trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 10.0,
  "eval_steps": 500,
  "global_step": 40,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.25,
      "grad_norm": 36.80842208862305,
      "learning_rate": 5e-06,
      "loss": 1.0201,
      "step": 1
    },
    {
      "epoch": 1.0,
      "eval_loss": 0.8701804876327515,
      "eval_runtime": 10.3098,
      "eval_samples_per_second": 10.378,
      "eval_steps_per_second": 0.388,
      "step": 4
    },
    {
      "epoch": 1.25,
      "grad_norm": 7.048046112060547,
      "learning_rate": 9.847001329696653e-06,
      "loss": 0.9576,
      "step": 5
    },
    {
      "epoch": 2.0,
      "eval_loss": 0.6512577533721924,
      "eval_runtime": 10.3536,
      "eval_samples_per_second": 10.335,
      "eval_steps_per_second": 0.386,
      "step": 8
    },
    {
      "epoch": 2.5,
      "grad_norm": 3.416456699371338,
      "learning_rate": 8.94570254698197e-06,
      "loss": 0.7027,
      "step": 10
    },
    {
      "epoch": 3.0,
      "eval_loss": 0.46684733033180237,
      "eval_runtime": 10.3209,
      "eval_samples_per_second": 10.367,
      "eval_steps_per_second": 0.388,
      "step": 12
    },
    {
      "epoch": 3.75,
      "grad_norm": 3.9423956871032715,
      "learning_rate": 7.379736965185369e-06,
      "loss": 0.4836,
      "step": 15
    },
    {
      "epoch": 4.0,
      "eval_loss": 0.27517324686050415,
      "eval_runtime": 10.3257,
      "eval_samples_per_second": 10.363,
      "eval_steps_per_second": 0.387,
      "step": 16
    },
    {
      "epoch": 5.0,
      "grad_norm": 5.219548225402832,
      "learning_rate": 5.412896727361663e-06,
      "loss": 0.2544,
      "step": 20
    },
    {
      "epoch": 5.0,
      "eval_loss": 0.14341801404953003,
      "eval_runtime": 10.3284,
      "eval_samples_per_second": 10.36,
      "eval_steps_per_second": 0.387,
      "step": 20
    },
    {
      "epoch": 6.0,
      "eval_loss": 0.07257690280675888,
      "eval_runtime": 10.3236,
      "eval_samples_per_second": 10.365,
      "eval_steps_per_second": 0.387,
      "step": 24
    },
    {
      "epoch": 6.25,
      "grad_norm": 2.7866573333740234,
      "learning_rate": 3.3765026539765832e-06,
      "loss": 0.109,
      "step": 25
    },
    {
      "epoch": 7.0,
      "eval_loss": 0.044053010642528534,
      "eval_runtime": 10.3422,
      "eval_samples_per_second": 10.346,
      "eval_steps_per_second": 0.387,
      "step": 28
    },
    {
      "epoch": 7.5,
      "grad_norm": 1.2933619022369385,
      "learning_rate": 1.6135921418712959e-06,
      "loss": 0.0544,
      "step": 30
    },
    {
      "epoch": 8.0,
      "eval_loss": 0.03408604860305786,
      "eval_runtime": 10.3496,
      "eval_samples_per_second": 10.339,
      "eval_steps_per_second": 0.386,
      "step": 32
    },
    {
      "epoch": 8.75,
      "grad_norm": 0.9404205679893494,
      "learning_rate": 4.211333667247125e-07,
      "loss": 0.0364,
      "step": 35
    },
    {
      "epoch": 9.0,
      "eval_loss": 0.02904023602604866,
      "eval_runtime": 10.3248,
      "eval_samples_per_second": 10.363,
      "eval_steps_per_second": 0.387,
      "step": 36
    },
    {
      "epoch": 10.0,
      "grad_norm": 0.5807708501815796,
      "learning_rate": 0.0,
      "loss": 0.0289,
      "step": 40
    },
    {
      "epoch": 10.0,
      "eval_loss": 0.027868157252669334,
      "eval_runtime": 10.2366,
      "eval_samples_per_second": 10.453,
      "eval_steps_per_second": 0.391,
      "step": 40
    },
    {
      "epoch": 10.0,
      "step": 40,
      "total_flos": 2.2368203314089165e+17,
      "train_loss": 0.3299283441156149,
      "train_runtime": 586.1928,
      "train_samples_per_second": 1.825,
      "train_steps_per_second": 0.068
    }
  ],
  "logging_steps": 5,
  "max_steps": 40,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 10,
  "save_steps": 100,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 2.2368203314089165e+17,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}