459,
+      "learning_rate": 9.605733904409335e-06,
+      "loss": 0.8159,
+      "step": 457
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.9768992833703017,
+      "learning_rate": 9.603611275956632e-06,
+      "loss": 0.9679,
+      "step": 458
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.5818237845489661,
+      "learning_rate": 9.601483184797827e-06,
+      "loss": 1.0072,
+      "step": 459
+    },
+    {
+      "epoch": 0.15,
+      "grad_norm": 0.6361684879296056,
+      "learning_rate": 9.599349633458145e-06,
+      "loss": 0.801,
+      "step": 460
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6280644884901198,
+      "learning_rate": 9.597210624469288e-06,
+      "loss": 0.9376,
+      "step": 461
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.519193784031376,
+      "learning_rate": 9.595066160369428e-06,
+      "loss": 0.9386,
+      "step": 462
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.4177007082293463,
+      "learning_rate": 9.592916243703217e-06,
+      "loss": 0.7797,
+      "step": 463
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5444046073235731,
+      "learning_rate": 9.590760877021775e-06,
+      "loss": 0.8565,
+      "step": 464
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6923944194433561,
+      "learning_rate": 9.588600062882686e-06,
+      "loss": 0.861,
+      "step": 465
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5261033091287904,
+      "learning_rate": 9.586433803850002e-06,
+      "loss": 0.9418,
+      "step": 466
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5201511869057559,
+      "learning_rate": 9.584262102494238e-06,
+      "loss": 0.9316,
+      "step": 467
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6016264207949443,
+      "learning_rate": 9.582084961392358e-06,
+      "loss": 1.0393,
+      "step": 468
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5422154914206374,
+      "learning_rate": 9.579902383127791e-06,
+      "loss": 0.9285,
+      "step": 469
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5288575200901744,
+      "learning_rate": 9.57771437029041e-06,
+      "loss": 1.0273,
+      "step": 470
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5716590461196561,
+      "learning_rate": 9.575520925476543e-06,
+      "loss": 0.8203,
+      "step": 471
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5223964972373102,
+      "learning_rate": 9.573322051288958e-06,
+      "loss": 0.7805,
+      "step": 472
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.42853959696993105,
+      "learning_rate": 9.57111775033687e-06,
+      "loss": 0.6512,
+      "step": 473
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6378210077902788,
+      "learning_rate": 9.56890802523593e-06,
+      "loss": 1.0117,
+      "step": 474
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6102680015772098,
+      "learning_rate": 9.566692878608229e-06,
+      "loss": 0.9969,
+      "step": 475
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6320462329098551,
+      "learning_rate": 9.564472313082288e-06,
+      "loss": 0.9413,
+      "step": 476
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5821068552902906,
+      "learning_rate": 9.56224633129306e-06,
+      "loss": 0.7915,
+      "step": 477
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.46626758468438007,
+      "learning_rate": 9.560014935881924e-06,
+      "loss": 1.0076,
+      "step": 478
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6592922041342202,
+      "learning_rate": 9.557778129496684e-06,
+      "loss": 0.8338,
+      "step": 479
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5972856289501323,
+      "learning_rate": 9.555535914791566e-06,
+      "loss": 0.9985,
+      "step": 480
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5722178292628372,
+      "learning_rate": 9.553288294427205e-06,
+      "loss": 0.9672,
+      "step": 481
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5325902284264654,
+      "learning_rate": 9.551035271070665e-06,
+      "loss": 0.7289,
+      "step": 482
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6141168999986168,
+      "learning_rate": 9.548776847395408e-06,
+      "loss": 0.8934,
+      "step": 483
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5922077787018413,
+      "learning_rate": 9.546513026081311e-06,
+      "loss": 0.8135,
+      "step": 484
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.594655020359658,
+      "learning_rate": 9.544243809814657e-06,
+      "loss": 1.0734,
+      "step": 485
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.7499142196766171,
+      "learning_rate": 9.541969201288123e-06,
+      "loss": 0.838,
+      "step": 486
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6517774725686335,
+      "learning_rate": 9.539689203200794e-06,
+      "loss": 0.9542,
+      "step": 487
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6079255977257327,
+      "learning_rate": 9.537403818258143e-06,
+      "loss": 0.823,
+      "step": 488
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.6052639410914262,
+      "learning_rate": 9.53511304917204e-06,
+      "loss": 1.1079,
+      "step": 489
+    },
+    {
+      "epoch": 0.16,
+      "grad_norm": 0.5591698339263936,
+      "learning_rate": 9.532816898660742e-06,
+      "loss": 0.6588,
+      "step": 490
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5032166425488378,
+      "learning_rate": 9.530515369448889e-06,
+      "loss": 0.8967,
+      "step": 491
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5592572154304799,
+      "learning_rate": 9.528208464267508e-06,
+      "loss": 0.9939,
+      "step": 492
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7741678083890957,
+      "learning_rate": 9.525896185854005e-06,
+      "loss": 1.0751,
+      "step": 493
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5894754722126929,
+      "learning_rate": 9.523578536952155e-06,
+      "loss": 0.8719,
+      "step": 494
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8964037270634329,
+      "learning_rate": 9.521255520312117e-06,
+      "loss": 0.8704,
+      "step": 495
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.714212627708992,
+      "learning_rate": 9.51892713869041e-06,
+      "loss": 0.8328,
+      "step": 496
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6535856058884094,
+      "learning_rate": 9.516593394849923e-06,
+      "loss": 1.133,
+      "step": 497
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7370791231455062,
+      "learning_rate": 9.514254291559905e-06,
+      "loss": 0.8555,
+      "step": 498
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.49662407770910477,
+      "learning_rate": 9.51190983159597e-06,
+      "loss": 0.6632,
+      "step": 499
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5407738292115046,
+      "learning_rate": 9.509560017740087e-06,
+      "loss": 0.8391,
+      "step": 500
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7140068420747377,
+      "learning_rate": 9.507204852780571e-06,
+      "loss": 1.0438,
+      "step": 501
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.4531367153147378,
+      "learning_rate": 9.504844339512096e-06,
+      "loss": 0.9677,
+      "step": 502
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.7068219495463897,
+      "learning_rate": 9.502478480735678e-06,
+      "loss": 0.8867,
+      "step": 503
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6037311983981587,
+      "learning_rate": 9.500107279258672e-06,
+      "loss": 0.9029,
+      "step": 504
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 2.149419346955673,
+      "learning_rate": 9.49773073789478e-06,
+      "loss": 0.9893,
+      "step": 505
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.9390184024672311,
+      "learning_rate": 9.495348859464042e-06,
+      "loss": 0.9556,
+      "step": 506
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6843537552981978,
+      "learning_rate": 9.492961646792819e-06,
+      "loss": 0.7224,
+      "step": 507
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6277057194870288,
+      "learning_rate": 9.490569102713814e-06,
+      "loss": 1.0444,
+      "step": 508
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6615802917512523,
+      "learning_rate": 9.488171230066048e-06,
+      "loss": 0.8472,
+      "step": 509
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6339933779540224,
+      "learning_rate": 9.485768031694872e-06,
+      "loss": 0.8606,
+      "step": 510
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5865254084812384,
+      "learning_rate": 9.48335951045195e-06,
+      "loss": 0.9123,
+      "step": 511
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.8710610881966934,
+      "learning_rate": 9.480945669195267e-06,
+      "loss": 0.9216,
+      "step": 512
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 1.2662110386831076,
+      "learning_rate": 9.47852651078912e-06,
+      "loss": 1.0412,
+      "step": 513
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.5972147607440076,
+      "learning_rate": 9.476102038104112e-06,
+      "loss": 0.849,
+      "step": 514
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.4185178564131485,
+      "learning_rate": 9.473672254017155e-06,
+      "loss": 0.9827,
+      "step": 515
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6660244599739163,
+      "learning_rate": 9.471237161411462e-06,
+      "loss": 1.0393,
+      "step": 516
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.45567915668562337,
+      "learning_rate": 9.468796763176549e-06,
+      "loss": 0.5901,
+      "step": 517
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6242491543075362,
+      "learning_rate": 9.466351062208223e-06,
+      "loss": 0.9807,
+      "step": 518
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6320498692522718,
+      "learning_rate": 9.463900061408585e-06,
+      "loss": 0.9381,
+      "step": 519
+    },
+    {
+      "epoch": 0.17,
+      "grad_norm": 0.6692884841028498,
+      "learning_rate": 9.461443763686027e-06,
+      "loss": 0.9966,
+      "step": 520
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5680567843402327,
+      "learning_rate": 9.45898217195522e-06,
+      "loss": 0.7574,
+      "step": 521
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6278983158991193,
+      "learning_rate": 9.456515289137127e-06,
+      "loss": 0.9181,
+      "step": 522
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6199989835429864,
+      "learning_rate": 9.454043118158979e-06,
+      "loss": 0.9868,
+      "step": 523
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5864762847881224,
+      "learning_rate": 9.45156566195429e-06,
+      "loss": 0.8429,
+      "step": 524
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.46181880520142093,
+      "learning_rate": 9.44908292346284e-06,
+      "loss": 0.7761,
+      "step": 525
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8046571870589957,
+      "learning_rate": 9.446594905630682e-06,
+      "loss": 0.7755,
+      "step": 526
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.4485206259850852,
+      "learning_rate": 9.444101611410127e-06,
+      "loss": 1.0147,
+      "step": 527
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6054278499287719,
+      "learning_rate": 9.441603043759756e-06,
+      "loss": 0.8811,
+      "step": 528
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6196728176702189,
+      "learning_rate": 9.4390992056444e-06,
+      "loss": 0.9349,
+      "step": 529
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.88334101733058,
+      "learning_rate": 9.436590100035145e-06,
+      "loss": 0.9792,
+      "step": 530
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5339295865381487,
+      "learning_rate": 9.43407572990933e-06,
+      "loss": 0.8036,
+      "step": 531
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.717405816667724,
+      "learning_rate": 9.43155609825054e-06,
+      "loss": 1.0081,
+      "step": 532
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5823634323940153,
+      "learning_rate": 9.429031208048604e-06,
+      "loss": 0.841,
+      "step": 533
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5861230505096734,
+      "learning_rate": 9.42650106229959e-06,
+      "loss": 0.9265,
+      "step": 534
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.474804188035922,
+      "learning_rate": 9.423965664005801e-06,
+      "loss": 0.6928,
+      "step": 535
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6678943625680691,
+      "learning_rate": 9.421425016175776e-06,
+      "loss": 0.8124,
+      "step": 536
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6624960547081093,
+      "learning_rate": 9.41887912182428e-06,
+      "loss": 0.9692,
+      "step": 537
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8692186263480175,
+      "learning_rate": 9.416327983972304e-06,
+      "loss": 0.8315,
+      "step": 538
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.539767181950713,
+      "learning_rate": 9.413771605647067e-06,
+      "loss": 0.9455,
+      "step": 539
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.36183533814511526,
+      "learning_rate": 9.411209989881995e-06,
+      "loss": 1.023,
+      "step": 540
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.8149997138525646,
+      "learning_rate": 9.408643139716737e-06,
+      "loss": 0.8341,
+      "step": 541
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5208065081674653,
+      "learning_rate": 9.406071058197154e-06,
+      "loss": 0.7365,
+      "step": 542
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5876120662184348,
+      "learning_rate": 9.403493748375311e-06,
+      "loss": 0.8438,
+      "step": 543
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6325527246521822,
+      "learning_rate": 9.400911213309478e-06,
+      "loss": 0.9949,
+      "step": 544
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7011517470225411,
+      "learning_rate": 9.398323456064124e-06,
+      "loss": 0.6921,
+      "step": 545
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5024833002240461,
+      "learning_rate": 9.395730479709916e-06,
+      "loss": 0.8688,
+      "step": 546
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.7034907259929679,
+      "learning_rate": 9.393132287323714e-06,
+      "loss": 1.078,
+      "step": 547
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.756797958378029,
+      "learning_rate": 9.39052888198857e-06,
+      "loss": 0.81,
+      "step": 548
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.5544651295035788,
+      "learning_rate": 9.387920266793718e-06,
+      "loss": 0.9431,
+      "step": 549
+    },
+    {
+      "epoch": 0.18,
+      "grad_norm": 0.6823842566433878,
+      "learning_rate": 9.385306444834573e-06,
+      "loss": 1.0263,
+      "step": 550
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.822935778848102,
+      "learning_rate": 9.382687419212733e-06,
+      "loss": 0.8462,
+      "step": 551
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6312311057889685,
+      "learning_rate": 9.380063193035968e-06,
+      "loss": 0.7445,
+      "step": 552
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7676546342947889,
+      "learning_rate": 9.377433769418218e-06,
+      "loss": 0.9325,
+      "step": 553
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.3774925008556089,
+      "learning_rate": 9.374799151479597e-06,
+      "loss": 1.0199,
+      "step": 554
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6908085139375705,
+      "learning_rate": 9.372159342346371e-06,
+      "loss": 0.779,
+      "step": 555
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7424443430127786,
+      "learning_rate": 9.369514345150977e-06,
+      "loss": 0.9924,
+      "step": 556
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6724292313052436,
+      "learning_rate": 9.366864163032e-06,
+      "loss": 0.7248,
+      "step": 557
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5976007578674862,
+      "learning_rate": 9.364208799134187e-06,
+      "loss": 0.7161,
+      "step": 558
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5478196011290102,
+      "learning_rate": 9.361548256608421e-06,
+      "loss": 0.9565,
+      "step": 559
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6065567697676643,
+      "learning_rate": 9.358882538611743e-06,
+      "loss": 0.8558,
+      "step": 560
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5416356043324321,
+      "learning_rate": 9.356211648307326e-06,
+      "loss": 0.7416,
+      "step": 561
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6577442525761232,
+      "learning_rate": 9.353535588864481e-06,
+      "loss": 0.8387,
+      "step": 562
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.562717482212086,
+      "learning_rate": 9.350854363458663e-06,
+      "loss": 1.0413,
+      "step": 563
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5245936370407582,
+      "learning_rate": 9.348167975271443e-06,
+      "loss": 0.9339,
+      "step": 564
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4361532228598306,
+      "learning_rate": 9.34547642749053e-06,
+      "loss": 0.9519,
+      "step": 565
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6711731795950429,
+      "learning_rate": 9.342779723309746e-06,
+      "loss": 0.9723,
+      "step": 566
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5936217518906253,
+      "learning_rate": 9.340077865929035e-06,
+      "loss": 0.8396,
+      "step": 567
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6197722646031352,
+      "learning_rate": 9.337370858554462e-06,
+      "loss": 0.9471,
+      "step": 568
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 2.647958131924657,
+      "learning_rate": 9.334658704398193e-06,
+      "loss": 0.6003,
+      "step": 569
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5693599489478767,
+      "learning_rate": 9.33194140667851e-06,
+      "loss": 0.8495,
+      "step": 570
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.7434738726689757,
+      "learning_rate": 9.329218968619793e-06,
+      "loss": 0.9663,
+      "step": 571
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.8177808910483898,
+      "learning_rate": 9.326491393452521e-06,
+      "loss": 0.8851,
+      "step": 572
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5875674432746177,
+      "learning_rate": 9.323758684413272e-06,
+      "loss": 1.1654,
+      "step": 573
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5510452907735479,
+      "learning_rate": 9.321020844744717e-06,
+      "loss": 0.9623,
+      "step": 574
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5242130687952804,
+      "learning_rate": 9.31827787769561e-06,
+      "loss": 0.9809,
+      "step": 575
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6061244221166413,
+      "learning_rate": 9.315529786520792e-06,
+      "loss": 0.8215,
+      "step": 576
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.4113510277055668,
+      "learning_rate": 9.312776574481187e-06,
+      "loss": 0.9333,
+      "step": 577
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.5621804797469198,
+      "learning_rate": 9.310018244843789e-06,
+      "loss": 0.662,
+      "step": 578
+    },
+    {
+      "epoch": 0.19,
+      "grad_norm": 0.6072415017334084,
+      "learning_rate": 9.307254800881672e-06,
+      "loss": 0.892,
+      "step": 579
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5085417377701305,
+      "learning_rate": 9.304486245873973e-06,
+      "loss": 0.8398,
+      "step": 580
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5548970352981448,
+      "learning_rate": 9.301712583105898e-06,
+      "loss": 0.8247,
+      "step": 581
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 1.0764129771108173,
+      "learning_rate": 9.29893381586871e-06,
+      "loss": 1.0682,
+      "step": 582
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.575476073787929,
+      "learning_rate": 9.296149947459731e-06,
+      "loss": 0.8848,
+      "step": 583
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6359001573213268,
+      "learning_rate": 9.293360981182338e-06,
+      "loss": 0.9099,
+      "step": 584
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5302434478553921,
+      "learning_rate": 9.290566920345954e-06,
+      "loss": 0.9143,
+      "step": 585
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5204816843052835,
+      "learning_rate": 9.287767768266046e-06,
+      "loss": 0.9067,
+      "step": 586
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6918697301649233,
+      "learning_rate": 9.284963528264133e-06,
+      "loss": 0.9985,
+      "step": 587
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7616639057347657,
+      "learning_rate": 9.282154203667754e-06,
+      "loss": 0.6037,
+      "step": 588
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.4760795235535698,
+      "learning_rate": 9.279339797810497e-06,
+      "loss": 0.9687,
+      "step": 589
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7463888388632527,
+      "learning_rate": 9.27652031403197e-06,
+      "loss": 0.874,
+      "step": 590
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.69803394521354,
+      "learning_rate": 9.27369575567781e-06,
+      "loss": 0.8614,
+      "step": 591
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5092602085888007,
+      "learning_rate": 9.270866126099679e-06,
+      "loss": 0.9723,
+      "step": 592
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8409349988933044,
+      "learning_rate": 9.268031428655248e-06,
+      "loss": 0.8109,
+      "step": 593
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6173200806115496,
+      "learning_rate": 9.26519166670821e-06,
+      "loss": 0.9086,
+      "step": 594
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5848038780231027,
+      "learning_rate": 9.26234684362826e-06,
+      "loss": 0.9979,
+      "step": 595
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.8028207928172314,
+      "learning_rate": 9.259496962791108e-06,
+      "loss": 0.6431,
+      "step": 596
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6119130117231941,
+      "learning_rate": 9.256642027578458e-06,
+      "loss": 0.9832,
+      "step": 597
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6452265323836353,
+      "learning_rate": 9.253782041378012e-06,
+      "loss": 0.9187,
+      "step": 598
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.694250189354051,
+      "learning_rate": 9.25091700758347e-06,
+      "loss": 1.0804,
+      "step": 599
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.649388947337623,
+      "learning_rate": 9.248046929594519e-06,
+      "loss": 0.8411,
+      "step": 600
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6439326756220767,
+      "learning_rate": 9.24517181081683e-06,
+      "loss": 0.8077,
+      "step": 601
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5766170097747314,
+      "learning_rate": 9.24229165466206e-06,
+      "loss": 0.9768,
+      "step": 602
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5033466286472357,
+      "learning_rate": 9.239406464547835e-06,
+      "loss": 0.8976,
+      "step": 603
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5570578661233653,
+      "learning_rate": 9.236516243897764e-06,
+      "loss": 0.7882,
+      "step": 604
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.619039307511655,
+      "learning_rate": 9.233620996141421e-06,
+      "loss": 0.8949,
+      "step": 605
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.5456180294223744,
+      "learning_rate": 9.230720724714345e-06,
+      "loss": 0.7904,
+      "step": 606
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.6170787304213075,
+      "learning_rate": 9.227815433058037e-06,
+      "loss": 0.8197,
+      "step": 607
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 0.7407974237440026,
+      "learning_rate": 9.22490512461995e-06,
+      "loss": 0.7807,
+      "step": 608
+    },
+    {
+      "epoch": 0.2,
+      "grad_norm": 3.0240470483049107,
+      "learning_rate": 9.2219898028535e-06,
+      "loss": 1.0784,
+      "step": 609
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7139967612502718,
+      "learning_rate": 9.219069471218045e-06,
+      "loss": 0.8259,
+      "step": 610
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5736487881030515,
+      "learning_rate": 9.216144133178889e-06,
+      "loss": 0.9655,
+      "step": 611
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5597535348192478,
+      "learning_rate": 9.213213792207275e-06,
+      "loss": 0.9286,
+      "step": 612
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5589466456909282,
+      "learning_rate": 9.210278451780386e-06,
+      "loss": 0.9786,
+      "step": 613
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.41388845565854837,
+      "learning_rate": 9.207338115381337e-06,
+      "loss": 0.5944,
+      "step": 614
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.4504920715663986,
+      "learning_rate": 9.204392786499168e-06,
+      "loss": 0.922,
+      "step": 615
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.56242972421395,
+      "learning_rate": 9.201442468628847e-06,
+      "loss": 1.0719,
+      "step": 616
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.579116979481909,
+      "learning_rate": 9.198487165271259e-06,
+      "loss": 0.7653,
+      "step": 617
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5453102109091742,
+      "learning_rate": 9.195526879933206e-06,
+      "loss": 1.0083,
+      "step": 618
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6182376049519114,
+      "learning_rate": 9.192561616127403e-06,
+      "loss": 0.8142,
+      "step": 619
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.577901337812579,
+      "learning_rate": 9.189591377372468e-06,
+      "loss": 0.866,
+      "step": 620
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5345806447132215,
+      "learning_rate": 9.18661616719293e-06,
+      "loss": 0.9445,
+      "step": 621
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.41864455223202507,
+      "learning_rate": 9.183635989119211e-06,
+      "loss": 0.7726,
+      "step": 622
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5470366421285914,
+      "learning_rate": 9.18065084668763e-06,
+      "loss": 0.9664,
+      "step": 623
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7794642975027475,
+      "learning_rate": 9.177660743440395e-06,
+      "loss": 0.8824,
+      "step": 624
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5679977642948043,
+      "learning_rate": 9.174665682925606e-06,
+      "loss": 0.969,
+      "step": 625
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5393353224291628,
+      "learning_rate": 9.171665668697236e-06,
+      "loss": 0.8265,
+      "step": 626
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.3657233762404947,
+      "learning_rate": 9.168660704315146e-06,
+      "loss": 0.9824,
+      "step": 627
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.8163737928644285,
+      "learning_rate": 9.165650793345064e-06,
+      "loss": 0.9702,
+      "step": 628
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6440036547956963,
+      "learning_rate": 9.162635939358593e-06,
+      "loss": 0.728,
+      "step": 629
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.7323442468307535,
+      "learning_rate": 9.159616145933194e-06,
+      "loss": 0.8814,
+      "step": 630
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6180501913833103,
+      "learning_rate": 9.156591416652198e-06,
+      "loss": 0.9048,
+      "step": 631
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.38734915767829553,
+      "learning_rate": 9.15356175510479e-06,
+      "loss": 0.8171,
+      "step": 632
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6225134995297875,
+      "learning_rate": 9.150527164886e-06,
+      "loss": 0.7875,
+      "step": 633
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5799861258927985,
+      "learning_rate": 9.14748764959672e-06,
+      "loss": 0.85,
+      "step": 634
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.4842120165141489,
+      "learning_rate": 9.144443212843675e-06,
+      "loss": 0.8955,
+      "step": 635
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.6832294039422693,
+      "learning_rate": 9.141393858239435e-06,
+      "loss": 0.8292,
+      "step": 636
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5648444489989328,
+      "learning_rate": 9.138339589402405e-06,
+      "loss": 0.9889,
+      "step": 637
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.5669014455463525,
+      "learning_rate": 9.135280409956819e-06,
+      "loss": 0.8352,
+      "step": 638
+    },
+    {
+      "epoch": 0.21,
+      "grad_norm": 0.4296311931767217,
+      "learning_rate": 9.132216323532743e-06,
+      "loss": 0.9795,
+      "step": 639
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.44191755455166937,
+      "learning_rate": 9.12914733376606e-06,
+      "loss": 0.6971,
+      "step": 640
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7862668242160047,
+      "learning_rate": 9.126073444298473e-06,
+      "loss": 0.8529,
+      "step": 641
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5636426624862624,
+      "learning_rate": 9.122994658777504e-06,
+      "loss": 0.8922,
+      "step": 642
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.538834093046837,
+      "learning_rate": 9.119910980856477e-06,
+      "loss": 0.8777,
+      "step": 643
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5812969869912092,
+      "learning_rate": 9.116822414194528e-06,
+      "loss": 0.9052,
+      "step": 644
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6098410274597778,
+      "learning_rate": 9.113728962456588e-06,
+      "loss": 0.956,
+      "step": 645
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5558231389592343,
+      "learning_rate": 9.110630629313388e-06,
+      "loss": 0.7453,
+      "step": 646
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5725464319729358,
+      "learning_rate": 9.107527418441455e-06,
+      "loss": 0.9061,
+      "step": 647
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.595794358749266,
+      "learning_rate": 9.104419333523096e-06,
+      "loss": 0.9625,
+      "step": 648
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.40271323064184605,
+      "learning_rate": 9.101306378246408e-06,
+      "loss": 0.7604,
+      "step": 649
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5406291081987763,
+      "learning_rate": 9.098188556305262e-06,
+      "loss": 0.8965,
+      "step": 650
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5917686778706932,
+      "learning_rate": 9.095065871399312e-06,
+      "loss": 0.8526,
+      "step": 651
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.4741360502741213,
+      "learning_rate": 9.091938327233973e-06,
+      "loss": 0.981,
+      "step": 652
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5812130642055352,
+      "learning_rate": 9.088805927520432e-06,
+      "loss": 0.7518,
+      "step": 653
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6608151169134449,
+      "learning_rate": 9.085668675975634e-06,
+      "loss": 0.98,
+      "step": 654
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5594124319508992,
+      "learning_rate": 9.082526576322286e-06,
+      "loss": 0.9215,
+      "step": 655
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5296871344564329,
+      "learning_rate": 9.079379632288842e-06,
+      "loss": 1.0115,
+      "step": 656
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.59224229719837,
+      "learning_rate": 9.076227847609513e-06,
+      "loss": 0.8123,
+      "step": 657
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6305101355190528,
+      "learning_rate": 9.073071226024242e-06,
+      "loss": 0.7884,
+      "step": 658
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.40778721206732627,
+      "learning_rate": 9.06990977127872e-06,
+      "loss": 0.6413,
+      "step": 659
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7410536554598425,
+      "learning_rate": 9.066743487124374e-06,
+      "loss": 1.0639,
+      "step": 660
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6144658159126587,
+      "learning_rate": 9.063572377318356e-06,
+      "loss": 0.8973,
+      "step": 661
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6289561146548285,
+      "learning_rate": 9.060396445623545e-06,
+      "loss": 0.9008,
+      "step": 662
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.5765227397815255,
+      "learning_rate": 9.057215695808548e-06,
+      "loss": 0.8721,
+      "step": 663
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.46188924175150287,
+      "learning_rate": 9.054030131647682e-06,
+      "loss": 0.9828,
+      "step": 664
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.540257164096004,
+      "learning_rate": 9.05083975692098e-06,
+      "loss": 0.8322,
+      "step": 665
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.8141755093492324,
+      "learning_rate": 9.047644575414184e-06,
+      "loss": 0.9102,
+      "step": 666
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.48460030567868884,
+      "learning_rate": 9.044444590918737e-06,
+      "loss": 0.7332,
+      "step": 667
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.6638101913983566,
+      "learning_rate": 9.041239807231785e-06,
+      "loss": 0.9804,
+      "step": 668
+    },
+    {
+      "epoch": 0.22,
+      "grad_norm": 0.7345279284747156,
+      "learning_rate": 9.038030228156165e-06,
+      "loss": 0.9136,
+      "step": 669
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9046439086188528,
+      "learning_rate": 9.034815857500407e-06,
+      "loss": 0.8003,
+      "step": 670
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6297810872709216,
+      "learning_rate": 9.031596699078727e-06,
+      "loss": 1.0511,
+      "step": 671
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6347483289881057,
+      "learning_rate": 9.028372756711017e-06,
+      "loss": 0.7848,
+      "step": 672
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6282256867781089,
+      "learning_rate": 9.025144034222853e-06,
+      "loss": 0.9527,
+      "step": 673
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5748654427160343,
+      "learning_rate": 9.021910535445479e-06,
+      "loss": 0.7824,
+      "step": 674
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6087097450685809,
+      "learning_rate": 9.018672264215808e-06,
+      "loss": 1.0383,
+      "step": 675
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.45132190825902846,
+      "learning_rate": 9.015429224376415e-06,
+      "loss": 0.6118,
+      "step": 676
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5285574540185501,
+      "learning_rate": 9.012181419775535e-06,
+      "loss": 0.9051,
+      "step": 677
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5569589668808301,
+      "learning_rate": 9.008928854267054e-06,
+      "loss": 1.0089,
+      "step": 678
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6073555459103063,
+      "learning_rate": 9.005671531710511e-06,
+      "loss": 0.9617,
+      "step": 679
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5469837338444202,
+      "learning_rate": 9.002409455971087e-06,
+      "loss": 0.9094,
+      "step": 680
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6290167622559641,
+      "learning_rate": 8.999142630919604e-06,
+      "loss": 0.9013,
+      "step": 681
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5905720563988278,
+      "learning_rate": 8.99587106043252e-06,
+      "loss": 0.7181,
+      "step": 682
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5541457487388559,
+      "learning_rate": 8.992594748391926e-06,
+      "loss": 1.1325,
+      "step": 683
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.8117262101417428,
+      "learning_rate": 8.989313698685536e-06,
+      "loss": 0.7656,
+      "step": 684
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.37596808984254504,
+      "learning_rate": 8.986027915206686e-06,
+      "loss": 0.678,
+      "step": 685
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.522421596330282,
+      "learning_rate": 8.982737401854328e-06,
+      "loss": 0.8121,
+      "step": 686
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.560224686799431,
+      "learning_rate": 8.979442162533036e-06,
+      "loss": 1.0024,
+      "step": 687
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.39692915864108647,
+      "learning_rate": 8.976142201152978e-06,
+      "loss": 0.9592,
+      "step": 688
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6437762410861644,
+      "learning_rate": 8.972837521629934e-06,
+      "loss": 0.7804,
+      "step": 689
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5812450588965643,
+      "learning_rate": 8.969528127885281e-06,
+      "loss": 0.9211,
+      "step": 690
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.9724504107073557,
+      "learning_rate": 8.966214023845989e-06,
+      "loss": 0.9491,
+      "step": 691
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5455716379233062,
+      "learning_rate": 8.962895213444618e-06,
+      "loss": 0.9741,
+      "step": 692
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.3692947277372023,
+      "learning_rate": 8.959571700619312e-06,
+      "loss": 0.6581,
+      "step": 693
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.542635297763208,
+      "learning_rate": 8.956243489313795e-06,
+      "loss": 0.9969,
+      "step": 694
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.5424170397025144,
+      "learning_rate": 8.952910583477366e-06,
+      "loss": 0.7535,
+      "step": 695
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6860319334148434,
+      "learning_rate": 8.949572987064896e-06,
+      "loss": 0.8215,
+      "step": 696
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6032832758398157,
+      "learning_rate": 8.946230704036824e-06,
+      "loss": 0.9378,
+      "step": 697
+    },
+    {
+      "epoch": 0.23,
+      "grad_norm": 0.6673759596083534,
+      "learning_rate": 8.942883738359142e-06,
+      "loss": 0.8804,
+      "step": 698
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7016183634425648,
+      "learning_rate": 8.939532094003409e-06,
+      "loss": 1.0123,
+      "step": 699
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5962508505904031,
+      "learning_rate": 8.936175774946726e-06,
+      "loss": 0.8308,
+      "step": 700
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.40820097316765314,
+      "learning_rate": 8.932814785171748e-06,
+      "loss": 1.0301,
+      "step": 701
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5184334174387534,
+      "learning_rate": 8.92944912866667e-06,
+      "loss": 0.7317,
+      "step": 702
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5559644082647761,
+      "learning_rate": 8.926078809425224e-06,
+      "loss": 0.753,
+      "step": 703
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6133269684761555,
+      "learning_rate": 8.922703831446675e-06,
+      "loss": 0.9782,
+      "step": 704
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5811961501196157,
+      "learning_rate": 8.919324198735817e-06,
+      "loss": 0.89,
+      "step": 705
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.629330563658317,
+      "learning_rate": 8.91593991530297e-06,
+      "loss": 0.9347,
+      "step": 706
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.4909771489307399,
+      "learning_rate": 8.912550985163963e-06,
+      "loss": 0.7566,
+      "step": 707
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.635866895310112,
+      "learning_rate": 8.90915741234015e-06,
+      "loss": 0.8639,
+      "step": 708
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5733101913273518,
+      "learning_rate": 8.905759200858388e-06,
+      "loss": 0.9243,
+      "step": 709
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5361017145509727,
+      "learning_rate": 8.902356354751042e-06,
+      "loss": 0.8455,
+      "step": 710
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.448066508191094,
+      "learning_rate": 8.898948878055971e-06,
+      "loss": 0.7253,
+      "step": 711
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.7257100438122076,
+      "learning_rate": 8.895536774816533e-06,
+      "loss": 0.8512,
+      "step": 712
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.35930271545645354,
+      "learning_rate": 8.892120049081577e-06,
+      "loss": 1.0241,
+      "step": 713
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.623849098749688,
+      "learning_rate": 8.888698704905431e-06,
+      "loss": 0.8767,
+      "step": 714
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.556776284122716,
+      "learning_rate": 8.885272746347912e-06,
+      "loss": 0.8533,
+      "step": 715
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6733519942651056,
+      "learning_rate": 8.881842177474304e-06,
+      "loss": 0.9306,
+      "step": 716
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5880433537479856,
+      "learning_rate": 8.878407002355367e-06,
+      "loss": 0.8593,
+      "step": 717
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6259899517055377,
+      "learning_rate": 8.874967225067325e-06,
+      "loss": 0.9912,
+      "step": 718
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5784139317782039,
+      "learning_rate": 8.871522849691863e-06,
+      "loss": 0.8091,
+      "step": 719
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.6091006480779,
+      "learning_rate": 8.868073880316125e-06,
+      "loss": 0.8978,
+      "step": 720
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.49822633673304567,
+      "learning_rate": 8.864620321032698e-06,
+      "loss": 0.6512,
+      "step": 721
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5832870249074441,
+      "learning_rate": 8.861162175939626e-06,
+      "loss": 0.7481,
+      "step": 722
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5957859069580258,
+      "learning_rate": 8.85769944914039e-06,
+      "loss": 0.9629,
+      "step": 723
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.522621918288453,
+      "learning_rate": 8.854232144743905e-06,
+      "loss": 0.8542,
+      "step": 724
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5197359994736168,
+      "learning_rate": 8.85076026686452e-06,
+      "loss": 0.9528,
+      "step": 725
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.35248125683824655,
+      "learning_rate": 8.847283819622015e-06,
+      "loss": 1.0264,
+      "step": 726
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.5716710161245896,
+      "learning_rate": 8.843802807141584e-06,
+      "loss": 0.7972,
+      "step": 727
+    },
+    {
+      "epoch": 0.24,
+      "grad_norm": 0.9189254633657306,
+      "learning_rate": 8.840317233553845e-06,
+      "loss": 0.7224,
+      "step": 728
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7812089625099218,
+      "learning_rate": 8.836827102994824e-06,
+      "loss": 0.8682,
+      "step": 729
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.8508103804195759,
+      "learning_rate": 8.83333241960596e-06,
+      "loss": 0.8797,
+      "step": 730
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5189984344226495,
+      "learning_rate": 8.829833187534086e-06,
+      "loss": 0.7243,
+      "step": 731
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5584119436366025,
+      "learning_rate": 8.82632941093144e-06,
+      "loss": 0.8613,
+      "step": 732
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1524703819338422,
+      "learning_rate": 8.822821093955646e-06,
+      "loss": 1.0308,
+      "step": 733
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5607078805090391,
+      "learning_rate": 8.819308240769726e-06,
+      "loss": 0.8518,
+      "step": 734
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5534896602743887,
+      "learning_rate": 8.81579085554207e-06,
+      "loss": 0.9268,
+      "step": 735
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5201893884590849,
+      "learning_rate": 8.812268942446461e-06,
+      "loss": 1.0568,
+      "step": 736
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.577827743267359,
+      "learning_rate": 8.808742505662045e-06,
+      "loss": 1.0107,
+      "step": 737
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1757090462667155,
+      "learning_rate": 8.805211549373335e-06,
+      "loss": 0.6532,
+      "step": 738
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5659763838207287,
+      "learning_rate": 8.801676077770217e-06,
+      "loss": 0.8408,
+      "step": 739
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.4018473708902758,
+      "learning_rate": 8.798136095047925e-06,
+      "loss": 1.001,
+      "step": 740
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.6299211476101874,
+      "learning_rate": 8.794591605407047e-06,
+      "loss": 0.7876,
+      "step": 741
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.571247171242798,
+      "learning_rate": 8.791042613053527e-06,
+      "loss": 0.9665,
+      "step": 742
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6695647305420357,
+      "learning_rate": 8.787489122198643e-06,
+      "loss": 0.7417,
+      "step": 743
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5962004759020868,
+      "learning_rate": 8.783931137059013e-06,
+      "loss": 0.7952,
+      "step": 744
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5488542441210196,
+      "learning_rate": 8.780368661856592e-06,
+      "loss": 0.9077,
+      "step": 745
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7424551536910317,
+      "learning_rate": 8.776801700818658e-06,
+      "loss": 0.8748,
+      "step": 746
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6118752578648774,
+      "learning_rate": 8.773230258177815e-06,
+      "loss": 0.7282,
+      "step": 747
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7773062173252425,
+      "learning_rate": 8.769654338171986e-06,
+      "loss": 0.8291,
+      "step": 748
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5619666605756146,
+      "learning_rate": 8.766073945044404e-06,
+      "loss": 1.0406,
+      "step": 749
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.4987527201407244,
+      "learning_rate": 8.76248908304361e-06,
+      "loss": 0.9434,
+      "step": 750
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.424305222215896,
+      "learning_rate": 8.758899756423448e-06,
+      "loss": 0.9368,
+      "step": 751
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5130831762372712,
+      "learning_rate": 8.755305969443062e-06,
+      "loss": 0.9131,
+      "step": 752
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.6641729752272197,
+      "learning_rate": 8.751707726366887e-06,
+      "loss": 0.8115,
+      "step": 753
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5066731670848554,
+      "learning_rate": 8.748105031464644e-06,
+      "loss": 0.9587,
+      "step": 754
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.59526934373539,
+      "learning_rate": 8.744497889011344e-06,
+      "loss": 0.5992,
+      "step": 755
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.5079186521914937,
+      "learning_rate": 8.740886303287263e-06,
+      "loss": 0.9528,
+      "step": 756
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 1.1498817169893505,
+      "learning_rate": 8.737270278577959e-06,
+      "loss": 0.8111,
+      "step": 757
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 0.7946795388260054,
+      "learning_rate": 8.733649819174257e-06,
+      "loss": 0.8072,
+      "step": 758
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.513137433976245,
+      "learning_rate": 8.730024929372241e-06,
+      "loss": 1.1289,
+      "step": 759
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5205091514669535,
+      "learning_rate": 8.726395613473251e-06,
+      "loss": 0.9409,
+      "step": 760
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5006562570450211,
+      "learning_rate": 8.722761875783885e-06,
+      "loss": 0.9669,
+      "step": 761
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.612016340027142,
+      "learning_rate": 8.71912372061598e-06,
+      "loss": 0.7532,
+      "step": 762
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.39967887458146256,
+      "learning_rate": 8.715481152286625e-06,
+      "loss": 0.9211,
+      "step": 763
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.44460944577431655,
+      "learning_rate": 8.711834175118132e-06,
+      "loss": 0.6857,
+      "step": 764
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5850737346970263,
+      "learning_rate": 8.708182793438059e-06,
+      "loss": 0.8179,
+      "step": 765
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6506520507490043,
+      "learning_rate": 8.704527011579181e-06,
+      "loss": 0.8462,
+      "step": 766
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7077192935931573,
+      "learning_rate": 8.700866833879497e-06,
+      "loss": 0.8153,
+      "step": 767
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.7316216471567178,
+      "learning_rate": 8.697202264682223e-06,
+      "loss": 0.9473,
+      "step": 768
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5870706024882952,
+      "learning_rate": 8.693533308335786e-06,
+      "loss": 0.8264,
+      "step": 769
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5650042188699568,
+      "learning_rate": 8.689859969193817e-06,
+      "loss": 0.8599,
+      "step": 770
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6245220391944163,
+      "learning_rate": 8.68618225161515e-06,
+      "loss": 0.9824,
+      "step": 771
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.49074424732812755,
+      "learning_rate": 8.682500159963812e-06,
+      "loss": 0.9645,
+      "step": 772
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.46617554688921287,
+      "learning_rate": 8.67881369860902e-06,
+      "loss": 0.6998,
+      "step": 773
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5517190884318978,
+      "learning_rate": 8.675122871925183e-06,
+      "loss": 0.8362,
+      "step": 774
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.4189910035855313,
+      "learning_rate": 8.671427684291882e-06,
+      "loss": 0.9542,
+      "step": 775
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5321406569832858,
+      "learning_rate": 8.667728140093876e-06,
+      "loss": 0.8663,
+      "step": 776
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5889908680446231,
+      "learning_rate": 8.664024243721092e-06,
+      "loss": 0.8632,
+      "step": 777
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.686115900026721,
+      "learning_rate": 8.660315999568623e-06,
+      "loss": 0.9613,
+      "step": 778
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.8015516994921479,
+      "learning_rate": 8.656603412036722e-06,
+      "loss": 0.752,
+      "step": 779
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5331293729655742,
+      "learning_rate": 8.652886485530791e-06,
+      "loss": 1.0221,
+      "step": 780
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.4590567813140477,
+      "learning_rate": 8.64916522446139e-06,
+      "loss": 0.5937,
+      "step": 781
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6348193273790612,
+      "learning_rate": 8.64543963324421e-06,
+      "loss": 0.8211,
+      "step": 782
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.552943477083374,
+      "learning_rate": 8.641709716300092e-06,
+      "loss": 1.0211,
+      "step": 783
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5273367631906747,
+      "learning_rate": 8.637975478055002e-06,
+      "loss": 0.9197,
+      "step": 784
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5574648815275085,
+      "learning_rate": 8.634236922940038e-06,
+      "loss": 0.9664,
+      "step": 785
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6040053362117375,
+      "learning_rate": 8.630494055391418e-06,
+      "loss": 0.7954,
+      "step": 786
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.5716543475371214,
+      "learning_rate": 8.626746879850479e-06,
+      "loss": 0.9636,
+      "step": 787
+    },
+    {
+      "epoch": 0.26,
+      "grad_norm": 0.6054452388421427,
+      "learning_rate": 8.62299540076367e-06,
+      "loss": 0.8162,
+      "step": 788
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.4210463211454794,
+      "learning_rate": 8.619239622582542e-06,
+      "loss": 0.9024,
+      "step": 789
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5211725206975543,
+      "learning_rate": 8.615479549763756e-06,
+      "loss": 0.8241,
+      "step": 790
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5239781036989998,
+      "learning_rate": 8.611715186769063e-06,
+      "loss": 0.8937,
+      "step": 791
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.4044974365338862,
+      "learning_rate": 8.607946538065301e-06,
+      "loss": 0.7644,
+      "step": 792
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5311728848755092,
+      "learning_rate": 8.604173608124405e-06,
+      "loss": 0.8031,
+      "step": 793
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6332881999766978,
+      "learning_rate": 8.600396401423382e-06,
+      "loss": 0.7911,
+      "step": 794
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.49051858511336205,
+      "learning_rate": 8.596614922444313e-06,
+      "loss": 1.0409,
+      "step": 795
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6857106265129076,
+      "learning_rate": 8.592829175674351e-06,
+      "loss": 0.7805,
+      "step": 796
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6729046153481091,
+      "learning_rate": 8.589039165605716e-06,
+      "loss": 0.9944,
+      "step": 797
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.606226874010581,
+      "learning_rate": 8.585244896735683e-06,
+      "loss": 0.883,
+      "step": 798
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5398366899212247,
+      "learning_rate": 8.581446373566581e-06,
+      "loss": 0.8971,
+      "step": 799
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.4437358309144129,
+      "learning_rate": 8.577643600605789e-06,
+      "loss": 0.5843,
+      "step": 800
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5065096949330108,
+      "learning_rate": 8.573836582365727e-06,
+      "loss": 0.9253,
+      "step": 801
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5600806698979193,
+      "learning_rate": 8.570025323363853e-06,
+      "loss": 1.0033,
+      "step": 802
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.668170309966901,
+      "learning_rate": 8.566209828122658e-06,
+      "loss": 0.9455,
+      "step": 803
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5238038092472526,
+      "learning_rate": 8.56239010116966e-06,
+      "loss": 0.9531,
+      "step": 804
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6194977803921942,
+      "learning_rate": 8.558566147037396e-06,
+      "loss": 0.7851,
+      "step": 805
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7476247395041725,
+      "learning_rate": 8.55473797026342e-06,
+      "loss": 0.9536,
+      "step": 806
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5595741384859675,
+      "learning_rate": 8.550905575390301e-06,
+      "loss": 0.9156,
+      "step": 807
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 1.026429715803227,
+      "learning_rate": 8.547068966965605e-06,
+      "loss": 0.644,
+      "step": 808
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5394030618101034,
+      "learning_rate": 8.543228149541905e-06,
+      "loss": 0.9685,
+      "step": 809
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.7220073507663487,
+      "learning_rate": 8.539383127676764e-06,
+      "loss": 0.8908,
+      "step": 810
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6102120096532525,
+      "learning_rate": 8.535533905932739e-06,
+      "loss": 1.0349,
+      "step": 811
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6406144145781414,
+      "learning_rate": 8.531680488877363e-06,
+      "loss": 0.8041,
+      "step": 812
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.4628963559002594,
+      "learning_rate": 8.527822881083157e-06,
+      "loss": 0.9822,
+      "step": 813
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.6851306414755971,
+      "learning_rate": 8.523961087127605e-06,
+      "loss": 0.8499,
+      "step": 814
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5995402958038584,
+      "learning_rate": 8.520095111593163e-06,
+      "loss": 0.8041,
+      "step": 815
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.5112969808921686,
+      "learning_rate": 8.516224959067253e-06,
+      "loss": 0.9278,
+      "step": 816
+    },
+    {
+      "epoch": 0.27,
+      "grad_norm": 0.63193487867569,
+      "learning_rate": 8.512350634142246e-06,
+      "loss": 0.7851,
+      "step": 817
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.473680495246771,
+      "learning_rate": 8.508472141415468e-06,
+      "loss": 0.7975,
+      "step": 818
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6495169967230109,
+      "learning_rate": 8.504589485489187e-06,
+      "loss": 0.7938,
+      "step": 819
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.612725194204173,
+      "learning_rate": 8.50070267097062e-06,
+      "loss": 0.817,
+      "step": 820
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.4977218955511234,
+      "learning_rate": 8.496811702471906e-06,
+      "loss": 0.9027,
+      "step": 821
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6503420502650409,
+      "learning_rate": 8.492916584610124e-06,
+      "loss": 0.7946,
+      "step": 822
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.55667109256535,
+      "learning_rate": 8.489017322007272e-06,
+      "loss": 1.0005,
+      "step": 823
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5406960115066629,
+      "learning_rate": 8.485113919290265e-06,
+      "loss": 0.927,
+      "step": 824
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.4036020837156698,
+      "learning_rate": 8.481206381090934e-06,
+      "loss": 0.9266,
+      "step": 825
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.4891032757799321,
+      "learning_rate": 8.477294712046015e-06,
+      "loss": 0.652,
+      "step": 826
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5907849672700929,
+      "learning_rate": 8.473378916797146e-06,
+      "loss": 0.8623,
+      "step": 827
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.4990926901510241,
+      "learning_rate": 8.46945899999086e-06,
+      "loss": 0.8721,
+      "step": 828
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5256095659295162,
+      "learning_rate": 8.465534966278586e-06,
+      "loss": 0.829,
+      "step": 829
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5166921275893014,
+      "learning_rate": 8.46160682031663e-06,
+      "loss": 1.0279,
+      "step": 830
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5615932821747313,
+      "learning_rate": 8.457674566766185e-06,
+      "loss": 0.863,
+      "step": 831
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5558768179670297,
+      "learning_rate": 8.453738210293316e-06,
+      "loss": 0.7064,
+      "step": 832
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5206175580577042,
+      "learning_rate": 8.449797755568952e-06,
+      "loss": 0.7935,
+      "step": 833
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.3925778814587729,
+      "learning_rate": 8.44585320726889e-06,
+      "loss": 0.6911,
+      "step": 834
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.531331326616415,
+      "learning_rate": 8.441904570073787e-06,
+      "loss": 1.0267,
+      "step": 835
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5326869481281266,
+      "learning_rate": 8.437951848669144e-06,
+      "loss": 0.825,
+      "step": 836
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6207914052749121,
+      "learning_rate": 8.433995047745313e-06,
+      "loss": 0.841,
+      "step": 837
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.42095057547829284,
+      "learning_rate": 8.430034171997487e-06,
+      "loss": 0.9952,
+      "step": 838
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6605401867407642,
+      "learning_rate": 8.426069226125695e-06,
+      "loss": 0.7084,
+      "step": 839
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.8734627360261285,
+      "learning_rate": 8.422100214834791e-06,
+      "loss": 0.9902,
+      "step": 840
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6397754962726314,
+      "learning_rate": 8.41812714283446e-06,
+      "loss": 0.9038,
+      "step": 841
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5128487862402931,
+      "learning_rate": 8.4141500148392e-06,
+      "loss": 1.0241,
+      "step": 842
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5806211373424224,
+      "learning_rate": 8.410168835568324e-06,
+      "loss": 0.7709,
+      "step": 843
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.7358050097985306,
+      "learning_rate": 8.406183609745953e-06,
+      "loss": 0.7611,
+      "step": 844
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.40474112263779355,
+      "learning_rate": 8.402194342101006e-06,
+      "loss": 0.6763,
+      "step": 845
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.5529921677878893,
+      "learning_rate": 8.398201037367202e-06,
+      "loss": 0.9914,
+      "step": 846
+    },
+    {
+      "epoch": 0.28,
+      "grad_norm": 0.6587241279884261,
+      "learning_rate": 8.39420370028305e-06,
+      "loss": 1.0721,
+      "step": 847
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5548507751478083,
+      "learning_rate": 8.390202335591845e-06,
+      "loss": 0.8621,
+      "step": 848
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5369831561321575,
+      "learning_rate": 8.386196948041657e-06,
+      "loss": 0.7983,
+      "step": 849
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.39610739852170496,
+      "learning_rate": 8.382187542385329e-06,
+      "loss": 0.9771,
+      "step": 850
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5529089546621883,
+      "learning_rate": 8.37817412338048e-06,
+      "loss": 0.76,
+      "step": 851
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8419166358050929,
+      "learning_rate": 8.374156695789487e-06,
+      "loss": 0.9503,
+      "step": 852
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5449560208202502,
+      "learning_rate": 8.370135264379475e-06,
+      "loss": 0.7086,
+      "step": 853
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.560948399226522,
+      "learning_rate": 8.366109833922335e-06,
+      "loss": 0.9499,
+      "step": 854
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6417304546298123,
+      "learning_rate": 8.362080409194694e-06,
+      "loss": 0.9093,
+      "step": 855
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5216079325683295,
+      "learning_rate": 8.35804699497792e-06,
+      "loss": 0.8784,
+      "step": 856
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5967904011232459,
+      "learning_rate": 8.354009596058116e-06,
+      "loss": 0.9962,
+      "step": 857
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 2.5300017312592846,
+      "learning_rate": 8.349968217226114e-06,
+      "loss": 0.788,
+      "step": 858
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6292951504032258,
+      "learning_rate": 8.345922863277466e-06,
+      "loss": 1.0093,
+      "step": 859
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.654840885925539,
+      "learning_rate": 8.341873539012443e-06,
+      "loss": 0.7508,
+      "step": 860
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5767264263710964,
+      "learning_rate": 8.337820249236028e-06,
+      "loss": 1.0086,
+      "step": 861
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.4481902331024401,
+      "learning_rate": 8.333762998757908e-06,
+      "loss": 0.6578,
+      "step": 862
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.45207284941848375,
+      "learning_rate": 8.329701792392468e-06,
+      "loss": 0.9165,
+      "step": 863
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.565271136468467,
+      "learning_rate": 8.325636634958793e-06,
+      "loss": 0.9534,
+      "step": 864
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5015507491622027,
+      "learning_rate": 8.32156753128065e-06,
+      "loss": 0.9774,
+      "step": 865
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6005819359948531,
+      "learning_rate": 8.31749448618649e-06,
+      "loss": 0.854,
+      "step": 866
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5587160265359955,
+      "learning_rate": 8.313417504509446e-06,
+      "loss": 0.8952,
+      "step": 867
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5909989921012672,
+      "learning_rate": 8.309336591087318e-06,
+      "loss": 0.726,
+      "step": 868
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5912082911099152,
+      "learning_rate": 8.305251750762568e-06,
+      "loss": 1.0914,
+      "step": 869
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6156141106292572,
+      "learning_rate": 8.301162988382325e-06,
+      "loss": 0.6341,
+      "step": 870
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.4182985224509654,
+      "learning_rate": 8.297070308798367e-06,
+      "loss": 0.7206,
+      "step": 871
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6033243979053533,
+      "learning_rate": 8.292973716867124e-06,
+      "loss": 0.8166,
+      "step": 872
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6025430788207411,
+      "learning_rate": 8.288873217449662e-06,
+      "loss": 0.9983,
+      "step": 873
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.4644321129034185,
+      "learning_rate": 8.284768815411693e-06,
+      "loss": 0.9548,
+      "step": 874
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.6396445172990071,
+      "learning_rate": 8.28066051562355e-06,
+      "loss": 0.7466,
+      "step": 875
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.5147279837000517,
+      "learning_rate": 8.2765483229602e-06,
+      "loss": 1.0441,
+      "step": 876
+    },
+    {
+      "epoch": 0.29,
+      "grad_norm": 0.8216512480499106,
+      "learning_rate": 8.272432242301223e-06,
+      "loss": 0.84,
+      "step": 877
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5186486113643917,
+      "learning_rate": 8.268312278530816e-06,
+      "loss": 0.9241,
+      "step": 878
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.40084048789994675,
+      "learning_rate": 8.26418843653778e-06,
+      "loss": 0.6617,
+      "step": 879
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6833496035272405,
+      "learning_rate": 8.260060721215527e-06,
+      "loss": 0.9978,
+      "step": 880
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5938552363809384,
+      "learning_rate": 8.255929137462049e-06,
+      "loss": 0.7701,
+      "step": 881
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6020858774603756,
+      "learning_rate": 8.251793690179947e-06,
+      "loss": 0.8468,
+      "step": 882
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7104128620010062,
+      "learning_rate": 8.247654384276391e-06,
+      "loss": 0.9332,
+      "step": 883
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.687512268902942,
+      "learning_rate": 8.243511224663141e-06,
+      "loss": 0.8334,
+      "step": 884
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6740374626127225,
+      "learning_rate": 8.239364216256521e-06,
+      "loss": 0.9925,
+      "step": 885
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.3894105519981566,
+      "learning_rate": 8.23521336397743e-06,
+      "loss": 0.9346,
+      "step": 886
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5558930325062331,
+      "learning_rate": 8.231058672751319e-06,
+      "loss": 1.0161,
+      "step": 887
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.38338091763445636,
+      "learning_rate": 8.226900147508205e-06,
+      "loss": 0.7255,
+      "step": 888
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5667691787544652,
+      "learning_rate": 8.222737793182648e-06,
+      "loss": 0.7183,
+      "step": 889
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5486507919663668,
+      "learning_rate": 8.21857161471375e-06,
+      "loss": 0.9409,
+      "step": 890
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6383933008760954,
+      "learning_rate": 8.214401617045155e-06,
+      "loss": 0.9407,
+      "step": 891
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5750483588547717,
+      "learning_rate": 8.210227805125039e-06,
+      "loss": 0.9252,
+      "step": 892
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.66152286488671,
+      "learning_rate": 8.206050183906102e-06,
+      "loss": 0.7448,
+      "step": 893
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6268384659262992,
+      "learning_rate": 8.201868758345561e-06,
+      "loss": 0.8207,
+      "step": 894
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5701617468628584,
+      "learning_rate": 8.197683533405156e-06,
+      "loss": 0.9816,
+      "step": 895
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7074814950422726,
+      "learning_rate": 8.193494514051128e-06,
+      "loss": 0.8136,
+      "step": 896
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.41310580338027714,
+      "learning_rate": 8.189301705254222e-06,
+      "loss": 0.7572,
+      "step": 897
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5985952503762904,
+      "learning_rate": 8.185105111989683e-06,
+      "loss": 0.8,
+      "step": 898
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.4340538183108925,
+      "learning_rate": 8.180904739237241e-06,
+      "loss": 1.0033,
+      "step": 899
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6732784863919254,
+      "learning_rate": 8.176700591981118e-06,
+      "loss": 0.8671,
+      "step": 900
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7813099980061881,
+      "learning_rate": 8.172492675210005e-06,
+      "loss": 0.8126,
+      "step": 901
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5858999759294196,
+      "learning_rate": 8.168280993917078e-06,
+      "loss": 0.952,
+      "step": 902
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.6703300332092279,
+      "learning_rate": 8.164065553099968e-06,
+      "loss": 0.8656,
+      "step": 903
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.7263675875879814,
+      "learning_rate": 8.159846357760776e-06,
+      "loss": 0.9554,
+      "step": 904
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5633150995451018,
+      "learning_rate": 8.155623412906058e-06,
+      "loss": 0.8357,
+      "step": 905
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.5602380449135007,
+      "learning_rate": 8.15139672354681e-06,
+      "loss": 0.8292,
+      "step": 906
+    },
+    {
+      "epoch": 0.3,
+      "grad_norm": 0.4453145705368918,
+      "learning_rate": 8.147166294698482e-06,
+      "loss": 0.6427,
+      "step": 907
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5907458349723954,
+      "learning_rate": 8.142932131380955e-06,
+      "loss": 0.7295,
+      "step": 908
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6035266268319889,
+      "learning_rate": 8.138694238618543e-06,
+      "loss": 1.0466,
+      "step": 909
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5254733094690656,
+      "learning_rate": 8.134452621439988e-06,
+      "loss": 0.8015,
+      "step": 910
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5423924748646352,
+      "learning_rate": 8.130207284878447e-06,
+      "loss": 1.0341,
+      "step": 911
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.34186596418215925,
+      "learning_rate": 8.125958233971494e-06,
+      "loss": 0.972,
+      "step": 912
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.681456237038538,
+      "learning_rate": 8.121705473761107e-06,
+      "loss": 0.7528,
+      "step": 913
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.46539489425498476,
+      "learning_rate": 8.117449009293668e-06,
+      "loss": 0.6676,
+      "step": 914
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7705956713886458,
+      "learning_rate": 8.113188845619957e-06,
+      "loss": 0.8636,
+      "step": 915
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.4946419333580816,
+      "learning_rate": 8.108924987795137e-06,
+      "loss": 0.909,
+      "step": 916
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5717787115328872,
+      "learning_rate": 8.10465744087876e-06,
+      "loss": 0.7144,
+      "step": 917
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5889968377350182,
+      "learning_rate": 8.100386209934754e-06,
+      "loss": 0.9963,
+      "step": 918
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6388165519484676,
+      "learning_rate": 8.096111300031417e-06,
+      "loss": 0.8739,
+      "step": 919
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5682579325592879,
+      "learning_rate": 8.091832716241414e-06,
+      "loss": 0.8061,
+      "step": 920
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5716976993966377,
+      "learning_rate": 8.08755046364177e-06,
+      "loss": 0.8663,
+      "step": 921
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5331971707258996,
+      "learning_rate": 8.083264547313863e-06,
+      "loss": 1.0478,
+      "step": 922
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6036590594042094,
+      "learning_rate": 8.078974972343414e-06,
+      "loss": 0.9749,
+      "step": 923
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5342263506349069,
+      "learning_rate": 8.074681743820496e-06,
+      "loss": 0.6353,
+      "step": 924
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.648212577192026,
+      "learning_rate": 8.070384866839504e-06,
+      "loss": 0.863,
+      "step": 925
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.3838405607983942,
+      "learning_rate": 8.066084346499176e-06,
+      "loss": 0.9787,
+      "step": 926
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.550559760678748,
+      "learning_rate": 8.061780187902562e-06,
+      "loss": 0.7414,
+      "step": 927
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5956760961222445,
+      "learning_rate": 8.057472396157034e-06,
+      "loss": 0.9648,
+      "step": 928
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7048170766212781,
+      "learning_rate": 8.053160976374275e-06,
+      "loss": 0.7348,
+      "step": 929
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.7374038251944228,
+      "learning_rate": 8.048845933670274e-06,
+      "loss": 0.6649,
+      "step": 930
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.570632936244177,
+      "learning_rate": 8.044527273165315e-06,
+      "loss": 0.9684,
+      "step": 931
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5571574804313917,
+      "learning_rate": 8.040204999983981e-06,
+      "loss": 0.8754,
+      "step": 932
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.5388733355003819,
+      "learning_rate": 8.035879119255134e-06,
+      "loss": 0.7158,
+      "step": 933
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.6650969296586566,
+      "learning_rate": 8.031549636111928e-06,
+      "loss": 0.7729,
+      "step": 934
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.4897911640440948,
+      "learning_rate": 8.027216555691781e-06,
+      "loss": 1.1003,
+      "step": 935
+    },
+    {
+      "epoch": 0.31,
+      "grad_norm": 0.623002702398435,
+      "learning_rate": 8.022879883136385e-06,
+      "loss": 0.7742,
+      "step": 936
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.4102136812610758,
+      "learning_rate": 8.01853962359169e-06,
+      "loss": 0.9568,
+      "step": 937
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5688870066771291,
+      "learning_rate": 8.01419578220791e-06,
+      "loss": 0.8319,
+      "step": 938
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6322895539188449,
+      "learning_rate": 8.0098483641395e-06,
+      "loss": 0.8078,
+      "step": 939
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.48641817175208973,
+      "learning_rate": 8.005497374545169e-06,
+      "loss": 0.873,
+      "step": 940
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5895648935971382,
+      "learning_rate": 8.001142818587853e-06,
+      "loss": 0.5596,
+      "step": 941
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.7882915347239713,
+      "learning_rate": 7.99678470143473e-06,
+      "loss": 0.9435,
+      "step": 942
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5946896994875447,
+      "learning_rate": 7.9924230282572e-06,
+      "loss": 0.775,
+      "step": 943
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.570817001947905,
+      "learning_rate": 7.988057804230878e-06,
+      "loss": 0.8885,
+      "step": 944
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5915922365409176,
+      "learning_rate": 7.9836890345356e-06,
+      "loss": 1.1545,
+      "step": 945
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.45631985072655973,
+      "learning_rate": 7.979316724355406e-06,
+      "loss": 0.9168,
+      "step": 946
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5004527473214415,
+      "learning_rate": 7.974940878878537e-06,
+      "loss": 0.9375,
+      "step": 947
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6196554119436458,
+      "learning_rate": 7.970561503297426e-06,
+      "loss": 0.7614,
+      "step": 948
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.42949270171131027,
+      "learning_rate": 7.966178602808703e-06,
+      "loss": 0.9437,
+      "step": 949
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5253888749675317,
+      "learning_rate": 7.96179218261317e-06,
+      "loss": 0.6737,
+      "step": 950
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5585131300698577,
+      "learning_rate": 7.957402247915817e-06,
+      "loss": 0.818,
+      "step": 951
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5409508616915775,
+      "learning_rate": 7.953008803925797e-06,
+      "loss": 0.85,
+      "step": 952
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6026299645738402,
+      "learning_rate": 7.948611855856426e-06,
+      "loss": 0.7709,
+      "step": 953
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5822439361836147,
+      "learning_rate": 7.944211408925184e-06,
+      "loss": 0.9405,
+      "step": 954
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.746915829161732,
+      "learning_rate": 7.939807468353697e-06,
+      "loss": 0.9126,
+      "step": 955
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6118144129484127,
+      "learning_rate": 7.935400039367742e-06,
+      "loss": 0.7822,
+      "step": 956
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5371625871745705,
+      "learning_rate": 7.930989127197232e-06,
+      "loss": 0.9698,
+      "step": 957
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.4911313299117015,
+      "learning_rate": 7.92657473707621e-06,
+      "loss": 0.9371,
+      "step": 958
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5531144368796467,
+      "learning_rate": 7.922156874242857e-06,
+      "loss": 0.6846,
+      "step": 959
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5179859551822413,
+      "learning_rate": 7.91773554393946e-06,
+      "loss": 0.7441,
+      "step": 960
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.3982890974475688,
+      "learning_rate": 7.913310751412431e-06,
+      "loss": 1.0066,
+      "step": 961
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5863311885348068,
+      "learning_rate": 7.908882501912289e-06,
+      "loss": 0.7481,
+      "step": 962
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6187477025609642,
+      "learning_rate": 7.904450800693652e-06,
+      "loss": 0.8454,
+      "step": 963
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.4972954097843683,
+      "learning_rate": 7.900015653015238e-06,
+      "loss": 0.9988,
+      "step": 964
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.6081419344269438,
+      "learning_rate": 7.895577064139847e-06,
+      "loss": 0.7342,
+      "step": 965
+    },
+    {
+      "epoch": 0.32,
+      "grad_norm": 0.5951050318500299,
+      "learning_rate": 7.89113503933437e-06,
+      "loss": 0.9466,
+      "step": 966
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5721307032095788,
+      "learning_rate": 7.886689583869773e-06,
+      "loss": 0.5904,
+      "step": 967
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5509454631874806,
+      "learning_rate": 7.882240703021092e-06,
+      "loss": 0.7949,
+      "step": 968
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.4568386730514535,
+      "learning_rate": 7.877788402067427e-06,
+      "loss": 1.0217,
+      "step": 969
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.7518864312307012,
+      "learning_rate": 7.873332686291939e-06,
+      "loss": 0.9002,
+      "step": 970
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.594696058333896,
+      "learning_rate": 7.868873560981837e-06,
+      "loss": 0.9874,
+      "step": 971
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5707116062404717,
+      "learning_rate": 7.864411031428379e-06,
+      "loss": 0.7956,
+      "step": 972
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.60464701309952,
+      "learning_rate": 7.859945102926862e-06,
+      "loss": 0.9924,
+      "step": 973
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6308745150582424,
+      "learning_rate": 7.855475780776618e-06,
+      "loss": 0.7782,
+      "step": 974
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.37170129459594575,
+      "learning_rate": 7.851003070281e-06,
+      "loss": 0.9324,
+      "step": 975
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.46627601533690577,
+      "learning_rate": 7.846526976747388e-06,
+      "loss": 0.7042,
+      "step": 976
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.48989073715728626,
+      "learning_rate": 7.842047505487173e-06,
+      "loss": 0.6371,
+      "step": 977
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5873060537660271,
+      "learning_rate": 7.837564661815755e-06,
+      "loss": 1.0429,
+      "step": 978
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5065159280020469,
+      "learning_rate": 7.833078451052537e-06,
+      "loss": 0.7961,
+      "step": 979
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6583870944747214,
+      "learning_rate": 7.828588878520911e-06,
+      "loss": 0.9129,
+      "step": 980
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.628277994339506,
+      "learning_rate": 7.824095949548269e-06,
+      "loss": 0.9409,
+      "step": 981
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.9319523168975155,
+      "learning_rate": 7.819599669465979e-06,
+      "loss": 0.8586,
+      "step": 982
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5394637261604093,
+      "learning_rate": 7.815100043609385e-06,
+      "loss": 1.017,
+      "step": 983
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5893696534517224,
+      "learning_rate": 7.810597077317802e-06,
+      "loss": 0.8729,
+      "step": 984
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5294439728327826,
+      "learning_rate": 7.806090775934513e-06,
+      "loss": 0.9,
+      "step": 985
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.3360130854815798,
+      "learning_rate": 7.801581144806752e-06,
+      "loss": 0.7996,
+      "step": 986
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6429695207703293,
+      "learning_rate": 7.797068189285705e-06,
+      "loss": 0.8249,
+      "step": 987
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.597263734079363,
+      "learning_rate": 7.792551914726509e-06,
+      "loss": 0.9295,
+      "step": 988
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5346852960060344,
+      "learning_rate": 7.788032326488233e-06,
+      "loss": 0.8797,
+      "step": 989
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5066243641371182,
+      "learning_rate": 7.78350942993388e-06,
+      "loss": 0.9222,
+      "step": 990
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6606388444684012,
+      "learning_rate": 7.778983230430382e-06,
+      "loss": 0.8201,
+      "step": 991
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6430544032783221,
+      "learning_rate": 7.77445373334858e-06,
+      "loss": 0.9855,
+      "step": 992
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5228331512591384,
+      "learning_rate": 7.769920944063244e-06,
+      "loss": 0.9464,
+      "step": 993
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.4818726370023612,
+      "learning_rate": 7.765384867953038e-06,
+      "loss": 0.6251,
+      "step": 994
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.5387014397274433,
+      "learning_rate": 7.76084551040053e-06,
+      "loss": 0.8868,
+      "step": 995
+    },
+    {
+      "epoch": 0.33,
+      "grad_norm": 0.6492543300465846,
+      "learning_rate": 7.75630287679218e-06,
+      "loss": 0.901,
+      "step": 996
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6452710521124598,
+      "learning_rate": 7.751756972518341e-06,
+      "loss": 0.9841,
+      "step": 997
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.9011413597561838,
+      "learning_rate": 7.74720780297324e-06,
+      "loss": 0.7872,
+      "step": 998
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.36167283150080254,
+      "learning_rate": 7.742655373554982e-06,
+      "loss": 0.9618,
+      "step": 999
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6268557038979575,
+      "learning_rate": 7.73809968966554e-06,
+      "loss": 0.8945,
+      "step": 1000
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.7206868704421296,
+      "learning_rate": 7.733540756710752e-06,
+      "loss": 0.7896,
+      "step": 1001
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5171799054021541,
+      "learning_rate": 7.728978580100304e-06,
+      "loss": 0.8965,
+      "step": 1002
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.514913216312941,
+      "learning_rate": 7.72441316524774e-06,
+      "loss": 0.569,
+      "step": 1003
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4965550280007811,
+      "learning_rate": 7.719844517570437e-06,
+      "loss": 0.9708,
+      "step": 1004
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5787009904944812,
+      "learning_rate": 7.715272642489612e-06,
+      "loss": 0.8198,
+      "step": 1005
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.584753473071653,
+      "learning_rate": 7.71069754543032e-06,
+      "loss": 0.8962,
+      "step": 1006
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5458549245754811,
+      "learning_rate": 7.706119231821423e-06,
+      "loss": 0.8086,
+      "step": 1007
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.65270354488186,
+      "learning_rate": 7.701537707095615e-06,
+      "loss": 0.7903,
+      "step": 1008
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5024630244718226,
+      "learning_rate": 7.69695297668939e-06,
+      "loss": 0.9895,
+      "step": 1009
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.3956585350291841,
+      "learning_rate": 7.692365046043053e-06,
+      "loss": 0.9752,
+      "step": 1010
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5495183407908699,
+      "learning_rate": 7.687773920600699e-06,
+      "loss": 0.9356,
+      "step": 1011
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5263010288802161,
+      "learning_rate": 7.683179605810221e-06,
+      "loss": 0.5007,
+      "step": 1012
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5957374587541919,
+      "learning_rate": 7.678582107123295e-06,
+      "loss": 0.8719,
+      "step": 1013
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5168831300505938,
+      "learning_rate": 7.673981429995372e-06,
+      "loss": 0.8922,
+      "step": 1014
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5834821837039692,
+      "learning_rate": 7.669377579885674e-06,
+      "loss": 0.813,
+      "step": 1015
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.46944296240544386,
+      "learning_rate": 7.664770562257197e-06,
+      "loss": 1.0103,
+      "step": 1016
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6300994219902402,
+      "learning_rate": 7.660160382576683e-06,
+      "loss": 0.8594,
+      "step": 1017
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.6769638639618092,
+      "learning_rate": 7.655547046314635e-06,
+      "loss": 0.7048,
+      "step": 1018
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5610160785079183,
+      "learning_rate": 7.650930558945299e-06,
+      "loss": 0.7871,
+      "step": 1019
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.4064545065316439,
+      "learning_rate": 7.646310925946659e-06,
+      "loss": 0.7113,
+      "step": 1020
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.589079903347669,
+      "learning_rate": 7.641688152800433e-06,
+      "loss": 1.043,
+      "step": 1021
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.5861049061013667,
+      "learning_rate": 7.637062244992065e-06,
+      "loss": 0.8353,
+      "step": 1022
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.591006320512785,
+      "learning_rate": 7.63243320801072e-06,
+      "loss": 0.9125,
+      "step": 1023
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.40821217675440935,
+      "learning_rate": 7.62780104734927e-06,
+      "loss": 0.9149,
+      "step": 1024
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 1.2561369163963982,
+      "learning_rate": 7.6231657685043036e-06,
+      "loss": 0.664,
+      "step": 1025
+    },
+    {
+      "epoch": 0.34,
+      "grad_norm": 0.563474726140817,
+      "learning_rate": 7.6185273769761015e-06,
+      "loss": 1.018,
+      "step": 1026
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5194338230924535,
+      "learning_rate": 7.61388587826864e-06,
+      "loss": 0.8183,
+      "step": 1027
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6338192453677742,
+      "learning_rate": 7.609241277889583e-06,
+      "loss": 0.9957,
+      "step": 1028
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5515637027776569,
+      "learning_rate": 7.6045935813502756e-06,
+      "loss": 0.8126,
+      "step": 1029
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5598327707734606,
+      "learning_rate": 7.599942794165738e-06,
+      "loss": 0.9287,
+      "step": 1030
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.39070076215284705,
+      "learning_rate": 7.595288921854651e-06,
+      "loss": 0.5377,
+      "step": 1031
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5300086580373395,
+      "learning_rate": 7.590631969939364e-06,
+      "loss": 0.9065,
+      "step": 1032
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.625366996611634,
+      "learning_rate": 7.58597194394588e-06,
+      "loss": 1.0249,
+      "step": 1033
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5795310667908754,
+      "learning_rate": 7.581308849403843e-06,
+      "loss": 0.7904,
+      "step": 1034
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5332714333788107,
+      "learning_rate": 7.5766426918465455e-06,
+      "loss": 0.9404,
+      "step": 1035
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.39429906511701823,
+      "learning_rate": 7.5719734768109125e-06,
+      "loss": 0.9183,
+      "step": 1036
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5293603777638457,
+      "learning_rate": 7.567301209837493e-06,
+      "loss": 0.8163,
+      "step": 1037
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5975326138393592,
+      "learning_rate": 7.5626258964704634e-06,
+      "loss": 0.936,
+      "step": 1038
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.4572075348078856,
+      "learning_rate": 7.557947542257613e-06,
+      "loss": 0.7007,
+      "step": 1039
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.668650621679814,
+      "learning_rate": 7.553266152750338e-06,
+      "loss": 0.9201,
+      "step": 1040
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9322252505623536,
+      "learning_rate": 7.548581733503636e-06,
+      "loss": 0.9334,
+      "step": 1041
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5553565769041645,
+      "learning_rate": 7.5438942900761035e-06,
+      "loss": 1.016,
+      "step": 1042
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5903455546807612,
+      "learning_rate": 7.539203828029919e-06,
+      "loss": 0.8545,
+      "step": 1043
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.549946240044186,
+      "learning_rate": 7.534510352930853e-06,
+      "loss": 0.7687,
+      "step": 1044
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5509971833474584,
+      "learning_rate": 7.529813870348238e-06,
+      "loss": 0.9725,
+      "step": 1045
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.50065528827927,
+      "learning_rate": 7.525114385854988e-06,
+      "loss": 0.7156,
+      "step": 1046
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6057514854994172,
+      "learning_rate": 7.520411905027571e-06,
+      "loss": 1.0472,
+      "step": 1047
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.33864483662172956,
+      "learning_rate": 7.515706433446013e-06,
+      "loss": 0.8004,
+      "step": 1048
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5818891232743176,
+      "learning_rate": 7.51099797669389e-06,
+      "loss": 0.8945,
+      "step": 1049
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 1.1175101030306798,
+      "learning_rate": 7.506286540358318e-06,
+      "loss": 0.7814,
+      "step": 1050
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.9151920100739886,
+      "learning_rate": 7.50157213002995e-06,
+      "loss": 0.9311,
+      "step": 1051
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6095990873942779,
+      "learning_rate": 7.4968547513029685e-06,
+      "loss": 0.8317,
+      "step": 1052
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.46976918272054524,
+      "learning_rate": 7.492134409775077e-06,
+      "loss": 0.8393,
+      "step": 1053
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.6811523390205392,
+      "learning_rate": 7.4874111110474955e-06,
+      "loss": 0.9254,
+      "step": 1054
+    },
+    {
+      "epoch": 0.35,
+      "grad_norm": 0.5313333330902266,
+      "learning_rate": 7.482684860724951e-06,
+      "loss": 0.9316,
+      "step": 1055
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.37561918479871825,
+      "learning_rate": 7.477955664415678e-06,
+      "loss": 0.5186,
+      "step": 1056
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.4584492936530272,
+      "learning_rate": 7.473223527731401e-06,
+      "loss": 0.9289,
+      "step": 1057
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5924327140049981,
+      "learning_rate": 7.4684884562873375e-06,
+      "loss": 0.8275,
+      "step": 1058
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5069297840063071,
+      "learning_rate": 7.463750455702185e-06,
+      "loss": 0.9215,
+      "step": 1059
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.4092207352844884,
+      "learning_rate": 7.459009531598118e-06,
+      "loss": 0.9102,
+      "step": 1060
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6131960535642291,
+      "learning_rate": 7.4542656896007804e-06,
+      "loss": 0.7462,
+      "step": 1061
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6551517148946325,
+      "learning_rate": 7.449518935339276e-06,
+      "loss": 1.0436,
+      "step": 1062
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.7922858306756169,
+      "learning_rate": 7.444769274446168e-06,
+      "loss": 0.7891,
+      "step": 1063
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.4584943517855498,
+      "learning_rate": 7.440016712557468e-06,
+      "loss": 0.8868,
+      "step": 1064
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.4025721638224008,
+      "learning_rate": 7.435261255312626e-06,
+      "loss": 0.6219,
+      "step": 1065
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.48331600801707003,
+      "learning_rate": 7.430502908354532e-06,
+      "loss": 0.9653,
+      "step": 1066
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5633362509883766,
+      "learning_rate": 7.4257416773295045e-06,
+      "loss": 0.7364,
+      "step": 1067
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 1.5006489599435686,
+      "learning_rate": 7.420977567887281e-06,
+      "loss": 0.7923,
+      "step": 1068
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5604002036073501,
+      "learning_rate": 7.416210585681019e-06,
+      "loss": 0.9895,
+      "step": 1069
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5928254933029005,
+      "learning_rate": 7.411440736367281e-06,
+      "loss": 0.8274,
+      "step": 1070
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.623424492735748,
+      "learning_rate": 7.406668025606035e-06,
+      "loss": 0.9246,
+      "step": 1071
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.39477759731243867,
+      "learning_rate": 7.401892459060641e-06,
+      "loss": 0.8785,
+      "step": 1072
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7279368964016807,
+      "learning_rate": 7.397114042397852e-06,
+      "loss": 0.9743,
+      "step": 1073
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.36512752932775616,
+      "learning_rate": 7.392332781287798e-06,
+      "loss": 0.7648,
+      "step": 1074
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6414406994249823,
+      "learning_rate": 7.387548681403986e-06,
+      "loss": 0.6624,
+      "step": 1075
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.584795648913772,
+      "learning_rate": 7.382761748423296e-06,
+      "loss": 0.9671,
+      "step": 1076
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.567397449988564,
+      "learning_rate": 7.377971988025964e-06,
+      "loss": 0.9957,
+      "step": 1077
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5987674972476548,
+      "learning_rate": 7.373179405895582e-06,
+      "loss": 0.8142,
+      "step": 1078
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5077512311893734,
+      "learning_rate": 7.368384007719093e-06,
+      "loss": 0.8089,
+      "step": 1079
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7078840080404868,
+      "learning_rate": 7.36358579918678e-06,
+      "loss": 0.8502,
+      "step": 1080
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.5266640259624071,
+      "learning_rate": 7.35878478599226e-06,
+      "loss": 0.9189,
+      "step": 1081
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.7097336048877754,
+      "learning_rate": 7.353980973832479e-06,
+      "loss": 0.7917,
+      "step": 1082
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.4003881609447993,
+      "learning_rate": 7.349174368407703e-06,
+      "loss": 0.7417,
+      "step": 1083
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.6608085666314166,
+      "learning_rate": 7.3443649754215175e-06,
+      "loss": 0.8227,
+      "step": 1084
+    },
+    {
+      "epoch": 0.36,
+      "grad_norm": 0.3758752498030171,
+      "learning_rate": 7.339552800580807e-06,
+      "loss": 1.0013,
+      "step": 1085
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6278997002779944,
+      "learning_rate": 7.3347378495957655e-06,
+      "loss": 0.7946,
+      "step": 1086
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5015133660658064,
+      "learning_rate": 7.329920128179875e-06,
+      "loss": 0.8453,
+      "step": 1087
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5559606982002677,
+      "learning_rate": 7.325099642049909e-06,
+      "loss": 0.9728,
+      "step": 1088
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.7131878516286082,
+      "learning_rate": 7.320276396925921e-06,
+      "loss": 0.8512,
+      "step": 1089
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5416883131718453,
+      "learning_rate": 7.3154503985312366e-06,
+      "loss": 0.9321,
+      "step": 1090
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5489992931384962,
+      "learning_rate": 7.310621652592449e-06,
+      "loss": 0.8449,
+      "step": 1091
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5221565726618351,
+      "learning_rate": 7.305790164839411e-06,
+      "loss": 0.946,
+      "step": 1092
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.4290290839735477,
+      "learning_rate": 7.300955941005232e-06,
+      "loss": 0.502,
+      "step": 1093
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5260178467040457,
+      "learning_rate": 7.296118986826266e-06,
+      "loss": 0.7985,
+      "step": 1094
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5184849063815847,
+      "learning_rate": 7.291279308042105e-06,
+      "loss": 0.9387,
+      "step": 1095
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5181925403310671,
+      "learning_rate": 7.286436910395579e-06,
+      "loss": 0.8288,
+      "step": 1096
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5109635307721284,
+      "learning_rate": 7.281591799632738e-06,
+      "loss": 1.0424,
+      "step": 1097
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.38827489949445787,
+      "learning_rate": 7.276743981502856e-06,
+      "loss": 0.9603,
+      "step": 1098
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5520892119575533,
+      "learning_rate": 7.271893461758418e-06,
+      "loss": 0.8737,
+      "step": 1099
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5099760570921432,
+      "learning_rate": 7.267040246155115e-06,
+      "loss": 0.5632,
+      "step": 1100
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6253076957655209,
+      "learning_rate": 7.262184340451838e-06,
+      "loss": 0.8633,
+      "step": 1101
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.4895104562601699,
+      "learning_rate": 7.2573257504106665e-06,
+      "loss": 0.8323,
+      "step": 1102
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5061385810713125,
+      "learning_rate": 7.252464481796869e-06,
+      "loss": 0.7636,
+      "step": 1103
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5089368253913088,
+      "learning_rate": 7.247600540378893e-06,
+      "loss": 1.026,
+      "step": 1104
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.609739883952128,
+      "learning_rate": 7.242733931928352e-06,
+      "loss": 0.8061,
+      "step": 1105
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.60873741931676,
+      "learning_rate": 7.237864662220032e-06,
+      "loss": 0.7971,
+      "step": 1106
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.6187567025942676,
+      "learning_rate": 7.232992737031871e-06,
+      "loss": 0.8741,
+      "step": 1107
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.457714209932252,
+      "learning_rate": 7.2281181621449594e-06,
+      "loss": 1.0314,
+      "step": 1108
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5862632101578414,
+      "learning_rate": 7.223240943343533e-06,
+      "loss": 0.6918,
+      "step": 1109
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5746871719059815,
+      "learning_rate": 7.2183610864149655e-06,
+      "loss": 0.7914,
+      "step": 1110
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5050942177239504,
+      "learning_rate": 7.213478597149758e-06,
+      "loss": 0.9998,
+      "step": 1111
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.32842825390645225,
+      "learning_rate": 7.208593481341536e-06,
+      "loss": 0.8951,
+      "step": 1112
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5224993553051833,
+      "learning_rate": 7.203705744787044e-06,
+      "loss": 0.7368,
+      "step": 1113
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5973652829111198,
+      "learning_rate": 7.198815393286136e-06,
+      "loss": 1.0312,
+      "step": 1114
+    },
+    {
+      "epoch": 0.37,
+      "grad_norm": 0.5527814476087755,
+      "learning_rate": 7.193922432641767e-06,
+      "loss": 0.6769,
+      "step": 1115
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5335879560487848,
+      "learning_rate": 7.1890268686599875e-06,
+      "loss": 0.8358,
+      "step": 1116
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5224928921127848,
+      "learning_rate": 7.18412870714994e-06,
+      "loss": 0.8017,
+      "step": 1117
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6124614522298713,
+      "learning_rate": 7.179227953923848e-06,
+      "loss": 0.9035,
+      "step": 1118
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5499377417407644,
+      "learning_rate": 7.1743246147970095e-06,
+      "loss": 0.6473,
+      "step": 1119
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.49247233341678726,
+      "learning_rate": 7.169418695587791e-06,
+      "loss": 0.8862,
+      "step": 1120
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.4743156754027042,
+      "learning_rate": 7.1645102021176225e-06,
+      "loss": 1.0672,
+      "step": 1121
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6252290087433129,
+      "learning_rate": 7.1595991402109865e-06,
+      "loss": 0.7857,
+      "step": 1122
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.35219321017852084,
+      "learning_rate": 7.154685515695415e-06,
+      "loss": 0.9344,
+      "step": 1123
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5059460382727503,
+      "learning_rate": 7.14976933440148e-06,
+      "loss": 0.8596,
+      "step": 1124
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5902056623836656,
+      "learning_rate": 7.144850602162785e-06,
+      "loss": 0.7351,
+      "step": 1125
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 1.4370719288486014,
+      "learning_rate": 7.139929324815965e-06,
+      "loss": 0.8809,
+      "step": 1126
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.47510664267726815,
+      "learning_rate": 7.135005508200673e-06,
+      "loss": 0.5632,
+      "step": 1127
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.45029784379861976,
+      "learning_rate": 7.130079158159573e-06,
+      "loss": 0.9397,
+      "step": 1128
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6459645380907941,
+      "learning_rate": 7.12515028053834e-06,
+      "loss": 0.7859,
+      "step": 1129
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5384436760106294,
+      "learning_rate": 7.120218881185644e-06,
+      "loss": 0.997,
+      "step": 1130
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5394925991989559,
+      "learning_rate": 7.115284965953147e-06,
+      "loss": 0.971,
+      "step": 1131
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.4317379807622066,
+      "learning_rate": 7.110348540695501e-06,
+      "loss": 0.8781,
+      "step": 1132
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.4761419310125354,
+      "learning_rate": 7.105409611270332e-06,
+      "loss": 0.933,
+      "step": 1133
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.6715319915382009,
+      "learning_rate": 7.100468183538241e-06,
+      "loss": 0.7524,
+      "step": 1134
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.3461629433226296,
+      "learning_rate": 7.095524263362787e-06,
+      "loss": 0.9603,
+      "step": 1135
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.45843143360302024,
+      "learning_rate": 7.090577856610497e-06,
+      "loss": 0.5007,
+      "step": 1136
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.49640632783113486,
+      "learning_rate": 7.0856289691508395e-06,
+      "loss": 0.827,
+      "step": 1137
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5385219907698342,
+      "learning_rate": 7.08067760685623e-06,
+      "loss": 0.8287,
+      "step": 1138
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5698237673667036,
+      "learning_rate": 7.07572377560202e-06,
+      "loss": 0.7261,
+      "step": 1139
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5813810027772864,
+      "learning_rate": 7.070767481266493e-06,
+      "loss": 1.0231,
+      "step": 1140
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5919608478863481,
+      "learning_rate": 7.065808729730853e-06,
+      "loss": 0.8854,
+      "step": 1141
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5295424001914607,
+      "learning_rate": 7.0608475268792186e-06,
+      "loss": 0.8728,
+      "step": 1142
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.4705941323230157,
+      "learning_rate": 7.055883878598622e-06,
+      "loss": 0.8234,
+      "step": 1143
+    },
+    {
+      "epoch": 0.38,
+      "grad_norm": 0.5333683642857805,
+      "learning_rate": 7.050917790778992e-06,
+      "loss": 0.9481,
+      "step": 1144
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4392390698663535,
+      "learning_rate": 7.045949269313153e-06,
+      "loss": 0.6012,
+      "step": 1145
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5330618186302982,
+      "learning_rate": 7.040978320096819e-06,
+      "loss": 0.8035,
+      "step": 1146
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4176689884221389,
+      "learning_rate": 7.036004949028587e-06,
+      "loss": 0.9634,
+      "step": 1147
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.608471264044503,
+      "learning_rate": 7.031029162009921e-06,
+      "loss": 0.7528,
+      "step": 1148
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.589191446906344,
+      "learning_rate": 7.026050964945158e-06,
+      "loss": 0.8818,
+      "step": 1149
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.47379192944595994,
+      "learning_rate": 7.021070363741492e-06,
+      "loss": 0.9263,
+      "step": 1150
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.6272402188583187,
+      "learning_rate": 7.016087364308971e-06,
+      "loss": 0.7436,
+      "step": 1151
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.4555966899339485,
+      "learning_rate": 7.0111019725604856e-06,
+      "loss": 0.9394,
+      "step": 1152
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5703490596117441,
+      "learning_rate": 7.006114194411772e-06,
+      "loss": 0.6058,
+      "step": 1153
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5557275316261898,
+      "learning_rate": 7.00112403578139e-06,
+      "loss": 0.9697,
+      "step": 1154
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.44557920795758643,
+      "learning_rate": 6.996131502590729e-06,
+      "loss": 0.9283,
+      "step": 1155
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.571509847100694,
+      "learning_rate": 6.991136600763997e-06,
+      "loss": 0.8868,
+      "step": 1156
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5451919876369528,
+      "learning_rate": 6.986139336228208e-06,
+      "loss": 0.9684,
+      "step": 1157
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5854047033287596,
+      "learning_rate": 6.9811397149131835e-06,
+      "loss": 0.8318,
+      "step": 1158
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.7126197252622675,
+      "learning_rate": 6.976137742751538e-06,
+      "loss": 0.9429,
+      "step": 1159
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.3799450767489363,
+      "learning_rate": 6.971133425678682e-06,
+      "loss": 0.9242,
+      "step": 1160
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5458240195414593,
+      "learning_rate": 6.9661267696328015e-06,
+      "loss": 0.9077,
+      "step": 1161
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.44644616494105016,
+      "learning_rate": 6.961117780554862e-06,
+      "loss": 0.7207,
+      "step": 1162
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.598659762324647,
+      "learning_rate": 6.956106464388596e-06,
+      "loss": 0.5789,
+      "step": 1163
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5248117524584618,
+      "learning_rate": 6.9510928270804975e-06,
+      "loss": 0.9959,
+      "step": 1164
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.49652719694535974,
+      "learning_rate": 6.9460768745798144e-06,
+      "loss": 0.7946,
+      "step": 1165
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5561257824068346,
+      "learning_rate": 6.941058612838544e-06,
+      "loss": 0.9345,
+      "step": 1166
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5198620701079795,
+      "learning_rate": 6.936038047811421e-06,
+      "loss": 0.9261,
+      "step": 1167
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8329637578261035,
+      "learning_rate": 6.931015185455915e-06,
+      "loss": 0.8659,
+      "step": 1168
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.502982353788205,
+      "learning_rate": 6.925990031732218e-06,
+      "loss": 0.9692,
+      "step": 1169
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 1.1475283912646639,
+      "learning_rate": 6.9209625926032485e-06,
+      "loss": 0.8699,
+      "step": 1170
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.49185518178453325,
+      "learning_rate": 6.91593287403463e-06,
+      "loss": 0.9156,
+      "step": 1171
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.3380663767229636,
+      "learning_rate": 6.910900881994692e-06,
+      "loss": 0.7858,
+      "step": 1172
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.5867430292541246,
+      "learning_rate": 6.905866622454466e-06,
+      "loss": 1.0,
+      "step": 1173
+    },
+    {
+      "epoch": 0.39,
+      "grad_norm": 0.8280673173566504,
+      "learning_rate": 6.900830101387667e-06,
+      "loss": 0.703,
+      "step": 1174
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5005266064399941,
+      "learning_rate": 6.895791324770702e-06,
+      "loss": 0.8717,
+      "step": 1175
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5245427106190206,
+      "learning_rate": 6.890750298582647e-06,
+      "loss": 0.8653,
+      "step": 1176
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6647250658867142,
+      "learning_rate": 6.885707028805253e-06,
+      "loss": 0.8869,
+      "step": 1177
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 1.2461188831964047,
+      "learning_rate": 6.8806615214229275e-06,
+      "loss": 0.9186,
+      "step": 1178
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5584637973194874,
+      "learning_rate": 6.875613782422737e-06,
+      "loss": 0.9423,
+      "step": 1179
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4091953154804003,
+      "learning_rate": 6.870563817794398e-06,
+      "loss": 0.6589,
+      "step": 1180
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5836691464048496,
+      "learning_rate": 6.865511633530263e-06,
+      "loss": 0.8404,
+      "step": 1181
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5947683677032507,
+      "learning_rate": 6.860457235625322e-06,
+      "loss": 0.8339,
+      "step": 1182
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 2.003949850446798,
+      "learning_rate": 6.855400630077189e-06,
+      "loss": 0.925,
+      "step": 1183
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.544038593015195,
+      "learning_rate": 6.8503418228861e-06,
+      "loss": 0.7418,
+      "step": 1184
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.34674764406553993,
+      "learning_rate": 6.845280820054904e-06,
+      "loss": 0.9727,
+      "step": 1185
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5657955169562142,
+      "learning_rate": 6.840217627589052e-06,
+      "loss": 0.8018,
+      "step": 1186
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.43914904852459224,
+      "learning_rate": 6.835152251496596e-06,
+      "loss": 0.6802,
+      "step": 1187
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.7266533928819141,
+      "learning_rate": 6.8300846977881785e-06,
+      "loss": 0.9821,
+      "step": 1188
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4474729408065013,
+      "learning_rate": 6.825014972477024e-06,
+      "loss": 0.5023,
+      "step": 1189
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.465293871826588,
+      "learning_rate": 6.819943081578939e-06,
+      "loss": 0.9561,
+      "step": 1190
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5563353753035327,
+      "learning_rate": 6.814869031112294e-06,
+      "loss": 0.8123,
+      "step": 1191
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5560550094973613,
+      "learning_rate": 6.809792827098025e-06,
+      "loss": 0.7914,
+      "step": 1192
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.478540098256999,
+      "learning_rate": 6.804714475559624e-06,
+      "loss": 0.8291,
+      "step": 1193
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.618034664869094,
+      "learning_rate": 6.799633982523128e-06,
+      "loss": 0.8969,
+      "step": 1194
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.47856380148050315,
+      "learning_rate": 6.794551354017118e-06,
+      "loss": 0.9558,
+      "step": 1195
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.3560759471901381,
+      "learning_rate": 6.7894665960727105e-06,
+      "loss": 0.946,
+      "step": 1196
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5572432563374421,
+      "learning_rate": 6.784379714723544e-06,
+      "loss": 0.9262,
+      "step": 1197
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6456925053786322,
+      "learning_rate": 6.7792907160057796e-06,
+      "loss": 0.549,
+      "step": 1198
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.5259674798174586,
+      "learning_rate": 6.774199605958091e-06,
+      "loss": 0.8322,
+      "step": 1199
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.45535780385896724,
+      "learning_rate": 6.7691063906216585e-06,
+      "loss": 0.8115,
+      "step": 1200
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.49316628484336933,
+      "learning_rate": 6.764011076040157e-06,
+      "loss": 0.9714,
+      "step": 1201
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4574959119415886,
+      "learning_rate": 6.758913668259753e-06,
+      "loss": 0.9347,
+      "step": 1202
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.6082873001782719,
+      "learning_rate": 6.7538141733291e-06,
+      "loss": 0.8679,
+      "step": 1203
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.49075291472466454,
+      "learning_rate": 6.748712597299327e-06,
+      "loss": 0.8014,
+      "step": 1204
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5128769037995072,
+      "learning_rate": 6.743608946224029e-06,
+      "loss": 0.6747,
+      "step": 1205
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6807914725525773,
+      "learning_rate": 6.738503226159269e-06,
+      "loss": 0.6579,
+      "step": 1206
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6008006618264438,
+      "learning_rate": 6.73339544316356e-06,
+      "loss": 0.9733,
+      "step": 1207
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5541893050222473,
+      "learning_rate": 6.728285603297865e-06,
+      "loss": 0.872,
+      "step": 1208
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6136148775971656,
+      "learning_rate": 6.723173712625588e-06,
+      "loss": 0.94,
+      "step": 1209
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4659792258813381,
+      "learning_rate": 6.7180597772125665e-06,
+      "loss": 0.883,
+      "step": 1210
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6142169562288466,
+      "learning_rate": 6.712943803127063e-06,
+      "loss": 0.6336,
+      "step": 1211
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5378424223130638,
+      "learning_rate": 6.707825796439761e-06,
+      "loss": 1.0395,
+      "step": 1212
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5255335945660673,
+      "learning_rate": 6.702705763223756e-06,
+      "loss": 0.803,
+      "step": 1213
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5559085783795369,
+      "learning_rate": 6.697583709554545e-06,
+      "loss": 1.0283,
+      "step": 1214
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5431630884842857,
+      "learning_rate": 6.692459641510025e-06,
+      "loss": 0.7814,
+      "step": 1215
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4860964254851905,
+      "learning_rate": 6.687333565170484e-06,
+      "loss": 0.6574,
+      "step": 1216
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6898531161422031,
+      "learning_rate": 6.682205486618592e-06,
+      "loss": 0.6778,
+      "step": 1217
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6085657418665649,
+      "learning_rate": 6.677075411939396e-06,
+      "loss": 0.9274,
+      "step": 1218
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5591949439863059,
+      "learning_rate": 6.671943347220308e-06,
+      "loss": 0.9448,
+      "step": 1219
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5430773382706926,
+      "learning_rate": 6.666809298551105e-06,
+      "loss": 0.8514,
+      "step": 1220
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7716090405680918,
+      "learning_rate": 6.661673272023917e-06,
+      "loss": 0.9307,
+      "step": 1221
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.35787829949209604,
+      "learning_rate": 6.656535273733222e-06,
+      "loss": 0.8939,
+      "step": 1222
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5948060821776634,
+      "learning_rate": 6.651395309775837e-06,
+      "loss": 0.9216,
+      "step": 1223
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5689827668540052,
+      "learning_rate": 6.646253386250909e-06,
+      "loss": 0.7841,
+      "step": 1224
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.3606010515959767,
+      "learning_rate": 6.641109509259916e-06,
+      "loss": 0.6898,
+      "step": 1225
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6009811680765854,
+      "learning_rate": 6.635963684906646e-06,
+      "loss": 0.9586,
+      "step": 1226
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6293269080216974,
+      "learning_rate": 6.630815919297205e-06,
+      "loss": 0.8171,
+      "step": 1227
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.44821401954557505,
+      "learning_rate": 6.625666218540001e-06,
+      "loss": 1.0311,
+      "step": 1228
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.7019649875287429,
+      "learning_rate": 6.620514588745736e-06,
+      "loss": 0.8325,
+      "step": 1229
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5182896982544559,
+      "learning_rate": 6.6153610360274014e-06,
+      "loss": 0.7918,
+      "step": 1230
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.6316034677776156,
+      "learning_rate": 6.610205566500272e-06,
+      "loss": 0.9007,
+      "step": 1231
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.5140621186379946,
+      "learning_rate": 6.605048186281895e-06,
+      "loss": 0.7742,
+      "step": 1232
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.4095569005346679,
+      "learning_rate": 6.5998889014920884e-06,
+      "loss": 0.7392,
+      "step": 1233
+    },
+    {
+      "epoch": 0.41,
+      "grad_norm": 0.39081196529426604,
+      "learning_rate": 6.594727718252925e-06,
+      "loss": 0.8877,
+      "step": 1234
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5525141103379567,
+      "learning_rate": 6.589564642688736e-06,
+      "loss": 0.9713,
+      "step": 1235
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 1.0507242507808083,
+      "learning_rate": 6.584399680926092e-06,
+      "loss": 0.7979,
+      "step": 1236
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.4586672833052734,
+      "learning_rate": 6.579232839093808e-06,
+      "loss": 0.9196,
+      "step": 1237
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5557543295173482,
+      "learning_rate": 6.574064123322925e-06,
+      "loss": 0.8396,
+      "step": 1238
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.9476317372240531,
+      "learning_rate": 6.56889353974671e-06,
+      "loss": 0.8534,
+      "step": 1239
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6190352429661059,
+      "learning_rate": 6.563721094500648e-06,
+      "loss": 0.9189,
+      "step": 1240
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6852325306392749,
+      "learning_rate": 6.558546793722429e-06,
+      "loss": 0.9623,
+      "step": 1241
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.37864057808213114,
+      "learning_rate": 6.5533706435519454e-06,
+      "loss": 0.6035,
+      "step": 1242
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5194682359196197,
+      "learning_rate": 6.548192650131289e-06,
+      "loss": 0.7721,
+      "step": 1243
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5643863767487095,
+      "learning_rate": 6.543012819604732e-06,
+      "loss": 0.8166,
+      "step": 1244
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6663557867612867,
+      "learning_rate": 6.537831158118733e-06,
+      "loss": 0.9442,
+      "step": 1245
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.3813985477900902,
+      "learning_rate": 6.5326476718219165e-06,
+      "loss": 0.9247,
+      "step": 1246
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7342712177370677,
+      "learning_rate": 6.5274623668650805e-06,
+      "loss": 0.8415,
+      "step": 1247
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5487537619799551,
+      "learning_rate": 6.522275249401173e-06,
+      "loss": 0.8909,
+      "step": 1248
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5911455200736817,
+      "learning_rate": 6.517086325585296e-06,
+      "loss": 0.7477,
+      "step": 1249
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5276351494357648,
+      "learning_rate": 6.511895601574698e-06,
+      "loss": 0.9222,
+      "step": 1250
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.38447982002677056,
+      "learning_rate": 6.5067030835287605e-06,
+      "loss": 0.6115,
+      "step": 1251
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5002556375154491,
+      "learning_rate": 6.50150877760899e-06,
+      "loss": 0.9763,
+      "step": 1252
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6407832084871841,
+      "learning_rate": 6.496312689979024e-06,
+      "loss": 0.7505,
+      "step": 1253
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5403358277434209,
+      "learning_rate": 6.491114826804607e-06,
+      "loss": 0.8864,
+      "step": 1254
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7544256062247786,
+      "learning_rate": 6.485915194253591e-06,
+      "loss": 0.8304,
+      "step": 1255
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7025031255710215,
+      "learning_rate": 6.48071379849593e-06,
+      "loss": 0.7618,
+      "step": 1256
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.6499761732055422,
+      "learning_rate": 6.475510645703669e-06,
+      "loss": 0.9607,
+      "step": 1257
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7030253277241286,
+      "learning_rate": 6.470305742050938e-06,
+      "loss": 0.899,
+      "step": 1258
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5715466455058523,
+      "learning_rate": 6.465099093713944e-06,
+      "loss": 1.0766,
+      "step": 1259
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.40562013653293416,
+      "learning_rate": 6.459890706870965e-06,
+      "loss": 0.7193,
+      "step": 1260
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7023838616436442,
+      "learning_rate": 6.454680587702342e-06,
+      "loss": 0.841,
+      "step": 1261
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.5449119716769367,
+      "learning_rate": 6.449468742390472e-06,
+      "loss": 0.9003,
+      "step": 1262
+    },
+    {
+      "epoch": 0.42,
+      "grad_norm": 0.7471574172185417,
+      "learning_rate": 6.444255177119799e-06,
+      "loss": 0.7648,
+      "step": 1263
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5058266657357556,
+      "learning_rate": 6.43903989807681e-06,
+      "loss": 0.7613,
+      "step": 1264
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6058987678493217,
+      "learning_rate": 6.433822911450022e-06,
+      "loss": 0.8457,
+      "step": 1265
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.650741985294376,
+      "learning_rate": 6.42860422342998e-06,
+      "loss": 0.9373,
+      "step": 1266
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5350792642838224,
+      "learning_rate": 6.423383840209251e-06,
+      "loss": 0.7836,
+      "step": 1267
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.2558471030693845,
+      "learning_rate": 6.41816176798241e-06,
+      "loss": 0.8436,
+      "step": 1268
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.3814095130156768,
+      "learning_rate": 6.412938012946035e-06,
+      "loss": 0.6708,
+      "step": 1269
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.633778821291055,
+      "learning_rate": 6.407712581298705e-06,
+      "loss": 0.7805,
+      "step": 1270
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.34597744303814243,
+      "learning_rate": 6.4024854792409864e-06,
+      "loss": 1.0032,
+      "step": 1271
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5918217236234016,
+      "learning_rate": 6.397256712975426e-06,
+      "loss": 0.7143,
+      "step": 1272
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5123216441928979,
+      "learning_rate": 6.392026288706549e-06,
+      "loss": 0.9912,
+      "step": 1273
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5134335218616224,
+      "learning_rate": 6.386794212640846e-06,
+      "loss": 0.7769,
+      "step": 1274
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 1.0053994274162352,
+      "learning_rate": 6.381560490986766e-06,
+      "loss": 0.8345,
+      "step": 1275
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5791468154672057,
+      "learning_rate": 6.376325129954712e-06,
+      "loss": 0.9252,
+      "step": 1276
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5732434566609952,
+      "learning_rate": 6.371088135757038e-06,
+      "loss": 0.6954,
+      "step": 1277
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.38659302138487395,
+      "learning_rate": 6.365849514608025e-06,
+      "loss": 0.6924,
+      "step": 1278
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5401032870481175,
+      "learning_rate": 6.36060927272389e-06,
+      "loss": 0.7102,
+      "step": 1279
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5428470761517648,
+      "learning_rate": 6.3553674163227786e-06,
+      "loss": 0.7485,
+      "step": 1280
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5883437733177318,
+      "learning_rate": 6.3501239516247424e-06,
+      "loss": 0.9299,
+      "step": 1281
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5673482121751694,
+      "learning_rate": 6.344878884851748e-06,
+      "loss": 0.7615,
+      "step": 1282
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5124705037228608,
+      "learning_rate": 6.339632222227661e-06,
+      "loss": 1.0372,
+      "step": 1283
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.34605001223878695,
+      "learning_rate": 6.33438396997824e-06,
+      "loss": 0.9594,
+      "step": 1284
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6191162784427793,
+      "learning_rate": 6.329134134331129e-06,
+      "loss": 0.8898,
+      "step": 1285
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.41156132700149417,
+      "learning_rate": 6.3238827215158575e-06,
+      "loss": 0.5699,
+      "step": 1286
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5377736895972479,
+      "learning_rate": 6.318629737763818e-06,
+      "loss": 0.8323,
+      "step": 1287
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6574328203274723,
+      "learning_rate": 6.313375189308269e-06,
+      "loss": 0.7811,
+      "step": 1288
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5177800752296974,
+      "learning_rate": 6.3081190823843295e-06,
+      "loss": 0.8133,
+      "step": 1289
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.6163522970048646,
+      "learning_rate": 6.302861423228967e-06,
+      "loss": 0.9916,
+      "step": 1290
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5535606732416045,
+      "learning_rate": 6.297602218080988e-06,
+      "loss": 0.7699,
+      "step": 1291
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.5658756137253763,
+      "learning_rate": 6.292341473181034e-06,
+      "loss": 0.7724,
+      "step": 1292
+    },
+    {
+      "epoch": 0.43,
+      "grad_norm": 0.9399132953729401,
+      "learning_rate": 6.287079194771576e-06,
+      "loss": 0.8513,
+      "step": 1293
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.4517568587555566,
+      "learning_rate": 6.281815389096903e-06,
+      "loss": 1.0102,
+      "step": 1294
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.44300407674452874,
+      "learning_rate": 6.276550062403115e-06,
+      "loss": 0.66,
+      "step": 1295
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5790246771187333,
+      "learning_rate": 6.271283220938119e-06,
+      "loss": 0.7831,
+      "step": 1296
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1196021830137914,
+      "learning_rate": 6.26601487095162e-06,
+      "loss": 0.9614,
+      "step": 1297
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.3931424396429415,
+      "learning_rate": 6.260745018695112e-06,
+      "loss": 0.926,
+      "step": 1298
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5322344223993547,
+      "learning_rate": 6.25547367042187e-06,
+      "loss": 0.7353,
+      "step": 1299
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6358563540241557,
+      "learning_rate": 6.250200832386947e-06,
+      "loss": 0.9894,
+      "step": 1300
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.7460358185279562,
+      "learning_rate": 6.244926510847162e-06,
+      "loss": 0.6811,
+      "step": 1301
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5136340379444415,
+      "learning_rate": 6.239650712061093e-06,
+      "loss": 0.8823,
+      "step": 1302
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5436312861742828,
+      "learning_rate": 6.2343734422890745e-06,
+      "loss": 0.8602,
+      "step": 1303
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6204723255132517,
+      "learning_rate": 6.229094707793185e-06,
+      "loss": 0.8334,
+      "step": 1304
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.4374780179521296,
+      "learning_rate": 6.223814514837241e-06,
+      "loss": 0.6599,
+      "step": 1305
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.8489205910126092,
+      "learning_rate": 6.2185328696867866e-06,
+      "loss": 0.89,
+      "step": 1306
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.4769584999147083,
+      "learning_rate": 6.2132497786090955e-06,
+      "loss": 1.0344,
+      "step": 1307
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6385058243564209,
+      "learning_rate": 6.207965247873151e-06,
+      "loss": 0.7906,
+      "step": 1308
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.38607267210973595,
+      "learning_rate": 6.202679283749647e-06,
+      "loss": 0.9558,
+      "step": 1309
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6381858802842635,
+      "learning_rate": 6.197391892510982e-06,
+      "loss": 0.8796,
+      "step": 1310
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6239801632371536,
+      "learning_rate": 6.192103080431242e-06,
+      "loss": 0.7237,
+      "step": 1311
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.010672857787697,
+      "learning_rate": 6.1868128537862005e-06,
+      "loss": 0.8672,
+      "step": 1312
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.4578307000914213,
+      "learning_rate": 6.181521218853312e-06,
+      "loss": 0.5634,
+      "step": 1313
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5897998504716209,
+      "learning_rate": 6.176228181911699e-06,
+      "loss": 0.9593,
+      "step": 1314
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 1.1160799956317637,
+      "learning_rate": 6.1709337492421515e-06,
+      "loss": 0.7836,
+      "step": 1315
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.6324976477935099,
+      "learning_rate": 6.165637927127111e-06,
+      "loss": 1.025,
+      "step": 1316
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5863955060368177,
+      "learning_rate": 6.1603407218506726e-06,
+      "loss": 0.9301,
+      "step": 1317
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5165329318301273,
+      "learning_rate": 6.155042139698568e-06,
+      "loss": 0.9558,
+      "step": 1318
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.4789640538188497,
+      "learning_rate": 6.149742186958166e-06,
+      "loss": 0.9226,
+      "step": 1319
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.961412403482475,
+      "learning_rate": 6.144440869918462e-06,
+      "loss": 0.7357,
+      "step": 1320
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.33582256622510404,
+      "learning_rate": 6.139138194870067e-06,
+      "loss": 0.9428,
+      "step": 1321
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.43477068914501477,
+      "learning_rate": 6.133834168105206e-06,
+      "loss": 0.5866,
+      "step": 1322
+    },
+    {
+      "epoch": 0.44,
+      "grad_norm": 0.5419271861418904,
+      "learning_rate": 6.128528795917708e-06,
+      "loss": 0.7588,
+      "step": 1323
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6476922054065702,
+      "learning_rate": 6.123222084602997e-06,
+      "loss": 0.7932,
+      "step": 1324
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5552372254793269,
+      "learning_rate": 6.117914040458087e-06,
+      "loss": 0.8084,
+      "step": 1325
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5725919649768952,
+      "learning_rate": 6.112604669781572e-06,
+      "loss": 0.9255,
+      "step": 1326
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5575639208127194,
+      "learning_rate": 6.107293978873626e-06,
+      "loss": 0.9065,
+      "step": 1327
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5591571888037937,
+      "learning_rate": 6.101981974035982e-06,
+      "loss": 0.8579,
+      "step": 1328
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.49469413858814665,
+      "learning_rate": 6.096668661571934e-06,
+      "loss": 0.84,
+      "step": 1329
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5350413708804319,
+      "learning_rate": 6.091354047786333e-06,
+      "loss": 0.9163,
+      "step": 1330
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.4901418200063408,
+      "learning_rate": 6.086038138985566e-06,
+      "loss": 0.6306,
+      "step": 1331
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6334410108447643,
+      "learning_rate": 6.080720941477564e-06,
+      "loss": 0.8053,
+      "step": 1332
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.3564866244929682,
+      "learning_rate": 6.075402461571781e-06,
+      "loss": 0.9549,
+      "step": 1333
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5944558585913549,
+      "learning_rate": 6.070082705579198e-06,
+      "loss": 0.6915,
+      "step": 1334
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5463622951099768,
+      "learning_rate": 6.0647616798123075e-06,
+      "loss": 1.0006,
+      "step": 1335
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5024266259562721,
+      "learning_rate": 6.0594393905851065e-06,
+      "loss": 0.7656,
+      "step": 1336
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5928425417709194,
+      "learning_rate": 6.054115844213099e-06,
+      "loss": 0.7331,
+      "step": 1337
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.498457214715271,
+      "learning_rate": 6.048791047013272e-06,
+      "loss": 0.9414,
+      "step": 1338
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.45218807021834384,
+      "learning_rate": 6.043465005304099e-06,
+      "loss": 0.6001,
+      "step": 1339
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.563708644812555,
+      "learning_rate": 6.038137725405535e-06,
+      "loss": 0.9685,
+      "step": 1340
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.48716470176799637,
+      "learning_rate": 6.032809213639e-06,
+      "loss": 0.91,
+      "step": 1341
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5573004398278839,
+      "learning_rate": 6.027479476327376e-06,
+      "loss": 0.8876,
+      "step": 1342
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5184107394421149,
+      "learning_rate": 6.0221485197949995e-06,
+      "loss": 0.9221,
+      "step": 1343
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5777256263156797,
+      "learning_rate": 6.016816350367655e-06,
+      "loss": 0.7244,
+      "step": 1344
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6882713450580209,
+      "learning_rate": 6.011482974372564e-06,
+      "loss": 0.9574,
+      "step": 1345
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.35543396887340506,
+      "learning_rate": 6.006148398138383e-06,
+      "loss": 0.8879,
+      "step": 1346
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5332967356348799,
+      "learning_rate": 6.00081262799519e-06,
+      "loss": 0.9303,
+      "step": 1347
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5071044306068199,
+      "learning_rate": 5.995475670274481e-06,
+      "loss": 0.7321,
+      "step": 1348
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.44432381663486004,
+      "learning_rate": 5.990137531309158e-06,
+      "loss": 0.5595,
+      "step": 1349
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.545516618834295,
+      "learning_rate": 5.9847982174335314e-06,
+      "loss": 0.9521,
+      "step": 1350
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5840018115456381,
+      "learning_rate": 5.9794577349833e-06,
+      "loss": 0.7948,
+      "step": 1351
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.6345549075248433,
+      "learning_rate": 5.974116090295548e-06,
+      "loss": 0.9775,
+      "step": 1352
+    },
+    {
+      "epoch": 0.45,
+      "grad_norm": 0.5078897235128643,
+      "learning_rate": 5.968773289708746e-06,
+      "loss": 0.9102,
+      "step": 1353
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.9207272832973242,
+      "learning_rate": 5.963429339562731e-06,
+      "loss": 0.8051,
+      "step": 1354
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6039449693878332,
+      "learning_rate": 5.958084246198704e-06,
+      "loss": 0.9743,
+      "step": 1355
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5463103761159442,
+      "learning_rate": 5.952738015959225e-06,
+      "loss": 0.8293,
+      "step": 1356
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.4928535190765737,
+      "learning_rate": 5.9473906551881985e-06,
+      "loss": 0.8464,
+      "step": 1357
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.33164959365449953,
+      "learning_rate": 5.942042170230879e-06,
+      "loss": 0.7797,
+      "step": 1358
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5220527218822149,
+      "learning_rate": 5.936692567433844e-06,
+      "loss": 0.9582,
+      "step": 1359
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.740655670920854,
+      "learning_rate": 5.931341853145005e-06,
+      "loss": 0.807,
+      "step": 1360
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5753425679855757,
+      "learning_rate": 5.925990033713591e-06,
+      "loss": 0.9251,
+      "step": 1361
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.49415556776119013,
+      "learning_rate": 5.920637115490142e-06,
+      "loss": 0.8274,
+      "step": 1362
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5502569337868357,
+      "learning_rate": 5.9152831048265005e-06,
+      "loss": 0.8404,
+      "step": 1363
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5467495009714272,
+      "learning_rate": 5.9099280080758085e-06,
+      "loss": 0.9286,
+      "step": 1364
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5378375528155436,
+      "learning_rate": 5.904571831592491e-06,
+      "loss": 0.9049,
+      "step": 1365
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.39521963637693314,
+      "learning_rate": 5.899214581732262e-06,
+      "loss": 0.6657,
+      "step": 1366
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5913697951121879,
+      "learning_rate": 5.893856264852103e-06,
+      "loss": 0.8441,
+      "step": 1367
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6923690424044437,
+      "learning_rate": 5.888496887310265e-06,
+      "loss": 0.7704,
+      "step": 1368
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.583057972329266,
+      "learning_rate": 5.883136455466254e-06,
+      "loss": 0.894,
+      "step": 1369
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.608389995051209,
+      "learning_rate": 5.877774975680831e-06,
+      "loss": 0.7557,
+      "step": 1370
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.32841168266327103,
+      "learning_rate": 5.872412454315999e-06,
+      "loss": 0.9587,
+      "step": 1371
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.6946430883176724,
+      "learning_rate": 5.867048897734997e-06,
+      "loss": 0.7777,
+      "step": 1372
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.5433699606895375,
+      "learning_rate": 5.8616843123022925e-06,
+      "loss": 0.8064,
+      "step": 1373
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7418882740073106,
+      "learning_rate": 5.856318704383572e-06,
+      "loss": 0.8794,
+      "step": 1374
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.42401735488500114,
+      "learning_rate": 5.850952080345739e-06,
+      "loss": 0.5026,
+      "step": 1375
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.46724250831668657,
+      "learning_rate": 5.845584446556899e-06,
+      "loss": 1.0099,
+      "step": 1376
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.512522512841429,
+      "learning_rate": 5.840215809386357e-06,
+      "loss": 0.7568,
+      "step": 1377
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.609165289671596,
+      "learning_rate": 5.834846175204612e-06,
+      "loss": 0.7967,
+      "step": 1378
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.4775375874469242,
+      "learning_rate": 5.829475550383339e-06,
+      "loss": 0.8201,
+      "step": 1379
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.8426654384315555,
+      "learning_rate": 5.824103941295394e-06,
+      "loss": 0.8789,
+      "step": 1380
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.7255722257078904,
+      "learning_rate": 5.818731354314801e-06,
+      "loss": 0.9665,
+      "step": 1381
+    },
+    {
+      "epoch": 0.46,
+      "grad_norm": 0.41180430444394917,
+      "learning_rate": 5.813357795816742e-06,
+      "loss": 0.9689,
+      "step": 1382
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.8345520555084504,
+      "learning_rate": 5.807983272177553e-06,
+      "loss": 0.8979,
+      "step": 1383
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.541366768452341,
+      "learning_rate": 5.802607789774713e-06,
+      "loss": 0.5614,
+      "step": 1384
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6772613090863823,
+      "learning_rate": 5.797231354986842e-06,
+      "loss": 0.843,
+      "step": 1385
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4619985593149905,
+      "learning_rate": 5.791853974193688e-06,
+      "loss": 0.8559,
+      "step": 1386
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.47993095860267987,
+      "learning_rate": 5.7864756537761245e-06,
+      "loss": 0.906,
+      "step": 1387
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5631892728649719,
+      "learning_rate": 5.781096400116135e-06,
+      "loss": 0.9906,
+      "step": 1388
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.9577081895167034,
+      "learning_rate": 5.775716219596815e-06,
+      "loss": 0.7451,
+      "step": 1389
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5719179010955436,
+      "learning_rate": 5.7703351186023575e-06,
+      "loss": 0.8718,
+      "step": 1390
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.640179523191946,
+      "learning_rate": 5.7649531035180476e-06,
+      "loss": 0.6801,
+      "step": 1391
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5178120251218077,
+      "learning_rate": 5.759570180730255e-06,
+      "loss": 0.645,
+      "step": 1392
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5529872047222608,
+      "learning_rate": 5.754186356626429e-06,
+      "loss": 0.9939,
+      "step": 1393
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6720452891650328,
+      "learning_rate": 5.748801637595085e-06,
+      "loss": 0.7907,
+      "step": 1394
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.600911526489819,
+      "learning_rate": 5.743416030025804e-06,
+      "loss": 0.9176,
+      "step": 1395
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.43888305118544657,
+      "learning_rate": 5.738029540309215e-06,
+      "loss": 0.9034,
+      "step": 1396
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6348263653204236,
+      "learning_rate": 5.732642174837003e-06,
+      "loss": 0.8154,
+      "step": 1397
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.48464521431573027,
+      "learning_rate": 5.727253940001884e-06,
+      "loss": 1.0064,
+      "step": 1398
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6318658400327235,
+      "learning_rate": 5.721864842197612e-06,
+      "loss": 0.788,
+      "step": 1399
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5962895397167132,
+      "learning_rate": 5.716474887818957e-06,
+      "loss": 1.0568,
+      "step": 1400
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6678916177821296,
+      "learning_rate": 5.711084083261716e-06,
+      "loss": 0.737,
+      "step": 1401
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.38768024021925296,
+      "learning_rate": 5.705692434922684e-06,
+      "loss": 0.589,
+      "step": 1402
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6279430419499268,
+      "learning_rate": 5.700299949199664e-06,
+      "loss": 0.8195,
+      "step": 1403
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5551606579267712,
+      "learning_rate": 5.694906632491452e-06,
+      "loss": 0.8937,
+      "step": 1404
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.7369778141402169,
+      "learning_rate": 5.689512491197827e-06,
+      "loss": 1.003,
+      "step": 1405
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4698220916258476,
+      "learning_rate": 5.684117531719552e-06,
+      "loss": 0.813,
+      "step": 1406
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6194116863126857,
+      "learning_rate": 5.678721760458352e-06,
+      "loss": 0.9739,
+      "step": 1407
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.3594086459988229,
+      "learning_rate": 5.673325183816927e-06,
+      "loss": 0.8788,
+      "step": 1408
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5171800209340425,
+      "learning_rate": 5.667927808198922e-06,
+      "loss": 0.9273,
+      "step": 1409
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.5480036518896935,
+      "learning_rate": 5.662529640008933e-06,
+      "loss": 0.7863,
+      "step": 1410
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.4579107553879061,
+      "learning_rate": 5.657130685652503e-06,
+      "loss": 0.6898,
+      "step": 1411
+    },
+    {
+      "epoch": 0.47,
+      "grad_norm": 0.6301989243717202,
+      "learning_rate": 5.6517309515361005e-06,
+      "loss": 0.9602,
+      "step": 1412
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.191094364177108,
+      "learning_rate": 5.646330444067121e-06,
+      "loss": 0.8332,
+      "step": 1413
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5466292362660937,
+      "learning_rate": 5.640929169653876e-06,
+      "loss": 0.9879,
+      "step": 1414
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6785290160162354,
+      "learning_rate": 5.635527134705593e-06,
+      "loss": 0.8109,
+      "step": 1415
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.5952563241525357,
+      "learning_rate": 5.630124345632396e-06,
+      "loss": 0.6847,
+      "step": 1416
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5900113601279945,
+      "learning_rate": 5.6247208088453065e-06,
+      "loss": 0.9235,
+      "step": 1417
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5747940434348776,
+      "learning_rate": 5.619316530756234e-06,
+      "loss": 0.7781,
+      "step": 1418
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.38754663669899453,
+      "learning_rate": 5.613911517777965e-06,
+      "loss": 0.7067,
+      "step": 1419
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4468148590488156,
+      "learning_rate": 5.608505776324158e-06,
+      "loss": 0.8945,
+      "step": 1420
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5655276805091501,
+      "learning_rate": 5.6030993128093415e-06,
+      "loss": 0.9522,
+      "step": 1421
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5613735541555306,
+      "learning_rate": 5.597692133648894e-06,
+      "loss": 0.7708,
+      "step": 1422
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4760117836429006,
+      "learning_rate": 5.592284245259047e-06,
+      "loss": 0.9129,
+      "step": 1423
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6116738123740251,
+      "learning_rate": 5.5868756540568704e-06,
+      "loss": 0.8643,
+      "step": 1424
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5241324347609684,
+      "learning_rate": 5.5814663664602735e-06,
+      "loss": 0.804,
+      "step": 1425
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6096111613075884,
+      "learning_rate": 5.576056388887985e-06,
+      "loss": 1.0071,
+      "step": 1426
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.7135205839469514,
+      "learning_rate": 5.570645727759558e-06,
+      "loss": 0.8182,
+      "step": 1427
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.3994727299977213,
+      "learning_rate": 5.565234389495354e-06,
+      "loss": 0.5671,
+      "step": 1428
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5779980244939685,
+      "learning_rate": 5.559822380516539e-06,
+      "loss": 0.8197,
+      "step": 1429
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5467741560227072,
+      "learning_rate": 5.554409707245076e-06,
+      "loss": 0.8176,
+      "step": 1430
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.2511254467883317,
+      "learning_rate": 5.548996376103712e-06,
+      "loss": 0.9088,
+      "step": 1431
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.3725624838503579,
+      "learning_rate": 5.543582393515978e-06,
+      "loss": 0.9309,
+      "step": 1432
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5597645145531849,
+      "learning_rate": 5.538167765906178e-06,
+      "loss": 0.8384,
+      "step": 1433
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4898654141840171,
+      "learning_rate": 5.532752499699381e-06,
+      "loss": 0.9009,
+      "step": 1434
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6143316199852962,
+      "learning_rate": 5.527336601321414e-06,
+      "loss": 0.8994,
+      "step": 1435
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5016658478200072,
+      "learning_rate": 5.521920077198853e-06,
+      "loss": 0.8965,
+      "step": 1436
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.4281230184194929,
+      "learning_rate": 5.516502933759017e-06,
+      "loss": 0.5161,
+      "step": 1437
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 1.1105512034378566,
+      "learning_rate": 5.511085177429961e-06,
+      "loss": 0.9787,
+      "step": 1438
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6395865200448326,
+      "learning_rate": 5.505666814640466e-06,
+      "loss": 0.7885,
+      "step": 1439
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.5661476649381292,
+      "learning_rate": 5.5002478518200316e-06,
+      "loss": 0.8698,
+      "step": 1440
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 0.6209103861496337,
+      "learning_rate": 5.494828295398874e-06,
+      "loss": 0.8273,
+      "step": 1441
+    },
+    {
+      "epoch": 0.48,
+      "grad_norm": 2.067807760662517,
+      "learning_rate": 5.489408151807908e-06,
+      "loss": 0.7243,
+      "step": 1442
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6154225140630487,
+      "learning_rate": 5.483987427478749e-06,
+      "loss": 0.9363,
+      "step": 1443
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.4862808040034081,
+      "learning_rate": 5.4785661288436985e-06,
+      "loss": 0.9389,
+      "step": 1444
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.4990616940890301,
+      "learning_rate": 5.473144262335744e-06,
+      "loss": 0.9017,
+      "step": 1445
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.3966060138149872,
+      "learning_rate": 5.467721834388543e-06,
+      "loss": 0.6778,
+      "step": 1446
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6173144344203854,
+      "learning_rate": 5.462298851436418e-06,
+      "loss": 0.8753,
+      "step": 1447
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.47848371695522196,
+      "learning_rate": 5.456875319914355e-06,
+      "loss": 0.8548,
+      "step": 1448
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7397729687375961,
+      "learning_rate": 5.451451246257989e-06,
+      "loss": 0.7174,
+      "step": 1449
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.500379361026962,
+      "learning_rate": 5.446026636903597e-06,
+      "loss": 0.8155,
+      "step": 1450
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6179922450613776,
+      "learning_rate": 5.440601498288088e-06,
+      "loss": 0.874,
+      "step": 1451
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6030670858504591,
+      "learning_rate": 5.43517583684901e-06,
+      "loss": 0.9339,
+      "step": 1452
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5194671430745754,
+      "learning_rate": 5.429749659024521e-06,
+      "loss": 0.7958,
+      "step": 1453
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6854392251487799,
+      "learning_rate": 5.424322971253395e-06,
+      "loss": 0.798,
+      "step": 1454
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.40690049480095913,
+      "learning_rate": 5.4188957799750145e-06,
+      "loss": 0.6697,
+      "step": 1455
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.8250980800156708,
+      "learning_rate": 5.413468091629353e-06,
+      "loss": 0.7517,
+      "step": 1456
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.3713228912428578,
+      "learning_rate": 5.40803991265698e-06,
+      "loss": 1.0066,
+      "step": 1457
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.7621810561885134,
+      "learning_rate": 5.402611249499042e-06,
+      "loss": 0.6708,
+      "step": 1458
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.590067431966973,
+      "learning_rate": 5.397182108597266e-06,
+      "loss": 1.0192,
+      "step": 1459
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5704103047792082,
+      "learning_rate": 5.391752496393941e-06,
+      "loss": 0.8064,
+      "step": 1460
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6400576718402666,
+      "learning_rate": 5.386322419331914e-06,
+      "loss": 0.8183,
+      "step": 1461
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5468400865728447,
+      "learning_rate": 5.380891883854591e-06,
+      "loss": 0.9353,
+      "step": 1462
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5952637104891704,
+      "learning_rate": 5.375460896405914e-06,
+      "loss": 0.7263,
+      "step": 1463
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.5146808109136807,
+      "learning_rate": 5.370029463430365e-06,
+      "loss": 0.7209,
+      "step": 1464
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6414143415135592,
+      "learning_rate": 5.364597591372957e-06,
+      "loss": 0.6897,
+      "step": 1465
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.9850473140972881,
+      "learning_rate": 5.359165286679218e-06,
+      "loss": 0.7354,
+      "step": 1466
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.49646541937532956,
+      "learning_rate": 5.3537325557951925e-06,
+      "loss": 0.9172,
+      "step": 1467
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.48169727890208935,
+      "learning_rate": 5.348299405167434e-06,
+      "loss": 0.936,
+      "step": 1468
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.654897265523849,
+      "learning_rate": 5.34286584124299e-06,
+      "loss": 1.0372,
+      "step": 1469
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.3230462722016132,
+      "learning_rate": 5.337431870469398e-06,
+      "loss": 0.9172,
+      "step": 1470
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.6714980737816897,
+      "learning_rate": 5.33199749929468e-06,
+      "loss": 0.9189,
+      "step": 1471
+    },
+    {
+      "epoch": 0.49,
+      "grad_norm": 0.4883836739438965,
+      "learning_rate": 5.3265627341673336e-06,
+      "loss": 0.5399,
+      "step": 1472
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5630449175088921,
+      "learning_rate": 5.321127581536323e-06,
+      "loss": 0.8117,
+      "step": 1473
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.4750122964929389,
+      "learning_rate": 5.31569204785107e-06,
+      "loss": 0.8034,
+      "step": 1474
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5696318898069855,
+      "learning_rate": 5.310256139561455e-06,
+      "loss": 0.7369,
+      "step": 1475
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5269139858862575,
+      "learning_rate": 5.304819863117796e-06,
+      "loss": 1.0204,
+      "step": 1476
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.700924681891728,
+      "learning_rate": 5.29938322497085e-06,
+      "loss": 0.7955,
+      "step": 1477
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5829565310612452,
+      "learning_rate": 5.293946231571806e-06,
+      "loss": 0.8156,
+      "step": 1478
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5671652711527967,
+      "learning_rate": 5.2885088893722725e-06,
+      "loss": 0.8558,
+      "step": 1479
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.48292776075164956,
+      "learning_rate": 5.283071204824269e-06,
+      "loss": 1.0109,
+      "step": 1480
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5625872113373949,
+      "learning_rate": 5.277633184380226e-06,
+      "loss": 0.7044,
+      "step": 1481
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5139583282996854,
+      "learning_rate": 5.27219483449297e-06,
+      "loss": 0.7214,
+      "step": 1482
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5923803536695029,
+      "learning_rate": 5.266756161615719e-06,
+      "loss": 1.0535,
+      "step": 1483
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.3468684972782718,
+      "learning_rate": 5.261317172202074e-06,
+      "loss": 0.915,
+      "step": 1484
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5384121108827326,
+      "learning_rate": 5.255877872706013e-06,
+      "loss": 0.7185,
+      "step": 1485
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.1888979154288495,
+      "learning_rate": 5.25043826958188e-06,
+      "loss": 0.9304,
+      "step": 1486
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5890887731822366,
+      "learning_rate": 5.244998369284379e-06,
+      "loss": 0.5729,
+      "step": 1487
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5372032182797931,
+      "learning_rate": 5.23955817826857e-06,
+      "loss": 0.9446,
+      "step": 1488
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 1.0859754909327335,
+      "learning_rate": 5.234117702989855e-06,
+      "loss": 0.822,
+      "step": 1489
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.49550700718857005,
+      "learning_rate": 5.228676949903974e-06,
+      "loss": 0.9355,
+      "step": 1490
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5734669526199275,
+      "learning_rate": 5.223235925466994e-06,
+      "loss": 0.5772,
+      "step": 1491
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6795276646424484,
+      "learning_rate": 5.217794636135311e-06,
+      "loss": 0.9386,
+      "step": 1492
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.476444878704206,
+      "learning_rate": 5.212353088365627e-06,
+      "loss": 1.0309,
+      "step": 1493
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5565595321984478,
+      "learning_rate": 5.2069112886149564e-06,
+      "loss": 0.7534,
+      "step": 1494
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.4397279718627829,
+      "learning_rate": 5.201469243340612e-06,
+      "loss": 0.9251,
+      "step": 1495
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.6574192869076606,
+      "learning_rate": 5.1960269590001965e-06,
+      "loss": 0.8305,
+      "step": 1496
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5275522573628336,
+      "learning_rate": 5.190584442051594e-06,
+      "loss": 0.9003,
+      "step": 1497
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.46954513567740147,
+      "learning_rate": 5.1851416989529705e-06,
+      "loss": 0.7734,
+      "step": 1498
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5456915984922197,
+      "learning_rate": 5.179698736162755e-06,
+      "loss": 0.5366,
+      "step": 1499
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5068803377986252,
+      "learning_rate": 5.17425556013964e-06,
+      "loss": 0.9247,
+      "step": 1500
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 0.5724980901006008,
+      "learning_rate": 5.168812177342568e-06,
+      "loss": 0.8053,
+      "step": 1501
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5765877599673561,
+      "learning_rate": 5.163368594230732e-06,
+      "loss": 0.9455,
+      "step": 1502
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.561468812181774,
+      "learning_rate": 5.15792481726356e-06,
+      "loss": 0.8868,
+      "step": 1503
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.49089237224290877,
+      "learning_rate": 5.1524808529007075e-06,
+      "loss": 0.9106,
+      "step": 1504
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.46560881134390814,
+      "learning_rate": 5.147036707602057e-06,
+      "loss": 0.8673,
+      "step": 1505
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5909067634312173,
+      "learning_rate": 5.141592387827701e-06,
+      "loss": 0.7243,
+      "step": 1506
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.7944239004681403,
+      "learning_rate": 5.136147900037942e-06,
+      "loss": 0.8092,
+      "step": 1507
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5498033460537294,
+      "learning_rate": 5.130703250693284e-06,
+      "loss": 0.9041,
+      "step": 1508
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.531891328409394,
+      "learning_rate": 5.1252584462544175e-06,
+      "loss": 0.7186,
+      "step": 1509
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.4659550701399199,
+      "learning_rate": 5.119813493182221e-06,
+      "loss": 0.8045,
+      "step": 1510
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.566558967826336,
+      "learning_rate": 5.114368397937744e-06,
+      "loss": 0.8038,
+      "step": 1511
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6058641442321939,
+      "learning_rate": 5.108923166982215e-06,
+      "loss": 0.9522,
+      "step": 1512
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6190946477526933,
+      "learning_rate": 5.103477806777012e-06,
+      "loss": 0.8704,
+      "step": 1513
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5376333191175304,
+      "learning_rate": 5.098032323783673e-06,
+      "loss": 0.8419,
+      "step": 1514
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5265219374450912,
+      "learning_rate": 5.092586724463881e-06,
+      "loss": 0.8646,
+      "step": 1515
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5268349741685618,
+      "learning_rate": 5.087141015279455e-06,
+      "loss": 0.8958,
+      "step": 1516
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.4683167492733695,
+      "learning_rate": 5.081695202692347e-06,
+      "loss": 0.63,
+      "step": 1517
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5006045263218769,
+      "learning_rate": 5.07624929316463e-06,
+      "loss": 0.7868,
+      "step": 1518
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.3318515964903951,
+      "learning_rate": 5.070803293158492e-06,
+      "loss": 0.9687,
+      "step": 1519
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5513521324944715,
+      "learning_rate": 5.06535720913623e-06,
+      "loss": 0.6867,
+      "step": 1520
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5101037931069548,
+      "learning_rate": 5.059911047560236e-06,
+      "loss": 0.9456,
+      "step": 1521
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5729349153525116,
+      "learning_rate": 5.054464814893001e-06,
+      "loss": 0.7712,
+      "step": 1522
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5661111047669664,
+      "learning_rate": 5.049018517597095e-06,
+      "loss": 0.7165,
+      "step": 1523
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5332088557695225,
+      "learning_rate": 5.043572162135166e-06,
+      "loss": 0.9051,
+      "step": 1524
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.48583287400527914,
+      "learning_rate": 5.038125754969933e-06,
+      "loss": 0.5821,
+      "step": 1525
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5343333054375786,
+      "learning_rate": 5.032679302564176e-06,
+      "loss": 0.9202,
+      "step": 1526
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.46167147710331574,
+      "learning_rate": 5.027232811380723e-06,
+      "loss": 0.9289,
+      "step": 1527
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5444793325800605,
+      "learning_rate": 5.021786287882458e-06,
+      "loss": 0.8608,
+      "step": 1528
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.48425848665712873,
+      "learning_rate": 5.0163397385322955e-06,
+      "loss": 0.865,
+      "step": 1529
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.5591432425703317,
+      "learning_rate": 5.010893169793182e-06,
+      "loss": 0.7429,
+      "step": 1530
+    },
+    {
+      "epoch": 0.51,
+      "grad_norm": 0.6817119701722063,
+      "learning_rate": 5.005446588128088e-06,
+      "loss": 1.006,
+      "step": 1531
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.3570547797232624,
+      "learning_rate": 5e-06,
+      "loss": 0.8847,
+      "step": 1532
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6672612862013213,
+      "learning_rate": 4.994553411871914e-06,
+      "loss": 0.8921,
+      "step": 1533
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.4715584680240374,
+      "learning_rate": 4.98910683020682e-06,
+      "loss": 0.751,
+      "step": 1534
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.46385639448012944,
+      "learning_rate": 4.983660261467706e-06,
+      "loss": 0.5223,
+      "step": 1535
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.4719636773678865,
+      "learning_rate": 4.978213712117543e-06,
+      "loss": 0.9361,
+      "step": 1536
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.4888539727862817,
+      "learning_rate": 4.972767188619278e-06,
+      "loss": 0.7603,
+      "step": 1537
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5371775729211952,
+      "learning_rate": 4.9673206974358254e-06,
+      "loss": 0.969,
+      "step": 1538
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5515754262398703,
+      "learning_rate": 4.9618742450300675e-06,
+      "loss": 0.923,
+      "step": 1539
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5130119515426325,
+      "learning_rate": 4.956427837864835e-06,
+      "loss": 0.8802,
+      "step": 1540
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5113968468566149,
+      "learning_rate": 4.9509814824029065e-06,
+      "loss": 0.8084,
+      "step": 1541
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6130687833841498,
+      "learning_rate": 4.945535185107e-06,
+      "loss": 0.8259,
+      "step": 1542
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5005508489006748,
+      "learning_rate": 4.940088952439765e-06,
+      "loss": 0.8922,
+      "step": 1543
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.33379234121645457,
+      "learning_rate": 4.934642790863772e-06,
+      "loss": 0.7462,
+      "step": 1544
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 1.1695712817467114,
+      "learning_rate": 4.9291967068415085e-06,
+      "loss": 1.0329,
+      "step": 1545
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5894327247014829,
+      "learning_rate": 4.923750706835371e-06,
+      "loss": 0.7179,
+      "step": 1546
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5418228355492859,
+      "learning_rate": 4.9183047973076546e-06,
+      "loss": 1.01,
+      "step": 1547
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5890682925862637,
+      "learning_rate": 4.912858984720546e-06,
+      "loss": 0.772,
+      "step": 1548
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5167808775652243,
+      "learning_rate": 4.90741327553612e-06,
+      "loss": 0.7182,
+      "step": 1549
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.78056498039351,
+      "learning_rate": 4.901967676216329e-06,
+      "loss": 1.0344,
+      "step": 1550
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.558314032657214,
+      "learning_rate": 4.89652219322299e-06,
+      "loss": 0.9438,
+      "step": 1551
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6124963087890729,
+      "learning_rate": 4.8910768330177875e-06,
+      "loss": 0.6782,
+      "step": 1552
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.7528767260995345,
+      "learning_rate": 4.8856316020622564e-06,
+      "loss": 0.8109,
+      "step": 1553
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5643881717222297,
+      "learning_rate": 4.880186506817781e-06,
+      "loss": 0.7851,
+      "step": 1554
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5936445164109526,
+      "learning_rate": 4.874741553745583e-06,
+      "loss": 0.9472,
+      "step": 1555
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.39985194743400126,
+      "learning_rate": 4.8692967493067175e-06,
+      "loss": 0.9135,
+      "step": 1556
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.48646820935610774,
+      "learning_rate": 4.8638520999620585e-06,
+      "loss": 0.9438,
+      "step": 1557
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.5493992336171958,
+      "learning_rate": 4.8584076121723e-06,
+      "loss": 0.8069,
+      "step": 1558
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.4856174126851461,
+      "learning_rate": 4.852963292397946e-06,
+      "loss": 0.7851,
+      "step": 1559
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.6981073230147626,
+      "learning_rate": 4.847519147099294e-06,
+      "loss": 0.8754,
+      "step": 1560
+    },
+    {
+      "epoch": 0.52,
+      "grad_norm": 0.43421362901769217,
+      "learning_rate": 4.8420751827364415e-06,
+      "loss": 0.6696,
+      "step": 1561
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6550707095852669,
+      "learning_rate": 4.8366314057692684e-06,
+      "loss": 0.8573,
+      "step": 1562
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5378655537926395,
+      "learning_rate": 4.831187822657434e-06,
+      "loss": 0.8076,
+      "step": 1563
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5990201024434038,
+      "learning_rate": 4.825744439860362e-06,
+      "loss": 0.8718,
+      "step": 1564
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5134510753466854,
+      "learning_rate": 4.8203012638372475e-06,
+      "loss": 0.7192,
+      "step": 1565
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6115286428921951,
+      "learning_rate": 4.814858301047031e-06,
+      "loss": 0.8612,
+      "step": 1566
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6193336558827097,
+      "learning_rate": 4.809415557948407e-06,
+      "loss": 0.9394,
+      "step": 1567
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.36899625533935004,
+      "learning_rate": 4.803973040999804e-06,
+      "loss": 0.9662,
+      "step": 1568
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5370113350983119,
+      "learning_rate": 4.798530756659389e-06,
+      "loss": 0.8521,
+      "step": 1569
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.491570988442801,
+      "learning_rate": 4.793088711385044e-06,
+      "loss": 0.5645,
+      "step": 1570
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.9336620558102691,
+      "learning_rate": 4.7876469116343745e-06,
+      "loss": 0.8143,
+      "step": 1571
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.41675840560875294,
+      "learning_rate": 4.782205363864692e-06,
+      "loss": 0.8557,
+      "step": 1572
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5213532558372046,
+      "learning_rate": 4.776764074533008e-06,
+      "loss": 0.8725,
+      "step": 1573
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.45745457116202687,
+      "learning_rate": 4.771323050096028e-06,
+      "loss": 0.9448,
+      "step": 1574
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5924890056501597,
+      "learning_rate": 4.765882297010147e-06,
+      "loss": 0.7314,
+      "step": 1575
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5184658393040683,
+      "learning_rate": 4.760441821731431e-06,
+      "loss": 0.8774,
+      "step": 1576
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.4603046854810551,
+      "learning_rate": 4.755001630715622e-06,
+      "loss": 0.781,
+      "step": 1577
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.48208501741948484,
+      "learning_rate": 4.749561730418121e-06,
+      "loss": 0.5261,
+      "step": 1578
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5268145620559865,
+      "learning_rate": 4.7441221272939886e-06,
+      "loss": 0.9917,
+      "step": 1579
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5423788469905657,
+      "learning_rate": 4.738682827797928e-06,
+      "loss": 0.75,
+      "step": 1580
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5504138036339412,
+      "learning_rate": 4.733243838384282e-06,
+      "loss": 0.9047,
+      "step": 1581
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.33126397288412796,
+      "learning_rate": 4.727805165507032e-06,
+      "loss": 0.8946,
+      "step": 1582
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6681683559071655,
+      "learning_rate": 4.722366815619776e-06,
+      "loss": 0.9323,
+      "step": 1583
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5158616325607743,
+      "learning_rate": 4.716928795175732e-06,
+      "loss": 0.954,
+      "step": 1584
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5274427336961677,
+      "learning_rate": 4.711491110627729e-06,
+      "loss": 0.7602,
+      "step": 1585
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.5122305094848724,
+      "learning_rate": 4.706053768428195e-06,
+      "loss": 0.969,
+      "step": 1586
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6143927116019613,
+      "learning_rate": 4.700616775029151e-06,
+      "loss": 0.7066,
+      "step": 1587
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.40156214696082076,
+      "learning_rate": 4.6951801368822055e-06,
+      "loss": 0.5636,
+      "step": 1588
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.78836971631889,
+      "learning_rate": 4.689743860438547e-06,
+      "loss": 0.8666,
+      "step": 1589
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6248022876546149,
+      "learning_rate": 4.684307952148931e-06,
+      "loss": 0.8849,
+      "step": 1590
+    },
+    {
+      "epoch": 0.53,
+      "grad_norm": 0.6315487006140595,
+      "learning_rate": 4.678872418463679e-06,
+      "loss": 0.9976,
+      "step": 1591
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.47908794018740286,
+      "learning_rate": 4.673437265832668e-06,
+      "loss": 0.8274,
+      "step": 1592
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.45513531262390383,
+      "learning_rate": 4.668002500705322e-06,
+      "loss": 0.9548,
+      "step": 1593
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.4988356376188356,
+      "learning_rate": 4.662568129530603e-06,
+      "loss": 0.6936,
+      "step": 1594
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5037530161575713,
+      "learning_rate": 4.6571341587570114e-06,
+      "loss": 0.9108,
+      "step": 1595
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.44557978509541735,
+      "learning_rate": 4.6517005948325674e-06,
+      "loss": 0.8842,
+      "step": 1596
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.4404348251309851,
+      "learning_rate": 4.646267444204809e-06,
+      "loss": 0.5389,
+      "step": 1597
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5316859268619631,
+      "learning_rate": 4.640834713320785e-06,
+      "loss": 0.9798,
+      "step": 1598
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5882965128916691,
+      "learning_rate": 4.635402408627046e-06,
+      "loss": 0.7693,
+      "step": 1599
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5489361251128632,
+      "learning_rate": 4.629970536569636e-06,
+      "loss": 0.9412,
+      "step": 1600
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5583623508559992,
+      "learning_rate": 4.624539103594087e-06,
+      "loss": 0.7851,
+      "step": 1601
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5215202926427825,
+      "learning_rate": 4.619108116145411e-06,
+      "loss": 0.8845,
+      "step": 1602
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6063899235753075,
+      "learning_rate": 4.613677580668087e-06,
+      "loss": 0.7761,
+      "step": 1603
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6021758071225181,
+      "learning_rate": 4.608247503606061e-06,
+      "loss": 0.7574,
+      "step": 1604
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.4356323212509351,
+      "learning_rate": 4.602817891402735e-06,
+      "loss": 0.7404,
+      "step": 1605
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.587028105309436,
+      "learning_rate": 4.597388750500959e-06,
+      "loss": 0.8629,
+      "step": 1606
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6023514139652903,
+      "learning_rate": 4.591960087343023e-06,
+      "loss": 0.9421,
+      "step": 1607
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6652702095634688,
+      "learning_rate": 4.586531908370648e-06,
+      "loss": 0.8097,
+      "step": 1608
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.4555239079204499,
+      "learning_rate": 4.581104220024988e-06,
+      "loss": 0.8897,
+      "step": 1609
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5292227195553245,
+      "learning_rate": 4.575677028746606e-06,
+      "loss": 0.8324,
+      "step": 1610
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.49820215268290646,
+      "learning_rate": 4.57025034097548e-06,
+      "loss": 0.8311,
+      "step": 1611
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.4924440711813656,
+      "learning_rate": 4.564824163150992e-06,
+      "loss": 1.0273,
+      "step": 1612
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.6369338342279315,
+      "learning_rate": 4.559398501711913e-06,
+      "loss": 0.7821,
+      "step": 1613
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.356239163036822,
+      "learning_rate": 4.553973363096405e-06,
+      "loss": 0.5576,
+      "step": 1614
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5196151023691132,
+      "learning_rate": 4.548548753742012e-06,
+      "loss": 0.8444,
+      "step": 1615
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5207231117260941,
+      "learning_rate": 4.5431246800856455e-06,
+      "loss": 0.816,
+      "step": 1616
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.5285617939263665,
+      "learning_rate": 4.537701148563584e-06,
+      "loss": 0.8815,
+      "step": 1617
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.34153314069915536,
+      "learning_rate": 4.532278165611459e-06,
+      "loss": 0.9244,
+      "step": 1618
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.657812470662135,
+      "learning_rate": 4.526855737664257e-06,
+      "loss": 0.872,
+      "step": 1619
+    },
+    {
+      "epoch": 0.54,
+      "grad_norm": 0.481518149621036,
+      "learning_rate": 4.521433871156302e-06,
+      "loss": 0.9042,
+      "step": 1620
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5262929778736092,
+      "learning_rate": 4.516012572521252e-06,
+      "loss": 0.8892,
+      "step": 1621
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.529277480187262,
+      "learning_rate": 4.510591848192093e-06,
+      "loss": 0.912,
+      "step": 1622
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.509668677009833,
+      "learning_rate": 4.505171704601128e-06,
+      "loss": 0.5365,
+      "step": 1623
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5081596645900809,
+      "learning_rate": 4.49975214817997e-06,
+      "loss": 0.9291,
+      "step": 1624
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6227167989210145,
+      "learning_rate": 4.494333185359536e-06,
+      "loss": 0.7799,
+      "step": 1625
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5350917561658534,
+      "learning_rate": 4.4889148225700406e-06,
+      "loss": 0.8868,
+      "step": 1626
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5816173056625777,
+      "learning_rate": 4.483497066240985e-06,
+      "loss": 0.7849,
+      "step": 1627
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6071145136524646,
+      "learning_rate": 4.4780799228011485e-06,
+      "loss": 0.7531,
+      "step": 1628
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5427321338903708,
+      "learning_rate": 4.472663398678587e-06,
+      "loss": 0.9477,
+      "step": 1629
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.4003048214565585,
+      "learning_rate": 4.467247500300621e-06,
+      "loss": 0.9261,
+      "step": 1630
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.4365609401677611,
+      "learning_rate": 4.461832234093823e-06,
+      "loss": 0.9833,
+      "step": 1631
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.4467806991079987,
+      "learning_rate": 4.456417606484023e-06,
+      "loss": 0.5682,
+      "step": 1632
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5217097348688442,
+      "learning_rate": 4.451003623896291e-06,
+      "loss": 0.8399,
+      "step": 1633
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6378158731753162,
+      "learning_rate": 4.445590292754927e-06,
+      "loss": 0.8805,
+      "step": 1634
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6662742332374774,
+      "learning_rate": 4.4401776194834615e-06,
+      "loss": 0.7773,
+      "step": 1635
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.46990087570122513,
+      "learning_rate": 4.434765610504647e-06,
+      "loss": 0.7552,
+      "step": 1636
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5298786196045242,
+      "learning_rate": 4.4293542722404435e-06,
+      "loss": 0.8153,
+      "step": 1637
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5632921424314683,
+      "learning_rate": 4.423943611112016e-06,
+      "loss": 0.8954,
+      "step": 1638
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6062475193488646,
+      "learning_rate": 4.418533633539728e-06,
+      "loss": 0.8595,
+      "step": 1639
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5790355828092847,
+      "learning_rate": 4.41312434594313e-06,
+      "loss": 0.9114,
+      "step": 1640
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.4750472955252488,
+      "learning_rate": 4.407715754740954e-06,
+      "loss": 0.5208,
+      "step": 1641
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5167587198663901,
+      "learning_rate": 4.402307866351107e-06,
+      "loss": 0.8998,
+      "step": 1642
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.3600702016641164,
+      "learning_rate": 4.396900687190659e-06,
+      "loss": 0.98,
+      "step": 1643
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5141593925007272,
+      "learning_rate": 4.391494223675843e-06,
+      "loss": 0.8177,
+      "step": 1644
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5295727589707778,
+      "learning_rate": 4.386088482222037e-06,
+      "loss": 0.893,
+      "step": 1645
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.5565869792343807,
+      "learning_rate": 4.380683469243768e-06,
+      "loss": 0.8321,
+      "step": 1646
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6243137014636156,
+      "learning_rate": 4.375279191154695e-06,
+      "loss": 0.8265,
+      "step": 1647
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.4803889778814455,
+      "learning_rate": 4.369875654367605e-06,
+      "loss": 0.9055,
+      "step": 1648
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.6305097036415725,
+      "learning_rate": 4.364472865294409e-06,
+      "loss": 0.7206,
+      "step": 1649
+    },
+    {
+      "epoch": 0.55,
+      "grad_norm": 0.4608768476098127,
+      "learning_rate": 4.359070830346126e-06,
+      "loss": 0.6497,
+      "step": 1650
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.654855756260135,
+      "learning_rate": 4.3536695559328816e-06,
+      "loss": 0.7242,
+      "step": 1651
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5939991935895634,
+      "learning_rate": 4.3482690484639e-06,
+      "loss": 0.7753,
+      "step": 1652
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.49741577589845465,
+      "learning_rate": 4.342869314347498e-06,
+      "loss": 0.8983,
+      "step": 1653
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5333509197466653,
+      "learning_rate": 4.337470359991068e-06,
+      "loss": 0.8919,
+      "step": 1654
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5573297143736206,
+      "learning_rate": 4.33207219180108e-06,
+      "loss": 1.0418,
+      "step": 1655
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.33750367105733453,
+      "learning_rate": 4.326674816183075e-06,
+      "loss": 0.9,
+      "step": 1656
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6397241535113349,
+      "learning_rate": 4.321278239541649e-06,
+      "loss": 0.9371,
+      "step": 1657
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5058084865114281,
+      "learning_rate": 4.31588246828045e-06,
+      "loss": 0.5232,
+      "step": 1658
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5174007183090739,
+      "learning_rate": 4.310487508802174e-06,
+      "loss": 0.8344,
+      "step": 1659
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.4696157440835379,
+      "learning_rate": 4.305093367508551e-06,
+      "loss": 0.7558,
+      "step": 1660
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5035792962973809,
+      "learning_rate": 4.299700050800337e-06,
+      "loss": 0.796,
+      "step": 1661
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5930662784582914,
+      "learning_rate": 4.294307565077318e-06,
+      "loss": 1.0367,
+      "step": 1662
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5748052614245083,
+      "learning_rate": 4.288915916738287e-06,
+      "loss": 0.7909,
+      "step": 1663
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5005428154124586,
+      "learning_rate": 4.283525112181044e-06,
+      "loss": 0.891,
+      "step": 1664
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6110034798927306,
+      "learning_rate": 4.278135157802389e-06,
+      "loss": 0.7243,
+      "step": 1665
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.4718117559806131,
+      "learning_rate": 4.272746059998117e-06,
+      "loss": 1.0389,
+      "step": 1666
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5311510585301572,
+      "learning_rate": 4.267357825162999e-06,
+      "loss": 0.6579,
+      "step": 1667
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5419840741413567,
+      "learning_rate": 4.2619704596907856e-06,
+      "loss": 0.76,
+      "step": 1668
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5393844683634438,
+      "learning_rate": 4.256583969974199e-06,
+      "loss": 0.9486,
+      "step": 1669
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.36552062257454926,
+      "learning_rate": 4.251198362404917e-06,
+      "loss": 0.9132,
+      "step": 1670
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 1.0416246257862691,
+      "learning_rate": 4.2458136433735726e-06,
+      "loss": 0.8655,
+      "step": 1671
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.6576709315022387,
+      "learning_rate": 4.240429819269746e-06,
+      "loss": 0.7652,
+      "step": 1672
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5258396846108064,
+      "learning_rate": 4.235046896481955e-06,
+      "loss": 0.6167,
+      "step": 1673
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5400199487164433,
+      "learning_rate": 4.229664881397645e-06,
+      "loss": 0.8613,
+      "step": 1674
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.72692759485707,
+      "learning_rate": 4.2242837804031855e-06,
+      "loss": 0.8165,
+      "step": 1675
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5520812128422019,
+      "learning_rate": 4.218903599883866e-06,
+      "loss": 1.0002,
+      "step": 1676
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.44495800555916926,
+      "learning_rate": 4.213524346223877e-06,
+      "loss": 0.5472,
+      "step": 1677
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.8789522607702016,
+      "learning_rate": 4.208146025806313e-06,
+      "loss": 0.9414,
+      "step": 1678
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.44530474685533356,
+      "learning_rate": 4.20276864501316e-06,
+      "loss": 1.0071,
+      "step": 1679
+    },
+    {
+      "epoch": 0.56,
+      "grad_norm": 0.5871637449233535,
+      "learning_rate": 4.197392210225289e-06,
+      "loss": 0.771,
+      "step": 1680
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3426723702023918,
+      "learning_rate": 4.192016727822449e-06,
+      "loss": 0.9177,
+      "step": 1681
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5762495660222858,
+      "learning_rate": 4.186642204183259e-06,
+      "loss": 0.8568,
+      "step": 1682
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5504454307957241,
+      "learning_rate": 4.1812686456852e-06,
+      "loss": 0.9695,
+      "step": 1683
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5722210158718394,
+      "learning_rate": 4.175896058704607e-06,
+      "loss": 0.7939,
+      "step": 1684
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.4100755466016813,
+      "learning_rate": 4.170524449616662e-06,
+      "loss": 0.5379,
+      "step": 1685
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5391646334841191,
+      "learning_rate": 4.165153824795391e-06,
+      "loss": 0.9284,
+      "step": 1686
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6206454780454271,
+      "learning_rate": 4.159784190613644e-06,
+      "loss": 0.7553,
+      "step": 1687
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5024240493579889,
+      "learning_rate": 4.154415553443103e-06,
+      "loss": 1.0579,
+      "step": 1688
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6533476153922252,
+      "learning_rate": 4.149047919654263e-06,
+      "loss": 0.8367,
+      "step": 1689
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.45821699462447446,
+      "learning_rate": 4.143681295616429e-06,
+      "loss": 0.9114,
+      "step": 1690
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5410590840117333,
+      "learning_rate": 4.138315687697708e-06,
+      "loss": 0.8965,
+      "step": 1691
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3504688669537082,
+      "learning_rate": 4.1329511022650035e-06,
+      "loss": 0.9134,
+      "step": 1692
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.39770652255574457,
+      "learning_rate": 4.127587545684002e-06,
+      "loss": 0.6199,
+      "step": 1693
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5553931405196691,
+      "learning_rate": 4.122225024319171e-06,
+      "loss": 0.8376,
+      "step": 1694
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5414292930655215,
+      "learning_rate": 4.116863544533747e-06,
+      "loss": 0.8626,
+      "step": 1695
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.47476694946146303,
+      "learning_rate": 4.111503112689737e-06,
+      "loss": 0.6773,
+      "step": 1696
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6774845194655786,
+      "learning_rate": 4.106143735147898e-06,
+      "loss": 0.7867,
+      "step": 1697
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6114438634556283,
+      "learning_rate": 4.100785418267739e-06,
+      "loss": 0.9407,
+      "step": 1698
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6090214728807547,
+      "learning_rate": 4.0954281684075096e-06,
+      "loss": 0.8808,
+      "step": 1699
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6318588072099844,
+      "learning_rate": 4.090071991924194e-06,
+      "loss": 0.8363,
+      "step": 1700
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 1.3103216732697365,
+      "learning_rate": 4.084716895173501e-06,
+      "loss": 0.873,
+      "step": 1701
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.4653664272689995,
+      "learning_rate": 4.0793628845098595e-06,
+      "loss": 0.9245,
+      "step": 1702
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.47130865147811324,
+      "learning_rate": 4.07400996628641e-06,
+      "loss": 0.5327,
+      "step": 1703
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.8029958450502058,
+      "learning_rate": 4.068658146854997e-06,
+      "loss": 0.7653,
+      "step": 1704
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.3929804156714689,
+      "learning_rate": 4.063307432566157e-06,
+      "loss": 0.9728,
+      "step": 1705
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6595009838266079,
+      "learning_rate": 4.057957829769123e-06,
+      "loss": 0.6838,
+      "step": 1706
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5090814687247464,
+      "learning_rate": 4.052609344811802e-06,
+      "loss": 0.9664,
+      "step": 1707
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.5204972716165095,
+      "learning_rate": 4.0472619840407765e-06,
+      "loss": 0.7437,
+      "step": 1708
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.6085589542844478,
+      "learning_rate": 4.041915753801297e-06,
+      "loss": 0.7196,
+      "step": 1709
+    },
+    {
+      "epoch": 0.57,
+      "grad_norm": 0.4809961761780804,
+      "learning_rate": 4.03657066043727e-06,
+      "loss": 0.9404,
+      "step": 1710
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5056159654822807,
+      "learning_rate": 4.031226710291255e-06,
+      "loss": 0.5589,
+      "step": 1711
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.4950908849284099,
+      "learning_rate": 4.025883909704453e-06,
+      "loss": 0.913,
+      "step": 1712
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.48129756896218445,
+      "learning_rate": 4.020542265016703e-06,
+      "loss": 0.9405,
+      "step": 1713
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5567766797982668,
+      "learning_rate": 4.015201782566471e-06,
+      "loss": 0.9483,
+      "step": 1714
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5610663050701714,
+      "learning_rate": 4.009862468690844e-06,
+      "loss": 0.765,
+      "step": 1715
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7741469963549095,
+      "learning_rate": 4.004524329725522e-06,
+      "loss": 0.8117,
+      "step": 1716
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5190392563601843,
+      "learning_rate": 3.999187372004812e-06,
+      "loss": 0.9693,
+      "step": 1717
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.41084148213202487,
+      "learning_rate": 3.993851601861618e-06,
+      "loss": 0.8671,
+      "step": 1718
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5076895692317938,
+      "learning_rate": 3.988517025627437e-06,
+      "loss": 0.8555,
+      "step": 1719
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.4971122268008823,
+      "learning_rate": 3.983183649632348e-06,
+      "loss": 0.7457,
+      "step": 1720
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5137821288742067,
+      "learning_rate": 3.977851480205003e-06,
+      "loss": 0.5961,
+      "step": 1721
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6743884054035312,
+      "learning_rate": 3.972520523672627e-06,
+      "loss": 0.9488,
+      "step": 1722
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.49076569539209464,
+      "learning_rate": 3.967190786361003e-06,
+      "loss": 0.72,
+      "step": 1723
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6006952514960138,
+      "learning_rate": 3.9618622745944675e-06,
+      "loss": 0.9236,
+      "step": 1724
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.596371665102166,
+      "learning_rate": 3.956534994695903e-06,
+      "loss": 0.9156,
+      "step": 1725
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5795182353851348,
+      "learning_rate": 3.951208952986731e-06,
+      "loss": 0.938,
+      "step": 1726
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 1.104840787883151,
+      "learning_rate": 3.945884155786905e-06,
+      "loss": 0.8653,
+      "step": 1727
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5982659933420477,
+      "learning_rate": 3.940560609414894e-06,
+      "loss": 0.8491,
+      "step": 1728
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.48981300443045217,
+      "learning_rate": 3.935238320187695e-06,
+      "loss": 0.8864,
+      "step": 1729
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.3190386067785862,
+      "learning_rate": 3.929917294420804e-06,
+      "loss": 0.7518,
+      "step": 1730
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5684159921872693,
+      "learning_rate": 3.9245975384282216e-06,
+      "loss": 1.0126,
+      "step": 1731
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6633923846986512,
+      "learning_rate": 3.919279058522438e-06,
+      "loss": 0.7637,
+      "step": 1732
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.5019914157624449,
+      "learning_rate": 3.913961861014436e-06,
+      "loss": 0.9712,
+      "step": 1733
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.4898614520038248,
+      "learning_rate": 3.90864595221367e-06,
+      "loss": 0.73,
+      "step": 1734
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.49704729195686215,
+      "learning_rate": 3.903331338428067e-06,
+      "loss": 0.7724,
+      "step": 1735
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.7751677801809973,
+      "learning_rate": 3.89801802596402e-06,
+      "loss": 0.9689,
+      "step": 1736
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.6236615442954563,
+      "learning_rate": 3.892706021126376e-06,
+      "loss": 0.8949,
+      "step": 1737
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.48631257584861187,
+      "learning_rate": 3.887395330218429e-06,
+      "loss": 0.6263,
+      "step": 1738
+    },
+    {
+      "epoch": 0.58,
+      "grad_norm": 0.4837453823511362,
+      "learning_rate": 3.882085959541916e-06,
+      "loss": 0.8238,
+      "step": 1739
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5300157512122439,
+      "learning_rate": 3.876777915397006e-06,
+      "loss": 0.7284,
+      "step": 1740
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5182794615482403,
+      "learning_rate": 3.8714712040822954e-06,
+      "loss": 0.9392,
+      "step": 1741
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.40520799703133076,
+      "learning_rate": 3.866165831894796e-06,
+      "loss": 0.9294,
+      "step": 1742
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.4729566811497528,
+      "learning_rate": 3.860861805129935e-06,
+      "loss": 0.9507,
+      "step": 1743
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 1.0551700335666627,
+      "learning_rate": 3.85555913008154e-06,
+      "loss": 0.7771,
+      "step": 1744
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.46855901103770026,
+      "learning_rate": 3.8502578130418345e-06,
+      "loss": 0.7867,
+      "step": 1745
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6299434711180859,
+      "learning_rate": 3.844957860301434e-06,
+      "loss": 0.7417,
+      "step": 1746
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.3838598863599968,
+      "learning_rate": 3.83965927814933e-06,
+      "loss": 0.6556,
+      "step": 1747
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.567160735268907,
+      "learning_rate": 3.83436207287289e-06,
+      "loss": 0.856,
+      "step": 1748
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.4696206876326631,
+      "learning_rate": 3.829066250757851e-06,
+      "loss": 0.7638,
+      "step": 1749
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5728305928908647,
+      "learning_rate": 3.823771818088303e-06,
+      "loss": 0.9326,
+      "step": 1750
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5752670390529224,
+      "learning_rate": 3.818478781146692e-06,
+      "loss": 0.7268,
+      "step": 1751
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5309094038119506,
+      "learning_rate": 3.813187146213802e-06,
+      "loss": 0.8318,
+      "step": 1752
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.543181315405188,
+      "learning_rate": 3.807896919568761e-06,
+      "loss": 0.9155,
+      "step": 1753
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.32582098986564256,
+      "learning_rate": 3.80260810748902e-06,
+      "loss": 0.9532,
+      "step": 1754
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.602886456589677,
+      "learning_rate": 3.7973207162503535e-06,
+      "loss": 0.8041,
+      "step": 1755
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.4484435848311491,
+      "learning_rate": 3.7920347521268514e-06,
+      "loss": 0.5636,
+      "step": 1756
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.7757367630243744,
+      "learning_rate": 3.7867502213909075e-06,
+      "loss": 1.0233,
+      "step": 1757
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.4383502078731391,
+      "learning_rate": 3.781467130313215e-06,
+      "loss": 0.7665,
+      "step": 1758
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.4699757606069799,
+      "learning_rate": 3.776185485162762e-06,
+      "loss": 0.8638,
+      "step": 1759
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.4978668982478577,
+      "learning_rate": 3.7709052922068174e-06,
+      "loss": 0.9504,
+      "step": 1760
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6200391979454231,
+      "learning_rate": 3.765626557710928e-06,
+      "loss": 0.6947,
+      "step": 1761
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5177293588149458,
+      "learning_rate": 3.7603492879389093e-06,
+      "loss": 0.8569,
+      "step": 1762
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.49056865694065527,
+      "learning_rate": 3.7550734891528413e-06,
+      "loss": 0.8473,
+      "step": 1763
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.44464761136719305,
+      "learning_rate": 3.7497991676130563e-06,
+      "loss": 0.6317,
+      "step": 1764
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5152733470493812,
+      "learning_rate": 3.744526329578132e-06,
+      "loss": 0.9208,
+      "step": 1765
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.6267475959503844,
+      "learning_rate": 3.73925498130489e-06,
+      "loss": 0.7136,
+      "step": 1766
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5306204363754503,
+      "learning_rate": 3.733985129048382e-06,
+      "loss": 0.9129,
+      "step": 1767
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.3348631196374087,
+      "learning_rate": 3.728716779061882e-06,
+      "loss": 0.8758,
+      "step": 1768
+    },
+    {
+      "epoch": 0.59,
+      "grad_norm": 0.5575559439181165,
+      "learning_rate": 3.7234499375968876e-06,
+      "loss": 0.9347,
+      "step": 1769
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.4826564510311279,
+      "learning_rate": 3.7181846109031007e-06,
+      "loss": 0.881,
+      "step": 1770
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.523162092629139,
+      "learning_rate": 3.712920805228427e-06,
+      "loss": 0.7164,
+      "step": 1771
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.48031942999269306,
+      "learning_rate": 3.707658526818968e-06,
+      "loss": 0.9519,
+      "step": 1772
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.8272739378212608,
+      "learning_rate": 3.7023977819190136e-06,
+      "loss": 0.7006,
+      "step": 1773
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.4375376397823873,
+      "learning_rate": 3.6971385767710345e-06,
+      "loss": 0.5658,
+      "step": 1774
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5635364784117248,
+      "learning_rate": 3.6918809176156713e-06,
+      "loss": 0.9221,
+      "step": 1775
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5516074523179556,
+      "learning_rate": 3.686624810691733e-06,
+      "loss": 0.9355,
+      "step": 1776
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.9450803580284826,
+      "learning_rate": 3.6813702622361858e-06,
+      "loss": 0.8668,
+      "step": 1777
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5142402918265898,
+      "learning_rate": 3.6761172784841446e-06,
+      "loss": 0.7809,
+      "step": 1778
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.34737334548936966,
+      "learning_rate": 3.6708658656688723e-06,
+      "loss": 0.9344,
+      "step": 1779
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5098827905720009,
+      "learning_rate": 3.6656160300217637e-06,
+      "loss": 0.7941,
+      "step": 1780
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.011370369940968,
+      "learning_rate": 3.6603677777723425e-06,
+      "loss": 0.942,
+      "step": 1781
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.4251071659328544,
+      "learning_rate": 3.655121115148254e-06,
+      "loss": 0.8533,
+      "step": 1782
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5962009042606876,
+      "learning_rate": 3.64987604837526e-06,
+      "loss": 0.6461,
+      "step": 1783
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6291501079857711,
+      "learning_rate": 3.6446325836772244e-06,
+      "loss": 0.8715,
+      "step": 1784
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6449008249075718,
+      "learning_rate": 3.639390727276111e-06,
+      "loss": 0.7854,
+      "step": 1785
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.48011901889269026,
+      "learning_rate": 3.6341504853919778e-06,
+      "loss": 0.9549,
+      "step": 1786
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5256832697747604,
+      "learning_rate": 3.6289118642429654e-06,
+      "loss": 0.7743,
+      "step": 1787
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6384089754774182,
+      "learning_rate": 3.6236748700452894e-06,
+      "loss": 0.8612,
+      "step": 1788
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5704594538836889,
+      "learning_rate": 3.618439509013236e-06,
+      "loss": 0.7515,
+      "step": 1789
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.62960178463794,
+      "learning_rate": 3.613205787359157e-06,
+      "loss": 0.8888,
+      "step": 1790
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5346459506189383,
+      "learning_rate": 3.6079737112934533e-06,
+      "loss": 0.6506,
+      "step": 1791
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.3465547530960333,
+      "learning_rate": 3.602743287024576e-06,
+      "loss": 0.858,
+      "step": 1792
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5555087676605183,
+      "learning_rate": 3.597514520759016e-06,
+      "loss": 0.8775,
+      "step": 1793
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.7529010557117054,
+      "learning_rate": 3.5922874187012977e-06,
+      "loss": 0.9143,
+      "step": 1794
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.4262321352202729,
+      "learning_rate": 3.587061987053967e-06,
+      "loss": 0.9114,
+      "step": 1795
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 1.0127486215061159,
+      "learning_rate": 3.581838232017593e-06,
+      "loss": 0.7504,
+      "step": 1796
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.4807454738227805,
+      "learning_rate": 3.576616159790751e-06,
+      "loss": 0.8519,
+      "step": 1797
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.5567753957804998,
+      "learning_rate": 3.5713957765700224e-06,
+      "loss": 1.0036,
+      "step": 1798
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 0.6189801598143124,
+      "learning_rate": 3.5661770885499808e-06,
+      "loss": 0.7549,
+      "step": 1799
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.37748578864875776,
+      "learning_rate": 3.5609601019231934e-06,
+      "loss": 0.5658,
+      "step": 1800
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8072328568957448,
+      "learning_rate": 3.5557448228802026e-06,
+      "loss": 0.8779,
+      "step": 1801
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5606673917545589,
+      "learning_rate": 3.5505312576095295e-06,
+      "loss": 0.9277,
+      "step": 1802
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.38317658772783064,
+      "learning_rate": 3.54531941229766e-06,
+      "loss": 0.9191,
+      "step": 1803
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5630254141175071,
+      "learning_rate": 3.5401092931290378e-06,
+      "loss": 0.7995,
+      "step": 1804
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7111975604383062,
+      "learning_rate": 3.5349009062860586e-06,
+      "loss": 0.8625,
+      "step": 1805
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.4617031966343969,
+      "learning_rate": 3.5296942579490645e-06,
+      "loss": 0.8627,
+      "step": 1806
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5257572184412697,
+      "learning_rate": 3.5244893542963332e-06,
+      "loss": 0.8993,
+      "step": 1807
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5548890766896545,
+      "learning_rate": 3.519286201504073e-06,
+      "loss": 0.8563,
+      "step": 1808
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.4480229841127608,
+      "learning_rate": 3.5140848057464107e-06,
+      "loss": 0.6205,
+      "step": 1809
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5381315893103058,
+      "learning_rate": 3.5088851731953956e-06,
+      "loss": 0.9439,
+      "step": 1810
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6621396658337527,
+      "learning_rate": 3.503687310020978e-06,
+      "loss": 0.7924,
+      "step": 1811
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.48835581352091967,
+      "learning_rate": 3.4984912223910105e-06,
+      "loss": 0.8441,
+      "step": 1812
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5734000823306113,
+      "learning_rate": 3.493296916471243e-06,
+      "loss": 0.7608,
+      "step": 1813
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6237914313274808,
+      "learning_rate": 3.488104398425304e-06,
+      "loss": 0.8072,
+      "step": 1814
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.7013368223275397,
+      "learning_rate": 3.482913674414705e-06,
+      "loss": 0.9123,
+      "step": 1815
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.34143918023947534,
+      "learning_rate": 3.47772475059883e-06,
+      "loss": 0.9201,
+      "step": 1816
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.43677421331111593,
+      "learning_rate": 3.4725376331349224e-06,
+      "loss": 0.9849,
+      "step": 1817
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.44758171067742575,
+      "learning_rate": 3.4673523281780856e-06,
+      "loss": 0.5412,
+      "step": 1818
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.47802753104803025,
+      "learning_rate": 3.46216884188127e-06,
+      "loss": 0.8337,
+      "step": 1819
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5807092139973227,
+      "learning_rate": 3.4569871803952702e-06,
+      "loss": 0.8786,
+      "step": 1820
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.6287239898856064,
+      "learning_rate": 3.4518073498687145e-06,
+      "loss": 0.9541,
+      "step": 1821
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.4818683936240644,
+      "learning_rate": 3.4466293564480562e-06,
+      "loss": 0.645,
+      "step": 1822
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5966965896522384,
+      "learning_rate": 3.4414532062775743e-06,
+      "loss": 0.861,
+      "step": 1823
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.527424190501117,
+      "learning_rate": 3.4362789054993544e-06,
+      "loss": 0.8838,
+      "step": 1824
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.8255468143237361,
+      "learning_rate": 3.4311064602532906e-06,
+      "loss": 0.8069,
+      "step": 1825
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5173562314854863,
+      "learning_rate": 3.425935876677077e-06,
+      "loss": 0.6389,
+      "step": 1826
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5197397067217455,
+      "learning_rate": 3.420767160906194e-06,
+      "loss": 0.7936,
+      "step": 1827
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.3199622469120037,
+      "learning_rate": 3.4156003190739108e-06,
+      "loss": 0.9455,
+      "step": 1828
+    },
+    {
+      "epoch": 0.61,
+      "grad_norm": 0.5305978084208749,
+      "learning_rate": 3.4104353573112662e-06,
+      "loss": 0.9973,
+      "step": 1829
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5718203913632003,
+      "learning_rate": 3.4052722817470767e-06,
+      "loss": 0.8111,
+      "step": 1830
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5690999099198052,
+      "learning_rate": 3.4001110985079145e-06,
+      "loss": 0.9052,
+      "step": 1831
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5698912785621042,
+      "learning_rate": 3.3949518137181058e-06,
+      "loss": 0.7835,
+      "step": 1832
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.7307202038708004,
+      "learning_rate": 3.38979443349973e-06,
+      "loss": 0.8955,
+      "step": 1833
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.498673177205537,
+      "learning_rate": 3.3846389639726007e-06,
+      "loss": 0.8303,
+      "step": 1834
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5179516847514546,
+      "learning_rate": 3.379485411254266e-06,
+      "loss": 0.7292,
+      "step": 1835
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3946016451481151,
+      "learning_rate": 3.374333781460001e-06,
+      "loss": 0.6036,
+      "step": 1836
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.598513543989879,
+      "learning_rate": 3.3691840807027965e-06,
+      "loss": 0.7623,
+      "step": 1837
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6540813505100903,
+      "learning_rate": 3.3640363150933574e-06,
+      "loss": 0.8654,
+      "step": 1838
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.48331320858379273,
+      "learning_rate": 3.3588904907400876e-06,
+      "loss": 0.7727,
+      "step": 1839
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.49276007656887816,
+      "learning_rate": 3.3537466137490937e-06,
+      "loss": 0.8915,
+      "step": 1840
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5195281524935695,
+      "learning_rate": 3.3486046902241663e-06,
+      "loss": 1.011,
+      "step": 1841
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.33837899651742687,
+      "learning_rate": 3.34346472626678e-06,
+      "loss": 0.9043,
+      "step": 1842
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.612034133859254,
+      "learning_rate": 3.3383267279760843e-06,
+      "loss": 0.9306,
+      "step": 1843
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.41101293408438294,
+      "learning_rate": 3.333190701448897e-06,
+      "loss": 0.5535,
+      "step": 1844
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5578810807160154,
+      "learning_rate": 3.328056652779694e-06,
+      "loss": 0.9232,
+      "step": 1845
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5030425184687053,
+      "learning_rate": 3.3229245880606063e-06,
+      "loss": 0.5651,
+      "step": 1846
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5710601831480298,
+      "learning_rate": 3.3177945133814093e-06,
+      "loss": 0.8618,
+      "step": 1847
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5639501123403139,
+      "learning_rate": 3.312666434829518e-06,
+      "loss": 0.9711,
+      "step": 1848
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6239011742667414,
+      "learning_rate": 3.307540358489977e-06,
+      "loss": 0.7882,
+      "step": 1849
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5050976539218537,
+      "learning_rate": 3.3024162904454584e-06,
+      "loss": 0.8698,
+      "step": 1850
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6163341487026115,
+      "learning_rate": 3.2972942367762473e-06,
+      "loss": 0.8869,
+      "step": 1851
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.48587036119961446,
+      "learning_rate": 3.2921742035602404e-06,
+      "loss": 0.9044,
+      "step": 1852
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.42015660763113055,
+      "learning_rate": 3.2870561968729388e-06,
+      "loss": 0.652,
+      "step": 1853
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.47637663089326204,
+      "learning_rate": 3.2819402227874364e-06,
+      "loss": 0.8308,
+      "step": 1854
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5607023240636131,
+      "learning_rate": 3.276826287374414e-06,
+      "loss": 0.9121,
+      "step": 1855
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.3848917309707498,
+      "learning_rate": 3.271714396702137e-06,
+      "loss": 0.8892,
+      "step": 1856
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.6060668348333976,
+      "learning_rate": 3.2666045568364423e-06,
+      "loss": 0.8466,
+      "step": 1857
+    },
+    {
+      "epoch": 0.62,
+      "grad_norm": 0.5292502119702298,
+      "learning_rate": 3.2614967738407332e-06,
+      "loss": 0.6268,
+      "step": 1858
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.595431980992499,
+      "learning_rate": 3.2563910537759723e-06,
+      "loss": 0.7141,
+      "step": 1859
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5445299217602267,
+      "learning_rate": 3.2512874027006753e-06,
+      "loss": 0.8558,
+      "step": 1860
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5993224878096026,
+      "learning_rate": 3.2461858266709017e-06,
+      "loss": 0.8876,
+      "step": 1861
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.47936135906865374,
+      "learning_rate": 3.2410863317402486e-06,
+      "loss": 0.8934,
+      "step": 1862
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.4380890709840872,
+      "learning_rate": 3.2359889239598456e-06,
+      "loss": 0.5011,
+      "step": 1863
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5192261605950195,
+      "learning_rate": 3.2308936093783436e-06,
+      "loss": 0.8954,
+      "step": 1864
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5821341564983684,
+      "learning_rate": 3.225800394041911e-06,
+      "loss": 0.9828,
+      "step": 1865
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5571514797245287,
+      "learning_rate": 3.220709283994222e-06,
+      "loss": 0.7816,
+      "step": 1866
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3312677059870761,
+      "learning_rate": 3.2156202852764584e-06,
+      "loss": 0.9352,
+      "step": 1867
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6298567154258303,
+      "learning_rate": 3.2105334039272924e-06,
+      "loss": 0.7854,
+      "step": 1868
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.4987605984536881,
+      "learning_rate": 3.205448645982883e-06,
+      "loss": 0.8621,
+      "step": 1869
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.46794152738961986,
+      "learning_rate": 3.2003660174768746e-06,
+      "loss": 0.5138,
+      "step": 1870
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5350412354782874,
+      "learning_rate": 3.195285524440379e-06,
+      "loss": 0.8317,
+      "step": 1871
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5702337176653438,
+      "learning_rate": 3.1902071729019767e-06,
+      "loss": 0.8383,
+      "step": 1872
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6810042046580277,
+      "learning_rate": 3.1851309688877074e-06,
+      "loss": 0.7792,
+      "step": 1873
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5093316328584271,
+      "learning_rate": 3.1800569184210627e-06,
+      "loss": 1.0397,
+      "step": 1874
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5017777267867418,
+      "learning_rate": 3.1749850275229777e-06,
+      "loss": 0.9246,
+      "step": 1875
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.49513268569761604,
+      "learning_rate": 3.1699153022118236e-06,
+      "loss": 0.8003,
+      "step": 1876
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.8814349951740265,
+      "learning_rate": 3.1648477485034062e-06,
+      "loss": 0.9437,
+      "step": 1877
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.3413893180517348,
+      "learning_rate": 3.15978237241095e-06,
+      "loss": 0.9032,
+      "step": 1878
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.4432456342190768,
+      "learning_rate": 3.1547191799450976e-06,
+      "loss": 0.5756,
+      "step": 1879
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5373337305065599,
+      "learning_rate": 3.1496581771139013e-06,
+      "loss": 0.7943,
+      "step": 1880
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.7189243824944811,
+      "learning_rate": 3.1445993699228132e-06,
+      "loss": 0.9619,
+      "step": 1881
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.4740013308443823,
+      "learning_rate": 3.1395427643746802e-06,
+      "loss": 0.6436,
+      "step": 1882
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.999644659074809,
+      "learning_rate": 3.134488366469739e-06,
+      "loss": 0.9534,
+      "step": 1883
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.560531910314043,
+      "learning_rate": 3.1294361822056048e-06,
+      "loss": 0.8153,
+      "step": 1884
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5403980356449571,
+      "learning_rate": 3.1243862175772656e-06,
+      "loss": 0.8422,
+      "step": 1885
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.5578988890380708,
+      "learning_rate": 3.1193384785770755e-06,
+      "loss": 0.8249,
+      "step": 1886
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.6169933264675781,
+      "learning_rate": 3.1142929711947502e-06,
+      "loss": 0.8752,
+      "step": 1887
+    },
+    {
+      "epoch": 0.63,
+      "grad_norm": 0.459891854118218,
+      "learning_rate": 3.109249701417355e-06,
+      "loss": 0.9022,
+      "step": 1888
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.49279102575159245,
+      "learning_rate": 3.1042086752292995e-06,
+      "loss": 0.5125,
+      "step": 1889
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3393436755492301,
+      "learning_rate": 3.0991698986123343e-06,
+      "loss": 0.9184,
+      "step": 1890
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.46190464415957266,
+      "learning_rate": 3.0941333775455367e-06,
+      "loss": 0.8299,
+      "step": 1891
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.7964899272467584,
+      "learning_rate": 3.0890991180053094e-06,
+      "loss": 0.7235,
+      "step": 1892
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.4658346609012984,
+      "learning_rate": 3.0840671259653714e-06,
+      "loss": 0.9616,
+      "step": 1893
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5693538316528126,
+      "learning_rate": 3.0790374073967523e-06,
+      "loss": 0.7154,
+      "step": 1894
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5932146683716092,
+      "learning_rate": 3.074009968267782e-06,
+      "loss": 0.9642,
+      "step": 1895
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 2.403683293904165,
+      "learning_rate": 3.068984814544087e-06,
+      "loss": 0.82,
+      "step": 1896
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.470475940953703,
+      "learning_rate": 3.0639619521885795e-06,
+      "loss": 0.5131,
+      "step": 1897
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.44558497316759244,
+      "learning_rate": 3.058941387161456e-06,
+      "loss": 0.9897,
+      "step": 1898
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.538082168249906,
+      "learning_rate": 3.0539231254201855e-06,
+      "loss": 0.8097,
+      "step": 1899
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5280669459051074,
+      "learning_rate": 3.0489071729195025e-06,
+      "loss": 0.942,
+      "step": 1900
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5918853176303486,
+      "learning_rate": 3.043893535611404e-06,
+      "loss": 0.7792,
+      "step": 1901
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 1.0230408647862979,
+      "learning_rate": 3.0388822194451385e-06,
+      "loss": 0.7824,
+      "step": 1902
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5420715803834221,
+      "learning_rate": 3.0338732303671993e-06,
+      "loss": 0.9314,
+      "step": 1903
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.33510207652426965,
+      "learning_rate": 3.0288665743213186e-06,
+      "loss": 0.8743,
+      "step": 1904
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5895406934349117,
+      "learning_rate": 3.023862257248462e-06,
+      "loss": 0.8782,
+      "step": 1905
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5728828351021552,
+      "learning_rate": 3.0188602850868186e-06,
+      "loss": 0.7121,
+      "step": 1906
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3916996715760614,
+      "learning_rate": 3.0138606637717925e-06,
+      "loss": 0.6927,
+      "step": 1907
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6068048157307404,
+      "learning_rate": 3.0088633992360037e-06,
+      "loss": 0.8416,
+      "step": 1908
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5047397222733783,
+      "learning_rate": 3.003868497409271e-06,
+      "loss": 0.7064,
+      "step": 1909
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.4424834981335316,
+      "learning_rate": 2.99887596421861e-06,
+      "loss": 0.9898,
+      "step": 1910
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.6239314015299207,
+      "learning_rate": 2.9938858055882287e-06,
+      "loss": 0.815,
+      "step": 1911
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.582422091020291,
+      "learning_rate": 2.988898027439514e-06,
+      "loss": 0.8833,
+      "step": 1912
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5714704589909797,
+      "learning_rate": 2.9839126356910308e-06,
+      "loss": 0.9195,
+      "step": 1913
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5300425116523337,
+      "learning_rate": 2.9789296362585084e-06,
+      "loss": 0.9353,
+      "step": 1914
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.4961636605614059,
+      "learning_rate": 2.973949035054843e-06,
+      "loss": 0.7582,
+      "step": 1915
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.3258971797950794,
+      "learning_rate": 2.96897083799008e-06,
+      "loss": 0.7428,
+      "step": 1916
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.5640451520123276,
+      "learning_rate": 2.9639950509714138e-06,
+      "loss": 1.0226,
+      "step": 1917
+    },
+    {
+      "epoch": 0.64,
+      "grad_norm": 0.546900376523398,
+      "learning_rate": 2.9590216799031814e-06,
+      "loss": 0.8171,
+      "step": 1918
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.45645314631997325,
+      "learning_rate": 2.954050730686848e-06,
+      "loss": 0.9346,
+      "step": 1919
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.501742446465402,
+      "learning_rate": 2.949082209221009e-06,
+      "loss": 0.743,
+      "step": 1920
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.547640898012896,
+      "learning_rate": 2.9441161214013787e-06,
+      "loss": 0.8157,
+      "step": 1921
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5822246635632062,
+      "learning_rate": 2.939152473120781e-06,
+      "loss": 0.9736,
+      "step": 1922
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.38198736384838655,
+      "learning_rate": 2.934191270269148e-06,
+      "loss": 0.6966,
+      "step": 1923
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5522918317273303,
+      "learning_rate": 2.929232518733507e-06,
+      "loss": 0.8801,
+      "step": 1924
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7368451308819847,
+      "learning_rate": 2.9242762243979804e-06,
+      "loss": 0.8529,
+      "step": 1925
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5371721601155082,
+      "learning_rate": 2.919322393143772e-06,
+      "loss": 0.8893,
+      "step": 1926
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.7199441724676955,
+      "learning_rate": 2.9143710308491625e-06,
+      "loss": 0.7326,
+      "step": 1927
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.3338415897793069,
+      "learning_rate": 2.909422143389504e-06,
+      "loss": 0.8957,
+      "step": 1928
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5156284395021431,
+      "learning_rate": 2.9044757366372123e-06,
+      "loss": 0.9805,
+      "step": 1929
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5619952878903333,
+      "learning_rate": 2.8995318164617614e-06,
+      "loss": 0.7045,
+      "step": 1930
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.4808778031825857,
+      "learning_rate": 2.8945903887296686e-06,
+      "loss": 0.8084,
+      "step": 1931
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5861772106018788,
+      "learning_rate": 2.889651459304499e-06,
+      "loss": 0.781,
+      "step": 1932
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.379730183607628,
+      "learning_rate": 2.884715034046854e-06,
+      "loss": 0.7148,
+      "step": 1933
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5527120725268724,
+      "learning_rate": 2.8797811188143572e-06,
+      "loss": 0.7211,
+      "step": 1934
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5095812023210308,
+      "learning_rate": 2.874849719461662e-06,
+      "loss": 0.7928,
+      "step": 1935
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.4639003327821117,
+      "learning_rate": 2.8699208418404273e-06,
+      "loss": 0.8298,
+      "step": 1936
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6938041630578211,
+      "learning_rate": 2.8649944917993278e-06,
+      "loss": 0.7968,
+      "step": 1937
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.8822191596546308,
+      "learning_rate": 2.860070675184036e-06,
+      "loss": 0.8483,
+      "step": 1938
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6094423300256213,
+      "learning_rate": 2.8551493978372158e-06,
+      "loss": 0.8589,
+      "step": 1939
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.34798436737699284,
+      "learning_rate": 2.8502306655985207e-06,
+      "loss": 0.9553,
+      "step": 1940
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.44624552342852614,
+      "learning_rate": 2.845314484304586e-06,
+      "loss": 0.5698,
+      "step": 1941
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.6048726537771186,
+      "learning_rate": 2.840400859789013e-06,
+      "loss": 0.8433,
+      "step": 1942
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5861507892303777,
+      "learning_rate": 2.835489797882377e-06,
+      "loss": 0.9103,
+      "step": 1943
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5336879314966877,
+      "learning_rate": 2.83058130441221e-06,
+      "loss": 0.7481,
+      "step": 1944
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.49816537087118956,
+      "learning_rate": 2.8256753852029917e-06,
+      "loss": 0.8734,
+      "step": 1945
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.4961878388729827,
+      "learning_rate": 2.8207720460761523e-06,
+      "loss": 0.9308,
+      "step": 1946
+    },
+    {
+      "epoch": 0.65,
+      "grad_norm": 0.5774212853036671,
+      "learning_rate": 2.8158712928500607e-06,
+      "loss": 0.7168,
+      "step": 1947
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.492460478300681,
+      "learning_rate": 2.810973131340013e-06,
+      "loss": 0.8061,
+      "step": 1948
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.469097382123378,
+      "learning_rate": 2.806077567358233e-06,
+      "loss": 0.878,
+      "step": 1949
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.4029117318093482,
+      "learning_rate": 2.8011846067138648e-06,
+      "loss": 0.6738,
+      "step": 1950
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5478973616344845,
+      "learning_rate": 2.7962942552129558e-06,
+      "loss": 0.7945,
+      "step": 1951
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.549228846783837,
+      "learning_rate": 2.7914065186584637e-06,
+      "loss": 0.7779,
+      "step": 1952
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.483421721212677,
+      "learning_rate": 2.786521402850244e-06,
+      "loss": 0.8602,
+      "step": 1953
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.32985093269606225,
+      "learning_rate": 2.7816389135850353e-06,
+      "loss": 0.8625,
+      "step": 1954
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5573709744688579,
+      "learning_rate": 2.776759056656468e-06,
+      "loss": 0.9669,
+      "step": 1955
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.4301358019790337,
+      "learning_rate": 2.771881837855041e-06,
+      "loss": 0.8798,
+      "step": 1956
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5544659163942873,
+      "learning_rate": 2.767007262968129e-06,
+      "loss": 0.9729,
+      "step": 1957
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5340477148013214,
+      "learning_rate": 2.762135337779969e-06,
+      "loss": 0.7712,
+      "step": 1958
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.45108973759002274,
+      "learning_rate": 2.757266068071648e-06,
+      "loss": 0.6752,
+      "step": 1959
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.35653488608483824,
+      "learning_rate": 2.752399459621108e-06,
+      "loss": 0.5459,
+      "step": 1960
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6533323304666467,
+      "learning_rate": 2.7475355182031315e-06,
+      "loss": 0.9381,
+      "step": 1961
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5604969525685909,
+      "learning_rate": 2.7426742495893343e-06,
+      "loss": 0.923,
+      "step": 1962
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.6732958804147491,
+      "learning_rate": 2.7378156595481625e-06,
+      "loss": 0.8947,
+      "step": 1963
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5617193639005219,
+      "learning_rate": 2.732959753844886e-06,
+      "loss": 0.7525,
+      "step": 1964
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.33012607198778804,
+      "learning_rate": 2.7281065382415822e-06,
+      "loss": 0.9396,
+      "step": 1965
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.4885539599925324,
+      "learning_rate": 2.7232560184971437e-06,
+      "loss": 0.7558,
+      "step": 1966
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5445071163642041,
+      "learning_rate": 2.718408200367263e-06,
+      "loss": 0.9151,
+      "step": 1967
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.43789295071280343,
+      "learning_rate": 2.713563089604422e-06,
+      "loss": 0.8975,
+      "step": 1968
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.517911045295225,
+      "learning_rate": 2.7087206919578932e-06,
+      "loss": 0.6327,
+      "step": 1969
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5461316457767759,
+      "learning_rate": 2.7038810131737346e-06,
+      "loss": 0.8574,
+      "step": 1970
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.75016661074732,
+      "learning_rate": 2.6990440589947676e-06,
+      "loss": 0.764,
+      "step": 1971
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5006843750148681,
+      "learning_rate": 2.6942098351605883e-06,
+      "loss": 1.0153,
+      "step": 1972
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5934586041637517,
+      "learning_rate": 2.689378347407553e-06,
+      "loss": 0.7256,
+      "step": 1973
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5850149695741528,
+      "learning_rate": 2.684549601468764e-06,
+      "loss": 0.8661,
+      "step": 1974
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5698917204365294,
+      "learning_rate": 2.6797236030740803e-06,
+      "loss": 0.7371,
+      "step": 1975
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.5574833267312963,
+      "learning_rate": 2.674900357950091e-06,
+      "loss": 1.0014,
+      "step": 1976
+    },
+    {
+      "epoch": 0.66,
+      "grad_norm": 0.4385691231914093,
+      "learning_rate": 2.670079871820125e-06,
+      "loss": 0.558,
+      "step": 1977
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.43706678202612986,
+      "learning_rate": 2.6652621504042366e-06,
+      "loss": 0.8431,
+      "step": 1978
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6205059403510563,
+      "learning_rate": 2.660447199419194e-06,
+      "loss": 0.915,
+      "step": 1979
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5782150469643013,
+      "learning_rate": 2.6556350245784833e-06,
+      "loss": 0.9862,
+      "step": 1980
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.45153606496773036,
+      "learning_rate": 2.6508256315922975e-06,
+      "loss": 0.8284,
+      "step": 1981
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5807334811924458,
+      "learning_rate": 2.6460190261675223e-06,
+      "loss": 0.8363,
+      "step": 1982
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5339517010985487,
+      "learning_rate": 2.6412152140077408e-06,
+      "loss": 0.7827,
+      "step": 1983
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5355624315300492,
+      "learning_rate": 2.6364142008132216e-06,
+      "loss": 0.9996,
+      "step": 1984
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.7156927477830018,
+      "learning_rate": 2.631615992280907e-06,
+      "loss": 0.8278,
+      "step": 1985
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.42228233000784815,
+      "learning_rate": 2.626820594104418e-06,
+      "loss": 0.5765,
+      "step": 1986
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.4865121323943808,
+      "learning_rate": 2.6220280119740376e-06,
+      "loss": 0.7715,
+      "step": 1987
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.572759071759359,
+      "learning_rate": 2.6172382515767046e-06,
+      "loss": 0.9555,
+      "step": 1988
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.32957984334418544,
+      "learning_rate": 2.6124513185960132e-06,
+      "loss": 0.9025,
+      "step": 1989
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5625099841820571,
+      "learning_rate": 2.6076672187122043e-06,
+      "loss": 0.8181,
+      "step": 1990
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.505616223430113,
+      "learning_rate": 2.6028859576021492e-06,
+      "loss": 0.747,
+      "step": 1991
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.49058851009612325,
+      "learning_rate": 2.5981075409393587e-06,
+      "loss": 0.9298,
+      "step": 1992
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5414019807809275,
+      "learning_rate": 2.593331974393966e-06,
+      "loss": 0.896,
+      "step": 1993
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.38314390819706295,
+      "learning_rate": 2.588559263632719e-06,
+      "loss": 0.5987,
+      "step": 1994
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.4550743751311197,
+      "learning_rate": 2.5837894143189825e-06,
+      "loss": 0.9275,
+      "step": 1995
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6144314546755157,
+      "learning_rate": 2.57902243211272e-06,
+      "loss": 0.8295,
+      "step": 1996
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.6657530980893746,
+      "learning_rate": 2.574258322670496e-06,
+      "loss": 0.7802,
+      "step": 1997
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5507700384847687,
+      "learning_rate": 2.5694970916454686e-06,
+      "loss": 0.8252,
+      "step": 1998
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5484152226804436,
+      "learning_rate": 2.5647387446873746e-06,
+      "loss": 0.7993,
+      "step": 1999
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.9614734243270157,
+      "learning_rate": 2.5599832874425325e-06,
+      "loss": 0.9905,
+      "step": 2000
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5061650213136163,
+      "learning_rate": 2.555230725553832e-06,
+      "loss": 0.7535,
+      "step": 2001
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.31157872627908306,
+      "learning_rate": 2.550481064660724e-06,
+      "loss": 0.906,
+      "step": 2002
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.4433468384706156,
+      "learning_rate": 2.54573431039922e-06,
+      "loss": 0.9743,
+      "step": 2003
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.41404930201025614,
+      "learning_rate": 2.540990468401883e-06,
+      "loss": 0.4887,
+      "step": 2004
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.512826602792586,
+      "learning_rate": 2.536249544297815e-06,
+      "loss": 0.9134,
+      "step": 2005
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5178124623897756,
+      "learning_rate": 2.531511543712662e-06,
+      "loss": 0.9071,
+      "step": 2006
+    },
+    {
+      "epoch": 0.67,
+      "grad_norm": 0.5944350920911026,
+      "learning_rate": 2.5267764722685995e-06,
+      "loss": 0.9112,
+      "step": 2007
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5222140965608532,
+      "learning_rate": 2.522044335584322e-06,
+      "loss": 0.6991,
+      "step": 2008
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.8524325640307062,
+      "learning_rate": 2.5173151392750477e-06,
+      "loss": 0.8388,
+      "step": 2009
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5881070297039829,
+      "learning_rate": 2.5125888889525057e-06,
+      "loss": 0.9017,
+      "step": 2010
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6837643732107077,
+      "learning_rate": 2.5078655902249233e-06,
+      "loss": 0.7433,
+      "step": 2011
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 2.059292547121155,
+      "learning_rate": 2.503145248697033e-06,
+      "loss": 0.6572,
+      "step": 2012
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.613634907138112,
+      "learning_rate": 2.4984278699700504e-06,
+      "loss": 0.7319,
+      "step": 2013
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.3506397034223148,
+      "learning_rate": 2.4937134596416823e-06,
+      "loss": 0.9978,
+      "step": 2014
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5892322502045387,
+      "learning_rate": 2.489002023306112e-06,
+      "loss": 0.8183,
+      "step": 2015
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5206630751896557,
+      "learning_rate": 2.4842935665539873e-06,
+      "loss": 0.8157,
+      "step": 2016
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5500676402974509,
+      "learning_rate": 2.4795880949724296e-06,
+      "loss": 0.9261,
+      "step": 2017
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5656523478497603,
+      "learning_rate": 2.4748856141450132e-06,
+      "loss": 0.7118,
+      "step": 2018
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6030626070369948,
+      "learning_rate": 2.4701861296517623e-06,
+      "loss": 0.9683,
+      "step": 2019
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5188565704866509,
+      "learning_rate": 2.465489647069148e-06,
+      "loss": 0.7919,
+      "step": 2020
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6100590732609551,
+      "learning_rate": 2.4607961719700806e-06,
+      "loss": 0.8395,
+      "step": 2021
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.37960001248473657,
+      "learning_rate": 2.4561057099238973e-06,
+      "loss": 0.5856,
+      "step": 2022
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5424096800062864,
+      "learning_rate": 2.4514182664963633e-06,
+      "loss": 0.7128,
+      "step": 2023
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5020077723596837,
+      "learning_rate": 2.4467338472496633e-06,
+      "loss": 0.8541,
+      "step": 2024
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.4822777364583031,
+      "learning_rate": 2.4420524577423872e-06,
+      "loss": 0.7694,
+      "step": 2025
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.4752102312231687,
+      "learning_rate": 2.4373741035295357e-06,
+      "loss": 0.8923,
+      "step": 2026
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.31223775686423055,
+      "learning_rate": 2.432698790162508e-06,
+      "loss": 0.9871,
+      "step": 2027
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.6874330442781562,
+      "learning_rate": 2.4280265231890887e-06,
+      "loss": 0.7572,
+      "step": 2028
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.44792042044647135,
+      "learning_rate": 2.423357308153454e-06,
+      "loss": 0.6539,
+      "step": 2029
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.542522614078808,
+      "learning_rate": 2.418691150596158e-06,
+      "loss": 0.7422,
+      "step": 2030
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.4643405625957246,
+      "learning_rate": 2.4140280560541214e-06,
+      "loss": 0.8933,
+      "step": 2031
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5283904930024123,
+      "learning_rate": 2.409368030060637e-06,
+      "loss": 0.6184,
+      "step": 2032
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.4839636084749083,
+      "learning_rate": 2.4047110781453498e-06,
+      "loss": 0.7827,
+      "step": 2033
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5762702438166024,
+      "learning_rate": 2.4000572058342637e-06,
+      "loss": 0.9777,
+      "step": 2034
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.5062918817103069,
+      "learning_rate": 2.3954064186497257e-06,
+      "loss": 0.7644,
+      "step": 2035
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.4587081349347354,
+      "learning_rate": 2.390758722110418e-06,
+      "loss": 0.8175,
+      "step": 2036
+    },
+    {
+      "epoch": 0.68,
+      "grad_norm": 0.4350430240825197,
+      "learning_rate": 2.386114121731361e-06,
+      "loss": 0.9979,
+      "step": 2037
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6032270871541832,
+      "learning_rate": 2.3814726230239006e-06,
+      "loss": 0.7663,
+      "step": 2038
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.4122013765164129,
+      "learning_rate": 2.3768342314956973e-06,
+      "loss": 0.6414,
+      "step": 2039
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5228905398915734,
+      "learning_rate": 2.3721989526507296e-06,
+      "loss": 0.8395,
+      "step": 2040
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.3110414878179147,
+      "learning_rate": 2.3675667919892827e-06,
+      "loss": 0.9272,
+      "step": 2041
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8329540391629542,
+      "learning_rate": 2.362937755007935e-06,
+      "loss": 0.6974,
+      "step": 2042
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5655281809068868,
+      "learning_rate": 2.358311847199567e-06,
+      "loss": 0.9304,
+      "step": 2043
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.688322968832214,
+      "learning_rate": 2.353689074053342e-06,
+      "loss": 0.6458,
+      "step": 2044
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5396603040749945,
+      "learning_rate": 2.3490694410547015e-06,
+      "loss": 0.6429,
+      "step": 2045
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5316893726470724,
+      "learning_rate": 2.3444529536853645e-06,
+      "loss": 0.8884,
+      "step": 2046
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5362044398661988,
+      "learning_rate": 2.339839617423318e-06,
+      "loss": 0.8445,
+      "step": 2047
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.4051843041241085,
+      "learning_rate": 2.3352294377428035e-06,
+      "loss": 0.6371,
+      "step": 2048
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6013322024554837,
+      "learning_rate": 2.3306224201143242e-06,
+      "loss": 0.7991,
+      "step": 2049
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5319376616933376,
+      "learning_rate": 2.3260185700046295e-06,
+      "loss": 0.9817,
+      "step": 2050
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.8676832964073671,
+      "learning_rate": 2.321417892876706e-06,
+      "loss": 0.8835,
+      "step": 2051
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.37437392195285984,
+      "learning_rate": 2.31682039418978e-06,
+      "loss": 0.9253,
+      "step": 2052
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5207617777137178,
+      "learning_rate": 2.312226079399302e-06,
+      "loss": 0.8756,
+      "step": 2053
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5453177059521175,
+      "learning_rate": 2.307634953956948e-06,
+      "loss": 0.7761,
+      "step": 2054
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.4805427814492055,
+      "learning_rate": 2.303047023310611e-06,
+      "loss": 0.8691,
+      "step": 2055
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.48348771424775094,
+      "learning_rate": 2.298462292904386e-06,
+      "loss": 0.5151,
+      "step": 2056
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5638494879677498,
+      "learning_rate": 2.2938807681785764e-06,
+      "loss": 0.8048,
+      "step": 2057
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6269820842684775,
+      "learning_rate": 2.2893024545696822e-06,
+      "loss": 0.8949,
+      "step": 2058
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5734491519104516,
+      "learning_rate": 2.2847273575103873e-06,
+      "loss": 0.801,
+      "step": 2059
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5253025379685686,
+      "learning_rate": 2.2801554824295646e-06,
+      "loss": 1.0996,
+      "step": 2060
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.44421146916461846,
+      "learning_rate": 2.2755868347522624e-06,
+      "loss": 0.8945,
+      "step": 2061
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5059710506836784,
+      "learning_rate": 2.271021419899696e-06,
+      "loss": 0.9278,
+      "step": 2062
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.6338038991027325,
+      "learning_rate": 2.2664592432892485e-06,
+      "loss": 0.7649,
+      "step": 2063
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.34699693365040607,
+      "learning_rate": 2.2619003103344607e-06,
+      "loss": 0.8778,
+      "step": 2064
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.4907655507929394,
+      "learning_rate": 2.2573446264450195e-06,
+      "loss": 0.5738,
+      "step": 2065
+    },
+    {
+      "epoch": 0.69,
+      "grad_norm": 0.5522420180918608,
+      "learning_rate": 2.2527921970267614e-06,
+      "loss": 0.8464,
+      "step": 2066
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5086557566613343,
+      "learning_rate": 2.2482430274816613e-06,
+      "loss": 0.7659,
+      "step": 2067
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.4928610510822706,
+      "learning_rate": 2.2436971232078204e-06,
+      "loss": 0.7926,
+      "step": 2068
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6082037825588199,
+      "learning_rate": 2.23915448959947e-06,
+      "loss": 0.977,
+      "step": 2069
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5396711416258541,
+      "learning_rate": 2.234615132046962e-06,
+      "loss": 0.7773,
+      "step": 2070
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.7992826350363434,
+      "learning_rate": 2.2300790559367553e-06,
+      "loss": 0.8347,
+      "step": 2071
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5360104419925196,
+      "learning_rate": 2.22554626665142e-06,
+      "loss": 0.854,
+      "step": 2072
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.4745461578117751,
+      "learning_rate": 2.2210167695696205e-06,
+      "loss": 0.8214,
+      "step": 2073
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5184507520098796,
+      "learning_rate": 2.21649057006612e-06,
+      "loss": 0.9175,
+      "step": 2074
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.46093582100067326,
+      "learning_rate": 2.211967673511769e-06,
+      "loss": 0.5091,
+      "step": 2075
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.33776039004153724,
+      "learning_rate": 2.2074480852734926e-06,
+      "loss": 0.913,
+      "step": 2076
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.4725493755509347,
+      "learning_rate": 2.2029318107142954e-06,
+      "loss": 0.7977,
+      "step": 2077
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6834529395265273,
+      "learning_rate": 2.1984188551932513e-06,
+      "loss": 0.7613,
+      "step": 2078
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.4655309439988653,
+      "learning_rate": 2.1939092240654886e-06,
+      "loss": 0.933,
+      "step": 2079
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6012844820866787,
+      "learning_rate": 2.1894029226821973e-06,
+      "loss": 0.7425,
+      "step": 2080
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5103439057461865,
+      "learning_rate": 2.1848999563906163e-06,
+      "loss": 0.8452,
+      "step": 2081
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5434119071921376,
+      "learning_rate": 2.1804003305340217e-06,
+      "loss": 0.9258,
+      "step": 2082
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.46027454477859053,
+      "learning_rate": 2.17590405045173e-06,
+      "loss": 0.5313,
+      "step": 2083
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.43780970802342184,
+      "learning_rate": 2.17141112147909e-06,
+      "loss": 0.9541,
+      "step": 2084
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6889345974827538,
+      "learning_rate": 2.166921548947466e-06,
+      "loss": 0.832,
+      "step": 2085
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6044094306294017,
+      "learning_rate": 2.1624353381842457e-06,
+      "loss": 1.0012,
+      "step": 2086
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.561797801597106,
+      "learning_rate": 2.1579524945128287e-06,
+      "loss": 0.7758,
+      "step": 2087
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.63693617218699,
+      "learning_rate": 2.153473023252613e-06,
+      "loss": 0.7486,
+      "step": 2088
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5622559429710952,
+      "learning_rate": 2.1489969297190017e-06,
+      "loss": 0.9268,
+      "step": 2089
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.3189920634830371,
+      "learning_rate": 2.144524219223383e-06,
+      "loss": 0.8729,
+      "step": 2090
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.4488052981696085,
+      "learning_rate": 2.1400548970731377e-06,
+      "loss": 0.709,
+      "step": 2091
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5075492253119379,
+      "learning_rate": 2.1355889685716225e-06,
+      "loss": 0.8165,
+      "step": 2092
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5663208317032055,
+      "learning_rate": 2.1311264390181647e-06,
+      "loss": 0.7001,
+      "step": 2093
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.5021592881928888,
+      "learning_rate": 2.126667313708062e-06,
+      "loss": 0.7374,
+      "step": 2094
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 0.6056826937105005,
+      "learning_rate": 2.1222115979325746e-06,
+      "loss": 0.6963,
+      "step": 2095
+    },
+    {
+      "epoch": 0.7,
+      "grad_norm": 1.4036396364113077,
+      "learning_rate": 2.1177592969789092e-06,
+      "loss": 1.025,
+      "step": 2096
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.8390480616513468,
+      "learning_rate": 2.113310416130227e-06,
+      "loss": 0.7848,
+      "step": 2097
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5493016225641482,
+      "learning_rate": 2.1088649606656313e-06,
+      "loss": 0.9101,
+      "step": 2098
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6005212643517278,
+      "learning_rate": 2.1044229358601543e-06,
+      "loss": 0.8996,
+      "step": 2099
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6339959094677109,
+      "learning_rate": 2.099984346984763e-06,
+      "loss": 0.8872,
+      "step": 2100
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.45760642023941456,
+      "learning_rate": 2.095549199306348e-06,
+      "loss": 0.5226,
+      "step": 2101
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.37785587103224666,
+      "learning_rate": 2.0911174980877106e-06,
+      "loss": 0.8631,
+      "step": 2102
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5461018426172461,
+      "learning_rate": 2.0866892485875684e-06,
+      "loss": 1.0328,
+      "step": 2103
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5838534493502139,
+      "learning_rate": 2.082264456060542e-06,
+      "loss": 0.6921,
+      "step": 2104
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.766722130843901,
+      "learning_rate": 2.0778431257571453e-06,
+      "loss": 0.9533,
+      "step": 2105
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5292394686172502,
+      "learning_rate": 2.0734252629237892e-06,
+      "loss": 0.7194,
+      "step": 2106
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5376736093208632,
+      "learning_rate": 2.06901087280277e-06,
+      "loss": 0.7954,
+      "step": 2107
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5729425816923203,
+      "learning_rate": 2.064599960632259e-06,
+      "loss": 0.9561,
+      "step": 2108
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.4062374259660298,
+      "learning_rate": 2.0601925316463047e-06,
+      "loss": 0.6746,
+      "step": 2109
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5166737664349593,
+      "learning_rate": 2.0557885910748177e-06,
+      "loss": 0.8844,
+      "step": 2110
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6234384056432394,
+      "learning_rate": 2.0513881441435748e-06,
+      "loss": 0.8501,
+      "step": 2111
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.523817576615301,
+      "learning_rate": 2.0469911960742055e-06,
+      "loss": 0.8999,
+      "step": 2112
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5205247110697834,
+      "learning_rate": 2.0425977520841837e-06,
+      "loss": 0.7519,
+      "step": 2113
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.3097106154301985,
+      "learning_rate": 2.03820781738683e-06,
+      "loss": 0.9359,
+      "step": 2114
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.515713790980987,
+      "learning_rate": 2.0338213971913e-06,
+      "loss": 0.883,
+      "step": 2115
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.668410726146474,
+      "learning_rate": 2.0294384967025748e-06,
+      "loss": 0.6758,
+      "step": 2116
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5244651444565004,
+      "learning_rate": 2.0250591211214647e-06,
+      "loss": 0.8351,
+      "step": 2117
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6231802011663488,
+      "learning_rate": 2.0206832756445954e-06,
+      "loss": 0.7916,
+      "step": 2118
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.41754424587807265,
+      "learning_rate": 2.0163109654644e-06,
+      "loss": 0.7539,
+      "step": 2119
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.6090913530045223,
+      "learning_rate": 2.011942195769122e-06,
+      "loss": 0.7422,
+      "step": 2120
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5463958360745396,
+      "learning_rate": 2.007576971742803e-06,
+      "loss": 0.752,
+      "step": 2121
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5862532511698234,
+      "learning_rate": 2.0032152985652708e-06,
+      "loss": 0.8174,
+      "step": 2122
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 1.5153905083946824,
+      "learning_rate": 1.9988571814121476e-06,
+      "loss": 0.8047,
+      "step": 2123
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5065050928371415,
+      "learning_rate": 1.994502625454834e-06,
+      "loss": 0.8994,
+      "step": 2124
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.5830211977385404,
+      "learning_rate": 1.9901516358605012e-06,
+      "loss": 0.7723,
+      "step": 2125
+    },
+    {
+      "epoch": 0.71,
+      "grad_norm": 0.33612596045093657,
+      "learning_rate": 1.9858042177920915e-06,
+      "loss": 0.941,
+      "step": 2126
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.40854466773009407,
+      "learning_rate": 1.9814603764083112e-06,
+      "loss": 0.5729,
+      "step": 2127
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5714032612147674,
+      "learning_rate": 1.977120116863617e-06,
+      "loss": 0.805,
+      "step": 2128
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5631461303461922,
+      "learning_rate": 1.972783444308221e-06,
+      "loss": 0.8409,
+      "step": 2129
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5161123459603311,
+      "learning_rate": 1.968450363888073e-06,
+      "loss": 0.8281,
+      "step": 2130
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.453002251882312,
+      "learning_rate": 1.964120880744866e-06,
+      "loss": 0.8772,
+      "step": 2131
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5234676509878602,
+      "learning_rate": 1.959795000016023e-06,
+      "loss": 0.9189,
+      "step": 2132
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.48958255210909074,
+      "learning_rate": 1.9554727268346876e-06,
+      "loss": 0.6851,
+      "step": 2133
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.68109891426371,
+      "learning_rate": 1.9511540663297284e-06,
+      "loss": 0.8297,
+      "step": 2134
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6003059597215256,
+      "learning_rate": 1.9468390236257277e-06,
+      "loss": 0.8338,
+      "step": 2135
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.448119260451637,
+      "learning_rate": 1.9425276038429684e-06,
+      "loss": 0.6855,
+      "step": 2136
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5812057617149113,
+      "learning_rate": 1.9382198120974395e-06,
+      "loss": 0.8076,
+      "step": 2137
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5961356923003247,
+      "learning_rate": 1.933915653500826e-06,
+      "loss": 0.767,
+      "step": 2138
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.370542421711581,
+      "learning_rate": 1.929615133160496e-06,
+      "loss": 0.9414,
+      "step": 2139
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5411204521585948,
+      "learning_rate": 1.925318256179505e-06,
+      "loss": 0.6647,
+      "step": 2140
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.984269076972637,
+      "learning_rate": 1.921025027656587e-06,
+      "loss": 0.9471,
+      "step": 2141
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.44996895942802795,
+      "learning_rate": 1.916735452686139e-06,
+      "loss": 0.8794,
+      "step": 2142
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6032403037155396,
+      "learning_rate": 1.9124495363582308e-06,
+      "loss": 0.9431,
+      "step": 2143
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5499655229176346,
+      "learning_rate": 1.9081672837585873e-06,
+      "loss": 0.7528,
+      "step": 2144
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5892974403198784,
+      "learning_rate": 1.903888699968584e-06,
+      "loss": 0.7633,
+      "step": 2145
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.37580271914361446,
+      "learning_rate": 1.8996137900652468e-06,
+      "loss": 0.5553,
+      "step": 2146
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.7231757805707497,
+      "learning_rate": 1.895342559121241e-06,
+      "loss": 0.9746,
+      "step": 2147
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6442902875520405,
+      "learning_rate": 1.8910750122048638e-06,
+      "loss": 0.8912,
+      "step": 2148
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6699156188354629,
+      "learning_rate": 1.8868111543800456e-06,
+      "loss": 0.891,
+      "step": 2149
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5573195047470441,
+      "learning_rate": 1.8825509907063328e-06,
+      "loss": 0.7979,
+      "step": 2150
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.6356449103910262,
+      "learning_rate": 1.8782945262388941e-06,
+      "loss": 0.9491,
+      "step": 2151
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5267728929120027,
+      "learning_rate": 1.8740417660285087e-06,
+      "loss": 0.7885,
+      "step": 2152
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5312157226016745,
+      "learning_rate": 1.8697927151215545e-06,
+      "loss": 0.8567,
+      "step": 2153
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.3584901380294739,
+      "learning_rate": 1.8655473785600125e-06,
+      "loss": 0.667,
+      "step": 2154
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.637730429301141,
+      "learning_rate": 1.8613057613814584e-06,
+      "loss": 0.8619,
+      "step": 2155
+    },
+    {
+      "epoch": 0.72,
+      "grad_norm": 0.5919617720075612,
+      "learning_rate": 1.8570678686190463e-06,
+      "loss": 0.8441,
+      "step": 2156
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 2.1918900498377054,
+      "learning_rate": 1.8528337053015188e-06,
+      "loss": 0.7243,
+      "step": 2157
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6622908040440342,
+      "learning_rate": 1.8486032764531918e-06,
+      "loss": 1.0017,
+      "step": 2158
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5417338338330788,
+      "learning_rate": 1.844376587093944e-06,
+      "loss": 0.7216,
+      "step": 2159
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8118880282344695,
+      "learning_rate": 1.8401536422392237e-06,
+      "loss": 0.8728,
+      "step": 2160
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5494547312934122,
+      "learning_rate": 1.8359344469000334e-06,
+      "loss": 0.7133,
+      "step": 2161
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5630156683628308,
+      "learning_rate": 1.8317190060829242e-06,
+      "loss": 0.9624,
+      "step": 2162
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.4050638402648249,
+      "learning_rate": 1.827507324789995e-06,
+      "loss": 0.5281,
+      "step": 2163
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.3298272911901062,
+      "learning_rate": 1.8232994080188848e-06,
+      "loss": 0.8613,
+      "step": 2164
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.48115274414263404,
+      "learning_rate": 1.8190952607627594e-06,
+      "loss": 0.9046,
+      "step": 2165
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7020691829754808,
+      "learning_rate": 1.8148948880103174e-06,
+      "loss": 0.9308,
+      "step": 2166
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5515256754859399,
+      "learning_rate": 1.8106982947457785e-06,
+      "loss": 0.8567,
+      "step": 2167
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5809375567676622,
+      "learning_rate": 1.8065054859488729e-06,
+      "loss": 0.8474,
+      "step": 2168
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5057601342748979,
+      "learning_rate": 1.8023164665948455e-06,
+      "loss": 0.6995,
+      "step": 2169
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.547126275768686,
+      "learning_rate": 1.7981312416544394e-06,
+      "loss": 1.0296,
+      "step": 2170
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.752046447473547,
+      "learning_rate": 1.7939498160938995e-06,
+      "loss": 0.7365,
+      "step": 2171
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.3361029256035591,
+      "learning_rate": 1.7897721948749625e-06,
+      "loss": 0.5837,
+      "step": 2172
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.4934256798376289,
+      "learning_rate": 1.7855983829548456e-06,
+      "loss": 0.747,
+      "step": 2173
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.4922594209475173,
+      "learning_rate": 1.7814283852862507e-06,
+      "loss": 0.9865,
+      "step": 2174
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.3303626571465021,
+      "learning_rate": 1.7772622068173544e-06,
+      "loss": 0.889,
+      "step": 2175
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6010693447884535,
+      "learning_rate": 1.773099852491796e-06,
+      "loss": 0.7565,
+      "step": 2176
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7262553534664486,
+      "learning_rate": 1.768941327248681e-06,
+      "loss": 0.8423,
+      "step": 2177
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.8171359061736324,
+      "learning_rate": 1.764786636022573e-06,
+      "loss": 0.885,
+      "step": 2178
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5245592297681129,
+      "learning_rate": 1.76063578374348e-06,
+      "loss": 0.9071,
+      "step": 2179
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.36339245528642294,
+      "learning_rate": 1.7564887753368604e-06,
+      "loss": 0.5888,
+      "step": 2180
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.4737629544802424,
+      "learning_rate": 1.7523456157236102e-06,
+      "loss": 0.9479,
+      "step": 2181
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5950252361156674,
+      "learning_rate": 1.7482063098200547e-06,
+      "loss": 0.764,
+      "step": 2182
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.7692332146449409,
+      "learning_rate": 1.7440708625379503e-06,
+      "loss": 0.7701,
+      "step": 2183
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.5269868701369955,
+      "learning_rate": 1.739939278784476e-06,
+      "loss": 0.8884,
+      "step": 2184
+    },
+    {
+      "epoch": 0.73,
+      "grad_norm": 0.6526267081821381,
+      "learning_rate": 1.7358115634622192e-06,
+      "loss": 0.8322,
+      "step": 2185
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5964593561478433,
+      "learning_rate": 1.7316877214691863e-06,
+      "loss": 0.9225,
+      "step": 2186
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5554671371856235,
+      "learning_rate": 1.727567757698778e-06,
+      "loss": 0.7822,
+      "step": 2187
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.327800799213394,
+      "learning_rate": 1.7234516770398008e-06,
+      "loss": 0.9516,
+      "step": 2188
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3242599607292603,
+      "learning_rate": 1.7193394843764516e-06,
+      "loss": 0.6529,
+      "step": 2189
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.8631120562762452,
+      "learning_rate": 1.7152311845883096e-06,
+      "loss": 0.7012,
+      "step": 2190
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5647334815263347,
+      "learning_rate": 1.711126782550339e-06,
+      "loss": 0.9553,
+      "step": 2191
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5679201182897488,
+      "learning_rate": 1.7070262831328799e-06,
+      "loss": 0.8699,
+      "step": 2192
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.613572286559702,
+      "learning_rate": 1.7029296912016346e-06,
+      "loss": 0.8581,
+      "step": 2193
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.461612440323199,
+      "learning_rate": 1.6988370116176766e-06,
+      "loss": 0.7185,
+      "step": 2194
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5067533858985829,
+      "learning_rate": 1.6947482492374346e-06,
+      "loss": 0.7969,
+      "step": 2195
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5591336172504564,
+      "learning_rate": 1.6906634089126844e-06,
+      "loss": 0.9176,
+      "step": 2196
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5744909018427462,
+      "learning_rate": 1.686582495490554e-06,
+      "loss": 0.8115,
+      "step": 2197
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.401169932086618,
+      "learning_rate": 1.6825055138135105e-06,
+      "loss": 0.6627,
+      "step": 2198
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6178504928536617,
+      "learning_rate": 1.6784324687193516e-06,
+      "loss": 0.7376,
+      "step": 2199
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3345513403402615,
+      "learning_rate": 1.6743633650412077e-06,
+      "loss": 0.9742,
+      "step": 2200
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5461458357886938,
+      "learning_rate": 1.6702982076075326e-06,
+      "loss": 0.8082,
+      "step": 2201
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.47970097984608456,
+      "learning_rate": 1.6662370012420931e-06,
+      "loss": 0.775,
+      "step": 2202
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5700998515875995,
+      "learning_rate": 1.662179750763972e-06,
+      "loss": 0.8875,
+      "step": 2203
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.46591574800964275,
+      "learning_rate": 1.658126460987558e-06,
+      "loss": 0.7777,
+      "step": 2204
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.6740003807744499,
+      "learning_rate": 1.654077136722535e-06,
+      "loss": 0.9228,
+      "step": 2205
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.541409575921345,
+      "learning_rate": 1.6500317827738887e-06,
+      "loss": 0.7481,
+      "step": 2206
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5620705174864981,
+      "learning_rate": 1.6459904039418856e-06,
+      "loss": 0.8604,
+      "step": 2207
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.4264403375439777,
+      "learning_rate": 1.641953005022081e-06,
+      "loss": 0.5779,
+      "step": 2208
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5556569412862551,
+      "learning_rate": 1.637919590805308e-06,
+      "loss": 0.6829,
+      "step": 2209
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5866613716712976,
+      "learning_rate": 1.6338901660776662e-06,
+      "loss": 0.9339,
+      "step": 2210
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.45703757189846866,
+      "learning_rate": 1.6298647356205255e-06,
+      "loss": 0.7887,
+      "step": 2211
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.46299727773286425,
+      "learning_rate": 1.6258433042105164e-06,
+      "loss": 0.8746,
+      "step": 2212
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.3160611854271421,
+      "learning_rate": 1.62182587661952e-06,
+      "loss": 0.9969,
+      "step": 2213
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.5623532720311555,
+      "learning_rate": 1.6178124576146708e-06,
+      "loss": 0.6939,
+      "step": 2214
+    },
+    {
+      "epoch": 0.74,
+      "grad_norm": 0.46068251552149897,
+      "learning_rate": 1.6138030519583458e-06,
+      "loss": 0.635,
+      "step": 2215
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5347111178120757,
+      "learning_rate": 1.6097976644081564e-06,
+      "loss": 0.7668,
+      "step": 2216
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5808272752948688,
+      "learning_rate": 1.605796299716949e-06,
+      "loss": 0.8886,
+      "step": 2217
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5165286703213693,
+      "learning_rate": 1.601798962632799e-06,
+      "loss": 0.6626,
+      "step": 2218
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5594190110668192,
+      "learning_rate": 1.5978056578989953e-06,
+      "loss": 0.7995,
+      "step": 2219
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5317143851358462,
+      "learning_rate": 1.5938163902540483e-06,
+      "loss": 0.9867,
+      "step": 2220
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.46688083062591884,
+      "learning_rate": 1.5898311644316772e-06,
+      "loss": 0.7962,
+      "step": 2221
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.49144950723504544,
+      "learning_rate": 1.5858499851608006e-06,
+      "loss": 0.8047,
+      "step": 2222
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.6648198225528845,
+      "learning_rate": 1.5818728571655394e-06,
+      "loss": 0.9946,
+      "step": 2223
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5172421742627559,
+      "learning_rate": 1.5778997851652095e-06,
+      "loss": 0.9136,
+      "step": 2224
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5730660366058762,
+      "learning_rate": 1.573930773874306e-06,
+      "loss": 0.5478,
+      "step": 2225
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.45626577511007926,
+      "learning_rate": 1.5699658280025143e-06,
+      "loss": 0.777,
+      "step": 2226
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.36167111260268875,
+      "learning_rate": 1.5660049522546883e-06,
+      "loss": 0.9556,
+      "step": 2227
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.532970860300864,
+      "learning_rate": 1.5620481513308571e-06,
+      "loss": 0.7455,
+      "step": 2228
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5100833393559174,
+      "learning_rate": 1.5580954299262146e-06,
+      "loss": 0.8747,
+      "step": 2229
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5662369879645915,
+      "learning_rate": 1.5541467927311093e-06,
+      "loss": 0.676,
+      "step": 2230
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5539928991557915,
+      "learning_rate": 1.5502022444310482e-06,
+      "loss": 0.6866,
+      "step": 2231
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5392845623842587,
+      "learning_rate": 1.5462617897066863e-06,
+      "loss": 0.8856,
+      "step": 2232
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.4845335270048156,
+      "learning_rate": 1.542325433233815e-06,
+      "loss": 0.8304,
+      "step": 2233
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.3731017815114956,
+      "learning_rate": 1.5383931796833702e-06,
+      "loss": 0.6415,
+      "step": 2234
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7232062983780918,
+      "learning_rate": 1.5344650337214163e-06,
+      "loss": 0.786,
+      "step": 2235
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.509432514284287,
+      "learning_rate": 1.5305410000091408e-06,
+      "loss": 0.9874,
+      "step": 2236
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.42999622716129254,
+      "learning_rate": 1.5266210832028556e-06,
+      "loss": 0.9006,
+      "step": 2237
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.38109880650276623,
+      "learning_rate": 1.5227052879539872e-06,
+      "loss": 0.9074,
+      "step": 2238
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5296994767069483,
+      "learning_rate": 1.5187936189090668e-06,
+      "loss": 0.8563,
+      "step": 2239
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.7370485042429333,
+      "learning_rate": 1.5148860807097348e-06,
+      "loss": 0.7261,
+      "step": 2240
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.45424076894485865,
+      "learning_rate": 1.5109826779927296e-06,
+      "loss": 0.8866,
+      "step": 2241
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.4505594629293681,
+      "learning_rate": 1.5070834153898766e-06,
+      "loss": 0.5126,
+      "step": 2242
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5218942656598463,
+      "learning_rate": 1.5031882975280942e-06,
+      "loss": 0.8901,
+      "step": 2243
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.5808362978068721,
+      "learning_rate": 1.4992973290293827e-06,
+      "loss": 0.7654,
+      "step": 2244
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 0.9720099591438109,
+      "learning_rate": 1.4954105145108133e-06,
+      "loss": 0.7699,
+      "step": 2245
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5192658280715803,
+      "learning_rate": 1.491527858584535e-06,
+      "loss": 1.0855,
+      "step": 2246
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.4185761354969329,
+      "learning_rate": 1.4876493658577557e-06,
+      "loss": 0.8995,
+      "step": 2247
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 1.009206964886309,
+      "learning_rate": 1.4837750409327478e-06,
+      "loss": 0.8975,
+      "step": 2248
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5624922148935104,
+      "learning_rate": 1.479904888406838e-06,
+      "loss": 0.6738,
+      "step": 2249
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.32592566358709985,
+      "learning_rate": 1.4760389128723968e-06,
+      "loss": 0.8824,
+      "step": 2250
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.43089754874747305,
+      "learning_rate": 1.4721771189168444e-06,
+      "loss": 0.5909,
+      "step": 2251
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7390115998179967,
+      "learning_rate": 1.4683195111226378e-06,
+      "loss": 0.7792,
+      "step": 2252
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.48490351842094764,
+      "learning_rate": 1.4644660940672628e-06,
+      "loss": 0.7983,
+      "step": 2253
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5030906199823055,
+      "learning_rate": 1.460616872323236e-06,
+      "loss": 0.7673,
+      "step": 2254
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.645693485228522,
+      "learning_rate": 1.456771850458097e-06,
+      "loss": 0.9352,
+      "step": 2255
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5632942024859234,
+      "learning_rate": 1.4529310330343964e-06,
+      "loss": 0.7764,
+      "step": 2256
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.49099029000394073,
+      "learning_rate": 1.4490944246097e-06,
+      "loss": 0.8271,
+      "step": 2257
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5369421609340504,
+      "learning_rate": 1.4452620297365804e-06,
+      "loss": 0.9091,
+      "step": 2258
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.46178790881650433,
+      "learning_rate": 1.4414338529626054e-06,
+      "loss": 0.9115,
+      "step": 2259
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.3947183077544402,
+      "learning_rate": 1.4376098988303406e-06,
+      "loss": 0.5888,
+      "step": 2260
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5989132244665587,
+      "learning_rate": 1.4337901718773427e-06,
+      "loss": 0.7667,
+      "step": 2261
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.3602059444350063,
+      "learning_rate": 1.4299746766361477e-06,
+      "loss": 0.9103,
+      "step": 2262
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5483049759010085,
+      "learning_rate": 1.4261634176342747e-06,
+      "loss": 0.851,
+      "step": 2263
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6302113658525994,
+      "learning_rate": 1.4223563993942124e-06,
+      "loss": 0.7991,
+      "step": 2264
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.7726555300988469,
+      "learning_rate": 1.41855362643342e-06,
+      "loss": 0.9133,
+      "step": 2265
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.686933761930828,
+      "learning_rate": 1.4147551032643192e-06,
+      "loss": 0.6927,
+      "step": 2266
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.4773068592670479,
+      "learning_rate": 1.4109608343942855e-06,
+      "loss": 0.9372,
+      "step": 2267
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.4163288450303881,
+      "learning_rate": 1.40717082432565e-06,
+      "loss": 0.53,
+      "step": 2268
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5970659113039093,
+      "learning_rate": 1.40338507755569e-06,
+      "loss": 0.7381,
+      "step": 2269
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.6894798339878889,
+      "learning_rate": 1.3996035985766205e-06,
+      "loss": 0.9829,
+      "step": 2270
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5080195793063105,
+      "learning_rate": 1.3958263918755955e-06,
+      "loss": 0.9025,
+      "step": 2271
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5657111208483911,
+      "learning_rate": 1.3920534619346998e-06,
+      "loss": 0.8906,
+      "step": 2272
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5786626560217686,
+      "learning_rate": 1.3882848132309395e-06,
+      "loss": 0.7729,
+      "step": 2273
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.8018533462188561,
+      "learning_rate": 1.3845204502362442e-06,
+      "loss": 0.915,
+      "step": 2274
+    },
+    {
+      "epoch": 0.76,
+      "grad_norm": 0.5922688164603227,
+      "learning_rate": 1.3807603774174582e-06,
+      "loss": 0.7891,
+      "step": 2275
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.31806948516291356,
+      "learning_rate": 1.3770045992363313e-06,
+      "loss": 0.8654,
+      "step": 2276
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5041628614398905,
+      "learning_rate": 1.3732531201495208e-06,
+      "loss": 0.8253,
+      "step": 2277
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6454694440527555,
+      "learning_rate": 1.369505944608583e-06,
+      "loss": 0.8838,
+      "step": 2278
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.43994084971228115,
+      "learning_rate": 1.3657630770599633e-06,
+      "loss": 0.6836,
+      "step": 2279
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5800898044280655,
+      "learning_rate": 1.3620245219449984e-06,
+      "loss": 0.7709,
+      "step": 2280
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5905885938755522,
+      "learning_rate": 1.35829028369991e-06,
+      "loss": 0.7462,
+      "step": 2281
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.4402788880667987,
+      "learning_rate": 1.3545603667557911e-06,
+      "loss": 0.9688,
+      "step": 2282
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6924680988738039,
+      "learning_rate": 1.350834775538613e-06,
+      "loss": 0.8019,
+      "step": 2283
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.49990935817813825,
+      "learning_rate": 1.34711351446921e-06,
+      "loss": 0.9388,
+      "step": 2284
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6910797543385605,
+      "learning_rate": 1.3433965879632799e-06,
+      "loss": 0.853,
+      "step": 2285
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5058906611727091,
+      "learning_rate": 1.3396840004313789e-06,
+      "loss": 0.8503,
+      "step": 2286
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5125370831610015,
+      "learning_rate": 1.3359757562789099e-06,
+      "loss": 0.5058,
+      "step": 2287
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.357537042008932,
+      "learning_rate": 1.3322718599061252e-06,
+      "loss": 0.8901,
+      "step": 2288
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5402349805811282,
+      "learning_rate": 1.3285723157081193e-06,
+      "loss": 1.0019,
+      "step": 2289
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.759916018365248,
+      "learning_rate": 1.3248771280748174e-06,
+      "loss": 0.8345,
+      "step": 2290
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5223628926491665,
+      "learning_rate": 1.3211863013909797e-06,
+      "loss": 0.8827,
+      "step": 2291
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5173041383447046,
+      "learning_rate": 1.3174998400361911e-06,
+      "loss": 0.7263,
+      "step": 2292
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.535659518491001,
+      "learning_rate": 1.3138177483848524e-06,
+      "loss": 0.9046,
+      "step": 2293
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5180542030908388,
+      "learning_rate": 1.310140030806184e-06,
+      "loss": 0.8875,
+      "step": 2294
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.4701298157642592,
+      "learning_rate": 1.306466691664216e-06,
+      "loss": 0.5389,
+      "step": 2295
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.49232188896043927,
+      "learning_rate": 1.302797735317778e-06,
+      "loss": 0.9028,
+      "step": 2296
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5916362084342457,
+      "learning_rate": 1.2991331661205036e-06,
+      "loss": 0.8032,
+      "step": 2297
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6920390297184765,
+      "learning_rate": 1.2954729884208212e-06,
+      "loss": 0.9584,
+      "step": 2298
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.5560130499303949,
+      "learning_rate": 1.2918172065619422e-06,
+      "loss": 0.7677,
+      "step": 2299
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.3761048123085909,
+      "learning_rate": 1.288165824881868e-06,
+      "loss": 0.9289,
+      "step": 2300
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6015385617326587,
+      "learning_rate": 1.284518847713378e-06,
+      "loss": 0.8005,
+      "step": 2301
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 1.8531122848811061,
+      "learning_rate": 1.28087627938402e-06,
+      "loss": 0.7234,
+      "step": 2302
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.4871133007746938,
+      "learning_rate": 1.277238124216117e-06,
+      "loss": 0.8539,
+      "step": 2303
+    },
+    {
+      "epoch": 0.77,
+      "grad_norm": 0.6073843818647819,
+      "learning_rate": 1.2736043865267495e-06,
+      "loss": 0.7213,
+      "step": 2304
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.34672626178114785,
+      "learning_rate": 1.2699750706277604e-06,
+      "loss": 0.7168,
+      "step": 2305
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5780667426034963,
+      "learning_rate": 1.2663501808257444e-06,
+      "loss": 0.7481,
+      "step": 2306
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5211451494112566,
+      "learning_rate": 1.2627297214220413e-06,
+      "loss": 0.7564,
+      "step": 2307
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.4535270016491472,
+      "learning_rate": 1.2591136967127383e-06,
+      "loss": 0.8471,
+      "step": 2308
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.57726557600696,
+      "learning_rate": 1.2555021109886589e-06,
+      "loss": 0.7225,
+      "step": 2309
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6068823101548866,
+      "learning_rate": 1.251894968535356e-06,
+      "loss": 0.9818,
+      "step": 2310
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5698390331955182,
+      "learning_rate": 1.2482922736331144e-06,
+      "loss": 0.8604,
+      "step": 2311
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.3205432336316307,
+      "learning_rate": 1.24469403055694e-06,
+      "loss": 0.9113,
+      "step": 2312
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.41693211550085924,
+      "learning_rate": 1.2411002435765535e-06,
+      "loss": 0.5915,
+      "step": 2313
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6134812804724566,
+      "learning_rate": 1.2375109169563915e-06,
+      "loss": 0.8121,
+      "step": 2314
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.666656743048381,
+      "learning_rate": 1.2339260549555976e-06,
+      "loss": 0.8613,
+      "step": 2315
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5309553377701748,
+      "learning_rate": 1.2303456618280141e-06,
+      "loss": 0.7799,
+      "step": 2316
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.4488112173726499,
+      "learning_rate": 1.226769741822184e-06,
+      "loss": 0.9798,
+      "step": 2317
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5774447921119675,
+      "learning_rate": 1.2231982991813428e-06,
+      "loss": 0.8483,
+      "step": 2318
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.4849274912373264,
+      "learning_rate": 1.2196313381434094e-06,
+      "loss": 0.6809,
+      "step": 2319
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5535487356815538,
+      "learning_rate": 1.2160688629409873e-06,
+      "loss": 0.8021,
+      "step": 2320
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.43830972153818565,
+      "learning_rate": 1.2125108778013588e-06,
+      "loss": 0.6164,
+      "step": 2321
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 1.2334131361609937,
+      "learning_rate": 1.2089573869464738e-06,
+      "loss": 0.9925,
+      "step": 2322
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.48558949374200727,
+      "learning_rate": 1.2054083945929534e-06,
+      "loss": 0.7608,
+      "step": 2323
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.8007709403686002,
+      "learning_rate": 1.201863904952077e-06,
+      "loss": 0.7467,
+      "step": 2324
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.47828518584558977,
+      "learning_rate": 1.1983239222297837e-06,
+      "loss": 0.9539,
+      "step": 2325
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.6752524142714916,
+      "learning_rate": 1.1947884506266655e-06,
+      "loss": 0.6588,
+      "step": 2326
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.584474124998159,
+      "learning_rate": 1.1912574943379574e-06,
+      "loss": 0.9525,
+      "step": 2327
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.41689555052688493,
+      "learning_rate": 1.1877310575535394e-06,
+      "loss": 0.8154,
+      "step": 2328
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5041961057113901,
+      "learning_rate": 1.1842091444579307e-06,
+      "loss": 0.9849,
+      "step": 2329
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5232363528879082,
+      "learning_rate": 1.1806917592302763e-06,
+      "loss": 0.7343,
+      "step": 2330
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5372336816949277,
+      "learning_rate": 1.1771789060443535e-06,
+      "loss": 0.7369,
+      "step": 2331
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.4796112573228635,
+      "learning_rate": 1.1736705890685623e-06,
+      "loss": 0.5786,
+      "step": 2332
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5625808429013301,
+      "learning_rate": 1.1701668124659148e-06,
+      "loss": 0.8836,
+      "step": 2333
+    },
+    {
+      "epoch": 0.78,
+      "grad_norm": 0.5883095666322794,
+      "learning_rate": 1.166667580394041e-06,
+      "loss": 1.0051,
+      "step": 2334
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5147826828894809,
+      "learning_rate": 1.163172897005176e-06,
+      "loss": 0.8146,
+      "step": 2335
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6875098318728239,
+      "learning_rate": 1.1596827664461558e-06,
+      "loss": 0.771,
+      "step": 2336
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.33198090796552066,
+      "learning_rate": 1.1561971928584158e-06,
+      "loss": 0.9274,
+      "step": 2337
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5372171452891453,
+      "learning_rate": 1.1527161803779868e-06,
+      "loss": 0.7452,
+      "step": 2338
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5652202289310566,
+      "learning_rate": 1.14923973313548e-06,
+      "loss": 0.938,
+      "step": 2339
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.3427408828096367,
+      "learning_rate": 1.1457678552560969e-06,
+      "loss": 0.6496,
+      "step": 2340
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7834291767797433,
+      "learning_rate": 1.142300550859612e-06,
+      "loss": 0.8666,
+      "step": 2341
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.7083472676572106,
+      "learning_rate": 1.1388378240603742e-06,
+      "loss": 0.8548,
+      "step": 2342
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5984771748555517,
+      "learning_rate": 1.1353796789673037e-06,
+      "loss": 0.8125,
+      "step": 2343
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5494706852370462,
+      "learning_rate": 1.1319261196838782e-06,
+      "loss": 0.9688,
+      "step": 2344
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6027322732261179,
+      "learning_rate": 1.1284771503081377e-06,
+      "loss": 0.7632,
+      "step": 2345
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6551603539195917,
+      "learning_rate": 1.1250327749326772e-06,
+      "loss": 0.9461,
+      "step": 2346
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5613082466191163,
+      "learning_rate": 1.121592997644635e-06,
+      "loss": 0.7165,
+      "step": 2347
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5521957426675163,
+      "learning_rate": 1.1181578225256973e-06,
+      "loss": 0.9446,
+      "step": 2348
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.4524897692199406,
+      "learning_rate": 1.1147272536520904e-06,
+      "loss": 0.6002,
+      "step": 2349
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.363310042968246,
+      "learning_rate": 1.1113012950945695e-06,
+      "loss": 0.8683,
+      "step": 2350
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5380807872110437,
+      "learning_rate": 1.1078799509184246e-06,
+      "loss": 0.9183,
+      "step": 2351
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5193790170424335,
+      "learning_rate": 1.1044632251834687e-06,
+      "loss": 0.9127,
+      "step": 2352
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.508932727773165,
+      "learning_rate": 1.1010511219440307e-06,
+      "loss": 0.8007,
+      "step": 2353
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.5514480600221711,
+      "learning_rate": 1.0976436452489592e-06,
+      "loss": 0.8349,
+      "step": 2354
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.540316142744408,
+      "learning_rate": 1.0942407991416132e-06,
+      "loss": 0.6948,
+      "step": 2355
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6321441965120509,
+      "learning_rate": 1.0908425876598512e-06,
+      "loss": 1.014,
+      "step": 2356
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6005724852246883,
+      "learning_rate": 1.0874490148360383e-06,
+      "loss": 0.5771,
+      "step": 2357
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.40042546259771333,
+      "learning_rate": 1.0840600846970333e-06,
+      "loss": 0.6414,
+      "step": 2358
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.48031030231751565,
+      "learning_rate": 1.0806758012641833e-06,
+      "loss": 0.7809,
+      "step": 2359
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6264351563770774,
+      "learning_rate": 1.0772961685533267e-06,
+      "loss": 0.9279,
+      "step": 2360
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.3150731056030197,
+      "learning_rate": 1.0739211905747777e-06,
+      "loss": 0.915,
+      "step": 2361
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6553820109807793,
+      "learning_rate": 1.0705508713333313e-06,
+      "loss": 0.7042,
+      "step": 2362
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6445821209796092,
+      "learning_rate": 1.0671852148282535e-06,
+      "loss": 0.9827,
+      "step": 2363
+    },
+    {
+      "epoch": 0.79,
+      "grad_norm": 0.6392859159145945,
+      "learning_rate": 1.0638242250532754e-06,
+      "loss": 0.763,
+      "step": 2364
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5213197159809989,
+      "learning_rate": 1.0604679059965923e-06,
+      "loss": 0.8724,
+      "step": 2365
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.3804790343887967,
+      "learning_rate": 1.0571162616408586e-06,
+      "loss": 0.606,
+      "step": 2366
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.49524007038238244,
+      "learning_rate": 1.0537692959631773e-06,
+      "loss": 0.9451,
+      "step": 2367
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.7752488095609149,
+      "learning_rate": 1.0504270129351035e-06,
+      "loss": 0.7085,
+      "step": 2368
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 1.0231723878739634,
+      "learning_rate": 1.0470894165226359e-06,
+      "loss": 0.7708,
+      "step": 2369
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6624082496947233,
+      "learning_rate": 1.0437565106862075e-06,
+      "loss": 0.8744,
+      "step": 2370
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5666967788702411,
+      "learning_rate": 1.0404282993806897e-06,
+      "loss": 0.7563,
+      "step": 2371
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6073100662400983,
+      "learning_rate": 1.0371047865553847e-06,
+      "loss": 0.9576,
+      "step": 2372
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.33820149503070007,
+      "learning_rate": 1.0337859761540125e-06,
+      "loss": 0.8959,
+      "step": 2373
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5357267528404536,
+      "learning_rate": 1.03047187211472e-06,
+      "loss": 0.9961,
+      "step": 2374
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.35536679201488697,
+      "learning_rate": 1.0271624783700678e-06,
+      "loss": 0.7102,
+      "step": 2375
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6962772228885178,
+      "learning_rate": 1.0238577988470233e-06,
+      "loss": 0.6635,
+      "step": 2376
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6864723356594338,
+      "learning_rate": 1.0205578374669655e-06,
+      "loss": 0.9003,
+      "step": 2377
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5885564837104114,
+      "learning_rate": 1.0172625981456723e-06,
+      "loss": 0.8471,
+      "step": 2378
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5710597195535327,
+      "learning_rate": 1.0139720847933166e-06,
+      "loss": 0.9234,
+      "step": 2379
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.46225721905768774,
+      "learning_rate": 1.0106863013144668e-06,
+      "loss": 0.671,
+      "step": 2380
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.586304855545277,
+      "learning_rate": 1.0074052516080756e-06,
+      "loss": 0.7768,
+      "step": 2381
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5156550653160442,
+      "learning_rate": 1.0041289395674802e-06,
+      "loss": 0.8899,
+      "step": 2382
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.629925425229625,
+      "learning_rate": 1.000857369080398e-06,
+      "loss": 0.7959,
+      "step": 2383
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.4093247150043599,
+      "learning_rate": 9.975905440289152e-07,
+      "loss": 0.6639,
+      "step": 2384
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5636306360527155,
+      "learning_rate": 9.943284682894905e-07,
+      "loss": 0.7352,
+      "step": 2385
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.28888985477057166,
+      "learning_rate": 9.91071145732948e-07,
+      "loss": 0.97,
+      "step": 2386
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5728441553890815,
+      "learning_rate": 9.878185802244666e-07,
+      "loss": 0.8232,
+      "step": 2387
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.4865858199313449,
+      "learning_rate": 9.84570775623585e-07,
+      "loss": 0.8128,
+      "step": 2388
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.584633626306075,
+      "learning_rate": 9.813277357841928e-07,
+      "loss": 0.9106,
+      "step": 2389
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.5616213277042632,
+      "learning_rate": 9.780894645545215e-07,
+      "loss": 0.8123,
+      "step": 2390
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.6057294795052606,
+      "learning_rate": 9.748559657771472e-07,
+      "loss": 0.8917,
+      "step": 2391
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.514482572221518,
+      "learning_rate": 9.716272432889845e-07,
+      "loss": 0.7897,
+      "step": 2392
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.529652689041722,
+      "learning_rate": 9.684033009212752e-07,
+      "loss": 0.7989,
+      "step": 2393
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.41290673120441146,
+      "learning_rate": 9.651841424995933e-07,
+      "loss": 0.5745,
+      "step": 2394
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6698834409707044,
+      "learning_rate": 9.619697718438358e-07,
+      "loss": 0.6618,
+      "step": 2395
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.1046814753976808,
+      "learning_rate": 9.587601927682162e-07,
+      "loss": 0.9944,
+      "step": 2396
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.4910190354676186,
+      "learning_rate": 9.55555409081263e-07,
+      "loss": 0.7734,
+      "step": 2397
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5300508310762102,
+      "learning_rate": 9.52355424585818e-07,
+      "loss": 0.9785,
+      "step": 2398
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.33991494828149094,
+      "learning_rate": 9.491602430790209e-07,
+      "loss": 0.9504,
+      "step": 2399
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6165417063391279,
+      "learning_rate": 9.459698683523205e-07,
+      "loss": 0.691,
+      "step": 2400
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.36552055906124176,
+      "learning_rate": 9.427843041914541e-07,
+      "loss": 0.6047,
+      "step": 2401
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6295462382184938,
+      "learning_rate": 9.396035543764559e-07,
+      "loss": 0.845,
+      "step": 2402
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.544788755725605,
+      "learning_rate": 9.364276226816476e-07,
+      "loss": 0.8828,
+      "step": 2403
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.42633952556720056,
+      "learning_rate": 9.332565128756283e-07,
+      "loss": 0.654,
+      "step": 2404
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5554308501541663,
+      "learning_rate": 9.300902287212804e-07,
+      "loss": 0.9387,
+      "step": 2405
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6508121422342547,
+      "learning_rate": 9.269287739757604e-07,
+      "loss": 0.8426,
+      "step": 2406
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 1.2576610044393384,
+      "learning_rate": 9.237721523904891e-07,
+      "loss": 0.7643,
+      "step": 2407
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.49604298693525356,
+      "learning_rate": 9.206203677111574e-07,
+      "loss": 0.8188,
+      "step": 2408
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6099033223171859,
+      "learning_rate": 9.174734236777156e-07,
+      "loss": 1.0256,
+      "step": 2409
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5586191344073941,
+      "learning_rate": 9.143313240243668e-07,
+      "loss": 0.9269,
+      "step": 2410
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.43775160954808756,
+      "learning_rate": 9.111940724795692e-07,
+      "loss": 0.5697,
+      "step": 2411
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6729801545132569,
+      "learning_rate": 9.080616727660286e-07,
+      "loss": 0.8203,
+      "step": 2412
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.37492942763700776,
+      "learning_rate": 9.049341286006891e-07,
+      "loss": 0.9262,
+      "step": 2413
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5916894429002136,
+      "learning_rate": 9.018114436947373e-07,
+      "loss": 0.6868,
+      "step": 2414
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5621510828266613,
+      "learning_rate": 8.986936217535941e-07,
+      "loss": 0.9601,
+      "step": 2415
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.6531552284037438,
+      "learning_rate": 8.955806664769051e-07,
+      "loss": 0.6997,
+      "step": 2416
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.56630399087382,
+      "learning_rate": 8.924725815585461e-07,
+      "loss": 0.6427,
+      "step": 2417
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.8686437453315595,
+      "learning_rate": 8.893693706866124e-07,
+      "loss": 0.9549,
+      "step": 2418
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5230650367977234,
+      "learning_rate": 8.862710375434136e-07,
+      "loss": 0.8497,
+      "step": 2419
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.4742727618091632,
+      "learning_rate": 8.831775858054747e-07,
+      "loss": 0.6181,
+      "step": 2420
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.7301256363280143,
+      "learning_rate": 8.80089019143524e-07,
+      "loss": 0.7585,
+      "step": 2421
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.4795692889380236,
+      "learning_rate": 8.770053412224972e-07,
+      "loss": 1.0932,
+      "step": 2422
+    },
+    {
+      "epoch": 0.81,
+      "grad_norm": 0.5215976833382091,
+      "learning_rate": 8.739265557015287e-07,
+      "loss": 0.7415,
+      "step": 2423
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.31884450132352865,
+      "learning_rate": 8.708526662339428e-07,
+      "loss": 0.896,
+      "step": 2424
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5069488678822675,
+      "learning_rate": 8.677836764672592e-07,
+      "loss": 0.8385,
+      "step": 2425
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8187850796046604,
+      "learning_rate": 8.647195900431832e-07,
+      "loss": 0.7695,
+      "step": 2426
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.4634221226502557,
+      "learning_rate": 8.616604105975973e-07,
+      "loss": 0.8651,
+      "step": 2427
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.43915561431769273,
+      "learning_rate": 8.586061417605668e-07,
+      "loss": 0.4958,
+      "step": 2428
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6428279289196709,
+      "learning_rate": 8.555567871563275e-07,
+      "loss": 0.862,
+      "step": 2429
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6958537274163064,
+      "learning_rate": 8.525123504032817e-07,
+      "loss": 0.7358,
+      "step": 2430
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6244326146521951,
+      "learning_rate": 8.494728351140003e-07,
+      "loss": 0.826,
+      "step": 2431
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5192039344304056,
+      "learning_rate": 8.46438244895213e-07,
+      "loss": 1.0847,
+      "step": 2432
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.43702106301054494,
+      "learning_rate": 8.434085833478023e-07,
+      "loss": 0.8958,
+      "step": 2433
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.4223840023342458,
+      "learning_rate": 8.403838540668058e-07,
+      "loss": 0.8381,
+      "step": 2434
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6523308244859327,
+      "learning_rate": 8.373640606414097e-07,
+      "loss": 0.7244,
+      "step": 2435
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.33205529022419933,
+      "learning_rate": 8.343492066549369e-07,
+      "loss": 0.8948,
+      "step": 2436
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.46250948118739105,
+      "learning_rate": 8.313392956848559e-07,
+      "loss": 0.5917,
+      "step": 2437
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5223188944484948,
+      "learning_rate": 8.283343313027654e-07,
+      "loss": 0.7804,
+      "step": 2438
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.7484528100190991,
+      "learning_rate": 8.25334317074396e-07,
+      "loss": 0.8055,
+      "step": 2439
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6130911960258897,
+      "learning_rate": 8.223392565596061e-07,
+      "loss": 0.7506,
+      "step": 2440
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5836142107645957,
+      "learning_rate": 8.193491533123715e-07,
+      "loss": 0.9444,
+      "step": 2441
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6229369287801725,
+      "learning_rate": 8.163640108807897e-07,
+      "loss": 0.8942,
+      "step": 2442
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.4950051363542739,
+      "learning_rate": 8.133838328070714e-07,
+      "loss": 0.7156,
+      "step": 2443
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 1.5369445780861388,
+      "learning_rate": 8.104086226275332e-07,
+      "loss": 0.9746,
+      "step": 2444
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.44049900018143634,
+      "learning_rate": 8.074383838725997e-07,
+      "loss": 0.9056,
+      "step": 2445
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.4103291756379522,
+      "learning_rate": 8.044731200667966e-07,
+      "loss": 0.616,
+      "step": 2446
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.544018764852163,
+      "learning_rate": 8.015128347287438e-07,
+      "loss": 0.7183,
+      "step": 2447
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.3459662022451798,
+      "learning_rate": 7.985575313711547e-07,
+      "loss": 0.9473,
+      "step": 2448
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5724945583053956,
+      "learning_rate": 7.956072135008336e-07,
+      "loss": 0.6653,
+      "step": 2449
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.8573516478348037,
+      "learning_rate": 7.926618846186646e-07,
+      "loss": 0.7884,
+      "step": 2450
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.5522724218012416,
+      "learning_rate": 7.897215482196146e-07,
+      "loss": 0.935,
+      "step": 2451
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.6470976983830455,
+      "learning_rate": 7.867862077927269e-07,
+      "loss": 0.6818,
+      "step": 2452
+    },
+    {
+      "epoch": 0.82,
+      "grad_norm": 0.47954224134562584,
+      "learning_rate": 7.838558668211132e-07,
+      "loss": 0.9252,
+      "step": 2453
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5101061088462782,
+      "learning_rate": 7.809305287819557e-07,
+      "loss": 0.4846,
+      "step": 2454
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.4852115336212447,
+      "learning_rate": 7.780101971465015e-07,
+      "loss": 0.7351,
+      "step": 2455
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.41879610102369325,
+      "learning_rate": 7.750948753800508e-07,
+      "loss": 0.962,
+      "step": 2456
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5627479305691072,
+      "learning_rate": 7.721845669419664e-07,
+      "loss": 0.8754,
+      "step": 2457
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5630085077323362,
+      "learning_rate": 7.692792752856564e-07,
+      "loss": 0.934,
+      "step": 2458
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6380879565667655,
+      "learning_rate": 7.663790038585794e-07,
+      "loss": 0.7593,
+      "step": 2459
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5929295151000629,
+      "learning_rate": 7.634837561022373e-07,
+      "loss": 0.8939,
+      "step": 2460
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6692296449827251,
+      "learning_rate": 7.605935354521665e-07,
+      "loss": 0.7681,
+      "step": 2461
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.33595157355322824,
+      "learning_rate": 7.577083453379425e-07,
+      "loss": 0.9081,
+      "step": 2462
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.48081914692817945,
+      "learning_rate": 7.548281891831715e-07,
+      "loss": 0.6978,
+      "step": 2463
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.3962956463866398,
+      "learning_rate": 7.519530704054828e-07,
+      "loss": 0.5802,
+      "step": 2464
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.547338327459814,
+      "learning_rate": 7.490829924165311e-07,
+      "loss": 0.9749,
+      "step": 2465
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.4617579924004818,
+      "learning_rate": 7.462179586219897e-07,
+      "loss": 0.7876,
+      "step": 2466
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5325936038231542,
+      "learning_rate": 7.433579724215445e-07,
+      "loss": 0.8613,
+      "step": 2467
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.4885214212083694,
+      "learning_rate": 7.405030372088934e-07,
+      "loss": 0.9277,
+      "step": 2468
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.9825670982030871,
+      "learning_rate": 7.376531563717415e-07,
+      "loss": 0.7857,
+      "step": 2469
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5257167197530623,
+      "learning_rate": 7.348083332917927e-07,
+      "loss": 0.9395,
+      "step": 2470
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6530691837830932,
+      "learning_rate": 7.319685713447533e-07,
+      "loss": 0.8603,
+      "step": 2471
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5106815785932725,
+      "learning_rate": 7.291338739003234e-07,
+      "loss": 0.8401,
+      "step": 2472
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.28543742142924566,
+      "learning_rate": 7.263042443221907e-07,
+      "loss": 0.7565,
+      "step": 2473
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5756246146562601,
+      "learning_rate": 7.23479685968031e-07,
+      "loss": 0.8072,
+      "step": 2474
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5439643503601778,
+      "learning_rate": 7.206602021895054e-07,
+      "loss": 0.9023,
+      "step": 2475
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.57073156465017,
+      "learning_rate": 7.178457963322471e-07,
+      "loss": 0.8623,
+      "step": 2476
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6752425118968287,
+      "learning_rate": 7.150364717358699e-07,
+      "loss": 0.9194,
+      "step": 2477
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6795556722756861,
+      "learning_rate": 7.122322317339542e-07,
+      "loss": 0.7683,
+      "step": 2478
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.595276659862919,
+      "learning_rate": 7.094330796540483e-07,
+      "loss": 0.933,
+      "step": 2479
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.510106908247762,
+      "learning_rate": 7.066390188176647e-07,
+      "loss": 0.9239,
+      "step": 2480
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.4548186148682848,
+      "learning_rate": 7.038500525402709e-07,
+      "loss": 0.5569,
+      "step": 2481
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.5630014456819465,
+      "learning_rate": 7.010661841312921e-07,
+      "loss": 0.8804,
+      "step": 2482
+    },
+    {
+      "epoch": 0.83,
+      "grad_norm": 0.6046131643087528,
+      "learning_rate": 6.98287416894105e-07,
+      "loss": 0.8808,
+      "step": 2483
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5816302370301653,
+      "learning_rate": 6.955137541260287e-07,
+      "loss": 0.8749,
+      "step": 2484
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6275729821893652,
+      "learning_rate": 6.927451991183292e-07,
+      "loss": 0.7546,
+      "step": 2485
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.30620693693132617,
+      "learning_rate": 6.899817551562127e-07,
+      "loss": 0.929,
+      "step": 2486
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6597707509759836,
+      "learning_rate": 6.872234255188154e-07,
+      "loss": 0.827,
+      "step": 2487
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6574908574285901,
+      "learning_rate": 6.844702134792086e-07,
+      "loss": 0.7277,
+      "step": 2488
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.4862181386485927,
+      "learning_rate": 6.817221223043918e-07,
+      "loss": 0.8203,
+      "step": 2489
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.37814238834573194,
+      "learning_rate": 6.789791552552838e-07,
+      "loss": 0.5201,
+      "step": 2490
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.4971243459596727,
+      "learning_rate": 6.762413155867276e-07,
+      "loss": 0.9201,
+      "step": 2491
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5695552942253646,
+      "learning_rate": 6.735086065474799e-07,
+      "loss": 0.7722,
+      "step": 2492
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5738083233225538,
+      "learning_rate": 6.707810313802077e-07,
+      "loss": 0.8048,
+      "step": 2493
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.4386734694733362,
+      "learning_rate": 6.680585933214895e-07,
+      "loss": 0.7805,
+      "step": 2494
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7266392530617899,
+      "learning_rate": 6.653412956018068e-07,
+      "loss": 0.7924,
+      "step": 2495
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5250415414227975,
+      "learning_rate": 6.626291414455388e-07,
+      "loss": 0.9552,
+      "step": 2496
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.3205883390545972,
+      "learning_rate": 6.599221340709661e-07,
+      "loss": 0.9433,
+      "step": 2497
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.2124957604230473,
+      "learning_rate": 6.572202766902569e-07,
+      "loss": 0.9318,
+      "step": 2498
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.4734497023193996,
+      "learning_rate": 6.545235725094723e-07,
+      "loss": 0.4513,
+      "step": 2499
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5129422438860668,
+      "learning_rate": 6.518320247285581e-07,
+      "loss": 0.8155,
+      "step": 2500
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5620982468152408,
+      "learning_rate": 6.49145636541339e-07,
+      "loss": 0.8408,
+      "step": 2501
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5762728395830209,
+      "learning_rate": 6.46464411135519e-07,
+      "loss": 0.7745,
+      "step": 2502
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.4453421613775968,
+      "learning_rate": 6.437883516926774e-07,
+      "loss": 0.9858,
+      "step": 2503
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.634120375683643,
+      "learning_rate": 6.411174613882593e-07,
+      "loss": 0.8101,
+      "step": 2504
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 1.053223938368053,
+      "learning_rate": 6.384517433915794e-07,
+      "loss": 0.6697,
+      "step": 2505
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.49059324637678153,
+      "learning_rate": 6.357912008658151e-07,
+      "loss": 0.7617,
+      "step": 2506
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.40494721099338815,
+      "learning_rate": 6.331358369680002e-07,
+      "loss": 0.6358,
+      "step": 2507
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.6347925439284512,
+      "learning_rate": 6.30485654849024e-07,
+      "loss": 0.992,
+      "step": 2508
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5664608725013189,
+      "learning_rate": 6.278406576536306e-07,
+      "loss": 0.804,
+      "step": 2509
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.7051948942692748,
+      "learning_rate": 6.252008485204053e-07,
+      "loss": 0.9063,
+      "step": 2510
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.33517078530569405,
+      "learning_rate": 6.22566230581782e-07,
+      "loss": 0.8987,
+      "step": 2511
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5916709369278125,
+      "learning_rate": 6.199368069640343e-07,
+      "loss": 0.6669,
+      "step": 2512
+    },
+    {
+      "epoch": 0.84,
+      "grad_norm": 0.5088555817002856,
+      "learning_rate": 6.173125807872687e-07,
+      "loss": 1.0009,
+      "step": 2513
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5755460287846117,
+      "learning_rate": 6.146935551654298e-07,
+      "loss": 0.7503,
+      "step": 2514
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6604449410209562,
+      "learning_rate": 6.12079733206285e-07,
+      "loss": 0.995,
+      "step": 2515
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6045259321515112,
+      "learning_rate": 6.09471118011431e-07,
+      "loss": 0.7731,
+      "step": 2516
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5577844721696263,
+      "learning_rate": 6.068677126762867e-07,
+      "loss": 0.8912,
+      "step": 2517
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.35035531112123613,
+      "learning_rate": 6.042695202900855e-07,
+      "loss": 0.4719,
+      "step": 2518
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5676559768477955,
+      "learning_rate": 6.016765439358774e-07,
+      "loss": 0.8985,
+      "step": 2519
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5680401752939954,
+      "learning_rate": 5.990887866905237e-07,
+      "loss": 0.9935,
+      "step": 2520
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5972485466752125,
+      "learning_rate": 5.965062516246895e-07,
+      "loss": 0.7927,
+      "step": 2521
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5370434397401336,
+      "learning_rate": 5.939289418028455e-07,
+      "loss": 0.9135,
+      "step": 2522
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.3966486635231103,
+      "learning_rate": 5.913568602832637e-07,
+      "loss": 0.8759,
+      "step": 2523
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5303224040868073,
+      "learning_rate": 5.887900101180067e-07,
+      "loss": 0.7873,
+      "step": 2524
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5922424753304288,
+      "learning_rate": 5.862283943529351e-07,
+      "loss": 0.917,
+      "step": 2525
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.320574760215839,
+      "learning_rate": 5.836720160276971e-07,
+      "loss": 0.6206,
+      "step": 2526
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5782927659562446,
+      "learning_rate": 5.811208781757221e-07,
+      "loss": 0.9201,
+      "step": 2527
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6025726732254959,
+      "learning_rate": 5.785749838242261e-07,
+      "loss": 0.8743,
+      "step": 2528
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5645621106831764,
+      "learning_rate": 5.760343359942011e-07,
+      "loss": 0.9344,
+      "step": 2529
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.7667480358989183,
+      "learning_rate": 5.734989377004119e-07,
+      "loss": 0.8304,
+      "step": 2530
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.48259700470550165,
+      "learning_rate": 5.709687919513974e-07,
+      "loss": 0.7737,
+      "step": 2531
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5431923532640464,
+      "learning_rate": 5.684439017494614e-07,
+      "loss": 0.8376,
+      "step": 2532
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5291490371496828,
+      "learning_rate": 5.659242700906719e-07,
+      "loss": 0.6786,
+      "step": 2533
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5790092231185603,
+      "learning_rate": 5.63409899964858e-07,
+      "loss": 0.9953,
+      "step": 2534
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.3066457632081577,
+      "learning_rate": 5.609007943556028e-07,
+      "loss": 0.7509,
+      "step": 2535
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5778931458992531,
+      "learning_rate": 5.583969562402447e-07,
+      "loss": 0.8171,
+      "step": 2536
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6760751756216472,
+      "learning_rate": 5.558983885898739e-07,
+      "loss": 0.7607,
+      "step": 2537
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.45067476606792556,
+      "learning_rate": 5.534050943693197e-07,
+      "loss": 0.9038,
+      "step": 2538
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.5408088100351446,
+      "learning_rate": 5.509170765371608e-07,
+      "loss": 0.7975,
+      "step": 2539
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.4862538183164935,
+      "learning_rate": 5.484343380457124e-07,
+      "loss": 0.8327,
+      "step": 2540
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6048067159579239,
+      "learning_rate": 5.459568818410221e-07,
+      "loss": 0.8959,
+      "step": 2541
+    },
+    {
+      "epoch": 0.85,
+      "grad_norm": 0.6862542456350015,
+      "learning_rate": 5.434847108628749e-07,
+      "loss": 0.8461,
+      "step": 2542
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.3806177582978116,
+      "learning_rate": 5.410178280447814e-07,
+      "loss": 0.4558,
+      "step": 2543
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6700985570419377,
+      "learning_rate": 5.385562363139757e-07,
+      "loss": 0.8881,
+      "step": 2544
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 1.047731098523148,
+      "learning_rate": 5.360999385914156e-07,
+      "loss": 0.7955,
+      "step": 2545
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5159691408197927,
+      "learning_rate": 5.336489377917786e-07,
+      "loss": 0.8851,
+      "step": 2546
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.31646799827610944,
+      "learning_rate": 5.312032368234527e-07,
+      "loss": 0.9039,
+      "step": 2547
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6051084689275906,
+      "learning_rate": 5.287628385885385e-07,
+      "loss": 0.7045,
+      "step": 2548
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7291561358613089,
+      "learning_rate": 5.263277459828475e-07,
+      "loss": 1.0152,
+      "step": 2549
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5742836326217083,
+      "learning_rate": 5.2389796189589e-07,
+      "loss": 0.7862,
+      "step": 2550
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.4723642028229018,
+      "learning_rate": 5.214734892108813e-07,
+      "loss": 0.8438,
+      "step": 2551
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.38776032063237803,
+      "learning_rate": 5.190543308047336e-07,
+      "loss": 0.5818,
+      "step": 2552
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.4763242388289529,
+      "learning_rate": 5.166404895480509e-07,
+      "loss": 0.9301,
+      "step": 2553
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5931274773574409,
+      "learning_rate": 5.1423196830513e-07,
+      "loss": 0.7344,
+      "step": 2554
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5367518062776178,
+      "learning_rate": 5.11828769933953e-07,
+      "loss": 0.745,
+      "step": 2555
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6537841265599356,
+      "learning_rate": 5.094308972861878e-07,
+      "loss": 0.93,
+      "step": 2556
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6430147077905457,
+      "learning_rate": 5.070383532071826e-07,
+      "loss": 0.7536,
+      "step": 2557
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.673697872010758,
+      "learning_rate": 5.0465114053596e-07,
+      "loss": 0.9083,
+      "step": 2558
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.3500092098694658,
+      "learning_rate": 5.022692621052189e-07,
+      "loss": 0.8611,
+      "step": 2559
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5254893223801428,
+      "learning_rate": 4.998927207413295e-07,
+      "loss": 0.9867,
+      "step": 2560
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.33390750489796495,
+      "learning_rate": 4.975215192643246e-07,
+      "loss": 0.7006,
+      "step": 2561
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5077258267146987,
+      "learning_rate": 4.951556604879049e-07,
+      "loss": 0.6688,
+      "step": 2562
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.486359838489138,
+      "learning_rate": 4.927951472194298e-07,
+      "loss": 0.8622,
+      "step": 2563
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5379492522392157,
+      "learning_rate": 4.904399822599143e-07,
+      "loss": 0.8688,
+      "step": 2564
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.5146689429716884,
+      "learning_rate": 4.88090168404029e-07,
+      "loss": 0.7561,
+      "step": 2565
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.8429351277639259,
+      "learning_rate": 4.857457084400957e-07,
+      "loss": 0.788,
+      "step": 2566
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.6721801302370304,
+      "learning_rate": 4.834066051500785e-07,
+      "loss": 0.8007,
+      "step": 2567
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.48812553569717776,
+      "learning_rate": 4.81072861309591e-07,
+      "loss": 0.8425,
+      "step": 2568
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7007833178055408,
+      "learning_rate": 4.787444796878843e-07,
+      "loss": 0.7655,
+      "step": 2569
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.4197258533476604,
+      "learning_rate": 4.7642146304784475e-07,
+      "loss": 0.6541,
+      "step": 2570
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.7271130355834613,
+      "learning_rate": 4.7410381414599635e-07,
+      "loss": 0.7674,
+      "step": 2571
+    },
+    {
+      "epoch": 0.86,
+      "grad_norm": 0.2924672239881524,
+      "learning_rate": 4.717915357324926e-07,
+      "loss": 0.9748,
+      "step": 2572
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5244044207110666,
+      "learning_rate": 4.6948463055111193e-07,
+      "loss": 0.6761,
+      "step": 2573
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.49819512977483993,
+      "learning_rate": 4.6718310133926084e-07,
+      "loss": 0.8143,
+      "step": 2574
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5561348228543783,
+      "learning_rate": 4.648869508279613e-07,
+      "loss": 0.9089,
+      "step": 2575
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.9561736429322735,
+      "learning_rate": 4.625961817418578e-07,
+      "loss": 0.7897,
+      "step": 2576
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.591844297376515,
+      "learning_rate": 4.603107967992082e-07,
+      "loss": 0.8964,
+      "step": 2577
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6414254845174617,
+      "learning_rate": 4.5803079871187816e-07,
+      "loss": 0.7873,
+      "step": 2578
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6363703024889703,
+      "learning_rate": 4.557561901853452e-07,
+      "loss": 0.928,
+      "step": 2579
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.3809183582017167,
+      "learning_rate": 4.5348697391869034e-07,
+      "loss": 0.456,
+      "step": 2580
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.4976429629039807,
+      "learning_rate": 4.5122315260459347e-07,
+      "loss": 0.6923,
+      "step": 2581
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6022937749148645,
+      "learning_rate": 4.4896472892933693e-07,
+      "loss": 0.9086,
+      "step": 2582
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5593877318367789,
+      "learning_rate": 4.4671170557279597e-07,
+      "loss": 0.8117,
+      "step": 2583
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.4918306921839903,
+      "learning_rate": 4.4446408520843696e-07,
+      "loss": 0.962,
+      "step": 2584
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.373386921429944,
+      "learning_rate": 4.4222187050331657e-07,
+      "loss": 0.9481,
+      "step": 2585
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5362846680426259,
+      "learning_rate": 4.39985064118077e-07,
+      "loss": 0.8142,
+      "step": 2586
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.41835956502163396,
+      "learning_rate": 4.3775366870694114e-07,
+      "loss": 0.4856,
+      "step": 2587
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7767684677355786,
+      "learning_rate": 4.355276869177127e-07,
+      "loss": 0.841,
+      "step": 2588
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5480823048018573,
+      "learning_rate": 4.333071213917722e-07,
+      "loss": 0.8026,
+      "step": 2589
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.48161303622610874,
+      "learning_rate": 4.310919747640707e-07,
+      "loss": 0.696,
+      "step": 2590
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.49392229716633823,
+      "learning_rate": 4.288822496631323e-07,
+      "loss": 1.009,
+      "step": 2591
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.7389799389363477,
+      "learning_rate": 4.266779487110434e-07,
+      "loss": 0.7511,
+      "step": 2592
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5740324933143502,
+      "learning_rate": 4.244790745234584e-07,
+      "loss": 0.7846,
+      "step": 2593
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5793189469733505,
+      "learning_rate": 4.222856297095912e-07,
+      "loss": 0.8506,
+      "step": 2594
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.43723681146085724,
+      "learning_rate": 4.200976168722104e-07,
+      "loss": 0.9839,
+      "step": 2595
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.3872443743014659,
+      "learning_rate": 4.179150386076425e-07,
+      "loss": 0.6081,
+      "step": 2596
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6180693230761537,
+      "learning_rate": 4.157378975057647e-07,
+      "loss": 0.766,
+      "step": 2597
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5331052120535124,
+      "learning_rate": 4.135661961499987e-07,
+      "loss": 1.0011,
+      "step": 2598
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.47032522828764634,
+      "learning_rate": 4.113999371173155e-07,
+      "loss": 0.8876,
+      "step": 2599
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.4945824548670708,
+      "learning_rate": 4.092391229782278e-07,
+      "loss": 0.6974,
+      "step": 2600
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.6170438618716793,
+      "learning_rate": 4.07083756296785e-07,
+      "loss": 0.9587,
+      "step": 2601
+    },
+    {
+      "epoch": 0.87,
+      "grad_norm": 0.5661032579261599,
+      "learning_rate": 4.0493383963057354e-07,
+      "loss": 0.6661,
+      "step": 2602
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.515186728222038,
+      "learning_rate": 4.027893755307144e-07,
+      "loss": 0.8138,
+      "step": 2603
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6576035374138204,
+      "learning_rate": 4.0065036654185543e-07,
+      "loss": 0.8238,
+      "step": 2604
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5091607077544018,
+      "learning_rate": 3.9851681520217346e-07,
+      "loss": 0.8583,
+      "step": 2605
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.3598929603027902,
+      "learning_rate": 3.9638872404337057e-07,
+      "loss": 0.6006,
+      "step": 2606
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.533852896601436,
+      "learning_rate": 3.942660955906652e-07,
+      "loss": 0.8929,
+      "step": 2607
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6042954831253619,
+      "learning_rate": 3.921489323627975e-07,
+      "loss": 1.052,
+      "step": 2608
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6581515205035079,
+      "learning_rate": 3.900372368720223e-07,
+      "loss": 0.7267,
+      "step": 2609
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.3765872367093276,
+      "learning_rate": 3.8793101162410417e-07,
+      "loss": 0.9065,
+      "step": 2610
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5761585252406775,
+      "learning_rate": 3.8583025911831885e-07,
+      "loss": 0.8134,
+      "step": 2611
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5871248885036193,
+      "learning_rate": 3.8373498184744564e-07,
+      "loss": 0.7251,
+      "step": 2612
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.4603562655311591,
+      "learning_rate": 3.816451822977696e-07,
+      "loss": 0.8227,
+      "step": 2613
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.8234901248009182,
+      "learning_rate": 3.795608629490738e-07,
+      "loss": 0.5194,
+      "step": 2614
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.48028276466805775,
+      "learning_rate": 3.774820262746398e-07,
+      "loss": 0.9033,
+      "step": 2615
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6487625003737979,
+      "learning_rate": 3.754086747412422e-07,
+      "loss": 0.7459,
+      "step": 2616
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6199249560743013,
+      "learning_rate": 3.733408108091485e-07,
+      "loss": 0.966,
+      "step": 2617
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6887346018337255,
+      "learning_rate": 3.712784369321121e-07,
+      "loss": 0.9371,
+      "step": 2618
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.44974618545272094,
+      "learning_rate": 3.6922155555737415e-07,
+      "loss": 0.8895,
+      "step": 2619
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.4291353931929911,
+      "learning_rate": 3.6717016912565575e-07,
+      "loss": 0.8956,
+      "step": 2620
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6313970018143807,
+      "learning_rate": 3.6512428007116076e-07,
+      "loss": 0.7333,
+      "step": 2621
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.3437442764755801,
+      "learning_rate": 3.6308389082156835e-07,
+      "loss": 0.933,
+      "step": 2622
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.4317845918406489,
+      "learning_rate": 3.6104900379802966e-07,
+      "loss": 0.4628,
+      "step": 2623
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.48037253598317586,
+      "learning_rate": 3.5901962141516975e-07,
+      "loss": 0.7873,
+      "step": 2624
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5205319833409534,
+      "learning_rate": 3.569957460810819e-07,
+      "loss": 0.8143,
+      "step": 2625
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5759767143912632,
+      "learning_rate": 3.549773801973211e-07,
+      "loss": 0.6851,
+      "step": 2626
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6791182960210822,
+      "learning_rate": 3.5296452615890763e-07,
+      "loss": 0.9511,
+      "step": 2627
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.49974977330839093,
+      "learning_rate": 3.5095718635432206e-07,
+      "loss": 0.8664,
+      "step": 2628
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.504254099127737,
+      "learning_rate": 3.4895536316549847e-07,
+      "loss": 0.8658,
+      "step": 2629
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.5578724657776929,
+      "learning_rate": 3.469590589678284e-07,
+      "loss": 0.7973,
+      "step": 2630
+    },
+    {
+      "epoch": 0.88,
+      "grad_norm": 0.6740023094022166,
+      "learning_rate": 3.4496827613015206e-07,
+      "loss": 0.9018,
+      "step": 2631
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.4077937846162993,
+      "learning_rate": 3.4298301701475855e-07,
+      "loss": 0.5391,
+      "step": 2632
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5388193893977923,
+      "learning_rate": 3.41003283977383e-07,
+      "loss": 0.7565,
+      "step": 2633
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.3055304304651456,
+      "learning_rate": 3.3902907936720353e-07,
+      "loss": 0.9532,
+      "step": 2634
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.0698694674633302,
+      "learning_rate": 3.370604055268356e-07,
+      "loss": 0.7257,
+      "step": 2635
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.656496897626994,
+      "learning_rate": 3.3509726479233506e-07,
+      "loss": 0.8527,
+      "step": 2636
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5566517428673374,
+      "learning_rate": 3.3313965949319124e-07,
+      "loss": 0.885,
+      "step": 2637
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 1.2178864673840164,
+      "learning_rate": 3.3118759195232273e-07,
+      "loss": 0.737,
+      "step": 2638
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5141441478937359,
+      "learning_rate": 3.2924106448607997e-07,
+      "loss": 0.9028,
+      "step": 2639
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.4061300877298855,
+      "learning_rate": 3.27300079404238e-07,
+      "loss": 0.542,
+      "step": 2640
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.610753487866224,
+      "learning_rate": 3.253646390099946e-07,
+      "loss": 0.9629,
+      "step": 2641
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.44399750186750264,
+      "learning_rate": 3.234347455999709e-07,
+      "loss": 0.899,
+      "step": 2642
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5212466931583958,
+      "learning_rate": 3.2151040146420174e-07,
+      "loss": 0.8354,
+      "step": 2643
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5137255842833025,
+      "learning_rate": 3.195916088861406e-07,
+      "loss": 0.8782,
+      "step": 2644
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.534755340464617,
+      "learning_rate": 3.176783701426528e-07,
+      "loss": 0.7683,
+      "step": 2645
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6704206978945372,
+      "learning_rate": 3.157706875040112e-07,
+      "loss": 0.9002,
+      "step": 2646
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.34931206383695645,
+      "learning_rate": 3.138685632338978e-07,
+      "loss": 0.8986,
+      "step": 2647
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5779403257178612,
+      "learning_rate": 3.119719995893994e-07,
+      "loss": 0.8715,
+      "step": 2648
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7236204104021513,
+      "learning_rate": 3.1008099882100183e-07,
+      "loss": 0.7135,
+      "step": 2649
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.37211929703479063,
+      "learning_rate": 3.0819556317259304e-07,
+      "loss": 0.5523,
+      "step": 2650
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.4699637909028294,
+      "learning_rate": 3.0631569488145605e-07,
+      "loss": 0.9343,
+      "step": 2651
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5224784072615004,
+      "learning_rate": 3.0444139617826605e-07,
+      "loss": 0.7446,
+      "step": 2652
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7474344998488728,
+      "learning_rate": 3.0257266928709174e-07,
+      "loss": 0.9426,
+      "step": 2653
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.43445648876371834,
+      "learning_rate": 3.0070951642538925e-07,
+      "loss": 0.8402,
+      "step": 2654
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.7612680541420405,
+      "learning_rate": 2.988519398040002e-07,
+      "loss": 0.808,
+      "step": 2655
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.49854578441517433,
+      "learning_rate": 2.969999416271496e-07,
+      "loss": 0.9506,
+      "step": 2656
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.6114586194320406,
+      "learning_rate": 2.95153524092443e-07,
+      "loss": 0.8214,
+      "step": 2657
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.48066178949246385,
+      "learning_rate": 2.9331268939086334e-07,
+      "loss": 0.9,
+      "step": 2658
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.3014154512856185,
+      "learning_rate": 2.91477439706771e-07,
+      "loss": 0.7368,
+      "step": 2659
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 0.5688359615152105,
+      "learning_rate": 2.896477772178957e-07,
+      "loss": 0.9593,
+      "step": 2660
+    },
+    {
+      "epoch": 0.89,
+      "grad_norm": 3.756136720407377,
+      "learning_rate": 2.878237040953402e-07,
+      "loss": 0.7149,
+      "step": 2661
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5511115252395008,
+      "learning_rate": 2.860052225035742e-07,
+      "loss": 0.842,
+      "step": 2662
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5593884302048667,
+      "learning_rate": 2.8419233460043106e-07,
+      "loss": 0.8301,
+      "step": 2663
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6131410264461488,
+      "learning_rate": 2.823850425371083e-07,
+      "loss": 0.8463,
+      "step": 2664
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5096129466674121,
+      "learning_rate": 2.8058334845816214e-07,
+      "loss": 0.8765,
+      "step": 2665
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5395431384452533,
+      "learning_rate": 2.787872545015069e-07,
+      "loss": 0.9022,
+      "step": 2666
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.394865889941057,
+      "learning_rate": 2.7699676279841104e-07,
+      "loss": 0.6269,
+      "step": 2667
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.7416638685570395,
+      "learning_rate": 2.752118754734967e-07,
+      "loss": 0.7931,
+      "step": 2668
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.627508260350181,
+      "learning_rate": 2.734325946447336e-07,
+      "loss": 0.8327,
+      "step": 2669
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5614341151310306,
+      "learning_rate": 2.716589224234406e-07,
+      "loss": 0.8966,
+      "step": 2670
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6167165897845092,
+      "learning_rate": 2.698908609142814e-07,
+      "loss": 0.7098,
+      "step": 2671
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.3125619853402098,
+      "learning_rate": 2.681284122152594e-07,
+      "loss": 0.936,
+      "step": 2672
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.626056748585571,
+      "learning_rate": 2.663715784177201e-07,
+      "loss": 0.7951,
+      "step": 2673
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.4558572354202087,
+      "learning_rate": 2.646203616063475e-07,
+      "loss": 0.6186,
+      "step": 2674
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5708358771366987,
+      "learning_rate": 2.628747638591561e-07,
+      "loss": 0.9796,
+      "step": 2675
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.40881920801321736,
+      "learning_rate": 2.6113478724749785e-07,
+      "loss": 0.4388,
+      "step": 2676
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.45213668808391305,
+      "learning_rate": 2.594004338360495e-07,
+      "loss": 0.9598,
+      "step": 2677
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6176419716671008,
+      "learning_rate": 2.576717056828193e-07,
+      "loss": 0.7317,
+      "step": 2678
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5652598668835487,
+      "learning_rate": 2.559486048391391e-07,
+      "loss": 0.7497,
+      "step": 2679
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.663679962469749,
+      "learning_rate": 2.542311333496622e-07,
+      "loss": 0.7982,
+      "step": 2680
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5798204278329373,
+      "learning_rate": 2.525192932523635e-07,
+      "loss": 0.794,
+      "step": 2681
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.4729945530069478,
+      "learning_rate": 2.5081308657853576e-07,
+      "loss": 0.899,
+      "step": 2682
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.31459803434340006,
+      "learning_rate": 2.4911251535278515e-07,
+      "loss": 0.9253,
+      "step": 2683
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5727203522966167,
+      "learning_rate": 2.474175815930324e-07,
+      "loss": 0.8776,
+      "step": 2684
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.4057606950493566,
+      "learning_rate": 2.457282873105099e-07,
+      "loss": 0.483,
+      "step": 2685
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5964104390574112,
+      "learning_rate": 2.4404463450975415e-07,
+      "loss": 0.7667,
+      "step": 2686
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 1.0090090353835717,
+      "learning_rate": 2.423666251886114e-07,
+      "loss": 0.7999,
+      "step": 2687
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5289143672703085,
+      "learning_rate": 2.406942613382296e-07,
+      "loss": 0.9637,
+      "step": 2688
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6584540794840685,
+      "learning_rate": 2.3902754494305734e-07,
+      "loss": 0.9206,
+      "step": 2689
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.6046330294413268,
+      "learning_rate": 2.3736647798084268e-07,
+      "loss": 0.8279,
+      "step": 2690
+    },
+    {
+      "epoch": 0.9,
+      "grad_norm": 0.5344683118119139,
+      "learning_rate": 2.3571106242262997e-07,
+      "loss": 0.8499,
+      "step": 2691
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.4778648509201441,
+      "learning_rate": 2.340613002327563e-07,
+      "loss": 0.6453,
+      "step": 2692
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.3783882238991578,
+      "learning_rate": 2.3241719336885182e-07,
+      "loss": 0.5873,
+      "step": 2693
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5148517884805066,
+      "learning_rate": 2.307787437818365e-07,
+      "loss": 0.9749,
+      "step": 2694
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6335051044919672,
+      "learning_rate": 2.2914595341591517e-07,
+      "loss": 0.8196,
+      "step": 2695
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.7309683784087092,
+      "learning_rate": 2.275188242085785e-07,
+      "loss": 0.8825,
+      "step": 2696
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.36259693831470197,
+      "learning_rate": 2.2589735809059986e-07,
+      "loss": 0.8605,
+      "step": 2697
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6007208878966803,
+      "learning_rate": 2.2428155698603182e-07,
+      "loss": 0.6786,
+      "step": 2698
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5072108923359918,
+      "learning_rate": 2.2267142281220678e-07,
+      "loss": 1.0108,
+      "step": 2699
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5467136068194405,
+      "learning_rate": 2.2106695747973027e-07,
+      "loss": 0.7708,
+      "step": 2700
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5861431976936672,
+      "learning_rate": 2.1946816289248163e-07,
+      "loss": 0.9796,
+      "step": 2701
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5802020214123531,
+      "learning_rate": 2.1787504094761268e-07,
+      "loss": 0.7382,
+      "step": 2702
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.398739193051506,
+      "learning_rate": 2.162875935355424e-07,
+      "loss": 0.5787,
+      "step": 2703
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5775172954531257,
+      "learning_rate": 2.147058225399573e-07,
+      "loss": 0.7038,
+      "step": 2704
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5767505277137507,
+      "learning_rate": 2.1312972983780823e-07,
+      "loss": 0.8806,
+      "step": 2705
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5495879985136222,
+      "learning_rate": 2.115593172993058e-07,
+      "loss": 0.9252,
+      "step": 2706
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5324884956644796,
+      "learning_rate": 2.0999458678792384e-07,
+      "loss": 0.8577,
+      "step": 2707
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5833257294146903,
+      "learning_rate": 2.0843554016039326e-07,
+      "loss": 0.915,
+      "step": 2708
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.36792917739296493,
+      "learning_rate": 2.0688217926669695e-07,
+      "loss": 0.8792,
+      "step": 2709
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.4884991412808175,
+      "learning_rate": 2.05334505950075e-07,
+      "loss": 0.8783,
+      "step": 2710
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6017604396049481,
+      "learning_rate": 2.0379252204701715e-07,
+      "loss": 0.7321,
+      "step": 2711
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.35730182662346804,
+      "learning_rate": 2.0225622938726097e-07,
+      "loss": 0.6666,
+      "step": 2712
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5040345491079155,
+      "learning_rate": 2.007256297937926e-07,
+      "loss": 0.8885,
+      "step": 2713
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6052359148304179,
+      "learning_rate": 1.9920072508284204e-07,
+      "loss": 0.7906,
+      "step": 2714
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.47100559356355576,
+      "learning_rate": 1.9768151706388016e-07,
+      "loss": 0.9535,
+      "step": 2715
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5904374362174968,
+      "learning_rate": 1.961680075396205e-07,
+      "loss": 0.7745,
+      "step": 2716
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.684797929882592,
+      "learning_rate": 1.9466019830601257e-07,
+      "loss": 0.7679,
+      "step": 2717
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.6643548861065802,
+      "learning_rate": 1.9315809115224348e-07,
+      "loss": 0.866,
+      "step": 2718
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.5259278383201087,
+      "learning_rate": 1.9166168786073303e-07,
+      "loss": 0.7433,
+      "step": 2719
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.4199309791845362,
+      "learning_rate": 1.9017099020713358e-07,
+      "loss": 0.6814,
+      "step": 2720
+    },
+    {
+      "epoch": 0.91,
+      "grad_norm": 0.29994982800052755,
+      "learning_rate": 1.8868599996032577e-07,
+      "loss": 0.8653,
+      "step": 2721
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5308648585661842,
+      "learning_rate": 1.8720671888242058e-07,
+      "loss": 0.9243,
+      "step": 2722
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.66187469027219,
+      "learning_rate": 1.857331487287506e-07,
+      "loss": 0.781,
+      "step": 2723
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.45562637199007056,
+      "learning_rate": 1.8426529124787372e-07,
+      "loss": 0.9237,
+      "step": 2724
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6340905964137458,
+      "learning_rate": 1.8280314818157118e-07,
+      "loss": 0.7977,
+      "step": 2725
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5563524444122532,
+      "learning_rate": 1.81346721264839e-07,
+      "loss": 0.8071,
+      "step": 2726
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6173040195133938,
+      "learning_rate": 1.7989601222589358e-07,
+      "loss": 0.8716,
+      "step": 2727
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.684162204335571,
+      "learning_rate": 1.7845102278616578e-07,
+      "loss": 0.9687,
+      "step": 2728
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.4188661269471277,
+      "learning_rate": 1.7701175466029895e-07,
+      "loss": 0.553,
+      "step": 2729
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7954204043372003,
+      "learning_rate": 1.75578209556147e-07,
+      "loss": 0.7603,
+      "step": 2730
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5223023135912356,
+      "learning_rate": 1.7415038917477478e-07,
+      "loss": 0.7696,
+      "step": 2731
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5825485651112952,
+      "learning_rate": 1.7272829521045142e-07,
+      "loss": 0.9254,
+      "step": 2732
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.33934403027088017,
+      "learning_rate": 1.7131192935065433e-07,
+      "loss": 0.9152,
+      "step": 2733
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5326109489587322,
+      "learning_rate": 1.699012932760602e-07,
+      "loss": 0.8379,
+      "step": 2734
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 1.145267986382217,
+      "learning_rate": 1.6849638866054896e-07,
+      "loss": 0.8801,
+      "step": 2735
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5839917998785583,
+      "learning_rate": 1.6709721717120042e-07,
+      "loss": 0.7341,
+      "step": 2736
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.4783225779541114,
+      "learning_rate": 1.657037804682876e-07,
+      "loss": 0.8723,
+      "step": 2737
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.36680364797601256,
+      "learning_rate": 1.6431608020528233e-07,
+      "loss": 0.5598,
+      "step": 2738
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.45618645339457875,
+      "learning_rate": 1.629341180288485e-07,
+      "loss": 0.925,
+      "step": 2739
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7009285203264635,
+      "learning_rate": 1.6155789557883994e-07,
+      "loss": 0.7396,
+      "step": 2740
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.49990846801691685,
+      "learning_rate": 1.6018741448830033e-07,
+      "loss": 0.8666,
+      "step": 2741
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7376883983070899,
+      "learning_rate": 1.5882267638346217e-07,
+      "loss": 0.8229,
+      "step": 2742
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5951082371800469,
+      "learning_rate": 1.574636828837395e-07,
+      "loss": 0.7169,
+      "step": 2743
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7453994938998925,
+      "learning_rate": 1.5611043560173345e-07,
+      "loss": 0.9324,
+      "step": 2744
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.3348605864294369,
+      "learning_rate": 1.5476293614322512e-07,
+      "loss": 0.8767,
+      "step": 2745
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5558525853131507,
+      "learning_rate": 1.5342118610717438e-07,
+      "loss": 1.0344,
+      "step": 2746
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.3189938665036729,
+      "learning_rate": 1.520851870857204e-07,
+      "loss": 0.6758,
+      "step": 2747
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.7138420764377298,
+      "learning_rate": 1.507549406641773e-07,
+      "loss": 0.7859,
+      "step": 2748
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.5470858092006471,
+      "learning_rate": 1.4943044842103238e-07,
+      "loss": 0.8653,
+      "step": 2749
+    },
+    {
+      "epoch": 0.92,
+      "grad_norm": 0.6369429617020219,
+      "learning_rate": 1.4811171192794628e-07,
+      "loss": 0.7239,
+      "step": 2750
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5235128762839479,
+      "learning_rate": 1.4679873274975e-07,
+      "loss": 0.7306,
+      "step": 2751
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5002887039270897,
+      "learning_rate": 1.4549151244444003e-07,
+      "loss": 0.8063,
+      "step": 2752
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6766288091746588,
+      "learning_rate": 1.4419005256318286e-07,
+      "loss": 0.9018,
+      "step": 2753
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5595510096070442,
+      "learning_rate": 1.42894354650307e-07,
+      "loss": 0.7615,
+      "step": 2754
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7621722771841607,
+      "learning_rate": 1.4160442024330533e-07,
+      "loss": 0.812,
+      "step": 2755
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.38713289776999676,
+      "learning_rate": 1.403202508728313e-07,
+      "loss": 0.637,
+      "step": 2756
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7191665360786699,
+      "learning_rate": 1.3904184806269705e-07,
+      "loss": 0.7405,
+      "step": 2757
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.33700108771310244,
+      "learning_rate": 1.3776921332987193e-07,
+      "loss": 0.9649,
+      "step": 2758
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6395848016058812,
+      "learning_rate": 1.3650234818448182e-07,
+      "loss": 0.6862,
+      "step": 2759
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.48245315508509773,
+      "learning_rate": 1.3524125412980482e-07,
+      "loss": 0.976,
+      "step": 2760
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5406018260229588,
+      "learning_rate": 1.3398593266227223e-07,
+      "loss": 0.7898,
+      "step": 2761
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.3213801701873649,
+      "learning_rate": 1.3273638527146638e-07,
+      "loss": 0.7949,
+      "step": 2762
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5258984485472166,
+      "learning_rate": 1.3149261344011456e-07,
+      "loss": 0.9123,
+      "step": 2763
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6522598831248193,
+      "learning_rate": 1.3025461864409395e-07,
+      "loss": 0.6967,
+      "step": 2764
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.36614358989274204,
+      "learning_rate": 1.2902240235242614e-07,
+      "loss": 0.6478,
+      "step": 2765
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6216542202064738,
+      "learning_rate": 1.277959660272743e-07,
+      "loss": 0.6803,
+      "step": 2766
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.552760807676953,
+      "learning_rate": 1.265753111239443e-07,
+      "loss": 0.7035,
+      "step": 2767
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.486142360441894,
+      "learning_rate": 1.253604390908819e-07,
+      "loss": 0.8822,
+      "step": 2768
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5536715796559076,
+      "learning_rate": 1.2415135136967015e-07,
+      "loss": 0.7929,
+      "step": 2769
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.48930802061625545,
+      "learning_rate": 1.2294804939502746e-07,
+      "loss": 1.0038,
+      "step": 2770
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.7246278176466333,
+      "learning_rate": 1.2175053459481e-07,
+      "loss": 0.9335,
+      "step": 2771
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.5932924440461,
+      "learning_rate": 1.2055880839000277e-07,
+      "loss": 0.8869,
+      "step": 2772
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.4014660586042811,
+      "learning_rate": 1.1937287219472515e-07,
+      "loss": 0.5171,
+      "step": 2773
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 1.0447965398064831,
+      "learning_rate": 1.1819272741622367e-07,
+      "loss": 0.7855,
+      "step": 2774
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.4936079280071705,
+      "learning_rate": 1.1701837545487426e-07,
+      "loss": 0.7675,
+      "step": 2775
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6341064346777373,
+      "learning_rate": 1.1584981770417891e-07,
+      "loss": 0.7782,
+      "step": 2776
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6714627880748988,
+      "learning_rate": 1.1468705555076287e-07,
+      "loss": 0.989,
+      "step": 2777
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.590836823723485,
+      "learning_rate": 1.1353009037437523e-07,
+      "loss": 0.7576,
+      "step": 2778
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.541088881921656,
+      "learning_rate": 1.1237892354788726e-07,
+      "loss": 0.7696,
+      "step": 2779
+    },
+    {
+      "epoch": 0.93,
+      "grad_norm": 0.6969024951307478,
+      "learning_rate": 1.1123355643728795e-07,
+      "loss": 0.8504,
+      "step": 2780
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.4515980162293798,
+      "learning_rate": 1.1009399040168455e-07,
+      "loss": 0.9956,
+      "step": 2781
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.3777708735717382,
+      "learning_rate": 1.0896022679330265e-07,
+      "loss": 0.6324,
+      "step": 2782
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5461872477154747,
+      "learning_rate": 1.078322669574794e-07,
+      "loss": 0.7656,
+      "step": 2783
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5408677699088975,
+      "learning_rate": 1.0671011223266803e-07,
+      "loss": 0.9151,
+      "step": 2784
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.3536671402440689,
+      "learning_rate": 1.0559376395043285e-07,
+      "loss": 0.8997,
+      "step": 2785
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5427965306660374,
+      "learning_rate": 1.0448322343544537e-07,
+      "loss": 0.7095,
+      "step": 2786
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6445010641487571,
+      "learning_rate": 1.0337849200548978e-07,
+      "loss": 0.9596,
+      "step": 2787
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.758709625052614,
+      "learning_rate": 1.0227957097145413e-07,
+      "loss": 0.6199,
+      "step": 2788
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5527698375701893,
+      "learning_rate": 1.0118646163733203e-07,
+      "loss": 0.8553,
+      "step": 2789
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5926962146494529,
+      "learning_rate": 1.0009916530022256e-07,
+      "loss": 0.7884,
+      "step": 2790
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5456692890250201,
+      "learning_rate": 9.901768325032535e-08,
+      "loss": 0.747,
+      "step": 2791
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.383207318946179,
+      "learning_rate": 9.794201677094162e-08,
+      "loss": 0.6286,
+      "step": 2792
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5273789158997206,
+      "learning_rate": 9.687216713847203e-08,
+      "loss": 0.8496,
+      "step": 2793
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.46281120062259756,
+      "learning_rate": 9.580813562241276e-08,
+      "loss": 1.0212,
+      "step": 2794
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6474041920021065,
+      "learning_rate": 9.474992348535884e-08,
+      "loss": 0.7684,
+      "step": 2795
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.30240909289586443,
+      "learning_rate": 9.369753198299859e-08,
+      "loss": 0.9091,
+      "step": 2796
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5279574123205776,
+      "learning_rate": 9.265096236411364e-08,
+      "loss": 0.8213,
+      "step": 2797
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.7830390735065886,
+      "learning_rate": 9.161021587057728e-08,
+      "loss": 0.7298,
+      "step": 2798
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.47374925691975567,
+      "learning_rate": 9.05752937373533e-08,
+      "loss": 0.8484,
+      "step": 2799
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.3825511787202101,
+      "learning_rate": 8.95461971924938e-08,
+      "loss": 0.5439,
+      "step": 2800
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6104025622376883,
+      "learning_rate": 8.852292745713864e-08,
+      "loss": 0.9212,
+      "step": 2801
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.8354912262490044,
+      "learning_rate": 8.75054857455132e-08,
+      "loss": 0.7612,
+      "step": 2802
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5379534541432136,
+      "learning_rate": 8.649387326492676e-08,
+      "loss": 0.989,
+      "step": 2803
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5290298732175548,
+      "learning_rate": 8.54880912157724e-08,
+      "loss": 0.9236,
+      "step": 2804
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.4762358250438234,
+      "learning_rate": 8.448814079152489e-08,
+      "loss": 0.9114,
+      "step": 2805
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.6147927526901038,
+      "learning_rate": 8.34940231787379e-08,
+      "loss": 0.9066,
+      "step": 2806
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.8632146201709755,
+      "learning_rate": 8.2505739557045e-08,
+      "loss": 0.7095,
+      "step": 2807
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.3817817690025299,
+      "learning_rate": 8.152329109915646e-08,
+      "loss": 0.9135,
+      "step": 2808
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5752370141176025,
+      "learning_rate": 8.054667897085867e-08,
+      "loss": 0.5747,
+      "step": 2809
+    },
+    {
+      "epoch": 0.94,
+      "grad_norm": 0.5614524355868884,
+      "learning_rate": 7.957590433101293e-08,
+      "loss": 0.6995,
+      "step": 2810
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.4904067455077929,
+      "learning_rate": 7.861096833155225e-08,
+      "loss": 0.7617,
+      "step": 2811
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.548168118287348,
+      "learning_rate": 7.765187211748348e-08,
+      "loss": 0.734,
+      "step": 2812
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5127516320716964,
+      "learning_rate": 7.66986168268824e-08,
+      "loss": 0.9404,
+      "step": 2813
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5575673544955835,
+      "learning_rate": 7.575120359089416e-08,
+      "loss": 0.8683,
+      "step": 2814
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7283427557885618,
+      "learning_rate": 7.480963353373227e-08,
+      "loss": 0.8319,
+      "step": 2815
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5133440452111558,
+      "learning_rate": 7.387390777267689e-08,
+      "loss": 0.8119,
+      "step": 2816
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.46365466724022564,
+      "learning_rate": 7.294402741807094e-08,
+      "loss": 0.8741,
+      "step": 2817
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.41740772231917256,
+      "learning_rate": 7.201999357332346e-08,
+      "loss": 0.5866,
+      "step": 2818
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7992768734106472,
+      "learning_rate": 7.110180733490624e-08,
+      "loss": 0.7475,
+      "step": 2819
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.30706178953898783,
+      "learning_rate": 7.018946979234997e-08,
+      "loss": 0.9433,
+      "step": 2820
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6045707513904737,
+      "learning_rate": 6.928298202824646e-08,
+      "loss": 0.6472,
+      "step": 2821
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.7885961455528099,
+      "learning_rate": 6.838234511824748e-08,
+      "loss": 0.9376,
+      "step": 2822
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5777305024596016,
+      "learning_rate": 6.748756013105928e-08,
+      "loss": 0.7733,
+      "step": 2823
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5497038149909921,
+      "learning_rate": 6.659862812844586e-08,
+      "loss": 0.6668,
+      "step": 2824
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5024749951245818,
+      "learning_rate": 6.571555016522624e-08,
+      "loss": 0.9132,
+      "step": 2825
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.42270819989690284,
+      "learning_rate": 6.483832728927219e-08,
+      "loss": 0.5567,
+      "step": 2826
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6646089211416473,
+      "learning_rate": 6.396696054150719e-08,
+      "loss": 0.9312,
+      "step": 2827
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.42310224757320997,
+      "learning_rate": 6.310145095590803e-08,
+      "loss": 0.8838,
+      "step": 2828
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5840849420851104,
+      "learning_rate": 6.224179955949871e-08,
+      "loss": 0.8837,
+      "step": 2829
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5959223478388346,
+      "learning_rate": 6.138800737235384e-08,
+      "loss": 0.9245,
+      "step": 2830
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 1.1926643781225912,
+      "learning_rate": 6.054007540759299e-08,
+      "loss": 0.7521,
+      "step": 2831
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6490179641613555,
+      "learning_rate": 5.96980046713852e-08,
+      "loss": 0.9436,
+      "step": 2832
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.35637696035360067,
+      "learning_rate": 5.8861796162942294e-08,
+      "loss": 0.8687,
+      "step": 2833
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.530844954567643,
+      "learning_rate": 5.803145087451945e-08,
+      "loss": 0.8854,
+      "step": 2834
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.8901379139239417,
+      "learning_rate": 5.7206969791416265e-08,
+      "loss": 0.6931,
+      "step": 2835
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.3702988342075755,
+      "learning_rate": 5.638835389197295e-08,
+      "loss": 0.5121,
+      "step": 2836
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5818437609997049,
+      "learning_rate": 5.557560414756913e-08,
+      "loss": 0.9538,
+      "step": 2837
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.5512147913624965,
+      "learning_rate": 5.476872152262558e-08,
+      "loss": 0.7978,
+      "step": 2838
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.6715747202596091,
+      "learning_rate": 5.396770697459919e-08,
+      "loss": 0.8949,
+      "step": 2839
+    },
+    {
+      "epoch": 0.95,
+      "grad_norm": 0.9216489937042177,
+      "learning_rate": 5.317256145398408e-08,
+      "loss": 0.8882,
+      "step": 2840
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7962262586524093,
+      "learning_rate": 5.238328590431163e-08,
+      "loss": 0.8265,
+      "step": 2841
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.590779804959573,
+      "learning_rate": 5.159988126214543e-08,
+      "loss": 0.9443,
+      "step": 2842
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6617636236865038,
+      "learning_rate": 5.082234845708467e-08,
+      "loss": 0.7911,
+      "step": 2843
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5702915759357834,
+      "learning_rate": 5.005068841175964e-08,
+      "loss": 0.8642,
+      "step": 2844
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3132605893733664,
+      "learning_rate": 4.928490204183234e-08,
+      "loss": 0.7641,
+      "step": 2845
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5315179101765419,
+      "learning_rate": 4.852499025599533e-08,
+      "loss": 0.9122,
+      "step": 2846
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5607957104197093,
+      "learning_rate": 4.7770953955969537e-08,
+      "loss": 0.6754,
+      "step": 2847
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6705453730053647,
+      "learning_rate": 4.702279403650534e-08,
+      "loss": 0.8666,
+      "step": 2848
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.9633659857473588,
+      "learning_rate": 4.628051138537759e-08,
+      "loss": 0.7947,
+      "step": 2849
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5934889493045076,
+      "learning_rate": 4.5544106883390614e-08,
+      "loss": 0.8544,
+      "step": 2850
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.56401692529092,
+      "learning_rate": 4.481358140437042e-08,
+      "loss": 0.9098,
+      "step": 2851
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.517408894660512,
+      "learning_rate": 4.4088935815168045e-08,
+      "loss": 0.9086,
+      "step": 2852
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.36974702950159416,
+      "learning_rate": 4.337017097565899e-08,
+      "loss": 0.635,
+      "step": 2853
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5588752662335646,
+      "learning_rate": 4.265728773873767e-08,
+      "loss": 0.8185,
+      "step": 2854
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.7616276089591572,
+      "learning_rate": 4.195028695032133e-08,
+      "loss": 0.7774,
+      "step": 2855
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5126363706441097,
+      "learning_rate": 4.124916944934665e-08,
+      "loss": 0.8592,
+      "step": 2856
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5380266947179672,
+      "learning_rate": 4.05539360677687e-08,
+      "loss": 0.7083,
+      "step": 2857
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.355642439967243,
+      "learning_rate": 3.986458763056089e-08,
+      "loss": 0.9481,
+      "step": 2858
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5759981660544827,
+      "learning_rate": 3.91811249557128e-08,
+      "loss": 0.7759,
+      "step": 2859
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.4529200510345702,
+      "learning_rate": 3.850354885423069e-08,
+      "loss": 0.7829,
+      "step": 2860
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.6322083397374256,
+      "learning_rate": 3.783186013013473e-08,
+      "loss": 0.84,
+      "step": 2861
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.3751255626101526,
+      "learning_rate": 3.716605958046071e-08,
+      "loss": 0.4459,
+      "step": 2862
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.430554899098449,
+      "learning_rate": 3.6506147995254985e-08,
+      "loss": 0.9665,
+      "step": 2863
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5256549870686379,
+      "learning_rate": 3.585212615757838e-08,
+      "loss": 0.7346,
+      "step": 2864
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.560472985489554,
+      "learning_rate": 3.5203994843501786e-08,
+      "loss": 0.7607,
+      "step": 2865
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.485988300378393,
+      "learning_rate": 3.456175482210611e-08,
+      "loss": 0.7817,
+      "step": 2866
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 1.0113269680385297,
+      "learning_rate": 3.392540685548229e-08,
+      "loss": 0.8788,
+      "step": 2867
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.5580822157849757,
+      "learning_rate": 3.329495169872854e-08,
+      "loss": 0.9217,
+      "step": 2868
+    },
+    {
+      "epoch": 0.96,
+      "grad_norm": 0.4444599297838316,
+      "learning_rate": 3.2670390099951985e-08,
+      "loss": 0.9428,
+      "step": 2869
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5277596873267654,
+      "learning_rate": 3.205172280026536e-08,
+      "loss": 0.8364,
+      "step": 2870
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.38480870349451207,
+      "learning_rate": 3.143895053378698e-08,
+      "loss": 0.4786,
+      "step": 2871
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5888117677577624,
+      "learning_rate": 3.0832074027641324e-08,
+      "loss": 0.8067,
+      "step": 2872
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.1980122184837356,
+      "learning_rate": 3.023109400195623e-08,
+      "loss": 0.8382,
+      "step": 2873
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.43780142416770745,
+      "learning_rate": 2.9636011169861812e-08,
+      "loss": 0.8517,
+      "step": 2874
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 1.6716451787872126,
+      "learning_rate": 2.9046826237491554e-08,
+      "loss": 0.9696,
+      "step": 2875
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5698656923994554,
+      "learning_rate": 2.846353990398065e-08,
+      "loss": 0.6602,
+      "step": 2876
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5416043091045207,
+      "learning_rate": 2.7886152861463233e-08,
+      "loss": 0.8579,
+      "step": 2877
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5529315813969775,
+      "learning_rate": 2.7314665795075135e-08,
+      "loss": 0.6338,
+      "step": 2878
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.47318349345231586,
+      "learning_rate": 2.6749079382950572e-08,
+      "loss": 0.6138,
+      "step": 2879
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5410530994417471,
+      "learning_rate": 2.6189394296221583e-08,
+      "loss": 1.0112,
+      "step": 2880
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.7178382425710329,
+      "learning_rate": 2.5635611199018028e-08,
+      "loss": 0.7648,
+      "step": 2881
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5777504945610893,
+      "learning_rate": 2.508773074846649e-08,
+      "loss": 0.8784,
+      "step": 2882
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.3250709918312288,
+      "learning_rate": 2.4545753594688582e-08,
+      "loss": 0.8722,
+      "step": 2883
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.790021130807426,
+      "learning_rate": 2.4009680380802093e-08,
+      "loss": 0.8016,
+      "step": 2884
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.4457693189683683,
+      "learning_rate": 2.347951174291818e-08,
+      "loss": 1.0001,
+      "step": 2885
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5673565455688695,
+      "learning_rate": 2.295524831014251e-08,
+      "loss": 0.773,
+      "step": 2886
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5864726051047383,
+      "learning_rate": 2.2436890704572444e-08,
+      "loss": 1.0348,
+      "step": 2887
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6929737208527538,
+      "learning_rate": 2.192443954129819e-08,
+      "loss": 0.7258,
+      "step": 2888
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.40499915271931564,
+      "learning_rate": 2.1417895428401114e-08,
+      "loss": 0.5428,
+      "step": 2889
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5952410452407629,
+      "learning_rate": 2.0917258966953735e-08,
+      "loss": 0.8428,
+      "step": 2890
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.756508987426255,
+      "learning_rate": 2.0422530751016968e-08,
+      "loss": 0.9072,
+      "step": 2891
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.562784988159828,
+      "learning_rate": 1.9933711367641772e-08,
+      "loss": 0.9605,
+      "step": 2892
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5223128362688587,
+      "learning_rate": 1.9450801396869167e-08,
+      "loss": 0.8164,
+      "step": 2893
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.5653653582729405,
+      "learning_rate": 1.8973801411724668e-08,
+      "loss": 0.953,
+      "step": 2894
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.34917538687107036,
+      "learning_rate": 1.8502711978223843e-08,
+      "loss": 0.8529,
+      "step": 2895
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.4813075310182657,
+      "learning_rate": 1.8037533655367866e-08,
+      "loss": 0.9,
+      "step": 2896
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.6363106147233571,
+      "learning_rate": 1.757826699514298e-08,
+      "loss": 0.7527,
+      "step": 2897
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.4891448626272258,
+      "learning_rate": 1.7124912542520468e-08,
+      "loss": 0.6354,
+      "step": 2898
+    },
+    {
+      "epoch": 0.97,
+      "grad_norm": 0.686109792414538,
+      "learning_rate": 1.6677470835458343e-08,
+      "loss": 0.9769,
+      "step": 2899
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6702405584522849,
+      "learning_rate": 1.623594240489579e-08,
+      "loss": 0.8038,
+      "step": 2900
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5247407404278372,
+      "learning_rate": 1.580032777475593e-08,
+      "loss": 0.9193,
+      "step": 2901
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5974032996666557,
+      "learning_rate": 1.537062746194584e-08,
+      "loss": 0.8117,
+      "step": 2902
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5683312495427636,
+      "learning_rate": 1.4946841976352654e-08,
+      "loss": 0.6763,
+      "step": 2903
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6165818768597658,
+      "learning_rate": 1.4528971820846894e-08,
+      "loss": 0.9232,
+      "step": 2904
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5682173522427186,
+      "learning_rate": 1.411701749127803e-08,
+      "loss": 0.737,
+      "step": 2905
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.4378978287350944,
+      "learning_rate": 1.3710979476476705e-08,
+      "loss": 0.6589,
+      "step": 2906
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.39876467616102745,
+      "learning_rate": 1.3310858258253068e-08,
+      "loss": 0.8713,
+      "step": 2907
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5538576926511212,
+      "learning_rate": 1.2916654311396214e-08,
+      "loss": 0.9257,
+      "step": 2908
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 2.667823612226655,
+      "learning_rate": 1.2528368103674194e-08,
+      "loss": 0.7635,
+      "step": 2909
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.43681938065620957,
+      "learning_rate": 1.2146000095831777e-08,
+      "loss": 0.9074,
+      "step": 2910
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5419585122385139,
+      "learning_rate": 1.176955074159214e-08,
+      "loss": 0.8252,
+      "step": 2911
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5101341631928583,
+      "learning_rate": 1.1399020487655176e-08,
+      "loss": 0.8067,
+      "step": 2912
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5180297393971065,
+      "learning_rate": 1.1034409773697519e-08,
+      "loss": 0.9846,
+      "step": 2913
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.7406018531605014,
+      "learning_rate": 1.0675719032370303e-08,
+      "loss": 0.7516,
+      "step": 2914
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.34415693513907875,
+      "learning_rate": 1.0322948689300838e-08,
+      "loss": 0.5358,
+      "step": 2915
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.4967939373942957,
+      "learning_rate": 9.976099163091502e-09,
+      "loss": 0.7684,
+      "step": 2916
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5119625536644818,
+      "learning_rate": 9.635170865318067e-09,
+      "loss": 0.8263,
+      "step": 2917
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5194497070005091,
+      "learning_rate": 9.300164200530815e-09,
+      "loss": 0.8883,
+      "step": 2918
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.3233261985662603,
+      "learning_rate": 8.971079566252871e-09,
+      "loss": 0.9326,
+      "step": 2919
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 1.0874198466139002,
+      "learning_rate": 8.647917352980762e-09,
+      "loss": 0.7847,
+      "step": 2920
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5332963126185502,
+      "learning_rate": 8.330677944183297e-09,
+      "loss": 0.9065,
+      "step": 2921
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5945198787483508,
+      "learning_rate": 8.019361716299912e-09,
+      "loss": 0.8564,
+      "step": 2922
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5055690949269347,
+      "learning_rate": 7.713969038743441e-09,
+      "loss": 0.8655,
+      "step": 2923
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.4164822740426262,
+      "learning_rate": 7.4145002738967856e-09,
+      "loss": 0.5236,
+      "step": 2924
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5809660853375601,
+      "learning_rate": 7.120955777112915e-09,
+      "loss": 0.9163,
+      "step": 2925
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.6054445795130737,
+      "learning_rate": 6.833335896716531e-09,
+      "loss": 0.7366,
+      "step": 2926
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5195043655088286,
+      "learning_rate": 6.551640974000739e-09,
+      "loss": 0.8784,
+      "step": 2927
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5553683855834188,
+      "learning_rate": 6.275871343228157e-09,
+      "loss": 0.777,
+      "step": 2928
+    },
+    {
+      "epoch": 0.98,
+      "grad_norm": 0.5927560758999204,
+      "learning_rate": 6.00602733163147e-09,
+      "loss": 0.7462,
+      "step": 2929
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5556989146859648,
+      "learning_rate": 5.7421092594101004e-09,
+      "loss": 0.9395,
+      "step": 2930
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.42033997900221653,
+      "learning_rate": 5.484117439734094e-09,
+      "loss": 0.9144,
+      "step": 2931
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.514516985165545,
+      "learning_rate": 5.232052178738567e-09,
+      "loss": 0.8971,
+      "step": 2932
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.3411371218295664,
+      "learning_rate": 4.985913775528151e-09,
+      "loss": 0.6542,
+      "step": 2933
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6121204021791266,
+      "learning_rate": 4.745702522174211e-09,
+      "loss": 0.8291,
+      "step": 2934
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5509455867984832,
+      "learning_rate": 4.511418703713743e-09,
+      "loss": 0.8905,
+      "step": 2935
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6680192663236202,
+      "learning_rate": 4.283062598152143e-09,
+      "loss": 0.6911,
+      "step": 2936
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.4891611918048787,
+      "learning_rate": 4.060634476458769e-09,
+      "loss": 0.8001,
+      "step": 2937
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5739982494423539,
+      "learning_rate": 3.844134602570826e-09,
+      "loss": 0.8511,
+      "step": 2938
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6006624892733244,
+      "learning_rate": 3.633563233388926e-09,
+      "loss": 0.9033,
+      "step": 2939
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5658649982135233,
+      "learning_rate": 3.428920618780418e-09,
+      "loss": 0.7867,
+      "step": 2940
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5338440913720207,
+      "learning_rate": 3.230207001576613e-09,
+      "loss": 0.7625,
+      "step": 2941
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.4172889173237741,
+      "learning_rate": 3.037422617573893e-09,
+      "loss": 0.6074,
+      "step": 2942
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5836433013653454,
+      "learning_rate": 2.8505676955326023e-09,
+      "loss": 0.7267,
+      "step": 2943
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.2898016072099733,
+      "learning_rate": 2.6696424571776015e-09,
+      "loss": 0.9846,
+      "step": 2944
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6255876228868129,
+      "learning_rate": 2.4946471171966025e-09,
+      "loss": 0.6615,
+      "step": 2945
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5623025850967939,
+      "learning_rate": 2.3255818832423894e-09,
+      "loss": 0.9632,
+      "step": 2946
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5767177177076792,
+      "learning_rate": 2.162446955928932e-09,
+      "loss": 0.7505,
+      "step": 2947
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.8870568444992752,
+      "learning_rate": 2.0052425288347166e-09,
+      "loss": 0.8136,
+      "step": 2948
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.5005192883477103,
+      "learning_rate": 1.8539687885010815e-09,
+      "loss": 0.8701,
+      "step": 2949
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6024951219227335,
+      "learning_rate": 1.7086259144305507e-09,
+      "loss": 0.6885,
+      "step": 2950
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.3943046652575907,
+      "learning_rate": 1.5692140790896093e-09,
+      "loss": 0.6857,
+      "step": 2951
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6317622792854868,
+      "learning_rate": 1.4357334479059292e-09,
+      "loss": 0.6465,
+      "step": 2952
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.6160618847305471,
+      "learning_rate": 1.3081841792694783e-09,
+      "loss": 0.7319,
+      "step": 2953
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.4679770902273626,
+      "learning_rate": 1.1865664245314101e-09,
+      "loss": 0.8964,
+      "step": 2954
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.4857595380712989,
+      "learning_rate": 1.0708803280057301e-09,
+      "loss": 0.8868,
+      "step": 2955
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.9507665749867376,
+      "learning_rate": 9.61126026966519e-10,
+      "loss": 1.0178,
+      "step": 2956
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.3555367065827477,
+      "learning_rate": 8.573036516501542e-10,
+      "loss": 0.9004,
+      "step": 2957
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.7070878958335391,
+      "learning_rate": 7.594133252530888e-10,
+      "loss": 0.8671,
+      "step": 2958
+    },
+    {
+      "epoch": 0.99,
+      "grad_norm": 0.3802460966745559,
+      "learning_rate": 6.674551639340721e-10,
+      "loss": 0.4913,
+      "step": 2959
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5268483485170355,
+      "learning_rate": 5.814292768108187e-10,
+      "loss": 0.7894,
+      "step": 2960
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.46830852575474263,
+      "learning_rate": 5.013357659638951e-10,
+      "loss": 0.7791,
+      "step": 2961
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6357394466377969,
+      "learning_rate": 4.2717472643227785e-10,
+      "loss": 0.7246,
+      "step": 2962
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5385898843337479,
+      "learning_rate": 3.5894624621723997e-10,
+      "loss": 1.0093,
+      "step": 2963
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5306539384405943,
+      "learning_rate": 2.9665040627901985e-10,
+      "loss": 0.7728,
+      "step": 2964
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.5950054220926712,
+      "learning_rate": 2.4028728053959725e-10,
+      "loss": 0.7793,
+      "step": 2965
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7628046325406455,
+      "learning_rate": 1.8985693587880715e-10,
+      "loss": 0.8275,
+      "step": 2966
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.44354496341099486,
+      "learning_rate": 1.453594321393359e-10,
+      "loss": 1.0025,
+      "step": 2967
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.47729181296882567,
+      "learning_rate": 1.067948221217252e-10,
+      "loss": 0.6729,
+      "step": 2968
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.4697258353492894,
+      "learning_rate": 7.416315158714771e-11,
+      "loss": 0.6968,
+      "step": 2969
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.600238324926368,
+      "learning_rate": 4.746445925740695e-11,
+      "loss": 0.9411,
+      "step": 2970
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.31791560116410567,
+      "learning_rate": 2.669877681271693e-11,
+      "loss": 0.8864,
+      "step": 2971
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.9963132133952572,
+      "learning_rate": 1.1866128895032803e-11,
+      "loss": 0.7059,
+      "step": 2972
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.7430222236552121,
+      "learning_rate": 2.9665331036099475e-12,
+      "loss": 0.9157,
+      "step": 2973
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 0.6051055822848231,
+      "learning_rate": 0.0,
+      "loss": 0.5361,
+      "step": 2974
+    },
+    {
+      "epoch": 1.0,
+      "step": 2974,
+      "total_flos": 0.0,
+      "train_loss": 0.08920118028314304,
+      "train_runtime": 13189.0335,
+      "train_samples_per_second": 462.409,
+      "train_steps_per_second": 0.225
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 2974,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 50,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": false,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 0.0,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}