checkpoint-500/trainer_state.json

{
  "best_metric": 0.4058004915714264,
  "best_model_checkpoint": "mikhail-panzo/ceb_b128_le4_s8000/checkpoint-500",
  "epoch": 39.603960396039604,
  "eval_steps": 500,
  "global_step": 500,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 3.9603960396039604,
      "grad_norm": 2.95651912689209,
      "learning_rate": 2.5e-06,
      "loss": 0.7707,
      "step": 50
    },
    {
      "epoch": 7.920792079207921,
      "grad_norm": 1.1273679733276367,
      "learning_rate": 5e-06,
      "loss": 0.6934,
      "step": 100
    },
    {
      "epoch": 11.881188118811881,
      "grad_norm": 2.1830034255981445,
      "learning_rate": 7.5e-06,
      "loss": 0.5983,
      "step": 150
    },
    {
      "epoch": 15.841584158415841,
      "grad_norm": 1.0733364820480347,
      "learning_rate": 1e-05,
      "loss": 0.5094,
      "step": 200
    },
    {
      "epoch": 19.801980198019802,
      "grad_norm": 1.2824387550354004,
      "learning_rate": 1.25e-05,
      "loss": 0.4913,
      "step": 250
    },
    {
      "epoch": 23.762376237623762,
      "grad_norm": 0.9164418578147888,
      "learning_rate": 1.5e-05,
      "loss": 0.4764,
      "step": 300
    },
    {
      "epoch": 27.722772277227723,
      "grad_norm": 0.8938279747962952,
      "learning_rate": 1.75e-05,
      "loss": 0.4648,
      "step": 350
    },
    {
      "epoch": 31.683168316831683,
      "grad_norm": 0.7855221033096313,
      "learning_rate": 2e-05,
      "loss": 0.459,
      "step": 400
    },
    {
      "epoch": 35.64356435643565,
      "grad_norm": 0.9247761964797974,
      "learning_rate": 2.25e-05,
      "loss": 0.4519,
      "step": 450
    },
    {
      "epoch": 39.603960396039604,
      "grad_norm": 1.1086606979370117,
      "learning_rate": 2.5e-05,
      "loss": 0.4432,
      "step": 500
    },
    {
      "epoch": 39.603960396039604,
      "eval_loss": 0.4058004915714264,
      "eval_runtime": 7.6054,
      "eval_samples_per_second": 23.667,
      "eval_steps_per_second": 3.024,
      "step": 500
    }
  ],
  "logging_steps": 50,
  "max_steps": 8000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 667,
  "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": 1.0803690034289064e+16,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}