{
  "best_metric": 1.091291069984436,
  "best_model_checkpoint": "vietnamese-emb-long-mlm/checkpoint-8000",
  "epoch": 5.0,
  "eval_steps": 500,
  "global_step": 8400,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.05952380952380952,
      "grad_norm": 7.383795261383057,
      "learning_rate": 4.940476190476191e-05,
      "loss": 6.9284,
      "step": 100
    },
    {
      "epoch": 0.11904761904761904,
      "grad_norm": 11.806465148925781,
      "learning_rate": 4.880952380952381e-05,
      "loss": 5.4103,
      "step": 200
    },
    {
      "epoch": 0.17857142857142858,
      "grad_norm": 10.51978588104248,
      "learning_rate": 4.8214285714285716e-05,
      "loss": 4.45,
      "step": 300
    },
    {
      "epoch": 0.23809523809523808,
      "grad_norm": 10.803540229797363,
      "learning_rate": 4.761904761904762e-05,
      "loss": 3.7495,
      "step": 400
    },
    {
      "epoch": 0.2976190476190476,
      "grad_norm": 10.87264347076416,
      "learning_rate": 4.7023809523809525e-05,
      "loss": 3.3683,
      "step": 500
    },
    {
      "epoch": 0.2976190476190476,
      "eval_loss": 2.989553928375244,
      "eval_runtime": 105.8686,
      "eval_samples_per_second": 112.838,
      "eval_steps_per_second": 1.766,
      "step": 500
    },
    {
      "epoch": 0.35714285714285715,
      "grad_norm": 9.709155082702637,
      "learning_rate": 4.642857142857143e-05,
      "loss": 3.0939,
      "step": 600
    },
    {
      "epoch": 0.4166666666666667,
      "grad_norm": 9.323984146118164,
      "learning_rate": 4.5833333333333334e-05,
      "loss": 2.8534,
      "step": 700
    },
    {
      "epoch": 0.47619047619047616,
      "grad_norm": 10.167367935180664,
      "learning_rate": 4.523809523809524e-05,
      "loss": 2.7091,
      "step": 800
    },
    {
      "epoch": 0.5357142857142857,
      "grad_norm": 10.095072746276855,
      "learning_rate": 4.464285714285715e-05,
      "loss": 2.5711,
      "step": 900
    },
    {
      "epoch": 0.5952380952380952,
      "grad_norm": 9.4666109085083,
      "learning_rate": 4.404761904761905e-05,
      "loss": 2.4663,
      "step": 1000
    },
    {
      "epoch": 0.5952380952380952,
      "eval_loss": 2.2190322875976562,
      "eval_runtime": 106.1834,
      "eval_samples_per_second": 112.504,
      "eval_steps_per_second": 1.761,
      "step": 1000
    },
    {
      "epoch": 0.6547619047619048,
      "grad_norm": 9.085387229919434,
      "learning_rate": 4.345238095238096e-05,
      "loss": 2.4001,
      "step": 1100
    },
    {
      "epoch": 0.7142857142857143,
      "grad_norm": 9.016592979431152,
      "learning_rate": 4.2857142857142856e-05,
      "loss": 2.3036,
      "step": 1200
    },
    {
      "epoch": 0.7738095238095238,
      "grad_norm": 10.062590599060059,
      "learning_rate": 4.226190476190476e-05,
      "loss": 2.2139,
      "step": 1300
    },
    {
      "epoch": 0.8333333333333334,
      "grad_norm": 9.198400497436523,
      "learning_rate": 4.166666666666667e-05,
      "loss": 2.1477,
      "step": 1400
    },
    {
      "epoch": 0.8928571428571429,
      "grad_norm": 9.555102348327637,
      "learning_rate": 4.107142857142857e-05,
      "loss": 2.078,
      "step": 1500
    },
    {
      "epoch": 0.8928571428571429,
      "eval_loss": 1.8398191928863525,
      "eval_runtime": 106.2078,
      "eval_samples_per_second": 112.478,
      "eval_steps_per_second": 1.761,
      "step": 1500
    },
    {
      "epoch": 0.9523809523809523,
      "grad_norm": 9.729957580566406,
      "learning_rate": 4.047619047619048e-05,
      "loss": 2.0146,
      "step": 1600
    },
    {
      "epoch": 1.0119047619047619,
      "grad_norm": 9.68885326385498,
      "learning_rate": 3.9880952380952386e-05,
      "loss": 1.9381,
      "step": 1700
    },
    {
      "epoch": 1.0714285714285714,
      "grad_norm": 8.330371856689453,
      "learning_rate": 3.928571428571429e-05,
      "loss": 1.8773,
      "step": 1800
    },
    {
      "epoch": 1.130952380952381,
      "grad_norm": 9.066886901855469,
      "learning_rate": 3.8690476190476195e-05,
      "loss": 1.8749,
      "step": 1900
    },
    {
      "epoch": 1.1904761904761905,
      "grad_norm": 10.275750160217285,
      "learning_rate": 3.809523809523809e-05,
      "loss": 1.8204,
      "step": 2000
    },
    {
      "epoch": 1.1904761904761905,
      "eval_loss": 1.667816400527954,
      "eval_runtime": 106.1061,
      "eval_samples_per_second": 112.585,
      "eval_steps_per_second": 1.762,
      "step": 2000
    },
    {
      "epoch": 1.25,
      "grad_norm": 9.68263053894043,
      "learning_rate": 3.7500000000000003e-05,
      "loss": 1.8137,
      "step": 2100
    },
    {
      "epoch": 1.3095238095238095,
      "grad_norm": 10.254537582397461,
      "learning_rate": 3.690476190476191e-05,
      "loss": 1.8096,
      "step": 2200
    },
    {
      "epoch": 1.369047619047619,
      "grad_norm": 8.361231803894043,
      "learning_rate": 3.630952380952381e-05,
      "loss": 1.7509,
      "step": 2300
    },
    {
      "epoch": 1.4285714285714286,
      "grad_norm": 9.450984954833984,
      "learning_rate": 3.571428571428572e-05,
      "loss": 1.7277,
      "step": 2400
    },
    {
      "epoch": 1.4880952380952381,
      "grad_norm": 8.3367280960083,
      "learning_rate": 3.511904761904762e-05,
      "loss": 1.7063,
      "step": 2500
    },
    {
      "epoch": 1.4880952380952381,
      "eval_loss": 1.55353844165802,
      "eval_runtime": 106.2169,
      "eval_samples_per_second": 112.468,
      "eval_steps_per_second": 1.761,
      "step": 2500
    },
    {
      "epoch": 1.5476190476190477,
      "grad_norm": 8.55213451385498,
      "learning_rate": 3.4523809523809526e-05,
      "loss": 1.6952,
      "step": 2600
    },
    {
      "epoch": 1.6071428571428572,
      "grad_norm": 8.386811256408691,
      "learning_rate": 3.392857142857143e-05,
      "loss": 1.6654,
      "step": 2700
    },
    {
      "epoch": 1.6666666666666665,
      "grad_norm": 9.841432571411133,
      "learning_rate": 3.3333333333333335e-05,
      "loss": 1.6662,
      "step": 2800
    },
    {
      "epoch": 1.7261904761904763,
      "grad_norm": 8.69044303894043,
      "learning_rate": 3.273809523809524e-05,
      "loss": 1.6291,
      "step": 2900
    },
    {
      "epoch": 1.7857142857142856,
      "grad_norm": 9.23985481262207,
      "learning_rate": 3.2142857142857144e-05,
      "loss": 1.5943,
      "step": 3000
    },
    {
      "epoch": 1.7857142857142856,
      "eval_loss": 1.464246392250061,
      "eval_runtime": 106.2444,
      "eval_samples_per_second": 112.439,
      "eval_steps_per_second": 1.76,
      "step": 3000
    },
    {
      "epoch": 1.8452380952380953,
      "grad_norm": 8.489648818969727,
      "learning_rate": 3.154761904761905e-05,
      "loss": 1.6024,
      "step": 3100
    },
    {
      "epoch": 1.9047619047619047,
      "grad_norm": 8.592900276184082,
      "learning_rate": 3.095238095238095e-05,
      "loss": 1.5711,
      "step": 3200
    },
    {
      "epoch": 1.9642857142857144,
      "grad_norm": 8.73353099822998,
      "learning_rate": 3.0357142857142857e-05,
      "loss": 1.561,
      "step": 3300
    },
    {
      "epoch": 2.0238095238095237,
      "grad_norm": 7.832152843475342,
      "learning_rate": 2.9761904761904762e-05,
      "loss": 1.4996,
      "step": 3400
    },
    {
      "epoch": 2.0833333333333335,
      "grad_norm": 8.37968921661377,
      "learning_rate": 2.916666666666667e-05,
      "loss": 1.5235,
      "step": 3500
    },
    {
      "epoch": 2.0833333333333335,
      "eval_loss": 1.3883955478668213,
      "eval_runtime": 106.3163,
      "eval_samples_per_second": 112.363,
      "eval_steps_per_second": 1.759,
      "step": 3500
    },
    {
      "epoch": 2.142857142857143,
      "grad_norm": 9.074413299560547,
      "learning_rate": 2.857142857142857e-05,
      "loss": 1.4867,
      "step": 3600
    },
    {
      "epoch": 2.2023809523809526,
      "grad_norm": 8.715752601623535,
      "learning_rate": 2.797619047619048e-05,
      "loss": 1.481,
      "step": 3700
    },
    {
      "epoch": 2.261904761904762,
      "grad_norm": 8.675792694091797,
      "learning_rate": 2.7380952380952383e-05,
      "loss": 1.4462,
      "step": 3800
    },
    {
      "epoch": 2.3214285714285716,
      "grad_norm": 8.826884269714355,
      "learning_rate": 2.6785714285714288e-05,
      "loss": 1.449,
      "step": 3900
    },
    {
      "epoch": 2.380952380952381,
      "grad_norm": 8.927966117858887,
      "learning_rate": 2.6190476190476192e-05,
      "loss": 1.4307,
      "step": 4000
    },
    {
      "epoch": 2.380952380952381,
      "eval_loss": 1.3017064332962036,
      "eval_runtime": 106.1765,
      "eval_samples_per_second": 112.511,
      "eval_steps_per_second": 1.761,
      "step": 4000
    },
    {
      "epoch": 2.4404761904761907,
      "grad_norm": 7.1491804122924805,
      "learning_rate": 2.5595238095238093e-05,
      "loss": 1.4789,
      "step": 4100
    },
    {
      "epoch": 2.5,
      "grad_norm": 7.50277853012085,
      "learning_rate": 2.5e-05,
      "loss": 1.4296,
      "step": 4200
    },
    {
      "epoch": 2.5595238095238093,
      "grad_norm": 8.132680892944336,
      "learning_rate": 2.4404761904761906e-05,
      "loss": 1.4177,
      "step": 4300
    },
    {
      "epoch": 2.619047619047619,
      "grad_norm": 8.233142852783203,
      "learning_rate": 2.380952380952381e-05,
      "loss": 1.3953,
      "step": 4400
    },
    {
      "epoch": 2.678571428571429,
      "grad_norm": 7.6400628089904785,
      "learning_rate": 2.3214285714285715e-05,
      "loss": 1.4063,
      "step": 4500
    },
    {
      "epoch": 2.678571428571429,
      "eval_loss": 1.2739932537078857,
      "eval_runtime": 106.1766,
      "eval_samples_per_second": 112.511,
      "eval_steps_per_second": 1.761,
      "step": 4500
    },
    {
      "epoch": 2.738095238095238,
      "grad_norm": 7.416648864746094,
      "learning_rate": 2.261904761904762e-05,
      "loss": 1.4008,
      "step": 4600
    },
    {
      "epoch": 2.7976190476190474,
      "grad_norm": 7.283276081085205,
      "learning_rate": 2.2023809523809524e-05,
      "loss": 1.3818,
      "step": 4700
    },
    {
      "epoch": 2.857142857142857,
      "grad_norm": 8.360782623291016,
      "learning_rate": 2.1428571428571428e-05,
      "loss": 1.379,
      "step": 4800
    },
    {
      "epoch": 2.9166666666666665,
      "grad_norm": 9.960053443908691,
      "learning_rate": 2.0833333333333336e-05,
      "loss": 1.3479,
      "step": 4900
    },
    {
      "epoch": 2.9761904761904763,
      "grad_norm": 8.903584480285645,
      "learning_rate": 2.023809523809524e-05,
      "loss": 1.3389,
      "step": 5000
    },
    {
      "epoch": 2.9761904761904763,
      "eval_loss": 1.2187227010726929,
      "eval_runtime": 106.0745,
      "eval_samples_per_second": 112.619,
      "eval_steps_per_second": 1.763,
      "step": 5000
    },
    {
      "epoch": 3.0357142857142856,
      "grad_norm": 8.299568176269531,
      "learning_rate": 1.9642857142857145e-05,
      "loss": 1.3336,
      "step": 5100
    },
    {
      "epoch": 3.0952380952380953,
      "grad_norm": 8.194002151489258,
      "learning_rate": 1.9047619047619046e-05,
      "loss": 1.3102,
      "step": 5200
    },
    {
      "epoch": 3.1547619047619047,
      "grad_norm": 7.804888725280762,
      "learning_rate": 1.8452380952380954e-05,
      "loss": 1.3354,
      "step": 5300
    },
    {
      "epoch": 3.2142857142857144,
      "grad_norm": 7.510008811950684,
      "learning_rate": 1.785714285714286e-05,
      "loss": 1.3382,
      "step": 5400
    },
    {
      "epoch": 3.2738095238095237,
      "grad_norm": 7.9406023025512695,
      "learning_rate": 1.7261904761904763e-05,
      "loss": 1.2996,
      "step": 5500
    },
    {
      "epoch": 3.2738095238095237,
      "eval_loss": 1.2163845300674438,
      "eval_runtime": 106.2374,
      "eval_samples_per_second": 112.446,
      "eval_steps_per_second": 1.76,
      "step": 5500
    },
    {
      "epoch": 3.3333333333333335,
      "grad_norm": 7.935986042022705,
      "learning_rate": 1.6666666666666667e-05,
      "loss": 1.2889,
      "step": 5600
    },
    {
      "epoch": 3.392857142857143,
      "grad_norm": 7.514946937561035,
      "learning_rate": 1.6071428571428572e-05,
      "loss": 1.3353,
      "step": 5700
    },
    {
      "epoch": 3.4523809523809526,
      "grad_norm": 8.081690788269043,
      "learning_rate": 1.5476190476190476e-05,
      "loss": 1.2959,
      "step": 5800
    },
    {
      "epoch": 3.511904761904762,
      "grad_norm": 8.02477741241455,
      "learning_rate": 1.4880952380952381e-05,
      "loss": 1.2467,
      "step": 5900
    },
    {
      "epoch": 3.571428571428571,
      "grad_norm": 8.462018013000488,
      "learning_rate": 1.4285714285714285e-05,
      "loss": 1.2899,
      "step": 6000
    },
    {
      "epoch": 3.571428571428571,
      "eval_loss": 1.1731750965118408,
      "eval_runtime": 106.1002,
      "eval_samples_per_second": 112.592,
      "eval_steps_per_second": 1.762,
      "step": 6000
    },
    {
      "epoch": 3.630952380952381,
      "grad_norm": 7.650486469268799,
      "learning_rate": 1.3690476190476192e-05,
      "loss": 1.2878,
      "step": 6100
    },
    {
      "epoch": 3.6904761904761907,
      "grad_norm": 8.889339447021484,
      "learning_rate": 1.3095238095238096e-05,
      "loss": 1.2704,
      "step": 6200
    },
    {
      "epoch": 3.75,
      "grad_norm": 9.502975463867188,
      "learning_rate": 1.25e-05,
      "loss": 1.2406,
      "step": 6300
    },
    {
      "epoch": 3.8095238095238093,
      "grad_norm": 9.480072975158691,
      "learning_rate": 1.1904761904761905e-05,
      "loss": 1.2707,
      "step": 6400
    },
    {
      "epoch": 3.869047619047619,
      "grad_norm": 7.378728866577148,
      "learning_rate": 1.130952380952381e-05,
      "loss": 1.2343,
      "step": 6500
    },
    {
      "epoch": 3.869047619047619,
      "eval_loss": 1.1631048917770386,
      "eval_runtime": 106.023,
      "eval_samples_per_second": 112.674,
      "eval_steps_per_second": 1.764,
      "step": 6500
    },
    {
      "epoch": 3.928571428571429,
      "grad_norm": 8.64068603515625,
      "learning_rate": 1.0714285714285714e-05,
      "loss": 1.2478,
      "step": 6600
    },
    {
      "epoch": 3.988095238095238,
      "grad_norm": 8.00733470916748,
      "learning_rate": 1.011904761904762e-05,
      "loss": 1.2414,
      "step": 6700
    },
    {
      "epoch": 4.0476190476190474,
      "grad_norm": 7.568075180053711,
      "learning_rate": 9.523809523809523e-06,
      "loss": 1.2533,
      "step": 6800
    },
    {
      "epoch": 4.107142857142857,
      "grad_norm": 7.190471649169922,
      "learning_rate": 8.92857142857143e-06,
      "loss": 1.2495,
      "step": 6900
    },
    {
      "epoch": 4.166666666666667,
      "grad_norm": 8.588766098022461,
      "learning_rate": 8.333333333333334e-06,
      "loss": 1.2095,
      "step": 7000
    },
    {
      "epoch": 4.166666666666667,
      "eval_loss": 1.0942662954330444,
      "eval_runtime": 106.0638,
      "eval_samples_per_second": 112.63,
      "eval_steps_per_second": 1.763,
      "step": 7000
    },
    {
      "epoch": 4.226190476190476,
      "grad_norm": 8.47437858581543,
      "learning_rate": 7.738095238095238e-06,
      "loss": 1.2134,
      "step": 7100
    },
    {
      "epoch": 4.285714285714286,
      "grad_norm": 7.481777191162109,
      "learning_rate": 7.142857142857143e-06,
      "loss": 1.1873,
      "step": 7200
    },
    {
      "epoch": 4.345238095238095,
      "grad_norm": 7.533321380615234,
      "learning_rate": 6.547619047619048e-06,
      "loss": 1.2022,
      "step": 7300
    },
    {
      "epoch": 4.404761904761905,
      "grad_norm": 8.106829643249512,
      "learning_rate": 5.9523809523809525e-06,
      "loss": 1.2162,
      "step": 7400
    },
    {
      "epoch": 4.464285714285714,
      "grad_norm": 7.893404483795166,
      "learning_rate": 5.357142857142857e-06,
      "loss": 1.2078,
      "step": 7500
    },
    {
      "epoch": 4.464285714285714,
      "eval_loss": 1.1191558837890625,
      "eval_runtime": 106.3099,
      "eval_samples_per_second": 112.37,
      "eval_steps_per_second": 1.759,
      "step": 7500
    },
    {
      "epoch": 4.523809523809524,
      "grad_norm": 8.90477466583252,
      "learning_rate": 4.7619047619047615e-06,
      "loss": 1.219,
      "step": 7600
    },
    {
      "epoch": 4.583333333333333,
      "grad_norm": 7.603702545166016,
      "learning_rate": 4.166666666666667e-06,
      "loss": 1.2,
      "step": 7700
    },
    {
      "epoch": 4.642857142857143,
      "grad_norm": 8.275321006774902,
      "learning_rate": 3.5714285714285714e-06,
      "loss": 1.1985,
      "step": 7800
    },
    {
      "epoch": 4.7023809523809526,
      "grad_norm": 8.652661323547363,
      "learning_rate": 2.9761904761904763e-06,
      "loss": 1.2421,
      "step": 7900
    },
    {
      "epoch": 4.761904761904762,
      "grad_norm": 8.569392204284668,
      "learning_rate": 2.3809523809523808e-06,
      "loss": 1.208,
      "step": 8000
    },
    {
      "epoch": 4.761904761904762,
      "eval_loss": 1.091291069984436,
      "eval_runtime": 106.5156,
      "eval_samples_per_second": 112.153,
      "eval_steps_per_second": 1.756,
      "step": 8000
    },
    {
      "epoch": 4.821428571428571,
      "grad_norm": 8.579488754272461,
      "learning_rate": 1.7857142857142857e-06,
      "loss": 1.2274,
      "step": 8100
    },
    {
      "epoch": 4.880952380952381,
      "grad_norm": 9.347880363464355,
      "learning_rate": 1.1904761904761904e-06,
      "loss": 1.1833,
      "step": 8200
    },
    {
      "epoch": 4.940476190476191,
      "grad_norm": 7.7748894691467285,
      "learning_rate": 5.952380952380952e-07,
      "loss": 1.1919,
      "step": 8300
    },
    {
      "epoch": 5.0,
      "grad_norm": 8.465712547302246,
      "learning_rate": 0.0,
      "loss": 1.2069,
      "step": 8400
    }
  ],
  "logging_steps": 100,
  "max_steps": 8400,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 2000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.5633072459126792e+16,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}