{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 1000,
  "global_step": 19181,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.02606746259319118,
      "grad_norm": 1.3942902088165283,
      "learning_rate": 4.869662687034044e-05,
      "loss": 4.3834,
      "step": 500
    },
    {
      "epoch": 0.05213492518638236,
      "grad_norm": 1.8239964246749878,
      "learning_rate": 4.7393253740680883e-05,
      "loss": 2.9072,
      "step": 1000
    },
    {
      "epoch": 0.05213492518638236,
      "eval_accuracy": 0.45563744280004975,
      "eval_loss": 2.416715145111084,
      "eval_runtime": 54.1547,
      "eval_samples_per_second": 114.007,
      "eval_steps_per_second": 3.564,
      "step": 1000
    },
    {
      "epoch": 0.07820238777957354,
      "grad_norm": 1.627564787864685,
      "learning_rate": 4.6089880611021324e-05,
      "loss": 2.2603,
      "step": 1500
    },
    {
      "epoch": 0.10426985037276472,
      "grad_norm": 1.6643654108047485,
      "learning_rate": 4.4786507481361764e-05,
      "loss": 1.9548,
      "step": 2000
    },
    {
      "epoch": 0.10426985037276472,
      "eval_accuracy": 0.573437753819584,
      "eval_loss": 1.7778319120407104,
      "eval_runtime": 53.9946,
      "eval_samples_per_second": 114.345,
      "eval_steps_per_second": 3.574,
      "step": 2000
    },
    {
      "epoch": 0.13033731296595588,
      "grad_norm": 1.399413824081421,
      "learning_rate": 4.348313435170221e-05,
      "loss": 1.8037,
      "step": 2500
    },
    {
      "epoch": 0.15640477555914709,
      "grad_norm": 1.3979791402816772,
      "learning_rate": 4.2179761222042645e-05,
      "loss": 1.707,
      "step": 3000
    },
    {
      "epoch": 0.15640477555914709,
      "eval_accuracy": 0.6059451849445266,
      "eval_loss": 1.5955791473388672,
      "eval_runtime": 53.7411,
      "eval_samples_per_second": 114.884,
      "eval_steps_per_second": 3.591,
      "step": 3000
    },
    {
      "epoch": 0.18247223815233826,
      "grad_norm": 1.3367459774017334,
      "learning_rate": 4.0876388092383086e-05,
      "loss": 1.6436,
      "step": 3500
    },
    {
      "epoch": 0.20853970074552944,
      "grad_norm": 1.2440216541290283,
      "learning_rate": 3.957301496272353e-05,
      "loss": 1.589,
      "step": 4000
    },
    {
      "epoch": 0.20853970074552944,
      "eval_accuracy": 0.6241351728514335,
      "eval_loss": 1.494871973991394,
      "eval_runtime": 53.6037,
      "eval_samples_per_second": 115.179,
      "eval_steps_per_second": 3.6,
      "step": 4000
    },
    {
      "epoch": 0.23460716333872061,
      "grad_norm": 1.3261712789535522,
      "learning_rate": 3.8269641833063974e-05,
      "loss": 1.5521,
      "step": 4500
    },
    {
      "epoch": 0.26067462593191176,
      "grad_norm": 1.1943484544754028,
      "learning_rate": 3.696626870340441e-05,
      "loss": 1.5186,
      "step": 5000
    },
    {
      "epoch": 0.26067462593191176,
      "eval_accuracy": 0.6353352959672907,
      "eval_loss": 1.4332515001296997,
      "eval_runtime": 54.1155,
      "eval_samples_per_second": 114.089,
      "eval_steps_per_second": 3.566,
      "step": 5000
    },
    {
      "epoch": 0.28674208852510297,
      "grad_norm": 1.2761880159378052,
      "learning_rate": 3.5662895573744855e-05,
      "loss": 1.4904,
      "step": 5500
    },
    {
      "epoch": 0.31280955111829417,
      "grad_norm": 1.2012956142425537,
      "learning_rate": 3.4359522444085295e-05,
      "loss": 1.463,
      "step": 6000
    },
    {
      "epoch": 0.31280955111829417,
      "eval_accuracy": 0.6444195236163637,
      "eval_loss": 1.3857548236846924,
      "eval_runtime": 53.8321,
      "eval_samples_per_second": 114.69,
      "eval_steps_per_second": 3.585,
      "step": 6000
    },
    {
      "epoch": 0.3388770137114853,
      "grad_norm": 1.0922613143920898,
      "learning_rate": 3.3056149314425736e-05,
      "loss": 1.4449,
      "step": 6500
    },
    {
      "epoch": 0.3649444763046765,
      "grad_norm": 1.140118956565857,
      "learning_rate": 3.1752776184766176e-05,
      "loss": 1.4279,
      "step": 7000
    },
    {
      "epoch": 0.3649444763046765,
      "eval_accuracy": 0.6507737014649457,
      "eval_loss": 1.3520748615264893,
      "eval_runtime": 53.6544,
      "eval_samples_per_second": 115.07,
      "eval_steps_per_second": 3.597,
      "step": 7000
    },
    {
      "epoch": 0.3910119388978677,
      "grad_norm": 1.226009726524353,
      "learning_rate": 3.0449403055106617e-05,
      "loss": 1.4112,
      "step": 7500
    },
    {
      "epoch": 0.4170794014910589,
      "grad_norm": 1.1891686916351318,
      "learning_rate": 2.914602992544706e-05,
      "loss": 1.3932,
      "step": 8000
    },
    {
      "epoch": 0.4170794014910589,
      "eval_accuracy": 0.6563536553661636,
      "eval_loss": 1.3245205879211426,
      "eval_runtime": 53.9661,
      "eval_samples_per_second": 114.405,
      "eval_steps_per_second": 3.576,
      "step": 8000
    },
    {
      "epoch": 0.44314686408425,
      "grad_norm": 1.1560375690460205,
      "learning_rate": 2.7842656795787498e-05,
      "loss": 1.3831,
      "step": 8500
    },
    {
      "epoch": 0.46921432667744123,
      "grad_norm": 1.2043641805648804,
      "learning_rate": 2.653928366612794e-05,
      "loss": 1.3687,
      "step": 9000
    },
    {
      "epoch": 0.46921432667744123,
      "eval_accuracy": 0.6603786699244237,
      "eval_loss": 1.3019030094146729,
      "eval_runtime": 53.6427,
      "eval_samples_per_second": 115.095,
      "eval_steps_per_second": 3.598,
      "step": 9000
    },
    {
      "epoch": 0.4952817892706324,
      "grad_norm": 1.1493725776672363,
      "learning_rate": 2.5235910536468382e-05,
      "loss": 1.3597,
      "step": 9500
    },
    {
      "epoch": 0.5213492518638235,
      "grad_norm": 1.1763168573379517,
      "learning_rate": 2.3932537406808823e-05,
      "loss": 1.3498,
      "step": 10000
    },
    {
      "epoch": 0.5213492518638235,
      "eval_accuracy": 0.6643971930344544,
      "eval_loss": 1.283115029335022,
      "eval_runtime": 53.7235,
      "eval_samples_per_second": 114.922,
      "eval_steps_per_second": 3.592,
      "step": 10000
    },
    {
      "epoch": 0.5474167144570148,
      "grad_norm": 1.0891131162643433,
      "learning_rate": 2.2629164277149263e-05,
      "loss": 1.3402,
      "step": 10500
    },
    {
      "epoch": 0.5734841770502059,
      "grad_norm": 1.1026585102081299,
      "learning_rate": 2.1325791147489704e-05,
      "loss": 1.3315,
      "step": 11000
    },
    {
      "epoch": 0.5734841770502059,
      "eval_accuracy": 0.6677052983833761,
      "eval_loss": 1.2660980224609375,
      "eval_runtime": 54.2389,
      "eval_samples_per_second": 113.83,
      "eval_steps_per_second": 3.558,
      "step": 11000
    },
    {
      "epoch": 0.5995516396433971,
      "grad_norm": 1.1274486780166626,
      "learning_rate": 2.0022418017830144e-05,
      "loss": 1.3221,
      "step": 11500
    },
    {
      "epoch": 0.6256191022365883,
      "grad_norm": 1.1317154169082642,
      "learning_rate": 1.8719044888170585e-05,
      "loss": 1.3141,
      "step": 12000
    },
    {
      "epoch": 0.6256191022365883,
      "eval_accuracy": 0.6706000093096868,
      "eval_loss": 1.2517062425613403,
      "eval_runtime": 54.6459,
      "eval_samples_per_second": 112.982,
      "eval_steps_per_second": 3.532,
      "step": 12000
    },
    {
      "epoch": 0.6516865648297795,
      "grad_norm": 1.124345064163208,
      "learning_rate": 1.7415671758511025e-05,
      "loss": 1.3056,
      "step": 12500
    },
    {
      "epoch": 0.6777540274229706,
      "grad_norm": 1.0935009717941284,
      "learning_rate": 1.611229862885147e-05,
      "loss": 1.3006,
      "step": 13000
    },
    {
      "epoch": 0.6777540274229706,
      "eval_accuracy": 0.6730620097967037,
      "eval_loss": 1.2389717102050781,
      "eval_runtime": 53.9277,
      "eval_samples_per_second": 114.487,
      "eval_steps_per_second": 3.579,
      "step": 13000
    },
    {
      "epoch": 0.7038214900161618,
      "grad_norm": 1.1120526790618896,
      "learning_rate": 1.4808925499191908e-05,
      "loss": 1.2937,
      "step": 13500
    },
    {
      "epoch": 0.729888952609353,
      "grad_norm": 1.1141753196716309,
      "learning_rate": 1.350555236953235e-05,
      "loss": 1.2872,
      "step": 14000
    },
    {
      "epoch": 0.729888952609353,
      "eval_accuracy": 0.675354599317733,
      "eval_loss": 1.227911353111267,
      "eval_runtime": 54.1923,
      "eval_samples_per_second": 113.928,
      "eval_steps_per_second": 3.561,
      "step": 14000
    },
    {
      "epoch": 0.7559564152025442,
      "grad_norm": 1.0864200592041016,
      "learning_rate": 1.220217923987279e-05,
      "loss": 1.284,
      "step": 14500
    },
    {
      "epoch": 0.7820238777957353,
      "grad_norm": 1.085408329963684,
      "learning_rate": 1.0898806110213231e-05,
      "loss": 1.2757,
      "step": 15000
    },
    {
      "epoch": 0.7820238777957353,
      "eval_accuracy": 0.6772790762257517,
      "eval_loss": 1.2187167406082153,
      "eval_runtime": 53.8812,
      "eval_samples_per_second": 114.585,
      "eval_steps_per_second": 3.582,
      "step": 15000
    },
    {
      "epoch": 0.8080913403889265,
      "grad_norm": 1.178449034690857,
      "learning_rate": 9.595432980553674e-06,
      "loss": 1.273,
      "step": 15500
    },
    {
      "epoch": 0.8341588029821178,
      "grad_norm": 1.114288330078125,
      "learning_rate": 8.292059850894114e-06,
      "loss": 1.2688,
      "step": 16000
    },
    {
      "epoch": 0.8341588029821178,
      "eval_accuracy": 0.6788172328001163,
      "eval_loss": 1.211661458015442,
      "eval_runtime": 54.4395,
      "eval_samples_per_second": 113.41,
      "eval_steps_per_second": 3.545,
      "step": 16000
    },
    {
      "epoch": 0.8602262655753089,
      "grad_norm": 1.1161280870437622,
      "learning_rate": 6.988686721234555e-06,
      "loss": 1.2609,
      "step": 16500
    },
    {
      "epoch": 0.8862937281685,
      "grad_norm": 1.1159133911132812,
      "learning_rate": 5.685313591574997e-06,
      "loss": 1.2609,
      "step": 17000
    },
    {
      "epoch": 0.8862937281685,
      "eval_accuracy": 0.6802594426030898,
      "eval_loss": 1.204089879989624,
      "eval_runtime": 53.5113,
      "eval_samples_per_second": 115.378,
      "eval_steps_per_second": 3.607,
      "step": 17000
    },
    {
      "epoch": 0.9123611907616913,
      "grad_norm": 1.0988874435424805,
      "learning_rate": 4.381940461915437e-06,
      "loss": 1.2575,
      "step": 17500
    },
    {
      "epoch": 0.9384286533548825,
      "grad_norm": 1.1227891445159912,
      "learning_rate": 3.078567332255878e-06,
      "loss": 1.2572,
      "step": 18000
    },
    {
      "epoch": 0.9384286533548825,
      "eval_accuracy": 0.6816663452608153,
      "eval_loss": 1.1983492374420166,
      "eval_runtime": 52.9054,
      "eval_samples_per_second": 116.699,
      "eval_steps_per_second": 3.648,
      "step": 18000
    },
    {
      "epoch": 0.9644961159480736,
      "grad_norm": 1.1137040853500366,
      "learning_rate": 1.7751942025963193e-06,
      "loss": 1.2525,
      "step": 18500
    },
    {
      "epoch": 0.9905635785412648,
      "grad_norm": 1.1135761737823486,
      "learning_rate": 4.7182107293676034e-07,
      "loss": 1.2517,
      "step": 19000
    },
    {
      "epoch": 0.9905635785412648,
      "eval_accuracy": 0.6824530771206215,
      "eval_loss": 1.1944977045059204,
      "eval_runtime": 52.9417,
      "eval_samples_per_second": 116.619,
      "eval_steps_per_second": 3.646,
      "step": 19000
    },
    {
      "epoch": 1.0,
      "step": 19181,
      "total_flos": 3.20748511297536e+17,
      "train_loss": 1.5379937968794295,
      "train_runtime": 7483.9627,
      "train_samples_per_second": 82.012,
      "train_steps_per_second": 2.563
    }
  ],
  "logging_steps": 500,
  "max_steps": 19181,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 1000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 3.20748511297536e+17,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": null
}