[
  {
    "loss": 1.9463,
    "grad_norm": 0.9043406844139099,
    "learning_rate": 1.8e-05,
    "entropy": 1.4898367643356323,
    "num_tokens": 239592.0,
    "mean_token_accuracy": 0.5726580291986465,
    "epoch": 0.016680567139282735,
    "step": 10
  },
  {
    "loss": 1.8033,
    "grad_norm": 0.6490057706832886,
    "learning_rate": 3.8e-05,
    "entropy": 1.7221096634864808,
    "num_tokens": 480339.0,
    "mean_token_accuracy": 0.5810627460479736,
    "epoch": 0.03336113427856547,
    "step": 20
  },
  {
    "loss": 1.5177,
    "grad_norm": 1.1159552335739136,
    "learning_rate": 5.8e-05,
    "entropy": 1.5653360843658448,
    "num_tokens": 719994.0,
    "mean_token_accuracy": 0.6268167346715927,
    "epoch": 0.05004170141784821,
    "step": 30
  },
  {
    "loss": 1.0739,
    "grad_norm": 0.4417094886302948,
    "learning_rate": 7.800000000000001e-05,
    "entropy": 1.053076034784317,
    "num_tokens": 957912.0,
    "mean_token_accuracy": 0.7257216989994049,
    "epoch": 0.06672226855713094,
    "step": 40
  },
  {
    "loss": 0.9505,
    "grad_norm": 0.3340432941913605,
    "learning_rate": 9.8e-05,
    "entropy": 0.9432537257671356,
    "num_tokens": 1197740.0,
    "mean_token_accuracy": 0.7573244959115982,
    "epoch": 0.08340283569641367,
    "step": 50
  },
  {
    "loss": 0.9094,
    "grad_norm": 0.3062295615673065,
    "learning_rate": 9.993394529660307e-05,
    "entropy": 0.902097898721695,
    "num_tokens": 1434364.0,
    "mean_token_accuracy": 0.761836588382721,
    "epoch": 0.10008340283569642,
    "step": 60
  },
  {
    "loss": 0.8602,
    "grad_norm": 0.302666038274765,
    "learning_rate": 9.970583211010007e-05,
    "entropy": 0.8552027255296707,
    "num_tokens": 1673721.0,
    "mean_token_accuracy": 0.7724049925804138,
    "epoch": 0.11676396997497915,
    "step": 70
  },
  {
    "loss": 0.8386,
    "grad_norm": 0.35731419920921326,
    "learning_rate": 9.931558883072403e-05,
    "entropy": 0.8345210403203964,
    "num_tokens": 1921434.0,
    "mean_token_accuracy": 0.7766192525625228,
    "epoch": 0.13344453711426188,
    "step": 80
  },
  {
    "loss": 0.8519,
    "grad_norm": 0.3715258538722992,
    "learning_rate": 9.87644883509393e-05,
    "entropy": 0.8489096939563752,
    "num_tokens": 2162407.0,
    "mean_token_accuracy": 0.7723332077264786,
    "epoch": 0.1501251042535446,
    "step": 90
  },
  {
    "loss": 0.766,
    "grad_norm": 0.3652157783508301,
    "learning_rate": 9.805432824596348e-05,
    "entropy": 0.7692611366510391,
    "num_tokens": 2399946.0,
    "mean_token_accuracy": 0.7895216166973114,
    "epoch": 0.16680567139282734,
    "step": 100
  },
  {
    "loss": 0.7927,
    "grad_norm": 0.3489046096801758,
    "learning_rate": 9.71874249104506e-05,
    "entropy": 0.7879800289869309,
    "num_tokens": 2643290.0,
    "mean_token_accuracy": 0.7858752965927124,
    "epoch": 0.1834862385321101,
    "step": 110
  },
  {
    "loss": 0.7939,
    "grad_norm": 0.3456013798713684,
    "learning_rate": 9.61666060028933e-05,
    "entropy": 0.8026239216327667,
    "num_tokens": 2890007.0,
    "mean_token_accuracy": 0.7865915298461914,
    "epoch": 0.20016680567139283,
    "step": 120
  },
  {
    "loss": 0.7367,
    "grad_norm": 0.390689492225647,
    "learning_rate": 9.499520122238845e-05,
    "entropy": 0.7422888517379761,
    "num_tokens": 3134139.0,
    "mean_token_accuracy": 0.7972356855869294,
    "epoch": 0.21684737281067556,
    "step": 130
  },
  {
    "loss": 0.7339,
    "grad_norm": 0.3555961549282074,
    "learning_rate": 9.367703144785096e-05,
    "entropy": 0.740860840678215,
    "num_tokens": 3382098.0,
    "mean_token_accuracy": 0.797610804438591,
    "epoch": 0.2335279399499583,
    "step": 140
  },
  {
    "loss": 0.7405,
    "grad_norm": 0.3681407570838928,
    "learning_rate": 9.221639627510076e-05,
    "entropy": 0.744793301820755,
    "num_tokens": 3622551.0,
    "mean_token_accuracy": 0.7954040169715881,
    "epoch": 0.25020850708924103,
    "step": 150
  },
  {
    "loss": 0.7253,
    "grad_norm": 0.32773861289024353,
    "learning_rate": 9.061805999247504e-05,
    "entropy": 0.7334282338619232,
    "num_tokens": 3865667.0,
    "mean_token_accuracy": 0.799143373966217,
    "epoch": 0.26688907422852376,
    "step": 160
  },
  {
    "loss": 0.711,
    "grad_norm": 0.32971468567848206,
    "learning_rate": 8.88872360407099e-05,
    "entropy": 0.7231753587722778,
    "num_tokens": 4101163.0,
    "mean_token_accuracy": 0.803010705113411,
    "epoch": 0.2835696413678065,
    "step": 170
  },
  {
    "loss": 0.7172,
    "grad_norm": 0.3492438495159149,
    "learning_rate": 8.70295700077803e-05,
    "entropy": 0.7242234885692597,
    "num_tokens": 4344116.0,
    "mean_token_accuracy": 0.8021188855171204,
    "epoch": 0.3002502085070892,
    "step": 180
  },
  {
    "loss": 0.7291,
    "grad_norm": 0.33426719903945923,
    "learning_rate": 8.505112121416555e-05,
    "entropy": 0.7333565503358841,
    "num_tokens": 4581952.0,
    "mean_token_accuracy": 0.7988870143890381,
    "epoch": 0.31693077564637195,
    "step": 190
  },
  {
    "loss": 0.7138,
    "grad_norm": 0.33009985089302063,
    "learning_rate": 8.295834294860535e-05,
    "entropy": 0.7183975011110306,
    "num_tokens": 4829123.0,
    "mean_token_accuracy": 0.8039891421794891,
    "epoch": 0.3336113427856547,
    "step": 200
  },
  {
    "loss": 0.685,
    "grad_norm": 0.35757768154144287,
    "learning_rate": 8.075806141881326e-05,
    "entropy": 0.6890431672334671,
    "num_tokens": 5067338.0,
    "mean_token_accuracy": 0.8091971307992936,
    "epoch": 0.3502919099249375,
    "step": 210
  },
  {
    "loss": 0.6954,
    "grad_norm": 0.3246421813964844,
    "learning_rate": 7.845745348580591e-05,
    "entropy": 0.6971507608890534,
    "num_tokens": 5311215.0,
    "mean_token_accuracy": 0.8067805171012878,
    "epoch": 0.3669724770642202,
    "step": 220
  },
  {
    "loss": 0.7073,
    "grad_norm": 0.3495083749294281,
    "learning_rate": 7.60640232544742e-05,
    "entropy": 0.7148334234952927,
    "num_tokens": 5554999.0,
    "mean_token_accuracy": 0.8036890238523483,
    "epoch": 0.38365304420350294,
    "step": 230
  },
  {
    "loss": 0.676,
    "grad_norm": 0.38008829951286316,
    "learning_rate": 7.358557759675284e-05,
    "entropy": 0.6754449099302292,
    "num_tokens": 5802860.0,
    "mean_token_accuracy": 0.8100622564554214,
    "epoch": 0.40033361134278567,
    "step": 240
  },
  {
    "loss": 0.682,
    "grad_norm": 0.3694831430912018,
    "learning_rate": 7.103020068722674e-05,
    "entropy": 0.6828817337751388,
    "num_tokens": 6043669.0,
    "mean_token_accuracy": 0.8099299311637879,
    "epoch": 0.4170141784820684,
    "step": 250
  },
  {
    "loss": 0.6754,
    "grad_norm": 0.34716424345970154,
    "learning_rate": 6.840622763423391e-05,
    "entropy": 0.6824066579341889,
    "num_tokens": 6279154.0,
    "mean_token_accuracy": 0.8119429558515548,
    "epoch": 0.43369474562135113,
    "step": 260
  },
  {
    "loss": 0.6567,
    "grad_norm": 0.3316477835178375,
    "learning_rate": 6.572221729247441e-05,
    "entropy": 0.6584962159395218,
    "num_tokens": 6514535.0,
    "mean_token_accuracy": 0.8146260052919387,
    "epoch": 0.45037531276063386,
    "step": 270
  },
  {
    "loss": 0.6816,
    "grad_norm": 0.36305058002471924,
    "learning_rate": 6.298692434580542e-05,
    "entropy": 0.6815447330474853,
    "num_tokens": 6757908.0,
    "mean_token_accuracy": 0.8089588433504105,
    "epoch": 0.4670558798999166,
    "step": 280
  },
  {
    "loss": 0.6897,
    "grad_norm": 0.33401617407798767,
    "learning_rate": 6.0209270751282165e-05,
    "entropy": 0.696434935927391,
    "num_tokens": 6996464.0,
    "mean_token_accuracy": 0.8082989573478698,
    "epoch": 0.4837364470391993,
    "step": 290
  },
  {
    "loss": 0.6921,
    "grad_norm": 0.34646838903427124,
    "learning_rate": 5.7398316637588336e-05,
    "entropy": 0.6913289308547974,
    "num_tokens": 7234102.0,
    "mean_token_accuracy": 0.8069221824407578,
    "epoch": 0.5004170141784821,
    "step": 300
  },
  {
    "loss": 0.6802,
    "grad_norm": 0.37350520491600037,
    "learning_rate": 5.4563230752779595e-05,
    "entropy": 0.6853398621082306,
    "num_tokens": 7479635.0,
    "mean_token_accuracy": 0.810760036110878,
    "epoch": 0.5170975813177648,
    "step": 310
  },
  {
    "loss": 0.6676,
    "grad_norm": 0.3853772282600403,
    "learning_rate": 5.1713260557733176e-05,
    "entropy": 0.6696277529001236,
    "num_tokens": 7718364.0,
    "mean_token_accuracy": 0.8130882978439331,
    "epoch": 0.5337781484570475,
    "step": 320
  },
  {
    "loss": 0.6537,
    "grad_norm": 0.338049054145813,
    "learning_rate": 4.885770206285252e-05,
    "entropy": 0.6589328497648239,
    "num_tokens": 7958054.0,
    "mean_token_accuracy": 0.8148987740278244,
    "epoch": 0.5504587155963303,
    "step": 330
  },
  {
    "loss": 0.6902,
    "grad_norm": 0.303418904542923,
    "learning_rate": 4.600586950641362e-05,
    "entropy": 0.692239272594452,
    "num_tokens": 8209298.0,
    "mean_token_accuracy": 0.8085194110870362,
    "epoch": 0.567139282735613,
    "step": 340
  },
  {
    "loss": 0.6659,
    "grad_norm": 0.38714390993118286,
    "learning_rate": 4.316706497345572e-05,
    "entropy": 0.6669029265642166,
    "num_tokens": 8455256.0,
    "mean_token_accuracy": 0.8131787478923798,
    "epoch": 0.5838198498748958,
    "step": 350
  },
  {
    "loss": 0.6579,
    "grad_norm": 0.35571643710136414,
    "learning_rate": 4.035054805431334e-05,
    "entropy": 0.6595758140087128,
    "num_tokens": 8695771.0,
    "mean_token_accuracy": 0.8157992452383042,
    "epoch": 0.6005004170141784,
    "step": 360
  },
  {
    "loss": 0.6641,
    "grad_norm": 0.3794392943382263,
    "learning_rate": 3.756550564175727e-05,
    "entropy": 0.6648996561765671,
    "num_tokens": 8944844.0,
    "mean_token_accuracy": 0.8132083296775818,
    "epoch": 0.6171809841534612,
    "step": 370
  },
  {
    "loss": 0.6577,
    "grad_norm": 0.3610455393791199,
    "learning_rate": 3.4821021965259904e-05,
    "entropy": 0.6610071390867234,
    "num_tokens": 9191203.0,
    "mean_token_accuracy": 0.8145314186811448,
    "epoch": 0.6338615512927439,
    "step": 380
  },
  {
    "loss": 0.659,
    "grad_norm": 0.3606182634830475,
    "learning_rate": 3.212604896012679e-05,
    "entropy": 0.660676684975624,
    "num_tokens": 9426243.0,
    "mean_token_accuracy": 0.8144401997327805,
    "epoch": 0.6505421184320267,
    "step": 390
  },
  {
    "loss": 0.6568,
    "grad_norm": 0.3522314727306366,
    "learning_rate": 2.948937706814442e-05,
    "entropy": 0.6625478237867355,
    "num_tokens": 9664795.0,
    "mean_token_accuracy": 0.8151305437088012,
    "epoch": 0.6672226855713094,
    "step": 400
  },
  {
    "loss": 0.6536,
    "grad_norm": 0.3643236756324768,
    "learning_rate": 2.691960656498621e-05,
    "entropy": 0.657148751616478,
    "num_tokens": 9904436.0,
    "mean_token_accuracy": 0.8148669481277466,
    "epoch": 0.6839032527105922,
    "step": 410
  },
  {
    "loss": 0.6704,
    "grad_norm": 0.3429484963417053,
    "learning_rate": 2.4425119507900813e-05,
    "entropy": 0.6674410969018936,
    "num_tokens": 10150234.0,
    "mean_token_accuracy": 0.8119391798973083,
    "epoch": 0.700583819849875,
    "step": 420
  },
  {
    "loss": 0.7034,
    "grad_norm": 0.3803344666957855,
    "learning_rate": 2.2014052395183627e-05,
    "entropy": 0.7082253098487854,
    "num_tokens": 10391148.0,
    "mean_token_accuracy": 0.8065144032239914,
    "epoch": 0.7172643869891576,
    "step": 430
  },
  {
    "loss": 0.6943,
    "grad_norm": 0.30864450335502625,
    "learning_rate": 1.969426962661059e-05,
    "entropy": 0.7002600938081741,
    "num_tokens": 10636950.0,
    "mean_token_accuracy": 0.8064252346754074,
    "epoch": 0.7339449541284404,
    "step": 440
  },
  {
    "loss": 0.6803,
    "grad_norm": 0.35500553250312805,
    "learning_rate": 1.747333785140066e-05,
    "entropy": 0.681396746635437,
    "num_tokens": 10879243.0,
    "mean_token_accuracy": 0.8101641505956649,
    "epoch": 0.7506255212677231,
    "step": 450
  },
  {
    "loss": 0.6421,
    "grad_norm": 0.36364850401878357,
    "learning_rate": 1.535850128737884e-05,
    "entropy": 0.6462778955698013,
    "num_tokens": 11114856.0,
    "mean_token_accuracy": 0.8183332979679108,
    "epoch": 0.7673060884070059,
    "step": 460
  },
  {
    "loss": 0.6766,
    "grad_norm": 0.33400440216064453,
    "learning_rate": 1.335665809184341e-05,
    "entropy": 0.6827076494693756,
    "num_tokens": 11360827.0,
    "mean_token_accuracy": 0.8112966924905777,
    "epoch": 0.7839866555462885,
    "step": 470
  },
  {
    "loss": 0.647,
    "grad_norm": 0.3601098954677582,
    "learning_rate": 1.1474337861210543e-05,
    "entropy": 0.6471425205469131,
    "num_tokens": 11601682.0,
    "mean_token_accuracy": 0.8174957513809205,
    "epoch": 0.8006672226855713,
    "step": 480
  },
  {
    "loss": 0.6936,
    "grad_norm": 0.36329320073127747,
    "learning_rate": 9.717680332828016e-06,
    "entropy": 0.7031383335590362,
    "num_tokens": 11844375.0,
    "mean_token_accuracy": 0.805937397480011,
    "epoch": 0.817347789824854,
    "step": 490
  },
  {
    "loss": 0.6743,
    "grad_norm": 0.3474838137626648,
    "learning_rate": 8.092415358428173e-06,
    "entropy": 0.6719307899475098,
    "num_tokens": 12085115.0,
    "mean_token_accuracy": 0.8131552457809448,
    "epoch": 0.8340283569641368,
    "step": 500
  },
  {
    "loss": 0.6344,
    "grad_norm": 0.35991644859313965,
    "learning_rate": 6.603844214542487e-06,
    "entropy": 0.637131878733635,
    "num_tokens": 12332095.0,
    "mean_token_accuracy": 0.8191376119852066,
    "epoch": 0.8507089241034195,
    "step": 510
  },
  {
    "loss": 0.6349,
    "grad_norm": 0.34625497460365295,
    "learning_rate": 5.256822310839405e-06,
    "entropy": 0.6384859681129456,
    "num_tokens": 12575942.0,
    "mean_token_accuracy": 0.8205374091863632,
    "epoch": 0.8673894912427023,
    "step": 520
  },
  {
    "loss": 0.6527,
    "grad_norm": 0.34414827823638916,
    "learning_rate": 4.055743352787267e-06,
    "entropy": 0.6543797552585602,
    "num_tokens": 12817224.0,
    "mean_token_accuracy": 0.8158955574035645,
    "epoch": 0.8840700583819849,
    "step": 530
  },
  {
    "loss": 0.6387,
    "grad_norm": 0.32114923000335693,
    "learning_rate": 3.0045250103002296e-06,
    "entropy": 0.6461419433355331,
    "num_tokens": 13055918.0,
    "mean_token_accuracy": 0.8180903941392899,
    "epoch": 0.9007506255212677,
    "step": 540
  },
  {
    "loss": 0.6749,
    "grad_norm": 0.38731274008750916,
    "learning_rate": 2.1065961391133702e-06,
    "entropy": 0.6730036258697509,
    "num_tokens": 13296781.0,
    "mean_token_accuracy": 0.8106796860694885,
    "epoch": 0.9174311926605505,
    "step": 550
  },
  {
    "loss": 0.63,
    "grad_norm": 0.3419639766216278,
    "learning_rate": 1.3648855965679497e-06,
    "entropy": 0.6305610507726669,
    "num_tokens": 13536924.0,
    "mean_token_accuracy": 0.8212851226329804,
    "epoch": 0.9341117597998332,
    "step": 560
  },
  {
    "loss": 0.641,
    "grad_norm": 0.3530997931957245,
    "learning_rate": 7.818126882875254e-07,
    "entropy": 0.647763478755951,
    "num_tokens": 13773817.0,
    "mean_token_accuracy": 0.8177464485168457,
    "epoch": 0.950792326939116,
    "step": 570
  },
  {
    "loss": 0.6724,
    "grad_norm": 0.3384498953819275,
    "learning_rate": 3.5927927690588283e-07,
    "entropy": 0.675562146306038,
    "num_tokens": 14012569.0,
    "mean_token_accuracy": 0.8116778939962387,
    "epoch": 0.9674728940783986,
    "step": 580
  },
  {
    "loss": 0.6523,
    "grad_norm": 0.3589385747909546,
    "learning_rate": 9.866357858642205e-08,
    "entropy": 0.6570082008838654,
    "num_tokens": 14254389.0,
    "mean_token_accuracy": 0.8173295229673385,
    "epoch": 0.9841534612176814,
    "step": 590
  },
  {
    "loss": 0.6664,
    "grad_norm": 0.5415430665016174,
    "learning_rate": 8.156675674941828e-10,
    "entropy": 0.6749062004842257,
    "num_tokens": 14485000.0,
    "mean_token_accuracy": 0.8125993891766197,
    "epoch": 1.0,
    "step": 600
  },
  {
    "train_runtime": 6029.5713,
    "train_samples_per_second": 3.181,
    "train_steps_per_second": 0.1,
    "total_flos": 6.620616404100956e+18,
    "train_loss": 0.7606559038162232,
    "epoch": 1.0,
    "step": 600
  }
]