{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 2.0, "eval_steps": 500, "global_step": 796, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.01256281407035176, "grad_norm": 59.5, "learning_rate": 1.6666666666666667e-06, "loss": 2.998407745361328, "step": 5 }, { "epoch": 0.02512562814070352, "grad_norm": 33.25, "learning_rate": 3.7500000000000005e-06, "loss": 2.837660789489746, "step": 10 }, { "epoch": 0.03768844221105527, "grad_norm": 13.75, "learning_rate": 5.833333333333334e-06, "loss": 2.7715581893920898, "step": 15 }, { "epoch": 0.05025125628140704, "grad_norm": 6.59375, "learning_rate": 7.916666666666667e-06, "loss": 2.556188201904297, "step": 20 }, { "epoch": 0.06281407035175879, "grad_norm": 7.21875, "learning_rate": 1e-05, "loss": 2.4427762985229493, "step": 25 }, { "epoch": 0.07537688442211055, "grad_norm": 4.28125, "learning_rate": 9.998965024239257e-06, "loss": 2.2074916839599608, "step": 30 }, { "epoch": 0.08793969849246232, "grad_norm": 5.375, "learning_rate": 9.995860525426954e-06, "loss": 2.110469627380371, "step": 35 }, { "epoch": 0.10050251256281408, "grad_norm": 3.203125, "learning_rate": 9.9906877887955e-06, "loss": 2.0038677215576173, "step": 40 }, { "epoch": 0.11306532663316583, "grad_norm": 3.0625, "learning_rate": 9.983448955807708e-06, "loss": 1.951383399963379, "step": 45 }, { "epoch": 0.12562814070351758, "grad_norm": 2.671875, "learning_rate": 9.974147023270248e-06, "loss": 1.893879508972168, "step": 50 }, { "epoch": 0.13819095477386933, "grad_norm": 2.421875, "learning_rate": 9.962785842093003e-06, "loss": 1.8473941802978515, "step": 55 }, { "epoch": 0.1507537688442211, "grad_norm": 2.265625, "learning_rate": 9.949370115694827e-06, "loss": 1.8276456832885741, "step": 60 }, { "epoch": 0.16331658291457288, "grad_norm": 2.140625, "learning_rate": 9.933905398056371e-06, "loss": 1.8131959915161133, "step": 65 }, { "epoch": 0.17587939698492464, "grad_norm": 2.09375, "learning_rate": 9.916398091420798e-06, "loss": 1.7640661239624023, "step": 70 }, { "epoch": 0.1884422110552764, "grad_norm": 2.171875, "learning_rate": 9.896855443643307e-06, "loss": 1.7705204010009765, "step": 75 }, { "epoch": 0.20100502512562815, "grad_norm": 2.0625, "learning_rate": 9.875285545190602e-06, "loss": 1.7359968185424806, "step": 80 }, { "epoch": 0.2135678391959799, "grad_norm": 2.140625, "learning_rate": 9.851697325791505e-06, "loss": 1.7584953308105469, "step": 85 }, { "epoch": 0.22613065326633167, "grad_norm": 1.8984375, "learning_rate": 9.826100550740144e-06, "loss": 1.7408344268798828, "step": 90 }, { "epoch": 0.23869346733668342, "grad_norm": 2.671875, "learning_rate": 9.798505816853207e-06, "loss": 1.7312856674194337, "step": 95 }, { "epoch": 0.25125628140703515, "grad_norm": 2.046875, "learning_rate": 9.768924548082978e-06, "loss": 1.6844066619873046, "step": 100 }, { "epoch": 0.2638190954773869, "grad_norm": 2.0625, "learning_rate": 9.737368990787917e-06, "loss": 1.6968770980834962, "step": 105 }, { "epoch": 0.27638190954773867, "grad_norm": 2.234375, "learning_rate": 9.703852208662788e-06, "loss": 1.7310028076171875, "step": 110 }, { "epoch": 0.2889447236180904, "grad_norm": 2.625, "learning_rate": 9.668388077330421e-06, "loss": 1.7414098739624024, "step": 115 }, { "epoch": 0.3015075376884422, "grad_norm": 2.03125, "learning_rate": 9.630991278597344e-06, "loss": 1.6551843643188477, "step": 120 }, { "epoch": 0.314070351758794, "grad_norm": 2.25, "learning_rate": 9.591677294375637e-06, "loss": 1.6384830474853516, "step": 125 }, { "epoch": 0.32663316582914576, "grad_norm": 2.578125, "learning_rate": 9.550462400273595e-06, "loss": 1.6961067199707032, "step": 130 }, { "epoch": 0.3391959798994975, "grad_norm": 2.6875, "learning_rate": 9.507363658857768e-06, "loss": 1.7122671127319335, "step": 135 }, { "epoch": 0.35175879396984927, "grad_norm": 2.359375, "learning_rate": 9.462398912589232e-06, "loss": 1.6428680419921875, "step": 140 }, { "epoch": 0.36432160804020103, "grad_norm": 2.515625, "learning_rate": 9.415586776436973e-06, "loss": 1.6411958694458009, "step": 145 }, { "epoch": 0.3768844221105528, "grad_norm": 2.078125, "learning_rate": 9.366946630171485e-06, "loss": 1.645296859741211, "step": 150 }, { "epoch": 0.38944723618090454, "grad_norm": 2.125, "learning_rate": 9.31649861034172e-06, "loss": 1.6568294525146485, "step": 155 }, { "epoch": 0.4020100502512563, "grad_norm": 2.046875, "learning_rate": 9.264263601938759e-06, "loss": 1.7317806243896485, "step": 160 }, { "epoch": 0.41457286432160806, "grad_norm": 2.75, "learning_rate": 9.210263229749626e-06, "loss": 1.6810962677001953, "step": 165 }, { "epoch": 0.4271356783919598, "grad_norm": 2.265625, "learning_rate": 9.154519849404833e-06, "loss": 1.6197080612182617, "step": 170 }, { "epoch": 0.4396984924623116, "grad_norm": 2.078125, "learning_rate": 9.097056538123376e-06, "loss": 1.6504512786865235, "step": 175 }, { "epoch": 0.45226130653266333, "grad_norm": 2.0625, "learning_rate": 9.037897085158977e-06, "loss": 1.6403568267822266, "step": 180 }, { "epoch": 0.4648241206030151, "grad_norm": 2.53125, "learning_rate": 8.977065981951567e-06, "loss": 1.6068395614624023, "step": 185 }, { "epoch": 0.47738693467336685, "grad_norm": 2.765625, "learning_rate": 8.914588411988079e-06, "loss": 1.6754827499389648, "step": 190 }, { "epoch": 0.4899497487437186, "grad_norm": 2.390625, "learning_rate": 8.85049024037671e-06, "loss": 1.6853302001953125, "step": 195 }, { "epoch": 0.5025125628140703, "grad_norm": 2.734375, "learning_rate": 8.784798003139035e-06, "loss": 1.603452491760254, "step": 200 }, { "epoch": 0.5150753768844221, "grad_norm": 2.15625, "learning_rate": 8.717538896224333e-06, "loss": 1.5766297340393067, "step": 205 }, { "epoch": 0.5276381909547738, "grad_norm": 2.5625, "learning_rate": 8.648740764250745e-06, "loss": 1.716385269165039, "step": 210 }, { "epoch": 0.5402010050251256, "grad_norm": 2.5, "learning_rate": 8.57843208897786e-06, "loss": 1.592827320098877, "step": 215 }, { "epoch": 0.5527638190954773, "grad_norm": 2.84375, "learning_rate": 8.50664197751555e-06, "loss": 1.615340232849121, "step": 220 }, { "epoch": 0.5653266331658291, "grad_norm": 2.21875, "learning_rate": 8.433400150273907e-06, "loss": 1.5731450080871583, "step": 225 }, { "epoch": 0.5778894472361809, "grad_norm": 2.5, "learning_rate": 8.358736928659275e-06, "loss": 1.7298618316650392, "step": 230 }, { "epoch": 0.5904522613065326, "grad_norm": 1.90625, "learning_rate": 8.282683222521491e-06, "loss": 1.6517925262451172, "step": 235 }, { "epoch": 0.6030150753768844, "grad_norm": 2.59375, "learning_rate": 8.205270517357503e-06, "loss": 1.6013933181762696, "step": 240 }, { "epoch": 0.6155778894472361, "grad_norm": 2.796875, "learning_rate": 8.126530861276677e-06, "loss": 1.584385585784912, "step": 245 }, { "epoch": 0.628140703517588, "grad_norm": 2.5, "learning_rate": 8.046496851733195e-06, "loss": 1.5994317054748535, "step": 250 }, { "epoch": 0.6407035175879398, "grad_norm": 3.0, "learning_rate": 7.96520162203102e-06, "loss": 1.590476131439209, "step": 255 }, { "epoch": 0.6532663316582915, "grad_norm": 2.28125, "learning_rate": 7.88267882760704e-06, "loss": 1.6111858367919922, "step": 260 }, { "epoch": 0.6658291457286433, "grad_norm": 2.28125, "learning_rate": 7.798962632098024e-06, "loss": 1.6274175643920898, "step": 265 }, { "epoch": 0.678391959798995, "grad_norm": 2.71875, "learning_rate": 7.714087693197228e-06, "loss": 1.6058385848999024, "step": 270 }, { "epoch": 0.6909547738693468, "grad_norm": 2.515625, "learning_rate": 7.628089148306434e-06, "loss": 1.5899338722229004, "step": 275 }, { "epoch": 0.7035175879396985, "grad_norm": 2.59375, "learning_rate": 7.541002599989411e-06, "loss": 1.5922243118286132, "step": 280 }, { "epoch": 0.7160804020100503, "grad_norm": 3.3125, "learning_rate": 7.452864101232798e-06, "loss": 1.5848384857177735, "step": 285 }, { "epoch": 0.7286432160804021, "grad_norm": 2.296875, "learning_rate": 7.363710140520515e-06, "loss": 1.60340576171875, "step": 290 }, { "epoch": 0.7412060301507538, "grad_norm": 1.9453125, "learning_rate": 7.273577626727884e-06, "loss": 1.6018304824829102, "step": 295 }, { "epoch": 0.7537688442211056, "grad_norm": 2.40625, "learning_rate": 7.182503873841722e-06, "loss": 1.5592435836791991, "step": 300 }, { "epoch": 0.7663316582914573, "grad_norm": 2.046875, "learning_rate": 7.090526585512696e-06, "loss": 1.6086605072021485, "step": 305 }, { "epoch": 0.7788944723618091, "grad_norm": 2.65625, "learning_rate": 6.997683839446392e-06, "loss": 1.6314334869384766, "step": 310 }, { "epoch": 0.7914572864321608, "grad_norm": 2.8125, "learning_rate": 6.904014071639503e-06, "loss": 1.5340024948120117, "step": 315 }, { "epoch": 0.8040201005025126, "grad_norm": 2.640625, "learning_rate": 6.80955606046771e-06, "loss": 1.5246624946594238, "step": 320 }, { "epoch": 0.8165829145728644, "grad_norm": 2.390625, "learning_rate": 6.7143489106318e-06, "loss": 1.5667107582092286, "step": 325 }, { "epoch": 0.8291457286432161, "grad_norm": 1.890625, "learning_rate": 6.618432036968705e-06, "loss": 1.555766487121582, "step": 330 }, { "epoch": 0.8417085427135679, "grad_norm": 2.28125, "learning_rate": 6.52184514813414e-06, "loss": 1.5755093574523926, "step": 335 }, { "epoch": 0.8542713567839196, "grad_norm": 2.265625, "learning_rate": 6.424628230163606e-06, "loss": 1.6115888595581054, "step": 340 }, { "epoch": 0.8668341708542714, "grad_norm": 2.3125, "learning_rate": 6.3268215299185545e-06, "loss": 1.582913589477539, "step": 345 }, { "epoch": 0.8793969849246231, "grad_norm": 2.640625, "learning_rate": 6.228465538424583e-06, "loss": 1.5411038398742676, "step": 350 }, { "epoch": 0.8919597989949749, "grad_norm": 2.5, "learning_rate": 6.129600974108538e-06, "loss": 1.5892680168151856, "step": 355 }, { "epoch": 0.9045226130653267, "grad_norm": 1.8984375, "learning_rate": 6.0302687659414904e-06, "loss": 1.5735843658447266, "step": 360 }, { "epoch": 0.9170854271356784, "grad_norm": 2.3125, "learning_rate": 5.93051003649452e-06, "loss": 1.5499893188476563, "step": 365 }, { "epoch": 0.9296482412060302, "grad_norm": 2.5625, "learning_rate": 5.83036608491439e-06, "loss": 1.6115238189697265, "step": 370 }, { "epoch": 0.9422110552763819, "grad_norm": 2.1875, "learning_rate": 5.729878369826087e-06, "loss": 1.5861882209777831, "step": 375 }, { "epoch": 0.9547738693467337, "grad_norm": 2.375, "learning_rate": 5.62908849216936e-06, "loss": 1.5845415115356445, "step": 380 }, { "epoch": 0.9673366834170855, "grad_norm": 2.03125, "learning_rate": 5.52803817797633e-06, "loss": 1.5787755012512208, "step": 385 }, { "epoch": 0.9798994974874372, "grad_norm": 2.75, "learning_rate": 5.426769261097318e-06, "loss": 1.5202445983886719, "step": 390 }, { "epoch": 0.992462311557789, "grad_norm": 2.375, "learning_rate": 5.325323665882039e-06, "loss": 1.5645788192749024, "step": 395 }, { "epoch": 1.0050251256281406, "grad_norm": 1.890625, "learning_rate": 5.223743389823327e-06, "loss": 1.5927098274230957, "step": 400 }, { "epoch": 1.0175879396984924, "grad_norm": 2.921875, "learning_rate": 5.1220704861705775e-06, "loss": 1.6001703262329101, "step": 405 }, { "epoch": 1.0301507537688441, "grad_norm": 2.421875, "learning_rate": 5.020347046520112e-06, "loss": 1.5346534729003907, "step": 410 }, { "epoch": 1.0427135678391959, "grad_norm": 2.03125, "learning_rate": 4.918615183389666e-06, "loss": 1.5630980491638184, "step": 415 }, { "epoch": 1.0552763819095476, "grad_norm": 2.59375, "learning_rate": 4.8169170127842135e-06, "loss": 1.523043441772461, "step": 420 }, { "epoch": 1.0678391959798994, "grad_norm": 2.34375, "learning_rate": 4.715294636760352e-06, "loss": 1.5776254653930664, "step": 425 }, { "epoch": 1.0804020100502512, "grad_norm": 2.046875, "learning_rate": 4.6137901259964515e-06, "loss": 1.5493414878845215, "step": 430 }, { "epoch": 1.092964824120603, "grad_norm": 2.59375, "learning_rate": 4.512445502375813e-06, "loss": 1.5309125900268554, "step": 435 }, { "epoch": 1.1055276381909547, "grad_norm": 2.265625, "learning_rate": 4.411302721590006e-06, "loss": 1.54165620803833, "step": 440 }, { "epoch": 1.1180904522613064, "grad_norm": 2.671875, "learning_rate": 4.310403655769629e-06, "loss": 1.5176856994628907, "step": 445 }, { "epoch": 1.1306532663316582, "grad_norm": 2.515625, "learning_rate": 4.209790076149644e-06, "loss": 1.5529046058654785, "step": 450 }, { "epoch": 1.1432160804020102, "grad_norm": 3.15625, "learning_rate": 4.109503635776492e-06, "loss": 1.565816593170166, "step": 455 }, { "epoch": 1.1557788944723617, "grad_norm": 2.140625, "learning_rate": 4.009585852264139e-06, "loss": 1.5432037353515624, "step": 460 }, { "epoch": 1.1683417085427137, "grad_norm": 2.53125, "learning_rate": 3.91007809060619e-06, "loss": 1.6062923431396485, "step": 465 }, { "epoch": 1.1809045226130652, "grad_norm": 2.0625, "learning_rate": 3.8110215460511697e-06, "loss": 1.5221455574035645, "step": 470 }, { "epoch": 1.1934673366834172, "grad_norm": 2.171875, "learning_rate": 3.7124572270481055e-06, "loss": 1.6717411041259767, "step": 475 }, { "epoch": 1.2060301507537687, "grad_norm": 2.734375, "learning_rate": 3.6144259382694115e-06, "loss": 1.5347406387329101, "step": 480 }, { "epoch": 1.2185929648241207, "grad_norm": 2.546875, "learning_rate": 3.516968263718159e-06, "loss": 1.6209028244018555, "step": 485 }, { "epoch": 1.2311557788944723, "grad_norm": 2.109375, "learning_rate": 3.420124549926693e-06, "loss": 1.5777502059936523, "step": 490 }, { "epoch": 1.2437185929648242, "grad_norm": 1.8515625, "learning_rate": 3.3239348892535562e-06, "loss": 1.6334873199462892, "step": 495 }, { "epoch": 1.2562814070351758, "grad_norm": 2.140625, "learning_rate": 3.228439103285641e-06, "loss": 1.5432828903198241, "step": 500 }, { "epoch": 1.2688442211055277, "grad_norm": 2.125, "learning_rate": 3.133676726352438e-06, "loss": 1.5457247734069823, "step": 505 }, { "epoch": 1.2814070351758793, "grad_norm": 2.0625, "learning_rate": 3.0396869891592096e-06, "loss": 1.6380638122558593, "step": 510 }, { "epoch": 1.2939698492462313, "grad_norm": 2.609375, "learning_rate": 2.946508802545859e-06, "loss": 1.562052059173584, "step": 515 }, { "epoch": 1.3065326633165828, "grad_norm": 2.265625, "learning_rate": 2.8541807413782142e-06, "loss": 1.5281488418579101, "step": 520 }, { "epoch": 1.3190954773869348, "grad_norm": 2.46875, "learning_rate": 2.7627410285784164e-06, "loss": 1.586146926879883, "step": 525 }, { "epoch": 1.3316582914572863, "grad_norm": 2.21875, "learning_rate": 2.6722275193009872e-06, "loss": 1.5366415977478027, "step": 530 }, { "epoch": 1.3442211055276383, "grad_norm": 2.875, "learning_rate": 2.582677685261179e-06, "loss": 1.5299442291259766, "step": 535 }, { "epoch": 1.3567839195979898, "grad_norm": 1.984375, "learning_rate": 2.4941285992220353e-06, "loss": 1.5406274795532227, "step": 540 }, { "epoch": 1.3693467336683418, "grad_norm": 2.453125, "learning_rate": 2.4066169196466326e-06, "loss": 1.538019561767578, "step": 545 }, { "epoch": 1.3819095477386933, "grad_norm": 2.625, "learning_rate": 2.3201788755218265e-06, "loss": 1.5008614540100098, "step": 550 }, { "epoch": 1.3944723618090453, "grad_norm": 2.25, "learning_rate": 2.2348502513598035e-06, "loss": 1.6197319030761719, "step": 555 }, { "epoch": 1.4070351758793969, "grad_norm": 2.40625, "learning_rate": 2.15066637238365e-06, "loss": 1.6248485565185546, "step": 560 }, { "epoch": 1.4195979899497488, "grad_norm": 2.421875, "learning_rate": 2.0676620899030393e-06, "loss": 1.5241939544677734, "step": 565 }, { "epoch": 1.4321608040201004, "grad_norm": 1.953125, "learning_rate": 1.985871766886136e-06, "loss": 1.532661533355713, "step": 570 }, { "epoch": 1.4447236180904524, "grad_norm": 2.53125, "learning_rate": 1.905329263733649e-06, "loss": 1.5913199424743651, "step": 575 }, { "epoch": 1.457286432160804, "grad_norm": 2.125, "learning_rate": 1.8260679242609703e-06, "loss": 1.534491729736328, "step": 580 }, { "epoch": 1.4698492462311559, "grad_norm": 2.015625, "learning_rate": 1.7481205618941472e-06, "loss": 1.5714677810668944, "step": 585 }, { "epoch": 1.4824120603015074, "grad_norm": 3.15625, "learning_rate": 1.6715194460854466e-06, "loss": 1.5404623031616211, "step": 590 }, { "epoch": 1.4949748743718594, "grad_norm": 2.0625, "learning_rate": 1.5962962889541106e-06, "loss": 1.5692951202392578, "step": 595 }, { "epoch": 1.507537688442211, "grad_norm": 2.03125, "learning_rate": 1.5224822321578481e-06, "loss": 1.5303547859191895, "step": 600 }, { "epoch": 1.520100502512563, "grad_norm": 2.46875, "learning_rate": 1.4501078340004954e-06, "loss": 1.5067039489746095, "step": 605 }, { "epoch": 1.5326633165829144, "grad_norm": 2.6875, "learning_rate": 1.3792030567811687e-06, "loss": 1.5673861503601074, "step": 610 }, { "epoch": 1.5452261306532664, "grad_norm": 2.6875, "learning_rate": 1.3097972543901672e-06, "loss": 1.5892368316650392, "step": 615 }, { "epoch": 1.557788944723618, "grad_norm": 2.3125, "learning_rate": 1.2419191601567409e-06, "loss": 1.5634510040283203, "step": 620 }, { "epoch": 1.57035175879397, "grad_norm": 2.359375, "learning_rate": 1.1755968749537755e-06, "loss": 1.5212089538574218, "step": 625 }, { "epoch": 1.5829145728643215, "grad_norm": 2.8125, "learning_rate": 1.1108578555643058e-06, "loss": 1.5546403884887696, "step": 630 }, { "epoch": 1.5954773869346734, "grad_norm": 2.25, "learning_rate": 1.0477289033146675e-06, "loss": 1.5283503532409668, "step": 635 }, { "epoch": 1.608040201005025, "grad_norm": 2.15625, "learning_rate": 9.86236152979015e-07, "loss": 1.5374995231628419, "step": 640 }, { "epoch": 1.620603015075377, "grad_norm": 2.296875, "learning_rate": 9.264050619597697e-07, "loss": 1.5790555000305175, "step": 645 }, { "epoch": 1.6331658291457285, "grad_norm": 2.953125, "learning_rate": 8.682603997485078e-07, "loss": 1.5679688453674316, "step": 650 }, { "epoch": 1.6457286432160805, "grad_norm": 2.265625, "learning_rate": 8.11826237671634e-07, "loss": 1.529183292388916, "step": 655 }, { "epoch": 1.658291457286432, "grad_norm": 1.90625, "learning_rate": 7.571259389250779e-07, "loss": 1.5451692581176757, "step": 660 }, { "epoch": 1.670854271356784, "grad_norm": 3.046875, "learning_rate": 7.041821489021639e-07, "loss": 1.5198535919189453, "step": 665 }, { "epoch": 1.6834170854271355, "grad_norm": 1.8984375, "learning_rate": 6.530167858186343e-07, "loss": 1.5573585510253907, "step": 670 }, { "epoch": 1.6959798994974875, "grad_norm": 2.171875, "learning_rate": 6.036510316387196e-07, "loss": 1.5550883293151856, "step": 675 }, { "epoch": 1.708542713567839, "grad_norm": 1.921875, "learning_rate": 5.561053233060154e-07, "loss": 1.5345685958862305, "step": 680 }, { "epoch": 1.721105527638191, "grad_norm": 2.046875, "learning_rate": 5.103993442827832e-07, "loss": 1.5294795989990235, "step": 685 }, { "epoch": 1.7336683417085426, "grad_norm": 1.96875, "learning_rate": 4.6655201640118775e-07, "loss": 1.6449104309082032, "step": 690 }, { "epoch": 1.7462311557788945, "grad_norm": 2.65625, "learning_rate": 4.2458149202984025e-07, "loss": 1.5543377876281739, "step": 695 }, { "epoch": 1.758793969849246, "grad_norm": 2.296875, "learning_rate": 3.845051465588961e-07, "loss": 1.5942207336425782, "step": 700 }, { "epoch": 1.771356783919598, "grad_norm": 2.265625, "learning_rate": 3.4633957120681294e-07, "loss": 1.5990875244140625, "step": 705 }, { "epoch": 1.7839195979899496, "grad_norm": 2.5, "learning_rate": 3.1010056615174367e-07, "loss": 1.5136988639831543, "step": 710 }, { "epoch": 1.7964824120603016, "grad_norm": 2.296875, "learning_rate": 2.75803133990421e-07, "loss": 1.6442916870117188, "step": 715 }, { "epoch": 1.809045226130653, "grad_norm": 1.96875, "learning_rate": 2.434614735272184e-07, "loss": 1.524810791015625, "step": 720 }, { "epoch": 1.821608040201005, "grad_norm": 2.234375, "learning_rate": 2.130889738959946e-07, "loss": 1.6057912826538085, "step": 725 }, { "epoch": 1.8341708542713566, "grad_norm": 2.390625, "learning_rate": 1.8469820901711345e-07, "loss": 1.5753741264343262, "step": 730 }, { "epoch": 1.8467336683417086, "grad_norm": 2.1875, "learning_rate": 1.5830093239196765e-07, "loss": 1.5886516571044922, "step": 735 }, { "epoch": 1.8592964824120601, "grad_norm": 2.203125, "learning_rate": 1.3390807223713887e-07, "loss": 1.5261211395263672, "step": 740 }, { "epoch": 1.8718592964824121, "grad_norm": 2.078125, "learning_rate": 1.1152972696022447e-07, "loss": 1.5142106056213378, "step": 745 }, { "epoch": 1.8844221105527639, "grad_norm": 2.296875, "learning_rate": 9.117516097919721e-08, "loss": 1.555587387084961, "step": 750 }, { "epoch": 1.8969849246231156, "grad_norm": 2.53125, "learning_rate": 7.285280088701996e-08, "loss": 1.5386908531188965, "step": 755 }, { "epoch": 1.9095477386934674, "grad_norm": 2.28125, "learning_rate": 5.657023196312394e-08, "loss": 1.5089382171630858, "step": 760 }, { "epoch": 1.9221105527638191, "grad_norm": 2.171875, "learning_rate": 4.233419503317182e-08, "loss": 1.5173494338989257, "step": 765 }, { "epoch": 1.934673366834171, "grad_norm": 2.828125, "learning_rate": 3.0150583678423825e-08, "loss": 1.5515380859375, "step": 770 }, { "epoch": 1.9472361809045227, "grad_norm": 2.625, "learning_rate": 2.002444179585339e-08, "loss": 1.529140090942383, "step": 775 }, { "epoch": 1.9597989949748744, "grad_norm": 2.0625, "learning_rate": 1.1959961510018547e-08, "loss": 1.5413886070251466, "step": 780 }, { "epoch": 1.9723618090452262, "grad_norm": 2.15625, "learning_rate": 5.9604814375685546e-09, "loss": 1.5065739631652832, "step": 785 }, { "epoch": 1.984924623115578, "grad_norm": 2.390625, "learning_rate": 2.028485305082817e-09, "loss": 1.575080680847168, "step": 790 }, { "epoch": 1.9974874371859297, "grad_norm": 2.171875, "learning_rate": 1.6560092083672818e-10, "loss": 1.5597107887268067, "step": 795 }, { "epoch": 2.0, "step": 796, "total_flos": 3.833069438698537e+17, "train_loss": 1.6445721892555754, "train_runtime": 6416.9422, "train_samples_per_second": 0.496, "train_steps_per_second": 0.124 } ], "logging_steps": 5, "max_steps": 796, "num_input_tokens_seen": 0, "num_train_epochs": 2, "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": 3.833069438698537e+17, "train_batch_size": 1, "trial_name": null, "trial_params": null }