{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 4.081632653061225,
  "eval_steps": 500,
  "global_step": 1000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.04081632653061224,
      "grad_norm": 7.114395618438721,
      "learning_rate": 9.981632653061225e-06,
      "loss": 0.7362,
      "step": 10
    },
    {
      "epoch": 0.08163265306122448,
      "grad_norm": 11.572301864624023,
      "learning_rate": 9.961224489795919e-06,
      "loss": 0.8729,
      "step": 20
    },
    {
      "epoch": 0.12244897959183673,
      "grad_norm": 9.383491516113281,
      "learning_rate": 9.940816326530614e-06,
      "loss": 0.773,
      "step": 30
    },
    {
      "epoch": 0.16326530612244897,
      "grad_norm": 7.83120059967041,
      "learning_rate": 9.920408163265307e-06,
      "loss": 0.7817,
      "step": 40
    },
    {
      "epoch": 0.20408163265306123,
      "grad_norm": 10.92087173461914,
      "learning_rate": 9.9e-06,
      "loss": 0.6256,
      "step": 50
    },
    {
      "epoch": 0.24489795918367346,
      "grad_norm": 3.8826725482940674,
      "learning_rate": 9.879591836734695e-06,
      "loss": 0.5759,
      "step": 60
    },
    {
      "epoch": 0.2857142857142857,
      "grad_norm": 11.15483283996582,
      "learning_rate": 9.859183673469388e-06,
      "loss": 0.7333,
      "step": 70
    },
    {
      "epoch": 0.32653061224489793,
      "grad_norm": 11.470726013183594,
      "learning_rate": 9.838775510204083e-06,
      "loss": 0.5943,
      "step": 80
    },
    {
      "epoch": 0.3673469387755102,
      "grad_norm": 13.159674644470215,
      "learning_rate": 9.818367346938777e-06,
      "loss": 0.7804,
      "step": 90
    },
    {
      "epoch": 0.40816326530612246,
      "grad_norm": 9.58558464050293,
      "learning_rate": 9.79795918367347e-06,
      "loss": 0.6491,
      "step": 100
    },
    {
      "epoch": 0.4489795918367347,
      "grad_norm": 9.653897285461426,
      "learning_rate": 9.777551020408163e-06,
      "loss": 0.5919,
      "step": 110
    },
    {
      "epoch": 0.4897959183673469,
      "grad_norm": 8.117432594299316,
      "learning_rate": 9.757142857142858e-06,
      "loss": 0.4571,
      "step": 120
    },
    {
      "epoch": 0.5306122448979592,
      "grad_norm": 6.9328460693359375,
      "learning_rate": 9.736734693877551e-06,
      "loss": 0.6597,
      "step": 130
    },
    {
      "epoch": 0.5714285714285714,
      "grad_norm": 7.962501049041748,
      "learning_rate": 9.716326530612246e-06,
      "loss": 0.5132,
      "step": 140
    },
    {
      "epoch": 0.6122448979591837,
      "grad_norm": 10.508763313293457,
      "learning_rate": 9.69591836734694e-06,
      "loss": 0.6893,
      "step": 150
    },
    {
      "epoch": 0.6530612244897959,
      "grad_norm": 7.637253761291504,
      "learning_rate": 9.675510204081635e-06,
      "loss": 0.6142,
      "step": 160
    },
    {
      "epoch": 0.6938775510204082,
      "grad_norm": 10.0332670211792,
      "learning_rate": 9.655102040816328e-06,
      "loss": 0.582,
      "step": 170
    },
    {
      "epoch": 0.7346938775510204,
      "grad_norm": 8.150875091552734,
      "learning_rate": 9.634693877551021e-06,
      "loss": 0.477,
      "step": 180
    },
    {
      "epoch": 0.7755102040816326,
      "grad_norm": 10.330913543701172,
      "learning_rate": 9.614285714285714e-06,
      "loss": 0.5916,
      "step": 190
    },
    {
      "epoch": 0.8163265306122449,
      "grad_norm": 11.654999732971191,
      "learning_rate": 9.593877551020408e-06,
      "loss": 0.6236,
      "step": 200
    },
    {
      "epoch": 0.8571428571428571,
      "grad_norm": 8.048078536987305,
      "learning_rate": 9.573469387755103e-06,
      "loss": 0.6142,
      "step": 210
    },
    {
      "epoch": 0.8979591836734694,
      "grad_norm": 9.869592666625977,
      "learning_rate": 9.553061224489798e-06,
      "loss": 0.625,
      "step": 220
    },
    {
      "epoch": 0.9387755102040817,
      "grad_norm": 8.321409225463867,
      "learning_rate": 9.532653061224491e-06,
      "loss": 0.5767,
      "step": 230
    },
    {
      "epoch": 0.9795918367346939,
      "grad_norm": 7.6769256591796875,
      "learning_rate": 9.512244897959184e-06,
      "loss": 0.5134,
      "step": 240
    },
    {
      "epoch": 1.0204081632653061,
      "grad_norm": 9.609123229980469,
      "learning_rate": 9.491836734693877e-06,
      "loss": 0.5868,
      "step": 250
    },
    {
      "epoch": 1.0612244897959184,
      "grad_norm": 9.19683837890625,
      "learning_rate": 9.471428571428572e-06,
      "loss": 0.5215,
      "step": 260
    },
    {
      "epoch": 1.1020408163265305,
      "grad_norm": 7.328164577484131,
      "learning_rate": 9.451020408163266e-06,
      "loss": 0.5422,
      "step": 270
    },
    {
      "epoch": 1.1428571428571428,
      "grad_norm": 6.913904190063477,
      "learning_rate": 9.430612244897959e-06,
      "loss": 0.5214,
      "step": 280
    },
    {
      "epoch": 1.183673469387755,
      "grad_norm": 9.28811264038086,
      "learning_rate": 9.410204081632654e-06,
      "loss": 0.5319,
      "step": 290
    },
    {
      "epoch": 1.2244897959183674,
      "grad_norm": 9.132966041564941,
      "learning_rate": 9.389795918367349e-06,
      "loss": 0.5581,
      "step": 300
    },
    {
      "epoch": 1.2653061224489797,
      "grad_norm": 6.9722065925598145,
      "learning_rate": 9.369387755102042e-06,
      "loss": 0.431,
      "step": 310
    },
    {
      "epoch": 1.306122448979592,
      "grad_norm": 5.06177282333374,
      "learning_rate": 9.348979591836736e-06,
      "loss": 0.4583,
      "step": 320
    },
    {
      "epoch": 1.346938775510204,
      "grad_norm": 7.732840538024902,
      "learning_rate": 9.328571428571429e-06,
      "loss": 0.4194,
      "step": 330
    },
    {
      "epoch": 1.3877551020408163,
      "grad_norm": 8.94101333618164,
      "learning_rate": 9.308163265306122e-06,
      "loss": 0.4519,
      "step": 340
    },
    {
      "epoch": 1.4285714285714286,
      "grad_norm": 7.5437750816345215,
      "learning_rate": 9.287755102040817e-06,
      "loss": 0.5095,
      "step": 350
    },
    {
      "epoch": 1.469387755102041,
      "grad_norm": 5.702700138092041,
      "learning_rate": 9.26734693877551e-06,
      "loss": 0.3936,
      "step": 360
    },
    {
      "epoch": 1.510204081632653,
      "grad_norm": 9.153871536254883,
      "learning_rate": 9.246938775510205e-06,
      "loss": 0.4566,
      "step": 370
    },
    {
      "epoch": 1.5510204081632653,
      "grad_norm": 13.249794006347656,
      "learning_rate": 9.226530612244899e-06,
      "loss": 0.5216,
      "step": 380
    },
    {
      "epoch": 1.5918367346938775,
      "grad_norm": 7.065913200378418,
      "learning_rate": 9.206122448979594e-06,
      "loss": 0.4562,
      "step": 390
    },
    {
      "epoch": 1.6326530612244898,
      "grad_norm": 7.559301853179932,
      "learning_rate": 9.185714285714287e-06,
      "loss": 0.3883,
      "step": 400
    },
    {
      "epoch": 1.6734693877551021,
      "grad_norm": 12.103629112243652,
      "learning_rate": 9.16530612244898e-06,
      "loss": 0.4149,
      "step": 410
    },
    {
      "epoch": 1.7142857142857144,
      "grad_norm": 7.9720072746276855,
      "learning_rate": 9.144897959183673e-06,
      "loss": 0.4718,
      "step": 420
    },
    {
      "epoch": 1.7551020408163265,
      "grad_norm": 4.845782279968262,
      "learning_rate": 9.124489795918368e-06,
      "loss": 0.4304,
      "step": 430
    },
    {
      "epoch": 1.7959183673469388,
      "grad_norm": 6.954368591308594,
      "learning_rate": 9.104081632653062e-06,
      "loss": 0.3436,
      "step": 440
    },
    {
      "epoch": 1.836734693877551,
      "grad_norm": 4.751299858093262,
      "learning_rate": 9.083673469387757e-06,
      "loss": 0.4366,
      "step": 450
    },
    {
      "epoch": 1.8775510204081631,
      "grad_norm": 6.507364273071289,
      "learning_rate": 9.06326530612245e-06,
      "loss": 0.5794,
      "step": 460
    },
    {
      "epoch": 1.9183673469387754,
      "grad_norm": 8.891802787780762,
      "learning_rate": 9.042857142857143e-06,
      "loss": 0.4616,
      "step": 470
    },
    {
      "epoch": 1.9591836734693877,
      "grad_norm": 10.056327819824219,
      "learning_rate": 9.022448979591838e-06,
      "loss": 0.4946,
      "step": 480
    },
    {
      "epoch": 2.0,
      "grad_norm": 7.899660110473633,
      "learning_rate": 9.002040816326531e-06,
      "loss": 0.4437,
      "step": 490
    },
    {
      "epoch": 2.0408163265306123,
      "grad_norm": 6.761326313018799,
      "learning_rate": 8.981632653061225e-06,
      "loss": 0.4303,
      "step": 500
    },
    {
      "epoch": 2.0816326530612246,
      "grad_norm": 8.639615058898926,
      "learning_rate": 8.96122448979592e-06,
      "loss": 0.3267,
      "step": 510
    },
    {
      "epoch": 2.122448979591837,
      "grad_norm": 7.710758209228516,
      "learning_rate": 8.940816326530613e-06,
      "loss": 0.3559,
      "step": 520
    },
    {
      "epoch": 2.163265306122449,
      "grad_norm": 6.812905311584473,
      "learning_rate": 8.920408163265308e-06,
      "loss": 0.4761,
      "step": 530
    },
    {
      "epoch": 2.204081632653061,
      "grad_norm": 7.2431511878967285,
      "learning_rate": 8.900000000000001e-06,
      "loss": 0.405,
      "step": 540
    },
    {
      "epoch": 2.2448979591836733,
      "grad_norm": 7.230724811553955,
      "learning_rate": 8.879591836734694e-06,
      "loss": 0.3638,
      "step": 550
    },
    {
      "epoch": 2.2857142857142856,
      "grad_norm": 9.520208358764648,
      "learning_rate": 8.859183673469388e-06,
      "loss": 0.3473,
      "step": 560
    },
    {
      "epoch": 2.326530612244898,
      "grad_norm": 7.048585414886475,
      "learning_rate": 8.838775510204083e-06,
      "loss": 0.3652,
      "step": 570
    },
    {
      "epoch": 2.36734693877551,
      "grad_norm": 6.979404449462891,
      "learning_rate": 8.818367346938776e-06,
      "loss": 0.3855,
      "step": 580
    },
    {
      "epoch": 2.4081632653061225,
      "grad_norm": 3.765305280685425,
      "learning_rate": 8.797959183673471e-06,
      "loss": 0.3452,
      "step": 590
    },
    {
      "epoch": 2.4489795918367347,
      "grad_norm": 10.533697128295898,
      "learning_rate": 8.777551020408164e-06,
      "loss": 0.3874,
      "step": 600
    },
    {
      "epoch": 2.489795918367347,
      "grad_norm": 8.108145713806152,
      "learning_rate": 8.757142857142858e-06,
      "loss": 0.3695,
      "step": 610
    },
    {
      "epoch": 2.5306122448979593,
      "grad_norm": 7.947360992431641,
      "learning_rate": 8.736734693877552e-06,
      "loss": 0.408,
      "step": 620
    },
    {
      "epoch": 2.571428571428571,
      "grad_norm": 7.8081374168396,
      "learning_rate": 8.716326530612246e-06,
      "loss": 0.4059,
      "step": 630
    },
    {
      "epoch": 2.612244897959184,
      "grad_norm": 8.579155921936035,
      "learning_rate": 8.695918367346939e-06,
      "loss": 0.3934,
      "step": 640
    },
    {
      "epoch": 2.6530612244897958,
      "grad_norm": 6.4387712478637695,
      "learning_rate": 8.675510204081632e-06,
      "loss": 0.4256,
      "step": 650
    },
    {
      "epoch": 2.693877551020408,
      "grad_norm": 8.415692329406738,
      "learning_rate": 8.655102040816327e-06,
      "loss": 0.3453,
      "step": 660
    },
    {
      "epoch": 2.7346938775510203,
      "grad_norm": 8.50904369354248,
      "learning_rate": 8.63469387755102e-06,
      "loss": 0.4766,
      "step": 670
    },
    {
      "epoch": 2.7755102040816326,
      "grad_norm": 4.662519931793213,
      "learning_rate": 8.614285714285716e-06,
      "loss": 0.3973,
      "step": 680
    },
    {
      "epoch": 2.816326530612245,
      "grad_norm": 6.288435935974121,
      "learning_rate": 8.593877551020409e-06,
      "loss": 0.4408,
      "step": 690
    },
    {
      "epoch": 2.857142857142857,
      "grad_norm": 6.625838279724121,
      "learning_rate": 8.573469387755102e-06,
      "loss": 0.2908,
      "step": 700
    },
    {
      "epoch": 2.8979591836734695,
      "grad_norm": 8.510032653808594,
      "learning_rate": 8.553061224489797e-06,
      "loss": 0.3813,
      "step": 710
    },
    {
      "epoch": 2.938775510204082,
      "grad_norm": 11.82463264465332,
      "learning_rate": 8.53265306122449e-06,
      "loss": 0.4352,
      "step": 720
    },
    {
      "epoch": 2.979591836734694,
      "grad_norm": 8.821819305419922,
      "learning_rate": 8.512244897959184e-06,
      "loss": 0.4318,
      "step": 730
    },
    {
      "epoch": 3.020408163265306,
      "grad_norm": 8.010713577270508,
      "learning_rate": 8.491836734693879e-06,
      "loss": 0.2323,
      "step": 740
    },
    {
      "epoch": 3.061224489795918,
      "grad_norm": 9.03991985321045,
      "learning_rate": 8.471428571428572e-06,
      "loss": 0.3603,
      "step": 750
    },
    {
      "epoch": 3.1020408163265305,
      "grad_norm": 10.94204044342041,
      "learning_rate": 8.451020408163267e-06,
      "loss": 0.3576,
      "step": 760
    },
    {
      "epoch": 3.142857142857143,
      "grad_norm": 7.89410924911499,
      "learning_rate": 8.43061224489796e-06,
      "loss": 0.2851,
      "step": 770
    },
    {
      "epoch": 3.183673469387755,
      "grad_norm": 6.53656005859375,
      "learning_rate": 8.410204081632653e-06,
      "loss": 0.318,
      "step": 780
    },
    {
      "epoch": 3.2244897959183674,
      "grad_norm": 6.487284183502197,
      "learning_rate": 8.389795918367347e-06,
      "loss": 0.317,
      "step": 790
    },
    {
      "epoch": 3.2653061224489797,
      "grad_norm": 6.947931289672852,
      "learning_rate": 8.369387755102042e-06,
      "loss": 0.2879,
      "step": 800
    },
    {
      "epoch": 3.306122448979592,
      "grad_norm": 4.166048526763916,
      "learning_rate": 8.348979591836735e-06,
      "loss": 0.3392,
      "step": 810
    },
    {
      "epoch": 3.3469387755102042,
      "grad_norm": 9.974846839904785,
      "learning_rate": 8.32857142857143e-06,
      "loss": 0.3663,
      "step": 820
    },
    {
      "epoch": 3.387755102040816,
      "grad_norm": 9.668428421020508,
      "learning_rate": 8.308163265306123e-06,
      "loss": 0.3212,
      "step": 830
    },
    {
      "epoch": 3.4285714285714284,
      "grad_norm": 11.81507396697998,
      "learning_rate": 8.287755102040816e-06,
      "loss": 0.3241,
      "step": 840
    },
    {
      "epoch": 3.4693877551020407,
      "grad_norm": 13.690321922302246,
      "learning_rate": 8.267346938775511e-06,
      "loss": 0.4535,
      "step": 850
    },
    {
      "epoch": 3.510204081632653,
      "grad_norm": 11.042778968811035,
      "learning_rate": 8.246938775510205e-06,
      "loss": 0.3826,
      "step": 860
    },
    {
      "epoch": 3.5510204081632653,
      "grad_norm": 8.57719612121582,
      "learning_rate": 8.226530612244898e-06,
      "loss": 0.3905,
      "step": 870
    },
    {
      "epoch": 3.5918367346938775,
      "grad_norm": 7.843425750732422,
      "learning_rate": 8.206122448979591e-06,
      "loss": 0.3125,
      "step": 880
    },
    {
      "epoch": 3.63265306122449,
      "grad_norm": 5.9236931800842285,
      "learning_rate": 8.185714285714286e-06,
      "loss": 0.3512,
      "step": 890
    },
    {
      "epoch": 3.673469387755102,
      "grad_norm": 8.213603973388672,
      "learning_rate": 8.165306122448981e-06,
      "loss": 0.4094,
      "step": 900
    },
    {
      "epoch": 3.7142857142857144,
      "grad_norm": 3.8083949089050293,
      "learning_rate": 8.144897959183674e-06,
      "loss": 0.2751,
      "step": 910
    },
    {
      "epoch": 3.7551020408163263,
      "grad_norm": 12.339240074157715,
      "learning_rate": 8.124489795918368e-06,
      "loss": 0.3296,
      "step": 920
    },
    {
      "epoch": 3.795918367346939,
      "grad_norm": 9.532052040100098,
      "learning_rate": 8.104081632653061e-06,
      "loss": 0.3033,
      "step": 930
    },
    {
      "epoch": 3.836734693877551,
      "grad_norm": 6.307032108306885,
      "learning_rate": 8.083673469387756e-06,
      "loss": 0.3765,
      "step": 940
    },
    {
      "epoch": 3.877551020408163,
      "grad_norm": 7.3003010749816895,
      "learning_rate": 8.06326530612245e-06,
      "loss": 0.2161,
      "step": 950
    },
    {
      "epoch": 3.9183673469387754,
      "grad_norm": 7.6572699546813965,
      "learning_rate": 8.042857142857143e-06,
      "loss": 0.2886,
      "step": 960
    },
    {
      "epoch": 3.9591836734693877,
      "grad_norm": 6.745776176452637,
      "learning_rate": 8.022448979591838e-06,
      "loss": 0.3376,
      "step": 970
    },
    {
      "epoch": 4.0,
      "grad_norm": 10.482270240783691,
      "learning_rate": 8.002040816326533e-06,
      "loss": 0.2657,
      "step": 980
    },
    {
      "epoch": 4.040816326530612,
      "grad_norm": 6.213717460632324,
      "learning_rate": 7.981632653061226e-06,
      "loss": 0.2596,
      "step": 990
    },
    {
      "epoch": 4.081632653061225,
      "grad_norm": 10.256094932556152,
      "learning_rate": 7.961224489795919e-06,
      "loss": 0.2338,
      "step": 1000
    }
  ],
  "logging_steps": 10,
  "max_steps": 4900,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 20,
  "save_steps": 1000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 0.0,
  "train_batch_size": 30,
  "trial_name": null,
  "trial_params": null
}