{
  "best_metric": 0.2653820216655731,
  "best_model_checkpoint": "/home/philipp.meier/author_writing_sentences/outputs/specific/ernie/macro/hard_512/checkpoint-3190",
  "epoch": 9.999530001566661,
  "eval_steps": 500,
  "global_step": 15950,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.31333228889237036,
      "grad_norm": 6.813873767852783,
      "learning_rate": 4.8432601880877745e-05,
      "loss": 0.3216,
      "step": 500
    },
    {
      "epoch": 0.6266645777847407,
      "grad_norm": 2.24408221244812,
      "learning_rate": 4.686520376175549e-05,
      "loss": 0.2793,
      "step": 1000
    },
    {
      "epoch": 0.9399968666771111,
      "grad_norm": 3.0516178607940674,
      "learning_rate": 4.529780564263323e-05,
      "loss": 0.2721,
      "step": 1500
    },
    {
      "epoch": 0.9995300015666615,
      "eval_f1": 0.7722205654410168,
      "eval_loss": 0.2868705093860626,
      "eval_runtime": 91.0612,
      "eval_samples_per_second": 57.577,
      "eval_steps_per_second": 14.397,
      "step": 1595
    },
    {
      "epoch": 1.25379915400282,
      "grad_norm": 2.2203149795532227,
      "learning_rate": 4.373040752351097e-05,
      "loss": 0.2338,
      "step": 2000
    },
    {
      "epoch": 1.5671314428951904,
      "grad_norm": 3.7226078510284424,
      "learning_rate": 4.2163009404388715e-05,
      "loss": 0.2156,
      "step": 2500
    },
    {
      "epoch": 1.8804637317875605,
      "grad_norm": 11.44078254699707,
      "learning_rate": 4.059561128526646e-05,
      "loss": 0.2212,
      "step": 3000
    },
    {
      "epoch": 1.9995300015666615,
      "eval_f1": 0.7341981355267706,
      "eval_loss": 0.2653820216655731,
      "eval_runtime": 91.0,
      "eval_samples_per_second": 57.615,
      "eval_steps_per_second": 14.407,
      "step": 3190
    },
    {
      "epoch": 2.1942660191132695,
      "grad_norm": 2.2864012718200684,
      "learning_rate": 3.90282131661442e-05,
      "loss": 0.183,
      "step": 3500
    },
    {
      "epoch": 2.50759830800564,
      "grad_norm": 10.732903480529785,
      "learning_rate": 3.746081504702195e-05,
      "loss": 0.1671,
      "step": 4000
    },
    {
      "epoch": 2.8209305968980103,
      "grad_norm": 5.939886093139648,
      "learning_rate": 3.5893416927899686e-05,
      "loss": 0.1721,
      "step": 4500
    },
    {
      "epoch": 2.9995300015666615,
      "eval_f1": 0.8214090906804821,
      "eval_loss": 0.3026501536369324,
      "eval_runtime": 90.9433,
      "eval_samples_per_second": 57.651,
      "eval_steps_per_second": 14.416,
      "step": 4785
    },
    {
      "epoch": 3.134732884223719,
      "grad_norm": 1.2383900880813599,
      "learning_rate": 3.4326018808777435e-05,
      "loss": 0.1467,
      "step": 5000
    },
    {
      "epoch": 3.4480651731160896,
      "grad_norm": 4.738731861114502,
      "learning_rate": 3.275862068965517e-05,
      "loss": 0.1177,
      "step": 5500
    },
    {
      "epoch": 3.76139746200846,
      "grad_norm": 8.59209156036377,
      "learning_rate": 3.119122257053292e-05,
      "loss": 0.1218,
      "step": 6000
    },
    {
      "epoch": 3.9995300015666615,
      "eval_f1": 0.8193411263336241,
      "eval_loss": 0.2795597016811371,
      "eval_runtime": 91.0456,
      "eval_samples_per_second": 57.587,
      "eval_steps_per_second": 14.399,
      "step": 6380
    },
    {
      "epoch": 4.0751997493341685,
      "grad_norm": 26.97426986694336,
      "learning_rate": 2.962382445141066e-05,
      "loss": 0.1189,
      "step": 6500
    },
    {
      "epoch": 4.388532038226539,
      "grad_norm": 3.5549023151397705,
      "learning_rate": 2.8056426332288405e-05,
      "loss": 0.0879,
      "step": 7000
    },
    {
      "epoch": 4.701864327118909,
      "grad_norm": 4.484798431396484,
      "learning_rate": 2.6489028213166144e-05,
      "loss": 0.0878,
      "step": 7500
    },
    {
      "epoch": 4.9995300015666615,
      "eval_f1": 0.8158898829210776,
      "eval_loss": 0.3778676986694336,
      "eval_runtime": 90.7155,
      "eval_samples_per_second": 57.796,
      "eval_steps_per_second": 14.452,
      "step": 7975
    },
    {
      "epoch": 5.015666614444618,
      "grad_norm": 0.11927812546491623,
      "learning_rate": 2.4921630094043887e-05,
      "loss": 0.0921,
      "step": 8000
    },
    {
      "epoch": 5.328998903336989,
      "grad_norm": 7.248912811279297,
      "learning_rate": 2.335423197492163e-05,
      "loss": 0.064,
      "step": 8500
    },
    {
      "epoch": 5.642331192229359,
      "grad_norm": 5.339770793914795,
      "learning_rate": 2.1786833855799376e-05,
      "loss": 0.0675,
      "step": 9000
    },
    {
      "epoch": 5.9556634811217295,
      "grad_norm": 7.253896713256836,
      "learning_rate": 2.0219435736677118e-05,
      "loss": 0.0731,
      "step": 9500
    },
    {
      "epoch": 5.9995300015666615,
      "eval_f1": 0.815418634651262,
      "eval_loss": 0.48566773533821106,
      "eval_runtime": 90.6913,
      "eval_samples_per_second": 57.811,
      "eval_steps_per_second": 14.456,
      "step": 9570
    },
    {
      "epoch": 6.269465768447438,
      "grad_norm": 11.255759239196777,
      "learning_rate": 1.865203761755486e-05,
      "loss": 0.0509,
      "step": 10000
    },
    {
      "epoch": 6.582798057339809,
      "grad_norm": 14.333255767822266,
      "learning_rate": 1.7084639498432603e-05,
      "loss": 0.0509,
      "step": 10500
    },
    {
      "epoch": 6.896130346232179,
      "grad_norm": 0.0382259301841259,
      "learning_rate": 1.5517241379310346e-05,
      "loss": 0.0459,
      "step": 11000
    },
    {
      "epoch": 6.9995300015666615,
      "eval_f1": 0.8059023505893753,
      "eval_loss": 0.5288071036338806,
      "eval_runtime": 90.738,
      "eval_samples_per_second": 57.782,
      "eval_steps_per_second": 14.448,
      "step": 11165
    },
    {
      "epoch": 7.209932633557888,
      "grad_norm": 0.03432682529091835,
      "learning_rate": 1.3949843260188088e-05,
      "loss": 0.0415,
      "step": 11500
    },
    {
      "epoch": 7.523264922450259,
      "grad_norm": 12.062665939331055,
      "learning_rate": 1.2382445141065831e-05,
      "loss": 0.0297,
      "step": 12000
    },
    {
      "epoch": 7.836597211342629,
      "grad_norm": 0.04464669153094292,
      "learning_rate": 1.0815047021943574e-05,
      "loss": 0.0362,
      "step": 12500
    },
    {
      "epoch": 7.9995300015666615,
      "eval_f1": 0.812358261029243,
      "eval_loss": 0.6266328692436218,
      "eval_runtime": 90.879,
      "eval_samples_per_second": 57.692,
      "eval_steps_per_second": 14.426,
      "step": 12760
    },
    {
      "epoch": 8.150399498668337,
      "grad_norm": 0.12006869912147522,
      "learning_rate": 9.247648902821316e-06,
      "loss": 0.025,
      "step": 13000
    },
    {
      "epoch": 8.463731787560707,
      "grad_norm": 0.046469058841466904,
      "learning_rate": 7.68025078369906e-06,
      "loss": 0.0245,
      "step": 13500
    },
    {
      "epoch": 8.777064076453078,
      "grad_norm": 0.018443206325173378,
      "learning_rate": 6.112852664576803e-06,
      "loss": 0.0234,
      "step": 14000
    },
    {
      "epoch": 8.999530001566661,
      "eval_f1": 0.8142940244894366,
      "eval_loss": 0.6539496779441833,
      "eval_runtime": 90.9198,
      "eval_samples_per_second": 57.666,
      "eval_steps_per_second": 14.419,
      "step": 14355
    },
    {
      "epoch": 9.090866363778787,
      "grad_norm": 0.011548267677426338,
      "learning_rate": 4.5454545454545455e-06,
      "loss": 0.0201,
      "step": 14500
    },
    {
      "epoch": 9.404198652671157,
      "grad_norm": 0.01802981086075306,
      "learning_rate": 2.9780564263322885e-06,
      "loss": 0.0193,
      "step": 15000
    },
    {
      "epoch": 9.717530941563528,
      "grad_norm": 0.10749288648366928,
      "learning_rate": 1.4106583072100315e-06,
      "loss": 0.0176,
      "step": 15500
    },
    {
      "epoch": 9.999530001566661,
      "eval_f1": 0.8171148524282092,
      "eval_loss": 0.723384439945221,
      "eval_runtime": 90.7517,
      "eval_samples_per_second": 57.773,
      "eval_steps_per_second": 14.446,
      "step": 15950
    }
  ],
  "logging_steps": 500,
  "max_steps": 15950,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 10,
  "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": 1.343392426457088e+17,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}