{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 3.380776902532202, "eval_steps": 500, "global_step": 100000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "epoch": 0.0003380776902532202, "grad_norm": 5.1544413566589355, "learning_rate": 1.8e-07, "loss": 0.6157, "step": 10 }, { "epoch": 0.0006761553805064404, "grad_norm": 4.660217761993408, "learning_rate": 3.8e-07, "loss": 0.6349, "step": 20 }, { "epoch": 0.0010142330707596605, "grad_norm": 5.005427360534668, "learning_rate": 5.8e-07, "loss": 0.6634, "step": 30 }, { "epoch": 0.0013523107610128808, "grad_norm": 3.0059070587158203, "learning_rate": 7.8e-07, "loss": 0.5849, "step": 40 }, { "epoch": 0.001690388451266101, "grad_norm": 3.709200620651245, "learning_rate": 9.8e-07, "loss": 0.5686, "step": 50 }, { "epoch": 0.002028466141519321, "grad_norm": 1.9112032651901245, "learning_rate": 1.18e-06, "loss": 0.5392, "step": 60 }, { "epoch": 0.0023665438317725415, "grad_norm": 2.1092212200164795, "learning_rate": 1.3800000000000001e-06, "loss": 0.475, "step": 70 }, { "epoch": 0.0027046215220257615, "grad_norm": 1.263190746307373, "learning_rate": 1.5800000000000003e-06, "loss": 0.3987, "step": 80 }, { "epoch": 0.0030426992122789816, "grad_norm": 1.043360710144043, "learning_rate": 1.7800000000000001e-06, "loss": 0.3447, "step": 90 }, { "epoch": 0.003380776902532202, "grad_norm": 0.8224862217903137, "learning_rate": 1.98e-06, "loss": 0.2965, "step": 100 }, { "epoch": 0.003718854592785422, "grad_norm": 0.9064249396324158, "learning_rate": 2.1800000000000003e-06, "loss": 0.2707, "step": 110 }, { "epoch": 0.004056932283038642, "grad_norm": 0.7632829546928406, "learning_rate": 2.38e-06, "loss": 0.2516, "step": 120 }, { "epoch": 0.004395009973291862, "grad_norm": 0.8271092772483826, "learning_rate": 2.5800000000000003e-06, "loss": 0.2168, "step": 130 }, { "epoch": 0.004733087663545083, "grad_norm": 0.664084255695343, "learning_rate": 2.78e-06, "loss": 0.2003, "step": 140 }, { "epoch": 0.005071165353798303, "grad_norm": 0.9221041798591614, "learning_rate": 2.9800000000000003e-06, "loss": 0.1976, "step": 150 }, { "epoch": 0.005409243044051523, "grad_norm": 0.7425479292869568, "learning_rate": 3.1800000000000005e-06, "loss": 0.188, "step": 160 }, { "epoch": 0.005747320734304743, "grad_norm": 0.7374719977378845, "learning_rate": 3.38e-06, "loss": 0.1731, "step": 170 }, { "epoch": 0.006085398424557963, "grad_norm": 1.2100754976272583, "learning_rate": 3.58e-06, "loss": 0.1556, "step": 180 }, { "epoch": 0.006423476114811184, "grad_norm": 1.2192373275756836, "learning_rate": 3.7800000000000002e-06, "loss": 0.1671, "step": 190 }, { "epoch": 0.006761553805064404, "grad_norm": 1.0434337854385376, "learning_rate": 3.98e-06, "loss": 0.1501, "step": 200 }, { "epoch": 0.007099631495317624, "grad_norm": 2.4590256214141846, "learning_rate": 4.18e-06, "loss": 0.1625, "step": 210 }, { "epoch": 0.007437709185570844, "grad_norm": 1.1473290920257568, "learning_rate": 4.38e-06, "loss": 0.1514, "step": 220 }, { "epoch": 0.007775786875824064, "grad_norm": 0.6452512741088867, "learning_rate": 4.58e-06, "loss": 0.1442, "step": 230 }, { "epoch": 0.008113864566077284, "grad_norm": 1.0163077116012573, "learning_rate": 4.780000000000001e-06, "loss": 0.1395, "step": 240 }, { "epoch": 0.008451942256330504, "grad_norm": 0.8165188431739807, "learning_rate": 4.98e-06, "loss": 0.1338, "step": 250 }, { "epoch": 0.008790019946583724, "grad_norm": 0.9194470643997192, "learning_rate": 5.18e-06, "loss": 0.1386, "step": 260 }, { "epoch": 0.009128097636836946, "grad_norm": 1.2051600217819214, "learning_rate": 5.38e-06, "loss": 0.1324, "step": 270 }, { "epoch": 0.009466175327090166, "grad_norm": 1.1965700387954712, "learning_rate": 5.580000000000001e-06, "loss": 0.133, "step": 280 }, { "epoch": 0.009804253017343386, "grad_norm": 1.3689740896224976, "learning_rate": 5.78e-06, "loss": 0.1144, "step": 290 }, { "epoch": 0.010142330707596606, "grad_norm": 1.0175467729568481, "learning_rate": 5.98e-06, "loss": 0.1313, "step": 300 }, { "epoch": 0.010480408397849826, "grad_norm": 1.3773566484451294, "learning_rate": 6.18e-06, "loss": 0.1222, "step": 310 }, { "epoch": 0.010818486088103046, "grad_norm": 1.0046466588974, "learning_rate": 6.38e-06, "loss": 0.1236, "step": 320 }, { "epoch": 0.011156563778356266, "grad_norm": 1.7571063041687012, "learning_rate": 6.58e-06, "loss": 0.117, "step": 330 }, { "epoch": 0.011494641468609486, "grad_norm": 1.0244174003601074, "learning_rate": 6.78e-06, "loss": 0.1213, "step": 340 }, { "epoch": 0.011832719158862706, "grad_norm": 1.5028676986694336, "learning_rate": 6.98e-06, "loss": 0.1111, "step": 350 }, { "epoch": 0.012170796849115926, "grad_norm": 1.1787211894989014, "learning_rate": 7.180000000000001e-06, "loss": 0.1147, "step": 360 }, { "epoch": 0.012508874539369146, "grad_norm": 1.4157413244247437, "learning_rate": 7.3800000000000005e-06, "loss": 0.1144, "step": 370 }, { "epoch": 0.012846952229622368, "grad_norm": 1.920854926109314, "learning_rate": 7.580000000000001e-06, "loss": 0.1138, "step": 380 }, { "epoch": 0.013185029919875588, "grad_norm": 1.057614803314209, "learning_rate": 7.78e-06, "loss": 0.1101, "step": 390 }, { "epoch": 0.013523107610128808, "grad_norm": 0.7947180271148682, "learning_rate": 7.98e-06, "loss": 0.1168, "step": 400 }, { "epoch": 0.013861185300382028, "grad_norm": 0.8044916391372681, "learning_rate": 8.18e-06, "loss": 0.111, "step": 410 }, { "epoch": 0.014199262990635248, "grad_norm": 0.6675881147384644, "learning_rate": 8.380000000000001e-06, "loss": 0.1126, "step": 420 }, { "epoch": 0.014537340680888468, "grad_norm": 0.5505193471908569, "learning_rate": 8.580000000000001e-06, "loss": 0.1002, "step": 430 }, { "epoch": 0.014875418371141688, "grad_norm": 1.0551629066467285, "learning_rate": 8.78e-06, "loss": 0.1061, "step": 440 }, { "epoch": 0.015213496061394908, "grad_norm": 1.9566686153411865, "learning_rate": 8.98e-06, "loss": 0.1063, "step": 450 }, { "epoch": 0.015551573751648128, "grad_norm": 1.1958558559417725, "learning_rate": 9.180000000000002e-06, "loss": 0.1122, "step": 460 }, { "epoch": 0.01588965144190135, "grad_norm": 0.8292704224586487, "learning_rate": 9.38e-06, "loss": 0.0988, "step": 470 }, { "epoch": 0.01622772913215457, "grad_norm": 0.8114256858825684, "learning_rate": 9.58e-06, "loss": 0.1048, "step": 480 }, { "epoch": 0.01656580682240779, "grad_norm": 1.1357492208480835, "learning_rate": 9.78e-06, "loss": 0.1, "step": 490 }, { "epoch": 0.01690388451266101, "grad_norm": 0.8385120034217834, "learning_rate": 9.980000000000001e-06, "loss": 0.1024, "step": 500 }, { "epoch": 0.01724196220291423, "grad_norm": 0.8120285868644714, "learning_rate": 1.018e-05, "loss": 0.1031, "step": 510 }, { "epoch": 0.01758003989316745, "grad_norm": 0.8348689675331116, "learning_rate": 1.038e-05, "loss": 0.1019, "step": 520 }, { "epoch": 0.01791811758342067, "grad_norm": 0.9193062782287598, "learning_rate": 1.058e-05, "loss": 0.1092, "step": 530 }, { "epoch": 0.018256195273673892, "grad_norm": 0.7177342176437378, "learning_rate": 1.0780000000000002e-05, "loss": 0.1003, "step": 540 }, { "epoch": 0.01859427296392711, "grad_norm": 0.7342193722724915, "learning_rate": 1.098e-05, "loss": 0.097, "step": 550 }, { "epoch": 0.018932350654180332, "grad_norm": 1.6237269639968872, "learning_rate": 1.118e-05, "loss": 0.0998, "step": 560 }, { "epoch": 0.01927042834443355, "grad_norm": 0.5930098295211792, "learning_rate": 1.1380000000000001e-05, "loss": 0.1095, "step": 570 }, { "epoch": 0.019608506034686772, "grad_norm": 0.5552340745925903, "learning_rate": 1.1580000000000001e-05, "loss": 0.0963, "step": 580 }, { "epoch": 0.01994658372493999, "grad_norm": 1.0610443353652954, "learning_rate": 1.178e-05, "loss": 0.0924, "step": 590 }, { "epoch": 0.020284661415193212, "grad_norm": 0.60377037525177, "learning_rate": 1.198e-05, "loss": 0.1023, "step": 600 }, { "epoch": 0.02062273910544643, "grad_norm": 0.7435840368270874, "learning_rate": 1.2180000000000002e-05, "loss": 0.0978, "step": 610 }, { "epoch": 0.020960816795699652, "grad_norm": 0.828460693359375, "learning_rate": 1.238e-05, "loss": 0.1125, "step": 620 }, { "epoch": 0.02129889448595287, "grad_norm": 0.767156183719635, "learning_rate": 1.258e-05, "loss": 0.1103, "step": 630 }, { "epoch": 0.021636972176206092, "grad_norm": 0.5024586319923401, "learning_rate": 1.278e-05, "loss": 0.096, "step": 640 }, { "epoch": 0.021975049866459314, "grad_norm": 0.5025770664215088, "learning_rate": 1.2980000000000001e-05, "loss": 0.0953, "step": 650 }, { "epoch": 0.022313127556712532, "grad_norm": 1.1363472938537598, "learning_rate": 1.3180000000000001e-05, "loss": 0.0971, "step": 660 }, { "epoch": 0.022651205246965754, "grad_norm": 0.4046609401702881, "learning_rate": 1.338e-05, "loss": 0.0933, "step": 670 }, { "epoch": 0.022989282937218972, "grad_norm": 0.5971781015396118, "learning_rate": 1.358e-05, "loss": 0.0896, "step": 680 }, { "epoch": 0.023327360627472194, "grad_norm": 0.6176867485046387, "learning_rate": 1.3780000000000002e-05, "loss": 0.0864, "step": 690 }, { "epoch": 0.023665438317725412, "grad_norm": 0.7840899229049683, "learning_rate": 1.3980000000000002e-05, "loss": 0.0829, "step": 700 }, { "epoch": 0.024003516007978634, "grad_norm": 0.7420234680175781, "learning_rate": 1.4180000000000001e-05, "loss": 0.0936, "step": 710 }, { "epoch": 0.024341593698231852, "grad_norm": 1.0565346479415894, "learning_rate": 1.4380000000000001e-05, "loss": 0.0943, "step": 720 }, { "epoch": 0.024679671388485074, "grad_norm": 0.8061684370040894, "learning_rate": 1.4580000000000003e-05, "loss": 0.101, "step": 730 }, { "epoch": 0.025017749078738293, "grad_norm": 0.542325496673584, "learning_rate": 1.4779999999999999e-05, "loss": 0.0912, "step": 740 }, { "epoch": 0.025355826768991514, "grad_norm": 0.9542072415351868, "learning_rate": 1.4979999999999999e-05, "loss": 0.0976, "step": 750 }, { "epoch": 0.025693904459244736, "grad_norm": 0.6654790043830872, "learning_rate": 1.518e-05, "loss": 0.088, "step": 760 }, { "epoch": 0.026031982149497954, "grad_norm": 0.7393090128898621, "learning_rate": 1.538e-05, "loss": 0.096, "step": 770 }, { "epoch": 0.026370059839751176, "grad_norm": 0.8434017896652222, "learning_rate": 1.558e-05, "loss": 0.0946, "step": 780 }, { "epoch": 0.026708137530004394, "grad_norm": 0.7358160018920898, "learning_rate": 1.578e-05, "loss": 0.0925, "step": 790 }, { "epoch": 0.027046215220257616, "grad_norm": 0.6388516426086426, "learning_rate": 1.598e-05, "loss": 0.087, "step": 800 }, { "epoch": 0.027384292910510834, "grad_norm": 0.7495275735855103, "learning_rate": 1.618e-05, "loss": 0.085, "step": 810 }, { "epoch": 0.027722370600764056, "grad_norm": 0.44026559591293335, "learning_rate": 1.6380000000000002e-05, "loss": 0.0832, "step": 820 }, { "epoch": 0.028060448291017275, "grad_norm": 0.5595347881317139, "learning_rate": 1.658e-05, "loss": 0.0881, "step": 830 }, { "epoch": 0.028398525981270496, "grad_norm": 0.7202157974243164, "learning_rate": 1.6780000000000002e-05, "loss": 0.0955, "step": 840 }, { "epoch": 0.028736603671523715, "grad_norm": 0.7819251418113708, "learning_rate": 1.698e-05, "loss": 0.0903, "step": 850 }, { "epoch": 0.029074681361776936, "grad_norm": 0.7201398015022278, "learning_rate": 1.718e-05, "loss": 0.0941, "step": 860 }, { "epoch": 0.029412759052030158, "grad_norm": 0.7965492606163025, "learning_rate": 1.7380000000000003e-05, "loss": 0.0916, "step": 870 }, { "epoch": 0.029750836742283376, "grad_norm": 0.840337872505188, "learning_rate": 1.758e-05, "loss": 0.0879, "step": 880 }, { "epoch": 0.030088914432536598, "grad_norm": 0.6120259761810303, "learning_rate": 1.7780000000000003e-05, "loss": 0.0883, "step": 890 }, { "epoch": 0.030426992122789816, "grad_norm": 1.18967604637146, "learning_rate": 1.798e-05, "loss": 0.0884, "step": 900 }, { "epoch": 0.030765069813043038, "grad_norm": 0.5516895055770874, "learning_rate": 1.818e-05, "loss": 0.0842, "step": 910 }, { "epoch": 0.031103147503296256, "grad_norm": 0.8786369562149048, "learning_rate": 1.838e-05, "loss": 0.0835, "step": 920 }, { "epoch": 0.03144122519354948, "grad_norm": 0.5816625952720642, "learning_rate": 1.858e-05, "loss": 0.0887, "step": 930 }, { "epoch": 0.0317793028838027, "grad_norm": 0.5245797038078308, "learning_rate": 1.878e-05, "loss": 0.0868, "step": 940 }, { "epoch": 0.032117380574055915, "grad_norm": 0.7586566805839539, "learning_rate": 1.898e-05, "loss": 0.0927, "step": 950 }, { "epoch": 0.03245545826430914, "grad_norm": 0.7981254458427429, "learning_rate": 1.918e-05, "loss": 0.0811, "step": 960 }, { "epoch": 0.03279353595456236, "grad_norm": 0.812903881072998, "learning_rate": 1.938e-05, "loss": 0.0924, "step": 970 }, { "epoch": 0.03313161364481558, "grad_norm": 0.499580442905426, "learning_rate": 1.9580000000000002e-05, "loss": 0.0881, "step": 980 }, { "epoch": 0.0334696913350688, "grad_norm": 0.6908408999443054, "learning_rate": 1.978e-05, "loss": 0.0968, "step": 990 }, { "epoch": 0.03380776902532202, "grad_norm": 0.7725153565406799, "learning_rate": 1.9980000000000002e-05, "loss": 0.0861, "step": 1000 }, { "epoch": 0.03414584671557524, "grad_norm": 0.733169674873352, "learning_rate": 2.0180000000000003e-05, "loss": 0.0844, "step": 1010 }, { "epoch": 0.03448392440582846, "grad_norm": 0.756138801574707, "learning_rate": 2.038e-05, "loss": 0.0924, "step": 1020 }, { "epoch": 0.03482200209608168, "grad_norm": 0.5602319240570068, "learning_rate": 2.0580000000000003e-05, "loss": 0.0913, "step": 1030 }, { "epoch": 0.0351600797863349, "grad_norm": 1.3122602701187134, "learning_rate": 2.078e-05, "loss": 0.085, "step": 1040 }, { "epoch": 0.03549815747658812, "grad_norm": 0.536663830280304, "learning_rate": 2.098e-05, "loss": 0.0917, "step": 1050 }, { "epoch": 0.03583623516684134, "grad_norm": 0.5672929286956787, "learning_rate": 2.118e-05, "loss": 0.0833, "step": 1060 }, { "epoch": 0.03617431285709456, "grad_norm": 0.7348262071609497, "learning_rate": 2.138e-05, "loss": 0.0814, "step": 1070 }, { "epoch": 0.036512390547347784, "grad_norm": 0.549435555934906, "learning_rate": 2.158e-05, "loss": 0.0888, "step": 1080 }, { "epoch": 0.036850468237601, "grad_norm": 0.5489999651908875, "learning_rate": 2.178e-05, "loss": 0.0907, "step": 1090 }, { "epoch": 0.03718854592785422, "grad_norm": 0.5305888652801514, "learning_rate": 2.198e-05, "loss": 0.0897, "step": 1100 }, { "epoch": 0.03752662361810744, "grad_norm": 0.8519772887229919, "learning_rate": 2.218e-05, "loss": 0.0855, "step": 1110 }, { "epoch": 0.037864701308360664, "grad_norm": 0.4052690863609314, "learning_rate": 2.2380000000000003e-05, "loss": 0.0785, "step": 1120 }, { "epoch": 0.03820277899861388, "grad_norm": 0.5775979161262512, "learning_rate": 2.258e-05, "loss": 0.0759, "step": 1130 }, { "epoch": 0.0385408566888671, "grad_norm": 0.5628293752670288, "learning_rate": 2.2780000000000002e-05, "loss": 0.0823, "step": 1140 }, { "epoch": 0.03887893437912032, "grad_norm": 0.80622398853302, "learning_rate": 2.298e-05, "loss": 0.0886, "step": 1150 }, { "epoch": 0.039217012069373544, "grad_norm": 0.7505769729614258, "learning_rate": 2.318e-05, "loss": 0.0897, "step": 1160 }, { "epoch": 0.03955508975962676, "grad_norm": 0.4784063398838043, "learning_rate": 2.3380000000000003e-05, "loss": 0.0878, "step": 1170 }, { "epoch": 0.03989316744987998, "grad_norm": 0.5376147627830505, "learning_rate": 2.358e-05, "loss": 0.0876, "step": 1180 }, { "epoch": 0.0402312451401332, "grad_norm": 0.5154613256454468, "learning_rate": 2.3780000000000003e-05, "loss": 0.082, "step": 1190 }, { "epoch": 0.040569322830386424, "grad_norm": 0.39426231384277344, "learning_rate": 2.398e-05, "loss": 0.0873, "step": 1200 }, { "epoch": 0.040907400520639646, "grad_norm": 0.7480712532997131, "learning_rate": 2.418e-05, "loss": 0.0796, "step": 1210 }, { "epoch": 0.04124547821089286, "grad_norm": 1.081687092781067, "learning_rate": 2.438e-05, "loss": 0.0821, "step": 1220 }, { "epoch": 0.04158355590114608, "grad_norm": 0.5666053295135498, "learning_rate": 2.4580000000000002e-05, "loss": 0.0884, "step": 1230 }, { "epoch": 0.041921633591399304, "grad_norm": 0.4692493975162506, "learning_rate": 2.478e-05, "loss": 0.0872, "step": 1240 }, { "epoch": 0.042259711281652526, "grad_norm": 0.9542802572250366, "learning_rate": 2.498e-05, "loss": 0.0846, "step": 1250 }, { "epoch": 0.04259778897190574, "grad_norm": 0.7030493021011353, "learning_rate": 2.5180000000000003e-05, "loss": 0.087, "step": 1260 }, { "epoch": 0.04293586666215896, "grad_norm": 0.7812115550041199, "learning_rate": 2.5380000000000004e-05, "loss": 0.0843, "step": 1270 }, { "epoch": 0.043273944352412184, "grad_norm": 0.6428888440132141, "learning_rate": 2.5580000000000002e-05, "loss": 0.0939, "step": 1280 }, { "epoch": 0.043612022042665406, "grad_norm": 0.4388526976108551, "learning_rate": 2.5779999999999997e-05, "loss": 0.08, "step": 1290 }, { "epoch": 0.04395009973291863, "grad_norm": 0.5861950516700745, "learning_rate": 2.598e-05, "loss": 0.0836, "step": 1300 }, { "epoch": 0.04428817742317184, "grad_norm": 0.5309887528419495, "learning_rate": 2.618e-05, "loss": 0.0839, "step": 1310 }, { "epoch": 0.044626255113425065, "grad_norm": 0.5871047973632812, "learning_rate": 2.6379999999999998e-05, "loss": 0.0828, "step": 1320 }, { "epoch": 0.044964332803678286, "grad_norm": 0.44460949301719666, "learning_rate": 2.658e-05, "loss": 0.0855, "step": 1330 }, { "epoch": 0.04530241049393151, "grad_norm": 0.7911015748977661, "learning_rate": 2.678e-05, "loss": 0.0879, "step": 1340 }, { "epoch": 0.04564048818418472, "grad_norm": 0.48826906085014343, "learning_rate": 2.698e-05, "loss": 0.0835, "step": 1350 }, { "epoch": 0.045978565874437945, "grad_norm": 0.8864228129386902, "learning_rate": 2.718e-05, "loss": 0.0846, "step": 1360 }, { "epoch": 0.046316643564691166, "grad_norm": 0.6374914050102234, "learning_rate": 2.738e-05, "loss": 0.0776, "step": 1370 }, { "epoch": 0.04665472125494439, "grad_norm": 0.38513702154159546, "learning_rate": 2.758e-05, "loss": 0.0795, "step": 1380 }, { "epoch": 0.0469927989451976, "grad_norm": 0.4453953802585602, "learning_rate": 2.778e-05, "loss": 0.079, "step": 1390 }, { "epoch": 0.047330876635450825, "grad_norm": 0.5258950591087341, "learning_rate": 2.798e-05, "loss": 0.0798, "step": 1400 }, { "epoch": 0.04766895432570405, "grad_norm": 0.5197187066078186, "learning_rate": 2.818e-05, "loss": 0.0869, "step": 1410 }, { "epoch": 0.04800703201595727, "grad_norm": 0.4635797441005707, "learning_rate": 2.8380000000000003e-05, "loss": 0.08, "step": 1420 }, { "epoch": 0.04834510970621049, "grad_norm": 0.38641539216041565, "learning_rate": 2.858e-05, "loss": 0.0827, "step": 1430 }, { "epoch": 0.048683187396463705, "grad_norm": 0.5362265706062317, "learning_rate": 2.8780000000000002e-05, "loss": 0.0787, "step": 1440 }, { "epoch": 0.04902126508671693, "grad_norm": 0.7841898798942566, "learning_rate": 2.898e-05, "loss": 0.0834, "step": 1450 }, { "epoch": 0.04935934277697015, "grad_norm": 0.7519902586936951, "learning_rate": 2.9180000000000002e-05, "loss": 0.0905, "step": 1460 }, { "epoch": 0.04969742046722337, "grad_norm": 0.7080463767051697, "learning_rate": 2.9380000000000003e-05, "loss": 0.0919, "step": 1470 }, { "epoch": 0.050035498157476585, "grad_norm": 0.5496295690536499, "learning_rate": 2.958e-05, "loss": 0.0808, "step": 1480 }, { "epoch": 0.05037357584772981, "grad_norm": 0.4630584120750427, "learning_rate": 2.9780000000000003e-05, "loss": 0.0726, "step": 1490 }, { "epoch": 0.05071165353798303, "grad_norm": 0.5246527791023254, "learning_rate": 2.998e-05, "loss": 0.0747, "step": 1500 }, { "epoch": 0.05104973122823625, "grad_norm": 0.5106741786003113, "learning_rate": 3.0180000000000002e-05, "loss": 0.0839, "step": 1510 }, { "epoch": 0.05138780891848947, "grad_norm": 0.6252930760383606, "learning_rate": 3.0380000000000004e-05, "loss": 0.0799, "step": 1520 }, { "epoch": 0.05172588660874269, "grad_norm": 0.544367253780365, "learning_rate": 3.058e-05, "loss": 0.0768, "step": 1530 }, { "epoch": 0.05206396429899591, "grad_norm": 0.5484617352485657, "learning_rate": 3.078e-05, "loss": 0.086, "step": 1540 }, { "epoch": 0.05240204198924913, "grad_norm": 0.48075827956199646, "learning_rate": 3.0980000000000005e-05, "loss": 0.0799, "step": 1550 }, { "epoch": 0.05274011967950235, "grad_norm": 0.4036509692668915, "learning_rate": 3.118e-05, "loss": 0.0768, "step": 1560 }, { "epoch": 0.05307819736975557, "grad_norm": 0.37452998757362366, "learning_rate": 3.138e-05, "loss": 0.0782, "step": 1570 }, { "epoch": 0.05341627506000879, "grad_norm": 0.5780650973320007, "learning_rate": 3.1580000000000006e-05, "loss": 0.0763, "step": 1580 }, { "epoch": 0.05375435275026201, "grad_norm": 0.7501440048217773, "learning_rate": 3.1780000000000004e-05, "loss": 0.0815, "step": 1590 }, { "epoch": 0.05409243044051523, "grad_norm": 0.5264964699745178, "learning_rate": 3.198e-05, "loss": 0.0731, "step": 1600 }, { "epoch": 0.05443050813076845, "grad_norm": 0.6407552361488342, "learning_rate": 3.218e-05, "loss": 0.0816, "step": 1610 }, { "epoch": 0.05476858582102167, "grad_norm": 0.5762882828712463, "learning_rate": 3.238e-05, "loss": 0.0871, "step": 1620 }, { "epoch": 0.05510666351127489, "grad_norm": 0.5342503786087036, "learning_rate": 3.2579999999999996e-05, "loss": 0.0755, "step": 1630 }, { "epoch": 0.05544474120152811, "grad_norm": 0.4961090087890625, "learning_rate": 3.278e-05, "loss": 0.0749, "step": 1640 }, { "epoch": 0.055782818891781334, "grad_norm": 0.4516008794307709, "learning_rate": 3.298e-05, "loss": 0.0746, "step": 1650 }, { "epoch": 0.05612089658203455, "grad_norm": 0.47903338074684143, "learning_rate": 3.318e-05, "loss": 0.0769, "step": 1660 }, { "epoch": 0.05645897427228777, "grad_norm": 0.4267156720161438, "learning_rate": 3.338e-05, "loss": 0.0729, "step": 1670 }, { "epoch": 0.05679705196254099, "grad_norm": 0.6040109395980835, "learning_rate": 3.358e-05, "loss": 0.0791, "step": 1680 }, { "epoch": 0.057135129652794214, "grad_norm": 0.4421294629573822, "learning_rate": 3.378e-05, "loss": 0.0842, "step": 1690 }, { "epoch": 0.05747320734304743, "grad_norm": 0.5595395565032959, "learning_rate": 3.398e-05, "loss": 0.0851, "step": 1700 }, { "epoch": 0.05781128503330065, "grad_norm": 0.8433418273925781, "learning_rate": 3.418e-05, "loss": 0.0833, "step": 1710 }, { "epoch": 0.05814936272355387, "grad_norm": 0.4289488196372986, "learning_rate": 3.438e-05, "loss": 0.0843, "step": 1720 }, { "epoch": 0.058487440413807094, "grad_norm": 0.510057270526886, "learning_rate": 3.4580000000000004e-05, "loss": 0.0829, "step": 1730 }, { "epoch": 0.058825518104060316, "grad_norm": 0.38354504108428955, "learning_rate": 3.478e-05, "loss": 0.081, "step": 1740 }, { "epoch": 0.05916359579431353, "grad_norm": 0.6776582598686218, "learning_rate": 3.498e-05, "loss": 0.081, "step": 1750 }, { "epoch": 0.05950167348456675, "grad_norm": 0.504172146320343, "learning_rate": 3.518e-05, "loss": 0.0793, "step": 1760 }, { "epoch": 0.059839751174819975, "grad_norm": 0.5406721234321594, "learning_rate": 3.5380000000000003e-05, "loss": 0.0787, "step": 1770 }, { "epoch": 0.060177828865073196, "grad_norm": 0.5150426030158997, "learning_rate": 3.558e-05, "loss": 0.0767, "step": 1780 }, { "epoch": 0.06051590655532641, "grad_norm": 0.5520379543304443, "learning_rate": 3.578e-05, "loss": 0.0762, "step": 1790 }, { "epoch": 0.06085398424557963, "grad_norm": 0.5183428525924683, "learning_rate": 3.5980000000000004e-05, "loss": 0.0817, "step": 1800 }, { "epoch": 0.061192061935832855, "grad_norm": 0.5228794813156128, "learning_rate": 3.618e-05, "loss": 0.0782, "step": 1810 }, { "epoch": 0.061530139626086076, "grad_norm": 0.5221309065818787, "learning_rate": 3.638e-05, "loss": 0.0842, "step": 1820 }, { "epoch": 0.0618682173163393, "grad_norm": 0.5121737718582153, "learning_rate": 3.6580000000000006e-05, "loss": 0.0883, "step": 1830 }, { "epoch": 0.06220629500659251, "grad_norm": 0.4922197163105011, "learning_rate": 3.6780000000000004e-05, "loss": 0.0796, "step": 1840 }, { "epoch": 0.06254437269684573, "grad_norm": 0.5732623338699341, "learning_rate": 3.698e-05, "loss": 0.0791, "step": 1850 }, { "epoch": 0.06288245038709896, "grad_norm": 0.6485913991928101, "learning_rate": 3.7180000000000007e-05, "loss": 0.0879, "step": 1860 }, { "epoch": 0.06322052807735218, "grad_norm": 0.957889974117279, "learning_rate": 3.7380000000000005e-05, "loss": 0.0894, "step": 1870 }, { "epoch": 0.0635586057676054, "grad_norm": 0.5377280116081238, "learning_rate": 3.758e-05, "loss": 0.0864, "step": 1880 }, { "epoch": 0.06389668345785862, "grad_norm": 0.8683066964149475, "learning_rate": 3.778000000000001e-05, "loss": 0.0846, "step": 1890 }, { "epoch": 0.06423476114811183, "grad_norm": 0.44290274381637573, "learning_rate": 3.7980000000000006e-05, "loss": 0.0793, "step": 1900 }, { "epoch": 0.06457283883836505, "grad_norm": 0.6479418277740479, "learning_rate": 3.818e-05, "loss": 0.0853, "step": 1910 }, { "epoch": 0.06491091652861827, "grad_norm": 0.5280694365501404, "learning_rate": 3.838e-05, "loss": 0.0755, "step": 1920 }, { "epoch": 0.0652489942188715, "grad_norm": 0.7745735049247742, "learning_rate": 3.858e-05, "loss": 0.0765, "step": 1930 }, { "epoch": 0.06558707190912472, "grad_norm": 0.6081345677375793, "learning_rate": 3.878e-05, "loss": 0.0753, "step": 1940 }, { "epoch": 0.06592514959937794, "grad_norm": 0.5984487533569336, "learning_rate": 3.898e-05, "loss": 0.0754, "step": 1950 }, { "epoch": 0.06626322728963116, "grad_norm": 0.3525026738643646, "learning_rate": 3.918e-05, "loss": 0.077, "step": 1960 }, { "epoch": 0.06660130497988438, "grad_norm": 0.3330334424972534, "learning_rate": 3.938e-05, "loss": 0.0782, "step": 1970 }, { "epoch": 0.0669393826701376, "grad_norm": 0.4915262460708618, "learning_rate": 3.958e-05, "loss": 0.0789, "step": 1980 }, { "epoch": 0.06727746036039081, "grad_norm": 0.45514610409736633, "learning_rate": 3.978e-05, "loss": 0.075, "step": 1990 }, { "epoch": 0.06761553805064403, "grad_norm": 0.3629968464374542, "learning_rate": 3.998e-05, "loss": 0.0812, "step": 2000 }, { "epoch": 0.06795361574089726, "grad_norm": 0.4811018407344818, "learning_rate": 4.018e-05, "loss": 0.0768, "step": 2010 }, { "epoch": 0.06829169343115048, "grad_norm": 0.9736223220825195, "learning_rate": 4.038e-05, "loss": 0.0771, "step": 2020 }, { "epoch": 0.0686297711214037, "grad_norm": 0.4855140745639801, "learning_rate": 4.058e-05, "loss": 0.0769, "step": 2030 }, { "epoch": 0.06896784881165692, "grad_norm": 0.4808104336261749, "learning_rate": 4.078e-05, "loss": 0.0729, "step": 2040 }, { "epoch": 0.06930592650191014, "grad_norm": 0.4791993498802185, "learning_rate": 4.0980000000000004e-05, "loss": 0.0789, "step": 2050 }, { "epoch": 0.06964400419216336, "grad_norm": 0.4556324779987335, "learning_rate": 4.118e-05, "loss": 0.0759, "step": 2060 }, { "epoch": 0.06998208188241659, "grad_norm": 0.5070098638534546, "learning_rate": 4.138e-05, "loss": 0.079, "step": 2070 }, { "epoch": 0.0703201595726698, "grad_norm": 0.37392356991767883, "learning_rate": 4.1580000000000005e-05, "loss": 0.0759, "step": 2080 }, { "epoch": 0.07065823726292302, "grad_norm": 0.4467257559299469, "learning_rate": 4.178e-05, "loss": 0.077, "step": 2090 }, { "epoch": 0.07099631495317624, "grad_norm": 0.5423240661621094, "learning_rate": 4.198e-05, "loss": 0.0751, "step": 2100 }, { "epoch": 0.07133439264342946, "grad_norm": 0.5800175070762634, "learning_rate": 4.2180000000000006e-05, "loss": 0.0743, "step": 2110 }, { "epoch": 0.07167247033368268, "grad_norm": 0.36543741822242737, "learning_rate": 4.2380000000000004e-05, "loss": 0.0724, "step": 2120 }, { "epoch": 0.0720105480239359, "grad_norm": 0.4895193576812744, "learning_rate": 4.258e-05, "loss": 0.089, "step": 2130 }, { "epoch": 0.07234862571418912, "grad_norm": 0.37282228469848633, "learning_rate": 4.278e-05, "loss": 0.0799, "step": 2140 }, { "epoch": 0.07268670340444235, "grad_norm": 0.478254497051239, "learning_rate": 4.2980000000000005e-05, "loss": 0.076, "step": 2150 }, { "epoch": 0.07302478109469557, "grad_norm": 0.3850748538970947, "learning_rate": 4.318e-05, "loss": 0.0712, "step": 2160 }, { "epoch": 0.07336285878494878, "grad_norm": 0.6348583102226257, "learning_rate": 4.338e-05, "loss": 0.0769, "step": 2170 }, { "epoch": 0.073700936475202, "grad_norm": 0.44916725158691406, "learning_rate": 4.3580000000000006e-05, "loss": 0.0731, "step": 2180 }, { "epoch": 0.07403901416545522, "grad_norm": 0.380626380443573, "learning_rate": 4.3780000000000004e-05, "loss": 0.0725, "step": 2190 }, { "epoch": 0.07437709185570844, "grad_norm": 0.6776061654090881, "learning_rate": 4.398e-05, "loss": 0.0817, "step": 2200 }, { "epoch": 0.07471516954596166, "grad_norm": 0.36932286620140076, "learning_rate": 4.418000000000001e-05, "loss": 0.0805, "step": 2210 }, { "epoch": 0.07505324723621488, "grad_norm": 0.4477256238460541, "learning_rate": 4.438e-05, "loss": 0.0807, "step": 2220 }, { "epoch": 0.0753913249264681, "grad_norm": 0.4106690287590027, "learning_rate": 4.458e-05, "loss": 0.0764, "step": 2230 }, { "epoch": 0.07572940261672133, "grad_norm": 0.34603413939476013, "learning_rate": 4.478e-05, "loss": 0.0682, "step": 2240 }, { "epoch": 0.07606748030697454, "grad_norm": 0.6244999766349792, "learning_rate": 4.498e-05, "loss": 0.0812, "step": 2250 }, { "epoch": 0.07640555799722776, "grad_norm": 0.4542495012283325, "learning_rate": 4.518e-05, "loss": 0.0818, "step": 2260 }, { "epoch": 0.07674363568748098, "grad_norm": 0.40962281823158264, "learning_rate": 4.538e-05, "loss": 0.079, "step": 2270 }, { "epoch": 0.0770817133777342, "grad_norm": 0.6181157827377319, "learning_rate": 4.558e-05, "loss": 0.0761, "step": 2280 }, { "epoch": 0.07741979106798742, "grad_norm": 0.6565207242965698, "learning_rate": 4.578e-05, "loss": 0.0785, "step": 2290 }, { "epoch": 0.07775786875824064, "grad_norm": 0.7621930241584778, "learning_rate": 4.5980000000000004e-05, "loss": 0.0763, "step": 2300 }, { "epoch": 0.07809594644849387, "grad_norm": 0.5904865860939026, "learning_rate": 4.618e-05, "loss": 0.0826, "step": 2310 }, { "epoch": 0.07843402413874709, "grad_norm": 0.3834878206253052, "learning_rate": 4.638e-05, "loss": 0.0809, "step": 2320 }, { "epoch": 0.07877210182900031, "grad_norm": 0.7073374390602112, "learning_rate": 4.6580000000000005e-05, "loss": 0.0767, "step": 2330 }, { "epoch": 0.07911017951925352, "grad_norm": 0.7107484340667725, "learning_rate": 4.678e-05, "loss": 0.0724, "step": 2340 }, { "epoch": 0.07944825720950674, "grad_norm": 0.45734140276908875, "learning_rate": 4.698e-05, "loss": 0.0792, "step": 2350 }, { "epoch": 0.07978633489975996, "grad_norm": 0.35506176948547363, "learning_rate": 4.718e-05, "loss": 0.0735, "step": 2360 }, { "epoch": 0.08012441259001318, "grad_norm": 0.5154622793197632, "learning_rate": 4.7380000000000004e-05, "loss": 0.0719, "step": 2370 }, { "epoch": 0.0804624902802664, "grad_norm": 0.38686102628707886, "learning_rate": 4.758e-05, "loss": 0.0755, "step": 2380 }, { "epoch": 0.08080056797051963, "grad_norm": 0.4182424545288086, "learning_rate": 4.778e-05, "loss": 0.0746, "step": 2390 }, { "epoch": 0.08113864566077285, "grad_norm": 0.5345888137817383, "learning_rate": 4.7980000000000005e-05, "loss": 0.0796, "step": 2400 }, { "epoch": 0.08147672335102607, "grad_norm": 0.48386722803115845, "learning_rate": 4.818e-05, "loss": 0.0745, "step": 2410 }, { "epoch": 0.08181480104127929, "grad_norm": 0.3767995834350586, "learning_rate": 4.838e-05, "loss": 0.0738, "step": 2420 }, { "epoch": 0.0821528787315325, "grad_norm": 0.3065585196018219, "learning_rate": 4.8580000000000006e-05, "loss": 0.0788, "step": 2430 }, { "epoch": 0.08249095642178572, "grad_norm": 0.4055224061012268, "learning_rate": 4.8780000000000004e-05, "loss": 0.0782, "step": 2440 }, { "epoch": 0.08282903411203894, "grad_norm": 0.4253605604171753, "learning_rate": 4.898e-05, "loss": 0.0783, "step": 2450 }, { "epoch": 0.08316711180229217, "grad_norm": 0.39734354615211487, "learning_rate": 4.918000000000001e-05, "loss": 0.0802, "step": 2460 }, { "epoch": 0.08350518949254539, "grad_norm": 0.4253120720386505, "learning_rate": 4.9380000000000005e-05, "loss": 0.0747, "step": 2470 }, { "epoch": 0.08384326718279861, "grad_norm": 0.4678979516029358, "learning_rate": 4.958e-05, "loss": 0.0774, "step": 2480 }, { "epoch": 0.08418134487305183, "grad_norm": 0.3591063618659973, "learning_rate": 4.978e-05, "loss": 0.0841, "step": 2490 }, { "epoch": 0.08451942256330505, "grad_norm": 0.43916839361190796, "learning_rate": 4.9980000000000006e-05, "loss": 0.0754, "step": 2500 }, { "epoch": 0.08485750025355827, "grad_norm": 0.37657347321510315, "learning_rate": 5.0180000000000004e-05, "loss": 0.071, "step": 2510 }, { "epoch": 0.08519557794381148, "grad_norm": 0.3553829491138458, "learning_rate": 5.038e-05, "loss": 0.074, "step": 2520 }, { "epoch": 0.0855336556340647, "grad_norm": 0.5154293179512024, "learning_rate": 5.058000000000001e-05, "loss": 0.0794, "step": 2530 }, { "epoch": 0.08587173332431793, "grad_norm": 0.4263910949230194, "learning_rate": 5.0780000000000005e-05, "loss": 0.0754, "step": 2540 }, { "epoch": 0.08620981101457115, "grad_norm": 0.3905929923057556, "learning_rate": 5.098e-05, "loss": 0.0788, "step": 2550 }, { "epoch": 0.08654788870482437, "grad_norm": 0.6267892122268677, "learning_rate": 5.118000000000001e-05, "loss": 0.0777, "step": 2560 }, { "epoch": 0.08688596639507759, "grad_norm": 0.5951860547065735, "learning_rate": 5.1380000000000006e-05, "loss": 0.0755, "step": 2570 }, { "epoch": 0.08722404408533081, "grad_norm": 0.3124784827232361, "learning_rate": 5.1580000000000004e-05, "loss": 0.0697, "step": 2580 }, { "epoch": 0.08756212177558403, "grad_norm": 0.38159099221229553, "learning_rate": 5.178000000000001e-05, "loss": 0.0757, "step": 2590 }, { "epoch": 0.08790019946583726, "grad_norm": 0.3150147795677185, "learning_rate": 5.198000000000001e-05, "loss": 0.0818, "step": 2600 }, { "epoch": 0.08823827715609046, "grad_norm": 0.2932322025299072, "learning_rate": 5.2180000000000005e-05, "loss": 0.0807, "step": 2610 }, { "epoch": 0.08857635484634369, "grad_norm": 0.30974718928337097, "learning_rate": 5.238000000000001e-05, "loss": 0.0817, "step": 2620 }, { "epoch": 0.08891443253659691, "grad_norm": 0.3877490162849426, "learning_rate": 5.258000000000001e-05, "loss": 0.0785, "step": 2630 }, { "epoch": 0.08925251022685013, "grad_norm": 0.4057486057281494, "learning_rate": 5.2780000000000006e-05, "loss": 0.073, "step": 2640 }, { "epoch": 0.08959058791710335, "grad_norm": 0.6063042879104614, "learning_rate": 5.2980000000000004e-05, "loss": 0.078, "step": 2650 }, { "epoch": 0.08992866560735657, "grad_norm": 0.3359815776348114, "learning_rate": 5.318000000000001e-05, "loss": 0.0718, "step": 2660 }, { "epoch": 0.0902667432976098, "grad_norm": 0.44830065965652466, "learning_rate": 5.338000000000001e-05, "loss": 0.0782, "step": 2670 }, { "epoch": 0.09060482098786302, "grad_norm": 0.47540560364723206, "learning_rate": 5.3580000000000005e-05, "loss": 0.0728, "step": 2680 }, { "epoch": 0.09094289867811624, "grad_norm": 0.5152238011360168, "learning_rate": 5.378e-05, "loss": 0.0854, "step": 2690 }, { "epoch": 0.09128097636836945, "grad_norm": 0.4432814419269562, "learning_rate": 5.3979999999999995e-05, "loss": 0.0663, "step": 2700 }, { "epoch": 0.09161905405862267, "grad_norm": 0.5342723727226257, "learning_rate": 5.418e-05, "loss": 0.0701, "step": 2710 }, { "epoch": 0.09195713174887589, "grad_norm": 0.3759232461452484, "learning_rate": 5.438e-05, "loss": 0.0709, "step": 2720 }, { "epoch": 0.09229520943912911, "grad_norm": 0.5573447942733765, "learning_rate": 5.4579999999999996e-05, "loss": 0.0834, "step": 2730 }, { "epoch": 0.09263328712938233, "grad_norm": 0.45181775093078613, "learning_rate": 5.478e-05, "loss": 0.0789, "step": 2740 }, { "epoch": 0.09297136481963555, "grad_norm": 0.4092378318309784, "learning_rate": 5.498e-05, "loss": 0.0676, "step": 2750 }, { "epoch": 0.09330944250988878, "grad_norm": 0.7554413080215454, "learning_rate": 5.518e-05, "loss": 0.0747, "step": 2760 }, { "epoch": 0.093647520200142, "grad_norm": 0.3709760010242462, "learning_rate": 5.538e-05, "loss": 0.0731, "step": 2770 }, { "epoch": 0.0939855978903952, "grad_norm": 0.5422531366348267, "learning_rate": 5.558e-05, "loss": 0.0723, "step": 2780 }, { "epoch": 0.09432367558064843, "grad_norm": 0.5003961324691772, "learning_rate": 5.578e-05, "loss": 0.0732, "step": 2790 }, { "epoch": 0.09466175327090165, "grad_norm": 0.3397095501422882, "learning_rate": 5.5979999999999996e-05, "loss": 0.0792, "step": 2800 }, { "epoch": 0.09499983096115487, "grad_norm": 0.395230770111084, "learning_rate": 5.618e-05, "loss": 0.0783, "step": 2810 }, { "epoch": 0.0953379086514081, "grad_norm": 0.5867420434951782, "learning_rate": 5.638e-05, "loss": 0.0831, "step": 2820 }, { "epoch": 0.09567598634166131, "grad_norm": 0.2957838475704193, "learning_rate": 5.658e-05, "loss": 0.0736, "step": 2830 }, { "epoch": 0.09601406403191454, "grad_norm": 0.3842826783657074, "learning_rate": 5.678e-05, "loss": 0.0771, "step": 2840 }, { "epoch": 0.09635214172216776, "grad_norm": 0.5651164650917053, "learning_rate": 5.698e-05, "loss": 0.0765, "step": 2850 }, { "epoch": 0.09669021941242098, "grad_norm": 0.5920230746269226, "learning_rate": 5.718e-05, "loss": 0.0775, "step": 2860 }, { "epoch": 0.09702829710267419, "grad_norm": 0.31480997800827026, "learning_rate": 5.738e-05, "loss": 0.0722, "step": 2870 }, { "epoch": 0.09736637479292741, "grad_norm": 0.4435073733329773, "learning_rate": 5.758e-05, "loss": 0.072, "step": 2880 }, { "epoch": 0.09770445248318063, "grad_norm": 0.38070207834243774, "learning_rate": 5.778e-05, "loss": 0.0792, "step": 2890 }, { "epoch": 0.09804253017343385, "grad_norm": 0.36227935552597046, "learning_rate": 5.7980000000000004e-05, "loss": 0.068, "step": 2900 }, { "epoch": 0.09838060786368708, "grad_norm": 0.46968507766723633, "learning_rate": 5.818e-05, "loss": 0.0713, "step": 2910 }, { "epoch": 0.0987186855539403, "grad_norm": 0.5469895005226135, "learning_rate": 5.838e-05, "loss": 0.0756, "step": 2920 }, { "epoch": 0.09905676324419352, "grad_norm": 0.34613287448883057, "learning_rate": 5.858e-05, "loss": 0.0817, "step": 2930 }, { "epoch": 0.09939484093444674, "grad_norm": 0.5897564888000488, "learning_rate": 5.878e-05, "loss": 0.0906, "step": 2940 }, { "epoch": 0.09973291862469996, "grad_norm": 0.5745919346809387, "learning_rate": 5.898e-05, "loss": 0.0743, "step": 2950 }, { "epoch": 0.10007099631495317, "grad_norm": 0.5692716240882874, "learning_rate": 5.918e-05, "loss": 0.0792, "step": 2960 }, { "epoch": 0.10040907400520639, "grad_norm": 1.1950650215148926, "learning_rate": 5.9380000000000004e-05, "loss": 0.0842, "step": 2970 }, { "epoch": 0.10074715169545961, "grad_norm": 0.34542161226272583, "learning_rate": 5.958e-05, "loss": 0.0713, "step": 2980 }, { "epoch": 0.10108522938571284, "grad_norm": 0.4618072211742401, "learning_rate": 5.978e-05, "loss": 0.0774, "step": 2990 }, { "epoch": 0.10142330707596606, "grad_norm": 0.4451849162578583, "learning_rate": 5.9980000000000005e-05, "loss": 0.0773, "step": 3000 }, { "epoch": 0.10176138476621928, "grad_norm": 0.2640022337436676, "learning_rate": 6.018e-05, "loss": 0.0831, "step": 3010 }, { "epoch": 0.1020994624564725, "grad_norm": 0.8933588266372681, "learning_rate": 6.038e-05, "loss": 0.0819, "step": 3020 }, { "epoch": 0.10243754014672572, "grad_norm": 0.3305746912956238, "learning_rate": 6.0580000000000006e-05, "loss": 0.0739, "step": 3030 }, { "epoch": 0.10277561783697894, "grad_norm": 0.3749232292175293, "learning_rate": 6.0780000000000004e-05, "loss": 0.0711, "step": 3040 }, { "epoch": 0.10311369552723215, "grad_norm": 0.45918208360671997, "learning_rate": 6.098e-05, "loss": 0.0714, "step": 3050 }, { "epoch": 0.10345177321748537, "grad_norm": 0.3409630358219147, "learning_rate": 6.118000000000001e-05, "loss": 0.0772, "step": 3060 }, { "epoch": 0.1037898509077386, "grad_norm": 0.33093300461769104, "learning_rate": 6.138e-05, "loss": 0.0728, "step": 3070 }, { "epoch": 0.10412792859799182, "grad_norm": 0.23260198533535004, "learning_rate": 6.158e-05, "loss": 0.0739, "step": 3080 }, { "epoch": 0.10446600628824504, "grad_norm": 0.3054370582103729, "learning_rate": 6.178000000000001e-05, "loss": 0.0726, "step": 3090 }, { "epoch": 0.10480408397849826, "grad_norm": 0.3102496564388275, "learning_rate": 6.198e-05, "loss": 0.0735, "step": 3100 }, { "epoch": 0.10514216166875148, "grad_norm": 0.26065874099731445, "learning_rate": 6.218e-05, "loss": 0.0674, "step": 3110 }, { "epoch": 0.1054802393590047, "grad_norm": 0.43858322501182556, "learning_rate": 6.238000000000001e-05, "loss": 0.0677, "step": 3120 }, { "epoch": 0.10581831704925793, "grad_norm": 0.37846335768699646, "learning_rate": 6.258e-05, "loss": 0.0791, "step": 3130 }, { "epoch": 0.10615639473951113, "grad_norm": 0.34692928194999695, "learning_rate": 6.278e-05, "loss": 0.0744, "step": 3140 }, { "epoch": 0.10649447242976436, "grad_norm": 0.3406008780002594, "learning_rate": 6.298000000000001e-05, "loss": 0.0751, "step": 3150 }, { "epoch": 0.10683255012001758, "grad_norm": 0.3644615709781647, "learning_rate": 6.318e-05, "loss": 0.0769, "step": 3160 }, { "epoch": 0.1071706278102708, "grad_norm": 0.5038279891014099, "learning_rate": 6.338e-05, "loss": 0.0752, "step": 3170 }, { "epoch": 0.10750870550052402, "grad_norm": 0.26699498295783997, "learning_rate": 6.358000000000001e-05, "loss": 0.0756, "step": 3180 }, { "epoch": 0.10784678319077724, "grad_norm": 0.3993644118309021, "learning_rate": 6.378e-05, "loss": 0.0785, "step": 3190 }, { "epoch": 0.10818486088103046, "grad_norm": 0.3712429702281952, "learning_rate": 6.398000000000001e-05, "loss": 0.077, "step": 3200 }, { "epoch": 0.10852293857128369, "grad_norm": 0.37765830755233765, "learning_rate": 6.418000000000001e-05, "loss": 0.0682, "step": 3210 }, { "epoch": 0.1088610162615369, "grad_norm": 0.6895607113838196, "learning_rate": 6.438e-05, "loss": 0.0744, "step": 3220 }, { "epoch": 0.10919909395179012, "grad_norm": 0.2872347831726074, "learning_rate": 6.458000000000001e-05, "loss": 0.0701, "step": 3230 }, { "epoch": 0.10953717164204334, "grad_norm": 0.4203350245952606, "learning_rate": 6.478000000000001e-05, "loss": 0.0792, "step": 3240 }, { "epoch": 0.10987524933229656, "grad_norm": 0.5929892659187317, "learning_rate": 6.498e-05, "loss": 0.0812, "step": 3250 }, { "epoch": 0.11021332702254978, "grad_norm": 0.45972442626953125, "learning_rate": 6.518000000000001e-05, "loss": 0.0801, "step": 3260 }, { "epoch": 0.110551404712803, "grad_norm": 0.8272041082382202, "learning_rate": 6.538000000000001e-05, "loss": 0.0739, "step": 3270 }, { "epoch": 0.11088948240305622, "grad_norm": 0.547297477722168, "learning_rate": 6.558e-05, "loss": 0.0722, "step": 3280 }, { "epoch": 0.11122756009330945, "grad_norm": 0.40436115860939026, "learning_rate": 6.578000000000001e-05, "loss": 0.0794, "step": 3290 }, { "epoch": 0.11156563778356267, "grad_norm": 0.4837801456451416, "learning_rate": 6.598e-05, "loss": 0.0752, "step": 3300 }, { "epoch": 0.11190371547381588, "grad_norm": 0.3052874505519867, "learning_rate": 6.618e-05, "loss": 0.0669, "step": 3310 }, { "epoch": 0.1122417931640691, "grad_norm": 0.4450998902320862, "learning_rate": 6.638e-05, "loss": 0.0758, "step": 3320 }, { "epoch": 0.11257987085432232, "grad_norm": 0.6632838249206543, "learning_rate": 6.658e-05, "loss": 0.0685, "step": 3330 }, { "epoch": 0.11291794854457554, "grad_norm": 0.38884830474853516, "learning_rate": 6.678e-05, "loss": 0.0794, "step": 3340 }, { "epoch": 0.11325602623482876, "grad_norm": 0.38972580432891846, "learning_rate": 6.698e-05, "loss": 0.0787, "step": 3350 }, { "epoch": 0.11359410392508199, "grad_norm": 0.41326814889907837, "learning_rate": 6.718e-05, "loss": 0.0757, "step": 3360 }, { "epoch": 0.1139321816153352, "grad_norm": 0.5817738175392151, "learning_rate": 6.738e-05, "loss": 0.0775, "step": 3370 }, { "epoch": 0.11427025930558843, "grad_norm": 0.3365069329738617, "learning_rate": 6.758e-05, "loss": 0.0745, "step": 3380 }, { "epoch": 0.11460833699584165, "grad_norm": 0.4304780960083008, "learning_rate": 6.778e-05, "loss": 0.0709, "step": 3390 }, { "epoch": 0.11494641468609486, "grad_norm": 0.32200026512145996, "learning_rate": 6.798e-05, "loss": 0.0714, "step": 3400 }, { "epoch": 0.11528449237634808, "grad_norm": 0.3095634877681732, "learning_rate": 6.818e-05, "loss": 0.0761, "step": 3410 }, { "epoch": 0.1156225700666013, "grad_norm": 0.43450385332107544, "learning_rate": 6.838e-05, "loss": 0.079, "step": 3420 }, { "epoch": 0.11596064775685452, "grad_norm": 0.3670126795768738, "learning_rate": 6.858e-05, "loss": 0.0721, "step": 3430 }, { "epoch": 0.11629872544710775, "grad_norm": 0.32832735776901245, "learning_rate": 6.878e-05, "loss": 0.0745, "step": 3440 }, { "epoch": 0.11663680313736097, "grad_norm": 0.43540292978286743, "learning_rate": 6.898e-05, "loss": 0.0698, "step": 3450 }, { "epoch": 0.11697488082761419, "grad_norm": 0.3082524836063385, "learning_rate": 6.918e-05, "loss": 0.0746, "step": 3460 }, { "epoch": 0.11731295851786741, "grad_norm": 0.3263588547706604, "learning_rate": 6.938e-05, "loss": 0.0704, "step": 3470 }, { "epoch": 0.11765103620812063, "grad_norm": 0.33594200015068054, "learning_rate": 6.958e-05, "loss": 0.0748, "step": 3480 }, { "epoch": 0.11798911389837384, "grad_norm": 0.365037739276886, "learning_rate": 6.978e-05, "loss": 0.0849, "step": 3490 }, { "epoch": 0.11832719158862706, "grad_norm": 0.30669164657592773, "learning_rate": 6.998e-05, "loss": 0.0662, "step": 3500 }, { "epoch": 0.11866526927888028, "grad_norm": 0.3116722106933594, "learning_rate": 7.018e-05, "loss": 0.0723, "step": 3510 }, { "epoch": 0.1190033469691335, "grad_norm": 0.39391177892684937, "learning_rate": 7.038e-05, "loss": 0.0761, "step": 3520 }, { "epoch": 0.11934142465938673, "grad_norm": 0.3317806124687195, "learning_rate": 7.058e-05, "loss": 0.0736, "step": 3530 }, { "epoch": 0.11967950234963995, "grad_norm": 0.32468655705451965, "learning_rate": 7.078e-05, "loss": 0.0689, "step": 3540 }, { "epoch": 0.12001758003989317, "grad_norm": 0.294173002243042, "learning_rate": 7.098e-05, "loss": 0.0682, "step": 3550 }, { "epoch": 0.12035565773014639, "grad_norm": 0.3228987157344818, "learning_rate": 7.118e-05, "loss": 0.0801, "step": 3560 }, { "epoch": 0.12069373542039961, "grad_norm": 0.4147019386291504, "learning_rate": 7.138e-05, "loss": 0.0785, "step": 3570 }, { "epoch": 0.12103181311065282, "grad_norm": 0.3206121623516083, "learning_rate": 7.158e-05, "loss": 0.0785, "step": 3580 }, { "epoch": 0.12136989080090604, "grad_norm": 0.36448371410369873, "learning_rate": 7.178000000000001e-05, "loss": 0.0709, "step": 3590 }, { "epoch": 0.12170796849115927, "grad_norm": 0.4014313220977783, "learning_rate": 7.198e-05, "loss": 0.0713, "step": 3600 }, { "epoch": 0.12204604618141249, "grad_norm": 0.3942115902900696, "learning_rate": 7.218e-05, "loss": 0.0756, "step": 3610 }, { "epoch": 0.12238412387166571, "grad_norm": 0.44879254698753357, "learning_rate": 7.238000000000001e-05, "loss": 0.0692, "step": 3620 }, { "epoch": 0.12272220156191893, "grad_norm": 0.5951998829841614, "learning_rate": 7.258e-05, "loss": 0.0734, "step": 3630 }, { "epoch": 0.12306027925217215, "grad_norm": 0.35129934549331665, "learning_rate": 7.278e-05, "loss": 0.0713, "step": 3640 }, { "epoch": 0.12339835694242537, "grad_norm": 0.3403385281562805, "learning_rate": 7.298000000000001e-05, "loss": 0.075, "step": 3650 }, { "epoch": 0.1237364346326786, "grad_norm": 0.31749650835990906, "learning_rate": 7.318e-05, "loss": 0.076, "step": 3660 }, { "epoch": 0.1240745123229318, "grad_norm": 0.5782597064971924, "learning_rate": 7.338e-05, "loss": 0.0781, "step": 3670 }, { "epoch": 0.12441259001318503, "grad_norm": 0.41831475496292114, "learning_rate": 7.358000000000001e-05, "loss": 0.0734, "step": 3680 }, { "epoch": 0.12475066770343825, "grad_norm": 0.30791422724723816, "learning_rate": 7.378e-05, "loss": 0.0714, "step": 3690 }, { "epoch": 0.12508874539369147, "grad_norm": 0.28813356161117554, "learning_rate": 7.398e-05, "loss": 0.073, "step": 3700 }, { "epoch": 0.12542682308394468, "grad_norm": 0.3316778242588043, "learning_rate": 7.418000000000001e-05, "loss": 0.0771, "step": 3710 }, { "epoch": 0.1257649007741979, "grad_norm": 0.31787824630737305, "learning_rate": 7.438e-05, "loss": 0.0757, "step": 3720 }, { "epoch": 0.12610297846445112, "grad_norm": 0.35258379578590393, "learning_rate": 7.458000000000001e-05, "loss": 0.073, "step": 3730 }, { "epoch": 0.12644105615470436, "grad_norm": 0.43723785877227783, "learning_rate": 7.478e-05, "loss": 0.0779, "step": 3740 }, { "epoch": 0.12677913384495756, "grad_norm": 0.29566067457199097, "learning_rate": 7.498e-05, "loss": 0.0815, "step": 3750 }, { "epoch": 0.1271172115352108, "grad_norm": 0.29778629541397095, "learning_rate": 7.518000000000001e-05, "loss": 0.0779, "step": 3760 }, { "epoch": 0.127455289225464, "grad_norm": 0.3922228515148163, "learning_rate": 7.538e-05, "loss": 0.072, "step": 3770 }, { "epoch": 0.12779336691571724, "grad_norm": 0.3091994524002075, "learning_rate": 7.558e-05, "loss": 0.0698, "step": 3780 }, { "epoch": 0.12813144460597045, "grad_norm": 0.3170596957206726, "learning_rate": 7.578000000000001e-05, "loss": 0.0731, "step": 3790 }, { "epoch": 0.12846952229622366, "grad_norm": 0.40072116255760193, "learning_rate": 7.598e-05, "loss": 0.0742, "step": 3800 }, { "epoch": 0.1288075999864769, "grad_norm": 0.3226798474788666, "learning_rate": 7.618e-05, "loss": 0.0714, "step": 3810 }, { "epoch": 0.1291456776767301, "grad_norm": 0.4607282876968384, "learning_rate": 7.638000000000001e-05, "loss": 0.074, "step": 3820 }, { "epoch": 0.12948375536698334, "grad_norm": 0.3254934251308441, "learning_rate": 7.658e-05, "loss": 0.0731, "step": 3830 }, { "epoch": 0.12982183305723655, "grad_norm": 0.3416624963283539, "learning_rate": 7.678000000000001e-05, "loss": 0.072, "step": 3840 }, { "epoch": 0.13015991074748978, "grad_norm": 0.4433768391609192, "learning_rate": 7.698000000000001e-05, "loss": 0.0734, "step": 3850 }, { "epoch": 0.130497988437743, "grad_norm": 0.29301467537879944, "learning_rate": 7.718e-05, "loss": 0.0774, "step": 3860 }, { "epoch": 0.13083606612799623, "grad_norm": 0.3702963590621948, "learning_rate": 7.738000000000001e-05, "loss": 0.0767, "step": 3870 }, { "epoch": 0.13117414381824943, "grad_norm": 0.3125069737434387, "learning_rate": 7.758000000000001e-05, "loss": 0.0768, "step": 3880 }, { "epoch": 0.13151222150850264, "grad_norm": 0.3570840060710907, "learning_rate": 7.778e-05, "loss": 0.068, "step": 3890 }, { "epoch": 0.13185029919875588, "grad_norm": 0.36525288224220276, "learning_rate": 7.798000000000001e-05, "loss": 0.073, "step": 3900 }, { "epoch": 0.13218837688900908, "grad_norm": 0.30403387546539307, "learning_rate": 7.818000000000001e-05, "loss": 0.0687, "step": 3910 }, { "epoch": 0.13252645457926232, "grad_norm": 0.5276205539703369, "learning_rate": 7.838e-05, "loss": 0.0791, "step": 3920 }, { "epoch": 0.13286453226951553, "grad_norm": 0.45514851808547974, "learning_rate": 7.858000000000001e-05, "loss": 0.0782, "step": 3930 }, { "epoch": 0.13320260995976876, "grad_norm": 0.41816315054893494, "learning_rate": 7.878e-05, "loss": 0.0795, "step": 3940 }, { "epoch": 0.13354068765002197, "grad_norm": 0.6430835723876953, "learning_rate": 7.897999999999999e-05, "loss": 0.0807, "step": 3950 }, { "epoch": 0.1338787653402752, "grad_norm": 0.47484880685806274, "learning_rate": 7.918e-05, "loss": 0.0699, "step": 3960 }, { "epoch": 0.13421684303052842, "grad_norm": 0.44311991333961487, "learning_rate": 7.938e-05, "loss": 0.079, "step": 3970 }, { "epoch": 0.13455492072078162, "grad_norm": 0.5242193937301636, "learning_rate": 7.958e-05, "loss": 0.0823, "step": 3980 }, { "epoch": 0.13489299841103486, "grad_norm": 0.3329663574695587, "learning_rate": 7.978e-05, "loss": 0.0876, "step": 3990 }, { "epoch": 0.13523107610128807, "grad_norm": 0.2657085359096527, "learning_rate": 7.998e-05, "loss": 0.0761, "step": 4000 }, { "epoch": 0.1355691537915413, "grad_norm": 0.2799423336982727, "learning_rate": 8.018e-05, "loss": 0.0793, "step": 4010 }, { "epoch": 0.1359072314817945, "grad_norm": 0.36355796456336975, "learning_rate": 8.038e-05, "loss": 0.0763, "step": 4020 }, { "epoch": 0.13624530917204775, "grad_norm": 0.4824463427066803, "learning_rate": 8.058e-05, "loss": 0.0678, "step": 4030 }, { "epoch": 0.13658338686230095, "grad_norm": 0.24391520023345947, "learning_rate": 8.078e-05, "loss": 0.0806, "step": 4040 }, { "epoch": 0.1369214645525542, "grad_norm": 0.3150074779987335, "learning_rate": 8.098e-05, "loss": 0.0775, "step": 4050 }, { "epoch": 0.1372595422428074, "grad_norm": 0.30168285965919495, "learning_rate": 8.118e-05, "loss": 0.0716, "step": 4060 }, { "epoch": 0.1375976199330606, "grad_norm": 0.2897312343120575, "learning_rate": 8.138e-05, "loss": 0.0673, "step": 4070 }, { "epoch": 0.13793569762331384, "grad_norm": 0.33143433928489685, "learning_rate": 8.158e-05, "loss": 0.07, "step": 4080 }, { "epoch": 0.13827377531356705, "grad_norm": 0.28139081597328186, "learning_rate": 8.178e-05, "loss": 0.0735, "step": 4090 }, { "epoch": 0.13861185300382028, "grad_norm": 0.42792409658432007, "learning_rate": 8.198e-05, "loss": 0.0683, "step": 4100 }, { "epoch": 0.1389499306940735, "grad_norm": 0.36554667353630066, "learning_rate": 8.218e-05, "loss": 0.0718, "step": 4110 }, { "epoch": 0.13928800838432673, "grad_norm": 0.38332927227020264, "learning_rate": 8.238000000000001e-05, "loss": 0.0659, "step": 4120 }, { "epoch": 0.13962608607457994, "grad_norm": 0.600671648979187, "learning_rate": 8.258e-05, "loss": 0.0788, "step": 4130 }, { "epoch": 0.13996416376483317, "grad_norm": 0.31780001521110535, "learning_rate": 8.278e-05, "loss": 0.0774, "step": 4140 }, { "epoch": 0.14030224145508638, "grad_norm": 0.29676365852355957, "learning_rate": 8.298000000000001e-05, "loss": 0.073, "step": 4150 }, { "epoch": 0.1406403191453396, "grad_norm": 0.27564018964767456, "learning_rate": 8.318e-05, "loss": 0.0713, "step": 4160 }, { "epoch": 0.14097839683559282, "grad_norm": 0.2956712543964386, "learning_rate": 8.338e-05, "loss": 0.0654, "step": 4170 }, { "epoch": 0.14131647452584603, "grad_norm": 0.34908249974250793, "learning_rate": 8.358e-05, "loss": 0.0727, "step": 4180 }, { "epoch": 0.14165455221609927, "grad_norm": 0.3421865701675415, "learning_rate": 8.378e-05, "loss": 0.0772, "step": 4190 }, { "epoch": 0.14199262990635247, "grad_norm": 0.3968649208545685, "learning_rate": 8.398e-05, "loss": 0.0829, "step": 4200 }, { "epoch": 0.1423307075966057, "grad_norm": 0.4347580671310425, "learning_rate": 8.418e-05, "loss": 0.0758, "step": 4210 }, { "epoch": 0.14266878528685892, "grad_norm": 0.2766113579273224, "learning_rate": 8.438e-05, "loss": 0.0731, "step": 4220 }, { "epoch": 0.14300686297711215, "grad_norm": 0.32069364190101624, "learning_rate": 8.458e-05, "loss": 0.0651, "step": 4230 }, { "epoch": 0.14334494066736536, "grad_norm": 0.39663228392601013, "learning_rate": 8.478e-05, "loss": 0.0773, "step": 4240 }, { "epoch": 0.14368301835761857, "grad_norm": 0.40389519929885864, "learning_rate": 8.498e-05, "loss": 0.0714, "step": 4250 }, { "epoch": 0.1440210960478718, "grad_norm": 0.20296162366867065, "learning_rate": 8.518000000000001e-05, "loss": 0.068, "step": 4260 }, { "epoch": 0.144359173738125, "grad_norm": 0.3106459677219391, "learning_rate": 8.538e-05, "loss": 0.0722, "step": 4270 }, { "epoch": 0.14469725142837825, "grad_norm": 0.443673700094223, "learning_rate": 8.558e-05, "loss": 0.0732, "step": 4280 }, { "epoch": 0.14503532911863146, "grad_norm": 0.2329217791557312, "learning_rate": 8.578000000000001e-05, "loss": 0.0726, "step": 4290 }, { "epoch": 0.1453734068088847, "grad_norm": 0.33638668060302734, "learning_rate": 8.598e-05, "loss": 0.0762, "step": 4300 }, { "epoch": 0.1457114844991379, "grad_norm": 0.29221877455711365, "learning_rate": 8.618e-05, "loss": 0.0681, "step": 4310 }, { "epoch": 0.14604956218939114, "grad_norm": 0.29248324036598206, "learning_rate": 8.638000000000001e-05, "loss": 0.073, "step": 4320 }, { "epoch": 0.14638763987964434, "grad_norm": 0.3009721636772156, "learning_rate": 8.658e-05, "loss": 0.0714, "step": 4330 }, { "epoch": 0.14672571756989755, "grad_norm": 0.38130199909210205, "learning_rate": 8.678e-05, "loss": 0.0732, "step": 4340 }, { "epoch": 0.1470637952601508, "grad_norm": 0.3862585425376892, "learning_rate": 8.698000000000001e-05, "loss": 0.0709, "step": 4350 }, { "epoch": 0.147401872950404, "grad_norm": 0.2681243419647217, "learning_rate": 8.718e-05, "loss": 0.0751, "step": 4360 }, { "epoch": 0.14773995064065723, "grad_norm": 0.5259104371070862, "learning_rate": 8.738000000000001e-05, "loss": 0.0747, "step": 4370 }, { "epoch": 0.14807802833091044, "grad_norm": 0.2444477379322052, "learning_rate": 8.758000000000001e-05, "loss": 0.074, "step": 4380 }, { "epoch": 0.14841610602116367, "grad_norm": 0.2894861102104187, "learning_rate": 8.778e-05, "loss": 0.0686, "step": 4390 }, { "epoch": 0.14875418371141688, "grad_norm": 0.28520745038986206, "learning_rate": 8.798000000000001e-05, "loss": 0.0703, "step": 4400 }, { "epoch": 0.14909226140167012, "grad_norm": 0.4411102831363678, "learning_rate": 8.818000000000001e-05, "loss": 0.0769, "step": 4410 }, { "epoch": 0.14943033909192333, "grad_norm": 0.2692822813987732, "learning_rate": 8.838e-05, "loss": 0.0722, "step": 4420 }, { "epoch": 0.14976841678217653, "grad_norm": 0.30478665232658386, "learning_rate": 8.858000000000001e-05, "loss": 0.0761, "step": 4430 }, { "epoch": 0.15010649447242977, "grad_norm": 0.3496260344982147, "learning_rate": 8.878000000000001e-05, "loss": 0.0739, "step": 4440 }, { "epoch": 0.15044457216268298, "grad_norm": 0.44935792684555054, "learning_rate": 8.898e-05, "loss": 0.0764, "step": 4450 }, { "epoch": 0.1507826498529362, "grad_norm": 0.5045684576034546, "learning_rate": 8.918000000000001e-05, "loss": 0.0651, "step": 4460 }, { "epoch": 0.15112072754318942, "grad_norm": 0.33032628893852234, "learning_rate": 8.938e-05, "loss": 0.07, "step": 4470 }, { "epoch": 0.15145880523344266, "grad_norm": 0.2828603982925415, "learning_rate": 8.958e-05, "loss": 0.0761, "step": 4480 }, { "epoch": 0.15179688292369586, "grad_norm": 0.29891130328178406, "learning_rate": 8.978000000000001e-05, "loss": 0.0742, "step": 4490 }, { "epoch": 0.15213496061394907, "grad_norm": 0.31146350502967834, "learning_rate": 8.998e-05, "loss": 0.0753, "step": 4500 }, { "epoch": 0.1524730383042023, "grad_norm": 0.3337556719779968, "learning_rate": 9.018000000000001e-05, "loss": 0.075, "step": 4510 }, { "epoch": 0.15281111599445552, "grad_norm": 0.41367578506469727, "learning_rate": 9.038000000000001e-05, "loss": 0.0754, "step": 4520 }, { "epoch": 0.15314919368470875, "grad_norm": 0.30758926272392273, "learning_rate": 9.058e-05, "loss": 0.0733, "step": 4530 }, { "epoch": 0.15348727137496196, "grad_norm": 0.3515576720237732, "learning_rate": 9.078000000000001e-05, "loss": 0.073, "step": 4540 }, { "epoch": 0.1538253490652152, "grad_norm": 0.2731901705265045, "learning_rate": 9.098000000000001e-05, "loss": 0.0685, "step": 4550 }, { "epoch": 0.1541634267554684, "grad_norm": 0.38069191575050354, "learning_rate": 9.118e-05, "loss": 0.0826, "step": 4560 }, { "epoch": 0.15450150444572164, "grad_norm": 0.4795483946800232, "learning_rate": 9.138e-05, "loss": 0.0699, "step": 4570 }, { "epoch": 0.15483958213597485, "grad_norm": 0.3426046073436737, "learning_rate": 9.158e-05, "loss": 0.0786, "step": 4580 }, { "epoch": 0.15517765982622805, "grad_norm": 0.3715868592262268, "learning_rate": 9.178e-05, "loss": 0.0687, "step": 4590 }, { "epoch": 0.1555157375164813, "grad_norm": 0.3310057818889618, "learning_rate": 9.198e-05, "loss": 0.0791, "step": 4600 }, { "epoch": 0.1558538152067345, "grad_norm": 0.4246353805065155, "learning_rate": 9.218e-05, "loss": 0.0869, "step": 4610 }, { "epoch": 0.15619189289698773, "grad_norm": 0.40421831607818604, "learning_rate": 9.238e-05, "loss": 0.0721, "step": 4620 }, { "epoch": 0.15652997058724094, "grad_norm": 0.48379719257354736, "learning_rate": 9.258e-05, "loss": 0.0689, "step": 4630 }, { "epoch": 0.15686804827749418, "grad_norm": 0.3584086000919342, "learning_rate": 9.278e-05, "loss": 0.0734, "step": 4640 }, { "epoch": 0.15720612596774738, "grad_norm": 0.3032270073890686, "learning_rate": 9.298e-05, "loss": 0.0806, "step": 4650 }, { "epoch": 0.15754420365800062, "grad_norm": 0.26300492882728577, "learning_rate": 9.318e-05, "loss": 0.0696, "step": 4660 }, { "epoch": 0.15788228134825383, "grad_norm": 0.3209935426712036, "learning_rate": 9.338e-05, "loss": 0.0638, "step": 4670 }, { "epoch": 0.15822035903850704, "grad_norm": 0.399688184261322, "learning_rate": 9.358e-05, "loss": 0.0656, "step": 4680 }, { "epoch": 0.15855843672876027, "grad_norm": 0.4112197458744049, "learning_rate": 9.378e-05, "loss": 0.0782, "step": 4690 }, { "epoch": 0.15889651441901348, "grad_norm": 0.42907506227493286, "learning_rate": 9.398e-05, "loss": 0.0728, "step": 4700 }, { "epoch": 0.15923459210926671, "grad_norm": 0.5378241539001465, "learning_rate": 9.418e-05, "loss": 0.0825, "step": 4710 }, { "epoch": 0.15957266979951992, "grad_norm": 0.27174854278564453, "learning_rate": 9.438e-05, "loss": 0.0737, "step": 4720 }, { "epoch": 0.15991074748977316, "grad_norm": 0.3370060920715332, "learning_rate": 9.458e-05, "loss": 0.0738, "step": 4730 }, { "epoch": 0.16024882518002637, "grad_norm": 0.3107588589191437, "learning_rate": 9.478e-05, "loss": 0.0641, "step": 4740 }, { "epoch": 0.1605869028702796, "grad_norm": 0.4257258176803589, "learning_rate": 9.498e-05, "loss": 0.0652, "step": 4750 }, { "epoch": 0.1609249805605328, "grad_norm": 0.27592888474464417, "learning_rate": 9.518000000000001e-05, "loss": 0.0755, "step": 4760 }, { "epoch": 0.16126305825078602, "grad_norm": 0.2824955880641937, "learning_rate": 9.538e-05, "loss": 0.0706, "step": 4770 }, { "epoch": 0.16160113594103925, "grad_norm": 0.37835875153541565, "learning_rate": 9.558e-05, "loss": 0.0757, "step": 4780 }, { "epoch": 0.16193921363129246, "grad_norm": 0.2965067923069, "learning_rate": 9.578000000000001e-05, "loss": 0.0746, "step": 4790 }, { "epoch": 0.1622772913215457, "grad_norm": 0.2533070743083954, "learning_rate": 9.598e-05, "loss": 0.0753, "step": 4800 }, { "epoch": 0.1626153690117989, "grad_norm": 0.41766658425331116, "learning_rate": 9.618e-05, "loss": 0.0722, "step": 4810 }, { "epoch": 0.16295344670205214, "grad_norm": 0.2553923428058624, "learning_rate": 9.638000000000001e-05, "loss": 0.074, "step": 4820 }, { "epoch": 0.16329152439230535, "grad_norm": 0.3421885073184967, "learning_rate": 9.658e-05, "loss": 0.0733, "step": 4830 }, { "epoch": 0.16362960208255858, "grad_norm": 0.38999953866004944, "learning_rate": 9.678e-05, "loss": 0.0767, "step": 4840 }, { "epoch": 0.1639676797728118, "grad_norm": 0.2755145728588104, "learning_rate": 9.698000000000001e-05, "loss": 0.0776, "step": 4850 }, { "epoch": 0.164305757463065, "grad_norm": 0.337912380695343, "learning_rate": 9.718e-05, "loss": 0.0743, "step": 4860 }, { "epoch": 0.16464383515331824, "grad_norm": 0.2793712615966797, "learning_rate": 9.738e-05, "loss": 0.0751, "step": 4870 }, { "epoch": 0.16498191284357144, "grad_norm": 0.27084192633628845, "learning_rate": 9.758000000000001e-05, "loss": 0.0735, "step": 4880 }, { "epoch": 0.16531999053382468, "grad_norm": 0.37362033128738403, "learning_rate": 9.778e-05, "loss": 0.0821, "step": 4890 }, { "epoch": 0.1656580682240779, "grad_norm": 0.24072812497615814, "learning_rate": 9.798000000000001e-05, "loss": 0.0768, "step": 4900 }, { "epoch": 0.16599614591433112, "grad_norm": 0.32531240582466125, "learning_rate": 9.818000000000001e-05, "loss": 0.0737, "step": 4910 }, { "epoch": 0.16633422360458433, "grad_norm": 0.33840736746788025, "learning_rate": 9.838e-05, "loss": 0.0725, "step": 4920 }, { "epoch": 0.16667230129483757, "grad_norm": 0.4578930735588074, "learning_rate": 9.858000000000001e-05, "loss": 0.0774, "step": 4930 }, { "epoch": 0.16701037898509077, "grad_norm": 0.3930746018886566, "learning_rate": 9.878e-05, "loss": 0.0712, "step": 4940 }, { "epoch": 0.16734845667534398, "grad_norm": 0.34766003489494324, "learning_rate": 9.898e-05, "loss": 0.0802, "step": 4950 }, { "epoch": 0.16768653436559722, "grad_norm": 0.30499187111854553, "learning_rate": 9.918000000000001e-05, "loss": 0.0767, "step": 4960 }, { "epoch": 0.16802461205585043, "grad_norm": 0.36769378185272217, "learning_rate": 9.938e-05, "loss": 0.0667, "step": 4970 }, { "epoch": 0.16836268974610366, "grad_norm": 0.3651949465274811, "learning_rate": 9.958e-05, "loss": 0.0754, "step": 4980 }, { "epoch": 0.16870076743635687, "grad_norm": 0.5768172740936279, "learning_rate": 9.978000000000001e-05, "loss": 0.0789, "step": 4990 }, { "epoch": 0.1690388451266101, "grad_norm": 0.278088241815567, "learning_rate": 9.998e-05, "loss": 0.0796, "step": 5000 }, { "epoch": 0.1693769228168633, "grad_norm": 0.33908846974372864, "learning_rate": 9.999999778549045e-05, "loss": 0.0717, "step": 5010 }, { "epoch": 0.16971500050711655, "grad_norm": 0.39626362919807434, "learning_rate": 9.999999013039593e-05, "loss": 0.074, "step": 5020 }, { "epoch": 0.17005307819736976, "grad_norm": 0.337375283241272, "learning_rate": 9.999997700737766e-05, "loss": 0.0754, "step": 5030 }, { "epoch": 0.17039115588762296, "grad_norm": 0.42100855708122253, "learning_rate": 9.999995841643709e-05, "loss": 0.0712, "step": 5040 }, { "epoch": 0.1707292335778762, "grad_norm": 0.2976936995983124, "learning_rate": 9.999993435757623e-05, "loss": 0.0692, "step": 5050 }, { "epoch": 0.1710673112681294, "grad_norm": 0.3678324818611145, "learning_rate": 9.999990483079773e-05, "loss": 0.071, "step": 5060 }, { "epoch": 0.17140538895838264, "grad_norm": 0.2941516041755676, "learning_rate": 9.999986983610481e-05, "loss": 0.0718, "step": 5070 }, { "epoch": 0.17174346664863585, "grad_norm": 0.43334805965423584, "learning_rate": 9.99998293735013e-05, "loss": 0.0733, "step": 5080 }, { "epoch": 0.1720815443388891, "grad_norm": 0.3685532510280609, "learning_rate": 9.999978344299161e-05, "loss": 0.0727, "step": 5090 }, { "epoch": 0.1724196220291423, "grad_norm": 0.32830914855003357, "learning_rate": 9.99997320445808e-05, "loss": 0.0701, "step": 5100 }, { "epoch": 0.17275769971939553, "grad_norm": 0.3395450711250305, "learning_rate": 9.999967517827444e-05, "loss": 0.0734, "step": 5110 }, { "epoch": 0.17309577740964874, "grad_norm": 0.5821679830551147, "learning_rate": 9.999961284407879e-05, "loss": 0.0744, "step": 5120 }, { "epoch": 0.17343385509990195, "grad_norm": 0.2886611819267273, "learning_rate": 9.999954504200067e-05, "loss": 0.0712, "step": 5130 }, { "epoch": 0.17377193279015518, "grad_norm": 0.3980024755001068, "learning_rate": 9.999947177204744e-05, "loss": 0.0706, "step": 5140 }, { "epoch": 0.1741100104804084, "grad_norm": 0.2367020547389984, "learning_rate": 9.999939303422718e-05, "loss": 0.0726, "step": 5150 }, { "epoch": 0.17444808817066162, "grad_norm": 0.46387195587158203, "learning_rate": 9.999930882854847e-05, "loss": 0.0716, "step": 5160 }, { "epoch": 0.17478616586091483, "grad_norm": 0.3762591481208801, "learning_rate": 9.999921915502051e-05, "loss": 0.0671, "step": 5170 }, { "epoch": 0.17512424355116807, "grad_norm": 0.258821964263916, "learning_rate": 9.99991240136531e-05, "loss": 0.0696, "step": 5180 }, { "epoch": 0.17546232124142128, "grad_norm": 0.2499101310968399, "learning_rate": 9.999902340445668e-05, "loss": 0.0744, "step": 5190 }, { "epoch": 0.1758003989316745, "grad_norm": 0.32544031739234924, "learning_rate": 9.999891732744224e-05, "loss": 0.0746, "step": 5200 }, { "epoch": 0.17613847662192772, "grad_norm": 0.3041284382343292, "learning_rate": 9.999880578262135e-05, "loss": 0.0778, "step": 5210 }, { "epoch": 0.17647655431218093, "grad_norm": 0.3040120601654053, "learning_rate": 9.999868877000624e-05, "loss": 0.0746, "step": 5220 }, { "epoch": 0.17681463200243416, "grad_norm": 0.3857915997505188, "learning_rate": 9.99985662896097e-05, "loss": 0.0713, "step": 5230 }, { "epoch": 0.17715270969268737, "grad_norm": 0.5731173157691956, "learning_rate": 9.999843834144513e-05, "loss": 0.0772, "step": 5240 }, { "epoch": 0.1774907873829406, "grad_norm": 0.279183566570282, "learning_rate": 9.99983049255265e-05, "loss": 0.0816, "step": 5250 }, { "epoch": 0.17782886507319381, "grad_norm": 0.26652610301971436, "learning_rate": 9.999816604186843e-05, "loss": 0.0733, "step": 5260 }, { "epoch": 0.17816694276344705, "grad_norm": 0.40532392263412476, "learning_rate": 9.999802169048609e-05, "loss": 0.0687, "step": 5270 }, { "epoch": 0.17850502045370026, "grad_norm": 0.28691184520721436, "learning_rate": 9.999787187139527e-05, "loss": 0.0688, "step": 5280 }, { "epoch": 0.1788430981439535, "grad_norm": 0.3130979835987091, "learning_rate": 9.999771658461234e-05, "loss": 0.0768, "step": 5290 }, { "epoch": 0.1791811758342067, "grad_norm": 0.3436627686023712, "learning_rate": 9.999755583015431e-05, "loss": 0.0714, "step": 5300 }, { "epoch": 0.1795192535244599, "grad_norm": 0.3477271795272827, "learning_rate": 9.999738960803874e-05, "loss": 0.0719, "step": 5310 }, { "epoch": 0.17985733121471315, "grad_norm": 0.4096946716308594, "learning_rate": 9.99972179182838e-05, "loss": 0.0732, "step": 5320 }, { "epoch": 0.18019540890496635, "grad_norm": 0.27843230962753296, "learning_rate": 9.99970407609083e-05, "loss": 0.078, "step": 5330 }, { "epoch": 0.1805334865952196, "grad_norm": 0.22837239503860474, "learning_rate": 9.999685813593159e-05, "loss": 0.0824, "step": 5340 }, { "epoch": 0.1808715642854728, "grad_norm": 0.25569117069244385, "learning_rate": 9.999667004337362e-05, "loss": 0.071, "step": 5350 }, { "epoch": 0.18120964197572603, "grad_norm": 0.24395819008350372, "learning_rate": 9.9996476483255e-05, "loss": 0.0703, "step": 5360 }, { "epoch": 0.18154771966597924, "grad_norm": 0.31748825311660767, "learning_rate": 9.999627745559688e-05, "loss": 0.0734, "step": 5370 }, { "epoch": 0.18188579735623248, "grad_norm": 0.24732328951358795, "learning_rate": 9.999607296042101e-05, "loss": 0.0698, "step": 5380 }, { "epoch": 0.18222387504648568, "grad_norm": 0.3803701400756836, "learning_rate": 9.99958629977498e-05, "loss": 0.0648, "step": 5390 }, { "epoch": 0.1825619527367389, "grad_norm": 0.37889108061790466, "learning_rate": 9.999564756760615e-05, "loss": 0.0764, "step": 5400 }, { "epoch": 0.18290003042699213, "grad_norm": 0.44724246859550476, "learning_rate": 9.999542667001366e-05, "loss": 0.0792, "step": 5410 }, { "epoch": 0.18323810811724534, "grad_norm": 0.25769853591918945, "learning_rate": 9.999520030499647e-05, "loss": 0.0636, "step": 5420 }, { "epoch": 0.18357618580749857, "grad_norm": 0.36541786789894104, "learning_rate": 9.999496847257936e-05, "loss": 0.072, "step": 5430 }, { "epoch": 0.18391426349775178, "grad_norm": 0.29828113317489624, "learning_rate": 9.999473117278764e-05, "loss": 0.0699, "step": 5440 }, { "epoch": 0.18425234118800501, "grad_norm": 0.3085501790046692, "learning_rate": 9.999448840564731e-05, "loss": 0.0742, "step": 5450 }, { "epoch": 0.18459041887825822, "grad_norm": 0.369045615196228, "learning_rate": 9.999424017118488e-05, "loss": 0.0733, "step": 5460 }, { "epoch": 0.18492849656851143, "grad_norm": 0.30274221301078796, "learning_rate": 9.999398646942751e-05, "loss": 0.0691, "step": 5470 }, { "epoch": 0.18526657425876467, "grad_norm": 0.2965167462825775, "learning_rate": 9.999372730040296e-05, "loss": 0.0692, "step": 5480 }, { "epoch": 0.18560465194901787, "grad_norm": 0.28566762804985046, "learning_rate": 9.999346266413953e-05, "loss": 0.0761, "step": 5490 }, { "epoch": 0.1859427296392711, "grad_norm": 0.2685447633266449, "learning_rate": 9.99931925606662e-05, "loss": 0.0761, "step": 5500 }, { "epoch": 0.18628080732952432, "grad_norm": 0.4200141131877899, "learning_rate": 9.99929169900125e-05, "loss": 0.0669, "step": 5510 }, { "epoch": 0.18661888501977755, "grad_norm": 0.2768501937389374, "learning_rate": 9.999263595220855e-05, "loss": 0.0701, "step": 5520 }, { "epoch": 0.18695696271003076, "grad_norm": 0.28311896324157715, "learning_rate": 9.99923494472851e-05, "loss": 0.0708, "step": 5530 }, { "epoch": 0.187295040400284, "grad_norm": 0.4469127655029297, "learning_rate": 9.999205747527348e-05, "loss": 0.0666, "step": 5540 }, { "epoch": 0.1876331180905372, "grad_norm": 0.22945570945739746, "learning_rate": 9.999176003620561e-05, "loss": 0.073, "step": 5550 }, { "epoch": 0.1879711957807904, "grad_norm": 0.37366223335266113, "learning_rate": 9.999145713011405e-05, "loss": 0.0742, "step": 5560 }, { "epoch": 0.18830927347104365, "grad_norm": 0.3038610517978668, "learning_rate": 9.999114875703186e-05, "loss": 0.0778, "step": 5570 }, { "epoch": 0.18864735116129686, "grad_norm": 0.3108358383178711, "learning_rate": 9.999083491699281e-05, "loss": 0.0712, "step": 5580 }, { "epoch": 0.1889854288515501, "grad_norm": 0.41795578598976135, "learning_rate": 9.999051561003123e-05, "loss": 0.0668, "step": 5590 }, { "epoch": 0.1893235065418033, "grad_norm": 0.4996907114982605, "learning_rate": 9.999019083618202e-05, "loss": 0.0773, "step": 5600 }, { "epoch": 0.18966158423205653, "grad_norm": 0.2887808382511139, "learning_rate": 9.99898605954807e-05, "loss": 0.0729, "step": 5610 }, { "epoch": 0.18999966192230974, "grad_norm": 0.45248347520828247, "learning_rate": 9.998952488796338e-05, "loss": 0.0696, "step": 5620 }, { "epoch": 0.19033773961256298, "grad_norm": 0.3473447263240814, "learning_rate": 9.998918371366676e-05, "loss": 0.0774, "step": 5630 }, { "epoch": 0.1906758173028162, "grad_norm": 0.24485202133655548, "learning_rate": 9.99888370726282e-05, "loss": 0.0708, "step": 5640 }, { "epoch": 0.1910138949930694, "grad_norm": 0.33855026960372925, "learning_rate": 9.998848496488556e-05, "loss": 0.0697, "step": 5650 }, { "epoch": 0.19135197268332263, "grad_norm": 0.2769596576690674, "learning_rate": 9.998812739047736e-05, "loss": 0.0754, "step": 5660 }, { "epoch": 0.19169005037357584, "grad_norm": 0.3779509365558624, "learning_rate": 9.99877643494427e-05, "loss": 0.0767, "step": 5670 }, { "epoch": 0.19202812806382907, "grad_norm": 0.2492722123861313, "learning_rate": 9.998739584182128e-05, "loss": 0.0787, "step": 5680 }, { "epoch": 0.19236620575408228, "grad_norm": 0.27111199498176575, "learning_rate": 9.998702186765342e-05, "loss": 0.0782, "step": 5690 }, { "epoch": 0.19270428344433552, "grad_norm": 0.2731785178184509, "learning_rate": 9.998664242698e-05, "loss": 0.0685, "step": 5700 }, { "epoch": 0.19304236113458872, "grad_norm": 0.30142447352409363, "learning_rate": 9.998625751984251e-05, "loss": 0.0703, "step": 5710 }, { "epoch": 0.19338043882484196, "grad_norm": 0.2822786867618561, "learning_rate": 9.998586714628307e-05, "loss": 0.0744, "step": 5720 }, { "epoch": 0.19371851651509517, "grad_norm": 0.2764637768268585, "learning_rate": 9.998547130634432e-05, "loss": 0.0755, "step": 5730 }, { "epoch": 0.19405659420534838, "grad_norm": 0.5768522024154663, "learning_rate": 9.99850700000696e-05, "loss": 0.0715, "step": 5740 }, { "epoch": 0.1943946718956016, "grad_norm": 0.32532280683517456, "learning_rate": 9.998466322750278e-05, "loss": 0.0738, "step": 5750 }, { "epoch": 0.19473274958585482, "grad_norm": 0.3604487180709839, "learning_rate": 9.998425098868834e-05, "loss": 0.0765, "step": 5760 }, { "epoch": 0.19507082727610806, "grad_norm": 0.3616235852241516, "learning_rate": 9.998383328367136e-05, "loss": 0.0736, "step": 5770 }, { "epoch": 0.19540890496636126, "grad_norm": 0.3065974712371826, "learning_rate": 9.99834101124975e-05, "loss": 0.0682, "step": 5780 }, { "epoch": 0.1957469826566145, "grad_norm": 0.2839737832546234, "learning_rate": 9.998298147521309e-05, "loss": 0.072, "step": 5790 }, { "epoch": 0.1960850603468677, "grad_norm": 0.3466457724571228, "learning_rate": 9.998254737186496e-05, "loss": 0.0675, "step": 5800 }, { "epoch": 0.19642313803712094, "grad_norm": 0.27706313133239746, "learning_rate": 9.99821078025006e-05, "loss": 0.0731, "step": 5810 }, { "epoch": 0.19676121572737415, "grad_norm": 0.36418336629867554, "learning_rate": 9.998166276716807e-05, "loss": 0.0756, "step": 5820 }, { "epoch": 0.19709929341762736, "grad_norm": 0.3604094386100769, "learning_rate": 9.998121226591606e-05, "loss": 0.0738, "step": 5830 }, { "epoch": 0.1974373711078806, "grad_norm": 0.19388002157211304, "learning_rate": 9.998075629879382e-05, "loss": 0.0651, "step": 5840 }, { "epoch": 0.1977754487981338, "grad_norm": 0.24212785065174103, "learning_rate": 9.99802948658512e-05, "loss": 0.0746, "step": 5850 }, { "epoch": 0.19811352648838704, "grad_norm": 0.269398033618927, "learning_rate": 9.99798279671387e-05, "loss": 0.0729, "step": 5860 }, { "epoch": 0.19845160417864025, "grad_norm": 0.41089001297950745, "learning_rate": 9.997935560270734e-05, "loss": 0.0695, "step": 5870 }, { "epoch": 0.19878968186889348, "grad_norm": 0.31110262870788574, "learning_rate": 9.997887777260879e-05, "loss": 0.0718, "step": 5880 }, { "epoch": 0.1991277595591467, "grad_norm": 0.31255748867988586, "learning_rate": 9.997839447689532e-05, "loss": 0.069, "step": 5890 }, { "epoch": 0.19946583724939992, "grad_norm": 0.3748161494731903, "learning_rate": 9.997790571561978e-05, "loss": 0.069, "step": 5900 }, { "epoch": 0.19980391493965313, "grad_norm": 0.5098368525505066, "learning_rate": 9.99774114888356e-05, "loss": 0.0759, "step": 5910 }, { "epoch": 0.20014199262990634, "grad_norm": 0.2687198519706726, "learning_rate": 9.997691179659684e-05, "loss": 0.0674, "step": 5920 }, { "epoch": 0.20048007032015958, "grad_norm": 0.3475379943847656, "learning_rate": 9.997640663895815e-05, "loss": 0.0696, "step": 5930 }, { "epoch": 0.20081814801041278, "grad_norm": 0.3172144889831543, "learning_rate": 9.997589601597477e-05, "loss": 0.0742, "step": 5940 }, { "epoch": 0.20115622570066602, "grad_norm": 0.2738882005214691, "learning_rate": 9.997537992770252e-05, "loss": 0.0764, "step": 5950 }, { "epoch": 0.20149430339091923, "grad_norm": 0.2271127551794052, "learning_rate": 9.997485837419788e-05, "loss": 0.0683, "step": 5960 }, { "epoch": 0.20183238108117246, "grad_norm": 0.3551950454711914, "learning_rate": 9.997433135551786e-05, "loss": 0.0664, "step": 5970 }, { "epoch": 0.20217045877142567, "grad_norm": 0.2643931806087494, "learning_rate": 9.997379887172009e-05, "loss": 0.076, "step": 5980 }, { "epoch": 0.2025085364616789, "grad_norm": 0.3288592994213104, "learning_rate": 9.997326092286281e-05, "loss": 0.0698, "step": 5990 }, { "epoch": 0.20284661415193211, "grad_norm": 0.3992771804332733, "learning_rate": 9.997271750900486e-05, "loss": 0.0658, "step": 6000 }, { "epoch": 0.20318469184218532, "grad_norm": 0.25612160563468933, "learning_rate": 9.997216863020565e-05, "loss": 0.0732, "step": 6010 }, { "epoch": 0.20352276953243856, "grad_norm": 0.27502644062042236, "learning_rate": 9.99716142865252e-05, "loss": 0.0738, "step": 6020 }, { "epoch": 0.20386084722269177, "grad_norm": 0.34494638442993164, "learning_rate": 9.997105447802415e-05, "loss": 0.0713, "step": 6030 }, { "epoch": 0.204198924912945, "grad_norm": 0.24398721754550934, "learning_rate": 9.997048920476373e-05, "loss": 0.0697, "step": 6040 }, { "epoch": 0.2045370026031982, "grad_norm": 0.23828962445259094, "learning_rate": 9.996991846680572e-05, "loss": 0.0741, "step": 6050 }, { "epoch": 0.20487508029345144, "grad_norm": 0.286625474691391, "learning_rate": 9.996934226421257e-05, "loss": 0.0747, "step": 6060 }, { "epoch": 0.20521315798370465, "grad_norm": 0.26749807596206665, "learning_rate": 9.996876059704726e-05, "loss": 0.0731, "step": 6070 }, { "epoch": 0.2055512356739579, "grad_norm": 0.27017098665237427, "learning_rate": 9.996817346537343e-05, "loss": 0.0798, "step": 6080 }, { "epoch": 0.2058893133642111, "grad_norm": 0.22116237878799438, "learning_rate": 9.996758086925526e-05, "loss": 0.0747, "step": 6090 }, { "epoch": 0.2062273910544643, "grad_norm": 0.29707708954811096, "learning_rate": 9.996698280875759e-05, "loss": 0.0701, "step": 6100 }, { "epoch": 0.20656546874471754, "grad_norm": 0.2895086705684662, "learning_rate": 9.99663792839458e-05, "loss": 0.0737, "step": 6110 }, { "epoch": 0.20690354643497075, "grad_norm": 0.25795572996139526, "learning_rate": 9.99657702948859e-05, "loss": 0.0727, "step": 6120 }, { "epoch": 0.20724162412522398, "grad_norm": 0.2968361973762512, "learning_rate": 9.996515584164448e-05, "loss": 0.0693, "step": 6130 }, { "epoch": 0.2075797018154772, "grad_norm": 0.21267658472061157, "learning_rate": 9.996453592428873e-05, "loss": 0.0683, "step": 6140 }, { "epoch": 0.20791777950573043, "grad_norm": 0.36477407813072205, "learning_rate": 9.996391054288646e-05, "loss": 0.0673, "step": 6150 }, { "epoch": 0.20825585719598363, "grad_norm": 0.2890692353248596, "learning_rate": 9.996327969750605e-05, "loss": 0.0772, "step": 6160 }, { "epoch": 0.20859393488623687, "grad_norm": 0.2331053465604782, "learning_rate": 9.996264338821649e-05, "loss": 0.0733, "step": 6170 }, { "epoch": 0.20893201257649008, "grad_norm": 0.28383898735046387, "learning_rate": 9.996200161508735e-05, "loss": 0.0705, "step": 6180 }, { "epoch": 0.20927009026674329, "grad_norm": 0.3739914894104004, "learning_rate": 9.996135437818885e-05, "loss": 0.0713, "step": 6190 }, { "epoch": 0.20960816795699652, "grad_norm": 0.3262585401535034, "learning_rate": 9.996070167759175e-05, "loss": 0.0804, "step": 6200 }, { "epoch": 0.20994624564724973, "grad_norm": 0.26181405782699585, "learning_rate": 9.996004351336743e-05, "loss": 0.0697, "step": 6210 }, { "epoch": 0.21028432333750297, "grad_norm": 0.24734626710414886, "learning_rate": 9.995937988558785e-05, "loss": 0.074, "step": 6220 }, { "epoch": 0.21062240102775617, "grad_norm": 0.263501912355423, "learning_rate": 9.995871079432561e-05, "loss": 0.0707, "step": 6230 }, { "epoch": 0.2109604787180094, "grad_norm": 0.2888759970664978, "learning_rate": 9.995803623965389e-05, "loss": 0.0743, "step": 6240 }, { "epoch": 0.21129855640826262, "grad_norm": 0.29325467348098755, "learning_rate": 9.995735622164641e-05, "loss": 0.0737, "step": 6250 }, { "epoch": 0.21163663409851585, "grad_norm": 0.22984938323497772, "learning_rate": 9.995667074037758e-05, "loss": 0.0717, "step": 6260 }, { "epoch": 0.21197471178876906, "grad_norm": 0.2728085219860077, "learning_rate": 9.995597979592232e-05, "loss": 0.0711, "step": 6270 }, { "epoch": 0.21231278947902227, "grad_norm": 0.2926950454711914, "learning_rate": 9.995528338835625e-05, "loss": 0.0636, "step": 6280 }, { "epoch": 0.2126508671692755, "grad_norm": 0.2808513939380646, "learning_rate": 9.995458151775547e-05, "loss": 0.0707, "step": 6290 }, { "epoch": 0.2129889448595287, "grad_norm": 0.35583028197288513, "learning_rate": 9.995387418419677e-05, "loss": 0.0703, "step": 6300 }, { "epoch": 0.21332702254978195, "grad_norm": 0.38055524230003357, "learning_rate": 9.99531613877575e-05, "loss": 0.0761, "step": 6310 }, { "epoch": 0.21366510024003516, "grad_norm": 0.3732640743255615, "learning_rate": 9.995244312851559e-05, "loss": 0.0733, "step": 6320 }, { "epoch": 0.2140031779302884, "grad_norm": 0.37943267822265625, "learning_rate": 9.995171940654961e-05, "loss": 0.075, "step": 6330 }, { "epoch": 0.2143412556205416, "grad_norm": 0.469853937625885, "learning_rate": 9.995099022193871e-05, "loss": 0.0741, "step": 6340 }, { "epoch": 0.21467933331079483, "grad_norm": 0.26190176606178284, "learning_rate": 9.995025557476261e-05, "loss": 0.0727, "step": 6350 }, { "epoch": 0.21501741100104804, "grad_norm": 0.33285030722618103, "learning_rate": 9.994951546510165e-05, "loss": 0.0732, "step": 6360 }, { "epoch": 0.21535548869130125, "grad_norm": 0.2859611213207245, "learning_rate": 9.994876989303679e-05, "loss": 0.0718, "step": 6370 }, { "epoch": 0.21569356638155449, "grad_norm": 0.259711891412735, "learning_rate": 9.994801885864955e-05, "loss": 0.0668, "step": 6380 }, { "epoch": 0.2160316440718077, "grad_norm": 0.31474071741104126, "learning_rate": 9.994726236202205e-05, "loss": 0.0743, "step": 6390 }, { "epoch": 0.21636972176206093, "grad_norm": 0.22404947876930237, "learning_rate": 9.994650040323704e-05, "loss": 0.0728, "step": 6400 }, { "epoch": 0.21670779945231414, "grad_norm": 0.29106903076171875, "learning_rate": 9.994573298237784e-05, "loss": 0.0706, "step": 6410 }, { "epoch": 0.21704587714256737, "grad_norm": 0.24402545392513275, "learning_rate": 9.994496009952837e-05, "loss": 0.0669, "step": 6420 }, { "epoch": 0.21738395483282058, "grad_norm": 0.26126179099082947, "learning_rate": 9.994418175477316e-05, "loss": 0.0634, "step": 6430 }, { "epoch": 0.2177220325230738, "grad_norm": 0.4989783763885498, "learning_rate": 9.994339794819733e-05, "loss": 0.0659, "step": 6440 }, { "epoch": 0.21806011021332702, "grad_norm": 0.2632708251476288, "learning_rate": 9.994260867988658e-05, "loss": 0.072, "step": 6450 }, { "epoch": 0.21839818790358023, "grad_norm": 0.27124232053756714, "learning_rate": 9.994181394992723e-05, "loss": 0.0697, "step": 6460 }, { "epoch": 0.21873626559383347, "grad_norm": 0.3242019712924957, "learning_rate": 9.994101375840618e-05, "loss": 0.0779, "step": 6470 }, { "epoch": 0.21907434328408668, "grad_norm": 0.2791728973388672, "learning_rate": 9.994020810541098e-05, "loss": 0.0746, "step": 6480 }, { "epoch": 0.2194124209743399, "grad_norm": 0.16286632418632507, "learning_rate": 9.99393969910297e-05, "loss": 0.0699, "step": 6490 }, { "epoch": 0.21975049866459312, "grad_norm": 0.22159482538700104, "learning_rate": 9.993858041535104e-05, "loss": 0.0683, "step": 6500 }, { "epoch": 0.22008857635484635, "grad_norm": 0.439235657453537, "learning_rate": 9.99377583784643e-05, "loss": 0.0699, "step": 6510 }, { "epoch": 0.22042665404509956, "grad_norm": 0.3180888593196869, "learning_rate": 9.993693088045939e-05, "loss": 0.074, "step": 6520 }, { "epoch": 0.22076473173535277, "grad_norm": 0.24911080300807953, "learning_rate": 9.99360979214268e-05, "loss": 0.0682, "step": 6530 }, { "epoch": 0.221102809425606, "grad_norm": 0.6701829433441162, "learning_rate": 9.99352595014576e-05, "loss": 0.0723, "step": 6540 }, { "epoch": 0.22144088711585921, "grad_norm": 0.44183817505836487, "learning_rate": 9.993441562064354e-05, "loss": 0.0752, "step": 6550 }, { "epoch": 0.22177896480611245, "grad_norm": 0.1951495260000229, "learning_rate": 9.993356627907685e-05, "loss": 0.0707, "step": 6560 }, { "epoch": 0.22211704249636566, "grad_norm": 0.28763845562934875, "learning_rate": 9.99327114768504e-05, "loss": 0.0734, "step": 6570 }, { "epoch": 0.2224551201866189, "grad_norm": 0.2141682654619217, "learning_rate": 9.99318512140577e-05, "loss": 0.0713, "step": 6580 }, { "epoch": 0.2227931978768721, "grad_norm": 0.25968971848487854, "learning_rate": 9.993098549079284e-05, "loss": 0.0676, "step": 6590 }, { "epoch": 0.22313127556712534, "grad_norm": 0.24312198162078857, "learning_rate": 9.993011430715047e-05, "loss": 0.0732, "step": 6600 }, { "epoch": 0.22346935325737854, "grad_norm": 0.17371709644794464, "learning_rate": 9.992923766322586e-05, "loss": 0.0719, "step": 6610 }, { "epoch": 0.22380743094763175, "grad_norm": 0.4368915557861328, "learning_rate": 9.99283555591149e-05, "loss": 0.0713, "step": 6620 }, { "epoch": 0.224145508637885, "grad_norm": 0.28577205538749695, "learning_rate": 9.992746799491404e-05, "loss": 0.0751, "step": 6630 }, { "epoch": 0.2244835863281382, "grad_norm": 0.2546434998512268, "learning_rate": 9.992657497072033e-05, "loss": 0.0709, "step": 6640 }, { "epoch": 0.22482166401839143, "grad_norm": 0.2615114450454712, "learning_rate": 9.992567648663147e-05, "loss": 0.0706, "step": 6650 }, { "epoch": 0.22515974170864464, "grad_norm": 0.24910302460193634, "learning_rate": 9.992477254274568e-05, "loss": 0.0738, "step": 6660 }, { "epoch": 0.22549781939889788, "grad_norm": 0.20128867030143738, "learning_rate": 9.992386313916183e-05, "loss": 0.069, "step": 6670 }, { "epoch": 0.22583589708915108, "grad_norm": 0.22934269905090332, "learning_rate": 9.992294827597934e-05, "loss": 0.0692, "step": 6680 }, { "epoch": 0.22617397477940432, "grad_norm": 0.19564904272556305, "learning_rate": 9.992202795329831e-05, "loss": 0.0716, "step": 6690 }, { "epoch": 0.22651205246965753, "grad_norm": 0.25499585270881653, "learning_rate": 9.992110217121936e-05, "loss": 0.0743, "step": 6700 }, { "epoch": 0.22685013015991073, "grad_norm": 0.2843632996082306, "learning_rate": 9.992017092984372e-05, "loss": 0.0704, "step": 6710 }, { "epoch": 0.22718820785016397, "grad_norm": 0.2619327902793884, "learning_rate": 9.991923422927326e-05, "loss": 0.0687, "step": 6720 }, { "epoch": 0.22752628554041718, "grad_norm": 0.2755325436592102, "learning_rate": 9.991829206961037e-05, "loss": 0.0675, "step": 6730 }, { "epoch": 0.2278643632306704, "grad_norm": 0.2827800214290619, "learning_rate": 9.991734445095813e-05, "loss": 0.0692, "step": 6740 }, { "epoch": 0.22820244092092362, "grad_norm": 0.33203786611557007, "learning_rate": 9.991639137342015e-05, "loss": 0.0734, "step": 6750 }, { "epoch": 0.22854051861117686, "grad_norm": 0.2478916198015213, "learning_rate": 9.991543283710064e-05, "loss": 0.0664, "step": 6760 }, { "epoch": 0.22887859630143007, "grad_norm": 0.3694203495979309, "learning_rate": 9.991446884210445e-05, "loss": 0.0771, "step": 6770 }, { "epoch": 0.2292166739916833, "grad_norm": 0.26483672857284546, "learning_rate": 9.9913499388537e-05, "loss": 0.0697, "step": 6780 }, { "epoch": 0.2295547516819365, "grad_norm": 0.3373892903327942, "learning_rate": 9.99125244765043e-05, "loss": 0.0786, "step": 6790 }, { "epoch": 0.22989282937218972, "grad_norm": 0.2900245785713196, "learning_rate": 9.991154410611296e-05, "loss": 0.0704, "step": 6800 }, { "epoch": 0.23023090706244295, "grad_norm": 0.47388502955436707, "learning_rate": 9.99105582774702e-05, "loss": 0.0722, "step": 6810 }, { "epoch": 0.23056898475269616, "grad_norm": 0.2140796184539795, "learning_rate": 9.990956699068384e-05, "loss": 0.0674, "step": 6820 }, { "epoch": 0.2309070624429494, "grad_norm": 0.18166862428188324, "learning_rate": 9.990857024586224e-05, "loss": 0.0809, "step": 6830 }, { "epoch": 0.2312451401332026, "grad_norm": 0.25160861015319824, "learning_rate": 9.990756804311446e-05, "loss": 0.0629, "step": 6840 }, { "epoch": 0.23158321782345584, "grad_norm": 0.36977481842041016, "learning_rate": 9.990656038255006e-05, "loss": 0.0654, "step": 6850 }, { "epoch": 0.23192129551370905, "grad_norm": 0.32322216033935547, "learning_rate": 9.990554726427926e-05, "loss": 0.0647, "step": 6860 }, { "epoch": 0.23225937320396228, "grad_norm": 0.2903149127960205, "learning_rate": 9.990452868841284e-05, "loss": 0.0649, "step": 6870 }, { "epoch": 0.2325974508942155, "grad_norm": 0.3557782471179962, "learning_rate": 9.99035046550622e-05, "loss": 0.0774, "step": 6880 }, { "epoch": 0.2329355285844687, "grad_norm": 0.5014543533325195, "learning_rate": 9.99024751643393e-05, "loss": 0.064, "step": 6890 }, { "epoch": 0.23327360627472193, "grad_norm": 0.45782211422920227, "learning_rate": 9.990144021635677e-05, "loss": 0.0726, "step": 6900 }, { "epoch": 0.23361168396497514, "grad_norm": 0.2748390734195709, "learning_rate": 9.990039981122775e-05, "loss": 0.0706, "step": 6910 }, { "epoch": 0.23394976165522838, "grad_norm": 0.4175605773925781, "learning_rate": 9.989935394906602e-05, "loss": 0.0667, "step": 6920 }, { "epoch": 0.23428783934548159, "grad_norm": 0.3641161322593689, "learning_rate": 9.989830262998598e-05, "loss": 0.0753, "step": 6930 }, { "epoch": 0.23462591703573482, "grad_norm": 0.2656913697719574, "learning_rate": 9.989724585410259e-05, "loss": 0.0626, "step": 6940 }, { "epoch": 0.23496399472598803, "grad_norm": 0.3199130892753601, "learning_rate": 9.989618362153139e-05, "loss": 0.068, "step": 6950 }, { "epoch": 0.23530207241624126, "grad_norm": 0.22432108223438263, "learning_rate": 9.989511593238859e-05, "loss": 0.059, "step": 6960 }, { "epoch": 0.23564015010649447, "grad_norm": 0.39884206652641296, "learning_rate": 9.98940427867909e-05, "loss": 0.0711, "step": 6970 }, { "epoch": 0.23597822779674768, "grad_norm": 0.2159564197063446, "learning_rate": 9.989296418485573e-05, "loss": 0.0699, "step": 6980 }, { "epoch": 0.23631630548700092, "grad_norm": 0.29532933235168457, "learning_rate": 9.989188012670101e-05, "loss": 0.062, "step": 6990 }, { "epoch": 0.23665438317725412, "grad_norm": 0.20850323140621185, "learning_rate": 9.989079061244528e-05, "loss": 0.0684, "step": 7000 }, { "epoch": 0.23699246086750736, "grad_norm": 0.29833561182022095, "learning_rate": 9.988969564220769e-05, "loss": 0.0712, "step": 7010 }, { "epoch": 0.23733053855776057, "grad_norm": 0.26294365525245667, "learning_rate": 9.988859521610801e-05, "loss": 0.0713, "step": 7020 }, { "epoch": 0.2376686162480138, "grad_norm": 0.26738569140434265, "learning_rate": 9.988748933426656e-05, "loss": 0.0735, "step": 7030 }, { "epoch": 0.238006693938267, "grad_norm": 0.2760653793811798, "learning_rate": 9.988637799680428e-05, "loss": 0.068, "step": 7040 }, { "epoch": 0.23834477162852025, "grad_norm": 0.2530217170715332, "learning_rate": 9.98852612038427e-05, "loss": 0.0716, "step": 7050 }, { "epoch": 0.23868284931877345, "grad_norm": 0.2860371470451355, "learning_rate": 9.988413895550397e-05, "loss": 0.0697, "step": 7060 }, { "epoch": 0.23902092700902666, "grad_norm": 0.39485275745391846, "learning_rate": 9.98830112519108e-05, "loss": 0.0746, "step": 7070 }, { "epoch": 0.2393590046992799, "grad_norm": 0.2732797861099243, "learning_rate": 9.98818780931865e-05, "loss": 0.0678, "step": 7080 }, { "epoch": 0.2396970823895331, "grad_norm": 0.522678554058075, "learning_rate": 9.988073947945502e-05, "loss": 0.0662, "step": 7090 }, { "epoch": 0.24003516007978634, "grad_norm": 0.28076353669166565, "learning_rate": 9.987959541084087e-05, "loss": 0.0711, "step": 7100 }, { "epoch": 0.24037323777003955, "grad_norm": 0.4876171350479126, "learning_rate": 9.987844588746915e-05, "loss": 0.0753, "step": 7110 }, { "epoch": 0.24071131546029279, "grad_norm": 0.5187517404556274, "learning_rate": 9.987729090946558e-05, "loss": 0.0742, "step": 7120 }, { "epoch": 0.241049393150546, "grad_norm": 0.3778478801250458, "learning_rate": 9.987613047695647e-05, "loss": 0.0691, "step": 7130 }, { "epoch": 0.24138747084079923, "grad_norm": 0.2803983688354492, "learning_rate": 9.987496459006871e-05, "loss": 0.0743, "step": 7140 }, { "epoch": 0.24172554853105244, "grad_norm": 0.29494303464889526, "learning_rate": 9.987379324892982e-05, "loss": 0.0686, "step": 7150 }, { "epoch": 0.24206362622130564, "grad_norm": 0.2821930944919586, "learning_rate": 9.987261645366788e-05, "loss": 0.0695, "step": 7160 }, { "epoch": 0.24240170391155888, "grad_norm": 0.4102829098701477, "learning_rate": 9.987143420441158e-05, "loss": 0.0668, "step": 7170 }, { "epoch": 0.2427397816018121, "grad_norm": 0.21178914606571198, "learning_rate": 9.987024650129022e-05, "loss": 0.0646, "step": 7180 }, { "epoch": 0.24307785929206532, "grad_norm": 0.30715543031692505, "learning_rate": 9.986905334443368e-05, "loss": 0.065, "step": 7190 }, { "epoch": 0.24341593698231853, "grad_norm": 0.27089864015579224, "learning_rate": 9.986785473397245e-05, "loss": 0.0776, "step": 7200 }, { "epoch": 0.24375401467257177, "grad_norm": 0.30032920837402344, "learning_rate": 9.98666506700376e-05, "loss": 0.0683, "step": 7210 }, { "epoch": 0.24409209236282498, "grad_norm": 0.3071911334991455, "learning_rate": 9.986544115276081e-05, "loss": 0.0751, "step": 7220 }, { "epoch": 0.2444301700530782, "grad_norm": 0.3472267687320709, "learning_rate": 9.986422618227433e-05, "loss": 0.0752, "step": 7230 }, { "epoch": 0.24476824774333142, "grad_norm": 0.27325767278671265, "learning_rate": 9.986300575871106e-05, "loss": 0.0762, "step": 7240 }, { "epoch": 0.24510632543358463, "grad_norm": 0.29865768551826477, "learning_rate": 9.986177988220444e-05, "loss": 0.0703, "step": 7250 }, { "epoch": 0.24544440312383786, "grad_norm": 0.28185608983039856, "learning_rate": 9.986054855288856e-05, "loss": 0.0715, "step": 7260 }, { "epoch": 0.24578248081409107, "grad_norm": 0.2724192440509796, "learning_rate": 9.985931177089802e-05, "loss": 0.0723, "step": 7270 }, { "epoch": 0.2461205585043443, "grad_norm": 0.44990676641464233, "learning_rate": 9.985806953636814e-05, "loss": 0.0699, "step": 7280 }, { "epoch": 0.2464586361945975, "grad_norm": 0.3230467140674591, "learning_rate": 9.985682184943471e-05, "loss": 0.0691, "step": 7290 }, { "epoch": 0.24679671388485075, "grad_norm": 0.24998512864112854, "learning_rate": 9.98555687102342e-05, "loss": 0.0689, "step": 7300 }, { "epoch": 0.24713479157510396, "grad_norm": 0.24465008080005646, "learning_rate": 9.985431011890367e-05, "loss": 0.0713, "step": 7310 }, { "epoch": 0.2474728692653572, "grad_norm": 0.3574208617210388, "learning_rate": 9.985304607558075e-05, "loss": 0.07, "step": 7320 }, { "epoch": 0.2478109469556104, "grad_norm": 0.2650483548641205, "learning_rate": 9.985177658040364e-05, "loss": 0.0631, "step": 7330 }, { "epoch": 0.2481490246458636, "grad_norm": 0.24366912245750427, "learning_rate": 9.985050163351119e-05, "loss": 0.0676, "step": 7340 }, { "epoch": 0.24848710233611684, "grad_norm": 0.24212884902954102, "learning_rate": 9.984922123504286e-05, "loss": 0.068, "step": 7350 }, { "epoch": 0.24882518002637005, "grad_norm": 0.3357871174812317, "learning_rate": 9.984793538513862e-05, "loss": 0.0737, "step": 7360 }, { "epoch": 0.2491632577166233, "grad_norm": 0.26296862959861755, "learning_rate": 9.984664408393912e-05, "loss": 0.0724, "step": 7370 }, { "epoch": 0.2495013354068765, "grad_norm": 0.31365135312080383, "learning_rate": 9.984534733158556e-05, "loss": 0.0693, "step": 7380 }, { "epoch": 0.24983941309712973, "grad_norm": 0.3317549526691437, "learning_rate": 9.984404512821977e-05, "loss": 0.0735, "step": 7390 }, { "epoch": 0.25017749078738294, "grad_norm": 0.21680741012096405, "learning_rate": 9.984273747398411e-05, "loss": 0.0707, "step": 7400 }, { "epoch": 0.2505155684776362, "grad_norm": 0.2102835327386856, "learning_rate": 9.984142436902165e-05, "loss": 0.0651, "step": 7410 }, { "epoch": 0.25085364616788935, "grad_norm": 0.21778477728366852, "learning_rate": 9.984010581347596e-05, "loss": 0.065, "step": 7420 }, { "epoch": 0.2511917238581426, "grad_norm": 0.38680076599121094, "learning_rate": 9.983878180749121e-05, "loss": 0.0686, "step": 7430 }, { "epoch": 0.2515298015483958, "grad_norm": 0.2767406105995178, "learning_rate": 9.983745235121222e-05, "loss": 0.0671, "step": 7440 }, { "epoch": 0.25186787923864906, "grad_norm": 0.3458111584186554, "learning_rate": 9.983611744478438e-05, "loss": 0.0682, "step": 7450 }, { "epoch": 0.25220595692890224, "grad_norm": 0.5427563190460205, "learning_rate": 9.983477708835365e-05, "loss": 0.0732, "step": 7460 }, { "epoch": 0.2525440346191555, "grad_norm": 0.3029729127883911, "learning_rate": 9.983343128206664e-05, "loss": 0.0712, "step": 7470 }, { "epoch": 0.2528821123094087, "grad_norm": 0.44681745767593384, "learning_rate": 9.983208002607049e-05, "loss": 0.0636, "step": 7480 }, { "epoch": 0.25322018999966195, "grad_norm": 0.3338148891925812, "learning_rate": 9.9830723320513e-05, "loss": 0.0643, "step": 7490 }, { "epoch": 0.25355826768991513, "grad_norm": 0.19398586452007294, "learning_rate": 9.982936116554254e-05, "loss": 0.0671, "step": 7500 }, { "epoch": 0.25389634538016836, "grad_norm": 0.2262643277645111, "learning_rate": 9.982799356130803e-05, "loss": 0.0691, "step": 7510 }, { "epoch": 0.2542344230704216, "grad_norm": 0.29669803380966187, "learning_rate": 9.982662050795908e-05, "loss": 0.0685, "step": 7520 }, { "epoch": 0.2545725007606748, "grad_norm": 0.23391515016555786, "learning_rate": 9.982524200564583e-05, "loss": 0.0693, "step": 7530 }, { "epoch": 0.254910578450928, "grad_norm": 0.2912433445453644, "learning_rate": 9.982385805451901e-05, "loss": 0.0662, "step": 7540 }, { "epoch": 0.25524865614118125, "grad_norm": 0.2644866704940796, "learning_rate": 9.982246865472998e-05, "loss": 0.0673, "step": 7550 }, { "epoch": 0.2555867338314345, "grad_norm": 0.2345660924911499, "learning_rate": 9.982107380643069e-05, "loss": 0.0814, "step": 7560 }, { "epoch": 0.25592481152168767, "grad_norm": 0.20200121402740479, "learning_rate": 9.981967350977368e-05, "loss": 0.0692, "step": 7570 }, { "epoch": 0.2562628892119409, "grad_norm": 0.17885206639766693, "learning_rate": 9.981826776491208e-05, "loss": 0.0712, "step": 7580 }, { "epoch": 0.25660096690219414, "grad_norm": 0.19478273391723633, "learning_rate": 9.98168565719996e-05, "loss": 0.0722, "step": 7590 }, { "epoch": 0.2569390445924473, "grad_norm": 0.2075556218624115, "learning_rate": 9.98154399311906e-05, "loss": 0.074, "step": 7600 }, { "epoch": 0.25727712228270055, "grad_norm": 0.33065858483314514, "learning_rate": 9.981401784263997e-05, "loss": 0.0658, "step": 7610 }, { "epoch": 0.2576151999729538, "grad_norm": 0.2856706976890564, "learning_rate": 9.981259030650326e-05, "loss": 0.0675, "step": 7620 }, { "epoch": 0.257953277663207, "grad_norm": 0.365749329328537, "learning_rate": 9.981115732293655e-05, "loss": 0.0737, "step": 7630 }, { "epoch": 0.2582913553534602, "grad_norm": 0.20246708393096924, "learning_rate": 9.980971889209659e-05, "loss": 0.0645, "step": 7640 }, { "epoch": 0.25862943304371344, "grad_norm": 0.2975812554359436, "learning_rate": 9.980827501414064e-05, "loss": 0.0687, "step": 7650 }, { "epoch": 0.2589675107339667, "grad_norm": 0.3805651068687439, "learning_rate": 9.980682568922663e-05, "loss": 0.0672, "step": 7660 }, { "epoch": 0.2593055884242199, "grad_norm": 0.19961687922477722, "learning_rate": 9.980537091751304e-05, "loss": 0.0661, "step": 7670 }, { "epoch": 0.2596436661144731, "grad_norm": 0.3210427463054657, "learning_rate": 9.980391069915897e-05, "loss": 0.0728, "step": 7680 }, { "epoch": 0.25998174380472633, "grad_norm": 0.2363499253988266, "learning_rate": 9.98024450343241e-05, "loss": 0.0653, "step": 7690 }, { "epoch": 0.26031982149497956, "grad_norm": 0.4483868479728699, "learning_rate": 9.980097392316872e-05, "loss": 0.0673, "step": 7700 }, { "epoch": 0.26065789918523274, "grad_norm": 0.29984837770462036, "learning_rate": 9.97994973658537e-05, "loss": 0.0705, "step": 7710 }, { "epoch": 0.260995976875486, "grad_norm": 0.36199700832366943, "learning_rate": 9.979801536254054e-05, "loss": 0.0647, "step": 7720 }, { "epoch": 0.2613340545657392, "grad_norm": 0.29247891902923584, "learning_rate": 9.979652791339127e-05, "loss": 0.0642, "step": 7730 }, { "epoch": 0.26167213225599245, "grad_norm": 0.2598596513271332, "learning_rate": 9.97950350185686e-05, "loss": 0.0684, "step": 7740 }, { "epoch": 0.26201020994624563, "grad_norm": 0.26128703355789185, "learning_rate": 9.979353667823574e-05, "loss": 0.0707, "step": 7750 }, { "epoch": 0.26234828763649887, "grad_norm": 0.3034011423587799, "learning_rate": 9.979203289255658e-05, "loss": 0.0666, "step": 7760 }, { "epoch": 0.2626863653267521, "grad_norm": 0.34200555086135864, "learning_rate": 9.979052366169557e-05, "loss": 0.0706, "step": 7770 }, { "epoch": 0.2630244430170053, "grad_norm": 0.25287148356437683, "learning_rate": 9.978900898581775e-05, "loss": 0.0676, "step": 7780 }, { "epoch": 0.2633625207072585, "grad_norm": 0.27143847942352295, "learning_rate": 9.978748886508875e-05, "loss": 0.064, "step": 7790 }, { "epoch": 0.26370059839751175, "grad_norm": 0.24523311853408813, "learning_rate": 9.978596329967484e-05, "loss": 0.0691, "step": 7800 }, { "epoch": 0.264038676087765, "grad_norm": 0.23963351547718048, "learning_rate": 9.978443228974284e-05, "loss": 0.0728, "step": 7810 }, { "epoch": 0.26437675377801817, "grad_norm": 0.20946568250656128, "learning_rate": 9.978289583546015e-05, "loss": 0.0741, "step": 7820 }, { "epoch": 0.2647148314682714, "grad_norm": 0.23538878560066223, "learning_rate": 9.978135393699484e-05, "loss": 0.073, "step": 7830 }, { "epoch": 0.26505290915852464, "grad_norm": 0.21072177588939667, "learning_rate": 9.977980659451548e-05, "loss": 0.0678, "step": 7840 }, { "epoch": 0.2653909868487779, "grad_norm": 0.19761107861995697, "learning_rate": 9.977825380819135e-05, "loss": 0.0715, "step": 7850 }, { "epoch": 0.26572906453903106, "grad_norm": 0.2268187701702118, "learning_rate": 9.97766955781922e-05, "loss": 0.0694, "step": 7860 }, { "epoch": 0.2660671422292843, "grad_norm": 0.2280113697052002, "learning_rate": 9.977513190468848e-05, "loss": 0.0732, "step": 7870 }, { "epoch": 0.26640521991953753, "grad_norm": 0.3359900712966919, "learning_rate": 9.977356278785116e-05, "loss": 0.0639, "step": 7880 }, { "epoch": 0.2667432976097907, "grad_norm": 0.3139697313308716, "learning_rate": 9.977198822785184e-05, "loss": 0.0731, "step": 7890 }, { "epoch": 0.26708137530004394, "grad_norm": 0.2887623608112335, "learning_rate": 9.977040822486273e-05, "loss": 0.067, "step": 7900 }, { "epoch": 0.2674194529902972, "grad_norm": 0.3537541925907135, "learning_rate": 9.97688227790566e-05, "loss": 0.0821, "step": 7910 }, { "epoch": 0.2677575306805504, "grad_norm": 0.5124951601028442, "learning_rate": 9.976723189060684e-05, "loss": 0.0736, "step": 7920 }, { "epoch": 0.2680956083708036, "grad_norm": 0.36900579929351807, "learning_rate": 9.976563555968742e-05, "loss": 0.0761, "step": 7930 }, { "epoch": 0.26843368606105683, "grad_norm": 0.3262450098991394, "learning_rate": 9.976403378647292e-05, "loss": 0.0721, "step": 7940 }, { "epoch": 0.26877176375131007, "grad_norm": 0.31257641315460205, "learning_rate": 9.97624265711385e-05, "loss": 0.0752, "step": 7950 }, { "epoch": 0.26910984144156325, "grad_norm": 0.31997618079185486, "learning_rate": 9.976081391385993e-05, "loss": 0.0742, "step": 7960 }, { "epoch": 0.2694479191318165, "grad_norm": 0.2246505618095398, "learning_rate": 9.975919581481356e-05, "loss": 0.0705, "step": 7970 }, { "epoch": 0.2697859968220697, "grad_norm": 0.2933832108974457, "learning_rate": 9.975757227417634e-05, "loss": 0.07, "step": 7980 }, { "epoch": 0.27012407451232295, "grad_norm": 0.3213794529438019, "learning_rate": 9.975594329212586e-05, "loss": 0.067, "step": 7990 }, { "epoch": 0.27046215220257613, "grad_norm": 0.2593749165534973, "learning_rate": 9.97543088688402e-05, "loss": 0.0566, "step": 8000 }, { "epoch": 0.27080022989282937, "grad_norm": 0.366862952709198, "learning_rate": 9.975266900449814e-05, "loss": 0.0644, "step": 8010 }, { "epoch": 0.2711383075830826, "grad_norm": 0.29315000772476196, "learning_rate": 9.975102369927898e-05, "loss": 0.0662, "step": 8020 }, { "epoch": 0.2714763852733358, "grad_norm": 0.47880423069000244, "learning_rate": 9.974937295336269e-05, "loss": 0.0764, "step": 8030 }, { "epoch": 0.271814462963589, "grad_norm": 0.1656324416399002, "learning_rate": 9.974771676692975e-05, "loss": 0.0726, "step": 8040 }, { "epoch": 0.27215254065384226, "grad_norm": 0.3236614763736725, "learning_rate": 9.974605514016131e-05, "loss": 0.0687, "step": 8050 }, { "epoch": 0.2724906183440955, "grad_norm": 0.19669900834560394, "learning_rate": 9.974438807323907e-05, "loss": 0.0687, "step": 8060 }, { "epoch": 0.27282869603434867, "grad_norm": 0.2931502163410187, "learning_rate": 9.974271556634535e-05, "loss": 0.0602, "step": 8070 }, { "epoch": 0.2731667737246019, "grad_norm": 0.27651718258857727, "learning_rate": 9.974103761966302e-05, "loss": 0.0756, "step": 8080 }, { "epoch": 0.27350485141485514, "grad_norm": 0.3791002929210663, "learning_rate": 9.973935423337563e-05, "loss": 0.0684, "step": 8090 }, { "epoch": 0.2738429291051084, "grad_norm": 0.3200909197330475, "learning_rate": 9.973766540766722e-05, "loss": 0.0671, "step": 8100 }, { "epoch": 0.27418100679536156, "grad_norm": 0.27150267362594604, "learning_rate": 9.97359711427225e-05, "loss": 0.0711, "step": 8110 }, { "epoch": 0.2745190844856148, "grad_norm": 0.30944105982780457, "learning_rate": 9.973427143872677e-05, "loss": 0.0613, "step": 8120 }, { "epoch": 0.27485716217586803, "grad_norm": 0.2136119157075882, "learning_rate": 9.973256629586589e-05, "loss": 0.0665, "step": 8130 }, { "epoch": 0.2751952398661212, "grad_norm": 0.36854684352874756, "learning_rate": 9.973085571432632e-05, "loss": 0.0688, "step": 8140 }, { "epoch": 0.27553331755637445, "grad_norm": 0.31995728611946106, "learning_rate": 9.972913969429513e-05, "loss": 0.0685, "step": 8150 }, { "epoch": 0.2758713952466277, "grad_norm": 0.24564655125141144, "learning_rate": 9.972741823596e-05, "loss": 0.072, "step": 8160 }, { "epoch": 0.2762094729368809, "grad_norm": 0.2907475233078003, "learning_rate": 9.972569133950917e-05, "loss": 0.0689, "step": 8170 }, { "epoch": 0.2765475506271341, "grad_norm": 0.2906219959259033, "learning_rate": 9.972395900513151e-05, "loss": 0.0683, "step": 8180 }, { "epoch": 0.27688562831738733, "grad_norm": 0.40003257989883423, "learning_rate": 9.972222123301645e-05, "loss": 0.0671, "step": 8190 }, { "epoch": 0.27722370600764057, "grad_norm": 0.33690693974494934, "learning_rate": 9.972047802335403e-05, "loss": 0.0685, "step": 8200 }, { "epoch": 0.27756178369789375, "grad_norm": 0.3376626968383789, "learning_rate": 9.971872937633488e-05, "loss": 0.0639, "step": 8210 }, { "epoch": 0.277899861388147, "grad_norm": 0.32320886850357056, "learning_rate": 9.971697529215024e-05, "loss": 0.0652, "step": 8220 }, { "epoch": 0.2782379390784002, "grad_norm": 0.23526102304458618, "learning_rate": 9.971521577099192e-05, "loss": 0.0768, "step": 8230 }, { "epoch": 0.27857601676865346, "grad_norm": 0.24766480922698975, "learning_rate": 9.971345081305236e-05, "loss": 0.0723, "step": 8240 }, { "epoch": 0.27891409445890664, "grad_norm": 0.3086213767528534, "learning_rate": 9.971168041852456e-05, "loss": 0.0832, "step": 8250 }, { "epoch": 0.27925217214915987, "grad_norm": 0.30118274688720703, "learning_rate": 9.970990458760215e-05, "loss": 0.07, "step": 8260 }, { "epoch": 0.2795902498394131, "grad_norm": 0.35055410861968994, "learning_rate": 9.970812332047929e-05, "loss": 0.0675, "step": 8270 }, { "epoch": 0.27992832752966634, "grad_norm": 0.24675868451595306, "learning_rate": 9.97063366173508e-05, "loss": 0.0691, "step": 8280 }, { "epoch": 0.2802664052199195, "grad_norm": 0.5119824409484863, "learning_rate": 9.970454447841207e-05, "loss": 0.0727, "step": 8290 }, { "epoch": 0.28060448291017276, "grad_norm": 0.32606780529022217, "learning_rate": 9.970274690385909e-05, "loss": 0.0638, "step": 8300 }, { "epoch": 0.280942560600426, "grad_norm": 0.3248588442802429, "learning_rate": 9.970094389388844e-05, "loss": 0.0744, "step": 8310 }, { "epoch": 0.2812806382906792, "grad_norm": 0.23067130148410797, "learning_rate": 9.969913544869728e-05, "loss": 0.0722, "step": 8320 }, { "epoch": 0.2816187159809324, "grad_norm": 0.24566955864429474, "learning_rate": 9.96973215684834e-05, "loss": 0.0718, "step": 8330 }, { "epoch": 0.28195679367118565, "grad_norm": 0.26055988669395447, "learning_rate": 9.969550225344513e-05, "loss": 0.0681, "step": 8340 }, { "epoch": 0.2822948713614389, "grad_norm": 0.32180580496788025, "learning_rate": 9.969367750378147e-05, "loss": 0.0712, "step": 8350 }, { "epoch": 0.28263294905169206, "grad_norm": 0.3464031219482422, "learning_rate": 9.969184731969194e-05, "loss": 0.0737, "step": 8360 }, { "epoch": 0.2829710267419453, "grad_norm": 0.20567087829113007, "learning_rate": 9.96900117013767e-05, "loss": 0.0776, "step": 8370 }, { "epoch": 0.28330910443219853, "grad_norm": 0.28304219245910645, "learning_rate": 9.96881706490365e-05, "loss": 0.0663, "step": 8380 }, { "epoch": 0.2836471821224517, "grad_norm": 0.19533862173557281, "learning_rate": 9.968632416287265e-05, "loss": 0.0652, "step": 8390 }, { "epoch": 0.28398525981270495, "grad_norm": 0.43603187799453735, "learning_rate": 9.96844722430871e-05, "loss": 0.0684, "step": 8400 }, { "epoch": 0.2843233375029582, "grad_norm": 0.28536295890808105, "learning_rate": 9.968261488988235e-05, "loss": 0.0679, "step": 8410 }, { "epoch": 0.2846614151932114, "grad_norm": 0.34283149242401123, "learning_rate": 9.968075210346155e-05, "loss": 0.0622, "step": 8420 }, { "epoch": 0.2849994928834646, "grad_norm": 0.22736063599586487, "learning_rate": 9.967888388402839e-05, "loss": 0.0627, "step": 8430 }, { "epoch": 0.28533757057371784, "grad_norm": 0.19238626956939697, "learning_rate": 9.967701023178717e-05, "loss": 0.0678, "step": 8440 }, { "epoch": 0.28567564826397107, "grad_norm": 0.28308922052383423, "learning_rate": 9.967513114694282e-05, "loss": 0.0758, "step": 8450 }, { "epoch": 0.2860137259542243, "grad_norm": 0.3809327781200409, "learning_rate": 9.967324662970079e-05, "loss": 0.0667, "step": 8460 }, { "epoch": 0.2863518036444775, "grad_norm": 0.17512205243110657, "learning_rate": 9.96713566802672e-05, "loss": 0.064, "step": 8470 }, { "epoch": 0.2866898813347307, "grad_norm": 0.35101908445358276, "learning_rate": 9.966946129884873e-05, "loss": 0.0709, "step": 8480 }, { "epoch": 0.28702795902498396, "grad_norm": 0.2611229121685028, "learning_rate": 9.966756048565265e-05, "loss": 0.0662, "step": 8490 }, { "epoch": 0.28736603671523714, "grad_norm": 0.28014183044433594, "learning_rate": 9.966565424088681e-05, "loss": 0.0657, "step": 8500 }, { "epoch": 0.2877041144054904, "grad_norm": 0.29072579741477966, "learning_rate": 9.96637425647597e-05, "loss": 0.0674, "step": 8510 }, { "epoch": 0.2880421920957436, "grad_norm": 0.2881372272968292, "learning_rate": 9.966182545748038e-05, "loss": 0.0635, "step": 8520 }, { "epoch": 0.28838026978599685, "grad_norm": 0.27061745524406433, "learning_rate": 9.96599029192585e-05, "loss": 0.0645, "step": 8530 }, { "epoch": 0.28871834747625, "grad_norm": 0.33192676305770874, "learning_rate": 9.965797495030428e-05, "loss": 0.07, "step": 8540 }, { "epoch": 0.28905642516650326, "grad_norm": 0.3624013662338257, "learning_rate": 9.96560415508286e-05, "loss": 0.0641, "step": 8550 }, { "epoch": 0.2893945028567565, "grad_norm": 0.23219646513462067, "learning_rate": 9.965410272104286e-05, "loss": 0.0733, "step": 8560 }, { "epoch": 0.2897325805470097, "grad_norm": 0.5592046976089478, "learning_rate": 9.96521584611591e-05, "loss": 0.0803, "step": 8570 }, { "epoch": 0.2900706582372629, "grad_norm": 0.3483235836029053, "learning_rate": 9.965020877138994e-05, "loss": 0.0697, "step": 8580 }, { "epoch": 0.29040873592751615, "grad_norm": 0.25224336981773376, "learning_rate": 9.964825365194861e-05, "loss": 0.0715, "step": 8590 }, { "epoch": 0.2907468136177694, "grad_norm": 0.2634609639644623, "learning_rate": 9.96462931030489e-05, "loss": 0.0599, "step": 8600 }, { "epoch": 0.29108489130802256, "grad_norm": 0.2375643253326416, "learning_rate": 9.96443271249052e-05, "loss": 0.0692, "step": 8610 }, { "epoch": 0.2914229689982758, "grad_norm": 0.3916403651237488, "learning_rate": 9.964235571773255e-05, "loss": 0.0658, "step": 8620 }, { "epoch": 0.29176104668852904, "grad_norm": 0.22279614210128784, "learning_rate": 9.96403788817465e-05, "loss": 0.067, "step": 8630 }, { "epoch": 0.29209912437878227, "grad_norm": 0.39244478940963745, "learning_rate": 9.963839661716325e-05, "loss": 0.0646, "step": 8640 }, { "epoch": 0.29243720206903545, "grad_norm": 0.23448199033737183, "learning_rate": 9.963640892419958e-05, "loss": 0.0723, "step": 8650 }, { "epoch": 0.2927752797592887, "grad_norm": 0.4942508339881897, "learning_rate": 9.963441580307286e-05, "loss": 0.0697, "step": 8660 }, { "epoch": 0.2931133574495419, "grad_norm": 0.2973001301288605, "learning_rate": 9.963241725400104e-05, "loss": 0.0714, "step": 8670 }, { "epoch": 0.2934514351397951, "grad_norm": 0.3123125731945038, "learning_rate": 9.963041327720271e-05, "loss": 0.0706, "step": 8680 }, { "epoch": 0.29378951283004834, "grad_norm": 0.2228439301252365, "learning_rate": 9.962840387289697e-05, "loss": 0.0641, "step": 8690 }, { "epoch": 0.2941275905203016, "grad_norm": 0.2506275475025177, "learning_rate": 9.962638904130363e-05, "loss": 0.0722, "step": 8700 }, { "epoch": 0.2944656682105548, "grad_norm": 0.27913030982017517, "learning_rate": 9.962436878264298e-05, "loss": 0.0704, "step": 8710 }, { "epoch": 0.294803745900808, "grad_norm": 0.33232244849205017, "learning_rate": 9.962234309713598e-05, "loss": 0.0713, "step": 8720 }, { "epoch": 0.2951418235910612, "grad_norm": 0.27537429332733154, "learning_rate": 9.962031198500414e-05, "loss": 0.0697, "step": 8730 }, { "epoch": 0.29547990128131446, "grad_norm": 0.29940417408943176, "learning_rate": 9.961827544646958e-05, "loss": 0.0704, "step": 8740 }, { "epoch": 0.29581797897156764, "grad_norm": 0.24933797121047974, "learning_rate": 9.961623348175501e-05, "loss": 0.0621, "step": 8750 }, { "epoch": 0.2961560566618209, "grad_norm": 0.2564278542995453, "learning_rate": 9.961418609108377e-05, "loss": 0.0663, "step": 8760 }, { "epoch": 0.2964941343520741, "grad_norm": 0.32145625352859497, "learning_rate": 9.961213327467971e-05, "loss": 0.0706, "step": 8770 }, { "epoch": 0.29683221204232735, "grad_norm": 0.2666192948818207, "learning_rate": 9.961007503276736e-05, "loss": 0.0684, "step": 8780 }, { "epoch": 0.29717028973258053, "grad_norm": 0.27423545718193054, "learning_rate": 9.960801136557179e-05, "loss": 0.068, "step": 8790 }, { "epoch": 0.29750836742283376, "grad_norm": 0.32093337178230286, "learning_rate": 9.960594227331866e-05, "loss": 0.0599, "step": 8800 }, { "epoch": 0.297846445113087, "grad_norm": 0.298094779253006, "learning_rate": 9.960386775623429e-05, "loss": 0.0679, "step": 8810 }, { "epoch": 0.29818452280334024, "grad_norm": 0.31175267696380615, "learning_rate": 9.96017878145455e-05, "loss": 0.0719, "step": 8820 }, { "epoch": 0.2985226004935934, "grad_norm": 0.27904728055000305, "learning_rate": 9.959970244847977e-05, "loss": 0.0643, "step": 8830 }, { "epoch": 0.29886067818384665, "grad_norm": 0.2508603036403656, "learning_rate": 9.959761165826518e-05, "loss": 0.0713, "step": 8840 }, { "epoch": 0.2991987558740999, "grad_norm": 0.31259676814079285, "learning_rate": 9.959551544413033e-05, "loss": 0.0654, "step": 8850 }, { "epoch": 0.29953683356435307, "grad_norm": 0.3868718147277832, "learning_rate": 9.959341380630448e-05, "loss": 0.0693, "step": 8860 }, { "epoch": 0.2998749112546063, "grad_norm": 0.2984175980091095, "learning_rate": 9.959130674501746e-05, "loss": 0.0661, "step": 8870 }, { "epoch": 0.30021298894485954, "grad_norm": 0.1923610419034958, "learning_rate": 9.958919426049968e-05, "loss": 0.0686, "step": 8880 }, { "epoch": 0.3005510666351128, "grad_norm": 0.28355973958969116, "learning_rate": 9.958707635298219e-05, "loss": 0.067, "step": 8890 }, { "epoch": 0.30088914432536595, "grad_norm": 0.4078853726387024, "learning_rate": 9.958495302269657e-05, "loss": 0.0618, "step": 8900 }, { "epoch": 0.3012272220156192, "grad_norm": 0.3250052034854889, "learning_rate": 9.958282426987503e-05, "loss": 0.0626, "step": 8910 }, { "epoch": 0.3015652997058724, "grad_norm": 0.28466495871543884, "learning_rate": 9.95806900947504e-05, "loss": 0.0696, "step": 8920 }, { "epoch": 0.3019033773961256, "grad_norm": 0.22981996834278107, "learning_rate": 9.957855049755604e-05, "loss": 0.073, "step": 8930 }, { "epoch": 0.30224145508637884, "grad_norm": 0.312997430562973, "learning_rate": 9.957640547852593e-05, "loss": 0.0657, "step": 8940 }, { "epoch": 0.3025795327766321, "grad_norm": 0.22960761189460754, "learning_rate": 9.957425503789466e-05, "loss": 0.0631, "step": 8950 }, { "epoch": 0.3029176104668853, "grad_norm": 0.23992152512073517, "learning_rate": 9.957209917589738e-05, "loss": 0.0645, "step": 8960 }, { "epoch": 0.3032556881571385, "grad_norm": 0.5174907445907593, "learning_rate": 9.956993789276987e-05, "loss": 0.0741, "step": 8970 }, { "epoch": 0.30359376584739173, "grad_norm": 0.41880688071250916, "learning_rate": 9.956777118874847e-05, "loss": 0.0666, "step": 8980 }, { "epoch": 0.30393184353764496, "grad_norm": 0.225642129778862, "learning_rate": 9.956559906407016e-05, "loss": 0.0666, "step": 8990 }, { "epoch": 0.30426992122789814, "grad_norm": 0.3545491099357605, "learning_rate": 9.956342151897245e-05, "loss": 0.0639, "step": 9000 }, { "epoch": 0.3046079989181514, "grad_norm": 0.30429595708847046, "learning_rate": 9.956123855369346e-05, "loss": 0.0659, "step": 9010 }, { "epoch": 0.3049460766084046, "grad_norm": 0.31851711869239807, "learning_rate": 9.955905016847196e-05, "loss": 0.0728, "step": 9020 }, { "epoch": 0.30528415429865785, "grad_norm": 0.19244419038295746, "learning_rate": 9.955685636354723e-05, "loss": 0.0668, "step": 9030 }, { "epoch": 0.30562223198891103, "grad_norm": 0.5585060715675354, "learning_rate": 9.95546571391592e-05, "loss": 0.0652, "step": 9040 }, { "epoch": 0.30596030967916427, "grad_norm": 0.3278455436229706, "learning_rate": 9.955245249554837e-05, "loss": 0.0755, "step": 9050 }, { "epoch": 0.3062983873694175, "grad_norm": 0.3307695984840393, "learning_rate": 9.955024243295582e-05, "loss": 0.0625, "step": 9060 }, { "epoch": 0.30663646505967074, "grad_norm": 0.24799568951129913, "learning_rate": 9.954802695162328e-05, "loss": 0.0682, "step": 9070 }, { "epoch": 0.3069745427499239, "grad_norm": 0.26202788949012756, "learning_rate": 9.954580605179302e-05, "loss": 0.0647, "step": 9080 }, { "epoch": 0.30731262044017715, "grad_norm": 0.22012364864349365, "learning_rate": 9.954357973370788e-05, "loss": 0.0673, "step": 9090 }, { "epoch": 0.3076506981304304, "grad_norm": 0.20389577746391296, "learning_rate": 9.954134799761135e-05, "loss": 0.0701, "step": 9100 }, { "epoch": 0.30798877582068357, "grad_norm": 0.2546350955963135, "learning_rate": 9.953911084374748e-05, "loss": 0.0618, "step": 9110 }, { "epoch": 0.3083268535109368, "grad_norm": 0.36288973689079285, "learning_rate": 9.953686827236093e-05, "loss": 0.066, "step": 9120 }, { "epoch": 0.30866493120119004, "grad_norm": 0.21290133893489838, "learning_rate": 9.953462028369695e-05, "loss": 0.0669, "step": 9130 }, { "epoch": 0.3090030088914433, "grad_norm": 0.2830338776111603, "learning_rate": 9.953236687800136e-05, "loss": 0.0599, "step": 9140 }, { "epoch": 0.30934108658169646, "grad_norm": 0.2698884606361389, "learning_rate": 9.95301080555206e-05, "loss": 0.0682, "step": 9150 }, { "epoch": 0.3096791642719497, "grad_norm": 0.35926011204719543, "learning_rate": 9.952784381650171e-05, "loss": 0.0658, "step": 9160 }, { "epoch": 0.3100172419622029, "grad_norm": 0.2444918155670166, "learning_rate": 9.952557416119226e-05, "loss": 0.0693, "step": 9170 }, { "epoch": 0.3103553196524561, "grad_norm": 0.285641074180603, "learning_rate": 9.95232990898405e-05, "loss": 0.0693, "step": 9180 }, { "epoch": 0.31069339734270934, "grad_norm": 0.29376640915870667, "learning_rate": 9.95210186026952e-05, "loss": 0.0668, "step": 9190 }, { "epoch": 0.3110314750329626, "grad_norm": 0.24714279174804688, "learning_rate": 9.951873270000576e-05, "loss": 0.0659, "step": 9200 }, { "epoch": 0.3113695527232158, "grad_norm": 0.34052780270576477, "learning_rate": 9.951644138202216e-05, "loss": 0.0721, "step": 9210 }, { "epoch": 0.311707630413469, "grad_norm": 0.3125946521759033, "learning_rate": 9.951414464899498e-05, "loss": 0.0661, "step": 9220 }, { "epoch": 0.31204570810372223, "grad_norm": 0.3640249967575073, "learning_rate": 9.951184250117538e-05, "loss": 0.0688, "step": 9230 }, { "epoch": 0.31238378579397547, "grad_norm": 0.30916324257850647, "learning_rate": 9.950953493881513e-05, "loss": 0.0702, "step": 9240 }, { "epoch": 0.3127218634842287, "grad_norm": 0.30838435888290405, "learning_rate": 9.950722196216658e-05, "loss": 0.0798, "step": 9250 }, { "epoch": 0.3130599411744819, "grad_norm": 0.35430973768234253, "learning_rate": 9.950490357148265e-05, "loss": 0.0772, "step": 9260 }, { "epoch": 0.3133980188647351, "grad_norm": 0.41650304198265076, "learning_rate": 9.950257976701692e-05, "loss": 0.0758, "step": 9270 }, { "epoch": 0.31373609655498835, "grad_norm": 0.2631267309188843, "learning_rate": 9.950025054902348e-05, "loss": 0.0681, "step": 9280 }, { "epoch": 0.31407417424524153, "grad_norm": 0.2726444900035858, "learning_rate": 9.949791591775706e-05, "loss": 0.0692, "step": 9290 }, { "epoch": 0.31441225193549477, "grad_norm": 0.24659763276576996, "learning_rate": 9.949557587347298e-05, "loss": 0.0678, "step": 9300 }, { "epoch": 0.314750329625748, "grad_norm": 0.3714849054813385, "learning_rate": 9.949323041642713e-05, "loss": 0.0703, "step": 9310 }, { "epoch": 0.31508840731600124, "grad_norm": 0.26766183972358704, "learning_rate": 9.949087954687602e-05, "loss": 0.0648, "step": 9320 }, { "epoch": 0.3154264850062544, "grad_norm": 0.2589639127254486, "learning_rate": 9.948852326507672e-05, "loss": 0.0705, "step": 9330 }, { "epoch": 0.31576456269650766, "grad_norm": 0.376015841960907, "learning_rate": 9.948616157128694e-05, "loss": 0.0699, "step": 9340 }, { "epoch": 0.3161026403867609, "grad_norm": 0.627557098865509, "learning_rate": 9.948379446576493e-05, "loss": 0.0657, "step": 9350 }, { "epoch": 0.31644071807701407, "grad_norm": 0.25716444849967957, "learning_rate": 9.948142194876952e-05, "loss": 0.068, "step": 9360 }, { "epoch": 0.3167787957672673, "grad_norm": 0.35268718004226685, "learning_rate": 9.947904402056024e-05, "loss": 0.0716, "step": 9370 }, { "epoch": 0.31711687345752054, "grad_norm": 0.3706202208995819, "learning_rate": 9.947666068139708e-05, "loss": 0.0617, "step": 9380 }, { "epoch": 0.3174549511477738, "grad_norm": 0.217122882604599, "learning_rate": 9.947427193154071e-05, "loss": 0.0667, "step": 9390 }, { "epoch": 0.31779302883802696, "grad_norm": 0.3720645606517792, "learning_rate": 9.947187777125233e-05, "loss": 0.0696, "step": 9400 }, { "epoch": 0.3181311065282802, "grad_norm": 0.25597602128982544, "learning_rate": 9.946947820079377e-05, "loss": 0.0662, "step": 9410 }, { "epoch": 0.31846918421853343, "grad_norm": 0.18265794217586517, "learning_rate": 9.946707322042747e-05, "loss": 0.0609, "step": 9420 }, { "epoch": 0.31880726190878667, "grad_norm": 0.2218511998653412, "learning_rate": 9.94646628304164e-05, "loss": 0.062, "step": 9430 }, { "epoch": 0.31914533959903985, "grad_norm": 0.2530517578125, "learning_rate": 9.946224703102418e-05, "loss": 0.0615, "step": 9440 }, { "epoch": 0.3194834172892931, "grad_norm": 0.2940152585506439, "learning_rate": 9.945982582251498e-05, "loss": 0.0736, "step": 9450 }, { "epoch": 0.3198214949795463, "grad_norm": 0.20854493975639343, "learning_rate": 9.94573992051536e-05, "loss": 0.0668, "step": 9460 }, { "epoch": 0.3201595726697995, "grad_norm": 0.2556966245174408, "learning_rate": 9.94549671792054e-05, "loss": 0.0669, "step": 9470 }, { "epoch": 0.32049765036005273, "grad_norm": 0.20896096527576447, "learning_rate": 9.945252974493635e-05, "loss": 0.0658, "step": 9480 }, { "epoch": 0.32083572805030597, "grad_norm": 0.24719908833503723, "learning_rate": 9.9450086902613e-05, "loss": 0.065, "step": 9490 }, { "epoch": 0.3211738057405592, "grad_norm": 0.34286850690841675, "learning_rate": 9.944763865250248e-05, "loss": 0.0662, "step": 9500 }, { "epoch": 0.3215118834308124, "grad_norm": 0.4009264409542084, "learning_rate": 9.944518499487254e-05, "loss": 0.0737, "step": 9510 }, { "epoch": 0.3218499611210656, "grad_norm": 0.2889411151409149, "learning_rate": 9.944272592999151e-05, "loss": 0.0625, "step": 9520 }, { "epoch": 0.32218803881131886, "grad_norm": 0.2866365313529968, "learning_rate": 9.94402614581283e-05, "loss": 0.0703, "step": 9530 }, { "epoch": 0.32252611650157204, "grad_norm": 0.46227607131004333, "learning_rate": 9.943779157955244e-05, "loss": 0.0668, "step": 9540 }, { "epoch": 0.32286419419182527, "grad_norm": 0.3566271662712097, "learning_rate": 9.943531629453403e-05, "loss": 0.0681, "step": 9550 }, { "epoch": 0.3232022718820785, "grad_norm": 0.2085709124803543, "learning_rate": 9.943283560334375e-05, "loss": 0.0654, "step": 9560 }, { "epoch": 0.32354034957233174, "grad_norm": 0.2795707881450653, "learning_rate": 9.943034950625288e-05, "loss": 0.0627, "step": 9570 }, { "epoch": 0.3238784272625849, "grad_norm": 0.4976660907268524, "learning_rate": 9.942785800353332e-05, "loss": 0.0708, "step": 9580 }, { "epoch": 0.32421650495283816, "grad_norm": 0.25405043363571167, "learning_rate": 9.942536109545751e-05, "loss": 0.0705, "step": 9590 }, { "epoch": 0.3245545826430914, "grad_norm": 0.3206310272216797, "learning_rate": 9.942285878229853e-05, "loss": 0.0654, "step": 9600 }, { "epoch": 0.32489266033334463, "grad_norm": 0.2544882297515869, "learning_rate": 9.942035106433001e-05, "loss": 0.067, "step": 9610 }, { "epoch": 0.3252307380235978, "grad_norm": 0.16294042766094208, "learning_rate": 9.94178379418262e-05, "loss": 0.0594, "step": 9620 }, { "epoch": 0.32556881571385105, "grad_norm": 0.22676672041416168, "learning_rate": 9.941531941506194e-05, "loss": 0.0704, "step": 9630 }, { "epoch": 0.3259068934041043, "grad_norm": 0.17555323243141174, "learning_rate": 9.941279548431263e-05, "loss": 0.0623, "step": 9640 }, { "epoch": 0.32624497109435746, "grad_norm": 0.3828623294830322, "learning_rate": 9.941026614985431e-05, "loss": 0.0703, "step": 9650 }, { "epoch": 0.3265830487846107, "grad_norm": 0.26067548990249634, "learning_rate": 9.940773141196357e-05, "loss": 0.0658, "step": 9660 }, { "epoch": 0.32692112647486393, "grad_norm": 0.54731285572052, "learning_rate": 9.94051912709176e-05, "loss": 0.0716, "step": 9670 }, { "epoch": 0.32725920416511717, "grad_norm": 0.32888010144233704, "learning_rate": 9.940264572699421e-05, "loss": 0.0691, "step": 9680 }, { "epoch": 0.32759728185537035, "grad_norm": 0.27305108308792114, "learning_rate": 9.940009478047174e-05, "loss": 0.0684, "step": 9690 }, { "epoch": 0.3279353595456236, "grad_norm": 0.2779880166053772, "learning_rate": 9.939753843162918e-05, "loss": 0.068, "step": 9700 }, { "epoch": 0.3282734372358768, "grad_norm": 0.20178762078285217, "learning_rate": 9.939497668074609e-05, "loss": 0.066, "step": 9710 }, { "epoch": 0.32861151492613, "grad_norm": 0.30610203742980957, "learning_rate": 9.93924095281026e-05, "loss": 0.0685, "step": 9720 }, { "epoch": 0.32894959261638324, "grad_norm": 0.2685268223285675, "learning_rate": 9.938983697397948e-05, "loss": 0.0723, "step": 9730 }, { "epoch": 0.32928767030663647, "grad_norm": 0.2895526885986328, "learning_rate": 9.938725901865805e-05, "loss": 0.0668, "step": 9740 }, { "epoch": 0.3296257479968897, "grad_norm": 0.25326773524284363, "learning_rate": 9.93846756624202e-05, "loss": 0.0679, "step": 9750 }, { "epoch": 0.3299638256871429, "grad_norm": 0.2011786848306656, "learning_rate": 9.938208690554849e-05, "loss": 0.0717, "step": 9760 }, { "epoch": 0.3303019033773961, "grad_norm": 0.32082369923591614, "learning_rate": 9.9379492748326e-05, "loss": 0.0684, "step": 9770 }, { "epoch": 0.33063998106764936, "grad_norm": 0.30796417593955994, "learning_rate": 9.937689319103641e-05, "loss": 0.0697, "step": 9780 }, { "epoch": 0.3309780587579026, "grad_norm": 0.2760114073753357, "learning_rate": 9.937428823396404e-05, "loss": 0.0707, "step": 9790 }, { "epoch": 0.3313161364481558, "grad_norm": 0.22138498723506927, "learning_rate": 9.937167787739372e-05, "loss": 0.0665, "step": 9800 }, { "epoch": 0.331654214138409, "grad_norm": 0.2658599615097046, "learning_rate": 9.936906212161095e-05, "loss": 0.0641, "step": 9810 }, { "epoch": 0.33199229182866224, "grad_norm": 0.2546182870864868, "learning_rate": 9.936644096690176e-05, "loss": 0.062, "step": 9820 }, { "epoch": 0.3323303695189154, "grad_norm": 0.24399778246879578, "learning_rate": 9.936381441355282e-05, "loss": 0.0569, "step": 9830 }, { "epoch": 0.33266844720916866, "grad_norm": 0.22657454013824463, "learning_rate": 9.936118246185136e-05, "loss": 0.0628, "step": 9840 }, { "epoch": 0.3330065248994219, "grad_norm": 0.22495992481708527, "learning_rate": 9.935854511208518e-05, "loss": 0.0643, "step": 9850 }, { "epoch": 0.33334460258967513, "grad_norm": 0.3779880106449127, "learning_rate": 9.935590236454272e-05, "loss": 0.0668, "step": 9860 }, { "epoch": 0.3336826802799283, "grad_norm": 0.335381418466568, "learning_rate": 9.935325421951298e-05, "loss": 0.0692, "step": 9870 }, { "epoch": 0.33402075797018155, "grad_norm": 0.42830905318260193, "learning_rate": 9.935060067728557e-05, "loss": 0.0653, "step": 9880 }, { "epoch": 0.3343588356604348, "grad_norm": 0.2719191610813141, "learning_rate": 9.934794173815067e-05, "loss": 0.0721, "step": 9890 }, { "epoch": 0.33469691335068796, "grad_norm": 0.5790141224861145, "learning_rate": 9.934527740239906e-05, "loss": 0.076, "step": 9900 }, { "epoch": 0.3350349910409412, "grad_norm": 0.2484777420759201, "learning_rate": 9.934260767032209e-05, "loss": 0.0661, "step": 9910 }, { "epoch": 0.33537306873119443, "grad_norm": 0.4684801697731018, "learning_rate": 9.933993254221172e-05, "loss": 0.0736, "step": 9920 }, { "epoch": 0.33571114642144767, "grad_norm": 0.2938686013221741, "learning_rate": 9.933725201836053e-05, "loss": 0.0735, "step": 9930 }, { "epoch": 0.33604922411170085, "grad_norm": 0.37747395038604736, "learning_rate": 9.933456609906162e-05, "loss": 0.0682, "step": 9940 }, { "epoch": 0.3363873018019541, "grad_norm": 0.2423977255821228, "learning_rate": 9.933187478460875e-05, "loss": 0.0681, "step": 9950 }, { "epoch": 0.3367253794922073, "grad_norm": 0.19334401190280914, "learning_rate": 9.93291780752962e-05, "loss": 0.0643, "step": 9960 }, { "epoch": 0.3370634571824605, "grad_norm": 0.19282039999961853, "learning_rate": 9.932647597141893e-05, "loss": 0.0673, "step": 9970 }, { "epoch": 0.33740153487271374, "grad_norm": 0.3805603086948395, "learning_rate": 9.932376847327239e-05, "loss": 0.0663, "step": 9980 }, { "epoch": 0.337739612562967, "grad_norm": 0.19717508554458618, "learning_rate": 9.932105558115268e-05, "loss": 0.059, "step": 9990 }, { "epoch": 0.3380776902532202, "grad_norm": 0.4182654321193695, "learning_rate": 9.931833729535651e-05, "loss": 0.0756, "step": 10000 }, { "epoch": 0.3384157679434734, "grad_norm": 0.4178314805030823, "learning_rate": 9.931561361618111e-05, "loss": 0.0698, "step": 10010 }, { "epoch": 0.3387538456337266, "grad_norm": 0.4297015070915222, "learning_rate": 9.931288454392435e-05, "loss": 0.0661, "step": 10020 }, { "epoch": 0.33909192332397986, "grad_norm": 0.2583359479904175, "learning_rate": 9.931015007888467e-05, "loss": 0.066, "step": 10030 }, { "epoch": 0.3394300010142331, "grad_norm": 0.21202024817466736, "learning_rate": 9.930741022136112e-05, "loss": 0.0641, "step": 10040 }, { "epoch": 0.3397680787044863, "grad_norm": 0.22018177807331085, "learning_rate": 9.930466497165333e-05, "loss": 0.0647, "step": 10050 }, { "epoch": 0.3401061563947395, "grad_norm": 0.24853487312793732, "learning_rate": 9.93019143300615e-05, "loss": 0.0733, "step": 10060 }, { "epoch": 0.34044423408499275, "grad_norm": 0.3426136076450348, "learning_rate": 9.929915829688644e-05, "loss": 0.0626, "step": 10070 }, { "epoch": 0.3407823117752459, "grad_norm": 0.22028203308582306, "learning_rate": 9.929639687242955e-05, "loss": 0.0711, "step": 10080 }, { "epoch": 0.34112038946549916, "grad_norm": 0.3124333918094635, "learning_rate": 9.929363005699281e-05, "loss": 0.0674, "step": 10090 }, { "epoch": 0.3414584671557524, "grad_norm": 0.2904198467731476, "learning_rate": 9.92908578508788e-05, "loss": 0.0673, "step": 10100 }, { "epoch": 0.34179654484600563, "grad_norm": 0.22305889427661896, "learning_rate": 9.928808025439069e-05, "loss": 0.0631, "step": 10110 }, { "epoch": 0.3421346225362588, "grad_norm": 0.25962600111961365, "learning_rate": 9.928529726783223e-05, "loss": 0.0696, "step": 10120 }, { "epoch": 0.34247270022651205, "grad_norm": 0.21835380792617798, "learning_rate": 9.928250889150774e-05, "loss": 0.0685, "step": 10130 }, { "epoch": 0.3428107779167653, "grad_norm": 0.23592329025268555, "learning_rate": 9.92797151257222e-05, "loss": 0.0614, "step": 10140 }, { "epoch": 0.34314885560701847, "grad_norm": 0.30474114418029785, "learning_rate": 9.927691597078108e-05, "loss": 0.0601, "step": 10150 }, { "epoch": 0.3434869332972717, "grad_norm": 0.5403532385826111, "learning_rate": 9.927411142699053e-05, "loss": 0.065, "step": 10160 }, { "epoch": 0.34382501098752494, "grad_norm": 0.3082452118396759, "learning_rate": 9.927130149465725e-05, "loss": 0.0678, "step": 10170 }, { "epoch": 0.3441630886777782, "grad_norm": 0.2466082125902176, "learning_rate": 9.92684861740885e-05, "loss": 0.0699, "step": 10180 }, { "epoch": 0.34450116636803135, "grad_norm": 0.39729392528533936, "learning_rate": 9.926566546559217e-05, "loss": 0.0596, "step": 10190 }, { "epoch": 0.3448392440582846, "grad_norm": 0.28792017698287964, "learning_rate": 9.926283936947673e-05, "loss": 0.0697, "step": 10200 }, { "epoch": 0.3451773217485378, "grad_norm": 0.26265233755111694, "learning_rate": 9.926000788605126e-05, "loss": 0.0663, "step": 10210 }, { "epoch": 0.34551539943879106, "grad_norm": 0.27492597699165344, "learning_rate": 9.92571710156254e-05, "loss": 0.0673, "step": 10220 }, { "epoch": 0.34585347712904424, "grad_norm": 0.20172841846942902, "learning_rate": 9.925432875850936e-05, "loss": 0.069, "step": 10230 }, { "epoch": 0.3461915548192975, "grad_norm": 0.2378712296485901, "learning_rate": 9.925148111501396e-05, "loss": 0.0656, "step": 10240 }, { "epoch": 0.3465296325095507, "grad_norm": 0.2186318039894104, "learning_rate": 9.924862808545066e-05, "loss": 0.0685, "step": 10250 }, { "epoch": 0.3468677101998039, "grad_norm": 0.24618859589099884, "learning_rate": 9.924576967013141e-05, "loss": 0.0733, "step": 10260 }, { "epoch": 0.3472057878900571, "grad_norm": 0.19118042290210724, "learning_rate": 9.924290586936887e-05, "loss": 0.0659, "step": 10270 }, { "epoch": 0.34754386558031036, "grad_norm": 0.2656804919242859, "learning_rate": 9.924003668347614e-05, "loss": 0.0656, "step": 10280 }, { "epoch": 0.3478819432705636, "grad_norm": 0.2512388825416565, "learning_rate": 9.923716211276704e-05, "loss": 0.0616, "step": 10290 }, { "epoch": 0.3482200209608168, "grad_norm": 0.448503702878952, "learning_rate": 9.923428215755594e-05, "loss": 0.0706, "step": 10300 }, { "epoch": 0.34855809865107, "grad_norm": 0.3173198699951172, "learning_rate": 9.923139681815775e-05, "loss": 0.0748, "step": 10310 }, { "epoch": 0.34889617634132325, "grad_norm": 0.2842460870742798, "learning_rate": 9.922850609488801e-05, "loss": 0.0634, "step": 10320 }, { "epoch": 0.34923425403157643, "grad_norm": 0.26409977674484253, "learning_rate": 9.922560998806287e-05, "loss": 0.0705, "step": 10330 }, { "epoch": 0.34957233172182967, "grad_norm": 0.3313313126564026, "learning_rate": 9.922270849799905e-05, "loss": 0.0636, "step": 10340 }, { "epoch": 0.3499104094120829, "grad_norm": 0.27127477526664734, "learning_rate": 9.92198016250138e-05, "loss": 0.0649, "step": 10350 }, { "epoch": 0.35024848710233614, "grad_norm": 0.17609821259975433, "learning_rate": 9.921688936942506e-05, "loss": 0.0721, "step": 10360 }, { "epoch": 0.3505865647925893, "grad_norm": 0.23657482862472534, "learning_rate": 9.921397173155129e-05, "loss": 0.0651, "step": 10370 }, { "epoch": 0.35092464248284255, "grad_norm": 0.3633139431476593, "learning_rate": 9.921104871171157e-05, "loss": 0.0641, "step": 10380 }, { "epoch": 0.3512627201730958, "grad_norm": 0.4363609552383423, "learning_rate": 9.920812031022554e-05, "loss": 0.064, "step": 10390 }, { "epoch": 0.351600797863349, "grad_norm": 0.2576075494289398, "learning_rate": 9.920518652741348e-05, "loss": 0.0648, "step": 10400 }, { "epoch": 0.3519388755536022, "grad_norm": 0.2897357940673828, "learning_rate": 9.920224736359618e-05, "loss": 0.0742, "step": 10410 }, { "epoch": 0.35227695324385544, "grad_norm": 0.23442433774471283, "learning_rate": 9.91993028190951e-05, "loss": 0.066, "step": 10420 }, { "epoch": 0.3526150309341087, "grad_norm": 0.2411980926990509, "learning_rate": 9.919635289423222e-05, "loss": 0.0631, "step": 10430 }, { "epoch": 0.35295310862436186, "grad_norm": 0.27827581763267517, "learning_rate": 9.919339758933015e-05, "loss": 0.0631, "step": 10440 }, { "epoch": 0.3532911863146151, "grad_norm": 0.3361443877220154, "learning_rate": 9.919043690471209e-05, "loss": 0.0692, "step": 10450 }, { "epoch": 0.3536292640048683, "grad_norm": 0.2671384811401367, "learning_rate": 9.91874708407018e-05, "loss": 0.0757, "step": 10460 }, { "epoch": 0.35396734169512156, "grad_norm": 0.25611215829849243, "learning_rate": 9.918449939762367e-05, "loss": 0.063, "step": 10470 }, { "epoch": 0.35430541938537474, "grad_norm": 0.38408979773521423, "learning_rate": 9.91815225758026e-05, "loss": 0.071, "step": 10480 }, { "epoch": 0.354643497075628, "grad_norm": 0.30271708965301514, "learning_rate": 9.917854037556419e-05, "loss": 0.0636, "step": 10490 }, { "epoch": 0.3549815747658812, "grad_norm": 0.20465339720249176, "learning_rate": 9.917555279723454e-05, "loss": 0.0671, "step": 10500 }, { "epoch": 0.3553196524561344, "grad_norm": 0.1898021548986435, "learning_rate": 9.917255984114036e-05, "loss": 0.059, "step": 10510 }, { "epoch": 0.35565773014638763, "grad_norm": 0.31746092438697815, "learning_rate": 9.916956150760896e-05, "loss": 0.0676, "step": 10520 }, { "epoch": 0.35599580783664087, "grad_norm": 0.27967527508735657, "learning_rate": 9.916655779696826e-05, "loss": 0.0687, "step": 10530 }, { "epoch": 0.3563338855268941, "grad_norm": 0.22121426463127136, "learning_rate": 9.916354870954671e-05, "loss": 0.0649, "step": 10540 }, { "epoch": 0.3566719632171473, "grad_norm": 0.20852287113666534, "learning_rate": 9.91605342456734e-05, "loss": 0.0623, "step": 10550 }, { "epoch": 0.3570100409074005, "grad_norm": 0.19180309772491455, "learning_rate": 9.915751440567795e-05, "loss": 0.065, "step": 10560 }, { "epoch": 0.35734811859765375, "grad_norm": 0.23481303453445435, "learning_rate": 9.915448918989066e-05, "loss": 0.0671, "step": 10570 }, { "epoch": 0.357686196287907, "grad_norm": 0.25013333559036255, "learning_rate": 9.915145859864232e-05, "loss": 0.0655, "step": 10580 }, { "epoch": 0.35802427397816017, "grad_norm": 0.20276281237602234, "learning_rate": 9.914842263226437e-05, "loss": 0.0592, "step": 10590 }, { "epoch": 0.3583623516684134, "grad_norm": 0.3309014141559601, "learning_rate": 9.914538129108882e-05, "loss": 0.0691, "step": 10600 }, { "epoch": 0.35870042935866664, "grad_norm": 0.22842568159103394, "learning_rate": 9.914233457544825e-05, "loss": 0.0591, "step": 10610 }, { "epoch": 0.3590385070489198, "grad_norm": 0.261481374502182, "learning_rate": 9.913928248567586e-05, "loss": 0.0656, "step": 10620 }, { "epoch": 0.35937658473917306, "grad_norm": 0.24565444886684418, "learning_rate": 9.913622502210542e-05, "loss": 0.0623, "step": 10630 }, { "epoch": 0.3597146624294263, "grad_norm": 0.2565050423145294, "learning_rate": 9.913316218507128e-05, "loss": 0.0697, "step": 10640 }, { "epoch": 0.3600527401196795, "grad_norm": 0.35536935925483704, "learning_rate": 9.91300939749084e-05, "loss": 0.0621, "step": 10650 }, { "epoch": 0.3603908178099327, "grad_norm": 0.33121320605278015, "learning_rate": 9.91270203919523e-05, "loss": 0.0605, "step": 10660 }, { "epoch": 0.36072889550018594, "grad_norm": 0.2505170702934265, "learning_rate": 9.912394143653912e-05, "loss": 0.0769, "step": 10670 }, { "epoch": 0.3610669731904392, "grad_norm": 0.40925756096839905, "learning_rate": 9.912085710900555e-05, "loss": 0.0727, "step": 10680 }, { "epoch": 0.36140505088069236, "grad_norm": 0.3129495680332184, "learning_rate": 9.911776740968892e-05, "loss": 0.0703, "step": 10690 }, { "epoch": 0.3617431285709456, "grad_norm": 0.31231001019477844, "learning_rate": 9.911467233892709e-05, "loss": 0.0626, "step": 10700 }, { "epoch": 0.36208120626119883, "grad_norm": 0.4341208338737488, "learning_rate": 9.911157189705853e-05, "loss": 0.0656, "step": 10710 }, { "epoch": 0.36241928395145206, "grad_norm": 0.41050806641578674, "learning_rate": 9.910846608442229e-05, "loss": 0.0645, "step": 10720 }, { "epoch": 0.36275736164170524, "grad_norm": 0.2184297889471054, "learning_rate": 9.910535490135805e-05, "loss": 0.0639, "step": 10730 }, { "epoch": 0.3630954393319585, "grad_norm": 0.24957038462162018, "learning_rate": 9.910223834820603e-05, "loss": 0.0651, "step": 10740 }, { "epoch": 0.3634335170222117, "grad_norm": 0.22088199853897095, "learning_rate": 9.909911642530703e-05, "loss": 0.0687, "step": 10750 }, { "epoch": 0.36377159471246495, "grad_norm": 0.2785934507846832, "learning_rate": 9.909598913300249e-05, "loss": 0.0698, "step": 10760 }, { "epoch": 0.36410967240271813, "grad_norm": 0.16997328400611877, "learning_rate": 9.909285647163438e-05, "loss": 0.0619, "step": 10770 }, { "epoch": 0.36444775009297137, "grad_norm": 0.3759918808937073, "learning_rate": 9.908971844154531e-05, "loss": 0.0644, "step": 10780 }, { "epoch": 0.3647858277832246, "grad_norm": 0.21590539813041687, "learning_rate": 9.908657504307843e-05, "loss": 0.0665, "step": 10790 }, { "epoch": 0.3651239054734778, "grad_norm": 0.25994059443473816, "learning_rate": 9.908342627657751e-05, "loss": 0.0647, "step": 10800 }, { "epoch": 0.365461983163731, "grad_norm": 0.2543034255504608, "learning_rate": 9.908027214238689e-05, "loss": 0.0629, "step": 10810 }, { "epoch": 0.36580006085398425, "grad_norm": 0.24501121044158936, "learning_rate": 9.90771126408515e-05, "loss": 0.0653, "step": 10820 }, { "epoch": 0.3661381385442375, "grad_norm": 0.2037029266357422, "learning_rate": 9.907394777231685e-05, "loss": 0.0661, "step": 10830 }, { "epoch": 0.36647621623449067, "grad_norm": 0.26200661063194275, "learning_rate": 9.907077753712905e-05, "loss": 0.0635, "step": 10840 }, { "epoch": 0.3668142939247439, "grad_norm": 0.21207310259342194, "learning_rate": 9.906760193563482e-05, "loss": 0.0701, "step": 10850 }, { "epoch": 0.36715237161499714, "grad_norm": 0.25575530529022217, "learning_rate": 9.906442096818139e-05, "loss": 0.0608, "step": 10860 }, { "epoch": 0.3674904493052503, "grad_norm": 0.21460814774036407, "learning_rate": 9.906123463511665e-05, "loss": 0.0692, "step": 10870 }, { "epoch": 0.36782852699550356, "grad_norm": 0.2929055094718933, "learning_rate": 9.905804293678907e-05, "loss": 0.0699, "step": 10880 }, { "epoch": 0.3681666046857568, "grad_norm": 0.272061824798584, "learning_rate": 9.905484587354766e-05, "loss": 0.061, "step": 10890 }, { "epoch": 0.36850468237601003, "grad_norm": 0.2579471468925476, "learning_rate": 9.905164344574205e-05, "loss": 0.064, "step": 10900 }, { "epoch": 0.3688427600662632, "grad_norm": 0.21158096194267273, "learning_rate": 9.904843565372248e-05, "loss": 0.0627, "step": 10910 }, { "epoch": 0.36918083775651644, "grad_norm": 0.240170419216156, "learning_rate": 9.904522249783972e-05, "loss": 0.0564, "step": 10920 }, { "epoch": 0.3695189154467697, "grad_norm": 0.2690487802028656, "learning_rate": 9.904200397844517e-05, "loss": 0.0631, "step": 10930 }, { "epoch": 0.36985699313702286, "grad_norm": 0.3092730641365051, "learning_rate": 9.903878009589078e-05, "loss": 0.0641, "step": 10940 }, { "epoch": 0.3701950708272761, "grad_norm": 0.3107215166091919, "learning_rate": 9.903555085052915e-05, "loss": 0.0602, "step": 10950 }, { "epoch": 0.37053314851752933, "grad_norm": 0.2123943567276001, "learning_rate": 9.903231624271338e-05, "loss": 0.0608, "step": 10960 }, { "epoch": 0.37087122620778257, "grad_norm": 0.2379707545042038, "learning_rate": 9.902907627279724e-05, "loss": 0.0618, "step": 10970 }, { "epoch": 0.37120930389803575, "grad_norm": 0.19018316268920898, "learning_rate": 9.902583094113504e-05, "loss": 0.0685, "step": 10980 }, { "epoch": 0.371547381588289, "grad_norm": 0.329655259847641, "learning_rate": 9.902258024808168e-05, "loss": 0.0716, "step": 10990 }, { "epoch": 0.3718854592785422, "grad_norm": 0.2899000346660614, "learning_rate": 9.901932419399264e-05, "loss": 0.0704, "step": 11000 }, { "epoch": 0.37222353696879545, "grad_norm": 0.2837768495082855, "learning_rate": 9.9016062779224e-05, "loss": 0.0662, "step": 11010 }, { "epoch": 0.37256161465904863, "grad_norm": 0.33574751019477844, "learning_rate": 9.901279600413242e-05, "loss": 0.0696, "step": 11020 }, { "epoch": 0.37289969234930187, "grad_norm": 0.24153515696525574, "learning_rate": 9.900952386907518e-05, "loss": 0.0689, "step": 11030 }, { "epoch": 0.3732377700395551, "grad_norm": 0.16926966607570648, "learning_rate": 9.90062463744101e-05, "loss": 0.0633, "step": 11040 }, { "epoch": 0.3735758477298083, "grad_norm": 0.20609652996063232, "learning_rate": 9.900296352049558e-05, "loss": 0.0648, "step": 11050 }, { "epoch": 0.3739139254200615, "grad_norm": 0.360872745513916, "learning_rate": 9.899967530769065e-05, "loss": 0.0692, "step": 11060 }, { "epoch": 0.37425200311031476, "grad_norm": 0.3331858217716217, "learning_rate": 9.899638173635489e-05, "loss": 0.0658, "step": 11070 }, { "epoch": 0.374590080800568, "grad_norm": 0.3059482276439667, "learning_rate": 9.899308280684849e-05, "loss": 0.0601, "step": 11080 }, { "epoch": 0.3749281584908212, "grad_norm": 0.2681943476200104, "learning_rate": 9.898977851953222e-05, "loss": 0.0632, "step": 11090 }, { "epoch": 0.3752662361810744, "grad_norm": 0.19332417845726013, "learning_rate": 9.898646887476741e-05, "loss": 0.0611, "step": 11100 }, { "epoch": 0.37560431387132764, "grad_norm": 0.3427911400794983, "learning_rate": 9.898315387291603e-05, "loss": 0.0645, "step": 11110 }, { "epoch": 0.3759423915615808, "grad_norm": 0.23623421788215637, "learning_rate": 9.89798335143406e-05, "loss": 0.0668, "step": 11120 }, { "epoch": 0.37628046925183406, "grad_norm": 0.25491073727607727, "learning_rate": 9.897650779940419e-05, "loss": 0.0652, "step": 11130 }, { "epoch": 0.3766185469420873, "grad_norm": 0.2781706154346466, "learning_rate": 9.897317672847054e-05, "loss": 0.0624, "step": 11140 }, { "epoch": 0.37695662463234053, "grad_norm": 0.18792177736759186, "learning_rate": 9.89698403019039e-05, "loss": 0.058, "step": 11150 }, { "epoch": 0.3772947023225937, "grad_norm": 0.223466157913208, "learning_rate": 9.896649852006917e-05, "loss": 0.0674, "step": 11160 }, { "epoch": 0.37763278001284695, "grad_norm": 0.2416124939918518, "learning_rate": 9.896315138333177e-05, "loss": 0.0691, "step": 11170 }, { "epoch": 0.3779708577031002, "grad_norm": 0.22435049712657928, "learning_rate": 9.895979889205774e-05, "loss": 0.064, "step": 11180 }, { "epoch": 0.3783089353933534, "grad_norm": 0.21478870511054993, "learning_rate": 9.895644104661372e-05, "loss": 0.0622, "step": 11190 }, { "epoch": 0.3786470130836066, "grad_norm": 0.2396262288093567, "learning_rate": 9.895307784736691e-05, "loss": 0.0679, "step": 11200 }, { "epoch": 0.37898509077385983, "grad_norm": 0.277463436126709, "learning_rate": 9.894970929468512e-05, "loss": 0.0628, "step": 11210 }, { "epoch": 0.37932316846411307, "grad_norm": 0.19877520203590393, "learning_rate": 9.89463353889367e-05, "loss": 0.0646, "step": 11220 }, { "epoch": 0.37966124615436625, "grad_norm": 0.19154761731624603, "learning_rate": 9.894295613049065e-05, "loss": 0.0673, "step": 11230 }, { "epoch": 0.3799993238446195, "grad_norm": 0.26708167791366577, "learning_rate": 9.893957151971649e-05, "loss": 0.0696, "step": 11240 }, { "epoch": 0.3803374015348727, "grad_norm": 0.25653472542762756, "learning_rate": 9.893618155698436e-05, "loss": 0.0661, "step": 11250 }, { "epoch": 0.38067547922512596, "grad_norm": 0.2568112313747406, "learning_rate": 9.8932786242665e-05, "loss": 0.0592, "step": 11260 }, { "epoch": 0.38101355691537914, "grad_norm": 0.35422733426094055, "learning_rate": 9.89293855771297e-05, "loss": 0.0689, "step": 11270 }, { "epoch": 0.3813516346056324, "grad_norm": 0.2804291248321533, "learning_rate": 9.892597956075036e-05, "loss": 0.0646, "step": 11280 }, { "epoch": 0.3816897122958856, "grad_norm": 0.3174343705177307, "learning_rate": 9.892256819389947e-05, "loss": 0.0712, "step": 11290 }, { "epoch": 0.3820277899861388, "grad_norm": 0.36576494574546814, "learning_rate": 9.891915147695006e-05, "loss": 0.0668, "step": 11300 }, { "epoch": 0.382365867676392, "grad_norm": 0.2463778406381607, "learning_rate": 9.891572941027577e-05, "loss": 0.0699, "step": 11310 }, { "epoch": 0.38270394536664526, "grad_norm": 0.2517305612564087, "learning_rate": 9.89123019942509e-05, "loss": 0.0699, "step": 11320 }, { "epoch": 0.3830420230568985, "grad_norm": 0.2663063704967499, "learning_rate": 9.89088692292502e-05, "loss": 0.0618, "step": 11330 }, { "epoch": 0.3833801007471517, "grad_norm": 0.22920794785022736, "learning_rate": 9.89054311156491e-05, "loss": 0.0635, "step": 11340 }, { "epoch": 0.3837181784374049, "grad_norm": 0.18819403648376465, "learning_rate": 9.890198765382357e-05, "loss": 0.0591, "step": 11350 }, { "epoch": 0.38405625612765815, "grad_norm": 0.2922394275665283, "learning_rate": 9.889853884415021e-05, "loss": 0.0695, "step": 11360 }, { "epoch": 0.3843943338179114, "grad_norm": 0.20644819736480713, "learning_rate": 9.889508468700614e-05, "loss": 0.0593, "step": 11370 }, { "epoch": 0.38473241150816456, "grad_norm": 0.23944637179374695, "learning_rate": 9.889162518276915e-05, "loss": 0.0655, "step": 11380 }, { "epoch": 0.3850704891984178, "grad_norm": 0.3435584008693695, "learning_rate": 9.888816033181752e-05, "loss": 0.0726, "step": 11390 }, { "epoch": 0.38540856688867103, "grad_norm": 0.23808497190475464, "learning_rate": 9.888469013453018e-05, "loss": 0.0725, "step": 11400 }, { "epoch": 0.3857466445789242, "grad_norm": 0.22190339863300323, "learning_rate": 9.888121459128663e-05, "loss": 0.0747, "step": 11410 }, { "epoch": 0.38608472226917745, "grad_norm": 0.2819700837135315, "learning_rate": 9.887773370246693e-05, "loss": 0.0587, "step": 11420 }, { "epoch": 0.3864227999594307, "grad_norm": 0.2538506090641022, "learning_rate": 9.887424746845177e-05, "loss": 0.0658, "step": 11430 }, { "epoch": 0.3867608776496839, "grad_norm": 0.47724926471710205, "learning_rate": 9.887075588962239e-05, "loss": 0.0673, "step": 11440 }, { "epoch": 0.3870989553399371, "grad_norm": 0.3744862377643585, "learning_rate": 9.88672589663606e-05, "loss": 0.0589, "step": 11450 }, { "epoch": 0.38743703303019034, "grad_norm": 0.2573641538619995, "learning_rate": 9.886375669904886e-05, "loss": 0.0685, "step": 11460 }, { "epoch": 0.38777511072044357, "grad_norm": 0.2908255159854889, "learning_rate": 9.886024908807014e-05, "loss": 0.0642, "step": 11470 }, { "epoch": 0.38811318841069675, "grad_norm": 0.4010554552078247, "learning_rate": 9.885673613380806e-05, "loss": 0.0642, "step": 11480 }, { "epoch": 0.38845126610095, "grad_norm": 0.2305164784193039, "learning_rate": 9.885321783664676e-05, "loss": 0.0624, "step": 11490 }, { "epoch": 0.3887893437912032, "grad_norm": 0.366383820772171, "learning_rate": 9.884969419697101e-05, "loss": 0.0666, "step": 11500 }, { "epoch": 0.38912742148145646, "grad_norm": 0.3550264239311218, "learning_rate": 9.884616521516614e-05, "loss": 0.0622, "step": 11510 }, { "epoch": 0.38946549917170964, "grad_norm": 0.2912687063217163, "learning_rate": 9.88426308916181e-05, "loss": 0.0609, "step": 11520 }, { "epoch": 0.3898035768619629, "grad_norm": 0.36751261353492737, "learning_rate": 9.883909122671335e-05, "loss": 0.067, "step": 11530 }, { "epoch": 0.3901416545522161, "grad_norm": 0.2704203426837921, "learning_rate": 9.883554622083904e-05, "loss": 0.0663, "step": 11540 }, { "epoch": 0.39047973224246935, "grad_norm": 0.209276482462883, "learning_rate": 9.88319958743828e-05, "loss": 0.0682, "step": 11550 }, { "epoch": 0.3908178099327225, "grad_norm": 0.297916442155838, "learning_rate": 9.882844018773291e-05, "loss": 0.0661, "step": 11560 }, { "epoch": 0.39115588762297576, "grad_norm": 0.2719772756099701, "learning_rate": 9.882487916127823e-05, "loss": 0.0669, "step": 11570 }, { "epoch": 0.391493965313229, "grad_norm": 0.30649250745773315, "learning_rate": 9.882131279540815e-05, "loss": 0.0682, "step": 11580 }, { "epoch": 0.3918320430034822, "grad_norm": 0.30287766456604004, "learning_rate": 9.881774109051271e-05, "loss": 0.0639, "step": 11590 }, { "epoch": 0.3921701206937354, "grad_norm": 0.2833282947540283, "learning_rate": 9.881416404698252e-05, "loss": 0.0673, "step": 11600 }, { "epoch": 0.39250819838398865, "grad_norm": 0.38232308626174927, "learning_rate": 9.881058166520873e-05, "loss": 0.0707, "step": 11610 }, { "epoch": 0.3928462760742419, "grad_norm": 0.38649052381515503, "learning_rate": 9.880699394558311e-05, "loss": 0.0716, "step": 11620 }, { "epoch": 0.39318435376449506, "grad_norm": 0.2190086543560028, "learning_rate": 9.880340088849801e-05, "loss": 0.0584, "step": 11630 }, { "epoch": 0.3935224314547483, "grad_norm": 0.31795167922973633, "learning_rate": 9.879980249434637e-05, "loss": 0.0722, "step": 11640 }, { "epoch": 0.39386050914500154, "grad_norm": 0.21717602014541626, "learning_rate": 9.879619876352168e-05, "loss": 0.0603, "step": 11650 }, { "epoch": 0.3941985868352547, "grad_norm": 0.2778547704219818, "learning_rate": 9.879258969641809e-05, "loss": 0.0652, "step": 11660 }, { "epoch": 0.39453666452550795, "grad_norm": 0.2672930359840393, "learning_rate": 9.878897529343023e-05, "loss": 0.0673, "step": 11670 }, { "epoch": 0.3948747422157612, "grad_norm": 0.24393804371356964, "learning_rate": 9.878535555495338e-05, "loss": 0.0679, "step": 11680 }, { "epoch": 0.3952128199060144, "grad_norm": 0.22602489590644836, "learning_rate": 9.87817304813834e-05, "loss": 0.0639, "step": 11690 }, { "epoch": 0.3955508975962676, "grad_norm": 0.32263368368148804, "learning_rate": 9.877810007311671e-05, "loss": 0.0715, "step": 11700 }, { "epoch": 0.39588897528652084, "grad_norm": 0.21765218675136566, "learning_rate": 9.877446433055035e-05, "loss": 0.0603, "step": 11710 }, { "epoch": 0.3962270529767741, "grad_norm": 0.29271644353866577, "learning_rate": 9.877082325408191e-05, "loss": 0.0598, "step": 11720 }, { "epoch": 0.3965651306670273, "grad_norm": 0.31264355778694153, "learning_rate": 9.876717684410954e-05, "loss": 0.0703, "step": 11730 }, { "epoch": 0.3969032083572805, "grad_norm": 0.2027726024389267, "learning_rate": 9.876352510103204e-05, "loss": 0.0687, "step": 11740 }, { "epoch": 0.3972412860475337, "grad_norm": 0.3029574453830719, "learning_rate": 9.875986802524875e-05, "loss": 0.0659, "step": 11750 }, { "epoch": 0.39757936373778696, "grad_norm": 0.24753832817077637, "learning_rate": 9.87562056171596e-05, "loss": 0.0693, "step": 11760 }, { "epoch": 0.39791744142804014, "grad_norm": 0.27864429354667664, "learning_rate": 9.875253787716511e-05, "loss": 0.068, "step": 11770 }, { "epoch": 0.3982555191182934, "grad_norm": 0.24487286806106567, "learning_rate": 9.874886480566637e-05, "loss": 0.0667, "step": 11780 }, { "epoch": 0.3985935968085466, "grad_norm": 0.23065045475959778, "learning_rate": 9.874518640306507e-05, "loss": 0.0645, "step": 11790 }, { "epoch": 0.39893167449879985, "grad_norm": 0.1949985772371292, "learning_rate": 9.874150266976347e-05, "loss": 0.062, "step": 11800 }, { "epoch": 0.39926975218905303, "grad_norm": 0.2276434600353241, "learning_rate": 9.873781360616443e-05, "loss": 0.0687, "step": 11810 }, { "epoch": 0.39960782987930626, "grad_norm": 0.23668965697288513, "learning_rate": 9.873411921267137e-05, "loss": 0.0619, "step": 11820 }, { "epoch": 0.3999459075695595, "grad_norm": 0.27957528829574585, "learning_rate": 9.873041948968829e-05, "loss": 0.0693, "step": 11830 }, { "epoch": 0.4002839852598127, "grad_norm": 0.20571576058864594, "learning_rate": 9.872671443761981e-05, "loss": 0.0616, "step": 11840 }, { "epoch": 0.4006220629500659, "grad_norm": 0.34532997012138367, "learning_rate": 9.872300405687109e-05, "loss": 0.0591, "step": 11850 }, { "epoch": 0.40096014064031915, "grad_norm": 0.1799800544977188, "learning_rate": 9.871928834784792e-05, "loss": 0.0595, "step": 11860 }, { "epoch": 0.4012982183305724, "grad_norm": 0.30291658639907837, "learning_rate": 9.871556731095661e-05, "loss": 0.0679, "step": 11870 }, { "epoch": 0.40163629602082557, "grad_norm": 0.19407640397548676, "learning_rate": 9.871184094660411e-05, "loss": 0.0612, "step": 11880 }, { "epoch": 0.4019743737110788, "grad_norm": 0.28767913579940796, "learning_rate": 9.870810925519791e-05, "loss": 0.0671, "step": 11890 }, { "epoch": 0.40231245140133204, "grad_norm": 0.3440547585487366, "learning_rate": 9.870437223714612e-05, "loss": 0.0647, "step": 11900 }, { "epoch": 0.4026505290915852, "grad_norm": 0.3100317418575287, "learning_rate": 9.87006298928574e-05, "loss": 0.0684, "step": 11910 }, { "epoch": 0.40298860678183845, "grad_norm": 0.22189277410507202, "learning_rate": 9.869688222274103e-05, "loss": 0.0646, "step": 11920 }, { "epoch": 0.4033266844720917, "grad_norm": 0.3080390989780426, "learning_rate": 9.869312922720681e-05, "loss": 0.0728, "step": 11930 }, { "epoch": 0.4036647621623449, "grad_norm": 0.2163276970386505, "learning_rate": 9.868937090666521e-05, "loss": 0.0651, "step": 11940 }, { "epoch": 0.4040028398525981, "grad_norm": 0.4609780013561249, "learning_rate": 9.86856072615272e-05, "loss": 0.0631, "step": 11950 }, { "epoch": 0.40434091754285134, "grad_norm": 0.17945276200771332, "learning_rate": 9.868183829220438e-05, "loss": 0.0683, "step": 11960 }, { "epoch": 0.4046789952331046, "grad_norm": 0.21513532102108002, "learning_rate": 9.867806399910893e-05, "loss": 0.0652, "step": 11970 }, { "epoch": 0.4050170729233578, "grad_norm": 0.2575118839740753, "learning_rate": 9.867428438265356e-05, "loss": 0.0689, "step": 11980 }, { "epoch": 0.405355150613611, "grad_norm": 0.1673227846622467, "learning_rate": 9.867049944325165e-05, "loss": 0.0664, "step": 11990 }, { "epoch": 0.40569322830386423, "grad_norm": 0.22120174765586853, "learning_rate": 9.86667091813171e-05, "loss": 0.0713, "step": 12000 }, { "epoch": 0.40603130599411746, "grad_norm": 0.2037106305360794, "learning_rate": 9.866291359726438e-05, "loss": 0.0636, "step": 12010 }, { "epoch": 0.40636938368437064, "grad_norm": 0.2884880304336548, "learning_rate": 9.865911269150861e-05, "loss": 0.0723, "step": 12020 }, { "epoch": 0.4067074613746239, "grad_norm": 0.2296300083398819, "learning_rate": 9.865530646446544e-05, "loss": 0.0583, "step": 12030 }, { "epoch": 0.4070455390648771, "grad_norm": 0.24948328733444214, "learning_rate": 9.86514949165511e-05, "loss": 0.0608, "step": 12040 }, { "epoch": 0.40738361675513035, "grad_norm": 0.2569175660610199, "learning_rate": 9.864767804818243e-05, "loss": 0.0674, "step": 12050 }, { "epoch": 0.40772169444538353, "grad_norm": 0.1805688589811325, "learning_rate": 9.86438558597768e-05, "loss": 0.0645, "step": 12060 }, { "epoch": 0.40805977213563677, "grad_norm": 0.2146618515253067, "learning_rate": 9.864002835175225e-05, "loss": 0.0579, "step": 12070 }, { "epoch": 0.40839784982589, "grad_norm": 0.30334052443504333, "learning_rate": 9.863619552452734e-05, "loss": 0.0653, "step": 12080 }, { "epoch": 0.4087359275161432, "grad_norm": 0.32339048385620117, "learning_rate": 9.863235737852119e-05, "loss": 0.0628, "step": 12090 }, { "epoch": 0.4090740052063964, "grad_norm": 0.2653566002845764, "learning_rate": 9.862851391415356e-05, "loss": 0.0666, "step": 12100 }, { "epoch": 0.40941208289664965, "grad_norm": 0.23860955238342285, "learning_rate": 9.862466513184477e-05, "loss": 0.0618, "step": 12110 }, { "epoch": 0.4097501605869029, "grad_norm": 0.2449835240840912, "learning_rate": 9.86208110320157e-05, "loss": 0.0611, "step": 12120 }, { "epoch": 0.41008823827715607, "grad_norm": 0.2328367680311203, "learning_rate": 9.861695161508784e-05, "loss": 0.0684, "step": 12130 }, { "epoch": 0.4104263159674093, "grad_norm": 0.2685941755771637, "learning_rate": 9.861308688148324e-05, "loss": 0.0625, "step": 12140 }, { "epoch": 0.41076439365766254, "grad_norm": 0.28468018770217896, "learning_rate": 9.860921683162455e-05, "loss": 0.0649, "step": 12150 }, { "epoch": 0.4111024713479158, "grad_norm": 0.27446064352989197, "learning_rate": 9.860534146593499e-05, "loss": 0.0715, "step": 12160 }, { "epoch": 0.41144054903816896, "grad_norm": 0.23087449371814728, "learning_rate": 9.860146078483836e-05, "loss": 0.0636, "step": 12170 }, { "epoch": 0.4117786267284222, "grad_norm": 0.33747076988220215, "learning_rate": 9.859757478875905e-05, "loss": 0.0671, "step": 12180 }, { "epoch": 0.41211670441867543, "grad_norm": 0.30559873580932617, "learning_rate": 9.859368347812204e-05, "loss": 0.0644, "step": 12190 }, { "epoch": 0.4124547821089286, "grad_norm": 0.33484989404678345, "learning_rate": 9.858978685335285e-05, "loss": 0.0646, "step": 12200 }, { "epoch": 0.41279285979918184, "grad_norm": 0.2670334279537201, "learning_rate": 9.858588491487763e-05, "loss": 0.0663, "step": 12210 }, { "epoch": 0.4131309374894351, "grad_norm": 0.24165114760398865, "learning_rate": 9.858197766312308e-05, "loss": 0.065, "step": 12220 }, { "epoch": 0.4134690151796883, "grad_norm": 0.25609272718429565, "learning_rate": 9.857806509851649e-05, "loss": 0.0699, "step": 12230 }, { "epoch": 0.4138070928699415, "grad_norm": 0.3126818835735321, "learning_rate": 9.857414722148574e-05, "loss": 0.0702, "step": 12240 }, { "epoch": 0.41414517056019473, "grad_norm": 0.3205353617668152, "learning_rate": 9.857022403245928e-05, "loss": 0.0589, "step": 12250 }, { "epoch": 0.41448324825044797, "grad_norm": 0.34930047392845154, "learning_rate": 9.856629553186615e-05, "loss": 0.0609, "step": 12260 }, { "epoch": 0.41482132594070115, "grad_norm": 0.3817729949951172, "learning_rate": 9.856236172013595e-05, "loss": 0.0641, "step": 12270 }, { "epoch": 0.4151594036309544, "grad_norm": 0.30430877208709717, "learning_rate": 9.85584225976989e-05, "loss": 0.0742, "step": 12280 }, { "epoch": 0.4154974813212076, "grad_norm": 0.25751689076423645, "learning_rate": 9.855447816498575e-05, "loss": 0.0607, "step": 12290 }, { "epoch": 0.41583555901146085, "grad_norm": 0.17483504116535187, "learning_rate": 9.855052842242787e-05, "loss": 0.0617, "step": 12300 }, { "epoch": 0.41617363670171403, "grad_norm": 0.25836604833602905, "learning_rate": 9.85465733704572e-05, "loss": 0.0737, "step": 12310 }, { "epoch": 0.41651171439196727, "grad_norm": 0.24795588850975037, "learning_rate": 9.854261300950624e-05, "loss": 0.0685, "step": 12320 }, { "epoch": 0.4168497920822205, "grad_norm": 0.22427800297737122, "learning_rate": 9.853864734000813e-05, "loss": 0.0644, "step": 12330 }, { "epoch": 0.41718786977247374, "grad_norm": 0.2433728128671646, "learning_rate": 9.85346763623965e-05, "loss": 0.0633, "step": 12340 }, { "epoch": 0.4175259474627269, "grad_norm": 0.25110846757888794, "learning_rate": 9.853070007710564e-05, "loss": 0.0604, "step": 12350 }, { "epoch": 0.41786402515298016, "grad_norm": 0.2527155876159668, "learning_rate": 9.85267184845704e-05, "loss": 0.0663, "step": 12360 }, { "epoch": 0.4182021028432334, "grad_norm": 0.21965283155441284, "learning_rate": 9.852273158522616e-05, "loss": 0.0677, "step": 12370 }, { "epoch": 0.41854018053348657, "grad_norm": 0.3626271188259125, "learning_rate": 9.851873937950896e-05, "loss": 0.0676, "step": 12380 }, { "epoch": 0.4188782582237398, "grad_norm": 0.3823398947715759, "learning_rate": 9.851474186785537e-05, "loss": 0.061, "step": 12390 }, { "epoch": 0.41921633591399304, "grad_norm": 0.25438815355300903, "learning_rate": 9.851073905070254e-05, "loss": 0.0709, "step": 12400 }, { "epoch": 0.4195544136042463, "grad_norm": 0.27185049653053284, "learning_rate": 9.850673092848824e-05, "loss": 0.0628, "step": 12410 }, { "epoch": 0.41989249129449946, "grad_norm": 0.3287135660648346, "learning_rate": 9.850271750165077e-05, "loss": 0.0638, "step": 12420 }, { "epoch": 0.4202305689847527, "grad_norm": 0.32051029801368713, "learning_rate": 9.849869877062902e-05, "loss": 0.0646, "step": 12430 }, { "epoch": 0.42056864667500593, "grad_norm": 0.2896265983581543, "learning_rate": 9.849467473586252e-05, "loss": 0.0645, "step": 12440 }, { "epoch": 0.4209067243652591, "grad_norm": 0.2978745996952057, "learning_rate": 9.849064539779127e-05, "loss": 0.065, "step": 12450 }, { "epoch": 0.42124480205551235, "grad_norm": 0.22354881465435028, "learning_rate": 9.848661075685594e-05, "loss": 0.0662, "step": 12460 }, { "epoch": 0.4215828797457656, "grad_norm": 0.27632325887680054, "learning_rate": 9.848257081349778e-05, "loss": 0.0637, "step": 12470 }, { "epoch": 0.4219209574360188, "grad_norm": 0.24977725744247437, "learning_rate": 9.847852556815856e-05, "loss": 0.0662, "step": 12480 }, { "epoch": 0.422259035126272, "grad_norm": 0.1818186640739441, "learning_rate": 9.847447502128067e-05, "loss": 0.0657, "step": 12490 }, { "epoch": 0.42259711281652523, "grad_norm": 0.234344944357872, "learning_rate": 9.847041917330708e-05, "loss": 0.0631, "step": 12500 }, { "epoch": 0.42293519050677847, "grad_norm": 0.21886368095874786, "learning_rate": 9.846635802468132e-05, "loss": 0.0666, "step": 12510 }, { "epoch": 0.4232732681970317, "grad_norm": 0.20884647965431213, "learning_rate": 9.84622915758475e-05, "loss": 0.0601, "step": 12520 }, { "epoch": 0.4236113458872849, "grad_norm": 0.30955880880355835, "learning_rate": 9.845821982725034e-05, "loss": 0.0587, "step": 12530 }, { "epoch": 0.4239494235775381, "grad_norm": 0.22576174139976501, "learning_rate": 9.845414277933514e-05, "loss": 0.0561, "step": 12540 }, { "epoch": 0.42428750126779136, "grad_norm": 0.3268166184425354, "learning_rate": 9.845006043254771e-05, "loss": 0.0714, "step": 12550 }, { "epoch": 0.42462557895804454, "grad_norm": 0.2012188881635666, "learning_rate": 9.844597278733451e-05, "loss": 0.066, "step": 12560 }, { "epoch": 0.42496365664829777, "grad_norm": 0.2737833857536316, "learning_rate": 9.844187984414259e-05, "loss": 0.0653, "step": 12570 }, { "epoch": 0.425301734338551, "grad_norm": 0.20493954420089722, "learning_rate": 9.84377816034195e-05, "loss": 0.063, "step": 12580 }, { "epoch": 0.42563981202880424, "grad_norm": 0.21125182509422302, "learning_rate": 9.843367806561345e-05, "loss": 0.0603, "step": 12590 }, { "epoch": 0.4259778897190574, "grad_norm": 0.2744312882423401, "learning_rate": 9.842956923117317e-05, "loss": 0.0608, "step": 12600 }, { "epoch": 0.42631596740931066, "grad_norm": 0.27000322937965393, "learning_rate": 9.842545510054802e-05, "loss": 0.0635, "step": 12610 }, { "epoch": 0.4266540450995639, "grad_norm": 0.20206111669540405, "learning_rate": 9.842133567418792e-05, "loss": 0.0675, "step": 12620 }, { "epoch": 0.4269921227898171, "grad_norm": 0.234543576836586, "learning_rate": 9.841721095254333e-05, "loss": 0.0676, "step": 12630 }, { "epoch": 0.4273302004800703, "grad_norm": 0.2019244134426117, "learning_rate": 9.841308093606537e-05, "loss": 0.065, "step": 12640 }, { "epoch": 0.42766827817032355, "grad_norm": 0.3071616291999817, "learning_rate": 9.840894562520565e-05, "loss": 0.0628, "step": 12650 }, { "epoch": 0.4280063558605768, "grad_norm": 0.23040884733200073, "learning_rate": 9.840480502041642e-05, "loss": 0.0569, "step": 12660 }, { "epoch": 0.42834443355082996, "grad_norm": 0.27103209495544434, "learning_rate": 9.840065912215049e-05, "loss": 0.0709, "step": 12670 }, { "epoch": 0.4286825112410832, "grad_norm": 0.3324100971221924, "learning_rate": 9.839650793086124e-05, "loss": 0.0653, "step": 12680 }, { "epoch": 0.42902058893133643, "grad_norm": 0.31937795877456665, "learning_rate": 9.839235144700265e-05, "loss": 0.0681, "step": 12690 }, { "epoch": 0.42935866662158967, "grad_norm": 0.28958287835121155, "learning_rate": 9.838818967102926e-05, "loss": 0.0623, "step": 12700 }, { "epoch": 0.42969674431184285, "grad_norm": 0.25965794920921326, "learning_rate": 9.83840226033962e-05, "loss": 0.0619, "step": 12710 }, { "epoch": 0.4300348220020961, "grad_norm": 0.27102112770080566, "learning_rate": 9.837985024455918e-05, "loss": 0.064, "step": 12720 }, { "epoch": 0.4303728996923493, "grad_norm": 0.28392457962036133, "learning_rate": 9.837567259497447e-05, "loss": 0.0716, "step": 12730 }, { "epoch": 0.4307109773826025, "grad_norm": 0.1803884357213974, "learning_rate": 9.837148965509894e-05, "loss": 0.0554, "step": 12740 }, { "epoch": 0.43104905507285574, "grad_norm": 0.247506245970726, "learning_rate": 9.836730142539001e-05, "loss": 0.065, "step": 12750 }, { "epoch": 0.43138713276310897, "grad_norm": 0.30706241726875305, "learning_rate": 9.836310790630574e-05, "loss": 0.0699, "step": 12760 }, { "epoch": 0.4317252104533622, "grad_norm": 0.3019953966140747, "learning_rate": 9.83589090983047e-05, "loss": 0.0644, "step": 12770 }, { "epoch": 0.4320632881436154, "grad_norm": 0.31266939640045166, "learning_rate": 9.835470500184605e-05, "loss": 0.0645, "step": 12780 }, { "epoch": 0.4324013658338686, "grad_norm": 0.21602493524551392, "learning_rate": 9.835049561738957e-05, "loss": 0.0681, "step": 12790 }, { "epoch": 0.43273944352412186, "grad_norm": 0.3611225187778473, "learning_rate": 9.834628094539558e-05, "loss": 0.0717, "step": 12800 }, { "epoch": 0.43307752121437504, "grad_norm": 0.30372822284698486, "learning_rate": 9.834206098632499e-05, "loss": 0.0667, "step": 12810 }, { "epoch": 0.4334155989046283, "grad_norm": 0.23606547713279724, "learning_rate": 9.833783574063931e-05, "loss": 0.0608, "step": 12820 }, { "epoch": 0.4337536765948815, "grad_norm": 0.2611715793609619, "learning_rate": 9.833360520880058e-05, "loss": 0.0619, "step": 12830 }, { "epoch": 0.43409175428513475, "grad_norm": 0.22426237165927887, "learning_rate": 9.832936939127144e-05, "loss": 0.0668, "step": 12840 }, { "epoch": 0.4344298319753879, "grad_norm": 0.20895370841026306, "learning_rate": 9.832512828851515e-05, "loss": 0.0754, "step": 12850 }, { "epoch": 0.43476790966564116, "grad_norm": 0.24559535086154938, "learning_rate": 9.832088190099546e-05, "loss": 0.0669, "step": 12860 }, { "epoch": 0.4351059873558944, "grad_norm": 0.2674482762813568, "learning_rate": 9.831663022917679e-05, "loss": 0.0714, "step": 12870 }, { "epoch": 0.4354440650461476, "grad_norm": 0.3103368580341339, "learning_rate": 9.831237327352407e-05, "loss": 0.0659, "step": 12880 }, { "epoch": 0.4357821427364008, "grad_norm": 0.2759416699409485, "learning_rate": 9.830811103450286e-05, "loss": 0.0692, "step": 12890 }, { "epoch": 0.43612022042665405, "grad_norm": 0.29831644892692566, "learning_rate": 9.830384351257924e-05, "loss": 0.0597, "step": 12900 }, { "epoch": 0.4364582981169073, "grad_norm": 0.24793262779712677, "learning_rate": 9.829957070821993e-05, "loss": 0.0651, "step": 12910 }, { "epoch": 0.43679637580716046, "grad_norm": 0.21799816191196442, "learning_rate": 9.829529262189218e-05, "loss": 0.0758, "step": 12920 }, { "epoch": 0.4371344534974137, "grad_norm": 0.29416316747665405, "learning_rate": 9.829100925406385e-05, "loss": 0.0637, "step": 12930 }, { "epoch": 0.43747253118766694, "grad_norm": 0.18872475624084473, "learning_rate": 9.828672060520333e-05, "loss": 0.0604, "step": 12940 }, { "epoch": 0.43781060887792017, "grad_norm": 0.26927125453948975, "learning_rate": 9.828242667577966e-05, "loss": 0.0656, "step": 12950 }, { "epoch": 0.43814868656817335, "grad_norm": 0.2079368680715561, "learning_rate": 9.82781274662624e-05, "loss": 0.0627, "step": 12960 }, { "epoch": 0.4384867642584266, "grad_norm": 0.31324681639671326, "learning_rate": 9.82738229771217e-05, "loss": 0.0645, "step": 12970 }, { "epoch": 0.4388248419486798, "grad_norm": 0.3073675334453583, "learning_rate": 9.826951320882829e-05, "loss": 0.0673, "step": 12980 }, { "epoch": 0.439162919638933, "grad_norm": 0.3585650622844696, "learning_rate": 9.826519816185351e-05, "loss": 0.0661, "step": 12990 }, { "epoch": 0.43950099732918624, "grad_norm": 0.2080330103635788, "learning_rate": 9.826087783666921e-05, "loss": 0.0728, "step": 13000 }, { "epoch": 0.4398390750194395, "grad_norm": 0.25111594796180725, "learning_rate": 9.825655223374787e-05, "loss": 0.0644, "step": 13010 }, { "epoch": 0.4401771527096927, "grad_norm": 0.26305845379829407, "learning_rate": 9.825222135356253e-05, "loss": 0.066, "step": 13020 }, { "epoch": 0.4405152303999459, "grad_norm": 0.30759555101394653, "learning_rate": 9.82478851965868e-05, "loss": 0.0651, "step": 13030 }, { "epoch": 0.4408533080901991, "grad_norm": 0.18199104070663452, "learning_rate": 9.82435437632949e-05, "loss": 0.0695, "step": 13040 }, { "epoch": 0.44119138578045236, "grad_norm": 0.2967785894870758, "learning_rate": 9.823919705416158e-05, "loss": 0.0677, "step": 13050 }, { "epoch": 0.44152946347070554, "grad_norm": 0.4656156599521637, "learning_rate": 9.82348450696622e-05, "loss": 0.065, "step": 13060 }, { "epoch": 0.4418675411609588, "grad_norm": 0.21515437960624695, "learning_rate": 9.823048781027268e-05, "loss": 0.0589, "step": 13070 }, { "epoch": 0.442205618851212, "grad_norm": 0.18071579933166504, "learning_rate": 9.822612527646953e-05, "loss": 0.0589, "step": 13080 }, { "epoch": 0.44254369654146525, "grad_norm": 0.2353813648223877, "learning_rate": 9.822175746872984e-05, "loss": 0.064, "step": 13090 }, { "epoch": 0.44288177423171843, "grad_norm": 0.1719016581773758, "learning_rate": 9.821738438753123e-05, "loss": 0.07, "step": 13100 }, { "epoch": 0.44321985192197166, "grad_norm": 0.32913699746131897, "learning_rate": 9.821300603335196e-05, "loss": 0.0702, "step": 13110 }, { "epoch": 0.4435579296122249, "grad_norm": 0.2600681781768799, "learning_rate": 9.820862240667085e-05, "loss": 0.0659, "step": 13120 }, { "epoch": 0.44389600730247813, "grad_norm": 0.2797924280166626, "learning_rate": 9.820423350796726e-05, "loss": 0.0607, "step": 13130 }, { "epoch": 0.4442340849927313, "grad_norm": 0.5116257667541504, "learning_rate": 9.819983933772118e-05, "loss": 0.0644, "step": 13140 }, { "epoch": 0.44457216268298455, "grad_norm": 0.3061426877975464, "learning_rate": 9.819543989641314e-05, "loss": 0.0684, "step": 13150 }, { "epoch": 0.4449102403732378, "grad_norm": 0.22922007739543915, "learning_rate": 9.819103518452423e-05, "loss": 0.0698, "step": 13160 }, { "epoch": 0.44524831806349097, "grad_norm": 0.2395579069852829, "learning_rate": 9.818662520253618e-05, "loss": 0.0669, "step": 13170 }, { "epoch": 0.4455863957537442, "grad_norm": 0.2227507382631302, "learning_rate": 9.818220995093126e-05, "loss": 0.0614, "step": 13180 }, { "epoch": 0.44592447344399744, "grad_norm": 0.2216266393661499, "learning_rate": 9.817778943019228e-05, "loss": 0.0637, "step": 13190 }, { "epoch": 0.4462625511342507, "grad_norm": 0.4890085756778717, "learning_rate": 9.81733636408027e-05, "loss": 0.0628, "step": 13200 }, { "epoch": 0.44660062882450385, "grad_norm": 0.2533320188522339, "learning_rate": 9.816893258324649e-05, "loss": 0.0645, "step": 13210 }, { "epoch": 0.4469387065147571, "grad_norm": 0.29557284712791443, "learning_rate": 9.816449625800823e-05, "loss": 0.0692, "step": 13220 }, { "epoch": 0.4472767842050103, "grad_norm": 0.1984381228685379, "learning_rate": 9.816005466557308e-05, "loss": 0.0647, "step": 13230 }, { "epoch": 0.4476148618952635, "grad_norm": 0.3841463029384613, "learning_rate": 9.815560780642674e-05, "loss": 0.0588, "step": 13240 }, { "epoch": 0.44795293958551674, "grad_norm": 0.16880467534065247, "learning_rate": 9.815115568105555e-05, "loss": 0.0653, "step": 13250 }, { "epoch": 0.44829101727577, "grad_norm": 0.25194406509399414, "learning_rate": 9.814669828994638e-05, "loss": 0.0646, "step": 13260 }, { "epoch": 0.4486290949660232, "grad_norm": 0.2789205014705658, "learning_rate": 9.814223563358665e-05, "loss": 0.0668, "step": 13270 }, { "epoch": 0.4489671726562764, "grad_norm": 0.3348623812198639, "learning_rate": 9.813776771246443e-05, "loss": 0.0652, "step": 13280 }, { "epoch": 0.4493052503465296, "grad_norm": 0.2806498110294342, "learning_rate": 9.813329452706829e-05, "loss": 0.0648, "step": 13290 }, { "epoch": 0.44964332803678286, "grad_norm": 0.3057667315006256, "learning_rate": 9.812881607788744e-05, "loss": 0.0627, "step": 13300 }, { "epoch": 0.4499814057270361, "grad_norm": 0.26996153593063354, "learning_rate": 9.812433236541163e-05, "loss": 0.0618, "step": 13310 }, { "epoch": 0.4503194834172893, "grad_norm": 0.22563403844833374, "learning_rate": 9.811984339013116e-05, "loss": 0.0761, "step": 13320 }, { "epoch": 0.4506575611075425, "grad_norm": 0.4048364758491516, "learning_rate": 9.811534915253698e-05, "loss": 0.0634, "step": 13330 }, { "epoch": 0.45099563879779575, "grad_norm": 0.22270143032073975, "learning_rate": 9.811084965312056e-05, "loss": 0.0702, "step": 13340 }, { "epoch": 0.45133371648804893, "grad_norm": 0.29325029253959656, "learning_rate": 9.810634489237396e-05, "loss": 0.067, "step": 13350 }, { "epoch": 0.45167179417830217, "grad_norm": 0.19029855728149414, "learning_rate": 9.81018348707898e-05, "loss": 0.0705, "step": 13360 }, { "epoch": 0.4520098718685554, "grad_norm": 0.30169877409935, "learning_rate": 9.809731958886131e-05, "loss": 0.0665, "step": 13370 }, { "epoch": 0.45234794955880864, "grad_norm": 0.4025127589702606, "learning_rate": 9.809279904708224e-05, "loss": 0.0598, "step": 13380 }, { "epoch": 0.4526860272490618, "grad_norm": 0.25361788272857666, "learning_rate": 9.808827324594699e-05, "loss": 0.0637, "step": 13390 }, { "epoch": 0.45302410493931505, "grad_norm": 0.32730942964553833, "learning_rate": 9.808374218595046e-05, "loss": 0.0655, "step": 13400 }, { "epoch": 0.4533621826295683, "grad_norm": 0.3320581018924713, "learning_rate": 9.80792058675882e-05, "loss": 0.0617, "step": 13410 }, { "epoch": 0.45370026031982147, "grad_norm": 0.2333044707775116, "learning_rate": 9.807466429135627e-05, "loss": 0.0562, "step": 13420 }, { "epoch": 0.4540383380100747, "grad_norm": 0.3063654899597168, "learning_rate": 9.807011745775132e-05, "loss": 0.0676, "step": 13430 }, { "epoch": 0.45437641570032794, "grad_norm": 0.18614712357521057, "learning_rate": 9.806556536727061e-05, "loss": 0.0653, "step": 13440 }, { "epoch": 0.4547144933905812, "grad_norm": 0.24144840240478516, "learning_rate": 9.806100802041193e-05, "loss": 0.0628, "step": 13450 }, { "epoch": 0.45505257108083436, "grad_norm": 0.1752190738916397, "learning_rate": 9.805644541767368e-05, "loss": 0.0662, "step": 13460 }, { "epoch": 0.4553906487710876, "grad_norm": 0.2591942846775055, "learning_rate": 9.805187755955478e-05, "loss": 0.07, "step": 13470 }, { "epoch": 0.4557287264613408, "grad_norm": 0.26762905716896057, "learning_rate": 9.804730444655483e-05, "loss": 0.0607, "step": 13480 }, { "epoch": 0.45606680415159406, "grad_norm": 0.24515782296657562, "learning_rate": 9.804272607917388e-05, "loss": 0.061, "step": 13490 }, { "epoch": 0.45640488184184724, "grad_norm": 0.19092626869678497, "learning_rate": 9.803814245791265e-05, "loss": 0.0654, "step": 13500 }, { "epoch": 0.4567429595321005, "grad_norm": 0.22793692350387573, "learning_rate": 9.803355358327239e-05, "loss": 0.0652, "step": 13510 }, { "epoch": 0.4570810372223537, "grad_norm": 0.27156785130500793, "learning_rate": 9.802895945575492e-05, "loss": 0.0541, "step": 13520 }, { "epoch": 0.4574191149126069, "grad_norm": 0.2617112696170807, "learning_rate": 9.802436007586266e-05, "loss": 0.0563, "step": 13530 }, { "epoch": 0.45775719260286013, "grad_norm": 0.2339746654033661, "learning_rate": 9.801975544409858e-05, "loss": 0.0633, "step": 13540 }, { "epoch": 0.45809527029311337, "grad_norm": 0.3093540370464325, "learning_rate": 9.801514556096625e-05, "loss": 0.069, "step": 13550 }, { "epoch": 0.4584333479833666, "grad_norm": 0.23567695915699005, "learning_rate": 9.801053042696977e-05, "loss": 0.0612, "step": 13560 }, { "epoch": 0.4587714256736198, "grad_norm": 0.23234719038009644, "learning_rate": 9.800591004261388e-05, "loss": 0.0714, "step": 13570 }, { "epoch": 0.459109503363873, "grad_norm": 0.4201902449131012, "learning_rate": 9.800128440840385e-05, "loss": 0.0649, "step": 13580 }, { "epoch": 0.45944758105412625, "grad_norm": 0.2811574339866638, "learning_rate": 9.799665352484552e-05, "loss": 0.0632, "step": 13590 }, { "epoch": 0.45978565874437943, "grad_norm": 0.30768921971321106, "learning_rate": 9.799201739244532e-05, "loss": 0.0726, "step": 13600 }, { "epoch": 0.46012373643463267, "grad_norm": 0.1871376633644104, "learning_rate": 9.798737601171025e-05, "loss": 0.0667, "step": 13610 }, { "epoch": 0.4604618141248859, "grad_norm": 0.29681724309921265, "learning_rate": 9.79827293831479e-05, "loss": 0.0685, "step": 13620 }, { "epoch": 0.46079989181513914, "grad_norm": 0.23206773400306702, "learning_rate": 9.797807750726638e-05, "loss": 0.0645, "step": 13630 }, { "epoch": 0.4611379695053923, "grad_norm": 0.3600550591945648, "learning_rate": 9.797342038457446e-05, "loss": 0.0573, "step": 13640 }, { "epoch": 0.46147604719564556, "grad_norm": 0.25226303935050964, "learning_rate": 9.796875801558141e-05, "loss": 0.0646, "step": 13650 }, { "epoch": 0.4618141248858988, "grad_norm": 0.19560915231704712, "learning_rate": 9.79640904007971e-05, "loss": 0.0689, "step": 13660 }, { "epoch": 0.462152202576152, "grad_norm": 0.3117327094078064, "learning_rate": 9.795941754073199e-05, "loss": 0.0661, "step": 13670 }, { "epoch": 0.4624902802664052, "grad_norm": 0.21343861520290375, "learning_rate": 9.795473943589705e-05, "loss": 0.062, "step": 13680 }, { "epoch": 0.46282835795665844, "grad_norm": 0.3240487575531006, "learning_rate": 9.795005608680394e-05, "loss": 0.0646, "step": 13690 }, { "epoch": 0.4631664356469117, "grad_norm": 0.41215917468070984, "learning_rate": 9.794536749396477e-05, "loss": 0.0635, "step": 13700 }, { "epoch": 0.46350451333716486, "grad_norm": 0.2200150191783905, "learning_rate": 9.79406736578923e-05, "loss": 0.0635, "step": 13710 }, { "epoch": 0.4638425910274181, "grad_norm": 0.23550191521644592, "learning_rate": 9.793597457909984e-05, "loss": 0.062, "step": 13720 }, { "epoch": 0.46418066871767133, "grad_norm": 0.23860955238342285, "learning_rate": 9.793127025810127e-05, "loss": 0.0695, "step": 13730 }, { "epoch": 0.46451874640792457, "grad_norm": 0.22217904031276703, "learning_rate": 9.792656069541104e-05, "loss": 0.0674, "step": 13740 }, { "epoch": 0.46485682409817775, "grad_norm": 0.3917473256587982, "learning_rate": 9.79218458915442e-05, "loss": 0.0603, "step": 13750 }, { "epoch": 0.465194901788431, "grad_norm": 0.2076631337404251, "learning_rate": 9.791712584701634e-05, "loss": 0.0639, "step": 13760 }, { "epoch": 0.4655329794786842, "grad_norm": 0.2699761390686035, "learning_rate": 9.791240056234364e-05, "loss": 0.0601, "step": 13770 }, { "epoch": 0.4658710571689374, "grad_norm": 0.35980546474456787, "learning_rate": 9.790767003804283e-05, "loss": 0.0605, "step": 13780 }, { "epoch": 0.46620913485919063, "grad_norm": 0.22798073291778564, "learning_rate": 9.790293427463126e-05, "loss": 0.0623, "step": 13790 }, { "epoch": 0.46654721254944387, "grad_norm": 0.22291529178619385, "learning_rate": 9.789819327262684e-05, "loss": 0.0636, "step": 13800 }, { "epoch": 0.4668852902396971, "grad_norm": 0.2605602741241455, "learning_rate": 9.7893447032548e-05, "loss": 0.0654, "step": 13810 }, { "epoch": 0.4672233679299503, "grad_norm": 0.3548091948032379, "learning_rate": 9.78886955549138e-05, "loss": 0.0601, "step": 13820 }, { "epoch": 0.4675614456202035, "grad_norm": 0.20622502267360687, "learning_rate": 9.788393884024387e-05, "loss": 0.0681, "step": 13830 }, { "epoch": 0.46789952331045676, "grad_norm": 0.42631861567497253, "learning_rate": 9.787917688905836e-05, "loss": 0.0701, "step": 13840 }, { "epoch": 0.46823760100070994, "grad_norm": 0.24787433445453644, "learning_rate": 9.787440970187807e-05, "loss": 0.0615, "step": 13850 }, { "epoch": 0.46857567869096317, "grad_norm": 0.3058626055717468, "learning_rate": 9.786963727922429e-05, "loss": 0.0656, "step": 13860 }, { "epoch": 0.4689137563812164, "grad_norm": 0.2482173591852188, "learning_rate": 9.786485962161897e-05, "loss": 0.0656, "step": 13870 }, { "epoch": 0.46925183407146964, "grad_norm": 0.2308058738708496, "learning_rate": 9.786007672958455e-05, "loss": 0.0656, "step": 13880 }, { "epoch": 0.4695899117617228, "grad_norm": 0.435903400182724, "learning_rate": 9.78552886036441e-05, "loss": 0.0587, "step": 13890 }, { "epoch": 0.46992798945197606, "grad_norm": 0.2048713117837906, "learning_rate": 9.785049524432124e-05, "loss": 0.0588, "step": 13900 }, { "epoch": 0.4702660671422293, "grad_norm": 0.2756282389163971, "learning_rate": 9.784569665214016e-05, "loss": 0.0674, "step": 13910 }, { "epoch": 0.47060414483248253, "grad_norm": 0.24436643719673157, "learning_rate": 9.784089282762563e-05, "loss": 0.0638, "step": 13920 }, { "epoch": 0.4709422225227357, "grad_norm": 0.23691818118095398, "learning_rate": 9.7836083771303e-05, "loss": 0.0636, "step": 13930 }, { "epoch": 0.47128030021298895, "grad_norm": 0.20565223693847656, "learning_rate": 9.783126948369817e-05, "loss": 0.0659, "step": 13940 }, { "epoch": 0.4716183779032422, "grad_norm": 0.23576122522354126, "learning_rate": 9.78264499653376e-05, "loss": 0.0665, "step": 13950 }, { "epoch": 0.47195645559349536, "grad_norm": 0.2553093135356903, "learning_rate": 9.782162521674838e-05, "loss": 0.0675, "step": 13960 }, { "epoch": 0.4722945332837486, "grad_norm": 0.22466923296451569, "learning_rate": 9.781679523845812e-05, "loss": 0.0631, "step": 13970 }, { "epoch": 0.47263261097400183, "grad_norm": 0.33145251870155334, "learning_rate": 9.781196003099502e-05, "loss": 0.0595, "step": 13980 }, { "epoch": 0.47297068866425507, "grad_norm": 0.26820969581604004, "learning_rate": 9.780711959488786e-05, "loss": 0.0564, "step": 13990 }, { "epoch": 0.47330876635450825, "grad_norm": 0.2288629412651062, "learning_rate": 9.780227393066599e-05, "loss": 0.0599, "step": 14000 }, { "epoch": 0.4736468440447615, "grad_norm": 0.3067095875740051, "learning_rate": 9.77974230388593e-05, "loss": 0.0643, "step": 14010 }, { "epoch": 0.4739849217350147, "grad_norm": 0.2170427292585373, "learning_rate": 9.779256691999829e-05, "loss": 0.0688, "step": 14020 }, { "epoch": 0.4743229994252679, "grad_norm": 0.29353076219558716, "learning_rate": 9.778770557461403e-05, "loss": 0.0606, "step": 14030 }, { "epoch": 0.47466107711552114, "grad_norm": 0.24649584293365479, "learning_rate": 9.778283900323812e-05, "loss": 0.0629, "step": 14040 }, { "epoch": 0.47499915480577437, "grad_norm": 0.21706920862197876, "learning_rate": 9.777796720640277e-05, "loss": 0.0674, "step": 14050 }, { "epoch": 0.4753372324960276, "grad_norm": 0.2496504932641983, "learning_rate": 9.777309018464078e-05, "loss": 0.0625, "step": 14060 }, { "epoch": 0.4756753101862808, "grad_norm": 0.408113956451416, "learning_rate": 9.776820793848547e-05, "loss": 0.0615, "step": 14070 }, { "epoch": 0.476013387876534, "grad_norm": 0.28152239322662354, "learning_rate": 9.776332046847075e-05, "loss": 0.0641, "step": 14080 }, { "epoch": 0.47635146556678726, "grad_norm": 0.3109959661960602, "learning_rate": 9.775842777513111e-05, "loss": 0.0618, "step": 14090 }, { "epoch": 0.4766895432570405, "grad_norm": 0.22787342965602875, "learning_rate": 9.775352985900163e-05, "loss": 0.064, "step": 14100 }, { "epoch": 0.4770276209472937, "grad_norm": 0.3011524975299835, "learning_rate": 9.774862672061791e-05, "loss": 0.0604, "step": 14110 }, { "epoch": 0.4773656986375469, "grad_norm": 0.419100284576416, "learning_rate": 9.774371836051616e-05, "loss": 0.0578, "step": 14120 }, { "epoch": 0.47770377632780014, "grad_norm": 0.5408785939216614, "learning_rate": 9.773880477923315e-05, "loss": 0.0634, "step": 14130 }, { "epoch": 0.4780418540180533, "grad_norm": 0.37305527925491333, "learning_rate": 9.773388597730623e-05, "loss": 0.0612, "step": 14140 }, { "epoch": 0.47837993170830656, "grad_norm": 0.39990079402923584, "learning_rate": 9.77289619552733e-05, "loss": 0.0689, "step": 14150 }, { "epoch": 0.4787180093985598, "grad_norm": 0.24241384863853455, "learning_rate": 9.772403271367285e-05, "loss": 0.0643, "step": 14160 }, { "epoch": 0.47905608708881303, "grad_norm": 0.20217172801494598, "learning_rate": 9.771909825304396e-05, "loss": 0.0611, "step": 14170 }, { "epoch": 0.4793941647790662, "grad_norm": 0.36995062232017517, "learning_rate": 9.771415857392619e-05, "loss": 0.0678, "step": 14180 }, { "epoch": 0.47973224246931945, "grad_norm": 0.18396973609924316, "learning_rate": 9.770921367685978e-05, "loss": 0.0561, "step": 14190 }, { "epoch": 0.4800703201595727, "grad_norm": 0.19569656252861023, "learning_rate": 9.770426356238551e-05, "loss": 0.0649, "step": 14200 }, { "epoch": 0.48040839784982586, "grad_norm": 0.3392314612865448, "learning_rate": 9.769930823104469e-05, "loss": 0.0669, "step": 14210 }, { "epoch": 0.4807464755400791, "grad_norm": 0.26514285802841187, "learning_rate": 9.769434768337926e-05, "loss": 0.0642, "step": 14220 }, { "epoch": 0.48108455323033233, "grad_norm": 0.19695697724819183, "learning_rate": 9.768938191993164e-05, "loss": 0.0637, "step": 14230 }, { "epoch": 0.48142263092058557, "grad_norm": 0.49390023946762085, "learning_rate": 9.768441094124494e-05, "loss": 0.0668, "step": 14240 }, { "epoch": 0.48176070861083875, "grad_norm": 0.3074679672718048, "learning_rate": 9.767943474786275e-05, "loss": 0.07, "step": 14250 }, { "epoch": 0.482098786301092, "grad_norm": 0.24571257829666138, "learning_rate": 9.767445334032923e-05, "loss": 0.0606, "step": 14260 }, { "epoch": 0.4824368639913452, "grad_norm": 0.3457341492176056, "learning_rate": 9.766946671918919e-05, "loss": 0.0623, "step": 14270 }, { "epoch": 0.48277494168159846, "grad_norm": 0.20184814929962158, "learning_rate": 9.766447488498796e-05, "loss": 0.0635, "step": 14280 }, { "epoch": 0.48311301937185164, "grad_norm": 0.2989766001701355, "learning_rate": 9.765947783827139e-05, "loss": 0.059, "step": 14290 }, { "epoch": 0.4834510970621049, "grad_norm": 0.28648626804351807, "learning_rate": 9.765447557958599e-05, "loss": 0.0653, "step": 14300 }, { "epoch": 0.4837891747523581, "grad_norm": 0.26028844714164734, "learning_rate": 9.764946810947879e-05, "loss": 0.0694, "step": 14310 }, { "epoch": 0.4841272524426113, "grad_norm": 0.3277009129524231, "learning_rate": 9.764445542849738e-05, "loss": 0.065, "step": 14320 }, { "epoch": 0.4844653301328645, "grad_norm": 0.23610100150108337, "learning_rate": 9.763943753718998e-05, "loss": 0.0529, "step": 14330 }, { "epoch": 0.48480340782311776, "grad_norm": 0.20101957023143768, "learning_rate": 9.76344144361053e-05, "loss": 0.0619, "step": 14340 }, { "epoch": 0.485141485513371, "grad_norm": 0.27223631739616394, "learning_rate": 9.762938612579269e-05, "loss": 0.0576, "step": 14350 }, { "epoch": 0.4854795632036242, "grad_norm": 0.2743830680847168, "learning_rate": 9.762435260680202e-05, "loss": 0.0588, "step": 14360 }, { "epoch": 0.4858176408938774, "grad_norm": 0.3635033965110779, "learning_rate": 9.761931387968373e-05, "loss": 0.0595, "step": 14370 }, { "epoch": 0.48615571858413065, "grad_norm": 0.249695286154747, "learning_rate": 9.76142699449889e-05, "loss": 0.0624, "step": 14380 }, { "epoch": 0.4864937962743838, "grad_norm": 0.2848186194896698, "learning_rate": 9.760922080326908e-05, "loss": 0.064, "step": 14390 }, { "epoch": 0.48683187396463706, "grad_norm": 0.3358220160007477, "learning_rate": 9.760416645507644e-05, "loss": 0.0569, "step": 14400 }, { "epoch": 0.4871699516548903, "grad_norm": 0.15568917989730835, "learning_rate": 9.759910690096375e-05, "loss": 0.0576, "step": 14410 }, { "epoch": 0.48750802934514353, "grad_norm": 0.34807467460632324, "learning_rate": 9.759404214148429e-05, "loss": 0.0689, "step": 14420 }, { "epoch": 0.4878461070353967, "grad_norm": 0.2883445620536804, "learning_rate": 9.758897217719191e-05, "loss": 0.0601, "step": 14430 }, { "epoch": 0.48818418472564995, "grad_norm": 0.29611074924468994, "learning_rate": 9.758389700864113e-05, "loss": 0.0586, "step": 14440 }, { "epoch": 0.4885222624159032, "grad_norm": 0.2927727699279785, "learning_rate": 9.757881663638688e-05, "loss": 0.064, "step": 14450 }, { "epoch": 0.4888603401061564, "grad_norm": 0.23114904761314392, "learning_rate": 9.757373106098478e-05, "loss": 0.0635, "step": 14460 }, { "epoch": 0.4891984177964096, "grad_norm": 0.21926121413707733, "learning_rate": 9.756864028299097e-05, "loss": 0.0627, "step": 14470 }, { "epoch": 0.48953649548666284, "grad_norm": 0.25626131892204285, "learning_rate": 9.75635443029622e-05, "loss": 0.0651, "step": 14480 }, { "epoch": 0.4898745731769161, "grad_norm": 0.908906877040863, "learning_rate": 9.755844312145572e-05, "loss": 0.071, "step": 14490 }, { "epoch": 0.49021265086716925, "grad_norm": 0.19263431429862976, "learning_rate": 9.755333673902941e-05, "loss": 0.066, "step": 14500 }, { "epoch": 0.4905507285574225, "grad_norm": 0.2238280475139618, "learning_rate": 9.75482251562417e-05, "loss": 0.0708, "step": 14510 }, { "epoch": 0.4908888062476757, "grad_norm": 0.3805108070373535, "learning_rate": 9.754310837365155e-05, "loss": 0.0652, "step": 14520 }, { "epoch": 0.49122688393792896, "grad_norm": 0.2804245948791504, "learning_rate": 9.753798639181856e-05, "loss": 0.0656, "step": 14530 }, { "epoch": 0.49156496162818214, "grad_norm": 0.21990659832954407, "learning_rate": 9.753285921130286e-05, "loss": 0.0667, "step": 14540 }, { "epoch": 0.4919030393184354, "grad_norm": 0.29184624552726746, "learning_rate": 9.752772683266512e-05, "loss": 0.0673, "step": 14550 }, { "epoch": 0.4922411170086886, "grad_norm": 0.20076435804367065, "learning_rate": 9.752258925646665e-05, "loss": 0.0594, "step": 14560 }, { "epoch": 0.4925791946989418, "grad_norm": 0.30340418219566345, "learning_rate": 9.751744648326926e-05, "loss": 0.0704, "step": 14570 }, { "epoch": 0.492917272389195, "grad_norm": 0.25658515095710754, "learning_rate": 9.751229851363536e-05, "loss": 0.0655, "step": 14580 }, { "epoch": 0.49325535007944826, "grad_norm": 0.29428696632385254, "learning_rate": 9.750714534812793e-05, "loss": 0.0605, "step": 14590 }, { "epoch": 0.4935934277697015, "grad_norm": 0.19453264772891998, "learning_rate": 9.750198698731053e-05, "loss": 0.0634, "step": 14600 }, { "epoch": 0.4939315054599547, "grad_norm": 0.25169017910957336, "learning_rate": 9.749682343174722e-05, "loss": 0.0669, "step": 14610 }, { "epoch": 0.4942695831502079, "grad_norm": 0.2752957046031952, "learning_rate": 9.749165468200272e-05, "loss": 0.0645, "step": 14620 }, { "epoch": 0.49460766084046115, "grad_norm": 0.4781162142753601, "learning_rate": 9.748648073864229e-05, "loss": 0.0624, "step": 14630 }, { "epoch": 0.4949457385307144, "grad_norm": 0.26602286100387573, "learning_rate": 9.748130160223168e-05, "loss": 0.0664, "step": 14640 }, { "epoch": 0.49528381622096757, "grad_norm": 0.20746201276779175, "learning_rate": 9.747611727333734e-05, "loss": 0.0653, "step": 14650 }, { "epoch": 0.4956218939112208, "grad_norm": 0.1916465312242508, "learning_rate": 9.74709277525262e-05, "loss": 0.0596, "step": 14660 }, { "epoch": 0.49595997160147404, "grad_norm": 0.2597532272338867, "learning_rate": 9.746573304036576e-05, "loss": 0.0627, "step": 14670 }, { "epoch": 0.4962980492917272, "grad_norm": 0.24047234654426575, "learning_rate": 9.746053313742412e-05, "loss": 0.0614, "step": 14680 }, { "epoch": 0.49663612698198045, "grad_norm": 0.26762405037879944, "learning_rate": 9.745532804426994e-05, "loss": 0.0656, "step": 14690 }, { "epoch": 0.4969742046722337, "grad_norm": 0.2149885594844818, "learning_rate": 9.745011776147242e-05, "loss": 0.0531, "step": 14700 }, { "epoch": 0.4973122823624869, "grad_norm": 0.2935393452644348, "learning_rate": 9.744490228960138e-05, "loss": 0.0594, "step": 14710 }, { "epoch": 0.4976503600527401, "grad_norm": 0.3495163917541504, "learning_rate": 9.743968162922713e-05, "loss": 0.0631, "step": 14720 }, { "epoch": 0.49798843774299334, "grad_norm": 0.3233014941215515, "learning_rate": 9.743445578092064e-05, "loss": 0.0638, "step": 14730 }, { "epoch": 0.4983265154332466, "grad_norm": 0.30493780970573425, "learning_rate": 9.742922474525338e-05, "loss": 0.0627, "step": 14740 }, { "epoch": 0.49866459312349976, "grad_norm": 0.31959375739097595, "learning_rate": 9.742398852279741e-05, "loss": 0.0634, "step": 14750 }, { "epoch": 0.499002670813753, "grad_norm": 0.25574928522109985, "learning_rate": 9.741874711412535e-05, "loss": 0.0641, "step": 14760 }, { "epoch": 0.4993407485040062, "grad_norm": 0.2677786350250244, "learning_rate": 9.741350051981042e-05, "loss": 0.0629, "step": 14770 }, { "epoch": 0.49967882619425946, "grad_norm": 0.41169753670692444, "learning_rate": 9.740824874042633e-05, "loss": 0.0645, "step": 14780 }, { "epoch": 0.5000169038845127, "grad_norm": 0.21094343066215515, "learning_rate": 9.740299177654746e-05, "loss": 0.0611, "step": 14790 }, { "epoch": 0.5003549815747659, "grad_norm": 0.23178927600383759, "learning_rate": 9.739772962874867e-05, "loss": 0.0672, "step": 14800 }, { "epoch": 0.5006930592650191, "grad_norm": 0.32601091265678406, "learning_rate": 9.739246229760541e-05, "loss": 0.0612, "step": 14810 }, { "epoch": 0.5010311369552723, "grad_norm": 0.3603608012199402, "learning_rate": 9.738718978369376e-05, "loss": 0.0578, "step": 14820 }, { "epoch": 0.5013692146455255, "grad_norm": 0.29324594140052795, "learning_rate": 9.738191208759025e-05, "loss": 0.0636, "step": 14830 }, { "epoch": 0.5017072923357787, "grad_norm": 0.2900864779949188, "learning_rate": 9.73766292098721e-05, "loss": 0.0599, "step": 14840 }, { "epoch": 0.502045370026032, "grad_norm": 0.18012186884880066, "learning_rate": 9.737134115111699e-05, "loss": 0.0642, "step": 14850 }, { "epoch": 0.5023834477162852, "grad_norm": 0.19782723486423492, "learning_rate": 9.736604791190323e-05, "loss": 0.0607, "step": 14860 }, { "epoch": 0.5027215254065385, "grad_norm": 0.22381357848644257, "learning_rate": 9.73607494928097e-05, "loss": 0.0651, "step": 14870 }, { "epoch": 0.5030596030967917, "grad_norm": 0.22741274535655975, "learning_rate": 9.735544589441581e-05, "loss": 0.0628, "step": 14880 }, { "epoch": 0.5033976807870448, "grad_norm": 0.25889259576797485, "learning_rate": 9.735013711730154e-05, "loss": 0.0649, "step": 14890 }, { "epoch": 0.5037357584772981, "grad_norm": 0.21859371662139893, "learning_rate": 9.734482316204747e-05, "loss": 0.0654, "step": 14900 }, { "epoch": 0.5040738361675513, "grad_norm": 0.3193960189819336, "learning_rate": 9.733950402923473e-05, "loss": 0.0604, "step": 14910 }, { "epoch": 0.5044119138578045, "grad_norm": 0.25842002034187317, "learning_rate": 9.7334179719445e-05, "loss": 0.0692, "step": 14920 }, { "epoch": 0.5047499915480578, "grad_norm": 0.25601035356521606, "learning_rate": 9.732885023326053e-05, "loss": 0.0675, "step": 14930 }, { "epoch": 0.505088069238311, "grad_norm": 0.2028033584356308, "learning_rate": 9.732351557126418e-05, "loss": 0.0644, "step": 14940 }, { "epoch": 0.5054261469285641, "grad_norm": 0.358032763004303, "learning_rate": 9.731817573403929e-05, "loss": 0.0593, "step": 14950 }, { "epoch": 0.5057642246188174, "grad_norm": 0.24062494933605194, "learning_rate": 9.731283072216985e-05, "loss": 0.0588, "step": 14960 }, { "epoch": 0.5061023023090706, "grad_norm": 0.255098819732666, "learning_rate": 9.730748053624039e-05, "loss": 0.0631, "step": 14970 }, { "epoch": 0.5064403799993239, "grad_norm": 0.23988819122314453, "learning_rate": 9.730212517683598e-05, "loss": 0.0628, "step": 14980 }, { "epoch": 0.5067784576895771, "grad_norm": 0.2769767642021179, "learning_rate": 9.729676464454228e-05, "loss": 0.0562, "step": 14990 }, { "epoch": 0.5071165353798303, "grad_norm": 0.28953346610069275, "learning_rate": 9.72913989399455e-05, "loss": 0.0572, "step": 15000 }, { "epoch": 0.5074546130700835, "grad_norm": 0.2518891394138336, "learning_rate": 9.728602806363242e-05, "loss": 0.0642, "step": 15010 }, { "epoch": 0.5077926907603367, "grad_norm": 0.23917026817798615, "learning_rate": 9.728065201619043e-05, "loss": 0.061, "step": 15020 }, { "epoch": 0.5081307684505899, "grad_norm": 0.24747364223003387, "learning_rate": 9.727527079820742e-05, "loss": 0.0567, "step": 15030 }, { "epoch": 0.5084688461408432, "grad_norm": 0.24486270546913147, "learning_rate": 9.726988441027186e-05, "loss": 0.0606, "step": 15040 }, { "epoch": 0.5088069238310964, "grad_norm": 0.2148405909538269, "learning_rate": 9.726449285297281e-05, "loss": 0.0623, "step": 15050 }, { "epoch": 0.5091450015213496, "grad_norm": 0.18832777440547943, "learning_rate": 9.72590961268999e-05, "loss": 0.0649, "step": 15060 }, { "epoch": 0.5094830792116029, "grad_norm": 0.2663850486278534, "learning_rate": 9.725369423264328e-05, "loss": 0.0667, "step": 15070 }, { "epoch": 0.509821156901856, "grad_norm": 0.26787450909614563, "learning_rate": 9.72482871707937e-05, "loss": 0.0634, "step": 15080 }, { "epoch": 0.5101592345921092, "grad_norm": 0.30558717250823975, "learning_rate": 9.724287494194247e-05, "loss": 0.0685, "step": 15090 }, { "epoch": 0.5104973122823625, "grad_norm": 0.28323182463645935, "learning_rate": 9.723745754668147e-05, "loss": 0.0629, "step": 15100 }, { "epoch": 0.5108353899726157, "grad_norm": 0.2165553867816925, "learning_rate": 9.723203498560313e-05, "loss": 0.0693, "step": 15110 }, { "epoch": 0.511173467662869, "grad_norm": 0.1815594881772995, "learning_rate": 9.722660725930046e-05, "loss": 0.0655, "step": 15120 }, { "epoch": 0.5115115453531222, "grad_norm": 0.32606515288352966, "learning_rate": 9.722117436836702e-05, "loss": 0.0676, "step": 15130 }, { "epoch": 0.5118496230433753, "grad_norm": 0.2917892336845398, "learning_rate": 9.721573631339696e-05, "loss": 0.0599, "step": 15140 }, { "epoch": 0.5121877007336286, "grad_norm": 0.21962232887744904, "learning_rate": 9.721029309498494e-05, "loss": 0.0624, "step": 15150 }, { "epoch": 0.5125257784238818, "grad_norm": 0.24679848551750183, "learning_rate": 9.720484471372627e-05, "loss": 0.0626, "step": 15160 }, { "epoch": 0.512863856114135, "grad_norm": 0.23117919266223907, "learning_rate": 9.719939117021673e-05, "loss": 0.0645, "step": 15170 }, { "epoch": 0.5132019338043883, "grad_norm": 0.23733678460121155, "learning_rate": 9.719393246505275e-05, "loss": 0.0646, "step": 15180 }, { "epoch": 0.5135400114946415, "grad_norm": 0.20117589831352234, "learning_rate": 9.718846859883128e-05, "loss": 0.0645, "step": 15190 }, { "epoch": 0.5138780891848946, "grad_norm": 0.3570111393928528, "learning_rate": 9.718299957214982e-05, "loss": 0.0635, "step": 15200 }, { "epoch": 0.5142161668751479, "grad_norm": 0.18193642795085907, "learning_rate": 9.717752538560646e-05, "loss": 0.0608, "step": 15210 }, { "epoch": 0.5145542445654011, "grad_norm": 0.3822164237499237, "learning_rate": 9.717204603979986e-05, "loss": 0.0578, "step": 15220 }, { "epoch": 0.5148923222556544, "grad_norm": 0.3215552568435669, "learning_rate": 9.716656153532922e-05, "loss": 0.0633, "step": 15230 }, { "epoch": 0.5152303999459076, "grad_norm": 0.25757837295532227, "learning_rate": 9.716107187279434e-05, "loss": 0.0677, "step": 15240 }, { "epoch": 0.5155684776361608, "grad_norm": 0.26841795444488525, "learning_rate": 9.715557705279555e-05, "loss": 0.057, "step": 15250 }, { "epoch": 0.515906555326414, "grad_norm": 0.2917717695236206, "learning_rate": 9.715007707593372e-05, "loss": 0.0625, "step": 15260 }, { "epoch": 0.5162446330166672, "grad_norm": 0.39705848693847656, "learning_rate": 9.714457194281036e-05, "loss": 0.0666, "step": 15270 }, { "epoch": 0.5165827107069204, "grad_norm": 0.2182631641626358, "learning_rate": 9.713906165402751e-05, "loss": 0.0643, "step": 15280 }, { "epoch": 0.5169207883971737, "grad_norm": 0.2122381031513214, "learning_rate": 9.713354621018774e-05, "loss": 0.0632, "step": 15290 }, { "epoch": 0.5172588660874269, "grad_norm": 0.2556014955043793, "learning_rate": 9.712802561189422e-05, "loss": 0.0686, "step": 15300 }, { "epoch": 0.5175969437776801, "grad_norm": 0.3984125852584839, "learning_rate": 9.712249985975069e-05, "loss": 0.0587, "step": 15310 }, { "epoch": 0.5179350214679334, "grad_norm": 0.22065460681915283, "learning_rate": 9.71169689543614e-05, "loss": 0.0561, "step": 15320 }, { "epoch": 0.5182730991581865, "grad_norm": 0.33049309253692627, "learning_rate": 9.711143289633123e-05, "loss": 0.0682, "step": 15330 }, { "epoch": 0.5186111768484398, "grad_norm": 0.2525547444820404, "learning_rate": 9.710589168626561e-05, "loss": 0.059, "step": 15340 }, { "epoch": 0.518949254538693, "grad_norm": 0.2365540713071823, "learning_rate": 9.710034532477048e-05, "loss": 0.0626, "step": 15350 }, { "epoch": 0.5192873322289462, "grad_norm": 0.28680720925331116, "learning_rate": 9.709479381245239e-05, "loss": 0.0663, "step": 15360 }, { "epoch": 0.5196254099191995, "grad_norm": 0.2184801697731018, "learning_rate": 9.708923714991847e-05, "loss": 0.0618, "step": 15370 }, { "epoch": 0.5199634876094527, "grad_norm": 0.2384341061115265, "learning_rate": 9.708367533777638e-05, "loss": 0.0657, "step": 15380 }, { "epoch": 0.5203015652997058, "grad_norm": 0.19493362307548523, "learning_rate": 9.707810837663431e-05, "loss": 0.0681, "step": 15390 }, { "epoch": 0.5206396429899591, "grad_norm": 0.3230590224266052, "learning_rate": 9.707253626710113e-05, "loss": 0.0627, "step": 15400 }, { "epoch": 0.5209777206802123, "grad_norm": 0.26784271001815796, "learning_rate": 9.706695900978613e-05, "loss": 0.0603, "step": 15410 }, { "epoch": 0.5213157983704655, "grad_norm": 0.2463972568511963, "learning_rate": 9.706137660529926e-05, "loss": 0.0579, "step": 15420 }, { "epoch": 0.5216538760607188, "grad_norm": 0.19143176078796387, "learning_rate": 9.705578905425101e-05, "loss": 0.0651, "step": 15430 }, { "epoch": 0.521991953750972, "grad_norm": 0.1470576673746109, "learning_rate": 9.705019635725241e-05, "loss": 0.0609, "step": 15440 }, { "epoch": 0.5223300314412251, "grad_norm": 0.3389839828014374, "learning_rate": 9.704459851491508e-05, "loss": 0.0666, "step": 15450 }, { "epoch": 0.5226681091314784, "grad_norm": 0.22256337106227875, "learning_rate": 9.703899552785118e-05, "loss": 0.0565, "step": 15460 }, { "epoch": 0.5230061868217316, "grad_norm": 0.31028860807418823, "learning_rate": 9.703338739667346e-05, "loss": 0.0596, "step": 15470 }, { "epoch": 0.5233442645119849, "grad_norm": 0.30793634057044983, "learning_rate": 9.70277741219952e-05, "loss": 0.0697, "step": 15480 }, { "epoch": 0.5236823422022381, "grad_norm": 0.2851279377937317, "learning_rate": 9.702215570443027e-05, "loss": 0.0595, "step": 15490 }, { "epoch": 0.5240204198924913, "grad_norm": 0.20042894780635834, "learning_rate": 9.701653214459309e-05, "loss": 0.0642, "step": 15500 }, { "epoch": 0.5243584975827446, "grad_norm": 0.18594205379486084, "learning_rate": 9.701090344309865e-05, "loss": 0.0573, "step": 15510 }, { "epoch": 0.5246965752729977, "grad_norm": 0.314948171377182, "learning_rate": 9.700526960056247e-05, "loss": 0.0606, "step": 15520 }, { "epoch": 0.5250346529632509, "grad_norm": 0.41373926401138306, "learning_rate": 9.699963061760068e-05, "loss": 0.0603, "step": 15530 }, { "epoch": 0.5253727306535042, "grad_norm": 0.4168444871902466, "learning_rate": 9.699398649482997e-05, "loss": 0.0618, "step": 15540 }, { "epoch": 0.5257108083437574, "grad_norm": 0.22239649295806885, "learning_rate": 9.698833723286753e-05, "loss": 0.063, "step": 15550 }, { "epoch": 0.5260488860340106, "grad_norm": 0.28099897503852844, "learning_rate": 9.698268283233118e-05, "loss": 0.0636, "step": 15560 }, { "epoch": 0.5263869637242639, "grad_norm": 0.16068165004253387, "learning_rate": 9.697702329383929e-05, "loss": 0.0552, "step": 15570 }, { "epoch": 0.526725041414517, "grad_norm": 0.22055283188819885, "learning_rate": 9.697135861801074e-05, "loss": 0.0601, "step": 15580 }, { "epoch": 0.5270631191047703, "grad_norm": 0.25449150800704956, "learning_rate": 9.696568880546505e-05, "loss": 0.0645, "step": 15590 }, { "epoch": 0.5274011967950235, "grad_norm": 0.30970314145088196, "learning_rate": 9.696001385682223e-05, "loss": 0.0622, "step": 15600 }, { "epoch": 0.5277392744852767, "grad_norm": 0.21868190169334412, "learning_rate": 9.695433377270291e-05, "loss": 0.058, "step": 15610 }, { "epoch": 0.52807735217553, "grad_norm": 0.26228857040405273, "learning_rate": 9.694864855372824e-05, "loss": 0.0579, "step": 15620 }, { "epoch": 0.5284154298657832, "grad_norm": 0.2728894054889679, "learning_rate": 9.694295820051995e-05, "loss": 0.0639, "step": 15630 }, { "epoch": 0.5287535075560363, "grad_norm": 0.30406951904296875, "learning_rate": 9.693726271370032e-05, "loss": 0.0617, "step": 15640 }, { "epoch": 0.5290915852462896, "grad_norm": 0.2568521201610565, "learning_rate": 9.693156209389221e-05, "loss": 0.0593, "step": 15650 }, { "epoch": 0.5294296629365428, "grad_norm": 0.22522474825382233, "learning_rate": 9.692585634171905e-05, "loss": 0.0553, "step": 15660 }, { "epoch": 0.529767740626796, "grad_norm": 0.19905968010425568, "learning_rate": 9.692014545780476e-05, "loss": 0.062, "step": 15670 }, { "epoch": 0.5301058183170493, "grad_norm": 0.2231067568063736, "learning_rate": 9.691442944277393e-05, "loss": 0.0638, "step": 15680 }, { "epoch": 0.5304438960073025, "grad_norm": 0.21162660419940948, "learning_rate": 9.690870829725162e-05, "loss": 0.0613, "step": 15690 }, { "epoch": 0.5307819736975558, "grad_norm": 0.20240433514118195, "learning_rate": 9.69029820218635e-05, "loss": 0.059, "step": 15700 }, { "epoch": 0.5311200513878089, "grad_norm": 0.342303603887558, "learning_rate": 9.689725061723579e-05, "loss": 0.0665, "step": 15710 }, { "epoch": 0.5314581290780621, "grad_norm": 0.40522316098213196, "learning_rate": 9.689151408399527e-05, "loss": 0.0619, "step": 15720 }, { "epoch": 0.5317962067683154, "grad_norm": 0.17561925947666168, "learning_rate": 9.688577242276924e-05, "loss": 0.0671, "step": 15730 }, { "epoch": 0.5321342844585686, "grad_norm": 0.24257366359233856, "learning_rate": 9.688002563418566e-05, "loss": 0.0581, "step": 15740 }, { "epoch": 0.5324723621488218, "grad_norm": 0.19780229032039642, "learning_rate": 9.687427371887293e-05, "loss": 0.0765, "step": 15750 }, { "epoch": 0.5328104398390751, "grad_norm": 0.21384009718894958, "learning_rate": 9.686851667746012e-05, "loss": 0.0614, "step": 15760 }, { "epoch": 0.5331485175293282, "grad_norm": 0.1984397917985916, "learning_rate": 9.686275451057677e-05, "loss": 0.062, "step": 15770 }, { "epoch": 0.5334865952195814, "grad_norm": 0.34466442465782166, "learning_rate": 9.685698721885308e-05, "loss": 0.0571, "step": 15780 }, { "epoch": 0.5338246729098347, "grad_norm": 0.25687161087989807, "learning_rate": 9.68512148029197e-05, "loss": 0.0626, "step": 15790 }, { "epoch": 0.5341627506000879, "grad_norm": 0.22040772438049316, "learning_rate": 9.684543726340791e-05, "loss": 0.0632, "step": 15800 }, { "epoch": 0.5345008282903411, "grad_norm": 0.32397517561912537, "learning_rate": 9.683965460094952e-05, "loss": 0.0676, "step": 15810 }, { "epoch": 0.5348389059805944, "grad_norm": 0.2593892216682434, "learning_rate": 9.683386681617694e-05, "loss": 0.0645, "step": 15820 }, { "epoch": 0.5351769836708475, "grad_norm": 0.33299627900123596, "learning_rate": 9.68280739097231e-05, "loss": 0.0644, "step": 15830 }, { "epoch": 0.5355150613611008, "grad_norm": 0.18758590519428253, "learning_rate": 9.682227588222148e-05, "loss": 0.0595, "step": 15840 }, { "epoch": 0.535853139051354, "grad_norm": 0.29039210081100464, "learning_rate": 9.681647273430618e-05, "loss": 0.0615, "step": 15850 }, { "epoch": 0.5361912167416072, "grad_norm": 0.2803020477294922, "learning_rate": 9.681066446661182e-05, "loss": 0.07, "step": 15860 }, { "epoch": 0.5365292944318605, "grad_norm": 0.22870057821273804, "learning_rate": 9.680485107977357e-05, "loss": 0.0585, "step": 15870 }, { "epoch": 0.5368673721221137, "grad_norm": 0.32002443075180054, "learning_rate": 9.679903257442716e-05, "loss": 0.0609, "step": 15880 }, { "epoch": 0.5372054498123668, "grad_norm": 0.3312406837940216, "learning_rate": 9.679320895120891e-05, "loss": 0.0638, "step": 15890 }, { "epoch": 0.5375435275026201, "grad_norm": 0.297636479139328, "learning_rate": 9.67873802107557e-05, "loss": 0.0656, "step": 15900 }, { "epoch": 0.5378816051928733, "grad_norm": 0.2672008275985718, "learning_rate": 9.67815463537049e-05, "loss": 0.057, "step": 15910 }, { "epoch": 0.5382196828831265, "grad_norm": 0.17040212452411652, "learning_rate": 9.677570738069457e-05, "loss": 0.0594, "step": 15920 }, { "epoch": 0.5385577605733798, "grad_norm": 0.3694249093532562, "learning_rate": 9.676986329236318e-05, "loss": 0.0655, "step": 15930 }, { "epoch": 0.538895838263633, "grad_norm": 0.22118721902370453, "learning_rate": 9.676401408934987e-05, "loss": 0.0644, "step": 15940 }, { "epoch": 0.5392339159538863, "grad_norm": 0.2894432544708252, "learning_rate": 9.675815977229428e-05, "loss": 0.064, "step": 15950 }, { "epoch": 0.5395719936441394, "grad_norm": 0.41510382294654846, "learning_rate": 9.675230034183664e-05, "loss": 0.0606, "step": 15960 }, { "epoch": 0.5399100713343926, "grad_norm": 0.34010040760040283, "learning_rate": 9.674643579861773e-05, "loss": 0.0672, "step": 15970 }, { "epoch": 0.5402481490246459, "grad_norm": 0.23306632041931152, "learning_rate": 9.674056614327886e-05, "loss": 0.0662, "step": 15980 }, { "epoch": 0.5405862267148991, "grad_norm": 0.30145126581192017, "learning_rate": 9.673469137646198e-05, "loss": 0.0604, "step": 15990 }, { "epoch": 0.5409243044051523, "grad_norm": 0.28346899151802063, "learning_rate": 9.67288114988095e-05, "loss": 0.0604, "step": 16000 }, { "epoch": 0.5412623820954056, "grad_norm": 0.3118065595626831, "learning_rate": 9.672292651096447e-05, "loss": 0.0626, "step": 16010 }, { "epoch": 0.5416004597856587, "grad_norm": 0.3335067629814148, "learning_rate": 9.671703641357042e-05, "loss": 0.0641, "step": 16020 }, { "epoch": 0.5419385374759119, "grad_norm": 0.2387707382440567, "learning_rate": 9.67111412072715e-05, "loss": 0.0561, "step": 16030 }, { "epoch": 0.5422766151661652, "grad_norm": 0.2084580659866333, "learning_rate": 9.670524089271242e-05, "loss": 0.061, "step": 16040 }, { "epoch": 0.5426146928564184, "grad_norm": 0.26723459362983704, "learning_rate": 9.669933547053842e-05, "loss": 0.0581, "step": 16050 }, { "epoch": 0.5429527705466716, "grad_norm": 0.21436697244644165, "learning_rate": 9.669342494139531e-05, "loss": 0.0661, "step": 16060 }, { "epoch": 0.5432908482369249, "grad_norm": 0.2186778336763382, "learning_rate": 9.668750930592943e-05, "loss": 0.0587, "step": 16070 }, { "epoch": 0.543628925927178, "grad_norm": 0.6784757375717163, "learning_rate": 9.668158856478775e-05, "loss": 0.0662, "step": 16080 }, { "epoch": 0.5439670036174313, "grad_norm": 0.24549689888954163, "learning_rate": 9.66756627186177e-05, "loss": 0.0649, "step": 16090 }, { "epoch": 0.5443050813076845, "grad_norm": 0.2389453947544098, "learning_rate": 9.666973176806737e-05, "loss": 0.0589, "step": 16100 }, { "epoch": 0.5446431589979377, "grad_norm": 0.18839038908481598, "learning_rate": 9.666379571378534e-05, "loss": 0.0559, "step": 16110 }, { "epoch": 0.544981236688191, "grad_norm": 0.4650140702724457, "learning_rate": 9.665785455642076e-05, "loss": 0.0676, "step": 16120 }, { "epoch": 0.5453193143784442, "grad_norm": 0.3069145381450653, "learning_rate": 9.665190829662337e-05, "loss": 0.0637, "step": 16130 }, { "epoch": 0.5456573920686973, "grad_norm": 0.20443467795848846, "learning_rate": 9.664595693504342e-05, "loss": 0.0646, "step": 16140 }, { "epoch": 0.5459954697589506, "grad_norm": 0.22630777955055237, "learning_rate": 9.664000047233175e-05, "loss": 0.0636, "step": 16150 }, { "epoch": 0.5463335474492038, "grad_norm": 0.20740719139575958, "learning_rate": 9.663403890913976e-05, "loss": 0.0605, "step": 16160 }, { "epoch": 0.546671625139457, "grad_norm": 0.3458676338195801, "learning_rate": 9.662807224611938e-05, "loss": 0.0603, "step": 16170 }, { "epoch": 0.5470097028297103, "grad_norm": 0.27012956142425537, "learning_rate": 9.662210048392311e-05, "loss": 0.0629, "step": 16180 }, { "epoch": 0.5473477805199635, "grad_norm": 0.23244138062000275, "learning_rate": 9.661612362320405e-05, "loss": 0.0628, "step": 16190 }, { "epoch": 0.5476858582102168, "grad_norm": 0.21142733097076416, "learning_rate": 9.661014166461579e-05, "loss": 0.063, "step": 16200 }, { "epoch": 0.5480239359004699, "grad_norm": 0.2458500862121582, "learning_rate": 9.66041546088125e-05, "loss": 0.0547, "step": 16210 }, { "epoch": 0.5483620135907231, "grad_norm": 0.2478354126214981, "learning_rate": 9.659816245644895e-05, "loss": 0.0567, "step": 16220 }, { "epoch": 0.5487000912809764, "grad_norm": 0.1799921840429306, "learning_rate": 9.65921652081804e-05, "loss": 0.065, "step": 16230 }, { "epoch": 0.5490381689712296, "grad_norm": 0.24844039976596832, "learning_rate": 9.658616286466271e-05, "loss": 0.0619, "step": 16240 }, { "epoch": 0.5493762466614828, "grad_norm": 0.2067224234342575, "learning_rate": 9.65801554265523e-05, "loss": 0.0614, "step": 16250 }, { "epoch": 0.5497143243517361, "grad_norm": 0.21342967450618744, "learning_rate": 9.657414289450612e-05, "loss": 0.066, "step": 16260 }, { "epoch": 0.5500524020419892, "grad_norm": 0.23814216256141663, "learning_rate": 9.656812526918171e-05, "loss": 0.0609, "step": 16270 }, { "epoch": 0.5503904797322424, "grad_norm": 0.18984417617321014, "learning_rate": 9.656210255123712e-05, "loss": 0.057, "step": 16280 }, { "epoch": 0.5507285574224957, "grad_norm": 0.19806860387325287, "learning_rate": 9.6556074741331e-05, "loss": 0.0578, "step": 16290 }, { "epoch": 0.5510666351127489, "grad_norm": 0.23422269523143768, "learning_rate": 9.655004184012256e-05, "loss": 0.0638, "step": 16300 }, { "epoch": 0.5514047128030022, "grad_norm": 0.21438567340373993, "learning_rate": 9.654400384827152e-05, "loss": 0.0582, "step": 16310 }, { "epoch": 0.5517427904932554, "grad_norm": 0.1623321920633316, "learning_rate": 9.653796076643818e-05, "loss": 0.0543, "step": 16320 }, { "epoch": 0.5520808681835085, "grad_norm": 0.2547948956489563, "learning_rate": 9.653191259528344e-05, "loss": 0.0634, "step": 16330 }, { "epoch": 0.5524189458737618, "grad_norm": 0.25285711884498596, "learning_rate": 9.65258593354687e-05, "loss": 0.0591, "step": 16340 }, { "epoch": 0.552757023564015, "grad_norm": 0.30073970556259155, "learning_rate": 9.651980098765591e-05, "loss": 0.0595, "step": 16350 }, { "epoch": 0.5530951012542682, "grad_norm": 0.24977314472198486, "learning_rate": 9.651373755250765e-05, "loss": 0.0585, "step": 16360 }, { "epoch": 0.5534331789445215, "grad_norm": 0.20518961548805237, "learning_rate": 9.650766903068697e-05, "loss": 0.0606, "step": 16370 }, { "epoch": 0.5537712566347747, "grad_norm": 0.29045370221138, "learning_rate": 9.650159542285753e-05, "loss": 0.0619, "step": 16380 }, { "epoch": 0.5541093343250278, "grad_norm": 0.2899528741836548, "learning_rate": 9.649551672968353e-05, "loss": 0.0609, "step": 16390 }, { "epoch": 0.5544474120152811, "grad_norm": 0.17746444046497345, "learning_rate": 9.648943295182973e-05, "loss": 0.0635, "step": 16400 }, { "epoch": 0.5547854897055343, "grad_norm": 0.2557869255542755, "learning_rate": 9.648334408996144e-05, "loss": 0.0587, "step": 16410 }, { "epoch": 0.5551235673957875, "grad_norm": 0.2977246344089508, "learning_rate": 9.647725014474452e-05, "loss": 0.0605, "step": 16420 }, { "epoch": 0.5554616450860408, "grad_norm": 0.39629125595092773, "learning_rate": 9.64711511168454e-05, "loss": 0.0656, "step": 16430 }, { "epoch": 0.555799722776294, "grad_norm": 0.3751155734062195, "learning_rate": 9.646504700693108e-05, "loss": 0.0643, "step": 16440 }, { "epoch": 0.5561378004665473, "grad_norm": 0.30383768677711487, "learning_rate": 9.645893781566907e-05, "loss": 0.0586, "step": 16450 }, { "epoch": 0.5564758781568004, "grad_norm": 0.3798382580280304, "learning_rate": 9.645282354372744e-05, "loss": 0.0613, "step": 16460 }, { "epoch": 0.5568139558470536, "grad_norm": 0.29440170526504517, "learning_rate": 9.644670419177491e-05, "loss": 0.0594, "step": 16470 }, { "epoch": 0.5571520335373069, "grad_norm": 0.3147638440132141, "learning_rate": 9.644057976048062e-05, "loss": 0.0618, "step": 16480 }, { "epoch": 0.5574901112275601, "grad_norm": 0.19819800555706024, "learning_rate": 9.643445025051435e-05, "loss": 0.065, "step": 16490 }, { "epoch": 0.5578281889178133, "grad_norm": 0.2040507197380066, "learning_rate": 9.642831566254641e-05, "loss": 0.0711, "step": 16500 }, { "epoch": 0.5581662666080666, "grad_norm": 0.23151880502700806, "learning_rate": 9.642217599724769e-05, "loss": 0.0626, "step": 16510 }, { "epoch": 0.5585043442983197, "grad_norm": 0.22866404056549072, "learning_rate": 9.64160312552896e-05, "loss": 0.0586, "step": 16520 }, { "epoch": 0.5588424219885729, "grad_norm": 0.22885587811470032, "learning_rate": 9.64098814373441e-05, "loss": 0.0583, "step": 16530 }, { "epoch": 0.5591804996788262, "grad_norm": 0.25523266196250916, "learning_rate": 9.640372654408374e-05, "loss": 0.0652, "step": 16540 }, { "epoch": 0.5595185773690794, "grad_norm": 0.19418451189994812, "learning_rate": 9.639756657618162e-05, "loss": 0.0671, "step": 16550 }, { "epoch": 0.5598566550593327, "grad_norm": 0.24616746604442596, "learning_rate": 9.639140153431138e-05, "loss": 0.0634, "step": 16560 }, { "epoch": 0.5601947327495859, "grad_norm": 0.19098550081253052, "learning_rate": 9.638523141914721e-05, "loss": 0.0614, "step": 16570 }, { "epoch": 0.560532810439839, "grad_norm": 0.2607516944408417, "learning_rate": 9.637905623136388e-05, "loss": 0.0631, "step": 16580 }, { "epoch": 0.5608708881300923, "grad_norm": 0.2892821431159973, "learning_rate": 9.637287597163669e-05, "loss": 0.0627, "step": 16590 }, { "epoch": 0.5612089658203455, "grad_norm": 0.288777619600296, "learning_rate": 9.63666906406415e-05, "loss": 0.0602, "step": 16600 }, { "epoch": 0.5615470435105987, "grad_norm": 0.224053755402565, "learning_rate": 9.636050023905473e-05, "loss": 0.0604, "step": 16610 }, { "epoch": 0.561885121200852, "grad_norm": 0.27503469586372375, "learning_rate": 9.635430476755336e-05, "loss": 0.061, "step": 16620 }, { "epoch": 0.5622231988911052, "grad_norm": 0.3050718605518341, "learning_rate": 9.63481042268149e-05, "loss": 0.0646, "step": 16630 }, { "epoch": 0.5625612765813583, "grad_norm": 0.296471506357193, "learning_rate": 9.634189861751745e-05, "loss": 0.0614, "step": 16640 }, { "epoch": 0.5628993542716116, "grad_norm": 0.2032569795846939, "learning_rate": 9.633568794033967e-05, "loss": 0.0608, "step": 16650 }, { "epoch": 0.5632374319618648, "grad_norm": 0.28469371795654297, "learning_rate": 9.63294721959607e-05, "loss": 0.0617, "step": 16660 }, { "epoch": 0.5635755096521181, "grad_norm": 0.2721666097640991, "learning_rate": 9.63232513850603e-05, "loss": 0.0622, "step": 16670 }, { "epoch": 0.5639135873423713, "grad_norm": 0.40959393978118896, "learning_rate": 9.631702550831878e-05, "loss": 0.0593, "step": 16680 }, { "epoch": 0.5642516650326245, "grad_norm": 0.395059734582901, "learning_rate": 9.631079456641698e-05, "loss": 0.0664, "step": 16690 }, { "epoch": 0.5645897427228778, "grad_norm": 0.2597099244594574, "learning_rate": 9.630455856003632e-05, "loss": 0.0663, "step": 16700 }, { "epoch": 0.5649278204131309, "grad_norm": 0.2312452644109726, "learning_rate": 9.629831748985876e-05, "loss": 0.0665, "step": 16710 }, { "epoch": 0.5652658981033841, "grad_norm": 0.39793524146080017, "learning_rate": 9.629207135656679e-05, "loss": 0.0627, "step": 16720 }, { "epoch": 0.5656039757936374, "grad_norm": 0.30830201506614685, "learning_rate": 9.628582016084353e-05, "loss": 0.0606, "step": 16730 }, { "epoch": 0.5659420534838906, "grad_norm": 0.3700035810470581, "learning_rate": 9.627956390337254e-05, "loss": 0.0559, "step": 16740 }, { "epoch": 0.5662801311741438, "grad_norm": 0.33752813935279846, "learning_rate": 9.627330258483802e-05, "loss": 0.0684, "step": 16750 }, { "epoch": 0.5666182088643971, "grad_norm": 0.19562101364135742, "learning_rate": 9.62670362059247e-05, "loss": 0.0711, "step": 16760 }, { "epoch": 0.5669562865546502, "grad_norm": 0.2663753628730774, "learning_rate": 9.626076476731786e-05, "loss": 0.0699, "step": 16770 }, { "epoch": 0.5672943642449034, "grad_norm": 0.2861669361591339, "learning_rate": 9.625448826970336e-05, "loss": 0.0601, "step": 16780 }, { "epoch": 0.5676324419351567, "grad_norm": 0.24444778263568878, "learning_rate": 9.624820671376755e-05, "loss": 0.0662, "step": 16790 }, { "epoch": 0.5679705196254099, "grad_norm": 0.16997374594211578, "learning_rate": 9.62419201001974e-05, "loss": 0.0579, "step": 16800 }, { "epoch": 0.5683085973156632, "grad_norm": 0.21485555171966553, "learning_rate": 9.623562842968037e-05, "loss": 0.0585, "step": 16810 }, { "epoch": 0.5686466750059164, "grad_norm": 0.2086489498615265, "learning_rate": 9.622933170290454e-05, "loss": 0.0632, "step": 16820 }, { "epoch": 0.5689847526961695, "grad_norm": 0.21802952885627747, "learning_rate": 9.622302992055849e-05, "loss": 0.0552, "step": 16830 }, { "epoch": 0.5693228303864228, "grad_norm": 0.2936471700668335, "learning_rate": 9.62167230833314e-05, "loss": 0.0614, "step": 16840 }, { "epoch": 0.569660908076676, "grad_norm": 0.19449873268604279, "learning_rate": 9.621041119191295e-05, "loss": 0.0633, "step": 16850 }, { "epoch": 0.5699989857669292, "grad_norm": 0.2638244926929474, "learning_rate": 9.620409424699342e-05, "loss": 0.064, "step": 16860 }, { "epoch": 0.5703370634571825, "grad_norm": 0.208283931016922, "learning_rate": 9.619777224926359e-05, "loss": 0.0615, "step": 16870 }, { "epoch": 0.5706751411474357, "grad_norm": 0.3397578299045563, "learning_rate": 9.619144519941485e-05, "loss": 0.0594, "step": 16880 }, { "epoch": 0.5710132188376889, "grad_norm": 0.24007323384284973, "learning_rate": 9.618511309813912e-05, "loss": 0.0581, "step": 16890 }, { "epoch": 0.5713512965279421, "grad_norm": 0.3077954649925232, "learning_rate": 9.617877594612886e-05, "loss": 0.0646, "step": 16900 }, { "epoch": 0.5716893742181953, "grad_norm": 0.3171347677707672, "learning_rate": 9.617243374407707e-05, "loss": 0.0602, "step": 16910 }, { "epoch": 0.5720274519084486, "grad_norm": 0.3563286066055298, "learning_rate": 9.616608649267736e-05, "loss": 0.0656, "step": 16920 }, { "epoch": 0.5723655295987018, "grad_norm": 0.25954097509384155, "learning_rate": 9.615973419262385e-05, "loss": 0.0638, "step": 16930 }, { "epoch": 0.572703607288955, "grad_norm": 0.19893549382686615, "learning_rate": 9.615337684461119e-05, "loss": 0.0595, "step": 16940 }, { "epoch": 0.5730416849792083, "grad_norm": 0.34226900339126587, "learning_rate": 9.614701444933465e-05, "loss": 0.0593, "step": 16950 }, { "epoch": 0.5733797626694614, "grad_norm": 0.257919043302536, "learning_rate": 9.614064700748997e-05, "loss": 0.0564, "step": 16960 }, { "epoch": 0.5737178403597146, "grad_norm": 0.1678401678800583, "learning_rate": 9.613427451977352e-05, "loss": 0.0594, "step": 16970 }, { "epoch": 0.5740559180499679, "grad_norm": 0.2244446724653244, "learning_rate": 9.612789698688216e-05, "loss": 0.0587, "step": 16980 }, { "epoch": 0.5743939957402211, "grad_norm": 0.19622333347797394, "learning_rate": 9.612151440951334e-05, "loss": 0.0613, "step": 16990 }, { "epoch": 0.5747320734304743, "grad_norm": 0.3196960389614105, "learning_rate": 9.611512678836506e-05, "loss": 0.0602, "step": 17000 }, { "epoch": 0.5750701511207276, "grad_norm": 0.36043471097946167, "learning_rate": 9.610873412413584e-05, "loss": 0.0601, "step": 17010 }, { "epoch": 0.5754082288109807, "grad_norm": 0.21871821582317352, "learning_rate": 9.610233641752476e-05, "loss": 0.0644, "step": 17020 }, { "epoch": 0.5757463065012339, "grad_norm": 0.29127347469329834, "learning_rate": 9.609593366923151e-05, "loss": 0.0663, "step": 17030 }, { "epoch": 0.5760843841914872, "grad_norm": 0.2946361005306244, "learning_rate": 9.608952587995625e-05, "loss": 0.0605, "step": 17040 }, { "epoch": 0.5764224618817404, "grad_norm": 0.22194428741931915, "learning_rate": 9.608311305039972e-05, "loss": 0.0673, "step": 17050 }, { "epoch": 0.5767605395719937, "grad_norm": 0.27819058299064636, "learning_rate": 9.607669518126326e-05, "loss": 0.0603, "step": 17060 }, { "epoch": 0.5770986172622469, "grad_norm": 0.28373709321022034, "learning_rate": 9.607027227324866e-05, "loss": 0.0629, "step": 17070 }, { "epoch": 0.5774366949525, "grad_norm": 0.2503490149974823, "learning_rate": 9.606384432705837e-05, "loss": 0.0642, "step": 17080 }, { "epoch": 0.5777747726427533, "grad_norm": 0.20308037102222443, "learning_rate": 9.60574113433953e-05, "loss": 0.0636, "step": 17090 }, { "epoch": 0.5781128503330065, "grad_norm": 0.28401097655296326, "learning_rate": 9.6050973322963e-05, "loss": 0.0628, "step": 17100 }, { "epoch": 0.5784509280232597, "grad_norm": 0.23342221975326538, "learning_rate": 9.604453026646547e-05, "loss": 0.0667, "step": 17110 }, { "epoch": 0.578789005713513, "grad_norm": 0.25598421692848206, "learning_rate": 9.603808217460735e-05, "loss": 0.062, "step": 17120 }, { "epoch": 0.5791270834037662, "grad_norm": 0.31231001019477844, "learning_rate": 9.603162904809377e-05, "loss": 0.069, "step": 17130 }, { "epoch": 0.5794651610940194, "grad_norm": 0.1759355366230011, "learning_rate": 9.602517088763045e-05, "loss": 0.0579, "step": 17140 }, { "epoch": 0.5798032387842726, "grad_norm": 0.23432399332523346, "learning_rate": 9.601870769392365e-05, "loss": 0.0614, "step": 17150 }, { "epoch": 0.5801413164745258, "grad_norm": 0.2175590991973877, "learning_rate": 9.601223946768017e-05, "loss": 0.0592, "step": 17160 }, { "epoch": 0.5804793941647791, "grad_norm": 0.1737203747034073, "learning_rate": 9.600576620960734e-05, "loss": 0.0607, "step": 17170 }, { "epoch": 0.5808174718550323, "grad_norm": 0.3028026819229126, "learning_rate": 9.599928792041308e-05, "loss": 0.0556, "step": 17180 }, { "epoch": 0.5811555495452855, "grad_norm": 0.32373887300491333, "learning_rate": 9.599280460080587e-05, "loss": 0.0618, "step": 17190 }, { "epoch": 0.5814936272355388, "grad_norm": 0.23744894564151764, "learning_rate": 9.59863162514947e-05, "loss": 0.0591, "step": 17200 }, { "epoch": 0.581831704925792, "grad_norm": 0.2765289545059204, "learning_rate": 9.597982287318911e-05, "loss": 0.0695, "step": 17210 }, { "epoch": 0.5821697826160451, "grad_norm": 0.22298270463943481, "learning_rate": 9.597332446659923e-05, "loss": 0.0651, "step": 17220 }, { "epoch": 0.5825078603062984, "grad_norm": 0.2550606429576874, "learning_rate": 9.59668210324357e-05, "loss": 0.0583, "step": 17230 }, { "epoch": 0.5828459379965516, "grad_norm": 0.21300029754638672, "learning_rate": 9.596031257140974e-05, "loss": 0.0544, "step": 17240 }, { "epoch": 0.5831840156868048, "grad_norm": 0.22604182362556458, "learning_rate": 9.59537990842331e-05, "loss": 0.0606, "step": 17250 }, { "epoch": 0.5835220933770581, "grad_norm": 0.22288046777248383, "learning_rate": 9.594728057161806e-05, "loss": 0.0594, "step": 17260 }, { "epoch": 0.5838601710673113, "grad_norm": 0.22451840341091156, "learning_rate": 9.594075703427752e-05, "loss": 0.0554, "step": 17270 }, { "epoch": 0.5841982487575645, "grad_norm": 0.34244799613952637, "learning_rate": 9.593422847292486e-05, "loss": 0.0559, "step": 17280 }, { "epoch": 0.5845363264478177, "grad_norm": 0.29905685782432556, "learning_rate": 9.592769488827402e-05, "loss": 0.0651, "step": 17290 }, { "epoch": 0.5848744041380709, "grad_norm": 0.2809232473373413, "learning_rate": 9.592115628103952e-05, "loss": 0.0605, "step": 17300 }, { "epoch": 0.5852124818283242, "grad_norm": 0.20451965928077698, "learning_rate": 9.591461265193643e-05, "loss": 0.0619, "step": 17310 }, { "epoch": 0.5855505595185774, "grad_norm": 0.3593059480190277, "learning_rate": 9.590806400168032e-05, "loss": 0.0625, "step": 17320 }, { "epoch": 0.5858886372088306, "grad_norm": 0.3884523808956146, "learning_rate": 9.590151033098735e-05, "loss": 0.0613, "step": 17330 }, { "epoch": 0.5862267148990838, "grad_norm": 0.25320112705230713, "learning_rate": 9.589495164057423e-05, "loss": 0.0556, "step": 17340 }, { "epoch": 0.586564792589337, "grad_norm": 0.18820376694202423, "learning_rate": 9.58883879311582e-05, "loss": 0.0641, "step": 17350 }, { "epoch": 0.5869028702795902, "grad_norm": 0.23400141298770905, "learning_rate": 9.588181920345705e-05, "loss": 0.0642, "step": 17360 }, { "epoch": 0.5872409479698435, "grad_norm": 0.24691030383110046, "learning_rate": 9.587524545818913e-05, "loss": 0.0666, "step": 17370 }, { "epoch": 0.5875790256600967, "grad_norm": 0.24796828627586365, "learning_rate": 9.586866669607335e-05, "loss": 0.0602, "step": 17380 }, { "epoch": 0.5879171033503499, "grad_norm": 0.23755845427513123, "learning_rate": 9.586208291782915e-05, "loss": 0.06, "step": 17390 }, { "epoch": 0.5882551810406031, "grad_norm": 0.3065384030342102, "learning_rate": 9.58554941241765e-05, "loss": 0.0632, "step": 17400 }, { "epoch": 0.5885932587308563, "grad_norm": 0.20776119828224182, "learning_rate": 9.584890031583596e-05, "loss": 0.066, "step": 17410 }, { "epoch": 0.5889313364211096, "grad_norm": 0.2523539662361145, "learning_rate": 9.584230149352861e-05, "loss": 0.0642, "step": 17420 }, { "epoch": 0.5892694141113628, "grad_norm": 0.23922675848007202, "learning_rate": 9.58356976579761e-05, "loss": 0.068, "step": 17430 }, { "epoch": 0.589607491801616, "grad_norm": 0.30821654200553894, "learning_rate": 9.58290888099006e-05, "loss": 0.0668, "step": 17440 }, { "epoch": 0.5899455694918693, "grad_norm": 0.45358866453170776, "learning_rate": 9.582247495002486e-05, "loss": 0.0662, "step": 17450 }, { "epoch": 0.5902836471821225, "grad_norm": 0.23609179258346558, "learning_rate": 9.581585607907214e-05, "loss": 0.0651, "step": 17460 }, { "epoch": 0.5906217248723756, "grad_norm": 0.23739689588546753, "learning_rate": 9.580923219776628e-05, "loss": 0.0654, "step": 17470 }, { "epoch": 0.5909598025626289, "grad_norm": 0.2141401320695877, "learning_rate": 9.580260330683167e-05, "loss": 0.0622, "step": 17480 }, { "epoch": 0.5912978802528821, "grad_norm": 0.3752254843711853, "learning_rate": 9.579596940699322e-05, "loss": 0.07, "step": 17490 }, { "epoch": 0.5916359579431353, "grad_norm": 0.3858216106891632, "learning_rate": 9.578933049897643e-05, "loss": 0.06, "step": 17500 }, { "epoch": 0.5919740356333886, "grad_norm": 0.2240355759859085, "learning_rate": 9.578268658350728e-05, "loss": 0.0586, "step": 17510 }, { "epoch": 0.5923121133236418, "grad_norm": 0.20591047406196594, "learning_rate": 9.577603766131235e-05, "loss": 0.0658, "step": 17520 }, { "epoch": 0.592650191013895, "grad_norm": 0.3072914183139801, "learning_rate": 9.576938373311878e-05, "loss": 0.0627, "step": 17530 }, { "epoch": 0.5929882687041482, "grad_norm": 0.21667861938476562, "learning_rate": 9.576272479965421e-05, "loss": 0.0562, "step": 17540 }, { "epoch": 0.5933263463944014, "grad_norm": 0.6494945883750916, "learning_rate": 9.575606086164687e-05, "loss": 0.0628, "step": 17550 }, { "epoch": 0.5936644240846547, "grad_norm": 0.29204821586608887, "learning_rate": 9.57493919198255e-05, "loss": 0.0657, "step": 17560 }, { "epoch": 0.5940025017749079, "grad_norm": 0.17932415008544922, "learning_rate": 9.57427179749194e-05, "loss": 0.0671, "step": 17570 }, { "epoch": 0.5943405794651611, "grad_norm": 0.24307265877723694, "learning_rate": 9.573603902765846e-05, "loss": 0.0647, "step": 17580 }, { "epoch": 0.5946786571554143, "grad_norm": 0.36420148611068726, "learning_rate": 9.572935507877304e-05, "loss": 0.0681, "step": 17590 }, { "epoch": 0.5950167348456675, "grad_norm": 0.2849215567111969, "learning_rate": 9.57226661289941e-05, "loss": 0.0623, "step": 17600 }, { "epoch": 0.5953548125359207, "grad_norm": 0.38382819294929504, "learning_rate": 9.571597217905315e-05, "loss": 0.0561, "step": 17610 }, { "epoch": 0.595692890226174, "grad_norm": 0.31254249811172485, "learning_rate": 9.57092732296822e-05, "loss": 0.0618, "step": 17620 }, { "epoch": 0.5960309679164272, "grad_norm": 0.216435968875885, "learning_rate": 9.570256928161385e-05, "loss": 0.0636, "step": 17630 }, { "epoch": 0.5963690456066805, "grad_norm": 0.36668282747268677, "learning_rate": 9.569586033558126e-05, "loss": 0.0637, "step": 17640 }, { "epoch": 0.5967071232969337, "grad_norm": 0.30737048387527466, "learning_rate": 9.568914639231807e-05, "loss": 0.059, "step": 17650 }, { "epoch": 0.5970452009871868, "grad_norm": 0.2597217857837677, "learning_rate": 9.568242745255852e-05, "loss": 0.0618, "step": 17660 }, { "epoch": 0.5973832786774401, "grad_norm": 0.2791704535484314, "learning_rate": 9.567570351703739e-05, "loss": 0.0653, "step": 17670 }, { "epoch": 0.5977213563676933, "grad_norm": 0.3220174014568329, "learning_rate": 9.566897458649001e-05, "loss": 0.0573, "step": 17680 }, { "epoch": 0.5980594340579465, "grad_norm": 0.3698379695415497, "learning_rate": 9.566224066165221e-05, "loss": 0.0679, "step": 17690 }, { "epoch": 0.5983975117481998, "grad_norm": 0.2416396290063858, "learning_rate": 9.565550174326043e-05, "loss": 0.0584, "step": 17700 }, { "epoch": 0.598735589438453, "grad_norm": 0.2936091125011444, "learning_rate": 9.564875783205162e-05, "loss": 0.0557, "step": 17710 }, { "epoch": 0.5990736671287061, "grad_norm": 0.3583836853504181, "learning_rate": 9.564200892876328e-05, "loss": 0.0625, "step": 17720 }, { "epoch": 0.5994117448189594, "grad_norm": 0.2509722411632538, "learning_rate": 9.563525503413348e-05, "loss": 0.0578, "step": 17730 }, { "epoch": 0.5997498225092126, "grad_norm": 0.34192466735839844, "learning_rate": 9.562849614890079e-05, "loss": 0.0588, "step": 17740 }, { "epoch": 0.6000879001994658, "grad_norm": 0.3604625463485718, "learning_rate": 9.562173227380436e-05, "loss": 0.0563, "step": 17750 }, { "epoch": 0.6004259778897191, "grad_norm": 0.3875628113746643, "learning_rate": 9.561496340958389e-05, "loss": 0.0668, "step": 17760 }, { "epoch": 0.6007640555799723, "grad_norm": 0.36825549602508545, "learning_rate": 9.560818955697959e-05, "loss": 0.0596, "step": 17770 }, { "epoch": 0.6011021332702255, "grad_norm": 0.32232820987701416, "learning_rate": 9.560141071673228e-05, "loss": 0.0613, "step": 17780 }, { "epoch": 0.6014402109604787, "grad_norm": 0.2510055899620056, "learning_rate": 9.559462688958323e-05, "loss": 0.056, "step": 17790 }, { "epoch": 0.6017782886507319, "grad_norm": 0.2269163876771927, "learning_rate": 9.558783807627434e-05, "loss": 0.0628, "step": 17800 }, { "epoch": 0.6021163663409852, "grad_norm": 0.30316075682640076, "learning_rate": 9.558104427754801e-05, "loss": 0.0586, "step": 17810 }, { "epoch": 0.6024544440312384, "grad_norm": 0.2537743151187897, "learning_rate": 9.557424549414722e-05, "loss": 0.0602, "step": 17820 }, { "epoch": 0.6027925217214916, "grad_norm": 0.2539571225643158, "learning_rate": 9.556744172681546e-05, "loss": 0.0589, "step": 17830 }, { "epoch": 0.6031305994117448, "grad_norm": 0.33342596888542175, "learning_rate": 9.556063297629677e-05, "loss": 0.063, "step": 17840 }, { "epoch": 0.603468677101998, "grad_norm": 0.23931409418582916, "learning_rate": 9.555381924333578e-05, "loss": 0.056, "step": 17850 }, { "epoch": 0.6038067547922512, "grad_norm": 0.20951873064041138, "learning_rate": 9.554700052867758e-05, "loss": 0.0636, "step": 17860 }, { "epoch": 0.6041448324825045, "grad_norm": 0.30328986048698425, "learning_rate": 9.554017683306789e-05, "loss": 0.0608, "step": 17870 }, { "epoch": 0.6044829101727577, "grad_norm": 0.38558197021484375, "learning_rate": 9.553334815725294e-05, "loss": 0.0618, "step": 17880 }, { "epoch": 0.604820987863011, "grad_norm": 0.3902948200702667, "learning_rate": 9.552651450197949e-05, "loss": 0.0564, "step": 17890 }, { "epoch": 0.6051590655532642, "grad_norm": 0.3414498567581177, "learning_rate": 9.551967586799486e-05, "loss": 0.0592, "step": 17900 }, { "epoch": 0.6054971432435173, "grad_norm": 0.23548875749111176, "learning_rate": 9.551283225604692e-05, "loss": 0.0613, "step": 17910 }, { "epoch": 0.6058352209337706, "grad_norm": 0.2979731261730194, "learning_rate": 9.550598366688406e-05, "loss": 0.0643, "step": 17920 }, { "epoch": 0.6061732986240238, "grad_norm": 0.22313664853572845, "learning_rate": 9.549913010125526e-05, "loss": 0.0523, "step": 17930 }, { "epoch": 0.606511376314277, "grad_norm": 0.24644701182842255, "learning_rate": 9.549227155990999e-05, "loss": 0.0615, "step": 17940 }, { "epoch": 0.6068494540045303, "grad_norm": 0.3171100914478302, "learning_rate": 9.548540804359828e-05, "loss": 0.06, "step": 17950 }, { "epoch": 0.6071875316947835, "grad_norm": 0.23095738887786865, "learning_rate": 9.547853955307077e-05, "loss": 0.059, "step": 17960 }, { "epoch": 0.6075256093850366, "grad_norm": 0.18253444135189056, "learning_rate": 9.547166608907853e-05, "loss": 0.0577, "step": 17970 }, { "epoch": 0.6078636870752899, "grad_norm": 0.40901947021484375, "learning_rate": 9.546478765237326e-05, "loss": 0.0639, "step": 17980 }, { "epoch": 0.6082017647655431, "grad_norm": 0.22906911373138428, "learning_rate": 9.545790424370715e-05, "loss": 0.0561, "step": 17990 }, { "epoch": 0.6085398424557963, "grad_norm": 0.21298158168792725, "learning_rate": 9.5451015863833e-05, "loss": 0.0623, "step": 18000 }, { "epoch": 0.6088779201460496, "grad_norm": 0.22497732937335968, "learning_rate": 9.544412251350408e-05, "loss": 0.0615, "step": 18010 }, { "epoch": 0.6092159978363028, "grad_norm": 0.2503424286842346, "learning_rate": 9.543722419347422e-05, "loss": 0.0538, "step": 18020 }, { "epoch": 0.609554075526556, "grad_norm": 0.3468102216720581, "learning_rate": 9.543032090449788e-05, "loss": 0.062, "step": 18030 }, { "epoch": 0.6098921532168092, "grad_norm": 0.24065682291984558, "learning_rate": 9.542341264732992e-05, "loss": 0.0603, "step": 18040 }, { "epoch": 0.6102302309070624, "grad_norm": 0.14289416372776031, "learning_rate": 9.541649942272585e-05, "loss": 0.0635, "step": 18050 }, { "epoch": 0.6105683085973157, "grad_norm": 0.17401234805583954, "learning_rate": 9.54095812314417e-05, "loss": 0.058, "step": 18060 }, { "epoch": 0.6109063862875689, "grad_norm": 0.23895980417728424, "learning_rate": 9.540265807423401e-05, "loss": 0.0566, "step": 18070 }, { "epoch": 0.6112444639778221, "grad_norm": 0.44208675622940063, "learning_rate": 9.53957299518599e-05, "loss": 0.0572, "step": 18080 }, { "epoch": 0.6115825416680754, "grad_norm": 0.2666286528110504, "learning_rate": 9.5388796865077e-05, "loss": 0.0646, "step": 18090 }, { "epoch": 0.6119206193583285, "grad_norm": 0.23729707300662994, "learning_rate": 9.538185881464353e-05, "loss": 0.0587, "step": 18100 }, { "epoch": 0.6122586970485817, "grad_norm": 0.3094271719455719, "learning_rate": 9.537491580131821e-05, "loss": 0.0636, "step": 18110 }, { "epoch": 0.612596774738835, "grad_norm": 0.3255549669265747, "learning_rate": 9.53679678258603e-05, "loss": 0.0599, "step": 18120 }, { "epoch": 0.6129348524290882, "grad_norm": 0.1986939162015915, "learning_rate": 9.536101488902966e-05, "loss": 0.0657, "step": 18130 }, { "epoch": 0.6132729301193415, "grad_norm": 0.2356698215007782, "learning_rate": 9.535405699158663e-05, "loss": 0.0628, "step": 18140 }, { "epoch": 0.6136110078095947, "grad_norm": 0.24066519737243652, "learning_rate": 9.53470941342921e-05, "loss": 0.0604, "step": 18150 }, { "epoch": 0.6139490854998478, "grad_norm": 0.19546493887901306, "learning_rate": 9.534012631790756e-05, "loss": 0.0541, "step": 18160 }, { "epoch": 0.6142871631901011, "grad_norm": 0.18053382635116577, "learning_rate": 9.533315354319494e-05, "loss": 0.0632, "step": 18170 }, { "epoch": 0.6146252408803543, "grad_norm": 0.21892161667346954, "learning_rate": 9.532617581091682e-05, "loss": 0.0647, "step": 18180 }, { "epoch": 0.6149633185706075, "grad_norm": 0.210064098238945, "learning_rate": 9.531919312183629e-05, "loss": 0.0592, "step": 18190 }, { "epoch": 0.6153013962608608, "grad_norm": 0.19733919203281403, "learning_rate": 9.531220547671688e-05, "loss": 0.0655, "step": 18200 }, { "epoch": 0.615639473951114, "grad_norm": 0.27758657932281494, "learning_rate": 9.530521287632285e-05, "loss": 0.0624, "step": 18210 }, { "epoch": 0.6159775516413671, "grad_norm": 0.2496555894613266, "learning_rate": 9.529821532141884e-05, "loss": 0.0593, "step": 18220 }, { "epoch": 0.6163156293316204, "grad_norm": 0.2713402211666107, "learning_rate": 9.52912128127701e-05, "loss": 0.0646, "step": 18230 }, { "epoch": 0.6166537070218736, "grad_norm": 0.31096526980400085, "learning_rate": 9.528420535114244e-05, "loss": 0.0636, "step": 18240 }, { "epoch": 0.6169917847121269, "grad_norm": 0.23423077166080475, "learning_rate": 9.527719293730215e-05, "loss": 0.0547, "step": 18250 }, { "epoch": 0.6173298624023801, "grad_norm": 0.18392357230186462, "learning_rate": 9.527017557201611e-05, "loss": 0.0592, "step": 18260 }, { "epoch": 0.6176679400926333, "grad_norm": 0.23166212439537048, "learning_rate": 9.526315325605176e-05, "loss": 0.06, "step": 18270 }, { "epoch": 0.6180060177828866, "grad_norm": 0.27771303057670593, "learning_rate": 9.525612599017699e-05, "loss": 0.0616, "step": 18280 }, { "epoch": 0.6183440954731397, "grad_norm": 0.2060299962759018, "learning_rate": 9.524909377516033e-05, "loss": 0.0608, "step": 18290 }, { "epoch": 0.6186821731633929, "grad_norm": 0.30983445048332214, "learning_rate": 9.524205661177081e-05, "loss": 0.0674, "step": 18300 }, { "epoch": 0.6190202508536462, "grad_norm": 0.18407684564590454, "learning_rate": 9.523501450077801e-05, "loss": 0.0602, "step": 18310 }, { "epoch": 0.6193583285438994, "grad_norm": 0.1882108747959137, "learning_rate": 9.522796744295202e-05, "loss": 0.0584, "step": 18320 }, { "epoch": 0.6196964062341526, "grad_norm": 0.2501167058944702, "learning_rate": 9.522091543906352e-05, "loss": 0.064, "step": 18330 }, { "epoch": 0.6200344839244059, "grad_norm": 0.26179420948028564, "learning_rate": 9.521385848988369e-05, "loss": 0.0615, "step": 18340 }, { "epoch": 0.620372561614659, "grad_norm": 0.17486649751663208, "learning_rate": 9.520679659618428e-05, "loss": 0.0581, "step": 18350 }, { "epoch": 0.6207106393049122, "grad_norm": 0.41943827271461487, "learning_rate": 9.519972975873754e-05, "loss": 0.0569, "step": 18360 }, { "epoch": 0.6210487169951655, "grad_norm": 0.21993888914585114, "learning_rate": 9.519265797831633e-05, "loss": 0.0605, "step": 18370 }, { "epoch": 0.6213867946854187, "grad_norm": 0.2505302131175995, "learning_rate": 9.518558125569399e-05, "loss": 0.0633, "step": 18380 }, { "epoch": 0.621724872375672, "grad_norm": 0.3030998706817627, "learning_rate": 9.517849959164442e-05, "loss": 0.0599, "step": 18390 }, { "epoch": 0.6220629500659252, "grad_norm": 0.41188743710517883, "learning_rate": 9.517141298694205e-05, "loss": 0.0543, "step": 18400 }, { "epoch": 0.6224010277561783, "grad_norm": 0.3145572543144226, "learning_rate": 9.516432144236188e-05, "loss": 0.0654, "step": 18410 }, { "epoch": 0.6227391054464316, "grad_norm": 0.4287378787994385, "learning_rate": 9.515722495867941e-05, "loss": 0.0578, "step": 18420 }, { "epoch": 0.6230771831366848, "grad_norm": 0.2512681186199188, "learning_rate": 9.515012353667072e-05, "loss": 0.0628, "step": 18430 }, { "epoch": 0.623415260826938, "grad_norm": 0.29811370372772217, "learning_rate": 9.51430171771124e-05, "loss": 0.06, "step": 18440 }, { "epoch": 0.6237533385171913, "grad_norm": 0.2537689805030823, "learning_rate": 9.513590588078159e-05, "loss": 0.0637, "step": 18450 }, { "epoch": 0.6240914162074445, "grad_norm": 0.23951967060565948, "learning_rate": 9.512878964845597e-05, "loss": 0.0605, "step": 18460 }, { "epoch": 0.6244294938976976, "grad_norm": 0.22071562707424164, "learning_rate": 9.512166848091377e-05, "loss": 0.0652, "step": 18470 }, { "epoch": 0.6247675715879509, "grad_norm": 0.22763046622276306, "learning_rate": 9.511454237893376e-05, "loss": 0.0588, "step": 18480 }, { "epoch": 0.6251056492782041, "grad_norm": 0.21130509674549103, "learning_rate": 9.51074113432952e-05, "loss": 0.0562, "step": 18490 }, { "epoch": 0.6254437269684574, "grad_norm": 0.2055767923593521, "learning_rate": 9.510027537477797e-05, "loss": 0.0539, "step": 18500 }, { "epoch": 0.6257818046587106, "grad_norm": 0.4361979365348816, "learning_rate": 9.509313447416242e-05, "loss": 0.0599, "step": 18510 }, { "epoch": 0.6261198823489638, "grad_norm": 0.21173948049545288, "learning_rate": 9.508598864222949e-05, "loss": 0.0658, "step": 18520 }, { "epoch": 0.626457960039217, "grad_norm": 0.21574601531028748, "learning_rate": 9.507883787976062e-05, "loss": 0.0631, "step": 18530 }, { "epoch": 0.6267960377294702, "grad_norm": 0.21891994774341583, "learning_rate": 9.507168218753781e-05, "loss": 0.0644, "step": 18540 }, { "epoch": 0.6271341154197234, "grad_norm": 0.23046749830245972, "learning_rate": 9.506452156634362e-05, "loss": 0.0627, "step": 18550 }, { "epoch": 0.6274721931099767, "grad_norm": 0.2833339273929596, "learning_rate": 9.505735601696109e-05, "loss": 0.0608, "step": 18560 }, { "epoch": 0.6278102708002299, "grad_norm": 0.25121986865997314, "learning_rate": 9.505018554017385e-05, "loss": 0.0697, "step": 18570 }, { "epoch": 0.6281483484904831, "grad_norm": 0.20258305966854095, "learning_rate": 9.504301013676604e-05, "loss": 0.0567, "step": 18580 }, { "epoch": 0.6284864261807364, "grad_norm": 0.34840628504753113, "learning_rate": 9.503582980752238e-05, "loss": 0.0604, "step": 18590 }, { "epoch": 0.6288245038709895, "grad_norm": 0.1962379813194275, "learning_rate": 9.502864455322809e-05, "loss": 0.0601, "step": 18600 }, { "epoch": 0.6291625815612428, "grad_norm": 0.21465884149074554, "learning_rate": 9.502145437466891e-05, "loss": 0.0618, "step": 18610 }, { "epoch": 0.629500659251496, "grad_norm": 0.294341504573822, "learning_rate": 9.501425927263116e-05, "loss": 0.0681, "step": 18620 }, { "epoch": 0.6298387369417492, "grad_norm": 0.3889791667461395, "learning_rate": 9.500705924790172e-05, "loss": 0.0658, "step": 18630 }, { "epoch": 0.6301768146320025, "grad_norm": 0.29794418811798096, "learning_rate": 9.499985430126794e-05, "loss": 0.0595, "step": 18640 }, { "epoch": 0.6305148923222557, "grad_norm": 0.2346891462802887, "learning_rate": 9.499264443351775e-05, "loss": 0.0632, "step": 18650 }, { "epoch": 0.6308529700125088, "grad_norm": 0.39675450325012207, "learning_rate": 9.498542964543961e-05, "loss": 0.0589, "step": 18660 }, { "epoch": 0.6311910477027621, "grad_norm": 0.3217610716819763, "learning_rate": 9.497820993782252e-05, "loss": 0.0589, "step": 18670 }, { "epoch": 0.6315291253930153, "grad_norm": 0.23996014893054962, "learning_rate": 9.497098531145601e-05, "loss": 0.0618, "step": 18680 }, { "epoch": 0.6318672030832685, "grad_norm": 0.3031204044818878, "learning_rate": 9.496375576713017e-05, "loss": 0.061, "step": 18690 }, { "epoch": 0.6322052807735218, "grad_norm": 0.22553019225597382, "learning_rate": 9.49565213056356e-05, "loss": 0.0595, "step": 18700 }, { "epoch": 0.632543358463775, "grad_norm": 0.30266493558883667, "learning_rate": 9.494928192776342e-05, "loss": 0.0575, "step": 18710 }, { "epoch": 0.6328814361540281, "grad_norm": 0.21118928492069244, "learning_rate": 9.494203763430538e-05, "loss": 0.0608, "step": 18720 }, { "epoch": 0.6332195138442814, "grad_norm": 0.37275227904319763, "learning_rate": 9.493478842605366e-05, "loss": 0.0578, "step": 18730 }, { "epoch": 0.6335575915345346, "grad_norm": 0.24279505014419556, "learning_rate": 9.492753430380105e-05, "loss": 0.0624, "step": 18740 }, { "epoch": 0.6338956692247879, "grad_norm": 0.212949737906456, "learning_rate": 9.492027526834083e-05, "loss": 0.0559, "step": 18750 }, { "epoch": 0.6342337469150411, "grad_norm": 0.2268417477607727, "learning_rate": 9.491301132046684e-05, "loss": 0.0601, "step": 18760 }, { "epoch": 0.6345718246052943, "grad_norm": 0.19118645787239075, "learning_rate": 9.490574246097345e-05, "loss": 0.0611, "step": 18770 }, { "epoch": 0.6349099022955476, "grad_norm": 0.40871453285217285, "learning_rate": 9.48984686906556e-05, "loss": 0.0608, "step": 18780 }, { "epoch": 0.6352479799858007, "grad_norm": 0.29732194542884827, "learning_rate": 9.489119001030871e-05, "loss": 0.0584, "step": 18790 }, { "epoch": 0.6355860576760539, "grad_norm": 0.26100143790245056, "learning_rate": 9.488390642072878e-05, "loss": 0.0541, "step": 18800 }, { "epoch": 0.6359241353663072, "grad_norm": 0.29548072814941406, "learning_rate": 9.48766179227123e-05, "loss": 0.0622, "step": 18810 }, { "epoch": 0.6362622130565604, "grad_norm": 0.3063375949859619, "learning_rate": 9.486932451705636e-05, "loss": 0.0586, "step": 18820 }, { "epoch": 0.6366002907468136, "grad_norm": 0.2529986798763275, "learning_rate": 9.486202620455857e-05, "loss": 0.0661, "step": 18830 }, { "epoch": 0.6369383684370669, "grad_norm": 0.287996768951416, "learning_rate": 9.485472298601704e-05, "loss": 0.0641, "step": 18840 }, { "epoch": 0.63727644612732, "grad_norm": 0.465963214635849, "learning_rate": 9.484741486223043e-05, "loss": 0.067, "step": 18850 }, { "epoch": 0.6376145238175733, "grad_norm": 0.2767977714538574, "learning_rate": 9.484010183399797e-05, "loss": 0.0604, "step": 18860 }, { "epoch": 0.6379526015078265, "grad_norm": 0.19334231317043304, "learning_rate": 9.483278390211938e-05, "loss": 0.0551, "step": 18870 }, { "epoch": 0.6382906791980797, "grad_norm": 0.1962452381849289, "learning_rate": 9.482546106739496e-05, "loss": 0.0606, "step": 18880 }, { "epoch": 0.638628756888333, "grad_norm": 0.2070644348859787, "learning_rate": 9.48181333306255e-05, "loss": 0.0616, "step": 18890 }, { "epoch": 0.6389668345785862, "grad_norm": 0.19812262058258057, "learning_rate": 9.481080069261237e-05, "loss": 0.0578, "step": 18900 }, { "epoch": 0.6393049122688393, "grad_norm": 0.23152050375938416, "learning_rate": 9.480346315415745e-05, "loss": 0.0573, "step": 18910 }, { "epoch": 0.6396429899590926, "grad_norm": 0.25973382592201233, "learning_rate": 9.479612071606314e-05, "loss": 0.0681, "step": 18920 }, { "epoch": 0.6399810676493458, "grad_norm": 0.31159526109695435, "learning_rate": 9.478877337913244e-05, "loss": 0.0582, "step": 18930 }, { "epoch": 0.640319145339599, "grad_norm": 0.1856435090303421, "learning_rate": 9.478142114416881e-05, "loss": 0.0647, "step": 18940 }, { "epoch": 0.6406572230298523, "grad_norm": 0.2701447904109955, "learning_rate": 9.47740640119763e-05, "loss": 0.0638, "step": 18950 }, { "epoch": 0.6409953007201055, "grad_norm": 0.20411120355129242, "learning_rate": 9.476670198335947e-05, "loss": 0.0635, "step": 18960 }, { "epoch": 0.6413333784103586, "grad_norm": 0.2079637199640274, "learning_rate": 9.47593350591234e-05, "loss": 0.0633, "step": 18970 }, { "epoch": 0.6416714561006119, "grad_norm": 0.16938017308712006, "learning_rate": 9.475196324007376e-05, "loss": 0.0561, "step": 18980 }, { "epoch": 0.6420095337908651, "grad_norm": 0.24475672841072083, "learning_rate": 9.474458652701669e-05, "loss": 0.0599, "step": 18990 }, { "epoch": 0.6423476114811184, "grad_norm": 0.2551565170288086, "learning_rate": 9.473720492075892e-05, "loss": 0.0531, "step": 19000 }, { "epoch": 0.6426856891713716, "grad_norm": 0.4028608202934265, "learning_rate": 9.472981842210768e-05, "loss": 0.0634, "step": 19010 }, { "epoch": 0.6430237668616248, "grad_norm": 0.17839035391807556, "learning_rate": 9.472242703187074e-05, "loss": 0.0526, "step": 19020 }, { "epoch": 0.6433618445518781, "grad_norm": 0.19511418044567108, "learning_rate": 9.471503075085643e-05, "loss": 0.0661, "step": 19030 }, { "epoch": 0.6436999222421312, "grad_norm": 0.28697752952575684, "learning_rate": 9.470762957987359e-05, "loss": 0.0578, "step": 19040 }, { "epoch": 0.6440379999323844, "grad_norm": 0.3410000503063202, "learning_rate": 9.470022351973158e-05, "loss": 0.0589, "step": 19050 }, { "epoch": 0.6443760776226377, "grad_norm": 0.3032234311103821, "learning_rate": 9.469281257124034e-05, "loss": 0.0657, "step": 19060 }, { "epoch": 0.6447141553128909, "grad_norm": 0.23957785964012146, "learning_rate": 9.46853967352103e-05, "loss": 0.0646, "step": 19070 }, { "epoch": 0.6450522330031441, "grad_norm": 0.2432117909193039, "learning_rate": 9.467797601245246e-05, "loss": 0.0644, "step": 19080 }, { "epoch": 0.6453903106933974, "grad_norm": 0.20474131405353546, "learning_rate": 9.467055040377834e-05, "loss": 0.0618, "step": 19090 }, { "epoch": 0.6457283883836505, "grad_norm": 0.3047933876514435, "learning_rate": 9.466311990999999e-05, "loss": 0.0669, "step": 19100 }, { "epoch": 0.6460664660739038, "grad_norm": 0.24474628269672394, "learning_rate": 9.465568453193e-05, "loss": 0.0618, "step": 19110 }, { "epoch": 0.646404543764157, "grad_norm": 0.2846909761428833, "learning_rate": 9.464824427038148e-05, "loss": 0.0655, "step": 19120 }, { "epoch": 0.6467426214544102, "grad_norm": 0.3530585467815399, "learning_rate": 9.46407991261681e-05, "loss": 0.0614, "step": 19130 }, { "epoch": 0.6470806991446635, "grad_norm": 0.26658427715301514, "learning_rate": 9.463334910010404e-05, "loss": 0.063, "step": 19140 }, { "epoch": 0.6474187768349167, "grad_norm": 0.18876303732395172, "learning_rate": 9.462589419300403e-05, "loss": 0.0601, "step": 19150 }, { "epoch": 0.6477568545251698, "grad_norm": 0.2209435999393463, "learning_rate": 9.461843440568333e-05, "loss": 0.0573, "step": 19160 }, { "epoch": 0.6480949322154231, "grad_norm": 0.2611802816390991, "learning_rate": 9.461096973895773e-05, "loss": 0.0612, "step": 19170 }, { "epoch": 0.6484330099056763, "grad_norm": 0.23731815814971924, "learning_rate": 9.460350019364355e-05, "loss": 0.0614, "step": 19180 }, { "epoch": 0.6487710875959295, "grad_norm": 0.236213818192482, "learning_rate": 9.459602577055764e-05, "loss": 0.0616, "step": 19190 }, { "epoch": 0.6491091652861828, "grad_norm": 0.20963265001773834, "learning_rate": 9.45885464705174e-05, "loss": 0.057, "step": 19200 }, { "epoch": 0.649447242976436, "grad_norm": 0.22823476791381836, "learning_rate": 9.458106229434076e-05, "loss": 0.0612, "step": 19210 }, { "epoch": 0.6497853206666893, "grad_norm": 0.19896966218948364, "learning_rate": 9.457357324284617e-05, "loss": 0.0624, "step": 19220 }, { "epoch": 0.6501233983569424, "grad_norm": 0.309696227312088, "learning_rate": 9.456607931685262e-05, "loss": 0.0629, "step": 19230 }, { "epoch": 0.6504614760471956, "grad_norm": 0.2586222290992737, "learning_rate": 9.455858051717965e-05, "loss": 0.0571, "step": 19240 }, { "epoch": 0.6507995537374489, "grad_norm": 0.42397457361221313, "learning_rate": 9.45510768446473e-05, "loss": 0.0603, "step": 19250 }, { "epoch": 0.6511376314277021, "grad_norm": 0.29106608033180237, "learning_rate": 9.454356830007618e-05, "loss": 0.065, "step": 19260 }, { "epoch": 0.6514757091179553, "grad_norm": 0.32703614234924316, "learning_rate": 9.45360548842874e-05, "loss": 0.0582, "step": 19270 }, { "epoch": 0.6518137868082086, "grad_norm": 0.5757505893707275, "learning_rate": 9.452853659810261e-05, "loss": 0.0678, "step": 19280 }, { "epoch": 0.6521518644984617, "grad_norm": 0.3802233040332794, "learning_rate": 9.452101344234401e-05, "loss": 0.0651, "step": 19290 }, { "epoch": 0.6524899421887149, "grad_norm": 0.18061433732509613, "learning_rate": 9.451348541783431e-05, "loss": 0.0627, "step": 19300 }, { "epoch": 0.6528280198789682, "grad_norm": 0.2671999931335449, "learning_rate": 9.450595252539678e-05, "loss": 0.0602, "step": 19310 }, { "epoch": 0.6531660975692214, "grad_norm": 0.2065574675798416, "learning_rate": 9.449841476585518e-05, "loss": 0.0609, "step": 19320 }, { "epoch": 0.6535041752594746, "grad_norm": 0.19015179574489594, "learning_rate": 9.449087214003384e-05, "loss": 0.0525, "step": 19330 }, { "epoch": 0.6538422529497279, "grad_norm": 0.2115701586008072, "learning_rate": 9.448332464875765e-05, "loss": 0.0617, "step": 19340 }, { "epoch": 0.654180330639981, "grad_norm": 0.18239884078502655, "learning_rate": 9.447577229285192e-05, "loss": 0.0592, "step": 19350 }, { "epoch": 0.6545184083302343, "grad_norm": 0.2854866683483124, "learning_rate": 9.446821507314261e-05, "loss": 0.0658, "step": 19360 }, { "epoch": 0.6548564860204875, "grad_norm": 0.23061524331569672, "learning_rate": 9.446065299045617e-05, "loss": 0.0718, "step": 19370 }, { "epoch": 0.6551945637107407, "grad_norm": 0.25603586435317993, "learning_rate": 9.445308604561955e-05, "loss": 0.0614, "step": 19380 }, { "epoch": 0.655532641400994, "grad_norm": 0.2593643367290497, "learning_rate": 9.444551423946028e-05, "loss": 0.0589, "step": 19390 }, { "epoch": 0.6558707190912472, "grad_norm": 0.30877718329429626, "learning_rate": 9.443793757280638e-05, "loss": 0.0645, "step": 19400 }, { "epoch": 0.6562087967815003, "grad_norm": 0.2752875089645386, "learning_rate": 9.443035604648646e-05, "loss": 0.0637, "step": 19410 }, { "epoch": 0.6565468744717536, "grad_norm": 0.25952696800231934, "learning_rate": 9.44227696613296e-05, "loss": 0.057, "step": 19420 }, { "epoch": 0.6568849521620068, "grad_norm": 0.2630195617675781, "learning_rate": 9.441517841816542e-05, "loss": 0.0618, "step": 19430 }, { "epoch": 0.65722302985226, "grad_norm": 0.24196545779705048, "learning_rate": 9.440758231782413e-05, "loss": 0.0639, "step": 19440 }, { "epoch": 0.6575611075425133, "grad_norm": 0.30449339747428894, "learning_rate": 9.439998136113639e-05, "loss": 0.0573, "step": 19450 }, { "epoch": 0.6578991852327665, "grad_norm": 0.23092642426490784, "learning_rate": 9.439237554893344e-05, "loss": 0.0527, "step": 19460 }, { "epoch": 0.6582372629230198, "grad_norm": 0.16432934999465942, "learning_rate": 9.438476488204705e-05, "loss": 0.0548, "step": 19470 }, { "epoch": 0.6585753406132729, "grad_norm": 0.29576778411865234, "learning_rate": 9.43771493613095e-05, "loss": 0.0619, "step": 19480 }, { "epoch": 0.6589134183035261, "grad_norm": 0.39652031660079956, "learning_rate": 9.436952898755362e-05, "loss": 0.0587, "step": 19490 }, { "epoch": 0.6592514959937794, "grad_norm": 0.25875669717788696, "learning_rate": 9.436190376161276e-05, "loss": 0.0564, "step": 19500 }, { "epoch": 0.6595895736840326, "grad_norm": 0.31773144006729126, "learning_rate": 9.43542736843208e-05, "loss": 0.0654, "step": 19510 }, { "epoch": 0.6599276513742858, "grad_norm": 0.22054935991764069, "learning_rate": 9.434663875651216e-05, "loss": 0.0543, "step": 19520 }, { "epoch": 0.6602657290645391, "grad_norm": 0.2816828787326813, "learning_rate": 9.433899897902177e-05, "loss": 0.0609, "step": 19530 }, { "epoch": 0.6606038067547922, "grad_norm": 0.28581154346466064, "learning_rate": 9.433135435268511e-05, "loss": 0.0538, "step": 19540 }, { "epoch": 0.6609418844450454, "grad_norm": 0.2902570068836212, "learning_rate": 9.432370487833819e-05, "loss": 0.0565, "step": 19550 }, { "epoch": 0.6612799621352987, "grad_norm": 0.21723830699920654, "learning_rate": 9.431605055681756e-05, "loss": 0.0605, "step": 19560 }, { "epoch": 0.6616180398255519, "grad_norm": 0.25123468041419983, "learning_rate": 9.430839138896026e-05, "loss": 0.0604, "step": 19570 }, { "epoch": 0.6619561175158052, "grad_norm": 0.16056126356124878, "learning_rate": 9.43007273756039e-05, "loss": 0.0546, "step": 19580 }, { "epoch": 0.6622941952060584, "grad_norm": 0.17940278351306915, "learning_rate": 9.429305851758658e-05, "loss": 0.0609, "step": 19590 }, { "epoch": 0.6626322728963115, "grad_norm": 0.24279317259788513, "learning_rate": 9.428538481574699e-05, "loss": 0.0627, "step": 19600 }, { "epoch": 0.6629703505865648, "grad_norm": 0.3137940764427185, "learning_rate": 9.42777062709243e-05, "loss": 0.0637, "step": 19610 }, { "epoch": 0.663308428276818, "grad_norm": 0.31613579392433167, "learning_rate": 9.427002288395821e-05, "loss": 0.0552, "step": 19620 }, { "epoch": 0.6636465059670712, "grad_norm": 0.29353538155555725, "learning_rate": 9.426233465568898e-05, "loss": 0.0591, "step": 19630 }, { "epoch": 0.6639845836573245, "grad_norm": 0.2080269604921341, "learning_rate": 9.42546415869574e-05, "loss": 0.0547, "step": 19640 }, { "epoch": 0.6643226613475777, "grad_norm": 0.216646209359169, "learning_rate": 9.424694367860473e-05, "loss": 0.0596, "step": 19650 }, { "epoch": 0.6646607390378308, "grad_norm": 0.377424955368042, "learning_rate": 9.423924093147284e-05, "loss": 0.0663, "step": 19660 }, { "epoch": 0.6649988167280841, "grad_norm": 0.2409677654504776, "learning_rate": 9.423153334640407e-05, "loss": 0.0625, "step": 19670 }, { "epoch": 0.6653368944183373, "grad_norm": 0.2680356502532959, "learning_rate": 9.42238209242413e-05, "loss": 0.0601, "step": 19680 }, { "epoch": 0.6656749721085905, "grad_norm": 0.22119778394699097, "learning_rate": 9.421610366582798e-05, "loss": 0.0582, "step": 19690 }, { "epoch": 0.6660130497988438, "grad_norm": 0.37965139746665955, "learning_rate": 9.420838157200803e-05, "loss": 0.0611, "step": 19700 }, { "epoch": 0.666351127489097, "grad_norm": 0.3260696530342102, "learning_rate": 9.420065464362594e-05, "loss": 0.0586, "step": 19710 }, { "epoch": 0.6666892051793503, "grad_norm": 0.18967591226100922, "learning_rate": 9.419292288152673e-05, "loss": 0.0585, "step": 19720 }, { "epoch": 0.6670272828696034, "grad_norm": 0.31210264563560486, "learning_rate": 9.418518628655588e-05, "loss": 0.0594, "step": 19730 }, { "epoch": 0.6673653605598566, "grad_norm": 0.30114036798477173, "learning_rate": 9.417744485955951e-05, "loss": 0.0627, "step": 19740 }, { "epoch": 0.6677034382501099, "grad_norm": 0.21884390711784363, "learning_rate": 9.41696986013842e-05, "loss": 0.0665, "step": 19750 }, { "epoch": 0.6680415159403631, "grad_norm": 0.2956778407096863, "learning_rate": 9.416194751287705e-05, "loss": 0.0636, "step": 19760 }, { "epoch": 0.6683795936306163, "grad_norm": 0.23483753204345703, "learning_rate": 9.415419159488572e-05, "loss": 0.0629, "step": 19770 }, { "epoch": 0.6687176713208696, "grad_norm": 0.3231548070907593, "learning_rate": 9.414643084825837e-05, "loss": 0.0552, "step": 19780 }, { "epoch": 0.6690557490111227, "grad_norm": 0.2583918273448944, "learning_rate": 9.413866527384372e-05, "loss": 0.066, "step": 19790 }, { "epoch": 0.6693938267013759, "grad_norm": 0.3130502998828888, "learning_rate": 9.4130894872491e-05, "loss": 0.0557, "step": 19800 }, { "epoch": 0.6697319043916292, "grad_norm": 0.2396242618560791, "learning_rate": 9.412311964504998e-05, "loss": 0.0661, "step": 19810 }, { "epoch": 0.6700699820818824, "grad_norm": 0.239772766828537, "learning_rate": 9.411533959237091e-05, "loss": 0.0576, "step": 19820 }, { "epoch": 0.6704080597721357, "grad_norm": 0.30743035674095154, "learning_rate": 9.410755471530464e-05, "loss": 0.0669, "step": 19830 }, { "epoch": 0.6707461374623889, "grad_norm": 0.17426976561546326, "learning_rate": 9.40997650147025e-05, "loss": 0.0603, "step": 19840 }, { "epoch": 0.671084215152642, "grad_norm": 0.27207693457603455, "learning_rate": 9.409197049141637e-05, "loss": 0.0606, "step": 19850 }, { "epoch": 0.6714222928428953, "grad_norm": 0.3273475468158722, "learning_rate": 9.408417114629863e-05, "loss": 0.0665, "step": 19860 }, { "epoch": 0.6717603705331485, "grad_norm": 0.27780139446258545, "learning_rate": 9.40763669802022e-05, "loss": 0.0652, "step": 19870 }, { "epoch": 0.6720984482234017, "grad_norm": 0.4160791039466858, "learning_rate": 9.406855799398056e-05, "loss": 0.0599, "step": 19880 }, { "epoch": 0.672436525913655, "grad_norm": 0.24062076210975647, "learning_rate": 9.406074418848767e-05, "loss": 0.062, "step": 19890 }, { "epoch": 0.6727746036039082, "grad_norm": 0.26014238595962524, "learning_rate": 9.405292556457805e-05, "loss": 0.0545, "step": 19900 }, { "epoch": 0.6731126812941614, "grad_norm": 0.2600368857383728, "learning_rate": 9.404510212310671e-05, "loss": 0.0601, "step": 19910 }, { "epoch": 0.6734507589844146, "grad_norm": 0.34329983592033386, "learning_rate": 9.403727386492924e-05, "loss": 0.0623, "step": 19920 }, { "epoch": 0.6737888366746678, "grad_norm": 0.3957080841064453, "learning_rate": 9.40294407909017e-05, "loss": 0.0645, "step": 19930 }, { "epoch": 0.674126914364921, "grad_norm": 0.1919858306646347, "learning_rate": 9.40216029018807e-05, "loss": 0.0603, "step": 19940 }, { "epoch": 0.6744649920551743, "grad_norm": 0.3809646964073181, "learning_rate": 9.401376019872338e-05, "loss": 0.0604, "step": 19950 }, { "epoch": 0.6748030697454275, "grad_norm": 0.20411884784698486, "learning_rate": 9.400591268228746e-05, "loss": 0.0588, "step": 19960 }, { "epoch": 0.6751411474356808, "grad_norm": 0.19823335111141205, "learning_rate": 9.399806035343106e-05, "loss": 0.0616, "step": 19970 }, { "epoch": 0.675479225125934, "grad_norm": 0.22040748596191406, "learning_rate": 9.399020321301294e-05, "loss": 0.0618, "step": 19980 }, { "epoch": 0.6758173028161871, "grad_norm": 0.18124938011169434, "learning_rate": 9.398234126189234e-05, "loss": 0.0604, "step": 19990 }, { "epoch": 0.6761553805064404, "grad_norm": 0.20426228642463684, "learning_rate": 9.397447450092902e-05, "loss": 0.0633, "step": 20000 }, { "epoch": 0.6764934581966936, "grad_norm": 0.46652835607528687, "learning_rate": 9.39666029309833e-05, "loss": 0.0634, "step": 20010 }, { "epoch": 0.6768315358869468, "grad_norm": 0.4960470199584961, "learning_rate": 9.395872655291596e-05, "loss": 0.0638, "step": 20020 }, { "epoch": 0.6771696135772001, "grad_norm": 0.2085620015859604, "learning_rate": 9.395084536758838e-05, "loss": 0.0572, "step": 20030 }, { "epoch": 0.6775076912674532, "grad_norm": 0.25039348006248474, "learning_rate": 9.394295937586243e-05, "loss": 0.0577, "step": 20040 }, { "epoch": 0.6778457689577064, "grad_norm": 0.18687184154987335, "learning_rate": 9.393506857860052e-05, "loss": 0.0648, "step": 20050 }, { "epoch": 0.6781838466479597, "grad_norm": 0.25663572549819946, "learning_rate": 9.392717297666555e-05, "loss": 0.0665, "step": 20060 }, { "epoch": 0.6785219243382129, "grad_norm": 0.25064659118652344, "learning_rate": 9.391927257092101e-05, "loss": 0.0626, "step": 20070 }, { "epoch": 0.6788600020284662, "grad_norm": 0.19300009310245514, "learning_rate": 9.391136736223085e-05, "loss": 0.0624, "step": 20080 }, { "epoch": 0.6791980797187194, "grad_norm": 0.14544691145420074, "learning_rate": 9.390345735145956e-05, "loss": 0.0523, "step": 20090 }, { "epoch": 0.6795361574089726, "grad_norm": 0.23133160173892975, "learning_rate": 9.389554253947219e-05, "loss": 0.0616, "step": 20100 }, { "epoch": 0.6798742350992258, "grad_norm": 0.22363373637199402, "learning_rate": 9.388762292713428e-05, "loss": 0.0551, "step": 20110 }, { "epoch": 0.680212312789479, "grad_norm": 0.2522121071815491, "learning_rate": 9.38796985153119e-05, "loss": 0.0565, "step": 20120 }, { "epoch": 0.6805503904797322, "grad_norm": 0.2722201347351074, "learning_rate": 9.387176930487169e-05, "loss": 0.0688, "step": 20130 }, { "epoch": 0.6808884681699855, "grad_norm": 0.2621610760688782, "learning_rate": 9.386383529668072e-05, "loss": 0.0554, "step": 20140 }, { "epoch": 0.6812265458602387, "grad_norm": 0.25701409578323364, "learning_rate": 9.385589649160669e-05, "loss": 0.062, "step": 20150 }, { "epoch": 0.6815646235504919, "grad_norm": 0.23322628438472748, "learning_rate": 9.384795289051775e-05, "loss": 0.0644, "step": 20160 }, { "epoch": 0.6819027012407451, "grad_norm": 0.255185604095459, "learning_rate": 9.384000449428261e-05, "loss": 0.0559, "step": 20170 }, { "epoch": 0.6822407789309983, "grad_norm": 0.23658397793769836, "learning_rate": 9.383205130377048e-05, "loss": 0.0631, "step": 20180 }, { "epoch": 0.6825788566212516, "grad_norm": 0.25497257709503174, "learning_rate": 9.382409331985114e-05, "loss": 0.0615, "step": 20190 }, { "epoch": 0.6829169343115048, "grad_norm": 0.21774712204933167, "learning_rate": 9.381613054339482e-05, "loss": 0.0611, "step": 20200 }, { "epoch": 0.683255012001758, "grad_norm": 0.35543110966682434, "learning_rate": 9.380816297527235e-05, "loss": 0.055, "step": 20210 }, { "epoch": 0.6835930896920113, "grad_norm": 0.22611431777477264, "learning_rate": 9.380019061635506e-05, "loss": 0.0563, "step": 20220 }, { "epoch": 0.6839311673822644, "grad_norm": 0.3333672881126404, "learning_rate": 9.379221346751474e-05, "loss": 0.0534, "step": 20230 }, { "epoch": 0.6842692450725176, "grad_norm": 0.21446946263313293, "learning_rate": 9.378423152962382e-05, "loss": 0.0617, "step": 20240 }, { "epoch": 0.6846073227627709, "grad_norm": 0.2727629244327545, "learning_rate": 9.377624480355517e-05, "loss": 0.0572, "step": 20250 }, { "epoch": 0.6849454004530241, "grad_norm": 0.2223530113697052, "learning_rate": 9.376825329018219e-05, "loss": 0.0599, "step": 20260 }, { "epoch": 0.6852834781432773, "grad_norm": 0.2833592891693115, "learning_rate": 9.376025699037884e-05, "loss": 0.061, "step": 20270 }, { "epoch": 0.6856215558335306, "grad_norm": 0.3716922104358673, "learning_rate": 9.37522559050196e-05, "loss": 0.0656, "step": 20280 }, { "epoch": 0.6859596335237838, "grad_norm": 0.42107611894607544, "learning_rate": 9.37442500349794e-05, "loss": 0.0628, "step": 20290 }, { "epoch": 0.6862977112140369, "grad_norm": 0.26481595635414124, "learning_rate": 9.373623938113381e-05, "loss": 0.0629, "step": 20300 }, { "epoch": 0.6866357889042902, "grad_norm": 0.2742433249950409, "learning_rate": 9.372822394435883e-05, "loss": 0.0594, "step": 20310 }, { "epoch": 0.6869738665945434, "grad_norm": 0.34463611245155334, "learning_rate": 9.372020372553102e-05, "loss": 0.06, "step": 20320 }, { "epoch": 0.6873119442847967, "grad_norm": 0.2748490571975708, "learning_rate": 9.371217872552746e-05, "loss": 0.0668, "step": 20330 }, { "epoch": 0.6876500219750499, "grad_norm": 0.2889922559261322, "learning_rate": 9.370414894522576e-05, "loss": 0.0604, "step": 20340 }, { "epoch": 0.687988099665303, "grad_norm": 0.22571241855621338, "learning_rate": 9.369611438550406e-05, "loss": 0.0572, "step": 20350 }, { "epoch": 0.6883261773555563, "grad_norm": 0.41933706402778625, "learning_rate": 9.368807504724095e-05, "loss": 0.0638, "step": 20360 }, { "epoch": 0.6886642550458095, "grad_norm": 0.3583625555038452, "learning_rate": 9.368003093131565e-05, "loss": 0.0638, "step": 20370 }, { "epoch": 0.6890023327360627, "grad_norm": 0.23267456889152527, "learning_rate": 9.367198203860785e-05, "loss": 0.0598, "step": 20380 }, { "epoch": 0.689340410426316, "grad_norm": 0.39782941341400146, "learning_rate": 9.366392836999774e-05, "loss": 0.0593, "step": 20390 }, { "epoch": 0.6896784881165692, "grad_norm": 0.4506528377532959, "learning_rate": 9.365586992636607e-05, "loss": 0.0614, "step": 20400 }, { "epoch": 0.6900165658068224, "grad_norm": 0.225088968873024, "learning_rate": 9.364780670859412e-05, "loss": 0.0603, "step": 20410 }, { "epoch": 0.6903546434970756, "grad_norm": 0.20227248966693878, "learning_rate": 9.363973871756364e-05, "loss": 0.0579, "step": 20420 }, { "epoch": 0.6906927211873288, "grad_norm": 0.19001296162605286, "learning_rate": 9.363166595415696e-05, "loss": 0.0569, "step": 20430 }, { "epoch": 0.6910307988775821, "grad_norm": 0.3070548176765442, "learning_rate": 9.362358841925686e-05, "loss": 0.0599, "step": 20440 }, { "epoch": 0.6913688765678353, "grad_norm": 0.369111031293869, "learning_rate": 9.361550611374674e-05, "loss": 0.0619, "step": 20450 }, { "epoch": 0.6917069542580885, "grad_norm": 0.2709929347038269, "learning_rate": 9.360741903851043e-05, "loss": 0.0608, "step": 20460 }, { "epoch": 0.6920450319483418, "grad_norm": 0.29187577962875366, "learning_rate": 9.359932719443236e-05, "loss": 0.0577, "step": 20470 }, { "epoch": 0.692383109638595, "grad_norm": 0.3540118634700775, "learning_rate": 9.35912305823974e-05, "loss": 0.0641, "step": 20480 }, { "epoch": 0.6927211873288481, "grad_norm": 0.28495925664901733, "learning_rate": 9.358312920329101e-05, "loss": 0.0568, "step": 20490 }, { "epoch": 0.6930592650191014, "grad_norm": 0.25084003806114197, "learning_rate": 9.357502305799914e-05, "loss": 0.0578, "step": 20500 }, { "epoch": 0.6933973427093546, "grad_norm": 0.25663629174232483, "learning_rate": 9.356691214740824e-05, "loss": 0.0595, "step": 20510 }, { "epoch": 0.6937354203996078, "grad_norm": 0.29903095960617065, "learning_rate": 9.355879647240535e-05, "loss": 0.0597, "step": 20520 }, { "epoch": 0.6940734980898611, "grad_norm": 0.20102769136428833, "learning_rate": 9.355067603387798e-05, "loss": 0.0612, "step": 20530 }, { "epoch": 0.6944115757801143, "grad_norm": 0.2894580662250519, "learning_rate": 9.354255083271412e-05, "loss": 0.0691, "step": 20540 }, { "epoch": 0.6947496534703675, "grad_norm": 0.23709508776664734, "learning_rate": 9.353442086980239e-05, "loss": 0.0633, "step": 20550 }, { "epoch": 0.6950877311606207, "grad_norm": 0.3456977903842926, "learning_rate": 9.352628614603185e-05, "loss": 0.0616, "step": 20560 }, { "epoch": 0.6954258088508739, "grad_norm": 0.2799602448940277, "learning_rate": 9.351814666229209e-05, "loss": 0.0592, "step": 20570 }, { "epoch": 0.6957638865411272, "grad_norm": 0.19956159591674805, "learning_rate": 9.351000241947324e-05, "loss": 0.0694, "step": 20580 }, { "epoch": 0.6961019642313804, "grad_norm": 0.3358197808265686, "learning_rate": 9.350185341846594e-05, "loss": 0.0737, "step": 20590 }, { "epoch": 0.6964400419216336, "grad_norm": 0.16201788187026978, "learning_rate": 9.349369966016134e-05, "loss": 0.0564, "step": 20600 }, { "epoch": 0.6967781196118868, "grad_norm": 0.289393812417984, "learning_rate": 9.348554114545117e-05, "loss": 0.0618, "step": 20610 }, { "epoch": 0.69711619730214, "grad_norm": 0.23874495923519135, "learning_rate": 9.347737787522758e-05, "loss": 0.0572, "step": 20620 }, { "epoch": 0.6974542749923932, "grad_norm": 0.2645680606365204, "learning_rate": 9.346920985038332e-05, "loss": 0.0685, "step": 20630 }, { "epoch": 0.6977923526826465, "grad_norm": 0.3430393934249878, "learning_rate": 9.346103707181162e-05, "loss": 0.0591, "step": 20640 }, { "epoch": 0.6981304303728997, "grad_norm": 0.2837093472480774, "learning_rate": 9.345285954040626e-05, "loss": 0.0561, "step": 20650 }, { "epoch": 0.6984685080631529, "grad_norm": 0.1992119997739792, "learning_rate": 9.34446772570615e-05, "loss": 0.0627, "step": 20660 }, { "epoch": 0.6988065857534062, "grad_norm": 0.2338920682668686, "learning_rate": 9.343649022267214e-05, "loss": 0.0592, "step": 20670 }, { "epoch": 0.6991446634436593, "grad_norm": 0.17939722537994385, "learning_rate": 9.342829843813353e-05, "loss": 0.0599, "step": 20680 }, { "epoch": 0.6994827411339126, "grad_norm": 0.23931008577346802, "learning_rate": 9.342010190434149e-05, "loss": 0.0627, "step": 20690 }, { "epoch": 0.6998208188241658, "grad_norm": 0.2622728645801544, "learning_rate": 9.34119006221924e-05, "loss": 0.0607, "step": 20700 }, { "epoch": 0.700158896514419, "grad_norm": 0.2758987545967102, "learning_rate": 9.340369459258313e-05, "loss": 0.0632, "step": 20710 }, { "epoch": 0.7004969742046723, "grad_norm": 0.3525405526161194, "learning_rate": 9.339548381641106e-05, "loss": 0.0571, "step": 20720 }, { "epoch": 0.7008350518949255, "grad_norm": 0.20662280917167664, "learning_rate": 9.338726829457413e-05, "loss": 0.0604, "step": 20730 }, { "epoch": 0.7011731295851786, "grad_norm": 0.3380522131919861, "learning_rate": 9.337904802797078e-05, "loss": 0.0559, "step": 20740 }, { "epoch": 0.7015112072754319, "grad_norm": 0.30200767517089844, "learning_rate": 9.337082301749993e-05, "loss": 0.0584, "step": 20750 }, { "epoch": 0.7018492849656851, "grad_norm": 0.3726314902305603, "learning_rate": 9.336259326406109e-05, "loss": 0.0614, "step": 20760 }, { "epoch": 0.7021873626559383, "grad_norm": 0.32210031151771545, "learning_rate": 9.335435876855427e-05, "loss": 0.0636, "step": 20770 }, { "epoch": 0.7025254403461916, "grad_norm": 0.3507435917854309, "learning_rate": 9.334611953187994e-05, "loss": 0.0695, "step": 20780 }, { "epoch": 0.7028635180364448, "grad_norm": 0.29990315437316895, "learning_rate": 9.333787555493914e-05, "loss": 0.0584, "step": 20790 }, { "epoch": 0.703201595726698, "grad_norm": 0.31083571910858154, "learning_rate": 9.332962683863345e-05, "loss": 0.0579, "step": 20800 }, { "epoch": 0.7035396734169512, "grad_norm": 0.2521316409111023, "learning_rate": 9.332137338386489e-05, "loss": 0.0583, "step": 20810 }, { "epoch": 0.7038777511072044, "grad_norm": 0.44622087478637695, "learning_rate": 9.33131151915361e-05, "loss": 0.061, "step": 20820 }, { "epoch": 0.7042158287974577, "grad_norm": 0.20754332840442657, "learning_rate": 9.330485226255012e-05, "loss": 0.0625, "step": 20830 }, { "epoch": 0.7045539064877109, "grad_norm": 0.25672799348831177, "learning_rate": 9.329658459781061e-05, "loss": 0.0604, "step": 20840 }, { "epoch": 0.7048919841779641, "grad_norm": 0.23151415586471558, "learning_rate": 9.328831219822172e-05, "loss": 0.0586, "step": 20850 }, { "epoch": 0.7052300618682174, "grad_norm": 0.42778369784355164, "learning_rate": 9.328003506468808e-05, "loss": 0.0606, "step": 20860 }, { "epoch": 0.7055681395584705, "grad_norm": 0.24242043495178223, "learning_rate": 9.327175319811488e-05, "loss": 0.0639, "step": 20870 }, { "epoch": 0.7059062172487237, "grad_norm": 0.19952288269996643, "learning_rate": 9.326346659940781e-05, "loss": 0.0546, "step": 20880 }, { "epoch": 0.706244294938977, "grad_norm": 0.32667386531829834, "learning_rate": 9.325517526947308e-05, "loss": 0.0615, "step": 20890 }, { "epoch": 0.7065823726292302, "grad_norm": 0.22731123864650726, "learning_rate": 9.32468792092174e-05, "loss": 0.0596, "step": 20900 }, { "epoch": 0.7069204503194834, "grad_norm": 0.2461690604686737, "learning_rate": 9.323857841954803e-05, "loss": 0.0593, "step": 20910 }, { "epoch": 0.7072585280097367, "grad_norm": 0.2924388647079468, "learning_rate": 9.323027290137276e-05, "loss": 0.0568, "step": 20920 }, { "epoch": 0.7075966056999898, "grad_norm": 0.36475229263305664, "learning_rate": 9.322196265559981e-05, "loss": 0.0573, "step": 20930 }, { "epoch": 0.7079346833902431, "grad_norm": 0.22771339118480682, "learning_rate": 9.321364768313803e-05, "loss": 0.061, "step": 20940 }, { "epoch": 0.7082727610804963, "grad_norm": 0.2554689347743988, "learning_rate": 9.32053279848967e-05, "loss": 0.0578, "step": 20950 }, { "epoch": 0.7086108387707495, "grad_norm": 0.24381229281425476, "learning_rate": 9.319700356178567e-05, "loss": 0.0551, "step": 20960 }, { "epoch": 0.7089489164610028, "grad_norm": 0.40705350041389465, "learning_rate": 9.318867441471527e-05, "loss": 0.0617, "step": 20970 }, { "epoch": 0.709286994151256, "grad_norm": 0.3422195613384247, "learning_rate": 9.318034054459637e-05, "loss": 0.0569, "step": 20980 }, { "epoch": 0.7096250718415091, "grad_norm": 0.18335507810115814, "learning_rate": 9.317200195234034e-05, "loss": 0.0596, "step": 20990 }, { "epoch": 0.7099631495317624, "grad_norm": 0.20883606374263763, "learning_rate": 9.316365863885909e-05, "loss": 0.0671, "step": 21000 }, { "epoch": 0.7103012272220156, "grad_norm": 0.24249468743801117, "learning_rate": 9.315531060506502e-05, "loss": 0.0664, "step": 21010 }, { "epoch": 0.7106393049122688, "grad_norm": 0.2359536588191986, "learning_rate": 9.314695785187108e-05, "loss": 0.0553, "step": 21020 }, { "epoch": 0.7109773826025221, "grad_norm": 0.22976000607013702, "learning_rate": 9.313860038019069e-05, "loss": 0.0614, "step": 21030 }, { "epoch": 0.7113154602927753, "grad_norm": 0.23082581162452698, "learning_rate": 9.313023819093782e-05, "loss": 0.069, "step": 21040 }, { "epoch": 0.7116535379830286, "grad_norm": 0.17525285482406616, "learning_rate": 9.312187128502695e-05, "loss": 0.0581, "step": 21050 }, { "epoch": 0.7119916156732817, "grad_norm": 0.3072092533111572, "learning_rate": 9.311349966337307e-05, "loss": 0.0619, "step": 21060 }, { "epoch": 0.7123296933635349, "grad_norm": 0.32168933749198914, "learning_rate": 9.310512332689169e-05, "loss": 0.06, "step": 21070 }, { "epoch": 0.7126677710537882, "grad_norm": 0.3081899583339691, "learning_rate": 9.309674227649883e-05, "loss": 0.0623, "step": 21080 }, { "epoch": 0.7130058487440414, "grad_norm": 0.3497020900249481, "learning_rate": 9.308835651311103e-05, "loss": 0.0655, "step": 21090 }, { "epoch": 0.7133439264342946, "grad_norm": 0.28495049476623535, "learning_rate": 9.307996603764533e-05, "loss": 0.0658, "step": 21100 }, { "epoch": 0.7136820041245479, "grad_norm": 0.3233051300048828, "learning_rate": 9.307157085101932e-05, "loss": 0.0572, "step": 21110 }, { "epoch": 0.714020081814801, "grad_norm": 0.5376464128494263, "learning_rate": 9.306317095415109e-05, "loss": 0.0594, "step": 21120 }, { "epoch": 0.7143581595050542, "grad_norm": 0.2761007249355316, "learning_rate": 9.305476634795922e-05, "loss": 0.0622, "step": 21130 }, { "epoch": 0.7146962371953075, "grad_norm": 0.25474220514297485, "learning_rate": 9.304635703336284e-05, "loss": 0.0557, "step": 21140 }, { "epoch": 0.7150343148855607, "grad_norm": 0.18886519968509674, "learning_rate": 9.303794301128157e-05, "loss": 0.0583, "step": 21150 }, { "epoch": 0.715372392575814, "grad_norm": 0.2953031361103058, "learning_rate": 9.302952428263555e-05, "loss": 0.0586, "step": 21160 }, { "epoch": 0.7157104702660672, "grad_norm": 0.258989155292511, "learning_rate": 9.302110084834545e-05, "loss": 0.0564, "step": 21170 }, { "epoch": 0.7160485479563203, "grad_norm": 0.3761790692806244, "learning_rate": 9.301267270933245e-05, "loss": 0.058, "step": 21180 }, { "epoch": 0.7163866256465736, "grad_norm": 0.21764717996120453, "learning_rate": 9.300423986651823e-05, "loss": 0.0571, "step": 21190 }, { "epoch": 0.7167247033368268, "grad_norm": 0.24089083075523376, "learning_rate": 9.299580232082501e-05, "loss": 0.0576, "step": 21200 }, { "epoch": 0.71706278102708, "grad_norm": 0.33856531977653503, "learning_rate": 9.298736007317547e-05, "loss": 0.0545, "step": 21210 }, { "epoch": 0.7174008587173333, "grad_norm": 0.2991529703140259, "learning_rate": 9.297891312449288e-05, "loss": 0.0582, "step": 21220 }, { "epoch": 0.7177389364075865, "grad_norm": 0.261785089969635, "learning_rate": 9.297046147570094e-05, "loss": 0.0576, "step": 21230 }, { "epoch": 0.7180770140978396, "grad_norm": 0.18606704473495483, "learning_rate": 9.296200512772396e-05, "loss": 0.0586, "step": 21240 }, { "epoch": 0.7184150917880929, "grad_norm": 0.4054858386516571, "learning_rate": 9.295354408148668e-05, "loss": 0.0601, "step": 21250 }, { "epoch": 0.7187531694783461, "grad_norm": 0.28763294219970703, "learning_rate": 9.294507833791441e-05, "loss": 0.0602, "step": 21260 }, { "epoch": 0.7190912471685993, "grad_norm": 0.22255483269691467, "learning_rate": 9.293660789793295e-05, "loss": 0.053, "step": 21270 }, { "epoch": 0.7194293248588526, "grad_norm": 0.2424289584159851, "learning_rate": 9.292813276246858e-05, "loss": 0.0584, "step": 21280 }, { "epoch": 0.7197674025491058, "grad_norm": 0.3932754695415497, "learning_rate": 9.291965293244816e-05, "loss": 0.0595, "step": 21290 }, { "epoch": 0.720105480239359, "grad_norm": 0.3870551884174347, "learning_rate": 9.291116840879904e-05, "loss": 0.0562, "step": 21300 }, { "epoch": 0.7204435579296122, "grad_norm": 0.3350449204444885, "learning_rate": 9.290267919244904e-05, "loss": 0.0643, "step": 21310 }, { "epoch": 0.7207816356198654, "grad_norm": 0.2024562954902649, "learning_rate": 9.289418528432655e-05, "loss": 0.0552, "step": 21320 }, { "epoch": 0.7211197133101187, "grad_norm": 0.2273927628993988, "learning_rate": 9.288568668536045e-05, "loss": 0.064, "step": 21330 }, { "epoch": 0.7214577910003719, "grad_norm": 0.2525302469730377, "learning_rate": 9.287718339648013e-05, "loss": 0.0539, "step": 21340 }, { "epoch": 0.7217958686906251, "grad_norm": 0.23310799896717072, "learning_rate": 9.28686754186155e-05, "loss": 0.0606, "step": 21350 }, { "epoch": 0.7221339463808784, "grad_norm": 0.2335798740386963, "learning_rate": 9.286016275269698e-05, "loss": 0.0577, "step": 21360 }, { "epoch": 0.7224720240711315, "grad_norm": 0.21027012169361115, "learning_rate": 9.285164539965551e-05, "loss": 0.0597, "step": 21370 }, { "epoch": 0.7228101017613847, "grad_norm": 0.21753095090389252, "learning_rate": 9.284312336042251e-05, "loss": 0.0532, "step": 21380 }, { "epoch": 0.723148179451638, "grad_norm": 0.512520432472229, "learning_rate": 9.283459663592996e-05, "loss": 0.0592, "step": 21390 }, { "epoch": 0.7234862571418912, "grad_norm": 0.28759095072746277, "learning_rate": 9.282606522711033e-05, "loss": 0.0591, "step": 21400 }, { "epoch": 0.7238243348321445, "grad_norm": 0.25274476408958435, "learning_rate": 9.281752913489657e-05, "loss": 0.0604, "step": 21410 }, { "epoch": 0.7241624125223977, "grad_norm": 0.2595599591732025, "learning_rate": 9.280898836022222e-05, "loss": 0.0534, "step": 21420 }, { "epoch": 0.7245004902126508, "grad_norm": 0.315261572599411, "learning_rate": 9.280044290402126e-05, "loss": 0.0641, "step": 21430 }, { "epoch": 0.7248385679029041, "grad_norm": 0.26976969838142395, "learning_rate": 9.279189276722821e-05, "loss": 0.0612, "step": 21440 }, { "epoch": 0.7251766455931573, "grad_norm": 0.554343044757843, "learning_rate": 9.278333795077812e-05, "loss": 0.0647, "step": 21450 }, { "epoch": 0.7255147232834105, "grad_norm": 0.2379419207572937, "learning_rate": 9.27747784556065e-05, "loss": 0.0598, "step": 21460 }, { "epoch": 0.7258528009736638, "grad_norm": 0.29268938302993774, "learning_rate": 9.276621428264942e-05, "loss": 0.0596, "step": 21470 }, { "epoch": 0.726190878663917, "grad_norm": 0.421866774559021, "learning_rate": 9.275764543284345e-05, "loss": 0.055, "step": 21480 }, { "epoch": 0.7265289563541701, "grad_norm": 0.4211269021034241, "learning_rate": 9.274907190712566e-05, "loss": 0.0548, "step": 21490 }, { "epoch": 0.7268670340444234, "grad_norm": 0.3370710015296936, "learning_rate": 9.274049370643363e-05, "loss": 0.0609, "step": 21500 }, { "epoch": 0.7272051117346766, "grad_norm": 0.2591937184333801, "learning_rate": 9.273191083170547e-05, "loss": 0.0535, "step": 21510 }, { "epoch": 0.7275431894249299, "grad_norm": 0.30871906876564026, "learning_rate": 9.27233232838798e-05, "loss": 0.0548, "step": 21520 }, { "epoch": 0.7278812671151831, "grad_norm": 0.30924755334854126, "learning_rate": 9.27147310638957e-05, "loss": 0.057, "step": 21530 }, { "epoch": 0.7282193448054363, "grad_norm": 0.17964774370193481, "learning_rate": 9.270613417269286e-05, "loss": 0.0574, "step": 21540 }, { "epoch": 0.7285574224956896, "grad_norm": 0.20989222824573517, "learning_rate": 9.269753261121138e-05, "loss": 0.0629, "step": 21550 }, { "epoch": 0.7288955001859427, "grad_norm": 0.3878730237483978, "learning_rate": 9.268892638039194e-05, "loss": 0.0613, "step": 21560 }, { "epoch": 0.7292335778761959, "grad_norm": 0.2635476589202881, "learning_rate": 9.268031548117569e-05, "loss": 0.0602, "step": 21570 }, { "epoch": 0.7295716555664492, "grad_norm": 0.2498933970928192, "learning_rate": 9.26716999145043e-05, "loss": 0.06, "step": 21580 }, { "epoch": 0.7299097332567024, "grad_norm": 0.18389195203781128, "learning_rate": 9.266307968131998e-05, "loss": 0.0623, "step": 21590 }, { "epoch": 0.7302478109469556, "grad_norm": 0.17684808373451233, "learning_rate": 9.26544547825654e-05, "loss": 0.0546, "step": 21600 }, { "epoch": 0.7305858886372089, "grad_norm": 0.28554797172546387, "learning_rate": 9.264582521918376e-05, "loss": 0.0609, "step": 21610 }, { "epoch": 0.730923966327462, "grad_norm": 0.2916090488433838, "learning_rate": 9.263719099211881e-05, "loss": 0.0633, "step": 21620 }, { "epoch": 0.7312620440177152, "grad_norm": 0.250898540019989, "learning_rate": 9.262855210231476e-05, "loss": 0.0651, "step": 21630 }, { "epoch": 0.7316001217079685, "grad_norm": 0.19925624132156372, "learning_rate": 9.261990855071633e-05, "loss": 0.0587, "step": 21640 }, { "epoch": 0.7319381993982217, "grad_norm": 0.3131753206253052, "learning_rate": 9.261126033826878e-05, "loss": 0.0633, "step": 21650 }, { "epoch": 0.732276277088475, "grad_norm": 0.26051101088523865, "learning_rate": 9.260260746591786e-05, "loss": 0.0602, "step": 21660 }, { "epoch": 0.7326143547787282, "grad_norm": 0.1890035718679428, "learning_rate": 9.259394993460985e-05, "loss": 0.064, "step": 21670 }, { "epoch": 0.7329524324689813, "grad_norm": 0.14868685603141785, "learning_rate": 9.258528774529151e-05, "loss": 0.0594, "step": 21680 }, { "epoch": 0.7332905101592346, "grad_norm": 0.3597698211669922, "learning_rate": 9.257662089891013e-05, "loss": 0.058, "step": 21690 }, { "epoch": 0.7336285878494878, "grad_norm": 0.20057925581932068, "learning_rate": 9.25679493964135e-05, "loss": 0.0583, "step": 21700 }, { "epoch": 0.733966665539741, "grad_norm": 0.22795704007148743, "learning_rate": 9.255927323874994e-05, "loss": 0.0607, "step": 21710 }, { "epoch": 0.7343047432299943, "grad_norm": 0.302321195602417, "learning_rate": 9.255059242686822e-05, "loss": 0.0625, "step": 21720 }, { "epoch": 0.7346428209202475, "grad_norm": 0.31557682156562805, "learning_rate": 9.254190696171769e-05, "loss": 0.0605, "step": 21730 }, { "epoch": 0.7349808986105006, "grad_norm": 0.23084093630313873, "learning_rate": 9.25332168442482e-05, "loss": 0.0594, "step": 21740 }, { "epoch": 0.7353189763007539, "grad_norm": 0.2981213927268982, "learning_rate": 9.252452207541004e-05, "loss": 0.0592, "step": 21750 }, { "epoch": 0.7356570539910071, "grad_norm": 0.25315216183662415, "learning_rate": 9.251582265615409e-05, "loss": 0.0596, "step": 21760 }, { "epoch": 0.7359951316812604, "grad_norm": 0.2554958164691925, "learning_rate": 9.250711858743169e-05, "loss": 0.0582, "step": 21770 }, { "epoch": 0.7363332093715136, "grad_norm": 0.22210641205310822, "learning_rate": 9.24984098701947e-05, "loss": 0.0664, "step": 21780 }, { "epoch": 0.7366712870617668, "grad_norm": 0.22282101213932037, "learning_rate": 9.248969650539552e-05, "loss": 0.0663, "step": 21790 }, { "epoch": 0.7370093647520201, "grad_norm": 0.21258527040481567, "learning_rate": 9.2480978493987e-05, "loss": 0.0561, "step": 21800 }, { "epoch": 0.7373474424422732, "grad_norm": 0.19982048869132996, "learning_rate": 9.247225583692256e-05, "loss": 0.059, "step": 21810 }, { "epoch": 0.7376855201325264, "grad_norm": 0.44936785101890564, "learning_rate": 9.246352853515607e-05, "loss": 0.0596, "step": 21820 }, { "epoch": 0.7380235978227797, "grad_norm": 0.3207259774208069, "learning_rate": 9.245479658964194e-05, "loss": 0.0618, "step": 21830 }, { "epoch": 0.7383616755130329, "grad_norm": 0.31005704402923584, "learning_rate": 9.244606000133507e-05, "loss": 0.0551, "step": 21840 }, { "epoch": 0.7386997532032861, "grad_norm": 0.1923457235097885, "learning_rate": 9.24373187711909e-05, "loss": 0.0547, "step": 21850 }, { "epoch": 0.7390378308935394, "grad_norm": 0.28310051560401917, "learning_rate": 9.242857290016537e-05, "loss": 0.058, "step": 21860 }, { "epoch": 0.7393759085837925, "grad_norm": 0.17628835141658783, "learning_rate": 9.241982238921488e-05, "loss": 0.0606, "step": 21870 }, { "epoch": 0.7397139862740457, "grad_norm": 0.28826674818992615, "learning_rate": 9.24110672392964e-05, "loss": 0.0714, "step": 21880 }, { "epoch": 0.740052063964299, "grad_norm": 0.20909181237220764, "learning_rate": 9.240230745136737e-05, "loss": 0.0598, "step": 21890 }, { "epoch": 0.7403901416545522, "grad_norm": 0.3375549614429474, "learning_rate": 9.239354302638575e-05, "loss": 0.0573, "step": 21900 }, { "epoch": 0.7407282193448055, "grad_norm": 0.24067994952201843, "learning_rate": 9.238477396531e-05, "loss": 0.067, "step": 21910 }, { "epoch": 0.7410662970350587, "grad_norm": 0.2668810188770294, "learning_rate": 9.23760002690991e-05, "loss": 0.0513, "step": 21920 }, { "epoch": 0.7414043747253118, "grad_norm": 0.23998263478279114, "learning_rate": 9.236722193871252e-05, "loss": 0.0639, "step": 21930 }, { "epoch": 0.7417424524155651, "grad_norm": 0.2010355144739151, "learning_rate": 9.235843897511023e-05, "loss": 0.0643, "step": 21940 }, { "epoch": 0.7420805301058183, "grad_norm": 0.27033448219299316, "learning_rate": 9.234965137925276e-05, "loss": 0.0633, "step": 21950 }, { "epoch": 0.7424186077960715, "grad_norm": 0.347785621881485, "learning_rate": 9.234085915210108e-05, "loss": 0.0553, "step": 21960 }, { "epoch": 0.7427566854863248, "grad_norm": 0.2098660171031952, "learning_rate": 9.23320622946167e-05, "loss": 0.0635, "step": 21970 }, { "epoch": 0.743094763176578, "grad_norm": 0.4058968126773834, "learning_rate": 9.232326080776163e-05, "loss": 0.0579, "step": 21980 }, { "epoch": 0.7434328408668311, "grad_norm": 0.22697614133358002, "learning_rate": 9.23144546924984e-05, "loss": 0.0596, "step": 21990 }, { "epoch": 0.7437709185570844, "grad_norm": 0.5741044282913208, "learning_rate": 9.230564394979e-05, "loss": 0.0641, "step": 22000 }, { "epoch": 0.7441089962473376, "grad_norm": 0.3304191529750824, "learning_rate": 9.22968285806e-05, "loss": 0.0567, "step": 22010 }, { "epoch": 0.7444470739375909, "grad_norm": 0.26356276869773865, "learning_rate": 9.228800858589242e-05, "loss": 0.0654, "step": 22020 }, { "epoch": 0.7447851516278441, "grad_norm": 0.28696581721305847, "learning_rate": 9.227918396663179e-05, "loss": 0.0611, "step": 22030 }, { "epoch": 0.7451232293180973, "grad_norm": 0.23814429342746735, "learning_rate": 9.227035472378319e-05, "loss": 0.0565, "step": 22040 }, { "epoch": 0.7454613070083506, "grad_norm": 0.3054503798484802, "learning_rate": 9.226152085831213e-05, "loss": 0.0589, "step": 22050 }, { "epoch": 0.7457993846986037, "grad_norm": 0.27004119753837585, "learning_rate": 9.22526823711847e-05, "loss": 0.055, "step": 22060 }, { "epoch": 0.7461374623888569, "grad_norm": 0.24826477468013763, "learning_rate": 9.224383926336745e-05, "loss": 0.0577, "step": 22070 }, { "epoch": 0.7464755400791102, "grad_norm": 0.2045440971851349, "learning_rate": 9.223499153582744e-05, "loss": 0.0566, "step": 22080 }, { "epoch": 0.7468136177693634, "grad_norm": 0.3642074465751648, "learning_rate": 9.222613918953226e-05, "loss": 0.0604, "step": 22090 }, { "epoch": 0.7471516954596166, "grad_norm": 0.428373247385025, "learning_rate": 9.221728222544999e-05, "loss": 0.0625, "step": 22100 }, { "epoch": 0.7474897731498699, "grad_norm": 0.19878122210502625, "learning_rate": 9.22084206445492e-05, "loss": 0.0597, "step": 22110 }, { "epoch": 0.747827850840123, "grad_norm": 0.21259169280529022, "learning_rate": 9.2199554447799e-05, "loss": 0.0571, "step": 22120 }, { "epoch": 0.7481659285303763, "grad_norm": 0.2894558310508728, "learning_rate": 9.219068363616897e-05, "loss": 0.0534, "step": 22130 }, { "epoch": 0.7485040062206295, "grad_norm": 0.37792733311653137, "learning_rate": 9.218180821062919e-05, "loss": 0.0597, "step": 22140 }, { "epoch": 0.7488420839108827, "grad_norm": 0.35417765378952026, "learning_rate": 9.21729281721503e-05, "loss": 0.0613, "step": 22150 }, { "epoch": 0.749180161601136, "grad_norm": 0.25242164731025696, "learning_rate": 9.216404352170339e-05, "loss": 0.0666, "step": 22160 }, { "epoch": 0.7495182392913892, "grad_norm": 0.3989343047142029, "learning_rate": 9.215515426026007e-05, "loss": 0.063, "step": 22170 }, { "epoch": 0.7498563169816423, "grad_norm": 0.2290535420179367, "learning_rate": 9.214626038879246e-05, "loss": 0.0613, "step": 22180 }, { "epoch": 0.7501943946718956, "grad_norm": 0.2429257184267044, "learning_rate": 9.21373619082732e-05, "loss": 0.0588, "step": 22190 }, { "epoch": 0.7505324723621488, "grad_norm": 0.1966559737920761, "learning_rate": 9.212845881967535e-05, "loss": 0.0614, "step": 22200 }, { "epoch": 0.750870550052402, "grad_norm": 0.23444212973117828, "learning_rate": 9.211955112397262e-05, "loss": 0.06, "step": 22210 }, { "epoch": 0.7512086277426553, "grad_norm": 0.2924877107143402, "learning_rate": 9.211063882213909e-05, "loss": 0.0615, "step": 22220 }, { "epoch": 0.7515467054329085, "grad_norm": 0.26931676268577576, "learning_rate": 9.210172191514942e-05, "loss": 0.0597, "step": 22230 }, { "epoch": 0.7518847831231616, "grad_norm": 0.28901270031929016, "learning_rate": 9.209280040397874e-05, "loss": 0.0603, "step": 22240 }, { "epoch": 0.7522228608134149, "grad_norm": 0.19736237823963165, "learning_rate": 9.208387428960268e-05, "loss": 0.057, "step": 22250 }, { "epoch": 0.7525609385036681, "grad_norm": 0.16481591761112213, "learning_rate": 9.20749435729974e-05, "loss": 0.0586, "step": 22260 }, { "epoch": 0.7528990161939214, "grad_norm": 0.2022933065891266, "learning_rate": 9.206600825513957e-05, "loss": 0.054, "step": 22270 }, { "epoch": 0.7532370938841746, "grad_norm": 0.22745651006698608, "learning_rate": 9.20570683370063e-05, "loss": 0.0558, "step": 22280 }, { "epoch": 0.7535751715744278, "grad_norm": 0.21314512193202972, "learning_rate": 9.204812381957528e-05, "loss": 0.0572, "step": 22290 }, { "epoch": 0.7539132492646811, "grad_norm": 0.4527108073234558, "learning_rate": 9.203917470382465e-05, "loss": 0.0594, "step": 22300 }, { "epoch": 0.7542513269549342, "grad_norm": 0.3078381419181824, "learning_rate": 9.203022099073309e-05, "loss": 0.062, "step": 22310 }, { "epoch": 0.7545894046451874, "grad_norm": 0.38845473527908325, "learning_rate": 9.202126268127976e-05, "loss": 0.0581, "step": 22320 }, { "epoch": 0.7549274823354407, "grad_norm": 0.21393665671348572, "learning_rate": 9.20122997764443e-05, "loss": 0.0632, "step": 22330 }, { "epoch": 0.7552655600256939, "grad_norm": 0.21182286739349365, "learning_rate": 9.200333227720692e-05, "loss": 0.0601, "step": 22340 }, { "epoch": 0.7556036377159471, "grad_norm": 0.27002575993537903, "learning_rate": 9.199436018454826e-05, "loss": 0.0647, "step": 22350 }, { "epoch": 0.7559417154062004, "grad_norm": 0.22772428393363953, "learning_rate": 9.198538349944952e-05, "loss": 0.056, "step": 22360 }, { "epoch": 0.7562797930964535, "grad_norm": 0.2892777621746063, "learning_rate": 9.197640222289234e-05, "loss": 0.0612, "step": 22370 }, { "epoch": 0.7566178707867068, "grad_norm": 0.32169872522354126, "learning_rate": 9.196741635585895e-05, "loss": 0.0581, "step": 22380 }, { "epoch": 0.75695594847696, "grad_norm": 0.2296375334262848, "learning_rate": 9.195842589933199e-05, "loss": 0.0584, "step": 22390 }, { "epoch": 0.7572940261672132, "grad_norm": 0.24574925005435944, "learning_rate": 9.194943085429466e-05, "loss": 0.0641, "step": 22400 }, { "epoch": 0.7576321038574665, "grad_norm": 0.4108474850654602, "learning_rate": 9.194043122173065e-05, "loss": 0.0626, "step": 22410 }, { "epoch": 0.7579701815477197, "grad_norm": 0.2980457842350006, "learning_rate": 9.193142700262413e-05, "loss": 0.0619, "step": 22420 }, { "epoch": 0.7583082592379728, "grad_norm": 0.20577716827392578, "learning_rate": 9.192241819795979e-05, "loss": 0.0606, "step": 22430 }, { "epoch": 0.7586463369282261, "grad_norm": 0.2088327556848526, "learning_rate": 9.191340480872284e-05, "loss": 0.0622, "step": 22440 }, { "epoch": 0.7589844146184793, "grad_norm": 0.26821139454841614, "learning_rate": 9.190438683589895e-05, "loss": 0.0627, "step": 22450 }, { "epoch": 0.7593224923087325, "grad_norm": 0.1959112286567688, "learning_rate": 9.189536428047432e-05, "loss": 0.0586, "step": 22460 }, { "epoch": 0.7596605699989858, "grad_norm": 0.2092367708683014, "learning_rate": 9.188633714343564e-05, "loss": 0.0574, "step": 22470 }, { "epoch": 0.759998647689239, "grad_norm": 0.2559226453304291, "learning_rate": 9.18773054257701e-05, "loss": 0.0616, "step": 22480 }, { "epoch": 0.7603367253794923, "grad_norm": 0.2574523389339447, "learning_rate": 9.18682691284654e-05, "loss": 0.0545, "step": 22490 }, { "epoch": 0.7606748030697454, "grad_norm": 0.19138358533382416, "learning_rate": 9.185922825250974e-05, "loss": 0.0619, "step": 22500 }, { "epoch": 0.7610128807599986, "grad_norm": 0.23528268933296204, "learning_rate": 9.185018279889181e-05, "loss": 0.0586, "step": 22510 }, { "epoch": 0.7613509584502519, "grad_norm": 0.33987727761268616, "learning_rate": 9.184113276860082e-05, "loss": 0.065, "step": 22520 }, { "epoch": 0.7616890361405051, "grad_norm": 0.23607468605041504, "learning_rate": 9.183207816262645e-05, "loss": 0.0532, "step": 22530 }, { "epoch": 0.7620271138307583, "grad_norm": 0.24754171073436737, "learning_rate": 9.182301898195891e-05, "loss": 0.065, "step": 22540 }, { "epoch": 0.7623651915210116, "grad_norm": 0.2888484001159668, "learning_rate": 9.181395522758889e-05, "loss": 0.065, "step": 22550 }, { "epoch": 0.7627032692112647, "grad_norm": 0.16225308179855347, "learning_rate": 9.180488690050759e-05, "loss": 0.063, "step": 22560 }, { "epoch": 0.7630413469015179, "grad_norm": 0.24121570587158203, "learning_rate": 9.179581400170671e-05, "loss": 0.061, "step": 22570 }, { "epoch": 0.7633794245917712, "grad_norm": 0.28718557953834534, "learning_rate": 9.178673653217845e-05, "loss": 0.0571, "step": 22580 }, { "epoch": 0.7637175022820244, "grad_norm": 0.20955964922904968, "learning_rate": 9.177765449291551e-05, "loss": 0.0547, "step": 22590 }, { "epoch": 0.7640555799722776, "grad_norm": 0.2357119917869568, "learning_rate": 9.176856788491109e-05, "loss": 0.0592, "step": 22600 }, { "epoch": 0.7643936576625309, "grad_norm": 0.2840127646923065, "learning_rate": 9.175947670915887e-05, "loss": 0.0637, "step": 22610 }, { "epoch": 0.764731735352784, "grad_norm": 0.5430931448936462, "learning_rate": 9.175038096665309e-05, "loss": 0.0567, "step": 22620 }, { "epoch": 0.7650698130430373, "grad_norm": 0.25070780515670776, "learning_rate": 9.17412806583884e-05, "loss": 0.0608, "step": 22630 }, { "epoch": 0.7654078907332905, "grad_norm": 0.3658379912376404, "learning_rate": 9.173217578536002e-05, "loss": 0.0655, "step": 22640 }, { "epoch": 0.7657459684235437, "grad_norm": 0.20699714124202728, "learning_rate": 9.172306634856362e-05, "loss": 0.0613, "step": 22650 }, { "epoch": 0.766084046113797, "grad_norm": 0.2752435505390167, "learning_rate": 9.171395234899545e-05, "loss": 0.0624, "step": 22660 }, { "epoch": 0.7664221238040502, "grad_norm": 0.2566186785697937, "learning_rate": 9.170483378765214e-05, "loss": 0.0612, "step": 22670 }, { "epoch": 0.7667602014943034, "grad_norm": 0.2647065818309784, "learning_rate": 9.169571066553091e-05, "loss": 0.0562, "step": 22680 }, { "epoch": 0.7670982791845566, "grad_norm": 0.25358277559280396, "learning_rate": 9.168658298362946e-05, "loss": 0.058, "step": 22690 }, { "epoch": 0.7674363568748098, "grad_norm": 0.1882782280445099, "learning_rate": 9.167745074294598e-05, "loss": 0.0585, "step": 22700 }, { "epoch": 0.767774434565063, "grad_norm": 0.24664632976055145, "learning_rate": 9.166831394447913e-05, "loss": 0.0645, "step": 22710 }, { "epoch": 0.7681125122553163, "grad_norm": 0.23213115334510803, "learning_rate": 9.165917258922812e-05, "loss": 0.0585, "step": 22720 }, { "epoch": 0.7684505899455695, "grad_norm": 0.3694913685321808, "learning_rate": 9.165002667819262e-05, "loss": 0.0605, "step": 22730 }, { "epoch": 0.7687886676358228, "grad_norm": 0.2455226480960846, "learning_rate": 9.164087621237282e-05, "loss": 0.0602, "step": 22740 }, { "epoch": 0.769126745326076, "grad_norm": 0.25459229946136475, "learning_rate": 9.163172119276942e-05, "loss": 0.0633, "step": 22750 }, { "epoch": 0.7694648230163291, "grad_norm": 0.19977781176567078, "learning_rate": 9.162256162038358e-05, "loss": 0.0602, "step": 22760 }, { "epoch": 0.7698029007065824, "grad_norm": 0.21512912213802338, "learning_rate": 9.161339749621698e-05, "loss": 0.0596, "step": 22770 }, { "epoch": 0.7701409783968356, "grad_norm": 0.29518917202949524, "learning_rate": 9.160422882127177e-05, "loss": 0.0555, "step": 22780 }, { "epoch": 0.7704790560870888, "grad_norm": 0.234594464302063, "learning_rate": 9.159505559655069e-05, "loss": 0.0544, "step": 22790 }, { "epoch": 0.7708171337773421, "grad_norm": 0.19613686203956604, "learning_rate": 9.158587782305684e-05, "loss": 0.0621, "step": 22800 }, { "epoch": 0.7711552114675952, "grad_norm": 0.3736155331134796, "learning_rate": 9.157669550179391e-05, "loss": 0.058, "step": 22810 }, { "epoch": 0.7714932891578484, "grad_norm": 0.208290234208107, "learning_rate": 9.156750863376609e-05, "loss": 0.0565, "step": 22820 }, { "epoch": 0.7718313668481017, "grad_norm": 0.146162748336792, "learning_rate": 9.155831721997801e-05, "loss": 0.0588, "step": 22830 }, { "epoch": 0.7721694445383549, "grad_norm": 0.2842203974723816, "learning_rate": 9.154912126143484e-05, "loss": 0.0609, "step": 22840 }, { "epoch": 0.7725075222286081, "grad_norm": 0.2894570529460907, "learning_rate": 9.153992075914224e-05, "loss": 0.0594, "step": 22850 }, { "epoch": 0.7728455999188614, "grad_norm": 0.29238298535346985, "learning_rate": 9.153071571410635e-05, "loss": 0.0516, "step": 22860 }, { "epoch": 0.7731836776091146, "grad_norm": 0.19259214401245117, "learning_rate": 9.152150612733384e-05, "loss": 0.0582, "step": 22870 }, { "epoch": 0.7735217552993678, "grad_norm": 0.27568140625953674, "learning_rate": 9.151229199983184e-05, "loss": 0.0599, "step": 22880 }, { "epoch": 0.773859832989621, "grad_norm": 0.18233126401901245, "learning_rate": 9.150307333260802e-05, "loss": 0.0562, "step": 22890 }, { "epoch": 0.7741979106798742, "grad_norm": 0.24721713364124298, "learning_rate": 9.149385012667048e-05, "loss": 0.0579, "step": 22900 }, { "epoch": 0.7745359883701275, "grad_norm": 0.41879913210868835, "learning_rate": 9.148462238302788e-05, "loss": 0.0524, "step": 22910 }, { "epoch": 0.7748740660603807, "grad_norm": 0.18139295279979706, "learning_rate": 9.147539010268936e-05, "loss": 0.0573, "step": 22920 }, { "epoch": 0.7752121437506339, "grad_norm": 0.22838547825813293, "learning_rate": 9.14661532866645e-05, "loss": 0.0575, "step": 22930 }, { "epoch": 0.7755502214408871, "grad_norm": 0.24186430871486664, "learning_rate": 9.145691193596348e-05, "loss": 0.0579, "step": 22940 }, { "epoch": 0.7758882991311403, "grad_norm": 0.2071639597415924, "learning_rate": 9.144766605159691e-05, "loss": 0.0646, "step": 22950 }, { "epoch": 0.7762263768213935, "grad_norm": 0.20914189517498016, "learning_rate": 9.14384156345759e-05, "loss": 0.0674, "step": 22960 }, { "epoch": 0.7765644545116468, "grad_norm": 0.2199511080980301, "learning_rate": 9.142916068591204e-05, "loss": 0.0632, "step": 22970 }, { "epoch": 0.7769025322019, "grad_norm": 0.33733320236206055, "learning_rate": 9.141990120661746e-05, "loss": 0.058, "step": 22980 }, { "epoch": 0.7772406098921533, "grad_norm": 0.24030464887619019, "learning_rate": 9.141063719770475e-05, "loss": 0.0635, "step": 22990 }, { "epoch": 0.7775786875824064, "grad_norm": 0.3222767412662506, "learning_rate": 9.140136866018704e-05, "loss": 0.057, "step": 23000 }, { "epoch": 0.7779167652726596, "grad_norm": 0.26353657245635986, "learning_rate": 9.139209559507788e-05, "loss": 0.0598, "step": 23010 }, { "epoch": 0.7782548429629129, "grad_norm": 0.2449493110179901, "learning_rate": 9.13828180033914e-05, "loss": 0.0616, "step": 23020 }, { "epoch": 0.7785929206531661, "grad_norm": 0.28312066197395325, "learning_rate": 9.137353588614212e-05, "loss": 0.0567, "step": 23030 }, { "epoch": 0.7789309983434193, "grad_norm": 0.18645238876342773, "learning_rate": 9.136424924434519e-05, "loss": 0.0592, "step": 23040 }, { "epoch": 0.7792690760336726, "grad_norm": 0.32635530829429626, "learning_rate": 9.135495807901615e-05, "loss": 0.0546, "step": 23050 }, { "epoch": 0.7796071537239257, "grad_norm": 0.21970981359481812, "learning_rate": 9.134566239117108e-05, "loss": 0.0586, "step": 23060 }, { "epoch": 0.7799452314141789, "grad_norm": 0.21213573217391968, "learning_rate": 9.13363621818265e-05, "loss": 0.0552, "step": 23070 }, { "epoch": 0.7802833091044322, "grad_norm": 0.21429438889026642, "learning_rate": 9.132705745199953e-05, "loss": 0.0648, "step": 23080 }, { "epoch": 0.7806213867946854, "grad_norm": 0.30116552114486694, "learning_rate": 9.131774820270768e-05, "loss": 0.0566, "step": 23090 }, { "epoch": 0.7809594644849387, "grad_norm": 0.21313196420669556, "learning_rate": 9.130843443496901e-05, "loss": 0.0575, "step": 23100 }, { "epoch": 0.7812975421751919, "grad_norm": 0.19629648327827454, "learning_rate": 9.129911614980206e-05, "loss": 0.063, "step": 23110 }, { "epoch": 0.781635619865445, "grad_norm": 0.1994994729757309, "learning_rate": 9.128979334822584e-05, "loss": 0.061, "step": 23120 }, { "epoch": 0.7819736975556983, "grad_norm": 0.25501614809036255, "learning_rate": 9.128046603125992e-05, "loss": 0.0561, "step": 23130 }, { "epoch": 0.7823117752459515, "grad_norm": 0.3065541982650757, "learning_rate": 9.12711341999243e-05, "loss": 0.0579, "step": 23140 }, { "epoch": 0.7826498529362047, "grad_norm": 0.8281217217445374, "learning_rate": 9.12617978552395e-05, "loss": 0.0682, "step": 23150 }, { "epoch": 0.782987930626458, "grad_norm": 0.20850370824337006, "learning_rate": 9.12524569982265e-05, "loss": 0.064, "step": 23160 }, { "epoch": 0.7833260083167112, "grad_norm": 0.2224050760269165, "learning_rate": 9.124311162990684e-05, "loss": 0.059, "step": 23170 }, { "epoch": 0.7836640860069644, "grad_norm": 0.3445676565170288, "learning_rate": 9.12337617513025e-05, "loss": 0.0578, "step": 23180 }, { "epoch": 0.7840021636972176, "grad_norm": 0.2645328938961029, "learning_rate": 9.122440736343596e-05, "loss": 0.0608, "step": 23190 }, { "epoch": 0.7843402413874708, "grad_norm": 0.27747976779937744, "learning_rate": 9.12150484673302e-05, "loss": 0.0591, "step": 23200 }, { "epoch": 0.784678319077724, "grad_norm": 0.2797681987285614, "learning_rate": 9.120568506400873e-05, "loss": 0.0603, "step": 23210 }, { "epoch": 0.7850163967679773, "grad_norm": 0.16944219172000885, "learning_rate": 9.119631715449548e-05, "loss": 0.0591, "step": 23220 }, { "epoch": 0.7853544744582305, "grad_norm": 0.21275480091571808, "learning_rate": 9.118694473981493e-05, "loss": 0.0534, "step": 23230 }, { "epoch": 0.7856925521484838, "grad_norm": 0.2749585807323456, "learning_rate": 9.117756782099203e-05, "loss": 0.0542, "step": 23240 }, { "epoch": 0.786030629838737, "grad_norm": 0.2741071879863739, "learning_rate": 9.11681863990522e-05, "loss": 0.0589, "step": 23250 }, { "epoch": 0.7863687075289901, "grad_norm": 0.24571874737739563, "learning_rate": 9.115880047502142e-05, "loss": 0.0628, "step": 23260 }, { "epoch": 0.7867067852192434, "grad_norm": 0.2042318731546402, "learning_rate": 9.114941004992609e-05, "loss": 0.0537, "step": 23270 }, { "epoch": 0.7870448629094966, "grad_norm": 0.24090184271335602, "learning_rate": 9.114001512479317e-05, "loss": 0.0554, "step": 23280 }, { "epoch": 0.7873829405997498, "grad_norm": 0.30542710423469543, "learning_rate": 9.113061570065003e-05, "loss": 0.0564, "step": 23290 }, { "epoch": 0.7877210182900031, "grad_norm": 0.6658942103385925, "learning_rate": 9.112121177852459e-05, "loss": 0.0581, "step": 23300 }, { "epoch": 0.7880590959802563, "grad_norm": 0.2135474681854248, "learning_rate": 9.111180335944527e-05, "loss": 0.0633, "step": 23310 }, { "epoch": 0.7883971736705094, "grad_norm": 0.5933997631072998, "learning_rate": 9.110239044444093e-05, "loss": 0.0546, "step": 23320 }, { "epoch": 0.7887352513607627, "grad_norm": 0.39621061086654663, "learning_rate": 9.109297303454099e-05, "loss": 0.0631, "step": 23330 }, { "epoch": 0.7890733290510159, "grad_norm": 0.23989298939704895, "learning_rate": 9.108355113077526e-05, "loss": 0.0585, "step": 23340 }, { "epoch": 0.7894114067412692, "grad_norm": 0.2116885930299759, "learning_rate": 9.107412473417419e-05, "loss": 0.0603, "step": 23350 }, { "epoch": 0.7897494844315224, "grad_norm": 0.35804665088653564, "learning_rate": 9.106469384576858e-05, "loss": 0.0583, "step": 23360 }, { "epoch": 0.7900875621217756, "grad_norm": 0.23995791375637054, "learning_rate": 9.105525846658978e-05, "loss": 0.0556, "step": 23370 }, { "epoch": 0.7904256398120288, "grad_norm": 0.22252580523490906, "learning_rate": 9.104581859766965e-05, "loss": 0.0539, "step": 23380 }, { "epoch": 0.790763717502282, "grad_norm": 0.4259137213230133, "learning_rate": 9.10363742400405e-05, "loss": 0.0513, "step": 23390 }, { "epoch": 0.7911017951925352, "grad_norm": 0.24561749398708344, "learning_rate": 9.102692539473518e-05, "loss": 0.0555, "step": 23400 }, { "epoch": 0.7914398728827885, "grad_norm": 0.3080173134803772, "learning_rate": 9.101747206278697e-05, "loss": 0.0534, "step": 23410 }, { "epoch": 0.7917779505730417, "grad_norm": 0.29890745878219604, "learning_rate": 9.100801424522968e-05, "loss": 0.055, "step": 23420 }, { "epoch": 0.7921160282632949, "grad_norm": 0.23578275740146637, "learning_rate": 9.099855194309762e-05, "loss": 0.0591, "step": 23430 }, { "epoch": 0.7924541059535481, "grad_norm": 0.31159383058547974, "learning_rate": 9.098908515742554e-05, "loss": 0.0634, "step": 23440 }, { "epoch": 0.7927921836438013, "grad_norm": 0.2241581678390503, "learning_rate": 9.097961388924873e-05, "loss": 0.055, "step": 23450 }, { "epoch": 0.7931302613340546, "grad_norm": 0.24501119554042816, "learning_rate": 9.097013813960298e-05, "loss": 0.0602, "step": 23460 }, { "epoch": 0.7934683390243078, "grad_norm": 0.23562286794185638, "learning_rate": 9.09606579095245e-05, "loss": 0.0624, "step": 23470 }, { "epoch": 0.793806416714561, "grad_norm": 0.35604146122932434, "learning_rate": 9.095117320005008e-05, "loss": 0.0611, "step": 23480 }, { "epoch": 0.7941444944048143, "grad_norm": 0.6863433718681335, "learning_rate": 9.094168401221691e-05, "loss": 0.0548, "step": 23490 }, { "epoch": 0.7944825720950675, "grad_norm": 0.33549219369888306, "learning_rate": 9.093219034706273e-05, "loss": 0.0614, "step": 23500 }, { "epoch": 0.7948206497853206, "grad_norm": 0.4423505663871765, "learning_rate": 9.092269220562577e-05, "loss": 0.0738, "step": 23510 }, { "epoch": 0.7951587274755739, "grad_norm": 0.24223600327968597, "learning_rate": 9.09131895889447e-05, "loss": 0.0567, "step": 23520 }, { "epoch": 0.7954968051658271, "grad_norm": 0.21003736555576324, "learning_rate": 9.090368249805873e-05, "loss": 0.0637, "step": 23530 }, { "epoch": 0.7958348828560803, "grad_norm": 0.24546977877616882, "learning_rate": 9.089417093400754e-05, "loss": 0.0689, "step": 23540 }, { "epoch": 0.7961729605463336, "grad_norm": 0.30981817841529846, "learning_rate": 9.088465489783131e-05, "loss": 0.0591, "step": 23550 }, { "epoch": 0.7965110382365868, "grad_norm": 0.19695866107940674, "learning_rate": 9.087513439057068e-05, "loss": 0.0582, "step": 23560 }, { "epoch": 0.7968491159268399, "grad_norm": 0.208661749958992, "learning_rate": 9.08656094132668e-05, "loss": 0.0656, "step": 23570 }, { "epoch": 0.7971871936170932, "grad_norm": 0.21197329461574554, "learning_rate": 9.085607996696134e-05, "loss": 0.0573, "step": 23580 }, { "epoch": 0.7975252713073464, "grad_norm": 0.33638623356819153, "learning_rate": 9.084654605269639e-05, "loss": 0.055, "step": 23590 }, { "epoch": 0.7978633489975997, "grad_norm": 0.3383524417877197, "learning_rate": 9.083700767151457e-05, "loss": 0.0615, "step": 23600 }, { "epoch": 0.7982014266878529, "grad_norm": 0.15952497720718384, "learning_rate": 9.082746482445898e-05, "loss": 0.0592, "step": 23610 }, { "epoch": 0.7985395043781061, "grad_norm": 0.26078882813453674, "learning_rate": 9.081791751257325e-05, "loss": 0.0614, "step": 23620 }, { "epoch": 0.7988775820683593, "grad_norm": 0.24691076576709747, "learning_rate": 9.080836573690142e-05, "loss": 0.0628, "step": 23630 }, { "epoch": 0.7992156597586125, "grad_norm": 0.20663334429264069, "learning_rate": 9.079880949848805e-05, "loss": 0.0598, "step": 23640 }, { "epoch": 0.7995537374488657, "grad_norm": 0.19621779024600983, "learning_rate": 9.078924879837822e-05, "loss": 0.0565, "step": 23650 }, { "epoch": 0.799891815139119, "grad_norm": 0.19541054964065552, "learning_rate": 9.077968363761747e-05, "loss": 0.0596, "step": 23660 }, { "epoch": 0.8002298928293722, "grad_norm": 0.250815212726593, "learning_rate": 9.077011401725182e-05, "loss": 0.057, "step": 23670 }, { "epoch": 0.8005679705196254, "grad_norm": 0.18292413651943207, "learning_rate": 9.07605399383278e-05, "loss": 0.0588, "step": 23680 }, { "epoch": 0.8009060482098787, "grad_norm": 0.2883249521255493, "learning_rate": 9.075096140189243e-05, "loss": 0.0586, "step": 23690 }, { "epoch": 0.8012441259001318, "grad_norm": 0.3418813645839691, "learning_rate": 9.074137840899318e-05, "loss": 0.064, "step": 23700 }, { "epoch": 0.8015822035903851, "grad_norm": 0.21498699486255646, "learning_rate": 9.073179096067804e-05, "loss": 0.0585, "step": 23710 }, { "epoch": 0.8019202812806383, "grad_norm": 0.23336772620677948, "learning_rate": 9.072219905799549e-05, "loss": 0.0594, "step": 23720 }, { "epoch": 0.8022583589708915, "grad_norm": 0.28742894530296326, "learning_rate": 9.071260270199447e-05, "loss": 0.0599, "step": 23730 }, { "epoch": 0.8025964366611448, "grad_norm": 0.24729542434215546, "learning_rate": 9.070300189372441e-05, "loss": 0.063, "step": 23740 }, { "epoch": 0.802934514351398, "grad_norm": 0.22741109132766724, "learning_rate": 9.069339663423528e-05, "loss": 0.0611, "step": 23750 }, { "epoch": 0.8032725920416511, "grad_norm": 0.19725771248340607, "learning_rate": 9.068378692457747e-05, "loss": 0.0549, "step": 23760 }, { "epoch": 0.8036106697319044, "grad_norm": 0.24206256866455078, "learning_rate": 9.067417276580189e-05, "loss": 0.0578, "step": 23770 }, { "epoch": 0.8039487474221576, "grad_norm": 0.2981005609035492, "learning_rate": 9.066455415895993e-05, "loss": 0.0594, "step": 23780 }, { "epoch": 0.8042868251124108, "grad_norm": 0.18581551313400269, "learning_rate": 9.065493110510346e-05, "loss": 0.0506, "step": 23790 }, { "epoch": 0.8046249028026641, "grad_norm": 0.19254644215106964, "learning_rate": 9.064530360528484e-05, "loss": 0.0577, "step": 23800 }, { "epoch": 0.8049629804929173, "grad_norm": 0.22225956618785858, "learning_rate": 9.063567166055695e-05, "loss": 0.0598, "step": 23810 }, { "epoch": 0.8053010581831704, "grad_norm": 0.19973556697368622, "learning_rate": 9.062603527197308e-05, "loss": 0.0537, "step": 23820 }, { "epoch": 0.8056391358734237, "grad_norm": 0.15714919567108154, "learning_rate": 9.06163944405871e-05, "loss": 0.0546, "step": 23830 }, { "epoch": 0.8059772135636769, "grad_norm": 0.23490846157073975, "learning_rate": 9.060674916745327e-05, "loss": 0.0605, "step": 23840 }, { "epoch": 0.8063152912539302, "grad_norm": 0.2500006854534149, "learning_rate": 9.05970994536264e-05, "loss": 0.048, "step": 23850 }, { "epoch": 0.8066533689441834, "grad_norm": 0.300422728061676, "learning_rate": 9.05874453001618e-05, "loss": 0.0597, "step": 23860 }, { "epoch": 0.8069914466344366, "grad_norm": 0.23553822934627533, "learning_rate": 9.057778670811517e-05, "loss": 0.0503, "step": 23870 }, { "epoch": 0.8073295243246899, "grad_norm": 0.6440113186836243, "learning_rate": 9.056812367854281e-05, "loss": 0.0581, "step": 23880 }, { "epoch": 0.807667602014943, "grad_norm": 0.648158609867096, "learning_rate": 9.055845621250143e-05, "loss": 0.0642, "step": 23890 }, { "epoch": 0.8080056797051962, "grad_norm": 0.1692105382680893, "learning_rate": 9.054878431104825e-05, "loss": 0.0591, "step": 23900 }, { "epoch": 0.8083437573954495, "grad_norm": 0.4355582594871521, "learning_rate": 9.0539107975241e-05, "loss": 0.0559, "step": 23910 }, { "epoch": 0.8086818350857027, "grad_norm": 0.20173823833465576, "learning_rate": 9.052942720613784e-05, "loss": 0.0543, "step": 23920 }, { "epoch": 0.8090199127759559, "grad_norm": 0.22774651646614075, "learning_rate": 9.051974200479745e-05, "loss": 0.0577, "step": 23930 }, { "epoch": 0.8093579904662092, "grad_norm": 0.2820984721183777, "learning_rate": 9.051005237227901e-05, "loss": 0.0538, "step": 23940 }, { "epoch": 0.8096960681564623, "grad_norm": 0.18553434312343597, "learning_rate": 9.050035830964215e-05, "loss": 0.0577, "step": 23950 }, { "epoch": 0.8100341458467156, "grad_norm": 0.2236989140510559, "learning_rate": 9.049065981794698e-05, "loss": 0.0579, "step": 23960 }, { "epoch": 0.8103722235369688, "grad_norm": 0.4164803922176361, "learning_rate": 9.048095689825414e-05, "loss": 0.0604, "step": 23970 }, { "epoch": 0.810710301227222, "grad_norm": 0.19385960698127747, "learning_rate": 9.047124955162472e-05, "loss": 0.0581, "step": 23980 }, { "epoch": 0.8110483789174753, "grad_norm": 0.3190581500530243, "learning_rate": 9.046153777912028e-05, "loss": 0.0605, "step": 23990 }, { "epoch": 0.8113864566077285, "grad_norm": 0.23022079467773438, "learning_rate": 9.045182158180292e-05, "loss": 0.0646, "step": 24000 }, { "epoch": 0.8117245342979816, "grad_norm": 0.2405940741300583, "learning_rate": 9.044210096073516e-05, "loss": 0.0597, "step": 24010 }, { "epoch": 0.8120626119882349, "grad_norm": 0.24095846712589264, "learning_rate": 9.043237591698004e-05, "loss": 0.0585, "step": 24020 }, { "epoch": 0.8124006896784881, "grad_norm": 0.3069983124732971, "learning_rate": 9.04226464516011e-05, "loss": 0.0613, "step": 24030 }, { "epoch": 0.8127387673687413, "grad_norm": 0.22982637584209442, "learning_rate": 9.041291256566229e-05, "loss": 0.0619, "step": 24040 }, { "epoch": 0.8130768450589946, "grad_norm": 0.23923610150814056, "learning_rate": 9.040317426022814e-05, "loss": 0.058, "step": 24050 }, { "epoch": 0.8134149227492478, "grad_norm": 0.20329128205776215, "learning_rate": 9.03934315363636e-05, "loss": 0.0596, "step": 24060 }, { "epoch": 0.813753000439501, "grad_norm": 0.23459619283676147, "learning_rate": 9.038368439513409e-05, "loss": 0.0568, "step": 24070 }, { "epoch": 0.8140910781297542, "grad_norm": 0.2217312604188919, "learning_rate": 9.03739328376056e-05, "loss": 0.0616, "step": 24080 }, { "epoch": 0.8144291558200074, "grad_norm": 0.28600946068763733, "learning_rate": 9.036417686484451e-05, "loss": 0.0618, "step": 24090 }, { "epoch": 0.8147672335102607, "grad_norm": 0.2642429769039154, "learning_rate": 9.035441647791773e-05, "loss": 0.0581, "step": 24100 }, { "epoch": 0.8151053112005139, "grad_norm": 0.22815769910812378, "learning_rate": 9.034465167789263e-05, "loss": 0.0615, "step": 24110 }, { "epoch": 0.8154433888907671, "grad_norm": 0.26502662897109985, "learning_rate": 9.033488246583706e-05, "loss": 0.0531, "step": 24120 }, { "epoch": 0.8157814665810204, "grad_norm": 0.22851350903511047, "learning_rate": 9.032510884281941e-05, "loss": 0.0629, "step": 24130 }, { "epoch": 0.8161195442712735, "grad_norm": 0.22480477392673492, "learning_rate": 9.031533080990848e-05, "loss": 0.0581, "step": 24140 }, { "epoch": 0.8164576219615267, "grad_norm": 0.2696411907672882, "learning_rate": 9.030554836817358e-05, "loss": 0.0599, "step": 24150 }, { "epoch": 0.81679569965178, "grad_norm": 0.22490862011909485, "learning_rate": 9.029576151868451e-05, "loss": 0.0567, "step": 24160 }, { "epoch": 0.8171337773420332, "grad_norm": 0.24871960282325745, "learning_rate": 9.028597026251155e-05, "loss": 0.0589, "step": 24170 }, { "epoch": 0.8174718550322864, "grad_norm": 0.2811647057533264, "learning_rate": 9.027617460072547e-05, "loss": 0.0556, "step": 24180 }, { "epoch": 0.8178099327225397, "grad_norm": 0.22360220551490784, "learning_rate": 9.026637453439745e-05, "loss": 0.0554, "step": 24190 }, { "epoch": 0.8181480104127928, "grad_norm": 0.3045805096626282, "learning_rate": 9.025657006459927e-05, "loss": 0.0611, "step": 24200 }, { "epoch": 0.8184860881030461, "grad_norm": 0.2598656117916107, "learning_rate": 9.024676119240311e-05, "loss": 0.0595, "step": 24210 }, { "epoch": 0.8188241657932993, "grad_norm": 0.24028801918029785, "learning_rate": 9.023694791888166e-05, "loss": 0.0595, "step": 24220 }, { "epoch": 0.8191622434835525, "grad_norm": 0.2286306917667389, "learning_rate": 9.022713024510808e-05, "loss": 0.0571, "step": 24230 }, { "epoch": 0.8195003211738058, "grad_norm": 0.20822232961654663, "learning_rate": 9.021730817215601e-05, "loss": 0.0606, "step": 24240 }, { "epoch": 0.819838398864059, "grad_norm": 0.41037991642951965, "learning_rate": 9.02074817010996e-05, "loss": 0.0526, "step": 24250 }, { "epoch": 0.8201764765543121, "grad_norm": 0.26084038615226746, "learning_rate": 9.019765083301342e-05, "loss": 0.0575, "step": 24260 }, { "epoch": 0.8205145542445654, "grad_norm": 0.35683301091194153, "learning_rate": 9.01878155689726e-05, "loss": 0.0598, "step": 24270 }, { "epoch": 0.8208526319348186, "grad_norm": 0.2615116834640503, "learning_rate": 9.017797591005268e-05, "loss": 0.0633, "step": 24280 }, { "epoch": 0.8211907096250718, "grad_norm": 0.20001627504825592, "learning_rate": 9.016813185732972e-05, "loss": 0.0596, "step": 24290 }, { "epoch": 0.8215287873153251, "grad_norm": 0.24333858489990234, "learning_rate": 9.015828341188027e-05, "loss": 0.06, "step": 24300 }, { "epoch": 0.8218668650055783, "grad_norm": 0.38437598943710327, "learning_rate": 9.01484305747813e-05, "loss": 0.0672, "step": 24310 }, { "epoch": 0.8222049426958316, "grad_norm": 0.44854649901390076, "learning_rate": 9.013857334711033e-05, "loss": 0.0603, "step": 24320 }, { "epoch": 0.8225430203860847, "grad_norm": 0.25958776473999023, "learning_rate": 9.012871172994534e-05, "loss": 0.0548, "step": 24330 }, { "epoch": 0.8228810980763379, "grad_norm": 0.2251851111650467, "learning_rate": 9.011884572436476e-05, "loss": 0.0564, "step": 24340 }, { "epoch": 0.8232191757665912, "grad_norm": 0.3574685752391815, "learning_rate": 9.010897533144754e-05, "loss": 0.0573, "step": 24350 }, { "epoch": 0.8235572534568444, "grad_norm": 0.3334948420524597, "learning_rate": 9.009910055227306e-05, "loss": 0.0573, "step": 24360 }, { "epoch": 0.8238953311470976, "grad_norm": 0.2946285009384155, "learning_rate": 9.008922138792124e-05, "loss": 0.0593, "step": 24370 }, { "epoch": 0.8242334088373509, "grad_norm": 0.21995748579502106, "learning_rate": 9.007933783947244e-05, "loss": 0.0608, "step": 24380 }, { "epoch": 0.824571486527604, "grad_norm": 0.2898619771003723, "learning_rate": 9.006944990800752e-05, "loss": 0.058, "step": 24390 }, { "epoch": 0.8249095642178572, "grad_norm": 0.30183151364326477, "learning_rate": 9.005955759460779e-05, "loss": 0.0645, "step": 24400 }, { "epoch": 0.8252476419081105, "grad_norm": 0.417275071144104, "learning_rate": 9.004966090035508e-05, "loss": 0.0617, "step": 24410 }, { "epoch": 0.8255857195983637, "grad_norm": 0.3025469481945038, "learning_rate": 9.003975982633166e-05, "loss": 0.0636, "step": 24420 }, { "epoch": 0.825923797288617, "grad_norm": 0.21696969866752625, "learning_rate": 9.00298543736203e-05, "loss": 0.0652, "step": 24430 }, { "epoch": 0.8262618749788702, "grad_norm": 0.19369034469127655, "learning_rate": 9.001994454330427e-05, "loss": 0.0616, "step": 24440 }, { "epoch": 0.8265999526691233, "grad_norm": 0.30254724621772766, "learning_rate": 9.001003033646727e-05, "loss": 0.0657, "step": 24450 }, { "epoch": 0.8269380303593766, "grad_norm": 0.18644587695598602, "learning_rate": 9.00001117541935e-05, "loss": 0.0564, "step": 24460 }, { "epoch": 0.8272761080496298, "grad_norm": 0.24900731444358826, "learning_rate": 8.999018879756764e-05, "loss": 0.058, "step": 24470 }, { "epoch": 0.827614185739883, "grad_norm": 0.21098825335502625, "learning_rate": 8.998026146767487e-05, "loss": 0.0592, "step": 24480 }, { "epoch": 0.8279522634301363, "grad_norm": 0.27870163321495056, "learning_rate": 8.99703297656008e-05, "loss": 0.0602, "step": 24490 }, { "epoch": 0.8282903411203895, "grad_norm": 0.20774973928928375, "learning_rate": 8.996039369243156e-05, "loss": 0.0571, "step": 24500 }, { "epoch": 0.8286284188106426, "grad_norm": 0.25612401962280273, "learning_rate": 8.995045324925378e-05, "loss": 0.0611, "step": 24510 }, { "epoch": 0.8289664965008959, "grad_norm": 0.2679705023765564, "learning_rate": 8.994050843715448e-05, "loss": 0.0554, "step": 24520 }, { "epoch": 0.8293045741911491, "grad_norm": 0.229207381606102, "learning_rate": 8.993055925722121e-05, "loss": 0.0649, "step": 24530 }, { "epoch": 0.8296426518814023, "grad_norm": 0.20834040641784668, "learning_rate": 8.992060571054202e-05, "loss": 0.0549, "step": 24540 }, { "epoch": 0.8299807295716556, "grad_norm": 0.2048489898443222, "learning_rate": 8.991064779820542e-05, "loss": 0.0593, "step": 24550 }, { "epoch": 0.8303188072619088, "grad_norm": 0.4427662193775177, "learning_rate": 8.990068552130036e-05, "loss": 0.0497, "step": 24560 }, { "epoch": 0.8306568849521621, "grad_norm": 0.31621137261390686, "learning_rate": 8.989071888091634e-05, "loss": 0.0551, "step": 24570 }, { "epoch": 0.8309949626424152, "grad_norm": 0.18479108810424805, "learning_rate": 8.988074787814329e-05, "loss": 0.0622, "step": 24580 }, { "epoch": 0.8313330403326684, "grad_norm": 0.2760178744792938, "learning_rate": 8.987077251407158e-05, "loss": 0.0562, "step": 24590 }, { "epoch": 0.8316711180229217, "grad_norm": 0.23410017788410187, "learning_rate": 8.986079278979216e-05, "loss": 0.0547, "step": 24600 }, { "epoch": 0.8320091957131749, "grad_norm": 0.23590916395187378, "learning_rate": 8.985080870639635e-05, "loss": 0.0586, "step": 24610 }, { "epoch": 0.8323472734034281, "grad_norm": 0.19442017376422882, "learning_rate": 8.984082026497603e-05, "loss": 0.0632, "step": 24620 }, { "epoch": 0.8326853510936814, "grad_norm": 0.2517504096031189, "learning_rate": 8.98308274666235e-05, "loss": 0.0592, "step": 24630 }, { "epoch": 0.8330234287839345, "grad_norm": 0.21778184175491333, "learning_rate": 8.982083031243155e-05, "loss": 0.0549, "step": 24640 }, { "epoch": 0.8333615064741877, "grad_norm": 0.31795957684516907, "learning_rate": 8.98108288034935e-05, "loss": 0.0599, "step": 24650 }, { "epoch": 0.833699584164441, "grad_norm": 0.21692270040512085, "learning_rate": 8.980082294090305e-05, "loss": 0.056, "step": 24660 }, { "epoch": 0.8340376618546942, "grad_norm": 0.22988204658031464, "learning_rate": 8.979081272575443e-05, "loss": 0.0609, "step": 24670 }, { "epoch": 0.8343757395449475, "grad_norm": 0.2124883234500885, "learning_rate": 8.978079815914236e-05, "loss": 0.0606, "step": 24680 }, { "epoch": 0.8347138172352007, "grad_norm": 0.2510266900062561, "learning_rate": 8.977077924216202e-05, "loss": 0.0549, "step": 24690 }, { "epoch": 0.8350518949254538, "grad_norm": 0.2713707387447357, "learning_rate": 8.976075597590905e-05, "loss": 0.0548, "step": 24700 }, { "epoch": 0.8353899726157071, "grad_norm": 0.2987319529056549, "learning_rate": 8.975072836147958e-05, "loss": 0.0614, "step": 24710 }, { "epoch": 0.8357280503059603, "grad_norm": 0.3595227599143982, "learning_rate": 8.974069639997025e-05, "loss": 0.0552, "step": 24720 }, { "epoch": 0.8360661279962135, "grad_norm": 0.20386934280395508, "learning_rate": 8.973066009247808e-05, "loss": 0.053, "step": 24730 }, { "epoch": 0.8364042056864668, "grad_norm": 0.18996590375900269, "learning_rate": 8.972061944010066e-05, "loss": 0.0613, "step": 24740 }, { "epoch": 0.83674228337672, "grad_norm": 0.40496689081192017, "learning_rate": 8.971057444393603e-05, "loss": 0.0568, "step": 24750 }, { "epoch": 0.8370803610669731, "grad_norm": 0.19821958243846893, "learning_rate": 8.970052510508268e-05, "loss": 0.057, "step": 24760 }, { "epoch": 0.8374184387572264, "grad_norm": 0.17627297341823578, "learning_rate": 8.969047142463959e-05, "loss": 0.0595, "step": 24770 }, { "epoch": 0.8377565164474796, "grad_norm": 0.363459974527359, "learning_rate": 8.968041340370621e-05, "loss": 0.058, "step": 24780 }, { "epoch": 0.8380945941377328, "grad_norm": 0.2656196653842926, "learning_rate": 8.96703510433825e-05, "loss": 0.0661, "step": 24790 }, { "epoch": 0.8384326718279861, "grad_norm": 0.30022433400154114, "learning_rate": 8.966028434476883e-05, "loss": 0.0596, "step": 24800 }, { "epoch": 0.8387707495182393, "grad_norm": 0.2902466654777527, "learning_rate": 8.96502133089661e-05, "loss": 0.0615, "step": 24810 }, { "epoch": 0.8391088272084926, "grad_norm": 0.2631717026233673, "learning_rate": 8.964013793707564e-05, "loss": 0.0583, "step": 24820 }, { "epoch": 0.8394469048987457, "grad_norm": 0.30447596311569214, "learning_rate": 8.963005823019932e-05, "loss": 0.0589, "step": 24830 }, { "epoch": 0.8397849825889989, "grad_norm": 0.34884876012802124, "learning_rate": 8.961997418943939e-05, "loss": 0.0549, "step": 24840 }, { "epoch": 0.8401230602792522, "grad_norm": 0.3149174749851227, "learning_rate": 8.960988581589865e-05, "loss": 0.0564, "step": 24850 }, { "epoch": 0.8404611379695054, "grad_norm": 0.40912094712257385, "learning_rate": 8.959979311068037e-05, "loss": 0.0701, "step": 24860 }, { "epoch": 0.8407992156597586, "grad_norm": 0.2815586030483246, "learning_rate": 8.958969607488823e-05, "loss": 0.056, "step": 24870 }, { "epoch": 0.8411372933500119, "grad_norm": 0.24713921546936035, "learning_rate": 8.957959470962647e-05, "loss": 0.0616, "step": 24880 }, { "epoch": 0.841475371040265, "grad_norm": 0.2640742361545563, "learning_rate": 8.956948901599971e-05, "loss": 0.0644, "step": 24890 }, { "epoch": 0.8418134487305182, "grad_norm": 0.2830218970775604, "learning_rate": 8.955937899511315e-05, "loss": 0.0631, "step": 24900 }, { "epoch": 0.8421515264207715, "grad_norm": 0.3403528332710266, "learning_rate": 8.954926464807238e-05, "loss": 0.0558, "step": 24910 }, { "epoch": 0.8424896041110247, "grad_norm": 0.2133941650390625, "learning_rate": 8.953914597598347e-05, "loss": 0.0538, "step": 24920 }, { "epoch": 0.842827681801278, "grad_norm": 0.24637746810913086, "learning_rate": 8.952902297995303e-05, "loss": 0.061, "step": 24930 }, { "epoch": 0.8431657594915312, "grad_norm": 0.3004252314567566, "learning_rate": 8.951889566108804e-05, "loss": 0.0543, "step": 24940 }, { "epoch": 0.8435038371817843, "grad_norm": 0.2649867832660675, "learning_rate": 8.950876402049606e-05, "loss": 0.0554, "step": 24950 }, { "epoch": 0.8438419148720376, "grad_norm": 0.4240155518054962, "learning_rate": 8.949862805928504e-05, "loss": 0.0616, "step": 24960 }, { "epoch": 0.8441799925622908, "grad_norm": 0.18801982700824738, "learning_rate": 8.948848777856343e-05, "loss": 0.057, "step": 24970 }, { "epoch": 0.844518070252544, "grad_norm": 0.1859777718782425, "learning_rate": 8.947834317944017e-05, "loss": 0.0528, "step": 24980 }, { "epoch": 0.8448561479427973, "grad_norm": 0.2757047414779663, "learning_rate": 8.946819426302466e-05, "loss": 0.0628, "step": 24990 }, { "epoch": 0.8451942256330505, "grad_norm": 0.5902528762817383, "learning_rate": 8.945804103042676e-05, "loss": 0.0608, "step": 25000 }, { "epoch": 0.8455323033233036, "grad_norm": 0.24561384320259094, "learning_rate": 8.944788348275681e-05, "loss": 0.0576, "step": 25010 }, { "epoch": 0.8458703810135569, "grad_norm": 0.25641676783561707, "learning_rate": 8.943772162112565e-05, "loss": 0.0575, "step": 25020 }, { "epoch": 0.8462084587038101, "grad_norm": 0.2568347454071045, "learning_rate": 8.942755544664454e-05, "loss": 0.0566, "step": 25030 }, { "epoch": 0.8465465363940634, "grad_norm": 0.3726033568382263, "learning_rate": 8.941738496042525e-05, "loss": 0.0596, "step": 25040 }, { "epoch": 0.8468846140843166, "grad_norm": 0.23178139328956604, "learning_rate": 8.940721016357999e-05, "loss": 0.0622, "step": 25050 }, { "epoch": 0.8472226917745698, "grad_norm": 0.23093923926353455, "learning_rate": 8.939703105722148e-05, "loss": 0.0563, "step": 25060 }, { "epoch": 0.8475607694648231, "grad_norm": 0.22096994519233704, "learning_rate": 8.93868476424629e-05, "loss": 0.0564, "step": 25070 }, { "epoch": 0.8478988471550762, "grad_norm": 0.4677508771419525, "learning_rate": 8.937665992041786e-05, "loss": 0.0637, "step": 25080 }, { "epoch": 0.8482369248453294, "grad_norm": 0.18464241921901703, "learning_rate": 8.93664678922005e-05, "loss": 0.0513, "step": 25090 }, { "epoch": 0.8485750025355827, "grad_norm": 0.22569121420383453, "learning_rate": 8.93562715589254e-05, "loss": 0.0567, "step": 25100 }, { "epoch": 0.8489130802258359, "grad_norm": 0.3434806168079376, "learning_rate": 8.934607092170762e-05, "loss": 0.0574, "step": 25110 }, { "epoch": 0.8492511579160891, "grad_norm": 0.24460795521736145, "learning_rate": 8.933586598166266e-05, "loss": 0.0545, "step": 25120 }, { "epoch": 0.8495892356063424, "grad_norm": 0.3126770257949829, "learning_rate": 8.932565673990655e-05, "loss": 0.0595, "step": 25130 }, { "epoch": 0.8499273132965955, "grad_norm": 0.24801142513751984, "learning_rate": 8.931544319755574e-05, "loss": 0.0608, "step": 25140 }, { "epoch": 0.8502653909868487, "grad_norm": 0.25230872631073, "learning_rate": 8.930522535572718e-05, "loss": 0.0571, "step": 25150 }, { "epoch": 0.850603468677102, "grad_norm": 0.23323071002960205, "learning_rate": 8.929500321553826e-05, "loss": 0.0595, "step": 25160 }, { "epoch": 0.8509415463673552, "grad_norm": 0.18415845930576324, "learning_rate": 8.928477677810686e-05, "loss": 0.0548, "step": 25170 }, { "epoch": 0.8512796240576085, "grad_norm": 0.2614712417125702, "learning_rate": 8.927454604455137e-05, "loss": 0.0594, "step": 25180 }, { "epoch": 0.8516177017478617, "grad_norm": 0.34500065445899963, "learning_rate": 8.926431101599053e-05, "loss": 0.0552, "step": 25190 }, { "epoch": 0.8519557794381148, "grad_norm": 0.29195842146873474, "learning_rate": 8.925407169354369e-05, "loss": 0.0582, "step": 25200 }, { "epoch": 0.8522938571283681, "grad_norm": 0.3395450711250305, "learning_rate": 8.92438280783306e-05, "loss": 0.0545, "step": 25210 }, { "epoch": 0.8526319348186213, "grad_norm": 0.20834259688854218, "learning_rate": 8.923358017147146e-05, "loss": 0.0568, "step": 25220 }, { "epoch": 0.8529700125088745, "grad_norm": 0.22557154297828674, "learning_rate": 8.922332797408697e-05, "loss": 0.0579, "step": 25230 }, { "epoch": 0.8533080901991278, "grad_norm": 0.2915285527706146, "learning_rate": 8.921307148729831e-05, "loss": 0.06, "step": 25240 }, { "epoch": 0.853646167889381, "grad_norm": 0.27937451004981995, "learning_rate": 8.920281071222712e-05, "loss": 0.0646, "step": 25250 }, { "epoch": 0.8539842455796341, "grad_norm": 0.24632489681243896, "learning_rate": 8.919254564999548e-05, "loss": 0.0614, "step": 25260 }, { "epoch": 0.8543223232698874, "grad_norm": 0.21923351287841797, "learning_rate": 8.918227630172598e-05, "loss": 0.0602, "step": 25270 }, { "epoch": 0.8546604009601406, "grad_norm": 0.2202175408601761, "learning_rate": 8.917200266854165e-05, "loss": 0.0601, "step": 25280 }, { "epoch": 0.8549984786503939, "grad_norm": 0.2225247174501419, "learning_rate": 8.9161724751566e-05, "loss": 0.062, "step": 25290 }, { "epoch": 0.8553365563406471, "grad_norm": 0.2857019305229187, "learning_rate": 8.915144255192302e-05, "loss": 0.062, "step": 25300 }, { "epoch": 0.8556746340309003, "grad_norm": 0.36922726035118103, "learning_rate": 8.914115607073714e-05, "loss": 0.0548, "step": 25310 }, { "epoch": 0.8560127117211536, "grad_norm": 0.20323359966278076, "learning_rate": 8.913086530913327e-05, "loss": 0.0581, "step": 25320 }, { "epoch": 0.8563507894114067, "grad_norm": 0.21260569989681244, "learning_rate": 8.912057026823681e-05, "loss": 0.0605, "step": 25330 }, { "epoch": 0.8566888671016599, "grad_norm": 0.40166786313056946, "learning_rate": 8.91102709491736e-05, "loss": 0.0577, "step": 25340 }, { "epoch": 0.8570269447919132, "grad_norm": 0.42376968264579773, "learning_rate": 8.909996735306996e-05, "loss": 0.0636, "step": 25350 }, { "epoch": 0.8573650224821664, "grad_norm": 0.33262625336647034, "learning_rate": 8.908965948105268e-05, "loss": 0.055, "step": 25360 }, { "epoch": 0.8577031001724196, "grad_norm": 0.35860297083854675, "learning_rate": 8.907934733424901e-05, "loss": 0.0601, "step": 25370 }, { "epoch": 0.8580411778626729, "grad_norm": 0.29828906059265137, "learning_rate": 8.906903091378666e-05, "loss": 0.0592, "step": 25380 }, { "epoch": 0.858379255552926, "grad_norm": 0.17574773728847504, "learning_rate": 8.905871022079384e-05, "loss": 0.058, "step": 25390 }, { "epoch": 0.8587173332431793, "grad_norm": 0.1503233164548874, "learning_rate": 8.90483852563992e-05, "loss": 0.0535, "step": 25400 }, { "epoch": 0.8590554109334325, "grad_norm": 0.23033197224140167, "learning_rate": 8.903805602173185e-05, "loss": 0.0667, "step": 25410 }, { "epoch": 0.8593934886236857, "grad_norm": 0.25999370217323303, "learning_rate": 8.902772251792137e-05, "loss": 0.0603, "step": 25420 }, { "epoch": 0.859731566313939, "grad_norm": 0.19054023921489716, "learning_rate": 8.901738474609786e-05, "loss": 0.0617, "step": 25430 }, { "epoch": 0.8600696440041922, "grad_norm": 0.20691853761672974, "learning_rate": 8.900704270739179e-05, "loss": 0.054, "step": 25440 }, { "epoch": 0.8604077216944453, "grad_norm": 0.4458543658256531, "learning_rate": 8.89966964029342e-05, "loss": 0.0548, "step": 25450 }, { "epoch": 0.8607457993846986, "grad_norm": 0.2812923789024353, "learning_rate": 8.898634583385652e-05, "loss": 0.0626, "step": 25460 }, { "epoch": 0.8610838770749518, "grad_norm": 0.23285417258739471, "learning_rate": 8.897599100129065e-05, "loss": 0.0602, "step": 25470 }, { "epoch": 0.861421954765205, "grad_norm": 0.22374704480171204, "learning_rate": 8.896563190636903e-05, "loss": 0.0644, "step": 25480 }, { "epoch": 0.8617600324554583, "grad_norm": 0.25884902477264404, "learning_rate": 8.895526855022448e-05, "loss": 0.0587, "step": 25490 }, { "epoch": 0.8620981101457115, "grad_norm": 0.3120608925819397, "learning_rate": 8.894490093399033e-05, "loss": 0.0578, "step": 25500 }, { "epoch": 0.8624361878359647, "grad_norm": 0.23969155550003052, "learning_rate": 8.893452905880035e-05, "loss": 0.0569, "step": 25510 }, { "epoch": 0.8627742655262179, "grad_norm": 0.26977482438087463, "learning_rate": 8.892415292578883e-05, "loss": 0.0605, "step": 25520 }, { "epoch": 0.8631123432164711, "grad_norm": 0.2683321237564087, "learning_rate": 8.891377253609046e-05, "loss": 0.0603, "step": 25530 }, { "epoch": 0.8634504209067244, "grad_norm": 0.20962035655975342, "learning_rate": 8.890338789084043e-05, "loss": 0.057, "step": 25540 }, { "epoch": 0.8637884985969776, "grad_norm": 0.19900467991828918, "learning_rate": 8.88929989911744e-05, "loss": 0.0608, "step": 25550 }, { "epoch": 0.8641265762872308, "grad_norm": 0.24831214547157288, "learning_rate": 8.888260583822847e-05, "loss": 0.0565, "step": 25560 }, { "epoch": 0.8644646539774841, "grad_norm": 0.509679913520813, "learning_rate": 8.887220843313921e-05, "loss": 0.0572, "step": 25570 }, { "epoch": 0.8648027316677372, "grad_norm": 0.24649013578891754, "learning_rate": 8.88618067770437e-05, "loss": 0.0536, "step": 25580 }, { "epoch": 0.8651408093579904, "grad_norm": 0.30858683586120605, "learning_rate": 8.885140087107942e-05, "loss": 0.0567, "step": 25590 }, { "epoch": 0.8654788870482437, "grad_norm": 0.41389966011047363, "learning_rate": 8.884099071638436e-05, "loss": 0.0523, "step": 25600 }, { "epoch": 0.8658169647384969, "grad_norm": 0.21455831825733185, "learning_rate": 8.883057631409695e-05, "loss": 0.0498, "step": 25610 }, { "epoch": 0.8661550424287501, "grad_norm": 0.42179468274116516, "learning_rate": 8.882015766535608e-05, "loss": 0.0574, "step": 25620 }, { "epoch": 0.8664931201190034, "grad_norm": 0.23379085958003998, "learning_rate": 8.880973477130115e-05, "loss": 0.0599, "step": 25630 }, { "epoch": 0.8668311978092565, "grad_norm": 0.24607986211776733, "learning_rate": 8.879930763307197e-05, "loss": 0.0568, "step": 25640 }, { "epoch": 0.8671692754995098, "grad_norm": 0.24855440855026245, "learning_rate": 8.878887625180884e-05, "loss": 0.0511, "step": 25650 }, { "epoch": 0.867507353189763, "grad_norm": 0.36046159267425537, "learning_rate": 8.877844062865253e-05, "loss": 0.0565, "step": 25660 }, { "epoch": 0.8678454308800162, "grad_norm": 0.23500677943229675, "learning_rate": 8.876800076474424e-05, "loss": 0.057, "step": 25670 }, { "epoch": 0.8681835085702695, "grad_norm": 0.28343653678894043, "learning_rate": 8.875755666122568e-05, "loss": 0.0606, "step": 25680 }, { "epoch": 0.8685215862605227, "grad_norm": 0.2538602650165558, "learning_rate": 8.8747108319239e-05, "loss": 0.0555, "step": 25690 }, { "epoch": 0.8688596639507759, "grad_norm": 0.3285179138183594, "learning_rate": 8.87366557399268e-05, "loss": 0.0645, "step": 25700 }, { "epoch": 0.8691977416410291, "grad_norm": 0.26401033997535706, "learning_rate": 8.872619892443217e-05, "loss": 0.0571, "step": 25710 }, { "epoch": 0.8695358193312823, "grad_norm": 0.3511100113391876, "learning_rate": 8.871573787389865e-05, "loss": 0.0573, "step": 25720 }, { "epoch": 0.8698738970215355, "grad_norm": 0.21448132395744324, "learning_rate": 8.870527258947024e-05, "loss": 0.0565, "step": 25730 }, { "epoch": 0.8702119747117888, "grad_norm": 0.18653346598148346, "learning_rate": 8.869480307229143e-05, "loss": 0.0609, "step": 25740 }, { "epoch": 0.870550052402042, "grad_norm": 0.20377510786056519, "learning_rate": 8.868432932350712e-05, "loss": 0.0542, "step": 25750 }, { "epoch": 0.8708881300922952, "grad_norm": 0.18894577026367188, "learning_rate": 8.867385134426272e-05, "loss": 0.0577, "step": 25760 }, { "epoch": 0.8712262077825484, "grad_norm": 0.2278454303741455, "learning_rate": 8.866336913570407e-05, "loss": 0.0622, "step": 25770 }, { "epoch": 0.8715642854728016, "grad_norm": 0.30225691199302673, "learning_rate": 8.865288269897751e-05, "loss": 0.0564, "step": 25780 }, { "epoch": 0.8719023631630549, "grad_norm": 0.21969005465507507, "learning_rate": 8.864239203522981e-05, "loss": 0.062, "step": 25790 }, { "epoch": 0.8722404408533081, "grad_norm": 0.28337156772613525, "learning_rate": 8.863189714560822e-05, "loss": 0.0563, "step": 25800 }, { "epoch": 0.8725785185435613, "grad_norm": 0.18123555183410645, "learning_rate": 8.862139803126043e-05, "loss": 0.0603, "step": 25810 }, { "epoch": 0.8729165962338146, "grad_norm": 0.2479667365550995, "learning_rate": 8.861089469333463e-05, "loss": 0.0544, "step": 25820 }, { "epoch": 0.8732546739240677, "grad_norm": 0.35426759719848633, "learning_rate": 8.860038713297944e-05, "loss": 0.0574, "step": 25830 }, { "epoch": 0.8735927516143209, "grad_norm": 0.2116418480873108, "learning_rate": 8.858987535134394e-05, "loss": 0.0559, "step": 25840 }, { "epoch": 0.8739308293045742, "grad_norm": 0.24792702496051788, "learning_rate": 8.857935934957769e-05, "loss": 0.0557, "step": 25850 }, { "epoch": 0.8742689069948274, "grad_norm": 0.26624831557273865, "learning_rate": 8.856883912883071e-05, "loss": 0.0537, "step": 25860 }, { "epoch": 0.8746069846850806, "grad_norm": 0.21702136099338531, "learning_rate": 8.855831469025346e-05, "loss": 0.0563, "step": 25870 }, { "epoch": 0.8749450623753339, "grad_norm": 0.21764768660068512, "learning_rate": 8.854778603499689e-05, "loss": 0.0629, "step": 25880 }, { "epoch": 0.875283140065587, "grad_norm": 0.3460659384727478, "learning_rate": 8.85372531642124e-05, "loss": 0.0519, "step": 25890 }, { "epoch": 0.8756212177558403, "grad_norm": 0.21284529566764832, "learning_rate": 8.852671607905185e-05, "loss": 0.057, "step": 25900 }, { "epoch": 0.8759592954460935, "grad_norm": 0.2407177835702896, "learning_rate": 8.851617478066754e-05, "loss": 0.0597, "step": 25910 }, { "epoch": 0.8762973731363467, "grad_norm": 0.20006735622882843, "learning_rate": 8.850562927021227e-05, "loss": 0.0614, "step": 25920 }, { "epoch": 0.8766354508266, "grad_norm": 0.20730161666870117, "learning_rate": 8.849507954883928e-05, "loss": 0.0601, "step": 25930 }, { "epoch": 0.8769735285168532, "grad_norm": 0.1814379245042801, "learning_rate": 8.848452561770226e-05, "loss": 0.0574, "step": 25940 }, { "epoch": 0.8773116062071064, "grad_norm": 0.23188120126724243, "learning_rate": 8.847396747795538e-05, "loss": 0.0597, "step": 25950 }, { "epoch": 0.8776496838973596, "grad_norm": 0.25159624218940735, "learning_rate": 8.846340513075327e-05, "loss": 0.0599, "step": 25960 }, { "epoch": 0.8779877615876128, "grad_norm": 0.325583815574646, "learning_rate": 8.845283857725099e-05, "loss": 0.0571, "step": 25970 }, { "epoch": 0.878325839277866, "grad_norm": 0.2555578351020813, "learning_rate": 8.844226781860409e-05, "loss": 0.0548, "step": 25980 }, { "epoch": 0.8786639169681193, "grad_norm": 0.38468918204307556, "learning_rate": 8.84316928559686e-05, "loss": 0.0624, "step": 25990 }, { "epoch": 0.8790019946583725, "grad_norm": 0.2102966606616974, "learning_rate": 8.842111369050094e-05, "loss": 0.0559, "step": 26000 }, { "epoch": 0.8793400723486258, "grad_norm": 0.3016993999481201, "learning_rate": 8.841053032335808e-05, "loss": 0.0582, "step": 26010 }, { "epoch": 0.879678150038879, "grad_norm": 1.359771966934204, "learning_rate": 8.839994275569735e-05, "loss": 0.0593, "step": 26020 }, { "epoch": 0.8800162277291321, "grad_norm": 0.49327969551086426, "learning_rate": 8.838935098867662e-05, "loss": 0.0593, "step": 26030 }, { "epoch": 0.8803543054193854, "grad_norm": 0.1844254434108734, "learning_rate": 8.837875502345418e-05, "loss": 0.0562, "step": 26040 }, { "epoch": 0.8806923831096386, "grad_norm": 0.193365678191185, "learning_rate": 8.83681548611888e-05, "loss": 0.0574, "step": 26050 }, { "epoch": 0.8810304607998918, "grad_norm": 0.24533413350582123, "learning_rate": 8.835755050303969e-05, "loss": 0.0597, "step": 26060 }, { "epoch": 0.8813685384901451, "grad_norm": 0.21705596148967743, "learning_rate": 8.834694195016653e-05, "loss": 0.0567, "step": 26070 }, { "epoch": 0.8817066161803983, "grad_norm": 0.1665138602256775, "learning_rate": 8.833632920372942e-05, "loss": 0.0547, "step": 26080 }, { "epoch": 0.8820446938706514, "grad_norm": 0.9872521162033081, "learning_rate": 8.832571226488903e-05, "loss": 0.0574, "step": 26090 }, { "epoch": 0.8823827715609047, "grad_norm": 0.22430163621902466, "learning_rate": 8.831509113480634e-05, "loss": 0.0543, "step": 26100 }, { "epoch": 0.8827208492511579, "grad_norm": 0.20293429493904114, "learning_rate": 8.83044658146429e-05, "loss": 0.0571, "step": 26110 }, { "epoch": 0.8830589269414111, "grad_norm": 0.36389273405075073, "learning_rate": 8.829383630556067e-05, "loss": 0.0547, "step": 26120 }, { "epoch": 0.8833970046316644, "grad_norm": 0.2158312350511551, "learning_rate": 8.828320260872207e-05, "loss": 0.055, "step": 26130 }, { "epoch": 0.8837350823219176, "grad_norm": 0.19819419085979462, "learning_rate": 8.827256472529e-05, "loss": 0.0555, "step": 26140 }, { "epoch": 0.8840731600121708, "grad_norm": 0.26256057620048523, "learning_rate": 8.826192265642778e-05, "loss": 0.0667, "step": 26150 }, { "epoch": 0.884411237702424, "grad_norm": 0.2551274001598358, "learning_rate": 8.825127640329923e-05, "loss": 0.059, "step": 26160 }, { "epoch": 0.8847493153926772, "grad_norm": 0.22327610850334167, "learning_rate": 8.824062596706861e-05, "loss": 0.0636, "step": 26170 }, { "epoch": 0.8850873930829305, "grad_norm": 0.2414255291223526, "learning_rate": 8.822997134890062e-05, "loss": 0.0605, "step": 26180 }, { "epoch": 0.8854254707731837, "grad_norm": 0.33282235264778137, "learning_rate": 8.821931254996044e-05, "loss": 0.0564, "step": 26190 }, { "epoch": 0.8857635484634369, "grad_norm": 0.23793409764766693, "learning_rate": 8.82086495714137e-05, "loss": 0.0628, "step": 26200 }, { "epoch": 0.8861016261536901, "grad_norm": 0.17363187670707703, "learning_rate": 8.81979824144265e-05, "loss": 0.058, "step": 26210 }, { "epoch": 0.8864397038439433, "grad_norm": 0.22150495648384094, "learning_rate": 8.818731108016536e-05, "loss": 0.0556, "step": 26220 }, { "epoch": 0.8867777815341965, "grad_norm": 0.3359498679637909, "learning_rate": 8.81766355697973e-05, "loss": 0.0618, "step": 26230 }, { "epoch": 0.8871158592244498, "grad_norm": 0.4445522129535675, "learning_rate": 8.816595588448977e-05, "loss": 0.056, "step": 26240 }, { "epoch": 0.887453936914703, "grad_norm": 0.4556458294391632, "learning_rate": 8.81552720254107e-05, "loss": 0.0571, "step": 26250 }, { "epoch": 0.8877920146049563, "grad_norm": 0.43448710441589355, "learning_rate": 8.814458399372842e-05, "loss": 0.0667, "step": 26260 }, { "epoch": 0.8881300922952095, "grad_norm": 0.29349538683891296, "learning_rate": 8.813389179061181e-05, "loss": 0.0584, "step": 26270 }, { "epoch": 0.8884681699854626, "grad_norm": 0.31874021887779236, "learning_rate": 8.812319541723012e-05, "loss": 0.0612, "step": 26280 }, { "epoch": 0.8888062476757159, "grad_norm": 0.23099233210086823, "learning_rate": 8.811249487475309e-05, "loss": 0.0563, "step": 26290 }, { "epoch": 0.8891443253659691, "grad_norm": 0.22423464059829712, "learning_rate": 8.810179016435092e-05, "loss": 0.0561, "step": 26300 }, { "epoch": 0.8894824030562223, "grad_norm": 0.19755107164382935, "learning_rate": 8.809108128719428e-05, "loss": 0.0696, "step": 26310 }, { "epoch": 0.8898204807464756, "grad_norm": 0.20826660096645355, "learning_rate": 8.808036824445424e-05, "loss": 0.0648, "step": 26320 }, { "epoch": 0.8901585584367288, "grad_norm": 0.5505613088607788, "learning_rate": 8.806965103730238e-05, "loss": 0.0577, "step": 26330 }, { "epoch": 0.8904966361269819, "grad_norm": 0.9967947602272034, "learning_rate": 8.805892966691074e-05, "loss": 0.0606, "step": 26340 }, { "epoch": 0.8908347138172352, "grad_norm": 0.19844265282154083, "learning_rate": 8.804820413445175e-05, "loss": 0.0598, "step": 26350 }, { "epoch": 0.8911727915074884, "grad_norm": 0.2937750518321991, "learning_rate": 8.803747444109837e-05, "loss": 0.0535, "step": 26360 }, { "epoch": 0.8915108691977417, "grad_norm": 0.2892059087753296, "learning_rate": 8.802674058802399e-05, "loss": 0.0584, "step": 26370 }, { "epoch": 0.8918489468879949, "grad_norm": 0.2804025709629059, "learning_rate": 8.801600257640241e-05, "loss": 0.0573, "step": 26380 }, { "epoch": 0.8921870245782481, "grad_norm": 0.3490004539489746, "learning_rate": 8.800526040740795e-05, "loss": 0.063, "step": 26390 }, { "epoch": 0.8925251022685013, "grad_norm": 0.29368090629577637, "learning_rate": 8.799451408221535e-05, "loss": 0.063, "step": 26400 }, { "epoch": 0.8928631799587545, "grad_norm": 0.17443183064460754, "learning_rate": 8.798376360199982e-05, "loss": 0.057, "step": 26410 }, { "epoch": 0.8932012576490077, "grad_norm": 0.2344209849834442, "learning_rate": 8.797300896793701e-05, "loss": 0.0587, "step": 26420 }, { "epoch": 0.893539335339261, "grad_norm": 0.23575511574745178, "learning_rate": 8.796225018120302e-05, "loss": 0.0585, "step": 26430 }, { "epoch": 0.8938774130295142, "grad_norm": 0.25182458758354187, "learning_rate": 8.795148724297444e-05, "loss": 0.0615, "step": 26440 }, { "epoch": 0.8942154907197674, "grad_norm": 0.13556568324565887, "learning_rate": 8.794072015442825e-05, "loss": 0.0533, "step": 26450 }, { "epoch": 0.8945535684100206, "grad_norm": 0.3440661132335663, "learning_rate": 8.792994891674198e-05, "loss": 0.0625, "step": 26460 }, { "epoch": 0.8948916461002738, "grad_norm": 0.19319893419742584, "learning_rate": 8.79191735310935e-05, "loss": 0.061, "step": 26470 }, { "epoch": 0.895229723790527, "grad_norm": 0.24853551387786865, "learning_rate": 8.790839399866122e-05, "loss": 0.0554, "step": 26480 }, { "epoch": 0.8955678014807803, "grad_norm": 0.2938297986984253, "learning_rate": 8.789761032062397e-05, "loss": 0.0598, "step": 26490 }, { "epoch": 0.8959058791710335, "grad_norm": 0.2816154360771179, "learning_rate": 8.788682249816103e-05, "loss": 0.0544, "step": 26500 }, { "epoch": 0.8962439568612868, "grad_norm": 0.20757001638412476, "learning_rate": 8.787603053245215e-05, "loss": 0.0617, "step": 26510 }, { "epoch": 0.89658203455154, "grad_norm": 0.29476043581962585, "learning_rate": 8.78652344246775e-05, "loss": 0.0542, "step": 26520 }, { "epoch": 0.8969201122417931, "grad_norm": 0.21815434098243713, "learning_rate": 8.785443417601776e-05, "loss": 0.055, "step": 26530 }, { "epoch": 0.8972581899320464, "grad_norm": 0.24317479133605957, "learning_rate": 8.784362978765401e-05, "loss": 0.0609, "step": 26540 }, { "epoch": 0.8975962676222996, "grad_norm": 0.3272753357887268, "learning_rate": 8.783282126076779e-05, "loss": 0.0599, "step": 26550 }, { "epoch": 0.8979343453125528, "grad_norm": 0.3300723433494568, "learning_rate": 8.782200859654112e-05, "loss": 0.055, "step": 26560 }, { "epoch": 0.8982724230028061, "grad_norm": 0.2320994734764099, "learning_rate": 8.781119179615646e-05, "loss": 0.0574, "step": 26570 }, { "epoch": 0.8986105006930593, "grad_norm": 0.4767119586467743, "learning_rate": 8.780037086079674e-05, "loss": 0.0603, "step": 26580 }, { "epoch": 0.8989485783833124, "grad_norm": 0.2684296667575836, "learning_rate": 8.778954579164527e-05, "loss": 0.0556, "step": 26590 }, { "epoch": 0.8992866560735657, "grad_norm": 0.14779484272003174, "learning_rate": 8.777871658988588e-05, "loss": 0.0562, "step": 26600 }, { "epoch": 0.8996247337638189, "grad_norm": 0.23269431293010712, "learning_rate": 8.776788325670285e-05, "loss": 0.0606, "step": 26610 }, { "epoch": 0.8999628114540722, "grad_norm": 0.297000527381897, "learning_rate": 8.775704579328089e-05, "loss": 0.0589, "step": 26620 }, { "epoch": 0.9003008891443254, "grad_norm": 0.2957291305065155, "learning_rate": 8.774620420080517e-05, "loss": 0.0564, "step": 26630 }, { "epoch": 0.9006389668345786, "grad_norm": 0.3207296133041382, "learning_rate": 8.773535848046131e-05, "loss": 0.0559, "step": 26640 }, { "epoch": 0.9009770445248318, "grad_norm": 0.38556432723999023, "learning_rate": 8.772450863343538e-05, "loss": 0.0566, "step": 26650 }, { "epoch": 0.901315122215085, "grad_norm": 0.15317858755588531, "learning_rate": 8.77136546609139e-05, "loss": 0.0542, "step": 26660 }, { "epoch": 0.9016531999053382, "grad_norm": 0.28898099064826965, "learning_rate": 8.770279656408385e-05, "loss": 0.0546, "step": 26670 }, { "epoch": 0.9019912775955915, "grad_norm": 0.3930251896381378, "learning_rate": 8.769193434413265e-05, "loss": 0.0586, "step": 26680 }, { "epoch": 0.9023293552858447, "grad_norm": 0.18833288550376892, "learning_rate": 8.76810680022482e-05, "loss": 0.0543, "step": 26690 }, { "epoch": 0.9026674329760979, "grad_norm": 0.19344255328178406, "learning_rate": 8.767019753961878e-05, "loss": 0.0592, "step": 26700 }, { "epoch": 0.9030055106663512, "grad_norm": 0.2935497760772705, "learning_rate": 8.765932295743321e-05, "loss": 0.0515, "step": 26710 }, { "epoch": 0.9033435883566043, "grad_norm": 0.21191908419132233, "learning_rate": 8.764844425688068e-05, "loss": 0.0632, "step": 26720 }, { "epoch": 0.9036816660468575, "grad_norm": 0.27752938866615295, "learning_rate": 8.763756143915092e-05, "loss": 0.0593, "step": 26730 }, { "epoch": 0.9040197437371108, "grad_norm": 0.2078903764486313, "learning_rate": 8.7626674505434e-05, "loss": 0.054, "step": 26740 }, { "epoch": 0.904357821427364, "grad_norm": 0.2998602092266083, "learning_rate": 8.761578345692053e-05, "loss": 0.0598, "step": 26750 }, { "epoch": 0.9046958991176173, "grad_norm": 0.6614800691604614, "learning_rate": 8.760488829480156e-05, "loss": 0.0609, "step": 26760 }, { "epoch": 0.9050339768078705, "grad_norm": 0.22993753850460052, "learning_rate": 8.759398902026854e-05, "loss": 0.0563, "step": 26770 }, { "epoch": 0.9053720544981236, "grad_norm": 0.20592202246189117, "learning_rate": 8.758308563451339e-05, "loss": 0.0536, "step": 26780 }, { "epoch": 0.9057101321883769, "grad_norm": 0.21883781254291534, "learning_rate": 8.75721781387285e-05, "loss": 0.0578, "step": 26790 }, { "epoch": 0.9060482098786301, "grad_norm": 0.30977651476860046, "learning_rate": 8.75612665341067e-05, "loss": 0.0583, "step": 26800 }, { "epoch": 0.9063862875688833, "grad_norm": 0.4004599452018738, "learning_rate": 8.755035082184126e-05, "loss": 0.0607, "step": 26810 }, { "epoch": 0.9067243652591366, "grad_norm": 0.36826109886169434, "learning_rate": 8.753943100312592e-05, "loss": 0.0666, "step": 26820 }, { "epoch": 0.9070624429493898, "grad_norm": 0.25326523184776306, "learning_rate": 8.752850707915484e-05, "loss": 0.0577, "step": 26830 }, { "epoch": 0.9074005206396429, "grad_norm": 0.28956982493400574, "learning_rate": 8.751757905112264e-05, "loss": 0.0609, "step": 26840 }, { "epoch": 0.9077385983298962, "grad_norm": 0.2855215072631836, "learning_rate": 8.75066469202244e-05, "loss": 0.0559, "step": 26850 }, { "epoch": 0.9080766760201494, "grad_norm": 0.3143850862979889, "learning_rate": 8.749571068765567e-05, "loss": 0.0599, "step": 26860 }, { "epoch": 0.9084147537104027, "grad_norm": 0.2726300358772278, "learning_rate": 8.748477035461238e-05, "loss": 0.0552, "step": 26870 }, { "epoch": 0.9087528314006559, "grad_norm": 0.21600031852722168, "learning_rate": 8.747382592229095e-05, "loss": 0.0592, "step": 26880 }, { "epoch": 0.9090909090909091, "grad_norm": 0.2675752341747284, "learning_rate": 8.746287739188828e-05, "loss": 0.0636, "step": 26890 }, { "epoch": 0.9094289867811624, "grad_norm": 0.21656131744384766, "learning_rate": 8.745192476460165e-05, "loss": 0.0566, "step": 26900 }, { "epoch": 0.9097670644714155, "grad_norm": 0.24122540652751923, "learning_rate": 8.744096804162882e-05, "loss": 0.0554, "step": 26910 }, { "epoch": 0.9101051421616687, "grad_norm": 0.2316509336233139, "learning_rate": 8.743000722416804e-05, "loss": 0.0565, "step": 26920 }, { "epoch": 0.910443219851922, "grad_norm": 0.30969175696372986, "learning_rate": 8.741904231341793e-05, "loss": 0.0597, "step": 26930 }, { "epoch": 0.9107812975421752, "grad_norm": 0.3526761829853058, "learning_rate": 8.740807331057762e-05, "loss": 0.0559, "step": 26940 }, { "epoch": 0.9111193752324284, "grad_norm": 0.18193185329437256, "learning_rate": 8.739710021684667e-05, "loss": 0.0548, "step": 26950 }, { "epoch": 0.9114574529226817, "grad_norm": 0.26799431443214417, "learning_rate": 8.738612303342503e-05, "loss": 0.0564, "step": 26960 }, { "epoch": 0.9117955306129348, "grad_norm": 0.24514509737491608, "learning_rate": 8.73751417615132e-05, "loss": 0.0557, "step": 26970 }, { "epoch": 0.9121336083031881, "grad_norm": 0.37778812646865845, "learning_rate": 8.736415640231208e-05, "loss": 0.0598, "step": 26980 }, { "epoch": 0.9124716859934413, "grad_norm": 0.2082204669713974, "learning_rate": 8.735316695702297e-05, "loss": 0.0536, "step": 26990 }, { "epoch": 0.9128097636836945, "grad_norm": 0.1693698912858963, "learning_rate": 8.734217342684769e-05, "loss": 0.0559, "step": 27000 }, { "epoch": 0.9131478413739478, "grad_norm": 0.22167624533176422, "learning_rate": 8.733117581298847e-05, "loss": 0.0571, "step": 27010 }, { "epoch": 0.913485919064201, "grad_norm": 0.4448226988315582, "learning_rate": 8.732017411664796e-05, "loss": 0.0565, "step": 27020 }, { "epoch": 0.9138239967544541, "grad_norm": 0.21933391690254211, "learning_rate": 8.730916833902936e-05, "loss": 0.0564, "step": 27030 }, { "epoch": 0.9141620744447074, "grad_norm": 0.44050559401512146, "learning_rate": 8.729815848133618e-05, "loss": 0.0543, "step": 27040 }, { "epoch": 0.9145001521349606, "grad_norm": 0.2666349411010742, "learning_rate": 8.728714454477247e-05, "loss": 0.0552, "step": 27050 }, { "epoch": 0.9148382298252138, "grad_norm": 0.23728755116462708, "learning_rate": 8.727612653054269e-05, "loss": 0.0509, "step": 27060 }, { "epoch": 0.9151763075154671, "grad_norm": 0.2189275622367859, "learning_rate": 8.726510443985176e-05, "loss": 0.0588, "step": 27070 }, { "epoch": 0.9155143852057203, "grad_norm": 0.17322643101215363, "learning_rate": 8.725407827390503e-05, "loss": 0.057, "step": 27080 }, { "epoch": 0.9158524628959734, "grad_norm": 0.2746286690235138, "learning_rate": 8.724304803390833e-05, "loss": 0.0556, "step": 27090 }, { "epoch": 0.9161905405862267, "grad_norm": 0.21563194692134857, "learning_rate": 8.723201372106788e-05, "loss": 0.0554, "step": 27100 }, { "epoch": 0.9165286182764799, "grad_norm": 0.3425794541835785, "learning_rate": 8.722097533659038e-05, "loss": 0.0554, "step": 27110 }, { "epoch": 0.9168666959667332, "grad_norm": 0.3719121515750885, "learning_rate": 8.720993288168299e-05, "loss": 0.0617, "step": 27120 }, { "epoch": 0.9172047736569864, "grad_norm": 0.43559902906417847, "learning_rate": 8.719888635755327e-05, "loss": 0.0524, "step": 27130 }, { "epoch": 0.9175428513472396, "grad_norm": 0.27547720074653625, "learning_rate": 8.718783576540928e-05, "loss": 0.0591, "step": 27140 }, { "epoch": 0.9178809290374929, "grad_norm": 0.19882656633853912, "learning_rate": 8.717678110645948e-05, "loss": 0.0578, "step": 27150 }, { "epoch": 0.918219006727746, "grad_norm": 0.316353976726532, "learning_rate": 8.716572238191279e-05, "loss": 0.0598, "step": 27160 }, { "epoch": 0.9185570844179992, "grad_norm": 0.20970125496387482, "learning_rate": 8.715465959297857e-05, "loss": 0.0547, "step": 27170 }, { "epoch": 0.9188951621082525, "grad_norm": 0.3110063672065735, "learning_rate": 8.714359274086665e-05, "loss": 0.0625, "step": 27180 }, { "epoch": 0.9192332397985057, "grad_norm": 0.32079437375068665, "learning_rate": 8.713252182678726e-05, "loss": 0.0606, "step": 27190 }, { "epoch": 0.9195713174887589, "grad_norm": 0.3009950816631317, "learning_rate": 8.712144685195112e-05, "loss": 0.0643, "step": 27200 }, { "epoch": 0.9199093951790122, "grad_norm": 0.2974127233028412, "learning_rate": 8.711036781756936e-05, "loss": 0.0614, "step": 27210 }, { "epoch": 0.9202474728692653, "grad_norm": 0.24163280427455902, "learning_rate": 8.709928472485357e-05, "loss": 0.0547, "step": 27220 }, { "epoch": 0.9205855505595186, "grad_norm": 0.2808991074562073, "learning_rate": 8.708819757501579e-05, "loss": 0.062, "step": 27230 }, { "epoch": 0.9209236282497718, "grad_norm": 0.2541268765926361, "learning_rate": 8.707710636926846e-05, "loss": 0.0565, "step": 27240 }, { "epoch": 0.921261705940025, "grad_norm": 0.21222436428070068, "learning_rate": 8.706601110882455e-05, "loss": 0.0528, "step": 27250 }, { "epoch": 0.9215997836302783, "grad_norm": 0.2804566025733948, "learning_rate": 8.705491179489738e-05, "loss": 0.0564, "step": 27260 }, { "epoch": 0.9219378613205315, "grad_norm": 0.30954673886299133, "learning_rate": 8.704380842870077e-05, "loss": 0.0599, "step": 27270 }, { "epoch": 0.9222759390107846, "grad_norm": 0.2983124256134033, "learning_rate": 8.703270101144895e-05, "loss": 0.0639, "step": 27280 }, { "epoch": 0.9226140167010379, "grad_norm": 0.1812448650598526, "learning_rate": 8.702158954435664e-05, "loss": 0.0595, "step": 27290 }, { "epoch": 0.9229520943912911, "grad_norm": 0.23794174194335938, "learning_rate": 8.701047402863896e-05, "loss": 0.0593, "step": 27300 }, { "epoch": 0.9232901720815443, "grad_norm": 0.31037217378616333, "learning_rate": 8.699935446551148e-05, "loss": 0.0549, "step": 27310 }, { "epoch": 0.9236282497717976, "grad_norm": 0.346793532371521, "learning_rate": 8.698823085619022e-05, "loss": 0.057, "step": 27320 }, { "epoch": 0.9239663274620508, "grad_norm": 0.22993585467338562, "learning_rate": 8.697710320189166e-05, "loss": 0.0628, "step": 27330 }, { "epoch": 0.924304405152304, "grad_norm": 0.37710022926330566, "learning_rate": 8.696597150383268e-05, "loss": 0.055, "step": 27340 }, { "epoch": 0.9246424828425572, "grad_norm": 0.23568011820316315, "learning_rate": 8.695483576323063e-05, "loss": 0.0573, "step": 27350 }, { "epoch": 0.9249805605328104, "grad_norm": 0.284767210483551, "learning_rate": 8.69436959813033e-05, "loss": 0.0588, "step": 27360 }, { "epoch": 0.9253186382230637, "grad_norm": 0.21936537325382233, "learning_rate": 8.693255215926892e-05, "loss": 0.0648, "step": 27370 }, { "epoch": 0.9256567159133169, "grad_norm": 0.8427550792694092, "learning_rate": 8.692140429834617e-05, "loss": 0.0604, "step": 27380 }, { "epoch": 0.9259947936035701, "grad_norm": 0.423505961894989, "learning_rate": 8.691025239975415e-05, "loss": 0.0605, "step": 27390 }, { "epoch": 0.9263328712938234, "grad_norm": 0.22770507633686066, "learning_rate": 8.689909646471243e-05, "loss": 0.0498, "step": 27400 }, { "epoch": 0.9266709489840765, "grad_norm": 0.23345716297626495, "learning_rate": 8.688793649444099e-05, "loss": 0.0572, "step": 27410 }, { "epoch": 0.9270090266743297, "grad_norm": 0.2577124536037445, "learning_rate": 8.687677249016029e-05, "loss": 0.0538, "step": 27420 }, { "epoch": 0.927347104364583, "grad_norm": 0.20168113708496094, "learning_rate": 8.686560445309118e-05, "loss": 0.0576, "step": 27430 }, { "epoch": 0.9276851820548362, "grad_norm": 0.200599804520607, "learning_rate": 8.685443238445499e-05, "loss": 0.0646, "step": 27440 }, { "epoch": 0.9280232597450894, "grad_norm": 0.19799000024795532, "learning_rate": 8.68432562854735e-05, "loss": 0.0586, "step": 27450 }, { "epoch": 0.9283613374353427, "grad_norm": 0.27697956562042236, "learning_rate": 8.683207615736887e-05, "loss": 0.0588, "step": 27460 }, { "epoch": 0.9286994151255958, "grad_norm": 0.36460381746292114, "learning_rate": 8.682089200136379e-05, "loss": 0.0584, "step": 27470 }, { "epoch": 0.9290374928158491, "grad_norm": 0.2758576571941376, "learning_rate": 8.680970381868132e-05, "loss": 0.0486, "step": 27480 }, { "epoch": 0.9293755705061023, "grad_norm": 0.292339026927948, "learning_rate": 8.679851161054498e-05, "loss": 0.0614, "step": 27490 }, { "epoch": 0.9297136481963555, "grad_norm": 0.22039590775966644, "learning_rate": 8.678731537817873e-05, "loss": 0.0594, "step": 27500 }, { "epoch": 0.9300517258866088, "grad_norm": 0.22703947126865387, "learning_rate": 8.677611512280697e-05, "loss": 0.0546, "step": 27510 }, { "epoch": 0.930389803576862, "grad_norm": 0.22003188729286194, "learning_rate": 8.676491084565457e-05, "loss": 0.0613, "step": 27520 }, { "epoch": 0.9307278812671151, "grad_norm": 0.32317593693733215, "learning_rate": 8.675370254794678e-05, "loss": 0.056, "step": 27530 }, { "epoch": 0.9310659589573684, "grad_norm": 0.45636898279190063, "learning_rate": 8.674249023090935e-05, "loss": 0.0588, "step": 27540 }, { "epoch": 0.9314040366476216, "grad_norm": 0.18722409009933472, "learning_rate": 8.673127389576843e-05, "loss": 0.0584, "step": 27550 }, { "epoch": 0.9317421143378748, "grad_norm": 0.21962788701057434, "learning_rate": 8.67200535437506e-05, "loss": 0.058, "step": 27560 }, { "epoch": 0.9320801920281281, "grad_norm": 0.23530589044094086, "learning_rate": 8.670882917608296e-05, "loss": 0.0572, "step": 27570 }, { "epoch": 0.9324182697183813, "grad_norm": 0.2782084047794342, "learning_rate": 8.669760079399292e-05, "loss": 0.0577, "step": 27580 }, { "epoch": 0.9327563474086346, "grad_norm": 0.2620440721511841, "learning_rate": 8.668636839870845e-05, "loss": 0.0633, "step": 27590 }, { "epoch": 0.9330944250988877, "grad_norm": 0.35038280487060547, "learning_rate": 8.667513199145789e-05, "loss": 0.0561, "step": 27600 }, { "epoch": 0.9334325027891409, "grad_norm": 0.18315400183200836, "learning_rate": 8.666389157347002e-05, "loss": 0.0547, "step": 27610 }, { "epoch": 0.9337705804793942, "grad_norm": 0.2494194656610489, "learning_rate": 8.66526471459741e-05, "loss": 0.0615, "step": 27620 }, { "epoch": 0.9341086581696474, "grad_norm": 0.30585533380508423, "learning_rate": 8.66413987101998e-05, "loss": 0.0569, "step": 27630 }, { "epoch": 0.9344467358599006, "grad_norm": 0.20275908708572388, "learning_rate": 8.663014626737723e-05, "loss": 0.058, "step": 27640 }, { "epoch": 0.9347848135501539, "grad_norm": 0.5967563390731812, "learning_rate": 8.661888981873691e-05, "loss": 0.0567, "step": 27650 }, { "epoch": 0.935122891240407, "grad_norm": 0.1817665547132492, "learning_rate": 8.660762936550988e-05, "loss": 0.0595, "step": 27660 }, { "epoch": 0.9354609689306602, "grad_norm": 0.24626824259757996, "learning_rate": 8.659636490892753e-05, "loss": 0.0637, "step": 27670 }, { "epoch": 0.9357990466209135, "grad_norm": 0.3503853380680084, "learning_rate": 8.658509645022174e-05, "loss": 0.066, "step": 27680 }, { "epoch": 0.9361371243111667, "grad_norm": 0.20311133563518524, "learning_rate": 8.657382399062481e-05, "loss": 0.0574, "step": 27690 }, { "epoch": 0.9364752020014199, "grad_norm": 0.42610207200050354, "learning_rate": 8.656254753136946e-05, "loss": 0.0582, "step": 27700 }, { "epoch": 0.9368132796916732, "grad_norm": 0.21016359329223633, "learning_rate": 8.655126707368891e-05, "loss": 0.0556, "step": 27710 }, { "epoch": 0.9371513573819263, "grad_norm": 0.23693545162677765, "learning_rate": 8.653998261881672e-05, "loss": 0.0627, "step": 27720 }, { "epoch": 0.9374894350721796, "grad_norm": 0.2802076041698456, "learning_rate": 8.652869416798699e-05, "loss": 0.0653, "step": 27730 }, { "epoch": 0.9378275127624328, "grad_norm": 0.2382393330335617, "learning_rate": 8.651740172243417e-05, "loss": 0.0503, "step": 27740 }, { "epoch": 0.938165590452686, "grad_norm": 0.2633489966392517, "learning_rate": 8.65061052833932e-05, "loss": 0.0549, "step": 27750 }, { "epoch": 0.9385036681429393, "grad_norm": 0.1972370147705078, "learning_rate": 8.649480485209945e-05, "loss": 0.0599, "step": 27760 }, { "epoch": 0.9388417458331925, "grad_norm": 0.3204479515552521, "learning_rate": 8.64835004297887e-05, "loss": 0.06, "step": 27770 }, { "epoch": 0.9391798235234456, "grad_norm": 0.2054765224456787, "learning_rate": 8.64721920176972e-05, "loss": 0.0625, "step": 27780 }, { "epoch": 0.9395179012136989, "grad_norm": 0.18247488141059875, "learning_rate": 8.646087961706164e-05, "loss": 0.0591, "step": 27790 }, { "epoch": 0.9398559789039521, "grad_norm": 0.2896445691585541, "learning_rate": 8.644956322911908e-05, "loss": 0.0589, "step": 27800 }, { "epoch": 0.9401940565942053, "grad_norm": 0.3489110767841339, "learning_rate": 8.643824285510709e-05, "loss": 0.0536, "step": 27810 }, { "epoch": 0.9405321342844586, "grad_norm": 0.308946818113327, "learning_rate": 8.642691849626364e-05, "loss": 0.0558, "step": 27820 }, { "epoch": 0.9408702119747118, "grad_norm": 0.20222942531108856, "learning_rate": 8.641559015382717e-05, "loss": 0.0636, "step": 27830 }, { "epoch": 0.9412082896649651, "grad_norm": 0.22089342772960663, "learning_rate": 8.640425782903649e-05, "loss": 0.0515, "step": 27840 }, { "epoch": 0.9415463673552182, "grad_norm": 0.26765844225883484, "learning_rate": 8.639292152313091e-05, "loss": 0.0526, "step": 27850 }, { "epoch": 0.9418844450454714, "grad_norm": 0.23093800246715546, "learning_rate": 8.638158123735015e-05, "loss": 0.0556, "step": 27860 }, { "epoch": 0.9422225227357247, "grad_norm": 0.20345008373260498, "learning_rate": 8.637023697293436e-05, "loss": 0.0602, "step": 27870 }, { "epoch": 0.9425606004259779, "grad_norm": 0.2767925262451172, "learning_rate": 8.635888873112414e-05, "loss": 0.0632, "step": 27880 }, { "epoch": 0.9428986781162311, "grad_norm": 0.19982445240020752, "learning_rate": 8.634753651316052e-05, "loss": 0.0545, "step": 27890 }, { "epoch": 0.9432367558064844, "grad_norm": 0.28779879212379456, "learning_rate": 8.633618032028496e-05, "loss": 0.0652, "step": 27900 }, { "epoch": 0.9435748334967375, "grad_norm": 0.27541065216064453, "learning_rate": 8.632482015373934e-05, "loss": 0.0616, "step": 27910 }, { "epoch": 0.9439129111869907, "grad_norm": 0.20982608199119568, "learning_rate": 8.6313456014766e-05, "loss": 0.0548, "step": 27920 }, { "epoch": 0.944250988877244, "grad_norm": 0.32799044251441956, "learning_rate": 8.630208790460771e-05, "loss": 0.0583, "step": 27930 }, { "epoch": 0.9445890665674972, "grad_norm": 0.2855225205421448, "learning_rate": 8.629071582450768e-05, "loss": 0.0621, "step": 27940 }, { "epoch": 0.9449271442577505, "grad_norm": 0.3093525767326355, "learning_rate": 8.62793397757095e-05, "loss": 0.0555, "step": 27950 }, { "epoch": 0.9452652219480037, "grad_norm": 0.22889477014541626, "learning_rate": 8.626795975945729e-05, "loss": 0.0585, "step": 27960 }, { "epoch": 0.9456032996382568, "grad_norm": 0.2857290506362915, "learning_rate": 8.625657577699551e-05, "loss": 0.0574, "step": 27970 }, { "epoch": 0.9459413773285101, "grad_norm": 0.25313788652420044, "learning_rate": 8.624518782956914e-05, "loss": 0.0593, "step": 27980 }, { "epoch": 0.9462794550187633, "grad_norm": 0.24900245666503906, "learning_rate": 8.62337959184235e-05, "loss": 0.0626, "step": 27990 }, { "epoch": 0.9466175327090165, "grad_norm": 0.34239399433135986, "learning_rate": 8.622240004480441e-05, "loss": 0.0596, "step": 28000 }, { "epoch": 0.9469556103992698, "grad_norm": 0.21662788093090057, "learning_rate": 8.621100020995814e-05, "loss": 0.0602, "step": 28010 }, { "epoch": 0.947293688089523, "grad_norm": 0.2789100706577301, "learning_rate": 8.619959641513132e-05, "loss": 0.0587, "step": 28020 }, { "epoch": 0.9476317657797761, "grad_norm": 0.21488188207149506, "learning_rate": 8.618818866157105e-05, "loss": 0.0592, "step": 28030 }, { "epoch": 0.9479698434700294, "grad_norm": 0.21636086702346802, "learning_rate": 8.617677695052487e-05, "loss": 0.0562, "step": 28040 }, { "epoch": 0.9483079211602826, "grad_norm": 0.25161978602409363, "learning_rate": 8.616536128324078e-05, "loss": 0.0538, "step": 28050 }, { "epoch": 0.9486459988505358, "grad_norm": 0.2812119722366333, "learning_rate": 8.615394166096712e-05, "loss": 0.0585, "step": 28060 }, { "epoch": 0.9489840765407891, "grad_norm": 0.27190256118774414, "learning_rate": 8.614251808495279e-05, "loss": 0.0635, "step": 28070 }, { "epoch": 0.9493221542310423, "grad_norm": 0.2675907015800476, "learning_rate": 8.6131090556447e-05, "loss": 0.0584, "step": 28080 }, { "epoch": 0.9496602319212956, "grad_norm": 0.1972612887620926, "learning_rate": 8.611965907669947e-05, "loss": 0.0531, "step": 28090 }, { "epoch": 0.9499983096115487, "grad_norm": 0.2589345872402191, "learning_rate": 8.610822364696034e-05, "loss": 0.0574, "step": 28100 }, { "epoch": 0.9503363873018019, "grad_norm": 0.3297954201698303, "learning_rate": 8.609678426848015e-05, "loss": 0.0584, "step": 28110 }, { "epoch": 0.9506744649920552, "grad_norm": 0.31643810868263245, "learning_rate": 8.60853409425099e-05, "loss": 0.0595, "step": 28120 }, { "epoch": 0.9510125426823084, "grad_norm": 0.516297459602356, "learning_rate": 8.607389367030104e-05, "loss": 0.0534, "step": 28130 }, { "epoch": 0.9513506203725616, "grad_norm": 0.25619322061538696, "learning_rate": 8.606244245310538e-05, "loss": 0.0595, "step": 28140 }, { "epoch": 0.9516886980628149, "grad_norm": 0.23392635583877563, "learning_rate": 8.605098729217525e-05, "loss": 0.0632, "step": 28150 }, { "epoch": 0.952026775753068, "grad_norm": 0.3756990432739258, "learning_rate": 8.603952818876335e-05, "loss": 0.0533, "step": 28160 }, { "epoch": 0.9523648534433212, "grad_norm": 0.3093583583831787, "learning_rate": 8.602806514412281e-05, "loss": 0.0567, "step": 28170 }, { "epoch": 0.9527029311335745, "grad_norm": 0.24159274995326996, "learning_rate": 8.601659815950726e-05, "loss": 0.058, "step": 28180 }, { "epoch": 0.9530410088238277, "grad_norm": 0.297303169965744, "learning_rate": 8.600512723617067e-05, "loss": 0.0533, "step": 28190 }, { "epoch": 0.953379086514081, "grad_norm": 0.1882724016904831, "learning_rate": 8.59936523753675e-05, "loss": 0.0584, "step": 28200 }, { "epoch": 0.9537171642043342, "grad_norm": 0.5258207321166992, "learning_rate": 8.598217357835264e-05, "loss": 0.0579, "step": 28210 }, { "epoch": 0.9540552418945873, "grad_norm": 0.21015997231006622, "learning_rate": 8.597069084638135e-05, "loss": 0.0587, "step": 28220 }, { "epoch": 0.9543933195848406, "grad_norm": 0.2752506732940674, "learning_rate": 8.595920418070939e-05, "loss": 0.0658, "step": 28230 }, { "epoch": 0.9547313972750938, "grad_norm": 0.270167738199234, "learning_rate": 8.594771358259295e-05, "loss": 0.0574, "step": 28240 }, { "epoch": 0.955069474965347, "grad_norm": 0.19482193887233734, "learning_rate": 8.593621905328858e-05, "loss": 0.0574, "step": 28250 }, { "epoch": 0.9554075526556003, "grad_norm": 0.24729593098163605, "learning_rate": 8.592472059405333e-05, "loss": 0.0583, "step": 28260 }, { "epoch": 0.9557456303458535, "grad_norm": 0.25680410861968994, "learning_rate": 8.591321820614464e-05, "loss": 0.0629, "step": 28270 }, { "epoch": 0.9560837080361066, "grad_norm": 0.1892891824245453, "learning_rate": 8.590171189082041e-05, "loss": 0.0601, "step": 28280 }, { "epoch": 0.9564217857263599, "grad_norm": 0.18092714250087738, "learning_rate": 8.589020164933894e-05, "loss": 0.0594, "step": 28290 }, { "epoch": 0.9567598634166131, "grad_norm": 0.26537683606147766, "learning_rate": 8.587868748295898e-05, "loss": 0.0563, "step": 28300 }, { "epoch": 0.9570979411068664, "grad_norm": 0.3288348913192749, "learning_rate": 8.586716939293971e-05, "loss": 0.0635, "step": 28310 }, { "epoch": 0.9574360187971196, "grad_norm": 0.17725031077861786, "learning_rate": 8.58556473805407e-05, "loss": 0.0625, "step": 28320 }, { "epoch": 0.9577740964873728, "grad_norm": 0.3068852722644806, "learning_rate": 8.584412144702202e-05, "loss": 0.0569, "step": 28330 }, { "epoch": 0.9581121741776261, "grad_norm": 0.19381064176559448, "learning_rate": 8.58325915936441e-05, "loss": 0.0597, "step": 28340 }, { "epoch": 0.9584502518678792, "grad_norm": 0.34944063425064087, "learning_rate": 8.582105782166783e-05, "loss": 0.061, "step": 28350 }, { "epoch": 0.9587883295581324, "grad_norm": 0.3196231722831726, "learning_rate": 8.580952013235455e-05, "loss": 0.0594, "step": 28360 }, { "epoch": 0.9591264072483857, "grad_norm": 0.24443010985851288, "learning_rate": 8.579797852696596e-05, "loss": 0.0553, "step": 28370 }, { "epoch": 0.9594644849386389, "grad_norm": 0.30051514506340027, "learning_rate": 8.578643300676428e-05, "loss": 0.0527, "step": 28380 }, { "epoch": 0.9598025626288921, "grad_norm": 0.21578706800937653, "learning_rate": 8.577488357301209e-05, "loss": 0.0549, "step": 28390 }, { "epoch": 0.9601406403191454, "grad_norm": 0.19271942973136902, "learning_rate": 8.576333022697242e-05, "loss": 0.0527, "step": 28400 }, { "epoch": 0.9604787180093985, "grad_norm": 0.22369535267353058, "learning_rate": 8.575177296990873e-05, "loss": 0.06, "step": 28410 }, { "epoch": 0.9608167956996517, "grad_norm": 0.18468786776065826, "learning_rate": 8.574021180308489e-05, "loss": 0.0548, "step": 28420 }, { "epoch": 0.961154873389905, "grad_norm": 0.28831401467323303, "learning_rate": 8.572864672776523e-05, "loss": 0.0567, "step": 28430 }, { "epoch": 0.9614929510801582, "grad_norm": 0.18711143732070923, "learning_rate": 8.571707774521447e-05, "loss": 0.0547, "step": 28440 }, { "epoch": 0.9618310287704115, "grad_norm": 0.21363304555416107, "learning_rate": 8.57055048566978e-05, "loss": 0.0576, "step": 28450 }, { "epoch": 0.9621691064606647, "grad_norm": 0.2177598476409912, "learning_rate": 8.569392806348078e-05, "loss": 0.0568, "step": 28460 }, { "epoch": 0.9625071841509178, "grad_norm": 0.2504737973213196, "learning_rate": 8.568234736682947e-05, "loss": 0.0552, "step": 28470 }, { "epoch": 0.9628452618411711, "grad_norm": 0.22747676074504852, "learning_rate": 8.567076276801029e-05, "loss": 0.0614, "step": 28480 }, { "epoch": 0.9631833395314243, "grad_norm": 0.1578865945339203, "learning_rate": 8.565917426829013e-05, "loss": 0.0562, "step": 28490 }, { "epoch": 0.9635214172216775, "grad_norm": 0.2876586616039276, "learning_rate": 8.564758186893628e-05, "loss": 0.0567, "step": 28500 }, { "epoch": 0.9638594949119308, "grad_norm": 0.23777469992637634, "learning_rate": 8.563598557121649e-05, "loss": 0.0537, "step": 28510 }, { "epoch": 0.964197572602184, "grad_norm": 0.2586100399494171, "learning_rate": 8.562438537639888e-05, "loss": 0.0552, "step": 28520 }, { "epoch": 0.9645356502924372, "grad_norm": 0.40869832038879395, "learning_rate": 8.561278128575206e-05, "loss": 0.0526, "step": 28530 }, { "epoch": 0.9648737279826904, "grad_norm": 0.22499260306358337, "learning_rate": 8.5601173300545e-05, "loss": 0.0576, "step": 28540 }, { "epoch": 0.9652118056729436, "grad_norm": 0.449351042509079, "learning_rate": 8.558956142204717e-05, "loss": 0.0635, "step": 28550 }, { "epoch": 0.9655498833631969, "grad_norm": 0.20321573317050934, "learning_rate": 8.55779456515284e-05, "loss": 0.0623, "step": 28560 }, { "epoch": 0.9658879610534501, "grad_norm": 0.30158939957618713, "learning_rate": 8.556632599025898e-05, "loss": 0.0609, "step": 28570 }, { "epoch": 0.9662260387437033, "grad_norm": 0.33875033259391785, "learning_rate": 8.555470243950964e-05, "loss": 0.0622, "step": 28580 }, { "epoch": 0.9665641164339566, "grad_norm": 0.2104620486497879, "learning_rate": 8.554307500055148e-05, "loss": 0.0504, "step": 28590 }, { "epoch": 0.9669021941242097, "grad_norm": 0.23271724581718445, "learning_rate": 8.553144367465609e-05, "loss": 0.0612, "step": 28600 }, { "epoch": 0.9672402718144629, "grad_norm": 0.17206287384033203, "learning_rate": 8.551980846309544e-05, "loss": 0.0559, "step": 28610 }, { "epoch": 0.9675783495047162, "grad_norm": 0.2430161088705063, "learning_rate": 8.550816936714193e-05, "loss": 0.0592, "step": 28620 }, { "epoch": 0.9679164271949694, "grad_norm": 0.3285054564476013, "learning_rate": 8.549652638806841e-05, "loss": 0.0566, "step": 28630 }, { "epoch": 0.9682545048852226, "grad_norm": 0.18241873383522034, "learning_rate": 8.548487952714812e-05, "loss": 0.0549, "step": 28640 }, { "epoch": 0.9685925825754759, "grad_norm": 0.2923746705055237, "learning_rate": 8.547322878565478e-05, "loss": 0.0532, "step": 28650 }, { "epoch": 0.968930660265729, "grad_norm": 0.23598721623420715, "learning_rate": 8.546157416486245e-05, "loss": 0.0582, "step": 28660 }, { "epoch": 0.9692687379559823, "grad_norm": 0.29205015301704407, "learning_rate": 8.54499156660457e-05, "loss": 0.0574, "step": 28670 }, { "epoch": 0.9696068156462355, "grad_norm": 0.2681199312210083, "learning_rate": 8.543825329047947e-05, "loss": 0.0591, "step": 28680 }, { "epoch": 0.9699448933364887, "grad_norm": 0.20231126248836517, "learning_rate": 8.542658703943913e-05, "loss": 0.0621, "step": 28690 }, { "epoch": 0.970282971026742, "grad_norm": 0.23815090954303741, "learning_rate": 8.541491691420051e-05, "loss": 0.0572, "step": 28700 }, { "epoch": 0.9706210487169952, "grad_norm": 0.2119116485118866, "learning_rate": 8.54032429160398e-05, "loss": 0.0536, "step": 28710 }, { "epoch": 0.9709591264072484, "grad_norm": 0.20698465406894684, "learning_rate": 8.539156504623369e-05, "loss": 0.0483, "step": 28720 }, { "epoch": 0.9712972040975016, "grad_norm": 0.2880479395389557, "learning_rate": 8.537988330605923e-05, "loss": 0.0542, "step": 28730 }, { "epoch": 0.9716352817877548, "grad_norm": 0.23971110582351685, "learning_rate": 8.536819769679393e-05, "loss": 0.0607, "step": 28740 }, { "epoch": 0.971973359478008, "grad_norm": 0.5753219127655029, "learning_rate": 8.53565082197157e-05, "loss": 0.0528, "step": 28750 }, { "epoch": 0.9723114371682613, "grad_norm": 0.19717025756835938, "learning_rate": 8.534481487610289e-05, "loss": 0.0552, "step": 28760 }, { "epoch": 0.9726495148585145, "grad_norm": 0.22981220483779907, "learning_rate": 8.533311766723428e-05, "loss": 0.054, "step": 28770 }, { "epoch": 0.9729875925487677, "grad_norm": 0.24007798731327057, "learning_rate": 8.532141659438901e-05, "loss": 0.0568, "step": 28780 }, { "epoch": 0.973325670239021, "grad_norm": 0.24591046571731567, "learning_rate": 8.530971165884675e-05, "loss": 0.059, "step": 28790 }, { "epoch": 0.9736637479292741, "grad_norm": 0.2933906614780426, "learning_rate": 8.529800286188752e-05, "loss": 0.0569, "step": 28800 }, { "epoch": 0.9740018256195274, "grad_norm": 0.24541348218917847, "learning_rate": 8.528629020479175e-05, "loss": 0.0603, "step": 28810 }, { "epoch": 0.9743399033097806, "grad_norm": 0.25873762369155884, "learning_rate": 8.527457368884033e-05, "loss": 0.0564, "step": 28820 }, { "epoch": 0.9746779810000338, "grad_norm": 0.33243122696876526, "learning_rate": 8.526285331531458e-05, "loss": 0.0567, "step": 28830 }, { "epoch": 0.9750160586902871, "grad_norm": 0.2765473425388336, "learning_rate": 8.525112908549621e-05, "loss": 0.0644, "step": 28840 }, { "epoch": 0.9753541363805402, "grad_norm": 0.32350486516952515, "learning_rate": 8.523940100066735e-05, "loss": 0.0566, "step": 28850 }, { "epoch": 0.9756922140707934, "grad_norm": 0.429124116897583, "learning_rate": 8.52276690621106e-05, "loss": 0.0646, "step": 28860 }, { "epoch": 0.9760302917610467, "grad_norm": 0.3338325619697571, "learning_rate": 8.521593327110889e-05, "loss": 0.0595, "step": 28870 }, { "epoch": 0.9763683694512999, "grad_norm": 0.46927258372306824, "learning_rate": 8.520419362894569e-05, "loss": 0.0637, "step": 28880 }, { "epoch": 0.9767064471415531, "grad_norm": 0.34619781374931335, "learning_rate": 8.51924501369048e-05, "loss": 0.0523, "step": 28890 }, { "epoch": 0.9770445248318064, "grad_norm": 0.2514932155609131, "learning_rate": 8.518070279627047e-05, "loss": 0.0535, "step": 28900 }, { "epoch": 0.9773826025220596, "grad_norm": 0.3143792450428009, "learning_rate": 8.516895160832737e-05, "loss": 0.0664, "step": 28910 }, { "epoch": 0.9777206802123128, "grad_norm": 0.20537064969539642, "learning_rate": 8.515719657436061e-05, "loss": 0.0522, "step": 28920 }, { "epoch": 0.978058757902566, "grad_norm": 0.21513788402080536, "learning_rate": 8.514543769565568e-05, "loss": 0.049, "step": 28930 }, { "epoch": 0.9783968355928192, "grad_norm": 0.17775952816009521, "learning_rate": 8.513367497349853e-05, "loss": 0.0539, "step": 28940 }, { "epoch": 0.9787349132830725, "grad_norm": 0.3320559859275818, "learning_rate": 8.51219084091755e-05, "loss": 0.0621, "step": 28950 }, { "epoch": 0.9790729909733257, "grad_norm": 0.37809690833091736, "learning_rate": 8.511013800397338e-05, "loss": 0.0602, "step": 28960 }, { "epoch": 0.9794110686635789, "grad_norm": 0.2751343846321106, "learning_rate": 8.509836375917937e-05, "loss": 0.0621, "step": 28970 }, { "epoch": 0.9797491463538321, "grad_norm": 0.22609800100326538, "learning_rate": 8.508658567608104e-05, "loss": 0.0533, "step": 28980 }, { "epoch": 0.9800872240440853, "grad_norm": 0.2775111794471741, "learning_rate": 8.507480375596647e-05, "loss": 0.0573, "step": 28990 }, { "epoch": 0.9804253017343385, "grad_norm": 0.20662124454975128, "learning_rate": 8.506301800012408e-05, "loss": 0.0538, "step": 29000 }, { "epoch": 0.9807633794245918, "grad_norm": 0.2533809244632721, "learning_rate": 8.505122840984278e-05, "loss": 0.0588, "step": 29010 }, { "epoch": 0.981101457114845, "grad_norm": 0.24276135861873627, "learning_rate": 8.503943498641182e-05, "loss": 0.0575, "step": 29020 }, { "epoch": 0.9814395348050982, "grad_norm": 0.36221280694007874, "learning_rate": 8.502763773112095e-05, "loss": 0.0576, "step": 29030 }, { "epoch": 0.9817776124953514, "grad_norm": 0.25838467478752136, "learning_rate": 8.501583664526026e-05, "loss": 0.0604, "step": 29040 }, { "epoch": 0.9821156901856046, "grad_norm": 0.26990994811058044, "learning_rate": 8.500403173012032e-05, "loss": 0.0574, "step": 29050 }, { "epoch": 0.9824537678758579, "grad_norm": 0.24814358353614807, "learning_rate": 8.499222298699211e-05, "loss": 0.0633, "step": 29060 }, { "epoch": 0.9827918455661111, "grad_norm": 0.19438640773296356, "learning_rate": 8.498041041716701e-05, "loss": 0.0623, "step": 29070 }, { "epoch": 0.9831299232563643, "grad_norm": 0.16618147492408752, "learning_rate": 8.496859402193681e-05, "loss": 0.0531, "step": 29080 }, { "epoch": 0.9834680009466176, "grad_norm": 0.31890538334846497, "learning_rate": 8.495677380259374e-05, "loss": 0.062, "step": 29090 }, { "epoch": 0.9838060786368708, "grad_norm": 0.22196617722511292, "learning_rate": 8.494494976043045e-05, "loss": 0.0562, "step": 29100 }, { "epoch": 0.9841441563271239, "grad_norm": 0.24087682366371155, "learning_rate": 8.493312189673998e-05, "loss": 0.0512, "step": 29110 }, { "epoch": 0.9844822340173772, "grad_norm": 0.3915475010871887, "learning_rate": 8.492129021281584e-05, "loss": 0.0526, "step": 29120 }, { "epoch": 0.9848203117076304, "grad_norm": 0.2642568051815033, "learning_rate": 8.490945470995188e-05, "loss": 0.0633, "step": 29130 }, { "epoch": 0.9851583893978836, "grad_norm": 0.17761334776878357, "learning_rate": 8.489761538944247e-05, "loss": 0.061, "step": 29140 }, { "epoch": 0.9854964670881369, "grad_norm": 0.35002636909484863, "learning_rate": 8.48857722525823e-05, "loss": 0.0602, "step": 29150 }, { "epoch": 0.98583454477839, "grad_norm": 0.2412700057029724, "learning_rate": 8.487392530066652e-05, "loss": 0.057, "step": 29160 }, { "epoch": 0.9861726224686433, "grad_norm": 0.1912599354982376, "learning_rate": 8.486207453499069e-05, "loss": 0.0571, "step": 29170 }, { "epoch": 0.9865107001588965, "grad_norm": 0.23229798674583435, "learning_rate": 8.485021995685082e-05, "loss": 0.0574, "step": 29180 }, { "epoch": 0.9868487778491497, "grad_norm": 0.28355157375335693, "learning_rate": 8.483836156754328e-05, "loss": 0.0625, "step": 29190 }, { "epoch": 0.987186855539403, "grad_norm": 0.3354129195213318, "learning_rate": 8.482649936836491e-05, "loss": 0.0549, "step": 29200 }, { "epoch": 0.9875249332296562, "grad_norm": 0.22971990704536438, "learning_rate": 8.481463336061293e-05, "loss": 0.0605, "step": 29210 }, { "epoch": 0.9878630109199094, "grad_norm": 0.2419731169939041, "learning_rate": 8.480276354558496e-05, "loss": 0.0529, "step": 29220 }, { "epoch": 0.9882010886101626, "grad_norm": 0.2559937536716461, "learning_rate": 8.479088992457913e-05, "loss": 0.0552, "step": 29230 }, { "epoch": 0.9885391663004158, "grad_norm": 0.24748553335666656, "learning_rate": 8.477901249889387e-05, "loss": 0.0563, "step": 29240 }, { "epoch": 0.988877243990669, "grad_norm": 0.185343936085701, "learning_rate": 8.47671312698281e-05, "loss": 0.0556, "step": 29250 }, { "epoch": 0.9892153216809223, "grad_norm": 0.2993527948856354, "learning_rate": 8.475524623868112e-05, "loss": 0.058, "step": 29260 }, { "epoch": 0.9895533993711755, "grad_norm": 0.25667330622673035, "learning_rate": 8.474335740675266e-05, "loss": 0.0601, "step": 29270 }, { "epoch": 0.9898914770614288, "grad_norm": 0.3065277338027954, "learning_rate": 8.473146477534289e-05, "loss": 0.0596, "step": 29280 }, { "epoch": 0.990229554751682, "grad_norm": 0.24253907799720764, "learning_rate": 8.471956834575232e-05, "loss": 0.06, "step": 29290 }, { "epoch": 0.9905676324419351, "grad_norm": 0.23751549422740936, "learning_rate": 8.470766811928197e-05, "loss": 0.0667, "step": 29300 }, { "epoch": 0.9909057101321884, "grad_norm": 0.2114502489566803, "learning_rate": 8.469576409723323e-05, "loss": 0.0581, "step": 29310 }, { "epoch": 0.9912437878224416, "grad_norm": 0.4128386080265045, "learning_rate": 8.468385628090788e-05, "loss": 0.056, "step": 29320 }, { "epoch": 0.9915818655126948, "grad_norm": 0.24114902317523956, "learning_rate": 8.467194467160815e-05, "loss": 0.0568, "step": 29330 }, { "epoch": 0.9919199432029481, "grad_norm": 0.2377741038799286, "learning_rate": 8.466002927063667e-05, "loss": 0.0533, "step": 29340 }, { "epoch": 0.9922580208932013, "grad_norm": 0.29987382888793945, "learning_rate": 8.464811007929651e-05, "loss": 0.0535, "step": 29350 }, { "epoch": 0.9925960985834544, "grad_norm": 0.27130553126335144, "learning_rate": 8.463618709889114e-05, "loss": 0.0588, "step": 29360 }, { "epoch": 0.9929341762737077, "grad_norm": 0.3875846564769745, "learning_rate": 8.462426033072442e-05, "loss": 0.0577, "step": 29370 }, { "epoch": 0.9932722539639609, "grad_norm": 0.30050912499427795, "learning_rate": 8.461232977610061e-05, "loss": 0.0572, "step": 29380 }, { "epoch": 0.9936103316542141, "grad_norm": 0.3484613597393036, "learning_rate": 8.46003954363245e-05, "loss": 0.0585, "step": 29390 }, { "epoch": 0.9939484093444674, "grad_norm": 0.4068176746368408, "learning_rate": 8.458845731270115e-05, "loss": 0.0523, "step": 29400 }, { "epoch": 0.9942864870347206, "grad_norm": 0.2021772712469101, "learning_rate": 8.45765154065361e-05, "loss": 0.0589, "step": 29410 }, { "epoch": 0.9946245647249738, "grad_norm": 0.23124484717845917, "learning_rate": 8.456456971913532e-05, "loss": 0.057, "step": 29420 }, { "epoch": 0.994962642415227, "grad_norm": 0.30816903710365295, "learning_rate": 8.455262025180517e-05, "loss": 0.0588, "step": 29430 }, { "epoch": 0.9953007201054802, "grad_norm": 0.34645578265190125, "learning_rate": 8.454066700585242e-05, "loss": 0.0594, "step": 29440 }, { "epoch": 0.9956387977957335, "grad_norm": 0.18364933133125305, "learning_rate": 8.452870998258423e-05, "loss": 0.0551, "step": 29450 }, { "epoch": 0.9959768754859867, "grad_norm": 0.29195091128349304, "learning_rate": 8.451674918330825e-05, "loss": 0.0536, "step": 29460 }, { "epoch": 0.9963149531762399, "grad_norm": 0.20182816684246063, "learning_rate": 8.450478460933246e-05, "loss": 0.0566, "step": 29470 }, { "epoch": 0.9966530308664932, "grad_norm": 0.17246517539024353, "learning_rate": 8.449281626196532e-05, "loss": 0.0591, "step": 29480 }, { "epoch": 0.9969911085567463, "grad_norm": 0.18027426302433014, "learning_rate": 8.448084414251564e-05, "loss": 0.0536, "step": 29490 }, { "epoch": 0.9973291862469995, "grad_norm": 0.24290399253368378, "learning_rate": 8.446886825229271e-05, "loss": 0.058, "step": 29500 }, { "epoch": 0.9976672639372528, "grad_norm": 0.34822094440460205, "learning_rate": 8.445688859260615e-05, "loss": 0.0573, "step": 29510 }, { "epoch": 0.998005341627506, "grad_norm": 0.17778699100017548, "learning_rate": 8.444490516476606e-05, "loss": 0.056, "step": 29520 }, { "epoch": 0.9983434193177593, "grad_norm": 0.21070852875709534, "learning_rate": 8.443291797008293e-05, "loss": 0.0558, "step": 29530 }, { "epoch": 0.9986814970080125, "grad_norm": 0.3271201252937317, "learning_rate": 8.442092700986765e-05, "loss": 0.0616, "step": 29540 }, { "epoch": 0.9990195746982656, "grad_norm": 0.503162145614624, "learning_rate": 8.440893228543156e-05, "loss": 0.0568, "step": 29550 }, { "epoch": 0.9993576523885189, "grad_norm": 0.25710949301719666, "learning_rate": 8.439693379808638e-05, "loss": 0.057, "step": 29560 }, { "epoch": 0.9996957300787721, "grad_norm": 0.34421196579933167, "learning_rate": 8.43849315491442e-05, "loss": 0.0594, "step": 29570 }, { "epoch": 1.0000338077690254, "grad_norm": 0.31338146328926086, "learning_rate": 8.437292553991763e-05, "loss": 0.0522, "step": 29580 }, { "epoch": 1.0003718854592785, "grad_norm": 0.22379431128501892, "learning_rate": 8.436091577171959e-05, "loss": 0.0589, "step": 29590 }, { "epoch": 1.0007099631495318, "grad_norm": 0.32499879598617554, "learning_rate": 8.434890224586347e-05, "loss": 0.0531, "step": 29600 }, { "epoch": 1.001048040839785, "grad_norm": 0.4452666640281677, "learning_rate": 8.433688496366303e-05, "loss": 0.0558, "step": 29610 }, { "epoch": 1.0013861185300381, "grad_norm": 0.19225937128067017, "learning_rate": 8.432486392643248e-05, "loss": 0.0542, "step": 29620 }, { "epoch": 1.0017241962202914, "grad_norm": 0.25763794779777527, "learning_rate": 8.431283913548643e-05, "loss": 0.0542, "step": 29630 }, { "epoch": 1.0020622739105447, "grad_norm": 0.2373788058757782, "learning_rate": 8.430081059213985e-05, "loss": 0.0528, "step": 29640 }, { "epoch": 1.0024003516007978, "grad_norm": 0.22337885200977325, "learning_rate": 8.428877829770823e-05, "loss": 0.0549, "step": 29650 }, { "epoch": 1.002738429291051, "grad_norm": 0.22774910926818848, "learning_rate": 8.427674225350735e-05, "loss": 0.0521, "step": 29660 }, { "epoch": 1.0030765069813043, "grad_norm": 0.17805255949497223, "learning_rate": 8.426470246085347e-05, "loss": 0.0579, "step": 29670 }, { "epoch": 1.0034145846715574, "grad_norm": 0.27217423915863037, "learning_rate": 8.425265892106324e-05, "loss": 0.0604, "step": 29680 }, { "epoch": 1.0037526623618107, "grad_norm": 0.20270851254463196, "learning_rate": 8.424061163545374e-05, "loss": 0.0556, "step": 29690 }, { "epoch": 1.004090740052064, "grad_norm": 0.21686801314353943, "learning_rate": 8.422856060534243e-05, "loss": 0.053, "step": 29700 }, { "epoch": 1.0044288177423173, "grad_norm": 0.18692874908447266, "learning_rate": 8.421650583204718e-05, "loss": 0.0575, "step": 29710 }, { "epoch": 1.0047668954325704, "grad_norm": 0.2342093139886856, "learning_rate": 8.420444731688633e-05, "loss": 0.0494, "step": 29720 }, { "epoch": 1.0051049731228237, "grad_norm": 0.37772250175476074, "learning_rate": 8.419238506117852e-05, "loss": 0.0544, "step": 29730 }, { "epoch": 1.005443050813077, "grad_norm": 0.2375873625278473, "learning_rate": 8.418031906624289e-05, "loss": 0.0525, "step": 29740 }, { "epoch": 1.00578112850333, "grad_norm": 0.31944817304611206, "learning_rate": 8.416824933339898e-05, "loss": 0.0605, "step": 29750 }, { "epoch": 1.0061192061935833, "grad_norm": 0.25266075134277344, "learning_rate": 8.415617586396667e-05, "loss": 0.058, "step": 29760 }, { "epoch": 1.0064572838838366, "grad_norm": 0.18980145454406738, "learning_rate": 8.414409865926632e-05, "loss": 0.0589, "step": 29770 }, { "epoch": 1.0067953615740897, "grad_norm": 0.31625449657440186, "learning_rate": 8.413201772061867e-05, "loss": 0.0619, "step": 29780 }, { "epoch": 1.007133439264343, "grad_norm": 0.31263867020606995, "learning_rate": 8.411993304934488e-05, "loss": 0.0535, "step": 29790 }, { "epoch": 1.0074715169545962, "grad_norm": 0.4078647494316101, "learning_rate": 8.410784464676654e-05, "loss": 0.0612, "step": 29800 }, { "epoch": 1.0078095946448493, "grad_norm": 0.2959587574005127, "learning_rate": 8.409575251420556e-05, "loss": 0.0581, "step": 29810 }, { "epoch": 1.0081476723351026, "grad_norm": 0.2525216042995453, "learning_rate": 8.408365665298435e-05, "loss": 0.0558, "step": 29820 }, { "epoch": 1.008485750025356, "grad_norm": 0.2571093738079071, "learning_rate": 8.40715570644257e-05, "loss": 0.0514, "step": 29830 }, { "epoch": 1.008823827715609, "grad_norm": 0.2271224707365036, "learning_rate": 8.40594537498528e-05, "loss": 0.0619, "step": 29840 }, { "epoch": 1.0091619054058623, "grad_norm": 0.4147769808769226, "learning_rate": 8.404734671058924e-05, "loss": 0.0553, "step": 29850 }, { "epoch": 1.0094999830961155, "grad_norm": 0.25654464960098267, "learning_rate": 8.403523594795902e-05, "loss": 0.053, "step": 29860 }, { "epoch": 1.0098380607863686, "grad_norm": 0.26181483268737793, "learning_rate": 8.402312146328659e-05, "loss": 0.0547, "step": 29870 }, { "epoch": 1.010176138476622, "grad_norm": 0.3016165792942047, "learning_rate": 8.401100325789675e-05, "loss": 0.061, "step": 29880 }, { "epoch": 1.0105142161668752, "grad_norm": 0.2329055219888687, "learning_rate": 8.399888133311472e-05, "loss": 0.06, "step": 29890 }, { "epoch": 1.0108522938571283, "grad_norm": 0.1980208158493042, "learning_rate": 8.398675569026613e-05, "loss": 0.0525, "step": 29900 }, { "epoch": 1.0111903715473816, "grad_norm": 0.2627086937427521, "learning_rate": 8.397462633067705e-05, "loss": 0.053, "step": 29910 }, { "epoch": 1.0115284492376349, "grad_norm": 0.35110634565353394, "learning_rate": 8.396249325567392e-05, "loss": 0.063, "step": 29920 }, { "epoch": 1.011866526927888, "grad_norm": 0.2694043219089508, "learning_rate": 8.395035646658357e-05, "loss": 0.0622, "step": 29930 }, { "epoch": 1.0122046046181412, "grad_norm": 0.42206767201423645, "learning_rate": 8.39382159647333e-05, "loss": 0.0574, "step": 29940 }, { "epoch": 1.0125426823083945, "grad_norm": 0.42480364441871643, "learning_rate": 8.392607175145075e-05, "loss": 0.0557, "step": 29950 }, { "epoch": 1.0128807599986478, "grad_norm": 0.27873238921165466, "learning_rate": 8.3913923828064e-05, "loss": 0.0552, "step": 29960 }, { "epoch": 1.0132188376889009, "grad_norm": 0.18601368367671967, "learning_rate": 8.390177219590152e-05, "loss": 0.0564, "step": 29970 }, { "epoch": 1.0135569153791542, "grad_norm": 0.3021586537361145, "learning_rate": 8.388961685629222e-05, "loss": 0.0594, "step": 29980 }, { "epoch": 1.0138949930694074, "grad_norm": 0.36900416016578674, "learning_rate": 8.387745781056536e-05, "loss": 0.0608, "step": 29990 }, { "epoch": 1.0142330707596605, "grad_norm": 0.4049617350101471, "learning_rate": 8.386529506005065e-05, "loss": 0.0579, "step": 30000 }, { "epoch": 1.0145711484499138, "grad_norm": 0.3008628487586975, "learning_rate": 8.38531286060782e-05, "loss": 0.0587, "step": 30010 }, { "epoch": 1.014909226140167, "grad_norm": 0.23667190968990326, "learning_rate": 8.384095844997849e-05, "loss": 0.0621, "step": 30020 }, { "epoch": 1.0152473038304202, "grad_norm": 0.2903751730918884, "learning_rate": 8.382878459308245e-05, "loss": 0.0625, "step": 30030 }, { "epoch": 1.0155853815206735, "grad_norm": 0.19539564847946167, "learning_rate": 8.381660703672138e-05, "loss": 0.0574, "step": 30040 }, { "epoch": 1.0159234592109267, "grad_norm": 0.288714200258255, "learning_rate": 8.380442578222702e-05, "loss": 0.0558, "step": 30050 }, { "epoch": 1.0162615369011798, "grad_norm": 0.29645538330078125, "learning_rate": 8.379224083093146e-05, "loss": 0.057, "step": 30060 }, { "epoch": 1.016599614591433, "grad_norm": 0.37345314025878906, "learning_rate": 8.378005218416727e-05, "loss": 0.0615, "step": 30070 }, { "epoch": 1.0169376922816864, "grad_norm": 0.270611047744751, "learning_rate": 8.376785984326735e-05, "loss": 0.0627, "step": 30080 }, { "epoch": 1.0172757699719395, "grad_norm": 0.4002110958099365, "learning_rate": 8.375566380956506e-05, "loss": 0.0593, "step": 30090 }, { "epoch": 1.0176138476621928, "grad_norm": 0.34460118412971497, "learning_rate": 8.374346408439411e-05, "loss": 0.0634, "step": 30100 }, { "epoch": 1.017951925352446, "grad_norm": 0.21797610819339752, "learning_rate": 8.373126066908868e-05, "loss": 0.0549, "step": 30110 }, { "epoch": 1.0182900030426991, "grad_norm": 0.2865511476993561, "learning_rate": 8.371905356498326e-05, "loss": 0.0537, "step": 30120 }, { "epoch": 1.0186280807329524, "grad_norm": 0.24037304520606995, "learning_rate": 8.370684277341288e-05, "loss": 0.0564, "step": 30130 }, { "epoch": 1.0189661584232057, "grad_norm": 0.18267810344696045, "learning_rate": 8.369462829571282e-05, "loss": 0.0533, "step": 30140 }, { "epoch": 1.0193042361134588, "grad_norm": 0.3526636064052582, "learning_rate": 8.36824101332189e-05, "loss": 0.0554, "step": 30150 }, { "epoch": 1.019642313803712, "grad_norm": 0.18829916417598724, "learning_rate": 8.367018828726721e-05, "loss": 0.0567, "step": 30160 }, { "epoch": 1.0199803914939654, "grad_norm": 0.2943969666957855, "learning_rate": 8.365796275919438e-05, "loss": 0.0565, "step": 30170 }, { "epoch": 1.0203184691842186, "grad_norm": 0.24244864284992218, "learning_rate": 8.364573355033734e-05, "loss": 0.0565, "step": 30180 }, { "epoch": 1.0206565468744717, "grad_norm": 0.21778951585292816, "learning_rate": 8.363350066203346e-05, "loss": 0.0534, "step": 30190 }, { "epoch": 1.020994624564725, "grad_norm": 0.22159166634082794, "learning_rate": 8.362126409562053e-05, "loss": 0.0635, "step": 30200 }, { "epoch": 1.0213327022549783, "grad_norm": 0.1867523342370987, "learning_rate": 8.360902385243667e-05, "loss": 0.0576, "step": 30210 }, { "epoch": 1.0216707799452314, "grad_norm": 0.4878412187099457, "learning_rate": 8.359677993382052e-05, "loss": 0.0545, "step": 30220 }, { "epoch": 1.0220088576354847, "grad_norm": 0.2890920341014862, "learning_rate": 8.358453234111103e-05, "loss": 0.0561, "step": 30230 }, { "epoch": 1.022346935325738, "grad_norm": 0.3217635154724121, "learning_rate": 8.357228107564756e-05, "loss": 0.0543, "step": 30240 }, { "epoch": 1.022685013015991, "grad_norm": 0.23452994227409363, "learning_rate": 8.356002613876993e-05, "loss": 0.0576, "step": 30250 }, { "epoch": 1.0230230907062443, "grad_norm": 0.31305399537086487, "learning_rate": 8.35477675318183e-05, "loss": 0.0525, "step": 30260 }, { "epoch": 1.0233611683964976, "grad_norm": 0.21782951056957245, "learning_rate": 8.353550525613323e-05, "loss": 0.0554, "step": 30270 }, { "epoch": 1.0236992460867507, "grad_norm": 0.20276950299739838, "learning_rate": 8.352323931305572e-05, "loss": 0.0522, "step": 30280 }, { "epoch": 1.024037323777004, "grad_norm": 0.3241713047027588, "learning_rate": 8.351096970392717e-05, "loss": 0.0585, "step": 30290 }, { "epoch": 1.0243754014672573, "grad_norm": 0.2027338296175003, "learning_rate": 8.349869643008937e-05, "loss": 0.0658, "step": 30300 }, { "epoch": 1.0247134791575103, "grad_norm": 0.29332631826400757, "learning_rate": 8.348641949288449e-05, "loss": 0.0614, "step": 30310 }, { "epoch": 1.0250515568477636, "grad_norm": 0.2801888883113861, "learning_rate": 8.34741388936551e-05, "loss": 0.0581, "step": 30320 }, { "epoch": 1.025389634538017, "grad_norm": 0.18328742682933807, "learning_rate": 8.346185463374423e-05, "loss": 0.0566, "step": 30330 }, { "epoch": 1.02572771222827, "grad_norm": 0.2191256731748581, "learning_rate": 8.344956671449524e-05, "loss": 0.0531, "step": 30340 }, { "epoch": 1.0260657899185233, "grad_norm": 0.22319386899471283, "learning_rate": 8.343727513725192e-05, "loss": 0.0525, "step": 30350 }, { "epoch": 1.0264038676087766, "grad_norm": 0.20702305436134338, "learning_rate": 8.342497990335847e-05, "loss": 0.0539, "step": 30360 }, { "epoch": 1.0267419452990296, "grad_norm": 0.24536588788032532, "learning_rate": 8.341268101415946e-05, "loss": 0.0596, "step": 30370 }, { "epoch": 1.027080022989283, "grad_norm": 0.1974267065525055, "learning_rate": 8.34003784709999e-05, "loss": 0.0561, "step": 30380 }, { "epoch": 1.0274181006795362, "grad_norm": 0.23244448006153107, "learning_rate": 8.338807227522516e-05, "loss": 0.0571, "step": 30390 }, { "epoch": 1.0277561783697893, "grad_norm": 0.3268997073173523, "learning_rate": 8.337576242818103e-05, "loss": 0.0591, "step": 30400 }, { "epoch": 1.0280942560600426, "grad_norm": 0.2919386625289917, "learning_rate": 8.336344893121372e-05, "loss": 0.0591, "step": 30410 }, { "epoch": 1.0284323337502959, "grad_norm": 0.28808414936065674, "learning_rate": 8.335113178566977e-05, "loss": 0.0583, "step": 30420 }, { "epoch": 1.028770411440549, "grad_norm": 0.30076348781585693, "learning_rate": 8.33388109928962e-05, "loss": 0.0562, "step": 30430 }, { "epoch": 1.0291084891308022, "grad_norm": 0.2831997573375702, "learning_rate": 8.33264865542404e-05, "loss": 0.0535, "step": 30440 }, { "epoch": 1.0294465668210555, "grad_norm": 0.2503221035003662, "learning_rate": 8.331415847105013e-05, "loss": 0.0539, "step": 30450 }, { "epoch": 1.0297846445113088, "grad_norm": 0.24918358027935028, "learning_rate": 8.330182674467357e-05, "loss": 0.0537, "step": 30460 }, { "epoch": 1.0301227222015619, "grad_norm": 0.2822079062461853, "learning_rate": 8.32894913764593e-05, "loss": 0.0546, "step": 30470 }, { "epoch": 1.0304607998918152, "grad_norm": 0.22754895687103271, "learning_rate": 8.327715236775633e-05, "loss": 0.0544, "step": 30480 }, { "epoch": 1.0307988775820685, "grad_norm": 0.2886335849761963, "learning_rate": 8.326480971991398e-05, "loss": 0.055, "step": 30490 }, { "epoch": 1.0311369552723215, "grad_norm": 0.588821291923523, "learning_rate": 8.325246343428206e-05, "loss": 0.0641, "step": 30500 }, { "epoch": 1.0314750329625748, "grad_norm": 0.2689781188964844, "learning_rate": 8.324011351221072e-05, "loss": 0.0567, "step": 30510 }, { "epoch": 1.031813110652828, "grad_norm": 0.2751968204975128, "learning_rate": 8.322775995505057e-05, "loss": 0.0565, "step": 30520 }, { "epoch": 1.0321511883430812, "grad_norm": 0.16389794647693634, "learning_rate": 8.321540276415254e-05, "loss": 0.0521, "step": 30530 }, { "epoch": 1.0324892660333345, "grad_norm": 0.3266281485557556, "learning_rate": 8.320304194086798e-05, "loss": 0.0586, "step": 30540 }, { "epoch": 1.0328273437235878, "grad_norm": 0.31227347254753113, "learning_rate": 8.31906774865487e-05, "loss": 0.0589, "step": 30550 }, { "epoch": 1.0331654214138408, "grad_norm": 0.18953527510166168, "learning_rate": 8.317830940254682e-05, "loss": 0.0554, "step": 30560 }, { "epoch": 1.0335034991040941, "grad_norm": 0.33481013774871826, "learning_rate": 8.316593769021491e-05, "loss": 0.0584, "step": 30570 }, { "epoch": 1.0338415767943474, "grad_norm": 0.34845617413520813, "learning_rate": 8.315356235090592e-05, "loss": 0.0556, "step": 30580 }, { "epoch": 1.0341796544846005, "grad_norm": 0.1666039228439331, "learning_rate": 8.314118338597319e-05, "loss": 0.0585, "step": 30590 }, { "epoch": 1.0345177321748538, "grad_norm": 0.20304784178733826, "learning_rate": 8.312880079677048e-05, "loss": 0.0616, "step": 30600 }, { "epoch": 1.034855809865107, "grad_norm": 0.29050347208976746, "learning_rate": 8.311641458465191e-05, "loss": 0.0519, "step": 30610 }, { "epoch": 1.0351938875553601, "grad_norm": 0.2757691442966461, "learning_rate": 8.310402475097205e-05, "loss": 0.0595, "step": 30620 }, { "epoch": 1.0355319652456134, "grad_norm": 0.2555352747440338, "learning_rate": 8.309163129708581e-05, "loss": 0.0491, "step": 30630 }, { "epoch": 1.0358700429358667, "grad_norm": 0.48888182640075684, "learning_rate": 8.307923422434852e-05, "loss": 0.0605, "step": 30640 }, { "epoch": 1.0362081206261198, "grad_norm": 0.24198013544082642, "learning_rate": 8.30668335341159e-05, "loss": 0.056, "step": 30650 }, { "epoch": 1.036546198316373, "grad_norm": 0.306094229221344, "learning_rate": 8.305442922774408e-05, "loss": 0.057, "step": 30660 }, { "epoch": 1.0368842760066264, "grad_norm": 0.195087268948555, "learning_rate": 8.304202130658959e-05, "loss": 0.0651, "step": 30670 }, { "epoch": 1.0372223536968797, "grad_norm": 0.23395191133022308, "learning_rate": 8.302960977200931e-05, "loss": 0.0546, "step": 30680 }, { "epoch": 1.0375604313871327, "grad_norm": 0.21878044307231903, "learning_rate": 8.301719462536058e-05, "loss": 0.0549, "step": 30690 }, { "epoch": 1.037898509077386, "grad_norm": 0.2347203940153122, "learning_rate": 8.300477586800108e-05, "loss": 0.056, "step": 30700 }, { "epoch": 1.0382365867676393, "grad_norm": 0.221777081489563, "learning_rate": 8.299235350128892e-05, "loss": 0.056, "step": 30710 }, { "epoch": 1.0385746644578924, "grad_norm": 0.19730053842067719, "learning_rate": 8.297992752658258e-05, "loss": 0.0612, "step": 30720 }, { "epoch": 1.0389127421481457, "grad_norm": 0.22856998443603516, "learning_rate": 8.296749794524094e-05, "loss": 0.0597, "step": 30730 }, { "epoch": 1.039250819838399, "grad_norm": 0.3479428291320801, "learning_rate": 8.295506475862332e-05, "loss": 0.0551, "step": 30740 }, { "epoch": 1.039588897528652, "grad_norm": 0.20436257123947144, "learning_rate": 8.294262796808933e-05, "loss": 0.0588, "step": 30750 }, { "epoch": 1.0399269752189053, "grad_norm": 0.19106614589691162, "learning_rate": 8.293018757499909e-05, "loss": 0.0574, "step": 30760 }, { "epoch": 1.0402650529091586, "grad_norm": 0.17349481582641602, "learning_rate": 8.291774358071305e-05, "loss": 0.0585, "step": 30770 }, { "epoch": 1.0406031305994117, "grad_norm": 0.4163411855697632, "learning_rate": 8.290529598659205e-05, "loss": 0.0538, "step": 30780 }, { "epoch": 1.040941208289665, "grad_norm": 0.2661783993244171, "learning_rate": 8.289284479399738e-05, "loss": 0.0537, "step": 30790 }, { "epoch": 1.0412792859799183, "grad_norm": 0.2708529233932495, "learning_rate": 8.288039000429064e-05, "loss": 0.0549, "step": 30800 }, { "epoch": 1.0416173636701713, "grad_norm": 0.25650259852409363, "learning_rate": 8.286793161883388e-05, "loss": 0.0541, "step": 30810 }, { "epoch": 1.0419554413604246, "grad_norm": 0.32846206426620483, "learning_rate": 8.285546963898954e-05, "loss": 0.0588, "step": 30820 }, { "epoch": 1.042293519050678, "grad_norm": 0.26338133215904236, "learning_rate": 8.284300406612044e-05, "loss": 0.0564, "step": 30830 }, { "epoch": 1.042631596740931, "grad_norm": 0.2836932837963104, "learning_rate": 8.283053490158978e-05, "loss": 0.0561, "step": 30840 }, { "epoch": 1.0429696744311843, "grad_norm": 0.19523917138576508, "learning_rate": 8.28180621467612e-05, "loss": 0.0569, "step": 30850 }, { "epoch": 1.0433077521214376, "grad_norm": 0.2524062991142273, "learning_rate": 8.280558580299867e-05, "loss": 0.0535, "step": 30860 }, { "epoch": 1.0436458298116906, "grad_norm": 0.2896188795566559, "learning_rate": 8.279310587166661e-05, "loss": 0.0579, "step": 30870 }, { "epoch": 1.043983907501944, "grad_norm": 0.2057177722454071, "learning_rate": 8.278062235412978e-05, "loss": 0.0603, "step": 30880 }, { "epoch": 1.0443219851921972, "grad_norm": 0.29478299617767334, "learning_rate": 8.276813525175339e-05, "loss": 0.0592, "step": 30890 }, { "epoch": 1.0446600628824503, "grad_norm": 0.16743747889995575, "learning_rate": 8.2755644565903e-05, "loss": 0.0603, "step": 30900 }, { "epoch": 1.0449981405727036, "grad_norm": 0.28074371814727783, "learning_rate": 8.274315029794454e-05, "loss": 0.0571, "step": 30910 }, { "epoch": 1.0453362182629569, "grad_norm": 0.255472332239151, "learning_rate": 8.273065244924443e-05, "loss": 0.0539, "step": 30920 }, { "epoch": 1.0456742959532102, "grad_norm": 0.1744445562362671, "learning_rate": 8.271815102116936e-05, "loss": 0.0553, "step": 30930 }, { "epoch": 1.0460123736434632, "grad_norm": 0.27736812829971313, "learning_rate": 8.270564601508648e-05, "loss": 0.0574, "step": 30940 }, { "epoch": 1.0463504513337165, "grad_norm": 0.23908917605876923, "learning_rate": 8.269313743236333e-05, "loss": 0.0598, "step": 30950 }, { "epoch": 1.0466885290239698, "grad_norm": 0.29007866978645325, "learning_rate": 8.268062527436783e-05, "loss": 0.0513, "step": 30960 }, { "epoch": 1.0470266067142229, "grad_norm": 0.32702022790908813, "learning_rate": 8.266810954246828e-05, "loss": 0.051, "step": 30970 }, { "epoch": 1.0473646844044762, "grad_norm": 0.18675345182418823, "learning_rate": 8.265559023803338e-05, "loss": 0.053, "step": 30980 }, { "epoch": 1.0477027620947295, "grad_norm": 0.23919904232025146, "learning_rate": 8.264306736243223e-05, "loss": 0.0524, "step": 30990 }, { "epoch": 1.0480408397849825, "grad_norm": 0.30619803071022034, "learning_rate": 8.263054091703432e-05, "loss": 0.0534, "step": 31000 }, { "epoch": 1.0483789174752358, "grad_norm": 0.33538252115249634, "learning_rate": 8.26180109032095e-05, "loss": 0.0544, "step": 31010 }, { "epoch": 1.048716995165489, "grad_norm": 0.29679805040359497, "learning_rate": 8.260547732232807e-05, "loss": 0.0581, "step": 31020 }, { "epoch": 1.0490550728557422, "grad_norm": 0.4268900454044342, "learning_rate": 8.259294017576063e-05, "loss": 0.0591, "step": 31030 }, { "epoch": 1.0493931505459955, "grad_norm": 0.20972375571727753, "learning_rate": 8.258039946487828e-05, "loss": 0.0583, "step": 31040 }, { "epoch": 1.0497312282362488, "grad_norm": 0.3013782501220703, "learning_rate": 8.256785519105241e-05, "loss": 0.0589, "step": 31050 }, { "epoch": 1.0500693059265018, "grad_norm": 0.2807738780975342, "learning_rate": 8.255530735565488e-05, "loss": 0.065, "step": 31060 }, { "epoch": 1.0504073836167551, "grad_norm": 0.2196795791387558, "learning_rate": 8.254275596005787e-05, "loss": 0.051, "step": 31070 }, { "epoch": 1.0507454613070084, "grad_norm": 0.20163580775260925, "learning_rate": 8.2530201005634e-05, "loss": 0.059, "step": 31080 }, { "epoch": 1.0510835389972615, "grad_norm": 0.2351241260766983, "learning_rate": 8.251764249375626e-05, "loss": 0.054, "step": 31090 }, { "epoch": 1.0514216166875148, "grad_norm": 0.4377899765968323, "learning_rate": 8.250508042579803e-05, "loss": 0.0521, "step": 31100 }, { "epoch": 1.051759694377768, "grad_norm": 0.27140119671821594, "learning_rate": 8.249251480313307e-05, "loss": 0.0561, "step": 31110 }, { "epoch": 1.0520977720680211, "grad_norm": 0.3360067903995514, "learning_rate": 8.247994562713555e-05, "loss": 0.0564, "step": 31120 }, { "epoch": 1.0524358497582744, "grad_norm": 0.4189335107803345, "learning_rate": 8.246737289917998e-05, "loss": 0.0589, "step": 31130 }, { "epoch": 1.0527739274485277, "grad_norm": 0.22891020774841309, "learning_rate": 8.245479662064135e-05, "loss": 0.0546, "step": 31140 }, { "epoch": 1.053112005138781, "grad_norm": 0.33072414994239807, "learning_rate": 8.244221679289496e-05, "loss": 0.0541, "step": 31150 }, { "epoch": 1.053450082829034, "grad_norm": 0.29684337973594666, "learning_rate": 8.242963341731652e-05, "loss": 0.0617, "step": 31160 }, { "epoch": 1.0537881605192874, "grad_norm": 0.2578258216381073, "learning_rate": 8.24170464952821e-05, "loss": 0.0583, "step": 31170 }, { "epoch": 1.0541262382095407, "grad_norm": 0.27093860507011414, "learning_rate": 8.240445602816824e-05, "loss": 0.0587, "step": 31180 }, { "epoch": 1.0544643158997937, "grad_norm": 0.4010903835296631, "learning_rate": 8.239186201735179e-05, "loss": 0.0617, "step": 31190 }, { "epoch": 1.054802393590047, "grad_norm": 0.22918859124183655, "learning_rate": 8.237926446420998e-05, "loss": 0.061, "step": 31200 }, { "epoch": 1.0551404712803003, "grad_norm": 0.34324631094932556, "learning_rate": 8.236666337012052e-05, "loss": 0.0562, "step": 31210 }, { "epoch": 1.0554785489705534, "grad_norm": 0.4086182713508606, "learning_rate": 8.23540587364614e-05, "loss": 0.0581, "step": 31220 }, { "epoch": 1.0558166266608067, "grad_norm": 0.5261779427528381, "learning_rate": 8.234145056461107e-05, "loss": 0.0605, "step": 31230 }, { "epoch": 1.05615470435106, "grad_norm": 0.27415934205055237, "learning_rate": 8.232883885594832e-05, "loss": 0.0518, "step": 31240 }, { "epoch": 1.056492782041313, "grad_norm": 0.35709497332572937, "learning_rate": 8.231622361185236e-05, "loss": 0.0609, "step": 31250 }, { "epoch": 1.0568308597315663, "grad_norm": 0.2316531538963318, "learning_rate": 8.230360483370278e-05, "loss": 0.0544, "step": 31260 }, { "epoch": 1.0571689374218196, "grad_norm": 0.2146332710981369, "learning_rate": 8.229098252287953e-05, "loss": 0.0547, "step": 31270 }, { "epoch": 1.0575070151120727, "grad_norm": 0.2430523782968521, "learning_rate": 8.2278356680763e-05, "loss": 0.0504, "step": 31280 }, { "epoch": 1.057845092802326, "grad_norm": 0.32912543416023254, "learning_rate": 8.22657273087339e-05, "loss": 0.0579, "step": 31290 }, { "epoch": 1.0581831704925793, "grad_norm": 0.2584492862224579, "learning_rate": 8.225309440817336e-05, "loss": 0.0646, "step": 31300 }, { "epoch": 1.0585212481828323, "grad_norm": 0.25136685371398926, "learning_rate": 8.224045798046293e-05, "loss": 0.0613, "step": 31310 }, { "epoch": 1.0588593258730856, "grad_norm": 0.21091529726982117, "learning_rate": 8.22278180269845e-05, "loss": 0.0576, "step": 31320 }, { "epoch": 1.059197403563339, "grad_norm": 0.2730776071548462, "learning_rate": 8.221517454912031e-05, "loss": 0.0581, "step": 31330 }, { "epoch": 1.059535481253592, "grad_norm": 0.232318714261055, "learning_rate": 8.220252754825308e-05, "loss": 0.0574, "step": 31340 }, { "epoch": 1.0598735589438453, "grad_norm": 0.260644793510437, "learning_rate": 8.218987702576586e-05, "loss": 0.057, "step": 31350 }, { "epoch": 1.0602116366340986, "grad_norm": 0.2646980285644531, "learning_rate": 8.217722298304207e-05, "loss": 0.0593, "step": 31360 }, { "epoch": 1.0605497143243516, "grad_norm": 0.20936952531337738, "learning_rate": 8.216456542146557e-05, "loss": 0.059, "step": 31370 }, { "epoch": 1.060887792014605, "grad_norm": 0.5886946320533752, "learning_rate": 8.215190434242055e-05, "loss": 0.0561, "step": 31380 }, { "epoch": 1.0612258697048582, "grad_norm": 0.19989784061908722, "learning_rate": 8.213923974729161e-05, "loss": 0.0563, "step": 31390 }, { "epoch": 1.0615639473951113, "grad_norm": 0.24776850640773773, "learning_rate": 8.212657163746373e-05, "loss": 0.0541, "step": 31400 }, { "epoch": 1.0619020250853646, "grad_norm": 0.34436509013175964, "learning_rate": 8.211390001432227e-05, "loss": 0.0674, "step": 31410 }, { "epoch": 1.0622401027756179, "grad_norm": 0.2808530628681183, "learning_rate": 8.210122487925297e-05, "loss": 0.0565, "step": 31420 }, { "epoch": 1.0625781804658712, "grad_norm": 0.2650662064552307, "learning_rate": 8.208854623364202e-05, "loss": 0.0591, "step": 31430 }, { "epoch": 1.0629162581561242, "grad_norm": 0.2245766818523407, "learning_rate": 8.207586407887589e-05, "loss": 0.0497, "step": 31440 }, { "epoch": 1.0632543358463775, "grad_norm": 0.1943754404783249, "learning_rate": 8.206317841634148e-05, "loss": 0.0516, "step": 31450 }, { "epoch": 1.0635924135366308, "grad_norm": 0.23897802829742432, "learning_rate": 8.205048924742609e-05, "loss": 0.058, "step": 31460 }, { "epoch": 1.0639304912268839, "grad_norm": 0.2458626925945282, "learning_rate": 8.203779657351738e-05, "loss": 0.0562, "step": 31470 }, { "epoch": 1.0642685689171372, "grad_norm": 0.1798367202281952, "learning_rate": 8.20251003960034e-05, "loss": 0.056, "step": 31480 }, { "epoch": 1.0646066466073905, "grad_norm": 0.24192999303340912, "learning_rate": 8.201240071627258e-05, "loss": 0.0541, "step": 31490 }, { "epoch": 1.0649447242976435, "grad_norm": 0.2043362259864807, "learning_rate": 8.199969753571377e-05, "loss": 0.0527, "step": 31500 }, { "epoch": 1.0652828019878968, "grad_norm": 0.26901328563690186, "learning_rate": 8.198699085571615e-05, "loss": 0.0569, "step": 31510 }, { "epoch": 1.0656208796781501, "grad_norm": 0.2571665048599243, "learning_rate": 8.197428067766928e-05, "loss": 0.0533, "step": 31520 }, { "epoch": 1.0659589573684032, "grad_norm": 0.26625698804855347, "learning_rate": 8.196156700296316e-05, "loss": 0.064, "step": 31530 }, { "epoch": 1.0662970350586565, "grad_norm": 0.19997094571590424, "learning_rate": 8.194884983298814e-05, "loss": 0.0565, "step": 31540 }, { "epoch": 1.0666351127489098, "grad_norm": 0.22635948657989502, "learning_rate": 8.193612916913491e-05, "loss": 0.0586, "step": 31550 }, { "epoch": 1.0669731904391628, "grad_norm": 0.2650618255138397, "learning_rate": 8.192340501279463e-05, "loss": 0.0544, "step": 31560 }, { "epoch": 1.0673112681294161, "grad_norm": 0.22157081961631775, "learning_rate": 8.191067736535876e-05, "loss": 0.0567, "step": 31570 }, { "epoch": 1.0676493458196694, "grad_norm": 0.2138378620147705, "learning_rate": 8.18979462282192e-05, "loss": 0.0526, "step": 31580 }, { "epoch": 1.0679874235099225, "grad_norm": 0.16685393452644348, "learning_rate": 8.188521160276819e-05, "loss": 0.0575, "step": 31590 }, { "epoch": 1.0683255012001758, "grad_norm": 0.3618558645248413, "learning_rate": 8.187247349039837e-05, "loss": 0.0493, "step": 31600 }, { "epoch": 1.068663578890429, "grad_norm": 0.2689863443374634, "learning_rate": 8.185973189250278e-05, "loss": 0.0553, "step": 31610 }, { "epoch": 1.0690016565806824, "grad_norm": 0.1937102973461151, "learning_rate": 8.184698681047482e-05, "loss": 0.052, "step": 31620 }, { "epoch": 1.0693397342709354, "grad_norm": 0.13429434597492218, "learning_rate": 8.183423824570827e-05, "loss": 0.0514, "step": 31630 }, { "epoch": 1.0696778119611887, "grad_norm": 0.30382996797561646, "learning_rate": 8.182148619959725e-05, "loss": 0.0583, "step": 31640 }, { "epoch": 1.070015889651442, "grad_norm": 0.3473210334777832, "learning_rate": 8.180873067353636e-05, "loss": 0.0553, "step": 31650 }, { "epoch": 1.070353967341695, "grad_norm": 0.2879530191421509, "learning_rate": 8.179597166892052e-05, "loss": 0.0568, "step": 31660 }, { "epoch": 1.0706920450319484, "grad_norm": 0.2623620927333832, "learning_rate": 8.178320918714501e-05, "loss": 0.0571, "step": 31670 }, { "epoch": 1.0710301227222017, "grad_norm": 0.41011738777160645, "learning_rate": 8.177044322960554e-05, "loss": 0.056, "step": 31680 }, { "epoch": 1.0713682004124547, "grad_norm": 0.25561633706092834, "learning_rate": 8.175767379769816e-05, "loss": 0.0562, "step": 31690 }, { "epoch": 1.071706278102708, "grad_norm": 0.2170923501253128, "learning_rate": 8.174490089281932e-05, "loss": 0.0636, "step": 31700 }, { "epoch": 1.0720443557929613, "grad_norm": 0.3915815055370331, "learning_rate": 8.173212451636584e-05, "loss": 0.0561, "step": 31710 }, { "epoch": 1.0723824334832144, "grad_norm": 0.21967163681983948, "learning_rate": 8.171934466973493e-05, "loss": 0.0527, "step": 31720 }, { "epoch": 1.0727205111734677, "grad_norm": 0.2199716717004776, "learning_rate": 8.170656135432418e-05, "loss": 0.0582, "step": 31730 }, { "epoch": 1.073058588863721, "grad_norm": 0.35515061020851135, "learning_rate": 8.169377457153155e-05, "loss": 0.0599, "step": 31740 }, { "epoch": 1.073396666553974, "grad_norm": 0.21048623323440552, "learning_rate": 8.168098432275539e-05, "loss": 0.054, "step": 31750 }, { "epoch": 1.0737347442442273, "grad_norm": 0.16504009068012238, "learning_rate": 8.166819060939442e-05, "loss": 0.0518, "step": 31760 }, { "epoch": 1.0740728219344806, "grad_norm": 0.3614059388637543, "learning_rate": 8.165539343284772e-05, "loss": 0.0622, "step": 31770 }, { "epoch": 1.0744108996247337, "grad_norm": 0.26059818267822266, "learning_rate": 8.16425927945148e-05, "loss": 0.0554, "step": 31780 }, { "epoch": 1.074748977314987, "grad_norm": 0.3459230959415436, "learning_rate": 8.162978869579551e-05, "loss": 0.0535, "step": 31790 }, { "epoch": 1.0750870550052403, "grad_norm": 0.31423091888427734, "learning_rate": 8.161698113809007e-05, "loss": 0.0591, "step": 31800 }, { "epoch": 1.0754251326954933, "grad_norm": 0.34614861011505127, "learning_rate": 8.160417012279911e-05, "loss": 0.0508, "step": 31810 }, { "epoch": 1.0757632103857466, "grad_norm": 0.24332350492477417, "learning_rate": 8.159135565132363e-05, "loss": 0.0592, "step": 31820 }, { "epoch": 1.076101288076, "grad_norm": 0.23227092623710632, "learning_rate": 8.157853772506498e-05, "loss": 0.0605, "step": 31830 }, { "epoch": 1.076439365766253, "grad_norm": 0.18496157228946686, "learning_rate": 8.156571634542494e-05, "loss": 0.0561, "step": 31840 }, { "epoch": 1.0767774434565063, "grad_norm": 0.23095856606960297, "learning_rate": 8.15528915138056e-05, "loss": 0.061, "step": 31850 }, { "epoch": 1.0771155211467596, "grad_norm": 0.3532733619213104, "learning_rate": 8.154006323160949e-05, "loss": 0.0582, "step": 31860 }, { "epoch": 1.0774535988370126, "grad_norm": 0.29202181100845337, "learning_rate": 8.152723150023949e-05, "loss": 0.0574, "step": 31870 }, { "epoch": 1.077791676527266, "grad_norm": 0.2807808220386505, "learning_rate": 8.151439632109886e-05, "loss": 0.0541, "step": 31880 }, { "epoch": 1.0781297542175192, "grad_norm": 0.3227086663246155, "learning_rate": 8.150155769559122e-05, "loss": 0.0589, "step": 31890 }, { "epoch": 1.0784678319077723, "grad_norm": 0.1951363980770111, "learning_rate": 8.148871562512058e-05, "loss": 0.0545, "step": 31900 }, { "epoch": 1.0788059095980256, "grad_norm": 0.2686103284358978, "learning_rate": 8.147587011109136e-05, "loss": 0.0535, "step": 31910 }, { "epoch": 1.0791439872882789, "grad_norm": 0.3926309645175934, "learning_rate": 8.14630211549083e-05, "loss": 0.0545, "step": 31920 }, { "epoch": 1.0794820649785322, "grad_norm": 0.42142850160598755, "learning_rate": 8.145016875797655e-05, "loss": 0.0554, "step": 31930 }, { "epoch": 1.0798201426687852, "grad_norm": 0.246910959482193, "learning_rate": 8.143731292170164e-05, "loss": 0.0565, "step": 31940 }, { "epoch": 1.0801582203590385, "grad_norm": 0.18926702439785004, "learning_rate": 8.142445364748944e-05, "loss": 0.0614, "step": 31950 }, { "epoch": 1.0804962980492918, "grad_norm": 0.2027309685945511, "learning_rate": 8.141159093674624e-05, "loss": 0.0514, "step": 31960 }, { "epoch": 1.0808343757395449, "grad_norm": 0.17181678116321564, "learning_rate": 8.139872479087869e-05, "loss": 0.0568, "step": 31970 }, { "epoch": 1.0811724534297982, "grad_norm": 0.20810455083847046, "learning_rate": 8.13858552112938e-05, "loss": 0.0506, "step": 31980 }, { "epoch": 1.0815105311200515, "grad_norm": 0.2696342170238495, "learning_rate": 8.137298219939895e-05, "loss": 0.0558, "step": 31990 }, { "epoch": 1.0818486088103045, "grad_norm": 0.31295520067214966, "learning_rate": 8.136010575660196e-05, "loss": 0.0573, "step": 32000 }, { "epoch": 1.0821866865005578, "grad_norm": 0.33707213401794434, "learning_rate": 8.134722588431095e-05, "loss": 0.0543, "step": 32010 }, { "epoch": 1.0825247641908111, "grad_norm": 0.2559635043144226, "learning_rate": 8.133434258393444e-05, "loss": 0.0614, "step": 32020 }, { "epoch": 1.0828628418810642, "grad_norm": 0.2238919883966446, "learning_rate": 8.132145585688134e-05, "loss": 0.0603, "step": 32030 }, { "epoch": 1.0832009195713175, "grad_norm": 0.233486145734787, "learning_rate": 8.130856570456093e-05, "loss": 0.0537, "step": 32040 }, { "epoch": 1.0835389972615708, "grad_norm": 0.22335779666900635, "learning_rate": 8.129567212838283e-05, "loss": 0.0608, "step": 32050 }, { "epoch": 1.0838770749518238, "grad_norm": 0.2831674814224243, "learning_rate": 8.128277512975708e-05, "loss": 0.0549, "step": 32060 }, { "epoch": 1.0842151526420771, "grad_norm": 0.18280720710754395, "learning_rate": 8.126987471009408e-05, "loss": 0.0509, "step": 32070 }, { "epoch": 1.0845532303323304, "grad_norm": 0.26704326272010803, "learning_rate": 8.125697087080459e-05, "loss": 0.0544, "step": 32080 }, { "epoch": 1.0848913080225835, "grad_norm": 0.24323305487632751, "learning_rate": 8.124406361329976e-05, "loss": 0.0546, "step": 32090 }, { "epoch": 1.0852293857128368, "grad_norm": 0.30920371413230896, "learning_rate": 8.123115293899111e-05, "loss": 0.0542, "step": 32100 }, { "epoch": 1.08556746340309, "grad_norm": 0.180341899394989, "learning_rate": 8.121823884929055e-05, "loss": 0.0597, "step": 32110 }, { "epoch": 1.0859055410933434, "grad_norm": 0.27669697999954224, "learning_rate": 8.12053213456103e-05, "loss": 0.0544, "step": 32120 }, { "epoch": 1.0862436187835964, "grad_norm": 0.20944854617118835, "learning_rate": 8.119240042936303e-05, "loss": 0.052, "step": 32130 }, { "epoch": 1.0865816964738497, "grad_norm": 0.2321566939353943, "learning_rate": 8.117947610196175e-05, "loss": 0.0535, "step": 32140 }, { "epoch": 1.086919774164103, "grad_norm": 0.2522382438182831, "learning_rate": 8.116654836481982e-05, "loss": 0.053, "step": 32150 }, { "epoch": 1.087257851854356, "grad_norm": 0.2554924786090851, "learning_rate": 8.115361721935106e-05, "loss": 0.0563, "step": 32160 }, { "epoch": 1.0875959295446094, "grad_norm": 0.314400315284729, "learning_rate": 8.114068266696953e-05, "loss": 0.0535, "step": 32170 }, { "epoch": 1.0879340072348627, "grad_norm": 0.20788440108299255, "learning_rate": 8.112774470908978e-05, "loss": 0.0603, "step": 32180 }, { "epoch": 1.0882720849251157, "grad_norm": 0.33038321137428284, "learning_rate": 8.111480334712665e-05, "loss": 0.0512, "step": 32190 }, { "epoch": 1.088610162615369, "grad_norm": 0.23857229948043823, "learning_rate": 8.110185858249542e-05, "loss": 0.0558, "step": 32200 }, { "epoch": 1.0889482403056223, "grad_norm": 0.26391616463661194, "learning_rate": 8.108891041661168e-05, "loss": 0.0531, "step": 32210 }, { "epoch": 1.0892863179958754, "grad_norm": 0.3141528069972992, "learning_rate": 8.107595885089146e-05, "loss": 0.0538, "step": 32220 }, { "epoch": 1.0896243956861287, "grad_norm": 0.24000675976276398, "learning_rate": 8.106300388675108e-05, "loss": 0.0499, "step": 32230 }, { "epoch": 1.089962473376382, "grad_norm": 0.35347700119018555, "learning_rate": 8.105004552560731e-05, "loss": 0.062, "step": 32240 }, { "epoch": 1.090300551066635, "grad_norm": 0.19143764674663544, "learning_rate": 8.103708376887724e-05, "loss": 0.0545, "step": 32250 }, { "epoch": 1.0906386287568883, "grad_norm": 0.4377743899822235, "learning_rate": 8.102411861797836e-05, "loss": 0.0547, "step": 32260 }, { "epoch": 1.0909767064471416, "grad_norm": 0.26839420199394226, "learning_rate": 8.101115007432851e-05, "loss": 0.0584, "step": 32270 }, { "epoch": 1.0913147841373947, "grad_norm": 0.2796750068664551, "learning_rate": 8.09981781393459e-05, "loss": 0.0546, "step": 32280 }, { "epoch": 1.091652861827648, "grad_norm": 0.20809637010097504, "learning_rate": 8.098520281444915e-05, "loss": 0.0548, "step": 32290 }, { "epoch": 1.0919909395179013, "grad_norm": 0.2844539284706116, "learning_rate": 8.09722241010572e-05, "loss": 0.0563, "step": 32300 }, { "epoch": 1.0923290172081543, "grad_norm": 0.21541066467761993, "learning_rate": 8.095924200058939e-05, "loss": 0.0591, "step": 32310 }, { "epoch": 1.0926670948984076, "grad_norm": 0.24495577812194824, "learning_rate": 8.094625651446541e-05, "loss": 0.0566, "step": 32320 }, { "epoch": 1.093005172588661, "grad_norm": 0.28771522641181946, "learning_rate": 8.093326764410536e-05, "loss": 0.0566, "step": 32330 }, { "epoch": 1.093343250278914, "grad_norm": 0.15804757177829742, "learning_rate": 8.092027539092966e-05, "loss": 0.0568, "step": 32340 }, { "epoch": 1.0936813279691673, "grad_norm": 0.2579987645149231, "learning_rate": 8.090727975635913e-05, "loss": 0.0587, "step": 32350 }, { "epoch": 1.0940194056594206, "grad_norm": 0.3637463450431824, "learning_rate": 8.089428074181497e-05, "loss": 0.053, "step": 32360 }, { "epoch": 1.0943574833496736, "grad_norm": 0.220474511384964, "learning_rate": 8.088127834871871e-05, "loss": 0.0557, "step": 32370 }, { "epoch": 1.094695561039927, "grad_norm": 0.2456025779247284, "learning_rate": 8.086827257849226e-05, "loss": 0.0586, "step": 32380 }, { "epoch": 1.0950336387301802, "grad_norm": 0.2848506569862366, "learning_rate": 8.085526343255795e-05, "loss": 0.0562, "step": 32390 }, { "epoch": 1.0953717164204335, "grad_norm": 0.25993627309799194, "learning_rate": 8.084225091233842e-05, "loss": 0.0587, "step": 32400 }, { "epoch": 1.0957097941106866, "grad_norm": 0.20490212738513947, "learning_rate": 8.082923501925668e-05, "loss": 0.0599, "step": 32410 }, { "epoch": 1.0960478718009399, "grad_norm": 0.30517569184303284, "learning_rate": 8.081621575473617e-05, "loss": 0.0642, "step": 32420 }, { "epoch": 1.0963859494911932, "grad_norm": 0.19974297285079956, "learning_rate": 8.080319312020064e-05, "loss": 0.0571, "step": 32430 }, { "epoch": 1.0967240271814462, "grad_norm": 0.2167779505252838, "learning_rate": 8.079016711707421e-05, "loss": 0.056, "step": 32440 }, { "epoch": 1.0970621048716995, "grad_norm": 0.2493787705898285, "learning_rate": 8.077713774678139e-05, "loss": 0.0535, "step": 32450 }, { "epoch": 1.0974001825619528, "grad_norm": 0.3003794550895691, "learning_rate": 8.076410501074707e-05, "loss": 0.054, "step": 32460 }, { "epoch": 1.0977382602522059, "grad_norm": 0.24080343544483185, "learning_rate": 8.075106891039647e-05, "loss": 0.0547, "step": 32470 }, { "epoch": 1.0980763379424592, "grad_norm": 0.2773653566837311, "learning_rate": 8.073802944715523e-05, "loss": 0.0582, "step": 32480 }, { "epoch": 1.0984144156327125, "grad_norm": 0.30073288083076477, "learning_rate": 8.072498662244929e-05, "loss": 0.0581, "step": 32490 }, { "epoch": 1.0987524933229655, "grad_norm": 0.35650140047073364, "learning_rate": 8.0711940437705e-05, "loss": 0.0587, "step": 32500 }, { "epoch": 1.0990905710132188, "grad_norm": 0.16216668486595154, "learning_rate": 8.06988908943491e-05, "loss": 0.0572, "step": 32510 }, { "epoch": 1.0994286487034721, "grad_norm": 0.2902977168560028, "learning_rate": 8.068583799380863e-05, "loss": 0.0543, "step": 32520 }, { "epoch": 1.0997667263937252, "grad_norm": 0.3109583854675293, "learning_rate": 8.067278173751104e-05, "loss": 0.0599, "step": 32530 }, { "epoch": 1.1001048040839785, "grad_norm": 0.24490559101104736, "learning_rate": 8.065972212688417e-05, "loss": 0.0583, "step": 32540 }, { "epoch": 1.1004428817742318, "grad_norm": 0.21703854203224182, "learning_rate": 8.064665916335618e-05, "loss": 0.0481, "step": 32550 }, { "epoch": 1.1007809594644848, "grad_norm": 0.19497297704219818, "learning_rate": 8.063359284835564e-05, "loss": 0.0574, "step": 32560 }, { "epoch": 1.1011190371547381, "grad_norm": 0.18779288232326508, "learning_rate": 8.062052318331142e-05, "loss": 0.0621, "step": 32570 }, { "epoch": 1.1014571148449914, "grad_norm": 0.3593255877494812, "learning_rate": 8.060745016965283e-05, "loss": 0.0531, "step": 32580 }, { "epoch": 1.1017951925352447, "grad_norm": 0.2162729948759079, "learning_rate": 8.059437380880952e-05, "loss": 0.0541, "step": 32590 }, { "epoch": 1.1021332702254978, "grad_norm": 0.2544594705104828, "learning_rate": 8.058129410221146e-05, "loss": 0.0587, "step": 32600 }, { "epoch": 1.102471347915751, "grad_norm": 0.20709790289402008, "learning_rate": 8.056821105128908e-05, "loss": 0.0523, "step": 32610 }, { "epoch": 1.1028094256060044, "grad_norm": 0.4142126441001892, "learning_rate": 8.05551246574731e-05, "loss": 0.0569, "step": 32620 }, { "epoch": 1.1031475032962574, "grad_norm": 0.2926779091358185, "learning_rate": 8.05420349221946e-05, "loss": 0.0606, "step": 32630 }, { "epoch": 1.1034855809865107, "grad_norm": 0.8847958445549011, "learning_rate": 8.05289418468851e-05, "loss": 0.0534, "step": 32640 }, { "epoch": 1.103823658676764, "grad_norm": 0.23706549406051636, "learning_rate": 8.051584543297642e-05, "loss": 0.0534, "step": 32650 }, { "epoch": 1.104161736367017, "grad_norm": 0.25338485836982727, "learning_rate": 8.050274568190074e-05, "loss": 0.0527, "step": 32660 }, { "epoch": 1.1044998140572704, "grad_norm": 0.26438090205192566, "learning_rate": 8.048964259509067e-05, "loss": 0.0582, "step": 32670 }, { "epoch": 1.1048378917475237, "grad_norm": 0.20469819009304047, "learning_rate": 8.047653617397914e-05, "loss": 0.0566, "step": 32680 }, { "epoch": 1.1051759694377767, "grad_norm": 0.41622474789619446, "learning_rate": 8.046342641999941e-05, "loss": 0.0527, "step": 32690 }, { "epoch": 1.10551404712803, "grad_norm": 0.40054744482040405, "learning_rate": 8.045031333458517e-05, "loss": 0.052, "step": 32700 }, { "epoch": 1.1058521248182833, "grad_norm": 0.34032273292541504, "learning_rate": 8.043719691917047e-05, "loss": 0.06, "step": 32710 }, { "epoch": 1.1061902025085364, "grad_norm": 0.28473687171936035, "learning_rate": 8.042407717518966e-05, "loss": 0.0611, "step": 32720 }, { "epoch": 1.1065282801987897, "grad_norm": 0.19117262959480286, "learning_rate": 8.041095410407751e-05, "loss": 0.0591, "step": 32730 }, { "epoch": 1.106866357889043, "grad_norm": 0.2540346682071686, "learning_rate": 8.039782770726913e-05, "loss": 0.0575, "step": 32740 }, { "epoch": 1.107204435579296, "grad_norm": 0.3887478709220886, "learning_rate": 8.038469798620004e-05, "loss": 0.0551, "step": 32750 }, { "epoch": 1.1075425132695493, "grad_norm": 0.18887533247470856, "learning_rate": 8.037156494230604e-05, "loss": 0.0472, "step": 32760 }, { "epoch": 1.1078805909598026, "grad_norm": 0.30410444736480713, "learning_rate": 8.035842857702338e-05, "loss": 0.0554, "step": 32770 }, { "epoch": 1.1082186686500557, "grad_norm": 0.3062264621257782, "learning_rate": 8.03452888917886e-05, "loss": 0.0529, "step": 32780 }, { "epoch": 1.108556746340309, "grad_norm": 0.21994464099407196, "learning_rate": 8.033214588803866e-05, "loss": 0.0543, "step": 32790 }, { "epoch": 1.1088948240305623, "grad_norm": 0.4183480441570282, "learning_rate": 8.031899956721083e-05, "loss": 0.0484, "step": 32800 }, { "epoch": 1.1092329017208153, "grad_norm": 0.28903383016586304, "learning_rate": 8.030584993074282e-05, "loss": 0.0541, "step": 32810 }, { "epoch": 1.1095709794110686, "grad_norm": 0.31903669238090515, "learning_rate": 8.02926969800726e-05, "loss": 0.0545, "step": 32820 }, { "epoch": 1.109909057101322, "grad_norm": 0.25246021151542664, "learning_rate": 8.027954071663859e-05, "loss": 0.0541, "step": 32830 }, { "epoch": 1.110247134791575, "grad_norm": 0.3903939425945282, "learning_rate": 8.026638114187954e-05, "loss": 0.0584, "step": 32840 }, { "epoch": 1.1105852124818283, "grad_norm": 0.25615519285202026, "learning_rate": 8.025321825723456e-05, "loss": 0.061, "step": 32850 }, { "epoch": 1.1109232901720816, "grad_norm": 0.29751765727996826, "learning_rate": 8.02400520641431e-05, "loss": 0.0591, "step": 32860 }, { "epoch": 1.1112613678623346, "grad_norm": 0.2565911114215851, "learning_rate": 8.022688256404501e-05, "loss": 0.0581, "step": 32870 }, { "epoch": 1.111599445552588, "grad_norm": 0.31566134095191956, "learning_rate": 8.02137097583805e-05, "loss": 0.0551, "step": 32880 }, { "epoch": 1.1119375232428412, "grad_norm": 0.30602559447288513, "learning_rate": 8.02005336485901e-05, "loss": 0.055, "step": 32890 }, { "epoch": 1.1122756009330945, "grad_norm": 0.2322775274515152, "learning_rate": 8.018735423611476e-05, "loss": 0.0648, "step": 32900 }, { "epoch": 1.1126136786233476, "grad_norm": 0.29900646209716797, "learning_rate": 8.017417152239574e-05, "loss": 0.057, "step": 32910 }, { "epoch": 1.1129517563136009, "grad_norm": 0.34871265292167664, "learning_rate": 8.01609855088747e-05, "loss": 0.0598, "step": 32920 }, { "epoch": 1.1132898340038542, "grad_norm": 0.39051494002342224, "learning_rate": 8.014779619699362e-05, "loss": 0.0583, "step": 32930 }, { "epoch": 1.1136279116941072, "grad_norm": 0.2253064662218094, "learning_rate": 8.013460358819489e-05, "loss": 0.0535, "step": 32940 }, { "epoch": 1.1139659893843605, "grad_norm": 0.2935599386692047, "learning_rate": 8.01214076839212e-05, "loss": 0.0564, "step": 32950 }, { "epoch": 1.1143040670746138, "grad_norm": 0.3100142180919647, "learning_rate": 8.010820848561565e-05, "loss": 0.0632, "step": 32960 }, { "epoch": 1.114642144764867, "grad_norm": 0.3072890341281891, "learning_rate": 8.009500599472171e-05, "loss": 0.0516, "step": 32970 }, { "epoch": 1.1149802224551202, "grad_norm": 0.17683815956115723, "learning_rate": 8.008180021268314e-05, "loss": 0.0582, "step": 32980 }, { "epoch": 1.1153183001453735, "grad_norm": 0.30328062176704407, "learning_rate": 8.006859114094414e-05, "loss": 0.0536, "step": 32990 }, { "epoch": 1.1156563778356265, "grad_norm": 0.26193785667419434, "learning_rate": 8.005537878094921e-05, "loss": 0.0605, "step": 33000 }, { "epoch": 1.1159944555258798, "grad_norm": 0.2478582113981247, "learning_rate": 8.004216313414323e-05, "loss": 0.0483, "step": 33010 }, { "epoch": 1.1163325332161331, "grad_norm": 0.23285531997680664, "learning_rate": 8.002894420197149e-05, "loss": 0.0502, "step": 33020 }, { "epoch": 1.1166706109063862, "grad_norm": 0.2678227126598358, "learning_rate": 8.001572198587954e-05, "loss": 0.0569, "step": 33030 }, { "epoch": 1.1170086885966395, "grad_norm": 0.2543814182281494, "learning_rate": 8.000249648731338e-05, "loss": 0.0502, "step": 33040 }, { "epoch": 1.1173467662868928, "grad_norm": 0.2086949348449707, "learning_rate": 7.998926770771928e-05, "loss": 0.0521, "step": 33050 }, { "epoch": 1.1176848439771458, "grad_norm": 0.3196662366390228, "learning_rate": 7.997603564854397e-05, "loss": 0.0504, "step": 33060 }, { "epoch": 1.1180229216673991, "grad_norm": 0.3130880296230316, "learning_rate": 7.996280031123448e-05, "loss": 0.0495, "step": 33070 }, { "epoch": 1.1183609993576524, "grad_norm": 0.2474190890789032, "learning_rate": 7.994956169723818e-05, "loss": 0.0551, "step": 33080 }, { "epoch": 1.1186990770479057, "grad_norm": 0.2477315217256546, "learning_rate": 7.993631980800285e-05, "loss": 0.0608, "step": 33090 }, { "epoch": 1.1190371547381588, "grad_norm": 0.22686193883419037, "learning_rate": 7.992307464497659e-05, "loss": 0.0493, "step": 33100 }, { "epoch": 1.119375232428412, "grad_norm": 0.4935896098613739, "learning_rate": 7.990982620960787e-05, "loss": 0.0585, "step": 33110 }, { "epoch": 1.1197133101186654, "grad_norm": 0.38255757093429565, "learning_rate": 7.989657450334554e-05, "loss": 0.0588, "step": 33120 }, { "epoch": 1.1200513878089184, "grad_norm": 0.24623386561870575, "learning_rate": 7.988331952763877e-05, "loss": 0.0558, "step": 33130 }, { "epoch": 1.1203894654991717, "grad_norm": 0.35602283477783203, "learning_rate": 7.987006128393709e-05, "loss": 0.0523, "step": 33140 }, { "epoch": 1.120727543189425, "grad_norm": 0.2697746157646179, "learning_rate": 7.985679977369043e-05, "loss": 0.0592, "step": 33150 }, { "epoch": 1.121065620879678, "grad_norm": 0.22584395110607147, "learning_rate": 7.984353499834902e-05, "loss": 0.0494, "step": 33160 }, { "epoch": 1.1214036985699314, "grad_norm": 0.4308967590332031, "learning_rate": 7.983026695936351e-05, "loss": 0.0528, "step": 33170 }, { "epoch": 1.1217417762601847, "grad_norm": 0.2759913504123688, "learning_rate": 7.981699565818486e-05, "loss": 0.0543, "step": 33180 }, { "epoch": 1.1220798539504377, "grad_norm": 0.23076213896274567, "learning_rate": 7.980372109626437e-05, "loss": 0.0626, "step": 33190 }, { "epoch": 1.122417931640691, "grad_norm": 0.23533868789672852, "learning_rate": 7.979044327505375e-05, "loss": 0.0558, "step": 33200 }, { "epoch": 1.1227560093309443, "grad_norm": 0.2896820902824402, "learning_rate": 7.977716219600506e-05, "loss": 0.0571, "step": 33210 }, { "epoch": 1.1230940870211974, "grad_norm": 0.19762642681598663, "learning_rate": 7.97638778605707e-05, "loss": 0.0519, "step": 33220 }, { "epoch": 1.1234321647114507, "grad_norm": 0.2354164868593216, "learning_rate": 7.975059027020338e-05, "loss": 0.0566, "step": 33230 }, { "epoch": 1.123770242401704, "grad_norm": 0.24263490736484528, "learning_rate": 7.973729942635623e-05, "loss": 0.051, "step": 33240 }, { "epoch": 1.124108320091957, "grad_norm": 0.26952260732650757, "learning_rate": 7.972400533048273e-05, "loss": 0.0626, "step": 33250 }, { "epoch": 1.1244463977822103, "grad_norm": 0.3430621922016144, "learning_rate": 7.97107079840367e-05, "loss": 0.0551, "step": 33260 }, { "epoch": 1.1247844754724636, "grad_norm": 0.2071102261543274, "learning_rate": 7.969740738847231e-05, "loss": 0.0532, "step": 33270 }, { "epoch": 1.1251225531627167, "grad_norm": 0.2434695065021515, "learning_rate": 7.968410354524411e-05, "loss": 0.0523, "step": 33280 }, { "epoch": 1.12546063085297, "grad_norm": 0.5177187919616699, "learning_rate": 7.967079645580697e-05, "loss": 0.0573, "step": 33290 }, { "epoch": 1.1257987085432233, "grad_norm": 0.2781270742416382, "learning_rate": 7.965748612161612e-05, "loss": 0.0518, "step": 33300 }, { "epoch": 1.1261367862334764, "grad_norm": 0.3056272268295288, "learning_rate": 7.96441725441272e-05, "loss": 0.048, "step": 33310 }, { "epoch": 1.1264748639237296, "grad_norm": 0.37883424758911133, "learning_rate": 7.963085572479614e-05, "loss": 0.0621, "step": 33320 }, { "epoch": 1.126812941613983, "grad_norm": 0.3852507472038269, "learning_rate": 7.961753566507924e-05, "loss": 0.0584, "step": 33330 }, { "epoch": 1.127151019304236, "grad_norm": 0.43696457147598267, "learning_rate": 7.960421236643316e-05, "loss": 0.0549, "step": 33340 }, { "epoch": 1.1274890969944893, "grad_norm": 0.2291078418493271, "learning_rate": 7.959088583031496e-05, "loss": 0.0577, "step": 33350 }, { "epoch": 1.1278271746847426, "grad_norm": 0.46580979228019714, "learning_rate": 7.957755605818194e-05, "loss": 0.0517, "step": 33360 }, { "epoch": 1.1281652523749957, "grad_norm": 0.3149864971637726, "learning_rate": 7.956422305149185e-05, "loss": 0.0521, "step": 33370 }, { "epoch": 1.128503330065249, "grad_norm": 0.2437591850757599, "learning_rate": 7.95508868117028e-05, "loss": 0.0544, "step": 33380 }, { "epoch": 1.1288414077555022, "grad_norm": 0.3054518401622772, "learning_rate": 7.953754734027318e-05, "loss": 0.057, "step": 33390 }, { "epoch": 1.1291794854457555, "grad_norm": 0.316851407289505, "learning_rate": 7.952420463866182e-05, "loss": 0.0532, "step": 33400 }, { "epoch": 1.1295175631360086, "grad_norm": 0.2303343415260315, "learning_rate": 7.951085870832782e-05, "loss": 0.0531, "step": 33410 }, { "epoch": 1.1298556408262619, "grad_norm": 0.22508589923381805, "learning_rate": 7.949750955073067e-05, "loss": 0.0515, "step": 33420 }, { "epoch": 1.1301937185165152, "grad_norm": 0.29507091641426086, "learning_rate": 7.948415716733022e-05, "loss": 0.0621, "step": 33430 }, { "epoch": 1.1305317962067682, "grad_norm": 0.4242004454135895, "learning_rate": 7.947080155958669e-05, "loss": 0.059, "step": 33440 }, { "epoch": 1.1308698738970215, "grad_norm": 0.5234391689300537, "learning_rate": 7.94574427289606e-05, "loss": 0.0592, "step": 33450 }, { "epoch": 1.1312079515872748, "grad_norm": 0.2650260329246521, "learning_rate": 7.944408067691284e-05, "loss": 0.0593, "step": 33460 }, { "epoch": 1.131546029277528, "grad_norm": 0.385843425989151, "learning_rate": 7.943071540490473e-05, "loss": 0.0577, "step": 33470 }, { "epoch": 1.1318841069677812, "grad_norm": 0.20409341156482697, "learning_rate": 7.94173469143978e-05, "loss": 0.0583, "step": 33480 }, { "epoch": 1.1322221846580345, "grad_norm": 0.17937231063842773, "learning_rate": 7.940397520685406e-05, "loss": 0.0575, "step": 33490 }, { "epoch": 1.1325602623482875, "grad_norm": 0.19298900663852692, "learning_rate": 7.939060028373577e-05, "loss": 0.0506, "step": 33500 }, { "epoch": 1.1328983400385408, "grad_norm": 0.23751646280288696, "learning_rate": 7.937722214650565e-05, "loss": 0.0492, "step": 33510 }, { "epoch": 1.1332364177287941, "grad_norm": 0.37948060035705566, "learning_rate": 7.936384079662666e-05, "loss": 0.0531, "step": 33520 }, { "epoch": 1.1335744954190472, "grad_norm": 0.26388052105903625, "learning_rate": 7.93504562355622e-05, "loss": 0.0525, "step": 33530 }, { "epoch": 1.1339125731093005, "grad_norm": 0.22936923801898956, "learning_rate": 7.933706846477599e-05, "loss": 0.0523, "step": 33540 }, { "epoch": 1.1342506507995538, "grad_norm": 0.26890888810157776, "learning_rate": 7.932367748573206e-05, "loss": 0.0595, "step": 33550 }, { "epoch": 1.134588728489807, "grad_norm": 0.16282673180103302, "learning_rate": 7.931028329989485e-05, "loss": 0.0584, "step": 33560 }, { "epoch": 1.1349268061800601, "grad_norm": 0.30871060490608215, "learning_rate": 7.929688590872913e-05, "loss": 0.0558, "step": 33570 }, { "epoch": 1.1352648838703134, "grad_norm": 0.21050550043582916, "learning_rate": 7.928348531370003e-05, "loss": 0.0525, "step": 33580 }, { "epoch": 1.1356029615605667, "grad_norm": 0.2691861689090729, "learning_rate": 7.927008151627297e-05, "loss": 0.0538, "step": 33590 }, { "epoch": 1.1359410392508198, "grad_norm": 0.293673038482666, "learning_rate": 7.925667451791383e-05, "loss": 0.0518, "step": 33600 }, { "epoch": 1.136279116941073, "grad_norm": 0.16478201746940613, "learning_rate": 7.924326432008874e-05, "loss": 0.0529, "step": 33610 }, { "epoch": 1.1366171946313264, "grad_norm": 0.20562170445919037, "learning_rate": 7.922985092426422e-05, "loss": 0.058, "step": 33620 }, { "epoch": 1.1369552723215794, "grad_norm": 0.2695719301700592, "learning_rate": 7.921643433190717e-05, "loss": 0.0511, "step": 33630 }, { "epoch": 1.1372933500118327, "grad_norm": 0.3665218949317932, "learning_rate": 7.920301454448478e-05, "loss": 0.0585, "step": 33640 }, { "epoch": 1.137631427702086, "grad_norm": 0.20084111392498016, "learning_rate": 7.918959156346461e-05, "loss": 0.0587, "step": 33650 }, { "epoch": 1.137969505392339, "grad_norm": 0.9430009126663208, "learning_rate": 7.91761653903146e-05, "loss": 0.0556, "step": 33660 }, { "epoch": 1.1383075830825924, "grad_norm": 0.2973940968513489, "learning_rate": 7.916273602650302e-05, "loss": 0.0538, "step": 33670 }, { "epoch": 1.1386456607728457, "grad_norm": 0.18406443297863007, "learning_rate": 7.914930347349847e-05, "loss": 0.0581, "step": 33680 }, { "epoch": 1.1389837384630987, "grad_norm": 0.3025643527507782, "learning_rate": 7.913586773276992e-05, "loss": 0.0567, "step": 33690 }, { "epoch": 1.139321816153352, "grad_norm": 0.21036125719547272, "learning_rate": 7.912242880578667e-05, "loss": 0.0546, "step": 33700 }, { "epoch": 1.1396598938436053, "grad_norm": 0.23135441541671753, "learning_rate": 7.910898669401839e-05, "loss": 0.0521, "step": 33710 }, { "epoch": 1.1399979715338584, "grad_norm": 0.2652775049209595, "learning_rate": 7.909554139893511e-05, "loss": 0.0533, "step": 33720 }, { "epoch": 1.1403360492241117, "grad_norm": 0.2276260107755661, "learning_rate": 7.908209292200715e-05, "loss": 0.0511, "step": 33730 }, { "epoch": 1.140674126914365, "grad_norm": 0.2175161987543106, "learning_rate": 7.906864126470523e-05, "loss": 0.0579, "step": 33740 }, { "epoch": 1.141012204604618, "grad_norm": 0.24981874227523804, "learning_rate": 7.905518642850041e-05, "loss": 0.0557, "step": 33750 }, { "epoch": 1.1413502822948713, "grad_norm": 0.23478126525878906, "learning_rate": 7.904172841486409e-05, "loss": 0.0577, "step": 33760 }, { "epoch": 1.1416883599851246, "grad_norm": 0.3392138183116913, "learning_rate": 7.902826722526801e-05, "loss": 0.0471, "step": 33770 }, { "epoch": 1.1420264376753777, "grad_norm": 0.2715267539024353, "learning_rate": 7.901480286118427e-05, "loss": 0.0583, "step": 33780 }, { "epoch": 1.142364515365631, "grad_norm": 0.25953400135040283, "learning_rate": 7.900133532408531e-05, "loss": 0.0609, "step": 33790 }, { "epoch": 1.1427025930558843, "grad_norm": 0.22027873992919922, "learning_rate": 7.898786461544395e-05, "loss": 0.0549, "step": 33800 }, { "epoch": 1.1430406707461374, "grad_norm": 0.1856301873922348, "learning_rate": 7.897439073673325e-05, "loss": 0.0496, "step": 33810 }, { "epoch": 1.1433787484363906, "grad_norm": 0.27925124764442444, "learning_rate": 7.896091368942677e-05, "loss": 0.0559, "step": 33820 }, { "epoch": 1.143716826126644, "grad_norm": 0.28504353761672974, "learning_rate": 7.894743347499832e-05, "loss": 0.0534, "step": 33830 }, { "epoch": 1.144054903816897, "grad_norm": 0.16769427061080933, "learning_rate": 7.893395009492203e-05, "loss": 0.0554, "step": 33840 }, { "epoch": 1.1443929815071503, "grad_norm": 0.2372005581855774, "learning_rate": 7.892046355067248e-05, "loss": 0.0493, "step": 33850 }, { "epoch": 1.1447310591974036, "grad_norm": 0.224487766623497, "learning_rate": 7.890697384372451e-05, "loss": 0.0566, "step": 33860 }, { "epoch": 1.1450691368876569, "grad_norm": 0.24835321307182312, "learning_rate": 7.889348097555336e-05, "loss": 0.0563, "step": 33870 }, { "epoch": 1.14540721457791, "grad_norm": 0.28102079033851624, "learning_rate": 7.887998494763455e-05, "loss": 0.0528, "step": 33880 }, { "epoch": 1.1457452922681632, "grad_norm": 0.1764947474002838, "learning_rate": 7.886648576144404e-05, "loss": 0.0534, "step": 33890 }, { "epoch": 1.1460833699584165, "grad_norm": 0.2190176546573639, "learning_rate": 7.885298341845802e-05, "loss": 0.0607, "step": 33900 }, { "epoch": 1.1464214476486696, "grad_norm": 0.5785945057868958, "learning_rate": 7.883947792015311e-05, "loss": 0.0526, "step": 33910 }, { "epoch": 1.146759525338923, "grad_norm": 0.24770820140838623, "learning_rate": 7.882596926800628e-05, "loss": 0.0544, "step": 33920 }, { "epoch": 1.1470976030291762, "grad_norm": 0.3102057874202728, "learning_rate": 7.881245746349477e-05, "loss": 0.0579, "step": 33930 }, { "epoch": 1.1474356807194293, "grad_norm": 0.291506826877594, "learning_rate": 7.879894250809623e-05, "loss": 0.0595, "step": 33940 }, { "epoch": 1.1477737584096825, "grad_norm": 0.2323945164680481, "learning_rate": 7.878542440328865e-05, "loss": 0.0552, "step": 33950 }, { "epoch": 1.1481118360999358, "grad_norm": 0.25342628359794617, "learning_rate": 7.877190315055031e-05, "loss": 0.0539, "step": 33960 }, { "epoch": 1.148449913790189, "grad_norm": 0.29345396161079407, "learning_rate": 7.875837875135991e-05, "loss": 0.0677, "step": 33970 }, { "epoch": 1.1487879914804422, "grad_norm": 0.18372267484664917, "learning_rate": 7.874485120719646e-05, "loss": 0.0587, "step": 33980 }, { "epoch": 1.1491260691706955, "grad_norm": 0.18869388103485107, "learning_rate": 7.873132051953928e-05, "loss": 0.0573, "step": 33990 }, { "epoch": 1.1494641468609486, "grad_norm": 0.2241271287202835, "learning_rate": 7.87177866898681e-05, "loss": 0.0582, "step": 34000 }, { "epoch": 1.1498022245512018, "grad_norm": 0.3156338930130005, "learning_rate": 7.870424971966294e-05, "loss": 0.0517, "step": 34010 }, { "epoch": 1.1501403022414551, "grad_norm": 0.3545479476451874, "learning_rate": 7.869070961040419e-05, "loss": 0.0549, "step": 34020 }, { "epoch": 1.1504783799317084, "grad_norm": 0.30931776762008667, "learning_rate": 7.867716636357257e-05, "loss": 0.0581, "step": 34030 }, { "epoch": 1.1508164576219615, "grad_norm": 0.36123335361480713, "learning_rate": 7.866361998064915e-05, "loss": 0.0535, "step": 34040 }, { "epoch": 1.1511545353122148, "grad_norm": 0.27207887172698975, "learning_rate": 7.865007046311534e-05, "loss": 0.0588, "step": 34050 }, { "epoch": 1.151492613002468, "grad_norm": 0.3311230540275574, "learning_rate": 7.86365178124529e-05, "loss": 0.0556, "step": 34060 }, { "epoch": 1.1518306906927211, "grad_norm": 0.31971070170402527, "learning_rate": 7.862296203014394e-05, "loss": 0.0539, "step": 34070 }, { "epoch": 1.1521687683829744, "grad_norm": 0.2378312647342682, "learning_rate": 7.860940311767088e-05, "loss": 0.0592, "step": 34080 }, { "epoch": 1.1525068460732277, "grad_norm": 0.19608178734779358, "learning_rate": 7.85958410765165e-05, "loss": 0.0582, "step": 34090 }, { "epoch": 1.1528449237634808, "grad_norm": 0.2059326022863388, "learning_rate": 7.858227590816394e-05, "loss": 0.0596, "step": 34100 }, { "epoch": 1.153183001453734, "grad_norm": 0.20221740007400513, "learning_rate": 7.856870761409664e-05, "loss": 0.0589, "step": 34110 }, { "epoch": 1.1535210791439874, "grad_norm": 0.33082494139671326, "learning_rate": 7.855513619579846e-05, "loss": 0.0568, "step": 34120 }, { "epoch": 1.1538591568342405, "grad_norm": 0.40738645195961, "learning_rate": 7.85415616547535e-05, "loss": 0.05, "step": 34130 }, { "epoch": 1.1541972345244937, "grad_norm": 0.35039374232292175, "learning_rate": 7.852798399244627e-05, "loss": 0.0568, "step": 34140 }, { "epoch": 1.154535312214747, "grad_norm": 0.4134216606616974, "learning_rate": 7.851440321036161e-05, "loss": 0.0472, "step": 34150 }, { "epoch": 1.154873389905, "grad_norm": 0.3892120122909546, "learning_rate": 7.850081930998468e-05, "loss": 0.0509, "step": 34160 }, { "epoch": 1.1552114675952534, "grad_norm": 0.36543795466423035, "learning_rate": 7.8487232292801e-05, "loss": 0.0619, "step": 34170 }, { "epoch": 1.1555495452855067, "grad_norm": 0.25026005506515503, "learning_rate": 7.847364216029642e-05, "loss": 0.0505, "step": 34180 }, { "epoch": 1.1558876229757598, "grad_norm": 0.22869116067886353, "learning_rate": 7.846004891395716e-05, "loss": 0.0538, "step": 34190 }, { "epoch": 1.156225700666013, "grad_norm": 0.2022005170583725, "learning_rate": 7.844645255526972e-05, "loss": 0.0592, "step": 34200 }, { "epoch": 1.1565637783562663, "grad_norm": 0.34690871834754944, "learning_rate": 7.843285308572101e-05, "loss": 0.0578, "step": 34210 }, { "epoch": 1.1569018560465194, "grad_norm": 0.1776241958141327, "learning_rate": 7.841925050679823e-05, "loss": 0.0562, "step": 34220 }, { "epoch": 1.1572399337367727, "grad_norm": 0.3586002290248871, "learning_rate": 7.840564481998895e-05, "loss": 0.0511, "step": 34230 }, { "epoch": 1.157578011427026, "grad_norm": 0.217389777302742, "learning_rate": 7.839203602678105e-05, "loss": 0.0501, "step": 34240 }, { "epoch": 1.157916089117279, "grad_norm": 0.3581806421279907, "learning_rate": 7.837842412866279e-05, "loss": 0.0565, "step": 34250 }, { "epoch": 1.1582541668075323, "grad_norm": 0.3202519416809082, "learning_rate": 7.836480912712272e-05, "loss": 0.0533, "step": 34260 }, { "epoch": 1.1585922444977856, "grad_norm": 0.20308814942836761, "learning_rate": 7.835119102364976e-05, "loss": 0.051, "step": 34270 }, { "epoch": 1.1589303221880387, "grad_norm": 0.19461318850517273, "learning_rate": 7.83375698197332e-05, "loss": 0.059, "step": 34280 }, { "epoch": 1.159268399878292, "grad_norm": 0.5369237065315247, "learning_rate": 7.83239455168626e-05, "loss": 0.0629, "step": 34290 }, { "epoch": 1.1596064775685453, "grad_norm": 0.20266707241535187, "learning_rate": 7.83103181165279e-05, "loss": 0.061, "step": 34300 }, { "epoch": 1.1599445552587984, "grad_norm": 0.41588711738586426, "learning_rate": 7.829668762021937e-05, "loss": 0.058, "step": 34310 }, { "epoch": 1.1602826329490517, "grad_norm": 0.30450090765953064, "learning_rate": 7.828305402942764e-05, "loss": 0.0611, "step": 34320 }, { "epoch": 1.160620710639305, "grad_norm": 0.16033510863780975, "learning_rate": 7.826941734564363e-05, "loss": 0.0554, "step": 34330 }, { "epoch": 1.160958788329558, "grad_norm": 0.23618844151496887, "learning_rate": 7.825577757035865e-05, "loss": 0.0547, "step": 34340 }, { "epoch": 1.1612968660198113, "grad_norm": 0.2673484981060028, "learning_rate": 7.824213470506431e-05, "loss": 0.056, "step": 34350 }, { "epoch": 1.1616349437100646, "grad_norm": 0.44811391830444336, "learning_rate": 7.822848875125257e-05, "loss": 0.0564, "step": 34360 }, { "epoch": 1.1619730214003179, "grad_norm": 0.5029431581497192, "learning_rate": 7.821483971041576e-05, "loss": 0.0533, "step": 34370 }, { "epoch": 1.162311099090571, "grad_norm": 0.44127729535102844, "learning_rate": 7.820118758404649e-05, "loss": 0.0613, "step": 34380 }, { "epoch": 1.1626491767808242, "grad_norm": 0.24449798464775085, "learning_rate": 7.818753237363776e-05, "loss": 0.0559, "step": 34390 }, { "epoch": 1.1629872544710775, "grad_norm": 0.4075545370578766, "learning_rate": 7.817387408068286e-05, "loss": 0.0591, "step": 34400 }, { "epoch": 1.1633253321613306, "grad_norm": 0.19425800442695618, "learning_rate": 7.816021270667544e-05, "loss": 0.0564, "step": 34410 }, { "epoch": 1.163663409851584, "grad_norm": 0.3363246023654938, "learning_rate": 7.81465482531095e-05, "loss": 0.0532, "step": 34420 }, { "epoch": 1.1640014875418372, "grad_norm": 0.16833613812923431, "learning_rate": 7.813288072147938e-05, "loss": 0.054, "step": 34430 }, { "epoch": 1.1643395652320903, "grad_norm": 0.21299058198928833, "learning_rate": 7.811921011327972e-05, "loss": 0.0589, "step": 34440 }, { "epoch": 1.1646776429223435, "grad_norm": 0.2206108421087265, "learning_rate": 7.810553643000549e-05, "loss": 0.0545, "step": 34450 }, { "epoch": 1.1650157206125968, "grad_norm": 0.21495579183101654, "learning_rate": 7.809185967315206e-05, "loss": 0.0486, "step": 34460 }, { "epoch": 1.16535379830285, "grad_norm": 0.29266276955604553, "learning_rate": 7.80781798442151e-05, "loss": 0.0543, "step": 34470 }, { "epoch": 1.1656918759931032, "grad_norm": 0.2199344038963318, "learning_rate": 7.806449694469061e-05, "loss": 0.0538, "step": 34480 }, { "epoch": 1.1660299536833565, "grad_norm": 0.30705663561820984, "learning_rate": 7.805081097607492e-05, "loss": 0.0533, "step": 34490 }, { "epoch": 1.1663680313736096, "grad_norm": 0.27856919169425964, "learning_rate": 7.803712193986474e-05, "loss": 0.0516, "step": 34500 }, { "epoch": 1.1667061090638629, "grad_norm": 0.3928166925907135, "learning_rate": 7.802342983755702e-05, "loss": 0.0569, "step": 34510 }, { "epoch": 1.1670441867541161, "grad_norm": 0.25692352652549744, "learning_rate": 7.800973467064918e-05, "loss": 0.0573, "step": 34520 }, { "epoch": 1.1673822644443694, "grad_norm": 0.26086917519569397, "learning_rate": 7.799603644063884e-05, "loss": 0.0561, "step": 34530 }, { "epoch": 1.1677203421346225, "grad_norm": 0.1834735870361328, "learning_rate": 7.798233514902405e-05, "loss": 0.0515, "step": 34540 }, { "epoch": 1.1680584198248758, "grad_norm": 0.280857115983963, "learning_rate": 7.796863079730318e-05, "loss": 0.057, "step": 34550 }, { "epoch": 1.168396497515129, "grad_norm": 0.33533167839050293, "learning_rate": 7.795492338697488e-05, "loss": 0.0522, "step": 34560 }, { "epoch": 1.1687345752053822, "grad_norm": 0.23430410027503967, "learning_rate": 7.794121291953819e-05, "loss": 0.0537, "step": 34570 }, { "epoch": 1.1690726528956354, "grad_norm": 0.23052369058132172, "learning_rate": 7.792749939649246e-05, "loss": 0.0515, "step": 34580 }, { "epoch": 1.1694107305858887, "grad_norm": 0.9137349128723145, "learning_rate": 7.79137828193374e-05, "loss": 0.0605, "step": 34590 }, { "epoch": 1.1697488082761418, "grad_norm": 0.2716549336910248, "learning_rate": 7.790006318957301e-05, "loss": 0.0562, "step": 34600 }, { "epoch": 1.170086885966395, "grad_norm": 0.22722570598125458, "learning_rate": 7.788634050869965e-05, "loss": 0.0531, "step": 34610 }, { "epoch": 1.1704249636566484, "grad_norm": 0.47030505537986755, "learning_rate": 7.787261477821803e-05, "loss": 0.0582, "step": 34620 }, { "epoch": 1.1707630413469015, "grad_norm": 0.21646273136138916, "learning_rate": 7.785888599962916e-05, "loss": 0.0542, "step": 34630 }, { "epoch": 1.1711011190371547, "grad_norm": 0.2868708074092865, "learning_rate": 7.784515417443439e-05, "loss": 0.0536, "step": 34640 }, { "epoch": 1.171439196727408, "grad_norm": 0.22809754312038422, "learning_rate": 7.783141930413545e-05, "loss": 0.0549, "step": 34650 }, { "epoch": 1.171777274417661, "grad_norm": 0.20400087535381317, "learning_rate": 7.78176813902343e-05, "loss": 0.0522, "step": 34660 }, { "epoch": 1.1721153521079144, "grad_norm": 0.26866430044174194, "learning_rate": 7.780394043423336e-05, "loss": 0.0482, "step": 34670 }, { "epoch": 1.1724534297981677, "grad_norm": 0.3160463571548462, "learning_rate": 7.77901964376353e-05, "loss": 0.0591, "step": 34680 }, { "epoch": 1.1727915074884208, "grad_norm": 0.4285029172897339, "learning_rate": 7.777644940194316e-05, "loss": 0.0572, "step": 34690 }, { "epoch": 1.173129585178674, "grad_norm": 0.2597242593765259, "learning_rate": 7.776269932866023e-05, "loss": 0.0524, "step": 34700 }, { "epoch": 1.1734676628689273, "grad_norm": 0.23743891716003418, "learning_rate": 7.774894621929026e-05, "loss": 0.057, "step": 34710 }, { "epoch": 1.1738057405591804, "grad_norm": 0.2657506465911865, "learning_rate": 7.773519007533725e-05, "loss": 0.0524, "step": 34720 }, { "epoch": 1.1741438182494337, "grad_norm": 0.2867894172668457, "learning_rate": 7.772143089830556e-05, "loss": 0.0518, "step": 34730 }, { "epoch": 1.174481895939687, "grad_norm": 0.5348312854766846, "learning_rate": 7.770766868969985e-05, "loss": 0.0538, "step": 34740 }, { "epoch": 1.17481997362994, "grad_norm": 0.2698118984699249, "learning_rate": 7.769390345102518e-05, "loss": 0.0598, "step": 34750 }, { "epoch": 1.1751580513201934, "grad_norm": 0.31774213910102844, "learning_rate": 7.768013518378683e-05, "loss": 0.0563, "step": 34760 }, { "epoch": 1.1754961290104466, "grad_norm": 0.5828033089637756, "learning_rate": 7.766636388949053e-05, "loss": 0.0531, "step": 34770 }, { "epoch": 1.1758342067006997, "grad_norm": 0.30947384238243103, "learning_rate": 7.765258956964229e-05, "loss": 0.0539, "step": 34780 }, { "epoch": 1.176172284390953, "grad_norm": 0.3540792763233185, "learning_rate": 7.76388122257484e-05, "loss": 0.0531, "step": 34790 }, { "epoch": 1.1765103620812063, "grad_norm": 0.4563525915145874, "learning_rate": 7.762503185931558e-05, "loss": 0.0573, "step": 34800 }, { "epoch": 1.1768484397714594, "grad_norm": 0.1911778450012207, "learning_rate": 7.76112484718508e-05, "loss": 0.0597, "step": 34810 }, { "epoch": 1.1771865174617127, "grad_norm": 0.29575470089912415, "learning_rate": 7.75974620648614e-05, "loss": 0.0624, "step": 34820 }, { "epoch": 1.177524595151966, "grad_norm": 0.4941204786300659, "learning_rate": 7.758367263985503e-05, "loss": 0.0511, "step": 34830 }, { "epoch": 1.1778626728422192, "grad_norm": 0.24307893216609955, "learning_rate": 7.75698801983397e-05, "loss": 0.0591, "step": 34840 }, { "epoch": 1.1782007505324723, "grad_norm": 0.22649899125099182, "learning_rate": 7.755608474182372e-05, "loss": 0.0557, "step": 34850 }, { "epoch": 1.1785388282227256, "grad_norm": 0.19158872961997986, "learning_rate": 7.754228627181574e-05, "loss": 0.0605, "step": 34860 }, { "epoch": 1.178876905912979, "grad_norm": 0.2429138422012329, "learning_rate": 7.752848478982476e-05, "loss": 0.0502, "step": 34870 }, { "epoch": 1.179214983603232, "grad_norm": 0.30495551228523254, "learning_rate": 7.751468029736006e-05, "loss": 0.0556, "step": 34880 }, { "epoch": 1.1795530612934852, "grad_norm": 0.2444792240858078, "learning_rate": 7.750087279593129e-05, "loss": 0.052, "step": 34890 }, { "epoch": 1.1798911389837385, "grad_norm": 0.25772368907928467, "learning_rate": 7.748706228704843e-05, "loss": 0.0509, "step": 34900 }, { "epoch": 1.1802292166739916, "grad_norm": 0.29475679993629456, "learning_rate": 7.747324877222176e-05, "loss": 0.0536, "step": 34910 }, { "epoch": 1.180567294364245, "grad_norm": 0.3355693519115448, "learning_rate": 7.745943225296188e-05, "loss": 0.0554, "step": 34920 }, { "epoch": 1.1809053720544982, "grad_norm": 0.29621291160583496, "learning_rate": 7.744561273077981e-05, "loss": 0.0548, "step": 34930 }, { "epoch": 1.1812434497447513, "grad_norm": 0.22347308695316315, "learning_rate": 7.743179020718678e-05, "loss": 0.0535, "step": 34940 }, { "epoch": 1.1815815274350046, "grad_norm": 0.3569125235080719, "learning_rate": 7.741796468369443e-05, "loss": 0.0527, "step": 34950 }, { "epoch": 1.1819196051252578, "grad_norm": 0.22408737242221832, "learning_rate": 7.740413616181466e-05, "loss": 0.0554, "step": 34960 }, { "epoch": 1.182257682815511, "grad_norm": 0.19341841340065002, "learning_rate": 7.739030464305978e-05, "loss": 0.0598, "step": 34970 }, { "epoch": 1.1825957605057642, "grad_norm": 0.22846277058124542, "learning_rate": 7.737647012894235e-05, "loss": 0.0538, "step": 34980 }, { "epoch": 1.1829338381960175, "grad_norm": 0.3455890119075775, "learning_rate": 7.736263262097532e-05, "loss": 0.0582, "step": 34990 }, { "epoch": 1.1832719158862708, "grad_norm": 0.2978130877017975, "learning_rate": 7.734879212067192e-05, "loss": 0.0556, "step": 35000 }, { "epoch": 1.1836099935765239, "grad_norm": 0.2925918400287628, "learning_rate": 7.733494862954573e-05, "loss": 0.0589, "step": 35010 }, { "epoch": 1.1839480712667771, "grad_norm": 0.272754967212677, "learning_rate": 7.732110214911066e-05, "loss": 0.0544, "step": 35020 }, { "epoch": 1.1842861489570304, "grad_norm": 0.3018021881580353, "learning_rate": 7.730725268088094e-05, "loss": 0.0599, "step": 35030 }, { "epoch": 1.1846242266472835, "grad_norm": 0.19235853850841522, "learning_rate": 7.729340022637112e-05, "loss": 0.0555, "step": 35040 }, { "epoch": 1.1849623043375368, "grad_norm": 0.4539581835269928, "learning_rate": 7.727954478709607e-05, "loss": 0.0542, "step": 35050 }, { "epoch": 1.18530038202779, "grad_norm": 0.34614455699920654, "learning_rate": 7.726568636457103e-05, "loss": 0.0515, "step": 35060 }, { "epoch": 1.1856384597180432, "grad_norm": 0.2721632719039917, "learning_rate": 7.725182496031153e-05, "loss": 0.053, "step": 35070 }, { "epoch": 1.1859765374082964, "grad_norm": 0.20669880509376526, "learning_rate": 7.723796057583341e-05, "loss": 0.058, "step": 35080 }, { "epoch": 1.1863146150985497, "grad_norm": 0.3491286039352417, "learning_rate": 7.722409321265287e-05, "loss": 0.0587, "step": 35090 }, { "epoch": 1.1866526927888028, "grad_norm": 0.16965682804584503, "learning_rate": 7.721022287228645e-05, "loss": 0.0546, "step": 35100 }, { "epoch": 1.186990770479056, "grad_norm": 0.27842971682548523, "learning_rate": 7.719634955625094e-05, "loss": 0.0584, "step": 35110 }, { "epoch": 1.1873288481693094, "grad_norm": 0.31961122155189514, "learning_rate": 7.718247326606355e-05, "loss": 0.0509, "step": 35120 }, { "epoch": 1.1876669258595625, "grad_norm": 0.2931060791015625, "learning_rate": 7.716859400324175e-05, "loss": 0.0496, "step": 35130 }, { "epoch": 1.1880050035498158, "grad_norm": 0.20686674118041992, "learning_rate": 7.715471176930335e-05, "loss": 0.0479, "step": 35140 }, { "epoch": 1.188343081240069, "grad_norm": 0.3077336847782135, "learning_rate": 7.714082656576651e-05, "loss": 0.0562, "step": 35150 }, { "epoch": 1.1886811589303221, "grad_norm": 0.28611505031585693, "learning_rate": 7.712693839414968e-05, "loss": 0.0622, "step": 35160 }, { "epoch": 1.1890192366205754, "grad_norm": 0.20982570946216583, "learning_rate": 7.711304725597164e-05, "loss": 0.0542, "step": 35170 }, { "epoch": 1.1893573143108287, "grad_norm": 0.3443072736263275, "learning_rate": 7.709915315275151e-05, "loss": 0.0565, "step": 35180 }, { "epoch": 1.1896953920010818, "grad_norm": 0.1747412085533142, "learning_rate": 7.708525608600876e-05, "loss": 0.0452, "step": 35190 }, { "epoch": 1.190033469691335, "grad_norm": 0.24992048740386963, "learning_rate": 7.707135605726311e-05, "loss": 0.056, "step": 35200 }, { "epoch": 1.1903715473815883, "grad_norm": 0.3503495156764984, "learning_rate": 7.705745306803466e-05, "loss": 0.0499, "step": 35210 }, { "epoch": 1.1907096250718414, "grad_norm": 0.3587808907032013, "learning_rate": 7.704354711984383e-05, "loss": 0.0527, "step": 35220 }, { "epoch": 1.1910477027620947, "grad_norm": 0.17966461181640625, "learning_rate": 7.702963821421133e-05, "loss": 0.0534, "step": 35230 }, { "epoch": 1.191385780452348, "grad_norm": 0.22482743859291077, "learning_rate": 7.701572635265826e-05, "loss": 0.0553, "step": 35240 }, { "epoch": 1.191723858142601, "grad_norm": 0.18880872428417206, "learning_rate": 7.700181153670596e-05, "loss": 0.0512, "step": 35250 }, { "epoch": 1.1920619358328544, "grad_norm": 0.39648622274398804, "learning_rate": 7.698789376787616e-05, "loss": 0.0504, "step": 35260 }, { "epoch": 1.1924000135231076, "grad_norm": 0.21210075914859772, "learning_rate": 7.697397304769085e-05, "loss": 0.0587, "step": 35270 }, { "epoch": 1.1927380912133607, "grad_norm": 0.3780267834663391, "learning_rate": 7.696004937767241e-05, "loss": 0.06, "step": 35280 }, { "epoch": 1.193076168903614, "grad_norm": 0.29439422488212585, "learning_rate": 7.694612275934352e-05, "loss": 0.0502, "step": 35290 }, { "epoch": 1.1934142465938673, "grad_norm": 0.2989833652973175, "learning_rate": 7.693219319422714e-05, "loss": 0.0557, "step": 35300 }, { "epoch": 1.1937523242841204, "grad_norm": 0.3429073095321655, "learning_rate": 7.69182606838466e-05, "loss": 0.0501, "step": 35310 }, { "epoch": 1.1940904019743737, "grad_norm": 0.36007341742515564, "learning_rate": 7.690432522972558e-05, "loss": 0.0496, "step": 35320 }, { "epoch": 1.194428479664627, "grad_norm": 0.3509940803050995, "learning_rate": 7.689038683338796e-05, "loss": 0.0551, "step": 35330 }, { "epoch": 1.1947665573548802, "grad_norm": 0.20705029368400574, "learning_rate": 7.687644549635808e-05, "loss": 0.048, "step": 35340 }, { "epoch": 1.1951046350451333, "grad_norm": 0.3865252435207367, "learning_rate": 7.686250122016053e-05, "loss": 0.0581, "step": 35350 }, { "epoch": 1.1954427127353866, "grad_norm": 0.3429955244064331, "learning_rate": 7.684855400632022e-05, "loss": 0.0555, "step": 35360 }, { "epoch": 1.19578079042564, "grad_norm": 0.2028629034757614, "learning_rate": 7.683460385636243e-05, "loss": 0.0553, "step": 35370 }, { "epoch": 1.196118868115893, "grad_norm": 0.20636315643787384, "learning_rate": 7.68206507718127e-05, "loss": 0.0553, "step": 35380 }, { "epoch": 1.1964569458061463, "grad_norm": 0.28076785802841187, "learning_rate": 7.680669475419693e-05, "loss": 0.0564, "step": 35390 }, { "epoch": 1.1967950234963995, "grad_norm": 0.5287385582923889, "learning_rate": 7.679273580504132e-05, "loss": 0.0545, "step": 35400 }, { "epoch": 1.1971331011866526, "grad_norm": 0.22573411464691162, "learning_rate": 7.67787739258724e-05, "loss": 0.0554, "step": 35410 }, { "epoch": 1.197471178876906, "grad_norm": 0.19360411167144775, "learning_rate": 7.676480911821705e-05, "loss": 0.0548, "step": 35420 }, { "epoch": 1.1978092565671592, "grad_norm": 0.21578973531723022, "learning_rate": 7.675084138360238e-05, "loss": 0.0535, "step": 35430 }, { "epoch": 1.1981473342574123, "grad_norm": 0.2526601552963257, "learning_rate": 7.673687072355592e-05, "loss": 0.0566, "step": 35440 }, { "epoch": 1.1984854119476656, "grad_norm": 0.22162334620952606, "learning_rate": 7.67228971396055e-05, "loss": 0.0592, "step": 35450 }, { "epoch": 1.1988234896379188, "grad_norm": 0.24896661937236786, "learning_rate": 7.670892063327922e-05, "loss": 0.0573, "step": 35460 }, { "epoch": 1.199161567328172, "grad_norm": 0.6121139526367188, "learning_rate": 7.669494120610552e-05, "loss": 0.0522, "step": 35470 }, { "epoch": 1.1994996450184252, "grad_norm": 0.24267280101776123, "learning_rate": 7.66809588596132e-05, "loss": 0.0551, "step": 35480 }, { "epoch": 1.1998377227086785, "grad_norm": 0.282640665769577, "learning_rate": 7.666697359533132e-05, "loss": 0.0507, "step": 35490 }, { "epoch": 1.2001758003989318, "grad_norm": 0.8950628042221069, "learning_rate": 7.665298541478932e-05, "loss": 0.0526, "step": 35500 }, { "epoch": 1.2005138780891849, "grad_norm": 0.20859995484352112, "learning_rate": 7.66389943195169e-05, "loss": 0.0549, "step": 35510 }, { "epoch": 1.2008519557794382, "grad_norm": 0.1961134821176529, "learning_rate": 7.66250003110441e-05, "loss": 0.0527, "step": 35520 }, { "epoch": 1.2011900334696914, "grad_norm": 0.29327619075775146, "learning_rate": 7.66110033909013e-05, "loss": 0.0555, "step": 35530 }, { "epoch": 1.2015281111599445, "grad_norm": 0.22936104238033295, "learning_rate": 7.659700356061918e-05, "loss": 0.055, "step": 35540 }, { "epoch": 1.2018661888501978, "grad_norm": 0.2558988332748413, "learning_rate": 7.658300082172875e-05, "loss": 0.0503, "step": 35550 }, { "epoch": 1.202204266540451, "grad_norm": 0.26354238390922546, "learning_rate": 7.65689951757613e-05, "loss": 0.0555, "step": 35560 }, { "epoch": 1.2025423442307042, "grad_norm": 0.2737906277179718, "learning_rate": 7.655498662424849e-05, "loss": 0.0465, "step": 35570 }, { "epoch": 1.2028804219209575, "grad_norm": 0.29871317744255066, "learning_rate": 7.654097516872227e-05, "loss": 0.0581, "step": 35580 }, { "epoch": 1.2032184996112107, "grad_norm": 0.320660799741745, "learning_rate": 7.65269608107149e-05, "loss": 0.0549, "step": 35590 }, { "epoch": 1.2035565773014638, "grad_norm": 0.20985427498817444, "learning_rate": 7.6512943551759e-05, "loss": 0.0575, "step": 35600 }, { "epoch": 1.203894654991717, "grad_norm": 0.2054758369922638, "learning_rate": 7.649892339338743e-05, "loss": 0.0551, "step": 35610 }, { "epoch": 1.2042327326819704, "grad_norm": 0.24178947508335114, "learning_rate": 7.648490033713344e-05, "loss": 0.051, "step": 35620 }, { "epoch": 1.2045708103722235, "grad_norm": 0.22430413961410522, "learning_rate": 7.647087438453058e-05, "loss": 0.0549, "step": 35630 }, { "epoch": 1.2049088880624768, "grad_norm": 0.21585653722286224, "learning_rate": 7.64568455371127e-05, "loss": 0.0528, "step": 35640 }, { "epoch": 1.20524696575273, "grad_norm": 0.25504130125045776, "learning_rate": 7.644281379641396e-05, "loss": 0.0517, "step": 35650 }, { "epoch": 1.2055850434429831, "grad_norm": 0.22177304327487946, "learning_rate": 7.642877916396887e-05, "loss": 0.0557, "step": 35660 }, { "epoch": 1.2059231211332364, "grad_norm": 0.25807711482048035, "learning_rate": 7.641474164131221e-05, "loss": 0.0601, "step": 35670 }, { "epoch": 1.2062611988234897, "grad_norm": 0.2162979543209076, "learning_rate": 7.640070122997913e-05, "loss": 0.0541, "step": 35680 }, { "epoch": 1.2065992765137428, "grad_norm": 0.18722431361675262, "learning_rate": 7.638665793150505e-05, "loss": 0.0555, "step": 35690 }, { "epoch": 1.206937354203996, "grad_norm": 0.2578164339065552, "learning_rate": 7.637261174742574e-05, "loss": 0.053, "step": 35700 }, { "epoch": 1.2072754318942494, "grad_norm": 0.36632606387138367, "learning_rate": 7.635856267927725e-05, "loss": 0.0541, "step": 35710 }, { "epoch": 1.2076135095845024, "grad_norm": 0.18670432269573212, "learning_rate": 7.634451072859597e-05, "loss": 0.0564, "step": 35720 }, { "epoch": 1.2079515872747557, "grad_norm": 0.32668426632881165, "learning_rate": 7.633045589691863e-05, "loss": 0.0499, "step": 35730 }, { "epoch": 1.208289664965009, "grad_norm": 0.3056575357913971, "learning_rate": 7.63163981857822e-05, "loss": 0.0583, "step": 35740 }, { "epoch": 1.208627742655262, "grad_norm": 0.5106516480445862, "learning_rate": 7.630233759672403e-05, "loss": 0.0563, "step": 35750 }, { "epoch": 1.2089658203455154, "grad_norm": 0.3427746295928955, "learning_rate": 7.628827413128177e-05, "loss": 0.0545, "step": 35760 }, { "epoch": 1.2093038980357687, "grad_norm": 0.21214430034160614, "learning_rate": 7.627420779099338e-05, "loss": 0.0477, "step": 35770 }, { "epoch": 1.2096419757260217, "grad_norm": 0.2890525460243225, "learning_rate": 7.626013857739711e-05, "loss": 0.0556, "step": 35780 }, { "epoch": 1.209980053416275, "grad_norm": 0.33664214611053467, "learning_rate": 7.624606649203158e-05, "loss": 0.0511, "step": 35790 }, { "epoch": 1.2103181311065283, "grad_norm": 0.29065608978271484, "learning_rate": 7.623199153643569e-05, "loss": 0.0586, "step": 35800 }, { "epoch": 1.2106562087967816, "grad_norm": 0.32291579246520996, "learning_rate": 7.621791371214863e-05, "loss": 0.0541, "step": 35810 }, { "epoch": 1.2109942864870347, "grad_norm": 0.37642747163772583, "learning_rate": 7.620383302070995e-05, "loss": 0.0576, "step": 35820 }, { "epoch": 1.211332364177288, "grad_norm": 0.209851935505867, "learning_rate": 7.61897494636595e-05, "loss": 0.0508, "step": 35830 }, { "epoch": 1.2116704418675412, "grad_norm": 0.2649427354335785, "learning_rate": 7.617566304253739e-05, "loss": 0.0564, "step": 35840 }, { "epoch": 1.2120085195577943, "grad_norm": 0.23237283527851105, "learning_rate": 7.616157375888416e-05, "loss": 0.0538, "step": 35850 }, { "epoch": 1.2123465972480476, "grad_norm": 0.3539905250072479, "learning_rate": 7.614748161424053e-05, "loss": 0.0508, "step": 35860 }, { "epoch": 1.212684674938301, "grad_norm": 0.325747549533844, "learning_rate": 7.613338661014763e-05, "loss": 0.0584, "step": 35870 }, { "epoch": 1.213022752628554, "grad_norm": 0.2847708463668823, "learning_rate": 7.611928874814686e-05, "loss": 0.051, "step": 35880 }, { "epoch": 1.2133608303188073, "grad_norm": 0.29057225584983826, "learning_rate": 7.610518802977996e-05, "loss": 0.0531, "step": 35890 }, { "epoch": 1.2136989080090606, "grad_norm": 0.19609495997428894, "learning_rate": 7.609108445658893e-05, "loss": 0.0489, "step": 35900 }, { "epoch": 1.2140369856993136, "grad_norm": 0.3256232738494873, "learning_rate": 7.607697803011612e-05, "loss": 0.0489, "step": 35910 }, { "epoch": 1.214375063389567, "grad_norm": 0.4146779477596283, "learning_rate": 7.606286875190421e-05, "loss": 0.0564, "step": 35920 }, { "epoch": 1.2147131410798202, "grad_norm": 0.4358370006084442, "learning_rate": 7.604875662349617e-05, "loss": 0.0583, "step": 35930 }, { "epoch": 1.2150512187700733, "grad_norm": 0.205158531665802, "learning_rate": 7.603464164643525e-05, "loss": 0.0607, "step": 35940 }, { "epoch": 1.2153892964603266, "grad_norm": 0.19800938665866852, "learning_rate": 7.602052382226507e-05, "loss": 0.0549, "step": 35950 }, { "epoch": 1.2157273741505799, "grad_norm": 0.2945767045021057, "learning_rate": 7.600640315252954e-05, "loss": 0.0583, "step": 35960 }, { "epoch": 1.2160654518408331, "grad_norm": 0.1974865198135376, "learning_rate": 7.599227963877284e-05, "loss": 0.0513, "step": 35970 }, { "epoch": 1.2164035295310862, "grad_norm": 0.3312009274959564, "learning_rate": 7.597815328253954e-05, "loss": 0.0541, "step": 35980 }, { "epoch": 1.2167416072213395, "grad_norm": 0.29669731855392456, "learning_rate": 7.596402408537443e-05, "loss": 0.0492, "step": 35990 }, { "epoch": 1.2170796849115928, "grad_norm": 0.2887393534183502, "learning_rate": 7.59498920488227e-05, "loss": 0.0561, "step": 36000 }, { "epoch": 1.2174177626018459, "grad_norm": 0.3521360754966736, "learning_rate": 7.593575717442979e-05, "loss": 0.0515, "step": 36010 }, { "epoch": 1.2177558402920992, "grad_norm": 0.2529674172401428, "learning_rate": 7.592161946374147e-05, "loss": 0.0505, "step": 36020 }, { "epoch": 1.2180939179823524, "grad_norm": 0.26093173027038574, "learning_rate": 7.590747891830381e-05, "loss": 0.0521, "step": 36030 }, { "epoch": 1.2184319956726055, "grad_norm": 0.22551614046096802, "learning_rate": 7.58933355396632e-05, "loss": 0.0511, "step": 36040 }, { "epoch": 1.2187700733628588, "grad_norm": 0.2652382552623749, "learning_rate": 7.587918932936636e-05, "loss": 0.0564, "step": 36050 }, { "epoch": 1.219108151053112, "grad_norm": 0.22306722402572632, "learning_rate": 7.586504028896028e-05, "loss": 0.0524, "step": 36060 }, { "epoch": 1.2194462287433652, "grad_norm": 0.32717832922935486, "learning_rate": 7.585088841999228e-05, "loss": 0.055, "step": 36070 }, { "epoch": 1.2197843064336185, "grad_norm": 0.2527722716331482, "learning_rate": 7.583673372400999e-05, "loss": 0.0525, "step": 36080 }, { "epoch": 1.2201223841238718, "grad_norm": 0.3229109048843384, "learning_rate": 7.582257620256134e-05, "loss": 0.0526, "step": 36090 }, { "epoch": 1.2204604618141248, "grad_norm": 0.20171299576759338, "learning_rate": 7.580841585719458e-05, "loss": 0.0507, "step": 36100 }, { "epoch": 1.220798539504378, "grad_norm": 0.2915162444114685, "learning_rate": 7.579425268945825e-05, "loss": 0.0497, "step": 36110 }, { "epoch": 1.2211366171946314, "grad_norm": 0.3632347881793976, "learning_rate": 7.578008670090127e-05, "loss": 0.0484, "step": 36120 }, { "epoch": 1.2214746948848845, "grad_norm": 0.22610566020011902, "learning_rate": 7.576591789307272e-05, "loss": 0.0563, "step": 36130 }, { "epoch": 1.2218127725751378, "grad_norm": 0.253335565328598, "learning_rate": 7.575174626752216e-05, "loss": 0.0534, "step": 36140 }, { "epoch": 1.222150850265391, "grad_norm": 0.5518985390663147, "learning_rate": 7.573757182579934e-05, "loss": 0.0599, "step": 36150 }, { "epoch": 1.2224889279556441, "grad_norm": 0.32133862376213074, "learning_rate": 7.572339456945435e-05, "loss": 0.0514, "step": 36160 }, { "epoch": 1.2228270056458974, "grad_norm": 0.25227048993110657, "learning_rate": 7.570921450003762e-05, "loss": 0.0534, "step": 36170 }, { "epoch": 1.2231650833361507, "grad_norm": 0.300131231546402, "learning_rate": 7.569503161909983e-05, "loss": 0.0535, "step": 36180 }, { "epoch": 1.2235031610264038, "grad_norm": 0.18744541704654694, "learning_rate": 7.568084592819202e-05, "loss": 0.0503, "step": 36190 }, { "epoch": 1.223841238716657, "grad_norm": 0.24215202033519745, "learning_rate": 7.566665742886551e-05, "loss": 0.0548, "step": 36200 }, { "epoch": 1.2241793164069104, "grad_norm": 0.39674314856529236, "learning_rate": 7.565246612267194e-05, "loss": 0.0528, "step": 36210 }, { "epoch": 1.2245173940971634, "grad_norm": 0.22982901334762573, "learning_rate": 7.563827201116325e-05, "loss": 0.0604, "step": 36220 }, { "epoch": 1.2248554717874167, "grad_norm": 0.22344297170639038, "learning_rate": 7.562407509589164e-05, "loss": 0.0595, "step": 36230 }, { "epoch": 1.22519354947767, "grad_norm": 0.2196234166622162, "learning_rate": 7.560987537840973e-05, "loss": 0.062, "step": 36240 }, { "epoch": 1.225531627167923, "grad_norm": 0.18538956344127655, "learning_rate": 7.559567286027036e-05, "loss": 0.0491, "step": 36250 }, { "epoch": 1.2258697048581764, "grad_norm": 0.2770664691925049, "learning_rate": 7.558146754302668e-05, "loss": 0.0578, "step": 36260 }, { "epoch": 1.2262077825484297, "grad_norm": 0.23136505484580994, "learning_rate": 7.556725942823217e-05, "loss": 0.056, "step": 36270 }, { "epoch": 1.2265458602386827, "grad_norm": 0.30267342925071716, "learning_rate": 7.555304851744061e-05, "loss": 0.0557, "step": 36280 }, { "epoch": 1.226883937928936, "grad_norm": 0.25048384070396423, "learning_rate": 7.553883481220608e-05, "loss": 0.0581, "step": 36290 }, { "epoch": 1.2272220156191893, "grad_norm": 0.3135499954223633, "learning_rate": 7.552461831408298e-05, "loss": 0.0582, "step": 36300 }, { "epoch": 1.2275600933094426, "grad_norm": 0.19371765851974487, "learning_rate": 7.551039902462599e-05, "loss": 0.053, "step": 36310 }, { "epoch": 1.2278981709996957, "grad_norm": 0.364955335855484, "learning_rate": 7.549617694539011e-05, "loss": 0.0595, "step": 36320 }, { "epoch": 1.228236248689949, "grad_norm": 0.2483416050672531, "learning_rate": 7.548195207793066e-05, "loss": 0.0601, "step": 36330 }, { "epoch": 1.2285743263802023, "grad_norm": 0.30101948976516724, "learning_rate": 7.546772442380323e-05, "loss": 0.0493, "step": 36340 }, { "epoch": 1.2289124040704553, "grad_norm": 0.3591409921646118, "learning_rate": 7.545349398456376e-05, "loss": 0.0616, "step": 36350 }, { "epoch": 1.2292504817607086, "grad_norm": 0.3237159848213196, "learning_rate": 7.543926076176845e-05, "loss": 0.0547, "step": 36360 }, { "epoch": 1.229588559450962, "grad_norm": 0.34577739238739014, "learning_rate": 7.542502475697385e-05, "loss": 0.0589, "step": 36370 }, { "epoch": 1.229926637141215, "grad_norm": 0.38498562574386597, "learning_rate": 7.541078597173675e-05, "loss": 0.0582, "step": 36380 }, { "epoch": 1.2302647148314683, "grad_norm": 0.2443719506263733, "learning_rate": 7.539654440761431e-05, "loss": 0.0574, "step": 36390 }, { "epoch": 1.2306027925217216, "grad_norm": 0.2598191499710083, "learning_rate": 7.538230006616395e-05, "loss": 0.0572, "step": 36400 }, { "epoch": 1.2309408702119746, "grad_norm": 0.20964466035366058, "learning_rate": 7.536805294894342e-05, "loss": 0.054, "step": 36410 }, { "epoch": 1.231278947902228, "grad_norm": 0.2791498899459839, "learning_rate": 7.535380305751076e-05, "loss": 0.0559, "step": 36420 }, { "epoch": 1.2316170255924812, "grad_norm": 0.27843520045280457, "learning_rate": 7.533955039342431e-05, "loss": 0.0582, "step": 36430 }, { "epoch": 1.2319551032827343, "grad_norm": 0.30971044301986694, "learning_rate": 7.532529495824274e-05, "loss": 0.0517, "step": 36440 }, { "epoch": 1.2322931809729876, "grad_norm": 0.23759211599826813, "learning_rate": 7.531103675352497e-05, "loss": 0.0545, "step": 36450 }, { "epoch": 1.2326312586632409, "grad_norm": 0.21631158888339996, "learning_rate": 7.529677578083029e-05, "loss": 0.0576, "step": 36460 }, { "epoch": 1.2329693363534941, "grad_norm": 0.17709249258041382, "learning_rate": 7.528251204171823e-05, "loss": 0.0511, "step": 36470 }, { "epoch": 1.2333074140437472, "grad_norm": 0.2900589406490326, "learning_rate": 7.526824553774867e-05, "loss": 0.0609, "step": 36480 }, { "epoch": 1.2336454917340005, "grad_norm": 0.27302131056785583, "learning_rate": 7.525397627048178e-05, "loss": 0.0605, "step": 36490 }, { "epoch": 1.2339835694242538, "grad_norm": 0.24260006844997406, "learning_rate": 7.523970424147802e-05, "loss": 0.0578, "step": 36500 }, { "epoch": 1.2343216471145069, "grad_norm": 0.9954864382743835, "learning_rate": 7.522542945229813e-05, "loss": 0.0465, "step": 36510 }, { "epoch": 1.2346597248047602, "grad_norm": 0.18474125862121582, "learning_rate": 7.521115190450321e-05, "loss": 0.0536, "step": 36520 }, { "epoch": 1.2349978024950135, "grad_norm": 0.2589479684829712, "learning_rate": 7.519687159965462e-05, "loss": 0.0544, "step": 36530 }, { "epoch": 1.2353358801852665, "grad_norm": 0.23638558387756348, "learning_rate": 7.518258853931403e-05, "loss": 0.0537, "step": 36540 }, { "epoch": 1.2356739578755198, "grad_norm": 0.2661285102367401, "learning_rate": 7.516830272504342e-05, "loss": 0.0544, "step": 36550 }, { "epoch": 1.236012035565773, "grad_norm": 0.2742113769054413, "learning_rate": 7.515401415840509e-05, "loss": 0.0529, "step": 36560 }, { "epoch": 1.2363501132560262, "grad_norm": 0.4347390830516815, "learning_rate": 7.513972284096155e-05, "loss": 0.0526, "step": 36570 }, { "epoch": 1.2366881909462795, "grad_norm": 0.22300441563129425, "learning_rate": 7.512542877427574e-05, "loss": 0.0533, "step": 36580 }, { "epoch": 1.2370262686365328, "grad_norm": 0.1827065348625183, "learning_rate": 7.51111319599108e-05, "loss": 0.0537, "step": 36590 }, { "epoch": 1.2373643463267858, "grad_norm": 0.20243024826049805, "learning_rate": 7.509683239943024e-05, "loss": 0.0533, "step": 36600 }, { "epoch": 1.2377024240170391, "grad_norm": 0.23941615223884583, "learning_rate": 7.50825300943978e-05, "loss": 0.0533, "step": 36610 }, { "epoch": 1.2380405017072924, "grad_norm": 0.4674052894115448, "learning_rate": 7.50682250463776e-05, "loss": 0.0594, "step": 36620 }, { "epoch": 1.2383785793975455, "grad_norm": 0.17833703756332397, "learning_rate": 7.5053917256934e-05, "loss": 0.056, "step": 36630 }, { "epoch": 1.2387166570877988, "grad_norm": 0.3050706088542938, "learning_rate": 7.503960672763166e-05, "loss": 0.0597, "step": 36640 }, { "epoch": 1.239054734778052, "grad_norm": 0.22491398453712463, "learning_rate": 7.502529346003559e-05, "loss": 0.0553, "step": 36650 }, { "epoch": 1.2393928124683051, "grad_norm": 0.2034171223640442, "learning_rate": 7.501097745571106e-05, "loss": 0.0498, "step": 36660 }, { "epoch": 1.2397308901585584, "grad_norm": 0.23324820399284363, "learning_rate": 7.499665871622361e-05, "loss": 0.0585, "step": 36670 }, { "epoch": 1.2400689678488117, "grad_norm": 0.37648725509643555, "learning_rate": 7.498233724313917e-05, "loss": 0.0569, "step": 36680 }, { "epoch": 1.2404070455390648, "grad_norm": 0.32367652654647827, "learning_rate": 7.496801303802387e-05, "loss": 0.0558, "step": 36690 }, { "epoch": 1.240745123229318, "grad_norm": 0.3318970203399658, "learning_rate": 7.495368610244423e-05, "loss": 0.054, "step": 36700 }, { "epoch": 1.2410832009195714, "grad_norm": 0.28647688031196594, "learning_rate": 7.493935643796697e-05, "loss": 0.0606, "step": 36710 }, { "epoch": 1.2414212786098244, "grad_norm": 0.23632168769836426, "learning_rate": 7.49250240461592e-05, "loss": 0.0507, "step": 36720 }, { "epoch": 1.2417593563000777, "grad_norm": 0.2484893798828125, "learning_rate": 7.491068892858826e-05, "loss": 0.0512, "step": 36730 }, { "epoch": 1.242097433990331, "grad_norm": 0.28194311261177063, "learning_rate": 7.489635108682184e-05, "loss": 0.0534, "step": 36740 }, { "epoch": 1.242435511680584, "grad_norm": 0.3329428434371948, "learning_rate": 7.48820105224279e-05, "loss": 0.0542, "step": 36750 }, { "epoch": 1.2427735893708374, "grad_norm": 0.25384098291397095, "learning_rate": 7.48676672369747e-05, "loss": 0.055, "step": 36760 }, { "epoch": 1.2431116670610907, "grad_norm": 0.21519851684570312, "learning_rate": 7.485332123203079e-05, "loss": 0.0578, "step": 36770 }, { "epoch": 1.243449744751344, "grad_norm": 0.32487326860427856, "learning_rate": 7.483897250916504e-05, "loss": 0.0545, "step": 36780 }, { "epoch": 1.243787822441597, "grad_norm": 0.5113327503204346, "learning_rate": 7.48246210699466e-05, "loss": 0.0574, "step": 36790 }, { "epoch": 1.2441259001318503, "grad_norm": 0.28647515177726746, "learning_rate": 7.481026691594492e-05, "loss": 0.0536, "step": 36800 }, { "epoch": 1.2444639778221036, "grad_norm": 0.26662394404411316, "learning_rate": 7.479591004872975e-05, "loss": 0.0527, "step": 36810 }, { "epoch": 1.2448020555123567, "grad_norm": 0.4767603576183319, "learning_rate": 7.478155046987114e-05, "loss": 0.0523, "step": 36820 }, { "epoch": 1.24514013320261, "grad_norm": 0.24972453713417053, "learning_rate": 7.476718818093944e-05, "loss": 0.0541, "step": 36830 }, { "epoch": 1.2454782108928633, "grad_norm": 0.20952551066875458, "learning_rate": 7.475282318350527e-05, "loss": 0.0561, "step": 36840 }, { "epoch": 1.2458162885831163, "grad_norm": 0.27630415558815, "learning_rate": 7.473845547913957e-05, "loss": 0.0493, "step": 36850 }, { "epoch": 1.2461543662733696, "grad_norm": 0.29575544595718384, "learning_rate": 7.472408506941357e-05, "loss": 0.0589, "step": 36860 }, { "epoch": 1.246492443963623, "grad_norm": 0.353204607963562, "learning_rate": 7.470971195589881e-05, "loss": 0.0583, "step": 36870 }, { "epoch": 1.246830521653876, "grad_norm": 0.2218170315027237, "learning_rate": 7.46953361401671e-05, "loss": 0.0561, "step": 36880 }, { "epoch": 1.2471685993441293, "grad_norm": 0.23701642453670502, "learning_rate": 7.468095762379055e-05, "loss": 0.0557, "step": 36890 }, { "epoch": 1.2475066770343826, "grad_norm": 0.31177031993865967, "learning_rate": 7.466657640834158e-05, "loss": 0.0627, "step": 36900 }, { "epoch": 1.2478447547246356, "grad_norm": 0.20775236189365387, "learning_rate": 7.46521924953929e-05, "loss": 0.0535, "step": 36910 }, { "epoch": 1.248182832414889, "grad_norm": 0.32582005858421326, "learning_rate": 7.463780588651752e-05, "loss": 0.0535, "step": 36920 }, { "epoch": 1.2485209101051422, "grad_norm": 0.4705912470817566, "learning_rate": 7.462341658328873e-05, "loss": 0.0572, "step": 36930 }, { "epoch": 1.2488589877953955, "grad_norm": 0.28151872754096985, "learning_rate": 7.460902458728012e-05, "loss": 0.0545, "step": 36940 }, { "epoch": 1.2491970654856486, "grad_norm": 0.17798475921154022, "learning_rate": 7.459462990006558e-05, "loss": 0.0505, "step": 36950 }, { "epoch": 1.2495351431759019, "grad_norm": 0.22797587513923645, "learning_rate": 7.458023252321929e-05, "loss": 0.0568, "step": 36960 }, { "epoch": 1.2498732208661552, "grad_norm": 0.24319656193256378, "learning_rate": 7.456583245831573e-05, "loss": 0.0569, "step": 36970 }, { "epoch": 1.2502112985564082, "grad_norm": 0.2968210279941559, "learning_rate": 7.455142970692968e-05, "loss": 0.0558, "step": 36980 }, { "epoch": 1.2505493762466615, "grad_norm": 0.21005868911743164, "learning_rate": 7.453702427063615e-05, "loss": 0.0549, "step": 36990 }, { "epoch": 1.2508874539369148, "grad_norm": 0.323465496301651, "learning_rate": 7.452261615101057e-05, "loss": 0.0535, "step": 37000 }, { "epoch": 1.2512255316271679, "grad_norm": 0.19860079884529114, "learning_rate": 7.450820534962856e-05, "loss": 0.062, "step": 37010 }, { "epoch": 1.2515636093174212, "grad_norm": 0.3991100788116455, "learning_rate": 7.449379186806604e-05, "loss": 0.0501, "step": 37020 }, { "epoch": 1.2519016870076745, "grad_norm": 0.34061399102211, "learning_rate": 7.44793757078993e-05, "loss": 0.0561, "step": 37030 }, { "epoch": 1.2522397646979275, "grad_norm": 0.2469596564769745, "learning_rate": 7.44649568707048e-05, "loss": 0.0589, "step": 37040 }, { "epoch": 1.2525778423881808, "grad_norm": 0.25261563062667847, "learning_rate": 7.445053535805942e-05, "loss": 0.0528, "step": 37050 }, { "epoch": 1.252915920078434, "grad_norm": 0.28861501812934875, "learning_rate": 7.443611117154024e-05, "loss": 0.0602, "step": 37060 }, { "epoch": 1.2532539977686872, "grad_norm": 0.22406983375549316, "learning_rate": 7.442168431272469e-05, "loss": 0.0549, "step": 37070 }, { "epoch": 1.2535920754589405, "grad_norm": 0.3536509573459625, "learning_rate": 7.440725478319044e-05, "loss": 0.0529, "step": 37080 }, { "epoch": 1.2539301531491938, "grad_norm": 0.244282066822052, "learning_rate": 7.439282258451553e-05, "loss": 0.0515, "step": 37090 }, { "epoch": 1.2542682308394468, "grad_norm": 0.32130929827690125, "learning_rate": 7.43783877182782e-05, "loss": 0.052, "step": 37100 }, { "epoch": 1.2546063085297001, "grad_norm": 0.3807469606399536, "learning_rate": 7.436395018605707e-05, "loss": 0.0554, "step": 37110 }, { "epoch": 1.2549443862199534, "grad_norm": 0.19267776608467102, "learning_rate": 7.434950998943094e-05, "loss": 0.056, "step": 37120 }, { "epoch": 1.2552824639102065, "grad_norm": 0.19789855182170868, "learning_rate": 7.433506712997902e-05, "loss": 0.055, "step": 37130 }, { "epoch": 1.2556205416004598, "grad_norm": 0.24135830998420715, "learning_rate": 7.432062160928077e-05, "loss": 0.0558, "step": 37140 }, { "epoch": 1.255958619290713, "grad_norm": 0.21604669094085693, "learning_rate": 7.430617342891588e-05, "loss": 0.0496, "step": 37150 }, { "epoch": 1.2562966969809661, "grad_norm": 0.2643934488296509, "learning_rate": 7.429172259046442e-05, "loss": 0.0535, "step": 37160 }, { "epoch": 1.2566347746712194, "grad_norm": 0.5606066584587097, "learning_rate": 7.42772690955067e-05, "loss": 0.0532, "step": 37170 }, { "epoch": 1.2569728523614727, "grad_norm": 0.19311103224754333, "learning_rate": 7.426281294562332e-05, "loss": 0.0543, "step": 37180 }, { "epoch": 1.2573109300517258, "grad_norm": 0.18100404739379883, "learning_rate": 7.424835414239519e-05, "loss": 0.0579, "step": 37190 }, { "epoch": 1.257649007741979, "grad_norm": 0.21632349491119385, "learning_rate": 7.423389268740352e-05, "loss": 0.0533, "step": 37200 }, { "epoch": 1.2579870854322324, "grad_norm": 0.2167925238609314, "learning_rate": 7.42194285822298e-05, "loss": 0.0558, "step": 37210 }, { "epoch": 1.2583251631224854, "grad_norm": 0.23676711320877075, "learning_rate": 7.420496182845576e-05, "loss": 0.0516, "step": 37220 }, { "epoch": 1.2586632408127387, "grad_norm": 0.1953379064798355, "learning_rate": 7.419049242766348e-05, "loss": 0.0528, "step": 37230 }, { "epoch": 1.259001318502992, "grad_norm": 0.36601629853248596, "learning_rate": 7.417602038143533e-05, "loss": 0.0532, "step": 37240 }, { "epoch": 1.259339396193245, "grad_norm": 0.46395519375801086, "learning_rate": 7.416154569135393e-05, "loss": 0.0595, "step": 37250 }, { "epoch": 1.2596774738834984, "grad_norm": 0.1571323126554489, "learning_rate": 7.414706835900222e-05, "loss": 0.0562, "step": 37260 }, { "epoch": 1.2600155515737517, "grad_norm": 0.2511904835700989, "learning_rate": 7.413258838596345e-05, "loss": 0.0533, "step": 37270 }, { "epoch": 1.2603536292640047, "grad_norm": 0.29210522770881653, "learning_rate": 7.411810577382106e-05, "loss": 0.0561, "step": 37280 }, { "epoch": 1.260691706954258, "grad_norm": 0.2161874920129776, "learning_rate": 7.410362052415889e-05, "loss": 0.0565, "step": 37290 }, { "epoch": 1.2610297846445113, "grad_norm": 0.33218443393707275, "learning_rate": 7.408913263856102e-05, "loss": 0.0505, "step": 37300 }, { "epoch": 1.2613678623347646, "grad_norm": 0.19950781762599945, "learning_rate": 7.407464211861183e-05, "loss": 0.0509, "step": 37310 }, { "epoch": 1.2617059400250177, "grad_norm": 0.23817819356918335, "learning_rate": 7.406014896589597e-05, "loss": 0.0526, "step": 37320 }, { "epoch": 1.262044017715271, "grad_norm": 0.28937724232673645, "learning_rate": 7.404565318199838e-05, "loss": 0.0559, "step": 37330 }, { "epoch": 1.2623820954055243, "grad_norm": 0.28352946043014526, "learning_rate": 7.403115476850432e-05, "loss": 0.049, "step": 37340 }, { "epoch": 1.2627201730957773, "grad_norm": 0.3871764838695526, "learning_rate": 7.401665372699932e-05, "loss": 0.0511, "step": 37350 }, { "epoch": 1.2630582507860306, "grad_norm": 0.2284667193889618, "learning_rate": 7.400215005906916e-05, "loss": 0.0559, "step": 37360 }, { "epoch": 1.263396328476284, "grad_norm": 0.18690407276153564, "learning_rate": 7.398764376629996e-05, "loss": 0.0537, "step": 37370 }, { "epoch": 1.2637344061665372, "grad_norm": 0.2609858810901642, "learning_rate": 7.39731348502781e-05, "loss": 0.0525, "step": 37380 }, { "epoch": 1.2640724838567903, "grad_norm": 0.20454493165016174, "learning_rate": 7.395862331259028e-05, "loss": 0.052, "step": 37390 }, { "epoch": 1.2644105615470436, "grad_norm": 0.44043827056884766, "learning_rate": 7.394410915482341e-05, "loss": 0.0533, "step": 37400 }, { "epoch": 1.2647486392372969, "grad_norm": 0.22579815983772278, "learning_rate": 7.392959237856476e-05, "loss": 0.0533, "step": 37410 }, { "epoch": 1.26508671692755, "grad_norm": 0.32801511883735657, "learning_rate": 7.391507298540187e-05, "loss": 0.0566, "step": 37420 }, { "epoch": 1.2654247946178032, "grad_norm": 0.3226386606693268, "learning_rate": 7.390055097692257e-05, "loss": 0.0528, "step": 37430 }, { "epoch": 1.2657628723080565, "grad_norm": 0.27907317876815796, "learning_rate": 7.388602635471494e-05, "loss": 0.0499, "step": 37440 }, { "epoch": 1.2661009499983096, "grad_norm": 0.3043903708457947, "learning_rate": 7.387149912036738e-05, "loss": 0.05, "step": 37450 }, { "epoch": 1.2664390276885629, "grad_norm": 0.3729109466075897, "learning_rate": 7.385696927546854e-05, "loss": 0.0487, "step": 37460 }, { "epoch": 1.2667771053788162, "grad_norm": 0.33781197667121887, "learning_rate": 7.384243682160743e-05, "loss": 0.0544, "step": 37470 }, { "epoch": 1.2671151830690692, "grad_norm": 0.2633560001850128, "learning_rate": 7.382790176037329e-05, "loss": 0.0534, "step": 37480 }, { "epoch": 1.2674532607593225, "grad_norm": 0.22622355818748474, "learning_rate": 7.381336409335562e-05, "loss": 0.0521, "step": 37490 }, { "epoch": 1.2677913384495758, "grad_norm": 0.20415940880775452, "learning_rate": 7.379882382214426e-05, "loss": 0.0486, "step": 37500 }, { "epoch": 1.2681294161398289, "grad_norm": 0.3991987705230713, "learning_rate": 7.378428094832931e-05, "loss": 0.0611, "step": 37510 }, { "epoch": 1.2684674938300822, "grad_norm": 0.2704184651374817, "learning_rate": 7.376973547350114e-05, "loss": 0.0566, "step": 37520 }, { "epoch": 1.2688055715203355, "grad_norm": 0.22279757261276245, "learning_rate": 7.375518739925046e-05, "loss": 0.053, "step": 37530 }, { "epoch": 1.2691436492105885, "grad_norm": 0.22321408987045288, "learning_rate": 7.374063672716818e-05, "loss": 0.0535, "step": 37540 }, { "epoch": 1.2694817269008418, "grad_norm": 0.20176273584365845, "learning_rate": 7.372608345884558e-05, "loss": 0.0484, "step": 37550 }, { "epoch": 1.2698198045910951, "grad_norm": 0.2129514515399933, "learning_rate": 7.371152759587415e-05, "loss": 0.0557, "step": 37560 }, { "epoch": 1.2701578822813482, "grad_norm": 0.2357541173696518, "learning_rate": 7.369696913984572e-05, "loss": 0.0544, "step": 37570 }, { "epoch": 1.2704959599716015, "grad_norm": 0.27247336506843567, "learning_rate": 7.368240809235237e-05, "loss": 0.0543, "step": 37580 }, { "epoch": 1.2708340376618548, "grad_norm": 0.2771092355251312, "learning_rate": 7.366784445498649e-05, "loss": 0.0574, "step": 37590 }, { "epoch": 1.2711721153521078, "grad_norm": 0.30067288875579834, "learning_rate": 7.36532782293407e-05, "loss": 0.056, "step": 37600 }, { "epoch": 1.2715101930423611, "grad_norm": 0.2969907522201538, "learning_rate": 7.363870941700797e-05, "loss": 0.0546, "step": 37610 }, { "epoch": 1.2718482707326144, "grad_norm": 0.21145953238010406, "learning_rate": 7.362413801958152e-05, "loss": 0.0552, "step": 37620 }, { "epoch": 1.2721863484228675, "grad_norm": 0.20895551145076752, "learning_rate": 7.360956403865486e-05, "loss": 0.0578, "step": 37630 }, { "epoch": 1.2725244261131208, "grad_norm": 0.2256636768579483, "learning_rate": 7.359498747582177e-05, "loss": 0.0508, "step": 37640 }, { "epoch": 1.272862503803374, "grad_norm": 0.25911852717399597, "learning_rate": 7.358040833267634e-05, "loss": 0.0499, "step": 37650 }, { "epoch": 1.2732005814936271, "grad_norm": 0.5241990089416504, "learning_rate": 7.356582661081288e-05, "loss": 0.0492, "step": 37660 }, { "epoch": 1.2735386591838804, "grad_norm": 0.24908407032489777, "learning_rate": 7.355124231182607e-05, "loss": 0.0543, "step": 37670 }, { "epoch": 1.2738767368741337, "grad_norm": 0.21813131868839264, "learning_rate": 7.353665543731083e-05, "loss": 0.0509, "step": 37680 }, { "epoch": 1.2742148145643868, "grad_norm": 0.22155669331550598, "learning_rate": 7.352206598886229e-05, "loss": 0.0582, "step": 37690 }, { "epoch": 1.27455289225464, "grad_norm": 0.387868732213974, "learning_rate": 7.350747396807601e-05, "loss": 0.058, "step": 37700 }, { "epoch": 1.2748909699448934, "grad_norm": 0.39346760511398315, "learning_rate": 7.349287937654772e-05, "loss": 0.0616, "step": 37710 }, { "epoch": 1.2752290476351464, "grad_norm": 0.1794426143169403, "learning_rate": 7.347828221587345e-05, "loss": 0.0533, "step": 37720 }, { "epoch": 1.2755671253253997, "grad_norm": 0.26029330492019653, "learning_rate": 7.346368248764955e-05, "loss": 0.0561, "step": 37730 }, { "epoch": 1.275905203015653, "grad_norm": 0.25802335143089294, "learning_rate": 7.34490801934726e-05, "loss": 0.0604, "step": 37740 }, { "epoch": 1.276243280705906, "grad_norm": 0.22798261046409607, "learning_rate": 7.343447533493947e-05, "loss": 0.0496, "step": 37750 }, { "epoch": 1.2765813583961594, "grad_norm": 0.2115008533000946, "learning_rate": 7.341986791364739e-05, "loss": 0.0553, "step": 37760 }, { "epoch": 1.2769194360864127, "grad_norm": 0.3039129972457886, "learning_rate": 7.340525793119374e-05, "loss": 0.0567, "step": 37770 }, { "epoch": 1.2772575137766657, "grad_norm": 0.27002161741256714, "learning_rate": 7.33906453891763e-05, "loss": 0.0613, "step": 37780 }, { "epoch": 1.277595591466919, "grad_norm": 0.4861888587474823, "learning_rate": 7.337603028919303e-05, "loss": 0.0528, "step": 37790 }, { "epoch": 1.2779336691571723, "grad_norm": 0.22501540184020996, "learning_rate": 7.336141263284225e-05, "loss": 0.0538, "step": 37800 }, { "epoch": 1.2782717468474256, "grad_norm": 0.19920869171619415, "learning_rate": 7.334679242172249e-05, "loss": 0.0601, "step": 37810 }, { "epoch": 1.2786098245376787, "grad_norm": 0.34829673171043396, "learning_rate": 7.33321696574326e-05, "loss": 0.0548, "step": 37820 }, { "epoch": 1.278947902227932, "grad_norm": 0.24457219243049622, "learning_rate": 7.331754434157174e-05, "loss": 0.0571, "step": 37830 }, { "epoch": 1.2792859799181853, "grad_norm": 0.5915870666503906, "learning_rate": 7.330291647573928e-05, "loss": 0.0561, "step": 37840 }, { "epoch": 1.2796240576084383, "grad_norm": 0.244615375995636, "learning_rate": 7.32882860615349e-05, "loss": 0.0543, "step": 37850 }, { "epoch": 1.2799621352986916, "grad_norm": 0.2603270709514618, "learning_rate": 7.327365310055858e-05, "loss": 0.0515, "step": 37860 }, { "epoch": 1.280300212988945, "grad_norm": 0.22869186103343964, "learning_rate": 7.325901759441055e-05, "loss": 0.0604, "step": 37870 }, { "epoch": 1.2806382906791982, "grad_norm": 0.2842046916484833, "learning_rate": 7.324437954469132e-05, "loss": 0.0585, "step": 37880 }, { "epoch": 1.2809763683694513, "grad_norm": 0.2428411841392517, "learning_rate": 7.32297389530017e-05, "loss": 0.0454, "step": 37890 }, { "epoch": 1.2813144460597046, "grad_norm": 0.2498951107263565, "learning_rate": 7.321509582094274e-05, "loss": 0.0568, "step": 37900 }, { "epoch": 1.2816525237499579, "grad_norm": 0.31425613164901733, "learning_rate": 7.320045015011582e-05, "loss": 0.0577, "step": 37910 }, { "epoch": 1.281990601440211, "grad_norm": 0.3753446340560913, "learning_rate": 7.318580194212253e-05, "loss": 0.0547, "step": 37920 }, { "epoch": 1.2823286791304642, "grad_norm": 0.3649551570415497, "learning_rate": 7.317115119856483e-05, "loss": 0.0523, "step": 37930 }, { "epoch": 1.2826667568207175, "grad_norm": 0.36134159564971924, "learning_rate": 7.315649792104485e-05, "loss": 0.0599, "step": 37940 }, { "epoch": 1.2830048345109706, "grad_norm": 0.31220927834510803, "learning_rate": 7.314184211116507e-05, "loss": 0.0542, "step": 37950 }, { "epoch": 1.2833429122012239, "grad_norm": 0.39567315578460693, "learning_rate": 7.312718377052823e-05, "loss": 0.058, "step": 37960 }, { "epoch": 1.2836809898914772, "grad_norm": 0.3196766972541809, "learning_rate": 7.311252290073737e-05, "loss": 0.0536, "step": 37970 }, { "epoch": 1.2840190675817302, "grad_norm": 0.2528131902217865, "learning_rate": 7.309785950339572e-05, "loss": 0.0524, "step": 37980 }, { "epoch": 1.2843571452719835, "grad_norm": 0.276751309633255, "learning_rate": 7.308319358010692e-05, "loss": 0.0523, "step": 37990 }, { "epoch": 1.2846952229622368, "grad_norm": 0.20710361003875732, "learning_rate": 7.306852513247474e-05, "loss": 0.0553, "step": 38000 }, { "epoch": 1.2850333006524899, "grad_norm": 0.22985802590847015, "learning_rate": 7.305385416210336e-05, "loss": 0.0554, "step": 38010 }, { "epoch": 1.2853713783427432, "grad_norm": 0.21299688518047333, "learning_rate": 7.303918067059711e-05, "loss": 0.0541, "step": 38020 }, { "epoch": 1.2857094560329965, "grad_norm": 0.2645083963871002, "learning_rate": 7.302450465956075e-05, "loss": 0.0506, "step": 38030 }, { "epoch": 1.2860475337232495, "grad_norm": 0.23467956483364105, "learning_rate": 7.300982613059914e-05, "loss": 0.0534, "step": 38040 }, { "epoch": 1.2863856114135028, "grad_norm": 0.27293795347213745, "learning_rate": 7.299514508531757e-05, "loss": 0.0495, "step": 38050 }, { "epoch": 1.2867236891037561, "grad_norm": 0.27657952904701233, "learning_rate": 7.298046152532148e-05, "loss": 0.0522, "step": 38060 }, { "epoch": 1.2870617667940092, "grad_norm": 0.2240726500749588, "learning_rate": 7.296577545221668e-05, "loss": 0.0495, "step": 38070 }, { "epoch": 1.2873998444842625, "grad_norm": 0.3758934438228607, "learning_rate": 7.29510868676092e-05, "loss": 0.0553, "step": 38080 }, { "epoch": 1.2877379221745158, "grad_norm": 0.2931390702724457, "learning_rate": 7.293639577310538e-05, "loss": 0.0516, "step": 38090 }, { "epoch": 1.2880759998647688, "grad_norm": 0.265567421913147, "learning_rate": 7.292170217031179e-05, "loss": 0.0547, "step": 38100 }, { "epoch": 1.2884140775550221, "grad_norm": 0.2458907961845398, "learning_rate": 7.290700606083532e-05, "loss": 0.0576, "step": 38110 }, { "epoch": 1.2887521552452754, "grad_norm": 0.1982714980840683, "learning_rate": 7.28923074462831e-05, "loss": 0.0535, "step": 38120 }, { "epoch": 1.2890902329355285, "grad_norm": 0.178690567612648, "learning_rate": 7.287760632826255e-05, "loss": 0.0511, "step": 38130 }, { "epoch": 1.2894283106257818, "grad_norm": 0.45369404554367065, "learning_rate": 7.286290270838139e-05, "loss": 0.0514, "step": 38140 }, { "epoch": 1.289766388316035, "grad_norm": 0.2775188386440277, "learning_rate": 7.284819658824756e-05, "loss": 0.0501, "step": 38150 }, { "epoch": 1.2901044660062881, "grad_norm": 0.2902385890483856, "learning_rate": 7.283348796946929e-05, "loss": 0.0605, "step": 38160 }, { "epoch": 1.2904425436965414, "grad_norm": 0.35200658440589905, "learning_rate": 7.28187768536551e-05, "loss": 0.0526, "step": 38170 }, { "epoch": 1.2907806213867947, "grad_norm": 0.27446797490119934, "learning_rate": 7.280406324241379e-05, "loss": 0.0603, "step": 38180 }, { "epoch": 1.2911186990770478, "grad_norm": 0.2786177694797516, "learning_rate": 7.278934713735441e-05, "loss": 0.0559, "step": 38190 }, { "epoch": 1.291456776767301, "grad_norm": 0.25655755400657654, "learning_rate": 7.277462854008629e-05, "loss": 0.0507, "step": 38200 }, { "epoch": 1.2917948544575544, "grad_norm": 0.18652144074440002, "learning_rate": 7.275990745221904e-05, "loss": 0.051, "step": 38210 }, { "epoch": 1.2921329321478074, "grad_norm": 0.3004320561885834, "learning_rate": 7.274518387536252e-05, "loss": 0.0589, "step": 38220 }, { "epoch": 1.2924710098380607, "grad_norm": 0.2720760405063629, "learning_rate": 7.273045781112689e-05, "loss": 0.0547, "step": 38230 }, { "epoch": 1.292809087528314, "grad_norm": 0.2565475404262543, "learning_rate": 7.271572926112257e-05, "loss": 0.0529, "step": 38240 }, { "epoch": 1.293147165218567, "grad_norm": 0.5993428230285645, "learning_rate": 7.270099822696024e-05, "loss": 0.0496, "step": 38250 }, { "epoch": 1.2934852429088204, "grad_norm": 0.3825119435787201, "learning_rate": 7.26862647102509e-05, "loss": 0.0547, "step": 38260 }, { "epoch": 1.2938233205990737, "grad_norm": 0.2007426917552948, "learning_rate": 7.267152871260573e-05, "loss": 0.0516, "step": 38270 }, { "epoch": 1.294161398289327, "grad_norm": 0.3013485074043274, "learning_rate": 7.265679023563629e-05, "loss": 0.0546, "step": 38280 }, { "epoch": 1.29449947597958, "grad_norm": 0.31488311290740967, "learning_rate": 7.264204928095433e-05, "loss": 0.0516, "step": 38290 }, { "epoch": 1.2948375536698333, "grad_norm": 0.2573113739490509, "learning_rate": 7.262730585017188e-05, "loss": 0.055, "step": 38300 }, { "epoch": 1.2951756313600866, "grad_norm": 0.42437323927879333, "learning_rate": 7.261255994490131e-05, "loss": 0.0523, "step": 38310 }, { "epoch": 1.2955137090503397, "grad_norm": 0.22487077116966248, "learning_rate": 7.259781156675517e-05, "loss": 0.0515, "step": 38320 }, { "epoch": 1.295851786740593, "grad_norm": 0.315104603767395, "learning_rate": 7.258306071734633e-05, "loss": 0.0585, "step": 38330 }, { "epoch": 1.2961898644308463, "grad_norm": 0.27415838837623596, "learning_rate": 7.256830739828792e-05, "loss": 0.057, "step": 38340 }, { "epoch": 1.2965279421210996, "grad_norm": 0.8321232199668884, "learning_rate": 7.255355161119336e-05, "loss": 0.0548, "step": 38350 }, { "epoch": 1.2968660198113526, "grad_norm": 0.32829633355140686, "learning_rate": 7.253879335767628e-05, "loss": 0.0499, "step": 38360 }, { "epoch": 1.297204097501606, "grad_norm": 0.3691498935222626, "learning_rate": 7.252403263935067e-05, "loss": 0.0528, "step": 38370 }, { "epoch": 1.2975421751918592, "grad_norm": 0.4320649206638336, "learning_rate": 7.25092694578307e-05, "loss": 0.0533, "step": 38380 }, { "epoch": 1.2978802528821123, "grad_norm": 0.29000794887542725, "learning_rate": 7.249450381473087e-05, "loss": 0.0525, "step": 38390 }, { "epoch": 1.2982183305723656, "grad_norm": 0.9410483241081238, "learning_rate": 7.247973571166593e-05, "loss": 0.0547, "step": 38400 }, { "epoch": 1.2985564082626189, "grad_norm": 0.24113599956035614, "learning_rate": 7.246496515025089e-05, "loss": 0.055, "step": 38410 }, { "epoch": 1.298894485952872, "grad_norm": 0.2541462182998657, "learning_rate": 7.245019213210101e-05, "loss": 0.0495, "step": 38420 }, { "epoch": 1.2992325636431252, "grad_norm": 0.28676140308380127, "learning_rate": 7.24354166588319e-05, "loss": 0.0547, "step": 38430 }, { "epoch": 1.2995706413333785, "grad_norm": 0.2772211730480194, "learning_rate": 7.242063873205935e-05, "loss": 0.0527, "step": 38440 }, { "epoch": 1.2999087190236316, "grad_norm": 0.2762100398540497, "learning_rate": 7.240585835339946e-05, "loss": 0.0571, "step": 38450 }, { "epoch": 1.3002467967138849, "grad_norm": 0.36982545256614685, "learning_rate": 7.239107552446857e-05, "loss": 0.0538, "step": 38460 }, { "epoch": 1.3005848744041382, "grad_norm": 0.2919953763484955, "learning_rate": 7.237629024688334e-05, "loss": 0.0568, "step": 38470 }, { "epoch": 1.3009229520943912, "grad_norm": 0.5046866536140442, "learning_rate": 7.236150252226064e-05, "loss": 0.0565, "step": 38480 }, { "epoch": 1.3012610297846445, "grad_norm": 0.19521240890026093, "learning_rate": 7.234671235221765e-05, "loss": 0.0506, "step": 38490 }, { "epoch": 1.3015991074748978, "grad_norm": 0.17899392545223236, "learning_rate": 7.233191973837179e-05, "loss": 0.0557, "step": 38500 }, { "epoch": 1.301937185165151, "grad_norm": 0.27990812063217163, "learning_rate": 7.231712468234078e-05, "loss": 0.0563, "step": 38510 }, { "epoch": 1.3022752628554042, "grad_norm": 0.18463610112667084, "learning_rate": 7.230232718574254e-05, "loss": 0.0519, "step": 38520 }, { "epoch": 1.3026133405456575, "grad_norm": 0.27255675196647644, "learning_rate": 7.228752725019535e-05, "loss": 0.0469, "step": 38530 }, { "epoch": 1.3029514182359105, "grad_norm": 0.18448463082313538, "learning_rate": 7.227272487731769e-05, "loss": 0.0512, "step": 38540 }, { "epoch": 1.3032894959261638, "grad_norm": 0.22513873875141144, "learning_rate": 7.225792006872831e-05, "loss": 0.0649, "step": 38550 }, { "epoch": 1.3036275736164171, "grad_norm": 0.187239870429039, "learning_rate": 7.224311282604628e-05, "loss": 0.057, "step": 38560 }, { "epoch": 1.3039656513066702, "grad_norm": 0.18793296813964844, "learning_rate": 7.222830315089085e-05, "loss": 0.0534, "step": 38570 }, { "epoch": 1.3043037289969235, "grad_norm": 0.3316923975944519, "learning_rate": 7.22134910448816e-05, "loss": 0.0534, "step": 38580 }, { "epoch": 1.3046418066871768, "grad_norm": 0.20812490582466125, "learning_rate": 7.219867650963839e-05, "loss": 0.0567, "step": 38590 }, { "epoch": 1.3049798843774298, "grad_norm": 0.2131303995847702, "learning_rate": 7.21838595467813e-05, "loss": 0.0519, "step": 38600 }, { "epoch": 1.3053179620676831, "grad_norm": 0.20063407719135284, "learning_rate": 7.216904015793067e-05, "loss": 0.0528, "step": 38610 }, { "epoch": 1.3056560397579364, "grad_norm": 0.19464024901390076, "learning_rate": 7.215421834470713e-05, "loss": 0.0542, "step": 38620 }, { "epoch": 1.3059941174481895, "grad_norm": 0.20434033870697021, "learning_rate": 7.213939410873158e-05, "loss": 0.0538, "step": 38630 }, { "epoch": 1.3063321951384428, "grad_norm": 0.23405073583126068, "learning_rate": 7.212456745162519e-05, "loss": 0.0573, "step": 38640 }, { "epoch": 1.306670272828696, "grad_norm": 0.24220238626003265, "learning_rate": 7.210973837500937e-05, "loss": 0.0549, "step": 38650 }, { "epoch": 1.3070083505189491, "grad_norm": 0.4205106496810913, "learning_rate": 7.209490688050578e-05, "loss": 0.0521, "step": 38660 }, { "epoch": 1.3073464282092024, "grad_norm": 0.18572086095809937, "learning_rate": 7.208007296973641e-05, "loss": 0.0496, "step": 38670 }, { "epoch": 1.3076845058994557, "grad_norm": 0.21917922794818878, "learning_rate": 7.206523664432345e-05, "loss": 0.0604, "step": 38680 }, { "epoch": 1.3080225835897088, "grad_norm": 0.228939026594162, "learning_rate": 7.205039790588939e-05, "loss": 0.0508, "step": 38690 }, { "epoch": 1.308360661279962, "grad_norm": 0.22895380854606628, "learning_rate": 7.203555675605697e-05, "loss": 0.0559, "step": 38700 }, { "epoch": 1.3086987389702154, "grad_norm": 0.1838909238576889, "learning_rate": 7.202071319644917e-05, "loss": 0.0549, "step": 38710 }, { "epoch": 1.3090368166604684, "grad_norm": 0.19072091579437256, "learning_rate": 7.200586722868932e-05, "loss": 0.0564, "step": 38720 }, { "epoch": 1.3093748943507217, "grad_norm": 0.3443347215652466, "learning_rate": 7.199101885440088e-05, "loss": 0.0604, "step": 38730 }, { "epoch": 1.309712972040975, "grad_norm": 0.2371784746646881, "learning_rate": 7.197616807520771e-05, "loss": 0.0524, "step": 38740 }, { "epoch": 1.310051049731228, "grad_norm": 0.23068726062774658, "learning_rate": 7.196131489273381e-05, "loss": 0.0557, "step": 38750 }, { "epoch": 1.3103891274214814, "grad_norm": 0.2664562165737152, "learning_rate": 7.194645930860355e-05, "loss": 0.0546, "step": 38760 }, { "epoch": 1.3107272051117347, "grad_norm": 0.293930321931839, "learning_rate": 7.19316013244415e-05, "loss": 0.0591, "step": 38770 }, { "epoch": 1.311065282801988, "grad_norm": 0.24215346574783325, "learning_rate": 7.191674094187248e-05, "loss": 0.0559, "step": 38780 }, { "epoch": 1.311403360492241, "grad_norm": 0.3426001965999603, "learning_rate": 7.190187816252165e-05, "loss": 0.0535, "step": 38790 }, { "epoch": 1.3117414381824943, "grad_norm": 0.26191243529319763, "learning_rate": 7.188701298801435e-05, "loss": 0.0521, "step": 38800 }, { "epoch": 1.3120795158727476, "grad_norm": 0.2950555980205536, "learning_rate": 7.18721454199762e-05, "loss": 0.0558, "step": 38810 }, { "epoch": 1.3124175935630007, "grad_norm": 0.16793528199195862, "learning_rate": 7.185727546003312e-05, "loss": 0.0536, "step": 38820 }, { "epoch": 1.312755671253254, "grad_norm": 0.17776371538639069, "learning_rate": 7.184240310981126e-05, "loss": 0.0509, "step": 38830 }, { "epoch": 1.3130937489435073, "grad_norm": 0.2097329944372177, "learning_rate": 7.182752837093704e-05, "loss": 0.0556, "step": 38840 }, { "epoch": 1.3134318266337606, "grad_norm": 0.16892825067043304, "learning_rate": 7.181265124503711e-05, "loss": 0.053, "step": 38850 }, { "epoch": 1.3137699043240136, "grad_norm": 0.2328767031431198, "learning_rate": 7.179777173373847e-05, "loss": 0.0523, "step": 38860 }, { "epoch": 1.314107982014267, "grad_norm": 0.42147552967071533, "learning_rate": 7.178288983866826e-05, "loss": 0.059, "step": 38870 }, { "epoch": 1.3144460597045202, "grad_norm": 0.3175809383392334, "learning_rate": 7.176800556145397e-05, "loss": 0.0593, "step": 38880 }, { "epoch": 1.3147841373947733, "grad_norm": 0.23064906895160675, "learning_rate": 7.175311890372334e-05, "loss": 0.0627, "step": 38890 }, { "epoch": 1.3151222150850266, "grad_norm": 0.24398300051689148, "learning_rate": 7.17382298671043e-05, "loss": 0.0573, "step": 38900 }, { "epoch": 1.3154602927752799, "grad_norm": 0.2417798489332199, "learning_rate": 7.172333845322515e-05, "loss": 0.0478, "step": 38910 }, { "epoch": 1.315798370465533, "grad_norm": 0.22831203043460846, "learning_rate": 7.170844466371436e-05, "loss": 0.0558, "step": 38920 }, { "epoch": 1.3161364481557862, "grad_norm": 0.40262213349342346, "learning_rate": 7.16935485002007e-05, "loss": 0.0557, "step": 38930 }, { "epoch": 1.3164745258460395, "grad_norm": 0.19972217082977295, "learning_rate": 7.167864996431319e-05, "loss": 0.0498, "step": 38940 }, { "epoch": 1.3168126035362926, "grad_norm": 0.3793812692165375, "learning_rate": 7.166374905768111e-05, "loss": 0.0531, "step": 38950 }, { "epoch": 1.3171506812265459, "grad_norm": 0.34607112407684326, "learning_rate": 7.164884578193401e-05, "loss": 0.0576, "step": 38960 }, { "epoch": 1.3174887589167992, "grad_norm": 0.2042425572872162, "learning_rate": 7.163394013870168e-05, "loss": 0.0511, "step": 38970 }, { "epoch": 1.3178268366070522, "grad_norm": 0.1913008838891983, "learning_rate": 7.161903212961419e-05, "loss": 0.0541, "step": 38980 }, { "epoch": 1.3181649142973055, "grad_norm": 0.2575196921825409, "learning_rate": 7.160412175630183e-05, "loss": 0.0539, "step": 38990 }, { "epoch": 1.3185029919875588, "grad_norm": 0.266491174697876, "learning_rate": 7.158920902039521e-05, "loss": 0.0545, "step": 39000 }, { "epoch": 1.318841069677812, "grad_norm": 0.2500446140766144, "learning_rate": 7.157429392352514e-05, "loss": 0.0563, "step": 39010 }, { "epoch": 1.3191791473680652, "grad_norm": 0.33366572856903076, "learning_rate": 7.155937646732275e-05, "loss": 0.0558, "step": 39020 }, { "epoch": 1.3195172250583185, "grad_norm": 0.24492448568344116, "learning_rate": 7.154445665341933e-05, "loss": 0.0564, "step": 39030 }, { "epoch": 1.3198553027485715, "grad_norm": 0.2674331068992615, "learning_rate": 7.152953448344654e-05, "loss": 0.0523, "step": 39040 }, { "epoch": 1.3201933804388248, "grad_norm": 0.29507577419281006, "learning_rate": 7.151460995903624e-05, "loss": 0.0518, "step": 39050 }, { "epoch": 1.3205314581290781, "grad_norm": 0.23625685274600983, "learning_rate": 7.149968308182052e-05, "loss": 0.0502, "step": 39060 }, { "epoch": 1.3208695358193312, "grad_norm": 0.2187422215938568, "learning_rate": 7.14847538534318e-05, "loss": 0.0526, "step": 39070 }, { "epoch": 1.3212076135095845, "grad_norm": 0.2635059654712677, "learning_rate": 7.14698222755027e-05, "loss": 0.0522, "step": 39080 }, { "epoch": 1.3215456911998378, "grad_norm": 0.24703194200992584, "learning_rate": 7.14548883496661e-05, "loss": 0.0527, "step": 39090 }, { "epoch": 1.3218837688900908, "grad_norm": 0.8572195768356323, "learning_rate": 7.143995207755517e-05, "loss": 0.0527, "step": 39100 }, { "epoch": 1.3222218465803441, "grad_norm": 0.22454270720481873, "learning_rate": 7.142501346080333e-05, "loss": 0.0566, "step": 39110 }, { "epoch": 1.3225599242705974, "grad_norm": 0.22731874883174896, "learning_rate": 7.141007250104421e-05, "loss": 0.0522, "step": 39120 }, { "epoch": 1.3228980019608505, "grad_norm": 0.22316904366016388, "learning_rate": 7.139512919991176e-05, "loss": 0.0459, "step": 39130 }, { "epoch": 1.3232360796511038, "grad_norm": 0.26745256781578064, "learning_rate": 7.138018355904014e-05, "loss": 0.0534, "step": 39140 }, { "epoch": 1.323574157341357, "grad_norm": 0.29654422402381897, "learning_rate": 7.13652355800638e-05, "loss": 0.0574, "step": 39150 }, { "epoch": 1.3239122350316102, "grad_norm": 0.1611323207616806, "learning_rate": 7.135028526461743e-05, "loss": 0.0504, "step": 39160 }, { "epoch": 1.3242503127218634, "grad_norm": 0.39325445890426636, "learning_rate": 7.133533261433594e-05, "loss": 0.0504, "step": 39170 }, { "epoch": 1.3245883904121167, "grad_norm": 0.2737731337547302, "learning_rate": 7.132037763085457e-05, "loss": 0.0589, "step": 39180 }, { "epoch": 1.3249264681023698, "grad_norm": 0.20617379248142242, "learning_rate": 7.130542031580875e-05, "loss": 0.0569, "step": 39190 }, { "epoch": 1.325264545792623, "grad_norm": 0.1823807656764984, "learning_rate": 7.12904606708342e-05, "loss": 0.0491, "step": 39200 }, { "epoch": 1.3256026234828764, "grad_norm": 0.21429549157619476, "learning_rate": 7.127549869756687e-05, "loss": 0.0508, "step": 39210 }, { "epoch": 1.3259407011731295, "grad_norm": 0.3389512002468109, "learning_rate": 7.126053439764299e-05, "loss": 0.051, "step": 39220 }, { "epoch": 1.3262787788633827, "grad_norm": 0.21652723848819733, "learning_rate": 7.124556777269904e-05, "loss": 0.056, "step": 39230 }, { "epoch": 1.326616856553636, "grad_norm": 0.17475920915603638, "learning_rate": 7.123059882437174e-05, "loss": 0.0545, "step": 39240 }, { "epoch": 1.3269549342438893, "grad_norm": 0.28795677423477173, "learning_rate": 7.121562755429807e-05, "loss": 0.0648, "step": 39250 }, { "epoch": 1.3272930119341424, "grad_norm": 0.2051832526922226, "learning_rate": 7.120065396411528e-05, "loss": 0.0578, "step": 39260 }, { "epoch": 1.3276310896243957, "grad_norm": 0.23523882031440735, "learning_rate": 7.118567805546084e-05, "loss": 0.0582, "step": 39270 }, { "epoch": 1.327969167314649, "grad_norm": 0.2691175043582916, "learning_rate": 7.117069982997248e-05, "loss": 0.0557, "step": 39280 }, { "epoch": 1.328307245004902, "grad_norm": 0.709047257900238, "learning_rate": 7.115571928928825e-05, "loss": 0.0523, "step": 39290 }, { "epoch": 1.3286453226951553, "grad_norm": 0.22721867263317108, "learning_rate": 7.114073643504635e-05, "loss": 0.0547, "step": 39300 }, { "epoch": 1.3289834003854086, "grad_norm": 0.1644282191991806, "learning_rate": 7.11257512688853e-05, "loss": 0.0564, "step": 39310 }, { "epoch": 1.329321478075662, "grad_norm": 0.3480994999408722, "learning_rate": 7.111076379244384e-05, "loss": 0.0524, "step": 39320 }, { "epoch": 1.329659555765915, "grad_norm": 0.3459396958351135, "learning_rate": 7.109577400736101e-05, "loss": 0.0584, "step": 39330 }, { "epoch": 1.3299976334561683, "grad_norm": 0.21615809202194214, "learning_rate": 7.108078191527605e-05, "loss": 0.0517, "step": 39340 }, { "epoch": 1.3303357111464216, "grad_norm": 0.3021867871284485, "learning_rate": 7.106578751782847e-05, "loss": 0.0526, "step": 39350 }, { "epoch": 1.3306737888366746, "grad_norm": 0.2402232438325882, "learning_rate": 7.105079081665803e-05, "loss": 0.0544, "step": 39360 }, { "epoch": 1.331011866526928, "grad_norm": 0.27394911646842957, "learning_rate": 7.103579181340476e-05, "loss": 0.0511, "step": 39370 }, { "epoch": 1.3313499442171812, "grad_norm": 0.19686195254325867, "learning_rate": 7.102079050970893e-05, "loss": 0.055, "step": 39380 }, { "epoch": 1.3316880219074343, "grad_norm": 0.360771119594574, "learning_rate": 7.100578690721105e-05, "loss": 0.0476, "step": 39390 }, { "epoch": 1.3320260995976876, "grad_norm": 0.16643688082695007, "learning_rate": 7.09907810075519e-05, "loss": 0.0537, "step": 39400 }, { "epoch": 1.3323641772879409, "grad_norm": 0.18105222284793854, "learning_rate": 7.097577281237249e-05, "loss": 0.049, "step": 39410 }, { "epoch": 1.332702254978194, "grad_norm": 0.17101389169692993, "learning_rate": 7.09607623233141e-05, "loss": 0.0589, "step": 39420 }, { "epoch": 1.3330403326684472, "grad_norm": 0.2813888192176819, "learning_rate": 7.094574954201827e-05, "loss": 0.0517, "step": 39430 }, { "epoch": 1.3333784103587005, "grad_norm": 0.1575462967157364, "learning_rate": 7.093073447012675e-05, "loss": 0.0487, "step": 39440 }, { "epoch": 1.3337164880489536, "grad_norm": 0.28526243567466736, "learning_rate": 7.09157171092816e-05, "loss": 0.0535, "step": 39450 }, { "epoch": 1.3340545657392069, "grad_norm": 0.2885506749153137, "learning_rate": 7.090069746112504e-05, "loss": 0.0518, "step": 39460 }, { "epoch": 1.3343926434294602, "grad_norm": 0.16486208140850067, "learning_rate": 7.088567552729965e-05, "loss": 0.0539, "step": 39470 }, { "epoch": 1.3347307211197132, "grad_norm": 0.3104313910007477, "learning_rate": 7.087065130944818e-05, "loss": 0.0568, "step": 39480 }, { "epoch": 1.3350687988099665, "grad_norm": 0.19453762471675873, "learning_rate": 7.085562480921366e-05, "loss": 0.0505, "step": 39490 }, { "epoch": 1.3354068765002198, "grad_norm": 0.2192394882440567, "learning_rate": 7.084059602823937e-05, "loss": 0.0532, "step": 39500 }, { "epoch": 1.335744954190473, "grad_norm": 0.2380325347185135, "learning_rate": 7.082556496816882e-05, "loss": 0.055, "step": 39510 }, { "epoch": 1.3360830318807262, "grad_norm": 0.20002645254135132, "learning_rate": 7.081053163064582e-05, "loss": 0.0534, "step": 39520 }, { "epoch": 1.3364211095709795, "grad_norm": 0.14830689132213593, "learning_rate": 7.079549601731434e-05, "loss": 0.0512, "step": 39530 }, { "epoch": 1.3367591872612326, "grad_norm": 0.14930135011672974, "learning_rate": 7.07804581298187e-05, "loss": 0.0543, "step": 39540 }, { "epoch": 1.3370972649514858, "grad_norm": 0.28859829902648926, "learning_rate": 7.07654179698034e-05, "loss": 0.0547, "step": 39550 }, { "epoch": 1.3374353426417391, "grad_norm": 0.26160505414009094, "learning_rate": 7.075037553891321e-05, "loss": 0.0519, "step": 39560 }, { "epoch": 1.3377734203319922, "grad_norm": 0.3711322247982025, "learning_rate": 7.073533083879315e-05, "loss": 0.0526, "step": 39570 }, { "epoch": 1.3381114980222455, "grad_norm": 0.34547194838523865, "learning_rate": 7.072028387108849e-05, "loss": 0.0547, "step": 39580 }, { "epoch": 1.3384495757124988, "grad_norm": 0.2386353313922882, "learning_rate": 7.070523463744472e-05, "loss": 0.0515, "step": 39590 }, { "epoch": 1.3387876534027519, "grad_norm": 0.2055506557226181, "learning_rate": 7.069018313950763e-05, "loss": 0.0521, "step": 39600 }, { "epoch": 1.3391257310930051, "grad_norm": 0.32485541701316833, "learning_rate": 7.067512937892322e-05, "loss": 0.056, "step": 39610 }, { "epoch": 1.3394638087832584, "grad_norm": 0.3235771059989929, "learning_rate": 7.066007335733775e-05, "loss": 0.0488, "step": 39620 }, { "epoch": 1.3398018864735115, "grad_norm": 0.26307377219200134, "learning_rate": 7.064501507639772e-05, "loss": 0.0569, "step": 39630 }, { "epoch": 1.3401399641637648, "grad_norm": 0.2014208287000656, "learning_rate": 7.062995453774987e-05, "loss": 0.0504, "step": 39640 }, { "epoch": 1.340478041854018, "grad_norm": 0.2085805982351303, "learning_rate": 7.061489174304121e-05, "loss": 0.0561, "step": 39650 }, { "epoch": 1.3408161195442712, "grad_norm": 0.2237151563167572, "learning_rate": 7.0599826693919e-05, "loss": 0.0516, "step": 39660 }, { "epoch": 1.3411541972345244, "grad_norm": 0.3073526620864868, "learning_rate": 7.05847593920307e-05, "loss": 0.0527, "step": 39670 }, { "epoch": 1.3414922749247777, "grad_norm": 0.2818675935268402, "learning_rate": 7.056968983902406e-05, "loss": 0.056, "step": 39680 }, { "epoch": 1.3418303526150308, "grad_norm": 0.29576992988586426, "learning_rate": 7.055461803654706e-05, "loss": 0.0625, "step": 39690 }, { "epoch": 1.342168430305284, "grad_norm": 0.2934156656265259, "learning_rate": 7.053954398624794e-05, "loss": 0.0557, "step": 39700 }, { "epoch": 1.3425065079955374, "grad_norm": 0.3021095097064972, "learning_rate": 7.052446768977518e-05, "loss": 0.0532, "step": 39710 }, { "epoch": 1.3428445856857905, "grad_norm": 0.20120182633399963, "learning_rate": 7.050938914877748e-05, "loss": 0.0549, "step": 39720 }, { "epoch": 1.3431826633760438, "grad_norm": 0.21550016105175018, "learning_rate": 7.049430836490382e-05, "loss": 0.0532, "step": 39730 }, { "epoch": 1.343520741066297, "grad_norm": 0.21427729725837708, "learning_rate": 7.047922533980342e-05, "loss": 0.055, "step": 39740 }, { "epoch": 1.3438588187565503, "grad_norm": 0.3157675266265869, "learning_rate": 7.046414007512571e-05, "loss": 0.0566, "step": 39750 }, { "epoch": 1.3441968964468034, "grad_norm": 0.19141261279582977, "learning_rate": 7.044905257252042e-05, "loss": 0.0543, "step": 39760 }, { "epoch": 1.3445349741370567, "grad_norm": 0.3340175747871399, "learning_rate": 7.04339628336375e-05, "loss": 0.0569, "step": 39770 }, { "epoch": 1.34487305182731, "grad_norm": 0.20393472909927368, "learning_rate": 7.041887086012711e-05, "loss": 0.0511, "step": 39780 }, { "epoch": 1.345211129517563, "grad_norm": 1.1673678159713745, "learning_rate": 7.040377665363969e-05, "loss": 0.0596, "step": 39790 }, { "epoch": 1.3455492072078163, "grad_norm": 0.22962026298046112, "learning_rate": 7.038868021582594e-05, "loss": 0.0559, "step": 39800 }, { "epoch": 1.3458872848980696, "grad_norm": 0.23897196352481842, "learning_rate": 7.037358154833679e-05, "loss": 0.0544, "step": 39810 }, { "epoch": 1.346225362588323, "grad_norm": 0.23552405834197998, "learning_rate": 7.035848065282339e-05, "loss": 0.0603, "step": 39820 }, { "epoch": 1.346563440278576, "grad_norm": 0.28066980838775635, "learning_rate": 7.034337753093714e-05, "loss": 0.0497, "step": 39830 }, { "epoch": 1.3469015179688293, "grad_norm": 0.277307391166687, "learning_rate": 7.032827218432972e-05, "loss": 0.0511, "step": 39840 }, { "epoch": 1.3472395956590826, "grad_norm": 0.3482230603694916, "learning_rate": 7.031316461465302e-05, "loss": 0.0532, "step": 39850 }, { "epoch": 1.3475776733493356, "grad_norm": 0.31175148487091064, "learning_rate": 7.029805482355915e-05, "loss": 0.0549, "step": 39860 }, { "epoch": 1.347915751039589, "grad_norm": 0.29867395758628845, "learning_rate": 7.028294281270055e-05, "loss": 0.0472, "step": 39870 }, { "epoch": 1.3482538287298422, "grad_norm": 0.3496028184890747, "learning_rate": 7.026782858372982e-05, "loss": 0.0477, "step": 39880 }, { "epoch": 1.3485919064200953, "grad_norm": 0.3593824803829193, "learning_rate": 7.025271213829982e-05, "loss": 0.0601, "step": 39890 }, { "epoch": 1.3489299841103486, "grad_norm": 0.28661003708839417, "learning_rate": 7.023759347806366e-05, "loss": 0.0583, "step": 39900 }, { "epoch": 1.3492680618006019, "grad_norm": 0.24414601922035217, "learning_rate": 7.022247260467472e-05, "loss": 0.0531, "step": 39910 }, { "epoch": 1.349606139490855, "grad_norm": 0.3451871871948242, "learning_rate": 7.020734951978658e-05, "loss": 0.0477, "step": 39920 }, { "epoch": 1.3499442171811082, "grad_norm": 0.20800870656967163, "learning_rate": 7.019222422505307e-05, "loss": 0.05, "step": 39930 }, { "epoch": 1.3502822948713615, "grad_norm": 0.20570024847984314, "learning_rate": 7.017709672212827e-05, "loss": 0.0517, "step": 39940 }, { "epoch": 1.3506203725616146, "grad_norm": 0.3960403800010681, "learning_rate": 7.016196701266652e-05, "loss": 0.049, "step": 39950 }, { "epoch": 1.350958450251868, "grad_norm": 0.30484437942504883, "learning_rate": 7.014683509832238e-05, "loss": 0.0548, "step": 39960 }, { "epoch": 1.3512965279421212, "grad_norm": 0.3906497359275818, "learning_rate": 7.013170098075063e-05, "loss": 0.0531, "step": 39970 }, { "epoch": 1.3516346056323743, "grad_norm": 0.2901833653450012, "learning_rate": 7.011656466160632e-05, "loss": 0.0513, "step": 39980 }, { "epoch": 1.3519726833226275, "grad_norm": 0.3036821484565735, "learning_rate": 7.010142614254475e-05, "loss": 0.0546, "step": 39990 }, { "epoch": 1.3523107610128808, "grad_norm": 0.27281004190444946, "learning_rate": 7.008628542522147e-05, "loss": 0.0501, "step": 40000 }, { "epoch": 1.352648838703134, "grad_norm": 0.26109251379966736, "learning_rate": 7.007114251129218e-05, "loss": 0.0583, "step": 40010 }, { "epoch": 1.3529869163933872, "grad_norm": 0.3299400806427002, "learning_rate": 7.005599740241293e-05, "loss": 0.0527, "step": 40020 }, { "epoch": 1.3533249940836405, "grad_norm": 0.1883242428302765, "learning_rate": 7.004085010023996e-05, "loss": 0.0595, "step": 40030 }, { "epoch": 1.3536630717738936, "grad_norm": 0.19648891687393188, "learning_rate": 7.002570060642976e-05, "loss": 0.0543, "step": 40040 }, { "epoch": 1.3540011494641468, "grad_norm": 0.35042673349380493, "learning_rate": 7.001054892263903e-05, "loss": 0.0619, "step": 40050 }, { "epoch": 1.3543392271544001, "grad_norm": 0.21801289916038513, "learning_rate": 6.999539505052477e-05, "loss": 0.0549, "step": 40060 }, { "epoch": 1.3546773048446532, "grad_norm": 0.2026006132364273, "learning_rate": 6.998023899174418e-05, "loss": 0.052, "step": 40070 }, { "epoch": 1.3550153825349065, "grad_norm": 0.28480178117752075, "learning_rate": 6.996508074795467e-05, "loss": 0.055, "step": 40080 }, { "epoch": 1.3553534602251598, "grad_norm": 0.28156623244285583, "learning_rate": 6.994992032081396e-05, "loss": 0.0518, "step": 40090 }, { "epoch": 1.3556915379154129, "grad_norm": 0.3103553056716919, "learning_rate": 6.993475771197995e-05, "loss": 0.0581, "step": 40100 }, { "epoch": 1.3560296156056661, "grad_norm": 0.1877496838569641, "learning_rate": 6.991959292311082e-05, "loss": 0.0538, "step": 40110 }, { "epoch": 1.3563676932959194, "grad_norm": 0.35408344864845276, "learning_rate": 6.990442595586495e-05, "loss": 0.05, "step": 40120 }, { "epoch": 1.3567057709861725, "grad_norm": 0.326157808303833, "learning_rate": 6.988925681190098e-05, "loss": 0.0565, "step": 40130 }, { "epoch": 1.3570438486764258, "grad_norm": 0.22357282042503357, "learning_rate": 6.987408549287778e-05, "loss": 0.048, "step": 40140 }, { "epoch": 1.357381926366679, "grad_norm": 0.32725754380226135, "learning_rate": 6.985891200045449e-05, "loss": 0.0514, "step": 40150 }, { "epoch": 1.3577200040569322, "grad_norm": 0.30025115609169006, "learning_rate": 6.984373633629045e-05, "loss": 0.0527, "step": 40160 }, { "epoch": 1.3580580817471855, "grad_norm": 0.5561268329620361, "learning_rate": 6.98285585020452e-05, "loss": 0.0526, "step": 40170 }, { "epoch": 1.3583961594374387, "grad_norm": 0.21020609140396118, "learning_rate": 6.981337849937864e-05, "loss": 0.0503, "step": 40180 }, { "epoch": 1.3587342371276918, "grad_norm": 0.2360849380493164, "learning_rate": 6.979819632995078e-05, "loss": 0.0549, "step": 40190 }, { "epoch": 1.359072314817945, "grad_norm": 0.19340795278549194, "learning_rate": 6.978301199542193e-05, "loss": 0.0493, "step": 40200 }, { "epoch": 1.3594103925081984, "grad_norm": 0.19565021991729736, "learning_rate": 6.976782549745263e-05, "loss": 0.0542, "step": 40210 }, { "epoch": 1.3597484701984517, "grad_norm": 0.18475241959095, "learning_rate": 6.975263683770367e-05, "loss": 0.0555, "step": 40220 }, { "epoch": 1.3600865478887048, "grad_norm": 0.3629065155982971, "learning_rate": 6.973744601783604e-05, "loss": 0.0503, "step": 40230 }, { "epoch": 1.360424625578958, "grad_norm": 0.3204191327095032, "learning_rate": 6.972225303951097e-05, "loss": 0.0529, "step": 40240 }, { "epoch": 1.3607627032692113, "grad_norm": 0.14900057017803192, "learning_rate": 6.970705790438998e-05, "loss": 0.0505, "step": 40250 }, { "epoch": 1.3611007809594644, "grad_norm": 0.2951807677745819, "learning_rate": 6.969186061413477e-05, "loss": 0.056, "step": 40260 }, { "epoch": 1.3614388586497177, "grad_norm": 0.3002043664455414, "learning_rate": 6.967666117040727e-05, "loss": 0.0497, "step": 40270 }, { "epoch": 1.361776936339971, "grad_norm": 0.3716284930706024, "learning_rate": 6.966145957486972e-05, "loss": 0.0522, "step": 40280 }, { "epoch": 1.3621150140302243, "grad_norm": 0.19679507613182068, "learning_rate": 6.964625582918449e-05, "loss": 0.0543, "step": 40290 }, { "epoch": 1.3624530917204773, "grad_norm": 0.39674070477485657, "learning_rate": 6.963104993501425e-05, "loss": 0.0546, "step": 40300 }, { "epoch": 1.3627911694107306, "grad_norm": 0.3066030442714691, "learning_rate": 6.961584189402192e-05, "loss": 0.0552, "step": 40310 }, { "epoch": 1.363129247100984, "grad_norm": 0.4088115990161896, "learning_rate": 6.96006317078706e-05, "loss": 0.0573, "step": 40320 }, { "epoch": 1.363467324791237, "grad_norm": 0.2813878655433655, "learning_rate": 6.958541937822367e-05, "loss": 0.0542, "step": 40330 }, { "epoch": 1.3638054024814903, "grad_norm": 0.28345951437950134, "learning_rate": 6.957020490674473e-05, "loss": 0.048, "step": 40340 }, { "epoch": 1.3641434801717436, "grad_norm": 0.218834787607193, "learning_rate": 6.95549882950976e-05, "loss": 0.0546, "step": 40350 }, { "epoch": 1.3644815578619967, "grad_norm": 0.3164375126361847, "learning_rate": 6.953976954494635e-05, "loss": 0.0497, "step": 40360 }, { "epoch": 1.36481963555225, "grad_norm": 0.34236717224121094, "learning_rate": 6.952454865795528e-05, "loss": 0.0521, "step": 40370 }, { "epoch": 1.3651577132425032, "grad_norm": 0.18422366678714752, "learning_rate": 6.950932563578892e-05, "loss": 0.0513, "step": 40380 }, { "epoch": 1.3654957909327563, "grad_norm": 0.34058496356010437, "learning_rate": 6.949410048011206e-05, "loss": 0.0514, "step": 40390 }, { "epoch": 1.3658338686230096, "grad_norm": 0.17034561932086945, "learning_rate": 6.947887319258966e-05, "loss": 0.0518, "step": 40400 }, { "epoch": 1.3661719463132629, "grad_norm": 0.2880447804927826, "learning_rate": 6.946364377488696e-05, "loss": 0.0548, "step": 40410 }, { "epoch": 1.366510024003516, "grad_norm": 0.4138709008693695, "learning_rate": 6.944841222866947e-05, "loss": 0.0534, "step": 40420 }, { "epoch": 1.3668481016937692, "grad_norm": 0.15263207256793976, "learning_rate": 6.943317855560284e-05, "loss": 0.0493, "step": 40430 }, { "epoch": 1.3671861793840225, "grad_norm": 0.20107463002204895, "learning_rate": 6.941794275735302e-05, "loss": 0.054, "step": 40440 }, { "epoch": 1.3675242570742756, "grad_norm": 0.21814671158790588, "learning_rate": 6.94027048355862e-05, "loss": 0.0477, "step": 40450 }, { "epoch": 1.367862334764529, "grad_norm": 0.21055711805820465, "learning_rate": 6.938746479196871e-05, "loss": 0.0524, "step": 40460 }, { "epoch": 1.3682004124547822, "grad_norm": 0.23551230132579803, "learning_rate": 6.937222262816724e-05, "loss": 0.0503, "step": 40470 }, { "epoch": 1.3685384901450353, "grad_norm": 0.31223157048225403, "learning_rate": 6.935697834584865e-05, "loss": 0.0564, "step": 40480 }, { "epoch": 1.3688765678352885, "grad_norm": 0.36151349544525146, "learning_rate": 6.934173194667998e-05, "loss": 0.0539, "step": 40490 }, { "epoch": 1.3692146455255418, "grad_norm": 0.29971420764923096, "learning_rate": 6.93264834323286e-05, "loss": 0.0609, "step": 40500 }, { "epoch": 1.369552723215795, "grad_norm": 0.2534381151199341, "learning_rate": 6.931123280446206e-05, "loss": 0.0559, "step": 40510 }, { "epoch": 1.3698908009060482, "grad_norm": 0.21639899909496307, "learning_rate": 6.929598006474811e-05, "loss": 0.0557, "step": 40520 }, { "epoch": 1.3702288785963015, "grad_norm": 0.20844189822673798, "learning_rate": 6.92807252148548e-05, "loss": 0.0564, "step": 40530 }, { "epoch": 1.3705669562865546, "grad_norm": 0.23917265236377716, "learning_rate": 6.926546825645037e-05, "loss": 0.0585, "step": 40540 }, { "epoch": 1.3709050339768079, "grad_norm": 0.1915089637041092, "learning_rate": 6.92502091912033e-05, "loss": 0.0547, "step": 40550 }, { "epoch": 1.3712431116670611, "grad_norm": 0.2145102471113205, "learning_rate": 6.923494802078229e-05, "loss": 0.0507, "step": 40560 }, { "epoch": 1.3715811893573142, "grad_norm": 0.2692126929759979, "learning_rate": 6.921968474685628e-05, "loss": 0.0507, "step": 40570 }, { "epoch": 1.3719192670475675, "grad_norm": 0.21070335805416107, "learning_rate": 6.920441937109445e-05, "loss": 0.0522, "step": 40580 }, { "epoch": 1.3722573447378208, "grad_norm": 0.34044188261032104, "learning_rate": 6.91891518951662e-05, "loss": 0.0516, "step": 40590 }, { "epoch": 1.3725954224280739, "grad_norm": 0.23375841975212097, "learning_rate": 6.917388232074114e-05, "loss": 0.0533, "step": 40600 }, { "epoch": 1.3729335001183272, "grad_norm": 0.19910892844200134, "learning_rate": 6.915861064948914e-05, "loss": 0.051, "step": 40610 }, { "epoch": 1.3732715778085804, "grad_norm": 0.33700835704803467, "learning_rate": 6.914333688308025e-05, "loss": 0.0512, "step": 40620 }, { "epoch": 1.3736096554988335, "grad_norm": 0.20587855577468872, "learning_rate": 6.912806102318487e-05, "loss": 0.0487, "step": 40630 }, { "epoch": 1.3739477331890868, "grad_norm": 0.250358521938324, "learning_rate": 6.911278307147347e-05, "loss": 0.0532, "step": 40640 }, { "epoch": 1.37428581087934, "grad_norm": 0.45613086223602295, "learning_rate": 6.909750302961684e-05, "loss": 0.0543, "step": 40650 }, { "epoch": 1.3746238885695932, "grad_norm": 0.24616529047489166, "learning_rate": 6.9082220899286e-05, "loss": 0.0476, "step": 40660 }, { "epoch": 1.3749619662598465, "grad_norm": 0.24848847091197968, "learning_rate": 6.906693668215219e-05, "loss": 0.0498, "step": 40670 }, { "epoch": 1.3753000439500997, "grad_norm": 0.30765172839164734, "learning_rate": 6.905165037988683e-05, "loss": 0.0515, "step": 40680 }, { "epoch": 1.3756381216403528, "grad_norm": 0.24206958711147308, "learning_rate": 6.903636199416164e-05, "loss": 0.0534, "step": 40690 }, { "epoch": 1.375976199330606, "grad_norm": 0.25459933280944824, "learning_rate": 6.902107152664851e-05, "loss": 0.054, "step": 40700 }, { "epoch": 1.3763142770208594, "grad_norm": 0.3804142475128174, "learning_rate": 6.90057789790196e-05, "loss": 0.0576, "step": 40710 }, { "epoch": 1.3766523547111127, "grad_norm": 0.22458888590335846, "learning_rate": 6.899048435294728e-05, "loss": 0.0583, "step": 40720 }, { "epoch": 1.3769904324013658, "grad_norm": 0.23702549934387207, "learning_rate": 6.897518765010415e-05, "loss": 0.051, "step": 40730 }, { "epoch": 1.377328510091619, "grad_norm": 0.31418633460998535, "learning_rate": 6.895988887216302e-05, "loss": 0.0485, "step": 40740 }, { "epoch": 1.3776665877818723, "grad_norm": 0.2008305788040161, "learning_rate": 6.894458802079694e-05, "loss": 0.0458, "step": 40750 }, { "epoch": 1.3780046654721254, "grad_norm": 0.35688039660453796, "learning_rate": 6.892928509767922e-05, "loss": 0.0571, "step": 40760 }, { "epoch": 1.3783427431623787, "grad_norm": 0.4294707179069519, "learning_rate": 6.891398010448333e-05, "loss": 0.051, "step": 40770 }, { "epoch": 1.378680820852632, "grad_norm": 0.27349621057510376, "learning_rate": 6.8898673042883e-05, "loss": 0.0586, "step": 40780 }, { "epoch": 1.3790188985428853, "grad_norm": 0.559869110584259, "learning_rate": 6.888336391455222e-05, "loss": 0.0518, "step": 40790 }, { "epoch": 1.3793569762331384, "grad_norm": 0.5982683300971985, "learning_rate": 6.886805272116513e-05, "loss": 0.0598, "step": 40800 }, { "epoch": 1.3796950539233916, "grad_norm": 0.2953031659126282, "learning_rate": 6.885273946439617e-05, "loss": 0.0529, "step": 40810 }, { "epoch": 1.380033131613645, "grad_norm": 0.27817875146865845, "learning_rate": 6.883742414591998e-05, "loss": 0.051, "step": 40820 }, { "epoch": 1.380371209303898, "grad_norm": 0.24164743721485138, "learning_rate": 6.88221067674114e-05, "loss": 0.0635, "step": 40830 }, { "epoch": 1.3807092869941513, "grad_norm": 0.24972467124462128, "learning_rate": 6.88067873305455e-05, "loss": 0.054, "step": 40840 }, { "epoch": 1.3810473646844046, "grad_norm": 0.41183871030807495, "learning_rate": 6.879146583699765e-05, "loss": 0.0589, "step": 40850 }, { "epoch": 1.3813854423746577, "grad_norm": 0.23522500693798065, "learning_rate": 6.877614228844334e-05, "loss": 0.0577, "step": 40860 }, { "epoch": 1.381723520064911, "grad_norm": 0.2555164396762848, "learning_rate": 6.876081668655832e-05, "loss": 0.0569, "step": 40870 }, { "epoch": 1.3820615977551642, "grad_norm": 0.2506631016731262, "learning_rate": 6.874548903301861e-05, "loss": 0.055, "step": 40880 }, { "epoch": 1.3823996754454173, "grad_norm": 0.1985449343919754, "learning_rate": 6.873015932950039e-05, "loss": 0.0568, "step": 40890 }, { "epoch": 1.3827377531356706, "grad_norm": 0.20007221400737762, "learning_rate": 6.871482757768012e-05, "loss": 0.0519, "step": 40900 }, { "epoch": 1.383075830825924, "grad_norm": 0.17841219902038574, "learning_rate": 6.869949377923442e-05, "loss": 0.0477, "step": 40910 }, { "epoch": 1.383413908516177, "grad_norm": 0.22346645593643188, "learning_rate": 6.868415793584022e-05, "loss": 0.0489, "step": 40920 }, { "epoch": 1.3837519862064303, "grad_norm": 0.20268025994300842, "learning_rate": 6.86688200491746e-05, "loss": 0.0542, "step": 40930 }, { "epoch": 1.3840900638966835, "grad_norm": 0.22991451621055603, "learning_rate": 6.865348012091488e-05, "loss": 0.0531, "step": 40940 }, { "epoch": 1.3844281415869366, "grad_norm": 0.23200401663780212, "learning_rate": 6.86381381527386e-05, "loss": 0.0514, "step": 40950 }, { "epoch": 1.38476621927719, "grad_norm": 0.19767647981643677, "learning_rate": 6.862279414632359e-05, "loss": 0.0516, "step": 40960 }, { "epoch": 1.3851042969674432, "grad_norm": 0.27488386631011963, "learning_rate": 6.860744810334778e-05, "loss": 0.0538, "step": 40970 }, { "epoch": 1.3854423746576963, "grad_norm": 0.2497398555278778, "learning_rate": 6.859210002548943e-05, "loss": 0.0584, "step": 40980 }, { "epoch": 1.3857804523479496, "grad_norm": 0.24343496561050415, "learning_rate": 6.8576749914427e-05, "loss": 0.0537, "step": 40990 }, { "epoch": 1.3861185300382028, "grad_norm": 0.6792377829551697, "learning_rate": 6.85613977718391e-05, "loss": 0.0521, "step": 41000 }, { "epoch": 1.386456607728456, "grad_norm": 0.2747415602207184, "learning_rate": 6.854604359940464e-05, "loss": 0.0548, "step": 41010 }, { "epoch": 1.3867946854187092, "grad_norm": 0.2615818381309509, "learning_rate": 6.853068739880276e-05, "loss": 0.0498, "step": 41020 }, { "epoch": 1.3871327631089625, "grad_norm": 0.28167563676834106, "learning_rate": 6.851532917171276e-05, "loss": 0.0553, "step": 41030 }, { "epoch": 1.3874708407992156, "grad_norm": 0.2193196564912796, "learning_rate": 6.849996891981418e-05, "loss": 0.056, "step": 41040 }, { "epoch": 1.3878089184894689, "grad_norm": 0.3211081027984619, "learning_rate": 6.848460664478684e-05, "loss": 0.0614, "step": 41050 }, { "epoch": 1.3881469961797221, "grad_norm": 0.26892659068107605, "learning_rate": 6.846924234831069e-05, "loss": 0.0509, "step": 41060 }, { "epoch": 1.3884850738699752, "grad_norm": 0.27105212211608887, "learning_rate": 6.845387603206597e-05, "loss": 0.0462, "step": 41070 }, { "epoch": 1.3888231515602285, "grad_norm": 0.20665644109249115, "learning_rate": 6.843850769773312e-05, "loss": 0.0575, "step": 41080 }, { "epoch": 1.3891612292504818, "grad_norm": 0.1917082518339157, "learning_rate": 6.842313734699278e-05, "loss": 0.057, "step": 41090 }, { "epoch": 1.3894993069407349, "grad_norm": 0.2569008469581604, "learning_rate": 6.840776498152584e-05, "loss": 0.0537, "step": 41100 }, { "epoch": 1.3898373846309882, "grad_norm": 0.2410045564174652, "learning_rate": 6.83923906030134e-05, "loss": 0.0562, "step": 41110 }, { "epoch": 1.3901754623212415, "grad_norm": 0.34097957611083984, "learning_rate": 6.837701421313677e-05, "loss": 0.0579, "step": 41120 }, { "epoch": 1.3905135400114945, "grad_norm": 0.27012452483177185, "learning_rate": 6.83616358135775e-05, "loss": 0.054, "step": 41130 }, { "epoch": 1.3908516177017478, "grad_norm": 0.3206806480884552, "learning_rate": 6.834625540601734e-05, "loss": 0.0532, "step": 41140 }, { "epoch": 1.391189695392001, "grad_norm": 0.22769030928611755, "learning_rate": 6.833087299213829e-05, "loss": 0.0461, "step": 41150 }, { "epoch": 1.3915277730822542, "grad_norm": 0.2939351499080658, "learning_rate": 6.83154885736225e-05, "loss": 0.0536, "step": 41160 }, { "epoch": 1.3918658507725075, "grad_norm": 0.2384679615497589, "learning_rate": 6.830010215215244e-05, "loss": 0.0532, "step": 41170 }, { "epoch": 1.3922039284627608, "grad_norm": 0.19158560037612915, "learning_rate": 6.828471372941072e-05, "loss": 0.0555, "step": 41180 }, { "epoch": 1.392542006153014, "grad_norm": 0.5504269003868103, "learning_rate": 6.826932330708018e-05, "loss": 0.0557, "step": 41190 }, { "epoch": 1.3928800838432671, "grad_norm": 0.45098453760147095, "learning_rate": 6.825393088684393e-05, "loss": 0.0611, "step": 41200 }, { "epoch": 1.3932181615335204, "grad_norm": 0.20004945993423462, "learning_rate": 6.823853647038524e-05, "loss": 0.0547, "step": 41210 }, { "epoch": 1.3935562392237737, "grad_norm": 0.24834780395030975, "learning_rate": 6.822314005938762e-05, "loss": 0.0542, "step": 41220 }, { "epoch": 1.3938943169140268, "grad_norm": 0.21322229504585266, "learning_rate": 6.820774165553479e-05, "loss": 0.0509, "step": 41230 }, { "epoch": 1.39423239460428, "grad_norm": 0.28617799282073975, "learning_rate": 6.819234126051072e-05, "loss": 0.0502, "step": 41240 }, { "epoch": 1.3945704722945333, "grad_norm": 0.19076257944107056, "learning_rate": 6.817693887599956e-05, "loss": 0.054, "step": 41250 }, { "epoch": 1.3949085499847866, "grad_norm": 0.21476800739765167, "learning_rate": 6.816153450368568e-05, "loss": 0.0615, "step": 41260 }, { "epoch": 1.3952466276750397, "grad_norm": 0.17996886372566223, "learning_rate": 6.814612814525372e-05, "loss": 0.0542, "step": 41270 }, { "epoch": 1.395584705365293, "grad_norm": 0.19740332663059235, "learning_rate": 6.813071980238846e-05, "loss": 0.0473, "step": 41280 }, { "epoch": 1.3959227830555463, "grad_norm": 0.15909065306186676, "learning_rate": 6.811530947677492e-05, "loss": 0.0529, "step": 41290 }, { "epoch": 1.3962608607457994, "grad_norm": 0.34287235140800476, "learning_rate": 6.809989717009839e-05, "loss": 0.0572, "step": 41300 }, { "epoch": 1.3965989384360527, "grad_norm": 0.31373661756515503, "learning_rate": 6.808448288404431e-05, "loss": 0.058, "step": 41310 }, { "epoch": 1.396937016126306, "grad_norm": 0.33056381344795227, "learning_rate": 6.806906662029838e-05, "loss": 0.0554, "step": 41320 }, { "epoch": 1.397275093816559, "grad_norm": 0.16548511385917664, "learning_rate": 6.805364838054647e-05, "loss": 0.0542, "step": 41330 }, { "epoch": 1.3976131715068123, "grad_norm": 0.15984053909778595, "learning_rate": 6.803822816647474e-05, "loss": 0.0541, "step": 41340 }, { "epoch": 1.3979512491970656, "grad_norm": 0.2504027187824249, "learning_rate": 6.802280597976949e-05, "loss": 0.0492, "step": 41350 }, { "epoch": 1.3982893268873187, "grad_norm": 0.3637964129447937, "learning_rate": 6.800738182211727e-05, "loss": 0.0569, "step": 41360 }, { "epoch": 1.398627404577572, "grad_norm": 0.2393486052751541, "learning_rate": 6.799195569520485e-05, "loss": 0.0615, "step": 41370 }, { "epoch": 1.3989654822678252, "grad_norm": 0.3253112733364105, "learning_rate": 6.79765276007192e-05, "loss": 0.0545, "step": 41380 }, { "epoch": 1.3993035599580783, "grad_norm": 0.27871137857437134, "learning_rate": 6.796109754034753e-05, "loss": 0.0575, "step": 41390 }, { "epoch": 1.3996416376483316, "grad_norm": 0.23615938425064087, "learning_rate": 6.794566551577724e-05, "loss": 0.0548, "step": 41400 }, { "epoch": 1.399979715338585, "grad_norm": 0.36538606882095337, "learning_rate": 6.793023152869594e-05, "loss": 0.0537, "step": 41410 }, { "epoch": 1.400317793028838, "grad_norm": 0.19501446187496185, "learning_rate": 6.791479558079147e-05, "loss": 0.0548, "step": 41420 }, { "epoch": 1.4006558707190913, "grad_norm": 0.3523041605949402, "learning_rate": 6.789935767375191e-05, "loss": 0.0543, "step": 41430 }, { "epoch": 1.4009939484093445, "grad_norm": 0.19307097792625427, "learning_rate": 6.78839178092655e-05, "loss": 0.0505, "step": 41440 }, { "epoch": 1.4013320260995976, "grad_norm": 0.20674152672290802, "learning_rate": 6.786847598902072e-05, "loss": 0.0551, "step": 41450 }, { "epoch": 1.401670103789851, "grad_norm": 0.384167343378067, "learning_rate": 6.785303221470627e-05, "loss": 0.0552, "step": 41460 }, { "epoch": 1.4020081814801042, "grad_norm": 0.272231787443161, "learning_rate": 6.783758648801108e-05, "loss": 0.0544, "step": 41470 }, { "epoch": 1.4023462591703573, "grad_norm": 0.21878482401371002, "learning_rate": 6.782213881062422e-05, "loss": 0.0554, "step": 41480 }, { "epoch": 1.4026843368606106, "grad_norm": 0.25849637389183044, "learning_rate": 6.780668918423508e-05, "loss": 0.0533, "step": 41490 }, { "epoch": 1.4030224145508639, "grad_norm": 0.2939956784248352, "learning_rate": 6.779123761053317e-05, "loss": 0.0515, "step": 41500 }, { "epoch": 1.403360492241117, "grad_norm": 0.17284061014652252, "learning_rate": 6.777578409120826e-05, "loss": 0.0551, "step": 41510 }, { "epoch": 1.4036985699313702, "grad_norm": 0.31848976016044617, "learning_rate": 6.776032862795033e-05, "loss": 0.0479, "step": 41520 }, { "epoch": 1.4040366476216235, "grad_norm": 0.21047604084014893, "learning_rate": 6.774487122244956e-05, "loss": 0.0536, "step": 41530 }, { "epoch": 1.4043747253118766, "grad_norm": 0.279758095741272, "learning_rate": 6.772941187639637e-05, "loss": 0.0515, "step": 41540 }, { "epoch": 1.4047128030021299, "grad_norm": 0.2767636775970459, "learning_rate": 6.771395059148134e-05, "loss": 0.0561, "step": 41550 }, { "epoch": 1.4050508806923832, "grad_norm": 0.35553663969039917, "learning_rate": 6.76984873693953e-05, "loss": 0.0525, "step": 41560 }, { "epoch": 1.4053889583826362, "grad_norm": 0.2494102120399475, "learning_rate": 6.768302221182931e-05, "loss": 0.0516, "step": 41570 }, { "epoch": 1.4057270360728895, "grad_norm": 0.276713103055954, "learning_rate": 6.766755512047457e-05, "loss": 0.0539, "step": 41580 }, { "epoch": 1.4060651137631428, "grad_norm": 0.340522825717926, "learning_rate": 6.765208609702259e-05, "loss": 0.0516, "step": 41590 }, { "epoch": 1.4064031914533959, "grad_norm": 0.17236460745334625, "learning_rate": 6.763661514316499e-05, "loss": 0.0516, "step": 41600 }, { "epoch": 1.4067412691436492, "grad_norm": 0.31888723373413086, "learning_rate": 6.76211422605937e-05, "loss": 0.0494, "step": 41610 }, { "epoch": 1.4070793468339025, "grad_norm": 0.27733975648880005, "learning_rate": 6.760566745100076e-05, "loss": 0.0505, "step": 41620 }, { "epoch": 1.4074174245241555, "grad_norm": 0.1823762059211731, "learning_rate": 6.75901907160785e-05, "loss": 0.0513, "step": 41630 }, { "epoch": 1.4077555022144088, "grad_norm": 0.2783457636833191, "learning_rate": 6.757471205751943e-05, "loss": 0.051, "step": 41640 }, { "epoch": 1.408093579904662, "grad_norm": 0.19232332706451416, "learning_rate": 6.75592314770163e-05, "loss": 0.0529, "step": 41650 }, { "epoch": 1.4084316575949152, "grad_norm": 0.2554399073123932, "learning_rate": 6.7543748976262e-05, "loss": 0.0486, "step": 41660 }, { "epoch": 1.4087697352851685, "grad_norm": 0.17816978693008423, "learning_rate": 6.752826455694968e-05, "loss": 0.0495, "step": 41670 }, { "epoch": 1.4091078129754218, "grad_norm": 0.2149929404258728, "learning_rate": 6.751277822077271e-05, "loss": 0.0537, "step": 41680 }, { "epoch": 1.409445890665675, "grad_norm": 0.16803984344005585, "learning_rate": 6.749728996942465e-05, "loss": 0.0539, "step": 41690 }, { "epoch": 1.4097839683559281, "grad_norm": 0.3847033381462097, "learning_rate": 6.748179980459924e-05, "loss": 0.0551, "step": 41700 }, { "epoch": 1.4101220460461814, "grad_norm": 0.20495513081550598, "learning_rate": 6.746630772799052e-05, "loss": 0.0527, "step": 41710 }, { "epoch": 1.4104601237364347, "grad_norm": 0.21710358560085297, "learning_rate": 6.745081374129262e-05, "loss": 0.0584, "step": 41720 }, { "epoch": 1.4107982014266878, "grad_norm": 0.21646276116371155, "learning_rate": 6.743531784619998e-05, "loss": 0.051, "step": 41730 }, { "epoch": 1.411136279116941, "grad_norm": 0.16480694711208344, "learning_rate": 6.741982004440719e-05, "loss": 0.0535, "step": 41740 }, { "epoch": 1.4114743568071944, "grad_norm": 0.2065192013978958, "learning_rate": 6.740432033760907e-05, "loss": 0.0498, "step": 41750 }, { "epoch": 1.4118124344974476, "grad_norm": 0.36490872502326965, "learning_rate": 6.738881872750066e-05, "loss": 0.0521, "step": 41760 }, { "epoch": 1.4121505121877007, "grad_norm": 0.2320992648601532, "learning_rate": 6.737331521577715e-05, "loss": 0.0497, "step": 41770 }, { "epoch": 1.412488589877954, "grad_norm": 0.236847922205925, "learning_rate": 6.735780980413403e-05, "loss": 0.058, "step": 41780 }, { "epoch": 1.4128266675682073, "grad_norm": 0.29274091124534607, "learning_rate": 6.734230249426692e-05, "loss": 0.0506, "step": 41790 }, { "epoch": 1.4131647452584604, "grad_norm": 0.21035031974315643, "learning_rate": 6.732679328787168e-05, "loss": 0.0462, "step": 41800 }, { "epoch": 1.4135028229487137, "grad_norm": 0.24471759796142578, "learning_rate": 6.731128218664438e-05, "loss": 0.0527, "step": 41810 }, { "epoch": 1.413840900638967, "grad_norm": 0.20487689971923828, "learning_rate": 6.729576919228129e-05, "loss": 0.052, "step": 41820 }, { "epoch": 1.41417897832922, "grad_norm": 0.34915903210639954, "learning_rate": 6.728025430647888e-05, "loss": 0.0551, "step": 41830 }, { "epoch": 1.4145170560194733, "grad_norm": 0.38000354170799255, "learning_rate": 6.726473753093383e-05, "loss": 0.0509, "step": 41840 }, { "epoch": 1.4148551337097266, "grad_norm": 0.23806047439575195, "learning_rate": 6.724921886734305e-05, "loss": 0.0533, "step": 41850 }, { "epoch": 1.4151932113999797, "grad_norm": 0.2172636091709137, "learning_rate": 6.723369831740363e-05, "loss": 0.0527, "step": 41860 }, { "epoch": 1.415531289090233, "grad_norm": 0.3269088864326477, "learning_rate": 6.721817588281288e-05, "loss": 0.0527, "step": 41870 }, { "epoch": 1.4158693667804862, "grad_norm": 0.2961299419403076, "learning_rate": 6.720265156526828e-05, "loss": 0.0567, "step": 41880 }, { "epoch": 1.4162074444707393, "grad_norm": 0.25089195370674133, "learning_rate": 6.718712536646758e-05, "loss": 0.0564, "step": 41890 }, { "epoch": 1.4165455221609926, "grad_norm": 0.26546990871429443, "learning_rate": 6.71715972881087e-05, "loss": 0.057, "step": 41900 }, { "epoch": 1.416883599851246, "grad_norm": 0.23306335508823395, "learning_rate": 6.715606733188974e-05, "loss": 0.0581, "step": 41910 }, { "epoch": 1.417221677541499, "grad_norm": 0.2651614248752594, "learning_rate": 6.714053549950909e-05, "loss": 0.0538, "step": 41920 }, { "epoch": 1.4175597552317523, "grad_norm": 0.2568824887275696, "learning_rate": 6.71250017926652e-05, "loss": 0.0487, "step": 41930 }, { "epoch": 1.4178978329220056, "grad_norm": 0.3890722393989563, "learning_rate": 6.71094662130569e-05, "loss": 0.0532, "step": 41940 }, { "epoch": 1.4182359106122586, "grad_norm": 0.47645050287246704, "learning_rate": 6.709392876238307e-05, "loss": 0.0514, "step": 41950 }, { "epoch": 1.418573988302512, "grad_norm": 0.2416907548904419, "learning_rate": 6.707838944234289e-05, "loss": 0.0571, "step": 41960 }, { "epoch": 1.4189120659927652, "grad_norm": 0.18262077867984772, "learning_rate": 6.706284825463574e-05, "loss": 0.0513, "step": 41970 }, { "epoch": 1.4192501436830183, "grad_norm": 0.4089345932006836, "learning_rate": 6.704730520096114e-05, "loss": 0.0524, "step": 41980 }, { "epoch": 1.4195882213732716, "grad_norm": 0.3403709828853607, "learning_rate": 6.703176028301888e-05, "loss": 0.0553, "step": 41990 }, { "epoch": 1.4199262990635249, "grad_norm": 0.31875357031822205, "learning_rate": 6.701621350250892e-05, "loss": 0.0558, "step": 42000 }, { "epoch": 1.420264376753778, "grad_norm": 0.19606377184391022, "learning_rate": 6.700066486113144e-05, "loss": 0.0539, "step": 42010 }, { "epoch": 1.4206024544440312, "grad_norm": 0.21401362121105194, "learning_rate": 6.69851143605868e-05, "loss": 0.0501, "step": 42020 }, { "epoch": 1.4209405321342845, "grad_norm": 0.17136111855506897, "learning_rate": 6.696956200257561e-05, "loss": 0.0547, "step": 42030 }, { "epoch": 1.4212786098245376, "grad_norm": 0.2959972321987152, "learning_rate": 6.695400778879863e-05, "loss": 0.0463, "step": 42040 }, { "epoch": 1.4216166875147909, "grad_norm": 0.19119596481323242, "learning_rate": 6.693845172095683e-05, "loss": 0.0475, "step": 42050 }, { "epoch": 1.4219547652050442, "grad_norm": 0.26675641536712646, "learning_rate": 6.692289380075142e-05, "loss": 0.0547, "step": 42060 }, { "epoch": 1.4222928428952972, "grad_norm": 0.3834295868873596, "learning_rate": 6.690733402988379e-05, "loss": 0.061, "step": 42070 }, { "epoch": 1.4226309205855505, "grad_norm": 0.32454943656921387, "learning_rate": 6.689177241005553e-05, "loss": 0.0559, "step": 42080 }, { "epoch": 1.4229689982758038, "grad_norm": 0.20274341106414795, "learning_rate": 6.687620894296844e-05, "loss": 0.059, "step": 42090 }, { "epoch": 1.4233070759660569, "grad_norm": 0.2460133135318756, "learning_rate": 6.686064363032451e-05, "loss": 0.0526, "step": 42100 }, { "epoch": 1.4236451536563102, "grad_norm": 0.1839759647846222, "learning_rate": 6.684507647382596e-05, "loss": 0.0488, "step": 42110 }, { "epoch": 1.4239832313465635, "grad_norm": 0.1978742927312851, "learning_rate": 6.682950747517513e-05, "loss": 0.0477, "step": 42120 }, { "epoch": 1.4243213090368165, "grad_norm": 0.25363293290138245, "learning_rate": 6.68139366360747e-05, "loss": 0.0512, "step": 42130 }, { "epoch": 1.4246593867270698, "grad_norm": 0.36267632246017456, "learning_rate": 6.679836395822744e-05, "loss": 0.0593, "step": 42140 }, { "epoch": 1.4249974644173231, "grad_norm": 0.41601115465164185, "learning_rate": 6.678278944333633e-05, "loss": 0.0585, "step": 42150 }, { "epoch": 1.4253355421075764, "grad_norm": 0.20013223588466644, "learning_rate": 6.676721309310462e-05, "loss": 0.0539, "step": 42160 }, { "epoch": 1.4256736197978295, "grad_norm": 0.23276670277118683, "learning_rate": 6.67516349092357e-05, "loss": 0.0509, "step": 42170 }, { "epoch": 1.4260116974880828, "grad_norm": 0.18384914100170135, "learning_rate": 6.673605489343315e-05, "loss": 0.0571, "step": 42180 }, { "epoch": 1.426349775178336, "grad_norm": 0.26164790987968445, "learning_rate": 6.672047304740081e-05, "loss": 0.051, "step": 42190 }, { "epoch": 1.4266878528685891, "grad_norm": 0.2225315272808075, "learning_rate": 6.670488937284268e-05, "loss": 0.0563, "step": 42200 }, { "epoch": 1.4270259305588424, "grad_norm": 0.2781025171279907, "learning_rate": 6.668930387146294e-05, "loss": 0.0542, "step": 42210 }, { "epoch": 1.4273640082490957, "grad_norm": 0.19401425123214722, "learning_rate": 6.667371654496605e-05, "loss": 0.0524, "step": 42220 }, { "epoch": 1.427702085939349, "grad_norm": 0.2698245048522949, "learning_rate": 6.665812739505659e-05, "loss": 0.0475, "step": 42230 }, { "epoch": 1.428040163629602, "grad_norm": 0.15731768310070038, "learning_rate": 6.664253642343935e-05, "loss": 0.0523, "step": 42240 }, { "epoch": 1.4283782413198554, "grad_norm": 0.2265152931213379, "learning_rate": 6.662694363181935e-05, "loss": 0.0476, "step": 42250 }, { "epoch": 1.4287163190101086, "grad_norm": 0.4247860610485077, "learning_rate": 6.66113490219018e-05, "loss": 0.0566, "step": 42260 }, { "epoch": 1.4290543967003617, "grad_norm": 0.369806170463562, "learning_rate": 6.659575259539207e-05, "loss": 0.0518, "step": 42270 }, { "epoch": 1.429392474390615, "grad_norm": 0.20672692358493805, "learning_rate": 6.65801543539958e-05, "loss": 0.0522, "step": 42280 }, { "epoch": 1.4297305520808683, "grad_norm": 0.3102329671382904, "learning_rate": 6.65645542994188e-05, "loss": 0.0559, "step": 42290 }, { "epoch": 1.4300686297711214, "grad_norm": 0.2886853516101837, "learning_rate": 6.654895243336702e-05, "loss": 0.0508, "step": 42300 }, { "epoch": 1.4304067074613747, "grad_norm": 0.22626405954360962, "learning_rate": 6.65333487575467e-05, "loss": 0.0488, "step": 42310 }, { "epoch": 1.430744785151628, "grad_norm": 0.3035784363746643, "learning_rate": 6.651774327366422e-05, "loss": 0.0511, "step": 42320 }, { "epoch": 1.431082862841881, "grad_norm": 0.34021997451782227, "learning_rate": 6.650213598342613e-05, "loss": 0.0548, "step": 42330 }, { "epoch": 1.4314209405321343, "grad_norm": 0.23748083412647247, "learning_rate": 6.648652688853931e-05, "loss": 0.05, "step": 42340 }, { "epoch": 1.4317590182223876, "grad_norm": 0.4290870726108551, "learning_rate": 6.647091599071066e-05, "loss": 0.0553, "step": 42350 }, { "epoch": 1.4320970959126407, "grad_norm": 0.1802455633878708, "learning_rate": 6.645530329164743e-05, "loss": 0.0538, "step": 42360 }, { "epoch": 1.432435173602894, "grad_norm": 0.29167675971984863, "learning_rate": 6.643968879305694e-05, "loss": 0.0538, "step": 42370 }, { "epoch": 1.4327732512931473, "grad_norm": 0.2868972420692444, "learning_rate": 6.642407249664683e-05, "loss": 0.0542, "step": 42380 }, { "epoch": 1.4331113289834003, "grad_norm": 0.3010658323764801, "learning_rate": 6.640845440412483e-05, "loss": 0.0542, "step": 42390 }, { "epoch": 1.4334494066736536, "grad_norm": 0.21430447697639465, "learning_rate": 6.639283451719893e-05, "loss": 0.0526, "step": 42400 }, { "epoch": 1.433787484363907, "grad_norm": 0.2398717850446701, "learning_rate": 6.637721283757729e-05, "loss": 0.0575, "step": 42410 }, { "epoch": 1.43412556205416, "grad_norm": 0.19720369577407837, "learning_rate": 6.636158936696828e-05, "loss": 0.0512, "step": 42420 }, { "epoch": 1.4344636397444133, "grad_norm": 0.22142598032951355, "learning_rate": 6.634596410708047e-05, "loss": 0.0588, "step": 42430 }, { "epoch": 1.4348017174346666, "grad_norm": 0.2144172191619873, "learning_rate": 6.633033705962257e-05, "loss": 0.054, "step": 42440 }, { "epoch": 1.4351397951249196, "grad_norm": 0.32784774899482727, "learning_rate": 6.631470822630359e-05, "loss": 0.0495, "step": 42450 }, { "epoch": 1.435477872815173, "grad_norm": 0.20762452483177185, "learning_rate": 6.629907760883263e-05, "loss": 0.0522, "step": 42460 }, { "epoch": 1.4358159505054262, "grad_norm": 0.2508732080459595, "learning_rate": 6.628344520891905e-05, "loss": 0.0529, "step": 42470 }, { "epoch": 1.4361540281956793, "grad_norm": 0.25236520171165466, "learning_rate": 6.62678110282724e-05, "loss": 0.053, "step": 42480 }, { "epoch": 1.4364921058859326, "grad_norm": 0.3528525233268738, "learning_rate": 6.625217506860237e-05, "loss": 0.0508, "step": 42490 }, { "epoch": 1.4368301835761859, "grad_norm": 0.313615083694458, "learning_rate": 6.623653733161892e-05, "loss": 0.0597, "step": 42500 }, { "epoch": 1.437168261266439, "grad_norm": 0.21298475563526154, "learning_rate": 6.622089781903217e-05, "loss": 0.0527, "step": 42510 }, { "epoch": 1.4375063389566922, "grad_norm": 0.23565126955509186, "learning_rate": 6.620525653255241e-05, "loss": 0.0561, "step": 42520 }, { "epoch": 1.4378444166469455, "grad_norm": 0.2441009134054184, "learning_rate": 6.618961347389015e-05, "loss": 0.0588, "step": 42530 }, { "epoch": 1.4381824943371986, "grad_norm": 0.2935851514339447, "learning_rate": 6.617396864475613e-05, "loss": 0.0565, "step": 42540 }, { "epoch": 1.4385205720274519, "grad_norm": 0.24666370451450348, "learning_rate": 6.61583220468612e-05, "loss": 0.0561, "step": 42550 }, { "epoch": 1.4388586497177052, "grad_norm": 0.1804620325565338, "learning_rate": 6.614267368191645e-05, "loss": 0.0582, "step": 42560 }, { "epoch": 1.4391967274079582, "grad_norm": 0.3383384346961975, "learning_rate": 6.61270235516332e-05, "loss": 0.0539, "step": 42570 }, { "epoch": 1.4395348050982115, "grad_norm": 0.20928189158439636, "learning_rate": 6.611137165772287e-05, "loss": 0.0554, "step": 42580 }, { "epoch": 1.4398728827884648, "grad_norm": 0.34767845273017883, "learning_rate": 6.60957180018972e-05, "loss": 0.0589, "step": 42590 }, { "epoch": 1.4402109604787179, "grad_norm": 0.4282507598400116, "learning_rate": 6.608006258586797e-05, "loss": 0.0552, "step": 42600 }, { "epoch": 1.4405490381689712, "grad_norm": 0.7675418853759766, "learning_rate": 6.606440541134731e-05, "loss": 0.0515, "step": 42610 }, { "epoch": 1.4408871158592245, "grad_norm": 0.28358325362205505, "learning_rate": 6.60487464800474e-05, "loss": 0.0483, "step": 42620 }, { "epoch": 1.4412251935494775, "grad_norm": 0.29401999711990356, "learning_rate": 6.603308579368071e-05, "loss": 0.0529, "step": 42630 }, { "epoch": 1.4415632712397308, "grad_norm": 0.20418037474155426, "learning_rate": 6.601742335395987e-05, "loss": 0.0553, "step": 42640 }, { "epoch": 1.4419013489299841, "grad_norm": 0.22914226353168488, "learning_rate": 6.600175916259769e-05, "loss": 0.0543, "step": 42650 }, { "epoch": 1.4422394266202374, "grad_norm": 0.2833709120750427, "learning_rate": 6.598609322130718e-05, "loss": 0.06, "step": 42660 }, { "epoch": 1.4425775043104905, "grad_norm": 0.3538619875907898, "learning_rate": 6.597042553180155e-05, "loss": 0.049, "step": 42670 }, { "epoch": 1.4429155820007438, "grad_norm": 0.23798684775829315, "learning_rate": 6.59547560957942e-05, "loss": 0.0515, "step": 42680 }, { "epoch": 1.443253659690997, "grad_norm": 0.15105508267879486, "learning_rate": 6.59390849149987e-05, "loss": 0.0578, "step": 42690 }, { "epoch": 1.4435917373812501, "grad_norm": 0.3825224041938782, "learning_rate": 6.592341199112886e-05, "loss": 0.0514, "step": 42700 }, { "epoch": 1.4439298150715034, "grad_norm": 0.22801490128040314, "learning_rate": 6.590773732589861e-05, "loss": 0.0603, "step": 42710 }, { "epoch": 1.4442678927617567, "grad_norm": 0.21042221784591675, "learning_rate": 6.589206092102213e-05, "loss": 0.0558, "step": 42720 }, { "epoch": 1.44460597045201, "grad_norm": 0.23667654395103455, "learning_rate": 6.587638277821375e-05, "loss": 0.0565, "step": 42730 }, { "epoch": 1.444944048142263, "grad_norm": 0.3521539270877838, "learning_rate": 6.586070289918803e-05, "loss": 0.0485, "step": 42740 }, { "epoch": 1.4452821258325164, "grad_norm": 0.42672199010849, "learning_rate": 6.584502128565968e-05, "loss": 0.0528, "step": 42750 }, { "epoch": 1.4456202035227697, "grad_norm": 0.20282545685768127, "learning_rate": 6.582933793934363e-05, "loss": 0.0597, "step": 42760 }, { "epoch": 1.4459582812130227, "grad_norm": 0.16473419964313507, "learning_rate": 6.581365286195498e-05, "loss": 0.0543, "step": 42770 }, { "epoch": 1.446296358903276, "grad_norm": 0.3297501802444458, "learning_rate": 6.579796605520903e-05, "loss": 0.0514, "step": 42780 }, { "epoch": 1.4466344365935293, "grad_norm": 0.30342909693717957, "learning_rate": 6.578227752082127e-05, "loss": 0.0503, "step": 42790 }, { "epoch": 1.4469725142837824, "grad_norm": 0.24096503853797913, "learning_rate": 6.576658726050735e-05, "loss": 0.0566, "step": 42800 }, { "epoch": 1.4473105919740357, "grad_norm": 0.20403257012367249, "learning_rate": 6.575089527598316e-05, "loss": 0.0589, "step": 42810 }, { "epoch": 1.447648669664289, "grad_norm": 0.23512613773345947, "learning_rate": 6.573520156896476e-05, "loss": 0.057, "step": 42820 }, { "epoch": 1.447986747354542, "grad_norm": 0.2405591458082199, "learning_rate": 6.571950614116835e-05, "loss": 0.0532, "step": 42830 }, { "epoch": 1.4483248250447953, "grad_norm": 0.3026515543460846, "learning_rate": 6.570380899431039e-05, "loss": 0.0504, "step": 42840 }, { "epoch": 1.4486629027350486, "grad_norm": 0.2605462372303009, "learning_rate": 6.568811013010749e-05, "loss": 0.0565, "step": 42850 }, { "epoch": 1.4490009804253017, "grad_norm": 0.2597769796848297, "learning_rate": 6.567240955027645e-05, "loss": 0.0556, "step": 42860 }, { "epoch": 1.449339058115555, "grad_norm": 0.3308729827404022, "learning_rate": 6.565670725653427e-05, "loss": 0.0567, "step": 42870 }, { "epoch": 1.4496771358058083, "grad_norm": 0.38650214672088623, "learning_rate": 6.564100325059812e-05, "loss": 0.0502, "step": 42880 }, { "epoch": 1.4500152134960613, "grad_norm": 0.14830127358436584, "learning_rate": 6.562529753418538e-05, "loss": 0.0494, "step": 42890 }, { "epoch": 1.4503532911863146, "grad_norm": 0.3253079950809479, "learning_rate": 6.56095901090136e-05, "loss": 0.0503, "step": 42900 }, { "epoch": 1.450691368876568, "grad_norm": 0.4593920111656189, "learning_rate": 6.55938809768005e-05, "loss": 0.0507, "step": 42910 }, { "epoch": 1.451029446566821, "grad_norm": 0.2662472426891327, "learning_rate": 6.557817013926402e-05, "loss": 0.0576, "step": 42920 }, { "epoch": 1.4513675242570743, "grad_norm": 0.2785075306892395, "learning_rate": 6.556245759812227e-05, "loss": 0.0574, "step": 42930 }, { "epoch": 1.4517056019473276, "grad_norm": 0.3122890591621399, "learning_rate": 6.554674335509357e-05, "loss": 0.0485, "step": 42940 }, { "epoch": 1.4520436796375806, "grad_norm": 0.2346867024898529, "learning_rate": 6.553102741189638e-05, "loss": 0.0493, "step": 42950 }, { "epoch": 1.452381757327834, "grad_norm": 0.23244506120681763, "learning_rate": 6.55153097702494e-05, "loss": 0.0515, "step": 42960 }, { "epoch": 1.4527198350180872, "grad_norm": 0.19276157021522522, "learning_rate": 6.549959043187146e-05, "loss": 0.051, "step": 42970 }, { "epoch": 1.4530579127083403, "grad_norm": 0.3033429980278015, "learning_rate": 6.54838693984816e-05, "loss": 0.0475, "step": 42980 }, { "epoch": 1.4533959903985936, "grad_norm": 0.21440407633781433, "learning_rate": 6.546814667179909e-05, "loss": 0.0505, "step": 42990 }, { "epoch": 1.4537340680888469, "grad_norm": 0.37057867646217346, "learning_rate": 6.545242225354328e-05, "loss": 0.0526, "step": 43000 }, { "epoch": 1.4540721457791, "grad_norm": 0.3352847397327423, "learning_rate": 6.543669614543383e-05, "loss": 0.0503, "step": 43010 }, { "epoch": 1.4544102234693532, "grad_norm": 0.19150100648403168, "learning_rate": 6.542096834919049e-05, "loss": 0.0547, "step": 43020 }, { "epoch": 1.4547483011596065, "grad_norm": 0.30482691526412964, "learning_rate": 6.540523886653324e-05, "loss": 0.05, "step": 43030 }, { "epoch": 1.4550863788498596, "grad_norm": 0.25575578212738037, "learning_rate": 6.538950769918222e-05, "loss": 0.0527, "step": 43040 }, { "epoch": 1.4554244565401129, "grad_norm": 0.33227336406707764, "learning_rate": 6.537377484885779e-05, "loss": 0.0513, "step": 43050 }, { "epoch": 1.4557625342303662, "grad_norm": 0.31205856800079346, "learning_rate": 6.535804031728045e-05, "loss": 0.0493, "step": 43060 }, { "epoch": 1.4561006119206192, "grad_norm": 0.19515898823738098, "learning_rate": 6.534230410617092e-05, "loss": 0.0585, "step": 43070 }, { "epoch": 1.4564386896108725, "grad_norm": 0.29254090785980225, "learning_rate": 6.532656621725007e-05, "loss": 0.0526, "step": 43080 }, { "epoch": 1.4567767673011258, "grad_norm": 0.2028902918100357, "learning_rate": 6.531082665223898e-05, "loss": 0.053, "step": 43090 }, { "epoch": 1.4571148449913789, "grad_norm": 0.39587312936782837, "learning_rate": 6.529508541285889e-05, "loss": 0.0557, "step": 43100 }, { "epoch": 1.4574529226816322, "grad_norm": 0.22128872573375702, "learning_rate": 6.527934250083128e-05, "loss": 0.0491, "step": 43110 }, { "epoch": 1.4577910003718855, "grad_norm": 0.3620302677154541, "learning_rate": 6.526359791787775e-05, "loss": 0.0556, "step": 43120 }, { "epoch": 1.4581290780621388, "grad_norm": 0.21347938477993011, "learning_rate": 6.524785166572009e-05, "loss": 0.0504, "step": 43130 }, { "epoch": 1.4584671557523918, "grad_norm": 0.39739274978637695, "learning_rate": 6.523210374608029e-05, "loss": 0.0515, "step": 43140 }, { "epoch": 1.4588052334426451, "grad_norm": 0.21008777618408203, "learning_rate": 6.521635416068054e-05, "loss": 0.0561, "step": 43150 }, { "epoch": 1.4591433111328984, "grad_norm": 0.2833227217197418, "learning_rate": 6.520060291124317e-05, "loss": 0.0541, "step": 43160 }, { "epoch": 1.4594813888231515, "grad_norm": 0.4936813414096832, "learning_rate": 6.518484999949072e-05, "loss": 0.0552, "step": 43170 }, { "epoch": 1.4598194665134048, "grad_norm": 0.2529200315475464, "learning_rate": 6.516909542714591e-05, "loss": 0.0542, "step": 43180 }, { "epoch": 1.460157544203658, "grad_norm": 0.2921448051929474, "learning_rate": 6.515333919593161e-05, "loss": 0.0518, "step": 43190 }, { "epoch": 1.4604956218939114, "grad_norm": 0.4926743507385254, "learning_rate": 6.513758130757094e-05, "loss": 0.0531, "step": 43200 }, { "epoch": 1.4608336995841644, "grad_norm": 0.25554436445236206, "learning_rate": 6.512182176378713e-05, "loss": 0.0484, "step": 43210 }, { "epoch": 1.4611717772744177, "grad_norm": 0.1848013997077942, "learning_rate": 6.510606056630362e-05, "loss": 0.0511, "step": 43220 }, { "epoch": 1.461509854964671, "grad_norm": 0.18154779076576233, "learning_rate": 6.509029771684405e-05, "loss": 0.0513, "step": 43230 }, { "epoch": 1.461847932654924, "grad_norm": 0.2862186133861542, "learning_rate": 6.50745332171322e-05, "loss": 0.0514, "step": 43240 }, { "epoch": 1.4621860103451774, "grad_norm": 0.4366658926010132, "learning_rate": 6.505876706889207e-05, "loss": 0.0546, "step": 43250 }, { "epoch": 1.4625240880354307, "grad_norm": 0.19469071924686432, "learning_rate": 6.504299927384781e-05, "loss": 0.0521, "step": 43260 }, { "epoch": 1.4628621657256837, "grad_norm": 0.2574637830257416, "learning_rate": 6.502722983372378e-05, "loss": 0.0607, "step": 43270 }, { "epoch": 1.463200243415937, "grad_norm": 1.1112233400344849, "learning_rate": 6.501145875024446e-05, "loss": 0.0499, "step": 43280 }, { "epoch": 1.4635383211061903, "grad_norm": 0.22651071846485138, "learning_rate": 6.499568602513459e-05, "loss": 0.0483, "step": 43290 }, { "epoch": 1.4638763987964434, "grad_norm": 0.4298270642757416, "learning_rate": 6.497991166011903e-05, "loss": 0.0498, "step": 43300 }, { "epoch": 1.4642144764866967, "grad_norm": 0.5971812009811401, "learning_rate": 6.49641356569229e-05, "loss": 0.0546, "step": 43310 }, { "epoch": 1.46455255417695, "grad_norm": 0.8908324241638184, "learning_rate": 6.494835801727135e-05, "loss": 0.0515, "step": 43320 }, { "epoch": 1.464890631867203, "grad_norm": 0.25777292251586914, "learning_rate": 6.493257874288986e-05, "loss": 0.0592, "step": 43330 }, { "epoch": 1.4652287095574563, "grad_norm": 0.3913516402244568, "learning_rate": 6.491679783550399e-05, "loss": 0.0573, "step": 43340 }, { "epoch": 1.4655667872477096, "grad_norm": 0.2031877189874649, "learning_rate": 6.490101529683957e-05, "loss": 0.0557, "step": 43350 }, { "epoch": 1.4659048649379627, "grad_norm": 0.5346220135688782, "learning_rate": 6.488523112862248e-05, "loss": 0.0523, "step": 43360 }, { "epoch": 1.466242942628216, "grad_norm": 0.34495341777801514, "learning_rate": 6.486944533257891e-05, "loss": 0.0519, "step": 43370 }, { "epoch": 1.4665810203184693, "grad_norm": 0.2264033555984497, "learning_rate": 6.485365791043517e-05, "loss": 0.0468, "step": 43380 }, { "epoch": 1.4669190980087223, "grad_norm": 0.2824539840221405, "learning_rate": 6.483786886391772e-05, "loss": 0.0576, "step": 43390 }, { "epoch": 1.4672571756989756, "grad_norm": 0.8429773449897766, "learning_rate": 6.482207819475323e-05, "loss": 0.0514, "step": 43400 }, { "epoch": 1.467595253389229, "grad_norm": 0.3979548215866089, "learning_rate": 6.480628590466857e-05, "loss": 0.0551, "step": 43410 }, { "epoch": 1.467933331079482, "grad_norm": 0.30775222182273865, "learning_rate": 6.479049199539074e-05, "loss": 0.0551, "step": 43420 }, { "epoch": 1.4682714087697353, "grad_norm": 0.20858687162399292, "learning_rate": 6.477469646864693e-05, "loss": 0.0512, "step": 43430 }, { "epoch": 1.4686094864599886, "grad_norm": 0.23183225095272064, "learning_rate": 6.475889932616454e-05, "loss": 0.0516, "step": 43440 }, { "epoch": 1.4689475641502416, "grad_norm": 0.4046320915222168, "learning_rate": 6.474310056967111e-05, "loss": 0.0533, "step": 43450 }, { "epoch": 1.469285641840495, "grad_norm": 0.7227317094802856, "learning_rate": 6.472730020089437e-05, "loss": 0.0457, "step": 43460 }, { "epoch": 1.4696237195307482, "grad_norm": 0.540837824344635, "learning_rate": 6.471149822156222e-05, "loss": 0.0513, "step": 43470 }, { "epoch": 1.4699617972210013, "grad_norm": 0.2716916501522064, "learning_rate": 6.469569463340275e-05, "loss": 0.0535, "step": 43480 }, { "epoch": 1.4702998749112546, "grad_norm": 0.2351144701242447, "learning_rate": 6.467988943814418e-05, "loss": 0.0608, "step": 43490 }, { "epoch": 1.4706379526015079, "grad_norm": 0.5300149321556091, "learning_rate": 6.4664082637515e-05, "loss": 0.0532, "step": 43500 }, { "epoch": 1.470976030291761, "grad_norm": 0.22445161640644073, "learning_rate": 6.464827423324377e-05, "loss": 0.0509, "step": 43510 }, { "epoch": 1.4713141079820142, "grad_norm": 0.2649249732494354, "learning_rate": 6.463246422705931e-05, "loss": 0.0511, "step": 43520 }, { "epoch": 1.4716521856722675, "grad_norm": 0.16771849989891052, "learning_rate": 6.461665262069055e-05, "loss": 0.0534, "step": 43530 }, { "epoch": 1.4719902633625206, "grad_norm": 0.1822470873594284, "learning_rate": 6.460083941586665e-05, "loss": 0.0491, "step": 43540 }, { "epoch": 1.4723283410527739, "grad_norm": 0.21395638585090637, "learning_rate": 6.45850246143169e-05, "loss": 0.0557, "step": 43550 }, { "epoch": 1.4726664187430272, "grad_norm": 0.2914879620075226, "learning_rate": 6.456920821777077e-05, "loss": 0.0569, "step": 43560 }, { "epoch": 1.4730044964332802, "grad_norm": 0.1687290221452713, "learning_rate": 6.455339022795796e-05, "loss": 0.0496, "step": 43570 }, { "epoch": 1.4733425741235335, "grad_norm": 0.19348442554473877, "learning_rate": 6.453757064660825e-05, "loss": 0.0448, "step": 43580 }, { "epoch": 1.4736806518137868, "grad_norm": 0.3111813962459564, "learning_rate": 6.452174947545169e-05, "loss": 0.0493, "step": 43590 }, { "epoch": 1.47401872950404, "grad_norm": 0.23508590459823608, "learning_rate": 6.450592671621842e-05, "loss": 0.055, "step": 43600 }, { "epoch": 1.4743568071942932, "grad_norm": 0.2147103101015091, "learning_rate": 6.449010237063882e-05, "loss": 0.0584, "step": 43610 }, { "epoch": 1.4746948848845465, "grad_norm": 0.18013858795166016, "learning_rate": 6.447427644044343e-05, "loss": 0.0523, "step": 43620 }, { "epoch": 1.4750329625747998, "grad_norm": 0.2577231824398041, "learning_rate": 6.44584489273629e-05, "loss": 0.0566, "step": 43630 }, { "epoch": 1.4753710402650528, "grad_norm": 0.2714589834213257, "learning_rate": 6.444261983312817e-05, "loss": 0.0504, "step": 43640 }, { "epoch": 1.4757091179553061, "grad_norm": 0.2049589306116104, "learning_rate": 6.442678915947023e-05, "loss": 0.0516, "step": 43650 }, { "epoch": 1.4760471956455594, "grad_norm": 0.28307682275772095, "learning_rate": 6.441095690812032e-05, "loss": 0.0527, "step": 43660 }, { "epoch": 1.4763852733358125, "grad_norm": 0.245884969830513, "learning_rate": 6.439512308080982e-05, "loss": 0.0567, "step": 43670 }, { "epoch": 1.4767233510260658, "grad_norm": 0.2764834761619568, "learning_rate": 6.43792876792703e-05, "loss": 0.0504, "step": 43680 }, { "epoch": 1.477061428716319, "grad_norm": 0.25341662764549255, "learning_rate": 6.43634507052335e-05, "loss": 0.0496, "step": 43690 }, { "epoch": 1.4773995064065724, "grad_norm": 0.1739863157272339, "learning_rate": 6.434761216043135e-05, "loss": 0.0527, "step": 43700 }, { "epoch": 1.4777375840968254, "grad_norm": 0.25564688444137573, "learning_rate": 6.433177204659589e-05, "loss": 0.0477, "step": 43710 }, { "epoch": 1.4780756617870787, "grad_norm": 0.2451440393924713, "learning_rate": 6.431593036545938e-05, "loss": 0.0458, "step": 43720 }, { "epoch": 1.478413739477332, "grad_norm": 0.21671162545681, "learning_rate": 6.430008711875426e-05, "loss": 0.0478, "step": 43730 }, { "epoch": 1.478751817167585, "grad_norm": 0.23036733269691467, "learning_rate": 6.42842423082131e-05, "loss": 0.0499, "step": 43740 }, { "epoch": 1.4790898948578384, "grad_norm": 0.3720935881137848, "learning_rate": 6.426839593556868e-05, "loss": 0.0538, "step": 43750 }, { "epoch": 1.4794279725480917, "grad_norm": 0.18640461564064026, "learning_rate": 6.425254800255392e-05, "loss": 0.0506, "step": 43760 }, { "epoch": 1.4797660502383447, "grad_norm": 0.27229198813438416, "learning_rate": 6.423669851090196e-05, "loss": 0.0511, "step": 43770 }, { "epoch": 1.480104127928598, "grad_norm": 0.2461899071931839, "learning_rate": 6.422084746234604e-05, "loss": 0.0532, "step": 43780 }, { "epoch": 1.4804422056188513, "grad_norm": 0.3254064917564392, "learning_rate": 6.420499485861965e-05, "loss": 0.0559, "step": 43790 }, { "epoch": 1.4807802833091044, "grad_norm": 0.2626393437385559, "learning_rate": 6.418914070145636e-05, "loss": 0.0513, "step": 43800 }, { "epoch": 1.4811183609993577, "grad_norm": 0.3132181465625763, "learning_rate": 6.417328499258996e-05, "loss": 0.0543, "step": 43810 }, { "epoch": 1.481456438689611, "grad_norm": 0.1818085014820099, "learning_rate": 6.415742773375446e-05, "loss": 0.0447, "step": 43820 }, { "epoch": 1.481794516379864, "grad_norm": 0.44559434056282043, "learning_rate": 6.414156892668393e-05, "loss": 0.0481, "step": 43830 }, { "epoch": 1.4821325940701173, "grad_norm": 0.2525460422039032, "learning_rate": 6.412570857311267e-05, "loss": 0.0502, "step": 43840 }, { "epoch": 1.4824706717603706, "grad_norm": 0.357358455657959, "learning_rate": 6.410984667477518e-05, "loss": 0.0503, "step": 43850 }, { "epoch": 1.4828087494506237, "grad_norm": 0.32728511095046997, "learning_rate": 6.409398323340607e-05, "loss": 0.0511, "step": 43860 }, { "epoch": 1.483146827140877, "grad_norm": 0.3562602400779724, "learning_rate": 6.407811825074012e-05, "loss": 0.0566, "step": 43870 }, { "epoch": 1.4834849048311303, "grad_norm": 0.40589195489883423, "learning_rate": 6.406225172851234e-05, "loss": 0.0527, "step": 43880 }, { "epoch": 1.4838229825213833, "grad_norm": 0.37065479159355164, "learning_rate": 6.404638366845786e-05, "loss": 0.0579, "step": 43890 }, { "epoch": 1.4841610602116366, "grad_norm": 0.2979867160320282, "learning_rate": 6.403051407231196e-05, "loss": 0.0495, "step": 43900 }, { "epoch": 1.48449913790189, "grad_norm": 0.19294843077659607, "learning_rate": 6.401464294181016e-05, "loss": 0.0483, "step": 43910 }, { "epoch": 1.484837215592143, "grad_norm": 0.243707537651062, "learning_rate": 6.399877027868808e-05, "loss": 0.0568, "step": 43920 }, { "epoch": 1.4851752932823963, "grad_norm": 0.18236175179481506, "learning_rate": 6.39828960846815e-05, "loss": 0.0494, "step": 43930 }, { "epoch": 1.4855133709726496, "grad_norm": 0.21260355412960052, "learning_rate": 6.396702036152644e-05, "loss": 0.054, "step": 43940 }, { "epoch": 1.4858514486629026, "grad_norm": 0.2507442831993103, "learning_rate": 6.395114311095905e-05, "loss": 0.0544, "step": 43950 }, { "epoch": 1.486189526353156, "grad_norm": 0.25350525975227356, "learning_rate": 6.393526433471559e-05, "loss": 0.0523, "step": 43960 }, { "epoch": 1.4865276040434092, "grad_norm": 0.3025490343570709, "learning_rate": 6.391938403453262e-05, "loss": 0.0482, "step": 43970 }, { "epoch": 1.4868656817336623, "grad_norm": 0.2377127856016159, "learning_rate": 6.390350221214671e-05, "loss": 0.0567, "step": 43980 }, { "epoch": 1.4872037594239156, "grad_norm": 0.25803619623184204, "learning_rate": 6.38876188692947e-05, "loss": 0.0555, "step": 43990 }, { "epoch": 1.4875418371141689, "grad_norm": 0.1658814549446106, "learning_rate": 6.387173400771358e-05, "loss": 0.0512, "step": 44000 }, { "epoch": 1.487879914804422, "grad_norm": 0.19762273132801056, "learning_rate": 6.385584762914047e-05, "loss": 0.0519, "step": 44010 }, { "epoch": 1.4882179924946752, "grad_norm": 0.24954544007778168, "learning_rate": 6.38399597353127e-05, "loss": 0.0522, "step": 44020 }, { "epoch": 1.4885560701849285, "grad_norm": 0.45403704047203064, "learning_rate": 6.382407032796775e-05, "loss": 0.0548, "step": 44030 }, { "epoch": 1.4888941478751816, "grad_norm": 0.24927300214767456, "learning_rate": 6.380817940884325e-05, "loss": 0.0506, "step": 44040 }, { "epoch": 1.4892322255654349, "grad_norm": 0.5923009514808655, "learning_rate": 6.379228697967702e-05, "loss": 0.0558, "step": 44050 }, { "epoch": 1.4895703032556882, "grad_norm": 0.19234168529510498, "learning_rate": 6.3776393042207e-05, "loss": 0.0522, "step": 44060 }, { "epoch": 1.4899083809459412, "grad_norm": 0.23427070677280426, "learning_rate": 6.376049759817137e-05, "loss": 0.0551, "step": 44070 }, { "epoch": 1.4902464586361945, "grad_norm": 0.24344907701015472, "learning_rate": 6.37446006493084e-05, "loss": 0.0497, "step": 44080 }, { "epoch": 1.4905845363264478, "grad_norm": 0.4190671145915985, "learning_rate": 6.372870219735656e-05, "loss": 0.051, "step": 44090 }, { "epoch": 1.4909226140167011, "grad_norm": 0.4737456738948822, "learning_rate": 6.371280224405452e-05, "loss": 0.0528, "step": 44100 }, { "epoch": 1.4912606917069542, "grad_norm": 0.41422146558761597, "learning_rate": 6.369690079114101e-05, "loss": 0.0555, "step": 44110 }, { "epoch": 1.4915987693972075, "grad_norm": 0.20226189494132996, "learning_rate": 6.368099784035504e-05, "loss": 0.0537, "step": 44120 }, { "epoch": 1.4919368470874608, "grad_norm": 0.2911636531352997, "learning_rate": 6.366509339343572e-05, "loss": 0.0556, "step": 44130 }, { "epoch": 1.4922749247777138, "grad_norm": 0.23712384700775146, "learning_rate": 6.364918745212232e-05, "loss": 0.0518, "step": 44140 }, { "epoch": 1.4926130024679671, "grad_norm": 0.3361506462097168, "learning_rate": 6.363328001815431e-05, "loss": 0.0582, "step": 44150 }, { "epoch": 1.4929510801582204, "grad_norm": 0.20324155688285828, "learning_rate": 6.361737109327128e-05, "loss": 0.0568, "step": 44160 }, { "epoch": 1.4932891578484737, "grad_norm": 0.6164899468421936, "learning_rate": 6.360146067921301e-05, "loss": 0.0564, "step": 44170 }, { "epoch": 1.4936272355387268, "grad_norm": 0.20093467831611633, "learning_rate": 6.35855487777195e-05, "loss": 0.0556, "step": 44180 }, { "epoch": 1.49396531322898, "grad_norm": 0.3336327373981476, "learning_rate": 6.356963539053076e-05, "loss": 0.0572, "step": 44190 }, { "epoch": 1.4943033909192334, "grad_norm": 0.2584336996078491, "learning_rate": 6.355372051938711e-05, "loss": 0.0511, "step": 44200 }, { "epoch": 1.4946414686094864, "grad_norm": 0.31963562965393066, "learning_rate": 6.353780416602894e-05, "loss": 0.0544, "step": 44210 }, { "epoch": 1.4949795462997397, "grad_norm": 0.24222704768180847, "learning_rate": 6.352188633219689e-05, "loss": 0.0526, "step": 44220 }, { "epoch": 1.495317623989993, "grad_norm": 0.203379824757576, "learning_rate": 6.350596701963166e-05, "loss": 0.0569, "step": 44230 }, { "epoch": 1.495655701680246, "grad_norm": 0.2928885817527771, "learning_rate": 6.349004623007419e-05, "loss": 0.0529, "step": 44240 }, { "epoch": 1.4959937793704994, "grad_norm": 0.23897434771060944, "learning_rate": 6.347412396526555e-05, "loss": 0.0616, "step": 44250 }, { "epoch": 1.4963318570607527, "grad_norm": 0.1845543086528778, "learning_rate": 6.345820022694696e-05, "loss": 0.0447, "step": 44260 }, { "epoch": 1.4966699347510057, "grad_norm": 0.5821138620376587, "learning_rate": 6.344227501685984e-05, "loss": 0.0605, "step": 44270 }, { "epoch": 1.497008012441259, "grad_norm": 0.35901695489883423, "learning_rate": 6.342634833674572e-05, "loss": 0.0556, "step": 44280 }, { "epoch": 1.4973460901315123, "grad_norm": 0.1981239914894104, "learning_rate": 6.341042018834635e-05, "loss": 0.0473, "step": 44290 }, { "epoch": 1.4976841678217654, "grad_norm": 0.4810437262058258, "learning_rate": 6.339449057340359e-05, "loss": 0.0544, "step": 44300 }, { "epoch": 1.4980222455120187, "grad_norm": 0.2321382462978363, "learning_rate": 6.337855949365945e-05, "loss": 0.0493, "step": 44310 }, { "epoch": 1.498360323202272, "grad_norm": 0.20433825254440308, "learning_rate": 6.336262695085619e-05, "loss": 0.0521, "step": 44320 }, { "epoch": 1.498698400892525, "grad_norm": 0.2626059949398041, "learning_rate": 6.334669294673612e-05, "loss": 0.0515, "step": 44330 }, { "epoch": 1.4990364785827783, "grad_norm": 0.7672393918037415, "learning_rate": 6.333075748304179e-05, "loss": 0.0565, "step": 44340 }, { "epoch": 1.4993745562730316, "grad_norm": 0.6449467539787292, "learning_rate": 6.331482056151585e-05, "loss": 0.0524, "step": 44350 }, { "epoch": 1.4997126339632847, "grad_norm": 0.21062889695167542, "learning_rate": 6.329888218390117e-05, "loss": 0.0552, "step": 44360 }, { "epoch": 1.500050711653538, "grad_norm": 0.23707793653011322, "learning_rate": 6.328294235194072e-05, "loss": 0.0513, "step": 44370 }, { "epoch": 1.5003887893437913, "grad_norm": 0.2566407024860382, "learning_rate": 6.326700106737765e-05, "loss": 0.0477, "step": 44380 }, { "epoch": 1.5007268670340443, "grad_norm": 0.18717548251152039, "learning_rate": 6.325105833195531e-05, "loss": 0.0555, "step": 44390 }, { "epoch": 1.5010649447242976, "grad_norm": 0.42597922682762146, "learning_rate": 6.323511414741715e-05, "loss": 0.0547, "step": 44400 }, { "epoch": 1.501403022414551, "grad_norm": 0.38993948698043823, "learning_rate": 6.321916851550678e-05, "loss": 0.0527, "step": 44410 }, { "epoch": 1.501741100104804, "grad_norm": 0.23852066695690155, "learning_rate": 6.320322143796806e-05, "loss": 0.0538, "step": 44420 }, { "epoch": 1.5020791777950573, "grad_norm": 0.3481338322162628, "learning_rate": 6.318727291654488e-05, "loss": 0.0549, "step": 44430 }, { "epoch": 1.5024172554853106, "grad_norm": 0.2925010621547699, "learning_rate": 6.317132295298134e-05, "loss": 0.0537, "step": 44440 }, { "epoch": 1.5027553331755636, "grad_norm": 0.18835565447807312, "learning_rate": 6.315537154902173e-05, "loss": 0.0487, "step": 44450 }, { "epoch": 1.503093410865817, "grad_norm": 0.48728904128074646, "learning_rate": 6.313941870641048e-05, "loss": 0.0507, "step": 44460 }, { "epoch": 1.5034314885560702, "grad_norm": 0.26497673988342285, "learning_rate": 6.312346442689214e-05, "loss": 0.0524, "step": 44470 }, { "epoch": 1.5037695662463233, "grad_norm": 0.3831683397293091, "learning_rate": 6.310750871221147e-05, "loss": 0.0615, "step": 44480 }, { "epoch": 1.5041076439365766, "grad_norm": 0.3018195629119873, "learning_rate": 6.309155156411335e-05, "loss": 0.0465, "step": 44490 }, { "epoch": 1.5044457216268299, "grad_norm": 0.38626670837402344, "learning_rate": 6.307559298434284e-05, "loss": 0.0535, "step": 44500 }, { "epoch": 1.504783799317083, "grad_norm": 0.3059965670108795, "learning_rate": 6.305963297464512e-05, "loss": 0.0512, "step": 44510 }, { "epoch": 1.5051218770073362, "grad_norm": 0.24986010789871216, "learning_rate": 6.304367153676561e-05, "loss": 0.054, "step": 44520 }, { "epoch": 1.5054599546975895, "grad_norm": 0.2750425338745117, "learning_rate": 6.302770867244978e-05, "loss": 0.0565, "step": 44530 }, { "epoch": 1.5057980323878426, "grad_norm": 0.2869846224784851, "learning_rate": 6.301174438344328e-05, "loss": 0.0531, "step": 44540 }, { "epoch": 1.506136110078096, "grad_norm": 0.2188904732465744, "learning_rate": 6.299577867149202e-05, "loss": 0.0527, "step": 44550 }, { "epoch": 1.5064741877683492, "grad_norm": 0.30422332882881165, "learning_rate": 6.297981153834192e-05, "loss": 0.0529, "step": 44560 }, { "epoch": 1.5068122654586023, "grad_norm": 0.2582066059112549, "learning_rate": 6.296384298573916e-05, "loss": 0.0562, "step": 44570 }, { "epoch": 1.5071503431488558, "grad_norm": 0.288006454706192, "learning_rate": 6.294787301543001e-05, "loss": 0.0564, "step": 44580 }, { "epoch": 1.5074884208391088, "grad_norm": 0.2886621654033661, "learning_rate": 6.293190162916095e-05, "loss": 0.057, "step": 44590 }, { "epoch": 1.507826498529362, "grad_norm": 0.31649836897850037, "learning_rate": 6.291592882867855e-05, "loss": 0.0579, "step": 44600 }, { "epoch": 1.5081645762196154, "grad_norm": 0.26575329899787903, "learning_rate": 6.28999546157296e-05, "loss": 0.0506, "step": 44610 }, { "epoch": 1.5085026539098685, "grad_norm": 0.36359262466430664, "learning_rate": 6.288397899206102e-05, "loss": 0.0489, "step": 44620 }, { "epoch": 1.5088407316001216, "grad_norm": 0.3561828136444092, "learning_rate": 6.286800195941984e-05, "loss": 0.0527, "step": 44630 }, { "epoch": 1.509178809290375, "grad_norm": 0.2176324427127838, "learning_rate": 6.285202351955334e-05, "loss": 0.0485, "step": 44640 }, { "epoch": 1.5095168869806281, "grad_norm": 0.5476887822151184, "learning_rate": 6.283604367420887e-05, "loss": 0.0591, "step": 44650 }, { "epoch": 1.5098549646708812, "grad_norm": 0.2201712429523468, "learning_rate": 6.282006242513394e-05, "loss": 0.0551, "step": 44660 }, { "epoch": 1.5101930423611347, "grad_norm": 0.18022461235523224, "learning_rate": 6.280407977407628e-05, "loss": 0.0502, "step": 44670 }, { "epoch": 1.5105311200513878, "grad_norm": 0.2585912048816681, "learning_rate": 6.27880957227837e-05, "loss": 0.0571, "step": 44680 }, { "epoch": 1.510869197741641, "grad_norm": 0.25630104541778564, "learning_rate": 6.27721102730042e-05, "loss": 0.0553, "step": 44690 }, { "epoch": 1.5112072754318944, "grad_norm": 0.316026508808136, "learning_rate": 6.275612342648592e-05, "loss": 0.058, "step": 44700 }, { "epoch": 1.5115453531221474, "grad_norm": 0.3681083619594574, "learning_rate": 6.274013518497716e-05, "loss": 0.0524, "step": 44710 }, { "epoch": 1.5118834308124007, "grad_norm": 0.3708231747150421, "learning_rate": 6.272414555022636e-05, "loss": 0.055, "step": 44720 }, { "epoch": 1.512221508502654, "grad_norm": 0.34846481680870056, "learning_rate": 6.270815452398215e-05, "loss": 0.051, "step": 44730 }, { "epoch": 1.512559586192907, "grad_norm": 0.32797861099243164, "learning_rate": 6.269216210799326e-05, "loss": 0.0549, "step": 44740 }, { "epoch": 1.5128976638831604, "grad_norm": 0.22355446219444275, "learning_rate": 6.26761683040086e-05, "loss": 0.0521, "step": 44750 }, { "epoch": 1.5132357415734137, "grad_norm": 0.38043442368507385, "learning_rate": 6.266017311377723e-05, "loss": 0.0499, "step": 44760 }, { "epoch": 1.5135738192636667, "grad_norm": 0.3143051266670227, "learning_rate": 6.264417653904839e-05, "loss": 0.0476, "step": 44770 }, { "epoch": 1.51391189695392, "grad_norm": 0.2837350070476532, "learning_rate": 6.262817858157139e-05, "loss": 0.051, "step": 44780 }, { "epoch": 1.5142499746441733, "grad_norm": 0.32731369137763977, "learning_rate": 6.261217924309576e-05, "loss": 0.0524, "step": 44790 }, { "epoch": 1.5145880523344264, "grad_norm": 0.49958229064941406, "learning_rate": 6.259617852537118e-05, "loss": 0.0512, "step": 44800 }, { "epoch": 1.5149261300246797, "grad_norm": 0.2620829939842224, "learning_rate": 6.258017643014747e-05, "loss": 0.0498, "step": 44810 }, { "epoch": 1.515264207714933, "grad_norm": 0.2826054096221924, "learning_rate": 6.256417295917456e-05, "loss": 0.0532, "step": 44820 }, { "epoch": 1.515602285405186, "grad_norm": 0.2077564001083374, "learning_rate": 6.254816811420258e-05, "loss": 0.0465, "step": 44830 }, { "epoch": 1.5159403630954393, "grad_norm": 0.27940651774406433, "learning_rate": 6.253216189698183e-05, "loss": 0.0491, "step": 44840 }, { "epoch": 1.5162784407856926, "grad_norm": 0.2854672074317932, "learning_rate": 6.251615430926267e-05, "loss": 0.0523, "step": 44850 }, { "epoch": 1.5166165184759457, "grad_norm": 0.22673442959785461, "learning_rate": 6.25001453527957e-05, "loss": 0.0579, "step": 44860 }, { "epoch": 1.516954596166199, "grad_norm": 0.26709574460983276, "learning_rate": 6.248413502933164e-05, "loss": 0.0564, "step": 44870 }, { "epoch": 1.5172926738564523, "grad_norm": 0.2650003135204315, "learning_rate": 6.246812334062133e-05, "loss": 0.054, "step": 44880 }, { "epoch": 1.5176307515467053, "grad_norm": 0.2169342041015625, "learning_rate": 6.245211028841579e-05, "loss": 0.0539, "step": 44890 }, { "epoch": 1.5179688292369586, "grad_norm": 0.2666597366333008, "learning_rate": 6.24360958744662e-05, "loss": 0.0539, "step": 44900 }, { "epoch": 1.518306906927212, "grad_norm": 0.3329383432865143, "learning_rate": 6.242008010052387e-05, "loss": 0.0494, "step": 44910 }, { "epoch": 1.518644984617465, "grad_norm": 0.2024952471256256, "learning_rate": 6.240406296834024e-05, "loss": 0.0533, "step": 44920 }, { "epoch": 1.5189830623077183, "grad_norm": 0.24123871326446533, "learning_rate": 6.238804447966694e-05, "loss": 0.0517, "step": 44930 }, { "epoch": 1.5193211399979716, "grad_norm": 0.41348448395729065, "learning_rate": 6.237202463625573e-05, "loss": 0.0499, "step": 44940 }, { "epoch": 1.5196592176882247, "grad_norm": 0.20540885627269745, "learning_rate": 6.235600343985848e-05, "loss": 0.0537, "step": 44950 }, { "epoch": 1.519997295378478, "grad_norm": 0.22143389284610748, "learning_rate": 6.233998089222729e-05, "loss": 0.0548, "step": 44960 }, { "epoch": 1.5203353730687312, "grad_norm": 0.2976893186569214, "learning_rate": 6.232395699511433e-05, "loss": 0.0549, "step": 44970 }, { "epoch": 1.5206734507589843, "grad_norm": 0.28201839327812195, "learning_rate": 6.230793175027199e-05, "loss": 0.0528, "step": 44980 }, { "epoch": 1.5210115284492376, "grad_norm": 0.25823426246643066, "learning_rate": 6.22919051594527e-05, "loss": 0.0483, "step": 44990 }, { "epoch": 1.5213496061394909, "grad_norm": 0.3056069314479828, "learning_rate": 6.227587722440917e-05, "loss": 0.0486, "step": 45000 }, { "epoch": 1.521687683829744, "grad_norm": 0.3459729254245758, "learning_rate": 6.225984794689414e-05, "loss": 0.0532, "step": 45010 }, { "epoch": 1.5220257615199972, "grad_norm": 0.2218107283115387, "learning_rate": 6.22438173286606e-05, "loss": 0.0532, "step": 45020 }, { "epoch": 1.5223638392102505, "grad_norm": 0.30518195033073425, "learning_rate": 6.222778537146159e-05, "loss": 0.0516, "step": 45030 }, { "epoch": 1.5227019169005036, "grad_norm": 0.5024914145469666, "learning_rate": 6.221175207705037e-05, "loss": 0.0558, "step": 45040 }, { "epoch": 1.5230399945907571, "grad_norm": 0.6531257033348083, "learning_rate": 6.219571744718027e-05, "loss": 0.0605, "step": 45050 }, { "epoch": 1.5233780722810102, "grad_norm": 0.49024859070777893, "learning_rate": 6.217968148360487e-05, "loss": 0.0572, "step": 45060 }, { "epoch": 1.5237161499712633, "grad_norm": 0.19555258750915527, "learning_rate": 6.216364418807782e-05, "loss": 0.0526, "step": 45070 }, { "epoch": 1.5240542276615168, "grad_norm": 0.3539108633995056, "learning_rate": 6.214760556235292e-05, "loss": 0.0593, "step": 45080 }, { "epoch": 1.5243923053517698, "grad_norm": 0.2541548013687134, "learning_rate": 6.213156560818414e-05, "loss": 0.0535, "step": 45090 }, { "epoch": 1.524730383042023, "grad_norm": 0.2196616530418396, "learning_rate": 6.211552432732559e-05, "loss": 0.0533, "step": 45100 }, { "epoch": 1.5250684607322764, "grad_norm": 0.23026026785373688, "learning_rate": 6.20994817215315e-05, "loss": 0.0525, "step": 45110 }, { "epoch": 1.5254065384225295, "grad_norm": 0.21334786713123322, "learning_rate": 6.208343779255628e-05, "loss": 0.0535, "step": 45120 }, { "epoch": 1.5257446161127826, "grad_norm": 0.383856862783432, "learning_rate": 6.206739254215449e-05, "loss": 0.0564, "step": 45130 }, { "epoch": 1.526082693803036, "grad_norm": 0.19172902405261993, "learning_rate": 6.205134597208077e-05, "loss": 0.0532, "step": 45140 }, { "epoch": 1.5264207714932891, "grad_norm": 0.23717589676380157, "learning_rate": 6.203529808409e-05, "loss": 0.0468, "step": 45150 }, { "epoch": 1.5267588491835424, "grad_norm": 0.24040164053440094, "learning_rate": 6.201924887993712e-05, "loss": 0.0494, "step": 45160 }, { "epoch": 1.5270969268737957, "grad_norm": 0.38615700602531433, "learning_rate": 6.200319836137725e-05, "loss": 0.0488, "step": 45170 }, { "epoch": 1.5274350045640488, "grad_norm": 0.3126423954963684, "learning_rate": 6.198714653016565e-05, "loss": 0.0541, "step": 45180 }, { "epoch": 1.527773082254302, "grad_norm": 0.21565115451812744, "learning_rate": 6.197109338805774e-05, "loss": 0.0571, "step": 45190 }, { "epoch": 1.5281111599445554, "grad_norm": 0.21967005729675293, "learning_rate": 6.195503893680903e-05, "loss": 0.0504, "step": 45200 }, { "epoch": 1.5284492376348084, "grad_norm": 0.1639147698879242, "learning_rate": 6.193898317817524e-05, "loss": 0.0505, "step": 45210 }, { "epoch": 1.5287873153250617, "grad_norm": 0.16171757876873016, "learning_rate": 6.192292611391221e-05, "loss": 0.0502, "step": 45220 }, { "epoch": 1.529125393015315, "grad_norm": 0.15169969201087952, "learning_rate": 6.190686774577591e-05, "loss": 0.05, "step": 45230 }, { "epoch": 1.529463470705568, "grad_norm": 0.26306018233299255, "learning_rate": 6.189080807552245e-05, "loss": 0.0514, "step": 45240 }, { "epoch": 1.5298015483958214, "grad_norm": 0.286445677280426, "learning_rate": 6.187474710490809e-05, "loss": 0.0503, "step": 45250 }, { "epoch": 1.5301396260860747, "grad_norm": 0.2136252224445343, "learning_rate": 6.185868483568926e-05, "loss": 0.0478, "step": 45260 }, { "epoch": 1.5304777037763277, "grad_norm": 0.44931772351264954, "learning_rate": 6.184262126962245e-05, "loss": 0.0567, "step": 45270 }, { "epoch": 1.530815781466581, "grad_norm": 0.21061192452907562, "learning_rate": 6.182655640846442e-05, "loss": 0.0543, "step": 45280 }, { "epoch": 1.5311538591568343, "grad_norm": 0.18565231561660767, "learning_rate": 6.181049025397196e-05, "loss": 0.0522, "step": 45290 }, { "epoch": 1.5314919368470874, "grad_norm": 0.3038399815559387, "learning_rate": 6.179442280790202e-05, "loss": 0.0485, "step": 45300 }, { "epoch": 1.5318300145373407, "grad_norm": 0.26860126852989197, "learning_rate": 6.177835407201174e-05, "loss": 0.0513, "step": 45310 }, { "epoch": 1.532168092227594, "grad_norm": 0.22165748476982117, "learning_rate": 6.176228404805839e-05, "loss": 0.0546, "step": 45320 }, { "epoch": 1.532506169917847, "grad_norm": 0.2659934163093567, "learning_rate": 6.174621273779932e-05, "loss": 0.0455, "step": 45330 }, { "epoch": 1.5328442476081003, "grad_norm": 0.260530948638916, "learning_rate": 6.17301401429921e-05, "loss": 0.0512, "step": 45340 }, { "epoch": 1.5331823252983536, "grad_norm": 0.26327258348464966, "learning_rate": 6.171406626539436e-05, "loss": 0.0458, "step": 45350 }, { "epoch": 1.5335204029886067, "grad_norm": 0.2160942554473877, "learning_rate": 6.169799110676398e-05, "loss": 0.0518, "step": 45360 }, { "epoch": 1.53385848067886, "grad_norm": 0.257505863904953, "learning_rate": 6.168191466885885e-05, "loss": 0.0504, "step": 45370 }, { "epoch": 1.5341965583691133, "grad_norm": 0.39598339796066284, "learning_rate": 6.166583695343711e-05, "loss": 0.0583, "step": 45380 }, { "epoch": 1.5345346360593664, "grad_norm": 0.27444446086883545, "learning_rate": 6.164975796225698e-05, "loss": 0.0499, "step": 45390 }, { "epoch": 1.5348727137496196, "grad_norm": 0.37085747718811035, "learning_rate": 6.163367769707683e-05, "loss": 0.0503, "step": 45400 }, { "epoch": 1.535210791439873, "grad_norm": 0.3134312331676483, "learning_rate": 6.161759615965519e-05, "loss": 0.0515, "step": 45410 }, { "epoch": 1.535548869130126, "grad_norm": 0.22522595524787903, "learning_rate": 6.16015133517507e-05, "loss": 0.0544, "step": 45420 }, { "epoch": 1.5358869468203793, "grad_norm": 0.3047979176044464, "learning_rate": 6.158542927512214e-05, "loss": 0.0538, "step": 45430 }, { "epoch": 1.5362250245106326, "grad_norm": 0.21836811304092407, "learning_rate": 6.156934393152846e-05, "loss": 0.0511, "step": 45440 }, { "epoch": 1.5365631022008857, "grad_norm": 0.18902109563350677, "learning_rate": 6.15532573227287e-05, "loss": 0.0555, "step": 45450 }, { "epoch": 1.536901179891139, "grad_norm": 0.25116798281669617, "learning_rate": 6.153716945048212e-05, "loss": 0.0564, "step": 45460 }, { "epoch": 1.5372392575813922, "grad_norm": 0.2637465000152588, "learning_rate": 6.152108031654802e-05, "loss": 0.0548, "step": 45470 }, { "epoch": 1.5375773352716453, "grad_norm": 0.20935918390750885, "learning_rate": 6.15049899226859e-05, "loss": 0.0485, "step": 45480 }, { "epoch": 1.5379154129618986, "grad_norm": 0.1683182418346405, "learning_rate": 6.148889827065537e-05, "loss": 0.0521, "step": 45490 }, { "epoch": 1.538253490652152, "grad_norm": 0.18225005269050598, "learning_rate": 6.147280536221622e-05, "loss": 0.0472, "step": 45500 }, { "epoch": 1.538591568342405, "grad_norm": 0.23783434927463531, "learning_rate": 6.145671119912832e-05, "loss": 0.0459, "step": 45510 }, { "epoch": 1.5389296460326582, "grad_norm": 0.19002492725849152, "learning_rate": 6.144061578315169e-05, "loss": 0.0554, "step": 45520 }, { "epoch": 1.5392677237229115, "grad_norm": 0.23711414635181427, "learning_rate": 6.142451911604654e-05, "loss": 0.0535, "step": 45530 }, { "epoch": 1.5396058014131646, "grad_norm": 0.24506635963916779, "learning_rate": 6.140842119957315e-05, "loss": 0.0473, "step": 45540 }, { "epoch": 1.5399438791034181, "grad_norm": 0.26025390625, "learning_rate": 6.139232203549197e-05, "loss": 0.0553, "step": 45550 }, { "epoch": 1.5402819567936712, "grad_norm": 0.17053940892219543, "learning_rate": 6.137622162556357e-05, "loss": 0.05, "step": 45560 }, { "epoch": 1.5406200344839243, "grad_norm": 0.2040601223707199, "learning_rate": 6.13601199715487e-05, "loss": 0.0534, "step": 45570 }, { "epoch": 1.5409581121741778, "grad_norm": 0.25989338755607605, "learning_rate": 6.134401707520816e-05, "loss": 0.0508, "step": 45580 }, { "epoch": 1.5412961898644308, "grad_norm": 0.20081128180027008, "learning_rate": 6.132791293830298e-05, "loss": 0.0559, "step": 45590 }, { "epoch": 1.541634267554684, "grad_norm": 0.21059384942054749, "learning_rate": 6.131180756259428e-05, "loss": 0.052, "step": 45600 }, { "epoch": 1.5419723452449374, "grad_norm": 0.2083752304315567, "learning_rate": 6.129570094984331e-05, "loss": 0.0474, "step": 45610 }, { "epoch": 1.5423104229351905, "grad_norm": 1.02300226688385, "learning_rate": 6.127959310181145e-05, "loss": 0.0443, "step": 45620 }, { "epoch": 1.5426485006254436, "grad_norm": 0.18290379643440247, "learning_rate": 6.126348402026026e-05, "loss": 0.0554, "step": 45630 }, { "epoch": 1.542986578315697, "grad_norm": 0.3408017158508301, "learning_rate": 6.12473737069514e-05, "loss": 0.0513, "step": 45640 }, { "epoch": 1.5433246560059501, "grad_norm": 0.21515385806560516, "learning_rate": 6.123126216364665e-05, "loss": 0.0559, "step": 45650 }, { "epoch": 1.5436627336962034, "grad_norm": 0.43864384293556213, "learning_rate": 6.121514939210797e-05, "loss": 0.053, "step": 45660 }, { "epoch": 1.5440008113864567, "grad_norm": 0.18441468477249146, "learning_rate": 6.119903539409741e-05, "loss": 0.0544, "step": 45670 }, { "epoch": 1.5443388890767098, "grad_norm": 0.18740230798721313, "learning_rate": 6.118292017137716e-05, "loss": 0.0492, "step": 45680 }, { "epoch": 1.544676966766963, "grad_norm": 0.31495195627212524, "learning_rate": 6.116680372570959e-05, "loss": 0.0558, "step": 45690 }, { "epoch": 1.5450150444572164, "grad_norm": 0.20757724344730377, "learning_rate": 6.115068605885713e-05, "loss": 0.053, "step": 45700 }, { "epoch": 1.5453531221474694, "grad_norm": 0.27320975065231323, "learning_rate": 6.113456717258243e-05, "loss": 0.0569, "step": 45710 }, { "epoch": 1.5456911998377227, "grad_norm": 0.21598538756370544, "learning_rate": 6.11184470686482e-05, "loss": 0.0484, "step": 45720 }, { "epoch": 1.546029277527976, "grad_norm": 0.15702040493488312, "learning_rate": 6.110232574881731e-05, "loss": 0.056, "step": 45730 }, { "epoch": 1.546367355218229, "grad_norm": 0.2675706446170807, "learning_rate": 6.108620321485277e-05, "loss": 0.0535, "step": 45740 }, { "epoch": 1.5467054329084824, "grad_norm": 0.23428502678871155, "learning_rate": 6.107007946851773e-05, "loss": 0.0577, "step": 45750 }, { "epoch": 1.5470435105987357, "grad_norm": 0.4890652000904083, "learning_rate": 6.105395451157542e-05, "loss": 0.0527, "step": 45760 }, { "epoch": 1.5473815882889888, "grad_norm": 0.299135684967041, "learning_rate": 6.103782834578928e-05, "loss": 0.0513, "step": 45770 }, { "epoch": 1.547719665979242, "grad_norm": 0.3171231746673584, "learning_rate": 6.102170097292281e-05, "loss": 0.0532, "step": 45780 }, { "epoch": 1.5480577436694953, "grad_norm": 0.35661500692367554, "learning_rate": 6.1005572394739716e-05, "loss": 0.0591, "step": 45790 }, { "epoch": 1.5483958213597484, "grad_norm": 0.1880647987127304, "learning_rate": 6.0989442613003765e-05, "loss": 0.0514, "step": 45800 }, { "epoch": 1.5487338990500017, "grad_norm": 0.31544291973114014, "learning_rate": 6.097331162947888e-05, "loss": 0.0496, "step": 45810 }, { "epoch": 1.549071976740255, "grad_norm": 0.38401827216148376, "learning_rate": 6.095717944592914e-05, "loss": 0.0455, "step": 45820 }, { "epoch": 1.549410054430508, "grad_norm": 0.26056474447250366, "learning_rate": 6.094104606411873e-05, "loss": 0.0467, "step": 45830 }, { "epoch": 1.5497481321207613, "grad_norm": 0.35207399725914, "learning_rate": 6.0924911485811966e-05, "loss": 0.055, "step": 45840 }, { "epoch": 1.5500862098110146, "grad_norm": 0.21197494864463806, "learning_rate": 6.090877571277331e-05, "loss": 0.0514, "step": 45850 }, { "epoch": 1.5504242875012677, "grad_norm": 0.24432918429374695, "learning_rate": 6.089263874676736e-05, "loss": 0.0514, "step": 45860 }, { "epoch": 1.550762365191521, "grad_norm": 0.2125774621963501, "learning_rate": 6.0876500589558796e-05, "loss": 0.0495, "step": 45870 }, { "epoch": 1.5511004428817743, "grad_norm": 0.25757309794425964, "learning_rate": 6.086036124291248e-05, "loss": 0.0551, "step": 45880 }, { "epoch": 1.5514385205720274, "grad_norm": 0.24203675985336304, "learning_rate": 6.084422070859339e-05, "loss": 0.0577, "step": 45890 }, { "epoch": 1.5517765982622806, "grad_norm": 0.2867441177368164, "learning_rate": 6.082807898836663e-05, "loss": 0.0513, "step": 45900 }, { "epoch": 1.552114675952534, "grad_norm": 0.1762501746416092, "learning_rate": 6.081193608399742e-05, "loss": 0.0526, "step": 45910 }, { "epoch": 1.552452753642787, "grad_norm": 0.22835186123847961, "learning_rate": 6.0795791997251164e-05, "loss": 0.0563, "step": 45920 }, { "epoch": 1.5527908313330403, "grad_norm": 0.22181066870689392, "learning_rate": 6.0779646729893294e-05, "loss": 0.049, "step": 45930 }, { "epoch": 1.5531289090232936, "grad_norm": 0.23539280891418457, "learning_rate": 6.0763500283689476e-05, "loss": 0.0525, "step": 45940 }, { "epoch": 1.5534669867135467, "grad_norm": 0.3301730155944824, "learning_rate": 6.0747352660405455e-05, "loss": 0.0573, "step": 45950 }, { "epoch": 1.5538050644038, "grad_norm": 0.23670189082622528, "learning_rate": 6.073120386180709e-05, "loss": 0.0543, "step": 45960 }, { "epoch": 1.5541431420940532, "grad_norm": 0.45076823234558105, "learning_rate": 6.0715053889660425e-05, "loss": 0.051, "step": 45970 }, { "epoch": 1.5544812197843063, "grad_norm": 0.30248838663101196, "learning_rate": 6.069890274573157e-05, "loss": 0.0525, "step": 45980 }, { "epoch": 1.5548192974745596, "grad_norm": 0.3943815529346466, "learning_rate": 6.068275043178679e-05, "loss": 0.0474, "step": 45990 }, { "epoch": 1.555157375164813, "grad_norm": 0.4301259517669678, "learning_rate": 6.066659694959248e-05, "loss": 0.0494, "step": 46000 }, { "epoch": 1.555495452855066, "grad_norm": 0.23054932057857513, "learning_rate": 6.0650442300915176e-05, "loss": 0.0484, "step": 46010 }, { "epoch": 1.5558335305453195, "grad_norm": 0.2455025166273117, "learning_rate": 6.063428648752152e-05, "loss": 0.0543, "step": 46020 }, { "epoch": 1.5561716082355725, "grad_norm": 0.3315044939517975, "learning_rate": 6.0618129511178266e-05, "loss": 0.0496, "step": 46030 }, { "epoch": 1.5565096859258256, "grad_norm": 0.17694848775863647, "learning_rate": 6.060197137365234e-05, "loss": 0.0546, "step": 46040 }, { "epoch": 1.5568477636160791, "grad_norm": 0.23771755397319794, "learning_rate": 6.058581207671077e-05, "loss": 0.0578, "step": 46050 }, { "epoch": 1.5571858413063322, "grad_norm": 0.19637568295001984, "learning_rate": 6.056965162212072e-05, "loss": 0.0529, "step": 46060 }, { "epoch": 1.5575239189965853, "grad_norm": 0.3219752609729767, "learning_rate": 6.055349001164943e-05, "loss": 0.0521, "step": 46070 }, { "epoch": 1.5578619966868388, "grad_norm": 0.19366031885147095, "learning_rate": 6.0537327247064347e-05, "loss": 0.0545, "step": 46080 }, { "epoch": 1.5582000743770918, "grad_norm": 0.30927708745002747, "learning_rate": 6.0521163330133014e-05, "loss": 0.0469, "step": 46090 }, { "epoch": 1.558538152067345, "grad_norm": 0.2266431599855423, "learning_rate": 6.050499826262306e-05, "loss": 0.0551, "step": 46100 }, { "epoch": 1.5588762297575984, "grad_norm": 0.1705072671175003, "learning_rate": 6.0488832046302294e-05, "loss": 0.0519, "step": 46110 }, { "epoch": 1.5592143074478515, "grad_norm": 0.24936749041080475, "learning_rate": 6.0472664682938626e-05, "loss": 0.0522, "step": 46120 }, { "epoch": 1.5595523851381048, "grad_norm": 0.3352516293525696, "learning_rate": 6.045649617430009e-05, "loss": 0.0533, "step": 46130 }, { "epoch": 1.559890462828358, "grad_norm": 0.2512085437774658, "learning_rate": 6.0440326522154866e-05, "loss": 0.0534, "step": 46140 }, { "epoch": 1.5602285405186112, "grad_norm": 0.20229659974575043, "learning_rate": 6.0424155728271224e-05, "loss": 0.0475, "step": 46150 }, { "epoch": 1.5605666182088644, "grad_norm": 0.23526576161384583, "learning_rate": 6.040798379441758e-05, "loss": 0.0556, "step": 46160 }, { "epoch": 1.5609046958991177, "grad_norm": 0.41924622654914856, "learning_rate": 6.0391810722362485e-05, "loss": 0.0525, "step": 46170 }, { "epoch": 1.5612427735893708, "grad_norm": 0.23300538957118988, "learning_rate": 6.037563651387458e-05, "loss": 0.0535, "step": 46180 }, { "epoch": 1.561580851279624, "grad_norm": 0.22613050043582916, "learning_rate": 6.0359461170722666e-05, "loss": 0.0565, "step": 46190 }, { "epoch": 1.5619189289698774, "grad_norm": 0.21054710447788239, "learning_rate": 6.034328469467566e-05, "loss": 0.0481, "step": 46200 }, { "epoch": 1.5622570066601305, "grad_norm": 0.17681871354579926, "learning_rate": 6.0327107087502596e-05, "loss": 0.0533, "step": 46210 }, { "epoch": 1.5625950843503837, "grad_norm": 0.16309548914432526, "learning_rate": 6.031092835097262e-05, "loss": 0.0474, "step": 46220 }, { "epoch": 1.562933162040637, "grad_norm": 0.1903093159198761, "learning_rate": 6.0294748486855024e-05, "loss": 0.0476, "step": 46230 }, { "epoch": 1.56327123973089, "grad_norm": 0.1964290589094162, "learning_rate": 6.0278567496919216e-05, "loss": 0.0572, "step": 46240 }, { "epoch": 1.5636093174211434, "grad_norm": 0.6502242684364319, "learning_rate": 6.026238538293472e-05, "loss": 0.0505, "step": 46250 }, { "epoch": 1.5639473951113967, "grad_norm": 0.21349544823169708, "learning_rate": 6.024620214667118e-05, "loss": 0.0516, "step": 46260 }, { "epoch": 1.5642854728016498, "grad_norm": 0.2527642250061035, "learning_rate": 6.0230017789898384e-05, "loss": 0.0499, "step": 46270 }, { "epoch": 1.564623550491903, "grad_norm": 0.23720483481884003, "learning_rate": 6.021383231438622e-05, "loss": 0.0497, "step": 46280 }, { "epoch": 1.5649616281821563, "grad_norm": 0.21727056801319122, "learning_rate": 6.0197645721904704e-05, "loss": 0.0551, "step": 46290 }, { "epoch": 1.5652997058724094, "grad_norm": 0.22481174767017365, "learning_rate": 6.0181458014224e-05, "loss": 0.053, "step": 46300 }, { "epoch": 1.5656377835626627, "grad_norm": 0.20443512499332428, "learning_rate": 6.016526919311435e-05, "loss": 0.0498, "step": 46310 }, { "epoch": 1.565975861252916, "grad_norm": 0.30880680680274963, "learning_rate": 6.014907926034613e-05, "loss": 0.0501, "step": 46320 }, { "epoch": 1.566313938943169, "grad_norm": 0.31659379601478577, "learning_rate": 6.0132888217689875e-05, "loss": 0.0536, "step": 46330 }, { "epoch": 1.5666520166334224, "grad_norm": 0.22040599584579468, "learning_rate": 6.01166960669162e-05, "loss": 0.0502, "step": 46340 }, { "epoch": 1.5669900943236756, "grad_norm": 0.31953999400138855, "learning_rate": 6.0100502809795845e-05, "loss": 0.0505, "step": 46350 }, { "epoch": 1.5673281720139287, "grad_norm": 0.18220750987529755, "learning_rate": 6.00843084480997e-05, "loss": 0.0529, "step": 46360 }, { "epoch": 1.567666249704182, "grad_norm": 0.3188861906528473, "learning_rate": 6.0068112983598736e-05, "loss": 0.051, "step": 46370 }, { "epoch": 1.5680043273944353, "grad_norm": 0.2110041379928589, "learning_rate": 6.005191641806407e-05, "loss": 0.0518, "step": 46380 }, { "epoch": 1.5683424050846884, "grad_norm": 0.24205748736858368, "learning_rate": 6.003571875326694e-05, "loss": 0.0556, "step": 46390 }, { "epoch": 1.5686804827749417, "grad_norm": 0.40398192405700684, "learning_rate": 6.00195199909787e-05, "loss": 0.0508, "step": 46400 }, { "epoch": 1.569018560465195, "grad_norm": 0.1753416210412979, "learning_rate": 6.000332013297082e-05, "loss": 0.0522, "step": 46410 }, { "epoch": 1.569356638155448, "grad_norm": 0.2566024363040924, "learning_rate": 5.998711918101487e-05, "loss": 0.0557, "step": 46420 }, { "epoch": 1.5696947158457013, "grad_norm": 0.2318906933069229, "learning_rate": 5.997091713688261e-05, "loss": 0.0506, "step": 46430 }, { "epoch": 1.5700327935359546, "grad_norm": 0.2979162931442261, "learning_rate": 5.9954714002345836e-05, "loss": 0.0546, "step": 46440 }, { "epoch": 1.5703708712262077, "grad_norm": 0.245913565158844, "learning_rate": 5.993850977917649e-05, "loss": 0.0501, "step": 46450 }, { "epoch": 1.570708948916461, "grad_norm": 0.19601793587207794, "learning_rate": 5.992230446914667e-05, "loss": 0.0525, "step": 46460 }, { "epoch": 1.5710470266067142, "grad_norm": 0.3240184485912323, "learning_rate": 5.990609807402855e-05, "loss": 0.0528, "step": 46470 }, { "epoch": 1.5713851042969673, "grad_norm": 0.4201679527759552, "learning_rate": 5.988989059559443e-05, "loss": 0.0536, "step": 46480 }, { "epoch": 1.5717231819872208, "grad_norm": 0.26811647415161133, "learning_rate": 5.987368203561675e-05, "loss": 0.0565, "step": 46490 }, { "epoch": 1.572061259677474, "grad_norm": 0.9190875887870789, "learning_rate": 5.9857472395868055e-05, "loss": 0.0501, "step": 46500 }, { "epoch": 1.572399337367727, "grad_norm": 0.351563423871994, "learning_rate": 5.9841261678120983e-05, "loss": 0.0579, "step": 46510 }, { "epoch": 1.5727374150579805, "grad_norm": 0.3017578423023224, "learning_rate": 5.982504988414834e-05, "loss": 0.0505, "step": 46520 }, { "epoch": 1.5730754927482336, "grad_norm": 0.23190630972385406, "learning_rate": 5.9808837015723015e-05, "loss": 0.0505, "step": 46530 }, { "epoch": 1.5734135704384866, "grad_norm": 0.20491287112236023, "learning_rate": 5.9792623074618016e-05, "loss": 0.0546, "step": 46540 }, { "epoch": 1.5737516481287401, "grad_norm": 1.091196060180664, "learning_rate": 5.977640806260648e-05, "loss": 0.0566, "step": 46550 }, { "epoch": 1.5740897258189932, "grad_norm": 0.2096744179725647, "learning_rate": 5.976019198146166e-05, "loss": 0.0427, "step": 46560 }, { "epoch": 1.5744278035092463, "grad_norm": 0.3778506815433502, "learning_rate": 5.974397483295692e-05, "loss": 0.0497, "step": 46570 }, { "epoch": 1.5747658811994998, "grad_norm": 0.32292112708091736, "learning_rate": 5.972775661886575e-05, "loss": 0.0507, "step": 46580 }, { "epoch": 1.5751039588897529, "grad_norm": 0.34360799193382263, "learning_rate": 5.971153734096173e-05, "loss": 0.0559, "step": 46590 }, { "epoch": 1.575442036580006, "grad_norm": 0.2151806652545929, "learning_rate": 5.969531700101859e-05, "loss": 0.0501, "step": 46600 }, { "epoch": 1.5757801142702594, "grad_norm": 0.21280911564826965, "learning_rate": 5.9679095600810155e-05, "loss": 0.0485, "step": 46610 }, { "epoch": 1.5761181919605125, "grad_norm": 0.18500381708145142, "learning_rate": 5.9662873142110384e-05, "loss": 0.0542, "step": 46620 }, { "epoch": 1.5764562696507658, "grad_norm": 0.3415907621383667, "learning_rate": 5.964664962669333e-05, "loss": 0.0511, "step": 46630 }, { "epoch": 1.576794347341019, "grad_norm": 0.387545108795166, "learning_rate": 5.9630425056333186e-05, "loss": 0.0557, "step": 46640 }, { "epoch": 1.5771324250312722, "grad_norm": 0.32680872082710266, "learning_rate": 5.961419943280422e-05, "loss": 0.0492, "step": 46650 }, { "epoch": 1.5774705027215254, "grad_norm": 0.23707225918769836, "learning_rate": 5.959797275788087e-05, "loss": 0.052, "step": 46660 }, { "epoch": 1.5778085804117787, "grad_norm": 0.33616453409194946, "learning_rate": 5.958174503333765e-05, "loss": 0.0511, "step": 46670 }, { "epoch": 1.5781466581020318, "grad_norm": 0.24937768280506134, "learning_rate": 5.9565516260949195e-05, "loss": 0.0558, "step": 46680 }, { "epoch": 1.578484735792285, "grad_norm": 0.4163190424442291, "learning_rate": 5.9549286442490273e-05, "loss": 0.0597, "step": 46690 }, { "epoch": 1.5788228134825384, "grad_norm": 0.3241402804851532, "learning_rate": 5.953305557973572e-05, "loss": 0.0503, "step": 46700 }, { "epoch": 1.5791608911727915, "grad_norm": 0.40172842144966125, "learning_rate": 5.9516823674460564e-05, "loss": 0.057, "step": 46710 }, { "epoch": 1.5794989688630447, "grad_norm": 0.26638245582580566, "learning_rate": 5.9500590728439875e-05, "loss": 0.0502, "step": 46720 }, { "epoch": 1.579837046553298, "grad_norm": 0.19629697501659393, "learning_rate": 5.948435674344886e-05, "loss": 0.0561, "step": 46730 }, { "epoch": 1.580175124243551, "grad_norm": 0.281006783246994, "learning_rate": 5.946812172126285e-05, "loss": 0.0548, "step": 46740 }, { "epoch": 1.5805132019338044, "grad_norm": 0.23296305537223816, "learning_rate": 5.945188566365729e-05, "loss": 0.0482, "step": 46750 }, { "epoch": 1.5808512796240577, "grad_norm": 0.2692458927631378, "learning_rate": 5.943564857240773e-05, "loss": 0.0511, "step": 46760 }, { "epoch": 1.5811893573143108, "grad_norm": 0.21256966888904572, "learning_rate": 5.941941044928981e-05, "loss": 0.055, "step": 46770 }, { "epoch": 1.581527435004564, "grad_norm": 0.33701345324516296, "learning_rate": 5.940317129607935e-05, "loss": 0.0527, "step": 46780 }, { "epoch": 1.5818655126948173, "grad_norm": 0.8101651668548584, "learning_rate": 5.9386931114552204e-05, "loss": 0.0514, "step": 46790 }, { "epoch": 1.5822035903850704, "grad_norm": 0.33625528216362, "learning_rate": 5.937068990648438e-05, "loss": 0.0553, "step": 46800 }, { "epoch": 1.5825416680753237, "grad_norm": 0.20018410682678223, "learning_rate": 5.935444767365199e-05, "loss": 0.0484, "step": 46810 }, { "epoch": 1.582879745765577, "grad_norm": 0.20765890181064606, "learning_rate": 5.933820441783129e-05, "loss": 0.0536, "step": 46820 }, { "epoch": 1.58321782345583, "grad_norm": 0.25124993920326233, "learning_rate": 5.932196014079857e-05, "loss": 0.0505, "step": 46830 }, { "epoch": 1.5835559011460834, "grad_norm": 0.2101142555475235, "learning_rate": 5.930571484433032e-05, "loss": 0.0474, "step": 46840 }, { "epoch": 1.5838939788363366, "grad_norm": 0.4003397524356842, "learning_rate": 5.928946853020309e-05, "loss": 0.0549, "step": 46850 }, { "epoch": 1.5842320565265897, "grad_norm": 0.2639347314834595, "learning_rate": 5.927322120019352e-05, "loss": 0.051, "step": 46860 }, { "epoch": 1.584570134216843, "grad_norm": 0.21418413519859314, "learning_rate": 5.925697285607844e-05, "loss": 0.051, "step": 46870 }, { "epoch": 1.5849082119070963, "grad_norm": 0.23492205142974854, "learning_rate": 5.924072349963472e-05, "loss": 0.0528, "step": 46880 }, { "epoch": 1.5852462895973494, "grad_norm": 0.2822231352329254, "learning_rate": 5.922447313263938e-05, "loss": 0.0493, "step": 46890 }, { "epoch": 1.5855843672876027, "grad_norm": 0.41723841428756714, "learning_rate": 5.920822175686952e-05, "loss": 0.0579, "step": 46900 }, { "epoch": 1.585922444977856, "grad_norm": 0.18280208110809326, "learning_rate": 5.919196937410239e-05, "loss": 0.05, "step": 46910 }, { "epoch": 1.586260522668109, "grad_norm": 0.22029165923595428, "learning_rate": 5.917571598611529e-05, "loss": 0.0512, "step": 46920 }, { "epoch": 1.5865986003583623, "grad_norm": 0.2570682168006897, "learning_rate": 5.9159461594685686e-05, "loss": 0.0474, "step": 46930 }, { "epoch": 1.5869366780486156, "grad_norm": 0.25163891911506653, "learning_rate": 5.914320620159114e-05, "loss": 0.0524, "step": 46940 }, { "epoch": 1.5872747557388687, "grad_norm": 0.33901169896125793, "learning_rate": 5.912694980860931e-05, "loss": 0.0487, "step": 46950 }, { "epoch": 1.587612833429122, "grad_norm": 0.23840028047561646, "learning_rate": 5.9110692417517964e-05, "loss": 0.0495, "step": 46960 }, { "epoch": 1.5879509111193753, "grad_norm": 0.32549044489860535, "learning_rate": 5.9094434030095e-05, "loss": 0.0479, "step": 46970 }, { "epoch": 1.5882889888096283, "grad_norm": 0.3029212951660156, "learning_rate": 5.907817464811841e-05, "loss": 0.0529, "step": 46980 }, { "epoch": 1.5886270664998818, "grad_norm": 0.31520262360572815, "learning_rate": 5.906191427336627e-05, "loss": 0.0523, "step": 46990 }, { "epoch": 1.588965144190135, "grad_norm": 0.20287668704986572, "learning_rate": 5.9045652907616835e-05, "loss": 0.0554, "step": 47000 }, { "epoch": 1.589303221880388, "grad_norm": 0.31677570939064026, "learning_rate": 5.902939055264838e-05, "loss": 0.056, "step": 47010 }, { "epoch": 1.5896412995706415, "grad_norm": 0.3171969950199127, "learning_rate": 5.901312721023935e-05, "loss": 0.0552, "step": 47020 }, { "epoch": 1.5899793772608946, "grad_norm": 0.2761155664920807, "learning_rate": 5.8996862882168294e-05, "loss": 0.0517, "step": 47030 }, { "epoch": 1.5903174549511476, "grad_norm": 0.24110279977321625, "learning_rate": 5.8980597570213826e-05, "loss": 0.0509, "step": 47040 }, { "epoch": 1.5906555326414011, "grad_norm": 0.1987263560295105, "learning_rate": 5.896433127615471e-05, "loss": 0.0558, "step": 47050 }, { "epoch": 1.5909936103316542, "grad_norm": 0.3392605185508728, "learning_rate": 5.894806400176981e-05, "loss": 0.0512, "step": 47060 }, { "epoch": 1.5913316880219073, "grad_norm": 0.19057826697826385, "learning_rate": 5.893179574883808e-05, "loss": 0.0528, "step": 47070 }, { "epoch": 1.5916697657121608, "grad_norm": 0.22779174149036407, "learning_rate": 5.8915526519138585e-05, "loss": 0.0556, "step": 47080 }, { "epoch": 1.5920078434024139, "grad_norm": 0.3910217583179474, "learning_rate": 5.889925631445053e-05, "loss": 0.0509, "step": 47090 }, { "epoch": 1.5923459210926671, "grad_norm": 0.21833567321300507, "learning_rate": 5.888298513655318e-05, "loss": 0.0509, "step": 47100 }, { "epoch": 1.5926839987829204, "grad_norm": 0.3685681223869324, "learning_rate": 5.886671298722592e-05, "loss": 0.0507, "step": 47110 }, { "epoch": 1.5930220764731735, "grad_norm": 0.17622673511505127, "learning_rate": 5.885043986824828e-05, "loss": 0.0499, "step": 47120 }, { "epoch": 1.5933601541634268, "grad_norm": 0.17020083963871002, "learning_rate": 5.883416578139982e-05, "loss": 0.0512, "step": 47130 }, { "epoch": 1.59369823185368, "grad_norm": 0.7295982837677002, "learning_rate": 5.881789072846029e-05, "loss": 0.0515, "step": 47140 }, { "epoch": 1.5940363095439332, "grad_norm": 0.19519415497779846, "learning_rate": 5.8801614711209474e-05, "loss": 0.0535, "step": 47150 }, { "epoch": 1.5943743872341865, "grad_norm": 0.26166170835494995, "learning_rate": 5.878533773142732e-05, "loss": 0.0537, "step": 47160 }, { "epoch": 1.5947124649244397, "grad_norm": 0.2583980858325958, "learning_rate": 5.8769059790893834e-05, "loss": 0.0558, "step": 47170 }, { "epoch": 1.5950505426146928, "grad_norm": 0.49483728408813477, "learning_rate": 5.8752780891389145e-05, "loss": 0.0568, "step": 47180 }, { "epoch": 1.595388620304946, "grad_norm": 0.2735242247581482, "learning_rate": 5.873650103469351e-05, "loss": 0.0516, "step": 47190 }, { "epoch": 1.5957266979951994, "grad_norm": 0.28214359283447266, "learning_rate": 5.872022022258725e-05, "loss": 0.0505, "step": 47200 }, { "epoch": 1.5960647756854525, "grad_norm": 0.3161773979663849, "learning_rate": 5.8703938456850814e-05, "loss": 0.0492, "step": 47210 }, { "epoch": 1.5964028533757058, "grad_norm": 0.2411506623029709, "learning_rate": 5.868765573926477e-05, "loss": 0.0519, "step": 47220 }, { "epoch": 1.596740931065959, "grad_norm": 0.29831618070602417, "learning_rate": 5.867137207160974e-05, "loss": 0.0514, "step": 47230 }, { "epoch": 1.5970790087562121, "grad_norm": 0.18128323554992676, "learning_rate": 5.86550874556665e-05, "loss": 0.0571, "step": 47240 }, { "epoch": 1.5974170864464654, "grad_norm": 0.3033638894557953, "learning_rate": 5.863880189321592e-05, "loss": 0.0524, "step": 47250 }, { "epoch": 1.5977551641367187, "grad_norm": 0.3955126702785492, "learning_rate": 5.862251538603894e-05, "loss": 0.0534, "step": 47260 }, { "epoch": 1.5980932418269718, "grad_norm": 0.17707625031471252, "learning_rate": 5.8606227935916656e-05, "loss": 0.0483, "step": 47270 }, { "epoch": 1.598431319517225, "grad_norm": 0.2788477838039398, "learning_rate": 5.858993954463021e-05, "loss": 0.0539, "step": 47280 }, { "epoch": 1.5987693972074783, "grad_norm": 0.2618483006954193, "learning_rate": 5.8573650213960896e-05, "loss": 0.0496, "step": 47290 }, { "epoch": 1.5991074748977314, "grad_norm": 0.30303874611854553, "learning_rate": 5.8557359945690084e-05, "loss": 0.0514, "step": 47300 }, { "epoch": 1.5994455525879847, "grad_norm": 0.19989483058452606, "learning_rate": 5.8541068741599246e-05, "loss": 0.0511, "step": 47310 }, { "epoch": 1.599783630278238, "grad_norm": 0.2358366996049881, "learning_rate": 5.8524776603469985e-05, "loss": 0.0531, "step": 47320 }, { "epoch": 1.600121707968491, "grad_norm": 0.3018917143344879, "learning_rate": 5.850848353308397e-05, "loss": 0.0521, "step": 47330 }, { "epoch": 1.6004597856587444, "grad_norm": 0.19989606738090515, "learning_rate": 5.849218953222297e-05, "loss": 0.0471, "step": 47340 }, { "epoch": 1.6007978633489977, "grad_norm": 0.22434908151626587, "learning_rate": 5.847589460266891e-05, "loss": 0.0505, "step": 47350 }, { "epoch": 1.6011359410392507, "grad_norm": 0.2277209460735321, "learning_rate": 5.845959874620375e-05, "loss": 0.0541, "step": 47360 }, { "epoch": 1.601474018729504, "grad_norm": 0.22832533717155457, "learning_rate": 5.84433019646096e-05, "loss": 0.0506, "step": 47370 }, { "epoch": 1.6018120964197573, "grad_norm": 0.15719006955623627, "learning_rate": 5.842700425966863e-05, "loss": 0.0497, "step": 47380 }, { "epoch": 1.6021501741100104, "grad_norm": 0.265609472990036, "learning_rate": 5.841070563316315e-05, "loss": 0.054, "step": 47390 }, { "epoch": 1.6024882518002637, "grad_norm": 0.2039768397808075, "learning_rate": 5.839440608687554e-05, "loss": 0.0552, "step": 47400 }, { "epoch": 1.602826329490517, "grad_norm": 0.2072603851556778, "learning_rate": 5.837810562258831e-05, "loss": 0.0501, "step": 47410 }, { "epoch": 1.60316440718077, "grad_norm": 0.13862371444702148, "learning_rate": 5.836180424208405e-05, "loss": 0.0452, "step": 47420 }, { "epoch": 1.6035024848710233, "grad_norm": 0.37092992663383484, "learning_rate": 5.834550194714543e-05, "loss": 0.0524, "step": 47430 }, { "epoch": 1.6038405625612766, "grad_norm": 0.21552450954914093, "learning_rate": 5.832919873955528e-05, "loss": 0.0522, "step": 47440 }, { "epoch": 1.6041786402515297, "grad_norm": 0.7593179941177368, "learning_rate": 5.831289462109648e-05, "loss": 0.0569, "step": 47450 }, { "epoch": 1.6045167179417832, "grad_norm": 0.23183275759220123, "learning_rate": 5.829658959355202e-05, "loss": 0.0505, "step": 47460 }, { "epoch": 1.6048547956320363, "grad_norm": 0.1911809742450714, "learning_rate": 5.8280283658704995e-05, "loss": 0.0548, "step": 47470 }, { "epoch": 1.6051928733222893, "grad_norm": 0.39346835017204285, "learning_rate": 5.826397681833859e-05, "loss": 0.0518, "step": 47480 }, { "epoch": 1.6055309510125428, "grad_norm": 0.22059348225593567, "learning_rate": 5.824766907423612e-05, "loss": 0.0521, "step": 47490 }, { "epoch": 1.605869028702796, "grad_norm": 0.2853274941444397, "learning_rate": 5.8231360428180945e-05, "loss": 0.0489, "step": 47500 }, { "epoch": 1.606207106393049, "grad_norm": 0.2545781433582306, "learning_rate": 5.821505088195658e-05, "loss": 0.051, "step": 47510 }, { "epoch": 1.6065451840833025, "grad_norm": 0.287570983171463, "learning_rate": 5.819874043734661e-05, "loss": 0.0518, "step": 47520 }, { "epoch": 1.6068832617735556, "grad_norm": 0.2732270658016205, "learning_rate": 5.8182429096134695e-05, "loss": 0.0554, "step": 47530 }, { "epoch": 1.6072213394638086, "grad_norm": 0.5261900424957275, "learning_rate": 5.816611686010465e-05, "loss": 0.0578, "step": 47540 }, { "epoch": 1.6075594171540621, "grad_norm": 0.22510574758052826, "learning_rate": 5.814980373104033e-05, "loss": 0.0491, "step": 47550 }, { "epoch": 1.6078974948443152, "grad_norm": 0.34596744179725647, "learning_rate": 5.813348971072572e-05, "loss": 0.0546, "step": 47560 }, { "epoch": 1.6082355725345683, "grad_norm": 0.34633591771125793, "learning_rate": 5.811717480094492e-05, "loss": 0.0521, "step": 47570 }, { "epoch": 1.6085736502248218, "grad_norm": 0.1827404797077179, "learning_rate": 5.810085900348209e-05, "loss": 0.0542, "step": 47580 }, { "epoch": 1.6089117279150749, "grad_norm": 0.26060670614242554, "learning_rate": 5.8084542320121483e-05, "loss": 0.0533, "step": 47590 }, { "epoch": 1.6092498056053282, "grad_norm": 0.18783089518547058, "learning_rate": 5.8068224752647497e-05, "loss": 0.0489, "step": 47600 }, { "epoch": 1.6095878832955814, "grad_norm": 0.2056625634431839, "learning_rate": 5.805190630284457e-05, "loss": 0.054, "step": 47610 }, { "epoch": 1.6099259609858345, "grad_norm": 0.27383142709732056, "learning_rate": 5.803558697249729e-05, "loss": 0.0525, "step": 47620 }, { "epoch": 1.6102640386760878, "grad_norm": 0.18563631176948547, "learning_rate": 5.8019266763390276e-05, "loss": 0.0512, "step": 47630 }, { "epoch": 1.610602116366341, "grad_norm": 0.34220749139785767, "learning_rate": 5.800294567730833e-05, "loss": 0.0471, "step": 47640 }, { "epoch": 1.6109401940565942, "grad_norm": 0.24980153143405914, "learning_rate": 5.798662371603628e-05, "loss": 0.0487, "step": 47650 }, { "epoch": 1.6112782717468475, "grad_norm": 0.2329213172197342, "learning_rate": 5.797030088135904e-05, "loss": 0.0481, "step": 47660 }, { "epoch": 1.6116163494371007, "grad_norm": 0.2105407416820526, "learning_rate": 5.7953977175061705e-05, "loss": 0.0584, "step": 47670 }, { "epoch": 1.6119544271273538, "grad_norm": 0.21098630130290985, "learning_rate": 5.793765259892938e-05, "loss": 0.0531, "step": 47680 }, { "epoch": 1.612292504817607, "grad_norm": 0.2862888276576996, "learning_rate": 5.792132715474729e-05, "loss": 0.0505, "step": 47690 }, { "epoch": 1.6126305825078604, "grad_norm": 0.33634647727012634, "learning_rate": 5.790500084430078e-05, "loss": 0.0539, "step": 47700 }, { "epoch": 1.6129686601981135, "grad_norm": 0.18454647064208984, "learning_rate": 5.7888673669375274e-05, "loss": 0.058, "step": 47710 }, { "epoch": 1.6133067378883668, "grad_norm": 0.22048187255859375, "learning_rate": 5.787234563175625e-05, "loss": 0.0561, "step": 47720 }, { "epoch": 1.61364481557862, "grad_norm": 0.18953590095043182, "learning_rate": 5.7856016733229355e-05, "loss": 0.0559, "step": 47730 }, { "epoch": 1.6139828932688731, "grad_norm": 0.19651702046394348, "learning_rate": 5.7839686975580297e-05, "loss": 0.0563, "step": 47740 }, { "epoch": 1.6143209709591264, "grad_norm": 0.17629612982273102, "learning_rate": 5.782335636059484e-05, "loss": 0.0489, "step": 47750 }, { "epoch": 1.6146590486493797, "grad_norm": 0.2557413876056671, "learning_rate": 5.780702489005889e-05, "loss": 0.05, "step": 47760 }, { "epoch": 1.6149971263396328, "grad_norm": 0.2745135426521301, "learning_rate": 5.779069256575845e-05, "loss": 0.0509, "step": 47770 }, { "epoch": 1.615335204029886, "grad_norm": 0.22641214728355408, "learning_rate": 5.7774359389479574e-05, "loss": 0.0487, "step": 47780 }, { "epoch": 1.6156732817201394, "grad_norm": 0.22095650434494019, "learning_rate": 5.775802536300845e-05, "loss": 0.0458, "step": 47790 }, { "epoch": 1.6160113594103924, "grad_norm": 0.23742257058620453, "learning_rate": 5.774169048813134e-05, "loss": 0.0475, "step": 47800 }, { "epoch": 1.6163494371006457, "grad_norm": 0.29771167039871216, "learning_rate": 5.77253547666346e-05, "loss": 0.0517, "step": 47810 }, { "epoch": 1.616687514790899, "grad_norm": 0.42170339822769165, "learning_rate": 5.770901820030465e-05, "loss": 0.051, "step": 47820 }, { "epoch": 1.617025592481152, "grad_norm": 0.237478107213974, "learning_rate": 5.769268079092809e-05, "loss": 0.0503, "step": 47830 }, { "epoch": 1.6173636701714054, "grad_norm": 0.20211409032344818, "learning_rate": 5.767634254029151e-05, "loss": 0.0517, "step": 47840 }, { "epoch": 1.6177017478616587, "grad_norm": 0.2136366069316864, "learning_rate": 5.7660003450181655e-05, "loss": 0.041, "step": 47850 }, { "epoch": 1.6180398255519117, "grad_norm": 0.3343557119369507, "learning_rate": 5.764366352238534e-05, "loss": 0.0515, "step": 47860 }, { "epoch": 1.618377903242165, "grad_norm": 0.2991087734699249, "learning_rate": 5.7627322758689474e-05, "loss": 0.0521, "step": 47870 }, { "epoch": 1.6187159809324183, "grad_norm": 0.2572799026966095, "learning_rate": 5.761098116088105e-05, "loss": 0.0485, "step": 47880 }, { "epoch": 1.6190540586226714, "grad_norm": 0.1976356953382492, "learning_rate": 5.759463873074717e-05, "loss": 0.0538, "step": 47890 }, { "epoch": 1.6193921363129247, "grad_norm": 0.2703512907028198, "learning_rate": 5.757829547007504e-05, "loss": 0.0505, "step": 47900 }, { "epoch": 1.619730214003178, "grad_norm": 0.2933197319507599, "learning_rate": 5.756195138065189e-05, "loss": 0.0515, "step": 47910 }, { "epoch": 1.620068291693431, "grad_norm": 0.2124345302581787, "learning_rate": 5.754560646426511e-05, "loss": 0.0482, "step": 47920 }, { "epoch": 1.6204063693836843, "grad_norm": 0.2259070873260498, "learning_rate": 5.752926072270216e-05, "loss": 0.0534, "step": 47930 }, { "epoch": 1.6207444470739376, "grad_norm": 0.21418921649456024, "learning_rate": 5.7512914157750563e-05, "loss": 0.0517, "step": 47940 }, { "epoch": 1.6210825247641907, "grad_norm": 0.22709190845489502, "learning_rate": 5.749656677119798e-05, "loss": 0.0539, "step": 47950 }, { "epoch": 1.6214206024544442, "grad_norm": 0.23907485604286194, "learning_rate": 5.7480218564832125e-05, "loss": 0.0469, "step": 47960 }, { "epoch": 1.6217586801446973, "grad_norm": 0.16612829267978668, "learning_rate": 5.746386954044082e-05, "loss": 0.05, "step": 47970 }, { "epoch": 1.6220967578349503, "grad_norm": 0.43613114953041077, "learning_rate": 5.744751969981195e-05, "loss": 0.0519, "step": 47980 }, { "epoch": 1.6224348355252038, "grad_norm": 0.1634749472141266, "learning_rate": 5.7431169044733526e-05, "loss": 0.0474, "step": 47990 }, { "epoch": 1.622772913215457, "grad_norm": 0.20058226585388184, "learning_rate": 5.741481757699364e-05, "loss": 0.0524, "step": 48000 }, { "epoch": 1.62311099090571, "grad_norm": 0.25471505522727966, "learning_rate": 5.7398465298380434e-05, "loss": 0.0528, "step": 48010 }, { "epoch": 1.6234490685959635, "grad_norm": 0.23504097759723663, "learning_rate": 5.7382112210682193e-05, "loss": 0.0611, "step": 48020 }, { "epoch": 1.6237871462862166, "grad_norm": 0.27268242835998535, "learning_rate": 5.7365758315687266e-05, "loss": 0.0513, "step": 48030 }, { "epoch": 1.6241252239764696, "grad_norm": 0.40060773491859436, "learning_rate": 5.734940361518407e-05, "loss": 0.054, "step": 48040 }, { "epoch": 1.6244633016667231, "grad_norm": 0.3498547673225403, "learning_rate": 5.733304811096116e-05, "loss": 0.0518, "step": 48050 }, { "epoch": 1.6248013793569762, "grad_norm": 0.30670326948165894, "learning_rate": 5.731669180480713e-05, "loss": 0.0513, "step": 48060 }, { "epoch": 1.6251394570472295, "grad_norm": 0.3585165739059448, "learning_rate": 5.730033469851067e-05, "loss": 0.0558, "step": 48070 }, { "epoch": 1.6254775347374828, "grad_norm": 0.16467072069644928, "learning_rate": 5.7283976793860607e-05, "loss": 0.0488, "step": 48080 }, { "epoch": 1.6258156124277359, "grad_norm": 0.5903405547142029, "learning_rate": 5.726761809264577e-05, "loss": 0.0494, "step": 48090 }, { "epoch": 1.6261536901179892, "grad_norm": 0.31041064858436584, "learning_rate": 5.7251258596655155e-05, "loss": 0.0554, "step": 48100 }, { "epoch": 1.6264917678082425, "grad_norm": 0.26250848174095154, "learning_rate": 5.72348983076778e-05, "loss": 0.047, "step": 48110 }, { "epoch": 1.6268298454984955, "grad_norm": 0.5655842423439026, "learning_rate": 5.7218537227502854e-05, "loss": 0.0466, "step": 48120 }, { "epoch": 1.6271679231887488, "grad_norm": 0.14530475437641144, "learning_rate": 5.720217535791951e-05, "loss": 0.0454, "step": 48130 }, { "epoch": 1.627506000879002, "grad_norm": 0.2002089023590088, "learning_rate": 5.718581270071711e-05, "loss": 0.0517, "step": 48140 }, { "epoch": 1.6278440785692552, "grad_norm": 0.26385432481765747, "learning_rate": 5.716944925768505e-05, "loss": 0.0535, "step": 48150 }, { "epoch": 1.6281821562595085, "grad_norm": 0.27324873208999634, "learning_rate": 5.7153085030612786e-05, "loss": 0.0512, "step": 48160 }, { "epoch": 1.6285202339497618, "grad_norm": 0.24257944524288177, "learning_rate": 5.713672002128989e-05, "loss": 0.047, "step": 48170 }, { "epoch": 1.6288583116400148, "grad_norm": 0.22250373661518097, "learning_rate": 5.712035423150602e-05, "loss": 0.0551, "step": 48180 }, { "epoch": 1.6291963893302681, "grad_norm": 0.2163468897342682, "learning_rate": 5.710398766305094e-05, "loss": 0.0508, "step": 48190 }, { "epoch": 1.6295344670205214, "grad_norm": 0.31012606620788574, "learning_rate": 5.708762031771442e-05, "loss": 0.0574, "step": 48200 }, { "epoch": 1.6298725447107745, "grad_norm": 0.3376739025115967, "learning_rate": 5.707125219728642e-05, "loss": 0.0524, "step": 48210 }, { "epoch": 1.6302106224010278, "grad_norm": 0.19406229257583618, "learning_rate": 5.7054883303556905e-05, "loss": 0.0487, "step": 48220 }, { "epoch": 1.630548700091281, "grad_norm": 0.2989690601825714, "learning_rate": 5.703851363831595e-05, "loss": 0.0547, "step": 48230 }, { "epoch": 1.6308867777815341, "grad_norm": 0.23401102423667908, "learning_rate": 5.702214320335374e-05, "loss": 0.0503, "step": 48240 }, { "epoch": 1.6312248554717874, "grad_norm": 0.1934530884027481, "learning_rate": 5.7005772000460514e-05, "loss": 0.0529, "step": 48250 }, { "epoch": 1.6315629331620407, "grad_norm": 0.2752409279346466, "learning_rate": 5.698940003142656e-05, "loss": 0.0504, "step": 48260 }, { "epoch": 1.6319010108522938, "grad_norm": 0.2758052349090576, "learning_rate": 5.697302729804236e-05, "loss": 0.0583, "step": 48270 }, { "epoch": 1.632239088542547, "grad_norm": 0.33180806040763855, "learning_rate": 5.695665380209837e-05, "loss": 0.0495, "step": 48280 }, { "epoch": 1.6325771662328004, "grad_norm": 0.16469806432724, "learning_rate": 5.6940279545385165e-05, "loss": 0.0587, "step": 48290 }, { "epoch": 1.6329152439230534, "grad_norm": 0.20641380548477173, "learning_rate": 5.692390452969344e-05, "loss": 0.0554, "step": 48300 }, { "epoch": 1.6332533216133067, "grad_norm": 0.19054055213928223, "learning_rate": 5.690752875681392e-05, "loss": 0.0466, "step": 48310 }, { "epoch": 1.63359139930356, "grad_norm": 0.43224114179611206, "learning_rate": 5.6891152228537435e-05, "loss": 0.0493, "step": 48320 }, { "epoch": 1.633929476993813, "grad_norm": 0.2853250801563263, "learning_rate": 5.687477494665492e-05, "loss": 0.0489, "step": 48330 }, { "epoch": 1.6342675546840664, "grad_norm": 0.30791765451431274, "learning_rate": 5.685839691295733e-05, "loss": 0.0514, "step": 48340 }, { "epoch": 1.6346056323743197, "grad_norm": 0.39797648787498474, "learning_rate": 5.6842018129235786e-05, "loss": 0.0502, "step": 48350 }, { "epoch": 1.6349437100645727, "grad_norm": 0.3785780668258667, "learning_rate": 5.6825638597281404e-05, "loss": 0.053, "step": 48360 }, { "epoch": 1.635281787754826, "grad_norm": 0.4288260042667389, "learning_rate": 5.680925831888546e-05, "loss": 0.0526, "step": 48370 }, { "epoch": 1.6356198654450793, "grad_norm": 0.22467972338199615, "learning_rate": 5.6792877295839274e-05, "loss": 0.056, "step": 48380 }, { "epoch": 1.6359579431353324, "grad_norm": 0.5291633605957031, "learning_rate": 5.6776495529934224e-05, "loss": 0.0466, "step": 48390 }, { "epoch": 1.6362960208255857, "grad_norm": 0.23037734627723694, "learning_rate": 5.6760113022961824e-05, "loss": 0.0483, "step": 48400 }, { "epoch": 1.636634098515839, "grad_norm": 0.15598386526107788, "learning_rate": 5.6743729776713617e-05, "loss": 0.0519, "step": 48410 }, { "epoch": 1.636972176206092, "grad_norm": 0.23614872992038727, "learning_rate": 5.672734579298126e-05, "loss": 0.0556, "step": 48420 }, { "epoch": 1.6373102538963455, "grad_norm": 0.3400670886039734, "learning_rate": 5.671096107355649e-05, "loss": 0.0467, "step": 48430 }, { "epoch": 1.6376483315865986, "grad_norm": 0.26473286747932434, "learning_rate": 5.66945756202311e-05, "loss": 0.0505, "step": 48440 }, { "epoch": 1.6379864092768517, "grad_norm": 0.25715869665145874, "learning_rate": 5.667818943479699e-05, "loss": 0.0485, "step": 48450 }, { "epoch": 1.6383244869671052, "grad_norm": 0.37413671612739563, "learning_rate": 5.666180251904612e-05, "loss": 0.0488, "step": 48460 }, { "epoch": 1.6386625646573583, "grad_norm": 0.20983602106571198, "learning_rate": 5.6645414874770555e-05, "loss": 0.0497, "step": 48470 }, { "epoch": 1.6390006423476113, "grad_norm": 0.2504423260688782, "learning_rate": 5.66290265037624e-05, "loss": 0.0489, "step": 48480 }, { "epoch": 1.6393387200378648, "grad_norm": 0.24480876326560974, "learning_rate": 5.661263740781386e-05, "loss": 0.05, "step": 48490 }, { "epoch": 1.639676797728118, "grad_norm": 0.44420868158340454, "learning_rate": 5.6596247588717254e-05, "loss": 0.0473, "step": 48500 }, { "epoch": 1.640014875418371, "grad_norm": 0.20643506944179535, "learning_rate": 5.6579857048264926e-05, "loss": 0.0486, "step": 48510 }, { "epoch": 1.6403529531086245, "grad_norm": 0.1605074405670166, "learning_rate": 5.6563465788249314e-05, "loss": 0.0483, "step": 48520 }, { "epoch": 1.6406910307988776, "grad_norm": 0.4023255407810211, "learning_rate": 5.6547073810462956e-05, "loss": 0.0482, "step": 48530 }, { "epoch": 1.6410291084891306, "grad_norm": 0.2711000144481659, "learning_rate": 5.653068111669846e-05, "loss": 0.0543, "step": 48540 }, { "epoch": 1.6413671861793842, "grad_norm": 0.39235907793045044, "learning_rate": 5.651428770874848e-05, "loss": 0.0541, "step": 48550 }, { "epoch": 1.6417052638696372, "grad_norm": 0.25702670216560364, "learning_rate": 5.64978935884058e-05, "loss": 0.0512, "step": 48560 }, { "epoch": 1.6420433415598905, "grad_norm": 0.4366973638534546, "learning_rate": 5.6481498757463244e-05, "loss": 0.0527, "step": 48570 }, { "epoch": 1.6423814192501438, "grad_norm": 0.2255077362060547, "learning_rate": 5.646510321771373e-05, "loss": 0.0564, "step": 48580 }, { "epoch": 1.6427194969403969, "grad_norm": 0.21094568073749542, "learning_rate": 5.644870697095024e-05, "loss": 0.0454, "step": 48590 }, { "epoch": 1.6430575746306502, "grad_norm": 0.3068980574607849, "learning_rate": 5.643231001896586e-05, "loss": 0.0518, "step": 48600 }, { "epoch": 1.6433956523209035, "grad_norm": 0.29011431336402893, "learning_rate": 5.6415912363553726e-05, "loss": 0.0491, "step": 48610 }, { "epoch": 1.6437337300111565, "grad_norm": 0.20795032382011414, "learning_rate": 5.639951400650706e-05, "loss": 0.0474, "step": 48620 }, { "epoch": 1.6440718077014098, "grad_norm": 0.19798466563224792, "learning_rate": 5.6383114949619165e-05, "loss": 0.0509, "step": 48630 }, { "epoch": 1.644409885391663, "grad_norm": 0.3186815083026886, "learning_rate": 5.636671519468342e-05, "loss": 0.0531, "step": 48640 }, { "epoch": 1.6447479630819162, "grad_norm": 0.20924794673919678, "learning_rate": 5.635031474349327e-05, "loss": 0.0523, "step": 48650 }, { "epoch": 1.6450860407721695, "grad_norm": 0.16868238151073456, "learning_rate": 5.6333913597842246e-05, "loss": 0.0476, "step": 48660 }, { "epoch": 1.6454241184624228, "grad_norm": 0.2601069211959839, "learning_rate": 5.6317511759523955e-05, "loss": 0.0499, "step": 48670 }, { "epoch": 1.6457621961526758, "grad_norm": 0.4015967845916748, "learning_rate": 5.630110923033207e-05, "loss": 0.0518, "step": 48680 }, { "epoch": 1.6461002738429291, "grad_norm": 0.2026623636484146, "learning_rate": 5.628470601206036e-05, "loss": 0.0526, "step": 48690 }, { "epoch": 1.6464383515331824, "grad_norm": 0.18215656280517578, "learning_rate": 5.626830210650266e-05, "loss": 0.0535, "step": 48700 }, { "epoch": 1.6467764292234355, "grad_norm": 2.1196978092193604, "learning_rate": 5.625189751545285e-05, "loss": 0.0514, "step": 48710 }, { "epoch": 1.6471145069136888, "grad_norm": 0.21492668986320496, "learning_rate": 5.6235492240704936e-05, "loss": 0.0563, "step": 48720 }, { "epoch": 1.647452584603942, "grad_norm": 0.16791918873786926, "learning_rate": 5.621908628405296e-05, "loss": 0.0555, "step": 48730 }, { "epoch": 1.6477906622941951, "grad_norm": 0.20635981857776642, "learning_rate": 5.620267964729106e-05, "loss": 0.0516, "step": 48740 }, { "epoch": 1.6481287399844484, "grad_norm": 0.16621968150138855, "learning_rate": 5.618627233221344e-05, "loss": 0.0467, "step": 48750 }, { "epoch": 1.6484668176747017, "grad_norm": 0.1344289928674698, "learning_rate": 5.616986434061438e-05, "loss": 0.0541, "step": 48760 }, { "epoch": 1.6488048953649548, "grad_norm": 0.48861560225486755, "learning_rate": 5.615345567428822e-05, "loss": 0.0503, "step": 48770 }, { "epoch": 1.649142973055208, "grad_norm": 0.3346039950847626, "learning_rate": 5.613704633502941e-05, "loss": 0.0502, "step": 48780 }, { "epoch": 1.6494810507454614, "grad_norm": 0.3885571360588074, "learning_rate": 5.612063632463245e-05, "loss": 0.0493, "step": 48790 }, { "epoch": 1.6498191284357144, "grad_norm": 0.2290022075176239, "learning_rate": 5.610422564489188e-05, "loss": 0.0587, "step": 48800 }, { "epoch": 1.6501572061259677, "grad_norm": 0.32353341579437256, "learning_rate": 5.608781429760239e-05, "loss": 0.0518, "step": 48810 }, { "epoch": 1.650495283816221, "grad_norm": 0.2692195475101471, "learning_rate": 5.607140228455866e-05, "loss": 0.053, "step": 48820 }, { "epoch": 1.650833361506474, "grad_norm": 0.22778424620628357, "learning_rate": 5.605498960755553e-05, "loss": 0.0557, "step": 48830 }, { "epoch": 1.6511714391967274, "grad_norm": 0.35802048444747925, "learning_rate": 5.603857626838782e-05, "loss": 0.0497, "step": 48840 }, { "epoch": 1.6515095168869807, "grad_norm": 0.2576430141925812, "learning_rate": 5.60221622688505e-05, "loss": 0.0527, "step": 48850 }, { "epoch": 1.6518475945772337, "grad_norm": 0.30456724762916565, "learning_rate": 5.6005747610738565e-05, "loss": 0.0509, "step": 48860 }, { "epoch": 1.652185672267487, "grad_norm": 0.2468867003917694, "learning_rate": 5.59893322958471e-05, "loss": 0.0452, "step": 48870 }, { "epoch": 1.6525237499577403, "grad_norm": 0.27886784076690674, "learning_rate": 5.5972916325971256e-05, "loss": 0.0559, "step": 48880 }, { "epoch": 1.6528618276479934, "grad_norm": 0.21568679809570312, "learning_rate": 5.595649970290628e-05, "loss": 0.0475, "step": 48890 }, { "epoch": 1.6531999053382467, "grad_norm": 0.2290380746126175, "learning_rate": 5.5940082428447426e-05, "loss": 0.0466, "step": 48900 }, { "epoch": 1.6535379830285, "grad_norm": 0.15539516508579254, "learning_rate": 5.592366450439012e-05, "loss": 0.0492, "step": 48910 }, { "epoch": 1.653876060718753, "grad_norm": 0.25803548097610474, "learning_rate": 5.590724593252975e-05, "loss": 0.0537, "step": 48920 }, { "epoch": 1.6542141384090066, "grad_norm": 0.28027307987213135, "learning_rate": 5.589082671466184e-05, "loss": 0.0539, "step": 48930 }, { "epoch": 1.6545522160992596, "grad_norm": 0.46433359384536743, "learning_rate": 5.587440685258199e-05, "loss": 0.0516, "step": 48940 }, { "epoch": 1.6548902937895127, "grad_norm": 0.19624163210391998, "learning_rate": 5.585798634808583e-05, "loss": 0.054, "step": 48950 }, { "epoch": 1.6552283714797662, "grad_norm": 0.2973482310771942, "learning_rate": 5.584156520296909e-05, "loss": 0.054, "step": 48960 }, { "epoch": 1.6555664491700193, "grad_norm": 0.2604665756225586, "learning_rate": 5.582514341902757e-05, "loss": 0.0567, "step": 48970 }, { "epoch": 1.6559045268602723, "grad_norm": 0.17930257320404053, "learning_rate": 5.580872099805713e-05, "loss": 0.0538, "step": 48980 }, { "epoch": 1.6562426045505259, "grad_norm": 0.20118458569049835, "learning_rate": 5.5792297941853674e-05, "loss": 0.0518, "step": 48990 }, { "epoch": 1.656580682240779, "grad_norm": 0.39002344012260437, "learning_rate": 5.5775874252213247e-05, "loss": 0.0476, "step": 49000 }, { "epoch": 1.656918759931032, "grad_norm": 0.27620917558670044, "learning_rate": 5.575944993093189e-05, "loss": 0.0559, "step": 49010 }, { "epoch": 1.6572568376212855, "grad_norm": 0.2546897828578949, "learning_rate": 5.574302497980574e-05, "loss": 0.0569, "step": 49020 }, { "epoch": 1.6575949153115386, "grad_norm": 0.2316572517156601, "learning_rate": 5.5726599400631005e-05, "loss": 0.0477, "step": 49030 }, { "epoch": 1.6579329930017919, "grad_norm": 0.20751945674419403, "learning_rate": 5.571017319520399e-05, "loss": 0.0545, "step": 49040 }, { "epoch": 1.6582710706920452, "grad_norm": 0.24539947509765625, "learning_rate": 5.5693746365320984e-05, "loss": 0.0496, "step": 49050 }, { "epoch": 1.6586091483822982, "grad_norm": 0.25618603825569153, "learning_rate": 5.567731891277848e-05, "loss": 0.0536, "step": 49060 }, { "epoch": 1.6589472260725515, "grad_norm": 0.3394943177700043, "learning_rate": 5.566089083937288e-05, "loss": 0.0605, "step": 49070 }, { "epoch": 1.6592853037628048, "grad_norm": 0.23040789365768433, "learning_rate": 5.564446214690079e-05, "loss": 0.0518, "step": 49080 }, { "epoch": 1.6596233814530579, "grad_norm": 0.2095382660627365, "learning_rate": 5.56280328371588e-05, "loss": 0.0473, "step": 49090 }, { "epoch": 1.6599614591433112, "grad_norm": 0.2353113293647766, "learning_rate": 5.5611602911943596e-05, "loss": 0.0505, "step": 49100 }, { "epoch": 1.6602995368335645, "grad_norm": 0.2183997631072998, "learning_rate": 5.559517237305194e-05, "loss": 0.0541, "step": 49110 }, { "epoch": 1.6606376145238175, "grad_norm": 0.30029019713401794, "learning_rate": 5.557874122228064e-05, "loss": 0.05, "step": 49120 }, { "epoch": 1.6609756922140708, "grad_norm": 0.27985233068466187, "learning_rate": 5.55623094614266e-05, "loss": 0.0517, "step": 49130 }, { "epoch": 1.6613137699043241, "grad_norm": 0.27865663170814514, "learning_rate": 5.5545877092286743e-05, "loss": 0.053, "step": 49140 }, { "epoch": 1.6616518475945772, "grad_norm": 0.21699878573417664, "learning_rate": 5.552944411665811e-05, "loss": 0.0471, "step": 49150 }, { "epoch": 1.6619899252848305, "grad_norm": 0.21113808453083038, "learning_rate": 5.5513010536337795e-05, "loss": 0.0487, "step": 49160 }, { "epoch": 1.6623280029750838, "grad_norm": 0.2180628776550293, "learning_rate": 5.5496576353122934e-05, "loss": 0.053, "step": 49170 }, { "epoch": 1.6626660806653368, "grad_norm": 0.32675498723983765, "learning_rate": 5.548014156881074e-05, "loss": 0.0514, "step": 49180 }, { "epoch": 1.6630041583555901, "grad_norm": 0.22626882791519165, "learning_rate": 5.5463706185198494e-05, "loss": 0.0525, "step": 49190 }, { "epoch": 1.6633422360458434, "grad_norm": 0.22580143809318542, "learning_rate": 5.544727020408358e-05, "loss": 0.0539, "step": 49200 }, { "epoch": 1.6636803137360965, "grad_norm": 0.26233816146850586, "learning_rate": 5.5430833627263366e-05, "loss": 0.0522, "step": 49210 }, { "epoch": 1.6640183914263498, "grad_norm": 0.19803547859191895, "learning_rate": 5.5414396456535354e-05, "loss": 0.0466, "step": 49220 }, { "epoch": 1.664356469116603, "grad_norm": 0.21558186411857605, "learning_rate": 5.539795869369709e-05, "loss": 0.0475, "step": 49230 }, { "epoch": 1.6646945468068561, "grad_norm": 0.275055468082428, "learning_rate": 5.538152034054618e-05, "loss": 0.0535, "step": 49240 }, { "epoch": 1.6650326244971094, "grad_norm": 0.2397138476371765, "learning_rate": 5.536508139888028e-05, "loss": 0.0535, "step": 49250 }, { "epoch": 1.6653707021873627, "grad_norm": 0.2569384276866913, "learning_rate": 5.534864187049716e-05, "loss": 0.0464, "step": 49260 }, { "epoch": 1.6657087798776158, "grad_norm": 0.18121063709259033, "learning_rate": 5.5332201757194615e-05, "loss": 0.0502, "step": 49270 }, { "epoch": 1.666046857567869, "grad_norm": 0.24751579761505127, "learning_rate": 5.531576106077049e-05, "loss": 0.0474, "step": 49280 }, { "epoch": 1.6663849352581224, "grad_norm": 0.23000939190387726, "learning_rate": 5.529931978302272e-05, "loss": 0.0461, "step": 49290 }, { "epoch": 1.6667230129483754, "grad_norm": 0.3178247809410095, "learning_rate": 5.528287792574932e-05, "loss": 0.0524, "step": 49300 }, { "epoch": 1.6670610906386287, "grad_norm": 0.2587776780128479, "learning_rate": 5.5266435490748324e-05, "loss": 0.0513, "step": 49310 }, { "epoch": 1.667399168328882, "grad_norm": 0.3067704737186432, "learning_rate": 5.524999247981787e-05, "loss": 0.0516, "step": 49320 }, { "epoch": 1.667737246019135, "grad_norm": 0.2581906318664551, "learning_rate": 5.523354889475613e-05, "loss": 0.0533, "step": 49330 }, { "epoch": 1.6680753237093884, "grad_norm": 0.5439001321792603, "learning_rate": 5.521710473736134e-05, "loss": 0.0524, "step": 49340 }, { "epoch": 1.6684134013996417, "grad_norm": 0.17797081172466278, "learning_rate": 5.5200660009431826e-05, "loss": 0.0511, "step": 49350 }, { "epoch": 1.6687514790898947, "grad_norm": 0.6793456077575684, "learning_rate": 5.518421471276596e-05, "loss": 0.0523, "step": 49360 }, { "epoch": 1.669089556780148, "grad_norm": 0.3052690029144287, "learning_rate": 5.5167768849162174e-05, "loss": 0.0545, "step": 49370 }, { "epoch": 1.6694276344704013, "grad_norm": 0.18659287691116333, "learning_rate": 5.515132242041893e-05, "loss": 0.0546, "step": 49380 }, { "epoch": 1.6697657121606544, "grad_norm": 0.16547800600528717, "learning_rate": 5.513487542833483e-05, "loss": 0.0511, "step": 49390 }, { "epoch": 1.670103789850908, "grad_norm": 0.2412237823009491, "learning_rate": 5.5118427874708466e-05, "loss": 0.0454, "step": 49400 }, { "epoch": 1.670441867541161, "grad_norm": 0.21504881978034973, "learning_rate": 5.510197976133853e-05, "loss": 0.0546, "step": 49410 }, { "epoch": 1.670779945231414, "grad_norm": 0.20323875546455383, "learning_rate": 5.508553109002376e-05, "loss": 0.0483, "step": 49420 }, { "epoch": 1.6711180229216676, "grad_norm": 0.37184739112854004, "learning_rate": 5.5069081862562957e-05, "loss": 0.0499, "step": 49430 }, { "epoch": 1.6714561006119206, "grad_norm": 0.386621356010437, "learning_rate": 5.5052632080754965e-05, "loss": 0.0516, "step": 49440 }, { "epoch": 1.6717941783021737, "grad_norm": 0.26308372616767883, "learning_rate": 5.503618174639874e-05, "loss": 0.0458, "step": 49450 }, { "epoch": 1.6721322559924272, "grad_norm": 0.286995530128479, "learning_rate": 5.501973086129325e-05, "loss": 0.05, "step": 49460 }, { "epoch": 1.6724703336826803, "grad_norm": 0.820740818977356, "learning_rate": 5.500327942723753e-05, "loss": 0.0533, "step": 49470 }, { "epoch": 1.6728084113729333, "grad_norm": 0.4453868865966797, "learning_rate": 5.498682744603071e-05, "loss": 0.0533, "step": 49480 }, { "epoch": 1.6731464890631869, "grad_norm": 0.20358119904994965, "learning_rate": 5.497037491947194e-05, "loss": 0.0494, "step": 49490 }, { "epoch": 1.67348456675344, "grad_norm": 0.2689347267150879, "learning_rate": 5.4953921849360424e-05, "loss": 0.0567, "step": 49500 }, { "epoch": 1.673822644443693, "grad_norm": 0.1824740767478943, "learning_rate": 5.493746823749547e-05, "loss": 0.0493, "step": 49510 }, { "epoch": 1.6741607221339465, "grad_norm": 0.14807191491127014, "learning_rate": 5.49210140856764e-05, "loss": 0.0523, "step": 49520 }, { "epoch": 1.6744987998241996, "grad_norm": 0.2425273060798645, "learning_rate": 5.4904559395702647e-05, "loss": 0.0462, "step": 49530 }, { "epoch": 1.6748368775144529, "grad_norm": 0.2583755552768707, "learning_rate": 5.488810416937364e-05, "loss": 0.0501, "step": 49540 }, { "epoch": 1.6751749552047062, "grad_norm": 0.21092098951339722, "learning_rate": 5.4871648408488926e-05, "loss": 0.0563, "step": 49550 }, { "epoch": 1.6755130328949592, "grad_norm": 0.15891103446483612, "learning_rate": 5.485519211484807e-05, "loss": 0.0514, "step": 49560 }, { "epoch": 1.6758511105852125, "grad_norm": 0.2422749251127243, "learning_rate": 5.4838735290250676e-05, "loss": 0.0488, "step": 49570 }, { "epoch": 1.6761891882754658, "grad_norm": 0.2666306793689728, "learning_rate": 5.482227793649649e-05, "loss": 0.0446, "step": 49580 }, { "epoch": 1.6765272659657189, "grad_norm": 0.2592788338661194, "learning_rate": 5.480582005538524e-05, "loss": 0.057, "step": 49590 }, { "epoch": 1.6768653436559722, "grad_norm": 0.19129005074501038, "learning_rate": 5.478936164871671e-05, "loss": 0.0485, "step": 49600 }, { "epoch": 1.6772034213462255, "grad_norm": 0.17050836980342865, "learning_rate": 5.477290271829081e-05, "loss": 0.0525, "step": 49610 }, { "epoch": 1.6775414990364785, "grad_norm": 0.25840291380882263, "learning_rate": 5.475644326590744e-05, "loss": 0.0544, "step": 49620 }, { "epoch": 1.6778795767267318, "grad_norm": 0.4388199746608734, "learning_rate": 5.473998329336658e-05, "loss": 0.0549, "step": 49630 }, { "epoch": 1.6782176544169851, "grad_norm": 0.29867446422576904, "learning_rate": 5.4723522802468286e-05, "loss": 0.0551, "step": 49640 }, { "epoch": 1.6785557321072382, "grad_norm": 0.2528788149356842, "learning_rate": 5.470706179501264e-05, "loss": 0.0507, "step": 49650 }, { "epoch": 1.6788938097974915, "grad_norm": 0.2788027226924896, "learning_rate": 5.4690600272799776e-05, "loss": 0.0533, "step": 49660 }, { "epoch": 1.6792318874877448, "grad_norm": 0.2049158811569214, "learning_rate": 5.467413823762993e-05, "loss": 0.0478, "step": 49670 }, { "epoch": 1.6795699651779978, "grad_norm": 0.21643711626529694, "learning_rate": 5.465767569130335e-05, "loss": 0.0501, "step": 49680 }, { "epoch": 1.6799080428682511, "grad_norm": 0.19861215353012085, "learning_rate": 5.464121263562036e-05, "loss": 0.0493, "step": 49690 }, { "epoch": 1.6802461205585044, "grad_norm": 0.34168386459350586, "learning_rate": 5.4624749072381343e-05, "loss": 0.053, "step": 49700 }, { "epoch": 1.6805841982487575, "grad_norm": 0.17275983095169067, "learning_rate": 5.460828500338672e-05, "loss": 0.0505, "step": 49710 }, { "epoch": 1.6809222759390108, "grad_norm": 0.38513970375061035, "learning_rate": 5.459182043043698e-05, "loss": 0.0516, "step": 49720 }, { "epoch": 1.681260353629264, "grad_norm": 0.34152165055274963, "learning_rate": 5.457535535533265e-05, "loss": 0.0552, "step": 49730 }, { "epoch": 1.6815984313195171, "grad_norm": 0.21786528825759888, "learning_rate": 5.4558889779874334e-05, "loss": 0.0475, "step": 49740 }, { "epoch": 1.6819365090097704, "grad_norm": 0.2758769690990448, "learning_rate": 5.454242370586269e-05, "loss": 0.0531, "step": 49750 }, { "epoch": 1.6822745867000237, "grad_norm": 0.18679894506931305, "learning_rate": 5.452595713509843e-05, "loss": 0.0473, "step": 49760 }, { "epoch": 1.6826126643902768, "grad_norm": 0.3997807502746582, "learning_rate": 5.4509490069382274e-05, "loss": 0.0524, "step": 49770 }, { "epoch": 1.68295074208053, "grad_norm": 0.42224666476249695, "learning_rate": 5.449302251051509e-05, "loss": 0.0547, "step": 49780 }, { "epoch": 1.6832888197707834, "grad_norm": 0.20235411822795868, "learning_rate": 5.44765544602977e-05, "loss": 0.0492, "step": 49790 }, { "epoch": 1.6836268974610364, "grad_norm": 0.25578030943870544, "learning_rate": 5.446008592053107e-05, "loss": 0.0559, "step": 49800 }, { "epoch": 1.6839649751512897, "grad_norm": 0.17687253654003143, "learning_rate": 5.4443616893016135e-05, "loss": 0.0505, "step": 49810 }, { "epoch": 1.684303052841543, "grad_norm": 0.3279340863227844, "learning_rate": 5.4427147379553934e-05, "loss": 0.0507, "step": 49820 }, { "epoch": 1.684641130531796, "grad_norm": 0.21393197774887085, "learning_rate": 5.441067738194556e-05, "loss": 0.0496, "step": 49830 }, { "epoch": 1.6849792082220494, "grad_norm": 0.1765405684709549, "learning_rate": 5.439420690199214e-05, "loss": 0.0576, "step": 49840 }, { "epoch": 1.6853172859123027, "grad_norm": 0.3030707836151123, "learning_rate": 5.437773594149484e-05, "loss": 0.0512, "step": 49850 }, { "epoch": 1.6856553636025557, "grad_norm": 0.2081577181816101, "learning_rate": 5.436126450225495e-05, "loss": 0.0534, "step": 49860 }, { "epoch": 1.685993441292809, "grad_norm": 0.34999051690101624, "learning_rate": 5.434479258607371e-05, "loss": 0.0504, "step": 49870 }, { "epoch": 1.6863315189830623, "grad_norm": 0.28189095854759216, "learning_rate": 5.43283201947525e-05, "loss": 0.0539, "step": 49880 }, { "epoch": 1.6866695966733154, "grad_norm": 0.22227050364017487, "learning_rate": 5.431184733009268e-05, "loss": 0.0495, "step": 49890 }, { "epoch": 1.687007674363569, "grad_norm": 0.18868771195411682, "learning_rate": 5.4295373993895736e-05, "loss": 0.0462, "step": 49900 }, { "epoch": 1.687345752053822, "grad_norm": 0.20537343621253967, "learning_rate": 5.4278900187963157e-05, "loss": 0.0528, "step": 49910 }, { "epoch": 1.687683829744075, "grad_norm": 0.2888089418411255, "learning_rate": 5.426242591409646e-05, "loss": 0.0461, "step": 49920 }, { "epoch": 1.6880219074343286, "grad_norm": 0.19840216636657715, "learning_rate": 5.42459511740973e-05, "loss": 0.0541, "step": 49930 }, { "epoch": 1.6883599851245816, "grad_norm": 0.2715575397014618, "learning_rate": 5.422947596976729e-05, "loss": 0.0583, "step": 49940 }, { "epoch": 1.6886980628148347, "grad_norm": 0.192665234208107, "learning_rate": 5.4213000302908134e-05, "loss": 0.0526, "step": 49950 }, { "epoch": 1.6890361405050882, "grad_norm": 0.22229544818401337, "learning_rate": 5.419652417532162e-05, "loss": 0.0497, "step": 49960 }, { "epoch": 1.6893742181953413, "grad_norm": 0.3208460807800293, "learning_rate": 5.4180047588809534e-05, "loss": 0.0523, "step": 49970 }, { "epoch": 1.6897122958855944, "grad_norm": 0.3477281332015991, "learning_rate": 5.4163570545173704e-05, "loss": 0.0481, "step": 49980 }, { "epoch": 1.6900503735758479, "grad_norm": 0.3140910863876343, "learning_rate": 5.414709304621608e-05, "loss": 0.0513, "step": 49990 }, { "epoch": 1.690388451266101, "grad_norm": 0.34527358412742615, "learning_rate": 5.413061509373858e-05, "loss": 0.0505, "step": 50000 }, { "epoch": 1.6907265289563542, "grad_norm": 0.1895660161972046, "learning_rate": 5.411413668954324e-05, "loss": 0.0518, "step": 50010 }, { "epoch": 1.6910646066466075, "grad_norm": 0.25206899642944336, "learning_rate": 5.4097657835432083e-05, "loss": 0.0518, "step": 50020 }, { "epoch": 1.6914026843368606, "grad_norm": 0.33477142453193665, "learning_rate": 5.408117853320723e-05, "loss": 0.0464, "step": 50030 }, { "epoch": 1.6917407620271139, "grad_norm": 0.21458449959754944, "learning_rate": 5.406469878467084e-05, "loss": 0.0532, "step": 50040 }, { "epoch": 1.6920788397173672, "grad_norm": 0.2445717453956604, "learning_rate": 5.404821859162509e-05, "loss": 0.0485, "step": 50050 }, { "epoch": 1.6924169174076202, "grad_norm": 0.1940104067325592, "learning_rate": 5.403173795587225e-05, "loss": 0.0558, "step": 50060 }, { "epoch": 1.6927549950978735, "grad_norm": 0.1902856081724167, "learning_rate": 5.4015256879214606e-05, "loss": 0.0543, "step": 50070 }, { "epoch": 1.6930930727881268, "grad_norm": 0.2237536609172821, "learning_rate": 5.39987753634545e-05, "loss": 0.0525, "step": 50080 }, { "epoch": 1.6934311504783799, "grad_norm": 0.3323976993560791, "learning_rate": 5.3982293410394335e-05, "loss": 0.053, "step": 50090 }, { "epoch": 1.6937692281686332, "grad_norm": 0.33434897661209106, "learning_rate": 5.396581102183655e-05, "loss": 0.0449, "step": 50100 }, { "epoch": 1.6941073058588865, "grad_norm": 1.0828543901443481, "learning_rate": 5.3949328199583634e-05, "loss": 0.0532, "step": 50110 }, { "epoch": 1.6944453835491395, "grad_norm": 0.23657742142677307, "learning_rate": 5.393284494543813e-05, "loss": 0.0527, "step": 50120 }, { "epoch": 1.6947834612393928, "grad_norm": 0.17047235369682312, "learning_rate": 5.391636126120262e-05, "loss": 0.0508, "step": 50130 }, { "epoch": 1.6951215389296461, "grad_norm": 0.2550262212753296, "learning_rate": 5.389987714867971e-05, "loss": 0.0575, "step": 50140 }, { "epoch": 1.6954596166198992, "grad_norm": 0.255138635635376, "learning_rate": 5.3883392609672114e-05, "loss": 0.0551, "step": 50150 }, { "epoch": 1.6957976943101525, "grad_norm": 0.6528826355934143, "learning_rate": 5.3866907645982545e-05, "loss": 0.0482, "step": 50160 }, { "epoch": 1.6961357720004058, "grad_norm": 0.2606407403945923, "learning_rate": 5.385042225941375e-05, "loss": 0.0521, "step": 50170 }, { "epoch": 1.6964738496906588, "grad_norm": 0.5255205035209656, "learning_rate": 5.3833936451768585e-05, "loss": 0.0533, "step": 50180 }, { "epoch": 1.6968119273809121, "grad_norm": 0.18848811089992523, "learning_rate": 5.3817450224849884e-05, "loss": 0.0508, "step": 50190 }, { "epoch": 1.6971500050711654, "grad_norm": 0.2973146438598633, "learning_rate": 5.380096358046056e-05, "loss": 0.0533, "step": 50200 }, { "epoch": 1.6974880827614185, "grad_norm": 0.19721050560474396, "learning_rate": 5.378447652040359e-05, "loss": 0.0491, "step": 50210 }, { "epoch": 1.6978261604516718, "grad_norm": 0.19778649508953094, "learning_rate": 5.376798904648195e-05, "loss": 0.0496, "step": 50220 }, { "epoch": 1.698164238141925, "grad_norm": 0.29202333092689514, "learning_rate": 5.375150116049869e-05, "loss": 0.0466, "step": 50230 }, { "epoch": 1.6985023158321781, "grad_norm": 0.21162575483322144, "learning_rate": 5.373501286425691e-05, "loss": 0.046, "step": 50240 }, { "epoch": 1.6988403935224314, "grad_norm": 0.1622617095708847, "learning_rate": 5.3718524159559726e-05, "loss": 0.0523, "step": 50250 }, { "epoch": 1.6991784712126847, "grad_norm": 0.1483754962682724, "learning_rate": 5.370203504821034e-05, "loss": 0.0538, "step": 50260 }, { "epoch": 1.6995165489029378, "grad_norm": 0.25340932607650757, "learning_rate": 5.368554553201196e-05, "loss": 0.0539, "step": 50270 }, { "epoch": 1.699854626593191, "grad_norm": 0.23199881613254547, "learning_rate": 5.3669055612767874e-05, "loss": 0.0494, "step": 50280 }, { "epoch": 1.7001927042834444, "grad_norm": 0.1777285635471344, "learning_rate": 5.3652565292281376e-05, "loss": 0.049, "step": 50290 }, { "epoch": 1.7005307819736974, "grad_norm": 0.3930206000804901, "learning_rate": 5.3636074572355824e-05, "loss": 0.0542, "step": 50300 }, { "epoch": 1.7008688596639507, "grad_norm": 0.36021992564201355, "learning_rate": 5.361958345479463e-05, "loss": 0.0491, "step": 50310 }, { "epoch": 1.701206937354204, "grad_norm": 0.2846003770828247, "learning_rate": 5.360309194140123e-05, "loss": 0.0524, "step": 50320 }, { "epoch": 1.701545015044457, "grad_norm": 0.27891674637794495, "learning_rate": 5.3586600033979106e-05, "loss": 0.0542, "step": 50330 }, { "epoch": 1.7018830927347104, "grad_norm": 0.2641310691833496, "learning_rate": 5.357010773433181e-05, "loss": 0.0465, "step": 50340 }, { "epoch": 1.7022211704249637, "grad_norm": 0.28930550813674927, "learning_rate": 5.3553615044262895e-05, "loss": 0.0478, "step": 50350 }, { "epoch": 1.7025592481152168, "grad_norm": 0.27514874935150146, "learning_rate": 5.3537121965575964e-05, "loss": 0.0531, "step": 50360 }, { "epoch": 1.7028973258054703, "grad_norm": 0.21736595034599304, "learning_rate": 5.352062850007471e-05, "loss": 0.0521, "step": 50370 }, { "epoch": 1.7032354034957233, "grad_norm": 0.1989249438047409, "learning_rate": 5.350413464956282e-05, "loss": 0.0488, "step": 50380 }, { "epoch": 1.7035734811859764, "grad_norm": 0.4184495806694031, "learning_rate": 5.348764041584403e-05, "loss": 0.0503, "step": 50390 }, { "epoch": 1.70391155887623, "grad_norm": 0.24275299906730652, "learning_rate": 5.347114580072212e-05, "loss": 0.0532, "step": 50400 }, { "epoch": 1.704249636566483, "grad_norm": 0.249223530292511, "learning_rate": 5.345465080600093e-05, "loss": 0.0519, "step": 50410 }, { "epoch": 1.704587714256736, "grad_norm": 0.14491239190101624, "learning_rate": 5.3438155433484326e-05, "loss": 0.0469, "step": 50420 }, { "epoch": 1.7049257919469896, "grad_norm": 0.31545913219451904, "learning_rate": 5.3421659684976197e-05, "loss": 0.0508, "step": 50430 }, { "epoch": 1.7052638696372426, "grad_norm": 0.24049413204193115, "learning_rate": 5.340516356228052e-05, "loss": 0.0475, "step": 50440 }, { "epoch": 1.7056019473274957, "grad_norm": 0.24817371368408203, "learning_rate": 5.338866706720128e-05, "loss": 0.0534, "step": 50450 }, { "epoch": 1.7059400250177492, "grad_norm": 0.49619805812835693, "learning_rate": 5.337217020154249e-05, "loss": 0.0566, "step": 50460 }, { "epoch": 1.7062781027080023, "grad_norm": 0.15917323529720306, "learning_rate": 5.335567296710825e-05, "loss": 0.06, "step": 50470 }, { "epoch": 1.7066161803982554, "grad_norm": 0.3084448575973511, "learning_rate": 5.333917536570265e-05, "loss": 0.0516, "step": 50480 }, { "epoch": 1.7069542580885089, "grad_norm": 0.2817935645580292, "learning_rate": 5.332267739912986e-05, "loss": 0.0507, "step": 50490 }, { "epoch": 1.707292335778762, "grad_norm": 0.1876005232334137, "learning_rate": 5.330617906919405e-05, "loss": 0.0482, "step": 50500 }, { "epoch": 1.7076304134690152, "grad_norm": 0.4087553024291992, "learning_rate": 5.328968037769949e-05, "loss": 0.0549, "step": 50510 }, { "epoch": 1.7079684911592685, "grad_norm": 0.18925587832927704, "learning_rate": 5.327318132645042e-05, "loss": 0.0495, "step": 50520 }, { "epoch": 1.7083065688495216, "grad_norm": 0.3481404781341553, "learning_rate": 5.325668191725116e-05, "loss": 0.0451, "step": 50530 }, { "epoch": 1.7086446465397749, "grad_norm": 0.4141717553138733, "learning_rate": 5.324018215190606e-05, "loss": 0.0494, "step": 50540 }, { "epoch": 1.7089827242300282, "grad_norm": 0.3343644142150879, "learning_rate": 5.3223682032219515e-05, "loss": 0.0501, "step": 50550 }, { "epoch": 1.7093208019202812, "grad_norm": 0.24674180150032043, "learning_rate": 5.320718155999595e-05, "loss": 0.0512, "step": 50560 }, { "epoch": 1.7096588796105345, "grad_norm": 0.23430830240249634, "learning_rate": 5.3190680737039835e-05, "loss": 0.0559, "step": 50570 }, { "epoch": 1.7099969573007878, "grad_norm": 0.16879147291183472, "learning_rate": 5.317417956515567e-05, "loss": 0.0505, "step": 50580 }, { "epoch": 1.710335034991041, "grad_norm": 0.39059963822364807, "learning_rate": 5.315767804614798e-05, "loss": 0.049, "step": 50590 }, { "epoch": 1.7106731126812942, "grad_norm": 0.5215779542922974, "learning_rate": 5.3141176181821395e-05, "loss": 0.0509, "step": 50600 }, { "epoch": 1.7110111903715475, "grad_norm": 0.3673407733440399, "learning_rate": 5.31246739739805e-05, "loss": 0.0502, "step": 50610 }, { "epoch": 1.7113492680618005, "grad_norm": 0.2853563725948334, "learning_rate": 5.310817142442994e-05, "loss": 0.0502, "step": 50620 }, { "epoch": 1.7116873457520538, "grad_norm": 0.23891988396644592, "learning_rate": 5.309166853497445e-05, "loss": 0.0511, "step": 50630 }, { "epoch": 1.7120254234423071, "grad_norm": 0.5838897228240967, "learning_rate": 5.307516530741873e-05, "loss": 0.0543, "step": 50640 }, { "epoch": 1.7123635011325602, "grad_norm": 0.48292601108551025, "learning_rate": 5.305866174356754e-05, "loss": 0.0575, "step": 50650 }, { "epoch": 1.7127015788228135, "grad_norm": 0.2964670956134796, "learning_rate": 5.304215784522571e-05, "loss": 0.048, "step": 50660 }, { "epoch": 1.7130396565130668, "grad_norm": 0.1669609695672989, "learning_rate": 5.302565361419808e-05, "loss": 0.0542, "step": 50670 }, { "epoch": 1.7133777342033198, "grad_norm": 0.28107860684394836, "learning_rate": 5.3009149052289507e-05, "loss": 0.055, "step": 50680 }, { "epoch": 1.7137158118935731, "grad_norm": 0.29662996530532837, "learning_rate": 5.299264416130493e-05, "loss": 0.0489, "step": 50690 }, { "epoch": 1.7140538895838264, "grad_norm": 0.23905731737613678, "learning_rate": 5.297613894304928e-05, "loss": 0.054, "step": 50700 }, { "epoch": 1.7143919672740795, "grad_norm": 0.29263103008270264, "learning_rate": 5.2959633399327534e-05, "loss": 0.0508, "step": 50710 }, { "epoch": 1.7147300449643328, "grad_norm": 0.28624027967453003, "learning_rate": 5.294312753194476e-05, "loss": 0.0526, "step": 50720 }, { "epoch": 1.715068122654586, "grad_norm": 0.24833987653255463, "learning_rate": 5.292662134270596e-05, "loss": 0.0543, "step": 50730 }, { "epoch": 1.7154062003448391, "grad_norm": 0.3465258777141571, "learning_rate": 5.291011483341626e-05, "loss": 0.045, "step": 50740 }, { "epoch": 1.7157442780350924, "grad_norm": 0.3716975152492523, "learning_rate": 5.2893608005880767e-05, "loss": 0.0492, "step": 50750 }, { "epoch": 1.7160823557253457, "grad_norm": 0.3118753731250763, "learning_rate": 5.287710086190467e-05, "loss": 0.0479, "step": 50760 }, { "epoch": 1.7164204334155988, "grad_norm": 0.1645665466785431, "learning_rate": 5.286059340329314e-05, "loss": 0.049, "step": 50770 }, { "epoch": 1.716758511105852, "grad_norm": 0.19453175365924835, "learning_rate": 5.284408563185141e-05, "loss": 0.0496, "step": 50780 }, { "epoch": 1.7170965887961054, "grad_norm": 0.30400174856185913, "learning_rate": 5.282757754938476e-05, "loss": 0.0506, "step": 50790 }, { "epoch": 1.7174346664863585, "grad_norm": 0.4939267933368683, "learning_rate": 5.281106915769849e-05, "loss": 0.0505, "step": 50800 }, { "epoch": 1.7177727441766117, "grad_norm": 0.2531362473964691, "learning_rate": 5.2794560458597897e-05, "loss": 0.046, "step": 50810 }, { "epoch": 1.718110821866865, "grad_norm": 0.31937310099601746, "learning_rate": 5.27780514538884e-05, "loss": 0.0455, "step": 50820 }, { "epoch": 1.718448899557118, "grad_norm": 0.2832302749156952, "learning_rate": 5.2761542145375365e-05, "loss": 0.0527, "step": 50830 }, { "epoch": 1.7187869772473714, "grad_norm": 0.2194003015756607, "learning_rate": 5.274503253486421e-05, "loss": 0.0488, "step": 50840 }, { "epoch": 1.7191250549376247, "grad_norm": 0.3177240490913391, "learning_rate": 5.272852262416046e-05, "loss": 0.051, "step": 50850 }, { "epoch": 1.7194631326278778, "grad_norm": 0.6998414993286133, "learning_rate": 5.2712012415069555e-05, "loss": 0.0596, "step": 50860 }, { "epoch": 1.7198012103181313, "grad_norm": 0.1915493756532669, "learning_rate": 5.269550190939705e-05, "loss": 0.0475, "step": 50870 }, { "epoch": 1.7201392880083843, "grad_norm": 0.2631566822528839, "learning_rate": 5.267899110894852e-05, "loss": 0.05, "step": 50880 }, { "epoch": 1.7204773656986374, "grad_norm": 0.25850385427474976, "learning_rate": 5.266248001552955e-05, "loss": 0.0509, "step": 50890 }, { "epoch": 1.720815443388891, "grad_norm": 0.21881815791130066, "learning_rate": 5.264596863094575e-05, "loss": 0.0463, "step": 50900 }, { "epoch": 1.721153521079144, "grad_norm": 0.275564044713974, "learning_rate": 5.262945695700282e-05, "loss": 0.0477, "step": 50910 }, { "epoch": 1.721491598769397, "grad_norm": 0.29226207733154297, "learning_rate": 5.261294499550643e-05, "loss": 0.049, "step": 50920 }, { "epoch": 1.7218296764596506, "grad_norm": 0.22005389630794525, "learning_rate": 5.25964327482623e-05, "loss": 0.0499, "step": 50930 }, { "epoch": 1.7221677541499036, "grad_norm": 0.30553507804870605, "learning_rate": 5.257992021707617e-05, "loss": 0.0532, "step": 50940 }, { "epoch": 1.7225058318401567, "grad_norm": 0.20595861971378326, "learning_rate": 5.256340740375387e-05, "loss": 0.0503, "step": 50950 }, { "epoch": 1.7228439095304102, "grad_norm": 0.28133198618888855, "learning_rate": 5.254689431010117e-05, "loss": 0.0436, "step": 50960 }, { "epoch": 1.7231819872206633, "grad_norm": 0.36864182353019714, "learning_rate": 5.253038093792395e-05, "loss": 0.049, "step": 50970 }, { "epoch": 1.7235200649109166, "grad_norm": 0.2652837038040161, "learning_rate": 5.251386728902806e-05, "loss": 0.0523, "step": 50980 }, { "epoch": 1.7238581426011699, "grad_norm": 0.28450241684913635, "learning_rate": 5.2497353365219446e-05, "loss": 0.0501, "step": 50990 }, { "epoch": 1.724196220291423, "grad_norm": 0.18087172508239746, "learning_rate": 5.2480839168304e-05, "loss": 0.0524, "step": 51000 }, { "epoch": 1.7245342979816762, "grad_norm": 0.2904239594936371, "learning_rate": 5.246432470008773e-05, "loss": 0.047, "step": 51010 }, { "epoch": 1.7248723756719295, "grad_norm": 0.2538665533065796, "learning_rate": 5.2447809962376616e-05, "loss": 0.0494, "step": 51020 }, { "epoch": 1.7252104533621826, "grad_norm": 0.2824520468711853, "learning_rate": 5.243129495697666e-05, "loss": 0.053, "step": 51030 }, { "epoch": 1.7255485310524359, "grad_norm": 0.40925225615501404, "learning_rate": 5.241477968569397e-05, "loss": 0.0477, "step": 51040 }, { "epoch": 1.7258866087426892, "grad_norm": 0.1774028092622757, "learning_rate": 5.23982641503346e-05, "loss": 0.0512, "step": 51050 }, { "epoch": 1.7262246864329422, "grad_norm": 0.304525226354599, "learning_rate": 5.238174835270465e-05, "loss": 0.0522, "step": 51060 }, { "epoch": 1.7265627641231955, "grad_norm": 0.21619316935539246, "learning_rate": 5.23652322946103e-05, "loss": 0.0464, "step": 51070 }, { "epoch": 1.7269008418134488, "grad_norm": 0.1908291131258011, "learning_rate": 5.234871597785771e-05, "loss": 0.0507, "step": 51080 }, { "epoch": 1.727238919503702, "grad_norm": 0.1446256786584854, "learning_rate": 5.233219940425306e-05, "loss": 0.0526, "step": 51090 }, { "epoch": 1.7275769971939552, "grad_norm": 0.17800255119800568, "learning_rate": 5.2315682575602576e-05, "loss": 0.0433, "step": 51100 }, { "epoch": 1.7279150748842085, "grad_norm": 0.207963228225708, "learning_rate": 5.229916549371254e-05, "loss": 0.0554, "step": 51110 }, { "epoch": 1.7282531525744615, "grad_norm": 0.2858288884162903, "learning_rate": 5.228264816038924e-05, "loss": 0.0484, "step": 51120 }, { "epoch": 1.7285912302647148, "grad_norm": 0.19945567846298218, "learning_rate": 5.2266130577438935e-05, "loss": 0.0467, "step": 51130 }, { "epoch": 1.7289293079549681, "grad_norm": 0.2957676649093628, "learning_rate": 5.224961274666802e-05, "loss": 0.0526, "step": 51140 }, { "epoch": 1.7292673856452212, "grad_norm": 0.32403603196144104, "learning_rate": 5.223309466988283e-05, "loss": 0.0529, "step": 51150 }, { "epoch": 1.7296054633354745, "grad_norm": 0.33242639899253845, "learning_rate": 5.221657634888976e-05, "loss": 0.051, "step": 51160 }, { "epoch": 1.7299435410257278, "grad_norm": 0.2021932452917099, "learning_rate": 5.2200057785495235e-05, "loss": 0.0532, "step": 51170 }, { "epoch": 1.7302816187159809, "grad_norm": 0.2599889934062958, "learning_rate": 5.2183538981505706e-05, "loss": 0.0551, "step": 51180 }, { "epoch": 1.7306196964062341, "grad_norm": 0.20542411506175995, "learning_rate": 5.216701993872762e-05, "loss": 0.0512, "step": 51190 }, { "epoch": 1.7309577740964874, "grad_norm": 0.4985724687576294, "learning_rate": 5.215050065896752e-05, "loss": 0.0517, "step": 51200 }, { "epoch": 1.7312958517867405, "grad_norm": 0.5081021785736084, "learning_rate": 5.2133981144031875e-05, "loss": 0.0452, "step": 51210 }, { "epoch": 1.7316339294769938, "grad_norm": 0.3707551658153534, "learning_rate": 5.211746139572727e-05, "loss": 0.0536, "step": 51220 }, { "epoch": 1.731972007167247, "grad_norm": 0.26527443528175354, "learning_rate": 5.210094141586026e-05, "loss": 0.0534, "step": 51230 }, { "epoch": 1.7323100848575002, "grad_norm": 0.16835036873817444, "learning_rate": 5.2084421206237476e-05, "loss": 0.0488, "step": 51240 }, { "epoch": 1.7326481625477534, "grad_norm": 0.16987088322639465, "learning_rate": 5.20679007686655e-05, "loss": 0.0505, "step": 51250 }, { "epoch": 1.7329862402380067, "grad_norm": 0.18901896476745605, "learning_rate": 5.205138010495102e-05, "loss": 0.0495, "step": 51260 }, { "epoch": 1.7333243179282598, "grad_norm": 0.28195369243621826, "learning_rate": 5.20348592169007e-05, "loss": 0.0484, "step": 51270 }, { "epoch": 1.733662395618513, "grad_norm": 0.42605677247047424, "learning_rate": 5.201833810632123e-05, "loss": 0.0507, "step": 51280 }, { "epoch": 1.7340004733087664, "grad_norm": 0.21179841458797455, "learning_rate": 5.200181677501933e-05, "loss": 0.0482, "step": 51290 }, { "epoch": 1.7343385509990195, "grad_norm": 0.4755268394947052, "learning_rate": 5.1985295224801777e-05, "loss": 0.0477, "step": 51300 }, { "epoch": 1.7346766286892727, "grad_norm": 0.22394385933876038, "learning_rate": 5.196877345747531e-05, "loss": 0.0521, "step": 51310 }, { "epoch": 1.735014706379526, "grad_norm": 0.3606009781360626, "learning_rate": 5.195225147484674e-05, "loss": 0.0493, "step": 51320 }, { "epoch": 1.735352784069779, "grad_norm": 0.27173277735710144, "learning_rate": 5.193572927872288e-05, "loss": 0.0484, "step": 51330 }, { "epoch": 1.7356908617600326, "grad_norm": 0.2058146446943283, "learning_rate": 5.191920687091059e-05, "loss": 0.0507, "step": 51340 }, { "epoch": 1.7360289394502857, "grad_norm": 0.34884610772132874, "learning_rate": 5.1902684253216695e-05, "loss": 0.0508, "step": 51350 }, { "epoch": 1.7363670171405388, "grad_norm": 0.19587276875972748, "learning_rate": 5.1886161427448134e-05, "loss": 0.0489, "step": 51360 }, { "epoch": 1.7367050948307923, "grad_norm": 0.3558475971221924, "learning_rate": 5.186963839541179e-05, "loss": 0.0467, "step": 51370 }, { "epoch": 1.7370431725210453, "grad_norm": 0.1117275133728981, "learning_rate": 5.185311515891459e-05, "loss": 0.0511, "step": 51380 }, { "epoch": 1.7373812502112984, "grad_norm": 0.18755225837230682, "learning_rate": 5.1836591719763516e-05, "loss": 0.0501, "step": 51390 }, { "epoch": 1.737719327901552, "grad_norm": 0.20797036588191986, "learning_rate": 5.182006807976554e-05, "loss": 0.0467, "step": 51400 }, { "epoch": 1.738057405591805, "grad_norm": 0.2745417356491089, "learning_rate": 5.180354424072763e-05, "loss": 0.0494, "step": 51410 }, { "epoch": 1.738395483282058, "grad_norm": 0.3631114661693573, "learning_rate": 5.178702020445686e-05, "loss": 0.0499, "step": 51420 }, { "epoch": 1.7387335609723116, "grad_norm": 0.25265178084373474, "learning_rate": 5.177049597276024e-05, "loss": 0.0549, "step": 51430 }, { "epoch": 1.7390716386625646, "grad_norm": 0.18192490935325623, "learning_rate": 5.175397154744482e-05, "loss": 0.0491, "step": 51440 }, { "epoch": 1.7394097163528177, "grad_norm": 0.2662868797779083, "learning_rate": 5.1737446930317736e-05, "loss": 0.0501, "step": 51450 }, { "epoch": 1.7397477940430712, "grad_norm": 0.2614988386631012, "learning_rate": 5.172092212318604e-05, "loss": 0.05, "step": 51460 }, { "epoch": 1.7400858717333243, "grad_norm": 0.24596451222896576, "learning_rate": 5.170439712785692e-05, "loss": 0.0476, "step": 51470 }, { "epoch": 1.7404239494235776, "grad_norm": 0.16356469690799713, "learning_rate": 5.1687871946137467e-05, "loss": 0.0488, "step": 51480 }, { "epoch": 1.7407620271138309, "grad_norm": 0.17966662347316742, "learning_rate": 5.16713465798349e-05, "loss": 0.0508, "step": 51490 }, { "epoch": 1.741100104804084, "grad_norm": 0.22833925485610962, "learning_rate": 5.165482103075637e-05, "loss": 0.0511, "step": 51500 }, { "epoch": 1.7414381824943372, "grad_norm": 0.24285858869552612, "learning_rate": 5.163829530070909e-05, "loss": 0.0506, "step": 51510 }, { "epoch": 1.7417762601845905, "grad_norm": 0.2648835778236389, "learning_rate": 5.162176939150032e-05, "loss": 0.0461, "step": 51520 }, { "epoch": 1.7421143378748436, "grad_norm": 0.3085269629955292, "learning_rate": 5.160524330493728e-05, "loss": 0.0542, "step": 51530 }, { "epoch": 1.742452415565097, "grad_norm": 0.21426159143447876, "learning_rate": 5.158871704282724e-05, "loss": 0.0548, "step": 51540 }, { "epoch": 1.7427904932553502, "grad_norm": 0.3271176815032959, "learning_rate": 5.15721906069775e-05, "loss": 0.0532, "step": 51550 }, { "epoch": 1.7431285709456033, "grad_norm": 0.24200263619422913, "learning_rate": 5.155566399919536e-05, "loss": 0.0529, "step": 51560 }, { "epoch": 1.7434666486358565, "grad_norm": 0.3237006366252899, "learning_rate": 5.153913722128813e-05, "loss": 0.0529, "step": 51570 }, { "epoch": 1.7438047263261098, "grad_norm": 0.28321242332458496, "learning_rate": 5.1522610275063176e-05, "loss": 0.051, "step": 51580 }, { "epoch": 1.744142804016363, "grad_norm": 0.2784445285797119, "learning_rate": 5.150608316232786e-05, "loss": 0.0467, "step": 51590 }, { "epoch": 1.7444808817066162, "grad_norm": 0.34794872999191284, "learning_rate": 5.148955588488955e-05, "loss": 0.058, "step": 51600 }, { "epoch": 1.7448189593968695, "grad_norm": 0.1837255209684372, "learning_rate": 5.147302844455564e-05, "loss": 0.0563, "step": 51610 }, { "epoch": 1.7451570370871226, "grad_norm": 0.3086700439453125, "learning_rate": 5.145650084313356e-05, "loss": 0.0471, "step": 51620 }, { "epoch": 1.7454951147773758, "grad_norm": 0.37276604771614075, "learning_rate": 5.1439973082430734e-05, "loss": 0.054, "step": 51630 }, { "epoch": 1.7458331924676291, "grad_norm": 0.17654463648796082, "learning_rate": 5.142344516425461e-05, "loss": 0.055, "step": 51640 }, { "epoch": 1.7461712701578822, "grad_norm": 0.19341841340065002, "learning_rate": 5.1406917090412673e-05, "loss": 0.051, "step": 51650 }, { "epoch": 1.7465093478481355, "grad_norm": 0.1619529128074646, "learning_rate": 5.1390388862712404e-05, "loss": 0.0516, "step": 51660 }, { "epoch": 1.7468474255383888, "grad_norm": 0.30329591035842896, "learning_rate": 5.137386048296129e-05, "loss": 0.0541, "step": 51670 }, { "epoch": 1.7471855032286419, "grad_norm": 0.25831884145736694, "learning_rate": 5.135733195296686e-05, "loss": 0.05, "step": 51680 }, { "epoch": 1.7475235809188951, "grad_norm": 0.20855146646499634, "learning_rate": 5.134080327453664e-05, "loss": 0.0463, "step": 51690 }, { "epoch": 1.7478616586091484, "grad_norm": 0.2771151661872864, "learning_rate": 5.1324274449478204e-05, "loss": 0.0464, "step": 51700 }, { "epoch": 1.7481997362994015, "grad_norm": 0.3102002739906311, "learning_rate": 5.130774547959909e-05, "loss": 0.055, "step": 51710 }, { "epoch": 1.7485378139896548, "grad_norm": 0.28375545144081116, "learning_rate": 5.129121636670692e-05, "loss": 0.0539, "step": 51720 }, { "epoch": 1.748875891679908, "grad_norm": 0.19430752098560333, "learning_rate": 5.1274687112609246e-05, "loss": 0.0558, "step": 51730 }, { "epoch": 1.7492139693701612, "grad_norm": 0.21141815185546875, "learning_rate": 5.125815771911372e-05, "loss": 0.053, "step": 51740 }, { "epoch": 1.7495520470604145, "grad_norm": 0.17117954790592194, "learning_rate": 5.1241628188027956e-05, "loss": 0.0533, "step": 51750 }, { "epoch": 1.7498901247506677, "grad_norm": 0.2685895562171936, "learning_rate": 5.1225098521159585e-05, "loss": 0.0487, "step": 51760 }, { "epoch": 1.7502282024409208, "grad_norm": 0.2800663709640503, "learning_rate": 5.12085687203163e-05, "loss": 0.0507, "step": 51770 }, { "epoch": 1.750566280131174, "grad_norm": 0.36716192960739136, "learning_rate": 5.119203878730575e-05, "loss": 0.0536, "step": 51780 }, { "epoch": 1.7509043578214274, "grad_norm": 0.31561458110809326, "learning_rate": 5.117550872393564e-05, "loss": 0.0532, "step": 51790 }, { "epoch": 1.7512424355116805, "grad_norm": 0.2232443243265152, "learning_rate": 5.1158978532013645e-05, "loss": 0.049, "step": 51800 }, { "epoch": 1.7515805132019338, "grad_norm": 0.21431010961532593, "learning_rate": 5.114244821334751e-05, "loss": 0.0481, "step": 51810 }, { "epoch": 1.751918590892187, "grad_norm": 0.28450390696525574, "learning_rate": 5.112591776974497e-05, "loss": 0.051, "step": 51820 }, { "epoch": 1.7522566685824401, "grad_norm": 0.2846607267856598, "learning_rate": 5.110938720301374e-05, "loss": 0.0435, "step": 51830 }, { "epoch": 1.7525947462726936, "grad_norm": 0.1532105654478073, "learning_rate": 5.1092856514961595e-05, "loss": 0.0528, "step": 51840 }, { "epoch": 1.7529328239629467, "grad_norm": 0.16230586171150208, "learning_rate": 5.107632570739631e-05, "loss": 0.0515, "step": 51850 }, { "epoch": 1.7532709016531998, "grad_norm": 0.24141409993171692, "learning_rate": 5.1059794782125645e-05, "loss": 0.045, "step": 51860 }, { "epoch": 1.7536089793434533, "grad_norm": 0.20667022466659546, "learning_rate": 5.104326374095744e-05, "loss": 0.0486, "step": 51870 }, { "epoch": 1.7539470570337063, "grad_norm": 0.2674682140350342, "learning_rate": 5.102673258569947e-05, "loss": 0.051, "step": 51880 }, { "epoch": 1.7542851347239594, "grad_norm": 0.20111516118049622, "learning_rate": 5.101020131815956e-05, "loss": 0.0432, "step": 51890 }, { "epoch": 1.754623212414213, "grad_norm": 0.20822089910507202, "learning_rate": 5.099366994014556e-05, "loss": 0.0513, "step": 51900 }, { "epoch": 1.754961290104466, "grad_norm": 0.25790977478027344, "learning_rate": 5.097713845346532e-05, "loss": 0.0487, "step": 51910 }, { "epoch": 1.755299367794719, "grad_norm": 0.23435823619365692, "learning_rate": 5.096060685992668e-05, "loss": 0.0463, "step": 51920 }, { "epoch": 1.7556374454849726, "grad_norm": 0.20560356974601746, "learning_rate": 5.0944075161337514e-05, "loss": 0.0504, "step": 51930 }, { "epoch": 1.7559755231752256, "grad_norm": 0.21941526234149933, "learning_rate": 5.09275433595057e-05, "loss": 0.0476, "step": 51940 }, { "epoch": 1.756313600865479, "grad_norm": 0.2112642228603363, "learning_rate": 5.0911011456239157e-05, "loss": 0.05, "step": 51950 }, { "epoch": 1.7566516785557322, "grad_norm": 0.23349694907665253, "learning_rate": 5.089447945334574e-05, "loss": 0.0463, "step": 51960 }, { "epoch": 1.7569897562459853, "grad_norm": 0.2179936319589615, "learning_rate": 5.087794735263343e-05, "loss": 0.0571, "step": 51970 }, { "epoch": 1.7573278339362386, "grad_norm": 0.21555079519748688, "learning_rate": 5.086141515591011e-05, "loss": 0.0549, "step": 51980 }, { "epoch": 1.7576659116264919, "grad_norm": 0.33881667256355286, "learning_rate": 5.084488286498371e-05, "loss": 0.051, "step": 51990 }, { "epoch": 1.758003989316745, "grad_norm": 0.21658073365688324, "learning_rate": 5.08283504816622e-05, "loss": 0.0524, "step": 52000 }, { "epoch": 1.7583420670069982, "grad_norm": 0.21882271766662598, "learning_rate": 5.081181800775353e-05, "loss": 0.0511, "step": 52010 }, { "epoch": 1.7586801446972515, "grad_norm": 0.19983068108558655, "learning_rate": 5.079528544506565e-05, "loss": 0.046, "step": 52020 }, { "epoch": 1.7590182223875046, "grad_norm": 0.2678769528865814, "learning_rate": 5.077875279540657e-05, "loss": 0.0474, "step": 52030 }, { "epoch": 1.759356300077758, "grad_norm": 0.39806264638900757, "learning_rate": 5.076222006058424e-05, "loss": 0.0533, "step": 52040 }, { "epoch": 1.7596943777680112, "grad_norm": 0.6667680740356445, "learning_rate": 5.074568724240667e-05, "loss": 0.0543, "step": 52050 }, { "epoch": 1.7600324554582643, "grad_norm": 0.5924772620201111, "learning_rate": 5.072915434268187e-05, "loss": 0.0503, "step": 52060 }, { "epoch": 1.7603705331485175, "grad_norm": 0.23835231363773346, "learning_rate": 5.071262136321785e-05, "loss": 0.0494, "step": 52070 }, { "epoch": 1.7607086108387708, "grad_norm": 0.17668788135051727, "learning_rate": 5.069608830582262e-05, "loss": 0.0496, "step": 52080 }, { "epoch": 1.761046688529024, "grad_norm": 0.20485606789588928, "learning_rate": 5.067955517230423e-05, "loss": 0.0509, "step": 52090 }, { "epoch": 1.7613847662192772, "grad_norm": 0.3239727318286896, "learning_rate": 5.06630219644707e-05, "loss": 0.0476, "step": 52100 }, { "epoch": 1.7617228439095305, "grad_norm": 0.16699719429016113, "learning_rate": 5.064648868413008e-05, "loss": 0.0536, "step": 52110 }, { "epoch": 1.7620609215997836, "grad_norm": 0.18284666538238525, "learning_rate": 5.062995533309044e-05, "loss": 0.0504, "step": 52120 }, { "epoch": 1.7623989992900368, "grad_norm": 0.26306724548339844, "learning_rate": 5.061342191315983e-05, "loss": 0.0467, "step": 52130 }, { "epoch": 1.7627370769802901, "grad_norm": 0.22249677777290344, "learning_rate": 5.0596888426146325e-05, "loss": 0.0494, "step": 52140 }, { "epoch": 1.7630751546705432, "grad_norm": 0.2695463299751282, "learning_rate": 5.058035487385798e-05, "loss": 0.0551, "step": 52150 }, { "epoch": 1.7634132323607965, "grad_norm": 0.16800521314144135, "learning_rate": 5.0563821258102915e-05, "loss": 0.0483, "step": 52160 }, { "epoch": 1.7637513100510498, "grad_norm": 0.38513728976249695, "learning_rate": 5.05472875806892e-05, "loss": 0.0532, "step": 52170 }, { "epoch": 1.7640893877413029, "grad_norm": 0.21064110100269318, "learning_rate": 5.053075384342494e-05, "loss": 0.0508, "step": 52180 }, { "epoch": 1.7644274654315562, "grad_norm": 0.19888073205947876, "learning_rate": 5.0514220048118234e-05, "loss": 0.0499, "step": 52190 }, { "epoch": 1.7647655431218094, "grad_norm": 0.2837306559085846, "learning_rate": 5.049768619657721e-05, "loss": 0.0549, "step": 52200 }, { "epoch": 1.7651036208120625, "grad_norm": 0.2739114761352539, "learning_rate": 5.048115229060996e-05, "loss": 0.0543, "step": 52210 }, { "epoch": 1.7654416985023158, "grad_norm": 0.20082709193229675, "learning_rate": 5.046461833202462e-05, "loss": 0.0519, "step": 52220 }, { "epoch": 1.765779776192569, "grad_norm": 0.2654235363006592, "learning_rate": 5.044808432262932e-05, "loss": 0.0546, "step": 52230 }, { "epoch": 1.7661178538828222, "grad_norm": 0.21104127168655396, "learning_rate": 5.043155026423219e-05, "loss": 0.0468, "step": 52240 }, { "epoch": 1.7664559315730755, "grad_norm": 0.26768624782562256, "learning_rate": 5.041501615864137e-05, "loss": 0.048, "step": 52250 }, { "epoch": 1.7667940092633287, "grad_norm": 0.3016619384288788, "learning_rate": 5.039848200766502e-05, "loss": 0.0517, "step": 52260 }, { "epoch": 1.7671320869535818, "grad_norm": 0.2657352685928345, "learning_rate": 5.038194781311125e-05, "loss": 0.0492, "step": 52270 }, { "epoch": 1.767470164643835, "grad_norm": 0.3009538948535919, "learning_rate": 5.0365413576788264e-05, "loss": 0.0556, "step": 52280 }, { "epoch": 1.7678082423340884, "grad_norm": 0.17529281973838806, "learning_rate": 5.0348879300504194e-05, "loss": 0.0527, "step": 52290 }, { "epoch": 1.7681463200243415, "grad_norm": 0.20165680348873138, "learning_rate": 5.033234498606721e-05, "loss": 0.0515, "step": 52300 }, { "epoch": 1.768484397714595, "grad_norm": 0.1453322023153305, "learning_rate": 5.0315810635285476e-05, "loss": 0.049, "step": 52310 }, { "epoch": 1.768822475404848, "grad_norm": 0.4690219461917877, "learning_rate": 5.0299276249967164e-05, "loss": 0.0476, "step": 52320 }, { "epoch": 1.7691605530951011, "grad_norm": 0.2292616069316864, "learning_rate": 5.0282741831920454e-05, "loss": 0.0505, "step": 52330 }, { "epoch": 1.7694986307853546, "grad_norm": 0.2752959430217743, "learning_rate": 5.026620738295351e-05, "loss": 0.0481, "step": 52340 }, { "epoch": 1.7698367084756077, "grad_norm": 0.4117007553577423, "learning_rate": 5.024967290487455e-05, "loss": 0.051, "step": 52350 }, { "epoch": 1.7701747861658608, "grad_norm": 0.27302756905555725, "learning_rate": 5.023313839949172e-05, "loss": 0.0558, "step": 52360 }, { "epoch": 1.7705128638561143, "grad_norm": 0.27054640650749207, "learning_rate": 5.0216603868613224e-05, "loss": 0.045, "step": 52370 }, { "epoch": 1.7708509415463674, "grad_norm": 0.19991327822208405, "learning_rate": 5.020006931404726e-05, "loss": 0.0508, "step": 52380 }, { "epoch": 1.7711890192366204, "grad_norm": 0.20086002349853516, "learning_rate": 5.0183534737602014e-05, "loss": 0.0458, "step": 52390 }, { "epoch": 1.771527096926874, "grad_norm": 0.3093338906764984, "learning_rate": 5.0167000141085686e-05, "loss": 0.0559, "step": 52400 }, { "epoch": 1.771865174617127, "grad_norm": 0.19004911184310913, "learning_rate": 5.015046552630647e-05, "loss": 0.0463, "step": 52410 }, { "epoch": 1.77220325230738, "grad_norm": 0.3056585192680359, "learning_rate": 5.013393089507258e-05, "loss": 0.0541, "step": 52420 }, { "epoch": 1.7725413299976336, "grad_norm": 0.20447659492492676, "learning_rate": 5.011739624919219e-05, "loss": 0.0497, "step": 52430 }, { "epoch": 1.7728794076878867, "grad_norm": 0.281510591506958, "learning_rate": 5.0100861590473545e-05, "loss": 0.0513, "step": 52440 }, { "epoch": 1.77321748537814, "grad_norm": 0.25120967626571655, "learning_rate": 5.008432692072482e-05, "loss": 0.0517, "step": 52450 }, { "epoch": 1.7735555630683932, "grad_norm": 0.323747843503952, "learning_rate": 5.006779224175423e-05, "loss": 0.0487, "step": 52460 }, { "epoch": 1.7738936407586463, "grad_norm": 0.22103522717952728, "learning_rate": 5.005125755536998e-05, "loss": 0.0487, "step": 52470 }, { "epoch": 1.7742317184488996, "grad_norm": 0.21783222258090973, "learning_rate": 5.003472286338029e-05, "loss": 0.0468, "step": 52480 }, { "epoch": 1.7745697961391529, "grad_norm": 0.2239476889371872, "learning_rate": 5.001818816759335e-05, "loss": 0.0477, "step": 52490 }, { "epoch": 1.774907873829406, "grad_norm": 0.31048649549484253, "learning_rate": 5.000165346981738e-05, "loss": 0.0476, "step": 52500 }, { "epoch": 1.7752459515196592, "grad_norm": 0.3306252360343933, "learning_rate": 4.9985118771860595e-05, "loss": 0.052, "step": 52510 }, { "epoch": 1.7755840292099125, "grad_norm": 0.28960973024368286, "learning_rate": 4.996858407553119e-05, "loss": 0.0516, "step": 52520 }, { "epoch": 1.7759221069001656, "grad_norm": 0.321113258600235, "learning_rate": 4.99520493826374e-05, "loss": 0.0506, "step": 52530 }, { "epoch": 1.776260184590419, "grad_norm": 0.162899911403656, "learning_rate": 4.9935514694987384e-05, "loss": 0.0453, "step": 52540 }, { "epoch": 1.7765982622806722, "grad_norm": 0.2536066174507141, "learning_rate": 4.991898001438941e-05, "loss": 0.0519, "step": 52550 }, { "epoch": 1.7769363399709253, "grad_norm": 0.2469719648361206, "learning_rate": 4.990244534265166e-05, "loss": 0.0526, "step": 52560 }, { "epoch": 1.7772744176611786, "grad_norm": 0.17859482765197754, "learning_rate": 4.988591068158234e-05, "loss": 0.0481, "step": 52570 }, { "epoch": 1.7776124953514318, "grad_norm": 0.2027435004711151, "learning_rate": 4.986937603298965e-05, "loss": 0.0518, "step": 52580 }, { "epoch": 1.777950573041685, "grad_norm": 0.2865588068962097, "learning_rate": 4.985284139868181e-05, "loss": 0.0518, "step": 52590 }, { "epoch": 1.7782886507319382, "grad_norm": 0.27379778027534485, "learning_rate": 4.9836306780467e-05, "loss": 0.0509, "step": 52600 }, { "epoch": 1.7786267284221915, "grad_norm": 0.3018706440925598, "learning_rate": 4.981977218015343e-05, "loss": 0.0509, "step": 52610 }, { "epoch": 1.7789648061124446, "grad_norm": 0.33421972393989563, "learning_rate": 4.980323759954932e-05, "loss": 0.0523, "step": 52620 }, { "epoch": 1.7793028838026979, "grad_norm": 0.1695573925971985, "learning_rate": 4.978670304046284e-05, "loss": 0.0503, "step": 52630 }, { "epoch": 1.7796409614929511, "grad_norm": 0.19120816886425018, "learning_rate": 4.9770168504702195e-05, "loss": 0.0549, "step": 52640 }, { "epoch": 1.7799790391832042, "grad_norm": 0.2936623990535736, "learning_rate": 4.975363399407559e-05, "loss": 0.047, "step": 52650 }, { "epoch": 1.7803171168734575, "grad_norm": 0.2355411797761917, "learning_rate": 4.9737099510391194e-05, "loss": 0.0506, "step": 52660 }, { "epoch": 1.7806551945637108, "grad_norm": 0.34498482942581177, "learning_rate": 4.972056505545719e-05, "loss": 0.0463, "step": 52670 }, { "epoch": 1.7809932722539639, "grad_norm": 0.19521883130073547, "learning_rate": 4.9704030631081775e-05, "loss": 0.0462, "step": 52680 }, { "epoch": 1.7813313499442172, "grad_norm": 0.14637750387191772, "learning_rate": 4.968749623907313e-05, "loss": 0.0474, "step": 52690 }, { "epoch": 1.7816694276344704, "grad_norm": 0.21326716244220734, "learning_rate": 4.967096188123943e-05, "loss": 0.0531, "step": 52700 }, { "epoch": 1.7820075053247235, "grad_norm": 0.1686764657497406, "learning_rate": 4.965442755938884e-05, "loss": 0.0504, "step": 52710 }, { "epoch": 1.7823455830149768, "grad_norm": 0.15621337294578552, "learning_rate": 4.963789327532954e-05, "loss": 0.0497, "step": 52720 }, { "epoch": 1.78268366070523, "grad_norm": 0.7257527112960815, "learning_rate": 4.962135903086968e-05, "loss": 0.0531, "step": 52730 }, { "epoch": 1.7830217383954832, "grad_norm": 0.20042771100997925, "learning_rate": 4.960482482781741e-05, "loss": 0.0546, "step": 52740 }, { "epoch": 1.7833598160857365, "grad_norm": 0.19406336545944214, "learning_rate": 4.958829066798092e-05, "loss": 0.049, "step": 52750 }, { "epoch": 1.7836978937759898, "grad_norm": 0.40526798367500305, "learning_rate": 4.957175655316835e-05, "loss": 0.0508, "step": 52760 }, { "epoch": 1.7840359714662428, "grad_norm": 0.4144473075866699, "learning_rate": 4.9555222485187834e-05, "loss": 0.0513, "step": 52770 }, { "epoch": 1.7843740491564961, "grad_norm": 0.3313446342945099, "learning_rate": 4.953868846584752e-05, "loss": 0.0576, "step": 52780 }, { "epoch": 1.7847121268467494, "grad_norm": 0.19260801374912262, "learning_rate": 4.9522154496955544e-05, "loss": 0.0474, "step": 52790 }, { "epoch": 1.7850502045370025, "grad_norm": 0.25384223461151123, "learning_rate": 4.950562058032001e-05, "loss": 0.0531, "step": 52800 }, { "epoch": 1.785388282227256, "grad_norm": 0.30256184935569763, "learning_rate": 4.948908671774909e-05, "loss": 0.0519, "step": 52810 }, { "epoch": 1.785726359917509, "grad_norm": 0.2520129084587097, "learning_rate": 4.947255291105087e-05, "loss": 0.0594, "step": 52820 }, { "epoch": 1.7860644376077621, "grad_norm": 0.44114401936531067, "learning_rate": 4.945601916203348e-05, "loss": 0.0475, "step": 52830 }, { "epoch": 1.7864025152980156, "grad_norm": 0.29026997089385986, "learning_rate": 4.9439485472505016e-05, "loss": 0.0529, "step": 52840 }, { "epoch": 1.7867405929882687, "grad_norm": 0.19008022546768188, "learning_rate": 4.942295184427356e-05, "loss": 0.0484, "step": 52850 }, { "epoch": 1.7870786706785218, "grad_norm": 0.28232336044311523, "learning_rate": 4.9406418279147234e-05, "loss": 0.0497, "step": 52860 }, { "epoch": 1.7874167483687753, "grad_norm": 0.20268312096595764, "learning_rate": 4.9389884778934116e-05, "loss": 0.0483, "step": 52870 }, { "epoch": 1.7877548260590284, "grad_norm": 0.340220183134079, "learning_rate": 4.937335134544228e-05, "loss": 0.0506, "step": 52880 }, { "epoch": 1.7880929037492814, "grad_norm": 0.18962128460407257, "learning_rate": 4.935681798047979e-05, "loss": 0.0481, "step": 52890 }, { "epoch": 1.788430981439535, "grad_norm": 0.25615113973617554, "learning_rate": 4.934028468585473e-05, "loss": 0.0517, "step": 52900 }, { "epoch": 1.788769059129788, "grad_norm": 0.355262815952301, "learning_rate": 4.932375146337514e-05, "loss": 0.0534, "step": 52910 }, { "epoch": 1.7891071368200413, "grad_norm": 0.18748335540294647, "learning_rate": 4.930721831484908e-05, "loss": 0.0425, "step": 52920 }, { "epoch": 1.7894452145102946, "grad_norm": 0.32092538475990295, "learning_rate": 4.929068524208456e-05, "loss": 0.0454, "step": 52930 }, { "epoch": 1.7897832922005477, "grad_norm": 0.23445408046245575, "learning_rate": 4.927415224688965e-05, "loss": 0.0518, "step": 52940 }, { "epoch": 1.790121369890801, "grad_norm": 0.20706906914710999, "learning_rate": 4.9257619331072375e-05, "loss": 0.049, "step": 52950 }, { "epoch": 1.7904594475810542, "grad_norm": 0.3012883961200714, "learning_rate": 4.924108649644072e-05, "loss": 0.0502, "step": 52960 }, { "epoch": 1.7907975252713073, "grad_norm": 0.3200875520706177, "learning_rate": 4.922455374480271e-05, "loss": 0.0467, "step": 52970 }, { "epoch": 1.7911356029615606, "grad_norm": 0.18109817802906036, "learning_rate": 4.920802107796634e-05, "loss": 0.0476, "step": 52980 }, { "epoch": 1.791473680651814, "grad_norm": 0.1746438890695572, "learning_rate": 4.9191488497739564e-05, "loss": 0.0486, "step": 52990 }, { "epoch": 1.791811758342067, "grad_norm": 0.2736721336841583, "learning_rate": 4.917495600593042e-05, "loss": 0.0471, "step": 53000 }, { "epoch": 1.7921498360323203, "grad_norm": 0.28628525137901306, "learning_rate": 4.915842360434684e-05, "loss": 0.0466, "step": 53010 }, { "epoch": 1.7924879137225735, "grad_norm": 0.2079821079969406, "learning_rate": 4.914189129479679e-05, "loss": 0.0495, "step": 53020 }, { "epoch": 1.7928259914128266, "grad_norm": 0.2661987543106079, "learning_rate": 4.912535907908822e-05, "loss": 0.0457, "step": 53030 }, { "epoch": 1.79316406910308, "grad_norm": 0.26114606857299805, "learning_rate": 4.910882695902907e-05, "loss": 0.0477, "step": 53040 }, { "epoch": 1.7935021467933332, "grad_norm": 0.3069137632846832, "learning_rate": 4.9092294936427234e-05, "loss": 0.0486, "step": 53050 }, { "epoch": 1.7938402244835863, "grad_norm": 0.22539348900318146, "learning_rate": 4.9075763013090685e-05, "loss": 0.0516, "step": 53060 }, { "epoch": 1.7941783021738396, "grad_norm": 0.3730063736438751, "learning_rate": 4.9059231190827305e-05, "loss": 0.0571, "step": 53070 }, { "epoch": 1.7945163798640928, "grad_norm": 0.24196502566337585, "learning_rate": 4.904269947144499e-05, "loss": 0.0506, "step": 53080 }, { "epoch": 1.794854457554346, "grad_norm": 0.22031353414058685, "learning_rate": 4.902616785675161e-05, "loss": 0.0468, "step": 53090 }, { "epoch": 1.7951925352445992, "grad_norm": 0.33517953753471375, "learning_rate": 4.900963634855505e-05, "loss": 0.0482, "step": 53100 }, { "epoch": 1.7955306129348525, "grad_norm": 0.2591230273246765, "learning_rate": 4.8993104948663173e-05, "loss": 0.0531, "step": 53110 }, { "epoch": 1.7958686906251056, "grad_norm": 0.2427440732717514, "learning_rate": 4.8976573658883823e-05, "loss": 0.0477, "step": 53120 }, { "epoch": 1.7962067683153589, "grad_norm": 0.15082165598869324, "learning_rate": 4.896004248102484e-05, "loss": 0.0496, "step": 53130 }, { "epoch": 1.7965448460056122, "grad_norm": 0.20572017133235931, "learning_rate": 4.8943511416894045e-05, "loss": 0.0477, "step": 53140 }, { "epoch": 1.7968829236958652, "grad_norm": 0.23668983578681946, "learning_rate": 4.892698046829926e-05, "loss": 0.0516, "step": 53150 }, { "epoch": 1.7972210013861185, "grad_norm": 0.19864092767238617, "learning_rate": 4.8910449637048275e-05, "loss": 0.0526, "step": 53160 }, { "epoch": 1.7975590790763718, "grad_norm": 0.24881044030189514, "learning_rate": 4.8893918924948884e-05, "loss": 0.0539, "step": 53170 }, { "epoch": 1.7978971567666249, "grad_norm": 0.19067475199699402, "learning_rate": 4.8877388333808835e-05, "loss": 0.0501, "step": 53180 }, { "epoch": 1.7982352344568782, "grad_norm": 0.17796668410301208, "learning_rate": 4.8860857865435925e-05, "loss": 0.0469, "step": 53190 }, { "epoch": 1.7985733121471315, "grad_norm": 0.2723631262779236, "learning_rate": 4.884432752163789e-05, "loss": 0.0498, "step": 53200 }, { "epoch": 1.7989113898373845, "grad_norm": 0.23618845641613007, "learning_rate": 4.8827797304222466e-05, "loss": 0.0445, "step": 53210 }, { "epoch": 1.7992494675276378, "grad_norm": 0.2014966607093811, "learning_rate": 4.8811267214997366e-05, "loss": 0.0485, "step": 53220 }, { "epoch": 1.799587545217891, "grad_norm": 0.21689347922801971, "learning_rate": 4.879473725577029e-05, "loss": 0.0491, "step": 53230 }, { "epoch": 1.7999256229081442, "grad_norm": 0.23643527925014496, "learning_rate": 4.877820742834891e-05, "loss": 0.0437, "step": 53240 }, { "epoch": 1.8002637005983975, "grad_norm": 0.2758178412914276, "learning_rate": 4.8761677734540956e-05, "loss": 0.0465, "step": 53250 }, { "epoch": 1.8006017782886508, "grad_norm": 0.17771954834461212, "learning_rate": 4.874514817615407e-05, "loss": 0.0467, "step": 53260 }, { "epoch": 1.8009398559789038, "grad_norm": 0.25545045733451843, "learning_rate": 4.872861875499588e-05, "loss": 0.0469, "step": 53270 }, { "epoch": 1.8012779336691573, "grad_norm": 0.24907425045967102, "learning_rate": 4.871208947287404e-05, "loss": 0.0487, "step": 53280 }, { "epoch": 1.8016160113594104, "grad_norm": 0.39406144618988037, "learning_rate": 4.869556033159615e-05, "loss": 0.0519, "step": 53290 }, { "epoch": 1.8019540890496635, "grad_norm": 0.26209744811058044, "learning_rate": 4.8679031332969835e-05, "loss": 0.0474, "step": 53300 }, { "epoch": 1.802292166739917, "grad_norm": 0.1709473878145218, "learning_rate": 4.8662502478802635e-05, "loss": 0.047, "step": 53310 }, { "epoch": 1.80263024443017, "grad_norm": 0.22918307781219482, "learning_rate": 4.8645973770902176e-05, "loss": 0.0506, "step": 53320 }, { "epoch": 1.8029683221204231, "grad_norm": 0.32022976875305176, "learning_rate": 4.862944521107599e-05, "loss": 0.0498, "step": 53330 }, { "epoch": 1.8033063998106766, "grad_norm": 0.3017251491546631, "learning_rate": 4.861291680113162e-05, "loss": 0.0505, "step": 53340 }, { "epoch": 1.8036444775009297, "grad_norm": 0.27783238887786865, "learning_rate": 4.8596388542876563e-05, "loss": 0.048, "step": 53350 }, { "epoch": 1.8039825551911828, "grad_norm": 0.3091418147087097, "learning_rate": 4.857986043811836e-05, "loss": 0.0487, "step": 53360 }, { "epoch": 1.8043206328814363, "grad_norm": 0.24389919638633728, "learning_rate": 4.856333248866448e-05, "loss": 0.0491, "step": 53370 }, { "epoch": 1.8046587105716894, "grad_norm": 0.2574332058429718, "learning_rate": 4.85468046963224e-05, "loss": 0.0496, "step": 53380 }, { "epoch": 1.8049967882619424, "grad_norm": 0.5217694640159607, "learning_rate": 4.853027706289957e-05, "loss": 0.0526, "step": 53390 }, { "epoch": 1.805334865952196, "grad_norm": 0.27498000860214233, "learning_rate": 4.851374959020344e-05, "loss": 0.0452, "step": 53400 }, { "epoch": 1.805672943642449, "grad_norm": 0.2824588716030121, "learning_rate": 4.849722228004141e-05, "loss": 0.0497, "step": 53410 }, { "epoch": 1.8060110213327023, "grad_norm": 0.3323604166507721, "learning_rate": 4.848069513422091e-05, "loss": 0.0476, "step": 53420 }, { "epoch": 1.8063490990229556, "grad_norm": 0.2164333611726761, "learning_rate": 4.8464168154549294e-05, "loss": 0.0464, "step": 53430 }, { "epoch": 1.8066871767132087, "grad_norm": 0.16431954503059387, "learning_rate": 4.844764134283392e-05, "loss": 0.0462, "step": 53440 }, { "epoch": 1.807025254403462, "grad_norm": 0.25134602189064026, "learning_rate": 4.843111470088218e-05, "loss": 0.0522, "step": 53450 }, { "epoch": 1.8073633320937152, "grad_norm": 0.29980331659317017, "learning_rate": 4.841458823050137e-05, "loss": 0.0498, "step": 53460 }, { "epoch": 1.8077014097839683, "grad_norm": 0.18079204857349396, "learning_rate": 4.8398061933498816e-05, "loss": 0.0487, "step": 53470 }, { "epoch": 1.8080394874742216, "grad_norm": 0.5436781048774719, "learning_rate": 4.83815358116818e-05, "loss": 0.0551, "step": 53480 }, { "epoch": 1.808377565164475, "grad_norm": 0.3423801064491272, "learning_rate": 4.8365009866857584e-05, "loss": 0.0554, "step": 53490 }, { "epoch": 1.808715642854728, "grad_norm": 0.20970219373703003, "learning_rate": 4.834848410083342e-05, "loss": 0.0552, "step": 53500 }, { "epoch": 1.8090537205449813, "grad_norm": 0.2127183973789215, "learning_rate": 4.833195851541657e-05, "loss": 0.049, "step": 53510 }, { "epoch": 1.8093917982352345, "grad_norm": 0.23635438084602356, "learning_rate": 4.831543311241423e-05, "loss": 0.0522, "step": 53520 }, { "epoch": 1.8097298759254876, "grad_norm": 0.19467027485370636, "learning_rate": 4.829890789363359e-05, "loss": 0.0502, "step": 53530 }, { "epoch": 1.810067953615741, "grad_norm": 0.1838967651128769, "learning_rate": 4.828238286088182e-05, "loss": 0.0524, "step": 53540 }, { "epoch": 1.8104060313059942, "grad_norm": 0.38970112800598145, "learning_rate": 4.826585801596609e-05, "loss": 0.0511, "step": 53550 }, { "epoch": 1.8107441089962473, "grad_norm": 0.39498406648635864, "learning_rate": 4.8249333360693486e-05, "loss": 0.0496, "step": 53560 }, { "epoch": 1.8110821866865006, "grad_norm": 0.2513996362686157, "learning_rate": 4.8232808896871185e-05, "loss": 0.0477, "step": 53570 }, { "epoch": 1.8114202643767539, "grad_norm": 0.2047843337059021, "learning_rate": 4.821628462630624e-05, "loss": 0.0565, "step": 53580 }, { "epoch": 1.811758342067007, "grad_norm": 0.26182571053504944, "learning_rate": 4.8199760550805735e-05, "loss": 0.0518, "step": 53590 }, { "epoch": 1.8120964197572602, "grad_norm": 0.2566318213939667, "learning_rate": 4.818323667217669e-05, "loss": 0.0475, "step": 53600 }, { "epoch": 1.8124344974475135, "grad_norm": 0.2537088096141815, "learning_rate": 4.8166712992226165e-05, "loss": 0.0492, "step": 53610 }, { "epoch": 1.8127725751377666, "grad_norm": 0.1652657389640808, "learning_rate": 4.8150189512761154e-05, "loss": 0.0513, "step": 53620 }, { "epoch": 1.8131106528280199, "grad_norm": 0.18703815340995789, "learning_rate": 4.813366623558862e-05, "loss": 0.047, "step": 53630 }, { "epoch": 1.8134487305182732, "grad_norm": 0.20364896953105927, "learning_rate": 4.811714316251554e-05, "loss": 0.0483, "step": 53640 }, { "epoch": 1.8137868082085262, "grad_norm": 0.21829569339752197, "learning_rate": 4.810062029534886e-05, "loss": 0.0522, "step": 53650 }, { "epoch": 1.8141248858987795, "grad_norm": 0.29965391755104065, "learning_rate": 4.8084097635895505e-05, "loss": 0.0484, "step": 53660 }, { "epoch": 1.8144629635890328, "grad_norm": 0.17637521028518677, "learning_rate": 4.8067575185962335e-05, "loss": 0.0508, "step": 53670 }, { "epoch": 1.8148010412792859, "grad_norm": 0.20942699909210205, "learning_rate": 4.805105294735625e-05, "loss": 0.0484, "step": 53680 }, { "epoch": 1.8151391189695392, "grad_norm": 0.21968770027160645, "learning_rate": 4.803453092188406e-05, "loss": 0.0413, "step": 53690 }, { "epoch": 1.8154771966597925, "grad_norm": 0.2147306501865387, "learning_rate": 4.801800911135263e-05, "loss": 0.0536, "step": 53700 }, { "epoch": 1.8158152743500455, "grad_norm": 0.20704656839370728, "learning_rate": 4.8001487517568744e-05, "loss": 0.0475, "step": 53710 }, { "epoch": 1.8161533520402988, "grad_norm": 0.3296665549278259, "learning_rate": 4.798496614233918e-05, "loss": 0.0513, "step": 53720 }, { "epoch": 1.816491429730552, "grad_norm": 0.4261503219604492, "learning_rate": 4.796844498747069e-05, "loss": 0.0519, "step": 53730 }, { "epoch": 1.8168295074208052, "grad_norm": 0.2866780757904053, "learning_rate": 4.795192405477e-05, "loss": 0.053, "step": 53740 }, { "epoch": 1.8171675851110585, "grad_norm": 0.3778426945209503, "learning_rate": 4.79354033460438e-05, "loss": 0.0487, "step": 53750 }, { "epoch": 1.8175056628013118, "grad_norm": 0.21953511238098145, "learning_rate": 4.79188828630988e-05, "loss": 0.0479, "step": 53760 }, { "epoch": 1.8178437404915648, "grad_norm": 0.2814556360244751, "learning_rate": 4.790236260774165e-05, "loss": 0.0477, "step": 53770 }, { "epoch": 1.8181818181818183, "grad_norm": 0.31923121213912964, "learning_rate": 4.788584258177896e-05, "loss": 0.0507, "step": 53780 }, { "epoch": 1.8185198958720714, "grad_norm": 0.24330492317676544, "learning_rate": 4.7869322787017356e-05, "loss": 0.0466, "step": 53790 }, { "epoch": 1.8188579735623245, "grad_norm": 0.2535816729068756, "learning_rate": 4.78528032252634e-05, "loss": 0.0547, "step": 53800 }, { "epoch": 1.819196051252578, "grad_norm": 0.17709273099899292, "learning_rate": 4.783628389832366e-05, "loss": 0.0523, "step": 53810 }, { "epoch": 1.819534128942831, "grad_norm": 0.42579540610313416, "learning_rate": 4.7819764808004644e-05, "loss": 0.0542, "step": 53820 }, { "epoch": 1.8198722066330841, "grad_norm": 0.16069507598876953, "learning_rate": 4.780324595611289e-05, "loss": 0.0504, "step": 53830 }, { "epoch": 1.8202102843233376, "grad_norm": 0.16076849400997162, "learning_rate": 4.7786727344454854e-05, "loss": 0.054, "step": 53840 }, { "epoch": 1.8205483620135907, "grad_norm": 0.20503847301006317, "learning_rate": 4.777020897483697e-05, "loss": 0.0539, "step": 53850 }, { "epoch": 1.8208864397038438, "grad_norm": 0.2856680750846863, "learning_rate": 4.775369084906569e-05, "loss": 0.0474, "step": 53860 }, { "epoch": 1.8212245173940973, "grad_norm": 0.1807159185409546, "learning_rate": 4.773717296894742e-05, "loss": 0.0562, "step": 53870 }, { "epoch": 1.8215625950843504, "grad_norm": 0.20875027775764465, "learning_rate": 4.772065533628848e-05, "loss": 0.0505, "step": 53880 }, { "epoch": 1.8219006727746037, "grad_norm": 0.36027058959007263, "learning_rate": 4.7704137952895265e-05, "loss": 0.0476, "step": 53890 }, { "epoch": 1.822238750464857, "grad_norm": 0.24817544221878052, "learning_rate": 4.768762082057407e-05, "loss": 0.058, "step": 53900 }, { "epoch": 1.82257682815511, "grad_norm": 0.2301124483346939, "learning_rate": 4.7671103941131186e-05, "loss": 0.0493, "step": 53910 }, { "epoch": 1.8229149058453633, "grad_norm": 0.2555182874202728, "learning_rate": 4.765458731637287e-05, "loss": 0.0441, "step": 53920 }, { "epoch": 1.8232529835356166, "grad_norm": 0.25118255615234375, "learning_rate": 4.763807094810537e-05, "loss": 0.0483, "step": 53930 }, { "epoch": 1.8235910612258697, "grad_norm": 0.1771467626094818, "learning_rate": 4.762155483813487e-05, "loss": 0.0507, "step": 53940 }, { "epoch": 1.823929138916123, "grad_norm": 0.32850876450538635, "learning_rate": 4.760503898826754e-05, "loss": 0.0489, "step": 53950 }, { "epoch": 1.8242672166063763, "grad_norm": 0.1811007708311081, "learning_rate": 4.758852340030957e-05, "loss": 0.0485, "step": 53960 }, { "epoch": 1.8246052942966293, "grad_norm": 0.22849392890930176, "learning_rate": 4.757200807606705e-05, "loss": 0.0546, "step": 53970 }, { "epoch": 1.8249433719868826, "grad_norm": 0.37158751487731934, "learning_rate": 4.755549301734607e-05, "loss": 0.0517, "step": 53980 }, { "epoch": 1.825281449677136, "grad_norm": 0.2638264298439026, "learning_rate": 4.753897822595271e-05, "loss": 0.051, "step": 53990 }, { "epoch": 1.825619527367389, "grad_norm": 0.32903021574020386, "learning_rate": 4.7522463703692994e-05, "loss": 0.0543, "step": 54000 }, { "epoch": 1.8259576050576423, "grad_norm": 0.22754181921482086, "learning_rate": 4.75059494523729e-05, "loss": 0.0485, "step": 54010 }, { "epoch": 1.8262956827478956, "grad_norm": 0.25947442650794983, "learning_rate": 4.748943547379843e-05, "loss": 0.0542, "step": 54020 }, { "epoch": 1.8266337604381486, "grad_norm": 0.17198413610458374, "learning_rate": 4.7472921769775535e-05, "loss": 0.0531, "step": 54030 }, { "epoch": 1.826971838128402, "grad_norm": 0.26635849475860596, "learning_rate": 4.745640834211011e-05, "loss": 0.0493, "step": 54040 }, { "epoch": 1.8273099158186552, "grad_norm": 0.25635209679603577, "learning_rate": 4.7439895192608054e-05, "loss": 0.0506, "step": 54050 }, { "epoch": 1.8276479935089083, "grad_norm": 0.2163657397031784, "learning_rate": 4.7423382323075186e-05, "loss": 0.0502, "step": 54060 }, { "epoch": 1.8279860711991616, "grad_norm": 0.26710745692253113, "learning_rate": 4.740686973531737e-05, "loss": 0.0479, "step": 54070 }, { "epoch": 1.8283241488894149, "grad_norm": 0.30697518587112427, "learning_rate": 4.7390357431140375e-05, "loss": 0.0544, "step": 54080 }, { "epoch": 1.828662226579668, "grad_norm": 0.4453373849391937, "learning_rate": 4.737384541234998e-05, "loss": 0.0563, "step": 54090 }, { "epoch": 1.8290003042699212, "grad_norm": 0.21555089950561523, "learning_rate": 4.735733368075189e-05, "loss": 0.0543, "step": 54100 }, { "epoch": 1.8293383819601745, "grad_norm": 0.2601558268070221, "learning_rate": 4.734082223815182e-05, "loss": 0.0508, "step": 54110 }, { "epoch": 1.8296764596504276, "grad_norm": 0.24269451200962067, "learning_rate": 4.7324311086355436e-05, "loss": 0.0462, "step": 54120 }, { "epoch": 1.8300145373406809, "grad_norm": 0.17986659705638885, "learning_rate": 4.730780022716836e-05, "loss": 0.0511, "step": 54130 }, { "epoch": 1.8303526150309342, "grad_norm": 0.27824845910072327, "learning_rate": 4.729128966239619e-05, "loss": 0.0503, "step": 54140 }, { "epoch": 1.8306906927211872, "grad_norm": 0.29251858592033386, "learning_rate": 4.7274779393844536e-05, "loss": 0.0464, "step": 54150 }, { "epoch": 1.8310287704114405, "grad_norm": 0.2453414797782898, "learning_rate": 4.725826942331891e-05, "loss": 0.051, "step": 54160 }, { "epoch": 1.8313668481016938, "grad_norm": 0.23161691427230835, "learning_rate": 4.724175975262481e-05, "loss": 0.0527, "step": 54170 }, { "epoch": 1.8317049257919469, "grad_norm": 0.19234545528888702, "learning_rate": 4.722525038356772e-05, "loss": 0.0465, "step": 54180 }, { "epoch": 1.8320430034822002, "grad_norm": 1.6208956241607666, "learning_rate": 4.7208741317953075e-05, "loss": 0.0514, "step": 54190 }, { "epoch": 1.8323810811724535, "grad_norm": 0.7362587451934814, "learning_rate": 4.7192232557586264e-05, "loss": 0.052, "step": 54200 }, { "epoch": 1.8327191588627065, "grad_norm": 0.2903003692626953, "learning_rate": 4.717572410427269e-05, "loss": 0.0496, "step": 54210 }, { "epoch": 1.8330572365529598, "grad_norm": 0.2937648296356201, "learning_rate": 4.71592159598177e-05, "loss": 0.0532, "step": 54220 }, { "epoch": 1.8333953142432131, "grad_norm": 0.344906210899353, "learning_rate": 4.714270812602657e-05, "loss": 0.0517, "step": 54230 }, { "epoch": 1.8337333919334662, "grad_norm": 0.2877674102783203, "learning_rate": 4.712620060470458e-05, "loss": 0.0456, "step": 54240 }, { "epoch": 1.8340714696237197, "grad_norm": 0.37561464309692383, "learning_rate": 4.7109693397656986e-05, "loss": 0.0532, "step": 54250 }, { "epoch": 1.8344095473139728, "grad_norm": 0.3479851186275482, "learning_rate": 4.709318650668894e-05, "loss": 0.0525, "step": 54260 }, { "epoch": 1.8347476250042258, "grad_norm": 0.3024939000606537, "learning_rate": 4.7076679933605676e-05, "loss": 0.0518, "step": 54270 }, { "epoch": 1.8350857026944793, "grad_norm": 0.40410754084587097, "learning_rate": 4.70601736802123e-05, "loss": 0.0485, "step": 54280 }, { "epoch": 1.8354237803847324, "grad_norm": 0.24211932718753815, "learning_rate": 4.704366774831391e-05, "loss": 0.0518, "step": 54290 }, { "epoch": 1.8357618580749855, "grad_norm": 0.22601455450057983, "learning_rate": 4.702716213971557e-05, "loss": 0.0453, "step": 54300 }, { "epoch": 1.836099935765239, "grad_norm": 0.22631198167800903, "learning_rate": 4.70106568562223e-05, "loss": 0.0502, "step": 54310 }, { "epoch": 1.836438013455492, "grad_norm": 0.31409093737602234, "learning_rate": 4.6994151899639105e-05, "loss": 0.054, "step": 54320 }, { "epoch": 1.8367760911457451, "grad_norm": 0.32322749495506287, "learning_rate": 4.697764727177093e-05, "loss": 0.0505, "step": 54330 }, { "epoch": 1.8371141688359987, "grad_norm": 0.338888555765152, "learning_rate": 4.696114297442272e-05, "loss": 0.0509, "step": 54340 }, { "epoch": 1.8374522465262517, "grad_norm": 0.33328357338905334, "learning_rate": 4.694463900939933e-05, "loss": 0.0515, "step": 54350 }, { "epoch": 1.8377903242165048, "grad_norm": 0.16805708408355713, "learning_rate": 4.692813537850564e-05, "loss": 0.0457, "step": 54360 }, { "epoch": 1.8381284019067583, "grad_norm": 0.2819092273712158, "learning_rate": 4.6911632083546445e-05, "loss": 0.0532, "step": 54370 }, { "epoch": 1.8384664795970114, "grad_norm": 0.6537904739379883, "learning_rate": 4.6895129126326525e-05, "loss": 0.0508, "step": 54380 }, { "epoch": 1.8388045572872647, "grad_norm": 0.20980384945869446, "learning_rate": 4.687862650865059e-05, "loss": 0.0511, "step": 54390 }, { "epoch": 1.839142634977518, "grad_norm": 0.2027452141046524, "learning_rate": 4.686212423232339e-05, "loss": 0.051, "step": 54400 }, { "epoch": 1.839480712667771, "grad_norm": 0.3502214252948761, "learning_rate": 4.6845622299149565e-05, "loss": 0.0519, "step": 54410 }, { "epoch": 1.8398187903580243, "grad_norm": 0.2484167218208313, "learning_rate": 4.682912071093374e-05, "loss": 0.0457, "step": 54420 }, { "epoch": 1.8401568680482776, "grad_norm": 0.22790376842021942, "learning_rate": 4.681261946948052e-05, "loss": 0.0519, "step": 54430 }, { "epoch": 1.8404949457385307, "grad_norm": 0.20377540588378906, "learning_rate": 4.679611857659443e-05, "loss": 0.0582, "step": 54440 }, { "epoch": 1.840833023428784, "grad_norm": 0.4336772561073303, "learning_rate": 4.6779618034079975e-05, "loss": 0.0504, "step": 54450 }, { "epoch": 1.8411711011190373, "grad_norm": 0.20725658535957336, "learning_rate": 4.6763117843741675e-05, "loss": 0.046, "step": 54460 }, { "epoch": 1.8415091788092903, "grad_norm": 0.16634343564510345, "learning_rate": 4.674661800738394e-05, "loss": 0.0509, "step": 54470 }, { "epoch": 1.8418472564995436, "grad_norm": 0.264618843793869, "learning_rate": 4.673011852681117e-05, "loss": 0.0466, "step": 54480 }, { "epoch": 1.842185334189797, "grad_norm": 0.263171911239624, "learning_rate": 4.671361940382772e-05, "loss": 0.0524, "step": 54490 }, { "epoch": 1.84252341188005, "grad_norm": 0.196387380361557, "learning_rate": 4.669712064023791e-05, "loss": 0.0445, "step": 54500 }, { "epoch": 1.8428614895703033, "grad_norm": 0.4764033257961273, "learning_rate": 4.668062223784602e-05, "loss": 0.0536, "step": 54510 }, { "epoch": 1.8431995672605566, "grad_norm": 0.44276338815689087, "learning_rate": 4.6664124198456274e-05, "loss": 0.0501, "step": 54520 }, { "epoch": 1.8435376449508096, "grad_norm": 0.25590506196022034, "learning_rate": 4.6647626523872905e-05, "loss": 0.0488, "step": 54530 }, { "epoch": 1.843875722641063, "grad_norm": 0.27538880705833435, "learning_rate": 4.6631129215900066e-05, "loss": 0.0553, "step": 54540 }, { "epoch": 1.8442138003313162, "grad_norm": 0.3285161554813385, "learning_rate": 4.661463227634187e-05, "loss": 0.0459, "step": 54550 }, { "epoch": 1.8445518780215693, "grad_norm": 0.4876512885093689, "learning_rate": 4.659813570700238e-05, "loss": 0.0492, "step": 54560 }, { "epoch": 1.8448899557118226, "grad_norm": 0.22268857061862946, "learning_rate": 4.658163950968568e-05, "loss": 0.0465, "step": 54570 }, { "epoch": 1.8452280334020759, "grad_norm": 0.6063520312309265, "learning_rate": 4.6565143686195726e-05, "loss": 0.0488, "step": 54580 }, { "epoch": 1.845566111092329, "grad_norm": 0.22581209242343903, "learning_rate": 4.654864823833649e-05, "loss": 0.0462, "step": 54590 }, { "epoch": 1.8459041887825822, "grad_norm": 0.2024303823709488, "learning_rate": 4.65321531679119e-05, "loss": 0.0491, "step": 54600 }, { "epoch": 1.8462422664728355, "grad_norm": 0.32728803157806396, "learning_rate": 4.6515658476725834e-05, "loss": 0.047, "step": 54610 }, { "epoch": 1.8465803441630886, "grad_norm": 0.19193635880947113, "learning_rate": 4.649916416658212e-05, "loss": 0.0503, "step": 54620 }, { "epoch": 1.8469184218533419, "grad_norm": 0.3512113094329834, "learning_rate": 4.648267023928455e-05, "loss": 0.0455, "step": 54630 }, { "epoch": 1.8472564995435952, "grad_norm": 0.24591000378131866, "learning_rate": 4.646617669663687e-05, "loss": 0.0472, "step": 54640 }, { "epoch": 1.8475945772338482, "grad_norm": 0.23380081355571747, "learning_rate": 4.644968354044277e-05, "loss": 0.0515, "step": 54650 }, { "epoch": 1.8479326549241015, "grad_norm": 0.23949043452739716, "learning_rate": 4.643319077250596e-05, "loss": 0.0462, "step": 54660 }, { "epoch": 1.8482707326143548, "grad_norm": 0.4028727114200592, "learning_rate": 4.6416698394630055e-05, "loss": 0.0542, "step": 54670 }, { "epoch": 1.8486088103046079, "grad_norm": 0.35087525844573975, "learning_rate": 4.640020640861862e-05, "loss": 0.0441, "step": 54680 }, { "epoch": 1.8489468879948612, "grad_norm": 0.22560149431228638, "learning_rate": 4.6383714816275195e-05, "loss": 0.0478, "step": 54690 }, { "epoch": 1.8492849656851145, "grad_norm": 0.19927261769771576, "learning_rate": 4.636722361940329e-05, "loss": 0.0459, "step": 54700 }, { "epoch": 1.8496230433753675, "grad_norm": 0.2615146338939667, "learning_rate": 4.635073281980632e-05, "loss": 0.0468, "step": 54710 }, { "epoch": 1.8499611210656208, "grad_norm": 0.29012879729270935, "learning_rate": 4.633424241928775e-05, "loss": 0.0471, "step": 54720 }, { "epoch": 1.8502991987558741, "grad_norm": 0.24579091370105743, "learning_rate": 4.6317752419650914e-05, "loss": 0.0524, "step": 54730 }, { "epoch": 1.8506372764461272, "grad_norm": 0.21982723474502563, "learning_rate": 4.630126282269914e-05, "loss": 0.0515, "step": 54740 }, { "epoch": 1.8509753541363807, "grad_norm": 0.2154659926891327, "learning_rate": 4.628477363023569e-05, "loss": 0.0482, "step": 54750 }, { "epoch": 1.8513134318266338, "grad_norm": 0.16394676268100739, "learning_rate": 4.626828484406383e-05, "loss": 0.0545, "step": 54760 }, { "epoch": 1.8516515095168868, "grad_norm": 0.202952578663826, "learning_rate": 4.625179646598669e-05, "loss": 0.0512, "step": 54770 }, { "epoch": 1.8519895872071404, "grad_norm": 0.21441419422626495, "learning_rate": 4.6235308497807487e-05, "loss": 0.0471, "step": 54780 }, { "epoch": 1.8523276648973934, "grad_norm": 0.21697849035263062, "learning_rate": 4.6218820941329276e-05, "loss": 0.0491, "step": 54790 }, { "epoch": 1.8526657425876465, "grad_norm": 0.21150954067707062, "learning_rate": 4.620233379835513e-05, "loss": 0.0473, "step": 54800 }, { "epoch": 1.8530038202779, "grad_norm": 0.2887779176235199, "learning_rate": 4.618584707068804e-05, "loss": 0.0479, "step": 54810 }, { "epoch": 1.853341897968153, "grad_norm": 0.31587234139442444, "learning_rate": 4.616936076013099e-05, "loss": 0.0522, "step": 54820 }, { "epoch": 1.8536799756584061, "grad_norm": 0.6722875237464905, "learning_rate": 4.615287486848688e-05, "loss": 0.0561, "step": 54830 }, { "epoch": 1.8540180533486597, "grad_norm": 0.18637140095233917, "learning_rate": 4.6136389397558584e-05, "loss": 0.0452, "step": 54840 }, { "epoch": 1.8543561310389127, "grad_norm": 0.22069187462329865, "learning_rate": 4.6119904349148935e-05, "loss": 0.0504, "step": 54850 }, { "epoch": 1.854694208729166, "grad_norm": 0.2645856738090515, "learning_rate": 4.6103419725060726e-05, "loss": 0.0456, "step": 54860 }, { "epoch": 1.8550322864194193, "grad_norm": 0.21705159544944763, "learning_rate": 4.608693552709667e-05, "loss": 0.0498, "step": 54870 }, { "epoch": 1.8553703641096724, "grad_norm": 0.2045959085226059, "learning_rate": 4.607045175705947e-05, "loss": 0.0453, "step": 54880 }, { "epoch": 1.8557084417999257, "grad_norm": 0.2968853712081909, "learning_rate": 4.605396841675176e-05, "loss": 0.0494, "step": 54890 }, { "epoch": 1.856046519490179, "grad_norm": 0.20790882408618927, "learning_rate": 4.6037485507976094e-05, "loss": 0.0488, "step": 54900 }, { "epoch": 1.856384597180432, "grad_norm": 0.3751811683177948, "learning_rate": 4.602100303253509e-05, "loss": 0.0448, "step": 54910 }, { "epoch": 1.8567226748706853, "grad_norm": 0.25703921914100647, "learning_rate": 4.600452099223122e-05, "loss": 0.0513, "step": 54920 }, { "epoch": 1.8570607525609386, "grad_norm": 0.1917123943567276, "learning_rate": 4.598803938886691e-05, "loss": 0.0492, "step": 54930 }, { "epoch": 1.8573988302511917, "grad_norm": 0.2596842646598816, "learning_rate": 4.597155822424459e-05, "loss": 0.0462, "step": 54940 }, { "epoch": 1.857736907941445, "grad_norm": 0.18793538212776184, "learning_rate": 4.59550775001666e-05, "loss": 0.0446, "step": 54950 }, { "epoch": 1.8580749856316983, "grad_norm": 0.1715470254421234, "learning_rate": 4.5938597218435234e-05, "loss": 0.0483, "step": 54960 }, { "epoch": 1.8584130633219513, "grad_norm": 0.29323145747184753, "learning_rate": 4.592211738085279e-05, "loss": 0.0528, "step": 54970 }, { "epoch": 1.8587511410122046, "grad_norm": 0.3126136064529419, "learning_rate": 4.590563798922146e-05, "loss": 0.0512, "step": 54980 }, { "epoch": 1.859089218702458, "grad_norm": 0.1905408352613449, "learning_rate": 4.5889159045343404e-05, "loss": 0.0468, "step": 54990 }, { "epoch": 1.859427296392711, "grad_norm": 0.17413410544395447, "learning_rate": 4.5872680551020734e-05, "loss": 0.0539, "step": 55000 }, { "epoch": 1.8597653740829643, "grad_norm": 0.1902073323726654, "learning_rate": 4.585620250805552e-05, "loss": 0.0515, "step": 55010 }, { "epoch": 1.8601034517732176, "grad_norm": 0.23820121586322784, "learning_rate": 4.583972491824975e-05, "loss": 0.0527, "step": 55020 }, { "epoch": 1.8604415294634706, "grad_norm": 0.2875961363315582, "learning_rate": 4.582324778340541e-05, "loss": 0.058, "step": 55030 }, { "epoch": 1.860779607153724, "grad_norm": 0.5008675456047058, "learning_rate": 4.5806771105324424e-05, "loss": 0.0523, "step": 55040 }, { "epoch": 1.8611176848439772, "grad_norm": 0.24211359024047852, "learning_rate": 4.579029488580865e-05, "loss": 0.051, "step": 55050 }, { "epoch": 1.8614557625342303, "grad_norm": 0.2591157555580139, "learning_rate": 4.5773819126659876e-05, "loss": 0.0468, "step": 55060 }, { "epoch": 1.8617938402244836, "grad_norm": 0.18798011541366577, "learning_rate": 4.575734382967991e-05, "loss": 0.0475, "step": 55070 }, { "epoch": 1.8621319179147369, "grad_norm": 0.15589886903762817, "learning_rate": 4.574086899667044e-05, "loss": 0.0468, "step": 55080 }, { "epoch": 1.86246999560499, "grad_norm": 0.143880695104599, "learning_rate": 4.5724394629433106e-05, "loss": 0.047, "step": 55090 }, { "epoch": 1.8628080732952432, "grad_norm": 0.22504495084285736, "learning_rate": 4.570792072976958e-05, "loss": 0.0463, "step": 55100 }, { "epoch": 1.8631461509854965, "grad_norm": 0.2776404321193695, "learning_rate": 4.569144729948138e-05, "loss": 0.0416, "step": 55110 }, { "epoch": 1.8634842286757496, "grad_norm": 0.36827173829078674, "learning_rate": 4.5674974340370033e-05, "loss": 0.0486, "step": 55120 }, { "epoch": 1.8638223063660029, "grad_norm": 0.25141069293022156, "learning_rate": 4.565850185423699e-05, "loss": 0.0503, "step": 55130 }, { "epoch": 1.8641603840562562, "grad_norm": 0.2158418446779251, "learning_rate": 4.564202984288365e-05, "loss": 0.0526, "step": 55140 }, { "epoch": 1.8644984617465092, "grad_norm": 0.2461155503988266, "learning_rate": 4.5625558308111356e-05, "loss": 0.0482, "step": 55150 }, { "epoch": 1.8648365394367625, "grad_norm": 0.441546767950058, "learning_rate": 4.560908725172145e-05, "loss": 0.0511, "step": 55160 }, { "epoch": 1.8651746171270158, "grad_norm": 0.44299253821372986, "learning_rate": 4.559261667551516e-05, "loss": 0.0495, "step": 55170 }, { "epoch": 1.865512694817269, "grad_norm": 0.2851177155971527, "learning_rate": 4.5576146581293685e-05, "loss": 0.0504, "step": 55180 }, { "epoch": 1.8658507725075222, "grad_norm": 0.3517584800720215, "learning_rate": 4.5559676970858164e-05, "loss": 0.0529, "step": 55190 }, { "epoch": 1.8661888501977755, "grad_norm": 0.32950520515441895, "learning_rate": 4.5543207846009705e-05, "loss": 0.052, "step": 55200 }, { "epoch": 1.8665269278880285, "grad_norm": 0.26304760575294495, "learning_rate": 4.552673920854933e-05, "loss": 0.0519, "step": 55210 }, { "epoch": 1.866865005578282, "grad_norm": 0.2537195682525635, "learning_rate": 4.5510271060278006e-05, "loss": 0.0505, "step": 55220 }, { "epoch": 1.8672030832685351, "grad_norm": 0.26987224817276, "learning_rate": 4.5493803402996704e-05, "loss": 0.052, "step": 55230 }, { "epoch": 1.8675411609587882, "grad_norm": 0.2265157848596573, "learning_rate": 4.5477336238506295e-05, "loss": 0.0525, "step": 55240 }, { "epoch": 1.8678792386490417, "grad_norm": 0.4186253249645233, "learning_rate": 4.546086956860759e-05, "loss": 0.052, "step": 55250 }, { "epoch": 1.8682173163392948, "grad_norm": 0.3003932237625122, "learning_rate": 4.5444403395101374e-05, "loss": 0.0449, "step": 55260 }, { "epoch": 1.8685553940295478, "grad_norm": 0.15548457205295563, "learning_rate": 4.542793771978834e-05, "loss": 0.0487, "step": 55270 }, { "epoch": 1.8688934717198014, "grad_norm": 0.19298477470874786, "learning_rate": 4.541147254446916e-05, "loss": 0.0525, "step": 55280 }, { "epoch": 1.8692315494100544, "grad_norm": 0.21825741231441498, "learning_rate": 4.539500787094445e-05, "loss": 0.0514, "step": 55290 }, { "epoch": 1.8695696271003075, "grad_norm": 0.2636115252971649, "learning_rate": 4.537854370101475e-05, "loss": 0.0541, "step": 55300 }, { "epoch": 1.869907704790561, "grad_norm": 0.2121591866016388, "learning_rate": 4.5362080036480566e-05, "loss": 0.0479, "step": 55310 }, { "epoch": 1.870245782480814, "grad_norm": 0.2612915635108948, "learning_rate": 4.534561687914235e-05, "loss": 0.05, "step": 55320 }, { "epoch": 1.8705838601710671, "grad_norm": 0.22407297790050507, "learning_rate": 4.5329154230800466e-05, "loss": 0.0542, "step": 55330 }, { "epoch": 1.8709219378613207, "grad_norm": 0.21842797100543976, "learning_rate": 4.531269209325525e-05, "loss": 0.0517, "step": 55340 }, { "epoch": 1.8712600155515737, "grad_norm": 0.22744983434677124, "learning_rate": 4.529623046830696e-05, "loss": 0.0508, "step": 55350 }, { "epoch": 1.871598093241827, "grad_norm": 0.4175945222377777, "learning_rate": 4.5279769357755866e-05, "loss": 0.0458, "step": 55360 }, { "epoch": 1.8719361709320803, "grad_norm": 0.9526165723800659, "learning_rate": 4.5263308763402084e-05, "loss": 0.049, "step": 55370 }, { "epoch": 1.8722742486223334, "grad_norm": 0.23624609410762787, "learning_rate": 4.524684868704574e-05, "loss": 0.0475, "step": 55380 }, { "epoch": 1.8726123263125867, "grad_norm": 0.1655803769826889, "learning_rate": 4.523038913048688e-05, "loss": 0.0507, "step": 55390 }, { "epoch": 1.87295040400284, "grad_norm": 0.323164165019989, "learning_rate": 4.521393009552549e-05, "loss": 0.0499, "step": 55400 }, { "epoch": 1.873288481693093, "grad_norm": 0.25103214383125305, "learning_rate": 4.5197471583961484e-05, "loss": 0.0499, "step": 55410 }, { "epoch": 1.8736265593833463, "grad_norm": 0.3499433398246765, "learning_rate": 4.51810135975948e-05, "loss": 0.0488, "step": 55420 }, { "epoch": 1.8739646370735996, "grad_norm": 0.19186931848526, "learning_rate": 4.516455613822521e-05, "loss": 0.0465, "step": 55430 }, { "epoch": 1.8743027147638527, "grad_norm": 0.261237233877182, "learning_rate": 4.514809920765249e-05, "loss": 0.0446, "step": 55440 }, { "epoch": 1.874640792454106, "grad_norm": 0.3193915784358978, "learning_rate": 4.5131642807676356e-05, "loss": 0.0536, "step": 55450 }, { "epoch": 1.8749788701443593, "grad_norm": 0.1811460703611374, "learning_rate": 4.511518694009644e-05, "loss": 0.0496, "step": 55460 }, { "epoch": 1.8753169478346123, "grad_norm": 0.23011210560798645, "learning_rate": 4.509873160671231e-05, "loss": 0.0466, "step": 55470 }, { "epoch": 1.8756550255248656, "grad_norm": 0.48315542936325073, "learning_rate": 4.508227680932356e-05, "loss": 0.0485, "step": 55480 }, { "epoch": 1.875993103215119, "grad_norm": 0.1981344074010849, "learning_rate": 4.5065822549729605e-05, "loss": 0.0499, "step": 55490 }, { "epoch": 1.876331180905372, "grad_norm": 0.25854796171188354, "learning_rate": 4.504936882972989e-05, "loss": 0.0483, "step": 55500 }, { "epoch": 1.8766692585956253, "grad_norm": 0.27053558826446533, "learning_rate": 4.503291565112374e-05, "loss": 0.0515, "step": 55510 }, { "epoch": 1.8770073362858786, "grad_norm": 0.1928807646036148, "learning_rate": 4.501646301571048e-05, "loss": 0.0513, "step": 55520 }, { "epoch": 1.8773454139761316, "grad_norm": 0.2001979947090149, "learning_rate": 4.500001092528933e-05, "loss": 0.0459, "step": 55530 }, { "epoch": 1.877683491666385, "grad_norm": 0.3767967224121094, "learning_rate": 4.4983559381659455e-05, "loss": 0.0445, "step": 55540 }, { "epoch": 1.8780215693566382, "grad_norm": 0.2184707075357437, "learning_rate": 4.496710838661999e-05, "loss": 0.0499, "step": 55550 }, { "epoch": 1.8783596470468913, "grad_norm": 0.25029218196868896, "learning_rate": 4.495065794196999e-05, "loss": 0.0513, "step": 55560 }, { "epoch": 1.8786977247371446, "grad_norm": 0.25747284293174744, "learning_rate": 4.4934208049508455e-05, "loss": 0.0561, "step": 55570 }, { "epoch": 1.8790358024273979, "grad_norm": 0.2866532802581787, "learning_rate": 4.4917758711034303e-05, "loss": 0.0475, "step": 55580 }, { "epoch": 1.879373880117651, "grad_norm": 0.38809317350387573, "learning_rate": 4.490130992834642e-05, "loss": 0.0497, "step": 55590 }, { "epoch": 1.8797119578079042, "grad_norm": 0.2576427757740021, "learning_rate": 4.488486170324359e-05, "loss": 0.0521, "step": 55600 }, { "epoch": 1.8800500354981575, "grad_norm": 0.25018635392189026, "learning_rate": 4.486841403752463e-05, "loss": 0.047, "step": 55610 }, { "epoch": 1.8803881131884106, "grad_norm": 0.2690996825695038, "learning_rate": 4.4851966932988174e-05, "loss": 0.0442, "step": 55620 }, { "epoch": 1.8807261908786639, "grad_norm": 0.328209787607193, "learning_rate": 4.4835520391432894e-05, "loss": 0.0517, "step": 55630 }, { "epoch": 1.8810642685689172, "grad_norm": 0.2972628176212311, "learning_rate": 4.481907441465733e-05, "loss": 0.0533, "step": 55640 }, { "epoch": 1.8814023462591702, "grad_norm": 0.20708180963993073, "learning_rate": 4.480262900446e-05, "loss": 0.0452, "step": 55650 }, { "epoch": 1.8817404239494235, "grad_norm": 0.25052008032798767, "learning_rate": 4.478618416263933e-05, "loss": 0.0505, "step": 55660 }, { "epoch": 1.8820785016396768, "grad_norm": 0.4490766227245331, "learning_rate": 4.476973989099373e-05, "loss": 0.0586, "step": 55670 }, { "epoch": 1.88241657932993, "grad_norm": 0.35739341378211975, "learning_rate": 4.475329619132153e-05, "loss": 0.0543, "step": 55680 }, { "epoch": 1.8827546570201832, "grad_norm": 0.4838310182094574, "learning_rate": 4.4736853065420956e-05, "loss": 0.0451, "step": 55690 }, { "epoch": 1.8830927347104365, "grad_norm": 0.315243661403656, "learning_rate": 4.472041051509023e-05, "loss": 0.0512, "step": 55700 }, { "epoch": 1.8834308124006895, "grad_norm": 0.19526976346969604, "learning_rate": 4.470396854212746e-05, "loss": 0.0468, "step": 55710 }, { "epoch": 1.883768890090943, "grad_norm": 0.2458336353302002, "learning_rate": 4.468752714833072e-05, "loss": 0.0457, "step": 55720 }, { "epoch": 1.8841069677811961, "grad_norm": 0.2024015635251999, "learning_rate": 4.467108633549803e-05, "loss": 0.0542, "step": 55730 }, { "epoch": 1.8844450454714492, "grad_norm": 0.2747388184070587, "learning_rate": 4.4654646105427335e-05, "loss": 0.0518, "step": 55740 }, { "epoch": 1.8847831231617027, "grad_norm": 0.2339678555727005, "learning_rate": 4.463820645991651e-05, "loss": 0.0568, "step": 55750 }, { "epoch": 1.8851212008519558, "grad_norm": 0.23022280633449554, "learning_rate": 4.462176740076334e-05, "loss": 0.0506, "step": 55760 }, { "epoch": 1.8854592785422088, "grad_norm": 0.3797459602355957, "learning_rate": 4.4605328929765624e-05, "loss": 0.0456, "step": 55770 }, { "epoch": 1.8857973562324624, "grad_norm": 0.5234701037406921, "learning_rate": 4.458889104872102e-05, "loss": 0.0506, "step": 55780 }, { "epoch": 1.8861354339227154, "grad_norm": 0.29130449891090393, "learning_rate": 4.457245375942715e-05, "loss": 0.0469, "step": 55790 }, { "epoch": 1.8864735116129685, "grad_norm": 0.2808367609977722, "learning_rate": 4.455601706368157e-05, "loss": 0.0514, "step": 55800 }, { "epoch": 1.886811589303222, "grad_norm": 0.2194249927997589, "learning_rate": 4.453958096328179e-05, "loss": 0.0527, "step": 55810 }, { "epoch": 1.887149666993475, "grad_norm": 0.3130941689014435, "learning_rate": 4.452314546002523e-05, "loss": 0.0471, "step": 55820 }, { "epoch": 1.8874877446837284, "grad_norm": 0.215606689453125, "learning_rate": 4.4506710555709246e-05, "loss": 0.049, "step": 55830 }, { "epoch": 1.8878258223739817, "grad_norm": 0.35088175535202026, "learning_rate": 4.449027625213114e-05, "loss": 0.0497, "step": 55840 }, { "epoch": 1.8881639000642347, "grad_norm": 0.25479862093925476, "learning_rate": 4.4473842551088135e-05, "loss": 0.0477, "step": 55850 }, { "epoch": 1.888501977754488, "grad_norm": 0.19563715159893036, "learning_rate": 4.445740945437738e-05, "loss": 0.0464, "step": 55860 }, { "epoch": 1.8888400554447413, "grad_norm": 0.1980934590101242, "learning_rate": 4.444097696379602e-05, "loss": 0.0483, "step": 55870 }, { "epoch": 1.8891781331349944, "grad_norm": 0.295041024684906, "learning_rate": 4.442454508114106e-05, "loss": 0.0486, "step": 55880 }, { "epoch": 1.8895162108252477, "grad_norm": 0.2070726603269577, "learning_rate": 4.4408113808209465e-05, "loss": 0.0454, "step": 55890 }, { "epoch": 1.889854288515501, "grad_norm": 0.25094422698020935, "learning_rate": 4.439168314679813e-05, "loss": 0.0463, "step": 55900 }, { "epoch": 1.890192366205754, "grad_norm": 0.32362091541290283, "learning_rate": 4.43752530987039e-05, "loss": 0.0504, "step": 55910 }, { "epoch": 1.8905304438960073, "grad_norm": 0.33091652393341064, "learning_rate": 4.4358823665723515e-05, "loss": 0.0516, "step": 55920 }, { "epoch": 1.8908685215862606, "grad_norm": 0.17006686329841614, "learning_rate": 4.434239484965371e-05, "loss": 0.0437, "step": 55930 }, { "epoch": 1.8912065992765137, "grad_norm": 0.15259061753749847, "learning_rate": 4.4325966652291103e-05, "loss": 0.0511, "step": 55940 }, { "epoch": 1.891544676966767, "grad_norm": 0.27005741000175476, "learning_rate": 4.430953907543225e-05, "loss": 0.0483, "step": 55950 }, { "epoch": 1.8918827546570203, "grad_norm": 0.33113163709640503, "learning_rate": 4.4293112120873645e-05, "loss": 0.0495, "step": 55960 }, { "epoch": 1.8922208323472733, "grad_norm": 0.25578010082244873, "learning_rate": 4.427668579041172e-05, "loss": 0.0477, "step": 55970 }, { "epoch": 1.8925589100375266, "grad_norm": 0.33273378014564514, "learning_rate": 4.4260260085842816e-05, "loss": 0.0495, "step": 55980 }, { "epoch": 1.89289698772778, "grad_norm": 0.2842034101486206, "learning_rate": 4.4243835008963256e-05, "loss": 0.0484, "step": 55990 }, { "epoch": 1.893235065418033, "grad_norm": 0.23374557495117188, "learning_rate": 4.422741056156925e-05, "loss": 0.051, "step": 56000 }, { "epoch": 1.8935731431082863, "grad_norm": 0.23300103843212128, "learning_rate": 4.421098674545693e-05, "loss": 0.049, "step": 56010 }, { "epoch": 1.8939112207985396, "grad_norm": 0.22333493828773499, "learning_rate": 4.419456356242241e-05, "loss": 0.0477, "step": 56020 }, { "epoch": 1.8942492984887926, "grad_norm": 0.24403023719787598, "learning_rate": 4.4178141014261685e-05, "loss": 0.0459, "step": 56030 }, { "epoch": 1.894587376179046, "grad_norm": 0.16638650000095367, "learning_rate": 4.416171910277071e-05, "loss": 0.0542, "step": 56040 }, { "epoch": 1.8949254538692992, "grad_norm": 0.26852452754974365, "learning_rate": 4.414529782974533e-05, "loss": 0.0564, "step": 56050 }, { "epoch": 1.8952635315595523, "grad_norm": 0.2673509418964386, "learning_rate": 4.41288771969814e-05, "loss": 0.0506, "step": 56060 }, { "epoch": 1.8956016092498056, "grad_norm": 0.24225516617298126, "learning_rate": 4.411245720627464e-05, "loss": 0.0491, "step": 56070 }, { "epoch": 1.8959396869400589, "grad_norm": 0.27996498346328735, "learning_rate": 4.40960378594207e-05, "loss": 0.0511, "step": 56080 }, { "epoch": 1.896277764630312, "grad_norm": 0.3476378917694092, "learning_rate": 4.407961915821519e-05, "loss": 0.0562, "step": 56090 }, { "epoch": 1.8966158423205652, "grad_norm": 0.1768382340669632, "learning_rate": 4.406320110445362e-05, "loss": 0.0508, "step": 56100 }, { "epoch": 1.8969539200108185, "grad_norm": 0.3524153232574463, "learning_rate": 4.404678369993144e-05, "loss": 0.0496, "step": 56110 }, { "epoch": 1.8972919977010716, "grad_norm": 0.21401086449623108, "learning_rate": 4.403036694644406e-05, "loss": 0.054, "step": 56120 }, { "epoch": 1.8976300753913249, "grad_norm": 0.2419598400592804, "learning_rate": 4.4013950845786764e-05, "loss": 0.0473, "step": 56130 }, { "epoch": 1.8979681530815782, "grad_norm": 0.1937231570482254, "learning_rate": 4.399753539975482e-05, "loss": 0.0518, "step": 56140 }, { "epoch": 1.8983062307718312, "grad_norm": 0.3521648645401001, "learning_rate": 4.398112061014337e-05, "loss": 0.0528, "step": 56150 }, { "epoch": 1.8986443084620845, "grad_norm": 0.2989485561847687, "learning_rate": 4.396470647874753e-05, "loss": 0.0515, "step": 56160 }, { "epoch": 1.8989823861523378, "grad_norm": 0.3812243640422821, "learning_rate": 4.394829300736229e-05, "loss": 0.0501, "step": 56170 }, { "epoch": 1.899320463842591, "grad_norm": 0.19469523429870605, "learning_rate": 4.393188019778265e-05, "loss": 0.0475, "step": 56180 }, { "epoch": 1.8996585415328444, "grad_norm": 0.35856735706329346, "learning_rate": 4.391546805180347e-05, "loss": 0.0432, "step": 56190 }, { "epoch": 1.8999966192230975, "grad_norm": 0.2798653841018677, "learning_rate": 4.389905657121955e-05, "loss": 0.0496, "step": 56200 }, { "epoch": 1.9003346969133506, "grad_norm": 0.6145756244659424, "learning_rate": 4.3882645757825635e-05, "loss": 0.0526, "step": 56210 }, { "epoch": 1.900672774603604, "grad_norm": 0.29704996943473816, "learning_rate": 4.386623561341637e-05, "loss": 0.0529, "step": 56220 }, { "epoch": 1.9010108522938571, "grad_norm": 0.2233390510082245, "learning_rate": 4.384982613978637e-05, "loss": 0.0465, "step": 56230 }, { "epoch": 1.9013489299841102, "grad_norm": 0.6027306914329529, "learning_rate": 4.383341733873012e-05, "loss": 0.0484, "step": 56240 }, { "epoch": 1.9016870076743637, "grad_norm": 0.4278462529182434, "learning_rate": 4.381700921204209e-05, "loss": 0.0489, "step": 56250 }, { "epoch": 1.9020250853646168, "grad_norm": 0.23092441260814667, "learning_rate": 4.380060176151663e-05, "loss": 0.0499, "step": 56260 }, { "epoch": 1.9023631630548699, "grad_norm": 0.24171318113803864, "learning_rate": 4.378419498894805e-05, "loss": 0.0511, "step": 56270 }, { "epoch": 1.9027012407451234, "grad_norm": 0.4497642517089844, "learning_rate": 4.376778889613056e-05, "loss": 0.0522, "step": 56280 }, { "epoch": 1.9030393184353764, "grad_norm": 0.21913789212703705, "learning_rate": 4.375138348485831e-05, "loss": 0.0466, "step": 56290 }, { "epoch": 1.9033773961256295, "grad_norm": 0.35806673765182495, "learning_rate": 4.3734978756925346e-05, "loss": 0.0499, "step": 56300 }, { "epoch": 1.903715473815883, "grad_norm": 0.21785789728164673, "learning_rate": 4.3718574714125706e-05, "loss": 0.0488, "step": 56310 }, { "epoch": 1.904053551506136, "grad_norm": 0.24651043117046356, "learning_rate": 4.370217135825329e-05, "loss": 0.0511, "step": 56320 }, { "epoch": 1.9043916291963894, "grad_norm": 0.19965502619743347, "learning_rate": 4.368576869110194e-05, "loss": 0.0527, "step": 56330 }, { "epoch": 1.9047297068866427, "grad_norm": 0.27968958020210266, "learning_rate": 4.366936671446544e-05, "loss": 0.0475, "step": 56340 }, { "epoch": 1.9050677845768957, "grad_norm": 0.18154382705688477, "learning_rate": 4.3652965430137474e-05, "loss": 0.0456, "step": 56350 }, { "epoch": 1.905405862267149, "grad_norm": 0.2938777804374695, "learning_rate": 4.3636564839911646e-05, "loss": 0.049, "step": 56360 }, { "epoch": 1.9057439399574023, "grad_norm": 0.18781428039073944, "learning_rate": 4.3620164945581545e-05, "loss": 0.0504, "step": 56370 }, { "epoch": 1.9060820176476554, "grad_norm": 0.23881745338439941, "learning_rate": 4.360376574894061e-05, "loss": 0.0515, "step": 56380 }, { "epoch": 1.9064200953379087, "grad_norm": 0.30101171135902405, "learning_rate": 4.358736725178224e-05, "loss": 0.0497, "step": 56390 }, { "epoch": 1.906758173028162, "grad_norm": 0.1935708224773407, "learning_rate": 4.357096945589974e-05, "loss": 0.0516, "step": 56400 }, { "epoch": 1.907096250718415, "grad_norm": 0.24185259640216827, "learning_rate": 4.355457236308636e-05, "loss": 0.0453, "step": 56410 }, { "epoch": 1.9074343284086683, "grad_norm": 0.15530039370059967, "learning_rate": 4.353817597513526e-05, "loss": 0.0503, "step": 56420 }, { "epoch": 1.9077724060989216, "grad_norm": 0.43207669258117676, "learning_rate": 4.352178029383948e-05, "loss": 0.0543, "step": 56430 }, { "epoch": 1.9081104837891747, "grad_norm": 0.19109168648719788, "learning_rate": 4.3505385320992105e-05, "loss": 0.0497, "step": 56440 }, { "epoch": 1.908448561479428, "grad_norm": 0.2940759062767029, "learning_rate": 4.348899105838602e-05, "loss": 0.0512, "step": 56450 }, { "epoch": 1.9087866391696813, "grad_norm": 0.22147785127162933, "learning_rate": 4.3472597507814087e-05, "loss": 0.0466, "step": 56460 }, { "epoch": 1.9091247168599343, "grad_norm": 0.16660529375076294, "learning_rate": 4.3456204671069066e-05, "loss": 0.0499, "step": 56470 }, { "epoch": 1.9094627945501876, "grad_norm": 0.35328730940818787, "learning_rate": 4.343981254994367e-05, "loss": 0.0483, "step": 56480 }, { "epoch": 1.909800872240441, "grad_norm": 0.4774778187274933, "learning_rate": 4.34234211462305e-05, "loss": 0.0473, "step": 56490 }, { "epoch": 1.910138949930694, "grad_norm": 0.2405986785888672, "learning_rate": 4.340703046172213e-05, "loss": 0.0495, "step": 56500 }, { "epoch": 1.9104770276209473, "grad_norm": 0.3112931549549103, "learning_rate": 4.339064049821097e-05, "loss": 0.0496, "step": 56510 }, { "epoch": 1.9108151053112006, "grad_norm": 0.3673633337020874, "learning_rate": 4.3374251257489446e-05, "loss": 0.0451, "step": 56520 }, { "epoch": 1.9111531830014536, "grad_norm": 0.3287216126918793, "learning_rate": 4.335786274134984e-05, "loss": 0.0525, "step": 56530 }, { "epoch": 1.911491260691707, "grad_norm": 0.18998464941978455, "learning_rate": 4.334147495158438e-05, "loss": 0.0493, "step": 56540 }, { "epoch": 1.9118293383819602, "grad_norm": 0.18954460322856903, "learning_rate": 4.33250878899852e-05, "loss": 0.052, "step": 56550 }, { "epoch": 1.9121674160722133, "grad_norm": 0.21283508837223053, "learning_rate": 4.330870155834437e-05, "loss": 0.0493, "step": 56560 }, { "epoch": 1.9125054937624666, "grad_norm": 0.4134674668312073, "learning_rate": 4.3292315958453884e-05, "loss": 0.0533, "step": 56570 }, { "epoch": 1.9128435714527199, "grad_norm": 0.41534534096717834, "learning_rate": 4.3275931092105646e-05, "loss": 0.0507, "step": 56580 }, { "epoch": 1.913181649142973, "grad_norm": 0.1891532689332962, "learning_rate": 4.325954696109147e-05, "loss": 0.0479, "step": 56590 }, { "epoch": 1.9135197268332262, "grad_norm": 0.2332625836133957, "learning_rate": 4.32431635672031e-05, "loss": 0.048, "step": 56600 }, { "epoch": 1.9138578045234795, "grad_norm": 0.3569364845752716, "learning_rate": 4.3226780912232214e-05, "loss": 0.0442, "step": 56610 }, { "epoch": 1.9141958822137326, "grad_norm": 0.27390074729919434, "learning_rate": 4.321039899797036e-05, "loss": 0.0559, "step": 56620 }, { "epoch": 1.914533959903986, "grad_norm": 0.19170379638671875, "learning_rate": 4.319401782620908e-05, "loss": 0.049, "step": 56630 }, { "epoch": 1.9148720375942392, "grad_norm": 0.359307199716568, "learning_rate": 4.317763739873978e-05, "loss": 0.0476, "step": 56640 }, { "epoch": 1.9152101152844923, "grad_norm": 0.24575646221637726, "learning_rate": 4.31612577173538e-05, "loss": 0.0475, "step": 56650 }, { "epoch": 1.9155481929747455, "grad_norm": 0.19502152502536774, "learning_rate": 4.314487878384239e-05, "loss": 0.053, "step": 56660 }, { "epoch": 1.9158862706649988, "grad_norm": 0.3598827123641968, "learning_rate": 4.3128500599996736e-05, "loss": 0.0519, "step": 56670 }, { "epoch": 1.916224348355252, "grad_norm": 0.30316075682640076, "learning_rate": 4.31121231676079e-05, "loss": 0.048, "step": 56680 }, { "epoch": 1.9165624260455054, "grad_norm": 0.2787904739379883, "learning_rate": 4.309574648846694e-05, "loss": 0.0482, "step": 56690 }, { "epoch": 1.9169005037357585, "grad_norm": 0.3242732286453247, "learning_rate": 4.3079370564364755e-05, "loss": 0.048, "step": 56700 }, { "epoch": 1.9172385814260116, "grad_norm": 0.3038822114467621, "learning_rate": 4.306299539709221e-05, "loss": 0.0549, "step": 56710 }, { "epoch": 1.917576659116265, "grad_norm": 0.26631462574005127, "learning_rate": 4.3046620988440046e-05, "loss": 0.0503, "step": 56720 }, { "epoch": 1.9179147368065181, "grad_norm": 0.3854822516441345, "learning_rate": 4.303024734019897e-05, "loss": 0.0542, "step": 56730 }, { "epoch": 1.9182528144967712, "grad_norm": 0.20339888334274292, "learning_rate": 4.3013874454159565e-05, "loss": 0.0485, "step": 56740 }, { "epoch": 1.9185908921870247, "grad_norm": 0.21026833355426788, "learning_rate": 4.299750233211233e-05, "loss": 0.0452, "step": 56750 }, { "epoch": 1.9189289698772778, "grad_norm": 0.20530690252780914, "learning_rate": 4.2981130975847715e-05, "loss": 0.0496, "step": 56760 }, { "epoch": 1.9192670475675309, "grad_norm": 0.7569129467010498, "learning_rate": 4.2964760387156076e-05, "loss": 0.0502, "step": 56770 }, { "epoch": 1.9196051252577844, "grad_norm": 0.20491468906402588, "learning_rate": 4.294839056782767e-05, "loss": 0.0476, "step": 56780 }, { "epoch": 1.9199432029480374, "grad_norm": 0.3969484865665436, "learning_rate": 4.293202151965266e-05, "loss": 0.056, "step": 56790 }, { "epoch": 1.9202812806382907, "grad_norm": 0.4873866140842438, "learning_rate": 4.291565324442115e-05, "loss": 0.0467, "step": 56800 }, { "epoch": 1.920619358328544, "grad_norm": 0.23378190398216248, "learning_rate": 4.2899285743923135e-05, "loss": 0.0495, "step": 56810 }, { "epoch": 1.920957436018797, "grad_norm": 0.3967688977718353, "learning_rate": 4.288291901994857e-05, "loss": 0.0503, "step": 56820 }, { "epoch": 1.9212955137090504, "grad_norm": 0.18573687970638275, "learning_rate": 4.286655307428729e-05, "loss": 0.049, "step": 56830 }, { "epoch": 1.9216335913993037, "grad_norm": 0.4247567653656006, "learning_rate": 4.285018790872903e-05, "loss": 0.0484, "step": 56840 }, { "epoch": 1.9219716690895567, "grad_norm": 0.27699583768844604, "learning_rate": 4.283382352506347e-05, "loss": 0.052, "step": 56850 }, { "epoch": 1.92230974677981, "grad_norm": 0.2213541567325592, "learning_rate": 4.2817459925080205e-05, "loss": 0.0503, "step": 56860 }, { "epoch": 1.9226478244700633, "grad_norm": 0.20792953670024872, "learning_rate": 4.28010971105687e-05, "loss": 0.0443, "step": 56870 }, { "epoch": 1.9229859021603164, "grad_norm": 0.2907000482082367, "learning_rate": 4.278473508331841e-05, "loss": 0.0509, "step": 56880 }, { "epoch": 1.9233239798505697, "grad_norm": 0.22661936283111572, "learning_rate": 4.276837384511864e-05, "loss": 0.0507, "step": 56890 }, { "epoch": 1.923662057540823, "grad_norm": 0.254207968711853, "learning_rate": 4.275201339775864e-05, "loss": 0.0495, "step": 56900 }, { "epoch": 1.924000135231076, "grad_norm": 0.29642701148986816, "learning_rate": 4.273565374302756e-05, "loss": 0.0501, "step": 56910 }, { "epoch": 1.9243382129213293, "grad_norm": 0.33979538083076477, "learning_rate": 4.2719294882714454e-05, "loss": 0.0512, "step": 56920 }, { "epoch": 1.9246762906115826, "grad_norm": 0.28563323616981506, "learning_rate": 4.270293681860831e-05, "loss": 0.0471, "step": 56930 }, { "epoch": 1.9250143683018357, "grad_norm": 0.19604431092739105, "learning_rate": 4.268657955249802e-05, "loss": 0.0473, "step": 56940 }, { "epoch": 1.925352445992089, "grad_norm": 0.22654989361763, "learning_rate": 4.267022308617241e-05, "loss": 0.0487, "step": 56950 }, { "epoch": 1.9256905236823423, "grad_norm": 0.3127986192703247, "learning_rate": 4.265386742142018e-05, "loss": 0.0441, "step": 56960 }, { "epoch": 1.9260286013725954, "grad_norm": 0.35903164744377136, "learning_rate": 4.263751256002996e-05, "loss": 0.0484, "step": 56970 }, { "epoch": 1.9263666790628486, "grad_norm": 0.19036023318767548, "learning_rate": 4.26211585037903e-05, "loss": 0.0453, "step": 56980 }, { "epoch": 1.926704756753102, "grad_norm": 0.26647719740867615, "learning_rate": 4.2604805254489656e-05, "loss": 0.0469, "step": 56990 }, { "epoch": 1.927042834443355, "grad_norm": 0.16942285001277924, "learning_rate": 4.258845281391638e-05, "loss": 0.045, "step": 57000 }, { "epoch": 1.9273809121336083, "grad_norm": 0.2314932942390442, "learning_rate": 4.257210118385876e-05, "loss": 0.0477, "step": 57010 }, { "epoch": 1.9277189898238616, "grad_norm": 0.20229098200798035, "learning_rate": 4.2555750366104994e-05, "loss": 0.05, "step": 57020 }, { "epoch": 1.9280570675141147, "grad_norm": 0.3050214350223541, "learning_rate": 4.2539400362443184e-05, "loss": 0.0505, "step": 57030 }, { "epoch": 1.928395145204368, "grad_norm": 0.22478702664375305, "learning_rate": 4.252305117466133e-05, "loss": 0.0474, "step": 57040 }, { "epoch": 1.9287332228946212, "grad_norm": 0.239334836602211, "learning_rate": 4.2506702804547356e-05, "loss": 0.0455, "step": 57050 }, { "epoch": 1.9290713005848743, "grad_norm": 0.27322471141815186, "learning_rate": 4.249035525388907e-05, "loss": 0.0482, "step": 57060 }, { "epoch": 1.9294093782751276, "grad_norm": 0.3405623435974121, "learning_rate": 4.247400852447427e-05, "loss": 0.0494, "step": 57070 }, { "epoch": 1.9297474559653809, "grad_norm": 0.27415961027145386, "learning_rate": 4.245766261809059e-05, "loss": 0.0473, "step": 57080 }, { "epoch": 1.930085533655634, "grad_norm": 0.30395743250846863, "learning_rate": 4.244131753652558e-05, "loss": 0.0523, "step": 57090 }, { "epoch": 1.9304236113458872, "grad_norm": 0.1796710044145584, "learning_rate": 4.242497328156672e-05, "loss": 0.0503, "step": 57100 }, { "epoch": 1.9307616890361405, "grad_norm": 0.23031246662139893, "learning_rate": 4.240862985500139e-05, "loss": 0.0476, "step": 57110 }, { "epoch": 1.9310997667263936, "grad_norm": 0.3179086744785309, "learning_rate": 4.2392287258616894e-05, "loss": 0.0524, "step": 57120 }, { "epoch": 1.931437844416647, "grad_norm": 0.2580122649669647, "learning_rate": 4.2375945494200396e-05, "loss": 0.0518, "step": 57130 }, { "epoch": 1.9317759221069002, "grad_norm": 0.3122889995574951, "learning_rate": 4.2359604563539054e-05, "loss": 0.0489, "step": 57140 }, { "epoch": 1.9321139997971533, "grad_norm": 0.32996663451194763, "learning_rate": 4.2343264468419885e-05, "loss": 0.0432, "step": 57150 }, { "epoch": 1.9324520774874068, "grad_norm": 0.41108131408691406, "learning_rate": 4.232692521062979e-05, "loss": 0.0505, "step": 57160 }, { "epoch": 1.9327901551776598, "grad_norm": 0.23942546546459198, "learning_rate": 4.231058679195562e-05, "loss": 0.0538, "step": 57170 }, { "epoch": 1.933128232867913, "grad_norm": 0.19896072149276733, "learning_rate": 4.229424921418411e-05, "loss": 0.0513, "step": 57180 }, { "epoch": 1.9334663105581664, "grad_norm": 0.2629457712173462, "learning_rate": 4.2277912479101916e-05, "loss": 0.0528, "step": 57190 }, { "epoch": 1.9338043882484195, "grad_norm": 0.4428122043609619, "learning_rate": 4.226157658849562e-05, "loss": 0.0531, "step": 57200 }, { "epoch": 1.9341424659386726, "grad_norm": 0.2946465313434601, "learning_rate": 4.2245241544151664e-05, "loss": 0.0541, "step": 57210 }, { "epoch": 1.934480543628926, "grad_norm": 0.28056105971336365, "learning_rate": 4.2228907347856425e-05, "loss": 0.0478, "step": 57220 }, { "epoch": 1.9348186213191791, "grad_norm": 0.21605318784713745, "learning_rate": 4.2212574001396214e-05, "loss": 0.0492, "step": 57230 }, { "epoch": 1.9351566990094322, "grad_norm": 0.26036420464515686, "learning_rate": 4.21962415065572e-05, "loss": 0.049, "step": 57240 }, { "epoch": 1.9354947766996857, "grad_norm": 0.3615540564060211, "learning_rate": 4.217990986512548e-05, "loss": 0.0521, "step": 57250 }, { "epoch": 1.9358328543899388, "grad_norm": 0.18691669404506683, "learning_rate": 4.216357907888704e-05, "loss": 0.0439, "step": 57260 }, { "epoch": 1.9361709320801919, "grad_norm": 0.16667576134204865, "learning_rate": 4.2147249149627824e-05, "loss": 0.0459, "step": 57270 }, { "epoch": 1.9365090097704454, "grad_norm": 0.23251405358314514, "learning_rate": 4.213092007913364e-05, "loss": 0.0501, "step": 57280 }, { "epoch": 1.9368470874606984, "grad_norm": 0.1764395534992218, "learning_rate": 4.2114591869190205e-05, "loss": 0.0434, "step": 57290 }, { "epoch": 1.9371851651509517, "grad_norm": 0.24863585829734802, "learning_rate": 4.209826452158315e-05, "loss": 0.0474, "step": 57300 }, { "epoch": 1.937523242841205, "grad_norm": 0.20679955184459686, "learning_rate": 4.2081938038098e-05, "loss": 0.049, "step": 57310 }, { "epoch": 1.937861320531458, "grad_norm": 0.2879149317741394, "learning_rate": 4.2065612420520185e-05, "loss": 0.049, "step": 57320 }, { "epoch": 1.9381993982217114, "grad_norm": 0.26015689969062805, "learning_rate": 4.204928767063509e-05, "loss": 0.0471, "step": 57330 }, { "epoch": 1.9385374759119647, "grad_norm": 0.32662320137023926, "learning_rate": 4.2032963790227944e-05, "loss": 0.05, "step": 57340 }, { "epoch": 1.9388755536022177, "grad_norm": 0.37732014060020447, "learning_rate": 4.2016640781083887e-05, "loss": 0.0528, "step": 57350 }, { "epoch": 1.939213631292471, "grad_norm": 0.25908172130584717, "learning_rate": 4.2000318644988004e-05, "loss": 0.0511, "step": 57360 }, { "epoch": 1.9395517089827243, "grad_norm": 0.6336226463317871, "learning_rate": 4.198399738372524e-05, "loss": 0.0504, "step": 57370 }, { "epoch": 1.9398897866729774, "grad_norm": 0.21006442606449127, "learning_rate": 4.196767699908045e-05, "loss": 0.0492, "step": 57380 }, { "epoch": 1.9402278643632307, "grad_norm": 0.2824147939682007, "learning_rate": 4.195135749283845e-05, "loss": 0.0516, "step": 57390 }, { "epoch": 1.940565942053484, "grad_norm": 0.29170432686805725, "learning_rate": 4.1935038866783894e-05, "loss": 0.0534, "step": 57400 }, { "epoch": 1.940904019743737, "grad_norm": 0.20812159776687622, "learning_rate": 4.1918721122701355e-05, "loss": 0.0525, "step": 57410 }, { "epoch": 1.9412420974339903, "grad_norm": 0.25555822253227234, "learning_rate": 4.190240426237533e-05, "loss": 0.0468, "step": 57420 }, { "epoch": 1.9415801751242436, "grad_norm": 0.28654441237449646, "learning_rate": 4.188608828759019e-05, "loss": 0.0487, "step": 57430 }, { "epoch": 1.9419182528144967, "grad_norm": 0.27007874846458435, "learning_rate": 4.1869773200130246e-05, "loss": 0.0524, "step": 57440 }, { "epoch": 1.94225633050475, "grad_norm": 0.29501059651374817, "learning_rate": 4.185345900177967e-05, "loss": 0.0531, "step": 57450 }, { "epoch": 1.9425944081950033, "grad_norm": 0.285900741815567, "learning_rate": 4.183714569432258e-05, "loss": 0.0438, "step": 57460 }, { "epoch": 1.9429324858852564, "grad_norm": 0.21717539429664612, "learning_rate": 4.182083327954295e-05, "loss": 0.0468, "step": 57470 }, { "epoch": 1.9432705635755096, "grad_norm": 0.2385929822921753, "learning_rate": 4.1804521759224714e-05, "loss": 0.0492, "step": 57480 }, { "epoch": 1.943608641265763, "grad_norm": 0.18422728776931763, "learning_rate": 4.178821113515165e-05, "loss": 0.0494, "step": 57490 }, { "epoch": 1.943946718956016, "grad_norm": 0.30890533328056335, "learning_rate": 4.1771901409107476e-05, "loss": 0.048, "step": 57500 }, { "epoch": 1.9442847966462693, "grad_norm": 0.15021906793117523, "learning_rate": 4.175559258287577e-05, "loss": 0.0469, "step": 57510 }, { "epoch": 1.9446228743365226, "grad_norm": 0.33830368518829346, "learning_rate": 4.173928465824008e-05, "loss": 0.0494, "step": 57520 }, { "epoch": 1.9449609520267757, "grad_norm": 0.2151188850402832, "learning_rate": 4.17229776369838e-05, "loss": 0.0465, "step": 57530 }, { "epoch": 1.945299029717029, "grad_norm": 0.27228274941444397, "learning_rate": 4.170667152089025e-05, "loss": 0.0454, "step": 57540 }, { "epoch": 1.9456371074072822, "grad_norm": 0.2861119508743286, "learning_rate": 4.1690366311742626e-05, "loss": 0.0468, "step": 57550 }, { "epoch": 1.9459751850975353, "grad_norm": 0.45445841550827026, "learning_rate": 4.1674062011324055e-05, "loss": 0.0471, "step": 57560 }, { "epoch": 1.9463132627877886, "grad_norm": 0.19579914212226868, "learning_rate": 4.165775862141752e-05, "loss": 0.0436, "step": 57570 }, { "epoch": 1.946651340478042, "grad_norm": 0.2787173390388489, "learning_rate": 4.164145614380598e-05, "loss": 0.0499, "step": 57580 }, { "epoch": 1.946989418168295, "grad_norm": 0.2708403468132019, "learning_rate": 4.162515458027223e-05, "loss": 0.0429, "step": 57590 }, { "epoch": 1.9473274958585483, "grad_norm": 0.21161898970603943, "learning_rate": 4.160885393259899e-05, "loss": 0.0534, "step": 57600 }, { "epoch": 1.9476655735488015, "grad_norm": 0.24754264950752258, "learning_rate": 4.159255420256886e-05, "loss": 0.0536, "step": 57610 }, { "epoch": 1.9480036512390546, "grad_norm": 0.2063947170972824, "learning_rate": 4.157625539196436e-05, "loss": 0.0494, "step": 57620 }, { "epoch": 1.948341728929308, "grad_norm": 0.293686181306839, "learning_rate": 4.155995750256791e-05, "loss": 0.0512, "step": 57630 }, { "epoch": 1.9486798066195612, "grad_norm": 0.31929993629455566, "learning_rate": 4.15436605361618e-05, "loss": 0.0492, "step": 57640 }, { "epoch": 1.9490178843098143, "grad_norm": 0.1951218992471695, "learning_rate": 4.152736449452827e-05, "loss": 0.0458, "step": 57650 }, { "epoch": 1.9493559620000678, "grad_norm": 0.2120881825685501, "learning_rate": 4.151106937944943e-05, "loss": 0.0507, "step": 57660 }, { "epoch": 1.9496940396903208, "grad_norm": 0.33675751090049744, "learning_rate": 4.149477519270727e-05, "loss": 0.0507, "step": 57670 }, { "epoch": 1.950032117380574, "grad_norm": 0.2591921389102936, "learning_rate": 4.147848193608371e-05, "loss": 0.0496, "step": 57680 }, { "epoch": 1.9503701950708274, "grad_norm": 0.3513115644454956, "learning_rate": 4.146218961136056e-05, "loss": 0.0457, "step": 57690 }, { "epoch": 1.9507082727610805, "grad_norm": 0.27600643038749695, "learning_rate": 4.14458982203195e-05, "loss": 0.0533, "step": 57700 }, { "epoch": 1.9510463504513336, "grad_norm": 0.28606075048446655, "learning_rate": 4.1429607764742165e-05, "loss": 0.052, "step": 57710 }, { "epoch": 1.951384428141587, "grad_norm": 0.2372482419013977, "learning_rate": 4.141331824641003e-05, "loss": 0.0443, "step": 57720 }, { "epoch": 1.9517225058318401, "grad_norm": 0.1870453953742981, "learning_rate": 4.139702966710452e-05, "loss": 0.0504, "step": 57730 }, { "epoch": 1.9520605835220932, "grad_norm": 0.7904488444328308, "learning_rate": 4.13807420286069e-05, "loss": 0.0495, "step": 57740 }, { "epoch": 1.9523986612123467, "grad_norm": 0.25766101479530334, "learning_rate": 4.136445533269837e-05, "loss": 0.0505, "step": 57750 }, { "epoch": 1.9527367389025998, "grad_norm": 0.2804143726825714, "learning_rate": 4.134816958116003e-05, "loss": 0.053, "step": 57760 }, { "epoch": 1.953074816592853, "grad_norm": 0.26005131006240845, "learning_rate": 4.133188477577283e-05, "loss": 0.0485, "step": 57770 }, { "epoch": 1.9534128942831064, "grad_norm": 0.207711324095726, "learning_rate": 4.131560091831771e-05, "loss": 0.0509, "step": 57780 }, { "epoch": 1.9537509719733595, "grad_norm": 0.25487446784973145, "learning_rate": 4.1299318010575405e-05, "loss": 0.0507, "step": 57790 }, { "epoch": 1.9540890496636127, "grad_norm": 0.8686109781265259, "learning_rate": 4.128303605432661e-05, "loss": 0.0485, "step": 57800 }, { "epoch": 1.954427127353866, "grad_norm": 0.21486155688762665, "learning_rate": 4.126675505135188e-05, "loss": 0.0465, "step": 57810 }, { "epoch": 1.954765205044119, "grad_norm": 0.23648309707641602, "learning_rate": 4.125047500343169e-05, "loss": 0.0507, "step": 57820 }, { "epoch": 1.9551032827343724, "grad_norm": 0.25249171257019043, "learning_rate": 4.1234195912346376e-05, "loss": 0.0515, "step": 57830 }, { "epoch": 1.9554413604246257, "grad_norm": 0.40622827410697937, "learning_rate": 4.1217917779876235e-05, "loss": 0.0507, "step": 57840 }, { "epoch": 1.9557794381148788, "grad_norm": 0.5983372926712036, "learning_rate": 4.120164060780141e-05, "loss": 0.0563, "step": 57850 }, { "epoch": 1.956117515805132, "grad_norm": 0.2830110490322113, "learning_rate": 4.1185364397901935e-05, "loss": 0.0514, "step": 57860 }, { "epoch": 1.9564555934953853, "grad_norm": 0.3464638292789459, "learning_rate": 4.116908915195776e-05, "loss": 0.0451, "step": 57870 }, { "epoch": 1.9567936711856384, "grad_norm": 0.2105841040611267, "learning_rate": 4.115281487174872e-05, "loss": 0.0482, "step": 57880 }, { "epoch": 1.9571317488758917, "grad_norm": 0.5313984155654907, "learning_rate": 4.113654155905452e-05, "loss": 0.0517, "step": 57890 }, { "epoch": 1.957469826566145, "grad_norm": 0.22898666560649872, "learning_rate": 4.112026921565483e-05, "loss": 0.0505, "step": 57900 }, { "epoch": 1.957807904256398, "grad_norm": 0.1440759301185608, "learning_rate": 4.1103997843329147e-05, "loss": 0.0435, "step": 57910 }, { "epoch": 1.9581459819466513, "grad_norm": 0.28300806879997253, "learning_rate": 4.108772744385689e-05, "loss": 0.0498, "step": 57920 }, { "epoch": 1.9584840596369046, "grad_norm": 0.27919283509254456, "learning_rate": 4.1071458019017346e-05, "loss": 0.0403, "step": 57930 }, { "epoch": 1.9588221373271577, "grad_norm": 0.19537784159183502, "learning_rate": 4.1055189570589734e-05, "loss": 0.0553, "step": 57940 }, { "epoch": 1.959160215017411, "grad_norm": 0.3602067232131958, "learning_rate": 4.103892210035315e-05, "loss": 0.042, "step": 57950 }, { "epoch": 1.9594982927076643, "grad_norm": 0.22795099020004272, "learning_rate": 4.102265561008656e-05, "loss": 0.0531, "step": 57960 }, { "epoch": 1.9598363703979174, "grad_norm": 0.3539789021015167, "learning_rate": 4.100639010156885e-05, "loss": 0.0498, "step": 57970 }, { "epoch": 1.9601744480881707, "grad_norm": 0.35524263978004456, "learning_rate": 4.099012557657882e-05, "loss": 0.0505, "step": 57980 }, { "epoch": 1.960512525778424, "grad_norm": 0.30079615116119385, "learning_rate": 4.09738620368951e-05, "loss": 0.0505, "step": 57990 }, { "epoch": 1.960850603468677, "grad_norm": 2.054940938949585, "learning_rate": 4.095759948429626e-05, "loss": 0.0451, "step": 58000 }, { "epoch": 1.9611886811589303, "grad_norm": 0.29943862557411194, "learning_rate": 4.094133792056075e-05, "loss": 0.0467, "step": 58010 }, { "epoch": 1.9615267588491836, "grad_norm": 0.15587618947029114, "learning_rate": 4.092507734746688e-05, "loss": 0.0539, "step": 58020 }, { "epoch": 1.9618648365394367, "grad_norm": 0.2569241225719452, "learning_rate": 4.090881776679293e-05, "loss": 0.0487, "step": 58030 }, { "epoch": 1.96220291422969, "grad_norm": 0.33914896845817566, "learning_rate": 4.0892559180316994e-05, "loss": 0.0505, "step": 58040 }, { "epoch": 1.9625409919199432, "grad_norm": 0.2282528579235077, "learning_rate": 4.08763015898171e-05, "loss": 0.0536, "step": 58050 }, { "epoch": 1.9628790696101963, "grad_norm": 0.2079096883535385, "learning_rate": 4.0860044997071145e-05, "loss": 0.0458, "step": 58060 }, { "epoch": 1.9632171473004496, "grad_norm": 0.14153501391410828, "learning_rate": 4.084378940385693e-05, "loss": 0.0471, "step": 58070 }, { "epoch": 1.963555224990703, "grad_norm": 0.1452440321445465, "learning_rate": 4.082753481195211e-05, "loss": 0.0479, "step": 58080 }, { "epoch": 1.963893302680956, "grad_norm": 0.3053872585296631, "learning_rate": 4.0811281223134324e-05, "loss": 0.0451, "step": 58090 }, { "epoch": 1.9642313803712093, "grad_norm": 0.24133332073688507, "learning_rate": 4.0795028639181e-05, "loss": 0.0518, "step": 58100 }, { "epoch": 1.9645694580614625, "grad_norm": 0.2925991714000702, "learning_rate": 4.077877706186951e-05, "loss": 0.04, "step": 58110 }, { "epoch": 1.9649075357517156, "grad_norm": 0.30390825867652893, "learning_rate": 4.076252649297709e-05, "loss": 0.0491, "step": 58120 }, { "epoch": 1.9652456134419691, "grad_norm": 0.24634554982185364, "learning_rate": 4.07462769342809e-05, "loss": 0.0459, "step": 58130 }, { "epoch": 1.9655836911322222, "grad_norm": 0.27182209491729736, "learning_rate": 4.073002838755793e-05, "loss": 0.0469, "step": 58140 }, { "epoch": 1.9659217688224753, "grad_norm": 0.1679777204990387, "learning_rate": 4.0713780854585115e-05, "loss": 0.0461, "step": 58150 }, { "epoch": 1.9662598465127288, "grad_norm": 0.1603960245847702, "learning_rate": 4.069753433713927e-05, "loss": 0.0511, "step": 58160 }, { "epoch": 1.9665979242029819, "grad_norm": 0.18495826423168182, "learning_rate": 4.068128883699709e-05, "loss": 0.0474, "step": 58170 }, { "epoch": 1.966936001893235, "grad_norm": 0.29414501786231995, "learning_rate": 4.0665044355935155e-05, "loss": 0.0454, "step": 58180 }, { "epoch": 1.9672740795834884, "grad_norm": 0.24033576250076294, "learning_rate": 4.064880089572993e-05, "loss": 0.0475, "step": 58190 }, { "epoch": 1.9676121572737415, "grad_norm": 0.23317724466323853, "learning_rate": 4.063255845815779e-05, "loss": 0.0505, "step": 58200 }, { "epoch": 1.9679502349639946, "grad_norm": 0.2725328803062439, "learning_rate": 4.0616317044994945e-05, "loss": 0.0546, "step": 58210 }, { "epoch": 1.968288312654248, "grad_norm": 0.21537211537361145, "learning_rate": 4.0600076658017585e-05, "loss": 0.0461, "step": 58220 }, { "epoch": 1.9686263903445012, "grad_norm": 0.2206646353006363, "learning_rate": 4.0583837299001706e-05, "loss": 0.0452, "step": 58230 }, { "epoch": 1.9689644680347542, "grad_norm": 0.20876815915107727, "learning_rate": 4.056759896972322e-05, "loss": 0.046, "step": 58240 }, { "epoch": 1.9693025457250077, "grad_norm": 0.2570383846759796, "learning_rate": 4.055136167195793e-05, "loss": 0.0515, "step": 58250 }, { "epoch": 1.9696406234152608, "grad_norm": 0.1853460818529129, "learning_rate": 4.053512540748152e-05, "loss": 0.0485, "step": 58260 }, { "epoch": 1.969978701105514, "grad_norm": 0.3720983564853668, "learning_rate": 4.051889017806955e-05, "loss": 0.043, "step": 58270 }, { "epoch": 1.9703167787957674, "grad_norm": 0.2451154887676239, "learning_rate": 4.0502655985497516e-05, "loss": 0.0468, "step": 58280 }, { "epoch": 1.9706548564860205, "grad_norm": 0.16943605244159698, "learning_rate": 4.048642283154074e-05, "loss": 0.0437, "step": 58290 }, { "epoch": 1.9709929341762737, "grad_norm": 0.15205155313014984, "learning_rate": 4.0470190717974464e-05, "loss": 0.0481, "step": 58300 }, { "epoch": 1.971331011866527, "grad_norm": 0.2653583884239197, "learning_rate": 4.045395964657381e-05, "loss": 0.0491, "step": 58310 }, { "epoch": 1.97166908955678, "grad_norm": 0.20335431396961212, "learning_rate": 4.043772961911376e-05, "loss": 0.0449, "step": 58320 }, { "epoch": 1.9720071672470334, "grad_norm": 0.2894182801246643, "learning_rate": 4.042150063736924e-05, "loss": 0.0491, "step": 58330 }, { "epoch": 1.9723452449372867, "grad_norm": 0.24552388489246368, "learning_rate": 4.0405272703114976e-05, "loss": 0.0467, "step": 58340 }, { "epoch": 1.9726833226275398, "grad_norm": 0.1968293935060501, "learning_rate": 4.038904581812569e-05, "loss": 0.0455, "step": 58350 }, { "epoch": 1.973021400317793, "grad_norm": 0.2894025146961212, "learning_rate": 4.03728199841759e-05, "loss": 0.0472, "step": 58360 }, { "epoch": 1.9733594780080463, "grad_norm": 0.2676451802253723, "learning_rate": 4.0356595203040044e-05, "loss": 0.0498, "step": 58370 }, { "epoch": 1.9736975556982994, "grad_norm": 0.34956690669059753, "learning_rate": 4.0340371476492426e-05, "loss": 0.0469, "step": 58380 }, { "epoch": 1.9740356333885527, "grad_norm": 0.44550201296806335, "learning_rate": 4.032414880630727e-05, "loss": 0.0541, "step": 58390 }, { "epoch": 1.974373711078806, "grad_norm": 0.6787275671958923, "learning_rate": 4.0307927194258646e-05, "loss": 0.055, "step": 58400 }, { "epoch": 1.974711788769059, "grad_norm": 0.34062856435775757, "learning_rate": 4.029170664212054e-05, "loss": 0.0471, "step": 58410 }, { "epoch": 1.9750498664593124, "grad_norm": 0.3524474799633026, "learning_rate": 4.0275487151666794e-05, "loss": 0.0485, "step": 58420 }, { "epoch": 1.9753879441495656, "grad_norm": 0.1669236421585083, "learning_rate": 4.025926872467116e-05, "loss": 0.0502, "step": 58430 }, { "epoch": 1.9757260218398187, "grad_norm": 0.18437886238098145, "learning_rate": 4.024305136290725e-05, "loss": 0.0504, "step": 58440 }, { "epoch": 1.976064099530072, "grad_norm": 0.5145938396453857, "learning_rate": 4.022683506814858e-05, "loss": 0.0423, "step": 58450 }, { "epoch": 1.9764021772203253, "grad_norm": 0.14525893330574036, "learning_rate": 4.021061984216853e-05, "loss": 0.0521, "step": 58460 }, { "epoch": 1.9767402549105784, "grad_norm": 0.20680290460586548, "learning_rate": 4.019440568674035e-05, "loss": 0.0504, "step": 58470 }, { "epoch": 1.9770783326008317, "grad_norm": 0.15972666442394257, "learning_rate": 4.017819260363724e-05, "loss": 0.048, "step": 58480 }, { "epoch": 1.977416410291085, "grad_norm": 0.6004859805107117, "learning_rate": 4.016198059463223e-05, "loss": 0.0512, "step": 58490 }, { "epoch": 1.977754487981338, "grad_norm": 0.1846664696931839, "learning_rate": 4.0145769661498233e-05, "loss": 0.0402, "step": 58500 }, { "epoch": 1.9780925656715913, "grad_norm": 0.19147789478302002, "learning_rate": 4.012955980600804e-05, "loss": 0.0497, "step": 58510 }, { "epoch": 1.9784306433618446, "grad_norm": 0.39225155115127563, "learning_rate": 4.0113351029934355e-05, "loss": 0.0483, "step": 58520 }, { "epoch": 1.9787687210520977, "grad_norm": 0.28673553466796875, "learning_rate": 4.0097143335049714e-05, "loss": 0.0516, "step": 58530 }, { "epoch": 1.979106798742351, "grad_norm": 0.18101143836975098, "learning_rate": 4.0080936723126606e-05, "loss": 0.0505, "step": 58540 }, { "epoch": 1.9794448764326043, "grad_norm": 0.2931762933731079, "learning_rate": 4.006473119593735e-05, "loss": 0.0555, "step": 58550 }, { "epoch": 1.9797829541228573, "grad_norm": 0.3368929326534271, "learning_rate": 4.004852675525415e-05, "loss": 0.054, "step": 58560 }, { "epoch": 1.9801210318131106, "grad_norm": 0.2268727868795395, "learning_rate": 4.00323234028491e-05, "loss": 0.0477, "step": 58570 }, { "epoch": 1.980459109503364, "grad_norm": 0.2961399555206299, "learning_rate": 4.0016121140494186e-05, "loss": 0.0489, "step": 58580 }, { "epoch": 1.980797187193617, "grad_norm": 0.21740266680717468, "learning_rate": 3.999991996996122e-05, "loss": 0.0469, "step": 58590 }, { "epoch": 1.9811352648838703, "grad_norm": 0.19032901525497437, "learning_rate": 3.9983719893022e-05, "loss": 0.0423, "step": 58600 }, { "epoch": 1.9814733425741236, "grad_norm": 0.18569247424602509, "learning_rate": 3.996752091144812e-05, "loss": 0.0452, "step": 58610 }, { "epoch": 1.9818114202643766, "grad_norm": 0.32981637120246887, "learning_rate": 3.9951323027011055e-05, "loss": 0.0499, "step": 58620 }, { "epoch": 1.9821494979546301, "grad_norm": 0.3798151910305023, "learning_rate": 3.993512624148219e-05, "loss": 0.0581, "step": 58630 }, { "epoch": 1.9824875756448832, "grad_norm": 0.3156028687953949, "learning_rate": 3.99189305566328e-05, "loss": 0.0498, "step": 58640 }, { "epoch": 1.9828256533351363, "grad_norm": 0.36497804522514343, "learning_rate": 3.990273597423401e-05, "loss": 0.0464, "step": 58650 }, { "epoch": 1.9831637310253898, "grad_norm": 0.3937445282936096, "learning_rate": 3.988654249605681e-05, "loss": 0.0471, "step": 58660 }, { "epoch": 1.9835018087156429, "grad_norm": 0.2096559852361679, "learning_rate": 3.987035012387213e-05, "loss": 0.0538, "step": 58670 }, { "epoch": 1.983839886405896, "grad_norm": 0.2163160741329193, "learning_rate": 3.9854158859450735e-05, "loss": 0.0522, "step": 58680 }, { "epoch": 1.9841779640961494, "grad_norm": 0.3687051832675934, "learning_rate": 3.983796870456328e-05, "loss": 0.0531, "step": 58690 }, { "epoch": 1.9845160417864025, "grad_norm": 0.18113617599010468, "learning_rate": 3.982177966098028e-05, "loss": 0.0436, "step": 58700 }, { "epoch": 1.9848541194766556, "grad_norm": 0.18730802834033966, "learning_rate": 3.980559173047216e-05, "loss": 0.0438, "step": 58710 }, { "epoch": 1.985192197166909, "grad_norm": 0.5595775246620178, "learning_rate": 3.9789404914809176e-05, "loss": 0.0493, "step": 58720 }, { "epoch": 1.9855302748571622, "grad_norm": 0.20765872299671173, "learning_rate": 3.9773219215761544e-05, "loss": 0.0473, "step": 58730 }, { "epoch": 1.9858683525474154, "grad_norm": 0.20990991592407227, "learning_rate": 3.975703463509928e-05, "loss": 0.0476, "step": 58740 }, { "epoch": 1.9862064302376687, "grad_norm": 0.2639017701148987, "learning_rate": 3.974085117459231e-05, "loss": 0.0413, "step": 58750 }, { "epoch": 1.9865445079279218, "grad_norm": 0.21612010896205902, "learning_rate": 3.972466883601043e-05, "loss": 0.0442, "step": 58760 }, { "epoch": 1.986882585618175, "grad_norm": 0.26305854320526123, "learning_rate": 3.9708487621123316e-05, "loss": 0.0466, "step": 58770 }, { "epoch": 1.9872206633084284, "grad_norm": 0.24797150492668152, "learning_rate": 3.969230753170051e-05, "loss": 0.0463, "step": 58780 }, { "epoch": 1.9875587409986815, "grad_norm": 0.32022660970687866, "learning_rate": 3.967612856951146e-05, "loss": 0.0515, "step": 58790 }, { "epoch": 1.9878968186889348, "grad_norm": 0.3155270218849182, "learning_rate": 3.9659950736325476e-05, "loss": 0.0525, "step": 58800 }, { "epoch": 1.988234896379188, "grad_norm": 0.1919345110654831, "learning_rate": 3.964377403391174e-05, "loss": 0.0473, "step": 58810 }, { "epoch": 1.9885729740694411, "grad_norm": 0.2302958220243454, "learning_rate": 3.9627598464039303e-05, "loss": 0.0506, "step": 58820 }, { "epoch": 1.9889110517596944, "grad_norm": 0.43164053559303284, "learning_rate": 3.9611424028477096e-05, "loss": 0.0448, "step": 58830 }, { "epoch": 1.9892491294499477, "grad_norm": 0.29276737570762634, "learning_rate": 3.959525072899394e-05, "loss": 0.0462, "step": 58840 }, { "epoch": 1.9895872071402008, "grad_norm": 0.2253478318452835, "learning_rate": 3.9579078567358506e-05, "loss": 0.0479, "step": 58850 }, { "epoch": 1.989925284830454, "grad_norm": 0.29235902428627014, "learning_rate": 3.956290754533939e-05, "loss": 0.0473, "step": 58860 }, { "epoch": 1.9902633625207073, "grad_norm": 0.2205764800310135, "learning_rate": 3.954673766470501e-05, "loss": 0.0511, "step": 58870 }, { "epoch": 1.9906014402109604, "grad_norm": 0.3267636001110077, "learning_rate": 3.9530568927223674e-05, "loss": 0.0491, "step": 58880 }, { "epoch": 1.9909395179012137, "grad_norm": 0.28700482845306396, "learning_rate": 3.95144013346636e-05, "loss": 0.047, "step": 58890 }, { "epoch": 1.991277595591467, "grad_norm": 0.3686405420303345, "learning_rate": 3.949823488879282e-05, "loss": 0.0509, "step": 58900 }, { "epoch": 1.99161567328172, "grad_norm": 0.6919625997543335, "learning_rate": 3.9482069591379276e-05, "loss": 0.0497, "step": 58910 }, { "epoch": 1.9919537509719734, "grad_norm": 0.24731716513633728, "learning_rate": 3.9465905444190784e-05, "loss": 0.0507, "step": 58920 }, { "epoch": 1.9922918286622266, "grad_norm": 0.3056337237358093, "learning_rate": 3.944974244899505e-05, "loss": 0.0509, "step": 58930 }, { "epoch": 1.9926299063524797, "grad_norm": 0.1971297562122345, "learning_rate": 3.9433580607559616e-05, "loss": 0.049, "step": 58940 }, { "epoch": 1.992967984042733, "grad_norm": 0.2180284708738327, "learning_rate": 3.941741992165193e-05, "loss": 0.0512, "step": 58950 }, { "epoch": 1.9933060617329863, "grad_norm": 0.3060995638370514, "learning_rate": 3.940126039303928e-05, "loss": 0.0465, "step": 58960 }, { "epoch": 1.9936441394232394, "grad_norm": 0.2411472052335739, "learning_rate": 3.938510202348886e-05, "loss": 0.0479, "step": 58970 }, { "epoch": 1.9939822171134927, "grad_norm": 0.8241722583770752, "learning_rate": 3.93689448147677e-05, "loss": 0.0492, "step": 58980 }, { "epoch": 1.994320294803746, "grad_norm": 0.3763771653175354, "learning_rate": 3.935278876864278e-05, "loss": 0.0468, "step": 58990 }, { "epoch": 1.994658372493999, "grad_norm": 0.1762845814228058, "learning_rate": 3.933663388688087e-05, "loss": 0.0424, "step": 59000 }, { "epoch": 1.9949964501842523, "grad_norm": 0.1538168489933014, "learning_rate": 3.9320480171248645e-05, "loss": 0.0437, "step": 59010 }, { "epoch": 1.9953345278745056, "grad_norm": 0.20479464530944824, "learning_rate": 3.930432762351265e-05, "loss": 0.0517, "step": 59020 }, { "epoch": 1.9956726055647587, "grad_norm": 0.2797839641571045, "learning_rate": 3.928817624543931e-05, "loss": 0.0424, "step": 59030 }, { "epoch": 1.996010683255012, "grad_norm": 0.3313273787498474, "learning_rate": 3.9272026038794885e-05, "loss": 0.0504, "step": 59040 }, { "epoch": 1.9963487609452653, "grad_norm": 0.21202830970287323, "learning_rate": 3.9255877005345586e-05, "loss": 0.0503, "step": 59050 }, { "epoch": 1.9966868386355183, "grad_norm": 0.3581709861755371, "learning_rate": 3.923972914685743e-05, "loss": 0.0462, "step": 59060 }, { "epoch": 1.9970249163257716, "grad_norm": 0.26422175765037537, "learning_rate": 3.922358246509631e-05, "loss": 0.0454, "step": 59070 }, { "epoch": 1.997362994016025, "grad_norm": 0.2494363635778427, "learning_rate": 3.920743696182801e-05, "loss": 0.0504, "step": 59080 }, { "epoch": 1.997701071706278, "grad_norm": 0.24952363967895508, "learning_rate": 3.9191292638818164e-05, "loss": 0.0428, "step": 59090 }, { "epoch": 1.9980391493965315, "grad_norm": 0.29318657517433167, "learning_rate": 3.9175149497832316e-05, "loss": 0.0492, "step": 59100 }, { "epoch": 1.9983772270867846, "grad_norm": 0.2554952800273895, "learning_rate": 3.915900754063584e-05, "loss": 0.0509, "step": 59110 }, { "epoch": 1.9987153047770376, "grad_norm": 0.23342551290988922, "learning_rate": 3.9142866768994014e-05, "loss": 0.0452, "step": 59120 }, { "epoch": 1.9990533824672911, "grad_norm": 0.4030650556087494, "learning_rate": 3.912672718467193e-05, "loss": 0.0478, "step": 59130 }, { "epoch": 1.9993914601575442, "grad_norm": 0.16198982298374176, "learning_rate": 3.911058878943463e-05, "loss": 0.0523, "step": 59140 }, { "epoch": 1.9997295378477973, "grad_norm": 0.27891477942466736, "learning_rate": 3.9094451585046974e-05, "loss": 0.0441, "step": 59150 }, { "epoch": 2.000067615538051, "grad_norm": 0.2510248124599457, "learning_rate": 3.90783155732737e-05, "loss": 0.0453, "step": 59160 }, { "epoch": 2.000405693228304, "grad_norm": 0.38348841667175293, "learning_rate": 3.9062180755879373e-05, "loss": 0.0395, "step": 59170 }, { "epoch": 2.000743770918557, "grad_norm": 0.23441874980926514, "learning_rate": 3.904604713462855e-05, "loss": 0.0509, "step": 59180 }, { "epoch": 2.0010818486088104, "grad_norm": 0.1685921549797058, "learning_rate": 3.902991471128554e-05, "loss": 0.0368, "step": 59190 }, { "epoch": 2.0014199262990635, "grad_norm": 0.340258926153183, "learning_rate": 3.901378348761456e-05, "loss": 0.0515, "step": 59200 }, { "epoch": 2.0017580039893166, "grad_norm": 0.23856429755687714, "learning_rate": 3.899765346537971e-05, "loss": 0.047, "step": 59210 }, { "epoch": 2.00209608167957, "grad_norm": 0.3347926437854767, "learning_rate": 3.898152464634493e-05, "loss": 0.0552, "step": 59220 }, { "epoch": 2.002434159369823, "grad_norm": 0.17945663630962372, "learning_rate": 3.896539703227402e-05, "loss": 0.0433, "step": 59230 }, { "epoch": 2.0027722370600762, "grad_norm": 1.3334746360778809, "learning_rate": 3.894927062493073e-05, "loss": 0.0464, "step": 59240 }, { "epoch": 2.0031103147503297, "grad_norm": 0.4279509484767914, "learning_rate": 3.893314542607859e-05, "loss": 0.0451, "step": 59250 }, { "epoch": 2.003448392440583, "grad_norm": 0.2979460656642914, "learning_rate": 3.891702143748103e-05, "loss": 0.0526, "step": 59260 }, { "epoch": 2.003786470130836, "grad_norm": 0.2573980987071991, "learning_rate": 3.890089866090133e-05, "loss": 0.0503, "step": 59270 }, { "epoch": 2.0041245478210894, "grad_norm": 0.258450984954834, "learning_rate": 3.888477709810267e-05, "loss": 0.0499, "step": 59280 }, { "epoch": 2.0044626255113425, "grad_norm": 0.2536048889160156, "learning_rate": 3.886865675084806e-05, "loss": 0.0538, "step": 59290 }, { "epoch": 2.0048007032015955, "grad_norm": 0.20180074870586395, "learning_rate": 3.885253762090043e-05, "loss": 0.0492, "step": 59300 }, { "epoch": 2.005138780891849, "grad_norm": 0.2743423581123352, "learning_rate": 3.8836419710022534e-05, "loss": 0.0466, "step": 59310 }, { "epoch": 2.005476858582102, "grad_norm": 0.20299535989761353, "learning_rate": 3.882030301997698e-05, "loss": 0.0492, "step": 59320 }, { "epoch": 2.005814936272355, "grad_norm": 0.26495683193206787, "learning_rate": 3.880418755252629e-05, "loss": 0.0412, "step": 59330 }, { "epoch": 2.0061530139626087, "grad_norm": 0.23177188634872437, "learning_rate": 3.87880733094328e-05, "loss": 0.0459, "step": 59340 }, { "epoch": 2.0064910916528618, "grad_norm": 0.24210457503795624, "learning_rate": 3.877196029245877e-05, "loss": 0.0475, "step": 59350 }, { "epoch": 2.006829169343115, "grad_norm": 0.2465326488018036, "learning_rate": 3.875584850336627e-05, "loss": 0.0504, "step": 59360 }, { "epoch": 2.0071672470333684, "grad_norm": 0.17507679760456085, "learning_rate": 3.873973794391728e-05, "loss": 0.0532, "step": 59370 }, { "epoch": 2.0075053247236214, "grad_norm": 0.24775531888008118, "learning_rate": 3.872362861587361e-05, "loss": 0.0475, "step": 59380 }, { "epoch": 2.0078434024138745, "grad_norm": 0.4437977969646454, "learning_rate": 3.870752052099698e-05, "loss": 0.0472, "step": 59390 }, { "epoch": 2.008181480104128, "grad_norm": 0.20917150378227234, "learning_rate": 3.8691413661048916e-05, "loss": 0.0461, "step": 59400 }, { "epoch": 2.008519557794381, "grad_norm": 0.14811454713344574, "learning_rate": 3.867530803779085e-05, "loss": 0.0461, "step": 59410 }, { "epoch": 2.0088576354846346, "grad_norm": 0.19781841337680817, "learning_rate": 3.8659203652984054e-05, "loss": 0.0486, "step": 59420 }, { "epoch": 2.0091957131748877, "grad_norm": 0.27048322558403015, "learning_rate": 3.864310050838972e-05, "loss": 0.0531, "step": 59430 }, { "epoch": 2.0095337908651407, "grad_norm": 0.2641300559043884, "learning_rate": 3.8626998605768835e-05, "loss": 0.05, "step": 59440 }, { "epoch": 2.0098718685553942, "grad_norm": 0.21834364533424377, "learning_rate": 3.8610897946882294e-05, "loss": 0.0452, "step": 59450 }, { "epoch": 2.0102099462456473, "grad_norm": 0.20394758880138397, "learning_rate": 3.859479853349082e-05, "loss": 0.0501, "step": 59460 }, { "epoch": 2.0105480239359004, "grad_norm": 0.23308517038822174, "learning_rate": 3.857870036735503e-05, "loss": 0.0516, "step": 59470 }, { "epoch": 2.010886101626154, "grad_norm": 0.1733718365430832, "learning_rate": 3.856260345023538e-05, "loss": 0.0485, "step": 59480 }, { "epoch": 2.011224179316407, "grad_norm": 0.2124934047460556, "learning_rate": 3.854650778389224e-05, "loss": 0.0492, "step": 59490 }, { "epoch": 2.01156225700666, "grad_norm": 0.3171559274196625, "learning_rate": 3.8530413370085786e-05, "loss": 0.0446, "step": 59500 }, { "epoch": 2.0119003346969135, "grad_norm": 0.5932653546333313, "learning_rate": 3.851432021057608e-05, "loss": 0.0423, "step": 59510 }, { "epoch": 2.0122384123871666, "grad_norm": 0.388690710067749, "learning_rate": 3.849822830712306e-05, "loss": 0.0417, "step": 59520 }, { "epoch": 2.0125764900774197, "grad_norm": 0.21131527423858643, "learning_rate": 3.848213766148649e-05, "loss": 0.0497, "step": 59530 }, { "epoch": 2.012914567767673, "grad_norm": 0.15006928145885468, "learning_rate": 3.846604827542602e-05, "loss": 0.0496, "step": 59540 }, { "epoch": 2.0132526454579263, "grad_norm": 0.43168774247169495, "learning_rate": 3.844996015070117e-05, "loss": 0.0525, "step": 59550 }, { "epoch": 2.0135907231481793, "grad_norm": 0.23701438307762146, "learning_rate": 3.843387328907132e-05, "loss": 0.0502, "step": 59560 }, { "epoch": 2.013928800838433, "grad_norm": 0.1925005465745926, "learning_rate": 3.8417787692295696e-05, "loss": 0.0431, "step": 59570 }, { "epoch": 2.014266878528686, "grad_norm": 0.2524566948413849, "learning_rate": 3.84017033621334e-05, "loss": 0.0473, "step": 59580 }, { "epoch": 2.014604956218939, "grad_norm": 0.19786520302295685, "learning_rate": 3.838562030034338e-05, "loss": 0.0505, "step": 59590 }, { "epoch": 2.0149430339091925, "grad_norm": 0.2719980478286743, "learning_rate": 3.836953850868447e-05, "loss": 0.05, "step": 59600 }, { "epoch": 2.0152811115994456, "grad_norm": 0.24437233805656433, "learning_rate": 3.835345798891533e-05, "loss": 0.0449, "step": 59610 }, { "epoch": 2.0156191892896986, "grad_norm": 0.21955570578575134, "learning_rate": 3.833737874279452e-05, "loss": 0.0471, "step": 59620 }, { "epoch": 2.015957266979952, "grad_norm": 0.24585272371768951, "learning_rate": 3.8321300772080433e-05, "loss": 0.0444, "step": 59630 }, { "epoch": 2.016295344670205, "grad_norm": 0.1899004727602005, "learning_rate": 3.8305224078531345e-05, "loss": 0.0461, "step": 59640 }, { "epoch": 2.0166334223604583, "grad_norm": 0.25491395592689514, "learning_rate": 3.828914866390537e-05, "loss": 0.0443, "step": 59650 }, { "epoch": 2.016971500050712, "grad_norm": 0.28252771496772766, "learning_rate": 3.827307452996048e-05, "loss": 0.0487, "step": 59660 }, { "epoch": 2.017309577740965, "grad_norm": 0.2738393545150757, "learning_rate": 3.825700167845454e-05, "loss": 0.0525, "step": 59670 }, { "epoch": 2.017647655431218, "grad_norm": 0.16650192439556122, "learning_rate": 3.8240930111145214e-05, "loss": 0.0491, "step": 59680 }, { "epoch": 2.0179857331214714, "grad_norm": 0.2269277274608612, "learning_rate": 3.8224859829790116e-05, "loss": 0.0435, "step": 59690 }, { "epoch": 2.0183238108117245, "grad_norm": 0.20055720210075378, "learning_rate": 3.820879083614664e-05, "loss": 0.0475, "step": 59700 }, { "epoch": 2.0186618885019776, "grad_norm": 0.19100548326969147, "learning_rate": 3.819272313197208e-05, "loss": 0.0451, "step": 59710 }, { "epoch": 2.018999966192231, "grad_norm": 0.1737716943025589, "learning_rate": 3.817665671902355e-05, "loss": 0.0438, "step": 59720 }, { "epoch": 2.019338043882484, "grad_norm": 0.658863365650177, "learning_rate": 3.816059159905808e-05, "loss": 0.0539, "step": 59730 }, { "epoch": 2.0196761215727372, "grad_norm": 0.4727608859539032, "learning_rate": 3.8144527773832476e-05, "loss": 0.0436, "step": 59740 }, { "epoch": 2.0200141992629908, "grad_norm": 0.3021847605705261, "learning_rate": 3.812846524510352e-05, "loss": 0.0505, "step": 59750 }, { "epoch": 2.020352276953244, "grad_norm": 0.24120813608169556, "learning_rate": 3.811240401462775e-05, "loss": 0.053, "step": 59760 }, { "epoch": 2.020690354643497, "grad_norm": 0.19827452301979065, "learning_rate": 3.809634408416162e-05, "loss": 0.0505, "step": 59770 }, { "epoch": 2.0210284323337504, "grad_norm": 0.24908863008022308, "learning_rate": 3.8080285455461394e-05, "loss": 0.0497, "step": 59780 }, { "epoch": 2.0213665100240035, "grad_norm": 0.18984371423721313, "learning_rate": 3.806422813028323e-05, "loss": 0.0484, "step": 59790 }, { "epoch": 2.0217045877142565, "grad_norm": 1.8522876501083374, "learning_rate": 3.8048172110383114e-05, "loss": 0.0525, "step": 59800 }, { "epoch": 2.02204266540451, "grad_norm": 0.28675681352615356, "learning_rate": 3.8032117397516944e-05, "loss": 0.0468, "step": 59810 }, { "epoch": 2.022380743094763, "grad_norm": 0.2870938181877136, "learning_rate": 3.801606399344042e-05, "loss": 0.049, "step": 59820 }, { "epoch": 2.022718820785016, "grad_norm": 0.19541887938976288, "learning_rate": 3.800001189990913e-05, "loss": 0.0515, "step": 59830 }, { "epoch": 2.0230568984752697, "grad_norm": 0.2596045434474945, "learning_rate": 3.798396111867847e-05, "loss": 0.046, "step": 59840 }, { "epoch": 2.0233949761655228, "grad_norm": 0.25724437832832336, "learning_rate": 3.796791165150378e-05, "loss": 0.0458, "step": 59850 }, { "epoch": 2.023733053855776, "grad_norm": 0.2317829132080078, "learning_rate": 3.795186350014018e-05, "loss": 0.0506, "step": 59860 }, { "epoch": 2.0240711315460294, "grad_norm": 0.28329166769981384, "learning_rate": 3.793581666634266e-05, "loss": 0.0503, "step": 59870 }, { "epoch": 2.0244092092362824, "grad_norm": 0.629423975944519, "learning_rate": 3.791977115186609e-05, "loss": 0.0421, "step": 59880 }, { "epoch": 2.0247472869265355, "grad_norm": 0.3869278132915497, "learning_rate": 3.79037269584652e-05, "loss": 0.0504, "step": 59890 }, { "epoch": 2.025085364616789, "grad_norm": 0.19867180287837982, "learning_rate": 3.7887684087894545e-05, "loss": 0.0481, "step": 59900 }, { "epoch": 2.025423442307042, "grad_norm": 0.2193138301372528, "learning_rate": 3.787164254190854e-05, "loss": 0.0482, "step": 59910 }, { "epoch": 2.0257615199972956, "grad_norm": 0.42918115854263306, "learning_rate": 3.7855602322261473e-05, "loss": 0.0438, "step": 59920 }, { "epoch": 2.0260995976875487, "grad_norm": 0.2795867919921875, "learning_rate": 3.783956343070746e-05, "loss": 0.0473, "step": 59930 }, { "epoch": 2.0264376753778017, "grad_norm": 0.23745262622833252, "learning_rate": 3.7823525869000526e-05, "loss": 0.0486, "step": 59940 }, { "epoch": 2.0267757530680552, "grad_norm": 0.32501181960105896, "learning_rate": 3.78074896388945e-05, "loss": 0.0521, "step": 59950 }, { "epoch": 2.0271138307583083, "grad_norm": 0.367501437664032, "learning_rate": 3.7791454742143075e-05, "loss": 0.0455, "step": 59960 }, { "epoch": 2.0274519084485614, "grad_norm": 0.5616281628608704, "learning_rate": 3.7775421180499795e-05, "loss": 0.0514, "step": 59970 }, { "epoch": 2.027789986138815, "grad_norm": 0.30234670639038086, "learning_rate": 3.7759388955718086e-05, "loss": 0.0452, "step": 59980 }, { "epoch": 2.028128063829068, "grad_norm": 0.20024579763412476, "learning_rate": 3.774335806955117e-05, "loss": 0.0487, "step": 59990 }, { "epoch": 2.028466141519321, "grad_norm": 0.17433346807956696, "learning_rate": 3.7727328523752215e-05, "loss": 0.0485, "step": 60000 }, { "epoch": 2.0288042192095745, "grad_norm": 0.3637717664241791, "learning_rate": 3.771130032007415e-05, "loss": 0.0511, "step": 60010 }, { "epoch": 2.0291422968998276, "grad_norm": 0.2721973955631256, "learning_rate": 3.769527346026982e-05, "loss": 0.049, "step": 60020 }, { "epoch": 2.0294803745900807, "grad_norm": 0.2915843427181244, "learning_rate": 3.767924794609188e-05, "loss": 0.049, "step": 60030 }, { "epoch": 2.029818452280334, "grad_norm": 0.23398852348327637, "learning_rate": 3.7663223779292854e-05, "loss": 0.0443, "step": 60040 }, { "epoch": 2.0301565299705873, "grad_norm": 0.2729816138744354, "learning_rate": 3.764720096162512e-05, "loss": 0.0489, "step": 60050 }, { "epoch": 2.0304946076608403, "grad_norm": 0.3277338147163391, "learning_rate": 3.763117949484092e-05, "loss": 0.0468, "step": 60060 }, { "epoch": 2.030832685351094, "grad_norm": 0.18956156075000763, "learning_rate": 3.7615159380692344e-05, "loss": 0.0505, "step": 60070 }, { "epoch": 2.031170763041347, "grad_norm": 0.2061312347650528, "learning_rate": 3.759914062093132e-05, "loss": 0.0446, "step": 60080 }, { "epoch": 2.0315088407316, "grad_norm": 0.18496093153953552, "learning_rate": 3.7583123217309615e-05, "loss": 0.045, "step": 60090 }, { "epoch": 2.0318469184218535, "grad_norm": 0.3479357659816742, "learning_rate": 3.7567107171578904e-05, "loss": 0.0549, "step": 60100 }, { "epoch": 2.0321849961121066, "grad_norm": 0.19584187865257263, "learning_rate": 3.755109248549066e-05, "loss": 0.0515, "step": 60110 }, { "epoch": 2.0325230738023596, "grad_norm": 0.23770713806152344, "learning_rate": 3.75350791607962e-05, "loss": 0.048, "step": 60120 }, { "epoch": 2.032861151492613, "grad_norm": 0.2623923420906067, "learning_rate": 3.751906719924676e-05, "loss": 0.0523, "step": 60130 }, { "epoch": 2.033199229182866, "grad_norm": 0.22106285393238068, "learning_rate": 3.750305660259337e-05, "loss": 0.0472, "step": 60140 }, { "epoch": 2.0335373068731193, "grad_norm": 0.17054781317710876, "learning_rate": 3.748704737258693e-05, "loss": 0.0437, "step": 60150 }, { "epoch": 2.033875384563373, "grad_norm": 0.24079467356204987, "learning_rate": 3.747103951097816e-05, "loss": 0.0491, "step": 60160 }, { "epoch": 2.034213462253626, "grad_norm": 0.33151280879974365, "learning_rate": 3.745503301951768e-05, "loss": 0.0481, "step": 60170 }, { "epoch": 2.034551539943879, "grad_norm": 0.4256715476512909, "learning_rate": 3.743902789995592e-05, "loss": 0.0498, "step": 60180 }, { "epoch": 2.0348896176341325, "grad_norm": 0.15048839151859283, "learning_rate": 3.7423024154043183e-05, "loss": 0.048, "step": 60190 }, { "epoch": 2.0352276953243855, "grad_norm": 0.3013647794723511, "learning_rate": 3.740702178352963e-05, "loss": 0.0499, "step": 60200 }, { "epoch": 2.0355657730146386, "grad_norm": 0.19053691625595093, "learning_rate": 3.739102079016523e-05, "loss": 0.0463, "step": 60210 }, { "epoch": 2.035903850704892, "grad_norm": 0.22679553925991058, "learning_rate": 3.7375021175699846e-05, "loss": 0.0452, "step": 60220 }, { "epoch": 2.036241928395145, "grad_norm": 0.2882853150367737, "learning_rate": 3.735902294188317e-05, "loss": 0.0472, "step": 60230 }, { "epoch": 2.0365800060853982, "grad_norm": 0.18168288469314575, "learning_rate": 3.734302609046473e-05, "loss": 0.0449, "step": 60240 }, { "epoch": 2.0369180837756518, "grad_norm": 0.2630862295627594, "learning_rate": 3.732703062319391e-05, "loss": 0.0467, "step": 60250 }, { "epoch": 2.037256161465905, "grad_norm": 0.6503587961196899, "learning_rate": 3.731103654181999e-05, "loss": 0.054, "step": 60260 }, { "epoch": 2.037594239156158, "grad_norm": 1.0607470273971558, "learning_rate": 3.729504384809203e-05, "loss": 0.0492, "step": 60270 }, { "epoch": 2.0379323168464114, "grad_norm": 0.2536330223083496, "learning_rate": 3.727905254375898e-05, "loss": 0.0449, "step": 60280 }, { "epoch": 2.0382703945366645, "grad_norm": 0.19498462975025177, "learning_rate": 3.7263062630569614e-05, "loss": 0.0439, "step": 60290 }, { "epoch": 2.0386084722269175, "grad_norm": 0.1592760682106018, "learning_rate": 3.724707411027256e-05, "loss": 0.0459, "step": 60300 }, { "epoch": 2.038946549917171, "grad_norm": 0.38032177090644836, "learning_rate": 3.723108698461631e-05, "loss": 0.0457, "step": 60310 }, { "epoch": 2.039284627607424, "grad_norm": 0.17130537331104279, "learning_rate": 3.72151012553492e-05, "loss": 0.0524, "step": 60320 }, { "epoch": 2.039622705297677, "grad_norm": 0.2303621917963028, "learning_rate": 3.719911692421939e-05, "loss": 0.0471, "step": 60330 }, { "epoch": 2.0399607829879307, "grad_norm": 0.2474292516708374, "learning_rate": 3.71831339929749e-05, "loss": 0.0491, "step": 60340 }, { "epoch": 2.040298860678184, "grad_norm": 0.39822426438331604, "learning_rate": 3.716715246336362e-05, "loss": 0.0407, "step": 60350 }, { "epoch": 2.0406369383684373, "grad_norm": 0.2459993213415146, "learning_rate": 3.715117233713324e-05, "loss": 0.047, "step": 60360 }, { "epoch": 2.0409750160586904, "grad_norm": 0.28209030628204346, "learning_rate": 3.713519361603135e-05, "loss": 0.052, "step": 60370 }, { "epoch": 2.0413130937489434, "grad_norm": 0.1607365757226944, "learning_rate": 3.711921630180532e-05, "loss": 0.046, "step": 60380 }, { "epoch": 2.041651171439197, "grad_norm": 0.20879854261875153, "learning_rate": 3.710324039620245e-05, "loss": 0.051, "step": 60390 }, { "epoch": 2.04198924912945, "grad_norm": 0.1281188577413559, "learning_rate": 3.708726590096982e-05, "loss": 0.0432, "step": 60400 }, { "epoch": 2.042327326819703, "grad_norm": 0.21171773970127106, "learning_rate": 3.707129281785437e-05, "loss": 0.0454, "step": 60410 }, { "epoch": 2.0426654045099566, "grad_norm": 0.27944478392601013, "learning_rate": 3.705532114860291e-05, "loss": 0.046, "step": 60420 }, { "epoch": 2.0430034822002097, "grad_norm": 0.32198643684387207, "learning_rate": 3.703935089496207e-05, "loss": 0.0464, "step": 60430 }, { "epoch": 2.0433415598904627, "grad_norm": 0.23996931314468384, "learning_rate": 3.702338205867829e-05, "loss": 0.0524, "step": 60440 }, { "epoch": 2.0436796375807162, "grad_norm": 0.3687593638896942, "learning_rate": 3.7007414641497976e-05, "loss": 0.0478, "step": 60450 }, { "epoch": 2.0440177152709693, "grad_norm": 0.17578014731407166, "learning_rate": 3.699144864516726e-05, "loss": 0.0461, "step": 60460 }, { "epoch": 2.0443557929612224, "grad_norm": 0.30276578664779663, "learning_rate": 3.6975484071432153e-05, "loss": 0.0503, "step": 60470 }, { "epoch": 2.044693870651476, "grad_norm": 0.23341141641139984, "learning_rate": 3.695952092203854e-05, "loss": 0.0504, "step": 60480 }, { "epoch": 2.045031948341729, "grad_norm": 0.7443271279335022, "learning_rate": 3.69435591987321e-05, "loss": 0.0465, "step": 60490 }, { "epoch": 2.045370026031982, "grad_norm": 0.19399479031562805, "learning_rate": 3.6927598903258374e-05, "loss": 0.0486, "step": 60500 }, { "epoch": 2.0457081037222355, "grad_norm": 0.38596057891845703, "learning_rate": 3.691164003736281e-05, "loss": 0.0526, "step": 60510 }, { "epoch": 2.0460461814124886, "grad_norm": 0.4005574584007263, "learning_rate": 3.689568260279061e-05, "loss": 0.0418, "step": 60520 }, { "epoch": 2.0463842591027417, "grad_norm": 0.24841998517513275, "learning_rate": 3.687972660128686e-05, "loss": 0.0498, "step": 60530 }, { "epoch": 2.046722336792995, "grad_norm": 0.20374688506126404, "learning_rate": 3.686377203459648e-05, "loss": 0.0445, "step": 60540 }, { "epoch": 2.0470604144832483, "grad_norm": 0.28211653232574463, "learning_rate": 3.6847818904464226e-05, "loss": 0.0484, "step": 60550 }, { "epoch": 2.0473984921735013, "grad_norm": 0.45973342657089233, "learning_rate": 3.683186721263474e-05, "loss": 0.0536, "step": 60560 }, { "epoch": 2.047736569863755, "grad_norm": 0.21698997914791107, "learning_rate": 3.6815916960852435e-05, "loss": 0.047, "step": 60570 }, { "epoch": 2.048074647554008, "grad_norm": 0.15852317214012146, "learning_rate": 3.679996815086165e-05, "loss": 0.0489, "step": 60580 }, { "epoch": 2.048412725244261, "grad_norm": 0.18931938707828522, "learning_rate": 3.678402078440649e-05, "loss": 0.04, "step": 60590 }, { "epoch": 2.0487508029345145, "grad_norm": 0.25371986627578735, "learning_rate": 3.676807486323096e-05, "loss": 0.046, "step": 60600 }, { "epoch": 2.0490888806247676, "grad_norm": 0.3628247082233429, "learning_rate": 3.6752130389078864e-05, "loss": 0.0518, "step": 60610 }, { "epoch": 2.0494269583150206, "grad_norm": 0.22296161949634552, "learning_rate": 3.673618736369388e-05, "loss": 0.0435, "step": 60620 }, { "epoch": 2.049765036005274, "grad_norm": 0.1691124588251114, "learning_rate": 3.6720245788819474e-05, "loss": 0.0479, "step": 60630 }, { "epoch": 2.0501031136955272, "grad_norm": 0.3256784677505493, "learning_rate": 3.6704305666199044e-05, "loss": 0.0456, "step": 60640 }, { "epoch": 2.0504411913857803, "grad_norm": 0.45855048298835754, "learning_rate": 3.668836699757576e-05, "loss": 0.0515, "step": 60650 }, { "epoch": 2.050779269076034, "grad_norm": 0.3077513873577118, "learning_rate": 3.667242978469265e-05, "loss": 0.0428, "step": 60660 }, { "epoch": 2.051117346766287, "grad_norm": 0.22682498395442963, "learning_rate": 3.6656494029292575e-05, "loss": 0.0503, "step": 60670 }, { "epoch": 2.05145542445654, "grad_norm": 0.1928618848323822, "learning_rate": 3.664055973311825e-05, "loss": 0.0467, "step": 60680 }, { "epoch": 2.0517935021467935, "grad_norm": 0.5832582116127014, "learning_rate": 3.662462689791221e-05, "loss": 0.0513, "step": 60690 }, { "epoch": 2.0521315798370465, "grad_norm": 0.2283712923526764, "learning_rate": 3.660869552541689e-05, "loss": 0.0404, "step": 60700 }, { "epoch": 2.0524696575272996, "grad_norm": 0.27300891280174255, "learning_rate": 3.659276561737448e-05, "loss": 0.0455, "step": 60710 }, { "epoch": 2.052807735217553, "grad_norm": 0.31649112701416016, "learning_rate": 3.657683717552708e-05, "loss": 0.0492, "step": 60720 }, { "epoch": 2.053145812907806, "grad_norm": 1.026240348815918, "learning_rate": 3.6560910201616574e-05, "loss": 0.0413, "step": 60730 }, { "epoch": 2.0534838905980592, "grad_norm": 0.3388078212738037, "learning_rate": 3.654498469738472e-05, "loss": 0.0469, "step": 60740 }, { "epoch": 2.0538219682883128, "grad_norm": 0.18180735409259796, "learning_rate": 3.6529060664573126e-05, "loss": 0.0464, "step": 60750 }, { "epoch": 2.054160045978566, "grad_norm": 0.2254558801651001, "learning_rate": 3.6513138104923176e-05, "loss": 0.0468, "step": 60760 }, { "epoch": 2.054498123668819, "grad_norm": 0.2265077531337738, "learning_rate": 3.6497217020176176e-05, "loss": 0.0424, "step": 60770 }, { "epoch": 2.0548362013590724, "grad_norm": 0.18373245000839233, "learning_rate": 3.648129741207323e-05, "loss": 0.0476, "step": 60780 }, { "epoch": 2.0551742790493255, "grad_norm": 0.24448376893997192, "learning_rate": 3.646537928235527e-05, "loss": 0.046, "step": 60790 }, { "epoch": 2.0555123567395785, "grad_norm": 0.3062826991081238, "learning_rate": 3.6449462632763075e-05, "loss": 0.0458, "step": 60800 }, { "epoch": 2.055850434429832, "grad_norm": 0.3856217563152313, "learning_rate": 3.643354746503729e-05, "loss": 0.0481, "step": 60810 }, { "epoch": 2.056188512120085, "grad_norm": 0.22359417378902435, "learning_rate": 3.6417633780918345e-05, "loss": 0.0457, "step": 60820 }, { "epoch": 2.056526589810338, "grad_norm": 0.2743578553199768, "learning_rate": 3.640172158214655e-05, "loss": 0.0459, "step": 60830 }, { "epoch": 2.0568646675005917, "grad_norm": 0.2714056968688965, "learning_rate": 3.6385810870462035e-05, "loss": 0.0484, "step": 60840 }, { "epoch": 2.057202745190845, "grad_norm": 0.40877628326416016, "learning_rate": 3.636990164760479e-05, "loss": 0.0449, "step": 60850 }, { "epoch": 2.057540822881098, "grad_norm": 0.38871607184410095, "learning_rate": 3.6353993915314596e-05, "loss": 0.0542, "step": 60860 }, { "epoch": 2.0578789005713514, "grad_norm": 0.3893790543079376, "learning_rate": 3.6338087675331124e-05, "loss": 0.0462, "step": 60870 }, { "epoch": 2.0582169782616044, "grad_norm": 0.23782950639724731, "learning_rate": 3.632218292939383e-05, "loss": 0.0514, "step": 60880 }, { "epoch": 2.058555055951858, "grad_norm": 0.24096697568893433, "learning_rate": 3.6306279679242033e-05, "loss": 0.0459, "step": 60890 }, { "epoch": 2.058893133642111, "grad_norm": 0.21308691799640656, "learning_rate": 3.629037792661492e-05, "loss": 0.0445, "step": 60900 }, { "epoch": 2.059231211332364, "grad_norm": 0.23575836420059204, "learning_rate": 3.627447767325147e-05, "loss": 0.0498, "step": 60910 }, { "epoch": 2.0595692890226176, "grad_norm": 0.3061276972293854, "learning_rate": 3.62585789208905e-05, "loss": 0.042, "step": 60920 }, { "epoch": 2.0599073667128707, "grad_norm": 0.20343375205993652, "learning_rate": 3.624268167127068e-05, "loss": 0.0425, "step": 60930 }, { "epoch": 2.0602454444031237, "grad_norm": 0.24804654717445374, "learning_rate": 3.622678592613051e-05, "loss": 0.0489, "step": 60940 }, { "epoch": 2.0605835220933773, "grad_norm": 0.27933964133262634, "learning_rate": 3.6210891687208304e-05, "loss": 0.0488, "step": 60950 }, { "epoch": 2.0609215997836303, "grad_norm": 0.19031547009944916, "learning_rate": 3.619499895624228e-05, "loss": 0.0508, "step": 60960 }, { "epoch": 2.0612596774738834, "grad_norm": 0.2724190354347229, "learning_rate": 3.61791077349704e-05, "loss": 0.0454, "step": 60970 }, { "epoch": 2.061597755164137, "grad_norm": 0.34341520071029663, "learning_rate": 3.616321802513053e-05, "loss": 0.0468, "step": 60980 }, { "epoch": 2.06193583285439, "grad_norm": 0.2703244686126709, "learning_rate": 3.6147329828460325e-05, "loss": 0.0456, "step": 60990 }, { "epoch": 2.062273910544643, "grad_norm": 0.21983590722084045, "learning_rate": 3.6131443146697306e-05, "loss": 0.0437, "step": 61000 }, { "epoch": 2.0626119882348966, "grad_norm": 0.21848304569721222, "learning_rate": 3.611555798157879e-05, "loss": 0.0497, "step": 61010 }, { "epoch": 2.0629500659251496, "grad_norm": 0.30408933758735657, "learning_rate": 3.6099674334842e-05, "loss": 0.049, "step": 61020 }, { "epoch": 2.0632881436154027, "grad_norm": 0.3538467288017273, "learning_rate": 3.6083792208223934e-05, "loss": 0.0469, "step": 61030 }, { "epoch": 2.063626221305656, "grad_norm": 0.2739993631839752, "learning_rate": 3.606791160346142e-05, "loss": 0.0492, "step": 61040 }, { "epoch": 2.0639642989959093, "grad_norm": 0.22166842222213745, "learning_rate": 3.6052032522291134e-05, "loss": 0.0485, "step": 61050 }, { "epoch": 2.0643023766861623, "grad_norm": 0.3011789917945862, "learning_rate": 3.603615496644962e-05, "loss": 0.049, "step": 61060 }, { "epoch": 2.064640454376416, "grad_norm": 0.28995445370674133, "learning_rate": 3.60202789376732e-05, "loss": 0.0475, "step": 61070 }, { "epoch": 2.064978532066669, "grad_norm": 0.21980620920658112, "learning_rate": 3.6004404437698036e-05, "loss": 0.0439, "step": 61080 }, { "epoch": 2.065316609756922, "grad_norm": 0.19032971560955048, "learning_rate": 3.5988531468260184e-05, "loss": 0.0498, "step": 61090 }, { "epoch": 2.0656546874471755, "grad_norm": 0.16904407739639282, "learning_rate": 3.597266003109546e-05, "loss": 0.0477, "step": 61100 }, { "epoch": 2.0659927651374286, "grad_norm": 0.1837310940027237, "learning_rate": 3.5956790127939557e-05, "loss": 0.0501, "step": 61110 }, { "epoch": 2.0663308428276816, "grad_norm": 0.22333228588104248, "learning_rate": 3.594092176052796e-05, "loss": 0.051, "step": 61120 }, { "epoch": 2.066668920517935, "grad_norm": 0.197676420211792, "learning_rate": 3.592505493059603e-05, "loss": 0.047, "step": 61130 }, { "epoch": 2.0670069982081882, "grad_norm": 0.4391210079193115, "learning_rate": 3.5909189639878905e-05, "loss": 0.0454, "step": 61140 }, { "epoch": 2.0673450758984413, "grad_norm": 0.3424026072025299, "learning_rate": 3.5893325890111626e-05, "loss": 0.0477, "step": 61150 }, { "epoch": 2.067683153588695, "grad_norm": 0.18574802577495575, "learning_rate": 3.5877463683029034e-05, "loss": 0.0459, "step": 61160 }, { "epoch": 2.068021231278948, "grad_norm": 0.4842887818813324, "learning_rate": 3.586160302036578e-05, "loss": 0.0503, "step": 61170 }, { "epoch": 2.068359308969201, "grad_norm": 0.26409196853637695, "learning_rate": 3.584574390385636e-05, "loss": 0.0504, "step": 61180 }, { "epoch": 2.0686973866594545, "grad_norm": 0.1513180136680603, "learning_rate": 3.582988633523511e-05, "loss": 0.0482, "step": 61190 }, { "epoch": 2.0690354643497075, "grad_norm": 0.453046053647995, "learning_rate": 3.581403031623616e-05, "loss": 0.0471, "step": 61200 }, { "epoch": 2.0693735420399606, "grad_norm": 0.32875797152519226, "learning_rate": 3.579817584859354e-05, "loss": 0.0515, "step": 61210 }, { "epoch": 2.069711619730214, "grad_norm": 0.24517247080802917, "learning_rate": 3.578232293404107e-05, "loss": 0.0482, "step": 61220 }, { "epoch": 2.070049697420467, "grad_norm": 0.3973124027252197, "learning_rate": 3.5766471574312385e-05, "loss": 0.0515, "step": 61230 }, { "epoch": 2.0703877751107203, "grad_norm": 0.20921830832958221, "learning_rate": 3.5750621771140964e-05, "loss": 0.048, "step": 61240 }, { "epoch": 2.0707258528009738, "grad_norm": 0.27220475673675537, "learning_rate": 3.5734773526260115e-05, "loss": 0.0444, "step": 61250 }, { "epoch": 2.071063930491227, "grad_norm": 0.29224100708961487, "learning_rate": 3.5718926841402996e-05, "loss": 0.0497, "step": 61260 }, { "epoch": 2.07140200818148, "grad_norm": 0.27869778871536255, "learning_rate": 3.5703081718302555e-05, "loss": 0.0461, "step": 61270 }, { "epoch": 2.0717400858717334, "grad_norm": 0.15280595421791077, "learning_rate": 3.568723815869162e-05, "loss": 0.0502, "step": 61280 }, { "epoch": 2.0720781635619865, "grad_norm": 0.18762856721878052, "learning_rate": 3.5671396164302777e-05, "loss": 0.0475, "step": 61290 }, { "epoch": 2.0724162412522396, "grad_norm": 0.3144676387310028, "learning_rate": 3.565555573686853e-05, "loss": 0.0541, "step": 61300 }, { "epoch": 2.072754318942493, "grad_norm": 0.2536471486091614, "learning_rate": 3.563971687812113e-05, "loss": 0.0445, "step": 61310 }, { "epoch": 2.073092396632746, "grad_norm": 0.3404861092567444, "learning_rate": 3.5623879589792706e-05, "loss": 0.0462, "step": 61320 }, { "epoch": 2.0734304743229997, "grad_norm": 0.33084508776664734, "learning_rate": 3.5608043873615185e-05, "loss": 0.0454, "step": 61330 }, { "epoch": 2.0737685520132527, "grad_norm": 0.2138340324163437, "learning_rate": 3.559220973132035e-05, "loss": 0.0441, "step": 61340 }, { "epoch": 2.074106629703506, "grad_norm": 0.18596619367599487, "learning_rate": 3.557637716463981e-05, "loss": 0.048, "step": 61350 }, { "epoch": 2.0744447073937593, "grad_norm": 0.2298964411020279, "learning_rate": 3.556054617530497e-05, "loss": 0.0459, "step": 61360 }, { "epoch": 2.0747827850840124, "grad_norm": 0.32149627804756165, "learning_rate": 3.5544716765047106e-05, "loss": 0.0457, "step": 61370 }, { "epoch": 2.0751208627742654, "grad_norm": 0.2758904695510864, "learning_rate": 3.5528888935597274e-05, "loss": 0.0492, "step": 61380 }, { "epoch": 2.075458940464519, "grad_norm": 0.23228614032268524, "learning_rate": 3.551306268868636e-05, "loss": 0.0472, "step": 61390 }, { "epoch": 2.075797018154772, "grad_norm": 0.23017920553684235, "learning_rate": 3.5497238026045166e-05, "loss": 0.0515, "step": 61400 }, { "epoch": 2.076135095845025, "grad_norm": 0.2524104714393616, "learning_rate": 3.548141494940421e-05, "loss": 0.0483, "step": 61410 }, { "epoch": 2.0764731735352786, "grad_norm": 0.34582704305648804, "learning_rate": 3.546559346049388e-05, "loss": 0.0455, "step": 61420 }, { "epoch": 2.0768112512255317, "grad_norm": 0.4055998921394348, "learning_rate": 3.544977356104441e-05, "loss": 0.0504, "step": 61430 }, { "epoch": 2.0771493289157847, "grad_norm": 0.2600761651992798, "learning_rate": 3.543395525278582e-05, "loss": 0.0445, "step": 61440 }, { "epoch": 2.0774874066060383, "grad_norm": 0.23334234952926636, "learning_rate": 3.5418138537447985e-05, "loss": 0.048, "step": 61450 }, { "epoch": 2.0778254842962913, "grad_norm": 0.20962412655353546, "learning_rate": 3.5402323416760574e-05, "loss": 0.0479, "step": 61460 }, { "epoch": 2.0781635619865444, "grad_norm": 0.2320316880941391, "learning_rate": 3.538650989245316e-05, "loss": 0.0482, "step": 61470 }, { "epoch": 2.078501639676798, "grad_norm": 0.2713623046875, "learning_rate": 3.537069796625504e-05, "loss": 0.0466, "step": 61480 }, { "epoch": 2.078839717367051, "grad_norm": 0.19932805001735687, "learning_rate": 3.53548876398954e-05, "loss": 0.0469, "step": 61490 }, { "epoch": 2.079177795057304, "grad_norm": 0.22553059458732605, "learning_rate": 3.5339078915103214e-05, "loss": 0.0472, "step": 61500 }, { "epoch": 2.0795158727475576, "grad_norm": 0.4501071870326996, "learning_rate": 3.532327179360733e-05, "loss": 0.0444, "step": 61510 }, { "epoch": 2.0798539504378106, "grad_norm": 0.5431408286094666, "learning_rate": 3.530746627713636e-05, "loss": 0.047, "step": 61520 }, { "epoch": 2.0801920281280637, "grad_norm": 0.2846461236476898, "learning_rate": 3.5291662367418805e-05, "loss": 0.0549, "step": 61530 }, { "epoch": 2.080530105818317, "grad_norm": 0.26149606704711914, "learning_rate": 3.527586006618293e-05, "loss": 0.0456, "step": 61540 }, { "epoch": 2.0808681835085703, "grad_norm": 0.15982399880886078, "learning_rate": 3.526005937515687e-05, "loss": 0.0455, "step": 61550 }, { "epoch": 2.0812062611988233, "grad_norm": 0.26851513981819153, "learning_rate": 3.524426029606856e-05, "loss": 0.046, "step": 61560 }, { "epoch": 2.081544338889077, "grad_norm": 0.4457615613937378, "learning_rate": 3.5228462830645746e-05, "loss": 0.0538, "step": 61570 }, { "epoch": 2.08188241657933, "grad_norm": 0.2876826226711273, "learning_rate": 3.5212666980616036e-05, "loss": 0.0493, "step": 61580 }, { "epoch": 2.082220494269583, "grad_norm": 0.33759135007858276, "learning_rate": 3.519687274770681e-05, "loss": 0.0441, "step": 61590 }, { "epoch": 2.0825585719598365, "grad_norm": 0.21108406782150269, "learning_rate": 3.518108013364535e-05, "loss": 0.048, "step": 61600 }, { "epoch": 2.0828966496500896, "grad_norm": 0.2616758942604065, "learning_rate": 3.516528914015868e-05, "loss": 0.0491, "step": 61610 }, { "epoch": 2.0832347273403427, "grad_norm": 0.2014596164226532, "learning_rate": 3.5149499768973694e-05, "loss": 0.0493, "step": 61620 }, { "epoch": 2.083572805030596, "grad_norm": 0.5487959384918213, "learning_rate": 3.513371202181708e-05, "loss": 0.0483, "step": 61630 }, { "epoch": 2.0839108827208492, "grad_norm": 0.30855342745780945, "learning_rate": 3.511792590041537e-05, "loss": 0.0429, "step": 61640 }, { "epoch": 2.0842489604111023, "grad_norm": 0.33822154998779297, "learning_rate": 3.510214140649488e-05, "loss": 0.0488, "step": 61650 }, { "epoch": 2.084587038101356, "grad_norm": 0.3408454954624176, "learning_rate": 3.508635854178184e-05, "loss": 0.0528, "step": 61660 }, { "epoch": 2.084925115791609, "grad_norm": 0.23436889052391052, "learning_rate": 3.50705773080022e-05, "loss": 0.0496, "step": 61670 }, { "epoch": 2.085263193481862, "grad_norm": 0.42864373326301575, "learning_rate": 3.505479770688178e-05, "loss": 0.0547, "step": 61680 }, { "epoch": 2.0856012711721155, "grad_norm": 0.6164041757583618, "learning_rate": 3.503901974014622e-05, "loss": 0.0492, "step": 61690 }, { "epoch": 2.0859393488623685, "grad_norm": 0.45657554268836975, "learning_rate": 3.502324340952095e-05, "loss": 0.0488, "step": 61700 }, { "epoch": 2.0862774265526216, "grad_norm": 0.21091774106025696, "learning_rate": 3.500746871673126e-05, "loss": 0.0504, "step": 61710 }, { "epoch": 2.086615504242875, "grad_norm": 0.270982563495636, "learning_rate": 3.499169566350227e-05, "loss": 0.0442, "step": 61720 }, { "epoch": 2.086953581933128, "grad_norm": 0.22977088391780853, "learning_rate": 3.4975924251558876e-05, "loss": 0.0507, "step": 61730 }, { "epoch": 2.0872916596233813, "grad_norm": 0.2753596007823944, "learning_rate": 3.496015448262582e-05, "loss": 0.0566, "step": 61740 }, { "epoch": 2.0876297373136348, "grad_norm": 0.21414241194725037, "learning_rate": 3.494438635842764e-05, "loss": 0.0536, "step": 61750 }, { "epoch": 2.087967815003888, "grad_norm": 0.26211991906166077, "learning_rate": 3.492861988068874e-05, "loss": 0.0508, "step": 61760 }, { "epoch": 2.088305892694141, "grad_norm": 0.21973076462745667, "learning_rate": 3.491285505113332e-05, "loss": 0.044, "step": 61770 }, { "epoch": 2.0886439703843944, "grad_norm": 0.29656293988227844, "learning_rate": 3.4897091871485376e-05, "loss": 0.0526, "step": 61780 }, { "epoch": 2.0889820480746475, "grad_norm": 0.3849656283855438, "learning_rate": 3.488133034346875e-05, "loss": 0.0492, "step": 61790 }, { "epoch": 2.0893201257649006, "grad_norm": 0.2521686255931854, "learning_rate": 3.486557046880713e-05, "loss": 0.05, "step": 61800 }, { "epoch": 2.089658203455154, "grad_norm": 0.2542586326599121, "learning_rate": 3.484981224922395e-05, "loss": 0.0476, "step": 61810 }, { "epoch": 2.089996281145407, "grad_norm": 0.2788228392601013, "learning_rate": 3.4834055686442544e-05, "loss": 0.0447, "step": 61820 }, { "epoch": 2.09033435883566, "grad_norm": 0.22673755884170532, "learning_rate": 3.4818300782186e-05, "loss": 0.048, "step": 61830 }, { "epoch": 2.0906724365259137, "grad_norm": 0.23947758972644806, "learning_rate": 3.4802547538177236e-05, "loss": 0.0484, "step": 61840 }, { "epoch": 2.091010514216167, "grad_norm": 0.2549045979976654, "learning_rate": 3.478679595613905e-05, "loss": 0.0462, "step": 61850 }, { "epoch": 2.0913485919064203, "grad_norm": 0.20358632504940033, "learning_rate": 3.4771046037793975e-05, "loss": 0.0454, "step": 61860 }, { "epoch": 2.0916866695966734, "grad_norm": 0.40053296089172363, "learning_rate": 3.4755297784864416e-05, "loss": 0.0477, "step": 61870 }, { "epoch": 2.0920247472869264, "grad_norm": 0.26264533400535583, "learning_rate": 3.473955119907257e-05, "loss": 0.0523, "step": 61880 }, { "epoch": 2.09236282497718, "grad_norm": 0.18912307918071747, "learning_rate": 3.472380628214046e-05, "loss": 0.0444, "step": 61890 }, { "epoch": 2.092700902667433, "grad_norm": 0.26243650913238525, "learning_rate": 3.4708063035789906e-05, "loss": 0.0502, "step": 61900 }, { "epoch": 2.093038980357686, "grad_norm": 0.3226442337036133, "learning_rate": 3.4692321461742614e-05, "loss": 0.0462, "step": 61910 }, { "epoch": 2.0933770580479396, "grad_norm": 0.18020622432231903, "learning_rate": 3.4676581561720023e-05, "loss": 0.04, "step": 61920 }, { "epoch": 2.0937151357381927, "grad_norm": 0.18399572372436523, "learning_rate": 3.466084333744344e-05, "loss": 0.0493, "step": 61930 }, { "epoch": 2.0940532134284457, "grad_norm": 0.14736929535865784, "learning_rate": 3.4645106790633964e-05, "loss": 0.0452, "step": 61940 }, { "epoch": 2.0943912911186993, "grad_norm": 0.35724881291389465, "learning_rate": 3.4629371923012526e-05, "loss": 0.0465, "step": 61950 }, { "epoch": 2.0947293688089523, "grad_norm": 0.5005120635032654, "learning_rate": 3.4613638736299856e-05, "loss": 0.0487, "step": 61960 }, { "epoch": 2.0950674464992054, "grad_norm": 0.33651429414749146, "learning_rate": 3.459790723221652e-05, "loss": 0.0454, "step": 61970 }, { "epoch": 2.095405524189459, "grad_norm": 0.17195774614810944, "learning_rate": 3.45821774124829e-05, "loss": 0.0425, "step": 61980 }, { "epoch": 2.095743601879712, "grad_norm": 0.2390294224023819, "learning_rate": 3.456644927881918e-05, "loss": 0.0468, "step": 61990 }, { "epoch": 2.096081679569965, "grad_norm": 0.21493075788021088, "learning_rate": 3.4550722832945354e-05, "loss": 0.043, "step": 62000 }, { "epoch": 2.0964197572602186, "grad_norm": 0.2701682448387146, "learning_rate": 3.453499807658126e-05, "loss": 0.0466, "step": 62010 }, { "epoch": 2.0967578349504716, "grad_norm": 0.23751762509346008, "learning_rate": 3.451927501144653e-05, "loss": 0.0428, "step": 62020 }, { "epoch": 2.0970959126407247, "grad_norm": 1.0388801097869873, "learning_rate": 3.450355363926059e-05, "loss": 0.0478, "step": 62030 }, { "epoch": 2.097433990330978, "grad_norm": 0.24080844223499298, "learning_rate": 3.448783396174273e-05, "loss": 0.047, "step": 62040 }, { "epoch": 2.0977720680212313, "grad_norm": 0.265471488237381, "learning_rate": 3.4472115980612034e-05, "loss": 0.051, "step": 62050 }, { "epoch": 2.0981101457114844, "grad_norm": 0.42284131050109863, "learning_rate": 3.4456399697587396e-05, "loss": 0.0464, "step": 62060 }, { "epoch": 2.098448223401738, "grad_norm": 0.2483140379190445, "learning_rate": 3.4440685114387506e-05, "loss": 0.0475, "step": 62070 }, { "epoch": 2.098786301091991, "grad_norm": 0.2603396475315094, "learning_rate": 3.442497223273091e-05, "loss": 0.0555, "step": 62080 }, { "epoch": 2.099124378782244, "grad_norm": 0.2512236535549164, "learning_rate": 3.4409261054335926e-05, "loss": 0.0503, "step": 62090 }, { "epoch": 2.0994624564724975, "grad_norm": 0.20403432846069336, "learning_rate": 3.439355158092069e-05, "loss": 0.0472, "step": 62100 }, { "epoch": 2.0998005341627506, "grad_norm": 0.2839076817035675, "learning_rate": 3.437784381420322e-05, "loss": 0.0463, "step": 62110 }, { "epoch": 2.1001386118530037, "grad_norm": 0.25112617015838623, "learning_rate": 3.436213775590126e-05, "loss": 0.0502, "step": 62120 }, { "epoch": 2.100476689543257, "grad_norm": 0.28720834851264954, "learning_rate": 3.4346433407732405e-05, "loss": 0.0401, "step": 62130 }, { "epoch": 2.1008147672335102, "grad_norm": 1.3216298818588257, "learning_rate": 3.433073077141405e-05, "loss": 0.0424, "step": 62140 }, { "epoch": 2.1011528449237633, "grad_norm": 0.35177081823349, "learning_rate": 3.431502984866343e-05, "loss": 0.0461, "step": 62150 }, { "epoch": 2.101490922614017, "grad_norm": 0.41158148646354675, "learning_rate": 3.429933064119753e-05, "loss": 0.0477, "step": 62160 }, { "epoch": 2.10182900030427, "grad_norm": 0.19215089082717896, "learning_rate": 3.4283633150733255e-05, "loss": 0.0444, "step": 62170 }, { "epoch": 2.102167077994523, "grad_norm": 0.19146928191184998, "learning_rate": 3.4267937378987225e-05, "loss": 0.0552, "step": 62180 }, { "epoch": 2.1025051556847765, "grad_norm": 0.24398203194141388, "learning_rate": 3.425224332767592e-05, "loss": 0.0527, "step": 62190 }, { "epoch": 2.1028432333750295, "grad_norm": 0.3338927626609802, "learning_rate": 3.4236550998515595e-05, "loss": 0.0464, "step": 62200 }, { "epoch": 2.1031813110652826, "grad_norm": 0.4212121069431305, "learning_rate": 3.422086039322235e-05, "loss": 0.0481, "step": 62210 }, { "epoch": 2.103519388755536, "grad_norm": 0.3342052698135376, "learning_rate": 3.420517151351209e-05, "loss": 0.0471, "step": 62220 }, { "epoch": 2.103857466445789, "grad_norm": 0.29448452591896057, "learning_rate": 3.418948436110053e-05, "loss": 0.0484, "step": 62230 }, { "epoch": 2.1041955441360423, "grad_norm": 0.7019675374031067, "learning_rate": 3.417379893770319e-05, "loss": 0.0475, "step": 62240 }, { "epoch": 2.1045336218262958, "grad_norm": 0.26887211203575134, "learning_rate": 3.41581152450354e-05, "loss": 0.0491, "step": 62250 }, { "epoch": 2.104871699516549, "grad_norm": 0.2097654491662979, "learning_rate": 3.414243328481231e-05, "loss": 0.0454, "step": 62260 }, { "epoch": 2.105209777206802, "grad_norm": 0.25593242049217224, "learning_rate": 3.412675305874888e-05, "loss": 0.0439, "step": 62270 }, { "epoch": 2.1055478548970554, "grad_norm": 0.2707069218158722, "learning_rate": 3.4111074568559876e-05, "loss": 0.0481, "step": 62280 }, { "epoch": 2.1058859325873085, "grad_norm": 0.29937782883644104, "learning_rate": 3.4095397815959846e-05, "loss": 0.0509, "step": 62290 }, { "epoch": 2.106224010277562, "grad_norm": 0.5209721922874451, "learning_rate": 3.407972280266322e-05, "loss": 0.0492, "step": 62300 }, { "epoch": 2.106562087967815, "grad_norm": 0.24661745131015778, "learning_rate": 3.4064049530384166e-05, "loss": 0.0499, "step": 62310 }, { "epoch": 2.106900165658068, "grad_norm": 0.2974960505962372, "learning_rate": 3.40483780008367e-05, "loss": 0.0479, "step": 62320 }, { "epoch": 2.1072382433483217, "grad_norm": 0.2509472966194153, "learning_rate": 3.4032708215734635e-05, "loss": 0.0448, "step": 62330 }, { "epoch": 2.1075763210385747, "grad_norm": 0.31399691104888916, "learning_rate": 3.4017040176791584e-05, "loss": 0.0472, "step": 62340 }, { "epoch": 2.107914398728828, "grad_norm": 0.20744386315345764, "learning_rate": 3.4001373885720976e-05, "loss": 0.0486, "step": 62350 }, { "epoch": 2.1082524764190813, "grad_norm": 0.21963416039943695, "learning_rate": 3.398570934423608e-05, "loss": 0.0487, "step": 62360 }, { "epoch": 2.1085905541093344, "grad_norm": 0.256449431180954, "learning_rate": 3.397004655404994e-05, "loss": 0.0458, "step": 62370 }, { "epoch": 2.1089286317995874, "grad_norm": 0.6108964085578918, "learning_rate": 3.395438551687541e-05, "loss": 0.046, "step": 62380 }, { "epoch": 2.109266709489841, "grad_norm": 0.26330748200416565, "learning_rate": 3.393872623442515e-05, "loss": 0.0494, "step": 62390 }, { "epoch": 2.109604787180094, "grad_norm": 0.15870101749897003, "learning_rate": 3.392306870841164e-05, "loss": 0.0476, "step": 62400 }, { "epoch": 2.109942864870347, "grad_norm": 0.27205562591552734, "learning_rate": 3.390741294054715e-05, "loss": 0.0401, "step": 62410 }, { "epoch": 2.1102809425606006, "grad_norm": 0.2720963656902313, "learning_rate": 3.38917589325438e-05, "loss": 0.0456, "step": 62420 }, { "epoch": 2.1106190202508537, "grad_norm": 0.19080185890197754, "learning_rate": 3.387610668611349e-05, "loss": 0.0408, "step": 62430 }, { "epoch": 2.1109570979411068, "grad_norm": 0.24509857594966888, "learning_rate": 3.38604562029679e-05, "loss": 0.0525, "step": 62440 }, { "epoch": 2.1112951756313603, "grad_norm": 0.2876695990562439, "learning_rate": 3.384480748481855e-05, "loss": 0.0548, "step": 62450 }, { "epoch": 2.1116332533216133, "grad_norm": 0.22377103567123413, "learning_rate": 3.382916053337676e-05, "loss": 0.0487, "step": 62460 }, { "epoch": 2.1119713310118664, "grad_norm": 0.36994677782058716, "learning_rate": 3.381351535035366e-05, "loss": 0.049, "step": 62470 }, { "epoch": 2.11230940870212, "grad_norm": 0.18343423306941986, "learning_rate": 3.379787193746019e-05, "loss": 0.0498, "step": 62480 }, { "epoch": 2.112647486392373, "grad_norm": 0.17128793895244598, "learning_rate": 3.378223029640709e-05, "loss": 0.0514, "step": 62490 }, { "epoch": 2.112985564082626, "grad_norm": 0.1846720576286316, "learning_rate": 3.376659042890489e-05, "loss": 0.0479, "step": 62500 }, { "epoch": 2.1133236417728796, "grad_norm": 0.29382774233818054, "learning_rate": 3.375095233666397e-05, "loss": 0.0476, "step": 62510 }, { "epoch": 2.1136617194631326, "grad_norm": 0.3272798955440521, "learning_rate": 3.373531602139447e-05, "loss": 0.043, "step": 62520 }, { "epoch": 2.1139997971533857, "grad_norm": 0.7288669347763062, "learning_rate": 3.371968148480635e-05, "loss": 0.0494, "step": 62530 }, { "epoch": 2.114337874843639, "grad_norm": 0.27151164412498474, "learning_rate": 3.370404872860937e-05, "loss": 0.0449, "step": 62540 }, { "epoch": 2.1146759525338923, "grad_norm": 0.1993815302848816, "learning_rate": 3.3688417754513145e-05, "loss": 0.0451, "step": 62550 }, { "epoch": 2.1150140302241454, "grad_norm": 0.2084423005580902, "learning_rate": 3.367278856422703e-05, "loss": 0.0482, "step": 62560 }, { "epoch": 2.115352107914399, "grad_norm": 0.23350316286087036, "learning_rate": 3.365716115946021e-05, "loss": 0.0476, "step": 62570 }, { "epoch": 2.115690185604652, "grad_norm": 0.21258936822414398, "learning_rate": 3.3641535541921666e-05, "loss": 0.0453, "step": 62580 }, { "epoch": 2.116028263294905, "grad_norm": 0.2081781029701233, "learning_rate": 3.36259117133202e-05, "loss": 0.0469, "step": 62590 }, { "epoch": 2.1163663409851585, "grad_norm": 0.2756292521953583, "learning_rate": 3.36102896753644e-05, "loss": 0.0475, "step": 62600 }, { "epoch": 2.1167044186754116, "grad_norm": 0.17962804436683655, "learning_rate": 3.35946694297627e-05, "loss": 0.046, "step": 62610 }, { "epoch": 2.1170424963656647, "grad_norm": 0.21528096497058868, "learning_rate": 3.357905097822328e-05, "loss": 0.0453, "step": 62620 }, { "epoch": 2.117380574055918, "grad_norm": 0.26643913984298706, "learning_rate": 3.356343432245416e-05, "loss": 0.0508, "step": 62630 }, { "epoch": 2.1177186517461712, "grad_norm": 0.29824718832969666, "learning_rate": 3.354781946416315e-05, "loss": 0.0464, "step": 62640 }, { "epoch": 2.1180567294364243, "grad_norm": 0.18913401663303375, "learning_rate": 3.3532206405057864e-05, "loss": 0.047, "step": 62650 }, { "epoch": 2.118394807126678, "grad_norm": 0.2236822545528412, "learning_rate": 3.351659514684573e-05, "loss": 0.0486, "step": 62660 }, { "epoch": 2.118732884816931, "grad_norm": 0.37195268273353577, "learning_rate": 3.350098569123395e-05, "loss": 0.0527, "step": 62670 }, { "epoch": 2.119070962507184, "grad_norm": 0.17386391758918762, "learning_rate": 3.348537803992958e-05, "loss": 0.0442, "step": 62680 }, { "epoch": 2.1194090401974375, "grad_norm": 0.5467410087585449, "learning_rate": 3.346977219463945e-05, "loss": 0.045, "step": 62690 }, { "epoch": 2.1197471178876905, "grad_norm": 0.37773066759109497, "learning_rate": 3.345416815707017e-05, "loss": 0.048, "step": 62700 }, { "epoch": 2.1200851955779436, "grad_norm": 0.3247371017932892, "learning_rate": 3.3438565928928185e-05, "loss": 0.0509, "step": 62710 }, { "epoch": 2.120423273268197, "grad_norm": 0.24055549502372742, "learning_rate": 3.3422965511919736e-05, "loss": 0.0502, "step": 62720 }, { "epoch": 2.12076135095845, "grad_norm": 0.29687371850013733, "learning_rate": 3.3407366907750846e-05, "loss": 0.0518, "step": 62730 }, { "epoch": 2.1210994286487033, "grad_norm": 0.18831142783164978, "learning_rate": 3.339177011812737e-05, "loss": 0.0465, "step": 62740 }, { "epoch": 2.121437506338957, "grad_norm": 0.4484289288520813, "learning_rate": 3.3376175144754935e-05, "loss": 0.0527, "step": 62750 }, { "epoch": 2.12177558402921, "grad_norm": 0.32936573028564453, "learning_rate": 3.3360581989339004e-05, "loss": 0.0437, "step": 62760 }, { "epoch": 2.122113661719463, "grad_norm": 0.17929163575172424, "learning_rate": 3.3344990653584804e-05, "loss": 0.0478, "step": 62770 }, { "epoch": 2.1224517394097164, "grad_norm": 0.17757003009319305, "learning_rate": 3.332940113919739e-05, "loss": 0.0525, "step": 62780 }, { "epoch": 2.1227898170999695, "grad_norm": 0.22165833413600922, "learning_rate": 3.331381344788161e-05, "loss": 0.0517, "step": 62790 }, { "epoch": 2.1231278947902226, "grad_norm": 0.23641233146190643, "learning_rate": 3.3298227581342076e-05, "loss": 0.047, "step": 62800 }, { "epoch": 2.123465972480476, "grad_norm": 0.3088141977787018, "learning_rate": 3.328264354128329e-05, "loss": 0.0453, "step": 62810 }, { "epoch": 2.123804050170729, "grad_norm": 0.22431354224681854, "learning_rate": 3.3267061329409465e-05, "loss": 0.0441, "step": 62820 }, { "epoch": 2.1241421278609827, "grad_norm": 0.23134104907512665, "learning_rate": 3.325148094742466e-05, "loss": 0.0477, "step": 62830 }, { "epoch": 2.1244802055512357, "grad_norm": 0.5844553112983704, "learning_rate": 3.323590239703271e-05, "loss": 0.046, "step": 62840 }, { "epoch": 2.124818283241489, "grad_norm": 0.449449747800827, "learning_rate": 3.322032567993728e-05, "loss": 0.0467, "step": 62850 }, { "epoch": 2.1251563609317423, "grad_norm": 0.23513710498809814, "learning_rate": 3.3204750797841774e-05, "loss": 0.047, "step": 62860 }, { "epoch": 2.1254944386219954, "grad_norm": 0.1987108588218689, "learning_rate": 3.31891777524495e-05, "loss": 0.0453, "step": 62870 }, { "epoch": 2.1258325163122485, "grad_norm": 0.4080137610435486, "learning_rate": 3.317360654546347e-05, "loss": 0.0493, "step": 62880 }, { "epoch": 2.126170594002502, "grad_norm": 0.23681488633155823, "learning_rate": 3.3158037178586534e-05, "loss": 0.0468, "step": 62890 }, { "epoch": 2.126508671692755, "grad_norm": 0.21277712285518646, "learning_rate": 3.3142469653521316e-05, "loss": 0.0451, "step": 62900 }, { "epoch": 2.126846749383008, "grad_norm": 0.3521278500556946, "learning_rate": 3.3126903971970284e-05, "loss": 0.0459, "step": 62910 }, { "epoch": 2.1271848270732616, "grad_norm": 0.2818748652935028, "learning_rate": 3.311134013563564e-05, "loss": 0.0468, "step": 62920 }, { "epoch": 2.1275229047635147, "grad_norm": 0.2807506024837494, "learning_rate": 3.309577814621947e-05, "loss": 0.0467, "step": 62930 }, { "epoch": 2.1278609824537678, "grad_norm": 0.23935721814632416, "learning_rate": 3.308021800542359e-05, "loss": 0.0465, "step": 62940 }, { "epoch": 2.1281990601440213, "grad_norm": 0.4212414622306824, "learning_rate": 3.306465971494963e-05, "loss": 0.0451, "step": 62950 }, { "epoch": 2.1285371378342743, "grad_norm": 0.17370164394378662, "learning_rate": 3.304910327649901e-05, "loss": 0.052, "step": 62960 }, { "epoch": 2.1288752155245274, "grad_norm": 0.9857081174850464, "learning_rate": 3.303354869177297e-05, "loss": 0.0437, "step": 62970 }, { "epoch": 2.129213293214781, "grad_norm": 0.3804190158843994, "learning_rate": 3.301799596247255e-05, "loss": 0.0496, "step": 62980 }, { "epoch": 2.129551370905034, "grad_norm": 0.2827680706977844, "learning_rate": 3.300244509029854e-05, "loss": 0.045, "step": 62990 }, { "epoch": 2.129889448595287, "grad_norm": 0.3377861976623535, "learning_rate": 3.2986896076951585e-05, "loss": 0.0529, "step": 63000 }, { "epoch": 2.1302275262855406, "grad_norm": 0.2109483778476715, "learning_rate": 3.297134892413211e-05, "loss": 0.0461, "step": 63010 }, { "epoch": 2.1305656039757936, "grad_norm": 0.525897204875946, "learning_rate": 3.295580363354031e-05, "loss": 0.0519, "step": 63020 }, { "epoch": 2.1309036816660467, "grad_norm": 0.2500293254852295, "learning_rate": 3.2940260206876194e-05, "loss": 0.048, "step": 63030 }, { "epoch": 2.1312417593563002, "grad_norm": 0.15464040637016296, "learning_rate": 3.292471864583958e-05, "loss": 0.0473, "step": 63040 }, { "epoch": 2.1315798370465533, "grad_norm": 0.3739108741283417, "learning_rate": 3.290917895213004e-05, "loss": 0.0525, "step": 63050 }, { "epoch": 2.1319179147368064, "grad_norm": 0.26423677802085876, "learning_rate": 3.289364112744701e-05, "loss": 0.043, "step": 63060 }, { "epoch": 2.13225599242706, "grad_norm": 0.2384280562400818, "learning_rate": 3.287810517348967e-05, "loss": 0.0441, "step": 63070 }, { "epoch": 2.132594070117313, "grad_norm": 0.39331334829330444, "learning_rate": 3.2862571091957014e-05, "loss": 0.0548, "step": 63080 }, { "epoch": 2.132932147807566, "grad_norm": 0.20052985846996307, "learning_rate": 3.284703888454781e-05, "loss": 0.0436, "step": 63090 }, { "epoch": 2.1332702254978195, "grad_norm": 0.24581222236156464, "learning_rate": 3.2831508552960646e-05, "loss": 0.0513, "step": 63100 }, { "epoch": 2.1336083031880726, "grad_norm": 0.24140091240406036, "learning_rate": 3.281598009889387e-05, "loss": 0.0436, "step": 63110 }, { "epoch": 2.1339463808783257, "grad_norm": 0.6209373474121094, "learning_rate": 3.28004535240457e-05, "loss": 0.0471, "step": 63120 }, { "epoch": 2.134284458568579, "grad_norm": 0.2508503794670105, "learning_rate": 3.278492883011407e-05, "loss": 0.0442, "step": 63130 }, { "epoch": 2.1346225362588322, "grad_norm": 0.25351160764694214, "learning_rate": 3.276940601879675e-05, "loss": 0.0478, "step": 63140 }, { "epoch": 2.1349606139490853, "grad_norm": 0.27216437458992004, "learning_rate": 3.275388509179127e-05, "loss": 0.0512, "step": 63150 }, { "epoch": 2.135298691639339, "grad_norm": 0.37697476148605347, "learning_rate": 3.2738366050794985e-05, "loss": 0.0494, "step": 63160 }, { "epoch": 2.135636769329592, "grad_norm": 0.2802790105342865, "learning_rate": 3.2722848897505034e-05, "loss": 0.0421, "step": 63170 }, { "epoch": 2.135974847019845, "grad_norm": 0.15724705159664154, "learning_rate": 3.2707333633618356e-05, "loss": 0.0448, "step": 63180 }, { "epoch": 2.1363129247100985, "grad_norm": 0.26369407773017883, "learning_rate": 3.269182026083167e-05, "loss": 0.0437, "step": 63190 }, { "epoch": 2.1366510024003516, "grad_norm": 0.1608438640832901, "learning_rate": 3.2676308780841516e-05, "loss": 0.0447, "step": 63200 }, { "epoch": 2.1369890800906046, "grad_norm": 0.17680896818637848, "learning_rate": 3.266079919534416e-05, "loss": 0.0485, "step": 63210 }, { "epoch": 2.137327157780858, "grad_norm": 0.32235464453697205, "learning_rate": 3.264529150603575e-05, "loss": 0.0473, "step": 63220 }, { "epoch": 2.137665235471111, "grad_norm": 0.7050651907920837, "learning_rate": 3.262978571461217e-05, "loss": 0.0506, "step": 63230 }, { "epoch": 2.1380033131613647, "grad_norm": 0.28498852252960205, "learning_rate": 3.26142818227691e-05, "loss": 0.0435, "step": 63240 }, { "epoch": 2.138341390851618, "grad_norm": 0.4590641260147095, "learning_rate": 3.259877983220203e-05, "loss": 0.0437, "step": 63250 }, { "epoch": 2.138679468541871, "grad_norm": 0.2071194052696228, "learning_rate": 3.2583279744606246e-05, "loss": 0.0457, "step": 63260 }, { "epoch": 2.1390175462321244, "grad_norm": 0.23453551530838013, "learning_rate": 3.25677815616768e-05, "loss": 0.0467, "step": 63270 }, { "epoch": 2.1393556239223774, "grad_norm": 0.17753705382347107, "learning_rate": 3.255228528510856e-05, "loss": 0.0475, "step": 63280 }, { "epoch": 2.1396937016126305, "grad_norm": 0.166425421833992, "learning_rate": 3.253679091659617e-05, "loss": 0.047, "step": 63290 }, { "epoch": 2.140031779302884, "grad_norm": 0.18892620503902435, "learning_rate": 3.252129845783404e-05, "loss": 0.0444, "step": 63300 }, { "epoch": 2.140369856993137, "grad_norm": 0.16726894676685333, "learning_rate": 3.2505807910516456e-05, "loss": 0.0454, "step": 63310 }, { "epoch": 2.14070793468339, "grad_norm": 0.20522262156009674, "learning_rate": 3.249031927633742e-05, "loss": 0.044, "step": 63320 }, { "epoch": 2.1410460123736437, "grad_norm": 0.3590947687625885, "learning_rate": 3.2474832556990737e-05, "loss": 0.0456, "step": 63330 }, { "epoch": 2.1413840900638967, "grad_norm": 0.2423250824213028, "learning_rate": 3.245934775417002e-05, "loss": 0.0473, "step": 63340 }, { "epoch": 2.14172216775415, "grad_norm": 0.30527251958847046, "learning_rate": 3.244386486956866e-05, "loss": 0.0452, "step": 63350 }, { "epoch": 2.1420602454444033, "grad_norm": 0.28790247440338135, "learning_rate": 3.2428383904879846e-05, "loss": 0.0484, "step": 63360 }, { "epoch": 2.1423983231346564, "grad_norm": 0.28396132588386536, "learning_rate": 3.241290486179652e-05, "loss": 0.049, "step": 63370 }, { "epoch": 2.1427364008249095, "grad_norm": 0.30449533462524414, "learning_rate": 3.239742774201151e-05, "loss": 0.0468, "step": 63380 }, { "epoch": 2.143074478515163, "grad_norm": 0.16703884303569794, "learning_rate": 3.238195254721733e-05, "loss": 0.0472, "step": 63390 }, { "epoch": 2.143412556205416, "grad_norm": 0.18689846992492676, "learning_rate": 3.236647927910633e-05, "loss": 0.0458, "step": 63400 }, { "epoch": 2.143750633895669, "grad_norm": 0.24578584730625153, "learning_rate": 3.2351007939370646e-05, "loss": 0.0479, "step": 63410 }, { "epoch": 2.1440887115859226, "grad_norm": 0.29975223541259766, "learning_rate": 3.2335538529702184e-05, "loss": 0.0486, "step": 63420 }, { "epoch": 2.1444267892761757, "grad_norm": 0.2950405478477478, "learning_rate": 3.232007105179268e-05, "loss": 0.0458, "step": 63430 }, { "epoch": 2.1447648669664288, "grad_norm": 0.2547647953033447, "learning_rate": 3.230460550733363e-05, "loss": 0.047, "step": 63440 }, { "epoch": 2.1451029446566823, "grad_norm": 0.16563406586647034, "learning_rate": 3.2289141898016315e-05, "loss": 0.0444, "step": 63450 }, { "epoch": 2.1454410223469353, "grad_norm": 0.24788175523281097, "learning_rate": 3.227368022553181e-05, "loss": 0.0447, "step": 63460 }, { "epoch": 2.1457791000371884, "grad_norm": 1.0106909275054932, "learning_rate": 3.225822049157099e-05, "loss": 0.0437, "step": 63470 }, { "epoch": 2.146117177727442, "grad_norm": 0.36492443084716797, "learning_rate": 3.2242762697824494e-05, "loss": 0.0461, "step": 63480 }, { "epoch": 2.146455255417695, "grad_norm": 0.16962853074073792, "learning_rate": 3.222730684598279e-05, "loss": 0.0401, "step": 63490 }, { "epoch": 2.146793333107948, "grad_norm": 0.3621997535228729, "learning_rate": 3.221185293773605e-05, "loss": 0.0459, "step": 63500 }, { "epoch": 2.1471314107982016, "grad_norm": 0.3111076354980469, "learning_rate": 3.2196400974774354e-05, "loss": 0.0462, "step": 63510 }, { "epoch": 2.1474694884884546, "grad_norm": 0.22309409081935883, "learning_rate": 3.2180950958787485e-05, "loss": 0.0475, "step": 63520 }, { "epoch": 2.1478075661787077, "grad_norm": 0.4831605553627014, "learning_rate": 3.2165502891465024e-05, "loss": 0.0427, "step": 63530 }, { "epoch": 2.1481456438689612, "grad_norm": 0.2502540946006775, "learning_rate": 3.215005677449636e-05, "loss": 0.0459, "step": 63540 }, { "epoch": 2.1484837215592143, "grad_norm": 0.35650908946990967, "learning_rate": 3.2134612609570644e-05, "loss": 0.0467, "step": 63550 }, { "epoch": 2.1488217992494674, "grad_norm": 0.7221079468727112, "learning_rate": 3.211917039837682e-05, "loss": 0.0425, "step": 63560 }, { "epoch": 2.149159876939721, "grad_norm": 0.27063101530075073, "learning_rate": 3.210373014260365e-05, "loss": 0.0505, "step": 63570 }, { "epoch": 2.149497954629974, "grad_norm": 0.27978143095970154, "learning_rate": 3.208829184393965e-05, "loss": 0.0444, "step": 63580 }, { "epoch": 2.149836032320227, "grad_norm": 0.2744961082935333, "learning_rate": 3.207285550407312e-05, "loss": 0.0454, "step": 63590 }, { "epoch": 2.1501741100104805, "grad_norm": 1.6569019556045532, "learning_rate": 3.205742112469217e-05, "loss": 0.0495, "step": 63600 }, { "epoch": 2.1505121877007336, "grad_norm": 0.34305912256240845, "learning_rate": 3.204198870748465e-05, "loss": 0.0486, "step": 63610 }, { "epoch": 2.1508502653909867, "grad_norm": 0.2982756197452545, "learning_rate": 3.202655825413824e-05, "loss": 0.0524, "step": 63620 }, { "epoch": 2.15118834308124, "grad_norm": 0.20962642133235931, "learning_rate": 3.2011129766340416e-05, "loss": 0.0488, "step": 63630 }, { "epoch": 2.1515264207714933, "grad_norm": 0.2557244896888733, "learning_rate": 3.1995703245778384e-05, "loss": 0.0506, "step": 63640 }, { "epoch": 2.1518644984617463, "grad_norm": 0.22517092525959015, "learning_rate": 3.1980278694139186e-05, "loss": 0.0448, "step": 63650 }, { "epoch": 2.152202576152, "grad_norm": 0.2718103229999542, "learning_rate": 3.196485611310962e-05, "loss": 0.0463, "step": 63660 }, { "epoch": 2.152540653842253, "grad_norm": 0.33696892857551575, "learning_rate": 3.1949435504376257e-05, "loss": 0.0454, "step": 63670 }, { "epoch": 2.152878731532506, "grad_norm": 0.2901248335838318, "learning_rate": 3.1934016869625505e-05, "loss": 0.0489, "step": 63680 }, { "epoch": 2.1532168092227595, "grad_norm": 0.14423836767673492, "learning_rate": 3.191860021054349e-05, "loss": 0.0509, "step": 63690 }, { "epoch": 2.1535548869130126, "grad_norm": 0.16027763485908508, "learning_rate": 3.190318552881618e-05, "loss": 0.046, "step": 63700 }, { "epoch": 2.1538929646032656, "grad_norm": 0.28829970955848694, "learning_rate": 3.1887772826129285e-05, "loss": 0.0454, "step": 63710 }, { "epoch": 2.154231042293519, "grad_norm": 0.15518240630626678, "learning_rate": 3.1872362104168335e-05, "loss": 0.0491, "step": 63720 }, { "epoch": 2.154569119983772, "grad_norm": 0.2602142095565796, "learning_rate": 3.185695336461861e-05, "loss": 0.0465, "step": 63730 }, { "epoch": 2.1549071976740253, "grad_norm": 0.19574834406375885, "learning_rate": 3.184154660916519e-05, "loss": 0.0447, "step": 63740 }, { "epoch": 2.155245275364279, "grad_norm": 0.414628267288208, "learning_rate": 3.1826141839492906e-05, "loss": 0.0459, "step": 63750 }, { "epoch": 2.155583353054532, "grad_norm": 0.23804831504821777, "learning_rate": 3.181073905728644e-05, "loss": 0.0461, "step": 63760 }, { "epoch": 2.155921430744785, "grad_norm": 0.2977009117603302, "learning_rate": 3.1795338264230214e-05, "loss": 0.0498, "step": 63770 }, { "epoch": 2.1562595084350384, "grad_norm": 0.26579275727272034, "learning_rate": 3.1779939462008426e-05, "loss": 0.0481, "step": 63780 }, { "epoch": 2.1565975861252915, "grad_norm": 0.23701316118240356, "learning_rate": 3.1764542652305066e-05, "loss": 0.0439, "step": 63790 }, { "epoch": 2.1569356638155446, "grad_norm": 0.37562301754951477, "learning_rate": 3.17491478368039e-05, "loss": 0.0451, "step": 63800 }, { "epoch": 2.157273741505798, "grad_norm": 0.21533504128456116, "learning_rate": 3.173375501718847e-05, "loss": 0.0455, "step": 63810 }, { "epoch": 2.157611819196051, "grad_norm": 0.1843150407075882, "learning_rate": 3.171836419514215e-05, "loss": 0.0447, "step": 63820 }, { "epoch": 2.1579498968863047, "grad_norm": 0.21982401609420776, "learning_rate": 3.170297537234803e-05, "loss": 0.0469, "step": 63830 }, { "epoch": 2.1582879745765577, "grad_norm": 0.26732704043388367, "learning_rate": 3.168758855048901e-05, "loss": 0.0452, "step": 63840 }, { "epoch": 2.158626052266811, "grad_norm": 0.214333176612854, "learning_rate": 3.1672203731247785e-05, "loss": 0.0477, "step": 63850 }, { "epoch": 2.1589641299570643, "grad_norm": 0.3811390995979309, "learning_rate": 3.16568209163068e-05, "loss": 0.0519, "step": 63860 }, { "epoch": 2.1593022076473174, "grad_norm": 0.3297458589076996, "learning_rate": 3.1641440107348295e-05, "loss": 0.0479, "step": 63870 }, { "epoch": 2.1596402853375705, "grad_norm": 0.21765479445457458, "learning_rate": 3.162606130605429e-05, "loss": 0.0517, "step": 63880 }, { "epoch": 2.159978363027824, "grad_norm": 0.3249824047088623, "learning_rate": 3.16106845141066e-05, "loss": 0.0473, "step": 63890 }, { "epoch": 2.160316440718077, "grad_norm": 0.21443958580493927, "learning_rate": 3.159530973318682e-05, "loss": 0.0413, "step": 63900 }, { "epoch": 2.16065451840833, "grad_norm": 0.1649116426706314, "learning_rate": 3.1579936964976286e-05, "loss": 0.0437, "step": 63910 }, { "epoch": 2.1609925960985836, "grad_norm": 0.2331812083721161, "learning_rate": 3.156456621115615e-05, "loss": 0.0473, "step": 63920 }, { "epoch": 2.1613306737888367, "grad_norm": 0.12853173911571503, "learning_rate": 3.154919747340734e-05, "loss": 0.0483, "step": 63930 }, { "epoch": 2.1616687514790898, "grad_norm": 0.22270210087299347, "learning_rate": 3.153383075341054e-05, "loss": 0.0469, "step": 63940 }, { "epoch": 2.1620068291693433, "grad_norm": 0.31589528918266296, "learning_rate": 3.1518466052846264e-05, "loss": 0.048, "step": 63950 }, { "epoch": 2.1623449068595963, "grad_norm": 0.2533048391342163, "learning_rate": 3.150310337339474e-05, "loss": 0.0475, "step": 63960 }, { "epoch": 2.1626829845498494, "grad_norm": 0.36035576462745667, "learning_rate": 3.148774271673603e-05, "loss": 0.0489, "step": 63970 }, { "epoch": 2.163021062240103, "grad_norm": 0.2381783127784729, "learning_rate": 3.147238408454995e-05, "loss": 0.0476, "step": 63980 }, { "epoch": 2.163359139930356, "grad_norm": 0.236551433801651, "learning_rate": 3.1457027478516086e-05, "loss": 0.0423, "step": 63990 }, { "epoch": 2.163697217620609, "grad_norm": 0.2722644805908203, "learning_rate": 3.144167290031382e-05, "loss": 0.0481, "step": 64000 }, { "epoch": 2.1640352953108626, "grad_norm": 0.2474113404750824, "learning_rate": 3.142632035162227e-05, "loss": 0.0476, "step": 64010 }, { "epoch": 2.1643733730011157, "grad_norm": 0.1924929916858673, "learning_rate": 3.141096983412044e-05, "loss": 0.0453, "step": 64020 }, { "epoch": 2.1647114506913687, "grad_norm": 0.16830620169639587, "learning_rate": 3.139562134948699e-05, "loss": 0.0492, "step": 64030 }, { "epoch": 2.1650495283816222, "grad_norm": 0.18702317774295807, "learning_rate": 3.1380274899400424e-05, "loss": 0.0459, "step": 64040 }, { "epoch": 2.1653876060718753, "grad_norm": 0.3445228040218353, "learning_rate": 3.1364930485539e-05, "loss": 0.0455, "step": 64050 }, { "epoch": 2.1657256837621284, "grad_norm": 0.26873403787612915, "learning_rate": 3.134958810958076e-05, "loss": 0.0482, "step": 64060 }, { "epoch": 2.166063761452382, "grad_norm": 0.2684028446674347, "learning_rate": 3.13342477732035e-05, "loss": 0.0469, "step": 64070 }, { "epoch": 2.166401839142635, "grad_norm": 0.23684538900852203, "learning_rate": 3.131890947808487e-05, "loss": 0.0457, "step": 64080 }, { "epoch": 2.166739916832888, "grad_norm": 0.22028321027755737, "learning_rate": 3.13035732259022e-05, "loss": 0.0469, "step": 64090 }, { "epoch": 2.1670779945231415, "grad_norm": 0.3265134394168854, "learning_rate": 3.1288239018332656e-05, "loss": 0.0516, "step": 64100 }, { "epoch": 2.1674160722133946, "grad_norm": 0.24336235225200653, "learning_rate": 3.1272906857053164e-05, "loss": 0.0456, "step": 64110 }, { "epoch": 2.1677541499036477, "grad_norm": 0.2043561339378357, "learning_rate": 3.125757674374042e-05, "loss": 0.0456, "step": 64120 }, { "epoch": 2.168092227593901, "grad_norm": 0.1951088160276413, "learning_rate": 3.124224868007088e-05, "loss": 0.0467, "step": 64130 }, { "epoch": 2.1684303052841543, "grad_norm": 0.2907179892063141, "learning_rate": 3.1226922667720854e-05, "loss": 0.047, "step": 64140 }, { "epoch": 2.1687683829744073, "grad_norm": 0.444381982088089, "learning_rate": 3.121159870836633e-05, "loss": 0.0497, "step": 64150 }, { "epoch": 2.169106460664661, "grad_norm": 0.2637602984905243, "learning_rate": 3.119627680368312e-05, "loss": 0.0517, "step": 64160 }, { "epoch": 2.169444538354914, "grad_norm": 0.23501485586166382, "learning_rate": 3.1180956955346816e-05, "loss": 0.0476, "step": 64170 }, { "epoch": 2.169782616045167, "grad_norm": 1.0738847255706787, "learning_rate": 3.116563916503276e-05, "loss": 0.0444, "step": 64180 }, { "epoch": 2.1701206937354205, "grad_norm": 0.24602217972278595, "learning_rate": 3.1150323434416095e-05, "loss": 0.0425, "step": 64190 }, { "epoch": 2.1704587714256736, "grad_norm": 0.19360218942165375, "learning_rate": 3.1135009765171696e-05, "loss": 0.0524, "step": 64200 }, { "epoch": 2.170796849115927, "grad_norm": 0.24905899167060852, "learning_rate": 3.111969815897429e-05, "loss": 0.0457, "step": 64210 }, { "epoch": 2.17113492680618, "grad_norm": 0.23936432600021362, "learning_rate": 3.1104388617498295e-05, "loss": 0.0398, "step": 64220 }, { "epoch": 2.171473004496433, "grad_norm": 0.21838532388210297, "learning_rate": 3.1089081142417966e-05, "loss": 0.0503, "step": 64230 }, { "epoch": 2.1718110821866867, "grad_norm": 0.4936612844467163, "learning_rate": 3.1073775735407285e-05, "loss": 0.0457, "step": 64240 }, { "epoch": 2.17214915987694, "grad_norm": 0.34257975220680237, "learning_rate": 3.1058472398140035e-05, "loss": 0.0531, "step": 64250 }, { "epoch": 2.172487237567193, "grad_norm": 0.23582673072814941, "learning_rate": 3.1043171132289736e-05, "loss": 0.0411, "step": 64260 }, { "epoch": 2.1728253152574464, "grad_norm": 0.21909989416599274, "learning_rate": 3.1027871939529774e-05, "loss": 0.0446, "step": 64270 }, { "epoch": 2.1731633929476994, "grad_norm": 0.19068989157676697, "learning_rate": 3.1012574821533214e-05, "loss": 0.0448, "step": 64280 }, { "epoch": 2.1735014706379525, "grad_norm": 0.27988550066947937, "learning_rate": 3.099727977997291e-05, "loss": 0.0468, "step": 64290 }, { "epoch": 2.173839548328206, "grad_norm": 0.21885403990745544, "learning_rate": 3.098198681652154e-05, "loss": 0.0489, "step": 64300 }, { "epoch": 2.174177626018459, "grad_norm": 0.2515628933906555, "learning_rate": 3.0966695932851475e-05, "loss": 0.0464, "step": 64310 }, { "epoch": 2.174515703708712, "grad_norm": 0.22773601114749908, "learning_rate": 3.0951407130634926e-05, "loss": 0.0443, "step": 64320 }, { "epoch": 2.1748537813989657, "grad_norm": 0.22456462681293488, "learning_rate": 3.0936120411543864e-05, "loss": 0.0448, "step": 64330 }, { "epoch": 2.1751918590892187, "grad_norm": 0.18244655430316925, "learning_rate": 3.0920835777250015e-05, "loss": 0.0421, "step": 64340 }, { "epoch": 2.175529936779472, "grad_norm": 0.19786617159843445, "learning_rate": 3.090555322942488e-05, "loss": 0.0484, "step": 64350 }, { "epoch": 2.1758680144697253, "grad_norm": 0.2833915650844574, "learning_rate": 3.089027276973973e-05, "loss": 0.053, "step": 64360 }, { "epoch": 2.1762060921599784, "grad_norm": 0.4264156222343445, "learning_rate": 3.0874994399865615e-05, "loss": 0.0479, "step": 64370 }, { "epoch": 2.1765441698502315, "grad_norm": 0.28629258275032043, "learning_rate": 3.085971812147337e-05, "loss": 0.0488, "step": 64380 }, { "epoch": 2.176882247540485, "grad_norm": 0.2686164677143097, "learning_rate": 3.0844443936233557e-05, "loss": 0.047, "step": 64390 }, { "epoch": 2.177220325230738, "grad_norm": 0.3224620223045349, "learning_rate": 3.082917184581657e-05, "loss": 0.0468, "step": 64400 }, { "epoch": 2.177558402920991, "grad_norm": 0.46112096309661865, "learning_rate": 3.0813901851892513e-05, "loss": 0.0471, "step": 64410 }, { "epoch": 2.1778964806112446, "grad_norm": 0.19727744162082672, "learning_rate": 3.0798633956131316e-05, "loss": 0.0481, "step": 64420 }, { "epoch": 2.1782345583014977, "grad_norm": 0.3233482837677002, "learning_rate": 3.078336816020264e-05, "loss": 0.0454, "step": 64430 }, { "epoch": 2.1785726359917508, "grad_norm": 0.2736036479473114, "learning_rate": 3.0768104465775935e-05, "loss": 0.046, "step": 64440 }, { "epoch": 2.1789107136820043, "grad_norm": 0.26890113949775696, "learning_rate": 3.0752842874520396e-05, "loss": 0.0478, "step": 64450 }, { "epoch": 2.1792487913722574, "grad_norm": 0.23919077217578888, "learning_rate": 3.073758338810503e-05, "loss": 0.0448, "step": 64460 }, { "epoch": 2.1795868690625104, "grad_norm": 0.26471883058547974, "learning_rate": 3.072232600819859e-05, "loss": 0.0447, "step": 64470 }, { "epoch": 2.179924946752764, "grad_norm": 0.49327147006988525, "learning_rate": 3.070707073646959e-05, "loss": 0.0463, "step": 64480 }, { "epoch": 2.180263024443017, "grad_norm": 0.2716486155986786, "learning_rate": 3.069181757458633e-05, "loss": 0.0525, "step": 64490 }, { "epoch": 2.18060110213327, "grad_norm": 0.1745721399784088, "learning_rate": 3.067656652421687e-05, "loss": 0.0476, "step": 64500 }, { "epoch": 2.1809391798235236, "grad_norm": 0.203475683927536, "learning_rate": 3.0661317587029025e-05, "loss": 0.0486, "step": 64510 }, { "epoch": 2.1812772575137767, "grad_norm": 0.1836307793855667, "learning_rate": 3.064607076469043e-05, "loss": 0.0483, "step": 64520 }, { "epoch": 2.1816153352040297, "grad_norm": 0.21592746675014496, "learning_rate": 3.063082605886845e-05, "loss": 0.049, "step": 64530 }, { "epoch": 2.1819534128942832, "grad_norm": 0.28961777687072754, "learning_rate": 3.061558347123019e-05, "loss": 0.0447, "step": 64540 }, { "epoch": 2.1822914905845363, "grad_norm": 0.23910164833068848, "learning_rate": 3.060034300344259e-05, "loss": 0.0516, "step": 64550 }, { "epoch": 2.1826295682747894, "grad_norm": 0.3113442361354828, "learning_rate": 3.058510465717231e-05, "loss": 0.0442, "step": 64560 }, { "epoch": 2.182967645965043, "grad_norm": 0.2103687971830368, "learning_rate": 3.0569868434085804e-05, "loss": 0.0459, "step": 64570 }, { "epoch": 2.183305723655296, "grad_norm": 0.2483210563659668, "learning_rate": 3.055463433584924e-05, "loss": 0.0509, "step": 64580 }, { "epoch": 2.183643801345549, "grad_norm": 0.16796141862869263, "learning_rate": 3.0539402364128656e-05, "loss": 0.0475, "step": 64590 }, { "epoch": 2.1839818790358025, "grad_norm": 0.28343987464904785, "learning_rate": 3.052417252058977e-05, "loss": 0.0461, "step": 64600 }, { "epoch": 2.1843199567260556, "grad_norm": 0.35017484426498413, "learning_rate": 3.0508944806898092e-05, "loss": 0.0488, "step": 64610 }, { "epoch": 2.1846580344163087, "grad_norm": 0.3427406847476959, "learning_rate": 3.04937192247189e-05, "loss": 0.0455, "step": 64620 }, { "epoch": 2.184996112106562, "grad_norm": 0.19338254630565643, "learning_rate": 3.047849577571725e-05, "loss": 0.0492, "step": 64630 }, { "epoch": 2.1853341897968153, "grad_norm": 0.3018707036972046, "learning_rate": 3.046327446155795e-05, "loss": 0.0512, "step": 64640 }, { "epoch": 2.1856722674870683, "grad_norm": 0.1388191282749176, "learning_rate": 3.0448055283905587e-05, "loss": 0.0427, "step": 64650 }, { "epoch": 2.186010345177322, "grad_norm": 0.20055818557739258, "learning_rate": 3.0432838244424494e-05, "loss": 0.0479, "step": 64660 }, { "epoch": 2.186348422867575, "grad_norm": 0.5625666975975037, "learning_rate": 3.04176233447788e-05, "loss": 0.0504, "step": 64670 }, { "epoch": 2.186686500557828, "grad_norm": 0.19456347823143005, "learning_rate": 3.040241058663238e-05, "loss": 0.0471, "step": 64680 }, { "epoch": 2.1870245782480815, "grad_norm": 0.29037898778915405, "learning_rate": 3.038719997164887e-05, "loss": 0.0458, "step": 64690 }, { "epoch": 2.1873626559383346, "grad_norm": 0.22832943499088287, "learning_rate": 3.0371991501491683e-05, "loss": 0.048, "step": 64700 }, { "epoch": 2.1877007336285876, "grad_norm": 0.40503090620040894, "learning_rate": 3.0356785177823975e-05, "loss": 0.0492, "step": 64710 }, { "epoch": 2.188038811318841, "grad_norm": 0.2003420889377594, "learning_rate": 3.0341581002308726e-05, "loss": 0.0446, "step": 64720 }, { "epoch": 2.188376889009094, "grad_norm": 0.26032939553260803, "learning_rate": 3.0326378976608627e-05, "loss": 0.0409, "step": 64730 }, { "epoch": 2.1887149666993473, "grad_norm": 0.28460609912872314, "learning_rate": 3.0311179102386146e-05, "loss": 0.0506, "step": 64740 }, { "epoch": 2.189053044389601, "grad_norm": 0.25752925872802734, "learning_rate": 3.0295981381303513e-05, "loss": 0.0497, "step": 64750 }, { "epoch": 2.189391122079854, "grad_norm": 0.6425094604492188, "learning_rate": 3.0280785815022727e-05, "loss": 0.0476, "step": 64760 }, { "epoch": 2.189729199770107, "grad_norm": 0.3402535915374756, "learning_rate": 3.0265592405205544e-05, "loss": 0.0492, "step": 64770 }, { "epoch": 2.1900672774603605, "grad_norm": 0.46561354398727417, "learning_rate": 3.0250401153513517e-05, "loss": 0.045, "step": 64780 }, { "epoch": 2.1904053551506135, "grad_norm": 0.2318601906299591, "learning_rate": 3.0235212061607936e-05, "loss": 0.0452, "step": 64790 }, { "epoch": 2.190743432840867, "grad_norm": 0.22683754563331604, "learning_rate": 3.0220025131149842e-05, "loss": 0.0498, "step": 64800 }, { "epoch": 2.19108151053112, "grad_norm": 0.22551026940345764, "learning_rate": 3.020484036380007e-05, "loss": 0.0417, "step": 64810 }, { "epoch": 2.191419588221373, "grad_norm": 0.1935538500547409, "learning_rate": 3.0189657761219188e-05, "loss": 0.0431, "step": 64820 }, { "epoch": 2.1917576659116267, "grad_norm": 0.3045368790626526, "learning_rate": 3.017447732506753e-05, "loss": 0.044, "step": 64830 }, { "epoch": 2.1920957436018798, "grad_norm": 0.2705247700214386, "learning_rate": 3.015929905700524e-05, "loss": 0.05, "step": 64840 }, { "epoch": 2.192433821292133, "grad_norm": 0.21145717799663544, "learning_rate": 3.014412295869218e-05, "loss": 0.0439, "step": 64850 }, { "epoch": 2.1927718989823863, "grad_norm": 0.2403673231601715, "learning_rate": 3.0128949031787977e-05, "loss": 0.0506, "step": 64860 }, { "epoch": 2.1931099766726394, "grad_norm": 0.16815803945064545, "learning_rate": 3.0113777277952022e-05, "loss": 0.0478, "step": 64870 }, { "epoch": 2.1934480543628925, "grad_norm": 0.1733953207731247, "learning_rate": 3.0098607698843496e-05, "loss": 0.0463, "step": 64880 }, { "epoch": 2.193786132053146, "grad_norm": 0.27015945315361023, "learning_rate": 3.0083440296121312e-05, "loss": 0.0426, "step": 64890 }, { "epoch": 2.194124209743399, "grad_norm": 0.17927445471286774, "learning_rate": 3.0068275071444142e-05, "loss": 0.0434, "step": 64900 }, { "epoch": 2.194462287433652, "grad_norm": 0.40095028281211853, "learning_rate": 3.0053112026470442e-05, "loss": 0.0468, "step": 64910 }, { "epoch": 2.1948003651239056, "grad_norm": 0.2405870109796524, "learning_rate": 3.0037951162858434e-05, "loss": 0.0498, "step": 64920 }, { "epoch": 2.1951384428141587, "grad_norm": 0.22163569927215576, "learning_rate": 3.002279248226607e-05, "loss": 0.0452, "step": 64930 }, { "epoch": 2.1954765205044118, "grad_norm": 0.3187890350818634, "learning_rate": 3.000763598635109e-05, "loss": 0.0484, "step": 64940 }, { "epoch": 2.1958145981946653, "grad_norm": 0.16061045229434967, "learning_rate": 2.999248167677098e-05, "loss": 0.0438, "step": 64950 }, { "epoch": 2.1961526758849184, "grad_norm": 0.3045499324798584, "learning_rate": 2.9977329555182976e-05, "loss": 0.0402, "step": 64960 }, { "epoch": 2.1964907535751714, "grad_norm": 0.30361640453338623, "learning_rate": 2.9962179623244124e-05, "loss": 0.0455, "step": 64970 }, { "epoch": 2.196828831265425, "grad_norm": 0.23210397362709045, "learning_rate": 2.9947031882611175e-05, "loss": 0.049, "step": 64980 }, { "epoch": 2.197166908955678, "grad_norm": 0.20723287761211395, "learning_rate": 2.9931886334940685e-05, "loss": 0.0447, "step": 64990 }, { "epoch": 2.197504986645931, "grad_norm": 0.210697740316391, "learning_rate": 2.9916742981888923e-05, "loss": 0.0464, "step": 65000 }, { "epoch": 2.1978430643361846, "grad_norm": 0.24028527736663818, "learning_rate": 2.990160182511196e-05, "loss": 0.0441, "step": 65010 }, { "epoch": 2.1981811420264377, "grad_norm": 0.3005659878253937, "learning_rate": 2.9886462866265574e-05, "loss": 0.0513, "step": 65020 }, { "epoch": 2.1985192197166907, "grad_norm": 0.2659548819065094, "learning_rate": 2.9871326107005392e-05, "loss": 0.0411, "step": 65030 }, { "epoch": 2.1988572974069442, "grad_norm": 0.25817039608955383, "learning_rate": 2.9856191548986735e-05, "loss": 0.051, "step": 65040 }, { "epoch": 2.1991953750971973, "grad_norm": 0.31295403838157654, "learning_rate": 2.9841059193864672e-05, "loss": 0.0433, "step": 65050 }, { "epoch": 2.1995334527874504, "grad_norm": 0.14925719797611237, "learning_rate": 2.982592904329407e-05, "loss": 0.0418, "step": 65060 }, { "epoch": 2.199871530477704, "grad_norm": 0.8971416354179382, "learning_rate": 2.981080109892954e-05, "loss": 0.0491, "step": 65070 }, { "epoch": 2.200209608167957, "grad_norm": 0.25443539023399353, "learning_rate": 2.9795675362425434e-05, "loss": 0.0445, "step": 65080 }, { "epoch": 2.20054768585821, "grad_norm": 0.5215486884117126, "learning_rate": 2.97805518354359e-05, "loss": 0.0484, "step": 65090 }, { "epoch": 2.2008857635484635, "grad_norm": 0.3445832133293152, "learning_rate": 2.9765430519614825e-05, "loss": 0.0439, "step": 65100 }, { "epoch": 2.2012238412387166, "grad_norm": 0.23459492623806, "learning_rate": 2.975031141661584e-05, "loss": 0.0452, "step": 65110 }, { "epoch": 2.2015619189289697, "grad_norm": 0.21544727683067322, "learning_rate": 2.9735194528092347e-05, "loss": 0.0454, "step": 65120 }, { "epoch": 2.201899996619223, "grad_norm": 0.22487325966358185, "learning_rate": 2.9720079855697525e-05, "loss": 0.0456, "step": 65130 }, { "epoch": 2.2022380743094763, "grad_norm": 0.2564537525177002, "learning_rate": 2.970496740108427e-05, "loss": 0.0472, "step": 65140 }, { "epoch": 2.2025761519997293, "grad_norm": 0.1769905835390091, "learning_rate": 2.968985716590527e-05, "loss": 0.046, "step": 65150 }, { "epoch": 2.202914229689983, "grad_norm": 0.21474474668502808, "learning_rate": 2.9674749151812953e-05, "loss": 0.0444, "step": 65160 }, { "epoch": 2.203252307380236, "grad_norm": 0.3516477048397064, "learning_rate": 2.965964336045951e-05, "loss": 0.0489, "step": 65170 }, { "epoch": 2.2035903850704894, "grad_norm": 0.24493427574634552, "learning_rate": 2.9644539793496894e-05, "loss": 0.0397, "step": 65180 }, { "epoch": 2.2039284627607425, "grad_norm": 0.780190646648407, "learning_rate": 2.9629438452576807e-05, "loss": 0.0439, "step": 65190 }, { "epoch": 2.2042665404509956, "grad_norm": 0.37356215715408325, "learning_rate": 2.9614339339350698e-05, "loss": 0.0436, "step": 65200 }, { "epoch": 2.204604618141249, "grad_norm": 0.3264528810977936, "learning_rate": 2.959924245546979e-05, "loss": 0.0512, "step": 65210 }, { "epoch": 2.204942695831502, "grad_norm": 0.2373240441083908, "learning_rate": 2.9584147802585038e-05, "loss": 0.0431, "step": 65220 }, { "epoch": 2.2052807735217552, "grad_norm": 0.217348113656044, "learning_rate": 2.956905538234721e-05, "loss": 0.0512, "step": 65230 }, { "epoch": 2.2056188512120087, "grad_norm": 0.283059298992157, "learning_rate": 2.9553965196406764e-05, "loss": 0.041, "step": 65240 }, { "epoch": 2.205956928902262, "grad_norm": 0.23230315744876862, "learning_rate": 2.9538877246413943e-05, "loss": 0.05, "step": 65250 }, { "epoch": 2.206295006592515, "grad_norm": 0.2578750550746918, "learning_rate": 2.952379153401874e-05, "loss": 0.0433, "step": 65260 }, { "epoch": 2.2066330842827684, "grad_norm": 0.23050853610038757, "learning_rate": 2.9508708060870914e-05, "loss": 0.0395, "step": 65270 }, { "epoch": 2.2069711619730215, "grad_norm": 0.3213701844215393, "learning_rate": 2.9493626828619948e-05, "loss": 0.0491, "step": 65280 }, { "epoch": 2.2073092396632745, "grad_norm": 0.5708706974983215, "learning_rate": 2.9478547838915137e-05, "loss": 0.0514, "step": 65290 }, { "epoch": 2.207647317353528, "grad_norm": 0.6202360987663269, "learning_rate": 2.946347109340548e-05, "loss": 0.0487, "step": 65300 }, { "epoch": 2.207985395043781, "grad_norm": 0.2110653817653656, "learning_rate": 2.944839659373976e-05, "loss": 0.0436, "step": 65310 }, { "epoch": 2.208323472734034, "grad_norm": 0.21476627886295319, "learning_rate": 2.943332434156647e-05, "loss": 0.0491, "step": 65320 }, { "epoch": 2.2086615504242877, "grad_norm": 0.19456954300403595, "learning_rate": 2.9418254338533912e-05, "loss": 0.0455, "step": 65330 }, { "epoch": 2.2089996281145408, "grad_norm": 0.21786092221736908, "learning_rate": 2.940318658629011e-05, "loss": 0.0456, "step": 65340 }, { "epoch": 2.209337705804794, "grad_norm": 0.17392003536224365, "learning_rate": 2.938812108648287e-05, "loss": 0.0395, "step": 65350 }, { "epoch": 2.2096757834950473, "grad_norm": 0.34690061211586, "learning_rate": 2.937305784075972e-05, "loss": 0.0465, "step": 65360 }, { "epoch": 2.2100138611853004, "grad_norm": 0.2796071171760559, "learning_rate": 2.9357996850767937e-05, "loss": 0.0438, "step": 65370 }, { "epoch": 2.2103519388755535, "grad_norm": 0.3122609555721283, "learning_rate": 2.9342938118154595e-05, "loss": 0.0431, "step": 65380 }, { "epoch": 2.210690016565807, "grad_norm": 0.19966530799865723, "learning_rate": 2.9327881644566485e-05, "loss": 0.0426, "step": 65390 }, { "epoch": 2.21102809425606, "grad_norm": 0.23554359376430511, "learning_rate": 2.9312827431650157e-05, "loss": 0.0468, "step": 65400 }, { "epoch": 2.211366171946313, "grad_norm": 0.17114798724651337, "learning_rate": 2.92977754810519e-05, "loss": 0.0463, "step": 65410 }, { "epoch": 2.2117042496365666, "grad_norm": 0.2754601538181305, "learning_rate": 2.928272579441781e-05, "loss": 0.051, "step": 65420 }, { "epoch": 2.2120423273268197, "grad_norm": 0.16763131320476532, "learning_rate": 2.9267678373393676e-05, "loss": 0.0443, "step": 65430 }, { "epoch": 2.212380405017073, "grad_norm": 0.34151917695999146, "learning_rate": 2.925263321962507e-05, "loss": 0.0441, "step": 65440 }, { "epoch": 2.2127184827073263, "grad_norm": 0.304755836725235, "learning_rate": 2.9237590334757292e-05, "loss": 0.0505, "step": 65450 }, { "epoch": 2.2130565603975794, "grad_norm": 0.2615790069103241, "learning_rate": 2.922254972043543e-05, "loss": 0.0478, "step": 65460 }, { "epoch": 2.2133946380878324, "grad_norm": 0.32723748683929443, "learning_rate": 2.9207511378304265e-05, "loss": 0.0458, "step": 65470 }, { "epoch": 2.213732715778086, "grad_norm": 0.1437925398349762, "learning_rate": 2.9192475310008428e-05, "loss": 0.0503, "step": 65480 }, { "epoch": 2.214070793468339, "grad_norm": 0.36432838439941406, "learning_rate": 2.917744151719219e-05, "loss": 0.0491, "step": 65490 }, { "epoch": 2.214408871158592, "grad_norm": 0.34519386291503906, "learning_rate": 2.916241000149964e-05, "loss": 0.0458, "step": 65500 }, { "epoch": 2.2147469488488456, "grad_norm": 0.26177355647087097, "learning_rate": 2.9147380764574622e-05, "loss": 0.0488, "step": 65510 }, { "epoch": 2.2150850265390987, "grad_norm": 0.31098735332489014, "learning_rate": 2.9132353808060674e-05, "loss": 0.0515, "step": 65520 }, { "epoch": 2.2154231042293517, "grad_norm": 0.28529492020606995, "learning_rate": 2.9117329133601134e-05, "loss": 0.048, "step": 65530 }, { "epoch": 2.2157611819196052, "grad_norm": 0.3276045024394989, "learning_rate": 2.9102306742839093e-05, "loss": 0.0518, "step": 65540 }, { "epoch": 2.2160992596098583, "grad_norm": 0.22255083918571472, "learning_rate": 2.9087286637417375e-05, "loss": 0.0413, "step": 65550 }, { "epoch": 2.2164373373001114, "grad_norm": 0.16098296642303467, "learning_rate": 2.9072268818978537e-05, "loss": 0.0457, "step": 65560 }, { "epoch": 2.216775414990365, "grad_norm": 0.3154144287109375, "learning_rate": 2.905725328916493e-05, "loss": 0.0444, "step": 65570 }, { "epoch": 2.217113492680618, "grad_norm": 0.3628511130809784, "learning_rate": 2.9042240049618608e-05, "loss": 0.0415, "step": 65580 }, { "epoch": 2.217451570370871, "grad_norm": 0.43074318766593933, "learning_rate": 2.9027229101981413e-05, "loss": 0.0421, "step": 65590 }, { "epoch": 2.2177896480611246, "grad_norm": 0.3197891414165497, "learning_rate": 2.901222044789488e-05, "loss": 0.0477, "step": 65600 }, { "epoch": 2.2181277257513776, "grad_norm": 0.2943876087665558, "learning_rate": 2.89972140890004e-05, "loss": 0.0442, "step": 65610 }, { "epoch": 2.2184658034416307, "grad_norm": 0.27015620470046997, "learning_rate": 2.8982210026938988e-05, "loss": 0.047, "step": 65620 }, { "epoch": 2.218803881131884, "grad_norm": 0.21234990656375885, "learning_rate": 2.896720826335151e-05, "loss": 0.0489, "step": 65630 }, { "epoch": 2.2191419588221373, "grad_norm": 0.17253436148166656, "learning_rate": 2.8952208799878493e-05, "loss": 0.047, "step": 65640 }, { "epoch": 2.2194800365123903, "grad_norm": 0.9537333250045776, "learning_rate": 2.89372116381603e-05, "loss": 0.043, "step": 65650 }, { "epoch": 2.219818114202644, "grad_norm": 1.0723567008972168, "learning_rate": 2.8922216779836932e-05, "loss": 0.0436, "step": 65660 }, { "epoch": 2.220156191892897, "grad_norm": 0.21475011110305786, "learning_rate": 2.8907224226548285e-05, "loss": 0.0429, "step": 65670 }, { "epoch": 2.22049426958315, "grad_norm": 0.2409128099679947, "learning_rate": 2.8892233979933868e-05, "loss": 0.0436, "step": 65680 }, { "epoch": 2.2208323472734035, "grad_norm": 0.30520138144493103, "learning_rate": 2.8877246041633026e-05, "loss": 0.0469, "step": 65690 }, { "epoch": 2.2211704249636566, "grad_norm": 0.16426222026348114, "learning_rate": 2.886226041328478e-05, "loss": 0.0445, "step": 65700 }, { "epoch": 2.2215085026539096, "grad_norm": 0.3444041907787323, "learning_rate": 2.8847277096527957e-05, "loss": 0.0461, "step": 65710 }, { "epoch": 2.221846580344163, "grad_norm": 0.3098791241645813, "learning_rate": 2.8832296093001105e-05, "loss": 0.0429, "step": 65720 }, { "epoch": 2.2221846580344162, "grad_norm": 0.37406304478645325, "learning_rate": 2.8817317404342547e-05, "loss": 0.0531, "step": 65730 }, { "epoch": 2.2225227357246693, "grad_norm": 1.5281122922897339, "learning_rate": 2.880234103219029e-05, "loss": 0.0487, "step": 65740 }, { "epoch": 2.222860813414923, "grad_norm": 0.3894523084163666, "learning_rate": 2.878736697818215e-05, "loss": 0.0433, "step": 65750 }, { "epoch": 2.223198891105176, "grad_norm": 0.22614045441150665, "learning_rate": 2.877239524395568e-05, "loss": 0.0471, "step": 65760 }, { "epoch": 2.2235369687954294, "grad_norm": 0.17626158893108368, "learning_rate": 2.8757425831148134e-05, "loss": 0.0429, "step": 65770 }, { "epoch": 2.2238750464856825, "grad_norm": 0.2999320924282074, "learning_rate": 2.874245874139658e-05, "loss": 0.0525, "step": 65780 }, { "epoch": 2.2242131241759355, "grad_norm": 0.2486218959093094, "learning_rate": 2.8727493976337737e-05, "loss": 0.0461, "step": 65790 }, { "epoch": 2.224551201866189, "grad_norm": 0.19922097027301788, "learning_rate": 2.8712531537608218e-05, "loss": 0.0425, "step": 65800 }, { "epoch": 2.224889279556442, "grad_norm": 0.36812180280685425, "learning_rate": 2.8697571426844217e-05, "loss": 0.0439, "step": 65810 }, { "epoch": 2.225227357246695, "grad_norm": 0.17894549667835236, "learning_rate": 2.86826136456818e-05, "loss": 0.0486, "step": 65820 }, { "epoch": 2.2255654349369487, "grad_norm": 0.468442440032959, "learning_rate": 2.8667658195756685e-05, "loss": 0.0494, "step": 65830 }, { "epoch": 2.2259035126272018, "grad_norm": 0.18726779520511627, "learning_rate": 2.8652705078704412e-05, "loss": 0.0479, "step": 65840 }, { "epoch": 2.226241590317455, "grad_norm": 0.2264174073934555, "learning_rate": 2.863775429616018e-05, "loss": 0.0487, "step": 65850 }, { "epoch": 2.2265796680077083, "grad_norm": 0.35032838582992554, "learning_rate": 2.862280584975906e-05, "loss": 0.0428, "step": 65860 }, { "epoch": 2.2269177456979614, "grad_norm": 0.22444039583206177, "learning_rate": 2.8607859741135735e-05, "loss": 0.0427, "step": 65870 }, { "epoch": 2.2272558233882145, "grad_norm": 0.2561662793159485, "learning_rate": 2.8592915971924723e-05, "loss": 0.0512, "step": 65880 }, { "epoch": 2.227593901078468, "grad_norm": 0.24147821962833405, "learning_rate": 2.8577974543760212e-05, "loss": 0.0463, "step": 65890 }, { "epoch": 2.227931978768721, "grad_norm": 0.21077726781368256, "learning_rate": 2.8563035458276222e-05, "loss": 0.0444, "step": 65900 }, { "epoch": 2.228270056458974, "grad_norm": 0.26101765036582947, "learning_rate": 2.8548098717106424e-05, "loss": 0.0432, "step": 65910 }, { "epoch": 2.2286081341492276, "grad_norm": 0.3032151460647583, "learning_rate": 2.85331643218843e-05, "loss": 0.0449, "step": 65920 }, { "epoch": 2.2289462118394807, "grad_norm": 0.2208864986896515, "learning_rate": 2.8518232274243046e-05, "loss": 0.0448, "step": 65930 }, { "epoch": 2.229284289529734, "grad_norm": 0.1637348085641861, "learning_rate": 2.8503302575815638e-05, "loss": 0.0421, "step": 65940 }, { "epoch": 2.2296223672199873, "grad_norm": 0.6126111745834351, "learning_rate": 2.8488375228234716e-05, "loss": 0.048, "step": 65950 }, { "epoch": 2.2299604449102404, "grad_norm": 0.21007436513900757, "learning_rate": 2.8473450233132738e-05, "loss": 0.044, "step": 65960 }, { "epoch": 2.2302985226004934, "grad_norm": 0.19261430203914642, "learning_rate": 2.8458527592141898e-05, "loss": 0.0493, "step": 65970 }, { "epoch": 2.230636600290747, "grad_norm": 0.3859582245349884, "learning_rate": 2.8443607306894073e-05, "loss": 0.0437, "step": 65980 }, { "epoch": 2.230974677981, "grad_norm": 0.14945875108242035, "learning_rate": 2.8428689379020946e-05, "loss": 0.0448, "step": 65990 }, { "epoch": 2.231312755671253, "grad_norm": 0.6859586834907532, "learning_rate": 2.841377381015391e-05, "loss": 0.0519, "step": 66000 }, { "epoch": 2.2316508333615066, "grad_norm": 0.14985254406929016, "learning_rate": 2.8398860601924144e-05, "loss": 0.0399, "step": 66010 }, { "epoch": 2.2319889110517597, "grad_norm": 0.27812832593917847, "learning_rate": 2.838394975596248e-05, "loss": 0.0493, "step": 66020 }, { "epoch": 2.2323269887420127, "grad_norm": 0.26171451807022095, "learning_rate": 2.83690412738996e-05, "loss": 0.0443, "step": 66030 }, { "epoch": 2.2326650664322663, "grad_norm": 0.35900864005088806, "learning_rate": 2.835413515736579e-05, "loss": 0.0486, "step": 66040 }, { "epoch": 2.2330031441225193, "grad_norm": 0.20244506001472473, "learning_rate": 2.8339231407991267e-05, "loss": 0.047, "step": 66050 }, { "epoch": 2.2333412218127724, "grad_norm": 0.21969878673553467, "learning_rate": 2.832433002740581e-05, "loss": 0.0493, "step": 66060 }, { "epoch": 2.233679299503026, "grad_norm": 0.22672776877880096, "learning_rate": 2.8309431017239047e-05, "loss": 0.0513, "step": 66070 }, { "epoch": 2.234017377193279, "grad_norm": 0.2769089937210083, "learning_rate": 2.829453437912028e-05, "loss": 0.0493, "step": 66080 }, { "epoch": 2.234355454883532, "grad_norm": 0.24823898077011108, "learning_rate": 2.8279640114678617e-05, "loss": 0.0492, "step": 66090 }, { "epoch": 2.2346935325737856, "grad_norm": 0.3763965666294098, "learning_rate": 2.8264748225542836e-05, "loss": 0.046, "step": 66100 }, { "epoch": 2.2350316102640386, "grad_norm": 0.23926959931850433, "learning_rate": 2.8249858713341516e-05, "loss": 0.0481, "step": 66110 }, { "epoch": 2.2353696879542917, "grad_norm": 0.27817970514297485, "learning_rate": 2.823497157970294e-05, "loss": 0.0485, "step": 66120 }, { "epoch": 2.235707765644545, "grad_norm": 0.3040577173233032, "learning_rate": 2.8220086826255164e-05, "loss": 0.0465, "step": 66130 }, { "epoch": 2.2360458433347983, "grad_norm": 0.31465739011764526, "learning_rate": 2.8205204454625934e-05, "loss": 0.0475, "step": 66140 }, { "epoch": 2.236383921025052, "grad_norm": 0.25443536043167114, "learning_rate": 2.8190324466442798e-05, "loss": 0.0498, "step": 66150 }, { "epoch": 2.236721998715305, "grad_norm": 0.2412964105606079, "learning_rate": 2.8175446863332965e-05, "loss": 0.0474, "step": 66160 }, { "epoch": 2.237060076405558, "grad_norm": 0.19330507516860962, "learning_rate": 2.816057164692345e-05, "loss": 0.0427, "step": 66170 }, { "epoch": 2.2373981540958114, "grad_norm": 0.2013363093137741, "learning_rate": 2.8145698818840992e-05, "loss": 0.045, "step": 66180 }, { "epoch": 2.2377362317860645, "grad_norm": 0.2622204124927521, "learning_rate": 2.8130828380712065e-05, "loss": 0.045, "step": 66190 }, { "epoch": 2.2380743094763176, "grad_norm": 0.35345762968063354, "learning_rate": 2.811596033416285e-05, "loss": 0.045, "step": 66200 }, { "epoch": 2.238412387166571, "grad_norm": 0.21244199573993683, "learning_rate": 2.8101094680819307e-05, "loss": 0.0479, "step": 66210 }, { "epoch": 2.238750464856824, "grad_norm": 0.1694752722978592, "learning_rate": 2.808623142230715e-05, "loss": 0.0451, "step": 66220 }, { "epoch": 2.2390885425470772, "grad_norm": 0.6131268739700317, "learning_rate": 2.8071370560251757e-05, "loss": 0.0459, "step": 66230 }, { "epoch": 2.2394266202373307, "grad_norm": 0.4852622449398041, "learning_rate": 2.805651209627831e-05, "loss": 0.0424, "step": 66240 }, { "epoch": 2.239764697927584, "grad_norm": 0.3439071774482727, "learning_rate": 2.804165603201171e-05, "loss": 0.0489, "step": 66250 }, { "epoch": 2.240102775617837, "grad_norm": 0.2499481588602066, "learning_rate": 2.802680236907661e-05, "loss": 0.0486, "step": 66260 }, { "epoch": 2.2404408533080904, "grad_norm": 0.25755566358566284, "learning_rate": 2.8011951109097344e-05, "loss": 0.0451, "step": 66270 }, { "epoch": 2.2407789309983435, "grad_norm": 0.2384946048259735, "learning_rate": 2.7997102253698072e-05, "loss": 0.0485, "step": 66280 }, { "epoch": 2.2411170086885965, "grad_norm": 0.32899969816207886, "learning_rate": 2.7982255804502588e-05, "loss": 0.0477, "step": 66290 }, { "epoch": 2.24145508637885, "grad_norm": 0.21600580215454102, "learning_rate": 2.7967411763134516e-05, "loss": 0.0409, "step": 66300 }, { "epoch": 2.241793164069103, "grad_norm": 0.22627486288547516, "learning_rate": 2.7952570131217166e-05, "loss": 0.0436, "step": 66310 }, { "epoch": 2.242131241759356, "grad_norm": 0.1400861144065857, "learning_rate": 2.7937730910373618e-05, "loss": 0.0467, "step": 66320 }, { "epoch": 2.2424693194496097, "grad_norm": 0.4060005843639374, "learning_rate": 2.7922894102226628e-05, "loss": 0.0447, "step": 66330 }, { "epoch": 2.2428073971398628, "grad_norm": 0.23108991980552673, "learning_rate": 2.790805970839876e-05, "loss": 0.0456, "step": 66340 }, { "epoch": 2.243145474830116, "grad_norm": 0.21021831035614014, "learning_rate": 2.789322773051225e-05, "loss": 0.0522, "step": 66350 }, { "epoch": 2.2434835525203694, "grad_norm": 0.3651450276374817, "learning_rate": 2.7878398170189124e-05, "loss": 0.0475, "step": 66360 }, { "epoch": 2.2438216302106224, "grad_norm": 0.20245695114135742, "learning_rate": 2.7863571029051112e-05, "loss": 0.0462, "step": 66370 }, { "epoch": 2.2441597079008755, "grad_norm": 0.20590244233608246, "learning_rate": 2.784874630871971e-05, "loss": 0.0467, "step": 66380 }, { "epoch": 2.244497785591129, "grad_norm": 0.20821517705917358, "learning_rate": 2.7833924010816086e-05, "loss": 0.0437, "step": 66390 }, { "epoch": 2.244835863281382, "grad_norm": 0.17698867619037628, "learning_rate": 2.7819104136961226e-05, "loss": 0.0452, "step": 66400 }, { "epoch": 2.245173940971635, "grad_norm": 0.3489745557308197, "learning_rate": 2.780428668877577e-05, "loss": 0.0466, "step": 66410 }, { "epoch": 2.2455120186618887, "grad_norm": 0.8312954902648926, "learning_rate": 2.7789471667880152e-05, "loss": 0.0442, "step": 66420 }, { "epoch": 2.2458500963521417, "grad_norm": 0.2610629200935364, "learning_rate": 2.7774659075894523e-05, "loss": 0.0468, "step": 66430 }, { "epoch": 2.246188174042395, "grad_norm": 0.2827656865119934, "learning_rate": 2.7759848914438778e-05, "loss": 0.0469, "step": 66440 }, { "epoch": 2.2465262517326483, "grad_norm": 0.19354236125946045, "learning_rate": 2.7745041185132496e-05, "loss": 0.0432, "step": 66450 }, { "epoch": 2.2468643294229014, "grad_norm": 0.1968930959701538, "learning_rate": 2.773023588959506e-05, "loss": 0.0463, "step": 66460 }, { "epoch": 2.2472024071131544, "grad_norm": 0.37545013427734375, "learning_rate": 2.7715433029445555e-05, "loss": 0.05, "step": 66470 }, { "epoch": 2.247540484803408, "grad_norm": 0.34640195965766907, "learning_rate": 2.7700632606302766e-05, "loss": 0.0479, "step": 66480 }, { "epoch": 2.247878562493661, "grad_norm": 0.21420444548130035, "learning_rate": 2.768583462178528e-05, "loss": 0.046, "step": 66490 }, { "epoch": 2.248216640183914, "grad_norm": 0.20287290215492249, "learning_rate": 2.767103907751136e-05, "loss": 0.0417, "step": 66500 }, { "epoch": 2.2485547178741676, "grad_norm": 0.31916552782058716, "learning_rate": 2.7656245975099065e-05, "loss": 0.0464, "step": 66510 }, { "epoch": 2.2488927955644207, "grad_norm": 0.17081019282341003, "learning_rate": 2.7641455316166093e-05, "loss": 0.0458, "step": 66520 }, { "epoch": 2.2492308732546737, "grad_norm": 0.26066190004348755, "learning_rate": 2.7626667102329973e-05, "loss": 0.0467, "step": 66530 }, { "epoch": 2.2495689509449273, "grad_norm": 0.2847929298877716, "learning_rate": 2.761188133520788e-05, "loss": 0.0463, "step": 66540 }, { "epoch": 2.2499070286351803, "grad_norm": 0.2342757135629654, "learning_rate": 2.7597098016416784e-05, "loss": 0.0452, "step": 66550 }, { "epoch": 2.2502451063254334, "grad_norm": 0.2022620290517807, "learning_rate": 2.7582317147573368e-05, "loss": 0.042, "step": 66560 }, { "epoch": 2.250583184015687, "grad_norm": 0.2853148281574249, "learning_rate": 2.7567538730294064e-05, "loss": 0.0434, "step": 66570 }, { "epoch": 2.25092126170594, "grad_norm": 0.3631168603897095, "learning_rate": 2.7552762766194973e-05, "loss": 0.0462, "step": 66580 }, { "epoch": 2.251259339396193, "grad_norm": 0.21028363704681396, "learning_rate": 2.753798925689202e-05, "loss": 0.0461, "step": 66590 }, { "epoch": 2.2515974170864466, "grad_norm": 0.2717556059360504, "learning_rate": 2.7523218204000774e-05, "loss": 0.0462, "step": 66600 }, { "epoch": 2.2519354947766996, "grad_norm": 0.19948841631412506, "learning_rate": 2.7508449609136612e-05, "loss": 0.0483, "step": 66610 }, { "epoch": 2.2522735724669527, "grad_norm": 0.2604043781757355, "learning_rate": 2.749368347391454e-05, "loss": 0.0422, "step": 66620 }, { "epoch": 2.252611650157206, "grad_norm": 0.19681300222873688, "learning_rate": 2.7478919799949443e-05, "loss": 0.0441, "step": 66630 }, { "epoch": 2.2529497278474593, "grad_norm": 0.30819597840309143, "learning_rate": 2.7464158588855803e-05, "loss": 0.0506, "step": 66640 }, { "epoch": 2.2532878055377124, "grad_norm": 0.21515119075775146, "learning_rate": 2.7449399842247918e-05, "loss": 0.0493, "step": 66650 }, { "epoch": 2.253625883227966, "grad_norm": 0.29874369502067566, "learning_rate": 2.7434643561739736e-05, "loss": 0.0445, "step": 66660 }, { "epoch": 2.253963960918219, "grad_norm": 0.20684079825878143, "learning_rate": 2.7419889748945016e-05, "loss": 0.0461, "step": 66670 }, { "epoch": 2.254302038608472, "grad_norm": 0.14532825350761414, "learning_rate": 2.7405138405477203e-05, "loss": 0.0433, "step": 66680 }, { "epoch": 2.2546401162987255, "grad_norm": 0.22581498324871063, "learning_rate": 2.73903895329495e-05, "loss": 0.043, "step": 66690 }, { "epoch": 2.2549781939889786, "grad_norm": 0.24267859756946564, "learning_rate": 2.737564313297478e-05, "loss": 0.0503, "step": 66700 }, { "epoch": 2.2553162716792317, "grad_norm": 0.26450079679489136, "learning_rate": 2.736089920716572e-05, "loss": 0.046, "step": 66710 }, { "epoch": 2.255654349369485, "grad_norm": 0.516638457775116, "learning_rate": 2.7346157757134706e-05, "loss": 0.0446, "step": 66720 }, { "epoch": 2.2559924270597382, "grad_norm": 0.27942273020744324, "learning_rate": 2.733141878449379e-05, "loss": 0.05, "step": 66730 }, { "epoch": 2.2563305047499913, "grad_norm": 0.21025097370147705, "learning_rate": 2.7316682290854844e-05, "loss": 0.0433, "step": 66740 }, { "epoch": 2.256668582440245, "grad_norm": 0.6201611757278442, "learning_rate": 2.7301948277829414e-05, "loss": 0.0476, "step": 66750 }, { "epoch": 2.257006660130498, "grad_norm": 0.2163926661014557, "learning_rate": 2.7287216747028814e-05, "loss": 0.0486, "step": 66760 }, { "epoch": 2.2573447378207514, "grad_norm": 0.21210382878780365, "learning_rate": 2.7272487700064025e-05, "loss": 0.0437, "step": 66770 }, { "epoch": 2.2576828155110045, "grad_norm": 0.21026282012462616, "learning_rate": 2.725776113854583e-05, "loss": 0.0468, "step": 66780 }, { "epoch": 2.2580208932012575, "grad_norm": 0.3017377555370331, "learning_rate": 2.7243037064084664e-05, "loss": 0.0465, "step": 66790 }, { "epoch": 2.258358970891511, "grad_norm": 0.41546744108200073, "learning_rate": 2.7228315478290766e-05, "loss": 0.0444, "step": 66800 }, { "epoch": 2.258697048581764, "grad_norm": 0.31412163376808167, "learning_rate": 2.721359638277402e-05, "loss": 0.0417, "step": 66810 }, { "epoch": 2.259035126272017, "grad_norm": 0.2244536280632019, "learning_rate": 2.7198879779144148e-05, "loss": 0.0433, "step": 66820 }, { "epoch": 2.2593732039622707, "grad_norm": 0.19210998713970184, "learning_rate": 2.7184165669010485e-05, "loss": 0.0476, "step": 66830 }, { "epoch": 2.2597112816525238, "grad_norm": 0.30566123127937317, "learning_rate": 2.716945405398218e-05, "loss": 0.0472, "step": 66840 }, { "epoch": 2.260049359342777, "grad_norm": 0.23536811769008636, "learning_rate": 2.715474493566803e-05, "loss": 0.0426, "step": 66850 }, { "epoch": 2.2603874370330304, "grad_norm": 0.2679847478866577, "learning_rate": 2.7140038315676652e-05, "loss": 0.0427, "step": 66860 }, { "epoch": 2.2607255147232834, "grad_norm": 0.1968529373407364, "learning_rate": 2.7125334195616275e-05, "loss": 0.0465, "step": 66870 }, { "epoch": 2.2610635924135365, "grad_norm": 0.2832529842853546, "learning_rate": 2.7110632577094997e-05, "loss": 0.0504, "step": 66880 }, { "epoch": 2.26140167010379, "grad_norm": 0.22662171721458435, "learning_rate": 2.709593346172051e-05, "loss": 0.0471, "step": 66890 }, { "epoch": 2.261739747794043, "grad_norm": 0.1505359262228012, "learning_rate": 2.708123685110032e-05, "loss": 0.046, "step": 66900 }, { "epoch": 2.262077825484296, "grad_norm": 0.1983962059020996, "learning_rate": 2.7066542746841588e-05, "loss": 0.0533, "step": 66910 }, { "epoch": 2.2624159031745497, "grad_norm": 0.16968882083892822, "learning_rate": 2.7051851150551267e-05, "loss": 0.0491, "step": 66920 }, { "epoch": 2.2627539808648027, "grad_norm": 0.2345036417245865, "learning_rate": 2.7037162063835996e-05, "loss": 0.0478, "step": 66930 }, { "epoch": 2.263092058555056, "grad_norm": 0.40038853883743286, "learning_rate": 2.702247548830218e-05, "loss": 0.0433, "step": 66940 }, { "epoch": 2.2634301362453093, "grad_norm": 0.23750588297843933, "learning_rate": 2.7007791425555883e-05, "loss": 0.0425, "step": 66950 }, { "epoch": 2.2637682139355624, "grad_norm": 0.2696927785873413, "learning_rate": 2.6993109877202945e-05, "loss": 0.0467, "step": 66960 }, { "epoch": 2.2641062916258154, "grad_norm": 0.27039965987205505, "learning_rate": 2.6978430844848935e-05, "loss": 0.0418, "step": 66970 }, { "epoch": 2.264444369316069, "grad_norm": 0.18382082879543304, "learning_rate": 2.69637543300991e-05, "loss": 0.0396, "step": 66980 }, { "epoch": 2.264782447006322, "grad_norm": 0.18888312578201294, "learning_rate": 2.694908033455848e-05, "loss": 0.048, "step": 66990 }, { "epoch": 2.265120524696575, "grad_norm": 0.28577739000320435, "learning_rate": 2.6934408859831733e-05, "loss": 0.0464, "step": 67000 }, { "epoch": 2.2654586023868286, "grad_norm": 0.314182311296463, "learning_rate": 2.69197399075234e-05, "loss": 0.0467, "step": 67010 }, { "epoch": 2.2657966800770817, "grad_norm": 0.2487298995256424, "learning_rate": 2.6905073479237584e-05, "loss": 0.0469, "step": 67020 }, { "epoch": 2.2661347577673348, "grad_norm": 0.23529531061649323, "learning_rate": 2.6890409576578235e-05, "loss": 0.045, "step": 67030 }, { "epoch": 2.2664728354575883, "grad_norm": 0.2221117913722992, "learning_rate": 2.687574820114892e-05, "loss": 0.047, "step": 67040 }, { "epoch": 2.2668109131478413, "grad_norm": 0.2161986529827118, "learning_rate": 2.686108935455305e-05, "loss": 0.0457, "step": 67050 }, { "epoch": 2.2671489908380944, "grad_norm": 0.1726965457201004, "learning_rate": 2.6846433038393603e-05, "loss": 0.0422, "step": 67060 }, { "epoch": 2.267487068528348, "grad_norm": 0.25641101598739624, "learning_rate": 2.6831779254273482e-05, "loss": 0.0443, "step": 67070 }, { "epoch": 2.267825146218601, "grad_norm": 0.36412596702575684, "learning_rate": 2.6817128003795122e-05, "loss": 0.0433, "step": 67080 }, { "epoch": 2.2681632239088545, "grad_norm": 0.1900564730167389, "learning_rate": 2.680247928856081e-05, "loss": 0.0419, "step": 67090 }, { "epoch": 2.2685013015991076, "grad_norm": 0.3394818603992462, "learning_rate": 2.6787833110172466e-05, "loss": 0.0444, "step": 67100 }, { "epoch": 2.2688393792893606, "grad_norm": 0.4223085939884186, "learning_rate": 2.6773189470231807e-05, "loss": 0.0431, "step": 67110 }, { "epoch": 2.269177456979614, "grad_norm": 0.32300180196762085, "learning_rate": 2.675854837034022e-05, "loss": 0.0435, "step": 67120 }, { "epoch": 2.269515534669867, "grad_norm": 0.24216753244400024, "learning_rate": 2.6743909812098833e-05, "loss": 0.0479, "step": 67130 }, { "epoch": 2.2698536123601203, "grad_norm": 0.34795132279396057, "learning_rate": 2.67292737971085e-05, "loss": 0.0467, "step": 67140 }, { "epoch": 2.270191690050374, "grad_norm": 0.26209017634391785, "learning_rate": 2.671464032696982e-05, "loss": 0.0436, "step": 67150 }, { "epoch": 2.270529767740627, "grad_norm": 0.2170151174068451, "learning_rate": 2.6700009403283055e-05, "loss": 0.0447, "step": 67160 }, { "epoch": 2.27086784543088, "grad_norm": 0.18615806102752686, "learning_rate": 2.6685381027648214e-05, "loss": 0.0472, "step": 67170 }, { "epoch": 2.2712059231211335, "grad_norm": 0.15373282134532928, "learning_rate": 2.6670755201665086e-05, "loss": 0.0436, "step": 67180 }, { "epoch": 2.2715440008113865, "grad_norm": 0.22497832775115967, "learning_rate": 2.6656131926933066e-05, "loss": 0.0454, "step": 67190 }, { "epoch": 2.2718820785016396, "grad_norm": 0.28716686367988586, "learning_rate": 2.664151120505136e-05, "loss": 0.0502, "step": 67200 }, { "epoch": 2.272220156191893, "grad_norm": 0.19929037988185883, "learning_rate": 2.6626893037618873e-05, "loss": 0.0415, "step": 67210 }, { "epoch": 2.272558233882146, "grad_norm": 0.20663605630397797, "learning_rate": 2.6612277426234243e-05, "loss": 0.0463, "step": 67220 }, { "epoch": 2.2728963115723992, "grad_norm": 0.2159481793642044, "learning_rate": 2.659766437249577e-05, "loss": 0.0457, "step": 67230 }, { "epoch": 2.2732343892626528, "grad_norm": 0.19168739020824432, "learning_rate": 2.6583053878001557e-05, "loss": 0.0462, "step": 67240 }, { "epoch": 2.273572466952906, "grad_norm": 0.3101035952568054, "learning_rate": 2.6568445944349323e-05, "loss": 0.0468, "step": 67250 }, { "epoch": 2.273910544643159, "grad_norm": 0.3167805075645447, "learning_rate": 2.6553840573136657e-05, "loss": 0.0524, "step": 67260 }, { "epoch": 2.2742486223334124, "grad_norm": 0.18819743394851685, "learning_rate": 2.6539237765960706e-05, "loss": 0.0392, "step": 67270 }, { "epoch": 2.2745867000236655, "grad_norm": 0.23409444093704224, "learning_rate": 2.6524637524418465e-05, "loss": 0.0443, "step": 67280 }, { "epoch": 2.2749247777139185, "grad_norm": 0.23197874426841736, "learning_rate": 2.651003985010655e-05, "loss": 0.049, "step": 67290 }, { "epoch": 2.275262855404172, "grad_norm": 0.2825290560722351, "learning_rate": 2.649544474462138e-05, "loss": 0.0456, "step": 67300 }, { "epoch": 2.275600933094425, "grad_norm": 0.3449375033378601, "learning_rate": 2.6480852209559015e-05, "loss": 0.0416, "step": 67310 }, { "epoch": 2.275939010784678, "grad_norm": 0.21460886299610138, "learning_rate": 2.6466262246515296e-05, "loss": 0.0442, "step": 67320 }, { "epoch": 2.2762770884749317, "grad_norm": 0.23591774702072144, "learning_rate": 2.6451674857085746e-05, "loss": 0.0471, "step": 67330 }, { "epoch": 2.276615166165185, "grad_norm": 0.19812339544296265, "learning_rate": 2.6437090042865655e-05, "loss": 0.0417, "step": 67340 }, { "epoch": 2.276953243855438, "grad_norm": 0.3011910021305084, "learning_rate": 2.6422507805449947e-05, "loss": 0.0483, "step": 67350 }, { "epoch": 2.2772913215456914, "grad_norm": 0.2690633535385132, "learning_rate": 2.6407928146433357e-05, "loss": 0.0433, "step": 67360 }, { "epoch": 2.2776293992359444, "grad_norm": 0.251495897769928, "learning_rate": 2.639335106741026e-05, "loss": 0.0392, "step": 67370 }, { "epoch": 2.2779674769261975, "grad_norm": 0.2485750913619995, "learning_rate": 2.6378776569974796e-05, "loss": 0.0481, "step": 67380 }, { "epoch": 2.278305554616451, "grad_norm": 0.3746238350868225, "learning_rate": 2.6364204655720814e-05, "loss": 0.0466, "step": 67390 }, { "epoch": 2.278643632306704, "grad_norm": 0.370719313621521, "learning_rate": 2.6349635326241894e-05, "loss": 0.0484, "step": 67400 }, { "epoch": 2.278981709996957, "grad_norm": 0.41142264008522034, "learning_rate": 2.633506858313127e-05, "loss": 0.0473, "step": 67410 }, { "epoch": 2.2793197876872107, "grad_norm": 0.13849356770515442, "learning_rate": 2.6320504427981975e-05, "loss": 0.0446, "step": 67420 }, { "epoch": 2.2796578653774637, "grad_norm": 0.2920859754085541, "learning_rate": 2.6305942862386734e-05, "loss": 0.0461, "step": 67430 }, { "epoch": 2.279995943067717, "grad_norm": 0.27739471197128296, "learning_rate": 2.629138388793793e-05, "loss": 0.0457, "step": 67440 }, { "epoch": 2.2803340207579703, "grad_norm": 0.2570948898792267, "learning_rate": 2.6276827506227743e-05, "loss": 0.0487, "step": 67450 }, { "epoch": 2.2806720984482234, "grad_norm": 0.17353719472885132, "learning_rate": 2.6262273718848037e-05, "loss": 0.0455, "step": 67460 }, { "epoch": 2.2810101761384765, "grad_norm": 0.27941298484802246, "learning_rate": 2.6247722527390407e-05, "loss": 0.047, "step": 67470 }, { "epoch": 2.28134825382873, "grad_norm": 0.28553643822669983, "learning_rate": 2.6233173933446104e-05, "loss": 0.0413, "step": 67480 }, { "epoch": 2.281686331518983, "grad_norm": 0.3759055435657501, "learning_rate": 2.6218627938606188e-05, "loss": 0.042, "step": 67490 }, { "epoch": 2.282024409209236, "grad_norm": 0.26247069239616394, "learning_rate": 2.620408454446135e-05, "loss": 0.0507, "step": 67500 }, { "epoch": 2.2823624868994896, "grad_norm": 0.2601357400417328, "learning_rate": 2.6189543752602046e-05, "loss": 0.047, "step": 67510 }, { "epoch": 2.2827005645897427, "grad_norm": 0.2536573112010956, "learning_rate": 2.6175005564618444e-05, "loss": 0.0477, "step": 67520 }, { "epoch": 2.2830386422799958, "grad_norm": 0.19518199563026428, "learning_rate": 2.6160469982100428e-05, "loss": 0.0397, "step": 67530 }, { "epoch": 2.2833767199702493, "grad_norm": 0.40827494859695435, "learning_rate": 2.6145937006637555e-05, "loss": 0.0472, "step": 67540 }, { "epoch": 2.2837147976605023, "grad_norm": 0.23483110964298248, "learning_rate": 2.6131406639819166e-05, "loss": 0.0484, "step": 67550 }, { "epoch": 2.2840528753507554, "grad_norm": 0.2728031575679779, "learning_rate": 2.611687888323424e-05, "loss": 0.0494, "step": 67560 }, { "epoch": 2.284390953041009, "grad_norm": 0.47825947403907776, "learning_rate": 2.610235373847153e-05, "loss": 0.0487, "step": 67570 }, { "epoch": 2.284729030731262, "grad_norm": 0.17309434711933136, "learning_rate": 2.6087831207119494e-05, "loss": 0.046, "step": 67580 }, { "epoch": 2.285067108421515, "grad_norm": 0.20996136963367462, "learning_rate": 2.60733112907663e-05, "loss": 0.0446, "step": 67590 }, { "epoch": 2.2854051861117686, "grad_norm": 0.26536983251571655, "learning_rate": 2.6058793990999786e-05, "loss": 0.0478, "step": 67600 }, { "epoch": 2.2857432638020216, "grad_norm": 0.3102436661720276, "learning_rate": 2.6044279309407593e-05, "loss": 0.0513, "step": 67610 }, { "epoch": 2.2860813414922747, "grad_norm": 0.37550604343414307, "learning_rate": 2.6029767247576976e-05, "loss": 0.0475, "step": 67620 }, { "epoch": 2.2864194191825282, "grad_norm": 0.46216487884521484, "learning_rate": 2.601525780709497e-05, "loss": 0.0465, "step": 67630 }, { "epoch": 2.2867574968727813, "grad_norm": 0.2985437512397766, "learning_rate": 2.6000750989548318e-05, "loss": 0.0441, "step": 67640 }, { "epoch": 2.2870955745630344, "grad_norm": 0.1872827708721161, "learning_rate": 2.5986246796523462e-05, "loss": 0.0461, "step": 67650 }, { "epoch": 2.287433652253288, "grad_norm": 0.32726162672042847, "learning_rate": 2.5971745229606546e-05, "loss": 0.0426, "step": 67660 }, { "epoch": 2.287771729943541, "grad_norm": 0.22692671418190002, "learning_rate": 2.5957246290383442e-05, "loss": 0.0397, "step": 67670 }, { "epoch": 2.288109807633794, "grad_norm": 0.21911075711250305, "learning_rate": 2.594274998043975e-05, "loss": 0.0488, "step": 67680 }, { "epoch": 2.2884478853240475, "grad_norm": 0.20410796999931335, "learning_rate": 2.5928256301360733e-05, "loss": 0.0428, "step": 67690 }, { "epoch": 2.2887859630143006, "grad_norm": 0.41481050848960876, "learning_rate": 2.591376525473142e-05, "loss": 0.0489, "step": 67700 }, { "epoch": 2.2891240407045537, "grad_norm": 0.2864656448364258, "learning_rate": 2.5899276842136523e-05, "loss": 0.0465, "step": 67710 }, { "epoch": 2.289462118394807, "grad_norm": 0.1866173893213272, "learning_rate": 2.5884791065160495e-05, "loss": 0.0475, "step": 67720 }, { "epoch": 2.2898001960850602, "grad_norm": 0.19816716015338898, "learning_rate": 2.5870307925387436e-05, "loss": 0.0448, "step": 67730 }, { "epoch": 2.2901382737753138, "grad_norm": 0.17470836639404297, "learning_rate": 2.585582742440125e-05, "loss": 0.046, "step": 67740 }, { "epoch": 2.290476351465567, "grad_norm": 0.27702319622039795, "learning_rate": 2.5841349563785455e-05, "loss": 0.0427, "step": 67750 }, { "epoch": 2.29081442915582, "grad_norm": 0.22651216387748718, "learning_rate": 2.5826874345123355e-05, "loss": 0.0453, "step": 67760 }, { "epoch": 2.2911525068460734, "grad_norm": 0.21364331245422363, "learning_rate": 2.581240176999792e-05, "loss": 0.0425, "step": 67770 }, { "epoch": 2.2914905845363265, "grad_norm": 0.24778743088245392, "learning_rate": 2.579793183999189e-05, "loss": 0.045, "step": 67780 }, { "epoch": 2.2918286622265795, "grad_norm": 0.2828594148159027, "learning_rate": 2.5783464556687627e-05, "loss": 0.0431, "step": 67790 }, { "epoch": 2.292166739916833, "grad_norm": 0.2729300856590271, "learning_rate": 2.576899992166729e-05, "loss": 0.0515, "step": 67800 }, { "epoch": 2.292504817607086, "grad_norm": 0.23446761071681976, "learning_rate": 2.5754537936512668e-05, "loss": 0.0439, "step": 67810 }, { "epoch": 2.292842895297339, "grad_norm": 0.2828332781791687, "learning_rate": 2.574007860280535e-05, "loss": 0.0422, "step": 67820 }, { "epoch": 2.2931809729875927, "grad_norm": 0.27723345160484314, "learning_rate": 2.5725621922126518e-05, "loss": 0.0463, "step": 67830 }, { "epoch": 2.293519050677846, "grad_norm": 0.19416125118732452, "learning_rate": 2.571116789605722e-05, "loss": 0.0439, "step": 67840 }, { "epoch": 2.293857128368099, "grad_norm": 0.21644246578216553, "learning_rate": 2.5696716526178054e-05, "loss": 0.0462, "step": 67850 }, { "epoch": 2.2941952060583524, "grad_norm": 0.49845248460769653, "learning_rate": 2.568226781406946e-05, "loss": 0.0466, "step": 67860 }, { "epoch": 2.2945332837486054, "grad_norm": 0.2829638719558716, "learning_rate": 2.5667821761311474e-05, "loss": 0.0459, "step": 67870 }, { "epoch": 2.2948713614388585, "grad_norm": 0.2701898217201233, "learning_rate": 2.5653378369483917e-05, "loss": 0.0513, "step": 67880 }, { "epoch": 2.295209439129112, "grad_norm": 0.1589718610048294, "learning_rate": 2.5638937640166292e-05, "loss": 0.0551, "step": 67890 }, { "epoch": 2.295547516819365, "grad_norm": 0.18322907388210297, "learning_rate": 2.5624499574937842e-05, "loss": 0.0488, "step": 67900 }, { "epoch": 2.295885594509618, "grad_norm": 0.2195008248090744, "learning_rate": 2.5610064175377456e-05, "loss": 0.0443, "step": 67910 }, { "epoch": 2.2962236721998717, "grad_norm": 0.24161463975906372, "learning_rate": 2.5595631443063782e-05, "loss": 0.0435, "step": 67920 }, { "epoch": 2.2965617498901247, "grad_norm": 0.558134913444519, "learning_rate": 2.5581201379575177e-05, "loss": 0.0473, "step": 67930 }, { "epoch": 2.296899827580378, "grad_norm": 0.48172619938850403, "learning_rate": 2.5566773986489655e-05, "loss": 0.0442, "step": 67940 }, { "epoch": 2.2972379052706313, "grad_norm": 0.36344611644744873, "learning_rate": 2.5552349265385e-05, "loss": 0.0471, "step": 67950 }, { "epoch": 2.2975759829608844, "grad_norm": 0.45984023809432983, "learning_rate": 2.5537927217838675e-05, "loss": 0.0424, "step": 67960 }, { "epoch": 2.2979140606511375, "grad_norm": 0.19134803116321564, "learning_rate": 2.5523507845427864e-05, "loss": 0.0459, "step": 67970 }, { "epoch": 2.298252138341391, "grad_norm": 0.23425649106502533, "learning_rate": 2.5509091149729413e-05, "loss": 0.045, "step": 67980 }, { "epoch": 2.298590216031644, "grad_norm": 0.2667112648487091, "learning_rate": 2.5494677132319962e-05, "loss": 0.0475, "step": 67990 }, { "epoch": 2.298928293721897, "grad_norm": 0.29367318749427795, "learning_rate": 2.548026579477575e-05, "loss": 0.0518, "step": 68000 }, { "epoch": 2.2992663714121506, "grad_norm": 0.1875874102115631, "learning_rate": 2.5465857138672834e-05, "loss": 0.051, "step": 68010 }, { "epoch": 2.2996044491024037, "grad_norm": 0.49771738052368164, "learning_rate": 2.5451451165586848e-05, "loss": 0.048, "step": 68020 }, { "epoch": 2.2999425267926568, "grad_norm": 0.2035440057516098, "learning_rate": 2.543704787709329e-05, "loss": 0.0437, "step": 68030 }, { "epoch": 2.3002806044829103, "grad_norm": 0.2968571186065674, "learning_rate": 2.542264727476723e-05, "loss": 0.0395, "step": 68040 }, { "epoch": 2.3006186821731633, "grad_norm": 0.1914462149143219, "learning_rate": 2.5408249360183535e-05, "loss": 0.048, "step": 68050 }, { "epoch": 2.300956759863417, "grad_norm": 0.19630835950374603, "learning_rate": 2.5393854134916694e-05, "loss": 0.0434, "step": 68060 }, { "epoch": 2.30129483755367, "grad_norm": 0.19516542553901672, "learning_rate": 2.5379461600540993e-05, "loss": 0.0413, "step": 68070 }, { "epoch": 2.301632915243923, "grad_norm": 0.2139747589826584, "learning_rate": 2.5365071758630317e-05, "loss": 0.0447, "step": 68080 }, { "epoch": 2.3019709929341765, "grad_norm": 0.15854759514331818, "learning_rate": 2.535068461075839e-05, "loss": 0.0457, "step": 68090 }, { "epoch": 2.3023090706244296, "grad_norm": 0.2607887387275696, "learning_rate": 2.5336300158498516e-05, "loss": 0.0537, "step": 68100 }, { "epoch": 2.3026471483146826, "grad_norm": 0.3004482686519623, "learning_rate": 2.5321918403423793e-05, "loss": 0.0489, "step": 68110 }, { "epoch": 2.302985226004936, "grad_norm": 0.2568541169166565, "learning_rate": 2.5307539347106957e-05, "loss": 0.0477, "step": 68120 }, { "epoch": 2.3033233036951892, "grad_norm": 0.422230988740921, "learning_rate": 2.529316299112049e-05, "loss": 0.043, "step": 68130 }, { "epoch": 2.3036613813854423, "grad_norm": 0.21841026842594147, "learning_rate": 2.5278789337036567e-05, "loss": 0.0481, "step": 68140 }, { "epoch": 2.303999459075696, "grad_norm": 0.5972616672515869, "learning_rate": 2.52644183864271e-05, "loss": 0.0471, "step": 68150 }, { "epoch": 2.304337536765949, "grad_norm": 0.2743380665779114, "learning_rate": 2.525005014086363e-05, "loss": 0.0486, "step": 68160 }, { "epoch": 2.304675614456202, "grad_norm": 0.2717818021774292, "learning_rate": 2.5235684601917465e-05, "loss": 0.0537, "step": 68170 }, { "epoch": 2.3050136921464555, "grad_norm": 0.23266470432281494, "learning_rate": 2.5221321771159618e-05, "loss": 0.043, "step": 68180 }, { "epoch": 2.3053517698367085, "grad_norm": 0.22157296538352966, "learning_rate": 2.5206961650160743e-05, "loss": 0.0395, "step": 68190 }, { "epoch": 2.3056898475269616, "grad_norm": 0.37737396359443665, "learning_rate": 2.519260424049129e-05, "loss": 0.045, "step": 68200 }, { "epoch": 2.306027925217215, "grad_norm": 0.2921846807003021, "learning_rate": 2.5178249543721304e-05, "loss": 0.0456, "step": 68210 }, { "epoch": 2.306366002907468, "grad_norm": 0.4207260012626648, "learning_rate": 2.5163897561420668e-05, "loss": 0.0467, "step": 68220 }, { "epoch": 2.3067040805977213, "grad_norm": 0.38830143213272095, "learning_rate": 2.514954829515883e-05, "loss": 0.0447, "step": 68230 }, { "epoch": 2.3070421582879748, "grad_norm": 0.24371109902858734, "learning_rate": 2.513520174650505e-05, "loss": 0.0445, "step": 68240 }, { "epoch": 2.307380235978228, "grad_norm": 0.18466855585575104, "learning_rate": 2.5120857917028207e-05, "loss": 0.0525, "step": 68250 }, { "epoch": 2.307718313668481, "grad_norm": 0.18967628479003906, "learning_rate": 2.5106516808296953e-05, "loss": 0.0462, "step": 68260 }, { "epoch": 2.3080563913587344, "grad_norm": 0.2038426250219345, "learning_rate": 2.5092178421879552e-05, "loss": 0.0446, "step": 68270 }, { "epoch": 2.3083944690489875, "grad_norm": 0.16377294063568115, "learning_rate": 2.5077842759344116e-05, "loss": 0.0451, "step": 68280 }, { "epoch": 2.3087325467392406, "grad_norm": 0.18730895221233368, "learning_rate": 2.5063509822258304e-05, "loss": 0.0446, "step": 68290 }, { "epoch": 2.309070624429494, "grad_norm": 0.2856316566467285, "learning_rate": 2.5049179612189576e-05, "loss": 0.0449, "step": 68300 }, { "epoch": 2.309408702119747, "grad_norm": 0.32995009422302246, "learning_rate": 2.5034852130705043e-05, "loss": 0.0468, "step": 68310 }, { "epoch": 2.30974677981, "grad_norm": 0.1575736701488495, "learning_rate": 2.5020527379371572e-05, "loss": 0.0456, "step": 68320 }, { "epoch": 2.3100848575002537, "grad_norm": 0.23365549743175507, "learning_rate": 2.500620535975564e-05, "loss": 0.0426, "step": 68330 }, { "epoch": 2.310422935190507, "grad_norm": 0.2188318520784378, "learning_rate": 2.4991886073423515e-05, "loss": 0.0456, "step": 68340 }, { "epoch": 2.31076101288076, "grad_norm": 0.2644204795360565, "learning_rate": 2.4977569521941134e-05, "loss": 0.0476, "step": 68350 }, { "epoch": 2.3110990905710134, "grad_norm": 0.4498644173145294, "learning_rate": 2.4963255706874143e-05, "loss": 0.0491, "step": 68360 }, { "epoch": 2.3114371682612664, "grad_norm": 0.558316171169281, "learning_rate": 2.4948944629787853e-05, "loss": 0.0435, "step": 68370 }, { "epoch": 2.3117752459515195, "grad_norm": 0.26318812370300293, "learning_rate": 2.493463629224731e-05, "loss": 0.047, "step": 68380 }, { "epoch": 2.312113323641773, "grad_norm": 0.20763298869132996, "learning_rate": 2.492033069581728e-05, "loss": 0.0445, "step": 68390 }, { "epoch": 2.312451401332026, "grad_norm": 0.16155299544334412, "learning_rate": 2.4906027842062137e-05, "loss": 0.0423, "step": 68400 }, { "epoch": 2.312789479022279, "grad_norm": 0.13831186294555664, "learning_rate": 2.48917277325461e-05, "loss": 0.0382, "step": 68410 }, { "epoch": 2.3131275567125327, "grad_norm": 0.19531495869159698, "learning_rate": 2.487743036883295e-05, "loss": 0.0474, "step": 68420 }, { "epoch": 2.3134656344027857, "grad_norm": 0.3233228027820587, "learning_rate": 2.486313575248626e-05, "loss": 0.0455, "step": 68430 }, { "epoch": 2.313803712093039, "grad_norm": 0.15606443583965302, "learning_rate": 2.4848843885069235e-05, "loss": 0.0432, "step": 68440 }, { "epoch": 2.3141417897832923, "grad_norm": 0.2106475830078125, "learning_rate": 2.483455476814484e-05, "loss": 0.0443, "step": 68450 }, { "epoch": 2.3144798674735454, "grad_norm": 0.15878482162952423, "learning_rate": 2.4820268403275672e-05, "loss": 0.042, "step": 68460 }, { "epoch": 2.3148179451637985, "grad_norm": 0.3062530755996704, "learning_rate": 2.480598479202413e-05, "loss": 0.0478, "step": 68470 }, { "epoch": 2.315156022854052, "grad_norm": 0.178421288728714, "learning_rate": 2.4791703935952193e-05, "loss": 0.0421, "step": 68480 }, { "epoch": 2.315494100544305, "grad_norm": 0.2567492723464966, "learning_rate": 2.477742583662163e-05, "loss": 0.0399, "step": 68490 }, { "epoch": 2.315832178234558, "grad_norm": 0.26088789105415344, "learning_rate": 2.476315049559384e-05, "loss": 0.0461, "step": 68500 }, { "epoch": 2.3161702559248116, "grad_norm": 0.25607791543006897, "learning_rate": 2.4748877914429987e-05, "loss": 0.0464, "step": 68510 }, { "epoch": 2.3165083336150647, "grad_norm": 0.42536649107933044, "learning_rate": 2.4734608094690857e-05, "loss": 0.0475, "step": 68520 }, { "epoch": 2.3168464113053178, "grad_norm": 0.23901736736297607, "learning_rate": 2.4720341037937006e-05, "loss": 0.0514, "step": 68530 }, { "epoch": 2.3171844889955713, "grad_norm": 0.15133202075958252, "learning_rate": 2.4706076745728645e-05, "loss": 0.0406, "step": 68540 }, { "epoch": 2.3175225666858243, "grad_norm": 0.2156643271446228, "learning_rate": 2.469181521962573e-05, "loss": 0.0429, "step": 68550 }, { "epoch": 2.3178606443760774, "grad_norm": 0.2325645387172699, "learning_rate": 2.467755646118783e-05, "loss": 0.0419, "step": 68560 }, { "epoch": 2.318198722066331, "grad_norm": 0.29390406608581543, "learning_rate": 2.46633004719743e-05, "loss": 0.0429, "step": 68570 }, { "epoch": 2.318536799756584, "grad_norm": 0.3667547106742859, "learning_rate": 2.4649047253544123e-05, "loss": 0.0454, "step": 68580 }, { "epoch": 2.318874877446837, "grad_norm": 0.4937690496444702, "learning_rate": 2.4634796807456016e-05, "loss": 0.0498, "step": 68590 }, { "epoch": 2.3192129551370906, "grad_norm": 1.0800929069519043, "learning_rate": 2.46205491352684e-05, "loss": 0.044, "step": 68600 }, { "epoch": 2.3195510328273437, "grad_norm": 0.4539974629878998, "learning_rate": 2.4606304238539375e-05, "loss": 0.0465, "step": 68610 }, { "epoch": 2.3198891105175967, "grad_norm": 0.22488383948802948, "learning_rate": 2.4592062118826753e-05, "loss": 0.0454, "step": 68620 }, { "epoch": 2.3202271882078502, "grad_norm": 1.1132776737213135, "learning_rate": 2.4577822777688004e-05, "loss": 0.046, "step": 68630 }, { "epoch": 2.3205652658981033, "grad_norm": 0.3364245891571045, "learning_rate": 2.4563586216680347e-05, "loss": 0.0434, "step": 68640 }, { "epoch": 2.3209033435883564, "grad_norm": 0.5046570301055908, "learning_rate": 2.4549352437360622e-05, "loss": 0.0455, "step": 68650 }, { "epoch": 2.32124142127861, "grad_norm": 0.2097141444683075, "learning_rate": 2.4535121441285493e-05, "loss": 0.0445, "step": 68660 }, { "epoch": 2.321579498968863, "grad_norm": 0.2863748073577881, "learning_rate": 2.4520893230011172e-05, "loss": 0.0441, "step": 68670 }, { "epoch": 2.321917576659116, "grad_norm": 0.17570005357265472, "learning_rate": 2.450666780509368e-05, "loss": 0.0462, "step": 68680 }, { "epoch": 2.3222556543493695, "grad_norm": 0.2530352771282196, "learning_rate": 2.4492445168088656e-05, "loss": 0.0464, "step": 68690 }, { "epoch": 2.3225937320396226, "grad_norm": 0.24476832151412964, "learning_rate": 2.447822532055149e-05, "loss": 0.0467, "step": 68700 }, { "epoch": 2.322931809729876, "grad_norm": 0.19954189658164978, "learning_rate": 2.4464008264037212e-05, "loss": 0.046, "step": 68710 }, { "epoch": 2.323269887420129, "grad_norm": 0.15059222280979156, "learning_rate": 2.4449794000100605e-05, "loss": 0.0387, "step": 68720 }, { "epoch": 2.3236079651103823, "grad_norm": 0.17511604726314545, "learning_rate": 2.4435582530296108e-05, "loss": 0.0463, "step": 68730 }, { "epoch": 2.3239460428006358, "grad_norm": 0.15028707683086395, "learning_rate": 2.4421373856177887e-05, "loss": 0.0455, "step": 68740 }, { "epoch": 2.324284120490889, "grad_norm": 0.2047250121831894, "learning_rate": 2.440716797929975e-05, "loss": 0.0448, "step": 68750 }, { "epoch": 2.324622198181142, "grad_norm": 0.30470147728919983, "learning_rate": 2.439296490121526e-05, "loss": 0.0447, "step": 68760 }, { "epoch": 2.3249602758713954, "grad_norm": 0.17060749232769012, "learning_rate": 2.4378764623477608e-05, "loss": 0.0423, "step": 68770 }, { "epoch": 2.3252983535616485, "grad_norm": 1.295270562171936, "learning_rate": 2.436456714763974e-05, "loss": 0.0493, "step": 68780 }, { "epoch": 2.3256364312519016, "grad_norm": 0.24290025234222412, "learning_rate": 2.4350372475254262e-05, "loss": 0.049, "step": 68790 }, { "epoch": 2.325974508942155, "grad_norm": 0.34918832778930664, "learning_rate": 2.433618060787351e-05, "loss": 0.0487, "step": 68800 }, { "epoch": 2.326312586632408, "grad_norm": 0.22623342275619507, "learning_rate": 2.432199154704944e-05, "loss": 0.0495, "step": 68810 }, { "epoch": 2.326650664322661, "grad_norm": 0.208973228931427, "learning_rate": 2.430780529433377e-05, "loss": 0.0411, "step": 68820 }, { "epoch": 2.3269887420129147, "grad_norm": 0.17159424722194672, "learning_rate": 2.42936218512779e-05, "loss": 0.0455, "step": 68830 }, { "epoch": 2.327326819703168, "grad_norm": 0.1711660772562027, "learning_rate": 2.427944121943288e-05, "loss": 0.0446, "step": 68840 }, { "epoch": 2.327664897393421, "grad_norm": 0.25278422236442566, "learning_rate": 2.4265263400349497e-05, "loss": 0.0459, "step": 68850 }, { "epoch": 2.3280029750836744, "grad_norm": 0.2612024247646332, "learning_rate": 2.4251088395578213e-05, "loss": 0.0465, "step": 68860 }, { "epoch": 2.3283410527739274, "grad_norm": 0.19584335386753082, "learning_rate": 2.4236916206669214e-05, "loss": 0.0434, "step": 68870 }, { "epoch": 2.3286791304641805, "grad_norm": 0.23105940222740173, "learning_rate": 2.4222746835172304e-05, "loss": 0.0476, "step": 68880 }, { "epoch": 2.329017208154434, "grad_norm": 0.23327352106571198, "learning_rate": 2.4208580282637065e-05, "loss": 0.0475, "step": 68890 }, { "epoch": 2.329355285844687, "grad_norm": 0.40154126286506653, "learning_rate": 2.4194416550612687e-05, "loss": 0.0481, "step": 68900 }, { "epoch": 2.32969336353494, "grad_norm": 0.2375151664018631, "learning_rate": 2.418025564064812e-05, "loss": 0.045, "step": 68910 }, { "epoch": 2.3300314412251937, "grad_norm": 0.17818111181259155, "learning_rate": 2.4166097554291978e-05, "loss": 0.0471, "step": 68920 }, { "epoch": 2.3303695189154467, "grad_norm": 0.22156009078025818, "learning_rate": 2.4151942293092583e-05, "loss": 0.0438, "step": 68930 }, { "epoch": 2.3307075966057, "grad_norm": 0.3510289788246155, "learning_rate": 2.41377898585979e-05, "loss": 0.0489, "step": 68940 }, { "epoch": 2.3310456742959533, "grad_norm": 0.2001827508211136, "learning_rate": 2.4123640252355652e-05, "loss": 0.0461, "step": 68950 }, { "epoch": 2.3313837519862064, "grad_norm": 0.44761866331100464, "learning_rate": 2.410949347591318e-05, "loss": 0.0519, "step": 68960 }, { "epoch": 2.3317218296764595, "grad_norm": 0.21981742978096008, "learning_rate": 2.4095349530817578e-05, "loss": 0.0448, "step": 68970 }, { "epoch": 2.332059907366713, "grad_norm": 0.20409004390239716, "learning_rate": 2.4081208418615603e-05, "loss": 0.0448, "step": 68980 }, { "epoch": 2.332397985056966, "grad_norm": 0.1737351417541504, "learning_rate": 2.4067070140853725e-05, "loss": 0.0446, "step": 68990 }, { "epoch": 2.332736062747219, "grad_norm": 0.46724775433540344, "learning_rate": 2.405293469907805e-05, "loss": 0.0477, "step": 69000 }, { "epoch": 2.3330741404374726, "grad_norm": 1.4912644624710083, "learning_rate": 2.4038802094834446e-05, "loss": 0.0458, "step": 69010 }, { "epoch": 2.3334122181277257, "grad_norm": 0.27030888199806213, "learning_rate": 2.4024672329668395e-05, "loss": 0.0513, "step": 69020 }, { "epoch": 2.333750295817979, "grad_norm": 0.37878984212875366, "learning_rate": 2.4010545405125125e-05, "loss": 0.0467, "step": 69030 }, { "epoch": 2.3340883735082323, "grad_norm": 0.260993629693985, "learning_rate": 2.3996421322749528e-05, "loss": 0.0451, "step": 69040 }, { "epoch": 2.3344264511984854, "grad_norm": 0.24436020851135254, "learning_rate": 2.3982300084086224e-05, "loss": 0.0458, "step": 69050 }, { "epoch": 2.334764528888739, "grad_norm": 0.30753061175346375, "learning_rate": 2.396818169067944e-05, "loss": 0.0447, "step": 69060 }, { "epoch": 2.335102606578992, "grad_norm": 0.48113495111465454, "learning_rate": 2.3954066144073178e-05, "loss": 0.0476, "step": 69070 }, { "epoch": 2.335440684269245, "grad_norm": 0.21974676847457886, "learning_rate": 2.39399534458111e-05, "loss": 0.0439, "step": 69080 }, { "epoch": 2.3357787619594985, "grad_norm": 0.342398464679718, "learning_rate": 2.392584359743651e-05, "loss": 0.0422, "step": 69090 }, { "epoch": 2.3361168396497516, "grad_norm": 0.19230598211288452, "learning_rate": 2.3911736600492463e-05, "loss": 0.0496, "step": 69100 }, { "epoch": 2.3364549173400047, "grad_norm": 0.25258854031562805, "learning_rate": 2.3897632456521672e-05, "loss": 0.0427, "step": 69110 }, { "epoch": 2.336792995030258, "grad_norm": 0.1373070627450943, "learning_rate": 2.388353116706658e-05, "loss": 0.0404, "step": 69120 }, { "epoch": 2.3371310727205112, "grad_norm": 0.226657897233963, "learning_rate": 2.3869432733669232e-05, "loss": 0.045, "step": 69130 }, { "epoch": 2.3374691504107643, "grad_norm": 0.16891872882843018, "learning_rate": 2.385533715787145e-05, "loss": 0.0458, "step": 69140 }, { "epoch": 2.337807228101018, "grad_norm": 0.24115167558193207, "learning_rate": 2.3841244441214673e-05, "loss": 0.0454, "step": 69150 }, { "epoch": 2.338145305791271, "grad_norm": 0.257889062166214, "learning_rate": 2.3827154585240073e-05, "loss": 0.0504, "step": 69160 }, { "epoch": 2.338483383481524, "grad_norm": 0.18229857087135315, "learning_rate": 2.3813067591488498e-05, "loss": 0.0435, "step": 69170 }, { "epoch": 2.3388214611717775, "grad_norm": 0.2758171558380127, "learning_rate": 2.3798983461500494e-05, "loss": 0.0444, "step": 69180 }, { "epoch": 2.3391595388620305, "grad_norm": 0.26370328664779663, "learning_rate": 2.3784902196816256e-05, "loss": 0.045, "step": 69190 }, { "epoch": 2.3394976165522836, "grad_norm": 0.34479326009750366, "learning_rate": 2.3770823798975716e-05, "loss": 0.0473, "step": 69200 }, { "epoch": 2.339835694242537, "grad_norm": 0.20632517337799072, "learning_rate": 2.3756748269518424e-05, "loss": 0.0433, "step": 69210 }, { "epoch": 2.34017377193279, "grad_norm": 0.2183917760848999, "learning_rate": 2.374267560998371e-05, "loss": 0.0496, "step": 69220 }, { "epoch": 2.3405118496230433, "grad_norm": 0.26800134778022766, "learning_rate": 2.3728605821910477e-05, "loss": 0.0478, "step": 69230 }, { "epoch": 2.3408499273132968, "grad_norm": 0.6343790292739868, "learning_rate": 2.3714538906837453e-05, "loss": 0.0449, "step": 69240 }, { "epoch": 2.34118800500355, "grad_norm": 0.17326362431049347, "learning_rate": 2.3700474866302913e-05, "loss": 0.046, "step": 69250 }, { "epoch": 2.341526082693803, "grad_norm": 0.19217537343502045, "learning_rate": 2.3686413701844917e-05, "loss": 0.0491, "step": 69260 }, { "epoch": 2.3418641603840564, "grad_norm": 0.19886380434036255, "learning_rate": 2.367235541500114e-05, "loss": 0.0453, "step": 69270 }, { "epoch": 2.3422022380743095, "grad_norm": 0.3592665195465088, "learning_rate": 2.3658300007308993e-05, "loss": 0.0481, "step": 69280 }, { "epoch": 2.3425403157645626, "grad_norm": 0.2483767867088318, "learning_rate": 2.364424748030555e-05, "loss": 0.0459, "step": 69290 }, { "epoch": 2.342878393454816, "grad_norm": 0.24569883942604065, "learning_rate": 2.3630197835527595e-05, "loss": 0.0459, "step": 69300 }, { "epoch": 2.343216471145069, "grad_norm": 0.15856827795505524, "learning_rate": 2.3616151074511546e-05, "loss": 0.0409, "step": 69310 }, { "epoch": 2.343554548835322, "grad_norm": 0.1906375288963318, "learning_rate": 2.3602107198793545e-05, "loss": 0.047, "step": 69320 }, { "epoch": 2.3438926265255757, "grad_norm": 0.2246849685907364, "learning_rate": 2.3588066209909427e-05, "loss": 0.0372, "step": 69330 }, { "epoch": 2.344230704215829, "grad_norm": 0.16246549785137177, "learning_rate": 2.3574028109394662e-05, "loss": 0.0503, "step": 69340 }, { "epoch": 2.344568781906082, "grad_norm": 0.33347174525260925, "learning_rate": 2.3559992898784454e-05, "loss": 0.0415, "step": 69350 }, { "epoch": 2.3449068595963354, "grad_norm": 0.25855445861816406, "learning_rate": 2.354596057961367e-05, "loss": 0.0481, "step": 69360 }, { "epoch": 2.3452449372865884, "grad_norm": 0.3226396143436432, "learning_rate": 2.353193115341688e-05, "loss": 0.0442, "step": 69370 }, { "epoch": 2.3455830149768415, "grad_norm": 0.2409164309501648, "learning_rate": 2.3517904621728294e-05, "loss": 0.0382, "step": 69380 }, { "epoch": 2.345921092667095, "grad_norm": 0.1952819675207138, "learning_rate": 2.3503880986081855e-05, "loss": 0.0419, "step": 69390 }, { "epoch": 2.346259170357348, "grad_norm": 0.4173488914966583, "learning_rate": 2.3489860248011136e-05, "loss": 0.0537, "step": 69400 }, { "epoch": 2.346597248047601, "grad_norm": 0.2128240317106247, "learning_rate": 2.3475842409049464e-05, "loss": 0.0421, "step": 69410 }, { "epoch": 2.3469353257378547, "grad_norm": 0.2969109117984772, "learning_rate": 2.346182747072975e-05, "loss": 0.0437, "step": 69420 }, { "epoch": 2.3472734034281078, "grad_norm": 0.4415285885334015, "learning_rate": 2.3447815434584718e-05, "loss": 0.0464, "step": 69430 }, { "epoch": 2.347611481118361, "grad_norm": 0.2922050356864929, "learning_rate": 2.3433806302146656e-05, "loss": 0.049, "step": 69440 }, { "epoch": 2.3479495588086143, "grad_norm": 0.4718913435935974, "learning_rate": 2.3419800074947612e-05, "loss": 0.0483, "step": 69450 }, { "epoch": 2.3482876364988674, "grad_norm": 0.2219076305627823, "learning_rate": 2.3405796754519248e-05, "loss": 0.0473, "step": 69460 }, { "epoch": 2.3486257141891205, "grad_norm": 0.14677578210830688, "learning_rate": 2.3391796342392992e-05, "loss": 0.0422, "step": 69470 }, { "epoch": 2.348963791879374, "grad_norm": 0.45417171716690063, "learning_rate": 2.337779884009984e-05, "loss": 0.0417, "step": 69480 }, { "epoch": 2.349301869569627, "grad_norm": 0.250332772731781, "learning_rate": 2.3363804249170624e-05, "loss": 0.0471, "step": 69490 }, { "epoch": 2.34963994725988, "grad_norm": 0.30969715118408203, "learning_rate": 2.3349812571135703e-05, "loss": 0.045, "step": 69500 }, { "epoch": 2.3499780249501336, "grad_norm": 0.20361165702342987, "learning_rate": 2.333582380752523e-05, "loss": 0.0494, "step": 69510 }, { "epoch": 2.3503161026403867, "grad_norm": 0.28587454557418823, "learning_rate": 2.3321837959868963e-05, "loss": 0.0478, "step": 69520 }, { "epoch": 2.3506541803306398, "grad_norm": 0.2712309956550598, "learning_rate": 2.3307855029696383e-05, "loss": 0.0468, "step": 69530 }, { "epoch": 2.3509922580208933, "grad_norm": 0.24534063041210175, "learning_rate": 2.3293875018536648e-05, "loss": 0.0458, "step": 69540 }, { "epoch": 2.3513303357111464, "grad_norm": 0.333372563123703, "learning_rate": 2.327989792791861e-05, "loss": 0.046, "step": 69550 }, { "epoch": 2.3516684134013994, "grad_norm": 0.2537592649459839, "learning_rate": 2.326592375937074e-05, "loss": 0.0462, "step": 69560 }, { "epoch": 2.352006491091653, "grad_norm": 0.1623113602399826, "learning_rate": 2.3251952514421248e-05, "loss": 0.0407, "step": 69570 }, { "epoch": 2.352344568781906, "grad_norm": 0.1909283995628357, "learning_rate": 2.323798419459804e-05, "loss": 0.0432, "step": 69580 }, { "epoch": 2.352682646472159, "grad_norm": 0.1777103990316391, "learning_rate": 2.3224018801428625e-05, "loss": 0.0456, "step": 69590 }, { "epoch": 2.3530207241624126, "grad_norm": 0.5003172159194946, "learning_rate": 2.3210056336440273e-05, "loss": 0.0459, "step": 69600 }, { "epoch": 2.3533588018526657, "grad_norm": 0.5241135954856873, "learning_rate": 2.3196096801159844e-05, "loss": 0.039, "step": 69610 }, { "epoch": 2.3536968795429187, "grad_norm": 0.3877508342266083, "learning_rate": 2.3182140197114012e-05, "loss": 0.0449, "step": 69620 }, { "epoch": 2.3540349572331722, "grad_norm": 0.22569073736667633, "learning_rate": 2.3168186525828985e-05, "loss": 0.0433, "step": 69630 }, { "epoch": 2.3543730349234253, "grad_norm": 0.28650107979774475, "learning_rate": 2.3154235788830757e-05, "loss": 0.0449, "step": 69640 }, { "epoch": 2.3547111126136784, "grad_norm": 0.6722433567047119, "learning_rate": 2.3140287987644927e-05, "loss": 0.0472, "step": 69650 }, { "epoch": 2.355049190303932, "grad_norm": 0.24537678062915802, "learning_rate": 2.3126343123796833e-05, "loss": 0.0373, "step": 69660 }, { "epoch": 2.355387267994185, "grad_norm": 0.1730126589536667, "learning_rate": 2.3112401198811418e-05, "loss": 0.0448, "step": 69670 }, { "epoch": 2.3557253456844385, "grad_norm": 0.28292974829673767, "learning_rate": 2.3098462214213418e-05, "loss": 0.0405, "step": 69680 }, { "epoch": 2.3560634233746915, "grad_norm": 0.20778445899486542, "learning_rate": 2.308452617152712e-05, "loss": 0.04, "step": 69690 }, { "epoch": 2.3564015010649446, "grad_norm": 0.18975256383419037, "learning_rate": 2.3070593072276603e-05, "loss": 0.0436, "step": 69700 }, { "epoch": 2.356739578755198, "grad_norm": 0.256538063287735, "learning_rate": 2.3056662917985518e-05, "loss": 0.0414, "step": 69710 }, { "epoch": 2.357077656445451, "grad_norm": 0.32587432861328125, "learning_rate": 2.3042735710177283e-05, "loss": 0.0464, "step": 69720 }, { "epoch": 2.3574157341357043, "grad_norm": 0.2859850823879242, "learning_rate": 2.302881145037492e-05, "loss": 0.0471, "step": 69730 }, { "epoch": 2.357753811825958, "grad_norm": 0.6611744165420532, "learning_rate": 2.3014890140101192e-05, "loss": 0.0446, "step": 69740 }, { "epoch": 2.358091889516211, "grad_norm": 0.14735794067382812, "learning_rate": 2.30009717808785e-05, "loss": 0.0479, "step": 69750 }, { "epoch": 2.358429967206464, "grad_norm": 0.24628157913684845, "learning_rate": 2.2987056374228967e-05, "loss": 0.0413, "step": 69760 }, { "epoch": 2.3587680448967174, "grad_norm": 0.19226032495498657, "learning_rate": 2.297314392167431e-05, "loss": 0.0461, "step": 69770 }, { "epoch": 2.3591061225869705, "grad_norm": 0.22666020691394806, "learning_rate": 2.295923442473601e-05, "loss": 0.0392, "step": 69780 }, { "epoch": 2.3594442002772236, "grad_norm": 0.28481525182724, "learning_rate": 2.294532788493519e-05, "loss": 0.0477, "step": 69790 }, { "epoch": 2.359782277967477, "grad_norm": 0.2395782619714737, "learning_rate": 2.293142430379262e-05, "loss": 0.0439, "step": 69800 }, { "epoch": 2.36012035565773, "grad_norm": 0.3760385513305664, "learning_rate": 2.291752368282879e-05, "loss": 0.0485, "step": 69810 }, { "epoch": 2.360458433347983, "grad_norm": 0.48739978671073914, "learning_rate": 2.290362602356385e-05, "loss": 0.0466, "step": 69820 }, { "epoch": 2.3607965110382367, "grad_norm": 0.2632945477962494, "learning_rate": 2.288973132751765e-05, "loss": 0.043, "step": 69830 }, { "epoch": 2.36113458872849, "grad_norm": 0.23123525083065033, "learning_rate": 2.287583959620965e-05, "loss": 0.0468, "step": 69840 }, { "epoch": 2.361472666418743, "grad_norm": 0.15856006741523743, "learning_rate": 2.2861950831159074e-05, "loss": 0.0426, "step": 69850 }, { "epoch": 2.3618107441089964, "grad_norm": 0.1578330248594284, "learning_rate": 2.284806503388472e-05, "loss": 0.0494, "step": 69860 }, { "epoch": 2.3621488217992495, "grad_norm": 0.8487514853477478, "learning_rate": 2.283418220590518e-05, "loss": 0.0422, "step": 69870 }, { "epoch": 2.3624868994895025, "grad_norm": 0.3283424377441406, "learning_rate": 2.282030234873862e-05, "loss": 0.0443, "step": 69880 }, { "epoch": 2.362824977179756, "grad_norm": 0.2069002091884613, "learning_rate": 2.2806425463902948e-05, "loss": 0.0424, "step": 69890 }, { "epoch": 2.363163054870009, "grad_norm": 0.3332176208496094, "learning_rate": 2.279255155291568e-05, "loss": 0.0478, "step": 69900 }, { "epoch": 2.363501132560262, "grad_norm": 0.2762090861797333, "learning_rate": 2.277868061729409e-05, "loss": 0.0419, "step": 69910 }, { "epoch": 2.3638392102505157, "grad_norm": 0.2831190824508667, "learning_rate": 2.276481265855504e-05, "loss": 0.0519, "step": 69920 }, { "epoch": 2.3641772879407688, "grad_norm": 0.3355293869972229, "learning_rate": 2.2750947678215128e-05, "loss": 0.0433, "step": 69930 }, { "epoch": 2.364515365631022, "grad_norm": 0.2580673098564148, "learning_rate": 2.2737085677790616e-05, "loss": 0.0465, "step": 69940 }, { "epoch": 2.3648534433212753, "grad_norm": 0.3028489351272583, "learning_rate": 2.2723226658797443e-05, "loss": 0.0482, "step": 69950 }, { "epoch": 2.3651915210115284, "grad_norm": 0.26725128293037415, "learning_rate": 2.270937062275117e-05, "loss": 0.0464, "step": 69960 }, { "epoch": 2.3655295987017815, "grad_norm": 0.38585972785949707, "learning_rate": 2.2695517571167113e-05, "loss": 0.0421, "step": 69970 }, { "epoch": 2.365867676392035, "grad_norm": 0.1651400923728943, "learning_rate": 2.268166750556019e-05, "loss": 0.0487, "step": 69980 }, { "epoch": 2.366205754082288, "grad_norm": 0.23605339229106903, "learning_rate": 2.266782042744503e-05, "loss": 0.0461, "step": 69990 }, { "epoch": 2.3665438317725416, "grad_norm": 0.2858721613883972, "learning_rate": 2.2653976338335935e-05, "loss": 0.0504, "step": 70000 }, { "epoch": 2.3668819094627946, "grad_norm": 0.22437192499637604, "learning_rate": 2.2640135239746895e-05, "loss": 0.0412, "step": 70010 }, { "epoch": 2.3672199871530477, "grad_norm": 0.22805479168891907, "learning_rate": 2.2626297133191515e-05, "loss": 0.0473, "step": 70020 }, { "epoch": 2.3675580648433012, "grad_norm": 0.19944410026073456, "learning_rate": 2.261246202018312e-05, "loss": 0.048, "step": 70030 }, { "epoch": 2.3678961425335543, "grad_norm": 0.20929178595542908, "learning_rate": 2.2598629902234718e-05, "loss": 0.0442, "step": 70040 }, { "epoch": 2.3682342202238074, "grad_norm": 0.38184818625450134, "learning_rate": 2.258480078085894e-05, "loss": 0.0402, "step": 70050 }, { "epoch": 2.368572297914061, "grad_norm": 0.5894629955291748, "learning_rate": 2.2570974657568126e-05, "loss": 0.0444, "step": 70060 }, { "epoch": 2.368910375604314, "grad_norm": 0.19752781093120575, "learning_rate": 2.2557151533874286e-05, "loss": 0.0434, "step": 70070 }, { "epoch": 2.369248453294567, "grad_norm": 0.18633025884628296, "learning_rate": 2.2543331411289116e-05, "loss": 0.0434, "step": 70080 }, { "epoch": 2.3695865309848205, "grad_norm": 0.3237267732620239, "learning_rate": 2.252951429132392e-05, "loss": 0.0442, "step": 70090 }, { "epoch": 2.3699246086750736, "grad_norm": 0.19005520641803741, "learning_rate": 2.2515700175489762e-05, "loss": 0.0393, "step": 70100 }, { "epoch": 2.3702626863653267, "grad_norm": 0.1662079393863678, "learning_rate": 2.2501889065297288e-05, "loss": 0.0468, "step": 70110 }, { "epoch": 2.37060076405558, "grad_norm": 0.3012407124042511, "learning_rate": 2.2488080962256887e-05, "loss": 0.0461, "step": 70120 }, { "epoch": 2.3709388417458332, "grad_norm": 0.22607964277267456, "learning_rate": 2.2474275867878585e-05, "loss": 0.0456, "step": 70130 }, { "epoch": 2.3712769194360863, "grad_norm": 0.28664302825927734, "learning_rate": 2.246047378367211e-05, "loss": 0.0457, "step": 70140 }, { "epoch": 2.37161499712634, "grad_norm": 0.34620851278305054, "learning_rate": 2.2446674711146798e-05, "loss": 0.045, "step": 70150 }, { "epoch": 2.371953074816593, "grad_norm": 0.2834206521511078, "learning_rate": 2.2432878651811734e-05, "loss": 0.0497, "step": 70160 }, { "epoch": 2.372291152506846, "grad_norm": 0.2196609228849411, "learning_rate": 2.2419085607175594e-05, "loss": 0.0408, "step": 70170 }, { "epoch": 2.3726292301970995, "grad_norm": 0.22916585206985474, "learning_rate": 2.240529557874678e-05, "loss": 0.0506, "step": 70180 }, { "epoch": 2.3729673078873526, "grad_norm": 0.31619691848754883, "learning_rate": 2.2391508568033358e-05, "loss": 0.0409, "step": 70190 }, { "epoch": 2.3733053855776056, "grad_norm": 0.19828730821609497, "learning_rate": 2.237772457654307e-05, "loss": 0.0422, "step": 70200 }, { "epoch": 2.373643463267859, "grad_norm": 0.28314778208732605, "learning_rate": 2.236394360578327e-05, "loss": 0.0463, "step": 70210 }, { "epoch": 2.373981540958112, "grad_norm": 0.1761840283870697, "learning_rate": 2.2350165657261073e-05, "loss": 0.0474, "step": 70220 }, { "epoch": 2.3743196186483653, "grad_norm": 0.25362348556518555, "learning_rate": 2.2336390732483164e-05, "loss": 0.0514, "step": 70230 }, { "epoch": 2.374657696338619, "grad_norm": 0.18549184501171112, "learning_rate": 2.2322618832955976e-05, "loss": 0.0447, "step": 70240 }, { "epoch": 2.374995774028872, "grad_norm": 0.3067251741886139, "learning_rate": 2.2308849960185578e-05, "loss": 0.0443, "step": 70250 }, { "epoch": 2.375333851719125, "grad_norm": 0.24560070037841797, "learning_rate": 2.229508411567774e-05, "loss": 0.0473, "step": 70260 }, { "epoch": 2.3756719294093784, "grad_norm": 0.2252061665058136, "learning_rate": 2.2281321300937823e-05, "loss": 0.0424, "step": 70270 }, { "epoch": 2.3760100070996315, "grad_norm": 0.4275560677051544, "learning_rate": 2.2267561517470937e-05, "loss": 0.0484, "step": 70280 }, { "epoch": 2.3763480847898846, "grad_norm": 0.1901884227991104, "learning_rate": 2.2253804766781845e-05, "loss": 0.0451, "step": 70290 }, { "epoch": 2.376686162480138, "grad_norm": 0.23139353096485138, "learning_rate": 2.2240051050374933e-05, "loss": 0.0403, "step": 70300 }, { "epoch": 2.377024240170391, "grad_norm": 0.27716389298439026, "learning_rate": 2.22263003697543e-05, "loss": 0.0533, "step": 70310 }, { "epoch": 2.3773623178606442, "grad_norm": 0.4974682033061981, "learning_rate": 2.2212552726423702e-05, "loss": 0.0403, "step": 70320 }, { "epoch": 2.3777003955508977, "grad_norm": 0.269062340259552, "learning_rate": 2.2198808121886578e-05, "loss": 0.05, "step": 70330 }, { "epoch": 2.378038473241151, "grad_norm": 0.309589147567749, "learning_rate": 2.218506655764598e-05, "loss": 0.0444, "step": 70340 }, { "epoch": 2.378376550931404, "grad_norm": 0.2370036393404007, "learning_rate": 2.2171328035204704e-05, "loss": 0.0433, "step": 70350 }, { "epoch": 2.3787146286216574, "grad_norm": 0.26138922572135925, "learning_rate": 2.2157592556065133e-05, "loss": 0.046, "step": 70360 }, { "epoch": 2.3790527063119105, "grad_norm": 0.36890822649002075, "learning_rate": 2.2143860121729382e-05, "loss": 0.0426, "step": 70370 }, { "epoch": 2.3793907840021635, "grad_norm": 0.4051574766635895, "learning_rate": 2.2130130733699206e-05, "loss": 0.0503, "step": 70380 }, { "epoch": 2.379728861692417, "grad_norm": 0.18410269916057587, "learning_rate": 2.2116404393476052e-05, "loss": 0.0451, "step": 70390 }, { "epoch": 2.38006693938267, "grad_norm": 0.2548163831233978, "learning_rate": 2.2102681102560967e-05, "loss": 0.045, "step": 70400 }, { "epoch": 2.380405017072923, "grad_norm": 0.2591051161289215, "learning_rate": 2.208896086245476e-05, "loss": 0.0499, "step": 70410 }, { "epoch": 2.3807430947631767, "grad_norm": 0.2840065062046051, "learning_rate": 2.2075243674657808e-05, "loss": 0.0507, "step": 70420 }, { "epoch": 2.3810811724534298, "grad_norm": 0.29542091488838196, "learning_rate": 2.206152954067024e-05, "loss": 0.0499, "step": 70430 }, { "epoch": 2.381419250143683, "grad_norm": 0.3131503760814667, "learning_rate": 2.2047818461991758e-05, "loss": 0.0466, "step": 70440 }, { "epoch": 2.3817573278339363, "grad_norm": 0.12753841280937195, "learning_rate": 2.2034110440121864e-05, "loss": 0.0383, "step": 70450 }, { "epoch": 2.3820954055241894, "grad_norm": 0.24711501598358154, "learning_rate": 2.2020405476559585e-05, "loss": 0.0411, "step": 70460 }, { "epoch": 2.3824334832144425, "grad_norm": 0.21072040498256683, "learning_rate": 2.2006703572803722e-05, "loss": 0.0506, "step": 70470 }, { "epoch": 2.382771560904696, "grad_norm": 0.26721689105033875, "learning_rate": 2.1993004730352646e-05, "loss": 0.0505, "step": 70480 }, { "epoch": 2.383109638594949, "grad_norm": 0.23072868585586548, "learning_rate": 2.1979308950704464e-05, "loss": 0.0432, "step": 70490 }, { "epoch": 2.383447716285202, "grad_norm": 0.2413851022720337, "learning_rate": 2.1965616235356923e-05, "loss": 0.0462, "step": 70500 }, { "epoch": 2.3837857939754556, "grad_norm": 0.271067351102829, "learning_rate": 2.1951926585807465e-05, "loss": 0.0444, "step": 70510 }, { "epoch": 2.3841238716657087, "grad_norm": 0.14720886945724487, "learning_rate": 2.193824000355313e-05, "loss": 0.0408, "step": 70520 }, { "epoch": 2.384461949355962, "grad_norm": 0.1618478000164032, "learning_rate": 2.192455649009067e-05, "loss": 0.0506, "step": 70530 }, { "epoch": 2.3848000270462153, "grad_norm": 0.15294788777828217, "learning_rate": 2.1910876046916523e-05, "loss": 0.0456, "step": 70540 }, { "epoch": 2.3851381047364684, "grad_norm": 0.2372964769601822, "learning_rate": 2.1897198675526725e-05, "loss": 0.0403, "step": 70550 }, { "epoch": 2.3854761824267214, "grad_norm": 0.3938552737236023, "learning_rate": 2.1883524377417024e-05, "loss": 0.0455, "step": 70560 }, { "epoch": 2.385814260116975, "grad_norm": 0.2818809449672699, "learning_rate": 2.1869853154082825e-05, "loss": 0.0461, "step": 70570 }, { "epoch": 2.386152337807228, "grad_norm": 0.21885792911052704, "learning_rate": 2.1856185007019215e-05, "loss": 0.0488, "step": 70580 }, { "epoch": 2.386490415497481, "grad_norm": 0.390112042427063, "learning_rate": 2.1842519937720874e-05, "loss": 0.0485, "step": 70590 }, { "epoch": 2.3868284931877346, "grad_norm": 0.34057778120040894, "learning_rate": 2.182885794768224e-05, "loss": 0.0446, "step": 70600 }, { "epoch": 2.3871665708779877, "grad_norm": 0.7784305214881897, "learning_rate": 2.1815199038397326e-05, "loss": 0.0447, "step": 70610 }, { "epoch": 2.3875046485682407, "grad_norm": 1.0971794128417969, "learning_rate": 2.1801543211359893e-05, "loss": 0.0451, "step": 70620 }, { "epoch": 2.3878427262584943, "grad_norm": 0.20827849209308624, "learning_rate": 2.178789046806326e-05, "loss": 0.0455, "step": 70630 }, { "epoch": 2.3881808039487473, "grad_norm": 0.4053884446620941, "learning_rate": 2.1774240810000547e-05, "loss": 0.044, "step": 70640 }, { "epoch": 2.388518881639001, "grad_norm": 0.2981721758842468, "learning_rate": 2.17605942386644e-05, "loss": 0.0502, "step": 70650 }, { "epoch": 2.388856959329254, "grad_norm": 0.26626691222190857, "learning_rate": 2.174695075554723e-05, "loss": 0.0471, "step": 70660 }, { "epoch": 2.389195037019507, "grad_norm": 0.2813066244125366, "learning_rate": 2.173331036214103e-05, "loss": 0.0499, "step": 70670 }, { "epoch": 2.3895331147097605, "grad_norm": 0.1393313705921173, "learning_rate": 2.171967305993753e-05, "loss": 0.0468, "step": 70680 }, { "epoch": 2.3898711924000136, "grad_norm": 0.2847149074077606, "learning_rate": 2.1706038850428024e-05, "loss": 0.0503, "step": 70690 }, { "epoch": 2.3902092700902666, "grad_norm": 0.25633272528648376, "learning_rate": 2.169240773510361e-05, "loss": 0.0473, "step": 70700 }, { "epoch": 2.39054734778052, "grad_norm": 0.17201659083366394, "learning_rate": 2.1678779715454906e-05, "loss": 0.0443, "step": 70710 }, { "epoch": 2.390885425470773, "grad_norm": 0.3040620684623718, "learning_rate": 2.166515479297229e-05, "loss": 0.0504, "step": 70720 }, { "epoch": 2.3912235031610263, "grad_norm": 0.16420459747314453, "learning_rate": 2.1651532969145732e-05, "loss": 0.0388, "step": 70730 }, { "epoch": 2.39156158085128, "grad_norm": 0.2460121214389801, "learning_rate": 2.1637914245464903e-05, "loss": 0.0515, "step": 70740 }, { "epoch": 2.391899658541533, "grad_norm": 0.31137216091156006, "learning_rate": 2.1624298623419133e-05, "loss": 0.0437, "step": 70750 }, { "epoch": 2.392237736231786, "grad_norm": 0.2524365484714508, "learning_rate": 2.1610686104497413e-05, "loss": 0.0433, "step": 70760 }, { "epoch": 2.3925758139220394, "grad_norm": 0.3283083736896515, "learning_rate": 2.1597076690188366e-05, "loss": 0.0445, "step": 70770 }, { "epoch": 2.3929138916122925, "grad_norm": 0.23514433205127716, "learning_rate": 2.1583470381980315e-05, "loss": 0.0471, "step": 70780 }, { "epoch": 2.3932519693025456, "grad_norm": 0.1451496034860611, "learning_rate": 2.156986718136123e-05, "loss": 0.0465, "step": 70790 }, { "epoch": 2.393590046992799, "grad_norm": 0.21279099583625793, "learning_rate": 2.155626708981871e-05, "loss": 0.0472, "step": 70800 }, { "epoch": 2.393928124683052, "grad_norm": 0.15442974865436554, "learning_rate": 2.1542670108840075e-05, "loss": 0.05, "step": 70810 }, { "epoch": 2.3942662023733052, "grad_norm": 0.3224918842315674, "learning_rate": 2.152907623991221e-05, "loss": 0.0396, "step": 70820 }, { "epoch": 2.3946042800635587, "grad_norm": 0.2243264615535736, "learning_rate": 2.1515485484521802e-05, "loss": 0.0442, "step": 70830 }, { "epoch": 2.394942357753812, "grad_norm": 0.22169454395771027, "learning_rate": 2.150189784415506e-05, "loss": 0.0452, "step": 70840 }, { "epoch": 2.395280435444065, "grad_norm": 0.19273126125335693, "learning_rate": 2.148831332029794e-05, "loss": 0.0458, "step": 70850 }, { "epoch": 2.3956185131343184, "grad_norm": 0.1783614456653595, "learning_rate": 2.147473191443599e-05, "loss": 0.0443, "step": 70860 }, { "epoch": 2.3959565908245715, "grad_norm": 0.20485518872737885, "learning_rate": 2.1461153628054492e-05, "loss": 0.0437, "step": 70870 }, { "epoch": 2.3962946685148245, "grad_norm": 0.7020265460014343, "learning_rate": 2.1447578462638284e-05, "loss": 0.0412, "step": 70880 }, { "epoch": 2.396632746205078, "grad_norm": 0.35365378856658936, "learning_rate": 2.143400641967201e-05, "loss": 0.0416, "step": 70890 }, { "epoch": 2.396970823895331, "grad_norm": 0.22836638987064362, "learning_rate": 2.1420437500639823e-05, "loss": 0.0442, "step": 70900 }, { "epoch": 2.397308901585584, "grad_norm": 0.3111383020877838, "learning_rate": 2.1406871707025646e-05, "loss": 0.0491, "step": 70910 }, { "epoch": 2.3976469792758377, "grad_norm": 0.23777322471141815, "learning_rate": 2.1393309040312963e-05, "loss": 0.0444, "step": 70920 }, { "epoch": 2.3979850569660908, "grad_norm": 0.1687544584274292, "learning_rate": 2.137974950198502e-05, "loss": 0.0425, "step": 70930 }, { "epoch": 2.398323134656344, "grad_norm": 0.17659638822078705, "learning_rate": 2.1366193093524617e-05, "loss": 0.0459, "step": 70940 }, { "epoch": 2.3986612123465973, "grad_norm": 0.3001417815685272, "learning_rate": 2.1352639816414288e-05, "loss": 0.0474, "step": 70950 }, { "epoch": 2.3989992900368504, "grad_norm": 0.5870402455329895, "learning_rate": 2.1339089672136202e-05, "loss": 0.0487, "step": 70960 }, { "epoch": 2.399337367727104, "grad_norm": 0.21806403994560242, "learning_rate": 2.1325542662172193e-05, "loss": 0.04, "step": 70970 }, { "epoch": 2.399675445417357, "grad_norm": 0.2196637988090515, "learning_rate": 2.1311998788003712e-05, "loss": 0.0418, "step": 70980 }, { "epoch": 2.40001352310761, "grad_norm": 0.505089521408081, "learning_rate": 2.1298458051111907e-05, "loss": 0.0421, "step": 70990 }, { "epoch": 2.4003516007978636, "grad_norm": 0.4568188786506653, "learning_rate": 2.1284920452977602e-05, "loss": 0.0412, "step": 71000 }, { "epoch": 2.4006896784881167, "grad_norm": 0.37663447856903076, "learning_rate": 2.12713859950812e-05, "loss": 0.0445, "step": 71010 }, { "epoch": 2.4010277561783697, "grad_norm": 0.23538793623447418, "learning_rate": 2.1257854678902832e-05, "loss": 0.0465, "step": 71020 }, { "epoch": 2.4013658338686232, "grad_norm": 0.21837885677814484, "learning_rate": 2.124432650592227e-05, "loss": 0.0431, "step": 71030 }, { "epoch": 2.4017039115588763, "grad_norm": 0.2646960914134979, "learning_rate": 2.1230801477618937e-05, "loss": 0.046, "step": 71040 }, { "epoch": 2.4020419892491294, "grad_norm": 0.16228055953979492, "learning_rate": 2.121727959547189e-05, "loss": 0.041, "step": 71050 }, { "epoch": 2.402380066939383, "grad_norm": 0.4742072820663452, "learning_rate": 2.1203760860959886e-05, "loss": 0.0511, "step": 71060 }, { "epoch": 2.402718144629636, "grad_norm": 0.18213093280792236, "learning_rate": 2.119024527556127e-05, "loss": 0.0447, "step": 71070 }, { "epoch": 2.403056222319889, "grad_norm": 0.21358907222747803, "learning_rate": 2.117673284075415e-05, "loss": 0.0453, "step": 71080 }, { "epoch": 2.4033943000101425, "grad_norm": 0.158165842294693, "learning_rate": 2.1163223558016165e-05, "loss": 0.0427, "step": 71090 }, { "epoch": 2.4037323777003956, "grad_norm": 0.17339465022087097, "learning_rate": 2.1149717428824722e-05, "loss": 0.0474, "step": 71100 }, { "epoch": 2.4040704553906487, "grad_norm": 0.2123880684375763, "learning_rate": 2.1136214454656784e-05, "loss": 0.0441, "step": 71110 }, { "epoch": 2.404408533080902, "grad_norm": 0.3382090926170349, "learning_rate": 2.1122714636989056e-05, "loss": 0.0416, "step": 71120 }, { "epoch": 2.4047466107711553, "grad_norm": 0.14701268076896667, "learning_rate": 2.1109217977297823e-05, "loss": 0.0432, "step": 71130 }, { "epoch": 2.4050846884614083, "grad_norm": 0.4255571663379669, "learning_rate": 2.1095724477059077e-05, "loss": 0.0427, "step": 71140 }, { "epoch": 2.405422766151662, "grad_norm": 0.16755923628807068, "learning_rate": 2.1082234137748434e-05, "loss": 0.0405, "step": 71150 }, { "epoch": 2.405760843841915, "grad_norm": 0.1641847789287567, "learning_rate": 2.106874696084122e-05, "loss": 0.0409, "step": 71160 }, { "epoch": 2.406098921532168, "grad_norm": 0.17814773321151733, "learning_rate": 2.1055262947812315e-05, "loss": 0.0465, "step": 71170 }, { "epoch": 2.4064369992224215, "grad_norm": 0.17240867018699646, "learning_rate": 2.104178210013636e-05, "loss": 0.0475, "step": 71180 }, { "epoch": 2.4067750769126746, "grad_norm": 0.2985071837902069, "learning_rate": 2.1028304419287554e-05, "loss": 0.0463, "step": 71190 }, { "epoch": 2.4071131546029276, "grad_norm": 0.21270249783992767, "learning_rate": 2.101482990673982e-05, "loss": 0.0431, "step": 71200 }, { "epoch": 2.407451232293181, "grad_norm": 0.16752685606479645, "learning_rate": 2.1001358563966707e-05, "loss": 0.0451, "step": 71210 }, { "epoch": 2.407789309983434, "grad_norm": 0.2364354282617569, "learning_rate": 2.098789039244144e-05, "loss": 0.0442, "step": 71220 }, { "epoch": 2.4081273876736873, "grad_norm": 0.3265824019908905, "learning_rate": 2.097442539363685e-05, "loss": 0.044, "step": 71230 }, { "epoch": 2.408465465363941, "grad_norm": 0.226030632853508, "learning_rate": 2.0960963569025448e-05, "loss": 0.0461, "step": 71240 }, { "epoch": 2.408803543054194, "grad_norm": 0.2916080951690674, "learning_rate": 2.0947504920079435e-05, "loss": 0.0491, "step": 71250 }, { "epoch": 2.409141620744447, "grad_norm": 0.26010045409202576, "learning_rate": 2.093404944827059e-05, "loss": 0.0393, "step": 71260 }, { "epoch": 2.4094796984347004, "grad_norm": 0.5056188106536865, "learning_rate": 2.0920597155070394e-05, "loss": 0.0443, "step": 71270 }, { "epoch": 2.4098177761249535, "grad_norm": 0.24498428404331207, "learning_rate": 2.090714804194997e-05, "loss": 0.0422, "step": 71280 }, { "epoch": 2.4101558538152066, "grad_norm": 0.16473253071308136, "learning_rate": 2.0893702110380114e-05, "loss": 0.0481, "step": 71290 }, { "epoch": 2.41049393150546, "grad_norm": 0.4666130840778351, "learning_rate": 2.0880259361831216e-05, "loss": 0.0465, "step": 71300 }, { "epoch": 2.410832009195713, "grad_norm": 0.3881249725818634, "learning_rate": 2.0866819797773397e-05, "loss": 0.0476, "step": 71310 }, { "epoch": 2.4111700868859662, "grad_norm": 0.39714547991752625, "learning_rate": 2.085338341967634e-05, "loss": 0.0452, "step": 71320 }, { "epoch": 2.4115081645762197, "grad_norm": 0.3285658359527588, "learning_rate": 2.083995022900946e-05, "loss": 0.0419, "step": 71330 }, { "epoch": 2.411846242266473, "grad_norm": 0.35045087337493896, "learning_rate": 2.0826520227241768e-05, "loss": 0.0458, "step": 71340 }, { "epoch": 2.412184319956726, "grad_norm": 0.2948610782623291, "learning_rate": 2.081309341584199e-05, "loss": 0.0479, "step": 71350 }, { "epoch": 2.4125223976469794, "grad_norm": 0.19455483555793762, "learning_rate": 2.0799669796278413e-05, "loss": 0.0456, "step": 71360 }, { "epoch": 2.4128604753372325, "grad_norm": 0.32420188188552856, "learning_rate": 2.078624937001906e-05, "loss": 0.0484, "step": 71370 }, { "epoch": 2.4131985530274855, "grad_norm": 0.20572781562805176, "learning_rate": 2.0772832138531544e-05, "loss": 0.0497, "step": 71380 }, { "epoch": 2.413536630717739, "grad_norm": 0.29148194193840027, "learning_rate": 2.0759418103283158e-05, "loss": 0.0424, "step": 71390 }, { "epoch": 2.413874708407992, "grad_norm": 0.2541332244873047, "learning_rate": 2.0746007265740846e-05, "loss": 0.0479, "step": 71400 }, { "epoch": 2.414212786098245, "grad_norm": 0.3303065598011017, "learning_rate": 2.0732599627371215e-05, "loss": 0.0456, "step": 71410 }, { "epoch": 2.4145508637884987, "grad_norm": 0.1798112690448761, "learning_rate": 2.0719195189640468e-05, "loss": 0.0413, "step": 71420 }, { "epoch": 2.4148889414787518, "grad_norm": 1.4435843229293823, "learning_rate": 2.0705793954014528e-05, "loss": 0.0414, "step": 71430 }, { "epoch": 2.415227019169005, "grad_norm": 0.22668394446372986, "learning_rate": 2.06923959219589e-05, "loss": 0.0442, "step": 71440 }, { "epoch": 2.4155650968592584, "grad_norm": 0.1535138040781021, "learning_rate": 2.0679001094938793e-05, "loss": 0.048, "step": 71450 }, { "epoch": 2.4159031745495114, "grad_norm": 0.2935989499092102, "learning_rate": 2.066560947441904e-05, "loss": 0.0479, "step": 71460 }, { "epoch": 2.4162412522397645, "grad_norm": 0.2087496817111969, "learning_rate": 2.0652221061864152e-05, "loss": 0.0529, "step": 71470 }, { "epoch": 2.416579329930018, "grad_norm": 0.2574697434902191, "learning_rate": 2.0638835858738226e-05, "loss": 0.044, "step": 71480 }, { "epoch": 2.416917407620271, "grad_norm": 0.18746817111968994, "learning_rate": 2.0625453866505068e-05, "loss": 0.0446, "step": 71490 }, { "epoch": 2.417255485310524, "grad_norm": 0.27281588315963745, "learning_rate": 2.0612075086628123e-05, "loss": 0.0434, "step": 71500 }, { "epoch": 2.4175935630007777, "grad_norm": 0.24875959753990173, "learning_rate": 2.059869952057045e-05, "loss": 0.0472, "step": 71510 }, { "epoch": 2.4179316406910307, "grad_norm": 0.5197808742523193, "learning_rate": 2.0585327169794793e-05, "loss": 0.042, "step": 71520 }, { "epoch": 2.418269718381284, "grad_norm": 0.2702605724334717, "learning_rate": 2.0571958035763534e-05, "loss": 0.0467, "step": 71530 }, { "epoch": 2.4186077960715373, "grad_norm": 1.5145081281661987, "learning_rate": 2.0558592119938715e-05, "loss": 0.0446, "step": 71540 }, { "epoch": 2.4189458737617904, "grad_norm": 0.23908229172229767, "learning_rate": 2.054522942378198e-05, "loss": 0.0408, "step": 71550 }, { "epoch": 2.4192839514520434, "grad_norm": 0.21703416109085083, "learning_rate": 2.053186994875469e-05, "loss": 0.0437, "step": 71560 }, { "epoch": 2.419622029142297, "grad_norm": 0.2993945777416229, "learning_rate": 2.0518513696317775e-05, "loss": 0.0451, "step": 71570 }, { "epoch": 2.41996010683255, "grad_norm": 0.25314685702323914, "learning_rate": 2.050516066793188e-05, "loss": 0.0507, "step": 71580 }, { "epoch": 2.420298184522803, "grad_norm": 0.18224310874938965, "learning_rate": 2.0491810865057276e-05, "loss": 0.047, "step": 71590 }, { "epoch": 2.4206362622130566, "grad_norm": 0.27343568205833435, "learning_rate": 2.0478464289153882e-05, "loss": 0.0545, "step": 71600 }, { "epoch": 2.4209743399033097, "grad_norm": 0.23742122948169708, "learning_rate": 2.0465120941681232e-05, "loss": 0.0451, "step": 71610 }, { "epoch": 2.421312417593563, "grad_norm": 0.2954413592815399, "learning_rate": 2.0451780824098566e-05, "loss": 0.0446, "step": 71620 }, { "epoch": 2.4216504952838163, "grad_norm": 0.19412527978420258, "learning_rate": 2.043844393786471e-05, "loss": 0.0422, "step": 71630 }, { "epoch": 2.4219885729740693, "grad_norm": 0.6013870239257812, "learning_rate": 2.0425110284438193e-05, "loss": 0.0442, "step": 71640 }, { "epoch": 2.422326650664323, "grad_norm": 0.2033672332763672, "learning_rate": 2.041177986527712e-05, "loss": 0.0414, "step": 71650 }, { "epoch": 2.422664728354576, "grad_norm": 0.33499443531036377, "learning_rate": 2.0398452681839347e-05, "loss": 0.0494, "step": 71660 }, { "epoch": 2.423002806044829, "grad_norm": 0.1848234087228775, "learning_rate": 2.0385128735582272e-05, "loss": 0.0417, "step": 71670 }, { "epoch": 2.4233408837350825, "grad_norm": 0.2807208001613617, "learning_rate": 2.0371808027963003e-05, "loss": 0.0482, "step": 71680 }, { "epoch": 2.4236789614253356, "grad_norm": 0.17332395911216736, "learning_rate": 2.035849056043825e-05, "loss": 0.0412, "step": 71690 }, { "epoch": 2.4240170391155886, "grad_norm": 0.17574460804462433, "learning_rate": 2.03451763344644e-05, "loss": 0.0413, "step": 71700 }, { "epoch": 2.424355116805842, "grad_norm": 0.27600035071372986, "learning_rate": 2.0331865351497482e-05, "loss": 0.0421, "step": 71710 }, { "epoch": 2.424693194496095, "grad_norm": 0.3268083333969116, "learning_rate": 2.0318557612993185e-05, "loss": 0.0448, "step": 71720 }, { "epoch": 2.4250312721863483, "grad_norm": 0.2313673496246338, "learning_rate": 2.0305253120406782e-05, "loss": 0.0447, "step": 71730 }, { "epoch": 2.425369349876602, "grad_norm": 0.2086864411830902, "learning_rate": 2.029195187519325e-05, "loss": 0.0402, "step": 71740 }, { "epoch": 2.425707427566855, "grad_norm": 0.3830759823322296, "learning_rate": 2.0278653878807217e-05, "loss": 0.0426, "step": 71750 }, { "epoch": 2.426045505257108, "grad_norm": 0.3339085280895233, "learning_rate": 2.0265359132702892e-05, "loss": 0.0476, "step": 71760 }, { "epoch": 2.4263835829473615, "grad_norm": 0.17336252331733704, "learning_rate": 2.0252067638334182e-05, "loss": 0.0434, "step": 71770 }, { "epoch": 2.4267216606376145, "grad_norm": 0.1837518960237503, "learning_rate": 2.023877939715464e-05, "loss": 0.0504, "step": 71780 }, { "epoch": 2.4270597383278676, "grad_norm": 0.24863755702972412, "learning_rate": 2.0225494410617448e-05, "loss": 0.0417, "step": 71790 }, { "epoch": 2.427397816018121, "grad_norm": 0.5706419348716736, "learning_rate": 2.02122126801754e-05, "loss": 0.046, "step": 71800 }, { "epoch": 2.427735893708374, "grad_norm": 0.23804275691509247, "learning_rate": 2.0198934207281018e-05, "loss": 0.0419, "step": 71810 }, { "epoch": 2.4280739713986272, "grad_norm": 0.2122502326965332, "learning_rate": 2.0185658993386357e-05, "loss": 0.0481, "step": 71820 }, { "epoch": 2.4284120490888808, "grad_norm": 0.21859866380691528, "learning_rate": 2.0172387039943226e-05, "loss": 0.0466, "step": 71830 }, { "epoch": 2.428750126779134, "grad_norm": 0.19044196605682373, "learning_rate": 2.0159118348402968e-05, "loss": 0.0445, "step": 71840 }, { "epoch": 2.429088204469387, "grad_norm": 0.33453330397605896, "learning_rate": 2.0145852920216697e-05, "loss": 0.0515, "step": 71850 }, { "epoch": 2.4294262821596404, "grad_norm": 0.3331321179866791, "learning_rate": 2.0132590756835045e-05, "loss": 0.0474, "step": 71860 }, { "epoch": 2.4297643598498935, "grad_norm": 0.19942308962345123, "learning_rate": 2.0119331859708386e-05, "loss": 0.0427, "step": 71870 }, { "epoch": 2.4301024375401465, "grad_norm": 0.1813281774520874, "learning_rate": 2.0106076230286654e-05, "loss": 0.0419, "step": 71880 }, { "epoch": 2.4304405152304, "grad_norm": 0.18543553352355957, "learning_rate": 2.0092823870019488e-05, "loss": 0.0454, "step": 71890 }, { "epoch": 2.430778592920653, "grad_norm": 0.2295648604631424, "learning_rate": 2.0079574780356115e-05, "loss": 0.0439, "step": 71900 }, { "epoch": 2.431116670610906, "grad_norm": 0.4689696431159973, "learning_rate": 2.0066328962745486e-05, "loss": 0.0399, "step": 71910 }, { "epoch": 2.4314547483011597, "grad_norm": 0.19099672138690948, "learning_rate": 2.00530864186361e-05, "loss": 0.0441, "step": 71920 }, { "epoch": 2.4317928259914128, "grad_norm": 0.2339724451303482, "learning_rate": 2.003984714947618e-05, "loss": 0.046, "step": 71930 }, { "epoch": 2.4321309036816663, "grad_norm": 0.2733651399612427, "learning_rate": 2.0026611156713515e-05, "loss": 0.044, "step": 71940 }, { "epoch": 2.4324689813719194, "grad_norm": 0.2037072777748108, "learning_rate": 2.0013378441795584e-05, "loss": 0.0466, "step": 71950 }, { "epoch": 2.4328070590621724, "grad_norm": 0.19580601155757904, "learning_rate": 2.0000149006169506e-05, "loss": 0.0434, "step": 71960 }, { "epoch": 2.433145136752426, "grad_norm": 0.2930159568786621, "learning_rate": 1.9986922851282048e-05, "loss": 0.0436, "step": 71970 }, { "epoch": 2.433483214442679, "grad_norm": 0.22640059888362885, "learning_rate": 1.997369997857956e-05, "loss": 0.0484, "step": 71980 }, { "epoch": 2.433821292132932, "grad_norm": 0.16749489307403564, "learning_rate": 1.9960480389508103e-05, "loss": 0.0472, "step": 71990 }, { "epoch": 2.4341593698231856, "grad_norm": 0.21939101815223694, "learning_rate": 1.9947264085513363e-05, "loss": 0.0486, "step": 72000 }, { "epoch": 2.4344974475134387, "grad_norm": 0.29564106464385986, "learning_rate": 1.993405106804062e-05, "loss": 0.0436, "step": 72010 }, { "epoch": 2.4348355252036917, "grad_norm": 0.3070595860481262, "learning_rate": 1.9920841338534867e-05, "loss": 0.0427, "step": 72020 }, { "epoch": 2.4351736028939452, "grad_norm": 0.2590324878692627, "learning_rate": 1.9907634898440646e-05, "loss": 0.0439, "step": 72030 }, { "epoch": 2.4355116805841983, "grad_norm": 0.34381377696990967, "learning_rate": 1.989443174920227e-05, "loss": 0.043, "step": 72040 }, { "epoch": 2.4358497582744514, "grad_norm": 0.19045065343379974, "learning_rate": 1.988123189226355e-05, "loss": 0.049, "step": 72050 }, { "epoch": 2.436187835964705, "grad_norm": 0.3112999200820923, "learning_rate": 1.986803532906805e-05, "loss": 0.0464, "step": 72060 }, { "epoch": 2.436525913654958, "grad_norm": 0.17115630209445953, "learning_rate": 1.985484206105889e-05, "loss": 0.0392, "step": 72070 }, { "epoch": 2.436863991345211, "grad_norm": 0.18638946115970612, "learning_rate": 1.9841652089678887e-05, "loss": 0.0471, "step": 72080 }, { "epoch": 2.4372020690354645, "grad_norm": 0.24095533788204193, "learning_rate": 1.9828465416370435e-05, "loss": 0.0424, "step": 72090 }, { "epoch": 2.4375401467257176, "grad_norm": 0.21121521294116974, "learning_rate": 1.9815282042575685e-05, "loss": 0.04, "step": 72100 }, { "epoch": 2.4378782244159707, "grad_norm": 0.14823828637599945, "learning_rate": 1.9802101969736285e-05, "loss": 0.0482, "step": 72110 }, { "epoch": 2.438216302106224, "grad_norm": 0.4104308485984802, "learning_rate": 1.9788925199293633e-05, "loss": 0.0471, "step": 72120 }, { "epoch": 2.4385543797964773, "grad_norm": 0.3349996507167816, "learning_rate": 1.977575173268868e-05, "loss": 0.0429, "step": 72130 }, { "epoch": 2.4388924574867303, "grad_norm": 0.14953851699829102, "learning_rate": 1.9762581571362094e-05, "loss": 0.0412, "step": 72140 }, { "epoch": 2.439230535176984, "grad_norm": 0.4359360337257385, "learning_rate": 1.9749414716754112e-05, "loss": 0.0479, "step": 72150 }, { "epoch": 2.439568612867237, "grad_norm": 0.2523094117641449, "learning_rate": 1.973625117030465e-05, "loss": 0.0413, "step": 72160 }, { "epoch": 2.43990669055749, "grad_norm": 0.22422918677330017, "learning_rate": 1.972309093345326e-05, "loss": 0.0445, "step": 72170 }, { "epoch": 2.4402447682477435, "grad_norm": 0.3848881721496582, "learning_rate": 1.970993400763915e-05, "loss": 0.0427, "step": 72180 }, { "epoch": 2.4405828459379966, "grad_norm": 0.6305657625198364, "learning_rate": 1.969678039430109e-05, "loss": 0.0469, "step": 72190 }, { "epoch": 2.4409209236282496, "grad_norm": 0.20985375344753265, "learning_rate": 1.9683630094877564e-05, "loss": 0.0471, "step": 72200 }, { "epoch": 2.441259001318503, "grad_norm": 0.333485871553421, "learning_rate": 1.9670483110806696e-05, "loss": 0.0461, "step": 72210 }, { "epoch": 2.441597079008756, "grad_norm": 0.3639271557331085, "learning_rate": 1.9657339443526172e-05, "loss": 0.0435, "step": 72220 }, { "epoch": 2.4419351566990093, "grad_norm": 0.26557016372680664, "learning_rate": 1.964419909447339e-05, "loss": 0.0438, "step": 72230 }, { "epoch": 2.442273234389263, "grad_norm": 0.35772520303726196, "learning_rate": 1.963106206508536e-05, "loss": 0.0498, "step": 72240 }, { "epoch": 2.442611312079516, "grad_norm": 0.2920409142971039, "learning_rate": 1.9617928356798738e-05, "loss": 0.0483, "step": 72250 }, { "epoch": 2.442949389769769, "grad_norm": 0.4030861556529999, "learning_rate": 1.960479797104977e-05, "loss": 0.047, "step": 72260 }, { "epoch": 2.4432874674600225, "grad_norm": 0.25051939487457275, "learning_rate": 1.9591670909274418e-05, "loss": 0.0447, "step": 72270 }, { "epoch": 2.4436255451502755, "grad_norm": 0.2355535924434662, "learning_rate": 1.9578547172908184e-05, "loss": 0.0447, "step": 72280 }, { "epoch": 2.4439636228405286, "grad_norm": 0.23698292672634125, "learning_rate": 1.9565426763386325e-05, "loss": 0.0411, "step": 72290 }, { "epoch": 2.444301700530782, "grad_norm": 0.32231783866882324, "learning_rate": 1.9552309682143622e-05, "loss": 0.0482, "step": 72300 }, { "epoch": 2.444639778221035, "grad_norm": 0.20895716547966003, "learning_rate": 1.9539195930614574e-05, "loss": 0.0454, "step": 72310 }, { "epoch": 2.4449778559112882, "grad_norm": 0.1750454604625702, "learning_rate": 1.952608551023324e-05, "loss": 0.0485, "step": 72320 }, { "epoch": 2.4453159336015418, "grad_norm": 0.29341718554496765, "learning_rate": 1.9512978422433397e-05, "loss": 0.0467, "step": 72330 }, { "epoch": 2.445654011291795, "grad_norm": 0.1971593201160431, "learning_rate": 1.9499874668648376e-05, "loss": 0.0442, "step": 72340 }, { "epoch": 2.445992088982048, "grad_norm": 0.15773646533489227, "learning_rate": 1.9486774250311213e-05, "loss": 0.0428, "step": 72350 }, { "epoch": 2.4463301666723014, "grad_norm": 0.2507302165031433, "learning_rate": 1.9473677168854537e-05, "loss": 0.0443, "step": 72360 }, { "epoch": 2.4466682443625545, "grad_norm": 0.45029500126838684, "learning_rate": 1.946058342571065e-05, "loss": 0.0453, "step": 72370 }, { "epoch": 2.4470063220528075, "grad_norm": 0.46965327858924866, "learning_rate": 1.9447493022311424e-05, "loss": 0.0491, "step": 72380 }, { "epoch": 2.447344399743061, "grad_norm": 0.2252815067768097, "learning_rate": 1.9434405960088442e-05, "loss": 0.0422, "step": 72390 }, { "epoch": 2.447682477433314, "grad_norm": 0.17883171141147614, "learning_rate": 1.9421322240472857e-05, "loss": 0.0438, "step": 72400 }, { "epoch": 2.448020555123567, "grad_norm": 0.23173128068447113, "learning_rate": 1.9408241864895494e-05, "loss": 0.0404, "step": 72410 }, { "epoch": 2.4483586328138207, "grad_norm": 0.401210218667984, "learning_rate": 1.939516483478681e-05, "loss": 0.0486, "step": 72420 }, { "epoch": 2.448696710504074, "grad_norm": 0.26922598481178284, "learning_rate": 1.9382091151576902e-05, "loss": 0.0427, "step": 72430 }, { "epoch": 2.449034788194327, "grad_norm": 0.40352216362953186, "learning_rate": 1.9369020816695454e-05, "loss": 0.0455, "step": 72440 }, { "epoch": 2.4493728658845804, "grad_norm": 0.18278303742408752, "learning_rate": 1.935595383157184e-05, "loss": 0.0462, "step": 72450 }, { "epoch": 2.4497109435748334, "grad_norm": 0.17298032343387604, "learning_rate": 1.9342890197635057e-05, "loss": 0.0469, "step": 72460 }, { "epoch": 2.4500490212650865, "grad_norm": 0.24651363492012024, "learning_rate": 1.9329829916313685e-05, "loss": 0.0443, "step": 72470 }, { "epoch": 2.45038709895534, "grad_norm": 0.19194678962230682, "learning_rate": 1.9316772989036013e-05, "loss": 0.0407, "step": 72480 }, { "epoch": 2.450725176645593, "grad_norm": 0.3108009696006775, "learning_rate": 1.930371941722991e-05, "loss": 0.0451, "step": 72490 }, { "epoch": 2.451063254335846, "grad_norm": 0.18031732738018036, "learning_rate": 1.929066920232292e-05, "loss": 0.0448, "step": 72500 }, { "epoch": 2.4514013320260997, "grad_norm": 0.24281039834022522, "learning_rate": 1.9277622345742152e-05, "loss": 0.0449, "step": 72510 }, { "epoch": 2.4517394097163527, "grad_norm": 0.20557598769664764, "learning_rate": 1.926457884891444e-05, "loss": 0.0447, "step": 72520 }, { "epoch": 2.452077487406606, "grad_norm": 0.35651296377182007, "learning_rate": 1.9251538713266147e-05, "loss": 0.0455, "step": 72530 }, { "epoch": 2.4524155650968593, "grad_norm": 0.21364618837833405, "learning_rate": 1.9238501940223358e-05, "loss": 0.0459, "step": 72540 }, { "epoch": 2.4527536427871124, "grad_norm": 0.4012526273727417, "learning_rate": 1.922546853121175e-05, "loss": 0.0465, "step": 72550 }, { "epoch": 2.4530917204773655, "grad_norm": 0.2878887355327606, "learning_rate": 1.9212438487656647e-05, "loss": 0.0446, "step": 72560 }, { "epoch": 2.453429798167619, "grad_norm": 0.2614641487598419, "learning_rate": 1.9199411810982965e-05, "loss": 0.0466, "step": 72570 }, { "epoch": 2.453767875857872, "grad_norm": 0.16805624961853027, "learning_rate": 1.918638850261532e-05, "loss": 0.0383, "step": 72580 }, { "epoch": 2.4541059535481256, "grad_norm": 0.22177769243717194, "learning_rate": 1.9173368563977885e-05, "loss": 0.0467, "step": 72590 }, { "epoch": 2.4544440312383786, "grad_norm": 0.23849360644817352, "learning_rate": 1.9160351996494513e-05, "loss": 0.0458, "step": 72600 }, { "epoch": 2.4547821089286317, "grad_norm": 0.32198232412338257, "learning_rate": 1.9147338801588677e-05, "loss": 0.044, "step": 72610 }, { "epoch": 2.455120186618885, "grad_norm": 0.9999159574508667, "learning_rate": 1.913432898068351e-05, "loss": 0.0479, "step": 72620 }, { "epoch": 2.4554582643091383, "grad_norm": 0.26764458417892456, "learning_rate": 1.912132253520169e-05, "loss": 0.0448, "step": 72630 }, { "epoch": 2.4557963419993913, "grad_norm": 0.2510094940662384, "learning_rate": 1.910831946656564e-05, "loss": 0.0469, "step": 72640 }, { "epoch": 2.456134419689645, "grad_norm": 0.25228753685951233, "learning_rate": 1.909531977619731e-05, "loss": 0.0437, "step": 72650 }, { "epoch": 2.456472497379898, "grad_norm": 0.2899746298789978, "learning_rate": 1.908232346551834e-05, "loss": 0.0436, "step": 72660 }, { "epoch": 2.456810575070151, "grad_norm": 0.22529032826423645, "learning_rate": 1.9069330535949993e-05, "loss": 0.0449, "step": 72670 }, { "epoch": 2.4571486527604045, "grad_norm": 0.2290017157793045, "learning_rate": 1.9056340988913167e-05, "loss": 0.0441, "step": 72680 }, { "epoch": 2.4574867304506576, "grad_norm": 0.27901753783226013, "learning_rate": 1.9043354825828357e-05, "loss": 0.0476, "step": 72690 }, { "epoch": 2.4578248081409106, "grad_norm": 0.2475261241197586, "learning_rate": 1.9030372048115724e-05, "loss": 0.0482, "step": 72700 }, { "epoch": 2.458162885831164, "grad_norm": 0.18777279555797577, "learning_rate": 1.9017392657195056e-05, "loss": 0.0381, "step": 72710 }, { "epoch": 2.4585009635214172, "grad_norm": 0.404863178730011, "learning_rate": 1.9004416654485724e-05, "loss": 0.0487, "step": 72720 }, { "epoch": 2.4588390412116703, "grad_norm": 0.16318558156490326, "learning_rate": 1.8991444041406785e-05, "loss": 0.0418, "step": 72730 }, { "epoch": 2.459177118901924, "grad_norm": 0.5204675197601318, "learning_rate": 1.8978474819376912e-05, "loss": 0.0473, "step": 72740 }, { "epoch": 2.459515196592177, "grad_norm": 0.239881694316864, "learning_rate": 1.8965508989814406e-05, "loss": 0.0453, "step": 72750 }, { "epoch": 2.45985327428243, "grad_norm": 0.35720404982566833, "learning_rate": 1.8952546554137158e-05, "loss": 0.0458, "step": 72760 }, { "epoch": 2.4601913519726835, "grad_norm": 0.3405437469482422, "learning_rate": 1.8939587513762757e-05, "loss": 0.0363, "step": 72770 }, { "epoch": 2.4605294296629365, "grad_norm": 0.1541287750005722, "learning_rate": 1.892663187010835e-05, "loss": 0.0493, "step": 72780 }, { "epoch": 2.4608675073531896, "grad_norm": 0.3178357779979706, "learning_rate": 1.891367962459077e-05, "loss": 0.0451, "step": 72790 }, { "epoch": 2.461205585043443, "grad_norm": 0.303072452545166, "learning_rate": 1.890073077862644e-05, "loss": 0.044, "step": 72800 }, { "epoch": 2.461543662733696, "grad_norm": 0.21965467929840088, "learning_rate": 1.888778533363146e-05, "loss": 0.0427, "step": 72810 }, { "epoch": 2.4618817404239492, "grad_norm": 0.3947591483592987, "learning_rate": 1.8874843291021478e-05, "loss": 0.0417, "step": 72820 }, { "epoch": 2.4622198181142028, "grad_norm": 0.3374454975128174, "learning_rate": 1.8861904652211853e-05, "loss": 0.0443, "step": 72830 }, { "epoch": 2.462557895804456, "grad_norm": 0.35741791129112244, "learning_rate": 1.8848969418617507e-05, "loss": 0.0454, "step": 72840 }, { "epoch": 2.462895973494709, "grad_norm": 0.2886112630367279, "learning_rate": 1.8836037591653044e-05, "loss": 0.0422, "step": 72850 }, { "epoch": 2.4632340511849624, "grad_norm": 0.2770111560821533, "learning_rate": 1.882310917273262e-05, "loss": 0.046, "step": 72860 }, { "epoch": 2.4635721288752155, "grad_norm": 0.24483510851860046, "learning_rate": 1.881018416327014e-05, "loss": 0.0422, "step": 72870 }, { "epoch": 2.4639102065654686, "grad_norm": 0.2102653831243515, "learning_rate": 1.8797262564679014e-05, "loss": 0.045, "step": 72880 }, { "epoch": 2.464248284255722, "grad_norm": 0.28260618448257446, "learning_rate": 1.8784344378372355e-05, "loss": 0.0376, "step": 72890 }, { "epoch": 2.464586361945975, "grad_norm": 0.2105375975370407, "learning_rate": 1.8771429605762843e-05, "loss": 0.0461, "step": 72900 }, { "epoch": 2.4649244396362286, "grad_norm": 0.5246991515159607, "learning_rate": 1.8758518248262835e-05, "loss": 0.0451, "step": 72910 }, { "epoch": 2.4652625173264817, "grad_norm": 0.4567252993583679, "learning_rate": 1.8745610307284305e-05, "loss": 0.0399, "step": 72920 }, { "epoch": 2.465600595016735, "grad_norm": 0.34868693351745605, "learning_rate": 1.8732705784238862e-05, "loss": 0.0435, "step": 72930 }, { "epoch": 2.4659386727069883, "grad_norm": 0.2004821002483368, "learning_rate": 1.8719804680537682e-05, "loss": 0.0431, "step": 72940 }, { "epoch": 2.4662767503972414, "grad_norm": 0.1612299531698227, "learning_rate": 1.870690699759164e-05, "loss": 0.0453, "step": 72950 }, { "epoch": 2.4666148280874944, "grad_norm": 0.3148420751094818, "learning_rate": 1.8694012736811213e-05, "loss": 0.042, "step": 72960 }, { "epoch": 2.466952905777748, "grad_norm": 0.2035340815782547, "learning_rate": 1.8681121899606463e-05, "loss": 0.0475, "step": 72970 }, { "epoch": 2.467290983468001, "grad_norm": 0.44639140367507935, "learning_rate": 1.8668234487387144e-05, "loss": 0.0469, "step": 72980 }, { "epoch": 2.467629061158254, "grad_norm": 0.5505692958831787, "learning_rate": 1.8655350501562592e-05, "loss": 0.0457, "step": 72990 }, { "epoch": 2.4679671388485076, "grad_norm": 0.3107708990573883, "learning_rate": 1.8642469943541796e-05, "loss": 0.0477, "step": 73000 }, { "epoch": 2.4683052165387607, "grad_norm": 0.20000606775283813, "learning_rate": 1.8629592814733328e-05, "loss": 0.0455, "step": 73010 }, { "epoch": 2.4686432942290137, "grad_norm": 0.350417822599411, "learning_rate": 1.8616719116545438e-05, "loss": 0.0459, "step": 73020 }, { "epoch": 2.4689813719192673, "grad_norm": 0.19981101155281067, "learning_rate": 1.8603848850385952e-05, "loss": 0.0466, "step": 73030 }, { "epoch": 2.4693194496095203, "grad_norm": 0.2974056005477905, "learning_rate": 1.859098201766236e-05, "loss": 0.0441, "step": 73040 }, { "epoch": 2.4696575272997734, "grad_norm": 0.217257559299469, "learning_rate": 1.8578118619781726e-05, "loss": 0.0433, "step": 73050 }, { "epoch": 2.469995604990027, "grad_norm": 0.17414921522140503, "learning_rate": 1.856525865815083e-05, "loss": 0.0402, "step": 73060 }, { "epoch": 2.47033368268028, "grad_norm": 0.19256560504436493, "learning_rate": 1.8552402134175966e-05, "loss": 0.0427, "step": 73070 }, { "epoch": 2.470671760370533, "grad_norm": 0.23392917215824127, "learning_rate": 1.8539549049263143e-05, "loss": 0.0406, "step": 73080 }, { "epoch": 2.4710098380607866, "grad_norm": 0.29396378993988037, "learning_rate": 1.8526699404817917e-05, "loss": 0.0469, "step": 73090 }, { "epoch": 2.4713479157510396, "grad_norm": 0.2847110629081726, "learning_rate": 1.8513853202245546e-05, "loss": 0.0471, "step": 73100 }, { "epoch": 2.4716859934412927, "grad_norm": 0.13105010986328125, "learning_rate": 1.850101044295081e-05, "loss": 0.045, "step": 73110 }, { "epoch": 2.472024071131546, "grad_norm": 0.29095402359962463, "learning_rate": 1.848817112833825e-05, "loss": 0.0501, "step": 73120 }, { "epoch": 2.4723621488217993, "grad_norm": 0.19983942806720734, "learning_rate": 1.8475335259811905e-05, "loss": 0.0449, "step": 73130 }, { "epoch": 2.4727002265120523, "grad_norm": 0.18241317570209503, "learning_rate": 1.8462502838775508e-05, "loss": 0.0471, "step": 73140 }, { "epoch": 2.473038304202306, "grad_norm": 0.2691495418548584, "learning_rate": 1.8449673866632378e-05, "loss": 0.0459, "step": 73150 }, { "epoch": 2.473376381892559, "grad_norm": 0.21840286254882812, "learning_rate": 1.8436848344785473e-05, "loss": 0.0481, "step": 73160 }, { "epoch": 2.473714459582812, "grad_norm": 0.2288208156824112, "learning_rate": 1.8424026274637378e-05, "loss": 0.0484, "step": 73170 }, { "epoch": 2.4740525372730655, "grad_norm": 0.23607145249843597, "learning_rate": 1.8411207657590312e-05, "loss": 0.0482, "step": 73180 }, { "epoch": 2.4743906149633186, "grad_norm": 0.26512444019317627, "learning_rate": 1.839839249504607e-05, "loss": 0.0408, "step": 73190 }, { "epoch": 2.4747286926535716, "grad_norm": 0.2849084436893463, "learning_rate": 1.8385580788406108e-05, "loss": 0.0461, "step": 73200 }, { "epoch": 2.475066770343825, "grad_norm": 0.2270137369632721, "learning_rate": 1.837277253907152e-05, "loss": 0.0438, "step": 73210 }, { "epoch": 2.4754048480340782, "grad_norm": 0.23683394491672516, "learning_rate": 1.8359967748442957e-05, "loss": 0.0505, "step": 73220 }, { "epoch": 2.4757429257243313, "grad_norm": 0.21814526617527008, "learning_rate": 1.8347166417920774e-05, "loss": 0.046, "step": 73230 }, { "epoch": 2.476081003414585, "grad_norm": 0.17215335369110107, "learning_rate": 1.833436854890484e-05, "loss": 0.0405, "step": 73240 }, { "epoch": 2.476419081104838, "grad_norm": 0.2502337694168091, "learning_rate": 1.8321574142794796e-05, "loss": 0.0405, "step": 73250 }, { "epoch": 2.476757158795091, "grad_norm": 0.3383854031562805, "learning_rate": 1.8308783200989748e-05, "loss": 0.0402, "step": 73260 }, { "epoch": 2.4770952364853445, "grad_norm": 0.3571563959121704, "learning_rate": 1.829599572488855e-05, "loss": 0.051, "step": 73270 }, { "epoch": 2.4774333141755975, "grad_norm": 0.36340638995170593, "learning_rate": 1.828321171588957e-05, "loss": 0.048, "step": 73280 }, { "epoch": 2.4777713918658506, "grad_norm": 0.2788114547729492, "learning_rate": 1.827043117539089e-05, "loss": 0.0475, "step": 73290 }, { "epoch": 2.478109469556104, "grad_norm": 0.14677928388118744, "learning_rate": 1.8257654104790116e-05, "loss": 0.0419, "step": 73300 }, { "epoch": 2.478447547246357, "grad_norm": 0.20824618637561798, "learning_rate": 1.82448805054846e-05, "loss": 0.0411, "step": 73310 }, { "epoch": 2.4787856249366103, "grad_norm": 0.24844419956207275, "learning_rate": 1.823211037887119e-05, "loss": 0.0491, "step": 73320 }, { "epoch": 2.4791237026268638, "grad_norm": 0.17021867632865906, "learning_rate": 1.821934372634645e-05, "loss": 0.0393, "step": 73330 }, { "epoch": 2.479461780317117, "grad_norm": 0.22334623336791992, "learning_rate": 1.8206580549306472e-05, "loss": 0.0441, "step": 73340 }, { "epoch": 2.47979985800737, "grad_norm": 0.9245617389678955, "learning_rate": 1.819382084914707e-05, "loss": 0.0418, "step": 73350 }, { "epoch": 2.4801379356976234, "grad_norm": 0.22449690103530884, "learning_rate": 1.8181064627263577e-05, "loss": 0.0498, "step": 73360 }, { "epoch": 2.4804760133878765, "grad_norm": 0.40298327803611755, "learning_rate": 1.8168311885051022e-05, "loss": 0.0452, "step": 73370 }, { "epoch": 2.4808140910781296, "grad_norm": 0.16956686973571777, "learning_rate": 1.815556262390402e-05, "loss": 0.042, "step": 73380 }, { "epoch": 2.481152168768383, "grad_norm": 0.23482227325439453, "learning_rate": 1.8142816845216826e-05, "loss": 0.0437, "step": 73390 }, { "epoch": 2.481490246458636, "grad_norm": 0.23530903458595276, "learning_rate": 1.8130074550383276e-05, "loss": 0.0449, "step": 73400 }, { "epoch": 2.481828324148889, "grad_norm": 0.19175241887569427, "learning_rate": 1.8117335740796848e-05, "loss": 0.0477, "step": 73410 }, { "epoch": 2.4821664018391427, "grad_norm": 0.2040984183549881, "learning_rate": 1.810460041785067e-05, "loss": 0.0442, "step": 73420 }, { "epoch": 2.482504479529396, "grad_norm": 0.16720478236675262, "learning_rate": 1.809186858293742e-05, "loss": 0.0455, "step": 73430 }, { "epoch": 2.482842557219649, "grad_norm": 0.284829318523407, "learning_rate": 1.8079140237449448e-05, "loss": 0.0438, "step": 73440 }, { "epoch": 2.4831806349099024, "grad_norm": 0.2680082619190216, "learning_rate": 1.8066415382778707e-05, "loss": 0.0465, "step": 73450 }, { "epoch": 2.4835187126001554, "grad_norm": 0.27564889192581177, "learning_rate": 1.8053694020316786e-05, "loss": 0.04, "step": 73460 }, { "epoch": 2.4838567902904085, "grad_norm": 0.25482118129730225, "learning_rate": 1.8040976151454837e-05, "loss": 0.045, "step": 73470 }, { "epoch": 2.484194867980662, "grad_norm": 0.21217487752437592, "learning_rate": 1.8028261777583704e-05, "loss": 0.0433, "step": 73480 }, { "epoch": 2.484532945670915, "grad_norm": 0.23965026438236237, "learning_rate": 1.8015550900093758e-05, "loss": 0.0379, "step": 73490 }, { "epoch": 2.484871023361168, "grad_norm": 0.2041042447090149, "learning_rate": 1.800284352037512e-05, "loss": 0.0464, "step": 73500 }, { "epoch": 2.4852091010514217, "grad_norm": 0.24893920123577118, "learning_rate": 1.799013963981738e-05, "loss": 0.0399, "step": 73510 }, { "epoch": 2.4855471787416747, "grad_norm": 0.2455557882785797, "learning_rate": 1.797743925980987e-05, "loss": 0.0513, "step": 73520 }, { "epoch": 2.485885256431928, "grad_norm": 0.26725924015045166, "learning_rate": 1.7964742381741435e-05, "loss": 0.049, "step": 73530 }, { "epoch": 2.4862233341221813, "grad_norm": 0.16259752213954926, "learning_rate": 1.7952049007000627e-05, "loss": 0.0437, "step": 73540 }, { "epoch": 2.4865614118124344, "grad_norm": 0.16860318183898926, "learning_rate": 1.7939359136975535e-05, "loss": 0.0442, "step": 73550 }, { "epoch": 2.486899489502688, "grad_norm": 0.3160824477672577, "learning_rate": 1.7926672773053932e-05, "loss": 0.0457, "step": 73560 }, { "epoch": 2.487237567192941, "grad_norm": 0.2858237326145172, "learning_rate": 1.791398991662317e-05, "loss": 0.0383, "step": 73570 }, { "epoch": 2.487575644883194, "grad_norm": 1.067071795463562, "learning_rate": 1.7901310569070244e-05, "loss": 0.0431, "step": 73580 }, { "epoch": 2.4879137225734476, "grad_norm": 0.2565690279006958, "learning_rate": 1.7888634731781724e-05, "loss": 0.0403, "step": 73590 }, { "epoch": 2.4882518002637006, "grad_norm": 0.2474776953458786, "learning_rate": 1.7875962406143844e-05, "loss": 0.0453, "step": 73600 }, { "epoch": 2.4885898779539537, "grad_norm": 0.29748213291168213, "learning_rate": 1.7863293593542403e-05, "loss": 0.0439, "step": 73610 }, { "epoch": 2.488927955644207, "grad_norm": 0.4520995318889618, "learning_rate": 1.7850628295362853e-05, "loss": 0.043, "step": 73620 }, { "epoch": 2.4892660333344603, "grad_norm": 0.20187455415725708, "learning_rate": 1.7837966512990263e-05, "loss": 0.0413, "step": 73630 }, { "epoch": 2.4896041110247134, "grad_norm": 0.1943005621433258, "learning_rate": 1.782530824780931e-05, "loss": 0.0484, "step": 73640 }, { "epoch": 2.489942188714967, "grad_norm": 0.22924669086933136, "learning_rate": 1.7812653501204263e-05, "loss": 0.0432, "step": 73650 }, { "epoch": 2.49028026640522, "grad_norm": 0.17278213798999786, "learning_rate": 1.780000227455903e-05, "loss": 0.0452, "step": 73660 }, { "epoch": 2.490618344095473, "grad_norm": 0.30302301049232483, "learning_rate": 1.778735456925716e-05, "loss": 0.0419, "step": 73670 }, { "epoch": 2.4909564217857265, "grad_norm": 0.4574267864227295, "learning_rate": 1.7774710386681737e-05, "loss": 0.0515, "step": 73680 }, { "epoch": 2.4912944994759796, "grad_norm": 0.16151021420955658, "learning_rate": 1.7762069728215545e-05, "loss": 0.0376, "step": 73690 }, { "epoch": 2.4916325771662327, "grad_norm": 0.23765131831169128, "learning_rate": 1.7749432595240932e-05, "loss": 0.0401, "step": 73700 }, { "epoch": 2.491970654856486, "grad_norm": 0.13873782753944397, "learning_rate": 1.773679898913991e-05, "loss": 0.0461, "step": 73710 }, { "epoch": 2.4923087325467392, "grad_norm": 0.1786404699087143, "learning_rate": 1.7724168911294015e-05, "loss": 0.0446, "step": 73720 }, { "epoch": 2.4926468102369923, "grad_norm": 0.32133862376213074, "learning_rate": 1.7711542363084504e-05, "loss": 0.0451, "step": 73730 }, { "epoch": 2.492984887927246, "grad_norm": 0.22467108070850372, "learning_rate": 1.7698919345892158e-05, "loss": 0.042, "step": 73740 }, { "epoch": 2.493322965617499, "grad_norm": 0.2123631089925766, "learning_rate": 1.768629986109743e-05, "loss": 0.0458, "step": 73750 }, { "epoch": 2.493661043307752, "grad_norm": 0.32569238543510437, "learning_rate": 1.7673683910080357e-05, "loss": 0.0472, "step": 73760 }, { "epoch": 2.4939991209980055, "grad_norm": 0.25957247614860535, "learning_rate": 1.7661071494220638e-05, "loss": 0.0435, "step": 73770 }, { "epoch": 2.4943371986882585, "grad_norm": 0.7918885350227356, "learning_rate": 1.7648462614897498e-05, "loss": 0.0464, "step": 73780 }, { "epoch": 2.4946752763785116, "grad_norm": 0.37627482414245605, "learning_rate": 1.7635857273489863e-05, "loss": 0.0462, "step": 73790 }, { "epoch": 2.495013354068765, "grad_norm": 0.3172354996204376, "learning_rate": 1.7623255471376198e-05, "loss": 0.0448, "step": 73800 }, { "epoch": 2.495351431759018, "grad_norm": 0.5648075342178345, "learning_rate": 1.7610657209934646e-05, "loss": 0.044, "step": 73810 }, { "epoch": 2.4956895094492713, "grad_norm": 0.21921242773532867, "learning_rate": 1.7598062490542915e-05, "loss": 0.0461, "step": 73820 }, { "epoch": 2.4960275871395248, "grad_norm": 0.23416019976139069, "learning_rate": 1.7585471314578377e-05, "loss": 0.0465, "step": 73830 }, { "epoch": 2.496365664829778, "grad_norm": 0.2713267207145691, "learning_rate": 1.7572883683417945e-05, "loss": 0.0388, "step": 73840 }, { "epoch": 2.496703742520031, "grad_norm": 0.23641277849674225, "learning_rate": 1.7560299598438212e-05, "loss": 0.047, "step": 73850 }, { "epoch": 2.4970418202102844, "grad_norm": 0.36146241426467896, "learning_rate": 1.7547719061015334e-05, "loss": 0.0443, "step": 73860 }, { "epoch": 2.4973798979005375, "grad_norm": 0.21185052394866943, "learning_rate": 1.75351420725251e-05, "loss": 0.0473, "step": 73870 }, { "epoch": 2.497717975590791, "grad_norm": 0.22677235305309296, "learning_rate": 1.7522568634342923e-05, "loss": 0.0412, "step": 73880 }, { "epoch": 2.498056053281044, "grad_norm": 0.3595428168773651, "learning_rate": 1.7509998747843825e-05, "loss": 0.0488, "step": 73890 }, { "epoch": 2.498394130971297, "grad_norm": 0.25142213702201843, "learning_rate": 1.7497432414402403e-05, "loss": 0.0399, "step": 73900 }, { "epoch": 2.4987322086615507, "grad_norm": 0.2703346908092499, "learning_rate": 1.7484869635392903e-05, "loss": 0.0454, "step": 73910 }, { "epoch": 2.4990702863518037, "grad_norm": 0.23488907516002655, "learning_rate": 1.747231041218919e-05, "loss": 0.0468, "step": 73920 }, { "epoch": 2.499408364042057, "grad_norm": 0.22440975904464722, "learning_rate": 1.7459754746164692e-05, "loss": 0.0479, "step": 73930 }, { "epoch": 2.4997464417323103, "grad_norm": 0.2454466074705124, "learning_rate": 1.744720263869249e-05, "loss": 0.0431, "step": 73940 }, { "epoch": 2.5000845194225634, "grad_norm": 0.2506387233734131, "learning_rate": 1.7434654091145263e-05, "loss": 0.0461, "step": 73950 }, { "epoch": 2.5004225971128164, "grad_norm": 0.2094084918498993, "learning_rate": 1.742210910489532e-05, "loss": 0.0419, "step": 73960 }, { "epoch": 2.50076067480307, "grad_norm": 0.3541281819343567, "learning_rate": 1.7409567681314532e-05, "loss": 0.0433, "step": 73970 }, { "epoch": 2.501098752493323, "grad_norm": 0.22047661244869232, "learning_rate": 1.7397029821774434e-05, "loss": 0.0455, "step": 73980 }, { "epoch": 2.501436830183576, "grad_norm": 0.242490753531456, "learning_rate": 1.7384495527646126e-05, "loss": 0.0433, "step": 73990 }, { "epoch": 2.5017749078738296, "grad_norm": 0.606072187423706, "learning_rate": 1.737196480030035e-05, "loss": 0.0469, "step": 74000 }, { "epoch": 2.5021129855640827, "grad_norm": 0.34416159987449646, "learning_rate": 1.7359437641107447e-05, "loss": 0.0429, "step": 74010 }, { "epoch": 2.5024510632543358, "grad_norm": 0.22246193885803223, "learning_rate": 1.7346914051437387e-05, "loss": 0.0395, "step": 74020 }, { "epoch": 2.5027891409445893, "grad_norm": 0.297091007232666, "learning_rate": 1.733439403265969e-05, "loss": 0.0471, "step": 74030 }, { "epoch": 2.5031272186348423, "grad_norm": 0.2537921667098999, "learning_rate": 1.732187758614357e-05, "loss": 0.0442, "step": 74040 }, { "epoch": 2.5034652963250954, "grad_norm": 0.23343099653720856, "learning_rate": 1.730936471325777e-05, "loss": 0.0419, "step": 74050 }, { "epoch": 2.503803374015349, "grad_norm": 0.1885959506034851, "learning_rate": 1.7296855415370718e-05, "loss": 0.0406, "step": 74060 }, { "epoch": 2.504141451705602, "grad_norm": 0.22905440628528595, "learning_rate": 1.728434969385035e-05, "loss": 0.0423, "step": 74070 }, { "epoch": 2.504479529395855, "grad_norm": 0.2016594558954239, "learning_rate": 1.7271847550064346e-05, "loss": 0.0427, "step": 74080 }, { "epoch": 2.5048176070861086, "grad_norm": 0.32130834460258484, "learning_rate": 1.725934898537987e-05, "loss": 0.0423, "step": 74090 }, { "epoch": 2.5051556847763616, "grad_norm": 0.3064986765384674, "learning_rate": 1.7246854001163783e-05, "loss": 0.0465, "step": 74100 }, { "epoch": 2.5054937624666147, "grad_norm": 0.24766655266284943, "learning_rate": 1.7234362598782476e-05, "loss": 0.0404, "step": 74110 }, { "epoch": 2.505831840156868, "grad_norm": 0.30312129855155945, "learning_rate": 1.722187477960202e-05, "loss": 0.0446, "step": 74120 }, { "epoch": 2.5061699178471213, "grad_norm": 0.21854670345783234, "learning_rate": 1.7209390544988057e-05, "loss": 0.0422, "step": 74130 }, { "epoch": 2.5065079955373744, "grad_norm": 0.20561322569847107, "learning_rate": 1.7196909896305856e-05, "loss": 0.0396, "step": 74140 }, { "epoch": 2.506846073227628, "grad_norm": 0.1869289129972458, "learning_rate": 1.718443283492026e-05, "loss": 0.0455, "step": 74150 }, { "epoch": 2.507184150917881, "grad_norm": 0.1937311440706253, "learning_rate": 1.7171959362195746e-05, "loss": 0.0478, "step": 74160 }, { "epoch": 2.507522228608134, "grad_norm": 0.2416459619998932, "learning_rate": 1.715948947949642e-05, "loss": 0.0448, "step": 74170 }, { "epoch": 2.5078603062983875, "grad_norm": 0.3504108488559723, "learning_rate": 1.714702318818593e-05, "loss": 0.0438, "step": 74180 }, { "epoch": 2.5081983839886406, "grad_norm": 0.2257089763879776, "learning_rate": 1.7134560489627592e-05, "loss": 0.0437, "step": 74190 }, { "epoch": 2.5085364616788937, "grad_norm": 0.16464334726333618, "learning_rate": 1.712210138518431e-05, "loss": 0.0415, "step": 74200 }, { "epoch": 2.508874539369147, "grad_norm": 1.0162222385406494, "learning_rate": 1.71096458762186e-05, "loss": 0.0407, "step": 74210 }, { "epoch": 2.5092126170594002, "grad_norm": 0.23687906563282013, "learning_rate": 1.7097193964092552e-05, "loss": 0.0403, "step": 74220 }, { "epoch": 2.5095506947496533, "grad_norm": 0.36927399039268494, "learning_rate": 1.708474565016792e-05, "loss": 0.0455, "step": 74230 }, { "epoch": 2.509888772439907, "grad_norm": 0.7702164053916931, "learning_rate": 1.7072300935806e-05, "loss": 0.045, "step": 74240 }, { "epoch": 2.51022685013016, "grad_norm": 0.1831032782793045, "learning_rate": 1.705985982236776e-05, "loss": 0.049, "step": 74250 }, { "epoch": 2.510564927820413, "grad_norm": 0.5778676867485046, "learning_rate": 1.7047422311213697e-05, "loss": 0.0411, "step": 74260 }, { "epoch": 2.5109030055106665, "grad_norm": 0.7970768213272095, "learning_rate": 1.703498840370401e-05, "loss": 0.0514, "step": 74270 }, { "epoch": 2.5112410832009195, "grad_norm": 0.21974776685237885, "learning_rate": 1.702255810119842e-05, "loss": 0.0429, "step": 74280 }, { "epoch": 2.5115791608911726, "grad_norm": 0.2859838902950287, "learning_rate": 1.7010131405056305e-05, "loss": 0.0464, "step": 74290 }, { "epoch": 2.511917238581426, "grad_norm": 0.30782192945480347, "learning_rate": 1.6997708316636613e-05, "loss": 0.0421, "step": 74300 }, { "epoch": 2.512255316271679, "grad_norm": 0.41911280155181885, "learning_rate": 1.6985288837297932e-05, "loss": 0.0457, "step": 74310 }, { "epoch": 2.5125933939619323, "grad_norm": 0.4898199737071991, "learning_rate": 1.6972872968398394e-05, "loss": 0.0458, "step": 74320 }, { "epoch": 2.5129314716521858, "grad_norm": 0.30003198981285095, "learning_rate": 1.6960460711295845e-05, "loss": 0.0471, "step": 74330 }, { "epoch": 2.513269549342439, "grad_norm": 0.2737480700016022, "learning_rate": 1.6948052067347623e-05, "loss": 0.0472, "step": 74340 }, { "epoch": 2.513607627032692, "grad_norm": 0.18595919013023376, "learning_rate": 1.693564703791075e-05, "loss": 0.0418, "step": 74350 }, { "epoch": 2.5139457047229454, "grad_norm": 0.16426822543144226, "learning_rate": 1.6923245624341793e-05, "loss": 0.0408, "step": 74360 }, { "epoch": 2.5142837824131985, "grad_norm": 0.38067007064819336, "learning_rate": 1.6910847827996962e-05, "loss": 0.0457, "step": 74370 }, { "epoch": 2.5146218601034516, "grad_norm": 0.3332165479660034, "learning_rate": 1.689845365023206e-05, "loss": 0.0414, "step": 74380 }, { "epoch": 2.514959937793705, "grad_norm": 0.23898983001708984, "learning_rate": 1.6886063092402505e-05, "loss": 0.049, "step": 74390 }, { "epoch": 2.515298015483958, "grad_norm": 0.1852799355983734, "learning_rate": 1.6873676155863317e-05, "loss": 0.0462, "step": 74400 }, { "epoch": 2.515636093174211, "grad_norm": 0.2125551402568817, "learning_rate": 1.6861292841969085e-05, "loss": 0.0489, "step": 74410 }, { "epoch": 2.5159741708644647, "grad_norm": 0.19449974596500397, "learning_rate": 1.684891315207406e-05, "loss": 0.0435, "step": 74420 }, { "epoch": 2.516312248554718, "grad_norm": 0.17072749137878418, "learning_rate": 1.6836537087532034e-05, "loss": 0.042, "step": 74430 }, { "epoch": 2.516650326244971, "grad_norm": 0.2751351594924927, "learning_rate": 1.6824164649696465e-05, "loss": 0.0451, "step": 74440 }, { "epoch": 2.5169884039352244, "grad_norm": 0.24412360787391663, "learning_rate": 1.681179583992034e-05, "loss": 0.0455, "step": 74450 }, { "epoch": 2.5173264816254775, "grad_norm": 0.5527705550193787, "learning_rate": 1.679943065955636e-05, "loss": 0.0412, "step": 74460 }, { "epoch": 2.5176645593157305, "grad_norm": 0.2965412735939026, "learning_rate": 1.678706910995671e-05, "loss": 0.0435, "step": 74470 }, { "epoch": 2.518002637005984, "grad_norm": 0.28285226225852966, "learning_rate": 1.6774711192473262e-05, "loss": 0.0453, "step": 74480 }, { "epoch": 2.518340714696237, "grad_norm": 0.20894835889339447, "learning_rate": 1.676235690845743e-05, "loss": 0.0409, "step": 74490 }, { "epoch": 2.51867879238649, "grad_norm": 0.26516640186309814, "learning_rate": 1.6750006259260285e-05, "loss": 0.0395, "step": 74500 }, { "epoch": 2.5190168700767437, "grad_norm": 0.2521314024925232, "learning_rate": 1.673765924623244e-05, "loss": 0.0406, "step": 74510 }, { "epoch": 2.5193549477669968, "grad_norm": 0.18314608931541443, "learning_rate": 1.67253158707242e-05, "loss": 0.0438, "step": 74520 }, { "epoch": 2.51969302545725, "grad_norm": 0.6071022152900696, "learning_rate": 1.671297613408537e-05, "loss": 0.0427, "step": 74530 }, { "epoch": 2.5200311031475033, "grad_norm": 0.3448241055011749, "learning_rate": 1.6700640037665444e-05, "loss": 0.0451, "step": 74540 }, { "epoch": 2.5203691808377564, "grad_norm": 0.373060405254364, "learning_rate": 1.668830758281344e-05, "loss": 0.0481, "step": 74550 }, { "epoch": 2.5207072585280095, "grad_norm": 0.11608992516994476, "learning_rate": 1.6675978770878052e-05, "loss": 0.0442, "step": 74560 }, { "epoch": 2.521045336218263, "grad_norm": 0.22737416625022888, "learning_rate": 1.6663653603207506e-05, "loss": 0.0419, "step": 74570 }, { "epoch": 2.521383413908516, "grad_norm": 0.15773513913154602, "learning_rate": 1.6651332081149685e-05, "loss": 0.0466, "step": 74580 }, { "epoch": 2.521721491598769, "grad_norm": 0.2200840413570404, "learning_rate": 1.6639014206052046e-05, "loss": 0.0482, "step": 74590 }, { "epoch": 2.5220595692890226, "grad_norm": 0.2660362422466278, "learning_rate": 1.6626699979261657e-05, "loss": 0.0444, "step": 74600 }, { "epoch": 2.5223976469792757, "grad_norm": 0.16089844703674316, "learning_rate": 1.6614389402125197e-05, "loss": 0.0448, "step": 74610 }, { "epoch": 2.5227357246695292, "grad_norm": 0.21322715282440186, "learning_rate": 1.66020824759889e-05, "loss": 0.0456, "step": 74620 }, { "epoch": 2.5230738023597823, "grad_norm": 1.4606213569641113, "learning_rate": 1.6589779202198675e-05, "loss": 0.0469, "step": 74630 }, { "epoch": 2.5234118800500354, "grad_norm": 0.291259229183197, "learning_rate": 1.6577479582099924e-05, "loss": 0.0426, "step": 74640 }, { "epoch": 2.523749957740289, "grad_norm": 0.24979349970817566, "learning_rate": 1.6565183617037788e-05, "loss": 0.0412, "step": 74650 }, { "epoch": 2.524088035430542, "grad_norm": 0.5908018350601196, "learning_rate": 1.6552891308356894e-05, "loss": 0.0463, "step": 74660 }, { "epoch": 2.524426113120795, "grad_norm": 0.4281908869743347, "learning_rate": 1.6540602657401526e-05, "loss": 0.0457, "step": 74670 }, { "epoch": 2.5247641908110485, "grad_norm": 0.27991098165512085, "learning_rate": 1.6528317665515537e-05, "loss": 0.0447, "step": 74680 }, { "epoch": 2.5251022685013016, "grad_norm": 0.20517674088478088, "learning_rate": 1.6516036334042416e-05, "loss": 0.0444, "step": 74690 }, { "epoch": 2.5254403461915547, "grad_norm": 0.2272598296403885, "learning_rate": 1.650375866432519e-05, "loss": 0.0412, "step": 74700 }, { "epoch": 2.525778423881808, "grad_norm": 0.31578531861305237, "learning_rate": 1.649148465770659e-05, "loss": 0.0393, "step": 74710 }, { "epoch": 2.5261165015720612, "grad_norm": 0.4073300063610077, "learning_rate": 1.647921431552883e-05, "loss": 0.0447, "step": 74720 }, { "epoch": 2.5264545792623143, "grad_norm": 0.46541959047317505, "learning_rate": 1.646694763913381e-05, "loss": 0.0434, "step": 74730 }, { "epoch": 2.526792656952568, "grad_norm": 0.29969942569732666, "learning_rate": 1.6454684629862966e-05, "loss": 0.044, "step": 74740 }, { "epoch": 2.527130734642821, "grad_norm": 0.3023318350315094, "learning_rate": 1.6442425289057388e-05, "loss": 0.0406, "step": 74750 }, { "epoch": 2.5274688123330744, "grad_norm": 0.2083696722984314, "learning_rate": 1.6430169618057718e-05, "loss": 0.0463, "step": 74760 }, { "epoch": 2.5278068900233275, "grad_norm": 0.2227182388305664, "learning_rate": 1.641791761820422e-05, "loss": 0.0432, "step": 74770 }, { "epoch": 2.5281449677135805, "grad_norm": 0.22109241783618927, "learning_rate": 1.640566929083676e-05, "loss": 0.0492, "step": 74780 }, { "epoch": 2.528483045403834, "grad_norm": 0.19353552162647247, "learning_rate": 1.6393424637294823e-05, "loss": 0.0477, "step": 74790 }, { "epoch": 2.528821123094087, "grad_norm": 0.1539042592048645, "learning_rate": 1.638118365891742e-05, "loss": 0.0434, "step": 74800 }, { "epoch": 2.52915920078434, "grad_norm": 0.3501183092594147, "learning_rate": 1.636894635704323e-05, "loss": 0.0463, "step": 74810 }, { "epoch": 2.5294972784745937, "grad_norm": 0.14162079989910126, "learning_rate": 1.6356712733010517e-05, "loss": 0.0378, "step": 74820 }, { "epoch": 2.529835356164847, "grad_norm": 0.3070678114891052, "learning_rate": 1.6344482788157096e-05, "loss": 0.0436, "step": 74830 }, { "epoch": 2.5301734338551, "grad_norm": 0.13979819416999817, "learning_rate": 1.6332256523820444e-05, "loss": 0.0429, "step": 74840 }, { "epoch": 2.5305115115453534, "grad_norm": 0.25806349515914917, "learning_rate": 1.6320033941337597e-05, "loss": 0.0422, "step": 74850 }, { "epoch": 2.5308495892356064, "grad_norm": 0.4639405906200409, "learning_rate": 1.6307815042045216e-05, "loss": 0.0476, "step": 74860 }, { "epoch": 2.5311876669258595, "grad_norm": 0.18020154535770416, "learning_rate": 1.6295599827279513e-05, "loss": 0.0395, "step": 74870 }, { "epoch": 2.531525744616113, "grad_norm": 0.264091432094574, "learning_rate": 1.6283388298376356e-05, "loss": 0.044, "step": 74880 }, { "epoch": 2.531863822306366, "grad_norm": 0.19918511807918549, "learning_rate": 1.6271180456671125e-05, "loss": 0.0447, "step": 74890 }, { "epoch": 2.532201899996619, "grad_norm": 0.19486767053604126, "learning_rate": 1.625897630349893e-05, "loss": 0.0447, "step": 74900 }, { "epoch": 2.5325399776868727, "grad_norm": 0.34834060072898865, "learning_rate": 1.6246775840194335e-05, "loss": 0.0441, "step": 74910 }, { "epoch": 2.5328780553771257, "grad_norm": 0.327173113822937, "learning_rate": 1.6234579068091604e-05, "loss": 0.0436, "step": 74920 }, { "epoch": 2.533216133067379, "grad_norm": 0.6120354533195496, "learning_rate": 1.6222385988524524e-05, "loss": 0.0447, "step": 74930 }, { "epoch": 2.5335542107576323, "grad_norm": 0.2125144600868225, "learning_rate": 1.6210196602826544e-05, "loss": 0.0455, "step": 74940 }, { "epoch": 2.5338922884478854, "grad_norm": 0.20197683572769165, "learning_rate": 1.6198010912330646e-05, "loss": 0.0444, "step": 74950 }, { "epoch": 2.5342303661381385, "grad_norm": 0.4726075530052185, "learning_rate": 1.6185828918369456e-05, "loss": 0.0445, "step": 74960 }, { "epoch": 2.534568443828392, "grad_norm": 0.2124045044183731, "learning_rate": 1.617365062227517e-05, "loss": 0.0379, "step": 74970 }, { "epoch": 2.534906521518645, "grad_norm": 0.15887191891670227, "learning_rate": 1.616147602537962e-05, "loss": 0.044, "step": 74980 }, { "epoch": 2.535244599208898, "grad_norm": 0.37259167432785034, "learning_rate": 1.614930512901415e-05, "loss": 0.0462, "step": 74990 }, { "epoch": 2.5355826768991516, "grad_norm": 0.26111090183258057, "learning_rate": 1.6137137934509804e-05, "loss": 0.0457, "step": 75000 }, { "epoch": 2.5359207545894047, "grad_norm": 0.15757332742214203, "learning_rate": 1.6124974443197123e-05, "loss": 0.0367, "step": 75010 }, { "epoch": 2.5362588322796578, "grad_norm": 1.0923981666564941, "learning_rate": 1.6112814656406306e-05, "loss": 0.0489, "step": 75020 }, { "epoch": 2.5365969099699113, "grad_norm": 0.22423598170280457, "learning_rate": 1.6100658575467125e-05, "loss": 0.0456, "step": 75030 }, { "epoch": 2.5369349876601643, "grad_norm": 0.2232462763786316, "learning_rate": 1.6088506201708974e-05, "loss": 0.0422, "step": 75040 }, { "epoch": 2.5372730653504174, "grad_norm": 0.2898537218570709, "learning_rate": 1.6076357536460785e-05, "loss": 0.042, "step": 75050 }, { "epoch": 2.537611143040671, "grad_norm": 0.2868436574935913, "learning_rate": 1.606421258105113e-05, "loss": 0.0488, "step": 75060 }, { "epoch": 2.537949220730924, "grad_norm": 0.2173348069190979, "learning_rate": 1.6052071336808183e-05, "loss": 0.0397, "step": 75070 }, { "epoch": 2.538287298421177, "grad_norm": 0.18108537793159485, "learning_rate": 1.6039933805059658e-05, "loss": 0.0419, "step": 75080 }, { "epoch": 2.5386253761114306, "grad_norm": 0.6578805446624756, "learning_rate": 1.6027799987132915e-05, "loss": 0.0424, "step": 75090 }, { "epoch": 2.5389634538016836, "grad_norm": 0.22852551937103271, "learning_rate": 1.6015669884354888e-05, "loss": 0.0468, "step": 75100 }, { "epoch": 2.5393015314919367, "grad_norm": 0.20937563478946686, "learning_rate": 1.6003543498052125e-05, "loss": 0.0399, "step": 75110 }, { "epoch": 2.5396396091821902, "grad_norm": 0.2505326271057129, "learning_rate": 1.5991420829550715e-05, "loss": 0.0427, "step": 75120 }, { "epoch": 2.5399776868724433, "grad_norm": 0.2080872803926468, "learning_rate": 1.5979301880176405e-05, "loss": 0.0453, "step": 75130 }, { "epoch": 2.5403157645626964, "grad_norm": 0.4857563078403473, "learning_rate": 1.5967186651254485e-05, "loss": 0.0457, "step": 75140 }, { "epoch": 2.54065384225295, "grad_norm": 0.1911708563566208, "learning_rate": 1.5955075144109853e-05, "loss": 0.0414, "step": 75150 }, { "epoch": 2.540991919943203, "grad_norm": 0.1694248765707016, "learning_rate": 1.594296736006703e-05, "loss": 0.0384, "step": 75160 }, { "epoch": 2.541329997633456, "grad_norm": 0.1719195395708084, "learning_rate": 1.59308633004501e-05, "loss": 0.0463, "step": 75170 }, { "epoch": 2.5416680753237095, "grad_norm": 0.2543559968471527, "learning_rate": 1.5918762966582724e-05, "loss": 0.0427, "step": 75180 }, { "epoch": 2.5420061530139626, "grad_norm": 0.3559230864048004, "learning_rate": 1.5906666359788203e-05, "loss": 0.0444, "step": 75190 }, { "epoch": 2.5423442307042157, "grad_norm": 0.1872677206993103, "learning_rate": 1.589457348138938e-05, "loss": 0.045, "step": 75200 }, { "epoch": 2.542682308394469, "grad_norm": 0.20565031468868256, "learning_rate": 1.588248433270872e-05, "loss": 0.04, "step": 75210 }, { "epoch": 2.5430203860847223, "grad_norm": 0.2204008549451828, "learning_rate": 1.587039891506828e-05, "loss": 0.0468, "step": 75220 }, { "epoch": 2.5433584637749753, "grad_norm": 0.2164212465286255, "learning_rate": 1.5858317229789715e-05, "loss": 0.0434, "step": 75230 }, { "epoch": 2.543696541465229, "grad_norm": 0.3111855089664459, "learning_rate": 1.584623927819423e-05, "loss": 0.0451, "step": 75240 }, { "epoch": 2.544034619155482, "grad_norm": 0.2309415340423584, "learning_rate": 1.583416506160269e-05, "loss": 0.0473, "step": 75250 }, { "epoch": 2.544372696845735, "grad_norm": 0.38705307245254517, "learning_rate": 1.5822094581335472e-05, "loss": 0.0452, "step": 75260 }, { "epoch": 2.5447107745359885, "grad_norm": 0.1636429876089096, "learning_rate": 1.5810027838712604e-05, "loss": 0.0461, "step": 75270 }, { "epoch": 2.5450488522262416, "grad_norm": 0.38886985182762146, "learning_rate": 1.5797964835053687e-05, "loss": 0.0465, "step": 75280 }, { "epoch": 2.5453869299164946, "grad_norm": 0.1747608184814453, "learning_rate": 1.5785905571677933e-05, "loss": 0.046, "step": 75290 }, { "epoch": 2.545725007606748, "grad_norm": 0.19192852079868317, "learning_rate": 1.5773850049904086e-05, "loss": 0.0406, "step": 75300 }, { "epoch": 2.546063085297001, "grad_norm": 0.16658593714237213, "learning_rate": 1.5761798271050543e-05, "loss": 0.0462, "step": 75310 }, { "epoch": 2.5464011629872543, "grad_norm": 0.46533381938934326, "learning_rate": 1.5749750236435277e-05, "loss": 0.0514, "step": 75320 }, { "epoch": 2.546739240677508, "grad_norm": 0.17997731268405914, "learning_rate": 1.5737705947375815e-05, "loss": 0.0415, "step": 75330 }, { "epoch": 2.547077318367761, "grad_norm": 0.2648165225982666, "learning_rate": 1.5725665405189322e-05, "loss": 0.0423, "step": 75340 }, { "epoch": 2.547415396058014, "grad_norm": 0.23690594732761383, "learning_rate": 1.571362861119253e-05, "loss": 0.0472, "step": 75350 }, { "epoch": 2.5477534737482674, "grad_norm": 0.16530515253543854, "learning_rate": 1.570159556670177e-05, "loss": 0.0461, "step": 75360 }, { "epoch": 2.5480915514385205, "grad_norm": 0.26735273003578186, "learning_rate": 1.5689566273032946e-05, "loss": 0.0418, "step": 75370 }, { "epoch": 2.5484296291287736, "grad_norm": 0.18793855607509613, "learning_rate": 1.5677540731501583e-05, "loss": 0.0481, "step": 75380 }, { "epoch": 2.548767706819027, "grad_norm": 0.18619078397750854, "learning_rate": 1.5665518943422746e-05, "loss": 0.0486, "step": 75390 }, { "epoch": 2.54910578450928, "grad_norm": 0.5633594393730164, "learning_rate": 1.5653500910111134e-05, "loss": 0.0451, "step": 75400 }, { "epoch": 2.5494438621995332, "grad_norm": 0.46904459595680237, "learning_rate": 1.5641486632881023e-05, "loss": 0.0462, "step": 75410 }, { "epoch": 2.5497819398897867, "grad_norm": 0.19811893999576569, "learning_rate": 1.56294761130463e-05, "loss": 0.0418, "step": 75420 }, { "epoch": 2.55012001758004, "grad_norm": 0.3610197603702545, "learning_rate": 1.5617469351920373e-05, "loss": 0.0421, "step": 75430 }, { "epoch": 2.550458095270293, "grad_norm": 0.17407311499118805, "learning_rate": 1.5605466350816315e-05, "loss": 0.0452, "step": 75440 }, { "epoch": 2.5507961729605464, "grad_norm": 0.48615944385528564, "learning_rate": 1.559346711104674e-05, "loss": 0.0454, "step": 75450 }, { "epoch": 2.5511342506507995, "grad_norm": 0.2595774233341217, "learning_rate": 1.5581471633923883e-05, "loss": 0.0405, "step": 75460 }, { "epoch": 2.5514723283410525, "grad_norm": 0.15572936832904816, "learning_rate": 1.5569479920759505e-05, "loss": 0.0423, "step": 75470 }, { "epoch": 2.551810406031306, "grad_norm": 0.26678159832954407, "learning_rate": 1.5557491972865072e-05, "loss": 0.0436, "step": 75480 }, { "epoch": 2.552148483721559, "grad_norm": 0.5535783171653748, "learning_rate": 1.554550779155152e-05, "loss": 0.0464, "step": 75490 }, { "epoch": 2.552486561411812, "grad_norm": 0.30114808678627014, "learning_rate": 1.5533527378129453e-05, "loss": 0.0461, "step": 75500 }, { "epoch": 2.5528246391020657, "grad_norm": 0.2508147060871124, "learning_rate": 1.552155073390899e-05, "loss": 0.0451, "step": 75510 }, { "epoch": 2.5531627167923188, "grad_norm": 0.2361958622932434, "learning_rate": 1.5509577860199903e-05, "loss": 0.0478, "step": 75520 }, { "epoch": 2.553500794482572, "grad_norm": 0.17120636999607086, "learning_rate": 1.5497608758311525e-05, "loss": 0.042, "step": 75530 }, { "epoch": 2.5538388721728253, "grad_norm": 0.26407745480537415, "learning_rate": 1.5485643429552797e-05, "loss": 0.042, "step": 75540 }, { "epoch": 2.5541769498630784, "grad_norm": 0.38151973485946655, "learning_rate": 1.5473681875232197e-05, "loss": 0.0484, "step": 75550 }, { "epoch": 2.5545150275533315, "grad_norm": 0.3298543691635132, "learning_rate": 1.5461724096657837e-05, "loss": 0.0457, "step": 75560 }, { "epoch": 2.554853105243585, "grad_norm": 0.16946107149124146, "learning_rate": 1.5449770095137416e-05, "loss": 0.0408, "step": 75570 }, { "epoch": 2.555191182933838, "grad_norm": 0.22626620531082153, "learning_rate": 1.5437819871978178e-05, "loss": 0.0394, "step": 75580 }, { "epoch": 2.5555292606240916, "grad_norm": 0.22359681129455566, "learning_rate": 1.5425873428486992e-05, "loss": 0.0445, "step": 75590 }, { "epoch": 2.5558673383143447, "grad_norm": 0.21787258982658386, "learning_rate": 1.5413930765970312e-05, "loss": 0.0399, "step": 75600 }, { "epoch": 2.5562054160045977, "grad_norm": 0.26834309101104736, "learning_rate": 1.5401991885734175e-05, "loss": 0.044, "step": 75610 }, { "epoch": 2.5565434936948512, "grad_norm": 0.2134740650653839, "learning_rate": 1.5390056789084174e-05, "loss": 0.0417, "step": 75620 }, { "epoch": 2.5568815713851043, "grad_norm": 0.18840090930461884, "learning_rate": 1.5378125477325545e-05, "loss": 0.048, "step": 75630 }, { "epoch": 2.5572196490753574, "grad_norm": 0.3351964056491852, "learning_rate": 1.5366197951763046e-05, "loss": 0.0463, "step": 75640 }, { "epoch": 2.557557726765611, "grad_norm": 0.2703741490840912, "learning_rate": 1.5354274213701075e-05, "loss": 0.0456, "step": 75650 }, { "epoch": 2.557895804455864, "grad_norm": 0.21132344007492065, "learning_rate": 1.5342354264443566e-05, "loss": 0.0476, "step": 75660 }, { "epoch": 2.558233882146117, "grad_norm": 0.39334481954574585, "learning_rate": 1.533043810529411e-05, "loss": 0.0428, "step": 75670 }, { "epoch": 2.5585719598363705, "grad_norm": 0.26749613881111145, "learning_rate": 1.5318525737555807e-05, "loss": 0.0449, "step": 75680 }, { "epoch": 2.5589100375266236, "grad_norm": 0.24966786801815033, "learning_rate": 1.53066171625314e-05, "loss": 0.048, "step": 75690 }, { "epoch": 2.5592481152168767, "grad_norm": 0.15414826571941376, "learning_rate": 1.529471238152317e-05, "loss": 0.0506, "step": 75700 }, { "epoch": 2.55958619290713, "grad_norm": 0.1895897537469864, "learning_rate": 1.5282811395833023e-05, "loss": 0.0442, "step": 75710 }, { "epoch": 2.5599242705973833, "grad_norm": 0.5072552561759949, "learning_rate": 1.5270914206762397e-05, "loss": 0.0483, "step": 75720 }, { "epoch": 2.5602623482876368, "grad_norm": 0.30707186460494995, "learning_rate": 1.525902081561241e-05, "loss": 0.0464, "step": 75730 }, { "epoch": 2.56060042597789, "grad_norm": 0.2547428011894226, "learning_rate": 1.5247131223683664e-05, "loss": 0.0443, "step": 75740 }, { "epoch": 2.560938503668143, "grad_norm": 0.1379653513431549, "learning_rate": 1.5235245432276408e-05, "loss": 0.0413, "step": 75750 }, { "epoch": 2.5612765813583964, "grad_norm": 0.24890103936195374, "learning_rate": 1.5223363442690424e-05, "loss": 0.0437, "step": 75760 }, { "epoch": 2.5616146590486495, "grad_norm": 0.2238738089799881, "learning_rate": 1.5211485256225133e-05, "loss": 0.0499, "step": 75770 }, { "epoch": 2.5619527367389026, "grad_norm": 0.19370897114276886, "learning_rate": 1.519961087417951e-05, "loss": 0.04, "step": 75780 }, { "epoch": 2.562290814429156, "grad_norm": 0.2224390059709549, "learning_rate": 1.5187740297852137e-05, "loss": 0.0432, "step": 75790 }, { "epoch": 2.562628892119409, "grad_norm": 0.37971851229667664, "learning_rate": 1.517587352854113e-05, "loss": 0.0434, "step": 75800 }, { "epoch": 2.562966969809662, "grad_norm": 0.2945347726345062, "learning_rate": 1.516401056754423e-05, "loss": 0.0425, "step": 75810 }, { "epoch": 2.5633050474999157, "grad_norm": 0.3422018885612488, "learning_rate": 1.5152151416158783e-05, "loss": 0.0451, "step": 75820 }, { "epoch": 2.563643125190169, "grad_norm": 0.174502432346344, "learning_rate": 1.5140296075681642e-05, "loss": 0.0439, "step": 75830 }, { "epoch": 2.563981202880422, "grad_norm": 0.2665809988975525, "learning_rate": 1.5128444547409331e-05, "loss": 0.0418, "step": 75840 }, { "epoch": 2.5643192805706754, "grad_norm": 0.372564435005188, "learning_rate": 1.5116596832637864e-05, "loss": 0.0466, "step": 75850 }, { "epoch": 2.5646573582609284, "grad_norm": 0.1314748078584671, "learning_rate": 1.5104752932662952e-05, "loss": 0.0394, "step": 75860 }, { "epoch": 2.5649954359511815, "grad_norm": 0.22557437419891357, "learning_rate": 1.5092912848779778e-05, "loss": 0.0409, "step": 75870 }, { "epoch": 2.565333513641435, "grad_norm": 0.29127463698387146, "learning_rate": 1.5081076582283194e-05, "loss": 0.0429, "step": 75880 }, { "epoch": 2.565671591331688, "grad_norm": 0.1778339445590973, "learning_rate": 1.5069244134467553e-05, "loss": 0.0383, "step": 75890 }, { "epoch": 2.566009669021941, "grad_norm": 0.36468833684921265, "learning_rate": 1.5057415506626876e-05, "loss": 0.0385, "step": 75900 }, { "epoch": 2.5663477467121947, "grad_norm": 1.416163444519043, "learning_rate": 1.5045590700054668e-05, "loss": 0.0484, "step": 75910 }, { "epoch": 2.5666858244024477, "grad_norm": 0.48423129320144653, "learning_rate": 1.5033769716044144e-05, "loss": 0.0361, "step": 75920 }, { "epoch": 2.567023902092701, "grad_norm": 0.4130205512046814, "learning_rate": 1.5021952555887975e-05, "loss": 0.0469, "step": 75930 }, { "epoch": 2.5673619797829543, "grad_norm": 0.3287448287010193, "learning_rate": 1.50101392208785e-05, "loss": 0.0399, "step": 75940 }, { "epoch": 2.5677000574732074, "grad_norm": 0.38326290249824524, "learning_rate": 1.4998329712307579e-05, "loss": 0.049, "step": 75950 }, { "epoch": 2.5680381351634605, "grad_norm": 0.525902509689331, "learning_rate": 1.4986524031466708e-05, "loss": 0.04, "step": 75960 }, { "epoch": 2.568376212853714, "grad_norm": 0.2965480089187622, "learning_rate": 1.4974722179646916e-05, "loss": 0.0427, "step": 75970 }, { "epoch": 2.568714290543967, "grad_norm": 0.17312858998775482, "learning_rate": 1.4962924158138847e-05, "loss": 0.0443, "step": 75980 }, { "epoch": 2.56905236823422, "grad_norm": 0.2300005555152893, "learning_rate": 1.4951129968232714e-05, "loss": 0.0447, "step": 75990 }, { "epoch": 2.5693904459244736, "grad_norm": 0.3400602638721466, "learning_rate": 1.4939339611218329e-05, "loss": 0.0459, "step": 76000 }, { "epoch": 2.5697285236147267, "grad_norm": 0.49840253591537476, "learning_rate": 1.4927553088385032e-05, "loss": 0.0451, "step": 76010 }, { "epoch": 2.5700666013049798, "grad_norm": 0.19731153547763824, "learning_rate": 1.4915770401021801e-05, "loss": 0.043, "step": 76020 }, { "epoch": 2.5704046789952333, "grad_norm": 0.17908908426761627, "learning_rate": 1.4903991550417185e-05, "loss": 0.0463, "step": 76030 }, { "epoch": 2.5707427566854864, "grad_norm": 0.7106303572654724, "learning_rate": 1.4892216537859277e-05, "loss": 0.0468, "step": 76040 }, { "epoch": 2.5710808343757394, "grad_norm": 0.22445809841156006, "learning_rate": 1.4880445364635775e-05, "loss": 0.0443, "step": 76050 }, { "epoch": 2.571418912065993, "grad_norm": 0.28599250316619873, "learning_rate": 1.4868678032033973e-05, "loss": 0.0427, "step": 76060 }, { "epoch": 2.571756989756246, "grad_norm": 0.25771310925483704, "learning_rate": 1.4856914541340739e-05, "loss": 0.046, "step": 76070 }, { "epoch": 2.572095067446499, "grad_norm": 0.23394861817359924, "learning_rate": 1.4845154893842472e-05, "loss": 0.0417, "step": 76080 }, { "epoch": 2.5724331451367526, "grad_norm": 0.2266700565814972, "learning_rate": 1.4833399090825234e-05, "loss": 0.0419, "step": 76090 }, { "epoch": 2.5727712228270057, "grad_norm": 0.23264877498149872, "learning_rate": 1.4821647133574562e-05, "loss": 0.0412, "step": 76100 }, { "epoch": 2.5731093005172587, "grad_norm": 0.22976483404636383, "learning_rate": 1.4809899023375706e-05, "loss": 0.0431, "step": 76110 }, { "epoch": 2.5734473782075122, "grad_norm": 0.21419279277324677, "learning_rate": 1.4798154761513366e-05, "loss": 0.0449, "step": 76120 }, { "epoch": 2.5737854558977653, "grad_norm": 0.21604971587657928, "learning_rate": 1.4786414349271916e-05, "loss": 0.0451, "step": 76130 }, { "epoch": 2.5741235335880184, "grad_norm": 0.25044646859169006, "learning_rate": 1.4774677787935226e-05, "loss": 0.0424, "step": 76140 }, { "epoch": 2.574461611278272, "grad_norm": 0.2574043273925781, "learning_rate": 1.4762945078786838e-05, "loss": 0.0438, "step": 76150 }, { "epoch": 2.574799688968525, "grad_norm": 0.5126864910125732, "learning_rate": 1.4751216223109777e-05, "loss": 0.0448, "step": 76160 }, { "epoch": 2.575137766658778, "grad_norm": 0.21711158752441406, "learning_rate": 1.4739491222186714e-05, "loss": 0.0387, "step": 76170 }, { "epoch": 2.5754758443490315, "grad_norm": 0.2325010746717453, "learning_rate": 1.4727770077299879e-05, "loss": 0.041, "step": 76180 }, { "epoch": 2.5758139220392846, "grad_norm": 0.30482667684555054, "learning_rate": 1.4716052789731094e-05, "loss": 0.0406, "step": 76190 }, { "epoch": 2.5761519997295377, "grad_norm": 0.26631104946136475, "learning_rate": 1.470433936076171e-05, "loss": 0.0425, "step": 76200 }, { "epoch": 2.576490077419791, "grad_norm": 0.2330067902803421, "learning_rate": 1.4692629791672719e-05, "loss": 0.0447, "step": 76210 }, { "epoch": 2.5768281551100443, "grad_norm": 0.2693004608154297, "learning_rate": 1.468092408374464e-05, "loss": 0.0454, "step": 76220 }, { "epoch": 2.5771662328002973, "grad_norm": 0.21788689494132996, "learning_rate": 1.4669222238257602e-05, "loss": 0.0432, "step": 76230 }, { "epoch": 2.577504310490551, "grad_norm": 0.2710466980934143, "learning_rate": 1.4657524256491301e-05, "loss": 0.0418, "step": 76240 }, { "epoch": 2.577842388180804, "grad_norm": 0.22406353056430817, "learning_rate": 1.464583013972502e-05, "loss": 0.0429, "step": 76250 }, { "epoch": 2.578180465871057, "grad_norm": 0.2656194567680359, "learning_rate": 1.4634139889237591e-05, "loss": 0.0492, "step": 76260 }, { "epoch": 2.5785185435613105, "grad_norm": 0.4876675307750702, "learning_rate": 1.4622453506307448e-05, "loss": 0.0459, "step": 76270 }, { "epoch": 2.5788566212515636, "grad_norm": 0.2537546157836914, "learning_rate": 1.461077099221262e-05, "loss": 0.0433, "step": 76280 }, { "epoch": 2.5791946989418166, "grad_norm": 0.22668811678886414, "learning_rate": 1.4599092348230652e-05, "loss": 0.0411, "step": 76290 }, { "epoch": 2.57953277663207, "grad_norm": 0.35316649079322815, "learning_rate": 1.4587417575638724e-05, "loss": 0.045, "step": 76300 }, { "epoch": 2.579870854322323, "grad_norm": 0.14535515010356903, "learning_rate": 1.4575746675713563e-05, "loss": 0.0407, "step": 76310 }, { "epoch": 2.5802089320125763, "grad_norm": 0.2634105682373047, "learning_rate": 1.4564079649731504e-05, "loss": 0.05, "step": 76320 }, { "epoch": 2.58054700970283, "grad_norm": 0.3548717796802521, "learning_rate": 1.4552416498968402e-05, "loss": 0.0435, "step": 76330 }, { "epoch": 2.580885087393083, "grad_norm": 0.211412250995636, "learning_rate": 1.4540757224699747e-05, "loss": 0.0437, "step": 76340 }, { "epoch": 2.581223165083336, "grad_norm": 0.2135988175868988, "learning_rate": 1.452910182820056e-05, "loss": 0.0433, "step": 76350 }, { "epoch": 2.5815612427735894, "grad_norm": 0.46689847111701965, "learning_rate": 1.4517450310745468e-05, "loss": 0.0437, "step": 76360 }, { "epoch": 2.5818993204638425, "grad_norm": 1.8699398040771484, "learning_rate": 1.450580267360866e-05, "loss": 0.0457, "step": 76370 }, { "epoch": 2.5822373981540956, "grad_norm": 0.3159376382827759, "learning_rate": 1.4494158918063922e-05, "loss": 0.0389, "step": 76380 }, { "epoch": 2.582575475844349, "grad_norm": 0.17048197984695435, "learning_rate": 1.4482519045384568e-05, "loss": 0.0425, "step": 76390 }, { "epoch": 2.582913553534602, "grad_norm": 0.668311595916748, "learning_rate": 1.4470883056843548e-05, "loss": 0.0484, "step": 76400 }, { "epoch": 2.5832516312248552, "grad_norm": 0.2994537949562073, "learning_rate": 1.4459250953713322e-05, "loss": 0.0435, "step": 76410 }, { "epoch": 2.5835897089151088, "grad_norm": 0.15431354939937592, "learning_rate": 1.444762273726598e-05, "loss": 0.0421, "step": 76420 }, { "epoch": 2.583927786605362, "grad_norm": 0.21531282365322113, "learning_rate": 1.443599840877316e-05, "loss": 0.0456, "step": 76430 }, { "epoch": 2.584265864295615, "grad_norm": 0.2106827348470688, "learning_rate": 1.4424377969506103e-05, "loss": 0.0424, "step": 76440 }, { "epoch": 2.5846039419858684, "grad_norm": 0.3949810862541199, "learning_rate": 1.4412761420735565e-05, "loss": 0.0436, "step": 76450 }, { "epoch": 2.5849420196761215, "grad_norm": 0.16996139287948608, "learning_rate": 1.4401148763731953e-05, "loss": 0.0384, "step": 76460 }, { "epoch": 2.5852800973663745, "grad_norm": 0.2014690339565277, "learning_rate": 1.4389539999765173e-05, "loss": 0.0425, "step": 76470 }, { "epoch": 2.585618175056628, "grad_norm": 0.19722747802734375, "learning_rate": 1.4377935130104758e-05, "loss": 0.0439, "step": 76480 }, { "epoch": 2.585956252746881, "grad_norm": 0.24764199554920197, "learning_rate": 1.4366334156019801e-05, "loss": 0.0444, "step": 76490 }, { "epoch": 2.586294330437134, "grad_norm": 0.1379082351922989, "learning_rate": 1.435473707877898e-05, "loss": 0.0443, "step": 76500 }, { "epoch": 2.5866324081273877, "grad_norm": 0.2241821140050888, "learning_rate": 1.4343143899650508e-05, "loss": 0.0457, "step": 76510 }, { "epoch": 2.5869704858176408, "grad_norm": 0.26930075883865356, "learning_rate": 1.4331554619902204e-05, "loss": 0.0445, "step": 76520 }, { "epoch": 2.587308563507894, "grad_norm": 0.20299111306667328, "learning_rate": 1.4319969240801484e-05, "loss": 0.0421, "step": 76530 }, { "epoch": 2.5876466411981474, "grad_norm": 0.3269229531288147, "learning_rate": 1.430838776361526e-05, "loss": 0.0452, "step": 76540 }, { "epoch": 2.5879847188884004, "grad_norm": 0.3268565535545349, "learning_rate": 1.4296810189610093e-05, "loss": 0.0422, "step": 76550 }, { "epoch": 2.588322796578654, "grad_norm": 0.3460668921470642, "learning_rate": 1.4285236520052087e-05, "loss": 0.0441, "step": 76560 }, { "epoch": 2.588660874268907, "grad_norm": 0.2429773062467575, "learning_rate": 1.4273666756206933e-05, "loss": 0.0485, "step": 76570 }, { "epoch": 2.58899895195916, "grad_norm": 0.3316822350025177, "learning_rate": 1.4262100899339848e-05, "loss": 0.046, "step": 76580 }, { "epoch": 2.5893370296494136, "grad_norm": 0.2034349888563156, "learning_rate": 1.4250538950715696e-05, "loss": 0.0472, "step": 76590 }, { "epoch": 2.5896751073396667, "grad_norm": 0.22495798766613007, "learning_rate": 1.4238980911598842e-05, "loss": 0.0414, "step": 76600 }, { "epoch": 2.5900131850299197, "grad_norm": 0.31822094321250916, "learning_rate": 1.4227426783253267e-05, "loss": 0.0394, "step": 76610 }, { "epoch": 2.5903512627201732, "grad_norm": 0.21683064103126526, "learning_rate": 1.4215876566942521e-05, "loss": 0.0414, "step": 76620 }, { "epoch": 2.5906893404104263, "grad_norm": 0.26302894949913025, "learning_rate": 1.4204330263929721e-05, "loss": 0.0418, "step": 76630 }, { "epoch": 2.5910274181006794, "grad_norm": 0.2052471786737442, "learning_rate": 1.4192787875477537e-05, "loss": 0.0463, "step": 76640 }, { "epoch": 2.591365495790933, "grad_norm": 0.23422221839427948, "learning_rate": 1.4181249402848246e-05, "loss": 0.0431, "step": 76650 }, { "epoch": 2.591703573481186, "grad_norm": 0.4775873124599457, "learning_rate": 1.4169714847303656e-05, "loss": 0.0457, "step": 76660 }, { "epoch": 2.592041651171439, "grad_norm": 0.29939913749694824, "learning_rate": 1.4158184210105197e-05, "loss": 0.0408, "step": 76670 }, { "epoch": 2.5923797288616925, "grad_norm": 0.26570940017700195, "learning_rate": 1.4146657492513787e-05, "loss": 0.0413, "step": 76680 }, { "epoch": 2.5927178065519456, "grad_norm": 0.19723287224769592, "learning_rate": 1.4135134695790047e-05, "loss": 0.0427, "step": 76690 }, { "epoch": 2.593055884242199, "grad_norm": 0.32831108570098877, "learning_rate": 1.4123615821194037e-05, "loss": 0.0489, "step": 76700 }, { "epoch": 2.593393961932452, "grad_norm": 0.2560344338417053, "learning_rate": 1.4112100869985473e-05, "loss": 0.0423, "step": 76710 }, { "epoch": 2.5937320396227053, "grad_norm": 0.4019911587238312, "learning_rate": 1.4100589843423589e-05, "loss": 0.0441, "step": 76720 }, { "epoch": 2.594070117312959, "grad_norm": 0.19939430058002472, "learning_rate": 1.408908274276723e-05, "loss": 0.046, "step": 76730 }, { "epoch": 2.594408195003212, "grad_norm": 0.4184168875217438, "learning_rate": 1.4077579569274786e-05, "loss": 0.0519, "step": 76740 }, { "epoch": 2.594746272693465, "grad_norm": 0.23894140124320984, "learning_rate": 1.4066080324204245e-05, "loss": 0.04, "step": 76750 }, { "epoch": 2.5950843503837184, "grad_norm": 0.1471218317747116, "learning_rate": 1.405458500881312e-05, "loss": 0.0407, "step": 76760 }, { "epoch": 2.5954224280739715, "grad_norm": 0.4693823456764221, "learning_rate": 1.4043093624358538e-05, "loss": 0.0404, "step": 76770 }, { "epoch": 2.5957605057642246, "grad_norm": 0.20999310910701752, "learning_rate": 1.4031606172097189e-05, "loss": 0.043, "step": 76780 }, { "epoch": 2.596098583454478, "grad_norm": 0.17526791989803314, "learning_rate": 1.40201226532853e-05, "loss": 0.0428, "step": 76790 }, { "epoch": 2.596436661144731, "grad_norm": 0.1697191745042801, "learning_rate": 1.4008643069178701e-05, "loss": 0.0415, "step": 76800 }, { "epoch": 2.596774738834984, "grad_norm": 0.18534430861473083, "learning_rate": 1.399716742103278e-05, "loss": 0.0425, "step": 76810 }, { "epoch": 2.5971128165252377, "grad_norm": 0.5624651312828064, "learning_rate": 1.398569571010252e-05, "loss": 0.0464, "step": 76820 }, { "epoch": 2.597450894215491, "grad_norm": 0.19787484407424927, "learning_rate": 1.3974227937642414e-05, "loss": 0.0429, "step": 76830 }, { "epoch": 2.597788971905744, "grad_norm": 0.4119083285331726, "learning_rate": 1.3962764104906595e-05, "loss": 0.0407, "step": 76840 }, { "epoch": 2.5981270495959974, "grad_norm": 0.40869373083114624, "learning_rate": 1.3951304213148696e-05, "loss": 0.0487, "step": 76850 }, { "epoch": 2.5984651272862505, "grad_norm": 0.2459126114845276, "learning_rate": 1.3939848263621979e-05, "loss": 0.0463, "step": 76860 }, { "epoch": 2.5988032049765035, "grad_norm": 0.325031042098999, "learning_rate": 1.3928396257579219e-05, "loss": 0.045, "step": 76870 }, { "epoch": 2.599141282666757, "grad_norm": 0.3365964889526367, "learning_rate": 1.3916948196272833e-05, "loss": 0.0431, "step": 76880 }, { "epoch": 2.59947936035701, "grad_norm": 0.46099042892456055, "learning_rate": 1.3905504080954734e-05, "loss": 0.0425, "step": 76890 }, { "epoch": 2.599817438047263, "grad_norm": 0.44478532671928406, "learning_rate": 1.3894063912876454e-05, "loss": 0.0437, "step": 76900 }, { "epoch": 2.6001555157375167, "grad_norm": 0.21270479261875153, "learning_rate": 1.3882627693289046e-05, "loss": 0.0478, "step": 76910 }, { "epoch": 2.6004935934277698, "grad_norm": 0.3586030900478363, "learning_rate": 1.3871195423443183e-05, "loss": 0.0462, "step": 76920 }, { "epoch": 2.600831671118023, "grad_norm": 0.17507946491241455, "learning_rate": 1.3859767104589044e-05, "loss": 0.0425, "step": 76930 }, { "epoch": 2.6011697488082763, "grad_norm": 0.20807379484176636, "learning_rate": 1.3848342737976467e-05, "loss": 0.0459, "step": 76940 }, { "epoch": 2.6015078264985294, "grad_norm": 0.6265398263931274, "learning_rate": 1.3836922324854757e-05, "loss": 0.0489, "step": 76950 }, { "epoch": 2.6018459041887825, "grad_norm": 0.15994134545326233, "learning_rate": 1.382550586647287e-05, "loss": 0.0448, "step": 76960 }, { "epoch": 2.602183981879036, "grad_norm": 0.18202270567417145, "learning_rate": 1.3814093364079257e-05, "loss": 0.0463, "step": 76970 }, { "epoch": 2.602522059569289, "grad_norm": 0.251382440328598, "learning_rate": 1.3802684818921991e-05, "loss": 0.0443, "step": 76980 }, { "epoch": 2.602860137259542, "grad_norm": 0.5840492248535156, "learning_rate": 1.3791280232248687e-05, "loss": 0.0457, "step": 76990 }, { "epoch": 2.6031982149497956, "grad_norm": 0.1671764850616455, "learning_rate": 1.3779879605306555e-05, "loss": 0.0407, "step": 77000 }, { "epoch": 2.6035362926400487, "grad_norm": 0.7421486377716064, "learning_rate": 1.376848293934232e-05, "loss": 0.0464, "step": 77010 }, { "epoch": 2.603874370330302, "grad_norm": 0.18293587863445282, "learning_rate": 1.3757090235602316e-05, "loss": 0.042, "step": 77020 }, { "epoch": 2.6042124480205553, "grad_norm": 0.2811943292617798, "learning_rate": 1.3745701495332447e-05, "loss": 0.0397, "step": 77030 }, { "epoch": 2.6045505257108084, "grad_norm": 0.5314447283744812, "learning_rate": 1.3734316719778145e-05, "loss": 0.043, "step": 77040 }, { "epoch": 2.6048886034010614, "grad_norm": 0.2644058167934418, "learning_rate": 1.3722935910184454e-05, "loss": 0.0502, "step": 77050 }, { "epoch": 2.605226681091315, "grad_norm": 0.31463733315467834, "learning_rate": 1.3711559067795921e-05, "loss": 0.0438, "step": 77060 }, { "epoch": 2.605564758781568, "grad_norm": 0.2142648547887802, "learning_rate": 1.3700186193856756e-05, "loss": 0.0442, "step": 77070 }, { "epoch": 2.605902836471821, "grad_norm": 0.33653566241264343, "learning_rate": 1.368881728961064e-05, "loss": 0.0412, "step": 77080 }, { "epoch": 2.6062409141620746, "grad_norm": 0.39983147382736206, "learning_rate": 1.3677452356300885e-05, "loss": 0.0399, "step": 77090 }, { "epoch": 2.6065789918523277, "grad_norm": 0.2983890175819397, "learning_rate": 1.3666091395170321e-05, "loss": 0.0453, "step": 77100 }, { "epoch": 2.6069170695425807, "grad_norm": 0.17502814531326294, "learning_rate": 1.3654734407461384e-05, "loss": 0.0425, "step": 77110 }, { "epoch": 2.6072551472328342, "grad_norm": 0.2967052161693573, "learning_rate": 1.3643381394416022e-05, "loss": 0.0468, "step": 77120 }, { "epoch": 2.6075932249230873, "grad_norm": 0.33506566286087036, "learning_rate": 1.3632032357275838e-05, "loss": 0.0412, "step": 77130 }, { "epoch": 2.6079313026133404, "grad_norm": 0.18707016110420227, "learning_rate": 1.3620687297281898e-05, "loss": 0.0411, "step": 77140 }, { "epoch": 2.608269380303594, "grad_norm": 0.16986800730228424, "learning_rate": 1.3609346215674917e-05, "loss": 0.0441, "step": 77150 }, { "epoch": 2.608607457993847, "grad_norm": 0.2511519491672516, "learning_rate": 1.3598009113695109e-05, "loss": 0.0431, "step": 77160 }, { "epoch": 2.6089455356841, "grad_norm": 0.139705628156662, "learning_rate": 1.358667599258231e-05, "loss": 0.0423, "step": 77170 }, { "epoch": 2.6092836133743535, "grad_norm": 0.18107137084007263, "learning_rate": 1.357534685357586e-05, "loss": 0.0417, "step": 77180 }, { "epoch": 2.6096216910646066, "grad_norm": 0.2048598676919937, "learning_rate": 1.3564021697914715e-05, "loss": 0.0454, "step": 77190 }, { "epoch": 2.6099597687548597, "grad_norm": 0.19681040942668915, "learning_rate": 1.3552700526837381e-05, "loss": 0.0441, "step": 77200 }, { "epoch": 2.610297846445113, "grad_norm": 0.20291662216186523, "learning_rate": 1.3541383341581925e-05, "loss": 0.0409, "step": 77210 }, { "epoch": 2.6106359241353663, "grad_norm": 0.23437033593654633, "learning_rate": 1.3530070143385965e-05, "loss": 0.05, "step": 77220 }, { "epoch": 2.6109740018256193, "grad_norm": 0.20888347923755646, "learning_rate": 1.3518760933486702e-05, "loss": 0.0483, "step": 77230 }, { "epoch": 2.611312079515873, "grad_norm": 0.3318253755569458, "learning_rate": 1.3507455713120908e-05, "loss": 0.0503, "step": 77240 }, { "epoch": 2.611650157206126, "grad_norm": 0.15997381508350372, "learning_rate": 1.3496154483524876e-05, "loss": 0.0455, "step": 77250 }, { "epoch": 2.611988234896379, "grad_norm": 0.153541699051857, "learning_rate": 1.3484857245934506e-05, "loss": 0.0453, "step": 77260 }, { "epoch": 2.6123263125866325, "grad_norm": 0.2577654719352722, "learning_rate": 1.3473564001585248e-05, "loss": 0.0418, "step": 77270 }, { "epoch": 2.6126643902768856, "grad_norm": 0.26329243183135986, "learning_rate": 1.3462274751712129e-05, "loss": 0.0432, "step": 77280 }, { "epoch": 2.6130024679671386, "grad_norm": 0.20352669060230255, "learning_rate": 1.3450989497549698e-05, "loss": 0.0386, "step": 77290 }, { "epoch": 2.613340545657392, "grad_norm": 0.34606242179870605, "learning_rate": 1.3439708240332115e-05, "loss": 0.0417, "step": 77300 }, { "epoch": 2.6136786233476452, "grad_norm": 0.2631070613861084, "learning_rate": 1.3428430981293043e-05, "loss": 0.0453, "step": 77310 }, { "epoch": 2.6140167010378983, "grad_norm": 0.2865139842033386, "learning_rate": 1.3417157721665808e-05, "loss": 0.0413, "step": 77320 }, { "epoch": 2.614354778728152, "grad_norm": 0.226011261343956, "learning_rate": 1.3405888462683186e-05, "loss": 0.0433, "step": 77330 }, { "epoch": 2.614692856418405, "grad_norm": 0.3547355830669403, "learning_rate": 1.3394623205577595e-05, "loss": 0.0405, "step": 77340 }, { "epoch": 2.615030934108658, "grad_norm": 0.27182736992836, "learning_rate": 1.3383361951580963e-05, "loss": 0.0448, "step": 77350 }, { "epoch": 2.6153690117989115, "grad_norm": 0.20368073880672455, "learning_rate": 1.337210470192483e-05, "loss": 0.0481, "step": 77360 }, { "epoch": 2.6157070894891645, "grad_norm": 0.3066059350967407, "learning_rate": 1.3360851457840245e-05, "loss": 0.0456, "step": 77370 }, { "epoch": 2.6160451671794176, "grad_norm": 0.1950124204158783, "learning_rate": 1.3349602220557856e-05, "loss": 0.0469, "step": 77380 }, { "epoch": 2.616383244869671, "grad_norm": 0.3550020456314087, "learning_rate": 1.3338356991307865e-05, "loss": 0.0446, "step": 77390 }, { "epoch": 2.616721322559924, "grad_norm": 0.23572123050689697, "learning_rate": 1.3327115771320048e-05, "loss": 0.0417, "step": 77400 }, { "epoch": 2.6170594002501772, "grad_norm": 0.32809123396873474, "learning_rate": 1.3315878561823697e-05, "loss": 0.0414, "step": 77410 }, { "epoch": 2.6173974779404308, "grad_norm": 0.5062118172645569, "learning_rate": 1.330464536404773e-05, "loss": 0.0441, "step": 77420 }, { "epoch": 2.617735555630684, "grad_norm": 0.21755141019821167, "learning_rate": 1.3293416179220563e-05, "loss": 0.0447, "step": 77430 }, { "epoch": 2.618073633320937, "grad_norm": 0.2015734314918518, "learning_rate": 1.328219100857021e-05, "loss": 0.0405, "step": 77440 }, { "epoch": 2.6184117110111904, "grad_norm": 0.21285296976566315, "learning_rate": 1.327096985332425e-05, "loss": 0.0401, "step": 77450 }, { "epoch": 2.6187497887014435, "grad_norm": 0.30842727422714233, "learning_rate": 1.3259752714709816e-05, "loss": 0.0417, "step": 77460 }, { "epoch": 2.6190878663916966, "grad_norm": 0.4381183385848999, "learning_rate": 1.3248539593953569e-05, "loss": 0.0456, "step": 77470 }, { "epoch": 2.61942594408195, "grad_norm": 0.37026163935661316, "learning_rate": 1.323733049228178e-05, "loss": 0.0422, "step": 77480 }, { "epoch": 2.619764021772203, "grad_norm": 0.2905544936656952, "learning_rate": 1.3226125410920276e-05, "loss": 0.0452, "step": 77490 }, { "epoch": 2.620102099462456, "grad_norm": 0.24351027607917786, "learning_rate": 1.3214924351094388e-05, "loss": 0.0451, "step": 77500 }, { "epoch": 2.6204401771527097, "grad_norm": 0.2512839436531067, "learning_rate": 1.3203727314029069e-05, "loss": 0.0453, "step": 77510 }, { "epoch": 2.620778254842963, "grad_norm": 0.38117125630378723, "learning_rate": 1.3192534300948811e-05, "loss": 0.0464, "step": 77520 }, { "epoch": 2.6211163325332163, "grad_norm": 0.18506236374378204, "learning_rate": 1.3181345313077675e-05, "loss": 0.0436, "step": 77530 }, { "epoch": 2.6214544102234694, "grad_norm": 0.1993182748556137, "learning_rate": 1.3170160351639243e-05, "loss": 0.0397, "step": 77540 }, { "epoch": 2.6217924879137224, "grad_norm": 0.6126166582107544, "learning_rate": 1.3158979417856721e-05, "loss": 0.0469, "step": 77550 }, { "epoch": 2.622130565603976, "grad_norm": 0.3628544211387634, "learning_rate": 1.31478025129528e-05, "loss": 0.0444, "step": 77560 }, { "epoch": 2.622468643294229, "grad_norm": 0.3317379355430603, "learning_rate": 1.3136629638149794e-05, "loss": 0.0435, "step": 77570 }, { "epoch": 2.622806720984482, "grad_norm": 0.41379478573799133, "learning_rate": 1.3125460794669547e-05, "loss": 0.0513, "step": 77580 }, { "epoch": 2.6231447986747356, "grad_norm": 0.2182399332523346, "learning_rate": 1.3114295983733483e-05, "loss": 0.0409, "step": 77590 }, { "epoch": 2.6234828763649887, "grad_norm": 0.22499974071979523, "learning_rate": 1.3103135206562534e-05, "loss": 0.0414, "step": 77600 }, { "epoch": 2.6238209540552417, "grad_norm": 0.2413894534111023, "learning_rate": 1.3091978464377264e-05, "loss": 0.0443, "step": 77610 }, { "epoch": 2.6241590317454953, "grad_norm": 0.7918470501899719, "learning_rate": 1.3080825758397724e-05, "loss": 0.0462, "step": 77620 }, { "epoch": 2.6244971094357483, "grad_norm": 0.13872739672660828, "learning_rate": 1.3069677089843568e-05, "loss": 0.0415, "step": 77630 }, { "epoch": 2.6248351871260014, "grad_norm": 0.40486669540405273, "learning_rate": 1.305853245993401e-05, "loss": 0.0482, "step": 77640 }, { "epoch": 2.625173264816255, "grad_norm": 0.16362658143043518, "learning_rate": 1.3047391869887809e-05, "loss": 0.0436, "step": 77650 }, { "epoch": 2.625511342506508, "grad_norm": 0.4416141211986542, "learning_rate": 1.3036255320923263e-05, "loss": 0.0461, "step": 77660 }, { "epoch": 2.6258494201967615, "grad_norm": 0.2781197726726532, "learning_rate": 1.3025122814258273e-05, "loss": 0.0467, "step": 77670 }, { "epoch": 2.6261874978870146, "grad_norm": 0.21012508869171143, "learning_rate": 1.3013994351110249e-05, "loss": 0.0424, "step": 77680 }, { "epoch": 2.6265255755772676, "grad_norm": 0.21850641071796417, "learning_rate": 1.3002869932696194e-05, "loss": 0.0484, "step": 77690 }, { "epoch": 2.626863653267521, "grad_norm": 0.7698867917060852, "learning_rate": 1.2991749560232664e-05, "loss": 0.0482, "step": 77700 }, { "epoch": 2.627201730957774, "grad_norm": 0.22009895741939545, "learning_rate": 1.2980633234935768e-05, "loss": 0.0442, "step": 77710 }, { "epoch": 2.6275398086480273, "grad_norm": 0.5050392150878906, "learning_rate": 1.2969520958021153e-05, "loss": 0.0411, "step": 77720 }, { "epoch": 2.627877886338281, "grad_norm": 0.2700328826904297, "learning_rate": 1.295841273070405e-05, "loss": 0.0454, "step": 77730 }, { "epoch": 2.628215964028534, "grad_norm": 0.8572322726249695, "learning_rate": 1.2947308554199255e-05, "loss": 0.041, "step": 77740 }, { "epoch": 2.628554041718787, "grad_norm": 1.0629897117614746, "learning_rate": 1.2936208429721075e-05, "loss": 0.0397, "step": 77750 }, { "epoch": 2.6288921194090404, "grad_norm": 0.5709351897239685, "learning_rate": 1.2925112358483416e-05, "loss": 0.0424, "step": 77760 }, { "epoch": 2.6292301970992935, "grad_norm": 0.2600272297859192, "learning_rate": 1.2914020341699728e-05, "loss": 0.0451, "step": 77770 }, { "epoch": 2.6295682747895466, "grad_norm": 0.2266615629196167, "learning_rate": 1.2902932380583032e-05, "loss": 0.0408, "step": 77780 }, { "epoch": 2.6299063524798, "grad_norm": 0.37134283781051636, "learning_rate": 1.2891848476345863e-05, "loss": 0.0449, "step": 77790 }, { "epoch": 2.630244430170053, "grad_norm": 0.9159934520721436, "learning_rate": 1.2880768630200368e-05, "loss": 0.0442, "step": 77800 }, { "epoch": 2.6305825078603062, "grad_norm": 0.6248215436935425, "learning_rate": 1.286969284335819e-05, "loss": 0.0457, "step": 77810 }, { "epoch": 2.6309205855505597, "grad_norm": 2.911583662033081, "learning_rate": 1.2858621117030578e-05, "loss": 0.0419, "step": 77820 }, { "epoch": 2.631258663240813, "grad_norm": 0.22069337964057922, "learning_rate": 1.2847553452428324e-05, "loss": 0.0461, "step": 77830 }, { "epoch": 2.631596740931066, "grad_norm": 0.20927388966083527, "learning_rate": 1.283648985076178e-05, "loss": 0.0474, "step": 77840 }, { "epoch": 2.6319348186213194, "grad_norm": 0.21620899438858032, "learning_rate": 1.2825430313240817e-05, "loss": 0.0402, "step": 77850 }, { "epoch": 2.6322728963115725, "grad_norm": 0.1881638467311859, "learning_rate": 1.281437484107491e-05, "loss": 0.0427, "step": 77860 }, { "epoch": 2.6326109740018255, "grad_norm": 0.407741516828537, "learning_rate": 1.2803323435473058e-05, "loss": 0.0445, "step": 77870 }, { "epoch": 2.632949051692079, "grad_norm": 0.16040244698524475, "learning_rate": 1.2792276097643841e-05, "loss": 0.0425, "step": 77880 }, { "epoch": 2.633287129382332, "grad_norm": 0.34694820642471313, "learning_rate": 1.2781232828795342e-05, "loss": 0.0404, "step": 77890 }, { "epoch": 2.633625207072585, "grad_norm": 0.21874450147151947, "learning_rate": 1.2770193630135292e-05, "loss": 0.0431, "step": 77900 }, { "epoch": 2.6339632847628387, "grad_norm": 0.15237122774124146, "learning_rate": 1.2759158502870877e-05, "loss": 0.0414, "step": 77910 }, { "epoch": 2.6343013624530918, "grad_norm": 0.3126939535140991, "learning_rate": 1.2748127448208913e-05, "loss": 0.0413, "step": 77920 }, { "epoch": 2.634639440143345, "grad_norm": 0.237232968211174, "learning_rate": 1.2737100467355706e-05, "loss": 0.0378, "step": 77930 }, { "epoch": 2.6349775178335983, "grad_norm": 0.2150094211101532, "learning_rate": 1.2726077561517169e-05, "loss": 0.0375, "step": 77940 }, { "epoch": 2.6353155955238514, "grad_norm": 0.168190598487854, "learning_rate": 1.2715058731898755e-05, "loss": 0.0437, "step": 77950 }, { "epoch": 2.6356536732141045, "grad_norm": 0.22035276889801025, "learning_rate": 1.2704043979705471e-05, "loss": 0.0424, "step": 77960 }, { "epoch": 2.635991750904358, "grad_norm": 0.20681844651699066, "learning_rate": 1.269303330614185e-05, "loss": 0.044, "step": 77970 }, { "epoch": 2.636329828594611, "grad_norm": 0.25214436650276184, "learning_rate": 1.2682026712412015e-05, "loss": 0.0468, "step": 77980 }, { "epoch": 2.636667906284864, "grad_norm": 0.24796529114246368, "learning_rate": 1.2671024199719644e-05, "loss": 0.0487, "step": 77990 }, { "epoch": 2.6370059839751177, "grad_norm": 0.25899943709373474, "learning_rate": 1.2660025769267936e-05, "loss": 0.0456, "step": 78000 }, { "epoch": 2.6373440616653707, "grad_norm": 0.2690454423427582, "learning_rate": 1.2649031422259667e-05, "loss": 0.044, "step": 78010 }, { "epoch": 2.637682139355624, "grad_norm": 0.2909177541732788, "learning_rate": 1.2638041159897163e-05, "loss": 0.047, "step": 78020 }, { "epoch": 2.6380202170458773, "grad_norm": 0.2651711106300354, "learning_rate": 1.262705498338232e-05, "loss": 0.0427, "step": 78030 }, { "epoch": 2.6383582947361304, "grad_norm": 0.19555464386940002, "learning_rate": 1.2616072893916542e-05, "loss": 0.0425, "step": 78040 }, { "epoch": 2.6386963724263834, "grad_norm": 0.5615649819374084, "learning_rate": 1.2605094892700836e-05, "loss": 0.0443, "step": 78050 }, { "epoch": 2.639034450116637, "grad_norm": 0.23194296658039093, "learning_rate": 1.259412098093572e-05, "loss": 0.0488, "step": 78060 }, { "epoch": 2.63937252780689, "grad_norm": 0.25896960496902466, "learning_rate": 1.258315115982131e-05, "loss": 0.0435, "step": 78070 }, { "epoch": 2.639710605497143, "grad_norm": 0.21887540817260742, "learning_rate": 1.2572185430557203e-05, "loss": 0.0378, "step": 78080 }, { "epoch": 2.6400486831873966, "grad_norm": 0.18812181055545807, "learning_rate": 1.2561223794342659e-05, "loss": 0.0421, "step": 78090 }, { "epoch": 2.6403867608776497, "grad_norm": 0.2342350035905838, "learning_rate": 1.255026625237637e-05, "loss": 0.0442, "step": 78100 }, { "epoch": 2.6407248385679027, "grad_norm": 0.32237517833709717, "learning_rate": 1.2539312805856678e-05, "loss": 0.0443, "step": 78110 }, { "epoch": 2.6410629162581563, "grad_norm": 0.4363294839859009, "learning_rate": 1.25283634559814e-05, "loss": 0.0465, "step": 78120 }, { "epoch": 2.6414009939484093, "grad_norm": 0.279563844203949, "learning_rate": 1.2517418203947972e-05, "loss": 0.0417, "step": 78130 }, { "epoch": 2.6417390716386624, "grad_norm": 0.26061511039733887, "learning_rate": 1.2506477050953308e-05, "loss": 0.0439, "step": 78140 }, { "epoch": 2.642077149328916, "grad_norm": 0.5726866126060486, "learning_rate": 1.2495539998193972e-05, "loss": 0.0432, "step": 78150 }, { "epoch": 2.642415227019169, "grad_norm": 0.2383071631193161, "learning_rate": 1.2484607046865975e-05, "loss": 0.0458, "step": 78160 }, { "epoch": 2.642753304709422, "grad_norm": 0.18863485753536224, "learning_rate": 1.2473678198164968e-05, "loss": 0.0412, "step": 78170 }, { "epoch": 2.6430913823996756, "grad_norm": 0.13746777176856995, "learning_rate": 1.2462753453286075e-05, "loss": 0.0412, "step": 78180 }, { "epoch": 2.6434294600899286, "grad_norm": 0.28578612208366394, "learning_rate": 1.2451832813424031e-05, "loss": 0.0467, "step": 78190 }, { "epoch": 2.6437675377801817, "grad_norm": 0.28737160563468933, "learning_rate": 1.2440916279773095e-05, "loss": 0.0421, "step": 78200 }, { "epoch": 2.644105615470435, "grad_norm": 0.20765681564807892, "learning_rate": 1.2430003853527101e-05, "loss": 0.0395, "step": 78210 }, { "epoch": 2.6444436931606883, "grad_norm": 0.17542214691638947, "learning_rate": 1.2419095535879383e-05, "loss": 0.0393, "step": 78220 }, { "epoch": 2.6447817708509413, "grad_norm": 0.22036868333816528, "learning_rate": 1.2408191328022878e-05, "loss": 0.046, "step": 78230 }, { "epoch": 2.645119848541195, "grad_norm": 0.26102039217948914, "learning_rate": 1.2397291231150066e-05, "loss": 0.0491, "step": 78240 }, { "epoch": 2.645457926231448, "grad_norm": 0.33487963676452637, "learning_rate": 1.2386395246452937e-05, "loss": 0.0406, "step": 78250 }, { "epoch": 2.645796003921701, "grad_norm": 0.27001121640205383, "learning_rate": 1.2375503375123083e-05, "loss": 0.041, "step": 78260 }, { "epoch": 2.6461340816119545, "grad_norm": 0.3970032036304474, "learning_rate": 1.2364615618351583e-05, "loss": 0.0413, "step": 78270 }, { "epoch": 2.6464721593022076, "grad_norm": 0.1930571347475052, "learning_rate": 1.2353731977329169e-05, "loss": 0.0419, "step": 78280 }, { "epoch": 2.6468102369924607, "grad_norm": 0.35185128450393677, "learning_rate": 1.2342852453246007e-05, "loss": 0.0454, "step": 78290 }, { "epoch": 2.647148314682714, "grad_norm": 0.25912755727767944, "learning_rate": 1.2331977047291898e-05, "loss": 0.0441, "step": 78300 }, { "epoch": 2.6474863923729672, "grad_norm": 0.2849721610546112, "learning_rate": 1.2321105760656137e-05, "loss": 0.044, "step": 78310 }, { "epoch": 2.6478244700632203, "grad_norm": 0.1508031040430069, "learning_rate": 1.2310238594527613e-05, "loss": 0.0439, "step": 78320 }, { "epoch": 2.648162547753474, "grad_norm": 0.4252881705760956, "learning_rate": 1.2299375550094693e-05, "loss": 0.0407, "step": 78330 }, { "epoch": 2.648500625443727, "grad_norm": 0.4428544044494629, "learning_rate": 1.2288516628545421e-05, "loss": 0.0452, "step": 78340 }, { "epoch": 2.64883870313398, "grad_norm": 0.3288096785545349, "learning_rate": 1.2277661831067255e-05, "loss": 0.0378, "step": 78350 }, { "epoch": 2.6491767808242335, "grad_norm": 0.28381699323654175, "learning_rate": 1.2266811158847286e-05, "loss": 0.0443, "step": 78360 }, { "epoch": 2.6495148585144865, "grad_norm": 0.20010079443454742, "learning_rate": 1.2255964613072107e-05, "loss": 0.0451, "step": 78370 }, { "epoch": 2.6498529362047396, "grad_norm": 0.23104603588581085, "learning_rate": 1.2245122194927905e-05, "loss": 0.0381, "step": 78380 }, { "epoch": 2.650191013894993, "grad_norm": 0.18250174820423126, "learning_rate": 1.2234283905600364e-05, "loss": 0.0422, "step": 78390 }, { "epoch": 2.650529091585246, "grad_norm": 0.19130776822566986, "learning_rate": 1.222344974627475e-05, "loss": 0.0453, "step": 78400 }, { "epoch": 2.6508671692754993, "grad_norm": 0.15767158567905426, "learning_rate": 1.2212619718135875e-05, "loss": 0.0403, "step": 78410 }, { "epoch": 2.6512052469657528, "grad_norm": 0.439903199672699, "learning_rate": 1.2201793822368118e-05, "loss": 0.0542, "step": 78420 }, { "epoch": 2.651543324656006, "grad_norm": 0.2545819878578186, "learning_rate": 1.2190972060155337e-05, "loss": 0.042, "step": 78430 }, { "epoch": 2.651881402346259, "grad_norm": 0.23461377620697021, "learning_rate": 1.2180154432681013e-05, "loss": 0.0432, "step": 78440 }, { "epoch": 2.6522194800365124, "grad_norm": 0.2402041256427765, "learning_rate": 1.2169340941128148e-05, "loss": 0.0423, "step": 78450 }, { "epoch": 2.6525575577267655, "grad_norm": 0.31257471442222595, "learning_rate": 1.2158531586679267e-05, "loss": 0.044, "step": 78460 }, { "epoch": 2.6528956354170186, "grad_norm": 1.7921979427337646, "learning_rate": 1.2147726370516477e-05, "loss": 0.0396, "step": 78470 }, { "epoch": 2.653233713107272, "grad_norm": 0.3468717336654663, "learning_rate": 1.2136925293821422e-05, "loss": 0.0439, "step": 78480 }, { "epoch": 2.653571790797525, "grad_norm": 0.4230455160140991, "learning_rate": 1.2126128357775306e-05, "loss": 0.043, "step": 78490 }, { "epoch": 2.6539098684877787, "grad_norm": 0.1931847631931305, "learning_rate": 1.2115335563558838e-05, "loss": 0.0414, "step": 78500 }, { "epoch": 2.6542479461780317, "grad_norm": 0.24915266036987305, "learning_rate": 1.2104546912352327e-05, "loss": 0.044, "step": 78510 }, { "epoch": 2.654586023868285, "grad_norm": 0.2513826787471771, "learning_rate": 1.2093762405335557e-05, "loss": 0.0441, "step": 78520 }, { "epoch": 2.6549241015585383, "grad_norm": 0.17986458539962769, "learning_rate": 1.2082982043687974e-05, "loss": 0.0445, "step": 78530 }, { "epoch": 2.6552621792487914, "grad_norm": 0.1735353171825409, "learning_rate": 1.207220582858845e-05, "loss": 0.0475, "step": 78540 }, { "epoch": 2.6556002569390444, "grad_norm": 0.31239256262779236, "learning_rate": 1.206143376121549e-05, "loss": 0.0388, "step": 78550 }, { "epoch": 2.655938334629298, "grad_norm": 0.20140500366687775, "learning_rate": 1.2050665842747078e-05, "loss": 0.0453, "step": 78560 }, { "epoch": 2.656276412319551, "grad_norm": 0.2672976553440094, "learning_rate": 1.2039902074360804e-05, "loss": 0.0406, "step": 78570 }, { "epoch": 2.656614490009804, "grad_norm": 0.6575401425361633, "learning_rate": 1.2029142457233755e-05, "loss": 0.0427, "step": 78580 }, { "epoch": 2.6569525677000576, "grad_norm": 0.1567719578742981, "learning_rate": 1.2018386992542601e-05, "loss": 0.0444, "step": 78590 }, { "epoch": 2.6572906453903107, "grad_norm": 0.35314327478408813, "learning_rate": 1.2007635681463542e-05, "loss": 0.0457, "step": 78600 }, { "epoch": 2.6576287230805637, "grad_norm": 0.21675845980644226, "learning_rate": 1.1996888525172334e-05, "loss": 0.0431, "step": 78610 }, { "epoch": 2.6579668007708173, "grad_norm": 0.36053556203842163, "learning_rate": 1.198614552484425e-05, "loss": 0.0455, "step": 78620 }, { "epoch": 2.6583048784610703, "grad_norm": 0.3264201581478119, "learning_rate": 1.197540668165415e-05, "loss": 0.0403, "step": 78630 }, { "epoch": 2.658642956151324, "grad_norm": 0.165141761302948, "learning_rate": 1.1964671996776395e-05, "loss": 0.0371, "step": 78640 }, { "epoch": 2.658981033841577, "grad_norm": 0.46353879570961, "learning_rate": 1.1953941471384922e-05, "loss": 0.0464, "step": 78650 }, { "epoch": 2.65931911153183, "grad_norm": 0.2081671804189682, "learning_rate": 1.1943215106653205e-05, "loss": 0.0447, "step": 78660 }, { "epoch": 2.6596571892220835, "grad_norm": 0.2045496553182602, "learning_rate": 1.193249290375429e-05, "loss": 0.0412, "step": 78670 }, { "epoch": 2.6599952669123366, "grad_norm": 0.26549234986305237, "learning_rate": 1.1921774863860696e-05, "loss": 0.0437, "step": 78680 }, { "epoch": 2.6603333446025896, "grad_norm": 0.22036896646022797, "learning_rate": 1.1911060988144556e-05, "loss": 0.0463, "step": 78690 }, { "epoch": 2.660671422292843, "grad_norm": 0.2153007537126541, "learning_rate": 1.1900351277777533e-05, "loss": 0.0419, "step": 78700 }, { "epoch": 2.661009499983096, "grad_norm": 0.2873241603374481, "learning_rate": 1.1889645733930798e-05, "loss": 0.046, "step": 78710 }, { "epoch": 2.6613475776733493, "grad_norm": 0.31996065378189087, "learning_rate": 1.1878944357775107e-05, "loss": 0.0423, "step": 78720 }, { "epoch": 2.661685655363603, "grad_norm": 0.27642202377319336, "learning_rate": 1.1868247150480749e-05, "loss": 0.0464, "step": 78730 }, { "epoch": 2.662023733053856, "grad_norm": 0.3219708204269409, "learning_rate": 1.1857554113217567e-05, "loss": 0.0407, "step": 78740 }, { "epoch": 2.662361810744109, "grad_norm": 0.27459517121315, "learning_rate": 1.1846865247154903e-05, "loss": 0.0441, "step": 78750 }, { "epoch": 2.6626998884343624, "grad_norm": 0.15694116055965424, "learning_rate": 1.1836180553461711e-05, "loss": 0.0456, "step": 78760 }, { "epoch": 2.6630379661246155, "grad_norm": 0.32897794246673584, "learning_rate": 1.1825500033306419e-05, "loss": 0.0469, "step": 78770 }, { "epoch": 2.6633760438148686, "grad_norm": 0.26103079319000244, "learning_rate": 1.1814823687857052e-05, "loss": 0.043, "step": 78780 }, { "epoch": 2.663714121505122, "grad_norm": 0.3023120164871216, "learning_rate": 1.1804151518281153e-05, "loss": 0.0463, "step": 78790 }, { "epoch": 2.664052199195375, "grad_norm": 0.21760819852352142, "learning_rate": 1.179348352574583e-05, "loss": 0.042, "step": 78800 }, { "epoch": 2.6643902768856282, "grad_norm": 0.21539421379566193, "learning_rate": 1.1782819711417697e-05, "loss": 0.0437, "step": 78810 }, { "epoch": 2.6647283545758818, "grad_norm": 0.24273300170898438, "learning_rate": 1.1772160076462951e-05, "loss": 0.0476, "step": 78820 }, { "epoch": 2.665066432266135, "grad_norm": 0.17047272622585297, "learning_rate": 1.1761504622047293e-05, "loss": 0.0403, "step": 78830 }, { "epoch": 2.665404509956388, "grad_norm": 0.3113851249217987, "learning_rate": 1.1750853349335999e-05, "loss": 0.0491, "step": 78840 }, { "epoch": 2.6657425876466414, "grad_norm": 0.28787049651145935, "learning_rate": 1.1740206259493874e-05, "loss": 0.044, "step": 78850 }, { "epoch": 2.6660806653368945, "grad_norm": 0.14802186191082, "learning_rate": 1.1729563353685286e-05, "loss": 0.0448, "step": 78860 }, { "epoch": 2.6664187430271475, "grad_norm": 0.6043316125869751, "learning_rate": 1.1718924633074091e-05, "loss": 0.0451, "step": 78870 }, { "epoch": 2.666756820717401, "grad_norm": 0.643591582775116, "learning_rate": 1.1708290098823765e-05, "loss": 0.0448, "step": 78880 }, { "epoch": 2.667094898407654, "grad_norm": 0.19638174772262573, "learning_rate": 1.1697659752097245e-05, "loss": 0.0439, "step": 78890 }, { "epoch": 2.667432976097907, "grad_norm": 0.1885453462600708, "learning_rate": 1.1687033594057068e-05, "loss": 0.0436, "step": 78900 }, { "epoch": 2.6677710537881607, "grad_norm": 0.31901517510414124, "learning_rate": 1.1676411625865297e-05, "loss": 0.049, "step": 78910 }, { "epoch": 2.6681091314784138, "grad_norm": 0.29426953196525574, "learning_rate": 1.1665793848683542e-05, "loss": 0.0486, "step": 78920 }, { "epoch": 2.668447209168667, "grad_norm": 0.1575455218553543, "learning_rate": 1.1655180263672927e-05, "loss": 0.041, "step": 78930 }, { "epoch": 2.6687852868589204, "grad_norm": 0.32502639293670654, "learning_rate": 1.1644570871994142e-05, "loss": 0.0428, "step": 78940 }, { "epoch": 2.6691233645491734, "grad_norm": 0.2609647512435913, "learning_rate": 1.163396567480744e-05, "loss": 0.0446, "step": 78950 }, { "epoch": 2.6694614422394265, "grad_norm": 0.2342866212129593, "learning_rate": 1.1623364673272552e-05, "loss": 0.0432, "step": 78960 }, { "epoch": 2.66979951992968, "grad_norm": 0.49849534034729004, "learning_rate": 1.1612767868548807e-05, "loss": 0.0404, "step": 78970 }, { "epoch": 2.670137597619933, "grad_norm": 0.6026780605316162, "learning_rate": 1.1602175261795057e-05, "loss": 0.0461, "step": 78980 }, { "epoch": 2.670475675310186, "grad_norm": 0.18345755338668823, "learning_rate": 1.1591586854169696e-05, "loss": 0.0437, "step": 78990 }, { "epoch": 2.6708137530004397, "grad_norm": 0.2521032691001892, "learning_rate": 1.158100264683064e-05, "loss": 0.0448, "step": 79000 }, { "epoch": 2.6711518306906927, "grad_norm": 0.18360109627246857, "learning_rate": 1.1570422640935386e-05, "loss": 0.0429, "step": 79010 }, { "epoch": 2.671489908380946, "grad_norm": 0.21208472549915314, "learning_rate": 1.155984683764092e-05, "loss": 0.0475, "step": 79020 }, { "epoch": 2.6718279860711993, "grad_norm": 0.4243779480457306, "learning_rate": 1.1549275238103812e-05, "loss": 0.0431, "step": 79030 }, { "epoch": 2.6721660637614524, "grad_norm": 0.18586495518684387, "learning_rate": 1.1538707843480152e-05, "loss": 0.0441, "step": 79040 }, { "epoch": 2.6725041414517055, "grad_norm": 0.22912172973155975, "learning_rate": 1.1528144654925593e-05, "loss": 0.0435, "step": 79050 }, { "epoch": 2.672842219141959, "grad_norm": 0.21463578939437866, "learning_rate": 1.1517585673595277e-05, "loss": 0.0449, "step": 79060 }, { "epoch": 2.673180296832212, "grad_norm": 0.3263290822505951, "learning_rate": 1.150703090064395e-05, "loss": 0.0448, "step": 79070 }, { "epoch": 2.673518374522465, "grad_norm": 0.2365133911371231, "learning_rate": 1.1496480337225835e-05, "loss": 0.041, "step": 79080 }, { "epoch": 2.6738564522127186, "grad_norm": 0.5124918818473816, "learning_rate": 1.1485933984494751e-05, "loss": 0.0402, "step": 79090 }, { "epoch": 2.6741945299029717, "grad_norm": 0.4175597131252289, "learning_rate": 1.1475391843604e-05, "loss": 0.0423, "step": 79100 }, { "epoch": 2.6745326075932248, "grad_norm": 0.29221367835998535, "learning_rate": 1.1464853915706503e-05, "loss": 0.0451, "step": 79110 }, { "epoch": 2.6748706852834783, "grad_norm": 0.19359874725341797, "learning_rate": 1.1454320201954626e-05, "loss": 0.045, "step": 79120 }, { "epoch": 2.6752087629737313, "grad_norm": 0.36220628023147583, "learning_rate": 1.1443790703500356e-05, "loss": 0.045, "step": 79130 }, { "epoch": 2.6755468406639844, "grad_norm": 0.2994817793369293, "learning_rate": 1.1433265421495154e-05, "loss": 0.0463, "step": 79140 }, { "epoch": 2.675884918354238, "grad_norm": 0.28033149242401123, "learning_rate": 1.1422744357090065e-05, "loss": 0.0489, "step": 79150 }, { "epoch": 2.676222996044491, "grad_norm": 0.4125634431838989, "learning_rate": 1.1412227511435647e-05, "loss": 0.0473, "step": 79160 }, { "epoch": 2.676561073734744, "grad_norm": 0.8365867733955383, "learning_rate": 1.1401714885682025e-05, "loss": 0.0441, "step": 79170 }, { "epoch": 2.6768991514249976, "grad_norm": 0.18414992094039917, "learning_rate": 1.1391206480978823e-05, "loss": 0.0439, "step": 79180 }, { "epoch": 2.6772372291152506, "grad_norm": 0.47984910011291504, "learning_rate": 1.1380702298475227e-05, "loss": 0.0433, "step": 79190 }, { "epoch": 2.6775753068055037, "grad_norm": 0.22464890778064728, "learning_rate": 1.1370202339319985e-05, "loss": 0.0458, "step": 79200 }, { "epoch": 2.677913384495757, "grad_norm": 0.3669357895851135, "learning_rate": 1.1359706604661319e-05, "loss": 0.0419, "step": 79210 }, { "epoch": 2.6782514621860103, "grad_norm": 0.21753068268299103, "learning_rate": 1.1349215095647042e-05, "loss": 0.036, "step": 79220 }, { "epoch": 2.6785895398762634, "grad_norm": 0.24864690005779266, "learning_rate": 1.1338727813424493e-05, "loss": 0.0463, "step": 79230 }, { "epoch": 2.678927617566517, "grad_norm": 0.19402912259101868, "learning_rate": 1.132824475914056e-05, "loss": 0.0411, "step": 79240 }, { "epoch": 2.67926569525677, "grad_norm": 0.2729549705982208, "learning_rate": 1.131776593394162e-05, "loss": 0.0471, "step": 79250 }, { "epoch": 2.679603772947023, "grad_norm": 0.15869735181331635, "learning_rate": 1.1307291338973657e-05, "loss": 0.0452, "step": 79260 }, { "epoch": 2.6799418506372765, "grad_norm": 0.22429236769676208, "learning_rate": 1.1296820975382121e-05, "loss": 0.0444, "step": 79270 }, { "epoch": 2.6802799283275296, "grad_norm": 0.16970382630825043, "learning_rate": 1.1286354844312074e-05, "loss": 0.0431, "step": 79280 }, { "epoch": 2.6806180060177827, "grad_norm": 0.2709791660308838, "learning_rate": 1.1275892946908023e-05, "loss": 0.0404, "step": 79290 }, { "epoch": 2.680956083708036, "grad_norm": 0.1968575417995453, "learning_rate": 1.126543528431413e-05, "loss": 0.0425, "step": 79300 }, { "epoch": 2.6812941613982892, "grad_norm": 0.24112877249717712, "learning_rate": 1.125498185767398e-05, "loss": 0.0405, "step": 79310 }, { "epoch": 2.6816322390885423, "grad_norm": 0.1516471654176712, "learning_rate": 1.1244532668130781e-05, "loss": 0.04, "step": 79320 }, { "epoch": 2.681970316778796, "grad_norm": 0.23579297959804535, "learning_rate": 1.1234087716827203e-05, "loss": 0.0482, "step": 79330 }, { "epoch": 2.682308394469049, "grad_norm": 0.24126072227954865, "learning_rate": 1.1223647004905524e-05, "loss": 0.0436, "step": 79340 }, { "epoch": 2.682646472159302, "grad_norm": 0.3964070677757263, "learning_rate": 1.1213210533507485e-05, "loss": 0.0473, "step": 79350 }, { "epoch": 2.6829845498495555, "grad_norm": 0.15482617914676666, "learning_rate": 1.1202778303774447e-05, "loss": 0.0438, "step": 79360 }, { "epoch": 2.6833226275398085, "grad_norm": 0.23578685522079468, "learning_rate": 1.1192350316847234e-05, "loss": 0.0396, "step": 79370 }, { "epoch": 2.6836607052300616, "grad_norm": 0.2441917210817337, "learning_rate": 1.1181926573866258e-05, "loss": 0.0439, "step": 79380 }, { "epoch": 2.683998782920315, "grad_norm": 0.29992759227752686, "learning_rate": 1.1171507075971416e-05, "loss": 0.0392, "step": 79390 }, { "epoch": 2.684336860610568, "grad_norm": 0.270175963640213, "learning_rate": 1.1161091824302177e-05, "loss": 0.0418, "step": 79400 }, { "epoch": 2.6846749383008213, "grad_norm": 0.19424985349178314, "learning_rate": 1.115068081999755e-05, "loss": 0.0436, "step": 79410 }, { "epoch": 2.685013015991075, "grad_norm": 0.22155195474624634, "learning_rate": 1.1140274064196071e-05, "loss": 0.0396, "step": 79420 }, { "epoch": 2.685351093681328, "grad_norm": 0.25989270210266113, "learning_rate": 1.112987155803578e-05, "loss": 0.0408, "step": 79430 }, { "epoch": 2.685689171371581, "grad_norm": 0.2572536766529083, "learning_rate": 1.1119473302654298e-05, "loss": 0.0437, "step": 79440 }, { "epoch": 2.6860272490618344, "grad_norm": 0.2566964030265808, "learning_rate": 1.1109079299188769e-05, "loss": 0.0447, "step": 79450 }, { "epoch": 2.6863653267520875, "grad_norm": 0.7792751789093018, "learning_rate": 1.1098689548775847e-05, "loss": 0.0465, "step": 79460 }, { "epoch": 2.686703404442341, "grad_norm": 0.2234724462032318, "learning_rate": 1.1088304052551762e-05, "loss": 0.0431, "step": 79470 }, { "epoch": 2.687041482132594, "grad_norm": 0.2115156203508377, "learning_rate": 1.1077922811652215e-05, "loss": 0.0436, "step": 79480 }, { "epoch": 2.687379559822847, "grad_norm": 0.22931882739067078, "learning_rate": 1.1067545827212534e-05, "loss": 0.0423, "step": 79490 }, { "epoch": 2.6877176375131007, "grad_norm": 0.31614115834236145, "learning_rate": 1.1057173100367496e-05, "loss": 0.0495, "step": 79500 }, { "epoch": 2.6880557152033537, "grad_norm": 0.29286590218544006, "learning_rate": 1.1046804632251474e-05, "loss": 0.0447, "step": 79510 }, { "epoch": 2.688393792893607, "grad_norm": 0.21395643055438995, "learning_rate": 1.1036440423998318e-05, "loss": 0.0446, "step": 79520 }, { "epoch": 2.6887318705838603, "grad_norm": 0.19277037680149078, "learning_rate": 1.1026080476741469e-05, "loss": 0.0471, "step": 79530 }, { "epoch": 2.6890699482741134, "grad_norm": 0.31424668431282043, "learning_rate": 1.1015724791613836e-05, "loss": 0.0462, "step": 79540 }, { "epoch": 2.6894080259643665, "grad_norm": 0.2673393487930298, "learning_rate": 1.1005373369747951e-05, "loss": 0.0455, "step": 79550 }, { "epoch": 2.68974610365462, "grad_norm": 0.1765664964914322, "learning_rate": 1.09950262122758e-05, "loss": 0.0443, "step": 79560 }, { "epoch": 2.690084181344873, "grad_norm": 0.3000929057598114, "learning_rate": 1.0984683320328948e-05, "loss": 0.0471, "step": 79570 }, { "epoch": 2.690422259035126, "grad_norm": 0.21066586673259735, "learning_rate": 1.0974344695038458e-05, "loss": 0.0394, "step": 79580 }, { "epoch": 2.6907603367253796, "grad_norm": 0.23075249791145325, "learning_rate": 1.0964010337534968e-05, "loss": 0.045, "step": 79590 }, { "epoch": 2.6910984144156327, "grad_norm": 0.20246285200119019, "learning_rate": 1.0953680248948611e-05, "loss": 0.0477, "step": 79600 }, { "epoch": 2.691436492105886, "grad_norm": 0.161606103181839, "learning_rate": 1.0943354430409069e-05, "loss": 0.0419, "step": 79610 }, { "epoch": 2.6917745697961393, "grad_norm": 0.21769413352012634, "learning_rate": 1.0933032883045574e-05, "loss": 0.0444, "step": 79620 }, { "epoch": 2.6921126474863923, "grad_norm": 0.36528193950653076, "learning_rate": 1.0922715607986872e-05, "loss": 0.0412, "step": 79630 }, { "epoch": 2.692450725176646, "grad_norm": 0.4200785160064697, "learning_rate": 1.0912402606361234e-05, "loss": 0.0497, "step": 79640 }, { "epoch": 2.692788802866899, "grad_norm": 0.2682165503501892, "learning_rate": 1.090209387929647e-05, "loss": 0.0432, "step": 79650 }, { "epoch": 2.693126880557152, "grad_norm": 0.3570471405982971, "learning_rate": 1.0891789427919957e-05, "loss": 0.0387, "step": 79660 }, { "epoch": 2.6934649582474055, "grad_norm": 0.4799223840236664, "learning_rate": 1.088148925335854e-05, "loss": 0.0463, "step": 79670 }, { "epoch": 2.6938030359376586, "grad_norm": 0.6573587656021118, "learning_rate": 1.087119335673864e-05, "loss": 0.0453, "step": 79680 }, { "epoch": 2.6941411136279116, "grad_norm": 0.17964495718479156, "learning_rate": 1.0860901739186208e-05, "loss": 0.0462, "step": 79690 }, { "epoch": 2.694479191318165, "grad_norm": 0.46783605217933655, "learning_rate": 1.085061440182673e-05, "loss": 0.042, "step": 79700 }, { "epoch": 2.6948172690084182, "grad_norm": 0.43944162130355835, "learning_rate": 1.0840331345785193e-05, "loss": 0.042, "step": 79710 }, { "epoch": 2.6951553466986713, "grad_norm": 0.2171030342578888, "learning_rate": 1.083005257218615e-05, "loss": 0.0436, "step": 79720 }, { "epoch": 2.695493424388925, "grad_norm": 0.31072551012039185, "learning_rate": 1.0819778082153642e-05, "loss": 0.0432, "step": 79730 }, { "epoch": 2.695831502079178, "grad_norm": 0.38285475969314575, "learning_rate": 1.0809507876811325e-05, "loss": 0.0429, "step": 79740 }, { "epoch": 2.696169579769431, "grad_norm": 0.2016444355249405, "learning_rate": 1.0799241957282292e-05, "loss": 0.0452, "step": 79750 }, { "epoch": 2.6965076574596845, "grad_norm": 0.20603516697883606, "learning_rate": 1.078898032468923e-05, "loss": 0.0469, "step": 79760 }, { "epoch": 2.6968457351499375, "grad_norm": 0.2137315571308136, "learning_rate": 1.0778722980154315e-05, "loss": 0.0463, "step": 79770 }, { "epoch": 2.6971838128401906, "grad_norm": 0.2809702455997467, "learning_rate": 1.0768469924799302e-05, "loss": 0.0437, "step": 79780 }, { "epoch": 2.697521890530444, "grad_norm": 0.2026291936635971, "learning_rate": 1.0758221159745419e-05, "loss": 0.0441, "step": 79790 }, { "epoch": 2.697859968220697, "grad_norm": 0.25312843918800354, "learning_rate": 1.074797668611347e-05, "loss": 0.0411, "step": 79800 }, { "epoch": 2.6981980459109502, "grad_norm": 0.1835792362689972, "learning_rate": 1.0737736505023777e-05, "loss": 0.0453, "step": 79810 }, { "epoch": 2.6985361236012038, "grad_norm": 0.306263267993927, "learning_rate": 1.0727500617596203e-05, "loss": 0.0429, "step": 79820 }, { "epoch": 2.698874201291457, "grad_norm": 0.1261250078678131, "learning_rate": 1.0717269024950104e-05, "loss": 0.0375, "step": 79830 }, { "epoch": 2.69921227898171, "grad_norm": 0.36536771059036255, "learning_rate": 1.070704172820442e-05, "loss": 0.0444, "step": 79840 }, { "epoch": 2.6995503566719634, "grad_norm": 0.19592587649822235, "learning_rate": 1.0696818728477559e-05, "loss": 0.0445, "step": 79850 }, { "epoch": 2.6998884343622165, "grad_norm": 0.43920662999153137, "learning_rate": 1.0686600026887511e-05, "loss": 0.0512, "step": 79860 }, { "epoch": 2.7002265120524696, "grad_norm": 0.4491542875766754, "learning_rate": 1.0676385624551777e-05, "loss": 0.0446, "step": 79870 }, { "epoch": 2.700564589742723, "grad_norm": 0.2499234825372696, "learning_rate": 1.0666175522587402e-05, "loss": 0.0438, "step": 79880 }, { "epoch": 2.700902667432976, "grad_norm": 0.17728528380393982, "learning_rate": 1.0655969722110925e-05, "loss": 0.0415, "step": 79890 }, { "epoch": 2.701240745123229, "grad_norm": 0.29554957151412964, "learning_rate": 1.0645768224238439e-05, "loss": 0.0443, "step": 79900 }, { "epoch": 2.7015788228134827, "grad_norm": 0.38618382811546326, "learning_rate": 1.063557103008559e-05, "loss": 0.0407, "step": 79910 }, { "epoch": 2.701916900503736, "grad_norm": 0.27474769949913025, "learning_rate": 1.0625378140767494e-05, "loss": 0.0485, "step": 79920 }, { "epoch": 2.702254978193989, "grad_norm": 0.17422766983509064, "learning_rate": 1.0615189557398846e-05, "loss": 0.0429, "step": 79930 }, { "epoch": 2.7025930558842424, "grad_norm": 0.3753844201564789, "learning_rate": 1.060500528109386e-05, "loss": 0.0442, "step": 79940 }, { "epoch": 2.7029311335744954, "grad_norm": 0.5429325103759766, "learning_rate": 1.0594825312966272e-05, "loss": 0.0446, "step": 79950 }, { "epoch": 2.7032692112647485, "grad_norm": 0.3659505546092987, "learning_rate": 1.0584649654129336e-05, "loss": 0.0449, "step": 79960 }, { "epoch": 2.703607288955002, "grad_norm": 0.22472655773162842, "learning_rate": 1.0574478305695863e-05, "loss": 0.0402, "step": 79970 }, { "epoch": 2.703945366645255, "grad_norm": 0.20976221561431885, "learning_rate": 1.0564311268778154e-05, "loss": 0.0468, "step": 79980 }, { "epoch": 2.704283444335508, "grad_norm": 0.39189064502716064, "learning_rate": 1.0554148544488073e-05, "loss": 0.0486, "step": 79990 }, { "epoch": 2.7046215220257617, "grad_norm": 0.2867249846458435, "learning_rate": 1.0543990133937004e-05, "loss": 0.0476, "step": 80000 }, { "epoch": 2.7049595997160147, "grad_norm": 0.16298900544643402, "learning_rate": 1.0533836038235868e-05, "loss": 0.0457, "step": 80010 }, { "epoch": 2.705297677406268, "grad_norm": 0.16189338266849518, "learning_rate": 1.052368625849507e-05, "loss": 0.0507, "step": 80020 }, { "epoch": 2.7056357550965213, "grad_norm": 0.2722153663635254, "learning_rate": 1.05135407958246e-05, "loss": 0.0468, "step": 80030 }, { "epoch": 2.7059738327867744, "grad_norm": 0.2734020948410034, "learning_rate": 1.0503399651333934e-05, "loss": 0.0463, "step": 80040 }, { "epoch": 2.7063119104770275, "grad_norm": 0.360198974609375, "learning_rate": 1.04932628261321e-05, "loss": 0.0421, "step": 80050 }, { "epoch": 2.706649988167281, "grad_norm": 0.3925951421260834, "learning_rate": 1.0483130321327644e-05, "loss": 0.0442, "step": 80060 }, { "epoch": 2.706988065857534, "grad_norm": 0.1904754340648651, "learning_rate": 1.0473002138028654e-05, "loss": 0.0458, "step": 80070 }, { "epoch": 2.707326143547787, "grad_norm": 0.22366346418857574, "learning_rate": 1.0462878277342713e-05, "loss": 0.0424, "step": 80080 }, { "epoch": 2.7076642212380406, "grad_norm": 0.3112306594848633, "learning_rate": 1.0452758740376972e-05, "loss": 0.0474, "step": 80090 }, { "epoch": 2.7080022989282937, "grad_norm": 0.24702240526676178, "learning_rate": 1.0442643528238067e-05, "loss": 0.045, "step": 80100 }, { "epoch": 2.7083403766185468, "grad_norm": 0.6887887716293335, "learning_rate": 1.0432532642032188e-05, "loss": 0.0466, "step": 80110 }, { "epoch": 2.7086784543088003, "grad_norm": 0.2019536942243576, "learning_rate": 1.042242608286505e-05, "loss": 0.0445, "step": 80120 }, { "epoch": 2.7090165319990533, "grad_norm": 0.6234080791473389, "learning_rate": 1.041232385184191e-05, "loss": 0.0466, "step": 80130 }, { "epoch": 2.7093546096893064, "grad_norm": 0.1597805768251419, "learning_rate": 1.0402225950067495e-05, "loss": 0.045, "step": 80140 }, { "epoch": 2.70969268737956, "grad_norm": 0.1868775486946106, "learning_rate": 1.0392132378646119e-05, "loss": 0.0497, "step": 80150 }, { "epoch": 2.710030765069813, "grad_norm": 0.17307689785957336, "learning_rate": 1.0382043138681607e-05, "loss": 0.0429, "step": 80160 }, { "epoch": 2.710368842760066, "grad_norm": 0.21474239230155945, "learning_rate": 1.0371958231277284e-05, "loss": 0.0412, "step": 80170 }, { "epoch": 2.7107069204503196, "grad_norm": 0.14751218259334564, "learning_rate": 1.0361877657536024e-05, "loss": 0.0368, "step": 80180 }, { "epoch": 2.7110449981405726, "grad_norm": 0.1776517778635025, "learning_rate": 1.035180141856023e-05, "loss": 0.0453, "step": 80190 }, { "epoch": 2.7113830758308257, "grad_norm": 0.14908573031425476, "learning_rate": 1.0341729515451836e-05, "loss": 0.0479, "step": 80200 }, { "epoch": 2.7117211535210792, "grad_norm": 0.28621992468833923, "learning_rate": 1.0331661949312265e-05, "loss": 0.0416, "step": 80210 }, { "epoch": 2.7120592312113323, "grad_norm": 0.20243380963802338, "learning_rate": 1.0321598721242514e-05, "loss": 0.0396, "step": 80220 }, { "epoch": 2.7123973089015854, "grad_norm": 0.26092925667762756, "learning_rate": 1.0311539832343064e-05, "loss": 0.0407, "step": 80230 }, { "epoch": 2.712735386591839, "grad_norm": 0.49986523389816284, "learning_rate": 1.0301485283713942e-05, "loss": 0.0441, "step": 80240 }, { "epoch": 2.713073464282092, "grad_norm": 0.16054843366146088, "learning_rate": 1.0291435076454703e-05, "loss": 0.0393, "step": 80250 }, { "epoch": 2.713411541972345, "grad_norm": 0.5446014404296875, "learning_rate": 1.028138921166444e-05, "loss": 0.0426, "step": 80260 }, { "epoch": 2.7137496196625985, "grad_norm": 0.19790595769882202, "learning_rate": 1.0271347690441719e-05, "loss": 0.0378, "step": 80270 }, { "epoch": 2.7140876973528516, "grad_norm": 0.21631106734275818, "learning_rate": 1.0261310513884698e-05, "loss": 0.0427, "step": 80280 }, { "epoch": 2.7144257750431047, "grad_norm": 0.2281777411699295, "learning_rate": 1.0251277683090998e-05, "loss": 0.0432, "step": 80290 }, { "epoch": 2.714763852733358, "grad_norm": 0.29577165842056274, "learning_rate": 1.0241249199157826e-05, "loss": 0.0456, "step": 80300 }, { "epoch": 2.7151019304236113, "grad_norm": 0.22442717850208282, "learning_rate": 1.0231225063181837e-05, "loss": 0.0439, "step": 80310 }, { "epoch": 2.7154400081138643, "grad_norm": 0.6584829092025757, "learning_rate": 1.0221205276259316e-05, "loss": 0.0431, "step": 80320 }, { "epoch": 2.715778085804118, "grad_norm": 0.2313523292541504, "learning_rate": 1.0211189839485958e-05, "loss": 0.0388, "step": 80330 }, { "epoch": 2.716116163494371, "grad_norm": 0.36675402522087097, "learning_rate": 1.0201178753957075e-05, "loss": 0.0439, "step": 80340 }, { "epoch": 2.716454241184624, "grad_norm": 0.14403192698955536, "learning_rate": 1.0191172020767432e-05, "loss": 0.0436, "step": 80350 }, { "epoch": 2.7167923188748775, "grad_norm": 0.36626484990119934, "learning_rate": 1.0181169641011362e-05, "loss": 0.0438, "step": 80360 }, { "epoch": 2.7171303965651306, "grad_norm": 0.22545917332172394, "learning_rate": 1.0171171615782721e-05, "loss": 0.0416, "step": 80370 }, { "epoch": 2.7174684742553836, "grad_norm": 0.30599385499954224, "learning_rate": 1.0161177946174866e-05, "loss": 0.0482, "step": 80380 }, { "epoch": 2.717806551945637, "grad_norm": 0.1735605150461197, "learning_rate": 1.0151188633280706e-05, "loss": 0.0378, "step": 80390 }, { "epoch": 2.71814462963589, "grad_norm": 0.5344139933586121, "learning_rate": 1.0141203678192635e-05, "loss": 0.0449, "step": 80400 }, { "epoch": 2.7184827073261433, "grad_norm": 0.2596575915813446, "learning_rate": 1.0131223082002612e-05, "loss": 0.0444, "step": 80410 }, { "epoch": 2.718820785016397, "grad_norm": 0.32804644107818604, "learning_rate": 1.012124684580208e-05, "loss": 0.0484, "step": 80420 }, { "epoch": 2.71915886270665, "grad_norm": 0.21404053270816803, "learning_rate": 1.0111274970682039e-05, "loss": 0.043, "step": 80430 }, { "epoch": 2.7194969403969034, "grad_norm": 0.33621037006378174, "learning_rate": 1.010130745773299e-05, "loss": 0.0453, "step": 80440 }, { "epoch": 2.7198350180871564, "grad_norm": 0.25166308879852295, "learning_rate": 1.0091344308044986e-05, "loss": 0.0478, "step": 80450 }, { "epoch": 2.7201730957774095, "grad_norm": 0.1888912171125412, "learning_rate": 1.0081385522707554e-05, "loss": 0.038, "step": 80460 }, { "epoch": 2.720511173467663, "grad_norm": 0.35078415274620056, "learning_rate": 1.0071431102809803e-05, "loss": 0.0411, "step": 80470 }, { "epoch": 2.720849251157916, "grad_norm": 0.26355665922164917, "learning_rate": 1.0061481049440297e-05, "loss": 0.042, "step": 80480 }, { "epoch": 2.721187328848169, "grad_norm": 0.254099577665329, "learning_rate": 1.0051535363687187e-05, "loss": 0.0446, "step": 80490 }, { "epoch": 2.7215254065384227, "grad_norm": 0.2001776099205017, "learning_rate": 1.0041594046638087e-05, "loss": 0.0397, "step": 80500 }, { "epoch": 2.7218634842286757, "grad_norm": 0.2594408392906189, "learning_rate": 1.0031657099380204e-05, "loss": 0.042, "step": 80510 }, { "epoch": 2.722201561918929, "grad_norm": 0.19033221900463104, "learning_rate": 1.0021724523000204e-05, "loss": 0.0456, "step": 80520 }, { "epoch": 2.7225396396091823, "grad_norm": 0.564913272857666, "learning_rate": 1.001179631858431e-05, "loss": 0.0418, "step": 80530 }, { "epoch": 2.7228777172994354, "grad_norm": 0.3853836953639984, "learning_rate": 1.0001872487218245e-05, "loss": 0.04, "step": 80540 }, { "epoch": 2.7232157949896885, "grad_norm": 0.32962697744369507, "learning_rate": 9.991953029987272e-06, "loss": 0.0432, "step": 80550 }, { "epoch": 2.723553872679942, "grad_norm": 0.2968129813671112, "learning_rate": 9.982037947976147e-06, "loss": 0.0416, "step": 80560 }, { "epoch": 2.723891950370195, "grad_norm": 0.6023221015930176, "learning_rate": 9.972127242269208e-06, "loss": 0.041, "step": 80570 }, { "epoch": 2.7242300280604486, "grad_norm": 0.24868938326835632, "learning_rate": 9.962220913950238e-06, "loss": 0.0476, "step": 80580 }, { "epoch": 2.7245681057507016, "grad_norm": 0.37414324283599854, "learning_rate": 9.952318964102591e-06, "loss": 0.0459, "step": 80590 }, { "epoch": 2.7249061834409547, "grad_norm": 0.14908793568611145, "learning_rate": 9.942421393809148e-06, "loss": 0.0397, "step": 80600 }, { "epoch": 2.725244261131208, "grad_norm": 0.17207714915275574, "learning_rate": 9.932528204152264e-06, "loss": 0.036, "step": 80610 }, { "epoch": 2.7255823388214613, "grad_norm": 0.658112108707428, "learning_rate": 9.922639396213856e-06, "loss": 0.0476, "step": 80620 }, { "epoch": 2.7259204165117144, "grad_norm": 0.16129635274410248, "learning_rate": 9.912754971075344e-06, "loss": 0.0441, "step": 80630 }, { "epoch": 2.726258494201968, "grad_norm": 0.16743725538253784, "learning_rate": 9.902874929817696e-06, "loss": 0.0436, "step": 80640 }, { "epoch": 2.726596571892221, "grad_norm": 0.2387302666902542, "learning_rate": 9.892999273521342e-06, "loss": 0.0408, "step": 80650 }, { "epoch": 2.726934649582474, "grad_norm": 0.2221618890762329, "learning_rate": 9.883128003266302e-06, "loss": 0.0396, "step": 80660 }, { "epoch": 2.7272727272727275, "grad_norm": 0.31096065044403076, "learning_rate": 9.87326112013206e-06, "loss": 0.0462, "step": 80670 }, { "epoch": 2.7276108049629806, "grad_norm": 0.129270538687706, "learning_rate": 9.863398625197662e-06, "loss": 0.0361, "step": 80680 }, { "epoch": 2.7279488826532337, "grad_norm": 0.18884684145450592, "learning_rate": 9.853540519541615e-06, "loss": 0.0488, "step": 80690 }, { "epoch": 2.728286960343487, "grad_norm": 0.2388160526752472, "learning_rate": 9.843686804242053e-06, "loss": 0.0429, "step": 80700 }, { "epoch": 2.7286250380337402, "grad_norm": 0.1786745935678482, "learning_rate": 9.833837480376506e-06, "loss": 0.0444, "step": 80710 }, { "epoch": 2.7289631157239933, "grad_norm": 0.27530544996261597, "learning_rate": 9.82399254902212e-06, "loss": 0.0393, "step": 80720 }, { "epoch": 2.729301193414247, "grad_norm": 0.2949689030647278, "learning_rate": 9.814152011255494e-06, "loss": 0.0447, "step": 80730 }, { "epoch": 2.7296392711045, "grad_norm": 0.3365923762321472, "learning_rate": 9.804315868152797e-06, "loss": 0.0395, "step": 80740 }, { "epoch": 2.729977348794753, "grad_norm": 0.18603278696537018, "learning_rate": 9.79448412078966e-06, "loss": 0.0417, "step": 80750 }, { "epoch": 2.7303154264850065, "grad_norm": 0.27876943349838257, "learning_rate": 9.784656770241312e-06, "loss": 0.0408, "step": 80760 }, { "epoch": 2.7306535041752595, "grad_norm": 0.17479228973388672, "learning_rate": 9.77483381758243e-06, "loss": 0.0371, "step": 80770 }, { "epoch": 2.7309915818655126, "grad_norm": 0.3012482821941376, "learning_rate": 9.76501526388725e-06, "loss": 0.0417, "step": 80780 }, { "epoch": 2.731329659555766, "grad_norm": 0.25833630561828613, "learning_rate": 9.7552011102295e-06, "loss": 0.0392, "step": 80790 }, { "epoch": 2.731667737246019, "grad_norm": 0.1466396152973175, "learning_rate": 9.745391357682449e-06, "loss": 0.044, "step": 80800 }, { "epoch": 2.7320058149362723, "grad_norm": 0.2679176926612854, "learning_rate": 9.735586007318875e-06, "loss": 0.0449, "step": 80810 }, { "epoch": 2.7323438926265258, "grad_norm": 0.2220270037651062, "learning_rate": 9.725785060211096e-06, "loss": 0.0408, "step": 80820 }, { "epoch": 2.732681970316779, "grad_norm": 0.22378294169902802, "learning_rate": 9.715988517430896e-06, "loss": 0.0384, "step": 80830 }, { "epoch": 2.733020048007032, "grad_norm": 0.2681393027305603, "learning_rate": 9.706196380049625e-06, "loss": 0.0446, "step": 80840 }, { "epoch": 2.7333581256972854, "grad_norm": 0.35011398792266846, "learning_rate": 9.69640864913815e-06, "loss": 0.0466, "step": 80850 }, { "epoch": 2.7336962033875385, "grad_norm": 0.33504799008369446, "learning_rate": 9.686625325766818e-06, "loss": 0.0432, "step": 80860 }, { "epoch": 2.7340342810777916, "grad_norm": 0.43274444341659546, "learning_rate": 9.67684641100554e-06, "loss": 0.0413, "step": 80870 }, { "epoch": 2.734372358768045, "grad_norm": 0.2699003219604492, "learning_rate": 9.667071905923692e-06, "loss": 0.0408, "step": 80880 }, { "epoch": 2.734710436458298, "grad_norm": 0.21000167727470398, "learning_rate": 9.657301811590242e-06, "loss": 0.0467, "step": 80890 }, { "epoch": 2.735048514148551, "grad_norm": 0.17582398653030396, "learning_rate": 9.647536129073603e-06, "loss": 0.039, "step": 80900 }, { "epoch": 2.7353865918388047, "grad_norm": 0.2874990999698639, "learning_rate": 9.637774859441751e-06, "loss": 0.0432, "step": 80910 }, { "epoch": 2.735724669529058, "grad_norm": 0.29671046137809753, "learning_rate": 9.628018003762152e-06, "loss": 0.0489, "step": 80920 }, { "epoch": 2.736062747219311, "grad_norm": 0.27211472392082214, "learning_rate": 9.618265563101813e-06, "loss": 0.0428, "step": 80930 }, { "epoch": 2.7364008249095644, "grad_norm": 0.1595434695482254, "learning_rate": 9.608517538527211e-06, "loss": 0.0417, "step": 80940 }, { "epoch": 2.7367389025998174, "grad_norm": 0.18616430461406708, "learning_rate": 9.598773931104431e-06, "loss": 0.043, "step": 80950 }, { "epoch": 2.7370769802900705, "grad_norm": 0.22547535598278046, "learning_rate": 9.589034741898983e-06, "loss": 0.0429, "step": 80960 }, { "epoch": 2.737415057980324, "grad_norm": 0.2030317634344101, "learning_rate": 9.579299971975947e-06, "loss": 0.0455, "step": 80970 }, { "epoch": 2.737753135670577, "grad_norm": 0.25925368070602417, "learning_rate": 9.569569622399882e-06, "loss": 0.0399, "step": 80980 }, { "epoch": 2.73809121336083, "grad_norm": 0.3816986083984375, "learning_rate": 9.559843694234916e-06, "loss": 0.044, "step": 80990 }, { "epoch": 2.7384292910510837, "grad_norm": 0.3504936397075653, "learning_rate": 9.550122188544625e-06, "loss": 0.0436, "step": 81000 }, { "epoch": 2.7387673687413367, "grad_norm": 0.30082929134368896, "learning_rate": 9.540405106392152e-06, "loss": 0.0448, "step": 81010 }, { "epoch": 2.73910544643159, "grad_norm": 0.3533640503883362, "learning_rate": 9.53069244884015e-06, "loss": 0.0478, "step": 81020 }, { "epoch": 2.7394435241218433, "grad_norm": 0.126921147108078, "learning_rate": 9.52098421695079e-06, "loss": 0.0409, "step": 81030 }, { "epoch": 2.7397816018120964, "grad_norm": 0.3357575535774231, "learning_rate": 9.511280411785728e-06, "loss": 0.0409, "step": 81040 }, { "epoch": 2.7401196795023495, "grad_norm": 0.19838447868824005, "learning_rate": 9.50158103440616e-06, "loss": 0.0458, "step": 81050 }, { "epoch": 2.740457757192603, "grad_norm": 0.37960708141326904, "learning_rate": 9.491886085872815e-06, "loss": 0.046, "step": 81060 }, { "epoch": 2.740795834882856, "grad_norm": 0.21143589913845062, "learning_rate": 9.482195567245895e-06, "loss": 0.0495, "step": 81070 }, { "epoch": 2.741133912573109, "grad_norm": 0.3963867723941803, "learning_rate": 9.472509479585151e-06, "loss": 0.0418, "step": 81080 }, { "epoch": 2.7414719902633626, "grad_norm": 0.15250703692436218, "learning_rate": 9.462827823949833e-06, "loss": 0.043, "step": 81090 }, { "epoch": 2.7418100679536157, "grad_norm": 0.2751472294330597, "learning_rate": 9.453150601398735e-06, "loss": 0.0402, "step": 81100 }, { "epoch": 2.7421481456438688, "grad_norm": 0.2903114855289459, "learning_rate": 9.44347781299011e-06, "loss": 0.04, "step": 81110 }, { "epoch": 2.7424862233341223, "grad_norm": 0.29178258776664734, "learning_rate": 9.43380945978179e-06, "loss": 0.0412, "step": 81120 }, { "epoch": 2.7428243010243754, "grad_norm": 0.25229936838150024, "learning_rate": 9.424145542831047e-06, "loss": 0.0433, "step": 81130 }, { "epoch": 2.7431623787146284, "grad_norm": 0.15033014118671417, "learning_rate": 9.414486063194778e-06, "loss": 0.044, "step": 81140 }, { "epoch": 2.743500456404882, "grad_norm": 0.526412308216095, "learning_rate": 9.404831021929273e-06, "loss": 0.0427, "step": 81150 }, { "epoch": 2.743838534095135, "grad_norm": 0.23543108999729156, "learning_rate": 9.395180420090427e-06, "loss": 0.0459, "step": 81160 }, { "epoch": 2.744176611785388, "grad_norm": 0.4335332214832306, "learning_rate": 9.385534258733591e-06, "loss": 0.0456, "step": 81170 }, { "epoch": 2.7445146894756416, "grad_norm": 0.281642884016037, "learning_rate": 9.375892538913677e-06, "loss": 0.0391, "step": 81180 }, { "epoch": 2.7448527671658947, "grad_norm": 0.2354382872581482, "learning_rate": 9.36625526168507e-06, "loss": 0.0433, "step": 81190 }, { "epoch": 2.7451908448561477, "grad_norm": 0.28669556975364685, "learning_rate": 9.356622428101686e-06, "loss": 0.0415, "step": 81200 }, { "epoch": 2.7455289225464012, "grad_norm": 0.25480303168296814, "learning_rate": 9.346994039216972e-06, "loss": 0.0473, "step": 81210 }, { "epoch": 2.7458670002366543, "grad_norm": 0.282235324382782, "learning_rate": 9.337370096083886e-06, "loss": 0.041, "step": 81220 }, { "epoch": 2.7462050779269074, "grad_norm": 0.3467077314853668, "learning_rate": 9.327750599754854e-06, "loss": 0.0495, "step": 81230 }, { "epoch": 2.746543155617161, "grad_norm": 0.17766565084457397, "learning_rate": 9.318135551281875e-06, "loss": 0.0397, "step": 81240 }, { "epoch": 2.746881233307414, "grad_norm": 0.350238561630249, "learning_rate": 9.308524951716424e-06, "loss": 0.0425, "step": 81250 }, { "epoch": 2.747219310997667, "grad_norm": 0.1842733770608902, "learning_rate": 9.298918802109497e-06, "loss": 0.0435, "step": 81260 }, { "epoch": 2.7475573886879205, "grad_norm": 0.20329609513282776, "learning_rate": 9.289317103511625e-06, "loss": 0.0442, "step": 81270 }, { "epoch": 2.7478954663781736, "grad_norm": 0.24559509754180908, "learning_rate": 9.279719856972835e-06, "loss": 0.042, "step": 81280 }, { "epoch": 2.7482335440684267, "grad_norm": 0.32457610964775085, "learning_rate": 9.270127063542645e-06, "loss": 0.0423, "step": 81290 }, { "epoch": 2.74857162175868, "grad_norm": 0.25511103868484497, "learning_rate": 9.260538724270124e-06, "loss": 0.0392, "step": 81300 }, { "epoch": 2.7489096994489333, "grad_norm": 0.2448628842830658, "learning_rate": 9.250954840203852e-06, "loss": 0.0449, "step": 81310 }, { "epoch": 2.7492477771391863, "grad_norm": 0.24788086116313934, "learning_rate": 9.241375412391878e-06, "loss": 0.0399, "step": 81320 }, { "epoch": 2.74958585482944, "grad_norm": 0.22846215963363647, "learning_rate": 9.23180044188181e-06, "loss": 0.0441, "step": 81330 }, { "epoch": 2.749923932519693, "grad_norm": 0.21823759377002716, "learning_rate": 9.222229929720755e-06, "loss": 0.0451, "step": 81340 }, { "epoch": 2.750262010209946, "grad_norm": 0.23660874366760254, "learning_rate": 9.212663876955336e-06, "loss": 0.0444, "step": 81350 }, { "epoch": 2.7506000879001995, "grad_norm": 0.28710147738456726, "learning_rate": 9.20310228463166e-06, "loss": 0.0465, "step": 81360 }, { "epoch": 2.7509381655904526, "grad_norm": 0.41678982973098755, "learning_rate": 9.19354515379539e-06, "loss": 0.0481, "step": 81370 }, { "epoch": 2.7512762432807056, "grad_norm": 0.2359856218099594, "learning_rate": 9.183992485491666e-06, "loss": 0.0406, "step": 81380 }, { "epoch": 2.751614320970959, "grad_norm": 0.18820282816886902, "learning_rate": 9.174444280765154e-06, "loss": 0.0448, "step": 81390 }, { "epoch": 2.751952398661212, "grad_norm": 0.4240472912788391, "learning_rate": 9.164900540660032e-06, "loss": 0.0475, "step": 81400 }, { "epoch": 2.7522904763514657, "grad_norm": 0.15582230687141418, "learning_rate": 9.155361266220008e-06, "loss": 0.0408, "step": 81410 }, { "epoch": 2.752628554041719, "grad_norm": 0.19680540263652802, "learning_rate": 9.145826458488254e-06, "loss": 0.0433, "step": 81420 }, { "epoch": 2.752966631731972, "grad_norm": 0.34032967686653137, "learning_rate": 9.136296118507509e-06, "loss": 0.0416, "step": 81430 }, { "epoch": 2.7533047094222254, "grad_norm": 0.2486681044101715, "learning_rate": 9.126770247319966e-06, "loss": 0.0433, "step": 81440 }, { "epoch": 2.7536427871124785, "grad_norm": 0.1905098706483841, "learning_rate": 9.11724884596738e-06, "loss": 0.0438, "step": 81450 }, { "epoch": 2.7539808648027315, "grad_norm": 0.18923622369766235, "learning_rate": 9.107731915490996e-06, "loss": 0.0407, "step": 81460 }, { "epoch": 2.754318942492985, "grad_norm": 0.28110450506210327, "learning_rate": 9.098219456931579e-06, "loss": 0.0427, "step": 81470 }, { "epoch": 2.754657020183238, "grad_norm": 0.2226594090461731, "learning_rate": 9.088711471329376e-06, "loss": 0.0445, "step": 81480 }, { "epoch": 2.754995097873491, "grad_norm": 0.20560623705387115, "learning_rate": 9.079207959724184e-06, "loss": 0.0383, "step": 81490 }, { "epoch": 2.7553331755637447, "grad_norm": 0.2874965965747833, "learning_rate": 9.069708923155279e-06, "loss": 0.0451, "step": 81500 }, { "epoch": 2.7556712532539978, "grad_norm": 0.3519202172756195, "learning_rate": 9.060214362661468e-06, "loss": 0.0433, "step": 81510 }, { "epoch": 2.756009330944251, "grad_norm": 0.25860920548439026, "learning_rate": 9.050724279281064e-06, "loss": 0.0428, "step": 81520 }, { "epoch": 2.7563474086345043, "grad_norm": 0.22374653816223145, "learning_rate": 9.041238674051894e-06, "loss": 0.0402, "step": 81530 }, { "epoch": 2.7566854863247574, "grad_norm": 0.16075824201107025, "learning_rate": 9.031757548011277e-06, "loss": 0.0378, "step": 81540 }, { "epoch": 2.757023564015011, "grad_norm": 0.236663818359375, "learning_rate": 9.022280902196051e-06, "loss": 0.0442, "step": 81550 }, { "epoch": 2.757361641705264, "grad_norm": 0.35202503204345703, "learning_rate": 9.012808737642592e-06, "loss": 0.0388, "step": 81560 }, { "epoch": 2.757699719395517, "grad_norm": 0.27057120203971863, "learning_rate": 9.003341055386739e-06, "loss": 0.0431, "step": 81570 }, { "epoch": 2.7580377970857706, "grad_norm": 0.24095921218395233, "learning_rate": 8.993877856463862e-06, "loss": 0.0433, "step": 81580 }, { "epoch": 2.7583758747760236, "grad_norm": 0.1998460739850998, "learning_rate": 8.984419141908856e-06, "loss": 0.0415, "step": 81590 }, { "epoch": 2.7587139524662767, "grad_norm": 0.23707452416419983, "learning_rate": 8.974964912756117e-06, "loss": 0.0423, "step": 81600 }, { "epoch": 2.75905203015653, "grad_norm": 0.20815999805927277, "learning_rate": 8.965515170039523e-06, "loss": 0.0451, "step": 81610 }, { "epoch": 2.7593901078467833, "grad_norm": 0.17625021934509277, "learning_rate": 8.956069914792514e-06, "loss": 0.047, "step": 81620 }, { "epoch": 2.7597281855370364, "grad_norm": 0.2390134334564209, "learning_rate": 8.946629148047969e-06, "loss": 0.0397, "step": 81630 }, { "epoch": 2.76006626322729, "grad_norm": 0.19694560766220093, "learning_rate": 8.93719287083834e-06, "loss": 0.0454, "step": 81640 }, { "epoch": 2.760404340917543, "grad_norm": 0.19435790181159973, "learning_rate": 8.927761084195569e-06, "loss": 0.0394, "step": 81650 }, { "epoch": 2.760742418607796, "grad_norm": 0.2900277376174927, "learning_rate": 8.918333789151096e-06, "loss": 0.0416, "step": 81660 }, { "epoch": 2.7610804962980495, "grad_norm": 0.19911602139472961, "learning_rate": 8.90891098673587e-06, "loss": 0.0404, "step": 81670 }, { "epoch": 2.7614185739883026, "grad_norm": 0.19686581194400787, "learning_rate": 8.89949267798037e-06, "loss": 0.0404, "step": 81680 }, { "epoch": 2.7617566516785557, "grad_norm": 0.21232084929943085, "learning_rate": 8.890078863914542e-06, "loss": 0.0387, "step": 81690 }, { "epoch": 2.762094729368809, "grad_norm": 0.3496401906013489, "learning_rate": 8.880669545567894e-06, "loss": 0.0438, "step": 81700 }, { "epoch": 2.7624328070590622, "grad_norm": 0.3612663149833679, "learning_rate": 8.871264723969381e-06, "loss": 0.0413, "step": 81710 }, { "epoch": 2.7627708847493153, "grad_norm": 0.23245245218276978, "learning_rate": 8.86186440014754e-06, "loss": 0.0407, "step": 81720 }, { "epoch": 2.763108962439569, "grad_norm": 0.361293762922287, "learning_rate": 8.852468575130351e-06, "loss": 0.0442, "step": 81730 }, { "epoch": 2.763447040129822, "grad_norm": 0.21221153438091278, "learning_rate": 8.843077249945347e-06, "loss": 0.0407, "step": 81740 }, { "epoch": 2.763785117820075, "grad_norm": 0.23476144671440125, "learning_rate": 8.833690425619523e-06, "loss": 0.0412, "step": 81750 }, { "epoch": 2.7641231955103285, "grad_norm": 0.25305503606796265, "learning_rate": 8.82430810317942e-06, "loss": 0.0421, "step": 81760 }, { "epoch": 2.7644612732005815, "grad_norm": 0.21199707686901093, "learning_rate": 8.814930283651074e-06, "loss": 0.0413, "step": 81770 }, { "epoch": 2.7647993508908346, "grad_norm": 0.2457672506570816, "learning_rate": 8.805556968060048e-06, "loss": 0.0435, "step": 81780 }, { "epoch": 2.765137428581088, "grad_norm": 0.35423219203948975, "learning_rate": 8.796188157431362e-06, "loss": 0.0438, "step": 81790 }, { "epoch": 2.765475506271341, "grad_norm": 0.26479262113571167, "learning_rate": 8.786823852789595e-06, "loss": 0.0475, "step": 81800 }, { "epoch": 2.7658135839615943, "grad_norm": 0.28431108593940735, "learning_rate": 8.777464055158813e-06, "loss": 0.0417, "step": 81810 }, { "epoch": 2.766151661651848, "grad_norm": 0.17952820658683777, "learning_rate": 8.768108765562578e-06, "loss": 0.0382, "step": 81820 }, { "epoch": 2.766489739342101, "grad_norm": 0.18488048017024994, "learning_rate": 8.758757985023975e-06, "loss": 0.0388, "step": 81830 }, { "epoch": 2.766827817032354, "grad_norm": 0.41263413429260254, "learning_rate": 8.749411714565598e-06, "loss": 0.0438, "step": 81840 }, { "epoch": 2.7671658947226074, "grad_norm": 0.20829610526561737, "learning_rate": 8.74006995520954e-06, "loss": 0.0436, "step": 81850 }, { "epoch": 2.7675039724128605, "grad_norm": 0.25130516290664673, "learning_rate": 8.730732707977397e-06, "loss": 0.0425, "step": 81860 }, { "epoch": 2.7678420501031136, "grad_norm": 0.17329376935958862, "learning_rate": 8.721399973890281e-06, "loss": 0.0453, "step": 81870 }, { "epoch": 2.768180127793367, "grad_norm": 0.2434372752904892, "learning_rate": 8.712071753968792e-06, "loss": 0.0444, "step": 81880 }, { "epoch": 2.76851820548362, "grad_norm": 0.2125571072101593, "learning_rate": 8.702748049233072e-06, "loss": 0.0456, "step": 81890 }, { "epoch": 2.7688562831738732, "grad_norm": 0.238667294383049, "learning_rate": 8.693428860702718e-06, "loss": 0.0396, "step": 81900 }, { "epoch": 2.7691943608641267, "grad_norm": 0.2762479782104492, "learning_rate": 8.684114189396898e-06, "loss": 0.0476, "step": 81910 }, { "epoch": 2.76953243855438, "grad_norm": 0.22329023480415344, "learning_rate": 8.674804036334222e-06, "loss": 0.0458, "step": 81920 }, { "epoch": 2.769870516244633, "grad_norm": 0.14517244696617126, "learning_rate": 8.665498402532857e-06, "loss": 0.037, "step": 81930 }, { "epoch": 2.7702085939348864, "grad_norm": 0.12456519901752472, "learning_rate": 8.656197289010432e-06, "loss": 0.044, "step": 81940 }, { "epoch": 2.7705466716251395, "grad_norm": 0.25476500391960144, "learning_rate": 8.646900696784122e-06, "loss": 0.0442, "step": 81950 }, { "epoch": 2.7708847493153925, "grad_norm": 0.2058236002922058, "learning_rate": 8.637608626870553e-06, "loss": 0.0418, "step": 81960 }, { "epoch": 2.771222827005646, "grad_norm": 0.22934024035930634, "learning_rate": 8.628321080285944e-06, "loss": 0.0401, "step": 81970 }, { "epoch": 2.771560904695899, "grad_norm": 0.28243544697761536, "learning_rate": 8.619038058045925e-06, "loss": 0.0365, "step": 81980 }, { "epoch": 2.771898982386152, "grad_norm": 0.1803734302520752, "learning_rate": 8.609759561165704e-06, "loss": 0.0446, "step": 81990 }, { "epoch": 2.7722370600764057, "grad_norm": 1.5539296865463257, "learning_rate": 8.600485590659934e-06, "loss": 0.0441, "step": 82000 }, { "epoch": 2.7725751377666588, "grad_norm": 0.16535674035549164, "learning_rate": 8.591216147542813e-06, "loss": 0.0459, "step": 82010 }, { "epoch": 2.772913215456912, "grad_norm": 0.21118056774139404, "learning_rate": 8.581951232828045e-06, "loss": 0.0488, "step": 82020 }, { "epoch": 2.7732512931471653, "grad_norm": 0.2597762644290924, "learning_rate": 8.572690847528825e-06, "loss": 0.0433, "step": 82030 }, { "epoch": 2.7735893708374184, "grad_norm": 0.3130304515361786, "learning_rate": 8.563434992657838e-06, "loss": 0.0431, "step": 82040 }, { "epoch": 2.7739274485276715, "grad_norm": 0.24373593926429749, "learning_rate": 8.554183669227307e-06, "loss": 0.0459, "step": 82050 }, { "epoch": 2.774265526217925, "grad_norm": 0.19057314097881317, "learning_rate": 8.544936878248937e-06, "loss": 0.0434, "step": 82060 }, { "epoch": 2.774603603908178, "grad_norm": 0.23078899085521698, "learning_rate": 8.535694620733942e-06, "loss": 0.0421, "step": 82070 }, { "epoch": 2.774941681598431, "grad_norm": 0.18950937688350677, "learning_rate": 8.526456897693053e-06, "loss": 0.0439, "step": 82080 }, { "epoch": 2.7752797592886846, "grad_norm": 0.2714363932609558, "learning_rate": 8.517223710136456e-06, "loss": 0.0432, "step": 82090 }, { "epoch": 2.7756178369789377, "grad_norm": 0.22285550832748413, "learning_rate": 8.507995059073931e-06, "loss": 0.0413, "step": 82100 }, { "epoch": 2.775955914669191, "grad_norm": 0.21759441494941711, "learning_rate": 8.498770945514673e-06, "loss": 0.0464, "step": 82110 }, { "epoch": 2.7762939923594443, "grad_norm": 0.2631154954433441, "learning_rate": 8.489551370467436e-06, "loss": 0.0418, "step": 82120 }, { "epoch": 2.7766320700496974, "grad_norm": 0.8363839387893677, "learning_rate": 8.480336334940442e-06, "loss": 0.0465, "step": 82130 }, { "epoch": 2.7769701477399504, "grad_norm": 0.2063169777393341, "learning_rate": 8.471125839941458e-06, "loss": 0.0477, "step": 82140 }, { "epoch": 2.777308225430204, "grad_norm": 0.23962117731571198, "learning_rate": 8.461919886477682e-06, "loss": 0.0443, "step": 82150 }, { "epoch": 2.777646303120457, "grad_norm": 0.31417718529701233, "learning_rate": 8.452718475555927e-06, "loss": 0.0448, "step": 82160 }, { "epoch": 2.77798438081071, "grad_norm": 0.22070562839508057, "learning_rate": 8.443521608182398e-06, "loss": 0.0428, "step": 82170 }, { "epoch": 2.7783224585009636, "grad_norm": 0.17786425352096558, "learning_rate": 8.43432928536288e-06, "loss": 0.0445, "step": 82180 }, { "epoch": 2.7786605361912167, "grad_norm": 0.34418344497680664, "learning_rate": 8.425141508102607e-06, "loss": 0.0494, "step": 82190 }, { "epoch": 2.7789986138814697, "grad_norm": 0.23760949075222015, "learning_rate": 8.415958277406365e-06, "loss": 0.0403, "step": 82200 }, { "epoch": 2.7793366915717233, "grad_norm": 0.14757569134235382, "learning_rate": 8.406779594278392e-06, "loss": 0.0393, "step": 82210 }, { "epoch": 2.7796747692619763, "grad_norm": 0.4894288182258606, "learning_rate": 8.397605459722468e-06, "loss": 0.045, "step": 82220 }, { "epoch": 2.7800128469522294, "grad_norm": 0.21836641430854797, "learning_rate": 8.388435874741863e-06, "loss": 0.0409, "step": 82230 }, { "epoch": 2.780350924642483, "grad_norm": 0.45299482345581055, "learning_rate": 8.37927084033936e-06, "loss": 0.0455, "step": 82240 }, { "epoch": 2.780689002332736, "grad_norm": 0.24257586896419525, "learning_rate": 8.370110357517208e-06, "loss": 0.0522, "step": 82250 }, { "epoch": 2.781027080022989, "grad_norm": 0.19453777372837067, "learning_rate": 8.360954427277201e-06, "loss": 0.0492, "step": 82260 }, { "epoch": 2.7813651577132426, "grad_norm": 0.19001491367816925, "learning_rate": 8.351803050620627e-06, "loss": 0.0431, "step": 82270 }, { "epoch": 2.7817032354034956, "grad_norm": 0.14851437509059906, "learning_rate": 8.34265622854824e-06, "loss": 0.0411, "step": 82280 }, { "epoch": 2.7820413130937487, "grad_norm": 0.2270556539297104, "learning_rate": 8.333513962060335e-06, "loss": 0.0386, "step": 82290 }, { "epoch": 2.782379390784002, "grad_norm": 0.185926154255867, "learning_rate": 8.324376252156701e-06, "loss": 0.0419, "step": 82300 }, { "epoch": 2.7827174684742553, "grad_norm": 0.24770353734493256, "learning_rate": 8.315243099836633e-06, "loss": 0.0418, "step": 82310 }, { "epoch": 2.7830555461645083, "grad_norm": 0.3059336543083191, "learning_rate": 8.3061145060989e-06, "loss": 0.0444, "step": 82320 }, { "epoch": 2.783393623854762, "grad_norm": 0.27518659830093384, "learning_rate": 8.296990471941806e-06, "loss": 0.0409, "step": 82330 }, { "epoch": 2.783731701545015, "grad_norm": 0.29448017477989197, "learning_rate": 8.287870998363112e-06, "loss": 0.045, "step": 82340 }, { "epoch": 2.784069779235268, "grad_norm": 0.27893710136413574, "learning_rate": 8.278756086360156e-06, "loss": 0.0438, "step": 82350 }, { "epoch": 2.7844078569255215, "grad_norm": 0.28365084528923035, "learning_rate": 8.269645736929693e-06, "loss": 0.0437, "step": 82360 }, { "epoch": 2.7847459346157746, "grad_norm": 0.3007795512676239, "learning_rate": 8.260539951068048e-06, "loss": 0.0535, "step": 82370 }, { "epoch": 2.785084012306028, "grad_norm": 0.4307478666305542, "learning_rate": 8.251438729770983e-06, "loss": 0.0443, "step": 82380 }, { "epoch": 2.785422089996281, "grad_norm": 0.29181283712387085, "learning_rate": 8.242342074033826e-06, "loss": 0.0472, "step": 82390 }, { "epoch": 2.7857601676865342, "grad_norm": 0.32962700724601746, "learning_rate": 8.233249984851344e-06, "loss": 0.0416, "step": 82400 }, { "epoch": 2.7860982453767877, "grad_norm": 0.1651894599199295, "learning_rate": 8.224162463217849e-06, "loss": 0.0459, "step": 82410 }, { "epoch": 2.786436323067041, "grad_norm": 0.19045644998550415, "learning_rate": 8.215079510127139e-06, "loss": 0.0409, "step": 82420 }, { "epoch": 2.786774400757294, "grad_norm": 0.2046862691640854, "learning_rate": 8.206001126572521e-06, "loss": 0.0444, "step": 82430 }, { "epoch": 2.7871124784475474, "grad_norm": 0.23040591180324554, "learning_rate": 8.196927313546776e-06, "loss": 0.0413, "step": 82440 }, { "epoch": 2.7874505561378005, "grad_norm": 0.31785866618156433, "learning_rate": 8.18785807204222e-06, "loss": 0.0485, "step": 82450 }, { "epoch": 2.7877886338280535, "grad_norm": 0.3052109181880951, "learning_rate": 8.178793403050627e-06, "loss": 0.0425, "step": 82460 }, { "epoch": 2.788126711518307, "grad_norm": 0.16669513285160065, "learning_rate": 8.16973330756331e-06, "loss": 0.0428, "step": 82470 }, { "epoch": 2.78846478920856, "grad_norm": 0.19462630152702332, "learning_rate": 8.16067778657107e-06, "loss": 0.0445, "step": 82480 }, { "epoch": 2.788802866898813, "grad_norm": 0.2854686677455902, "learning_rate": 8.151626841064214e-06, "loss": 0.0519, "step": 82490 }, { "epoch": 2.7891409445890667, "grad_norm": 0.2596639096736908, "learning_rate": 8.142580472032518e-06, "loss": 0.046, "step": 82500 }, { "epoch": 2.7894790222793198, "grad_norm": 0.3212991952896118, "learning_rate": 8.133538680465285e-06, "loss": 0.0428, "step": 82510 }, { "epoch": 2.7898170999695733, "grad_norm": 0.20584966242313385, "learning_rate": 8.124501467351336e-06, "loss": 0.045, "step": 82520 }, { "epoch": 2.7901551776598263, "grad_norm": 0.3248353600502014, "learning_rate": 8.11546883367893e-06, "loss": 0.042, "step": 82530 }, { "epoch": 2.7904932553500794, "grad_norm": 0.34263715147972107, "learning_rate": 8.106440780435882e-06, "loss": 0.0439, "step": 82540 }, { "epoch": 2.790831333040333, "grad_norm": 0.39747315645217896, "learning_rate": 8.097417308609485e-06, "loss": 0.0426, "step": 82550 }, { "epoch": 2.791169410730586, "grad_norm": 0.2599131464958191, "learning_rate": 8.088398419186544e-06, "loss": 0.0395, "step": 82560 }, { "epoch": 2.791507488420839, "grad_norm": 0.23761725425720215, "learning_rate": 8.079384113153321e-06, "loss": 0.044, "step": 82570 }, { "epoch": 2.7918455661110926, "grad_norm": 0.25373363494873047, "learning_rate": 8.070374391495644e-06, "loss": 0.0458, "step": 82580 }, { "epoch": 2.7921836438013456, "grad_norm": 0.1536857783794403, "learning_rate": 8.061369255198764e-06, "loss": 0.0407, "step": 82590 }, { "epoch": 2.7925217214915987, "grad_norm": 0.16494733095169067, "learning_rate": 8.052368705247493e-06, "loss": 0.0418, "step": 82600 }, { "epoch": 2.7928597991818522, "grad_norm": 0.31028226017951965, "learning_rate": 8.043372742626114e-06, "loss": 0.0458, "step": 82610 }, { "epoch": 2.7931978768721053, "grad_norm": 0.2903081178665161, "learning_rate": 8.034381368318422e-06, "loss": 0.0461, "step": 82620 }, { "epoch": 2.7935359545623584, "grad_norm": 0.32549962401390076, "learning_rate": 8.025394583307677e-06, "loss": 0.045, "step": 82630 }, { "epoch": 2.793874032252612, "grad_norm": 0.36212047934532166, "learning_rate": 8.016412388576683e-06, "loss": 0.0447, "step": 82640 }, { "epoch": 2.794212109942865, "grad_norm": 0.2303982675075531, "learning_rate": 8.007434785107699e-06, "loss": 0.0421, "step": 82650 }, { "epoch": 2.794550187633118, "grad_norm": 0.3714626431465149, "learning_rate": 7.998461773882509e-06, "loss": 0.0407, "step": 82660 }, { "epoch": 2.7948882653233715, "grad_norm": 0.16193008422851562, "learning_rate": 7.989493355882393e-06, "loss": 0.0381, "step": 82670 }, { "epoch": 2.7952263430136246, "grad_norm": 0.2825600504875183, "learning_rate": 7.980529532088137e-06, "loss": 0.0399, "step": 82680 }, { "epoch": 2.7955644207038777, "grad_norm": 0.26102039217948914, "learning_rate": 7.97157030347998e-06, "loss": 0.0504, "step": 82690 }, { "epoch": 2.795902498394131, "grad_norm": 0.2281108945608139, "learning_rate": 7.96261567103772e-06, "loss": 0.0402, "step": 82700 }, { "epoch": 2.7962405760843843, "grad_norm": 0.44652059674263, "learning_rate": 7.953665635740593e-06, "loss": 0.0461, "step": 82710 }, { "epoch": 2.7965786537746373, "grad_norm": 0.2822057604789734, "learning_rate": 7.94472019856738e-06, "loss": 0.0414, "step": 82720 }, { "epoch": 2.796916731464891, "grad_norm": 0.18903814256191254, "learning_rate": 7.935779360496337e-06, "loss": 0.04, "step": 82730 }, { "epoch": 2.797254809155144, "grad_norm": 0.2147555649280548, "learning_rate": 7.92684312250523e-06, "loss": 0.0443, "step": 82740 }, { "epoch": 2.797592886845397, "grad_norm": 0.16065111756324768, "learning_rate": 7.917911485571289e-06, "loss": 0.0414, "step": 82750 }, { "epoch": 2.7979309645356505, "grad_norm": 0.36076757311820984, "learning_rate": 7.90898445067128e-06, "loss": 0.0425, "step": 82760 }, { "epoch": 2.7982690422259036, "grad_norm": 0.45221471786499023, "learning_rate": 7.900062018781462e-06, "loss": 0.0444, "step": 82770 }, { "epoch": 2.7986071199161566, "grad_norm": 0.2690753936767578, "learning_rate": 7.891144190877553e-06, "loss": 0.0426, "step": 82780 }, { "epoch": 2.79894519760641, "grad_norm": 0.30879083275794983, "learning_rate": 7.882230967934806e-06, "loss": 0.0411, "step": 82790 }, { "epoch": 2.799283275296663, "grad_norm": 0.29810020327568054, "learning_rate": 7.873322350927959e-06, "loss": 0.0407, "step": 82800 }, { "epoch": 2.7996213529869163, "grad_norm": 0.2862408459186554, "learning_rate": 7.864418340831253e-06, "loss": 0.043, "step": 82810 }, { "epoch": 2.79995943067717, "grad_norm": 0.23073531687259674, "learning_rate": 7.855518938618395e-06, "loss": 0.0382, "step": 82820 }, { "epoch": 2.800297508367423, "grad_norm": 0.18416188657283783, "learning_rate": 7.846624145262633e-06, "loss": 0.0409, "step": 82830 }, { "epoch": 2.800635586057676, "grad_norm": 0.30784112215042114, "learning_rate": 7.837733961736672e-06, "loss": 0.044, "step": 82840 }, { "epoch": 2.8009736637479294, "grad_norm": 0.19864703714847565, "learning_rate": 7.828848389012727e-06, "loss": 0.0408, "step": 82850 }, { "epoch": 2.8013117414381825, "grad_norm": 0.2971931993961334, "learning_rate": 7.819967428062524e-06, "loss": 0.0403, "step": 82860 }, { "epoch": 2.8016498191284356, "grad_norm": 0.2331208437681198, "learning_rate": 7.811091079857274e-06, "loss": 0.0403, "step": 82870 }, { "epoch": 2.801987896818689, "grad_norm": 0.3281390070915222, "learning_rate": 7.802219345367667e-06, "loss": 0.0448, "step": 82880 }, { "epoch": 2.802325974508942, "grad_norm": 0.17661836743354797, "learning_rate": 7.793352225563916e-06, "loss": 0.0455, "step": 82890 }, { "epoch": 2.8026640521991952, "grad_norm": 0.15159878134727478, "learning_rate": 7.7844897214157e-06, "loss": 0.0422, "step": 82900 }, { "epoch": 2.8030021298894487, "grad_norm": 0.19122807681560516, "learning_rate": 7.775631833892228e-06, "loss": 0.0427, "step": 82910 }, { "epoch": 2.803340207579702, "grad_norm": 0.4375275671482086, "learning_rate": 7.76677856396215e-06, "loss": 0.0433, "step": 82920 }, { "epoch": 2.803678285269955, "grad_norm": 0.29454222321510315, "learning_rate": 7.757929912593697e-06, "loss": 0.044, "step": 82930 }, { "epoch": 2.8040163629602084, "grad_norm": 0.1549697369337082, "learning_rate": 7.749085880754509e-06, "loss": 0.042, "step": 82940 }, { "epoch": 2.8043544406504615, "grad_norm": 0.21058255434036255, "learning_rate": 7.740246469411777e-06, "loss": 0.0458, "step": 82950 }, { "epoch": 2.8046925183407145, "grad_norm": 0.2557981610298157, "learning_rate": 7.731411679532142e-06, "loss": 0.043, "step": 82960 }, { "epoch": 2.805030596030968, "grad_norm": 0.2726046144962311, "learning_rate": 7.722581512081779e-06, "loss": 0.0371, "step": 82970 }, { "epoch": 2.805368673721221, "grad_norm": 0.1654396504163742, "learning_rate": 7.713755968026331e-06, "loss": 0.0427, "step": 82980 }, { "epoch": 2.805706751411474, "grad_norm": 0.22398215532302856, "learning_rate": 7.704935048330969e-06, "loss": 0.0498, "step": 82990 }, { "epoch": 2.8060448291017277, "grad_norm": 0.2747970521450043, "learning_rate": 7.696118753960313e-06, "loss": 0.0467, "step": 83000 }, { "epoch": 2.8063829067919808, "grad_norm": 0.2331901639699936, "learning_rate": 7.687307085878503e-06, "loss": 0.0421, "step": 83010 }, { "epoch": 2.806720984482234, "grad_norm": 0.29441094398498535, "learning_rate": 7.678500045049186e-06, "loss": 0.0403, "step": 83020 }, { "epoch": 2.8070590621724874, "grad_norm": 0.1544760912656784, "learning_rate": 7.669697632435463e-06, "loss": 0.0442, "step": 83030 }, { "epoch": 2.8073971398627404, "grad_norm": 0.1892900913953781, "learning_rate": 7.660899848999963e-06, "loss": 0.044, "step": 83040 }, { "epoch": 2.8077352175529935, "grad_norm": 0.1803523749113083, "learning_rate": 7.652106695704803e-06, "loss": 0.0399, "step": 83050 }, { "epoch": 2.808073295243247, "grad_norm": 0.1554846614599228, "learning_rate": 7.643318173511598e-06, "loss": 0.0387, "step": 83060 }, { "epoch": 2.8084113729335, "grad_norm": 0.24434535205364227, "learning_rate": 7.634534283381423e-06, "loss": 0.0416, "step": 83070 }, { "epoch": 2.808749450623753, "grad_norm": 0.1598326861858368, "learning_rate": 7.625755026274889e-06, "loss": 0.0419, "step": 83080 }, { "epoch": 2.8090875283140067, "grad_norm": 0.2407798320055008, "learning_rate": 7.616980403152069e-06, "loss": 0.044, "step": 83090 }, { "epoch": 2.8094256060042597, "grad_norm": 0.435696005821228, "learning_rate": 7.608210414972561e-06, "loss": 0.04, "step": 83100 }, { "epoch": 2.809763683694513, "grad_norm": 0.17483080923557281, "learning_rate": 7.599445062695404e-06, "loss": 0.0403, "step": 83110 }, { "epoch": 2.8101017613847663, "grad_norm": 0.3886816203594208, "learning_rate": 7.590684347279214e-06, "loss": 0.0453, "step": 83120 }, { "epoch": 2.8104398390750194, "grad_norm": 0.42786315083503723, "learning_rate": 7.581928269682004e-06, "loss": 0.0414, "step": 83130 }, { "epoch": 2.8107779167652724, "grad_norm": 0.1904345452785492, "learning_rate": 7.573176830861361e-06, "loss": 0.0391, "step": 83140 }, { "epoch": 2.811115994455526, "grad_norm": 0.21519270539283752, "learning_rate": 7.5644300317743e-06, "loss": 0.0473, "step": 83150 }, { "epoch": 2.811454072145779, "grad_norm": 0.17082762718200684, "learning_rate": 7.555687873377376e-06, "loss": 0.0431, "step": 83160 }, { "epoch": 2.811792149836032, "grad_norm": 0.19768624007701874, "learning_rate": 7.546950356626592e-06, "loss": 0.039, "step": 83170 }, { "epoch": 2.8121302275262856, "grad_norm": 0.2531498074531555, "learning_rate": 7.538217482477516e-06, "loss": 0.0433, "step": 83180 }, { "epoch": 2.8124683052165387, "grad_norm": 0.4024569094181061, "learning_rate": 7.529489251885119e-06, "loss": 0.0435, "step": 83190 }, { "epoch": 2.8128063829067917, "grad_norm": 0.1601000726222992, "learning_rate": 7.520765665803941e-06, "loss": 0.0438, "step": 83200 }, { "epoch": 2.8131444605970453, "grad_norm": 0.3154531717300415, "learning_rate": 7.512046725187949e-06, "loss": 0.0397, "step": 83210 }, { "epoch": 2.8134825382872983, "grad_norm": 0.29329484701156616, "learning_rate": 7.503332430990645e-06, "loss": 0.0502, "step": 83220 }, { "epoch": 2.8138206159775514, "grad_norm": 0.3192998170852661, "learning_rate": 7.494622784165017e-06, "loss": 0.0433, "step": 83230 }, { "epoch": 2.814158693667805, "grad_norm": 0.26386019587516785, "learning_rate": 7.485917785663543e-06, "loss": 0.0437, "step": 83240 }, { "epoch": 2.814496771358058, "grad_norm": 0.18383213877677917, "learning_rate": 7.477217436438172e-06, "loss": 0.0426, "step": 83250 }, { "epoch": 2.814834849048311, "grad_norm": 0.38734671473503113, "learning_rate": 7.46852173744037e-06, "loss": 0.0471, "step": 83260 }, { "epoch": 2.8151729267385646, "grad_norm": 0.21029409766197205, "learning_rate": 7.4598306896211004e-06, "loss": 0.0451, "step": 83270 }, { "epoch": 2.8155110044288176, "grad_norm": 0.2560451328754425, "learning_rate": 7.451144293930773e-06, "loss": 0.041, "step": 83280 }, { "epoch": 2.8158490821190707, "grad_norm": 0.15969951450824738, "learning_rate": 7.4424625513193455e-06, "loss": 0.0387, "step": 83290 }, { "epoch": 2.816187159809324, "grad_norm": 0.22999966144561768, "learning_rate": 7.433785462736209e-06, "loss": 0.0413, "step": 83300 }, { "epoch": 2.8165252374995773, "grad_norm": 0.18989118933677673, "learning_rate": 7.425113029130315e-06, "loss": 0.0485, "step": 83310 }, { "epoch": 2.8168633151898304, "grad_norm": 0.185896635055542, "learning_rate": 7.416445251450044e-06, "loss": 0.0366, "step": 83320 }, { "epoch": 2.817201392880084, "grad_norm": 0.2451857030391693, "learning_rate": 7.4077821306433035e-06, "loss": 0.04, "step": 83330 }, { "epoch": 2.817539470570337, "grad_norm": 0.23399090766906738, "learning_rate": 7.399123667657465e-06, "loss": 0.044, "step": 83340 }, { "epoch": 2.8178775482605904, "grad_norm": 0.18344257771968842, "learning_rate": 7.390469863439426e-06, "loss": 0.0409, "step": 83350 }, { "epoch": 2.8182156259508435, "grad_norm": 0.20370212197303772, "learning_rate": 7.381820718935512e-06, "loss": 0.0436, "step": 83360 }, { "epoch": 2.8185537036410966, "grad_norm": 0.36885079741477966, "learning_rate": 7.373176235091633e-06, "loss": 0.0449, "step": 83370 }, { "epoch": 2.81889178133135, "grad_norm": 0.23396919667720795, "learning_rate": 7.364536412853101e-06, "loss": 0.0461, "step": 83380 }, { "epoch": 2.819229859021603, "grad_norm": 0.1580374836921692, "learning_rate": 7.355901253164776e-06, "loss": 0.0419, "step": 83390 }, { "epoch": 2.8195679367118562, "grad_norm": 0.24639776349067688, "learning_rate": 7.34727075697097e-06, "loss": 0.0403, "step": 83400 }, { "epoch": 2.8199060144021098, "grad_norm": 0.17722627520561218, "learning_rate": 7.3386449252155095e-06, "loss": 0.043, "step": 83410 }, { "epoch": 2.820244092092363, "grad_norm": 0.3018106520175934, "learning_rate": 7.330023758841692e-06, "loss": 0.0463, "step": 83420 }, { "epoch": 2.820582169782616, "grad_norm": 0.20431235432624817, "learning_rate": 7.321407258792323e-06, "loss": 0.0443, "step": 83430 }, { "epoch": 2.8209202474728694, "grad_norm": 0.28431424498558044, "learning_rate": 7.312795426009694e-06, "loss": 0.0455, "step": 83440 }, { "epoch": 2.8212583251631225, "grad_norm": 0.16835075616836548, "learning_rate": 7.304188261435585e-06, "loss": 0.0393, "step": 83450 }, { "epoch": 2.8215964028533755, "grad_norm": 0.15645326673984528, "learning_rate": 7.295585766011242e-06, "loss": 0.0416, "step": 83460 }, { "epoch": 2.821934480543629, "grad_norm": 0.21616239845752716, "learning_rate": 7.286987940677436e-06, "loss": 0.0433, "step": 83470 }, { "epoch": 2.822272558233882, "grad_norm": 0.27376940846443176, "learning_rate": 7.278394786374426e-06, "loss": 0.0416, "step": 83480 }, { "epoch": 2.8226106359241356, "grad_norm": 0.29163655638694763, "learning_rate": 7.269806304041915e-06, "loss": 0.0431, "step": 83490 }, { "epoch": 2.8229487136143887, "grad_norm": 0.16421110928058624, "learning_rate": 7.261222494619147e-06, "loss": 0.0442, "step": 83500 }, { "epoch": 2.8232867913046418, "grad_norm": 0.2602076232433319, "learning_rate": 7.252643359044825e-06, "loss": 0.0471, "step": 83510 }, { "epoch": 2.8236248689948953, "grad_norm": 0.21071584522724152, "learning_rate": 7.244068898257172e-06, "loss": 0.0424, "step": 83520 }, { "epoch": 2.8239629466851484, "grad_norm": 0.2236775904893875, "learning_rate": 7.2354991131938495e-06, "loss": 0.0441, "step": 83530 }, { "epoch": 2.8243010243754014, "grad_norm": 0.17779409885406494, "learning_rate": 7.2269340047920574e-06, "loss": 0.0364, "step": 83540 }, { "epoch": 2.824639102065655, "grad_norm": 0.1960013210773468, "learning_rate": 7.218373573988435e-06, "loss": 0.042, "step": 83550 }, { "epoch": 2.824977179755908, "grad_norm": 0.4338124692440033, "learning_rate": 7.209817821719178e-06, "loss": 0.0427, "step": 83560 }, { "epoch": 2.825315257446161, "grad_norm": 0.1557174175977707, "learning_rate": 7.201266748919894e-06, "loss": 0.0417, "step": 83570 }, { "epoch": 2.8256533351364146, "grad_norm": 0.1629246026277542, "learning_rate": 7.19272035652575e-06, "loss": 0.0407, "step": 83580 }, { "epoch": 2.8259914128266677, "grad_norm": 0.23909686505794525, "learning_rate": 7.184178645471334e-06, "loss": 0.0414, "step": 83590 }, { "epoch": 2.8263294905169207, "grad_norm": 0.34666454792022705, "learning_rate": 7.175641616690775e-06, "loss": 0.0395, "step": 83600 }, { "epoch": 2.8266675682071742, "grad_norm": 0.21679072082042694, "learning_rate": 7.167109271117661e-06, "loss": 0.0435, "step": 83610 }, { "epoch": 2.8270056458974273, "grad_norm": 0.30624818801879883, "learning_rate": 7.158581609685072e-06, "loss": 0.0393, "step": 83620 }, { "epoch": 2.8273437235876804, "grad_norm": 0.20033767819404602, "learning_rate": 7.150058633325591e-06, "loss": 0.0386, "step": 83630 }, { "epoch": 2.827681801277934, "grad_norm": 2.739067316055298, "learning_rate": 7.141540342971287e-06, "loss": 0.039, "step": 83640 }, { "epoch": 2.828019878968187, "grad_norm": 0.1891656219959259, "learning_rate": 7.133026739553677e-06, "loss": 0.0447, "step": 83650 }, { "epoch": 2.82835795665844, "grad_norm": 0.2440343052148819, "learning_rate": 7.124517824003829e-06, "loss": 0.0401, "step": 83660 }, { "epoch": 2.8286960343486935, "grad_norm": 0.27925625443458557, "learning_rate": 7.116013597252236e-06, "loss": 0.0422, "step": 83670 }, { "epoch": 2.8290341120389466, "grad_norm": 0.33099251985549927, "learning_rate": 7.107514060228921e-06, "loss": 0.04, "step": 83680 }, { "epoch": 2.8293721897291997, "grad_norm": 0.2680320143699646, "learning_rate": 7.099019213863378e-06, "loss": 0.0444, "step": 83690 }, { "epoch": 2.829710267419453, "grad_norm": 0.3123638927936554, "learning_rate": 7.090529059084611e-06, "loss": 0.0456, "step": 83700 }, { "epoch": 2.8300483451097063, "grad_norm": 0.22917777299880981, "learning_rate": 7.082043596821058e-06, "loss": 0.0408, "step": 83710 }, { "epoch": 2.8303864227999593, "grad_norm": 0.2973829507827759, "learning_rate": 7.07356282800069e-06, "loss": 0.0503, "step": 83720 }, { "epoch": 2.830724500490213, "grad_norm": 0.2521131634712219, "learning_rate": 7.0650867535509645e-06, "loss": 0.0427, "step": 83730 }, { "epoch": 2.831062578180466, "grad_norm": 0.2139766812324524, "learning_rate": 7.0566153743987895e-06, "loss": 0.0391, "step": 83740 }, { "epoch": 2.831400655870719, "grad_norm": 0.1718721240758896, "learning_rate": 7.04814869147059e-06, "loss": 0.0423, "step": 83750 }, { "epoch": 2.8317387335609725, "grad_norm": 0.3608884811401367, "learning_rate": 7.0396867056922725e-06, "loss": 0.0354, "step": 83760 }, { "epoch": 2.8320768112512256, "grad_norm": 0.2047964483499527, "learning_rate": 7.031229417989244e-06, "loss": 0.0379, "step": 83770 }, { "epoch": 2.8324148889414786, "grad_norm": 0.24452711641788483, "learning_rate": 7.022776829286348e-06, "loss": 0.0483, "step": 83780 }, { "epoch": 2.832752966631732, "grad_norm": 0.7134111523628235, "learning_rate": 7.014328940507975e-06, "loss": 0.0404, "step": 83790 }, { "epoch": 2.833091044321985, "grad_norm": 0.3246191740036011, "learning_rate": 7.005885752577945e-06, "loss": 0.0411, "step": 83800 }, { "epoch": 2.8334291220122383, "grad_norm": 0.22146767377853394, "learning_rate": 6.997447266419615e-06, "loss": 0.0403, "step": 83810 }, { "epoch": 2.833767199702492, "grad_norm": 0.18635129928588867, "learning_rate": 6.989013482955797e-06, "loss": 0.0447, "step": 83820 }, { "epoch": 2.834105277392745, "grad_norm": 0.15916083753108978, "learning_rate": 6.980584403108808e-06, "loss": 0.0462, "step": 83830 }, { "epoch": 2.834443355082998, "grad_norm": 0.1756868213415146, "learning_rate": 6.972160027800417e-06, "loss": 0.0454, "step": 83840 }, { "epoch": 2.8347814327732515, "grad_norm": 0.2802339792251587, "learning_rate": 6.963740357951931e-06, "loss": 0.0414, "step": 83850 }, { "epoch": 2.8351195104635045, "grad_norm": 0.16788502037525177, "learning_rate": 6.95532539448408e-06, "loss": 0.04, "step": 83860 }, { "epoch": 2.8354575881537576, "grad_norm": 0.4646252691745758, "learning_rate": 6.946915138317128e-06, "loss": 0.0438, "step": 83870 }, { "epoch": 2.835795665844011, "grad_norm": 0.14697538316249847, "learning_rate": 6.9385095903708065e-06, "loss": 0.0441, "step": 83880 }, { "epoch": 2.836133743534264, "grad_norm": 0.6675686240196228, "learning_rate": 6.930108751564346e-06, "loss": 0.0557, "step": 83890 }, { "epoch": 2.8364718212245172, "grad_norm": 0.2904934585094452, "learning_rate": 6.921712622816428e-06, "loss": 0.0406, "step": 83900 }, { "epoch": 2.8368098989147708, "grad_norm": 0.24997705221176147, "learning_rate": 6.913321205045259e-06, "loss": 0.044, "step": 83910 }, { "epoch": 2.837147976605024, "grad_norm": 0.2647833228111267, "learning_rate": 6.904934499168491e-06, "loss": 0.0415, "step": 83920 }, { "epoch": 2.837486054295277, "grad_norm": 0.23692114651203156, "learning_rate": 6.896552506103299e-06, "loss": 0.0485, "step": 83930 }, { "epoch": 2.8378241319855304, "grad_norm": 0.2209695428609848, "learning_rate": 6.888175226766313e-06, "loss": 0.0433, "step": 83940 }, { "epoch": 2.8381622096757835, "grad_norm": 0.1857345998287201, "learning_rate": 6.879802662073686e-06, "loss": 0.0413, "step": 83950 }, { "epoch": 2.8385002873660365, "grad_norm": 0.24819263815879822, "learning_rate": 6.871434812940991e-06, "loss": 0.0395, "step": 83960 }, { "epoch": 2.83883836505629, "grad_norm": 0.3975434899330139, "learning_rate": 6.863071680283345e-06, "loss": 0.0439, "step": 83970 }, { "epoch": 2.839176442746543, "grad_norm": 0.2032661736011505, "learning_rate": 6.854713265015339e-06, "loss": 0.0404, "step": 83980 }, { "epoch": 2.839514520436796, "grad_norm": 0.15238633751869202, "learning_rate": 6.8463595680510025e-06, "loss": 0.0417, "step": 83990 }, { "epoch": 2.8398525981270497, "grad_norm": 0.20867864787578583, "learning_rate": 6.838010590303911e-06, "loss": 0.0428, "step": 84000 }, { "epoch": 2.840190675817303, "grad_norm": 0.19414947926998138, "learning_rate": 6.829666332687085e-06, "loss": 0.0447, "step": 84010 }, { "epoch": 2.840528753507556, "grad_norm": 0.4660297930240631, "learning_rate": 6.821326796113059e-06, "loss": 0.0399, "step": 84020 }, { "epoch": 2.8408668311978094, "grad_norm": 0.3080141544342041, "learning_rate": 6.812991981493799e-06, "loss": 0.0399, "step": 84030 }, { "epoch": 2.8412049088880624, "grad_norm": 0.19488708674907684, "learning_rate": 6.804661889740821e-06, "loss": 0.0403, "step": 84040 }, { "epoch": 2.8415429865783155, "grad_norm": 0.2624748945236206, "learning_rate": 6.7963365217650625e-06, "loss": 0.0468, "step": 84050 }, { "epoch": 2.841881064268569, "grad_norm": 0.40036338567733765, "learning_rate": 6.788015878476983e-06, "loss": 0.0475, "step": 84060 }, { "epoch": 2.842219141958822, "grad_norm": 0.5085367560386658, "learning_rate": 6.779699960786523e-06, "loss": 0.0403, "step": 84070 }, { "epoch": 2.842557219649075, "grad_norm": 0.14993126690387726, "learning_rate": 6.771388769603104e-06, "loss": 0.0395, "step": 84080 }, { "epoch": 2.8428952973393287, "grad_norm": 0.24791815876960754, "learning_rate": 6.763082305835605e-06, "loss": 0.0424, "step": 84090 }, { "epoch": 2.8432333750295817, "grad_norm": 0.26680639386177063, "learning_rate": 6.754780570392433e-06, "loss": 0.041, "step": 84100 }, { "epoch": 2.843571452719835, "grad_norm": 0.40241947770118713, "learning_rate": 6.7464835641814215e-06, "loss": 0.0423, "step": 84110 }, { "epoch": 2.8439095304100883, "grad_norm": 0.1757325679063797, "learning_rate": 6.7381912881099515e-06, "loss": 0.038, "step": 84120 }, { "epoch": 2.8442476081003414, "grad_norm": 0.6057904362678528, "learning_rate": 6.72990374308482e-06, "loss": 0.0449, "step": 84130 }, { "epoch": 2.8445856857905945, "grad_norm": 0.23344656825065613, "learning_rate": 6.721620930012379e-06, "loss": 0.0423, "step": 84140 }, { "epoch": 2.844923763480848, "grad_norm": 0.3366641402244568, "learning_rate": 6.713342849798393e-06, "loss": 0.042, "step": 84150 }, { "epoch": 2.845261841171101, "grad_norm": 0.13317692279815674, "learning_rate": 6.705069503348166e-06, "loss": 0.0379, "step": 84160 }, { "epoch": 2.845599918861354, "grad_norm": 0.352498322725296, "learning_rate": 6.696800891566435e-06, "loss": 0.0472, "step": 84170 }, { "epoch": 2.8459379965516076, "grad_norm": 0.6037773489952087, "learning_rate": 6.688537015357449e-06, "loss": 0.0456, "step": 84180 }, { "epoch": 2.8462760742418607, "grad_norm": 0.2218054085969925, "learning_rate": 6.680277875624941e-06, "loss": 0.0416, "step": 84190 }, { "epoch": 2.8466141519321138, "grad_norm": 0.2561713457107544, "learning_rate": 6.672023473272121e-06, "loss": 0.0379, "step": 84200 }, { "epoch": 2.8469522296223673, "grad_norm": 0.2520774006843567, "learning_rate": 6.663773809201667e-06, "loss": 0.0424, "step": 84210 }, { "epoch": 2.8472903073126203, "grad_norm": 0.2687004804611206, "learning_rate": 6.655528884315748e-06, "loss": 0.0446, "step": 84220 }, { "epoch": 2.8476283850028734, "grad_norm": 0.2225504368543625, "learning_rate": 6.647288699516036e-06, "loss": 0.0404, "step": 84230 }, { "epoch": 2.847966462693127, "grad_norm": 0.19048815965652466, "learning_rate": 6.639053255703642e-06, "loss": 0.042, "step": 84240 }, { "epoch": 2.84830454038338, "grad_norm": 0.19932891428470612, "learning_rate": 6.6308225537791925e-06, "loss": 0.0408, "step": 84250 }, { "epoch": 2.848642618073633, "grad_norm": 0.19528868794441223, "learning_rate": 6.622596594642777e-06, "loss": 0.0386, "step": 84260 }, { "epoch": 2.8489806957638866, "grad_norm": 0.23197592794895172, "learning_rate": 6.614375379193999e-06, "loss": 0.0371, "step": 84270 }, { "epoch": 2.8493187734541396, "grad_norm": 0.2496892213821411, "learning_rate": 6.606158908331889e-06, "loss": 0.0397, "step": 84280 }, { "epoch": 2.8496568511443927, "grad_norm": 0.34572967886924744, "learning_rate": 6.597947182955005e-06, "loss": 0.0456, "step": 84290 }, { "epoch": 2.8499949288346462, "grad_norm": 0.2022150456905365, "learning_rate": 6.589740203961359e-06, "loss": 0.0426, "step": 84300 }, { "epoch": 2.8503330065248993, "grad_norm": 0.1983128935098648, "learning_rate": 6.581537972248464e-06, "loss": 0.043, "step": 84310 }, { "epoch": 2.850671084215153, "grad_norm": 0.20125775039196014, "learning_rate": 6.573340488713275e-06, "loss": 0.0441, "step": 84320 }, { "epoch": 2.851009161905406, "grad_norm": 0.3149738013744354, "learning_rate": 6.565147754252304e-06, "loss": 0.0448, "step": 84330 }, { "epoch": 2.851347239595659, "grad_norm": 0.2198375016450882, "learning_rate": 6.556959769761462e-06, "loss": 0.045, "step": 84340 }, { "epoch": 2.8516853172859125, "grad_norm": 0.24592649936676025, "learning_rate": 6.548776536136192e-06, "loss": 0.039, "step": 84350 }, { "epoch": 2.8520233949761655, "grad_norm": 0.30928540229797363, "learning_rate": 6.54059805427138e-06, "loss": 0.047, "step": 84360 }, { "epoch": 2.8523614726664186, "grad_norm": 0.20406101644039154, "learning_rate": 6.532424325061437e-06, "loss": 0.0399, "step": 84370 }, { "epoch": 2.852699550356672, "grad_norm": 1.2225266695022583, "learning_rate": 6.5242553494001935e-06, "loss": 0.0431, "step": 84380 }, { "epoch": 2.853037628046925, "grad_norm": 0.2456398904323578, "learning_rate": 6.516091128181046e-06, "loss": 0.0424, "step": 84390 }, { "epoch": 2.8533757057371782, "grad_norm": 0.37746891379356384, "learning_rate": 6.507931662296785e-06, "loss": 0.0388, "step": 84400 }, { "epoch": 2.8537137834274318, "grad_norm": 0.1899624764919281, "learning_rate": 6.4997769526397354e-06, "loss": 0.0415, "step": 84410 }, { "epoch": 2.854051861117685, "grad_norm": 0.2568485140800476, "learning_rate": 6.491627000101663e-06, "loss": 0.0416, "step": 84420 }, { "epoch": 2.854389938807938, "grad_norm": 0.25863656401634216, "learning_rate": 6.4834818055738545e-06, "loss": 0.0423, "step": 84430 }, { "epoch": 2.8547280164981914, "grad_norm": 0.3391711711883545, "learning_rate": 6.4753413699470465e-06, "loss": 0.043, "step": 84440 }, { "epoch": 2.8550660941884445, "grad_norm": 0.3329007029533386, "learning_rate": 6.467205694111478e-06, "loss": 0.0413, "step": 84450 }, { "epoch": 2.855404171878698, "grad_norm": 0.2609030604362488, "learning_rate": 6.459074778956836e-06, "loss": 0.0421, "step": 84460 }, { "epoch": 2.855742249568951, "grad_norm": 0.21550211310386658, "learning_rate": 6.450948625372311e-06, "loss": 0.0423, "step": 84470 }, { "epoch": 2.856080327259204, "grad_norm": 0.34940266609191895, "learning_rate": 6.442827234246585e-06, "loss": 0.0427, "step": 84480 }, { "epoch": 2.8564184049494576, "grad_norm": 0.20555193722248077, "learning_rate": 6.43471060646777e-06, "loss": 0.0466, "step": 84490 }, { "epoch": 2.8567564826397107, "grad_norm": 0.1500038057565689, "learning_rate": 6.426598742923512e-06, "loss": 0.0382, "step": 84500 }, { "epoch": 2.857094560329964, "grad_norm": 0.17023055255413055, "learning_rate": 6.418491644500885e-06, "loss": 0.0396, "step": 84510 }, { "epoch": 2.8574326380202173, "grad_norm": 0.1955903321504593, "learning_rate": 6.410389312086512e-06, "loss": 0.0397, "step": 84520 }, { "epoch": 2.8577707157104704, "grad_norm": 0.5742412209510803, "learning_rate": 6.4022917465664135e-06, "loss": 0.0441, "step": 84530 }, { "epoch": 2.8581087934007234, "grad_norm": 0.27053210139274597, "learning_rate": 6.3941989488261575e-06, "loss": 0.0428, "step": 84540 }, { "epoch": 2.858446871090977, "grad_norm": 0.2135850489139557, "learning_rate": 6.386110919750726e-06, "loss": 0.0391, "step": 84550 }, { "epoch": 2.85878494878123, "grad_norm": 0.15189296007156372, "learning_rate": 6.378027660224651e-06, "loss": 0.038, "step": 84560 }, { "epoch": 2.859123026471483, "grad_norm": 0.5093490481376648, "learning_rate": 6.369949171131856e-06, "loss": 0.0435, "step": 84570 }, { "epoch": 2.8594611041617366, "grad_norm": 0.17033106088638306, "learning_rate": 6.361875453355848e-06, "loss": 0.041, "step": 84580 }, { "epoch": 2.8597991818519897, "grad_norm": 0.3025341331958771, "learning_rate": 6.353806507779525e-06, "loss": 0.0424, "step": 84590 }, { "epoch": 2.8601372595422427, "grad_norm": 0.23715975880622864, "learning_rate": 6.3457423352853044e-06, "loss": 0.0433, "step": 84600 }, { "epoch": 2.8604753372324963, "grad_norm": 0.1489495187997818, "learning_rate": 6.337682936755062e-06, "loss": 0.0395, "step": 84610 }, { "epoch": 2.8608134149227493, "grad_norm": 0.33644551038742065, "learning_rate": 6.329628313070179e-06, "loss": 0.0443, "step": 84620 }, { "epoch": 2.8611514926130024, "grad_norm": 0.24572059512138367, "learning_rate": 6.321578465111477e-06, "loss": 0.0461, "step": 84630 }, { "epoch": 2.861489570303256, "grad_norm": 0.27934396266937256, "learning_rate": 6.31353339375928e-06, "loss": 0.043, "step": 84640 }, { "epoch": 2.861827647993509, "grad_norm": 0.21293388307094574, "learning_rate": 6.305493099893395e-06, "loss": 0.0478, "step": 84650 }, { "epoch": 2.862165725683762, "grad_norm": 0.5644688010215759, "learning_rate": 6.297457584393096e-06, "loss": 0.0436, "step": 84660 }, { "epoch": 2.8625038033740156, "grad_norm": 0.36863017082214355, "learning_rate": 6.289426848137126e-06, "loss": 0.0438, "step": 84670 }, { "epoch": 2.8628418810642686, "grad_norm": 0.3359792232513428, "learning_rate": 6.281400892003713e-06, "loss": 0.0452, "step": 84680 }, { "epoch": 2.8631799587545217, "grad_norm": 0.35163745284080505, "learning_rate": 6.273379716870581e-06, "loss": 0.0396, "step": 84690 }, { "epoch": 2.863518036444775, "grad_norm": 0.22150762379169464, "learning_rate": 6.265363323614887e-06, "loss": 0.045, "step": 84700 }, { "epoch": 2.8638561141350283, "grad_norm": 0.2666744589805603, "learning_rate": 6.257351713113308e-06, "loss": 0.044, "step": 84710 }, { "epoch": 2.8641941918252813, "grad_norm": 0.2741228938102722, "learning_rate": 6.249344886241981e-06, "loss": 0.0435, "step": 84720 }, { "epoch": 2.864532269515535, "grad_norm": 0.3843180239200592, "learning_rate": 6.2413428438765295e-06, "loss": 0.0427, "step": 84730 }, { "epoch": 2.864870347205788, "grad_norm": 0.20142734050750732, "learning_rate": 6.233345586892026e-06, "loss": 0.043, "step": 84740 }, { "epoch": 2.865208424896041, "grad_norm": 0.2815329432487488, "learning_rate": 6.225353116163052e-06, "loss": 0.0407, "step": 84750 }, { "epoch": 2.8655465025862945, "grad_norm": 0.21805568039417267, "learning_rate": 6.217365432563627e-06, "loss": 0.0431, "step": 84760 }, { "epoch": 2.8658845802765476, "grad_norm": 0.2017555981874466, "learning_rate": 6.209382536967318e-06, "loss": 0.0378, "step": 84770 }, { "epoch": 2.8662226579668006, "grad_norm": 0.3493207097053528, "learning_rate": 6.201404430247082e-06, "loss": 0.044, "step": 84780 }, { "epoch": 2.866560735657054, "grad_norm": 0.4050888419151306, "learning_rate": 6.1934311132754165e-06, "loss": 0.0444, "step": 84790 }, { "epoch": 2.8668988133473072, "grad_norm": 0.21783538162708282, "learning_rate": 6.185462586924257e-06, "loss": 0.0445, "step": 84800 }, { "epoch": 2.8672368910375603, "grad_norm": 0.4706352651119232, "learning_rate": 6.177498852065039e-06, "loss": 0.0432, "step": 84810 }, { "epoch": 2.867574968727814, "grad_norm": 0.28265833854675293, "learning_rate": 6.169539909568655e-06, "loss": 0.0455, "step": 84820 }, { "epoch": 2.867913046418067, "grad_norm": 0.3184395730495453, "learning_rate": 6.161585760305488e-06, "loss": 0.0459, "step": 84830 }, { "epoch": 2.86825112410832, "grad_norm": 0.3036477267742157, "learning_rate": 6.153636405145391e-06, "loss": 0.0426, "step": 84840 }, { "epoch": 2.8685892017985735, "grad_norm": 0.26016271114349365, "learning_rate": 6.145691844957702e-06, "loss": 0.0418, "step": 84850 }, { "epoch": 2.8689272794888265, "grad_norm": 0.22538623213768005, "learning_rate": 6.1377520806112145e-06, "loss": 0.0437, "step": 84860 }, { "epoch": 2.8692653571790796, "grad_norm": 0.2134377658367157, "learning_rate": 6.129817112974218e-06, "loss": 0.0478, "step": 84870 }, { "epoch": 2.869603434869333, "grad_norm": 0.2702528238296509, "learning_rate": 6.121886942914457e-06, "loss": 0.0377, "step": 84880 }, { "epoch": 2.869941512559586, "grad_norm": 0.20774683356285095, "learning_rate": 6.11396157129917e-06, "loss": 0.0431, "step": 84890 }, { "epoch": 2.8702795902498393, "grad_norm": 0.22515630722045898, "learning_rate": 6.106040998995066e-06, "loss": 0.0455, "step": 84900 }, { "epoch": 2.8706176679400928, "grad_norm": 0.20816069841384888, "learning_rate": 6.098125226868329e-06, "loss": 0.0419, "step": 84910 }, { "epoch": 2.870955745630346, "grad_norm": 0.23350460827350616, "learning_rate": 6.090214255784604e-06, "loss": 0.043, "step": 84920 }, { "epoch": 2.871293823320599, "grad_norm": 0.14257267117500305, "learning_rate": 6.0823080866090346e-06, "loss": 0.041, "step": 84930 }, { "epoch": 2.8716319010108524, "grad_norm": 0.2250775694847107, "learning_rate": 6.074406720206233e-06, "loss": 0.0413, "step": 84940 }, { "epoch": 2.8719699787011055, "grad_norm": 0.22348853945732117, "learning_rate": 6.066510157440259e-06, "loss": 0.0426, "step": 84950 }, { "epoch": 2.8723080563913586, "grad_norm": 0.20809316635131836, "learning_rate": 6.058618399174676e-06, "loss": 0.0439, "step": 84960 }, { "epoch": 2.872646134081612, "grad_norm": 0.26619526743888855, "learning_rate": 6.050731446272528e-06, "loss": 0.0398, "step": 84970 }, { "epoch": 2.872984211771865, "grad_norm": 0.28624841570854187, "learning_rate": 6.042849299596321e-06, "loss": 0.0489, "step": 84980 }, { "epoch": 2.873322289462118, "grad_norm": 0.31334972381591797, "learning_rate": 6.034971960008013e-06, "loss": 0.0446, "step": 84990 }, { "epoch": 2.8736603671523717, "grad_norm": 0.24506691098213196, "learning_rate": 6.027099428369082e-06, "loss": 0.0395, "step": 85000 }, { "epoch": 2.873998444842625, "grad_norm": 0.23171888291835785, "learning_rate": 6.019231705540435e-06, "loss": 0.0423, "step": 85010 }, { "epoch": 2.874336522532878, "grad_norm": 0.22548362612724304, "learning_rate": 6.011368792382488e-06, "loss": 0.044, "step": 85020 }, { "epoch": 2.8746746002231314, "grad_norm": 0.26117560267448425, "learning_rate": 6.00351068975511e-06, "loss": 0.0439, "step": 85030 }, { "epoch": 2.8750126779133844, "grad_norm": 0.27842897176742554, "learning_rate": 5.9956573985176674e-06, "loss": 0.0436, "step": 85040 }, { "epoch": 2.8753507556036375, "grad_norm": 0.4157659709453583, "learning_rate": 5.987808919528959e-06, "loss": 0.0418, "step": 85050 }, { "epoch": 2.875688833293891, "grad_norm": 0.19202712178230286, "learning_rate": 5.979965253647307e-06, "loss": 0.0408, "step": 85060 }, { "epoch": 2.876026910984144, "grad_norm": 0.2852787971496582, "learning_rate": 5.9721264017304655e-06, "loss": 0.0433, "step": 85070 }, { "epoch": 2.876364988674397, "grad_norm": 0.15968981385231018, "learning_rate": 5.964292364635682e-06, "loss": 0.0442, "step": 85080 }, { "epoch": 2.8767030663646507, "grad_norm": 0.2381751835346222, "learning_rate": 5.956463143219682e-06, "loss": 0.0435, "step": 85090 }, { "epoch": 2.8770411440549037, "grad_norm": 0.21855777502059937, "learning_rate": 5.948638738338663e-06, "loss": 0.0388, "step": 85100 }, { "epoch": 2.877379221745157, "grad_norm": 0.15991634130477905, "learning_rate": 5.940819150848271e-06, "loss": 0.0373, "step": 85110 }, { "epoch": 2.8777172994354103, "grad_norm": 0.27453532814979553, "learning_rate": 5.933004381603663e-06, "loss": 0.0368, "step": 85120 }, { "epoch": 2.8780553771256634, "grad_norm": 0.16933000087738037, "learning_rate": 5.92519443145943e-06, "loss": 0.0455, "step": 85130 }, { "epoch": 2.8783934548159165, "grad_norm": 0.350647896528244, "learning_rate": 5.917389301269671e-06, "loss": 0.0433, "step": 85140 }, { "epoch": 2.87873153250617, "grad_norm": 0.23649737238883972, "learning_rate": 5.9095889918879435e-06, "loss": 0.0409, "step": 85150 }, { "epoch": 2.879069610196423, "grad_norm": 0.32161423563957214, "learning_rate": 5.901793504167286e-06, "loss": 0.0425, "step": 85160 }, { "epoch": 2.879407687886676, "grad_norm": 0.24969196319580078, "learning_rate": 5.894002838960172e-06, "loss": 0.0382, "step": 85170 }, { "epoch": 2.8797457655769296, "grad_norm": 0.23692984879016876, "learning_rate": 5.886216997118604e-06, "loss": 0.0445, "step": 85180 }, { "epoch": 2.8800838432671827, "grad_norm": 0.2530852258205414, "learning_rate": 5.878435979494023e-06, "loss": 0.0432, "step": 85190 }, { "epoch": 2.8804219209574358, "grad_norm": 0.20330189168453217, "learning_rate": 5.870659786937344e-06, "loss": 0.0438, "step": 85200 }, { "epoch": 2.8807599986476893, "grad_norm": 0.21245300769805908, "learning_rate": 5.862888420298962e-06, "loss": 0.0405, "step": 85210 }, { "epoch": 2.8810980763379423, "grad_norm": 0.17673981189727783, "learning_rate": 5.855121880428743e-06, "loss": 0.0441, "step": 85220 }, { "epoch": 2.8814361540281954, "grad_norm": 0.18896134197711945, "learning_rate": 5.84736016817603e-06, "loss": 0.038, "step": 85230 }, { "epoch": 2.881774231718449, "grad_norm": 0.2526026964187622, "learning_rate": 5.839603284389622e-06, "loss": 0.0456, "step": 85240 }, { "epoch": 2.882112309408702, "grad_norm": 0.2247627079486847, "learning_rate": 5.831851229917812e-06, "loss": 0.0455, "step": 85250 }, { "epoch": 2.882450387098955, "grad_norm": 0.7171489000320435, "learning_rate": 5.824104005608338e-06, "loss": 0.0426, "step": 85260 }, { "epoch": 2.8827884647892086, "grad_norm": 0.17647942900657654, "learning_rate": 5.81636161230843e-06, "loss": 0.0415, "step": 85270 }, { "epoch": 2.8831265424794617, "grad_norm": 0.2698588967323303, "learning_rate": 5.808624050864791e-06, "loss": 0.0414, "step": 85280 }, { "epoch": 2.883464620169715, "grad_norm": 0.29797783493995667, "learning_rate": 5.800891322123592e-06, "loss": 0.0471, "step": 85290 }, { "epoch": 2.8838026978599682, "grad_norm": 0.20285840332508087, "learning_rate": 5.793163426930454e-06, "loss": 0.0399, "step": 85300 }, { "epoch": 2.8841407755502213, "grad_norm": 0.1814260184764862, "learning_rate": 5.785440366130512e-06, "loss": 0.0404, "step": 85310 }, { "epoch": 2.884478853240475, "grad_norm": 0.2119099199771881, "learning_rate": 5.777722140568326e-06, "loss": 0.0434, "step": 85320 }, { "epoch": 2.884816930930728, "grad_norm": 0.2846077084541321, "learning_rate": 5.7700087510879685e-06, "loss": 0.0382, "step": 85330 }, { "epoch": 2.885155008620981, "grad_norm": 0.20481503009796143, "learning_rate": 5.762300198532933e-06, "loss": 0.0502, "step": 85340 }, { "epoch": 2.8854930863112345, "grad_norm": 0.22229304909706116, "learning_rate": 5.7545964837462606e-06, "loss": 0.0426, "step": 85350 }, { "epoch": 2.8858311640014875, "grad_norm": 0.35794928669929504, "learning_rate": 5.746897607570384e-06, "loss": 0.0459, "step": 85360 }, { "epoch": 2.8861692416917406, "grad_norm": 0.1959337294101715, "learning_rate": 5.7392035708472626e-06, "loss": 0.0477, "step": 85370 }, { "epoch": 2.886507319381994, "grad_norm": 0.31843605637550354, "learning_rate": 5.731514374418278e-06, "loss": 0.0482, "step": 85380 }, { "epoch": 2.886845397072247, "grad_norm": 0.2884267568588257, "learning_rate": 5.7238300191243325e-06, "loss": 0.0387, "step": 85390 }, { "epoch": 2.8871834747625003, "grad_norm": 0.18626375496387482, "learning_rate": 5.716150505805762e-06, "loss": 0.0433, "step": 85400 }, { "epoch": 2.8875215524527538, "grad_norm": 0.2589564621448517, "learning_rate": 5.708475835302407e-06, "loss": 0.0399, "step": 85410 }, { "epoch": 2.887859630143007, "grad_norm": 0.15570200979709625, "learning_rate": 5.70080600845353e-06, "loss": 0.0427, "step": 85420 }, { "epoch": 2.8881977078332604, "grad_norm": 0.1724643111228943, "learning_rate": 5.693141026097909e-06, "loss": 0.0418, "step": 85430 }, { "epoch": 2.8885357855235134, "grad_norm": 0.2134552150964737, "learning_rate": 5.6854808890737745e-06, "loss": 0.0378, "step": 85440 }, { "epoch": 2.8888738632137665, "grad_norm": 0.17193126678466797, "learning_rate": 5.6778255982188225e-06, "loss": 0.0403, "step": 85450 }, { "epoch": 2.88921194090402, "grad_norm": 0.20889656245708466, "learning_rate": 5.670175154370217e-06, "loss": 0.0466, "step": 85460 }, { "epoch": 2.889550018594273, "grad_norm": 0.33136966824531555, "learning_rate": 5.662529558364616e-06, "loss": 0.0419, "step": 85470 }, { "epoch": 2.889888096284526, "grad_norm": 0.2137003242969513, "learning_rate": 5.654888811038128e-06, "loss": 0.041, "step": 85480 }, { "epoch": 2.8902261739747797, "grad_norm": 0.19881115853786469, "learning_rate": 5.647252913226314e-06, "loss": 0.0476, "step": 85490 }, { "epoch": 2.8905642516650327, "grad_norm": 0.46967971324920654, "learning_rate": 5.639621865764255e-06, "loss": 0.0461, "step": 85500 }, { "epoch": 2.890902329355286, "grad_norm": 0.2724800109863281, "learning_rate": 5.63199566948644e-06, "loss": 0.0463, "step": 85510 }, { "epoch": 2.8912404070455393, "grad_norm": 0.4140506088733673, "learning_rate": 5.6243743252268835e-06, "loss": 0.0452, "step": 85520 }, { "epoch": 2.8915784847357924, "grad_norm": 0.376371294260025, "learning_rate": 5.616757833819009e-06, "loss": 0.041, "step": 85530 }, { "epoch": 2.8919165624260454, "grad_norm": 0.2331562638282776, "learning_rate": 5.609146196095794e-06, "loss": 0.0405, "step": 85540 }, { "epoch": 2.892254640116299, "grad_norm": 0.3346235156059265, "learning_rate": 5.6015394128896005e-06, "loss": 0.0458, "step": 85550 }, { "epoch": 2.892592717806552, "grad_norm": 0.6273505091667175, "learning_rate": 5.593937485032308e-06, "loss": 0.0447, "step": 85560 }, { "epoch": 2.892930795496805, "grad_norm": 0.7813854217529297, "learning_rate": 5.586340413355246e-06, "loss": 0.0419, "step": 85570 }, { "epoch": 2.8932688731870586, "grad_norm": 0.25005972385406494, "learning_rate": 5.578748198689227e-06, "loss": 0.0485, "step": 85580 }, { "epoch": 2.8936069508773117, "grad_norm": 0.20471207797527313, "learning_rate": 5.571160841864504e-06, "loss": 0.043, "step": 85590 }, { "epoch": 2.8939450285675647, "grad_norm": 0.27530980110168457, "learning_rate": 5.563578343710846e-06, "loss": 0.0434, "step": 85600 }, { "epoch": 2.8942831062578183, "grad_norm": 0.12114102393388748, "learning_rate": 5.556000705057446e-06, "loss": 0.0443, "step": 85610 }, { "epoch": 2.8946211839480713, "grad_norm": 0.2084575742483139, "learning_rate": 5.548427926732996e-06, "loss": 0.0395, "step": 85620 }, { "epoch": 2.8949592616383244, "grad_norm": 0.2311195731163025, "learning_rate": 5.540860009565629e-06, "loss": 0.0445, "step": 85630 }, { "epoch": 2.895297339328578, "grad_norm": 0.19413721561431885, "learning_rate": 5.533296954382966e-06, "loss": 0.0431, "step": 85640 }, { "epoch": 2.895635417018831, "grad_norm": 0.30769792199134827, "learning_rate": 5.525738762012095e-06, "loss": 0.0446, "step": 85650 }, { "epoch": 2.895973494709084, "grad_norm": 0.17919978499412537, "learning_rate": 5.518185433279577e-06, "loss": 0.042, "step": 85660 }, { "epoch": 2.8963115723993376, "grad_norm": 0.29286620020866394, "learning_rate": 5.510636969011407e-06, "loss": 0.0401, "step": 85670 }, { "epoch": 2.8966496500895906, "grad_norm": 0.20196394622325897, "learning_rate": 5.503093370033096e-06, "loss": 0.0426, "step": 85680 }, { "epoch": 2.8969877277798437, "grad_norm": 0.23892588913440704, "learning_rate": 5.495554637169603e-06, "loss": 0.0389, "step": 85690 }, { "epoch": 2.897325805470097, "grad_norm": 0.1980934739112854, "learning_rate": 5.488020771245328e-06, "loss": 0.0437, "step": 85700 }, { "epoch": 2.8976638831603503, "grad_norm": 0.30160486698150635, "learning_rate": 5.480491773084195e-06, "loss": 0.044, "step": 85710 }, { "epoch": 2.8980019608506034, "grad_norm": 0.15108513832092285, "learning_rate": 5.472967643509525e-06, "loss": 0.0411, "step": 85720 }, { "epoch": 2.898340038540857, "grad_norm": 0.2782422602176666, "learning_rate": 5.46544838334419e-06, "loss": 0.0428, "step": 85730 }, { "epoch": 2.89867811623111, "grad_norm": 0.3160097599029541, "learning_rate": 5.4579339934104525e-06, "loss": 0.0408, "step": 85740 }, { "epoch": 2.899016193921363, "grad_norm": 0.18017135560512543, "learning_rate": 5.450424474530091e-06, "loss": 0.0429, "step": 85750 }, { "epoch": 2.8993542716116165, "grad_norm": 0.26381969451904297, "learning_rate": 5.442919827524323e-06, "loss": 0.045, "step": 85760 }, { "epoch": 2.8996923493018696, "grad_norm": 0.24204625189304352, "learning_rate": 5.435420053213863e-06, "loss": 0.042, "step": 85770 }, { "epoch": 2.9000304269921227, "grad_norm": 0.20881307125091553, "learning_rate": 5.427925152418845e-06, "loss": 0.0423, "step": 85780 }, { "epoch": 2.900368504682376, "grad_norm": 0.22870071232318878, "learning_rate": 5.420435125958934e-06, "loss": 0.0438, "step": 85790 }, { "epoch": 2.9007065823726292, "grad_norm": 0.16115352511405945, "learning_rate": 5.412949974653209e-06, "loss": 0.0408, "step": 85800 }, { "epoch": 2.9010446600628823, "grad_norm": 0.2219133973121643, "learning_rate": 5.405469699320248e-06, "loss": 0.0397, "step": 85810 }, { "epoch": 2.901382737753136, "grad_norm": 0.2568829655647278, "learning_rate": 5.397994300778064e-06, "loss": 0.0425, "step": 85820 }, { "epoch": 2.901720815443389, "grad_norm": 0.261480450630188, "learning_rate": 5.39052377984417e-06, "loss": 0.0409, "step": 85830 }, { "epoch": 2.902058893133642, "grad_norm": 0.42402738332748413, "learning_rate": 5.383058137335523e-06, "loss": 0.0434, "step": 85840 }, { "epoch": 2.9023969708238955, "grad_norm": 0.1951465904712677, "learning_rate": 5.375597374068553e-06, "loss": 0.0392, "step": 85850 }, { "epoch": 2.9027350485141485, "grad_norm": 0.25864288210868835, "learning_rate": 5.368141490859163e-06, "loss": 0.0416, "step": 85860 }, { "epoch": 2.9030731262044016, "grad_norm": 0.2140623927116394, "learning_rate": 5.360690488522729e-06, "loss": 0.0441, "step": 85870 }, { "epoch": 2.903411203894655, "grad_norm": 0.17019455134868622, "learning_rate": 5.353244367874055e-06, "loss": 0.0401, "step": 85880 }, { "epoch": 2.903749281584908, "grad_norm": 0.2917027175426483, "learning_rate": 5.345803129727456e-06, "loss": 0.0413, "step": 85890 }, { "epoch": 2.9040873592751613, "grad_norm": 0.22372780740261078, "learning_rate": 5.338366774896697e-06, "loss": 0.0418, "step": 85900 }, { "epoch": 2.9044254369654148, "grad_norm": 0.24487370252609253, "learning_rate": 5.330935304194995e-06, "loss": 0.0401, "step": 85910 }, { "epoch": 2.904763514655668, "grad_norm": 0.27491915225982666, "learning_rate": 5.323508718435044e-06, "loss": 0.0433, "step": 85920 }, { "epoch": 2.905101592345921, "grad_norm": 0.245924711227417, "learning_rate": 5.316087018429011e-06, "loss": 0.0401, "step": 85930 }, { "epoch": 2.9054396700361744, "grad_norm": 0.2847166955471039, "learning_rate": 5.30867020498853e-06, "loss": 0.0432, "step": 85940 }, { "epoch": 2.9057777477264275, "grad_norm": 0.20514898002147675, "learning_rate": 5.30125827892467e-06, "loss": 0.0405, "step": 85950 }, { "epoch": 2.9061158254166806, "grad_norm": 0.23580259084701538, "learning_rate": 5.293851241048015e-06, "loss": 0.0422, "step": 85960 }, { "epoch": 2.906453903106934, "grad_norm": 0.5873736143112183, "learning_rate": 5.286449092168555e-06, "loss": 0.0414, "step": 85970 }, { "epoch": 2.906791980797187, "grad_norm": 0.17816683650016785, "learning_rate": 5.2790518330958105e-06, "loss": 0.0481, "step": 85980 }, { "epoch": 2.90713005848744, "grad_norm": 0.15973134338855743, "learning_rate": 5.271659464638717e-06, "loss": 0.04, "step": 85990 }, { "epoch": 2.9074681361776937, "grad_norm": 0.19318217039108276, "learning_rate": 5.264271987605701e-06, "loss": 0.0396, "step": 86000 }, { "epoch": 2.907806213867947, "grad_norm": 0.18517756462097168, "learning_rate": 5.256889402804632e-06, "loss": 0.0385, "step": 86010 }, { "epoch": 2.9081442915582, "grad_norm": 0.2686929702758789, "learning_rate": 5.249511711042881e-06, "loss": 0.0427, "step": 86020 }, { "epoch": 2.9084823692484534, "grad_norm": 0.17656974494457245, "learning_rate": 5.242138913127237e-06, "loss": 0.0379, "step": 86030 }, { "epoch": 2.9088204469387064, "grad_norm": 0.54448002576828, "learning_rate": 5.234771009863987e-06, "loss": 0.0458, "step": 86040 }, { "epoch": 2.9091585246289595, "grad_norm": 0.2777174413204193, "learning_rate": 5.227408002058881e-06, "loss": 0.0419, "step": 86050 }, { "epoch": 2.909496602319213, "grad_norm": 0.20259219408035278, "learning_rate": 5.220049890517126e-06, "loss": 0.0482, "step": 86060 }, { "epoch": 2.909834680009466, "grad_norm": 0.2514716386795044, "learning_rate": 5.212696676043377e-06, "loss": 0.0411, "step": 86070 }, { "epoch": 2.910172757699719, "grad_norm": 0.368064820766449, "learning_rate": 5.205348359441798e-06, "loss": 0.0434, "step": 86080 }, { "epoch": 2.9105108353899727, "grad_norm": 0.23958927392959595, "learning_rate": 5.198004941515966e-06, "loss": 0.0456, "step": 86090 }, { "epoch": 2.9108489130802258, "grad_norm": 0.3012673258781433, "learning_rate": 5.190666423068952e-06, "loss": 0.0438, "step": 86100 }, { "epoch": 2.911186990770479, "grad_norm": 0.2025042176246643, "learning_rate": 5.183332804903284e-06, "loss": 0.0482, "step": 86110 }, { "epoch": 2.9115250684607323, "grad_norm": 0.7313923239707947, "learning_rate": 5.176004087820979e-06, "loss": 0.0399, "step": 86120 }, { "epoch": 2.9118631461509854, "grad_norm": 0.24315257370471954, "learning_rate": 5.168680272623455e-06, "loss": 0.0466, "step": 86130 }, { "epoch": 2.9122012238412385, "grad_norm": 0.2273818850517273, "learning_rate": 5.161361360111661e-06, "loss": 0.0399, "step": 86140 }, { "epoch": 2.912539301531492, "grad_norm": 0.20261763036251068, "learning_rate": 5.154047351085984e-06, "loss": 0.04, "step": 86150 }, { "epoch": 2.912877379221745, "grad_norm": 0.6031486988067627, "learning_rate": 5.1467382463462385e-06, "loss": 0.0452, "step": 86160 }, { "epoch": 2.913215456911998, "grad_norm": 0.20975735783576965, "learning_rate": 5.139434046691788e-06, "loss": 0.0405, "step": 86170 }, { "epoch": 2.9135535346022516, "grad_norm": 0.1690593957901001, "learning_rate": 5.13213475292137e-06, "loss": 0.0435, "step": 86180 }, { "epoch": 2.9138916122925047, "grad_norm": 0.19048619270324707, "learning_rate": 5.12484036583325e-06, "loss": 0.0422, "step": 86190 }, { "epoch": 2.9142296899827578, "grad_norm": 0.23153673112392426, "learning_rate": 5.117550886225109e-06, "loss": 0.0413, "step": 86200 }, { "epoch": 2.9145677676730113, "grad_norm": 0.2503672242164612, "learning_rate": 5.11026631489413e-06, "loss": 0.0453, "step": 86210 }, { "epoch": 2.9149058453632644, "grad_norm": 0.18549273908138275, "learning_rate": 5.102986652636926e-06, "loss": 0.0455, "step": 86220 }, { "epoch": 2.9152439230535174, "grad_norm": 0.35506707429885864, "learning_rate": 5.095711900249606e-06, "loss": 0.0402, "step": 86230 }, { "epoch": 2.915582000743771, "grad_norm": 0.441000759601593, "learning_rate": 5.088442058527715e-06, "loss": 0.0434, "step": 86240 }, { "epoch": 2.915920078434024, "grad_norm": 0.2404244989156723, "learning_rate": 5.08117712826629e-06, "loss": 0.0426, "step": 86250 }, { "epoch": 2.9162581561242775, "grad_norm": 0.4182519316673279, "learning_rate": 5.073917110259785e-06, "loss": 0.0449, "step": 86260 }, { "epoch": 2.9165962338145306, "grad_norm": 0.1580563485622406, "learning_rate": 5.066662005302175e-06, "loss": 0.0458, "step": 86270 }, { "epoch": 2.9169343115047837, "grad_norm": 0.3870774507522583, "learning_rate": 5.059411814186838e-06, "loss": 0.0427, "step": 86280 }, { "epoch": 2.917272389195037, "grad_norm": 0.27111056447029114, "learning_rate": 5.05216653770666e-06, "loss": 0.0412, "step": 86290 }, { "epoch": 2.9176104668852902, "grad_norm": 0.26100119948387146, "learning_rate": 5.044926176653969e-06, "loss": 0.0422, "step": 86300 }, { "epoch": 2.9179485445755433, "grad_norm": 0.17420503497123718, "learning_rate": 5.037690731820571e-06, "loss": 0.0484, "step": 86310 }, { "epoch": 2.918286622265797, "grad_norm": 0.16321665048599243, "learning_rate": 5.030460203997711e-06, "loss": 0.0408, "step": 86320 }, { "epoch": 2.91862469995605, "grad_norm": 1.0003923177719116, "learning_rate": 5.0232345939761174e-06, "loss": 0.0407, "step": 86330 }, { "epoch": 2.918962777646303, "grad_norm": 0.15404663980007172, "learning_rate": 5.016013902545957e-06, "loss": 0.0459, "step": 86340 }, { "epoch": 2.9193008553365565, "grad_norm": 0.2360449880361557, "learning_rate": 5.008798130496889e-06, "loss": 0.0393, "step": 86350 }, { "epoch": 2.9196389330268095, "grad_norm": 0.1664857566356659, "learning_rate": 5.001587278618008e-06, "loss": 0.0388, "step": 86360 }, { "epoch": 2.9199770107170626, "grad_norm": 0.1542678028345108, "learning_rate": 4.994381347697891e-06, "loss": 0.0438, "step": 86370 }, { "epoch": 2.920315088407316, "grad_norm": 0.5478786826133728, "learning_rate": 4.987180338524572e-06, "loss": 0.0404, "step": 86380 }, { "epoch": 2.920653166097569, "grad_norm": 0.27070853114128113, "learning_rate": 4.9799842518855275e-06, "loss": 0.0402, "step": 86390 }, { "epoch": 2.9209912437878227, "grad_norm": 0.42666536569595337, "learning_rate": 4.9727930885677285e-06, "loss": 0.0409, "step": 86400 }, { "epoch": 2.921329321478076, "grad_norm": 0.4242175221443176, "learning_rate": 4.965606849357573e-06, "loss": 0.0499, "step": 86410 }, { "epoch": 2.921667399168329, "grad_norm": 0.24240750074386597, "learning_rate": 4.9584255350409445e-06, "loss": 0.0423, "step": 86420 }, { "epoch": 2.9220054768585824, "grad_norm": 0.22896060347557068, "learning_rate": 4.951249146403175e-06, "loss": 0.0494, "step": 86430 }, { "epoch": 2.9223435545488354, "grad_norm": 0.31297948956489563, "learning_rate": 4.9440776842290815e-06, "loss": 0.0442, "step": 86440 }, { "epoch": 2.9226816322390885, "grad_norm": 0.27559444308280945, "learning_rate": 4.936911149302903e-06, "loss": 0.0488, "step": 86450 }, { "epoch": 2.923019709929342, "grad_norm": 0.26911935210227966, "learning_rate": 4.929749542408379e-06, "loss": 0.0388, "step": 86460 }, { "epoch": 2.923357787619595, "grad_norm": 0.3071913719177246, "learning_rate": 4.922592864328673e-06, "loss": 0.0443, "step": 86470 }, { "epoch": 2.923695865309848, "grad_norm": 0.1874462217092514, "learning_rate": 4.915441115846436e-06, "loss": 0.0442, "step": 86480 }, { "epoch": 2.9240339430001017, "grad_norm": 0.19210100173950195, "learning_rate": 4.908294297743782e-06, "loss": 0.0401, "step": 86490 }, { "epoch": 2.9243720206903547, "grad_norm": 0.31695595383644104, "learning_rate": 4.901152410802273e-06, "loss": 0.0431, "step": 86500 }, { "epoch": 2.924710098380608, "grad_norm": 0.7913495302200317, "learning_rate": 4.894015455802919e-06, "loss": 0.042, "step": 86510 }, { "epoch": 2.9250481760708613, "grad_norm": 0.3047991394996643, "learning_rate": 4.8868834335262346e-06, "loss": 0.0416, "step": 86520 }, { "epoch": 2.9253862537611144, "grad_norm": 0.3026181757450104, "learning_rate": 4.87975634475214e-06, "loss": 0.0429, "step": 86530 }, { "epoch": 2.9257243314513675, "grad_norm": 0.8727990984916687, "learning_rate": 4.8726341902600614e-06, "loss": 0.0497, "step": 86540 }, { "epoch": 2.926062409141621, "grad_norm": 0.19254404306411743, "learning_rate": 4.865516970828843e-06, "loss": 0.0401, "step": 86550 }, { "epoch": 2.926400486831874, "grad_norm": 0.3862435519695282, "learning_rate": 4.85840468723685e-06, "loss": 0.0392, "step": 86560 }, { "epoch": 2.926738564522127, "grad_norm": 0.27247053384780884, "learning_rate": 4.85129734026184e-06, "loss": 0.0412, "step": 86570 }, { "epoch": 2.9270766422123806, "grad_norm": 0.215585395693779, "learning_rate": 4.844194930681073e-06, "loss": 0.0406, "step": 86580 }, { "epoch": 2.9274147199026337, "grad_norm": 0.20303380489349365, "learning_rate": 4.837097459271267e-06, "loss": 0.0495, "step": 86590 }, { "epoch": 2.9277527975928868, "grad_norm": 0.31301578879356384, "learning_rate": 4.830004926808573e-06, "loss": 0.0388, "step": 86600 }, { "epoch": 2.9280908752831403, "grad_norm": 0.2721492052078247, "learning_rate": 4.822917334068627e-06, "loss": 0.0455, "step": 86610 }, { "epoch": 2.9284289529733933, "grad_norm": 0.27625221014022827, "learning_rate": 4.81583468182652e-06, "loss": 0.0415, "step": 86620 }, { "epoch": 2.9287670306636464, "grad_norm": 0.21392619609832764, "learning_rate": 4.808756970856809e-06, "loss": 0.0464, "step": 86630 }, { "epoch": 2.9291051083539, "grad_norm": 0.29178380966186523, "learning_rate": 4.801684201933476e-06, "loss": 0.0431, "step": 86640 }, { "epoch": 2.929443186044153, "grad_norm": 0.17188258469104767, "learning_rate": 4.794616375830019e-06, "loss": 0.0433, "step": 86650 }, { "epoch": 2.929781263734406, "grad_norm": 0.20479080080986023, "learning_rate": 4.787553493319341e-06, "loss": 0.044, "step": 86660 }, { "epoch": 2.9301193414246596, "grad_norm": 0.29512643814086914, "learning_rate": 4.780495555173836e-06, "loss": 0.0382, "step": 86670 }, { "epoch": 2.9304574191149126, "grad_norm": 0.29527372121810913, "learning_rate": 4.7734425621653475e-06, "loss": 0.0452, "step": 86680 }, { "epoch": 2.9307954968051657, "grad_norm": 0.38150736689567566, "learning_rate": 4.766394515065198e-06, "loss": 0.0448, "step": 86690 }, { "epoch": 2.9311335744954192, "grad_norm": 0.19622910022735596, "learning_rate": 4.75935141464412e-06, "loss": 0.0477, "step": 86700 }, { "epoch": 2.9314716521856723, "grad_norm": 0.5559358596801758, "learning_rate": 4.75231326167237e-06, "loss": 0.0422, "step": 86710 }, { "epoch": 2.9318097298759254, "grad_norm": 0.17228926718235016, "learning_rate": 4.745280056919599e-06, "loss": 0.0382, "step": 86720 }, { "epoch": 2.932147807566179, "grad_norm": 0.21629872918128967, "learning_rate": 4.738251801154975e-06, "loss": 0.0404, "step": 86730 }, { "epoch": 2.932485885256432, "grad_norm": 1.1817653179168701, "learning_rate": 4.731228495147061e-06, "loss": 0.0384, "step": 86740 }, { "epoch": 2.932823962946685, "grad_norm": 0.3028629720211029, "learning_rate": 4.7242101396639595e-06, "loss": 0.0429, "step": 86750 }, { "epoch": 2.9331620406369385, "grad_norm": 0.1961383819580078, "learning_rate": 4.717196735473151e-06, "loss": 0.0367, "step": 86760 }, { "epoch": 2.9335001183271916, "grad_norm": 0.43743056058883667, "learning_rate": 4.710188283341643e-06, "loss": 0.0397, "step": 86770 }, { "epoch": 2.9338381960174447, "grad_norm": 0.2149888575077057, "learning_rate": 4.703184784035835e-06, "loss": 0.0432, "step": 86780 }, { "epoch": 2.934176273707698, "grad_norm": 0.31391167640686035, "learning_rate": 4.696186238321642e-06, "loss": 0.0427, "step": 86790 }, { "epoch": 2.9345143513979512, "grad_norm": 0.28692615032196045, "learning_rate": 4.689192646964408e-06, "loss": 0.0378, "step": 86800 }, { "epoch": 2.9348524290882043, "grad_norm": 0.27982383966445923, "learning_rate": 4.682204010728952e-06, "loss": 0.0423, "step": 86810 }, { "epoch": 2.935190506778458, "grad_norm": 0.33050772547721863, "learning_rate": 4.67522033037952e-06, "loss": 0.0414, "step": 86820 }, { "epoch": 2.935528584468711, "grad_norm": 0.19424250721931458, "learning_rate": 4.668241606679846e-06, "loss": 0.0474, "step": 86830 }, { "epoch": 2.935866662158964, "grad_norm": 0.2342027723789215, "learning_rate": 4.66126784039313e-06, "loss": 0.0484, "step": 86840 }, { "epoch": 2.9362047398492175, "grad_norm": 0.3081638514995575, "learning_rate": 4.6542990322819836e-06, "loss": 0.0432, "step": 86850 }, { "epoch": 2.9365428175394706, "grad_norm": 0.7609987258911133, "learning_rate": 4.64733518310852e-06, "loss": 0.0481, "step": 86860 }, { "epoch": 2.9368808952297236, "grad_norm": 0.12868891656398773, "learning_rate": 4.640376293634291e-06, "loss": 0.0374, "step": 86870 }, { "epoch": 2.937218972919977, "grad_norm": 0.16666880249977112, "learning_rate": 4.633422364620327e-06, "loss": 0.0434, "step": 86880 }, { "epoch": 2.93755705061023, "grad_norm": 0.29427096247673035, "learning_rate": 4.6264733968270745e-06, "loss": 0.0422, "step": 86890 }, { "epoch": 2.9378951283004833, "grad_norm": 0.2846672832965851, "learning_rate": 4.619529391014482e-06, "loss": 0.0435, "step": 86900 }, { "epoch": 2.938233205990737, "grad_norm": 0.2985491454601288, "learning_rate": 4.61259034794192e-06, "loss": 0.0445, "step": 86910 }, { "epoch": 2.93857128368099, "grad_norm": 0.19674429297447205, "learning_rate": 4.605656268368247e-06, "loss": 0.0408, "step": 86920 }, { "epoch": 2.938909361371243, "grad_norm": 0.25879284739494324, "learning_rate": 4.598727153051735e-06, "loss": 0.0425, "step": 86930 }, { "epoch": 2.9392474390614964, "grad_norm": 0.1653289943933487, "learning_rate": 4.591803002750178e-06, "loss": 0.0331, "step": 86940 }, { "epoch": 2.9395855167517495, "grad_norm": 0.22928935289382935, "learning_rate": 4.58488381822077e-06, "loss": 0.0395, "step": 86950 }, { "epoch": 2.9399235944420026, "grad_norm": 0.2025580257177353, "learning_rate": 4.5779696002201946e-06, "loss": 0.0434, "step": 86960 }, { "epoch": 2.940261672132256, "grad_norm": 0.8074488043785095, "learning_rate": 4.571060349504563e-06, "loss": 0.0452, "step": 86970 }, { "epoch": 2.940599749822509, "grad_norm": 0.39700305461883545, "learning_rate": 4.564156066829478e-06, "loss": 0.0523, "step": 86980 }, { "epoch": 2.9409378275127622, "grad_norm": 0.25342148542404175, "learning_rate": 4.557256752949957e-06, "loss": 0.0416, "step": 86990 }, { "epoch": 2.9412759052030157, "grad_norm": 0.20828777551651, "learning_rate": 4.550362408620534e-06, "loss": 0.0445, "step": 87000 }, { "epoch": 2.941613982893269, "grad_norm": 0.1745227724313736, "learning_rate": 4.543473034595136e-06, "loss": 0.0431, "step": 87010 }, { "epoch": 2.941952060583522, "grad_norm": 0.1369548887014389, "learning_rate": 4.536588631627192e-06, "loss": 0.041, "step": 87020 }, { "epoch": 2.9422901382737754, "grad_norm": 0.2294950783252716, "learning_rate": 4.5297092004695565e-06, "loss": 0.0461, "step": 87030 }, { "epoch": 2.9426282159640285, "grad_norm": 0.25917884707450867, "learning_rate": 4.522834741874555e-06, "loss": 0.042, "step": 87040 }, { "epoch": 2.9429662936542815, "grad_norm": 0.17893879115581512, "learning_rate": 4.515965256593974e-06, "loss": 0.0413, "step": 87050 }, { "epoch": 2.943304371344535, "grad_norm": 0.29846349358558655, "learning_rate": 4.509100745379058e-06, "loss": 0.0425, "step": 87060 }, { "epoch": 2.943642449034788, "grad_norm": 0.3771840035915375, "learning_rate": 4.502241208980479e-06, "loss": 0.0416, "step": 87070 }, { "epoch": 2.943980526725041, "grad_norm": 0.22680050134658813, "learning_rate": 4.495386648148398e-06, "loss": 0.043, "step": 87080 }, { "epoch": 2.9443186044152947, "grad_norm": 0.19721776247024536, "learning_rate": 4.488537063632431e-06, "loss": 0.0386, "step": 87090 }, { "epoch": 2.9446566821055478, "grad_norm": 0.24672821164131165, "learning_rate": 4.481692456181608e-06, "loss": 0.0454, "step": 87100 }, { "epoch": 2.944994759795801, "grad_norm": 0.23649290204048157, "learning_rate": 4.474852826544479e-06, "loss": 0.0452, "step": 87110 }, { "epoch": 2.9453328374860543, "grad_norm": 0.24946974217891693, "learning_rate": 4.468018175468974e-06, "loss": 0.0448, "step": 87120 }, { "epoch": 2.9456709151763074, "grad_norm": 0.280923455953598, "learning_rate": 4.461188503702568e-06, "loss": 0.0413, "step": 87130 }, { "epoch": 2.9460089928665605, "grad_norm": 0.2788771688938141, "learning_rate": 4.454363811992107e-06, "loss": 0.0417, "step": 87140 }, { "epoch": 2.946347070556814, "grad_norm": 0.39420178532600403, "learning_rate": 4.447544101083956e-06, "loss": 0.0397, "step": 87150 }, { "epoch": 2.946685148247067, "grad_norm": 0.2755165994167328, "learning_rate": 4.440729371723884e-06, "loss": 0.0406, "step": 87160 }, { "epoch": 2.94702322593732, "grad_norm": 0.21472382545471191, "learning_rate": 4.433919624657162e-06, "loss": 0.04, "step": 87170 }, { "epoch": 2.9473613036275736, "grad_norm": 0.22284561395645142, "learning_rate": 4.427114860628462e-06, "loss": 0.0428, "step": 87180 }, { "epoch": 2.9476993813178267, "grad_norm": 0.2637052834033966, "learning_rate": 4.420315080381982e-06, "loss": 0.0438, "step": 87190 }, { "epoch": 2.94803745900808, "grad_norm": 0.131662517786026, "learning_rate": 4.4135202846613045e-06, "loss": 0.0399, "step": 87200 }, { "epoch": 2.9483755366983333, "grad_norm": 0.33097654581069946, "learning_rate": 4.406730474209525e-06, "loss": 0.0432, "step": 87210 }, { "epoch": 2.9487136143885864, "grad_norm": 0.3831000328063965, "learning_rate": 4.399945649769138e-06, "loss": 0.0441, "step": 87220 }, { "epoch": 2.94905169207884, "grad_norm": 0.185815691947937, "learning_rate": 4.393165812082151e-06, "loss": 0.0454, "step": 87230 }, { "epoch": 2.949389769769093, "grad_norm": 0.25120607018470764, "learning_rate": 4.386390961889969e-06, "loss": 0.0414, "step": 87240 }, { "epoch": 2.949727847459346, "grad_norm": 0.17431654036045074, "learning_rate": 4.3796210999334895e-06, "loss": 0.0455, "step": 87250 }, { "epoch": 2.9500659251495995, "grad_norm": 1.695639967918396, "learning_rate": 4.372856226953065e-06, "loss": 0.0448, "step": 87260 }, { "epoch": 2.9504040028398526, "grad_norm": 0.20150837302207947, "learning_rate": 4.366096343688486e-06, "loss": 0.0426, "step": 87270 }, { "epoch": 2.9507420805301057, "grad_norm": 0.19821414351463318, "learning_rate": 4.359341450878995e-06, "loss": 0.0448, "step": 87280 }, { "epoch": 2.951080158220359, "grad_norm": 0.2556212842464447, "learning_rate": 4.352591549263302e-06, "loss": 0.0429, "step": 87290 }, { "epoch": 2.9514182359106123, "grad_norm": 0.30998578667640686, "learning_rate": 4.345846639579582e-06, "loss": 0.0471, "step": 87300 }, { "epoch": 2.9517563136008653, "grad_norm": 0.2210935652256012, "learning_rate": 4.339106722565417e-06, "loss": 0.0425, "step": 87310 }, { "epoch": 2.952094391291119, "grad_norm": 0.2403521090745926, "learning_rate": 4.332371798957896e-06, "loss": 0.0426, "step": 87320 }, { "epoch": 2.952432468981372, "grad_norm": 0.13427643477916718, "learning_rate": 4.3256418694935295e-06, "loss": 0.0405, "step": 87330 }, { "epoch": 2.952770546671625, "grad_norm": 0.19893234968185425, "learning_rate": 4.3189169349083125e-06, "loss": 0.0449, "step": 87340 }, { "epoch": 2.9531086243618785, "grad_norm": 0.22608120739459991, "learning_rate": 4.312196995937645e-06, "loss": 0.0466, "step": 87350 }, { "epoch": 2.9534467020521316, "grad_norm": 0.15120002627372742, "learning_rate": 4.305482053316434e-06, "loss": 0.0396, "step": 87360 }, { "epoch": 2.953784779742385, "grad_norm": 0.2276599109172821, "learning_rate": 4.2987721077789925e-06, "loss": 0.0494, "step": 87370 }, { "epoch": 2.954122857432638, "grad_norm": 0.2128850370645523, "learning_rate": 4.2920671600591325e-06, "loss": 0.0455, "step": 87380 }, { "epoch": 2.954460935122891, "grad_norm": 0.39875030517578125, "learning_rate": 4.285367210890084e-06, "loss": 0.0453, "step": 87390 }, { "epoch": 2.9547990128131447, "grad_norm": 0.3433224856853485, "learning_rate": 4.278672261004551e-06, "loss": 0.037, "step": 87400 }, { "epoch": 2.955137090503398, "grad_norm": 0.14469748735427856, "learning_rate": 4.2719823111346704e-06, "loss": 0.039, "step": 87410 }, { "epoch": 2.955475168193651, "grad_norm": 0.3222343325614929, "learning_rate": 4.265297362012061e-06, "loss": 0.0376, "step": 87420 }, { "epoch": 2.9558132458839044, "grad_norm": 0.3074503242969513, "learning_rate": 4.258617414367766e-06, "loss": 0.04, "step": 87430 }, { "epoch": 2.9561513235741574, "grad_norm": 0.28190094232559204, "learning_rate": 4.251942468932296e-06, "loss": 0.0442, "step": 87440 }, { "epoch": 2.9564894012644105, "grad_norm": 0.9504743814468384, "learning_rate": 4.2452725264356175e-06, "loss": 0.0442, "step": 87450 }, { "epoch": 2.956827478954664, "grad_norm": 0.17750895023345947, "learning_rate": 4.2386075876071575e-06, "loss": 0.0437, "step": 87460 }, { "epoch": 2.957165556644917, "grad_norm": 0.38171520829200745, "learning_rate": 4.231947653175755e-06, "loss": 0.0424, "step": 87470 }, { "epoch": 2.95750363433517, "grad_norm": 0.3181806802749634, "learning_rate": 4.225292723869761e-06, "loss": 0.0436, "step": 87480 }, { "epoch": 2.9578417120254237, "grad_norm": 0.2752739191055298, "learning_rate": 4.218642800416916e-06, "loss": 0.0435, "step": 87490 }, { "epoch": 2.9581797897156767, "grad_norm": 0.30683672428131104, "learning_rate": 4.211997883544472e-06, "loss": 0.0436, "step": 87500 }, { "epoch": 2.95851786740593, "grad_norm": 0.3229459226131439, "learning_rate": 4.205357973979096e-06, "loss": 0.0414, "step": 87510 }, { "epoch": 2.9588559450961833, "grad_norm": 0.1645316779613495, "learning_rate": 4.198723072446925e-06, "loss": 0.0441, "step": 87520 }, { "epoch": 2.9591940227864364, "grad_norm": 0.4201529622077942, "learning_rate": 4.192093179673534e-06, "loss": 0.0456, "step": 87530 }, { "epoch": 2.9595321004766895, "grad_norm": 0.18042689561843872, "learning_rate": 4.185468296383959e-06, "loss": 0.0393, "step": 87540 }, { "epoch": 2.959870178166943, "grad_norm": 0.36536964774131775, "learning_rate": 4.1788484233027005e-06, "loss": 0.0476, "step": 87550 }, { "epoch": 2.960208255857196, "grad_norm": 0.35543012619018555, "learning_rate": 4.172233561153677e-06, "loss": 0.0428, "step": 87560 }, { "epoch": 2.960546333547449, "grad_norm": 0.33514344692230225, "learning_rate": 4.165623710660288e-06, "loss": 0.0456, "step": 87570 }, { "epoch": 2.9608844112377026, "grad_norm": 0.30461376905441284, "learning_rate": 4.1590188725453835e-06, "loss": 0.0451, "step": 87580 }, { "epoch": 2.9612224889279557, "grad_norm": 0.21137668192386627, "learning_rate": 4.152419047531259e-06, "loss": 0.0358, "step": 87590 }, { "epoch": 2.9615605666182088, "grad_norm": 0.15875343978405, "learning_rate": 4.145824236339651e-06, "loss": 0.0395, "step": 87600 }, { "epoch": 2.9618986443084623, "grad_norm": 0.23106245696544647, "learning_rate": 4.1392344396917695e-06, "loss": 0.0472, "step": 87610 }, { "epoch": 2.9622367219987153, "grad_norm": 0.1561436653137207, "learning_rate": 4.132649658308252e-06, "loss": 0.0471, "step": 87620 }, { "epoch": 2.9625747996889684, "grad_norm": 0.16896095871925354, "learning_rate": 4.126069892909213e-06, "loss": 0.0406, "step": 87630 }, { "epoch": 2.962912877379222, "grad_norm": 0.24020476639270782, "learning_rate": 4.119495144214191e-06, "loss": 0.0434, "step": 87640 }, { "epoch": 2.963250955069475, "grad_norm": 0.19115519523620605, "learning_rate": 4.112925412942214e-06, "loss": 0.0435, "step": 87650 }, { "epoch": 2.963589032759728, "grad_norm": 0.20225007832050323, "learning_rate": 4.106360699811712e-06, "loss": 0.0416, "step": 87660 }, { "epoch": 2.9639271104499816, "grad_norm": 0.19643373787403107, "learning_rate": 4.099801005540616e-06, "loss": 0.0417, "step": 87670 }, { "epoch": 2.9642651881402347, "grad_norm": 0.2726350724697113, "learning_rate": 4.093246330846256e-06, "loss": 0.0437, "step": 87680 }, { "epoch": 2.9646032658304877, "grad_norm": 0.47543174028396606, "learning_rate": 4.086696676445467e-06, "loss": 0.0456, "step": 87690 }, { "epoch": 2.9649413435207412, "grad_norm": 0.2614290714263916, "learning_rate": 4.0801520430544914e-06, "loss": 0.0392, "step": 87700 }, { "epoch": 2.9652794212109943, "grad_norm": 0.21657520532608032, "learning_rate": 4.073612431389062e-06, "loss": 0.0406, "step": 87710 }, { "epoch": 2.9656174989012474, "grad_norm": 0.29527273774147034, "learning_rate": 4.067077842164319e-06, "loss": 0.0385, "step": 87720 }, { "epoch": 2.965955576591501, "grad_norm": 0.22736018896102905, "learning_rate": 4.0605482760948945e-06, "loss": 0.0473, "step": 87730 }, { "epoch": 2.966293654281754, "grad_norm": 0.530207633972168, "learning_rate": 4.054023733894829e-06, "loss": 0.0455, "step": 87740 }, { "epoch": 2.966631731972007, "grad_norm": 0.24469348788261414, "learning_rate": 4.047504216277653e-06, "loss": 0.0423, "step": 87750 }, { "epoch": 2.9669698096622605, "grad_norm": 0.21704885363578796, "learning_rate": 4.040989723956323e-06, "loss": 0.0392, "step": 87760 }, { "epoch": 2.9673078873525136, "grad_norm": 0.24880342185497284, "learning_rate": 4.03448025764327e-06, "loss": 0.0396, "step": 87770 }, { "epoch": 2.9676459650427667, "grad_norm": 0.19057916104793549, "learning_rate": 4.027975818050339e-06, "loss": 0.0384, "step": 87780 }, { "epoch": 2.96798404273302, "grad_norm": 0.305269330739975, "learning_rate": 4.021476405888858e-06, "loss": 0.0392, "step": 87790 }, { "epoch": 2.9683221204232733, "grad_norm": 0.29267600178718567, "learning_rate": 4.01498202186959e-06, "loss": 0.043, "step": 87800 }, { "epoch": 2.9686601981135263, "grad_norm": 0.39810582995414734, "learning_rate": 4.008492666702751e-06, "loss": 0.0417, "step": 87810 }, { "epoch": 2.96899827580378, "grad_norm": 0.23264552652835846, "learning_rate": 4.002008341098001e-06, "loss": 0.0391, "step": 87820 }, { "epoch": 2.969336353494033, "grad_norm": 0.2850041687488556, "learning_rate": 3.995529045764462e-06, "loss": 0.0435, "step": 87830 }, { "epoch": 2.969674431184286, "grad_norm": 0.2388620376586914, "learning_rate": 3.989054781410712e-06, "loss": 0.0411, "step": 87840 }, { "epoch": 2.9700125088745395, "grad_norm": 0.37435030937194824, "learning_rate": 3.982585548744744e-06, "loss": 0.0404, "step": 87850 }, { "epoch": 2.9703505865647926, "grad_norm": 0.18527784943580627, "learning_rate": 3.9761213484740435e-06, "loss": 0.0419, "step": 87860 }, { "epoch": 2.9706886642550456, "grad_norm": 0.2948632538318634, "learning_rate": 3.9696621813055055e-06, "loss": 0.0372, "step": 87870 }, { "epoch": 2.971026741945299, "grad_norm": 0.26989343762397766, "learning_rate": 3.9632080479455044e-06, "loss": 0.0458, "step": 87880 }, { "epoch": 2.971364819635552, "grad_norm": 0.47681429982185364, "learning_rate": 3.956758949099859e-06, "loss": 0.0445, "step": 87890 }, { "epoch": 2.9717028973258053, "grad_norm": 0.3540748059749603, "learning_rate": 3.9503148854738385e-06, "loss": 0.0449, "step": 87900 }, { "epoch": 2.972040975016059, "grad_norm": 0.47680333256721497, "learning_rate": 3.943875857772139e-06, "loss": 0.0414, "step": 87910 }, { "epoch": 2.972379052706312, "grad_norm": 0.22057053446769714, "learning_rate": 3.9374418666989375e-06, "loss": 0.0424, "step": 87920 }, { "epoch": 2.972717130396565, "grad_norm": 0.19032859802246094, "learning_rate": 3.931012912957838e-06, "loss": 0.0452, "step": 87930 }, { "epoch": 2.9730552080868184, "grad_norm": 0.22243434190750122, "learning_rate": 3.9245889972519045e-06, "loss": 0.04, "step": 87940 }, { "epoch": 2.9733932857770715, "grad_norm": 0.34537050127983093, "learning_rate": 3.9181701202836265e-06, "loss": 0.0455, "step": 87950 }, { "epoch": 2.9737313634673246, "grad_norm": 0.277309775352478, "learning_rate": 3.911756282755003e-06, "loss": 0.0428, "step": 87960 }, { "epoch": 2.974069441157578, "grad_norm": 0.216602623462677, "learning_rate": 3.905347485367411e-06, "loss": 0.0385, "step": 87970 }, { "epoch": 2.974407518847831, "grad_norm": 0.19106467068195343, "learning_rate": 3.898943728821725e-06, "loss": 0.05, "step": 87980 }, { "epoch": 2.9747455965380842, "grad_norm": 0.18904399871826172, "learning_rate": 3.892545013818227e-06, "loss": 0.0396, "step": 87990 }, { "epoch": 2.9750836742283377, "grad_norm": 0.4909493029117584, "learning_rate": 3.886151341056687e-06, "loss": 0.0424, "step": 88000 }, { "epoch": 2.975421751918591, "grad_norm": 0.2892451882362366, "learning_rate": 3.8797627112363066e-06, "loss": 0.04, "step": 88010 }, { "epoch": 2.975759829608844, "grad_norm": 0.30779024958610535, "learning_rate": 3.873379125055743e-06, "loss": 0.045, "step": 88020 }, { "epoch": 2.9760979072990974, "grad_norm": 0.19477374851703644, "learning_rate": 3.867000583213082e-06, "loss": 0.0404, "step": 88030 }, { "epoch": 2.9764359849893505, "grad_norm": 0.23940563201904297, "learning_rate": 3.860627086405871e-06, "loss": 0.0377, "step": 88040 }, { "epoch": 2.9767740626796035, "grad_norm": 0.17332221567630768, "learning_rate": 3.854258635331126e-06, "loss": 0.038, "step": 88050 }, { "epoch": 2.977112140369857, "grad_norm": 0.20366427302360535, "learning_rate": 3.8478952306852665e-06, "loss": 0.041, "step": 88060 }, { "epoch": 2.97745021806011, "grad_norm": 0.2635685205459595, "learning_rate": 3.841536873164198e-06, "loss": 0.0411, "step": 88070 }, { "epoch": 2.977788295750363, "grad_norm": 0.34040600061416626, "learning_rate": 3.8351835634632626e-06, "loss": 0.0427, "step": 88080 }, { "epoch": 2.9781263734406167, "grad_norm": 0.15673531591892242, "learning_rate": 3.828835302277245e-06, "loss": 0.0425, "step": 88090 }, { "epoch": 2.9784644511308698, "grad_norm": 0.21174319088459015, "learning_rate": 3.822492090300378e-06, "loss": 0.0417, "step": 88100 }, { "epoch": 2.978802528821123, "grad_norm": 0.33797746896743774, "learning_rate": 3.8161539282263566e-06, "loss": 0.048, "step": 88110 }, { "epoch": 2.9791406065113764, "grad_norm": 0.321470707654953, "learning_rate": 3.8098208167482995e-06, "loss": 0.0428, "step": 88120 }, { "epoch": 2.9794786842016294, "grad_norm": 0.30186060070991516, "learning_rate": 3.803492756558796e-06, "loss": 0.0441, "step": 88130 }, { "epoch": 2.9798167618918825, "grad_norm": 0.27208298444747925, "learning_rate": 3.797169748349855e-06, "loss": 0.0466, "step": 88140 }, { "epoch": 2.980154839582136, "grad_norm": 0.18950709700584412, "learning_rate": 3.790851792812977e-06, "loss": 0.0384, "step": 88150 }, { "epoch": 2.980492917272389, "grad_norm": 0.19178332388401031, "learning_rate": 3.784538890639072e-06, "loss": 0.0422, "step": 88160 }, { "epoch": 2.980830994962642, "grad_norm": 0.2530968487262726, "learning_rate": 3.7782310425185153e-06, "loss": 0.0425, "step": 88170 }, { "epoch": 2.9811690726528957, "grad_norm": 1.0510993003845215, "learning_rate": 3.77192824914111e-06, "loss": 0.0482, "step": 88180 }, { "epoch": 2.9815071503431487, "grad_norm": 0.2358223795890808, "learning_rate": 3.7656305111961333e-06, "loss": 0.0404, "step": 88190 }, { "epoch": 2.9818452280334022, "grad_norm": 0.26163703203201294, "learning_rate": 3.7593378293722782e-06, "loss": 0.0393, "step": 88200 }, { "epoch": 2.9821833057236553, "grad_norm": 0.15953823924064636, "learning_rate": 3.7530502043577333e-06, "loss": 0.0412, "step": 88210 }, { "epoch": 2.9825213834139084, "grad_norm": 0.3539066016674042, "learning_rate": 3.746767636840076e-06, "loss": 0.0392, "step": 88220 }, { "epoch": 2.982859461104162, "grad_norm": 0.26957863569259644, "learning_rate": 3.74049012750638e-06, "loss": 0.0442, "step": 88230 }, { "epoch": 2.983197538794415, "grad_norm": 0.2877125144004822, "learning_rate": 3.734217677043128e-06, "loss": 0.042, "step": 88240 }, { "epoch": 2.983535616484668, "grad_norm": 0.30491387844085693, "learning_rate": 3.727950286136267e-06, "loss": 0.0469, "step": 88250 }, { "epoch": 2.9838736941749215, "grad_norm": 0.419202983379364, "learning_rate": 3.7216879554711925e-06, "loss": 0.04, "step": 88260 }, { "epoch": 2.9842117718651746, "grad_norm": 0.33715611696243286, "learning_rate": 3.715430685732757e-06, "loss": 0.0414, "step": 88270 }, { "epoch": 2.9845498495554277, "grad_norm": 0.49595484137535095, "learning_rate": 3.7091784776052243e-06, "loss": 0.0379, "step": 88280 }, { "epoch": 2.984887927245681, "grad_norm": 0.30889981985092163, "learning_rate": 3.7029313317723356e-06, "loss": 0.0429, "step": 88290 }, { "epoch": 2.9852260049359343, "grad_norm": 0.16981537640094757, "learning_rate": 3.6966892489172787e-06, "loss": 0.0448, "step": 88300 }, { "epoch": 2.9855640826261873, "grad_norm": 0.15438444912433624, "learning_rate": 3.6904522297226575e-06, "loss": 0.0369, "step": 88310 }, { "epoch": 2.985902160316441, "grad_norm": 0.29727286100387573, "learning_rate": 3.6842202748705647e-06, "loss": 0.0408, "step": 88320 }, { "epoch": 2.986240238006694, "grad_norm": 0.2277444750070572, "learning_rate": 3.677993385042494e-06, "loss": 0.0413, "step": 88330 }, { "epoch": 2.9865783156969474, "grad_norm": 0.2263087034225464, "learning_rate": 3.67177156091944e-06, "loss": 0.0399, "step": 88340 }, { "epoch": 2.9869163933872005, "grad_norm": 0.37335965037345886, "learning_rate": 3.6655548031817866e-06, "loss": 0.0439, "step": 88350 }, { "epoch": 2.9872544710774536, "grad_norm": 0.25082287192344666, "learning_rate": 3.6593431125094e-06, "loss": 0.0391, "step": 88360 }, { "epoch": 2.987592548767707, "grad_norm": 0.34112995862960815, "learning_rate": 3.6531364895815766e-06, "loss": 0.0442, "step": 88370 }, { "epoch": 2.98793062645796, "grad_norm": 0.25527992844581604, "learning_rate": 3.6469349350770677e-06, "loss": 0.0391, "step": 88380 }, { "epoch": 2.988268704148213, "grad_norm": 0.1556139886379242, "learning_rate": 3.6407384496740525e-06, "loss": 0.0393, "step": 88390 }, { "epoch": 2.9886067818384667, "grad_norm": 0.20863641798496246, "learning_rate": 3.6345470340501954e-06, "loss": 0.0426, "step": 88400 }, { "epoch": 2.98894485952872, "grad_norm": 0.31080541014671326, "learning_rate": 3.6283606888825537e-06, "loss": 0.0457, "step": 88410 }, { "epoch": 2.989282937218973, "grad_norm": 0.2521391808986664, "learning_rate": 3.622179414847676e-06, "loss": 0.0458, "step": 88420 }, { "epoch": 2.9896210149092264, "grad_norm": 0.2727523446083069, "learning_rate": 3.616003212621527e-06, "loss": 0.0426, "step": 88430 }, { "epoch": 2.9899590925994795, "grad_norm": 0.451140433549881, "learning_rate": 3.609832082879533e-06, "loss": 0.0409, "step": 88440 }, { "epoch": 2.9902971702897325, "grad_norm": 0.2776201069355011, "learning_rate": 3.6036660262965485e-06, "loss": 0.0468, "step": 88450 }, { "epoch": 2.990635247979986, "grad_norm": 0.3044722378253937, "learning_rate": 3.597505043546895e-06, "loss": 0.0419, "step": 88460 }, { "epoch": 2.990973325670239, "grad_norm": 0.21831972897052765, "learning_rate": 3.5913491353043227e-06, "loss": 0.0437, "step": 88470 }, { "epoch": 2.991311403360492, "grad_norm": 0.29854199290275574, "learning_rate": 3.5851983022420375e-06, "loss": 0.0391, "step": 88480 }, { "epoch": 2.9916494810507457, "grad_norm": 0.24185411632061005, "learning_rate": 3.5790525450326797e-06, "loss": 0.0423, "step": 88490 }, { "epoch": 2.9919875587409988, "grad_norm": 0.2354544848203659, "learning_rate": 3.572911864348344e-06, "loss": 0.0359, "step": 88500 }, { "epoch": 2.992325636431252, "grad_norm": 0.4113193154335022, "learning_rate": 3.5667762608605717e-06, "loss": 0.0395, "step": 88510 }, { "epoch": 2.9926637141215053, "grad_norm": 0.23928622901439667, "learning_rate": 3.5606457352403255e-06, "loss": 0.041, "step": 88520 }, { "epoch": 2.9930017918117584, "grad_norm": 0.1442195624113083, "learning_rate": 3.554520288158042e-06, "loss": 0.0426, "step": 88530 }, { "epoch": 2.9933398695020115, "grad_norm": 0.3063587546348572, "learning_rate": 3.5483999202835903e-06, "loss": 0.0384, "step": 88540 }, { "epoch": 2.993677947192265, "grad_norm": 0.3810461461544037, "learning_rate": 3.5422846322862913e-06, "loss": 0.038, "step": 88550 }, { "epoch": 2.994016024882518, "grad_norm": 0.2873041033744812, "learning_rate": 3.5361744248348925e-06, "loss": 0.0428, "step": 88560 }, { "epoch": 2.994354102572771, "grad_norm": 0.3787585198879242, "learning_rate": 3.5300692985976047e-06, "loss": 0.0397, "step": 88570 }, { "epoch": 2.9946921802630246, "grad_norm": 0.19570372998714447, "learning_rate": 3.52396925424206e-06, "loss": 0.0415, "step": 88580 }, { "epoch": 2.9950302579532777, "grad_norm": 0.19514936208724976, "learning_rate": 3.5178742924353747e-06, "loss": 0.0386, "step": 88590 }, { "epoch": 2.9953683356435308, "grad_norm": 0.31023889780044556, "learning_rate": 3.511784413844066e-06, "loss": 0.0474, "step": 88600 }, { "epoch": 2.9957064133337843, "grad_norm": 2.115987539291382, "learning_rate": 3.5056996191341287e-06, "loss": 0.0443, "step": 88610 }, { "epoch": 2.9960444910240374, "grad_norm": 0.2646861672401428, "learning_rate": 3.4996199089709692e-06, "loss": 0.0441, "step": 88620 }, { "epoch": 2.9963825687142904, "grad_norm": 0.21765856444835663, "learning_rate": 3.4935452840194725e-06, "loss": 0.0392, "step": 88630 }, { "epoch": 2.996720646404544, "grad_norm": 0.22848765552043915, "learning_rate": 3.48747574494393e-06, "loss": 0.0462, "step": 88640 }, { "epoch": 2.997058724094797, "grad_norm": 0.31198063492774963, "learning_rate": 3.4814112924081154e-06, "loss": 0.0403, "step": 88650 }, { "epoch": 2.99739680178505, "grad_norm": 0.16584721207618713, "learning_rate": 3.475351927075221e-06, "loss": 0.0383, "step": 88660 }, { "epoch": 2.9977348794753036, "grad_norm": 0.20694585144519806, "learning_rate": 3.469297649607894e-06, "loss": 0.0425, "step": 88670 }, { "epoch": 2.9980729571655567, "grad_norm": 0.171977698802948, "learning_rate": 3.463248460668217e-06, "loss": 0.0397, "step": 88680 }, { "epoch": 2.9984110348558097, "grad_norm": 0.41173362731933594, "learning_rate": 3.457204360917726e-06, "loss": 0.0423, "step": 88690 }, { "epoch": 2.9987491125460632, "grad_norm": 0.2975669205188751, "learning_rate": 3.4511653510173827e-06, "loss": 0.0416, "step": 88700 }, { "epoch": 2.9990871902363163, "grad_norm": 0.20189958810806274, "learning_rate": 3.4451314316276074e-06, "loss": 0.0425, "step": 88710 }, { "epoch": 2.9994252679265694, "grad_norm": 0.2893112897872925, "learning_rate": 3.4391026034082674e-06, "loss": 0.043, "step": 88720 }, { "epoch": 2.999763345616823, "grad_norm": 0.3038763701915741, "learning_rate": 3.433078867018674e-06, "loss": 0.0433, "step": 88730 }, { "epoch": 3.000101423307076, "grad_norm": 0.24559183418750763, "learning_rate": 3.427060223117551e-06, "loss": 0.0407, "step": 88740 }, { "epoch": 3.000439500997329, "grad_norm": 0.30083099007606506, "learning_rate": 3.4210466723631042e-06, "loss": 0.0385, "step": 88750 }, { "epoch": 3.0007775786875825, "grad_norm": 0.27635762095451355, "learning_rate": 3.415038215412969e-06, "loss": 0.0453, "step": 88760 }, { "epoch": 3.0011156563778356, "grad_norm": 0.21822789311408997, "learning_rate": 3.4090348529242088e-06, "loss": 0.0372, "step": 88770 }, { "epoch": 3.0014537340680887, "grad_norm": 0.34003862738609314, "learning_rate": 3.4030365855533487e-06, "loss": 0.0469, "step": 88780 }, { "epoch": 3.001791811758342, "grad_norm": 0.441511869430542, "learning_rate": 3.397043413956347e-06, "loss": 0.0444, "step": 88790 }, { "epoch": 3.0021298894485953, "grad_norm": 0.24413655698299408, "learning_rate": 3.3910553387886237e-06, "loss": 0.0385, "step": 88800 }, { "epoch": 3.0024679671388483, "grad_norm": 0.519739031791687, "learning_rate": 3.3850723607049996e-06, "loss": 0.0384, "step": 88810 }, { "epoch": 3.002806044829102, "grad_norm": 0.42916813492774963, "learning_rate": 3.3790944803597902e-06, "loss": 0.0439, "step": 88820 }, { "epoch": 3.003144122519355, "grad_norm": 0.2900143563747406, "learning_rate": 3.373121698406706e-06, "loss": 0.0401, "step": 88830 }, { "epoch": 3.003482200209608, "grad_norm": 0.2958313226699829, "learning_rate": 3.3671540154989302e-06, "loss": 0.0395, "step": 88840 }, { "epoch": 3.0038202778998615, "grad_norm": 0.17000022530555725, "learning_rate": 3.361191432289079e-06, "loss": 0.038, "step": 88850 }, { "epoch": 3.0041583555901146, "grad_norm": 0.26283523440361023, "learning_rate": 3.3552339494292207e-06, "loss": 0.0449, "step": 88860 }, { "epoch": 3.0044964332803676, "grad_norm": 0.2124853879213333, "learning_rate": 3.349281567570839e-06, "loss": 0.0431, "step": 88870 }, { "epoch": 3.004834510970621, "grad_norm": 0.21521420776844025, "learning_rate": 3.3433342873648965e-06, "loss": 0.042, "step": 88880 }, { "epoch": 3.0051725886608742, "grad_norm": 0.1762826144695282, "learning_rate": 3.337392109461762e-06, "loss": 0.0439, "step": 88890 }, { "epoch": 3.0055106663511273, "grad_norm": 0.21271543204784393, "learning_rate": 3.331455034511266e-06, "loss": 0.0408, "step": 88900 }, { "epoch": 3.005848744041381, "grad_norm": 0.1841895431280136, "learning_rate": 3.325523063162689e-06, "loss": 0.0399, "step": 88910 }, { "epoch": 3.006186821731634, "grad_norm": 0.389858603477478, "learning_rate": 3.3195961960647393e-06, "loss": 0.0459, "step": 88920 }, { "epoch": 3.006524899421887, "grad_norm": 0.31275710463523865, "learning_rate": 3.31367443386556e-06, "loss": 0.0477, "step": 88930 }, { "epoch": 3.0068629771121405, "grad_norm": 0.21258406341075897, "learning_rate": 3.3077577772127556e-06, "loss": 0.041, "step": 88940 }, { "epoch": 3.0072010548023935, "grad_norm": 0.4251815974712372, "learning_rate": 3.301846226753358e-06, "loss": 0.0412, "step": 88950 }, { "epoch": 3.0075391324926466, "grad_norm": 0.18675002455711365, "learning_rate": 3.2959397831338444e-06, "loss": 0.0397, "step": 88960 }, { "epoch": 3.0078772101829, "grad_norm": 0.2649921774864197, "learning_rate": 3.2900384470001432e-06, "loss": 0.0407, "step": 88970 }, { "epoch": 3.008215287873153, "grad_norm": 0.2110077142715454, "learning_rate": 3.28414221899761e-06, "loss": 0.0417, "step": 88980 }, { "epoch": 3.0085533655634067, "grad_norm": 0.22107450664043427, "learning_rate": 3.278251099771046e-06, "loss": 0.0334, "step": 88990 }, { "epoch": 3.0088914432536598, "grad_norm": 0.2506832480430603, "learning_rate": 3.2723650899646906e-06, "loss": 0.0402, "step": 89000 }, { "epoch": 3.009229520943913, "grad_norm": 0.22678673267364502, "learning_rate": 3.2664841902222466e-06, "loss": 0.0439, "step": 89010 }, { "epoch": 3.0095675986341663, "grad_norm": 0.16902925074100494, "learning_rate": 3.2606084011868208e-06, "loss": 0.0403, "step": 89020 }, { "epoch": 3.0099056763244194, "grad_norm": 0.8302085399627686, "learning_rate": 3.254737723500989e-06, "loss": 0.0444, "step": 89030 }, { "epoch": 3.0102437540146725, "grad_norm": 0.3579614758491516, "learning_rate": 3.2488721578067595e-06, "loss": 0.0446, "step": 89040 }, { "epoch": 3.010581831704926, "grad_norm": 0.17028553783893585, "learning_rate": 3.243011704745591e-06, "loss": 0.0381, "step": 89050 }, { "epoch": 3.010919909395179, "grad_norm": 0.8707995414733887, "learning_rate": 3.237156364958355e-06, "loss": 0.0469, "step": 89060 }, { "epoch": 3.011257987085432, "grad_norm": 0.2831965386867523, "learning_rate": 3.231306139085394e-06, "loss": 0.0382, "step": 89070 }, { "epoch": 3.0115960647756856, "grad_norm": 0.19559568166732788, "learning_rate": 3.2254610277664797e-06, "loss": 0.0408, "step": 89080 }, { "epoch": 3.0119341424659387, "grad_norm": 0.3453342318534851, "learning_rate": 3.2196210316408183e-06, "loss": 0.0435, "step": 89090 }, { "epoch": 3.012272220156192, "grad_norm": 0.22627508640289307, "learning_rate": 3.2137861513470644e-06, "loss": 0.0378, "step": 89100 }, { "epoch": 3.0126102978464453, "grad_norm": 0.23718732595443726, "learning_rate": 3.2079563875233252e-06, "loss": 0.0427, "step": 89110 }, { "epoch": 3.0129483755366984, "grad_norm": 0.1735113561153412, "learning_rate": 3.2021317408071184e-06, "loss": 0.0428, "step": 89120 }, { "epoch": 3.0132864532269514, "grad_norm": 0.2533327043056488, "learning_rate": 3.196312211835434e-06, "loss": 0.0429, "step": 89130 }, { "epoch": 3.013624530917205, "grad_norm": 0.23448356986045837, "learning_rate": 3.1904978012446686e-06, "loss": 0.0413, "step": 89140 }, { "epoch": 3.013962608607458, "grad_norm": 0.4229069650173187, "learning_rate": 3.184688509670691e-06, "loss": 0.0397, "step": 89150 }, { "epoch": 3.014300686297711, "grad_norm": 0.17513003945350647, "learning_rate": 3.178884337748783e-06, "loss": 0.0436, "step": 89160 }, { "epoch": 3.0146387639879646, "grad_norm": 0.3837519586086273, "learning_rate": 3.173085286113697e-06, "loss": 0.0429, "step": 89170 }, { "epoch": 3.0149768416782177, "grad_norm": 0.2228938490152359, "learning_rate": 3.1672913553995986e-06, "loss": 0.0456, "step": 89180 }, { "epoch": 3.0153149193684707, "grad_norm": 0.23465894162654877, "learning_rate": 3.1615025462401138e-06, "loss": 0.0474, "step": 89190 }, { "epoch": 3.0156529970587242, "grad_norm": 0.21014919877052307, "learning_rate": 3.1557188592682808e-06, "loss": 0.0389, "step": 89200 }, { "epoch": 3.0159910747489773, "grad_norm": 0.24932560324668884, "learning_rate": 3.1499402951166045e-06, "loss": 0.0401, "step": 89210 }, { "epoch": 3.0163291524392304, "grad_norm": 0.13481900095939636, "learning_rate": 3.144166854417019e-06, "loss": 0.0401, "step": 89220 }, { "epoch": 3.016667230129484, "grad_norm": 0.25581878423690796, "learning_rate": 3.1383985378009073e-06, "loss": 0.0422, "step": 89230 }, { "epoch": 3.017005307819737, "grad_norm": 0.2919316589832306, "learning_rate": 3.1326353458990654e-06, "loss": 0.043, "step": 89240 }, { "epoch": 3.01734338550999, "grad_norm": 0.2135496735572815, "learning_rate": 3.126877279341761e-06, "loss": 0.038, "step": 89250 }, { "epoch": 3.0176814632002436, "grad_norm": 0.24312420189380646, "learning_rate": 3.1211243387586964e-06, "loss": 0.04, "step": 89260 }, { "epoch": 3.0180195408904966, "grad_norm": 1.3426042795181274, "learning_rate": 3.115376524778979e-06, "loss": 0.0436, "step": 89270 }, { "epoch": 3.0183576185807497, "grad_norm": 0.29167473316192627, "learning_rate": 3.109633838031195e-06, "loss": 0.042, "step": 89280 }, { "epoch": 3.018695696271003, "grad_norm": 0.18846501410007477, "learning_rate": 3.103896279143359e-06, "loss": 0.0391, "step": 89290 }, { "epoch": 3.0190337739612563, "grad_norm": 0.23021696507930756, "learning_rate": 3.098163848742924e-06, "loss": 0.0492, "step": 89300 }, { "epoch": 3.0193718516515093, "grad_norm": 0.25831931829452515, "learning_rate": 3.0924365474567675e-06, "loss": 0.0412, "step": 89310 }, { "epoch": 3.019709929341763, "grad_norm": 0.3447628319263458, "learning_rate": 3.086714375911237e-06, "loss": 0.0438, "step": 89320 }, { "epoch": 3.020048007032016, "grad_norm": 0.21652336418628693, "learning_rate": 3.0809973347320775e-06, "loss": 0.0405, "step": 89330 }, { "epoch": 3.020386084722269, "grad_norm": 0.38386765122413635, "learning_rate": 3.075285424544516e-06, "loss": 0.0378, "step": 89340 }, { "epoch": 3.0207241624125225, "grad_norm": 0.3059595823287964, "learning_rate": 3.069578645973176e-06, "loss": 0.041, "step": 89350 }, { "epoch": 3.0210622401027756, "grad_norm": 0.207561194896698, "learning_rate": 3.0638769996421735e-06, "loss": 0.0411, "step": 89360 }, { "epoch": 3.0214003177930286, "grad_norm": 0.22952036559581757, "learning_rate": 3.058180486175005e-06, "loss": 0.0388, "step": 89370 }, { "epoch": 3.021738395483282, "grad_norm": 0.31737643480300903, "learning_rate": 3.052489106194645e-06, "loss": 0.0399, "step": 89380 }, { "epoch": 3.0220764731735352, "grad_norm": 0.20529325306415558, "learning_rate": 3.0468028603234946e-06, "loss": 0.0489, "step": 89390 }, { "epoch": 3.0224145508637883, "grad_norm": 0.9503926634788513, "learning_rate": 3.04112174918339e-06, "loss": 0.0411, "step": 89400 }, { "epoch": 3.022752628554042, "grad_norm": 0.3156145215034485, "learning_rate": 3.0354457733955954e-06, "loss": 0.0353, "step": 89410 }, { "epoch": 3.023090706244295, "grad_norm": 0.22277019917964935, "learning_rate": 3.029774933580859e-06, "loss": 0.0379, "step": 89420 }, { "epoch": 3.023428783934548, "grad_norm": 0.16677962243556976, "learning_rate": 3.0241092303593065e-06, "loss": 0.0373, "step": 89430 }, { "epoch": 3.0237668616248015, "grad_norm": 0.18941737711429596, "learning_rate": 3.018448664350554e-06, "loss": 0.0463, "step": 89440 }, { "epoch": 3.0241049393150545, "grad_norm": 0.2467663288116455, "learning_rate": 3.0127932361736055e-06, "loss": 0.0419, "step": 89450 }, { "epoch": 3.024443017005308, "grad_norm": 0.20790229737758636, "learning_rate": 3.00714294644695e-06, "loss": 0.0387, "step": 89460 }, { "epoch": 3.024781094695561, "grad_norm": 0.8275421261787415, "learning_rate": 3.0014977957884828e-06, "loss": 0.0407, "step": 89470 }, { "epoch": 3.025119172385814, "grad_norm": 0.2507433593273163, "learning_rate": 2.99585778481557e-06, "loss": 0.038, "step": 89480 }, { "epoch": 3.0254572500760677, "grad_norm": 0.25020408630371094, "learning_rate": 2.990222914144969e-06, "loss": 0.0474, "step": 89490 }, { "epoch": 3.0257953277663208, "grad_norm": 0.31329989433288574, "learning_rate": 2.984593184392914e-06, "loss": 0.0396, "step": 89500 }, { "epoch": 3.026133405456574, "grad_norm": 0.2556997537612915, "learning_rate": 2.978968596175069e-06, "loss": 0.0404, "step": 89510 }, { "epoch": 3.0264714831468273, "grad_norm": 0.23866552114486694, "learning_rate": 2.9733491501065137e-06, "loss": 0.0382, "step": 89520 }, { "epoch": 3.0268095608370804, "grad_norm": 0.2819077968597412, "learning_rate": 2.9677348468018005e-06, "loss": 0.038, "step": 89530 }, { "epoch": 3.0271476385273335, "grad_norm": 0.30814024806022644, "learning_rate": 2.9621256868748726e-06, "loss": 0.0425, "step": 89540 }, { "epoch": 3.027485716217587, "grad_norm": 0.2398994266986847, "learning_rate": 2.956521670939183e-06, "loss": 0.0418, "step": 89550 }, { "epoch": 3.02782379390784, "grad_norm": 0.47055044770240784, "learning_rate": 2.950922799607542e-06, "loss": 0.0373, "step": 89560 }, { "epoch": 3.028161871598093, "grad_norm": 0.31759482622146606, "learning_rate": 2.9453290734922535e-06, "loss": 0.0426, "step": 89570 }, { "epoch": 3.0284999492883466, "grad_norm": 0.15069042146205902, "learning_rate": 2.939740493205029e-06, "loss": 0.0375, "step": 89580 }, { "epoch": 3.0288380269785997, "grad_norm": 0.19887542724609375, "learning_rate": 2.9341570593570345e-06, "loss": 0.0464, "step": 89590 }, { "epoch": 3.029176104668853, "grad_norm": 0.2942316234111786, "learning_rate": 2.928578772558843e-06, "loss": 0.0412, "step": 89600 }, { "epoch": 3.0295141823591063, "grad_norm": 0.2181767225265503, "learning_rate": 2.923005633420528e-06, "loss": 0.0367, "step": 89610 }, { "epoch": 3.0298522600493594, "grad_norm": 0.324688583612442, "learning_rate": 2.9174376425515237e-06, "loss": 0.0405, "step": 89620 }, { "epoch": 3.0301903377396124, "grad_norm": 0.3436424136161804, "learning_rate": 2.91187480056076e-06, "loss": 0.0476, "step": 89630 }, { "epoch": 3.030528415429866, "grad_norm": 0.37110191583633423, "learning_rate": 2.9063171080565664e-06, "loss": 0.0424, "step": 89640 }, { "epoch": 3.030866493120119, "grad_norm": 0.14905959367752075, "learning_rate": 2.900764565646741e-06, "loss": 0.0412, "step": 89650 }, { "epoch": 3.031204570810372, "grad_norm": 0.19922596216201782, "learning_rate": 2.895217173938475e-06, "loss": 0.0464, "step": 89660 }, { "epoch": 3.0315426485006256, "grad_norm": 0.27209630608558655, "learning_rate": 2.8896749335384445e-06, "loss": 0.0415, "step": 89670 }, { "epoch": 3.0318807261908787, "grad_norm": 0.6716121435165405, "learning_rate": 2.884137845052731e-06, "loss": 0.0386, "step": 89680 }, { "epoch": 3.0322188038811317, "grad_norm": 0.5491514801979065, "learning_rate": 2.8786059090868734e-06, "loss": 0.0451, "step": 89690 }, { "epoch": 3.0325568815713853, "grad_norm": 0.2588368356227875, "learning_rate": 2.87307912624582e-06, "loss": 0.0443, "step": 89700 }, { "epoch": 3.0328949592616383, "grad_norm": 0.769250214099884, "learning_rate": 2.867557497133977e-06, "loss": 0.0385, "step": 89710 }, { "epoch": 3.0332330369518914, "grad_norm": 0.2973039448261261, "learning_rate": 2.862041022355194e-06, "loss": 0.0422, "step": 89720 }, { "epoch": 3.033571114642145, "grad_norm": 0.2800840735435486, "learning_rate": 2.8565297025127226e-06, "loss": 0.0435, "step": 89730 }, { "epoch": 3.033909192332398, "grad_norm": 0.2728886306285858, "learning_rate": 2.8510235382092853e-06, "loss": 0.0393, "step": 89740 }, { "epoch": 3.034247270022651, "grad_norm": 0.22000792622566223, "learning_rate": 2.845522530047029e-06, "loss": 0.0397, "step": 89750 }, { "epoch": 3.0345853477129046, "grad_norm": 0.1825149655342102, "learning_rate": 2.8400266786275387e-06, "loss": 0.0405, "step": 89760 }, { "epoch": 3.0349234254031576, "grad_norm": 0.1795767992734909, "learning_rate": 2.834535984551817e-06, "loss": 0.042, "step": 89770 }, { "epoch": 3.0352615030934107, "grad_norm": 0.2840324342250824, "learning_rate": 2.8290504484203397e-06, "loss": 0.0425, "step": 89780 }, { "epoch": 3.035599580783664, "grad_norm": 0.23641248047351837, "learning_rate": 2.8235700708329705e-06, "loss": 0.0442, "step": 89790 }, { "epoch": 3.0359376584739173, "grad_norm": 0.2869318425655365, "learning_rate": 2.818094852389058e-06, "loss": 0.0411, "step": 89800 }, { "epoch": 3.0362757361641703, "grad_norm": 0.25854232907295227, "learning_rate": 2.8126247936873516e-06, "loss": 0.0449, "step": 89810 }, { "epoch": 3.036613813854424, "grad_norm": 0.3320878744125366, "learning_rate": 2.8071598953260614e-06, "loss": 0.0466, "step": 89820 }, { "epoch": 3.036951891544677, "grad_norm": 0.21456660330295563, "learning_rate": 2.8017001579028035e-06, "loss": 0.0392, "step": 89830 }, { "epoch": 3.03728996923493, "grad_norm": 0.24754616618156433, "learning_rate": 2.7962455820146617e-06, "loss": 0.0446, "step": 89840 }, { "epoch": 3.0376280469251835, "grad_norm": 0.2699606716632843, "learning_rate": 2.7907961682581253e-06, "loss": 0.0432, "step": 89850 }, { "epoch": 3.0379661246154366, "grad_norm": 0.2728879451751709, "learning_rate": 2.7853519172291453e-06, "loss": 0.0423, "step": 89860 }, { "epoch": 3.0383042023056896, "grad_norm": 0.3730008602142334, "learning_rate": 2.7799128295230904e-06, "loss": 0.0411, "step": 89870 }, { "epoch": 3.038642279995943, "grad_norm": 0.5052222609519958, "learning_rate": 2.774478905734784e-06, "loss": 0.045, "step": 89880 }, { "epoch": 3.0389803576861962, "grad_norm": 0.23930539190769196, "learning_rate": 2.7690501464584516e-06, "loss": 0.0405, "step": 89890 }, { "epoch": 3.0393184353764493, "grad_norm": 0.13173575699329376, "learning_rate": 2.763626552287796e-06, "loss": 0.0418, "step": 89900 }, { "epoch": 3.039656513066703, "grad_norm": 0.3182491362094879, "learning_rate": 2.7582081238159083e-06, "loss": 0.04, "step": 89910 }, { "epoch": 3.039994590756956, "grad_norm": 0.1846388578414917, "learning_rate": 2.7527948616353548e-06, "loss": 0.0433, "step": 89920 }, { "epoch": 3.040332668447209, "grad_norm": 0.17024917900562286, "learning_rate": 2.7473867663381226e-06, "loss": 0.0422, "step": 89930 }, { "epoch": 3.0406707461374625, "grad_norm": 0.25715482234954834, "learning_rate": 2.7419838385156383e-06, "loss": 0.0366, "step": 89940 }, { "epoch": 3.0410088238277155, "grad_norm": 0.21710814535617828, "learning_rate": 2.7365860787587407e-06, "loss": 0.0392, "step": 89950 }, { "epoch": 3.0413469015179686, "grad_norm": 0.18540385365486145, "learning_rate": 2.7311934876577295e-06, "loss": 0.0421, "step": 89960 }, { "epoch": 3.041684979208222, "grad_norm": 0.23611173033714294, "learning_rate": 2.725806065802339e-06, "loss": 0.0431, "step": 89970 }, { "epoch": 3.042023056898475, "grad_norm": 0.277955025434494, "learning_rate": 2.720423813781714e-06, "loss": 0.0404, "step": 89980 }, { "epoch": 3.0423611345887287, "grad_norm": 0.15694352984428406, "learning_rate": 2.715046732184462e-06, "loss": 0.0427, "step": 89990 }, { "epoch": 3.0426992122789818, "grad_norm": 0.2175624817609787, "learning_rate": 2.7096748215986013e-06, "loss": 0.0424, "step": 90000 }, { "epoch": 3.043037289969235, "grad_norm": 0.40859898924827576, "learning_rate": 2.704308082611612e-06, "loss": 0.038, "step": 90010 }, { "epoch": 3.0433753676594884, "grad_norm": 0.17581824958324432, "learning_rate": 2.698946515810374e-06, "loss": 0.0438, "step": 90020 }, { "epoch": 3.0437134453497414, "grad_norm": 0.17850090563297272, "learning_rate": 2.6935901217812367e-06, "loss": 0.0407, "step": 90030 }, { "epoch": 3.0440515230399945, "grad_norm": 0.22872208058834076, "learning_rate": 2.688238901109952e-06, "loss": 0.0465, "step": 90040 }, { "epoch": 3.044389600730248, "grad_norm": 0.17643606662750244, "learning_rate": 2.68289285438173e-06, "loss": 0.0416, "step": 90050 }, { "epoch": 3.044727678420501, "grad_norm": 0.1998656541109085, "learning_rate": 2.6775519821812033e-06, "loss": 0.0409, "step": 90060 }, { "epoch": 3.045065756110754, "grad_norm": 0.3196231722831726, "learning_rate": 2.672216285092449e-06, "loss": 0.0412, "step": 90070 }, { "epoch": 3.0454038338010077, "grad_norm": 0.28514477610588074, "learning_rate": 2.666885763698962e-06, "loss": 0.0391, "step": 90080 }, { "epoch": 3.0457419114912607, "grad_norm": 0.30062147974967957, "learning_rate": 2.661560418583692e-06, "loss": 0.0373, "step": 90090 }, { "epoch": 3.046079989181514, "grad_norm": 0.20777587592601776, "learning_rate": 2.656240250328995e-06, "loss": 0.043, "step": 90100 }, { "epoch": 3.0464180668717673, "grad_norm": 0.31752994656562805, "learning_rate": 2.650925259516679e-06, "loss": 0.0383, "step": 90110 }, { "epoch": 3.0467561445620204, "grad_norm": 0.2930091619491577, "learning_rate": 2.645615446727995e-06, "loss": 0.041, "step": 90120 }, { "epoch": 3.0470942222522734, "grad_norm": 0.19902917742729187, "learning_rate": 2.6403108125436115e-06, "loss": 0.0396, "step": 90130 }, { "epoch": 3.047432299942527, "grad_norm": 0.32581761479377747, "learning_rate": 2.6350113575436264e-06, "loss": 0.0412, "step": 90140 }, { "epoch": 3.04777037763278, "grad_norm": 0.24839366972446442, "learning_rate": 2.629717082307598e-06, "loss": 0.0434, "step": 90150 }, { "epoch": 3.048108455323033, "grad_norm": 0.2548542022705078, "learning_rate": 2.6244279874144793e-06, "loss": 0.048, "step": 90160 }, { "epoch": 3.0484465330132866, "grad_norm": 0.25299012660980225, "learning_rate": 2.6191440734426862e-06, "loss": 0.0463, "step": 90170 }, { "epoch": 3.0487846107035397, "grad_norm": 0.3559145927429199, "learning_rate": 2.613865340970062e-06, "loss": 0.0421, "step": 90180 }, { "epoch": 3.0491226883937927, "grad_norm": 0.22488634288311005, "learning_rate": 2.608591790573889e-06, "loss": 0.0515, "step": 90190 }, { "epoch": 3.0494607660840463, "grad_norm": 0.19074243307113647, "learning_rate": 2.603323422830861e-06, "loss": 0.0402, "step": 90200 }, { "epoch": 3.0497988437742993, "grad_norm": 0.3017450273036957, "learning_rate": 2.5980602383171183e-06, "loss": 0.0388, "step": 90210 }, { "epoch": 3.0501369214645524, "grad_norm": 0.20424005389213562, "learning_rate": 2.5928022376082496e-06, "loss": 0.042, "step": 90220 }, { "epoch": 3.050474999154806, "grad_norm": 0.18105238676071167, "learning_rate": 2.5875494212792505e-06, "loss": 0.0429, "step": 90230 }, { "epoch": 3.050813076845059, "grad_norm": 0.22117164731025696, "learning_rate": 2.5823017899045565e-06, "loss": 0.0479, "step": 90240 }, { "epoch": 3.051151154535312, "grad_norm": 0.23960193991661072, "learning_rate": 2.577059344058047e-06, "loss": 0.0449, "step": 90250 }, { "epoch": 3.0514892322255656, "grad_norm": 0.18274468183517456, "learning_rate": 2.5718220843130415e-06, "loss": 0.042, "step": 90260 }, { "epoch": 3.0518273099158186, "grad_norm": 0.18494918942451477, "learning_rate": 2.5665900112422535e-06, "loss": 0.0398, "step": 90270 }, { "epoch": 3.0521653876060717, "grad_norm": 0.19514980912208557, "learning_rate": 2.561363125417876e-06, "loss": 0.0425, "step": 90280 }, { "epoch": 3.052503465296325, "grad_norm": 0.31790125370025635, "learning_rate": 2.556141427411496e-06, "loss": 0.0392, "step": 90290 }, { "epoch": 3.0528415429865783, "grad_norm": 0.26967763900756836, "learning_rate": 2.550924917794156e-06, "loss": 0.0405, "step": 90300 }, { "epoch": 3.0531796206768314, "grad_norm": 0.9476869106292725, "learning_rate": 2.545713597136329e-06, "loss": 0.0431, "step": 90310 }, { "epoch": 3.053517698367085, "grad_norm": 0.23791804909706116, "learning_rate": 2.5405074660079298e-06, "loss": 0.0412, "step": 90320 }, { "epoch": 3.053855776057338, "grad_norm": 0.2251162976026535, "learning_rate": 2.5353065249782647e-06, "loss": 0.0379, "step": 90330 }, { "epoch": 3.054193853747591, "grad_norm": 0.33219635486602783, "learning_rate": 2.530110774616129e-06, "loss": 0.0417, "step": 90340 }, { "epoch": 3.0545319314378445, "grad_norm": 0.23243461549282074, "learning_rate": 2.524920215489701e-06, "loss": 0.042, "step": 90350 }, { "epoch": 3.0548700091280976, "grad_norm": 0.1858716905117035, "learning_rate": 2.519734848166627e-06, "loss": 0.0459, "step": 90360 }, { "epoch": 3.0552080868183507, "grad_norm": 0.6286957859992981, "learning_rate": 2.514554673213948e-06, "loss": 0.0415, "step": 90370 }, { "epoch": 3.055546164508604, "grad_norm": 0.15949225425720215, "learning_rate": 2.5093796911981936e-06, "loss": 0.042, "step": 90380 }, { "epoch": 3.0558842421988572, "grad_norm": 0.4733610153198242, "learning_rate": 2.5042099026852672e-06, "loss": 0.0447, "step": 90390 }, { "epoch": 3.0562223198891103, "grad_norm": 0.17223531007766724, "learning_rate": 2.499045308240544e-06, "loss": 0.0411, "step": 90400 }, { "epoch": 3.056560397579364, "grad_norm": 0.18808604776859283, "learning_rate": 2.4938859084288003e-06, "loss": 0.0463, "step": 90410 }, { "epoch": 3.056898475269617, "grad_norm": 0.30520129203796387, "learning_rate": 2.4887317038142734e-06, "loss": 0.0394, "step": 90420 }, { "epoch": 3.0572365529598704, "grad_norm": 0.1976616233587265, "learning_rate": 2.4835826949606124e-06, "loss": 0.0438, "step": 90430 }, { "epoch": 3.0575746306501235, "grad_norm": 0.2658534049987793, "learning_rate": 2.4784388824309167e-06, "loss": 0.0462, "step": 90440 }, { "epoch": 3.0579127083403765, "grad_norm": 0.22565755248069763, "learning_rate": 2.473300266787687e-06, "loss": 0.0451, "step": 90450 }, { "epoch": 3.05825078603063, "grad_norm": 0.2368251085281372, "learning_rate": 2.46816684859289e-06, "loss": 0.043, "step": 90460 }, { "epoch": 3.058588863720883, "grad_norm": 0.41737475991249084, "learning_rate": 2.4630386284079043e-06, "loss": 0.0437, "step": 90470 }, { "epoch": 3.058926941411136, "grad_norm": 0.28068673610687256, "learning_rate": 2.4579156067935427e-06, "loss": 0.0459, "step": 90480 }, { "epoch": 3.0592650191013897, "grad_norm": 0.20318205654621124, "learning_rate": 2.452797784310046e-06, "loss": 0.0427, "step": 90490 }, { "epoch": 3.0596030967916428, "grad_norm": 0.2895333468914032, "learning_rate": 2.447685161517105e-06, "loss": 0.0431, "step": 90500 }, { "epoch": 3.059941174481896, "grad_norm": 0.25535041093826294, "learning_rate": 2.4425777389738224e-06, "loss": 0.0408, "step": 90510 }, { "epoch": 3.0602792521721494, "grad_norm": 0.4957916736602783, "learning_rate": 2.43747551723873e-06, "loss": 0.042, "step": 90520 }, { "epoch": 3.0606173298624024, "grad_norm": 0.7075740098953247, "learning_rate": 2.4323784968698136e-06, "loss": 0.0439, "step": 90530 }, { "epoch": 3.0609554075526555, "grad_norm": 0.19747552275657654, "learning_rate": 2.4272866784244565e-06, "loss": 0.0405, "step": 90540 }, { "epoch": 3.061293485242909, "grad_norm": 0.30872946977615356, "learning_rate": 2.422200062459518e-06, "loss": 0.0458, "step": 90550 }, { "epoch": 3.061631562933162, "grad_norm": 0.22901849448680878, "learning_rate": 2.41711864953123e-06, "loss": 0.0418, "step": 90560 }, { "epoch": 3.061969640623415, "grad_norm": 0.21380257606506348, "learning_rate": 2.412042440195322e-06, "loss": 0.0364, "step": 90570 }, { "epoch": 3.0623077183136687, "grad_norm": 0.326642245054245, "learning_rate": 2.4069714350068993e-06, "loss": 0.0446, "step": 90580 }, { "epoch": 3.0626457960039217, "grad_norm": 0.2988021671772003, "learning_rate": 2.4019056345205294e-06, "loss": 0.0415, "step": 90590 }, { "epoch": 3.062983873694175, "grad_norm": 0.30677923560142517, "learning_rate": 2.396845039290191e-06, "loss": 0.039, "step": 90600 }, { "epoch": 3.0633219513844283, "grad_norm": 0.22353553771972656, "learning_rate": 2.3917896498693194e-06, "loss": 0.0362, "step": 90610 }, { "epoch": 3.0636600290746814, "grad_norm": 0.25961610674858093, "learning_rate": 2.3867394668107335e-06, "loss": 0.0425, "step": 90620 }, { "epoch": 3.0639981067649344, "grad_norm": 0.17829902470111847, "learning_rate": 2.3816944906667527e-06, "loss": 0.0356, "step": 90630 }, { "epoch": 3.064336184455188, "grad_norm": 0.2167445868253708, "learning_rate": 2.3766547219890634e-06, "loss": 0.043, "step": 90640 }, { "epoch": 3.064674262145441, "grad_norm": 0.24828623235225677, "learning_rate": 2.371620161328819e-06, "loss": 0.0401, "step": 90650 }, { "epoch": 3.065012339835694, "grad_norm": 0.2684709131717682, "learning_rate": 2.366590809236574e-06, "loss": 0.0407, "step": 90660 }, { "epoch": 3.0653504175259476, "grad_norm": 0.18063807487487793, "learning_rate": 2.3615666662623494e-06, "loss": 0.0433, "step": 90670 }, { "epoch": 3.0656884952162007, "grad_norm": 0.27388182282447815, "learning_rate": 2.3565477329555664e-06, "loss": 0.042, "step": 90680 }, { "epoch": 3.0660265729064538, "grad_norm": 0.17410309612751007, "learning_rate": 2.3515340098650975e-06, "loss": 0.0436, "step": 90690 }, { "epoch": 3.0663646505967073, "grad_norm": 0.20318377017974854, "learning_rate": 2.3465254975392258e-06, "loss": 0.0427, "step": 90700 }, { "epoch": 3.0667027282869603, "grad_norm": 0.38050907850265503, "learning_rate": 2.3415221965256805e-06, "loss": 0.0433, "step": 90710 }, { "epoch": 3.0670408059772134, "grad_norm": 0.17290852963924408, "learning_rate": 2.3365241073716237e-06, "loss": 0.0391, "step": 90720 }, { "epoch": 3.067378883667467, "grad_norm": 0.3553224802017212, "learning_rate": 2.3315312306236183e-06, "loss": 0.0422, "step": 90730 }, { "epoch": 3.06771696135772, "grad_norm": 0.26671135425567627, "learning_rate": 2.326543566827699e-06, "loss": 0.0397, "step": 90740 }, { "epoch": 3.068055039047973, "grad_norm": 0.12282824516296387, "learning_rate": 2.3215611165292862e-06, "loss": 0.0407, "step": 90750 }, { "epoch": 3.0683931167382266, "grad_norm": 0.18922336399555206, "learning_rate": 2.3165838802732765e-06, "loss": 0.041, "step": 90760 }, { "epoch": 3.0687311944284796, "grad_norm": 0.22758498787879944, "learning_rate": 2.311611858603957e-06, "loss": 0.0393, "step": 90770 }, { "epoch": 3.0690692721187327, "grad_norm": 0.18405921757221222, "learning_rate": 2.3066450520650763e-06, "loss": 0.0434, "step": 90780 }, { "epoch": 3.069407349808986, "grad_norm": 0.20986144244670868, "learning_rate": 2.3016834611997773e-06, "loss": 0.0474, "step": 90790 }, { "epoch": 3.0697454274992393, "grad_norm": 0.1313217431306839, "learning_rate": 2.296727086550665e-06, "loss": 0.0445, "step": 90800 }, { "epoch": 3.0700835051894924, "grad_norm": 0.19658130407333374, "learning_rate": 2.2917759286597507e-06, "loss": 0.0447, "step": 90810 }, { "epoch": 3.070421582879746, "grad_norm": 0.17689301073551178, "learning_rate": 2.2868299880685006e-06, "loss": 0.042, "step": 90820 }, { "epoch": 3.070759660569999, "grad_norm": 0.16920673847198486, "learning_rate": 2.2818892653177827e-06, "loss": 0.0354, "step": 90830 }, { "epoch": 3.071097738260252, "grad_norm": 0.23959459364414215, "learning_rate": 2.2769537609479196e-06, "loss": 0.0438, "step": 90840 }, { "epoch": 3.0714358159505055, "grad_norm": 0.1621294617652893, "learning_rate": 2.2720234754986357e-06, "loss": 0.0415, "step": 90850 }, { "epoch": 3.0717738936407586, "grad_norm": 0.1509730964899063, "learning_rate": 2.2670984095091108e-06, "loss": 0.0425, "step": 90860 }, { "epoch": 3.0721119713310117, "grad_norm": 0.29765358567237854, "learning_rate": 2.262178563517936e-06, "loss": 0.0439, "step": 90870 }, { "epoch": 3.072450049021265, "grad_norm": 0.1686282902956009, "learning_rate": 2.257263938063137e-06, "loss": 0.0412, "step": 90880 }, { "epoch": 3.0727881267115182, "grad_norm": 0.18199115991592407, "learning_rate": 2.2523545336821715e-06, "loss": 0.0394, "step": 90890 }, { "epoch": 3.0731262044017713, "grad_norm": 0.25308898091316223, "learning_rate": 2.247450350911939e-06, "loss": 0.0366, "step": 90900 }, { "epoch": 3.073464282092025, "grad_norm": 0.3325130045413971, "learning_rate": 2.2425513902887328e-06, "loss": 0.0414, "step": 90910 }, { "epoch": 3.073802359782278, "grad_norm": 0.34996289014816284, "learning_rate": 2.2376576523483015e-06, "loss": 0.0381, "step": 90920 }, { "epoch": 3.074140437472531, "grad_norm": 0.2734073996543884, "learning_rate": 2.232769137625829e-06, "loss": 0.0425, "step": 90930 }, { "epoch": 3.0744785151627845, "grad_norm": 0.15422028303146362, "learning_rate": 2.227885846655903e-06, "loss": 0.0386, "step": 90940 }, { "epoch": 3.0748165928530375, "grad_norm": 0.30045753717422485, "learning_rate": 2.2230077799725523e-06, "loss": 0.045, "step": 90950 }, { "epoch": 3.075154670543291, "grad_norm": 0.2518872618675232, "learning_rate": 2.2181349381092396e-06, "loss": 0.0413, "step": 90960 }, { "epoch": 3.075492748233544, "grad_norm": 0.17449551820755005, "learning_rate": 2.2132673215988552e-06, "loss": 0.0393, "step": 90970 }, { "epoch": 3.075830825923797, "grad_norm": 0.16491679847240448, "learning_rate": 2.2084049309737065e-06, "loss": 0.0451, "step": 90980 }, { "epoch": 3.0761689036140507, "grad_norm": 0.3310369849205017, "learning_rate": 2.2035477667655513e-06, "loss": 0.0389, "step": 90990 }, { "epoch": 3.076506981304304, "grad_norm": 0.2857823073863983, "learning_rate": 2.1986958295055315e-06, "loss": 0.0412, "step": 91000 }, { "epoch": 3.076845058994557, "grad_norm": 0.3350639343261719, "learning_rate": 2.193849119724284e-06, "loss": 0.0462, "step": 91010 }, { "epoch": 3.0771831366848104, "grad_norm": 0.18420666456222534, "learning_rate": 2.1890076379518175e-06, "loss": 0.039, "step": 91020 }, { "epoch": 3.0775212143750634, "grad_norm": 0.3432978093624115, "learning_rate": 2.1841713847175925e-06, "loss": 0.0475, "step": 91030 }, { "epoch": 3.0778592920653165, "grad_norm": 0.2718091905117035, "learning_rate": 2.1793403605504968e-06, "loss": 0.0406, "step": 91040 }, { "epoch": 3.07819736975557, "grad_norm": 0.22094859182834625, "learning_rate": 2.1745145659788413e-06, "loss": 0.0403, "step": 91050 }, { "epoch": 3.078535447445823, "grad_norm": 0.24457311630249023, "learning_rate": 2.1696940015303645e-06, "loss": 0.0439, "step": 91060 }, { "epoch": 3.078873525136076, "grad_norm": 0.717007577419281, "learning_rate": 2.1648786677322397e-06, "loss": 0.0389, "step": 91070 }, { "epoch": 3.0792116028263297, "grad_norm": 0.37033987045288086, "learning_rate": 2.1600685651110673e-06, "loss": 0.0464, "step": 91080 }, { "epoch": 3.0795496805165827, "grad_norm": 0.24490877985954285, "learning_rate": 2.1552636941928717e-06, "loss": 0.0436, "step": 91090 }, { "epoch": 3.079887758206836, "grad_norm": 0.21474558115005493, "learning_rate": 2.1504640555031043e-06, "loss": 0.0415, "step": 91100 }, { "epoch": 3.0802258358970893, "grad_norm": 0.24362532794475555, "learning_rate": 2.1456696495666506e-06, "loss": 0.0342, "step": 91110 }, { "epoch": 3.0805639135873424, "grad_norm": 0.30915576219558716, "learning_rate": 2.1408804769078084e-06, "loss": 0.037, "step": 91120 }, { "epoch": 3.0809019912775955, "grad_norm": 0.20985805988311768, "learning_rate": 2.136096538050325e-06, "loss": 0.0386, "step": 91130 }, { "epoch": 3.081240068967849, "grad_norm": 0.20226813852787018, "learning_rate": 2.13131783351736e-06, "loss": 0.04, "step": 91140 }, { "epoch": 3.081578146658102, "grad_norm": 0.21705757081508636, "learning_rate": 2.126544363831512e-06, "loss": 0.0388, "step": 91150 }, { "epoch": 3.081916224348355, "grad_norm": 0.15926122665405273, "learning_rate": 2.121776129514791e-06, "loss": 0.0388, "step": 91160 }, { "epoch": 3.0822543020386086, "grad_norm": 0.2877260446548462, "learning_rate": 2.1170131310886523e-06, "loss": 0.038, "step": 91170 }, { "epoch": 3.0825923797288617, "grad_norm": 0.22633953392505646, "learning_rate": 2.112255369073968e-06, "loss": 0.0438, "step": 91180 }, { "epoch": 3.0829304574191148, "grad_norm": 0.19251972436904907, "learning_rate": 2.1075028439910328e-06, "loss": 0.0428, "step": 91190 }, { "epoch": 3.0832685351093683, "grad_norm": 0.20501486957073212, "learning_rate": 2.1027555563595813e-06, "loss": 0.0397, "step": 91200 }, { "epoch": 3.0836066127996213, "grad_norm": 0.16308404505252838, "learning_rate": 2.0980135066987706e-06, "loss": 0.0465, "step": 91210 }, { "epoch": 3.0839446904898744, "grad_norm": 0.23630917072296143, "learning_rate": 2.093276695527191e-06, "loss": 0.041, "step": 91220 }, { "epoch": 3.084282768180128, "grad_norm": 0.14055360853672028, "learning_rate": 2.0885451233628396e-06, "loss": 0.0461, "step": 91230 }, { "epoch": 3.084620845870381, "grad_norm": 0.2534147799015045, "learning_rate": 2.0838187907231633e-06, "loss": 0.0459, "step": 91240 }, { "epoch": 3.084958923560634, "grad_norm": 0.1800488531589508, "learning_rate": 2.0790976981250153e-06, "loss": 0.043, "step": 91250 }, { "epoch": 3.0852970012508876, "grad_norm": 0.26828253269195557, "learning_rate": 2.0743818460847e-06, "loss": 0.0424, "step": 91260 }, { "epoch": 3.0856350789411406, "grad_norm": 0.325531005859375, "learning_rate": 2.069671235117926e-06, "loss": 0.0422, "step": 91270 }, { "epoch": 3.0859731566313937, "grad_norm": 0.26203781366348267, "learning_rate": 2.0649658657398484e-06, "loss": 0.0376, "step": 91280 }, { "epoch": 3.0863112343216472, "grad_norm": 0.24000152945518494, "learning_rate": 2.060265738465034e-06, "loss": 0.0447, "step": 91290 }, { "epoch": 3.0866493120119003, "grad_norm": 0.36075156927108765, "learning_rate": 2.055570853807487e-06, "loss": 0.0424, "step": 91300 }, { "epoch": 3.0869873897021534, "grad_norm": 0.3435310423374176, "learning_rate": 2.0508812122806196e-06, "loss": 0.0379, "step": 91310 }, { "epoch": 3.087325467392407, "grad_norm": 0.21103984117507935, "learning_rate": 2.046196814397294e-06, "loss": 0.0416, "step": 91320 }, { "epoch": 3.08766354508266, "grad_norm": 0.2378229945898056, "learning_rate": 2.0415176606697896e-06, "loss": 0.0443, "step": 91330 }, { "epoch": 3.088001622772913, "grad_norm": 0.17734654247760773, "learning_rate": 2.0368437516098126e-06, "loss": 0.044, "step": 91340 }, { "epoch": 3.0883397004631665, "grad_norm": 0.19070498645305634, "learning_rate": 2.032175087728483e-06, "loss": 0.0428, "step": 91350 }, { "epoch": 3.0886777781534196, "grad_norm": 0.1615464985370636, "learning_rate": 2.027511669536375e-06, "loss": 0.0448, "step": 91360 }, { "epoch": 3.0890158558436727, "grad_norm": 0.18224403262138367, "learning_rate": 2.0228534975434643e-06, "loss": 0.0353, "step": 91370 }, { "epoch": 3.089353933533926, "grad_norm": 0.49063828587532043, "learning_rate": 2.0182005722591548e-06, "loss": 0.0441, "step": 91380 }, { "epoch": 3.0896920112241792, "grad_norm": 0.2641732394695282, "learning_rate": 2.0135528941922944e-06, "loss": 0.0455, "step": 91390 }, { "epoch": 3.0900300889144328, "grad_norm": 0.3788398206233978, "learning_rate": 2.008910463851149e-06, "loss": 0.0436, "step": 91400 }, { "epoch": 3.090368166604686, "grad_norm": 0.3980165123939514, "learning_rate": 2.0042732817433963e-06, "loss": 0.042, "step": 91410 }, { "epoch": 3.090706244294939, "grad_norm": 0.19049349427223206, "learning_rate": 1.999641348376158e-06, "loss": 0.0417, "step": 91420 }, { "epoch": 3.0910443219851924, "grad_norm": 0.26462918519973755, "learning_rate": 1.9950146642559787e-06, "loss": 0.0421, "step": 91430 }, { "epoch": 3.0913823996754455, "grad_norm": 0.1907852739095688, "learning_rate": 1.990393229888815e-06, "loss": 0.0413, "step": 91440 }, { "epoch": 3.0917204773656985, "grad_norm": 0.24064567685127258, "learning_rate": 1.9857770457800684e-06, "loss": 0.0435, "step": 91450 }, { "epoch": 3.092058555055952, "grad_norm": 0.1946636140346527, "learning_rate": 1.9811661124345517e-06, "loss": 0.0391, "step": 91460 }, { "epoch": 3.092396632746205, "grad_norm": 0.2550148367881775, "learning_rate": 1.976560430356522e-06, "loss": 0.039, "step": 91470 }, { "epoch": 3.092734710436458, "grad_norm": 0.2652076482772827, "learning_rate": 1.9719600000496383e-06, "loss": 0.0399, "step": 91480 }, { "epoch": 3.0930727881267117, "grad_norm": 0.24604153633117676, "learning_rate": 1.967364822017004e-06, "loss": 0.0393, "step": 91490 }, { "epoch": 3.093410865816965, "grad_norm": 0.27133628726005554, "learning_rate": 1.962774896761127e-06, "loss": 0.0437, "step": 91500 }, { "epoch": 3.093748943507218, "grad_norm": 0.4193904399871826, "learning_rate": 1.958190224783968e-06, "loss": 0.046, "step": 91510 }, { "epoch": 3.0940870211974714, "grad_norm": 0.3201743960380554, "learning_rate": 1.9536108065868975e-06, "loss": 0.0435, "step": 91520 }, { "epoch": 3.0944250988877244, "grad_norm": 0.3453175723552704, "learning_rate": 1.949036642670715e-06, "loss": 0.0452, "step": 91530 }, { "epoch": 3.0947631765779775, "grad_norm": 0.24355663359165192, "learning_rate": 1.9444677335356377e-06, "loss": 0.0432, "step": 91540 }, { "epoch": 3.095101254268231, "grad_norm": 0.22877587378025055, "learning_rate": 1.939904079681326e-06, "loss": 0.0416, "step": 91550 }, { "epoch": 3.095439331958484, "grad_norm": 0.30271539092063904, "learning_rate": 1.935345681606837e-06, "loss": 0.0384, "step": 91560 }, { "epoch": 3.095777409648737, "grad_norm": 0.25255444645881653, "learning_rate": 1.930792539810694e-06, "loss": 0.0435, "step": 91570 }, { "epoch": 3.0961154873389907, "grad_norm": 0.3143899440765381, "learning_rate": 1.9262446547907877e-06, "loss": 0.0387, "step": 91580 }, { "epoch": 3.0964535650292437, "grad_norm": 0.4870624542236328, "learning_rate": 1.921702027044503e-06, "loss": 0.0393, "step": 91590 }, { "epoch": 3.096791642719497, "grad_norm": 0.20592287182807922, "learning_rate": 1.917164657068593e-06, "loss": 0.0414, "step": 91600 }, { "epoch": 3.0971297204097503, "grad_norm": 0.18993812799453735, "learning_rate": 1.912632545359272e-06, "loss": 0.0384, "step": 91610 }, { "epoch": 3.0974677981000034, "grad_norm": 0.27917662262916565, "learning_rate": 1.9081056924121488e-06, "loss": 0.0379, "step": 91620 }, { "epoch": 3.0978058757902565, "grad_norm": 0.18026100099086761, "learning_rate": 1.9035840987222887e-06, "loss": 0.0422, "step": 91630 }, { "epoch": 3.09814395348051, "grad_norm": 0.13665850460529327, "learning_rate": 1.8990677647841515e-06, "loss": 0.0429, "step": 91640 }, { "epoch": 3.098482031170763, "grad_norm": 0.3867708444595337, "learning_rate": 1.8945566910916534e-06, "loss": 0.0473, "step": 91650 }, { "epoch": 3.098820108861016, "grad_norm": 0.2234116792678833, "learning_rate": 1.8900508781381054e-06, "loss": 0.0432, "step": 91660 }, { "epoch": 3.0991581865512696, "grad_norm": 0.18746474385261536, "learning_rate": 1.8855503264162633e-06, "loss": 0.0427, "step": 91670 }, { "epoch": 3.0994962642415227, "grad_norm": 0.302247017621994, "learning_rate": 1.8810550364182998e-06, "loss": 0.0405, "step": 91680 }, { "epoch": 3.0998343419317758, "grad_norm": 0.4819391369819641, "learning_rate": 1.8765650086358112e-06, "loss": 0.0396, "step": 91690 }, { "epoch": 3.1001724196220293, "grad_norm": 0.14294973015785217, "learning_rate": 1.8720802435598151e-06, "loss": 0.0386, "step": 91700 }, { "epoch": 3.1005104973122823, "grad_norm": 0.5416302680969238, "learning_rate": 1.8676007416807694e-06, "loss": 0.0408, "step": 91710 }, { "epoch": 3.1008485750025354, "grad_norm": 0.2994559407234192, "learning_rate": 1.8631265034885437e-06, "loss": 0.0404, "step": 91720 }, { "epoch": 3.101186652692789, "grad_norm": 0.21087534725666046, "learning_rate": 1.858657529472424e-06, "loss": 0.0392, "step": 91730 }, { "epoch": 3.101524730383042, "grad_norm": 0.4282723367214203, "learning_rate": 1.8541938201211473e-06, "loss": 0.0396, "step": 91740 }, { "epoch": 3.101862808073295, "grad_norm": 0.43849092721939087, "learning_rate": 1.8497353759228398e-06, "loss": 0.0474, "step": 91750 }, { "epoch": 3.1022008857635486, "grad_norm": 0.33443230390548706, "learning_rate": 1.8452821973650835e-06, "loss": 0.0475, "step": 91760 }, { "epoch": 3.1025389634538016, "grad_norm": 0.19693079590797424, "learning_rate": 1.840834284934856e-06, "loss": 0.0386, "step": 91770 }, { "epoch": 3.1028770411440547, "grad_norm": 0.2730579674243927, "learning_rate": 1.8363916391185954e-06, "loss": 0.0431, "step": 91780 }, { "epoch": 3.1032151188343082, "grad_norm": 0.20242707431316376, "learning_rate": 1.8319542604021245e-06, "loss": 0.0404, "step": 91790 }, { "epoch": 3.1035531965245613, "grad_norm": 0.23597106337547302, "learning_rate": 1.8275221492707273e-06, "loss": 0.0382, "step": 91800 }, { "epoch": 3.1038912742148144, "grad_norm": 0.1414879709482193, "learning_rate": 1.8230953062090717e-06, "loss": 0.0349, "step": 91810 }, { "epoch": 3.104229351905068, "grad_norm": 0.2053748071193695, "learning_rate": 1.8186737317012926e-06, "loss": 0.0421, "step": 91820 }, { "epoch": 3.104567429595321, "grad_norm": 0.698711633682251, "learning_rate": 1.8142574262308977e-06, "loss": 0.0445, "step": 91830 }, { "epoch": 3.104905507285574, "grad_norm": 0.17563480138778687, "learning_rate": 1.8098463902808727e-06, "loss": 0.0372, "step": 91840 }, { "epoch": 3.1052435849758275, "grad_norm": 0.2570711374282837, "learning_rate": 1.805440624333593e-06, "loss": 0.046, "step": 91850 }, { "epoch": 3.1055816626660806, "grad_norm": 0.2474663406610489, "learning_rate": 1.8010401288708734e-06, "loss": 0.0397, "step": 91860 }, { "epoch": 3.1059197403563337, "grad_norm": 0.1854906529188156, "learning_rate": 1.7966449043739343e-06, "loss": 0.0408, "step": 91870 }, { "epoch": 3.106257818046587, "grad_norm": 0.16996310651302338, "learning_rate": 1.79225495132343e-06, "loss": 0.0431, "step": 91880 }, { "epoch": 3.1065958957368403, "grad_norm": 0.16096283495426178, "learning_rate": 1.7878702701994487e-06, "loss": 0.0416, "step": 91890 }, { "epoch": 3.1069339734270933, "grad_norm": 0.27843594551086426, "learning_rate": 1.783490861481496e-06, "loss": 0.0367, "step": 91900 }, { "epoch": 3.107272051117347, "grad_norm": 0.2417401224374771, "learning_rate": 1.7791167256484832e-06, "loss": 0.039, "step": 91910 }, { "epoch": 3.1076101288076, "grad_norm": 0.22797925770282745, "learning_rate": 1.774747863178766e-06, "loss": 0.0418, "step": 91920 }, { "epoch": 3.1079482064978534, "grad_norm": 0.22425620257854462, "learning_rate": 1.7703842745501186e-06, "loss": 0.0427, "step": 91930 }, { "epoch": 3.1082862841881065, "grad_norm": 0.18677708506584167, "learning_rate": 1.7660259602397366e-06, "loss": 0.0462, "step": 91940 }, { "epoch": 3.1086243618783596, "grad_norm": 0.2071586400270462, "learning_rate": 1.7616729207242388e-06, "loss": 0.0399, "step": 91950 }, { "epoch": 3.108962439568613, "grad_norm": 0.2784126102924347, "learning_rate": 1.7573251564796556e-06, "loss": 0.0432, "step": 91960 }, { "epoch": 3.109300517258866, "grad_norm": 0.31144019961357117, "learning_rate": 1.7529826679814676e-06, "loss": 0.0414, "step": 91970 }, { "epoch": 3.109638594949119, "grad_norm": 0.28355973958969116, "learning_rate": 1.7486454557045562e-06, "loss": 0.0431, "step": 91980 }, { "epoch": 3.1099766726393727, "grad_norm": 0.27466022968292236, "learning_rate": 1.7443135201232364e-06, "loss": 0.0394, "step": 91990 }, { "epoch": 3.110314750329626, "grad_norm": 0.3143549859523773, "learning_rate": 1.7399868617112347e-06, "loss": 0.0426, "step": 92000 }, { "epoch": 3.110652828019879, "grad_norm": 0.3890441060066223, "learning_rate": 1.7356654809417171e-06, "loss": 0.0396, "step": 92010 }, { "epoch": 3.1109909057101324, "grad_norm": 0.20009365677833557, "learning_rate": 1.7313493782872448e-06, "loss": 0.042, "step": 92020 }, { "epoch": 3.1113289834003854, "grad_norm": 0.3227671980857849, "learning_rate": 1.7270385542198453e-06, "loss": 0.0401, "step": 92030 }, { "epoch": 3.1116670610906385, "grad_norm": 0.1673886924982071, "learning_rate": 1.7227330092109307e-06, "loss": 0.0458, "step": 92040 }, { "epoch": 3.112005138780892, "grad_norm": 0.2823336124420166, "learning_rate": 1.718432743731352e-06, "loss": 0.0436, "step": 92050 }, { "epoch": 3.112343216471145, "grad_norm": 0.2256903201341629, "learning_rate": 1.7141377582513718e-06, "loss": 0.0415, "step": 92060 }, { "epoch": 3.112681294161398, "grad_norm": 0.27517855167388916, "learning_rate": 1.709848053240698e-06, "loss": 0.0409, "step": 92070 }, { "epoch": 3.1130193718516517, "grad_norm": 0.30000054836273193, "learning_rate": 1.7055636291684328e-06, "loss": 0.0426, "step": 92080 }, { "epoch": 3.1133574495419047, "grad_norm": 0.19212932884693146, "learning_rate": 1.7012844865031185e-06, "loss": 0.0433, "step": 92090 }, { "epoch": 3.113695527232158, "grad_norm": 1.341652274131775, "learning_rate": 1.6970106257127138e-06, "loss": 0.0394, "step": 92100 }, { "epoch": 3.1140336049224113, "grad_norm": 0.3690451979637146, "learning_rate": 1.6927420472646117e-06, "loss": 0.0445, "step": 92110 }, { "epoch": 3.1143716826126644, "grad_norm": 0.3214012086391449, "learning_rate": 1.6884787516256106e-06, "loss": 0.0392, "step": 92120 }, { "epoch": 3.1147097603029175, "grad_norm": 0.24522173404693604, "learning_rate": 1.6842207392619325e-06, "loss": 0.0405, "step": 92130 }, { "epoch": 3.115047837993171, "grad_norm": 0.24871252477169037, "learning_rate": 1.679968010639238e-06, "loss": 0.0445, "step": 92140 }, { "epoch": 3.115385915683424, "grad_norm": 0.16649119555950165, "learning_rate": 1.6757205662225829e-06, "loss": 0.0416, "step": 92150 }, { "epoch": 3.115723993373677, "grad_norm": 0.3996776044368744, "learning_rate": 1.6714784064764844e-06, "loss": 0.0425, "step": 92160 }, { "epoch": 3.1160620710639306, "grad_norm": 0.2975854277610779, "learning_rate": 1.6672415318648383e-06, "loss": 0.0446, "step": 92170 }, { "epoch": 3.1164001487541837, "grad_norm": 0.40164101123809814, "learning_rate": 1.6630099428510016e-06, "loss": 0.0462, "step": 92180 }, { "epoch": 3.1167382264444368, "grad_norm": 0.2099616974592209, "learning_rate": 1.6587836398977153e-06, "loss": 0.0428, "step": 92190 }, { "epoch": 3.1170763041346903, "grad_norm": 0.16480764746665955, "learning_rate": 1.6545626234671762e-06, "loss": 0.0436, "step": 92200 }, { "epoch": 3.1174143818249433, "grad_norm": 0.19272743165493011, "learning_rate": 1.650346894020971e-06, "loss": 0.0378, "step": 92210 }, { "epoch": 3.1177524595151964, "grad_norm": 0.18733498454093933, "learning_rate": 1.6461364520201472e-06, "loss": 0.0454, "step": 92220 }, { "epoch": 3.11809053720545, "grad_norm": 0.2962642312049866, "learning_rate": 1.6419312979251366e-06, "loss": 0.0401, "step": 92230 }, { "epoch": 3.118428614895703, "grad_norm": 0.2802219092845917, "learning_rate": 1.6377314321958214e-06, "loss": 0.0434, "step": 92240 }, { "epoch": 3.118766692585956, "grad_norm": 0.17126137018203735, "learning_rate": 1.6335368552914787e-06, "loss": 0.0435, "step": 92250 }, { "epoch": 3.1191047702762096, "grad_norm": 0.278073251247406, "learning_rate": 1.6293475676708358e-06, "loss": 0.0407, "step": 92260 }, { "epoch": 3.1194428479664627, "grad_norm": 0.6304609179496765, "learning_rate": 1.6251635697920097e-06, "loss": 0.042, "step": 92270 }, { "epoch": 3.1197809256567157, "grad_norm": 0.33613789081573486, "learning_rate": 1.6209848621125624e-06, "loss": 0.0396, "step": 92280 }, { "epoch": 3.1201190033469692, "grad_norm": 0.35935699939727783, "learning_rate": 1.6168114450894778e-06, "loss": 0.0463, "step": 92290 }, { "epoch": 3.1204570810372223, "grad_norm": 0.23247402906417847, "learning_rate": 1.612643319179158e-06, "loss": 0.0445, "step": 92300 }, { "epoch": 3.1207951587274754, "grad_norm": 1.1595700979232788, "learning_rate": 1.6084804848374046e-06, "loss": 0.0414, "step": 92310 }, { "epoch": 3.121133236417729, "grad_norm": 0.36117351055145264, "learning_rate": 1.6043229425194817e-06, "loss": 0.0388, "step": 92320 }, { "epoch": 3.121471314107982, "grad_norm": 0.42524006962776184, "learning_rate": 1.6001706926800308e-06, "loss": 0.0482, "step": 92330 }, { "epoch": 3.121809391798235, "grad_norm": 0.25726598501205444, "learning_rate": 1.5960237357731444e-06, "loss": 0.0445, "step": 92340 }, { "epoch": 3.1221474694884885, "grad_norm": 0.1905679553747177, "learning_rate": 1.5918820722523264e-06, "loss": 0.0402, "step": 92350 }, { "epoch": 3.1224855471787416, "grad_norm": 0.28119558095932007, "learning_rate": 1.587745702570509e-06, "loss": 0.0427, "step": 92360 }, { "epoch": 3.122823624868995, "grad_norm": 0.3061910569667816, "learning_rate": 1.5836146271800356e-06, "loss": 0.04, "step": 92370 }, { "epoch": 3.123161702559248, "grad_norm": 1.7714767456054688, "learning_rate": 1.5794888465326728e-06, "loss": 0.0424, "step": 92380 }, { "epoch": 3.1234997802495013, "grad_norm": 0.21635985374450684, "learning_rate": 1.5753683610796155e-06, "loss": 0.0399, "step": 92390 }, { "epoch": 3.1238378579397548, "grad_norm": 0.3452156186103821, "learning_rate": 1.5712531712714528e-06, "loss": 0.0417, "step": 92400 }, { "epoch": 3.124175935630008, "grad_norm": 0.2219020277261734, "learning_rate": 1.5671432775582474e-06, "loss": 0.0427, "step": 92410 }, { "epoch": 3.124514013320261, "grad_norm": 0.2051781564950943, "learning_rate": 1.563038680389428e-06, "loss": 0.0401, "step": 92420 }, { "epoch": 3.1248520910105144, "grad_norm": 0.9484410881996155, "learning_rate": 1.5589393802138808e-06, "loss": 0.0394, "step": 92430 }, { "epoch": 3.1251901687007675, "grad_norm": 0.2772109806537628, "learning_rate": 1.554845377479891e-06, "loss": 0.046, "step": 92440 }, { "epoch": 3.1255282463910206, "grad_norm": 0.7832998633384705, "learning_rate": 1.550756672635173e-06, "loss": 0.0382, "step": 92450 }, { "epoch": 3.125866324081274, "grad_norm": 0.19629840552806854, "learning_rate": 1.5466732661268635e-06, "loss": 0.044, "step": 92460 }, { "epoch": 3.126204401771527, "grad_norm": 0.5575363039970398, "learning_rate": 1.5425951584015163e-06, "loss": 0.046, "step": 92470 }, { "epoch": 3.12654247946178, "grad_norm": 0.2466670572757721, "learning_rate": 1.5385223499051026e-06, "loss": 0.0382, "step": 92480 }, { "epoch": 3.1268805571520337, "grad_norm": 0.32541605830192566, "learning_rate": 1.5344548410830328e-06, "loss": 0.0419, "step": 92490 }, { "epoch": 3.127218634842287, "grad_norm": 0.20695877075195312, "learning_rate": 1.5303926323801121e-06, "loss": 0.0412, "step": 92500 }, { "epoch": 3.12755671253254, "grad_norm": 0.43923482298851013, "learning_rate": 1.5263357242405795e-06, "loss": 0.0382, "step": 92510 }, { "epoch": 3.1278947902227934, "grad_norm": 0.5461732149124146, "learning_rate": 1.5222841171080915e-06, "loss": 0.0449, "step": 92520 }, { "epoch": 3.1282328679130464, "grad_norm": 0.19431357085704803, "learning_rate": 1.5182378114257268e-06, "loss": 0.041, "step": 92530 }, { "epoch": 3.1285709456032995, "grad_norm": 0.2253303974866867, "learning_rate": 1.5141968076359869e-06, "loss": 0.0385, "step": 92540 }, { "epoch": 3.128909023293553, "grad_norm": 0.1777988076210022, "learning_rate": 1.5101611061807908e-06, "loss": 0.0408, "step": 92550 }, { "epoch": 3.129247100983806, "grad_norm": 0.3106941878795624, "learning_rate": 1.506130707501463e-06, "loss": 0.04, "step": 92560 }, { "epoch": 3.129585178674059, "grad_norm": 0.2761642336845398, "learning_rate": 1.5021056120387734e-06, "loss": 0.0391, "step": 92570 }, { "epoch": 3.1299232563643127, "grad_norm": 0.18992473185062408, "learning_rate": 1.4980858202329084e-06, "loss": 0.0481, "step": 92580 }, { "epoch": 3.1302613340545657, "grad_norm": 0.3617629408836365, "learning_rate": 1.4940713325234445e-06, "loss": 0.0455, "step": 92590 }, { "epoch": 3.130599411744819, "grad_norm": 0.3197515308856964, "learning_rate": 1.4900621493494137e-06, "loss": 0.0376, "step": 92600 }, { "epoch": 3.1309374894350723, "grad_norm": 0.3323434889316559, "learning_rate": 1.4860582711492544e-06, "loss": 0.0419, "step": 92610 }, { "epoch": 3.1312755671253254, "grad_norm": 0.2811444401741028, "learning_rate": 1.4820596983608325e-06, "loss": 0.0441, "step": 92620 }, { "epoch": 3.1316136448155785, "grad_norm": 0.1898435652256012, "learning_rate": 1.4780664314214044e-06, "loss": 0.0388, "step": 92630 }, { "epoch": 3.131951722505832, "grad_norm": 0.18300825357437134, "learning_rate": 1.474078470767687e-06, "loss": 0.0408, "step": 92640 }, { "epoch": 3.132289800196085, "grad_norm": 0.24310652911663055, "learning_rate": 1.470095816835787e-06, "loss": 0.0412, "step": 92650 }, { "epoch": 3.132627877886338, "grad_norm": 0.29113760590553284, "learning_rate": 1.4661184700612395e-06, "loss": 0.0435, "step": 92660 }, { "epoch": 3.1329659555765916, "grad_norm": 0.2763058841228485, "learning_rate": 1.4621464308790133e-06, "loss": 0.0468, "step": 92670 }, { "epoch": 3.1333040332668447, "grad_norm": 0.18682992458343506, "learning_rate": 1.4581796997234775e-06, "loss": 0.0402, "step": 92680 }, { "epoch": 3.1336421109570978, "grad_norm": 0.21229659020900726, "learning_rate": 1.4542182770284297e-06, "loss": 0.0406, "step": 92690 }, { "epoch": 3.1339801886473513, "grad_norm": 0.2788633406162262, "learning_rate": 1.450262163227084e-06, "loss": 0.037, "step": 92700 }, { "epoch": 3.1343182663376044, "grad_norm": 0.18530188500881195, "learning_rate": 1.4463113587520671e-06, "loss": 0.0368, "step": 92710 }, { "epoch": 3.1346563440278574, "grad_norm": 0.2180977165699005, "learning_rate": 1.4423658640354443e-06, "loss": 0.0419, "step": 92720 }, { "epoch": 3.134994421718111, "grad_norm": 0.25964441895484924, "learning_rate": 1.438425679508687e-06, "loss": 0.0434, "step": 92730 }, { "epoch": 3.135332499408364, "grad_norm": 0.2333279699087143, "learning_rate": 1.4344908056026895e-06, "loss": 0.0401, "step": 92740 }, { "epoch": 3.135670577098617, "grad_norm": 0.2954440414905548, "learning_rate": 1.4305612427477521e-06, "loss": 0.0431, "step": 92750 }, { "epoch": 3.1360086547888706, "grad_norm": 0.252905011177063, "learning_rate": 1.42663699137362e-06, "loss": 0.041, "step": 92760 }, { "epoch": 3.1363467324791237, "grad_norm": 0.25946664810180664, "learning_rate": 1.4227180519094386e-06, "loss": 0.0473, "step": 92770 }, { "epoch": 3.1366848101693767, "grad_norm": 0.16769158840179443, "learning_rate": 1.4188044247837706e-06, "loss": 0.0421, "step": 92780 }, { "epoch": 3.1370228878596302, "grad_norm": 0.2532590329647064, "learning_rate": 1.4148961104246128e-06, "loss": 0.0402, "step": 92790 }, { "epoch": 3.1373609655498833, "grad_norm": 0.15724927186965942, "learning_rate": 1.4109931092593731e-06, "loss": 0.0416, "step": 92800 }, { "epoch": 3.1376990432401364, "grad_norm": 0.2998181879520416, "learning_rate": 1.4070954217148658e-06, "loss": 0.0408, "step": 92810 }, { "epoch": 3.13803712093039, "grad_norm": 0.5504156947135925, "learning_rate": 1.4032030482173497e-06, "loss": 0.0402, "step": 92820 }, { "epoch": 3.138375198620643, "grad_norm": 0.2839210033416748, "learning_rate": 1.3993159891924901e-06, "loss": 0.0452, "step": 92830 }, { "epoch": 3.138713276310896, "grad_norm": 0.49030759930610657, "learning_rate": 1.3954342450653524e-06, "loss": 0.0425, "step": 92840 }, { "epoch": 3.1390513540011495, "grad_norm": 0.22863009572029114, "learning_rate": 1.3915578162604582e-06, "loss": 0.0407, "step": 92850 }, { "epoch": 3.1393894316914026, "grad_norm": 0.20604631304740906, "learning_rate": 1.3876867032017126e-06, "loss": 0.044, "step": 92860 }, { "epoch": 3.1397275093816557, "grad_norm": 0.19017890095710754, "learning_rate": 1.383820906312472e-06, "loss": 0.044, "step": 92870 }, { "epoch": 3.140065587071909, "grad_norm": 0.17947730422019958, "learning_rate": 1.3799604260154752e-06, "loss": 0.0407, "step": 92880 }, { "epoch": 3.1404036647621623, "grad_norm": 0.2236737608909607, "learning_rate": 1.376105262732913e-06, "loss": 0.0378, "step": 92890 }, { "epoch": 3.1407417424524153, "grad_norm": 0.37980613112449646, "learning_rate": 1.37225541688637e-06, "loss": 0.0415, "step": 92900 }, { "epoch": 3.141079820142669, "grad_norm": 0.5110695958137512, "learning_rate": 1.3684108888968594e-06, "loss": 0.0459, "step": 92910 }, { "epoch": 3.141417897832922, "grad_norm": 0.4471436142921448, "learning_rate": 1.364571679184823e-06, "loss": 0.0414, "step": 92920 }, { "epoch": 3.1417559755231754, "grad_norm": 0.1903192698955536, "learning_rate": 1.3607377881701077e-06, "loss": 0.0399, "step": 92930 }, { "epoch": 3.1420940532134285, "grad_norm": 0.2257426530122757, "learning_rate": 1.3569092162719787e-06, "loss": 0.0413, "step": 92940 }, { "epoch": 3.1424321309036816, "grad_norm": 0.19745002686977386, "learning_rate": 1.353085963909123e-06, "loss": 0.0376, "step": 92950 }, { "epoch": 3.142770208593935, "grad_norm": 0.17284740507602692, "learning_rate": 1.3492680314996454e-06, "loss": 0.0419, "step": 92960 }, { "epoch": 3.143108286284188, "grad_norm": 0.21539251506328583, "learning_rate": 1.345455419461078e-06, "loss": 0.0435, "step": 92970 }, { "epoch": 3.143446363974441, "grad_norm": 0.4391278922557831, "learning_rate": 1.3416481282103431e-06, "loss": 0.0399, "step": 92980 }, { "epoch": 3.1437844416646947, "grad_norm": 0.17037786543369293, "learning_rate": 1.337846158163819e-06, "loss": 0.0404, "step": 92990 }, { "epoch": 3.144122519354948, "grad_norm": 0.2859622538089752, "learning_rate": 1.334049509737273e-06, "loss": 0.0415, "step": 93000 }, { "epoch": 3.144460597045201, "grad_norm": 0.32533952593803406, "learning_rate": 1.3302581833459116e-06, "loss": 0.0388, "step": 93010 }, { "epoch": 3.1447986747354544, "grad_norm": 0.19082137942314148, "learning_rate": 1.3264721794043366e-06, "loss": 0.0368, "step": 93020 }, { "epoch": 3.1451367524257074, "grad_norm": 0.6552472114562988, "learning_rate": 1.3226914983265836e-06, "loss": 0.0464, "step": 93030 }, { "epoch": 3.1454748301159605, "grad_norm": 0.376371294260025, "learning_rate": 1.3189161405260997e-06, "loss": 0.0449, "step": 93040 }, { "epoch": 3.145812907806214, "grad_norm": 0.32026931643486023, "learning_rate": 1.3151461064157656e-06, "loss": 0.0419, "step": 93050 }, { "epoch": 3.146150985496467, "grad_norm": 0.21505428850650787, "learning_rate": 1.3113813964078459e-06, "loss": 0.0449, "step": 93060 }, { "epoch": 3.14648906318672, "grad_norm": 0.2893633246421814, "learning_rate": 1.3076220109140614e-06, "loss": 0.0371, "step": 93070 }, { "epoch": 3.1468271408769737, "grad_norm": 0.22097988426685333, "learning_rate": 1.3038679503455275e-06, "loss": 0.0438, "step": 93080 }, { "epoch": 3.1471652185672268, "grad_norm": 0.22160938382148743, "learning_rate": 1.3001192151127773e-06, "loss": 0.0428, "step": 93090 }, { "epoch": 3.14750329625748, "grad_norm": 0.1659425050020218, "learning_rate": 1.2963758056257657e-06, "loss": 0.0394, "step": 93100 }, { "epoch": 3.1478413739477333, "grad_norm": 0.24058306217193604, "learning_rate": 1.2926377222938768e-06, "loss": 0.042, "step": 93110 }, { "epoch": 3.1481794516379864, "grad_norm": 0.48050299286842346, "learning_rate": 1.2889049655259e-06, "loss": 0.0426, "step": 93120 }, { "epoch": 3.1485175293282395, "grad_norm": 0.24581220746040344, "learning_rate": 1.2851775357300312e-06, "loss": 0.0462, "step": 93130 }, { "epoch": 3.148855607018493, "grad_norm": 0.22357773780822754, "learning_rate": 1.281455433313916e-06, "loss": 0.0409, "step": 93140 }, { "epoch": 3.149193684708746, "grad_norm": 0.18713924288749695, "learning_rate": 1.2777386586845796e-06, "loss": 0.0493, "step": 93150 }, { "epoch": 3.149531762398999, "grad_norm": 0.17995604872703552, "learning_rate": 1.2740272122484964e-06, "loss": 0.0442, "step": 93160 }, { "epoch": 3.1498698400892526, "grad_norm": 0.21450123190879822, "learning_rate": 1.2703210944115308e-06, "loss": 0.0391, "step": 93170 }, { "epoch": 3.1502079177795057, "grad_norm": 0.26144278049468994, "learning_rate": 1.2666203055789916e-06, "loss": 0.0373, "step": 93180 }, { "epoch": 3.1505459954697588, "grad_norm": 0.1608700305223465, "learning_rate": 1.2629248461555832e-06, "loss": 0.0378, "step": 93190 }, { "epoch": 3.1508840731600123, "grad_norm": 0.351310133934021, "learning_rate": 1.2592347165454431e-06, "loss": 0.0457, "step": 93200 }, { "epoch": 3.1512221508502654, "grad_norm": 0.31135129928588867, "learning_rate": 1.2555499171521102e-06, "loss": 0.0421, "step": 93210 }, { "epoch": 3.1515602285405184, "grad_norm": 0.3442924916744232, "learning_rate": 1.251870448378556e-06, "loss": 0.0374, "step": 93220 }, { "epoch": 3.151898306230772, "grad_norm": 0.19674572348594666, "learning_rate": 1.2481963106271478e-06, "loss": 0.0433, "step": 93230 }, { "epoch": 3.152236383921025, "grad_norm": 0.24511639773845673, "learning_rate": 1.2445275042997084e-06, "loss": 0.0418, "step": 93240 }, { "epoch": 3.152574461611278, "grad_norm": 0.16141290962696075, "learning_rate": 1.2408640297974283e-06, "loss": 0.0376, "step": 93250 }, { "epoch": 3.1529125393015316, "grad_norm": 0.19989223778247833, "learning_rate": 1.2372058875209535e-06, "loss": 0.0414, "step": 93260 }, { "epoch": 3.1532506169917847, "grad_norm": 0.275266170501709, "learning_rate": 1.2335530778703252e-06, "loss": 0.0426, "step": 93270 }, { "epoch": 3.1535886946820377, "grad_norm": 0.30920785665512085, "learning_rate": 1.2299056012450182e-06, "loss": 0.0402, "step": 93280 }, { "epoch": 3.1539267723722912, "grad_norm": 0.19258321821689606, "learning_rate": 1.2262634580439026e-06, "loss": 0.042, "step": 93290 }, { "epoch": 3.1542648500625443, "grad_norm": 0.17466330528259277, "learning_rate": 1.2226266486652926e-06, "loss": 0.0378, "step": 93300 }, { "epoch": 3.154602927752798, "grad_norm": 0.332206130027771, "learning_rate": 1.2189951735068927e-06, "loss": 0.0392, "step": 93310 }, { "epoch": 3.154941005443051, "grad_norm": 0.2018096148967743, "learning_rate": 1.2153690329658407e-06, "loss": 0.0457, "step": 93320 }, { "epoch": 3.155279083133304, "grad_norm": 0.1934659630060196, "learning_rate": 1.2117482274386859e-06, "loss": 0.042, "step": 93330 }, { "epoch": 3.1556171608235575, "grad_norm": 0.207834854722023, "learning_rate": 1.2081327573213953e-06, "loss": 0.0415, "step": 93340 }, { "epoch": 3.1559552385138105, "grad_norm": 0.21503286063671112, "learning_rate": 1.2045226230093465e-06, "loss": 0.0392, "step": 93350 }, { "epoch": 3.1562933162040636, "grad_norm": 0.2848395109176636, "learning_rate": 1.2009178248973296e-06, "loss": 0.0396, "step": 93360 }, { "epoch": 3.156631393894317, "grad_norm": 0.22474299371242523, "learning_rate": 1.1973183633795849e-06, "loss": 0.038, "step": 93370 }, { "epoch": 3.15696947158457, "grad_norm": 0.40811604261398315, "learning_rate": 1.193724238849725e-06, "loss": 0.0431, "step": 93380 }, { "epoch": 3.1573075492748233, "grad_norm": 0.26393553614616394, "learning_rate": 1.190135451700808e-06, "loss": 0.0381, "step": 93390 }, { "epoch": 3.157645626965077, "grad_norm": 0.25736284255981445, "learning_rate": 1.186552002325292e-06, "loss": 0.0455, "step": 93400 }, { "epoch": 3.15798370465533, "grad_norm": 0.22490131855010986, "learning_rate": 1.1829738911150578e-06, "loss": 0.0412, "step": 93410 }, { "epoch": 3.158321782345583, "grad_norm": 0.278689444065094, "learning_rate": 1.1794011184613985e-06, "loss": 0.0425, "step": 93420 }, { "epoch": 3.1586598600358364, "grad_norm": 0.26345011591911316, "learning_rate": 1.17583368475504e-06, "loss": 0.0434, "step": 93430 }, { "epoch": 3.1589979377260895, "grad_norm": 0.1519816666841507, "learning_rate": 1.1722715903860982e-06, "loss": 0.0419, "step": 93440 }, { "epoch": 3.1593360154163426, "grad_norm": 0.20188571512699127, "learning_rate": 1.168714835744128e-06, "loss": 0.0411, "step": 93450 }, { "epoch": 3.159674093106596, "grad_norm": 0.3200221061706543, "learning_rate": 1.165163421218085e-06, "loss": 0.0394, "step": 93460 }, { "epoch": 3.160012170796849, "grad_norm": 0.22341066598892212, "learning_rate": 1.1616173471963477e-06, "loss": 0.0365, "step": 93470 }, { "epoch": 3.160350248487102, "grad_norm": 0.2293197065591812, "learning_rate": 1.1580766140667109e-06, "loss": 0.0379, "step": 93480 }, { "epoch": 3.1606883261773557, "grad_norm": 0.6966031789779663, "learning_rate": 1.1545412222163764e-06, "loss": 0.0455, "step": 93490 }, { "epoch": 3.161026403867609, "grad_norm": 0.36856719851493835, "learning_rate": 1.1510111720319795e-06, "loss": 0.0407, "step": 93500 }, { "epoch": 3.161364481557862, "grad_norm": 0.2685832679271698, "learning_rate": 1.1474864638995663e-06, "loss": 0.0419, "step": 93510 }, { "epoch": 3.1617025592481154, "grad_norm": 0.4132108688354492, "learning_rate": 1.1439670982045735e-06, "loss": 0.0425, "step": 93520 }, { "epoch": 3.1620406369383685, "grad_norm": 0.23296739161014557, "learning_rate": 1.1404530753318877e-06, "loss": 0.0442, "step": 93530 }, { "epoch": 3.1623787146286215, "grad_norm": 0.2834162414073944, "learning_rate": 1.136944395665801e-06, "loss": 0.0398, "step": 93540 }, { "epoch": 3.162716792318875, "grad_norm": 0.2612655758857727, "learning_rate": 1.1334410595900014e-06, "loss": 0.0461, "step": 93550 }, { "epoch": 3.163054870009128, "grad_norm": 0.284110963344574, "learning_rate": 1.129943067487621e-06, "loss": 0.0433, "step": 93560 }, { "epoch": 3.163392947699381, "grad_norm": 0.18618246912956238, "learning_rate": 1.126450419741193e-06, "loss": 0.0449, "step": 93570 }, { "epoch": 3.1637310253896347, "grad_norm": 0.16362500190734863, "learning_rate": 1.1229631167326671e-06, "loss": 0.0398, "step": 93580 }, { "epoch": 3.1640691030798878, "grad_norm": 0.22445914149284363, "learning_rate": 1.1194811588434052e-06, "loss": 0.0411, "step": 93590 }, { "epoch": 3.164407180770141, "grad_norm": 0.4209457337856293, "learning_rate": 1.1160045464542023e-06, "loss": 0.0417, "step": 93600 }, { "epoch": 3.1647452584603943, "grad_norm": 0.35407519340515137, "learning_rate": 1.1125332799452325e-06, "loss": 0.0453, "step": 93610 }, { "epoch": 3.1650833361506474, "grad_norm": 0.24725885689258575, "learning_rate": 1.109067359696131e-06, "loss": 0.0434, "step": 93620 }, { "epoch": 3.1654214138409005, "grad_norm": 0.2572175860404968, "learning_rate": 1.1056067860859166e-06, "loss": 0.0466, "step": 93630 }, { "epoch": 3.165759491531154, "grad_norm": 0.30373916029930115, "learning_rate": 1.102151559493031e-06, "loss": 0.0418, "step": 93640 }, { "epoch": 3.166097569221407, "grad_norm": 0.2732129991054535, "learning_rate": 1.0987016802953332e-06, "loss": 0.0391, "step": 93650 }, { "epoch": 3.16643564691166, "grad_norm": 0.2431250363588333, "learning_rate": 1.0952571488701048e-06, "loss": 0.0454, "step": 93660 }, { "epoch": 3.1667737246019136, "grad_norm": 0.22794422507286072, "learning_rate": 1.0918179655940163e-06, "loss": 0.0472, "step": 93670 }, { "epoch": 3.1671118022921667, "grad_norm": 0.2441645711660385, "learning_rate": 1.088384130843184e-06, "loss": 0.046, "step": 93680 }, { "epoch": 3.16744987998242, "grad_norm": 0.33837464451789856, "learning_rate": 1.0849556449931296e-06, "loss": 0.0422, "step": 93690 }, { "epoch": 3.1677879576726733, "grad_norm": 0.19767481088638306, "learning_rate": 1.0815325084187866e-06, "loss": 0.0409, "step": 93700 }, { "epoch": 3.1681260353629264, "grad_norm": 0.22361505031585693, "learning_rate": 1.0781147214944943e-06, "loss": 0.0341, "step": 93710 }, { "epoch": 3.1684641130531794, "grad_norm": 1.0324863195419312, "learning_rate": 1.0747022845940313e-06, "loss": 0.041, "step": 93720 }, { "epoch": 3.168802190743433, "grad_norm": 0.32201942801475525, "learning_rate": 1.0712951980905605e-06, "loss": 0.0399, "step": 93730 }, { "epoch": 3.169140268433686, "grad_norm": 0.3225153684616089, "learning_rate": 1.0678934623566839e-06, "loss": 0.0437, "step": 93740 }, { "epoch": 3.169478346123939, "grad_norm": 0.8411832451820374, "learning_rate": 1.0644970777644091e-06, "loss": 0.0457, "step": 93750 }, { "epoch": 3.1698164238141926, "grad_norm": 0.23382797837257385, "learning_rate": 1.0611060446851673e-06, "loss": 0.0407, "step": 93760 }, { "epoch": 3.1701545015044457, "grad_norm": 0.25023287534713745, "learning_rate": 1.0577203634897836e-06, "loss": 0.0431, "step": 93770 }, { "epoch": 3.1704925791946987, "grad_norm": 0.1959560513496399, "learning_rate": 1.0543400345485178e-06, "loss": 0.0418, "step": 93780 }, { "epoch": 3.1708306568849522, "grad_norm": 0.33148688077926636, "learning_rate": 1.0509650582310404e-06, "loss": 0.038, "step": 93790 }, { "epoch": 3.1711687345752053, "grad_norm": 0.32666298747062683, "learning_rate": 1.0475954349064287e-06, "loss": 0.0363, "step": 93800 }, { "epoch": 3.1715068122654584, "grad_norm": 0.18979871273040771, "learning_rate": 1.0442311649431825e-06, "loss": 0.0402, "step": 93810 }, { "epoch": 3.171844889955712, "grad_norm": 0.3022692799568176, "learning_rate": 1.0408722487092071e-06, "loss": 0.0399, "step": 93820 }, { "epoch": 3.172182967645965, "grad_norm": 0.16643773019313812, "learning_rate": 1.0375186865718424e-06, "loss": 0.0447, "step": 93830 }, { "epoch": 3.172521045336218, "grad_norm": 0.2456437200307846, "learning_rate": 1.0341704788978113e-06, "loss": 0.0435, "step": 93840 }, { "epoch": 3.1728591230264716, "grad_norm": 0.5910263061523438, "learning_rate": 1.0308276260532879e-06, "loss": 0.0389, "step": 93850 }, { "epoch": 3.1731972007167246, "grad_norm": 0.21731217205524445, "learning_rate": 1.0274901284038241e-06, "loss": 0.0412, "step": 93860 }, { "epoch": 3.1735352784069777, "grad_norm": 0.3236415684223175, "learning_rate": 1.0241579863144114e-06, "loss": 0.0438, "step": 93870 }, { "epoch": 3.173873356097231, "grad_norm": 0.17530617117881775, "learning_rate": 1.0208312001494413e-06, "loss": 0.0421, "step": 93880 }, { "epoch": 3.1742114337874843, "grad_norm": 0.19517575204372406, "learning_rate": 1.0175097702727453e-06, "loss": 0.0434, "step": 93890 }, { "epoch": 3.174549511477738, "grad_norm": 0.26132145524024963, "learning_rate": 1.0141936970475275e-06, "loss": 0.045, "step": 93900 }, { "epoch": 3.174887589167991, "grad_norm": 0.2850603461265564, "learning_rate": 1.0108829808364417e-06, "loss": 0.0461, "step": 93910 }, { "epoch": 3.175225666858244, "grad_norm": 0.27552032470703125, "learning_rate": 1.0075776220015376e-06, "loss": 0.0386, "step": 93920 }, { "epoch": 3.1755637445484974, "grad_norm": 0.4694962501525879, "learning_rate": 1.0042776209042816e-06, "loss": 0.0411, "step": 93930 }, { "epoch": 3.1759018222387505, "grad_norm": 0.22633562982082367, "learning_rate": 1.000982977905568e-06, "loss": 0.0395, "step": 93940 }, { "epoch": 3.1762398999290036, "grad_norm": 0.7657662630081177, "learning_rate": 9.976936933656867e-07, "loss": 0.0426, "step": 93950 }, { "epoch": 3.176577977619257, "grad_norm": 0.3980804979801178, "learning_rate": 9.944097676443442e-07, "loss": 0.0501, "step": 93960 }, { "epoch": 3.17691605530951, "grad_norm": 0.3166660964488983, "learning_rate": 9.911312011006757e-07, "loss": 0.0408, "step": 93970 }, { "epoch": 3.1772541329997632, "grad_norm": 0.3617798984050751, "learning_rate": 9.878579940932109e-07, "loss": 0.0468, "step": 93980 }, { "epoch": 3.1775922106900167, "grad_norm": 0.17606663703918457, "learning_rate": 9.84590146979908e-07, "loss": 0.0414, "step": 93990 }, { "epoch": 3.17793028838027, "grad_norm": 0.3821044862270355, "learning_rate": 9.813276601181365e-07, "loss": 0.0401, "step": 94000 }, { "epoch": 3.178268366070523, "grad_norm": 0.34301310777664185, "learning_rate": 9.780705338646724e-07, "loss": 0.0387, "step": 94010 }, { "epoch": 3.1786064437607764, "grad_norm": 0.7191110849380493, "learning_rate": 9.748187685757083e-07, "loss": 0.0418, "step": 94020 }, { "epoch": 3.1789445214510295, "grad_norm": 0.2840895652770996, "learning_rate": 9.715723646068543e-07, "loss": 0.0422, "step": 94030 }, { "epoch": 3.1792825991412825, "grad_norm": 0.3178679347038269, "learning_rate": 9.68331322313143e-07, "loss": 0.042, "step": 94040 }, { "epoch": 3.179620676831536, "grad_norm": 0.19427454471588135, "learning_rate": 9.650956420489909e-07, "loss": 0.046, "step": 94050 }, { "epoch": 3.179958754521789, "grad_norm": 0.1937624216079712, "learning_rate": 9.618653241682595e-07, "loss": 0.0413, "step": 94060 }, { "epoch": 3.180296832212042, "grad_norm": 0.3130143880844116, "learning_rate": 9.586403690242053e-07, "loss": 0.0364, "step": 94070 }, { "epoch": 3.1806349099022957, "grad_norm": 0.25075992941856384, "learning_rate": 9.554207769695122e-07, "loss": 0.0417, "step": 94080 }, { "epoch": 3.1809729875925488, "grad_norm": 0.17019975185394287, "learning_rate": 9.522065483562659e-07, "loss": 0.0395, "step": 94090 }, { "epoch": 3.181311065282802, "grad_norm": 0.42783939838409424, "learning_rate": 9.489976835359681e-07, "loss": 0.0442, "step": 94100 }, { "epoch": 3.1816491429730553, "grad_norm": 0.3246503174304962, "learning_rate": 9.457941828595329e-07, "loss": 0.0419, "step": 94110 }, { "epoch": 3.1819872206633084, "grad_norm": 0.299472838640213, "learning_rate": 9.425960466772965e-07, "loss": 0.0367, "step": 94120 }, { "epoch": 3.1823252983535615, "grad_norm": 0.40827473998069763, "learning_rate": 9.394032753390014e-07, "loss": 0.0392, "step": 94130 }, { "epoch": 3.182663376043815, "grad_norm": 0.3081861436367035, "learning_rate": 9.362158691938017e-07, "loss": 0.037, "step": 94140 }, { "epoch": 3.183001453734068, "grad_norm": 0.2520028352737427, "learning_rate": 9.330338285902685e-07, "loss": 0.0403, "step": 94150 }, { "epoch": 3.183339531424321, "grad_norm": 0.20285382866859436, "learning_rate": 9.298571538763845e-07, "loss": 0.041, "step": 94160 }, { "epoch": 3.1836776091145746, "grad_norm": 0.2326357513666153, "learning_rate": 9.266858453995441e-07, "loss": 0.0399, "step": 94170 }, { "epoch": 3.1840156868048277, "grad_norm": 0.22927738726139069, "learning_rate": 9.235199035065645e-07, "loss": 0.044, "step": 94180 }, { "epoch": 3.184353764495081, "grad_norm": 0.5194789171218872, "learning_rate": 9.203593285436574e-07, "loss": 0.0418, "step": 94190 }, { "epoch": 3.1846918421853343, "grad_norm": 0.3359185755252838, "learning_rate": 9.172041208564686e-07, "loss": 0.0462, "step": 94200 }, { "epoch": 3.1850299198755874, "grad_norm": 0.2348368763923645, "learning_rate": 9.140542807900332e-07, "loss": 0.0442, "step": 94210 }, { "epoch": 3.1853679975658404, "grad_norm": 0.17918901145458221, "learning_rate": 9.109098086888313e-07, "loss": 0.0415, "step": 94220 }, { "epoch": 3.185706075256094, "grad_norm": 0.19637955725193024, "learning_rate": 9.077707048967265e-07, "loss": 0.0369, "step": 94230 }, { "epoch": 3.186044152946347, "grad_norm": 0.28671079874038696, "learning_rate": 9.046369697570057e-07, "loss": 0.0399, "step": 94240 }, { "epoch": 3.1863822306366, "grad_norm": 0.12806996703147888, "learning_rate": 9.015086036123721e-07, "loss": 0.0438, "step": 94250 }, { "epoch": 3.1867203083268536, "grad_norm": 0.2105044424533844, "learning_rate": 8.983856068049413e-07, "loss": 0.0422, "step": 94260 }, { "epoch": 3.1870583860171067, "grad_norm": 0.2632889151573181, "learning_rate": 8.952679796762398e-07, "loss": 0.0425, "step": 94270 }, { "epoch": 3.18739646370736, "grad_norm": 0.2974603772163391, "learning_rate": 8.921557225671951e-07, "loss": 0.0454, "step": 94280 }, { "epoch": 3.1877345413976133, "grad_norm": 0.4081517159938812, "learning_rate": 8.890488358181792e-07, "loss": 0.0401, "step": 94290 }, { "epoch": 3.1880726190878663, "grad_norm": 0.174160897731781, "learning_rate": 8.859473197689372e-07, "loss": 0.0373, "step": 94300 }, { "epoch": 3.18841069677812, "grad_norm": 0.3109252452850342, "learning_rate": 8.828511747586588e-07, "loss": 0.0415, "step": 94310 }, { "epoch": 3.188748774468373, "grad_norm": 0.2069474756717682, "learning_rate": 8.797604011259287e-07, "loss": 0.0467, "step": 94320 }, { "epoch": 3.189086852158626, "grad_norm": 0.2941955327987671, "learning_rate": 8.766749992087487e-07, "loss": 0.0409, "step": 94330 }, { "epoch": 3.1894249298488795, "grad_norm": 0.5605082511901855, "learning_rate": 8.735949693445378e-07, "loss": 0.0359, "step": 94340 }, { "epoch": 3.1897630075391326, "grad_norm": 0.3682219684123993, "learning_rate": 8.705203118701211e-07, "loss": 0.0435, "step": 94350 }, { "epoch": 3.1901010852293856, "grad_norm": 0.19604690372943878, "learning_rate": 8.674510271217295e-07, "loss": 0.0413, "step": 94360 }, { "epoch": 3.190439162919639, "grad_norm": 0.1392562836408615, "learning_rate": 8.643871154350336e-07, "loss": 0.0367, "step": 94370 }, { "epoch": 3.190777240609892, "grad_norm": 0.2291465401649475, "learning_rate": 8.613285771450818e-07, "loss": 0.0427, "step": 94380 }, { "epoch": 3.1911153183001453, "grad_norm": 0.2733204960823059, "learning_rate": 8.582754125863623e-07, "loss": 0.0443, "step": 94390 }, { "epoch": 3.191453395990399, "grad_norm": 0.38856351375579834, "learning_rate": 8.55227622092758e-07, "loss": 0.0405, "step": 94400 }, { "epoch": 3.191791473680652, "grad_norm": 0.41135600209236145, "learning_rate": 8.521852059975743e-07, "loss": 0.0399, "step": 94410 }, { "epoch": 3.192129551370905, "grad_norm": 0.24404321610927582, "learning_rate": 8.491481646335231e-07, "loss": 0.0429, "step": 94420 }, { "epoch": 3.1924676290611584, "grad_norm": 0.32900241017341614, "learning_rate": 8.46116498332733e-07, "loss": 0.0421, "step": 94430 }, { "epoch": 3.1928057067514115, "grad_norm": 0.22668316960334778, "learning_rate": 8.430902074267333e-07, "loss": 0.0453, "step": 94440 }, { "epoch": 3.1931437844416646, "grad_norm": 0.25184166431427, "learning_rate": 8.400692922464925e-07, "loss": 0.0427, "step": 94450 }, { "epoch": 3.193481862131918, "grad_norm": 0.29709547758102417, "learning_rate": 8.370537531223577e-07, "loss": 0.0434, "step": 94460 }, { "epoch": 3.193819939822171, "grad_norm": 0.16192109882831573, "learning_rate": 8.340435903841092e-07, "loss": 0.0419, "step": 94470 }, { "epoch": 3.1941580175124242, "grad_norm": 0.3115587532520294, "learning_rate": 8.310388043609341e-07, "loss": 0.0399, "step": 94480 }, { "epoch": 3.1944960952026777, "grad_norm": 0.30711084604263306, "learning_rate": 8.280393953814303e-07, "loss": 0.0424, "step": 94490 }, { "epoch": 3.194834172892931, "grad_norm": 0.1975693702697754, "learning_rate": 8.250453637736077e-07, "loss": 0.0434, "step": 94500 }, { "epoch": 3.195172250583184, "grad_norm": 0.4564582407474518, "learning_rate": 8.22056709864899e-07, "loss": 0.0448, "step": 94510 }, { "epoch": 3.1955103282734374, "grad_norm": 0.5023984909057617, "learning_rate": 8.190734339821204e-07, "loss": 0.0412, "step": 94520 }, { "epoch": 3.1958484059636905, "grad_norm": 0.39176270365715027, "learning_rate": 8.160955364515333e-07, "loss": 0.0434, "step": 94530 }, { "epoch": 3.1961864836539435, "grad_norm": 0.1759483516216278, "learning_rate": 8.131230175987936e-07, "loss": 0.0421, "step": 94540 }, { "epoch": 3.196524561344197, "grad_norm": 0.2270253300666809, "learning_rate": 8.101558777489693e-07, "loss": 0.044, "step": 94550 }, { "epoch": 3.19686263903445, "grad_norm": 0.19753241539001465, "learning_rate": 8.071941172265396e-07, "loss": 0.0428, "step": 94560 }, { "epoch": 3.197200716724703, "grad_norm": 0.2656221389770508, "learning_rate": 8.042377363553955e-07, "loss": 0.0418, "step": 94570 }, { "epoch": 3.1975387944149567, "grad_norm": 0.2087084949016571, "learning_rate": 8.01286735458856e-07, "loss": 0.0415, "step": 94580 }, { "epoch": 3.1978768721052098, "grad_norm": 0.18726927042007446, "learning_rate": 7.983411148596298e-07, "loss": 0.0414, "step": 94590 }, { "epoch": 3.198214949795463, "grad_norm": 0.5584458112716675, "learning_rate": 7.954008748798425e-07, "loss": 0.043, "step": 94600 }, { "epoch": 3.1985530274857163, "grad_norm": 0.23926791548728943, "learning_rate": 7.924660158410369e-07, "loss": 0.0434, "step": 94610 }, { "epoch": 3.1988911051759694, "grad_norm": 0.2710537016391754, "learning_rate": 7.895365380641728e-07, "loss": 0.0442, "step": 94620 }, { "epoch": 3.1992291828662225, "grad_norm": 0.3231509029865265, "learning_rate": 7.866124418695941e-07, "loss": 0.0432, "step": 94630 }, { "epoch": 3.199567260556476, "grad_norm": 0.15720489621162415, "learning_rate": 7.83693727577095e-07, "loss": 0.0373, "step": 94640 }, { "epoch": 3.199905338246729, "grad_norm": 0.23314917087554932, "learning_rate": 7.807803955058535e-07, "loss": 0.0402, "step": 94650 }, { "epoch": 3.200243415936982, "grad_norm": 0.30028513073921204, "learning_rate": 7.778724459744702e-07, "loss": 0.0391, "step": 94660 }, { "epoch": 3.2005814936272357, "grad_norm": 0.4179155230522156, "learning_rate": 7.749698793009464e-07, "loss": 0.044, "step": 94670 }, { "epoch": 3.2009195713174887, "grad_norm": 0.13842488825321198, "learning_rate": 7.720726958027169e-07, "loss": 0.0419, "step": 94680 }, { "epoch": 3.201257649007742, "grad_norm": 0.3026379644870758, "learning_rate": 7.691808957965951e-07, "loss": 0.0474, "step": 94690 }, { "epoch": 3.2015957266979953, "grad_norm": 0.34994640946388245, "learning_rate": 7.662944795988336e-07, "loss": 0.0488, "step": 94700 }, { "epoch": 3.2019338043882484, "grad_norm": 0.42870745062828064, "learning_rate": 7.634134475250854e-07, "loss": 0.0381, "step": 94710 }, { "epoch": 3.2022718820785014, "grad_norm": 0.24121461808681488, "learning_rate": 7.605377998904262e-07, "loss": 0.037, "step": 94720 }, { "epoch": 3.202609959768755, "grad_norm": 0.3221009373664856, "learning_rate": 7.5766753700931e-07, "loss": 0.0405, "step": 94730 }, { "epoch": 3.202948037459008, "grad_norm": 0.29779818654060364, "learning_rate": 7.54802659195647e-07, "loss": 0.0435, "step": 94740 }, { "epoch": 3.203286115149261, "grad_norm": 0.19596998393535614, "learning_rate": 7.519431667627253e-07, "loss": 0.0403, "step": 94750 }, { "epoch": 3.2036241928395146, "grad_norm": 0.1848125159740448, "learning_rate": 7.490890600232503e-07, "loss": 0.0428, "step": 94760 }, { "epoch": 3.2039622705297677, "grad_norm": 0.2831145226955414, "learning_rate": 7.462403392893502e-07, "loss": 0.0412, "step": 94770 }, { "epoch": 3.2043003482200207, "grad_norm": 0.4613724648952484, "learning_rate": 7.433970048725536e-07, "loss": 0.038, "step": 94780 }, { "epoch": 3.2046384259102743, "grad_norm": 0.33721959590911865, "learning_rate": 7.405590570838062e-07, "loss": 0.0412, "step": 94790 }, { "epoch": 3.2049765036005273, "grad_norm": 0.16665318608283997, "learning_rate": 7.377264962334596e-07, "loss": 0.0391, "step": 94800 }, { "epoch": 3.2053145812907804, "grad_norm": 0.23512142896652222, "learning_rate": 7.348993226312773e-07, "loss": 0.0506, "step": 94810 }, { "epoch": 3.205652658981034, "grad_norm": 0.13419045507907867, "learning_rate": 7.32077536586434e-07, "loss": 0.0408, "step": 94820 }, { "epoch": 3.205990736671287, "grad_norm": 0.17613859474658966, "learning_rate": 7.292611384075221e-07, "loss": 0.0437, "step": 94830 }, { "epoch": 3.20632881436154, "grad_norm": 0.36336106061935425, "learning_rate": 7.264501284025338e-07, "loss": 0.0407, "step": 94840 }, { "epoch": 3.2066668920517936, "grad_norm": 0.27334150671958923, "learning_rate": 7.236445068788789e-07, "loss": 0.0403, "step": 94850 }, { "epoch": 3.2070049697420466, "grad_norm": 0.1800774335861206, "learning_rate": 7.20844274143373e-07, "loss": 0.04, "step": 94860 }, { "epoch": 3.2073430474323, "grad_norm": 0.22540491819381714, "learning_rate": 7.180494305022489e-07, "loss": 0.0446, "step": 94870 }, { "epoch": 3.207681125122553, "grad_norm": 0.21393164992332458, "learning_rate": 7.152599762611456e-07, "loss": 0.0432, "step": 94880 }, { "epoch": 3.2080192028128063, "grad_norm": 0.36132991313934326, "learning_rate": 7.124759117251079e-07, "loss": 0.0371, "step": 94890 }, { "epoch": 3.20835728050306, "grad_norm": 0.15315568447113037, "learning_rate": 7.096972371986033e-07, "loss": 0.0436, "step": 94900 }, { "epoch": 3.208695358193313, "grad_norm": 0.15281987190246582, "learning_rate": 7.069239529855054e-07, "loss": 0.0403, "step": 94910 }, { "epoch": 3.209033435883566, "grad_norm": 0.2921268045902252, "learning_rate": 7.041560593890939e-07, "loss": 0.0436, "step": 94920 }, { "epoch": 3.2093715135738194, "grad_norm": 0.3267890214920044, "learning_rate": 7.013935567120599e-07, "loss": 0.0396, "step": 94930 }, { "epoch": 3.2097095912640725, "grad_norm": 0.30055415630340576, "learning_rate": 6.986364452565009e-07, "loss": 0.0415, "step": 94940 }, { "epoch": 3.2100476689543256, "grad_norm": 0.23744933307170868, "learning_rate": 6.95884725323942e-07, "loss": 0.0432, "step": 94950 }, { "epoch": 3.210385746644579, "grad_norm": 0.48477816581726074, "learning_rate": 6.931383972152982e-07, "loss": 0.0398, "step": 94960 }, { "epoch": 3.210723824334832, "grad_norm": 0.14851544797420502, "learning_rate": 6.903974612309127e-07, "loss": 0.0366, "step": 94970 }, { "epoch": 3.2110619020250852, "grad_norm": 0.20819468796253204, "learning_rate": 6.876619176705179e-07, "loss": 0.0392, "step": 94980 }, { "epoch": 3.2113999797153387, "grad_norm": 0.5384671688079834, "learning_rate": 6.8493176683328e-07, "loss": 0.0395, "step": 94990 }, { "epoch": 3.211738057405592, "grad_norm": 0.22535261511802673, "learning_rate": 6.822070090177657e-07, "loss": 0.0415, "step": 95000 }, { "epoch": 3.212076135095845, "grad_norm": 0.16466157138347626, "learning_rate": 6.794876445219367e-07, "loss": 0.0395, "step": 95010 }, { "epoch": 3.2124142127860984, "grad_norm": 0.8301717638969421, "learning_rate": 6.767736736431885e-07, "loss": 0.0408, "step": 95020 }, { "epoch": 3.2127522904763515, "grad_norm": 0.15283149480819702, "learning_rate": 6.740650966783113e-07, "loss": 0.0408, "step": 95030 }, { "epoch": 3.2130903681666045, "grad_norm": 0.1554657220840454, "learning_rate": 6.713619139235239e-07, "loss": 0.0416, "step": 95040 }, { "epoch": 3.213428445856858, "grad_norm": 0.22322753071784973, "learning_rate": 6.686641256744286e-07, "loss": 0.0408, "step": 95050 }, { "epoch": 3.213766523547111, "grad_norm": 0.20785754919052124, "learning_rate": 6.659717322260562e-07, "loss": 0.0401, "step": 95060 }, { "epoch": 3.214104601237364, "grad_norm": 0.2523755133152008, "learning_rate": 6.632847338728377e-07, "loss": 0.0404, "step": 95070 }, { "epoch": 3.2144426789276177, "grad_norm": 0.2321605533361435, "learning_rate": 6.60603130908627e-07, "loss": 0.0375, "step": 95080 }, { "epoch": 3.2147807566178708, "grad_norm": 0.30566003918647766, "learning_rate": 6.579269236266783e-07, "loss": 0.0422, "step": 95090 }, { "epoch": 3.215118834308124, "grad_norm": 0.24290673434734344, "learning_rate": 6.552561123196632e-07, "loss": 0.0382, "step": 95100 }, { "epoch": 3.2154569119983774, "grad_norm": 0.33509954810142517, "learning_rate": 6.525906972796425e-07, "loss": 0.041, "step": 95110 }, { "epoch": 3.2157949896886304, "grad_norm": 0.28953254222869873, "learning_rate": 6.499306787981108e-07, "loss": 0.0423, "step": 95120 }, { "epoch": 3.2161330673788835, "grad_norm": 0.3186982274055481, "learning_rate": 6.472760571659631e-07, "loss": 0.0404, "step": 95130 }, { "epoch": 3.216471145069137, "grad_norm": 0.21949201822280884, "learning_rate": 6.446268326735061e-07, "loss": 0.044, "step": 95140 }, { "epoch": 3.21680922275939, "grad_norm": 0.2595660984516144, "learning_rate": 6.41983005610458e-07, "loss": 0.0415, "step": 95150 }, { "epoch": 3.217147300449643, "grad_norm": 0.218453511595726, "learning_rate": 6.393445762659378e-07, "loss": 0.0485, "step": 95160 }, { "epoch": 3.2174853781398967, "grad_norm": 0.25851890444755554, "learning_rate": 6.36711544928481e-07, "loss": 0.0426, "step": 95170 }, { "epoch": 3.2178234558301497, "grad_norm": 0.40701785683631897, "learning_rate": 6.340839118860298e-07, "loss": 0.0378, "step": 95180 }, { "epoch": 3.218161533520403, "grad_norm": 0.18165083229541779, "learning_rate": 6.314616774259485e-07, "loss": 0.0412, "step": 95190 }, { "epoch": 3.2184996112106563, "grad_norm": 0.24051165580749512, "learning_rate": 6.288448418349858e-07, "loss": 0.0429, "step": 95200 }, { "epoch": 3.2188376889009094, "grad_norm": 0.20427116751670837, "learning_rate": 6.262334053993235e-07, "loss": 0.0403, "step": 95210 }, { "epoch": 3.2191757665911624, "grad_norm": 0.27905797958374023, "learning_rate": 6.236273684045501e-07, "loss": 0.0438, "step": 95220 }, { "epoch": 3.219513844281416, "grad_norm": 0.25313153862953186, "learning_rate": 6.210267311356488e-07, "loss": 0.041, "step": 95230 }, { "epoch": 3.219851921971669, "grad_norm": 0.1647363007068634, "learning_rate": 6.184314938770197e-07, "loss": 0.0452, "step": 95240 }, { "epoch": 3.2201899996619225, "grad_norm": 0.22054527699947357, "learning_rate": 6.158416569124858e-07, "loss": 0.0428, "step": 95250 }, { "epoch": 3.2205280773521756, "grad_norm": 0.2891022264957428, "learning_rate": 6.132572205252541e-07, "loss": 0.0429, "step": 95260 }, { "epoch": 3.2208661550424287, "grad_norm": 0.2325027734041214, "learning_rate": 6.106781849979648e-07, "loss": 0.0428, "step": 95270 }, { "epoch": 3.221204232732682, "grad_norm": 0.29895108938217163, "learning_rate": 6.081045506126482e-07, "loss": 0.0423, "step": 95280 }, { "epoch": 3.2215423104229353, "grad_norm": 0.4201810956001282, "learning_rate": 6.055363176507678e-07, "loss": 0.0405, "step": 95290 }, { "epoch": 3.2218803881131883, "grad_norm": 0.27525997161865234, "learning_rate": 6.029734863931602e-07, "loss": 0.0447, "step": 95300 }, { "epoch": 3.222218465803442, "grad_norm": 0.40185776352882385, "learning_rate": 6.004160571201179e-07, "loss": 0.038, "step": 95310 }, { "epoch": 3.222556543493695, "grad_norm": 0.22875946760177612, "learning_rate": 5.978640301112947e-07, "loss": 0.0415, "step": 95320 }, { "epoch": 3.222894621183948, "grad_norm": 0.19782033562660217, "learning_rate": 5.9531740564579e-07, "loss": 0.0465, "step": 95330 }, { "epoch": 3.2232326988742015, "grad_norm": 0.36147764325141907, "learning_rate": 5.927761840020918e-07, "loss": 0.0434, "step": 95340 }, { "epoch": 3.2235707765644546, "grad_norm": 0.18150407075881958, "learning_rate": 5.902403654581113e-07, "loss": 0.0421, "step": 95350 }, { "epoch": 3.2239088542547076, "grad_norm": 0.17762888967990875, "learning_rate": 5.877099502911542e-07, "loss": 0.0424, "step": 95360 }, { "epoch": 3.224246931944961, "grad_norm": 0.21415235102176666, "learning_rate": 5.851849387779496e-07, "loss": 0.0447, "step": 95370 }, { "epoch": 3.224585009635214, "grad_norm": 0.2510358393192291, "learning_rate": 5.826653311946262e-07, "loss": 0.0466, "step": 95380 }, { "epoch": 3.2249230873254673, "grad_norm": 0.24011588096618652, "learning_rate": 5.801511278167249e-07, "loss": 0.0395, "step": 95390 }, { "epoch": 3.225261165015721, "grad_norm": 0.2565416991710663, "learning_rate": 5.776423289191868e-07, "loss": 0.0424, "step": 95400 }, { "epoch": 3.225599242705974, "grad_norm": 0.5066211819648743, "learning_rate": 5.751389347763869e-07, "loss": 0.0435, "step": 95410 }, { "epoch": 3.225937320396227, "grad_norm": 0.1840100735425949, "learning_rate": 5.726409456620841e-07, "loss": 0.039, "step": 95420 }, { "epoch": 3.2262753980864805, "grad_norm": 0.14624381065368652, "learning_rate": 5.701483618494541e-07, "loss": 0.037, "step": 95430 }, { "epoch": 3.2266134757767335, "grad_norm": 0.22956176102161407, "learning_rate": 5.67661183611079e-07, "loss": 0.0432, "step": 95440 }, { "epoch": 3.2269515534669866, "grad_norm": 0.2948589026927948, "learning_rate": 5.651794112189579e-07, "loss": 0.0431, "step": 95450 }, { "epoch": 3.22728963115724, "grad_norm": 0.24811992049217224, "learning_rate": 5.627030449444903e-07, "loss": 0.0393, "step": 95460 }, { "epoch": 3.227627708847493, "grad_norm": 0.21298828721046448, "learning_rate": 5.60232085058493e-07, "loss": 0.0445, "step": 95470 }, { "epoch": 3.2279657865377462, "grad_norm": 0.24883590638637543, "learning_rate": 5.577665318311776e-07, "loss": 0.0425, "step": 95480 }, { "epoch": 3.2283038642279998, "grad_norm": 0.3022897243499756, "learning_rate": 5.553063855321838e-07, "loss": 0.0407, "step": 95490 }, { "epoch": 3.228641941918253, "grad_norm": 0.21896669268608093, "learning_rate": 5.528516464305467e-07, "loss": 0.04, "step": 95500 }, { "epoch": 3.228980019608506, "grad_norm": 0.3323383331298828, "learning_rate": 5.504023147947068e-07, "loss": 0.0409, "step": 95510 }, { "epoch": 3.2293180972987594, "grad_norm": 0.22531946003437042, "learning_rate": 5.479583908925279e-07, "loss": 0.0418, "step": 95520 }, { "epoch": 3.2296561749890125, "grad_norm": 0.1323307752609253, "learning_rate": 5.455198749912626e-07, "loss": 0.046, "step": 95530 }, { "epoch": 3.2299942526792655, "grad_norm": 0.27057600021362305, "learning_rate": 5.430867673575979e-07, "loss": 0.046, "step": 95540 }, { "epoch": 3.230332330369519, "grad_norm": 0.29253914952278137, "learning_rate": 5.406590682575985e-07, "loss": 0.042, "step": 95550 }, { "epoch": 3.230670408059772, "grad_norm": 0.17207269370555878, "learning_rate": 5.382367779567687e-07, "loss": 0.042, "step": 95560 }, { "epoch": 3.231008485750025, "grad_norm": 0.28670239448547363, "learning_rate": 5.358198967200023e-07, "loss": 0.0394, "step": 95570 }, { "epoch": 3.2313465634402787, "grad_norm": 0.3121492266654968, "learning_rate": 5.334084248115989e-07, "loss": 0.0426, "step": 95580 }, { "epoch": 3.2316846411305318, "grad_norm": 0.2413039207458496, "learning_rate": 5.310023624952754e-07, "loss": 0.0379, "step": 95590 }, { "epoch": 3.232022718820785, "grad_norm": 0.36944183707237244, "learning_rate": 5.286017100341656e-07, "loss": 0.0424, "step": 95600 }, { "epoch": 3.2323607965110384, "grad_norm": 0.1853713095188141, "learning_rate": 5.262064676907874e-07, "loss": 0.0401, "step": 95610 }, { "epoch": 3.2326988742012914, "grad_norm": 0.2405305653810501, "learning_rate": 5.238166357270924e-07, "loss": 0.0487, "step": 95620 }, { "epoch": 3.2330369518915445, "grad_norm": 0.2349671870470047, "learning_rate": 5.214322144044215e-07, "loss": 0.0428, "step": 95630 }, { "epoch": 3.233375029581798, "grad_norm": 0.19652359187602997, "learning_rate": 5.190532039835327e-07, "loss": 0.0423, "step": 95640 }, { "epoch": 3.233713107272051, "grad_norm": 0.41006311774253845, "learning_rate": 5.166796047245903e-07, "loss": 0.0465, "step": 95650 }, { "epoch": 3.234051184962304, "grad_norm": 0.2211756557226181, "learning_rate": 5.143114168871699e-07, "loss": 0.0439, "step": 95660 }, { "epoch": 3.2343892626525577, "grad_norm": 0.31338533759117126, "learning_rate": 5.119486407302532e-07, "loss": 0.0371, "step": 95670 }, { "epoch": 3.2347273403428107, "grad_norm": 0.2288980633020401, "learning_rate": 5.095912765122335e-07, "loss": 0.0377, "step": 95680 }, { "epoch": 3.235065418033064, "grad_norm": 0.20500658452510834, "learning_rate": 5.07239324490899e-07, "loss": 0.0411, "step": 95690 }, { "epoch": 3.2354034957233173, "grad_norm": 0.2222631424665451, "learning_rate": 5.048927849234608e-07, "loss": 0.0377, "step": 95700 }, { "epoch": 3.2357415734135704, "grad_norm": 0.3140859305858612, "learning_rate": 5.025516580665358e-07, "loss": 0.044, "step": 95710 }, { "epoch": 3.2360796511038235, "grad_norm": 0.19090454280376434, "learning_rate": 5.002159441761412e-07, "loss": 0.0386, "step": 95720 }, { "epoch": 3.236417728794077, "grad_norm": 0.3798080384731293, "learning_rate": 4.978856435077062e-07, "loss": 0.0413, "step": 95730 }, { "epoch": 3.23675580648433, "grad_norm": 0.3802123963832855, "learning_rate": 4.955607563160769e-07, "loss": 0.0399, "step": 95740 }, { "epoch": 3.237093884174583, "grad_norm": 0.22128936648368835, "learning_rate": 4.932412828554944e-07, "loss": 0.0477, "step": 95750 }, { "epoch": 3.2374319618648366, "grad_norm": 0.13546635210514069, "learning_rate": 4.909272233796114e-07, "loss": 0.0427, "step": 95760 }, { "epoch": 3.2377700395550897, "grad_norm": 0.322106271982193, "learning_rate": 4.88618578141492e-07, "loss": 0.0415, "step": 95770 }, { "epoch": 3.2381081172453428, "grad_norm": 0.12923462688922882, "learning_rate": 4.863153473936011e-07, "loss": 0.045, "step": 95780 }, { "epoch": 3.2384461949355963, "grad_norm": 0.26745477318763733, "learning_rate": 4.840175313878259e-07, "loss": 0.0434, "step": 95790 }, { "epoch": 3.2387842726258493, "grad_norm": 0.283595472574234, "learning_rate": 4.817251303754433e-07, "loss": 0.0422, "step": 95800 }, { "epoch": 3.2391223503161024, "grad_norm": 0.35212674736976624, "learning_rate": 4.794381446071583e-07, "loss": 0.0419, "step": 95810 }, { "epoch": 3.239460428006356, "grad_norm": 0.19164469838142395, "learning_rate": 4.771565743330597e-07, "loss": 0.0432, "step": 95820 }, { "epoch": 3.239798505696609, "grad_norm": 0.2672572731971741, "learning_rate": 4.7488041980266463e-07, "loss": 0.0392, "step": 95830 }, { "epoch": 3.2401365833868625, "grad_norm": 0.25418317317962646, "learning_rate": 4.726096812648795e-07, "loss": 0.0447, "step": 95840 }, { "epoch": 3.2404746610771156, "grad_norm": 0.4064464867115021, "learning_rate": 4.703443589680445e-07, "loss": 0.0461, "step": 95850 }, { "epoch": 3.2408127387673686, "grad_norm": 0.2723939120769501, "learning_rate": 4.6808445315987826e-07, "loss": 0.0445, "step": 95860 }, { "epoch": 3.241150816457622, "grad_norm": 0.49006199836730957, "learning_rate": 4.658299640875274e-07, "loss": 0.0451, "step": 95870 }, { "epoch": 3.241488894147875, "grad_norm": 0.2158183753490448, "learning_rate": 4.635808919975393e-07, "loss": 0.0473, "step": 95880 }, { "epoch": 3.2418269718381283, "grad_norm": 0.2260562926530838, "learning_rate": 4.6133723713586706e-07, "loss": 0.045, "step": 95890 }, { "epoch": 3.242165049528382, "grad_norm": 0.23879051208496094, "learning_rate": 4.590989997478756e-07, "loss": 0.042, "step": 95900 }, { "epoch": 3.242503127218635, "grad_norm": 0.27609723806381226, "learning_rate": 4.5686618007833027e-07, "loss": 0.0402, "step": 95910 }, { "epoch": 3.242841204908888, "grad_norm": 0.2395886331796646, "learning_rate": 4.54638778371419e-07, "loss": 0.0449, "step": 95920 }, { "epoch": 3.2431792825991415, "grad_norm": 0.6104511618614197, "learning_rate": 4.524167948707192e-07, "loss": 0.0416, "step": 95930 }, { "epoch": 3.2435173602893945, "grad_norm": 0.2720244824886322, "learning_rate": 4.5020022981921986e-07, "loss": 0.042, "step": 95940 }, { "epoch": 3.2438554379796476, "grad_norm": 0.20385994017124176, "learning_rate": 4.47989083459327e-07, "loss": 0.0387, "step": 95950 }, { "epoch": 3.244193515669901, "grad_norm": 0.22714117169380188, "learning_rate": 4.457833560328528e-07, "loss": 0.0402, "step": 95960 }, { "epoch": 3.244531593360154, "grad_norm": 0.3110351860523224, "learning_rate": 4.4358304778100434e-07, "loss": 0.0416, "step": 95970 }, { "epoch": 3.2448696710504072, "grad_norm": 0.16904997825622559, "learning_rate": 4.413881589444058e-07, "loss": 0.0408, "step": 95980 }, { "epoch": 3.2452077487406608, "grad_norm": 0.19932985305786133, "learning_rate": 4.391986897630873e-07, "loss": 0.0364, "step": 95990 }, { "epoch": 3.245545826430914, "grad_norm": 0.2614150941371918, "learning_rate": 4.370146404764908e-07, "loss": 0.0451, "step": 96000 }, { "epoch": 3.245883904121167, "grad_norm": 0.1309342235326767, "learning_rate": 4.3483601132345285e-07, "loss": 0.0387, "step": 96010 }, { "epoch": 3.2462219818114204, "grad_norm": 0.43853703141212463, "learning_rate": 4.32662802542233e-07, "loss": 0.0388, "step": 96020 }, { "epoch": 3.2465600595016735, "grad_norm": 0.18932680785655975, "learning_rate": 4.3049501437047444e-07, "loss": 0.0377, "step": 96030 }, { "epoch": 3.2468981371919265, "grad_norm": 0.2539174556732178, "learning_rate": 4.283326470452653e-07, "loss": 0.0355, "step": 96040 }, { "epoch": 3.24723621488218, "grad_norm": 0.18991544842720032, "learning_rate": 4.2617570080306647e-07, "loss": 0.0386, "step": 96050 }, { "epoch": 3.247574292572433, "grad_norm": 0.24932169914245605, "learning_rate": 4.2402417587975585e-07, "loss": 0.0427, "step": 96060 }, { "epoch": 3.247912370262686, "grad_norm": 0.15170523524284363, "learning_rate": 4.218780725106286e-07, "loss": 0.0449, "step": 96070 }, { "epoch": 3.2482504479529397, "grad_norm": 0.24851122498512268, "learning_rate": 4.197373909303748e-07, "loss": 0.0427, "step": 96080 }, { "epoch": 3.248588525643193, "grad_norm": 0.3297574520111084, "learning_rate": 4.17602131373096e-07, "loss": 0.0408, "step": 96090 }, { "epoch": 3.248926603333446, "grad_norm": 0.21493418514728546, "learning_rate": 4.154722940723055e-07, "loss": 0.0413, "step": 96100 }, { "epoch": 3.2492646810236994, "grad_norm": 0.2392336130142212, "learning_rate": 4.1334787926091136e-07, "loss": 0.0413, "step": 96110 }, { "epoch": 3.2496027587139524, "grad_norm": 0.15047089755535126, "learning_rate": 4.1122888717124444e-07, "loss": 0.0374, "step": 96120 }, { "epoch": 3.2499408364042055, "grad_norm": 0.41299137473106384, "learning_rate": 4.091153180350249e-07, "loss": 0.0381, "step": 96130 }, { "epoch": 3.250278914094459, "grad_norm": 0.22900056838989258, "learning_rate": 4.0700717208340125e-07, "loss": 0.0423, "step": 96140 }, { "epoch": 3.250616991784712, "grad_norm": 0.22086495161056519, "learning_rate": 4.049044495469112e-07, "loss": 0.0423, "step": 96150 }, { "epoch": 3.250955069474965, "grad_norm": 0.19958560168743134, "learning_rate": 4.0280715065549866e-07, "loss": 0.0406, "step": 96160 }, { "epoch": 3.2512931471652187, "grad_norm": 0.24756942689418793, "learning_rate": 4.0071527563853575e-07, "loss": 0.0427, "step": 96170 }, { "epoch": 3.2516312248554717, "grad_norm": 0.24513758718967438, "learning_rate": 3.9862882472477823e-07, "loss": 0.0387, "step": 96180 }, { "epoch": 3.2519693025457252, "grad_norm": 0.37327656149864197, "learning_rate": 3.9654779814239375e-07, "loss": 0.0441, "step": 96190 }, { "epoch": 3.2523073802359783, "grad_norm": 0.27704185247421265, "learning_rate": 3.944721961189668e-07, "loss": 0.0401, "step": 96200 }, { "epoch": 3.2526454579262314, "grad_norm": 0.1948571652173996, "learning_rate": 3.9240201888147697e-07, "loss": 0.0444, "step": 96210 }, { "epoch": 3.252983535616485, "grad_norm": 0.37196236848831177, "learning_rate": 3.9033726665632096e-07, "loss": 0.044, "step": 96220 }, { "epoch": 3.253321613306738, "grad_norm": 1.127863883972168, "learning_rate": 3.882779396692959e-07, "loss": 0.0387, "step": 96230 }, { "epoch": 3.253659690996991, "grad_norm": 0.3846845030784607, "learning_rate": 3.8622403814559947e-07, "loss": 0.0411, "step": 96240 }, { "epoch": 3.2539977686872446, "grad_norm": 0.23586329817771912, "learning_rate": 3.841755623098575e-07, "loss": 0.0376, "step": 96250 }, { "epoch": 3.2543358463774976, "grad_norm": 1.1148185729980469, "learning_rate": 3.821325123860742e-07, "loss": 0.0454, "step": 96260 }, { "epoch": 3.2546739240677507, "grad_norm": 0.2424718141555786, "learning_rate": 3.800948885976818e-07, "loss": 0.0407, "step": 96270 }, { "epoch": 3.255012001758004, "grad_norm": 0.4730411469936371, "learning_rate": 3.780626911675078e-07, "loss": 0.0441, "step": 96280 }, { "epoch": 3.2553500794482573, "grad_norm": 0.28749820590019226, "learning_rate": 3.76035920317791e-07, "loss": 0.0371, "step": 96290 }, { "epoch": 3.2556881571385103, "grad_norm": 0.3266129195690155, "learning_rate": 3.740145762701819e-07, "loss": 0.0415, "step": 96300 }, { "epoch": 3.256026234828764, "grad_norm": 0.18400461971759796, "learning_rate": 3.7199865924572585e-07, "loss": 0.0413, "step": 96310 }, { "epoch": 3.256364312519017, "grad_norm": 0.3078025281429291, "learning_rate": 3.699881694648799e-07, "loss": 0.0393, "step": 96320 }, { "epoch": 3.25670239020927, "grad_norm": 0.35863372683525085, "learning_rate": 3.679831071475126e-07, "loss": 0.0449, "step": 96330 }, { "epoch": 3.2570404678995235, "grad_norm": 0.18432122468948364, "learning_rate": 3.65983472512893e-07, "loss": 0.04, "step": 96340 }, { "epoch": 3.2573785455897766, "grad_norm": 0.36677029728889465, "learning_rate": 3.6398926577969616e-07, "loss": 0.0405, "step": 96350 }, { "epoch": 3.2577166232800296, "grad_norm": 0.2741076350212097, "learning_rate": 3.620004871660032e-07, "loss": 0.0376, "step": 96360 }, { "epoch": 3.258054700970283, "grad_norm": 0.32085421681404114, "learning_rate": 3.600171368893124e-07, "loss": 0.0436, "step": 96370 }, { "epoch": 3.2583927786605362, "grad_norm": 0.5205184817314148, "learning_rate": 3.580392151665113e-07, "loss": 0.0435, "step": 96380 }, { "epoch": 3.2587308563507893, "grad_norm": 0.3524744510650635, "learning_rate": 3.5606672221391024e-07, "loss": 0.0444, "step": 96390 }, { "epoch": 3.259068934041043, "grad_norm": 0.3584645390510559, "learning_rate": 3.540996582472089e-07, "loss": 0.044, "step": 96400 }, { "epoch": 3.259407011731296, "grad_norm": 0.3098980486392975, "learning_rate": 3.521380234815297e-07, "loss": 0.0449, "step": 96410 }, { "epoch": 3.259745089421549, "grad_norm": 0.2316523939371109, "learning_rate": 3.5018181813138985e-07, "loss": 0.0414, "step": 96420 }, { "epoch": 3.2600831671118025, "grad_norm": 0.24618127942085266, "learning_rate": 3.482310424107238e-07, "loss": 0.0436, "step": 96430 }, { "epoch": 3.2604212448020555, "grad_norm": 0.3354799747467041, "learning_rate": 3.4628569653285537e-07, "loss": 0.0384, "step": 96440 }, { "epoch": 3.2607593224923086, "grad_norm": 0.131836399435997, "learning_rate": 3.44345780710531e-07, "loss": 0.0402, "step": 96450 }, { "epoch": 3.261097400182562, "grad_norm": 0.19664902985095978, "learning_rate": 3.424112951558978e-07, "loss": 0.0435, "step": 96460 }, { "epoch": 3.261435477872815, "grad_norm": 0.24266140162944794, "learning_rate": 3.40482240080503e-07, "loss": 0.0405, "step": 96470 }, { "epoch": 3.2617735555630682, "grad_norm": 0.2796211540699005, "learning_rate": 3.3855861569531133e-07, "loss": 0.0395, "step": 96480 }, { "epoch": 3.2621116332533218, "grad_norm": 0.32929274439811707, "learning_rate": 3.3664042221068226e-07, "loss": 0.041, "step": 96490 }, { "epoch": 3.262449710943575, "grad_norm": 0.32014036178588867, "learning_rate": 3.347276598363924e-07, "loss": 0.0418, "step": 96500 }, { "epoch": 3.262787788633828, "grad_norm": 0.33645665645599365, "learning_rate": 3.328203287816134e-07, "loss": 0.0514, "step": 96510 }, { "epoch": 3.2631258663240814, "grad_norm": 0.2213595062494278, "learning_rate": 3.3091842925493386e-07, "loss": 0.0351, "step": 96520 }, { "epoch": 3.2634639440143345, "grad_norm": 0.24936141073703766, "learning_rate": 3.2902196146433196e-07, "loss": 0.0425, "step": 96530 }, { "epoch": 3.2638020217045876, "grad_norm": 0.16847726702690125, "learning_rate": 3.2713092561720836e-07, "loss": 0.0377, "step": 96540 }, { "epoch": 3.264140099394841, "grad_norm": 0.4953534007072449, "learning_rate": 3.2524532192036996e-07, "loss": 0.036, "step": 96550 }, { "epoch": 3.264478177085094, "grad_norm": 0.2705223262310028, "learning_rate": 3.2336515058001836e-07, "loss": 0.0396, "step": 96560 }, { "epoch": 3.264816254775347, "grad_norm": 0.19220341742038727, "learning_rate": 3.2149041180176143e-07, "loss": 0.0398, "step": 96570 }, { "epoch": 3.2651543324656007, "grad_norm": 0.23001812398433685, "learning_rate": 3.1962110579062954e-07, "loss": 0.0406, "step": 96580 }, { "epoch": 3.265492410155854, "grad_norm": 0.26985177397727966, "learning_rate": 3.177572327510314e-07, "loss": 0.0372, "step": 96590 }, { "epoch": 3.265830487846107, "grad_norm": 0.7109028100967407, "learning_rate": 3.158987928868151e-07, "loss": 0.0402, "step": 96600 }, { "epoch": 3.2661685655363604, "grad_norm": 0.34355002641677856, "learning_rate": 3.1404578640120144e-07, "loss": 0.0463, "step": 96610 }, { "epoch": 3.2665066432266134, "grad_norm": 0.38499316573143005, "learning_rate": 3.1219821349683933e-07, "loss": 0.0419, "step": 96620 }, { "epoch": 3.2668447209168665, "grad_norm": 0.2535040080547333, "learning_rate": 3.103560743757783e-07, "loss": 0.037, "step": 96630 }, { "epoch": 3.26718279860712, "grad_norm": 0.5022460222244263, "learning_rate": 3.0851936923946834e-07, "loss": 0.0436, "step": 96640 }, { "epoch": 3.267520876297373, "grad_norm": 0.21329063177108765, "learning_rate": 3.06688098288771e-07, "loss": 0.0403, "step": 96650 }, { "epoch": 3.267858953987626, "grad_norm": 0.2710123360157013, "learning_rate": 3.0486226172394825e-07, "loss": 0.0388, "step": 96660 }, { "epoch": 3.2681970316778797, "grad_norm": 0.1703164130449295, "learning_rate": 3.030418597446738e-07, "loss": 0.0391, "step": 96670 }, { "epoch": 3.2685351093681327, "grad_norm": 0.25611603260040283, "learning_rate": 3.012268925500272e-07, "loss": 0.0403, "step": 96680 }, { "epoch": 3.268873187058386, "grad_norm": 0.20235900580883026, "learning_rate": 2.994173603384831e-07, "loss": 0.0386, "step": 96690 }, { "epoch": 3.2692112647486393, "grad_norm": 0.2001018524169922, "learning_rate": 2.9761326330793317e-07, "loss": 0.0434, "step": 96700 }, { "epoch": 3.2695493424388924, "grad_norm": 0.23213498294353485, "learning_rate": 2.958146016556751e-07, "loss": 0.042, "step": 96710 }, { "epoch": 3.2698874201291455, "grad_norm": 0.28176525235176086, "learning_rate": 2.94021375578396e-07, "loss": 0.0384, "step": 96720 }, { "epoch": 3.270225497819399, "grad_norm": 0.1572154313325882, "learning_rate": 2.9223358527221135e-07, "loss": 0.0423, "step": 96730 }, { "epoch": 3.270563575509652, "grad_norm": 0.24990373849868774, "learning_rate": 2.904512309326257e-07, "loss": 0.0443, "step": 96740 }, { "epoch": 3.270901653199905, "grad_norm": 0.18488797545433044, "learning_rate": 2.8867431275455545e-07, "loss": 0.0452, "step": 96750 }, { "epoch": 3.2712397308901586, "grad_norm": 0.2594158947467804, "learning_rate": 2.86902830932323e-07, "loss": 0.0447, "step": 96760 }, { "epoch": 3.2715778085804117, "grad_norm": 0.4684549868106842, "learning_rate": 2.851367856596565e-07, "loss": 0.0386, "step": 96770 }, { "epoch": 3.2719158862706648, "grad_norm": 0.35253116488456726, "learning_rate": 2.83376177129685e-07, "loss": 0.0379, "step": 96780 }, { "epoch": 3.2722539639609183, "grad_norm": 0.2782133221626282, "learning_rate": 2.8162100553494884e-07, "loss": 0.0402, "step": 96790 }, { "epoch": 3.2725920416511713, "grad_norm": 0.22224003076553345, "learning_rate": 2.798712710673834e-07, "loss": 0.0418, "step": 96800 }, { "epoch": 3.2729301193414244, "grad_norm": 0.20011891424655914, "learning_rate": 2.7812697391834676e-07, "loss": 0.0455, "step": 96810 }, { "epoch": 3.273268197031678, "grad_norm": 0.3212184011936188, "learning_rate": 2.7638811427858627e-07, "loss": 0.0469, "step": 96820 }, { "epoch": 3.273606274721931, "grad_norm": 0.22863705456256866, "learning_rate": 2.746546923382665e-07, "loss": 0.0401, "step": 96830 }, { "epoch": 3.2739443524121845, "grad_norm": 0.1889420747756958, "learning_rate": 2.729267082869469e-07, "loss": 0.0462, "step": 96840 }, { "epoch": 3.2742824301024376, "grad_norm": 0.22164247930049896, "learning_rate": 2.712041623135986e-07, "loss": 0.0422, "step": 96850 }, { "epoch": 3.2746205077926906, "grad_norm": 0.19225256145000458, "learning_rate": 2.694870546065931e-07, "loss": 0.0448, "step": 96860 }, { "epoch": 3.274958585482944, "grad_norm": 0.1823880970478058, "learning_rate": 2.677753853537246e-07, "loss": 0.0399, "step": 96870 }, { "epoch": 3.2752966631731972, "grad_norm": 0.273857444524765, "learning_rate": 2.660691547421601e-07, "loss": 0.0413, "step": 96880 }, { "epoch": 3.2756347408634503, "grad_norm": 0.260576456785202, "learning_rate": 2.6436836295850033e-07, "loss": 0.0434, "step": 96890 }, { "epoch": 3.275972818553704, "grad_norm": 0.31766819953918457, "learning_rate": 2.6267301018874093e-07, "loss": 0.041, "step": 96900 }, { "epoch": 3.276310896243957, "grad_norm": 0.25269103050231934, "learning_rate": 2.6098309661828355e-07, "loss": 0.0418, "step": 96910 }, { "epoch": 3.27664897393421, "grad_norm": 0.2760675847530365, "learning_rate": 2.592986224319305e-07, "loss": 0.0424, "step": 96920 }, { "epoch": 3.2769870516244635, "grad_norm": 0.35276493430137634, "learning_rate": 2.576195878139009e-07, "loss": 0.0428, "step": 96930 }, { "epoch": 3.2773251293147165, "grad_norm": 0.8328345417976379, "learning_rate": 2.5594599294780364e-07, "loss": 0.0374, "step": 96940 }, { "epoch": 3.2776632070049696, "grad_norm": 0.2456035315990448, "learning_rate": 2.542778380166644e-07, "loss": 0.045, "step": 96950 }, { "epoch": 3.278001284695223, "grad_norm": 0.42888444662094116, "learning_rate": 2.526151232029095e-07, "loss": 0.0432, "step": 96960 }, { "epoch": 3.278339362385476, "grad_norm": 0.365597665309906, "learning_rate": 2.5095784868836567e-07, "loss": 0.041, "step": 96970 }, { "epoch": 3.2786774400757293, "grad_norm": 0.2762519121170044, "learning_rate": 2.493060146542825e-07, "loss": 0.0458, "step": 96980 }, { "epoch": 3.2790155177659828, "grad_norm": 0.17596110701560974, "learning_rate": 2.476596212812876e-07, "loss": 0.0444, "step": 96990 }, { "epoch": 3.279353595456236, "grad_norm": 0.14813552796840668, "learning_rate": 2.4601866874943703e-07, "loss": 0.0366, "step": 97000 }, { "epoch": 3.279691673146489, "grad_norm": 0.15108069777488708, "learning_rate": 2.4438315723818163e-07, "loss": 0.0426, "step": 97010 }, { "epoch": 3.2800297508367424, "grad_norm": 0.3062770664691925, "learning_rate": 2.4275308692637833e-07, "loss": 0.0402, "step": 97020 }, { "epoch": 3.2803678285269955, "grad_norm": 0.21367456018924713, "learning_rate": 2.411284579922901e-07, "loss": 0.0408, "step": 97030 }, { "epoch": 3.2807059062172486, "grad_norm": 0.18805846571922302, "learning_rate": 2.395092706135804e-07, "loss": 0.044, "step": 97040 }, { "epoch": 3.281043983907502, "grad_norm": 0.3292165994644165, "learning_rate": 2.3789552496731872e-07, "loss": 0.04, "step": 97050 }, { "epoch": 3.281382061597755, "grad_norm": 0.6153271794319153, "learning_rate": 2.3628722122999158e-07, "loss": 0.0424, "step": 97060 }, { "epoch": 3.281720139288008, "grad_norm": 0.3171014189720154, "learning_rate": 2.3468435957747503e-07, "loss": 0.0429, "step": 97070 }, { "epoch": 3.2820582169782617, "grad_norm": 0.8827620148658752, "learning_rate": 2.330869401850566e-07, "loss": 0.036, "step": 97080 }, { "epoch": 3.282396294668515, "grad_norm": 0.2812298536300659, "learning_rate": 2.3149496322742437e-07, "loss": 0.0415, "step": 97090 }, { "epoch": 3.282734372358768, "grad_norm": 0.17955677211284637, "learning_rate": 2.2990842887868346e-07, "loss": 0.046, "step": 97100 }, { "epoch": 3.2830724500490214, "grad_norm": 0.2941180169582367, "learning_rate": 2.2832733731232292e-07, "loss": 0.04, "step": 97110 }, { "epoch": 3.2834105277392744, "grad_norm": 0.44322940707206726, "learning_rate": 2.2675168870124886e-07, "loss": 0.0436, "step": 97120 }, { "epoch": 3.2837486054295275, "grad_norm": 0.11150599271059036, "learning_rate": 2.2518148321778455e-07, "loss": 0.0395, "step": 97130 }, { "epoch": 3.284086683119781, "grad_norm": 0.24157120287418365, "learning_rate": 2.2361672103363706e-07, "loss": 0.0398, "step": 97140 }, { "epoch": 3.284424760810034, "grad_norm": 0.3277466595172882, "learning_rate": 2.220574023199251e-07, "loss": 0.0412, "step": 97150 }, { "epoch": 3.2847628385002876, "grad_norm": 0.15932612121105194, "learning_rate": 2.2050352724717894e-07, "loss": 0.0397, "step": 97160 }, { "epoch": 3.2851009161905407, "grad_norm": 0.17532433569431305, "learning_rate": 2.1895509598532372e-07, "loss": 0.0427, "step": 97170 }, { "epoch": 3.2854389938807937, "grad_norm": 0.1702956110239029, "learning_rate": 2.1741210870369066e-07, "loss": 0.0403, "step": 97180 }, { "epoch": 3.2857770715710473, "grad_norm": 0.3668205440044403, "learning_rate": 2.1587456557102814e-07, "loss": 0.0396, "step": 97190 }, { "epoch": 3.2861151492613003, "grad_norm": 0.33320194482803345, "learning_rate": 2.143424667554683e-07, "loss": 0.0384, "step": 97200 }, { "epoch": 3.2864532269515534, "grad_norm": 0.22360776364803314, "learning_rate": 2.128158124245716e-07, "loss": 0.0414, "step": 97210 }, { "epoch": 3.286791304641807, "grad_norm": 0.29841938614845276, "learning_rate": 2.1129460274527668e-07, "loss": 0.044, "step": 97220 }, { "epoch": 3.28712938233206, "grad_norm": 0.2330160140991211, "learning_rate": 2.0977883788395046e-07, "loss": 0.0377, "step": 97230 }, { "epoch": 3.287467460022313, "grad_norm": 0.2661716043949127, "learning_rate": 2.0826851800634928e-07, "loss": 0.0431, "step": 97240 }, { "epoch": 3.2878055377125666, "grad_norm": 0.34679627418518066, "learning_rate": 2.0676364327764653e-07, "loss": 0.0355, "step": 97250 }, { "epoch": 3.2881436154028196, "grad_norm": 0.2118871659040451, "learning_rate": 2.0526421386240502e-07, "loss": 0.0423, "step": 97260 }, { "epoch": 3.2884816930930727, "grad_norm": 0.1944638341665268, "learning_rate": 2.0377022992461025e-07, "loss": 0.0481, "step": 97270 }, { "epoch": 3.288819770783326, "grad_norm": 0.47506478428840637, "learning_rate": 2.0228169162763156e-07, "loss": 0.0451, "step": 97280 }, { "epoch": 3.2891578484735793, "grad_norm": 0.2326684147119522, "learning_rate": 2.0079859913425536e-07, "loss": 0.0406, "step": 97290 }, { "epoch": 3.2894959261638324, "grad_norm": 0.24432745575904846, "learning_rate": 1.9932095260667417e-07, "loss": 0.0425, "step": 97300 }, { "epoch": 3.289834003854086, "grad_norm": 0.2190612107515335, "learning_rate": 1.9784875220648092e-07, "loss": 0.047, "step": 97310 }, { "epoch": 3.290172081544339, "grad_norm": 0.27443426847457886, "learning_rate": 1.9638199809466907e-07, "loss": 0.0389, "step": 97320 }, { "epoch": 3.290510159234592, "grad_norm": 0.22257117927074432, "learning_rate": 1.9492069043164364e-07, "loss": 0.0487, "step": 97330 }, { "epoch": 3.2908482369248455, "grad_norm": 0.23439106345176697, "learning_rate": 1.9346482937721567e-07, "loss": 0.0401, "step": 97340 }, { "epoch": 3.2911863146150986, "grad_norm": 0.19580954313278198, "learning_rate": 1.920144150905856e-07, "loss": 0.0392, "step": 97350 }, { "epoch": 3.2915243923053517, "grad_norm": 0.22406142950057983, "learning_rate": 1.9056944773037655e-07, "loss": 0.0432, "step": 97360 }, { "epoch": 3.291862469995605, "grad_norm": 0.9404149651527405, "learning_rate": 1.8912992745460102e-07, "loss": 0.0462, "step": 97370 }, { "epoch": 3.2922005476858582, "grad_norm": 0.23124602437019348, "learning_rate": 1.876958544206886e-07, "loss": 0.0407, "step": 97380 }, { "epoch": 3.2925386253761113, "grad_norm": 0.15722782909870148, "learning_rate": 1.8626722878546942e-07, "loss": 0.0447, "step": 97390 }, { "epoch": 3.292876703066365, "grad_norm": 0.19558753073215485, "learning_rate": 1.8484405070516851e-07, "loss": 0.0391, "step": 97400 }, { "epoch": 3.293214780756618, "grad_norm": 0.29833659529685974, "learning_rate": 1.8342632033542805e-07, "loss": 0.0383, "step": 97410 }, { "epoch": 3.293552858446871, "grad_norm": 0.1263774335384369, "learning_rate": 1.8201403783129066e-07, "loss": 0.0381, "step": 97420 }, { "epoch": 3.2938909361371245, "grad_norm": 0.2334565371274948, "learning_rate": 1.806072033471884e-07, "loss": 0.0371, "step": 97430 }, { "epoch": 3.2942290138273775, "grad_norm": 0.3626197874546051, "learning_rate": 1.79205817036987e-07, "loss": 0.0441, "step": 97440 }, { "epoch": 3.2945670915176306, "grad_norm": 0.32688409090042114, "learning_rate": 1.7780987905393065e-07, "loss": 0.0511, "step": 97450 }, { "epoch": 3.294905169207884, "grad_norm": 0.3506099581718445, "learning_rate": 1.764193895506805e-07, "loss": 0.0429, "step": 97460 }, { "epoch": 3.295243246898137, "grad_norm": 0.29127174615859985, "learning_rate": 1.7503434867929824e-07, "loss": 0.0419, "step": 97470 }, { "epoch": 3.2955813245883903, "grad_norm": 0.5363515615463257, "learning_rate": 1.7365475659125165e-07, "loss": 0.043, "step": 97480 }, { "epoch": 3.2959194022786438, "grad_norm": 0.24956724047660828, "learning_rate": 1.7228061343740332e-07, "loss": 0.0374, "step": 97490 }, { "epoch": 3.296257479968897, "grad_norm": 0.16661906242370605, "learning_rate": 1.7091191936803308e-07, "loss": 0.0391, "step": 97500 }, { "epoch": 3.29659555765915, "grad_norm": 0.268497109413147, "learning_rate": 1.6954867453281563e-07, "loss": 0.0401, "step": 97510 }, { "epoch": 3.2969336353494034, "grad_norm": 0.1902364045381546, "learning_rate": 1.6819087908084285e-07, "loss": 0.0415, "step": 97520 }, { "epoch": 3.2972717130396565, "grad_norm": 0.17599326372146606, "learning_rate": 1.6683853316058485e-07, "loss": 0.0397, "step": 97530 }, { "epoch": 3.2976097907299096, "grad_norm": 0.282388836145401, "learning_rate": 1.6549163691995106e-07, "loss": 0.0446, "step": 97540 }, { "epoch": 3.297947868420163, "grad_norm": 0.2945542335510254, "learning_rate": 1.641501905062237e-07, "loss": 0.047, "step": 97550 }, { "epoch": 3.298285946110416, "grad_norm": 0.2741572856903076, "learning_rate": 1.628141940661021e-07, "loss": 0.0417, "step": 97560 }, { "epoch": 3.298624023800669, "grad_norm": 0.27753132581710815, "learning_rate": 1.6148364774569158e-07, "loss": 0.0396, "step": 97570 }, { "epoch": 3.2989621014909227, "grad_norm": 0.14892971515655518, "learning_rate": 1.6015855169050355e-07, "loss": 0.0383, "step": 97580 }, { "epoch": 3.299300179181176, "grad_norm": 0.37179824709892273, "learning_rate": 1.5883890604543873e-07, "loss": 0.0431, "step": 97590 }, { "epoch": 3.299638256871429, "grad_norm": 0.23236523568630219, "learning_rate": 1.5752471095482057e-07, "loss": 0.041, "step": 97600 }, { "epoch": 3.2999763345616824, "grad_norm": 0.22175459563732147, "learning_rate": 1.5621596656235638e-07, "loss": 0.0406, "step": 97610 }, { "epoch": 3.3003144122519354, "grad_norm": 0.30140966176986694, "learning_rate": 1.5491267301117608e-07, "loss": 0.0364, "step": 97620 }, { "epoch": 3.3006524899421885, "grad_norm": 0.27752646803855896, "learning_rate": 1.536148304438101e-07, "loss": 0.0377, "step": 97630 }, { "epoch": 3.300990567632442, "grad_norm": 0.12429256737232208, "learning_rate": 1.5232243900217823e-07, "loss": 0.0348, "step": 97640 }, { "epoch": 3.301328645322695, "grad_norm": 0.285245805978775, "learning_rate": 1.51035498827623e-07, "loss": 0.0428, "step": 97650 }, { "epoch": 3.301666723012948, "grad_norm": 0.2113341987133026, "learning_rate": 1.497540100608763e-07, "loss": 0.0408, "step": 97660 }, { "epoch": 3.3020048007032017, "grad_norm": 0.20926563441753387, "learning_rate": 1.4847797284208153e-07, "loss": 0.0387, "step": 97670 }, { "epoch": 3.3023428783934548, "grad_norm": 0.1670147031545639, "learning_rate": 1.472073873107882e-07, "loss": 0.0376, "step": 97680 }, { "epoch": 3.302680956083708, "grad_norm": 0.21101562678813934, "learning_rate": 1.459422536059407e-07, "loss": 0.0421, "step": 97690 }, { "epoch": 3.3030190337739613, "grad_norm": 0.18440893292427063, "learning_rate": 1.4468257186589508e-07, "loss": 0.0366, "step": 97700 }, { "epoch": 3.3033571114642144, "grad_norm": 0.18790479004383087, "learning_rate": 1.4342834222840217e-07, "loss": 0.0442, "step": 97710 }, { "epoch": 3.3036951891544675, "grad_norm": 0.21208599209785461, "learning_rate": 1.4217956483063566e-07, "loss": 0.047, "step": 97720 }, { "epoch": 3.304033266844721, "grad_norm": 1.5365554094314575, "learning_rate": 1.409362398091474e-07, "loss": 0.0423, "step": 97730 }, { "epoch": 3.304371344534974, "grad_norm": 0.17447209358215332, "learning_rate": 1.3969836729990638e-07, "loss": 0.0406, "step": 97740 }, { "epoch": 3.304709422225227, "grad_norm": 0.4216531217098236, "learning_rate": 1.384659474382932e-07, "loss": 0.0433, "step": 97750 }, { "epoch": 3.3050474999154806, "grad_norm": 0.6049134731292725, "learning_rate": 1.372389803590779e-07, "loss": 0.0425, "step": 97760 }, { "epoch": 3.3053855776057337, "grad_norm": 0.3449890911579132, "learning_rate": 1.360174661964364e-07, "loss": 0.0423, "step": 97770 }, { "epoch": 3.3057236552959868, "grad_norm": 0.21330519020557404, "learning_rate": 1.3480140508396188e-07, "loss": 0.042, "step": 97780 }, { "epoch": 3.3060617329862403, "grad_norm": 0.1735474169254303, "learning_rate": 1.335907971546313e-07, "loss": 0.0436, "step": 97790 }, { "epoch": 3.3063998106764934, "grad_norm": 0.18385198712348938, "learning_rate": 1.3238564254083875e-07, "loss": 0.0427, "step": 97800 }, { "epoch": 3.306737888366747, "grad_norm": 0.29818540811538696, "learning_rate": 1.3118594137437323e-07, "loss": 0.0436, "step": 97810 }, { "epoch": 3.307075966057, "grad_norm": 0.26083338260650635, "learning_rate": 1.2999169378644094e-07, "loss": 0.0408, "step": 97820 }, { "epoch": 3.307414043747253, "grad_norm": 0.25559282302856445, "learning_rate": 1.2880289990763738e-07, "loss": 0.0403, "step": 97830 }, { "epoch": 3.3077521214375065, "grad_norm": 0.5359532237052917, "learning_rate": 1.2761955986796968e-07, "loss": 0.0413, "step": 97840 }, { "epoch": 3.3080901991277596, "grad_norm": 0.20310577750205994, "learning_rate": 1.264416737968399e-07, "loss": 0.0407, "step": 97850 }, { "epoch": 3.3084282768180127, "grad_norm": 0.25445231795310974, "learning_rate": 1.2526924182307275e-07, "loss": 0.0414, "step": 97860 }, { "epoch": 3.308766354508266, "grad_norm": 0.24108374118804932, "learning_rate": 1.2410226407487124e-07, "loss": 0.0417, "step": 97870 }, { "epoch": 3.3091044321985192, "grad_norm": 0.16288506984710693, "learning_rate": 1.229407406798555e-07, "loss": 0.0404, "step": 97880 }, { "epoch": 3.3094425098887723, "grad_norm": 0.2651297450065613, "learning_rate": 1.2178467176505726e-07, "loss": 0.0466, "step": 97890 }, { "epoch": 3.309780587579026, "grad_norm": 0.28438058495521545, "learning_rate": 1.2063405745689205e-07, "loss": 0.041, "step": 97900 }, { "epoch": 3.310118665269279, "grad_norm": 0.23235487937927246, "learning_rate": 1.1948889788119812e-07, "loss": 0.0386, "step": 97910 }, { "epoch": 3.310456742959532, "grad_norm": 0.34357982873916626, "learning_rate": 1.1834919316320303e-07, "loss": 0.0468, "step": 97920 }, { "epoch": 3.3107948206497855, "grad_norm": 0.3048805296421051, "learning_rate": 1.1721494342754048e-07, "loss": 0.0403, "step": 97930 }, { "epoch": 3.3111328983400385, "grad_norm": 0.2155790776014328, "learning_rate": 1.1608614879825563e-07, "loss": 0.0418, "step": 97940 }, { "epoch": 3.3114709760302916, "grad_norm": 0.36306852102279663, "learning_rate": 1.1496280939879422e-07, "loss": 0.039, "step": 97950 }, { "epoch": 3.311809053720545, "grad_norm": 0.5925774574279785, "learning_rate": 1.1384492535199687e-07, "loss": 0.0475, "step": 97960 }, { "epoch": 3.312147131410798, "grad_norm": 0.2480255365371704, "learning_rate": 1.1273249678011578e-07, "loss": 0.0467, "step": 97970 }, { "epoch": 3.3124852091010513, "grad_norm": 0.13664653897285461, "learning_rate": 1.1162552380480362e-07, "loss": 0.0428, "step": 97980 }, { "epoch": 3.3128232867913048, "grad_norm": 0.16984546184539795, "learning_rate": 1.1052400654711359e-07, "loss": 0.0376, "step": 97990 }, { "epoch": 3.313161364481558, "grad_norm": 0.14983250200748444, "learning_rate": 1.0942794512751597e-07, "loss": 0.039, "step": 98000 }, { "epoch": 3.313499442171811, "grad_norm": 0.15123915672302246, "learning_rate": 1.0833733966587045e-07, "loss": 0.0439, "step": 98010 }, { "epoch": 3.3138375198620644, "grad_norm": 0.28280243277549744, "learning_rate": 1.0725219028143718e-07, "loss": 0.0426, "step": 98020 }, { "epoch": 3.3141755975523175, "grad_norm": 0.22436825931072235, "learning_rate": 1.0617249709289345e-07, "loss": 0.0388, "step": 98030 }, { "epoch": 3.3145136752425706, "grad_norm": 0.2615970969200134, "learning_rate": 1.0509826021831703e-07, "loss": 0.0416, "step": 98040 }, { "epoch": 3.314851752932824, "grad_norm": 0.18875941634178162, "learning_rate": 1.0402947977517508e-07, "loss": 0.0423, "step": 98050 }, { "epoch": 3.315189830623077, "grad_norm": 0.28564101457595825, "learning_rate": 1.0296615588035185e-07, "loss": 0.0364, "step": 98060 }, { "epoch": 3.31552790831333, "grad_norm": 0.2766350507736206, "learning_rate": 1.0190828865012659e-07, "loss": 0.0413, "step": 98070 }, { "epoch": 3.3158659860035837, "grad_norm": 0.19718915224075317, "learning_rate": 1.008558782001956e-07, "loss": 0.0395, "step": 98080 }, { "epoch": 3.316204063693837, "grad_norm": 0.16263490915298462, "learning_rate": 9.980892464563906e-08, "loss": 0.0442, "step": 98090 }, { "epoch": 3.31654214138409, "grad_norm": 0.17078205943107605, "learning_rate": 9.876742810095985e-08, "loss": 0.0385, "step": 98100 }, { "epoch": 3.3168802190743434, "grad_norm": 0.19950473308563232, "learning_rate": 9.77313886800446e-08, "loss": 0.044, "step": 98110 }, { "epoch": 3.3172182967645965, "grad_norm": 0.3016728162765503, "learning_rate": 9.670080649619717e-08, "loss": 0.0382, "step": 98120 }, { "epoch": 3.31755637445485, "grad_norm": 0.23417922854423523, "learning_rate": 9.567568166212737e-08, "loss": 0.0415, "step": 98130 }, { "epoch": 3.317894452145103, "grad_norm": 0.2629740238189697, "learning_rate": 9.465601428992888e-08, "loss": 0.0427, "step": 98140 }, { "epoch": 3.318232529835356, "grad_norm": 0.16760754585266113, "learning_rate": 9.364180449111803e-08, "loss": 0.039, "step": 98150 }, { "epoch": 3.3185706075256096, "grad_norm": 0.2857745289802551, "learning_rate": 9.263305237661169e-08, "loss": 0.0401, "step": 98160 }, { "epoch": 3.3189086852158627, "grad_norm": 0.21361024677753448, "learning_rate": 9.162975805671603e-08, "loss": 0.0364, "step": 98170 }, { "epoch": 3.3192467629061158, "grad_norm": 0.22175447642803192, "learning_rate": 9.063192164115442e-08, "loss": 0.0381, "step": 98180 }, { "epoch": 3.3195848405963693, "grad_norm": 0.21913379430770874, "learning_rate": 8.963954323904511e-08, "loss": 0.0426, "step": 98190 }, { "epoch": 3.3199229182866223, "grad_norm": 0.26530566811561584, "learning_rate": 8.865262295891796e-08, "loss": 0.0446, "step": 98200 }, { "epoch": 3.3202609959768754, "grad_norm": 0.27307960391044617, "learning_rate": 8.767116090870331e-08, "loss": 0.0454, "step": 98210 }, { "epoch": 3.320599073667129, "grad_norm": 0.26780471205711365, "learning_rate": 8.669515719572086e-08, "loss": 0.04, "step": 98220 }, { "epoch": 3.320937151357382, "grad_norm": 0.15274684131145477, "learning_rate": 8.572461192671855e-08, "loss": 0.0369, "step": 98230 }, { "epoch": 3.321275229047635, "grad_norm": 0.1647845059633255, "learning_rate": 8.475952520782815e-08, "loss": 0.0409, "step": 98240 }, { "epoch": 3.3216133067378886, "grad_norm": 0.27632394433021545, "learning_rate": 8.379989714458747e-08, "loss": 0.0456, "step": 98250 }, { "epoch": 3.3219513844281416, "grad_norm": 0.21102985739707947, "learning_rate": 8.284572784194034e-08, "loss": 0.0413, "step": 98260 }, { "epoch": 3.3222894621183947, "grad_norm": 0.16255858540534973, "learning_rate": 8.189701740424216e-08, "loss": 0.0371, "step": 98270 }, { "epoch": 3.3226275398086482, "grad_norm": 0.1539722979068756, "learning_rate": 8.095376593522663e-08, "loss": 0.0386, "step": 98280 }, { "epoch": 3.3229656174989013, "grad_norm": 0.22139713168144226, "learning_rate": 8.001597353806123e-08, "loss": 0.0437, "step": 98290 }, { "epoch": 3.3233036951891544, "grad_norm": 0.30235451459884644, "learning_rate": 7.90836403152917e-08, "loss": 0.0417, "step": 98300 }, { "epoch": 3.323641772879408, "grad_norm": 0.19774627685546875, "learning_rate": 7.815676636888092e-08, "loss": 0.0381, "step": 98310 }, { "epoch": 3.323979850569661, "grad_norm": 0.2401706576347351, "learning_rate": 7.723535180019226e-08, "loss": 0.0418, "step": 98320 }, { "epoch": 3.324317928259914, "grad_norm": 0.20165999233722687, "learning_rate": 7.6319396709984e-08, "loss": 0.0378, "step": 98330 }, { "epoch": 3.3246560059501675, "grad_norm": 0.3263760209083557, "learning_rate": 7.540890119843158e-08, "loss": 0.0421, "step": 98340 }, { "epoch": 3.3249940836404206, "grad_norm": 0.3238777816295624, "learning_rate": 7.450386536509979e-08, "loss": 0.0392, "step": 98350 }, { "epoch": 3.3253321613306737, "grad_norm": 0.1709139347076416, "learning_rate": 7.360428930895947e-08, "loss": 0.0401, "step": 98360 }, { "epoch": 3.325670239020927, "grad_norm": 0.8654248714447021, "learning_rate": 7.271017312839302e-08, "loss": 0.0352, "step": 98370 }, { "epoch": 3.3260083167111802, "grad_norm": 0.3661455512046814, "learning_rate": 7.182151692118333e-08, "loss": 0.0456, "step": 98380 }, { "epoch": 3.3263463944014333, "grad_norm": 0.3184182345867157, "learning_rate": 7.093832078449713e-08, "loss": 0.0448, "step": 98390 }, { "epoch": 3.326684472091687, "grad_norm": 0.23109985888004303, "learning_rate": 7.006058481493493e-08, "loss": 0.04, "step": 98400 }, { "epoch": 3.32702254978194, "grad_norm": 0.20492301881313324, "learning_rate": 6.918830910847552e-08, "loss": 0.0384, "step": 98410 }, { "epoch": 3.327360627472193, "grad_norm": 0.2273569107055664, "learning_rate": 6.832149376052033e-08, "loss": 0.0422, "step": 98420 }, { "epoch": 3.3276987051624465, "grad_norm": 0.2396049052476883, "learning_rate": 6.746013886584912e-08, "loss": 0.0418, "step": 98430 }, { "epoch": 3.3280367828526995, "grad_norm": 0.47308284044265747, "learning_rate": 6.660424451866431e-08, "loss": 0.0437, "step": 98440 }, { "epoch": 3.3283748605429526, "grad_norm": 0.20819833874702454, "learning_rate": 6.575381081256327e-08, "loss": 0.0422, "step": 98450 }, { "epoch": 3.328712938233206, "grad_norm": 0.22464264929294586, "learning_rate": 6.49088378405549e-08, "loss": 0.0412, "step": 98460 }, { "epoch": 3.329051015923459, "grad_norm": 0.15961776673793793, "learning_rate": 6.406932569503755e-08, "loss": 0.0408, "step": 98470 }, { "epoch": 3.3293890936137123, "grad_norm": 0.2147216945886612, "learning_rate": 6.323527446782662e-08, "loss": 0.0411, "step": 98480 }, { "epoch": 3.329727171303966, "grad_norm": 0.5311933159828186, "learning_rate": 6.240668425012142e-08, "loss": 0.0392, "step": 98490 }, { "epoch": 3.330065248994219, "grad_norm": 0.21968168020248413, "learning_rate": 6.158355513254388e-08, "loss": 0.0439, "step": 98500 }, { "epoch": 3.330403326684472, "grad_norm": 0.19364410638809204, "learning_rate": 6.076588720510534e-08, "loss": 0.0405, "step": 98510 }, { "epoch": 3.3307414043747254, "grad_norm": 0.25919318199157715, "learning_rate": 5.995368055722316e-08, "loss": 0.0382, "step": 98520 }, { "epoch": 3.3310794820649785, "grad_norm": 0.12048214673995972, "learning_rate": 5.914693527773185e-08, "loss": 0.0351, "step": 98530 }, { "epoch": 3.3314175597552316, "grad_norm": 0.2399102747440338, "learning_rate": 5.834565145483861e-08, "loss": 0.0415, "step": 98540 }, { "epoch": 3.331755637445485, "grad_norm": 0.18827120959758759, "learning_rate": 5.754982917618446e-08, "loss": 0.0423, "step": 98550 }, { "epoch": 3.332093715135738, "grad_norm": 1.1134757995605469, "learning_rate": 5.675946852878866e-08, "loss": 0.042, "step": 98560 }, { "epoch": 3.3324317928259912, "grad_norm": 0.1604790836572647, "learning_rate": 5.59745695990932e-08, "loss": 0.0391, "step": 98570 }, { "epoch": 3.3327698705162447, "grad_norm": 0.21257391571998596, "learning_rate": 5.51951324729294e-08, "loss": 0.0438, "step": 98580 }, { "epoch": 3.333107948206498, "grad_norm": 0.32162386178970337, "learning_rate": 5.4421157235529093e-08, "loss": 0.0454, "step": 98590 }, { "epoch": 3.333446025896751, "grad_norm": 1.503584861755371, "learning_rate": 5.365264397154679e-08, "loss": 0.0445, "step": 98600 }, { "epoch": 3.3337841035870044, "grad_norm": 0.29034703969955444, "learning_rate": 5.288959276501526e-08, "loss": 0.0416, "step": 98610 }, { "epoch": 3.3341221812772575, "grad_norm": 0.14853627979755402, "learning_rate": 5.2132003699378876e-08, "loss": 0.0338, "step": 98620 }, { "epoch": 3.3344602589675105, "grad_norm": 0.24639351665973663, "learning_rate": 5.137987685749912e-08, "loss": 0.0365, "step": 98630 }, { "epoch": 3.334798336657764, "grad_norm": 0.2235582172870636, "learning_rate": 5.0633212321610227e-08, "loss": 0.0391, "step": 98640 }, { "epoch": 3.335136414348017, "grad_norm": 0.39789721369743347, "learning_rate": 4.989201017338019e-08, "loss": 0.0376, "step": 98650 }, { "epoch": 3.33547449203827, "grad_norm": 0.20975591242313385, "learning_rate": 4.9156270493860846e-08, "loss": 0.0423, "step": 98660 }, { "epoch": 3.3358125697285237, "grad_norm": 0.2998920679092407, "learning_rate": 4.842599336351561e-08, "loss": 0.0445, "step": 98670 }, { "epoch": 3.3361506474187768, "grad_norm": 0.21519601345062256, "learning_rate": 4.7701178862197274e-08, "loss": 0.0406, "step": 98680 }, { "epoch": 3.33648872510903, "grad_norm": 0.3109104037284851, "learning_rate": 4.698182706918131e-08, "loss": 0.0398, "step": 98690 }, { "epoch": 3.3368268027992833, "grad_norm": 0.20117232203483582, "learning_rate": 4.6267938063121466e-08, "loss": 0.0401, "step": 98700 }, { "epoch": 3.3371648804895364, "grad_norm": 0.27069947123527527, "learning_rate": 4.555951192209973e-08, "loss": 0.0423, "step": 98710 }, { "epoch": 3.3375029581797895, "grad_norm": 0.13634967803955078, "learning_rate": 4.485654872358747e-08, "loss": 0.0395, "step": 98720 }, { "epoch": 3.337841035870043, "grad_norm": 0.25549283623695374, "learning_rate": 4.415904854445097e-08, "loss": 0.0399, "step": 98730 }, { "epoch": 3.338179113560296, "grad_norm": 0.14456778764724731, "learning_rate": 4.34670114609792e-08, "loss": 0.0409, "step": 98740 }, { "epoch": 3.338517191250549, "grad_norm": 0.20349518954753876, "learning_rate": 4.278043754884498e-08, "loss": 0.0458, "step": 98750 }, { "epoch": 3.3388552689408026, "grad_norm": 0.21127769351005554, "learning_rate": 4.209932688312712e-08, "loss": 0.0402, "step": 98760 }, { "epoch": 3.3391933466310557, "grad_norm": 0.4675067663192749, "learning_rate": 4.14236795383216e-08, "loss": 0.0436, "step": 98770 }, { "epoch": 3.3395314243213092, "grad_norm": 0.6374474167823792, "learning_rate": 4.075349558830821e-08, "loss": 0.0391, "step": 98780 }, { "epoch": 3.3398695020115623, "grad_norm": 0.31517723202705383, "learning_rate": 4.008877510638387e-08, "loss": 0.0404, "step": 98790 }, { "epoch": 3.3402075797018154, "grad_norm": 0.2806539237499237, "learning_rate": 3.942951816523488e-08, "loss": 0.044, "step": 98800 }, { "epoch": 3.340545657392069, "grad_norm": 0.20127399265766144, "learning_rate": 3.8775724836959125e-08, "loss": 0.0413, "step": 98810 }, { "epoch": 3.340883735082322, "grad_norm": 1.4760950803756714, "learning_rate": 3.812739519305497e-08, "loss": 0.0459, "step": 98820 }, { "epoch": 3.341221812772575, "grad_norm": 0.3444664776325226, "learning_rate": 3.748452930442126e-08, "loss": 0.0413, "step": 98830 }, { "epoch": 3.3415598904628285, "grad_norm": 0.3049556314945221, "learning_rate": 3.6847127241362855e-08, "loss": 0.0407, "step": 98840 }, { "epoch": 3.3418979681530816, "grad_norm": 0.17104266583919525, "learning_rate": 3.621518907358512e-08, "loss": 0.0397, "step": 98850 }, { "epoch": 3.3422360458433347, "grad_norm": 0.3245519995689392, "learning_rate": 3.558871487019388e-08, "loss": 0.0394, "step": 98860 }, { "epoch": 3.342574123533588, "grad_norm": 0.23495596647262573, "learning_rate": 3.4967704699701007e-08, "loss": 0.0491, "step": 98870 }, { "epoch": 3.3429122012238413, "grad_norm": 0.2358359843492508, "learning_rate": 3.435215863001884e-08, "loss": 0.0422, "step": 98880 }, { "epoch": 3.3432502789140943, "grad_norm": 0.26216137409210205, "learning_rate": 3.374207672846019e-08, "loss": 0.0446, "step": 98890 }, { "epoch": 3.343588356604348, "grad_norm": 0.2064887136220932, "learning_rate": 3.313745906174392e-08, "loss": 0.041, "step": 98900 }, { "epoch": 3.343926434294601, "grad_norm": 0.27483922243118286, "learning_rate": 3.253830569599492e-08, "loss": 0.0462, "step": 98910 }, { "epoch": 3.344264511984854, "grad_norm": 0.2731441259384155, "learning_rate": 3.1944616696727436e-08, "loss": 0.0376, "step": 98920 }, { "epoch": 3.3446025896751075, "grad_norm": 0.2831164002418518, "learning_rate": 3.135639212887287e-08, "loss": 0.0429, "step": 98930 }, { "epoch": 3.3449406673653606, "grad_norm": 0.20057302713394165, "learning_rate": 3.077363205675754e-08, "loss": 0.0375, "step": 98940 }, { "epoch": 3.3452787450556136, "grad_norm": 0.31997695565223694, "learning_rate": 3.0196336544113804e-08, "loss": 0.044, "step": 98950 }, { "epoch": 3.345616822745867, "grad_norm": 0.523038387298584, "learning_rate": 2.9624505654063383e-08, "loss": 0.0406, "step": 98960 }, { "epoch": 3.34595490043612, "grad_norm": 0.18396933376789093, "learning_rate": 2.90581394491507e-08, "loss": 0.0436, "step": 98970 }, { "epoch": 3.3462929781263733, "grad_norm": 0.45666518807411194, "learning_rate": 2.8497237991309545e-08, "loss": 0.0405, "step": 98980 }, { "epoch": 3.346631055816627, "grad_norm": 0.345095694065094, "learning_rate": 2.7941801341879735e-08, "loss": 0.043, "step": 98990 }, { "epoch": 3.34696913350688, "grad_norm": 0.1805475801229477, "learning_rate": 2.7391829561601578e-08, "loss": 0.0373, "step": 99000 }, { "epoch": 3.347307211197133, "grad_norm": 0.20785008370876312, "learning_rate": 2.6847322710621403e-08, "loss": 0.0395, "step": 99010 }, { "epoch": 3.3476452888873864, "grad_norm": 0.1766260266304016, "learning_rate": 2.6308280848486022e-08, "loss": 0.0434, "step": 99020 }, { "epoch": 3.3479833665776395, "grad_norm": 0.24600274860858917, "learning_rate": 2.5774704034137177e-08, "loss": 0.0382, "step": 99030 }, { "epoch": 3.3483214442678926, "grad_norm": 0.22389322519302368, "learning_rate": 2.524659232593929e-08, "loss": 0.0382, "step": 99040 }, { "epoch": 3.348659521958146, "grad_norm": 0.3047108054161072, "learning_rate": 2.472394578163506e-08, "loss": 0.0376, "step": 99050 }, { "epoch": 3.348997599648399, "grad_norm": 0.39854881167411804, "learning_rate": 2.4206764458378772e-08, "loss": 0.0432, "step": 99060 }, { "epoch": 3.3493356773386522, "grad_norm": 0.2625667452812195, "learning_rate": 2.3695048412736285e-08, "loss": 0.0386, "step": 99070 }, { "epoch": 3.3496737550289057, "grad_norm": 0.2765200138092041, "learning_rate": 2.3188797700668395e-08, "loss": 0.0472, "step": 99080 }, { "epoch": 3.350011832719159, "grad_norm": 0.21011431515216827, "learning_rate": 2.268801237753082e-08, "loss": 0.043, "step": 99090 }, { "epoch": 3.3503499104094123, "grad_norm": 0.2854979336261749, "learning_rate": 2.2192692498090862e-08, "loss": 0.0444, "step": 99100 }, { "epoch": 3.3506879880996654, "grad_norm": 0.22597692906856537, "learning_rate": 2.170283811652185e-08, "loss": 0.0403, "step": 99110 }, { "epoch": 3.3510260657899185, "grad_norm": 0.3860304057598114, "learning_rate": 2.12184492863865e-08, "loss": 0.0473, "step": 99120 }, { "epoch": 3.351364143480172, "grad_norm": 0.7499144077301025, "learning_rate": 2.073952606066465e-08, "loss": 0.0432, "step": 99130 }, { "epoch": 3.351702221170425, "grad_norm": 0.4750044643878937, "learning_rate": 2.026606849171997e-08, "loss": 0.0378, "step": 99140 }, { "epoch": 3.352040298860678, "grad_norm": 0.1429854929447174, "learning_rate": 1.9798076631333262e-08, "loss": 0.0433, "step": 99150 }, { "epoch": 3.3523783765509316, "grad_norm": 0.2462652623653412, "learning_rate": 1.933555053069136e-08, "loss": 0.0436, "step": 99160 }, { "epoch": 3.3527164542411847, "grad_norm": 0.19517497718334198, "learning_rate": 1.887849024036492e-08, "loss": 0.0396, "step": 99170 }, { "epoch": 3.3530545319314378, "grad_norm": 0.127557635307312, "learning_rate": 1.8426895810341736e-08, "loss": 0.0406, "step": 99180 }, { "epoch": 3.3533926096216913, "grad_norm": 0.23577328026294708, "learning_rate": 1.798076729000453e-08, "loss": 0.0343, "step": 99190 }, { "epoch": 3.3537306873119443, "grad_norm": 0.2696424424648285, "learning_rate": 1.75401047281476e-08, "loss": 0.0392, "step": 99200 }, { "epoch": 3.3540687650021974, "grad_norm": 0.1552143543958664, "learning_rate": 1.7104908172954626e-08, "loss": 0.0387, "step": 99210 }, { "epoch": 3.354406842692451, "grad_norm": 0.4034787714481354, "learning_rate": 1.6675177672020868e-08, "loss": 0.0369, "step": 99220 }, { "epoch": 3.354744920382704, "grad_norm": 0.25577983260154724, "learning_rate": 1.6250913272342073e-08, "loss": 0.0442, "step": 99230 }, { "epoch": 3.355082998072957, "grad_norm": 0.23611444234848022, "learning_rate": 1.5832115020314453e-08, "loss": 0.0395, "step": 99240 }, { "epoch": 3.3554210757632106, "grad_norm": 0.4700203537940979, "learning_rate": 1.5418782961734712e-08, "loss": 0.0444, "step": 99250 }, { "epoch": 3.3557591534534637, "grad_norm": 0.32994842529296875, "learning_rate": 1.501091714181113e-08, "loss": 0.0436, "step": 99260 }, { "epoch": 3.3560972311437167, "grad_norm": 0.3017423748970032, "learning_rate": 1.460851760513582e-08, "loss": 0.042, "step": 99270 }, { "epoch": 3.3564353088339702, "grad_norm": 0.22968697547912598, "learning_rate": 1.4211584395723566e-08, "loss": 0.041, "step": 99280 }, { "epoch": 3.3567733865242233, "grad_norm": 0.21122430264949799, "learning_rate": 1.382011755697854e-08, "loss": 0.0411, "step": 99290 }, { "epoch": 3.3571114642144764, "grad_norm": 0.26217320561408997, "learning_rate": 1.3434117131716494e-08, "loss": 0.0398, "step": 99300 }, { "epoch": 3.35744954190473, "grad_norm": 0.3194802701473236, "learning_rate": 1.3053583162142557e-08, "loss": 0.0386, "step": 99310 }, { "epoch": 3.357787619594983, "grad_norm": 0.3365517258644104, "learning_rate": 1.2678515689873438e-08, "loss": 0.0405, "step": 99320 }, { "epoch": 3.358125697285236, "grad_norm": 0.18789556622505188, "learning_rate": 1.2308914755931877e-08, "loss": 0.0407, "step": 99330 }, { "epoch": 3.3584637749754895, "grad_norm": 0.30112189054489136, "learning_rate": 1.1944780400729993e-08, "loss": 0.0397, "step": 99340 }, { "epoch": 3.3588018526657426, "grad_norm": 0.42892375588417053, "learning_rate": 1.1586112664085935e-08, "loss": 0.0429, "step": 99350 }, { "epoch": 3.3591399303559957, "grad_norm": 0.39894619584083557, "learning_rate": 1.1232911585234984e-08, "loss": 0.0456, "step": 99360 }, { "epoch": 3.359478008046249, "grad_norm": 0.25051388144493103, "learning_rate": 1.0885177202790697e-08, "loss": 0.041, "step": 99370 }, { "epoch": 3.3598160857365023, "grad_norm": 0.2580036222934723, "learning_rate": 1.0542909554789315e-08, "loss": 0.0437, "step": 99380 }, { "epoch": 3.3601541634267553, "grad_norm": 0.17460711300373077, "learning_rate": 1.0206108678656456e-08, "loss": 0.0408, "step": 99390 }, { "epoch": 3.360492241117009, "grad_norm": 0.31896865367889404, "learning_rate": 9.874774611223769e-09, "loss": 0.0389, "step": 99400 }, { "epoch": 3.360830318807262, "grad_norm": 0.18836773931980133, "learning_rate": 9.548907388728933e-09, "loss": 0.0426, "step": 99410 }, { "epoch": 3.361168396497515, "grad_norm": 0.2632828652858734, "learning_rate": 9.228507046804557e-09, "loss": 0.0401, "step": 99420 }, { "epoch": 3.3615064741877685, "grad_norm": 0.32160133123397827, "learning_rate": 8.91357362048928e-09, "loss": 0.042, "step": 99430 }, { "epoch": 3.3618445518780216, "grad_norm": 0.19986405968666077, "learning_rate": 8.604107144227769e-09, "loss": 0.0416, "step": 99440 }, { "epoch": 3.3621826295682746, "grad_norm": 0.2663193643093109, "learning_rate": 8.300107651859623e-09, "loss": 0.0447, "step": 99450 }, { "epoch": 3.362520707258528, "grad_norm": 0.4685139060020447, "learning_rate": 8.001575176630472e-09, "loss": 0.0431, "step": 99460 }, { "epoch": 3.362858784948781, "grad_norm": 0.408263236284256, "learning_rate": 7.708509751186422e-09, "loss": 0.0445, "step": 99470 }, { "epoch": 3.3631968626390343, "grad_norm": 0.4746311902999878, "learning_rate": 7.420911407579611e-09, "loss": 0.0405, "step": 99480 }, { "epoch": 3.363534940329288, "grad_norm": 0.3459494113922119, "learning_rate": 7.13878017725711e-09, "loss": 0.0463, "step": 99490 }, { "epoch": 3.363873018019541, "grad_norm": 0.18509609997272491, "learning_rate": 6.8621160910720125e-09, "loss": 0.0444, "step": 99500 }, { "epoch": 3.364211095709794, "grad_norm": 0.21404753625392914, "learning_rate": 6.59091917928345e-09, "loss": 0.0416, "step": 99510 }, { "epoch": 3.3645491734000474, "grad_norm": 0.25252169370651245, "learning_rate": 6.32518947155103e-09, "loss": 0.0385, "step": 99520 }, { "epoch": 3.3648872510903005, "grad_norm": 0.2536228895187378, "learning_rate": 6.064926996929288e-09, "loss": 0.0449, "step": 99530 }, { "epoch": 3.3652253287805536, "grad_norm": 0.19614069163799286, "learning_rate": 5.810131783884343e-09, "loss": 0.0406, "step": 99540 }, { "epoch": 3.365563406470807, "grad_norm": 0.5386055111885071, "learning_rate": 5.5608038602772415e-09, "loss": 0.0403, "step": 99550 }, { "epoch": 3.36590148416106, "grad_norm": 0.2584041655063629, "learning_rate": 5.316943253375062e-09, "loss": 0.0412, "step": 99560 }, { "epoch": 3.3662395618513132, "grad_norm": 0.18483678996562958, "learning_rate": 5.07854998984536e-09, "loss": 0.0406, "step": 99570 }, { "epoch": 3.3665776395415667, "grad_norm": 0.5434134602546692, "learning_rate": 4.845624095756174e-09, "loss": 0.0414, "step": 99580 }, { "epoch": 3.36691571723182, "grad_norm": 0.16514363884925842, "learning_rate": 4.618165596587121e-09, "loss": 0.0433, "step": 99590 }, { "epoch": 3.367253794922073, "grad_norm": 0.28719770908355713, "learning_rate": 4.396174517207197e-09, "loss": 0.0453, "step": 99600 }, { "epoch": 3.3675918726123264, "grad_norm": 0.29647570848464966, "learning_rate": 4.179650881896979e-09, "loss": 0.042, "step": 99610 }, { "epoch": 3.3679299503025795, "grad_norm": 0.20997463166713715, "learning_rate": 3.968594714331975e-09, "loss": 0.0493, "step": 99620 }, { "epoch": 3.3682680279928325, "grad_norm": 0.18570753931999207, "learning_rate": 3.763006037593719e-09, "loss": 0.0424, "step": 99630 }, { "epoch": 3.368606105683086, "grad_norm": 0.13745903968811035, "learning_rate": 3.5628848741586786e-09, "loss": 0.0442, "step": 99640 }, { "epoch": 3.368944183373339, "grad_norm": 0.2317495197057724, "learning_rate": 3.368231245926001e-09, "loss": 0.0385, "step": 99650 }, { "epoch": 3.369282261063592, "grad_norm": 0.24949465692043304, "learning_rate": 3.179045174167561e-09, "loss": 0.0397, "step": 99660 }, { "epoch": 3.3696203387538457, "grad_norm": 0.29749155044555664, "learning_rate": 2.9953266795834657e-09, "loss": 0.0401, "step": 99670 }, { "epoch": 3.3699584164440988, "grad_norm": 0.46071764826774597, "learning_rate": 2.817075782263201e-09, "loss": 0.0358, "step": 99680 }, { "epoch": 3.370296494134352, "grad_norm": 0.19312527775764465, "learning_rate": 2.6442925016967325e-09, "loss": 0.0422, "step": 99690 }, { "epoch": 3.3706345718246054, "grad_norm": 0.23380789160728455, "learning_rate": 2.4769768567800555e-09, "loss": 0.0405, "step": 99700 }, { "epoch": 3.3709726495148584, "grad_norm": 0.2635670304298401, "learning_rate": 2.315128865809646e-09, "loss": 0.04, "step": 99710 }, { "epoch": 3.3713107272051115, "grad_norm": 0.6335452198982239, "learning_rate": 2.1587485464880097e-09, "loss": 0.0438, "step": 99720 }, { "epoch": 3.371648804895365, "grad_norm": 0.30510255694389343, "learning_rate": 2.0078359159125813e-09, "loss": 0.0433, "step": 99730 }, { "epoch": 3.371986882585618, "grad_norm": 0.20325401425361633, "learning_rate": 1.8623909905923776e-09, "loss": 0.0351, "step": 99740 }, { "epoch": 3.3723249602758716, "grad_norm": 0.27566784620285034, "learning_rate": 1.7224137864257916e-09, "loss": 0.0396, "step": 99750 }, { "epoch": 3.3726630379661247, "grad_norm": 0.2742348611354828, "learning_rate": 1.5879043187283505e-09, "loss": 0.0376, "step": 99760 }, { "epoch": 3.3730011156563777, "grad_norm": 0.3691704273223877, "learning_rate": 1.458862602204958e-09, "loss": 0.0432, "step": 99770 }, { "epoch": 3.3733391933466312, "grad_norm": 0.46940940618515015, "learning_rate": 1.3352886509720996e-09, "loss": 0.0407, "step": 99780 }, { "epoch": 3.3736772710368843, "grad_norm": 0.29838109016418457, "learning_rate": 1.2171824785356389e-09, "loss": 0.0425, "step": 99790 }, { "epoch": 3.3740153487271374, "grad_norm": 0.16682776808738708, "learning_rate": 1.1045440978185718e-09, "loss": 0.0438, "step": 99800 }, { "epoch": 3.374353426417391, "grad_norm": 0.2634158134460449, "learning_rate": 9.973735211332714e-10, "loss": 0.042, "step": 99810 }, { "epoch": 3.374691504107644, "grad_norm": 0.26947787404060364, "learning_rate": 8.956707602036929e-10, "loss": 0.0436, "step": 99820 }, { "epoch": 3.375029581797897, "grad_norm": 0.2359260767698288, "learning_rate": 7.994358261542712e-10, "loss": 0.0393, "step": 99830 }, { "epoch": 3.3753676594881505, "grad_norm": 0.3374170660972595, "learning_rate": 7.086687295043692e-10, "loss": 0.0424, "step": 99840 }, { "epoch": 3.3757057371784036, "grad_norm": 0.2119787335395813, "learning_rate": 6.23369480179381e-10, "loss": 0.0407, "step": 99850 }, { "epoch": 3.3760438148686567, "grad_norm": 0.2935253083705902, "learning_rate": 5.43538087510731e-10, "loss": 0.0419, "step": 99860 }, { "epoch": 3.37638189255891, "grad_norm": 0.21375620365142822, "learning_rate": 4.691745602303233e-10, "loss": 0.0375, "step": 99870 }, { "epoch": 3.3767199702491633, "grad_norm": 0.13682244718074799, "learning_rate": 4.0027890646499034e-10, "loss": 0.045, "step": 99880 }, { "epoch": 3.3770580479394163, "grad_norm": 0.36278507113456726, "learning_rate": 3.368511337531466e-10, "loss": 0.0473, "step": 99890 }, { "epoch": 3.37739612562967, "grad_norm": 0.2559727132320404, "learning_rate": 2.788912490281348e-10, "loss": 0.0437, "step": 99900 }, { "epoch": 3.377734203319923, "grad_norm": 0.30648308992385864, "learning_rate": 2.2639925863487954e-10, "loss": 0.0371, "step": 99910 }, { "epoch": 3.378072281010176, "grad_norm": 0.17437951266765594, "learning_rate": 1.7937516830213163e-10, "loss": 0.0399, "step": 99920 }, { "epoch": 3.3784103587004295, "grad_norm": 0.16232313215732574, "learning_rate": 1.3781898318687702e-10, "loss": 0.0422, "step": 99930 }, { "epoch": 3.3787484363906826, "grad_norm": 0.20651674270629883, "learning_rate": 1.0173070781882566e-10, "loss": 0.0407, "step": 99940 }, { "epoch": 3.3790865140809356, "grad_norm": 0.1360999047756195, "learning_rate": 7.111034615592261e-11, "loss": 0.0429, "step": 99950 }, { "epoch": 3.379424591771189, "grad_norm": 0.371756911277771, "learning_rate": 4.595790153993918e-11, "loss": 0.0395, "step": 99960 }, { "epoch": 3.379762669461442, "grad_norm": 0.18135470151901245, "learning_rate": 2.6273376729779585e-11, "loss": 0.0365, "step": 99970 }, { "epoch": 3.3801007471516953, "grad_norm": 0.19698543846607208, "learning_rate": 1.2056773868174276e-11, "loss": 0.0368, "step": 99980 }, { "epoch": 3.380438824841949, "grad_norm": 0.31813400983810425, "learning_rate": 3.3080945149865926e-12, "loss": 0.045, "step": 99990 }, { "epoch": 3.380776902532202, "grad_norm": 0.3422926962375641, "learning_rate": 2.7339625008337977e-14, "loss": 0.0428, "step": 100000 }, { "epoch": 3.380776902532202, "step": 100000, "total_flos": 0.0, "train_loss": 0.05353647198677063, "train_runtime": 41174.1401, "train_samples_per_second": 155.437, "train_steps_per_second": 2.429 } ], "logging_steps": 10, "max_steps": 100000, "num_input_tokens_seen": 0, "num_train_epochs": 4, "save_steps": 20000, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": true }, "attributes": {} } }, "total_flos": 0.0, "train_batch_size": 64, "trial_name": null, "trial_params": null }