{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 4.149377593360996,
  "eval_steps": 500,
  "global_step": 1000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.04149377593360996,
      "grad_norm": 0.7534617185592651,
      "learning_rate": 1e-05,
      "loss": 1.3441,
      "step": 10
    },
    {
      "epoch": 0.08298755186721991,
      "grad_norm": 0.17292505502700806,
      "learning_rate": 2e-05,
      "loss": 1.2553,
      "step": 20
    },
    {
      "epoch": 0.12448132780082988,
      "grad_norm": 0.16635559499263763,
      "learning_rate": 3e-05,
      "loss": 1.2958,
      "step": 30
    },
    {
      "epoch": 0.16597510373443983,
      "grad_norm": 0.1318502277135849,
      "learning_rate": 4e-05,
      "loss": 1.1205,
      "step": 40
    },
    {
      "epoch": 0.2074688796680498,
      "grad_norm": 0.21048150956630707,
      "learning_rate": 5e-05,
      "loss": 1.0102,
      "step": 50
    },
    {
      "epoch": 0.24896265560165975,
      "grad_norm": 0.239803746342659,
      "learning_rate": 6e-05,
      "loss": 1.0016,
      "step": 60
    },
    {
      "epoch": 0.29045643153526973,
      "grad_norm": 0.16366511583328247,
      "learning_rate": 7e-05,
      "loss": 0.9466,
      "step": 70
    },
    {
      "epoch": 0.33195020746887965,
      "grad_norm": 0.24975861608982086,
      "learning_rate": 8e-05,
      "loss": 0.7838,
      "step": 80
    },
    {
      "epoch": 0.37344398340248963,
      "grad_norm": 0.282753586769104,
      "learning_rate": 9e-05,
      "loss": 0.7927,
      "step": 90
    },
    {
      "epoch": 0.4149377593360996,
      "grad_norm": 0.19942770898342133,
      "learning_rate": 0.0001,
      "loss": 0.7928,
      "step": 100
    },
    {
      "epoch": 0.45643153526970953,
      "grad_norm": 0.21815817058086395,
      "learning_rate": 9.888888888888889e-05,
      "loss": 0.58,
      "step": 110
    },
    {
      "epoch": 0.4979253112033195,
      "grad_norm": 0.19703078269958496,
      "learning_rate": 9.777777777777778e-05,
      "loss": 0.639,
      "step": 120
    },
    {
      "epoch": 0.5394190871369294,
      "grad_norm": 0.270720511674881,
      "learning_rate": 9.666666666666667e-05,
      "loss": 0.5981,
      "step": 130
    },
    {
      "epoch": 0.5809128630705395,
      "grad_norm": 0.2211589813232422,
      "learning_rate": 9.555555555555557e-05,
      "loss": 0.6611,
      "step": 140
    },
    {
      "epoch": 0.6224066390041494,
      "grad_norm": 0.22035779058933258,
      "learning_rate": 9.444444444444444e-05,
      "loss": 0.6056,
      "step": 150
    },
    {
      "epoch": 0.6639004149377593,
      "grad_norm": 0.3599834740161896,
      "learning_rate": 9.333333333333334e-05,
      "loss": 0.7573,
      "step": 160
    },
    {
      "epoch": 0.7053941908713693,
      "grad_norm": 0.2543361485004425,
      "learning_rate": 9.222222222222223e-05,
      "loss": 0.6592,
      "step": 170
    },
    {
      "epoch": 0.7468879668049793,
      "grad_norm": 0.26748475432395935,
      "learning_rate": 9.111111111111112e-05,
      "loss": 0.5471,
      "step": 180
    },
    {
      "epoch": 0.7883817427385892,
      "grad_norm": 0.3198457956314087,
      "learning_rate": 9e-05,
      "loss": 0.5802,
      "step": 190
    },
    {
      "epoch": 0.8298755186721992,
      "grad_norm": 0.26507750153541565,
      "learning_rate": 8.888888888888889e-05,
      "loss": 0.5088,
      "step": 200
    },
    {
      "epoch": 0.8713692946058091,
      "grad_norm": 0.32707032561302185,
      "learning_rate": 8.777777777777778e-05,
      "loss": 0.4706,
      "step": 210
    },
    {
      "epoch": 0.9128630705394191,
      "grad_norm": 0.20190617442131042,
      "learning_rate": 8.666666666666667e-05,
      "loss": 0.3719,
      "step": 220
    },
    {
      "epoch": 0.9543568464730291,
      "grad_norm": 0.270940363407135,
      "learning_rate": 8.555555555555556e-05,
      "loss": 0.5147,
      "step": 230
    },
    {
      "epoch": 0.995850622406639,
      "grad_norm": 0.34486109018325806,
      "learning_rate": 8.444444444444444e-05,
      "loss": 0.4241,
      "step": 240
    },
    {
      "epoch": 1.037344398340249,
      "grad_norm": 0.42535537481307983,
      "learning_rate": 8.333333333333334e-05,
      "loss": 0.5259,
      "step": 250
    },
    {
      "epoch": 1.0788381742738589,
      "grad_norm": 0.4141991138458252,
      "learning_rate": 8.222222222222222e-05,
      "loss": 0.3362,
      "step": 260
    },
    {
      "epoch": 1.120331950207469,
      "grad_norm": 0.38707345724105835,
      "learning_rate": 8.111111111111112e-05,
      "loss": 0.1921,
      "step": 270
    },
    {
      "epoch": 1.161825726141079,
      "grad_norm": 0.22714611887931824,
      "learning_rate": 8e-05,
      "loss": 0.5203,
      "step": 280
    },
    {
      "epoch": 1.2033195020746887,
      "grad_norm": 0.23145520687103271,
      "learning_rate": 7.88888888888889e-05,
      "loss": 0.3289,
      "step": 290
    },
    {
      "epoch": 1.2448132780082988,
      "grad_norm": 0.24606278538703918,
      "learning_rate": 7.777777777777778e-05,
      "loss": 0.4789,
      "step": 300
    },
    {
      "epoch": 1.2863070539419086,
      "grad_norm": 0.3598890006542206,
      "learning_rate": 7.666666666666667e-05,
      "loss": 0.3212,
      "step": 310
    },
    {
      "epoch": 1.3278008298755186,
      "grad_norm": 0.4746328592300415,
      "learning_rate": 7.555555555555556e-05,
      "loss": 0.4568,
      "step": 320
    },
    {
      "epoch": 1.3692946058091287,
      "grad_norm": 0.28307732939720154,
      "learning_rate": 7.444444444444444e-05,
      "loss": 0.3563,
      "step": 330
    },
    {
      "epoch": 1.4107883817427385,
      "grad_norm": 0.3323902189731598,
      "learning_rate": 7.333333333333333e-05,
      "loss": 0.5342,
      "step": 340
    },
    {
      "epoch": 1.4522821576763485,
      "grad_norm": 0.29897817969322205,
      "learning_rate": 7.222222222222222e-05,
      "loss": 0.3693,
      "step": 350
    },
    {
      "epoch": 1.4937759336099585,
      "grad_norm": 0.5410626530647278,
      "learning_rate": 7.111111111111112e-05,
      "loss": 0.4809,
      "step": 360
    },
    {
      "epoch": 1.5352697095435683,
      "grad_norm": 0.28330478072166443,
      "learning_rate": 7e-05,
      "loss": 0.3591,
      "step": 370
    },
    {
      "epoch": 1.5767634854771784,
      "grad_norm": 0.47028374671936035,
      "learning_rate": 6.88888888888889e-05,
      "loss": 0.3717,
      "step": 380
    },
    {
      "epoch": 1.6182572614107884,
      "grad_norm": 0.4893304109573364,
      "learning_rate": 6.777777777777778e-05,
      "loss": 0.3227,
      "step": 390
    },
    {
      "epoch": 1.6597510373443982,
      "grad_norm": 0.5082384347915649,
      "learning_rate": 6.666666666666667e-05,
      "loss": 0.3572,
      "step": 400
    },
    {
      "epoch": 1.7012448132780082,
      "grad_norm": 0.34136882424354553,
      "learning_rate": 6.555555555555556e-05,
      "loss": 0.4784,
      "step": 410
    },
    {
      "epoch": 1.7427385892116183,
      "grad_norm": 0.39626437425613403,
      "learning_rate": 6.444444444444446e-05,
      "loss": 0.4348,
      "step": 420
    },
    {
      "epoch": 1.784232365145228,
      "grad_norm": 0.5795237421989441,
      "learning_rate": 6.333333333333333e-05,
      "loss": 0.4211,
      "step": 430
    },
    {
      "epoch": 1.8257261410788381,
      "grad_norm": 0.6919090151786804,
      "learning_rate": 6.222222222222222e-05,
      "loss": 0.3886,
      "step": 440
    },
    {
      "epoch": 1.8672199170124482,
      "grad_norm": 0.3347764015197754,
      "learning_rate": 6.111111111111112e-05,
      "loss": 0.3753,
      "step": 450
    },
    {
      "epoch": 1.908713692946058,
      "grad_norm": 0.8023832440376282,
      "learning_rate": 6e-05,
      "loss": 0.3052,
      "step": 460
    },
    {
      "epoch": 1.950207468879668,
      "grad_norm": 0.23686052858829498,
      "learning_rate": 5.8888888888888896e-05,
      "loss": 0.4345,
      "step": 470
    },
    {
      "epoch": 1.991701244813278,
      "grad_norm": 0.38190993666648865,
      "learning_rate": 5.7777777777777776e-05,
      "loss": 0.3561,
      "step": 480
    },
    {
      "epoch": 2.033195020746888,
      "grad_norm": 0.38062775135040283,
      "learning_rate": 5.666666666666667e-05,
      "loss": 0.4208,
      "step": 490
    },
    {
      "epoch": 2.074688796680498,
      "grad_norm": 0.42297548055648804,
      "learning_rate": 5.555555555555556e-05,
      "loss": 0.2817,
      "step": 500
    },
    {
      "epoch": 2.116182572614108,
      "grad_norm": 0.5791902542114258,
      "learning_rate": 5.4444444444444446e-05,
      "loss": 0.2804,
      "step": 510
    },
    {
      "epoch": 2.1576763485477177,
      "grad_norm": 0.39327019453048706,
      "learning_rate": 5.333333333333333e-05,
      "loss": 0.2416,
      "step": 520
    },
    {
      "epoch": 2.199170124481328,
      "grad_norm": 0.49688029289245605,
      "learning_rate": 5.222222222222223e-05,
      "loss": 0.3241,
      "step": 530
    },
    {
      "epoch": 2.240663900414938,
      "grad_norm": 0.6079151630401611,
      "learning_rate": 5.111111111111111e-05,
      "loss": 0.383,
      "step": 540
    },
    {
      "epoch": 2.2821576763485476,
      "grad_norm": 0.2557106614112854,
      "learning_rate": 5e-05,
      "loss": 0.2163,
      "step": 550
    },
    {
      "epoch": 2.323651452282158,
      "grad_norm": 0.41979703307151794,
      "learning_rate": 4.888888888888889e-05,
      "loss": 0.3263,
      "step": 560
    },
    {
      "epoch": 2.3651452282157677,
      "grad_norm": 0.298300564289093,
      "learning_rate": 4.7777777777777784e-05,
      "loss": 0.178,
      "step": 570
    },
    {
      "epoch": 2.4066390041493775,
      "grad_norm": 0.33259597420692444,
      "learning_rate": 4.666666666666667e-05,
      "loss": 0.2444,
      "step": 580
    },
    {
      "epoch": 2.4481327800829877,
      "grad_norm": 0.3294433355331421,
      "learning_rate": 4.555555555555556e-05,
      "loss": 0.1401,
      "step": 590
    },
    {
      "epoch": 2.4896265560165975,
      "grad_norm": 0.4625357389450073,
      "learning_rate": 4.4444444444444447e-05,
      "loss": 0.3397,
      "step": 600
    },
    {
      "epoch": 2.5311203319502074,
      "grad_norm": 0.5396658778190613,
      "learning_rate": 4.3333333333333334e-05,
      "loss": 0.3147,
      "step": 610
    },
    {
      "epoch": 2.572614107883817,
      "grad_norm": 0.4528850317001343,
      "learning_rate": 4.222222222222222e-05,
      "loss": 0.2427,
      "step": 620
    },
    {
      "epoch": 2.6141078838174274,
      "grad_norm": 0.47281742095947266,
      "learning_rate": 4.111111111111111e-05,
      "loss": 0.2276,
      "step": 630
    },
    {
      "epoch": 2.6556016597510372,
      "grad_norm": 0.2738474905490875,
      "learning_rate": 4e-05,
      "loss": 0.2599,
      "step": 640
    },
    {
      "epoch": 2.6970954356846475,
      "grad_norm": 0.31404346227645874,
      "learning_rate": 3.888888888888889e-05,
      "loss": 0.2423,
      "step": 650
    },
    {
      "epoch": 2.7385892116182573,
      "grad_norm": 0.53602135181427,
      "learning_rate": 3.777777777777778e-05,
      "loss": 0.2915,
      "step": 660
    },
    {
      "epoch": 2.780082987551867,
      "grad_norm": 0.312480092048645,
      "learning_rate": 3.6666666666666666e-05,
      "loss": 0.3175,
      "step": 670
    },
    {
      "epoch": 2.821576763485477,
      "grad_norm": 0.2691058814525604,
      "learning_rate": 3.555555555555556e-05,
      "loss": 0.2499,
      "step": 680
    },
    {
      "epoch": 2.863070539419087,
      "grad_norm": 0.5619029998779297,
      "learning_rate": 3.444444444444445e-05,
      "loss": 0.3429,
      "step": 690
    },
    {
      "epoch": 2.904564315352697,
      "grad_norm": 0.625348687171936,
      "learning_rate": 3.3333333333333335e-05,
      "loss": 0.1967,
      "step": 700
    },
    {
      "epoch": 2.9460580912863072,
      "grad_norm": 0.20030611753463745,
      "learning_rate": 3.222222222222223e-05,
      "loss": 0.2182,
      "step": 710
    },
    {
      "epoch": 2.987551867219917,
      "grad_norm": 0.2181701362133026,
      "learning_rate": 3.111111111111111e-05,
      "loss": 0.3135,
      "step": 720
    },
    {
      "epoch": 3.029045643153527,
      "grad_norm": 0.5375049710273743,
      "learning_rate": 3e-05,
      "loss": 0.2967,
      "step": 730
    },
    {
      "epoch": 3.070539419087137,
      "grad_norm": 0.40118399262428284,
      "learning_rate": 2.8888888888888888e-05,
      "loss": 0.1545,
      "step": 740
    },
    {
      "epoch": 3.112033195020747,
      "grad_norm": 0.7035902738571167,
      "learning_rate": 2.777777777777778e-05,
      "loss": 0.2678,
      "step": 750
    },
    {
      "epoch": 3.1535269709543567,
      "grad_norm": 0.7388056516647339,
      "learning_rate": 2.6666666666666667e-05,
      "loss": 0.2665,
      "step": 760
    },
    {
      "epoch": 3.195020746887967,
      "grad_norm": 0.7555689811706543,
      "learning_rate": 2.5555555555555554e-05,
      "loss": 0.2337,
      "step": 770
    },
    {
      "epoch": 3.236514522821577,
      "grad_norm": 0.8938640356063843,
      "learning_rate": 2.4444444444444445e-05,
      "loss": 0.1993,
      "step": 780
    },
    {
      "epoch": 3.2780082987551866,
      "grad_norm": 0.7519901990890503,
      "learning_rate": 2.3333333333333336e-05,
      "loss": 0.2179,
      "step": 790
    },
    {
      "epoch": 3.3195020746887964,
      "grad_norm": 0.3737233281135559,
      "learning_rate": 2.2222222222222223e-05,
      "loss": 0.1721,
      "step": 800
    },
    {
      "epoch": 3.3609958506224067,
      "grad_norm": 0.6671114563941956,
      "learning_rate": 2.111111111111111e-05,
      "loss": 0.2366,
      "step": 810
    },
    {
      "epoch": 3.4024896265560165,
      "grad_norm": 0.5140301585197449,
      "learning_rate": 2e-05,
      "loss": 0.2073,
      "step": 820
    },
    {
      "epoch": 3.4439834024896268,
      "grad_norm": 0.2032405585050583,
      "learning_rate": 1.888888888888889e-05,
      "loss": 0.1978,
      "step": 830
    },
    {
      "epoch": 3.4854771784232366,
      "grad_norm": 0.5465924739837646,
      "learning_rate": 1.777777777777778e-05,
      "loss": 0.2103,
      "step": 840
    },
    {
      "epoch": 3.5269709543568464,
      "grad_norm": 0.49002525210380554,
      "learning_rate": 1.6666666666666667e-05,
      "loss": 0.2528,
      "step": 850
    },
    {
      "epoch": 3.568464730290456,
      "grad_norm": 0.37268325686454773,
      "learning_rate": 1.5555555555555555e-05,
      "loss": 0.2067,
      "step": 860
    },
    {
      "epoch": 3.6099585062240664,
      "grad_norm": 0.8844376802444458,
      "learning_rate": 1.4444444444444444e-05,
      "loss": 0.1711,
      "step": 870
    },
    {
      "epoch": 3.6514522821576763,
      "grad_norm": 0.4846467077732086,
      "learning_rate": 1.3333333333333333e-05,
      "loss": 0.1465,
      "step": 880
    },
    {
      "epoch": 3.6929460580912865,
      "grad_norm": 0.645414412021637,
      "learning_rate": 1.2222222222222222e-05,
      "loss": 0.2711,
      "step": 890
    },
    {
      "epoch": 3.7344398340248963,
      "grad_norm": 0.4424195885658264,
      "learning_rate": 1.1111111111111112e-05,
      "loss": 0.1949,
      "step": 900
    },
    {
      "epoch": 3.775933609958506,
      "grad_norm": 0.21110430359840393,
      "learning_rate": 1e-05,
      "loss": 0.1738,
      "step": 910
    },
    {
      "epoch": 3.817427385892116,
      "grad_norm": 0.3548027276992798,
      "learning_rate": 8.88888888888889e-06,
      "loss": 0.1363,
      "step": 920
    },
    {
      "epoch": 3.858921161825726,
      "grad_norm": 0.43183088302612305,
      "learning_rate": 7.777777777777777e-06,
      "loss": 0.205,
      "step": 930
    },
    {
      "epoch": 3.900414937759336,
      "grad_norm": 0.21933656930923462,
      "learning_rate": 6.666666666666667e-06,
      "loss": 0.2037,
      "step": 940
    },
    {
      "epoch": 3.9419087136929463,
      "grad_norm": 0.6654905080795288,
      "learning_rate": 5.555555555555556e-06,
      "loss": 0.209,
      "step": 950
    },
    {
      "epoch": 3.983402489626556,
      "grad_norm": 0.2863740026950836,
      "learning_rate": 4.444444444444445e-06,
      "loss": 0.2686,
      "step": 960
    },
    {
      "epoch": 4.024896265560166,
      "grad_norm": 0.23517794907093048,
      "learning_rate": 3.3333333333333333e-06,
      "loss": 0.1827,
      "step": 970
    },
    {
      "epoch": 4.066390041493776,
      "grad_norm": 0.7014562487602234,
      "learning_rate": 2.2222222222222225e-06,
      "loss": 0.2283,
      "step": 980
    },
    {
      "epoch": 4.1078838174273855,
      "grad_norm": 0.4174599051475525,
      "learning_rate": 1.1111111111111112e-06,
      "loss": 0.1267,
      "step": 990
    },
    {
      "epoch": 4.149377593360996,
      "grad_norm": 0.1902274340391159,
      "learning_rate": 0.0,
      "loss": 0.1947,
      "step": 1000
    }
  ],
  "logging_steps": 10,
  "max_steps": 1000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 7702734288592896.0,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}