{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.9986731534719153,
  "eval_steps": 142,
  "global_step": 1695,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0,
      "grad_norm": 4.582726955413818,
      "learning_rate": 3e-06,
      "loss": 3.3182,
      "step": 1
    },
    {
      "epoch": 0.0,
      "eval_loss": 3.3362529277801514,
      "eval_runtime": 14.4066,
      "eval_samples_per_second": 33.11,
      "eval_steps_per_second": 8.329,
      "step": 1
    },
    {
      "epoch": 0.0,
      "grad_norm": 4.511408805847168,
      "learning_rate": 6e-06,
      "loss": 3.2788,
      "step": 2
    },
    {
      "epoch": 0.01,
      "grad_norm": 4.698073387145996,
      "learning_rate": 9e-06,
      "loss": 3.3753,
      "step": 3
    },
    {
      "epoch": 0.01,
      "grad_norm": 4.63289213180542,
      "learning_rate": 1.2e-05,
      "loss": 3.3188,
      "step": 4
    },
    {
      "epoch": 0.01,
      "grad_norm": 4.522731781005859,
      "learning_rate": 1.5e-05,
      "loss": 3.3082,
      "step": 5
    },
    {
      "epoch": 0.01,
      "grad_norm": 4.691629409790039,
      "learning_rate": 1.8e-05,
      "loss": 3.2839,
      "step": 6
    },
    {
      "epoch": 0.01,
      "grad_norm": 4.7225470542907715,
      "learning_rate": 2.1e-05,
      "loss": 3.2686,
      "step": 7
    },
    {
      "epoch": 0.01,
      "grad_norm": 4.691238880157471,
      "learning_rate": 2.4e-05,
      "loss": 3.0782,
      "step": 8
    },
    {
      "epoch": 0.02,
      "grad_norm": 4.64508581161499,
      "learning_rate": 2.7000000000000002e-05,
      "loss": 2.9714,
      "step": 9
    },
    {
      "epoch": 0.02,
      "grad_norm": 4.355061054229736,
      "learning_rate": 3e-05,
      "loss": 2.6886,
      "step": 10
    },
    {
      "epoch": 0.02,
      "grad_norm": 4.206029891967773,
      "learning_rate": 2.9999973928796923e-05,
      "loss": 2.4829,
      "step": 11
    },
    {
      "epoch": 0.02,
      "grad_norm": 3.8142871856689453,
      "learning_rate": 2.999989571527831e-05,
      "loss": 2.0556,
      "step": 12
    },
    {
      "epoch": 0.02,
      "grad_norm": 3.5461277961730957,
      "learning_rate": 2.9999765359716046e-05,
      "loss": 1.6669,
      "step": 13
    },
    {
      "epoch": 0.02,
      "grad_norm": 3.3414511680603027,
      "learning_rate": 2.999958286256327e-05,
      "loss": 1.3396,
      "step": 14
    },
    {
      "epoch": 0.03,
      "grad_norm": 3.0794336795806885,
      "learning_rate": 2.9999348224454367e-05,
      "loss": 1.0494,
      "step": 15
    },
    {
      "epoch": 0.03,
      "grad_norm": 2.524055242538452,
      "learning_rate": 2.9999061446204985e-05,
      "loss": 0.7706,
      "step": 16
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.8878108263015747,
      "learning_rate": 2.9998722528812e-05,
      "loss": 0.5666,
      "step": 17
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.5863118171691895,
      "learning_rate": 2.9998331473453557e-05,
      "loss": 0.4017,
      "step": 18
    },
    {
      "epoch": 0.03,
      "grad_norm": 1.3763269186019897,
      "learning_rate": 2.9997888281489015e-05,
      "loss": 0.3645,
      "step": 19
    },
    {
      "epoch": 0.04,
      "grad_norm": 1.0631593465805054,
      "learning_rate": 2.9997392954458985e-05,
      "loss": 0.2886,
      "step": 20
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.9699187278747559,
      "learning_rate": 2.9996845494085306e-05,
      "loss": 0.1911,
      "step": 21
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.8020363450050354,
      "learning_rate": 2.999624590227103e-05,
      "loss": 0.1762,
      "step": 22
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.8209808468818665,
      "learning_rate": 2.9995594181100443e-05,
      "loss": 0.1449,
      "step": 23
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.39430710673332214,
      "learning_rate": 2.9994890332839027e-05,
      "loss": 0.1201,
      "step": 24
    },
    {
      "epoch": 0.04,
      "grad_norm": 0.9686388969421387,
      "learning_rate": 2.9994134359933476e-05,
      "loss": 0.1421,
      "step": 25
    },
    {
      "epoch": 0.05,
      "grad_norm": 1.3096848726272583,
      "learning_rate": 2.999332626501167e-05,
      "loss": 0.1178,
      "step": 26
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.7969554662704468,
      "learning_rate": 2.9992466050882673e-05,
      "loss": 0.1127,
      "step": 27
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.6500488519668579,
      "learning_rate": 2.9991553720536733e-05,
      "loss": 0.0909,
      "step": 28
    },
    {
      "epoch": 0.05,
      "grad_norm": 3.3015730381011963,
      "learning_rate": 2.9990589277145254e-05,
      "loss": 0.2658,
      "step": 29
    },
    {
      "epoch": 0.05,
      "grad_norm": 0.9849212169647217,
      "learning_rate": 2.9989572724060797e-05,
      "loss": 0.0956,
      "step": 30
    },
    {
      "epoch": 0.05,
      "grad_norm": 1.366385579109192,
      "learning_rate": 2.998850406481707e-05,
      "loss": 0.1276,
      "step": 31
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.42471978068351746,
      "learning_rate": 2.9987383303128887e-05,
      "loss": 0.0885,
      "step": 32
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.2538015842437744,
      "learning_rate": 2.9986210442892215e-05,
      "loss": 0.1246,
      "step": 33
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.9688710570335388,
      "learning_rate": 2.9984985488184086e-05,
      "loss": 0.147,
      "step": 34
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.4737780690193176,
      "learning_rate": 2.9983708443262656e-05,
      "loss": 0.1141,
      "step": 35
    },
    {
      "epoch": 0.06,
      "grad_norm": 0.33743003010749817,
      "learning_rate": 2.9982379312567126e-05,
      "loss": 0.119,
      "step": 36
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.49797967076301575,
      "learning_rate": 2.998099810071777e-05,
      "loss": 0.1319,
      "step": 37
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.8518417477607727,
      "learning_rate": 2.9979564812515906e-05,
      "loss": 0.1609,
      "step": 38
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.2824989855289459,
      "learning_rate": 2.9978079452943875e-05,
      "loss": 0.1126,
      "step": 39
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.2989768981933594,
      "learning_rate": 2.9976542027165016e-05,
      "loss": 0.1164,
      "step": 40
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.49210599064826965,
      "learning_rate": 2.9974952540523676e-05,
      "loss": 0.1175,
      "step": 41
    },
    {
      "epoch": 0.07,
      "grad_norm": 0.35604602098464966,
      "learning_rate": 2.997331099854516e-05,
      "loss": 0.0876,
      "step": 42
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.6768718957901001,
      "learning_rate": 2.9971617406935735e-05,
      "loss": 0.1649,
      "step": 43
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.4443073868751526,
      "learning_rate": 2.9969871771582596e-05,
      "loss": 0.0655,
      "step": 44
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.8104651570320129,
      "learning_rate": 2.996807409855385e-05,
      "loss": 0.1336,
      "step": 45
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.4232860803604126,
      "learning_rate": 2.99662243940985e-05,
      "loss": 0.0993,
      "step": 46
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.5418788194656372,
      "learning_rate": 2.9964322664646412e-05,
      "loss": 0.1131,
      "step": 47
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.39015039801597595,
      "learning_rate": 2.9962368916808308e-05,
      "loss": 0.13,
      "step": 48
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.40271061658859253,
      "learning_rate": 2.9960363157375724e-05,
      "loss": 0.1088,
      "step": 49
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.23656688630580902,
      "learning_rate": 2.9958305393321e-05,
      "loss": 0.1209,
      "step": 50
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.4160294234752655,
      "learning_rate": 2.995619563179726e-05,
      "loss": 0.1214,
      "step": 51
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.25202676653862,
      "learning_rate": 2.9954033880138368e-05,
      "loss": 0.115,
      "step": 52
    },
    {
      "epoch": 0.09,
      "grad_norm": 0.5302817821502686,
      "learning_rate": 2.995182014585892e-05,
      "loss": 0.124,
      "step": 53
    },
    {
      "epoch": 0.1,
      "grad_norm": 1.063620924949646,
      "learning_rate": 2.9949554436654215e-05,
      "loss": 0.1347,
      "step": 54
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.3588281273841858,
      "learning_rate": 2.994723676040022e-05,
      "loss": 0.0931,
      "step": 55
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.6446585059165955,
      "learning_rate": 2.9944867125153548e-05,
      "loss": 0.1108,
      "step": 56
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.31614363193511963,
      "learning_rate": 2.9942445539151432e-05,
      "loss": 0.0712,
      "step": 57
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.2502954304218292,
      "learning_rate": 2.9939972010811693e-05,
      "loss": 0.0914,
      "step": 58
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.33516696095466614,
      "learning_rate": 2.993744654873272e-05,
      "loss": 0.0923,
      "step": 59
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.2990712821483612,
      "learning_rate": 2.993486916169341e-05,
      "loss": 0.0899,
      "step": 60
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.40888020396232605,
      "learning_rate": 2.9932239858653183e-05,
      "loss": 0.0789,
      "step": 61
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.756536602973938,
      "learning_rate": 2.992955864875192e-05,
      "loss": 0.1707,
      "step": 62
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.3337976932525635,
      "learning_rate": 2.9926825541309928e-05,
      "loss": 0.0676,
      "step": 63
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.6163253784179688,
      "learning_rate": 2.9924040545827936e-05,
      "loss": 0.0961,
      "step": 64
    },
    {
      "epoch": 0.11,
      "grad_norm": 0.45453113317489624,
      "learning_rate": 2.9921203671987025e-05,
      "loss": 0.0724,
      "step": 65
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.64926677942276,
      "learning_rate": 2.9918314929648637e-05,
      "loss": 0.1015,
      "step": 66
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.37845009565353394,
      "learning_rate": 2.991537432885449e-05,
      "loss": 0.0964,
      "step": 67
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.3352985382080078,
      "learning_rate": 2.991238187982659e-05,
      "loss": 0.0862,
      "step": 68
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.485416442155838,
      "learning_rate": 2.9909337592967176e-05,
      "loss": 0.1095,
      "step": 69
    },
    {
      "epoch": 0.12,
      "grad_norm": 0.9582281708717346,
      "learning_rate": 2.9906241478858666e-05,
      "loss": 0.1314,
      "step": 70
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.3674233555793762,
      "learning_rate": 2.990309354826366e-05,
      "loss": 0.109,
      "step": 71
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.408796101808548,
      "learning_rate": 2.9899893812124862e-05,
      "loss": 0.126,
      "step": 72
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.5258140563964844,
      "learning_rate": 2.989664228156507e-05,
      "loss": 0.0596,
      "step": 73
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.3256467580795288,
      "learning_rate": 2.9893338967887128e-05,
      "loss": 0.0736,
      "step": 74
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.22679170966148376,
      "learning_rate": 2.988998388257388e-05,
      "loss": 0.0982,
      "step": 75
    },
    {
      "epoch": 0.13,
      "grad_norm": 0.2959941625595093,
      "learning_rate": 2.988657703728815e-05,
      "loss": 0.0515,
      "step": 76
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.26353922486305237,
      "learning_rate": 2.9883118443872662e-05,
      "loss": 0.0732,
      "step": 77
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.23671653866767883,
      "learning_rate": 2.9879608114350064e-05,
      "loss": 0.0865,
      "step": 78
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.415596067905426,
      "learning_rate": 2.9876046060922803e-05,
      "loss": 0.0871,
      "step": 79
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.5430477261543274,
      "learning_rate": 2.987243229597316e-05,
      "loss": 0.1385,
      "step": 80
    },
    {
      "epoch": 0.14,
      "grad_norm": 0.6044889092445374,
      "learning_rate": 2.9868766832063156e-05,
      "loss": 0.1279,
      "step": 81
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.452231764793396,
      "learning_rate": 2.986504968193454e-05,
      "loss": 0.0767,
      "step": 82
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.3864612579345703,
      "learning_rate": 2.9861280858508712e-05,
      "loss": 0.0728,
      "step": 83
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.3158702850341797,
      "learning_rate": 2.9857460374886717e-05,
      "loss": 0.087,
      "step": 84
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.448408842086792,
      "learning_rate": 2.985358824434916e-05,
      "loss": 0.078,
      "step": 85
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.4518391788005829,
      "learning_rate": 2.984966448035619e-05,
      "loss": 0.1013,
      "step": 86
    },
    {
      "epoch": 0.15,
      "grad_norm": 0.9920629858970642,
      "learning_rate": 2.9845689096547442e-05,
      "loss": 0.1145,
      "step": 87
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.5967357754707336,
      "learning_rate": 2.9841662106741986e-05,
      "loss": 0.1132,
      "step": 88
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.5016050338745117,
      "learning_rate": 2.983758352493829e-05,
      "loss": 0.094,
      "step": 89
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.300605833530426,
      "learning_rate": 2.983345336531415e-05,
      "loss": 0.0762,
      "step": 90
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.44358232617378235,
      "learning_rate": 2.9829271642226665e-05,
      "loss": 0.0969,
      "step": 91
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.9166831374168396,
      "learning_rate": 2.9825038370212183e-05,
      "loss": 0.1532,
      "step": 92
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.6240636706352234,
      "learning_rate": 2.982075356398623e-05,
      "loss": 0.1351,
      "step": 93
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.34054040908813477,
      "learning_rate": 2.9816417238443482e-05,
      "loss": 0.0803,
      "step": 94
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.3595755100250244,
      "learning_rate": 2.9812029408657698e-05,
      "loss": 0.0522,
      "step": 95
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.4143121540546417,
      "learning_rate": 2.9807590089881687e-05,
      "loss": 0.0816,
      "step": 96
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.2190728336572647,
      "learning_rate": 2.980309929754722e-05,
      "loss": 0.0606,
      "step": 97
    },
    {
      "epoch": 0.17,
      "grad_norm": 0.23459163308143616,
      "learning_rate": 2.9798557047265023e-05,
      "loss": 0.0841,
      "step": 98
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.24551355838775635,
      "learning_rate": 2.979396335482469e-05,
      "loss": 0.0545,
      "step": 99
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.471500426530838,
      "learning_rate": 2.9789318236194618e-05,
      "loss": 0.0948,
      "step": 100
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.5168260335922241,
      "learning_rate": 2.9784621707521993e-05,
      "loss": 0.1471,
      "step": 101
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.7932892441749573,
      "learning_rate": 2.97798737851327e-05,
      "loss": 0.1101,
      "step": 102
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.32231828570365906,
      "learning_rate": 2.977507448553128e-05,
      "loss": 0.0525,
      "step": 103
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.30824771523475647,
      "learning_rate": 2.9770223825400872e-05,
      "loss": 0.0965,
      "step": 104
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.5549985766410828,
      "learning_rate": 2.9765321821603143e-05,
      "loss": 0.1105,
      "step": 105
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.6406698822975159,
      "learning_rate": 2.9760368491178244e-05,
      "loss": 0.1111,
      "step": 106
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.3495526909828186,
      "learning_rate": 2.9755363851344753e-05,
      "loss": 0.0792,
      "step": 107
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.3962193429470062,
      "learning_rate": 2.9750307919499595e-05,
      "loss": 0.0894,
      "step": 108
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.5372034907341003,
      "learning_rate": 2.9745200713218002e-05,
      "loss": 0.1735,
      "step": 109
    },
    {
      "epoch": 0.19,
      "grad_norm": 0.22693391144275665,
      "learning_rate": 2.9740042250253443e-05,
      "loss": 0.0603,
      "step": 110
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.4251677393913269,
      "learning_rate": 2.973483254853756e-05,
      "loss": 0.1124,
      "step": 111
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.22285960614681244,
      "learning_rate": 2.9729571626180116e-05,
      "loss": 0.0677,
      "step": 112
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.18614518642425537,
      "learning_rate": 2.972425950146891e-05,
      "loss": 0.0854,
      "step": 113
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.3953099250793457,
      "learning_rate": 2.9718896192869758e-05,
      "loss": 0.0832,
      "step": 114
    },
    {
      "epoch": 0.2,
      "grad_norm": 0.2483922243118286,
      "learning_rate": 2.9713481719026368e-05,
      "loss": 0.0992,
      "step": 115
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.3185414969921112,
      "learning_rate": 2.970801609876032e-05,
      "loss": 0.0707,
      "step": 116
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.40958282351493835,
      "learning_rate": 2.9702499351070992e-05,
      "loss": 0.1019,
      "step": 117
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.5805838704109192,
      "learning_rate": 2.969693149513548e-05,
      "loss": 0.116,
      "step": 118
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.21251153945922852,
      "learning_rate": 2.969131255030855e-05,
      "loss": 0.0843,
      "step": 119
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.4262985587120056,
      "learning_rate": 2.9685642536122545e-05,
      "loss": 0.0706,
      "step": 120
    },
    {
      "epoch": 0.21,
      "grad_norm": 0.4474552869796753,
      "learning_rate": 2.9679921472287358e-05,
      "loss": 0.0859,
      "step": 121
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.3137888014316559,
      "learning_rate": 2.967414937869031e-05,
      "loss": 0.1316,
      "step": 122
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.2894512712955475,
      "learning_rate": 2.9668326275396133e-05,
      "loss": 0.0715,
      "step": 123
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.2738913595676422,
      "learning_rate": 2.966245218264687e-05,
      "loss": 0.0524,
      "step": 124
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.6975142955780029,
      "learning_rate": 2.96565271208618e-05,
      "loss": 0.1225,
      "step": 125
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.5777083039283752,
      "learning_rate": 2.9650551110637397e-05,
      "loss": 0.1309,
      "step": 126
    },
    {
      "epoch": 0.22,
      "grad_norm": 0.8783730864524841,
      "learning_rate": 2.964452417274723e-05,
      "loss": 0.0995,
      "step": 127
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.28593194484710693,
      "learning_rate": 2.96384463281419e-05,
      "loss": 0.0771,
      "step": 128
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.35265710949897766,
      "learning_rate": 2.9632317597948968e-05,
      "loss": 0.0696,
      "step": 129
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.4801347851753235,
      "learning_rate": 2.9626138003472884e-05,
      "loss": 0.0954,
      "step": 130
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.7368841171264648,
      "learning_rate": 2.9619907566194915e-05,
      "loss": 0.1493,
      "step": 131
    },
    {
      "epoch": 0.23,
      "grad_norm": 0.3211562931537628,
      "learning_rate": 2.9613626307773055e-05,
      "loss": 0.0721,
      "step": 132
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.2380824089050293,
      "learning_rate": 2.9607294250041965e-05,
      "loss": 0.0574,
      "step": 133
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.4969393312931061,
      "learning_rate": 2.96009114150129e-05,
      "loss": 0.0463,
      "step": 134
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.4968222379684448,
      "learning_rate": 2.959447782487361e-05,
      "loss": 0.1302,
      "step": 135
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.5577062368392944,
      "learning_rate": 2.9587993501988292e-05,
      "loss": 0.1239,
      "step": 136
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.5446056127548218,
      "learning_rate": 2.958145846889749e-05,
      "loss": 0.1169,
      "step": 137
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.3976798355579376,
      "learning_rate": 2.957487274831803e-05,
      "loss": 0.1012,
      "step": 138
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.6089766621589661,
      "learning_rate": 2.9568236363142927e-05,
      "loss": 0.0859,
      "step": 139
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.3312886655330658,
      "learning_rate": 2.9561549336441333e-05,
      "loss": 0.0996,
      "step": 140
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.3692180812358856,
      "learning_rate": 2.955481169145841e-05,
      "loss": 0.0958,
      "step": 141
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.23727117478847504,
      "learning_rate": 2.95480234516153e-05,
      "loss": 0.0815,
      "step": 142
    },
    {
      "epoch": 0.25,
      "eval_loss": 0.09037704020738602,
      "eval_runtime": 14.7372,
      "eval_samples_per_second": 32.367,
      "eval_steps_per_second": 8.143,
      "step": 142
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.44341254234313965,
      "learning_rate": 2.9541184640509015e-05,
      "loss": 0.108,
      "step": 143
    },
    {
      "epoch": 0.25,
      "grad_norm": 0.27407336235046387,
      "learning_rate": 2.953429528191236e-05,
      "loss": 0.1192,
      "step": 144
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.39919063448905945,
      "learning_rate": 2.9527355399773847e-05,
      "loss": 0.0832,
      "step": 145
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.3581640422344208,
      "learning_rate": 2.9520365018217622e-05,
      "loss": 0.0625,
      "step": 146
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.29779088497161865,
      "learning_rate": 2.951332416154337e-05,
      "loss": 0.111,
      "step": 147
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.21348236501216888,
      "learning_rate": 2.9506232854226242e-05,
      "loss": 0.092,
      "step": 148
    },
    {
      "epoch": 0.26,
      "grad_norm": 0.3025546073913574,
      "learning_rate": 2.9499091120916757e-05,
      "loss": 0.1217,
      "step": 149
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.24288436770439148,
      "learning_rate": 2.9491898986440726e-05,
      "loss": 0.0752,
      "step": 150
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.22768545150756836,
      "learning_rate": 2.9484656475799164e-05,
      "loss": 0.0964,
      "step": 151
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.7110497355461121,
      "learning_rate": 2.9477363614168197e-05,
      "loss": 0.1099,
      "step": 152
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.7041640281677246,
      "learning_rate": 2.9470020426898983e-05,
      "loss": 0.1599,
      "step": 153
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.4882669746875763,
      "learning_rate": 2.946262693951762e-05,
      "loss": 0.1059,
      "step": 154
    },
    {
      "epoch": 0.27,
      "grad_norm": 0.2939213812351227,
      "learning_rate": 2.9455183177725055e-05,
      "loss": 0.0776,
      "step": 155
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.25689342617988586,
      "learning_rate": 2.9447689167397e-05,
      "loss": 0.098,
      "step": 156
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.22836799919605255,
      "learning_rate": 2.944014493458383e-05,
      "loss": 0.0838,
      "step": 157
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.525640606880188,
      "learning_rate": 2.9432550505510516e-05,
      "loss": 0.0958,
      "step": 158
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.6011547446250916,
      "learning_rate": 2.942490590657651e-05,
      "loss": 0.1154,
      "step": 159
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.3123567998409271,
      "learning_rate": 2.9417211164355668e-05,
      "loss": 0.1097,
      "step": 160
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.39742252230644226,
      "learning_rate": 2.9409466305596135e-05,
      "loss": 0.0671,
      "step": 161
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.2927655279636383,
      "learning_rate": 2.9401671357220297e-05,
      "loss": 0.0752,
      "step": 162
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.6094531416893005,
      "learning_rate": 2.9393826346324634e-05,
      "loss": 0.1437,
      "step": 163
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.23876389861106873,
      "learning_rate": 2.9385931300179675e-05,
      "loss": 0.1064,
      "step": 164
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.26306307315826416,
      "learning_rate": 2.9377986246229853e-05,
      "loss": 0.065,
      "step": 165
    },
    {
      "epoch": 0.29,
      "grad_norm": 0.3406679332256317,
      "learning_rate": 2.9369991212093462e-05,
      "loss": 0.0939,
      "step": 166
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.5959851741790771,
      "learning_rate": 2.9361946225562516e-05,
      "loss": 0.1163,
      "step": 167
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.3201580047607422,
      "learning_rate": 2.9353851314602676e-05,
      "loss": 0.115,
      "step": 168
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.4657009243965149,
      "learning_rate": 2.9345706507353158e-05,
      "loss": 0.0877,
      "step": 169
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.313793420791626,
      "learning_rate": 2.9337511832126616e-05,
      "loss": 0.0574,
      "step": 170
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.2866755723953247,
      "learning_rate": 2.9329267317409053e-05,
      "loss": 0.0968,
      "step": 171
    },
    {
      "epoch": 0.3,
      "grad_norm": 0.2852758765220642,
      "learning_rate": 2.9320972991859728e-05,
      "loss": 0.1091,
      "step": 172
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.3556657135486603,
      "learning_rate": 2.9312628884311048e-05,
      "loss": 0.0716,
      "step": 173
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.3235069215297699,
      "learning_rate": 2.9304235023768465e-05,
      "loss": 0.0532,
      "step": 174
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.3486980199813843,
      "learning_rate": 2.9295791439410387e-05,
      "loss": 0.1146,
      "step": 175
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.319968044757843,
      "learning_rate": 2.9287298160588073e-05,
      "loss": 0.1007,
      "step": 176
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.48733624815940857,
      "learning_rate": 2.927875521682551e-05,
      "loss": 0.1164,
      "step": 177
    },
    {
      "epoch": 0.31,
      "grad_norm": 0.2947496473789215,
      "learning_rate": 2.9270162637819352e-05,
      "loss": 0.0434,
      "step": 178
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.35532888770103455,
      "learning_rate": 2.9261520453438775e-05,
      "loss": 0.0498,
      "step": 179
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.5203955173492432,
      "learning_rate": 2.9252828693725404e-05,
      "loss": 0.1061,
      "step": 180
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.4153119921684265,
      "learning_rate": 2.9244087388893187e-05,
      "loss": 0.0953,
      "step": 181
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.41368868947029114,
      "learning_rate": 2.9235296569328303e-05,
      "loss": 0.1015,
      "step": 182
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.4907214343547821,
      "learning_rate": 2.922645626558905e-05,
      "loss": 0.1213,
      "step": 183
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.2939632833003998,
      "learning_rate": 2.921756650840574e-05,
      "loss": 0.0757,
      "step": 184
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.36250534653663635,
      "learning_rate": 2.92086273286806e-05,
      "loss": 0.0983,
      "step": 185
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.34401020407676697,
      "learning_rate": 2.9199638757487648e-05,
      "loss": 0.115,
      "step": 186
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.24301297962665558,
      "learning_rate": 2.9190600826072603e-05,
      "loss": 0.0871,
      "step": 187
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.3000582754611969,
      "learning_rate": 2.918151356585276e-05,
      "loss": 0.0675,
      "step": 188
    },
    {
      "epoch": 0.33,
      "grad_norm": 0.23538866639137268,
      "learning_rate": 2.9172377008416898e-05,
      "loss": 0.0645,
      "step": 189
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.3890560567378998,
      "learning_rate": 2.916319118552515e-05,
      "loss": 0.1009,
      "step": 190
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.4634495973587036,
      "learning_rate": 2.9153956129108918e-05,
      "loss": 0.098,
      "step": 191
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.20195429027080536,
      "learning_rate": 2.9144671871270734e-05,
      "loss": 0.0894,
      "step": 192
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.42528125643730164,
      "learning_rate": 2.913533844428417e-05,
      "loss": 0.1398,
      "step": 193
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.36887580156326294,
      "learning_rate": 2.912595588059371e-05,
      "loss": 0.1066,
      "step": 194
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.17726579308509827,
      "learning_rate": 2.9116524212814653e-05,
      "loss": 0.0611,
      "step": 195
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.19946645200252533,
      "learning_rate": 2.9107043473733e-05,
      "loss": 0.0947,
      "step": 196
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.3372867703437805,
      "learning_rate": 2.9097513696305304e-05,
      "loss": 0.0952,
      "step": 197
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.3198525011539459,
      "learning_rate": 2.908793491365861e-05,
      "loss": 0.1155,
      "step": 198
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.3473115861415863,
      "learning_rate": 2.90783071590903e-05,
      "loss": 0.0752,
      "step": 199
    },
    {
      "epoch": 0.35,
      "grad_norm": 0.20476488769054413,
      "learning_rate": 2.9068630466067997e-05,
      "loss": 0.0708,
      "step": 200
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.3317318558692932,
      "learning_rate": 2.905890486822943e-05,
      "loss": 0.111,
      "step": 201
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.247785285115242,
      "learning_rate": 2.9049130399382345e-05,
      "loss": 0.0855,
      "step": 202
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.18749944865703583,
      "learning_rate": 2.903930709350436e-05,
      "loss": 0.0864,
      "step": 203
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.43616777658462524,
      "learning_rate": 2.9029434984742866e-05,
      "loss": 0.1307,
      "step": 204
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.23449763655662537,
      "learning_rate": 2.9019514107414888e-05,
      "loss": 0.086,
      "step": 205
    },
    {
      "epoch": 0.36,
      "grad_norm": 0.3940359055995941,
      "learning_rate": 2.9009544496006998e-05,
      "loss": 0.0711,
      "step": 206
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.8495943546295166,
      "learning_rate": 2.8999526185175155e-05,
      "loss": 0.1159,
      "step": 207
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.6435235142707825,
      "learning_rate": 2.898945920974462e-05,
      "loss": 0.0951,
      "step": 208
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.24877464771270752,
      "learning_rate": 2.8979343604709818e-05,
      "loss": 0.0472,
      "step": 209
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.3478896915912628,
      "learning_rate": 2.8969179405234202e-05,
      "loss": 0.1026,
      "step": 210
    },
    {
      "epoch": 0.37,
      "grad_norm": 0.4590359032154083,
      "learning_rate": 2.8958966646650172e-05,
      "loss": 0.1551,
      "step": 211
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.33144280314445496,
      "learning_rate": 2.894870536445891e-05,
      "loss": 0.069,
      "step": 212
    },
    {
      "epoch": 0.38,
      "grad_norm": 1.0096515417099,
      "learning_rate": 2.893839559433028e-05,
      "loss": 0.1919,
      "step": 213
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.4398234188556671,
      "learning_rate": 2.8928037372102698e-05,
      "loss": 0.1481,
      "step": 214
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.2270442694425583,
      "learning_rate": 2.8917630733783004e-05,
      "loss": 0.0811,
      "step": 215
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.30207884311676025,
      "learning_rate": 2.8907175715546337e-05,
      "loss": 0.091,
      "step": 216
    },
    {
      "epoch": 0.38,
      "grad_norm": 0.21270009875297546,
      "learning_rate": 2.889667235373603e-05,
      "loss": 0.092,
      "step": 217
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.2912154197692871,
      "learning_rate": 2.888612068486344e-05,
      "loss": 0.0834,
      "step": 218
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.23577392101287842,
      "learning_rate": 2.887552074560787e-05,
      "loss": 0.1215,
      "step": 219
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.16501927375793457,
      "learning_rate": 2.8864872572816407e-05,
      "loss": 0.0665,
      "step": 220
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.1898716688156128,
      "learning_rate": 2.885417620350381e-05,
      "loss": 0.0595,
      "step": 221
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.25912415981292725,
      "learning_rate": 2.8843431674852366e-05,
      "loss": 0.0455,
      "step": 222
    },
    {
      "epoch": 0.39,
      "grad_norm": 0.33551594614982605,
      "learning_rate": 2.883263902421179e-05,
      "loss": 0.0836,
      "step": 223
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.2037317305803299,
      "learning_rate": 2.8821798289099054e-05,
      "loss": 0.0974,
      "step": 224
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.29482024908065796,
      "learning_rate": 2.881090950719831e-05,
      "loss": 0.1051,
      "step": 225
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.20771493017673492,
      "learning_rate": 2.87999727163607e-05,
      "loss": 0.0761,
      "step": 226
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.2051924616098404,
      "learning_rate": 2.878898795460426e-05,
      "loss": 0.0553,
      "step": 227
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.2791610062122345,
      "learning_rate": 2.8777955260113794e-05,
      "loss": 0.0651,
      "step": 228
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.35246115922927856,
      "learning_rate": 2.8766874671240708e-05,
      "loss": 0.1042,
      "step": 229
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.29613715410232544,
      "learning_rate": 2.8755746226502914e-05,
      "loss": 0.0679,
      "step": 230
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.5915008783340454,
      "learning_rate": 2.874456996458467e-05,
      "loss": 0.1112,
      "step": 231
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.5234435200691223,
      "learning_rate": 2.8733345924336448e-05,
      "loss": 0.1,
      "step": 232
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.582992434501648,
      "learning_rate": 2.872207414477482e-05,
      "loss": 0.1505,
      "step": 233
    },
    {
      "epoch": 0.41,
      "grad_norm": 0.463219553232193,
      "learning_rate": 2.8710754665082295e-05,
      "loss": 0.1148,
      "step": 234
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.3954102694988251,
      "learning_rate": 2.8699387524607206e-05,
      "loss": 0.1252,
      "step": 235
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.2828240692615509,
      "learning_rate": 2.868797276286355e-05,
      "loss": 0.0916,
      "step": 236
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.6134512424468994,
      "learning_rate": 2.8676510419530875e-05,
      "loss": 0.0953,
      "step": 237
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.4864056408405304,
      "learning_rate": 2.866500053445412e-05,
      "loss": 0.0887,
      "step": 238
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.34841328859329224,
      "learning_rate": 2.86534431476435e-05,
      "loss": 0.0919,
      "step": 239
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.22810260951519012,
      "learning_rate": 2.8641838299274336e-05,
      "loss": 0.0515,
      "step": 240
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.7178006768226624,
      "learning_rate": 2.863018602968695e-05,
      "loss": 0.2241,
      "step": 241
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.3194310665130615,
      "learning_rate": 2.8618486379386496e-05,
      "loss": 0.111,
      "step": 242
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.15743738412857056,
      "learning_rate": 2.8606739389042838e-05,
      "loss": 0.0634,
      "step": 243
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.24189315736293793,
      "learning_rate": 2.8594945099490395e-05,
      "loss": 0.0644,
      "step": 244
    },
    {
      "epoch": 0.43,
      "grad_norm": 0.21816010773181915,
      "learning_rate": 2.8583103551728008e-05,
      "loss": 0.0852,
      "step": 245
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.282953679561615,
      "learning_rate": 2.857121478691881e-05,
      "loss": 0.0842,
      "step": 246
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.2257271707057953,
      "learning_rate": 2.855927884639004e-05,
      "loss": 0.0686,
      "step": 247
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.2596382796764374,
      "learning_rate": 2.854729577163294e-05,
      "loss": 0.0595,
      "step": 248
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.3601246178150177,
      "learning_rate": 2.8535265604302614e-05,
      "loss": 0.0781,
      "step": 249
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.22987039387226105,
      "learning_rate": 2.852318838621784e-05,
      "loss": 0.0712,
      "step": 250
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.2668680250644684,
      "learning_rate": 2.851106415936098e-05,
      "loss": 0.1015,
      "step": 251
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.3502819836139679,
      "learning_rate": 2.8498892965877776e-05,
      "loss": 0.1164,
      "step": 252
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.4412795603275299,
      "learning_rate": 2.848667484807726e-05,
      "loss": 0.0891,
      "step": 253
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.506323516368866,
      "learning_rate": 2.8474409848431562e-05,
      "loss": 0.1134,
      "step": 254
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.519180178642273,
      "learning_rate": 2.8462098009575793e-05,
      "loss": 0.1147,
      "step": 255
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.4656154215335846,
      "learning_rate": 2.8449739374307877e-05,
      "loss": 0.0939,
      "step": 256
    },
    {
      "epoch": 0.45,
      "grad_norm": 0.23170435428619385,
      "learning_rate": 2.8437333985588418e-05,
      "loss": 0.0365,
      "step": 257
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.4948550760746002,
      "learning_rate": 2.8424881886540527e-05,
      "loss": 0.127,
      "step": 258
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.2533349096775055,
      "learning_rate": 2.8412383120449707e-05,
      "loss": 0.0582,
      "step": 259
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.3986698091030121,
      "learning_rate": 2.839983773076367e-05,
      "loss": 0.0687,
      "step": 260
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.5078080296516418,
      "learning_rate": 2.8387245761092203e-05,
      "loss": 0.1166,
      "step": 261
    },
    {
      "epoch": 0.46,
      "grad_norm": 0.5207770466804504,
      "learning_rate": 2.8374607255207012e-05,
      "loss": 0.0615,
      "step": 262
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.578711211681366,
      "learning_rate": 2.8361922257041577e-05,
      "loss": 0.1259,
      "step": 263
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.3096364140510559,
      "learning_rate": 2.8349190810690977e-05,
      "loss": 0.0708,
      "step": 264
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.4309447407722473,
      "learning_rate": 2.8336412960411765e-05,
      "loss": 0.1351,
      "step": 265
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.4576749801635742,
      "learning_rate": 2.8323588750621802e-05,
      "loss": 0.0875,
      "step": 266
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.5308840274810791,
      "learning_rate": 2.8310718225900095e-05,
      "loss": 0.0803,
      "step": 267
    },
    {
      "epoch": 0.47,
      "grad_norm": 0.32970359921455383,
      "learning_rate": 2.8297801430986652e-05,
      "loss": 0.0774,
      "step": 268
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.2413233071565628,
      "learning_rate": 2.8284838410782327e-05,
      "loss": 0.056,
      "step": 269
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.45652082562446594,
      "learning_rate": 2.8271829210348657e-05,
      "loss": 0.165,
      "step": 270
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.36683741211891174,
      "learning_rate": 2.82587738749077e-05,
      "loss": 0.1019,
      "step": 271
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.3328053057193756,
      "learning_rate": 2.824567244984192e-05,
      "loss": 0.1022,
      "step": 272
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.40834105014801025,
      "learning_rate": 2.8232524980693947e-05,
      "loss": 0.0607,
      "step": 273
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.2283175140619278,
      "learning_rate": 2.8219331513166503e-05,
      "loss": 0.0772,
      "step": 274
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.3196583092212677,
      "learning_rate": 2.8206092093122195e-05,
      "loss": 0.0834,
      "step": 275
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.27474430203437805,
      "learning_rate": 2.8192806766583373e-05,
      "loss": 0.0749,
      "step": 276
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.3614490330219269,
      "learning_rate": 2.8179475579731967e-05,
      "loss": 0.0888,
      "step": 277
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.20857541263103485,
      "learning_rate": 2.8166098578909315e-05,
      "loss": 0.035,
      "step": 278
    },
    {
      "epoch": 0.49,
      "grad_norm": 0.27701324224472046,
      "learning_rate": 2.815267581061602e-05,
      "loss": 0.062,
      "step": 279
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.45753782987594604,
      "learning_rate": 2.8139207321511778e-05,
      "loss": 0.1062,
      "step": 280
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.3005506098270416,
      "learning_rate": 2.8125693158415217e-05,
      "loss": 0.1344,
      "step": 281
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.2129545509815216,
      "learning_rate": 2.8112133368303737e-05,
      "loss": 0.0382,
      "step": 282
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.37999290227890015,
      "learning_rate": 2.809852799831334e-05,
      "loss": 0.0812,
      "step": 283
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.1718507707118988,
      "learning_rate": 2.8084877095738477e-05,
      "loss": 0.022,
      "step": 284
    },
    {
      "epoch": 0.5,
      "eval_loss": 0.08163367211818695,
      "eval_runtime": 14.651,
      "eval_samples_per_second": 32.558,
      "eval_steps_per_second": 8.191,
      "step": 284
    },
    {
      "epoch": 0.5,
      "grad_norm": 0.27595949172973633,
      "learning_rate": 2.8071180708031874e-05,
      "loss": 0.0593,
      "step": 285
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.48061394691467285,
      "learning_rate": 2.8057438882804372e-05,
      "loss": 0.1139,
      "step": 286
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.32916727662086487,
      "learning_rate": 2.8043651667824767e-05,
      "loss": 0.0335,
      "step": 287
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.6290953755378723,
      "learning_rate": 2.8029819111019618e-05,
      "loss": 0.099,
      "step": 288
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.7197291851043701,
      "learning_rate": 2.8015941260473117e-05,
      "loss": 0.1165,
      "step": 289
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.4702732264995575,
      "learning_rate": 2.8002018164426896e-05,
      "loss": 0.0948,
      "step": 290
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.4018266201019287,
      "learning_rate": 2.798804987127988e-05,
      "loss": 0.0982,
      "step": 291
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.7587385773658752,
      "learning_rate": 2.7974036429588082e-05,
      "loss": 0.0915,
      "step": 292
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.36278435587882996,
      "learning_rate": 2.7959977888064484e-05,
      "loss": 0.0442,
      "step": 293
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.5261387228965759,
      "learning_rate": 2.7945874295578827e-05,
      "loss": 0.117,
      "step": 294
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.8715558648109436,
      "learning_rate": 2.7931725701157462e-05,
      "loss": 0.1461,
      "step": 295
    },
    {
      "epoch": 0.52,
      "grad_norm": 0.3017662465572357,
      "learning_rate": 2.7917532153983176e-05,
      "loss": 0.0829,
      "step": 296
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.23709601163864136,
      "learning_rate": 2.790329370339501e-05,
      "loss": 0.0693,
      "step": 297
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.23073746263980865,
      "learning_rate": 2.7889010398888104e-05,
      "loss": 0.0657,
      "step": 298
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.35000571608543396,
      "learning_rate": 2.7874682290113514e-05,
      "loss": 0.0774,
      "step": 299
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.2956024408340454,
      "learning_rate": 2.786030942687805e-05,
      "loss": 0.0836,
      "step": 300
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.35490185022354126,
      "learning_rate": 2.7845891859144088e-05,
      "loss": 0.1066,
      "step": 301
    },
    {
      "epoch": 0.53,
      "grad_norm": 0.5659674406051636,
      "learning_rate": 2.7831429637029402e-05,
      "loss": 0.1574,
      "step": 302
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.3070801794528961,
      "learning_rate": 2.7816922810807e-05,
      "loss": 0.0913,
      "step": 303
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.3739650547504425,
      "learning_rate": 2.7802371430904936e-05,
      "loss": 0.1343,
      "step": 304
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.6166901588439941,
      "learning_rate": 2.7787775547906142e-05,
      "loss": 0.0998,
      "step": 305
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.3676758408546448,
      "learning_rate": 2.7773135212548247e-05,
      "loss": 0.0673,
      "step": 306
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.4253583550453186,
      "learning_rate": 2.7758450475723405e-05,
      "loss": 0.1059,
      "step": 307
    },
    {
      "epoch": 0.54,
      "grad_norm": 0.31137579679489136,
      "learning_rate": 2.774372138847812e-05,
      "loss": 0.0884,
      "step": 308
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.23509562015533447,
      "learning_rate": 2.7728948002013054e-05,
      "loss": 0.0792,
      "step": 309
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.459160715341568,
      "learning_rate": 2.771413036768288e-05,
      "loss": 0.1156,
      "step": 310
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.34745368361473083,
      "learning_rate": 2.769926853699606e-05,
      "loss": 0.0971,
      "step": 311
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.30285558104515076,
      "learning_rate": 2.7684362561614714e-05,
      "loss": 0.0954,
      "step": 312
    },
    {
      "epoch": 0.55,
      "grad_norm": 0.30963224172592163,
      "learning_rate": 2.766941249335439e-05,
      "loss": 0.099,
      "step": 313
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.34839507937431335,
      "learning_rate": 2.765441838418393e-05,
      "loss": 0.1011,
      "step": 314
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.23209823668003082,
      "learning_rate": 2.7639380286225264e-05,
      "loss": 0.0662,
      "step": 315
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.3385458290576935,
      "learning_rate": 2.7624298251753232e-05,
      "loss": 0.1058,
      "step": 316
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.2462904453277588,
      "learning_rate": 2.7609172333195406e-05,
      "loss": 0.1128,
      "step": 317
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.3830464482307434,
      "learning_rate": 2.75940025831319e-05,
      "loss": 0.0927,
      "step": 318
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.23960565030574799,
      "learning_rate": 2.7578789054295202e-05,
      "loss": 0.0484,
      "step": 319
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.24503108859062195,
      "learning_rate": 2.7563531799569984e-05,
      "loss": 0.1001,
      "step": 320
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.3050799071788788,
      "learning_rate": 2.75482308719929e-05,
      "loss": 0.1168,
      "step": 321
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.4061020016670227,
      "learning_rate": 2.753288632475244e-05,
      "loss": 0.1424,
      "step": 322
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.30394184589385986,
      "learning_rate": 2.75174982111887e-05,
      "loss": 0.0774,
      "step": 323
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.22858074307441711,
      "learning_rate": 2.7502066584793243e-05,
      "loss": 0.0911,
      "step": 324
    },
    {
      "epoch": 0.57,
      "grad_norm": 0.31271883845329285,
      "learning_rate": 2.7486591499208867e-05,
      "loss": 0.0555,
      "step": 325
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.22895589470863342,
      "learning_rate": 2.7471073008229462e-05,
      "loss": 0.0741,
      "step": 326
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.22698134183883667,
      "learning_rate": 2.7455511165799783e-05,
      "loss": 0.0829,
      "step": 327
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.24353599548339844,
      "learning_rate": 2.7439906026015297e-05,
      "loss": 0.0965,
      "step": 328
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.19678641855716705,
      "learning_rate": 2.742425764312197e-05,
      "loss": 0.0822,
      "step": 329
    },
    {
      "epoch": 0.58,
      "grad_norm": 0.21248051524162292,
      "learning_rate": 2.7408566071516087e-05,
      "loss": 0.0713,
      "step": 330
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.47786301374435425,
      "learning_rate": 2.7392831365744074e-05,
      "loss": 0.0863,
      "step": 331
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.2300809770822525,
      "learning_rate": 2.7377053580502298e-05,
      "loss": 0.0864,
      "step": 332
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.34608194231987,
      "learning_rate": 2.7361232770636856e-05,
      "loss": 0.0869,
      "step": 333
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.37124645709991455,
      "learning_rate": 2.7345368991143433e-05,
      "loss": 0.1013,
      "step": 334
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.4118335545063019,
      "learning_rate": 2.732946229716707e-05,
      "loss": 0.1163,
      "step": 335
    },
    {
      "epoch": 0.59,
      "grad_norm": 0.5510608553886414,
      "learning_rate": 2.7313512744001982e-05,
      "loss": 0.1305,
      "step": 336
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.21711857616901398,
      "learning_rate": 2.7297520387091376e-05,
      "loss": 0.0668,
      "step": 337
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.24401220679283142,
      "learning_rate": 2.7281485282027252e-05,
      "loss": 0.0855,
      "step": 338
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.33350592851638794,
      "learning_rate": 2.7265407484550206e-05,
      "loss": 0.0632,
      "step": 339
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.7973533868789673,
      "learning_rate": 2.724928705054924e-05,
      "loss": 0.1894,
      "step": 340
    },
    {
      "epoch": 0.6,
      "grad_norm": 0.4574988782405853,
      "learning_rate": 2.7233124036061575e-05,
      "loss": 0.105,
      "step": 341
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.2864499092102051,
      "learning_rate": 2.7216918497272426e-05,
      "loss": 0.0826,
      "step": 342
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.2666662335395813,
      "learning_rate": 2.7200670490514865e-05,
      "loss": 0.0723,
      "step": 343
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.27098262310028076,
      "learning_rate": 2.7184380072269558e-05,
      "loss": 0.0579,
      "step": 344
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.22073043882846832,
      "learning_rate": 2.7168047299164614e-05,
      "loss": 0.0539,
      "step": 345
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.2872254550457001,
      "learning_rate": 2.7151672227975377e-05,
      "loss": 0.1239,
      "step": 346
    },
    {
      "epoch": 0.61,
      "grad_norm": 0.41940388083457947,
      "learning_rate": 2.7135254915624213e-05,
      "loss": 0.0755,
      "step": 347
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.2658102214336395,
      "learning_rate": 2.711879541918034e-05,
      "loss": 0.1158,
      "step": 348
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.23337627947330475,
      "learning_rate": 2.71022937958596e-05,
      "loss": 0.0814,
      "step": 349
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.26450300216674805,
      "learning_rate": 2.7085750103024296e-05,
      "loss": 0.0393,
      "step": 350
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.34251630306243896,
      "learning_rate": 2.7069164398182948e-05,
      "loss": 0.0718,
      "step": 351
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.3174653947353363,
      "learning_rate": 2.7052536738990125e-05,
      "loss": 0.0838,
      "step": 352
    },
    {
      "epoch": 0.62,
      "grad_norm": 0.2430477887392044,
      "learning_rate": 2.7035867183246247e-05,
      "loss": 0.1018,
      "step": 353
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.3248900771141052,
      "learning_rate": 2.7019155788897357e-05,
      "loss": 0.0764,
      "step": 354
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.24875646829605103,
      "learning_rate": 2.700240261403494e-05,
      "loss": 0.0731,
      "step": 355
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.209304079413414,
      "learning_rate": 2.6985607716895727e-05,
      "loss": 0.0482,
      "step": 356
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.24239419400691986,
      "learning_rate": 2.6968771155861464e-05,
      "loss": 0.0515,
      "step": 357
    },
    {
      "epoch": 0.63,
      "grad_norm": 0.5304663777351379,
      "learning_rate": 2.695189298945875e-05,
      "loss": 0.0919,
      "step": 358
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.21653291583061218,
      "learning_rate": 2.6934973276358792e-05,
      "loss": 0.0604,
      "step": 359
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.2541111409664154,
      "learning_rate": 2.6918012075377226e-05,
      "loss": 0.0823,
      "step": 360
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.3744794428348541,
      "learning_rate": 2.6901009445473912e-05,
      "loss": 0.1051,
      "step": 361
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.28307440876960754,
      "learning_rate": 2.6883965445752714e-05,
      "loss": 0.0425,
      "step": 362
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.49167400598526,
      "learning_rate": 2.6866880135461314e-05,
      "loss": 0.1167,
      "step": 363
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.3999198079109192,
      "learning_rate": 2.684975357399099e-05,
      "loss": 0.0547,
      "step": 364
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.23400184512138367,
      "learning_rate": 2.683258582087641e-05,
      "loss": 0.0545,
      "step": 365
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.1796039193868637,
      "learning_rate": 2.681537693579545e-05,
      "loss": 0.0346,
      "step": 366
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.5400248169898987,
      "learning_rate": 2.6798126978568942e-05,
      "loss": 0.1084,
      "step": 367
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.3125457167625427,
      "learning_rate": 2.6780836009160514e-05,
      "loss": 0.0445,
      "step": 368
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.48360738158226013,
      "learning_rate": 2.6763504087676346e-05,
      "loss": 0.0612,
      "step": 369
    },
    {
      "epoch": 0.65,
      "grad_norm": 0.9248375296592712,
      "learning_rate": 2.674613127436498e-05,
      "loss": 0.1119,
      "step": 370
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.4848339855670929,
      "learning_rate": 2.6728717629617093e-05,
      "loss": 0.1119,
      "step": 371
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.8126673102378845,
      "learning_rate": 2.671126321396532e-05,
      "loss": 0.1964,
      "step": 372
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.3666157126426697,
      "learning_rate": 2.6693768088083994e-05,
      "loss": 0.0688,
      "step": 373
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.33919093012809753,
      "learning_rate": 2.6676232312788998e-05,
      "loss": 0.087,
      "step": 374
    },
    {
      "epoch": 0.66,
      "grad_norm": 0.4218893349170685,
      "learning_rate": 2.6658655949037482e-05,
      "loss": 0.0704,
      "step": 375
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.2786632776260376,
      "learning_rate": 2.6641039057927724e-05,
      "loss": 0.0724,
      "step": 376
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.7727587819099426,
      "learning_rate": 2.662338170069885e-05,
      "loss": 0.1219,
      "step": 377
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.47525373101234436,
      "learning_rate": 2.6605683938730666e-05,
      "loss": 0.0894,
      "step": 378
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.6987584233283997,
      "learning_rate": 2.6587945833543432e-05,
      "loss": 0.0976,
      "step": 379
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.28406238555908203,
      "learning_rate": 2.6570167446797657e-05,
      "loss": 0.1048,
      "step": 380
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.2821677625179291,
      "learning_rate": 2.6552348840293856e-05,
      "loss": 0.0677,
      "step": 381
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.37019118666648865,
      "learning_rate": 2.6534490075972368e-05,
      "loss": 0.0901,
      "step": 382
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.3454105257987976,
      "learning_rate": 2.6516591215913118e-05,
      "loss": 0.0956,
      "step": 383
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.2895926535129547,
      "learning_rate": 2.6498652322335416e-05,
      "loss": 0.0539,
      "step": 384
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.4975215494632721,
      "learning_rate": 2.6480673457597737e-05,
      "loss": 0.0979,
      "step": 385
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.6093487739562988,
      "learning_rate": 2.646265468419749e-05,
      "loss": 0.146,
      "step": 386
    },
    {
      "epoch": 0.68,
      "grad_norm": 0.3748679459095001,
      "learning_rate": 2.6444596064770837e-05,
      "loss": 0.131,
      "step": 387
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.3992125391960144,
      "learning_rate": 2.6426497662092424e-05,
      "loss": 0.08,
      "step": 388
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.18994563817977905,
      "learning_rate": 2.6408359539075204e-05,
      "loss": 0.064,
      "step": 389
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.2519935965538025,
      "learning_rate": 2.6390181758770208e-05,
      "loss": 0.1035,
      "step": 390
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.1836027204990387,
      "learning_rate": 2.6371964384366305e-05,
      "loss": 0.0876,
      "step": 391
    },
    {
      "epoch": 0.69,
      "grad_norm": 0.21602681279182434,
      "learning_rate": 2.6353707479190022e-05,
      "loss": 0.0764,
      "step": 392
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.28096359968185425,
      "learning_rate": 2.6335411106705283e-05,
      "loss": 0.0598,
      "step": 393
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.3226209878921509,
      "learning_rate": 2.6317075330513212e-05,
      "loss": 0.1071,
      "step": 394
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.3776714503765106,
      "learning_rate": 2.6298700214351922e-05,
      "loss": 0.0576,
      "step": 395
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.5143963694572449,
      "learning_rate": 2.628028582209625e-05,
      "loss": 0.0927,
      "step": 396
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.5534598231315613,
      "learning_rate": 2.626183221775758e-05,
      "loss": 0.0813,
      "step": 397
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.3283978998661041,
      "learning_rate": 2.6243339465483605e-05,
      "loss": 0.091,
      "step": 398
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.26208245754241943,
      "learning_rate": 2.6224807629558094e-05,
      "loss": 0.0825,
      "step": 399
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.4244028925895691,
      "learning_rate": 2.6206236774400684e-05,
      "loss": 0.1077,
      "step": 400
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.39094752073287964,
      "learning_rate": 2.6187626964566644e-05,
      "loss": 0.1031,
      "step": 401
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.4204391539096832,
      "learning_rate": 2.6168978264746663e-05,
      "loss": 0.1153,
      "step": 402
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.3522421419620514,
      "learning_rate": 2.615029073976661e-05,
      "loss": 0.055,
      "step": 403
    },
    {
      "epoch": 0.71,
      "grad_norm": 0.28039833903312683,
      "learning_rate": 2.6131564454587316e-05,
      "loss": 0.1058,
      "step": 404
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.3698963224887848,
      "learning_rate": 2.611279947430436e-05,
      "loss": 0.0639,
      "step": 405
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.23384957015514374,
      "learning_rate": 2.609399586414782e-05,
      "loss": 0.0443,
      "step": 406
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.24013498425483704,
      "learning_rate": 2.607515368948206e-05,
      "loss": 0.0607,
      "step": 407
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.6072742342948914,
      "learning_rate": 2.60562730158055e-05,
      "loss": 0.1062,
      "step": 408
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.5543622970581055,
      "learning_rate": 2.6037353908750394e-05,
      "loss": 0.108,
      "step": 409
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.2501017153263092,
      "learning_rate": 2.601839643408259e-05,
      "loss": 0.0445,
      "step": 410
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.3118091821670532,
      "learning_rate": 2.5999400657701314e-05,
      "loss": 0.0982,
      "step": 411
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.424712210893631,
      "learning_rate": 2.598036664563893e-05,
      "loss": 0.0884,
      "step": 412
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.18902163207530975,
      "learning_rate": 2.596129446406072e-05,
      "loss": 0.0453,
      "step": 413
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.4779305160045624,
      "learning_rate": 2.594218417926464e-05,
      "loss": 0.1217,
      "step": 414
    },
    {
      "epoch": 0.73,
      "grad_norm": 0.3665257394313812,
      "learning_rate": 2.592303585768111e-05,
      "loss": 0.0791,
      "step": 415
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.22645382583141327,
      "learning_rate": 2.590384956587277e-05,
      "loss": 0.0384,
      "step": 416
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.3394840359687805,
      "learning_rate": 2.5884625370534242e-05,
      "loss": 0.0853,
      "step": 417
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.5719820857048035,
      "learning_rate": 2.5865363338491916e-05,
      "loss": 0.1058,
      "step": 418
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.5341434478759766,
      "learning_rate": 2.5846063536703706e-05,
      "loss": 0.0955,
      "step": 419
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.27542081475257874,
      "learning_rate": 2.582672603225882e-05,
      "loss": 0.0547,
      "step": 420
    },
    {
      "epoch": 0.74,
      "grad_norm": 0.4750405550003052,
      "learning_rate": 2.5807350892377517e-05,
      "loss": 0.152,
      "step": 421
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.292294979095459,
      "learning_rate": 2.5787938184410902e-05,
      "loss": 0.0877,
      "step": 422
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.33273130655288696,
      "learning_rate": 2.5768487975840655e-05,
      "loss": 0.0721,
      "step": 423
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.27887821197509766,
      "learning_rate": 2.5749000334278825e-05,
      "loss": 0.0603,
      "step": 424
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.35635143518447876,
      "learning_rate": 2.572947532746758e-05,
      "loss": 0.1011,
      "step": 425
    },
    {
      "epoch": 0.75,
      "grad_norm": 0.5215089321136475,
      "learning_rate": 2.570991302327897e-05,
      "loss": 0.1295,
      "step": 426
    },
    {
      "epoch": 0.75,
      "eval_loss": 0.07854457199573517,
      "eval_runtime": 14.6749,
      "eval_samples_per_second": 32.505,
      "eval_steps_per_second": 8.177,
      "step": 426
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.3576556444168091,
      "learning_rate": 2.569031348971471e-05,
      "loss": 0.0801,
      "step": 427
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.36309999227523804,
      "learning_rate": 2.5670676794905915e-05,
      "loss": 0.0694,
      "step": 428
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.6117097735404968,
      "learning_rate": 2.5651003007112892e-05,
      "loss": 0.111,
      "step": 429
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.217342346906662,
      "learning_rate": 2.5631292194724883e-05,
      "loss": 0.0346,
      "step": 430
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.4748762249946594,
      "learning_rate": 2.561154442625983e-05,
      "loss": 0.1427,
      "step": 431
    },
    {
      "epoch": 0.76,
      "grad_norm": 0.295284241437912,
      "learning_rate": 2.559175977036415e-05,
      "loss": 0.083,
      "step": 432
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.29578328132629395,
      "learning_rate": 2.5571938295812476e-05,
      "loss": 0.06,
      "step": 433
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.22991441190242767,
      "learning_rate": 2.555208007150743e-05,
      "loss": 0.092,
      "step": 434
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.23324204981327057,
      "learning_rate": 2.553218516647939e-05,
      "loss": 0.0972,
      "step": 435
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.40380072593688965,
      "learning_rate": 2.5512253649886237e-05,
      "loss": 0.0752,
      "step": 436
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.3805236518383026,
      "learning_rate": 2.5492285591013118e-05,
      "loss": 0.0998,
      "step": 437
    },
    {
      "epoch": 0.77,
      "grad_norm": 0.22794733941555023,
      "learning_rate": 2.5472281059272213e-05,
      "loss": 0.0747,
      "step": 438
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.22835515439510345,
      "learning_rate": 2.5452240124202477e-05,
      "loss": 0.0797,
      "step": 439
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.27887535095214844,
      "learning_rate": 2.5432162855469422e-05,
      "loss": 0.0499,
      "step": 440
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.2706398069858551,
      "learning_rate": 2.5412049322864847e-05,
      "loss": 0.052,
      "step": 441
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.22984421253204346,
      "learning_rate": 2.539189959630662e-05,
      "loss": 0.0735,
      "step": 442
    },
    {
      "epoch": 0.78,
      "grad_norm": 0.21597078442573547,
      "learning_rate": 2.537171374583843e-05,
      "loss": 0.0619,
      "step": 443
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.1989249587059021,
      "learning_rate": 2.535149184162952e-05,
      "loss": 0.0438,
      "step": 444
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.5477291345596313,
      "learning_rate": 2.533123395397448e-05,
      "loss": 0.1213,
      "step": 445
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.2946613132953644,
      "learning_rate": 2.5310940153292978e-05,
      "loss": 0.0751,
      "step": 446
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.3890973627567291,
      "learning_rate": 2.5290610510129518e-05,
      "loss": 0.0904,
      "step": 447
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.4777461588382721,
      "learning_rate": 2.5270245095153198e-05,
      "loss": 0.0436,
      "step": 448
    },
    {
      "epoch": 0.79,
      "grad_norm": 0.3443584442138672,
      "learning_rate": 2.524984397915747e-05,
      "loss": 0.0327,
      "step": 449
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.45741167664527893,
      "learning_rate": 2.5229407233059887e-05,
      "loss": 0.1213,
      "step": 450
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.3687816858291626,
      "learning_rate": 2.5208934927901857e-05,
      "loss": 0.1082,
      "step": 451
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.46452489495277405,
      "learning_rate": 2.5188427134848395e-05,
      "loss": 0.0841,
      "step": 452
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.23272164165973663,
      "learning_rate": 2.5167883925187878e-05,
      "loss": 0.0564,
      "step": 453
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.48221856355667114,
      "learning_rate": 2.51473053703318e-05,
      "loss": 0.12,
      "step": 454
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.5240137577056885,
      "learning_rate": 2.5126691541814518e-05,
      "loss": 0.0835,
      "step": 455
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.48311519622802734,
      "learning_rate": 2.510604251129301e-05,
      "loss": 0.1015,
      "step": 456
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.29986709356307983,
      "learning_rate": 2.5085358350546612e-05,
      "loss": 0.0827,
      "step": 457
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.4454290270805359,
      "learning_rate": 2.506463913147679e-05,
      "loss": 0.1309,
      "step": 458
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.305890828371048,
      "learning_rate": 2.5043884926106873e-05,
      "loss": 0.0636,
      "step": 459
    },
    {
      "epoch": 0.81,
      "grad_norm": 0.25751793384552,
      "learning_rate": 2.5023095806581802e-05,
      "loss": 0.0857,
      "step": 460
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.4034629166126251,
      "learning_rate": 2.5002271845167896e-05,
      "loss": 0.1383,
      "step": 461
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.2389378845691681,
      "learning_rate": 2.4981413114252588e-05,
      "loss": 0.0411,
      "step": 462
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.4818887412548065,
      "learning_rate": 2.4960519686344168e-05,
      "loss": 0.0837,
      "step": 463
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.32815420627593994,
      "learning_rate": 2.4939591634071544e-05,
      "loss": 0.0903,
      "step": 464
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.1801011711359024,
      "learning_rate": 2.491862903018398e-05,
      "loss": 0.0827,
      "step": 465
    },
    {
      "epoch": 0.82,
      "grad_norm": 0.23406219482421875,
      "learning_rate": 2.4897631947550857e-05,
      "loss": 0.0887,
      "step": 466
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.20487089455127716,
      "learning_rate": 2.4876600459161397e-05,
      "loss": 0.0944,
      "step": 467
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.2795143127441406,
      "learning_rate": 2.4855534638124427e-05,
      "loss": 0.0721,
      "step": 468
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.2514597475528717,
      "learning_rate": 2.4834434557668126e-05,
      "loss": 0.0798,
      "step": 469
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.24707554280757904,
      "learning_rate": 2.4813300291139754e-05,
      "loss": 0.0875,
      "step": 470
    },
    {
      "epoch": 0.83,
      "grad_norm": 0.2397790551185608,
      "learning_rate": 2.4792131912005407e-05,
      "loss": 0.0497,
      "step": 471
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.4203518331050873,
      "learning_rate": 2.4770929493849773e-05,
      "loss": 0.0789,
      "step": 472
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.23552100360393524,
      "learning_rate": 2.4749693110375856e-05,
      "loss": 0.0682,
      "step": 473
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.19143146276474,
      "learning_rate": 2.4728422835404735e-05,
      "loss": 0.0779,
      "step": 474
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.2365415096282959,
      "learning_rate": 2.4707118742875293e-05,
      "loss": 0.0442,
      "step": 475
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.36645641922950745,
      "learning_rate": 2.4685780906843975e-05,
      "loss": 0.1043,
      "step": 476
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.20781370997428894,
      "learning_rate": 2.4664409401484522e-05,
      "loss": 0.0797,
      "step": 477
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.27483153343200684,
      "learning_rate": 2.4643004301087716e-05,
      "loss": 0.0455,
      "step": 478
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.2820053696632385,
      "learning_rate": 2.462156568006112e-05,
      "loss": 0.0615,
      "step": 479
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.3784559369087219,
      "learning_rate": 2.4600093612928812e-05,
      "loss": 0.1338,
      "step": 480
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.46063724160194397,
      "learning_rate": 2.457858817433115e-05,
      "loss": 0.0897,
      "step": 481
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.320568323135376,
      "learning_rate": 2.4557049439024488e-05,
      "loss": 0.0747,
      "step": 482
    },
    {
      "epoch": 0.85,
      "grad_norm": 0.2996703088283539,
      "learning_rate": 2.4535477481880923e-05,
      "loss": 0.1219,
      "step": 483
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.46839097142219543,
      "learning_rate": 2.451387237788804e-05,
      "loss": 0.1104,
      "step": 484
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.3005930185317993,
      "learning_rate": 2.4492234202148643e-05,
      "loss": 0.0755,
      "step": 485
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.2897649109363556,
      "learning_rate": 2.44705630298805e-05,
      "loss": 0.0798,
      "step": 486
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.2890012860298157,
      "learning_rate": 2.4448858936416093e-05,
      "loss": 0.1019,
      "step": 487
    },
    {
      "epoch": 0.86,
      "grad_norm": 0.2596171200275421,
      "learning_rate": 2.442712199720232e-05,
      "loss": 0.0835,
      "step": 488
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.4037571847438812,
      "learning_rate": 2.4405352287800268e-05,
      "loss": 0.0822,
      "step": 489
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.337190181016922,
      "learning_rate": 2.4383549883884954e-05,
      "loss": 0.0356,
      "step": 490
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.3917272686958313,
      "learning_rate": 2.4361714861245017e-05,
      "loss": 0.0966,
      "step": 491
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.5287500023841858,
      "learning_rate": 2.4339847295782508e-05,
      "loss": 0.0706,
      "step": 492
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.29515090584754944,
      "learning_rate": 2.431794726351258e-05,
      "loss": 0.054,
      "step": 493
    },
    {
      "epoch": 0.87,
      "grad_norm": 0.30174991488456726,
      "learning_rate": 2.4296014840563266e-05,
      "loss": 0.0768,
      "step": 494
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.33137279748916626,
      "learning_rate": 2.4274050103175195e-05,
      "loss": 0.1008,
      "step": 495
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.19129472970962524,
      "learning_rate": 2.42520531277013e-05,
      "loss": 0.0556,
      "step": 496
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.3620724081993103,
      "learning_rate": 2.423002399060661e-05,
      "loss": 0.0604,
      "step": 497
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.28499534726142883,
      "learning_rate": 2.420796276846793e-05,
      "loss": 0.067,
      "step": 498
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.3519887328147888,
      "learning_rate": 2.4185869537973613e-05,
      "loss": 0.0973,
      "step": 499
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.34655264019966125,
      "learning_rate": 2.4163744375923272e-05,
      "loss": 0.0425,
      "step": 500
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.5716597437858582,
      "learning_rate": 2.4141587359227514e-05,
      "loss": 0.1659,
      "step": 501
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.32810407876968384,
      "learning_rate": 2.4119398564907685e-05,
      "loss": 0.116,
      "step": 502
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.32078537344932556,
      "learning_rate": 2.4097178070095602e-05,
      "loss": 0.0646,
      "step": 503
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.389475554227829,
      "learning_rate": 2.4074925952033263e-05,
      "loss": 0.1275,
      "step": 504
    },
    {
      "epoch": 0.89,
      "grad_norm": 0.15961997210979462,
      "learning_rate": 2.4052642288072594e-05,
      "loss": 0.0427,
      "step": 505
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.2679911255836487,
      "learning_rate": 2.4030327155675192e-05,
      "loss": 0.0868,
      "step": 506
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.2590488791465759,
      "learning_rate": 2.4007980632412034e-05,
      "loss": 0.0476,
      "step": 507
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.4470207989215851,
      "learning_rate": 2.398560279596323e-05,
      "loss": 0.0644,
      "step": 508
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.2822036147117615,
      "learning_rate": 2.3963193724117715e-05,
      "loss": 0.06,
      "step": 509
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.3097975552082062,
      "learning_rate": 2.394075349477302e-05,
      "loss": 0.0481,
      "step": 510
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.30676892399787903,
      "learning_rate": 2.3918282185934986e-05,
      "loss": 0.0794,
      "step": 511
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.23650003969669342,
      "learning_rate": 2.3895779875717486e-05,
      "loss": 0.0653,
      "step": 512
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.2657279968261719,
      "learning_rate": 2.3873246642342163e-05,
      "loss": 0.093,
      "step": 513
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.24548004567623138,
      "learning_rate": 2.3850682564138145e-05,
      "loss": 0.0659,
      "step": 514
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.5615881085395813,
      "learning_rate": 2.3828087719541787e-05,
      "loss": 0.1278,
      "step": 515
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.27736374735832214,
      "learning_rate": 2.3805462187096402e-05,
      "loss": 0.0575,
      "step": 516
    },
    {
      "epoch": 0.91,
      "grad_norm": 0.24056284129619598,
      "learning_rate": 2.3782806045451963e-05,
      "loss": 0.0452,
      "step": 517
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.4982932507991791,
      "learning_rate": 2.376011937336485e-05,
      "loss": 0.1689,
      "step": 518
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.29784631729125977,
      "learning_rate": 2.373740224969758e-05,
      "loss": 0.0938,
      "step": 519
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.43975648283958435,
      "learning_rate": 2.371465475341852e-05,
      "loss": 0.1245,
      "step": 520
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.30927592515945435,
      "learning_rate": 2.369187696360161e-05,
      "loss": 0.0619,
      "step": 521
    },
    {
      "epoch": 0.92,
      "grad_norm": 0.36457571387290955,
      "learning_rate": 2.3669068959426107e-05,
      "loss": 0.0859,
      "step": 522
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.25598183274269104,
      "learning_rate": 2.364623082017629e-05,
      "loss": 0.0461,
      "step": 523
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.44431087374687195,
      "learning_rate": 2.3623362625241193e-05,
      "loss": 0.1439,
      "step": 524
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.24602153897285461,
      "learning_rate": 2.3600464454114326e-05,
      "loss": 0.0408,
      "step": 525
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.2084829956293106,
      "learning_rate": 2.3577536386393416e-05,
      "loss": 0.0856,
      "step": 526
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.22140052914619446,
      "learning_rate": 2.35545785017801e-05,
      "loss": 0.0721,
      "step": 527
    },
    {
      "epoch": 0.93,
      "grad_norm": 0.24179723858833313,
      "learning_rate": 2.3531590880079663e-05,
      "loss": 0.0909,
      "step": 528
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.3095971345901489,
      "learning_rate": 2.3508573601200767e-05,
      "loss": 0.0804,
      "step": 529
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.2845352292060852,
      "learning_rate": 2.348552674515517e-05,
      "loss": 0.0462,
      "step": 530
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.4917078912258148,
      "learning_rate": 2.3462450392057437e-05,
      "loss": 0.1184,
      "step": 531
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.33279237151145935,
      "learning_rate": 2.343934462212467e-05,
      "loss": 0.0531,
      "step": 532
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.33190539479255676,
      "learning_rate": 2.341620951567624e-05,
      "loss": 0.1114,
      "step": 533
    },
    {
      "epoch": 0.94,
      "grad_norm": 0.5710054039955139,
      "learning_rate": 2.339304515313348e-05,
      "loss": 0.1531,
      "step": 534
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.20240336656570435,
      "learning_rate": 2.3369851615019436e-05,
      "loss": 0.0624,
      "step": 535
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.4915199279785156,
      "learning_rate": 2.3346628981958565e-05,
      "loss": 0.1463,
      "step": 536
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.29593560099601746,
      "learning_rate": 2.332337733467646e-05,
      "loss": 0.0789,
      "step": 537
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.1819988489151001,
      "learning_rate": 2.3300096753999585e-05,
      "loss": 0.0724,
      "step": 538
    },
    {
      "epoch": 0.95,
      "grad_norm": 0.39005154371261597,
      "learning_rate": 2.3276787320854967e-05,
      "loss": 0.083,
      "step": 539
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.4818272590637207,
      "learning_rate": 2.3253449116269937e-05,
      "loss": 0.0906,
      "step": 540
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.3471298813819885,
      "learning_rate": 2.3230082221371834e-05,
      "loss": 0.0952,
      "step": 541
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.2572653591632843,
      "learning_rate": 2.3206686717387742e-05,
      "loss": 0.115,
      "step": 542
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.407372385263443,
      "learning_rate": 2.3183262685644177e-05,
      "loss": 0.1084,
      "step": 543
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.16921907663345337,
      "learning_rate": 2.3159810207566832e-05,
      "loss": 0.0694,
      "step": 544
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.2879027724266052,
      "learning_rate": 2.3136329364680288e-05,
      "loss": 0.0927,
      "step": 545
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.23834727704524994,
      "learning_rate": 2.3112820238607716e-05,
      "loss": 0.0772,
      "step": 546
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.29006141424179077,
      "learning_rate": 2.3089282911070613e-05,
      "loss": 0.0934,
      "step": 547
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.4761680066585541,
      "learning_rate": 2.3065717463888505e-05,
      "loss": 0.1083,
      "step": 548
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.2773534953594208,
      "learning_rate": 2.3042123978978665e-05,
      "loss": 0.0802,
      "step": 549
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.24215322732925415,
      "learning_rate": 2.3018502538355827e-05,
      "loss": 0.0661,
      "step": 550
    },
    {
      "epoch": 0.97,
      "grad_norm": 0.21285729110240936,
      "learning_rate": 2.2994853224131915e-05,
      "loss": 0.0709,
      "step": 551
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.1974431574344635,
      "learning_rate": 2.2971176118515734e-05,
      "loss": 0.0857,
      "step": 552
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.2114921659231186,
      "learning_rate": 2.2947471303812708e-05,
      "loss": 0.0715,
      "step": 553
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.23644877970218658,
      "learning_rate": 2.2923738862424565e-05,
      "loss": 0.086,
      "step": 554
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.2784081697463989,
      "learning_rate": 2.2899978876849084e-05,
      "loss": 0.0392,
      "step": 555
    },
    {
      "epoch": 0.98,
      "grad_norm": 0.20647476613521576,
      "learning_rate": 2.287619142967979e-05,
      "loss": 0.0479,
      "step": 556
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.4792695641517639,
      "learning_rate": 2.285237660360566e-05,
      "loss": 0.1084,
      "step": 557
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.32381659746170044,
      "learning_rate": 2.2828534481410847e-05,
      "loss": 0.0916,
      "step": 558
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.5321642160415649,
      "learning_rate": 2.28046651459744e-05,
      "loss": 0.155,
      "step": 559
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.2742505669593811,
      "learning_rate": 2.278076868026995e-05,
      "loss": 0.0725,
      "step": 560
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.40313002467155457,
      "learning_rate": 2.2756845167365452e-05,
      "loss": 0.0856,
      "step": 561
    },
    {
      "epoch": 0.99,
      "grad_norm": 0.2948872745037079,
      "learning_rate": 2.273289469042287e-05,
      "loss": 0.0646,
      "step": 562
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.23548893630504608,
      "learning_rate": 2.2708917332697908e-05,
      "loss": 0.0451,
      "step": 563
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.25398510694503784,
      "learning_rate": 2.26849131775397e-05,
      "loss": 0.0647,
      "step": 564
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.30981212854385376,
      "learning_rate": 2.2660882308390547e-05,
      "loss": 0.077,
      "step": 565
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.2093246728181839,
      "learning_rate": 2.2636824808785604e-05,
      "loss": 0.0541,
      "step": 566
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.6050600409507751,
      "learning_rate": 2.2612740762352594e-05,
      "loss": 0.1058,
      "step": 567
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.3822791576385498,
      "learning_rate": 2.258863025281153e-05,
      "loss": 0.0869,
      "step": 568
    },
    {
      "epoch": 1.0,
      "eval_loss": 0.08007431030273438,
      "eval_runtime": 14.6613,
      "eval_samples_per_second": 32.535,
      "eval_steps_per_second": 8.185,
      "step": 568
    },
    {
      "epoch": 1.01,
      "grad_norm": 0.28725308179855347,
      "learning_rate": 2.2564493363974416e-05,
      "loss": 0.0437,
      "step": 569
    },
    {
      "epoch": 1.01,
      "grad_norm": 0.4120897650718689,
      "learning_rate": 2.2540330179744935e-05,
      "loss": 0.0868,
      "step": 570
    },
    {
      "epoch": 1.01,
      "grad_norm": 0.4469630718231201,
      "learning_rate": 2.2516140784118202e-05,
      "loss": 0.1007,
      "step": 571
    },
    {
      "epoch": 1.01,
      "grad_norm": 0.5416034460067749,
      "learning_rate": 2.2491925261180433e-05,
      "loss": 0.0854,
      "step": 572
    },
    {
      "epoch": 1.01,
      "grad_norm": 0.468612939119339,
      "learning_rate": 2.2467683695108676e-05,
      "loss": 0.1247,
      "step": 573
    },
    {
      "epoch": 1.02,
      "grad_norm": 0.3557000458240509,
      "learning_rate": 2.244341617017049e-05,
      "loss": 0.1331,
      "step": 574
    },
    {
      "epoch": 1.02,
      "grad_norm": 0.20697084069252014,
      "learning_rate": 2.24191227707237e-05,
      "loss": 0.0621,
      "step": 575
    },
    {
      "epoch": 1.02,
      "grad_norm": 0.3448030650615692,
      "learning_rate": 2.2394803581216047e-05,
      "loss": 0.0572,
      "step": 576
    },
    {
      "epoch": 1.02,
      "grad_norm": 0.3418349027633667,
      "learning_rate": 2.2370458686184945e-05,
      "loss": 0.0748,
      "step": 577
    },
    {
      "epoch": 1.02,
      "grad_norm": 0.2847341001033783,
      "learning_rate": 2.234608817025715e-05,
      "loss": 0.0755,
      "step": 578
    },
    {
      "epoch": 1.02,
      "grad_norm": 0.25333938002586365,
      "learning_rate": 2.232169211814848e-05,
      "loss": 0.0705,
      "step": 579
    },
    {
      "epoch": 1.03,
      "grad_norm": 0.257980078458786,
      "learning_rate": 2.2297270614663535e-05,
      "loss": 0.0726,
      "step": 580
    },
    {
      "epoch": 1.03,
      "grad_norm": 0.2288191020488739,
      "learning_rate": 2.2272823744695366e-05,
      "loss": 0.0512,
      "step": 581
    },
    {
      "epoch": 1.03,
      "grad_norm": 0.37362760305404663,
      "learning_rate": 2.2248351593225225e-05,
      "loss": 0.071,
      "step": 582
    },
    {
      "epoch": 1.03,
      "grad_norm": 0.2675292193889618,
      "learning_rate": 2.2223854245322236e-05,
      "loss": 0.0749,
      "step": 583
    },
    {
      "epoch": 1.03,
      "grad_norm": 0.27135464549064636,
      "learning_rate": 2.21993317861431e-05,
      "loss": 0.0912,
      "step": 584
    },
    {
      "epoch": 1.03,
      "grad_norm": 0.28075793385505676,
      "learning_rate": 2.217478430093183e-05,
      "loss": 0.0531,
      "step": 585
    },
    {
      "epoch": 1.04,
      "grad_norm": 0.2495741993188858,
      "learning_rate": 2.2150211875019417e-05,
      "loss": 0.0827,
      "step": 586
    },
    {
      "epoch": 1.04,
      "grad_norm": 0.36625027656555176,
      "learning_rate": 2.2125614593823556e-05,
      "loss": 0.0536,
      "step": 587
    },
    {
      "epoch": 1.04,
      "grad_norm": 0.24689775705337524,
      "learning_rate": 2.2100992542848344e-05,
      "loss": 0.0524,
      "step": 588
    },
    {
      "epoch": 1.04,
      "grad_norm": 0.2930172383785248,
      "learning_rate": 2.2076345807683975e-05,
      "loss": 0.1101,
      "step": 589
    },
    {
      "epoch": 1.04,
      "grad_norm": 0.29739370942115784,
      "learning_rate": 2.2051674474006457e-05,
      "loss": 0.0446,
      "step": 590
    },
    {
      "epoch": 1.05,
      "grad_norm": 0.24752137064933777,
      "learning_rate": 2.2026978627577303e-05,
      "loss": 0.0662,
      "step": 591
    },
    {
      "epoch": 1.05,
      "grad_norm": 0.6093515753746033,
      "learning_rate": 2.2002258354243242e-05,
      "loss": 0.0631,
      "step": 592
    },
    {
      "epoch": 1.05,
      "grad_norm": 0.23734183609485626,
      "learning_rate": 2.1977513739935897e-05,
      "loss": 0.0353,
      "step": 593
    },
    {
      "epoch": 1.05,
      "grad_norm": 0.3490619957447052,
      "learning_rate": 2.195274487067153e-05,
      "loss": 0.0893,
      "step": 594
    },
    {
      "epoch": 1.05,
      "grad_norm": 0.2838170528411865,
      "learning_rate": 2.1927951832550696e-05,
      "loss": 0.0674,
      "step": 595
    },
    {
      "epoch": 1.05,
      "grad_norm": 0.38547414541244507,
      "learning_rate": 2.1903134711757974e-05,
      "loss": 0.1,
      "step": 596
    },
    {
      "epoch": 1.06,
      "grad_norm": 0.3761378228664398,
      "learning_rate": 2.187829359456166e-05,
      "loss": 0.1023,
      "step": 597
    },
    {
      "epoch": 1.06,
      "grad_norm": 0.3699404299259186,
      "learning_rate": 2.1853428567313455e-05,
      "loss": 0.0965,
      "step": 598
    },
    {
      "epoch": 1.06,
      "grad_norm": 0.21764816343784332,
      "learning_rate": 2.1828539716448188e-05,
      "loss": 0.0383,
      "step": 599
    },
    {
      "epoch": 1.06,
      "grad_norm": 0.265080988407135,
      "learning_rate": 2.1803627128483497e-05,
      "loss": 0.0402,
      "step": 600
    },
    {
      "epoch": 1.06,
      "grad_norm": 0.6008972525596619,
      "learning_rate": 2.1778690890019535e-05,
      "loss": 0.1595,
      "step": 601
    },
    {
      "epoch": 1.07,
      "grad_norm": 0.16610845923423767,
      "learning_rate": 2.1753731087738657e-05,
      "loss": 0.0217,
      "step": 602
    },
    {
      "epoch": 1.07,
      "grad_norm": 0.2897780239582062,
      "learning_rate": 2.1728747808405154e-05,
      "loss": 0.0718,
      "step": 603
    },
    {
      "epoch": 1.07,
      "grad_norm": 0.2148287147283554,
      "learning_rate": 2.1703741138864908e-05,
      "loss": 0.0607,
      "step": 604
    },
    {
      "epoch": 1.07,
      "grad_norm": 0.3162532150745392,
      "learning_rate": 2.167871116604511e-05,
      "loss": 0.0607,
      "step": 605
    },
    {
      "epoch": 1.07,
      "grad_norm": 0.3892384171485901,
      "learning_rate": 2.1653657976953956e-05,
      "loss": 0.1097,
      "step": 606
    },
    {
      "epoch": 1.07,
      "grad_norm": 0.41244956851005554,
      "learning_rate": 2.1628581658680358e-05,
      "loss": 0.0668,
      "step": 607
    },
    {
      "epoch": 1.08,
      "grad_norm": 0.260682612657547,
      "learning_rate": 2.1603482298393622e-05,
      "loss": 0.0657,
      "step": 608
    },
    {
      "epoch": 1.08,
      "grad_norm": 0.2950402796268463,
      "learning_rate": 2.157835998334315e-05,
      "loss": 0.0482,
      "step": 609
    },
    {
      "epoch": 1.08,
      "grad_norm": 0.5721485018730164,
      "learning_rate": 2.155321480085813e-05,
      "loss": 0.135,
      "step": 610
    },
    {
      "epoch": 1.08,
      "grad_norm": 0.3857111632823944,
      "learning_rate": 2.1528046838347262e-05,
      "loss": 0.0824,
      "step": 611
    },
    {
      "epoch": 1.08,
      "grad_norm": 0.3081589937210083,
      "learning_rate": 2.1502856183298414e-05,
      "loss": 0.0812,
      "step": 612
    },
    {
      "epoch": 1.08,
      "grad_norm": 0.1889428198337555,
      "learning_rate": 2.147764292327834e-05,
      "loss": 0.0386,
      "step": 613
    },
    {
      "epoch": 1.09,
      "grad_norm": 0.23867502808570862,
      "learning_rate": 2.145240714593239e-05,
      "loss": 0.0662,
      "step": 614
    },
    {
      "epoch": 1.09,
      "grad_norm": 0.43273600935935974,
      "learning_rate": 2.142714893898416e-05,
      "loss": 0.0741,
      "step": 615
    },
    {
      "epoch": 1.09,
      "grad_norm": 0.19105297327041626,
      "learning_rate": 2.1401868390235236e-05,
      "loss": 0.0692,
      "step": 616
    },
    {
      "epoch": 1.09,
      "grad_norm": 0.4356454610824585,
      "learning_rate": 2.137656558756486e-05,
      "loss": 0.0821,
      "step": 617
    },
    {
      "epoch": 1.09,
      "grad_norm": 0.515384316444397,
      "learning_rate": 2.1351240618929633e-05,
      "loss": 0.1168,
      "step": 618
    },
    {
      "epoch": 1.1,
      "grad_norm": 0.36919349431991577,
      "learning_rate": 2.1325893572363205e-05,
      "loss": 0.0879,
      "step": 619
    },
    {
      "epoch": 1.1,
      "grad_norm": 0.37004175782203674,
      "learning_rate": 2.130052453597598e-05,
      "loss": 0.0933,
      "step": 620
    },
    {
      "epoch": 1.1,
      "grad_norm": 0.35473090410232544,
      "learning_rate": 2.1275133597954794e-05,
      "loss": 0.0949,
      "step": 621
    },
    {
      "epoch": 1.1,
      "grad_norm": 0.26554247736930847,
      "learning_rate": 2.1249720846562626e-05,
      "loss": 0.061,
      "step": 622
    },
    {
      "epoch": 1.1,
      "grad_norm": 0.3732711672782898,
      "learning_rate": 2.122428637013827e-05,
      "loss": 0.0985,
      "step": 623
    },
    {
      "epoch": 1.1,
      "grad_norm": 0.20980623364448547,
      "learning_rate": 2.1198830257096055e-05,
      "loss": 0.1016,
      "step": 624
    },
    {
      "epoch": 1.11,
      "grad_norm": 0.2785292863845825,
      "learning_rate": 2.1173352595925505e-05,
      "loss": 0.0562,
      "step": 625
    },
    {
      "epoch": 1.11,
      "grad_norm": 0.1760614961385727,
      "learning_rate": 2.114785347519106e-05,
      "loss": 0.0421,
      "step": 626
    },
    {
      "epoch": 1.11,
      "grad_norm": 0.40503427386283875,
      "learning_rate": 2.112233298353175e-05,
      "loss": 0.0588,
      "step": 627
    },
    {
      "epoch": 1.11,
      "grad_norm": 0.512732982635498,
      "learning_rate": 2.1096791209660905e-05,
      "loss": 0.1067,
      "step": 628
    },
    {
      "epoch": 1.11,
      "grad_norm": 0.22243398427963257,
      "learning_rate": 2.1071228242365826e-05,
      "loss": 0.0331,
      "step": 629
    },
    {
      "epoch": 1.11,
      "grad_norm": 0.5802137851715088,
      "learning_rate": 2.1045644170507486e-05,
      "loss": 0.0995,
      "step": 630
    },
    {
      "epoch": 1.12,
      "grad_norm": 0.2316664159297943,
      "learning_rate": 2.1020039083020222e-05,
      "loss": 0.056,
      "step": 631
    },
    {
      "epoch": 1.12,
      "grad_norm": 0.26103687286376953,
      "learning_rate": 2.0994413068911425e-05,
      "loss": 0.0551,
      "step": 632
    },
    {
      "epoch": 1.12,
      "grad_norm": 0.35766682028770447,
      "learning_rate": 2.0968766217261236e-05,
      "loss": 0.0705,
      "step": 633
    },
    {
      "epoch": 1.12,
      "grad_norm": 0.2776765823364258,
      "learning_rate": 2.0943098617222216e-05,
      "loss": 0.0768,
      "step": 634
    },
    {
      "epoch": 1.12,
      "grad_norm": 0.38086697459220886,
      "learning_rate": 2.0917410358019073e-05,
      "loss": 0.069,
      "step": 635
    },
    {
      "epoch": 1.13,
      "grad_norm": 0.3982766568660736,
      "learning_rate": 2.0891701528948302e-05,
      "loss": 0.0557,
      "step": 636
    },
    {
      "epoch": 1.13,
      "grad_norm": 0.32681411504745483,
      "learning_rate": 2.0865972219377935e-05,
      "loss": 0.0993,
      "step": 637
    },
    {
      "epoch": 1.13,
      "grad_norm": 0.4559233784675598,
      "learning_rate": 2.0840222518747162e-05,
      "loss": 0.1082,
      "step": 638
    },
    {
      "epoch": 1.13,
      "grad_norm": 0.42872411012649536,
      "learning_rate": 2.081445251656608e-05,
      "loss": 0.0717,
      "step": 639
    },
    {
      "epoch": 1.13,
      "grad_norm": 0.25050288438796997,
      "learning_rate": 2.0788662302415352e-05,
      "loss": 0.0593,
      "step": 640
    },
    {
      "epoch": 1.13,
      "grad_norm": 0.7354700565338135,
      "learning_rate": 2.0762851965945894e-05,
      "loss": 0.1242,
      "step": 641
    },
    {
      "epoch": 1.14,
      "grad_norm": 0.4167981445789337,
      "learning_rate": 2.073702159687858e-05,
      "loss": 0.0865,
      "step": 642
    },
    {
      "epoch": 1.14,
      "grad_norm": 0.36572423577308655,
      "learning_rate": 2.0711171285003917e-05,
      "loss": 0.0574,
      "step": 643
    },
    {
      "epoch": 1.14,
      "grad_norm": 0.25673985481262207,
      "learning_rate": 2.0685301120181737e-05,
      "loss": 0.0661,
      "step": 644
    },
    {
      "epoch": 1.14,
      "grad_norm": 0.6029005646705627,
      "learning_rate": 2.0659411192340875e-05,
      "loss": 0.0907,
      "step": 645
    },
    {
      "epoch": 1.14,
      "grad_norm": 0.22739049792289734,
      "learning_rate": 2.0633501591478886e-05,
      "loss": 0.0559,
      "step": 646
    },
    {
      "epoch": 1.14,
      "grad_norm": 0.3935805559158325,
      "learning_rate": 2.0607572407661683e-05,
      "loss": 0.1237,
      "step": 647
    },
    {
      "epoch": 1.15,
      "grad_norm": 0.30537936091423035,
      "learning_rate": 2.058162373102328e-05,
      "loss": 0.0939,
      "step": 648
    },
    {
      "epoch": 1.15,
      "grad_norm": 0.3546753227710724,
      "learning_rate": 2.0555655651765436e-05,
      "loss": 0.0584,
      "step": 649
    },
    {
      "epoch": 1.15,
      "grad_norm": 0.3305690884590149,
      "learning_rate": 2.0529668260157354e-05,
      "loss": 0.0593,
      "step": 650
    },
    {
      "epoch": 1.15,
      "grad_norm": 0.1683047115802765,
      "learning_rate": 2.050366164653539e-05,
      "loss": 0.0302,
      "step": 651
    },
    {
      "epoch": 1.15,
      "grad_norm": 0.3908555209636688,
      "learning_rate": 2.047763590130269e-05,
      "loss": 0.1198,
      "step": 652
    },
    {
      "epoch": 1.16,
      "grad_norm": 0.3852555751800537,
      "learning_rate": 2.045159111492893e-05,
      "loss": 0.084,
      "step": 653
    },
    {
      "epoch": 1.16,
      "grad_norm": 0.27726736664772034,
      "learning_rate": 2.042552737794996e-05,
      "loss": 0.0397,
      "step": 654
    },
    {
      "epoch": 1.16,
      "grad_norm": 0.16915327310562134,
      "learning_rate": 2.039944478096751e-05,
      "loss": 0.027,
      "step": 655
    },
    {
      "epoch": 1.16,
      "grad_norm": 0.37171369791030884,
      "learning_rate": 2.037334341464888e-05,
      "loss": 0.0883,
      "step": 656
    },
    {
      "epoch": 1.16,
      "grad_norm": 0.4880152940750122,
      "learning_rate": 2.034722336972659e-05,
      "loss": 0.1114,
      "step": 657
    },
    {
      "epoch": 1.16,
      "grad_norm": 0.3333703279495239,
      "learning_rate": 2.032108473699812e-05,
      "loss": 0.0995,
      "step": 658
    },
    {
      "epoch": 1.17,
      "grad_norm": 0.33865606784820557,
      "learning_rate": 2.0294927607325542e-05,
      "loss": 0.0825,
      "step": 659
    },
    {
      "epoch": 1.17,
      "grad_norm": 0.4168679118156433,
      "learning_rate": 2.0268752071635233e-05,
      "loss": 0.0959,
      "step": 660
    },
    {
      "epoch": 1.17,
      "grad_norm": 0.14923910796642303,
      "learning_rate": 2.0242558220917573e-05,
      "loss": 0.0149,
      "step": 661
    },
    {
      "epoch": 1.17,
      "grad_norm": 0.2838289439678192,
      "learning_rate": 2.021634614622656e-05,
      "loss": 0.0652,
      "step": 662
    },
    {
      "epoch": 1.17,
      "grad_norm": 0.3029157519340515,
      "learning_rate": 2.0190115938679594e-05,
      "loss": 0.0907,
      "step": 663
    },
    {
      "epoch": 1.17,
      "grad_norm": 0.2661125063896179,
      "learning_rate": 2.0163867689457072e-05,
      "loss": 0.0405,
      "step": 664
    },
    {
      "epoch": 1.18,
      "grad_norm": 0.4163992702960968,
      "learning_rate": 2.0137601489802126e-05,
      "loss": 0.0981,
      "step": 665
    },
    {
      "epoch": 1.18,
      "grad_norm": 0.4169825315475464,
      "learning_rate": 2.0111317431020272e-05,
      "loss": 0.0811,
      "step": 666
    },
    {
      "epoch": 1.18,
      "grad_norm": 0.2571697235107422,
      "learning_rate": 2.0085015604479113e-05,
      "loss": 0.0416,
      "step": 667
    },
    {
      "epoch": 1.18,
      "grad_norm": 0.2569189965724945,
      "learning_rate": 2.005869610160803e-05,
      "loss": 0.038,
      "step": 668
    },
    {
      "epoch": 1.18,
      "grad_norm": 0.3448300361633301,
      "learning_rate": 2.0032359013897825e-05,
      "loss": 0.0674,
      "step": 669
    },
    {
      "epoch": 1.19,
      "grad_norm": 0.3533551096916199,
      "learning_rate": 2.0006004432900446e-05,
      "loss": 0.1156,
      "step": 670
    },
    {
      "epoch": 1.19,
      "grad_norm": 0.3890987038612366,
      "learning_rate": 1.9979632450228632e-05,
      "loss": 0.076,
      "step": 671
    },
    {
      "epoch": 1.19,
      "grad_norm": 0.23361268639564514,
      "learning_rate": 1.9953243157555648e-05,
      "loss": 0.0448,
      "step": 672
    },
    {
      "epoch": 1.19,
      "grad_norm": 0.36684730648994446,
      "learning_rate": 1.9926836646614888e-05,
      "loss": 0.0487,
      "step": 673
    },
    {
      "epoch": 1.19,
      "grad_norm": 0.48742666840553284,
      "learning_rate": 1.9900413009199627e-05,
      "loss": 0.0627,
      "step": 674
    },
    {
      "epoch": 1.19,
      "grad_norm": 0.19356118142604828,
      "learning_rate": 1.987397233716267e-05,
      "loss": 0.0251,
      "step": 675
    },
    {
      "epoch": 1.2,
      "grad_norm": 0.5242466330528259,
      "learning_rate": 1.9847514722416028e-05,
      "loss": 0.0956,
      "step": 676
    },
    {
      "epoch": 1.2,
      "grad_norm": 0.3173238933086395,
      "learning_rate": 1.982104025693062e-05,
      "loss": 0.0678,
      "step": 677
    },
    {
      "epoch": 1.2,
      "grad_norm": 1.015837550163269,
      "learning_rate": 1.9794549032735935e-05,
      "loss": 0.1514,
      "step": 678
    },
    {
      "epoch": 1.2,
      "grad_norm": 0.21766427159309387,
      "learning_rate": 1.9768041141919713e-05,
      "loss": 0.0291,
      "step": 679
    },
    {
      "epoch": 1.2,
      "grad_norm": 0.39488041400909424,
      "learning_rate": 1.9741516676627632e-05,
      "loss": 0.1148,
      "step": 680
    },
    {
      "epoch": 1.2,
      "grad_norm": 0.17024394869804382,
      "learning_rate": 1.9714975729063e-05,
      "loss": 0.0227,
      "step": 681
    },
    {
      "epoch": 1.21,
      "grad_norm": 0.48008298873901367,
      "learning_rate": 1.96884183914864e-05,
      "loss": 0.0762,
      "step": 682
    },
    {
      "epoch": 1.21,
      "grad_norm": 0.3254008889198303,
      "learning_rate": 1.96618447562154e-05,
      "loss": 0.0536,
      "step": 683
    },
    {
      "epoch": 1.21,
      "grad_norm": 0.26731014251708984,
      "learning_rate": 1.963525491562421e-05,
      "loss": 0.0431,
      "step": 684
    },
    {
      "epoch": 1.21,
      "grad_norm": 0.4722350537776947,
      "learning_rate": 1.9608648962143395e-05,
      "loss": 0.0869,
      "step": 685
    },
    {
      "epoch": 1.21,
      "grad_norm": 0.5004008412361145,
      "learning_rate": 1.958202698825951e-05,
      "loss": 0.1011,
      "step": 686
    },
    {
      "epoch": 1.22,
      "grad_norm": 0.6916111707687378,
      "learning_rate": 1.9555389086514813e-05,
      "loss": 0.1184,
      "step": 687
    },
    {
      "epoch": 1.22,
      "grad_norm": 0.1500881463289261,
      "learning_rate": 1.9528735349506906e-05,
      "loss": 0.018,
      "step": 688
    },
    {
      "epoch": 1.22,
      "grad_norm": 0.2581406235694885,
      "learning_rate": 1.9502065869888473e-05,
      "loss": 0.0692,
      "step": 689
    },
    {
      "epoch": 1.22,
      "grad_norm": 0.23035988211631775,
      "learning_rate": 1.94753807403669e-05,
      "loss": 0.0566,
      "step": 690
    },
    {
      "epoch": 1.22,
      "grad_norm": 0.3322215676307678,
      "learning_rate": 1.9448680053703974e-05,
      "loss": 0.0878,
      "step": 691
    },
    {
      "epoch": 1.22,
      "grad_norm": 0.3694436252117157,
      "learning_rate": 1.9421963902715565e-05,
      "loss": 0.1127,
      "step": 692
    },
    {
      "epoch": 1.23,
      "grad_norm": 0.5266421437263489,
      "learning_rate": 1.9395232380271292e-05,
      "loss": 0.0902,
      "step": 693
    },
    {
      "epoch": 1.23,
      "grad_norm": 0.44570016860961914,
      "learning_rate": 1.9368485579294233e-05,
      "loss": 0.0707,
      "step": 694
    },
    {
      "epoch": 1.23,
      "grad_norm": 0.4436816871166229,
      "learning_rate": 1.9341723592760542e-05,
      "loss": 0.1019,
      "step": 695
    },
    {
      "epoch": 1.23,
      "grad_norm": 0.6907861828804016,
      "learning_rate": 1.931494651369919e-05,
      "loss": 0.12,
      "step": 696
    },
    {
      "epoch": 1.23,
      "grad_norm": 0.2745947241783142,
      "learning_rate": 1.9288154435191583e-05,
      "loss": 0.0648,
      "step": 697
    },
    {
      "epoch": 1.23,
      "grad_norm": 0.3967859148979187,
      "learning_rate": 1.92613474503713e-05,
      "loss": 0.1285,
      "step": 698
    },
    {
      "epoch": 1.24,
      "grad_norm": 0.28435570001602173,
      "learning_rate": 1.9234525652423714e-05,
      "loss": 0.0837,
      "step": 699
    },
    {
      "epoch": 1.24,
      "grad_norm": 0.2751648724079132,
      "learning_rate": 1.9207689134585698e-05,
      "loss": 0.0548,
      "step": 700
    },
    {
      "epoch": 1.24,
      "grad_norm": 0.33161744475364685,
      "learning_rate": 1.918083799014529e-05,
      "loss": 0.0791,
      "step": 701
    },
    {
      "epoch": 1.24,
      "grad_norm": 0.21666060388088226,
      "learning_rate": 1.9153972312441374e-05,
      "loss": 0.0589,
      "step": 702
    },
    {
      "epoch": 1.24,
      "grad_norm": 0.23783162236213684,
      "learning_rate": 1.912709219486336e-05,
      "loss": 0.0652,
      "step": 703
    },
    {
      "epoch": 1.25,
      "grad_norm": 0.40111228823661804,
      "learning_rate": 1.910019773085085e-05,
      "loss": 0.089,
      "step": 704
    },
    {
      "epoch": 1.25,
      "grad_norm": 0.24063000082969666,
      "learning_rate": 1.907328901389331e-05,
      "loss": 0.0637,
      "step": 705
    },
    {
      "epoch": 1.25,
      "grad_norm": 0.292782723903656,
      "learning_rate": 1.904636613752976e-05,
      "loss": 0.1015,
      "step": 706
    },
    {
      "epoch": 1.25,
      "grad_norm": 0.4749392569065094,
      "learning_rate": 1.9019429195348433e-05,
      "loss": 0.1259,
      "step": 707
    },
    {
      "epoch": 1.25,
      "grad_norm": 0.31627988815307617,
      "learning_rate": 1.8992478280986468e-05,
      "loss": 0.1111,
      "step": 708
    },
    {
      "epoch": 1.25,
      "grad_norm": 0.34547096490859985,
      "learning_rate": 1.896551348812956e-05,
      "loss": 0.0845,
      "step": 709
    },
    {
      "epoch": 1.26,
      "grad_norm": 0.3739677369594574,
      "learning_rate": 1.893853491051165e-05,
      "loss": 0.087,
      "step": 710
    },
    {
      "epoch": 1.26,
      "eval_loss": 0.07435023784637451,
      "eval_runtime": 14.6675,
      "eval_samples_per_second": 32.521,
      "eval_steps_per_second": 8.181,
      "step": 710
    },
    {
      "epoch": 1.26,
      "grad_norm": 0.34483763575553894,
      "learning_rate": 1.8911542641914617e-05,
      "loss": 0.1147,
      "step": 711
    },
    {
      "epoch": 1.26,
      "grad_norm": 0.3022269606590271,
      "learning_rate": 1.88845367761679e-05,
      "loss": 0.071,
      "step": 712
    },
    {
      "epoch": 1.26,
      "grad_norm": 0.2281966507434845,
      "learning_rate": 1.8857517407148236e-05,
      "loss": 0.0811,
      "step": 713
    },
    {
      "epoch": 1.26,
      "grad_norm": 0.5963127017021179,
      "learning_rate": 1.8830484628779267e-05,
      "loss": 0.1027,
      "step": 714
    },
    {
      "epoch": 1.26,
      "grad_norm": 0.43141207098960876,
      "learning_rate": 1.880343853503129e-05,
      "loss": 0.1366,
      "step": 715
    },
    {
      "epoch": 1.27,
      "grad_norm": 0.2259010374546051,
      "learning_rate": 1.8776379219920847e-05,
      "loss": 0.0529,
      "step": 716
    },
    {
      "epoch": 1.27,
      "grad_norm": 0.3586896061897278,
      "learning_rate": 1.8749306777510465e-05,
      "loss": 0.0976,
      "step": 717
    },
    {
      "epoch": 1.27,
      "grad_norm": 0.3314034640789032,
      "learning_rate": 1.8722221301908295e-05,
      "loss": 0.054,
      "step": 718
    },
    {
      "epoch": 1.27,
      "grad_norm": 0.2230258584022522,
      "learning_rate": 1.8695122887267788e-05,
      "loss": 0.0584,
      "step": 719
    },
    {
      "epoch": 1.27,
      "grad_norm": 0.3144403398036957,
      "learning_rate": 1.8668011627787398e-05,
      "loss": 0.07,
      "step": 720
    },
    {
      "epoch": 1.28,
      "grad_norm": 0.28598567843437195,
      "learning_rate": 1.86408876177102e-05,
      "loss": 0.0748,
      "step": 721
    },
    {
      "epoch": 1.28,
      "grad_norm": 0.34994563460350037,
      "learning_rate": 1.8613750951323602e-05,
      "loss": 0.0852,
      "step": 722
    },
    {
      "epoch": 1.28,
      "grad_norm": 0.2040272057056427,
      "learning_rate": 1.858660172295901e-05,
      "loss": 0.0562,
      "step": 723
    },
    {
      "epoch": 1.28,
      "grad_norm": 0.3063504099845886,
      "learning_rate": 1.8559440026991508e-05,
      "loss": 0.0679,
      "step": 724
    },
    {
      "epoch": 1.28,
      "grad_norm": 0.26353538036346436,
      "learning_rate": 1.8532265957839497e-05,
      "loss": 0.0659,
      "step": 725
    },
    {
      "epoch": 1.28,
      "grad_norm": 0.4694713056087494,
      "learning_rate": 1.8505079609964403e-05,
      "loss": 0.1353,
      "step": 726
    },
    {
      "epoch": 1.29,
      "grad_norm": 0.2791508138179779,
      "learning_rate": 1.847788107787033e-05,
      "loss": 0.0622,
      "step": 727
    },
    {
      "epoch": 1.29,
      "grad_norm": 0.26720237731933594,
      "learning_rate": 1.8450670456103743e-05,
      "loss": 0.0814,
      "step": 728
    },
    {
      "epoch": 1.29,
      "grad_norm": 0.2125985026359558,
      "learning_rate": 1.8423447839253128e-05,
      "loss": 0.0408,
      "step": 729
    },
    {
      "epoch": 1.29,
      "grad_norm": 0.2482578605413437,
      "learning_rate": 1.8396213321948662e-05,
      "loss": 0.0404,
      "step": 730
    },
    {
      "epoch": 1.29,
      "grad_norm": 0.3244602680206299,
      "learning_rate": 1.83689669988619e-05,
      "loss": 0.0954,
      "step": 731
    },
    {
      "epoch": 1.3,
      "grad_norm": 0.369554728269577,
      "learning_rate": 1.8341708964705434e-05,
      "loss": 0.0992,
      "step": 732
    },
    {
      "epoch": 1.3,
      "grad_norm": 0.507155179977417,
      "learning_rate": 1.831443931423256e-05,
      "loss": 0.095,
      "step": 733
    },
    {
      "epoch": 1.3,
      "grad_norm": 0.38941535353660583,
      "learning_rate": 1.828715814223696e-05,
      "loss": 0.0941,
      "step": 734
    },
    {
      "epoch": 1.3,
      "grad_norm": 0.2319694459438324,
      "learning_rate": 1.825986554355236e-05,
      "loss": 0.064,
      "step": 735
    },
    {
      "epoch": 1.3,
      "grad_norm": 0.2474977821111679,
      "learning_rate": 1.8232561613052214e-05,
      "loss": 0.0464,
      "step": 736
    },
    {
      "epoch": 1.3,
      "grad_norm": 0.25080692768096924,
      "learning_rate": 1.8205246445649366e-05,
      "loss": 0.0817,
      "step": 737
    },
    {
      "epoch": 1.31,
      "grad_norm": 0.3491278886795044,
      "learning_rate": 1.8177920136295715e-05,
      "loss": 0.067,
      "step": 738
    },
    {
      "epoch": 1.31,
      "grad_norm": 0.358746737241745,
      "learning_rate": 1.81505827799819e-05,
      "loss": 0.0903,
      "step": 739
    },
    {
      "epoch": 1.31,
      "grad_norm": 0.2560611367225647,
      "learning_rate": 1.8123234471736947e-05,
      "loss": 0.0368,
      "step": 740
    },
    {
      "epoch": 1.31,
      "grad_norm": 0.19439151883125305,
      "learning_rate": 1.809587530662798e-05,
      "loss": 0.0463,
      "step": 741
    },
    {
      "epoch": 1.31,
      "grad_norm": 0.25514161586761475,
      "learning_rate": 1.8068505379759828e-05,
      "loss": 0.0603,
      "step": 742
    },
    {
      "epoch": 1.31,
      "grad_norm": 0.26982584595680237,
      "learning_rate": 1.8041124786274757e-05,
      "loss": 0.0734,
      "step": 743
    },
    {
      "epoch": 1.32,
      "grad_norm": 0.25043055415153503,
      "learning_rate": 1.80137336213521e-05,
      "loss": 0.0515,
      "step": 744
    },
    {
      "epoch": 1.32,
      "grad_norm": 0.1994292140007019,
      "learning_rate": 1.7986331980207942e-05,
      "loss": 0.0426,
      "step": 745
    },
    {
      "epoch": 1.32,
      "grad_norm": 0.25462812185287476,
      "learning_rate": 1.795891995809478e-05,
      "loss": 0.0761,
      "step": 746
    },
    {
      "epoch": 1.32,
      "grad_norm": 0.2635657489299774,
      "learning_rate": 1.79314976503012e-05,
      "loss": 0.0495,
      "step": 747
    },
    {
      "epoch": 1.32,
      "grad_norm": 0.3107217252254486,
      "learning_rate": 1.7904065152151547e-05,
      "loss": 0.0723,
      "step": 748
    },
    {
      "epoch": 1.33,
      "grad_norm": 0.3569051921367645,
      "learning_rate": 1.7876622559005578e-05,
      "loss": 0.1097,
      "step": 749
    },
    {
      "epoch": 1.33,
      "grad_norm": 0.44406095147132874,
      "learning_rate": 1.7849169966258158e-05,
      "loss": 0.0709,
      "step": 750
    },
    {
      "epoch": 1.33,
      "grad_norm": 0.29121848940849304,
      "learning_rate": 1.7821707469338894e-05,
      "loss": 0.0571,
      "step": 751
    },
    {
      "epoch": 1.33,
      "grad_norm": 0.5233845114707947,
      "learning_rate": 1.7794235163711837e-05,
      "loss": 0.0925,
      "step": 752
    },
    {
      "epoch": 1.33,
      "grad_norm": 0.3112752437591553,
      "learning_rate": 1.776675314487512e-05,
      "loss": 0.0597,
      "step": 753
    },
    {
      "epoch": 1.33,
      "grad_norm": 0.415249764919281,
      "learning_rate": 1.7739261508360662e-05,
      "loss": 0.0805,
      "step": 754
    },
    {
      "epoch": 1.34,
      "grad_norm": 0.33063799142837524,
      "learning_rate": 1.7711760349733794e-05,
      "loss": 0.0415,
      "step": 755
    },
    {
      "epoch": 1.34,
      "grad_norm": 0.6214970946311951,
      "learning_rate": 1.7684249764592948e-05,
      "loss": 0.0753,
      "step": 756
    },
    {
      "epoch": 1.34,
      "grad_norm": 0.3045737147331238,
      "learning_rate": 1.7656729848569342e-05,
      "loss": 0.062,
      "step": 757
    },
    {
      "epoch": 1.34,
      "grad_norm": 0.31340816617012024,
      "learning_rate": 1.7629200697326617e-05,
      "loss": 0.0577,
      "step": 758
    },
    {
      "epoch": 1.34,
      "grad_norm": 0.3969898223876953,
      "learning_rate": 1.760166240656051e-05,
      "loss": 0.0522,
      "step": 759
    },
    {
      "epoch": 1.34,
      "grad_norm": 0.30962708592414856,
      "learning_rate": 1.757411507199855e-05,
      "loss": 0.0579,
      "step": 760
    },
    {
      "epoch": 1.35,
      "grad_norm": 0.30885669589042664,
      "learning_rate": 1.754655878939968e-05,
      "loss": 0.0459,
      "step": 761
    },
    {
      "epoch": 1.35,
      "grad_norm": 0.2914258539676666,
      "learning_rate": 1.7518993654553963e-05,
      "loss": 0.0749,
      "step": 762
    },
    {
      "epoch": 1.35,
      "grad_norm": 0.5590508580207825,
      "learning_rate": 1.7491419763282228e-05,
      "loss": 0.1082,
      "step": 763
    },
    {
      "epoch": 1.35,
      "grad_norm": 0.6015578508377075,
      "learning_rate": 1.7463837211435746e-05,
      "loss": 0.0759,
      "step": 764
    },
    {
      "epoch": 1.35,
      "grad_norm": 0.3625446557998657,
      "learning_rate": 1.7436246094895896e-05,
      "loss": 0.0628,
      "step": 765
    },
    {
      "epoch": 1.36,
      "grad_norm": 0.3535112142562866,
      "learning_rate": 1.7408646509573815e-05,
      "loss": 0.063,
      "step": 766
    },
    {
      "epoch": 1.36,
      "grad_norm": 0.6107704639434814,
      "learning_rate": 1.73810385514101e-05,
      "loss": 0.1272,
      "step": 767
    },
    {
      "epoch": 1.36,
      "grad_norm": 0.5027735233306885,
      "learning_rate": 1.7353422316374428e-05,
      "loss": 0.0683,
      "step": 768
    },
    {
      "epoch": 1.36,
      "grad_norm": 0.2779427468776703,
      "learning_rate": 1.732579790046528e-05,
      "loss": 0.0428,
      "step": 769
    },
    {
      "epoch": 1.36,
      "grad_norm": 0.6216077208518982,
      "learning_rate": 1.729816539970954e-05,
      "loss": 0.1563,
      "step": 770
    },
    {
      "epoch": 1.36,
      "grad_norm": 0.46196988224983215,
      "learning_rate": 1.727052491016222e-05,
      "loss": 0.1104,
      "step": 771
    },
    {
      "epoch": 1.37,
      "grad_norm": 0.43399301171302795,
      "learning_rate": 1.7242876527906096e-05,
      "loss": 0.1069,
      "step": 772
    },
    {
      "epoch": 1.37,
      "grad_norm": 0.21114163100719452,
      "learning_rate": 1.721522034905138e-05,
      "loss": 0.0334,
      "step": 773
    },
    {
      "epoch": 1.37,
      "grad_norm": 0.21999305486679077,
      "learning_rate": 1.718755646973539e-05,
      "loss": 0.0429,
      "step": 774
    },
    {
      "epoch": 1.37,
      "grad_norm": 0.36062875390052795,
      "learning_rate": 1.7159884986122196e-05,
      "loss": 0.0928,
      "step": 775
    },
    {
      "epoch": 1.37,
      "grad_norm": 0.21104684472084045,
      "learning_rate": 1.713220599440233e-05,
      "loss": 0.0441,
      "step": 776
    },
    {
      "epoch": 1.37,
      "grad_norm": 0.30601590871810913,
      "learning_rate": 1.7104519590792396e-05,
      "loss": 0.1,
      "step": 777
    },
    {
      "epoch": 1.38,
      "grad_norm": 0.19373103976249695,
      "learning_rate": 1.7076825871534775e-05,
      "loss": 0.0363,
      "step": 778
    },
    {
      "epoch": 1.38,
      "grad_norm": 0.2828684449195862,
      "learning_rate": 1.7049124932897278e-05,
      "loss": 0.0905,
      "step": 779
    },
    {
      "epoch": 1.38,
      "grad_norm": 0.2555391788482666,
      "learning_rate": 1.7021416871172817e-05,
      "loss": 0.0575,
      "step": 780
    },
    {
      "epoch": 1.38,
      "grad_norm": 0.2440968006849289,
      "learning_rate": 1.6993701782679056e-05,
      "loss": 0.0774,
      "step": 781
    },
    {
      "epoch": 1.38,
      "grad_norm": 0.31575918197631836,
      "learning_rate": 1.696597976375808e-05,
      "loss": 0.0788,
      "step": 782
    },
    {
      "epoch": 1.39,
      "grad_norm": 0.2803939878940582,
      "learning_rate": 1.6938250910776086e-05,
      "loss": 0.0715,
      "step": 783
    },
    {
      "epoch": 1.39,
      "grad_norm": 0.21095207333564758,
      "learning_rate": 1.691051532012301e-05,
      "loss": 0.0652,
      "step": 784
    },
    {
      "epoch": 1.39,
      "grad_norm": 0.26113957166671753,
      "learning_rate": 1.6882773088212213e-05,
      "loss": 0.0677,
      "step": 785
    },
    {
      "epoch": 1.39,
      "grad_norm": 0.3644431531429291,
      "learning_rate": 1.6855024311480147e-05,
      "loss": 0.0983,
      "step": 786
    },
    {
      "epoch": 1.39,
      "grad_norm": 0.2940008342266083,
      "learning_rate": 1.6827269086386006e-05,
      "loss": 0.0759,
      "step": 787
    },
    {
      "epoch": 1.39,
      "grad_norm": 0.3901608884334564,
      "learning_rate": 1.6799507509411408e-05,
      "loss": 0.0709,
      "step": 788
    },
    {
      "epoch": 1.4,
      "grad_norm": 0.4993228316307068,
      "learning_rate": 1.6771739677060042e-05,
      "loss": 0.0776,
      "step": 789
    },
    {
      "epoch": 1.4,
      "grad_norm": 0.23843778669834137,
      "learning_rate": 1.6743965685857362e-05,
      "loss": 0.0638,
      "step": 790
    },
    {
      "epoch": 1.4,
      "grad_norm": 0.26617488265037537,
      "learning_rate": 1.6716185632350204e-05,
      "loss": 0.0745,
      "step": 791
    },
    {
      "epoch": 1.4,
      "grad_norm": 0.24340148270130157,
      "learning_rate": 1.6688399613106488e-05,
      "loss": 0.0544,
      "step": 792
    },
    {
      "epoch": 1.4,
      "grad_norm": 0.2784338593482971,
      "learning_rate": 1.666060772471488e-05,
      "loss": 0.0311,
      "step": 793
    },
    {
      "epoch": 1.4,
      "grad_norm": 0.28367799520492554,
      "learning_rate": 1.663281006378445e-05,
      "loss": 0.0592,
      "step": 794
    },
    {
      "epoch": 1.41,
      "grad_norm": 0.43154534697532654,
      "learning_rate": 1.6605006726944317e-05,
      "loss": 0.0646,
      "step": 795
    },
    {
      "epoch": 1.41,
      "grad_norm": 0.4112074673175812,
      "learning_rate": 1.6577197810843332e-05,
      "loss": 0.0482,
      "step": 796
    },
    {
      "epoch": 1.41,
      "grad_norm": 0.7609788775444031,
      "learning_rate": 1.6549383412149764e-05,
      "loss": 0.0816,
      "step": 797
    },
    {
      "epoch": 1.41,
      "grad_norm": 0.5045328736305237,
      "learning_rate": 1.6521563627550915e-05,
      "loss": 0.1091,
      "step": 798
    },
    {
      "epoch": 1.41,
      "grad_norm": 0.7338900566101074,
      "learning_rate": 1.6493738553752822e-05,
      "loss": 0.105,
      "step": 799
    },
    {
      "epoch": 1.42,
      "grad_norm": 0.4961400628089905,
      "learning_rate": 1.646590828747991e-05,
      "loss": 0.0835,
      "step": 800
    },
    {
      "epoch": 1.42,
      "grad_norm": 0.19100184738636017,
      "learning_rate": 1.643807292547463e-05,
      "loss": 0.0367,
      "step": 801
    },
    {
      "epoch": 1.42,
      "grad_norm": 0.387093186378479,
      "learning_rate": 1.6410232564497188e-05,
      "loss": 0.0608,
      "step": 802
    },
    {
      "epoch": 1.42,
      "grad_norm": 0.23065884411334991,
      "learning_rate": 1.6382387301325136e-05,
      "loss": 0.0323,
      "step": 803
    },
    {
      "epoch": 1.42,
      "grad_norm": 0.3351573348045349,
      "learning_rate": 1.6354537232753072e-05,
      "loss": 0.0563,
      "step": 804
    },
    {
      "epoch": 1.42,
      "grad_norm": 0.47596344351768494,
      "learning_rate": 1.6326682455592305e-05,
      "loss": 0.0711,
      "step": 805
    },
    {
      "epoch": 1.43,
      "grad_norm": 0.25883013010025024,
      "learning_rate": 1.629882306667051e-05,
      "loss": 0.0499,
      "step": 806
    },
    {
      "epoch": 1.43,
      "grad_norm": 0.22424423694610596,
      "learning_rate": 1.627095916283139e-05,
      "loss": 0.0366,
      "step": 807
    },
    {
      "epoch": 1.43,
      "grad_norm": 0.5582901835441589,
      "learning_rate": 1.6243090840934344e-05,
      "loss": 0.1633,
      "step": 808
    },
    {
      "epoch": 1.43,
      "grad_norm": 0.34346896409988403,
      "learning_rate": 1.621521819785413e-05,
      "loss": 0.0677,
      "step": 809
    },
    {
      "epoch": 1.43,
      "grad_norm": 0.21031644940376282,
      "learning_rate": 1.618734133048052e-05,
      "loss": 0.0317,
      "step": 810
    },
    {
      "epoch": 1.43,
      "grad_norm": 0.681185781955719,
      "learning_rate": 1.6159460335717992e-05,
      "loss": 0.0939,
      "step": 811
    },
    {
      "epoch": 1.44,
      "grad_norm": 0.2531503140926361,
      "learning_rate": 1.6131575310485346e-05,
      "loss": 0.0573,
      "step": 812
    },
    {
      "epoch": 1.44,
      "grad_norm": 0.21248729526996613,
      "learning_rate": 1.6103686351715403e-05,
      "loss": 0.0446,
      "step": 813
    },
    {
      "epoch": 1.44,
      "grad_norm": 0.4461505711078644,
      "learning_rate": 1.6075793556354655e-05,
      "loss": 0.1028,
      "step": 814
    },
    {
      "epoch": 1.44,
      "grad_norm": 0.3272171914577484,
      "learning_rate": 1.604789702136294e-05,
      "loss": 0.0667,
      "step": 815
    },
    {
      "epoch": 1.44,
      "grad_norm": 0.3863116204738617,
      "learning_rate": 1.6019996843713088e-05,
      "loss": 0.0804,
      "step": 816
    },
    {
      "epoch": 1.45,
      "grad_norm": 0.36971715092658997,
      "learning_rate": 1.599209312039058e-05,
      "loss": 0.0829,
      "step": 817
    },
    {
      "epoch": 1.45,
      "grad_norm": 0.5918539762496948,
      "learning_rate": 1.5964185948393248e-05,
      "loss": 0.0891,
      "step": 818
    },
    {
      "epoch": 1.45,
      "grad_norm": 0.36615926027297974,
      "learning_rate": 1.5936275424730892e-05,
      "loss": 0.0929,
      "step": 819
    },
    {
      "epoch": 1.45,
      "grad_norm": 0.4229936897754669,
      "learning_rate": 1.5908361646424973e-05,
      "loss": 0.0677,
      "step": 820
    },
    {
      "epoch": 1.45,
      "grad_norm": 0.22287757694721222,
      "learning_rate": 1.5880444710508263e-05,
      "loss": 0.0394,
      "step": 821
    },
    {
      "epoch": 1.45,
      "grad_norm": 0.24906857311725616,
      "learning_rate": 1.5852524714024502e-05,
      "loss": 0.0681,
      "step": 822
    },
    {
      "epoch": 1.46,
      "grad_norm": 0.25015828013420105,
      "learning_rate": 1.5824601754028083e-05,
      "loss": 0.0773,
      "step": 823
    },
    {
      "epoch": 1.46,
      "grad_norm": 0.19321171939373016,
      "learning_rate": 1.579667592758369e-05,
      "loss": 0.0538,
      "step": 824
    },
    {
      "epoch": 1.46,
      "grad_norm": 0.23880505561828613,
      "learning_rate": 1.5768747331765975e-05,
      "loss": 0.0342,
      "step": 825
    },
    {
      "epoch": 1.46,
      "grad_norm": 0.27807632088661194,
      "learning_rate": 1.5740816063659226e-05,
      "loss": 0.0891,
      "step": 826
    },
    {
      "epoch": 1.46,
      "grad_norm": 0.4834766387939453,
      "learning_rate": 1.5712882220357e-05,
      "loss": 0.1325,
      "step": 827
    },
    {
      "epoch": 1.46,
      "grad_norm": 0.3581242263317108,
      "learning_rate": 1.5684945898961826e-05,
      "loss": 0.0721,
      "step": 828
    },
    {
      "epoch": 1.47,
      "grad_norm": 0.33834755420684814,
      "learning_rate": 1.5657007196584833e-05,
      "loss": 0.0731,
      "step": 829
    },
    {
      "epoch": 1.47,
      "grad_norm": 0.3898695707321167,
      "learning_rate": 1.562906621034543e-05,
      "loss": 0.1152,
      "step": 830
    },
    {
      "epoch": 1.47,
      "grad_norm": 0.2026025652885437,
      "learning_rate": 1.5601123037370977e-05,
      "loss": 0.0275,
      "step": 831
    },
    {
      "epoch": 1.47,
      "grad_norm": 0.2803891897201538,
      "learning_rate": 1.5573177774796415e-05,
      "loss": 0.0679,
      "step": 832
    },
    {
      "epoch": 1.47,
      "grad_norm": 0.3605427145957947,
      "learning_rate": 1.5545230519763967e-05,
      "loss": 0.0949,
      "step": 833
    },
    {
      "epoch": 1.48,
      "grad_norm": 0.27228227257728577,
      "learning_rate": 1.5517281369422767e-05,
      "loss": 0.0762,
      "step": 834
    },
    {
      "epoch": 1.48,
      "grad_norm": 0.29753464460372925,
      "learning_rate": 1.5489330420928554e-05,
      "loss": 0.0603,
      "step": 835
    },
    {
      "epoch": 1.48,
      "grad_norm": 0.2596915364265442,
      "learning_rate": 1.54613777714433e-05,
      "loss": 0.0843,
      "step": 836
    },
    {
      "epoch": 1.48,
      "grad_norm": 0.16269999742507935,
      "learning_rate": 1.5433423518134906e-05,
      "loss": 0.0308,
      "step": 837
    },
    {
      "epoch": 1.48,
      "grad_norm": 0.23855046927928925,
      "learning_rate": 1.5405467758176836e-05,
      "loss": 0.0685,
      "step": 838
    },
    {
      "epoch": 1.48,
      "grad_norm": 0.22682958841323853,
      "learning_rate": 1.53775105887478e-05,
      "loss": 0.0702,
      "step": 839
    },
    {
      "epoch": 1.49,
      "grad_norm": 0.18888510763645172,
      "learning_rate": 1.5349552107031397e-05,
      "loss": 0.0355,
      "step": 840
    },
    {
      "epoch": 1.49,
      "grad_norm": 0.344084769487381,
      "learning_rate": 1.5321592410215797e-05,
      "loss": 0.0943,
      "step": 841
    },
    {
      "epoch": 1.49,
      "grad_norm": 0.23934723436832428,
      "learning_rate": 1.5293631595493402e-05,
      "loss": 0.07,
      "step": 842
    },
    {
      "epoch": 1.49,
      "grad_norm": 0.36512523889541626,
      "learning_rate": 1.5265669760060465e-05,
      "loss": 0.0705,
      "step": 843
    },
    {
      "epoch": 1.49,
      "grad_norm": 0.29379802942276,
      "learning_rate": 1.5237707001116832e-05,
      "loss": 0.0942,
      "step": 844
    },
    {
      "epoch": 1.49,
      "grad_norm": 0.22728702425956726,
      "learning_rate": 1.5209743415865535e-05,
      "loss": 0.054,
      "step": 845
    },
    {
      "epoch": 1.5,
      "grad_norm": 0.23909953236579895,
      "learning_rate": 1.5181779101512479e-05,
      "loss": 0.0435,
      "step": 846
    },
    {
      "epoch": 1.5,
      "grad_norm": 0.28431251645088196,
      "learning_rate": 1.5153814155266105e-05,
      "loss": 0.0468,
      "step": 847
    },
    {
      "epoch": 1.5,
      "grad_norm": 0.2859126031398773,
      "learning_rate": 1.5125848674337052e-05,
      "loss": 0.0531,
      "step": 848
    },
    {
      "epoch": 1.5,
      "grad_norm": 0.2681651711463928,
      "learning_rate": 1.5097882755937824e-05,
      "loss": 0.0493,
      "step": 849
    },
    {
      "epoch": 1.5,
      "grad_norm": 0.41042011976242065,
      "learning_rate": 1.5069916497282432e-05,
      "loss": 0.101,
      "step": 850
    },
    {
      "epoch": 1.51,
      "grad_norm": 0.21335624158382416,
      "learning_rate": 1.504194999558608e-05,
      "loss": 0.0606,
      "step": 851
    },
    {
      "epoch": 1.51,
      "grad_norm": 0.309781014919281,
      "learning_rate": 1.5013983348064814e-05,
      "loss": 0.0724,
      "step": 852
    },
    {
      "epoch": 1.51,
      "eval_loss": 0.07102719694375992,
      "eval_runtime": 14.6891,
      "eval_samples_per_second": 32.473,
      "eval_steps_per_second": 8.169,
      "step": 852
    },
    {
      "epoch": 1.51,
      "grad_norm": 0.3069019317626953,
      "learning_rate": 1.4986016651935187e-05,
      "loss": 0.0455,
      "step": 853
    },
    {
      "epoch": 1.51,
      "grad_norm": 0.723366379737854,
      "learning_rate": 1.4958050004413922e-05,
      "loss": 0.0927,
      "step": 854
    },
    {
      "epoch": 1.51,
      "grad_norm": 0.20222370326519012,
      "learning_rate": 1.4930083502717573e-05,
      "loss": 0.0259,
      "step": 855
    },
    {
      "epoch": 1.51,
      "grad_norm": 0.476968377828598,
      "learning_rate": 1.490211724406218e-05,
      "loss": 0.105,
      "step": 856
    },
    {
      "epoch": 1.52,
      "grad_norm": 0.4652007520198822,
      "learning_rate": 1.4874151325662952e-05,
      "loss": 0.0628,
      "step": 857
    },
    {
      "epoch": 1.52,
      "grad_norm": 0.3219752013683319,
      "learning_rate": 1.4846185844733896e-05,
      "loss": 0.0699,
      "step": 858
    },
    {
      "epoch": 1.52,
      "grad_norm": 0.29580509662628174,
      "learning_rate": 1.4818220898487522e-05,
      "loss": 0.0427,
      "step": 859
    },
    {
      "epoch": 1.52,
      "grad_norm": 0.3006477355957031,
      "learning_rate": 1.4790256584134467e-05,
      "loss": 0.0568,
      "step": 860
    },
    {
      "epoch": 1.52,
      "grad_norm": 0.6600325703620911,
      "learning_rate": 1.4762292998883167e-05,
      "loss": 0.0738,
      "step": 861
    },
    {
      "epoch": 1.52,
      "grad_norm": 0.3576618432998657,
      "learning_rate": 1.4734330239939537e-05,
      "loss": 0.0256,
      "step": 862
    },
    {
      "epoch": 1.53,
      "grad_norm": 0.24733182787895203,
      "learning_rate": 1.4706368404506604e-05,
      "loss": 0.0352,
      "step": 863
    },
    {
      "epoch": 1.53,
      "grad_norm": 0.5404502749443054,
      "learning_rate": 1.4678407589784202e-05,
      "loss": 0.0658,
      "step": 864
    },
    {
      "epoch": 1.53,
      "grad_norm": 0.22657887637615204,
      "learning_rate": 1.4650447892968606e-05,
      "loss": 0.045,
      "step": 865
    },
    {
      "epoch": 1.53,
      "grad_norm": 0.543570876121521,
      "learning_rate": 1.4622489411252203e-05,
      "loss": 0.1212,
      "step": 866
    },
    {
      "epoch": 1.53,
      "grad_norm": 0.34453868865966797,
      "learning_rate": 1.4594532241823168e-05,
      "loss": 0.0518,
      "step": 867
    },
    {
      "epoch": 1.54,
      "grad_norm": 0.6955142021179199,
      "learning_rate": 1.4566576481865097e-05,
      "loss": 0.1413,
      "step": 868
    },
    {
      "epoch": 1.54,
      "grad_norm": 0.6116628646850586,
      "learning_rate": 1.4538622228556698e-05,
      "loss": 0.1609,
      "step": 869
    },
    {
      "epoch": 1.54,
      "grad_norm": 0.6426568627357483,
      "learning_rate": 1.4510669579071449e-05,
      "loss": 0.1168,
      "step": 870
    },
    {
      "epoch": 1.54,
      "grad_norm": 0.23240916430950165,
      "learning_rate": 1.4482718630577234e-05,
      "loss": 0.0407,
      "step": 871
    },
    {
      "epoch": 1.54,
      "grad_norm": 0.23364579677581787,
      "learning_rate": 1.4454769480236034e-05,
      "loss": 0.0548,
      "step": 872
    },
    {
      "epoch": 1.54,
      "grad_norm": 0.24377289414405823,
      "learning_rate": 1.442682222520359e-05,
      "loss": 0.0411,
      "step": 873
    },
    {
      "epoch": 1.55,
      "grad_norm": 0.31778621673583984,
      "learning_rate": 1.4398876962629026e-05,
      "loss": 0.0728,
      "step": 874
    },
    {
      "epoch": 1.55,
      "grad_norm": 0.23079492151737213,
      "learning_rate": 1.4370933789654571e-05,
      "loss": 0.0757,
      "step": 875
    },
    {
      "epoch": 1.55,
      "grad_norm": 0.19835864007472992,
      "learning_rate": 1.4342992803415171e-05,
      "loss": 0.0862,
      "step": 876
    },
    {
      "epoch": 1.55,
      "grad_norm": 0.2709265649318695,
      "learning_rate": 1.4315054101038176e-05,
      "loss": 0.0687,
      "step": 877
    },
    {
      "epoch": 1.55,
      "grad_norm": 0.25470414757728577,
      "learning_rate": 1.4287117779643005e-05,
      "loss": 0.0618,
      "step": 878
    },
    {
      "epoch": 1.56,
      "grad_norm": 0.31228628754615784,
      "learning_rate": 1.4259183936340775e-05,
      "loss": 0.1058,
      "step": 879
    },
    {
      "epoch": 1.56,
      "grad_norm": 0.30994561314582825,
      "learning_rate": 1.4231252668234025e-05,
      "loss": 0.0555,
      "step": 880
    },
    {
      "epoch": 1.56,
      "grad_norm": 0.21292701363563538,
      "learning_rate": 1.4203324072416313e-05,
      "loss": 0.0446,
      "step": 881
    },
    {
      "epoch": 1.56,
      "grad_norm": 0.21547499299049377,
      "learning_rate": 1.4175398245971921e-05,
      "loss": 0.0604,
      "step": 882
    },
    {
      "epoch": 1.56,
      "grad_norm": 0.4474267363548279,
      "learning_rate": 1.4147475285975504e-05,
      "loss": 0.0809,
      "step": 883
    },
    {
      "epoch": 1.56,
      "grad_norm": 0.23740467429161072,
      "learning_rate": 1.411955528949174e-05,
      "loss": 0.0661,
      "step": 884
    },
    {
      "epoch": 1.57,
      "grad_norm": 0.32307055592536926,
      "learning_rate": 1.4091638353575025e-05,
      "loss": 0.0634,
      "step": 885
    },
    {
      "epoch": 1.57,
      "grad_norm": 0.22688160836696625,
      "learning_rate": 1.4063724575269109e-05,
      "loss": 0.0496,
      "step": 886
    },
    {
      "epoch": 1.57,
      "grad_norm": 0.18514683842658997,
      "learning_rate": 1.4035814051606753e-05,
      "loss": 0.0638,
      "step": 887
    },
    {
      "epoch": 1.57,
      "grad_norm": 0.23845839500427246,
      "learning_rate": 1.4007906879609424e-05,
      "loss": 0.0533,
      "step": 888
    },
    {
      "epoch": 1.57,
      "grad_norm": 0.3531744182109833,
      "learning_rate": 1.3980003156286916e-05,
      "loss": 0.082,
      "step": 889
    },
    {
      "epoch": 1.57,
      "grad_norm": 0.216387540102005,
      "learning_rate": 1.3952102978637058e-05,
      "loss": 0.0692,
      "step": 890
    },
    {
      "epoch": 1.58,
      "grad_norm": 0.29363855719566345,
      "learning_rate": 1.3924206443645345e-05,
      "loss": 0.0803,
      "step": 891
    },
    {
      "epoch": 1.58,
      "grad_norm": 0.3476901650428772,
      "learning_rate": 1.38963136482846e-05,
      "loss": 0.1126,
      "step": 892
    },
    {
      "epoch": 1.58,
      "grad_norm": 0.4968585669994354,
      "learning_rate": 1.3868424689514658e-05,
      "loss": 0.1233,
      "step": 893
    },
    {
      "epoch": 1.58,
      "grad_norm": 0.31899258494377136,
      "learning_rate": 1.384053966428201e-05,
      "loss": 0.0671,
      "step": 894
    },
    {
      "epoch": 1.58,
      "grad_norm": 0.3615509271621704,
      "learning_rate": 1.3812658669519474e-05,
      "loss": 0.0668,
      "step": 895
    },
    {
      "epoch": 1.59,
      "grad_norm": 0.403689980506897,
      "learning_rate": 1.3784781802145875e-05,
      "loss": 0.1079,
      "step": 896
    },
    {
      "epoch": 1.59,
      "grad_norm": 0.42689767479896545,
      "learning_rate": 1.3756909159065659e-05,
      "loss": 0.1182,
      "step": 897
    },
    {
      "epoch": 1.59,
      "grad_norm": 0.27076485753059387,
      "learning_rate": 1.372904083716861e-05,
      "loss": 0.0305,
      "step": 898
    },
    {
      "epoch": 1.59,
      "grad_norm": 0.3220841884613037,
      "learning_rate": 1.3701176933329495e-05,
      "loss": 0.0902,
      "step": 899
    },
    {
      "epoch": 1.59,
      "grad_norm": 0.4248409867286682,
      "learning_rate": 1.3673317544407694e-05,
      "loss": 0.067,
      "step": 900
    },
    {
      "epoch": 1.59,
      "grad_norm": 0.45589709281921387,
      "learning_rate": 1.364546276724693e-05,
      "loss": 0.1396,
      "step": 901
    },
    {
      "epoch": 1.6,
      "grad_norm": 0.24416354298591614,
      "learning_rate": 1.3617612698674867e-05,
      "loss": 0.0537,
      "step": 902
    },
    {
      "epoch": 1.6,
      "grad_norm": 0.2165834754705429,
      "learning_rate": 1.3589767435502813e-05,
      "loss": 0.0468,
      "step": 903
    },
    {
      "epoch": 1.6,
      "grad_norm": 0.31254321336746216,
      "learning_rate": 1.3561927074525372e-05,
      "loss": 0.0689,
      "step": 904
    },
    {
      "epoch": 1.6,
      "grad_norm": 0.292851984500885,
      "learning_rate": 1.3534091712520097e-05,
      "loss": 0.0906,
      "step": 905
    },
    {
      "epoch": 1.6,
      "grad_norm": 0.36904656887054443,
      "learning_rate": 1.3506261446247182e-05,
      "loss": 0.0956,
      "step": 906
    },
    {
      "epoch": 1.6,
      "grad_norm": 0.25328972935676575,
      "learning_rate": 1.3478436372449089e-05,
      "loss": 0.0956,
      "step": 907
    },
    {
      "epoch": 1.61,
      "grad_norm": 0.33212894201278687,
      "learning_rate": 1.345061658785024e-05,
      "loss": 0.0771,
      "step": 908
    },
    {
      "epoch": 1.61,
      "grad_norm": 0.19586564600467682,
      "learning_rate": 1.3422802189156672e-05,
      "loss": 0.037,
      "step": 909
    },
    {
      "epoch": 1.61,
      "grad_norm": 0.2859354019165039,
      "learning_rate": 1.339499327305569e-05,
      "loss": 0.0637,
      "step": 910
    },
    {
      "epoch": 1.61,
      "grad_norm": 0.43225035071372986,
      "learning_rate": 1.3367189936215552e-05,
      "loss": 0.082,
      "step": 911
    },
    {
      "epoch": 1.61,
      "grad_norm": 0.42300665378570557,
      "learning_rate": 1.3339392275285118e-05,
      "loss": 0.08,
      "step": 912
    },
    {
      "epoch": 1.62,
      "grad_norm": 0.21360281109809875,
      "learning_rate": 1.3311600386893513e-05,
      "loss": 0.06,
      "step": 913
    },
    {
      "epoch": 1.62,
      "grad_norm": 0.30538302659988403,
      "learning_rate": 1.3283814367649804e-05,
      "loss": 0.0413,
      "step": 914
    },
    {
      "epoch": 1.62,
      "grad_norm": 0.22462303936481476,
      "learning_rate": 1.3256034314142639e-05,
      "loss": 0.0521,
      "step": 915
    },
    {
      "epoch": 1.62,
      "grad_norm": 0.29125386476516724,
      "learning_rate": 1.3228260322939955e-05,
      "loss": 0.0783,
      "step": 916
    },
    {
      "epoch": 1.62,
      "grad_norm": 0.3136795163154602,
      "learning_rate": 1.3200492490588596e-05,
      "loss": 0.0837,
      "step": 917
    },
    {
      "epoch": 1.62,
      "grad_norm": 0.28667670488357544,
      "learning_rate": 1.3172730913613996e-05,
      "loss": 0.0632,
      "step": 918
    },
    {
      "epoch": 1.63,
      "grad_norm": 0.3482688367366791,
      "learning_rate": 1.3144975688519859e-05,
      "loss": 0.0841,
      "step": 919
    },
    {
      "epoch": 1.63,
      "grad_norm": 0.4961785674095154,
      "learning_rate": 1.3117226911787791e-05,
      "loss": 0.1161,
      "step": 920
    },
    {
      "epoch": 1.63,
      "grad_norm": 0.47871673107147217,
      "learning_rate": 1.3089484679876989e-05,
      "loss": 0.1175,
      "step": 921
    },
    {
      "epoch": 1.63,
      "grad_norm": 0.47289371490478516,
      "learning_rate": 1.3061749089223914e-05,
      "loss": 0.0603,
      "step": 922
    },
    {
      "epoch": 1.63,
      "grad_norm": 0.31602954864501953,
      "learning_rate": 1.303402023624192e-05,
      "loss": 0.0579,
      "step": 923
    },
    {
      "epoch": 1.63,
      "grad_norm": 0.25430694222450256,
      "learning_rate": 1.3006298217320948e-05,
      "loss": 0.0468,
      "step": 924
    },
    {
      "epoch": 1.64,
      "grad_norm": 0.14246149361133575,
      "learning_rate": 1.2978583128827187e-05,
      "loss": 0.0216,
      "step": 925
    },
    {
      "epoch": 1.64,
      "grad_norm": 0.22487245500087738,
      "learning_rate": 1.2950875067102719e-05,
      "loss": 0.0649,
      "step": 926
    },
    {
      "epoch": 1.64,
      "grad_norm": 0.5525780916213989,
      "learning_rate": 1.2923174128465227e-05,
      "loss": 0.0886,
      "step": 927
    },
    {
      "epoch": 1.64,
      "grad_norm": 0.3489205241203308,
      "learning_rate": 1.2895480409207608e-05,
      "loss": 0.0724,
      "step": 928
    },
    {
      "epoch": 1.64,
      "grad_norm": 0.27638673782348633,
      "learning_rate": 1.286779400559767e-05,
      "loss": 0.0701,
      "step": 929
    },
    {
      "epoch": 1.65,
      "grad_norm": 0.1930011659860611,
      "learning_rate": 1.2840115013877803e-05,
      "loss": 0.0245,
      "step": 930
    },
    {
      "epoch": 1.65,
      "grad_norm": 0.3952576518058777,
      "learning_rate": 1.2812443530264611e-05,
      "loss": 0.0835,
      "step": 931
    },
    {
      "epoch": 1.65,
      "grad_norm": 0.6183737516403198,
      "learning_rate": 1.2784779650948622e-05,
      "loss": 0.1108,
      "step": 932
    },
    {
      "epoch": 1.65,
      "grad_norm": 0.46960824728012085,
      "learning_rate": 1.2757123472093904e-05,
      "loss": 0.0448,
      "step": 933
    },
    {
      "epoch": 1.65,
      "grad_norm": 0.374906986951828,
      "learning_rate": 1.272947508983778e-05,
      "loss": 0.0527,
      "step": 934
    },
    {
      "epoch": 1.65,
      "grad_norm": 0.28407779335975647,
      "learning_rate": 1.2701834600290465e-05,
      "loss": 0.0427,
      "step": 935
    },
    {
      "epoch": 1.66,
      "grad_norm": 0.30743423104286194,
      "learning_rate": 1.2674202099534725e-05,
      "loss": 0.0842,
      "step": 936
    },
    {
      "epoch": 1.66,
      "grad_norm": 0.5731581449508667,
      "learning_rate": 1.264657768362557e-05,
      "loss": 0.0585,
      "step": 937
    },
    {
      "epoch": 1.66,
      "grad_norm": 0.7583017945289612,
      "learning_rate": 1.2618961448589903e-05,
      "loss": 0.1045,
      "step": 938
    },
    {
      "epoch": 1.66,
      "grad_norm": 0.404668927192688,
      "learning_rate": 1.2591353490426185e-05,
      "loss": 0.0889,
      "step": 939
    },
    {
      "epoch": 1.66,
      "grad_norm": 0.37973326444625854,
      "learning_rate": 1.2563753905104108e-05,
      "loss": 0.092,
      "step": 940
    },
    {
      "epoch": 1.66,
      "grad_norm": 0.31521323323249817,
      "learning_rate": 1.2536162788564253e-05,
      "loss": 0.0605,
      "step": 941
    },
    {
      "epoch": 1.67,
      "grad_norm": 0.34035030007362366,
      "learning_rate": 1.250858023671777e-05,
      "loss": 0.0387,
      "step": 942
    },
    {
      "epoch": 1.67,
      "grad_norm": 0.45016026496887207,
      "learning_rate": 1.2481006345446039e-05,
      "loss": 0.0767,
      "step": 943
    },
    {
      "epoch": 1.67,
      "grad_norm": 0.2923746705055237,
      "learning_rate": 1.2453441210600323e-05,
      "loss": 0.065,
      "step": 944
    },
    {
      "epoch": 1.67,
      "grad_norm": 0.4382145404815674,
      "learning_rate": 1.2425884928001455e-05,
      "loss": 0.053,
      "step": 945
    },
    {
      "epoch": 1.67,
      "grad_norm": 0.29467448592185974,
      "learning_rate": 1.2398337593439491e-05,
      "loss": 0.0883,
      "step": 946
    },
    {
      "epoch": 1.68,
      "grad_norm": 0.1951567381620407,
      "learning_rate": 1.2370799302673386e-05,
      "loss": 0.0285,
      "step": 947
    },
    {
      "epoch": 1.68,
      "grad_norm": 0.6156413555145264,
      "learning_rate": 1.2343270151430659e-05,
      "loss": 0.0968,
      "step": 948
    },
    {
      "epoch": 1.68,
      "grad_norm": 0.42884185910224915,
      "learning_rate": 1.2315750235407053e-05,
      "loss": 0.0855,
      "step": 949
    },
    {
      "epoch": 1.68,
      "grad_norm": 0.3351009786128998,
      "learning_rate": 1.2288239650266211e-05,
      "loss": 0.086,
      "step": 950
    },
    {
      "epoch": 1.68,
      "grad_norm": 0.4797426164150238,
      "learning_rate": 1.2260738491639342e-05,
      "loss": 0.0699,
      "step": 951
    },
    {
      "epoch": 1.68,
      "grad_norm": 0.28032705187797546,
      "learning_rate": 1.2233246855124876e-05,
      "loss": 0.068,
      "step": 952
    },
    {
      "epoch": 1.69,
      "grad_norm": 0.3798501491546631,
      "learning_rate": 1.2205764836288167e-05,
      "loss": 0.0628,
      "step": 953
    },
    {
      "epoch": 1.69,
      "grad_norm": 0.3550231456756592,
      "learning_rate": 1.2178292530661108e-05,
      "loss": 0.0501,
      "step": 954
    },
    {
      "epoch": 1.69,
      "grad_norm": 0.20574571192264557,
      "learning_rate": 1.2150830033741845e-05,
      "loss": 0.0282,
      "step": 955
    },
    {
      "epoch": 1.69,
      "grad_norm": 0.33242711424827576,
      "learning_rate": 1.2123377440994424e-05,
      "loss": 0.0772,
      "step": 956
    },
    {
      "epoch": 1.69,
      "grad_norm": 0.2370976209640503,
      "learning_rate": 1.2095934847848454e-05,
      "loss": 0.0351,
      "step": 957
    },
    {
      "epoch": 1.69,
      "grad_norm": 0.2720176577568054,
      "learning_rate": 1.2068502349698806e-05,
      "loss": 0.0396,
      "step": 958
    },
    {
      "epoch": 1.7,
      "grad_norm": 0.3753195106983185,
      "learning_rate": 1.2041080041905221e-05,
      "loss": 0.0682,
      "step": 959
    },
    {
      "epoch": 1.7,
      "grad_norm": 0.34298160672187805,
      "learning_rate": 1.2013668019792059e-05,
      "loss": 0.062,
      "step": 960
    },
    {
      "epoch": 1.7,
      "grad_norm": 0.19010630249977112,
      "learning_rate": 1.1986266378647903e-05,
      "loss": 0.0443,
      "step": 961
    },
    {
      "epoch": 1.7,
      "grad_norm": 0.3464229702949524,
      "learning_rate": 1.1958875213725242e-05,
      "loss": 0.0536,
      "step": 962
    },
    {
      "epoch": 1.7,
      "grad_norm": 0.39320218563079834,
      "learning_rate": 1.1931494620240173e-05,
      "loss": 0.0616,
      "step": 963
    },
    {
      "epoch": 1.71,
      "grad_norm": 0.5615540146827698,
      "learning_rate": 1.1904124693372027e-05,
      "loss": 0.0908,
      "step": 964
    },
    {
      "epoch": 1.71,
      "grad_norm": 0.4105425477027893,
      "learning_rate": 1.1876765528263054e-05,
      "loss": 0.0769,
      "step": 965
    },
    {
      "epoch": 1.71,
      "grad_norm": 0.3218759000301361,
      "learning_rate": 1.1849417220018108e-05,
      "loss": 0.0666,
      "step": 966
    },
    {
      "epoch": 1.71,
      "grad_norm": 0.42983943223953247,
      "learning_rate": 1.1822079863704291e-05,
      "loss": 0.0978,
      "step": 967
    },
    {
      "epoch": 1.71,
      "grad_norm": 0.42350447177886963,
      "learning_rate": 1.1794753554350635e-05,
      "loss": 0.0587,
      "step": 968
    },
    {
      "epoch": 1.71,
      "grad_norm": 0.47886720299720764,
      "learning_rate": 1.1767438386947789e-05,
      "loss": 0.1,
      "step": 969
    },
    {
      "epoch": 1.72,
      "grad_norm": 0.3799029290676117,
      "learning_rate": 1.1740134456447642e-05,
      "loss": 0.057,
      "step": 970
    },
    {
      "epoch": 1.72,
      "grad_norm": 0.15734422206878662,
      "learning_rate": 1.1712841857763043e-05,
      "loss": 0.0201,
      "step": 971
    },
    {
      "epoch": 1.72,
      "grad_norm": 0.3751939535140991,
      "learning_rate": 1.1685560685767441e-05,
      "loss": 0.0722,
      "step": 972
    },
    {
      "epoch": 1.72,
      "grad_norm": 0.32888197898864746,
      "learning_rate": 1.1658291035294565e-05,
      "loss": 0.037,
      "step": 973
    },
    {
      "epoch": 1.72,
      "grad_norm": 0.2801758646965027,
      "learning_rate": 1.16310330011381e-05,
      "loss": 0.0686,
      "step": 974
    },
    {
      "epoch": 1.72,
      "grad_norm": 0.34503892064094543,
      "learning_rate": 1.160378667805134e-05,
      "loss": 0.0668,
      "step": 975
    },
    {
      "epoch": 1.73,
      "grad_norm": 0.42055317759513855,
      "learning_rate": 1.1576552160746876e-05,
      "loss": 0.1113,
      "step": 976
    },
    {
      "epoch": 1.73,
      "grad_norm": 0.3703528344631195,
      "learning_rate": 1.1549329543896261e-05,
      "loss": 0.0666,
      "step": 977
    },
    {
      "epoch": 1.73,
      "grad_norm": 0.35604336857795715,
      "learning_rate": 1.152211892212967e-05,
      "loss": 0.065,
      "step": 978
    },
    {
      "epoch": 1.73,
      "grad_norm": 0.2423248142004013,
      "learning_rate": 1.1494920390035603e-05,
      "loss": 0.0504,
      "step": 979
    },
    {
      "epoch": 1.73,
      "grad_norm": 0.22511935234069824,
      "learning_rate": 1.1467734042160506e-05,
      "loss": 0.0546,
      "step": 980
    },
    {
      "epoch": 1.74,
      "grad_norm": 0.3228350281715393,
      "learning_rate": 1.1440559973008493e-05,
      "loss": 0.0654,
      "step": 981
    },
    {
      "epoch": 1.74,
      "grad_norm": 0.3399416506290436,
      "learning_rate": 1.141339827704099e-05,
      "loss": 0.0672,
      "step": 982
    },
    {
      "epoch": 1.74,
      "grad_norm": 0.2496933490037918,
      "learning_rate": 1.1386249048676399e-05,
      "loss": 0.0402,
      "step": 983
    },
    {
      "epoch": 1.74,
      "grad_norm": 0.6407752633094788,
      "learning_rate": 1.1359112382289807e-05,
      "loss": 0.0714,
      "step": 984
    },
    {
      "epoch": 1.74,
      "grad_norm": 0.4333494305610657,
      "learning_rate": 1.1331988372212604e-05,
      "loss": 0.0452,
      "step": 985
    },
    {
      "epoch": 1.74,
      "grad_norm": 0.2763339579105377,
      "learning_rate": 1.1304877112732209e-05,
      "loss": 0.0467,
      "step": 986
    },
    {
      "epoch": 1.75,
      "grad_norm": 0.39962539076805115,
      "learning_rate": 1.1277778698091711e-05,
      "loss": 0.1254,
      "step": 987
    },
    {
      "epoch": 1.75,
      "grad_norm": 0.5673380494117737,
      "learning_rate": 1.1250693222489538e-05,
      "loss": 0.0889,
      "step": 988
    },
    {
      "epoch": 1.75,
      "grad_norm": 0.3895013928413391,
      "learning_rate": 1.1223620780079155e-05,
      "loss": 0.0657,
      "step": 989
    },
    {
      "epoch": 1.75,
      "grad_norm": 0.32049494981765747,
      "learning_rate": 1.1196561464968715e-05,
      "loss": 0.0451,
      "step": 990
    },
    {
      "epoch": 1.75,
      "grad_norm": 0.797935426235199,
      "learning_rate": 1.1169515371220729e-05,
      "loss": 0.1373,
      "step": 991
    },
    {
      "epoch": 1.75,
      "grad_norm": 0.1642259806394577,
      "learning_rate": 1.114248259285177e-05,
      "loss": 0.0253,
      "step": 992
    },
    {
      "epoch": 1.76,
      "grad_norm": 0.14906607568264008,
      "learning_rate": 1.1115463223832101e-05,
      "loss": 0.0221,
      "step": 993
    },
    {
      "epoch": 1.76,
      "grad_norm": 0.4249594807624817,
      "learning_rate": 1.1088457358085384e-05,
      "loss": 0.0663,
      "step": 994
    },
    {
      "epoch": 1.76,
      "eval_loss": 0.07590385526418686,
      "eval_runtime": 14.6825,
      "eval_samples_per_second": 32.488,
      "eval_steps_per_second": 8.173,
      "step": 994
    },
    {
      "epoch": 1.76,
      "grad_norm": 0.29456090927124023,
      "learning_rate": 1.106146508948835e-05,
      "loss": 0.0277,
      "step": 995
    },
    {
      "epoch": 1.76,
      "grad_norm": 0.42705532908439636,
      "learning_rate": 1.1034486511870444e-05,
      "loss": 0.0707,
      "step": 996
    },
    {
      "epoch": 1.76,
      "grad_norm": 0.43404194712638855,
      "learning_rate": 1.1007521719013538e-05,
      "loss": 0.1038,
      "step": 997
    },
    {
      "epoch": 1.77,
      "grad_norm": 0.17568662762641907,
      "learning_rate": 1.0980570804651568e-05,
      "loss": 0.0219,
      "step": 998
    },
    {
      "epoch": 1.77,
      "grad_norm": 0.42081278562545776,
      "learning_rate": 1.095363386247024e-05,
      "loss": 0.0787,
      "step": 999
    },
    {
      "epoch": 1.77,
      "grad_norm": 0.34417223930358887,
      "learning_rate": 1.0926710986106691e-05,
      "loss": 0.0368,
      "step": 1000
    },
    {
      "epoch": 1.77,
      "grad_norm": 0.4904881417751312,
      "learning_rate": 1.0899802269149151e-05,
      "loss": 0.1149,
      "step": 1001
    },
    {
      "epoch": 1.77,
      "grad_norm": 0.6708604097366333,
      "learning_rate": 1.0872907805136637e-05,
      "loss": 0.1035,
      "step": 1002
    },
    {
      "epoch": 1.77,
      "grad_norm": 0.2843027710914612,
      "learning_rate": 1.0846027687558627e-05,
      "loss": 0.0774,
      "step": 1003
    },
    {
      "epoch": 1.78,
      "grad_norm": 0.3383296728134155,
      "learning_rate": 1.081916200985471e-05,
      "loss": 0.0675,
      "step": 1004
    },
    {
      "epoch": 1.78,
      "grad_norm": 0.3500230014324188,
      "learning_rate": 1.0792310865414307e-05,
      "loss": 0.065,
      "step": 1005
    },
    {
      "epoch": 1.78,
      "grad_norm": 0.40395185351371765,
      "learning_rate": 1.0765474347576288e-05,
      "loss": 0.0966,
      "step": 1006
    },
    {
      "epoch": 1.78,
      "grad_norm": 0.3098698854446411,
      "learning_rate": 1.07386525496287e-05,
      "loss": 0.0543,
      "step": 1007
    },
    {
      "epoch": 1.78,
      "grad_norm": 0.3940174877643585,
      "learning_rate": 1.0711845564808418e-05,
      "loss": 0.1036,
      "step": 1008
    },
    {
      "epoch": 1.79,
      "grad_norm": 0.3571736216545105,
      "learning_rate": 1.0685053486300814e-05,
      "loss": 0.0853,
      "step": 1009
    },
    {
      "epoch": 1.79,
      "grad_norm": 0.36929792165756226,
      "learning_rate": 1.0658276407239462e-05,
      "loss": 0.0718,
      "step": 1010
    },
    {
      "epoch": 1.79,
      "grad_norm": 0.24736067652702332,
      "learning_rate": 1.063151442070577e-05,
      "loss": 0.0468,
      "step": 1011
    },
    {
      "epoch": 1.79,
      "grad_norm": 0.3313748836517334,
      "learning_rate": 1.0604767619728707e-05,
      "loss": 0.0501,
      "step": 1012
    },
    {
      "epoch": 1.79,
      "grad_norm": 0.2802097201347351,
      "learning_rate": 1.0578036097284442e-05,
      "loss": 0.0567,
      "step": 1013
    },
    {
      "epoch": 1.79,
      "grad_norm": 0.3235451281070709,
      "learning_rate": 1.0551319946296027e-05,
      "loss": 0.0625,
      "step": 1014
    },
    {
      "epoch": 1.8,
      "grad_norm": 0.5376477837562561,
      "learning_rate": 1.05246192596331e-05,
      "loss": 0.1033,
      "step": 1015
    },
    {
      "epoch": 1.8,
      "grad_norm": 0.21468472480773926,
      "learning_rate": 1.0497934130111526e-05,
      "loss": 0.0592,
      "step": 1016
    },
    {
      "epoch": 1.8,
      "grad_norm": 0.22704298794269562,
      "learning_rate": 1.0471264650493093e-05,
      "loss": 0.0434,
      "step": 1017
    },
    {
      "epoch": 1.8,
      "grad_norm": 0.35349923372268677,
      "learning_rate": 1.0444610913485195e-05,
      "loss": 0.0803,
      "step": 1018
    },
    {
      "epoch": 1.8,
      "grad_norm": 0.2493237853050232,
      "learning_rate": 1.0417973011740492e-05,
      "loss": 0.0463,
      "step": 1019
    },
    {
      "epoch": 1.8,
      "grad_norm": 0.6416348814964294,
      "learning_rate": 1.0391351037856604e-05,
      "loss": 0.1573,
      "step": 1020
    },
    {
      "epoch": 1.81,
      "grad_norm": 0.3609257638454437,
      "learning_rate": 1.036474508437579e-05,
      "loss": 0.0679,
      "step": 1021
    },
    {
      "epoch": 1.81,
      "grad_norm": 0.21030279994010925,
      "learning_rate": 1.0338155243784606e-05,
      "loss": 0.041,
      "step": 1022
    },
    {
      "epoch": 1.81,
      "grad_norm": 0.2578910291194916,
      "learning_rate": 1.0311581608513603e-05,
      "loss": 0.0619,
      "step": 1023
    },
    {
      "epoch": 1.81,
      "grad_norm": 0.476671427488327,
      "learning_rate": 1.0285024270937003e-05,
      "loss": 0.0773,
      "step": 1024
    },
    {
      "epoch": 1.81,
      "grad_norm": 0.318682998418808,
      "learning_rate": 1.0258483323372365e-05,
      "loss": 0.0889,
      "step": 1025
    },
    {
      "epoch": 1.82,
      "grad_norm": 0.24907314777374268,
      "learning_rate": 1.0231958858080291e-05,
      "loss": 0.0712,
      "step": 1026
    },
    {
      "epoch": 1.82,
      "grad_norm": 0.30080917477607727,
      "learning_rate": 1.0205450967264067e-05,
      "loss": 0.0513,
      "step": 1027
    },
    {
      "epoch": 1.82,
      "grad_norm": 0.33363106846809387,
      "learning_rate": 1.017895974306938e-05,
      "loss": 0.0784,
      "step": 1028
    },
    {
      "epoch": 1.82,
      "grad_norm": 0.48355334997177124,
      "learning_rate": 1.0152485277583975e-05,
      "loss": 0.1164,
      "step": 1029
    },
    {
      "epoch": 1.82,
      "grad_norm": 0.3023397624492645,
      "learning_rate": 1.0126027662837332e-05,
      "loss": 0.0474,
      "step": 1030
    },
    {
      "epoch": 1.82,
      "grad_norm": 0.41697564721107483,
      "learning_rate": 1.0099586990800375e-05,
      "loss": 0.0745,
      "step": 1031
    },
    {
      "epoch": 1.83,
      "grad_norm": 0.28540369868278503,
      "learning_rate": 1.0073163353385114e-05,
      "loss": 0.0348,
      "step": 1032
    },
    {
      "epoch": 1.83,
      "grad_norm": 0.3662286698818207,
      "learning_rate": 1.0046756842444356e-05,
      "loss": 0.0538,
      "step": 1033
    },
    {
      "epoch": 1.83,
      "grad_norm": 0.2947501242160797,
      "learning_rate": 1.0020367549771367e-05,
      "loss": 0.041,
      "step": 1034
    },
    {
      "epoch": 1.83,
      "grad_norm": 0.5739675760269165,
      "learning_rate": 9.993995567099557e-06,
      "loss": 0.1022,
      "step": 1035
    },
    {
      "epoch": 1.83,
      "grad_norm": 0.2689216434955597,
      "learning_rate": 9.967640986102181e-06,
      "loss": 0.0271,
      "step": 1036
    },
    {
      "epoch": 1.83,
      "grad_norm": 0.48024898767471313,
      "learning_rate": 9.941303898391973e-06,
      "loss": 0.1125,
      "step": 1037
    },
    {
      "epoch": 1.84,
      "grad_norm": 0.6236119866371155,
      "learning_rate": 9.914984395520885e-06,
      "loss": 0.1276,
      "step": 1038
    },
    {
      "epoch": 1.84,
      "grad_norm": 0.3640107214450836,
      "learning_rate": 9.888682568979735e-06,
      "loss": 0.0949,
      "step": 1039
    },
    {
      "epoch": 1.84,
      "grad_norm": 0.35603880882263184,
      "learning_rate": 9.862398510197875e-06,
      "loss": 0.0725,
      "step": 1040
    },
    {
      "epoch": 1.84,
      "grad_norm": 0.39341041445732117,
      "learning_rate": 9.836132310542926e-06,
      "loss": 0.077,
      "step": 1041
    },
    {
      "epoch": 1.84,
      "grad_norm": 0.3216860890388489,
      "learning_rate": 9.809884061320407e-06,
      "loss": 0.0753,
      "step": 1042
    },
    {
      "epoch": 1.85,
      "grad_norm": 0.3560468256473541,
      "learning_rate": 9.783653853773438e-06,
      "loss": 0.0741,
      "step": 1043
    },
    {
      "epoch": 1.85,
      "grad_norm": 0.2818620502948761,
      "learning_rate": 9.757441779082434e-06,
      "loss": 0.0368,
      "step": 1044
    },
    {
      "epoch": 1.85,
      "grad_norm": 0.3345241844654083,
      "learning_rate": 9.731247928364766e-06,
      "loss": 0.0643,
      "step": 1045
    },
    {
      "epoch": 1.85,
      "grad_norm": 0.5847212076187134,
      "learning_rate": 9.705072392674459e-06,
      "loss": 0.073,
      "step": 1046
    },
    {
      "epoch": 1.85,
      "grad_norm": 0.1948455423116684,
      "learning_rate": 9.678915263001884e-06,
      "loss": 0.0581,
      "step": 1047
    },
    {
      "epoch": 1.85,
      "grad_norm": 0.2879828214645386,
      "learning_rate": 9.65277663027341e-06,
      "loss": 0.0737,
      "step": 1048
    },
    {
      "epoch": 1.86,
      "grad_norm": 0.2977559268474579,
      "learning_rate": 9.626656585351127e-06,
      "loss": 0.0384,
      "step": 1049
    },
    {
      "epoch": 1.86,
      "grad_norm": 0.2820340394973755,
      "learning_rate": 9.600555219032493e-06,
      "loss": 0.0617,
      "step": 1050
    },
    {
      "epoch": 1.86,
      "grad_norm": 0.2548690736293793,
      "learning_rate": 9.574472622050037e-06,
      "loss": 0.0452,
      "step": 1051
    },
    {
      "epoch": 1.86,
      "grad_norm": 0.41625139117240906,
      "learning_rate": 9.548408885071071e-06,
      "loss": 0.059,
      "step": 1052
    },
    {
      "epoch": 1.86,
      "grad_norm": 0.35167020559310913,
      "learning_rate": 9.52236409869731e-06,
      "loss": 0.0694,
      "step": 1053
    },
    {
      "epoch": 1.86,
      "grad_norm": 0.3224475681781769,
      "learning_rate": 9.496338353464613e-06,
      "loss": 0.0964,
      "step": 1054
    },
    {
      "epoch": 1.87,
      "grad_norm": 0.23356600105762482,
      "learning_rate": 9.470331739842645e-06,
      "loss": 0.0463,
      "step": 1055
    },
    {
      "epoch": 1.87,
      "grad_norm": 0.29073721170425415,
      "learning_rate": 9.444344348234566e-06,
      "loss": 0.051,
      "step": 1056
    },
    {
      "epoch": 1.87,
      "grad_norm": 0.4316430985927582,
      "learning_rate": 9.418376268976727e-06,
      "loss": 0.1106,
      "step": 1057
    },
    {
      "epoch": 1.87,
      "grad_norm": 0.3090851306915283,
      "learning_rate": 9.392427592338318e-06,
      "loss": 0.0568,
      "step": 1058
    },
    {
      "epoch": 1.87,
      "grad_norm": 0.4424888789653778,
      "learning_rate": 9.36649840852112e-06,
      "loss": 0.0554,
      "step": 1059
    },
    {
      "epoch": 1.88,
      "grad_norm": 0.2993427515029907,
      "learning_rate": 9.340588807659128e-06,
      "loss": 0.0399,
      "step": 1060
    },
    {
      "epoch": 1.88,
      "grad_norm": 0.26381632685661316,
      "learning_rate": 9.314698879818264e-06,
      "loss": 0.035,
      "step": 1061
    },
    {
      "epoch": 1.88,
      "grad_norm": 0.31222546100616455,
      "learning_rate": 9.288828714996086e-06,
      "loss": 0.0812,
      "step": 1062
    },
    {
      "epoch": 1.88,
      "grad_norm": 0.26161107420921326,
      "learning_rate": 9.26297840312142e-06,
      "loss": 0.031,
      "step": 1063
    },
    {
      "epoch": 1.88,
      "grad_norm": 0.298797607421875,
      "learning_rate": 9.237148034054107e-06,
      "loss": 0.0428,
      "step": 1064
    },
    {
      "epoch": 1.88,
      "grad_norm": 0.46619024872779846,
      "learning_rate": 9.211337697584654e-06,
      "loss": 0.0677,
      "step": 1065
    },
    {
      "epoch": 1.89,
      "grad_norm": 0.3210253119468689,
      "learning_rate": 9.185547483433923e-06,
      "loss": 0.0312,
      "step": 1066
    },
    {
      "epoch": 1.89,
      "grad_norm": 0.6666526794433594,
      "learning_rate": 9.159777481252838e-06,
      "loss": 0.1368,
      "step": 1067
    },
    {
      "epoch": 1.89,
      "grad_norm": 0.36307379603385925,
      "learning_rate": 9.134027780622069e-06,
      "loss": 0.0714,
      "step": 1068
    },
    {
      "epoch": 1.89,
      "grad_norm": 0.6304948329925537,
      "learning_rate": 9.108298471051695e-06,
      "loss": 0.0744,
      "step": 1069
    },
    {
      "epoch": 1.89,
      "grad_norm": 0.5147308707237244,
      "learning_rate": 9.082589641980931e-06,
      "loss": 0.1044,
      "step": 1070
    },
    {
      "epoch": 1.89,
      "grad_norm": 0.22846105694770813,
      "learning_rate": 9.056901382777784e-06,
      "loss": 0.0356,
      "step": 1071
    },
    {
      "epoch": 1.9,
      "grad_norm": 0.6081435084342957,
      "learning_rate": 9.031233782738765e-06,
      "loss": 0.0522,
      "step": 1072
    },
    {
      "epoch": 1.9,
      "grad_norm": 0.5012447237968445,
      "learning_rate": 9.00558693108858e-06,
      "loss": 0.0851,
      "step": 1073
    },
    {
      "epoch": 1.9,
      "grad_norm": 0.43411707878112793,
      "learning_rate": 8.979960916979782e-06,
      "loss": 0.1041,
      "step": 1074
    },
    {
      "epoch": 1.9,
      "grad_norm": 0.5360767841339111,
      "learning_rate": 8.954355829492522e-06,
      "loss": 0.0737,
      "step": 1075
    },
    {
      "epoch": 1.9,
      "grad_norm": 0.4742479622364044,
      "learning_rate": 8.928771757634179e-06,
      "loss": 0.0596,
      "step": 1076
    },
    {
      "epoch": 1.91,
      "grad_norm": 0.4261029362678528,
      "learning_rate": 8.903208790339094e-06,
      "loss": 0.0951,
      "step": 1077
    },
    {
      "epoch": 1.91,
      "grad_norm": 0.371076375246048,
      "learning_rate": 8.877667016468251e-06,
      "loss": 0.0647,
      "step": 1078
    },
    {
      "epoch": 1.91,
      "grad_norm": 0.3911438286304474,
      "learning_rate": 8.852146524808945e-06,
      "loss": 0.0749,
      "step": 1079
    },
    {
      "epoch": 1.91,
      "grad_norm": 0.46002396941185,
      "learning_rate": 8.826647404074497e-06,
      "loss": 0.064,
      "step": 1080
    },
    {
      "epoch": 1.91,
      "grad_norm": 0.4756135046482086,
      "learning_rate": 8.80116974290395e-06,
      "loss": 0.0859,
      "step": 1081
    },
    {
      "epoch": 1.91,
      "grad_norm": 0.2722984850406647,
      "learning_rate": 8.775713629861726e-06,
      "loss": 0.0712,
      "step": 1082
    },
    {
      "epoch": 1.92,
      "grad_norm": 0.26421287655830383,
      "learning_rate": 8.750279153437375e-06,
      "loss": 0.0516,
      "step": 1083
    },
    {
      "epoch": 1.92,
      "grad_norm": 0.3331577777862549,
      "learning_rate": 8.72486640204521e-06,
      "loss": 0.0939,
      "step": 1084
    },
    {
      "epoch": 1.92,
      "grad_norm": 0.4229680597782135,
      "learning_rate": 8.699475464024022e-06,
      "loss": 0.0977,
      "step": 1085
    },
    {
      "epoch": 1.92,
      "grad_norm": 0.283701628446579,
      "learning_rate": 8.6741064276368e-06,
      "loss": 0.0516,
      "step": 1086
    },
    {
      "epoch": 1.92,
      "grad_norm": 0.24555976688861847,
      "learning_rate": 8.648759381070369e-06,
      "loss": 0.0374,
      "step": 1087
    },
    {
      "epoch": 1.92,
      "grad_norm": 0.461603581905365,
      "learning_rate": 8.623434412435144e-06,
      "loss": 0.0675,
      "step": 1088
    },
    {
      "epoch": 1.93,
      "grad_norm": 0.24331988394260406,
      "learning_rate": 8.59813160976477e-06,
      "loss": 0.0294,
      "step": 1089
    },
    {
      "epoch": 1.93,
      "grad_norm": 0.3258945047855377,
      "learning_rate": 8.572851061015842e-06,
      "loss": 0.0708,
      "step": 1090
    },
    {
      "epoch": 1.93,
      "grad_norm": 0.2624408006668091,
      "learning_rate": 8.547592854067616e-06,
      "loss": 0.0499,
      "step": 1091
    },
    {
      "epoch": 1.93,
      "grad_norm": 0.17459358274936676,
      "learning_rate": 8.522357076721658e-06,
      "loss": 0.0334,
      "step": 1092
    },
    {
      "epoch": 1.93,
      "grad_norm": 0.4228392243385315,
      "learning_rate": 8.497143816701592e-06,
      "loss": 0.0576,
      "step": 1093
    },
    {
      "epoch": 1.94,
      "grad_norm": 0.3577699661254883,
      "learning_rate": 8.471953161652744e-06,
      "loss": 0.0445,
      "step": 1094
    },
    {
      "epoch": 1.94,
      "grad_norm": 0.34772756695747375,
      "learning_rate": 8.44678519914187e-06,
      "loss": 0.0548,
      "step": 1095
    },
    {
      "epoch": 1.94,
      "grad_norm": 0.45824792981147766,
      "learning_rate": 8.421640016656857e-06,
      "loss": 0.0515,
      "step": 1096
    },
    {
      "epoch": 1.94,
      "grad_norm": 0.1746465563774109,
      "learning_rate": 8.396517701606377e-06,
      "loss": 0.0253,
      "step": 1097
    },
    {
      "epoch": 1.94,
      "grad_norm": 0.2786015272140503,
      "learning_rate": 8.371418341319643e-06,
      "loss": 0.0411,
      "step": 1098
    },
    {
      "epoch": 1.94,
      "grad_norm": 0.2789120674133301,
      "learning_rate": 8.346342023046049e-06,
      "loss": 0.0285,
      "step": 1099
    },
    {
      "epoch": 1.95,
      "grad_norm": 0.4477371275424957,
      "learning_rate": 8.321288833954896e-06,
      "loss": 0.0677,
      "step": 1100
    },
    {
      "epoch": 1.95,
      "grad_norm": 0.29438501596450806,
      "learning_rate": 8.296258861135098e-06,
      "loss": 0.0513,
      "step": 1101
    },
    {
      "epoch": 1.95,
      "grad_norm": 0.25353726744651794,
      "learning_rate": 8.271252191594846e-06,
      "loss": 0.034,
      "step": 1102
    },
    {
      "epoch": 1.95,
      "grad_norm": 0.39505115151405334,
      "learning_rate": 8.246268912261344e-06,
      "loss": 0.0748,
      "step": 1103
    },
    {
      "epoch": 1.95,
      "grad_norm": 0.2933278977870941,
      "learning_rate": 8.221309109980472e-06,
      "loss": 0.0497,
      "step": 1104
    },
    {
      "epoch": 1.95,
      "grad_norm": 0.31416723132133484,
      "learning_rate": 8.196372871516502e-06,
      "loss": 0.0696,
      "step": 1105
    },
    {
      "epoch": 1.96,
      "grad_norm": 0.5482780933380127,
      "learning_rate": 8.17146028355181e-06,
      "loss": 0.0726,
      "step": 1106
    },
    {
      "epoch": 1.96,
      "grad_norm": 0.7449930906295776,
      "learning_rate": 8.146571432686545e-06,
      "loss": 0.0913,
      "step": 1107
    },
    {
      "epoch": 1.96,
      "grad_norm": 0.17089535295963287,
      "learning_rate": 8.121706405438346e-06,
      "loss": 0.0167,
      "step": 1108
    },
    {
      "epoch": 1.96,
      "grad_norm": 0.261944979429245,
      "learning_rate": 8.09686528824203e-06,
      "loss": 0.0373,
      "step": 1109
    },
    {
      "epoch": 1.96,
      "grad_norm": 0.3364504277706146,
      "learning_rate": 8.072048167449306e-06,
      "loss": 0.0729,
      "step": 1110
    },
    {
      "epoch": 1.97,
      "grad_norm": 0.5227497220039368,
      "learning_rate": 8.047255129328467e-06,
      "loss": 0.1096,
      "step": 1111
    },
    {
      "epoch": 1.97,
      "grad_norm": 0.2813100814819336,
      "learning_rate": 8.0224862600641e-06,
      "loss": 0.0376,
      "step": 1112
    },
    {
      "epoch": 1.97,
      "grad_norm": 0.4989968538284302,
      "learning_rate": 7.997741645756764e-06,
      "loss": 0.0804,
      "step": 1113
    },
    {
      "epoch": 1.97,
      "grad_norm": 0.492436021566391,
      "learning_rate": 7.9730213724227e-06,
      "loss": 0.072,
      "step": 1114
    },
    {
      "epoch": 1.97,
      "grad_norm": 0.46163979172706604,
      "learning_rate": 7.948325525993545e-06,
      "loss": 0.0367,
      "step": 1115
    },
    {
      "epoch": 1.97,
      "grad_norm": 0.41931623220443726,
      "learning_rate": 7.923654192316023e-06,
      "loss": 0.0279,
      "step": 1116
    },
    {
      "epoch": 1.98,
      "grad_norm": 1.1422687768936157,
      "learning_rate": 7.899007457151658e-06,
      "loss": 0.1581,
      "step": 1117
    },
    {
      "epoch": 1.98,
      "grad_norm": 0.41459235548973083,
      "learning_rate": 7.874385406176445e-06,
      "loss": 0.1014,
      "step": 1118
    },
    {
      "epoch": 1.98,
      "grad_norm": 0.42844435572624207,
      "learning_rate": 7.849788124980583e-06,
      "loss": 0.0818,
      "step": 1119
    },
    {
      "epoch": 1.98,
      "grad_norm": 0.2925580143928528,
      "learning_rate": 7.82521569906817e-06,
      "loss": 0.0554,
      "step": 1120
    },
    {
      "epoch": 1.98,
      "grad_norm": 0.24082061648368835,
      "learning_rate": 7.800668213856897e-06,
      "loss": 0.0319,
      "step": 1121
    },
    {
      "epoch": 1.98,
      "grad_norm": 0.4475957453250885,
      "learning_rate": 7.776145754677773e-06,
      "loss": 0.0769,
      "step": 1122
    },
    {
      "epoch": 1.99,
      "grad_norm": 0.2008124142885208,
      "learning_rate": 7.751648406774778e-06,
      "loss": 0.0217,
      "step": 1123
    },
    {
      "epoch": 1.99,
      "grad_norm": 0.38693517446517944,
      "learning_rate": 7.727176255304633e-06,
      "loss": 0.1004,
      "step": 1124
    },
    {
      "epoch": 1.99,
      "grad_norm": 0.429131418466568,
      "learning_rate": 7.70272938533647e-06,
      "loss": 0.0592,
      "step": 1125
    },
    {
      "epoch": 1.99,
      "grad_norm": 0.3874954283237457,
      "learning_rate": 7.678307881851517e-06,
      "loss": 0.0673,
      "step": 1126
    },
    {
      "epoch": 1.99,
      "grad_norm": 0.41746100783348083,
      "learning_rate": 7.653911829742858e-06,
      "loss": 0.1241,
      "step": 1127
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.4857166111469269,
      "learning_rate": 7.629541313815057e-06,
      "loss": 0.0897,
      "step": 1128
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.24189823865890503,
      "learning_rate": 7.605196418783949e-06,
      "loss": 0.0356,
      "step": 1129
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.2707352936267853,
      "learning_rate": 7.580877229276302e-06,
      "loss": 0.0764,
      "step": 1130
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.2668614685535431,
      "learning_rate": 7.556583829829512e-06,
      "loss": 0.0697,
      "step": 1131
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.26246827840805054,
      "learning_rate": 7.532316304891327e-06,
      "loss": 0.0863,
      "step": 1132
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.2628766894340515,
      "learning_rate": 7.508074738819569e-06,
      "loss": 0.0413,
      "step": 1133
    },
    {
      "epoch": 2.01,
      "grad_norm": 0.2973581850528717,
      "learning_rate": 7.483859215881799e-06,
      "loss": 0.0349,
      "step": 1134
    },
    {
      "epoch": 2.01,
      "grad_norm": 0.506683886051178,
      "learning_rate": 7.459669820255068e-06,
      "loss": 0.0772,
      "step": 1135
    },
    {
      "epoch": 2.01,
      "grad_norm": 0.271662175655365,
      "learning_rate": 7.4355066360255916e-06,
      "loss": 0.0447,
      "step": 1136
    },
    {
      "epoch": 2.01,
      "eval_loss": 0.07106287032365799,
      "eval_runtime": 14.6705,
      "eval_samples_per_second": 32.514,
      "eval_steps_per_second": 8.18,
      "step": 1136
    },
    {
      "epoch": 2.01,
      "grad_norm": 0.5070781707763672,
      "learning_rate": 7.411369747188469e-06,
      "loss": 0.0572,
      "step": 1137
    },
    {
      "epoch": 2.01,
      "grad_norm": 0.2998686730861664,
      "learning_rate": 7.387259237647408e-06,
      "loss": 0.0451,
      "step": 1138
    },
    {
      "epoch": 2.02,
      "grad_norm": 0.4732572138309479,
      "learning_rate": 7.3631751912143954e-06,
      "loss": 0.0832,
      "step": 1139
    },
    {
      "epoch": 2.02,
      "grad_norm": 0.4027310311794281,
      "learning_rate": 7.339117691609454e-06,
      "loss": 0.0695,
      "step": 1140
    },
    {
      "epoch": 2.02,
      "grad_norm": 0.37893640995025635,
      "learning_rate": 7.3150868224603034e-06,
      "loss": 0.0562,
      "step": 1141
    },
    {
      "epoch": 2.02,
      "grad_norm": 0.3249414265155792,
      "learning_rate": 7.2910826673020954e-06,
      "loss": 0.0689,
      "step": 1142
    },
    {
      "epoch": 2.02,
      "grad_norm": 0.36183980107307434,
      "learning_rate": 7.267105309577133e-06,
      "loss": 0.0906,
      "step": 1143
    },
    {
      "epoch": 2.02,
      "grad_norm": 0.24284303188323975,
      "learning_rate": 7.243154832634547e-06,
      "loss": 0.0557,
      "step": 1144
    },
    {
      "epoch": 2.03,
      "grad_norm": 0.45840340852737427,
      "learning_rate": 7.21923131973005e-06,
      "loss": 0.0795,
      "step": 1145
    },
    {
      "epoch": 2.03,
      "grad_norm": 0.5642706751823425,
      "learning_rate": 7.195334854025607e-06,
      "loss": 0.1057,
      "step": 1146
    },
    {
      "epoch": 2.03,
      "grad_norm": 0.8529704213142395,
      "learning_rate": 7.171465518589153e-06,
      "loss": 0.1285,
      "step": 1147
    },
    {
      "epoch": 2.03,
      "grad_norm": 0.4283912777900696,
      "learning_rate": 7.147623396394347e-06,
      "loss": 0.0791,
      "step": 1148
    },
    {
      "epoch": 2.03,
      "grad_norm": 0.4282079041004181,
      "learning_rate": 7.123808570320212e-06,
      "loss": 0.0956,
      "step": 1149
    },
    {
      "epoch": 2.03,
      "grad_norm": 0.21502268314361572,
      "learning_rate": 7.100021123150918e-06,
      "loss": 0.0315,
      "step": 1150
    },
    {
      "epoch": 2.04,
      "grad_norm": 0.2664368450641632,
      "learning_rate": 7.076261137575441e-06,
      "loss": 0.0751,
      "step": 1151
    },
    {
      "epoch": 2.04,
      "grad_norm": 0.47574445605278015,
      "learning_rate": 7.052528696187296e-06,
      "loss": 0.0943,
      "step": 1152
    },
    {
      "epoch": 2.04,
      "grad_norm": 0.22526611387729645,
      "learning_rate": 7.028823881484269e-06,
      "loss": 0.0429,
      "step": 1153
    },
    {
      "epoch": 2.04,
      "grad_norm": 0.20425187051296234,
      "learning_rate": 7.005146775868086e-06,
      "loss": 0.0302,
      "step": 1154
    },
    {
      "epoch": 2.04,
      "grad_norm": 0.2875848412513733,
      "learning_rate": 6.981497461644177e-06,
      "loss": 0.0223,
      "step": 1155
    },
    {
      "epoch": 2.05,
      "grad_norm": 0.20630909502506256,
      "learning_rate": 6.957876021021344e-06,
      "loss": 0.0651,
      "step": 1156
    },
    {
      "epoch": 2.05,
      "grad_norm": 0.216980442404747,
      "learning_rate": 6.9342825361114975e-06,
      "loss": 0.0295,
      "step": 1157
    },
    {
      "epoch": 2.05,
      "grad_norm": 0.28700754046440125,
      "learning_rate": 6.910717088929385e-06,
      "loss": 0.0541,
      "step": 1158
    },
    {
      "epoch": 2.05,
      "grad_norm": 0.4029622972011566,
      "learning_rate": 6.887179761392283e-06,
      "loss": 0.0785,
      "step": 1159
    },
    {
      "epoch": 2.05,
      "grad_norm": 0.28533175587654114,
      "learning_rate": 6.863670635319715e-06,
      "loss": 0.0369,
      "step": 1160
    },
    {
      "epoch": 2.05,
      "grad_norm": 0.3500222861766815,
      "learning_rate": 6.8401897924331705e-06,
      "loss": 0.0386,
      "step": 1161
    },
    {
      "epoch": 2.06,
      "grad_norm": 0.24947623908519745,
      "learning_rate": 6.816737314355826e-06,
      "loss": 0.0387,
      "step": 1162
    },
    {
      "epoch": 2.06,
      "grad_norm": 0.2996102571487427,
      "learning_rate": 6.793313282612257e-06,
      "loss": 0.0345,
      "step": 1163
    },
    {
      "epoch": 2.06,
      "grad_norm": 0.39841368794441223,
      "learning_rate": 6.769917778628165e-06,
      "loss": 0.0715,
      "step": 1164
    },
    {
      "epoch": 2.06,
      "grad_norm": 0.3743310272693634,
      "learning_rate": 6.746550883730068e-06,
      "loss": 0.0588,
      "step": 1165
    },
    {
      "epoch": 2.06,
      "grad_norm": 0.3841390013694763,
      "learning_rate": 6.723212679145039e-06,
      "loss": 0.0622,
      "step": 1166
    },
    {
      "epoch": 2.06,
      "grad_norm": 0.503218412399292,
      "learning_rate": 6.699903246000418e-06,
      "loss": 0.0527,
      "step": 1167
    },
    {
      "epoch": 2.07,
      "grad_norm": 0.670925498008728,
      "learning_rate": 6.676622665323538e-06,
      "loss": 0.0668,
      "step": 1168
    },
    {
      "epoch": 2.07,
      "grad_norm": 0.4905701279640198,
      "learning_rate": 6.653371018041438e-06,
      "loss": 0.0961,
      "step": 1169
    },
    {
      "epoch": 2.07,
      "grad_norm": 0.5032178163528442,
      "learning_rate": 6.630148384980567e-06,
      "loss": 0.0959,
      "step": 1170
    },
    {
      "epoch": 2.07,
      "grad_norm": 0.5046835541725159,
      "learning_rate": 6.60695484686652e-06,
      "loss": 0.0575,
      "step": 1171
    },
    {
      "epoch": 2.07,
      "grad_norm": 0.3424021005630493,
      "learning_rate": 6.5837904843237624e-06,
      "loss": 0.0474,
      "step": 1172
    },
    {
      "epoch": 2.08,
      "grad_norm": 0.5579243302345276,
      "learning_rate": 6.560655377875328e-06,
      "loss": 0.0686,
      "step": 1173
    },
    {
      "epoch": 2.08,
      "grad_norm": 0.42658042907714844,
      "learning_rate": 6.537549607942572e-06,
      "loss": 0.1095,
      "step": 1174
    },
    {
      "epoch": 2.08,
      "grad_norm": 0.43908706307411194,
      "learning_rate": 6.514473254844833e-06,
      "loss": 0.088,
      "step": 1175
    },
    {
      "epoch": 2.08,
      "grad_norm": 0.28826475143432617,
      "learning_rate": 6.491426398799232e-06,
      "loss": 0.0598,
      "step": 1176
    },
    {
      "epoch": 2.08,
      "grad_norm": 0.5978088974952698,
      "learning_rate": 6.468409119920339e-06,
      "loss": 0.1118,
      "step": 1177
    },
    {
      "epoch": 2.08,
      "grad_norm": 0.4633474349975586,
      "learning_rate": 6.445421498219905e-06,
      "loss": 0.0841,
      "step": 1178
    },
    {
      "epoch": 2.09,
      "grad_norm": 0.2268844097852707,
      "learning_rate": 6.422463613606587e-06,
      "loss": 0.0484,
      "step": 1179
    },
    {
      "epoch": 2.09,
      "grad_norm": 0.6244954466819763,
      "learning_rate": 6.399535545885673e-06,
      "loss": 0.097,
      "step": 1180
    },
    {
      "epoch": 2.09,
      "grad_norm": 0.25237593054771423,
      "learning_rate": 6.376637374758807e-06,
      "loss": 0.061,
      "step": 1181
    },
    {
      "epoch": 2.09,
      "grad_norm": 0.4511301815509796,
      "learning_rate": 6.353769179823713e-06,
      "loss": 0.0661,
      "step": 1182
    },
    {
      "epoch": 2.09,
      "grad_norm": 0.24250440299510956,
      "learning_rate": 6.330931040573896e-06,
      "loss": 0.0559,
      "step": 1183
    },
    {
      "epoch": 2.09,
      "grad_norm": 0.40236830711364746,
      "learning_rate": 6.308123036398389e-06,
      "loss": 0.0569,
      "step": 1184
    },
    {
      "epoch": 2.1,
      "grad_norm": 0.29502615332603455,
      "learning_rate": 6.285345246581483e-06,
      "loss": 0.053,
      "step": 1185
    },
    {
      "epoch": 2.1,
      "grad_norm": 0.2569676339626312,
      "learning_rate": 6.2625977503024175e-06,
      "loss": 0.0307,
      "step": 1186
    },
    {
      "epoch": 2.1,
      "grad_norm": 0.2814861536026001,
      "learning_rate": 6.239880626635151e-06,
      "loss": 0.0427,
      "step": 1187
    },
    {
      "epoch": 2.1,
      "grad_norm": 0.32552242279052734,
      "learning_rate": 6.2171939545480424e-06,
      "loss": 0.0566,
      "step": 1188
    },
    {
      "epoch": 2.1,
      "grad_norm": 0.3663504123687744,
      "learning_rate": 6.194537812903599e-06,
      "loss": 0.0345,
      "step": 1189
    },
    {
      "epoch": 2.11,
      "grad_norm": 0.4621020257472992,
      "learning_rate": 6.171912280458214e-06,
      "loss": 0.0826,
      "step": 1190
    },
    {
      "epoch": 2.11,
      "grad_norm": 0.3509308993816376,
      "learning_rate": 6.1493174358618556e-06,
      "loss": 0.0613,
      "step": 1191
    },
    {
      "epoch": 2.11,
      "grad_norm": 0.35134875774383545,
      "learning_rate": 6.12675335765784e-06,
      "loss": 0.0402,
      "step": 1192
    },
    {
      "epoch": 2.11,
      "grad_norm": 0.3418435752391815,
      "learning_rate": 6.104220124282517e-06,
      "loss": 0.0783,
      "step": 1193
    },
    {
      "epoch": 2.11,
      "grad_norm": 0.23387284576892853,
      "learning_rate": 6.081717814065014e-06,
      "loss": 0.0369,
      "step": 1194
    },
    {
      "epoch": 2.11,
      "grad_norm": 0.41693243384361267,
      "learning_rate": 6.0592465052269845e-06,
      "loss": 0.0579,
      "step": 1195
    },
    {
      "epoch": 2.12,
      "grad_norm": 0.4878503978252411,
      "learning_rate": 6.0368062758822885e-06,
      "loss": 0.1005,
      "step": 1196
    },
    {
      "epoch": 2.12,
      "grad_norm": 0.4003542959690094,
      "learning_rate": 6.014397204036776e-06,
      "loss": 0.068,
      "step": 1197
    },
    {
      "epoch": 2.12,
      "grad_norm": 0.5766608119010925,
      "learning_rate": 5.992019367587968e-06,
      "loss": 0.0889,
      "step": 1198
    },
    {
      "epoch": 2.12,
      "grad_norm": 0.40670904517173767,
      "learning_rate": 5.96967284432481e-06,
      "loss": 0.079,
      "step": 1199
    },
    {
      "epoch": 2.12,
      "grad_norm": 0.3071047067642212,
      "learning_rate": 5.94735771192741e-06,
      "loss": 0.0572,
      "step": 1200
    },
    {
      "epoch": 2.12,
      "grad_norm": 0.4638030230998993,
      "learning_rate": 5.9250740479667405e-06,
      "loss": 0.0604,
      "step": 1201
    },
    {
      "epoch": 2.13,
      "grad_norm": 0.21257556974887848,
      "learning_rate": 5.902821929904401e-06,
      "loss": 0.0412,
      "step": 1202
    },
    {
      "epoch": 2.13,
      "grad_norm": 0.29504892230033875,
      "learning_rate": 5.880601435092316e-06,
      "loss": 0.0792,
      "step": 1203
    },
    {
      "epoch": 2.13,
      "grad_norm": 0.4328271150588989,
      "learning_rate": 5.858412640772488e-06,
      "loss": 0.1337,
      "step": 1204
    },
    {
      "epoch": 2.13,
      "grad_norm": 0.468470960855484,
      "learning_rate": 5.836255624076733e-06,
      "loss": 0.0499,
      "step": 1205
    },
    {
      "epoch": 2.13,
      "grad_norm": 0.335814505815506,
      "learning_rate": 5.814130462026388e-06,
      "loss": 0.049,
      "step": 1206
    },
    {
      "epoch": 2.14,
      "grad_norm": 0.7266606688499451,
      "learning_rate": 5.792037231532072e-06,
      "loss": 0.0794,
      "step": 1207
    },
    {
      "epoch": 2.14,
      "grad_norm": 0.319246381521225,
      "learning_rate": 5.769976009393396e-06,
      "loss": 0.0644,
      "step": 1208
    },
    {
      "epoch": 2.14,
      "grad_norm": 0.31407445669174194,
      "learning_rate": 5.7479468722987e-06,
      "loss": 0.0375,
      "step": 1209
    },
    {
      "epoch": 2.14,
      "grad_norm": 0.2565354108810425,
      "learning_rate": 5.725949896824805e-06,
      "loss": 0.0207,
      "step": 1210
    },
    {
      "epoch": 2.14,
      "grad_norm": 0.2970009744167328,
      "learning_rate": 5.703985159436732e-06,
      "loss": 0.0335,
      "step": 1211
    },
    {
      "epoch": 2.14,
      "grad_norm": 0.5138164758682251,
      "learning_rate": 5.682052736487422e-06,
      "loss": 0.0996,
      "step": 1212
    },
    {
      "epoch": 2.15,
      "grad_norm": 0.40743786096572876,
      "learning_rate": 5.6601527042175e-06,
      "loss": 0.0852,
      "step": 1213
    },
    {
      "epoch": 2.15,
      "grad_norm": 0.2411600947380066,
      "learning_rate": 5.638285138754984e-06,
      "loss": 0.0359,
      "step": 1214
    },
    {
      "epoch": 2.15,
      "grad_norm": 0.24094296991825104,
      "learning_rate": 5.616450116115045e-06,
      "loss": 0.0251,
      "step": 1215
    },
    {
      "epoch": 2.15,
      "grad_norm": 0.3407548666000366,
      "learning_rate": 5.59464771219973e-06,
      "loss": 0.0493,
      "step": 1216
    },
    {
      "epoch": 2.15,
      "grad_norm": 0.2684566080570221,
      "learning_rate": 5.572878002797686e-06,
      "loss": 0.0393,
      "step": 1217
    },
    {
      "epoch": 2.15,
      "grad_norm": 0.31322258710861206,
      "learning_rate": 5.551141063583915e-06,
      "loss": 0.0267,
      "step": 1218
    },
    {
      "epoch": 2.16,
      "grad_norm": 0.3750323951244354,
      "learning_rate": 5.529436970119501e-06,
      "loss": 0.0448,
      "step": 1219
    },
    {
      "epoch": 2.16,
      "grad_norm": 0.37540924549102783,
      "learning_rate": 5.5077657978513566e-06,
      "loss": 0.0282,
      "step": 1220
    },
    {
      "epoch": 2.16,
      "grad_norm": 0.6199687719345093,
      "learning_rate": 5.486127622111967e-06,
      "loss": 0.0756,
      "step": 1221
    },
    {
      "epoch": 2.16,
      "grad_norm": 0.5618672966957092,
      "learning_rate": 5.464522518119079e-06,
      "loss": 0.0949,
      "step": 1222
    },
    {
      "epoch": 2.16,
      "grad_norm": 0.37877845764160156,
      "learning_rate": 5.44295056097551e-06,
      "loss": 0.0537,
      "step": 1223
    },
    {
      "epoch": 2.17,
      "grad_norm": 0.19313400983810425,
      "learning_rate": 5.42141182566885e-06,
      "loss": 0.0188,
      "step": 1224
    },
    {
      "epoch": 2.17,
      "grad_norm": 0.5637698173522949,
      "learning_rate": 5.3999063870711865e-06,
      "loss": 0.0931,
      "step": 1225
    },
    {
      "epoch": 2.17,
      "grad_norm": 0.1735624372959137,
      "learning_rate": 5.378434319938889e-06,
      "loss": 0.0214,
      "step": 1226
    },
    {
      "epoch": 2.17,
      "grad_norm": 0.3521130084991455,
      "learning_rate": 5.356995698912287e-06,
      "loss": 0.0555,
      "step": 1227
    },
    {
      "epoch": 2.17,
      "grad_norm": 0.1801150143146515,
      "learning_rate": 5.3355905985154765e-06,
      "loss": 0.0184,
      "step": 1228
    },
    {
      "epoch": 2.17,
      "grad_norm": 0.3993035852909088,
      "learning_rate": 5.314219093156027e-06,
      "loss": 0.0573,
      "step": 1229
    },
    {
      "epoch": 2.18,
      "grad_norm": 0.5339346528053284,
      "learning_rate": 5.292881257124711e-06,
      "loss": 0.0666,
      "step": 1230
    },
    {
      "epoch": 2.18,
      "grad_norm": 0.4357282221317291,
      "learning_rate": 5.271577164595271e-06,
      "loss": 0.047,
      "step": 1231
    },
    {
      "epoch": 2.18,
      "grad_norm": 0.3784913122653961,
      "learning_rate": 5.250306889624147e-06,
      "loss": 0.0484,
      "step": 1232
    },
    {
      "epoch": 2.18,
      "grad_norm": 0.42714861035346985,
      "learning_rate": 5.229070506150227e-06,
      "loss": 0.0707,
      "step": 1233
    },
    {
      "epoch": 2.18,
      "grad_norm": 0.5577570796012878,
      "learning_rate": 5.207868087994596e-06,
      "loss": 0.0899,
      "step": 1234
    },
    {
      "epoch": 2.18,
      "grad_norm": 0.24935442209243774,
      "learning_rate": 5.186699708860253e-06,
      "loss": 0.0501,
      "step": 1235
    },
    {
      "epoch": 2.19,
      "grad_norm": 0.2705991566181183,
      "learning_rate": 5.165565442331876e-06,
      "loss": 0.0258,
      "step": 1236
    },
    {
      "epoch": 2.19,
      "grad_norm": 0.4349115192890167,
      "learning_rate": 5.144465361875574e-06,
      "loss": 0.0619,
      "step": 1237
    },
    {
      "epoch": 2.19,
      "grad_norm": 0.3423144817352295,
      "learning_rate": 5.123399540838603e-06,
      "loss": 0.0549,
      "step": 1238
    },
    {
      "epoch": 2.19,
      "grad_norm": 0.3189082741737366,
      "learning_rate": 5.102368052449145e-06,
      "loss": 0.0641,
      "step": 1239
    },
    {
      "epoch": 2.19,
      "grad_norm": 0.3039840757846832,
      "learning_rate": 5.0813709698160224e-06,
      "loss": 0.0304,
      "step": 1240
    },
    {
      "epoch": 2.2,
      "grad_norm": 0.41860231757164,
      "learning_rate": 5.060408365928459e-06,
      "loss": 0.1006,
      "step": 1241
    },
    {
      "epoch": 2.2,
      "grad_norm": 0.44600537419319153,
      "learning_rate": 5.039480313655836e-06,
      "loss": 0.0516,
      "step": 1242
    },
    {
      "epoch": 2.2,
      "grad_norm": 0.4928046762943268,
      "learning_rate": 5.018586885747414e-06,
      "loss": 0.0769,
      "step": 1243
    },
    {
      "epoch": 2.2,
      "grad_norm": 0.4866577088832855,
      "learning_rate": 4.997728154832106e-06,
      "loss": 0.0627,
      "step": 1244
    },
    {
      "epoch": 2.2,
      "grad_norm": 0.42723092436790466,
      "learning_rate": 4.976904193418204e-06,
      "loss": 0.067,
      "step": 1245
    },
    {
      "epoch": 2.2,
      "grad_norm": 0.3545394837856293,
      "learning_rate": 4.95611507389313e-06,
      "loss": 0.042,
      "step": 1246
    },
    {
      "epoch": 2.21,
      "grad_norm": 0.41975167393684387,
      "learning_rate": 4.935360868523213e-06,
      "loss": 0.0683,
      "step": 1247
    },
    {
      "epoch": 2.21,
      "grad_norm": 0.37999439239501953,
      "learning_rate": 4.914641649453387e-06,
      "loss": 0.071,
      "step": 1248
    },
    {
      "epoch": 2.21,
      "grad_norm": 0.4677155911922455,
      "learning_rate": 4.893957488706993e-06,
      "loss": 0.0572,
      "step": 1249
    },
    {
      "epoch": 2.21,
      "grad_norm": 0.34663069248199463,
      "learning_rate": 4.873308458185485e-06,
      "loss": 0.0621,
      "step": 1250
    },
    {
      "epoch": 2.21,
      "grad_norm": 0.30609679222106934,
      "learning_rate": 4.852694629668201e-06,
      "loss": 0.0315,
      "step": 1251
    },
    {
      "epoch": 2.21,
      "grad_norm": 0.3409930169582367,
      "learning_rate": 4.8321160748121245e-06,
      "loss": 0.0371,
      "step": 1252
    },
    {
      "epoch": 2.22,
      "grad_norm": 0.45600539445877075,
      "learning_rate": 4.811572865151605e-06,
      "loss": 0.1029,
      "step": 1253
    },
    {
      "epoch": 2.22,
      "grad_norm": 0.30145716667175293,
      "learning_rate": 4.791065072098145e-06,
      "loss": 0.0366,
      "step": 1254
    },
    {
      "epoch": 2.22,
      "grad_norm": 0.3910576403141022,
      "learning_rate": 4.770592766940116e-06,
      "loss": 0.0402,
      "step": 1255
    },
    {
      "epoch": 2.22,
      "grad_norm": 0.5734702348709106,
      "learning_rate": 4.75015602084253e-06,
      "loss": 0.0813,
      "step": 1256
    },
    {
      "epoch": 2.22,
      "grad_norm": 0.2789899706840515,
      "learning_rate": 4.729754904846806e-06,
      "loss": 0.0333,
      "step": 1257
    },
    {
      "epoch": 2.23,
      "grad_norm": 0.12205646932125092,
      "learning_rate": 4.7093894898704835e-06,
      "loss": 0.0118,
      "step": 1258
    },
    {
      "epoch": 2.23,
      "grad_norm": 0.43429574370384216,
      "learning_rate": 4.689059846707025e-06,
      "loss": 0.0729,
      "step": 1259
    },
    {
      "epoch": 2.23,
      "grad_norm": 0.34453603625297546,
      "learning_rate": 4.668766046025522e-06,
      "loss": 0.0519,
      "step": 1260
    },
    {
      "epoch": 2.23,
      "grad_norm": 0.6605038046836853,
      "learning_rate": 4.64850815837048e-06,
      "loss": 0.1086,
      "step": 1261
    },
    {
      "epoch": 2.23,
      "grad_norm": 0.43998420238494873,
      "learning_rate": 4.628286254161571e-06,
      "loss": 0.0842,
      "step": 1262
    },
    {
      "epoch": 2.23,
      "grad_norm": 0.47020792961120605,
      "learning_rate": 4.60810040369338e-06,
      "loss": 0.0579,
      "step": 1263
    },
    {
      "epoch": 2.24,
      "grad_norm": 0.628449022769928,
      "learning_rate": 4.587950677135157e-06,
      "loss": 0.0477,
      "step": 1264
    },
    {
      "epoch": 2.24,
      "grad_norm": 0.33337822556495667,
      "learning_rate": 4.567837144530585e-06,
      "loss": 0.0411,
      "step": 1265
    },
    {
      "epoch": 2.24,
      "grad_norm": 0.39417657256126404,
      "learning_rate": 4.547759875797524e-06,
      "loss": 0.0641,
      "step": 1266
    },
    {
      "epoch": 2.24,
      "grad_norm": 0.48267191648483276,
      "learning_rate": 4.5277189407277855e-06,
      "loss": 0.0872,
      "step": 1267
    },
    {
      "epoch": 2.24,
      "grad_norm": 0.19380518794059753,
      "learning_rate": 4.50771440898688e-06,
      "loss": 0.0154,
      "step": 1268
    },
    {
      "epoch": 2.25,
      "grad_norm": 0.5586434602737427,
      "learning_rate": 4.487746350113764e-06,
      "loss": 0.0446,
      "step": 1269
    },
    {
      "epoch": 2.25,
      "grad_norm": 0.5079402923583984,
      "learning_rate": 4.467814833520613e-06,
      "loss": 0.0964,
      "step": 1270
    },
    {
      "epoch": 2.25,
      "grad_norm": 0.228449746966362,
      "learning_rate": 4.447919928492572e-06,
      "loss": 0.0269,
      "step": 1271
    },
    {
      "epoch": 2.25,
      "grad_norm": 0.5465266704559326,
      "learning_rate": 4.428061704187525e-06,
      "loss": 0.1152,
      "step": 1272
    },
    {
      "epoch": 2.25,
      "grad_norm": 0.3933871388435364,
      "learning_rate": 4.408240229635855e-06,
      "loss": 0.0659,
      "step": 1273
    },
    {
      "epoch": 2.25,
      "grad_norm": 0.259215772151947,
      "learning_rate": 4.3884555737401705e-06,
      "loss": 0.0799,
      "step": 1274
    },
    {
      "epoch": 2.26,
      "grad_norm": 0.23984096944332123,
      "learning_rate": 4.368707805275116e-06,
      "loss": 0.0244,
      "step": 1275
    },
    {
      "epoch": 2.26,
      "grad_norm": 0.43554961681365967,
      "learning_rate": 4.348996992887107e-06,
      "loss": 0.0666,
      "step": 1276
    },
    {
      "epoch": 2.26,
      "grad_norm": 0.2799820601940155,
      "learning_rate": 4.329323205094083e-06,
      "loss": 0.0427,
      "step": 1277
    },
    {
      "epoch": 2.26,
      "grad_norm": 0.461411714553833,
      "learning_rate": 4.309686510285296e-06,
      "loss": 0.0402,
      "step": 1278
    },
    {
      "epoch": 2.26,
      "eval_loss": 0.07151434570550919,
      "eval_runtime": 14.6624,
      "eval_samples_per_second": 32.532,
      "eval_steps_per_second": 8.184,
      "step": 1278
    },
    {
      "epoch": 2.26,
      "grad_norm": 0.5163243412971497,
      "learning_rate": 4.2900869767210326e-06,
      "loss": 0.0544,
      "step": 1279
    },
    {
      "epoch": 2.26,
      "grad_norm": 0.2997105121612549,
      "learning_rate": 4.270524672532422e-06,
      "loss": 0.0385,
      "step": 1280
    },
    {
      "epoch": 2.27,
      "grad_norm": 0.4136078953742981,
      "learning_rate": 4.250999665721177e-06,
      "loss": 0.0744,
      "step": 1281
    },
    {
      "epoch": 2.27,
      "grad_norm": 0.3969447612762451,
      "learning_rate": 4.2315120241593486e-06,
      "loss": 0.0519,
      "step": 1282
    },
    {
      "epoch": 2.27,
      "grad_norm": 0.2855490744113922,
      "learning_rate": 4.212061815589103e-06,
      "loss": 0.0374,
      "step": 1283
    },
    {
      "epoch": 2.27,
      "grad_norm": 0.5917441248893738,
      "learning_rate": 4.192649107622485e-06,
      "loss": 0.0535,
      "step": 1284
    },
    {
      "epoch": 2.27,
      "grad_norm": 0.5109731554985046,
      "learning_rate": 4.173273967741184e-06,
      "loss": 0.082,
      "step": 1285
    },
    {
      "epoch": 2.28,
      "grad_norm": 0.37259694933891296,
      "learning_rate": 4.1539364632962955e-06,
      "loss": 0.0308,
      "step": 1286
    },
    {
      "epoch": 2.28,
      "grad_norm": 0.6265465021133423,
      "learning_rate": 4.134636661508088e-06,
      "loss": 0.068,
      "step": 1287
    },
    {
      "epoch": 2.28,
      "grad_norm": 0.2064184993505478,
      "learning_rate": 4.115374629465759e-06,
      "loss": 0.0225,
      "step": 1288
    },
    {
      "epoch": 2.28,
      "grad_norm": 0.5311080813407898,
      "learning_rate": 4.096150434127234e-06,
      "loss": 0.0914,
      "step": 1289
    },
    {
      "epoch": 2.28,
      "grad_norm": 0.23628176748752594,
      "learning_rate": 4.07696414231889e-06,
      "loss": 0.0254,
      "step": 1290
    },
    {
      "epoch": 2.28,
      "grad_norm": 0.40518784523010254,
      "learning_rate": 4.0578158207353615e-06,
      "loss": 0.0697,
      "step": 1291
    },
    {
      "epoch": 2.29,
      "grad_norm": 0.28831130266189575,
      "learning_rate": 4.038705535939285e-06,
      "loss": 0.0462,
      "step": 1292
    },
    {
      "epoch": 2.29,
      "grad_norm": 0.5238087773323059,
      "learning_rate": 4.019633354361071e-06,
      "loss": 0.0633,
      "step": 1293
    },
    {
      "epoch": 2.29,
      "grad_norm": 0.14449584484100342,
      "learning_rate": 4.000599342298688e-06,
      "loss": 0.014,
      "step": 1294
    },
    {
      "epoch": 2.29,
      "grad_norm": 0.5843258500099182,
      "learning_rate": 3.98160356591741e-06,
      "loss": 0.0779,
      "step": 1295
    },
    {
      "epoch": 2.29,
      "grad_norm": 0.3418098986148834,
      "learning_rate": 3.962646091249609e-06,
      "loss": 0.0282,
      "step": 1296
    },
    {
      "epoch": 2.29,
      "grad_norm": 0.2479211390018463,
      "learning_rate": 3.943726984194504e-06,
      "loss": 0.0427,
      "step": 1297
    },
    {
      "epoch": 2.3,
      "grad_norm": 0.5063907504081726,
      "learning_rate": 3.9248463105179434e-06,
      "loss": 0.0697,
      "step": 1298
    },
    {
      "epoch": 2.3,
      "grad_norm": 0.19487616419792175,
      "learning_rate": 3.9060041358521836e-06,
      "loss": 0.0222,
      "step": 1299
    },
    {
      "epoch": 2.3,
      "grad_norm": 0.34873899817466736,
      "learning_rate": 3.8872005256956384e-06,
      "loss": 0.0561,
      "step": 1300
    },
    {
      "epoch": 2.3,
      "grad_norm": 0.40151849389076233,
      "learning_rate": 3.868435545412683e-06,
      "loss": 0.0499,
      "step": 1301
    },
    {
      "epoch": 2.3,
      "grad_norm": 0.27045050263404846,
      "learning_rate": 3.849709260233394e-06,
      "loss": 0.0319,
      "step": 1302
    },
    {
      "epoch": 2.31,
      "grad_norm": 0.25688886642456055,
      "learning_rate": 3.831021735253338e-06,
      "loss": 0.0582,
      "step": 1303
    },
    {
      "epoch": 2.31,
      "grad_norm": 0.24792470037937164,
      "learning_rate": 3.812373035433358e-06,
      "loss": 0.0327,
      "step": 1304
    },
    {
      "epoch": 2.31,
      "grad_norm": 0.8629806041717529,
      "learning_rate": 3.7937632255993176e-06,
      "loss": 0.0637,
      "step": 1305
    },
    {
      "epoch": 2.31,
      "grad_norm": 0.1900147944688797,
      "learning_rate": 3.77519237044191e-06,
      "loss": 0.0148,
      "step": 1306
    },
    {
      "epoch": 2.31,
      "grad_norm": 0.24926882982254028,
      "learning_rate": 3.7566605345164014e-06,
      "loss": 0.0251,
      "step": 1307
    },
    {
      "epoch": 2.31,
      "grad_norm": 0.6645401120185852,
      "learning_rate": 3.738167782242423e-06,
      "loss": 0.0609,
      "step": 1308
    },
    {
      "epoch": 2.32,
      "grad_norm": 0.3478599488735199,
      "learning_rate": 3.7197141779037543e-06,
      "loss": 0.0492,
      "step": 1309
    },
    {
      "epoch": 2.32,
      "grad_norm": 0.40398237109184265,
      "learning_rate": 3.7012997856480797e-06,
      "loss": 0.063,
      "step": 1310
    },
    {
      "epoch": 2.32,
      "grad_norm": 0.3616901934146881,
      "learning_rate": 3.682924669486785e-06,
      "loss": 0.0487,
      "step": 1311
    },
    {
      "epoch": 2.32,
      "grad_norm": 0.5381997227668762,
      "learning_rate": 3.664588893294721e-06,
      "loss": 0.05,
      "step": 1312
    },
    {
      "epoch": 2.32,
      "grad_norm": 0.11271072179079056,
      "learning_rate": 3.6462925208099796e-06,
      "loss": 0.0085,
      "step": 1313
    },
    {
      "epoch": 2.32,
      "grad_norm": 0.350453644990921,
      "learning_rate": 3.6280356156336924e-06,
      "loss": 0.0374,
      "step": 1314
    },
    {
      "epoch": 2.33,
      "grad_norm": 0.16282452642917633,
      "learning_rate": 3.6098182412297943e-06,
      "loss": 0.0133,
      "step": 1315
    },
    {
      "epoch": 2.33,
      "grad_norm": 0.3517274558544159,
      "learning_rate": 3.591640460924797e-06,
      "loss": 0.0203,
      "step": 1316
    },
    {
      "epoch": 2.33,
      "grad_norm": 0.25263047218322754,
      "learning_rate": 3.573502337907579e-06,
      "loss": 0.0315,
      "step": 1317
    },
    {
      "epoch": 2.33,
      "grad_norm": 0.19249844551086426,
      "learning_rate": 3.5554039352291633e-06,
      "loss": 0.0164,
      "step": 1318
    },
    {
      "epoch": 2.33,
      "grad_norm": 0.2523667514324188,
      "learning_rate": 3.537345315802506e-06,
      "loss": 0.0224,
      "step": 1319
    },
    {
      "epoch": 2.34,
      "grad_norm": 0.42540183663368225,
      "learning_rate": 3.51932654240227e-06,
      "loss": 0.0455,
      "step": 1320
    },
    {
      "epoch": 2.34,
      "grad_norm": 0.22705115377902985,
      "learning_rate": 3.5013476776645874e-06,
      "loss": 0.01,
      "step": 1321
    },
    {
      "epoch": 2.34,
      "grad_norm": 0.4173358082771301,
      "learning_rate": 3.4834087840868883e-06,
      "loss": 0.0542,
      "step": 1322
    },
    {
      "epoch": 2.34,
      "grad_norm": 0.7138379812240601,
      "learning_rate": 3.4655099240276363e-06,
      "loss": 0.0758,
      "step": 1323
    },
    {
      "epoch": 2.34,
      "grad_norm": 0.6766006946563721,
      "learning_rate": 3.447651159706144e-06,
      "loss": 0.1455,
      "step": 1324
    },
    {
      "epoch": 2.34,
      "grad_norm": 0.6347004175186157,
      "learning_rate": 3.4298325532023494e-06,
      "loss": 0.1176,
      "step": 1325
    },
    {
      "epoch": 2.35,
      "grad_norm": 0.34710976481437683,
      "learning_rate": 3.4120541664565693e-06,
      "loss": 0.0391,
      "step": 1326
    },
    {
      "epoch": 2.35,
      "grad_norm": 0.3936313986778259,
      "learning_rate": 3.3943160612693374e-06,
      "loss": 0.0543,
      "step": 1327
    },
    {
      "epoch": 2.35,
      "grad_norm": 0.8079743385314941,
      "learning_rate": 3.376618299301156e-06,
      "loss": 0.1023,
      "step": 1328
    },
    {
      "epoch": 2.35,
      "grad_norm": 0.41258299350738525,
      "learning_rate": 3.3589609420722828e-06,
      "loss": 0.0453,
      "step": 1329
    },
    {
      "epoch": 2.35,
      "grad_norm": 0.4742684066295624,
      "learning_rate": 3.34134405096252e-06,
      "loss": 0.0527,
      "step": 1330
    },
    {
      "epoch": 2.35,
      "grad_norm": 0.32282182574272156,
      "learning_rate": 3.323767687211006e-06,
      "loss": 0.0345,
      "step": 1331
    },
    {
      "epoch": 2.36,
      "grad_norm": 0.3263574540615082,
      "learning_rate": 3.306231911916005e-06,
      "loss": 0.0338,
      "step": 1332
    },
    {
      "epoch": 2.36,
      "grad_norm": 0.3974379897117615,
      "learning_rate": 3.2887367860346858e-06,
      "loss": 0.0409,
      "step": 1333
    },
    {
      "epoch": 2.36,
      "grad_norm": 0.5523183345794678,
      "learning_rate": 3.271282370382912e-06,
      "loss": 0.0836,
      "step": 1334
    },
    {
      "epoch": 2.36,
      "grad_norm": 0.2913055717945099,
      "learning_rate": 3.2538687256350297e-06,
      "loss": 0.0721,
      "step": 1335
    },
    {
      "epoch": 2.36,
      "grad_norm": 0.25458353757858276,
      "learning_rate": 3.2364959123236575e-06,
      "loss": 0.0321,
      "step": 1336
    },
    {
      "epoch": 2.37,
      "grad_norm": 0.24113737046718597,
      "learning_rate": 3.2191639908394864e-06,
      "loss": 0.0261,
      "step": 1337
    },
    {
      "epoch": 2.37,
      "grad_norm": 0.40691906213760376,
      "learning_rate": 3.2018730214310594e-06,
      "loss": 0.0842,
      "step": 1338
    },
    {
      "epoch": 2.37,
      "grad_norm": 0.3553066849708557,
      "learning_rate": 3.184623064204556e-06,
      "loss": 0.0319,
      "step": 1339
    },
    {
      "epoch": 2.37,
      "grad_norm": 0.6767488121986389,
      "learning_rate": 3.1674141791235887e-06,
      "loss": 0.0835,
      "step": 1340
    },
    {
      "epoch": 2.37,
      "grad_norm": 0.3634691536426544,
      "learning_rate": 3.1502464260090145e-06,
      "loss": 0.0368,
      "step": 1341
    },
    {
      "epoch": 2.37,
      "grad_norm": 0.26596900820732117,
      "learning_rate": 3.1331198645386843e-06,
      "loss": 0.034,
      "step": 1342
    },
    {
      "epoch": 2.38,
      "grad_norm": 0.6405837535858154,
      "learning_rate": 3.116034554247285e-06,
      "loss": 0.1115,
      "step": 1343
    },
    {
      "epoch": 2.38,
      "grad_norm": 0.19923199713230133,
      "learning_rate": 3.0989905545260906e-06,
      "loss": 0.0158,
      "step": 1344
    },
    {
      "epoch": 2.38,
      "grad_norm": 0.48847952485084534,
      "learning_rate": 3.081987924622774e-06,
      "loss": 0.0709,
      "step": 1345
    },
    {
      "epoch": 2.38,
      "grad_norm": 0.39113959670066833,
      "learning_rate": 3.0650267236412107e-06,
      "loss": 0.0879,
      "step": 1346
    },
    {
      "epoch": 2.38,
      "grad_norm": 0.3358931839466095,
      "learning_rate": 3.0481070105412506e-06,
      "loss": 0.0521,
      "step": 1347
    },
    {
      "epoch": 2.38,
      "grad_norm": 0.4323839843273163,
      "learning_rate": 3.031228844138537e-06,
      "loss": 0.0576,
      "step": 1348
    },
    {
      "epoch": 2.39,
      "grad_norm": 0.515374481678009,
      "learning_rate": 3.01439228310428e-06,
      "loss": 0.0901,
      "step": 1349
    },
    {
      "epoch": 2.39,
      "grad_norm": 0.2964722216129303,
      "learning_rate": 2.997597385965062e-06,
      "loss": 0.0576,
      "step": 1350
    },
    {
      "epoch": 2.39,
      "grad_norm": 0.30420923233032227,
      "learning_rate": 2.980844211102649e-06,
      "loss": 0.0372,
      "step": 1351
    },
    {
      "epoch": 2.39,
      "grad_norm": 0.44702932238578796,
      "learning_rate": 2.9641328167537553e-06,
      "loss": 0.0646,
      "step": 1352
    },
    {
      "epoch": 2.39,
      "grad_norm": 0.6082139611244202,
      "learning_rate": 2.947463261009875e-06,
      "loss": 0.08,
      "step": 1353
    },
    {
      "epoch": 2.4,
      "grad_norm": 0.4607807397842407,
      "learning_rate": 2.9308356018170567e-06,
      "loss": 0.0584,
      "step": 1354
    },
    {
      "epoch": 2.4,
      "grad_norm": 0.2861080765724182,
      "learning_rate": 2.9142498969757046e-06,
      "loss": 0.0364,
      "step": 1355
    },
    {
      "epoch": 2.4,
      "grad_norm": 0.37672412395477295,
      "learning_rate": 2.8977062041404002e-06,
      "loss": 0.0419,
      "step": 1356
    },
    {
      "epoch": 2.4,
      "grad_norm": 0.5021412968635559,
      "learning_rate": 2.8812045808196625e-06,
      "loss": 0.0798,
      "step": 1357
    },
    {
      "epoch": 2.4,
      "grad_norm": 0.3646734058856964,
      "learning_rate": 2.86474508437579e-06,
      "loss": 0.0636,
      "step": 1358
    },
    {
      "epoch": 2.4,
      "grad_norm": 0.40267646312713623,
      "learning_rate": 2.848327772024628e-06,
      "loss": 0.1034,
      "step": 1359
    },
    {
      "epoch": 2.41,
      "grad_norm": 0.28698602318763733,
      "learning_rate": 2.8319527008353863e-06,
      "loss": 0.0372,
      "step": 1360
    },
    {
      "epoch": 2.41,
      "grad_norm": 0.5204796195030212,
      "learning_rate": 2.8156199277304445e-06,
      "loss": 0.0824,
      "step": 1361
    },
    {
      "epoch": 2.41,
      "grad_norm": 0.23750893771648407,
      "learning_rate": 2.799329509485136e-06,
      "loss": 0.0324,
      "step": 1362
    },
    {
      "epoch": 2.41,
      "grad_norm": 0.49522823095321655,
      "learning_rate": 2.7830815027275735e-06,
      "loss": 0.0926,
      "step": 1363
    },
    {
      "epoch": 2.41,
      "grad_norm": 0.3374737799167633,
      "learning_rate": 2.7668759639384315e-06,
      "loss": 0.0614,
      "step": 1364
    },
    {
      "epoch": 2.41,
      "grad_norm": 0.3192250728607178,
      "learning_rate": 2.75071294945076e-06,
      "loss": 0.0368,
      "step": 1365
    },
    {
      "epoch": 2.42,
      "grad_norm": 0.37366268038749695,
      "learning_rate": 2.7345925154497915e-06,
      "loss": 0.027,
      "step": 1366
    },
    {
      "epoch": 2.42,
      "grad_norm": 0.29517611861228943,
      "learning_rate": 2.7185147179727477e-06,
      "loss": 0.0476,
      "step": 1367
    },
    {
      "epoch": 2.42,
      "grad_norm": 0.46492999792099,
      "learning_rate": 2.702479612908624e-06,
      "loss": 0.0441,
      "step": 1368
    },
    {
      "epoch": 2.42,
      "grad_norm": 0.17148606479167938,
      "learning_rate": 2.6864872559980204e-06,
      "loss": 0.0163,
      "step": 1369
    },
    {
      "epoch": 2.42,
      "grad_norm": 0.3044341206550598,
      "learning_rate": 2.6705377028329314e-06,
      "loss": 0.0362,
      "step": 1370
    },
    {
      "epoch": 2.43,
      "grad_norm": 0.6719170212745667,
      "learning_rate": 2.6546310088565623e-06,
      "loss": 0.0799,
      "step": 1371
    },
    {
      "epoch": 2.43,
      "grad_norm": 0.3043392300605774,
      "learning_rate": 2.6387672293631455e-06,
      "loss": 0.027,
      "step": 1372
    },
    {
      "epoch": 2.43,
      "grad_norm": 0.47442537546157837,
      "learning_rate": 2.622946419497706e-06,
      "loss": 0.0548,
      "step": 1373
    },
    {
      "epoch": 2.43,
      "grad_norm": 0.4492048919200897,
      "learning_rate": 2.607168634255927e-06,
      "loss": 0.0504,
      "step": 1374
    },
    {
      "epoch": 2.43,
      "grad_norm": 0.18210428953170776,
      "learning_rate": 2.5914339284839143e-06,
      "loss": 0.0145,
      "step": 1375
    },
    {
      "epoch": 2.43,
      "grad_norm": 0.3385724127292633,
      "learning_rate": 2.5757423568780323e-06,
      "loss": 0.0615,
      "step": 1376
    },
    {
      "epoch": 2.44,
      "grad_norm": 0.3785318434238434,
      "learning_rate": 2.56009397398471e-06,
      "loss": 0.0971,
      "step": 1377
    },
    {
      "epoch": 2.44,
      "grad_norm": 0.23501533269882202,
      "learning_rate": 2.54448883420022e-06,
      "loss": 0.0368,
      "step": 1378
    },
    {
      "epoch": 2.44,
      "grad_norm": 0.442818284034729,
      "learning_rate": 2.528926991770541e-06,
      "loss": 0.067,
      "step": 1379
    },
    {
      "epoch": 2.44,
      "grad_norm": 0.4778684973716736,
      "learning_rate": 2.513408500791135e-06,
      "loss": 0.0692,
      "step": 1380
    },
    {
      "epoch": 2.44,
      "grad_norm": 0.5602303743362427,
      "learning_rate": 2.4979334152067625e-06,
      "loss": 0.092,
      "step": 1381
    },
    {
      "epoch": 2.44,
      "grad_norm": 0.24895040690898895,
      "learning_rate": 2.482501788811304e-06,
      "loss": 0.0218,
      "step": 1382
    },
    {
      "epoch": 2.45,
      "grad_norm": 0.5014092326164246,
      "learning_rate": 2.4671136752475647e-06,
      "loss": 0.082,
      "step": 1383
    },
    {
      "epoch": 2.45,
      "grad_norm": 0.40069711208343506,
      "learning_rate": 2.4517691280070996e-06,
      "loss": 0.0337,
      "step": 1384
    },
    {
      "epoch": 2.45,
      "grad_norm": 0.4973551630973816,
      "learning_rate": 2.4364682004300193e-06,
      "loss": 0.0436,
      "step": 1385
    },
    {
      "epoch": 2.45,
      "grad_norm": 0.312372624874115,
      "learning_rate": 2.4212109457048004e-06,
      "loss": 0.05,
      "step": 1386
    },
    {
      "epoch": 2.45,
      "grad_norm": 0.3463585078716278,
      "learning_rate": 2.405997416868105e-06,
      "loss": 0.0889,
      "step": 1387
    },
    {
      "epoch": 2.46,
      "grad_norm": 0.5272729396820068,
      "learning_rate": 2.3908276668045977e-06,
      "loss": 0.0556,
      "step": 1388
    },
    {
      "epoch": 2.46,
      "grad_norm": 0.38329455256462097,
      "learning_rate": 2.3757017482467686e-06,
      "loss": 0.0618,
      "step": 1389
    },
    {
      "epoch": 2.46,
      "grad_norm": 0.4668821692466736,
      "learning_rate": 2.3606197137747367e-06,
      "loss": 0.0627,
      "step": 1390
    },
    {
      "epoch": 2.46,
      "grad_norm": 0.4220178723335266,
      "learning_rate": 2.345581615816072e-06,
      "loss": 0.072,
      "step": 1391
    },
    {
      "epoch": 2.46,
      "grad_norm": 0.3023943603038788,
      "learning_rate": 2.330587506645611e-06,
      "loss": 0.0359,
      "step": 1392
    },
    {
      "epoch": 2.46,
      "grad_norm": 0.45773857831954956,
      "learning_rate": 2.3156374383852907e-06,
      "loss": 0.0806,
      "step": 1393
    },
    {
      "epoch": 2.47,
      "grad_norm": 0.2793668210506439,
      "learning_rate": 2.3007314630039384e-06,
      "loss": 0.0365,
      "step": 1394
    },
    {
      "epoch": 2.47,
      "grad_norm": 0.23663386702537537,
      "learning_rate": 2.2858696323171223e-06,
      "loss": 0.0284,
      "step": 1395
    },
    {
      "epoch": 2.47,
      "grad_norm": 0.5298436284065247,
      "learning_rate": 2.271051997986946e-06,
      "loss": 0.0793,
      "step": 1396
    },
    {
      "epoch": 2.47,
      "grad_norm": 0.48489508032798767,
      "learning_rate": 2.2562786115218815e-06,
      "loss": 0.0616,
      "step": 1397
    },
    {
      "epoch": 2.47,
      "grad_norm": 0.24025483429431915,
      "learning_rate": 2.2415495242765958e-06,
      "loss": 0.0358,
      "step": 1398
    },
    {
      "epoch": 2.48,
      "grad_norm": 0.34402361512184143,
      "learning_rate": 2.2268647874517532e-06,
      "loss": 0.0588,
      "step": 1399
    },
    {
      "epoch": 2.48,
      "grad_norm": 0.1925269067287445,
      "learning_rate": 2.212224452093859e-06,
      "loss": 0.0276,
      "step": 1400
    },
    {
      "epoch": 2.48,
      "grad_norm": 0.44379922747612,
      "learning_rate": 2.1976285690950664e-06,
      "loss": 0.0577,
      "step": 1401
    },
    {
      "epoch": 2.48,
      "grad_norm": 0.16387049853801727,
      "learning_rate": 2.1830771891930003e-06,
      "loss": 0.014,
      "step": 1402
    },
    {
      "epoch": 2.48,
      "grad_norm": 0.5020678043365479,
      "learning_rate": 2.1685703629705995e-06,
      "loss": 0.0624,
      "step": 1403
    },
    {
      "epoch": 2.48,
      "grad_norm": 0.4342516362667084,
      "learning_rate": 2.1541081408559142e-06,
      "loss": 0.0458,
      "step": 1404
    },
    {
      "epoch": 2.49,
      "grad_norm": 0.6095245480537415,
      "learning_rate": 2.1396905731219508e-06,
      "loss": 0.0953,
      "step": 1405
    },
    {
      "epoch": 2.49,
      "grad_norm": 0.36794888973236084,
      "learning_rate": 2.1253177098864886e-06,
      "loss": 0.0518,
      "step": 1406
    },
    {
      "epoch": 2.49,
      "grad_norm": 0.39768266677856445,
      "learning_rate": 2.1109896011118986e-06,
      "loss": 0.0299,
      "step": 1407
    },
    {
      "epoch": 2.49,
      "grad_norm": 0.3278927505016327,
      "learning_rate": 2.0967062966049943e-06,
      "loss": 0.0352,
      "step": 1408
    },
    {
      "epoch": 2.49,
      "grad_norm": 0.5671990513801575,
      "learning_rate": 2.0824678460168246e-06,
      "loss": 0.0441,
      "step": 1409
    },
    {
      "epoch": 2.49,
      "grad_norm": 0.3596339225769043,
      "learning_rate": 2.068274298842537e-06,
      "loss": 0.0259,
      "step": 1410
    },
    {
      "epoch": 2.5,
      "grad_norm": 1.04413640499115,
      "learning_rate": 2.054125704421174e-06,
      "loss": 0.0884,
      "step": 1411
    },
    {
      "epoch": 2.5,
      "grad_norm": 0.21271169185638428,
      "learning_rate": 2.040022111935516e-06,
      "loss": 0.0238,
      "step": 1412
    },
    {
      "epoch": 2.5,
      "grad_norm": 0.26020702719688416,
      "learning_rate": 2.025963570411919e-06,
      "loss": 0.0278,
      "step": 1413
    },
    {
      "epoch": 2.5,
      "grad_norm": 0.37250760197639465,
      "learning_rate": 2.0119501287201223e-06,
      "loss": 0.0436,
      "step": 1414
    },
    {
      "epoch": 2.5,
      "grad_norm": 0.423744797706604,
      "learning_rate": 1.997981835573102e-06,
      "loss": 0.0395,
      "step": 1415
    },
    {
      "epoch": 2.51,
      "grad_norm": 0.3542567193508148,
      "learning_rate": 1.9840587395268855e-06,
      "loss": 0.0466,
      "step": 1416
    },
    {
      "epoch": 2.51,
      "grad_norm": 0.393533855676651,
      "learning_rate": 1.9701808889803842e-06,
      "loss": 0.058,
      "step": 1417
    },
    {
      "epoch": 2.51,
      "grad_norm": 0.5265038013458252,
      "learning_rate": 1.9563483321752335e-06,
      "loss": 0.0645,
      "step": 1418
    },
    {
      "epoch": 2.51,
      "grad_norm": 0.4641624689102173,
      "learning_rate": 1.9425611171956254e-06,
      "loss": 0.0695,
      "step": 1419
    },
    {
      "epoch": 2.51,
      "grad_norm": 0.3584718108177185,
      "learning_rate": 1.9288192919681276e-06,
      "loss": 0.0623,
      "step": 1420
    },
    {
      "epoch": 2.51,
      "eval_loss": 0.07115022838115692,
      "eval_runtime": 14.6922,
      "eval_samples_per_second": 32.466,
      "eval_steps_per_second": 8.168,
      "step": 1420
    },
    {
      "epoch": 2.51,
      "grad_norm": 0.3901975452899933,
      "learning_rate": 1.915122904261527e-06,
      "loss": 0.0337,
      "step": 1421
    },
    {
      "epoch": 2.52,
      "grad_norm": 0.22733630239963531,
      "learning_rate": 1.9014720016866626e-06,
      "loss": 0.0188,
      "step": 1422
    },
    {
      "epoch": 2.52,
      "grad_norm": 0.21050037443637848,
      "learning_rate": 1.8878666316962644e-06,
      "loss": 0.0178,
      "step": 1423
    },
    {
      "epoch": 2.52,
      "grad_norm": 0.24884532392024994,
      "learning_rate": 1.8743068415847875e-06,
      "loss": 0.0343,
      "step": 1424
    },
    {
      "epoch": 2.52,
      "grad_norm": 0.7502188682556152,
      "learning_rate": 1.8607926784882234e-06,
      "loss": 0.1535,
      "step": 1425
    },
    {
      "epoch": 2.52,
      "grad_norm": 0.41273728013038635,
      "learning_rate": 1.8473241893839825e-06,
      "loss": 0.0502,
      "step": 1426
    },
    {
      "epoch": 2.52,
      "grad_norm": 0.12475810199975967,
      "learning_rate": 1.8339014210906858e-06,
      "loss": 0.0081,
      "step": 1427
    },
    {
      "epoch": 2.53,
      "grad_norm": 0.36283671855926514,
      "learning_rate": 1.8205244202680322e-06,
      "loss": 0.0216,
      "step": 1428
    },
    {
      "epoch": 2.53,
      "grad_norm": 0.29988062381744385,
      "learning_rate": 1.8071932334166286e-06,
      "loss": 0.0191,
      "step": 1429
    },
    {
      "epoch": 2.53,
      "grad_norm": 0.5531153678894043,
      "learning_rate": 1.7939079068778075e-06,
      "loss": 0.079,
      "step": 1430
    },
    {
      "epoch": 2.53,
      "grad_norm": 0.545619547367096,
      "learning_rate": 1.7806684868335005e-06,
      "loss": 0.0638,
      "step": 1431
    },
    {
      "epoch": 2.53,
      "grad_norm": 0.44630181789398193,
      "learning_rate": 1.7674750193060586e-06,
      "loss": 0.0638,
      "step": 1432
    },
    {
      "epoch": 2.54,
      "grad_norm": 0.434993177652359,
      "learning_rate": 1.754327550158087e-06,
      "loss": 0.0604,
      "step": 1433
    },
    {
      "epoch": 2.54,
      "grad_norm": 0.36365070939064026,
      "learning_rate": 1.7412261250923007e-06,
      "loss": 0.0398,
      "step": 1434
    },
    {
      "epoch": 2.54,
      "grad_norm": 0.483367383480072,
      "learning_rate": 1.7281707896513476e-06,
      "loss": 0.079,
      "step": 1435
    },
    {
      "epoch": 2.54,
      "grad_norm": 0.5866251587867737,
      "learning_rate": 1.7151615892176741e-06,
      "loss": 0.0764,
      "step": 1436
    },
    {
      "epoch": 2.54,
      "grad_norm": 0.6167349815368652,
      "learning_rate": 1.7021985690133496e-06,
      "loss": 0.0811,
      "step": 1437
    },
    {
      "epoch": 2.54,
      "grad_norm": 0.333730012178421,
      "learning_rate": 1.6892817740999084e-06,
      "loss": 0.0318,
      "step": 1438
    },
    {
      "epoch": 2.55,
      "grad_norm": 0.5943579077720642,
      "learning_rate": 1.676411249378202e-06,
      "loss": 0.1098,
      "step": 1439
    },
    {
      "epoch": 2.55,
      "grad_norm": 0.18848778307437897,
      "learning_rate": 1.6635870395882368e-06,
      "loss": 0.0135,
      "step": 1440
    },
    {
      "epoch": 2.55,
      "grad_norm": 0.5768080949783325,
      "learning_rate": 1.6508091893090238e-06,
      "loss": 0.0549,
      "step": 1441
    },
    {
      "epoch": 2.55,
      "grad_norm": 0.6475699543952942,
      "learning_rate": 1.6380777429584254e-06,
      "loss": 0.0899,
      "step": 1442
    },
    {
      "epoch": 2.55,
      "grad_norm": 0.21376800537109375,
      "learning_rate": 1.6253927447929883e-06,
      "loss": 0.0204,
      "step": 1443
    },
    {
      "epoch": 2.55,
      "grad_norm": 0.7223924398422241,
      "learning_rate": 1.612754238907797e-06,
      "loss": 0.0971,
      "step": 1444
    },
    {
      "epoch": 2.56,
      "grad_norm": 0.3582633137702942,
      "learning_rate": 1.6001622692363315e-06,
      "loss": 0.0518,
      "step": 1445
    },
    {
      "epoch": 2.56,
      "grad_norm": 0.6689372658729553,
      "learning_rate": 1.5876168795502931e-06,
      "loss": 0.0976,
      "step": 1446
    },
    {
      "epoch": 2.56,
      "grad_norm": 0.43327635526657104,
      "learning_rate": 1.5751181134594744e-06,
      "loss": 0.0586,
      "step": 1447
    },
    {
      "epoch": 2.56,
      "grad_norm": 0.4076501727104187,
      "learning_rate": 1.5626660144115873e-06,
      "loss": 0.0576,
      "step": 1448
    },
    {
      "epoch": 2.56,
      "grad_norm": 0.5978496670722961,
      "learning_rate": 1.550260625692122e-06,
      "loss": 0.0519,
      "step": 1449
    },
    {
      "epoch": 2.57,
      "grad_norm": 0.48387810587882996,
      "learning_rate": 1.5379019904242087e-06,
      "loss": 0.0809,
      "step": 1450
    },
    {
      "epoch": 2.57,
      "grad_norm": 0.3782576620578766,
      "learning_rate": 1.5255901515684373e-06,
      "loss": 0.0481,
      "step": 1451
    },
    {
      "epoch": 2.57,
      "grad_norm": 0.5015052556991577,
      "learning_rate": 1.5133251519227414e-06,
      "loss": 0.0526,
      "step": 1452
    },
    {
      "epoch": 2.57,
      "grad_norm": 0.5278434753417969,
      "learning_rate": 1.501107034122225e-06,
      "loss": 0.0697,
      "step": 1453
    },
    {
      "epoch": 2.57,
      "grad_norm": 0.3558174669742584,
      "learning_rate": 1.4889358406390218e-06,
      "loss": 0.0406,
      "step": 1454
    },
    {
      "epoch": 2.57,
      "grad_norm": 0.23327094316482544,
      "learning_rate": 1.4768116137821586e-06,
      "loss": 0.0158,
      "step": 1455
    },
    {
      "epoch": 2.58,
      "grad_norm": 0.3898582458496094,
      "learning_rate": 1.4647343956973869e-06,
      "loss": 0.0463,
      "step": 1456
    },
    {
      "epoch": 2.58,
      "grad_norm": 0.7152931690216064,
      "learning_rate": 1.4527042283670594e-06,
      "loss": 0.0575,
      "step": 1457
    },
    {
      "epoch": 2.58,
      "grad_norm": 0.3319093585014343,
      "learning_rate": 1.4407211536099662e-06,
      "loss": 0.0459,
      "step": 1458
    },
    {
      "epoch": 2.58,
      "grad_norm": 0.6415733098983765,
      "learning_rate": 1.4287852130811934e-06,
      "loss": 0.087,
      "step": 1459
    },
    {
      "epoch": 2.58,
      "grad_norm": 0.4185143709182739,
      "learning_rate": 1.4168964482719916e-06,
      "loss": 0.041,
      "step": 1460
    },
    {
      "epoch": 2.58,
      "grad_norm": 0.5387318730354309,
      "learning_rate": 1.4050549005096053e-06,
      "loss": 0.0415,
      "step": 1461
    },
    {
      "epoch": 2.59,
      "grad_norm": 0.437717080116272,
      "learning_rate": 1.3932606109571643e-06,
      "loss": 0.0866,
      "step": 1462
    },
    {
      "epoch": 2.59,
      "grad_norm": 0.6287564635276794,
      "learning_rate": 1.3815136206135065e-06,
      "loss": 0.0766,
      "step": 1463
    },
    {
      "epoch": 2.59,
      "grad_norm": 0.2929694652557373,
      "learning_rate": 1.3698139703130507e-06,
      "loss": 0.0253,
      "step": 1464
    },
    {
      "epoch": 2.59,
      "grad_norm": 0.40167632699012756,
      "learning_rate": 1.3581617007256647e-06,
      "loss": 0.0417,
      "step": 1465
    },
    {
      "epoch": 2.59,
      "grad_norm": 0.24478429555892944,
      "learning_rate": 1.3465568523565008e-06,
      "loss": 0.0266,
      "step": 1466
    },
    {
      "epoch": 2.6,
      "grad_norm": 0.7342913746833801,
      "learning_rate": 1.3349994655458786e-06,
      "loss": 0.1278,
      "step": 1467
    },
    {
      "epoch": 2.6,
      "grad_norm": 0.2442154437303543,
      "learning_rate": 1.3234895804691277e-06,
      "loss": 0.0307,
      "step": 1468
    },
    {
      "epoch": 2.6,
      "grad_norm": 0.27558204531669617,
      "learning_rate": 1.31202723713645e-06,
      "loss": 0.053,
      "step": 1469
    },
    {
      "epoch": 2.6,
      "grad_norm": 0.32540640234947205,
      "learning_rate": 1.3006124753927945e-06,
      "loss": 0.0315,
      "step": 1470
    },
    {
      "epoch": 2.6,
      "grad_norm": 0.315765917301178,
      "learning_rate": 1.2892453349177075e-06,
      "loss": 0.0288,
      "step": 1471
    },
    {
      "epoch": 2.6,
      "grad_norm": 0.1552322506904602,
      "learning_rate": 1.2779258552251826e-06,
      "loss": 0.0107,
      "step": 1472
    },
    {
      "epoch": 2.61,
      "grad_norm": 0.2357468605041504,
      "learning_rate": 1.2666540756635546e-06,
      "loss": 0.0224,
      "step": 1473
    },
    {
      "epoch": 2.61,
      "grad_norm": 0.24665170907974243,
      "learning_rate": 1.2554300354153325e-06,
      "loss": 0.0206,
      "step": 1474
    },
    {
      "epoch": 2.61,
      "grad_norm": 0.4270358979701996,
      "learning_rate": 1.2442537734970843e-06,
      "loss": 0.0543,
      "step": 1475
    },
    {
      "epoch": 2.61,
      "grad_norm": 0.39957016706466675,
      "learning_rate": 1.2331253287592931e-06,
      "loss": 0.0561,
      "step": 1476
    },
    {
      "epoch": 2.61,
      "grad_norm": 0.3666171431541443,
      "learning_rate": 1.2220447398862079e-06,
      "loss": 0.0477,
      "step": 1477
    },
    {
      "epoch": 2.61,
      "grad_norm": 0.17843380570411682,
      "learning_rate": 1.2110120453957407e-06,
      "loss": 0.0139,
      "step": 1478
    },
    {
      "epoch": 2.62,
      "grad_norm": 0.49627992510795593,
      "learning_rate": 1.2000272836393023e-06,
      "loss": 0.0604,
      "step": 1479
    },
    {
      "epoch": 2.62,
      "grad_norm": 0.49571043252944946,
      "learning_rate": 1.1890904928016928e-06,
      "loss": 0.0785,
      "step": 1480
    },
    {
      "epoch": 2.62,
      "grad_norm": 0.19603154063224792,
      "learning_rate": 1.1782017109009458e-06,
      "loss": 0.0117,
      "step": 1481
    },
    {
      "epoch": 2.62,
      "grad_norm": 0.5238936543464661,
      "learning_rate": 1.1673609757882152e-06,
      "loss": 0.0985,
      "step": 1482
    },
    {
      "epoch": 2.62,
      "grad_norm": 0.5575663447380066,
      "learning_rate": 1.1565683251476356e-06,
      "loss": 0.0714,
      "step": 1483
    },
    {
      "epoch": 2.63,
      "grad_norm": 0.2580691874027252,
      "learning_rate": 1.1458237964961943e-06,
      "loss": 0.0237,
      "step": 1484
    },
    {
      "epoch": 2.63,
      "grad_norm": 0.40892645716667175,
      "learning_rate": 1.1351274271835948e-06,
      "loss": 0.0648,
      "step": 1485
    },
    {
      "epoch": 2.63,
      "grad_norm": 0.2748018801212311,
      "learning_rate": 1.1244792543921328e-06,
      "loss": 0.0234,
      "step": 1486
    },
    {
      "epoch": 2.63,
      "grad_norm": 0.2547319829463959,
      "learning_rate": 1.113879315136561e-06,
      "loss": 0.0225,
      "step": 1487
    },
    {
      "epoch": 2.63,
      "grad_norm": 0.2713530659675598,
      "learning_rate": 1.1033276462639736e-06,
      "loss": 0.0278,
      "step": 1488
    },
    {
      "epoch": 2.63,
      "grad_norm": 0.5096026659011841,
      "learning_rate": 1.0928242844536646e-06,
      "loss": 0.0497,
      "step": 1489
    },
    {
      "epoch": 2.64,
      "grad_norm": 0.36748412251472473,
      "learning_rate": 1.0823692662170015e-06,
      "loss": 0.042,
      "step": 1490
    },
    {
      "epoch": 2.64,
      "grad_norm": 0.4302610158920288,
      "learning_rate": 1.071962627897306e-06,
      "loss": 0.0797,
      "step": 1491
    },
    {
      "epoch": 2.64,
      "grad_norm": 0.6664643287658691,
      "learning_rate": 1.0616044056697211e-06,
      "loss": 0.0656,
      "step": 1492
    },
    {
      "epoch": 2.64,
      "grad_norm": 0.35455599427223206,
      "learning_rate": 1.051294635541089e-06,
      "loss": 0.0425,
      "step": 1493
    },
    {
      "epoch": 2.64,
      "grad_norm": 0.19645701348781586,
      "learning_rate": 1.0410333533498274e-06,
      "loss": 0.0151,
      "step": 1494
    },
    {
      "epoch": 2.64,
      "grad_norm": 0.34883156418800354,
      "learning_rate": 1.0308205947657977e-06,
      "loss": 0.0387,
      "step": 1495
    },
    {
      "epoch": 2.65,
      "grad_norm": 0.28385984897613525,
      "learning_rate": 1.0206563952901843e-06,
      "loss": 0.0231,
      "step": 1496
    },
    {
      "epoch": 2.65,
      "grad_norm": 0.26889827847480774,
      "learning_rate": 1.0105407902553788e-06,
      "loss": 0.0178,
      "step": 1497
    },
    {
      "epoch": 2.65,
      "grad_norm": 0.3332523703575134,
      "learning_rate": 1.0004738148248433e-06,
      "loss": 0.0484,
      "step": 1498
    },
    {
      "epoch": 2.65,
      "grad_norm": 0.3340343236923218,
      "learning_rate": 9.904555039930019e-07,
      "loss": 0.0259,
      "step": 1499
    },
    {
      "epoch": 2.65,
      "grad_norm": 0.5199639201164246,
      "learning_rate": 9.804858925851124e-07,
      "loss": 0.1132,
      "step": 1500
    },
    {
      "epoch": 2.66,
      "grad_norm": 0.45809826254844666,
      "learning_rate": 9.70565015257136e-07,
      "loss": 0.0751,
      "step": 1501
    },
    {
      "epoch": 2.66,
      "grad_norm": 0.6508705019950867,
      "learning_rate": 9.6069290649564e-07,
      "loss": 0.0563,
      "step": 1502
    },
    {
      "epoch": 2.66,
      "grad_norm": 0.622417688369751,
      "learning_rate": 9.508696006176549e-07,
      "loss": 0.1071,
      "step": 1503
    },
    {
      "epoch": 2.66,
      "grad_norm": 0.26498842239379883,
      "learning_rate": 9.410951317705707e-07,
      "loss": 0.0242,
      "step": 1504
    },
    {
      "epoch": 2.66,
      "grad_norm": 0.5023782253265381,
      "learning_rate": 9.313695339320067e-07,
      "loss": 0.0521,
      "step": 1505
    },
    {
      "epoch": 2.66,
      "grad_norm": 0.511811375617981,
      "learning_rate": 9.216928409097009e-07,
      "loss": 0.0631,
      "step": 1506
    },
    {
      "epoch": 2.67,
      "grad_norm": 0.4455617070198059,
      "learning_rate": 9.120650863413926e-07,
      "loss": 0.0532,
      "step": 1507
    },
    {
      "epoch": 2.67,
      "grad_norm": 0.3804994225502014,
      "learning_rate": 9.024863036946973e-07,
      "loss": 0.0395,
      "step": 1508
    },
    {
      "epoch": 2.67,
      "grad_norm": 0.46066808700561523,
      "learning_rate": 8.929565262670031e-07,
      "loss": 0.05,
      "step": 1509
    },
    {
      "epoch": 2.67,
      "grad_norm": 0.5113961696624756,
      "learning_rate": 8.834757871853461e-07,
      "loss": 0.0384,
      "step": 1510
    },
    {
      "epoch": 2.67,
      "grad_norm": 0.5323219299316406,
      "learning_rate": 8.74044119406292e-07,
      "loss": 0.074,
      "step": 1511
    },
    {
      "epoch": 2.67,
      "grad_norm": 0.283601313829422,
      "learning_rate": 8.646615557158349e-07,
      "loss": 0.0301,
      "step": 1512
    },
    {
      "epoch": 2.68,
      "grad_norm": 0.6767367124557495,
      "learning_rate": 8.553281287292685e-07,
      "loss": 0.0762,
      "step": 1513
    },
    {
      "epoch": 2.68,
      "grad_norm": 0.3783128261566162,
      "learning_rate": 8.460438708910834e-07,
      "loss": 0.0471,
      "step": 1514
    },
    {
      "epoch": 2.68,
      "grad_norm": 0.42817193269729614,
      "learning_rate": 8.368088144748515e-07,
      "loss": 0.0422,
      "step": 1515
    },
    {
      "epoch": 2.68,
      "grad_norm": 0.8753883242607117,
      "learning_rate": 8.276229915831035e-07,
      "loss": 0.0891,
      "step": 1516
    },
    {
      "epoch": 2.68,
      "grad_norm": 0.5999267101287842,
      "learning_rate": 8.18486434147242e-07,
      "loss": 0.1057,
      "step": 1517
    },
    {
      "epoch": 2.69,
      "grad_norm": 0.4635411500930786,
      "learning_rate": 8.093991739273982e-07,
      "loss": 0.0677,
      "step": 1518
    },
    {
      "epoch": 2.69,
      "grad_norm": 0.30402085185050964,
      "learning_rate": 8.003612425123519e-07,
      "loss": 0.0168,
      "step": 1519
    },
    {
      "epoch": 2.69,
      "grad_norm": 0.4269234240055084,
      "learning_rate": 7.91372671319402e-07,
      "loss": 0.0857,
      "step": 1520
    },
    {
      "epoch": 2.69,
      "grad_norm": 0.28628554940223694,
      "learning_rate": 7.824334915942593e-07,
      "loss": 0.0359,
      "step": 1521
    },
    {
      "epoch": 2.69,
      "grad_norm": 0.44654789566993713,
      "learning_rate": 7.735437344109492e-07,
      "loss": 0.0534,
      "step": 1522
    },
    {
      "epoch": 2.69,
      "grad_norm": 0.2673017084598541,
      "learning_rate": 7.647034306716988e-07,
      "loss": 0.0261,
      "step": 1523
    },
    {
      "epoch": 2.7,
      "grad_norm": 0.6055803298950195,
      "learning_rate": 7.559126111068132e-07,
      "loss": 0.0597,
      "step": 1524
    },
    {
      "epoch": 2.7,
      "grad_norm": 0.5424272418022156,
      "learning_rate": 7.471713062745967e-07,
      "loss": 0.0613,
      "step": 1525
    },
    {
      "epoch": 2.7,
      "grad_norm": 0.33164864778518677,
      "learning_rate": 7.384795465612243e-07,
      "loss": 0.0534,
      "step": 1526
    },
    {
      "epoch": 2.7,
      "grad_norm": 0.5044467449188232,
      "learning_rate": 7.298373621806492e-07,
      "loss": 0.0304,
      "step": 1527
    },
    {
      "epoch": 2.7,
      "grad_norm": 0.4113554060459137,
      "learning_rate": 7.21244783174494e-07,
      "loss": 0.0747,
      "step": 1528
    },
    {
      "epoch": 2.7,
      "grad_norm": 0.39796996116638184,
      "learning_rate": 7.127018394119322e-07,
      "loss": 0.0505,
      "step": 1529
    },
    {
      "epoch": 2.71,
      "grad_norm": 0.6214816570281982,
      "learning_rate": 7.042085605896142e-07,
      "loss": 0.0923,
      "step": 1530
    },
    {
      "epoch": 2.71,
      "grad_norm": 0.4567388892173767,
      "learning_rate": 6.957649762315365e-07,
      "loss": 0.0318,
      "step": 1531
    },
    {
      "epoch": 2.71,
      "grad_norm": 0.4828348159790039,
      "learning_rate": 6.87371115688954e-07,
      "loss": 0.0436,
      "step": 1532
    },
    {
      "epoch": 2.71,
      "grad_norm": 0.5060093402862549,
      "learning_rate": 6.790270081402744e-07,
      "loss": 0.0474,
      "step": 1533
    },
    {
      "epoch": 2.71,
      "grad_norm": 0.4873170852661133,
      "learning_rate": 6.707326825909476e-07,
      "loss": 0.0421,
      "step": 1534
    },
    {
      "epoch": 2.72,
      "grad_norm": 0.37341755628585815,
      "learning_rate": 6.624881678733852e-07,
      "loss": 0.066,
      "step": 1535
    },
    {
      "epoch": 2.72,
      "grad_norm": 0.5217847228050232,
      "learning_rate": 6.542934926468436e-07,
      "loss": 0.0687,
      "step": 1536
    },
    {
      "epoch": 2.72,
      "grad_norm": 0.4019028842449188,
      "learning_rate": 6.461486853973264e-07,
      "loss": 0.0349,
      "step": 1537
    },
    {
      "epoch": 2.72,
      "grad_norm": 0.4829115569591522,
      "learning_rate": 6.380537744374892e-07,
      "loss": 0.1073,
      "step": 1538
    },
    {
      "epoch": 2.72,
      "grad_norm": 0.30531370639801025,
      "learning_rate": 6.300087879065397e-07,
      "loss": 0.0293,
      "step": 1539
    },
    {
      "epoch": 2.72,
      "grad_norm": 0.17552495002746582,
      "learning_rate": 6.220137537701459e-07,
      "loss": 0.016,
      "step": 1540
    },
    {
      "epoch": 2.73,
      "grad_norm": 0.43100669980049133,
      "learning_rate": 6.140686998203266e-07,
      "loss": 0.067,
      "step": 1541
    },
    {
      "epoch": 2.73,
      "grad_norm": 0.2270410656929016,
      "learning_rate": 6.061736536753643e-07,
      "loss": 0.0219,
      "step": 1542
    },
    {
      "epoch": 2.73,
      "grad_norm": 0.3404647409915924,
      "learning_rate": 5.983286427797058e-07,
      "loss": 0.0311,
      "step": 1543
    },
    {
      "epoch": 2.73,
      "grad_norm": 0.7673945426940918,
      "learning_rate": 5.905336944038652e-07,
      "loss": 0.1104,
      "step": 1544
    },
    {
      "epoch": 2.73,
      "grad_norm": 0.6210458874702454,
      "learning_rate": 5.827888356443362e-07,
      "loss": 0.0985,
      "step": 1545
    },
    {
      "epoch": 2.74,
      "grad_norm": 0.4982345402240753,
      "learning_rate": 5.750940934234894e-07,
      "loss": 0.0513,
      "step": 1546
    },
    {
      "epoch": 2.74,
      "grad_norm": 0.207786425948143,
      "learning_rate": 5.674494944894853e-07,
      "loss": 0.0123,
      "step": 1547
    },
    {
      "epoch": 2.74,
      "grad_norm": 0.3776586949825287,
      "learning_rate": 5.598550654161705e-07,
      "loss": 0.0451,
      "step": 1548
    },
    {
      "epoch": 2.74,
      "grad_norm": 0.49511584639549255,
      "learning_rate": 5.52310832603004e-07,
      "loss": 0.0984,
      "step": 1549
    },
    {
      "epoch": 2.74,
      "grad_norm": 0.4227202832698822,
      "learning_rate": 5.448168222749467e-07,
      "loss": 0.0515,
      "step": 1550
    },
    {
      "epoch": 2.74,
      "grad_norm": 0.7386292219161987,
      "learning_rate": 5.373730604823818e-07,
      "loss": 0.1768,
      "step": 1551
    },
    {
      "epoch": 2.75,
      "grad_norm": 0.2904682159423828,
      "learning_rate": 5.299795731010193e-07,
      "loss": 0.0368,
      "step": 1552
    },
    {
      "epoch": 2.75,
      "grad_norm": 0.5355067253112793,
      "learning_rate": 5.22636385831804e-07,
      "loss": 0.0915,
      "step": 1553
    },
    {
      "epoch": 2.75,
      "grad_norm": 0.47790849208831787,
      "learning_rate": 5.153435242008386e-07,
      "loss": 0.0453,
      "step": 1554
    },
    {
      "epoch": 2.75,
      "grad_norm": 0.46563971042633057,
      "learning_rate": 5.081010135592745e-07,
      "loss": 0.0772,
      "step": 1555
    },
    {
      "epoch": 2.75,
      "grad_norm": 0.5036541819572449,
      "learning_rate": 5.009088790832445e-07,
      "loss": 0.0936,
      "step": 1556
    },
    {
      "epoch": 2.75,
      "grad_norm": 0.30798229575157166,
      "learning_rate": 4.937671457737597e-07,
      "loss": 0.0332,
      "step": 1557
    },
    {
      "epoch": 2.76,
      "grad_norm": 0.6296735405921936,
      "learning_rate": 4.8667583845663e-07,
      "loss": 0.0987,
      "step": 1558
    },
    {
      "epoch": 2.76,
      "grad_norm": 0.28153422474861145,
      "learning_rate": 4.796349817823803e-07,
      "loss": 0.0279,
      "step": 1559
    },
    {
      "epoch": 2.76,
      "grad_norm": 0.2385409027338028,
      "learning_rate": 4.726446002261542e-07,
      "loss": 0.0329,
      "step": 1560
    },
    {
      "epoch": 2.76,
      "grad_norm": 0.6607958674430847,
      "learning_rate": 4.657047180876406e-07,
      "loss": 0.0986,
      "step": 1561
    },
    {
      "epoch": 2.76,
      "grad_norm": 0.37040072679519653,
      "learning_rate": 4.5881535949098553e-07,
      "loss": 0.0285,
      "step": 1562
    },
    {
      "epoch": 2.76,
      "eval_loss": 0.07119648903608322,
      "eval_runtime": 14.7013,
      "eval_samples_per_second": 32.446,
      "eval_steps_per_second": 8.163,
      "step": 1562
    },
    {
      "epoch": 2.77,
      "grad_norm": 0.3848133087158203,
      "learning_rate": 4.519765483847005e-07,
      "loss": 0.0631,
      "step": 1563
    },
    {
      "epoch": 2.77,
      "grad_norm": 0.4391275644302368,
      "learning_rate": 4.4518830854159253e-07,
      "loss": 0.0338,
      "step": 1564
    },
    {
      "epoch": 2.77,
      "grad_norm": 0.31983691453933716,
      "learning_rate": 4.3845066355867104e-07,
      "loss": 0.0444,
      "step": 1565
    },
    {
      "epoch": 2.77,
      "grad_norm": 0.42251527309417725,
      "learning_rate": 4.317636368570727e-07,
      "loss": 0.0376,
      "step": 1566
    },
    {
      "epoch": 2.77,
      "grad_norm": 0.513642430305481,
      "learning_rate": 4.251272516819732e-07,
      "loss": 0.0597,
      "step": 1567
    },
    {
      "epoch": 2.77,
      "grad_norm": 0.5827353596687317,
      "learning_rate": 4.1854153110251093e-07,
      "loss": 0.0846,
      "step": 1568
    },
    {
      "epoch": 2.78,
      "grad_norm": 0.4715573191642761,
      "learning_rate": 4.1200649801170996e-07,
      "loss": 0.0719,
      "step": 1569
    },
    {
      "epoch": 2.78,
      "grad_norm": 0.30255019664764404,
      "learning_rate": 4.0552217512639213e-07,
      "loss": 0.028,
      "step": 1570
    },
    {
      "epoch": 2.78,
      "grad_norm": 0.3282603621482849,
      "learning_rate": 3.9908858498710176e-07,
      "loss": 0.0328,
      "step": 1571
    },
    {
      "epoch": 2.78,
      "grad_norm": 0.5265703201293945,
      "learning_rate": 3.9270574995803446e-07,
      "loss": 0.0409,
      "step": 1572
    },
    {
      "epoch": 2.78,
      "grad_norm": 0.42628806829452515,
      "learning_rate": 3.863736922269451e-07,
      "loss": 0.1025,
      "step": 1573
    },
    {
      "epoch": 2.78,
      "grad_norm": 0.36057740449905396,
      "learning_rate": 3.8009243380508485e-07,
      "loss": 0.0776,
      "step": 1574
    },
    {
      "epoch": 2.79,
      "grad_norm": 0.9079022407531738,
      "learning_rate": 3.738619965271145e-07,
      "loss": 0.0848,
      "step": 1575
    },
    {
      "epoch": 2.79,
      "grad_norm": 0.5063414573669434,
      "learning_rate": 3.676824020510328e-07,
      "loss": 0.1095,
      "step": 1576
    },
    {
      "epoch": 2.79,
      "grad_norm": 0.6455457806587219,
      "learning_rate": 3.6155367185810137e-07,
      "loss": 0.1155,
      "step": 1577
    },
    {
      "epoch": 2.79,
      "grad_norm": 0.521200954914093,
      "learning_rate": 3.5547582725277026e-07,
      "loss": 0.0395,
      "step": 1578
    },
    {
      "epoch": 2.79,
      "grad_norm": 0.18381936848163605,
      "learning_rate": 3.4944888936260246e-07,
      "loss": 0.0117,
      "step": 1579
    },
    {
      "epoch": 2.8,
      "grad_norm": 0.4150390028953552,
      "learning_rate": 3.4347287913819915e-07,
      "loss": 0.0672,
      "step": 1580
    },
    {
      "epoch": 2.8,
      "grad_norm": 0.6446459889411926,
      "learning_rate": 3.3754781735313155e-07,
      "loss": 0.0991,
      "step": 1581
    },
    {
      "epoch": 2.8,
      "grad_norm": 0.5226888656616211,
      "learning_rate": 3.316737246038676e-07,
      "loss": 0.0868,
      "step": 1582
    },
    {
      "epoch": 2.8,
      "grad_norm": 0.16347049176692963,
      "learning_rate": 3.258506213096901e-07,
      "loss": 0.0154,
      "step": 1583
    },
    {
      "epoch": 2.8,
      "grad_norm": 0.574674129486084,
      "learning_rate": 3.2007852771264555e-07,
      "loss": 0.078,
      "step": 1584
    },
    {
      "epoch": 2.8,
      "grad_norm": 0.23604702949523926,
      "learning_rate": 3.1435746387745543e-07,
      "loss": 0.0223,
      "step": 1585
    },
    {
      "epoch": 2.81,
      "grad_norm": 0.5670037269592285,
      "learning_rate": 3.086874496914532e-07,
      "loss": 0.0698,
      "step": 1586
    },
    {
      "epoch": 2.81,
      "grad_norm": 0.4633881151676178,
      "learning_rate": 3.030685048645193e-07,
      "loss": 0.043,
      "step": 1587
    },
    {
      "epoch": 2.81,
      "grad_norm": 0.41053467988967896,
      "learning_rate": 2.975006489290094e-07,
      "loss": 0.0565,
      "step": 1588
    },
    {
      "epoch": 2.81,
      "grad_norm": 0.5542423725128174,
      "learning_rate": 2.9198390123968133e-07,
      "loss": 0.1034,
      "step": 1589
    },
    {
      "epoch": 2.81,
      "grad_norm": 0.48999693989753723,
      "learning_rate": 2.8651828097363664e-07,
      "loss": 0.0522,
      "step": 1590
    },
    {
      "epoch": 2.81,
      "grad_norm": 0.36279919743537903,
      "learning_rate": 2.8110380713024406e-07,
      "loss": 0.0692,
      "step": 1591
    },
    {
      "epoch": 2.82,
      "grad_norm": 0.3871041238307953,
      "learning_rate": 2.75740498531088e-07,
      "loss": 0.0409,
      "step": 1592
    },
    {
      "epoch": 2.82,
      "grad_norm": 0.5611832141876221,
      "learning_rate": 2.704283738198882e-07,
      "loss": 0.0566,
      "step": 1593
    },
    {
      "epoch": 2.82,
      "grad_norm": 0.37240684032440186,
      "learning_rate": 2.6516745146244193e-07,
      "loss": 0.0725,
      "step": 1594
    },
    {
      "epoch": 2.82,
      "grad_norm": 0.42195749282836914,
      "learning_rate": 2.599577497465605e-07,
      "loss": 0.0558,
      "step": 1595
    },
    {
      "epoch": 2.82,
      "grad_norm": 0.25024503469467163,
      "learning_rate": 2.547992867819993e-07,
      "loss": 0.0226,
      "step": 1596
    },
    {
      "epoch": 2.83,
      "grad_norm": 0.7625443339347839,
      "learning_rate": 2.496920805004044e-07,
      "loss": 0.0931,
      "step": 1597
    },
    {
      "epoch": 2.83,
      "grad_norm": 0.23752780258655548,
      "learning_rate": 2.446361486552462e-07,
      "loss": 0.0176,
      "step": 1598
    },
    {
      "epoch": 2.83,
      "grad_norm": 0.2915506064891815,
      "learning_rate": 2.396315088217543e-07,
      "loss": 0.0259,
      "step": 1599
    },
    {
      "epoch": 2.83,
      "grad_norm": 0.6136963963508606,
      "learning_rate": 2.346781783968577e-07,
      "loss": 0.0956,
      "step": 1600
    },
    {
      "epoch": 2.83,
      "grad_norm": 0.26477885246276855,
      "learning_rate": 2.2977617459912782e-07,
      "loss": 0.0381,
      "step": 1601
    },
    {
      "epoch": 2.83,
      "grad_norm": 0.33673399686813354,
      "learning_rate": 2.2492551446871735e-07,
      "loss": 0.0433,
      "step": 1602
    },
    {
      "epoch": 2.84,
      "grad_norm": 0.3527979850769043,
      "learning_rate": 2.2012621486730001e-07,
      "loss": 0.0431,
      "step": 1603
    },
    {
      "epoch": 2.84,
      "grad_norm": 0.6614082455635071,
      "learning_rate": 2.15378292478009e-07,
      "loss": 0.0463,
      "step": 1604
    },
    {
      "epoch": 2.84,
      "grad_norm": 0.44098663330078125,
      "learning_rate": 2.1068176380538372e-07,
      "loss": 0.0511,
      "step": 1605
    },
    {
      "epoch": 2.84,
      "grad_norm": 0.6739132404327393,
      "learning_rate": 2.0603664517531318e-07,
      "loss": 0.08,
      "step": 1606
    },
    {
      "epoch": 2.84,
      "grad_norm": 0.6040840744972229,
      "learning_rate": 2.0144295273497427e-07,
      "loss": 0.0832,
      "step": 1607
    },
    {
      "epoch": 2.84,
      "grad_norm": 0.14114706218242645,
      "learning_rate": 1.9690070245277858e-07,
      "loss": 0.0123,
      "step": 1608
    },
    {
      "epoch": 2.85,
      "grad_norm": 0.29719462990760803,
      "learning_rate": 1.9240991011831576e-07,
      "loss": 0.0303,
      "step": 1609
    },
    {
      "epoch": 2.85,
      "grad_norm": 0.43595388531684875,
      "learning_rate": 1.8797059134230188e-07,
      "loss": 0.0691,
      "step": 1610
    },
    {
      "epoch": 2.85,
      "grad_norm": 0.3607807755470276,
      "learning_rate": 1.8358276155651943e-07,
      "loss": 0.0286,
      "step": 1611
    },
    {
      "epoch": 2.85,
      "grad_norm": 0.2499689906835556,
      "learning_rate": 1.7924643601377078e-07,
      "loss": 0.0256,
      "step": 1612
    },
    {
      "epoch": 2.85,
      "grad_norm": 0.5863063335418701,
      "learning_rate": 1.7496162978781815e-07,
      "loss": 0.1147,
      "step": 1613
    },
    {
      "epoch": 2.86,
      "grad_norm": 0.5148590803146362,
      "learning_rate": 1.707283577733354e-07,
      "loss": 0.0404,
      "step": 1614
    },
    {
      "epoch": 2.86,
      "grad_norm": 0.5219923257827759,
      "learning_rate": 1.6654663468585296e-07,
      "loss": 0.0802,
      "step": 1615
    },
    {
      "epoch": 2.86,
      "grad_norm": 0.599203884601593,
      "learning_rate": 1.6241647506171465e-07,
      "loss": 0.0765,
      "step": 1616
    },
    {
      "epoch": 2.86,
      "grad_norm": 0.3104330599308014,
      "learning_rate": 1.583378932580143e-07,
      "loss": 0.0294,
      "step": 1617
    },
    {
      "epoch": 2.86,
      "grad_norm": 0.2936621606349945,
      "learning_rate": 1.5431090345255917e-07,
      "loss": 0.0556,
      "step": 1618
    },
    {
      "epoch": 2.86,
      "grad_norm": 0.380932092666626,
      "learning_rate": 1.5033551964381164e-07,
      "loss": 0.0484,
      "step": 1619
    },
    {
      "epoch": 2.87,
      "grad_norm": 0.7550992965698242,
      "learning_rate": 1.4641175565084263e-07,
      "loss": 0.1731,
      "step": 1620
    },
    {
      "epoch": 2.87,
      "grad_norm": 0.3216942250728607,
      "learning_rate": 1.4253962511328655e-07,
      "loss": 0.0307,
      "step": 1621
    },
    {
      "epoch": 2.87,
      "grad_norm": 0.7092807292938232,
      "learning_rate": 1.3871914149128805e-07,
      "loss": 0.1201,
      "step": 1622
    },
    {
      "epoch": 2.87,
      "grad_norm": 0.6073005199432373,
      "learning_rate": 1.3495031806546044e-07,
      "loss": 0.106,
      "step": 1623
    },
    {
      "epoch": 2.87,
      "grad_norm": 0.6765084266662598,
      "learning_rate": 1.3123316793684225e-07,
      "loss": 0.1752,
      "step": 1624
    },
    {
      "epoch": 2.87,
      "grad_norm": 0.3464655578136444,
      "learning_rate": 1.275677040268408e-07,
      "loss": 0.0229,
      "step": 1625
    },
    {
      "epoch": 2.88,
      "grad_norm": 0.451846718788147,
      "learning_rate": 1.239539390771971e-07,
      "loss": 0.0971,
      "step": 1626
    },
    {
      "epoch": 2.88,
      "grad_norm": 0.643297553062439,
      "learning_rate": 1.2039188564994086e-07,
      "loss": 0.0649,
      "step": 1627
    },
    {
      "epoch": 2.88,
      "grad_norm": 0.4810031056404114,
      "learning_rate": 1.1688155612733565e-07,
      "loss": 0.0679,
      "step": 1628
    },
    {
      "epoch": 2.88,
      "grad_norm": 0.44242003560066223,
      "learning_rate": 1.1342296271185381e-07,
      "loss": 0.0699,
      "step": 1629
    },
    {
      "epoch": 2.88,
      "grad_norm": 0.4847901165485382,
      "learning_rate": 1.1001611742611828e-07,
      "loss": 0.0633,
      "step": 1630
    },
    {
      "epoch": 2.89,
      "grad_norm": 0.4562082588672638,
      "learning_rate": 1.0666103211287248e-07,
      "loss": 0.0493,
      "step": 1631
    },
    {
      "epoch": 2.89,
      "grad_norm": 0.37047621607780457,
      "learning_rate": 1.0335771843493047e-07,
      "loss": 0.0545,
      "step": 1632
    },
    {
      "epoch": 2.89,
      "grad_norm": 0.2991613447666168,
      "learning_rate": 1.0010618787513859e-07,
      "loss": 0.0411,
      "step": 1633
    },
    {
      "epoch": 2.89,
      "grad_norm": 0.23022329807281494,
      "learning_rate": 9.69064517363405e-08,
      "loss": 0.0188,
      "step": 1634
    },
    {
      "epoch": 2.89,
      "grad_norm": 0.20215001702308655,
      "learning_rate": 9.375852114133221e-08,
      "loss": 0.0174,
      "step": 1635
    },
    {
      "epoch": 2.89,
      "grad_norm": 0.3517981767654419,
      "learning_rate": 9.066240703282547e-08,
      "loss": 0.0587,
      "step": 1636
    },
    {
      "epoch": 2.9,
      "grad_norm": 0.48674067854881287,
      "learning_rate": 8.761812017340776e-08,
      "loss": 0.1072,
      "step": 1637
    },
    {
      "epoch": 2.9,
      "grad_norm": 0.4294196367263794,
      "learning_rate": 8.462567114551067e-08,
      "loss": 0.0455,
      "step": 1638
    },
    {
      "epoch": 2.9,
      "grad_norm": 0.5136977434158325,
      "learning_rate": 8.168507035136663e-08,
      "loss": 0.0698,
      "step": 1639
    },
    {
      "epoch": 2.9,
      "grad_norm": 0.23223884403705597,
      "learning_rate": 7.879632801297387e-08,
      "loss": 0.0394,
      "step": 1640
    },
    {
      "epoch": 2.9,
      "grad_norm": 0.4619080424308777,
      "learning_rate": 7.595945417206818e-08,
      "loss": 0.0433,
      "step": 1641
    },
    {
      "epoch": 2.9,
      "grad_norm": 0.4181370735168457,
      "learning_rate": 7.317445869007288e-08,
      "loss": 0.0697,
      "step": 1642
    },
    {
      "epoch": 2.91,
      "grad_norm": 0.18920888006687164,
      "learning_rate": 7.044135124808227e-08,
      "loss": 0.0199,
      "step": 1643
    },
    {
      "epoch": 2.91,
      "grad_norm": 0.21702830493450165,
      "learning_rate": 6.776014134681652e-08,
      "loss": 0.0334,
      "step": 1644
    },
    {
      "epoch": 2.91,
      "grad_norm": 0.4063786268234253,
      "learning_rate": 6.513083830659016e-08,
      "loss": 0.0588,
      "step": 1645
    },
    {
      "epoch": 2.91,
      "grad_norm": 0.5082572102546692,
      "learning_rate": 6.255345126728373e-08,
      "loss": 0.0657,
      "step": 1646
    },
    {
      "epoch": 2.91,
      "grad_norm": 0.4424477219581604,
      "learning_rate": 6.002798918830543e-08,
      "loss": 0.0688,
      "step": 1647
    },
    {
      "epoch": 2.92,
      "grad_norm": 0.4119831621646881,
      "learning_rate": 5.755446084856786e-08,
      "loss": 0.056,
      "step": 1648
    },
    {
      "epoch": 2.92,
      "grad_norm": 0.5998245477676392,
      "learning_rate": 5.513287484645302e-08,
      "loss": 0.0865,
      "step": 1649
    },
    {
      "epoch": 2.92,
      "grad_norm": 0.50651615858078,
      "learning_rate": 5.2763239599782357e-08,
      "loss": 0.1007,
      "step": 1650
    },
    {
      "epoch": 2.92,
      "grad_norm": 0.6716424822807312,
      "learning_rate": 5.044556334578676e-08,
      "loss": 0.0563,
      "step": 1651
    },
    {
      "epoch": 2.92,
      "grad_norm": 0.23979637026786804,
      "learning_rate": 4.8179854141079947e-08,
      "loss": 0.0243,
      "step": 1652
    },
    {
      "epoch": 2.92,
      "grad_norm": 0.5607647895812988,
      "learning_rate": 4.5966119861635125e-08,
      "loss": 0.0902,
      "step": 1653
    },
    {
      "epoch": 2.93,
      "grad_norm": 0.5742710828781128,
      "learning_rate": 4.3804368202741695e-08,
      "loss": 0.0658,
      "step": 1654
    },
    {
      "epoch": 2.93,
      "grad_norm": 0.3109664022922516,
      "learning_rate": 4.169460667900027e-08,
      "loss": 0.0513,
      "step": 1655
    },
    {
      "epoch": 2.93,
      "grad_norm": 0.24484005570411682,
      "learning_rate": 3.963684262427769e-08,
      "loss": 0.0247,
      "step": 1656
    },
    {
      "epoch": 2.93,
      "grad_norm": 0.22879289090633392,
      "learning_rate": 3.7631083191692064e-08,
      "loss": 0.0337,
      "step": 1657
    },
    {
      "epoch": 2.93,
      "grad_norm": 0.3116399347782135,
      "learning_rate": 3.567733535358608e-08,
      "loss": 0.0385,
      "step": 1658
    },
    {
      "epoch": 2.93,
      "grad_norm": 0.509244978427887,
      "learning_rate": 3.3775605901498755e-08,
      "loss": 0.0368,
      "step": 1659
    },
    {
      "epoch": 2.94,
      "grad_norm": 0.3884599208831787,
      "learning_rate": 3.1925901446148707e-08,
      "loss": 0.0669,
      "step": 1660
    },
    {
      "epoch": 2.94,
      "grad_norm": 0.5966331958770752,
      "learning_rate": 3.012822841740426e-08,
      "loss": 0.0499,
      "step": 1661
    },
    {
      "epoch": 2.94,
      "grad_norm": 0.33274173736572266,
      "learning_rate": 2.838259306426505e-08,
      "loss": 0.0687,
      "step": 1662
    },
    {
      "epoch": 2.94,
      "grad_norm": 0.4982815384864807,
      "learning_rate": 2.6689001454840432e-08,
      "loss": 0.0394,
      "step": 1663
    },
    {
      "epoch": 2.94,
      "grad_norm": 0.3643025755882263,
      "learning_rate": 2.5047459476324475e-08,
      "loss": 0.0579,
      "step": 1664
    },
    {
      "epoch": 2.95,
      "grad_norm": 0.39075028896331787,
      "learning_rate": 2.34579728349843e-08,
      "loss": 0.0595,
      "step": 1665
    },
    {
      "epoch": 2.95,
      "grad_norm": 0.553180992603302,
      "learning_rate": 2.192054705612678e-08,
      "loss": 0.0594,
      "step": 1666
    },
    {
      "epoch": 2.95,
      "grad_norm": 0.497694194316864,
      "learning_rate": 2.04351874840919e-08,
      "loss": 0.0677,
      "step": 1667
    },
    {
      "epoch": 2.95,
      "grad_norm": 0.5634981393814087,
      "learning_rate": 1.900189928222773e-08,
      "loss": 0.0714,
      "step": 1668
    },
    {
      "epoch": 2.95,
      "grad_norm": 0.5375871062278748,
      "learning_rate": 1.7620687432873818e-08,
      "loss": 0.0909,
      "step": 1669
    },
    {
      "epoch": 2.95,
      "grad_norm": 0.23803794384002686,
      "learning_rate": 1.6291556737344505e-08,
      "loss": 0.0204,
      "step": 1670
    },
    {
      "epoch": 2.96,
      "grad_norm": 0.4865075945854187,
      "learning_rate": 1.501451181591229e-08,
      "loss": 0.0757,
      "step": 1671
    },
    {
      "epoch": 2.96,
      "grad_norm": 0.4838518798351288,
      "learning_rate": 1.3789557107787842e-08,
      "loss": 0.0482,
      "step": 1672
    },
    {
      "epoch": 2.96,
      "grad_norm": 0.343869686126709,
      "learning_rate": 1.2616696871111666e-08,
      "loss": 0.0387,
      "step": 1673
    },
    {
      "epoch": 2.96,
      "grad_norm": 0.6620836853981018,
      "learning_rate": 1.149593518293246e-08,
      "loss": 0.1368,
      "step": 1674
    },
    {
      "epoch": 2.96,
      "grad_norm": 0.18913516402244568,
      "learning_rate": 1.0427275939200453e-08,
      "loss": 0.031,
      "step": 1675
    },
    {
      "epoch": 2.97,
      "grad_norm": 0.38958144187927246,
      "learning_rate": 9.410722854744092e-09,
      "loss": 0.0402,
      "step": 1676
    },
    {
      "epoch": 2.97,
      "grad_norm": 0.15901847183704376,
      "learning_rate": 8.446279463266704e-09,
      "loss": 0.0142,
      "step": 1677
    },
    {
      "epoch": 2.97,
      "grad_norm": 0.28944361209869385,
      "learning_rate": 7.533949117328187e-09,
      "loss": 0.0384,
      "step": 1678
    },
    {
      "epoch": 2.97,
      "grad_norm": 0.28497225046157837,
      "learning_rate": 6.673734988333347e-09,
      "loss": 0.0328,
      "step": 1679
    },
    {
      "epoch": 2.97,
      "grad_norm": 0.24431495368480682,
      "learning_rate": 5.865640066525235e-09,
      "loss": 0.0343,
      "step": 1680
    },
    {
      "epoch": 2.97,
      "grad_norm": 0.3041365444660187,
      "learning_rate": 5.109667160973497e-09,
      "loss": 0.0511,
      "step": 1681
    },
    {
      "epoch": 2.98,
      "grad_norm": 0.5336529016494751,
      "learning_rate": 4.405818899557712e-09,
      "loss": 0.0385,
      "step": 1682
    },
    {
      "epoch": 2.98,
      "grad_norm": 0.3261311948299408,
      "learning_rate": 3.754097728970729e-09,
      "loss": 0.0429,
      "step": 1683
    },
    {
      "epoch": 2.98,
      "grad_norm": 0.33494555950164795,
      "learning_rate": 3.1545059146970147e-09,
      "loss": 0.0271,
      "step": 1684
    },
    {
      "epoch": 2.98,
      "grad_norm": 0.8140985369682312,
      "learning_rate": 2.607045541015984e-09,
      "loss": 0.1147,
      "step": 1685
    },
    {
      "epoch": 2.98,
      "grad_norm": 0.42994144558906555,
      "learning_rate": 2.11171851098535e-09,
      "loss": 0.0489,
      "step": 1686
    },
    {
      "epoch": 2.98,
      "grad_norm": 0.3397289514541626,
      "learning_rate": 1.6685265464444489e-09,
      "loss": 0.0393,
      "step": 1687
    },
    {
      "epoch": 2.99,
      "grad_norm": 0.3589901030063629,
      "learning_rate": 1.277471187997592e-09,
      "loss": 0.0638,
      "step": 1688
    },
    {
      "epoch": 2.99,
      "grad_norm": 0.2832041084766388,
      "learning_rate": 9.385537950173938e-10,
      "loss": 0.0363,
      "step": 1689
    },
    {
      "epoch": 2.99,
      "grad_norm": 0.2788923680782318,
      "learning_rate": 6.517755456331154e-10,
      "loss": 0.0508,
      "step": 1690
    },
    {
      "epoch": 2.99,
      "grad_norm": 0.48392757773399353,
      "learning_rate": 4.1713743673232886e-10,
      "loss": 0.0746,
      "step": 1691
    },
    {
      "epoch": 2.99,
      "grad_norm": 0.34901174902915955,
      "learning_rate": 2.346402839559225e-10,
      "loss": 0.0949,
      "step": 1692
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.4794461131095886,
      "learning_rate": 1.0428472169310422e-10,
      "loss": 0.0607,
      "step": 1693
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.3746906816959381,
      "learning_rate": 2.6071203079736094e-11,
      "loss": 0.0369,
      "step": 1694
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.326846718788147,
      "learning_rate": 0.0,
      "loss": 0.0342,
      "step": 1695
    }
  ],
  "logging_steps": 1,
  "max_steps": 1695,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 565,
  "total_flos": 1.549439947809751e+17,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}