{
  "best_global_step": 18000,
  "best_metric": 4.319030284881592,
  "best_model_checkpoint": "/kaggle/working/MASRIBERTv3_ckpts/checkpoint-18000",
  "epoch": 2.705625916043444,
  "eval_steps": 1000,
  "global_step": 18000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.03758126949528355,
      "grad_norm": 104.15485382080078,
      "learning_rate": 4.680451127819549e-06,
      "loss": 82.7376953125,
      "step": 250
    },
    {
      "epoch": 0.0751625389905671,
      "grad_norm": 95.12043762207031,
      "learning_rate": 9.3796992481203e-06,
      "loss": 71.7012734375,
      "step": 500
    },
    {
      "epoch": 0.11274380848585065,
      "grad_norm": 104.137451171875,
      "learning_rate": 1.4078947368421053e-05,
      "loss": 64.50540625,
      "step": 750
    },
    {
      "epoch": 0.1503250779811342,
      "grad_norm": 100.89773559570312,
      "learning_rate": 1.8778195488721807e-05,
      "loss": 61.0135390625,
      "step": 1000
    },
    {
      "epoch": 0.1503250779811342,
      "eval_loss": 7.187783241271973,
      "eval_runtime": 245.496,
      "eval_samples_per_second": 70.07,
      "eval_steps_per_second": 1.096,
      "step": 1000
    },
    {
      "epoch": 0.18790634747641774,
      "grad_norm": 108.36956024169922,
      "learning_rate": 2.3477443609022555e-05,
      "loss": 58.785421875,
      "step": 1250
    },
    {
      "epoch": 0.2254876169717013,
      "grad_norm": 111.00702667236328,
      "learning_rate": 2.8176691729323312e-05,
      "loss": 56.1699921875,
      "step": 1500
    },
    {
      "epoch": 0.26306888646698484,
      "grad_norm": 102.0089111328125,
      "learning_rate": 3.287593984962406e-05,
      "loss": 54.32422265625,
      "step": 1750
    },
    {
      "epoch": 0.3006501559622684,
      "grad_norm": 107.5756607055664,
      "learning_rate": 3.7575187969924814e-05,
      "loss": 52.7279453125,
      "step": 2000
    },
    {
      "epoch": 0.3006501559622684,
      "eval_loss": 6.232546329498291,
      "eval_runtime": 245.4165,
      "eval_samples_per_second": 70.093,
      "eval_steps_per_second": 1.096,
      "step": 2000
    },
    {
      "epoch": 0.33823142545755197,
      "grad_norm": 97.99393463134766,
      "learning_rate": 4.2274436090225565e-05,
      "loss": 51.351296875,
      "step": 2250
    },
    {
      "epoch": 0.3758126949528355,
      "grad_norm": 95.21705627441406,
      "learning_rate": 4.697368421052632e-05,
      "loss": 50.12950390625,
      "step": 2500
    },
    {
      "epoch": 0.41339396444811904,
      "grad_norm": 109.60403442382812,
      "learning_rate": 4.999829666027942e-05,
      "loss": 49.0492109375,
      "step": 2750
    },
    {
      "epoch": 0.4509752339434026,
      "grad_norm": 94.4717025756836,
      "learning_rate": 4.997529106447553e-05,
      "loss": 48.00233984375,
      "step": 3000
    },
    {
      "epoch": 0.4509752339434026,
      "eval_loss": 5.684965133666992,
      "eval_runtime": 245.4233,
      "eval_samples_per_second": 70.091,
      "eval_steps_per_second": 1.096,
      "step": 3000
    },
    {
      "epoch": 0.4885565034386862,
      "grad_norm": 100.90162658691406,
      "learning_rate": 4.9925434012594344e-05,
      "loss": 46.94662109375,
      "step": 3250
    },
    {
      "epoch": 0.5261377729339697,
      "grad_norm": 97.686767578125,
      "learning_rate": 4.9848779106991685e-05,
      "loss": 46.1941953125,
      "step": 3500
    },
    {
      "epoch": 0.5637190424292533,
      "grad_norm": 85.13678741455078,
      "learning_rate": 4.974540876095457e-05,
      "loss": 45.45590234375,
      "step": 3750
    },
    {
      "epoch": 0.6013003119245368,
      "grad_norm": 88.05296325683594,
      "learning_rate": 4.9615434110096976e-05,
      "loss": 44.72342578125,
      "step": 4000
    },
    {
      "epoch": 0.6013003119245368,
      "eval_loss": 5.337510108947754,
      "eval_runtime": 245.395,
      "eval_samples_per_second": 70.099,
      "eval_steps_per_second": 1.096,
      "step": 4000
    },
    {
      "epoch": 0.6388815814198203,
      "grad_norm": 85.22454833984375,
      "learning_rate": 4.6920337276199595e-05,
      "loss": 44.36359375,
      "step": 4250
    },
    {
      "epoch": 0.6764628509151039,
      "grad_norm": 89.28252410888672,
      "learning_rate": 4.6556074961301624e-05,
      "loss": 43.84180859375,
      "step": 4500
    },
    {
      "epoch": 0.7140441204103875,
      "grad_norm": 88.03447723388672,
      "learning_rate": 4.617303838350222e-05,
      "loss": 43.3140625,
      "step": 4750
    },
    {
      "epoch": 0.751625389905671,
      "grad_norm": 82.08790588378906,
      "learning_rate": 4.57715611484984e-05,
      "loss": 42.943515625,
      "step": 5000
    },
    {
      "epoch": 0.751625389905671,
      "eval_loss": 5.109297752380371,
      "eval_runtime": 246.4709,
      "eval_samples_per_second": 69.793,
      "eval_steps_per_second": 1.091,
      "step": 5000
    },
    {
      "epoch": 0.7892066594009546,
      "grad_norm": 86.16392517089844,
      "learning_rate": 4.535199292287709e-05,
      "loss": 42.626734375,
      "step": 5250
    },
    {
      "epoch": 0.8267879288962381,
      "grad_norm": 87.34744262695312,
      "learning_rate": 4.491469912957297e-05,
      "loss": 42.33,
      "step": 5500
    },
    {
      "epoch": 0.8643691983915217,
      "grad_norm": 79.0370101928711,
      "learning_rate": 4.4460060629603434e-05,
      "loss": 41.88755078125,
      "step": 5750
    },
    {
      "epoch": 0.9019504678868052,
      "grad_norm": 83.2159423828125,
      "learning_rate": 4.398847339035767e-05,
      "loss": 41.65294140625,
      "step": 6000
    },
    {
      "epoch": 0.9019504678868052,
      "eval_loss": 4.957474231719971,
      "eval_runtime": 239.2537,
      "eval_samples_per_second": 71.899,
      "eval_steps_per_second": 1.124,
      "step": 6000
    },
    {
      "epoch": 0.9395317373820887,
      "grad_norm": 82.08615112304688,
      "learning_rate": 4.350034814072887e-05,
      "loss": 41.38473828125,
      "step": 6250
    },
    {
      "epoch": 0.9771130068773723,
      "grad_norm": 86.85932922363281,
      "learning_rate": 4.299611001338994e-05,
      "loss": 41.04867578125,
      "step": 6500
    },
    {
      "epoch": 1.01458153256417,
      "grad_norm": 91.38688659667969,
      "learning_rate": 4.2476198174524166e-05,
      "loss": 40.55618359375,
      "step": 6750
    },
    {
      "epoch": 1.0521628020594536,
      "grad_norm": 87.19780731201172,
      "learning_rate": 4.194106544133346e-05,
      "loss": 40.355609375,
      "step": 7000
    },
    {
      "epoch": 1.0521628020594536,
      "eval_loss": 4.834552764892578,
      "eval_runtime": 239.8672,
      "eval_samples_per_second": 71.715,
      "eval_steps_per_second": 1.121,
      "step": 7000
    },
    {
      "epoch": 1.089744071554737,
      "grad_norm": 90.9477767944336,
      "learning_rate": 4.1391177887657266e-05,
      "loss": 40.2546953125,
      "step": 7250
    },
    {
      "epoch": 1.1273253410500206,
      "grad_norm": 77.87168884277344,
      "learning_rate": 4.08270144380455e-05,
      "loss": 39.96091015625,
      "step": 7500
    },
    {
      "epoch": 1.1649066105453043,
      "grad_norm": 90.72994995117188,
      "learning_rate": 4.024906645063925e-05,
      "loss": 39.95100390625,
      "step": 7750
    },
    {
      "epoch": 1.2024878800405878,
      "grad_norm": 78.02538299560547,
      "learning_rate": 3.965783728922245e-05,
      "loss": 39.549375,
      "step": 8000
    },
    {
      "epoch": 1.2024878800405878,
      "eval_loss": 4.758424758911133,
      "eval_runtime": 238.4738,
      "eval_samples_per_second": 72.134,
      "eval_steps_per_second": 1.128,
      "step": 8000
    },
    {
      "epoch": 1.2400691495358713,
      "grad_norm": 94.54589080810547,
      "learning_rate": 3.905384188481714e-05,
      "loss": 39.47437890625,
      "step": 8250
    },
    {
      "epoch": 1.2776504190311548,
      "grad_norm": 84.85308837890625,
      "learning_rate": 3.843760628720439e-05,
      "loss": 39.3814765625,
      "step": 8500
    },
    {
      "epoch": 1.3152316885264383,
      "grad_norm": 80.48727416992188,
      "learning_rate": 3.7809667206761164e-05,
      "loss": 39.18900390625,
      "step": 8750
    },
    {
      "epoch": 1.352812958021722,
      "grad_norm": 79.16352844238281,
      "learning_rate": 3.717057154701251e-05,
      "loss": 39.15230859375,
      "step": 9000
    },
    {
      "epoch": 1.352812958021722,
      "eval_loss": 4.669365406036377,
      "eval_runtime": 240.8713,
      "eval_samples_per_second": 71.416,
      "eval_steps_per_second": 1.117,
      "step": 9000
    },
    {
      "epoch": 1.3903942275170056,
      "grad_norm": 81.47567749023438,
      "learning_rate": 3.6520875928305904e-05,
      "loss": 38.7915625,
      "step": 9250
    },
    {
      "epoch": 1.427975497012289,
      "grad_norm": 79.67969512939453,
      "learning_rate": 3.58611462030228e-05,
      "loss": 38.70076953125,
      "step": 9500
    },
    {
      "epoch": 1.4655567665075726,
      "grad_norm": 78.06887817382812,
      "learning_rate": 3.5191956962749483e-05,
      "loss": 38.5454296875,
      "step": 9750
    },
    {
      "epoch": 1.5031380360028561,
      "grad_norm": 81.57176971435547,
      "learning_rate": 3.4513891037836586e-05,
      "loss": 38.56761328125,
      "step": 10000
    },
    {
      "epoch": 1.5031380360028561,
      "eval_loss": 4.5992112159729,
      "eval_runtime": 240.6841,
      "eval_samples_per_second": 71.471,
      "eval_steps_per_second": 1.118,
      "step": 10000
    },
    {
      "epoch": 1.5407193054981398,
      "grad_norm": 76.18035888671875,
      "learning_rate": 3.382753898978297e-05,
      "loss": 38.34355859375,
      "step": 10250
    },
    {
      "epoch": 1.5783005749934231,
      "grad_norm": 77.42598724365234,
      "learning_rate": 3.313349859688622e-05,
      "loss": 38.3613984375,
      "step": 10500
    },
    {
      "epoch": 1.6158818444887069,
      "grad_norm": 83.92522430419922,
      "learning_rate": 3.2432374333607674e-05,
      "loss": 38.15187890625,
      "step": 10750
    },
    {
      "epoch": 1.6534631139839904,
      "grad_norm": 80.59062957763672,
      "learning_rate": 3.172477684410535e-05,
      "loss": 38.12749609375,
      "step": 11000
    },
    {
      "epoch": 1.6534631139839904,
      "eval_loss": 4.554149150848389,
      "eval_runtime": 240.8324,
      "eval_samples_per_second": 71.427,
      "eval_steps_per_second": 1.117,
      "step": 11000
    },
    {
      "epoch": 1.6910443834792739,
      "grad_norm": 84.59385681152344,
      "learning_rate": 3.1011322410393507e-05,
      "loss": 38.04288671875,
      "step": 11250
    },
    {
      "epoch": 1.7286256529745576,
      "grad_norm": 78.90901947021484,
      "learning_rate": 3.0292632415591782e-05,
      "loss": 37.87243359375,
      "step": 11500
    },
    {
      "epoch": 1.766206922469841,
      "grad_norm": 82.36697387695312,
      "learning_rate": 2.95693328027316e-05,
      "loss": 37.75355859375,
      "step": 11750
    },
    {
      "epoch": 1.8037881919651246,
      "grad_norm": 75.0630874633789,
      "learning_rate": 2.884205352959113e-05,
      "loss": 37.58091796875,
      "step": 12000
    },
    {
      "epoch": 1.8037881919651246,
      "eval_loss": 4.493417263031006,
      "eval_runtime": 240.8882,
      "eval_samples_per_second": 71.411,
      "eval_steps_per_second": 1.117,
      "step": 12000
    },
    {
      "epoch": 1.8413694614604081,
      "grad_norm": 79.44371032714844,
      "learning_rate": 2.8111428020033515e-05,
      "loss": 37.55409375,
      "step": 12250
    },
    {
      "epoch": 1.8789507309556917,
      "grad_norm": 74.53401947021484,
      "learning_rate": 2.7378092612326463e-05,
      "loss": 37.37219140625,
      "step": 12500
    },
    {
      "epoch": 1.9165320004509754,
      "grad_norm": 84.87117767333984,
      "learning_rate": 2.6642686004923362e-05,
      "loss": 37.2856171875,
      "step": 12750
    },
    {
      "epoch": 1.9541132699462587,
      "grad_norm": 82.84407806396484,
      "learning_rate": 2.590584870018894e-05,
      "loss": 37.37483984375,
      "step": 13000
    },
    {
      "epoch": 1.9541132699462587,
      "eval_loss": 4.470274925231934,
      "eval_runtime": 211.1559,
      "eval_samples_per_second": 81.466,
      "eval_steps_per_second": 1.274,
      "step": 13000
    },
    {
      "epoch": 1.9916945394415424,
      "grad_norm": 77.57933807373047,
      "learning_rate": 2.5168222446553714e-05,
      "loss": 37.2361953125,
      "step": 13250
    },
    {
      "epoch": 2.02916306512834,
      "grad_norm": 89.35295104980469,
      "learning_rate": 2.443044967958324e-05,
      "loss": 36.6822734375,
      "step": 13500
    },
    {
      "epoch": 2.0667443346236234,
      "grad_norm": 79.35661315917969,
      "learning_rate": 2.3693172962448842e-05,
      "loss": 36.6494375,
      "step": 13750
    },
    {
      "epoch": 2.104325604118907,
      "grad_norm": 78.58153533935547,
      "learning_rate": 2.295703442628727e-05,
      "loss": 36.6395859375,
      "step": 14000
    },
    {
      "epoch": 2.104325604118907,
      "eval_loss": 4.428871154785156,
      "eval_runtime": 210.9857,
      "eval_samples_per_second": 81.532,
      "eval_steps_per_second": 1.275,
      "step": 14000
    },
    {
      "epoch": 2.141906873614191,
      "grad_norm": 83.40979766845703,
      "learning_rate": 2.2222675210936572e-05,
      "loss": 36.75696484375,
      "step": 14250
    },
    {
      "epoch": 2.179488143109474,
      "grad_norm": 82.87840270996094,
      "learning_rate": 2.14907349065354e-05,
      "loss": 36.61794921875,
      "step": 14500
    },
    {
      "epoch": 2.217069412604758,
      "grad_norm": 80.84054565429688,
      "learning_rate": 2.076185099647196e-05,
      "loss": 36.6228125,
      "step": 14750
    },
    {
      "epoch": 2.254650682100041,
      "grad_norm": 78.66047668457031,
      "learning_rate": 2.00366583021679e-05,
      "loss": 36.2935,
      "step": 15000
    },
    {
      "epoch": 2.254650682100041,
      "eval_loss": 4.390705585479736,
      "eval_runtime": 210.9905,
      "eval_samples_per_second": 81.53,
      "eval_steps_per_second": 1.275,
      "step": 15000
    },
    {
      "epoch": 2.292231951595325,
      "grad_norm": 75.72366333007812,
      "learning_rate": 1.931578843018062e-05,
      "loss": 36.5094375,
      "step": 15250
    },
    {
      "epoch": 2.3298132210906086,
      "grad_norm": 91.38578033447266,
      "learning_rate": 1.859986922210552e-05,
      "loss": 36.31355078125,
      "step": 15500
    },
    {
      "epoch": 2.367394490585892,
      "grad_norm": 78.77313995361328,
      "learning_rate": 1.78895242077574e-05,
      "loss": 36.16515234375,
      "step": 15750
    },
    {
      "epoch": 2.4049757600811756,
      "grad_norm": 84.32102966308594,
      "learning_rate": 1.7185372062107203e-05,
      "loss": 36.3475625,
      "step": 16000
    },
    {
      "epoch": 2.4049757600811756,
      "eval_loss": 4.355281829833984,
      "eval_runtime": 211.2504,
      "eval_samples_per_second": 81.429,
      "eval_steps_per_second": 1.273,
      "step": 16000
    },
    {
      "epoch": 2.442557029576459,
      "grad_norm": 80.85596466064453,
      "learning_rate": 1.6488026066447024e-05,
      "loss": 36.123234375,
      "step": 16250
    },
    {
      "epoch": 2.4801382990717427,
      "grad_norm": 84.26553344726562,
      "learning_rate": 1.5798093574252835e-05,
      "loss": 36.04619140625,
      "step": 16500
    },
    {
      "epoch": 2.517719568567026,
      "grad_norm": 83.43304443359375,
      "learning_rate": 1.5116175482209957e-05,
      "loss": 35.89179296875,
      "step": 16750
    },
    {
      "epoch": 2.5553008380623097,
      "grad_norm": 83.38468170166016,
      "learning_rate": 1.4442865706862096e-05,
      "loss": 36.0048515625,
      "step": 17000
    },
    {
      "epoch": 2.5553008380623097,
      "eval_loss": 4.32999324798584,
      "eval_runtime": 232.6354,
      "eval_samples_per_second": 73.944,
      "eval_steps_per_second": 1.156,
      "step": 17000
    },
    {
      "epoch": 2.5928821075575934,
      "grad_norm": 94.61381530761719,
      "learning_rate": 1.377875066733979e-05,
      "loss": 35.8738515625,
      "step": 17250
    },
    {
      "epoch": 2.6304633770528767,
      "grad_norm": 85.54600524902344,
      "learning_rate": 1.3124408774618657e-05,
      "loss": 35.88093359375,
      "step": 17500
    },
    {
      "epoch": 2.6680446465481604,
      "grad_norm": 83.1180191040039,
      "learning_rate": 1.2480409927752376e-05,
      "loss": 35.8044375,
      "step": 17750
    },
    {
      "epoch": 2.705625916043444,
      "grad_norm": 80.10919189453125,
      "learning_rate": 1.1847315017519101e-05,
      "loss": 35.79001171875,
      "step": 18000
    },
    {
      "epoch": 2.705625916043444,
      "eval_loss": 4.319030284881592,
      "eval_runtime": 232.6024,
      "eval_samples_per_second": 73.955,
      "eval_steps_per_second": 1.156,
      "step": 18000
    }
  ],
  "logging_steps": 250,
  "max_steps": 26612,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 4,
  "save_steps": 1000,
  "stateful_callbacks": {
    "EarlyStoppingCallback": {
      "args": {
        "early_stopping_patience": 3,
        "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": false
      },
      "attributes": {}
    }
  },
  "total_flos": 3.035725468495534e+17,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": null
}