{
  "best_metric": 0.1889,
  "best_model_checkpoint": "./runs/byt5-large-ocr-pc/checkpoint-1000",
  "epoch": 9.993210475266732,
  "eval_steps": 500,
  "global_step": 1280,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.3879728419010669,
      "grad_norm": 5.049915790557861,
      "learning_rate": 0.000390625,
      "loss": 2.0359,
      "step": 50
    },
    {
      "epoch": 0.7759456838021338,
      "grad_norm": 1.971483588218689,
      "learning_rate": 0.00048519736842105265,
      "loss": 0.4429,
      "step": 100
    },
    {
      "epoch": 1.1707080504364695,
      "grad_norm": 0.2933945655822754,
      "learning_rate": 0.00046463815789473686,
      "loss": 0.2582,
      "step": 150
    },
    {
      "epoch": 1.5586808923375364,
      "grad_norm": 0.18698160350322723,
      "learning_rate": 0.00044407894736842106,
      "loss": 0.2186,
      "step": 200
    },
    {
      "epoch": 1.9466537342386032,
      "grad_norm": 0.1569913625717163,
      "learning_rate": 0.00042351973684210526,
      "loss": 0.2997,
      "step": 250
    },
    {
      "epoch": 2.341416100872939,
      "grad_norm": 0.12357465922832489,
      "learning_rate": 0.0004029605263157895,
      "loss": 0.163,
      "step": 300
    },
    {
      "epoch": 2.7293889427740057,
      "grad_norm": 0.13612589240074158,
      "learning_rate": 0.00038240131578947366,
      "loss": 0.1606,
      "step": 350
    },
    {
      "epoch": 3.1241513094083415,
      "grad_norm": 0.10511065274477005,
      "learning_rate": 0.0003618421052631579,
      "loss": 0.1475,
      "step": 400
    },
    {
      "epoch": 3.5121241513094086,
      "grad_norm": 0.1341216266155243,
      "learning_rate": 0.0003412828947368421,
      "loss": 0.131,
      "step": 450
    },
    {
      "epoch": 3.900096993210475,
      "grad_norm": 0.11188370734453201,
      "learning_rate": 0.0003207236842105263,
      "loss": 0.1274,
      "step": 500
    },
    {
      "epoch": 3.900096993210475,
      "eval_bleu": 78.724,
      "eval_cer": 0.0847,
      "eval_cer_baseline": 0.1052,
      "eval_cer_reduction": 0.0205,
      "eval_loss": 0.1449187695980072,
      "eval_runtime": 470.9086,
      "eval_samples_per_second": 4.362,
      "eval_steps_per_second": 0.546,
      "eval_wer": 0.219,
      "eval_wer_baseline": 0.3262,
      "eval_wer_reduction": 0.1072,
      "step": 500
    },
    {
      "epoch": 4.294859359844811,
      "grad_norm": 0.09929317981004715,
      "learning_rate": 0.0003001644736842105,
      "loss": 0.1124,
      "step": 550
    },
    {
      "epoch": 4.682832201745878,
      "grad_norm": 0.09972019493579865,
      "learning_rate": 0.00027960526315789477,
      "loss": 0.112,
      "step": 600
    },
    {
      "epoch": 5.0775945683802135,
      "grad_norm": 0.07586631923913956,
      "learning_rate": 0.0002590460526315789,
      "loss": 0.104,
      "step": 650
    },
    {
      "epoch": 5.46556741028128,
      "grad_norm": 0.07480718195438385,
      "learning_rate": 0.00023848684210526317,
      "loss": 0.0915,
      "step": 700
    },
    {
      "epoch": 5.853540252182347,
      "grad_norm": 0.08518869429826736,
      "learning_rate": 0.00021792763157894737,
      "loss": 0.0919,
      "step": 750
    },
    {
      "epoch": 6.248302618816683,
      "grad_norm": 0.07508755475282669,
      "learning_rate": 0.00019736842105263157,
      "loss": 0.0847,
      "step": 800
    },
    {
      "epoch": 6.63627546071775,
      "grad_norm": 0.07648547738790512,
      "learning_rate": 0.0001768092105263158,
      "loss": 0.0795,
      "step": 850
    },
    {
      "epoch": 7.031037827352085,
      "grad_norm": 0.059742897748947144,
      "learning_rate": 0.00015625,
      "loss": 0.0804,
      "step": 900
    },
    {
      "epoch": 7.419010669253153,
      "grad_norm": 0.07011482119560242,
      "learning_rate": 0.0001356907894736842,
      "loss": 0.0672,
      "step": 950
    },
    {
      "epoch": 7.806983511154219,
      "grad_norm": 0.06633743643760681,
      "learning_rate": 0.00011513157894736842,
      "loss": 0.0672,
      "step": 1000
    },
    {
      "epoch": 7.806983511154219,
      "eval_bleu": 81.1256,
      "eval_cer": 0.0651,
      "eval_cer_baseline": 0.1047,
      "eval_cer_reduction": 0.0396,
      "eval_loss": 0.15159520506858826,
      "eval_runtime": 445.4937,
      "eval_samples_per_second": 4.611,
      "eval_steps_per_second": 0.577,
      "eval_wer": 0.1889,
      "eval_wer_baseline": 0.3254,
      "eval_wer_reduction": 0.1365,
      "step": 1000
    },
    {
      "epoch": 8.201745877788555,
      "grad_norm": 0.05455777049064636,
      "learning_rate": 9.457236842105263e-05,
      "loss": 0.066,
      "step": 1050
    },
    {
      "epoch": 8.589718719689621,
      "grad_norm": 0.05844505503773689,
      "learning_rate": 7.401315789473683e-05,
      "loss": 0.0585,
      "step": 1100
    },
    {
      "epoch": 8.977691561590689,
      "grad_norm": 0.0751623660326004,
      "learning_rate": 5.3453947368421055e-05,
      "loss": 0.0607,
      "step": 1150
    },
    {
      "epoch": 9.372453928225024,
      "grad_norm": 0.15131056308746338,
      "learning_rate": 3.289473684210526e-05,
      "loss": 0.0534,
      "step": 1200
    },
    {
      "epoch": 9.760426770126092,
      "grad_norm": 0.16118672490119934,
      "learning_rate": 1.2335526315789473e-05,
      "loss": 0.0554,
      "step": 1250
    },
    {
      "epoch": 9.993210475266732,
      "step": 1280,
      "total_flos": 1.162864585825321e+17,
      "train_loss": 0.203181495051831,
      "train_runtime": 3159.2527,
      "train_samples_per_second": 26.101,
      "train_steps_per_second": 0.405
    }
  ],
  "logging_steps": 50,
  "max_steps": 1280,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 10,
  "save_steps": 500,
  "stateful_callbacks": {
    "EarlyStoppingCallback": {
      "args": {
        "early_stopping_patience": 5,
        "early_stopping_threshold": 0.0
      },
      "attributes": {
        "early_stopping_patience_counter": 0
      }
    },
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.162864585825321e+17,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}