{
  "best_metric": 2.7392032146453857,
  "best_model_checkpoint": "/home/p318482/babyLM_controlled/models_trained_last/de_mlm/de_childes_30/checkpoint-76000",
  "epoch": 57.078482914006756,
  "eval_steps": 2000,
  "global_step": 76000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.5020653398422832,
      "eval_loss": 7.472109317779541,
      "eval_runtime": 2.5707,
      "eval_samples_per_second": 1291.498,
      "eval_steps_per_second": 80.913,
      "step": 2000
    },
    {
      "epoch": 3.0041306796845664,
      "grad_norm": 0.9529591798782349,
      "learning_rate": 1e-05,
      "loss": 7.3851,
      "step": 4000
    },
    {
      "epoch": 3.0041306796845664,
      "eval_loss": 6.4175896644592285,
      "eval_runtime": 2.5761,
      "eval_samples_per_second": 1288.76,
      "eval_steps_per_second": 80.742,
      "step": 4000
    },
    {
      "epoch": 4.50619601952685,
      "eval_loss": 6.3026041984558105,
      "eval_runtime": 2.5876,
      "eval_samples_per_second": 1283.023,
      "eval_steps_per_second": 80.382,
      "step": 6000
    },
    {
      "epoch": 6.008261359369133,
      "grad_norm": 1.5569523572921753,
      "learning_rate": 2e-05,
      "loss": 6.0706,
      "step": 8000
    },
    {
      "epoch": 6.008261359369133,
      "eval_loss": 6.200850486755371,
      "eval_runtime": 2.6505,
      "eval_samples_per_second": 1252.599,
      "eval_steps_per_second": 78.476,
      "step": 8000
    },
    {
      "epoch": 7.510326699211416,
      "eval_loss": 6.105844497680664,
      "eval_runtime": 2.5888,
      "eval_samples_per_second": 1282.425,
      "eval_steps_per_second": 80.345,
      "step": 10000
    },
    {
      "epoch": 9.0123920390537,
      "grad_norm": 2.3049111366271973,
      "learning_rate": 3e-05,
      "loss": 5.8733,
      "step": 12000
    },
    {
      "epoch": 9.0123920390537,
      "eval_loss": 6.0330352783203125,
      "eval_runtime": 2.5874,
      "eval_samples_per_second": 1283.132,
      "eval_steps_per_second": 80.389,
      "step": 12000
    },
    {
      "epoch": 10.514457378895981,
      "eval_loss": 5.968105792999268,
      "eval_runtime": 2.6066,
      "eval_samples_per_second": 1273.702,
      "eval_steps_per_second": 79.798,
      "step": 14000
    },
    {
      "epoch": 12.016522718738265,
      "grad_norm": 2.316237688064575,
      "learning_rate": 4e-05,
      "loss": 5.723,
      "step": 16000
    },
    {
      "epoch": 12.016522718738265,
      "eval_loss": 5.878941535949707,
      "eval_runtime": 2.5723,
      "eval_samples_per_second": 1290.696,
      "eval_steps_per_second": 80.863,
      "step": 16000
    },
    {
      "epoch": 13.518588058580548,
      "eval_loss": 5.8198137283325195,
      "eval_runtime": 2.5819,
      "eval_samples_per_second": 1285.857,
      "eval_steps_per_second": 80.56,
      "step": 18000
    },
    {
      "epoch": 15.020653398422832,
      "grad_norm": 2.299781322479248,
      "learning_rate": 5e-05,
      "loss": 5.6127,
      "step": 20000
    },
    {
      "epoch": 15.020653398422832,
      "eval_loss": 5.802139759063721,
      "eval_runtime": 2.6545,
      "eval_samples_per_second": 1250.713,
      "eval_steps_per_second": 78.358,
      "step": 20000
    },
    {
      "epoch": 16.522718738265116,
      "eval_loss": 5.7662129402160645,
      "eval_runtime": 2.5877,
      "eval_samples_per_second": 1282.983,
      "eval_steps_per_second": 80.38,
      "step": 22000
    },
    {
      "epoch": 18.0247840781074,
      "grad_norm": 2.8674566745758057,
      "learning_rate": 6e-05,
      "loss": 5.5325,
      "step": 24000
    },
    {
      "epoch": 18.0247840781074,
      "eval_loss": 5.731865406036377,
      "eval_runtime": 2.5813,
      "eval_samples_per_second": 1286.16,
      "eval_steps_per_second": 80.579,
      "step": 24000
    },
    {
      "epoch": 19.52684941794968,
      "eval_loss": 5.703192234039307,
      "eval_runtime": 2.5846,
      "eval_samples_per_second": 1284.544,
      "eval_steps_per_second": 80.477,
      "step": 26000
    },
    {
      "epoch": 21.028914757791963,
      "grad_norm": 4.573985576629639,
      "learning_rate": 7e-05,
      "loss": 5.4632,
      "step": 28000
    },
    {
      "epoch": 21.028914757791963,
      "eval_loss": 5.70460319519043,
      "eval_runtime": 2.5824,
      "eval_samples_per_second": 1285.631,
      "eval_steps_per_second": 80.546,
      "step": 28000
    },
    {
      "epoch": 22.53098009763425,
      "eval_loss": 5.570806503295898,
      "eval_runtime": 2.5775,
      "eval_samples_per_second": 1288.087,
      "eval_steps_per_second": 80.699,
      "step": 30000
    },
    {
      "epoch": 24.03304543747653,
      "grad_norm": 5.407034397125244,
      "learning_rate": 8e-05,
      "loss": 5.2633,
      "step": 32000
    },
    {
      "epoch": 24.03304543747653,
      "eval_loss": 4.99446439743042,
      "eval_runtime": 2.5969,
      "eval_samples_per_second": 1278.448,
      "eval_steps_per_second": 80.096,
      "step": 32000
    },
    {
      "epoch": 25.535110777318813,
      "eval_loss": 4.523237705230713,
      "eval_runtime": 2.5888,
      "eval_samples_per_second": 1282.462,
      "eval_steps_per_second": 80.347,
      "step": 34000
    },
    {
      "epoch": 27.037176117161096,
      "grad_norm": 5.287095069885254,
      "learning_rate": 9e-05,
      "loss": 4.3386,
      "step": 36000
    },
    {
      "epoch": 27.037176117161096,
      "eval_loss": 4.140077590942383,
      "eval_runtime": 2.5851,
      "eval_samples_per_second": 1284.299,
      "eval_steps_per_second": 80.462,
      "step": 36000
    },
    {
      "epoch": 28.539241457003378,
      "eval_loss": 3.857847213745117,
      "eval_runtime": 2.6038,
      "eval_samples_per_second": 1275.044,
      "eval_steps_per_second": 79.882,
      "step": 38000
    },
    {
      "epoch": 30.041306796845664,
      "grad_norm": 4.30466890335083,
      "learning_rate": 0.0001,
      "loss": 3.6823,
      "step": 40000
    },
    {
      "epoch": 30.041306796845664,
      "eval_loss": 3.6678993701934814,
      "eval_runtime": 2.7074,
      "eval_samples_per_second": 1226.275,
      "eval_steps_per_second": 76.827,
      "step": 40000
    },
    {
      "epoch": 31.543372136687946,
      "eval_loss": 3.5396695137023926,
      "eval_runtime": 4.9347,
      "eval_samples_per_second": 672.792,
      "eval_steps_per_second": 42.151,
      "step": 42000
    },
    {
      "epoch": 33.04543747653023,
      "grad_norm": 4.037199974060059,
      "learning_rate": 9.333333333333334e-05,
      "loss": 3.3424,
      "step": 44000
    },
    {
      "epoch": 33.04543747653023,
      "eval_loss": 3.4012534618377686,
      "eval_runtime": 4.8721,
      "eval_samples_per_second": 681.437,
      "eval_steps_per_second": 42.692,
      "step": 44000
    },
    {
      "epoch": 34.547502816372514,
      "eval_loss": 3.315351963043213,
      "eval_runtime": 4.9228,
      "eval_samples_per_second": 674.419,
      "eval_steps_per_second": 42.253,
      "step": 46000
    },
    {
      "epoch": 36.0495681562148,
      "grad_norm": 5.304887771606445,
      "learning_rate": 8.666666666666667e-05,
      "loss": 3.1266,
      "step": 48000
    },
    {
      "epoch": 36.0495681562148,
      "eval_loss": 3.2563161849975586,
      "eval_runtime": 4.958,
      "eval_samples_per_second": 669.632,
      "eval_steps_per_second": 41.953,
      "step": 48000
    },
    {
      "epoch": 37.55163349605708,
      "eval_loss": 3.1665167808532715,
      "eval_runtime": 4.695,
      "eval_samples_per_second": 707.142,
      "eval_steps_per_second": 44.303,
      "step": 50000
    },
    {
      "epoch": 39.05369883589936,
      "grad_norm": 4.668087005615234,
      "learning_rate": 8e-05,
      "loss": 2.9693,
      "step": 52000
    },
    {
      "epoch": 39.05369883589936,
      "eval_loss": 3.1341300010681152,
      "eval_runtime": 3.8879,
      "eval_samples_per_second": 853.931,
      "eval_steps_per_second": 53.499,
      "step": 52000
    },
    {
      "epoch": 40.55576417574164,
      "eval_loss": 3.0564303398132324,
      "eval_runtime": 4.2699,
      "eval_samples_per_second": 777.528,
      "eval_steps_per_second": 48.713,
      "step": 54000
    },
    {
      "epoch": 42.057829515583926,
      "grad_norm": 4.504222393035889,
      "learning_rate": 7.333333333333333e-05,
      "loss": 2.8544,
      "step": 56000
    },
    {
      "epoch": 42.057829515583926,
      "eval_loss": 3.0329349040985107,
      "eval_runtime": 4.8343,
      "eval_samples_per_second": 686.764,
      "eval_steps_per_second": 43.026,
      "step": 56000
    },
    {
      "epoch": 43.55989485542621,
      "eval_loss": 2.9552295207977295,
      "eval_runtime": 4.8628,
      "eval_samples_per_second": 682.729,
      "eval_steps_per_second": 42.773,
      "step": 58000
    },
    {
      "epoch": 45.0619601952685,
      "grad_norm": 4.24030065536499,
      "learning_rate": 6.666666666666667e-05,
      "loss": 2.758,
      "step": 60000
    },
    {
      "epoch": 45.0619601952685,
      "eval_loss": 2.9491591453552246,
      "eval_runtime": 4.8121,
      "eval_samples_per_second": 689.932,
      "eval_steps_per_second": 43.225,
      "step": 60000
    },
    {
      "epoch": 46.56402553511078,
      "eval_loss": 2.8936784267425537,
      "eval_runtime": 4.9154,
      "eval_samples_per_second": 675.43,
      "eval_steps_per_second": 42.316,
      "step": 62000
    },
    {
      "epoch": 48.06609087495306,
      "grad_norm": 4.06797456741333,
      "learning_rate": 6e-05,
      "loss": 2.684,
      "step": 64000
    },
    {
      "epoch": 48.06609087495306,
      "eval_loss": 2.866178512573242,
      "eval_runtime": 4.8779,
      "eval_samples_per_second": 680.615,
      "eval_steps_per_second": 42.641,
      "step": 64000
    },
    {
      "epoch": 49.568156214795344,
      "eval_loss": 2.8494906425476074,
      "eval_runtime": 4.8964,
      "eval_samples_per_second": 678.05,
      "eval_steps_per_second": 42.48,
      "step": 66000
    },
    {
      "epoch": 51.07022155463763,
      "grad_norm": 4.346592903137207,
      "learning_rate": 5.333333333333333e-05,
      "loss": 2.6223,
      "step": 68000
    },
    {
      "epoch": 51.07022155463763,
      "eval_loss": 2.825268507003784,
      "eval_runtime": 4.8304,
      "eval_samples_per_second": 687.313,
      "eval_steps_per_second": 43.061,
      "step": 68000
    },
    {
      "epoch": 52.57228689447991,
      "eval_loss": 2.78464937210083,
      "eval_runtime": 4.904,
      "eval_samples_per_second": 676.993,
      "eval_steps_per_second": 42.414,
      "step": 70000
    },
    {
      "epoch": 54.07435223432219,
      "grad_norm": 4.131275177001953,
      "learning_rate": 4.666666666666667e-05,
      "loss": 2.5704,
      "step": 72000
    },
    {
      "epoch": 54.07435223432219,
      "eval_loss": 2.7910172939300537,
      "eval_runtime": 4.8745,
      "eval_samples_per_second": 681.097,
      "eval_steps_per_second": 42.671,
      "step": 72000
    },
    {
      "epoch": 55.57641757416447,
      "eval_loss": 2.750303030014038,
      "eval_runtime": 2.5866,
      "eval_samples_per_second": 1283.529,
      "eval_steps_per_second": 80.414,
      "step": 74000
    },
    {
      "epoch": 57.078482914006756,
      "grad_norm": 4.2763261795043945,
      "learning_rate": 4e-05,
      "loss": 2.5235,
      "step": 76000
    },
    {
      "epoch": 57.078482914006756,
      "eval_loss": 2.7392032146453857,
      "eval_runtime": 4.7672,
      "eval_samples_per_second": 696.429,
      "eval_steps_per_second": 43.632,
      "step": 76000
    }
  ],
  "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": 2.370274257744691e+16,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}