{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.980891719745223,
  "eval_steps": 500,
  "global_step": 156,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.01910828025477707,
      "grad_norm": 7.547975540161133,
      "learning_rate": 6.25e-07,
      "loss": 1.2512,
      "step": 1
    },
    {
      "epoch": 0.03821656050955414,
      "grad_norm": 7.374826431274414,
      "learning_rate": 1.25e-06,
      "loss": 1.2318,
      "step": 2
    },
    {
      "epoch": 0.05732484076433121,
      "grad_norm": 7.294417381286621,
      "learning_rate": 1.8750000000000003e-06,
      "loss": 1.2433,
      "step": 3
    },
    {
      "epoch": 0.07643312101910828,
      "grad_norm": 7.174262523651123,
      "learning_rate": 2.5e-06,
      "loss": 1.2361,
      "step": 4
    },
    {
      "epoch": 0.09554140127388536,
      "grad_norm": 6.600234031677246,
      "learning_rate": 3.125e-06,
      "loss": 1.2129,
      "step": 5
    },
    {
      "epoch": 0.11464968152866242,
      "grad_norm": 4.914210796356201,
      "learning_rate": 3.7500000000000005e-06,
      "loss": 1.1534,
      "step": 6
    },
    {
      "epoch": 0.1337579617834395,
      "grad_norm": 2.7489099502563477,
      "learning_rate": 4.3750000000000005e-06,
      "loss": 1.1355,
      "step": 7
    },
    {
      "epoch": 0.15286624203821655,
      "grad_norm": 2.5778281688690186,
      "learning_rate": 5e-06,
      "loss": 1.0924,
      "step": 8
    },
    {
      "epoch": 0.17197452229299362,
      "grad_norm": 4.604168891906738,
      "learning_rate": 5.625e-06,
      "loss": 1.0943,
      "step": 9
    },
    {
      "epoch": 0.1910828025477707,
      "grad_norm": 4.840404987335205,
      "learning_rate": 6.25e-06,
      "loss": 1.0389,
      "step": 10
    },
    {
      "epoch": 0.21019108280254778,
      "grad_norm": 4.339366436004639,
      "learning_rate": 6.875e-06,
      "loss": 1.0224,
      "step": 11
    },
    {
      "epoch": 0.22929936305732485,
      "grad_norm": 3.31107234954834,
      "learning_rate": 7.500000000000001e-06,
      "loss": 1.0536,
      "step": 12
    },
    {
      "epoch": 0.2484076433121019,
      "grad_norm": 3.4537413120269775,
      "learning_rate": 8.125000000000001e-06,
      "loss": 1.0143,
      "step": 13
    },
    {
      "epoch": 0.267515923566879,
      "grad_norm": 2.662017345428467,
      "learning_rate": 8.750000000000001e-06,
      "loss": 0.9741,
      "step": 14
    },
    {
      "epoch": 0.28662420382165604,
      "grad_norm": 2.157660722732544,
      "learning_rate": 9.375000000000001e-06,
      "loss": 0.9494,
      "step": 15
    },
    {
      "epoch": 0.3057324840764331,
      "grad_norm": 2.1211636066436768,
      "learning_rate": 1e-05,
      "loss": 0.9275,
      "step": 16
    },
    {
      "epoch": 0.3248407643312102,
      "grad_norm": 1.895501732826233,
      "learning_rate": 9.998741174712534e-06,
      "loss": 0.9178,
      "step": 17
    },
    {
      "epoch": 0.34394904458598724,
      "grad_norm": 1.7070763111114502,
      "learning_rate": 9.994965332706574e-06,
      "loss": 0.8882,
      "step": 18
    },
    {
      "epoch": 0.3630573248407643,
      "grad_norm": 1.585816502571106,
      "learning_rate": 9.98867437523228e-06,
      "loss": 0.8857,
      "step": 19
    },
    {
      "epoch": 0.3821656050955414,
      "grad_norm": 1.6084542274475098,
      "learning_rate": 9.979871469976197e-06,
      "loss": 0.9158,
      "step": 20
    },
    {
      "epoch": 0.4012738853503185,
      "grad_norm": 1.507398009300232,
      "learning_rate": 9.968561049466214e-06,
      "loss": 0.9254,
      "step": 21
    },
    {
      "epoch": 0.42038216560509556,
      "grad_norm": 0.9898801445960999,
      "learning_rate": 9.954748808839675e-06,
      "loss": 0.9127,
      "step": 22
    },
    {
      "epoch": 0.4394904458598726,
      "grad_norm": 0.9775479435920715,
      "learning_rate": 9.938441702975689e-06,
      "loss": 0.8819,
      "step": 23
    },
    {
      "epoch": 0.4585987261146497,
      "grad_norm": 1.253936529159546,
      "learning_rate": 9.91964794299315e-06,
      "loss": 0.8983,
      "step": 24
    },
    {
      "epoch": 0.47770700636942676,
      "grad_norm": 1.239895224571228,
      "learning_rate": 9.898376992116179e-06,
      "loss": 0.9112,
      "step": 25
    },
    {
      "epoch": 0.4968152866242038,
      "grad_norm": 0.9475780129432678,
      "learning_rate": 9.874639560909118e-06,
      "loss": 0.8848,
      "step": 26
    },
    {
      "epoch": 0.5159235668789809,
      "grad_norm": 1.0773193836212158,
      "learning_rate": 9.848447601883436e-06,
      "loss": 0.8977,
      "step": 27
    },
    {
      "epoch": 0.535031847133758,
      "grad_norm": 1.1501091718673706,
      "learning_rate": 9.819814303479268e-06,
      "loss": 0.8912,
      "step": 28
    },
    {
      "epoch": 0.554140127388535,
      "grad_norm": 0.9324758648872375,
      "learning_rate": 9.788754083424654e-06,
      "loss": 0.8851,
      "step": 29
    },
    {
      "epoch": 0.5732484076433121,
      "grad_norm": 0.9901764988899231,
      "learning_rate": 9.755282581475769e-06,
      "loss": 0.8716,
      "step": 30
    },
    {
      "epoch": 0.5923566878980892,
      "grad_norm": 1.0067858695983887,
      "learning_rate": 9.719416651541839e-06,
      "loss": 0.8603,
      "step": 31
    },
    {
      "epoch": 0.6114649681528662,
      "grad_norm": 0.8844172358512878,
      "learning_rate": 9.681174353198687e-06,
      "loss": 0.8885,
      "step": 32
    },
    {
      "epoch": 0.6305732484076433,
      "grad_norm": 1.0087534189224243,
      "learning_rate": 9.640574942595195e-06,
      "loss": 0.8426,
      "step": 33
    },
    {
      "epoch": 0.6496815286624203,
      "grad_norm": 0.8578860759735107,
      "learning_rate": 9.597638862757255e-06,
      "loss": 0.8413,
      "step": 34
    },
    {
      "epoch": 0.6687898089171974,
      "grad_norm": 0.8848086595535278,
      "learning_rate": 9.552387733294081e-06,
      "loss": 0.8398,
      "step": 35
    },
    {
      "epoch": 0.6878980891719745,
      "grad_norm": 0.8705900311470032,
      "learning_rate": 9.504844339512096e-06,
      "loss": 0.8738,
      "step": 36
    },
    {
      "epoch": 0.7070063694267515,
      "grad_norm": 0.8821431398391724,
      "learning_rate": 9.45503262094184e-06,
      "loss": 0.8632,
      "step": 37
    },
    {
      "epoch": 0.7261146496815286,
      "grad_norm": 0.82889324426651,
      "learning_rate": 9.40297765928369e-06,
      "loss": 0.832,
      "step": 38
    },
    {
      "epoch": 0.7452229299363057,
      "grad_norm": 0.8025405406951904,
      "learning_rate": 9.348705665778479e-06,
      "loss": 0.8392,
      "step": 39
    },
    {
      "epoch": 0.7643312101910829,
      "grad_norm": 0.8777817487716675,
      "learning_rate": 9.292243968009332e-06,
      "loss": 0.8553,
      "step": 40
    },
    {
      "epoch": 0.7834394904458599,
      "grad_norm": 0.7455444931983948,
      "learning_rate": 9.233620996141421e-06,
      "loss": 0.8169,
      "step": 41
    },
    {
      "epoch": 0.802547770700637,
      "grad_norm": 0.7949671745300293,
      "learning_rate": 9.172866268606514e-06,
      "loss": 0.8268,
      "step": 42
    },
    {
      "epoch": 0.821656050955414,
      "grad_norm": 0.8802870512008667,
      "learning_rate": 9.110010377239552e-06,
      "loss": 0.8662,
      "step": 43
    },
    {
      "epoch": 0.8407643312101911,
      "grad_norm": 0.8528987765312195,
      "learning_rate": 9.045084971874738e-06,
      "loss": 0.8184,
      "step": 44
    },
    {
      "epoch": 0.8598726114649682,
      "grad_norm": 0.8329055309295654,
      "learning_rate": 8.978122744408905e-06,
      "loss": 0.8376,
      "step": 45
    },
    {
      "epoch": 0.8789808917197452,
      "grad_norm": 0.7796112895011902,
      "learning_rate": 8.90915741234015e-06,
      "loss": 0.8462,
      "step": 46
    },
    {
      "epoch": 0.8980891719745223,
      "grad_norm": 0.8821595311164856,
      "learning_rate": 8.838223701790057e-06,
      "loss": 0.8462,
      "step": 47
    },
    {
      "epoch": 0.9171974522292994,
      "grad_norm": 0.8074951171875,
      "learning_rate": 8.765357330018056e-06,
      "loss": 0.8161,
      "step": 48
    },
    {
      "epoch": 0.9363057324840764,
      "grad_norm": 0.8599761724472046,
      "learning_rate": 8.690594987436705e-06,
      "loss": 0.8612,
      "step": 49
    },
    {
      "epoch": 0.9554140127388535,
      "grad_norm": 0.8650784492492676,
      "learning_rate": 8.613974319136959e-06,
      "loss": 0.8474,
      "step": 50
    },
    {
      "epoch": 0.9745222929936306,
      "grad_norm": 0.7465563416481018,
      "learning_rate": 8.535533905932739e-06,
      "loss": 0.8366,
      "step": 51
    },
    {
      "epoch": 0.9936305732484076,
      "grad_norm": 0.8294474482536316,
      "learning_rate": 8.455313244934324e-06,
      "loss": 0.869,
      "step": 52
    },
    {
      "epoch": 1.0127388535031847,
      "grad_norm": 0.8520503044128418,
      "learning_rate": 8.373352729660373e-06,
      "loss": 0.7792,
      "step": 53
    },
    {
      "epoch": 1.0318471337579618,
      "grad_norm": 0.8020623326301575,
      "learning_rate": 8.289693629698564e-06,
      "loss": 0.7468,
      "step": 54
    },
    {
      "epoch": 1.0509554140127388,
      "grad_norm": 0.7566718459129333,
      "learning_rate": 8.204378069925121e-06,
      "loss": 0.7348,
      "step": 55
    },
    {
      "epoch": 1.070063694267516,
      "grad_norm": 0.7379488348960876,
      "learning_rate": 8.117449009293668e-06,
      "loss": 0.7509,
      "step": 56
    },
    {
      "epoch": 1.089171974522293,
      "grad_norm": 0.6847801208496094,
      "learning_rate": 8.0289502192041e-06,
      "loss": 0.6954,
      "step": 57
    },
    {
      "epoch": 1.10828025477707,
      "grad_norm": 0.784027636051178,
      "learning_rate": 7.938926261462366e-06,
      "loss": 0.7201,
      "step": 58
    },
    {
      "epoch": 1.127388535031847,
      "grad_norm": 0.7846508622169495,
      "learning_rate": 7.84742246584226e-06,
      "loss": 0.7429,
      "step": 59
    },
    {
      "epoch": 1.1464968152866242,
      "grad_norm": 0.764062225818634,
      "learning_rate": 7.754484907260513e-06,
      "loss": 0.7031,
      "step": 60
    },
    {
      "epoch": 1.1656050955414012,
      "grad_norm": 0.7919678092002869,
      "learning_rate": 7.660160382576683e-06,
      "loss": 0.7004,
      "step": 61
    },
    {
      "epoch": 1.1847133757961783,
      "grad_norm": 0.7915036678314209,
      "learning_rate": 7.564496387029532e-06,
      "loss": 0.7189,
      "step": 62
    },
    {
      "epoch": 1.2038216560509554,
      "grad_norm": 0.7123846411705017,
      "learning_rate": 7.467541090321735e-06,
      "loss": 0.7075,
      "step": 63
    },
    {
      "epoch": 1.2229299363057324,
      "grad_norm": 0.8034456372261047,
      "learning_rate": 7.369343312364994e-06,
      "loss": 0.7137,
      "step": 64
    },
    {
      "epoch": 1.2420382165605095,
      "grad_norm": 0.7536247968673706,
      "learning_rate": 7.269952498697734e-06,
      "loss": 0.7145,
      "step": 65
    },
    {
      "epoch": 1.2611464968152866,
      "grad_norm": 0.7744494080543518,
      "learning_rate": 7.169418695587791e-06,
      "loss": 0.7113,
      "step": 66
    },
    {
      "epoch": 1.2802547770700636,
      "grad_norm": 0.7711228132247925,
      "learning_rate": 7.067792524832604e-06,
      "loss": 0.7053,
      "step": 67
    },
    {
      "epoch": 1.2993630573248407,
      "grad_norm": 0.8071858882904053,
      "learning_rate": 6.965125158269619e-06,
      "loss": 0.7037,
      "step": 68
    },
    {
      "epoch": 1.3184713375796178,
      "grad_norm": 0.780947744846344,
      "learning_rate": 6.8614682920097265e-06,
      "loss": 0.7267,
      "step": 69
    },
    {
      "epoch": 1.3375796178343948,
      "grad_norm": 0.7653530240058899,
      "learning_rate": 6.7568741204067145e-06,
      "loss": 0.7474,
      "step": 70
    },
    {
      "epoch": 1.356687898089172,
      "grad_norm": 0.7996588945388794,
      "learning_rate": 6.651395309775837e-06,
      "loss": 0.7328,
      "step": 71
    },
    {
      "epoch": 1.3757961783439492,
      "grad_norm": 0.7123798131942749,
      "learning_rate": 6.545084971874738e-06,
      "loss": 0.6917,
      "step": 72
    },
    {
      "epoch": 1.394904458598726,
      "grad_norm": 0.6633033156394958,
      "learning_rate": 6.437996637160086e-06,
      "loss": 0.6991,
      "step": 73
    },
    {
      "epoch": 1.4140127388535033,
      "grad_norm": 0.8307355642318726,
      "learning_rate": 6.330184227833376e-06,
      "loss": 0.6909,
      "step": 74
    },
    {
      "epoch": 1.4331210191082802,
      "grad_norm": 0.6795753240585327,
      "learning_rate": 6.2217020306894705e-06,
      "loss": 0.7098,
      "step": 75
    },
    {
      "epoch": 1.4522292993630574,
      "grad_norm": 0.6153970956802368,
      "learning_rate": 6.112604669781572e-06,
      "loss": 0.6726,
      "step": 76
    },
    {
      "epoch": 1.4713375796178343,
      "grad_norm": 0.7344216108322144,
      "learning_rate": 6.002947078916365e-06,
      "loss": 0.7012,
      "step": 77
    },
    {
      "epoch": 1.4904458598726116,
      "grad_norm": 0.7123773097991943,
      "learning_rate": 5.892784473993184e-06,
      "loss": 0.6638,
      "step": 78
    },
    {
      "epoch": 1.5095541401273884,
      "grad_norm": 0.7022320032119751,
      "learning_rate": 5.782172325201155e-06,
      "loss": 0.6997,
      "step": 79
    },
    {
      "epoch": 1.5286624203821657,
      "grad_norm": 0.7084267139434814,
      "learning_rate": 5.671166329088278e-06,
      "loss": 0.692,
      "step": 80
    },
    {
      "epoch": 1.5477707006369426,
      "grad_norm": 0.68487948179245,
      "learning_rate": 5.559822380516539e-06,
      "loss": 0.6929,
      "step": 81
    },
    {
      "epoch": 1.5668789808917198,
      "grad_norm": 0.7181028723716736,
      "learning_rate": 5.448196544517168e-06,
      "loss": 0.7193,
      "step": 82
    },
    {
      "epoch": 1.5859872611464967,
      "grad_norm": 0.7298808693885803,
      "learning_rate": 5.336345028060199e-06,
      "loss": 0.7338,
      "step": 83
    },
    {
      "epoch": 1.605095541401274,
      "grad_norm": 0.6929397583007812,
      "learning_rate": 5.224324151752575e-06,
      "loss": 0.7025,
      "step": 84
    },
    {
      "epoch": 1.6242038216560508,
      "grad_norm": 0.6811756491661072,
      "learning_rate": 5.112190321479026e-06,
      "loss": 0.7052,
      "step": 85
    },
    {
      "epoch": 1.643312101910828,
      "grad_norm": 0.6826193332672119,
      "learning_rate": 5e-06,
      "loss": 0.6931,
      "step": 86
    },
    {
      "epoch": 1.662420382165605,
      "grad_norm": 0.654704749584198,
      "learning_rate": 4.887809678520976e-06,
      "loss": 0.6833,
      "step": 87
    },
    {
      "epoch": 1.6815286624203822,
      "grad_norm": 0.6750189065933228,
      "learning_rate": 4.775675848247427e-06,
      "loss": 0.7073,
      "step": 88
    },
    {
      "epoch": 1.700636942675159,
      "grad_norm": 0.6871541142463684,
      "learning_rate": 4.663654971939802e-06,
      "loss": 0.7184,
      "step": 89
    },
    {
      "epoch": 1.7197452229299364,
      "grad_norm": 0.6716470122337341,
      "learning_rate": 4.551803455482833e-06,
      "loss": 0.7035,
      "step": 90
    },
    {
      "epoch": 1.7388535031847132,
      "grad_norm": 0.6717052459716797,
      "learning_rate": 4.4401776194834615e-06,
      "loss": 0.6988,
      "step": 91
    },
    {
      "epoch": 1.7579617834394905,
      "grad_norm": 0.7127666473388672,
      "learning_rate": 4.3288336709117246e-06,
      "loss": 0.7121,
      "step": 92
    },
    {
      "epoch": 1.7770700636942676,
      "grad_norm": 0.6773775219917297,
      "learning_rate": 4.217827674798845e-06,
      "loss": 0.689,
      "step": 93
    },
    {
      "epoch": 1.7961783439490446,
      "grad_norm": 0.6683153510093689,
      "learning_rate": 4.107215526006818e-06,
      "loss": 0.6857,
      "step": 94
    },
    {
      "epoch": 1.8152866242038217,
      "grad_norm": 0.6821578741073608,
      "learning_rate": 3.997052921083637e-06,
      "loss": 0.6992,
      "step": 95
    },
    {
      "epoch": 1.8343949044585988,
      "grad_norm": 0.6637680530548096,
      "learning_rate": 3.887395330218429e-06,
      "loss": 0.6848,
      "step": 96
    },
    {
      "epoch": 1.8535031847133758,
      "grad_norm": 0.6395952105522156,
      "learning_rate": 3.778297969310529e-06,
      "loss": 0.6604,
      "step": 97
    },
    {
      "epoch": 1.872611464968153,
      "grad_norm": 0.6628071069717407,
      "learning_rate": 3.669815772166625e-06,
      "loss": 0.6961,
      "step": 98
    },
    {
      "epoch": 1.89171974522293,
      "grad_norm": 0.6521620154380798,
      "learning_rate": 3.562003362839914e-06,
      "loss": 0.71,
      "step": 99
    },
    {
      "epoch": 1.910828025477707,
      "grad_norm": 0.6818088889122009,
      "learning_rate": 3.4549150281252635e-06,
      "loss": 0.6946,
      "step": 100
    },
    {
      "epoch": 1.929936305732484,
      "grad_norm": 0.6266446113586426,
      "learning_rate": 3.3486046902241663e-06,
      "loss": 0.6786,
      "step": 101
    },
    {
      "epoch": 1.9490445859872612,
      "grad_norm": 0.6931262612342834,
      "learning_rate": 3.2431258795932863e-06,
      "loss": 0.699,
      "step": 102
    },
    {
      "epoch": 1.9681528662420382,
      "grad_norm": 0.644861102104187,
      "learning_rate": 3.1385317079902743e-06,
      "loss": 0.7014,
      "step": 103
    },
    {
      "epoch": 1.9872611464968153,
      "grad_norm": 0.67961186170578,
      "learning_rate": 3.0348748417303826e-06,
      "loss": 0.6903,
      "step": 104
    },
    {
      "epoch": 2.0063694267515926,
      "grad_norm": 0.7411230802536011,
      "learning_rate": 2.932207475167398e-06,
      "loss": 0.6627,
      "step": 105
    },
    {
      "epoch": 2.0254777070063694,
      "grad_norm": 0.7194402813911438,
      "learning_rate": 2.83058130441221e-06,
      "loss": 0.6365,
      "step": 106
    },
    {
      "epoch": 2.0445859872611467,
      "grad_norm": 0.6588802933692932,
      "learning_rate": 2.7300475013022666e-06,
      "loss": 0.6299,
      "step": 107
    },
    {
      "epoch": 2.0636942675159236,
      "grad_norm": 0.6904900074005127,
      "learning_rate": 2.6306566876350072e-06,
      "loss": 0.6289,
      "step": 108
    },
    {
      "epoch": 2.082802547770701,
      "grad_norm": 0.6685463786125183,
      "learning_rate": 2.532458909678266e-06,
      "loss": 0.6208,
      "step": 109
    },
    {
      "epoch": 2.1019108280254777,
      "grad_norm": 0.6020084023475647,
      "learning_rate": 2.43550361297047e-06,
      "loss": 0.6027,
      "step": 110
    },
    {
      "epoch": 2.121019108280255,
      "grad_norm": 0.6874512434005737,
      "learning_rate": 2.339839617423318e-06,
      "loss": 0.6021,
      "step": 111
    },
    {
      "epoch": 2.140127388535032,
      "grad_norm": 0.6943764090538025,
      "learning_rate": 2.245515092739488e-06,
      "loss": 0.6031,
      "step": 112
    },
    {
      "epoch": 2.159235668789809,
      "grad_norm": 0.7539130449295044,
      "learning_rate": 2.1525775341577404e-06,
      "loss": 0.6343,
      "step": 113
    },
    {
      "epoch": 2.178343949044586,
      "grad_norm": 0.64998459815979,
      "learning_rate": 2.061073738537635e-06,
      "loss": 0.574,
      "step": 114
    },
    {
      "epoch": 2.1974522292993632,
      "grad_norm": 0.6652138829231262,
      "learning_rate": 1.971049780795901e-06,
      "loss": 0.6022,
      "step": 115
    },
    {
      "epoch": 2.21656050955414,
      "grad_norm": 0.6873827576637268,
      "learning_rate": 1.8825509907063328e-06,
      "loss": 0.5928,
      "step": 116
    },
    {
      "epoch": 2.2356687898089174,
      "grad_norm": 0.6433504819869995,
      "learning_rate": 1.7956219300748796e-06,
      "loss": 0.607,
      "step": 117
    },
    {
      "epoch": 2.254777070063694,
      "grad_norm": 0.620965301990509,
      "learning_rate": 1.7103063703014372e-06,
      "loss": 0.606,
      "step": 118
    },
    {
      "epoch": 2.2738853503184715,
      "grad_norm": 0.6539048552513123,
      "learning_rate": 1.6266472703396286e-06,
      "loss": 0.6023,
      "step": 119
    },
    {
      "epoch": 2.2929936305732483,
      "grad_norm": 0.6151559948921204,
      "learning_rate": 1.544686755065677e-06,
      "loss": 0.6034,
      "step": 120
    },
    {
      "epoch": 2.3121019108280256,
      "grad_norm": 0.636107325553894,
      "learning_rate": 1.4644660940672628e-06,
      "loss": 0.5776,
      "step": 121
    },
    {
      "epoch": 2.3312101910828025,
      "grad_norm": 0.6516319513320923,
      "learning_rate": 1.3860256808630429e-06,
      "loss": 0.5965,
      "step": 122
    },
    {
      "epoch": 2.3503184713375798,
      "grad_norm": 0.6334967613220215,
      "learning_rate": 1.3094050125632973e-06,
      "loss": 0.6159,
      "step": 123
    },
    {
      "epoch": 2.3694267515923566,
      "grad_norm": 0.6599099636077881,
      "learning_rate": 1.234642669981946e-06,
      "loss": 0.6122,
      "step": 124
    },
    {
      "epoch": 2.388535031847134,
      "grad_norm": 0.6333508491516113,
      "learning_rate": 1.1617762982099446e-06,
      "loss": 0.5929,
      "step": 125
    },
    {
      "epoch": 2.4076433121019107,
      "grad_norm": 0.6153550148010254,
      "learning_rate": 1.0908425876598512e-06,
      "loss": 0.5706,
      "step": 126
    },
    {
      "epoch": 2.426751592356688,
      "grad_norm": 0.6106190085411072,
      "learning_rate": 1.0218772555910955e-06,
      "loss": 0.6076,
      "step": 127
    },
    {
      "epoch": 2.445859872611465,
      "grad_norm": 0.6668431162834167,
      "learning_rate": 9.549150281252633e-07,
      "loss": 0.6054,
      "step": 128
    },
    {
      "epoch": 2.464968152866242,
      "grad_norm": 0.6144554615020752,
      "learning_rate": 8.899896227604509e-07,
      "loss": 0.5989,
      "step": 129
    },
    {
      "epoch": 2.484076433121019,
      "grad_norm": 0.621407151222229,
      "learning_rate": 8.271337313934869e-07,
      "loss": 0.5972,
      "step": 130
    },
    {
      "epoch": 2.5031847133757963,
      "grad_norm": 0.6157580614089966,
      "learning_rate": 7.663790038585794e-07,
      "loss": 0.5954,
      "step": 131
    },
    {
      "epoch": 2.522292993630573,
      "grad_norm": 0.6083014607429504,
      "learning_rate": 7.077560319906696e-07,
      "loss": 0.6132,
      "step": 132
    },
    {
      "epoch": 2.5414012738853504,
      "grad_norm": 0.5859634876251221,
      "learning_rate": 6.512943342215234e-07,
      "loss": 0.6082,
      "step": 133
    },
    {
      "epoch": 2.5605095541401273,
      "grad_norm": 0.5937418341636658,
      "learning_rate": 5.9702234071631e-07,
      "loss": 0.5908,
      "step": 134
    },
    {
      "epoch": 2.5796178343949046,
      "grad_norm": 0.604749321937561,
      "learning_rate": 5.449673790581611e-07,
      "loss": 0.6164,
      "step": 135
    },
    {
      "epoch": 2.5987261146496814,
      "grad_norm": 0.599074125289917,
      "learning_rate": 4.951556604879049e-07,
      "loss": 0.6123,
      "step": 136
    },
    {
      "epoch": 2.6178343949044587,
      "grad_norm": 0.6109683513641357,
      "learning_rate": 4.4761226670592074e-07,
      "loss": 0.601,
      "step": 137
    },
    {
      "epoch": 2.6369426751592355,
      "grad_norm": 0.6154210567474365,
      "learning_rate": 4.0236113724274716e-07,
      "loss": 0.6018,
      "step": 138
    },
    {
      "epoch": 2.656050955414013,
      "grad_norm": 0.6092813611030579,
      "learning_rate": 3.5942505740480583e-07,
      "loss": 0.6081,
      "step": 139
    },
    {
      "epoch": 2.6751592356687897,
      "grad_norm": 0.6313538551330566,
      "learning_rate": 3.18825646801314e-07,
      "loss": 0.5801,
      "step": 140
    },
    {
      "epoch": 2.694267515923567,
      "grad_norm": 0.5989095568656921,
      "learning_rate": 2.8058334845816214e-07,
      "loss": 0.581,
      "step": 141
    },
    {
      "epoch": 2.713375796178344,
      "grad_norm": 0.5897513628005981,
      "learning_rate": 2.447174185242324e-07,
      "loss": 0.5949,
      "step": 142
    },
    {
      "epoch": 2.732484076433121,
      "grad_norm": 0.6216192245483398,
      "learning_rate": 2.1124591657534776e-07,
      "loss": 0.5946,
      "step": 143
    },
    {
      "epoch": 2.7515923566878984,
      "grad_norm": 0.6057581901550293,
      "learning_rate": 1.801856965207338e-07,
      "loss": 0.6176,
      "step": 144
    },
    {
      "epoch": 2.770700636942675,
      "grad_norm": 0.5893821120262146,
      "learning_rate": 1.5155239811656562e-07,
      "loss": 0.5866,
      "step": 145
    },
    {
      "epoch": 2.789808917197452,
      "grad_norm": 0.5728128552436829,
      "learning_rate": 1.253604390908819e-07,
      "loss": 0.5793,
      "step": 146
    },
    {
      "epoch": 2.8089171974522293,
      "grad_norm": 0.5864872932434082,
      "learning_rate": 1.0162300788382263e-07,
      "loss": 0.5808,
      "step": 147
    },
    {
      "epoch": 2.8280254777070066,
      "grad_norm": 0.5862218737602234,
      "learning_rate": 8.035205700685167e-08,
      "loss": 0.586,
      "step": 148
    },
    {
      "epoch": 2.8471337579617835,
      "grad_norm": 0.6472737789154053,
      "learning_rate": 6.15582970243117e-08,
      "loss": 0.5998,
      "step": 149
    },
    {
      "epoch": 2.8662420382165603,
      "grad_norm": 0.576315701007843,
      "learning_rate": 4.52511911603265e-08,
      "loss": 0.5839,
      "step": 150
    },
    {
      "epoch": 2.8853503184713376,
      "grad_norm": 0.5896339416503906,
      "learning_rate": 3.143895053378698e-08,
      "loss": 0.6035,
      "step": 151
    },
    {
      "epoch": 2.904458598726115,
      "grad_norm": 0.6027177572250366,
      "learning_rate": 2.012853002380466e-08,
      "loss": 0.6129,
      "step": 152
    },
    {
      "epoch": 2.9235668789808917,
      "grad_norm": 0.5746337175369263,
      "learning_rate": 1.132562476771959e-08,
      "loss": 0.5796,
      "step": 153
    },
    {
      "epoch": 2.9426751592356686,
      "grad_norm": 0.5898021459579468,
      "learning_rate": 5.034667293427053e-09,
      "loss": 0.574,
      "step": 154
    },
    {
      "epoch": 2.961783439490446,
      "grad_norm": 0.6063404083251953,
      "learning_rate": 1.2588252874673469e-09,
      "loss": 0.5974,
      "step": 155
    },
    {
      "epoch": 2.980891719745223,
      "grad_norm": 0.6394419074058533,
      "learning_rate": 0.0,
      "loss": 0.6357,
      "step": 156
    },
    {
      "epoch": 2.980891719745223,
      "step": 156,
      "total_flos": 9.25222495000658e+16,
      "train_loss": 0.7492156773805618,
      "train_runtime": 2914.5015,
      "train_samples_per_second": 5.147,
      "train_steps_per_second": 0.054
    }
  ],
  "logging_steps": 1,
  "max_steps": 156,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 9.25222495000658e+16,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}