{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.0,
  "eval_steps": 500,
  "global_step": 774,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.007751937984496124,
      "grad_norm": 2.372762680053711,
      "learning_rate": 1.935483870967742e-07,
      "loss": 4.068140983581543,
      "step": 2
    },
    {
      "epoch": 0.015503875968992248,
      "grad_norm": 0.1989186704158783,
      "learning_rate": 5.806451612903226e-07,
      "loss": 2.024549722671509,
      "step": 4
    },
    {
      "epoch": 0.023255813953488372,
      "grad_norm": 0.287252277135849,
      "learning_rate": 9.67741935483871e-07,
      "loss": 1.9417322874069214,
      "step": 6
    },
    {
      "epoch": 0.031007751937984496,
      "grad_norm": 0.3818705379962921,
      "learning_rate": 1.3548387096774193e-06,
      "loss": 1.9495620727539062,
      "step": 8
    },
    {
      "epoch": 0.03875968992248062,
      "grad_norm": 0.15045583248138428,
      "learning_rate": 1.7419354838709678e-06,
      "loss": 2.319340705871582,
      "step": 10
    },
    {
      "epoch": 0.046511627906976744,
      "grad_norm": 2.149204969406128,
      "learning_rate": 2.129032258064516e-06,
      "loss": 3.75003981590271,
      "step": 12
    },
    {
      "epoch": 0.05426356589147287,
      "grad_norm": 0.2591269314289093,
      "learning_rate": 2.516129032258065e-06,
      "loss": 1.8552712202072144,
      "step": 14
    },
    {
      "epoch": 0.06201550387596899,
      "grad_norm": 0.7406859397888184,
      "learning_rate": 2.9032258064516128e-06,
      "loss": 2.239339590072632,
      "step": 16
    },
    {
      "epoch": 0.06976744186046512,
      "grad_norm": 0.9925192594528198,
      "learning_rate": 3.2903225806451615e-06,
      "loss": 2.0903704166412354,
      "step": 18
    },
    {
      "epoch": 0.07751937984496124,
      "grad_norm": 0.29998043179512024,
      "learning_rate": 3.6774193548387094e-06,
      "loss": 2.208868980407715,
      "step": 20
    },
    {
      "epoch": 0.08527131782945736,
      "grad_norm": 2.925621747970581,
      "learning_rate": 4.064516129032259e-06,
      "loss": 3.090871810913086,
      "step": 22
    },
    {
      "epoch": 0.09302325581395349,
      "grad_norm": 0.2913108766078949,
      "learning_rate": 4.451612903225806e-06,
      "loss": 1.8751976490020752,
      "step": 24
    },
    {
      "epoch": 0.10077519379844961,
      "grad_norm": 0.645453929901123,
      "learning_rate": 4.838709677419355e-06,
      "loss": 1.761456847190857,
      "step": 26
    },
    {
      "epoch": 0.10852713178294573,
      "grad_norm": 0.7114430665969849,
      "learning_rate": 5.2258064516129035e-06,
      "loss": 1.8018805980682373,
      "step": 28
    },
    {
      "epoch": 0.11627906976744186,
      "grad_norm": 0.7619907259941101,
      "learning_rate": 5.612903225806452e-06,
      "loss": 1.5239174365997314,
      "step": 30
    },
    {
      "epoch": 0.12403100775193798,
      "grad_norm": 0.16530756652355194,
      "learning_rate": 6e-06,
      "loss": 1.5915422439575195,
      "step": 32
    },
    {
      "epoch": 0.13178294573643412,
      "grad_norm": 0.16290850937366486,
      "learning_rate": 6.3870967741935485e-06,
      "loss": 2.3932559490203857,
      "step": 34
    },
    {
      "epoch": 0.13953488372093023,
      "grad_norm": 0.23374152183532715,
      "learning_rate": 6.774193548387097e-06,
      "loss": 1.4996192455291748,
      "step": 36
    },
    {
      "epoch": 0.14728682170542637,
      "grad_norm": 0.26606833934783936,
      "learning_rate": 7.161290322580645e-06,
      "loss": 1.2402312755584717,
      "step": 38
    },
    {
      "epoch": 0.15503875968992248,
      "grad_norm": 0.12572336196899414,
      "learning_rate": 7.548387096774193e-06,
      "loss": 1.5890406370162964,
      "step": 40
    },
    {
      "epoch": 0.16279069767441862,
      "grad_norm": 0.7024409770965576,
      "learning_rate": 7.935483870967743e-06,
      "loss": 1.1217167377471924,
      "step": 42
    },
    {
      "epoch": 0.17054263565891473,
      "grad_norm": 0.1792139858007431,
      "learning_rate": 8.32258064516129e-06,
      "loss": 1.3888567686080933,
      "step": 44
    },
    {
      "epoch": 0.17829457364341086,
      "grad_norm": 0.16807198524475098,
      "learning_rate": 8.70967741935484e-06,
      "loss": 1.746845006942749,
      "step": 46
    },
    {
      "epoch": 0.18604651162790697,
      "grad_norm": 0.1879073679447174,
      "learning_rate": 9.096774193548387e-06,
      "loss": 0.9335119128227234,
      "step": 48
    },
    {
      "epoch": 0.1937984496124031,
      "grad_norm": 0.2711307406425476,
      "learning_rate": 9.483870967741934e-06,
      "loss": 0.6987147331237793,
      "step": 50
    },
    {
      "epoch": 0.20155038759689922,
      "grad_norm": 0.1467754989862442,
      "learning_rate": 9.870967741935485e-06,
      "loss": 1.1381912231445312,
      "step": 52
    },
    {
      "epoch": 0.20930232558139536,
      "grad_norm": 0.5655927062034607,
      "learning_rate": 1.0258064516129032e-05,
      "loss": 0.581324577331543,
      "step": 54
    },
    {
      "epoch": 0.21705426356589147,
      "grad_norm": 0.503146767616272,
      "learning_rate": 1.0645161290322582e-05,
      "loss": 1.3841195106506348,
      "step": 56
    },
    {
      "epoch": 0.2248062015503876,
      "grad_norm": 0.7577893733978271,
      "learning_rate": 1.1032258064516129e-05,
      "loss": 1.4041730165481567,
      "step": 58
    },
    {
      "epoch": 0.23255813953488372,
      "grad_norm": 0.7149530649185181,
      "learning_rate": 1.1419354838709677e-05,
      "loss": 1.1680717468261719,
      "step": 60
    },
    {
      "epoch": 0.24031007751937986,
      "grad_norm": 0.14106588065624237,
      "learning_rate": 1.1806451612903226e-05,
      "loss": 1.2515933513641357,
      "step": 62
    },
    {
      "epoch": 0.24806201550387597,
      "grad_norm": 0.19908978044986725,
      "learning_rate": 1.2193548387096773e-05,
      "loss": 1.310453176498413,
      "step": 64
    },
    {
      "epoch": 0.2558139534883721,
      "grad_norm": 0.2537965476512909,
      "learning_rate": 1.2580645161290324e-05,
      "loss": 1.2055346965789795,
      "step": 66
    },
    {
      "epoch": 0.26356589147286824,
      "grad_norm": 0.14298808574676514,
      "learning_rate": 1.2967741935483872e-05,
      "loss": 1.178272008895874,
      "step": 68
    },
    {
      "epoch": 0.2713178294573643,
      "grad_norm": 0.12258580327033997,
      "learning_rate": 1.335483870967742e-05,
      "loss": 1.33051335811615,
      "step": 70
    },
    {
      "epoch": 0.27906976744186046,
      "grad_norm": 0.1763555109500885,
      "learning_rate": 1.3741935483870968e-05,
      "loss": 1.0556892156600952,
      "step": 72
    },
    {
      "epoch": 0.2868217054263566,
      "grad_norm": 0.28760379552841187,
      "learning_rate": 1.4129032258064516e-05,
      "loss": 1.1506415605545044,
      "step": 74
    },
    {
      "epoch": 0.29457364341085274,
      "grad_norm": 0.3192862868309021,
      "learning_rate": 1.4516129032258065e-05,
      "loss": 1.1051461696624756,
      "step": 76
    },
    {
      "epoch": 0.3023255813953488,
      "grad_norm": 0.17246203124523163,
      "learning_rate": 1.4903225806451612e-05,
      "loss": 1.3540679216384888,
      "step": 78
    },
    {
      "epoch": 0.31007751937984496,
      "grad_norm": 0.07415715605020523,
      "learning_rate": 1.529032258064516e-05,
      "loss": 1.1473115682601929,
      "step": 80
    },
    {
      "epoch": 0.3178294573643411,
      "grad_norm": 0.16899384558200836,
      "learning_rate": 1.567741935483871e-05,
      "loss": 1.3307069540023804,
      "step": 82
    },
    {
      "epoch": 0.32558139534883723,
      "grad_norm": 0.16342990100383759,
      "learning_rate": 1.6064516129032258e-05,
      "loss": 0.9781491756439209,
      "step": 84
    },
    {
      "epoch": 0.3333333333333333,
      "grad_norm": 0.12549830973148346,
      "learning_rate": 1.6451612903225807e-05,
      "loss": 1.1001088619232178,
      "step": 86
    },
    {
      "epoch": 0.34108527131782945,
      "grad_norm": 0.13267701864242554,
      "learning_rate": 1.6838709677419356e-05,
      "loss": 1.399463415145874,
      "step": 88
    },
    {
      "epoch": 0.3488372093023256,
      "grad_norm": 0.4117628037929535,
      "learning_rate": 1.7225806451612906e-05,
      "loss": 0.8339037299156189,
      "step": 90
    },
    {
      "epoch": 0.35658914728682173,
      "grad_norm": 0.3214583992958069,
      "learning_rate": 1.761290322580645e-05,
      "loss": 1.1209487915039062,
      "step": 92
    },
    {
      "epoch": 0.3643410852713178,
      "grad_norm": 0.1995217651128769,
      "learning_rate": 1.8e-05,
      "loss": 1.0714277029037476,
      "step": 94
    },
    {
      "epoch": 0.37209302325581395,
      "grad_norm": 0.0989437848329544,
      "learning_rate": 1.838709677419355e-05,
      "loss": 1.192923903465271,
      "step": 96
    },
    {
      "epoch": 0.3798449612403101,
      "grad_norm": 0.1220928430557251,
      "learning_rate": 1.87741935483871e-05,
      "loss": 0.7750011086463928,
      "step": 98
    },
    {
      "epoch": 0.3875968992248062,
      "grad_norm": 0.13039161264896393,
      "learning_rate": 1.9161290322580645e-05,
      "loss": 0.6638532280921936,
      "step": 100
    },
    {
      "epoch": 0.3953488372093023,
      "grad_norm": 0.10157541930675507,
      "learning_rate": 1.9548387096774194e-05,
      "loss": 1.3122025728225708,
      "step": 102
    },
    {
      "epoch": 0.40310077519379844,
      "grad_norm": 0.08540091663599014,
      "learning_rate": 1.9935483870967743e-05,
      "loss": 1.203364372253418,
      "step": 104
    },
    {
      "epoch": 0.4108527131782946,
      "grad_norm": 0.1469828337430954,
      "learning_rate": 2.032258064516129e-05,
      "loss": 1.0173336267471313,
      "step": 106
    },
    {
      "epoch": 0.4186046511627907,
      "grad_norm": 0.09850579500198364,
      "learning_rate": 2.0709677419354838e-05,
      "loss": 1.0237982273101807,
      "step": 108
    },
    {
      "epoch": 0.4263565891472868,
      "grad_norm": 0.08415781706571579,
      "learning_rate": 2.1096774193548387e-05,
      "loss": 1.0086685419082642,
      "step": 110
    },
    {
      "epoch": 0.43410852713178294,
      "grad_norm": 0.08691313862800598,
      "learning_rate": 2.1483870967741936e-05,
      "loss": 0.9078419208526611,
      "step": 112
    },
    {
      "epoch": 0.4418604651162791,
      "grad_norm": 0.1410425454378128,
      "learning_rate": 2.1870967741935485e-05,
      "loss": 1.0696989297866821,
      "step": 114
    },
    {
      "epoch": 0.4496124031007752,
      "grad_norm": 0.11058277636766434,
      "learning_rate": 2.2258064516129034e-05,
      "loss": 1.0235304832458496,
      "step": 116
    },
    {
      "epoch": 0.4573643410852713,
      "grad_norm": 0.17280496656894684,
      "learning_rate": 2.2645161290322584e-05,
      "loss": 0.8430470824241638,
      "step": 118
    },
    {
      "epoch": 0.46511627906976744,
      "grad_norm": 0.0914328321814537,
      "learning_rate": 2.303225806451613e-05,
      "loss": 1.2717251777648926,
      "step": 120
    },
    {
      "epoch": 0.4728682170542636,
      "grad_norm": 0.10640335828065872,
      "learning_rate": 2.341935483870968e-05,
      "loss": 1.0239697694778442,
      "step": 122
    },
    {
      "epoch": 0.4806201550387597,
      "grad_norm": 0.10921759903430939,
      "learning_rate": 2.3806451612903228e-05,
      "loss": 1.2588107585906982,
      "step": 124
    },
    {
      "epoch": 0.4883720930232558,
      "grad_norm": 0.21347090601921082,
      "learning_rate": 2.4193548387096773e-05,
      "loss": 1.149160385131836,
      "step": 126
    },
    {
      "epoch": 0.49612403100775193,
      "grad_norm": 0.16161420941352844,
      "learning_rate": 2.4580645161290323e-05,
      "loss": 1.1978895664215088,
      "step": 128
    },
    {
      "epoch": 0.5038759689922481,
      "grad_norm": 0.11207153648138046,
      "learning_rate": 2.4967741935483872e-05,
      "loss": 1.1477546691894531,
      "step": 130
    },
    {
      "epoch": 0.5116279069767442,
      "grad_norm": 0.13203421235084534,
      "learning_rate": 2.535483870967742e-05,
      "loss": 0.684913694858551,
      "step": 132
    },
    {
      "epoch": 0.5193798449612403,
      "grad_norm": 0.16211755573749542,
      "learning_rate": 2.5741935483870967e-05,
      "loss": 1.135626196861267,
      "step": 134
    },
    {
      "epoch": 0.5271317829457365,
      "grad_norm": 0.19277559220790863,
      "learning_rate": 2.6129032258064516e-05,
      "loss": 1.1265920400619507,
      "step": 136
    },
    {
      "epoch": 0.5348837209302325,
      "grad_norm": 0.12664702534675598,
      "learning_rate": 2.6516129032258065e-05,
      "loss": 1.3438581228256226,
      "step": 138
    },
    {
      "epoch": 0.5426356589147286,
      "grad_norm": 0.07170073688030243,
      "learning_rate": 2.690322580645161e-05,
      "loss": 1.2673969268798828,
      "step": 140
    },
    {
      "epoch": 0.5503875968992248,
      "grad_norm": 0.1509476602077484,
      "learning_rate": 2.7290322580645163e-05,
      "loss": 0.6191468834877014,
      "step": 142
    },
    {
      "epoch": 0.5581395348837209,
      "grad_norm": 0.08468706905841827,
      "learning_rate": 2.7677419354838713e-05,
      "loss": 1.3681960105895996,
      "step": 144
    },
    {
      "epoch": 0.5658914728682171,
      "grad_norm": 0.08381734788417816,
      "learning_rate": 2.806451612903226e-05,
      "loss": 1.0066568851470947,
      "step": 146
    },
    {
      "epoch": 0.5736434108527132,
      "grad_norm": 0.0786685049533844,
      "learning_rate": 2.8451612903225808e-05,
      "loss": 0.993072509765625,
      "step": 148
    },
    {
      "epoch": 0.5813953488372093,
      "grad_norm": 0.16647084057331085,
      "learning_rate": 2.8838709677419357e-05,
      "loss": 0.9888545274734497,
      "step": 150
    },
    {
      "epoch": 0.5891472868217055,
      "grad_norm": 0.0718846470117569,
      "learning_rate": 2.9225806451612906e-05,
      "loss": 1.2481988668441772,
      "step": 152
    },
    {
      "epoch": 0.5968992248062015,
      "grad_norm": 0.054430440068244934,
      "learning_rate": 2.961290322580645e-05,
      "loss": 1.2282058000564575,
      "step": 154
    },
    {
      "epoch": 0.6046511627906976,
      "grad_norm": 0.08397103101015091,
      "learning_rate": 3e-05,
      "loss": 0.7997589111328125,
      "step": 156
    },
    {
      "epoch": 0.6124031007751938,
      "grad_norm": 0.16056904196739197,
      "learning_rate": 2.9994088909931925e-05,
      "loss": 0.9445045590400696,
      "step": 158
    },
    {
      "epoch": 0.6201550387596899,
      "grad_norm": 0.0699552372097969,
      "learning_rate": 2.9976361120666653e-05,
      "loss": 1.3003367185592651,
      "step": 160
    },
    {
      "epoch": 0.627906976744186,
      "grad_norm": 0.326346755027771,
      "learning_rate": 2.9946833069938937e-05,
      "loss": 1.0931003093719482,
      "step": 162
    },
    {
      "epoch": 0.6356589147286822,
      "grad_norm": 0.05613398924469948,
      "learning_rate": 2.990553213703777e-05,
      "loss": 1.252951741218567,
      "step": 164
    },
    {
      "epoch": 0.6434108527131783,
      "grad_norm": 0.06693823635578156,
      "learning_rate": 2.985249661741952e-05,
      "loss": 0.9843292236328125,
      "step": 166
    },
    {
      "epoch": 0.6511627906976745,
      "grad_norm": 0.05880317464470863,
      "learning_rate": 2.9787775687199204e-05,
      "loss": 1.2224986553192139,
      "step": 168
    },
    {
      "epoch": 0.6589147286821705,
      "grad_norm": 0.22196456789970398,
      "learning_rate": 2.9711429357552954e-05,
      "loss": 0.6978975534439087,
      "step": 170
    },
    {
      "epoch": 0.6666666666666666,
      "grad_norm": 0.11266355961561203,
      "learning_rate": 2.9623528419073845e-05,
      "loss": 1.1819331645965576,
      "step": 172
    },
    {
      "epoch": 0.6744186046511628,
      "grad_norm": 0.23993322253227234,
      "learning_rate": 2.952415437613278e-05,
      "loss": 0.9013431668281555,
      "step": 174
    },
    {
      "epoch": 0.6821705426356589,
      "grad_norm": 0.16824588179588318,
      "learning_rate": 2.9413399371305162e-05,
      "loss": 0.9518213272094727,
      "step": 176
    },
    {
      "epoch": 0.689922480620155,
      "grad_norm": 0.10848796367645264,
      "learning_rate": 2.9291366099933583e-05,
      "loss": 0.9333250522613525,
      "step": 178
    },
    {
      "epoch": 0.6976744186046512,
      "grad_norm": 0.07721789926290512,
      "learning_rate": 2.9158167714905638e-05,
      "loss": 1.0462111234664917,
      "step": 180
    },
    {
      "epoch": 0.7054263565891473,
      "grad_norm": 0.07368086278438568,
      "learning_rate": 2.9013927721735172e-05,
      "loss": 1.1160787343978882,
      "step": 182
    },
    {
      "epoch": 0.7131782945736435,
      "grad_norm": 0.12432865053415298,
      "learning_rate": 2.885877986404432e-05,
      "loss": 0.9108925461769104,
      "step": 184
    },
    {
      "epoch": 0.7209302325581395,
      "grad_norm": 0.0682280957698822,
      "learning_rate": 2.86928679995524e-05,
      "loss": 1.3199962377548218,
      "step": 186
    },
    {
      "epoch": 0.7286821705426356,
      "grad_norm": 0.0608358271420002,
      "learning_rate": 2.8516345966686766e-05,
      "loss": 1.0051369667053223,
      "step": 188
    },
    {
      "epoch": 0.7364341085271318,
      "grad_norm": 0.16042953729629517,
      "learning_rate": 2.832937744193922e-05,
      "loss": 1.0462324619293213,
      "step": 190
    },
    {
      "epoch": 0.7441860465116279,
      "grad_norm": 0.09749440848827362,
      "learning_rate": 2.8132135788100312e-05,
      "loss": 1.1752909421920776,
      "step": 192
    },
    {
      "epoch": 0.751937984496124,
      "grad_norm": 0.057570479810237885,
      "learning_rate": 2.7924803893512144e-05,
      "loss": 1.2336088418960571,
      "step": 194
    },
    {
      "epoch": 0.7596899224806202,
      "grad_norm": 0.05938956141471863,
      "learning_rate": 2.770757400248891e-05,
      "loss": 1.2357627153396606,
      "step": 196
    },
    {
      "epoch": 0.7674418604651163,
      "grad_norm": 0.1574079990386963,
      "learning_rate": 2.7480647537062183e-05,
      "loss": 0.6303899884223938,
      "step": 198
    },
    {
      "epoch": 0.7751937984496124,
      "grad_norm": 0.7175618410110474,
      "learning_rate": 2.7244234910216427e-05,
      "loss": 1.449771761894226,
      "step": 200
    },
    {
      "epoch": 0.7829457364341085,
      "grad_norm": 0.12645360827445984,
      "learning_rate": 2.699855533078781e-05,
      "loss": 1.136543869972229,
      "step": 202
    },
    {
      "epoch": 0.7906976744186046,
      "grad_norm": 0.1046365350484848,
      "learning_rate": 2.6743836600207206e-05,
      "loss": 1.2583595514297485,
      "step": 204
    },
    {
      "epoch": 0.7984496124031008,
      "grad_norm": 0.06069144234061241,
      "learning_rate": 2.6480314901275954e-05,
      "loss": 1.073771595954895,
      "step": 206
    },
    {
      "epoch": 0.8062015503875969,
      "grad_norm": 0.08813668042421341,
      "learning_rate": 2.620823457917009e-05,
      "loss": 0.8364643454551697,
      "step": 208
    },
    {
      "epoch": 0.813953488372093,
      "grad_norm": 0.1366778016090393,
      "learning_rate": 2.592784791487625e-05,
      "loss": 1.355994701385498,
      "step": 210
    },
    {
      "epoch": 0.8217054263565892,
      "grad_norm": 0.06360755860805511,
      "learning_rate": 2.5639414891269164e-05,
      "loss": 1.226666808128357,
      "step": 212
    },
    {
      "epoch": 0.8294573643410853,
      "grad_norm": 0.05345217138528824,
      "learning_rate": 2.534320295204785e-05,
      "loss": 0.9844304919242859,
      "step": 214
    },
    {
      "epoch": 0.8372093023255814,
      "grad_norm": 0.13714972138404846,
      "learning_rate": 2.5039486753753788e-05,
      "loss": 1.0947370529174805,
      "step": 216
    },
    {
      "epoch": 0.8449612403100775,
      "grad_norm": 0.08509152382612228,
      "learning_rate": 2.4728547911101212e-05,
      "loss": 1.341536283493042,
      "step": 218
    },
    {
      "epoch": 0.8527131782945736,
      "grad_norm": 0.06335103511810303,
      "learning_rate": 2.4410674735855555e-05,
      "loss": 1.2149854898452759,
      "step": 220
    },
    {
      "epoch": 0.8604651162790697,
      "grad_norm": 0.07615257799625397,
      "learning_rate": 2.4086161969502158e-05,
      "loss": 0.5183571577072144,
      "step": 222
    },
    {
      "epoch": 0.8682170542635659,
      "grad_norm": 0.05618668347597122,
      "learning_rate": 2.3755310509953208e-05,
      "loss": 1.2107856273651123,
      "step": 224
    },
    {
      "epoch": 0.875968992248062,
      "grad_norm": 0.08672861009836197,
      "learning_rate": 2.34184271325462e-05,
      "loss": 1.2980958223342896,
      "step": 226
    },
    {
      "epoch": 0.8837209302325582,
      "grad_norm": 0.2590503692626953,
      "learning_rate": 2.3075824205592707e-05,
      "loss": 0.9205237030982971,
      "step": 228
    },
    {
      "epoch": 0.8914728682170543,
      "grad_norm": 0.08758358657360077,
      "learning_rate": 2.2727819400741172e-05,
      "loss": 1.058215618133545,
      "step": 230
    },
    {
      "epoch": 0.8992248062015504,
      "grad_norm": 0.07180935889482498,
      "learning_rate": 2.2374735398422276e-05,
      "loss": 0.8909934759140015,
      "step": 232
    },
    {
      "epoch": 0.9069767441860465,
      "grad_norm": 0.07425983250141144,
      "learning_rate": 2.2016899588650032e-05,
      "loss": 1.3261955976486206,
      "step": 234
    },
    {
      "epoch": 0.9147286821705426,
      "grad_norm": 0.10620886832475662,
      "learning_rate": 2.165464376745598e-05,
      "loss": 0.7832080125808716,
      "step": 236
    },
    {
      "epoch": 0.9224806201550387,
      "grad_norm": 0.11008700728416443,
      "learning_rate": 2.1288303829238058e-05,
      "loss": 1.243245244026184,
      "step": 238
    },
    {
      "epoch": 0.9302325581395349,
      "grad_norm": 0.15069428086280823,
      "learning_rate": 2.091821945530925e-05,
      "loss": 1.0843262672424316,
      "step": 240
    },
    {
      "epoch": 0.937984496124031,
      "grad_norm": 0.20396780967712402,
      "learning_rate": 2.0544733798934988e-05,
      "loss": 1.2043696641921997,
      "step": 242
    },
    {
      "epoch": 0.9457364341085271,
      "grad_norm": 0.1518591046333313,
      "learning_rate": 2.0168193167151183e-05,
      "loss": 1.0192402601242065,
      "step": 244
    },
    {
      "epoch": 0.9534883720930233,
      "grad_norm": 0.08409406244754791,
      "learning_rate": 1.9788946699658032e-05,
      "loss": 0.7911899089813232,
      "step": 246
    },
    {
      "epoch": 0.9612403100775194,
      "grad_norm": 0.1363200843334198,
      "learning_rate": 1.9407346045087278e-05,
      "loss": 0.7742171883583069,
      "step": 248
    },
    {
      "epoch": 0.9689922480620154,
      "grad_norm": 0.07617780566215515,
      "learning_rate": 1.902374503494311e-05,
      "loss": 1.2457553148269653,
      "step": 250
    },
    {
      "epoch": 0.9767441860465116,
      "grad_norm": 0.0721699595451355,
      "learning_rate": 1.863849935551905e-05,
      "loss": 1.016923427581787,
      "step": 252
    },
    {
      "epoch": 0.9844961240310077,
      "grad_norm": 0.14891768991947174,
      "learning_rate": 1.825196621809499e-05,
      "loss": 0.9890077710151672,
      "step": 254
    },
    {
      "epoch": 0.9922480620155039,
      "grad_norm": 0.28671520948410034,
      "learning_rate": 1.7864504027720297e-05,
      "loss": 0.8091475367546082,
      "step": 256
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.04788220301270485,
      "learning_rate": 1.747647205088991e-05,
      "loss": 1.1878516674041748,
      "step": 258
    },
    {
      "epoch": 1.0077519379844961,
      "grad_norm": 0.06899169087409973,
      "learning_rate": 1.7088230082421763e-05,
      "loss": 0.7679018378257751,
      "step": 260
    },
    {
      "epoch": 1.0155038759689923,
      "grad_norm": 0.0531589537858963,
      "learning_rate": 1.67001381118443e-05,
      "loss": 1.128183126449585,
      "step": 262
    },
    {
      "epoch": 1.0232558139534884,
      "grad_norm": 0.36434194445610046,
      "learning_rate": 1.6312555989603446e-05,
      "loss": 0.7761940956115723,
      "step": 264
    },
    {
      "epoch": 1.0310077519379846,
      "grad_norm": 0.09103479981422424,
      "learning_rate": 1.5925843093398552e-05,
      "loss": 0.6045004725456238,
      "step": 266
    },
    {
      "epoch": 1.0387596899224807,
      "grad_norm": 0.13370856642723083,
      "learning_rate": 1.554035799495667e-05,
      "loss": 0.8422982096672058,
      "step": 268
    },
    {
      "epoch": 1.0465116279069768,
      "grad_norm": 0.0862463042140007,
      "learning_rate": 1.515645812755415e-05,
      "loss": 1.0801072120666504,
      "step": 270
    },
    {
      "epoch": 1.054263565891473,
      "grad_norm": 0.08025830239057541,
      "learning_rate": 1.4774499454593878e-05,
      "loss": 0.5319198966026306,
      "step": 272
    },
    {
      "epoch": 1.062015503875969,
      "grad_norm": 0.12001097202301025,
      "learning_rate": 1.4394836139545363e-05,
      "loss": 1.100441575050354,
      "step": 274
    },
    {
      "epoch": 1.069767441860465,
      "grad_norm": 0.0670214369893074,
      "learning_rate": 1.4017820217553832e-05,
      "loss": 0.9006233215332031,
      "step": 276
    },
    {
      "epoch": 1.0775193798449612,
      "grad_norm": 0.05677079036831856,
      "learning_rate": 1.3643801269022732e-05,
      "loss": 0.9063330292701721,
      "step": 278
    },
    {
      "epoch": 1.0852713178294573,
      "grad_norm": 0.06150190904736519,
      "learning_rate": 1.3273126095472361e-05,
      "loss": 0.6792492270469666,
      "step": 280
    },
    {
      "epoch": 1.0930232558139534,
      "grad_norm": 0.11940456926822662,
      "learning_rate": 1.2906138397975178e-05,
      "loss": 0.5880073308944702,
      "step": 282
    },
    {
      "epoch": 1.1007751937984496,
      "grad_norm": 0.059427861124277115,
      "learning_rate": 1.2543178458465887e-05,
      "loss": 0.6459496021270752,
      "step": 284
    },
    {
      "epoch": 1.1085271317829457,
      "grad_norm": 0.14358076453208923,
      "learning_rate": 1.2184582824221902e-05,
      "loss": 0.6029883027076721,
      "step": 286
    },
    {
      "epoch": 1.1162790697674418,
      "grad_norm": 0.12145813554525375,
      "learning_rate": 1.1830683995806694e-05,
      "loss": 0.5689815282821655,
      "step": 288
    },
    {
      "epoch": 1.124031007751938,
      "grad_norm": 0.06693974882364273,
      "learning_rate": 1.14818101187653e-05,
      "loss": 0.9926015138626099,
      "step": 290
    },
    {
      "epoch": 1.1317829457364341,
      "grad_norm": 0.05298677831888199,
      "learning_rate": 1.113828467935807e-05,
      "loss": 0.599225640296936,
      "step": 292
    },
    {
      "epoch": 1.1395348837209303,
      "grad_norm": 0.1428927630186081,
      "learning_rate": 1.080042620461448e-05,
      "loss": 1.0777664184570312,
      "step": 294
    },
    {
      "epoch": 1.1472868217054264,
      "grad_norm": 0.33007150888442993,
      "learning_rate": 1.0468547966985433e-05,
      "loss": 0.592754065990448,
      "step": 296
    },
    {
      "epoch": 1.1550387596899225,
      "grad_norm": 0.09632085263729095,
      "learning_rate": 1.0142957693867676e-05,
      "loss": 0.8235950469970703,
      "step": 298
    },
    {
      "epoch": 1.1627906976744187,
      "grad_norm": 0.05365417152643204,
      "learning_rate": 9.823957282269788e-06,
      "loss": 0.6445293426513672,
      "step": 300
    },
    {
      "epoch": 1.1705426356589148,
      "grad_norm": 0.05418482422828674,
      "learning_rate": 9.51184251888427e-06,
      "loss": 0.8274694085121155,
      "step": 302
    },
    {
      "epoch": 1.178294573643411,
      "grad_norm": 0.08142586797475815,
      "learning_rate": 9.206902805825313e-06,
      "loss": 0.6243199110031128,
      "step": 304
    },
    {
      "epoch": 1.1860465116279069,
      "grad_norm": 0.09854786843061447,
      "learning_rate": 8.909420892286522e-06,
      "loss": 0.9028869867324829,
      "step": 306
    },
    {
      "epoch": 1.193798449612403,
      "grad_norm": 0.12125203758478165,
      "learning_rate": 8.619672612367426e-06,
      "loss": 1.018742322921753,
      "step": 308
    },
    {
      "epoch": 1.2015503875968991,
      "grad_norm": 0.05682411789894104,
      "learning_rate": 8.337926629311901e-06,
      "loss": 0.8793404698371887,
      "step": 310
    },
    {
      "epoch": 1.2093023255813953,
      "grad_norm": 0.08018923550844193,
      "learning_rate": 8.064444186395577e-06,
      "loss": 0.5548562407493591,
      "step": 312
    },
    {
      "epoch": 1.2170542635658914,
      "grad_norm": 0.4036023020744324,
      "learning_rate": 7.799478864693305e-06,
      "loss": 0.40799465775489807,
      "step": 314
    },
    {
      "epoch": 1.2248062015503876,
      "grad_norm": 0.09002506732940674,
      "learning_rate": 7.543276347951186e-06,
      "loss": 0.854184627532959,
      "step": 316
    },
    {
      "epoch": 1.2325581395348837,
      "grad_norm": 0.05118729546666145,
      "learning_rate": 7.2960741947813325e-06,
      "loss": 0.7614866495132446,
      "step": 318
    },
    {
      "epoch": 1.2403100775193798,
      "grad_norm": 0.048252079635858536,
      "learning_rate": 7.058101618390395e-06,
      "loss": 0.9597308039665222,
      "step": 320
    },
    {
      "epoch": 1.248062015503876,
      "grad_norm": 0.08246824890375137,
      "learning_rate": 6.829579274046253e-06,
      "loss": 1.0625540018081665,
      "step": 322
    },
    {
      "epoch": 1.255813953488372,
      "grad_norm": 0.10916761308908463,
      "learning_rate": 6.61071905447989e-06,
      "loss": 0.5724548101425171,
      "step": 324
    },
    {
      "epoch": 1.2635658914728682,
      "grad_norm": 0.0720956102013588,
      "learning_rate": 6.40172389341212e-06,
      "loss": 0.9105185866355896,
      "step": 326
    },
    {
      "epoch": 1.2713178294573644,
      "grad_norm": 0.1327294260263443,
      "learning_rate": 6.202787577387384e-06,
      "loss": 0.6492634415626526,
      "step": 328
    },
    {
      "epoch": 1.2790697674418605,
      "grad_norm": 0.055963922291994095,
      "learning_rate": 6.014094566089114e-06,
      "loss": 0.6872971653938293,
      "step": 330
    },
    {
      "epoch": 1.2868217054263567,
      "grad_norm": 0.055926598608493805,
      "learning_rate": 5.835819821303189e-06,
      "loss": 0.8940346240997314,
      "step": 332
    },
    {
      "epoch": 1.2945736434108528,
      "grad_norm": 0.11941412091255188,
      "learning_rate": 5.6681286446881695e-06,
      "loss": 0.517791748046875,
      "step": 334
    },
    {
      "epoch": 1.302325581395349,
      "grad_norm": 0.05569684877991676,
      "learning_rate": 5.511176524502653e-06,
      "loss": 0.9667401313781738,
      "step": 336
    },
    {
      "epoch": 1.310077519379845,
      "grad_norm": 0.04035505652427673,
      "learning_rate": 5.36510899143194e-06,
      "loss": 0.6123257875442505,
      "step": 338
    },
    {
      "epoch": 1.3178294573643412,
      "grad_norm": 0.053604647517204285,
      "learning_rate": 5.2300614836476425e-06,
      "loss": 0.8501400947570801,
      "step": 340
    },
    {
      "epoch": 1.3255813953488373,
      "grad_norm": 0.08587184548377991,
      "learning_rate": 5.106159221225361e-06,
      "loss": 0.8949810266494751,
      "step": 342
    },
    {
      "epoch": 1.3333333333333333,
      "grad_norm": 0.08081138134002686,
      "learning_rate": 4.993517090036914e-06,
      "loss": 0.8400573134422302,
      "step": 344
    },
    {
      "epoch": 1.3410852713178294,
      "grad_norm": 0.5886430740356445,
      "learning_rate": 4.89223953522472e-06,
      "loss": 0.24276283383369446,
      "step": 346
    },
    {
      "epoch": 1.3488372093023255,
      "grad_norm": 0.09961400926113129,
      "learning_rate": 4.802420464357147e-06,
      "loss": 0.7755904793739319,
      "step": 348
    },
    {
      "epoch": 1.3565891472868217,
      "grad_norm": 0.07265698909759521,
      "learning_rate": 4.724143160354624e-06,
      "loss": 0.9579305648803711,
      "step": 350
    },
    {
      "epoch": 1.3643410852713178,
      "grad_norm": 0.057041656225919724,
      "learning_rate": 4.6574802042672275e-06,
      "loss": 0.6767364740371704,
      "step": 352
    },
    {
      "epoch": 1.372093023255814,
      "grad_norm": 0.17740975320339203,
      "learning_rate": 4.602493407975375e-06,
      "loss": 0.43329671025276184,
      "step": 354
    },
    {
      "epoch": 1.37984496124031,
      "grad_norm": 0.10877375304698944,
      "learning_rate": 4.559233756875995e-06,
      "loss": 0.9381350874900818,
      "step": 356
    },
    {
      "epoch": 1.3875968992248062,
      "grad_norm": 0.06163076311349869,
      "learning_rate": 4.5277413626073554e-06,
      "loss": 0.6927696466445923,
      "step": 358
    },
    {
      "epoch": 1.3953488372093024,
      "grad_norm": 0.05219251289963722,
      "learning_rate": 4.508045425856358e-06,
      "loss": 0.7913583517074585,
      "step": 360
    },
    {
      "epoch": 1.4031007751937985,
      "grad_norm": 0.09336534887552261,
      "learning_rate": 4.500164209282782e-06,
      "loss": 0.7886365056037903,
      "step": 362
    },
    {
      "epoch": 1.4108527131782946,
      "grad_norm": 0.05429435521364212,
      "learning_rate": 4.504105020585611e-06,
      "loss": 0.7922376394271851,
      "step": 364
    },
    {
      "epoch": 1.4186046511627908,
      "grad_norm": 0.07881563901901245,
      "learning_rate": 4.519864205727111e-06,
      "loss": 0.4149991273880005,
      "step": 366
    },
    {
      "epoch": 1.4263565891472867,
      "grad_norm": 0.050976116210222244,
      "learning_rate": 4.547427152320965e-06,
      "loss": 0.40345892310142517,
      "step": 368
    },
    {
      "epoch": 1.4341085271317828,
      "grad_norm": 0.13834504783153534,
      "learning_rate": 4.586768303181312e-06,
      "loss": 0.8296991586685181,
      "step": 370
    },
    {
      "epoch": 1.441860465116279,
      "grad_norm": 0.14809437096118927,
      "learning_rate": 4.637851180020136e-06,
      "loss": 0.7817572355270386,
      "step": 372
    },
    {
      "epoch": 1.449612403100775,
      "grad_norm": 0.04827815294265747,
      "learning_rate": 4.7006284172710145e-06,
      "loss": 0.7629826664924622,
      "step": 374
    },
    {
      "epoch": 1.4573643410852712,
      "grad_norm": 0.09472133964300156,
      "learning_rate": 4.775041806007891e-06,
      "loss": 0.6470780372619629,
      "step": 376
    },
    {
      "epoch": 1.4651162790697674,
      "grad_norm": 0.23365041613578796,
      "learning_rate": 4.861022347918125e-06,
      "loss": 0.608534038066864,
      "step": 378
    },
    {
      "epoch": 1.4728682170542635,
      "grad_norm": 0.04631495103240013,
      "learning_rate": 4.958490319279778e-06,
      "loss": 0.7042877078056335,
      "step": 380
    },
    {
      "epoch": 1.4806201550387597,
      "grad_norm": 0.10502998530864716,
      "learning_rate": 5.067355344883837e-06,
      "loss": 0.7883775234222412,
      "step": 382
    },
    {
      "epoch": 1.4883720930232558,
      "grad_norm": 0.1601502150297165,
      "learning_rate": 5.187516481832796e-06,
      "loss": 0.6707104444503784,
      "step": 384
    },
    {
      "epoch": 1.496124031007752,
      "grad_norm": 0.07128335535526276,
      "learning_rate": 5.318862313137916e-06,
      "loss": 0.5789697170257568,
      "step": 386
    },
    {
      "epoch": 1.503875968992248,
      "grad_norm": 0.051307253539562225,
      "learning_rate": 5.461271051028392e-06,
      "loss": 0.7344382405281067,
      "step": 388
    },
    {
      "epoch": 1.5116279069767442,
      "grad_norm": 0.06438810378313065,
      "learning_rate": 5.614610649876592e-06,
      "loss": 0.5874720811843872,
      "step": 390
    },
    {
      "epoch": 1.5193798449612403,
      "grad_norm": 0.12045904248952866,
      "learning_rate": 5.778738928634702e-06,
      "loss": 0.6812347769737244,
      "step": 392
    },
    {
      "epoch": 1.5271317829457365,
      "grad_norm": 0.12421064078807831,
      "learning_rate": 5.953503702669238e-06,
      "loss": 0.7551456689834595,
      "step": 394
    },
    {
      "epoch": 1.5348837209302326,
      "grad_norm": 0.09568459540605545,
      "learning_rate": 6.138742924871177e-06,
      "loss": 0.6949661374092102,
      "step": 396
    },
    {
      "epoch": 1.5426356589147288,
      "grad_norm": 0.1447361260652542,
      "learning_rate": 6.334284835910859e-06,
      "loss": 0.7189986705780029,
      "step": 398
    },
    {
      "epoch": 1.550387596899225,
      "grad_norm": 0.16750486195087433,
      "learning_rate": 6.53994812349836e-06,
      "loss": 0.9427580237388611,
      "step": 400
    },
    {
      "epoch": 1.558139534883721,
      "grad_norm": 0.171998992562294,
      "learning_rate": 6.755542090501669e-06,
      "loss": 0.5628240704536438,
      "step": 402
    },
    {
      "epoch": 1.5658914728682172,
      "grad_norm": 0.04393970966339111,
      "learning_rate": 6.98086683176673e-06,
      "loss": 0.8915067315101624,
      "step": 404
    },
    {
      "epoch": 1.5736434108527133,
      "grad_norm": 0.06884656846523285,
      "learning_rate": 7.215713419475466e-06,
      "loss": 0.8232840299606323,
      "step": 406
    },
    {
      "epoch": 1.5813953488372094,
      "grad_norm": 0.06587037444114685,
      "learning_rate": 7.459864096869907e-06,
      "loss": 0.2791080176830292,
      "step": 408
    },
    {
      "epoch": 1.5891472868217056,
      "grad_norm": 0.07907220721244812,
      "learning_rate": 7.713092480162712e-06,
      "loss": 0.7711836695671082,
      "step": 410
    },
    {
      "epoch": 1.5968992248062015,
      "grad_norm": 0.07207150012254715,
      "learning_rate": 7.975163768446994e-06,
      "loss": 0.7907366156578064,
      "step": 412
    },
    {
      "epoch": 1.6046511627906976,
      "grad_norm": 0.09880537539720535,
      "learning_rate": 8.245834961410696e-06,
      "loss": 0.49653783440589905,
      "step": 414
    },
    {
      "epoch": 1.6124031007751938,
      "grad_norm": 0.05552729591727257,
      "learning_rate": 8.524855084653766e-06,
      "loss": 0.44712162017822266,
      "step": 416
    },
    {
      "epoch": 1.62015503875969,
      "grad_norm": 0.05235432833433151,
      "learning_rate": 8.811965422399043e-06,
      "loss": 0.7449421882629395,
      "step": 418
    },
    {
      "epoch": 1.627906976744186,
      "grad_norm": 0.06001957878470421,
      "learning_rate": 9.106899757381288e-06,
      "loss": 0.7992556095123291,
      "step": 420
    },
    {
      "epoch": 1.6356589147286822,
      "grad_norm": 0.09635739773511887,
      "learning_rate": 9.409384617691782e-06,
      "loss": 0.5354418754577637,
      "step": 422
    },
    {
      "epoch": 1.6434108527131783,
      "grad_norm": 0.06418698281049728,
      "learning_rate": 9.719139530349551e-06,
      "loss": 0.8580225110054016,
      "step": 424
    },
    {
      "epoch": 1.6511627906976745,
      "grad_norm": 0.06972765177488327,
      "learning_rate": 1.0035877281364364e-05,
      "loss": 0.6384758949279785,
      "step": 426
    },
    {
      "epoch": 1.6589147286821704,
      "grad_norm": 0.08585374802350998,
      "learning_rate": 1.0359304182050086e-05,
      "loss": 0.88965904712677,
      "step": 428
    },
    {
      "epoch": 1.6666666666666665,
      "grad_norm": 0.09853579103946686,
      "learning_rate": 1.068912034134155e-05,
      "loss": 0.6674047708511353,
      "step": 430
    },
    {
      "epoch": 1.6744186046511627,
      "grad_norm": 0.09648997336626053,
      "learning_rate": 1.1025019943862524e-05,
      "loss": 0.8478899002075195,
      "step": 432
    },
    {
      "epoch": 1.6821705426356588,
      "grad_norm": 0.07512516528367996,
      "learning_rate": 1.1366691533486839e-05,
      "loss": 0.6321271061897278,
      "step": 434
    },
    {
      "epoch": 1.689922480620155,
      "grad_norm": 0.052200138568878174,
      "learning_rate": 1.171381830212979e-05,
      "loss": 0.6788074970245361,
      "step": 436
    },
    {
      "epoch": 1.697674418604651,
      "grad_norm": 0.05563069507479668,
      "learning_rate": 1.2066078383502049e-05,
      "loss": 0.6934828758239746,
      "step": 438
    },
    {
      "epoch": 1.7054263565891472,
      "grad_norm": 0.07190771400928497,
      "learning_rate": 1.242314515155367e-05,
      "loss": 0.617615282535553,
      "step": 440
    },
    {
      "epoch": 1.7131782945736433,
      "grad_norm": 0.0535132996737957,
      "learning_rate": 1.2784687523331521e-05,
      "loss": 0.7270308136940002,
      "step": 442
    },
    {
      "epoch": 1.7209302325581395,
      "grad_norm": 0.07645192742347717,
      "learning_rate": 1.3150370265969243e-05,
      "loss": 0.5209411978721619,
      "step": 444
    },
    {
      "epoch": 1.7286821705426356,
      "grad_norm": 0.05032151937484741,
      "learning_rate": 1.3519854307525166e-05,
      "loss": 0.8537179231643677,
      "step": 446
    },
    {
      "epoch": 1.7364341085271318,
      "grad_norm": 0.0833229050040245,
      "learning_rate": 1.3892797051379974e-05,
      "loss": 0.8147948980331421,
      "step": 448
    },
    {
      "epoch": 1.744186046511628,
      "grad_norm": 0.05039038509130478,
      "learning_rate": 1.4268852693902394e-05,
      "loss": 0.8937851190567017,
      "step": 450
    },
    {
      "epoch": 1.751937984496124,
      "grad_norm": 0.046379536390304565,
      "learning_rate": 1.4647672545088743e-05,
      "loss": 0.9495609402656555,
      "step": 452
    },
    {
      "epoch": 1.7596899224806202,
      "grad_norm": 0.15748457610607147,
      "learning_rate": 1.5028905351878626e-05,
      "loss": 0.6565471887588501,
      "step": 454
    },
    {
      "epoch": 1.7674418604651163,
      "grad_norm": 0.06636267155408859,
      "learning_rate": 1.5412197623847304e-05,
      "loss": 0.6486067175865173,
      "step": 456
    },
    {
      "epoch": 1.7751937984496124,
      "grad_norm": 0.2050524801015854,
      "learning_rate": 1.5797193960972498e-05,
      "loss": 0.44677096605300903,
      "step": 458
    },
    {
      "epoch": 1.7829457364341086,
      "grad_norm": 0.06217222660779953,
      "learning_rate": 1.6183537383171904e-05,
      "loss": 0.5162793397903442,
      "step": 460
    },
    {
      "epoch": 1.7906976744186047,
      "grad_norm": 0.040993690490722656,
      "learning_rate": 1.6570869661305794e-05,
      "loss": 0.6910752654075623,
      "step": 462
    },
    {
      "epoch": 1.7984496124031009,
      "grad_norm": 0.08061198890209198,
      "learning_rate": 1.6958831649337716e-05,
      "loss": 0.8616082668304443,
      "step": 464
    },
    {
      "epoch": 1.806201550387597,
      "grad_norm": 0.04601367563009262,
      "learning_rate": 1.7347063617345443e-05,
      "loss": 0.7291563153266907,
      "step": 466
    },
    {
      "epoch": 1.8139534883720931,
      "grad_norm": 0.047357648611068726,
      "learning_rate": 1.773520558507325e-05,
      "loss": 0.6221119165420532,
      "step": 468
    },
    {
      "epoch": 1.8217054263565893,
      "grad_norm": 0.051967669278383255,
      "learning_rate": 1.812289765571636e-05,
      "loss": 0.3711574971675873,
      "step": 470
    },
    {
      "epoch": 1.8294573643410854,
      "grad_norm": 0.06112854182720184,
      "learning_rate": 1.8509780349628006e-05,
      "loss": 1.0664476156234741,
      "step": 472
    },
    {
      "epoch": 1.8372093023255816,
      "grad_norm": 0.05628122389316559,
      "learning_rate": 1.889549493763964e-05,
      "loss": 0.517056405544281,
      "step": 474
    },
    {
      "epoch": 1.8449612403100775,
      "grad_norm": 0.08481237292289734,
      "learning_rate": 1.9279683773685313e-05,
      "loss": 0.6204915046691895,
      "step": 476
    },
    {
      "epoch": 1.8527131782945736,
      "grad_norm": 0.06115817651152611,
      "learning_rate": 1.9661990626421812e-05,
      "loss": 0.5971916913986206,
      "step": 478
    },
    {
      "epoch": 1.8604651162790697,
      "grad_norm": 0.05057134851813316,
      "learning_rate": 2.0042061009536956e-05,
      "loss": 0.8469266295433044,
      "step": 480
    },
    {
      "epoch": 1.8682170542635659,
      "grad_norm": 0.056389980018138885,
      "learning_rate": 2.041954251043988e-05,
      "loss": 0.5026958584785461,
      "step": 482
    },
    {
      "epoch": 1.875968992248062,
      "grad_norm": 0.24311350286006927,
      "learning_rate": 2.079408511702847e-05,
      "loss": 0.7969828248023987,
      "step": 484
    },
    {
      "epoch": 1.8837209302325582,
      "grad_norm": 0.06835257261991501,
      "learning_rate": 2.1165341542231086e-05,
      "loss": 0.8827776908874512,
      "step": 486
    },
    {
      "epoch": 1.8914728682170543,
      "grad_norm": 0.08240044862031937,
      "learning_rate": 2.15329675460214e-05,
      "loss": 0.5578112006187439,
      "step": 488
    },
    {
      "epoch": 1.8992248062015504,
      "grad_norm": 0.18629829585552216,
      "learning_rate": 2.189662225460808e-05,
      "loss": 0.6527209281921387,
      "step": 490
    },
    {
      "epoch": 1.9069767441860463,
      "grad_norm": 0.20755064487457275,
      "learning_rate": 2.2255968476503105e-05,
      "loss": 0.5279259085655212,
      "step": 492
    },
    {
      "epoch": 1.9147286821705425,
      "grad_norm": 0.10236147791147232,
      "learning_rate": 2.2610673015175764e-05,
      "loss": 0.5332280397415161,
      "step": 494
    },
    {
      "epoch": 1.9224806201550386,
      "grad_norm": 0.06894674897193909,
      "learning_rate": 2.296040697800243e-05,
      "loss": 0.5644171237945557,
      "step": 496
    },
    {
      "epoch": 1.9302325581395348,
      "grad_norm": 0.06815814226865768,
      "learning_rate": 2.3304846081225598e-05,
      "loss": 0.4328697621822357,
      "step": 498
    },
    {
      "epoch": 1.937984496124031,
      "grad_norm": 0.05157966539263725,
      "learning_rate": 2.3643670950639452e-05,
      "loss": 1.0288827419281006,
      "step": 500
    },
    {
      "epoch": 1.945736434108527,
      "grad_norm": 0.09934229403734207,
      "learning_rate": 2.3976567417723124e-05,
      "loss": 0.7295251488685608,
      "step": 502
    },
    {
      "epoch": 1.9534883720930232,
      "grad_norm": 0.048644986003637314,
      "learning_rate": 2.4303226810947168e-05,
      "loss": 0.6065406203269958,
      "step": 504
    },
    {
      "epoch": 1.9612403100775193,
      "grad_norm": 0.04483645409345627,
      "learning_rate": 2.462334624198297e-05,
      "loss": 0.9802693724632263,
      "step": 506
    },
    {
      "epoch": 1.9689922480620154,
      "grad_norm": 0.04317508637905121,
      "learning_rate": 2.4936628886549827e-05,
      "loss": 0.7508119344711304,
      "step": 508
    },
    {
      "epoch": 1.9767441860465116,
      "grad_norm": 0.04037319868803024,
      "learning_rate": 2.524278425963931e-05,
      "loss": 0.9507177472114563,
      "step": 510
    },
    {
      "epoch": 1.9844961240310077,
      "grad_norm": 0.09510686993598938,
      "learning_rate": 2.5541528484861597e-05,
      "loss": 0.7446141242980957,
      "step": 512
    },
    {
      "epoch": 1.9922480620155039,
      "grad_norm": 0.05790317803621292,
      "learning_rate": 2.583258455766412e-05,
      "loss": 1.0172169208526611,
      "step": 514
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.06235863268375397,
      "learning_rate": 2.6115682602178452e-05,
      "loss": 0.8593966364860535,
      "step": 516
    },
    {
      "epoch": 2.007751937984496,
      "grad_norm": 0.10687481611967087,
      "learning_rate": 2.6390560121457266e-05,
      "loss": 0.49814847111701965,
      "step": 518
    },
    {
      "epoch": 2.0155038759689923,
      "grad_norm": 0.06080131232738495,
      "learning_rate": 2.665696224086932e-05,
      "loss": 0.6426661610603333,
      "step": 520
    },
    {
      "epoch": 2.0232558139534884,
      "grad_norm": 0.05089246854186058,
      "learning_rate": 2.6914641944426832e-05,
      "loss": 0.3929331600666046,
      "step": 522
    },
    {
      "epoch": 2.0310077519379846,
      "grad_norm": 0.052005868405103683,
      "learning_rate": 2.7163360303826028e-05,
      "loss": 0.7454217672348022,
      "step": 524
    },
    {
      "epoch": 2.0387596899224807,
      "grad_norm": 0.08798030763864517,
      "learning_rate": 2.7402886699988654e-05,
      "loss": 0.555336594581604,
      "step": 526
    },
    {
      "epoch": 2.046511627906977,
      "grad_norm": 0.08520969748497009,
      "learning_rate": 2.7632999036898793e-05,
      "loss": 0.7960832118988037,
      "step": 528
    },
    {
      "epoch": 2.054263565891473,
      "grad_norm": 0.11796343326568604,
      "learning_rate": 2.7853483947536958e-05,
      "loss": 0.32746729254722595,
      "step": 530
    },
    {
      "epoch": 2.062015503875969,
      "grad_norm": 0.08845008909702301,
      "learning_rate": 2.806413699172034e-05,
      "loss": 0.6935387253761292,
      "step": 532
    },
    {
      "epoch": 2.0697674418604652,
      "grad_norm": 0.0798339694738388,
      "learning_rate": 2.8264762845665833e-05,
      "loss": 0.43687596917152405,
      "step": 534
    },
    {
      "epoch": 2.0775193798449614,
      "grad_norm": 0.13349516689777374,
      "learning_rate": 2.8455175483100068e-05,
      "loss": 0.5209518671035767,
      "step": 536
    },
    {
      "epoch": 2.0852713178294575,
      "grad_norm": 0.09369954466819763,
      "learning_rate": 2.8635198347748558e-05,
      "loss": 0.6955545544624329,
      "step": 538
    },
    {
      "epoch": 2.0930232558139537,
      "grad_norm": 0.05283479019999504,
      "learning_rate": 2.8804664517043884e-05,
      "loss": 0.5326330661773682,
      "step": 540
    },
    {
      "epoch": 2.10077519379845,
      "grad_norm": 0.08794603496789932,
      "learning_rate": 2.896341685690131e-05,
      "loss": 0.2991466224193573,
      "step": 542
    },
    {
      "epoch": 2.108527131782946,
      "grad_norm": 0.061321720480918884,
      "learning_rate": 2.911130816741817e-05,
      "loss": 0.6528932452201843,
      "step": 544
    },
    {
      "epoch": 2.116279069767442,
      "grad_norm": 0.054165348410606384,
      "learning_rate": 2.9248201319362058e-05,
      "loss": 0.4507254660129547,
      "step": 546
    },
    {
      "epoch": 2.124031007751938,
      "grad_norm": 0.049841977655887604,
      "learning_rate": 2.937396938132106e-05,
      "loss": 0.5575969219207764,
      "step": 548
    },
    {
      "epoch": 2.1317829457364343,
      "grad_norm": 0.0796908587217331,
      "learning_rate": 2.94884957373984e-05,
      "loss": 0.34075814485549927,
      "step": 550
    },
    {
      "epoch": 2.13953488372093,
      "grad_norm": 0.0508221872150898,
      "learning_rate": 2.959167419534217e-05,
      "loss": 0.6530574560165405,
      "step": 552
    },
    {
      "epoch": 2.147286821705426,
      "grad_norm": 0.04539173096418381,
      "learning_rate": 2.968340908500995e-05,
      "loss": 0.6907740831375122,
      "step": 554
    },
    {
      "epoch": 2.1550387596899223,
      "grad_norm": 0.058205924928188324,
      "learning_rate": 2.9763615347076983e-05,
      "loss": 0.44485923647880554,
      "step": 556
    },
    {
      "epoch": 2.1627906976744184,
      "grad_norm": 0.06929825246334076,
      "learning_rate": 2.9832218611905778e-05,
      "loss": 0.8006160259246826,
      "step": 558
    },
    {
      "epoch": 2.1705426356589146,
      "grad_norm": 0.04289000853896141,
      "learning_rate": 2.98891552685038e-05,
      "loss": 0.4540764391422272,
      "step": 560
    },
    {
      "epoch": 2.1782945736434107,
      "grad_norm": 0.09553489089012146,
      "learning_rate": 2.993437252350551e-05,
      "loss": 0.6768143177032471,
      "step": 562
    },
    {
      "epoch": 2.186046511627907,
      "grad_norm": 0.07250722497701645,
      "learning_rate": 2.9967828450123938e-05,
      "loss": 0.37812501192092896,
      "step": 564
    },
    {
      "epoch": 2.193798449612403,
      "grad_norm": 0.06566791236400604,
      "learning_rate": 2.998949202702644e-05,
      "loss": 0.42279544472694397,
      "step": 566
    },
    {
      "epoch": 2.201550387596899,
      "grad_norm": 0.04562269151210785,
      "learning_rate": 2.9999343167098628e-05,
      "loss": 0.5368779301643372,
      "step": 568
    },
    {
      "epoch": 2.2093023255813953,
      "grad_norm": 0.18373429775238037,
      "learning_rate": 2.999737273606972e-05,
      "loss": 0.2325422167778015,
      "step": 570
    },
    {
      "epoch": 2.2170542635658914,
      "grad_norm": 0.089279904961586,
      "learning_rate": 2.9983582560982107e-05,
      "loss": 0.5718221664428711,
      "step": 572
    },
    {
      "epoch": 2.2248062015503876,
      "grad_norm": 0.05974709987640381,
      "learning_rate": 2.9957985428497287e-05,
      "loss": 0.20123861730098724,
      "step": 574
    },
    {
      "epoch": 2.2325581395348837,
      "grad_norm": 0.09997298568487167,
      "learning_rate": 2.992060507303966e-05,
      "loss": 0.4844609797000885,
      "step": 576
    },
    {
      "epoch": 2.24031007751938,
      "grad_norm": 0.06058935075998306,
      "learning_rate": 2.98714761547893e-05,
      "loss": 0.6003465056419373,
      "step": 578
    },
    {
      "epoch": 2.248062015503876,
      "grad_norm": 0.10959988087415695,
      "learning_rate": 2.981064422754395e-05,
      "loss": 0.49536406993865967,
      "step": 580
    },
    {
      "epoch": 2.255813953488372,
      "grad_norm": 0.03991486132144928,
      "learning_rate": 2.9738165696480244e-05,
      "loss": 0.27629104256629944,
      "step": 582
    },
    {
      "epoch": 2.2635658914728682,
      "grad_norm": 0.054911598563194275,
      "learning_rate": 2.9654107765853097e-05,
      "loss": 0.8991298675537109,
      "step": 584
    },
    {
      "epoch": 2.2713178294573644,
      "grad_norm": 0.15896278619766235,
      "learning_rate": 2.955854837668194e-05,
      "loss": 0.4965202808380127,
      "step": 586
    },
    {
      "epoch": 2.2790697674418605,
      "grad_norm": 0.04574244096875191,
      "learning_rate": 2.9451576134481485e-05,
      "loss": 0.8411018252372742,
      "step": 588
    },
    {
      "epoch": 2.2868217054263567,
      "grad_norm": 0.049831654876470566,
      "learning_rate": 2.9333290227104026e-05,
      "loss": 0.39420267939567566,
      "step": 590
    },
    {
      "epoch": 2.294573643410853,
      "grad_norm": 0.04978864639997482,
      "learning_rate": 2.9203800332769538e-05,
      "loss": 0.6820889115333557,
      "step": 592
    },
    {
      "epoch": 2.302325581395349,
      "grad_norm": 0.047215450555086136,
      "learning_rate": 2.906322651836873e-05,
      "loss": 0.5932551622390747,
      "step": 594
    },
    {
      "epoch": 2.310077519379845,
      "grad_norm": 0.225591778755188,
      "learning_rate": 2.891169912813347e-05,
      "loss": 0.6146702170372009,
      "step": 596
    },
    {
      "epoch": 2.317829457364341,
      "grad_norm": 0.039774637669324875,
      "learning_rate": 2.8749358662777702e-05,
      "loss": 0.5908256769180298,
      "step": 598
    },
    {
      "epoch": 2.3255813953488373,
      "grad_norm": 0.04213051497936249,
      "learning_rate": 2.857635564922104e-05,
      "loss": 0.3715899884700775,
      "step": 600
    },
    {
      "epoch": 2.3333333333333335,
      "grad_norm": 0.15360774099826813,
      "learning_rate": 2.8392850501015662e-05,
      "loss": 0.5040525794029236,
      "step": 602
    },
    {
      "epoch": 2.3410852713178296,
      "grad_norm": 0.03753961622714996,
      "learning_rate": 2.819901336960611e-05,
      "loss": 0.6956257820129395,
      "step": 604
    },
    {
      "epoch": 2.3488372093023258,
      "grad_norm": 0.11748861521482468,
      "learning_rate": 2.7995023986559807e-05,
      "loss": 0.5580706000328064,
      "step": 606
    },
    {
      "epoch": 2.356589147286822,
      "grad_norm": 0.07652665674686432,
      "learning_rate": 2.7781071496914573e-05,
      "loss": 0.7142091989517212,
      "step": 608
    },
    {
      "epoch": 2.3643410852713176,
      "grad_norm": 0.08042655140161514,
      "learning_rate": 2.755735428379772e-05,
      "loss": 0.6314153075218201,
      "step": 610
    },
    {
      "epoch": 2.3720930232558137,
      "grad_norm": 0.07687707245349884,
      "learning_rate": 2.7324079784479287e-05,
      "loss": 0.37621596455574036,
      "step": 612
    },
    {
      "epoch": 2.37984496124031,
      "grad_norm": 0.0547901950776577,
      "learning_rate": 2.7081464298030026e-05,
      "loss": 0.43838873505592346,
      "step": 614
    },
    {
      "epoch": 2.387596899224806,
      "grad_norm": 0.046424973756074905,
      "learning_rate": 2.6829732784762465e-05,
      "loss": 0.6145251989364624,
      "step": 616
    },
    {
      "epoch": 2.395348837209302,
      "grad_norm": 0.054426416754722595,
      "learning_rate": 2.656911865764097e-05,
      "loss": 0.8150652050971985,
      "step": 618
    },
    {
      "epoch": 2.4031007751937983,
      "grad_norm": 0.04885483533143997,
      "learning_rate": 2.629986356585431e-05,
      "loss": 0.5932292342185974,
      "step": 620
    },
    {
      "epoch": 2.4108527131782944,
      "grad_norm": 0.08931045234203339,
      "learning_rate": 2.602221717075134e-05,
      "loss": 0.27623146772384644,
      "step": 622
    },
    {
      "epoch": 2.4186046511627906,
      "grad_norm": 0.198171466588974,
      "learning_rate": 2.5736436914347484e-05,
      "loss": 0.49001190066337585,
      "step": 624
    },
    {
      "epoch": 2.4263565891472867,
      "grad_norm": 0.06653870642185211,
      "learning_rate": 2.544278778061694e-05,
      "loss": 0.362051784992218,
      "step": 626
    },
    {
      "epoch": 2.434108527131783,
      "grad_norm": 0.045915715396404266,
      "learning_rate": 2.514154204979152e-05,
      "loss": 0.7866628766059875,
      "step": 628
    },
    {
      "epoch": 2.441860465116279,
      "grad_norm": 0.17191638052463531,
      "learning_rate": 2.483297904589437e-05,
      "loss": 0.4470142424106598,
      "step": 630
    },
    {
      "epoch": 2.449612403100775,
      "grad_norm": 0.05101427063345909,
      "learning_rate": 2.451738487774237e-05,
      "loss": 0.5275189876556396,
      "step": 632
    },
    {
      "epoch": 2.4573643410852712,
      "grad_norm": 0.21946240961551666,
      "learning_rate": 2.419505217365756e-05,
      "loss": 0.4540758728981018,
      "step": 634
    },
    {
      "epoch": 2.4651162790697674,
      "grad_norm": 0.04593559727072716,
      "learning_rate": 2.3866279810133385e-05,
      "loss": 0.6065298914909363,
      "step": 636
    },
    {
      "epoch": 2.4728682170542635,
      "grad_norm": 0.059740740805864334,
      "learning_rate": 2.3531372634707517e-05,
      "loss": 0.4904002249240875,
      "step": 638
    },
    {
      "epoch": 2.4806201550387597,
      "grad_norm": 0.050732191652059555,
      "learning_rate": 2.3190641183298133e-05,
      "loss": 0.5315344929695129,
      "step": 640
    },
    {
      "epoch": 2.488372093023256,
      "grad_norm": 0.08542405068874359,
      "learning_rate": 2.2844401392265773e-05,
      "loss": 0.5050536394119263,
      "step": 642
    },
    {
      "epoch": 2.496124031007752,
      "grad_norm": 0.05612539127469063,
      "learning_rate": 2.2492974305467652e-05,
      "loss": 0.6769103407859802,
      "step": 644
    },
    {
      "epoch": 2.503875968992248,
      "grad_norm": 0.08466104418039322,
      "learning_rate": 2.213668577657632e-05,
      "loss": 0.6569755673408508,
      "step": 646
    },
    {
      "epoch": 2.511627906976744,
      "grad_norm": 0.06182940676808357,
      "learning_rate": 2.177586616693837e-05,
      "loss": 0.8741461038589478,
      "step": 648
    },
    {
      "epoch": 2.5193798449612403,
      "grad_norm": 0.10157312452793121,
      "learning_rate": 2.141085003925353e-05,
      "loss": 0.25412437319755554,
      "step": 650
    },
    {
      "epoch": 2.5271317829457365,
      "grad_norm": 0.048401493579149246,
      "learning_rate": 2.1041975847358226e-05,
      "loss": 0.5729293823242188,
      "step": 652
    },
    {
      "epoch": 2.5348837209302326,
      "grad_norm": 0.04819415509700775,
      "learning_rate": 2.066958562240101e-05,
      "loss": 0.38875052332878113,
      "step": 654
    },
    {
      "epoch": 2.5426356589147288,
      "grad_norm": 0.055878717452287674,
      "learning_rate": 2.029402465570114e-05,
      "loss": 0.5112055540084839,
      "step": 656
    },
    {
      "epoch": 2.550387596899225,
      "grad_norm": 0.07780204713344574,
      "learning_rate": 1.9915641178584115e-05,
      "loss": 0.6484737992286682,
      "step": 658
    },
    {
      "epoch": 2.558139534883721,
      "grad_norm": 0.05842196196317673,
      "learning_rate": 1.953478603949128e-05,
      "loss": 0.6154912114143372,
      "step": 660
    },
    {
      "epoch": 2.565891472868217,
      "grad_norm": 0.060033708810806274,
      "learning_rate": 1.9151812378662645e-05,
      "loss": 0.6602538228034973,
      "step": 662
    },
    {
      "epoch": 2.5736434108527133,
      "grad_norm": 0.06244450435042381,
      "learning_rate": 1.8767075300694745e-05,
      "loss": 0.3542620837688446,
      "step": 664
    },
    {
      "epoch": 2.5813953488372094,
      "grad_norm": 0.04792879521846771,
      "learning_rate": 1.8380931545277047e-05,
      "loss": 0.5548843145370483,
      "step": 666
    },
    {
      "epoch": 2.5891472868217056,
      "grad_norm": 0.06219867989420891,
      "learning_rate": 1.7993739156412294e-05,
      "loss": 0.4414609968662262,
      "step": 668
    },
    {
      "epoch": 2.5968992248062017,
      "grad_norm": 0.20640681684017181,
      "learning_rate": 1.7605857150427447e-05,
      "loss": 0.5872040390968323,
      "step": 670
    },
    {
      "epoch": 2.604651162790698,
      "grad_norm": 0.10361964255571365,
      "learning_rate": 1.7217645183082965e-05,
      "loss": 0.42644187808036804,
      "step": 672
    },
    {
      "epoch": 2.612403100775194,
      "grad_norm": 0.04928445816040039,
      "learning_rate": 1.682946321608938e-05,
      "loss": 0.7362091541290283,
      "step": 674
    },
    {
      "epoch": 2.62015503875969,
      "grad_norm": 0.047589972615242004,
      "learning_rate": 1.6441671183339962e-05,
      "loss": 0.6339392066001892,
      "step": 676
    },
    {
      "epoch": 2.6279069767441863,
      "grad_norm": 0.0819360688328743,
      "learning_rate": 1.605462865716936e-05,
      "loss": 0.9860416650772095,
      "step": 678
    },
    {
      "epoch": 2.6356589147286824,
      "grad_norm": 0.04943787679076195,
      "learning_rate": 1.566869451494735e-05,
      "loss": 0.8548033237457275,
      "step": 680
    },
    {
      "epoch": 2.6434108527131785,
      "grad_norm": 0.04377022013068199,
      "learning_rate": 1.5284226606317093e-05,
      "loss": 0.44147443771362305,
      "step": 682
    },
    {
      "epoch": 2.6511627906976747,
      "grad_norm": 0.07998047769069672,
      "learning_rate": 1.4901581421386185e-05,
      "loss": 0.6076138615608215,
      "step": 684
    },
    {
      "epoch": 2.6589147286821704,
      "grad_norm": 0.06454984098672867,
      "learning_rate": 1.4521113760178527e-05,
      "loss": 0.5302262902259827,
      "step": 686
    },
    {
      "epoch": 2.6666666666666665,
      "grad_norm": 0.04626493155956268,
      "learning_rate": 1.414317640365306e-05,
      "loss": 0.34584927558898926,
      "step": 688
    },
    {
      "epoch": 2.6744186046511627,
      "grad_norm": 0.051621273159980774,
      "learning_rate": 1.3768119786594849e-05,
      "loss": 0.409042626619339,
      "step": 690
    },
    {
      "epoch": 2.682170542635659,
      "grad_norm": 0.04060710594058037,
      "learning_rate": 1.3396291672681374e-05,
      "loss": 0.6091200113296509,
      "step": 692
    },
    {
      "epoch": 2.689922480620155,
      "grad_norm": 0.08434680104255676,
      "learning_rate": 1.3028036832025761e-05,
      "loss": 0.5412120819091797,
      "step": 694
    },
    {
      "epoch": 2.697674418604651,
      "grad_norm": 0.08353015780448914,
      "learning_rate": 1.2663696721495657e-05,
      "loss": 0.4864289462566376,
      "step": 696
    },
    {
      "epoch": 2.705426356589147,
      "grad_norm": 0.05606602877378464,
      "learning_rate": 1.2303609168104158e-05,
      "loss": 0.6230313777923584,
      "step": 698
    },
    {
      "epoch": 2.7131782945736433,
      "grad_norm": 0.05251370370388031,
      "learning_rate": 1.1948108055766654e-05,
      "loss": 0.5568596124649048,
      "step": 700
    },
    {
      "epoch": 2.7209302325581395,
      "grad_norm": 0.06997047364711761,
      "learning_rate": 1.159752301571363e-05,
      "loss": 0.6038510799407959,
      "step": 702
    },
    {
      "epoch": 2.7286821705426356,
      "grad_norm": 0.055278290063142776,
      "learning_rate": 1.1252179120846778e-05,
      "loss": 0.4581022262573242,
      "step": 704
    },
    {
      "epoch": 2.7364341085271318,
      "grad_norm": 0.09902060031890869,
      "learning_rate": 1.0912396584321704e-05,
      "loss": 0.3432044982910156,
      "step": 706
    },
    {
      "epoch": 2.744186046511628,
      "grad_norm": 0.039606984704732895,
      "learning_rate": 1.0578490462636758e-05,
      "loss": 0.47076573967933655,
      "step": 708
    },
    {
      "epoch": 2.751937984496124,
      "grad_norm": 0.0477338470518589,
      "learning_rate": 1.0250770363503242e-05,
      "loss": 0.5192364454269409,
      "step": 710
    },
    {
      "epoch": 2.75968992248062,
      "grad_norm": 0.05086624622344971,
      "learning_rate": 9.929540158767916e-06,
      "loss": 0.5344876050949097,
      "step": 712
    },
    {
      "epoch": 2.7674418604651163,
      "grad_norm": 0.054383717477321625,
      "learning_rate": 9.615097702653961e-06,
      "loss": 0.43612009286880493,
      "step": 714
    },
    {
      "epoch": 2.7751937984496124,
      "grad_norm": 0.14214906096458435,
      "learning_rate": 9.307734555581662e-06,
      "loss": 0.3253142535686493,
      "step": 716
    },
    {
      "epoch": 2.7829457364341086,
      "grad_norm": 0.0564068928360939,
      "learning_rate": 9.007735713824893e-06,
      "loss": 0.6681756973266602,
      "step": 718
    },
    {
      "epoch": 2.7906976744186047,
      "grad_norm": 0.06708779186010361,
      "learning_rate": 8.715379345254077e-06,
      "loss": 0.5442537665367126,
      "step": 720
    },
    {
      "epoch": 2.798449612403101,
      "grad_norm": 0.0515008307993412,
      "learning_rate": 8.43093653141064e-06,
      "loss": 0.4479949176311493,
      "step": 722
    },
    {
      "epoch": 2.806201550387597,
      "grad_norm": 0.05638190731406212,
      "learning_rate": 8.15467101615213e-06,
      "loss": 0.5014258623123169,
      "step": 724
    },
    {
      "epoch": 2.813953488372093,
      "grad_norm": 0.10201210528612137,
      "learning_rate": 7.886838961101036e-06,
      "loss": 0.46058180928230286,
      "step": 726
    },
    {
      "epoch": 2.8217054263565893,
      "grad_norm": 0.04419756308197975,
      "learning_rate": 7.627688708124106e-06,
      "loss": 0.5526507496833801,
      "step": 728
    },
    {
      "epoch": 2.8294573643410854,
      "grad_norm": 0.04845201596617699,
      "learning_rate": 7.377460549062367e-06,
      "loss": 0.5275202989578247,
      "step": 730
    },
    {
      "epoch": 2.8372093023255816,
      "grad_norm": 0.04706037789583206,
      "learning_rate": 7.13638650292535e-06,
      "loss": 0.507082998752594,
      "step": 732
    },
    {
      "epoch": 2.8449612403100772,
      "grad_norm": 0.09409839659929276,
      "learning_rate": 6.904690100756159e-06,
      "loss": 0.47032397985458374,
      "step": 734
    },
    {
      "epoch": 2.8527131782945734,
      "grad_norm": 0.052645258605480194,
      "learning_rate": 6.682586178366833e-06,
      "loss": 0.5226138830184937,
      "step": 736
    },
    {
      "epoch": 2.8604651162790695,
      "grad_norm": 0.1135181188583374,
      "learning_rate": 6.470280677136127e-06,
      "loss": 0.47321808338165283,
      "step": 738
    },
    {
      "epoch": 2.8682170542635657,
      "grad_norm": 0.17483681440353394,
      "learning_rate": 6.267970453054588e-06,
      "loss": 0.3441992700099945,
      "step": 740
    },
    {
      "epoch": 2.875968992248062,
      "grad_norm": 0.0963001623749733,
      "learning_rate": 6.075843094193802e-06,
      "loss": 0.4783623516559601,
      "step": 742
    },
    {
      "epoch": 2.883720930232558,
      "grad_norm": 0.06458084285259247,
      "learning_rate": 5.894076746769099e-06,
      "loss": 0.55689537525177,
      "step": 744
    },
    {
      "epoch": 2.891472868217054,
      "grad_norm": 0.08305259048938751,
      "learning_rate": 5.722839949957098e-06,
      "loss": 0.46318957209587097,
      "step": 746
    },
    {
      "epoch": 2.89922480620155,
      "grad_norm": 0.04614809900522232,
      "learning_rate": 5.56229147962116e-06,
      "loss": 0.5878427624702454,
      "step": 748
    },
    {
      "epoch": 2.9069767441860463,
      "grad_norm": 0.06380455940961838,
      "learning_rate": 5.412580201089678e-06,
      "loss": 0.3637682795524597,
      "step": 750
    },
    {
      "epoch": 2.9147286821705425,
      "grad_norm": 0.08269240707159042,
      "learning_rate": 5.273844931123672e-06,
      "loss": 0.5682251453399658,
      "step": 752
    },
    {
      "epoch": 2.9224806201550386,
      "grad_norm": 0.04752229154109955,
      "learning_rate": 5.146214309201799e-06,
      "loss": 0.4354322850704193,
      "step": 754
    },
    {
      "epoch": 2.9302325581395348,
      "grad_norm": 0.08373939245939255,
      "learning_rate": 5.029806678241959e-06,
      "loss": 0.801591694355011,
      "step": 756
    },
    {
      "epoch": 2.937984496124031,
      "grad_norm": 0.047429829835891724,
      "learning_rate": 4.924729974870227e-06,
      "loss": 0.6289284229278564,
      "step": 758
    },
    {
      "epoch": 2.945736434108527,
      "grad_norm": 0.07736939191818237,
      "learning_rate": 4.831081629338789e-06,
      "loss": 0.6334935426712036,
      "step": 760
    },
    {
      "epoch": 2.953488372093023,
      "grad_norm": 0.04397986829280853,
      "learning_rate": 4.74894847518571e-06,
      "loss": 0.6884579658508301,
      "step": 762
    },
    {
      "epoch": 2.9612403100775193,
      "grad_norm": 0.05032786726951599,
      "learning_rate": 4.678406668720287e-06,
      "loss": 0.7347701191902161,
      "step": 764
    },
    {
      "epoch": 2.9689922480620154,
      "grad_norm": 0.11958435922861099,
      "learning_rate": 4.61952161840865e-06,
      "loss": 0.24074648320674896,
      "step": 766
    },
    {
      "epoch": 2.9767441860465116,
      "grad_norm": 0.078817218542099,
      "learning_rate": 4.572347924225084e-06,
      "loss": 0.32004672288894653,
      "step": 768
    },
    {
      "epoch": 2.9844961240310077,
      "grad_norm": 0.036485981196165085,
      "learning_rate": 4.536929327025298e-06,
      "loss": 0.3227121829986572,
      "step": 770
    },
    {
      "epoch": 2.992248062015504,
      "grad_norm": 0.041053567081689835,
      "learning_rate": 4.51329866798861e-06,
      "loss": 0.6057068109512329,
      "step": 772
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.11451369524002075,
      "learning_rate": 4.501477858166617e-06,
      "loss": 0.16319799423217773,
      "step": 774
    }
  ],
  "logging_steps": 2,
  "max_steps": 774,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 99999,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 3.2487544184132076e+18,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}