checkpoint-16000/trainer_state.json

{
  "best_metric": 5.1074652671813965,
  "best_model_checkpoint": "/home/p318482/babyLM_controlled/models_trained/en_clm/wikipedia_30/checkpoint-16000",
  "epoch": 8.414409676571129,
  "eval_steps": 2000,
  "global_step": 16000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0518012095713911,
      "eval_loss": 7.518920421600342,
      "eval_runtime": 2.2399,
      "eval_samples_per_second": 1422.808,
      "eval_steps_per_second": 89.288,
      "step": 2000
    },
    {
      "epoch": 2.1036024191427822,
      "grad_norm": 1.2947496175765991,
      "learning_rate": 1e-05,
      "loss": 7.6141,
      "step": 4000
    },
    {
      "epoch": 2.1036024191427822,
      "eval_loss": 6.568221092224121,
      "eval_runtime": 2.1236,
      "eval_samples_per_second": 1500.783,
      "eval_steps_per_second": 94.182,
      "step": 4000
    },
    {
      "epoch": 3.155403628714173,
      "eval_loss": 6.16925573348999,
      "eval_runtime": 2.1624,
      "eval_samples_per_second": 1473.825,
      "eval_steps_per_second": 92.49,
      "step": 6000
    },
    {
      "epoch": 4.2072048382855645,
      "grad_norm": 2.330080270767212,
      "learning_rate": 2e-05,
      "loss": 6.2396,
      "step": 8000
    },
    {
      "epoch": 4.2072048382855645,
      "eval_loss": 5.901575088500977,
      "eval_runtime": 2.1326,
      "eval_samples_per_second": 1494.388,
      "eval_steps_per_second": 93.78,
      "step": 8000
    },
    {
      "epoch": 5.259006047856955,
      "eval_loss": 5.67369270324707,
      "eval_runtime": 2.1364,
      "eval_samples_per_second": 1491.788,
      "eval_steps_per_second": 93.617,
      "step": 10000
    },
    {
      "epoch": 6.310807257428346,
      "grad_norm": 2.565220594406128,
      "learning_rate": 2.99925e-05,
      "loss": 5.7217,
      "step": 12000
    },
    {
      "epoch": 6.310807257428346,
      "eval_loss": 5.460005760192871,
      "eval_runtime": 2.1835,
      "eval_samples_per_second": 1459.563,
      "eval_steps_per_second": 91.595,
      "step": 12000
    },
    {
      "epoch": 7.362608466999737,
      "eval_loss": 5.272489547729492,
      "eval_runtime": 2.1833,
      "eval_samples_per_second": 1459.691,
      "eval_steps_per_second": 91.603,
      "step": 14000
    },
    {
      "epoch": 8.414409676571129,
      "grad_norm": 2.7334115505218506,
      "learning_rate": 3.99875e-05,
      "loss": 5.3064,
      "step": 16000
    },
    {
      "epoch": 8.414409676571129,
      "eval_loss": 5.1074652671813965,
      "eval_runtime": 2.2402,
      "eval_samples_per_second": 1422.652,
      "eval_steps_per_second": 89.278,
      "step": 16000
    }
  ],
  "logging_steps": 4000,
  "max_steps": 100000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 53,
  "save_steps": 4000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 4136084771023872.0,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}