last-checkpoint/trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.007343491830365339,
  "eval_steps": 3,
  "global_step": 10,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0007343491830365339,
      "grad_norm": 0.7365296483039856,
      "learning_rate": 2e-05,
      "loss": 0.6968,
      "step": 1
    },
    {
      "epoch": 0.0007343491830365339,
      "eval_loss": 0.6995853185653687,
      "eval_runtime": 172.2307,
      "eval_samples_per_second": 6.66,
      "eval_steps_per_second": 1.666,
      "step": 1
    },
    {
      "epoch": 0.0014686983660730677,
      "grad_norm": 0.7936408519744873,
      "learning_rate": 4e-05,
      "loss": 0.6825,
      "step": 2
    },
    {
      "epoch": 0.0022030475491096016,
      "grad_norm": 0.6955214142799377,
      "learning_rate": 6e-05,
      "loss": 0.6531,
      "step": 3
    },
    {
      "epoch": 0.0022030475491096016,
      "eval_loss": 0.67781662940979,
      "eval_runtime": 171.5728,
      "eval_samples_per_second": 6.685,
      "eval_steps_per_second": 1.673,
      "step": 3
    },
    {
      "epoch": 0.0029373967321461355,
      "grad_norm": 0.6376969218254089,
      "learning_rate": 8e-05,
      "loss": 0.5873,
      "step": 4
    },
    {
      "epoch": 0.0036717459151826694,
      "grad_norm": 0.6535074710845947,
      "learning_rate": 0.0001,
      "loss": 0.5735,
      "step": 5
    },
    {
      "epoch": 0.004406095098219203,
      "grad_norm": 0.7316965460777283,
      "learning_rate": 0.00012,
      "loss": 0.5308,
      "step": 6
    },
    {
      "epoch": 0.004406095098219203,
      "eval_loss": 0.46789583563804626,
      "eval_runtime": 170.6442,
      "eval_samples_per_second": 6.722,
      "eval_steps_per_second": 1.682,
      "step": 6
    },
    {
      "epoch": 0.005140444281255737,
      "grad_norm": 0.5175211429595947,
      "learning_rate": 0.00014,
      "loss": 0.4637,
      "step": 7
    },
    {
      "epoch": 0.005874793464292271,
      "grad_norm": 0.700370192527771,
      "learning_rate": 0.00016,
      "loss": 0.4532,
      "step": 8
    },
    {
      "epoch": 0.006609142647328804,
      "grad_norm": 0.6425027847290039,
      "learning_rate": 0.00018,
      "loss": 0.4004,
      "step": 9
    },
    {
      "epoch": 0.006609142647328804,
      "eval_loss": 0.3521825969219208,
      "eval_runtime": 170.8063,
      "eval_samples_per_second": 6.715,
      "eval_steps_per_second": 1.68,
      "step": 9
    },
    {
      "epoch": 0.007343491830365339,
      "grad_norm": 0.7014135718345642,
      "learning_rate": 0.0002,
      "loss": 0.3319,
      "step": 10
    }
  ],
  "logging_steps": 1,
  "max_steps": 10,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 3,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.317076820033536e+16,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}