trainer_state.json

{
  "best_global_step": 8000,
  "best_metric": 0.11000050604343414,
  "best_model_checkpoint": "./blip-vqa-lora-final-1/checkpoint-8000",
  "epoch": 1.8479039150017322,
  "eval_steps": 1000,
  "global_step": 8000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.23097355352812102,
      "grad_norm": 0.8786712884902954,
      "learning_rate": 0.00017701547701547702,
      "loss": 0.7467,
      "step": 1000
    },
    {
      "epoch": 0.23097355352812102,
      "eval_loss": 0.16969867050647736,
      "eval_runtime": 989.2451,
      "eval_samples_per_second": 6.564,
      "eval_steps_per_second": 0.821,
      "step": 1000
    },
    {
      "epoch": 0.46194710705624203,
      "grad_norm": 0.578987717628479,
      "learning_rate": 0.0001539154539154539,
      "loss": 0.1516,
      "step": 2000
    },
    {
      "epoch": 0.46194710705624203,
      "eval_loss": 0.146992489695549,
      "eval_runtime": 964.1326,
      "eval_samples_per_second": 6.735,
      "eval_steps_per_second": 0.842,
      "step": 2000
    },
    {
      "epoch": 0.6929206605843631,
      "grad_norm": 1.0143482685089111,
      "learning_rate": 0.00013081543081543083,
      "loss": 0.1343,
      "step": 3000
    },
    {
      "epoch": 0.6929206605843631,
      "eval_loss": 0.1319289207458496,
      "eval_runtime": 958.9433,
      "eval_samples_per_second": 6.771,
      "eval_steps_per_second": 0.847,
      "step": 3000
    },
    {
      "epoch": 0.9238942141124841,
      "grad_norm": 0.3963450789451599,
      "learning_rate": 0.00010771540771540771,
      "loss": 0.124,
      "step": 4000
    },
    {
      "epoch": 0.9238942141124841,
      "eval_loss": 0.1267184019088745,
      "eval_runtime": 1076.6607,
      "eval_samples_per_second": 6.031,
      "eval_steps_per_second": 0.754,
      "step": 4000
    },
    {
      "epoch": 1.1549832544173693,
      "grad_norm": 0.6118173003196716,
      "learning_rate": 8.461538461538461e-05,
      "loss": 0.1012,
      "step": 5000
    },
    {
      "epoch": 1.1549832544173693,
      "eval_loss": 0.12135909497737885,
      "eval_runtime": 1032.2551,
      "eval_samples_per_second": 6.29,
      "eval_steps_per_second": 0.787,
      "step": 5000
    },
    {
      "epoch": 1.3859568079454903,
      "grad_norm": 0.6660658121109009,
      "learning_rate": 6.151536151536152e-05,
      "loss": 0.0862,
      "step": 6000
    },
    {
      "epoch": 1.3859568079454903,
      "eval_loss": 0.11788077652454376,
      "eval_runtime": 1024.8116,
      "eval_samples_per_second": 6.336,
      "eval_steps_per_second": 0.792,
      "step": 6000
    },
    {
      "epoch": 1.6169303614736112,
      "grad_norm": 0.7107754349708557,
      "learning_rate": 3.841533841533842e-05,
      "loss": 0.0832,
      "step": 7000
    },
    {
      "epoch": 1.6169303614736112,
      "eval_loss": 0.11289198696613312,
      "eval_runtime": 1022.4664,
      "eval_samples_per_second": 6.35,
      "eval_steps_per_second": 0.794,
      "step": 7000
    },
    {
      "epoch": 1.8479039150017322,
      "grad_norm": 0.6195062398910522,
      "learning_rate": 1.5315315315315316e-05,
      "loss": 0.0809,
      "step": 8000
    },
    {
      "epoch": 1.8479039150017322,
      "eval_loss": 0.11000050604343414,
      "eval_runtime": 1027.0762,
      "eval_samples_per_second": 6.322,
      "eval_steps_per_second": 0.791,
      "step": 8000
    }
  ],
  "logging_steps": 1000,
  "max_steps": 8658,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 1000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 3.416643625195315e+16,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": null
}