{
  "best_metric": 2.6344850063323975,
  "best_model_checkpoint": "/home/p318482/babyLM_controlled/models_trained_last/de_mlm/de_childes_42/checkpoint-88000",
  "epoch": 66.1010138941044,
  "eval_steps": 2000,
  "global_step": 88000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.5020653398422832,
      "eval_loss": 7.505205154418945,
      "eval_runtime": 4.9012,
      "eval_samples_per_second": 677.383,
      "eval_steps_per_second": 42.438,
      "step": 2000
    },
    {
      "epoch": 3.0041306796845664,
      "grad_norm": 1.3569484949111938,
      "learning_rate": 1e-05,
      "loss": 7.4063,
      "step": 4000
    },
    {
      "epoch": 3.0041306796845664,
      "eval_loss": 6.433775901794434,
      "eval_runtime": 4.7916,
      "eval_samples_per_second": 692.873,
      "eval_steps_per_second": 43.409,
      "step": 4000
    },
    {
      "epoch": 4.50619601952685,
      "eval_loss": 6.305135250091553,
      "eval_runtime": 4.8722,
      "eval_samples_per_second": 681.421,
      "eval_steps_per_second": 42.691,
      "step": 6000
    },
    {
      "epoch": 6.008261359369133,
      "grad_norm": 1.548327088356018,
      "learning_rate": 2e-05,
      "loss": 6.0672,
      "step": 8000
    },
    {
      "epoch": 6.008261359369133,
      "eval_loss": 6.1662516593933105,
      "eval_runtime": 4.8572,
      "eval_samples_per_second": 683.527,
      "eval_steps_per_second": 42.823,
      "step": 8000
    },
    {
      "epoch": 7.510326699211416,
      "eval_loss": 6.095215797424316,
      "eval_runtime": 4.7932,
      "eval_samples_per_second": 692.655,
      "eval_steps_per_second": 43.395,
      "step": 10000
    },
    {
      "epoch": 9.0123920390537,
      "grad_norm": 1.639375925064087,
      "learning_rate": 3e-05,
      "loss": 5.8691,
      "step": 12000
    },
    {
      "epoch": 9.0123920390537,
      "eval_loss": 6.0121965408325195,
      "eval_runtime": 4.8217,
      "eval_samples_per_second": 688.56,
      "eval_steps_per_second": 43.139,
      "step": 12000
    },
    {
      "epoch": 10.514457378895981,
      "eval_loss": 5.951181888580322,
      "eval_runtime": 4.7668,
      "eval_samples_per_second": 696.478,
      "eval_steps_per_second": 43.635,
      "step": 14000
    },
    {
      "epoch": 12.016522718738265,
      "grad_norm": 1.6758147478103638,
      "learning_rate": 4e-05,
      "loss": 5.7209,
      "step": 16000
    },
    {
      "epoch": 12.016522718738265,
      "eval_loss": 5.869147300720215,
      "eval_runtime": 4.6953,
      "eval_samples_per_second": 707.092,
      "eval_steps_per_second": 44.3,
      "step": 16000
    },
    {
      "epoch": 13.518588058580548,
      "eval_loss": 5.852941513061523,
      "eval_runtime": 4.8028,
      "eval_samples_per_second": 691.267,
      "eval_steps_per_second": 43.308,
      "step": 18000
    },
    {
      "epoch": 15.020653398422832,
      "grad_norm": 2.3887548446655273,
      "learning_rate": 5e-05,
      "loss": 5.6105,
      "step": 20000
    },
    {
      "epoch": 15.020653398422832,
      "eval_loss": 5.796474933624268,
      "eval_runtime": 4.7859,
      "eval_samples_per_second": 693.704,
      "eval_steps_per_second": 43.461,
      "step": 20000
    },
    {
      "epoch": 16.522718738265116,
      "eval_loss": 5.740424156188965,
      "eval_runtime": 4.6987,
      "eval_samples_per_second": 706.584,
      "eval_steps_per_second": 44.268,
      "step": 22000
    },
    {
      "epoch": 18.0247840781074,
      "grad_norm": 4.52785587310791,
      "learning_rate": 6e-05,
      "loss": 5.5302,
      "step": 24000
    },
    {
      "epoch": 18.0247840781074,
      "eval_loss": 5.742419719696045,
      "eval_runtime": 3.8111,
      "eval_samples_per_second": 871.137,
      "eval_steps_per_second": 54.577,
      "step": 24000
    },
    {
      "epoch": 19.52684941794968,
      "eval_loss": 5.7255754470825195,
      "eval_runtime": 2.6268,
      "eval_samples_per_second": 1263.889,
      "eval_steps_per_second": 79.183,
      "step": 26000
    },
    {
      "epoch": 21.028914757791963,
      "grad_norm": 4.69941520690918,
      "learning_rate": 7e-05,
      "loss": 5.4587,
      "step": 28000
    },
    {
      "epoch": 21.028914757791963,
      "eval_loss": 5.6831231117248535,
      "eval_runtime": 2.5816,
      "eval_samples_per_second": 1286.034,
      "eval_steps_per_second": 80.571,
      "step": 28000
    },
    {
      "epoch": 22.53098009763425,
      "eval_loss": 5.396602153778076,
      "eval_runtime": 2.6907,
      "eval_samples_per_second": 1233.859,
      "eval_steps_per_second": 77.302,
      "step": 30000
    },
    {
      "epoch": 24.03304543747653,
      "grad_norm": 5.43873929977417,
      "learning_rate": 8e-05,
      "loss": 5.0899,
      "step": 32000
    },
    {
      "epoch": 24.03304543747653,
      "eval_loss": 4.704195022583008,
      "eval_runtime": 2.6893,
      "eval_samples_per_second": 1234.521,
      "eval_steps_per_second": 77.343,
      "step": 32000
    },
    {
      "epoch": 25.535110777318813,
      "eval_loss": 4.2316813468933105,
      "eval_runtime": 4.2325,
      "eval_samples_per_second": 784.414,
      "eval_steps_per_second": 49.144,
      "step": 34000
    },
    {
      "epoch": 27.037176117161096,
      "grad_norm": 4.492648124694824,
      "learning_rate": 9e-05,
      "loss": 4.0988,
      "step": 36000
    },
    {
      "epoch": 27.037176117161096,
      "eval_loss": 3.90928316116333,
      "eval_runtime": 4.8525,
      "eval_samples_per_second": 684.177,
      "eval_steps_per_second": 42.864,
      "step": 36000
    },
    {
      "epoch": 28.549380398047315,
      "eval_loss": 3.749643564224243,
      "eval_runtime": 4.8877,
      "eval_samples_per_second": 679.256,
      "eval_steps_per_second": 42.556,
      "step": 38000
    },
    {
      "epoch": 30.051445737889598,
      "grad_norm": 4.555140018463135,
      "learning_rate": 0.0001,
      "loss": 3.555,
      "step": 40000
    },
    {
      "epoch": 30.051445737889598,
      "eval_loss": 3.5960817337036133,
      "eval_runtime": 4.777,
      "eval_samples_per_second": 695.0,
      "eval_steps_per_second": 43.542,
      "step": 40000
    },
    {
      "epoch": 31.55351107773188,
      "eval_loss": 3.454169988632202,
      "eval_runtime": 4.8006,
      "eval_samples_per_second": 691.578,
      "eval_steps_per_second": 43.328,
      "step": 42000
    },
    {
      "epoch": 33.055576417574166,
      "grad_norm": 4.173329830169678,
      "learning_rate": 9.333333333333334e-05,
      "loss": 3.2522,
      "step": 44000
    },
    {
      "epoch": 33.055576417574166,
      "eval_loss": 3.3299686908721924,
      "eval_runtime": 4.544,
      "eval_samples_per_second": 730.638,
      "eval_steps_per_second": 45.775,
      "step": 44000
    },
    {
      "epoch": 34.55764175741645,
      "eval_loss": 3.2829904556274414,
      "eval_runtime": 4.8645,
      "eval_samples_per_second": 682.495,
      "eval_steps_per_second": 42.759,
      "step": 46000
    },
    {
      "epoch": 36.05970709725873,
      "grad_norm": 4.994029521942139,
      "learning_rate": 8.666666666666667e-05,
      "loss": 3.0484,
      "step": 48000
    },
    {
      "epoch": 36.05970709725873,
      "eval_loss": 3.1864311695098877,
      "eval_runtime": 4.5673,
      "eval_samples_per_second": 726.907,
      "eval_steps_per_second": 45.541,
      "step": 48000
    },
    {
      "epoch": 37.56177243710101,
      "eval_loss": 3.118920087814331,
      "eval_runtime": 4.2694,
      "eval_samples_per_second": 777.63,
      "eval_steps_per_second": 48.719,
      "step": 50000
    },
    {
      "epoch": 39.063837776943295,
      "grad_norm": 4.800024509429932,
      "learning_rate": 8e-05,
      "loss": 2.9026,
      "step": 52000
    },
    {
      "epoch": 39.063837776943295,
      "eval_loss": 3.0475075244903564,
      "eval_runtime": 4.6498,
      "eval_samples_per_second": 714.014,
      "eval_steps_per_second": 44.733,
      "step": 52000
    },
    {
      "epoch": 40.56590311678558,
      "eval_loss": 2.993332624435425,
      "eval_runtime": 4.9159,
      "eval_samples_per_second": 675.36,
      "eval_steps_per_second": 42.312,
      "step": 54000
    },
    {
      "epoch": 42.06796845662787,
      "grad_norm": 4.461485385894775,
      "learning_rate": 7.333333333333333e-05,
      "loss": 2.7874,
      "step": 56000
    },
    {
      "epoch": 42.06796845662787,
      "eval_loss": 2.9410743713378906,
      "eval_runtime": 4.855,
      "eval_samples_per_second": 683.83,
      "eval_steps_per_second": 42.842,
      "step": 56000
    },
    {
      "epoch": 43.57003379647015,
      "eval_loss": 2.935495138168335,
      "eval_runtime": 4.8864,
      "eval_samples_per_second": 679.431,
      "eval_steps_per_second": 42.567,
      "step": 58000
    },
    {
      "epoch": 45.07209913631243,
      "grad_norm": 4.862215518951416,
      "learning_rate": 6.666666666666667e-05,
      "loss": 2.7001,
      "step": 60000
    },
    {
      "epoch": 45.07209913631243,
      "eval_loss": 2.891265630722046,
      "eval_runtime": 4.7977,
      "eval_samples_per_second": 691.997,
      "eval_steps_per_second": 43.354,
      "step": 60000
    },
    {
      "epoch": 46.574164476154714,
      "eval_loss": 2.860131025314331,
      "eval_runtime": 4.8035,
      "eval_samples_per_second": 691.17,
      "eval_steps_per_second": 43.302,
      "step": 62000
    },
    {
      "epoch": 48.076229815996996,
      "grad_norm": 4.176759719848633,
      "learning_rate": 6e-05,
      "loss": 2.6298,
      "step": 64000
    },
    {
      "epoch": 48.076229815996996,
      "eval_loss": 2.8227334022521973,
      "eval_runtime": 4.8405,
      "eval_samples_per_second": 685.878,
      "eval_steps_per_second": 42.971,
      "step": 64000
    },
    {
      "epoch": 49.57829515583928,
      "eval_loss": 2.8202409744262695,
      "eval_runtime": 4.9107,
      "eval_samples_per_second": 676.074,
      "eval_steps_per_second": 42.356,
      "step": 66000
    },
    {
      "epoch": 51.08036049568156,
      "grad_norm": 4.324464321136475,
      "learning_rate": 5.333333333333333e-05,
      "loss": 2.5722,
      "step": 68000
    },
    {
      "epoch": 51.08036049568156,
      "eval_loss": 2.7873833179473877,
      "eval_runtime": 4.9614,
      "eval_samples_per_second": 669.165,
      "eval_steps_per_second": 41.924,
      "step": 68000
    },
    {
      "epoch": 52.58242583552384,
      "eval_loss": 2.7716081142425537,
      "eval_runtime": 4.9279,
      "eval_samples_per_second": 673.719,
      "eval_steps_per_second": 42.209,
      "step": 70000
    },
    {
      "epoch": 54.084491175366125,
      "grad_norm": 4.540297985076904,
      "learning_rate": 4.666666666666667e-05,
      "loss": 2.523,
      "step": 72000
    },
    {
      "epoch": 54.084491175366125,
      "eval_loss": 2.736281633377075,
      "eval_runtime": 5.0438,
      "eval_samples_per_second": 658.231,
      "eval_steps_per_second": 41.239,
      "step": 72000
    },
    {
      "epoch": 55.586556515208414,
      "eval_loss": 2.721235513687134,
      "eval_runtime": 4.9084,
      "eval_samples_per_second": 676.391,
      "eval_steps_per_second": 42.376,
      "step": 74000
    },
    {
      "epoch": 57.0886218550507,
      "grad_norm": 4.172296524047852,
      "learning_rate": 4e-05,
      "loss": 2.4788,
      "step": 76000
    },
    {
      "epoch": 57.0886218550507,
      "eval_loss": 2.6943581104278564,
      "eval_runtime": 4.972,
      "eval_samples_per_second": 667.739,
      "eval_steps_per_second": 41.834,
      "step": 76000
    },
    {
      "epoch": 58.59068719489298,
      "eval_loss": 2.6760997772216797,
      "eval_runtime": 4.8781,
      "eval_samples_per_second": 680.594,
      "eval_steps_per_second": 42.64,
      "step": 78000
    },
    {
      "epoch": 60.09275253473526,
      "grad_norm": 4.14224910736084,
      "learning_rate": 3.3333333333333335e-05,
      "loss": 2.4466,
      "step": 80000
    },
    {
      "epoch": 60.09275253473526,
      "eval_loss": 2.670499563217163,
      "eval_runtime": 4.6467,
      "eval_samples_per_second": 714.491,
      "eval_steps_per_second": 44.763,
      "step": 80000
    },
    {
      "epoch": 61.594817874577544,
      "eval_loss": 2.6551034450531006,
      "eval_runtime": 4.9275,
      "eval_samples_per_second": 673.77,
      "eval_steps_per_second": 42.212,
      "step": 82000
    },
    {
      "epoch": 63.096883214419826,
      "grad_norm": 4.495992183685303,
      "learning_rate": 2.6666666666666667e-05,
      "loss": 2.4122,
      "step": 84000
    },
    {
      "epoch": 63.096883214419826,
      "eval_loss": 2.6368117332458496,
      "eval_runtime": 4.8647,
      "eval_samples_per_second": 682.471,
      "eval_steps_per_second": 42.757,
      "step": 84000
    },
    {
      "epoch": 64.59894855426211,
      "eval_loss": 2.6423897743225098,
      "eval_runtime": 4.8657,
      "eval_samples_per_second": 682.327,
      "eval_steps_per_second": 42.748,
      "step": 86000
    },
    {
      "epoch": 66.1010138941044,
      "grad_norm": 3.998427152633667,
      "learning_rate": 2e-05,
      "loss": 2.3832,
      "step": 88000
    },
    {
      "epoch": 66.1010138941044,
      "eval_loss": 2.6344850063323975,
      "eval_runtime": 4.8088,
      "eval_samples_per_second": 690.4,
      "eval_steps_per_second": 43.254,
      "step": 88000
    }
  ],
  "logging_steps": 4000,
  "max_steps": 100000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 76,
  "save_steps": 4000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 2.7445280879149056e+16,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}