trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.9869281045751634,
  "eval_steps": 500,
  "global_step": 228,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.13071895424836602,
      "grad_norm": 13.640703803287584,
      "learning_rate": 5e-06,
      "loss": 0.9844,
      "step": 10
    },
    {
      "epoch": 0.26143790849673204,
      "grad_norm": 0.9447823765191696,
      "learning_rate": 5e-06,
      "loss": 0.8863,
      "step": 20
    },
    {
      "epoch": 0.39215686274509803,
      "grad_norm": 1.038467083575288,
      "learning_rate": 5e-06,
      "loss": 0.8449,
      "step": 30
    },
    {
      "epoch": 0.5228758169934641,
      "grad_norm": 1.6427820915875102,
      "learning_rate": 5e-06,
      "loss": 0.8257,
      "step": 40
    },
    {
      "epoch": 0.6535947712418301,
      "grad_norm": 1.7936488531165335,
      "learning_rate": 5e-06,
      "loss": 0.8169,
      "step": 50
    },
    {
      "epoch": 0.7843137254901961,
      "grad_norm": 0.7697816321414731,
      "learning_rate": 5e-06,
      "loss": 0.8056,
      "step": 60
    },
    {
      "epoch": 0.9150326797385621,
      "grad_norm": 1.250994290613137,
      "learning_rate": 5e-06,
      "loss": 0.7971,
      "step": 70
    },
    {
      "epoch": 1.0490196078431373,
      "grad_norm": 0.843433753244107,
      "learning_rate": 5e-06,
      "loss": 0.8367,
      "step": 80
    },
    {
      "epoch": 1.1797385620915033,
      "grad_norm": 0.6027656670101825,
      "learning_rate": 5e-06,
      "loss": 0.7483,
      "step": 90
    },
    {
      "epoch": 1.3104575163398693,
      "grad_norm": 0.8024576817469242,
      "learning_rate": 5e-06,
      "loss": 0.7428,
      "step": 100
    },
    {
      "epoch": 1.4411764705882353,
      "grad_norm": 0.7594730875779195,
      "learning_rate": 5e-06,
      "loss": 0.7401,
      "step": 110
    },
    {
      "epoch": 1.5718954248366013,
      "grad_norm": 0.5293998500966177,
      "learning_rate": 5e-06,
      "loss": 0.7426,
      "step": 120
    },
    {
      "epoch": 1.7026143790849673,
      "grad_norm": 0.9249042901353932,
      "learning_rate": 5e-06,
      "loss": 0.7406,
      "step": 130
    },
    {
      "epoch": 1.8333333333333335,
      "grad_norm": 0.5767932991870924,
      "learning_rate": 5e-06,
      "loss": 0.7379,
      "step": 140
    },
    {
      "epoch": 1.9640522875816995,
      "grad_norm": 0.5974379920218519,
      "learning_rate": 5e-06,
      "loss": 0.7337,
      "step": 150
    },
    {
      "epoch": 2.0980392156862746,
      "grad_norm": 1.3799060310730653,
      "learning_rate": 5e-06,
      "loss": 0.7541,
      "step": 160
    },
    {
      "epoch": 2.2287581699346406,
      "grad_norm": 0.8440697657265467,
      "learning_rate": 5e-06,
      "loss": 0.6849,
      "step": 170
    },
    {
      "epoch": 2.3594771241830066,
      "grad_norm": 0.6523664577578698,
      "learning_rate": 5e-06,
      "loss": 0.6828,
      "step": 180
    },
    {
      "epoch": 2.4901960784313726,
      "grad_norm": 0.5604368514967889,
      "learning_rate": 5e-06,
      "loss": 0.6833,
      "step": 190
    },
    {
      "epoch": 2.6209150326797386,
      "grad_norm": 0.651015676014187,
      "learning_rate": 5e-06,
      "loss": 0.6825,
      "step": 200
    },
    {
      "epoch": 2.7516339869281046,
      "grad_norm": 0.6331718263692562,
      "learning_rate": 5e-06,
      "loss": 0.6826,
      "step": 210
    },
    {
      "epoch": 2.8823529411764706,
      "grad_norm": 0.6748382635791591,
      "learning_rate": 5e-06,
      "loss": 0.6867,
      "step": 220
    },
    {
      "epoch": 2.9869281045751634,
      "step": 228,
      "total_flos": 381489732648960.0,
      "train_loss": 0.7633350188272041,
      "train_runtime": 3333.6415,
      "train_samples_per_second": 35.22,
      "train_steps_per_second": 0.068
    }
  ],
  "logging_steps": 10,
  "max_steps": 228,
  "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": 381489732648960.0,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}