run-16/checkpoint-2544/trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 8.0,
  "eval_steps": 500,
  "global_step": 2544,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.9968553459119497,
      "grad_norm": 0.7093676328659058,
      "learning_rate": 1.7507861635220126e-05,
      "loss": 0.4971,
      "step": 317
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.6170967741935484,
      "eval_loss": 0.2514326870441437,
      "eval_runtime": 1.427,
      "eval_samples_per_second": 2172.377,
      "eval_steps_per_second": 45.55,
      "step": 318
    },
    {
      "epoch": 1.9937106918238994,
      "grad_norm": 0.5753037333488464,
      "learning_rate": 1.5015723270440253e-05,
      "loss": 0.2041,
      "step": 634
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8416129032258064,
      "eval_loss": 0.11213530600070953,
      "eval_runtime": 1.4434,
      "eval_samples_per_second": 2147.677,
      "eval_steps_per_second": 45.032,
      "step": 636
    },
    {
      "epoch": 2.990566037735849,
      "grad_norm": 0.503776490688324,
      "learning_rate": 1.2523584905660378e-05,
      "loss": 0.1194,
      "step": 951
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8858064516129033,
      "eval_loss": 0.06934015452861786,
      "eval_runtime": 1.442,
      "eval_samples_per_second": 2149.749,
      "eval_steps_per_second": 45.075,
      "step": 954
    },
    {
      "epoch": 3.9874213836477987,
      "grad_norm": 0.45865872502326965,
      "learning_rate": 1.0031446540880504e-05,
      "loss": 0.0864,
      "step": 1268
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9045161290322581,
      "eval_loss": 0.05072392523288727,
      "eval_runtime": 1.4463,
      "eval_samples_per_second": 2143.474,
      "eval_steps_per_second": 44.944,
      "step": 1272
    },
    {
      "epoch": 4.984276729559748,
      "grad_norm": 0.35762810707092285,
      "learning_rate": 7.539308176100629e-06,
      "loss": 0.07,
      "step": 1585
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9125806451612903,
      "eval_loss": 0.04198624566197395,
      "eval_runtime": 1.6456,
      "eval_samples_per_second": 1883.86,
      "eval_steps_per_second": 39.5,
      "step": 1590
    },
    {
      "epoch": 5.981132075471698,
      "grad_norm": 0.4083423316478729,
      "learning_rate": 5.047169811320756e-06,
      "loss": 0.0612,
      "step": 1902
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9183870967741935,
      "eval_loss": 0.03691146895289421,
      "eval_runtime": 1.6409,
      "eval_samples_per_second": 1889.151,
      "eval_steps_per_second": 39.611,
      "step": 1908
    },
    {
      "epoch": 6.977987421383648,
      "grad_norm": 0.2706115245819092,
      "learning_rate": 2.555031446540881e-06,
      "loss": 0.0564,
      "step": 2219
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9229032258064516,
      "eval_loss": 0.03457155451178551,
      "eval_runtime": 1.6398,
      "eval_samples_per_second": 1890.51,
      "eval_steps_per_second": 39.64,
      "step": 2226
    },
    {
      "epoch": 7.9748427672955975,
      "grad_norm": 0.29390910267829895,
      "learning_rate": 6.289308176100629e-08,
      "loss": 0.0541,
      "step": 2536
    }
  ],
  "logging_steps": 317,
  "max_steps": 2544,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 8,
  "save_steps": 1000000000.0,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 660694498184616.0,
  "train_batch_size": 48,
  "trial_name": null,
  "trial_params": {
    "alpha": 0.4600551411814572,
    "num_train_epochs": 8,
    "temperature": 4
  }
}