{
  "best_metric": 0.9820159673690796,
  "best_model_checkpoint": "miner_id_24/checkpoint-200",
  "epoch": 0.38077106139933364,
  "eval_steps": 50,
  "global_step": 200,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0019038553069966682,
      "grad_norm": 23.7499942779541,
      "learning_rate": 7e-06,
      "loss": 7.21,
      "step": 1
    },
    {
      "epoch": 0.0019038553069966682,
      "eval_loss": 2.085144281387329,
      "eval_runtime": 64.4441,
      "eval_samples_per_second": 13.733,
      "eval_steps_per_second": 3.445,
      "step": 1
    },
    {
      "epoch": 0.0038077106139933364,
      "grad_norm": 27.487117767333984,
      "learning_rate": 1.4e-05,
      "loss": 7.3893,
      "step": 2
    },
    {
      "epoch": 0.005711565920990005,
      "grad_norm": 24.298641204833984,
      "learning_rate": 2.1e-05,
      "loss": 7.4764,
      "step": 3
    },
    {
      "epoch": 0.007615421227986673,
      "grad_norm": 20.819686889648438,
      "learning_rate": 2.8e-05,
      "loss": 7.2008,
      "step": 4
    },
    {
      "epoch": 0.00951927653498334,
      "grad_norm": 17.770313262939453,
      "learning_rate": 3.5e-05,
      "loss": 6.8417,
      "step": 5
    },
    {
      "epoch": 0.01142313184198001,
      "grad_norm": 17.92336654663086,
      "learning_rate": 4.2e-05,
      "loss": 6.5631,
      "step": 6
    },
    {
      "epoch": 0.013326987148976678,
      "grad_norm": 16.182708740234375,
      "learning_rate": 4.899999999999999e-05,
      "loss": 6.0062,
      "step": 7
    },
    {
      "epoch": 0.015230842455973346,
      "grad_norm": 14.703605651855469,
      "learning_rate": 5.6e-05,
      "loss": 6.5074,
      "step": 8
    },
    {
      "epoch": 0.017134697762970014,
      "grad_norm": 15.21464729309082,
      "learning_rate": 6.3e-05,
      "loss": 6.0345,
      "step": 9
    },
    {
      "epoch": 0.01903855306996668,
      "grad_norm": 13.005841255187988,
      "learning_rate": 7e-05,
      "loss": 5.738,
      "step": 10
    },
    {
      "epoch": 0.020942408376963352,
      "grad_norm": 13.063154220581055,
      "learning_rate": 6.999521567473641e-05,
      "loss": 5.8279,
      "step": 11
    },
    {
      "epoch": 0.02284626368396002,
      "grad_norm": 12.097558975219727,
      "learning_rate": 6.998086400693241e-05,
      "loss": 5.8253,
      "step": 12
    },
    {
      "epoch": 0.024750118990956686,
      "grad_norm": 11.619023323059082,
      "learning_rate": 6.995694892019065e-05,
      "loss": 5.7643,
      "step": 13
    },
    {
      "epoch": 0.026653974297953357,
      "grad_norm": 13.587701797485352,
      "learning_rate": 6.99234769526571e-05,
      "loss": 5.5242,
      "step": 14
    },
    {
      "epoch": 0.028557829604950024,
      "grad_norm": 11.041449546813965,
      "learning_rate": 6.988045725523343e-05,
      "loss": 5.3918,
      "step": 15
    },
    {
      "epoch": 0.03046168491194669,
      "grad_norm": 12.899657249450684,
      "learning_rate": 6.982790158907539e-05,
      "loss": 5.7544,
      "step": 16
    },
    {
      "epoch": 0.03236554021894336,
      "grad_norm": 13.852838516235352,
      "learning_rate": 6.976582432237733e-05,
      "loss": 5.3847,
      "step": 17
    },
    {
      "epoch": 0.03426939552594003,
      "grad_norm": 11.95140552520752,
      "learning_rate": 6.969424242644413e-05,
      "loss": 5.6636,
      "step": 18
    },
    {
      "epoch": 0.036173250832936696,
      "grad_norm": 11.032059669494629,
      "learning_rate": 6.961317547105138e-05,
      "loss": 5.0469,
      "step": 19
    },
    {
      "epoch": 0.03807710613993336,
      "grad_norm": 11.564652442932129,
      "learning_rate": 6.952264561909527e-05,
      "loss": 5.602,
      "step": 20
    },
    {
      "epoch": 0.03998096144693003,
      "grad_norm": 11.808911323547363,
      "learning_rate": 6.942267762053337e-05,
      "loss": 5.9484,
      "step": 21
    },
    {
      "epoch": 0.041884816753926704,
      "grad_norm": 10.792513847351074,
      "learning_rate": 6.931329880561832e-05,
      "loss": 5.3864,
      "step": 22
    },
    {
      "epoch": 0.04378867206092337,
      "grad_norm": 10.414179801940918,
      "learning_rate": 6.919453907742597e-05,
      "loss": 5.5784,
      "step": 23
    },
    {
      "epoch": 0.04569252736792004,
      "grad_norm": 10.658806800842285,
      "learning_rate": 6.90664309036802e-05,
      "loss": 5.2816,
      "step": 24
    },
    {
      "epoch": 0.047596382674916705,
      "grad_norm": 11.604804039001465,
      "learning_rate": 6.892900930787656e-05,
      "loss": 5.4807,
      "step": 25
    },
    {
      "epoch": 0.04950023798191337,
      "grad_norm": 10.047147750854492,
      "learning_rate": 6.87823118597072e-05,
      "loss": 5.3292,
      "step": 26
    },
    {
      "epoch": 0.05140409328891004,
      "grad_norm": 10.286588668823242,
      "learning_rate": 6.862637866478969e-05,
      "loss": 5.2082,
      "step": 27
    },
    {
      "epoch": 0.053307948595906714,
      "grad_norm": 10.427512168884277,
      "learning_rate": 6.846125235370252e-05,
      "loss": 5.3268,
      "step": 28
    },
    {
      "epoch": 0.05521180390290338,
      "grad_norm": 10.335875511169434,
      "learning_rate": 6.828697807033038e-05,
      "loss": 5.4725,
      "step": 29
    },
    {
      "epoch": 0.05711565920990005,
      "grad_norm": 10.073786735534668,
      "learning_rate": 6.81036034595222e-05,
      "loss": 5.3723,
      "step": 30
    },
    {
      "epoch": 0.059019514516896715,
      "grad_norm": 9.964614868164062,
      "learning_rate": 6.791117865406564e-05,
      "loss": 5.3471,
      "step": 31
    },
    {
      "epoch": 0.06092336982389338,
      "grad_norm": 9.527873992919922,
      "learning_rate": 6.770975626098112e-05,
      "loss": 5.1568,
      "step": 32
    },
    {
      "epoch": 0.06282722513089005,
      "grad_norm": 10.344468116760254,
      "learning_rate": 6.749939134713974e-05,
      "loss": 5.3934,
      "step": 33
    },
    {
      "epoch": 0.06473108043788672,
      "grad_norm": 10.566803932189941,
      "learning_rate": 6.728014142420846e-05,
      "loss": 5.5978,
      "step": 34
    },
    {
      "epoch": 0.06663493574488338,
      "grad_norm": 9.318648338317871,
      "learning_rate": 6.7052066432927e-05,
      "loss": 5.2326,
      "step": 35
    },
    {
      "epoch": 0.06853879105188006,
      "grad_norm": 10.439335823059082,
      "learning_rate": 6.681522872672069e-05,
      "loss": 5.4662,
      "step": 36
    },
    {
      "epoch": 0.07044264635887673,
      "grad_norm": 9.824296951293945,
      "learning_rate": 6.656969305465356e-05,
      "loss": 5.2866,
      "step": 37
    },
    {
      "epoch": 0.07234650166587339,
      "grad_norm": 9.726062774658203,
      "learning_rate": 6.631552654372672e-05,
      "loss": 5.2126,
      "step": 38
    },
    {
      "epoch": 0.07425035697287007,
      "grad_norm": 11.502147674560547,
      "learning_rate": 6.60527986805264e-05,
      "loss": 5.6166,
      "step": 39
    },
    {
      "epoch": 0.07615421227986673,
      "grad_norm": 10.295775413513184,
      "learning_rate": 6.578158129222711e-05,
      "loss": 5.5236,
      "step": 40
    },
    {
      "epoch": 0.0780580675868634,
      "grad_norm": 9.807427406311035,
      "learning_rate": 6.550194852695469e-05,
      "loss": 5.4277,
      "step": 41
    },
    {
      "epoch": 0.07996192289386006,
      "grad_norm": 10.789260864257812,
      "learning_rate": 6.521397683351509e-05,
      "loss": 5.74,
      "step": 42
    },
    {
      "epoch": 0.08186577820085673,
      "grad_norm": 10.294041633605957,
      "learning_rate": 6.491774494049386e-05,
      "loss": 5.4681,
      "step": 43
    },
    {
      "epoch": 0.08376963350785341,
      "grad_norm": 9.753371238708496,
      "learning_rate": 6.461333383473272e-05,
      "loss": 5.6462,
      "step": 44
    },
    {
      "epoch": 0.08567348881485007,
      "grad_norm": 9.903716087341309,
      "learning_rate": 6.430082673918849e-05,
      "loss": 5.2489,
      "step": 45
    },
    {
      "epoch": 0.08757734412184674,
      "grad_norm": 9.664334297180176,
      "learning_rate": 6.398030909018069e-05,
      "loss": 5.2939,
      "step": 46
    },
    {
      "epoch": 0.0894811994288434,
      "grad_norm": 10.391850471496582,
      "learning_rate": 6.365186851403423e-05,
      "loss": 5.8567,
      "step": 47
    },
    {
      "epoch": 0.09138505473584008,
      "grad_norm": 11.669763565063477,
      "learning_rate": 6.331559480312315e-05,
      "loss": 5.7577,
      "step": 48
    },
    {
      "epoch": 0.09328891004283675,
      "grad_norm": 9.692889213562012,
      "learning_rate": 6.297157989132236e-05,
      "loss": 5.4625,
      "step": 49
    },
    {
      "epoch": 0.09519276534983341,
      "grad_norm": 11.684773445129395,
      "learning_rate": 6.261991782887377e-05,
      "loss": 5.6681,
      "step": 50
    },
    {
      "epoch": 0.09519276534983341,
      "eval_loss": 1.3040165901184082,
      "eval_runtime": 65.4623,
      "eval_samples_per_second": 13.519,
      "eval_steps_per_second": 3.391,
      "step": 50
    },
    {
      "epoch": 0.09709662065683008,
      "grad_norm": 15.806927680969238,
      "learning_rate": 6.226070475667393e-05,
      "loss": 5.6676,
      "step": 51
    },
    {
      "epoch": 0.09900047596382675,
      "grad_norm": 11.198895454406738,
      "learning_rate": 6.189403887999006e-05,
      "loss": 5.4295,
      "step": 52
    },
    {
      "epoch": 0.10090433127082342,
      "grad_norm": 8.934762001037598,
      "learning_rate": 6.152002044161171e-05,
      "loss": 5.408,
      "step": 53
    },
    {
      "epoch": 0.10280818657782008,
      "grad_norm": 7.409307479858398,
      "learning_rate": 6.113875169444539e-05,
      "loss": 5.2086,
      "step": 54
    },
    {
      "epoch": 0.10471204188481675,
      "grad_norm": 8.708901405334473,
      "learning_rate": 6.0750336873559605e-05,
      "loss": 5.4806,
      "step": 55
    },
    {
      "epoch": 0.10661589719181343,
      "grad_norm": 8.244229316711426,
      "learning_rate": 6.035488216768811e-05,
      "loss": 5.1627,
      "step": 56
    },
    {
      "epoch": 0.10851975249881009,
      "grad_norm": 7.6382060050964355,
      "learning_rate": 5.9952495690198894e-05,
      "loss": 4.7832,
      "step": 57
    },
    {
      "epoch": 0.11042360780580676,
      "grad_norm": 7.794398784637451,
      "learning_rate": 5.954328744953709e-05,
      "loss": 4.759,
      "step": 58
    },
    {
      "epoch": 0.11232746311280342,
      "grad_norm": 8.159358978271484,
      "learning_rate": 5.91273693191498e-05,
      "loss": 4.8963,
      "step": 59
    },
    {
      "epoch": 0.1142313184198001,
      "grad_norm": 8.03155517578125,
      "learning_rate": 5.870485500690094e-05,
      "loss": 5.1858,
      "step": 60
    },
    {
      "epoch": 0.11613517372679677,
      "grad_norm": 9.014118194580078,
      "learning_rate": 5.827586002398468e-05,
      "loss": 5.2089,
      "step": 61
    },
    {
      "epoch": 0.11803902903379343,
      "grad_norm": 8.561113357543945,
      "learning_rate": 5.784050165334589e-05,
      "loss": 4.8377,
      "step": 62
    },
    {
      "epoch": 0.1199428843407901,
      "grad_norm": 9.359479904174805,
      "learning_rate": 5.739889891761608e-05,
      "loss": 4.8591,
      "step": 63
    },
    {
      "epoch": 0.12184673964778676,
      "grad_norm": 8.884666442871094,
      "learning_rate": 5.6951172546573794e-05,
      "loss": 4.8723,
      "step": 64
    },
    {
      "epoch": 0.12375059495478344,
      "grad_norm": 8.695745468139648,
      "learning_rate": 5.6497444944138376e-05,
      "loss": 5.0063,
      "step": 65
    },
    {
      "epoch": 0.1256544502617801,
      "grad_norm": 7.5747456550598145,
      "learning_rate": 5.603784015490587e-05,
      "loss": 4.7092,
      "step": 66
    },
    {
      "epoch": 0.1275583055687768,
      "grad_norm": 7.801674842834473,
      "learning_rate": 5.557248383023655e-05,
      "loss": 4.7957,
      "step": 67
    },
    {
      "epoch": 0.12946216087577345,
      "grad_norm": 8.105006217956543,
      "learning_rate": 5.510150319390302e-05,
      "loss": 4.4399,
      "step": 68
    },
    {
      "epoch": 0.1313660161827701,
      "grad_norm": 8.557613372802734,
      "learning_rate": 5.4625027007308546e-05,
      "loss": 4.885,
      "step": 69
    },
    {
      "epoch": 0.13326987148976677,
      "grad_norm": 11.274003028869629,
      "learning_rate": 5.414318553428494e-05,
      "loss": 4.8082,
      "step": 70
    },
    {
      "epoch": 0.13517372679676345,
      "grad_norm": 11.312886238098145,
      "learning_rate": 5.3656110505479776e-05,
      "loss": 4.7281,
      "step": 71
    },
    {
      "epoch": 0.13707758210376011,
      "grad_norm": 8.340498924255371,
      "learning_rate": 5.316393508234253e-05,
      "loss": 4.7314,
      "step": 72
    },
    {
      "epoch": 0.13898143741075678,
      "grad_norm": 8.157063484191895,
      "learning_rate": 5.266679382071953e-05,
      "loss": 4.9252,
      "step": 73
    },
    {
      "epoch": 0.14088529271775346,
      "grad_norm": 8.306255340576172,
      "learning_rate": 5.216482263406778e-05,
      "loss": 4.6616,
      "step": 74
    },
    {
      "epoch": 0.14278914802475012,
      "grad_norm": 8.908082962036133,
      "learning_rate": 5.1658158756297576e-05,
      "loss": 4.8742,
      "step": 75
    },
    {
      "epoch": 0.14469300333174678,
      "grad_norm": 7.921310901641846,
      "learning_rate": 5.114694070425407e-05,
      "loss": 4.4142,
      "step": 76
    },
    {
      "epoch": 0.14659685863874344,
      "grad_norm": 7.845428466796875,
      "learning_rate": 5.063130823984823e-05,
      "loss": 4.4547,
      "step": 77
    },
    {
      "epoch": 0.14850071394574013,
      "grad_norm": 8.892997741699219,
      "learning_rate": 5.011140233184724e-05,
      "loss": 5.108,
      "step": 78
    },
    {
      "epoch": 0.1504045692527368,
      "grad_norm": 9.606233596801758,
      "learning_rate": 4.958736511733516e-05,
      "loss": 4.6994,
      "step": 79
    },
    {
      "epoch": 0.15230842455973345,
      "grad_norm": 8.047012329101562,
      "learning_rate": 4.905933986285393e-05,
      "loss": 4.6938,
      "step": 80
    },
    {
      "epoch": 0.15421227986673014,
      "grad_norm": 9.193869590759277,
      "learning_rate": 4.8527470925235824e-05,
      "loss": 4.8014,
      "step": 81
    },
    {
      "epoch": 0.1561161351737268,
      "grad_norm": 7.83010196685791,
      "learning_rate": 4.799190371213772e-05,
      "loss": 4.6519,
      "step": 82
    },
    {
      "epoch": 0.15801999048072346,
      "grad_norm": 7.466574668884277,
      "learning_rate": 4.745278464228808e-05,
      "loss": 4.4398,
      "step": 83
    },
    {
      "epoch": 0.15992384578772012,
      "grad_norm": 8.107156753540039,
      "learning_rate": 4.69102611054575e-05,
      "loss": 4.6071,
      "step": 84
    },
    {
      "epoch": 0.1618277010947168,
      "grad_norm": 8.724653244018555,
      "learning_rate": 4.6364481422163926e-05,
      "loss": 4.9037,
      "step": 85
    },
    {
      "epoch": 0.16373155640171347,
      "grad_norm": 7.404033660888672,
      "learning_rate": 4.581559480312316e-05,
      "loss": 4.3696,
      "step": 86
    },
    {
      "epoch": 0.16563541170871013,
      "grad_norm": 8.051767349243164,
      "learning_rate": 4.526375130845627e-05,
      "loss": 4.5589,
      "step": 87
    },
    {
      "epoch": 0.16753926701570682,
      "grad_norm": 9.341611862182617,
      "learning_rate": 4.4709101806664554e-05,
      "loss": 4.7734,
      "step": 88
    },
    {
      "epoch": 0.16944312232270348,
      "grad_norm": 8.678064346313477,
      "learning_rate": 4.4151797933383685e-05,
      "loss": 4.6396,
      "step": 89
    },
    {
      "epoch": 0.17134697762970014,
      "grad_norm": 8.946483612060547,
      "learning_rate": 4.359199204992797e-05,
      "loss": 4.7698,
      "step": 90
    },
    {
      "epoch": 0.17325083293669682,
      "grad_norm": 8.687679290771484,
      "learning_rate": 4.30298372016363e-05,
      "loss": 4.7269,
      "step": 91
    },
    {
      "epoch": 0.17515468824369348,
      "grad_norm": 10.165061950683594,
      "learning_rate": 4.246548707603114e-05,
      "loss": 4.9481,
      "step": 92
    },
    {
      "epoch": 0.17705854355069015,
      "grad_norm": 8.327232360839844,
      "learning_rate": 4.1899095960801805e-05,
      "loss": 4.6007,
      "step": 93
    },
    {
      "epoch": 0.1789623988576868,
      "grad_norm": 8.452240943908691,
      "learning_rate": 4.133081870162385e-05,
      "loss": 4.6878,
      "step": 94
    },
    {
      "epoch": 0.1808662541646835,
      "grad_norm": 8.48222827911377,
      "learning_rate": 4.076081065982569e-05,
      "loss": 4.4604,
      "step": 95
    },
    {
      "epoch": 0.18277010947168015,
      "grad_norm": 9.326024055480957,
      "learning_rate": 4.018922766991447e-05,
      "loss": 4.7372,
      "step": 96
    },
    {
      "epoch": 0.1846739647786768,
      "grad_norm": 9.113911628723145,
      "learning_rate": 3.961622599697241e-05,
      "loss": 4.4883,
      "step": 97
    },
    {
      "epoch": 0.1865778200856735,
      "grad_norm": 8.056068420410156,
      "learning_rate": 3.9041962293935516e-05,
      "loss": 4.8074,
      "step": 98
    },
    {
      "epoch": 0.18848167539267016,
      "grad_norm": 8.778728485107422,
      "learning_rate": 3.84665935587662e-05,
      "loss": 4.4564,
      "step": 99
    },
    {
      "epoch": 0.19038553069966682,
      "grad_norm": 8.800410270690918,
      "learning_rate": 3.7890277091531636e-05,
      "loss": 4.6041,
      "step": 100
    },
    {
      "epoch": 0.19038553069966682,
      "eval_loss": 1.1414027214050293,
      "eval_runtime": 65.4902,
      "eval_samples_per_second": 13.513,
      "eval_steps_per_second": 3.39,
      "step": 100
    },
    {
      "epoch": 0.19228938600666348,
      "grad_norm": 8.70529556274414,
      "learning_rate": 3.7313170451399475e-05,
      "loss": 5.0098,
      "step": 101
    },
    {
      "epoch": 0.19419324131366017,
      "grad_norm": 8.006318092346191,
      "learning_rate": 3.673543141356278e-05,
      "loss": 5.0347,
      "step": 102
    },
    {
      "epoch": 0.19609709662065683,
      "grad_norm": 7.4283013343811035,
      "learning_rate": 3.6157217926105783e-05,
      "loss": 4.8268,
      "step": 103
    },
    {
      "epoch": 0.1980009519276535,
      "grad_norm": 6.7939605712890625,
      "learning_rate": 3.557868806682255e-05,
      "loss": 4.7183,
      "step": 104
    },
    {
      "epoch": 0.19990480723465018,
      "grad_norm": 6.694666862487793,
      "learning_rate": 3.5e-05,
      "loss": 4.6545,
      "step": 105
    },
    {
      "epoch": 0.20180866254164684,
      "grad_norm": 6.7515692710876465,
      "learning_rate": 3.442131193317745e-05,
      "loss": 4.5719,
      "step": 106
    },
    {
      "epoch": 0.2037125178486435,
      "grad_norm": 6.9047932624816895,
      "learning_rate": 3.384278207389421e-05,
      "loss": 4.3211,
      "step": 107
    },
    {
      "epoch": 0.20561637315564016,
      "grad_norm": 8.433969497680664,
      "learning_rate": 3.3264568586437216e-05,
      "loss": 4.414,
      "step": 108
    },
    {
      "epoch": 0.20752022846263685,
      "grad_norm": 6.5694050788879395,
      "learning_rate": 3.268682954860052e-05,
      "loss": 4.2368,
      "step": 109
    },
    {
      "epoch": 0.2094240837696335,
      "grad_norm": 7.584961891174316,
      "learning_rate": 3.210972290846837e-05,
      "loss": 4.6716,
      "step": 110
    },
    {
      "epoch": 0.21132793907663017,
      "grad_norm": 6.995484828948975,
      "learning_rate": 3.15334064412338e-05,
      "loss": 4.3203,
      "step": 111
    },
    {
      "epoch": 0.21323179438362685,
      "grad_norm": 7.1296467781066895,
      "learning_rate": 3.0958037706064485e-05,
      "loss": 4.5241,
      "step": 112
    },
    {
      "epoch": 0.21513564969062351,
      "grad_norm": 7.044524669647217,
      "learning_rate": 3.038377400302758e-05,
      "loss": 4.4863,
      "step": 113
    },
    {
      "epoch": 0.21703950499762018,
      "grad_norm": 7.043889045715332,
      "learning_rate": 2.9810772330085524e-05,
      "loss": 4.4081,
      "step": 114
    },
    {
      "epoch": 0.21894336030461684,
      "grad_norm": 8.038551330566406,
      "learning_rate": 2.9239189340174306e-05,
      "loss": 4.5455,
      "step": 115
    },
    {
      "epoch": 0.22084721561161352,
      "grad_norm": 7.743556499481201,
      "learning_rate": 2.8669181298376163e-05,
      "loss": 4.3879,
      "step": 116
    },
    {
      "epoch": 0.22275107091861018,
      "grad_norm": 7.689728260040283,
      "learning_rate": 2.8100904039198193e-05,
      "loss": 4.3637,
      "step": 117
    },
    {
      "epoch": 0.22465492622560684,
      "grad_norm": 7.642689228057861,
      "learning_rate": 2.7534512923968863e-05,
      "loss": 4.5594,
      "step": 118
    },
    {
      "epoch": 0.22655878153260353,
      "grad_norm": 8.036393165588379,
      "learning_rate": 2.6970162798363695e-05,
      "loss": 4.2413,
      "step": 119
    },
    {
      "epoch": 0.2284626368396002,
      "grad_norm": 7.012694358825684,
      "learning_rate": 2.640800795007203e-05,
      "loss": 4.2993,
      "step": 120
    },
    {
      "epoch": 0.23036649214659685,
      "grad_norm": 7.756348133087158,
      "learning_rate": 2.5848202066616305e-05,
      "loss": 4.1293,
      "step": 121
    },
    {
      "epoch": 0.23227034745359354,
      "grad_norm": 7.184790134429932,
      "learning_rate": 2.5290898193335446e-05,
      "loss": 4.0667,
      "step": 122
    },
    {
      "epoch": 0.2341742027605902,
      "grad_norm": 6.861097812652588,
      "learning_rate": 2.4736248691543736e-05,
      "loss": 3.9088,
      "step": 123
    },
    {
      "epoch": 0.23607805806758686,
      "grad_norm": 7.792126655578613,
      "learning_rate": 2.4184405196876842e-05,
      "loss": 4.0982,
      "step": 124
    },
    {
      "epoch": 0.23798191337458352,
      "grad_norm": 8.279322624206543,
      "learning_rate": 2.363551857783608e-05,
      "loss": 4.273,
      "step": 125
    },
    {
      "epoch": 0.2398857686815802,
      "grad_norm": 7.2576494216918945,
      "learning_rate": 2.308973889454249e-05,
      "loss": 4.133,
      "step": 126
    },
    {
      "epoch": 0.24178962398857687,
      "grad_norm": 8.2999906539917,
      "learning_rate": 2.2547215357711918e-05,
      "loss": 4.0224,
      "step": 127
    },
    {
      "epoch": 0.24369347929557353,
      "grad_norm": 8.445121765136719,
      "learning_rate": 2.2008096287862266e-05,
      "loss": 4.2589,
      "step": 128
    },
    {
      "epoch": 0.24559733460257022,
      "grad_norm": 7.673584938049316,
      "learning_rate": 2.1472529074764177e-05,
      "loss": 4.3057,
      "step": 129
    },
    {
      "epoch": 0.24750118990956688,
      "grad_norm": 7.786920547485352,
      "learning_rate": 2.0940660137146074e-05,
      "loss": 4.2528,
      "step": 130
    },
    {
      "epoch": 0.24940504521656354,
      "grad_norm": 7.61902379989624,
      "learning_rate": 2.041263488266484e-05,
      "loss": 3.7492,
      "step": 131
    },
    {
      "epoch": 0.2513089005235602,
      "grad_norm": 7.465391159057617,
      "learning_rate": 1.988859766815275e-05,
      "loss": 3.981,
      "step": 132
    },
    {
      "epoch": 0.25321275583055686,
      "grad_norm": 7.733811855316162,
      "learning_rate": 1.9368691760151773e-05,
      "loss": 4.1569,
      "step": 133
    },
    {
      "epoch": 0.2551166111375536,
      "grad_norm": 7.370545387268066,
      "learning_rate": 1.885305929574593e-05,
      "loss": 3.874,
      "step": 134
    },
    {
      "epoch": 0.25702046644455023,
      "grad_norm": 8.920854568481445,
      "learning_rate": 1.8341841243702424e-05,
      "loss": 4.3121,
      "step": 135
    },
    {
      "epoch": 0.2589243217515469,
      "grad_norm": 9.357240676879883,
      "learning_rate": 1.7835177365932225e-05,
      "loss": 4.524,
      "step": 136
    },
    {
      "epoch": 0.26082817705854355,
      "grad_norm": 8.812352180480957,
      "learning_rate": 1.7333206179280478e-05,
      "loss": 3.9994,
      "step": 137
    },
    {
      "epoch": 0.2627320323655402,
      "grad_norm": 8.463519096374512,
      "learning_rate": 1.6836064917657478e-05,
      "loss": 4.0683,
      "step": 138
    },
    {
      "epoch": 0.2646358876725369,
      "grad_norm": 7.832219123840332,
      "learning_rate": 1.6343889494520224e-05,
      "loss": 3.9524,
      "step": 139
    },
    {
      "epoch": 0.26653974297953353,
      "grad_norm": 8.509306907653809,
      "learning_rate": 1.5856814465715064e-05,
      "loss": 3.9898,
      "step": 140
    },
    {
      "epoch": 0.26844359828653025,
      "grad_norm": 8.911566734313965,
      "learning_rate": 1.5374972992691458e-05,
      "loss": 3.9925,
      "step": 141
    },
    {
      "epoch": 0.2703474535935269,
      "grad_norm": 9.420488357543945,
      "learning_rate": 1.4898496806096974e-05,
      "loss": 4.1377,
      "step": 142
    },
    {
      "epoch": 0.27225130890052357,
      "grad_norm": 8.05871868133545,
      "learning_rate": 1.4427516169763444e-05,
      "loss": 3.8602,
      "step": 143
    },
    {
      "epoch": 0.27415516420752023,
      "grad_norm": 8.699991226196289,
      "learning_rate": 1.396215984509412e-05,
      "loss": 4.0414,
      "step": 144
    },
    {
      "epoch": 0.2760590195145169,
      "grad_norm": 8.904756546020508,
      "learning_rate": 1.3502555055861625e-05,
      "loss": 4.4121,
      "step": 145
    },
    {
      "epoch": 0.27796287482151355,
      "grad_norm": 10.7162504196167,
      "learning_rate": 1.3048827453426203e-05,
      "loss": 4.5674,
      "step": 146
    },
    {
      "epoch": 0.2798667301285102,
      "grad_norm": 10.042287826538086,
      "learning_rate": 1.2601101082383917e-05,
      "loss": 4.6137,
      "step": 147
    },
    {
      "epoch": 0.2817705854355069,
      "grad_norm": 9.6063871383667,
      "learning_rate": 1.2159498346654094e-05,
      "loss": 4.24,
      "step": 148
    },
    {
      "epoch": 0.2836744407425036,
      "grad_norm": 11.02759838104248,
      "learning_rate": 1.1724139976015306e-05,
      "loss": 4.5189,
      "step": 149
    },
    {
      "epoch": 0.28557829604950025,
      "grad_norm": 9.639123916625977,
      "learning_rate": 1.1295144993099068e-05,
      "loss": 4.4337,
      "step": 150
    },
    {
      "epoch": 0.28557829604950025,
      "eval_loss": 1.0172055959701538,
      "eval_runtime": 65.4679,
      "eval_samples_per_second": 13.518,
      "eval_steps_per_second": 3.391,
      "step": 150
    },
    {
      "epoch": 0.2874821513564969,
      "grad_norm": 7.883493900299072,
      "learning_rate": 1.0872630680850196e-05,
      "loss": 4.6472,
      "step": 151
    },
    {
      "epoch": 0.28938600666349357,
      "grad_norm": 8.46251106262207,
      "learning_rate": 1.0456712550462898e-05,
      "loss": 4.3637,
      "step": 152
    },
    {
      "epoch": 0.2912898619704902,
      "grad_norm": 8.094184875488281,
      "learning_rate": 1.0047504309801104e-05,
      "loss": 4.2769,
      "step": 153
    },
    {
      "epoch": 0.2931937172774869,
      "grad_norm": 8.725149154663086,
      "learning_rate": 9.645117832311886e-06,
      "loss": 4.4069,
      "step": 154
    },
    {
      "epoch": 0.2950975725844836,
      "grad_norm": 7.509702205657959,
      "learning_rate": 9.249663126440394e-06,
      "loss": 3.7436,
      "step": 155
    },
    {
      "epoch": 0.29700142789148026,
      "grad_norm": 8.107771873474121,
      "learning_rate": 8.861248305554624e-06,
      "loss": 4.3415,
      "step": 156
    },
    {
      "epoch": 0.2989052831984769,
      "grad_norm": 7.017758369445801,
      "learning_rate": 8.47997955838829e-06,
      "loss": 4.0017,
      "step": 157
    },
    {
      "epoch": 0.3008091385054736,
      "grad_norm": 7.199747562408447,
      "learning_rate": 8.10596112000994e-06,
      "loss": 4.1551,
      "step": 158
    },
    {
      "epoch": 0.30271299381247024,
      "grad_norm": 6.826944351196289,
      "learning_rate": 7.739295243326067e-06,
      "loss": 3.8882,
      "step": 159
    },
    {
      "epoch": 0.3046168491194669,
      "grad_norm": 7.0702972412109375,
      "learning_rate": 7.380082171126228e-06,
      "loss": 4.1228,
      "step": 160
    },
    {
      "epoch": 0.30652070442646356,
      "grad_norm": 8.110949516296387,
      "learning_rate": 7.028420108677635e-06,
      "loss": 4.2441,
      "step": 161
    },
    {
      "epoch": 0.3084245597334603,
      "grad_norm": 6.716869354248047,
      "learning_rate": 6.684405196876842e-06,
      "loss": 3.8069,
      "step": 162
    },
    {
      "epoch": 0.31032841504045694,
      "grad_norm": 7.542235374450684,
      "learning_rate": 6.3481314859657675e-06,
      "loss": 4.1167,
      "step": 163
    },
    {
      "epoch": 0.3122322703474536,
      "grad_norm": 7.276134014129639,
      "learning_rate": 6.019690909819298e-06,
      "loss": 3.9686,
      "step": 164
    },
    {
      "epoch": 0.31413612565445026,
      "grad_norm": 7.606207370758057,
      "learning_rate": 5.6991732608115e-06,
      "loss": 3.8366,
      "step": 165
    },
    {
      "epoch": 0.3160399809614469,
      "grad_norm": 7.363119125366211,
      "learning_rate": 5.386666165267256e-06,
      "loss": 3.9421,
      "step": 166
    },
    {
      "epoch": 0.3179438362684436,
      "grad_norm": 7.743343830108643,
      "learning_rate": 5.08225505950613e-06,
      "loss": 4.0282,
      "step": 167
    },
    {
      "epoch": 0.31984769157544024,
      "grad_norm": 7.004614353179932,
      "learning_rate": 4.786023166484913e-06,
      "loss": 3.921,
      "step": 168
    },
    {
      "epoch": 0.32175154688243696,
      "grad_norm": 7.168651103973389,
      "learning_rate": 4.498051473045291e-06,
      "loss": 3.9259,
      "step": 169
    },
    {
      "epoch": 0.3236554021894336,
      "grad_norm": 7.944221496582031,
      "learning_rate": 4.218418707772886e-06,
      "loss": 3.9505,
      "step": 170
    },
    {
      "epoch": 0.3255592574964303,
      "grad_norm": 7.535454273223877,
      "learning_rate": 3.947201319473587e-06,
      "loss": 3.976,
      "step": 171
    },
    {
      "epoch": 0.32746311280342694,
      "grad_norm": 7.3209075927734375,
      "learning_rate": 3.684473456273278e-06,
      "loss": 3.9647,
      "step": 172
    },
    {
      "epoch": 0.3293669681104236,
      "grad_norm": 7.745491981506348,
      "learning_rate": 3.4303069453464383e-06,
      "loss": 3.6598,
      "step": 173
    },
    {
      "epoch": 0.33127082341742026,
      "grad_norm": 7.710866451263428,
      "learning_rate": 3.184771273279312e-06,
      "loss": 4.0528,
      "step": 174
    },
    {
      "epoch": 0.3331746787244169,
      "grad_norm": 7.628995895385742,
      "learning_rate": 2.947933567072987e-06,
      "loss": 4.0274,
      "step": 175
    },
    {
      "epoch": 0.33507853403141363,
      "grad_norm": 8.762227058410645,
      "learning_rate": 2.719858575791534e-06,
      "loss": 4.0017,
      "step": 176
    },
    {
      "epoch": 0.3369823893384103,
      "grad_norm": 7.491204738616943,
      "learning_rate": 2.500608652860256e-06,
      "loss": 4.0085,
      "step": 177
    },
    {
      "epoch": 0.33888624464540695,
      "grad_norm": 7.047388076782227,
      "learning_rate": 2.2902437390188737e-06,
      "loss": 3.5474,
      "step": 178
    },
    {
      "epoch": 0.3407900999524036,
      "grad_norm": 7.473215579986572,
      "learning_rate": 2.0888213459343587e-06,
      "loss": 3.9556,
      "step": 179
    },
    {
      "epoch": 0.3426939552594003,
      "grad_norm": 8.065407752990723,
      "learning_rate": 1.8963965404777875e-06,
      "loss": 3.7152,
      "step": 180
    },
    {
      "epoch": 0.34459781056639693,
      "grad_norm": 7.622908592224121,
      "learning_rate": 1.7130219296696263e-06,
      "loss": 3.9459,
      "step": 181
    },
    {
      "epoch": 0.34650166587339365,
      "grad_norm": 7.5590128898620605,
      "learning_rate": 1.5387476462974824e-06,
      "loss": 4.0831,
      "step": 182
    },
    {
      "epoch": 0.3484055211803903,
      "grad_norm": 7.376468658447266,
      "learning_rate": 1.3736213352103147e-06,
      "loss": 3.6449,
      "step": 183
    },
    {
      "epoch": 0.35030937648738697,
      "grad_norm": 8.028775215148926,
      "learning_rate": 1.2176881402928002e-06,
      "loss": 3.8625,
      "step": 184
    },
    {
      "epoch": 0.35221323179438363,
      "grad_norm": 8.449442863464355,
      "learning_rate": 1.0709906921234367e-06,
      "loss": 4.0228,
      "step": 185
    },
    {
      "epoch": 0.3541170871013803,
      "grad_norm": 7.713665008544922,
      "learning_rate": 9.33569096319799e-07,
      "loss": 3.9626,
      "step": 186
    },
    {
      "epoch": 0.35602094240837695,
      "grad_norm": 8.763521194458008,
      "learning_rate": 8.054609225740255e-07,
      "loss": 4.0449,
      "step": 187
    },
    {
      "epoch": 0.3579247977153736,
      "grad_norm": 7.908934593200684,
      "learning_rate": 6.867011943816724e-07,
      "loss": 3.8442,
      "step": 188
    },
    {
      "epoch": 0.3598286530223703,
      "grad_norm": 9.312561988830566,
      "learning_rate": 5.77322379466617e-07,
      "loss": 4.4175,
      "step": 189
    },
    {
      "epoch": 0.361732508329367,
      "grad_norm": 8.53190803527832,
      "learning_rate": 4.773543809047186e-07,
      "loss": 4.0274,
      "step": 190
    },
    {
      "epoch": 0.36363636363636365,
      "grad_norm": 7.50021505355835,
      "learning_rate": 3.868245289486027e-07,
      "loss": 3.4893,
      "step": 191
    },
    {
      "epoch": 0.3655402189433603,
      "grad_norm": 8.497869491577148,
      "learning_rate": 3.0575757355586817e-07,
      "loss": 4.0881,
      "step": 192
    },
    {
      "epoch": 0.36744407425035697,
      "grad_norm": 7.563754081726074,
      "learning_rate": 2.3417567762266497e-07,
      "loss": 3.9095,
      "step": 193
    },
    {
      "epoch": 0.3693479295573536,
      "grad_norm": 8.050283432006836,
      "learning_rate": 1.7209841092460043e-07,
      "loss": 3.9385,
      "step": 194
    },
    {
      "epoch": 0.3712517848643503,
      "grad_norm": 8.874429702758789,
      "learning_rate": 1.1954274476655534e-07,
      "loss": 4.371,
      "step": 195
    },
    {
      "epoch": 0.373155640171347,
      "grad_norm": 9.123085021972656,
      "learning_rate": 7.652304734289127e-08,
      "loss": 4.1202,
      "step": 196
    },
    {
      "epoch": 0.37505949547834366,
      "grad_norm": 9.050190925598145,
      "learning_rate": 4.30510798093342e-08,
      "loss": 4.452,
      "step": 197
    },
    {
      "epoch": 0.3769633507853403,
      "grad_norm": 10.816521644592285,
      "learning_rate": 1.9135993067588284e-08,
      "loss": 4.5865,
      "step": 198
    },
    {
      "epoch": 0.378867206092337,
      "grad_norm": 10.921710968017578,
      "learning_rate": 4.784325263584854e-09,
      "loss": 4.5009,
      "step": 199
    },
    {
      "epoch": 0.38077106139933364,
      "grad_norm": 12.262462615966797,
      "learning_rate": 0.0,
      "loss": 4.617,
      "step": 200
    },
    {
      "epoch": 0.38077106139933364,
      "eval_loss": 0.9820159673690796,
      "eval_runtime": 65.4623,
      "eval_samples_per_second": 13.519,
      "eval_steps_per_second": 3.391,
      "step": 200
    }
  ],
  "logging_steps": 1,
  "max_steps": 200,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 50,
  "stateful_callbacks": {
    "EarlyStoppingCallback": {
      "args": {
        "early_stopping_patience": 4,
        "early_stopping_threshold": 0.0
      },
      "attributes": {
        "early_stopping_patience_counter": 0
      }
    },
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 2.861999239200768e+17,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}