{
  "best_global_step": 1980,
  "best_metric": 0.7248677248677249,
  "best_model_checkpoint": "sentiment-model/model_v2\\checkpoint-1980",
  "epoch": 4.0,
  "eval_steps": 250,
  "global_step": 1980,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.10114632501685772,
      "grad_norm": 1.0693855285644531,
      "learning_rate": 0.02,
      "loss": 0.9859145355224609,
      "step": 50
    },
    {
      "epoch": 0.20229265003371544,
      "grad_norm": 0.6242259740829468,
      "learning_rate": 0.02,
      "loss": 0.8014328765869141,
      "step": 100
    },
    {
      "epoch": 0.30343897505057316,
      "grad_norm": 0.5784542560577393,
      "learning_rate": 0.02,
      "loss": 0.7533317565917969,
      "step": 150
    },
    {
      "epoch": 0.4045853000674309,
      "grad_norm": 0.44530612230300903,
      "learning_rate": 0.02,
      "loss": 0.7336380767822266,
      "step": 200
    },
    {
      "epoch": 0.5057316250842886,
      "grad_norm": 0.4625791013240814,
      "learning_rate": 0.02,
      "loss": 0.7139082336425782,
      "step": 250
    },
    {
      "epoch": 0.5057316250842886,
      "eval_accuracy": 0.671957671957672,
      "eval_f1_macro": 0.6568898505429338,
      "eval_f1_weighted": 0.6704089764553762,
      "eval_loss": 0.7160658836364746,
      "eval_runtime": 0.5115,
      "eval_samples_per_second": 369.534,
      "eval_steps_per_second": 3.91,
      "step": 250
    },
    {
      "epoch": 0.6068779501011463,
      "grad_norm": 0.4543777406215668,
      "learning_rate": 0.02,
      "loss": 0.707908935546875,
      "step": 300
    },
    {
      "epoch": 0.708024275118004,
      "grad_norm": 0.3702331781387329,
      "learning_rate": 0.02,
      "loss": 0.6974618530273438,
      "step": 350
    },
    {
      "epoch": 0.8091706001348618,
      "grad_norm": 0.359216570854187,
      "learning_rate": 0.02,
      "loss": 0.6783417510986328,
      "step": 400
    },
    {
      "epoch": 0.9103169251517195,
      "grad_norm": 0.5427055358886719,
      "learning_rate": 0.02,
      "loss": 0.6694049072265625,
      "step": 450
    },
    {
      "epoch": 1.0101146325016859,
      "grad_norm": 0.4603957235813141,
      "learning_rate": 0.02,
      "loss": 0.6679498291015625,
      "step": 500
    },
    {
      "epoch": 1.0101146325016859,
      "eval_accuracy": 0.6137566137566137,
      "eval_f1_macro": 0.6017957351290685,
      "eval_f1_weighted": 0.6163611422870683,
      "eval_loss": 0.7431102991104126,
      "eval_runtime": 0.4311,
      "eval_samples_per_second": 438.451,
      "eval_steps_per_second": 4.64,
      "step": 500
    },
    {
      "epoch": 1.1112609575185435,
      "grad_norm": 0.41513311862945557,
      "learning_rate": 0.02,
      "loss": 0.6447215270996094,
      "step": 550
    },
    {
      "epoch": 1.2124072825354013,
      "grad_norm": 0.5227681398391724,
      "learning_rate": 0.02,
      "loss": 0.6510934448242187,
      "step": 600
    },
    {
      "epoch": 1.313553607552259,
      "grad_norm": 0.45083895325660706,
      "learning_rate": 0.02,
      "loss": 0.6367919540405274,
      "step": 650
    },
    {
      "epoch": 1.4146999325691167,
      "grad_norm": 0.38931724429130554,
      "learning_rate": 0.02,
      "loss": 0.6353715133666992,
      "step": 700
    },
    {
      "epoch": 1.5158462575859744,
      "grad_norm": 0.5749202966690063,
      "learning_rate": 0.02,
      "loss": 0.6385128402709961,
      "step": 750
    },
    {
      "epoch": 1.5158462575859744,
      "eval_accuracy": 0.6825396825396826,
      "eval_f1_macro": 0.6571172796152046,
      "eval_f1_weighted": 0.6804525824646179,
      "eval_loss": 0.6635375618934631,
      "eval_runtime": 0.5071,
      "eval_samples_per_second": 372.683,
      "eval_steps_per_second": 3.944,
      "step": 750
    },
    {
      "epoch": 1.616992582602832,
      "grad_norm": 0.45581257343292236,
      "learning_rate": 0.02,
      "loss": 0.625518913269043,
      "step": 800
    },
    {
      "epoch": 1.7181389076196898,
      "grad_norm": 0.3377571702003479,
      "learning_rate": 0.02,
      "loss": 0.6252500915527344,
      "step": 850
    },
    {
      "epoch": 1.8192852326365476,
      "grad_norm": 0.5641970038414001,
      "learning_rate": 0.02,
      "loss": 0.6137918853759765,
      "step": 900
    },
    {
      "epoch": 1.9204315576534052,
      "grad_norm": 0.38811010122299194,
      "learning_rate": 0.02,
      "loss": 0.6167445373535156,
      "step": 950
    },
    {
      "epoch": 2.0202292650033717,
      "grad_norm": 0.4042782783508301,
      "learning_rate": 0.02,
      "loss": 0.6058655548095703,
      "step": 1000
    },
    {
      "epoch": 2.0202292650033717,
      "eval_accuracy": 0.6666666666666666,
      "eval_f1_macro": 0.6448093451550602,
      "eval_f1_weighted": 0.667169150765329,
      "eval_loss": 0.6820808053016663,
      "eval_runtime": 0.4736,
      "eval_samples_per_second": 399.042,
      "eval_steps_per_second": 4.223,
      "step": 1000
    },
    {
      "epoch": 2.121375590020229,
      "grad_norm": 0.6726216077804565,
      "learning_rate": 0.02,
      "loss": 0.5750845718383789,
      "step": 1050
    },
    {
      "epoch": 2.222521915037087,
      "grad_norm": 0.43511903285980225,
      "learning_rate": 0.02,
      "loss": 0.5747246170043945,
      "step": 1100
    },
    {
      "epoch": 2.323668240053945,
      "grad_norm": 0.4764536917209625,
      "learning_rate": 0.02,
      "loss": 0.5774239730834961,
      "step": 1150
    },
    {
      "epoch": 2.4248145650708026,
      "grad_norm": 0.3557627499103546,
      "learning_rate": 0.02,
      "loss": 0.5817196655273438,
      "step": 1200
    },
    {
      "epoch": 2.52596089008766,
      "grad_norm": 0.4189053773880005,
      "learning_rate": 0.02,
      "loss": 0.58625732421875,
      "step": 1250
    },
    {
      "epoch": 2.52596089008766,
      "eval_accuracy": 0.6666666666666666,
      "eval_f1_macro": 0.642448966034101,
      "eval_f1_weighted": 0.6673650887236832,
      "eval_loss": 0.6505153179168701,
      "eval_runtime": 0.5733,
      "eval_samples_per_second": 329.696,
      "eval_steps_per_second": 3.489,
      "step": 1250
    },
    {
      "epoch": 2.627107215104518,
      "grad_norm": 0.43322718143463135,
      "learning_rate": 0.02,
      "loss": 0.5713236999511718,
      "step": 1300
    },
    {
      "epoch": 2.7282535401213757,
      "grad_norm": 0.539417564868927,
      "learning_rate": 0.02,
      "loss": 0.5782129287719726,
      "step": 1350
    },
    {
      "epoch": 2.8293998651382335,
      "grad_norm": 0.5032817721366882,
      "learning_rate": 0.02,
      "loss": 0.5660055160522461,
      "step": 1400
    },
    {
      "epoch": 2.930546190155091,
      "grad_norm": 0.5244382619857788,
      "learning_rate": 0.02,
      "loss": 0.5648572158813476,
      "step": 1450
    },
    {
      "epoch": 3.030343897505057,
      "grad_norm": 0.443816602230072,
      "learning_rate": 0.02,
      "loss": 0.5666342926025391,
      "step": 1500
    },
    {
      "epoch": 3.030343897505057,
      "eval_accuracy": 0.7142857142857143,
      "eval_f1_macro": 0.701973951973952,
      "eval_f1_weighted": 0.7131567726805821,
      "eval_loss": 0.6522725224494934,
      "eval_runtime": 0.4984,
      "eval_samples_per_second": 379.246,
      "eval_steps_per_second": 4.013,
      "step": 1500
    },
    {
      "epoch": 3.131490222521915,
      "grad_norm": 0.3713579773902893,
      "learning_rate": 0.02,
      "loss": 0.5389602661132813,
      "step": 1550
    },
    {
      "epoch": 3.232636547538773,
      "grad_norm": 0.44091686606407166,
      "learning_rate": 0.019929278846732883,
      "loss": 0.5371050262451171,
      "step": 1600
    },
    {
      "epoch": 3.3337828725556307,
      "grad_norm": 0.4661828577518463,
      "learning_rate": 0.01869964766980079,
      "loss": 0.5366607666015625,
      "step": 1650
    },
    {
      "epoch": 3.434929197572488,
      "grad_norm": 0.42961224913597107,
      "learning_rate": 0.016119035924142468,
      "loss": 0.5338973236083985,
      "step": 1700
    },
    {
      "epoch": 3.536075522589346,
      "grad_norm": 0.4576632082462311,
      "learning_rate": 0.012588190451025208,
      "loss": 0.5357645797729492,
      "step": 1750
    },
    {
      "epoch": 3.536075522589346,
      "eval_accuracy": 0.7037037037037037,
      "eval_f1_macro": 0.6804698569404452,
      "eval_f1_weighted": 0.7027801537605459,
      "eval_loss": 0.6928421854972839,
      "eval_runtime": 0.5153,
      "eval_samples_per_second": 366.775,
      "eval_steps_per_second": 3.881,
      "step": 1750
    },
    {
      "epoch": 3.6372218476062037,
      "grad_norm": 0.4636659026145935,
      "learning_rate": 0.008655421228784682,
      "loss": 0.5186992263793946,
      "step": 1800
    },
    {
      "epoch": 3.738368172623061,
      "grad_norm": 0.38083815574645996,
      "learning_rate": 0.004931453542190172,
      "loss": 0.5256573867797851,
      "step": 1850
    },
    {
      "epoch": 3.839514497639919,
      "grad_norm": 0.37412029504776,
      "learning_rate": 0.001994587590756397,
      "loss": 0.5135831451416015,
      "step": 1900
    },
    {
      "epoch": 3.9406608226567768,
      "grad_norm": 0.35141825675964355,
      "learning_rate": 0.00030089343432194825,
      "loss": 0.5098935699462891,
      "step": 1950
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.7248677248677249,
      "eval_f1_macro": 0.7106796605474553,
      "eval_f1_weighted": 0.7248677248677249,
      "eval_loss": 0.6587123274803162,
      "eval_runtime": 0.4572,
      "eval_samples_per_second": 413.345,
      "eval_steps_per_second": 4.374,
      "step": 1980
    }
  ],
  "logging_steps": 50,
  "max_steps": 1980,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 4,
  "save_steps": 250,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 61833044017152.0,
  "train_batch_size": 128,
  "trial_name": null,
  "trial_params": null
}