trainer_state.json

{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.0,
  "eval_steps": 500,
  "global_step": 1044,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.28735632183908044,
      "grad_norm": 0.22320619225502014,
      "learning_rate": 3.787896253602306e-05,
      "loss": 4.3135,
      "step": 100
    },
    {
      "epoch": 0.5747126436781609,
      "grad_norm": 0.24471743404865265,
      "learning_rate": 3.557348703170029e-05,
      "loss": 3.7517,
      "step": 200
    },
    {
      "epoch": 0.8620689655172413,
      "grad_norm": 0.24077577888965607,
      "learning_rate": 3.326801152737752e-05,
      "loss": 3.7293,
      "step": 300
    },
    {
      "epoch": 1.0,
      "eval_loss": 3.7024362087249756,
      "eval_runtime": 234.2655,
      "eval_samples_per_second": 1.481,
      "eval_steps_per_second": 1.481,
      "step": 348
    },
    {
      "epoch": 1.1494252873563218,
      "grad_norm": 0.21085461974143982,
      "learning_rate": 3.096253602305476e-05,
      "loss": 3.7261,
      "step": 400
    },
    {
      "epoch": 1.4367816091954024,
      "grad_norm": 0.24059197306632996,
      "learning_rate": 2.865706051873199e-05,
      "loss": 3.7369,
      "step": 500
    },
    {
      "epoch": 1.7241379310344827,
      "grad_norm": 0.251012921333313,
      "learning_rate": 2.6351585014409226e-05,
      "loss": 3.7541,
      "step": 600
    },
    {
      "epoch": 2.0,
      "eval_loss": 3.6997647285461426,
      "eval_runtime": 232.9377,
      "eval_samples_per_second": 1.49,
      "eval_steps_per_second": 1.49,
      "step": 696
    },
    {
      "epoch": 2.0114942528735633,
      "grad_norm": 0.24482741951942444,
      "learning_rate": 2.4046109510086458e-05,
      "loss": 3.6996,
      "step": 700
    },
    {
      "epoch": 2.2988505747126435,
      "grad_norm": 0.30800557136535645,
      "learning_rate": 2.174063400576369e-05,
      "loss": 3.7372,
      "step": 800
    },
    {
      "epoch": 2.586206896551724,
      "grad_norm": 0.3067159056663513,
      "learning_rate": 1.943515850144092e-05,
      "loss": 3.6875,
      "step": 900
    },
    {
      "epoch": 2.873563218390805,
      "grad_norm": 0.3820016086101532,
      "learning_rate": 1.7129682997118157e-05,
      "loss": 3.6998,
      "step": 1000
    },
    {
      "epoch": 3.0,
      "eval_loss": 3.7038047313690186,
      "eval_runtime": 232.8777,
      "eval_samples_per_second": 1.49,
      "eval_steps_per_second": 1.49,
      "step": 1044
    }
  ],
  "logging_steps": 100,
  "max_steps": 1740,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 1.9328482134510797e+17,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}