{
  "best_global_step": 2000,
  "best_metric": 0.6840855479240417,
  "best_model_checkpoint": "./legal-bert-peft/checkpoint-2000",
  "epoch": 2.0,
  "eval_steps": 500,
  "global_step": 2000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.1,
      "grad_norm": 1.7681492567062378,
      "learning_rate": 3.960000000000001e-05,
      "loss": 1.9858,
      "step": 100
    },
    {
      "epoch": 0.2,
      "grad_norm": 2.879218578338623,
      "learning_rate": 7.960000000000001e-05,
      "loss": 1.5445,
      "step": 200
    },
    {
      "epoch": 0.3,
      "grad_norm": 3.320913076400757,
      "learning_rate": 0.00011960000000000001,
      "loss": 1.2561,
      "step": 300
    },
    {
      "epoch": 0.4,
      "grad_norm": 3.157815933227539,
      "learning_rate": 0.0001596,
      "loss": 1.1633,
      "step": 400
    },
    {
      "epoch": 0.5,
      "grad_norm": 2.4125008583068848,
      "learning_rate": 0.0001996,
      "loss": 1.0927,
      "step": 500
    },
    {
      "epoch": 0.5,
      "eval_loss": 0.9191896915435791,
      "eval_runtime": 32.1718,
      "eval_samples_per_second": 62.166,
      "eval_steps_per_second": 7.771,
      "step": 500
    },
    {
      "epoch": 0.6,
      "grad_norm": 2.8193042278289795,
      "learning_rate": 0.00019208000000000002,
      "loss": 1.0294,
      "step": 600
    },
    {
      "epoch": 0.7,
      "grad_norm": 2.2981388568878174,
      "learning_rate": 0.00018408,
      "loss": 0.9826,
      "step": 700
    },
    {
      "epoch": 0.8,
      "grad_norm": 2.191439151763916,
      "learning_rate": 0.00017608,
      "loss": 0.9392,
      "step": 800
    },
    {
      "epoch": 0.9,
      "grad_norm": 2.4926645755767822,
      "learning_rate": 0.00016808,
      "loss": 0.9081,
      "step": 900
    },
    {
      "epoch": 1.0,
      "grad_norm": 1.8044929504394531,
      "learning_rate": 0.00016008,
      "loss": 0.8846,
      "step": 1000
    },
    {
      "epoch": 1.0,
      "eval_loss": 0.7807372808456421,
      "eval_runtime": 32.3692,
      "eval_samples_per_second": 61.787,
      "eval_steps_per_second": 7.723,
      "step": 1000
    },
    {
      "epoch": 1.1,
      "grad_norm": 1.9347330331802368,
      "learning_rate": 0.00015208,
      "loss": 0.861,
      "step": 1100
    },
    {
      "epoch": 1.2,
      "grad_norm": 1.9204456806182861,
      "learning_rate": 0.00014408000000000002,
      "loss": 0.852,
      "step": 1200
    },
    {
      "epoch": 1.3,
      "grad_norm": 1.9173040390014648,
      "learning_rate": 0.00013608,
      "loss": 0.8211,
      "step": 1300
    },
    {
      "epoch": 1.4,
      "grad_norm": 1.7719258069992065,
      "learning_rate": 0.00012808,
      "loss": 0.8213,
      "step": 1400
    },
    {
      "epoch": 1.5,
      "grad_norm": 1.6444308757781982,
      "learning_rate": 0.00012008000000000002,
      "loss": 0.7994,
      "step": 1500
    },
    {
      "epoch": 1.5,
      "eval_loss": 0.7083392143249512,
      "eval_runtime": 32.4195,
      "eval_samples_per_second": 61.691,
      "eval_steps_per_second": 7.711,
      "step": 1500
    },
    {
      "epoch": 1.6,
      "grad_norm": 1.9344223737716675,
      "learning_rate": 0.00011208000000000001,
      "loss": 0.7955,
      "step": 1600
    },
    {
      "epoch": 1.7,
      "grad_norm": 2.0480849742889404,
      "learning_rate": 0.00010408,
      "loss": 0.7928,
      "step": 1700
    },
    {
      "epoch": 1.8,
      "grad_norm": 1.7024152278900146,
      "learning_rate": 9.608e-05,
      "loss": 0.7896,
      "step": 1800
    },
    {
      "epoch": 1.9,
      "grad_norm": 1.7481478452682495,
      "learning_rate": 8.808000000000001e-05,
      "loss": 0.7765,
      "step": 1900
    },
    {
      "epoch": 2.0,
      "grad_norm": 1.79092276096344,
      "learning_rate": 8.008e-05,
      "loss": 0.7777,
      "step": 2000
    },
    {
      "epoch": 2.0,
      "eval_loss": 0.6840855479240417,
      "eval_runtime": 32.1738,
      "eval_samples_per_second": 62.162,
      "eval_steps_per_second": 7.77,
      "step": 2000
    }
  ],
  "logging_steps": 100,
  "max_steps": 3000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 1000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 8554221600768000.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}