{
  "best_metric": 0.1112121045589447,
  "best_model_checkpoint": "./vit5_qqp/checkpoint-9855",
  "epoch": 3.0,
  "eval_steps": 500,
  "global_step": 9855,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.15220700152207,
      "grad_norm": 0.6250647902488708,
      "learning_rate": 4.7463216641298836e-05,
      "loss": 0.5593,
      "step": 500
    },
    {
      "epoch": 0.30441400304414,
      "grad_norm": 0.41860121488571167,
      "learning_rate": 4.492643328259767e-05,
      "loss": 0.2028,
      "step": 1000
    },
    {
      "epoch": 0.45662100456621,
      "grad_norm": 0.5042670369148254,
      "learning_rate": 4.2389649923896504e-05,
      "loss": 0.1702,
      "step": 1500
    },
    {
      "epoch": 0.60882800608828,
      "grad_norm": 0.357123464345932,
      "learning_rate": 3.985286656519534e-05,
      "loss": 0.1601,
      "step": 2000
    },
    {
      "epoch": 0.76103500761035,
      "grad_norm": 0.398662805557251,
      "learning_rate": 3.731608320649417e-05,
      "loss": 0.1551,
      "step": 2500
    },
    {
      "epoch": 0.91324200913242,
      "grad_norm": 1.139906644821167,
      "learning_rate": 3.4779299847793e-05,
      "loss": 0.1485,
      "step": 3000
    },
    {
      "epoch": 1.0,
      "eval_loss": 0.1306794434785843,
      "eval_runtime": 178.8964,
      "eval_samples_per_second": 73.16,
      "eval_steps_per_second": 2.035,
      "step": 3285
    },
    {
      "epoch": 1.06544901065449,
      "grad_norm": 0.4199409484863281,
      "learning_rate": 3.224251648909183e-05,
      "loss": 0.1363,
      "step": 3500
    },
    {
      "epoch": 1.21765601217656,
      "grad_norm": 0.313374787569046,
      "learning_rate": 2.9705733130390666e-05,
      "loss": 0.1246,
      "step": 4000
    },
    {
      "epoch": 1.36986301369863,
      "grad_norm": 0.43579331040382385,
      "learning_rate": 2.71689497716895e-05,
      "loss": 0.1224,
      "step": 4500
    },
    {
      "epoch": 1.5220700152207,
      "grad_norm": 0.43089091777801514,
      "learning_rate": 2.463216641298833e-05,
      "loss": 0.1223,
      "step": 5000
    },
    {
      "epoch": 1.6742770167427703,
      "grad_norm": 0.3335800766944885,
      "learning_rate": 2.2095383054287164e-05,
      "loss": 0.1193,
      "step": 5500
    },
    {
      "epoch": 1.82648401826484,
      "grad_norm": 0.4197755455970764,
      "learning_rate": 1.9558599695585997e-05,
      "loss": 0.1178,
      "step": 6000
    },
    {
      "epoch": 1.9786910197869103,
      "grad_norm": 0.4110891819000244,
      "learning_rate": 1.702181633688483e-05,
      "loss": 0.1156,
      "step": 6500
    },
    {
      "epoch": 2.0,
      "eval_loss": 0.11589660495519638,
      "eval_runtime": 178.8699,
      "eval_samples_per_second": 73.171,
      "eval_steps_per_second": 2.035,
      "step": 6570
    },
    {
      "epoch": 2.13089802130898,
      "grad_norm": 0.3453280031681061,
      "learning_rate": 1.4485032978183663e-05,
      "loss": 0.101,
      "step": 7000
    },
    {
      "epoch": 2.2831050228310503,
      "grad_norm": 0.3175284266471863,
      "learning_rate": 1.1948249619482495e-05,
      "loss": 0.097,
      "step": 7500
    },
    {
      "epoch": 2.43531202435312,
      "grad_norm": 0.42786237597465515,
      "learning_rate": 9.411466260781329e-06,
      "loss": 0.0959,
      "step": 8000
    },
    {
      "epoch": 2.5875190258751903,
      "grad_norm": 0.4301067590713501,
      "learning_rate": 6.874682902080162e-06,
      "loss": 0.095,
      "step": 8500
    },
    {
      "epoch": 2.73972602739726,
      "grad_norm": 0.3909034729003906,
      "learning_rate": 4.337899543378996e-06,
      "loss": 0.0958,
      "step": 9000
    },
    {
      "epoch": 2.8919330289193304,
      "grad_norm": 0.38906097412109375,
      "learning_rate": 1.8011161846778284e-06,
      "loss": 0.0939,
      "step": 9500
    },
    {
      "epoch": 3.0,
      "eval_loss": 0.1112121045589447,
      "eval_runtime": 179.4052,
      "eval_samples_per_second": 72.952,
      "eval_steps_per_second": 2.029,
      "step": 9855
    }
  ],
  "logging_steps": 500,
  "max_steps": 9855,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 7.50631108018176e+16,
  "train_batch_size": 36,
  "trial_name": null,
  "trial_params": null
}