{
  "best_metric": 1.5566551155643538e-05,
  "best_model_checkpoint": "saves/CADICA_qwenvl_direction_scale4/lora/sft/checkpoint-1150",
  "epoch": 0.6185938707185166,
  "eval_steps": 50,
  "global_step": 1201,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0025753283543651817,
      "grad_norm": 14.230540018977528,
      "learning_rate": 2.9411764705882355e-06,
      "loss": 1.0954,
      "num_input_tokens_seen": 49920,
      "step": 5
    },
    {
      "epoch": 0.0051506567087303634,
      "grad_norm": 12.725175128208063,
      "learning_rate": 5.882352941176471e-06,
      "loss": 0.9793,
      "num_input_tokens_seen": 99840,
      "step": 10
    },
    {
      "epoch": 0.007725985063095545,
      "grad_norm": 11.908760913870685,
      "learning_rate": 8.823529411764707e-06,
      "loss": 1.0964,
      "num_input_tokens_seen": 149760,
      "step": 15
    },
    {
      "epoch": 0.010301313417460727,
      "grad_norm": 7.5714179545602835,
      "learning_rate": 1.1764705882352942e-05,
      "loss": 0.7079,
      "num_input_tokens_seen": 199680,
      "step": 20
    },
    {
      "epoch": 0.012876641771825908,
      "grad_norm": 2.273110551179123,
      "learning_rate": 1.4705882352941177e-05,
      "loss": 0.4213,
      "num_input_tokens_seen": 249600,
      "step": 25
    },
    {
      "epoch": 0.01545197012619109,
      "grad_norm": 1.4511084067844011,
      "learning_rate": 1.7647058823529414e-05,
      "loss": 0.3359,
      "num_input_tokens_seen": 299520,
      "step": 30
    },
    {
      "epoch": 0.018027298480556272,
      "grad_norm": 1.7462649262033438,
      "learning_rate": 2.058823529411765e-05,
      "loss": 0.3463,
      "num_input_tokens_seen": 349440,
      "step": 35
    },
    {
      "epoch": 0.020602626834921454,
      "grad_norm": 1.2143595820366577,
      "learning_rate": 2.3529411764705884e-05,
      "loss": 0.28,
      "num_input_tokens_seen": 399360,
      "step": 40
    },
    {
      "epoch": 0.023177955189286635,
      "grad_norm": 0.8468719125045373,
      "learning_rate": 2.647058823529412e-05,
      "loss": 0.3442,
      "num_input_tokens_seen": 449280,
      "step": 45
    },
    {
      "epoch": 0.025753283543651816,
      "grad_norm": 0.9726867932660042,
      "learning_rate": 2.9411764705882354e-05,
      "loss": 0.3441,
      "num_input_tokens_seen": 499200,
      "step": 50
    },
    {
      "epoch": 0.025753283543651816,
      "eval_loss": 0.33834776282310486,
      "eval_runtime": 47.4467,
      "eval_samples_per_second": 1.265,
      "eval_steps_per_second": 0.316,
      "num_input_tokens_seen": 499200,
      "step": 50
    },
    {
      "epoch": 0.028328611898016998,
      "grad_norm": 1.4016556961184263,
      "learning_rate": 3.235294117647059e-05,
      "loss": 0.3182,
      "num_input_tokens_seen": 549120,
      "step": 55
    },
    {
      "epoch": 0.03090394025238218,
      "grad_norm": 0.6437613769459606,
      "learning_rate": 3.529411764705883e-05,
      "loss": 0.3294,
      "num_input_tokens_seen": 599040,
      "step": 60
    },
    {
      "epoch": 0.03347926860674736,
      "grad_norm": 0.7389008951321312,
      "learning_rate": 3.8235294117647055e-05,
      "loss": 0.3097,
      "num_input_tokens_seen": 648960,
      "step": 65
    },
    {
      "epoch": 0.036054596961112545,
      "grad_norm": 0.771553860801019,
      "learning_rate": 4.11764705882353e-05,
      "loss": 0.3008,
      "num_input_tokens_seen": 698880,
      "step": 70
    },
    {
      "epoch": 0.03862992531547772,
      "grad_norm": 0.6965369148334918,
      "learning_rate": 4.411764705882353e-05,
      "loss": 0.3278,
      "num_input_tokens_seen": 748800,
      "step": 75
    },
    {
      "epoch": 0.04120525366984291,
      "grad_norm": 0.912943461315541,
      "learning_rate": 4.705882352941177e-05,
      "loss": 0.3074,
      "num_input_tokens_seen": 798720,
      "step": 80
    },
    {
      "epoch": 0.043780582024208085,
      "grad_norm": 0.8407481737577445,
      "learning_rate": 5e-05,
      "loss": 0.3423,
      "num_input_tokens_seen": 848640,
      "step": 85
    },
    {
      "epoch": 0.04635591037857327,
      "grad_norm": 0.9112879058417015,
      "learning_rate": 5.294117647058824e-05,
      "loss": 0.3008,
      "num_input_tokens_seen": 898560,
      "step": 90
    },
    {
      "epoch": 0.04893123873293845,
      "grad_norm": 2.391489040464162,
      "learning_rate": 5.588235294117647e-05,
      "loss": 0.2815,
      "num_input_tokens_seen": 948480,
      "step": 95
    },
    {
      "epoch": 0.05150656708730363,
      "grad_norm": 2.155211791607199,
      "learning_rate": 5.882352941176471e-05,
      "loss": 0.2274,
      "num_input_tokens_seen": 998400,
      "step": 100
    },
    {
      "epoch": 0.05150656708730363,
      "eval_loss": 0.18663176894187927,
      "eval_runtime": 18.9199,
      "eval_samples_per_second": 3.171,
      "eval_steps_per_second": 0.793,
      "num_input_tokens_seen": 998400,
      "step": 100
    },
    {
      "epoch": 0.05408189544166881,
      "grad_norm": 2.2181531422996716,
      "learning_rate": 6.176470588235295e-05,
      "loss": 0.168,
      "num_input_tokens_seen": 1048320,
      "step": 105
    },
    {
      "epoch": 0.056657223796033995,
      "grad_norm": 3.1829920225573236,
      "learning_rate": 6.470588235294118e-05,
      "loss": 0.0709,
      "num_input_tokens_seen": 1098240,
      "step": 110
    },
    {
      "epoch": 0.05923255215039917,
      "grad_norm": 4.337350477588576,
      "learning_rate": 6.764705882352942e-05,
      "loss": 0.1609,
      "num_input_tokens_seen": 1148160,
      "step": 115
    },
    {
      "epoch": 0.06180788050476436,
      "grad_norm": 2.1010046045637365,
      "learning_rate": 7.058823529411765e-05,
      "loss": 0.0354,
      "num_input_tokens_seen": 1198080,
      "step": 120
    },
    {
      "epoch": 0.06438320885912954,
      "grad_norm": 2.232308844812103,
      "learning_rate": 7.352941176470589e-05,
      "loss": 0.1133,
      "num_input_tokens_seen": 1248000,
      "step": 125
    },
    {
      "epoch": 0.06695853721349472,
      "grad_norm": 5.641631090993415,
      "learning_rate": 7.647058823529411e-05,
      "loss": 0.0867,
      "num_input_tokens_seen": 1297920,
      "step": 130
    },
    {
      "epoch": 0.0695338655678599,
      "grad_norm": 1.5031437609685787,
      "learning_rate": 7.941176470588235e-05,
      "loss": 0.1352,
      "num_input_tokens_seen": 1347840,
      "step": 135
    },
    {
      "epoch": 0.07210919392222509,
      "grad_norm": 3.2992644431188465,
      "learning_rate": 8.23529411764706e-05,
      "loss": 0.101,
      "num_input_tokens_seen": 1397760,
      "step": 140
    },
    {
      "epoch": 0.07468452227659027,
      "grad_norm": 3.494236832758233,
      "learning_rate": 8.529411764705883e-05,
      "loss": 0.0334,
      "num_input_tokens_seen": 1447680,
      "step": 145
    },
    {
      "epoch": 0.07725985063095545,
      "grad_norm": 0.0602113869322109,
      "learning_rate": 8.823529411764706e-05,
      "loss": 0.0667,
      "num_input_tokens_seen": 1497600,
      "step": 150
    },
    {
      "epoch": 0.07725985063095545,
      "eval_loss": 0.09665286540985107,
      "eval_runtime": 19.2745,
      "eval_samples_per_second": 3.113,
      "eval_steps_per_second": 0.778,
      "num_input_tokens_seen": 1497600,
      "step": 150
    },
    {
      "epoch": 0.07983517898532062,
      "grad_norm": 6.096163706748617,
      "learning_rate": 9.11764705882353e-05,
      "loss": 0.06,
      "num_input_tokens_seen": 1547520,
      "step": 155
    },
    {
      "epoch": 0.08241050733968582,
      "grad_norm": 4.278069142242893,
      "learning_rate": 9.411764705882353e-05,
      "loss": 0.0265,
      "num_input_tokens_seen": 1597440,
      "step": 160
    },
    {
      "epoch": 0.08498583569405099,
      "grad_norm": 1.7183097652953412,
      "learning_rate": 9.705882352941177e-05,
      "loss": 0.1743,
      "num_input_tokens_seen": 1647360,
      "step": 165
    },
    {
      "epoch": 0.08756116404841617,
      "grad_norm": 4.95401899568707,
      "learning_rate": 0.0001,
      "loss": 0.051,
      "num_input_tokens_seen": 1697280,
      "step": 170
    },
    {
      "epoch": 0.09013649240278135,
      "grad_norm": 0.34558354886099124,
      "learning_rate": 9.999940874631277e-05,
      "loss": 0.0584,
      "num_input_tokens_seen": 1747200,
      "step": 175
    },
    {
      "epoch": 0.09271182075714654,
      "grad_norm": 3.6326401692458146,
      "learning_rate": 9.999763499923432e-05,
      "loss": 0.0704,
      "num_input_tokens_seen": 1797120,
      "step": 180
    },
    {
      "epoch": 0.09528714911151172,
      "grad_norm": 0.4999720825488852,
      "learning_rate": 9.999467880071402e-05,
      "loss": 0.0278,
      "num_input_tokens_seen": 1847040,
      "step": 185
    },
    {
      "epoch": 0.0978624774658769,
      "grad_norm": 0.5211964505880501,
      "learning_rate": 9.999054022066641e-05,
      "loss": 0.0862,
      "num_input_tokens_seen": 1896960,
      "step": 190
    },
    {
      "epoch": 0.10043780582024209,
      "grad_norm": 0.8767038751832389,
      "learning_rate": 9.998521935696953e-05,
      "loss": 0.0565,
      "num_input_tokens_seen": 1946880,
      "step": 195
    },
    {
      "epoch": 0.10301313417460727,
      "grad_norm": 0.1848827361202722,
      "learning_rate": 9.997871633546257e-05,
      "loss": 0.0459,
      "num_input_tokens_seen": 1996800,
      "step": 200
    },
    {
      "epoch": 0.10301313417460727,
      "eval_loss": 0.09957947582006454,
      "eval_runtime": 19.2651,
      "eval_samples_per_second": 3.114,
      "eval_steps_per_second": 0.779,
      "num_input_tokens_seen": 1996800,
      "step": 200
    },
    {
      "epoch": 0.10558846252897244,
      "grad_norm": 1.0969393658164421,
      "learning_rate": 9.997103130994296e-05,
      "loss": 0.0539,
      "num_input_tokens_seen": 2046720,
      "step": 205
    },
    {
      "epoch": 0.10816379088333762,
      "grad_norm": 2.885869194934028,
      "learning_rate": 9.996216446216267e-05,
      "loss": 0.0654,
      "num_input_tokens_seen": 2096640,
      "step": 210
    },
    {
      "epoch": 0.11073911923770281,
      "grad_norm": 0.5225257245731217,
      "learning_rate": 9.995211600182397e-05,
      "loss": 0.0316,
      "num_input_tokens_seen": 2146560,
      "step": 215
    },
    {
      "epoch": 0.11331444759206799,
      "grad_norm": 2.1553510734212797,
      "learning_rate": 9.994088616657444e-05,
      "loss": 0.1169,
      "num_input_tokens_seen": 2196480,
      "step": 220
    },
    {
      "epoch": 0.11588977594643317,
      "grad_norm": 1.1133884703723633,
      "learning_rate": 9.992847522200133e-05,
      "loss": 0.0382,
      "num_input_tokens_seen": 2246400,
      "step": 225
    },
    {
      "epoch": 0.11846510430079835,
      "grad_norm": 0.8875243341616034,
      "learning_rate": 9.99148834616253e-05,
      "loss": 0.0406,
      "num_input_tokens_seen": 2296320,
      "step": 230
    },
    {
      "epoch": 0.12104043265516354,
      "grad_norm": 1.81283533812695,
      "learning_rate": 9.990011120689351e-05,
      "loss": 0.0182,
      "num_input_tokens_seen": 2346240,
      "step": 235
    },
    {
      "epoch": 0.12361576100952872,
      "grad_norm": 3.873083258671571,
      "learning_rate": 9.988415880717194e-05,
      "loss": 0.0881,
      "num_input_tokens_seen": 2396160,
      "step": 240
    },
    {
      "epoch": 0.1261910893638939,
      "grad_norm": 3.427761103620865,
      "learning_rate": 9.986702663973722e-05,
      "loss": 0.0565,
      "num_input_tokens_seen": 2446080,
      "step": 245
    },
    {
      "epoch": 0.12876641771825909,
      "grad_norm": 1.531943599765959,
      "learning_rate": 9.98487151097676e-05,
      "loss": 0.0805,
      "num_input_tokens_seen": 2496000,
      "step": 250
    },
    {
      "epoch": 0.12876641771825909,
      "eval_loss": 0.055875860154628754,
      "eval_runtime": 19.5106,
      "eval_samples_per_second": 3.075,
      "eval_steps_per_second": 0.769,
      "num_input_tokens_seen": 2496000,
      "step": 250
    },
    {
      "epoch": 0.13134174607262425,
      "grad_norm": 0.845514751132877,
      "learning_rate": 9.98292246503335e-05,
      "loss": 0.0235,
      "num_input_tokens_seen": 2545920,
      "step": 255
    },
    {
      "epoch": 0.13391707442698944,
      "grad_norm": 1.7441538057875345,
      "learning_rate": 9.980855572238714e-05,
      "loss": 0.0137,
      "num_input_tokens_seen": 2595840,
      "step": 260
    },
    {
      "epoch": 0.13649240278135463,
      "grad_norm": 0.6628270146260998,
      "learning_rate": 9.978670881475172e-05,
      "loss": 0.0605,
      "num_input_tokens_seen": 2645760,
      "step": 265
    },
    {
      "epoch": 0.1390677311357198,
      "grad_norm": 4.49926360488578,
      "learning_rate": 9.976368444410985e-05,
      "loss": 0.0585,
      "num_input_tokens_seen": 2695680,
      "step": 270
    },
    {
      "epoch": 0.141643059490085,
      "grad_norm": 0.7159498856832776,
      "learning_rate": 9.973948315499126e-05,
      "loss": 0.0096,
      "num_input_tokens_seen": 2745600,
      "step": 275
    },
    {
      "epoch": 0.14421838784445018,
      "grad_norm": 4.09842502301765,
      "learning_rate": 9.971410551976002e-05,
      "loss": 0.1383,
      "num_input_tokens_seen": 2795520,
      "step": 280
    },
    {
      "epoch": 0.14679371619881534,
      "grad_norm": 1.0773897808674016,
      "learning_rate": 9.968755213860094e-05,
      "loss": 0.049,
      "num_input_tokens_seen": 2845440,
      "step": 285
    },
    {
      "epoch": 0.14936904455318054,
      "grad_norm": 1.7224091076306651,
      "learning_rate": 9.96598236395054e-05,
      "loss": 0.0336,
      "num_input_tokens_seen": 2895360,
      "step": 290
    },
    {
      "epoch": 0.1519443729075457,
      "grad_norm": 0.24158350585209873,
      "learning_rate": 9.96309206782565e-05,
      "loss": 0.019,
      "num_input_tokens_seen": 2945280,
      "step": 295
    },
    {
      "epoch": 0.1545197012619109,
      "grad_norm": 0.6614271496466305,
      "learning_rate": 9.960084393841355e-05,
      "loss": 0.0381,
      "num_input_tokens_seen": 2995200,
      "step": 300
    },
    {
      "epoch": 0.1545197012619109,
      "eval_loss": 0.030908752232789993,
      "eval_runtime": 19.8702,
      "eval_samples_per_second": 3.02,
      "eval_steps_per_second": 0.755,
      "num_input_tokens_seen": 2995200,
      "step": 300
    },
    {
      "epoch": 0.15709502961627608,
      "grad_norm": 0.2314383428097956,
      "learning_rate": 9.956959413129585e-05,
      "loss": 0.0245,
      "num_input_tokens_seen": 3045120,
      "step": 305
    },
    {
      "epoch": 0.15967035797064125,
      "grad_norm": 0.2035043692790888,
      "learning_rate": 9.953717199596598e-05,
      "loss": 0.0121,
      "num_input_tokens_seen": 3095040,
      "step": 310
    },
    {
      "epoch": 0.16224568632500644,
      "grad_norm": 0.014152925203425593,
      "learning_rate": 9.95035782992122e-05,
      "loss": 0.054,
      "num_input_tokens_seen": 3144960,
      "step": 315
    },
    {
      "epoch": 0.16482101467937163,
      "grad_norm": 1.0162363271291994,
      "learning_rate": 9.94688138355304e-05,
      "loss": 0.0849,
      "num_input_tokens_seen": 3194880,
      "step": 320
    },
    {
      "epoch": 0.1673963430337368,
      "grad_norm": 3.92714662464831,
      "learning_rate": 9.943287942710527e-05,
      "loss": 0.0503,
      "num_input_tokens_seen": 3244800,
      "step": 325
    },
    {
      "epoch": 0.16997167138810199,
      "grad_norm": 0.47450325213422423,
      "learning_rate": 9.939577592379088e-05,
      "loss": 0.0302,
      "num_input_tokens_seen": 3294720,
      "step": 330
    },
    {
      "epoch": 0.17254699974246718,
      "grad_norm": 1.6706834486788253,
      "learning_rate": 9.935750420309055e-05,
      "loss": 0.082,
      "num_input_tokens_seen": 3344640,
      "step": 335
    },
    {
      "epoch": 0.17512232809683234,
      "grad_norm": 0.16856461555999938,
      "learning_rate": 9.931806517013612e-05,
      "loss": 0.0411,
      "num_input_tokens_seen": 3394560,
      "step": 340
    },
    {
      "epoch": 0.17769765645119753,
      "grad_norm": 8.475859938133665,
      "learning_rate": 9.927745975766654e-05,
      "loss": 0.1158,
      "num_input_tokens_seen": 3444480,
      "step": 345
    },
    {
      "epoch": 0.1802729848055627,
      "grad_norm": 0.7891476176494399,
      "learning_rate": 9.923568892600578e-05,
      "loss": 0.1761,
      "num_input_tokens_seen": 3494400,
      "step": 350
    },
    {
      "epoch": 0.1802729848055627,
      "eval_loss": 0.043948542326688766,
      "eval_runtime": 19.3589,
      "eval_samples_per_second": 3.099,
      "eval_steps_per_second": 0.775,
      "num_input_tokens_seen": 3494400,
      "step": 350
    },
    {
      "epoch": 0.1828483131599279,
      "grad_norm": 0.3217035938609153,
      "learning_rate": 9.91927536630402e-05,
      "loss": 0.0267,
      "num_input_tokens_seen": 3544320,
      "step": 355
    },
    {
      "epoch": 0.18542364151429308,
      "grad_norm": 1.9146681345367045,
      "learning_rate": 9.91486549841951e-05,
      "loss": 0.0642,
      "num_input_tokens_seen": 3594240,
      "step": 360
    },
    {
      "epoch": 0.18799896986865824,
      "grad_norm": 0.053067921944857786,
      "learning_rate": 9.91033939324107e-05,
      "loss": 0.0302,
      "num_input_tokens_seen": 3644160,
      "step": 365
    },
    {
      "epoch": 0.19057429822302344,
      "grad_norm": 0.07277339857890938,
      "learning_rate": 9.905697157811761e-05,
      "loss": 0.0099,
      "num_input_tokens_seen": 3694080,
      "step": 370
    },
    {
      "epoch": 0.19314962657738863,
      "grad_norm": 4.117705571576552,
      "learning_rate": 9.900938901921131e-05,
      "loss": 0.0573,
      "num_input_tokens_seen": 3744000,
      "step": 375
    },
    {
      "epoch": 0.1957249549317538,
      "grad_norm": 2.4479374171708304,
      "learning_rate": 9.896064738102635e-05,
      "loss": 0.0302,
      "num_input_tokens_seen": 3793920,
      "step": 380
    },
    {
      "epoch": 0.19830028328611898,
      "grad_norm": 2.1751552571638153,
      "learning_rate": 9.891074781630966e-05,
      "loss": 0.0374,
      "num_input_tokens_seen": 3843840,
      "step": 385
    },
    {
      "epoch": 0.20087561164048418,
      "grad_norm": 1.4138134932494708,
      "learning_rate": 9.885969150519331e-05,
      "loss": 0.0227,
      "num_input_tokens_seen": 3893760,
      "step": 390
    },
    {
      "epoch": 0.20345093999484934,
      "grad_norm": 0.3312762825665537,
      "learning_rate": 9.88074796551666e-05,
      "loss": 0.0104,
      "num_input_tokens_seen": 3943680,
      "step": 395
    },
    {
      "epoch": 0.20602626834921453,
      "grad_norm": 0.06774783590394957,
      "learning_rate": 9.875411350104744e-05,
      "loss": 0.0146,
      "num_input_tokens_seen": 3993600,
      "step": 400
    },
    {
      "epoch": 0.20602626834921453,
      "eval_loss": 0.024433813989162445,
      "eval_runtime": 18.8143,
      "eval_samples_per_second": 3.189,
      "eval_steps_per_second": 0.797,
      "num_input_tokens_seen": 3993600,
      "step": 400
    },
    {
      "epoch": 0.2086015967035797,
      "grad_norm": 0.06416132017781931,
      "learning_rate": 9.86995943049533e-05,
      "loss": 0.0074,
      "num_input_tokens_seen": 4043520,
      "step": 405
    },
    {
      "epoch": 0.2111769250579449,
      "grad_norm": 0.04809215529224866,
      "learning_rate": 9.864392335627117e-05,
      "loss": 0.0375,
      "num_input_tokens_seen": 4093440,
      "step": 410
    },
    {
      "epoch": 0.21375225341231008,
      "grad_norm": 0.40943321947836153,
      "learning_rate": 9.858710197162721e-05,
      "loss": 0.0217,
      "num_input_tokens_seen": 4143360,
      "step": 415
    },
    {
      "epoch": 0.21632758176667524,
      "grad_norm": 0.08069011480339518,
      "learning_rate": 9.852913149485556e-05,
      "loss": 0.016,
      "num_input_tokens_seen": 4193280,
      "step": 420
    },
    {
      "epoch": 0.21890291012104043,
      "grad_norm": 1.806585526467194,
      "learning_rate": 9.847001329696653e-05,
      "loss": 0.0832,
      "num_input_tokens_seen": 4243200,
      "step": 425
    },
    {
      "epoch": 0.22147823847540563,
      "grad_norm": 3.122914329597603,
      "learning_rate": 9.840974877611422e-05,
      "loss": 0.0444,
      "num_input_tokens_seen": 4293120,
      "step": 430
    },
    {
      "epoch": 0.2240535668297708,
      "grad_norm": 1.0101510644418257,
      "learning_rate": 9.834833935756344e-05,
      "loss": 0.0465,
      "num_input_tokens_seen": 4343040,
      "step": 435
    },
    {
      "epoch": 0.22662889518413598,
      "grad_norm": 1.7770164472545809,
      "learning_rate": 9.828578649365601e-05,
      "loss": 0.0428,
      "num_input_tokens_seen": 4392960,
      "step": 440
    },
    {
      "epoch": 0.22920422353850115,
      "grad_norm": 0.9798524063329249,
      "learning_rate": 9.822209166377635e-05,
      "loss": 0.02,
      "num_input_tokens_seen": 4442880,
      "step": 445
    },
    {
      "epoch": 0.23177955189286634,
      "grad_norm": 0.22305429096692395,
      "learning_rate": 9.815725637431662e-05,
      "loss": 0.0157,
      "num_input_tokens_seen": 4492800,
      "step": 450
    },
    {
      "epoch": 0.23177955189286634,
      "eval_loss": 0.006690301466733217,
      "eval_runtime": 18.7725,
      "eval_samples_per_second": 3.196,
      "eval_steps_per_second": 0.799,
      "num_input_tokens_seen": 4492800,
      "step": 450
    },
    {
      "epoch": 0.23435488024723153,
      "grad_norm": 2.8137229294991064,
      "learning_rate": 9.809128215864097e-05,
      "loss": 0.0596,
      "num_input_tokens_seen": 4542720,
      "step": 455
    },
    {
      "epoch": 0.2369302086015967,
      "grad_norm": 0.045182426772616975,
      "learning_rate": 9.802417057704931e-05,
      "loss": 0.0238,
      "num_input_tokens_seen": 4592640,
      "step": 460
    },
    {
      "epoch": 0.23950553695596188,
      "grad_norm": 2.1659990503080424,
      "learning_rate": 9.795592321674045e-05,
      "loss": 0.0331,
      "num_input_tokens_seen": 4642560,
      "step": 465
    },
    {
      "epoch": 0.24208086531032708,
      "grad_norm": 0.3208179936186945,
      "learning_rate": 9.788654169177453e-05,
      "loss": 0.0177,
      "num_input_tokens_seen": 4692480,
      "step": 470
    },
    {
      "epoch": 0.24465619366469224,
      "grad_norm": 1.4150868022870595,
      "learning_rate": 9.781602764303487e-05,
      "loss": 0.028,
      "num_input_tokens_seen": 4742400,
      "step": 475
    },
    {
      "epoch": 0.24723152201905743,
      "grad_norm": 0.21015675310797993,
      "learning_rate": 9.774438273818911e-05,
      "loss": 0.0561,
      "num_input_tokens_seen": 4792320,
      "step": 480
    },
    {
      "epoch": 0.24980685037342262,
      "grad_norm": 2.6788940269709287,
      "learning_rate": 9.767160867164979e-05,
      "loss": 0.0242,
      "num_input_tokens_seen": 4842240,
      "step": 485
    },
    {
      "epoch": 0.2523821787277878,
      "grad_norm": 0.03933517764434036,
      "learning_rate": 9.759770716453436e-05,
      "loss": 0.0286,
      "num_input_tokens_seen": 4892160,
      "step": 490
    },
    {
      "epoch": 0.254957507082153,
      "grad_norm": 2.640299598300008,
      "learning_rate": 9.752267996462434e-05,
      "loss": 0.0394,
      "num_input_tokens_seen": 4942080,
      "step": 495
    },
    {
      "epoch": 0.25753283543651817,
      "grad_norm": 0.03143215150024672,
      "learning_rate": 9.744652884632406e-05,
      "loss": 0.0122,
      "num_input_tokens_seen": 4992000,
      "step": 500
    },
    {
      "epoch": 0.25753283543651817,
      "eval_loss": 0.007972972467541695,
      "eval_runtime": 18.8509,
      "eval_samples_per_second": 3.183,
      "eval_steps_per_second": 0.796,
      "num_input_tokens_seen": 4992000,
      "step": 500
    },
    {
      "epoch": 0.26010816379088336,
      "grad_norm": 0.3421379771468272,
      "learning_rate": 9.736925561061871e-05,
      "loss": 0.0184,
      "num_input_tokens_seen": 5041920,
      "step": 505
    },
    {
      "epoch": 0.2626834921452485,
      "grad_norm": 0.16336037200955245,
      "learning_rate": 9.729086208503174e-05,
      "loss": 0.0046,
      "num_input_tokens_seen": 5091840,
      "step": 510
    },
    {
      "epoch": 0.2652588204996137,
      "grad_norm": 0.7318312729236729,
      "learning_rate": 9.721135012358156e-05,
      "loss": 0.0226,
      "num_input_tokens_seen": 5141760,
      "step": 515
    },
    {
      "epoch": 0.2678341488539789,
      "grad_norm": 2.1703341770378333,
      "learning_rate": 9.713072160673777e-05,
      "loss": 0.0475,
      "num_input_tokens_seen": 5191680,
      "step": 520
    },
    {
      "epoch": 0.2704094772083441,
      "grad_norm": 1.6372493711976897,
      "learning_rate": 9.704897844137673e-05,
      "loss": 0.0572,
      "num_input_tokens_seen": 5241600,
      "step": 525
    },
    {
      "epoch": 0.27298480556270927,
      "grad_norm": 3.5840623375725804,
      "learning_rate": 9.696612256073633e-05,
      "loss": 0.025,
      "num_input_tokens_seen": 5291520,
      "step": 530
    },
    {
      "epoch": 0.2755601339170744,
      "grad_norm": 0.06141348661076509,
      "learning_rate": 9.688215592437039e-05,
      "loss": 0.0122,
      "num_input_tokens_seen": 5341440,
      "step": 535
    },
    {
      "epoch": 0.2781354622714396,
      "grad_norm": 0.04980492840656424,
      "learning_rate": 9.679708051810221e-05,
      "loss": 0.0045,
      "num_input_tokens_seen": 5391360,
      "step": 540
    },
    {
      "epoch": 0.2807107906258048,
      "grad_norm": 3.153079438616822,
      "learning_rate": 9.67108983539777e-05,
      "loss": 0.0431,
      "num_input_tokens_seen": 5441280,
      "step": 545
    },
    {
      "epoch": 0.28328611898017,
      "grad_norm": 1.6764625232464418,
      "learning_rate": 9.662361147021779e-05,
      "loss": 0.0339,
      "num_input_tokens_seen": 5491200,
      "step": 550
    },
    {
      "epoch": 0.28328611898017,
      "eval_loss": 0.0033526704646646976,
      "eval_runtime": 18.8387,
      "eval_samples_per_second": 3.185,
      "eval_steps_per_second": 0.796,
      "num_input_tokens_seen": 5491200,
      "step": 550
    },
    {
      "epoch": 0.28586144733453517,
      "grad_norm": 0.05427488923085365,
      "learning_rate": 9.653522193117013e-05,
      "loss": 0.0341,
      "num_input_tokens_seen": 5541120,
      "step": 555
    },
    {
      "epoch": 0.28843677568890036,
      "grad_norm": 1.538376531568411,
      "learning_rate": 9.644573182726035e-05,
      "loss": 0.0341,
      "num_input_tokens_seen": 5591040,
      "step": 560
    },
    {
      "epoch": 0.2910121040432655,
      "grad_norm": 0.04402865983757039,
      "learning_rate": 9.63551432749426e-05,
      "loss": 0.0453,
      "num_input_tokens_seen": 5640960,
      "step": 565
    },
    {
      "epoch": 0.2935874323976307,
      "grad_norm": 0.1469885874710427,
      "learning_rate": 9.626345841664953e-05,
      "loss": 0.0227,
      "num_input_tokens_seen": 5690880,
      "step": 570
    },
    {
      "epoch": 0.2961627607519959,
      "grad_norm": 0.3459204695983954,
      "learning_rate": 9.617067942074153e-05,
      "loss": 0.0135,
      "num_input_tokens_seen": 5740800,
      "step": 575
    },
    {
      "epoch": 0.29873808910636107,
      "grad_norm": 1.7604264611882452,
      "learning_rate": 9.607680848145558e-05,
      "loss": 0.0439,
      "num_input_tokens_seen": 5790720,
      "step": 580
    },
    {
      "epoch": 0.30131341746072626,
      "grad_norm": 2.1661707703583204,
      "learning_rate": 9.598184781885318e-05,
      "loss": 0.0344,
      "num_input_tokens_seen": 5840640,
      "step": 585
    },
    {
      "epoch": 0.3038887458150914,
      "grad_norm": 0.19674934066326408,
      "learning_rate": 9.588579967876806e-05,
      "loss": 0.0244,
      "num_input_tokens_seen": 5890560,
      "step": 590
    },
    {
      "epoch": 0.3064640741694566,
      "grad_norm": 4.463551986915384,
      "learning_rate": 9.578866633275288e-05,
      "loss": 0.022,
      "num_input_tokens_seen": 5940480,
      "step": 595
    },
    {
      "epoch": 0.3090394025238218,
      "grad_norm": 0.08358626804613765,
      "learning_rate": 9.569045007802559e-05,
      "loss": 0.0217,
      "num_input_tokens_seen": 5990400,
      "step": 600
    },
    {
      "epoch": 0.3090394025238218,
      "eval_loss": 0.013290103524923325,
      "eval_runtime": 18.8576,
      "eval_samples_per_second": 3.182,
      "eval_steps_per_second": 0.795,
      "num_input_tokens_seen": 5990400,
      "step": 600
    },
    {
      "epoch": 0.311614730878187,
      "grad_norm": 0.06328082378687645,
      "learning_rate": 9.55911532374151e-05,
      "loss": 0.0207,
      "num_input_tokens_seen": 6040320,
      "step": 605
    },
    {
      "epoch": 0.31419005923255217,
      "grad_norm": 1.846961607001629,
      "learning_rate": 9.549077815930636e-05,
      "loss": 0.023,
      "num_input_tokens_seen": 6090240,
      "step": 610
    },
    {
      "epoch": 0.31676538758691736,
      "grad_norm": 0.07122251919235507,
      "learning_rate": 9.538932721758474e-05,
      "loss": 0.017,
      "num_input_tokens_seen": 6140160,
      "step": 615
    },
    {
      "epoch": 0.3193407159412825,
      "grad_norm": 1.9564639611342651,
      "learning_rate": 9.528680281157999e-05,
      "loss": 0.009,
      "num_input_tokens_seen": 6190080,
      "step": 620
    },
    {
      "epoch": 0.3219160442956477,
      "grad_norm": 0.43628623054662674,
      "learning_rate": 9.518320736600943e-05,
      "loss": 0.0421,
      "num_input_tokens_seen": 6240000,
      "step": 625
    },
    {
      "epoch": 0.3244913726500129,
      "grad_norm": 0.7626652394047067,
      "learning_rate": 9.507854333092063e-05,
      "loss": 0.0324,
      "num_input_tokens_seen": 6289920,
      "step": 630
    },
    {
      "epoch": 0.32706670100437807,
      "grad_norm": 0.5823990007842583,
      "learning_rate": 9.497281318163346e-05,
      "loss": 0.0139,
      "num_input_tokens_seen": 6339840,
      "step": 635
    },
    {
      "epoch": 0.32964202935874326,
      "grad_norm": 1.6726620150276597,
      "learning_rate": 9.486601941868154e-05,
      "loss": 0.0375,
      "num_input_tokens_seen": 6389760,
      "step": 640
    },
    {
      "epoch": 0.3322173577131084,
      "grad_norm": 0.08456122780891835,
      "learning_rate": 9.475816456775313e-05,
      "loss": 0.0618,
      "num_input_tokens_seen": 6439680,
      "step": 645
    },
    {
      "epoch": 0.3347926860674736,
      "grad_norm": 0.3287256358007373,
      "learning_rate": 9.464925117963133e-05,
      "loss": 0.0327,
      "num_input_tokens_seen": 6489600,
      "step": 650
    },
    {
      "epoch": 0.3347926860674736,
      "eval_loss": 0.02095886506140232,
      "eval_runtime": 19.1993,
      "eval_samples_per_second": 3.125,
      "eval_steps_per_second": 0.781,
      "num_input_tokens_seen": 6489600,
      "step": 650
    },
    {
      "epoch": 0.3373680144218388,
      "grad_norm": 1.642912250249591,
      "learning_rate": 9.453928183013385e-05,
      "loss": 0.037,
      "num_input_tokens_seen": 6539520,
      "step": 655
    },
    {
      "epoch": 0.33994334277620397,
      "grad_norm": 1.6829194140719153,
      "learning_rate": 9.442825912005202e-05,
      "loss": 0.034,
      "num_input_tokens_seen": 6589440,
      "step": 660
    },
    {
      "epoch": 0.34251867113056916,
      "grad_norm": 1.7722553129499397,
      "learning_rate": 9.431618567508933e-05,
      "loss": 0.0584,
      "num_input_tokens_seen": 6639360,
      "step": 665
    },
    {
      "epoch": 0.34509399948493436,
      "grad_norm": 1.8398383207694327,
      "learning_rate": 9.420306414579925e-05,
      "loss": 0.059,
      "num_input_tokens_seen": 6689280,
      "step": 670
    },
    {
      "epoch": 0.3476693278392995,
      "grad_norm": 0.2451231053960763,
      "learning_rate": 9.408889720752266e-05,
      "loss": 0.023,
      "num_input_tokens_seen": 6739200,
      "step": 675
    },
    {
      "epoch": 0.3502446561936647,
      "grad_norm": 0.13166014546250449,
      "learning_rate": 9.397368756032445e-05,
      "loss": 0.0327,
      "num_input_tokens_seen": 6789120,
      "step": 680
    },
    {
      "epoch": 0.3528199845480299,
      "grad_norm": 1.9392864055798256,
      "learning_rate": 9.385743792892982e-05,
      "loss": 0.0241,
      "num_input_tokens_seen": 6839040,
      "step": 685
    },
    {
      "epoch": 0.35539531290239507,
      "grad_norm": 0.7574757741852987,
      "learning_rate": 9.374015106265968e-05,
      "loss": 0.0227,
      "num_input_tokens_seen": 6888960,
      "step": 690
    },
    {
      "epoch": 0.35797064125676026,
      "grad_norm": 1.1649279233149774,
      "learning_rate": 9.362182973536569e-05,
      "loss": 0.0171,
      "num_input_tokens_seen": 6938880,
      "step": 695
    },
    {
      "epoch": 0.3605459696111254,
      "grad_norm": 0.04561814221227265,
      "learning_rate": 9.35024767453647e-05,
      "loss": 0.0267,
      "num_input_tokens_seen": 6988800,
      "step": 700
    },
    {
      "epoch": 0.3605459696111254,
      "eval_loss": 0.005286735948175192,
      "eval_runtime": 19.2824,
      "eval_samples_per_second": 3.112,
      "eval_steps_per_second": 0.778,
      "num_input_tokens_seen": 6988800,
      "step": 700
    },
    {
      "epoch": 0.3631212979654906,
      "grad_norm": 1.7432145241229078,
      "learning_rate": 9.338209491537257e-05,
      "loss": 0.0335,
      "num_input_tokens_seen": 7038720,
      "step": 705
    },
    {
      "epoch": 0.3656966263198558,
      "grad_norm": 1.821635752664896,
      "learning_rate": 9.326068709243727e-05,
      "loss": 0.0108,
      "num_input_tokens_seen": 7088640,
      "step": 710
    },
    {
      "epoch": 0.36827195467422097,
      "grad_norm": 0.5782878283043246,
      "learning_rate": 9.313825614787177e-05,
      "loss": 0.0296,
      "num_input_tokens_seen": 7138560,
      "step": 715
    },
    {
      "epoch": 0.37084728302858616,
      "grad_norm": 2.0568373189070908,
      "learning_rate": 9.301480497718593e-05,
      "loss": 0.042,
      "num_input_tokens_seen": 7188480,
      "step": 720
    },
    {
      "epoch": 0.37342261138295135,
      "grad_norm": 0.20623581315280795,
      "learning_rate": 9.289033650001817e-05,
      "loss": 0.0083,
      "num_input_tokens_seen": 7238400,
      "step": 725
    },
    {
      "epoch": 0.3759979397373165,
      "grad_norm": 1.424505040979842,
      "learning_rate": 9.276485366006634e-05,
      "loss": 0.0109,
      "num_input_tokens_seen": 7288320,
      "step": 730
    },
    {
      "epoch": 0.3785732680916817,
      "grad_norm": 2.1248350485820584,
      "learning_rate": 9.263835942501807e-05,
      "loss": 0.0122,
      "num_input_tokens_seen": 7338240,
      "step": 735
    },
    {
      "epoch": 0.3811485964460469,
      "grad_norm": 0.357888255955218,
      "learning_rate": 9.251085678648072e-05,
      "loss": 0.017,
      "num_input_tokens_seen": 7388160,
      "step": 740
    },
    {
      "epoch": 0.38372392480041206,
      "grad_norm": 3.859418562917624,
      "learning_rate": 9.238234875991046e-05,
      "loss": 0.0236,
      "num_input_tokens_seen": 7438080,
      "step": 745
    },
    {
      "epoch": 0.38629925315477726,
      "grad_norm": 2.660896728773248,
      "learning_rate": 9.225283838454111e-05,
      "loss": 0.014,
      "num_input_tokens_seen": 7488000,
      "step": 750
    },
    {
      "epoch": 0.38629925315477726,
      "eval_loss": 0.005334165878593922,
      "eval_runtime": 19.3986,
      "eval_samples_per_second": 3.093,
      "eval_steps_per_second": 0.773,
      "num_input_tokens_seen": 7488000,
      "step": 750
    },
    {
      "epoch": 0.3888745815091424,
      "grad_norm": 3.511127189174474,
      "learning_rate": 9.21223287233121e-05,
      "loss": 0.0426,
      "num_input_tokens_seen": 7537920,
      "step": 755
    },
    {
      "epoch": 0.3914499098635076,
      "grad_norm": 2.858952603531337,
      "learning_rate": 9.199082286279622e-05,
      "loss": 0.0094,
      "num_input_tokens_seen": 7587840,
      "step": 760
    },
    {
      "epoch": 0.3940252382178728,
      "grad_norm": 1.9568584051828664,
      "learning_rate": 9.185832391312644e-05,
      "loss": 0.008,
      "num_input_tokens_seen": 7637760,
      "step": 765
    },
    {
      "epoch": 0.39660056657223797,
      "grad_norm": 1.7485145115139782,
      "learning_rate": 9.172483500792244e-05,
      "loss": 0.0297,
      "num_input_tokens_seen": 7687680,
      "step": 770
    },
    {
      "epoch": 0.39917589492660316,
      "grad_norm": 0.05161945290023201,
      "learning_rate": 9.159035930421658e-05,
      "loss": 0.0239,
      "num_input_tokens_seen": 7737600,
      "step": 775
    },
    {
      "epoch": 0.40175122328096835,
      "grad_norm": 5.343080934932771,
      "learning_rate": 9.145489998237902e-05,
      "loss": 0.0162,
      "num_input_tokens_seen": 7787520,
      "step": 780
    },
    {
      "epoch": 0.4043265516353335,
      "grad_norm": 0.020924688334471742,
      "learning_rate": 9.131846024604274e-05,
      "loss": 0.0309,
      "num_input_tokens_seen": 7837440,
      "step": 785
    },
    {
      "epoch": 0.4069018799896987,
      "grad_norm": 0.5607043400469908,
      "learning_rate": 9.11810433220276e-05,
      "loss": 0.0432,
      "num_input_tokens_seen": 7887360,
      "step": 790
    },
    {
      "epoch": 0.40947720834406387,
      "grad_norm": 0.05221845282486231,
      "learning_rate": 9.104265246026415e-05,
      "loss": 0.0022,
      "num_input_tokens_seen": 7937280,
      "step": 795
    },
    {
      "epoch": 0.41205253669842906,
      "grad_norm": 0.09825330413354653,
      "learning_rate": 9.090329093371666e-05,
      "loss": 0.0065,
      "num_input_tokens_seen": 7987200,
      "step": 800
    },
    {
      "epoch": 0.41205253669842906,
      "eval_loss": 0.006797688081860542,
      "eval_runtime": 19.3386,
      "eval_samples_per_second": 3.103,
      "eval_steps_per_second": 0.776,
      "num_input_tokens_seen": 7987200,
      "step": 800
    },
    {
      "epoch": 0.41462786505279425,
      "grad_norm": 0.016984614729716928,
      "learning_rate": 9.076296203830579e-05,
      "loss": 0.0025,
      "num_input_tokens_seen": 8037120,
      "step": 805
    },
    {
      "epoch": 0.4172031934071594,
      "grad_norm": 8.87156169406028,
      "learning_rate": 9.062166909283062e-05,
      "loss": 0.0338,
      "num_input_tokens_seen": 8087040,
      "step": 810
    },
    {
      "epoch": 0.4197785217615246,
      "grad_norm": 0.07123492673307173,
      "learning_rate": 9.047941543889014e-05,
      "loss": 0.0505,
      "num_input_tokens_seen": 8136960,
      "step": 815
    },
    {
      "epoch": 0.4223538501158898,
      "grad_norm": 0.3486611879887143,
      "learning_rate": 9.033620444080428e-05,
      "loss": 0.0335,
      "num_input_tokens_seen": 8186880,
      "step": 820
    },
    {
      "epoch": 0.42492917847025496,
      "grad_norm": 0.2658900406741178,
      "learning_rate": 9.019203948553422e-05,
      "loss": 0.0196,
      "num_input_tokens_seen": 8236800,
      "step": 825
    },
    {
      "epoch": 0.42750450682462016,
      "grad_norm": 0.15327530188348007,
      "learning_rate": 9.004692398260244e-05,
      "loss": 0.018,
      "num_input_tokens_seen": 8286720,
      "step": 830
    },
    {
      "epoch": 0.43007983517898535,
      "grad_norm": 5.292901102191953,
      "learning_rate": 8.9900861364012e-05,
      "loss": 0.012,
      "num_input_tokens_seen": 8336640,
      "step": 835
    },
    {
      "epoch": 0.4326551635333505,
      "grad_norm": 3.2113529029102375,
      "learning_rate": 8.975385508416532e-05,
      "loss": 0.0062,
      "num_input_tokens_seen": 8386560,
      "step": 840
    },
    {
      "epoch": 0.4352304918877157,
      "grad_norm": 0.3685880212421972,
      "learning_rate": 8.960590861978265e-05,
      "loss": 0.0232,
      "num_input_tokens_seen": 8436480,
      "step": 845
    },
    {
      "epoch": 0.43780582024208087,
      "grad_norm": 1.4858234531184304,
      "learning_rate": 8.945702546981969e-05,
      "loss": 0.0306,
      "num_input_tokens_seen": 8486400,
      "step": 850
    },
    {
      "epoch": 0.43780582024208087,
      "eval_loss": 0.007198736071586609,
      "eval_runtime": 19.7019,
      "eval_samples_per_second": 3.045,
      "eval_steps_per_second": 0.761,
      "num_input_tokens_seen": 8486400,
      "step": 850
    },
    {
      "epoch": 0.44038114859644606,
      "grad_norm": 0.22602261311014887,
      "learning_rate": 8.930720915538487e-05,
      "loss": 0.0255,
      "num_input_tokens_seen": 8536320,
      "step": 855
    },
    {
      "epoch": 0.44295647695081125,
      "grad_norm": 0.1515497727795517,
      "learning_rate": 8.915646321965614e-05,
      "loss": 0.0267,
      "num_input_tokens_seen": 8586240,
      "step": 860
    },
    {
      "epoch": 0.4455318053051764,
      "grad_norm": 0.039053785243136956,
      "learning_rate": 8.900479122779712e-05,
      "loss": 0.0107,
      "num_input_tokens_seen": 8636160,
      "step": 865
    },
    {
      "epoch": 0.4481071336595416,
      "grad_norm": 0.42882717357803835,
      "learning_rate": 8.885219676687277e-05,
      "loss": 0.0277,
      "num_input_tokens_seen": 8686080,
      "step": 870
    },
    {
      "epoch": 0.45068246201390677,
      "grad_norm": 5.519940082555776,
      "learning_rate": 8.869868344576459e-05,
      "loss": 0.0175,
      "num_input_tokens_seen": 8736000,
      "step": 875
    },
    {
      "epoch": 0.45325779036827196,
      "grad_norm": 0.15983443395533375,
      "learning_rate": 8.854425489508532e-05,
      "loss": 0.0065,
      "num_input_tokens_seen": 8785920,
      "step": 880
    },
    {
      "epoch": 0.45583311872263715,
      "grad_norm": 0.42885948370195226,
      "learning_rate": 8.838891476709288e-05,
      "loss": 0.0476,
      "num_input_tokens_seen": 8835840,
      "step": 885
    },
    {
      "epoch": 0.4584084470770023,
      "grad_norm": 0.18209352469474613,
      "learning_rate": 8.823266673560426e-05,
      "loss": 0.018,
      "num_input_tokens_seen": 8885760,
      "step": 890
    },
    {
      "epoch": 0.4609837754313675,
      "grad_norm": 1.7895291298260116,
      "learning_rate": 8.807551449590846e-05,
      "loss": 0.0337,
      "num_input_tokens_seen": 8935680,
      "step": 895
    },
    {
      "epoch": 0.4635591037857327,
      "grad_norm": 0.11579050055375595,
      "learning_rate": 8.791746176467907e-05,
      "loss": 0.0063,
      "num_input_tokens_seen": 8985600,
      "step": 900
    },
    {
      "epoch": 0.4635591037857327,
      "eval_loss": 0.010749292559921741,
      "eval_runtime": 19.2918,
      "eval_samples_per_second": 3.11,
      "eval_steps_per_second": 0.778,
      "num_input_tokens_seen": 8985600,
      "step": 900
    },
    {
      "epoch": 0.46613443214009787,
      "grad_norm": 2.090931980208863,
      "learning_rate": 8.775851227988656e-05,
      "loss": 0.0318,
      "num_input_tokens_seen": 9035520,
      "step": 905
    },
    {
      "epoch": 0.46870976049446306,
      "grad_norm": 1.7373295104389102,
      "learning_rate": 8.759866980070963e-05,
      "loss": 0.0635,
      "num_input_tokens_seen": 9085440,
      "step": 910
    },
    {
      "epoch": 0.47128508884882825,
      "grad_norm": 0.1557668148789241,
      "learning_rate": 8.743793810744654e-05,
      "loss": 0.0035,
      "num_input_tokens_seen": 9135360,
      "step": 915
    },
    {
      "epoch": 0.4738604172031934,
      "grad_norm": 0.5241949871459053,
      "learning_rate": 8.727632100142551e-05,
      "loss": 0.0047,
      "num_input_tokens_seen": 9185280,
      "step": 920
    },
    {
      "epoch": 0.4764357455575586,
      "grad_norm": 2.329131853129593,
      "learning_rate": 8.711382230491493e-05,
      "loss": 0.0194,
      "num_input_tokens_seen": 9235200,
      "step": 925
    },
    {
      "epoch": 0.47901107391192377,
      "grad_norm": 1.0158880737206768,
      "learning_rate": 8.695044586103296e-05,
      "loss": 0.0234,
      "num_input_tokens_seen": 9285120,
      "step": 930
    },
    {
      "epoch": 0.48158640226628896,
      "grad_norm": 0.016729230547496875,
      "learning_rate": 8.678619553365659e-05,
      "loss": 0.0253,
      "num_input_tokens_seen": 9335040,
      "step": 935
    },
    {
      "epoch": 0.48416173062065415,
      "grad_norm": 2.937987778998638,
      "learning_rate": 8.662107520733027e-05,
      "loss": 0.0191,
      "num_input_tokens_seen": 9384960,
      "step": 940
    },
    {
      "epoch": 0.4867370589750193,
      "grad_norm": 0.20640046184009653,
      "learning_rate": 8.64550887871741e-05,
      "loss": 0.0469,
      "num_input_tokens_seen": 9434880,
      "step": 945
    },
    {
      "epoch": 0.4893123873293845,
      "grad_norm": 1.9952606902760353,
      "learning_rate": 8.628824019879137e-05,
      "loss": 0.0415,
      "num_input_tokens_seen": 9484800,
      "step": 950
    },
    {
      "epoch": 0.4893123873293845,
      "eval_loss": 0.007216573692858219,
      "eval_runtime": 19.2716,
      "eval_samples_per_second": 3.113,
      "eval_steps_per_second": 0.778,
      "num_input_tokens_seen": 9484800,
      "step": 950
    },
    {
      "epoch": 0.49188771568374967,
      "grad_norm": 3.7704976412587468,
      "learning_rate": 8.612053338817581e-05,
      "loss": 0.0393,
      "num_input_tokens_seen": 9534720,
      "step": 955
    },
    {
      "epoch": 0.49446304403811486,
      "grad_norm": 0.2205817340051924,
      "learning_rate": 8.595197232161824e-05,
      "loss": 0.0102,
      "num_input_tokens_seen": 9584640,
      "step": 960
    },
    {
      "epoch": 0.49703837239248005,
      "grad_norm": 1.9646554268692766,
      "learning_rate": 8.578256098561275e-05,
      "loss": 0.0087,
      "num_input_tokens_seen": 9634560,
      "step": 965
    },
    {
      "epoch": 0.49961370074684525,
      "grad_norm": 2.08637430640156,
      "learning_rate": 8.561230338676239e-05,
      "loss": 0.0097,
      "num_input_tokens_seen": 9684480,
      "step": 970
    },
    {
      "epoch": 0.5021890291012104,
      "grad_norm": 0.018874732349478156,
      "learning_rate": 8.544120355168451e-05,
      "loss": 0.0013,
      "num_input_tokens_seen": 9734400,
      "step": 975
    },
    {
      "epoch": 0.5047643574555756,
      "grad_norm": 0.007747713318690538,
      "learning_rate": 8.526926552691544e-05,
      "loss": 0.0268,
      "num_input_tokens_seen": 9784320,
      "step": 980
    },
    {
      "epoch": 0.5073396858099408,
      "grad_norm": 4.2765032076143585,
      "learning_rate": 8.509649337881483e-05,
      "loss": 0.0418,
      "num_input_tokens_seen": 9834240,
      "step": 985
    },
    {
      "epoch": 0.509915014164306,
      "grad_norm": 0.016725809588179666,
      "learning_rate": 8.492289119346943e-05,
      "loss": 0.0011,
      "num_input_tokens_seen": 9884160,
      "step": 990
    },
    {
      "epoch": 0.5124903425186711,
      "grad_norm": 1.5189362175317798,
      "learning_rate": 8.474846307659658e-05,
      "loss": 0.0058,
      "num_input_tokens_seen": 9934080,
      "step": 995
    },
    {
      "epoch": 0.5150656708730363,
      "grad_norm": 0.5873213891971129,
      "learning_rate": 8.457321315344694e-05,
      "loss": 0.0547,
      "num_input_tokens_seen": 9984000,
      "step": 1000
    },
    {
      "epoch": 0.5150656708730363,
      "eval_loss": 0.0007357922149822116,
      "eval_runtime": 19.1915,
      "eval_samples_per_second": 3.126,
      "eval_steps_per_second": 0.782,
      "num_input_tokens_seen": 9984000,
      "step": 1000
    },
    {
      "epoch": 0.5176409992274015,
      "grad_norm": 0.06909777020752475,
      "learning_rate": 8.439714556870704e-05,
      "loss": 0.0042,
      "num_input_tokens_seen": 10033920,
      "step": 1005
    },
    {
      "epoch": 0.5202163275817667,
      "grad_norm": 0.15306455110846962,
      "learning_rate": 8.422026448640124e-05,
      "loss": 0.0419,
      "num_input_tokens_seen": 10083840,
      "step": 1010
    },
    {
      "epoch": 0.5227916559361319,
      "grad_norm": 0.06864588320363613,
      "learning_rate": 8.40425740897932e-05,
      "loss": 0.0313,
      "num_input_tokens_seen": 10133760,
      "step": 1015
    },
    {
      "epoch": 0.525366984290497,
      "grad_norm": 0.041539288845711536,
      "learning_rate": 8.386407858128706e-05,
      "loss": 0.005,
      "num_input_tokens_seen": 10183680,
      "step": 1020
    },
    {
      "epoch": 0.5279423126448622,
      "grad_norm": 3.866462743446183,
      "learning_rate": 8.368478218232787e-05,
      "loss": 0.0171,
      "num_input_tokens_seen": 10233600,
      "step": 1025
    },
    {
      "epoch": 0.5305176409992274,
      "grad_norm": 0.27036285469370147,
      "learning_rate": 8.350468913330192e-05,
      "loss": 0.0019,
      "num_input_tokens_seen": 10283520,
      "step": 1030
    },
    {
      "epoch": 0.5330929693535926,
      "grad_norm": 0.12025617159059553,
      "learning_rate": 8.33238036934364e-05,
      "loss": 0.0256,
      "num_input_tokens_seen": 10333440,
      "step": 1035
    },
    {
      "epoch": 0.5356682977079578,
      "grad_norm": 0.013500323699360073,
      "learning_rate": 8.31421301406986e-05,
      "loss": 0.0019,
      "num_input_tokens_seen": 10383360,
      "step": 1040
    },
    {
      "epoch": 0.5382436260623229,
      "grad_norm": 0.10118623585384687,
      "learning_rate": 8.29596727716949e-05,
      "loss": 0.018,
      "num_input_tokens_seen": 10433280,
      "step": 1045
    },
    {
      "epoch": 0.5408189544166881,
      "grad_norm": 0.23565575572231393,
      "learning_rate": 8.277643590156894e-05,
      "loss": 0.0007,
      "num_input_tokens_seen": 10483200,
      "step": 1050
    },
    {
      "epoch": 0.5408189544166881,
      "eval_loss": 0.05677948147058487,
      "eval_runtime": 19.0806,
      "eval_samples_per_second": 3.145,
      "eval_steps_per_second": 0.786,
      "num_input_tokens_seen": 10483200,
      "step": 1050
    },
    {
      "epoch": 0.5433942827710533,
      "grad_norm": 6.026032426572499,
      "learning_rate": 8.259242386389973e-05,
      "loss": 0.0436,
      "num_input_tokens_seen": 10533120,
      "step": 1055
    },
    {
      "epoch": 0.5459696111254185,
      "grad_norm": 9.168641384066685,
      "learning_rate": 8.240764101059912e-05,
      "loss": 0.0124,
      "num_input_tokens_seen": 10583040,
      "step": 1060
    },
    {
      "epoch": 0.5485449394797837,
      "grad_norm": 0.590720691746965,
      "learning_rate": 8.222209171180883e-05,
      "loss": 0.0155,
      "num_input_tokens_seen": 10632960,
      "step": 1065
    },
    {
      "epoch": 0.5511202678341488,
      "grad_norm": 8.790606959622112,
      "learning_rate": 8.203578035579715e-05,
      "loss": 0.0206,
      "num_input_tokens_seen": 10682880,
      "step": 1070
    },
    {
      "epoch": 0.553695596188514,
      "grad_norm": 0.11045279468612182,
      "learning_rate": 8.184871134885513e-05,
      "loss": 0.0237,
      "num_input_tokens_seen": 10732800,
      "step": 1075
    },
    {
      "epoch": 0.5562709245428792,
      "grad_norm": 0.004997644768678168,
      "learning_rate": 8.166088911519235e-05,
      "loss": 0.0018,
      "num_input_tokens_seen": 10782720,
      "step": 1080
    },
    {
      "epoch": 0.5588462528972444,
      "grad_norm": 2.249874392718823,
      "learning_rate": 8.147231809683236e-05,
      "loss": 0.0296,
      "num_input_tokens_seen": 10832640,
      "step": 1085
    },
    {
      "epoch": 0.5614215812516096,
      "grad_norm": 0.12969877743086014,
      "learning_rate": 8.128300275350756e-05,
      "loss": 0.0006,
      "num_input_tokens_seen": 10882560,
      "step": 1090
    },
    {
      "epoch": 0.5639969096059748,
      "grad_norm": 0.01592203313456207,
      "learning_rate": 8.109294756255375e-05,
      "loss": 0.0004,
      "num_input_tokens_seen": 10932480,
      "step": 1095
    },
    {
      "epoch": 0.56657223796034,
      "grad_norm": 0.048956329384631035,
      "learning_rate": 8.090215701880419e-05,
      "loss": 0.0056,
      "num_input_tokens_seen": 10982400,
      "step": 1100
    },
    {
      "epoch": 0.56657223796034,
      "eval_loss": 0.0003656313638202846,
      "eval_runtime": 19.3051,
      "eval_samples_per_second": 3.108,
      "eval_steps_per_second": 0.777,
      "num_input_tokens_seen": 10982400,
      "step": 1100
    },
    {
      "epoch": 0.5691475663147051,
      "grad_norm": 0.020540774848059624,
      "learning_rate": 8.07106356344834e-05,
      "loss": 0.0005,
      "num_input_tokens_seen": 11032320,
      "step": 1105
    },
    {
      "epoch": 0.5717228946690703,
      "grad_norm": 0.049646985711380834,
      "learning_rate": 8.051838793910038e-05,
      "loss": 0.0008,
      "num_input_tokens_seen": 11082240,
      "step": 1110
    },
    {
      "epoch": 0.5742982230234355,
      "grad_norm": 0.02410302194894019,
      "learning_rate": 8.032541847934146e-05,
      "loss": 0.001,
      "num_input_tokens_seen": 11132160,
      "step": 1115
    },
    {
      "epoch": 0.5768735513778007,
      "grad_norm": 0.03404095257940449,
      "learning_rate": 8.013173181896283e-05,
      "loss": 0.0219,
      "num_input_tokens_seen": 11182080,
      "step": 1120
    },
    {
      "epoch": 0.5794488797321659,
      "grad_norm": 0.0005611895912019751,
      "learning_rate": 7.993733253868256e-05,
      "loss": 0.0,
      "num_input_tokens_seen": 11232000,
      "step": 1125
    },
    {
      "epoch": 0.582024208086531,
      "grad_norm": 0.03678453984585236,
      "learning_rate": 7.974222523607236e-05,
      "loss": 0.0,
      "num_input_tokens_seen": 11281920,
      "step": 1130
    },
    {
      "epoch": 0.5845995364408962,
      "grad_norm": 0.002820815423322981,
      "learning_rate": 7.954641452544865e-05,
      "loss": 0.0001,
      "num_input_tokens_seen": 11331840,
      "step": 1135
    },
    {
      "epoch": 0.5871748647952614,
      "grad_norm": 0.003003123169716893,
      "learning_rate": 7.934990503776363e-05,
      "loss": 0.0,
      "num_input_tokens_seen": 11381760,
      "step": 1140
    },
    {
      "epoch": 0.5897501931496266,
      "grad_norm": 0.020018776153065302,
      "learning_rate": 7.915270142049566e-05,
      "loss": 0.0001,
      "num_input_tokens_seen": 11431680,
      "step": 1145
    },
    {
      "epoch": 0.5923255215039918,
      "grad_norm": 0.0005407295497242103,
      "learning_rate": 7.89548083375394e-05,
      "loss": 0.0127,
      "num_input_tokens_seen": 11481600,
      "step": 1150
    },
    {
      "epoch": 0.5923255215039918,
      "eval_loss": 1.5566551155643538e-05,
      "eval_runtime": 19.3425,
      "eval_samples_per_second": 3.102,
      "eval_steps_per_second": 0.775,
      "num_input_tokens_seen": 11481600,
      "step": 1150
    },
    {
      "epoch": 0.5949008498583569,
      "grad_norm": 0.00046149068264744525,
      "learning_rate": 7.875623046909544e-05,
      "loss": 0.0003,
      "num_input_tokens_seen": 11531520,
      "step": 1155
    },
    {
      "epoch": 0.5974761782127221,
      "grad_norm": 0.0007891686440962277,
      "learning_rate": 7.855697251155967e-05,
      "loss": 0.0,
      "num_input_tokens_seen": 11581440,
      "step": 1160
    },
    {
      "epoch": 0.6000515065670873,
      "grad_norm": 6.055667685356653,
      "learning_rate": 7.835703917741212e-05,
      "loss": 0.0414,
      "num_input_tokens_seen": 11631360,
      "step": 1165
    },
    {
      "epoch": 0.6026268349214525,
      "grad_norm": 3.9367530795344696,
      "learning_rate": 7.81564351951057e-05,
      "loss": 0.02,
      "num_input_tokens_seen": 11681280,
      "step": 1170
    },
    {
      "epoch": 0.6052021632758177,
      "grad_norm": 0.014148361508830113,
      "learning_rate": 7.795516530895414e-05,
      "loss": 0.0002,
      "num_input_tokens_seen": 11731200,
      "step": 1175
    },
    {
      "epoch": 0.6077774916301828,
      "grad_norm": 0.12219412030414617,
      "learning_rate": 7.775323427901993e-05,
      "loss": 0.0026,
      "num_input_tokens_seen": 11781120,
      "step": 1180
    },
    {
      "epoch": 0.610352819984548,
      "grad_norm": 5.108798582105785,
      "learning_rate": 7.755064688100171e-05,
      "loss": 0.0168,
      "num_input_tokens_seen": 11831040,
      "step": 1185
    },
    {
      "epoch": 0.6129281483389132,
      "grad_norm": 0.004932933139070266,
      "learning_rate": 7.734740790612136e-05,
      "loss": 0.0037,
      "num_input_tokens_seen": 11880960,
      "step": 1190
    },
    {
      "epoch": 0.6155034766932784,
      "grad_norm": 0.022109282268939556,
      "learning_rate": 7.714352216101055e-05,
      "loss": 0.0007,
      "num_input_tokens_seen": 11930880,
      "step": 1195
    },
    {
      "epoch": 0.6180788050476436,
      "grad_norm": 0.037426371340479145,
      "learning_rate": 7.693899446759727e-05,
      "loss": 0.0038,
      "num_input_tokens_seen": 11980800,
      "step": 1200
    },
    {
      "epoch": 0.6180788050476436,
      "eval_loss": 0.0022160401567816734,
      "eval_runtime": 18.725,
      "eval_samples_per_second": 3.204,
      "eval_steps_per_second": 0.801,
      "num_input_tokens_seen": 11980800,
      "step": 1200
    }
  ],
  "logging_steps": 5,
  "max_steps": 1200,
  "num_input_tokens_seen": 11990784,
  "num_train_epochs": 1,
  "save_steps": 50,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 791107965943808.0,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}