{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 0.6, "eval_steps": 500, "global_step": 60000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "grad_norm": 0.9511709809303284, "learning_rate": 1.8e-07, "loss": 1.4856, "step": 10 }, { "grad_norm": 0.8089783191680908, "learning_rate": 3.8e-07, "loss": 1.4899, "step": 20 }, { "grad_norm": 0.9533777832984924, "learning_rate": 5.8e-07, "loss": 1.4704, "step": 30 }, { "grad_norm": 0.8591621518135071, "learning_rate": 7.8e-07, "loss": 1.4639, "step": 40 }, { "grad_norm": 1.5313694477081299, "learning_rate": 9.8e-07, "loss": 1.4359, "step": 50 }, { "grad_norm": 15.992796897888184, "learning_rate": 1.18e-06, "loss": 1.3939, "step": 60 }, { "grad_norm": 1.8355778455734253, "learning_rate": 1.3800000000000001e-06, "loss": 1.3558, "step": 70 }, { "grad_norm": 1.1542338132858276, "learning_rate": 1.5800000000000003e-06, "loss": 1.3002, "step": 80 }, { "grad_norm": 1.9136810302734375, "learning_rate": 1.7800000000000001e-06, "loss": 1.3268, "step": 90 }, { "grad_norm": 1.8063944578170776, "learning_rate": 1.98e-06, "loss": 1.2526, "step": 100 }, { "grad_norm": 1.6720718145370483, "learning_rate": 2.1800000000000003e-06, "loss": 1.3091, "step": 110 }, { "grad_norm": 2.002459764480591, "learning_rate": 2.38e-06, "loss": 1.2685, "step": 120 }, { "grad_norm": 1.3225202560424805, "learning_rate": 2.5800000000000003e-06, "loss": 1.2972, "step": 130 }, { "grad_norm": 2.525108814239502, "learning_rate": 2.78e-06, "loss": 1.2536, "step": 140 }, { "grad_norm": 1.6202949285507202, "learning_rate": 2.9800000000000003e-06, "loss": 1.2321, "step": 150 }, { "grad_norm": 4.261291980743408, "learning_rate": 3.1800000000000005e-06, "loss": 1.2524, "step": 160 }, { "grad_norm": 2.624082326889038, "learning_rate": 3.38e-06, "loss": 1.2195, "step": 170 }, { "grad_norm": 2.420454263687134, "learning_rate": 3.58e-06, "loss": 1.2273, "step": 180 }, { "grad_norm": 1.9982647895812988, "learning_rate": 3.7800000000000002e-06, "loss": 1.1733, "step": 190 }, { "grad_norm": 2.0410892963409424, "learning_rate": 3.98e-06, "loss": 1.1522, "step": 200 }, { "grad_norm": 4.125710964202881, "learning_rate": 4.18e-06, "loss": 1.1193, "step": 210 }, { "grad_norm": 2.6023435592651367, "learning_rate": 4.38e-06, "loss": 1.1681, "step": 220 }, { "grad_norm": 1.5881658792495728, "learning_rate": 4.58e-06, "loss": 1.1234, "step": 230 }, { "grad_norm": 1.458034634590149, "learning_rate": 4.780000000000001e-06, "loss": 1.0935, "step": 240 }, { "grad_norm": 2.3354644775390625, "learning_rate": 4.98e-06, "loss": 1.1293, "step": 250 }, { "grad_norm": 1.812134027481079, "learning_rate": 5.18e-06, "loss": 1.1074, "step": 260 }, { "grad_norm": 2.79339599609375, "learning_rate": 5.38e-06, "loss": 1.1258, "step": 270 }, { "grad_norm": 2.508848190307617, "learning_rate": 5.580000000000001e-06, "loss": 1.0965, "step": 280 }, { "grad_norm": 2.007289409637451, "learning_rate": 5.78e-06, "loss": 1.0819, "step": 290 }, { "grad_norm": 2.6910762786865234, "learning_rate": 5.98e-06, "loss": 1.0869, "step": 300 }, { "grad_norm": 1.7632522583007812, "learning_rate": 6.18e-06, "loss": 1.1126, "step": 310 }, { "grad_norm": 1.239662766456604, "learning_rate": 6.38e-06, "loss": 1.0951, "step": 320 }, { "grad_norm": 1.7973089218139648, "learning_rate": 6.58e-06, "loss": 1.0973, "step": 330 }, { "grad_norm": 1.3363670110702515, "learning_rate": 6.78e-06, "loss": 1.0807, "step": 340 }, { "grad_norm": 1.7420942783355713, "learning_rate": 6.98e-06, "loss": 1.0916, "step": 350 }, { "grad_norm": 2.508195400238037, "learning_rate": 7.180000000000001e-06, "loss": 1.0859, "step": 360 }, { "grad_norm": 1.6706804037094116, "learning_rate": 7.3800000000000005e-06, "loss": 1.0859, "step": 370 }, { "grad_norm": 1.2607227563858032, "learning_rate": 7.580000000000001e-06, "loss": 1.0684, "step": 380 }, { "grad_norm": 1.8795640468597412, "learning_rate": 7.78e-06, "loss": 1.0746, "step": 390 }, { "grad_norm": 1.8169543743133545, "learning_rate": 7.98e-06, "loss": 1.0942, "step": 400 }, { "grad_norm": 1.9738904237747192, "learning_rate": 8.18e-06, "loss": 1.098, "step": 410 }, { "grad_norm": 1.4607094526290894, "learning_rate": 8.380000000000001e-06, "loss": 1.0717, "step": 420 }, { "grad_norm": 1.2731170654296875, "learning_rate": 8.580000000000001e-06, "loss": 1.0744, "step": 430 }, { "grad_norm": 10.590871810913086, "learning_rate": 8.78e-06, "loss": 1.1141, "step": 440 }, { "grad_norm": 1.5174068212509155, "learning_rate": 8.98e-06, "loss": 1.0636, "step": 450 }, { "grad_norm": 1.1776469945907593, "learning_rate": 9.180000000000002e-06, "loss": 1.085, "step": 460 }, { "grad_norm": 1.3851628303527832, "learning_rate": 9.38e-06, "loss": 1.0763, "step": 470 }, { "grad_norm": 1.3513115644454956, "learning_rate": 9.58e-06, "loss": 1.0924, "step": 480 }, { "grad_norm": 1.4795184135437012, "learning_rate": 9.78e-06, "loss": 1.0875, "step": 490 }, { "grad_norm": 1.5138522386550903, "learning_rate": 9.980000000000001e-06, "loss": 1.0818, "step": 500 }, { "grad_norm": 1.4966660737991333, "learning_rate": 1.018e-05, "loss": 1.0582, "step": 510 }, { "grad_norm": 1.4339085817337036, "learning_rate": 1.038e-05, "loss": 1.0631, "step": 520 }, { "grad_norm": 1.2118209600448608, "learning_rate": 1.058e-05, "loss": 1.0534, "step": 530 }, { "grad_norm": 1.400166392326355, "learning_rate": 1.0780000000000002e-05, "loss": 1.0559, "step": 540 }, { "grad_norm": 1.5643550157546997, "learning_rate": 1.098e-05, "loss": 1.0478, "step": 550 }, { "grad_norm": 1.5805484056472778, "learning_rate": 1.118e-05, "loss": 1.037, "step": 560 }, { "grad_norm": 1.5822973251342773, "learning_rate": 1.1380000000000001e-05, "loss": 1.0178, "step": 570 }, { "grad_norm": 1.847975730895996, "learning_rate": 1.1580000000000001e-05, "loss": 0.9919, "step": 580 }, { "grad_norm": 1.9303910732269287, "learning_rate": 1.178e-05, "loss": 0.991, "step": 590 }, { "grad_norm": 1.5539828538894653, "learning_rate": 1.198e-05, "loss": 0.9597, "step": 600 }, { "grad_norm": 1.7665153741836548, "learning_rate": 1.2180000000000002e-05, "loss": 0.9481, "step": 610 }, { "grad_norm": 2.671569347381592, "learning_rate": 1.238e-05, "loss": 0.9261, "step": 620 }, { "grad_norm": 1.980902910232544, "learning_rate": 1.258e-05, "loss": 0.8893, "step": 630 }, { "grad_norm": 2.0801076889038086, "learning_rate": 1.278e-05, "loss": 0.8715, "step": 640 }, { "grad_norm": 2.2111315727233887, "learning_rate": 1.2980000000000001e-05, "loss": 0.8331, "step": 650 }, { "grad_norm": 2.719674825668335, "learning_rate": 1.3180000000000001e-05, "loss": 0.8087, "step": 660 }, { "grad_norm": 1.9336645603179932, "learning_rate": 1.338e-05, "loss": 0.7743, "step": 670 }, { "grad_norm": 2.2799386978149414, "learning_rate": 1.358e-05, "loss": 0.7688, "step": 680 }, { "grad_norm": 2.006140947341919, "learning_rate": 1.3780000000000002e-05, "loss": 0.7601, "step": 690 }, { "grad_norm": 2.650465726852417, "learning_rate": 1.3980000000000002e-05, "loss": 0.7487, "step": 700 }, { "grad_norm": 2.215714931488037, "learning_rate": 1.4180000000000001e-05, "loss": 0.6858, "step": 710 }, { "grad_norm": 2.133212089538574, "learning_rate": 1.4380000000000001e-05, "loss": 0.6896, "step": 720 }, { "grad_norm": 2.333899974822998, "learning_rate": 1.4580000000000003e-05, "loss": 0.6501, "step": 730 }, { "grad_norm": 2.8487374782562256, "learning_rate": 1.4779999999999999e-05, "loss": 0.6036, "step": 740 }, { "grad_norm": 2.4035632610321045, "learning_rate": 1.4979999999999999e-05, "loss": 0.6071, "step": 750 }, { "grad_norm": 2.980544328689575, "learning_rate": 1.518e-05, "loss": 0.5846, "step": 760 }, { "grad_norm": 2.3575599193573, "learning_rate": 1.538e-05, "loss": 0.5512, "step": 770 }, { "grad_norm": 2.6987667083740234, "learning_rate": 1.558e-05, "loss": 0.5622, "step": 780 }, { "grad_norm": 2.2596445083618164, "learning_rate": 1.578e-05, "loss": 0.4873, "step": 790 }, { "grad_norm": 3.184044361114502, "learning_rate": 1.598e-05, "loss": 0.472, "step": 800 }, { "grad_norm": 3.210742712020874, "learning_rate": 1.618e-05, "loss": 0.4476, "step": 810 }, { "grad_norm": 2.9446568489074707, "learning_rate": 1.6380000000000002e-05, "loss": 0.4483, "step": 820 }, { "grad_norm": 2.974304676055908, "learning_rate": 1.658e-05, "loss": 0.4427, "step": 830 }, { "grad_norm": 3.075972318649292, "learning_rate": 1.6780000000000002e-05, "loss": 0.4282, "step": 840 }, { "grad_norm": 2.315190315246582, "learning_rate": 1.698e-05, "loss": 0.3878, "step": 850 }, { "grad_norm": 2.540806531906128, "learning_rate": 1.718e-05, "loss": 0.3575, "step": 860 }, { "grad_norm": 2.033179998397827, "learning_rate": 1.7380000000000003e-05, "loss": 0.3609, "step": 870 }, { "grad_norm": 2.78023362159729, "learning_rate": 1.758e-05, "loss": 0.354, "step": 880 }, { "grad_norm": 3.3053765296936035, "learning_rate": 1.7780000000000003e-05, "loss": 0.3356, "step": 890 }, { "grad_norm": 3.074331283569336, "learning_rate": 1.798e-05, "loss": 0.3347, "step": 900 }, { "grad_norm": 2.2514917850494385, "learning_rate": 1.818e-05, "loss": 0.2845, "step": 910 }, { "grad_norm": 2.1502065658569336, "learning_rate": 1.838e-05, "loss": 0.2479, "step": 920 }, { "grad_norm": 2.6326334476470947, "learning_rate": 1.858e-05, "loss": 0.2624, "step": 930 }, { "grad_norm": 3.5425188541412354, "learning_rate": 1.878e-05, "loss": 0.2486, "step": 940 }, { "grad_norm": 2.251235246658325, "learning_rate": 1.898e-05, "loss": 0.2153, "step": 950 }, { "grad_norm": 2.2244160175323486, "learning_rate": 1.918e-05, "loss": 0.2419, "step": 960 }, { "grad_norm": 2.968663215637207, "learning_rate": 1.938e-05, "loss": 0.2638, "step": 970 }, { "grad_norm": 2.237957239151001, "learning_rate": 1.9580000000000002e-05, "loss": 0.2151, "step": 980 }, { "grad_norm": 3.298628568649292, "learning_rate": 1.978e-05, "loss": 0.2062, "step": 990 }, { "grad_norm": 2.648430585861206, "learning_rate": 1.9980000000000002e-05, "loss": 0.2229, "step": 1000 }, { "grad_norm": 2.374605894088745, "learning_rate": 2.0180000000000003e-05, "loss": 0.2009, "step": 1010 }, { "grad_norm": 2.579284429550171, "learning_rate": 2.038e-05, "loss": 0.2372, "step": 1020 }, { "grad_norm": 1.9114998579025269, "learning_rate": 2.0580000000000003e-05, "loss": 0.1896, "step": 1030 }, { "grad_norm": 2.2677831649780273, "learning_rate": 2.078e-05, "loss": 0.1939, "step": 1040 }, { "grad_norm": 2.734956979751587, "learning_rate": 2.098e-05, "loss": 0.2103, "step": 1050 }, { "grad_norm": 3.556872606277466, "learning_rate": 2.118e-05, "loss": 0.2393, "step": 1060 }, { "grad_norm": 2.3170645236968994, "learning_rate": 2.138e-05, "loss": 0.2199, "step": 1070 }, { "grad_norm": 2.489729881286621, "learning_rate": 2.158e-05, "loss": 0.2209, "step": 1080 }, { "grad_norm": 2.237142562866211, "learning_rate": 2.178e-05, "loss": 0.2041, "step": 1090 }, { "grad_norm": 2.0124528408050537, "learning_rate": 2.198e-05, "loss": 0.1661, "step": 1100 }, { "grad_norm": 2.2334744930267334, "learning_rate": 2.218e-05, "loss": 0.2421, "step": 1110 }, { "grad_norm": 2.3322031497955322, "learning_rate": 2.2380000000000003e-05, "loss": 0.1611, "step": 1120 }, { "grad_norm": 2.4346137046813965, "learning_rate": 2.258e-05, "loss": 0.1874, "step": 1130 }, { "grad_norm": 2.681036949157715, "learning_rate": 2.2780000000000002e-05, "loss": 0.2318, "step": 1140 }, { "grad_norm": 3.046482801437378, "learning_rate": 2.298e-05, "loss": 0.2107, "step": 1150 }, { "grad_norm": 1.9949970245361328, "learning_rate": 2.318e-05, "loss": 0.1989, "step": 1160 }, { "grad_norm": 2.632946491241455, "learning_rate": 2.3380000000000003e-05, "loss": 0.1839, "step": 1170 }, { "grad_norm": 1.764267921447754, "learning_rate": 2.358e-05, "loss": 0.1592, "step": 1180 }, { "grad_norm": 2.853921413421631, "learning_rate": 2.3780000000000003e-05, "loss": 0.1717, "step": 1190 }, { "grad_norm": 3.136575937271118, "learning_rate": 2.398e-05, "loss": 0.1932, "step": 1200 }, { "grad_norm": 3.4540066719055176, "learning_rate": 2.418e-05, "loss": 0.2087, "step": 1210 }, { "grad_norm": 1.8468149900436401, "learning_rate": 2.438e-05, "loss": 0.178, "step": 1220 }, { "grad_norm": 2.328723669052124, "learning_rate": 2.4580000000000002e-05, "loss": 0.1841, "step": 1230 }, { "grad_norm": 1.592699408531189, "learning_rate": 2.478e-05, "loss": 0.1891, "step": 1240 }, { "grad_norm": 2.260723352432251, "learning_rate": 2.498e-05, "loss": 0.1602, "step": 1250 }, { "grad_norm": 2.6749157905578613, "learning_rate": 2.5180000000000003e-05, "loss": 0.1871, "step": 1260 }, { "grad_norm": 2.519312858581543, "learning_rate": 2.5380000000000004e-05, "loss": 0.1782, "step": 1270 }, { "grad_norm": 1.7435035705566406, "learning_rate": 2.5580000000000002e-05, "loss": 0.1943, "step": 1280 }, { "grad_norm": 2.0047168731689453, "learning_rate": 2.5779999999999997e-05, "loss": 0.1633, "step": 1290 }, { "grad_norm": 2.0540170669555664, "learning_rate": 2.598e-05, "loss": 0.1785, "step": 1300 }, { "grad_norm": 1.9522325992584229, "learning_rate": 2.618e-05, "loss": 0.1612, "step": 1310 }, { "grad_norm": 2.110589027404785, "learning_rate": 2.6379999999999998e-05, "loss": 0.1742, "step": 1320 }, { "grad_norm": 2.1972272396087646, "learning_rate": 2.658e-05, "loss": 0.1487, "step": 1330 }, { "grad_norm": 2.03448224067688, "learning_rate": 2.678e-05, "loss": 0.1492, "step": 1340 }, { "grad_norm": 1.914736270904541, "learning_rate": 2.698e-05, "loss": 0.1817, "step": 1350 }, { "grad_norm": 3.346343755722046, "learning_rate": 2.718e-05, "loss": 0.1893, "step": 1360 }, { "grad_norm": 2.4164679050445557, "learning_rate": 2.738e-05, "loss": 0.1504, "step": 1370 }, { "grad_norm": 2.1928634643554688, "learning_rate": 2.758e-05, "loss": 0.1565, "step": 1380 }, { "grad_norm": 2.0784053802490234, "learning_rate": 2.778e-05, "loss": 0.1665, "step": 1390 }, { "grad_norm": 1.8546178340911865, "learning_rate": 2.798e-05, "loss": 0.1506, "step": 1400 }, { "grad_norm": 1.532692551612854, "learning_rate": 2.818e-05, "loss": 0.159, "step": 1410 }, { "grad_norm": 2.003329277038574, "learning_rate": 2.8380000000000003e-05, "loss": 0.1617, "step": 1420 }, { "grad_norm": 1.9383817911148071, "learning_rate": 2.858e-05, "loss": 0.1383, "step": 1430 }, { "grad_norm": 2.0291101932525635, "learning_rate": 2.8780000000000002e-05, "loss": 0.1855, "step": 1440 }, { "grad_norm": 1.575304388999939, "learning_rate": 2.898e-05, "loss": 0.1633, "step": 1450 }, { "grad_norm": 1.586602807044983, "learning_rate": 2.9180000000000002e-05, "loss": 0.172, "step": 1460 }, { "grad_norm": 1.5125943422317505, "learning_rate": 2.9380000000000003e-05, "loss": 0.1642, "step": 1470 }, { "grad_norm": 1.7510510683059692, "learning_rate": 2.958e-05, "loss": 0.1589, "step": 1480 }, { "grad_norm": 1.7127574682235718, "learning_rate": 2.9780000000000003e-05, "loss": 0.1461, "step": 1490 }, { "grad_norm": 1.8046997785568237, "learning_rate": 2.998e-05, "loss": 0.1586, "step": 1500 }, { "grad_norm": 2.0884737968444824, "learning_rate": 3.0180000000000002e-05, "loss": 0.1461, "step": 1510 }, { "grad_norm": 1.9392865896224976, "learning_rate": 3.0380000000000004e-05, "loss": 0.1689, "step": 1520 }, { "grad_norm": 1.3286800384521484, "learning_rate": 3.058e-05, "loss": 0.1689, "step": 1530 }, { "grad_norm": 2.1085543632507324, "learning_rate": 3.078e-05, "loss": 0.1631, "step": 1540 }, { "grad_norm": 1.8939948081970215, "learning_rate": 3.0980000000000005e-05, "loss": 0.1498, "step": 1550 }, { "grad_norm": 2.1181821823120117, "learning_rate": 3.118e-05, "loss": 0.1727, "step": 1560 }, { "grad_norm": 1.7002955675125122, "learning_rate": 3.138e-05, "loss": 0.1474, "step": 1570 }, { "grad_norm": 1.7590913772583008, "learning_rate": 3.1580000000000006e-05, "loss": 0.1621, "step": 1580 }, { "grad_norm": 1.6942781209945679, "learning_rate": 3.1780000000000004e-05, "loss": 0.1608, "step": 1590 }, { "grad_norm": 1.7855643033981323, "learning_rate": 3.198e-05, "loss": 0.1541, "step": 1600 }, { "grad_norm": 1.698759913444519, "learning_rate": 3.218e-05, "loss": 0.1366, "step": 1610 }, { "grad_norm": 1.1855703592300415, "learning_rate": 3.238e-05, "loss": 0.1593, "step": 1620 }, { "grad_norm": 1.6936813592910767, "learning_rate": 3.2579999999999996e-05, "loss": 0.1761, "step": 1630 }, { "grad_norm": 1.6100504398345947, "learning_rate": 3.278e-05, "loss": 0.1473, "step": 1640 }, { "grad_norm": 1.400516390800476, "learning_rate": 3.298e-05, "loss": 0.1594, "step": 1650 }, { "grad_norm": 1.9690333604812622, "learning_rate": 3.318e-05, "loss": 0.1523, "step": 1660 }, { "grad_norm": 1.3166701793670654, "learning_rate": 3.338e-05, "loss": 0.142, "step": 1670 }, { "grad_norm": 1.7303414344787598, "learning_rate": 3.358e-05, "loss": 0.1402, "step": 1680 }, { "grad_norm": 1.316248893737793, "learning_rate": 3.378e-05, "loss": 0.1481, "step": 1690 }, { "grad_norm": 2.0378947257995605, "learning_rate": 3.398e-05, "loss": 0.1602, "step": 1700 }, { "grad_norm": 1.8722807168960571, "learning_rate": 3.418e-05, "loss": 0.1649, "step": 1710 }, { "grad_norm": 1.5514804124832153, "learning_rate": 3.438e-05, "loss": 0.1596, "step": 1720 }, { "grad_norm": 1.389766812324524, "learning_rate": 3.4580000000000004e-05, "loss": 0.1502, "step": 1730 }, { "grad_norm": 2.474884033203125, "learning_rate": 3.478e-05, "loss": 0.1401, "step": 1740 }, { "grad_norm": 1.6906168460845947, "learning_rate": 3.498e-05, "loss": 0.1556, "step": 1750 }, { "grad_norm": 2.0204007625579834, "learning_rate": 3.518e-05, "loss": 0.1496, "step": 1760 }, { "grad_norm": 1.3387794494628906, "learning_rate": 3.5380000000000003e-05, "loss": 0.1498, "step": 1770 }, { "grad_norm": 1.1275721788406372, "learning_rate": 3.558e-05, "loss": 0.1625, "step": 1780 }, { "grad_norm": 1.9130122661590576, "learning_rate": 3.578e-05, "loss": 0.1363, "step": 1790 }, { "grad_norm": 1.4939054250717163, "learning_rate": 3.5980000000000004e-05, "loss": 0.148, "step": 1800 }, { "grad_norm": 1.2533314228057861, "learning_rate": 3.618e-05, "loss": 0.136, "step": 1810 }, { "grad_norm": 1.6963413953781128, "learning_rate": 3.638e-05, "loss": 0.1452, "step": 1820 }, { "grad_norm": 1.7490293979644775, "learning_rate": 3.6580000000000006e-05, "loss": 0.1357, "step": 1830 }, { "grad_norm": 1.4599295854568481, "learning_rate": 3.6780000000000004e-05, "loss": 0.1486, "step": 1840 }, { "grad_norm": 1.4392956495285034, "learning_rate": 3.698e-05, "loss": 0.1849, "step": 1850 }, { "grad_norm": 1.8683152198791504, "learning_rate": 3.7180000000000007e-05, "loss": 0.1739, "step": 1860 }, { "grad_norm": 1.1515216827392578, "learning_rate": 3.7380000000000005e-05, "loss": 0.1449, "step": 1870 }, { "grad_norm": 1.5492677688598633, "learning_rate": 3.758e-05, "loss": 0.122, "step": 1880 }, { "grad_norm": 1.7967698574066162, "learning_rate": 3.778000000000001e-05, "loss": 0.1407, "step": 1890 }, { "grad_norm": 1.0091925859451294, "learning_rate": 3.7980000000000006e-05, "loss": 0.1424, "step": 1900 }, { "grad_norm": 1.715470790863037, "learning_rate": 3.818e-05, "loss": 0.1664, "step": 1910 }, { "grad_norm": 1.180148959159851, "learning_rate": 3.838e-05, "loss": 0.1519, "step": 1920 }, { "grad_norm": 1.0933903455734253, "learning_rate": 3.858e-05, "loss": 0.1342, "step": 1930 }, { "grad_norm": 1.643254041671753, "learning_rate": 3.878e-05, "loss": 0.1403, "step": 1940 }, { "grad_norm": 1.1347639560699463, "learning_rate": 3.898e-05, "loss": 0.1297, "step": 1950 }, { "grad_norm": 1.9953129291534424, "learning_rate": 3.918e-05, "loss": 0.1642, "step": 1960 }, { "grad_norm": 1.8393601179122925, "learning_rate": 3.938e-05, "loss": 0.1242, "step": 1970 }, { "grad_norm": 1.7340682744979858, "learning_rate": 3.958e-05, "loss": 0.1476, "step": 1980 }, { "grad_norm": 1.3368085622787476, "learning_rate": 3.978e-05, "loss": 0.1372, "step": 1990 }, { "grad_norm": 1.6145403385162354, "learning_rate": 3.998e-05, "loss": 0.137, "step": 2000 }, { "grad_norm": 1.4705896377563477, "learning_rate": 4.018e-05, "loss": 0.1744, "step": 2010 }, { "grad_norm": 1.714348316192627, "learning_rate": 4.038e-05, "loss": 0.1347, "step": 2020 }, { "grad_norm": 1.73570716381073, "learning_rate": 4.058e-05, "loss": 0.1153, "step": 2030 }, { "grad_norm": 1.0694289207458496, "learning_rate": 4.078e-05, "loss": 0.1351, "step": 2040 }, { "grad_norm": 1.5878716707229614, "learning_rate": 4.0980000000000004e-05, "loss": 0.1448, "step": 2050 }, { "grad_norm": 1.6058846712112427, "learning_rate": 4.118e-05, "loss": 0.1282, "step": 2060 }, { "grad_norm": 1.571274757385254, "learning_rate": 4.138e-05, "loss": 0.106, "step": 2070 }, { "grad_norm": 1.82014000415802, "learning_rate": 4.1580000000000005e-05, "loss": 0.1349, "step": 2080 }, { "grad_norm": 1.3415836095809937, "learning_rate": 4.178e-05, "loss": 0.1421, "step": 2090 }, { "grad_norm": 1.2755082845687866, "learning_rate": 4.198e-05, "loss": 0.1408, "step": 2100 }, { "grad_norm": 1.5193936824798584, "learning_rate": 4.2180000000000006e-05, "loss": 0.1746, "step": 2110 }, { "grad_norm": 1.4011003971099854, "learning_rate": 4.2380000000000004e-05, "loss": 0.14, "step": 2120 }, { "grad_norm": 1.114669919013977, "learning_rate": 4.258e-05, "loss": 0.1293, "step": 2130 }, { "grad_norm": 1.221988558769226, "learning_rate": 4.278e-05, "loss": 0.1072, "step": 2140 }, { "grad_norm": 1.3123196363449097, "learning_rate": 4.2980000000000005e-05, "loss": 0.1449, "step": 2150 }, { "grad_norm": 1.0453275442123413, "learning_rate": 4.318e-05, "loss": 0.1324, "step": 2160 }, { "grad_norm": 1.2893707752227783, "learning_rate": 4.338e-05, "loss": 0.1261, "step": 2170 }, { "grad_norm": 1.1385297775268555, "learning_rate": 4.3580000000000006e-05, "loss": 0.1514, "step": 2180 }, { "grad_norm": 1.1650900840759277, "learning_rate": 4.3780000000000004e-05, "loss": 0.1227, "step": 2190 }, { "grad_norm": 1.0900707244873047, "learning_rate": 4.398e-05, "loss": 0.1286, "step": 2200 }, { "grad_norm": 1.2437065839767456, "learning_rate": 4.418000000000001e-05, "loss": 0.1337, "step": 2210 }, { "grad_norm": 1.2069637775421143, "learning_rate": 4.438e-05, "loss": 0.1139, "step": 2220 }, { "grad_norm": 1.529842734336853, "learning_rate": 4.458e-05, "loss": 0.1365, "step": 2230 }, { "grad_norm": 1.2198429107666016, "learning_rate": 4.478e-05, "loss": 0.1198, "step": 2240 }, { "grad_norm": 1.0115875005722046, "learning_rate": 4.498e-05, "loss": 0.1385, "step": 2250 }, { "grad_norm": 1.0218024253845215, "learning_rate": 4.518e-05, "loss": 0.117, "step": 2260 }, { "grad_norm": 1.0345559120178223, "learning_rate": 4.538e-05, "loss": 0.1203, "step": 2270 }, { "grad_norm": 1.0485221147537231, "learning_rate": 4.558e-05, "loss": 0.1173, "step": 2280 }, { "grad_norm": 1.6061984300613403, "learning_rate": 4.578e-05, "loss": 0.1581, "step": 2290 }, { "grad_norm": 1.1830568313598633, "learning_rate": 4.5980000000000004e-05, "loss": 0.1102, "step": 2300 }, { "grad_norm": 1.2337013483047485, "learning_rate": 4.618e-05, "loss": 0.1384, "step": 2310 }, { "grad_norm": 0.8949767351150513, "learning_rate": 4.638e-05, "loss": 0.1174, "step": 2320 }, { "grad_norm": 1.0698639154434204, "learning_rate": 4.6580000000000005e-05, "loss": 0.1462, "step": 2330 }, { "grad_norm": 1.295883297920227, "learning_rate": 4.678e-05, "loss": 0.1283, "step": 2340 }, { "grad_norm": 1.0028705596923828, "learning_rate": 4.698e-05, "loss": 0.134, "step": 2350 }, { "grad_norm": 1.3857882022857666, "learning_rate": 4.718e-05, "loss": 0.1378, "step": 2360 }, { "grad_norm": 1.0203059911727905, "learning_rate": 4.7380000000000004e-05, "loss": 0.1498, "step": 2370 }, { "grad_norm": 0.9996606111526489, "learning_rate": 4.758e-05, "loss": 0.1377, "step": 2380 }, { "grad_norm": 0.8418086767196655, "learning_rate": 4.778e-05, "loss": 0.1452, "step": 2390 }, { "grad_norm": 1.2551019191741943, "learning_rate": 4.7980000000000005e-05, "loss": 0.1302, "step": 2400 }, { "grad_norm": 1.2393394708633423, "learning_rate": 4.818e-05, "loss": 0.1452, "step": 2410 }, { "grad_norm": 1.533171534538269, "learning_rate": 4.838e-05, "loss": 0.1811, "step": 2420 }, { "grad_norm": 1.3839366436004639, "learning_rate": 4.8580000000000006e-05, "loss": 0.137, "step": 2430 }, { "grad_norm": 1.533735990524292, "learning_rate": 4.8780000000000004e-05, "loss": 0.1464, "step": 2440 }, { "grad_norm": 1.291691541671753, "learning_rate": 4.898e-05, "loss": 0.1219, "step": 2450 }, { "grad_norm": 1.0277217626571655, "learning_rate": 4.918000000000001e-05, "loss": 0.1095, "step": 2460 }, { "grad_norm": 1.2336722612380981, "learning_rate": 4.9380000000000005e-05, "loss": 0.1116, "step": 2470 }, { "grad_norm": 1.1591224670410156, "learning_rate": 4.958e-05, "loss": 0.1481, "step": 2480 }, { "grad_norm": 1.2153942584991455, "learning_rate": 4.978e-05, "loss": 0.1128, "step": 2490 }, { "grad_norm": 1.014035701751709, "learning_rate": 4.9980000000000006e-05, "loss": 0.1185, "step": 2500 }, { "grad_norm": 1.2840336561203003, "learning_rate": 5.0180000000000004e-05, "loss": 0.1121, "step": 2510 }, { "grad_norm": 0.9230601787567139, "learning_rate": 5.038e-05, "loss": 0.1224, "step": 2520 }, { "grad_norm": 0.912327766418457, "learning_rate": 5.058000000000001e-05, "loss": 0.1121, "step": 2530 }, { "grad_norm": 1.0567822456359863, "learning_rate": 5.0780000000000005e-05, "loss": 0.1156, "step": 2540 }, { "grad_norm": 0.9102693200111389, "learning_rate": 5.098e-05, "loss": 0.1298, "step": 2550 }, { "grad_norm": 1.2006686925888062, "learning_rate": 5.118000000000001e-05, "loss": 0.1564, "step": 2560 }, { "grad_norm": 1.4650899171829224, "learning_rate": 5.1380000000000006e-05, "loss": 0.1248, "step": 2570 }, { "grad_norm": 1.0561851263046265, "learning_rate": 5.1580000000000004e-05, "loss": 0.1394, "step": 2580 }, { "grad_norm": 0.8095341324806213, "learning_rate": 5.178000000000001e-05, "loss": 0.1262, "step": 2590 }, { "grad_norm": 1.349546194076538, "learning_rate": 5.198000000000001e-05, "loss": 0.134, "step": 2600 }, { "grad_norm": 1.0753093957901, "learning_rate": 5.2180000000000005e-05, "loss": 0.1269, "step": 2610 }, { "grad_norm": 0.9891457557678223, "learning_rate": 5.238000000000001e-05, "loss": 0.116, "step": 2620 }, { "grad_norm": 1.0068366527557373, "learning_rate": 5.258000000000001e-05, "loss": 0.1239, "step": 2630 }, { "grad_norm": 1.3866342306137085, "learning_rate": 5.2780000000000006e-05, "loss": 0.1439, "step": 2640 }, { "grad_norm": 1.21674644947052, "learning_rate": 5.2980000000000004e-05, "loss": 0.1335, "step": 2650 }, { "grad_norm": 1.3749321699142456, "learning_rate": 5.318000000000001e-05, "loss": 0.1399, "step": 2660 }, { "grad_norm": 1.003164291381836, "learning_rate": 5.338000000000001e-05, "loss": 0.1492, "step": 2670 }, { "grad_norm": 1.2681872844696045, "learning_rate": 5.3580000000000005e-05, "loss": 0.153, "step": 2680 }, { "grad_norm": 1.037233591079712, "learning_rate": 5.378e-05, "loss": 0.1262, "step": 2690 }, { "grad_norm": 1.047550082206726, "learning_rate": 5.3979999999999995e-05, "loss": 0.1125, "step": 2700 }, { "grad_norm": 0.8866931200027466, "learning_rate": 5.418e-05, "loss": 0.1141, "step": 2710 }, { "grad_norm": 0.9175385236740112, "learning_rate": 5.438e-05, "loss": 0.1372, "step": 2720 }, { "grad_norm": 0.9836975336074829, "learning_rate": 5.4579999999999996e-05, "loss": 0.1252, "step": 2730 }, { "grad_norm": 1.1980557441711426, "learning_rate": 5.478e-05, "loss": 0.1265, "step": 2740 }, { "grad_norm": 1.112894058227539, "learning_rate": 5.498e-05, "loss": 0.1189, "step": 2750 }, { "grad_norm": 1.0806139707565308, "learning_rate": 5.518e-05, "loss": 0.1059, "step": 2760 }, { "grad_norm": 0.9796690940856934, "learning_rate": 5.538e-05, "loss": 0.1217, "step": 2770 }, { "grad_norm": 1.1660420894622803, "learning_rate": 5.558e-05, "loss": 0.1347, "step": 2780 }, { "grad_norm": 1.1473740339279175, "learning_rate": 5.578e-05, "loss": 0.1283, "step": 2790 }, { "grad_norm": 1.013344168663025, "learning_rate": 5.5979999999999996e-05, "loss": 0.1279, "step": 2800 }, { "grad_norm": 0.9554686546325684, "learning_rate": 5.618e-05, "loss": 0.1574, "step": 2810 }, { "grad_norm": 1.0622106790542603, "learning_rate": 5.638e-05, "loss": 0.1388, "step": 2820 }, { "grad_norm": 1.0678400993347168, "learning_rate": 5.658e-05, "loss": 0.1209, "step": 2830 }, { "grad_norm": 1.2139111757278442, "learning_rate": 5.678e-05, "loss": 0.1523, "step": 2840 }, { "grad_norm": 0.9777641892433167, "learning_rate": 5.698e-05, "loss": 0.1296, "step": 2850 }, { "grad_norm": 1.1673041582107544, "learning_rate": 5.718e-05, "loss": 0.1286, "step": 2860 }, { "grad_norm": 0.9848873615264893, "learning_rate": 5.738e-05, "loss": 0.106, "step": 2870 }, { "grad_norm": 0.8010454177856445, "learning_rate": 5.758e-05, "loss": 0.1092, "step": 2880 }, { "grad_norm": 1.0326006412506104, "learning_rate": 5.778e-05, "loss": 0.1275, "step": 2890 }, { "grad_norm": 0.9177785515785217, "learning_rate": 5.7980000000000004e-05, "loss": 0.1303, "step": 2900 }, { "grad_norm": 1.0886006355285645, "learning_rate": 5.818e-05, "loss": 0.1197, "step": 2910 }, { "grad_norm": 0.6738573312759399, "learning_rate": 5.838e-05, "loss": 0.1151, "step": 2920 }, { "grad_norm": 0.9993757605552673, "learning_rate": 5.858e-05, "loss": 0.0977, "step": 2930 }, { "grad_norm": 0.9890748858451843, "learning_rate": 5.878e-05, "loss": 0.0985, "step": 2940 }, { "grad_norm": 0.8689008355140686, "learning_rate": 5.898e-05, "loss": 0.0972, "step": 2950 }, { "grad_norm": 0.8470520377159119, "learning_rate": 5.918e-05, "loss": 0.1121, "step": 2960 }, { "grad_norm": 1.112988829612732, "learning_rate": 5.9380000000000004e-05, "loss": 0.1175, "step": 2970 }, { "grad_norm": 0.9986160397529602, "learning_rate": 5.958e-05, "loss": 0.1178, "step": 2980 }, { "grad_norm": 0.9575360417366028, "learning_rate": 5.978e-05, "loss": 0.1278, "step": 2990 }, { "grad_norm": 0.7282163500785828, "learning_rate": 5.9980000000000005e-05, "loss": 0.1289, "step": 3000 }, { "grad_norm": 1.00343656539917, "learning_rate": 6.018e-05, "loss": 0.108, "step": 3010 }, { "grad_norm": 0.7787696123123169, "learning_rate": 6.038e-05, "loss": 0.1215, "step": 3020 }, { "grad_norm": 1.164481282234192, "learning_rate": 6.0580000000000006e-05, "loss": 0.1163, "step": 3030 }, { "grad_norm": 1.0298131704330444, "learning_rate": 6.0780000000000004e-05, "loss": 0.1189, "step": 3040 }, { "grad_norm": 1.1403284072875977, "learning_rate": 6.098e-05, "loss": 0.1088, "step": 3050 }, { "grad_norm": 1.0276285409927368, "learning_rate": 6.118000000000001e-05, "loss": 0.1225, "step": 3060 }, { "grad_norm": 1.0896973609924316, "learning_rate": 6.138e-05, "loss": 0.1109, "step": 3070 }, { "grad_norm": 1.1300747394561768, "learning_rate": 6.158e-05, "loss": 0.1166, "step": 3080 }, { "grad_norm": 1.2934088706970215, "learning_rate": 6.178000000000001e-05, "loss": 0.1422, "step": 3090 }, { "grad_norm": 1.2242493629455566, "learning_rate": 6.198e-05, "loss": 0.1217, "step": 3100 }, { "grad_norm": 0.6924254894256592, "learning_rate": 6.218e-05, "loss": 0.1067, "step": 3110 }, { "grad_norm": 0.8460937142372131, "learning_rate": 6.238000000000001e-05, "loss": 0.0936, "step": 3120 }, { "grad_norm": 0.9960484504699707, "learning_rate": 6.258e-05, "loss": 0.1108, "step": 3130 }, { "grad_norm": 0.9967566728591919, "learning_rate": 6.278e-05, "loss": 0.116, "step": 3140 }, { "grad_norm": 0.8094844222068787, "learning_rate": 6.298000000000001e-05, "loss": 0.1283, "step": 3150 }, { "grad_norm": 1.0407308340072632, "learning_rate": 6.318e-05, "loss": 0.1139, "step": 3160 }, { "grad_norm": 0.873887836933136, "learning_rate": 6.338e-05, "loss": 0.1146, "step": 3170 }, { "grad_norm": 1.205177903175354, "learning_rate": 6.358000000000001e-05, "loss": 0.1398, "step": 3180 }, { "grad_norm": 0.6775646805763245, "learning_rate": 6.378e-05, "loss": 0.1267, "step": 3190 }, { "grad_norm": 0.9172083139419556, "learning_rate": 6.398000000000001e-05, "loss": 0.1204, "step": 3200 }, { "grad_norm": 1.026411771774292, "learning_rate": 6.418000000000001e-05, "loss": 0.1138, "step": 3210 }, { "grad_norm": 1.7333134412765503, "learning_rate": 6.438e-05, "loss": 0.1601, "step": 3220 }, { "grad_norm": 1.0319480895996094, "learning_rate": 6.458000000000001e-05, "loss": 0.1407, "step": 3230 }, { "grad_norm": 0.7345286011695862, "learning_rate": 6.478000000000001e-05, "loss": 0.1105, "step": 3240 }, { "grad_norm": 0.8060258030891418, "learning_rate": 6.498e-05, "loss": 0.1192, "step": 3250 }, { "grad_norm": 1.431876540184021, "learning_rate": 6.518000000000001e-05, "loss": 0.1261, "step": 3260 }, { "grad_norm": 0.9989150762557983, "learning_rate": 6.538000000000001e-05, "loss": 0.1176, "step": 3270 }, { "grad_norm": 1.2062376737594604, "learning_rate": 6.558e-05, "loss": 0.105, "step": 3280 }, { "grad_norm": 0.8581637740135193, "learning_rate": 6.578000000000001e-05, "loss": 0.1145, "step": 3290 }, { "grad_norm": 2.020275831222534, "learning_rate": 6.598e-05, "loss": 0.1262, "step": 3300 }, { "grad_norm": 1.1280765533447266, "learning_rate": 6.618e-05, "loss": 0.1233, "step": 3310 }, { "grad_norm": 0.6145249605178833, "learning_rate": 6.638e-05, "loss": 0.1501, "step": 3320 }, { "grad_norm": 0.844628632068634, "learning_rate": 6.658e-05, "loss": 0.1153, "step": 3330 }, { "grad_norm": 0.9398614764213562, "learning_rate": 6.678e-05, "loss": 0.126, "step": 3340 }, { "grad_norm": 0.6847575306892395, "learning_rate": 6.698e-05, "loss": 0.1205, "step": 3350 }, { "grad_norm": 1.0560985803604126, "learning_rate": 6.718e-05, "loss": 0.1093, "step": 3360 }, { "grad_norm": 0.7167390584945679, "learning_rate": 6.738e-05, "loss": 0.1066, "step": 3370 }, { "grad_norm": 0.8113360404968262, "learning_rate": 6.758e-05, "loss": 0.1533, "step": 3380 }, { "grad_norm": 0.8806983232498169, "learning_rate": 6.778e-05, "loss": 0.1215, "step": 3390 }, { "grad_norm": 0.9252278804779053, "learning_rate": 6.798e-05, "loss": 0.1272, "step": 3400 }, { "grad_norm": 1.279195785522461, "learning_rate": 6.818e-05, "loss": 0.1279, "step": 3410 }, { "grad_norm": 1.0971617698669434, "learning_rate": 6.838e-05, "loss": 0.1125, "step": 3420 }, { "grad_norm": 0.8646501898765564, "learning_rate": 6.858e-05, "loss": 0.1115, "step": 3430 }, { "grad_norm": 0.876218855381012, "learning_rate": 6.878e-05, "loss": 0.1247, "step": 3440 }, { "grad_norm": 0.7517340183258057, "learning_rate": 6.898e-05, "loss": 0.124, "step": 3450 }, { "grad_norm": 0.8869956731796265, "learning_rate": 6.918e-05, "loss": 0.1113, "step": 3460 }, { "grad_norm": 0.911651611328125, "learning_rate": 6.938e-05, "loss": 0.1232, "step": 3470 }, { "grad_norm": 0.7686192393302917, "learning_rate": 6.958e-05, "loss": 0.128, "step": 3480 }, { "grad_norm": 0.6786593794822693, "learning_rate": 6.978e-05, "loss": 0.1072, "step": 3490 }, { "grad_norm": 0.9375817179679871, "learning_rate": 6.998e-05, "loss": 0.0991, "step": 3500 }, { "grad_norm": 0.9730860590934753, "learning_rate": 7.018e-05, "loss": 0.1104, "step": 3510 }, { "grad_norm": 0.735240638256073, "learning_rate": 7.038e-05, "loss": 0.1081, "step": 3520 }, { "grad_norm": 0.8927597999572754, "learning_rate": 7.058e-05, "loss": 0.1111, "step": 3530 }, { "grad_norm": 0.7174538373947144, "learning_rate": 7.078e-05, "loss": 0.1033, "step": 3540 }, { "grad_norm": 0.9131749868392944, "learning_rate": 7.098e-05, "loss": 0.1195, "step": 3550 }, { "grad_norm": 0.9278097748756409, "learning_rate": 7.118e-05, "loss": 0.0913, "step": 3560 }, { "grad_norm": 0.8304092884063721, "learning_rate": 7.138e-05, "loss": 0.1259, "step": 3570 }, { "grad_norm": 0.9198570847511292, "learning_rate": 7.158e-05, "loss": 0.1379, "step": 3580 }, { "grad_norm": 1.0696516036987305, "learning_rate": 7.178000000000001e-05, "loss": 0.1267, "step": 3590 }, { "grad_norm": 0.8791998028755188, "learning_rate": 7.198e-05, "loss": 0.1218, "step": 3600 }, { "grad_norm": 0.7616946697235107, "learning_rate": 7.218e-05, "loss": 0.1052, "step": 3610 }, { "grad_norm": 0.7529808878898621, "learning_rate": 7.238000000000001e-05, "loss": 0.1077, "step": 3620 }, { "grad_norm": 0.7046818733215332, "learning_rate": 7.258e-05, "loss": 0.0945, "step": 3630 }, { "grad_norm": 1.0456006526947021, "learning_rate": 7.278e-05, "loss": 0.1078, "step": 3640 }, { "grad_norm": 1.0489368438720703, "learning_rate": 7.298000000000001e-05, "loss": 0.1198, "step": 3650 }, { "grad_norm": 0.991188645362854, "learning_rate": 7.318e-05, "loss": 0.1242, "step": 3660 }, { "grad_norm": 0.9548091292381287, "learning_rate": 7.338e-05, "loss": 0.1294, "step": 3670 }, { "grad_norm": 1.1672680377960205, "learning_rate": 7.358000000000001e-05, "loss": 0.1211, "step": 3680 }, { "grad_norm": 0.9176283478736877, "learning_rate": 7.378e-05, "loss": 0.1127, "step": 3690 }, { "grad_norm": 0.555199146270752, "learning_rate": 7.398e-05, "loss": 0.12, "step": 3700 }, { "grad_norm": 0.9543347954750061, "learning_rate": 7.418000000000001e-05, "loss": 0.1229, "step": 3710 }, { "grad_norm": 0.5928622484207153, "learning_rate": 7.438e-05, "loss": 0.1131, "step": 3720 }, { "grad_norm": 0.7619514465332031, "learning_rate": 7.458000000000001e-05, "loss": 0.0906, "step": 3730 }, { "grad_norm": 0.7846157550811768, "learning_rate": 7.478e-05, "loss": 0.1023, "step": 3740 }, { "grad_norm": 0.7935389280319214, "learning_rate": 7.498e-05, "loss": 0.1209, "step": 3750 }, { "grad_norm": 0.7776572108268738, "learning_rate": 7.518000000000001e-05, "loss": 0.0963, "step": 3760 }, { "grad_norm": 1.1459238529205322, "learning_rate": 7.538e-05, "loss": 0.1153, "step": 3770 }, { "grad_norm": 0.838996946811676, "learning_rate": 7.558e-05, "loss": 0.1109, "step": 3780 }, { "grad_norm": 0.7485544681549072, "learning_rate": 7.578000000000001e-05, "loss": 0.1125, "step": 3790 }, { "grad_norm": 1.001765489578247, "learning_rate": 7.598e-05, "loss": 0.1295, "step": 3800 }, { "grad_norm": 0.6430847644805908, "learning_rate": 7.618e-05, "loss": 0.1051, "step": 3810 }, { "grad_norm": 0.8795580267906189, "learning_rate": 7.638000000000001e-05, "loss": 0.1101, "step": 3820 }, { "grad_norm": 0.7451896071434021, "learning_rate": 7.658e-05, "loss": 0.1024, "step": 3830 }, { "grad_norm": 0.8010369539260864, "learning_rate": 7.678000000000001e-05, "loss": 0.1022, "step": 3840 }, { "grad_norm": 0.8008686900138855, "learning_rate": 7.698000000000001e-05, "loss": 0.1107, "step": 3850 }, { "grad_norm": 1.0054020881652832, "learning_rate": 7.718e-05, "loss": 0.1121, "step": 3860 }, { "grad_norm": 0.7650145888328552, "learning_rate": 7.738000000000001e-05, "loss": 0.0992, "step": 3870 }, { "grad_norm": 0.6757243275642395, "learning_rate": 7.758000000000001e-05, "loss": 0.0831, "step": 3880 }, { "grad_norm": 0.9322896599769592, "learning_rate": 7.778e-05, "loss": 0.1063, "step": 3890 }, { "grad_norm": 0.8216775059700012, "learning_rate": 7.798000000000001e-05, "loss": 0.1032, "step": 3900 }, { "grad_norm": 0.9977006316184998, "learning_rate": 7.818000000000001e-05, "loss": 0.0964, "step": 3910 }, { "grad_norm": 1.0798285007476807, "learning_rate": 7.838e-05, "loss": 0.117, "step": 3920 }, { "grad_norm": 0.7515882253646851, "learning_rate": 7.858000000000001e-05, "loss": 0.1018, "step": 3930 }, { "grad_norm": 0.8430335521697998, "learning_rate": 7.878e-05, "loss": 0.1094, "step": 3940 }, { "grad_norm": 0.8087600469589233, "learning_rate": 7.897999999999999e-05, "loss": 0.1005, "step": 3950 }, { "grad_norm": 0.8770368695259094, "learning_rate": 7.918e-05, "loss": 0.1127, "step": 3960 }, { "grad_norm": 0.7110810875892639, "learning_rate": 7.938e-05, "loss": 0.0854, "step": 3970 }, { "grad_norm": 0.6262394785881042, "learning_rate": 7.958e-05, "loss": 0.1206, "step": 3980 }, { "grad_norm": 0.9287399053573608, "learning_rate": 7.978e-05, "loss": 0.1092, "step": 3990 }, { "grad_norm": 1.0216073989868164, "learning_rate": 7.998e-05, "loss": 0.1113, "step": 4000 }, { "grad_norm": 0.689810037612915, "learning_rate": 8.018e-05, "loss": 0.0985, "step": 4010 }, { "grad_norm": 0.869611918926239, "learning_rate": 8.038e-05, "loss": 0.0975, "step": 4020 }, { "grad_norm": 0.5785909295082092, "learning_rate": 8.058e-05, "loss": 0.1072, "step": 4030 }, { "grad_norm": 1.0798239707946777, "learning_rate": 8.078e-05, "loss": 0.0887, "step": 4040 }, { "grad_norm": 0.7362409234046936, "learning_rate": 8.098e-05, "loss": 0.1096, "step": 4050 }, { "grad_norm": 0.7699953317642212, "learning_rate": 8.118e-05, "loss": 0.1065, "step": 4060 }, { "grad_norm": 0.8991296291351318, "learning_rate": 8.138e-05, "loss": 0.1051, "step": 4070 }, { "grad_norm": 0.7962868213653564, "learning_rate": 8.158e-05, "loss": 0.1193, "step": 4080 }, { "grad_norm": 0.8073585629463196, "learning_rate": 8.178e-05, "loss": 0.1006, "step": 4090 }, { "grad_norm": 0.8526430130004883, "learning_rate": 8.198e-05, "loss": 0.0987, "step": 4100 }, { "grad_norm": 0.7617095112800598, "learning_rate": 8.218e-05, "loss": 0.1027, "step": 4110 }, { "grad_norm": 0.9189274311065674, "learning_rate": 8.238000000000001e-05, "loss": 0.119, "step": 4120 }, { "grad_norm": 0.725437343120575, "learning_rate": 8.258e-05, "loss": 0.0964, "step": 4130 }, { "grad_norm": 0.6900134682655334, "learning_rate": 8.278e-05, "loss": 0.109, "step": 4140 }, { "grad_norm": 0.7917046546936035, "learning_rate": 8.298000000000001e-05, "loss": 0.1268, "step": 4150 }, { "grad_norm": 0.56847083568573, "learning_rate": 8.318e-05, "loss": 0.0834, "step": 4160 }, { "grad_norm": 0.7717142701148987, "learning_rate": 8.338e-05, "loss": 0.0959, "step": 4170 }, { "grad_norm": 0.7160118222236633, "learning_rate": 8.358e-05, "loss": 0.0967, "step": 4180 }, { "grad_norm": 1.0377726554870605, "learning_rate": 8.378e-05, "loss": 0.1128, "step": 4190 }, { "grad_norm": 0.6575815677642822, "learning_rate": 8.398e-05, "loss": 0.1058, "step": 4200 }, { "grad_norm": 0.622734546661377, "learning_rate": 8.418e-05, "loss": 0.122, "step": 4210 }, { "grad_norm": 0.9631053805351257, "learning_rate": 8.438e-05, "loss": 0.1329, "step": 4220 }, { "grad_norm": 0.9485803246498108, "learning_rate": 8.458e-05, "loss": 0.1353, "step": 4230 }, { "grad_norm": 0.8246829509735107, "learning_rate": 8.478e-05, "loss": 0.1253, "step": 4240 }, { "grad_norm": 0.7812141180038452, "learning_rate": 8.498e-05, "loss": 0.1203, "step": 4250 }, { "grad_norm": 0.8272555470466614, "learning_rate": 8.518000000000001e-05, "loss": 0.1312, "step": 4260 }, { "grad_norm": 0.5899245738983154, "learning_rate": 8.538e-05, "loss": 0.1062, "step": 4270 }, { "grad_norm": 0.9789780378341675, "learning_rate": 8.558e-05, "loss": 0.1197, "step": 4280 }, { "grad_norm": 0.7665773630142212, "learning_rate": 8.578000000000001e-05, "loss": 0.1146, "step": 4290 }, { "grad_norm": 0.7946252822875977, "learning_rate": 8.598e-05, "loss": 0.1191, "step": 4300 }, { "grad_norm": 1.0886846780776978, "learning_rate": 8.618e-05, "loss": 0.1215, "step": 4310 }, { "grad_norm": 0.8715952038764954, "learning_rate": 8.638000000000001e-05, "loss": 0.1457, "step": 4320 }, { "grad_norm": 0.9267996549606323, "learning_rate": 8.658e-05, "loss": 0.1078, "step": 4330 }, { "grad_norm": 0.6128423810005188, "learning_rate": 8.678e-05, "loss": 0.0961, "step": 4340 }, { "grad_norm": 0.7897301912307739, "learning_rate": 8.698000000000001e-05, "loss": 0.081, "step": 4350 }, { "grad_norm": 0.6687045097351074, "learning_rate": 8.718e-05, "loss": 0.1046, "step": 4360 }, { "grad_norm": 0.7629536390304565, "learning_rate": 8.738000000000001e-05, "loss": 0.103, "step": 4370 }, { "grad_norm": 0.7697046399116516, "learning_rate": 8.758000000000001e-05, "loss": 0.1106, "step": 4380 }, { "grad_norm": 0.9960927367210388, "learning_rate": 8.778e-05, "loss": 0.1036, "step": 4390 }, { "grad_norm": 0.6359614729881287, "learning_rate": 8.798000000000001e-05, "loss": 0.137, "step": 4400 }, { "grad_norm": 0.6596130728721619, "learning_rate": 8.818000000000001e-05, "loss": 0.1241, "step": 4410 }, { "grad_norm": 0.8426551818847656, "learning_rate": 8.838e-05, "loss": 0.1232, "step": 4420 }, { "grad_norm": 0.783476710319519, "learning_rate": 8.858000000000001e-05, "loss": 0.1192, "step": 4430 }, { "grad_norm": 0.9050313234329224, "learning_rate": 8.878000000000001e-05, "loss": 0.1028, "step": 4440 }, { "grad_norm": 0.6061708331108093, "learning_rate": 8.898e-05, "loss": 0.1298, "step": 4450 }, { "grad_norm": 0.7183692455291748, "learning_rate": 8.918000000000001e-05, "loss": 0.1212, "step": 4460 }, { "grad_norm": 0.6867538094520569, "learning_rate": 8.938e-05, "loss": 0.1137, "step": 4470 }, { "grad_norm": 0.8406731486320496, "learning_rate": 8.958e-05, "loss": 0.1072, "step": 4480 }, { "grad_norm": 0.9555157423019409, "learning_rate": 8.978000000000001e-05, "loss": 0.1312, "step": 4490 }, { "grad_norm": 0.8291189074516296, "learning_rate": 8.998e-05, "loss": 0.1206, "step": 4500 }, { "grad_norm": 0.7693631649017334, "learning_rate": 9.018000000000001e-05, "loss": 0.1301, "step": 4510 }, { "grad_norm": 0.9736393690109253, "learning_rate": 9.038000000000001e-05, "loss": 0.1115, "step": 4520 }, { "grad_norm": 0.6159675121307373, "learning_rate": 9.058e-05, "loss": 0.1108, "step": 4530 }, { "grad_norm": 0.8217610120773315, "learning_rate": 9.078000000000001e-05, "loss": 0.1321, "step": 4540 }, { "grad_norm": 0.6583220958709717, "learning_rate": 9.098000000000001e-05, "loss": 0.1317, "step": 4550 }, { "grad_norm": 0.6816418170928955, "learning_rate": 9.118e-05, "loss": 0.1037, "step": 4560 }, { "grad_norm": 0.6437064409255981, "learning_rate": 9.138e-05, "loss": 0.123, "step": 4570 }, { "grad_norm": 1.1537479162216187, "learning_rate": 9.158e-05, "loss": 0.1065, "step": 4580 }, { "grad_norm": 0.6960337162017822, "learning_rate": 9.178e-05, "loss": 0.1284, "step": 4590 }, { "grad_norm": 0.8225893378257751, "learning_rate": 9.198e-05, "loss": 0.1087, "step": 4600 }, { "grad_norm": 0.9905645847320557, "learning_rate": 9.218e-05, "loss": 0.1141, "step": 4610 }, { "grad_norm": 0.8048555254936218, "learning_rate": 9.238e-05, "loss": 0.1071, "step": 4620 }, { "grad_norm": 1.098063588142395, "learning_rate": 9.258e-05, "loss": 0.123, "step": 4630 }, { "grad_norm": 0.7546206712722778, "learning_rate": 9.278e-05, "loss": 0.1257, "step": 4640 }, { "grad_norm": 0.9112017750740051, "learning_rate": 9.298e-05, "loss": 0.1137, "step": 4650 }, { "grad_norm": 0.7669366598129272, "learning_rate": 9.318e-05, "loss": 0.1109, "step": 4660 }, { "grad_norm": 1.004346251487732, "learning_rate": 9.338e-05, "loss": 0.1013, "step": 4670 }, { "grad_norm": 0.8270790576934814, "learning_rate": 9.358e-05, "loss": 0.115, "step": 4680 }, { "grad_norm": 0.7456740736961365, "learning_rate": 9.378e-05, "loss": 0.1225, "step": 4690 }, { "grad_norm": 0.8754923343658447, "learning_rate": 9.398e-05, "loss": 0.1134, "step": 4700 }, { "grad_norm": 0.6692763566970825, "learning_rate": 9.418e-05, "loss": 0.1321, "step": 4710 }, { "grad_norm": 0.8302074670791626, "learning_rate": 9.438e-05, "loss": 0.1182, "step": 4720 }, { "grad_norm": 0.7491344213485718, "learning_rate": 9.458e-05, "loss": 0.1201, "step": 4730 }, { "grad_norm": 0.7187442183494568, "learning_rate": 9.478e-05, "loss": 0.1082, "step": 4740 }, { "grad_norm": 0.8242186903953552, "learning_rate": 9.498e-05, "loss": 0.1037, "step": 4750 }, { "grad_norm": 0.7524170875549316, "learning_rate": 9.518000000000001e-05, "loss": 0.1093, "step": 4760 }, { "grad_norm": 0.8877623677253723, "learning_rate": 9.538e-05, "loss": 0.116, "step": 4770 }, { "grad_norm": 1.158291220664978, "learning_rate": 9.558e-05, "loss": 0.1225, "step": 4780 }, { "grad_norm": 0.7551198601722717, "learning_rate": 9.578000000000001e-05, "loss": 0.1029, "step": 4790 }, { "grad_norm": 0.6464711427688599, "learning_rate": 9.598e-05, "loss": 0.1029, "step": 4800 }, { "grad_norm": 0.9193889498710632, "learning_rate": 9.618e-05, "loss": 0.1037, "step": 4810 }, { "grad_norm": 0.6666151881217957, "learning_rate": 9.638000000000001e-05, "loss": 0.1099, "step": 4820 }, { "grad_norm": 0.604324460029602, "learning_rate": 9.658e-05, "loss": 0.125, "step": 4830 }, { "grad_norm": 0.6216613054275513, "learning_rate": 9.678e-05, "loss": 0.1215, "step": 4840 }, { "grad_norm": 0.5685839653015137, "learning_rate": 9.698000000000001e-05, "loss": 0.1182, "step": 4850 }, { "grad_norm": 0.6165072321891785, "learning_rate": 9.718e-05, "loss": 0.0984, "step": 4860 }, { "grad_norm": 0.7066221237182617, "learning_rate": 9.738e-05, "loss": 0.1053, "step": 4870 }, { "grad_norm": 0.7000887393951416, "learning_rate": 9.758000000000001e-05, "loss": 0.1044, "step": 4880 }, { "grad_norm": 1.0005630254745483, "learning_rate": 9.778e-05, "loss": 0.1115, "step": 4890 }, { "grad_norm": 0.7909587025642395, "learning_rate": 9.798000000000001e-05, "loss": 0.1086, "step": 4900 }, { "grad_norm": 0.661105751991272, "learning_rate": 9.818000000000001e-05, "loss": 0.1214, "step": 4910 }, { "grad_norm": 0.7444678544998169, "learning_rate": 9.838e-05, "loss": 0.1295, "step": 4920 }, { "grad_norm": 1.0373756885528564, "learning_rate": 9.858000000000001e-05, "loss": 0.1168, "step": 4930 }, { "grad_norm": 0.6097736358642578, "learning_rate": 9.878e-05, "loss": 0.1062, "step": 4940 }, { "grad_norm": 0.6878100633621216, "learning_rate": 9.898e-05, "loss": 0.1234, "step": 4950 }, { "grad_norm": 0.7370219230651855, "learning_rate": 9.918000000000001e-05, "loss": 0.1113, "step": 4960 }, { "grad_norm": 1.0565142631530762, "learning_rate": 9.938e-05, "loss": 0.1321, "step": 4970 }, { "grad_norm": 0.8572822213172913, "learning_rate": 9.958e-05, "loss": 0.1199, "step": 4980 }, { "grad_norm": 0.6125404238700867, "learning_rate": 9.978000000000001e-05, "loss": 0.1149, "step": 4990 }, { "grad_norm": 0.840225100517273, "learning_rate": 9.998e-05, "loss": 0.1328, "step": 5000 }, { "grad_norm": 0.9344463348388672, "learning_rate": 9.999999778549045e-05, "loss": 0.1181, "step": 5010 }, { "grad_norm": 0.8454543948173523, "learning_rate": 9.999999013039593e-05, "loss": 0.1072, "step": 5020 }, { "grad_norm": 0.6857911348342896, "learning_rate": 9.999997700737766e-05, "loss": 0.1021, "step": 5030 }, { "grad_norm": 0.709067165851593, "learning_rate": 9.999995841643709e-05, "loss": 0.0952, "step": 5040 }, { "grad_norm": 0.5978896021842957, "learning_rate": 9.999993435757623e-05, "loss": 0.1214, "step": 5050 }, { "grad_norm": 0.7139127850532532, "learning_rate": 9.999990483079773e-05, "loss": 0.0827, "step": 5060 }, { "grad_norm": 0.728933572769165, "learning_rate": 9.999986983610481e-05, "loss": 0.1394, "step": 5070 }, { "grad_norm": 0.6932781934738159, "learning_rate": 9.99998293735013e-05, "loss": 0.1068, "step": 5080 }, { "grad_norm": 0.5593099594116211, "learning_rate": 9.999978344299161e-05, "loss": 0.0941, "step": 5090 }, { "grad_norm": 0.7351681590080261, "learning_rate": 9.99997320445808e-05, "loss": 0.0917, "step": 5100 }, { "grad_norm": 0.7798919677734375, "learning_rate": 9.999967517827444e-05, "loss": 0.1194, "step": 5110 }, { "grad_norm": 0.9062651991844177, "learning_rate": 9.999961284407879e-05, "loss": 0.1146, "step": 5120 }, { "grad_norm": 0.7182037234306335, "learning_rate": 9.999954504200067e-05, "loss": 0.1247, "step": 5130 }, { "grad_norm": 0.7708905339241028, "learning_rate": 9.999947177204744e-05, "loss": 0.1248, "step": 5140 }, { "grad_norm": 1.1626973152160645, "learning_rate": 9.999939303422718e-05, "loss": 0.1068, "step": 5150 }, { "grad_norm": 0.6320338845252991, "learning_rate": 9.999930882854847e-05, "loss": 0.1306, "step": 5160 }, { "grad_norm": 0.8189074397087097, "learning_rate": 9.999921915502051e-05, "loss": 0.1124, "step": 5170 }, { "grad_norm": 0.8293450474739075, "learning_rate": 9.99991240136531e-05, "loss": 0.0939, "step": 5180 }, { "grad_norm": 0.788169264793396, "learning_rate": 9.999902340445668e-05, "loss": 0.1305, "step": 5190 }, { "grad_norm": 0.6513338088989258, "learning_rate": 9.999891732744224e-05, "loss": 0.0911, "step": 5200 }, { "grad_norm": 2.169926166534424, "learning_rate": 9.999880578262135e-05, "loss": 0.1019, "step": 5210 }, { "grad_norm": 0.7161943316459656, "learning_rate": 9.999868877000624e-05, "loss": 0.1254, "step": 5220 }, { "grad_norm": 0.6481282711029053, "learning_rate": 9.99985662896097e-05, "loss": 0.0996, "step": 5230 }, { "grad_norm": 0.6162830591201782, "learning_rate": 9.999843834144513e-05, "loss": 0.1132, "step": 5240 }, { "grad_norm": 0.6669424772262573, "learning_rate": 9.99983049255265e-05, "loss": 0.1304, "step": 5250 }, { "grad_norm": 0.8260868191719055, "learning_rate": 9.999816604186843e-05, "loss": 0.0926, "step": 5260 }, { "grad_norm": 0.7663576602935791, "learning_rate": 9.999802169048609e-05, "loss": 0.1058, "step": 5270 }, { "grad_norm": 0.699521005153656, "learning_rate": 9.999787187139527e-05, "loss": 0.1213, "step": 5280 }, { "grad_norm": 0.8022137880325317, "learning_rate": 9.999771658461234e-05, "loss": 0.1239, "step": 5290 }, { "grad_norm": 0.6243753433227539, "learning_rate": 9.999755583015431e-05, "loss": 0.1024, "step": 5300 }, { "grad_norm": 0.6390677094459534, "learning_rate": 9.999738960803874e-05, "loss": 0.0863, "step": 5310 }, { "grad_norm": 0.6959462761878967, "learning_rate": 9.99972179182838e-05, "loss": 0.0984, "step": 5320 }, { "grad_norm": 0.8521237373352051, "learning_rate": 9.99970407609083e-05, "loss": 0.1157, "step": 5330 }, { "grad_norm": 0.640988290309906, "learning_rate": 9.999685813593159e-05, "loss": 0.0959, "step": 5340 }, { "grad_norm": 0.741902232170105, "learning_rate": 9.999667004337362e-05, "loss": 0.0954, "step": 5350 }, { "grad_norm": 0.6892989873886108, "learning_rate": 9.9996476483255e-05, "loss": 0.1066, "step": 5360 }, { "grad_norm": 0.8130404949188232, "learning_rate": 9.999627745559688e-05, "loss": 0.1082, "step": 5370 }, { "grad_norm": 0.6115046143531799, "learning_rate": 9.999607296042101e-05, "loss": 0.1215, "step": 5380 }, { "grad_norm": 0.6472975611686707, "learning_rate": 9.99958629977498e-05, "loss": 0.0905, "step": 5390 }, { "grad_norm": 1.0869901180267334, "learning_rate": 9.999564756760615e-05, "loss": 0.112, "step": 5400 }, { "grad_norm": 0.6310997605323792, "learning_rate": 9.999542667001366e-05, "loss": 0.0967, "step": 5410 }, { "grad_norm": 0.7383485436439514, "learning_rate": 9.999520030499647e-05, "loss": 0.1122, "step": 5420 }, { "grad_norm": 0.8050594329833984, "learning_rate": 9.999496847257936e-05, "loss": 0.1124, "step": 5430 }, { "grad_norm": 0.7535364031791687, "learning_rate": 9.999473117278764e-05, "loss": 0.1275, "step": 5440 }, { "grad_norm": 0.8712766170501709, "learning_rate": 9.999448840564731e-05, "loss": 0.0938, "step": 5450 }, { "grad_norm": 1.276078224182129, "learning_rate": 9.999424017118488e-05, "loss": 0.0998, "step": 5460 }, { "grad_norm": 0.8417381644248962, "learning_rate": 9.999398646942751e-05, "loss": 0.1004, "step": 5470 }, { "grad_norm": 0.6495277881622314, "learning_rate": 9.999372730040296e-05, "loss": 0.0959, "step": 5480 }, { "grad_norm": 0.6320420503616333, "learning_rate": 9.999346266413953e-05, "loss": 0.1005, "step": 5490 }, { "grad_norm": 0.6688899993896484, "learning_rate": 9.99931925606662e-05, "loss": 0.1148, "step": 5500 }, { "grad_norm": 0.6016878485679626, "learning_rate": 9.99929169900125e-05, "loss": 0.1073, "step": 5510 }, { "grad_norm": 0.656457781791687, "learning_rate": 9.999263595220855e-05, "loss": 0.0849, "step": 5520 }, { "grad_norm": 0.6307219862937927, "learning_rate": 9.99923494472851e-05, "loss": 0.0989, "step": 5530 }, { "grad_norm": 0.9449863433837891, "learning_rate": 9.999205747527348e-05, "loss": 0.0967, "step": 5540 }, { "grad_norm": 0.7478062510490417, "learning_rate": 9.999176003620561e-05, "loss": 0.1006, "step": 5550 }, { "grad_norm": 0.6899396777153015, "learning_rate": 9.999145713011405e-05, "loss": 0.1019, "step": 5560 }, { "grad_norm": 0.8057308197021484, "learning_rate": 9.999114875703186e-05, "loss": 0.1055, "step": 5570 }, { "grad_norm": 0.7194453477859497, "learning_rate": 9.999083491699281e-05, "loss": 0.1166, "step": 5580 }, { "grad_norm": 0.9707412123680115, "learning_rate": 9.999051561003123e-05, "loss": 0.0922, "step": 5590 }, { "grad_norm": 0.7936347723007202, "learning_rate": 9.999019083618202e-05, "loss": 0.0988, "step": 5600 }, { "grad_norm": 0.7210625410079956, "learning_rate": 9.99898605954807e-05, "loss": 0.1151, "step": 5610 }, { "grad_norm": 0.742637574672699, "learning_rate": 9.998952488796338e-05, "loss": 0.0972, "step": 5620 }, { "grad_norm": 0.586972177028656, "learning_rate": 9.998918371366676e-05, "loss": 0.095, "step": 5630 }, { "grad_norm": 0.6678465604782104, "learning_rate": 9.99888370726282e-05, "loss": 0.0978, "step": 5640 }, { "grad_norm": 0.5431642532348633, "learning_rate": 9.998848496488556e-05, "loss": 0.1197, "step": 5650 }, { "grad_norm": 0.639735221862793, "learning_rate": 9.998812739047736e-05, "loss": 0.0944, "step": 5660 }, { "grad_norm": 0.7479398250579834, "learning_rate": 9.99877643494427e-05, "loss": 0.1061, "step": 5670 }, { "grad_norm": 1.1641770601272583, "learning_rate": 9.998739584182128e-05, "loss": 0.0985, "step": 5680 }, { "grad_norm": 0.7182254195213318, "learning_rate": 9.998702186765342e-05, "loss": 0.1181, "step": 5690 }, { "grad_norm": 0.6322213411331177, "learning_rate": 9.998664242698e-05, "loss": 0.1048, "step": 5700 }, { "grad_norm": 0.5225303173065186, "learning_rate": 9.998625751984251e-05, "loss": 0.0979, "step": 5710 }, { "grad_norm": 0.6207631230354309, "learning_rate": 9.998586714628307e-05, "loss": 0.0929, "step": 5720 }, { "grad_norm": 0.5886924862861633, "learning_rate": 9.998547130634432e-05, "loss": 0.0929, "step": 5730 }, { "grad_norm": 0.6215883493423462, "learning_rate": 9.99850700000696e-05, "loss": 0.1145, "step": 5740 }, { "grad_norm": 0.6156638264656067, "learning_rate": 9.998466322750278e-05, "loss": 0.0896, "step": 5750 }, { "grad_norm": 0.5959202647209167, "learning_rate": 9.998425098868834e-05, "loss": 0.1076, "step": 5760 }, { "grad_norm": 0.597223699092865, "learning_rate": 9.998383328367136e-05, "loss": 0.0881, "step": 5770 }, { "grad_norm": 0.8025229573249817, "learning_rate": 9.99834101124975e-05, "loss": 0.1089, "step": 5780 }, { "grad_norm": 0.6464446187019348, "learning_rate": 9.998298147521309e-05, "loss": 0.0869, "step": 5790 }, { "grad_norm": 0.7543665766716003, "learning_rate": 9.998254737186496e-05, "loss": 0.11, "step": 5800 }, { "grad_norm": 0.734991729259491, "learning_rate": 9.99821078025006e-05, "loss": 0.0998, "step": 5810 }, { "grad_norm": 0.5860126614570618, "learning_rate": 9.998166276716807e-05, "loss": 0.1047, "step": 5820 }, { "grad_norm": 0.6159362196922302, "learning_rate": 9.998121226591606e-05, "loss": 0.0986, "step": 5830 }, { "grad_norm": 0.6159380078315735, "learning_rate": 9.998075629879382e-05, "loss": 0.1025, "step": 5840 }, { "grad_norm": 0.5794269442558289, "learning_rate": 9.99802948658512e-05, "loss": 0.0863, "step": 5850 }, { "grad_norm": 0.6017411351203918, "learning_rate": 9.99798279671387e-05, "loss": 0.1015, "step": 5860 }, { "grad_norm": 0.5499852895736694, "learning_rate": 9.997935560270734e-05, "loss": 0.0998, "step": 5870 }, { "grad_norm": 0.5946820378303528, "learning_rate": 9.997887777260879e-05, "loss": 0.1007, "step": 5880 }, { "grad_norm": 0.548430860042572, "learning_rate": 9.997839447689532e-05, "loss": 0.0919, "step": 5890 }, { "grad_norm": 0.6280621290206909, "learning_rate": 9.997790571561978e-05, "loss": 0.1133, "step": 5900 }, { "grad_norm": 0.44287627935409546, "learning_rate": 9.99774114888356e-05, "loss": 0.1066, "step": 5910 }, { "grad_norm": 0.7303943037986755, "learning_rate": 9.997691179659684e-05, "loss": 0.0885, "step": 5920 }, { "grad_norm": 0.5182363986968994, "learning_rate": 9.997640663895815e-05, "loss": 0.0915, "step": 5930 }, { "grad_norm": 0.6001327633857727, "learning_rate": 9.997589601597477e-05, "loss": 0.0895, "step": 5940 }, { "grad_norm": 0.5593392848968506, "learning_rate": 9.997537992770252e-05, "loss": 0.0919, "step": 5950 }, { "grad_norm": 1.1143929958343506, "learning_rate": 9.997485837419788e-05, "loss": 0.116, "step": 5960 }, { "grad_norm": 0.9431538581848145, "learning_rate": 9.997433135551786e-05, "loss": 0.1135, "step": 5970 }, { "grad_norm": 0.6709555387496948, "learning_rate": 9.997379887172009e-05, "loss": 0.1064, "step": 5980 }, { "grad_norm": 0.7475752830505371, "learning_rate": 9.997326092286281e-05, "loss": 0.1227, "step": 5990 }, { "grad_norm": 0.5449286103248596, "learning_rate": 9.997271750900486e-05, "loss": 0.1186, "step": 6000 }, { "grad_norm": 0.5808519124984741, "learning_rate": 9.997216863020565e-05, "loss": 0.097, "step": 6010 }, { "grad_norm": 0.6939573884010315, "learning_rate": 9.99716142865252e-05, "loss": 0.1192, "step": 6020 }, { "grad_norm": 0.6871560215950012, "learning_rate": 9.997105447802415e-05, "loss": 0.1021, "step": 6030 }, { "grad_norm": 0.7094020843505859, "learning_rate": 9.997048920476373e-05, "loss": 0.1133, "step": 6040 }, { "grad_norm": 0.6990440487861633, "learning_rate": 9.996991846680572e-05, "loss": 0.1109, "step": 6050 }, { "grad_norm": 0.5395070910453796, "learning_rate": 9.996934226421257e-05, "loss": 0.0935, "step": 6060 }, { "grad_norm": 0.6392314434051514, "learning_rate": 9.996876059704726e-05, "loss": 0.0938, "step": 6070 }, { "grad_norm": 0.5976810455322266, "learning_rate": 9.996817346537343e-05, "loss": 0.0834, "step": 6080 }, { "grad_norm": 0.6286873817443848, "learning_rate": 9.996758086925526e-05, "loss": 0.1005, "step": 6090 }, { "grad_norm": 0.6210547089576721, "learning_rate": 9.996698280875759e-05, "loss": 0.1242, "step": 6100 }, { "grad_norm": 0.7389043569564819, "learning_rate": 9.99663792839458e-05, "loss": 0.0884, "step": 6110 }, { "grad_norm": 0.6324407458305359, "learning_rate": 9.99657702948859e-05, "loss": 0.0991, "step": 6120 }, { "grad_norm": 0.6176057457923889, "learning_rate": 9.996515584164448e-05, "loss": 0.0761, "step": 6130 }, { "grad_norm": 0.8182070851325989, "learning_rate": 9.996453592428873e-05, "loss": 0.0871, "step": 6140 }, { "grad_norm": 0.6313467025756836, "learning_rate": 9.996391054288646e-05, "loss": 0.1137, "step": 6150 }, { "grad_norm": 0.6655222773551941, "learning_rate": 9.996327969750605e-05, "loss": 0.1283, "step": 6160 }, { "grad_norm": 0.6398155093193054, "learning_rate": 9.996264338821649e-05, "loss": 0.1144, "step": 6170 }, { "grad_norm": 0.6637073159217834, "learning_rate": 9.996200161508735e-05, "loss": 0.0837, "step": 6180 }, { "grad_norm": 0.4959658682346344, "learning_rate": 9.996135437818885e-05, "loss": 0.1085, "step": 6190 }, { "grad_norm": 0.6073696613311768, "learning_rate": 9.996070167759175e-05, "loss": 0.1021, "step": 6200 }, { "grad_norm": 1.0741926431655884, "learning_rate": 9.996004351336743e-05, "loss": 0.0999, "step": 6210 }, { "grad_norm": 0.9464169144630432, "learning_rate": 9.995937988558785e-05, "loss": 0.0967, "step": 6220 }, { "grad_norm": 0.7895137071609497, "learning_rate": 9.995871079432561e-05, "loss": 0.1053, "step": 6230 }, { "grad_norm": 0.6032257676124573, "learning_rate": 9.995803623965389e-05, "loss": 0.0914, "step": 6240 }, { "grad_norm": 0.5982276797294617, "learning_rate": 9.995735622164641e-05, "loss": 0.0939, "step": 6250 }, { "grad_norm": 0.6928361058235168, "learning_rate": 9.995667074037758e-05, "loss": 0.0814, "step": 6260 }, { "grad_norm": 0.5313943028450012, "learning_rate": 9.995597979592232e-05, "loss": 0.088, "step": 6270 }, { "grad_norm": 0.5755606293678284, "learning_rate": 9.995528338835625e-05, "loss": 0.0919, "step": 6280 }, { "grad_norm": 0.5477207899093628, "learning_rate": 9.995458151775547e-05, "loss": 0.0944, "step": 6290 }, { "grad_norm": 0.6281101703643799, "learning_rate": 9.995387418419677e-05, "loss": 0.0921, "step": 6300 }, { "grad_norm": 0.6162277460098267, "learning_rate": 9.99531613877575e-05, "loss": 0.0792, "step": 6310 }, { "grad_norm": 0.6463935375213623, "learning_rate": 9.995244312851559e-05, "loss": 0.0863, "step": 6320 }, { "grad_norm": 0.8800878524780273, "learning_rate": 9.995171940654961e-05, "loss": 0.0906, "step": 6330 }, { "grad_norm": 0.7124094367027283, "learning_rate": 9.995099022193871e-05, "loss": 0.1001, "step": 6340 }, { "grad_norm": 0.6971118450164795, "learning_rate": 9.995025557476261e-05, "loss": 0.0958, "step": 6350 }, { "grad_norm": 0.859200656414032, "learning_rate": 9.994951546510165e-05, "loss": 0.1038, "step": 6360 }, { "grad_norm": 0.8498514890670776, "learning_rate": 9.994876989303679e-05, "loss": 0.1047, "step": 6370 }, { "grad_norm": 0.6796415448188782, "learning_rate": 9.994801885864955e-05, "loss": 0.0904, "step": 6380 }, { "grad_norm": 0.7910406589508057, "learning_rate": 9.994726236202205e-05, "loss": 0.1076, "step": 6390 }, { "grad_norm": 0.5648146271705627, "learning_rate": 9.994650040323704e-05, "loss": 0.1038, "step": 6400 }, { "grad_norm": 0.6035485863685608, "learning_rate": 9.994573298237784e-05, "loss": 0.0856, "step": 6410 }, { "grad_norm": 0.767261803150177, "learning_rate": 9.994496009952837e-05, "loss": 0.0992, "step": 6420 }, { "grad_norm": 0.5865898132324219, "learning_rate": 9.994418175477316e-05, "loss": 0.0869, "step": 6430 }, { "grad_norm": 0.6001040935516357, "learning_rate": 9.994339794819733e-05, "loss": 0.0925, "step": 6440 }, { "grad_norm": 0.6108767986297607, "learning_rate": 9.994260867988658e-05, "loss": 0.09, "step": 6450 }, { "grad_norm": 0.7225428819656372, "learning_rate": 9.994181394992723e-05, "loss": 0.0882, "step": 6460 }, { "grad_norm": 0.6886321306228638, "learning_rate": 9.994101375840618e-05, "loss": 0.0868, "step": 6470 }, { "grad_norm": 0.8723170757293701, "learning_rate": 9.994020810541098e-05, "loss": 0.087, "step": 6480 }, { "grad_norm": 0.7053617835044861, "learning_rate": 9.99393969910297e-05, "loss": 0.108, "step": 6490 }, { "grad_norm": 0.5631186962127686, "learning_rate": 9.993858041535104e-05, "loss": 0.0926, "step": 6500 }, { "grad_norm": 0.7319157123565674, "learning_rate": 9.99377583784643e-05, "loss": 0.0835, "step": 6510 }, { "grad_norm": 0.7204568982124329, "learning_rate": 9.993693088045939e-05, "loss": 0.0816, "step": 6520 }, { "grad_norm": 0.509221613407135, "learning_rate": 9.99360979214268e-05, "loss": 0.0805, "step": 6530 }, { "grad_norm": 0.628598690032959, "learning_rate": 9.99352595014576e-05, "loss": 0.0995, "step": 6540 }, { "grad_norm": 0.7188753485679626, "learning_rate": 9.993441562064354e-05, "loss": 0.1, "step": 6550 }, { "grad_norm": 0.6015487313270569, "learning_rate": 9.993356627907685e-05, "loss": 0.0972, "step": 6560 }, { "grad_norm": 0.5774953365325928, "learning_rate": 9.99327114768504e-05, "loss": 0.0875, "step": 6570 }, { "grad_norm": 0.7213730216026306, "learning_rate": 9.99318512140577e-05, "loss": 0.0774, "step": 6580 }, { "grad_norm": 0.5342260599136353, "learning_rate": 9.993098549079284e-05, "loss": 0.1138, "step": 6590 }, { "grad_norm": 0.5593581199645996, "learning_rate": 9.993011430715047e-05, "loss": 0.1096, "step": 6600 }, { "grad_norm": 0.6003103256225586, "learning_rate": 9.992923766322586e-05, "loss": 0.0741, "step": 6610 }, { "grad_norm": 0.49705690145492554, "learning_rate": 9.99283555591149e-05, "loss": 0.0808, "step": 6620 }, { "grad_norm": 0.5781526565551758, "learning_rate": 9.992746799491404e-05, "loss": 0.0782, "step": 6630 }, { "grad_norm": 0.6301517486572266, "learning_rate": 9.992657497072033e-05, "loss": 0.0774, "step": 6640 }, { "grad_norm": 0.6145487427711487, "learning_rate": 9.992567648663147e-05, "loss": 0.1046, "step": 6650 }, { "grad_norm": 0.5142645835876465, "learning_rate": 9.992477254274568e-05, "loss": 0.0764, "step": 6660 }, { "grad_norm": 0.8233643174171448, "learning_rate": 9.992386313916183e-05, "loss": 0.0809, "step": 6670 }, { "grad_norm": 0.5860337615013123, "learning_rate": 9.992294827597934e-05, "loss": 0.0891, "step": 6680 }, { "grad_norm": 0.695533275604248, "learning_rate": 9.992202795329831e-05, "loss": 0.0831, "step": 6690 }, { "grad_norm": 1.079714298248291, "learning_rate": 9.992110217121936e-05, "loss": 0.1015, "step": 6700 }, { "grad_norm": 0.7836354374885559, "learning_rate": 9.992017092984372e-05, "loss": 0.0969, "step": 6710 }, { "grad_norm": 0.7782709002494812, "learning_rate": 9.991923422927326e-05, "loss": 0.1036, "step": 6720 }, { "grad_norm": 0.6841916441917419, "learning_rate": 9.991829206961037e-05, "loss": 0.1064, "step": 6730 }, { "grad_norm": 0.6787856817245483, "learning_rate": 9.991734445095813e-05, "loss": 0.1196, "step": 6740 }, { "grad_norm": 0.5861249566078186, "learning_rate": 9.991639137342015e-05, "loss": 0.0888, "step": 6750 }, { "grad_norm": 0.7220519185066223, "learning_rate": 9.991543283710064e-05, "loss": 0.11, "step": 6760 }, { "grad_norm": 0.7159816026687622, "learning_rate": 9.991446884210445e-05, "loss": 0.0883, "step": 6770 }, { "grad_norm": 0.7197673320770264, "learning_rate": 9.9913499388537e-05, "loss": 0.0971, "step": 6780 }, { "grad_norm": 0.5889588594436646, "learning_rate": 9.99125244765043e-05, "loss": 0.0905, "step": 6790 }, { "grad_norm": 0.526380181312561, "learning_rate": 9.991154410611296e-05, "loss": 0.0969, "step": 6800 }, { "grad_norm": 0.9491122364997864, "learning_rate": 9.99105582774702e-05, "loss": 0.0891, "step": 6810 }, { "grad_norm": 0.6235520243644714, "learning_rate": 9.990956699068384e-05, "loss": 0.1025, "step": 6820 }, { "grad_norm": 0.5501372814178467, "learning_rate": 9.990857024586224e-05, "loss": 0.0929, "step": 6830 }, { "grad_norm": 0.587453305721283, "learning_rate": 9.990756804311446e-05, "loss": 0.1036, "step": 6840 }, { "grad_norm": 0.5613220930099487, "learning_rate": 9.990656038255006e-05, "loss": 0.0862, "step": 6850 }, { "grad_norm": 0.5524001717567444, "learning_rate": 9.990554726427926e-05, "loss": 0.0879, "step": 6860 }, { "grad_norm": 0.6691156029701233, "learning_rate": 9.990452868841284e-05, "loss": 0.106, "step": 6870 }, { "grad_norm": 0.7269251346588135, "learning_rate": 9.99035046550622e-05, "loss": 0.1025, "step": 6880 }, { "grad_norm": 0.640803873538971, "learning_rate": 9.99024751643393e-05, "loss": 0.0892, "step": 6890 }, { "grad_norm": 0.5964599251747131, "learning_rate": 9.990144021635677e-05, "loss": 0.0806, "step": 6900 }, { "grad_norm": 0.5529967546463013, "learning_rate": 9.990039981122775e-05, "loss": 0.0814, "step": 6910 }, { "grad_norm": 0.5681519508361816, "learning_rate": 9.989935394906602e-05, "loss": 0.0835, "step": 6920 }, { "grad_norm": 0.5552298426628113, "learning_rate": 9.989830262998598e-05, "loss": 0.0949, "step": 6930 }, { "grad_norm": 0.5759883522987366, "learning_rate": 9.989724585410259e-05, "loss": 0.0731, "step": 6940 }, { "grad_norm": 0.6001555323600769, "learning_rate": 9.989618362153139e-05, "loss": 0.0869, "step": 6950 }, { "grad_norm": 0.6103785037994385, "learning_rate": 9.989511593238859e-05, "loss": 0.0889, "step": 6960 }, { "grad_norm": 0.5640770196914673, "learning_rate": 9.98940427867909e-05, "loss": 0.0925, "step": 6970 }, { "grad_norm": 1.1343307495117188, "learning_rate": 9.989296418485573e-05, "loss": 0.0854, "step": 6980 }, { "grad_norm": 0.7241509556770325, "learning_rate": 9.989188012670101e-05, "loss": 0.0861, "step": 6990 }, { "grad_norm": 0.835153341293335, "learning_rate": 9.989079061244528e-05, "loss": 0.1048, "step": 7000 }, { "grad_norm": 0.5046650767326355, "learning_rate": 9.988969564220769e-05, "loss": 0.1152, "step": 7010 }, { "grad_norm": 0.6019567251205444, "learning_rate": 9.988859521610801e-05, "loss": 0.0971, "step": 7020 }, { "grad_norm": 0.709644079208374, "learning_rate": 9.988748933426656e-05, "loss": 0.1157, "step": 7030 }, { "grad_norm": 0.706613302230835, "learning_rate": 9.988637799680428e-05, "loss": 0.0966, "step": 7040 }, { "grad_norm": 0.6303542852401733, "learning_rate": 9.98852612038427e-05, "loss": 0.0927, "step": 7050 }, { "grad_norm": 0.5849530100822449, "learning_rate": 9.988413895550397e-05, "loss": 0.0895, "step": 7060 }, { "grad_norm": 0.6105577945709229, "learning_rate": 9.98830112519108e-05, "loss": 0.0845, "step": 7070 }, { "grad_norm": 0.552455723285675, "learning_rate": 9.98818780931865e-05, "loss": 0.1071, "step": 7080 }, { "grad_norm": 0.732758641242981, "learning_rate": 9.988073947945502e-05, "loss": 0.1002, "step": 7090 }, { "grad_norm": 0.8039178252220154, "learning_rate": 9.987959541084087e-05, "loss": 0.087, "step": 7100 }, { "grad_norm": 0.5559234023094177, "learning_rate": 9.987844588746915e-05, "loss": 0.0975, "step": 7110 }, { "grad_norm": 0.5532280206680298, "learning_rate": 9.987729090946558e-05, "loss": 0.0816, "step": 7120 }, { "grad_norm": 0.5052943825721741, "learning_rate": 9.987613047695647e-05, "loss": 0.0858, "step": 7130 }, { "grad_norm": 0.5632840991020203, "learning_rate": 9.987496459006871e-05, "loss": 0.1014, "step": 7140 }, { "grad_norm": 0.6094858050346375, "learning_rate": 9.987379324892982e-05, "loss": 0.0869, "step": 7150 }, { "grad_norm": 0.8228130340576172, "learning_rate": 9.987261645366788e-05, "loss": 0.0998, "step": 7160 }, { "grad_norm": 0.5983760356903076, "learning_rate": 9.987143420441158e-05, "loss": 0.1014, "step": 7170 }, { "grad_norm": 0.4929274916648865, "learning_rate": 9.987024650129022e-05, "loss": 0.0851, "step": 7180 }, { "grad_norm": 0.5898838043212891, "learning_rate": 9.986905334443368e-05, "loss": 0.099, "step": 7190 }, { "grad_norm": 0.6748920679092407, "learning_rate": 9.986785473397245e-05, "loss": 0.0849, "step": 7200 }, { "grad_norm": 0.6831320524215698, "learning_rate": 9.98666506700376e-05, "loss": 0.1116, "step": 7210 }, { "grad_norm": 0.6684938073158264, "learning_rate": 9.986544115276081e-05, "loss": 0.0857, "step": 7220 }, { "grad_norm": 0.6268082857131958, "learning_rate": 9.986422618227433e-05, "loss": 0.0955, "step": 7230 }, { "grad_norm": 0.8245791792869568, "learning_rate": 9.986300575871106e-05, "loss": 0.0923, "step": 7240 }, { "grad_norm": 0.717117428779602, "learning_rate": 9.986177988220444e-05, "loss": 0.1026, "step": 7250 }, { "grad_norm": 0.6642735600471497, "learning_rate": 9.986054855288856e-05, "loss": 0.0951, "step": 7260 }, { "grad_norm": 0.6476512551307678, "learning_rate": 9.985931177089802e-05, "loss": 0.0932, "step": 7270 }, { "grad_norm": 0.5781366229057312, "learning_rate": 9.985806953636814e-05, "loss": 0.0837, "step": 7280 }, { "grad_norm": 0.7777464985847473, "learning_rate": 9.985682184943471e-05, "loss": 0.0923, "step": 7290 }, { "grad_norm": 0.4530040919780731, "learning_rate": 9.98555687102342e-05, "loss": 0.0877, "step": 7300 }, { "grad_norm": 0.7167714834213257, "learning_rate": 9.985431011890367e-05, "loss": 0.0963, "step": 7310 }, { "grad_norm": 0.5759075284004211, "learning_rate": 9.985304607558075e-05, "loss": 0.075, "step": 7320 }, { "grad_norm": 0.4501078426837921, "learning_rate": 9.985177658040364e-05, "loss": 0.0767, "step": 7330 }, { "grad_norm": 0.5518580675125122, "learning_rate": 9.985050163351119e-05, "loss": 0.0926, "step": 7340 }, { "grad_norm": 0.7057280540466309, "learning_rate": 9.984922123504286e-05, "loss": 0.0844, "step": 7350 }, { "grad_norm": 0.5221959948539734, "learning_rate": 9.984793538513862e-05, "loss": 0.0804, "step": 7360 }, { "grad_norm": 0.5746706128120422, "learning_rate": 9.984664408393912e-05, "loss": 0.0879, "step": 7370 }, { "grad_norm": 0.5351296067237854, "learning_rate": 9.984534733158556e-05, "loss": 0.0846, "step": 7380 }, { "grad_norm": 0.5552535653114319, "learning_rate": 9.984404512821977e-05, "loss": 0.084, "step": 7390 }, { "grad_norm": 0.6806361079216003, "learning_rate": 9.984273747398411e-05, "loss": 0.0948, "step": 7400 }, { "grad_norm": 0.6821239590644836, "learning_rate": 9.984142436902165e-05, "loss": 0.0954, "step": 7410 }, { "grad_norm": 0.7710086703300476, "learning_rate": 9.984010581347596e-05, "loss": 0.1114, "step": 7420 }, { "grad_norm": 0.7042375206947327, "learning_rate": 9.983878180749121e-05, "loss": 0.1168, "step": 7430 }, { "grad_norm": 0.5244135856628418, "learning_rate": 9.983745235121222e-05, "loss": 0.0986, "step": 7440 }, { "grad_norm": 0.4888067841529846, "learning_rate": 9.983611744478438e-05, "loss": 0.0937, "step": 7450 }, { "grad_norm": 0.43464556336402893, "learning_rate": 9.983477708835365e-05, "loss": 0.0906, "step": 7460 }, { "grad_norm": 0.7696208357810974, "learning_rate": 9.983343128206664e-05, "loss": 0.1063, "step": 7470 }, { "grad_norm": 0.6803582906723022, "learning_rate": 9.983208002607049e-05, "loss": 0.0905, "step": 7480 }, { "grad_norm": 0.69597989320755, "learning_rate": 9.9830723320513e-05, "loss": 0.0912, "step": 7490 }, { "grad_norm": 0.653900682926178, "learning_rate": 9.982936116554254e-05, "loss": 0.0795, "step": 7500 }, { "grad_norm": 1.1568065881729126, "learning_rate": 9.982799356130803e-05, "loss": 0.0821, "step": 7510 }, { "grad_norm": 0.45966944098472595, "learning_rate": 9.982662050795908e-05, "loss": 0.0781, "step": 7520 }, { "grad_norm": 0.4088380038738251, "learning_rate": 9.982524200564583e-05, "loss": 0.0709, "step": 7530 }, { "grad_norm": 0.5653975009918213, "learning_rate": 9.982385805451901e-05, "loss": 0.098, "step": 7540 }, { "grad_norm": 0.5648491382598877, "learning_rate": 9.982246865472998e-05, "loss": 0.1051, "step": 7550 }, { "grad_norm": 0.5093026161193848, "learning_rate": 9.982107380643069e-05, "loss": 0.0861, "step": 7560 }, { "grad_norm": 0.5383346676826477, "learning_rate": 9.981967350977368e-05, "loss": 0.099, "step": 7570 }, { "grad_norm": 0.5995389819145203, "learning_rate": 9.981826776491208e-05, "loss": 0.0919, "step": 7580 }, { "grad_norm": 0.6238002181053162, "learning_rate": 9.98168565719996e-05, "loss": 0.0945, "step": 7590 }, { "grad_norm": 0.6278378367424011, "learning_rate": 9.98154399311906e-05, "loss": 0.0999, "step": 7600 }, { "grad_norm": 0.4794835150241852, "learning_rate": 9.981401784263997e-05, "loss": 0.0937, "step": 7610 }, { "grad_norm": 0.5265167355537415, "learning_rate": 9.981259030650326e-05, "loss": 0.0857, "step": 7620 }, { "grad_norm": 0.600752055644989, "learning_rate": 9.981115732293655e-05, "loss": 0.074, "step": 7630 }, { "grad_norm": 0.7200352549552917, "learning_rate": 9.980971889209659e-05, "loss": 0.0792, "step": 7640 }, { "grad_norm": 0.6049966812133789, "learning_rate": 9.980827501414064e-05, "loss": 0.0731, "step": 7650 }, { "grad_norm": 0.7552555203437805, "learning_rate": 9.980682568922663e-05, "loss": 0.1033, "step": 7660 }, { "grad_norm": 0.49056515097618103, "learning_rate": 9.980537091751304e-05, "loss": 0.0896, "step": 7670 }, { "grad_norm": 0.5456610918045044, "learning_rate": 9.980391069915897e-05, "loss": 0.1145, "step": 7680 }, { "grad_norm": 0.647594153881073, "learning_rate": 9.98024450343241e-05, "loss": 0.0778, "step": 7690 }, { "grad_norm": 0.4912182688713074, "learning_rate": 9.980097392316872e-05, "loss": 0.0688, "step": 7700 }, { "grad_norm": 0.6144121289253235, "learning_rate": 9.97994973658537e-05, "loss": 0.0909, "step": 7710 }, { "grad_norm": 0.5892572402954102, "learning_rate": 9.979801536254054e-05, "loss": 0.0974, "step": 7720 }, { "grad_norm": 0.5770010948181152, "learning_rate": 9.979652791339127e-05, "loss": 0.0959, "step": 7730 }, { "grad_norm": 0.8433646559715271, "learning_rate": 9.97950350185686e-05, "loss": 0.0872, "step": 7740 }, { "grad_norm": 0.5716922879219055, "learning_rate": 9.979353667823574e-05, "loss": 0.0786, "step": 7750 }, { "grad_norm": 0.5058563351631165, "learning_rate": 9.979203289255658e-05, "loss": 0.0883, "step": 7760 }, { "grad_norm": 0.5043525099754333, "learning_rate": 9.979052366169557e-05, "loss": 0.0773, "step": 7770 }, { "grad_norm": 0.5974773168563843, "learning_rate": 9.978900898581775e-05, "loss": 0.0848, "step": 7780 }, { "grad_norm": 0.6181852221488953, "learning_rate": 9.978748886508875e-05, "loss": 0.08, "step": 7790 }, { "grad_norm": 0.6315383911132812, "learning_rate": 9.978596329967484e-05, "loss": 0.091, "step": 7800 }, { "grad_norm": 0.5356779098510742, "learning_rate": 9.978443228974284e-05, "loss": 0.0844, "step": 7810 }, { "grad_norm": 0.5648067593574524, "learning_rate": 9.978289583546015e-05, "loss": 0.0747, "step": 7820 }, { "grad_norm": 0.5941452383995056, "learning_rate": 9.978135393699484e-05, "loss": 0.0815, "step": 7830 }, { "grad_norm": 0.8585506081581116, "learning_rate": 9.977980659451548e-05, "loss": 0.0778, "step": 7840 }, { "grad_norm": 0.7565423250198364, "learning_rate": 9.977825380819135e-05, "loss": 0.0833, "step": 7850 }, { "grad_norm": 0.7215997576713562, "learning_rate": 9.97766955781922e-05, "loss": 0.0907, "step": 7860 }, { "grad_norm": 0.5297353863716125, "learning_rate": 9.977513190468848e-05, "loss": 0.094, "step": 7870 }, { "grad_norm": 0.47856584191322327, "learning_rate": 9.977356278785116e-05, "loss": 0.0994, "step": 7880 }, { "grad_norm": 0.5190351605415344, "learning_rate": 9.977198822785184e-05, "loss": 0.0781, "step": 7890 }, { "grad_norm": 0.6097343564033508, "learning_rate": 9.977040822486273e-05, "loss": 0.1003, "step": 7900 }, { "grad_norm": 0.4584157168865204, "learning_rate": 9.97688227790566e-05, "loss": 0.0955, "step": 7910 }, { "grad_norm": 0.5704355239868164, "learning_rate": 9.976723189060684e-05, "loss": 0.1214, "step": 7920 }, { "grad_norm": 0.7882084250450134, "learning_rate": 9.976563555968742e-05, "loss": 0.0971, "step": 7930 }, { "grad_norm": 0.6104371547698975, "learning_rate": 9.976403378647292e-05, "loss": 0.0999, "step": 7940 }, { "grad_norm": 0.6380048990249634, "learning_rate": 9.97624265711385e-05, "loss": 0.0817, "step": 7950 }, { "grad_norm": 0.8568858504295349, "learning_rate": 9.976081391385993e-05, "loss": 0.0973, "step": 7960 }, { "grad_norm": 0.6410380005836487, "learning_rate": 9.975919581481356e-05, "loss": 0.098, "step": 7970 }, { "grad_norm": 0.507700502872467, "learning_rate": 9.975757227417634e-05, "loss": 0.0996, "step": 7980 }, { "grad_norm": 0.5615932941436768, "learning_rate": 9.975594329212586e-05, "loss": 0.0867, "step": 7990 }, { "grad_norm": 0.5736196041107178, "learning_rate": 9.97543088688402e-05, "loss": 0.0941, "step": 8000 }, { "grad_norm": 0.49154600501060486, "learning_rate": 9.975266900449814e-05, "loss": 0.0744, "step": 8010 }, { "grad_norm": 0.8038212060928345, "learning_rate": 9.975102369927898e-05, "loss": 0.0805, "step": 8020 }, { "grad_norm": 0.5575769543647766, "learning_rate": 9.974937295336269e-05, "loss": 0.0724, "step": 8030 }, { "grad_norm": 0.614460825920105, "learning_rate": 9.974771676692975e-05, "loss": 0.0887, "step": 8040 }, { "grad_norm": 0.5546360015869141, "learning_rate": 9.974605514016131e-05, "loss": 0.0806, "step": 8050 }, { "grad_norm": 0.5371960997581482, "learning_rate": 9.974438807323907e-05, "loss": 0.0867, "step": 8060 }, { "grad_norm": 0.5367050766944885, "learning_rate": 9.974271556634535e-05, "loss": 0.0869, "step": 8070 }, { "grad_norm": 0.818469226360321, "learning_rate": 9.974103761966302e-05, "loss": 0.1104, "step": 8080 }, { "grad_norm": 0.5624365210533142, "learning_rate": 9.973935423337563e-05, "loss": 0.0964, "step": 8090 }, { "grad_norm": 0.6339496374130249, "learning_rate": 9.973766540766722e-05, "loss": 0.096, "step": 8100 }, { "grad_norm": 0.7220972180366516, "learning_rate": 9.97359711427225e-05, "loss": 0.1021, "step": 8110 }, { "grad_norm": 0.7198814749717712, "learning_rate": 9.973427143872677e-05, "loss": 0.0999, "step": 8120 }, { "grad_norm": 0.9616439342498779, "learning_rate": 9.973256629586589e-05, "loss": 0.1217, "step": 8130 }, { "grad_norm": 0.8153815269470215, "learning_rate": 9.973085571432632e-05, "loss": 0.0933, "step": 8140 }, { "grad_norm": 0.6969852447509766, "learning_rate": 9.972913969429513e-05, "loss": 0.0842, "step": 8150 }, { "grad_norm": 0.4862261712551117, "learning_rate": 9.972741823596e-05, "loss": 0.0938, "step": 8160 }, { "grad_norm": 0.6517384648323059, "learning_rate": 9.972569133950917e-05, "loss": 0.078, "step": 8170 }, { "grad_norm": 0.550464928150177, "learning_rate": 9.972395900513151e-05, "loss": 0.0745, "step": 8180 }, { "grad_norm": 0.6303491592407227, "learning_rate": 9.972222123301645e-05, "loss": 0.0814, "step": 8190 }, { "grad_norm": 0.5339958071708679, "learning_rate": 9.972047802335403e-05, "loss": 0.0846, "step": 8200 }, { "grad_norm": 0.43467599153518677, "learning_rate": 9.971872937633488e-05, "loss": 0.0708, "step": 8210 }, { "grad_norm": 1.2276984453201294, "learning_rate": 9.971697529215024e-05, "loss": 0.0833, "step": 8220 }, { "grad_norm": 0.5505781769752502, "learning_rate": 9.971521577099192e-05, "loss": 0.0883, "step": 8230 }, { "grad_norm": 0.5844340920448303, "learning_rate": 9.971345081305236e-05, "loss": 0.0832, "step": 8240 }, { "grad_norm": 0.5321600437164307, "learning_rate": 9.971168041852456e-05, "loss": 0.083, "step": 8250 }, { "grad_norm": 0.5813252925872803, "learning_rate": 9.970990458760215e-05, "loss": 0.0804, "step": 8260 }, { "grad_norm": 0.5529824495315552, "learning_rate": 9.970812332047929e-05, "loss": 0.0761, "step": 8270 }, { "grad_norm": 0.46416765451431274, "learning_rate": 9.97063366173508e-05, "loss": 0.0782, "step": 8280 }, { "grad_norm": 0.583499550819397, "learning_rate": 9.970454447841207e-05, "loss": 0.0767, "step": 8290 }, { "grad_norm": 0.4889448285102844, "learning_rate": 9.970274690385909e-05, "loss": 0.0864, "step": 8300 }, { "grad_norm": 0.6674699783325195, "learning_rate": 9.970094389388844e-05, "loss": 0.0806, "step": 8310 }, { "grad_norm": 0.5779807567596436, "learning_rate": 9.969913544869728e-05, "loss": 0.068, "step": 8320 }, { "grad_norm": 0.708966076374054, "learning_rate": 9.96973215684834e-05, "loss": 0.0979, "step": 8330 }, { "grad_norm": 0.5686944127082825, "learning_rate": 9.969550225344513e-05, "loss": 0.0885, "step": 8340 }, { "grad_norm": 0.8033183217048645, "learning_rate": 9.969367750378147e-05, "loss": 0.0878, "step": 8350 }, { "grad_norm": 0.6722063422203064, "learning_rate": 9.969184731969194e-05, "loss": 0.0904, "step": 8360 }, { "grad_norm": 0.5248117446899414, "learning_rate": 9.96900117013767e-05, "loss": 0.0929, "step": 8370 }, { "grad_norm": 0.7547135949134827, "learning_rate": 9.96881706490365e-05, "loss": 0.0842, "step": 8380 }, { "grad_norm": 0.7114159464836121, "learning_rate": 9.968632416287265e-05, "loss": 0.0874, "step": 8390 }, { "grad_norm": 0.5954317450523376, "learning_rate": 9.96844722430871e-05, "loss": 0.0887, "step": 8400 }, { "grad_norm": 0.6620914340019226, "learning_rate": 9.968261488988235e-05, "loss": 0.0781, "step": 8410 }, { "grad_norm": 0.5511741042137146, "learning_rate": 9.968075210346155e-05, "loss": 0.0723, "step": 8420 }, { "grad_norm": 0.7256964445114136, "learning_rate": 9.967888388402839e-05, "loss": 0.0919, "step": 8430 }, { "grad_norm": 0.510750949382782, "learning_rate": 9.967701023178717e-05, "loss": 0.0759, "step": 8440 }, { "grad_norm": 0.5589015483856201, "learning_rate": 9.967513114694282e-05, "loss": 0.0781, "step": 8450 }, { "grad_norm": 0.6148762106895447, "learning_rate": 9.967324662970079e-05, "loss": 0.0933, "step": 8460 }, { "grad_norm": 0.5473476052284241, "learning_rate": 9.96713566802672e-05, "loss": 0.0794, "step": 8470 }, { "grad_norm": 0.4770444929599762, "learning_rate": 9.966946129884873e-05, "loss": 0.0829, "step": 8480 }, { "grad_norm": 0.6981711983680725, "learning_rate": 9.966756048565265e-05, "loss": 0.0869, "step": 8490 }, { "grad_norm": 0.7836324572563171, "learning_rate": 9.966565424088681e-05, "loss": 0.098, "step": 8500 }, { "grad_norm": 0.5626245737075806, "learning_rate": 9.96637425647597e-05, "loss": 0.0786, "step": 8510 }, { "grad_norm": 0.7556378245353699, "learning_rate": 9.966182545748038e-05, "loss": 0.0818, "step": 8520 }, { "grad_norm": 0.5757916569709778, "learning_rate": 9.96599029192585e-05, "loss": 0.0887, "step": 8530 }, { "grad_norm": 0.677787184715271, "learning_rate": 9.965797495030428e-05, "loss": 0.0809, "step": 8540 }, { "grad_norm": 0.5104233622550964, "learning_rate": 9.96560415508286e-05, "loss": 0.0835, "step": 8550 }, { "grad_norm": 0.5013442039489746, "learning_rate": 9.965410272104286e-05, "loss": 0.0783, "step": 8560 }, { "grad_norm": 0.6512135863304138, "learning_rate": 9.96521584611591e-05, "loss": 0.143, "step": 8570 }, { "grad_norm": 0.6820481419563293, "learning_rate": 9.965020877138994e-05, "loss": 0.0909, "step": 8580 }, { "grad_norm": 0.9625211358070374, "learning_rate": 9.964825365194861e-05, "loss": 0.107, "step": 8590 }, { "grad_norm": 0.6238473057746887, "learning_rate": 9.96462931030489e-05, "loss": 0.1042, "step": 8600 }, { "grad_norm": 0.5531347393989563, "learning_rate": 9.96443271249052e-05, "loss": 0.0869, "step": 8610 }, { "grad_norm": 0.5471090078353882, "learning_rate": 9.964235571773255e-05, "loss": 0.0801, "step": 8620 }, { "grad_norm": 0.5937244892120361, "learning_rate": 9.96403788817465e-05, "loss": 0.0727, "step": 8630 }, { "grad_norm": 0.5986392498016357, "learning_rate": 9.963839661716325e-05, "loss": 0.08, "step": 8640 }, { "grad_norm": 0.5171228051185608, "learning_rate": 9.963640892419958e-05, "loss": 0.08, "step": 8650 }, { "grad_norm": 0.7024669051170349, "learning_rate": 9.963441580307286e-05, "loss": 0.0831, "step": 8660 }, { "grad_norm": 0.4808242917060852, "learning_rate": 9.963241725400104e-05, "loss": 0.0904, "step": 8670 }, { "grad_norm": 0.5157133340835571, "learning_rate": 9.963041327720271e-05, "loss": 0.0663, "step": 8680 }, { "grad_norm": 0.4741113781929016, "learning_rate": 9.962840387289697e-05, "loss": 0.0741, "step": 8690 }, { "grad_norm": 0.6514261364936829, "learning_rate": 9.962638904130363e-05, "loss": 0.0828, "step": 8700 }, { "grad_norm": 0.4696279466152191, "learning_rate": 9.962436878264298e-05, "loss": 0.0946, "step": 8710 }, { "grad_norm": 0.6351137161254883, "learning_rate": 9.962234309713598e-05, "loss": 0.0984, "step": 8720 }, { "grad_norm": 0.5678004026412964, "learning_rate": 9.962031198500414e-05, "loss": 0.0772, "step": 8730 }, { "grad_norm": 0.4291149973869324, "learning_rate": 9.961827544646958e-05, "loss": 0.0737, "step": 8740 }, { "grad_norm": 0.6950181126594543, "learning_rate": 9.961623348175501e-05, "loss": 0.1015, "step": 8750 }, { "grad_norm": 0.7033677101135254, "learning_rate": 9.961418609108377e-05, "loss": 0.0844, "step": 8760 }, { "grad_norm": 0.558192789554596, "learning_rate": 9.961213327467971e-05, "loss": 0.0831, "step": 8770 }, { "grad_norm": 0.6803210973739624, "learning_rate": 9.961007503276736e-05, "loss": 0.0861, "step": 8780 }, { "grad_norm": 0.7480688691139221, "learning_rate": 9.960801136557179e-05, "loss": 0.079, "step": 8790 }, { "grad_norm": 0.6284939646720886, "learning_rate": 9.960594227331866e-05, "loss": 0.0722, "step": 8800 }, { "grad_norm": 0.5831637382507324, "learning_rate": 9.960386775623429e-05, "loss": 0.0901, "step": 8810 }, { "grad_norm": 0.45898547768592834, "learning_rate": 9.96017878145455e-05, "loss": 0.074, "step": 8820 }, { "grad_norm": 0.703059732913971, "learning_rate": 9.959970244847977e-05, "loss": 0.0728, "step": 8830 }, { "grad_norm": 0.6807403564453125, "learning_rate": 9.959761165826518e-05, "loss": 0.0762, "step": 8840 }, { "grad_norm": 1.0877429246902466, "learning_rate": 9.959551544413033e-05, "loss": 0.0831, "step": 8850 }, { "grad_norm": 0.5633538365364075, "learning_rate": 9.959341380630448e-05, "loss": 0.0831, "step": 8860 }, { "grad_norm": 0.6401535272598267, "learning_rate": 9.959130674501746e-05, "loss": 0.077, "step": 8870 }, { "grad_norm": 0.535465657711029, "learning_rate": 9.958919426049968e-05, "loss": 0.0837, "step": 8880 }, { "grad_norm": 0.5019620060920715, "learning_rate": 9.958707635298219e-05, "loss": 0.0773, "step": 8890 }, { "grad_norm": 0.5818244218826294, "learning_rate": 9.958495302269657e-05, "loss": 0.0849, "step": 8900 }, { "grad_norm": 0.5925021767616272, "learning_rate": 9.958282426987503e-05, "loss": 0.0861, "step": 8910 }, { "grad_norm": 0.7157803773880005, "learning_rate": 9.95806900947504e-05, "loss": 0.0948, "step": 8920 }, { "grad_norm": 0.7909169793128967, "learning_rate": 9.957855049755604e-05, "loss": 0.0752, "step": 8930 }, { "grad_norm": 0.5332646369934082, "learning_rate": 9.957640547852593e-05, "loss": 0.0861, "step": 8940 }, { "grad_norm": 0.46253207325935364, "learning_rate": 9.957425503789466e-05, "loss": 0.0773, "step": 8950 }, { "grad_norm": 0.532977283000946, "learning_rate": 9.957209917589738e-05, "loss": 0.0682, "step": 8960 }, { "grad_norm": 0.5377448201179504, "learning_rate": 9.956993789276987e-05, "loss": 0.0704, "step": 8970 }, { "grad_norm": 0.6011037230491638, "learning_rate": 9.956777118874847e-05, "loss": 0.0895, "step": 8980 }, { "grad_norm": 0.6733790636062622, "learning_rate": 9.956559906407016e-05, "loss": 0.096, "step": 8990 }, { "grad_norm": 0.7057199478149414, "learning_rate": 9.956342151897245e-05, "loss": 0.089, "step": 9000 }, { "grad_norm": 0.5894232392311096, "learning_rate": 9.956123855369346e-05, "loss": 0.078, "step": 9010 }, { "grad_norm": 0.7026933431625366, "learning_rate": 9.955905016847196e-05, "loss": 0.0777, "step": 9020 }, { "grad_norm": 0.581170916557312, "learning_rate": 9.955685636354723e-05, "loss": 0.084, "step": 9030 }, { "grad_norm": 0.6691320538520813, "learning_rate": 9.95546571391592e-05, "loss": 0.0871, "step": 9040 }, { "grad_norm": 0.7555967569351196, "learning_rate": 9.955245249554837e-05, "loss": 0.0682, "step": 9050 }, { "grad_norm": 0.6996241211891174, "learning_rate": 9.955024243295582e-05, "loss": 0.102, "step": 9060 }, { "grad_norm": 0.6168289184570312, "learning_rate": 9.954802695162328e-05, "loss": 0.0828, "step": 9070 }, { "grad_norm": 0.86152583360672, "learning_rate": 9.954580605179302e-05, "loss": 0.0847, "step": 9080 }, { "grad_norm": 0.47002631425857544, "learning_rate": 9.954357973370788e-05, "loss": 0.0917, "step": 9090 }, { "grad_norm": 0.6128401756286621, "learning_rate": 9.954134799761135e-05, "loss": 0.1117, "step": 9100 }, { "grad_norm": 0.803985595703125, "learning_rate": 9.953911084374748e-05, "loss": 0.0863, "step": 9110 }, { "grad_norm": 0.4331008791923523, "learning_rate": 9.953686827236093e-05, "loss": 0.0993, "step": 9120 }, { "grad_norm": 0.558117151260376, "learning_rate": 9.953462028369695e-05, "loss": 0.0847, "step": 9130 }, { "grad_norm": 0.4966691732406616, "learning_rate": 9.953236687800136e-05, "loss": 0.0762, "step": 9140 }, { "grad_norm": 0.6115354895591736, "learning_rate": 9.95301080555206e-05, "loss": 0.0902, "step": 9150 }, { "grad_norm": 0.5847378373146057, "learning_rate": 9.952784381650171e-05, "loss": 0.0821, "step": 9160 }, { "grad_norm": 0.5681011080741882, "learning_rate": 9.952557416119226e-05, "loss": 0.0766, "step": 9170 }, { "grad_norm": 0.7579455971717834, "learning_rate": 9.95232990898405e-05, "loss": 0.0778, "step": 9180 }, { "grad_norm": 0.3923374116420746, "learning_rate": 9.95210186026952e-05, "loss": 0.0714, "step": 9190 }, { "grad_norm": 0.5703105926513672, "learning_rate": 9.951873270000576e-05, "loss": 0.0823, "step": 9200 }, { "grad_norm": 0.5554647445678711, "learning_rate": 9.951644138202216e-05, "loss": 0.0994, "step": 9210 }, { "grad_norm": 0.5211426019668579, "learning_rate": 9.951414464899498e-05, "loss": 0.0884, "step": 9220 }, { "grad_norm": 0.6519138216972351, "learning_rate": 9.951184250117538e-05, "loss": 0.0899, "step": 9230 }, { "grad_norm": 0.54067462682724, "learning_rate": 9.950953493881513e-05, "loss": 0.0858, "step": 9240 }, { "grad_norm": 0.5821452140808105, "learning_rate": 9.950722196216658e-05, "loss": 0.0684, "step": 9250 }, { "grad_norm": 0.6799665689468384, "learning_rate": 9.950490357148265e-05, "loss": 0.0787, "step": 9260 }, { "grad_norm": 0.6534043550491333, "learning_rate": 9.950257976701692e-05, "loss": 0.0919, "step": 9270 }, { "grad_norm": 0.6404492259025574, "learning_rate": 9.950025054902348e-05, "loss": 0.085, "step": 9280 }, { "grad_norm": 0.7988382577896118, "learning_rate": 9.949791591775706e-05, "loss": 0.0806, "step": 9290 }, { "grad_norm": 0.5996073484420776, "learning_rate": 9.949557587347298e-05, "loss": 0.0789, "step": 9300 }, { "grad_norm": 0.5949200987815857, "learning_rate": 9.949323041642713e-05, "loss": 0.0749, "step": 9310 }, { "grad_norm": 0.5433605909347534, "learning_rate": 9.949087954687602e-05, "loss": 0.093, "step": 9320 }, { "grad_norm": 0.6784794926643372, "learning_rate": 9.948852326507672e-05, "loss": 0.0857, "step": 9330 }, { "grad_norm": 0.43489882349967957, "learning_rate": 9.948616157128694e-05, "loss": 0.0859, "step": 9340 }, { "grad_norm": 0.5138662457466125, "learning_rate": 9.948379446576493e-05, "loss": 0.0785, "step": 9350 }, { "grad_norm": 0.5579209327697754, "learning_rate": 9.948142194876952e-05, "loss": 0.0839, "step": 9360 }, { "grad_norm": 0.5782290697097778, "learning_rate": 9.947904402056024e-05, "loss": 0.0917, "step": 9370 }, { "grad_norm": 0.49143320322036743, "learning_rate": 9.947666068139708e-05, "loss": 0.079, "step": 9380 }, { "grad_norm": 0.5885505676269531, "learning_rate": 9.947427193154071e-05, "loss": 0.0745, "step": 9390 }, { "grad_norm": 0.7427430748939514, "learning_rate": 9.947187777125233e-05, "loss": 0.0972, "step": 9400 }, { "grad_norm": 0.5372757911682129, "learning_rate": 9.946947820079377e-05, "loss": 0.0833, "step": 9410 }, { "grad_norm": 0.6733894944190979, "learning_rate": 9.946707322042747e-05, "loss": 0.0851, "step": 9420 }, { "grad_norm": 0.5971024632453918, "learning_rate": 9.94646628304164e-05, "loss": 0.0755, "step": 9430 }, { "grad_norm": 0.5710270404815674, "learning_rate": 9.946224703102418e-05, "loss": 0.0836, "step": 9440 }, { "grad_norm": 0.8392489552497864, "learning_rate": 9.945982582251498e-05, "loss": 0.0704, "step": 9450 }, { "grad_norm": 0.7076278924942017, "learning_rate": 9.94573992051536e-05, "loss": 0.0734, "step": 9460 }, { "grad_norm": 1.3969097137451172, "learning_rate": 9.94549671792054e-05, "loss": 0.1092, "step": 9470 }, { "grad_norm": 0.5306630730628967, "learning_rate": 9.945252974493635e-05, "loss": 0.0804, "step": 9480 }, { "grad_norm": 0.4921726584434509, "learning_rate": 9.9450086902613e-05, "loss": 0.0844, "step": 9490 }, { "grad_norm": 0.5995213985443115, "learning_rate": 9.944763865250248e-05, "loss": 0.0748, "step": 9500 }, { "grad_norm": 0.8468683958053589, "learning_rate": 9.944518499487254e-05, "loss": 0.0798, "step": 9510 }, { "grad_norm": 0.6486026048660278, "learning_rate": 9.944272592999151e-05, "loss": 0.0895, "step": 9520 }, { "grad_norm": 0.8157832026481628, "learning_rate": 9.94402614581283e-05, "loss": 0.0881, "step": 9530 }, { "grad_norm": 0.7663639187812805, "learning_rate": 9.943779157955244e-05, "loss": 0.0877, "step": 9540 }, { "grad_norm": 0.5907610654830933, "learning_rate": 9.943531629453403e-05, "loss": 0.0798, "step": 9550 }, { "grad_norm": 0.5865486264228821, "learning_rate": 9.943283560334375e-05, "loss": 0.09, "step": 9560 }, { "grad_norm": 0.5820397734642029, "learning_rate": 9.943034950625288e-05, "loss": 0.1051, "step": 9570 }, { "grad_norm": 0.6307183504104614, "learning_rate": 9.942785800353332e-05, "loss": 0.0826, "step": 9580 }, { "grad_norm": 0.6053169369697571, "learning_rate": 9.942536109545751e-05, "loss": 0.0859, "step": 9590 }, { "grad_norm": 0.6043580770492554, "learning_rate": 9.942285878229853e-05, "loss": 0.0638, "step": 9600 }, { "grad_norm": 1.151889443397522, "learning_rate": 9.942035106433001e-05, "loss": 0.0908, "step": 9610 }, { "grad_norm": 0.7551982402801514, "learning_rate": 9.94178379418262e-05, "loss": 0.0957, "step": 9620 }, { "grad_norm": 0.5326666235923767, "learning_rate": 9.941531941506194e-05, "loss": 0.0846, "step": 9630 }, { "grad_norm": 0.5983622670173645, "learning_rate": 9.941279548431263e-05, "loss": 0.087, "step": 9640 }, { "grad_norm": 0.5477142333984375, "learning_rate": 9.941026614985431e-05, "loss": 0.0865, "step": 9650 }, { "grad_norm": 0.4234887361526489, "learning_rate": 9.940773141196357e-05, "loss": 0.0791, "step": 9660 }, { "grad_norm": 0.6023424863815308, "learning_rate": 9.94051912709176e-05, "loss": 0.0814, "step": 9670 }, { "grad_norm": 0.4916042387485504, "learning_rate": 9.940264572699421e-05, "loss": 0.0789, "step": 9680 }, { "grad_norm": 0.6060595512390137, "learning_rate": 9.940009478047174e-05, "loss": 0.0687, "step": 9690 }, { "grad_norm": 0.4880523681640625, "learning_rate": 9.939753843162918e-05, "loss": 0.069, "step": 9700 }, { "grad_norm": 0.614329993724823, "learning_rate": 9.939497668074609e-05, "loss": 0.0807, "step": 9710 }, { "grad_norm": 0.6824995279312134, "learning_rate": 9.93924095281026e-05, "loss": 0.0811, "step": 9720 }, { "grad_norm": 0.7096372842788696, "learning_rate": 9.938983697397948e-05, "loss": 0.085, "step": 9730 }, { "grad_norm": 0.5199359059333801, "learning_rate": 9.938725901865805e-05, "loss": 0.0747, "step": 9740 }, { "grad_norm": 0.645908534526825, "learning_rate": 9.93846756624202e-05, "loss": 0.0811, "step": 9750 }, { "grad_norm": 0.645989716053009, "learning_rate": 9.938208690554849e-05, "loss": 0.0739, "step": 9760 }, { "grad_norm": 0.5671721696853638, "learning_rate": 9.9379492748326e-05, "loss": 0.0769, "step": 9770 }, { "grad_norm": 0.7403807640075684, "learning_rate": 9.937689319103641e-05, "loss": 0.0741, "step": 9780 }, { "grad_norm": 0.5140476822853088, "learning_rate": 9.937428823396404e-05, "loss": 0.0816, "step": 9790 }, { "grad_norm": 0.8084299564361572, "learning_rate": 9.937167787739372e-05, "loss": 0.0732, "step": 9800 }, { "grad_norm": 0.5931748151779175, "learning_rate": 9.936906212161095e-05, "loss": 0.07, "step": 9810 }, { "grad_norm": 0.512702226638794, "learning_rate": 9.936644096690176e-05, "loss": 0.0893, "step": 9820 }, { "grad_norm": 0.4576549530029297, "learning_rate": 9.936381441355282e-05, "loss": 0.0805, "step": 9830 }, { "grad_norm": 0.5657522678375244, "learning_rate": 9.936118246185136e-05, "loss": 0.0885, "step": 9840 }, { "grad_norm": 0.5811039209365845, "learning_rate": 9.935854511208518e-05, "loss": 0.0847, "step": 9850 }, { "grad_norm": 0.37243014574050903, "learning_rate": 9.935590236454272e-05, "loss": 0.0659, "step": 9860 }, { "grad_norm": 0.47705546021461487, "learning_rate": 9.935325421951298e-05, "loss": 0.07, "step": 9870 }, { "grad_norm": 0.4821339547634125, "learning_rate": 9.935060067728557e-05, "loss": 0.0689, "step": 9880 }, { "grad_norm": 0.4249475598335266, "learning_rate": 9.934794173815067e-05, "loss": 0.0574, "step": 9890 }, { "grad_norm": 0.5411922335624695, "learning_rate": 9.934527740239906e-05, "loss": 0.0796, "step": 9900 }, { "grad_norm": 0.4461168050765991, "learning_rate": 9.934260767032209e-05, "loss": 0.0909, "step": 9910 }, { "grad_norm": 0.604425847530365, "learning_rate": 9.933993254221172e-05, "loss": 0.0824, "step": 9920 }, { "grad_norm": 0.40903806686401367, "learning_rate": 9.933725201836053e-05, "loss": 0.0669, "step": 9930 }, { "grad_norm": 0.49353548884391785, "learning_rate": 9.933456609906162e-05, "loss": 0.0712, "step": 9940 }, { "grad_norm": 0.533912718296051, "learning_rate": 9.933187478460875e-05, "loss": 0.0739, "step": 9950 }, { "grad_norm": 0.7759056687355042, "learning_rate": 9.93291780752962e-05, "loss": 0.0745, "step": 9960 }, { "grad_norm": 0.40788745880126953, "learning_rate": 9.932647597141893e-05, "loss": 0.0884, "step": 9970 }, { "grad_norm": 0.5390570759773254, "learning_rate": 9.932376847327239e-05, "loss": 0.0761, "step": 9980 }, { "grad_norm": 0.6674067378044128, "learning_rate": 9.932105558115268e-05, "loss": 0.0755, "step": 9990 }, { "grad_norm": 0.7067263722419739, "learning_rate": 9.931833729535651e-05, "loss": 0.09, "step": 10000 }, { "grad_norm": 0.7123307585716248, "learning_rate": 9.931561361618111e-05, "loss": 0.0835, "step": 10010 }, { "grad_norm": 0.6803063154220581, "learning_rate": 9.931288454392435e-05, "loss": 0.0941, "step": 10020 }, { "grad_norm": 0.5586262345314026, "learning_rate": 9.931015007888467e-05, "loss": 0.0725, "step": 10030 }, { "grad_norm": 0.5998433232307434, "learning_rate": 9.930741022136112e-05, "loss": 0.0827, "step": 10040 }, { "grad_norm": 0.5680224895477295, "learning_rate": 9.930466497165333e-05, "loss": 0.0753, "step": 10050 }, { "grad_norm": 0.5910230875015259, "learning_rate": 9.93019143300615e-05, "loss": 0.0767, "step": 10060 }, { "grad_norm": 0.537956953048706, "learning_rate": 9.929915829688644e-05, "loss": 0.0618, "step": 10070 }, { "grad_norm": 0.625791072845459, "learning_rate": 9.929639687242955e-05, "loss": 0.0735, "step": 10080 }, { "grad_norm": 0.5375595688819885, "learning_rate": 9.929363005699281e-05, "loss": 0.0732, "step": 10090 }, { "grad_norm": 0.6920980215072632, "learning_rate": 9.92908578508788e-05, "loss": 0.0873, "step": 10100 }, { "grad_norm": 0.5693997144699097, "learning_rate": 9.928808025439069e-05, "loss": 0.0677, "step": 10110 }, { "grad_norm": 0.8081331253051758, "learning_rate": 9.928529726783223e-05, "loss": 0.0846, "step": 10120 }, { "grad_norm": 0.5973985195159912, "learning_rate": 9.928250889150774e-05, "loss": 0.0882, "step": 10130 }, { "grad_norm": 0.6765013337135315, "learning_rate": 9.92797151257222e-05, "loss": 0.0808, "step": 10140 }, { "grad_norm": 0.5078689455986023, "learning_rate": 9.927691597078108e-05, "loss": 0.0861, "step": 10150 }, { "grad_norm": 0.535650908946991, "learning_rate": 9.927411142699053e-05, "loss": 0.0813, "step": 10160 }, { "grad_norm": 0.4728032946586609, "learning_rate": 9.927130149465725e-05, "loss": 0.0686, "step": 10170 }, { "grad_norm": 0.6411539912223816, "learning_rate": 9.92684861740885e-05, "loss": 0.0894, "step": 10180 }, { "grad_norm": 0.6356725692749023, "learning_rate": 9.926566546559217e-05, "loss": 0.081, "step": 10190 }, { "grad_norm": 0.6369946599006653, "learning_rate": 9.926283936947673e-05, "loss": 0.0782, "step": 10200 }, { "grad_norm": 0.5701655745506287, "learning_rate": 9.926000788605126e-05, "loss": 0.0821, "step": 10210 }, { "grad_norm": 0.43951693177223206, "learning_rate": 9.92571710156254e-05, "loss": 0.0793, "step": 10220 }, { "grad_norm": 0.5672465562820435, "learning_rate": 9.925432875850936e-05, "loss": 0.0916, "step": 10230 }, { "grad_norm": 0.514807403087616, "learning_rate": 9.925148111501396e-05, "loss": 0.092, "step": 10240 }, { "grad_norm": 0.7467047572135925, "learning_rate": 9.924862808545066e-05, "loss": 0.0796, "step": 10250 }, { "grad_norm": 0.3846419155597687, "learning_rate": 9.924576967013141e-05, "loss": 0.0684, "step": 10260 }, { "grad_norm": 0.6358858346939087, "learning_rate": 9.924290586936887e-05, "loss": 0.0814, "step": 10270 }, { "grad_norm": 0.4678584635257721, "learning_rate": 9.924003668347614e-05, "loss": 0.0746, "step": 10280 }, { "grad_norm": 0.5607278943061829, "learning_rate": 9.923716211276704e-05, "loss": 0.0837, "step": 10290 }, { "grad_norm": 0.5194940567016602, "learning_rate": 9.923428215755594e-05, "loss": 0.0717, "step": 10300 }, { "grad_norm": 0.5324553847312927, "learning_rate": 9.923139681815775e-05, "loss": 0.0591, "step": 10310 }, { "grad_norm": 0.5128465294837952, "learning_rate": 9.922850609488801e-05, "loss": 0.0774, "step": 10320 }, { "grad_norm": 0.5531270503997803, "learning_rate": 9.922560998806287e-05, "loss": 0.0755, "step": 10330 }, { "grad_norm": 0.5568884611129761, "learning_rate": 9.922270849799905e-05, "loss": 0.0914, "step": 10340 }, { "grad_norm": 0.6035518646240234, "learning_rate": 9.92198016250138e-05, "loss": 0.0879, "step": 10350 }, { "grad_norm": 0.45805710554122925, "learning_rate": 9.921688936942506e-05, "loss": 0.078, "step": 10360 }, { "grad_norm": 0.49270129203796387, "learning_rate": 9.921397173155129e-05, "loss": 0.069, "step": 10370 }, { "grad_norm": 0.4925785958766937, "learning_rate": 9.921104871171157e-05, "loss": 0.0631, "step": 10380 }, { "grad_norm": 0.4014309048652649, "learning_rate": 9.920812031022554e-05, "loss": 0.0579, "step": 10390 }, { "grad_norm": 0.49642401933670044, "learning_rate": 9.920518652741348e-05, "loss": 0.0688, "step": 10400 }, { "grad_norm": 0.49141788482666016, "learning_rate": 9.920224736359618e-05, "loss": 0.0627, "step": 10410 }, { "grad_norm": 0.6116394996643066, "learning_rate": 9.91993028190951e-05, "loss": 0.0872, "step": 10420 }, { "grad_norm": 0.7265805602073669, "learning_rate": 9.919635289423222e-05, "loss": 0.074, "step": 10430 }, { "grad_norm": 0.4897650480270386, "learning_rate": 9.919339758933015e-05, "loss": 0.0671, "step": 10440 }, { "grad_norm": 0.6212243437767029, "learning_rate": 9.919043690471209e-05, "loss": 0.0822, "step": 10450 }, { "grad_norm": 0.5452858805656433, "learning_rate": 9.91874708407018e-05, "loss": 0.0782, "step": 10460 }, { "grad_norm": 0.5459113121032715, "learning_rate": 9.918449939762367e-05, "loss": 0.0744, "step": 10470 }, { "grad_norm": 0.49184608459472656, "learning_rate": 9.91815225758026e-05, "loss": 0.0721, "step": 10480 }, { "grad_norm": 0.568067193031311, "learning_rate": 9.917854037556419e-05, "loss": 0.0744, "step": 10490 }, { "grad_norm": 0.6048381924629211, "learning_rate": 9.917555279723454e-05, "loss": 0.0771, "step": 10500 }, { "grad_norm": 0.7778548002243042, "learning_rate": 9.917255984114036e-05, "loss": 0.0883, "step": 10510 }, { "grad_norm": 0.5084627866744995, "learning_rate": 9.916956150760896e-05, "loss": 0.0715, "step": 10520 }, { "grad_norm": 0.5440859198570251, "learning_rate": 9.916655779696826e-05, "loss": 0.0749, "step": 10530 }, { "grad_norm": 0.6512702107429504, "learning_rate": 9.916354870954671e-05, "loss": 0.0773, "step": 10540 }, { "grad_norm": 0.5467981100082397, "learning_rate": 9.91605342456734e-05, "loss": 0.0679, "step": 10550 }, { "grad_norm": 0.621918261051178, "learning_rate": 9.915751440567795e-05, "loss": 0.0696, "step": 10560 }, { "grad_norm": 0.5389869213104248, "learning_rate": 9.915448918989066e-05, "loss": 0.0689, "step": 10570 }, { "grad_norm": 0.7389652729034424, "learning_rate": 9.915145859864232e-05, "loss": 0.0989, "step": 10580 }, { "grad_norm": 0.9727239012718201, "learning_rate": 9.914842263226437e-05, "loss": 0.1075, "step": 10590 }, { "grad_norm": 0.6582130789756775, "learning_rate": 9.914538129108882e-05, "loss": 0.0945, "step": 10600 }, { "grad_norm": 0.9020976424217224, "learning_rate": 9.914233457544825e-05, "loss": 0.1014, "step": 10610 }, { "grad_norm": 0.639677882194519, "learning_rate": 9.913928248567586e-05, "loss": 0.0963, "step": 10620 }, { "grad_norm": 0.5003873705863953, "learning_rate": 9.913622502210542e-05, "loss": 0.077, "step": 10630 }, { "grad_norm": 0.5182608366012573, "learning_rate": 9.913316218507128e-05, "loss": 0.0889, "step": 10640 }, { "grad_norm": 0.43990057706832886, "learning_rate": 9.91300939749084e-05, "loss": 0.0749, "step": 10650 }, { "grad_norm": 0.7065050601959229, "learning_rate": 9.91270203919523e-05, "loss": 0.0782, "step": 10660 }, { "grad_norm": 0.6451457142829895, "learning_rate": 9.912394143653912e-05, "loss": 0.0804, "step": 10670 }, { "grad_norm": 0.5298787355422974, "learning_rate": 9.912085710900555e-05, "loss": 0.077, "step": 10680 }, { "grad_norm": 0.5608459115028381, "learning_rate": 9.911776740968892e-05, "loss": 0.0786, "step": 10690 }, { "grad_norm": 0.5705996155738831, "learning_rate": 9.911467233892709e-05, "loss": 0.1042, "step": 10700 }, { "grad_norm": 0.45749130845069885, "learning_rate": 9.911157189705853e-05, "loss": 0.0804, "step": 10710 }, { "grad_norm": 0.6915501356124878, "learning_rate": 9.910846608442229e-05, "loss": 0.0755, "step": 10720 }, { "grad_norm": 0.5832468271255493, "learning_rate": 9.910535490135805e-05, "loss": 0.0817, "step": 10730 }, { "grad_norm": 0.4716850221157074, "learning_rate": 9.910223834820603e-05, "loss": 0.0679, "step": 10740 }, { "grad_norm": 0.7776519060134888, "learning_rate": 9.909911642530703e-05, "loss": 0.0752, "step": 10750 }, { "grad_norm": 0.885272204875946, "learning_rate": 9.909598913300249e-05, "loss": 0.0869, "step": 10760 }, { "grad_norm": 0.5150632262229919, "learning_rate": 9.909285647163438e-05, "loss": 0.082, "step": 10770 }, { "grad_norm": 0.728115439414978, "learning_rate": 9.908971844154531e-05, "loss": 0.0813, "step": 10780 }, { "grad_norm": 0.7088465094566345, "learning_rate": 9.908657504307843e-05, "loss": 0.0756, "step": 10790 }, { "grad_norm": 0.5896149277687073, "learning_rate": 9.908342627657751e-05, "loss": 0.0856, "step": 10800 }, { "grad_norm": 0.4700723886489868, "learning_rate": 9.908027214238689e-05, "loss": 0.0717, "step": 10810 }, { "grad_norm": 0.44602304697036743, "learning_rate": 9.90771126408515e-05, "loss": 0.0874, "step": 10820 }, { "grad_norm": 0.7386049032211304, "learning_rate": 9.907394777231685e-05, "loss": 0.0837, "step": 10830 }, { "grad_norm": 0.5433297157287598, "learning_rate": 9.907077753712905e-05, "loss": 0.0939, "step": 10840 }, { "grad_norm": 0.6910134553909302, "learning_rate": 9.906760193563482e-05, "loss": 0.0811, "step": 10850 }, { "grad_norm": 0.5775553584098816, "learning_rate": 9.906442096818139e-05, "loss": 0.0818, "step": 10860 }, { "grad_norm": 0.5415511727333069, "learning_rate": 9.906123463511665e-05, "loss": 0.0705, "step": 10870 }, { "grad_norm": 0.47023671865463257, "learning_rate": 9.905804293678907e-05, "loss": 0.0804, "step": 10880 }, { "grad_norm": 0.5055596232414246, "learning_rate": 9.905484587354766e-05, "loss": 0.0726, "step": 10890 }, { "grad_norm": 0.4965604841709137, "learning_rate": 9.905164344574205e-05, "loss": 0.0778, "step": 10900 }, { "grad_norm": 0.4701477587223053, "learning_rate": 9.904843565372248e-05, "loss": 0.0878, "step": 10910 }, { "grad_norm": 0.6536282300949097, "learning_rate": 9.904522249783972e-05, "loss": 0.0758, "step": 10920 }, { "grad_norm": 0.4245617985725403, "learning_rate": 9.904200397844517e-05, "loss": 0.0876, "step": 10930 }, { "grad_norm": 0.5102784037590027, "learning_rate": 9.903878009589078e-05, "loss": 0.0882, "step": 10940 }, { "grad_norm": 0.7367872595787048, "learning_rate": 9.903555085052915e-05, "loss": 0.0741, "step": 10950 }, { "grad_norm": 0.8042119741439819, "learning_rate": 9.903231624271338e-05, "loss": 0.0616, "step": 10960 }, { "grad_norm": 0.7343879342079163, "learning_rate": 9.902907627279724e-05, "loss": 0.0776, "step": 10970 }, { "grad_norm": 0.7806215882301331, "learning_rate": 9.902583094113504e-05, "loss": 0.0779, "step": 10980 }, { "grad_norm": 0.6685734987258911, "learning_rate": 9.902258024808168e-05, "loss": 0.0794, "step": 10990 }, { "grad_norm": 0.4989832639694214, "learning_rate": 9.901932419399264e-05, "loss": 0.0702, "step": 11000 }, { "grad_norm": 0.6241717338562012, "learning_rate": 9.9016062779224e-05, "loss": 0.0815, "step": 11010 }, { "grad_norm": 0.5338150262832642, "learning_rate": 9.901279600413242e-05, "loss": 0.0801, "step": 11020 }, { "grad_norm": 0.5227404236793518, "learning_rate": 9.900952386907518e-05, "loss": 0.0586, "step": 11030 }, { "grad_norm": 0.4759729504585266, "learning_rate": 9.90062463744101e-05, "loss": 0.0755, "step": 11040 }, { "grad_norm": 0.6385480761528015, "learning_rate": 9.900296352049558e-05, "loss": 0.0724, "step": 11050 }, { "grad_norm": 0.7776791453361511, "learning_rate": 9.899967530769065e-05, "loss": 0.0953, "step": 11060 }, { "grad_norm": 0.7480636835098267, "learning_rate": 9.899638173635489e-05, "loss": 0.087, "step": 11070 }, { "grad_norm": 0.555286169052124, "learning_rate": 9.899308280684849e-05, "loss": 0.0886, "step": 11080 }, { "grad_norm": 1.2140151262283325, "learning_rate": 9.898977851953222e-05, "loss": 0.0858, "step": 11090 }, { "grad_norm": 0.596057116985321, "learning_rate": 9.898646887476741e-05, "loss": 0.081, "step": 11100 }, { "grad_norm": 0.5398393869400024, "learning_rate": 9.898315387291603e-05, "loss": 0.0774, "step": 11110 }, { "grad_norm": 0.6177268624305725, "learning_rate": 9.89798335143406e-05, "loss": 0.073, "step": 11120 }, { "grad_norm": 0.7566661834716797, "learning_rate": 9.897650779940419e-05, "loss": 0.0805, "step": 11130 }, { "grad_norm": 0.5976080298423767, "learning_rate": 9.897317672847054e-05, "loss": 0.0729, "step": 11140 }, { "grad_norm": 0.5680294036865234, "learning_rate": 9.89698403019039e-05, "loss": 0.0608, "step": 11150 }, { "grad_norm": 0.7355846762657166, "learning_rate": 9.896649852006917e-05, "loss": 0.0791, "step": 11160 }, { "grad_norm": 0.6693976521492004, "learning_rate": 9.896315138333177e-05, "loss": 0.0731, "step": 11170 }, { "grad_norm": 0.8390043377876282, "learning_rate": 9.895979889205774e-05, "loss": 0.0774, "step": 11180 }, { "grad_norm": 1.0186444520950317, "learning_rate": 9.895644104661372e-05, "loss": 0.0684, "step": 11190 }, { "grad_norm": 0.4555801749229431, "learning_rate": 9.895307784736691e-05, "loss": 0.0718, "step": 11200 }, { "grad_norm": 0.5245493650436401, "learning_rate": 9.894970929468512e-05, "loss": 0.0775, "step": 11210 }, { "grad_norm": 0.4557781517505646, "learning_rate": 9.89463353889367e-05, "loss": 0.0737, "step": 11220 }, { "grad_norm": 0.4357207715511322, "learning_rate": 9.894295613049065e-05, "loss": 0.0767, "step": 11230 }, { "grad_norm": 0.6113861799240112, "learning_rate": 9.893957151971649e-05, "loss": 0.068, "step": 11240 }, { "grad_norm": 0.5238082408905029, "learning_rate": 9.893618155698436e-05, "loss": 0.0759, "step": 11250 }, { "grad_norm": 0.455962598323822, "learning_rate": 9.8932786242665e-05, "loss": 0.0676, "step": 11260 }, { "grad_norm": 0.430656373500824, "learning_rate": 9.89293855771297e-05, "loss": 0.0654, "step": 11270 }, { "grad_norm": 0.6913371682167053, "learning_rate": 9.892597956075036e-05, "loss": 0.0787, "step": 11280 }, { "grad_norm": 0.5529530048370361, "learning_rate": 9.892256819389947e-05, "loss": 0.0719, "step": 11290 }, { "grad_norm": 0.693730354309082, "learning_rate": 9.891915147695006e-05, "loss": 0.0755, "step": 11300 }, { "grad_norm": 0.7339118719100952, "learning_rate": 9.891572941027577e-05, "loss": 0.0621, "step": 11310 }, { "grad_norm": 0.4475550055503845, "learning_rate": 9.89123019942509e-05, "loss": 0.0651, "step": 11320 }, { "grad_norm": 0.7286328673362732, "learning_rate": 9.89088692292502e-05, "loss": 0.0662, "step": 11330 }, { "grad_norm": 0.5658106803894043, "learning_rate": 9.89054311156491e-05, "loss": 0.0588, "step": 11340 }, { "grad_norm": 0.6263095736503601, "learning_rate": 9.890198765382357e-05, "loss": 0.0944, "step": 11350 }, { "grad_norm": 0.40705567598342896, "learning_rate": 9.889853884415021e-05, "loss": 0.0654, "step": 11360 }, { "grad_norm": 0.410675972700119, "learning_rate": 9.889508468700614e-05, "loss": 0.0634, "step": 11370 }, { "grad_norm": 0.34662947058677673, "learning_rate": 9.889162518276915e-05, "loss": 0.0542, "step": 11380 }, { "grad_norm": 0.5324862599372864, "learning_rate": 9.888816033181752e-05, "loss": 0.0743, "step": 11390 }, { "grad_norm": 0.4763200879096985, "learning_rate": 9.888469013453018e-05, "loss": 0.0714, "step": 11400 }, { "grad_norm": 0.6948539614677429, "learning_rate": 9.888121459128663e-05, "loss": 0.0718, "step": 11410 }, { "grad_norm": 0.45714664459228516, "learning_rate": 9.887773370246693e-05, "loss": 0.0851, "step": 11420 }, { "grad_norm": 0.8398704528808594, "learning_rate": 9.887424746845177e-05, "loss": 0.0739, "step": 11430 }, { "grad_norm": 0.4418770372867584, "learning_rate": 9.887075588962239e-05, "loss": 0.075, "step": 11440 }, { "grad_norm": 0.628872811794281, "learning_rate": 9.88672589663606e-05, "loss": 0.061, "step": 11450 }, { "grad_norm": 0.5142441391944885, "learning_rate": 9.886375669904886e-05, "loss": 0.0813, "step": 11460 }, { "grad_norm": 0.5129854083061218, "learning_rate": 9.886024908807014e-05, "loss": 0.0639, "step": 11470 }, { "grad_norm": 0.546919584274292, "learning_rate": 9.885673613380806e-05, "loss": 0.061, "step": 11480 }, { "grad_norm": 0.5046511292457581, "learning_rate": 9.885321783664676e-05, "loss": 0.0702, "step": 11490 }, { "grad_norm": 0.5589166283607483, "learning_rate": 9.884969419697101e-05, "loss": 0.0684, "step": 11500 }, { "grad_norm": 0.5961459875106812, "learning_rate": 9.884616521516614e-05, "loss": 0.0683, "step": 11510 }, { "grad_norm": 0.5157492756843567, "learning_rate": 9.88426308916181e-05, "loss": 0.0888, "step": 11520 }, { "grad_norm": 0.7174730896949768, "learning_rate": 9.883909122671335e-05, "loss": 0.0711, "step": 11530 }, { "grad_norm": 0.5195183157920837, "learning_rate": 9.883554622083904e-05, "loss": 0.082, "step": 11540 }, { "grad_norm": 0.6923317909240723, "learning_rate": 9.88319958743828e-05, "loss": 0.086, "step": 11550 }, { "grad_norm": 0.4580357074737549, "learning_rate": 9.882844018773291e-05, "loss": 0.0739, "step": 11560 }, { "grad_norm": 0.5478584170341492, "learning_rate": 9.882487916127823e-05, "loss": 0.0852, "step": 11570 }, { "grad_norm": 0.6073324680328369, "learning_rate": 9.882131279540815e-05, "loss": 0.0865, "step": 11580 }, { "grad_norm": 0.6523494124412537, "learning_rate": 9.881774109051271e-05, "loss": 0.0761, "step": 11590 }, { "grad_norm": 0.3902169167995453, "learning_rate": 9.881416404698252e-05, "loss": 0.0659, "step": 11600 }, { "grad_norm": 0.4179495871067047, "learning_rate": 9.881058166520873e-05, "loss": 0.0877, "step": 11610 }, { "grad_norm": 0.6857146620750427, "learning_rate": 9.880699394558311e-05, "loss": 0.0784, "step": 11620 }, { "grad_norm": 0.4707344174385071, "learning_rate": 9.880340088849801e-05, "loss": 0.0885, "step": 11630 }, { "grad_norm": 0.6428598165512085, "learning_rate": 9.879980249434637e-05, "loss": 0.0809, "step": 11640 }, { "grad_norm": 0.5641239881515503, "learning_rate": 9.879619876352168e-05, "loss": 0.0868, "step": 11650 }, { "grad_norm": 0.6403355002403259, "learning_rate": 9.879258969641809e-05, "loss": 0.0799, "step": 11660 }, { "grad_norm": 0.39525264501571655, "learning_rate": 9.878897529343023e-05, "loss": 0.0625, "step": 11670 }, { "grad_norm": 0.6227962970733643, "learning_rate": 9.878535555495338e-05, "loss": 0.0824, "step": 11680 }, { "grad_norm": 0.6605601906776428, "learning_rate": 9.87817304813834e-05, "loss": 0.0919, "step": 11690 }, { "grad_norm": 0.41811805963516235, "learning_rate": 9.877810007311671e-05, "loss": 0.0808, "step": 11700 }, { "grad_norm": 0.4930799603462219, "learning_rate": 9.877446433055035e-05, "loss": 0.0747, "step": 11710 }, { "grad_norm": 0.6361083388328552, "learning_rate": 9.877082325408191e-05, "loss": 0.0809, "step": 11720 }, { "grad_norm": 0.5996217727661133, "learning_rate": 9.876717684410954e-05, "loss": 0.0622, "step": 11730 }, { "grad_norm": 0.7414737939834595, "learning_rate": 9.876352510103204e-05, "loss": 0.0721, "step": 11740 }, { "grad_norm": 0.6087737679481506, "learning_rate": 9.875986802524875e-05, "loss": 0.0753, "step": 11750 }, { "grad_norm": 0.6100000143051147, "learning_rate": 9.87562056171596e-05, "loss": 0.0672, "step": 11760 }, { "grad_norm": 0.7520536780357361, "learning_rate": 9.875253787716511e-05, "loss": 0.0919, "step": 11770 }, { "grad_norm": 0.5435769557952881, "learning_rate": 9.874886480566637e-05, "loss": 0.0658, "step": 11780 }, { "grad_norm": 0.4301117956638336, "learning_rate": 9.874518640306507e-05, "loss": 0.0678, "step": 11790 }, { "grad_norm": 0.4250524044036865, "learning_rate": 9.874150266976347e-05, "loss": 0.067, "step": 11800 }, { "grad_norm": 0.5233409404754639, "learning_rate": 9.873781360616443e-05, "loss": 0.0639, "step": 11810 }, { "grad_norm": 0.31186068058013916, "learning_rate": 9.873411921267137e-05, "loss": 0.0623, "step": 11820 }, { "grad_norm": 0.8763093948364258, "learning_rate": 9.873041948968829e-05, "loss": 0.0806, "step": 11830 }, { "grad_norm": 0.33985406160354614, "learning_rate": 9.872671443761981e-05, "loss": 0.0608, "step": 11840 }, { "grad_norm": 0.6173772811889648, "learning_rate": 9.872300405687109e-05, "loss": 0.0691, "step": 11850 }, { "grad_norm": 0.604223906993866, "learning_rate": 9.871928834784792e-05, "loss": 0.078, "step": 11860 }, { "grad_norm": 0.3938983678817749, "learning_rate": 9.871556731095661e-05, "loss": 0.0633, "step": 11870 }, { "grad_norm": 0.5182603001594543, "learning_rate": 9.871184094660411e-05, "loss": 0.0835, "step": 11880 }, { "grad_norm": 0.8262428045272827, "learning_rate": 9.870810925519791e-05, "loss": 0.0701, "step": 11890 }, { "grad_norm": 0.6267531514167786, "learning_rate": 9.870437223714612e-05, "loss": 0.0647, "step": 11900 }, { "grad_norm": 0.5929664373397827, "learning_rate": 9.87006298928574e-05, "loss": 0.0792, "step": 11910 }, { "grad_norm": 0.4865091145038605, "learning_rate": 9.869688222274103e-05, "loss": 0.0866, "step": 11920 }, { "grad_norm": 0.416133850812912, "learning_rate": 9.869312922720681e-05, "loss": 0.0701, "step": 11930 }, { "grad_norm": 0.5156626105308533, "learning_rate": 9.868937090666521e-05, "loss": 0.0695, "step": 11940 }, { "grad_norm": 0.5431283116340637, "learning_rate": 9.86856072615272e-05, "loss": 0.0677, "step": 11950 }, { "grad_norm": 0.5772140026092529, "learning_rate": 9.868183829220438e-05, "loss": 0.065, "step": 11960 }, { "grad_norm": 0.6322351098060608, "learning_rate": 9.867806399910893e-05, "loss": 0.0681, "step": 11970 }, { "grad_norm": 0.6247822046279907, "learning_rate": 9.867428438265356e-05, "loss": 0.0774, "step": 11980 }, { "grad_norm": 0.4398997128009796, "learning_rate": 9.867049944325165e-05, "loss": 0.0753, "step": 11990 }, { "grad_norm": 0.5018728375434875, "learning_rate": 9.86667091813171e-05, "loss": 0.0619, "step": 12000 }, { "grad_norm": 0.5266463160514832, "learning_rate": 9.866291359726438e-05, "loss": 0.0634, "step": 12010 }, { "grad_norm": 1.1759061813354492, "learning_rate": 9.865911269150861e-05, "loss": 0.0668, "step": 12020 }, { "grad_norm": 0.432451069355011, "learning_rate": 9.865530646446544e-05, "loss": 0.0726, "step": 12030 }, { "grad_norm": 0.5971106290817261, "learning_rate": 9.86514949165511e-05, "loss": 0.0803, "step": 12040 }, { "grad_norm": 0.9156911969184875, "learning_rate": 9.864767804818243e-05, "loss": 0.0822, "step": 12050 }, { "grad_norm": 0.4246056377887726, "learning_rate": 9.86438558597768e-05, "loss": 0.0692, "step": 12060 }, { "grad_norm": 0.7141113877296448, "learning_rate": 9.864002835175225e-05, "loss": 0.0923, "step": 12070 }, { "grad_norm": 0.547774076461792, "learning_rate": 9.863619552452734e-05, "loss": 0.0716, "step": 12080 }, { "grad_norm": 0.46664759516716003, "learning_rate": 9.863235737852119e-05, "loss": 0.0579, "step": 12090 }, { "grad_norm": 0.6352298259735107, "learning_rate": 9.862851391415356e-05, "loss": 0.0691, "step": 12100 }, { "grad_norm": 0.3813004195690155, "learning_rate": 9.862466513184477e-05, "loss": 0.0714, "step": 12110 }, { "grad_norm": 0.7030959129333496, "learning_rate": 9.86208110320157e-05, "loss": 0.0576, "step": 12120 }, { "grad_norm": 0.4971124231815338, "learning_rate": 9.861695161508784e-05, "loss": 0.0668, "step": 12130 }, { "grad_norm": 0.3867605924606323, "learning_rate": 9.861308688148324e-05, "loss": 0.0525, "step": 12140 }, { "grad_norm": 0.5554335117340088, "learning_rate": 9.860921683162455e-05, "loss": 0.0618, "step": 12150 }, { "grad_norm": 0.45231592655181885, "learning_rate": 9.860534146593499e-05, "loss": 0.055, "step": 12160 }, { "grad_norm": 0.5772343277931213, "learning_rate": 9.860146078483836e-05, "loss": 0.0597, "step": 12170 }, { "grad_norm": 0.6503999829292297, "learning_rate": 9.859757478875905e-05, "loss": 0.0726, "step": 12180 }, { "grad_norm": 0.49790239334106445, "learning_rate": 9.859368347812204e-05, "loss": 0.0702, "step": 12190 }, { "grad_norm": 0.9372458457946777, "learning_rate": 9.858978685335285e-05, "loss": 0.0772, "step": 12200 }, { "grad_norm": 0.6109262704849243, "learning_rate": 9.858588491487763e-05, "loss": 0.0968, "step": 12210 }, { "grad_norm": 0.5129473805427551, "learning_rate": 9.858197766312308e-05, "loss": 0.0715, "step": 12220 }, { "grad_norm": 0.585886538028717, "learning_rate": 9.857806509851649e-05, "loss": 0.0709, "step": 12230 }, { "grad_norm": 0.4980422258377075, "learning_rate": 9.857414722148574e-05, "loss": 0.0672, "step": 12240 }, { "grad_norm": 0.5295401811599731, "learning_rate": 9.857022403245928e-05, "loss": 0.0738, "step": 12250 }, { "grad_norm": 0.6715447306632996, "learning_rate": 9.856629553186615e-05, "loss": 0.0877, "step": 12260 }, { "grad_norm": 0.6602984070777893, "learning_rate": 9.856236172013595e-05, "loss": 0.066, "step": 12270 }, { "grad_norm": 0.4873175323009491, "learning_rate": 9.85584225976989e-05, "loss": 0.0727, "step": 12280 }, { "grad_norm": 0.47545769810676575, "learning_rate": 9.855447816498575e-05, "loss": 0.0786, "step": 12290 }, { "grad_norm": 0.5418950319290161, "learning_rate": 9.855052842242787e-05, "loss": 0.0796, "step": 12300 }, { "grad_norm": 0.4399215281009674, "learning_rate": 9.85465733704572e-05, "loss": 0.061, "step": 12310 }, { "grad_norm": 0.7163896560668945, "learning_rate": 9.854261300950624e-05, "loss": 0.0728, "step": 12320 }, { "grad_norm": 0.5980188846588135, "learning_rate": 9.853864734000813e-05, "loss": 0.0794, "step": 12330 }, { "grad_norm": 0.497770756483078, "learning_rate": 9.85346763623965e-05, "loss": 0.0748, "step": 12340 }, { "grad_norm": 0.4594686031341553, "learning_rate": 9.853070007710564e-05, "loss": 0.0725, "step": 12350 }, { "grad_norm": 0.5075130462646484, "learning_rate": 9.85267184845704e-05, "loss": 0.0667, "step": 12360 }, { "grad_norm": 0.431944876909256, "learning_rate": 9.852273158522616e-05, "loss": 0.0567, "step": 12370 }, { "grad_norm": 0.5930433869361877, "learning_rate": 9.851873937950896e-05, "loss": 0.0724, "step": 12380 }, { "grad_norm": 0.6009594202041626, "learning_rate": 9.851474186785537e-05, "loss": 0.069, "step": 12390 }, { "grad_norm": 0.51096111536026, "learning_rate": 9.851073905070254e-05, "loss": 0.0784, "step": 12400 }, { "grad_norm": 0.6405578255653381, "learning_rate": 9.850673092848824e-05, "loss": 0.0763, "step": 12410 }, { "grad_norm": 0.3981204628944397, "learning_rate": 9.850271750165077e-05, "loss": 0.0717, "step": 12420 }, { "grad_norm": 0.6400072574615479, "learning_rate": 9.849869877062902e-05, "loss": 0.0787, "step": 12430 }, { "grad_norm": 0.5906952023506165, "learning_rate": 9.849467473586252e-05, "loss": 0.0848, "step": 12440 }, { "grad_norm": 0.47049909830093384, "learning_rate": 9.849064539779127e-05, "loss": 0.059, "step": 12450 }, { "grad_norm": 0.4102904796600342, "learning_rate": 9.848661075685594e-05, "loss": 0.0782, "step": 12460 }, { "grad_norm": 0.5688937306404114, "learning_rate": 9.848257081349778e-05, "loss": 0.0715, "step": 12470 }, { "grad_norm": 0.5072950124740601, "learning_rate": 9.847852556815856e-05, "loss": 0.0503, "step": 12480 }, { "grad_norm": 0.5126859545707703, "learning_rate": 9.847447502128067e-05, "loss": 0.0786, "step": 12490 }, { "grad_norm": 0.4508276581764221, "learning_rate": 9.847041917330708e-05, "loss": 0.0781, "step": 12500 }, { "grad_norm": 0.46519142389297485, "learning_rate": 9.846635802468132e-05, "loss": 0.0833, "step": 12510 }, { "grad_norm": 0.45741742849349976, "learning_rate": 9.84622915758475e-05, "loss": 0.058, "step": 12520 }, { "grad_norm": 0.7839412689208984, "learning_rate": 9.845821982725034e-05, "loss": 0.0676, "step": 12530 }, { "grad_norm": 0.4562146067619324, "learning_rate": 9.845414277933514e-05, "loss": 0.0667, "step": 12540 }, { "grad_norm": 0.6077532172203064, "learning_rate": 9.845006043254771e-05, "loss": 0.0633, "step": 12550 }, { "grad_norm": 0.5747917294502258, "learning_rate": 9.844597278733451e-05, "loss": 0.0632, "step": 12560 }, { "grad_norm": 0.7805191278457642, "learning_rate": 9.844187984414259e-05, "loss": 0.0654, "step": 12570 }, { "grad_norm": 0.3218630254268646, "learning_rate": 9.84377816034195e-05, "loss": 0.046, "step": 12580 }, { "grad_norm": 0.5411585569381714, "learning_rate": 9.843367806561345e-05, "loss": 0.0668, "step": 12590 }, { "grad_norm": 0.5162075161933899, "learning_rate": 9.842956923117317e-05, "loss": 0.0686, "step": 12600 }, { "grad_norm": 0.5815295577049255, "learning_rate": 9.842545510054802e-05, "loss": 0.0551, "step": 12610 }, { "grad_norm": 0.45352476835250854, "learning_rate": 9.842133567418792e-05, "loss": 0.0623, "step": 12620 }, { "grad_norm": 0.4857296943664551, "learning_rate": 9.841721095254333e-05, "loss": 0.0538, "step": 12630 }, { "grad_norm": 0.5885125994682312, "learning_rate": 9.841308093606537e-05, "loss": 0.0771, "step": 12640 }, { "grad_norm": 0.3466682732105255, "learning_rate": 9.840894562520565e-05, "loss": 0.064, "step": 12650 }, { "grad_norm": 0.5870630741119385, "learning_rate": 9.840480502041642e-05, "loss": 0.0832, "step": 12660 }, { "grad_norm": 0.6512677669525146, "learning_rate": 9.840065912215049e-05, "loss": 0.0923, "step": 12670 }, { "grad_norm": 0.49640265107154846, "learning_rate": 9.839650793086124e-05, "loss": 0.0912, "step": 12680 }, { "grad_norm": 0.4102204144001007, "learning_rate": 9.839235144700265e-05, "loss": 0.0669, "step": 12690 }, { "grad_norm": 0.43132883310317993, "learning_rate": 9.838818967102926e-05, "loss": 0.0662, "step": 12700 }, { "grad_norm": 0.5071401000022888, "learning_rate": 9.83840226033962e-05, "loss": 0.0617, "step": 12710 }, { "grad_norm": 0.7670071721076965, "learning_rate": 9.837985024455918e-05, "loss": 0.0819, "step": 12720 }, { "grad_norm": 0.5911761522293091, "learning_rate": 9.837567259497447e-05, "loss": 0.074, "step": 12730 }, { "grad_norm": 0.4662657380104065, "learning_rate": 9.837148965509894e-05, "loss": 0.0833, "step": 12740 }, { "grad_norm": 0.5740112662315369, "learning_rate": 9.836730142539001e-05, "loss": 0.0603, "step": 12750 }, { "grad_norm": 0.7249771952629089, "learning_rate": 9.836310790630574e-05, "loss": 0.0779, "step": 12760 }, { "grad_norm": 0.8082495331764221, "learning_rate": 9.83589090983047e-05, "loss": 0.0794, "step": 12770 }, { "grad_norm": 0.5250426530838013, "learning_rate": 9.835470500184605e-05, "loss": 0.0724, "step": 12780 }, { "grad_norm": 0.37121233344078064, "learning_rate": 9.835049561738957e-05, "loss": 0.0637, "step": 12790 }, { "grad_norm": 0.4642145335674286, "learning_rate": 9.834628094539558e-05, "loss": 0.0621, "step": 12800 }, { "grad_norm": 0.48728370666503906, "learning_rate": 9.834206098632499e-05, "loss": 0.0628, "step": 12810 }, { "grad_norm": 0.5077314376831055, "learning_rate": 9.833783574063931e-05, "loss": 0.0663, "step": 12820 }, { "grad_norm": 0.6503658294677734, "learning_rate": 9.833360520880058e-05, "loss": 0.0782, "step": 12830 }, { "grad_norm": 0.7018253207206726, "learning_rate": 9.832936939127144e-05, "loss": 0.062, "step": 12840 }, { "grad_norm": 0.7929916977882385, "learning_rate": 9.832512828851515e-05, "loss": 0.0621, "step": 12850 }, { "grad_norm": 0.4754112958908081, "learning_rate": 9.832088190099546e-05, "loss": 0.0642, "step": 12860 }, { "grad_norm": 0.4822908937931061, "learning_rate": 9.831663022917679e-05, "loss": 0.067, "step": 12870 }, { "grad_norm": 0.40018367767333984, "learning_rate": 9.831237327352407e-05, "loss": 0.0674, "step": 12880 }, { "grad_norm": 0.6495615243911743, "learning_rate": 9.830811103450286e-05, "loss": 0.071, "step": 12890 }, { "grad_norm": 0.38355544209480286, "learning_rate": 9.830384351257924e-05, "loss": 0.077, "step": 12900 }, { "grad_norm": 0.4951013922691345, "learning_rate": 9.829957070821993e-05, "loss": 0.0638, "step": 12910 }, { "grad_norm": 0.5249380469322205, "learning_rate": 9.829529262189218e-05, "loss": 0.0628, "step": 12920 }, { "grad_norm": 0.5868560671806335, "learning_rate": 9.829100925406385e-05, "loss": 0.0666, "step": 12930 }, { "grad_norm": 0.430285781621933, "learning_rate": 9.828672060520333e-05, "loss": 0.0688, "step": 12940 }, { "grad_norm": 0.3790825605392456, "learning_rate": 9.828242667577966e-05, "loss": 0.0535, "step": 12950 }, { "grad_norm": 0.4763283431529999, "learning_rate": 9.82781274662624e-05, "loss": 0.0534, "step": 12960 }, { "grad_norm": 1.0465670824050903, "learning_rate": 9.82738229771217e-05, "loss": 0.0602, "step": 12970 }, { "grad_norm": 0.5923973917961121, "learning_rate": 9.826951320882829e-05, "loss": 0.0658, "step": 12980 }, { "grad_norm": 0.3742401599884033, "learning_rate": 9.826519816185351e-05, "loss": 0.0532, "step": 12990 }, { "grad_norm": 0.7742959260940552, "learning_rate": 9.826087783666921e-05, "loss": 0.0758, "step": 13000 }, { "grad_norm": 0.5757840871810913, "learning_rate": 9.825655223374787e-05, "loss": 0.0869, "step": 13010 }, { "grad_norm": 0.6901115775108337, "learning_rate": 9.825222135356253e-05, "loss": 0.0667, "step": 13020 }, { "grad_norm": 0.582964301109314, "learning_rate": 9.82478851965868e-05, "loss": 0.0867, "step": 13030 }, { "grad_norm": 0.4973212778568268, "learning_rate": 9.82435437632949e-05, "loss": 0.0675, "step": 13040 }, { "grad_norm": 0.45223402976989746, "learning_rate": 9.823919705416158e-05, "loss": 0.0611, "step": 13050 }, { "grad_norm": 0.4284549355506897, "learning_rate": 9.82348450696622e-05, "loss": 0.0632, "step": 13060 }, { "grad_norm": 0.6329777240753174, "learning_rate": 9.823048781027268e-05, "loss": 0.0806, "step": 13070 }, { "grad_norm": 0.4345613718032837, "learning_rate": 9.822612527646953e-05, "loss": 0.0669, "step": 13080 }, { "grad_norm": 1.004307508468628, "learning_rate": 9.822175746872984e-05, "loss": 0.0658, "step": 13090 }, { "grad_norm": 1.493868112564087, "learning_rate": 9.821738438753123e-05, "loss": 0.0785, "step": 13100 }, { "grad_norm": 0.5389469861984253, "learning_rate": 9.821300603335196e-05, "loss": 0.0632, "step": 13110 }, { "grad_norm": 0.560727059841156, "learning_rate": 9.820862240667085e-05, "loss": 0.0681, "step": 13120 }, { "grad_norm": 0.4548585116863251, "learning_rate": 9.820423350796726e-05, "loss": 0.0628, "step": 13130 }, { "grad_norm": 0.4076622426509857, "learning_rate": 9.819983933772118e-05, "loss": 0.0617, "step": 13140 }, { "grad_norm": 0.5391562581062317, "learning_rate": 9.819543989641314e-05, "loss": 0.0656, "step": 13150 }, { "grad_norm": 0.37385493516921997, "learning_rate": 9.819103518452423e-05, "loss": 0.0564, "step": 13160 }, { "grad_norm": 0.4016643762588501, "learning_rate": 9.818662520253618e-05, "loss": 0.0694, "step": 13170 }, { "grad_norm": 0.49519550800323486, "learning_rate": 9.818220995093126e-05, "loss": 0.0622, "step": 13180 }, { "grad_norm": 0.4633079171180725, "learning_rate": 9.817778943019228e-05, "loss": 0.0634, "step": 13190 }, { "grad_norm": 0.4947296679019928, "learning_rate": 9.81733636408027e-05, "loss": 0.0712, "step": 13200 }, { "grad_norm": 0.5614418983459473, "learning_rate": 9.816893258324649e-05, "loss": 0.0553, "step": 13210 }, { "grad_norm": 0.6010323762893677, "learning_rate": 9.816449625800823e-05, "loss": 0.0595, "step": 13220 }, { "grad_norm": 0.4692857265472412, "learning_rate": 9.816005466557308e-05, "loss": 0.0667, "step": 13230 }, { "grad_norm": 0.5307237505912781, "learning_rate": 9.815560780642674e-05, "loss": 0.0758, "step": 13240 }, { "grad_norm": 0.5072485208511353, "learning_rate": 9.815115568105555e-05, "loss": 0.0709, "step": 13250 }, { "grad_norm": 0.6439869403839111, "learning_rate": 9.814669828994638e-05, "loss": 0.0772, "step": 13260 }, { "grad_norm": 0.6697062849998474, "learning_rate": 9.814223563358665e-05, "loss": 0.0694, "step": 13270 }, { "grad_norm": 0.7300432324409485, "learning_rate": 9.813776771246443e-05, "loss": 0.0598, "step": 13280 }, { "grad_norm": 0.4246239960193634, "learning_rate": 9.813329452706829e-05, "loss": 0.0559, "step": 13290 }, { "grad_norm": 0.6156641244888306, "learning_rate": 9.812881607788744e-05, "loss": 0.057, "step": 13300 }, { "grad_norm": 0.6251289248466492, "learning_rate": 9.812433236541163e-05, "loss": 0.0648, "step": 13310 }, { "grad_norm": 0.512965202331543, "learning_rate": 9.811984339013116e-05, "loss": 0.0733, "step": 13320 }, { "grad_norm": 0.5525036454200745, "learning_rate": 9.811534915253698e-05, "loss": 0.0656, "step": 13330 }, { "grad_norm": 0.45150554180145264, "learning_rate": 9.811084965312056e-05, "loss": 0.0663, "step": 13340 }, { "grad_norm": 0.563696026802063, "learning_rate": 9.810634489237396e-05, "loss": 0.0588, "step": 13350 }, { "grad_norm": 0.457223117351532, "learning_rate": 9.81018348707898e-05, "loss": 0.0679, "step": 13360 }, { "grad_norm": 0.404554158449173, "learning_rate": 9.809731958886131e-05, "loss": 0.0642, "step": 13370 }, { "grad_norm": 0.5515093803405762, "learning_rate": 9.809279904708224e-05, "loss": 0.0634, "step": 13380 }, { "grad_norm": 0.6236027479171753, "learning_rate": 9.808827324594699e-05, "loss": 0.0734, "step": 13390 }, { "grad_norm": 0.5447319746017456, "learning_rate": 9.808374218595046e-05, "loss": 0.0771, "step": 13400 }, { "grad_norm": 0.5586089491844177, "learning_rate": 9.80792058675882e-05, "loss": 0.0729, "step": 13410 }, { "grad_norm": 0.35861098766326904, "learning_rate": 9.807466429135627e-05, "loss": 0.0632, "step": 13420 }, { "grad_norm": 0.46285247802734375, "learning_rate": 9.807011745775132e-05, "loss": 0.0599, "step": 13430 }, { "grad_norm": 0.521842360496521, "learning_rate": 9.806556536727061e-05, "loss": 0.0544, "step": 13440 }, { "grad_norm": 0.5287321209907532, "learning_rate": 9.806100802041193e-05, "loss": 0.0708, "step": 13450 }, { "grad_norm": 0.610966682434082, "learning_rate": 9.805644541767368e-05, "loss": 0.0806, "step": 13460 }, { "grad_norm": 0.4858299195766449, "learning_rate": 9.805187755955478e-05, "loss": 0.082, "step": 13470 }, { "grad_norm": 0.432146817445755, "learning_rate": 9.804730444655483e-05, "loss": 0.0673, "step": 13480 }, { "grad_norm": 0.4005570113658905, "learning_rate": 9.804272607917388e-05, "loss": 0.0741, "step": 13490 }, { "grad_norm": 0.5158354640007019, "learning_rate": 9.803814245791265e-05, "loss": 0.0542, "step": 13500 }, { "grad_norm": 0.4376029074192047, "learning_rate": 9.803355358327239e-05, "loss": 0.0732, "step": 13510 }, { "grad_norm": 0.35872146487236023, "learning_rate": 9.802895945575492e-05, "loss": 0.0743, "step": 13520 }, { "grad_norm": 0.5341852307319641, "learning_rate": 9.802436007586266e-05, "loss": 0.0717, "step": 13530 }, { "grad_norm": 0.36208516359329224, "learning_rate": 9.801975544409858e-05, "loss": 0.0762, "step": 13540 }, { "grad_norm": 0.5965536832809448, "learning_rate": 9.801514556096625e-05, "loss": 0.0626, "step": 13550 }, { "grad_norm": 0.4081786870956421, "learning_rate": 9.801053042696977e-05, "loss": 0.0634, "step": 13560 }, { "grad_norm": 0.6182920932769775, "learning_rate": 9.800591004261388e-05, "loss": 0.0633, "step": 13570 }, { "grad_norm": 0.47403135895729065, "learning_rate": 9.800128440840385e-05, "loss": 0.0539, "step": 13580 }, { "grad_norm": 1.0080268383026123, "learning_rate": 9.799665352484552e-05, "loss": 0.0701, "step": 13590 }, { "grad_norm": 0.47135263681411743, "learning_rate": 9.799201739244532e-05, "loss": 0.0927, "step": 13600 }, { "grad_norm": 0.4001418352127075, "learning_rate": 9.798737601171025e-05, "loss": 0.0613, "step": 13610 }, { "grad_norm": 0.45838114619255066, "learning_rate": 9.79827293831479e-05, "loss": 0.0722, "step": 13620 }, { "grad_norm": 0.5279148817062378, "learning_rate": 9.797807750726638e-05, "loss": 0.0563, "step": 13630 }, { "grad_norm": 0.6085997819900513, "learning_rate": 9.797342038457446e-05, "loss": 0.0748, "step": 13640 }, { "grad_norm": 0.4740087687969208, "learning_rate": 9.796875801558141e-05, "loss": 0.068, "step": 13650 }, { "grad_norm": 0.5918585062026978, "learning_rate": 9.79640904007971e-05, "loss": 0.0686, "step": 13660 }, { "grad_norm": 0.4426991939544678, "learning_rate": 9.795941754073199e-05, "loss": 0.0534, "step": 13670 }, { "grad_norm": 0.9535490274429321, "learning_rate": 9.795473943589705e-05, "loss": 0.0602, "step": 13680 }, { "grad_norm": 0.5340253114700317, "learning_rate": 9.795005608680394e-05, "loss": 0.061, "step": 13690 }, { "grad_norm": 0.37351521849632263, "learning_rate": 9.794536749396477e-05, "loss": 0.0597, "step": 13700 }, { "grad_norm": 0.34907108545303345, "learning_rate": 9.79406736578923e-05, "loss": 0.0567, "step": 13710 }, { "grad_norm": 0.851003110408783, "learning_rate": 9.793597457909984e-05, "loss": 0.0562, "step": 13720 }, { "grad_norm": 0.47731420397758484, "learning_rate": 9.793127025810127e-05, "loss": 0.0706, "step": 13730 }, { "grad_norm": 0.6386615633964539, "learning_rate": 9.792656069541104e-05, "loss": 0.0834, "step": 13740 }, { "grad_norm": 0.6976563930511475, "learning_rate": 9.79218458915442e-05, "loss": 0.0755, "step": 13750 }, { "grad_norm": 0.5987343788146973, "learning_rate": 9.791712584701634e-05, "loss": 0.0816, "step": 13760 }, { "grad_norm": 0.611303448677063, "learning_rate": 9.791240056234364e-05, "loss": 0.0567, "step": 13770 }, { "grad_norm": 0.4363289773464203, "learning_rate": 9.790767003804283e-05, "loss": 0.0575, "step": 13780 }, { "grad_norm": 0.5187733173370361, "learning_rate": 9.790293427463126e-05, "loss": 0.0672, "step": 13790 }, { "grad_norm": 0.666678249835968, "learning_rate": 9.789819327262684e-05, "loss": 0.0798, "step": 13800 }, { "grad_norm": 0.5982012748718262, "learning_rate": 9.7893447032548e-05, "loss": 0.0746, "step": 13810 }, { "grad_norm": 0.5238581299781799, "learning_rate": 9.78886955549138e-05, "loss": 0.0708, "step": 13820 }, { "grad_norm": 0.7655274271965027, "learning_rate": 9.788393884024387e-05, "loss": 0.0685, "step": 13830 }, { "grad_norm": 1.3008910417556763, "learning_rate": 9.787917688905836e-05, "loss": 0.0617, "step": 13840 }, { "grad_norm": 0.6349837183952332, "learning_rate": 9.787440970187807e-05, "loss": 0.0632, "step": 13850 }, { "grad_norm": 0.4226491451263428, "learning_rate": 9.786963727922429e-05, "loss": 0.0616, "step": 13860 }, { "grad_norm": 0.4039466381072998, "learning_rate": 9.786485962161897e-05, "loss": 0.0767, "step": 13870 }, { "grad_norm": 0.49670591950416565, "learning_rate": 9.786007672958455e-05, "loss": 0.0504, "step": 13880 }, { "grad_norm": 0.5465351939201355, "learning_rate": 9.78552886036441e-05, "loss": 0.0631, "step": 13890 }, { "grad_norm": 0.4244658350944519, "learning_rate": 9.785049524432124e-05, "loss": 0.0499, "step": 13900 }, { "grad_norm": 0.6589184999465942, "learning_rate": 9.784569665214016e-05, "loss": 0.0693, "step": 13910 }, { "grad_norm": 0.48177486658096313, "learning_rate": 9.784089282762563e-05, "loss": 0.0801, "step": 13920 }, { "grad_norm": 0.4205303490161896, "learning_rate": 9.7836083771303e-05, "loss": 0.0681, "step": 13930 }, { "grad_norm": 0.49322864413261414, "learning_rate": 9.783126948369817e-05, "loss": 0.0767, "step": 13940 }, { "grad_norm": 0.4928606450557709, "learning_rate": 9.78264499653376e-05, "loss": 0.0641, "step": 13950 }, { "grad_norm": 0.383420467376709, "learning_rate": 9.782162521674838e-05, "loss": 0.0745, "step": 13960 }, { "grad_norm": 0.39530643820762634, "learning_rate": 9.781679523845812e-05, "loss": 0.0774, "step": 13970 }, { "grad_norm": 0.3721444606781006, "learning_rate": 9.781196003099502e-05, "loss": 0.0618, "step": 13980 }, { "grad_norm": 0.4222901165485382, "learning_rate": 9.780711959488786e-05, "loss": 0.0669, "step": 13990 }, { "grad_norm": 0.457940936088562, "learning_rate": 9.780227393066599e-05, "loss": 0.0622, "step": 14000 }, { "grad_norm": 0.43081071972846985, "learning_rate": 9.77974230388593e-05, "loss": 0.0638, "step": 14010 }, { "grad_norm": 0.754679799079895, "learning_rate": 9.779256691999829e-05, "loss": 0.0736, "step": 14020 }, { "grad_norm": 0.6305056810379028, "learning_rate": 9.778770557461403e-05, "loss": 0.073, "step": 14030 }, { "grad_norm": 0.7078084945678711, "learning_rate": 9.778283900323812e-05, "loss": 0.0797, "step": 14040 }, { "grad_norm": 0.508226752281189, "learning_rate": 9.777796720640277e-05, "loss": 0.0632, "step": 14050 }, { "grad_norm": 0.3752935230731964, "learning_rate": 9.777309018464078e-05, "loss": 0.0722, "step": 14060 }, { "grad_norm": 0.5950247049331665, "learning_rate": 9.776820793848547e-05, "loss": 0.0662, "step": 14070 }, { "grad_norm": 0.5499175190925598, "learning_rate": 9.776332046847075e-05, "loss": 0.0684, "step": 14080 }, { "grad_norm": 0.5468037128448486, "learning_rate": 9.775842777513111e-05, "loss": 0.066, "step": 14090 }, { "grad_norm": 1.070455551147461, "learning_rate": 9.775352985900163e-05, "loss": 0.0903, "step": 14100 }, { "grad_norm": 0.6176468133926392, "learning_rate": 9.774862672061791e-05, "loss": 0.0906, "step": 14110 }, { "grad_norm": 0.6963502168655396, "learning_rate": 9.774371836051616e-05, "loss": 0.0733, "step": 14120 }, { "grad_norm": 0.527359127998352, "learning_rate": 9.773880477923315e-05, "loss": 0.0785, "step": 14130 }, { "grad_norm": 0.3619128465652466, "learning_rate": 9.773388597730623e-05, "loss": 0.0613, "step": 14140 }, { "grad_norm": 0.4538627862930298, "learning_rate": 9.77289619552733e-05, "loss": 0.0581, "step": 14150 }, { "grad_norm": 0.6056627631187439, "learning_rate": 9.772403271367285e-05, "loss": 0.0692, "step": 14160 }, { "grad_norm": 0.5178154110908508, "learning_rate": 9.771909825304396e-05, "loss": 0.0663, "step": 14170 }, { "grad_norm": 0.5481258034706116, "learning_rate": 9.771415857392619e-05, "loss": 0.0722, "step": 14180 }, { "grad_norm": 0.4433809816837311, "learning_rate": 9.770921367685978e-05, "loss": 0.072, "step": 14190 }, { "grad_norm": 0.6207618117332458, "learning_rate": 9.770426356238551e-05, "loss": 0.0683, "step": 14200 }, { "grad_norm": 0.4351861774921417, "learning_rate": 9.769930823104469e-05, "loss": 0.0667, "step": 14210 }, { "grad_norm": 0.38231581449508667, "learning_rate": 9.769434768337926e-05, "loss": 0.0634, "step": 14220 }, { "grad_norm": 0.47289204597473145, "learning_rate": 9.768938191993164e-05, "loss": 0.0613, "step": 14230 }, { "grad_norm": 0.4314984381198883, "learning_rate": 9.768441094124494e-05, "loss": 0.0619, "step": 14240 }, { "grad_norm": 0.36895066499710083, "learning_rate": 9.767943474786275e-05, "loss": 0.0671, "step": 14250 }, { "grad_norm": 0.37962818145751953, "learning_rate": 9.767445334032923e-05, "loss": 0.0609, "step": 14260 }, { "grad_norm": 0.4306883215904236, "learning_rate": 9.766946671918919e-05, "loss": 0.0657, "step": 14270 }, { "grad_norm": 0.5062089562416077, "learning_rate": 9.766447488498796e-05, "loss": 0.0578, "step": 14280 }, { "grad_norm": 0.44479650259017944, "learning_rate": 9.765947783827139e-05, "loss": 0.0575, "step": 14290 }, { "grad_norm": 1.0335114002227783, "learning_rate": 9.765447557958599e-05, "loss": 0.0687, "step": 14300 }, { "grad_norm": 0.5461196899414062, "learning_rate": 9.764946810947879e-05, "loss": 0.0582, "step": 14310 }, { "grad_norm": 0.6632774472236633, "learning_rate": 9.764445542849738e-05, "loss": 0.0737, "step": 14320 }, { "grad_norm": 0.450095534324646, "learning_rate": 9.763943753718998e-05, "loss": 0.0613, "step": 14330 }, { "grad_norm": 0.4254568815231323, "learning_rate": 9.76344144361053e-05, "loss": 0.0693, "step": 14340 }, { "grad_norm": 0.5214806199073792, "learning_rate": 9.762938612579269e-05, "loss": 0.0581, "step": 14350 }, { "grad_norm": 0.3791298270225525, "learning_rate": 9.762435260680202e-05, "loss": 0.0535, "step": 14360 }, { "grad_norm": 0.4676028788089752, "learning_rate": 9.761931387968373e-05, "loss": 0.0616, "step": 14370 }, { "grad_norm": 0.4979185163974762, "learning_rate": 9.76142699449889e-05, "loss": 0.0562, "step": 14380 }, { "grad_norm": 0.37111547589302063, "learning_rate": 9.760922080326908e-05, "loss": 0.0731, "step": 14390 }, { "grad_norm": 0.9749168753623962, "learning_rate": 9.760416645507644e-05, "loss": 0.0646, "step": 14400 }, { "grad_norm": 0.5710306763648987, "learning_rate": 9.759910690096375e-05, "loss": 0.0793, "step": 14410 }, { "grad_norm": 0.4286741614341736, "learning_rate": 9.759404214148429e-05, "loss": 0.0733, "step": 14420 }, { "grad_norm": 0.4225502908229828, "learning_rate": 9.758897217719191e-05, "loss": 0.0702, "step": 14430 }, { "grad_norm": 0.516467273235321, "learning_rate": 9.758389700864113e-05, "loss": 0.0559, "step": 14440 }, { "grad_norm": 0.593730092048645, "learning_rate": 9.757881663638688e-05, "loss": 0.0656, "step": 14450 }, { "grad_norm": 0.7159096002578735, "learning_rate": 9.757373106098478e-05, "loss": 0.0636, "step": 14460 }, { "grad_norm": 1.077877163887024, "learning_rate": 9.756864028299097e-05, "loss": 0.071, "step": 14470 }, { "grad_norm": 0.6062937378883362, "learning_rate": 9.75635443029622e-05, "loss": 0.0597, "step": 14480 }, { "grad_norm": 0.6102299094200134, "learning_rate": 9.755844312145572e-05, "loss": 0.0704, "step": 14490 }, { "grad_norm": 0.46640679240226746, "learning_rate": 9.755333673902941e-05, "loss": 0.0808, "step": 14500 }, { "grad_norm": 0.45040011405944824, "learning_rate": 9.75482251562417e-05, "loss": 0.0574, "step": 14510 }, { "grad_norm": 0.571245551109314, "learning_rate": 9.754310837365155e-05, "loss": 0.055, "step": 14520 }, { "grad_norm": 0.5192873477935791, "learning_rate": 9.753798639181856e-05, "loss": 0.0774, "step": 14530 }, { "grad_norm": 0.40372762084007263, "learning_rate": 9.753285921130286e-05, "loss": 0.0623, "step": 14540 }, { "grad_norm": 0.4268222749233246, "learning_rate": 9.752772683266512e-05, "loss": 0.0608, "step": 14550 }, { "grad_norm": 0.4468863010406494, "learning_rate": 9.752258925646665e-05, "loss": 0.0597, "step": 14560 }, { "grad_norm": 0.5402755737304688, "learning_rate": 9.751744648326926e-05, "loss": 0.0638, "step": 14570 }, { "grad_norm": 0.4854623079299927, "learning_rate": 9.751229851363536e-05, "loss": 0.0684, "step": 14580 }, { "grad_norm": 1.2535096406936646, "learning_rate": 9.750714534812793e-05, "loss": 0.0928, "step": 14590 }, { "grad_norm": 0.5370067358016968, "learning_rate": 9.750198698731053e-05, "loss": 0.0839, "step": 14600 }, { "grad_norm": 0.4381142556667328, "learning_rate": 9.749682343174722e-05, "loss": 0.0637, "step": 14610 }, { "grad_norm": 0.5743894577026367, "learning_rate": 9.749165468200272e-05, "loss": 0.0712, "step": 14620 }, { "grad_norm": 0.7896475195884705, "learning_rate": 9.748648073864229e-05, "loss": 0.0861, "step": 14630 }, { "grad_norm": 0.4465954303741455, "learning_rate": 9.748130160223168e-05, "loss": 0.0688, "step": 14640 }, { "grad_norm": 0.5957369804382324, "learning_rate": 9.747611727333734e-05, "loss": 0.0759, "step": 14650 }, { "grad_norm": 0.3893662095069885, "learning_rate": 9.74709277525262e-05, "loss": 0.0652, "step": 14660 }, { "grad_norm": 0.6695943474769592, "learning_rate": 9.746573304036576e-05, "loss": 0.0676, "step": 14670 }, { "grad_norm": 0.4802410900592804, "learning_rate": 9.746053313742412e-05, "loss": 0.06, "step": 14680 }, { "grad_norm": 0.5460699200630188, "learning_rate": 9.745532804426994e-05, "loss": 0.0678, "step": 14690 }, { "grad_norm": 0.33392298221588135, "learning_rate": 9.745011776147242e-05, "loss": 0.0616, "step": 14700 }, { "grad_norm": 0.6957206726074219, "learning_rate": 9.744490228960138e-05, "loss": 0.0782, "step": 14710 }, { "grad_norm": 0.6632285714149475, "learning_rate": 9.743968162922713e-05, "loss": 0.0646, "step": 14720 }, { "grad_norm": 0.44090837240219116, "learning_rate": 9.743445578092064e-05, "loss": 0.075, "step": 14730 }, { "grad_norm": 0.7940802574157715, "learning_rate": 9.742922474525338e-05, "loss": 0.0652, "step": 14740 }, { "grad_norm": 0.6170717477798462, "learning_rate": 9.742398852279741e-05, "loss": 0.066, "step": 14750 }, { "grad_norm": 0.475851446390152, "learning_rate": 9.741874711412535e-05, "loss": 0.0754, "step": 14760 }, { "grad_norm": 0.40305888652801514, "learning_rate": 9.741350051981042e-05, "loss": 0.0584, "step": 14770 }, { "grad_norm": 0.6725048422813416, "learning_rate": 9.740824874042633e-05, "loss": 0.0629, "step": 14780 }, { "grad_norm": 0.3797786831855774, "learning_rate": 9.740299177654746e-05, "loss": 0.0533, "step": 14790 }, { "grad_norm": 0.5780371427536011, "learning_rate": 9.739772962874867e-05, "loss": 0.0632, "step": 14800 }, { "grad_norm": 0.37708011269569397, "learning_rate": 9.739246229760541e-05, "loss": 0.0593, "step": 14810 }, { "grad_norm": 0.6654151082038879, "learning_rate": 9.738718978369376e-05, "loss": 0.0662, "step": 14820 }, { "grad_norm": 0.5030716061592102, "learning_rate": 9.738191208759025e-05, "loss": 0.0714, "step": 14830 }, { "grad_norm": 0.44349151849746704, "learning_rate": 9.73766292098721e-05, "loss": 0.0788, "step": 14840 }, { "grad_norm": 0.6203278303146362, "learning_rate": 9.737134115111699e-05, "loss": 0.0757, "step": 14850 }, { "grad_norm": 0.5934670567512512, "learning_rate": 9.736604791190323e-05, "loss": 0.0575, "step": 14860 }, { "grad_norm": 0.3421987295150757, "learning_rate": 9.73607494928097e-05, "loss": 0.0535, "step": 14870 }, { "grad_norm": 0.4524281919002533, "learning_rate": 9.735544589441581e-05, "loss": 0.0681, "step": 14880 }, { "grad_norm": 0.36016014218330383, "learning_rate": 9.735013711730154e-05, "loss": 0.0659, "step": 14890 }, { "grad_norm": 0.4456963539123535, "learning_rate": 9.734482316204747e-05, "loss": 0.0442, "step": 14900 }, { "grad_norm": 0.4593248963356018, "learning_rate": 9.733950402923473e-05, "loss": 0.0615, "step": 14910 }, { "grad_norm": 0.48529455065727234, "learning_rate": 9.7334179719445e-05, "loss": 0.056, "step": 14920 }, { "grad_norm": 0.554641604423523, "learning_rate": 9.732885023326053e-05, "loss": 0.0553, "step": 14930 }, { "grad_norm": 0.43606460094451904, "learning_rate": 9.732351557126418e-05, "loss": 0.0768, "step": 14940 }, { "grad_norm": 0.5901069641113281, "learning_rate": 9.731817573403929e-05, "loss": 0.0776, "step": 14950 }, { "grad_norm": 0.4221612215042114, "learning_rate": 9.731283072216985e-05, "loss": 0.0634, "step": 14960 }, { "grad_norm": 0.5984055399894714, "learning_rate": 9.730748053624039e-05, "loss": 0.0594, "step": 14970 }, { "grad_norm": 0.4820452928543091, "learning_rate": 9.730212517683598e-05, "loss": 0.0793, "step": 14980 }, { "grad_norm": 0.5495629906654358, "learning_rate": 9.729676464454228e-05, "loss": 0.0566, "step": 14990 }, { "grad_norm": 0.46277502179145813, "learning_rate": 9.72913989399455e-05, "loss": 0.0611, "step": 15000 }, { "grad_norm": 0.4237927794456482, "learning_rate": 9.728602806363242e-05, "loss": 0.0518, "step": 15010 }, { "grad_norm": 0.4453878700733185, "learning_rate": 9.728065201619043e-05, "loss": 0.0649, "step": 15020 }, { "grad_norm": 0.45254477858543396, "learning_rate": 9.727527079820742e-05, "loss": 0.0474, "step": 15030 }, { "grad_norm": 0.4424811005592346, "learning_rate": 9.726988441027186e-05, "loss": 0.0612, "step": 15040 }, { "grad_norm": 0.545566201210022, "learning_rate": 9.726449285297281e-05, "loss": 0.0687, "step": 15050 }, { "grad_norm": 0.36009615659713745, "learning_rate": 9.72590961268999e-05, "loss": 0.0618, "step": 15060 }, { "grad_norm": 0.44177576899528503, "learning_rate": 9.725369423264328e-05, "loss": 0.0631, "step": 15070 }, { "grad_norm": 0.4511999785900116, "learning_rate": 9.72482871707937e-05, "loss": 0.0564, "step": 15080 }, { "grad_norm": 0.43395528197288513, "learning_rate": 9.724287494194247e-05, "loss": 0.0636, "step": 15090 }, { "grad_norm": 0.5359119772911072, "learning_rate": 9.723745754668147e-05, "loss": 0.0649, "step": 15100 }, { "grad_norm": 0.4999174177646637, "learning_rate": 9.723203498560313e-05, "loss": 0.0589, "step": 15110 }, { "grad_norm": 0.36744269728660583, "learning_rate": 9.722660725930046e-05, "loss": 0.0554, "step": 15120 }, { "grad_norm": 0.36687612533569336, "learning_rate": 9.722117436836702e-05, "loss": 0.0516, "step": 15130 }, { "grad_norm": 0.4039219617843628, "learning_rate": 9.721573631339696e-05, "loss": 0.0396, "step": 15140 }, { "grad_norm": 0.42915472388267517, "learning_rate": 9.721029309498494e-05, "loss": 0.0741, "step": 15150 }, { "grad_norm": 0.5215663313865662, "learning_rate": 9.720484471372627e-05, "loss": 0.0683, "step": 15160 }, { "grad_norm": 0.523251473903656, "learning_rate": 9.719939117021673e-05, "loss": 0.0758, "step": 15170 }, { "grad_norm": 0.6232030987739563, "learning_rate": 9.719393246505275e-05, "loss": 0.0649, "step": 15180 }, { "grad_norm": 0.687139093875885, "learning_rate": 9.718846859883128e-05, "loss": 0.0686, "step": 15190 }, { "grad_norm": 1.0318248271942139, "learning_rate": 9.718299957214982e-05, "loss": 0.0667, "step": 15200 }, { "grad_norm": 0.574714183807373, "learning_rate": 9.717752538560646e-05, "loss": 0.0602, "step": 15210 }, { "grad_norm": 0.6228794455528259, "learning_rate": 9.717204603979986e-05, "loss": 0.0581, "step": 15220 }, { "grad_norm": 0.45522579550743103, "learning_rate": 9.716656153532922e-05, "loss": 0.0594, "step": 15230 }, { "grad_norm": 0.46426939964294434, "learning_rate": 9.716107187279434e-05, "loss": 0.0553, "step": 15240 }, { "grad_norm": 0.6246171593666077, "learning_rate": 9.715557705279555e-05, "loss": 0.0744, "step": 15250 }, { "grad_norm": 0.6421952247619629, "learning_rate": 9.715007707593372e-05, "loss": 0.0783, "step": 15260 }, { "grad_norm": 0.5445294976234436, "learning_rate": 9.714457194281036e-05, "loss": 0.0617, "step": 15270 }, { "grad_norm": 0.34358540177345276, "learning_rate": 9.713906165402751e-05, "loss": 0.0556, "step": 15280 }, { "grad_norm": 0.5702515244483948, "learning_rate": 9.713354621018774e-05, "loss": 0.0738, "step": 15290 }, { "grad_norm": 0.6525587439537048, "learning_rate": 9.712802561189422e-05, "loss": 0.0749, "step": 15300 }, { "grad_norm": 0.3794575333595276, "learning_rate": 9.712249985975069e-05, "loss": 0.0587, "step": 15310 }, { "grad_norm": 0.4268910586833954, "learning_rate": 9.71169689543614e-05, "loss": 0.0575, "step": 15320 }, { "grad_norm": 0.5217090249061584, "learning_rate": 9.711143289633123e-05, "loss": 0.0554, "step": 15330 }, { "grad_norm": 0.38544389605522156, "learning_rate": 9.710589168626561e-05, "loss": 0.0787, "step": 15340 }, { "grad_norm": 0.4252174496650696, "learning_rate": 9.710034532477048e-05, "loss": 0.0735, "step": 15350 }, { "grad_norm": 0.49746984243392944, "learning_rate": 9.709479381245239e-05, "loss": 0.0648, "step": 15360 }, { "grad_norm": 0.49182116985321045, "learning_rate": 9.708923714991847e-05, "loss": 0.0618, "step": 15370 }, { "grad_norm": 0.35220399498939514, "learning_rate": 9.708367533777638e-05, "loss": 0.0632, "step": 15380 }, { "grad_norm": 0.558837890625, "learning_rate": 9.707810837663431e-05, "loss": 0.0531, "step": 15390 }, { "grad_norm": 0.5074422359466553, "learning_rate": 9.707253626710113e-05, "loss": 0.0609, "step": 15400 }, { "grad_norm": 0.4580691158771515, "learning_rate": 9.706695900978613e-05, "loss": 0.0636, "step": 15410 }, { "grad_norm": 0.6231019496917725, "learning_rate": 9.706137660529926e-05, "loss": 0.075, "step": 15420 }, { "grad_norm": 0.4000304937362671, "learning_rate": 9.705578905425101e-05, "loss": 0.0829, "step": 15430 }, { "grad_norm": 0.38613268733024597, "learning_rate": 9.705019635725241e-05, "loss": 0.0706, "step": 15440 }, { "grad_norm": 0.5060884952545166, "learning_rate": 9.704459851491508e-05, "loss": 0.0792, "step": 15450 }, { "grad_norm": 0.4013938307762146, "learning_rate": 9.703899552785118e-05, "loss": 0.0539, "step": 15460 }, { "grad_norm": 0.5400416851043701, "learning_rate": 9.703338739667346e-05, "loss": 0.064, "step": 15470 }, { "grad_norm": 0.539276123046875, "learning_rate": 9.70277741219952e-05, "loss": 0.0722, "step": 15480 }, { "grad_norm": 0.729126513004303, "learning_rate": 9.702215570443027e-05, "loss": 0.0665, "step": 15490 }, { "grad_norm": 0.553185760974884, "learning_rate": 9.701653214459309e-05, "loss": 0.0578, "step": 15500 }, { "grad_norm": 0.3713013231754303, "learning_rate": 9.701090344309865e-05, "loss": 0.0673, "step": 15510 }, { "grad_norm": 0.5616557002067566, "learning_rate": 9.700526960056247e-05, "loss": 0.0519, "step": 15520 }, { "grad_norm": 0.525009274482727, "learning_rate": 9.699963061760068e-05, "loss": 0.054, "step": 15530 }, { "grad_norm": 0.6400655508041382, "learning_rate": 9.699398649482997e-05, "loss": 0.058, "step": 15540 }, { "grad_norm": 0.4621807932853699, "learning_rate": 9.698833723286753e-05, "loss": 0.0491, "step": 15550 }, { "grad_norm": 0.4729324281215668, "learning_rate": 9.698268283233118e-05, "loss": 0.0647, "step": 15560 }, { "grad_norm": 0.39773082733154297, "learning_rate": 9.697702329383929e-05, "loss": 0.0591, "step": 15570 }, { "grad_norm": 0.373037725687027, "learning_rate": 9.697135861801074e-05, "loss": 0.0685, "step": 15580 }, { "grad_norm": 0.44337165355682373, "learning_rate": 9.696568880546505e-05, "loss": 0.0607, "step": 15590 }, { "grad_norm": 0.35969704389572144, "learning_rate": 9.696001385682223e-05, "loss": 0.0569, "step": 15600 }, { "grad_norm": 0.45952779054641724, "learning_rate": 9.695433377270291e-05, "loss": 0.0797, "step": 15610 }, { "grad_norm": 0.4325565695762634, "learning_rate": 9.694864855372824e-05, "loss": 0.0634, "step": 15620 }, { "grad_norm": 0.5210190415382385, "learning_rate": 9.694295820051995e-05, "loss": 0.0631, "step": 15630 }, { "grad_norm": 0.5447579622268677, "learning_rate": 9.693726271370032e-05, "loss": 0.0574, "step": 15640 }, { "grad_norm": 0.4191785454750061, "learning_rate": 9.693156209389221e-05, "loss": 0.052, "step": 15650 }, { "grad_norm": 0.40743371844291687, "learning_rate": 9.692585634171905e-05, "loss": 0.0708, "step": 15660 }, { "grad_norm": 0.4018363654613495, "learning_rate": 9.692014545780476e-05, "loss": 0.0653, "step": 15670 }, { "grad_norm": 0.4123808741569519, "learning_rate": 9.691442944277393e-05, "loss": 0.0567, "step": 15680 }, { "grad_norm": 0.4703851342201233, "learning_rate": 9.690870829725162e-05, "loss": 0.0719, "step": 15690 }, { "grad_norm": 0.9270809292793274, "learning_rate": 9.69029820218635e-05, "loss": 0.0719, "step": 15700 }, { "grad_norm": 0.405396968126297, "learning_rate": 9.689725061723579e-05, "loss": 0.0665, "step": 15710 }, { "grad_norm": 0.4319874346256256, "learning_rate": 9.689151408399527e-05, "loss": 0.0628, "step": 15720 }, { "grad_norm": 0.4881707727909088, "learning_rate": 9.688577242276924e-05, "loss": 0.0761, "step": 15730 }, { "grad_norm": 0.3597528338432312, "learning_rate": 9.688002563418566e-05, "loss": 0.0579, "step": 15740 }, { "grad_norm": 0.6190440654754639, "learning_rate": 9.687427371887293e-05, "loss": 0.0656, "step": 15750 }, { "grad_norm": 0.42621079087257385, "learning_rate": 9.686851667746012e-05, "loss": 0.0568, "step": 15760 }, { "grad_norm": 0.6161486506462097, "learning_rate": 9.686275451057677e-05, "loss": 0.0695, "step": 15770 }, { "grad_norm": 0.4850216805934906, "learning_rate": 9.685698721885308e-05, "loss": 0.0607, "step": 15780 }, { "grad_norm": 0.5054264068603516, "learning_rate": 9.68512148029197e-05, "loss": 0.0726, "step": 15790 }, { "grad_norm": 0.5032316446304321, "learning_rate": 9.684543726340791e-05, "loss": 0.0498, "step": 15800 }, { "grad_norm": 0.554323673248291, "learning_rate": 9.683965460094952e-05, "loss": 0.0568, "step": 15810 }, { "grad_norm": 0.6761610507965088, "learning_rate": 9.683386681617694e-05, "loss": 0.0664, "step": 15820 }, { "grad_norm": 0.4767880439758301, "learning_rate": 9.68280739097231e-05, "loss": 0.0562, "step": 15830 }, { "grad_norm": 0.3917367458343506, "learning_rate": 9.682227588222148e-05, "loss": 0.0587, "step": 15840 }, { "grad_norm": 0.4945228099822998, "learning_rate": 9.681647273430618e-05, "loss": 0.0624, "step": 15850 }, { "grad_norm": 0.3876928985118866, "learning_rate": 9.681066446661182e-05, "loss": 0.0613, "step": 15860 }, { "grad_norm": 0.4603184461593628, "learning_rate": 9.680485107977357e-05, "loss": 0.0644, "step": 15870 }, { "grad_norm": 0.7491199374198914, "learning_rate": 9.679903257442716e-05, "loss": 0.0743, "step": 15880 }, { "grad_norm": 0.4556189477443695, "learning_rate": 9.679320895120891e-05, "loss": 0.0679, "step": 15890 }, { "grad_norm": 0.39164629578590393, "learning_rate": 9.67873802107557e-05, "loss": 0.0598, "step": 15900 }, { "grad_norm": 0.48299264907836914, "learning_rate": 9.67815463537049e-05, "loss": 0.0696, "step": 15910 }, { "grad_norm": 0.4862929880619049, "learning_rate": 9.677570738069457e-05, "loss": 0.0699, "step": 15920 }, { "grad_norm": 0.4747142493724823, "learning_rate": 9.676986329236318e-05, "loss": 0.0672, "step": 15930 }, { "grad_norm": 0.33581072092056274, "learning_rate": 9.676401408934987e-05, "loss": 0.054, "step": 15940 }, { "grad_norm": 0.5448657870292664, "learning_rate": 9.675815977229428e-05, "loss": 0.0611, "step": 15950 }, { "grad_norm": 0.44726455211639404, "learning_rate": 9.675230034183664e-05, "loss": 0.0504, "step": 15960 }, { "grad_norm": 0.5225932002067566, "learning_rate": 9.674643579861773e-05, "loss": 0.0749, "step": 15970 }, { "grad_norm": 0.672227680683136, "learning_rate": 9.674056614327886e-05, "loss": 0.0499, "step": 15980 }, { "grad_norm": 0.5849981307983398, "learning_rate": 9.673469137646198e-05, "loss": 0.0721, "step": 15990 }, { "grad_norm": 0.35754793882369995, "learning_rate": 9.67288114988095e-05, "loss": 0.0533, "step": 16000 }, { "grad_norm": 0.603827953338623, "learning_rate": 9.672292651096447e-05, "loss": 0.0499, "step": 16010 }, { "grad_norm": 0.4250597357749939, "learning_rate": 9.671703641357042e-05, "loss": 0.0557, "step": 16020 }, { "grad_norm": 0.4038894474506378, "learning_rate": 9.67111412072715e-05, "loss": 0.0597, "step": 16030 }, { "grad_norm": 0.4196789860725403, "learning_rate": 9.670524089271242e-05, "loss": 0.0504, "step": 16040 }, { "grad_norm": 0.5581070184707642, "learning_rate": 9.669933547053842e-05, "loss": 0.0537, "step": 16050 }, { "grad_norm": 0.42782896757125854, "learning_rate": 9.669342494139531e-05, "loss": 0.0546, "step": 16060 }, { "grad_norm": 0.45520004630088806, "learning_rate": 9.668750930592943e-05, "loss": 0.056, "step": 16070 }, { "grad_norm": 0.48746457695961, "learning_rate": 9.668158856478775e-05, "loss": 0.0602, "step": 16080 }, { "grad_norm": 0.8867878317832947, "learning_rate": 9.66756627186177e-05, "loss": 0.0656, "step": 16090 }, { "grad_norm": 0.5778855085372925, "learning_rate": 9.666973176806737e-05, "loss": 0.0746, "step": 16100 }, { "grad_norm": 0.4991523325443268, "learning_rate": 9.666379571378534e-05, "loss": 0.0763, "step": 16110 }, { "grad_norm": 0.5980634093284607, "learning_rate": 9.665785455642076e-05, "loss": 0.2495, "step": 16120 }, { "grad_norm": 0.936424970626831, "learning_rate": 9.665190829662337e-05, "loss": 0.0823, "step": 16130 }, { "grad_norm": 0.47950056195259094, "learning_rate": 9.664595693504342e-05, "loss": 0.0769, "step": 16140 }, { "grad_norm": 0.5090413093566895, "learning_rate": 9.664000047233175e-05, "loss": 0.067, "step": 16150 }, { "grad_norm": 0.35307127237319946, "learning_rate": 9.663403890913976e-05, "loss": 0.0658, "step": 16160 }, { "grad_norm": 0.825936496257782, "learning_rate": 9.662807224611938e-05, "loss": 0.0672, "step": 16170 }, { "grad_norm": 0.383257120847702, "learning_rate": 9.662210048392311e-05, "loss": 0.0623, "step": 16180 }, { "grad_norm": 0.5905084013938904, "learning_rate": 9.661612362320405e-05, "loss": 0.0685, "step": 16190 }, { "grad_norm": 0.4000242352485657, "learning_rate": 9.661014166461579e-05, "loss": 0.0643, "step": 16200 }, { "grad_norm": 0.6693078875541687, "learning_rate": 9.66041546088125e-05, "loss": 0.0763, "step": 16210 }, { "grad_norm": 0.5263623595237732, "learning_rate": 9.659816245644895e-05, "loss": 0.0637, "step": 16220 }, { "grad_norm": 0.82764732837677, "learning_rate": 9.65921652081804e-05, "loss": 0.0705, "step": 16230 }, { "grad_norm": 0.5808859467506409, "learning_rate": 9.658616286466271e-05, "loss": 0.0681, "step": 16240 }, { "grad_norm": 0.5006952285766602, "learning_rate": 9.65801554265523e-05, "loss": 0.0625, "step": 16250 }, { "grad_norm": 0.8335576057434082, "learning_rate": 9.657414289450612e-05, "loss": 0.0691, "step": 16260 }, { "grad_norm": 0.9541276693344116, "learning_rate": 9.656812526918171e-05, "loss": 0.0579, "step": 16270 }, { "grad_norm": 0.46892958879470825, "learning_rate": 9.656210255123712e-05, "loss": 0.0653, "step": 16280 }, { "grad_norm": 0.44056496024131775, "learning_rate": 9.6556074741331e-05, "loss": 0.0634, "step": 16290 }, { "grad_norm": 0.4673784077167511, "learning_rate": 9.655004184012256e-05, "loss": 0.0568, "step": 16300 }, { "grad_norm": 0.5953832268714905, "learning_rate": 9.654400384827152e-05, "loss": 0.0725, "step": 16310 }, { "grad_norm": 0.5132274627685547, "learning_rate": 9.653796076643818e-05, "loss": 0.0596, "step": 16320 }, { "grad_norm": 0.5724490880966187, "learning_rate": 9.653191259528344e-05, "loss": 0.0632, "step": 16330 }, { "grad_norm": 0.44019803404808044, "learning_rate": 9.65258593354687e-05, "loss": 0.0536, "step": 16340 }, { "grad_norm": 0.542707085609436, "learning_rate": 9.651980098765591e-05, "loss": 0.0676, "step": 16350 }, { "grad_norm": 0.4435291886329651, "learning_rate": 9.651373755250765e-05, "loss": 0.0678, "step": 16360 }, { "grad_norm": 0.508118212223053, "learning_rate": 9.650766903068697e-05, "loss": 0.0637, "step": 16370 }, { "grad_norm": 0.37570512294769287, "learning_rate": 9.650159542285753e-05, "loss": 0.0527, "step": 16380 }, { "grad_norm": 0.4512949585914612, "learning_rate": 9.649551672968353e-05, "loss": 0.0662, "step": 16390 }, { "grad_norm": 0.38036322593688965, "learning_rate": 9.648943295182973e-05, "loss": 0.0663, "step": 16400 }, { "grad_norm": 0.4657583236694336, "learning_rate": 9.648334408996144e-05, "loss": 0.0612, "step": 16410 }, { "grad_norm": 0.41100597381591797, "learning_rate": 9.647725014474452e-05, "loss": 0.0618, "step": 16420 }, { "grad_norm": 0.493971049785614, "learning_rate": 9.64711511168454e-05, "loss": 0.0716, "step": 16430 }, { "grad_norm": 0.7014585733413696, "learning_rate": 9.646504700693108e-05, "loss": 0.0677, "step": 16440 }, { "grad_norm": 0.9370844960212708, "learning_rate": 9.645893781566907e-05, "loss": 0.067, "step": 16450 }, { "grad_norm": 0.7314720153808594, "learning_rate": 9.645282354372744e-05, "loss": 0.0672, "step": 16460 }, { "grad_norm": 0.4017263650894165, "learning_rate": 9.644670419177491e-05, "loss": 0.0641, "step": 16470 }, { "grad_norm": 0.5840994119644165, "learning_rate": 9.644057976048062e-05, "loss": 0.0715, "step": 16480 }, { "grad_norm": 0.47422143816947937, "learning_rate": 9.643445025051435e-05, "loss": 0.062, "step": 16490 }, { "grad_norm": 0.5236076712608337, "learning_rate": 9.642831566254641e-05, "loss": 0.0528, "step": 16500 }, { "grad_norm": 1.0773011445999146, "learning_rate": 9.642217599724769e-05, "loss": 0.0609, "step": 16510 }, { "grad_norm": 0.4386278986930847, "learning_rate": 9.64160312552896e-05, "loss": 0.0705, "step": 16520 }, { "grad_norm": 0.4934091567993164, "learning_rate": 9.64098814373441e-05, "loss": 0.0625, "step": 16530 }, { "grad_norm": 0.47554418444633484, "learning_rate": 9.640372654408374e-05, "loss": 0.0597, "step": 16540 }, { "grad_norm": 0.4871263802051544, "learning_rate": 9.639756657618162e-05, "loss": 0.0509, "step": 16550 }, { "grad_norm": 0.4930332601070404, "learning_rate": 9.639140153431138e-05, "loss": 0.0646, "step": 16560 }, { "grad_norm": 0.33450642228126526, "learning_rate": 9.638523141914721e-05, "loss": 0.0581, "step": 16570 }, { "grad_norm": 0.4430426359176636, "learning_rate": 9.637905623136388e-05, "loss": 0.066, "step": 16580 }, { "grad_norm": 0.3794594407081604, "learning_rate": 9.637287597163669e-05, "loss": 0.0623, "step": 16590 }, { "grad_norm": 0.4270378053188324, "learning_rate": 9.63666906406415e-05, "loss": 0.0503, "step": 16600 }, { "grad_norm": 0.4964475929737091, "learning_rate": 9.636050023905473e-05, "loss": 0.0622, "step": 16610 }, { "grad_norm": 0.42663028836250305, "learning_rate": 9.635430476755336e-05, "loss": 0.0654, "step": 16620 }, { "grad_norm": 0.5231679081916809, "learning_rate": 9.63481042268149e-05, "loss": 0.0644, "step": 16630 }, { "grad_norm": 0.5167319178581238, "learning_rate": 9.634189861751745e-05, "loss": 0.0724, "step": 16640 }, { "grad_norm": 0.608886182308197, "learning_rate": 9.633568794033967e-05, "loss": 0.0688, "step": 16650 }, { "grad_norm": 0.8422916531562805, "learning_rate": 9.63294721959607e-05, "loss": 0.0552, "step": 16660 }, { "grad_norm": 0.3928010165691376, "learning_rate": 9.63232513850603e-05, "loss": 0.0783, "step": 16670 }, { "grad_norm": 0.5675823092460632, "learning_rate": 9.631702550831878e-05, "loss": 0.0535, "step": 16680 }, { "grad_norm": 0.4745244085788727, "learning_rate": 9.631079456641698e-05, "loss": 0.0622, "step": 16690 }, { "grad_norm": 0.623656690120697, "learning_rate": 9.630455856003632e-05, "loss": 0.0696, "step": 16700 }, { "grad_norm": 0.4622122347354889, "learning_rate": 9.629831748985876e-05, "loss": 0.0689, "step": 16710 }, { "grad_norm": 0.5853526592254639, "learning_rate": 9.629207135656679e-05, "loss": 0.0931, "step": 16720 }, { "grad_norm": 0.5506648421287537, "learning_rate": 9.628582016084353e-05, "loss": 0.0838, "step": 16730 }, { "grad_norm": 0.6101503968238831, "learning_rate": 9.627956390337254e-05, "loss": 0.0631, "step": 16740 }, { "grad_norm": 0.35668936371803284, "learning_rate": 9.627330258483802e-05, "loss": 0.0764, "step": 16750 }, { "grad_norm": 0.6170410513877869, "learning_rate": 9.62670362059247e-05, "loss": 0.0621, "step": 16760 }, { "grad_norm": 0.3310771882534027, "learning_rate": 9.626076476731786e-05, "loss": 0.0559, "step": 16770 }, { "grad_norm": 0.4803693890571594, "learning_rate": 9.625448826970336e-05, "loss": 0.0593, "step": 16780 }, { "grad_norm": 0.6891173124313354, "learning_rate": 9.624820671376755e-05, "loss": 0.0542, "step": 16790 }, { "grad_norm": 0.46921026706695557, "learning_rate": 9.62419201001974e-05, "loss": 0.055, "step": 16800 }, { "grad_norm": 0.41078442335128784, "learning_rate": 9.623562842968037e-05, "loss": 0.0539, "step": 16810 }, { "grad_norm": 0.45745325088500977, "learning_rate": 9.622933170290454e-05, "loss": 0.0551, "step": 16820 }, { "grad_norm": 0.6148324012756348, "learning_rate": 9.622302992055849e-05, "loss": 0.053, "step": 16830 }, { "grad_norm": 0.6949450373649597, "learning_rate": 9.62167230833314e-05, "loss": 0.0466, "step": 16840 }, { "grad_norm": 0.7387850284576416, "learning_rate": 9.621041119191295e-05, "loss": 0.0503, "step": 16850 }, { "grad_norm": 0.44668132066726685, "learning_rate": 9.620409424699342e-05, "loss": 0.0661, "step": 16860 }, { "grad_norm": 0.48900821805000305, "learning_rate": 9.619777224926359e-05, "loss": 0.0579, "step": 16870 }, { "grad_norm": 0.5305954813957214, "learning_rate": 9.619144519941485e-05, "loss": 0.0605, "step": 16880 }, { "grad_norm": 0.4742506146430969, "learning_rate": 9.618511309813912e-05, "loss": 0.0668, "step": 16890 }, { "grad_norm": 0.4920089840888977, "learning_rate": 9.617877594612886e-05, "loss": 0.0538, "step": 16900 }, { "grad_norm": 0.5135745406150818, "learning_rate": 9.617243374407707e-05, "loss": 0.0526, "step": 16910 }, { "grad_norm": 0.4017592966556549, "learning_rate": 9.616608649267736e-05, "loss": 0.0559, "step": 16920 }, { "grad_norm": 0.49310073256492615, "learning_rate": 9.615973419262385e-05, "loss": 0.0647, "step": 16930 }, { "grad_norm": 0.3966994285583496, "learning_rate": 9.615337684461119e-05, "loss": 0.0486, "step": 16940 }, { "grad_norm": 0.6389589905738831, "learning_rate": 9.614701444933465e-05, "loss": 0.0549, "step": 16950 }, { "grad_norm": 0.5202091932296753, "learning_rate": 9.614064700748997e-05, "loss": 0.0535, "step": 16960 }, { "grad_norm": 0.4467572271823883, "learning_rate": 9.613427451977352e-05, "loss": 0.0567, "step": 16970 }, { "grad_norm": 0.37787699699401855, "learning_rate": 9.612789698688216e-05, "loss": 0.0591, "step": 16980 }, { "grad_norm": 0.5429306030273438, "learning_rate": 9.612151440951334e-05, "loss": 0.0647, "step": 16990 }, { "grad_norm": 0.5708123445510864, "learning_rate": 9.611512678836506e-05, "loss": 0.0494, "step": 17000 }, { "grad_norm": 0.6612259149551392, "learning_rate": 9.610873412413584e-05, "loss": 0.066, "step": 17010 }, { "grad_norm": 0.4315076172351837, "learning_rate": 9.610233641752476e-05, "loss": 0.0706, "step": 17020 }, { "grad_norm": 0.4887668192386627, "learning_rate": 9.609593366923151e-05, "loss": 0.047, "step": 17030 }, { "grad_norm": 0.34524062275886536, "learning_rate": 9.608952587995625e-05, "loss": 0.0527, "step": 17040 }, { "grad_norm": 1.2388074398040771, "learning_rate": 9.608311305039972e-05, "loss": 0.0624, "step": 17050 }, { "grad_norm": 0.6836798787117004, "learning_rate": 9.607669518126326e-05, "loss": 0.0641, "step": 17060 }, { "grad_norm": 0.6838039755821228, "learning_rate": 9.607027227324866e-05, "loss": 0.0655, "step": 17070 }, { "grad_norm": 0.4755205512046814, "learning_rate": 9.606384432705837e-05, "loss": 0.0655, "step": 17080 }, { "grad_norm": 0.2829541563987732, "learning_rate": 9.60574113433953e-05, "loss": 0.0545, "step": 17090 }, { "grad_norm": 0.7074893116950989, "learning_rate": 9.6050973322963e-05, "loss": 0.058, "step": 17100 }, { "grad_norm": 0.4884391725063324, "learning_rate": 9.604453026646547e-05, "loss": 0.0745, "step": 17110 }, { "grad_norm": 0.6025058627128601, "learning_rate": 9.603808217460735e-05, "loss": 0.0562, "step": 17120 }, { "grad_norm": 0.4362201988697052, "learning_rate": 9.603162904809377e-05, "loss": 0.0527, "step": 17130 }, { "grad_norm": 0.3696657419204712, "learning_rate": 9.602517088763045e-05, "loss": 0.0474, "step": 17140 }, { "grad_norm": 0.4653837978839874, "learning_rate": 9.601870769392365e-05, "loss": 0.0595, "step": 17150 }, { "grad_norm": 0.4189734160900116, "learning_rate": 9.601223946768017e-05, "loss": 0.0587, "step": 17160 }, { "grad_norm": 0.3925131559371948, "learning_rate": 9.600576620960734e-05, "loss": 0.0698, "step": 17170 }, { "grad_norm": 0.536102294921875, "learning_rate": 9.599928792041308e-05, "loss": 0.0593, "step": 17180 }, { "grad_norm": 0.5128871202468872, "learning_rate": 9.599280460080587e-05, "loss": 0.0561, "step": 17190 }, { "grad_norm": 0.40781480073928833, "learning_rate": 9.59863162514947e-05, "loss": 0.0581, "step": 17200 }, { "grad_norm": 0.425335556268692, "learning_rate": 9.597982287318911e-05, "loss": 0.0525, "step": 17210 }, { "grad_norm": 0.5480942130088806, "learning_rate": 9.597332446659923e-05, "loss": 0.0583, "step": 17220 }, { "grad_norm": 0.4975042939186096, "learning_rate": 9.59668210324357e-05, "loss": 0.05, "step": 17230 }, { "grad_norm": 0.37769946455955505, "learning_rate": 9.596031257140974e-05, "loss": 0.0547, "step": 17240 }, { "grad_norm": 0.5325322151184082, "learning_rate": 9.59537990842331e-05, "loss": 0.06, "step": 17250 }, { "grad_norm": 0.4383968412876129, "learning_rate": 9.594728057161806e-05, "loss": 0.0575, "step": 17260 }, { "grad_norm": 0.4399649500846863, "learning_rate": 9.594075703427752e-05, "loss": 0.0586, "step": 17270 }, { "grad_norm": 0.37542688846588135, "learning_rate": 9.593422847292486e-05, "loss": 0.0519, "step": 17280 }, { "grad_norm": 0.4711669087409973, "learning_rate": 9.592769488827402e-05, "loss": 0.0664, "step": 17290 }, { "grad_norm": 0.4922937750816345, "learning_rate": 9.592115628103952e-05, "loss": 0.0561, "step": 17300 }, { "grad_norm": 0.6039365530014038, "learning_rate": 9.591461265193643e-05, "loss": 0.0586, "step": 17310 }, { "grad_norm": 0.5368059277534485, "learning_rate": 9.590806400168032e-05, "loss": 0.0478, "step": 17320 }, { "grad_norm": 0.48309627175331116, "learning_rate": 9.590151033098735e-05, "loss": 0.0601, "step": 17330 }, { "grad_norm": 0.6760925054550171, "learning_rate": 9.589495164057423e-05, "loss": 0.0571, "step": 17340 }, { "grad_norm": 0.4273609519004822, "learning_rate": 9.58883879311582e-05, "loss": 0.0529, "step": 17350 }, { "grad_norm": 0.45452383160591125, "learning_rate": 9.588181920345705e-05, "loss": 0.0648, "step": 17360 }, { "grad_norm": 0.4177566468715668, "learning_rate": 9.587524545818913e-05, "loss": 0.0548, "step": 17370 }, { "grad_norm": 0.31606802344322205, "learning_rate": 9.586866669607335e-05, "loss": 0.0502, "step": 17380 }, { "grad_norm": 0.3305969536304474, "learning_rate": 9.586208291782915e-05, "loss": 0.0467, "step": 17390 }, { "grad_norm": 0.44323110580444336, "learning_rate": 9.58554941241765e-05, "loss": 0.0512, "step": 17400 }, { "grad_norm": 0.4314933717250824, "learning_rate": 9.584890031583596e-05, "loss": 0.0411, "step": 17410 }, { "grad_norm": 0.5693681240081787, "learning_rate": 9.584230149352861e-05, "loss": 0.06, "step": 17420 }, { "grad_norm": 0.637450098991394, "learning_rate": 9.58356976579761e-05, "loss": 0.0588, "step": 17430 }, { "grad_norm": 0.5062475204467773, "learning_rate": 9.58290888099006e-05, "loss": 0.0727, "step": 17440 }, { "grad_norm": 0.42289483547210693, "learning_rate": 9.582247495002486e-05, "loss": 0.0611, "step": 17450 }, { "grad_norm": 0.4147450029850006, "learning_rate": 9.581585607907214e-05, "loss": 0.0537, "step": 17460 }, { "grad_norm": 0.44731059670448303, "learning_rate": 9.580923219776628e-05, "loss": 0.0433, "step": 17470 }, { "grad_norm": 0.28521230816841125, "learning_rate": 9.580260330683167e-05, "loss": 0.0479, "step": 17480 }, { "grad_norm": 0.5068074464797974, "learning_rate": 9.579596940699322e-05, "loss": 0.054, "step": 17490 }, { "grad_norm": 0.3199733793735504, "learning_rate": 9.578933049897643e-05, "loss": 0.0472, "step": 17500 }, { "grad_norm": 0.3891640603542328, "learning_rate": 9.578268658350728e-05, "loss": 0.0589, "step": 17510 }, { "grad_norm": 0.5120144486427307, "learning_rate": 9.577603766131235e-05, "loss": 0.0527, "step": 17520 }, { "grad_norm": 0.3412737548351288, "learning_rate": 9.576938373311878e-05, "loss": 0.0493, "step": 17530 }, { "grad_norm": 0.5576606392860413, "learning_rate": 9.576272479965421e-05, "loss": 0.0633, "step": 17540 }, { "grad_norm": 0.4701825976371765, "learning_rate": 9.575606086164687e-05, "loss": 0.0509, "step": 17550 }, { "grad_norm": 0.4425925612449646, "learning_rate": 9.57493919198255e-05, "loss": 0.058, "step": 17560 }, { "grad_norm": 0.6969526410102844, "learning_rate": 9.57427179749194e-05, "loss": 0.0587, "step": 17570 }, { "grad_norm": 0.5200176239013672, "learning_rate": 9.573603902765846e-05, "loss": 0.0546, "step": 17580 }, { "grad_norm": 0.9411501884460449, "learning_rate": 9.572935507877304e-05, "loss": 0.0666, "step": 17590 }, { "grad_norm": 0.5400189757347107, "learning_rate": 9.57226661289941e-05, "loss": 0.0564, "step": 17600 }, { "grad_norm": 0.42382505536079407, "learning_rate": 9.571597217905315e-05, "loss": 0.0602, "step": 17610 }, { "grad_norm": 0.5571410655975342, "learning_rate": 9.57092732296822e-05, "loss": 0.0533, "step": 17620 }, { "grad_norm": 0.37942394614219666, "learning_rate": 9.570256928161385e-05, "loss": 0.0503, "step": 17630 }, { "grad_norm": 0.5316056609153748, "learning_rate": 9.569586033558126e-05, "loss": 0.0606, "step": 17640 }, { "grad_norm": 0.4075030982494354, "learning_rate": 9.568914639231807e-05, "loss": 0.0585, "step": 17650 }, { "grad_norm": 0.5536906719207764, "learning_rate": 9.568242745255852e-05, "loss": 0.0702, "step": 17660 }, { "grad_norm": 0.3182663023471832, "learning_rate": 9.567570351703739e-05, "loss": 0.0582, "step": 17670 }, { "grad_norm": 0.5421991348266602, "learning_rate": 9.566897458649001e-05, "loss": 0.063, "step": 17680 }, { "grad_norm": 0.4153401851654053, "learning_rate": 9.566224066165221e-05, "loss": 0.0554, "step": 17690 }, { "grad_norm": 0.48150894045829773, "learning_rate": 9.565550174326043e-05, "loss": 0.06, "step": 17700 }, { "grad_norm": 0.49738815426826477, "learning_rate": 9.564875783205162e-05, "loss": 0.0668, "step": 17710 }, { "grad_norm": 0.35858631134033203, "learning_rate": 9.564200892876328e-05, "loss": 0.0588, "step": 17720 }, { "grad_norm": 0.4668145477771759, "learning_rate": 9.563525503413348e-05, "loss": 0.0505, "step": 17730 }, { "grad_norm": 0.5451149940490723, "learning_rate": 9.562849614890079e-05, "loss": 0.0581, "step": 17740 }, { "grad_norm": 0.5510773062705994, "learning_rate": 9.562173227380436e-05, "loss": 0.0695, "step": 17750 }, { "grad_norm": 0.5236000418663025, "learning_rate": 9.561496340958389e-05, "loss": 0.0597, "step": 17760 }, { "grad_norm": 0.5128506422042847, "learning_rate": 9.560818955697959e-05, "loss": 0.0575, "step": 17770 }, { "grad_norm": 0.5163714289665222, "learning_rate": 9.560141071673228e-05, "loss": 0.0529, "step": 17780 }, { "grad_norm": 0.44092532992362976, "learning_rate": 9.559462688958323e-05, "loss": 0.0481, "step": 17790 }, { "grad_norm": 0.39476364850997925, "learning_rate": 9.558783807627434e-05, "loss": 0.0541, "step": 17800 }, { "grad_norm": 0.6173282861709595, "learning_rate": 9.558104427754801e-05, "loss": 0.064, "step": 17810 }, { "grad_norm": 0.57573401927948, "learning_rate": 9.557424549414722e-05, "loss": 0.0643, "step": 17820 }, { "grad_norm": 0.3682211637496948, "learning_rate": 9.556744172681546e-05, "loss": 0.048, "step": 17830 }, { "grad_norm": 0.4908471405506134, "learning_rate": 9.556063297629677e-05, "loss": 0.0577, "step": 17840 }, { "grad_norm": 0.42251524329185486, "learning_rate": 9.555381924333578e-05, "loss": 0.0554, "step": 17850 }, { "grad_norm": 0.5017793774604797, "learning_rate": 9.554700052867758e-05, "loss": 0.0575, "step": 17860 }, { "grad_norm": 0.5035987496376038, "learning_rate": 9.554017683306789e-05, "loss": 0.0555, "step": 17870 }, { "grad_norm": 0.36808109283447266, "learning_rate": 9.553334815725294e-05, "loss": 0.0586, "step": 17880 }, { "grad_norm": 0.6080914735794067, "learning_rate": 9.552651450197949e-05, "loss": 0.0644, "step": 17890 }, { "grad_norm": 0.6686139702796936, "learning_rate": 9.551967586799486e-05, "loss": 0.0481, "step": 17900 }, { "grad_norm": 0.8341468572616577, "learning_rate": 9.551283225604692e-05, "loss": 0.0661, "step": 17910 }, { "grad_norm": 0.4532622694969177, "learning_rate": 9.550598366688406e-05, "loss": 0.0655, "step": 17920 }, { "grad_norm": 0.6164504289627075, "learning_rate": 9.549913010125526e-05, "loss": 0.0709, "step": 17930 }, { "grad_norm": 0.5273715257644653, "learning_rate": 9.549227155990999e-05, "loss": 0.07, "step": 17940 }, { "grad_norm": 0.49360400438308716, "learning_rate": 9.548540804359828e-05, "loss": 0.0675, "step": 17950 }, { "grad_norm": 0.6348342895507812, "learning_rate": 9.547853955307077e-05, "loss": 0.0663, "step": 17960 }, { "grad_norm": 0.375787615776062, "learning_rate": 9.547166608907853e-05, "loss": 0.057, "step": 17970 }, { "grad_norm": 0.39592018723487854, "learning_rate": 9.546478765237326e-05, "loss": 0.0589, "step": 17980 }, { "grad_norm": 0.3740125596523285, "learning_rate": 9.545790424370715e-05, "loss": 0.038, "step": 17990 }, { "grad_norm": 0.493980348110199, "learning_rate": 9.5451015863833e-05, "loss": 0.0605, "step": 18000 }, { "grad_norm": 0.35274800658226013, "learning_rate": 9.544412251350408e-05, "loss": 0.0646, "step": 18010 }, { "grad_norm": 0.4889039397239685, "learning_rate": 9.543722419347422e-05, "loss": 0.0567, "step": 18020 }, { "grad_norm": 0.47698938846588135, "learning_rate": 9.543032090449788e-05, "loss": 0.0708, "step": 18030 }, { "grad_norm": 0.42650535702705383, "learning_rate": 9.542341264732992e-05, "loss": 0.0652, "step": 18040 }, { "grad_norm": 0.4860661029815674, "learning_rate": 9.541649942272585e-05, "loss": 0.0618, "step": 18050 }, { "grad_norm": 0.42616918683052063, "learning_rate": 9.54095812314417e-05, "loss": 0.0595, "step": 18060 }, { "grad_norm": 0.5421475172042847, "learning_rate": 9.540265807423401e-05, "loss": 0.0627, "step": 18070 }, { "grad_norm": 0.46351298689842224, "learning_rate": 9.53957299518599e-05, "loss": 0.0565, "step": 18080 }, { "grad_norm": 0.3697303533554077, "learning_rate": 9.5388796865077e-05, "loss": 0.0568, "step": 18090 }, { "grad_norm": 0.6281103491783142, "learning_rate": 9.538185881464353e-05, "loss": 0.0638, "step": 18100 }, { "grad_norm": 0.5261610746383667, "learning_rate": 9.537491580131821e-05, "loss": 0.0547, "step": 18110 }, { "grad_norm": 0.6082879304885864, "learning_rate": 9.53679678258603e-05, "loss": 0.0641, "step": 18120 }, { "grad_norm": 0.6527595520019531, "learning_rate": 9.536101488902966e-05, "loss": 0.0705, "step": 18130 }, { "grad_norm": 0.8859024047851562, "learning_rate": 9.535405699158663e-05, "loss": 0.0752, "step": 18140 }, { "grad_norm": 0.44088658690452576, "learning_rate": 9.53470941342921e-05, "loss": 0.0748, "step": 18150 }, { "grad_norm": 0.7530426979064941, "learning_rate": 9.534012631790756e-05, "loss": 0.0714, "step": 18160 }, { "grad_norm": 0.44644296169281006, "learning_rate": 9.533315354319494e-05, "loss": 0.0713, "step": 18170 }, { "grad_norm": 0.3377719819545746, "learning_rate": 9.532617581091682e-05, "loss": 0.0528, "step": 18180 }, { "grad_norm": 0.5088037848472595, "learning_rate": 9.531919312183629e-05, "loss": 0.0664, "step": 18190 }, { "grad_norm": 0.6188068389892578, "learning_rate": 9.531220547671688e-05, "loss": 0.0601, "step": 18200 }, { "grad_norm": 0.501137375831604, "learning_rate": 9.530521287632285e-05, "loss": 0.0529, "step": 18210 }, { "grad_norm": 0.42572858929634094, "learning_rate": 9.529821532141884e-05, "loss": 0.0479, "step": 18220 }, { "grad_norm": 0.2855079770088196, "learning_rate": 9.52912128127701e-05, "loss": 0.0451, "step": 18230 }, { "grad_norm": 0.511588990688324, "learning_rate": 9.528420535114244e-05, "loss": 0.0604, "step": 18240 }, { "grad_norm": 0.5652008056640625, "learning_rate": 9.527719293730215e-05, "loss": 0.0598, "step": 18250 }, { "grad_norm": 0.43536725640296936, "learning_rate": 9.527017557201611e-05, "loss": 0.0545, "step": 18260 }, { "grad_norm": 0.4741955101490021, "learning_rate": 9.526315325605176e-05, "loss": 0.0602, "step": 18270 }, { "grad_norm": 0.3859561085700989, "learning_rate": 9.525612599017699e-05, "loss": 0.0497, "step": 18280 }, { "grad_norm": 0.712252676486969, "learning_rate": 9.524909377516033e-05, "loss": 0.0656, "step": 18290 }, { "grad_norm": 0.5428673028945923, "learning_rate": 9.524205661177081e-05, "loss": 0.0506, "step": 18300 }, { "grad_norm": 0.379881352186203, "learning_rate": 9.523501450077801e-05, "loss": 0.0568, "step": 18310 }, { "grad_norm": 0.5165907144546509, "learning_rate": 9.522796744295202e-05, "loss": 0.0576, "step": 18320 }, { "grad_norm": 0.44569650292396545, "learning_rate": 9.522091543906352e-05, "loss": 0.0498, "step": 18330 }, { "grad_norm": 0.48295456171035767, "learning_rate": 9.521385848988369e-05, "loss": 0.0613, "step": 18340 }, { "grad_norm": 0.37825751304626465, "learning_rate": 9.520679659618428e-05, "loss": 0.0657, "step": 18350 }, { "grad_norm": 0.6834061741828918, "learning_rate": 9.519972975873754e-05, "loss": 0.0543, "step": 18360 }, { "grad_norm": 0.8941804766654968, "learning_rate": 9.519265797831633e-05, "loss": 0.0577, "step": 18370 }, { "grad_norm": 0.4436318278312683, "learning_rate": 9.518558125569399e-05, "loss": 0.0498, "step": 18380 }, { "grad_norm": 1.1160297393798828, "learning_rate": 9.517849959164442e-05, "loss": 0.0583, "step": 18390 }, { "grad_norm": 0.475518137216568, "learning_rate": 9.517141298694205e-05, "loss": 0.0647, "step": 18400 }, { "grad_norm": 0.8305209875106812, "learning_rate": 9.516432144236188e-05, "loss": 0.0672, "step": 18410 }, { "grad_norm": 0.4693242013454437, "learning_rate": 9.515722495867941e-05, "loss": 0.0704, "step": 18420 }, { "grad_norm": 0.3914080262184143, "learning_rate": 9.515012353667072e-05, "loss": 0.0621, "step": 18430 }, { "grad_norm": 0.4755915403366089, "learning_rate": 9.51430171771124e-05, "loss": 0.0513, "step": 18440 }, { "grad_norm": 0.4724309742450714, "learning_rate": 9.513590588078159e-05, "loss": 0.0823, "step": 18450 }, { "grad_norm": 0.6521891951560974, "learning_rate": 9.512878964845597e-05, "loss": 0.0709, "step": 18460 }, { "grad_norm": 0.5598706603050232, "learning_rate": 9.512166848091377e-05, "loss": 0.0681, "step": 18470 }, { "grad_norm": 0.37383636832237244, "learning_rate": 9.511454237893376e-05, "loss": 0.0583, "step": 18480 }, { "grad_norm": 0.5613808035850525, "learning_rate": 9.51074113432952e-05, "loss": 0.0555, "step": 18490 }, { "grad_norm": 0.550406277179718, "learning_rate": 9.510027537477797e-05, "loss": 0.0657, "step": 18500 }, { "grad_norm": 0.477683961391449, "learning_rate": 9.509313447416242e-05, "loss": 0.0529, "step": 18510 }, { "grad_norm": 0.4503922760486603, "learning_rate": 9.508598864222949e-05, "loss": 0.0525, "step": 18520 }, { "grad_norm": 0.6766983866691589, "learning_rate": 9.507883787976062e-05, "loss": 0.0585, "step": 18530 }, { "grad_norm": 0.39938393235206604, "learning_rate": 9.507168218753781e-05, "loss": 0.0523, "step": 18540 }, { "grad_norm": 0.3441644608974457, "learning_rate": 9.506452156634362e-05, "loss": 0.0579, "step": 18550 }, { "grad_norm": 0.6217454671859741, "learning_rate": 9.505735601696109e-05, "loss": 0.0744, "step": 18560 }, { "grad_norm": 0.7751559615135193, "learning_rate": 9.505018554017385e-05, "loss": 0.0738, "step": 18570 }, { "grad_norm": 0.564980149269104, "learning_rate": 9.504301013676604e-05, "loss": 0.0607, "step": 18580 }, { "grad_norm": 0.39157912135124207, "learning_rate": 9.503582980752238e-05, "loss": 0.0553, "step": 18590 }, { "grad_norm": 0.4909820258617401, "learning_rate": 9.502864455322809e-05, "loss": 0.0776, "step": 18600 }, { "grad_norm": 0.4499177038669586, "learning_rate": 9.502145437466891e-05, "loss": 0.0571, "step": 18610 }, { "grad_norm": 0.5241454243659973, "learning_rate": 9.501425927263116e-05, "loss": 0.0693, "step": 18620 }, { "grad_norm": 0.5931776165962219, "learning_rate": 9.500705924790172e-05, "loss": 0.0711, "step": 18630 }, { "grad_norm": 0.5789008736610413, "learning_rate": 9.499985430126794e-05, "loss": 0.0646, "step": 18640 }, { "grad_norm": 0.638454258441925, "learning_rate": 9.499264443351775e-05, "loss": 0.0672, "step": 18650 }, { "grad_norm": 0.4317209720611572, "learning_rate": 9.498542964543961e-05, "loss": 0.0605, "step": 18660 }, { "grad_norm": 0.5232232809066772, "learning_rate": 9.497820993782252e-05, "loss": 0.0659, "step": 18670 }, { "grad_norm": 0.46402204036712646, "learning_rate": 9.497098531145601e-05, "loss": 0.0544, "step": 18680 }, { "grad_norm": 0.4129827618598938, "learning_rate": 9.496375576713017e-05, "loss": 0.0562, "step": 18690 }, { "grad_norm": 0.32054412364959717, "learning_rate": 9.49565213056356e-05, "loss": 0.0489, "step": 18700 }, { "grad_norm": 0.5580883026123047, "learning_rate": 9.494928192776342e-05, "loss": 0.0551, "step": 18710 }, { "grad_norm": 0.538001298904419, "learning_rate": 9.494203763430538e-05, "loss": 0.0436, "step": 18720 }, { "grad_norm": 0.5329576730728149, "learning_rate": 9.493478842605366e-05, "loss": 0.09, "step": 18730 }, { "grad_norm": 0.4876662790775299, "learning_rate": 9.492753430380105e-05, "loss": 0.0585, "step": 18740 }, { "grad_norm": 0.44014984369277954, "learning_rate": 9.492027526834083e-05, "loss": 0.0516, "step": 18750 }, { "grad_norm": 0.6928979158401489, "learning_rate": 9.491301132046684e-05, "loss": 0.0551, "step": 18760 }, { "grad_norm": 0.364149808883667, "learning_rate": 9.490574246097345e-05, "loss": 0.0752, "step": 18770 }, { "grad_norm": 0.6232640743255615, "learning_rate": 9.48984686906556e-05, "loss": 0.0909, "step": 18780 }, { "grad_norm": 0.5196787714958191, "learning_rate": 9.489119001030871e-05, "loss": 0.0571, "step": 18790 }, { "grad_norm": 0.39655765891075134, "learning_rate": 9.488390642072878e-05, "loss": 0.0496, "step": 18800 }, { "grad_norm": 0.3336314857006073, "learning_rate": 9.48766179227123e-05, "loss": 0.0494, "step": 18810 }, { "grad_norm": 0.4368487596511841, "learning_rate": 9.486932451705636e-05, "loss": 0.0564, "step": 18820 }, { "grad_norm": 0.3379356265068054, "learning_rate": 9.486202620455857e-05, "loss": 0.0501, "step": 18830 }, { "grad_norm": 0.47537532448768616, "learning_rate": 9.485472298601704e-05, "loss": 0.0501, "step": 18840 }, { "grad_norm": 0.45051079988479614, "learning_rate": 9.484741486223043e-05, "loss": 0.0497, "step": 18850 }, { "grad_norm": 0.4151460826396942, "learning_rate": 9.484010183399797e-05, "loss": 0.0541, "step": 18860 }, { "grad_norm": 0.37396785616874695, "learning_rate": 9.483278390211938e-05, "loss": 0.057, "step": 18870 }, { "grad_norm": 0.5247480869293213, "learning_rate": 9.482546106739496e-05, "loss": 0.0552, "step": 18880 }, { "grad_norm": 0.6067299246788025, "learning_rate": 9.48181333306255e-05, "loss": 0.0501, "step": 18890 }, { "grad_norm": 0.5034573078155518, "learning_rate": 9.481080069261237e-05, "loss": 0.0778, "step": 18900 }, { "grad_norm": 0.5338325500488281, "learning_rate": 9.480346315415745e-05, "loss": 0.0622, "step": 18910 }, { "grad_norm": 0.5226472020149231, "learning_rate": 9.479612071606314e-05, "loss": 0.0514, "step": 18920 }, { "grad_norm": 0.4486217796802521, "learning_rate": 9.478877337913244e-05, "loss": 0.0503, "step": 18930 }, { "grad_norm": 0.48549625277519226, "learning_rate": 9.478142114416881e-05, "loss": 0.0458, "step": 18940 }, { "grad_norm": 0.3131078779697418, "learning_rate": 9.47740640119763e-05, "loss": 0.0468, "step": 18950 }, { "grad_norm": 0.4427940249443054, "learning_rate": 9.476670198335947e-05, "loss": 0.0543, "step": 18960 }, { "grad_norm": 0.4274751543998718, "learning_rate": 9.47593350591234e-05, "loss": 0.0576, "step": 18970 }, { "grad_norm": 0.5660244226455688, "learning_rate": 9.475196324007376e-05, "loss": 0.0623, "step": 18980 }, { "grad_norm": 0.5660751461982727, "learning_rate": 9.474458652701669e-05, "loss": 0.0673, "step": 18990 }, { "grad_norm": 0.4519835412502289, "learning_rate": 9.473720492075892e-05, "loss": 0.0689, "step": 19000 }, { "grad_norm": 0.5064630508422852, "learning_rate": 9.472981842210768e-05, "loss": 0.063, "step": 19010 }, { "grad_norm": 0.40779098868370056, "learning_rate": 9.472242703187074e-05, "loss": 0.0577, "step": 19020 }, { "grad_norm": 0.4070720970630646, "learning_rate": 9.471503075085643e-05, "loss": 0.0534, "step": 19030 }, { "grad_norm": 0.38814377784729004, "learning_rate": 9.470762957987359e-05, "loss": 0.064, "step": 19040 }, { "grad_norm": 0.30165740847587585, "learning_rate": 9.470022351973158e-05, "loss": 0.0536, "step": 19050 }, { "grad_norm": 0.3913382887840271, "learning_rate": 9.469281257124034e-05, "loss": 0.0562, "step": 19060 }, { "grad_norm": 0.2737761437892914, "learning_rate": 9.46853967352103e-05, "loss": 0.0601, "step": 19070 }, { "grad_norm": 0.530672013759613, "learning_rate": 9.467797601245246e-05, "loss": 0.0513, "step": 19080 }, { "grad_norm": 0.41995468735694885, "learning_rate": 9.467055040377834e-05, "loss": 0.0673, "step": 19090 }, { "grad_norm": 0.5921972393989563, "learning_rate": 9.466311990999999e-05, "loss": 0.061, "step": 19100 }, { "grad_norm": 0.34851086139678955, "learning_rate": 9.465568453193e-05, "loss": 0.0466, "step": 19110 }, { "grad_norm": 0.5890504121780396, "learning_rate": 9.464824427038148e-05, "loss": 0.0454, "step": 19120 }, { "grad_norm": 0.5304145216941833, "learning_rate": 9.46407991261681e-05, "loss": 0.066, "step": 19130 }, { "grad_norm": 0.6069457530975342, "learning_rate": 9.463334910010404e-05, "loss": 0.048, "step": 19140 }, { "grad_norm": 0.5535039305686951, "learning_rate": 9.462589419300403e-05, "loss": 0.0539, "step": 19150 }, { "grad_norm": 0.407905638217926, "learning_rate": 9.461843440568333e-05, "loss": 0.0469, "step": 19160 }, { "grad_norm": 0.4334525167942047, "learning_rate": 9.461096973895773e-05, "loss": 0.058, "step": 19170 }, { "grad_norm": 0.41562119126319885, "learning_rate": 9.460350019364355e-05, "loss": 0.048, "step": 19180 }, { "grad_norm": 0.4890945255756378, "learning_rate": 9.459602577055764e-05, "loss": 0.0807, "step": 19190 }, { "grad_norm": 0.4259917438030243, "learning_rate": 9.45885464705174e-05, "loss": 0.0682, "step": 19200 }, { "grad_norm": 0.49683061242103577, "learning_rate": 9.458106229434076e-05, "loss": 0.0644, "step": 19210 }, { "grad_norm": 0.4935692548751831, "learning_rate": 9.457357324284617e-05, "loss": 0.0515, "step": 19220 }, { "grad_norm": 0.5452241897583008, "learning_rate": 9.456607931685262e-05, "loss": 0.0671, "step": 19230 }, { "grad_norm": 0.5344727039337158, "learning_rate": 9.455858051717965e-05, "loss": 0.0577, "step": 19240 }, { "grad_norm": 0.3453068733215332, "learning_rate": 9.45510768446473e-05, "loss": 0.0478, "step": 19250 }, { "grad_norm": 0.5136892199516296, "learning_rate": 9.454356830007618e-05, "loss": 0.0468, "step": 19260 }, { "grad_norm": 0.6285949945449829, "learning_rate": 9.45360548842874e-05, "loss": 0.0501, "step": 19270 }, { "grad_norm": 0.3033589720726013, "learning_rate": 9.452853659810261e-05, "loss": 0.056, "step": 19280 }, { "grad_norm": 0.501711905002594, "learning_rate": 9.452101344234401e-05, "loss": 0.0551, "step": 19290 }, { "grad_norm": 0.41408804059028625, "learning_rate": 9.451348541783431e-05, "loss": 0.0593, "step": 19300 }, { "grad_norm": 0.4046952724456787, "learning_rate": 9.450595252539678e-05, "loss": 0.0527, "step": 19310 }, { "grad_norm": 0.3500345051288605, "learning_rate": 9.449841476585518e-05, "loss": 0.0581, "step": 19320 }, { "grad_norm": 0.6191453337669373, "learning_rate": 9.449087214003384e-05, "loss": 0.0655, "step": 19330 }, { "grad_norm": 0.612560510635376, "learning_rate": 9.448332464875765e-05, "loss": 0.0755, "step": 19340 }, { "grad_norm": 0.609778106212616, "learning_rate": 9.447577229285192e-05, "loss": 0.0628, "step": 19350 }, { "grad_norm": 0.4725516140460968, "learning_rate": 9.446821507314261e-05, "loss": 0.0486, "step": 19360 }, { "grad_norm": 0.45865270495414734, "learning_rate": 9.446065299045617e-05, "loss": 0.0575, "step": 19370 }, { "grad_norm": 0.6391687393188477, "learning_rate": 9.445308604561955e-05, "loss": 0.0617, "step": 19380 }, { "grad_norm": 0.2640931010246277, "learning_rate": 9.444551423946028e-05, "loss": 0.0601, "step": 19390 }, { "grad_norm": 0.3521830141544342, "learning_rate": 9.443793757280638e-05, "loss": 0.0573, "step": 19400 }, { "grad_norm": 0.3402632772922516, "learning_rate": 9.443035604648646e-05, "loss": 0.0499, "step": 19410 }, { "grad_norm": 0.5587541460990906, "learning_rate": 9.44227696613296e-05, "loss": 0.057, "step": 19420 }, { "grad_norm": 0.48346179723739624, "learning_rate": 9.441517841816542e-05, "loss": 0.0711, "step": 19430 }, { "grad_norm": 0.45536261796951294, "learning_rate": 9.440758231782413e-05, "loss": 0.0652, "step": 19440 }, { "grad_norm": 0.6013180017471313, "learning_rate": 9.439998136113639e-05, "loss": 0.0641, "step": 19450 }, { "grad_norm": 1.1037812232971191, "learning_rate": 9.439237554893344e-05, "loss": 0.0699, "step": 19460 }, { "grad_norm": 1.0909395217895508, "learning_rate": 9.438476488204705e-05, "loss": 0.0809, "step": 19470 }, { "grad_norm": 0.8715920448303223, "learning_rate": 9.43771493613095e-05, "loss": 0.0629, "step": 19480 }, { "grad_norm": 0.5793569684028625, "learning_rate": 9.436952898755362e-05, "loss": 0.0769, "step": 19490 }, { "grad_norm": 0.6088061928749084, "learning_rate": 9.436190376161276e-05, "loss": 0.0686, "step": 19500 }, { "grad_norm": 0.9537076354026794, "learning_rate": 9.43542736843208e-05, "loss": 0.0864, "step": 19510 }, { "grad_norm": 0.37083789706230164, "learning_rate": 9.434663875651216e-05, "loss": 0.0567, "step": 19520 }, { "grad_norm": 0.4966728389263153, "learning_rate": 9.433899897902177e-05, "loss": 0.0641, "step": 19530 }, { "grad_norm": 0.5184901356697083, "learning_rate": 9.433135435268511e-05, "loss": 0.0582, "step": 19540 }, { "grad_norm": 0.711897075176239, "learning_rate": 9.432370487833819e-05, "loss": 0.0724, "step": 19550 }, { "grad_norm": 0.9690349102020264, "learning_rate": 9.431605055681756e-05, "loss": 0.0666, "step": 19560 }, { "grad_norm": 0.656324565410614, "learning_rate": 9.430839138896026e-05, "loss": 0.0667, "step": 19570 }, { "grad_norm": 0.41827163100242615, "learning_rate": 9.43007273756039e-05, "loss": 0.067, "step": 19580 }, { "grad_norm": 0.4600698947906494, "learning_rate": 9.429305851758658e-05, "loss": 0.059, "step": 19590 }, { "grad_norm": 0.4918856620788574, "learning_rate": 9.428538481574699e-05, "loss": 0.0615, "step": 19600 }, { "grad_norm": 0.47812554240226746, "learning_rate": 9.42777062709243e-05, "loss": 0.0585, "step": 19610 }, { "grad_norm": 0.42008787393569946, "learning_rate": 9.427002288395821e-05, "loss": 0.0459, "step": 19620 }, { "grad_norm": 0.9598294496536255, "learning_rate": 9.426233465568898e-05, "loss": 0.0636, "step": 19630 }, { "grad_norm": 0.4548772871494293, "learning_rate": 9.42546415869574e-05, "loss": 0.0515, "step": 19640 }, { "grad_norm": 0.4778732657432556, "learning_rate": 9.424694367860473e-05, "loss": 0.059, "step": 19650 }, { "grad_norm": 0.3482990264892578, "learning_rate": 9.423924093147284e-05, "loss": 0.0502, "step": 19660 }, { "grad_norm": 0.8544325828552246, "learning_rate": 9.423153334640407e-05, "loss": 0.051, "step": 19670 }, { "grad_norm": 0.5114798545837402, "learning_rate": 9.42238209242413e-05, "loss": 0.0493, "step": 19680 }, { "grad_norm": 0.5858446359634399, "learning_rate": 9.421610366582798e-05, "loss": 0.0551, "step": 19690 }, { "grad_norm": 0.8319953083992004, "learning_rate": 9.420838157200803e-05, "loss": 0.0505, "step": 19700 }, { "grad_norm": 0.4490336775779724, "learning_rate": 9.420065464362594e-05, "loss": 0.0659, "step": 19710 }, { "grad_norm": 0.5549680590629578, "learning_rate": 9.419292288152673e-05, "loss": 0.0592, "step": 19720 }, { "grad_norm": 0.489534854888916, "learning_rate": 9.418518628655588e-05, "loss": 0.0639, "step": 19730 }, { "grad_norm": 0.3384355902671814, "learning_rate": 9.417744485955951e-05, "loss": 0.0563, "step": 19740 }, { "grad_norm": 0.8686873912811279, "learning_rate": 9.41696986013842e-05, "loss": 0.0581, "step": 19750 }, { "grad_norm": 0.44024530053138733, "learning_rate": 9.416194751287705e-05, "loss": 0.0636, "step": 19760 }, { "grad_norm": 0.6001033186912537, "learning_rate": 9.415419159488572e-05, "loss": 0.0553, "step": 19770 }, { "grad_norm": 0.5125467777252197, "learning_rate": 9.414643084825837e-05, "loss": 0.0647, "step": 19780 }, { "grad_norm": 0.3321883976459503, "learning_rate": 9.413866527384372e-05, "loss": 0.0693, "step": 19790 }, { "grad_norm": 0.3910486102104187, "learning_rate": 9.4130894872491e-05, "loss": 0.057, "step": 19800 }, { "grad_norm": 0.32126525044441223, "learning_rate": 9.412311964504998e-05, "loss": 0.0581, "step": 19810 }, { "grad_norm": 0.5243303179740906, "learning_rate": 9.411533959237091e-05, "loss": 0.0589, "step": 19820 }, { "grad_norm": 0.5507054328918457, "learning_rate": 9.410755471530464e-05, "loss": 0.0415, "step": 19830 }, { "grad_norm": 0.691003143787384, "learning_rate": 9.40997650147025e-05, "loss": 0.0533, "step": 19840 }, { "grad_norm": 0.5960753560066223, "learning_rate": 9.409197049141637e-05, "loss": 0.0589, "step": 19850 }, { "grad_norm": 1.8725082874298096, "learning_rate": 9.408417114629863e-05, "loss": 0.0566, "step": 19860 }, { "grad_norm": 0.3987361788749695, "learning_rate": 9.40763669802022e-05, "loss": 0.0455, "step": 19870 }, { "grad_norm": 0.8440999984741211, "learning_rate": 9.406855799398056e-05, "loss": 0.0674, "step": 19880 }, { "grad_norm": 0.8178975582122803, "learning_rate": 9.406074418848767e-05, "loss": 0.0543, "step": 19890 }, { "grad_norm": 0.5262519121170044, "learning_rate": 9.405292556457805e-05, "loss": 0.0691, "step": 19900 }, { "grad_norm": 0.34986260533332825, "learning_rate": 9.404510212310671e-05, "loss": 0.0558, "step": 19910 }, { "grad_norm": 0.5403884649276733, "learning_rate": 9.403727386492924e-05, "loss": 0.0528, "step": 19920 }, { "grad_norm": 0.5135267972946167, "learning_rate": 9.40294407909017e-05, "loss": 0.0574, "step": 19930 }, { "grad_norm": 0.4500488340854645, "learning_rate": 9.40216029018807e-05, "loss": 0.0578, "step": 19940 }, { "grad_norm": 0.35817471146583557, "learning_rate": 9.401376019872338e-05, "loss": 0.0501, "step": 19950 }, { "grad_norm": 0.39351406693458557, "learning_rate": 9.400591268228746e-05, "loss": 0.0598, "step": 19960 }, { "grad_norm": 0.5702764987945557, "learning_rate": 9.399806035343106e-05, "loss": 0.073, "step": 19970 }, { "grad_norm": 0.5305197834968567, "learning_rate": 9.399020321301294e-05, "loss": 0.0645, "step": 19980 }, { "grad_norm": 0.6009148955345154, "learning_rate": 9.398234126189234e-05, "loss": 0.0482, "step": 19990 }, { "grad_norm": 0.6949361562728882, "learning_rate": 9.397447450092902e-05, "loss": 0.0622, "step": 20000 }, { "grad_norm": 0.4668304920196533, "learning_rate": 9.39666029309833e-05, "loss": 0.0535, "step": 20010 }, { "grad_norm": 0.4138278365135193, "learning_rate": 9.395872655291596e-05, "loss": 0.0652, "step": 20020 }, { "grad_norm": 0.5185472965240479, "learning_rate": 9.395084536758838e-05, "loss": 0.0472, "step": 20030 }, { "grad_norm": 0.37026068568229675, "learning_rate": 9.394295937586243e-05, "loss": 0.0665, "step": 20040 }, { "grad_norm": 0.4480966627597809, "learning_rate": 9.393506857860052e-05, "loss": 0.0532, "step": 20050 }, { "grad_norm": 0.3478356599807739, "learning_rate": 9.392717297666555e-05, "loss": 0.0447, "step": 20060 }, { "grad_norm": 0.6284745931625366, "learning_rate": 9.391927257092101e-05, "loss": 0.0622, "step": 20070 }, { "grad_norm": 0.440682977437973, "learning_rate": 9.391136736223085e-05, "loss": 0.0566, "step": 20080 }, { "grad_norm": 0.47707250714302063, "learning_rate": 9.390345735145956e-05, "loss": 0.0612, "step": 20090 }, { "grad_norm": 0.3994075655937195, "learning_rate": 9.389554253947219e-05, "loss": 0.0502, "step": 20100 }, { "grad_norm": 0.7786524295806885, "learning_rate": 9.388762292713428e-05, "loss": 0.0584, "step": 20110 }, { "grad_norm": 0.3952798545360565, "learning_rate": 9.38796985153119e-05, "loss": 0.0534, "step": 20120 }, { "grad_norm": 0.4270094931125641, "learning_rate": 9.387176930487169e-05, "loss": 0.057, "step": 20130 }, { "grad_norm": 0.5131462812423706, "learning_rate": 9.386383529668072e-05, "loss": 0.0476, "step": 20140 }, { "grad_norm": 0.5007686018943787, "learning_rate": 9.385589649160669e-05, "loss": 0.0617, "step": 20150 }, { "grad_norm": 0.5838122963905334, "learning_rate": 9.384795289051775e-05, "loss": 0.0563, "step": 20160 }, { "grad_norm": 0.46584197878837585, "learning_rate": 9.384000449428261e-05, "loss": 0.0516, "step": 20170 }, { "grad_norm": 0.44297462701797485, "learning_rate": 9.383205130377048e-05, "loss": 0.0532, "step": 20180 }, { "grad_norm": 0.536625862121582, "learning_rate": 9.382409331985114e-05, "loss": 0.0512, "step": 20190 }, { "grad_norm": 0.5328351259231567, "learning_rate": 9.381613054339482e-05, "loss": 0.0449, "step": 20200 }, { "grad_norm": 0.5314372181892395, "learning_rate": 9.380816297527235e-05, "loss": 0.0497, "step": 20210 }, { "grad_norm": 0.6741223335266113, "learning_rate": 9.380019061635506e-05, "loss": 0.0577, "step": 20220 }, { "grad_norm": 0.5564643144607544, "learning_rate": 9.379221346751474e-05, "loss": 0.0677, "step": 20230 }, { "grad_norm": 0.40682366490364075, "learning_rate": 9.378423152962382e-05, "loss": 0.0516, "step": 20240 }, { "grad_norm": 0.3473639190196991, "learning_rate": 9.377624480355517e-05, "loss": 0.0444, "step": 20250 }, { "grad_norm": 0.6534020304679871, "learning_rate": 9.376825329018219e-05, "loss": 0.0453, "step": 20260 }, { "grad_norm": 0.39651018381118774, "learning_rate": 9.376025699037884e-05, "loss": 0.0523, "step": 20270 }, { "grad_norm": 0.44796139001846313, "learning_rate": 9.37522559050196e-05, "loss": 0.0436, "step": 20280 }, { "grad_norm": 0.33090880513191223, "learning_rate": 9.37442500349794e-05, "loss": 0.0544, "step": 20290 }, { "grad_norm": 0.4934203028678894, "learning_rate": 9.373623938113381e-05, "loss": 0.05, "step": 20300 }, { "grad_norm": 0.3654867708683014, "learning_rate": 9.372822394435883e-05, "loss": 0.0585, "step": 20310 }, { "grad_norm": 0.5411023497581482, "learning_rate": 9.372020372553102e-05, "loss": 0.0626, "step": 20320 }, { "grad_norm": 0.4383411109447479, "learning_rate": 9.371217872552746e-05, "loss": 0.06, "step": 20330 }, { "grad_norm": 0.8282859921455383, "learning_rate": 9.370414894522576e-05, "loss": 0.0607, "step": 20340 }, { "grad_norm": 0.448025107383728, "learning_rate": 9.369611438550406e-05, "loss": 0.0647, "step": 20350 }, { "grad_norm": 0.3422242999076843, "learning_rate": 9.368807504724095e-05, "loss": 0.059, "step": 20360 }, { "grad_norm": 0.45186328887939453, "learning_rate": 9.368003093131565e-05, "loss": 0.0627, "step": 20370 }, { "grad_norm": 0.474055677652359, "learning_rate": 9.367198203860785e-05, "loss": 0.07, "step": 20380 }, { "grad_norm": 0.503574550151825, "learning_rate": 9.366392836999774e-05, "loss": 0.0539, "step": 20390 }, { "grad_norm": 0.5807859301567078, "learning_rate": 9.365586992636607e-05, "loss": 0.0566, "step": 20400 }, { "grad_norm": 0.5265973806381226, "learning_rate": 9.364780670859412e-05, "loss": 0.0527, "step": 20410 }, { "grad_norm": 0.46080753207206726, "learning_rate": 9.363973871756364e-05, "loss": 0.0511, "step": 20420 }, { "grad_norm": 0.5999647974967957, "learning_rate": 9.363166595415696e-05, "loss": 0.0563, "step": 20430 }, { "grad_norm": 0.5145984888076782, "learning_rate": 9.362358841925686e-05, "loss": 0.068, "step": 20440 }, { "grad_norm": 0.3408719301223755, "learning_rate": 9.361550611374674e-05, "loss": 0.0538, "step": 20450 }, { "grad_norm": 0.35396668314933777, "learning_rate": 9.360741903851043e-05, "loss": 0.0709, "step": 20460 }, { "grad_norm": 0.38850370049476624, "learning_rate": 9.359932719443236e-05, "loss": 0.0594, "step": 20470 }, { "grad_norm": 0.4604242444038391, "learning_rate": 9.35912305823974e-05, "loss": 0.0495, "step": 20480 }, { "grad_norm": 0.46929270029067993, "learning_rate": 9.358312920329101e-05, "loss": 0.0583, "step": 20490 }, { "grad_norm": 0.42495155334472656, "learning_rate": 9.357502305799914e-05, "loss": 0.0498, "step": 20500 }, { "grad_norm": 0.44292208552360535, "learning_rate": 9.356691214740824e-05, "loss": 0.0629, "step": 20510 }, { "grad_norm": 0.3939334452152252, "learning_rate": 9.355879647240535e-05, "loss": 0.0516, "step": 20520 }, { "grad_norm": 0.364420622587204, "learning_rate": 9.355067603387798e-05, "loss": 0.057, "step": 20530 }, { "grad_norm": 0.37804678082466125, "learning_rate": 9.354255083271412e-05, "loss": 0.0495, "step": 20540 }, { "grad_norm": 0.39515796303749084, "learning_rate": 9.353442086980239e-05, "loss": 0.0507, "step": 20550 }, { "grad_norm": 0.5102770328521729, "learning_rate": 9.352628614603185e-05, "loss": 0.0523, "step": 20560 }, { "grad_norm": 0.5074473023414612, "learning_rate": 9.351814666229209e-05, "loss": 0.0433, "step": 20570 }, { "grad_norm": 0.5698148012161255, "learning_rate": 9.351000241947324e-05, "loss": 0.0532, "step": 20580 }, { "grad_norm": 0.6093165874481201, "learning_rate": 9.350185341846594e-05, "loss": 0.0496, "step": 20590 }, { "grad_norm": 0.6663399338722229, "learning_rate": 9.349369966016134e-05, "loss": 0.0609, "step": 20600 }, { "grad_norm": 0.563127338886261, "learning_rate": 9.348554114545117e-05, "loss": 0.0676, "step": 20610 }, { "grad_norm": 0.43984100222587585, "learning_rate": 9.347737787522758e-05, "loss": 0.0416, "step": 20620 }, { "grad_norm": 0.46982574462890625, "learning_rate": 9.346920985038332e-05, "loss": 0.0444, "step": 20630 }, { "grad_norm": 0.43486589193344116, "learning_rate": 9.346103707181162e-05, "loss": 0.0527, "step": 20640 }, { "grad_norm": 0.41244250535964966, "learning_rate": 9.345285954040626e-05, "loss": 0.054, "step": 20650 }, { "grad_norm": 0.9103245139122009, "learning_rate": 9.34446772570615e-05, "loss": 0.0529, "step": 20660 }, { "grad_norm": 0.9097773432731628, "learning_rate": 9.343649022267214e-05, "loss": 0.0431, "step": 20670 }, { "grad_norm": 0.6305363774299622, "learning_rate": 9.342829843813353e-05, "loss": 0.051, "step": 20680 }, { "grad_norm": 0.5183584094047546, "learning_rate": 9.342010190434149e-05, "loss": 0.0652, "step": 20690 }, { "grad_norm": 0.37540683150291443, "learning_rate": 9.34119006221924e-05, "loss": 0.0587, "step": 20700 }, { "grad_norm": 0.5851314067840576, "learning_rate": 9.340369459258313e-05, "loss": 0.0573, "step": 20710 }, { "grad_norm": 0.47655189037323, "learning_rate": 9.339548381641106e-05, "loss": 0.0522, "step": 20720 }, { "grad_norm": 0.38797155022621155, "learning_rate": 9.338726829457413e-05, "loss": 0.0557, "step": 20730 }, { "grad_norm": 0.6475501656532288, "learning_rate": 9.337904802797078e-05, "loss": 0.0475, "step": 20740 }, { "grad_norm": 0.6939681172370911, "learning_rate": 9.337082301749993e-05, "loss": 0.0579, "step": 20750 }, { "grad_norm": 0.47218018770217896, "learning_rate": 9.336259326406109e-05, "loss": 0.0588, "step": 20760 }, { "grad_norm": 0.4841625690460205, "learning_rate": 9.335435876855427e-05, "loss": 0.0657, "step": 20770 }, { "grad_norm": 0.8595733642578125, "learning_rate": 9.334611953187994e-05, "loss": 0.0668, "step": 20780 }, { "grad_norm": 0.3627639412879944, "learning_rate": 9.333787555493914e-05, "loss": 0.0432, "step": 20790 }, { "grad_norm": 0.42225760221481323, "learning_rate": 9.332962683863345e-05, "loss": 0.0537, "step": 20800 }, { "grad_norm": 0.43792009353637695, "learning_rate": 9.332137338386489e-05, "loss": 0.0515, "step": 20810 }, { "grad_norm": 0.448091596364975, "learning_rate": 9.33131151915361e-05, "loss": 0.056, "step": 20820 }, { "grad_norm": 0.43275633454322815, "learning_rate": 9.330485226255012e-05, "loss": 0.0466, "step": 20830 }, { "grad_norm": 0.4551967680454254, "learning_rate": 9.329658459781061e-05, "loss": 0.0558, "step": 20840 }, { "grad_norm": 0.4332959055900574, "learning_rate": 9.328831219822172e-05, "loss": 0.0552, "step": 20850 }, { "grad_norm": 0.7206501960754395, "learning_rate": 9.328003506468808e-05, "loss": 0.0524, "step": 20860 }, { "grad_norm": 0.4585782289505005, "learning_rate": 9.327175319811488e-05, "loss": 0.0527, "step": 20870 }, { "grad_norm": 0.5690492987632751, "learning_rate": 9.326346659940781e-05, "loss": 0.053, "step": 20880 }, { "grad_norm": 0.4250342547893524, "learning_rate": 9.325517526947308e-05, "loss": 0.0559, "step": 20890 }, { "grad_norm": 0.6107490658760071, "learning_rate": 9.32468792092174e-05, "loss": 0.0592, "step": 20900 }, { "grad_norm": 0.4106023609638214, "learning_rate": 9.323857841954803e-05, "loss": 0.0483, "step": 20910 }, { "grad_norm": 0.5758616328239441, "learning_rate": 9.323027290137276e-05, "loss": 0.0577, "step": 20920 }, { "grad_norm": 0.49303773045539856, "learning_rate": 9.322196265559981e-05, "loss": 0.0667, "step": 20930 }, { "grad_norm": 0.4408280551433563, "learning_rate": 9.321364768313803e-05, "loss": 0.0618, "step": 20940 }, { "grad_norm": 0.5553278923034668, "learning_rate": 9.32053279848967e-05, "loss": 0.0567, "step": 20950 }, { "grad_norm": 0.2937105894088745, "learning_rate": 9.319700356178567e-05, "loss": 0.0544, "step": 20960 }, { "grad_norm": 0.7107142210006714, "learning_rate": 9.318867441471527e-05, "loss": 0.0519, "step": 20970 }, { "grad_norm": 0.31234851479530334, "learning_rate": 9.318034054459637e-05, "loss": 0.0437, "step": 20980 }, { "grad_norm": 0.715519368648529, "learning_rate": 9.317200195234034e-05, "loss": 0.051, "step": 20990 }, { "grad_norm": 0.5615711808204651, "learning_rate": 9.316365863885909e-05, "loss": 0.0499, "step": 21000 }, { "grad_norm": 0.4478769600391388, "learning_rate": 9.315531060506502e-05, "loss": 0.0626, "step": 21010 }, { "grad_norm": 0.7711564898490906, "learning_rate": 9.314695785187108e-05, "loss": 0.0539, "step": 21020 }, { "grad_norm": 0.760779619216919, "learning_rate": 9.313860038019069e-05, "loss": 0.0618, "step": 21030 }, { "grad_norm": 1.0873892307281494, "learning_rate": 9.313023819093782e-05, "loss": 0.0594, "step": 21040 }, { "grad_norm": 0.6023936867713928, "learning_rate": 9.312187128502695e-05, "loss": 0.0695, "step": 21050 }, { "grad_norm": 0.8809281587600708, "learning_rate": 9.311349966337307e-05, "loss": 0.0568, "step": 21060 }, { "grad_norm": 0.42033135890960693, "learning_rate": 9.310512332689169e-05, "loss": 0.0555, "step": 21070 }, { "grad_norm": 0.29644128680229187, "learning_rate": 9.309674227649883e-05, "loss": 0.0477, "step": 21080 }, { "grad_norm": 0.4461668133735657, "learning_rate": 9.308835651311103e-05, "loss": 0.0657, "step": 21090 }, { "grad_norm": 0.43857166171073914, "learning_rate": 9.307996603764533e-05, "loss": 0.0574, "step": 21100 }, { "grad_norm": 0.4193372428417206, "learning_rate": 9.307157085101932e-05, "loss": 0.0466, "step": 21110 }, { "grad_norm": 0.5561701655387878, "learning_rate": 9.306317095415109e-05, "loss": 0.0483, "step": 21120 }, { "grad_norm": 0.5232712030410767, "learning_rate": 9.305476634795922e-05, "loss": 0.0495, "step": 21130 }, { "grad_norm": 0.3259900212287903, "learning_rate": 9.304635703336284e-05, "loss": 0.0492, "step": 21140 }, { "grad_norm": 0.3751900792121887, "learning_rate": 9.303794301128157e-05, "loss": 0.0432, "step": 21150 }, { "grad_norm": 0.3874107599258423, "learning_rate": 9.302952428263555e-05, "loss": 0.0451, "step": 21160 }, { "grad_norm": 0.6830415725708008, "learning_rate": 9.302110084834545e-05, "loss": 0.0421, "step": 21170 }, { "grad_norm": 0.39455655217170715, "learning_rate": 9.301267270933245e-05, "loss": 0.0443, "step": 21180 }, { "grad_norm": 0.5241729617118835, "learning_rate": 9.300423986651823e-05, "loss": 0.0503, "step": 21190 }, { "grad_norm": 0.4410437345504761, "learning_rate": 9.299580232082501e-05, "loss": 0.0521, "step": 21200 }, { "grad_norm": 0.5756122469902039, "learning_rate": 9.298736007317547e-05, "loss": 0.0535, "step": 21210 }, { "grad_norm": 0.4833616316318512, "learning_rate": 9.297891312449288e-05, "loss": 0.0605, "step": 21220 }, { "grad_norm": 0.3961345851421356, "learning_rate": 9.297046147570094e-05, "loss": 0.0587, "step": 21230 }, { "grad_norm": 0.5893582701683044, "learning_rate": 9.296200512772396e-05, "loss": 0.0552, "step": 21240 }, { "grad_norm": 0.3407323956489563, "learning_rate": 9.295354408148668e-05, "loss": 0.0466, "step": 21250 }, { "grad_norm": 0.3853147029876709, "learning_rate": 9.294507833791441e-05, "loss": 0.0683, "step": 21260 }, { "grad_norm": 0.5114762187004089, "learning_rate": 9.293660789793295e-05, "loss": 0.056, "step": 21270 }, { "grad_norm": 0.4605792760848999, "learning_rate": 9.292813276246858e-05, "loss": 0.0522, "step": 21280 }, { "grad_norm": 0.3948979079723358, "learning_rate": 9.291965293244816e-05, "loss": 0.0514, "step": 21290 }, { "grad_norm": 0.32866770029067993, "learning_rate": 9.291116840879904e-05, "loss": 0.0389, "step": 21300 }, { "grad_norm": 0.5126930475234985, "learning_rate": 9.290267919244904e-05, "loss": 0.0491, "step": 21310 }, { "grad_norm": 0.6545630097389221, "learning_rate": 9.289418528432655e-05, "loss": 0.0508, "step": 21320 }, { "grad_norm": 0.5942518711090088, "learning_rate": 9.288568668536045e-05, "loss": 0.0626, "step": 21330 }, { "grad_norm": 0.6234440207481384, "learning_rate": 9.287718339648013e-05, "loss": 0.0643, "step": 21340 }, { "grad_norm": 0.567956268787384, "learning_rate": 9.28686754186155e-05, "loss": 0.0656, "step": 21350 }, { "grad_norm": 0.5605396032333374, "learning_rate": 9.286016275269698e-05, "loss": 0.065, "step": 21360 }, { "grad_norm": 1.0280940532684326, "learning_rate": 9.285164539965551e-05, "loss": 0.063, "step": 21370 }, { "grad_norm": 0.47325599193573, "learning_rate": 9.284312336042251e-05, "loss": 0.0523, "step": 21380 }, { "grad_norm": 0.48567381501197815, "learning_rate": 9.283459663592996e-05, "loss": 0.0579, "step": 21390 }, { "grad_norm": 0.4084853529930115, "learning_rate": 9.282606522711033e-05, "loss": 0.0573, "step": 21400 }, { "grad_norm": 0.44133472442626953, "learning_rate": 9.281752913489657e-05, "loss": 0.0658, "step": 21410 }, { "grad_norm": 0.5874272584915161, "learning_rate": 9.280898836022222e-05, "loss": 0.0678, "step": 21420 }, { "grad_norm": 0.40837615728378296, "learning_rate": 9.280044290402126e-05, "loss": 0.0515, "step": 21430 }, { "grad_norm": 0.39665651321411133, "learning_rate": 9.279189276722821e-05, "loss": 0.0648, "step": 21440 }, { "grad_norm": 0.527549684047699, "learning_rate": 9.278333795077812e-05, "loss": 0.0684, "step": 21450 }, { "grad_norm": 0.4179525077342987, "learning_rate": 9.27747784556065e-05, "loss": 0.0577, "step": 21460 }, { "grad_norm": 0.38878318667411804, "learning_rate": 9.276621428264942e-05, "loss": 0.0647, "step": 21470 }, { "grad_norm": 0.6192681193351746, "learning_rate": 9.275764543284345e-05, "loss": 0.0511, "step": 21480 }, { "grad_norm": 0.4637540280818939, "learning_rate": 9.274907190712566e-05, "loss": 0.0552, "step": 21490 }, { "grad_norm": 0.4206330180168152, "learning_rate": 9.274049370643363e-05, "loss": 0.0557, "step": 21500 }, { "grad_norm": 0.47477224469184875, "learning_rate": 9.273191083170547e-05, "loss": 0.0643, "step": 21510 }, { "grad_norm": 0.32544568181037903, "learning_rate": 9.27233232838798e-05, "loss": 0.0569, "step": 21520 }, { "grad_norm": 0.6647737622261047, "learning_rate": 9.27147310638957e-05, "loss": 0.0559, "step": 21530 }, { "grad_norm": 0.49791544675827026, "learning_rate": 9.270613417269286e-05, "loss": 0.0508, "step": 21540 }, { "grad_norm": 0.5970211625099182, "learning_rate": 9.269753261121138e-05, "loss": 0.0508, "step": 21550 }, { "grad_norm": 0.507758617401123, "learning_rate": 9.268892638039194e-05, "loss": 0.0467, "step": 21560 }, { "grad_norm": 0.4168432950973511, "learning_rate": 9.268031548117569e-05, "loss": 0.0571, "step": 21570 }, { "grad_norm": 0.5217226147651672, "learning_rate": 9.26716999145043e-05, "loss": 0.0528, "step": 21580 }, { "grad_norm": 0.3984183073043823, "learning_rate": 9.266307968131998e-05, "loss": 0.0541, "step": 21590 }, { "grad_norm": 0.7300992012023926, "learning_rate": 9.26544547825654e-05, "loss": 0.0577, "step": 21600 }, { "grad_norm": 0.5742703080177307, "learning_rate": 9.264582521918376e-05, "loss": 0.0602, "step": 21610 }, { "grad_norm": 0.4188172221183777, "learning_rate": 9.263719099211881e-05, "loss": 0.0563, "step": 21620 }, { "grad_norm": 0.5176172256469727, "learning_rate": 9.262855210231476e-05, "loss": 0.0575, "step": 21630 }, { "grad_norm": 0.5253049731254578, "learning_rate": 9.261990855071633e-05, "loss": 0.048, "step": 21640 }, { "grad_norm": 0.3772651255130768, "learning_rate": 9.261126033826878e-05, "loss": 0.0512, "step": 21650 }, { "grad_norm": 0.3876833915710449, "learning_rate": 9.260260746591786e-05, "loss": 0.0567, "step": 21660 }, { "grad_norm": 0.35696887969970703, "learning_rate": 9.259394993460985e-05, "loss": 0.0458, "step": 21670 }, { "grad_norm": 0.29906296730041504, "learning_rate": 9.258528774529151e-05, "loss": 0.0502, "step": 21680 }, { "grad_norm": 0.4069351553916931, "learning_rate": 9.257662089891013e-05, "loss": 0.0474, "step": 21690 }, { "grad_norm": 0.4533713757991791, "learning_rate": 9.25679493964135e-05, "loss": 0.0619, "step": 21700 }, { "grad_norm": 0.4964284896850586, "learning_rate": 9.255927323874994e-05, "loss": 0.0512, "step": 21710 }, { "grad_norm": 0.426893413066864, "learning_rate": 9.255059242686822e-05, "loss": 0.0551, "step": 21720 }, { "grad_norm": 0.5111502408981323, "learning_rate": 9.254190696171769e-05, "loss": 0.0459, "step": 21730 }, { "grad_norm": 0.36651915311813354, "learning_rate": 9.25332168442482e-05, "loss": 0.0514, "step": 21740 }, { "grad_norm": 0.37870895862579346, "learning_rate": 9.252452207541004e-05, "loss": 0.0507, "step": 21750 }, { "grad_norm": 0.44594046473503113, "learning_rate": 9.251582265615409e-05, "loss": 0.0662, "step": 21760 }, { "grad_norm": 0.5865987539291382, "learning_rate": 9.250711858743169e-05, "loss": 0.0583, "step": 21770 }, { "grad_norm": 0.3721519708633423, "learning_rate": 9.24984098701947e-05, "loss": 0.073, "step": 21780 }, { "grad_norm": 0.6573216915130615, "learning_rate": 9.248969650539552e-05, "loss": 0.0599, "step": 21790 }, { "grad_norm": 0.5752791166305542, "learning_rate": 9.2480978493987e-05, "loss": 0.0556, "step": 21800 }, { "grad_norm": 0.5504024624824524, "learning_rate": 9.247225583692256e-05, "loss": 0.0448, "step": 21810 }, { "grad_norm": 0.29897540807724, "learning_rate": 9.246352853515607e-05, "loss": 0.0466, "step": 21820 }, { "grad_norm": 0.3949015438556671, "learning_rate": 9.245479658964194e-05, "loss": 0.0507, "step": 21830 }, { "grad_norm": 0.4909633994102478, "learning_rate": 9.244606000133507e-05, "loss": 0.0409, "step": 21840 }, { "grad_norm": 0.420773446559906, "learning_rate": 9.24373187711909e-05, "loss": 0.0461, "step": 21850 }, { "grad_norm": 0.41892603039741516, "learning_rate": 9.242857290016537e-05, "loss": 0.0508, "step": 21860 }, { "grad_norm": 0.493608295917511, "learning_rate": 9.241982238921488e-05, "loss": 0.0461, "step": 21870 }, { "grad_norm": 0.4923473000526428, "learning_rate": 9.24110672392964e-05, "loss": 0.0436, "step": 21880 }, { "grad_norm": 0.34000363945961, "learning_rate": 9.240230745136737e-05, "loss": 0.0532, "step": 21890 }, { "grad_norm": 0.3228115737438202, "learning_rate": 9.239354302638575e-05, "loss": 0.0465, "step": 21900 }, { "grad_norm": 0.5015596151351929, "learning_rate": 9.238477396531e-05, "loss": 0.0571, "step": 21910 }, { "grad_norm": 0.4045490026473999, "learning_rate": 9.23760002690991e-05, "loss": 0.0413, "step": 21920 }, { "grad_norm": 0.41350167989730835, "learning_rate": 9.236722193871252e-05, "loss": 0.0446, "step": 21930 }, { "grad_norm": 0.5962153077125549, "learning_rate": 9.235843897511023e-05, "loss": 0.0441, "step": 21940 }, { "grad_norm": 0.4577547013759613, "learning_rate": 9.234965137925276e-05, "loss": 0.0446, "step": 21950 }, { "grad_norm": 0.3796500265598297, "learning_rate": 9.234085915210108e-05, "loss": 0.0382, "step": 21960 }, { "grad_norm": 0.46755585074424744, "learning_rate": 9.23320622946167e-05, "loss": 0.0467, "step": 21970 }, { "grad_norm": 0.3343504071235657, "learning_rate": 9.232326080776163e-05, "loss": 0.0396, "step": 21980 }, { "grad_norm": 0.5451169013977051, "learning_rate": 9.23144546924984e-05, "loss": 0.0456, "step": 21990 }, { "grad_norm": 0.5173920392990112, "learning_rate": 9.230564394979e-05, "loss": 0.0625, "step": 22000 }, { "grad_norm": 0.45369836688041687, "learning_rate": 9.22968285806e-05, "loss": 0.0525, "step": 22010 }, { "grad_norm": 0.5459136366844177, "learning_rate": 9.228800858589242e-05, "loss": 0.0591, "step": 22020 }, { "grad_norm": 0.368893563747406, "learning_rate": 9.227918396663179e-05, "loss": 0.0506, "step": 22030 }, { "grad_norm": 0.46758967638015747, "learning_rate": 9.227035472378319e-05, "loss": 0.056, "step": 22040 }, { "grad_norm": 0.4807138442993164, "learning_rate": 9.226152085831213e-05, "loss": 0.0614, "step": 22050 }, { "grad_norm": 0.41450706124305725, "learning_rate": 9.22526823711847e-05, "loss": 0.0491, "step": 22060 }, { "grad_norm": 0.5738231539726257, "learning_rate": 9.224383926336745e-05, "loss": 0.0518, "step": 22070 }, { "grad_norm": 0.4167916178703308, "learning_rate": 9.223499153582744e-05, "loss": 0.0508, "step": 22080 }, { "grad_norm": 0.5132371187210083, "learning_rate": 9.222613918953226e-05, "loss": 0.0626, "step": 22090 }, { "grad_norm": 0.35694804787635803, "learning_rate": 9.221728222544999e-05, "loss": 0.044, "step": 22100 }, { "grad_norm": 0.6087024807929993, "learning_rate": 9.22084206445492e-05, "loss": 0.0554, "step": 22110 }, { "grad_norm": 0.5059858560562134, "learning_rate": 9.2199554447799e-05, "loss": 0.0685, "step": 22120 }, { "grad_norm": 0.38485077023506165, "learning_rate": 9.219068363616897e-05, "loss": 0.0535, "step": 22130 }, { "grad_norm": 0.3832987844944, "learning_rate": 9.218180821062919e-05, "loss": 0.05, "step": 22140 }, { "grad_norm": 0.35644304752349854, "learning_rate": 9.21729281721503e-05, "loss": 0.0524, "step": 22150 }, { "grad_norm": 0.6452879309654236, "learning_rate": 9.216404352170339e-05, "loss": 0.0518, "step": 22160 }, { "grad_norm": 0.6728871464729309, "learning_rate": 9.215515426026007e-05, "loss": 0.0502, "step": 22170 }, { "grad_norm": 0.40473881363868713, "learning_rate": 9.214626038879246e-05, "loss": 0.0474, "step": 22180 }, { "grad_norm": 0.6522594094276428, "learning_rate": 9.21373619082732e-05, "loss": 0.0518, "step": 22190 }, { "grad_norm": 0.47041261196136475, "learning_rate": 9.212845881967535e-05, "loss": 0.0569, "step": 22200 }, { "grad_norm": 0.4636712968349457, "learning_rate": 9.211955112397262e-05, "loss": 0.0574, "step": 22210 }, { "grad_norm": 0.9219340682029724, "learning_rate": 9.211063882213909e-05, "loss": 0.0432, "step": 22220 }, { "grad_norm": 0.33990147709846497, "learning_rate": 9.210172191514942e-05, "loss": 0.0434, "step": 22230 }, { "grad_norm": 0.48040351271629333, "learning_rate": 9.209280040397874e-05, "loss": 0.0458, "step": 22240 }, { "grad_norm": 0.42557865381240845, "learning_rate": 9.208387428960268e-05, "loss": 0.0422, "step": 22250 }, { "grad_norm": 0.3636239469051361, "learning_rate": 9.20749435729974e-05, "loss": 0.0456, "step": 22260 }, { "grad_norm": 0.31640443205833435, "learning_rate": 9.206600825513957e-05, "loss": 0.0465, "step": 22270 }, { "grad_norm": 0.49750661849975586, "learning_rate": 9.20570683370063e-05, "loss": 0.0481, "step": 22280 }, { "grad_norm": 0.8740575909614563, "learning_rate": 9.204812381957528e-05, "loss": 0.0493, "step": 22290 }, { "grad_norm": 0.3457021415233612, "learning_rate": 9.203917470382465e-05, "loss": 0.0387, "step": 22300 }, { "grad_norm": 0.4467380940914154, "learning_rate": 9.203022099073309e-05, "loss": 0.0587, "step": 22310 }, { "grad_norm": 0.386557936668396, "learning_rate": 9.202126268127976e-05, "loss": 0.0545, "step": 22320 }, { "grad_norm": 0.6734598278999329, "learning_rate": 9.20122997764443e-05, "loss": 0.0499, "step": 22330 }, { "grad_norm": 0.5722162127494812, "learning_rate": 9.200333227720692e-05, "loss": 0.0589, "step": 22340 }, { "grad_norm": 0.596542477607727, "learning_rate": 9.199436018454826e-05, "loss": 0.0583, "step": 22350 }, { "grad_norm": 0.48367857933044434, "learning_rate": 9.198538349944952e-05, "loss": 0.0525, "step": 22360 }, { "grad_norm": 0.4256386458873749, "learning_rate": 9.197640222289234e-05, "loss": 0.068, "step": 22370 }, { "grad_norm": 0.5061110854148865, "learning_rate": 9.196741635585895e-05, "loss": 0.0494, "step": 22380 }, { "grad_norm": 0.8984097242355347, "learning_rate": 9.195842589933199e-05, "loss": 0.0565, "step": 22390 }, { "grad_norm": 0.46808141469955444, "learning_rate": 9.194943085429466e-05, "loss": 0.0572, "step": 22400 }, { "grad_norm": 0.4861321449279785, "learning_rate": 9.194043122173065e-05, "loss": 0.0743, "step": 22410 }, { "grad_norm": 0.5429383516311646, "learning_rate": 9.193142700262413e-05, "loss": 0.0609, "step": 22420 }, { "grad_norm": 0.4391779899597168, "learning_rate": 9.192241819795979e-05, "loss": 0.0476, "step": 22430 }, { "grad_norm": 0.42640429735183716, "learning_rate": 9.191340480872284e-05, "loss": 0.0444, "step": 22440 }, { "grad_norm": 0.36533817648887634, "learning_rate": 9.190438683589895e-05, "loss": 0.0494, "step": 22450 }, { "grad_norm": 0.5410795211791992, "learning_rate": 9.189536428047432e-05, "loss": 0.0555, "step": 22460 }, { "grad_norm": 0.4041254222393036, "learning_rate": 9.188633714343564e-05, "loss": 0.0489, "step": 22470 }, { "grad_norm": 0.5356159210205078, "learning_rate": 9.18773054257701e-05, "loss": 0.0577, "step": 22480 }, { "grad_norm": 0.5200331807136536, "learning_rate": 9.18682691284654e-05, "loss": 0.0447, "step": 22490 }, { "grad_norm": 0.3226737380027771, "learning_rate": 9.185922825250974e-05, "loss": 0.0468, "step": 22500 }, { "grad_norm": 0.4512491524219513, "learning_rate": 9.185018279889181e-05, "loss": 0.0498, "step": 22510 }, { "grad_norm": 0.6669139266014099, "learning_rate": 9.184113276860082e-05, "loss": 0.0598, "step": 22520 }, { "grad_norm": 0.4563369154930115, "learning_rate": 9.183207816262645e-05, "loss": 0.0549, "step": 22530 }, { "grad_norm": 0.29837965965270996, "learning_rate": 9.182301898195891e-05, "loss": 0.0394, "step": 22540 }, { "grad_norm": 0.504048228263855, "learning_rate": 9.181395522758889e-05, "loss": 0.0495, "step": 22550 }, { "grad_norm": 0.5930325984954834, "learning_rate": 9.180488690050759e-05, "loss": 0.0648, "step": 22560 }, { "grad_norm": 0.35600197315216064, "learning_rate": 9.179581400170671e-05, "loss": 0.061, "step": 22570 }, { "grad_norm": 0.24287821352481842, "learning_rate": 9.178673653217845e-05, "loss": 0.0466, "step": 22580 }, { "grad_norm": 0.5136223435401917, "learning_rate": 9.177765449291551e-05, "loss": 0.052, "step": 22590 }, { "grad_norm": 0.4005897641181946, "learning_rate": 9.176856788491109e-05, "loss": 0.063, "step": 22600 }, { "grad_norm": 0.5517712831497192, "learning_rate": 9.175947670915887e-05, "loss": 0.0547, "step": 22610 }, { "grad_norm": 0.41282781958580017, "learning_rate": 9.175038096665309e-05, "loss": 0.0519, "step": 22620 }, { "grad_norm": 0.7772107124328613, "learning_rate": 9.17412806583884e-05, "loss": 0.0588, "step": 22630 }, { "grad_norm": 0.5078704953193665, "learning_rate": 9.173217578536002e-05, "loss": 0.0803, "step": 22640 }, { "grad_norm": 0.49067428708076477, "learning_rate": 9.172306634856362e-05, "loss": 0.0556, "step": 22650 }, { "grad_norm": 0.6021350622177124, "learning_rate": 9.171395234899545e-05, "loss": 0.0535, "step": 22660 }, { "grad_norm": 0.4938865602016449, "learning_rate": 9.170483378765214e-05, "loss": 0.0607, "step": 22670 }, { "grad_norm": 0.5849158763885498, "learning_rate": 9.169571066553091e-05, "loss": 0.056, "step": 22680 }, { "grad_norm": 0.5346100330352783, "learning_rate": 9.168658298362946e-05, "loss": 0.0469, "step": 22690 }, { "grad_norm": 0.5377492308616638, "learning_rate": 9.167745074294598e-05, "loss": 0.0505, "step": 22700 }, { "grad_norm": 0.4150577783584595, "learning_rate": 9.166831394447913e-05, "loss": 0.0629, "step": 22710 }, { "grad_norm": 0.47193148732185364, "learning_rate": 9.165917258922812e-05, "loss": 0.0457, "step": 22720 }, { "grad_norm": 0.861310601234436, "learning_rate": 9.165002667819262e-05, "loss": 0.0509, "step": 22730 }, { "grad_norm": 0.3883854150772095, "learning_rate": 9.164087621237282e-05, "loss": 0.0507, "step": 22740 }, { "grad_norm": 0.5148563981056213, "learning_rate": 9.163172119276942e-05, "loss": 0.0651, "step": 22750 }, { "grad_norm": 0.5247547030448914, "learning_rate": 9.162256162038358e-05, "loss": 0.052, "step": 22760 }, { "grad_norm": 0.48456335067749023, "learning_rate": 9.161339749621698e-05, "loss": 0.0551, "step": 22770 }, { "grad_norm": 0.4975188374519348, "learning_rate": 9.160422882127177e-05, "loss": 0.0447, "step": 22780 }, { "grad_norm": 0.4438733160495758, "learning_rate": 9.159505559655069e-05, "loss": 0.0437, "step": 22790 }, { "grad_norm": 0.34891027212142944, "learning_rate": 9.158587782305684e-05, "loss": 0.0437, "step": 22800 }, { "grad_norm": 0.5379161238670349, "learning_rate": 9.157669550179391e-05, "loss": 0.0465, "step": 22810 }, { "grad_norm": 0.47989127039909363, "learning_rate": 9.156750863376609e-05, "loss": 0.0505, "step": 22820 }, { "grad_norm": 0.6063101291656494, "learning_rate": 9.155831721997801e-05, "loss": 0.0467, "step": 22830 }, { "grad_norm": 0.3929460048675537, "learning_rate": 9.154912126143484e-05, "loss": 0.0525, "step": 22840 }, { "grad_norm": 0.49018123745918274, "learning_rate": 9.153992075914224e-05, "loss": 0.0757, "step": 22850 }, { "grad_norm": 0.35640963912010193, "learning_rate": 9.153071571410635e-05, "loss": 0.0401, "step": 22860 }, { "grad_norm": 0.6456179618835449, "learning_rate": 9.152150612733384e-05, "loss": 0.0543, "step": 22870 }, { "grad_norm": 0.43266206979751587, "learning_rate": 9.151229199983184e-05, "loss": 0.0568, "step": 22880 }, { "grad_norm": 0.471792995929718, "learning_rate": 9.150307333260802e-05, "loss": 0.0459, "step": 22890 }, { "grad_norm": 0.46565064787864685, "learning_rate": 9.149385012667048e-05, "loss": 0.0571, "step": 22900 }, { "grad_norm": 0.41543400287628174, "learning_rate": 9.148462238302788e-05, "loss": 0.0496, "step": 22910 }, { "grad_norm": 1.0679985284805298, "learning_rate": 9.147539010268936e-05, "loss": 0.0571, "step": 22920 }, { "grad_norm": 0.42208561301231384, "learning_rate": 9.14661532866645e-05, "loss": 0.0623, "step": 22930 }, { "grad_norm": 0.3989224433898926, "learning_rate": 9.145691193596348e-05, "loss": 0.0446, "step": 22940 }, { "grad_norm": 0.5809559226036072, "learning_rate": 9.144766605159691e-05, "loss": 0.0599, "step": 22950 }, { "grad_norm": 0.3817436695098877, "learning_rate": 9.14384156345759e-05, "loss": 0.0566, "step": 22960 }, { "grad_norm": 0.6412025690078735, "learning_rate": 9.142916068591204e-05, "loss": 0.042, "step": 22970 }, { "grad_norm": 0.8046175837516785, "learning_rate": 9.141990120661746e-05, "loss": 0.0593, "step": 22980 }, { "grad_norm": 0.4663873016834259, "learning_rate": 9.141063719770475e-05, "loss": 0.046, "step": 22990 }, { "grad_norm": 0.49721020460128784, "learning_rate": 9.140136866018704e-05, "loss": 0.0466, "step": 23000 }, { "grad_norm": 0.4090326726436615, "learning_rate": 9.139209559507788e-05, "loss": 0.0502, "step": 23010 }, { "grad_norm": 0.5012127161026001, "learning_rate": 9.13828180033914e-05, "loss": 0.047, "step": 23020 }, { "grad_norm": 0.39103829860687256, "learning_rate": 9.137353588614212e-05, "loss": 0.0498, "step": 23030 }, { "grad_norm": 0.37594565749168396, "learning_rate": 9.136424924434519e-05, "loss": 0.0505, "step": 23040 }, { "grad_norm": 0.45951300859451294, "learning_rate": 9.135495807901615e-05, "loss": 0.0618, "step": 23050 }, { "grad_norm": 0.4521200656890869, "learning_rate": 9.134566239117108e-05, "loss": 0.0467, "step": 23060 }, { "grad_norm": 0.9998456239700317, "learning_rate": 9.13363621818265e-05, "loss": 0.0597, "step": 23070 }, { "grad_norm": 0.47440844774246216, "learning_rate": 9.132705745199953e-05, "loss": 0.0541, "step": 23080 }, { "grad_norm": 0.37020164728164673, "learning_rate": 9.131774820270768e-05, "loss": 0.0496, "step": 23090 }, { "grad_norm": 0.49470576643943787, "learning_rate": 9.130843443496901e-05, "loss": 0.0646, "step": 23100 }, { "grad_norm": 0.3752119243144989, "learning_rate": 9.129911614980206e-05, "loss": 0.055, "step": 23110 }, { "grad_norm": 0.6015266180038452, "learning_rate": 9.128979334822584e-05, "loss": 0.0461, "step": 23120 }, { "grad_norm": 0.4802594780921936, "learning_rate": 9.128046603125992e-05, "loss": 0.042, "step": 23130 }, { "grad_norm": 0.3812987506389618, "learning_rate": 9.12711341999243e-05, "loss": 0.0455, "step": 23140 }, { "grad_norm": 0.4802057445049286, "learning_rate": 9.12617978552395e-05, "loss": 0.0725, "step": 23150 }, { "grad_norm": 0.6329435706138611, "learning_rate": 9.12524569982265e-05, "loss": 0.0597, "step": 23160 }, { "grad_norm": 0.6548169851303101, "learning_rate": 9.124311162990684e-05, "loss": 0.0529, "step": 23170 }, { "grad_norm": 0.4409189820289612, "learning_rate": 9.12337617513025e-05, "loss": 0.0517, "step": 23180 }, { "grad_norm": 0.5744642019271851, "learning_rate": 9.122440736343596e-05, "loss": 0.0487, "step": 23190 }, { "grad_norm": 0.8791731595993042, "learning_rate": 9.12150484673302e-05, "loss": 0.0572, "step": 23200 }, { "grad_norm": 0.344382643699646, "learning_rate": 9.120568506400873e-05, "loss": 0.0496, "step": 23210 }, { "grad_norm": 0.43915891647338867, "learning_rate": 9.119631715449548e-05, "loss": 0.0641, "step": 23220 }, { "grad_norm": 0.4832238554954529, "learning_rate": 9.118694473981493e-05, "loss": 0.0541, "step": 23230 }, { "grad_norm": 0.3233942687511444, "learning_rate": 9.117756782099203e-05, "loss": 0.0482, "step": 23240 }, { "grad_norm": 0.45964574813842773, "learning_rate": 9.11681863990522e-05, "loss": 0.0507, "step": 23250 }, { "grad_norm": 0.4563405513763428, "learning_rate": 9.115880047502142e-05, "loss": 0.0499, "step": 23260 }, { "grad_norm": 0.5325382351875305, "learning_rate": 9.114941004992609e-05, "loss": 0.0563, "step": 23270 }, { "grad_norm": 0.47371822595596313, "learning_rate": 9.114001512479317e-05, "loss": 0.051, "step": 23280 }, { "grad_norm": 0.8172397017478943, "learning_rate": 9.113061570065003e-05, "loss": 0.0759, "step": 23290 }, { "grad_norm": 0.5255565047264099, "learning_rate": 9.112121177852459e-05, "loss": 0.0624, "step": 23300 }, { "grad_norm": 0.5646945238113403, "learning_rate": 9.111180335944527e-05, "loss": 0.0545, "step": 23310 }, { "grad_norm": 0.6306105852127075, "learning_rate": 9.110239044444093e-05, "loss": 0.0492, "step": 23320 }, { "grad_norm": 0.386332631111145, "learning_rate": 9.109297303454099e-05, "loss": 0.0501, "step": 23330 }, { "grad_norm": 0.2889355421066284, "learning_rate": 9.108355113077526e-05, "loss": 0.0481, "step": 23340 }, { "grad_norm": 0.4155034124851227, "learning_rate": 9.107412473417419e-05, "loss": 0.042, "step": 23350 }, { "grad_norm": 0.4326240122318268, "learning_rate": 9.106469384576858e-05, "loss": 0.0473, "step": 23360 }, { "grad_norm": 0.4134761691093445, "learning_rate": 9.105525846658978e-05, "loss": 0.0478, "step": 23370 }, { "grad_norm": 0.5082457065582275, "learning_rate": 9.104581859766965e-05, "loss": 0.053, "step": 23380 }, { "grad_norm": 0.4800567328929901, "learning_rate": 9.10363742400405e-05, "loss": 0.0548, "step": 23390 }, { "grad_norm": 0.48515763878822327, "learning_rate": 9.102692539473518e-05, "loss": 0.0533, "step": 23400 }, { "grad_norm": 0.5132384300231934, "learning_rate": 9.101747206278697e-05, "loss": 0.0609, "step": 23410 }, { "grad_norm": 0.43065735697746277, "learning_rate": 9.100801424522968e-05, "loss": 0.04, "step": 23420 }, { "grad_norm": 0.418968141078949, "learning_rate": 9.099855194309762e-05, "loss": 0.0471, "step": 23430 }, { "grad_norm": 0.4134041368961334, "learning_rate": 9.098908515742554e-05, "loss": 0.0462, "step": 23440 }, { "grad_norm": 0.4339729845523834, "learning_rate": 9.097961388924873e-05, "loss": 0.0408, "step": 23450 }, { "grad_norm": 0.4299241900444031, "learning_rate": 9.097013813960298e-05, "loss": 0.0463, "step": 23460 }, { "grad_norm": 0.5127289891242981, "learning_rate": 9.09606579095245e-05, "loss": 0.0478, "step": 23470 }, { "grad_norm": 0.3813733160495758, "learning_rate": 9.095117320005008e-05, "loss": 0.0481, "step": 23480 }, { "grad_norm": 0.542149543762207, "learning_rate": 9.094168401221691e-05, "loss": 0.0494, "step": 23490 }, { "grad_norm": 0.409120112657547, "learning_rate": 9.093219034706273e-05, "loss": 0.0626, "step": 23500 }, { "grad_norm": 0.37285512685775757, "learning_rate": 9.092269220562577e-05, "loss": 0.0483, "step": 23510 }, { "grad_norm": 0.3561530113220215, "learning_rate": 9.09131895889447e-05, "loss": 0.0721, "step": 23520 }, { "grad_norm": 0.3678750693798065, "learning_rate": 9.090368249805873e-05, "loss": 0.0483, "step": 23530 }, { "grad_norm": 0.4260546565055847, "learning_rate": 9.089417093400754e-05, "loss": 0.0566, "step": 23540 }, { "grad_norm": 0.3304693400859833, "learning_rate": 9.088465489783131e-05, "loss": 0.0557, "step": 23550 }, { "grad_norm": 0.32267242670059204, "learning_rate": 9.087513439057068e-05, "loss": 0.055, "step": 23560 }, { "grad_norm": 0.31531035900115967, "learning_rate": 9.08656094132668e-05, "loss": 0.0555, "step": 23570 }, { "grad_norm": 0.5092661380767822, "learning_rate": 9.085607996696134e-05, "loss": 0.0516, "step": 23580 }, { "grad_norm": 0.3983056843280792, "learning_rate": 9.084654605269639e-05, "loss": 0.055, "step": 23590 }, { "grad_norm": 0.36737802624702454, "learning_rate": 9.083700767151457e-05, "loss": 0.0476, "step": 23600 }, { "grad_norm": 0.9417341947555542, "learning_rate": 9.082746482445898e-05, "loss": 0.0533, "step": 23610 }, { "grad_norm": 0.5585528016090393, "learning_rate": 9.081791751257325e-05, "loss": 0.0519, "step": 23620 }, { "grad_norm": 0.48618564009666443, "learning_rate": 9.080836573690142e-05, "loss": 0.0532, "step": 23630 }, { "grad_norm": 0.37021225690841675, "learning_rate": 9.079880949848805e-05, "loss": 0.0574, "step": 23640 }, { "grad_norm": 0.5012544989585876, "learning_rate": 9.078924879837822e-05, "loss": 0.0464, "step": 23650 }, { "grad_norm": 0.4226851761341095, "learning_rate": 9.077968363761747e-05, "loss": 0.0514, "step": 23660 }, { "grad_norm": 0.4206017851829529, "learning_rate": 9.077011401725182e-05, "loss": 0.0448, "step": 23670 }, { "grad_norm": 0.37445273995399475, "learning_rate": 9.07605399383278e-05, "loss": 0.0502, "step": 23680 }, { "grad_norm": 0.4951813817024231, "learning_rate": 9.075096140189243e-05, "loss": 0.051, "step": 23690 }, { "grad_norm": 0.4796120822429657, "learning_rate": 9.074137840899318e-05, "loss": 0.0598, "step": 23700 }, { "grad_norm": 0.41992971301078796, "learning_rate": 9.073179096067804e-05, "loss": 0.0405, "step": 23710 }, { "grad_norm": 0.5153414607048035, "learning_rate": 9.072219905799549e-05, "loss": 0.0484, "step": 23720 }, { "grad_norm": 0.5534383654594421, "learning_rate": 9.071260270199447e-05, "loss": 0.0543, "step": 23730 }, { "grad_norm": 0.5471685528755188, "learning_rate": 9.070300189372441e-05, "loss": 0.0479, "step": 23740 }, { "grad_norm": 0.5935088992118835, "learning_rate": 9.069339663423528e-05, "loss": 0.0428, "step": 23750 }, { "grad_norm": 0.2906002998352051, "learning_rate": 9.068378692457747e-05, "loss": 0.0411, "step": 23760 }, { "grad_norm": 0.3940313756465912, "learning_rate": 9.067417276580189e-05, "loss": 0.0455, "step": 23770 }, { "grad_norm": 0.3386039733886719, "learning_rate": 9.066455415895993e-05, "loss": 0.0462, "step": 23780 }, { "grad_norm": 0.3229844570159912, "learning_rate": 9.065493110510346e-05, "loss": 0.0401, "step": 23790 }, { "grad_norm": 0.38988807797431946, "learning_rate": 9.064530360528484e-05, "loss": 0.0433, "step": 23800 }, { "grad_norm": 0.4549877643585205, "learning_rate": 9.063567166055695e-05, "loss": 0.0533, "step": 23810 }, { "grad_norm": 0.4415609836578369, "learning_rate": 9.062603527197308e-05, "loss": 0.0467, "step": 23820 }, { "grad_norm": 0.381882905960083, "learning_rate": 9.06163944405871e-05, "loss": 0.052, "step": 23830 }, { "grad_norm": 0.31667274236679077, "learning_rate": 9.060674916745327e-05, "loss": 0.0517, "step": 23840 }, { "grad_norm": 0.4934476613998413, "learning_rate": 9.05970994536264e-05, "loss": 0.0421, "step": 23850 }, { "grad_norm": 0.4810314178466797, "learning_rate": 9.05874453001618e-05, "loss": 0.0442, "step": 23860 }, { "grad_norm": 0.41178202629089355, "learning_rate": 9.057778670811517e-05, "loss": 0.0449, "step": 23870 }, { "grad_norm": 0.5174605250358582, "learning_rate": 9.056812367854281e-05, "loss": 0.0446, "step": 23880 }, { "grad_norm": 0.4524012506008148, "learning_rate": 9.055845621250143e-05, "loss": 0.0425, "step": 23890 }, { "grad_norm": 0.6591270565986633, "learning_rate": 9.054878431104825e-05, "loss": 0.037, "step": 23900 }, { "grad_norm": 0.37848541140556335, "learning_rate": 9.0539107975241e-05, "loss": 0.0441, "step": 23910 }, { "grad_norm": 1.3025823831558228, "learning_rate": 9.052942720613784e-05, "loss": 0.0595, "step": 23920 }, { "grad_norm": 0.508029043674469, "learning_rate": 9.051974200479745e-05, "loss": 0.0461, "step": 23930 }, { "grad_norm": 0.4749718904495239, "learning_rate": 9.051005237227901e-05, "loss": 0.0535, "step": 23940 }, { "grad_norm": 0.7616560459136963, "learning_rate": 9.050035830964215e-05, "loss": 0.0445, "step": 23950 }, { "grad_norm": 0.4835522174835205, "learning_rate": 9.049065981794698e-05, "loss": 0.0556, "step": 23960 }, { "grad_norm": 0.4529847800731659, "learning_rate": 9.048095689825414e-05, "loss": 0.0588, "step": 23970 }, { "grad_norm": 0.3924643397331238, "learning_rate": 9.047124955162472e-05, "loss": 0.0467, "step": 23980 }, { "grad_norm": 0.6330369114875793, "learning_rate": 9.046153777912028e-05, "loss": 0.0637, "step": 23990 }, { "grad_norm": 0.5226704478263855, "learning_rate": 9.045182158180292e-05, "loss": 0.05, "step": 24000 }, { "grad_norm": 0.42188313603401184, "learning_rate": 9.044210096073516e-05, "loss": 0.0457, "step": 24010 }, { "grad_norm": 0.49646860361099243, "learning_rate": 9.043237591698004e-05, "loss": 0.0438, "step": 24020 }, { "grad_norm": 0.9279539585113525, "learning_rate": 9.04226464516011e-05, "loss": 0.0552, "step": 24030 }, { "grad_norm": 0.4248162806034088, "learning_rate": 9.041291256566229e-05, "loss": 0.0566, "step": 24040 }, { "grad_norm": 0.5383519530296326, "learning_rate": 9.040317426022814e-05, "loss": 0.0571, "step": 24050 }, { "grad_norm": 0.4084644913673401, "learning_rate": 9.03934315363636e-05, "loss": 0.0509, "step": 24060 }, { "grad_norm": 0.4241016209125519, "learning_rate": 9.038368439513409e-05, "loss": 0.0462, "step": 24070 }, { "grad_norm": 0.9063088297843933, "learning_rate": 9.03739328376056e-05, "loss": 0.0538, "step": 24080 }, { "grad_norm": 0.4569801986217499, "learning_rate": 9.036417686484451e-05, "loss": 0.0508, "step": 24090 }, { "grad_norm": 0.47610169649124146, "learning_rate": 9.035441647791773e-05, "loss": 0.0514, "step": 24100 }, { "grad_norm": 0.3879718780517578, "learning_rate": 9.034465167789263e-05, "loss": 0.0487, "step": 24110 }, { "grad_norm": 0.3733871579170227, "learning_rate": 9.033488246583706e-05, "loss": 0.043, "step": 24120 }, { "grad_norm": 0.3796119689941406, "learning_rate": 9.032510884281941e-05, "loss": 0.0646, "step": 24130 }, { "grad_norm": 0.48665711283683777, "learning_rate": 9.031533080990848e-05, "loss": 0.0484, "step": 24140 }, { "grad_norm": 0.33815905451774597, "learning_rate": 9.030554836817358e-05, "loss": 0.0431, "step": 24150 }, { "grad_norm": 0.9575137495994568, "learning_rate": 9.029576151868451e-05, "loss": 0.0442, "step": 24160 }, { "grad_norm": 0.4863518476486206, "learning_rate": 9.028597026251155e-05, "loss": 0.0542, "step": 24170 }, { "grad_norm": 0.5578774809837341, "learning_rate": 9.027617460072547e-05, "loss": 0.0504, "step": 24180 }, { "grad_norm": 0.7427324652671814, "learning_rate": 9.026637453439745e-05, "loss": 0.0481, "step": 24190 }, { "grad_norm": 0.30437371134757996, "learning_rate": 9.025657006459927e-05, "loss": 0.0519, "step": 24200 }, { "grad_norm": 0.6622511148452759, "learning_rate": 9.024676119240311e-05, "loss": 0.0579, "step": 24210 }, { "grad_norm": 0.3221023678779602, "learning_rate": 9.023694791888166e-05, "loss": 0.0549, "step": 24220 }, { "grad_norm": 0.4583461582660675, "learning_rate": 9.022713024510808e-05, "loss": 0.0612, "step": 24230 }, { "grad_norm": 0.3218613266944885, "learning_rate": 9.021730817215601e-05, "loss": 0.048, "step": 24240 }, { "grad_norm": 0.3160427510738373, "learning_rate": 9.02074817010996e-05, "loss": 0.0381, "step": 24250 }, { "grad_norm": 0.4849420189857483, "learning_rate": 9.019765083301342e-05, "loss": 0.0477, "step": 24260 }, { "grad_norm": 0.4592702388763428, "learning_rate": 9.01878155689726e-05, "loss": 0.0601, "step": 24270 }, { "grad_norm": 0.4273412525653839, "learning_rate": 9.017797591005268e-05, "loss": 0.0566, "step": 24280 }, { "grad_norm": 0.39211907982826233, "learning_rate": 9.016813185732972e-05, "loss": 0.0556, "step": 24290 }, { "grad_norm": 0.390764981508255, "learning_rate": 9.015828341188027e-05, "loss": 0.0589, "step": 24300 }, { "grad_norm": 0.6223324537277222, "learning_rate": 9.01484305747813e-05, "loss": 0.0626, "step": 24310 }, { "grad_norm": 0.5826742053031921, "learning_rate": 9.013857334711033e-05, "loss": 0.0484, "step": 24320 }, { "grad_norm": 0.3237043619155884, "learning_rate": 9.012871172994534e-05, "loss": 0.0411, "step": 24330 }, { "grad_norm": 0.4736843407154083, "learning_rate": 9.011884572436476e-05, "loss": 0.0514, "step": 24340 }, { "grad_norm": 0.39559271931648254, "learning_rate": 9.010897533144754e-05, "loss": 0.0477, "step": 24350 }, { "grad_norm": 0.4967421591281891, "learning_rate": 9.009910055227306e-05, "loss": 0.0535, "step": 24360 }, { "grad_norm": 0.6606371998786926, "learning_rate": 9.008922138792124e-05, "loss": 0.0558, "step": 24370 }, { "grad_norm": 0.5151311159133911, "learning_rate": 9.007933783947244e-05, "loss": 0.056, "step": 24380 }, { "grad_norm": 0.815518856048584, "learning_rate": 9.006944990800752e-05, "loss": 0.0491, "step": 24390 }, { "grad_norm": 0.53105628490448, "learning_rate": 9.005955759460779e-05, "loss": 0.0504, "step": 24400 }, { "grad_norm": 0.8976193070411682, "learning_rate": 9.004966090035508e-05, "loss": 0.0458, "step": 24410 }, { "grad_norm": 0.5969977378845215, "learning_rate": 9.003975982633166e-05, "loss": 0.0449, "step": 24420 }, { "grad_norm": 0.5624619126319885, "learning_rate": 9.00298543736203e-05, "loss": 0.055, "step": 24430 }, { "grad_norm": 0.493548184633255, "learning_rate": 9.001994454330427e-05, "loss": 0.0581, "step": 24440 }, { "grad_norm": 0.7431007027626038, "learning_rate": 9.001003033646727e-05, "loss": 0.0733, "step": 24450 }, { "grad_norm": 0.32027339935302734, "learning_rate": 9.00001117541935e-05, "loss": 0.0517, "step": 24460 }, { "grad_norm": 0.38349008560180664, "learning_rate": 8.999018879756764e-05, "loss": 0.0516, "step": 24470 }, { "grad_norm": 0.5371146202087402, "learning_rate": 8.998026146767487e-05, "loss": 0.0512, "step": 24480 }, { "grad_norm": 0.4838242530822754, "learning_rate": 8.99703297656008e-05, "loss": 0.0558, "step": 24490 }, { "grad_norm": 0.4796748161315918, "learning_rate": 8.996039369243156e-05, "loss": 0.0604, "step": 24500 }, { "grad_norm": 0.7382066249847412, "learning_rate": 8.995045324925378e-05, "loss": 0.0483, "step": 24510 }, { "grad_norm": 0.47724148631095886, "learning_rate": 8.994050843715448e-05, "loss": 0.0572, "step": 24520 }, { "grad_norm": 0.3748643696308136, "learning_rate": 8.993055925722121e-05, "loss": 0.0434, "step": 24530 }, { "grad_norm": 0.42071810364723206, "learning_rate": 8.992060571054202e-05, "loss": 0.0462, "step": 24540 }, { "grad_norm": 0.5495966076850891, "learning_rate": 8.991064779820542e-05, "loss": 0.046, "step": 24550 }, { "grad_norm": 0.40511468052864075, "learning_rate": 8.990068552130036e-05, "loss": 0.0346, "step": 24560 }, { "grad_norm": 0.3048292100429535, "learning_rate": 8.989071888091634e-05, "loss": 0.0449, "step": 24570 }, { "grad_norm": 0.6758398413658142, "learning_rate": 8.988074787814329e-05, "loss": 0.0454, "step": 24580 }, { "grad_norm": 0.4523245394229889, "learning_rate": 8.987077251407158e-05, "loss": 0.0505, "step": 24590 }, { "grad_norm": 0.40614569187164307, "learning_rate": 8.986079278979216e-05, "loss": 0.0475, "step": 24600 }, { "grad_norm": 0.5588261485099792, "learning_rate": 8.985080870639635e-05, "loss": 0.0613, "step": 24610 }, { "grad_norm": 0.27166709303855896, "learning_rate": 8.984082026497603e-05, "loss": 0.0463, "step": 24620 }, { "grad_norm": 0.3759322166442871, "learning_rate": 8.98308274666235e-05, "loss": 0.0575, "step": 24630 }, { "grad_norm": 0.41008517146110535, "learning_rate": 8.982083031243155e-05, "loss": 0.0454, "step": 24640 }, { "grad_norm": 0.4199810028076172, "learning_rate": 8.98108288034935e-05, "loss": 0.0479, "step": 24650 }, { "grad_norm": 0.7794217467308044, "learning_rate": 8.980082294090305e-05, "loss": 0.0523, "step": 24660 }, { "grad_norm": 0.5956878662109375, "learning_rate": 8.979081272575443e-05, "loss": 0.0594, "step": 24670 }, { "grad_norm": 0.4691842496395111, "learning_rate": 8.978079815914236e-05, "loss": 0.0432, "step": 24680 }, { "grad_norm": 0.525235652923584, "learning_rate": 8.977077924216202e-05, "loss": 0.0467, "step": 24690 }, { "grad_norm": 0.6984302401542664, "learning_rate": 8.976075597590905e-05, "loss": 0.0545, "step": 24700 }, { "grad_norm": 0.3442275822162628, "learning_rate": 8.975072836147958e-05, "loss": 0.049, "step": 24710 }, { "grad_norm": 0.6287702322006226, "learning_rate": 8.974069639997025e-05, "loss": 0.0457, "step": 24720 }, { "grad_norm": 0.45447537302970886, "learning_rate": 8.973066009247808e-05, "loss": 0.0513, "step": 24730 }, { "grad_norm": 0.43832460045814514, "learning_rate": 8.972061944010066e-05, "loss": 0.0548, "step": 24740 }, { "grad_norm": 0.4837369918823242, "learning_rate": 8.971057444393603e-05, "loss": 0.0561, "step": 24750 }, { "grad_norm": 0.5375605821609497, "learning_rate": 8.970052510508268e-05, "loss": 0.0599, "step": 24760 }, { "grad_norm": 0.33053043484687805, "learning_rate": 8.969047142463959e-05, "loss": 0.0447, "step": 24770 }, { "grad_norm": 0.5856555104255676, "learning_rate": 8.968041340370621e-05, "loss": 0.0529, "step": 24780 }, { "grad_norm": 0.30723777413368225, "learning_rate": 8.96703510433825e-05, "loss": 0.0334, "step": 24790 }, { "grad_norm": 0.6885170340538025, "learning_rate": 8.966028434476883e-05, "loss": 0.0514, "step": 24800 }, { "grad_norm": 0.4324670135974884, "learning_rate": 8.96502133089661e-05, "loss": 0.0424, "step": 24810 }, { "grad_norm": 0.6387101411819458, "learning_rate": 8.964013793707564e-05, "loss": 0.0519, "step": 24820 }, { "grad_norm": 0.3287966251373291, "learning_rate": 8.963005823019932e-05, "loss": 0.0493, "step": 24830 }, { "grad_norm": 0.4381820857524872, "learning_rate": 8.961997418943939e-05, "loss": 0.0404, "step": 24840 }, { "grad_norm": 0.45173314213752747, "learning_rate": 8.960988581589865e-05, "loss": 0.0486, "step": 24850 }, { "grad_norm": 0.4199291169643402, "learning_rate": 8.959979311068037e-05, "loss": 0.0545, "step": 24860 }, { "grad_norm": 0.4066380262374878, "learning_rate": 8.958969607488823e-05, "loss": 0.0584, "step": 24870 }, { "grad_norm": 0.5653148293495178, "learning_rate": 8.957959470962647e-05, "loss": 0.0506, "step": 24880 }, { "grad_norm": 0.42768505215644836, "learning_rate": 8.956948901599971e-05, "loss": 0.0462, "step": 24890 }, { "grad_norm": 0.43867918848991394, "learning_rate": 8.955937899511315e-05, "loss": 0.0513, "step": 24900 }, { "grad_norm": 0.41345512866973877, "learning_rate": 8.954926464807238e-05, "loss": 0.0509, "step": 24910 }, { "grad_norm": 0.36154934763908386, "learning_rate": 8.953914597598347e-05, "loss": 0.0444, "step": 24920 }, { "grad_norm": 0.4137035012245178, "learning_rate": 8.952902297995303e-05, "loss": 0.0381, "step": 24930 }, { "grad_norm": 0.4940926432609558, "learning_rate": 8.951889566108804e-05, "loss": 0.0447, "step": 24940 }, { "grad_norm": 0.9073460698127747, "learning_rate": 8.950876402049606e-05, "loss": 0.0577, "step": 24950 }, { "grad_norm": 0.5098775029182434, "learning_rate": 8.949862805928504e-05, "loss": 0.0433, "step": 24960 }, { "grad_norm": 0.45747146010398865, "learning_rate": 8.948848777856343e-05, "loss": 0.0518, "step": 24970 }, { "grad_norm": 0.3988809883594513, "learning_rate": 8.947834317944017e-05, "loss": 0.0507, "step": 24980 }, { "grad_norm": 0.4220636785030365, "learning_rate": 8.946819426302466e-05, "loss": 0.0657, "step": 24990 }, { "grad_norm": 0.5524183511734009, "learning_rate": 8.945804103042676e-05, "loss": 0.0569, "step": 25000 }, { "grad_norm": 0.39255449175834656, "learning_rate": 8.944788348275681e-05, "loss": 0.0464, "step": 25010 }, { "grad_norm": 0.4691571295261383, "learning_rate": 8.943772162112565e-05, "loss": 0.0531, "step": 25020 }, { "grad_norm": 0.6907101273536682, "learning_rate": 8.942755544664454e-05, "loss": 0.068, "step": 25030 }, { "grad_norm": 0.41705024242401123, "learning_rate": 8.941738496042525e-05, "loss": 0.0561, "step": 25040 }, { "grad_norm": 0.8377290368080139, "learning_rate": 8.940721016357999e-05, "loss": 0.0558, "step": 25050 }, { "grad_norm": 0.28411081433296204, "learning_rate": 8.939703105722148e-05, "loss": 0.0479, "step": 25060 }, { "grad_norm": 0.5035302042961121, "learning_rate": 8.93868476424629e-05, "loss": 0.0351, "step": 25070 }, { "grad_norm": 0.4736759066581726, "learning_rate": 8.937665992041786e-05, "loss": 0.0463, "step": 25080 }, { "grad_norm": 0.4737817049026489, "learning_rate": 8.93664678922005e-05, "loss": 0.0546, "step": 25090 }, { "grad_norm": 0.3232593536376953, "learning_rate": 8.93562715589254e-05, "loss": 0.0486, "step": 25100 }, { "grad_norm": 0.4814046621322632, "learning_rate": 8.934607092170762e-05, "loss": 0.0603, "step": 25110 }, { "grad_norm": 0.4881901741027832, "learning_rate": 8.933586598166266e-05, "loss": 0.0496, "step": 25120 }, { "grad_norm": 0.47486943006515503, "learning_rate": 8.932565673990655e-05, "loss": 0.0572, "step": 25130 }, { "grad_norm": 0.4549006521701813, "learning_rate": 8.931544319755574e-05, "loss": 0.0449, "step": 25140 }, { "grad_norm": 0.3365626335144043, "learning_rate": 8.930522535572718e-05, "loss": 0.047, "step": 25150 }, { "grad_norm": 0.47899672389030457, "learning_rate": 8.929500321553826e-05, "loss": 0.045, "step": 25160 }, { "grad_norm": 0.5297359228134155, "learning_rate": 8.928477677810686e-05, "loss": 0.0426, "step": 25170 }, { "grad_norm": 0.3716687560081482, "learning_rate": 8.927454604455137e-05, "loss": 0.0512, "step": 25180 }, { "grad_norm": 0.42391592264175415, "learning_rate": 8.926431101599053e-05, "loss": 0.0558, "step": 25190 }, { "grad_norm": 0.4651886820793152, "learning_rate": 8.925407169354369e-05, "loss": 0.0455, "step": 25200 }, { "grad_norm": 0.5852547287940979, "learning_rate": 8.92438280783306e-05, "loss": 0.0507, "step": 25210 }, { "grad_norm": 0.4509921669960022, "learning_rate": 8.923358017147146e-05, "loss": 0.0507, "step": 25220 }, { "grad_norm": 0.4644327461719513, "learning_rate": 8.922332797408697e-05, "loss": 0.0596, "step": 25230 }, { "grad_norm": 0.38124996423721313, "learning_rate": 8.921307148729831e-05, "loss": 0.0425, "step": 25240 }, { "grad_norm": 0.44727879762649536, "learning_rate": 8.920281071222712e-05, "loss": 0.0446, "step": 25250 }, { "grad_norm": 0.547166109085083, "learning_rate": 8.919254564999548e-05, "loss": 0.0466, "step": 25260 }, { "grad_norm": 0.48372939229011536, "learning_rate": 8.918227630172598e-05, "loss": 0.045, "step": 25270 }, { "grad_norm": 0.39953187108039856, "learning_rate": 8.917200266854165e-05, "loss": 0.0526, "step": 25280 }, { "grad_norm": 0.40391427278518677, "learning_rate": 8.9161724751566e-05, "loss": 0.0464, "step": 25290 }, { "grad_norm": 0.3928813338279724, "learning_rate": 8.915144255192302e-05, "loss": 0.04, "step": 25300 }, { "grad_norm": 0.38518664240837097, "learning_rate": 8.914115607073714e-05, "loss": 0.0491, "step": 25310 }, { "grad_norm": 0.33632683753967285, "learning_rate": 8.913086530913327e-05, "loss": 0.0444, "step": 25320 }, { "grad_norm": 0.3997834026813507, "learning_rate": 8.912057026823681e-05, "loss": 0.0397, "step": 25330 }, { "grad_norm": 0.42039626836776733, "learning_rate": 8.91102709491736e-05, "loss": 0.0416, "step": 25340 }, { "grad_norm": 0.460361123085022, "learning_rate": 8.909996735306996e-05, "loss": 0.0517, "step": 25350 }, { "grad_norm": 0.44686752557754517, "learning_rate": 8.908965948105268e-05, "loss": 0.0384, "step": 25360 }, { "grad_norm": 0.5728374123573303, "learning_rate": 8.907934733424901e-05, "loss": 0.0516, "step": 25370 }, { "grad_norm": 0.8602591156959534, "learning_rate": 8.906903091378666e-05, "loss": 0.0485, "step": 25380 }, { "grad_norm": 0.4838699996471405, "learning_rate": 8.905871022079384e-05, "loss": 0.0489, "step": 25390 }, { "grad_norm": 0.35340291261672974, "learning_rate": 8.90483852563992e-05, "loss": 0.0407, "step": 25400 }, { "grad_norm": 0.751585841178894, "learning_rate": 8.903805602173185e-05, "loss": 0.0415, "step": 25410 }, { "grad_norm": 0.7102996706962585, "learning_rate": 8.902772251792137e-05, "loss": 0.0443, "step": 25420 }, { "grad_norm": 0.6128044128417969, "learning_rate": 8.901738474609786e-05, "loss": 0.0568, "step": 25430 }, { "grad_norm": 0.7715702056884766, "learning_rate": 8.900704270739179e-05, "loss": 0.049, "step": 25440 }, { "grad_norm": 0.41871708631515503, "learning_rate": 8.89966964029342e-05, "loss": 0.0635, "step": 25450 }, { "grad_norm": 0.4373708963394165, "learning_rate": 8.898634583385652e-05, "loss": 0.0506, "step": 25460 }, { "grad_norm": 0.41949009895324707, "learning_rate": 8.897599100129065e-05, "loss": 0.0506, "step": 25470 }, { "grad_norm": 0.540843665599823, "learning_rate": 8.896563190636903e-05, "loss": 0.0474, "step": 25480 }, { "grad_norm": 0.46848708391189575, "learning_rate": 8.895526855022448e-05, "loss": 0.0467, "step": 25490 }, { "grad_norm": 0.5348319411277771, "learning_rate": 8.894490093399033e-05, "loss": 0.0537, "step": 25500 }, { "grad_norm": 0.3634887635707855, "learning_rate": 8.893452905880035e-05, "loss": 0.0535, "step": 25510 }, { "grad_norm": 0.4454406797885895, "learning_rate": 8.892415292578883e-05, "loss": 0.0437, "step": 25520 }, { "grad_norm": 0.37853989005088806, "learning_rate": 8.891377253609046e-05, "loss": 0.0474, "step": 25530 }, { "grad_norm": 0.42121657729148865, "learning_rate": 8.890338789084043e-05, "loss": 0.0407, "step": 25540 }, { "grad_norm": 0.49983617663383484, "learning_rate": 8.88929989911744e-05, "loss": 0.0499, "step": 25550 }, { "grad_norm": 0.5387226939201355, "learning_rate": 8.888260583822847e-05, "loss": 0.0513, "step": 25560 }, { "grad_norm": 0.3078398108482361, "learning_rate": 8.887220843313921e-05, "loss": 0.051, "step": 25570 }, { "grad_norm": 0.5189229249954224, "learning_rate": 8.88618067770437e-05, "loss": 0.049, "step": 25580 }, { "grad_norm": 0.698570966720581, "learning_rate": 8.885140087107942e-05, "loss": 0.0394, "step": 25590 }, { "grad_norm": 0.3890557885169983, "learning_rate": 8.884099071638436e-05, "loss": 0.0412, "step": 25600 }, { "grad_norm": 0.4057479500770569, "learning_rate": 8.883057631409695e-05, "loss": 0.0531, "step": 25610 }, { "grad_norm": 0.7452201247215271, "learning_rate": 8.882015766535608e-05, "loss": 0.0546, "step": 25620 }, { "grad_norm": 0.5584287047386169, "learning_rate": 8.880973477130115e-05, "loss": 0.0487, "step": 25630 }, { "grad_norm": 0.35018348693847656, "learning_rate": 8.879930763307197e-05, "loss": 0.0412, "step": 25640 }, { "grad_norm": 0.3181014358997345, "learning_rate": 8.878887625180884e-05, "loss": 0.0496, "step": 25650 }, { "grad_norm": 0.3693796694278717, "learning_rate": 8.877844062865253e-05, "loss": 0.041, "step": 25660 }, { "grad_norm": 0.4293934106826782, "learning_rate": 8.876800076474424e-05, "loss": 0.0586, "step": 25670 }, { "grad_norm": 0.5600841045379639, "learning_rate": 8.875755666122568e-05, "loss": 0.0518, "step": 25680 }, { "grad_norm": 0.37015920877456665, "learning_rate": 8.8747108319239e-05, "loss": 0.0436, "step": 25690 }, { "grad_norm": 0.43575143814086914, "learning_rate": 8.87366557399268e-05, "loss": 0.0513, "step": 25700 }, { "grad_norm": 0.5436950922012329, "learning_rate": 8.872619892443217e-05, "loss": 0.0522, "step": 25710 }, { "grad_norm": 0.43222588300704956, "learning_rate": 8.871573787389865e-05, "loss": 0.0512, "step": 25720 }, { "grad_norm": 0.4354841411113739, "learning_rate": 8.870527258947024e-05, "loss": 0.0527, "step": 25730 }, { "grad_norm": 0.41719794273376465, "learning_rate": 8.869480307229143e-05, "loss": 0.0391, "step": 25740 }, { "grad_norm": 0.3815498352050781, "learning_rate": 8.868432932350712e-05, "loss": 0.0386, "step": 25750 }, { "grad_norm": 0.540518581867218, "learning_rate": 8.867385134426272e-05, "loss": 0.0379, "step": 25760 }, { "grad_norm": 0.4770578444004059, "learning_rate": 8.866336913570407e-05, "loss": 0.0512, "step": 25770 }, { "grad_norm": 0.5073477625846863, "learning_rate": 8.865288269897751e-05, "loss": 0.0327, "step": 25780 }, { "grad_norm": 0.354473352432251, "learning_rate": 8.864239203522981e-05, "loss": 0.0522, "step": 25790 }, { "grad_norm": 0.5365971922874451, "learning_rate": 8.863189714560822e-05, "loss": 0.0466, "step": 25800 }, { "grad_norm": 0.34868183732032776, "learning_rate": 8.862139803126043e-05, "loss": 0.0403, "step": 25810 }, { "grad_norm": 0.8780081272125244, "learning_rate": 8.861089469333463e-05, "loss": 0.043, "step": 25820 }, { "grad_norm": 0.5529897212982178, "learning_rate": 8.860038713297944e-05, "loss": 0.0456, "step": 25830 }, { "grad_norm": 0.8162475824356079, "learning_rate": 8.858987535134394e-05, "loss": 0.0685, "step": 25840 }, { "grad_norm": 0.5728287696838379, "learning_rate": 8.857935934957769e-05, "loss": 0.0537, "step": 25850 }, { "grad_norm": 0.466460257768631, "learning_rate": 8.856883912883071e-05, "loss": 0.0487, "step": 25860 }, { "grad_norm": 0.8755597472190857, "learning_rate": 8.855831469025346e-05, "loss": 0.0492, "step": 25870 }, { "grad_norm": 0.4832601249217987, "learning_rate": 8.854778603499689e-05, "loss": 0.0558, "step": 25880 }, { "grad_norm": 0.46083709597587585, "learning_rate": 8.85372531642124e-05, "loss": 0.048, "step": 25890 }, { "grad_norm": 0.4288637936115265, "learning_rate": 8.852671607905185e-05, "loss": 0.0396, "step": 25900 }, { "grad_norm": 0.5325437188148499, "learning_rate": 8.851617478066754e-05, "loss": 0.0501, "step": 25910 }, { "grad_norm": 0.38289737701416016, "learning_rate": 8.850562927021227e-05, "loss": 0.0466, "step": 25920 }, { "grad_norm": 0.637108325958252, "learning_rate": 8.849507954883928e-05, "loss": 0.0515, "step": 25930 }, { "grad_norm": 0.5915395617485046, "learning_rate": 8.848452561770226e-05, "loss": 0.0438, "step": 25940 }, { "grad_norm": 0.3517693281173706, "learning_rate": 8.847396747795538e-05, "loss": 0.047, "step": 25950 }, { "grad_norm": 0.42795711755752563, "learning_rate": 8.846340513075327e-05, "loss": 0.0543, "step": 25960 }, { "grad_norm": 0.5579240918159485, "learning_rate": 8.845283857725099e-05, "loss": 0.0432, "step": 25970 }, { "grad_norm": 0.5399824976921082, "learning_rate": 8.844226781860409e-05, "loss": 0.0491, "step": 25980 }, { "grad_norm": 0.409812331199646, "learning_rate": 8.84316928559686e-05, "loss": 0.048, "step": 25990 }, { "grad_norm": 0.28493690490722656, "learning_rate": 8.842111369050094e-05, "loss": 0.0418, "step": 26000 }, { "grad_norm": 0.4258131682872772, "learning_rate": 8.841053032335808e-05, "loss": 0.0471, "step": 26010 }, { "grad_norm": 0.38404008746147156, "learning_rate": 8.839994275569735e-05, "loss": 0.041, "step": 26020 }, { "grad_norm": 0.5164791941642761, "learning_rate": 8.838935098867662e-05, "loss": 0.0537, "step": 26030 }, { "grad_norm": 0.5216286778450012, "learning_rate": 8.837875502345418e-05, "loss": 0.0609, "step": 26040 }, { "grad_norm": 0.4630051851272583, "learning_rate": 8.83681548611888e-05, "loss": 0.0413, "step": 26050 }, { "grad_norm": 0.32549986243247986, "learning_rate": 8.835755050303969e-05, "loss": 0.0458, "step": 26060 }, { "grad_norm": 0.39972612261772156, "learning_rate": 8.834694195016653e-05, "loss": 0.0491, "step": 26070 }, { "grad_norm": 0.6063929796218872, "learning_rate": 8.833632920372942e-05, "loss": 0.0694, "step": 26080 }, { "grad_norm": 0.44610241055488586, "learning_rate": 8.832571226488903e-05, "loss": 0.0695, "step": 26090 }, { "grad_norm": 0.4855516850948334, "learning_rate": 8.831509113480634e-05, "loss": 0.0575, "step": 26100 }, { "grad_norm": 0.46127450466156006, "learning_rate": 8.83044658146429e-05, "loss": 0.0525, "step": 26110 }, { "grad_norm": 0.37666526436805725, "learning_rate": 8.829383630556067e-05, "loss": 0.0463, "step": 26120 }, { "grad_norm": 0.3496986925601959, "learning_rate": 8.828320260872207e-05, "loss": 0.0485, "step": 26130 }, { "grad_norm": 0.5180019736289978, "learning_rate": 8.827256472529e-05, "loss": 0.0534, "step": 26140 }, { "grad_norm": 0.4194466471672058, "learning_rate": 8.826192265642778e-05, "loss": 0.05, "step": 26150 }, { "grad_norm": 0.7997886538505554, "learning_rate": 8.825127640329923e-05, "loss": 0.0538, "step": 26160 }, { "grad_norm": 0.5240404009819031, "learning_rate": 8.824062596706861e-05, "loss": 0.0527, "step": 26170 }, { "grad_norm": 0.42930087447166443, "learning_rate": 8.822997134890062e-05, "loss": 0.0486, "step": 26180 }, { "grad_norm": 0.5061740279197693, "learning_rate": 8.821931254996044e-05, "loss": 0.0588, "step": 26190 }, { "grad_norm": 0.5447565317153931, "learning_rate": 8.82086495714137e-05, "loss": 0.0525, "step": 26200 }, { "grad_norm": 0.4617311358451843, "learning_rate": 8.81979824144265e-05, "loss": 0.0526, "step": 26210 }, { "grad_norm": 0.8489009141921997, "learning_rate": 8.818731108016536e-05, "loss": 0.0574, "step": 26220 }, { "grad_norm": 0.4000844359397888, "learning_rate": 8.81766355697973e-05, "loss": 0.0542, "step": 26230 }, { "grad_norm": 0.4260730743408203, "learning_rate": 8.816595588448977e-05, "loss": 0.053, "step": 26240 }, { "grad_norm": 0.655647337436676, "learning_rate": 8.81552720254107e-05, "loss": 0.0592, "step": 26250 }, { "grad_norm": 0.45927858352661133, "learning_rate": 8.814458399372842e-05, "loss": 0.0468, "step": 26260 }, { "grad_norm": 0.49153584241867065, "learning_rate": 8.813389179061181e-05, "loss": 0.0435, "step": 26270 }, { "grad_norm": 0.51890629529953, "learning_rate": 8.812319541723012e-05, "loss": 0.0575, "step": 26280 }, { "grad_norm": 0.6037246584892273, "learning_rate": 8.811249487475309e-05, "loss": 0.0599, "step": 26290 }, { "grad_norm": 0.5007054805755615, "learning_rate": 8.810179016435092e-05, "loss": 0.0536, "step": 26300 }, { "grad_norm": 0.5095990300178528, "learning_rate": 8.809108128719428e-05, "loss": 0.0699, "step": 26310 }, { "grad_norm": 0.6779108047485352, "learning_rate": 8.808036824445424e-05, "loss": 0.0442, "step": 26320 }, { "grad_norm": 0.4109426438808441, "learning_rate": 8.806965103730238e-05, "loss": 0.0485, "step": 26330 }, { "grad_norm": 0.4556422829627991, "learning_rate": 8.805892966691074e-05, "loss": 0.0416, "step": 26340 }, { "grad_norm": 0.3497326076030731, "learning_rate": 8.804820413445175e-05, "loss": 0.0417, "step": 26350 }, { "grad_norm": 0.33925575017929077, "learning_rate": 8.803747444109837e-05, "loss": 0.0463, "step": 26360 }, { "grad_norm": 0.4789049029350281, "learning_rate": 8.802674058802399e-05, "loss": 0.0391, "step": 26370 }, { "grad_norm": 0.46441885828971863, "learning_rate": 8.801600257640241e-05, "loss": 0.0477, "step": 26380 }, { "grad_norm": 0.3664313852787018, "learning_rate": 8.800526040740795e-05, "loss": 0.0441, "step": 26390 }, { "grad_norm": 0.42592912912368774, "learning_rate": 8.799451408221535e-05, "loss": 0.0498, "step": 26400 }, { "grad_norm": 0.3381921648979187, "learning_rate": 8.798376360199982e-05, "loss": 0.0483, "step": 26410 }, { "grad_norm": 0.30406343936920166, "learning_rate": 8.797300896793701e-05, "loss": 0.0444, "step": 26420 }, { "grad_norm": 0.4729672372341156, "learning_rate": 8.796225018120302e-05, "loss": 0.0528, "step": 26430 }, { "grad_norm": 0.3066086173057556, "learning_rate": 8.795148724297444e-05, "loss": 0.0458, "step": 26440 }, { "grad_norm": 0.520540177822113, "learning_rate": 8.794072015442825e-05, "loss": 0.0491, "step": 26450 }, { "grad_norm": 0.41177067160606384, "learning_rate": 8.792994891674198e-05, "loss": 0.0536, "step": 26460 }, { "grad_norm": 0.6057921648025513, "learning_rate": 8.79191735310935e-05, "loss": 0.0424, "step": 26470 }, { "grad_norm": 0.45206964015960693, "learning_rate": 8.790839399866122e-05, "loss": 0.0498, "step": 26480 }, { "grad_norm": 0.6212587952613831, "learning_rate": 8.789761032062397e-05, "loss": 0.0618, "step": 26490 }, { "grad_norm": 0.5979846119880676, "learning_rate": 8.788682249816103e-05, "loss": 0.0608, "step": 26500 }, { "grad_norm": 0.6406435370445251, "learning_rate": 8.787603053245215e-05, "loss": 0.0594, "step": 26510 }, { "grad_norm": 0.41339266300201416, "learning_rate": 8.78652344246775e-05, "loss": 0.0415, "step": 26520 }, { "grad_norm": 0.48673292994499207, "learning_rate": 8.785443417601776e-05, "loss": 0.0488, "step": 26530 }, { "grad_norm": 0.4214744567871094, "learning_rate": 8.784362978765401e-05, "loss": 0.0593, "step": 26540 }, { "grad_norm": 0.5597013235092163, "learning_rate": 8.783282126076779e-05, "loss": 0.0452, "step": 26550 }, { "grad_norm": 0.41672447323799133, "learning_rate": 8.782200859654112e-05, "loss": 0.0523, "step": 26560 }, { "grad_norm": 1.1065798997879028, "learning_rate": 8.781119179615646e-05, "loss": 0.0473, "step": 26570 }, { "grad_norm": 0.465244859457016, "learning_rate": 8.780037086079674e-05, "loss": 0.0502, "step": 26580 }, { "grad_norm": 0.4396522343158722, "learning_rate": 8.778954579164527e-05, "loss": 0.0511, "step": 26590 }, { "grad_norm": 0.47233304381370544, "learning_rate": 8.777871658988588e-05, "loss": 0.0456, "step": 26600 }, { "grad_norm": 0.7077588438987732, "learning_rate": 8.776788325670285e-05, "loss": 0.0417, "step": 26610 }, { "grad_norm": 0.41102534532546997, "learning_rate": 8.775704579328089e-05, "loss": 0.0507, "step": 26620 }, { "grad_norm": 0.701504647731781, "learning_rate": 8.774620420080517e-05, "loss": 0.053, "step": 26630 }, { "grad_norm": 0.9224864840507507, "learning_rate": 8.773535848046131e-05, "loss": 0.0567, "step": 26640 }, { "grad_norm": 0.5139950513839722, "learning_rate": 8.772450863343538e-05, "loss": 0.046, "step": 26650 }, { "grad_norm": 0.4670935869216919, "learning_rate": 8.77136546609139e-05, "loss": 0.0629, "step": 26660 }, { "grad_norm": 0.3843276798725128, "learning_rate": 8.770279656408385e-05, "loss": 0.0428, "step": 26670 }, { "grad_norm": 0.5072686672210693, "learning_rate": 8.769193434413265e-05, "loss": 0.0554, "step": 26680 }, { "grad_norm": 0.4395225942134857, "learning_rate": 8.76810680022482e-05, "loss": 0.0505, "step": 26690 }, { "grad_norm": 0.9578447341918945, "learning_rate": 8.767019753961878e-05, "loss": 0.0649, "step": 26700 }, { "grad_norm": 0.4151153564453125, "learning_rate": 8.765932295743321e-05, "loss": 0.0527, "step": 26710 }, { "grad_norm": 0.2758937478065491, "learning_rate": 8.764844425688068e-05, "loss": 0.0452, "step": 26720 }, { "grad_norm": 0.806148111820221, "learning_rate": 8.763756143915092e-05, "loss": 0.0446, "step": 26730 }, { "grad_norm": 0.8419967293739319, "learning_rate": 8.7626674505434e-05, "loss": 0.0506, "step": 26740 }, { "grad_norm": 0.4908873736858368, "learning_rate": 8.761578345692053e-05, "loss": 0.0692, "step": 26750 }, { "grad_norm": 0.493073970079422, "learning_rate": 8.760488829480156e-05, "loss": 0.0556, "step": 26760 }, { "grad_norm": 0.360880047082901, "learning_rate": 8.759398902026854e-05, "loss": 0.0395, "step": 26770 }, { "grad_norm": 0.33357471227645874, "learning_rate": 8.758308563451339e-05, "loss": 0.0416, "step": 26780 }, { "grad_norm": 0.46522220969200134, "learning_rate": 8.75721781387285e-05, "loss": 0.043, "step": 26790 }, { "grad_norm": 0.29969632625579834, "learning_rate": 8.75612665341067e-05, "loss": 0.0396, "step": 26800 }, { "grad_norm": 0.2676364481449127, "learning_rate": 8.755035082184126e-05, "loss": 0.0391, "step": 26810 }, { "grad_norm": 0.3258746564388275, "learning_rate": 8.753943100312592e-05, "loss": 0.0384, "step": 26820 }, { "grad_norm": 0.34080177545547485, "learning_rate": 8.752850707915484e-05, "loss": 0.0365, "step": 26830 }, { "grad_norm": 0.7377243638038635, "learning_rate": 8.751757905112264e-05, "loss": 0.0625, "step": 26840 }, { "grad_norm": 0.4185936748981476, "learning_rate": 8.75066469202244e-05, "loss": 0.0384, "step": 26850 }, { "grad_norm": 0.3964887857437134, "learning_rate": 8.749571068765567e-05, "loss": 0.0506, "step": 26860 }, { "grad_norm": 0.5790490508079529, "learning_rate": 8.748477035461238e-05, "loss": 0.0439, "step": 26870 }, { "grad_norm": 0.4767417311668396, "learning_rate": 8.747382592229095e-05, "loss": 0.0524, "step": 26880 }, { "grad_norm": 0.5044834017753601, "learning_rate": 8.746287739188828e-05, "loss": 0.0519, "step": 26890 }, { "grad_norm": 0.4475903809070587, "learning_rate": 8.745192476460165e-05, "loss": 0.054, "step": 26900 }, { "grad_norm": 0.46878159046173096, "learning_rate": 8.744096804162882e-05, "loss": 0.0511, "step": 26910 }, { "grad_norm": 0.530643105506897, "learning_rate": 8.743000722416804e-05, "loss": 0.049, "step": 26920 }, { "grad_norm": 0.3059093952178955, "learning_rate": 8.741904231341793e-05, "loss": 0.0455, "step": 26930 }, { "grad_norm": 0.6566867232322693, "learning_rate": 8.740807331057762e-05, "loss": 0.0664, "step": 26940 }, { "grad_norm": 0.5822293162345886, "learning_rate": 8.739710021684667e-05, "loss": 0.0569, "step": 26950 }, { "grad_norm": 0.3730621337890625, "learning_rate": 8.738612303342503e-05, "loss": 0.0459, "step": 26960 }, { "grad_norm": 0.3333081901073456, "learning_rate": 8.73751417615132e-05, "loss": 0.0566, "step": 26970 }, { "grad_norm": 0.5325162410736084, "learning_rate": 8.736415640231208e-05, "loss": 0.0511, "step": 26980 }, { "grad_norm": 0.4876365661621094, "learning_rate": 8.735316695702297e-05, "loss": 0.0493, "step": 26990 }, { "grad_norm": 0.4769127666950226, "learning_rate": 8.734217342684769e-05, "loss": 0.0495, "step": 27000 }, { "grad_norm": 0.30868130922317505, "learning_rate": 8.733117581298847e-05, "loss": 0.0323, "step": 27010 }, { "grad_norm": 0.45643383264541626, "learning_rate": 8.732017411664796e-05, "loss": 0.0453, "step": 27020 }, { "grad_norm": 0.3696606457233429, "learning_rate": 8.730916833902936e-05, "loss": 0.0508, "step": 27030 }, { "grad_norm": 0.43122804164886475, "learning_rate": 8.729815848133618e-05, "loss": 0.0499, "step": 27040 }, { "grad_norm": 0.4698239266872406, "learning_rate": 8.728714454477247e-05, "loss": 0.0421, "step": 27050 }, { "grad_norm": 0.522671103477478, "learning_rate": 8.727612653054269e-05, "loss": 0.0484, "step": 27060 }, { "grad_norm": 0.42845284938812256, "learning_rate": 8.726510443985176e-05, "loss": 0.0442, "step": 27070 }, { "grad_norm": 0.5866246223449707, "learning_rate": 8.725407827390503e-05, "loss": 0.043, "step": 27080 }, { "grad_norm": 0.4357488453388214, "learning_rate": 8.724304803390833e-05, "loss": 0.0493, "step": 27090 }, { "grad_norm": 0.4797114431858063, "learning_rate": 8.723201372106788e-05, "loss": 0.0408, "step": 27100 }, { "grad_norm": 0.6364606618881226, "learning_rate": 8.722097533659038e-05, "loss": 0.0537, "step": 27110 }, { "grad_norm": 0.5742764472961426, "learning_rate": 8.720993288168299e-05, "loss": 0.0461, "step": 27120 }, { "grad_norm": 0.6298643350601196, "learning_rate": 8.719888635755327e-05, "loss": 0.0468, "step": 27130 }, { "grad_norm": 0.3662937581539154, "learning_rate": 8.718783576540928e-05, "loss": 0.0482, "step": 27140 }, { "grad_norm": 0.4702870845794678, "learning_rate": 8.717678110645948e-05, "loss": 0.0434, "step": 27150 }, { "grad_norm": 0.6243777871131897, "learning_rate": 8.716572238191279e-05, "loss": 0.0482, "step": 27160 }, { "grad_norm": 0.48543083667755127, "learning_rate": 8.715465959297857e-05, "loss": 0.0548, "step": 27170 }, { "grad_norm": 0.7357565760612488, "learning_rate": 8.714359274086665e-05, "loss": 0.0539, "step": 27180 }, { "grad_norm": 0.48504579067230225, "learning_rate": 8.713252182678726e-05, "loss": 0.0473, "step": 27190 }, { "grad_norm": 0.37637829780578613, "learning_rate": 8.712144685195112e-05, "loss": 0.0532, "step": 27200 }, { "grad_norm": 0.34336474537849426, "learning_rate": 8.711036781756936e-05, "loss": 0.0469, "step": 27210 }, { "grad_norm": 0.3671169579029083, "learning_rate": 8.709928472485357e-05, "loss": 0.0385, "step": 27220 }, { "grad_norm": 0.34983885288238525, "learning_rate": 8.708819757501579e-05, "loss": 0.0384, "step": 27230 }, { "grad_norm": 0.3057531416416168, "learning_rate": 8.707710636926846e-05, "loss": 0.043, "step": 27240 }, { "grad_norm": 0.5539920330047607, "learning_rate": 8.706601110882455e-05, "loss": 0.0567, "step": 27250 }, { "grad_norm": 0.48012280464172363, "learning_rate": 8.705491179489738e-05, "loss": 0.0413, "step": 27260 }, { "grad_norm": 0.42678922414779663, "learning_rate": 8.704380842870077e-05, "loss": 0.0435, "step": 27270 }, { "grad_norm": 0.3345668613910675, "learning_rate": 8.703270101144895e-05, "loss": 0.039, "step": 27280 }, { "grad_norm": 0.33935996890068054, "learning_rate": 8.702158954435664e-05, "loss": 0.0401, "step": 27290 }, { "grad_norm": 0.4982677102088928, "learning_rate": 8.701047402863896e-05, "loss": 0.0367, "step": 27300 }, { "grad_norm": 0.5026471018791199, "learning_rate": 8.699935446551148e-05, "loss": 0.0454, "step": 27310 }, { "grad_norm": 0.2986285090446472, "learning_rate": 8.698823085619022e-05, "loss": 0.0444, "step": 27320 }, { "grad_norm": 0.3882739543914795, "learning_rate": 8.697710320189166e-05, "loss": 0.0486, "step": 27330 }, { "grad_norm": 0.4579845070838928, "learning_rate": 8.696597150383268e-05, "loss": 0.0433, "step": 27340 }, { "grad_norm": 0.4962519705295563, "learning_rate": 8.695483576323063e-05, "loss": 0.048, "step": 27350 }, { "grad_norm": 0.376204252243042, "learning_rate": 8.69436959813033e-05, "loss": 0.058, "step": 27360 }, { "grad_norm": 0.40951037406921387, "learning_rate": 8.693255215926892e-05, "loss": 0.0455, "step": 27370 }, { "grad_norm": 0.39558809995651245, "learning_rate": 8.692140429834617e-05, "loss": 0.0487, "step": 27380 }, { "grad_norm": 0.45495641231536865, "learning_rate": 8.691025239975415e-05, "loss": 0.0533, "step": 27390 }, { "grad_norm": 0.46385475993156433, "learning_rate": 8.689909646471243e-05, "loss": 0.0407, "step": 27400 }, { "grad_norm": 0.4894523620605469, "learning_rate": 8.688793649444099e-05, "loss": 0.0599, "step": 27410 }, { "grad_norm": 0.5600683093070984, "learning_rate": 8.687677249016029e-05, "loss": 0.0548, "step": 27420 }, { "grad_norm": 0.5333238840103149, "learning_rate": 8.686560445309118e-05, "loss": 0.0557, "step": 27430 }, { "grad_norm": 0.6341641545295715, "learning_rate": 8.685443238445499e-05, "loss": 0.052, "step": 27440 }, { "grad_norm": 0.38394758105278015, "learning_rate": 8.68432562854735e-05, "loss": 0.0492, "step": 27450 }, { "grad_norm": 0.40482044219970703, "learning_rate": 8.683207615736887e-05, "loss": 0.0431, "step": 27460 }, { "grad_norm": 0.48395824432373047, "learning_rate": 8.682089200136379e-05, "loss": 0.0654, "step": 27470 }, { "grad_norm": 0.5031662583351135, "learning_rate": 8.680970381868132e-05, "loss": 0.0549, "step": 27480 }, { "grad_norm": 0.4008173942565918, "learning_rate": 8.679851161054498e-05, "loss": 0.052, "step": 27490 }, { "grad_norm": 0.5136702060699463, "learning_rate": 8.678731537817873e-05, "loss": 0.0458, "step": 27500 }, { "grad_norm": 1.2421821355819702, "learning_rate": 8.677611512280697e-05, "loss": 0.0486, "step": 27510 }, { "grad_norm": 0.3774801790714264, "learning_rate": 8.676491084565457e-05, "loss": 0.0551, "step": 27520 }, { "grad_norm": 0.5084092617034912, "learning_rate": 8.675370254794678e-05, "loss": 0.0593, "step": 27530 }, { "grad_norm": 0.3996627628803253, "learning_rate": 8.674249023090935e-05, "loss": 0.0504, "step": 27540 }, { "grad_norm": 0.4391305446624756, "learning_rate": 8.673127389576843e-05, "loss": 0.0518, "step": 27550 }, { "grad_norm": 0.4744311273097992, "learning_rate": 8.67200535437506e-05, "loss": 0.0387, "step": 27560 }, { "grad_norm": 0.4920186400413513, "learning_rate": 8.670882917608296e-05, "loss": 0.0537, "step": 27570 }, { "grad_norm": 0.5032496452331543, "learning_rate": 8.669760079399292e-05, "loss": 0.0417, "step": 27580 }, { "grad_norm": 0.32280322909355164, "learning_rate": 8.668636839870845e-05, "loss": 0.0404, "step": 27590 }, { "grad_norm": 0.4446963369846344, "learning_rate": 8.667513199145789e-05, "loss": 0.0513, "step": 27600 }, { "grad_norm": 0.4163692891597748, "learning_rate": 8.666389157347002e-05, "loss": 0.0513, "step": 27610 }, { "grad_norm": 0.3897112011909485, "learning_rate": 8.66526471459741e-05, "loss": 0.0437, "step": 27620 }, { "grad_norm": 0.5946875810623169, "learning_rate": 8.66413987101998e-05, "loss": 0.0436, "step": 27630 }, { "grad_norm": 0.3463907837867737, "learning_rate": 8.663014626737723e-05, "loss": 0.0391, "step": 27640 }, { "grad_norm": 0.357673317193985, "learning_rate": 8.661888981873691e-05, "loss": 0.0448, "step": 27650 }, { "grad_norm": 0.538374662399292, "learning_rate": 8.660762936550988e-05, "loss": 0.057, "step": 27660 }, { "grad_norm": 0.5414177775382996, "learning_rate": 8.659636490892753e-05, "loss": 0.0482, "step": 27670 }, { "grad_norm": 0.34652069211006165, "learning_rate": 8.658509645022174e-05, "loss": 0.0454, "step": 27680 }, { "grad_norm": 0.48793354630470276, "learning_rate": 8.657382399062481e-05, "loss": 0.0423, "step": 27690 }, { "grad_norm": 0.5308260321617126, "learning_rate": 8.656254753136946e-05, "loss": 0.0582, "step": 27700 }, { "grad_norm": 0.49928444623947144, "learning_rate": 8.655126707368891e-05, "loss": 0.0382, "step": 27710 }, { "grad_norm": 0.33113110065460205, "learning_rate": 8.653998261881672e-05, "loss": 0.0443, "step": 27720 }, { "grad_norm": 0.4206564724445343, "learning_rate": 8.652869416798699e-05, "loss": 0.0455, "step": 27730 }, { "grad_norm": 0.31947410106658936, "learning_rate": 8.651740172243417e-05, "loss": 0.0459, "step": 27740 }, { "grad_norm": 0.39684000611305237, "learning_rate": 8.65061052833932e-05, "loss": 0.0426, "step": 27750 }, { "grad_norm": 0.3223803639411926, "learning_rate": 8.649480485209945e-05, "loss": 0.0404, "step": 27760 }, { "grad_norm": 0.4101308584213257, "learning_rate": 8.64835004297887e-05, "loss": 0.0414, "step": 27770 }, { "grad_norm": 0.4254971444606781, "learning_rate": 8.64721920176972e-05, "loss": 0.0492, "step": 27780 }, { "grad_norm": 0.4287854731082916, "learning_rate": 8.646087961706164e-05, "loss": 0.0562, "step": 27790 }, { "grad_norm": 0.47203558683395386, "learning_rate": 8.644956322911908e-05, "loss": 0.0507, "step": 27800 }, { "grad_norm": 0.5649179220199585, "learning_rate": 8.643824285510709e-05, "loss": 0.0543, "step": 27810 }, { "grad_norm": 0.319826602935791, "learning_rate": 8.642691849626364e-05, "loss": 0.0446, "step": 27820 }, { "grad_norm": 0.36929160356521606, "learning_rate": 8.641559015382717e-05, "loss": 0.0431, "step": 27830 }, { "grad_norm": 0.8343546986579895, "learning_rate": 8.640425782903649e-05, "loss": 0.0571, "step": 27840 }, { "grad_norm": 0.6280784010887146, "learning_rate": 8.639292152313091e-05, "loss": 0.0544, "step": 27850 }, { "grad_norm": 0.49655893445014954, "learning_rate": 8.638158123735015e-05, "loss": 0.0505, "step": 27860 }, { "grad_norm": 0.6004173755645752, "learning_rate": 8.637023697293436e-05, "loss": 0.0411, "step": 27870 }, { "grad_norm": 0.5471030473709106, "learning_rate": 8.635888873112414e-05, "loss": 0.0593, "step": 27880 }, { "grad_norm": 0.48647066950798035, "learning_rate": 8.634753651316052e-05, "loss": 0.0411, "step": 27890 }, { "grad_norm": 0.2614755630493164, "learning_rate": 8.633618032028496e-05, "loss": 0.041, "step": 27900 }, { "grad_norm": 0.4129497706890106, "learning_rate": 8.632482015373934e-05, "loss": 0.0424, "step": 27910 }, { "grad_norm": 0.45121467113494873, "learning_rate": 8.6313456014766e-05, "loss": 0.0549, "step": 27920 }, { "grad_norm": 0.32887718081474304, "learning_rate": 8.630208790460771e-05, "loss": 0.052, "step": 27930 }, { "grad_norm": 0.4617141783237457, "learning_rate": 8.629071582450768e-05, "loss": 0.0484, "step": 27940 }, { "grad_norm": 0.5573215484619141, "learning_rate": 8.62793397757095e-05, "loss": 0.0478, "step": 27950 }, { "grad_norm": 0.3681383728981018, "learning_rate": 8.626795975945729e-05, "loss": 0.0613, "step": 27960 }, { "grad_norm": 0.47750240564346313, "learning_rate": 8.625657577699551e-05, "loss": 0.0423, "step": 27970 }, { "grad_norm": 0.34510114789009094, "learning_rate": 8.624518782956914e-05, "loss": 0.0439, "step": 27980 }, { "grad_norm": 0.31336653232574463, "learning_rate": 8.62337959184235e-05, "loss": 0.0485, "step": 27990 }, { "grad_norm": 0.5052139163017273, "learning_rate": 8.622240004480441e-05, "loss": 0.0542, "step": 28000 }, { "grad_norm": 0.3621792495250702, "learning_rate": 8.621100020995814e-05, "loss": 0.0432, "step": 28010 }, { "grad_norm": 0.45440489053726196, "learning_rate": 8.619959641513132e-05, "loss": 0.048, "step": 28020 }, { "grad_norm": 0.4225250482559204, "learning_rate": 8.618818866157105e-05, "loss": 0.0419, "step": 28030 }, { "grad_norm": 0.6300266981124878, "learning_rate": 8.617677695052487e-05, "loss": 0.0437, "step": 28040 }, { "grad_norm": 0.2495489865541458, "learning_rate": 8.616536128324078e-05, "loss": 0.0421, "step": 28050 }, { "grad_norm": 0.4828283190727234, "learning_rate": 8.615394166096712e-05, "loss": 0.042, "step": 28060 }, { "grad_norm": 0.6103078126907349, "learning_rate": 8.614251808495279e-05, "loss": 0.0479, "step": 28070 }, { "grad_norm": 0.3509331941604614, "learning_rate": 8.6131090556447e-05, "loss": 0.0426, "step": 28080 }, { "grad_norm": 0.8046698570251465, "learning_rate": 8.611965907669947e-05, "loss": 0.0487, "step": 28090 }, { "grad_norm": 0.49986132979393005, "learning_rate": 8.610822364696034e-05, "loss": 0.049, "step": 28100 }, { "grad_norm": 0.39973562955856323, "learning_rate": 8.609678426848015e-05, "loss": 0.0501, "step": 28110 }, { "grad_norm": 0.41180089116096497, "learning_rate": 8.60853409425099e-05, "loss": 0.0509, "step": 28120 }, { "grad_norm": 0.6791853308677673, "learning_rate": 8.607389367030104e-05, "loss": 0.0516, "step": 28130 }, { "grad_norm": 0.5049699544906616, "learning_rate": 8.606244245310538e-05, "loss": 0.0504, "step": 28140 }, { "grad_norm": 0.599846601486206, "learning_rate": 8.605098729217525e-05, "loss": 0.054, "step": 28150 }, { "grad_norm": 0.4478357434272766, "learning_rate": 8.603952818876335e-05, "loss": 0.0577, "step": 28160 }, { "grad_norm": 0.44812291860580444, "learning_rate": 8.602806514412281e-05, "loss": 0.0474, "step": 28170 }, { "grad_norm": 0.7771289944648743, "learning_rate": 8.601659815950726e-05, "loss": 0.0579, "step": 28180 }, { "grad_norm": 0.47068649530410767, "learning_rate": 8.600512723617067e-05, "loss": 0.0588, "step": 28190 }, { "grad_norm": 0.5219124555587769, "learning_rate": 8.59936523753675e-05, "loss": 0.0448, "step": 28200 }, { "grad_norm": 0.4336729347705841, "learning_rate": 8.598217357835264e-05, "loss": 0.0413, "step": 28210 }, { "grad_norm": 0.36898431181907654, "learning_rate": 8.597069084638135e-05, "loss": 0.0386, "step": 28220 }, { "grad_norm": 0.41362136602401733, "learning_rate": 8.595920418070939e-05, "loss": 0.0393, "step": 28230 }, { "grad_norm": 0.3817110061645508, "learning_rate": 8.594771358259295e-05, "loss": 0.0335, "step": 28240 }, { "grad_norm": 0.38529863953590393, "learning_rate": 8.593621905328858e-05, "loss": 0.0491, "step": 28250 }, { "grad_norm": 0.33637186884880066, "learning_rate": 8.592472059405333e-05, "loss": 0.0449, "step": 28260 }, { "grad_norm": 0.45994463562965393, "learning_rate": 8.591321820614464e-05, "loss": 0.0387, "step": 28270 }, { "grad_norm": 0.33317193388938904, "learning_rate": 8.590171189082041e-05, "loss": 0.043, "step": 28280 }, { "grad_norm": 0.3209112584590912, "learning_rate": 8.589020164933894e-05, "loss": 0.0408, "step": 28290 }, { "grad_norm": 0.37435540556907654, "learning_rate": 8.587868748295898e-05, "loss": 0.039, "step": 28300 }, { "grad_norm": 0.37682637572288513, "learning_rate": 8.586716939293971e-05, "loss": 0.0449, "step": 28310 }, { "grad_norm": 0.41529539227485657, "learning_rate": 8.58556473805407e-05, "loss": 0.0511, "step": 28320 }, { "grad_norm": 0.4546007812023163, "learning_rate": 8.584412144702202e-05, "loss": 0.045, "step": 28330 }, { "grad_norm": 0.4563957452774048, "learning_rate": 8.58325915936441e-05, "loss": 0.0471, "step": 28340 }, { "grad_norm": 0.6145401000976562, "learning_rate": 8.582105782166783e-05, "loss": 0.0449, "step": 28350 }, { "grad_norm": 0.409586638212204, "learning_rate": 8.580952013235455e-05, "loss": 0.037, "step": 28360 }, { "grad_norm": 0.30825796723365784, "learning_rate": 8.579797852696596e-05, "loss": 0.047, "step": 28370 }, { "grad_norm": 0.3195056915283203, "learning_rate": 8.578643300676428e-05, "loss": 0.0549, "step": 28380 }, { "grad_norm": 0.5752893686294556, "learning_rate": 8.577488357301209e-05, "loss": 0.0484, "step": 28390 }, { "grad_norm": 0.5758222937583923, "learning_rate": 8.576333022697242e-05, "loss": 0.042, "step": 28400 }, { "grad_norm": 0.4597967565059662, "learning_rate": 8.575177296990873e-05, "loss": 0.041, "step": 28410 }, { "grad_norm": 0.8062965869903564, "learning_rate": 8.574021180308489e-05, "loss": 0.0531, "step": 28420 }, { "grad_norm": 0.46586138010025024, "learning_rate": 8.572864672776523e-05, "loss": 0.0394, "step": 28430 }, { "grad_norm": 0.5187187790870667, "learning_rate": 8.571707774521447e-05, "loss": 0.0335, "step": 28440 }, { "grad_norm": 0.24073074758052826, "learning_rate": 8.57055048566978e-05, "loss": 0.0363, "step": 28450 }, { "grad_norm": 0.469059020280838, "learning_rate": 8.569392806348078e-05, "loss": 0.0412, "step": 28460 }, { "grad_norm": 0.3130365312099457, "learning_rate": 8.568234736682947e-05, "loss": 0.0323, "step": 28470 }, { "grad_norm": 0.5358191728591919, "learning_rate": 8.567076276801029e-05, "loss": 0.0542, "step": 28480 }, { "grad_norm": 0.7911509275436401, "learning_rate": 8.565917426829013e-05, "loss": 0.0363, "step": 28490 }, { "grad_norm": 0.31716489791870117, "learning_rate": 8.564758186893628e-05, "loss": 0.0408, "step": 28500 }, { "grad_norm": 0.49071574211120605, "learning_rate": 8.563598557121649e-05, "loss": 0.0513, "step": 28510 }, { "grad_norm": 0.4184907376766205, "learning_rate": 8.562438537639888e-05, "loss": 0.046, "step": 28520 }, { "grad_norm": 0.507368803024292, "learning_rate": 8.561278128575206e-05, "loss": 0.0735, "step": 28530 }, { "grad_norm": 0.35543158650398254, "learning_rate": 8.5601173300545e-05, "loss": 0.0434, "step": 28540 }, { "grad_norm": 0.6231217980384827, "learning_rate": 8.558956142204717e-05, "loss": 0.0567, "step": 28550 }, { "grad_norm": 0.8346620798110962, "learning_rate": 8.55779456515284e-05, "loss": 0.0501, "step": 28560 }, { "grad_norm": 0.49423739314079285, "learning_rate": 8.556632599025898e-05, "loss": 0.0523, "step": 28570 }, { "grad_norm": 0.4918862283229828, "learning_rate": 8.555470243950964e-05, "loss": 0.0501, "step": 28580 }, { "grad_norm": 0.4732723832130432, "learning_rate": 8.554307500055148e-05, "loss": 0.0495, "step": 28590 }, { "grad_norm": 0.4951348900794983, "learning_rate": 8.553144367465609e-05, "loss": 0.046, "step": 28600 }, { "grad_norm": 0.33634668588638306, "learning_rate": 8.551980846309544e-05, "loss": 0.0619, "step": 28610 }, { "grad_norm": 0.4863744378089905, "learning_rate": 8.550816936714193e-05, "loss": 0.047, "step": 28620 }, { "grad_norm": 0.42291268706321716, "learning_rate": 8.549652638806841e-05, "loss": 0.0467, "step": 28630 }, { "grad_norm": 0.2603558897972107, "learning_rate": 8.548487952714812e-05, "loss": 0.0419, "step": 28640 }, { "grad_norm": 0.30579498410224915, "learning_rate": 8.547322878565478e-05, "loss": 0.0419, "step": 28650 }, { "grad_norm": 0.3770729899406433, "learning_rate": 8.546157416486245e-05, "loss": 0.0647, "step": 28660 }, { "grad_norm": 0.4773498773574829, "learning_rate": 8.54499156660457e-05, "loss": 0.0391, "step": 28670 }, { "grad_norm": 0.4056383967399597, "learning_rate": 8.543825329047947e-05, "loss": 0.0562, "step": 28680 }, { "grad_norm": 0.7536836862564087, "learning_rate": 8.542658703943913e-05, "loss": 0.0539, "step": 28690 }, { "grad_norm": 0.29700079560279846, "learning_rate": 8.541491691420051e-05, "loss": 0.0424, "step": 28700 }, { "grad_norm": 0.37438011169433594, "learning_rate": 8.54032429160398e-05, "loss": 0.0412, "step": 28710 }, { "grad_norm": 0.24382205307483673, "learning_rate": 8.539156504623369e-05, "loss": 0.0414, "step": 28720 }, { "grad_norm": 0.45286762714385986, "learning_rate": 8.537988330605923e-05, "loss": 0.0718, "step": 28730 }, { "grad_norm": 0.5249919295310974, "learning_rate": 8.536819769679393e-05, "loss": 0.0473, "step": 28740 }, { "grad_norm": 0.37396878004074097, "learning_rate": 8.53565082197157e-05, "loss": 0.0497, "step": 28750 }, { "grad_norm": 1.0280011892318726, "learning_rate": 8.534481487610289e-05, "loss": 0.0528, "step": 28760 }, { "grad_norm": 0.3717027008533478, "learning_rate": 8.533311766723428e-05, "loss": 0.0388, "step": 28770 }, { "grad_norm": 0.5460662841796875, "learning_rate": 8.532141659438901e-05, "loss": 0.0557, "step": 28780 }, { "grad_norm": 0.2956964373588562, "learning_rate": 8.530971165884675e-05, "loss": 0.0522, "step": 28790 }, { "grad_norm": 0.4197459816932678, "learning_rate": 8.529800286188752e-05, "loss": 0.0449, "step": 28800 }, { "grad_norm": 0.4107639193534851, "learning_rate": 8.528629020479175e-05, "loss": 0.037, "step": 28810 }, { "grad_norm": 0.5774354338645935, "learning_rate": 8.527457368884033e-05, "loss": 0.0396, "step": 28820 }, { "grad_norm": 0.32444268465042114, "learning_rate": 8.526285331531458e-05, "loss": 0.0343, "step": 28830 }, { "grad_norm": 0.48847171664237976, "learning_rate": 8.525112908549621e-05, "loss": 0.0425, "step": 28840 }, { "grad_norm": 0.6404227614402771, "learning_rate": 8.523940100066735e-05, "loss": 0.0443, "step": 28850 }, { "grad_norm": 0.4010327160358429, "learning_rate": 8.52276690621106e-05, "loss": 0.0368, "step": 28860 }, { "grad_norm": 0.3939807415008545, "learning_rate": 8.521593327110889e-05, "loss": 0.0488, "step": 28870 }, { "grad_norm": 0.3973313271999359, "learning_rate": 8.520419362894569e-05, "loss": 0.0587, "step": 28880 }, { "grad_norm": 0.37789079546928406, "learning_rate": 8.51924501369048e-05, "loss": 0.0579, "step": 28890 }, { "grad_norm": 0.6481369733810425, "learning_rate": 8.518070279627047e-05, "loss": 0.0509, "step": 28900 }, { "grad_norm": 0.333546906709671, "learning_rate": 8.516895160832737e-05, "loss": 0.0545, "step": 28910 }, { "grad_norm": 0.5215402245521545, "learning_rate": 8.515719657436061e-05, "loss": 0.0461, "step": 28920 }, { "grad_norm": 0.3348951041698456, "learning_rate": 8.514543769565568e-05, "loss": 0.0458, "step": 28930 }, { "grad_norm": 0.3940006196498871, "learning_rate": 8.513367497349853e-05, "loss": 0.0507, "step": 28940 }, { "grad_norm": 0.3156372010707855, "learning_rate": 8.51219084091755e-05, "loss": 0.0489, "step": 28950 }, { "grad_norm": 0.4715906083583832, "learning_rate": 8.511013800397338e-05, "loss": 0.0483, "step": 28960 }, { "grad_norm": 0.4825160801410675, "learning_rate": 8.509836375917937e-05, "loss": 0.0507, "step": 28970 }, { "grad_norm": 0.6096702218055725, "learning_rate": 8.508658567608104e-05, "loss": 0.0489, "step": 28980 }, { "grad_norm": 0.5595227479934692, "learning_rate": 8.507480375596647e-05, "loss": 0.0455, "step": 28990 }, { "grad_norm": 0.5507151484489441, "learning_rate": 8.506301800012408e-05, "loss": 0.0528, "step": 29000 }, { "grad_norm": 0.4645865857601166, "learning_rate": 8.505122840984278e-05, "loss": 0.0505, "step": 29010 }, { "grad_norm": 0.42065316438674927, "learning_rate": 8.503943498641182e-05, "loss": 0.0504, "step": 29020 }, { "grad_norm": 0.4340366721153259, "learning_rate": 8.502763773112095e-05, "loss": 0.0452, "step": 29030 }, { "grad_norm": 0.7176566123962402, "learning_rate": 8.501583664526026e-05, "loss": 0.0477, "step": 29040 }, { "grad_norm": 0.4402715861797333, "learning_rate": 8.500403173012032e-05, "loss": 0.0542, "step": 29050 }, { "grad_norm": 0.40684500336647034, "learning_rate": 8.499222298699211e-05, "loss": 0.0474, "step": 29060 }, { "grad_norm": 0.40090832114219666, "learning_rate": 8.498041041716701e-05, "loss": 0.0482, "step": 29070 }, { "grad_norm": 0.273483008146286, "learning_rate": 8.496859402193681e-05, "loss": 0.0453, "step": 29080 }, { "grad_norm": 0.4663798213005066, "learning_rate": 8.495677380259374e-05, "loss": 0.042, "step": 29090 }, { "grad_norm": 0.2715776264667511, "learning_rate": 8.494494976043045e-05, "loss": 0.043, "step": 29100 }, { "grad_norm": 0.3681122958660126, "learning_rate": 8.493312189673998e-05, "loss": 0.0355, "step": 29110 }, { "grad_norm": 0.4180848002433777, "learning_rate": 8.492129021281584e-05, "loss": 0.0429, "step": 29120 }, { "grad_norm": 0.3393029272556305, "learning_rate": 8.490945470995188e-05, "loss": 0.0293, "step": 29130 }, { "grad_norm": 0.8143341541290283, "learning_rate": 8.489761538944247e-05, "loss": 0.0468, "step": 29140 }, { "grad_norm": 0.42033109068870544, "learning_rate": 8.48857722525823e-05, "loss": 0.0462, "step": 29150 }, { "grad_norm": 0.26176637411117554, "learning_rate": 8.487392530066652e-05, "loss": 0.0494, "step": 29160 }, { "grad_norm": 0.4914891719818115, "learning_rate": 8.486207453499069e-05, "loss": 0.0505, "step": 29170 }, { "grad_norm": 1.1855045557022095, "learning_rate": 8.485021995685082e-05, "loss": 0.0548, "step": 29180 }, { "grad_norm": 0.8688516020774841, "learning_rate": 8.483836156754328e-05, "loss": 0.0583, "step": 29190 }, { "grad_norm": 0.6954649686813354, "learning_rate": 8.482649936836491e-05, "loss": 0.0519, "step": 29200 }, { "grad_norm": 0.46095365285873413, "learning_rate": 8.481463336061293e-05, "loss": 0.0501, "step": 29210 }, { "grad_norm": 0.6010869145393372, "learning_rate": 8.480276354558496e-05, "loss": 0.0488, "step": 29220 }, { "grad_norm": 0.2881067097187042, "learning_rate": 8.479088992457913e-05, "loss": 0.0389, "step": 29230 }, { "grad_norm": 0.34964922070503235, "learning_rate": 8.477901249889387e-05, "loss": 0.0543, "step": 29240 }, { "grad_norm": 0.3364631235599518, "learning_rate": 8.47671312698281e-05, "loss": 0.0491, "step": 29250 }, { "grad_norm": 0.3399202823638916, "learning_rate": 8.475524623868112e-05, "loss": 0.034, "step": 29260 }, { "grad_norm": 0.41629284620285034, "learning_rate": 8.474335740675266e-05, "loss": 0.0411, "step": 29270 }, { "grad_norm": 0.4369162619113922, "learning_rate": 8.473146477534289e-05, "loss": 0.0395, "step": 29280 }, { "grad_norm": 0.3861738443374634, "learning_rate": 8.471956834575232e-05, "loss": 0.0451, "step": 29290 }, { "grad_norm": 0.39107346534729004, "learning_rate": 8.470766811928197e-05, "loss": 0.0414, "step": 29300 }, { "grad_norm": 0.3469196856021881, "learning_rate": 8.469576409723323e-05, "loss": 0.052, "step": 29310 }, { "grad_norm": 0.5771698951721191, "learning_rate": 8.468385628090788e-05, "loss": 0.0465, "step": 29320 }, { "grad_norm": 0.3855162560939789, "learning_rate": 8.467194467160815e-05, "loss": 0.0447, "step": 29330 }, { "grad_norm": 0.4886637330055237, "learning_rate": 8.466002927063667e-05, "loss": 0.041, "step": 29340 }, { "grad_norm": 0.5562793612480164, "learning_rate": 8.464811007929651e-05, "loss": 0.0411, "step": 29350 }, { "grad_norm": 0.5324130058288574, "learning_rate": 8.463618709889114e-05, "loss": 0.056, "step": 29360 }, { "grad_norm": 0.5495750904083252, "learning_rate": 8.462426033072442e-05, "loss": 0.056, "step": 29370 }, { "grad_norm": 0.4125517010688782, "learning_rate": 8.461232977610061e-05, "loss": 0.0381, "step": 29380 }, { "grad_norm": 0.468404620885849, "learning_rate": 8.46003954363245e-05, "loss": 0.0357, "step": 29390 }, { "grad_norm": 0.4608524739742279, "learning_rate": 8.458845731270115e-05, "loss": 0.0494, "step": 29400 }, { "grad_norm": 0.6597328782081604, "learning_rate": 8.45765154065361e-05, "loss": 0.0483, "step": 29410 }, { "grad_norm": 0.6423056125640869, "learning_rate": 8.456456971913532e-05, "loss": 0.0381, "step": 29420 }, { "grad_norm": 0.6484696865081787, "learning_rate": 8.455262025180517e-05, "loss": 0.0445, "step": 29430 }, { "grad_norm": 0.3635101318359375, "learning_rate": 8.454066700585242e-05, "loss": 0.0479, "step": 29440 }, { "grad_norm": 0.47309452295303345, "learning_rate": 8.452870998258423e-05, "loss": 0.0464, "step": 29450 }, { "grad_norm": 0.35774075984954834, "learning_rate": 8.451674918330825e-05, "loss": 0.0411, "step": 29460 }, { "grad_norm": 0.5556983351707458, "learning_rate": 8.450478460933246e-05, "loss": 0.0449, "step": 29470 }, { "grad_norm": 0.5034259557723999, "learning_rate": 8.449281626196532e-05, "loss": 0.0554, "step": 29480 }, { "grad_norm": 0.5206724405288696, "learning_rate": 8.448084414251564e-05, "loss": 0.0523, "step": 29490 }, { "grad_norm": 0.46574679017066956, "learning_rate": 8.446886825229271e-05, "loss": 0.0547, "step": 29500 }, { "grad_norm": 0.34941330552101135, "learning_rate": 8.445688859260615e-05, "loss": 0.0448, "step": 29510 }, { "grad_norm": 0.4868461489677429, "learning_rate": 8.444490516476606e-05, "loss": 0.0446, "step": 29520 }, { "grad_norm": 0.3452550172805786, "learning_rate": 8.443291797008293e-05, "loss": 0.0487, "step": 29530 }, { "grad_norm": 0.3957383334636688, "learning_rate": 8.442092700986765e-05, "loss": 0.0579, "step": 29540 }, { "grad_norm": 0.4894969165325165, "learning_rate": 8.440893228543156e-05, "loss": 0.0536, "step": 29550 }, { "grad_norm": 0.2871648371219635, "learning_rate": 8.439693379808638e-05, "loss": 0.0351, "step": 29560 }, { "grad_norm": 0.4391933083534241, "learning_rate": 8.43849315491442e-05, "loss": 0.0467, "step": 29570 }, { "grad_norm": 0.600469708442688, "learning_rate": 8.437292553991763e-05, "loss": 0.0387, "step": 29580 }, { "grad_norm": 0.38091370463371277, "learning_rate": 8.436091577171959e-05, "loss": 0.0423, "step": 29590 }, { "grad_norm": 0.8749332427978516, "learning_rate": 8.434890224586347e-05, "loss": 0.0486, "step": 29600 }, { "grad_norm": 0.27631548047065735, "learning_rate": 8.433688496366303e-05, "loss": 0.0354, "step": 29610 }, { "grad_norm": 0.46985840797424316, "learning_rate": 8.432486392643248e-05, "loss": 0.0469, "step": 29620 }, { "grad_norm": 0.44993823766708374, "learning_rate": 8.431283913548643e-05, "loss": 0.0551, "step": 29630 }, { "grad_norm": 0.5560656189918518, "learning_rate": 8.430081059213985e-05, "loss": 0.0455, "step": 29640 }, { "grad_norm": 0.5051944851875305, "learning_rate": 8.428877829770823e-05, "loss": 0.0573, "step": 29650 }, { "grad_norm": 0.4840853214263916, "learning_rate": 8.427674225350735e-05, "loss": 0.0506, "step": 29660 }, { "grad_norm": 0.44621002674102783, "learning_rate": 8.426470246085347e-05, "loss": 0.0425, "step": 29670 }, { "grad_norm": 0.26754167675971985, "learning_rate": 8.425265892106324e-05, "loss": 0.0347, "step": 29680 }, { "grad_norm": 0.42937231063842773, "learning_rate": 8.424061163545374e-05, "loss": 0.039, "step": 29690 }, { "grad_norm": 0.4516586661338806, "learning_rate": 8.422856060534243e-05, "loss": 0.0351, "step": 29700 }, { "grad_norm": 0.4970287084579468, "learning_rate": 8.421650583204718e-05, "loss": 0.0558, "step": 29710 }, { "grad_norm": 0.5073338747024536, "learning_rate": 8.420444731688633e-05, "loss": 0.0473, "step": 29720 }, { "grad_norm": 0.35951098799705505, "learning_rate": 8.419238506117852e-05, "loss": 0.0518, "step": 29730 }, { "grad_norm": 1.1182336807250977, "learning_rate": 8.418031906624289e-05, "loss": 0.0393, "step": 29740 }, { "grad_norm": 0.5077521204948425, "learning_rate": 8.416824933339898e-05, "loss": 0.0567, "step": 29750 }, { "grad_norm": 0.4876757264137268, "learning_rate": 8.415617586396667e-05, "loss": 0.0458, "step": 29760 }, { "grad_norm": 0.6464158892631531, "learning_rate": 8.414409865926632e-05, "loss": 0.0463, "step": 29770 }, { "grad_norm": 0.3385474383831024, "learning_rate": 8.413201772061867e-05, "loss": 0.0402, "step": 29780 }, { "grad_norm": 0.3602641820907593, "learning_rate": 8.411993304934488e-05, "loss": 0.0392, "step": 29790 }, { "grad_norm": 0.7781320214271545, "learning_rate": 8.410784464676654e-05, "loss": 0.0346, "step": 29800 }, { "grad_norm": 0.3561538755893707, "learning_rate": 8.409575251420556e-05, "loss": 0.0346, "step": 29810 }, { "grad_norm": 0.4162670969963074, "learning_rate": 8.408365665298435e-05, "loss": 0.0495, "step": 29820 }, { "grad_norm": 0.4989272952079773, "learning_rate": 8.40715570644257e-05, "loss": 0.0617, "step": 29830 }, { "grad_norm": 0.5656960010528564, "learning_rate": 8.40594537498528e-05, "loss": 0.0641, "step": 29840 }, { "grad_norm": 0.4848167896270752, "learning_rate": 8.404734671058924e-05, "loss": 0.0451, "step": 29850 }, { "grad_norm": 0.4521695673465729, "learning_rate": 8.403523594795902e-05, "loss": 0.0428, "step": 29860 }, { "grad_norm": 0.48126527667045593, "learning_rate": 8.402312146328659e-05, "loss": 0.0468, "step": 29870 }, { "grad_norm": 0.4051112234592438, "learning_rate": 8.401100325789675e-05, "loss": 0.0414, "step": 29880 }, { "grad_norm": 0.36503568291664124, "learning_rate": 8.399888133311472e-05, "loss": 0.0457, "step": 29890 }, { "grad_norm": 0.4248039424419403, "learning_rate": 8.398675569026613e-05, "loss": 0.0568, "step": 29900 }, { "grad_norm": 0.38183483481407166, "learning_rate": 8.397462633067705e-05, "loss": 0.0455, "step": 29910 }, { "grad_norm": 0.4629870653152466, "learning_rate": 8.396249325567392e-05, "loss": 0.0437, "step": 29920 }, { "grad_norm": 0.45033544301986694, "learning_rate": 8.395035646658357e-05, "loss": 0.0429, "step": 29930 }, { "grad_norm": 0.4109766483306885, "learning_rate": 8.39382159647333e-05, "loss": 0.0404, "step": 29940 }, { "grad_norm": 0.3987467885017395, "learning_rate": 8.392607175145075e-05, "loss": 0.0409, "step": 29950 }, { "grad_norm": 0.3649289608001709, "learning_rate": 8.3913923828064e-05, "loss": 0.0437, "step": 29960 }, { "grad_norm": 0.37328705191612244, "learning_rate": 8.390177219590152e-05, "loss": 0.037, "step": 29970 }, { "grad_norm": 0.37851130962371826, "learning_rate": 8.388961685629222e-05, "loss": 0.0389, "step": 29980 }, { "grad_norm": 0.3949648439884186, "learning_rate": 8.387745781056536e-05, "loss": 0.0352, "step": 29990 }, { "grad_norm": 0.49614953994750977, "learning_rate": 8.386529506005065e-05, "loss": 0.0468, "step": 30000 }, { "grad_norm": 0.5149168372154236, "learning_rate": 8.38531286060782e-05, "loss": 0.0402, "step": 30010 }, { "grad_norm": 0.3682870864868164, "learning_rate": 8.384095844997849e-05, "loss": 0.0466, "step": 30020 }, { "grad_norm": 0.5156713128089905, "learning_rate": 8.382878459308245e-05, "loss": 0.0431, "step": 30030 }, { "grad_norm": 0.39295637607574463, "learning_rate": 8.381660703672138e-05, "loss": 0.0441, "step": 30040 }, { "grad_norm": 0.40300989151000977, "learning_rate": 8.380442578222702e-05, "loss": 0.0609, "step": 30050 }, { "grad_norm": 0.5063155293464661, "learning_rate": 8.379224083093146e-05, "loss": 0.0574, "step": 30060 }, { "grad_norm": 0.48016345500946045, "learning_rate": 8.378005218416727e-05, "loss": 0.0438, "step": 30070 }, { "grad_norm": 0.7707872986793518, "learning_rate": 8.376785984326735e-05, "loss": 0.0435, "step": 30080 }, { "grad_norm": 0.4316428005695343, "learning_rate": 8.375566380956506e-05, "loss": 0.0371, "step": 30090 }, { "grad_norm": 0.4877875745296478, "learning_rate": 8.374346408439411e-05, "loss": 0.0441, "step": 30100 }, { "grad_norm": 0.6369553208351135, "learning_rate": 8.373126066908868e-05, "loss": 0.0594, "step": 30110 }, { "grad_norm": 0.3789628744125366, "learning_rate": 8.371905356498326e-05, "loss": 0.0409, "step": 30120 }, { "grad_norm": 0.42434030771255493, "learning_rate": 8.370684277341288e-05, "loss": 0.0469, "step": 30130 }, { "grad_norm": 0.2996809780597687, "learning_rate": 8.369462829571282e-05, "loss": 0.0471, "step": 30140 }, { "grad_norm": 0.3615964353084564, "learning_rate": 8.36824101332189e-05, "loss": 0.0375, "step": 30150 }, { "grad_norm": 0.5501866936683655, "learning_rate": 8.367018828726721e-05, "loss": 0.0386, "step": 30160 }, { "grad_norm": 0.44963040947914124, "learning_rate": 8.365796275919438e-05, "loss": 0.0473, "step": 30170 }, { "grad_norm": 0.28449901938438416, "learning_rate": 8.364573355033734e-05, "loss": 0.0358, "step": 30180 }, { "grad_norm": 0.32639414072036743, "learning_rate": 8.363350066203346e-05, "loss": 0.0398, "step": 30190 }, { "grad_norm": 0.34503719210624695, "learning_rate": 8.362126409562053e-05, "loss": 0.0381, "step": 30200 }, { "grad_norm": 0.4324975907802582, "learning_rate": 8.360902385243667e-05, "loss": 0.0421, "step": 30210 }, { "grad_norm": 0.45685070753097534, "learning_rate": 8.359677993382052e-05, "loss": 0.0422, "step": 30220 }, { "grad_norm": 0.7219868302345276, "learning_rate": 8.358453234111103e-05, "loss": 0.0457, "step": 30230 }, { "grad_norm": 0.37200719118118286, "learning_rate": 8.357228107564756e-05, "loss": 0.045, "step": 30240 }, { "grad_norm": 0.5104681849479675, "learning_rate": 8.356002613876993e-05, "loss": 0.0453, "step": 30250 }, { "grad_norm": 0.6338528990745544, "learning_rate": 8.35477675318183e-05, "loss": 0.0567, "step": 30260 }, { "grad_norm": 0.3781393766403198, "learning_rate": 8.353550525613323e-05, "loss": 0.0497, "step": 30270 }, { "grad_norm": 0.5006465911865234, "learning_rate": 8.352323931305572e-05, "loss": 0.0598, "step": 30280 }, { "grad_norm": 0.5630226135253906, "learning_rate": 8.351096970392717e-05, "loss": 0.0474, "step": 30290 }, { "grad_norm": 0.4441918730735779, "learning_rate": 8.349869643008937e-05, "loss": 0.0367, "step": 30300 }, { "grad_norm": 0.318427175283432, "learning_rate": 8.348641949288449e-05, "loss": 0.0482, "step": 30310 }, { "grad_norm": 0.3359156847000122, "learning_rate": 8.34741388936551e-05, "loss": 0.0519, "step": 30320 }, { "grad_norm": 0.3816338777542114, "learning_rate": 8.346185463374423e-05, "loss": 0.0652, "step": 30330 }, { "grad_norm": 0.31586021184921265, "learning_rate": 8.344956671449524e-05, "loss": 0.0544, "step": 30340 }, { "grad_norm": 0.3152840733528137, "learning_rate": 8.343727513725192e-05, "loss": 0.0444, "step": 30350 }, { "grad_norm": 0.4117121398448944, "learning_rate": 8.342497990335847e-05, "loss": 0.0432, "step": 30360 }, { "grad_norm": 0.39700213074684143, "learning_rate": 8.341268101415946e-05, "loss": 0.0505, "step": 30370 }, { "grad_norm": 0.7754237055778503, "learning_rate": 8.34003784709999e-05, "loss": 0.0466, "step": 30380 }, { "grad_norm": 0.3818400800228119, "learning_rate": 8.338807227522516e-05, "loss": 0.0617, "step": 30390 }, { "grad_norm": 0.5476786494255066, "learning_rate": 8.337576242818103e-05, "loss": 0.0557, "step": 30400 }, { "grad_norm": 0.4279285669326782, "learning_rate": 8.336344893121372e-05, "loss": 0.0388, "step": 30410 }, { "grad_norm": 0.505021333694458, "learning_rate": 8.335113178566977e-05, "loss": 0.0444, "step": 30420 }, { "grad_norm": 0.5148542523384094, "learning_rate": 8.33388109928962e-05, "loss": 0.042, "step": 30430 }, { "grad_norm": 0.38792842626571655, "learning_rate": 8.33264865542404e-05, "loss": 0.0405, "step": 30440 }, { "grad_norm": 0.3461867868900299, "learning_rate": 8.331415847105013e-05, "loss": 0.0415, "step": 30450 }, { "grad_norm": 0.5535471439361572, "learning_rate": 8.330182674467357e-05, "loss": 0.0476, "step": 30460 }, { "grad_norm": 0.3754257261753082, "learning_rate": 8.32894913764593e-05, "loss": 0.0395, "step": 30470 }, { "grad_norm": 0.43056970834732056, "learning_rate": 8.327715236775633e-05, "loss": 0.0473, "step": 30480 }, { "grad_norm": 0.48843055963516235, "learning_rate": 8.326480971991398e-05, "loss": 0.0501, "step": 30490 }, { "grad_norm": 0.46070316433906555, "learning_rate": 8.325246343428206e-05, "loss": 0.0472, "step": 30500 }, { "grad_norm": 0.32301098108291626, "learning_rate": 8.324011351221072e-05, "loss": 0.0385, "step": 30510 }, { "grad_norm": 0.643093466758728, "learning_rate": 8.322775995505057e-05, "loss": 0.0487, "step": 30520 }, { "grad_norm": 1.1144866943359375, "learning_rate": 8.321540276415254e-05, "loss": 0.0449, "step": 30530 }, { "grad_norm": 0.5214252471923828, "learning_rate": 8.320304194086798e-05, "loss": 0.0453, "step": 30540 }, { "grad_norm": 0.4967045783996582, "learning_rate": 8.31906774865487e-05, "loss": 0.0472, "step": 30550 }, { "grad_norm": 0.4993545413017273, "learning_rate": 8.317830940254682e-05, "loss": 0.0438, "step": 30560 }, { "grad_norm": 0.46190014481544495, "learning_rate": 8.316593769021491e-05, "loss": 0.0519, "step": 30570 }, { "grad_norm": 0.43875211477279663, "learning_rate": 8.315356235090592e-05, "loss": 0.0564, "step": 30580 }, { "grad_norm": 0.32704398036003113, "learning_rate": 8.314118338597319e-05, "loss": 0.0446, "step": 30590 }, { "grad_norm": 0.4631480872631073, "learning_rate": 8.312880079677048e-05, "loss": 0.0449, "step": 30600 }, { "grad_norm": 0.4478025734424591, "learning_rate": 8.311641458465191e-05, "loss": 0.0482, "step": 30610 }, { "grad_norm": 0.46619293093681335, "learning_rate": 8.310402475097205e-05, "loss": 0.0494, "step": 30620 }, { "grad_norm": 0.5448774695396423, "learning_rate": 8.309163129708581e-05, "loss": 0.0492, "step": 30630 }, { "grad_norm": 0.33899784088134766, "learning_rate": 8.307923422434852e-05, "loss": 0.0539, "step": 30640 }, { "grad_norm": 0.6560578942298889, "learning_rate": 8.30668335341159e-05, "loss": 0.0521, "step": 30650 }, { "grad_norm": 0.42127713561058044, "learning_rate": 8.305442922774408e-05, "loss": 0.0472, "step": 30660 }, { "grad_norm": 0.29747268557548523, "learning_rate": 8.304202130658959e-05, "loss": 0.0424, "step": 30670 }, { "grad_norm": 0.6299291253089905, "learning_rate": 8.302960977200931e-05, "loss": 0.0452, "step": 30680 }, { "grad_norm": 1.2836421728134155, "learning_rate": 8.301719462536058e-05, "loss": 0.0377, "step": 30690 }, { "grad_norm": 0.42640966176986694, "learning_rate": 8.300477586800108e-05, "loss": 0.0459, "step": 30700 }, { "grad_norm": 1.104718804359436, "learning_rate": 8.299235350128892e-05, "loss": 0.0498, "step": 30710 }, { "grad_norm": 0.571382462978363, "learning_rate": 8.297992752658258e-05, "loss": 0.0471, "step": 30720 }, { "grad_norm": 0.4205564558506012, "learning_rate": 8.296749794524094e-05, "loss": 0.043, "step": 30730 }, { "grad_norm": 0.5636090636253357, "learning_rate": 8.295506475862332e-05, "loss": 0.0451, "step": 30740 }, { "grad_norm": 0.44641268253326416, "learning_rate": 8.294262796808933e-05, "loss": 0.0418, "step": 30750 }, { "grad_norm": 0.3883179724216461, "learning_rate": 8.293018757499909e-05, "loss": 0.0406, "step": 30760 }, { "grad_norm": 0.3364674150943756, "learning_rate": 8.291774358071305e-05, "loss": 0.0427, "step": 30770 }, { "grad_norm": 0.543708324432373, "learning_rate": 8.290529598659205e-05, "loss": 0.0435, "step": 30780 }, { "grad_norm": 0.5595937371253967, "learning_rate": 8.289284479399738e-05, "loss": 0.0496, "step": 30790 }, { "grad_norm": 0.3229839503765106, "learning_rate": 8.288039000429064e-05, "loss": 0.0392, "step": 30800 }, { "grad_norm": 0.4316699206829071, "learning_rate": 8.286793161883388e-05, "loss": 0.044, "step": 30810 }, { "grad_norm": 0.606605052947998, "learning_rate": 8.285546963898954e-05, "loss": 0.0419, "step": 30820 }, { "grad_norm": 0.5121706128120422, "learning_rate": 8.284300406612044e-05, "loss": 0.0549, "step": 30830 }, { "grad_norm": 0.40931975841522217, "learning_rate": 8.283053490158978e-05, "loss": 0.0511, "step": 30840 }, { "grad_norm": 0.6393462419509888, "learning_rate": 8.28180621467612e-05, "loss": 0.0548, "step": 30850 }, { "grad_norm": 0.39974790811538696, "learning_rate": 8.280558580299867e-05, "loss": 0.0548, "step": 30860 }, { "grad_norm": 0.3864806890487671, "learning_rate": 8.279310587166661e-05, "loss": 0.038, "step": 30870 }, { "grad_norm": 0.39326953887939453, "learning_rate": 8.278062235412978e-05, "loss": 0.0403, "step": 30880 }, { "grad_norm": 0.6232218742370605, "learning_rate": 8.276813525175339e-05, "loss": 0.0516, "step": 30890 }, { "grad_norm": 0.2981538772583008, "learning_rate": 8.2755644565903e-05, "loss": 0.0477, "step": 30900 }, { "grad_norm": 0.4340137243270874, "learning_rate": 8.274315029794454e-05, "loss": 0.0504, "step": 30910 }, { "grad_norm": 0.3778117001056671, "learning_rate": 8.273065244924443e-05, "loss": 0.0522, "step": 30920 }, { "grad_norm": 0.39235448837280273, "learning_rate": 8.271815102116936e-05, "loss": 0.0363, "step": 30930 }, { "grad_norm": 0.33371642231941223, "learning_rate": 8.270564601508648e-05, "loss": 0.0546, "step": 30940 }, { "grad_norm": 0.3511925935745239, "learning_rate": 8.269313743236333e-05, "loss": 0.0424, "step": 30950 }, { "grad_norm": 0.3841322064399719, "learning_rate": 8.268062527436783e-05, "loss": 0.0371, "step": 30960 }, { "grad_norm": 0.49883997440338135, "learning_rate": 8.266810954246828e-05, "loss": 0.0407, "step": 30970 }, { "grad_norm": 0.4275018870830536, "learning_rate": 8.265559023803338e-05, "loss": 0.0389, "step": 30980 }, { "grad_norm": 0.3733880817890167, "learning_rate": 8.264306736243223e-05, "loss": 0.0405, "step": 30990 }, { "grad_norm": 0.35519513487815857, "learning_rate": 8.263054091703432e-05, "loss": 0.0415, "step": 31000 }, { "grad_norm": 0.33332687616348267, "learning_rate": 8.26180109032095e-05, "loss": 0.0431, "step": 31010 }, { "grad_norm": 0.403128445148468, "learning_rate": 8.260547732232807e-05, "loss": 0.0347, "step": 31020 }, { "grad_norm": 0.36191922426223755, "learning_rate": 8.259294017576063e-05, "loss": 0.0395, "step": 31030 }, { "grad_norm": 0.5511162281036377, "learning_rate": 8.258039946487828e-05, "loss": 0.0374, "step": 31040 }, { "grad_norm": 0.3903093636035919, "learning_rate": 8.256785519105241e-05, "loss": 0.0351, "step": 31050 }, { "grad_norm": 0.4576256573200226, "learning_rate": 8.255530735565488e-05, "loss": 0.0418, "step": 31060 }, { "grad_norm": 0.30532997846603394, "learning_rate": 8.254275596005787e-05, "loss": 0.0405, "step": 31070 }, { "grad_norm": 0.45256945490837097, "learning_rate": 8.2530201005634e-05, "loss": 0.0415, "step": 31080 }, { "grad_norm": 0.4052352011203766, "learning_rate": 8.251764249375626e-05, "loss": 0.0441, "step": 31090 }, { "grad_norm": 0.3870474100112915, "learning_rate": 8.250508042579803e-05, "loss": 0.0464, "step": 31100 }, { "grad_norm": 0.5216339826583862, "learning_rate": 8.249251480313307e-05, "loss": 0.0414, "step": 31110 }, { "grad_norm": 0.4365648031234741, "learning_rate": 8.247994562713555e-05, "loss": 0.0405, "step": 31120 }, { "grad_norm": 0.41709768772125244, "learning_rate": 8.246737289917998e-05, "loss": 0.0395, "step": 31130 }, { "grad_norm": 0.3597673177719116, "learning_rate": 8.245479662064135e-05, "loss": 0.0385, "step": 31140 }, { "grad_norm": 0.8986209630966187, "learning_rate": 8.244221679289496e-05, "loss": 0.0421, "step": 31150 }, { "grad_norm": 0.4432280957698822, "learning_rate": 8.242963341731652e-05, "loss": 0.0381, "step": 31160 }, { "grad_norm": 0.48140084743499756, "learning_rate": 8.24170464952821e-05, "loss": 0.0404, "step": 31170 }, { "grad_norm": 0.3030170202255249, "learning_rate": 8.240445602816824e-05, "loss": 0.0464, "step": 31180 }, { "grad_norm": 0.4379161298274994, "learning_rate": 8.239186201735179e-05, "loss": 0.0326, "step": 31190 }, { "grad_norm": 0.3603084683418274, "learning_rate": 8.237926446420998e-05, "loss": 0.0337, "step": 31200 }, { "grad_norm": 0.4601813852787018, "learning_rate": 8.236666337012052e-05, "loss": 0.0374, "step": 31210 }, { "grad_norm": 0.8397666215896606, "learning_rate": 8.23540587364614e-05, "loss": 0.0452, "step": 31220 }, { "grad_norm": 0.5458506345748901, "learning_rate": 8.234145056461107e-05, "loss": 0.0413, "step": 31230 }, { "grad_norm": 0.4704558253288269, "learning_rate": 8.232883885594832e-05, "loss": 0.0522, "step": 31240 }, { "grad_norm": 0.3814091682434082, "learning_rate": 8.231622361185236e-05, "loss": 0.0434, "step": 31250 }, { "grad_norm": 0.5960466861724854, "learning_rate": 8.230360483370278e-05, "loss": 0.0715, "step": 31260 }, { "grad_norm": 0.46820077300071716, "learning_rate": 8.229098252287953e-05, "loss": 0.0434, "step": 31270 }, { "grad_norm": 0.5924047231674194, "learning_rate": 8.2278356680763e-05, "loss": 0.0544, "step": 31280 }, { "grad_norm": 0.43478527665138245, "learning_rate": 8.22657273087339e-05, "loss": 0.0479, "step": 31290 }, { "grad_norm": 0.37755805253982544, "learning_rate": 8.225309440817336e-05, "loss": 0.0501, "step": 31300 }, { "grad_norm": 0.45138460397720337, "learning_rate": 8.224045798046293e-05, "loss": 0.0697, "step": 31310 }, { "grad_norm": 0.46962451934814453, "learning_rate": 8.22278180269845e-05, "loss": 0.0378, "step": 31320 }, { "grad_norm": 0.32819753885269165, "learning_rate": 8.221517454912031e-05, "loss": 0.0327, "step": 31330 }, { "grad_norm": 0.4143783152103424, "learning_rate": 8.220252754825308e-05, "loss": 0.0414, "step": 31340 }, { "grad_norm": 0.5345963835716248, "learning_rate": 8.218987702576586e-05, "loss": 0.0494, "step": 31350 }, { "grad_norm": 0.5332557559013367, "learning_rate": 8.217722298304207e-05, "loss": 0.0626, "step": 31360 }, { "grad_norm": 0.7342982292175293, "learning_rate": 8.216456542146557e-05, "loss": 0.0586, "step": 31370 }, { "grad_norm": 0.39186665415763855, "learning_rate": 8.215190434242055e-05, "loss": 0.0412, "step": 31380 }, { "grad_norm": 0.30297908186912537, "learning_rate": 8.213923974729161e-05, "loss": 0.041, "step": 31390 }, { "grad_norm": 0.4774925112724304, "learning_rate": 8.212657163746373e-05, "loss": 0.0519, "step": 31400 }, { "grad_norm": 0.36146625876426697, "learning_rate": 8.211390001432227e-05, "loss": 0.0377, "step": 31410 }, { "grad_norm": 0.3260194957256317, "learning_rate": 8.210122487925297e-05, "loss": 0.043, "step": 31420 }, { "grad_norm": 0.3827454745769501, "learning_rate": 8.208854623364202e-05, "loss": 0.0441, "step": 31430 }, { "grad_norm": 0.5808562636375427, "learning_rate": 8.207586407887589e-05, "loss": 0.0441, "step": 31440 }, { "grad_norm": 0.4536915123462677, "learning_rate": 8.206317841634148e-05, "loss": 0.0366, "step": 31450 }, { "grad_norm": 0.579215407371521, "learning_rate": 8.205048924742609e-05, "loss": 0.0437, "step": 31460 }, { "grad_norm": 0.3124238848686218, "learning_rate": 8.203779657351738e-05, "loss": 0.0408, "step": 31470 }, { "grad_norm": 0.3535503149032593, "learning_rate": 8.20251003960034e-05, "loss": 0.0432, "step": 31480 }, { "grad_norm": 0.44698354601860046, "learning_rate": 8.201240071627258e-05, "loss": 0.0458, "step": 31490 }, { "grad_norm": 0.3497445583343506, "learning_rate": 8.199969753571377e-05, "loss": 0.0457, "step": 31500 }, { "grad_norm": 0.32892316579818726, "learning_rate": 8.198699085571615e-05, "loss": 0.044, "step": 31510 }, { "grad_norm": 0.4693726897239685, "learning_rate": 8.197428067766928e-05, "loss": 0.0468, "step": 31520 }, { "grad_norm": 0.5692413449287415, "learning_rate": 8.196156700296316e-05, "loss": 0.0533, "step": 31530 }, { "grad_norm": 0.38564831018447876, "learning_rate": 8.194884983298814e-05, "loss": 0.0335, "step": 31540 }, { "grad_norm": 0.36894258856773376, "learning_rate": 8.193612916913491e-05, "loss": 0.0344, "step": 31550 }, { "grad_norm": 0.34820929169654846, "learning_rate": 8.192340501279463e-05, "loss": 0.038, "step": 31560 }, { "grad_norm": 0.7993165254592896, "learning_rate": 8.191067736535876e-05, "loss": 0.0496, "step": 31570 }, { "grad_norm": 0.4118540287017822, "learning_rate": 8.18979462282192e-05, "loss": 0.0461, "step": 31580 }, { "grad_norm": 0.4711077809333801, "learning_rate": 8.188521160276819e-05, "loss": 0.0434, "step": 31590 }, { "grad_norm": 0.4069857597351074, "learning_rate": 8.187247349039837e-05, "loss": 0.0408, "step": 31600 }, { "grad_norm": 0.47304102778434753, "learning_rate": 8.185973189250278e-05, "loss": 0.0522, "step": 31610 }, { "grad_norm": 0.34945082664489746, "learning_rate": 8.184698681047482e-05, "loss": 0.0403, "step": 31620 }, { "grad_norm": 0.41930603981018066, "learning_rate": 8.183423824570827e-05, "loss": 0.0401, "step": 31630 }, { "grad_norm": 0.4790724217891693, "learning_rate": 8.182148619959725e-05, "loss": 0.0445, "step": 31640 }, { "grad_norm": 0.37055760622024536, "learning_rate": 8.180873067353636e-05, "loss": 0.0404, "step": 31650 }, { "grad_norm": 0.517642080783844, "learning_rate": 8.179597166892052e-05, "loss": 0.0497, "step": 31660 }, { "grad_norm": 0.36972567439079285, "learning_rate": 8.178320918714501e-05, "loss": 0.0409, "step": 31670 }, { "grad_norm": 0.2668568193912506, "learning_rate": 8.177044322960554e-05, "loss": 0.0356, "step": 31680 }, { "grad_norm": 0.35934361815452576, "learning_rate": 8.175767379769816e-05, "loss": 0.0334, "step": 31690 }, { "grad_norm": 0.39386770129203796, "learning_rate": 8.174490089281932e-05, "loss": 0.0451, "step": 31700 }, { "grad_norm": 0.3635842502117157, "learning_rate": 8.173212451636584e-05, "loss": 0.049, "step": 31710 }, { "grad_norm": 0.5323770642280579, "learning_rate": 8.171934466973493e-05, "loss": 0.05, "step": 31720 }, { "grad_norm": 0.37478938698768616, "learning_rate": 8.170656135432418e-05, "loss": 0.0394, "step": 31730 }, { "grad_norm": 1.2834997177124023, "learning_rate": 8.169377457153155e-05, "loss": 0.0468, "step": 31740 }, { "grad_norm": 0.36040085554122925, "learning_rate": 8.168098432275539e-05, "loss": 0.0377, "step": 31750 }, { "grad_norm": 0.43759429454803467, "learning_rate": 8.166819060939442e-05, "loss": 0.0372, "step": 31760 }, { "grad_norm": 0.806529700756073, "learning_rate": 8.165539343284772e-05, "loss": 0.0505, "step": 31770 }, { "grad_norm": 0.4174671769142151, "learning_rate": 8.16425927945148e-05, "loss": 0.0455, "step": 31780 }, { "grad_norm": 0.4816261827945709, "learning_rate": 8.162978869579551e-05, "loss": 0.0606, "step": 31790 }, { "grad_norm": 0.37880411744117737, "learning_rate": 8.161698113809007e-05, "loss": 0.0456, "step": 31800 }, { "grad_norm": 0.3712918758392334, "learning_rate": 8.160417012279911e-05, "loss": 0.0675, "step": 31810 }, { "grad_norm": 0.3653475344181061, "learning_rate": 8.159135565132363e-05, "loss": 0.0439, "step": 31820 }, { "grad_norm": 0.2346002757549286, "learning_rate": 8.157853772506498e-05, "loss": 0.0411, "step": 31830 }, { "grad_norm": 0.5450911521911621, "learning_rate": 8.156571634542494e-05, "loss": 0.0579, "step": 31840 }, { "grad_norm": 0.5713192820549011, "learning_rate": 8.15528915138056e-05, "loss": 0.0464, "step": 31850 }, { "grad_norm": 0.35341358184814453, "learning_rate": 8.154006323160949e-05, "loss": 0.0515, "step": 31860 }, { "grad_norm": 0.4039926528930664, "learning_rate": 8.152723150023949e-05, "loss": 0.0539, "step": 31870 }, { "grad_norm": 0.5503909587860107, "learning_rate": 8.151439632109886e-05, "loss": 0.04, "step": 31880 }, { "grad_norm": 0.35590416193008423, "learning_rate": 8.150155769559122e-05, "loss": 0.052, "step": 31890 }, { "grad_norm": 0.34603017568588257, "learning_rate": 8.148871562512058e-05, "loss": 0.0402, "step": 31900 }, { "grad_norm": 0.3880552053451538, "learning_rate": 8.147587011109136e-05, "loss": 0.0494, "step": 31910 }, { "grad_norm": 0.40171918272972107, "learning_rate": 8.14630211549083e-05, "loss": 0.0555, "step": 31920 }, { "grad_norm": 0.4674794375896454, "learning_rate": 8.145016875797655e-05, "loss": 0.0353, "step": 31930 }, { "grad_norm": 0.3308548927307129, "learning_rate": 8.143731292170164e-05, "loss": 0.0381, "step": 31940 }, { "grad_norm": 0.40114477276802063, "learning_rate": 8.142445364748944e-05, "loss": 0.0478, "step": 31950 }, { "grad_norm": 0.41409051418304443, "learning_rate": 8.141159093674624e-05, "loss": 0.0451, "step": 31960 }, { "grad_norm": 0.3557128608226776, "learning_rate": 8.139872479087869e-05, "loss": 0.0309, "step": 31970 }, { "grad_norm": 0.3032536506652832, "learning_rate": 8.13858552112938e-05, "loss": 0.0307, "step": 31980 }, { "grad_norm": 0.4661358594894409, "learning_rate": 8.137298219939895e-05, "loss": 0.0484, "step": 31990 }, { "grad_norm": 0.3829984962940216, "learning_rate": 8.136010575660196e-05, "loss": 0.0364, "step": 32000 }, { "grad_norm": 0.41908779740333557, "learning_rate": 8.134722588431095e-05, "loss": 0.0363, "step": 32010 }, { "grad_norm": 0.39346152544021606, "learning_rate": 8.133434258393444e-05, "loss": 0.0421, "step": 32020 }, { "grad_norm": 0.7246062755584717, "learning_rate": 8.132145585688134e-05, "loss": 0.0517, "step": 32030 }, { "grad_norm": 0.3760046362876892, "learning_rate": 8.130856570456093e-05, "loss": 0.0447, "step": 32040 }, { "grad_norm": 0.38283008337020874, "learning_rate": 8.129567212838283e-05, "loss": 0.0445, "step": 32050 }, { "grad_norm": 0.6501092910766602, "learning_rate": 8.128277512975708e-05, "loss": 0.0522, "step": 32060 }, { "grad_norm": 0.6153934001922607, "learning_rate": 8.126987471009408e-05, "loss": 0.0516, "step": 32070 }, { "grad_norm": 0.34838294982910156, "learning_rate": 8.125697087080459e-05, "loss": 0.0389, "step": 32080 }, { "grad_norm": 0.3778747320175171, "learning_rate": 8.124406361329976e-05, "loss": 0.0428, "step": 32090 }, { "grad_norm": 0.49835580587387085, "learning_rate": 8.123115293899111e-05, "loss": 0.0419, "step": 32100 }, { "grad_norm": 0.3649153411388397, "learning_rate": 8.121823884929055e-05, "loss": 0.0371, "step": 32110 }, { "grad_norm": 0.44210904836654663, "learning_rate": 8.12053213456103e-05, "loss": 0.0484, "step": 32120 }, { "grad_norm": 0.3706584572792053, "learning_rate": 8.119240042936303e-05, "loss": 0.0468, "step": 32130 }, { "grad_norm": 0.49680274724960327, "learning_rate": 8.117947610196175e-05, "loss": 0.048, "step": 32140 }, { "grad_norm": 0.38998568058013916, "learning_rate": 8.116654836481982e-05, "loss": 0.035, "step": 32150 }, { "grad_norm": 0.46989402174949646, "learning_rate": 8.115361721935106e-05, "loss": 0.0394, "step": 32160 }, { "grad_norm": 0.31960925459861755, "learning_rate": 8.114068266696953e-05, "loss": 0.0468, "step": 32170 }, { "grad_norm": 0.4040053188800812, "learning_rate": 8.112774470908978e-05, "loss": 0.0417, "step": 32180 }, { "grad_norm": 0.2960280776023865, "learning_rate": 8.111480334712665e-05, "loss": 0.0403, "step": 32190 }, { "grad_norm": 0.35080257058143616, "learning_rate": 8.110185858249542e-05, "loss": 0.0382, "step": 32200 }, { "grad_norm": 0.54052734375, "learning_rate": 8.108891041661168e-05, "loss": 0.0442, "step": 32210 }, { "grad_norm": 0.46341022849082947, "learning_rate": 8.107595885089146e-05, "loss": 0.0489, "step": 32220 }, { "grad_norm": 0.4912430942058563, "learning_rate": 8.106300388675108e-05, "loss": 0.0352, "step": 32230 }, { "grad_norm": 0.46670252084732056, "learning_rate": 8.105004552560731e-05, "loss": 0.0466, "step": 32240 }, { "grad_norm": 0.6391008496284485, "learning_rate": 8.103708376887724e-05, "loss": 0.0463, "step": 32250 }, { "grad_norm": 0.5164236426353455, "learning_rate": 8.102411861797836e-05, "loss": 0.0457, "step": 32260 }, { "grad_norm": 0.5813243985176086, "learning_rate": 8.101115007432851e-05, "loss": 0.0406, "step": 32270 }, { "grad_norm": 0.5011415481567383, "learning_rate": 8.09981781393459e-05, "loss": 0.0326, "step": 32280 }, { "grad_norm": 0.3516954779624939, "learning_rate": 8.098520281444915e-05, "loss": 0.0453, "step": 32290 }, { "grad_norm": 0.6184502243995667, "learning_rate": 8.09722241010572e-05, "loss": 0.0418, "step": 32300 }, { "grad_norm": 0.46987539529800415, "learning_rate": 8.095924200058939e-05, "loss": 0.0482, "step": 32310 }, { "grad_norm": 0.4308284521102905, "learning_rate": 8.094625651446541e-05, "loss": 0.0391, "step": 32320 }, { "grad_norm": 0.42869648337364197, "learning_rate": 8.093326764410536e-05, "loss": 0.0356, "step": 32330 }, { "grad_norm": 0.4297792911529541, "learning_rate": 8.092027539092966e-05, "loss": 0.0429, "step": 32340 }, { "grad_norm": 0.26379942893981934, "learning_rate": 8.090727975635913e-05, "loss": 0.0313, "step": 32350 }, { "grad_norm": 0.3549271821975708, "learning_rate": 8.089428074181497e-05, "loss": 0.0385, "step": 32360 }, { "grad_norm": 0.31790176033973694, "learning_rate": 8.088127834871871e-05, "loss": 0.0383, "step": 32370 }, { "grad_norm": 0.5614716410636902, "learning_rate": 8.086827257849226e-05, "loss": 0.0399, "step": 32380 }, { "grad_norm": 0.38754400610923767, "learning_rate": 8.085526343255795e-05, "loss": 0.0354, "step": 32390 }, { "grad_norm": 0.5572485327720642, "learning_rate": 8.084225091233842e-05, "loss": 0.0472, "step": 32400 }, { "grad_norm": 0.41157326102256775, "learning_rate": 8.082923501925668e-05, "loss": 0.0393, "step": 32410 }, { "grad_norm": 0.3667776882648468, "learning_rate": 8.081621575473617e-05, "loss": 0.0478, "step": 32420 }, { "grad_norm": 0.3781217038631439, "learning_rate": 8.080319312020064e-05, "loss": 0.0484, "step": 32430 }, { "grad_norm": 0.696851909160614, "learning_rate": 8.079016711707421e-05, "loss": 0.0446, "step": 32440 }, { "grad_norm": 0.42272570729255676, "learning_rate": 8.077713774678139e-05, "loss": 0.0468, "step": 32450 }, { "grad_norm": 0.44530820846557617, "learning_rate": 8.076410501074707e-05, "loss": 0.0364, "step": 32460 }, { "grad_norm": 0.49857935309410095, "learning_rate": 8.075106891039647e-05, "loss": 0.0475, "step": 32470 }, { "grad_norm": 0.5660277605056763, "learning_rate": 8.073802944715523e-05, "loss": 0.0486, "step": 32480 }, { "grad_norm": 1.264568567276001, "learning_rate": 8.072498662244929e-05, "loss": 0.0406, "step": 32490 }, { "grad_norm": 0.4647604525089264, "learning_rate": 8.0711940437705e-05, "loss": 0.0508, "step": 32500 }, { "grad_norm": 0.4409876763820648, "learning_rate": 8.06988908943491e-05, "loss": 0.0404, "step": 32510 }, { "grad_norm": 0.30973196029663086, "learning_rate": 8.068583799380863e-05, "loss": 0.0439, "step": 32520 }, { "grad_norm": 0.4284253716468811, "learning_rate": 8.067278173751104e-05, "loss": 0.0394, "step": 32530 }, { "grad_norm": 0.621860921382904, "learning_rate": 8.065972212688417e-05, "loss": 0.0544, "step": 32540 }, { "grad_norm": 0.39154690504074097, "learning_rate": 8.064665916335618e-05, "loss": 0.0496, "step": 32550 }, { "grad_norm": 0.46572691202163696, "learning_rate": 8.063359284835564e-05, "loss": 0.0503, "step": 32560 }, { "grad_norm": 0.4331851899623871, "learning_rate": 8.062052318331142e-05, "loss": 0.047, "step": 32570 }, { "grad_norm": 0.56661456823349, "learning_rate": 8.060745016965283e-05, "loss": 0.0448, "step": 32580 }, { "grad_norm": 0.39025846123695374, "learning_rate": 8.059437380880952e-05, "loss": 0.0419, "step": 32590 }, { "grad_norm": 0.6026953458786011, "learning_rate": 8.058129410221146e-05, "loss": 0.0453, "step": 32600 }, { "grad_norm": 0.5602832436561584, "learning_rate": 8.056821105128908e-05, "loss": 0.0429, "step": 32610 }, { "grad_norm": 0.2952268719673157, "learning_rate": 8.05551246574731e-05, "loss": 0.0384, "step": 32620 }, { "grad_norm": 0.6528308987617493, "learning_rate": 8.05420349221946e-05, "loss": 0.0463, "step": 32630 }, { "grad_norm": 0.5603519678115845, "learning_rate": 8.05289418468851e-05, "loss": 0.0452, "step": 32640 }, { "grad_norm": 0.47381430864334106, "learning_rate": 8.051584543297642e-05, "loss": 0.0361, "step": 32650 }, { "grad_norm": 0.5127105116844177, "learning_rate": 8.050274568190074e-05, "loss": 0.0381, "step": 32660 }, { "grad_norm": 0.32924455404281616, "learning_rate": 8.048964259509067e-05, "loss": 0.0405, "step": 32670 }, { "grad_norm": 0.4887799322605133, "learning_rate": 8.047653617397914e-05, "loss": 0.0382, "step": 32680 }, { "grad_norm": 0.46369075775146484, "learning_rate": 8.046342641999941e-05, "loss": 0.0476, "step": 32690 }, { "grad_norm": 0.6333648562431335, "learning_rate": 8.045031333458517e-05, "loss": 0.0457, "step": 32700 }, { "grad_norm": 0.5218387246131897, "learning_rate": 8.043719691917047e-05, "loss": 0.0459, "step": 32710 }, { "grad_norm": 0.4048614501953125, "learning_rate": 8.042407717518966e-05, "loss": 0.0454, "step": 32720 }, { "grad_norm": 0.38909146189689636, "learning_rate": 8.041095410407751e-05, "loss": 0.0402, "step": 32730 }, { "grad_norm": 0.3410646915435791, "learning_rate": 8.039782770726913e-05, "loss": 0.0457, "step": 32740 }, { "grad_norm": 0.5519552826881409, "learning_rate": 8.038469798620004e-05, "loss": 0.0457, "step": 32750 }, { "grad_norm": 0.3200048506259918, "learning_rate": 8.037156494230604e-05, "loss": 0.049, "step": 32760 }, { "grad_norm": 0.3150765299797058, "learning_rate": 8.035842857702338e-05, "loss": 0.0429, "step": 32770 }, { "grad_norm": 0.5009952783584595, "learning_rate": 8.03452888917886e-05, "loss": 0.0449, "step": 32780 }, { "grad_norm": 0.3281596601009369, "learning_rate": 8.033214588803866e-05, "loss": 0.0378, "step": 32790 }, { "grad_norm": 0.45704081654548645, "learning_rate": 8.031899956721083e-05, "loss": 0.0479, "step": 32800 }, { "grad_norm": 0.49058860540390015, "learning_rate": 8.030584993074282e-05, "loss": 0.0496, "step": 32810 }, { "grad_norm": 0.33394700288772583, "learning_rate": 8.02926969800726e-05, "loss": 0.0424, "step": 32820 }, { "grad_norm": 1.0121718645095825, "learning_rate": 8.027954071663859e-05, "loss": 0.0547, "step": 32830 }, { "grad_norm": 0.41785362362861633, "learning_rate": 8.026638114187954e-05, "loss": 0.0409, "step": 32840 }, { "grad_norm": 0.46830904483795166, "learning_rate": 8.025321825723456e-05, "loss": 0.0477, "step": 32850 }, { "grad_norm": 0.459152489900589, "learning_rate": 8.02400520641431e-05, "loss": 0.0499, "step": 32860 }, { "grad_norm": 0.48230406641960144, "learning_rate": 8.022688256404501e-05, "loss": 0.0383, "step": 32870 }, { "grad_norm": 0.31917107105255127, "learning_rate": 8.02137097583805e-05, "loss": 0.0401, "step": 32880 }, { "grad_norm": 0.4103264808654785, "learning_rate": 8.02005336485901e-05, "loss": 0.0482, "step": 32890 }, { "grad_norm": 0.6640306115150452, "learning_rate": 8.018735423611476e-05, "loss": 0.0462, "step": 32900 }, { "grad_norm": 0.26666298508644104, "learning_rate": 8.017417152239574e-05, "loss": 0.0416, "step": 32910 }, { "grad_norm": 0.3651650846004486, "learning_rate": 8.01609855088747e-05, "loss": 0.052, "step": 32920 }, { "grad_norm": 0.29691821336746216, "learning_rate": 8.014779619699362e-05, "loss": 0.0322, "step": 32930 }, { "grad_norm": 0.3760974109172821, "learning_rate": 8.013460358819489e-05, "loss": 0.0428, "step": 32940 }, { "grad_norm": 0.5065377950668335, "learning_rate": 8.01214076839212e-05, "loss": 0.0392, "step": 32950 }, { "grad_norm": 0.5546466708183289, "learning_rate": 8.010820848561565e-05, "loss": 0.0378, "step": 32960 }, { "grad_norm": 0.5259812474250793, "learning_rate": 8.009500599472171e-05, "loss": 0.0488, "step": 32970 }, { "grad_norm": 0.38216787576675415, "learning_rate": 8.008180021268314e-05, "loss": 0.0288, "step": 32980 }, { "grad_norm": 0.4968043267726898, "learning_rate": 8.006859114094414e-05, "loss": 0.0374, "step": 32990 }, { "grad_norm": 0.4069444239139557, "learning_rate": 8.005537878094921e-05, "loss": 0.0385, "step": 33000 }, { "grad_norm": 0.3457048833370209, "learning_rate": 8.004216313414323e-05, "loss": 0.0395, "step": 33010 }, { "grad_norm": 0.3722517788410187, "learning_rate": 8.002894420197149e-05, "loss": 0.0387, "step": 33020 }, { "grad_norm": 0.37614431977272034, "learning_rate": 8.001572198587954e-05, "loss": 0.0456, "step": 33030 }, { "grad_norm": 0.4821080267429352, "learning_rate": 8.000249648731338e-05, "loss": 0.0392, "step": 33040 }, { "grad_norm": 0.4391201138496399, "learning_rate": 7.998926770771928e-05, "loss": 0.0589, "step": 33050 }, { "grad_norm": 0.31572404503822327, "learning_rate": 7.997603564854397e-05, "loss": 0.041, "step": 33060 }, { "grad_norm": 0.336956262588501, "learning_rate": 7.996280031123448e-05, "loss": 0.0343, "step": 33070 }, { "grad_norm": 0.298452764749527, "learning_rate": 7.994956169723818e-05, "loss": 0.0578, "step": 33080 }, { "grad_norm": 0.35482916235923767, "learning_rate": 7.993631980800285e-05, "loss": 0.0357, "step": 33090 }, { "grad_norm": 0.36340513825416565, "learning_rate": 7.992307464497659e-05, "loss": 0.0444, "step": 33100 }, { "grad_norm": 0.3432120084762573, "learning_rate": 7.990982620960787e-05, "loss": 0.0464, "step": 33110 }, { "grad_norm": 0.31437239050865173, "learning_rate": 7.989657450334554e-05, "loss": 0.0388, "step": 33120 }, { "grad_norm": 0.36759066581726074, "learning_rate": 7.988331952763877e-05, "loss": 0.045, "step": 33130 }, { "grad_norm": 0.47481459379196167, "learning_rate": 7.987006128393709e-05, "loss": 0.0507, "step": 33140 }, { "grad_norm": 0.40636146068573, "learning_rate": 7.985679977369043e-05, "loss": 0.0315, "step": 33150 }, { "grad_norm": 0.4346277117729187, "learning_rate": 7.984353499834902e-05, "loss": 0.0332, "step": 33160 }, { "grad_norm": 0.38617032766342163, "learning_rate": 7.983026695936351e-05, "loss": 0.0388, "step": 33170 }, { "grad_norm": 0.4031032919883728, "learning_rate": 7.981699565818486e-05, "loss": 0.0364, "step": 33180 }, { "grad_norm": 0.5170980095863342, "learning_rate": 7.980372109626437e-05, "loss": 0.0368, "step": 33190 }, { "grad_norm": 0.7737836241722107, "learning_rate": 7.979044327505375e-05, "loss": 0.0448, "step": 33200 }, { "grad_norm": 0.3826553523540497, "learning_rate": 7.977716219600506e-05, "loss": 0.0369, "step": 33210 }, { "grad_norm": 0.40686529874801636, "learning_rate": 7.97638778605707e-05, "loss": 0.0431, "step": 33220 }, { "grad_norm": 0.6203293800354004, "learning_rate": 7.975059027020338e-05, "loss": 0.065, "step": 33230 }, { "grad_norm": 0.5377521514892578, "learning_rate": 7.973729942635623e-05, "loss": 0.0425, "step": 33240 }, { "grad_norm": 0.40439218282699585, "learning_rate": 7.972400533048273e-05, "loss": 0.0607, "step": 33250 }, { "grad_norm": 0.5421104431152344, "learning_rate": 7.97107079840367e-05, "loss": 0.0437, "step": 33260 }, { "grad_norm": 0.3782821297645569, "learning_rate": 7.969740738847231e-05, "loss": 0.041, "step": 33270 }, { "grad_norm": 0.30899250507354736, "learning_rate": 7.968410354524411e-05, "loss": 0.0414, "step": 33280 }, { "grad_norm": 0.3159862160682678, "learning_rate": 7.967079645580697e-05, "loss": 0.041, "step": 33290 }, { "grad_norm": 0.366258442401886, "learning_rate": 7.965748612161612e-05, "loss": 0.0361, "step": 33300 }, { "grad_norm": 0.43111681938171387, "learning_rate": 7.96441725441272e-05, "loss": 0.0394, "step": 33310 }, { "grad_norm": 0.4305168390274048, "learning_rate": 7.963085572479614e-05, "loss": 0.0477, "step": 33320 }, { "grad_norm": 0.45927414298057556, "learning_rate": 7.961753566507924e-05, "loss": 0.0428, "step": 33330 }, { "grad_norm": 0.4867604970932007, "learning_rate": 7.960421236643316e-05, "loss": 0.052, "step": 33340 }, { "grad_norm": 0.4401456415653229, "learning_rate": 7.959088583031496e-05, "loss": 0.0409, "step": 33350 }, { "grad_norm": 0.37851202487945557, "learning_rate": 7.957755605818194e-05, "loss": 0.0363, "step": 33360 }, { "grad_norm": 0.27023088932037354, "learning_rate": 7.956422305149185e-05, "loss": 0.0532, "step": 33370 }, { "grad_norm": 0.5431543588638306, "learning_rate": 7.95508868117028e-05, "loss": 0.0484, "step": 33380 }, { "grad_norm": 0.45873549580574036, "learning_rate": 7.953754734027318e-05, "loss": 0.0514, "step": 33390 }, { "grad_norm": 0.4324743151664734, "learning_rate": 7.952420463866182e-05, "loss": 0.0366, "step": 33400 }, { "grad_norm": 0.4808467626571655, "learning_rate": 7.951085870832782e-05, "loss": 0.041, "step": 33410 }, { "grad_norm": 0.23960857093334198, "learning_rate": 7.949750955073067e-05, "loss": 0.0423, "step": 33420 }, { "grad_norm": 0.29406341910362244, "learning_rate": 7.948415716733022e-05, "loss": 0.0384, "step": 33430 }, { "grad_norm": 0.3856276273727417, "learning_rate": 7.947080155958669e-05, "loss": 0.044, "step": 33440 }, { "grad_norm": 0.40677088499069214, "learning_rate": 7.94574427289606e-05, "loss": 0.0393, "step": 33450 }, { "grad_norm": 0.3570919632911682, "learning_rate": 7.944408067691284e-05, "loss": 0.0309, "step": 33460 }, { "grad_norm": 0.3751624524593353, "learning_rate": 7.943071540490473e-05, "loss": 0.0445, "step": 33470 }, { "grad_norm": 0.3333591818809509, "learning_rate": 7.94173469143978e-05, "loss": 0.0385, "step": 33480 }, { "grad_norm": 0.6279962658882141, "learning_rate": 7.940397520685406e-05, "loss": 0.0416, "step": 33490 }, { "grad_norm": 0.5734893083572388, "learning_rate": 7.939060028373577e-05, "loss": 0.0565, "step": 33500 }, { "grad_norm": 0.4534696936607361, "learning_rate": 7.937722214650565e-05, "loss": 0.061, "step": 33510 }, { "grad_norm": 0.3157466650009155, "learning_rate": 7.936384079662666e-05, "loss": 0.0464, "step": 33520 }, { "grad_norm": 0.4939025640487671, "learning_rate": 7.93504562355622e-05, "loss": 0.0496, "step": 33530 }, { "grad_norm": 0.3866112232208252, "learning_rate": 7.933706846477599e-05, "loss": 0.0476, "step": 33540 }, { "grad_norm": 0.37813127040863037, "learning_rate": 7.932367748573206e-05, "loss": 0.041, "step": 33550 }, { "grad_norm": 0.38000988960266113, "learning_rate": 7.931028329989485e-05, "loss": 0.0384, "step": 33560 }, { "grad_norm": 1.0336012840270996, "learning_rate": 7.929688590872913e-05, "loss": 0.048, "step": 33570 }, { "grad_norm": 0.3686300814151764, "learning_rate": 7.928348531370003e-05, "loss": 0.0433, "step": 33580 }, { "grad_norm": 0.4506359398365021, "learning_rate": 7.927008151627297e-05, "loss": 0.0507, "step": 33590 }, { "grad_norm": 0.29009032249450684, "learning_rate": 7.925667451791383e-05, "loss": 0.0498, "step": 33600 }, { "grad_norm": 0.5042358040809631, "learning_rate": 7.924326432008874e-05, "loss": 0.0377, "step": 33610 }, { "grad_norm": 0.3859573006629944, "learning_rate": 7.922985092426422e-05, "loss": 0.045, "step": 33620 }, { "grad_norm": 0.3491133153438568, "learning_rate": 7.921643433190717e-05, "loss": 0.046, "step": 33630 }, { "grad_norm": 0.5922810435295105, "learning_rate": 7.920301454448478e-05, "loss": 0.0393, "step": 33640 }, { "grad_norm": 0.3144689202308655, "learning_rate": 7.918959156346461e-05, "loss": 0.0387, "step": 33650 }, { "grad_norm": 0.3476676642894745, "learning_rate": 7.91761653903146e-05, "loss": 0.0389, "step": 33660 }, { "grad_norm": 0.5393819212913513, "learning_rate": 7.916273602650302e-05, "loss": 0.0447, "step": 33670 }, { "grad_norm": 0.48732686042785645, "learning_rate": 7.914930347349847e-05, "loss": 0.038, "step": 33680 }, { "grad_norm": 0.3380139470100403, "learning_rate": 7.913586773276992e-05, "loss": 0.0376, "step": 33690 }, { "grad_norm": 0.5340874791145325, "learning_rate": 7.912242880578667e-05, "loss": 0.0331, "step": 33700 }, { "grad_norm": 0.4055168330669403, "learning_rate": 7.910898669401839e-05, "loss": 0.0461, "step": 33710 }, { "grad_norm": 0.37847501039505005, "learning_rate": 7.909554139893511e-05, "loss": 0.035, "step": 33720 }, { "grad_norm": 0.2896818518638611, "learning_rate": 7.908209292200715e-05, "loss": 0.0376, "step": 33730 }, { "grad_norm": 0.3258105218410492, "learning_rate": 7.906864126470523e-05, "loss": 0.0347, "step": 33740 }, { "grad_norm": 0.40129995346069336, "learning_rate": 7.905518642850041e-05, "loss": 0.0372, "step": 33750 }, { "grad_norm": 0.4160616993904114, "learning_rate": 7.904172841486409e-05, "loss": 0.0469, "step": 33760 }, { "grad_norm": 0.7731725573539734, "learning_rate": 7.902826722526801e-05, "loss": 0.0474, "step": 33770 }, { "grad_norm": 0.4043234884738922, "learning_rate": 7.901480286118427e-05, "loss": 0.0409, "step": 33780 }, { "grad_norm": 0.5309062004089355, "learning_rate": 7.900133532408531e-05, "loss": 0.0406, "step": 33790 }, { "grad_norm": 0.51707923412323, "learning_rate": 7.898786461544395e-05, "loss": 0.0455, "step": 33800 }, { "grad_norm": 0.4765395522117615, "learning_rate": 7.897439073673325e-05, "loss": 0.0442, "step": 33810 }, { "grad_norm": 0.3048049509525299, "learning_rate": 7.896091368942677e-05, "loss": 0.0385, "step": 33820 }, { "grad_norm": 0.37721338868141174, "learning_rate": 7.894743347499832e-05, "loss": 0.0411, "step": 33830 }, { "grad_norm": 0.5476840734481812, "learning_rate": 7.893395009492203e-05, "loss": 0.037, "step": 33840 }, { "grad_norm": 0.5144762992858887, "learning_rate": 7.892046355067248e-05, "loss": 0.0353, "step": 33850 }, { "grad_norm": 0.4246319830417633, "learning_rate": 7.890697384372451e-05, "loss": 0.0407, "step": 33860 }, { "grad_norm": 0.4177079200744629, "learning_rate": 7.889348097555336e-05, "loss": 0.0336, "step": 33870 }, { "grad_norm": 0.21729286015033722, "learning_rate": 7.887998494763455e-05, "loss": 0.0406, "step": 33880 }, { "grad_norm": 0.2949831485748291, "learning_rate": 7.886648576144404e-05, "loss": 0.0303, "step": 33890 }, { "grad_norm": 0.32690033316612244, "learning_rate": 7.885298341845802e-05, "loss": 0.029, "step": 33900 }, { "grad_norm": 0.4543393850326538, "learning_rate": 7.883947792015311e-05, "loss": 0.0452, "step": 33910 }, { "grad_norm": 0.6036893129348755, "learning_rate": 7.882596926800628e-05, "loss": 0.0612, "step": 33920 }, { "grad_norm": 0.3931399881839752, "learning_rate": 7.881245746349477e-05, "loss": 0.0415, "step": 33930 }, { "grad_norm": 0.4963158369064331, "learning_rate": 7.879894250809623e-05, "loss": 0.0423, "step": 33940 }, { "grad_norm": 0.6427702903747559, "learning_rate": 7.878542440328865e-05, "loss": 0.062, "step": 33950 }, { "grad_norm": 0.4937250316143036, "learning_rate": 7.877190315055031e-05, "loss": 0.0395, "step": 33960 }, { "grad_norm": 0.5584900379180908, "learning_rate": 7.875837875135991e-05, "loss": 0.0391, "step": 33970 }, { "grad_norm": 0.48951002955436707, "learning_rate": 7.874485120719646e-05, "loss": 0.0432, "step": 33980 }, { "grad_norm": 0.36868828535079956, "learning_rate": 7.873132051953928e-05, "loss": 0.0321, "step": 33990 }, { "grad_norm": 0.44415083527565, "learning_rate": 7.87177866898681e-05, "loss": 0.0486, "step": 34000 }, { "grad_norm": 0.4070533215999603, "learning_rate": 7.870424971966294e-05, "loss": 0.0352, "step": 34010 }, { "grad_norm": 0.20371191203594208, "learning_rate": 7.869070961040419e-05, "loss": 0.0396, "step": 34020 }, { "grad_norm": 0.5490823984146118, "learning_rate": 7.867716636357257e-05, "loss": 0.0421, "step": 34030 }, { "grad_norm": 0.5355221033096313, "learning_rate": 7.866361998064915e-05, "loss": 0.043, "step": 34040 }, { "grad_norm": 0.47921279072761536, "learning_rate": 7.865007046311534e-05, "loss": 0.0363, "step": 34050 }, { "grad_norm": 0.33936378359794617, "learning_rate": 7.86365178124529e-05, "loss": 0.039, "step": 34060 }, { "grad_norm": 0.4155634641647339, "learning_rate": 7.862296203014394e-05, "loss": 0.04, "step": 34070 }, { "grad_norm": 0.37026092410087585, "learning_rate": 7.860940311767088e-05, "loss": 0.0394, "step": 34080 }, { "grad_norm": 0.31801751255989075, "learning_rate": 7.85958410765165e-05, "loss": 0.0394, "step": 34090 }, { "grad_norm": 0.29819563031196594, "learning_rate": 7.858227590816394e-05, "loss": 0.0352, "step": 34100 }, { "grad_norm": 0.48842209577560425, "learning_rate": 7.856870761409664e-05, "loss": 0.0431, "step": 34110 }, { "grad_norm": 0.34938326478004456, "learning_rate": 7.855513619579846e-05, "loss": 0.037, "step": 34120 }, { "grad_norm": 0.4348839819431305, "learning_rate": 7.85415616547535e-05, "loss": 0.0422, "step": 34130 }, { "grad_norm": 0.4201577603816986, "learning_rate": 7.852798399244627e-05, "loss": 0.0448, "step": 34140 }, { "grad_norm": 0.36023375391960144, "learning_rate": 7.851440321036161e-05, "loss": 0.0317, "step": 34150 }, { "grad_norm": 0.3291534185409546, "learning_rate": 7.850081930998468e-05, "loss": 0.0448, "step": 34160 }, { "grad_norm": 0.28233787417411804, "learning_rate": 7.8487232292801e-05, "loss": 0.046, "step": 34170 }, { "grad_norm": 0.254090815782547, "learning_rate": 7.847364216029642e-05, "loss": 0.0507, "step": 34180 }, { "grad_norm": 0.4645380675792694, "learning_rate": 7.846004891395716e-05, "loss": 0.0431, "step": 34190 }, { "grad_norm": 0.27744850516319275, "learning_rate": 7.844645255526972e-05, "loss": 0.0389, "step": 34200 }, { "grad_norm": 0.3701971173286438, "learning_rate": 7.843285308572101e-05, "loss": 0.0395, "step": 34210 }, { "grad_norm": 0.35538116097450256, "learning_rate": 7.841925050679823e-05, "loss": 0.0403, "step": 34220 }, { "grad_norm": 0.3582930266857147, "learning_rate": 7.840564481998895e-05, "loss": 0.0432, "step": 34230 }, { "grad_norm": 0.8210463523864746, "learning_rate": 7.839203602678105e-05, "loss": 0.0435, "step": 34240 }, { "grad_norm": 0.4454965591430664, "learning_rate": 7.837842412866279e-05, "loss": 0.0574, "step": 34250 }, { "grad_norm": 0.6404107809066772, "learning_rate": 7.836480912712272e-05, "loss": 0.0522, "step": 34260 }, { "grad_norm": 0.6203315854072571, "learning_rate": 7.835119102364976e-05, "loss": 0.0405, "step": 34270 }, { "grad_norm": 0.44075867533683777, "learning_rate": 7.83375698197332e-05, "loss": 0.0477, "step": 34280 }, { "grad_norm": 0.41822996735572815, "learning_rate": 7.83239455168626e-05, "loss": 0.0445, "step": 34290 }, { "grad_norm": 0.4179961383342743, "learning_rate": 7.83103181165279e-05, "loss": 0.0393, "step": 34300 }, { "grad_norm": 0.4904674291610718, "learning_rate": 7.829668762021937e-05, "loss": 0.0376, "step": 34310 }, { "grad_norm": 0.4865662753582001, "learning_rate": 7.828305402942764e-05, "loss": 0.0388, "step": 34320 }, { "grad_norm": 0.43114155530929565, "learning_rate": 7.826941734564363e-05, "loss": 0.0349, "step": 34330 }, { "grad_norm": 0.34355631470680237, "learning_rate": 7.825577757035865e-05, "loss": 0.0317, "step": 34340 }, { "grad_norm": 0.3787676990032196, "learning_rate": 7.824213470506431e-05, "loss": 0.0436, "step": 34350 }, { "grad_norm": 0.41730988025665283, "learning_rate": 7.822848875125257e-05, "loss": 0.0473, "step": 34360 }, { "grad_norm": 0.3361479938030243, "learning_rate": 7.821483971041576e-05, "loss": 0.0406, "step": 34370 }, { "grad_norm": 0.9717593193054199, "learning_rate": 7.820118758404649e-05, "loss": 0.0418, "step": 34380 }, { "grad_norm": 0.3387829661369324, "learning_rate": 7.818753237363776e-05, "loss": 0.0371, "step": 34390 }, { "grad_norm": 0.40299656987190247, "learning_rate": 7.817387408068286e-05, "loss": 0.0441, "step": 34400 }, { "grad_norm": 0.5179889798164368, "learning_rate": 7.816021270667544e-05, "loss": 0.0415, "step": 34410 }, { "grad_norm": 0.857090175151825, "learning_rate": 7.81465482531095e-05, "loss": 0.0491, "step": 34420 }, { "grad_norm": 0.7253618240356445, "learning_rate": 7.813288072147938e-05, "loss": 0.0341, "step": 34430 }, { "grad_norm": 0.3865216374397278, "learning_rate": 7.811921011327972e-05, "loss": 0.0416, "step": 34440 }, { "grad_norm": 0.5023815035820007, "learning_rate": 7.810553643000549e-05, "loss": 0.0505, "step": 34450 }, { "grad_norm": 0.45435264706611633, "learning_rate": 7.809185967315206e-05, "loss": 0.0441, "step": 34460 }, { "grad_norm": 0.46994200348854065, "learning_rate": 7.80781798442151e-05, "loss": 0.0471, "step": 34470 }, { "grad_norm": 0.3073720633983612, "learning_rate": 7.806449694469061e-05, "loss": 0.0346, "step": 34480 }, { "grad_norm": 0.33085694909095764, "learning_rate": 7.805081097607492e-05, "loss": 0.0377, "step": 34490 }, { "grad_norm": 0.43567654490470886, "learning_rate": 7.803712193986474e-05, "loss": 0.0395, "step": 34500 }, { "grad_norm": 0.7183294892311096, "learning_rate": 7.802342983755702e-05, "loss": 0.0401, "step": 34510 }, { "grad_norm": 0.5286707878112793, "learning_rate": 7.800973467064918e-05, "loss": 0.0443, "step": 34520 }, { "grad_norm": 0.5287226438522339, "learning_rate": 7.799603644063884e-05, "loss": 0.0528, "step": 34530 }, { "grad_norm": 0.412658154964447, "learning_rate": 7.798233514902405e-05, "loss": 0.0472, "step": 34540 }, { "grad_norm": 0.38509973883628845, "learning_rate": 7.796863079730318e-05, "loss": 0.0473, "step": 34550 }, { "grad_norm": 0.31268519163131714, "learning_rate": 7.795492338697488e-05, "loss": 0.0422, "step": 34560 }, { "grad_norm": 0.37345048785209656, "learning_rate": 7.794121291953819e-05, "loss": 0.0466, "step": 34570 }, { "grad_norm": 1.4257304668426514, "learning_rate": 7.792749939649246e-05, "loss": 0.041, "step": 34580 }, { "grad_norm": 0.3985992968082428, "learning_rate": 7.79137828193374e-05, "loss": 0.0362, "step": 34590 }, { "grad_norm": 0.412584513425827, "learning_rate": 7.790006318957301e-05, "loss": 0.0492, "step": 34600 }, { "grad_norm": 0.3933919072151184, "learning_rate": 7.788634050869965e-05, "loss": 0.0425, "step": 34610 }, { "grad_norm": 0.35215142369270325, "learning_rate": 7.787261477821803e-05, "loss": 0.0398, "step": 34620 }, { "grad_norm": 0.6409521102905273, "learning_rate": 7.785888599962916e-05, "loss": 0.0423, "step": 34630 }, { "grad_norm": 0.41753560304641724, "learning_rate": 7.784515417443439e-05, "loss": 0.0399, "step": 34640 }, { "grad_norm": 0.3683805465698242, "learning_rate": 7.783141930413545e-05, "loss": 0.0415, "step": 34650 }, { "grad_norm": 0.5021305680274963, "learning_rate": 7.78176813902343e-05, "loss": 0.0441, "step": 34660 }, { "grad_norm": 0.3633524477481842, "learning_rate": 7.780394043423336e-05, "loss": 0.0404, "step": 34670 }, { "grad_norm": 0.4151079058647156, "learning_rate": 7.77901964376353e-05, "loss": 0.0484, "step": 34680 }, { "grad_norm": 0.338004469871521, "learning_rate": 7.777644940194316e-05, "loss": 0.0475, "step": 34690 }, { "grad_norm": 0.49561095237731934, "learning_rate": 7.776269932866023e-05, "loss": 0.0415, "step": 34700 }, { "grad_norm": 0.8389529585838318, "learning_rate": 7.774894621929026e-05, "loss": 0.0475, "step": 34710 }, { "grad_norm": 0.3863428831100464, "learning_rate": 7.773519007533725e-05, "loss": 0.0487, "step": 34720 }, { "grad_norm": 0.3590887784957886, "learning_rate": 7.772143089830556e-05, "loss": 0.0488, "step": 34730 }, { "grad_norm": 0.45702576637268066, "learning_rate": 7.770766868969985e-05, "loss": 0.0451, "step": 34740 }, { "grad_norm": 0.5642984509468079, "learning_rate": 7.769390345102518e-05, "loss": 0.0448, "step": 34750 }, { "grad_norm": 0.5011569857597351, "learning_rate": 7.768013518378683e-05, "loss": 0.0549, "step": 34760 }, { "grad_norm": 0.38036248087882996, "learning_rate": 7.766636388949053e-05, "loss": 0.0335, "step": 34770 }, { "grad_norm": 0.6788637042045593, "learning_rate": 7.765258956964229e-05, "loss": 0.0514, "step": 34780 }, { "grad_norm": 0.452029824256897, "learning_rate": 7.76388122257484e-05, "loss": 0.0416, "step": 34790 }, { "grad_norm": 0.45522341132164, "learning_rate": 7.762503185931558e-05, "loss": 0.0308, "step": 34800 }, { "grad_norm": 0.38751551508903503, "learning_rate": 7.76112484718508e-05, "loss": 0.0303, "step": 34810 }, { "grad_norm": 0.6823364496231079, "learning_rate": 7.75974620648614e-05, "loss": 0.0447, "step": 34820 }, { "grad_norm": 0.30200451612472534, "learning_rate": 7.758367263985503e-05, "loss": 0.035, "step": 34830 }, { "grad_norm": 0.5523493885993958, "learning_rate": 7.75698801983397e-05, "loss": 0.0341, "step": 34840 }, { "grad_norm": 0.4578806757926941, "learning_rate": 7.755608474182372e-05, "loss": 0.0422, "step": 34850 }, { "grad_norm": 0.4951346814632416, "learning_rate": 7.754228627181574e-05, "loss": 0.0371, "step": 34860 }, { "grad_norm": 0.6489666700363159, "learning_rate": 7.752848478982476e-05, "loss": 0.0387, "step": 34870 }, { "grad_norm": 0.5099566578865051, "learning_rate": 7.751468029736006e-05, "loss": 0.0412, "step": 34880 }, { "grad_norm": 0.30962464213371277, "learning_rate": 7.750087279593129e-05, "loss": 0.0476, "step": 34890 }, { "grad_norm": 0.2951168119907379, "learning_rate": 7.748706228704843e-05, "loss": 0.0341, "step": 34900 }, { "grad_norm": 0.350488543510437, "learning_rate": 7.747324877222176e-05, "loss": 0.0456, "step": 34910 }, { "grad_norm": 0.29233965277671814, "learning_rate": 7.745943225296188e-05, "loss": 0.04, "step": 34920 }, { "grad_norm": 0.3369850814342499, "learning_rate": 7.744561273077981e-05, "loss": 0.0395, "step": 34930 }, { "grad_norm": 0.29728081822395325, "learning_rate": 7.743179020718678e-05, "loss": 0.043, "step": 34940 }, { "grad_norm": 0.4247451722621918, "learning_rate": 7.741796468369443e-05, "loss": 0.0444, "step": 34950 }, { "grad_norm": 0.4778628647327423, "learning_rate": 7.740413616181466e-05, "loss": 0.04, "step": 34960 }, { "grad_norm": 0.400948166847229, "learning_rate": 7.739030464305978e-05, "loss": 0.0396, "step": 34970 }, { "grad_norm": 0.35540950298309326, "learning_rate": 7.737647012894235e-05, "loss": 0.0423, "step": 34980 }, { "grad_norm": 0.37515950202941895, "learning_rate": 7.736263262097532e-05, "loss": 0.0356, "step": 34990 }, { "grad_norm": 0.3118087947368622, "learning_rate": 7.734879212067192e-05, "loss": 0.0397, "step": 35000 }, { "grad_norm": 0.3048710823059082, "learning_rate": 7.733494862954573e-05, "loss": 0.0403, "step": 35010 }, { "grad_norm": 0.36715713143348694, "learning_rate": 7.732110214911066e-05, "loss": 0.0374, "step": 35020 }, { "grad_norm": 0.5565719604492188, "learning_rate": 7.730725268088094e-05, "loss": 0.0438, "step": 35030 }, { "grad_norm": 0.4197688400745392, "learning_rate": 7.729340022637112e-05, "loss": 0.043, "step": 35040 }, { "grad_norm": 0.4303574860095978, "learning_rate": 7.727954478709607e-05, "loss": 0.0323, "step": 35050 }, { "grad_norm": 0.3073008358478546, "learning_rate": 7.726568636457103e-05, "loss": 0.0365, "step": 35060 }, { "grad_norm": 0.5411195755004883, "learning_rate": 7.725182496031153e-05, "loss": 0.0332, "step": 35070 }, { "grad_norm": 0.30246198177337646, "learning_rate": 7.723796057583341e-05, "loss": 0.0435, "step": 35080 }, { "grad_norm": 0.3179005980491638, "learning_rate": 7.722409321265287e-05, "loss": 0.0451, "step": 35090 }, { "grad_norm": 0.39101600646972656, "learning_rate": 7.721022287228645e-05, "loss": 0.0517, "step": 35100 }, { "grad_norm": 0.4220447838306427, "learning_rate": 7.719634955625094e-05, "loss": 0.0473, "step": 35110 }, { "grad_norm": 0.38562822341918945, "learning_rate": 7.718247326606355e-05, "loss": 0.0323, "step": 35120 }, { "grad_norm": 0.36997589468955994, "learning_rate": 7.716859400324175e-05, "loss": 0.0332, "step": 35130 }, { "grad_norm": 0.2952788472175598, "learning_rate": 7.715471176930335e-05, "loss": 0.051, "step": 35140 }, { "grad_norm": 0.3924355208873749, "learning_rate": 7.714082656576651e-05, "loss": 0.0428, "step": 35150 }, { "grad_norm": 0.5133775472640991, "learning_rate": 7.712693839414968e-05, "loss": 0.0383, "step": 35160 }, { "grad_norm": 0.29502272605895996, "learning_rate": 7.711304725597164e-05, "loss": 0.0414, "step": 35170 }, { "grad_norm": 0.32445862889289856, "learning_rate": 7.709915315275151e-05, "loss": 0.0402, "step": 35180 }, { "grad_norm": 0.4603899419307709, "learning_rate": 7.708525608600876e-05, "loss": 0.0472, "step": 35190 }, { "grad_norm": 0.323322057723999, "learning_rate": 7.707135605726311e-05, "loss": 0.0411, "step": 35200 }, { "grad_norm": 0.38149309158325195, "learning_rate": 7.705745306803466e-05, "loss": 0.0566, "step": 35210 }, { "grad_norm": 0.36447566747665405, "learning_rate": 7.704354711984383e-05, "loss": 0.0415, "step": 35220 }, { "grad_norm": 0.3641757667064667, "learning_rate": 7.702963821421133e-05, "loss": 0.0372, "step": 35230 }, { "grad_norm": 0.32560697197914124, "learning_rate": 7.701572635265826e-05, "loss": 0.0404, "step": 35240 }, { "grad_norm": 0.3821530044078827, "learning_rate": 7.700181153670596e-05, "loss": 0.036, "step": 35250 }, { "grad_norm": 0.32932931184768677, "learning_rate": 7.698789376787616e-05, "loss": 0.0408, "step": 35260 }, { "grad_norm": 0.4805792272090912, "learning_rate": 7.697397304769085e-05, "loss": 0.0471, "step": 35270 }, { "grad_norm": 0.546843945980072, "learning_rate": 7.696004937767241e-05, "loss": 0.0505, "step": 35280 }, { "grad_norm": 0.6037040948867798, "learning_rate": 7.694612275934352e-05, "loss": 0.0371, "step": 35290 }, { "grad_norm": 0.4042285680770874, "learning_rate": 7.693219319422714e-05, "loss": 0.0414, "step": 35300 }, { "grad_norm": 0.5010522603988647, "learning_rate": 7.69182606838466e-05, "loss": 0.035, "step": 35310 }, { "grad_norm": 0.3584740459918976, "learning_rate": 7.690432522972558e-05, "loss": 0.0324, "step": 35320 }, { "grad_norm": 1.6052607297897339, "learning_rate": 7.689038683338796e-05, "loss": 0.0458, "step": 35330 }, { "grad_norm": 0.4434266984462738, "learning_rate": 7.687644549635808e-05, "loss": 0.0323, "step": 35340 }, { "grad_norm": 0.44863665103912354, "learning_rate": 7.686250122016053e-05, "loss": 0.0468, "step": 35350 }, { "grad_norm": 0.8649276494979858, "learning_rate": 7.684855400632022e-05, "loss": 0.045, "step": 35360 }, { "grad_norm": 0.5954293012619019, "learning_rate": 7.683460385636243e-05, "loss": 0.0413, "step": 35370 }, { "grad_norm": 0.5667187571525574, "learning_rate": 7.68206507718127e-05, "loss": 0.0473, "step": 35380 }, { "grad_norm": 1.4226223230361938, "learning_rate": 7.680669475419693e-05, "loss": 0.0582, "step": 35390 }, { "grad_norm": 0.34856995940208435, "learning_rate": 7.679273580504132e-05, "loss": 0.04, "step": 35400 }, { "grad_norm": 0.4290619194507599, "learning_rate": 7.67787739258724e-05, "loss": 0.0383, "step": 35410 }, { "grad_norm": 0.3104188144207001, "learning_rate": 7.676480911821705e-05, "loss": 0.0433, "step": 35420 }, { "grad_norm": 0.5403212904930115, "learning_rate": 7.675084138360238e-05, "loss": 0.0473, "step": 35430 }, { "grad_norm": 0.325070858001709, "learning_rate": 7.673687072355592e-05, "loss": 0.0379, "step": 35440 }, { "grad_norm": 0.5379425287246704, "learning_rate": 7.67228971396055e-05, "loss": 0.0347, "step": 35450 }, { "grad_norm": 0.4975610375404358, "learning_rate": 7.670892063327922e-05, "loss": 0.0428, "step": 35460 }, { "grad_norm": 0.2986542880535126, "learning_rate": 7.669494120610552e-05, "loss": 0.0321, "step": 35470 }, { "grad_norm": 0.3226126730442047, "learning_rate": 7.66809588596132e-05, "loss": 0.038, "step": 35480 }, { "grad_norm": 0.4709866940975189, "learning_rate": 7.666697359533132e-05, "loss": 0.0425, "step": 35490 }, { "grad_norm": 0.5104271173477173, "learning_rate": 7.665298541478932e-05, "loss": 0.0511, "step": 35500 }, { "grad_norm": 0.38124799728393555, "learning_rate": 7.66389943195169e-05, "loss": 0.0454, "step": 35510 }, { "grad_norm": 0.33047881722450256, "learning_rate": 7.66250003110441e-05, "loss": 0.0451, "step": 35520 }, { "grad_norm": 0.3615698516368866, "learning_rate": 7.66110033909013e-05, "loss": 0.0422, "step": 35530 }, { "grad_norm": 0.30161792039871216, "learning_rate": 7.659700356061918e-05, "loss": 0.0408, "step": 35540 }, { "grad_norm": 0.26913896203041077, "learning_rate": 7.658300082172875e-05, "loss": 0.0349, "step": 35550 }, { "grad_norm": 0.2964392602443695, "learning_rate": 7.65689951757613e-05, "loss": 0.0432, "step": 35560 }, { "grad_norm": 0.2655521035194397, "learning_rate": 7.655498662424849e-05, "loss": 0.0338, "step": 35570 }, { "grad_norm": 0.41567984223365784, "learning_rate": 7.654097516872227e-05, "loss": 0.033, "step": 35580 }, { "grad_norm": 0.5081698298454285, "learning_rate": 7.65269608107149e-05, "loss": 0.0461, "step": 35590 }, { "grad_norm": 0.39199721813201904, "learning_rate": 7.6512943551759e-05, "loss": 0.0487, "step": 35600 }, { "grad_norm": 0.4076448678970337, "learning_rate": 7.649892339338743e-05, "loss": 0.0316, "step": 35610 }, { "grad_norm": 0.3857921361923218, "learning_rate": 7.648490033713344e-05, "loss": 0.0423, "step": 35620 }, { "grad_norm": 0.6781319379806519, "learning_rate": 7.647087438453058e-05, "loss": 0.0304, "step": 35630 }, { "grad_norm": 0.4177210330963135, "learning_rate": 7.64568455371127e-05, "loss": 0.0417, "step": 35640 }, { "grad_norm": 0.41927874088287354, "learning_rate": 7.644281379641396e-05, "loss": 0.0369, "step": 35650 }, { "grad_norm": 0.3854503631591797, "learning_rate": 7.642877916396887e-05, "loss": 0.044, "step": 35660 }, { "grad_norm": 0.3002723753452301, "learning_rate": 7.641474164131221e-05, "loss": 0.0451, "step": 35670 }, { "grad_norm": 0.45257073640823364, "learning_rate": 7.640070122997913e-05, "loss": 0.0381, "step": 35680 }, { "grad_norm": 0.808081328868866, "learning_rate": 7.638665793150505e-05, "loss": 0.0396, "step": 35690 }, { "grad_norm": 0.4363688826560974, "learning_rate": 7.637261174742574e-05, "loss": 0.0404, "step": 35700 }, { "grad_norm": 0.40523990988731384, "learning_rate": 7.635856267927725e-05, "loss": 0.0364, "step": 35710 }, { "grad_norm": 0.48472341895103455, "learning_rate": 7.634451072859597e-05, "loss": 0.0346, "step": 35720 }, { "grad_norm": 0.4975217580795288, "learning_rate": 7.633045589691863e-05, "loss": 0.0436, "step": 35730 }, { "grad_norm": 0.37515953183174133, "learning_rate": 7.63163981857822e-05, "loss": 0.0311, "step": 35740 }, { "grad_norm": 0.45709511637687683, "learning_rate": 7.630233759672403e-05, "loss": 0.0378, "step": 35750 }, { "grad_norm": 0.3301997482776642, "learning_rate": 7.628827413128177e-05, "loss": 0.0278, "step": 35760 }, { "grad_norm": 0.3466668426990509, "learning_rate": 7.627420779099338e-05, "loss": 0.0402, "step": 35770 }, { "grad_norm": 0.3190325200557709, "learning_rate": 7.626013857739711e-05, "loss": 0.0437, "step": 35780 }, { "grad_norm": 0.29946577548980713, "learning_rate": 7.624606649203158e-05, "loss": 0.0379, "step": 35790 }, { "grad_norm": 0.44464582204818726, "learning_rate": 7.623199153643569e-05, "loss": 0.0466, "step": 35800 }, { "grad_norm": 0.32252010703086853, "learning_rate": 7.621791371214863e-05, "loss": 0.0382, "step": 35810 }, { "grad_norm": 0.8820437788963318, "learning_rate": 7.620383302070995e-05, "loss": 0.0401, "step": 35820 }, { "grad_norm": 0.6352372765541077, "learning_rate": 7.61897494636595e-05, "loss": 0.0375, "step": 35830 }, { "grad_norm": 0.42181551456451416, "learning_rate": 7.617566304253739e-05, "loss": 0.0403, "step": 35840 }, { "grad_norm": 0.9100844860076904, "learning_rate": 7.616157375888416e-05, "loss": 0.0395, "step": 35850 }, { "grad_norm": 0.5503950119018555, "learning_rate": 7.614748161424053e-05, "loss": 0.0368, "step": 35860 }, { "grad_norm": 0.4772377908229828, "learning_rate": 7.613338661014763e-05, "loss": 0.0382, "step": 35870 }, { "grad_norm": 0.3936614394187927, "learning_rate": 7.611928874814686e-05, "loss": 0.0446, "step": 35880 }, { "grad_norm": 0.3977995216846466, "learning_rate": 7.610518802977996e-05, "loss": 0.0312, "step": 35890 }, { "grad_norm": 0.3122358024120331, "learning_rate": 7.609108445658893e-05, "loss": 0.0386, "step": 35900 }, { "grad_norm": 0.3202524483203888, "learning_rate": 7.607697803011612e-05, "loss": 0.051, "step": 35910 }, { "grad_norm": 0.6029244065284729, "learning_rate": 7.606286875190421e-05, "loss": 0.0401, "step": 35920 }, { "grad_norm": 0.5602330565452576, "learning_rate": 7.604875662349617e-05, "loss": 0.0437, "step": 35930 }, { "grad_norm": 0.45534607768058777, "learning_rate": 7.603464164643525e-05, "loss": 0.0518, "step": 35940 }, { "grad_norm": 0.3681303560733795, "learning_rate": 7.602052382226507e-05, "loss": 0.0337, "step": 35950 }, { "grad_norm": 0.5538403391838074, "learning_rate": 7.600640315252954e-05, "loss": 0.0436, "step": 35960 }, { "grad_norm": 0.5899806022644043, "learning_rate": 7.599227963877284e-05, "loss": 0.0496, "step": 35970 }, { "grad_norm": 0.4251846969127655, "learning_rate": 7.597815328253954e-05, "loss": 0.0404, "step": 35980 }, { "grad_norm": 0.43052831292152405, "learning_rate": 7.596402408537443e-05, "loss": 0.0434, "step": 35990 }, { "grad_norm": 0.3592373728752136, "learning_rate": 7.59498920488227e-05, "loss": 0.043, "step": 36000 }, { "grad_norm": 0.5506091713905334, "learning_rate": 7.593575717442979e-05, "loss": 0.0382, "step": 36010 }, { "grad_norm": 0.5335658192634583, "learning_rate": 7.592161946374147e-05, "loss": 0.0466, "step": 36020 }, { "grad_norm": 0.44391414523124695, "learning_rate": 7.590747891830381e-05, "loss": 0.0377, "step": 36030 }, { "grad_norm": 0.3904511332511902, "learning_rate": 7.58933355396632e-05, "loss": 0.0425, "step": 36040 }, { "grad_norm": 0.3869834244251251, "learning_rate": 7.587918932936636e-05, "loss": 0.0354, "step": 36050 }, { "grad_norm": 0.4646347165107727, "learning_rate": 7.586504028896028e-05, "loss": 0.0395, "step": 36060 }, { "grad_norm": 0.3489551246166229, "learning_rate": 7.585088841999228e-05, "loss": 0.0353, "step": 36070 }, { "grad_norm": 0.53654545545578, "learning_rate": 7.583673372400999e-05, "loss": 0.0436, "step": 36080 }, { "grad_norm": 0.443418025970459, "learning_rate": 7.582257620256134e-05, "loss": 0.0429, "step": 36090 }, { "grad_norm": 0.3845427334308624, "learning_rate": 7.580841585719458e-05, "loss": 0.0419, "step": 36100 }, { "grad_norm": 0.5035152435302734, "learning_rate": 7.579425268945825e-05, "loss": 0.0313, "step": 36110 }, { "grad_norm": 0.4843977987766266, "learning_rate": 7.578008670090127e-05, "loss": 0.039, "step": 36120 }, { "grad_norm": 0.41212198138237, "learning_rate": 7.576591789307272e-05, "loss": 0.0355, "step": 36130 }, { "grad_norm": 0.4205881357192993, "learning_rate": 7.575174626752216e-05, "loss": 0.0438, "step": 36140 }, { "grad_norm": 0.32512328028678894, "learning_rate": 7.573757182579934e-05, "loss": 0.0388, "step": 36150 }, { "grad_norm": 0.47871771454811096, "learning_rate": 7.572339456945435e-05, "loss": 0.0398, "step": 36160 }, { "grad_norm": 0.5851303935050964, "learning_rate": 7.570921450003762e-05, "loss": 0.0385, "step": 36170 }, { "grad_norm": 0.38133400678634644, "learning_rate": 7.569503161909983e-05, "loss": 0.0405, "step": 36180 }, { "grad_norm": 0.40307989716529846, "learning_rate": 7.568084592819202e-05, "loss": 0.0453, "step": 36190 }, { "grad_norm": 0.45406174659729004, "learning_rate": 7.566665742886551e-05, "loss": 0.0469, "step": 36200 }, { "grad_norm": 0.49887964129447937, "learning_rate": 7.565246612267194e-05, "loss": 0.0443, "step": 36210 }, { "grad_norm": 0.3713148236274719, "learning_rate": 7.563827201116325e-05, "loss": 0.0354, "step": 36220 }, { "grad_norm": 0.3271808326244354, "learning_rate": 7.562407509589164e-05, "loss": 0.0388, "step": 36230 }, { "grad_norm": 0.32709187269210815, "learning_rate": 7.560987537840973e-05, "loss": 0.0249, "step": 36240 }, { "grad_norm": 0.3488519489765167, "learning_rate": 7.559567286027036e-05, "loss": 0.0301, "step": 36250 }, { "grad_norm": 0.30674222111701965, "learning_rate": 7.558146754302668e-05, "loss": 0.0351, "step": 36260 }, { "grad_norm": 0.38141578435897827, "learning_rate": 7.556725942823217e-05, "loss": 0.041, "step": 36270 }, { "grad_norm": 0.368673712015152, "learning_rate": 7.555304851744061e-05, "loss": 0.036, "step": 36280 }, { "grad_norm": 0.42334067821502686, "learning_rate": 7.553883481220608e-05, "loss": 0.0499, "step": 36290 }, { "grad_norm": 0.34182009100914, "learning_rate": 7.552461831408298e-05, "loss": 0.0315, "step": 36300 }, { "grad_norm": 0.3916948139667511, "learning_rate": 7.551039902462599e-05, "loss": 0.0337, "step": 36310 }, { "grad_norm": 0.5078794360160828, "learning_rate": 7.549617694539011e-05, "loss": 0.0337, "step": 36320 }, { "grad_norm": 0.3349403738975525, "learning_rate": 7.548195207793066e-05, "loss": 0.0351, "step": 36330 }, { "grad_norm": 0.3170545995235443, "learning_rate": 7.546772442380323e-05, "loss": 0.0341, "step": 36340 }, { "grad_norm": 0.3793262243270874, "learning_rate": 7.545349398456376e-05, "loss": 0.0352, "step": 36350 }, { "grad_norm": 0.5227200388908386, "learning_rate": 7.543926076176845e-05, "loss": 0.0355, "step": 36360 }, { "grad_norm": 0.39207780361175537, "learning_rate": 7.542502475697385e-05, "loss": 0.0456, "step": 36370 }, { "grad_norm": 0.355253130197525, "learning_rate": 7.541078597173675e-05, "loss": 0.0407, "step": 36380 }, { "grad_norm": 0.2834361791610718, "learning_rate": 7.539654440761431e-05, "loss": 0.0348, "step": 36390 }, { "grad_norm": 0.3292500078678131, "learning_rate": 7.538230006616395e-05, "loss": 0.0355, "step": 36400 }, { "grad_norm": 0.5766750574111938, "learning_rate": 7.536805294894342e-05, "loss": 0.0445, "step": 36410 }, { "grad_norm": 0.320888489484787, "learning_rate": 7.535380305751076e-05, "loss": 0.0342, "step": 36420 }, { "grad_norm": 0.4663909077644348, "learning_rate": 7.533955039342431e-05, "loss": 0.0451, "step": 36430 }, { "grad_norm": 0.32547011971473694, "learning_rate": 7.532529495824274e-05, "loss": 0.0448, "step": 36440 }, { "grad_norm": 0.5862382054328918, "learning_rate": 7.531103675352497e-05, "loss": 0.036, "step": 36450 }, { "grad_norm": 0.502633810043335, "learning_rate": 7.529677578083029e-05, "loss": 0.0436, "step": 36460 }, { "grad_norm": 0.6631824374198914, "learning_rate": 7.528251204171823e-05, "loss": 0.0371, "step": 36470 }, { "grad_norm": 0.3189784288406372, "learning_rate": 7.526824553774867e-05, "loss": 0.0468, "step": 36480 }, { "grad_norm": 0.3275604844093323, "learning_rate": 7.525397627048178e-05, "loss": 0.0412, "step": 36490 }, { "grad_norm": 0.34430697560310364, "learning_rate": 7.523970424147802e-05, "loss": 0.0339, "step": 36500 }, { "grad_norm": 0.4537894129753113, "learning_rate": 7.522542945229813e-05, "loss": 0.0426, "step": 36510 }, { "grad_norm": 0.3977561891078949, "learning_rate": 7.521115190450321e-05, "loss": 0.0265, "step": 36520 }, { "grad_norm": 0.32935723662376404, "learning_rate": 7.519687159965462e-05, "loss": 0.0409, "step": 36530 }, { "grad_norm": 0.29075437784194946, "learning_rate": 7.518258853931403e-05, "loss": 0.0346, "step": 36540 }, { "grad_norm": 0.37844565510749817, "learning_rate": 7.516830272504342e-05, "loss": 0.0351, "step": 36550 }, { "grad_norm": 3.4999687671661377, "learning_rate": 7.515401415840509e-05, "loss": 0.0391, "step": 36560 }, { "grad_norm": 0.44043031334877014, "learning_rate": 7.513972284096155e-05, "loss": 0.0366, "step": 36570 }, { "grad_norm": 0.39741867780685425, "learning_rate": 7.512542877427574e-05, "loss": 0.0461, "step": 36580 }, { "grad_norm": 0.7263071537017822, "learning_rate": 7.51111319599108e-05, "loss": 0.0491, "step": 36590 }, { "grad_norm": 0.3389054238796234, "learning_rate": 7.509683239943024e-05, "loss": 0.0437, "step": 36600 }, { "grad_norm": 0.48821836709976196, "learning_rate": 7.50825300943978e-05, "loss": 0.0419, "step": 36610 }, { "grad_norm": 0.3034723699092865, "learning_rate": 7.50682250463776e-05, "loss": 0.0377, "step": 36620 }, { "grad_norm": 0.7843981981277466, "learning_rate": 7.5053917256934e-05, "loss": 0.0466, "step": 36630 }, { "grad_norm": 0.35402917861938477, "learning_rate": 7.503960672763166e-05, "loss": 0.0353, "step": 36640 }, { "grad_norm": 0.529708743095398, "learning_rate": 7.502529346003559e-05, "loss": 0.0391, "step": 36650 }, { "grad_norm": 0.6695042252540588, "learning_rate": 7.501097745571106e-05, "loss": 0.0342, "step": 36660 }, { "grad_norm": 0.37902623414993286, "learning_rate": 7.499665871622361e-05, "loss": 0.0517, "step": 36670 }, { "grad_norm": 0.31431326270103455, "learning_rate": 7.498233724313917e-05, "loss": 0.0435, "step": 36680 }, { "grad_norm": 0.2996140420436859, "learning_rate": 7.496801303802387e-05, "loss": 0.0337, "step": 36690 }, { "grad_norm": 0.36184507608413696, "learning_rate": 7.495368610244423e-05, "loss": 0.0282, "step": 36700 }, { "grad_norm": 0.5556908249855042, "learning_rate": 7.493935643796697e-05, "loss": 0.0398, "step": 36710 }, { "grad_norm": 0.3320837914943695, "learning_rate": 7.49250240461592e-05, "loss": 0.0369, "step": 36720 }, { "grad_norm": 0.5646036267280579, "learning_rate": 7.491068892858826e-05, "loss": 0.0363, "step": 36730 }, { "grad_norm": 0.36297935247421265, "learning_rate": 7.489635108682184e-05, "loss": 0.0356, "step": 36740 }, { "grad_norm": 0.3605192005634308, "learning_rate": 7.48820105224279e-05, "loss": 0.0329, "step": 36750 }, { "grad_norm": 0.23804357647895813, "learning_rate": 7.48676672369747e-05, "loss": 0.0326, "step": 36760 }, { "grad_norm": 0.35866788029670715, "learning_rate": 7.485332123203079e-05, "loss": 0.0342, "step": 36770 }, { "grad_norm": 0.8251959085464478, "learning_rate": 7.483897250916504e-05, "loss": 0.0508, "step": 36780 }, { "grad_norm": 0.46212929487228394, "learning_rate": 7.48246210699466e-05, "loss": 0.04, "step": 36790 }, { "grad_norm": 0.3869829475879669, "learning_rate": 7.481026691594492e-05, "loss": 0.0354, "step": 36800 }, { "grad_norm": 0.4711456298828125, "learning_rate": 7.479591004872975e-05, "loss": 0.0514, "step": 36810 }, { "grad_norm": 0.3947973847389221, "learning_rate": 7.478155046987114e-05, "loss": 0.0556, "step": 36820 }, { "grad_norm": 0.3039699196815491, "learning_rate": 7.476718818093944e-05, "loss": 0.0422, "step": 36830 }, { "grad_norm": 0.4101152718067169, "learning_rate": 7.475282318350527e-05, "loss": 0.0365, "step": 36840 }, { "grad_norm": 0.44539183378219604, "learning_rate": 7.473845547913957e-05, "loss": 0.0403, "step": 36850 }, { "grad_norm": 0.4320415258407593, "learning_rate": 7.472408506941357e-05, "loss": 0.0512, "step": 36860 }, { "grad_norm": 0.8313024640083313, "learning_rate": 7.470971195589881e-05, "loss": 0.0438, "step": 36870 }, { "grad_norm": 0.3079584836959839, "learning_rate": 7.46953361401671e-05, "loss": 0.0447, "step": 36880 }, { "grad_norm": 0.41066381335258484, "learning_rate": 7.468095762379055e-05, "loss": 0.04, "step": 36890 }, { "grad_norm": 0.4298422634601593, "learning_rate": 7.466657640834158e-05, "loss": 0.0399, "step": 36900 }, { "grad_norm": 0.41570723056793213, "learning_rate": 7.46521924953929e-05, "loss": 0.0431, "step": 36910 }, { "grad_norm": 0.27958282828330994, "learning_rate": 7.463780588651752e-05, "loss": 0.0339, "step": 36920 }, { "grad_norm": 0.6593984365463257, "learning_rate": 7.462341658328873e-05, "loss": 0.0411, "step": 36930 }, { "grad_norm": 0.5122741460800171, "learning_rate": 7.460902458728012e-05, "loss": 0.0453, "step": 36940 }, { "grad_norm": 0.4028701186180115, "learning_rate": 7.459462990006558e-05, "loss": 0.0342, "step": 36950 }, { "grad_norm": 0.5148748159408569, "learning_rate": 7.458023252321929e-05, "loss": 0.041, "step": 36960 }, { "grad_norm": 0.4117031395435333, "learning_rate": 7.456583245831573e-05, "loss": 0.0338, "step": 36970 }, { "grad_norm": 0.4886506199836731, "learning_rate": 7.455142970692968e-05, "loss": 0.0485, "step": 36980 }, { "grad_norm": 0.389880508184433, "learning_rate": 7.453702427063615e-05, "loss": 0.0379, "step": 36990 }, { "grad_norm": 0.38161808252334595, "learning_rate": 7.452261615101057e-05, "loss": 0.0412, "step": 37000 }, { "grad_norm": 0.39487624168395996, "learning_rate": 7.450820534962856e-05, "loss": 0.0453, "step": 37010 }, { "grad_norm": 0.8237579464912415, "learning_rate": 7.449379186806604e-05, "loss": 0.0423, "step": 37020 }, { "grad_norm": 0.3977537155151367, "learning_rate": 7.44793757078993e-05, "loss": 0.0385, "step": 37030 }, { "grad_norm": 0.3953194320201874, "learning_rate": 7.44649568707048e-05, "loss": 0.0404, "step": 37040 }, { "grad_norm": 0.6327202916145325, "learning_rate": 7.445053535805942e-05, "loss": 0.0525, "step": 37050 }, { "grad_norm": 0.560516893863678, "learning_rate": 7.443611117154024e-05, "loss": 0.0442, "step": 37060 }, { "grad_norm": 1.0224313735961914, "learning_rate": 7.442168431272469e-05, "loss": 0.0398, "step": 37070 }, { "grad_norm": 0.36728811264038086, "learning_rate": 7.440725478319044e-05, "loss": 0.0321, "step": 37080 }, { "grad_norm": 0.38146016001701355, "learning_rate": 7.439282258451553e-05, "loss": 0.0376, "step": 37090 }, { "grad_norm": 0.43171426653862, "learning_rate": 7.43783877182782e-05, "loss": 0.0381, "step": 37100 }, { "grad_norm": 0.44903749227523804, "learning_rate": 7.436395018605707e-05, "loss": 0.0388, "step": 37110 }, { "grad_norm": 0.3469516336917877, "learning_rate": 7.434950998943094e-05, "loss": 0.0384, "step": 37120 }, { "grad_norm": 0.3824727237224579, "learning_rate": 7.433506712997902e-05, "loss": 0.0447, "step": 37130 }, { "grad_norm": 0.3290667235851288, "learning_rate": 7.432062160928077e-05, "loss": 0.0402, "step": 37140 }, { "grad_norm": 0.2999905049800873, "learning_rate": 7.430617342891588e-05, "loss": 0.0306, "step": 37150 }, { "grad_norm": 0.29463398456573486, "learning_rate": 7.429172259046442e-05, "loss": 0.0366, "step": 37160 }, { "grad_norm": 0.3438977301120758, "learning_rate": 7.42772690955067e-05, "loss": 0.0345, "step": 37170 }, { "grad_norm": 0.39269715547561646, "learning_rate": 7.426281294562332e-05, "loss": 0.0453, "step": 37180 }, { "grad_norm": 0.2783610224723816, "learning_rate": 7.424835414239519e-05, "loss": 0.0334, "step": 37190 }, { "grad_norm": 0.3316769301891327, "learning_rate": 7.423389268740352e-05, "loss": 0.0408, "step": 37200 }, { "grad_norm": 0.35996904969215393, "learning_rate": 7.42194285822298e-05, "loss": 0.0399, "step": 37210 }, { "grad_norm": 0.31219935417175293, "learning_rate": 7.420496182845576e-05, "loss": 0.0383, "step": 37220 }, { "grad_norm": 0.38230031728744507, "learning_rate": 7.419049242766348e-05, "loss": 0.0426, "step": 37230 }, { "grad_norm": 0.801348090171814, "learning_rate": 7.417602038143533e-05, "loss": 0.0346, "step": 37240 }, { "grad_norm": 0.4643697142601013, "learning_rate": 7.416154569135393e-05, "loss": 0.0422, "step": 37250 }, { "grad_norm": 0.9207540154457092, "learning_rate": 7.414706835900222e-05, "loss": 0.0389, "step": 37260 }, { "grad_norm": 1.4207420349121094, "learning_rate": 7.413258838596345e-05, "loss": 0.0373, "step": 37270 }, { "grad_norm": 0.6481617093086243, "learning_rate": 7.411810577382106e-05, "loss": 0.0473, "step": 37280 }, { "grad_norm": 0.49459347128868103, "learning_rate": 7.410362052415889e-05, "loss": 0.0447, "step": 37290 }, { "grad_norm": 0.7495505213737488, "learning_rate": 7.408913263856102e-05, "loss": 0.0415, "step": 37300 }, { "grad_norm": 0.36858901381492615, "learning_rate": 7.407464211861183e-05, "loss": 0.0445, "step": 37310 }, { "grad_norm": 0.4147140383720398, "learning_rate": 7.406014896589597e-05, "loss": 0.0484, "step": 37320 }, { "grad_norm": 0.6088116765022278, "learning_rate": 7.404565318199838e-05, "loss": 0.0366, "step": 37330 }, { "grad_norm": 0.4351584017276764, "learning_rate": 7.403115476850432e-05, "loss": 0.0498, "step": 37340 }, { "grad_norm": 0.3640419542789459, "learning_rate": 7.401665372699932e-05, "loss": 0.0439, "step": 37350 }, { "grad_norm": 0.28230276703834534, "learning_rate": 7.400215005906916e-05, "loss": 0.042, "step": 37360 }, { "grad_norm": 0.49948689341545105, "learning_rate": 7.398764376629996e-05, "loss": 0.0404, "step": 37370 }, { "grad_norm": 0.3052448034286499, "learning_rate": 7.39731348502781e-05, "loss": 0.0371, "step": 37380 }, { "grad_norm": 0.5976805090904236, "learning_rate": 7.395862331259028e-05, "loss": 0.0411, "step": 37390 }, { "grad_norm": 0.27502158284187317, "learning_rate": 7.394410915482341e-05, "loss": 0.0396, "step": 37400 }, { "grad_norm": 0.4656476080417633, "learning_rate": 7.392959237856476e-05, "loss": 0.0467, "step": 37410 }, { "grad_norm": 0.4150450527667999, "learning_rate": 7.391507298540187e-05, "loss": 0.0398, "step": 37420 }, { "grad_norm": 0.4396893382072449, "learning_rate": 7.390055097692257e-05, "loss": 0.0354, "step": 37430 }, { "grad_norm": 0.379886269569397, "learning_rate": 7.388602635471494e-05, "loss": 0.0545, "step": 37440 }, { "grad_norm": 0.45134684443473816, "learning_rate": 7.387149912036738e-05, "loss": 0.0421, "step": 37450 }, { "grad_norm": 0.3212227523326874, "learning_rate": 7.385696927546854e-05, "loss": 0.0337, "step": 37460 }, { "grad_norm": 0.27514031529426575, "learning_rate": 7.384243682160743e-05, "loss": 0.047, "step": 37470 }, { "grad_norm": 0.31371816992759705, "learning_rate": 7.382790176037329e-05, "loss": 0.0343, "step": 37480 }, { "grad_norm": 0.4643017053604126, "learning_rate": 7.381336409335562e-05, "loss": 0.0321, "step": 37490 }, { "grad_norm": 0.32513415813446045, "learning_rate": 7.379882382214426e-05, "loss": 0.0392, "step": 37500 }, { "grad_norm": 0.3035237789154053, "learning_rate": 7.378428094832931e-05, "loss": 0.043, "step": 37510 }, { "grad_norm": 0.431795209646225, "learning_rate": 7.376973547350114e-05, "loss": 0.0591, "step": 37520 }, { "grad_norm": 0.5131528377532959, "learning_rate": 7.375518739925046e-05, "loss": 0.0411, "step": 37530 }, { "grad_norm": 0.5032866597175598, "learning_rate": 7.374063672716818e-05, "loss": 0.0327, "step": 37540 }, { "grad_norm": 0.28829896450042725, "learning_rate": 7.372608345884558e-05, "loss": 0.0397, "step": 37550 }, { "grad_norm": 0.5818714499473572, "learning_rate": 7.371152759587415e-05, "loss": 0.0431, "step": 37560 }, { "grad_norm": 0.450773149728775, "learning_rate": 7.369696913984572e-05, "loss": 0.0351, "step": 37570 }, { "grad_norm": 0.350677490234375, "learning_rate": 7.368240809235237e-05, "loss": 0.0328, "step": 37580 }, { "grad_norm": 0.45607975125312805, "learning_rate": 7.366784445498649e-05, "loss": 0.032, "step": 37590 }, { "grad_norm": 0.3729853928089142, "learning_rate": 7.36532782293407e-05, "loss": 0.036, "step": 37600 }, { "grad_norm": 0.41714733839035034, "learning_rate": 7.363870941700797e-05, "loss": 0.0379, "step": 37610 }, { "grad_norm": 0.5375720858573914, "learning_rate": 7.362413801958152e-05, "loss": 0.0474, "step": 37620 }, { "grad_norm": 0.5621006488800049, "learning_rate": 7.360956403865486e-05, "loss": 0.0455, "step": 37630 }, { "grad_norm": 0.45303016901016235, "learning_rate": 7.359498747582177e-05, "loss": 0.0378, "step": 37640 }, { "grad_norm": 0.26400259137153625, "learning_rate": 7.358040833267634e-05, "loss": 0.0415, "step": 37650 }, { "grad_norm": 0.46706727147102356, "learning_rate": 7.356582661081288e-05, "loss": 0.0553, "step": 37660 }, { "grad_norm": 0.31031811237335205, "learning_rate": 7.355124231182607e-05, "loss": 0.0358, "step": 37670 }, { "grad_norm": 0.34508007764816284, "learning_rate": 7.353665543731083e-05, "loss": 0.047, "step": 37680 }, { "grad_norm": 0.3993712067604065, "learning_rate": 7.352206598886229e-05, "loss": 0.0389, "step": 37690 }, { "grad_norm": 0.39443686604499817, "learning_rate": 7.350747396807601e-05, "loss": 0.0369, "step": 37700 }, { "grad_norm": 0.4038415253162384, "learning_rate": 7.349287937654772e-05, "loss": 0.0374, "step": 37710 }, { "grad_norm": 0.40515831112861633, "learning_rate": 7.347828221587345e-05, "loss": 0.0452, "step": 37720 }, { "grad_norm": 0.42051997780799866, "learning_rate": 7.346368248764955e-05, "loss": 0.0347, "step": 37730 }, { "grad_norm": 0.28889164328575134, "learning_rate": 7.34490801934726e-05, "loss": 0.0421, "step": 37740 }, { "grad_norm": 0.43043166399002075, "learning_rate": 7.343447533493947e-05, "loss": 0.0429, "step": 37750 }, { "grad_norm": 0.367922306060791, "learning_rate": 7.341986791364739e-05, "loss": 0.0527, "step": 37760 }, { "grad_norm": 0.28601980209350586, "learning_rate": 7.340525793119374e-05, "loss": 0.0462, "step": 37770 }, { "grad_norm": 0.4659223258495331, "learning_rate": 7.33906453891763e-05, "loss": 0.0353, "step": 37780 }, { "grad_norm": 0.2676997184753418, "learning_rate": 7.337603028919303e-05, "loss": 0.0363, "step": 37790 }, { "grad_norm": 0.4817482829093933, "learning_rate": 7.336141263284225e-05, "loss": 0.0363, "step": 37800 }, { "grad_norm": 0.2675081789493561, "learning_rate": 7.334679242172249e-05, "loss": 0.0369, "step": 37810 }, { "grad_norm": 0.35317811369895935, "learning_rate": 7.33321696574326e-05, "loss": 0.0404, "step": 37820 }, { "grad_norm": 0.27673131227493286, "learning_rate": 7.331754434157174e-05, "loss": 0.0391, "step": 37830 }, { "grad_norm": 0.34441646933555603, "learning_rate": 7.330291647573928e-05, "loss": 0.0373, "step": 37840 }, { "grad_norm": 0.41367217898368835, "learning_rate": 7.32882860615349e-05, "loss": 0.0367, "step": 37850 }, { "grad_norm": 0.3379243016242981, "learning_rate": 7.327365310055858e-05, "loss": 0.0326, "step": 37860 }, { "grad_norm": 0.285475492477417, "learning_rate": 7.325901759441055e-05, "loss": 0.0313, "step": 37870 }, { "grad_norm": 0.3426942229270935, "learning_rate": 7.324437954469132e-05, "loss": 0.0513, "step": 37880 }, { "grad_norm": 0.5150179266929626, "learning_rate": 7.32297389530017e-05, "loss": 0.0399, "step": 37890 }, { "grad_norm": 0.4198213815689087, "learning_rate": 7.321509582094274e-05, "loss": 0.0319, "step": 37900 }, { "grad_norm": 0.34368211030960083, "learning_rate": 7.320045015011582e-05, "loss": 0.0417, "step": 37910 }, { "grad_norm": 0.3369218409061432, "learning_rate": 7.318580194212253e-05, "loss": 0.0298, "step": 37920 }, { "grad_norm": 0.4920037090778351, "learning_rate": 7.317115119856483e-05, "loss": 0.0367, "step": 37930 }, { "grad_norm": 0.40680038928985596, "learning_rate": 7.315649792104485e-05, "loss": 0.0353, "step": 37940 }, { "grad_norm": 0.24258747696876526, "learning_rate": 7.314184211116507e-05, "loss": 0.044, "step": 37950 }, { "grad_norm": 0.3541043996810913, "learning_rate": 7.312718377052823e-05, "loss": 0.03, "step": 37960 }, { "grad_norm": 0.4632737934589386, "learning_rate": 7.311252290073737e-05, "loss": 0.0405, "step": 37970 }, { "grad_norm": 0.3838469088077545, "learning_rate": 7.309785950339572e-05, "loss": 0.0387, "step": 37980 }, { "grad_norm": 1.0032697916030884, "learning_rate": 7.308319358010692e-05, "loss": 0.0396, "step": 37990 }, { "grad_norm": 0.311123251914978, "learning_rate": 7.306852513247474e-05, "loss": 0.0518, "step": 38000 }, { "grad_norm": 0.3394966125488281, "learning_rate": 7.305385416210336e-05, "loss": 0.0449, "step": 38010 }, { "grad_norm": 0.3946841359138489, "learning_rate": 7.303918067059711e-05, "loss": 0.0346, "step": 38020 }, { "grad_norm": 0.5502455234527588, "learning_rate": 7.302450465956075e-05, "loss": 0.0364, "step": 38030 }, { "grad_norm": 0.5624734163284302, "learning_rate": 7.300982613059914e-05, "loss": 0.0309, "step": 38040 }, { "grad_norm": 0.4039069712162018, "learning_rate": 7.299514508531757e-05, "loss": 0.0468, "step": 38050 }, { "grad_norm": 0.37771090865135193, "learning_rate": 7.298046152532148e-05, "loss": 0.0405, "step": 38060 }, { "grad_norm": 0.3825446665287018, "learning_rate": 7.296577545221668e-05, "loss": 0.0295, "step": 38070 }, { "grad_norm": 0.42428699135780334, "learning_rate": 7.29510868676092e-05, "loss": 0.0526, "step": 38080 }, { "grad_norm": 0.35135284066200256, "learning_rate": 7.293639577310538e-05, "loss": 0.0333, "step": 38090 }, { "grad_norm": 0.34152865409851074, "learning_rate": 7.292170217031179e-05, "loss": 0.0404, "step": 38100 }, { "grad_norm": 0.4253056049346924, "learning_rate": 7.290700606083532e-05, "loss": 0.035, "step": 38110 }, { "grad_norm": 0.4338434934616089, "learning_rate": 7.28923074462831e-05, "loss": 0.0383, "step": 38120 }, { "grad_norm": 0.3627757728099823, "learning_rate": 7.287760632826255e-05, "loss": 0.0386, "step": 38130 }, { "grad_norm": 0.6628098487854004, "learning_rate": 7.286290270838139e-05, "loss": 0.0513, "step": 38140 }, { "grad_norm": 0.44885116815567017, "learning_rate": 7.284819658824756e-05, "loss": 0.0408, "step": 38150 }, { "grad_norm": 0.3946666717529297, "learning_rate": 7.283348796946929e-05, "loss": 0.0427, "step": 38160 }, { "grad_norm": 0.30389347672462463, "learning_rate": 7.28187768536551e-05, "loss": 0.035, "step": 38170 }, { "grad_norm": 0.35746464133262634, "learning_rate": 7.280406324241379e-05, "loss": 0.0453, "step": 38180 }, { "grad_norm": 0.3676425814628601, "learning_rate": 7.278934713735441e-05, "loss": 0.0446, "step": 38190 }, { "grad_norm": 0.40648698806762695, "learning_rate": 7.277462854008629e-05, "loss": 0.0348, "step": 38200 }, { "grad_norm": 0.595745325088501, "learning_rate": 7.275990745221904e-05, "loss": 0.034, "step": 38210 }, { "grad_norm": 0.30979329347610474, "learning_rate": 7.274518387536252e-05, "loss": 0.0404, "step": 38220 }, { "grad_norm": 0.36603811383247375, "learning_rate": 7.273045781112689e-05, "loss": 0.0491, "step": 38230 }, { "grad_norm": 0.3844001293182373, "learning_rate": 7.271572926112257e-05, "loss": 0.0392, "step": 38240 }, { "grad_norm": 0.3982715904712677, "learning_rate": 7.270099822696024e-05, "loss": 0.0335, "step": 38250 }, { "grad_norm": 0.37311065196990967, "learning_rate": 7.26862647102509e-05, "loss": 0.0339, "step": 38260 }, { "grad_norm": 0.3419027626514435, "learning_rate": 7.267152871260573e-05, "loss": 0.0397, "step": 38270 }, { "grad_norm": 0.4173045754432678, "learning_rate": 7.265679023563629e-05, "loss": 0.039, "step": 38280 }, { "grad_norm": 0.3537473678588867, "learning_rate": 7.264204928095433e-05, "loss": 0.0346, "step": 38290 }, { "grad_norm": 0.32630255818367004, "learning_rate": 7.262730585017188e-05, "loss": 0.0429, "step": 38300 }, { "grad_norm": 0.3459918200969696, "learning_rate": 7.261255994490131e-05, "loss": 0.0374, "step": 38310 }, { "grad_norm": 0.31259772181510925, "learning_rate": 7.259781156675517e-05, "loss": 0.0419, "step": 38320 }, { "grad_norm": 0.3793736696243286, "learning_rate": 7.258306071734633e-05, "loss": 0.0422, "step": 38330 }, { "grad_norm": 0.3515962064266205, "learning_rate": 7.256830739828792e-05, "loss": 0.0405, "step": 38340 }, { "grad_norm": 0.26724308729171753, "learning_rate": 7.255355161119336e-05, "loss": 0.031, "step": 38350 }, { "grad_norm": 0.38367676734924316, "learning_rate": 7.253879335767628e-05, "loss": 0.0282, "step": 38360 }, { "grad_norm": 0.38616499304771423, "learning_rate": 7.252403263935067e-05, "loss": 0.0376, "step": 38370 }, { "grad_norm": 0.4274788498878479, "learning_rate": 7.25092694578307e-05, "loss": 0.0373, "step": 38380 }, { "grad_norm": 0.35050973296165466, "learning_rate": 7.249450381473087e-05, "loss": 0.036, "step": 38390 }, { "grad_norm": 0.3070213198661804, "learning_rate": 7.247973571166593e-05, "loss": 0.0398, "step": 38400 }, { "grad_norm": 0.38781288266181946, "learning_rate": 7.246496515025089e-05, "loss": 0.036, "step": 38410 }, { "grad_norm": 0.5045669078826904, "learning_rate": 7.245019213210101e-05, "loss": 0.0374, "step": 38420 }, { "grad_norm": 0.44524678587913513, "learning_rate": 7.24354166588319e-05, "loss": 0.0357, "step": 38430 }, { "grad_norm": 0.515188455581665, "learning_rate": 7.242063873205935e-05, "loss": 0.0515, "step": 38440 }, { "grad_norm": 0.3446197509765625, "learning_rate": 7.240585835339946e-05, "loss": 0.0368, "step": 38450 }, { "grad_norm": 0.48135852813720703, "learning_rate": 7.239107552446857e-05, "loss": 0.0406, "step": 38460 }, { "grad_norm": 0.3945661187171936, "learning_rate": 7.237629024688334e-05, "loss": 0.0349, "step": 38470 }, { "grad_norm": 0.33339056372642517, "learning_rate": 7.236150252226064e-05, "loss": 0.0331, "step": 38480 }, { "grad_norm": 1.21379554271698, "learning_rate": 7.234671235221765e-05, "loss": 0.0511, "step": 38490 }, { "grad_norm": 0.3535010516643524, "learning_rate": 7.233191973837179e-05, "loss": 0.0382, "step": 38500 }, { "grad_norm": 0.46740567684173584, "learning_rate": 7.231712468234078e-05, "loss": 0.036, "step": 38510 }, { "grad_norm": 0.3890911936759949, "learning_rate": 7.230232718574254e-05, "loss": 0.055, "step": 38520 }, { "grad_norm": 0.8552790880203247, "learning_rate": 7.228752725019535e-05, "loss": 0.0498, "step": 38530 }, { "grad_norm": 0.4785808324813843, "learning_rate": 7.227272487731769e-05, "loss": 0.0564, "step": 38540 }, { "grad_norm": 0.45258083939552307, "learning_rate": 7.225792006872831e-05, "loss": 0.039, "step": 38550 }, { "grad_norm": 0.5881657004356384, "learning_rate": 7.224311282604628e-05, "loss": 0.0426, "step": 38560 }, { "grad_norm": 1.3806713819503784, "learning_rate": 7.222830315089085e-05, "loss": 0.0455, "step": 38570 }, { "grad_norm": 0.3636183738708496, "learning_rate": 7.22134910448816e-05, "loss": 0.0393, "step": 38580 }, { "grad_norm": 0.4093351364135742, "learning_rate": 7.219867650963839e-05, "loss": 0.0464, "step": 38590 }, { "grad_norm": 0.37816163897514343, "learning_rate": 7.21838595467813e-05, "loss": 0.0456, "step": 38600 }, { "grad_norm": 0.26528236269950867, "learning_rate": 7.216904015793067e-05, "loss": 0.0329, "step": 38610 }, { "grad_norm": 0.30949363112449646, "learning_rate": 7.215421834470713e-05, "loss": 0.0325, "step": 38620 }, { "grad_norm": 0.5356205701828003, "learning_rate": 7.213939410873158e-05, "loss": 0.0441, "step": 38630 }, { "grad_norm": 0.41856345534324646, "learning_rate": 7.212456745162519e-05, "loss": 0.0306, "step": 38640 }, { "grad_norm": 0.4580269753932953, "learning_rate": 7.210973837500937e-05, "loss": 0.0415, "step": 38650 }, { "grad_norm": 0.29005327820777893, "learning_rate": 7.209490688050578e-05, "loss": 0.0357, "step": 38660 }, { "grad_norm": 0.38829442858695984, "learning_rate": 7.208007296973641e-05, "loss": 0.0373, "step": 38670 }, { "grad_norm": 1.3856098651885986, "learning_rate": 7.206523664432345e-05, "loss": 0.0507, "step": 38680 }, { "grad_norm": 0.8289972543716431, "learning_rate": 7.205039790588939e-05, "loss": 0.0359, "step": 38690 }, { "grad_norm": 0.3179164528846741, "learning_rate": 7.203555675605697e-05, "loss": 0.0423, "step": 38700 }, { "grad_norm": 0.4582727551460266, "learning_rate": 7.202071319644917e-05, "loss": 0.0505, "step": 38710 }, { "grad_norm": 0.4619578719139099, "learning_rate": 7.200586722868932e-05, "loss": 0.042, "step": 38720 }, { "grad_norm": 0.5720801949501038, "learning_rate": 7.199101885440088e-05, "loss": 0.0363, "step": 38730 }, { "grad_norm": 0.2868484556674957, "learning_rate": 7.197616807520771e-05, "loss": 0.035, "step": 38740 }, { "grad_norm": 0.4699828028678894, "learning_rate": 7.196131489273381e-05, "loss": 0.0348, "step": 38750 }, { "grad_norm": 0.48902004957199097, "learning_rate": 7.194645930860355e-05, "loss": 0.0404, "step": 38760 }, { "grad_norm": 0.5125347375869751, "learning_rate": 7.19316013244415e-05, "loss": 0.0458, "step": 38770 }, { "grad_norm": 0.45937198400497437, "learning_rate": 7.191674094187248e-05, "loss": 0.0315, "step": 38780 }, { "grad_norm": 0.41860321164131165, "learning_rate": 7.190187816252165e-05, "loss": 0.0412, "step": 38790 }, { "grad_norm": 0.35038962960243225, "learning_rate": 7.188701298801435e-05, "loss": 0.0325, "step": 38800 }, { "grad_norm": 0.4527704119682312, "learning_rate": 7.18721454199762e-05, "loss": 0.0419, "step": 38810 }, { "grad_norm": 0.278484582901001, "learning_rate": 7.185727546003312e-05, "loss": 0.0373, "step": 38820 }, { "grad_norm": 0.5411576628684998, "learning_rate": 7.184240310981126e-05, "loss": 0.0402, "step": 38830 }, { "grad_norm": 0.3306947350502014, "learning_rate": 7.182752837093704e-05, "loss": 0.0349, "step": 38840 }, { "grad_norm": 0.3882180154323578, "learning_rate": 7.181265124503711e-05, "loss": 0.0375, "step": 38850 }, { "grad_norm": 0.4799171984195709, "learning_rate": 7.179777173373847e-05, "loss": 0.0417, "step": 38860 }, { "grad_norm": 0.33933717012405396, "learning_rate": 7.178288983866826e-05, "loss": 0.0375, "step": 38870 }, { "grad_norm": 0.3748299777507782, "learning_rate": 7.176800556145397e-05, "loss": 0.0473, "step": 38880 }, { "grad_norm": 0.24526958167552948, "learning_rate": 7.175311890372334e-05, "loss": 0.0371, "step": 38890 }, { "grad_norm": 0.5577466487884521, "learning_rate": 7.17382298671043e-05, "loss": 0.0334, "step": 38900 }, { "grad_norm": 0.649067759513855, "learning_rate": 7.172333845322515e-05, "loss": 0.0406, "step": 38910 }, { "grad_norm": 0.3451182544231415, "learning_rate": 7.170844466371436e-05, "loss": 0.0384, "step": 38920 }, { "grad_norm": 0.32224351167678833, "learning_rate": 7.16935485002007e-05, "loss": 0.034, "step": 38930 }, { "grad_norm": 0.3539793789386749, "learning_rate": 7.167864996431319e-05, "loss": 0.0393, "step": 38940 }, { "grad_norm": 0.35036441683769226, "learning_rate": 7.166374905768111e-05, "loss": 0.0337, "step": 38950 }, { "grad_norm": 0.7779937982559204, "learning_rate": 7.164884578193401e-05, "loss": 0.0351, "step": 38960 }, { "grad_norm": 0.42483994364738464, "learning_rate": 7.163394013870168e-05, "loss": 0.0323, "step": 38970 }, { "grad_norm": 0.38286855816841125, "learning_rate": 7.161903212961419e-05, "loss": 0.0469, "step": 38980 }, { "grad_norm": 0.7029533982276917, "learning_rate": 7.160412175630183e-05, "loss": 0.0327, "step": 38990 }, { "grad_norm": 0.38380494713783264, "learning_rate": 7.158920902039521e-05, "loss": 0.0341, "step": 39000 }, { "grad_norm": 0.2603139579296112, "learning_rate": 7.157429392352514e-05, "loss": 0.0354, "step": 39010 }, { "grad_norm": 0.47916379570961, "learning_rate": 7.155937646732275e-05, "loss": 0.0344, "step": 39020 }, { "grad_norm": 1.355177402496338, "learning_rate": 7.154445665341933e-05, "loss": 0.0449, "step": 39030 }, { "grad_norm": 0.4097428023815155, "learning_rate": 7.152953448344654e-05, "loss": 0.0367, "step": 39040 }, { "grad_norm": 0.6769502758979797, "learning_rate": 7.151460995903624e-05, "loss": 0.0381, "step": 39050 }, { "grad_norm": 0.5431044697761536, "learning_rate": 7.149968308182052e-05, "loss": 0.0321, "step": 39060 }, { "grad_norm": 0.9870690107345581, "learning_rate": 7.14847538534318e-05, "loss": 0.0423, "step": 39070 }, { "grad_norm": 0.5366509556770325, "learning_rate": 7.14698222755027e-05, "loss": 0.0426, "step": 39080 }, { "grad_norm": 0.3864578902721405, "learning_rate": 7.14548883496661e-05, "loss": 0.0411, "step": 39090 }, { "grad_norm": 0.34006211161613464, "learning_rate": 7.143995207755517e-05, "loss": 0.0384, "step": 39100 }, { "grad_norm": 0.27862951159477234, "learning_rate": 7.142501346080333e-05, "loss": 0.0362, "step": 39110 }, { "grad_norm": 0.3688372075557709, "learning_rate": 7.141007250104421e-05, "loss": 0.0322, "step": 39120 }, { "grad_norm": 0.4076825976371765, "learning_rate": 7.139512919991176e-05, "loss": 0.0345, "step": 39130 }, { "grad_norm": 0.40666162967681885, "learning_rate": 7.138018355904014e-05, "loss": 0.031, "step": 39140 }, { "grad_norm": 0.5080729126930237, "learning_rate": 7.13652355800638e-05, "loss": 0.0476, "step": 39150 }, { "grad_norm": 0.49366605281829834, "learning_rate": 7.135028526461743e-05, "loss": 0.031, "step": 39160 }, { "grad_norm": 0.4561888873577118, "learning_rate": 7.133533261433594e-05, "loss": 0.0356, "step": 39170 }, { "grad_norm": 0.7767712473869324, "learning_rate": 7.132037763085457e-05, "loss": 0.0426, "step": 39180 }, { "grad_norm": 0.987847626209259, "learning_rate": 7.130542031580875e-05, "loss": 0.0404, "step": 39190 }, { "grad_norm": 0.4200859069824219, "learning_rate": 7.12904606708342e-05, "loss": 0.0498, "step": 39200 }, { "grad_norm": 0.46460017561912537, "learning_rate": 7.127549869756687e-05, "loss": 0.0456, "step": 39210 }, { "grad_norm": 0.7712892293930054, "learning_rate": 7.126053439764299e-05, "loss": 0.0332, "step": 39220 }, { "grad_norm": 0.7112690210342407, "learning_rate": 7.124556777269904e-05, "loss": 0.0376, "step": 39230 }, { "grad_norm": 0.2880001366138458, "learning_rate": 7.123059882437174e-05, "loss": 0.0424, "step": 39240 }, { "grad_norm": 0.26542553305625916, "learning_rate": 7.121562755429807e-05, "loss": 0.0307, "step": 39250 }, { "grad_norm": 0.22015441954135895, "learning_rate": 7.120065396411528e-05, "loss": 0.0293, "step": 39260 }, { "grad_norm": 0.23547250032424927, "learning_rate": 7.118567805546084e-05, "loss": 0.0298, "step": 39270 }, { "grad_norm": 0.4431305229663849, "learning_rate": 7.117069982997248e-05, "loss": 0.0394, "step": 39280 }, { "grad_norm": 0.41507789492607117, "learning_rate": 7.115571928928825e-05, "loss": 0.0513, "step": 39290 }, { "grad_norm": 0.3958519399166107, "learning_rate": 7.114073643504635e-05, "loss": 0.0318, "step": 39300 }, { "grad_norm": 0.33818450570106506, "learning_rate": 7.11257512688853e-05, "loss": 0.0321, "step": 39310 }, { "grad_norm": 0.24221648275852203, "learning_rate": 7.111076379244384e-05, "loss": 0.0375, "step": 39320 }, { "grad_norm": 0.32630783319473267, "learning_rate": 7.109577400736101e-05, "loss": 0.0396, "step": 39330 }, { "grad_norm": 0.2846633195877075, "learning_rate": 7.108078191527605e-05, "loss": 0.0345, "step": 39340 }, { "grad_norm": 0.4589211344718933, "learning_rate": 7.106578751782847e-05, "loss": 0.0396, "step": 39350 }, { "grad_norm": 0.4600798785686493, "learning_rate": 7.105079081665803e-05, "loss": 0.0356, "step": 39360 }, { "grad_norm": 0.4222017228603363, "learning_rate": 7.103579181340476e-05, "loss": 0.0364, "step": 39370 }, { "grad_norm": 0.3902132213115692, "learning_rate": 7.102079050970893e-05, "loss": 0.0348, "step": 39380 }, { "grad_norm": 0.4159368574619293, "learning_rate": 7.100578690721105e-05, "loss": 0.0501, "step": 39390 }, { "grad_norm": 0.32825541496276855, "learning_rate": 7.09907810075519e-05, "loss": 0.043, "step": 39400 }, { "grad_norm": 0.6423478126525879, "learning_rate": 7.097577281237249e-05, "loss": 0.0435, "step": 39410 }, { "grad_norm": 1.307809829711914, "learning_rate": 7.09607623233141e-05, "loss": 0.0317, "step": 39420 }, { "grad_norm": 0.7707601189613342, "learning_rate": 7.094574954201827e-05, "loss": 0.0392, "step": 39430 }, { "grad_norm": 0.26419347524642944, "learning_rate": 7.093073447012675e-05, "loss": 0.038, "step": 39440 }, { "grad_norm": 0.3701143264770508, "learning_rate": 7.09157171092816e-05, "loss": 0.0347, "step": 39450 }, { "grad_norm": 0.5773344039916992, "learning_rate": 7.090069746112504e-05, "loss": 0.0331, "step": 39460 }, { "grad_norm": 0.5189907550811768, "learning_rate": 7.088567552729965e-05, "loss": 0.0402, "step": 39470 }, { "grad_norm": 0.28490397334098816, "learning_rate": 7.087065130944818e-05, "loss": 0.0444, "step": 39480 }, { "grad_norm": 0.35214072465896606, "learning_rate": 7.085562480921366e-05, "loss": 0.0387, "step": 39490 }, { "grad_norm": 0.38344481587409973, "learning_rate": 7.084059602823937e-05, "loss": 0.0307, "step": 39500 }, { "grad_norm": 0.3647516369819641, "learning_rate": 7.082556496816882e-05, "loss": 0.0354, "step": 39510 }, { "grad_norm": 0.3978880047798157, "learning_rate": 7.081053163064582e-05, "loss": 0.0368, "step": 39520 }, { "grad_norm": 0.5418509244918823, "learning_rate": 7.079549601731434e-05, "loss": 0.0407, "step": 39530 }, { "grad_norm": 0.2580825090408325, "learning_rate": 7.07804581298187e-05, "loss": 0.029, "step": 39540 }, { "grad_norm": 0.312073677778244, "learning_rate": 7.07654179698034e-05, "loss": 0.0457, "step": 39550 }, { "grad_norm": 0.47887203097343445, "learning_rate": 7.075037553891321e-05, "loss": 0.0431, "step": 39560 }, { "grad_norm": 0.2985416054725647, "learning_rate": 7.073533083879315e-05, "loss": 0.0296, "step": 39570 }, { "grad_norm": 0.47535210847854614, "learning_rate": 7.072028387108849e-05, "loss": 0.0396, "step": 39580 }, { "grad_norm": 0.47003042697906494, "learning_rate": 7.070523463744472e-05, "loss": 0.0488, "step": 39590 }, { "grad_norm": 0.5724557638168335, "learning_rate": 7.069018313950763e-05, "loss": 0.0344, "step": 39600 }, { "grad_norm": 0.35290083289146423, "learning_rate": 7.067512937892322e-05, "loss": 0.0288, "step": 39610 }, { "grad_norm": 0.4342893362045288, "learning_rate": 7.066007335733775e-05, "loss": 0.0349, "step": 39620 }, { "grad_norm": 1.0605911016464233, "learning_rate": 7.064501507639772e-05, "loss": 0.0412, "step": 39630 }, { "grad_norm": 0.45850297808647156, "learning_rate": 7.062995453774987e-05, "loss": 0.0648, "step": 39640 }, { "grad_norm": 0.6482236981391907, "learning_rate": 7.061489174304121e-05, "loss": 0.0534, "step": 39650 }, { "grad_norm": 0.34287333488464355, "learning_rate": 7.0599826693919e-05, "loss": 0.0336, "step": 39660 }, { "grad_norm": 0.9120466709136963, "learning_rate": 7.05847593920307e-05, "loss": 0.0432, "step": 39670 }, { "grad_norm": 0.5505225658416748, "learning_rate": 7.056968983902406e-05, "loss": 0.0373, "step": 39680 }, { "grad_norm": 0.3634912967681885, "learning_rate": 7.055461803654706e-05, "loss": 0.0404, "step": 39690 }, { "grad_norm": 0.38915005326271057, "learning_rate": 7.053954398624794e-05, "loss": 0.0353, "step": 39700 }, { "grad_norm": 0.36597055196762085, "learning_rate": 7.052446768977518e-05, "loss": 0.0453, "step": 39710 }, { "grad_norm": 0.29200479388237, "learning_rate": 7.050938914877748e-05, "loss": 0.0364, "step": 39720 }, { "grad_norm": 0.5062122941017151, "learning_rate": 7.049430836490382e-05, "loss": 0.0553, "step": 39730 }, { "grad_norm": 0.44440603256225586, "learning_rate": 7.047922533980342e-05, "loss": 0.0337, "step": 39740 }, { "grad_norm": 0.39327847957611084, "learning_rate": 7.046414007512571e-05, "loss": 0.0468, "step": 39750 }, { "grad_norm": 0.21491406857967377, "learning_rate": 7.044905257252042e-05, "loss": 0.0383, "step": 39760 }, { "grad_norm": 0.23174065351486206, "learning_rate": 7.04339628336375e-05, "loss": 0.036, "step": 39770 }, { "grad_norm": 0.4257465898990631, "learning_rate": 7.041887086012711e-05, "loss": 0.0412, "step": 39780 }, { "grad_norm": 0.33689194917678833, "learning_rate": 7.040377665363969e-05, "loss": 0.0312, "step": 39790 }, { "grad_norm": 0.6818081736564636, "learning_rate": 7.038868021582594e-05, "loss": 0.0394, "step": 39800 }, { "grad_norm": 0.4079444706439972, "learning_rate": 7.037358154833679e-05, "loss": 0.0374, "step": 39810 }, { "grad_norm": 0.3344181776046753, "learning_rate": 7.035848065282339e-05, "loss": 0.0333, "step": 39820 }, { "grad_norm": 0.49375203251838684, "learning_rate": 7.034337753093714e-05, "loss": 0.0317, "step": 39830 }, { "grad_norm": 0.3333929777145386, "learning_rate": 7.032827218432972e-05, "loss": 0.0389, "step": 39840 }, { "grad_norm": 0.3292728066444397, "learning_rate": 7.031316461465302e-05, "loss": 0.0453, "step": 39850 }, { "grad_norm": 0.7394622564315796, "learning_rate": 7.029805482355915e-05, "loss": 0.0327, "step": 39860 }, { "grad_norm": 0.2579314410686493, "learning_rate": 7.028294281270055e-05, "loss": 0.0372, "step": 39870 }, { "grad_norm": 0.39556047320365906, "learning_rate": 7.026782858372982e-05, "loss": 0.0375, "step": 39880 }, { "grad_norm": 0.3940182626247406, "learning_rate": 7.025271213829982e-05, "loss": 0.0378, "step": 39890 }, { "grad_norm": 0.5266101956367493, "learning_rate": 7.023759347806366e-05, "loss": 0.0333, "step": 39900 }, { "grad_norm": 1.1151355504989624, "learning_rate": 7.022247260467472e-05, "loss": 0.0423, "step": 39910 }, { "grad_norm": 0.881941020488739, "learning_rate": 7.020734951978658e-05, "loss": 0.0349, "step": 39920 }, { "grad_norm": 0.5436200499534607, "learning_rate": 7.019222422505307e-05, "loss": 0.0422, "step": 39930 }, { "grad_norm": 0.466674268245697, "learning_rate": 7.017709672212827e-05, "loss": 0.0379, "step": 39940 }, { "grad_norm": 0.3383438289165497, "learning_rate": 7.016196701266652e-05, "loss": 0.0343, "step": 39950 }, { "grad_norm": 0.4343920946121216, "learning_rate": 7.014683509832238e-05, "loss": 0.0358, "step": 39960 }, { "grad_norm": 0.9630020260810852, "learning_rate": 7.013170098075063e-05, "loss": 0.0437, "step": 39970 }, { "grad_norm": 0.6584568023681641, "learning_rate": 7.011656466160632e-05, "loss": 0.045, "step": 39980 }, { "grad_norm": 0.3900730311870575, "learning_rate": 7.010142614254475e-05, "loss": 0.039, "step": 39990 }, { "grad_norm": 0.4855828881263733, "learning_rate": 7.008628542522147e-05, "loss": 0.0452, "step": 40000 }, { "grad_norm": 0.5133259892463684, "learning_rate": 7.007114251129218e-05, "loss": 0.0389, "step": 40010 }, { "grad_norm": 0.5747536420822144, "learning_rate": 7.005599740241293e-05, "loss": 0.0349, "step": 40020 }, { "grad_norm": 0.6178879141807556, "learning_rate": 7.004085010023996e-05, "loss": 0.0451, "step": 40030 }, { "grad_norm": 0.4303174316883087, "learning_rate": 7.002570060642976e-05, "loss": 0.0414, "step": 40040 }, { "grad_norm": 0.3995724320411682, "learning_rate": 7.001054892263903e-05, "loss": 0.0434, "step": 40050 }, { "grad_norm": 0.3277638256549835, "learning_rate": 6.999539505052477e-05, "loss": 0.0374, "step": 40060 }, { "grad_norm": 0.34994807839393616, "learning_rate": 6.998023899174418e-05, "loss": 0.0374, "step": 40070 }, { "grad_norm": 0.29659634828567505, "learning_rate": 6.996508074795467e-05, "loss": 0.0384, "step": 40080 }, { "grad_norm": 0.3880034387111664, "learning_rate": 6.994992032081396e-05, "loss": 0.0453, "step": 40090 }, { "grad_norm": 0.34424036741256714, "learning_rate": 6.993475771197995e-05, "loss": 0.0306, "step": 40100 }, { "grad_norm": 0.33026760816574097, "learning_rate": 6.991959292311082e-05, "loss": 0.0315, "step": 40110 }, { "grad_norm": 0.4152970612049103, "learning_rate": 6.990442595586495e-05, "loss": 0.0423, "step": 40120 }, { "grad_norm": 0.4394337236881256, "learning_rate": 6.988925681190098e-05, "loss": 0.041, "step": 40130 }, { "grad_norm": 0.8246396780014038, "learning_rate": 6.987408549287778e-05, "loss": 0.0481, "step": 40140 }, { "grad_norm": 0.6716520190238953, "learning_rate": 6.985891200045449e-05, "loss": 0.0416, "step": 40150 }, { "grad_norm": 0.34432247281074524, "learning_rate": 6.984373633629045e-05, "loss": 0.0387, "step": 40160 }, { "grad_norm": 0.2847477197647095, "learning_rate": 6.98285585020452e-05, "loss": 0.0311, "step": 40170 }, { "grad_norm": 0.28256285190582275, "learning_rate": 6.981337849937864e-05, "loss": 0.0285, "step": 40180 }, { "grad_norm": 0.2746754288673401, "learning_rate": 6.979819632995078e-05, "loss": 0.0365, "step": 40190 }, { "grad_norm": 0.3343116343021393, "learning_rate": 6.978301199542193e-05, "loss": 0.0238, "step": 40200 }, { "grad_norm": 0.3773445785045624, "learning_rate": 6.976782549745263e-05, "loss": 0.033, "step": 40210 }, { "grad_norm": 0.48948973417282104, "learning_rate": 6.975263683770367e-05, "loss": 0.0397, "step": 40220 }, { "grad_norm": 0.47861889004707336, "learning_rate": 6.973744601783604e-05, "loss": 0.0352, "step": 40230 }, { "grad_norm": 0.4039421081542969, "learning_rate": 6.972225303951097e-05, "loss": 0.0394, "step": 40240 }, { "grad_norm": 0.6235849261283875, "learning_rate": 6.970705790438998e-05, "loss": 0.0451, "step": 40250 }, { "grad_norm": 0.5556256771087646, "learning_rate": 6.969186061413477e-05, "loss": 0.0311, "step": 40260 }, { "grad_norm": 0.25380760431289673, "learning_rate": 6.967666117040727e-05, "loss": 0.0397, "step": 40270 }, { "grad_norm": 0.2657027840614319, "learning_rate": 6.966145957486972e-05, "loss": 0.0378, "step": 40280 }, { "grad_norm": 0.324872225522995, "learning_rate": 6.964625582918449e-05, "loss": 0.0367, "step": 40290 }, { "grad_norm": 0.3601202368736267, "learning_rate": 6.963104993501425e-05, "loss": 0.0366, "step": 40300 }, { "grad_norm": 0.6161816120147705, "learning_rate": 6.961584189402192e-05, "loss": 0.0397, "step": 40310 }, { "grad_norm": 0.3388277590274811, "learning_rate": 6.96006317078706e-05, "loss": 0.0327, "step": 40320 }, { "grad_norm": 0.33088570833206177, "learning_rate": 6.958541937822367e-05, "loss": 0.0456, "step": 40330 }, { "grad_norm": 0.378425657749176, "learning_rate": 6.957020490674473e-05, "loss": 0.035, "step": 40340 }, { "grad_norm": 0.36339157819747925, "learning_rate": 6.95549882950976e-05, "loss": 0.0382, "step": 40350 }, { "grad_norm": 0.5007419586181641, "learning_rate": 6.953976954494635e-05, "loss": 0.0416, "step": 40360 }, { "grad_norm": 0.39231258630752563, "learning_rate": 6.952454865795528e-05, "loss": 0.0342, "step": 40370 }, { "grad_norm": 0.30661773681640625, "learning_rate": 6.950932563578892e-05, "loss": 0.0331, "step": 40380 }, { "grad_norm": 0.33283567428588867, "learning_rate": 6.949410048011206e-05, "loss": 0.0312, "step": 40390 }, { "grad_norm": 0.2649807333946228, "learning_rate": 6.947887319258966e-05, "loss": 0.0409, "step": 40400 }, { "grad_norm": 0.33110034465789795, "learning_rate": 6.946364377488696e-05, "loss": 0.034, "step": 40410 }, { "grad_norm": 0.6168255805969238, "learning_rate": 6.944841222866947e-05, "loss": 0.0383, "step": 40420 }, { "grad_norm": 0.2807311415672302, "learning_rate": 6.943317855560284e-05, "loss": 0.0422, "step": 40430 }, { "grad_norm": 0.29399722814559937, "learning_rate": 6.941794275735302e-05, "loss": 0.0244, "step": 40440 }, { "grad_norm": 0.3312700688838959, "learning_rate": 6.94027048355862e-05, "loss": 0.0379, "step": 40450 }, { "grad_norm": 0.3151746094226837, "learning_rate": 6.938746479196871e-05, "loss": 0.038, "step": 40460 }, { "grad_norm": 0.4966231882572174, "learning_rate": 6.937222262816724e-05, "loss": 0.0314, "step": 40470 }, { "grad_norm": 0.372265487909317, "learning_rate": 6.935697834584865e-05, "loss": 0.0402, "step": 40480 }, { "grad_norm": 0.8084140419960022, "learning_rate": 6.934173194667998e-05, "loss": 0.0332, "step": 40490 }, { "grad_norm": 0.36826154589653015, "learning_rate": 6.93264834323286e-05, "loss": 0.031, "step": 40500 }, { "grad_norm": 0.31186768412590027, "learning_rate": 6.931123280446206e-05, "loss": 0.0325, "step": 40510 }, { "grad_norm": 0.39344310760498047, "learning_rate": 6.929598006474811e-05, "loss": 0.0354, "step": 40520 }, { "grad_norm": 0.5692847371101379, "learning_rate": 6.92807252148548e-05, "loss": 0.038, "step": 40530 }, { "grad_norm": 0.31132662296295166, "learning_rate": 6.926546825645037e-05, "loss": 0.0381, "step": 40540 }, { "grad_norm": 0.2635326385498047, "learning_rate": 6.92502091912033e-05, "loss": 0.0524, "step": 40550 }, { "grad_norm": 0.3914772570133209, "learning_rate": 6.923494802078229e-05, "loss": 0.0411, "step": 40560 }, { "grad_norm": 0.5826932787895203, "learning_rate": 6.921968474685628e-05, "loss": 0.0369, "step": 40570 }, { "grad_norm": 0.38451430201530457, "learning_rate": 6.920441937109445e-05, "loss": 0.0418, "step": 40580 }, { "grad_norm": 0.45521509647369385, "learning_rate": 6.91891518951662e-05, "loss": 0.0416, "step": 40590 }, { "grad_norm": 0.40797895193099976, "learning_rate": 6.917388232074114e-05, "loss": 0.0455, "step": 40600 }, { "grad_norm": 0.3704792261123657, "learning_rate": 6.915861064948914e-05, "loss": 0.0343, "step": 40610 }, { "grad_norm": 0.41279056668281555, "learning_rate": 6.914333688308025e-05, "loss": 0.0331, "step": 40620 }, { "grad_norm": 0.43127408623695374, "learning_rate": 6.912806102318487e-05, "loss": 0.0292, "step": 40630 }, { "grad_norm": 0.3278863728046417, "learning_rate": 6.911278307147347e-05, "loss": 0.0337, "step": 40640 }, { "grad_norm": 0.281156063079834, "learning_rate": 6.909750302961684e-05, "loss": 0.029, "step": 40650 }, { "grad_norm": 0.505062997341156, "learning_rate": 6.9082220899286e-05, "loss": 0.0492, "step": 40660 }, { "grad_norm": 0.39068421721458435, "learning_rate": 6.906693668215219e-05, "loss": 0.0385, "step": 40670 }, { "grad_norm": 0.8107537627220154, "learning_rate": 6.905165037988683e-05, "loss": 0.0307, "step": 40680 }, { "grad_norm": 0.45939674973487854, "learning_rate": 6.903636199416164e-05, "loss": 0.0311, "step": 40690 }, { "grad_norm": 0.4238709509372711, "learning_rate": 6.902107152664851e-05, "loss": 0.0347, "step": 40700 }, { "grad_norm": 0.5291559100151062, "learning_rate": 6.90057789790196e-05, "loss": 0.0395, "step": 40710 }, { "grad_norm": 0.3705650866031647, "learning_rate": 6.899048435294728e-05, "loss": 0.0436, "step": 40720 }, { "grad_norm": 0.4565955400466919, "learning_rate": 6.897518765010415e-05, "loss": 0.0426, "step": 40730 }, { "grad_norm": 0.4094790518283844, "learning_rate": 6.895988887216302e-05, "loss": 0.0419, "step": 40740 }, { "grad_norm": 0.8509475588798523, "learning_rate": 6.894458802079694e-05, "loss": 0.0423, "step": 40750 }, { "grad_norm": 0.43660861253738403, "learning_rate": 6.892928509767922e-05, "loss": 0.0315, "step": 40760 }, { "grad_norm": 0.614754319190979, "learning_rate": 6.891398010448333e-05, "loss": 0.0355, "step": 40770 }, { "grad_norm": 0.5665578842163086, "learning_rate": 6.8898673042883e-05, "loss": 0.048, "step": 40780 }, { "grad_norm": 0.36052560806274414, "learning_rate": 6.888336391455222e-05, "loss": 0.0459, "step": 40790 }, { "grad_norm": 0.33531954884529114, "learning_rate": 6.886805272116513e-05, "loss": 0.0405, "step": 40800 }, { "grad_norm": 0.6576297879219055, "learning_rate": 6.885273946439617e-05, "loss": 0.055, "step": 40810 }, { "grad_norm": 0.33738821744918823, "learning_rate": 6.883742414591998e-05, "loss": 0.0406, "step": 40820 }, { "grad_norm": 0.47461479902267456, "learning_rate": 6.88221067674114e-05, "loss": 0.0317, "step": 40830 }, { "grad_norm": 0.27811768651008606, "learning_rate": 6.88067873305455e-05, "loss": 0.0373, "step": 40840 }, { "grad_norm": 0.6493703722953796, "learning_rate": 6.879146583699765e-05, "loss": 0.0495, "step": 40850 }, { "grad_norm": 0.2386084645986557, "learning_rate": 6.877614228844334e-05, "loss": 0.0324, "step": 40860 }, { "grad_norm": 0.26774370670318604, "learning_rate": 6.876081668655832e-05, "loss": 0.0383, "step": 40870 }, { "grad_norm": 0.3905435800552368, "learning_rate": 6.874548903301861e-05, "loss": 0.0404, "step": 40880 }, { "grad_norm": 0.4775356352329254, "learning_rate": 6.873015932950039e-05, "loss": 0.0459, "step": 40890 }, { "grad_norm": 0.3460299074649811, "learning_rate": 6.871482757768012e-05, "loss": 0.0361, "step": 40900 }, { "grad_norm": 0.5393320322036743, "learning_rate": 6.869949377923442e-05, "loss": 0.0423, "step": 40910 }, { "grad_norm": 0.37419360876083374, "learning_rate": 6.868415793584022e-05, "loss": 0.0298, "step": 40920 }, { "grad_norm": 0.24934931099414825, "learning_rate": 6.86688200491746e-05, "loss": 0.0331, "step": 40930 }, { "grad_norm": 0.8538826107978821, "learning_rate": 6.865348012091488e-05, "loss": 0.0329, "step": 40940 }, { "grad_norm": 0.38580870628356934, "learning_rate": 6.86381381527386e-05, "loss": 0.0331, "step": 40950 }, { "grad_norm": 0.465094655752182, "learning_rate": 6.862279414632359e-05, "loss": 0.0307, "step": 40960 }, { "grad_norm": 0.5017273426055908, "learning_rate": 6.860744810334778e-05, "loss": 0.0403, "step": 40970 }, { "grad_norm": 0.22870880365371704, "learning_rate": 6.859210002548943e-05, "loss": 0.0263, "step": 40980 }, { "grad_norm": 0.2311299443244934, "learning_rate": 6.8576749914427e-05, "loss": 0.0314, "step": 40990 }, { "grad_norm": 0.40465089678764343, "learning_rate": 6.85613977718391e-05, "loss": 0.0311, "step": 41000 }, { "grad_norm": 0.36338961124420166, "learning_rate": 6.854604359940464e-05, "loss": 0.0299, "step": 41010 }, { "grad_norm": 0.4003283977508545, "learning_rate": 6.853068739880276e-05, "loss": 0.0305, "step": 41020 }, { "grad_norm": 0.5525258183479309, "learning_rate": 6.851532917171276e-05, "loss": 0.0423, "step": 41030 }, { "grad_norm": 0.6287146210670471, "learning_rate": 6.849996891981418e-05, "loss": 0.0389, "step": 41040 }, { "grad_norm": 0.5247761607170105, "learning_rate": 6.848460664478684e-05, "loss": 0.0435, "step": 41050 }, { "grad_norm": 0.21126823127269745, "learning_rate": 6.846924234831069e-05, "loss": 0.0336, "step": 41060 }, { "grad_norm": 0.271806925535202, "learning_rate": 6.845387603206597e-05, "loss": 0.0455, "step": 41070 }, { "grad_norm": 0.61652672290802, "learning_rate": 6.843850769773312e-05, "loss": 0.0407, "step": 41080 }, { "grad_norm": 0.4889322817325592, "learning_rate": 6.842313734699278e-05, "loss": 0.0491, "step": 41090 }, { "grad_norm": 0.21086612343788147, "learning_rate": 6.840776498152584e-05, "loss": 0.0344, "step": 41100 }, { "grad_norm": 1.0139950513839722, "learning_rate": 6.83923906030134e-05, "loss": 0.0438, "step": 41110 }, { "grad_norm": 0.3755675256252289, "learning_rate": 6.837701421313677e-05, "loss": 0.039, "step": 41120 }, { "grad_norm": 0.9015558362007141, "learning_rate": 6.83616358135775e-05, "loss": 0.0475, "step": 41130 }, { "grad_norm": 0.5621705651283264, "learning_rate": 6.834625540601734e-05, "loss": 0.0446, "step": 41140 }, { "grad_norm": 0.4817582964897156, "learning_rate": 6.833087299213829e-05, "loss": 0.0381, "step": 41150 }, { "grad_norm": 0.30224522948265076, "learning_rate": 6.83154885736225e-05, "loss": 0.0423, "step": 41160 }, { "grad_norm": 0.4587370753288269, "learning_rate": 6.830010215215244e-05, "loss": 0.0508, "step": 41170 }, { "grad_norm": 0.3627294600009918, "learning_rate": 6.828471372941072e-05, "loss": 0.0342, "step": 41180 }, { "grad_norm": 0.39373213052749634, "learning_rate": 6.826932330708018e-05, "loss": 0.0361, "step": 41190 }, { "grad_norm": 0.2997516691684723, "learning_rate": 6.825393088684393e-05, "loss": 0.0369, "step": 41200 }, { "grad_norm": 0.627615749835968, "learning_rate": 6.823853647038524e-05, "loss": 0.0352, "step": 41210 }, { "grad_norm": 0.28249168395996094, "learning_rate": 6.822314005938762e-05, "loss": 0.0297, "step": 41220 }, { "grad_norm": 0.3954353630542755, "learning_rate": 6.820774165553479e-05, "loss": 0.0462, "step": 41230 }, { "grad_norm": 0.6480016112327576, "learning_rate": 6.819234126051072e-05, "loss": 0.0295, "step": 41240 }, { "grad_norm": 0.3563222885131836, "learning_rate": 6.817693887599956e-05, "loss": 0.036, "step": 41250 }, { "grad_norm": 0.39238670468330383, "learning_rate": 6.816153450368568e-05, "loss": 0.0388, "step": 41260 }, { "grad_norm": 0.31643223762512207, "learning_rate": 6.814612814525372e-05, "loss": 0.0434, "step": 41270 }, { "grad_norm": 0.35530686378479004, "learning_rate": 6.813071980238846e-05, "loss": 0.0284, "step": 41280 }, { "grad_norm": 0.43703579902648926, "learning_rate": 6.811530947677492e-05, "loss": 0.0315, "step": 41290 }, { "grad_norm": 0.38863322138786316, "learning_rate": 6.809989717009839e-05, "loss": 0.0294, "step": 41300 }, { "grad_norm": 0.41640985012054443, "learning_rate": 6.808448288404431e-05, "loss": 0.0437, "step": 41310 }, { "grad_norm": 0.4508680999279022, "learning_rate": 6.806906662029838e-05, "loss": 0.0302, "step": 41320 }, { "grad_norm": 0.3608681857585907, "learning_rate": 6.805364838054647e-05, "loss": 0.0338, "step": 41330 }, { "grad_norm": 0.35670945048332214, "learning_rate": 6.803822816647474e-05, "loss": 0.0278, "step": 41340 }, { "grad_norm": 0.30310148000717163, "learning_rate": 6.802280597976949e-05, "loss": 0.0356, "step": 41350 }, { "grad_norm": 0.42832058668136597, "learning_rate": 6.800738182211727e-05, "loss": 0.0414, "step": 41360 }, { "grad_norm": 0.4113539755344391, "learning_rate": 6.799195569520485e-05, "loss": 0.0401, "step": 41370 }, { "grad_norm": 0.37099459767341614, "learning_rate": 6.79765276007192e-05, "loss": 0.0368, "step": 41380 }, { "grad_norm": 0.4618326723575592, "learning_rate": 6.796109754034753e-05, "loss": 0.0323, "step": 41390 }, { "grad_norm": 0.36143049597740173, "learning_rate": 6.794566551577724e-05, "loss": 0.0323, "step": 41400 }, { "grad_norm": 0.41619786620140076, "learning_rate": 6.793023152869594e-05, "loss": 0.0351, "step": 41410 }, { "grad_norm": 0.3758358657360077, "learning_rate": 6.791479558079147e-05, "loss": 0.0321, "step": 41420 }, { "grad_norm": 0.43892520666122437, "learning_rate": 6.789935767375191e-05, "loss": 0.0378, "step": 41430 }, { "grad_norm": 0.6966116428375244, "learning_rate": 6.78839178092655e-05, "loss": 0.0367, "step": 41440 }, { "grad_norm": 0.28585782647132874, "learning_rate": 6.786847598902072e-05, "loss": 0.0339, "step": 41450 }, { "grad_norm": 0.43302392959594727, "learning_rate": 6.785303221470627e-05, "loss": 0.0428, "step": 41460 }, { "grad_norm": 0.29339346289634705, "learning_rate": 6.783758648801108e-05, "loss": 0.0311, "step": 41470 }, { "grad_norm": 0.4680909216403961, "learning_rate": 6.782213881062422e-05, "loss": 0.0441, "step": 41480 }, { "grad_norm": 0.3946552574634552, "learning_rate": 6.780668918423508e-05, "loss": 0.0357, "step": 41490 }, { "grad_norm": 0.25303569436073303, "learning_rate": 6.779123761053317e-05, "loss": 0.0315, "step": 41500 }, { "grad_norm": 0.38882356882095337, "learning_rate": 6.777578409120826e-05, "loss": 0.0383, "step": 41510 }, { "grad_norm": 0.41644856333732605, "learning_rate": 6.776032862795033e-05, "loss": 0.0335, "step": 41520 }, { "grad_norm": 0.36934196949005127, "learning_rate": 6.774487122244956e-05, "loss": 0.0356, "step": 41530 }, { "grad_norm": 0.4058925211429596, "learning_rate": 6.772941187639637e-05, "loss": 0.036, "step": 41540 }, { "grad_norm": 0.46754664182662964, "learning_rate": 6.771395059148134e-05, "loss": 0.0319, "step": 41550 }, { "grad_norm": 0.2408631145954132, "learning_rate": 6.76984873693953e-05, "loss": 0.0308, "step": 41560 }, { "grad_norm": 0.518841564655304, "learning_rate": 6.768302221182931e-05, "loss": 0.0406, "step": 41570 }, { "grad_norm": 0.7368676662445068, "learning_rate": 6.766755512047457e-05, "loss": 0.0462, "step": 41580 }, { "grad_norm": 0.3120989501476288, "learning_rate": 6.765208609702259e-05, "loss": 0.0308, "step": 41590 }, { "grad_norm": 0.3811580240726471, "learning_rate": 6.763661514316499e-05, "loss": 0.0274, "step": 41600 }, { "grad_norm": 0.5511805415153503, "learning_rate": 6.76211422605937e-05, "loss": 0.0388, "step": 41610 }, { "grad_norm": 0.6940933465957642, "learning_rate": 6.760566745100076e-05, "loss": 0.0397, "step": 41620 }, { "grad_norm": 0.30723169445991516, "learning_rate": 6.75901907160785e-05, "loss": 0.0375, "step": 41630 }, { "grad_norm": 0.3408347964286804, "learning_rate": 6.757471205751943e-05, "loss": 0.0329, "step": 41640 }, { "grad_norm": 1.0121822357177734, "learning_rate": 6.75592314770163e-05, "loss": 0.0367, "step": 41650 }, { "grad_norm": 0.33835238218307495, "learning_rate": 6.7543748976262e-05, "loss": 0.0384, "step": 41660 }, { "grad_norm": 0.4358532130718231, "learning_rate": 6.752826455694968e-05, "loss": 0.0258, "step": 41670 }, { "grad_norm": 0.24895399808883667, "learning_rate": 6.751277822077271e-05, "loss": 0.0325, "step": 41680 }, { "grad_norm": 0.4124366343021393, "learning_rate": 6.749728996942465e-05, "loss": 0.0358, "step": 41690 }, { "grad_norm": 0.3477553427219391, "learning_rate": 6.748179980459924e-05, "loss": 0.0339, "step": 41700 }, { "grad_norm": 0.561305582523346, "learning_rate": 6.746630772799052e-05, "loss": 0.0353, "step": 41710 }, { "grad_norm": 0.4033742845058441, "learning_rate": 6.745081374129262e-05, "loss": 0.0498, "step": 41720 }, { "grad_norm": 0.6719545722007751, "learning_rate": 6.743531784619998e-05, "loss": 0.0401, "step": 41730 }, { "grad_norm": 0.3891531229019165, "learning_rate": 6.741982004440719e-05, "loss": 0.032, "step": 41740 }, { "grad_norm": 0.3704988360404968, "learning_rate": 6.740432033760907e-05, "loss": 0.0332, "step": 41750 }, { "grad_norm": 0.4727257192134857, "learning_rate": 6.738881872750066e-05, "loss": 0.0389, "step": 41760 }, { "grad_norm": 1.0504200458526611, "learning_rate": 6.737331521577715e-05, "loss": 0.0464, "step": 41770 }, { "grad_norm": 0.3859803080558777, "learning_rate": 6.735780980413403e-05, "loss": 0.0381, "step": 41780 }, { "grad_norm": 0.44404336810112, "learning_rate": 6.734230249426692e-05, "loss": 0.0415, "step": 41790 }, { "grad_norm": 0.2961881756782532, "learning_rate": 6.732679328787168e-05, "loss": 0.0358, "step": 41800 }, { "grad_norm": 0.32185354828834534, "learning_rate": 6.731128218664438e-05, "loss": 0.0314, "step": 41810 }, { "grad_norm": 0.5395655035972595, "learning_rate": 6.729576919228129e-05, "loss": 0.0361, "step": 41820 }, { "grad_norm": 0.3826359212398529, "learning_rate": 6.728025430647888e-05, "loss": 0.0308, "step": 41830 }, { "grad_norm": 0.49674299359321594, "learning_rate": 6.726473753093383e-05, "loss": 0.0457, "step": 41840 }, { "grad_norm": 0.3542211651802063, "learning_rate": 6.724921886734305e-05, "loss": 0.0301, "step": 41850 }, { "grad_norm": 0.27027440071105957, "learning_rate": 6.723369831740363e-05, "loss": 0.0428, "step": 41860 }, { "grad_norm": 0.34438568353652954, "learning_rate": 6.721817588281288e-05, "loss": 0.0337, "step": 41870 }, { "grad_norm": 0.4499269425868988, "learning_rate": 6.720265156526828e-05, "loss": 0.0362, "step": 41880 }, { "grad_norm": 0.3589136302471161, "learning_rate": 6.718712536646758e-05, "loss": 0.0307, "step": 41890 }, { "grad_norm": 0.4096646308898926, "learning_rate": 6.71715972881087e-05, "loss": 0.0486, "step": 41900 }, { "grad_norm": 0.4216625690460205, "learning_rate": 6.715606733188974e-05, "loss": 0.0388, "step": 41910 }, { "grad_norm": 0.4023616313934326, "learning_rate": 6.714053549950909e-05, "loss": 0.0288, "step": 41920 }, { "grad_norm": 0.34265440702438354, "learning_rate": 6.71250017926652e-05, "loss": 0.0418, "step": 41930 }, { "grad_norm": 0.322660356760025, "learning_rate": 6.71094662130569e-05, "loss": 0.0338, "step": 41940 }, { "grad_norm": 0.3554029166698456, "learning_rate": 6.709392876238307e-05, "loss": 0.0318, "step": 41950 }, { "grad_norm": 0.4701141119003296, "learning_rate": 6.707838944234289e-05, "loss": 0.043, "step": 41960 }, { "grad_norm": 0.4634595513343811, "learning_rate": 6.706284825463574e-05, "loss": 0.0444, "step": 41970 }, { "grad_norm": 0.40592366456985474, "learning_rate": 6.704730520096114e-05, "loss": 0.0328, "step": 41980 }, { "grad_norm": 0.5508264303207397, "learning_rate": 6.703176028301888e-05, "loss": 0.0439, "step": 41990 }, { "grad_norm": 0.3477446436882019, "learning_rate": 6.701621350250892e-05, "loss": 0.036, "step": 42000 }, { "grad_norm": 0.25534477829933167, "learning_rate": 6.700066486113144e-05, "loss": 0.0339, "step": 42010 }, { "grad_norm": 0.4421050548553467, "learning_rate": 6.69851143605868e-05, "loss": 0.036, "step": 42020 }, { "grad_norm": 0.3865491449832916, "learning_rate": 6.696956200257561e-05, "loss": 0.0351, "step": 42030 }, { "grad_norm": 0.22414137423038483, "learning_rate": 6.695400778879863e-05, "loss": 0.0404, "step": 42040 }, { "grad_norm": 0.44677624106407166, "learning_rate": 6.693845172095683e-05, "loss": 0.0485, "step": 42050 }, { "grad_norm": 0.4305347800254822, "learning_rate": 6.692289380075142e-05, "loss": 0.0347, "step": 42060 }, { "grad_norm": 0.25498276948928833, "learning_rate": 6.690733402988379e-05, "loss": 0.03, "step": 42070 }, { "grad_norm": 0.37887969613075256, "learning_rate": 6.689177241005553e-05, "loss": 0.0321, "step": 42080 }, { "grad_norm": 0.28067898750305176, "learning_rate": 6.687620894296844e-05, "loss": 0.0385, "step": 42090 }, { "grad_norm": 0.5012296438217163, "learning_rate": 6.686064363032451e-05, "loss": 0.0363, "step": 42100 }, { "grad_norm": 0.4380716383457184, "learning_rate": 6.684507647382596e-05, "loss": 0.037, "step": 42110 }, { "grad_norm": 0.3897198736667633, "learning_rate": 6.682950747517513e-05, "loss": 0.031, "step": 42120 }, { "grad_norm": 0.5268429517745972, "learning_rate": 6.68139366360747e-05, "loss": 0.0358, "step": 42130 }, { "grad_norm": 0.2897590100765228, "learning_rate": 6.679836395822744e-05, "loss": 0.0311, "step": 42140 }, { "grad_norm": 0.3580067753791809, "learning_rate": 6.678278944333633e-05, "loss": 0.0384, "step": 42150 }, { "grad_norm": 0.36726972460746765, "learning_rate": 6.676721309310462e-05, "loss": 0.0429, "step": 42160 }, { "grad_norm": 0.249911367893219, "learning_rate": 6.67516349092357e-05, "loss": 0.0309, "step": 42170 }, { "grad_norm": 0.2467634379863739, "learning_rate": 6.673605489343315e-05, "loss": 0.0335, "step": 42180 }, { "grad_norm": 0.2540282607078552, "learning_rate": 6.672047304740081e-05, "loss": 0.0337, "step": 42190 }, { "grad_norm": 0.4603140354156494, "learning_rate": 6.670488937284268e-05, "loss": 0.0346, "step": 42200 }, { "grad_norm": 0.3293549716472626, "learning_rate": 6.668930387146294e-05, "loss": 0.0333, "step": 42210 }, { "grad_norm": 0.4808288812637329, "learning_rate": 6.667371654496605e-05, "loss": 0.036, "step": 42220 }, { "grad_norm": 0.3479710519313812, "learning_rate": 6.665812739505659e-05, "loss": 0.0298, "step": 42230 }, { "grad_norm": 0.2686174213886261, "learning_rate": 6.664253642343935e-05, "loss": 0.0314, "step": 42240 }, { "grad_norm": 0.3958979845046997, "learning_rate": 6.662694363181935e-05, "loss": 0.0345, "step": 42250 }, { "grad_norm": 0.27839624881744385, "learning_rate": 6.66113490219018e-05, "loss": 0.0263, "step": 42260 }, { "grad_norm": 0.4336652457714081, "learning_rate": 6.659575259539207e-05, "loss": 0.0436, "step": 42270 }, { "grad_norm": 0.6400441527366638, "learning_rate": 6.65801543539958e-05, "loss": 0.0355, "step": 42280 }, { "grad_norm": 0.3530031740665436, "learning_rate": 6.65645542994188e-05, "loss": 0.0424, "step": 42290 }, { "grad_norm": 0.3823676109313965, "learning_rate": 6.654895243336702e-05, "loss": 0.0323, "step": 42300 }, { "grad_norm": 0.3994736075401306, "learning_rate": 6.65333487575467e-05, "loss": 0.0327, "step": 42310 }, { "grad_norm": 0.32628363370895386, "learning_rate": 6.651774327366422e-05, "loss": 0.0287, "step": 42320 }, { "grad_norm": 0.4303762912750244, "learning_rate": 6.650213598342613e-05, "loss": 0.0319, "step": 42330 }, { "grad_norm": 0.5956652760505676, "learning_rate": 6.648652688853931e-05, "loss": 0.0417, "step": 42340 }, { "grad_norm": 0.3245229721069336, "learning_rate": 6.647091599071066e-05, "loss": 0.0354, "step": 42350 }, { "grad_norm": 0.3643154203891754, "learning_rate": 6.645530329164743e-05, "loss": 0.0362, "step": 42360 }, { "grad_norm": 0.47269442677497864, "learning_rate": 6.643968879305694e-05, "loss": 0.0357, "step": 42370 }, { "grad_norm": 0.32773467898368835, "learning_rate": 6.642407249664683e-05, "loss": 0.0266, "step": 42380 }, { "grad_norm": 0.41212376952171326, "learning_rate": 6.640845440412483e-05, "loss": 0.0361, "step": 42390 }, { "grad_norm": 0.3069905638694763, "learning_rate": 6.639283451719893e-05, "loss": 0.0275, "step": 42400 }, { "grad_norm": 0.3653113842010498, "learning_rate": 6.637721283757729e-05, "loss": 0.0296, "step": 42410 }, { "grad_norm": 0.4281160831451416, "learning_rate": 6.636158936696828e-05, "loss": 0.031, "step": 42420 }, { "grad_norm": 0.380801796913147, "learning_rate": 6.634596410708047e-05, "loss": 0.0383, "step": 42430 }, { "grad_norm": 0.5436680316925049, "learning_rate": 6.633033705962257e-05, "loss": 0.0294, "step": 42440 }, { "grad_norm": 0.3678940534591675, "learning_rate": 6.631470822630359e-05, "loss": 0.033, "step": 42450 }, { "grad_norm": 0.8284432291984558, "learning_rate": 6.629907760883263e-05, "loss": 0.0392, "step": 42460 }, { "grad_norm": 0.531650185585022, "learning_rate": 6.628344520891905e-05, "loss": 0.0302, "step": 42470 }, { "grad_norm": 0.2698412537574768, "learning_rate": 6.62678110282724e-05, "loss": 0.0286, "step": 42480 }, { "grad_norm": 0.4282526671886444, "learning_rate": 6.625217506860237e-05, "loss": 0.0367, "step": 42490 }, { "grad_norm": 0.34412896633148193, "learning_rate": 6.623653733161892e-05, "loss": 0.0375, "step": 42500 }, { "grad_norm": 0.3758094310760498, "learning_rate": 6.622089781903217e-05, "loss": 0.0352, "step": 42510 }, { "grad_norm": 0.32698771357536316, "learning_rate": 6.620525653255241e-05, "loss": 0.0331, "step": 42520 }, { "grad_norm": 0.4116099178791046, "learning_rate": 6.618961347389015e-05, "loss": 0.0337, "step": 42530 }, { "grad_norm": 0.664498507976532, "learning_rate": 6.617396864475613e-05, "loss": 0.0335, "step": 42540 }, { "grad_norm": 0.3769332468509674, "learning_rate": 6.61583220468612e-05, "loss": 0.0291, "step": 42550 }, { "grad_norm": 0.3498988747596741, "learning_rate": 6.614267368191645e-05, "loss": 0.0411, "step": 42560 }, { "grad_norm": 0.5064662098884583, "learning_rate": 6.61270235516332e-05, "loss": 0.0331, "step": 42570 }, { "grad_norm": 0.3185432255268097, "learning_rate": 6.611137165772287e-05, "loss": 0.0357, "step": 42580 }, { "grad_norm": 0.3746669888496399, "learning_rate": 6.60957180018972e-05, "loss": 0.0398, "step": 42590 }, { "grad_norm": 0.5329744815826416, "learning_rate": 6.608006258586797e-05, "loss": 0.0296, "step": 42600 }, { "grad_norm": 0.4528406858444214, "learning_rate": 6.606440541134731e-05, "loss": 0.0388, "step": 42610 }, { "grad_norm": 0.2428710013628006, "learning_rate": 6.60487464800474e-05, "loss": 0.0419, "step": 42620 }, { "grad_norm": 0.49288102984428406, "learning_rate": 6.603308579368071e-05, "loss": 0.0297, "step": 42630 }, { "grad_norm": 0.6877891421318054, "learning_rate": 6.601742335395987e-05, "loss": 0.0356, "step": 42640 }, { "grad_norm": 0.39239075779914856, "learning_rate": 6.600175916259769e-05, "loss": 0.0323, "step": 42650 }, { "grad_norm": 0.2902522683143616, "learning_rate": 6.598609322130718e-05, "loss": 0.0318, "step": 42660 }, { "grad_norm": 0.25205090641975403, "learning_rate": 6.597042553180155e-05, "loss": 0.0317, "step": 42670 }, { "grad_norm": 0.3632351756095886, "learning_rate": 6.59547560957942e-05, "loss": 0.0322, "step": 42680 }, { "grad_norm": 0.6775850057601929, "learning_rate": 6.59390849149987e-05, "loss": 0.0453, "step": 42690 }, { "grad_norm": 0.4378271996974945, "learning_rate": 6.592341199112886e-05, "loss": 0.039, "step": 42700 }, { "grad_norm": 0.44902557134628296, "learning_rate": 6.590773732589861e-05, "loss": 0.0328, "step": 42710 }, { "grad_norm": 0.5060417056083679, "learning_rate": 6.589206092102213e-05, "loss": 0.047, "step": 42720 }, { "grad_norm": 0.3506792187690735, "learning_rate": 6.587638277821375e-05, "loss": 0.0424, "step": 42730 }, { "grad_norm": 0.38680267333984375, "learning_rate": 6.586070289918803e-05, "loss": 0.0356, "step": 42740 }, { "grad_norm": 0.36495500802993774, "learning_rate": 6.584502128565968e-05, "loss": 0.0383, "step": 42750 }, { "grad_norm": 0.39269933104515076, "learning_rate": 6.582933793934363e-05, "loss": 0.04, "step": 42760 }, { "grad_norm": 0.21816079318523407, "learning_rate": 6.581365286195498e-05, "loss": 0.0366, "step": 42770 }, { "grad_norm": 0.34957218170166016, "learning_rate": 6.579796605520903e-05, "loss": 0.0331, "step": 42780 }, { "grad_norm": 0.5542402267456055, "learning_rate": 6.578227752082127e-05, "loss": 0.0333, "step": 42790 }, { "grad_norm": 0.3968998193740845, "learning_rate": 6.576658726050735e-05, "loss": 0.029, "step": 42800 }, { "grad_norm": 0.28034865856170654, "learning_rate": 6.575089527598316e-05, "loss": 0.026, "step": 42810 }, { "grad_norm": 0.6394943594932556, "learning_rate": 6.573520156896476e-05, "loss": 0.0304, "step": 42820 }, { "grad_norm": 0.35434725880622864, "learning_rate": 6.571950614116835e-05, "loss": 0.0344, "step": 42830 }, { "grad_norm": 0.3920140564441681, "learning_rate": 6.570380899431039e-05, "loss": 0.052, "step": 42840 }, { "grad_norm": 0.5345960855484009, "learning_rate": 6.568811013010749e-05, "loss": 0.0338, "step": 42850 }, { "grad_norm": 0.24240151047706604, "learning_rate": 6.567240955027645e-05, "loss": 0.0335, "step": 42860 }, { "grad_norm": 0.3497033715248108, "learning_rate": 6.565670725653427e-05, "loss": 0.0299, "step": 42870 }, { "grad_norm": 0.2727303206920624, "learning_rate": 6.564100325059812e-05, "loss": 0.04, "step": 42880 }, { "grad_norm": 0.39497119188308716, "learning_rate": 6.562529753418538e-05, "loss": 0.0307, "step": 42890 }, { "grad_norm": 0.4056043326854706, "learning_rate": 6.56095901090136e-05, "loss": 0.0359, "step": 42900 }, { "grad_norm": 0.2612498104572296, "learning_rate": 6.55938809768005e-05, "loss": 0.0296, "step": 42910 }, { "grad_norm": 0.5540112853050232, "learning_rate": 6.557817013926402e-05, "loss": 0.0311, "step": 42920 }, { "grad_norm": 0.32908904552459717, "learning_rate": 6.556245759812227e-05, "loss": 0.0394, "step": 42930 }, { "grad_norm": 0.3931235373020172, "learning_rate": 6.554674335509357e-05, "loss": 0.0428, "step": 42940 }, { "grad_norm": 0.29880425333976746, "learning_rate": 6.553102741189638e-05, "loss": 0.0322, "step": 42950 }, { "grad_norm": 0.38200661540031433, "learning_rate": 6.55153097702494e-05, "loss": 0.0425, "step": 42960 }, { "grad_norm": 0.39823323488235474, "learning_rate": 6.549959043187146e-05, "loss": 0.0421, "step": 42970 }, { "grad_norm": 0.3929368853569031, "learning_rate": 6.54838693984816e-05, "loss": 0.0311, "step": 42980 }, { "grad_norm": 0.5794782638549805, "learning_rate": 6.546814667179909e-05, "loss": 0.0369, "step": 42990 }, { "grad_norm": 0.3484783172607422, "learning_rate": 6.545242225354328e-05, "loss": 0.0358, "step": 43000 }, { "grad_norm": 0.38530415296554565, "learning_rate": 6.543669614543383e-05, "loss": 0.0272, "step": 43010 }, { "grad_norm": 0.30787327885627747, "learning_rate": 6.542096834919049e-05, "loss": 0.0286, "step": 43020 }, { "grad_norm": 0.33527636528015137, "learning_rate": 6.540523886653324e-05, "loss": 0.0368, "step": 43030 }, { "grad_norm": 0.3895074427127838, "learning_rate": 6.538950769918222e-05, "loss": 0.0333, "step": 43040 }, { "grad_norm": 0.39333727955818176, "learning_rate": 6.537377484885779e-05, "loss": 0.0326, "step": 43050 }, { "grad_norm": 0.7291915416717529, "learning_rate": 6.535804031728045e-05, "loss": 0.032, "step": 43060 }, { "grad_norm": 0.32796505093574524, "learning_rate": 6.534230410617092e-05, "loss": 0.0393, "step": 43070 }, { "grad_norm": 0.20801320672035217, "learning_rate": 6.532656621725007e-05, "loss": 0.0366, "step": 43080 }, { "grad_norm": 0.6800782680511475, "learning_rate": 6.531082665223898e-05, "loss": 0.0431, "step": 43090 }, { "grad_norm": 0.3196486830711365, "learning_rate": 6.529508541285889e-05, "loss": 0.0274, "step": 43100 }, { "grad_norm": 0.40074366331100464, "learning_rate": 6.527934250083128e-05, "loss": 0.0411, "step": 43110 }, { "grad_norm": 0.5081363916397095, "learning_rate": 6.526359791787775e-05, "loss": 0.0367, "step": 43120 }, { "grad_norm": 0.357483834028244, "learning_rate": 6.524785166572009e-05, "loss": 0.0326, "step": 43130 }, { "grad_norm": 0.6490859389305115, "learning_rate": 6.523210374608029e-05, "loss": 0.0354, "step": 43140 }, { "grad_norm": 0.278080016374588, "learning_rate": 6.521635416068054e-05, "loss": 0.0329, "step": 43150 }, { "grad_norm": 0.5214020609855652, "learning_rate": 6.520060291124317e-05, "loss": 0.0421, "step": 43160 }, { "grad_norm": 0.806277334690094, "learning_rate": 6.518484999949072e-05, "loss": 0.0347, "step": 43170 }, { "grad_norm": 0.5475288033485413, "learning_rate": 6.516909542714591e-05, "loss": 0.04, "step": 43180 }, { "grad_norm": 0.4034004211425781, "learning_rate": 6.515333919593161e-05, "loss": 0.045, "step": 43190 }, { "grad_norm": 0.47877025604248047, "learning_rate": 6.513758130757094e-05, "loss": 0.0447, "step": 43200 }, { "grad_norm": 0.36875155568122864, "learning_rate": 6.512182176378713e-05, "loss": 0.0319, "step": 43210 }, { "grad_norm": 0.3770553767681122, "learning_rate": 6.510606056630362e-05, "loss": 0.0334, "step": 43220 }, { "grad_norm": 0.2396107017993927, "learning_rate": 6.509029771684405e-05, "loss": 0.0381, "step": 43230 }, { "grad_norm": 0.3377054035663605, "learning_rate": 6.50745332171322e-05, "loss": 0.0332, "step": 43240 }, { "grad_norm": 0.6485287547111511, "learning_rate": 6.505876706889207e-05, "loss": 0.0435, "step": 43250 }, { "grad_norm": 0.7508370876312256, "learning_rate": 6.504299927384781e-05, "loss": 0.0474, "step": 43260 }, { "grad_norm": 0.46334826946258545, "learning_rate": 6.502722983372378e-05, "loss": 0.0347, "step": 43270 }, { "grad_norm": 0.554998517036438, "learning_rate": 6.501145875024446e-05, "loss": 0.0401, "step": 43280 }, { "grad_norm": 0.30353039503097534, "learning_rate": 6.499568602513459e-05, "loss": 0.0375, "step": 43290 }, { "grad_norm": 0.3764610290527344, "learning_rate": 6.497991166011903e-05, "loss": 0.038, "step": 43300 }, { "grad_norm": 0.4294433891773224, "learning_rate": 6.49641356569229e-05, "loss": 0.0353, "step": 43310 }, { "grad_norm": 0.3541242182254791, "learning_rate": 6.494835801727135e-05, "loss": 0.0376, "step": 43320 }, { "grad_norm": 0.5568318963050842, "learning_rate": 6.493257874288986e-05, "loss": 0.0313, "step": 43330 }, { "grad_norm": 0.46075114607810974, "learning_rate": 6.491679783550399e-05, "loss": 0.0344, "step": 43340 }, { "grad_norm": 0.4461531639099121, "learning_rate": 6.490101529683957e-05, "loss": 0.0333, "step": 43350 }, { "grad_norm": 0.3126389682292938, "learning_rate": 6.488523112862248e-05, "loss": 0.036, "step": 43360 }, { "grad_norm": 0.3370460271835327, "learning_rate": 6.486944533257891e-05, "loss": 0.0488, "step": 43370 }, { "grad_norm": 0.41079118847846985, "learning_rate": 6.485365791043517e-05, "loss": 0.0396, "step": 43380 }, { "grad_norm": 0.5045949220657349, "learning_rate": 6.483786886391772e-05, "loss": 0.0321, "step": 43390 }, { "grad_norm": 0.2903783917427063, "learning_rate": 6.482207819475323e-05, "loss": 0.0301, "step": 43400 }, { "grad_norm": 0.2771800458431244, "learning_rate": 6.480628590466857e-05, "loss": 0.0372, "step": 43410 }, { "grad_norm": 0.43188533186912537, "learning_rate": 6.479049199539074e-05, "loss": 0.0361, "step": 43420 }, { "grad_norm": 0.43213507533073425, "learning_rate": 6.477469646864693e-05, "loss": 0.0359, "step": 43430 }, { "grad_norm": 0.36917582154273987, "learning_rate": 6.475889932616454e-05, "loss": 0.0329, "step": 43440 }, { "grad_norm": 0.34140923619270325, "learning_rate": 6.474310056967111e-05, "loss": 0.0347, "step": 43450 }, { "grad_norm": 0.48464223742485046, "learning_rate": 6.472730020089437e-05, "loss": 0.0286, "step": 43460 }, { "grad_norm": 0.44196823239326477, "learning_rate": 6.471149822156222e-05, "loss": 0.0299, "step": 43470 }, { "grad_norm": 0.3506810963153839, "learning_rate": 6.469569463340275e-05, "loss": 0.0287, "step": 43480 }, { "grad_norm": 0.18974611163139343, "learning_rate": 6.467988943814418e-05, "loss": 0.0297, "step": 43490 }, { "grad_norm": 0.5780832171440125, "learning_rate": 6.4664082637515e-05, "loss": 0.0365, "step": 43500 }, { "grad_norm": 0.3788785934448242, "learning_rate": 6.464827423324377e-05, "loss": 0.0332, "step": 43510 }, { "grad_norm": 0.2718498408794403, "learning_rate": 6.463246422705931e-05, "loss": 0.0303, "step": 43520 }, { "grad_norm": 0.29416006803512573, "learning_rate": 6.461665262069055e-05, "loss": 0.0285, "step": 43530 }, { "grad_norm": 0.2925877571105957, "learning_rate": 6.460083941586665e-05, "loss": 0.0337, "step": 43540 }, { "grad_norm": 0.4991433918476105, "learning_rate": 6.45850246143169e-05, "loss": 0.033, "step": 43550 }, { "grad_norm": 0.9373776912689209, "learning_rate": 6.456920821777077e-05, "loss": 0.0336, "step": 43560 }, { "grad_norm": 0.3004159927368164, "learning_rate": 6.455339022795796e-05, "loss": 0.0289, "step": 43570 }, { "grad_norm": 0.5381854772567749, "learning_rate": 6.453757064660825e-05, "loss": 0.031, "step": 43580 }, { "grad_norm": 0.6213778853416443, "learning_rate": 6.452174947545169e-05, "loss": 0.0266, "step": 43590 }, { "grad_norm": 0.6271640062332153, "learning_rate": 6.450592671621842e-05, "loss": 0.0315, "step": 43600 }, { "grad_norm": 0.26357394456863403, "learning_rate": 6.449010237063882e-05, "loss": 0.0319, "step": 43610 }, { "grad_norm": 0.3266136348247528, "learning_rate": 6.447427644044343e-05, "loss": 0.0343, "step": 43620 }, { "grad_norm": 0.3624243140220642, "learning_rate": 6.44584489273629e-05, "loss": 0.0414, "step": 43630 }, { "grad_norm": 0.5098437666893005, "learning_rate": 6.444261983312817e-05, "loss": 0.0333, "step": 43640 }, { "grad_norm": 0.45966973900794983, "learning_rate": 6.442678915947023e-05, "loss": 0.0364, "step": 43650 }, { "grad_norm": 0.40616488456726074, "learning_rate": 6.441095690812032e-05, "loss": 0.0238, "step": 43660 }, { "grad_norm": 0.33769601583480835, "learning_rate": 6.439512308080982e-05, "loss": 0.0373, "step": 43670 }, { "grad_norm": 0.7465749382972717, "learning_rate": 6.43792876792703e-05, "loss": 0.0334, "step": 43680 }, { "grad_norm": 0.7024111747741699, "learning_rate": 6.43634507052335e-05, "loss": 0.0381, "step": 43690 }, { "grad_norm": 0.39480775594711304, "learning_rate": 6.434761216043135e-05, "loss": 0.0327, "step": 43700 }, { "grad_norm": 0.427047461271286, "learning_rate": 6.433177204659589e-05, "loss": 0.0356, "step": 43710 }, { "grad_norm": 0.5241185426712036, "learning_rate": 6.431593036545938e-05, "loss": 0.0353, "step": 43720 }, { "grad_norm": 0.43305107951164246, "learning_rate": 6.430008711875426e-05, "loss": 0.0392, "step": 43730 }, { "grad_norm": 0.3179013431072235, "learning_rate": 6.42842423082131e-05, "loss": 0.0284, "step": 43740 }, { "grad_norm": 0.3409441411495209, "learning_rate": 6.426839593556868e-05, "loss": 0.0328, "step": 43750 }, { "grad_norm": 0.356086790561676, "learning_rate": 6.425254800255392e-05, "loss": 0.0342, "step": 43760 }, { "grad_norm": 0.46235722303390503, "learning_rate": 6.423669851090196e-05, "loss": 0.0477, "step": 43770 }, { "grad_norm": 0.3611503839492798, "learning_rate": 6.422084746234604e-05, "loss": 0.0323, "step": 43780 }, { "grad_norm": 0.32053402066230774, "learning_rate": 6.420499485861965e-05, "loss": 0.0297, "step": 43790 }, { "grad_norm": 0.8233073949813843, "learning_rate": 6.418914070145636e-05, "loss": 0.0359, "step": 43800 }, { "grad_norm": 0.6226935386657715, "learning_rate": 6.417328499258996e-05, "loss": 0.0346, "step": 43810 }, { "grad_norm": 0.26449066400527954, "learning_rate": 6.415742773375446e-05, "loss": 0.0321, "step": 43820 }, { "grad_norm": 0.28808584809303284, "learning_rate": 6.414156892668393e-05, "loss": 0.0315, "step": 43830 }, { "grad_norm": 0.4023433327674866, "learning_rate": 6.412570857311267e-05, "loss": 0.0286, "step": 43840 }, { "grad_norm": 0.41092967987060547, "learning_rate": 6.410984667477518e-05, "loss": 0.0342, "step": 43850 }, { "grad_norm": 0.3062065839767456, "learning_rate": 6.409398323340607e-05, "loss": 0.0292, "step": 43860 }, { "grad_norm": 0.35637784004211426, "learning_rate": 6.407811825074012e-05, "loss": 0.0284, "step": 43870 }, { "grad_norm": 0.2661358714103699, "learning_rate": 6.406225172851234e-05, "loss": 0.0371, "step": 43880 }, { "grad_norm": 0.5433562397956848, "learning_rate": 6.404638366845786e-05, "loss": 0.0368, "step": 43890 }, { "grad_norm": 0.3222522735595703, "learning_rate": 6.403051407231196e-05, "loss": 0.0308, "step": 43900 }, { "grad_norm": 0.38098612427711487, "learning_rate": 6.401464294181016e-05, "loss": 0.0339, "step": 43910 }, { "grad_norm": 0.33502283692359924, "learning_rate": 6.399877027868808e-05, "loss": 0.0343, "step": 43920 }, { "grad_norm": 0.39182838797569275, "learning_rate": 6.39828960846815e-05, "loss": 0.0398, "step": 43930 }, { "grad_norm": 0.36787036061286926, "learning_rate": 6.396702036152644e-05, "loss": 0.0309, "step": 43940 }, { "grad_norm": 0.29357653856277466, "learning_rate": 6.395114311095905e-05, "loss": 0.044, "step": 43950 }, { "grad_norm": 0.29129883646965027, "learning_rate": 6.393526433471559e-05, "loss": 0.033, "step": 43960 }, { "grad_norm": 0.7632303237915039, "learning_rate": 6.391938403453262e-05, "loss": 0.0425, "step": 43970 }, { "grad_norm": 0.45855188369750977, "learning_rate": 6.390350221214671e-05, "loss": 0.0377, "step": 43980 }, { "grad_norm": 0.3501278758049011, "learning_rate": 6.38876188692947e-05, "loss": 0.0474, "step": 43990 }, { "grad_norm": 0.503690242767334, "learning_rate": 6.387173400771358e-05, "loss": 0.039, "step": 44000 }, { "grad_norm": 0.4948585033416748, "learning_rate": 6.385584762914047e-05, "loss": 0.0331, "step": 44010 }, { "grad_norm": 0.26526013016700745, "learning_rate": 6.38399597353127e-05, "loss": 0.04, "step": 44020 }, { "grad_norm": 0.28100988268852234, "learning_rate": 6.382407032796775e-05, "loss": 0.035, "step": 44030 }, { "grad_norm": 0.6551303863525391, "learning_rate": 6.380817940884325e-05, "loss": 0.0363, "step": 44040 }, { "grad_norm": 0.4188297688961029, "learning_rate": 6.379228697967702e-05, "loss": 0.0403, "step": 44050 }, { "grad_norm": 0.3264245390892029, "learning_rate": 6.3776393042207e-05, "loss": 0.0338, "step": 44060 }, { "grad_norm": 0.431392639875412, "learning_rate": 6.376049759817137e-05, "loss": 0.0268, "step": 44070 }, { "grad_norm": 0.2514946460723877, "learning_rate": 6.37446006493084e-05, "loss": 0.0347, "step": 44080 }, { "grad_norm": 0.39275380969047546, "learning_rate": 6.372870219735656e-05, "loss": 0.0378, "step": 44090 }, { "grad_norm": 0.2853188216686249, "learning_rate": 6.371280224405452e-05, "loss": 0.0404, "step": 44100 }, { "grad_norm": 0.4299790859222412, "learning_rate": 6.369690079114101e-05, "loss": 0.0334, "step": 44110 }, { "grad_norm": 0.504658579826355, "learning_rate": 6.368099784035504e-05, "loss": 0.0368, "step": 44120 }, { "grad_norm": 0.4738781452178955, "learning_rate": 6.366509339343572e-05, "loss": 0.0349, "step": 44130 }, { "grad_norm": 0.3871033787727356, "learning_rate": 6.364918745212232e-05, "loss": 0.0276, "step": 44140 }, { "grad_norm": 0.5049381852149963, "learning_rate": 6.363328001815431e-05, "loss": 0.03, "step": 44150 }, { "grad_norm": 0.23147010803222656, "learning_rate": 6.361737109327128e-05, "loss": 0.0296, "step": 44160 }, { "grad_norm": 0.30930325388908386, "learning_rate": 6.360146067921301e-05, "loss": 0.0298, "step": 44170 }, { "grad_norm": 0.32318270206451416, "learning_rate": 6.35855487777195e-05, "loss": 0.0349, "step": 44180 }, { "grad_norm": 0.3712194263935089, "learning_rate": 6.356963539053076e-05, "loss": 0.0406, "step": 44190 }, { "grad_norm": 0.4761961102485657, "learning_rate": 6.355372051938711e-05, "loss": 0.0361, "step": 44200 }, { "grad_norm": 0.4590695798397064, "learning_rate": 6.353780416602894e-05, "loss": 0.0313, "step": 44210 }, { "grad_norm": 0.30450427532196045, "learning_rate": 6.352188633219689e-05, "loss": 0.0291, "step": 44220 }, { "grad_norm": 0.47648730874061584, "learning_rate": 6.350596701963166e-05, "loss": 0.0336, "step": 44230 }, { "grad_norm": 0.2321753352880478, "learning_rate": 6.349004623007419e-05, "loss": 0.0309, "step": 44240 }, { "grad_norm": 0.491190642118454, "learning_rate": 6.347412396526555e-05, "loss": 0.0364, "step": 44250 }, { "grad_norm": 0.3543434143066406, "learning_rate": 6.345820022694696e-05, "loss": 0.031, "step": 44260 }, { "grad_norm": 0.45984265208244324, "learning_rate": 6.344227501685984e-05, "loss": 0.0309, "step": 44270 }, { "grad_norm": 0.450426310300827, "learning_rate": 6.342634833674572e-05, "loss": 0.0327, "step": 44280 }, { "grad_norm": 0.4361113905906677, "learning_rate": 6.341042018834635e-05, "loss": 0.0386, "step": 44290 }, { "grad_norm": 0.2741334140300751, "learning_rate": 6.339449057340359e-05, "loss": 0.0382, "step": 44300 }, { "grad_norm": 0.3734683394432068, "learning_rate": 6.337855949365945e-05, "loss": 0.032, "step": 44310 }, { "grad_norm": 0.40710344910621643, "learning_rate": 6.336262695085619e-05, "loss": 0.0297, "step": 44320 }, { "grad_norm": 0.519195020198822, "learning_rate": 6.334669294673612e-05, "loss": 0.0355, "step": 44330 }, { "grad_norm": 0.3176940679550171, "learning_rate": 6.333075748304179e-05, "loss": 0.0366, "step": 44340 }, { "grad_norm": 0.3302747309207916, "learning_rate": 6.331482056151585e-05, "loss": 0.0432, "step": 44350 }, { "grad_norm": 0.8641408085823059, "learning_rate": 6.329888218390117e-05, "loss": 0.0332, "step": 44360 }, { "grad_norm": 0.61134272813797, "learning_rate": 6.328294235194072e-05, "loss": 0.0302, "step": 44370 }, { "grad_norm": 0.4974342882633209, "learning_rate": 6.326700106737765e-05, "loss": 0.0344, "step": 44380 }, { "grad_norm": 0.24751196801662445, "learning_rate": 6.325105833195531e-05, "loss": 0.0264, "step": 44390 }, { "grad_norm": 0.391739159822464, "learning_rate": 6.323511414741715e-05, "loss": 0.0383, "step": 44400 }, { "grad_norm": 0.6418096423149109, "learning_rate": 6.321916851550678e-05, "loss": 0.0432, "step": 44410 }, { "grad_norm": 0.45718836784362793, "learning_rate": 6.320322143796806e-05, "loss": 0.0455, "step": 44420 }, { "grad_norm": 0.2879917621612549, "learning_rate": 6.318727291654488e-05, "loss": 0.034, "step": 44430 }, { "grad_norm": 0.2615237236022949, "learning_rate": 6.317132295298134e-05, "loss": 0.034, "step": 44440 }, { "grad_norm": 1.00301992893219, "learning_rate": 6.315537154902173e-05, "loss": 0.0316, "step": 44450 }, { "grad_norm": 0.4758610725402832, "learning_rate": 6.313941870641048e-05, "loss": 0.0329, "step": 44460 }, { "grad_norm": 0.31256037950515747, "learning_rate": 6.312346442689214e-05, "loss": 0.0404, "step": 44470 }, { "grad_norm": 0.5565505027770996, "learning_rate": 6.310750871221147e-05, "loss": 0.0371, "step": 44480 }, { "grad_norm": 0.4073399305343628, "learning_rate": 6.309155156411335e-05, "loss": 0.0321, "step": 44490 }, { "grad_norm": 0.5484876036643982, "learning_rate": 6.307559298434284e-05, "loss": 0.0429, "step": 44500 }, { "grad_norm": 0.42776724696159363, "learning_rate": 6.305963297464512e-05, "loss": 0.0377, "step": 44510 }, { "grad_norm": 0.3458835184574127, "learning_rate": 6.304367153676561e-05, "loss": 0.037, "step": 44520 }, { "grad_norm": 0.42405301332473755, "learning_rate": 6.302770867244978e-05, "loss": 0.0389, "step": 44530 }, { "grad_norm": 0.28612399101257324, "learning_rate": 6.301174438344328e-05, "loss": 0.0327, "step": 44540 }, { "grad_norm": 0.3355843424797058, "learning_rate": 6.299577867149202e-05, "loss": 0.0423, "step": 44550 }, { "grad_norm": 0.3467482328414917, "learning_rate": 6.297981153834192e-05, "loss": 0.033, "step": 44560 }, { "grad_norm": 0.3891068994998932, "learning_rate": 6.296384298573916e-05, "loss": 0.0398, "step": 44570 }, { "grad_norm": 0.4009205102920532, "learning_rate": 6.294787301543001e-05, "loss": 0.0471, "step": 44580 }, { "grad_norm": 0.3143903613090515, "learning_rate": 6.293190162916095e-05, "loss": 0.0284, "step": 44590 }, { "grad_norm": 0.5415225028991699, "learning_rate": 6.291592882867855e-05, "loss": 0.0388, "step": 44600 }, { "grad_norm": 0.4116606116294861, "learning_rate": 6.28999546157296e-05, "loss": 0.0345, "step": 44610 }, { "grad_norm": 0.47534018754959106, "learning_rate": 6.288397899206102e-05, "loss": 0.0333, "step": 44620 }, { "grad_norm": 0.4800972044467926, "learning_rate": 6.286800195941984e-05, "loss": 0.0432, "step": 44630 }, { "grad_norm": 0.4507230818271637, "learning_rate": 6.285202351955334e-05, "loss": 0.0463, "step": 44640 }, { "grad_norm": 0.5180692076683044, "learning_rate": 6.283604367420887e-05, "loss": 0.0362, "step": 44650 }, { "grad_norm": 0.5743687748908997, "learning_rate": 6.282006242513394e-05, "loss": 0.0417, "step": 44660 }, { "grad_norm": 0.5114428997039795, "learning_rate": 6.280407977407628e-05, "loss": 0.0392, "step": 44670 }, { "grad_norm": 0.4131477177143097, "learning_rate": 6.27880957227837e-05, "loss": 0.0392, "step": 44680 }, { "grad_norm": 0.3796303868293762, "learning_rate": 6.27721102730042e-05, "loss": 0.0324, "step": 44690 }, { "grad_norm": 0.42333880066871643, "learning_rate": 6.275612342648592e-05, "loss": 0.0364, "step": 44700 }, { "grad_norm": 0.2867392897605896, "learning_rate": 6.274013518497716e-05, "loss": 0.0296, "step": 44710 }, { "grad_norm": 1.0788853168487549, "learning_rate": 6.272414555022636e-05, "loss": 0.0289, "step": 44720 }, { "grad_norm": 0.36042582988739014, "learning_rate": 6.270815452398215e-05, "loss": 0.0315, "step": 44730 }, { "grad_norm": 0.29202741384506226, "learning_rate": 6.269216210799326e-05, "loss": 0.0333, "step": 44740 }, { "grad_norm": 0.34146973490715027, "learning_rate": 6.26761683040086e-05, "loss": 0.0453, "step": 44750 }, { "grad_norm": 0.4982683062553406, "learning_rate": 6.266017311377723e-05, "loss": 0.046, "step": 44760 }, { "grad_norm": 0.3656432032585144, "learning_rate": 6.264417653904839e-05, "loss": 0.036, "step": 44770 }, { "grad_norm": 0.23815134167671204, "learning_rate": 6.262817858157139e-05, "loss": 0.0299, "step": 44780 }, { "grad_norm": 0.31278786063194275, "learning_rate": 6.261217924309576e-05, "loss": 0.0407, "step": 44790 }, { "grad_norm": 0.631803572177887, "learning_rate": 6.259617852537118e-05, "loss": 0.0295, "step": 44800 }, { "grad_norm": 0.47832849621772766, "learning_rate": 6.258017643014747e-05, "loss": 0.039, "step": 44810 }, { "grad_norm": 0.3869171738624573, "learning_rate": 6.256417295917456e-05, "loss": 0.0403, "step": 44820 }, { "grad_norm": 0.47458022832870483, "learning_rate": 6.254816811420258e-05, "loss": 0.0391, "step": 44830 }, { "grad_norm": 0.3862384855747223, "learning_rate": 6.253216189698183e-05, "loss": 0.0333, "step": 44840 }, { "grad_norm": 0.2769148647785187, "learning_rate": 6.251615430926267e-05, "loss": 0.0286, "step": 44850 }, { "grad_norm": 0.41007912158966064, "learning_rate": 6.25001453527957e-05, "loss": 0.0375, "step": 44860 }, { "grad_norm": 0.3517419397830963, "learning_rate": 6.248413502933164e-05, "loss": 0.0382, "step": 44870 }, { "grad_norm": 0.4499068558216095, "learning_rate": 6.246812334062133e-05, "loss": 0.0452, "step": 44880 }, { "grad_norm": 0.5372024178504944, "learning_rate": 6.245211028841579e-05, "loss": 0.0385, "step": 44890 }, { "grad_norm": 0.375620573759079, "learning_rate": 6.24360958744662e-05, "loss": 0.0311, "step": 44900 }, { "grad_norm": 0.215256929397583, "learning_rate": 6.242008010052387e-05, "loss": 0.0339, "step": 44910 }, { "grad_norm": 0.3658120930194855, "learning_rate": 6.240406296834024e-05, "loss": 0.0452, "step": 44920 }, { "grad_norm": 0.2559692859649658, "learning_rate": 6.238804447966694e-05, "loss": 0.0272, "step": 44930 }, { "grad_norm": 0.3491780757904053, "learning_rate": 6.237202463625573e-05, "loss": 0.0303, "step": 44940 }, { "grad_norm": 0.4053493142127991, "learning_rate": 6.235600343985848e-05, "loss": 0.0347, "step": 44950 }, { "grad_norm": 0.2561838924884796, "learning_rate": 6.233998089222729e-05, "loss": 0.0312, "step": 44960 }, { "grad_norm": 0.3581414818763733, "learning_rate": 6.232395699511433e-05, "loss": 0.043, "step": 44970 }, { "grad_norm": 0.31345388293266296, "learning_rate": 6.230793175027199e-05, "loss": 0.0342, "step": 44980 }, { "grad_norm": 0.38085857033729553, "learning_rate": 6.22919051594527e-05, "loss": 0.0263, "step": 44990 }, { "grad_norm": 0.3010728061199188, "learning_rate": 6.227587722440917e-05, "loss": 0.0385, "step": 45000 }, { "grad_norm": 0.3442159593105316, "learning_rate": 6.225984794689414e-05, "loss": 0.0335, "step": 45010 }, { "grad_norm": 0.34080490469932556, "learning_rate": 6.22438173286606e-05, "loss": 0.0393, "step": 45020 }, { "grad_norm": 0.43221837282180786, "learning_rate": 6.222778537146159e-05, "loss": 0.0379, "step": 45030 }, { "grad_norm": 0.7012179493904114, "learning_rate": 6.221175207705037e-05, "loss": 0.0396, "step": 45040 }, { "grad_norm": 0.37922027707099915, "learning_rate": 6.219571744718027e-05, "loss": 0.0308, "step": 45050 }, { "grad_norm": 0.8075982332229614, "learning_rate": 6.217968148360487e-05, "loss": 0.04, "step": 45060 }, { "grad_norm": 0.2975711524486542, "learning_rate": 6.216364418807782e-05, "loss": 0.0326, "step": 45070 }, { "grad_norm": 0.5645615458488464, "learning_rate": 6.214760556235292e-05, "loss": 0.0364, "step": 45080 }, { "grad_norm": 0.3290588855743408, "learning_rate": 6.213156560818414e-05, "loss": 0.0299, "step": 45090 }, { "grad_norm": 0.4168420433998108, "learning_rate": 6.211552432732559e-05, "loss": 0.0401, "step": 45100 }, { "grad_norm": 0.2184748500585556, "learning_rate": 6.20994817215315e-05, "loss": 0.0348, "step": 45110 }, { "grad_norm": 0.30656468868255615, "learning_rate": 6.208343779255628e-05, "loss": 0.0337, "step": 45120 }, { "grad_norm": 0.3918287754058838, "learning_rate": 6.206739254215449e-05, "loss": 0.0312, "step": 45130 }, { "grad_norm": 0.23824937641620636, "learning_rate": 6.205134597208077e-05, "loss": 0.0237, "step": 45140 }, { "grad_norm": 0.4854699671268463, "learning_rate": 6.203529808409e-05, "loss": 0.0404, "step": 45150 }, { "grad_norm": 0.2964397966861725, "learning_rate": 6.201924887993712e-05, "loss": 0.0479, "step": 45160 }, { "grad_norm": 0.40493130683898926, "learning_rate": 6.200319836137725e-05, "loss": 0.0365, "step": 45170 }, { "grad_norm": 0.22493614256381989, "learning_rate": 6.198714653016565e-05, "loss": 0.0324, "step": 45180 }, { "grad_norm": 0.30707651376724243, "learning_rate": 6.197109338805774e-05, "loss": 0.0368, "step": 45190 }, { "grad_norm": 0.25042134523391724, "learning_rate": 6.195503893680903e-05, "loss": 0.0381, "step": 45200 }, { "grad_norm": 0.3370530903339386, "learning_rate": 6.193898317817524e-05, "loss": 0.0311, "step": 45210 }, { "grad_norm": 0.22446751594543457, "learning_rate": 6.192292611391221e-05, "loss": 0.0267, "step": 45220 }, { "grad_norm": 0.43748193979263306, "learning_rate": 6.190686774577591e-05, "loss": 0.0412, "step": 45230 }, { "grad_norm": 0.20210465788841248, "learning_rate": 6.189080807552245e-05, "loss": 0.0454, "step": 45240 }, { "grad_norm": 0.2958391606807709, "learning_rate": 6.187474710490809e-05, "loss": 0.0264, "step": 45250 }, { "grad_norm": 0.29451367259025574, "learning_rate": 6.185868483568926e-05, "loss": 0.0271, "step": 45260 }, { "grad_norm": 0.26410147547721863, "learning_rate": 6.184262126962245e-05, "loss": 0.0318, "step": 45270 }, { "grad_norm": 0.446811705827713, "learning_rate": 6.182655640846442e-05, "loss": 0.0327, "step": 45280 }, { "grad_norm": 0.2802617847919464, "learning_rate": 6.181049025397196e-05, "loss": 0.033, "step": 45290 }, { "grad_norm": 0.3032892942428589, "learning_rate": 6.179442280790202e-05, "loss": 0.036, "step": 45300 }, { "grad_norm": 0.5358495116233826, "learning_rate": 6.177835407201174e-05, "loss": 0.0386, "step": 45310 }, { "grad_norm": 0.2810453772544861, "learning_rate": 6.176228404805839e-05, "loss": 0.0325, "step": 45320 }, { "grad_norm": 0.4235498905181885, "learning_rate": 6.174621273779932e-05, "loss": 0.0304, "step": 45330 }, { "grad_norm": 0.4771668314933777, "learning_rate": 6.17301401429921e-05, "loss": 0.0344, "step": 45340 }, { "grad_norm": 0.23973824083805084, "learning_rate": 6.171406626539436e-05, "loss": 0.031, "step": 45350 }, { "grad_norm": 0.5901950001716614, "learning_rate": 6.169799110676398e-05, "loss": 0.0445, "step": 45360 }, { "grad_norm": 0.46453240513801575, "learning_rate": 6.168191466885885e-05, "loss": 0.026, "step": 45370 }, { "grad_norm": 0.32192373275756836, "learning_rate": 6.166583695343711e-05, "loss": 0.0316, "step": 45380 }, { "grad_norm": 0.451629102230072, "learning_rate": 6.164975796225698e-05, "loss": 0.0372, "step": 45390 }, { "grad_norm": 0.314796507358551, "learning_rate": 6.163367769707683e-05, "loss": 0.0276, "step": 45400 }, { "grad_norm": 0.3362596333026886, "learning_rate": 6.161759615965519e-05, "loss": 0.0388, "step": 45410 }, { "grad_norm": 0.27245065569877625, "learning_rate": 6.16015133517507e-05, "loss": 0.034, "step": 45420 }, { "grad_norm": 0.2749735414981842, "learning_rate": 6.158542927512214e-05, "loss": 0.0347, "step": 45430 }, { "grad_norm": 0.38173162937164307, "learning_rate": 6.156934393152846e-05, "loss": 0.032, "step": 45440 }, { "grad_norm": 0.33989813923835754, "learning_rate": 6.15532573227287e-05, "loss": 0.0386, "step": 45450 }, { "grad_norm": 0.4084349274635315, "learning_rate": 6.153716945048212e-05, "loss": 0.0444, "step": 45460 }, { "grad_norm": 0.40381762385368347, "learning_rate": 6.152108031654802e-05, "loss": 0.0471, "step": 45470 }, { "grad_norm": 0.45853888988494873, "learning_rate": 6.15049899226859e-05, "loss": 0.033, "step": 45480 }, { "grad_norm": 0.41022056341171265, "learning_rate": 6.148889827065537e-05, "loss": 0.0411, "step": 45490 }, { "grad_norm": 0.6332377195358276, "learning_rate": 6.147280536221622e-05, "loss": 0.0366, "step": 45500 }, { "grad_norm": 0.6025694012641907, "learning_rate": 6.145671119912832e-05, "loss": 0.0361, "step": 45510 }, { "grad_norm": 0.406787246465683, "learning_rate": 6.144061578315169e-05, "loss": 0.0359, "step": 45520 }, { "grad_norm": 0.3376815915107727, "learning_rate": 6.142451911604654e-05, "loss": 0.028, "step": 45530 }, { "grad_norm": 0.3299280107021332, "learning_rate": 6.140842119957315e-05, "loss": 0.025, "step": 45540 }, { "grad_norm": 0.40781816840171814, "learning_rate": 6.139232203549197e-05, "loss": 0.0309, "step": 45550 }, { "grad_norm": 0.3303791880607605, "learning_rate": 6.137622162556357e-05, "loss": 0.033, "step": 45560 }, { "grad_norm": 0.3951131999492645, "learning_rate": 6.13601199715487e-05, "loss": 0.0312, "step": 45570 }, { "grad_norm": 0.300718754529953, "learning_rate": 6.134401707520816e-05, "loss": 0.0351, "step": 45580 }, { "grad_norm": 0.3159123361110687, "learning_rate": 6.132791293830298e-05, "loss": 0.0235, "step": 45590 }, { "grad_norm": 0.3451389670372009, "learning_rate": 6.131180756259428e-05, "loss": 0.0325, "step": 45600 }, { "grad_norm": 0.3523235023021698, "learning_rate": 6.129570094984331e-05, "loss": 0.0331, "step": 45610 }, { "grad_norm": 0.8100460767745972, "learning_rate": 6.127959310181145e-05, "loss": 0.0482, "step": 45620 }, { "grad_norm": 0.3161967396736145, "learning_rate": 6.126348402026026e-05, "loss": 0.031, "step": 45630 }, { "grad_norm": 0.31200528144836426, "learning_rate": 6.12473737069514e-05, "loss": 0.0367, "step": 45640 }, { "grad_norm": 0.30418023467063904, "learning_rate": 6.123126216364665e-05, "loss": 0.0323, "step": 45650 }, { "grad_norm": 0.4051944613456726, "learning_rate": 6.121514939210797e-05, "loss": 0.0331, "step": 45660 }, { "grad_norm": 0.3808932304382324, "learning_rate": 6.119903539409741e-05, "loss": 0.033, "step": 45670 }, { "grad_norm": 0.2443174421787262, "learning_rate": 6.118292017137716e-05, "loss": 0.0386, "step": 45680 }, { "grad_norm": 0.3851858377456665, "learning_rate": 6.116680372570959e-05, "loss": 0.0275, "step": 45690 }, { "grad_norm": 0.4625587463378906, "learning_rate": 6.115068605885713e-05, "loss": 0.0317, "step": 45700 }, { "grad_norm": 0.24865679442882538, "learning_rate": 6.113456717258243e-05, "loss": 0.032, "step": 45710 }, { "grad_norm": 0.41794803738594055, "learning_rate": 6.11184470686482e-05, "loss": 0.0305, "step": 45720 }, { "grad_norm": 0.2985776364803314, "learning_rate": 6.110232574881731e-05, "loss": 0.024, "step": 45730 }, { "grad_norm": 0.4646579921245575, "learning_rate": 6.108620321485277e-05, "loss": 0.0357, "step": 45740 }, { "grad_norm": 0.5313912630081177, "learning_rate": 6.107007946851773e-05, "loss": 0.0337, "step": 45750 }, { "grad_norm": 0.3384891748428345, "learning_rate": 6.105395451157542e-05, "loss": 0.044, "step": 45760 }, { "grad_norm": 0.40896010398864746, "learning_rate": 6.103782834578928e-05, "loss": 0.0328, "step": 45770 }, { "grad_norm": 0.3500524163246155, "learning_rate": 6.102170097292281e-05, "loss": 0.0292, "step": 45780 }, { "grad_norm": 0.4050636887550354, "learning_rate": 6.1005572394739716e-05, "loss": 0.0394, "step": 45790 }, { "grad_norm": 0.3658168613910675, "learning_rate": 6.0989442613003765e-05, "loss": 0.0409, "step": 45800 }, { "grad_norm": 0.43675169348716736, "learning_rate": 6.097331162947888e-05, "loss": 0.0404, "step": 45810 }, { "grad_norm": 0.5707777738571167, "learning_rate": 6.095717944592914e-05, "loss": 0.0306, "step": 45820 }, { "grad_norm": 0.25129029154777527, "learning_rate": 6.094104606411873e-05, "loss": 0.0503, "step": 45830 }, { "grad_norm": 0.3782576322555542, "learning_rate": 6.0924911485811966e-05, "loss": 0.0459, "step": 45840 }, { "grad_norm": 0.37379273772239685, "learning_rate": 6.090877571277331e-05, "loss": 0.0338, "step": 45850 }, { "grad_norm": 0.4243023097515106, "learning_rate": 6.089263874676736e-05, "loss": 0.0329, "step": 45860 }, { "grad_norm": 0.5262272357940674, "learning_rate": 6.0876500589558796e-05, "loss": 0.0371, "step": 45870 }, { "grad_norm": 0.4498524069786072, "learning_rate": 6.086036124291248e-05, "loss": 0.032, "step": 45880 }, { "grad_norm": 0.3317078948020935, "learning_rate": 6.084422070859339e-05, "loss": 0.0434, "step": 45890 }, { "grad_norm": 0.3438365161418915, "learning_rate": 6.082807898836663e-05, "loss": 0.0291, "step": 45900 }, { "grad_norm": 0.4511534571647644, "learning_rate": 6.081193608399742e-05, "loss": 0.0232, "step": 45910 }, { "grad_norm": 0.28290626406669617, "learning_rate": 6.0795791997251164e-05, "loss": 0.0249, "step": 45920 }, { "grad_norm": 0.33926936984062195, "learning_rate": 6.0779646729893294e-05, "loss": 0.0283, "step": 45930 }, { "grad_norm": 0.32470691204071045, "learning_rate": 6.0763500283689476e-05, "loss": 0.0257, "step": 45940 }, { "grad_norm": 0.41196608543395996, "learning_rate": 6.0747352660405455e-05, "loss": 0.0313, "step": 45950 }, { "grad_norm": 0.3812471032142639, "learning_rate": 6.073120386180709e-05, "loss": 0.031, "step": 45960 }, { "grad_norm": 0.37449613213539124, "learning_rate": 6.0715053889660425e-05, "loss": 0.0369, "step": 45970 }, { "grad_norm": 0.3375701308250427, "learning_rate": 6.069890274573157e-05, "loss": 0.0316, "step": 45980 }, { "grad_norm": 0.966271698474884, "learning_rate": 6.068275043178679e-05, "loss": 0.0416, "step": 45990 }, { "grad_norm": 0.3867020010948181, "learning_rate": 6.066659694959248e-05, "loss": 0.0432, "step": 46000 }, { "grad_norm": 0.27092641592025757, "learning_rate": 6.0650442300915176e-05, "loss": 0.0278, "step": 46010 }, { "grad_norm": 0.2991166114807129, "learning_rate": 6.063428648752152e-05, "loss": 0.032, "step": 46020 }, { "grad_norm": 0.41209670901298523, "learning_rate": 6.0618129511178266e-05, "loss": 0.0303, "step": 46030 }, { "grad_norm": 0.4886704385280609, "learning_rate": 6.060197137365234e-05, "loss": 0.0349, "step": 46040 }, { "grad_norm": 0.5686153173446655, "learning_rate": 6.058581207671077e-05, "loss": 0.0289, "step": 46050 }, { "grad_norm": 0.31571489572525024, "learning_rate": 6.056965162212072e-05, "loss": 0.0279, "step": 46060 }, { "grad_norm": 0.34952738881111145, "learning_rate": 6.055349001164943e-05, "loss": 0.0341, "step": 46070 }, { "grad_norm": 0.41830939054489136, "learning_rate": 6.0537327247064347e-05, "loss": 0.0349, "step": 46080 }, { "grad_norm": 0.6707746386528015, "learning_rate": 6.0521163330133014e-05, "loss": 0.0439, "step": 46090 }, { "grad_norm": 0.3615257740020752, "learning_rate": 6.050499826262306e-05, "loss": 0.0289, "step": 46100 }, { "grad_norm": 0.4513275921344757, "learning_rate": 6.0488832046302294e-05, "loss": 0.0292, "step": 46110 }, { "grad_norm": 0.25569310784339905, "learning_rate": 6.0472664682938626e-05, "loss": 0.0305, "step": 46120 }, { "grad_norm": 0.21833555400371552, "learning_rate": 6.045649617430009e-05, "loss": 0.0285, "step": 46130 }, { "grad_norm": 0.4960653781890869, "learning_rate": 6.0440326522154866e-05, "loss": 0.0304, "step": 46140 }, { "grad_norm": 0.565873384475708, "learning_rate": 6.0424155728271224e-05, "loss": 0.0398, "step": 46150 }, { "grad_norm": 0.5752214789390564, "learning_rate": 6.040798379441758e-05, "loss": 0.0319, "step": 46160 }, { "grad_norm": 0.30662018060684204, "learning_rate": 6.0391810722362485e-05, "loss": 0.0395, "step": 46170 }, { "grad_norm": 0.4778442680835724, "learning_rate": 6.037563651387458e-05, "loss": 0.0341, "step": 46180 }, { "grad_norm": 0.40740543603897095, "learning_rate": 6.0359461170722666e-05, "loss": 0.0341, "step": 46190 }, { "grad_norm": 0.39384350180625916, "learning_rate": 6.034328469467566e-05, "loss": 0.0404, "step": 46200 }, { "grad_norm": 0.4196702837944031, "learning_rate": 6.0327107087502596e-05, "loss": 0.0387, "step": 46210 }, { "grad_norm": 0.4656120836734772, "learning_rate": 6.031092835097262e-05, "loss": 0.0327, "step": 46220 }, { "grad_norm": 0.3460751175880432, "learning_rate": 6.0294748486855024e-05, "loss": 0.0314, "step": 46230 }, { "grad_norm": 0.31983643770217896, "learning_rate": 6.0278567496919216e-05, "loss": 0.0289, "step": 46240 }, { "grad_norm": 0.4817160964012146, "learning_rate": 6.026238538293472e-05, "loss": 0.0376, "step": 46250 }, { "grad_norm": 0.4624084532260895, "learning_rate": 6.024620214667118e-05, "loss": 0.0322, "step": 46260 }, { "grad_norm": 0.276117742061615, "learning_rate": 6.0230017789898384e-05, "loss": 0.0303, "step": 46270 }, { "grad_norm": 0.34058764576911926, "learning_rate": 6.021383231438622e-05, "loss": 0.0275, "step": 46280 }, { "grad_norm": 0.3300950229167938, "learning_rate": 6.0197645721904704e-05, "loss": 0.0309, "step": 46290 }, { "grad_norm": 0.3414401113986969, "learning_rate": 6.0181458014224e-05, "loss": 0.0362, "step": 46300 }, { "grad_norm": 0.5223149657249451, "learning_rate": 6.016526919311435e-05, "loss": 0.031, "step": 46310 }, { "grad_norm": 0.39035046100616455, "learning_rate": 6.014907926034613e-05, "loss": 0.0303, "step": 46320 }, { "grad_norm": 0.2654147148132324, "learning_rate": 6.0132888217689875e-05, "loss": 0.0319, "step": 46330 }, { "grad_norm": 0.4692043662071228, "learning_rate": 6.01166960669162e-05, "loss": 0.0375, "step": 46340 }, { "grad_norm": 0.25673413276672363, "learning_rate": 6.0100502809795845e-05, "loss": 0.0292, "step": 46350 }, { "grad_norm": 0.30907562375068665, "learning_rate": 6.00843084480997e-05, "loss": 0.0355, "step": 46360 }, { "grad_norm": 0.36672645807266235, "learning_rate": 6.0068112983598736e-05, "loss": 0.0319, "step": 46370 }, { "grad_norm": 0.28177425265312195, "learning_rate": 6.005191641806407e-05, "loss": 0.0263, "step": 46380 }, { "grad_norm": 0.4871598184108734, "learning_rate": 6.003571875326694e-05, "loss": 0.0367, "step": 46390 }, { "grad_norm": 0.28324973583221436, "learning_rate": 6.00195199909787e-05, "loss": 0.0277, "step": 46400 }, { "grad_norm": 0.31881117820739746, "learning_rate": 6.000332013297082e-05, "loss": 0.0343, "step": 46410 }, { "grad_norm": 0.5380931496620178, "learning_rate": 5.998711918101487e-05, "loss": 0.0353, "step": 46420 }, { "grad_norm": 0.33785778284072876, "learning_rate": 5.997091713688261e-05, "loss": 0.0268, "step": 46430 }, { "grad_norm": 0.41628047823905945, "learning_rate": 5.9954714002345836e-05, "loss": 0.0377, "step": 46440 }, { "grad_norm": 0.31284603476524353, "learning_rate": 5.993850977917649e-05, "loss": 0.034, "step": 46450 }, { "grad_norm": 0.3728019893169403, "learning_rate": 5.992230446914667e-05, "loss": 0.0379, "step": 46460 }, { "grad_norm": 0.37368345260620117, "learning_rate": 5.990609807402855e-05, "loss": 0.0355, "step": 46470 }, { "grad_norm": 0.22661447525024414, "learning_rate": 5.988989059559443e-05, "loss": 0.03, "step": 46480 }, { "grad_norm": 0.2834111154079437, "learning_rate": 5.987368203561675e-05, "loss": 0.0293, "step": 46490 }, { "grad_norm": 0.24011904001235962, "learning_rate": 5.9857472395868055e-05, "loss": 0.0268, "step": 46500 }, { "grad_norm": 0.3154459297657013, "learning_rate": 5.9841261678120983e-05, "loss": 0.0327, "step": 46510 }, { "grad_norm": 0.2971125841140747, "learning_rate": 5.982504988414834e-05, "loss": 0.0238, "step": 46520 }, { "grad_norm": 0.3838680386543274, "learning_rate": 5.9808837015723015e-05, "loss": 0.0462, "step": 46530 }, { "grad_norm": 0.36676159501075745, "learning_rate": 5.9792623074618016e-05, "loss": 0.0308, "step": 46540 }, { "grad_norm": 0.30691277980804443, "learning_rate": 5.977640806260648e-05, "loss": 0.0295, "step": 46550 }, { "grad_norm": 0.6611349582672119, "learning_rate": 5.976019198146166e-05, "loss": 0.0405, "step": 46560 }, { "grad_norm": 0.5587453842163086, "learning_rate": 5.974397483295692e-05, "loss": 0.0238, "step": 46570 }, { "grad_norm": 0.41938871145248413, "learning_rate": 5.972775661886575e-05, "loss": 0.0337, "step": 46580 }, { "grad_norm": 0.24318458139896393, "learning_rate": 5.971153734096173e-05, "loss": 0.0246, "step": 46590 }, { "grad_norm": 0.4362047016620636, "learning_rate": 5.969531700101859e-05, "loss": 0.0299, "step": 46600 }, { "grad_norm": 0.4125433564186096, "learning_rate": 5.9679095600810155e-05, "loss": 0.0373, "step": 46610 }, { "grad_norm": 0.7688597440719604, "learning_rate": 5.9662873142110384e-05, "loss": 0.0295, "step": 46620 }, { "grad_norm": 0.2110181450843811, "learning_rate": 5.964664962669333e-05, "loss": 0.0301, "step": 46630 }, { "grad_norm": 0.24672719836235046, "learning_rate": 5.9630425056333186e-05, "loss": 0.0355, "step": 46640 }, { "grad_norm": 0.3493923246860504, "learning_rate": 5.961419943280422e-05, "loss": 0.0377, "step": 46650 }, { "grad_norm": 0.46097859740257263, "learning_rate": 5.959797275788087e-05, "loss": 0.0314, "step": 46660 }, { "grad_norm": 0.367875337600708, "learning_rate": 5.958174503333765e-05, "loss": 0.0367, "step": 46670 }, { "grad_norm": 0.35661250352859497, "learning_rate": 5.9565516260949195e-05, "loss": 0.0338, "step": 46680 }, { "grad_norm": 0.2944066822528839, "learning_rate": 5.9549286442490273e-05, "loss": 0.0282, "step": 46690 }, { "grad_norm": 0.34269437193870544, "learning_rate": 5.953305557973572e-05, "loss": 0.0346, "step": 46700 }, { "grad_norm": 0.3804711401462555, "learning_rate": 5.9516823674460564e-05, "loss": 0.0361, "step": 46710 }, { "grad_norm": 0.4018082320690155, "learning_rate": 5.9500590728439875e-05, "loss": 0.0278, "step": 46720 }, { "grad_norm": 0.3087543547153473, "learning_rate": 5.948435674344886e-05, "loss": 0.0371, "step": 46730 }, { "grad_norm": 0.26842001080513, "learning_rate": 5.946812172126285e-05, "loss": 0.0214, "step": 46740 }, { "grad_norm": 0.9276534914970398, "learning_rate": 5.945188566365729e-05, "loss": 0.0381, "step": 46750 }, { "grad_norm": 0.4664374887943268, "learning_rate": 5.943564857240773e-05, "loss": 0.0358, "step": 46760 }, { "grad_norm": 0.3351638913154602, "learning_rate": 5.941941044928981e-05, "loss": 0.0331, "step": 46770 }, { "grad_norm": 0.2776596248149872, "learning_rate": 5.940317129607935e-05, "loss": 0.0282, "step": 46780 }, { "grad_norm": 0.2810329794883728, "learning_rate": 5.9386931114552204e-05, "loss": 0.0239, "step": 46790 }, { "grad_norm": 0.2833344340324402, "learning_rate": 5.937068990648438e-05, "loss": 0.032, "step": 46800 }, { "grad_norm": 0.188933864235878, "learning_rate": 5.935444767365199e-05, "loss": 0.0387, "step": 46810 }, { "grad_norm": 0.43924999237060547, "learning_rate": 5.933820441783129e-05, "loss": 0.037, "step": 46820 }, { "grad_norm": 0.3799751102924347, "learning_rate": 5.932196014079857e-05, "loss": 0.033, "step": 46830 }, { "grad_norm": 0.449569970369339, "learning_rate": 5.930571484433032e-05, "loss": 0.0393, "step": 46840 }, { "grad_norm": 0.23393061757087708, "learning_rate": 5.928946853020309e-05, "loss": 0.0291, "step": 46850 }, { "grad_norm": 0.39805954694747925, "learning_rate": 5.927322120019352e-05, "loss": 0.0335, "step": 46860 }, { "grad_norm": 0.2819327712059021, "learning_rate": 5.925697285607844e-05, "loss": 0.0305, "step": 46870 }, { "grad_norm": 0.44562461972236633, "learning_rate": 5.924072349963472e-05, "loss": 0.0283, "step": 46880 }, { "grad_norm": 0.20634321868419647, "learning_rate": 5.922447313263938e-05, "loss": 0.0264, "step": 46890 }, { "grad_norm": 0.4970458447933197, "learning_rate": 5.920822175686952e-05, "loss": 0.0348, "step": 46900 }, { "grad_norm": 0.3602643311023712, "learning_rate": 5.919196937410239e-05, "loss": 0.0353, "step": 46910 }, { "grad_norm": 0.2976166605949402, "learning_rate": 5.917571598611529e-05, "loss": 0.038, "step": 46920 }, { "grad_norm": 0.3932865858078003, "learning_rate": 5.9159461594685686e-05, "loss": 0.0334, "step": 46930 }, { "grad_norm": 0.3205997049808502, "learning_rate": 5.914320620159114e-05, "loss": 0.032, "step": 46940 }, { "grad_norm": 0.5550095438957214, "learning_rate": 5.912694980860931e-05, "loss": 0.0282, "step": 46950 }, { "grad_norm": 0.34768378734588623, "learning_rate": 5.9110692417517964e-05, "loss": 0.024, "step": 46960 }, { "grad_norm": 0.2560351490974426, "learning_rate": 5.9094434030095e-05, "loss": 0.0313, "step": 46970 }, { "grad_norm": 0.26229777932167053, "learning_rate": 5.907817464811841e-05, "loss": 0.036, "step": 46980 }, { "grad_norm": 0.40764862298965454, "learning_rate": 5.906191427336627e-05, "loss": 0.031, "step": 46990 }, { "grad_norm": 0.29073503613471985, "learning_rate": 5.9045652907616835e-05, "loss": 0.0301, "step": 47000 }, { "grad_norm": 0.31345948576927185, "learning_rate": 5.902939055264838e-05, "loss": 0.0331, "step": 47010 }, { "grad_norm": 0.2526688575744629, "learning_rate": 5.901312721023935e-05, "loss": 0.03, "step": 47020 }, { "grad_norm": 0.28170645236968994, "learning_rate": 5.8996862882168294e-05, "loss": 0.0339, "step": 47030 }, { "grad_norm": 0.2343430072069168, "learning_rate": 5.8980597570213826e-05, "loss": 0.0315, "step": 47040 }, { "grad_norm": 0.25696656107902527, "learning_rate": 5.896433127615471e-05, "loss": 0.0234, "step": 47050 }, { "grad_norm": 0.2854308485984802, "learning_rate": 5.894806400176981e-05, "loss": 0.0268, "step": 47060 }, { "grad_norm": 0.3152093291282654, "learning_rate": 5.893179574883808e-05, "loss": 0.0355, "step": 47070 }, { "grad_norm": 0.34492263197898865, "learning_rate": 5.8915526519138585e-05, "loss": 0.0285, "step": 47080 }, { "grad_norm": 0.49130892753601074, "learning_rate": 5.889925631445053e-05, "loss": 0.0279, "step": 47090 }, { "grad_norm": 0.3477287292480469, "learning_rate": 5.888298513655318e-05, "loss": 0.0249, "step": 47100 }, { "grad_norm": 0.5069206357002258, "learning_rate": 5.886671298722592e-05, "loss": 0.0279, "step": 47110 }, { "grad_norm": 0.26557886600494385, "learning_rate": 5.885043986824828e-05, "loss": 0.0239, "step": 47120 }, { "grad_norm": 0.8052173256874084, "learning_rate": 5.883416578139982e-05, "loss": 0.0297, "step": 47130 }, { "grad_norm": 0.3277823030948639, "learning_rate": 5.881789072846029e-05, "loss": 0.0359, "step": 47140 }, { "grad_norm": 0.33045920729637146, "learning_rate": 5.8801614711209474e-05, "loss": 0.0228, "step": 47150 }, { "grad_norm": 0.4718170166015625, "learning_rate": 5.878533773142732e-05, "loss": 0.0329, "step": 47160 }, { "grad_norm": 0.33427923917770386, "learning_rate": 5.8769059790893834e-05, "loss": 0.0286, "step": 47170 }, { "grad_norm": 0.2409915328025818, "learning_rate": 5.8752780891389145e-05, "loss": 0.0347, "step": 47180 }, { "grad_norm": 0.32846924662590027, "learning_rate": 5.873650103469351e-05, "loss": 0.0344, "step": 47190 }, { "grad_norm": 0.19930239021778107, "learning_rate": 5.872022022258725e-05, "loss": 0.0275, "step": 47200 }, { "grad_norm": 0.33251962065696716, "learning_rate": 5.8703938456850814e-05, "loss": 0.0304, "step": 47210 }, { "grad_norm": 0.3101442754268646, "learning_rate": 5.868765573926477e-05, "loss": 0.0373, "step": 47220 }, { "grad_norm": 0.47563570737838745, "learning_rate": 5.867137207160974e-05, "loss": 0.0356, "step": 47230 }, { "grad_norm": 0.36502790451049805, "learning_rate": 5.86550874556665e-05, "loss": 0.0304, "step": 47240 }, { "grad_norm": 0.3525463044643402, "learning_rate": 5.863880189321592e-05, "loss": 0.0306, "step": 47250 }, { "grad_norm": 0.48268839716911316, "learning_rate": 5.862251538603894e-05, "loss": 0.0483, "step": 47260 }, { "grad_norm": 0.45023220777511597, "learning_rate": 5.8606227935916656e-05, "loss": 0.0332, "step": 47270 }, { "grad_norm": 0.23746679723262787, "learning_rate": 5.858993954463021e-05, "loss": 0.0306, "step": 47280 }, { "grad_norm": 0.3378685712814331, "learning_rate": 5.8573650213960896e-05, "loss": 0.039, "step": 47290 }, { "grad_norm": 0.3299017548561096, "learning_rate": 5.8557359945690084e-05, "loss": 0.0305, "step": 47300 }, { "grad_norm": 0.3438926935195923, "learning_rate": 5.8541068741599246e-05, "loss": 0.0324, "step": 47310 }, { "grad_norm": 0.28957366943359375, "learning_rate": 5.8524776603469985e-05, "loss": 0.0271, "step": 47320 }, { "grad_norm": 0.4052351713180542, "learning_rate": 5.850848353308397e-05, "loss": 0.0258, "step": 47330 }, { "grad_norm": 0.34000977873802185, "learning_rate": 5.849218953222297e-05, "loss": 0.0309, "step": 47340 }, { "grad_norm": 0.3234298527240753, "learning_rate": 5.847589460266891e-05, "loss": 0.0332, "step": 47350 }, { "grad_norm": 0.32414376735687256, "learning_rate": 5.845959874620375e-05, "loss": 0.032, "step": 47360 }, { "grad_norm": 0.2628415524959564, "learning_rate": 5.84433019646096e-05, "loss": 0.0294, "step": 47370 }, { "grad_norm": 0.3680340051651001, "learning_rate": 5.842700425966863e-05, "loss": 0.035, "step": 47380 }, { "grad_norm": 0.24825629591941833, "learning_rate": 5.841070563316315e-05, "loss": 0.0261, "step": 47390 }, { "grad_norm": 0.3414041996002197, "learning_rate": 5.839440608687554e-05, "loss": 0.0324, "step": 47400 }, { "grad_norm": 0.3494284451007843, "learning_rate": 5.837810562258831e-05, "loss": 0.0251, "step": 47410 }, { "grad_norm": 0.3504273295402527, "learning_rate": 5.836180424208405e-05, "loss": 0.0359, "step": 47420 }, { "grad_norm": 0.3710666000843048, "learning_rate": 5.834550194714543e-05, "loss": 0.0475, "step": 47430 }, { "grad_norm": 0.3569755256175995, "learning_rate": 5.832919873955528e-05, "loss": 0.0391, "step": 47440 }, { "grad_norm": 0.31282714009284973, "learning_rate": 5.831289462109648e-05, "loss": 0.0279, "step": 47450 }, { "grad_norm": 0.43413200974464417, "learning_rate": 5.829658959355202e-05, "loss": 0.0457, "step": 47460 }, { "grad_norm": 0.599020779132843, "learning_rate": 5.8280283658704995e-05, "loss": 0.0401, "step": 47470 }, { "grad_norm": 0.4851572513580322, "learning_rate": 5.826397681833859e-05, "loss": 0.0336, "step": 47480 }, { "grad_norm": 0.42093467712402344, "learning_rate": 5.824766907423612e-05, "loss": 0.0234, "step": 47490 }, { "grad_norm": 0.3471878170967102, "learning_rate": 5.8231360428180945e-05, "loss": 0.0324, "step": 47500 }, { "grad_norm": 0.4544479250907898, "learning_rate": 5.821505088195658e-05, "loss": 0.0259, "step": 47510 }, { "grad_norm": 0.31180518865585327, "learning_rate": 5.819874043734661e-05, "loss": 0.0283, "step": 47520 }, { "grad_norm": 0.4470447897911072, "learning_rate": 5.8182429096134695e-05, "loss": 0.025, "step": 47530 }, { "grad_norm": 0.3411146402359009, "learning_rate": 5.816611686010465e-05, "loss": 0.033, "step": 47540 }, { "grad_norm": 0.36401885747909546, "learning_rate": 5.814980373104033e-05, "loss": 0.0323, "step": 47550 }, { "grad_norm": 0.49294495582580566, "learning_rate": 5.813348971072572e-05, "loss": 0.0394, "step": 47560 }, { "grad_norm": 0.3597768545150757, "learning_rate": 5.811717480094492e-05, "loss": 0.0294, "step": 47570 }, { "grad_norm": 0.3655100166797638, "learning_rate": 5.810085900348209e-05, "loss": 0.034, "step": 47580 }, { "grad_norm": 0.44154122471809387, "learning_rate": 5.8084542320121483e-05, "loss": 0.0342, "step": 47590 }, { "grad_norm": 0.38510048389434814, "learning_rate": 5.8068224752647497e-05, "loss": 0.0461, "step": 47600 }, { "grad_norm": 0.3406083285808563, "learning_rate": 5.805190630284457e-05, "loss": 0.0337, "step": 47610 }, { "grad_norm": 0.7653079032897949, "learning_rate": 5.803558697249729e-05, "loss": 0.0375, "step": 47620 }, { "grad_norm": 0.4208073616027832, "learning_rate": 5.8019266763390276e-05, "loss": 0.041, "step": 47630 }, { "grad_norm": 0.20827889442443848, "learning_rate": 5.800294567730833e-05, "loss": 0.0285, "step": 47640 }, { "grad_norm": 0.3223889172077179, "learning_rate": 5.798662371603628e-05, "loss": 0.0313, "step": 47650 }, { "grad_norm": 0.26139017939567566, "learning_rate": 5.797030088135904e-05, "loss": 0.0316, "step": 47660 }, { "grad_norm": 0.38314393162727356, "learning_rate": 5.7953977175061705e-05, "loss": 0.0271, "step": 47670 }, { "grad_norm": 0.41947367787361145, "learning_rate": 5.793765259892938e-05, "loss": 0.0319, "step": 47680 }, { "grad_norm": 0.20177572965621948, "learning_rate": 5.792132715474729e-05, "loss": 0.0292, "step": 47690 }, { "grad_norm": 0.3664211928844452, "learning_rate": 5.790500084430078e-05, "loss": 0.0285, "step": 47700 }, { "grad_norm": 0.48865407705307007, "learning_rate": 5.7888673669375274e-05, "loss": 0.0293, "step": 47710 }, { "grad_norm": 0.3226590156555176, "learning_rate": 5.787234563175625e-05, "loss": 0.0331, "step": 47720 }, { "grad_norm": 0.46333229541778564, "learning_rate": 5.7856016733229355e-05, "loss": 0.0282, "step": 47730 }, { "grad_norm": 0.37310200929641724, "learning_rate": 5.7839686975580297e-05, "loss": 0.0336, "step": 47740 }, { "grad_norm": 0.5535902380943298, "learning_rate": 5.782335636059484e-05, "loss": 0.0297, "step": 47750 }, { "grad_norm": 0.22768153250217438, "learning_rate": 5.780702489005889e-05, "loss": 0.0278, "step": 47760 }, { "grad_norm": 0.34030771255493164, "learning_rate": 5.779069256575845e-05, "loss": 0.0289, "step": 47770 }, { "grad_norm": 0.47272419929504395, "learning_rate": 5.7774359389479574e-05, "loss": 0.0411, "step": 47780 }, { "grad_norm": 0.3142959475517273, "learning_rate": 5.775802536300845e-05, "loss": 0.0265, "step": 47790 }, { "grad_norm": 0.26943573355674744, "learning_rate": 5.774169048813134e-05, "loss": 0.0241, "step": 47800 }, { "grad_norm": 0.3058566153049469, "learning_rate": 5.77253547666346e-05, "loss": 0.0341, "step": 47810 }, { "grad_norm": 0.4341321885585785, "learning_rate": 5.770901820030465e-05, "loss": 0.0367, "step": 47820 }, { "grad_norm": 0.2997325658798218, "learning_rate": 5.769268079092809e-05, "loss": 0.0321, "step": 47830 }, { "grad_norm": 0.46585002541542053, "learning_rate": 5.767634254029151e-05, "loss": 0.0317, "step": 47840 }, { "grad_norm": 0.43608081340789795, "learning_rate": 5.7660003450181655e-05, "loss": 0.0406, "step": 47850 }, { "grad_norm": 0.31463927030563354, "learning_rate": 5.764366352238534e-05, "loss": 0.0244, "step": 47860 }, { "grad_norm": 0.3014591336250305, "learning_rate": 5.7627322758689474e-05, "loss": 0.024, "step": 47870 }, { "grad_norm": 0.3674318492412567, "learning_rate": 5.761098116088105e-05, "loss": 0.0303, "step": 47880 }, { "grad_norm": 1.2698074579238892, "learning_rate": 5.759463873074717e-05, "loss": 0.034, "step": 47890 }, { "grad_norm": 0.3437529504299164, "learning_rate": 5.757829547007504e-05, "loss": 0.0318, "step": 47900 }, { "grad_norm": 0.453226238489151, "learning_rate": 5.756195138065189e-05, "loss": 0.0215, "step": 47910 }, { "grad_norm": 0.31160667538642883, "learning_rate": 5.754560646426511e-05, "loss": 0.0377, "step": 47920 }, { "grad_norm": 0.36723628640174866, "learning_rate": 5.752926072270216e-05, "loss": 0.0342, "step": 47930 }, { "grad_norm": 0.4027470648288727, "learning_rate": 5.7512914157750563e-05, "loss": 0.026, "step": 47940 }, { "grad_norm": 0.5200299620628357, "learning_rate": 5.749656677119798e-05, "loss": 0.0291, "step": 47950 }, { "grad_norm": 0.4031822383403778, "learning_rate": 5.7480218564832125e-05, "loss": 0.0398, "step": 47960 }, { "grad_norm": 0.9816375374794006, "learning_rate": 5.746386954044082e-05, "loss": 0.0295, "step": 47970 }, { "grad_norm": 0.3593901991844177, "learning_rate": 5.744751969981195e-05, "loss": 0.022, "step": 47980 }, { "grad_norm": 0.33520400524139404, "learning_rate": 5.7431169044733526e-05, "loss": 0.0392, "step": 47990 }, { "grad_norm": 0.26541000604629517, "learning_rate": 5.741481757699364e-05, "loss": 0.0337, "step": 48000 }, { "grad_norm": 0.37139901518821716, "learning_rate": 5.7398465298380434e-05, "loss": 0.0309, "step": 48010 }, { "grad_norm": 0.21759133040905, "learning_rate": 5.7382112210682193e-05, "loss": 0.0445, "step": 48020 }, { "grad_norm": 0.36264801025390625, "learning_rate": 5.7365758315687266e-05, "loss": 0.0365, "step": 48030 }, { "grad_norm": 0.38923075795173645, "learning_rate": 5.734940361518407e-05, "loss": 0.0367, "step": 48040 }, { "grad_norm": 0.4773526191711426, "learning_rate": 5.733304811096116e-05, "loss": 0.0262, "step": 48050 }, { "grad_norm": 0.25676828622817993, "learning_rate": 5.731669180480713e-05, "loss": 0.0346, "step": 48060 }, { "grad_norm": 0.47072991728782654, "learning_rate": 5.730033469851067e-05, "loss": 0.0434, "step": 48070 }, { "grad_norm": 0.568714439868927, "learning_rate": 5.7283976793860607e-05, "loss": 0.0342, "step": 48080 }, { "grad_norm": 0.5407252311706543, "learning_rate": 5.726761809264577e-05, "loss": 0.0321, "step": 48090 }, { "grad_norm": 0.46485504508018494, "learning_rate": 5.7251258596655155e-05, "loss": 0.0279, "step": 48100 }, { "grad_norm": 0.3059700131416321, "learning_rate": 5.72348983076778e-05, "loss": 0.0247, "step": 48110 }, { "grad_norm": 0.43523263931274414, "learning_rate": 5.7218537227502854e-05, "loss": 0.0304, "step": 48120 }, { "grad_norm": 0.4287128746509552, "learning_rate": 5.720217535791951e-05, "loss": 0.0269, "step": 48130 }, { "grad_norm": 0.37352439761161804, "learning_rate": 5.718581270071711e-05, "loss": 0.0315, "step": 48140 }, { "grad_norm": 0.3364638686180115, "learning_rate": 5.716944925768505e-05, "loss": 0.0298, "step": 48150 }, { "grad_norm": 0.3809998333454132, "learning_rate": 5.7153085030612786e-05, "loss": 0.0292, "step": 48160 }, { "grad_norm": 0.2521727979183197, "learning_rate": 5.713672002128989e-05, "loss": 0.0278, "step": 48170 }, { "grad_norm": 0.45722270011901855, "learning_rate": 5.712035423150602e-05, "loss": 0.0327, "step": 48180 }, { "grad_norm": 0.29960787296295166, "learning_rate": 5.710398766305094e-05, "loss": 0.0253, "step": 48190 }, { "grad_norm": 0.4066786766052246, "learning_rate": 5.708762031771442e-05, "loss": 0.0385, "step": 48200 }, { "grad_norm": 0.28366225957870483, "learning_rate": 5.707125219728642e-05, "loss": 0.0217, "step": 48210 }, { "grad_norm": 0.24448364973068237, "learning_rate": 5.7054883303556905e-05, "loss": 0.0248, "step": 48220 }, { "grad_norm": 0.3630671203136444, "learning_rate": 5.703851363831595e-05, "loss": 0.0312, "step": 48230 }, { "grad_norm": 0.7572551369667053, "learning_rate": 5.702214320335374e-05, "loss": 0.0482, "step": 48240 }, { "grad_norm": 0.37136122584342957, "learning_rate": 5.7005772000460514e-05, "loss": 0.0388, "step": 48250 }, { "grad_norm": 0.2918950319290161, "learning_rate": 5.698940003142656e-05, "loss": 0.0319, "step": 48260 }, { "grad_norm": 0.5454902052879333, "learning_rate": 5.697302729804236e-05, "loss": 0.0294, "step": 48270 }, { "grad_norm": 0.38664960861206055, "learning_rate": 5.695665380209837e-05, "loss": 0.0314, "step": 48280 }, { "grad_norm": 0.3638136386871338, "learning_rate": 5.6940279545385165e-05, "loss": 0.0291, "step": 48290 }, { "grad_norm": 0.2756560444831848, "learning_rate": 5.692390452969344e-05, "loss": 0.0293, "step": 48300 }, { "grad_norm": 0.35269415378570557, "learning_rate": 5.690752875681392e-05, "loss": 0.0292, "step": 48310 }, { "grad_norm": 0.29514631628990173, "learning_rate": 5.6891152228537435e-05, "loss": 0.0296, "step": 48320 }, { "grad_norm": 0.4650264382362366, "learning_rate": 5.687477494665492e-05, "loss": 0.0257, "step": 48330 }, { "grad_norm": 0.3456779420375824, "learning_rate": 5.685839691295733e-05, "loss": 0.0268, "step": 48340 }, { "grad_norm": 0.42882105708122253, "learning_rate": 5.6842018129235786e-05, "loss": 0.0376, "step": 48350 }, { "grad_norm": 0.5554303526878357, "learning_rate": 5.6825638597281404e-05, "loss": 0.0309, "step": 48360 }, { "grad_norm": 0.3078099191188812, "learning_rate": 5.680925831888546e-05, "loss": 0.0335, "step": 48370 }, { "grad_norm": 0.314014732837677, "learning_rate": 5.6792877295839274e-05, "loss": 0.0389, "step": 48380 }, { "grad_norm": 0.41265103220939636, "learning_rate": 5.6776495529934224e-05, "loss": 0.0338, "step": 48390 }, { "grad_norm": 0.26838424801826477, "learning_rate": 5.6760113022961824e-05, "loss": 0.0373, "step": 48400 }, { "grad_norm": 0.2530127465724945, "learning_rate": 5.6743729776713617e-05, "loss": 0.0216, "step": 48410 }, { "grad_norm": 0.2533119022846222, "learning_rate": 5.672734579298126e-05, "loss": 0.0319, "step": 48420 }, { "grad_norm": 0.4730961322784424, "learning_rate": 5.671096107355649e-05, "loss": 0.0357, "step": 48430 }, { "grad_norm": 0.4139457643032074, "learning_rate": 5.66945756202311e-05, "loss": 0.0288, "step": 48440 }, { "grad_norm": 0.27894127368927, "learning_rate": 5.667818943479699e-05, "loss": 0.0263, "step": 48450 }, { "grad_norm": 0.5204705595970154, "learning_rate": 5.666180251904612e-05, "loss": 0.0381, "step": 48460 }, { "grad_norm": 0.4732828438282013, "learning_rate": 5.6645414874770555e-05, "loss": 0.0358, "step": 48470 }, { "grad_norm": 0.3175666034221649, "learning_rate": 5.66290265037624e-05, "loss": 0.0324, "step": 48480 }, { "grad_norm": 0.4699462950229645, "learning_rate": 5.661263740781386e-05, "loss": 0.0373, "step": 48490 }, { "grad_norm": 0.3954308032989502, "learning_rate": 5.6596247588717254e-05, "loss": 0.0241, "step": 48500 }, { "grad_norm": 0.40555962920188904, "learning_rate": 5.6579857048264926e-05, "loss": 0.0308, "step": 48510 }, { "grad_norm": 0.33524295687675476, "learning_rate": 5.6563465788249314e-05, "loss": 0.0324, "step": 48520 }, { "grad_norm": 0.22640573978424072, "learning_rate": 5.6547073810462956e-05, "loss": 0.0268, "step": 48530 }, { "grad_norm": 0.23027579486370087, "learning_rate": 5.653068111669846e-05, "loss": 0.0266, "step": 48540 }, { "grad_norm": 0.2752290666103363, "learning_rate": 5.651428770874848e-05, "loss": 0.0299, "step": 48550 }, { "grad_norm": 0.23126164078712463, "learning_rate": 5.64978935884058e-05, "loss": 0.0276, "step": 48560 }, { "grad_norm": 0.45895394682884216, "learning_rate": 5.6481498757463244e-05, "loss": 0.0395, "step": 48570 }, { "grad_norm": 0.32974278926849365, "learning_rate": 5.646510321771373e-05, "loss": 0.0286, "step": 48580 }, { "grad_norm": 0.27013733983039856, "learning_rate": 5.644870697095024e-05, "loss": 0.0319, "step": 48590 }, { "grad_norm": 0.4600803852081299, "learning_rate": 5.643231001896586e-05, "loss": 0.032, "step": 48600 }, { "grad_norm": 0.2743256986141205, "learning_rate": 5.6415912363553726e-05, "loss": 0.0581, "step": 48610 }, { "grad_norm": 0.44755318760871887, "learning_rate": 5.639951400650706e-05, "loss": 0.0371, "step": 48620 }, { "grad_norm": 0.3645901381969452, "learning_rate": 5.6383114949619165e-05, "loss": 0.0334, "step": 48630 }, { "grad_norm": 0.3948754072189331, "learning_rate": 5.636671519468342e-05, "loss": 0.0307, "step": 48640 }, { "grad_norm": 0.3979956805706024, "learning_rate": 5.635031474349327e-05, "loss": 0.0421, "step": 48650 }, { "grad_norm": 0.43159717321395874, "learning_rate": 5.6333913597842246e-05, "loss": 0.0374, "step": 48660 }, { "grad_norm": 0.45044994354248047, "learning_rate": 5.6317511759523955e-05, "loss": 0.044, "step": 48670 }, { "grad_norm": 0.4369845688343048, "learning_rate": 5.630110923033207e-05, "loss": 0.0293, "step": 48680 }, { "grad_norm": 0.3530597686767578, "learning_rate": 5.628470601206036e-05, "loss": 0.0415, "step": 48690 }, { "grad_norm": 1.0302764177322388, "learning_rate": 5.626830210650266e-05, "loss": 0.0286, "step": 48700 }, { "grad_norm": 0.39554429054260254, "learning_rate": 5.625189751545285e-05, "loss": 0.0355, "step": 48710 }, { "grad_norm": 0.4377705156803131, "learning_rate": 5.6235492240704936e-05, "loss": 0.0369, "step": 48720 }, { "grad_norm": 0.5216577649116516, "learning_rate": 5.621908628405296e-05, "loss": 0.039, "step": 48730 }, { "grad_norm": 0.6584490537643433, "learning_rate": 5.620267964729106e-05, "loss": 0.0426, "step": 48740 }, { "grad_norm": 0.2933180630207062, "learning_rate": 5.618627233221344e-05, "loss": 0.0308, "step": 48750 }, { "grad_norm": 0.37584057450294495, "learning_rate": 5.616986434061438e-05, "loss": 0.0344, "step": 48760 }, { "grad_norm": 0.30644166469573975, "learning_rate": 5.615345567428822e-05, "loss": 0.0254, "step": 48770 }, { "grad_norm": 0.27518758177757263, "learning_rate": 5.613704633502941e-05, "loss": 0.0218, "step": 48780 }, { "grad_norm": 0.2257193624973297, "learning_rate": 5.612063632463245e-05, "loss": 0.0203, "step": 48790 }, { "grad_norm": 0.28410404920578003, "learning_rate": 5.610422564489188e-05, "loss": 0.0273, "step": 48800 }, { "grad_norm": 0.3986169993877411, "learning_rate": 5.608781429760239e-05, "loss": 0.0329, "step": 48810 }, { "grad_norm": 0.31623342633247375, "learning_rate": 5.607140228455866e-05, "loss": 0.0296, "step": 48820 }, { "grad_norm": 0.3077884018421173, "learning_rate": 5.605498960755553e-05, "loss": 0.0326, "step": 48830 }, { "grad_norm": 0.38683634996414185, "learning_rate": 5.603857626838782e-05, "loss": 0.0344, "step": 48840 }, { "grad_norm": 0.331497460603714, "learning_rate": 5.60221622688505e-05, "loss": 0.0344, "step": 48850 }, { "grad_norm": 0.5015192627906799, "learning_rate": 5.6005747610738565e-05, "loss": 0.0346, "step": 48860 }, { "grad_norm": 0.4695807099342346, "learning_rate": 5.59893322958471e-05, "loss": 0.0302, "step": 48870 }, { "grad_norm": 0.2885936200618744, "learning_rate": 5.5972916325971256e-05, "loss": 0.0372, "step": 48880 }, { "grad_norm": 0.3243655860424042, "learning_rate": 5.595649970290628e-05, "loss": 0.0393, "step": 48890 }, { "grad_norm": 0.3265816271305084, "learning_rate": 5.5940082428447426e-05, "loss": 0.0318, "step": 48900 }, { "grad_norm": 0.477897971868515, "learning_rate": 5.592366450439012e-05, "loss": 0.047, "step": 48910 }, { "grad_norm": 0.44036370515823364, "learning_rate": 5.590724593252975e-05, "loss": 0.0326, "step": 48920 }, { "grad_norm": 0.3312816321849823, "learning_rate": 5.589082671466184e-05, "loss": 0.032, "step": 48930 }, { "grad_norm": 0.23847201466560364, "learning_rate": 5.587440685258199e-05, "loss": 0.0292, "step": 48940 }, { "grad_norm": 0.2824300527572632, "learning_rate": 5.585798634808583e-05, "loss": 0.0221, "step": 48950 }, { "grad_norm": 0.4750296175479889, "learning_rate": 5.584156520296909e-05, "loss": 0.0286, "step": 48960 }, { "grad_norm": 0.26997387409210205, "learning_rate": 5.582514341902757e-05, "loss": 0.0227, "step": 48970 }, { "grad_norm": 0.34975072741508484, "learning_rate": 5.580872099805713e-05, "loss": 0.0243, "step": 48980 }, { "grad_norm": 0.33342355489730835, "learning_rate": 5.5792297941853674e-05, "loss": 0.0287, "step": 48990 }, { "grad_norm": 0.3264351487159729, "learning_rate": 5.5775874252213247e-05, "loss": 0.033, "step": 49000 }, { "grad_norm": 0.7176574468612671, "learning_rate": 5.575944993093189e-05, "loss": 0.0326, "step": 49010 }, { "grad_norm": 0.29946786165237427, "learning_rate": 5.574302497980574e-05, "loss": 0.0293, "step": 49020 }, { "grad_norm": 0.4656549394130707, "learning_rate": 5.5726599400631005e-05, "loss": 0.0312, "step": 49030 }, { "grad_norm": 0.4268670678138733, "learning_rate": 5.571017319520399e-05, "loss": 0.0248, "step": 49040 }, { "grad_norm": 0.3573026955127716, "learning_rate": 5.5693746365320984e-05, "loss": 0.0334, "step": 49050 }, { "grad_norm": 0.249522864818573, "learning_rate": 5.567731891277848e-05, "loss": 0.0298, "step": 49060 }, { "grad_norm": 0.44277071952819824, "learning_rate": 5.566089083937288e-05, "loss": 0.04, "step": 49070 }, { "grad_norm": 0.5670885443687439, "learning_rate": 5.564446214690079e-05, "loss": 0.0399, "step": 49080 }, { "grad_norm": 0.2420680671930313, "learning_rate": 5.56280328371588e-05, "loss": 0.0299, "step": 49090 }, { "grad_norm": 0.30999892950057983, "learning_rate": 5.5611602911943596e-05, "loss": 0.0342, "step": 49100 }, { "grad_norm": 0.29398664832115173, "learning_rate": 5.559517237305194e-05, "loss": 0.0449, "step": 49110 }, { "grad_norm": 0.4130973219871521, "learning_rate": 5.557874122228064e-05, "loss": 0.0335, "step": 49120 }, { "grad_norm": 0.28031593561172485, "learning_rate": 5.55623094614266e-05, "loss": 0.0335, "step": 49130 }, { "grad_norm": 0.26724302768707275, "learning_rate": 5.5545877092286743e-05, "loss": 0.0316, "step": 49140 }, { "grad_norm": 0.3518003225326538, "learning_rate": 5.552944411665811e-05, "loss": 0.0234, "step": 49150 }, { "grad_norm": 0.5475871562957764, "learning_rate": 5.5513010536337795e-05, "loss": 0.0339, "step": 49160 }, { "grad_norm": 0.36453568935394287, "learning_rate": 5.5496576353122934e-05, "loss": 0.0326, "step": 49170 }, { "grad_norm": 0.3212921917438507, "learning_rate": 5.548014156881074e-05, "loss": 0.0235, "step": 49180 }, { "grad_norm": 0.3700053095817566, "learning_rate": 5.5463706185198494e-05, "loss": 0.031, "step": 49190 }, { "grad_norm": 0.2777331471443176, "learning_rate": 5.544727020408358e-05, "loss": 0.0241, "step": 49200 }, { "grad_norm": 0.3112677335739136, "learning_rate": 5.5430833627263366e-05, "loss": 0.0313, "step": 49210 }, { "grad_norm": 0.3920641243457794, "learning_rate": 5.5414396456535354e-05, "loss": 0.0334, "step": 49220 }, { "grad_norm": 0.42978835105895996, "learning_rate": 5.539795869369709e-05, "loss": 0.0274, "step": 49230 }, { "grad_norm": 0.3622482419013977, "learning_rate": 5.538152034054618e-05, "loss": 0.03, "step": 49240 }, { "grad_norm": 0.5223897099494934, "learning_rate": 5.536508139888028e-05, "loss": 0.0349, "step": 49250 }, { "grad_norm": 0.28042489290237427, "learning_rate": 5.534864187049716e-05, "loss": 0.0284, "step": 49260 }, { "grad_norm": 0.7570464611053467, "learning_rate": 5.5332201757194615e-05, "loss": 0.035, "step": 49270 }, { "grad_norm": 0.4405370056629181, "learning_rate": 5.531576106077049e-05, "loss": 0.0334, "step": 49280 }, { "grad_norm": 0.4984023869037628, "learning_rate": 5.529931978302272e-05, "loss": 0.0332, "step": 49290 }, { "grad_norm": 1.9516804218292236, "learning_rate": 5.528287792574932e-05, "loss": 0.042, "step": 49300 }, { "grad_norm": 0.24157704412937164, "learning_rate": 5.5266435490748324e-05, "loss": 0.0305, "step": 49310 }, { "grad_norm": 0.3660617768764496, "learning_rate": 5.524999247981787e-05, "loss": 0.0334, "step": 49320 }, { "grad_norm": 0.3385765254497528, "learning_rate": 5.523354889475613e-05, "loss": 0.0329, "step": 49330 }, { "grad_norm": 0.34235522150993347, "learning_rate": 5.521710473736134e-05, "loss": 0.0332, "step": 49340 }, { "grad_norm": 0.18064440786838531, "learning_rate": 5.5200660009431826e-05, "loss": 0.0274, "step": 49350 }, { "grad_norm": 0.29550063610076904, "learning_rate": 5.518421471276596e-05, "loss": 0.0264, "step": 49360 }, { "grad_norm": 0.5988187789916992, "learning_rate": 5.5167768849162174e-05, "loss": 0.0307, "step": 49370 }, { "grad_norm": 0.2741139233112335, "learning_rate": 5.515132242041893e-05, "loss": 0.0353, "step": 49380 }, { "grad_norm": 0.48622214794158936, "learning_rate": 5.513487542833483e-05, "loss": 0.0297, "step": 49390 }, { "grad_norm": 0.3796064555644989, "learning_rate": 5.5118427874708466e-05, "loss": 0.0308, "step": 49400 }, { "grad_norm": 0.2801552414894104, "learning_rate": 5.510197976133853e-05, "loss": 0.0266, "step": 49410 }, { "grad_norm": 0.41168129444122314, "learning_rate": 5.508553109002376e-05, "loss": 0.0297, "step": 49420 }, { "grad_norm": 0.41165298223495483, "learning_rate": 5.5069081862562957e-05, "loss": 0.0281, "step": 49430 }, { "grad_norm": 0.30897292494773865, "learning_rate": 5.5052632080754965e-05, "loss": 0.0221, "step": 49440 }, { "grad_norm": 0.2509043216705322, "learning_rate": 5.503618174639874e-05, "loss": 0.0276, "step": 49450 }, { "grad_norm": 0.42915380001068115, "learning_rate": 5.501973086129325e-05, "loss": 0.0304, "step": 49460 }, { "grad_norm": 0.5831298232078552, "learning_rate": 5.500327942723753e-05, "loss": 0.0334, "step": 49470 }, { "grad_norm": 0.7623299956321716, "learning_rate": 5.498682744603071e-05, "loss": 0.0303, "step": 49480 }, { "grad_norm": 0.3162215054035187, "learning_rate": 5.497037491947194e-05, "loss": 0.033, "step": 49490 }, { "grad_norm": 0.25548210740089417, "learning_rate": 5.4953921849360424e-05, "loss": 0.0304, "step": 49500 }, { "grad_norm": 0.3248414993286133, "learning_rate": 5.493746823749547e-05, "loss": 0.0332, "step": 49510 }, { "grad_norm": 0.6449852585792542, "learning_rate": 5.49210140856764e-05, "loss": 0.0392, "step": 49520 }, { "grad_norm": 0.23395217955112457, "learning_rate": 5.4904559395702647e-05, "loss": 0.0241, "step": 49530 }, { "grad_norm": 0.5528903007507324, "learning_rate": 5.488810416937364e-05, "loss": 0.0257, "step": 49540 }, { "grad_norm": 0.4212830662727356, "learning_rate": 5.4871648408488926e-05, "loss": 0.0333, "step": 49550 }, { "grad_norm": 0.5028489232063293, "learning_rate": 5.485519211484807e-05, "loss": 0.0308, "step": 49560 }, { "grad_norm": 0.4601914584636688, "learning_rate": 5.4838735290250676e-05, "loss": 0.0354, "step": 49570 }, { "grad_norm": 1.045504093170166, "learning_rate": 5.482227793649649e-05, "loss": 0.0339, "step": 49580 }, { "grad_norm": 0.5473504066467285, "learning_rate": 5.480582005538524e-05, "loss": 0.0316, "step": 49590 }, { "grad_norm": 0.5026081204414368, "learning_rate": 5.478936164871671e-05, "loss": 0.0279, "step": 49600 }, { "grad_norm": 0.5032469630241394, "learning_rate": 5.477290271829081e-05, "loss": 0.0408, "step": 49610 }, { "grad_norm": 0.3932247757911682, "learning_rate": 5.475644326590744e-05, "loss": 0.0337, "step": 49620 }, { "grad_norm": 0.38997629284858704, "learning_rate": 5.473998329336658e-05, "loss": 0.0313, "step": 49630 }, { "grad_norm": 0.3467336893081665, "learning_rate": 5.4723522802468286e-05, "loss": 0.0281, "step": 49640 }, { "grad_norm": 0.3042188584804535, "learning_rate": 5.470706179501264e-05, "loss": 0.0294, "step": 49650 }, { "grad_norm": 0.2680201530456543, "learning_rate": 5.4690600272799776e-05, "loss": 0.0331, "step": 49660 }, { "grad_norm": 0.5247873067855835, "learning_rate": 5.467413823762993e-05, "loss": 0.0325, "step": 49670 }, { "grad_norm": 0.4100964367389679, "learning_rate": 5.465767569130335e-05, "loss": 0.027, "step": 49680 }, { "grad_norm": 0.3324335813522339, "learning_rate": 5.464121263562036e-05, "loss": 0.0307, "step": 49690 }, { "grad_norm": 0.4068828523159027, "learning_rate": 5.4624749072381343e-05, "loss": 0.0361, "step": 49700 }, { "grad_norm": 0.5506085753440857, "learning_rate": 5.460828500338672e-05, "loss": 0.0292, "step": 49710 }, { "grad_norm": 0.34246066212654114, "learning_rate": 5.459182043043698e-05, "loss": 0.0426, "step": 49720 }, { "grad_norm": 0.25340163707733154, "learning_rate": 5.457535535533265e-05, "loss": 0.0276, "step": 49730 }, { "grad_norm": 0.38420340418815613, "learning_rate": 5.4558889779874334e-05, "loss": 0.0354, "step": 49740 }, { "grad_norm": 0.5003181099891663, "learning_rate": 5.454242370586269e-05, "loss": 0.0364, "step": 49750 }, { "grad_norm": 0.25581490993499756, "learning_rate": 5.452595713509843e-05, "loss": 0.0338, "step": 49760 }, { "grad_norm": 0.795312762260437, "learning_rate": 5.4509490069382274e-05, "loss": 0.0252, "step": 49770 }, { "grad_norm": 0.24178056418895721, "learning_rate": 5.449302251051509e-05, "loss": 0.0268, "step": 49780 }, { "grad_norm": 0.31851619482040405, "learning_rate": 5.44765544602977e-05, "loss": 0.0358, "step": 49790 }, { "grad_norm": 0.5102514624595642, "learning_rate": 5.446008592053107e-05, "loss": 0.0341, "step": 49800 }, { "grad_norm": 0.34370744228363037, "learning_rate": 5.4443616893016135e-05, "loss": 0.0313, "step": 49810 }, { "grad_norm": 0.3915124833583832, "learning_rate": 5.4427147379553934e-05, "loss": 0.0305, "step": 49820 }, { "grad_norm": 0.2879994809627533, "learning_rate": 5.441067738194556e-05, "loss": 0.0295, "step": 49830 }, { "grad_norm": 0.3193698227405548, "learning_rate": 5.439420690199214e-05, "loss": 0.0344, "step": 49840 }, { "grad_norm": 0.322599858045578, "learning_rate": 5.437773594149484e-05, "loss": 0.0377, "step": 49850 }, { "grad_norm": 0.31581684947013855, "learning_rate": 5.436126450225495e-05, "loss": 0.027, "step": 49860 }, { "grad_norm": 0.32221662998199463, "learning_rate": 5.434479258607371e-05, "loss": 0.0323, "step": 49870 }, { "grad_norm": 0.8549134731292725, "learning_rate": 5.43283201947525e-05, "loss": 0.029, "step": 49880 }, { "grad_norm": 0.3501631021499634, "learning_rate": 5.431184733009268e-05, "loss": 0.0459, "step": 49890 }, { "grad_norm": 0.23508350551128387, "learning_rate": 5.4295373993895736e-05, "loss": 0.0281, "step": 49900 }, { "grad_norm": 0.40066927671432495, "learning_rate": 5.4278900187963157e-05, "loss": 0.0311, "step": 49910 }, { "grad_norm": 0.2963639199733734, "learning_rate": 5.426242591409646e-05, "loss": 0.0359, "step": 49920 }, { "grad_norm": 0.807601273059845, "learning_rate": 5.42459511740973e-05, "loss": 0.0303, "step": 49930 }, { "grad_norm": 0.2336299568414688, "learning_rate": 5.422947596976729e-05, "loss": 0.0294, "step": 49940 }, { "grad_norm": 0.3482988774776459, "learning_rate": 5.4213000302908134e-05, "loss": 0.0315, "step": 49950 }, { "grad_norm": 0.31709179282188416, "learning_rate": 5.419652417532162e-05, "loss": 0.027, "step": 49960 }, { "grad_norm": 0.37217608094215393, "learning_rate": 5.4180047588809534e-05, "loss": 0.0285, "step": 49970 }, { "grad_norm": 0.32189586758613586, "learning_rate": 5.4163570545173704e-05, "loss": 0.0271, "step": 49980 }, { "grad_norm": 0.42917174100875854, "learning_rate": 5.414709304621608e-05, "loss": 0.0368, "step": 49990 }, { "grad_norm": 1.0238460302352905, "learning_rate": 5.413061509373858e-05, "loss": 0.0315, "step": 50000 }, { "grad_norm": 0.6206430196762085, "learning_rate": 5.411413668954324e-05, "loss": 0.0346, "step": 50010 }, { "grad_norm": 0.3920648694038391, "learning_rate": 5.4097657835432083e-05, "loss": 0.0319, "step": 50020 }, { "grad_norm": 0.2252926230430603, "learning_rate": 5.408117853320723e-05, "loss": 0.0276, "step": 50030 }, { "grad_norm": 0.3834969103336334, "learning_rate": 5.406469878467084e-05, "loss": 0.0285, "step": 50040 }, { "grad_norm": 0.20033623278141022, "learning_rate": 5.404821859162509e-05, "loss": 0.0336, "step": 50050 }, { "grad_norm": 0.31459471583366394, "learning_rate": 5.403173795587225e-05, "loss": 0.0279, "step": 50060 }, { "grad_norm": 0.357838898897171, "learning_rate": 5.4015256879214606e-05, "loss": 0.0284, "step": 50070 }, { "grad_norm": 0.41938620805740356, "learning_rate": 5.39987753634545e-05, "loss": 0.0316, "step": 50080 }, { "grad_norm": 0.3352167010307312, "learning_rate": 5.3982293410394335e-05, "loss": 0.0383, "step": 50090 }, { "grad_norm": 0.40787097811698914, "learning_rate": 5.396581102183655e-05, "loss": 0.0336, "step": 50100 }, { "grad_norm": 0.38739433884620667, "learning_rate": 5.3949328199583634e-05, "loss": 0.0282, "step": 50110 }, { "grad_norm": 0.34079328179359436, "learning_rate": 5.393284494543813e-05, "loss": 0.0343, "step": 50120 }, { "grad_norm": 0.3170253038406372, "learning_rate": 5.391636126120262e-05, "loss": 0.0295, "step": 50130 }, { "grad_norm": 0.35575544834136963, "learning_rate": 5.389987714867971e-05, "loss": 0.023, "step": 50140 }, { "grad_norm": 0.25991204380989075, "learning_rate": 5.3883392609672114e-05, "loss": 0.0223, "step": 50150 }, { "grad_norm": 0.32065272331237793, "learning_rate": 5.3866907645982545e-05, "loss": 0.0296, "step": 50160 }, { "grad_norm": 0.2925384044647217, "learning_rate": 5.385042225941375e-05, "loss": 0.0265, "step": 50170 }, { "grad_norm": 0.3572738766670227, "learning_rate": 5.3833936451768585e-05, "loss": 0.0265, "step": 50180 }, { "grad_norm": 0.4740144610404968, "learning_rate": 5.3817450224849884e-05, "loss": 0.0337, "step": 50190 }, { "grad_norm": 0.2326626181602478, "learning_rate": 5.380096358046056e-05, "loss": 0.0399, "step": 50200 }, { "grad_norm": 0.3552899956703186, "learning_rate": 5.378447652040359e-05, "loss": 0.0332, "step": 50210 }, { "grad_norm": 0.31128472089767456, "learning_rate": 5.376798904648195e-05, "loss": 0.0299, "step": 50220 }, { "grad_norm": 0.35738882422447205, "learning_rate": 5.375150116049869e-05, "loss": 0.0297, "step": 50230 }, { "grad_norm": 0.7075569033622742, "learning_rate": 5.373501286425691e-05, "loss": 0.0312, "step": 50240 }, { "grad_norm": 0.282027930021286, "learning_rate": 5.3718524159559726e-05, "loss": 0.0258, "step": 50250 }, { "grad_norm": 0.41716906428337097, "learning_rate": 5.370203504821034e-05, "loss": 0.0314, "step": 50260 }, { "grad_norm": 0.24625073373317719, "learning_rate": 5.368554553201196e-05, "loss": 0.0295, "step": 50270 }, { "grad_norm": 0.9020963311195374, "learning_rate": 5.3669055612767874e-05, "loss": 0.0411, "step": 50280 }, { "grad_norm": 0.34510916471481323, "learning_rate": 5.3652565292281376e-05, "loss": 0.0319, "step": 50290 }, { "grad_norm": 0.5479210019111633, "learning_rate": 5.3636074572355824e-05, "loss": 0.0342, "step": 50300 }, { "grad_norm": 0.22686995565891266, "learning_rate": 5.361958345479463e-05, "loss": 0.0252, "step": 50310 }, { "grad_norm": 0.3077006936073303, "learning_rate": 5.360309194140123e-05, "loss": 0.031, "step": 50320 }, { "grad_norm": 0.3207424283027649, "learning_rate": 5.3586600033979106e-05, "loss": 0.0326, "step": 50330 }, { "grad_norm": 0.3751095235347748, "learning_rate": 5.357010773433181e-05, "loss": 0.035, "step": 50340 }, { "grad_norm": 0.5127211809158325, "learning_rate": 5.3553615044262895e-05, "loss": 0.0261, "step": 50350 }, { "grad_norm": 0.47026461362838745, "learning_rate": 5.3537121965575964e-05, "loss": 0.0231, "step": 50360 }, { "grad_norm": 0.356322705745697, "learning_rate": 5.352062850007471e-05, "loss": 0.0253, "step": 50370 }, { "grad_norm": 0.26009371876716614, "learning_rate": 5.350413464956282e-05, "loss": 0.0266, "step": 50380 }, { "grad_norm": 0.5010318756103516, "learning_rate": 5.348764041584403e-05, "loss": 0.0347, "step": 50390 }, { "grad_norm": 0.3264400362968445, "learning_rate": 5.347114580072212e-05, "loss": 0.0294, "step": 50400 }, { "grad_norm": 0.39519378542900085, "learning_rate": 5.345465080600093e-05, "loss": 0.0371, "step": 50410 }, { "grad_norm": 0.28712356090545654, "learning_rate": 5.3438155433484326e-05, "loss": 0.0312, "step": 50420 }, { "grad_norm": 0.37389370799064636, "learning_rate": 5.3421659684976197e-05, "loss": 0.0306, "step": 50430 }, { "grad_norm": 0.2373158186674118, "learning_rate": 5.340516356228052e-05, "loss": 0.0256, "step": 50440 }, { "grad_norm": 1.7233424186706543, "learning_rate": 5.338866706720128e-05, "loss": 0.032, "step": 50450 }, { "grad_norm": 0.225332111120224, "learning_rate": 5.337217020154249e-05, "loss": 0.0277, "step": 50460 }, { "grad_norm": 0.38228124380111694, "learning_rate": 5.335567296710825e-05, "loss": 0.0338, "step": 50470 }, { "grad_norm": 0.4875324070453644, "learning_rate": 5.333917536570265e-05, "loss": 0.027, "step": 50480 }, { "grad_norm": 0.2817456126213074, "learning_rate": 5.332267739912986e-05, "loss": 0.0312, "step": 50490 }, { "grad_norm": 1.2372167110443115, "learning_rate": 5.330617906919405e-05, "loss": 0.0335, "step": 50500 }, { "grad_norm": 0.22256121039390564, "learning_rate": 5.328968037769949e-05, "loss": 0.0317, "step": 50510 }, { "grad_norm": 0.3126170039176941, "learning_rate": 5.327318132645042e-05, "loss": 0.0304, "step": 50520 }, { "grad_norm": 0.2726093530654907, "learning_rate": 5.325668191725116e-05, "loss": 0.034, "step": 50530 }, { "grad_norm": 0.6743184328079224, "learning_rate": 5.324018215190606e-05, "loss": 0.0264, "step": 50540 }, { "grad_norm": 0.3415083885192871, "learning_rate": 5.3223682032219515e-05, "loss": 0.0264, "step": 50550 }, { "grad_norm": 0.17962239682674408, "learning_rate": 5.320718155999595e-05, "loss": 0.0305, "step": 50560 }, { "grad_norm": 0.3028618097305298, "learning_rate": 5.3190680737039835e-05, "loss": 0.0306, "step": 50570 }, { "grad_norm": 0.3306210935115814, "learning_rate": 5.317417956515567e-05, "loss": 0.0269, "step": 50580 }, { "grad_norm": 0.5025006532669067, "learning_rate": 5.315767804614798e-05, "loss": 0.034, "step": 50590 }, { "grad_norm": 0.3926761746406555, "learning_rate": 5.3141176181821395e-05, "loss": 0.0281, "step": 50600 }, { "grad_norm": 0.4803366959095001, "learning_rate": 5.31246739739805e-05, "loss": 0.0278, "step": 50610 }, { "grad_norm": 0.4236290752887726, "learning_rate": 5.310817142442994e-05, "loss": 0.0377, "step": 50620 }, { "grad_norm": 0.42039254307746887, "learning_rate": 5.309166853497445e-05, "loss": 0.0324, "step": 50630 }, { "grad_norm": 0.47172316908836365, "learning_rate": 5.307516530741873e-05, "loss": 0.0377, "step": 50640 }, { "grad_norm": 0.3962312340736389, "learning_rate": 5.305866174356754e-05, "loss": 0.0321, "step": 50650 }, { "grad_norm": 0.3841498792171478, "learning_rate": 5.304215784522571e-05, "loss": 0.0282, "step": 50660 }, { "grad_norm": 0.32038894295692444, "learning_rate": 5.302565361419808e-05, "loss": 0.0283, "step": 50670 }, { "grad_norm": 0.3255561888217926, "learning_rate": 5.3009149052289507e-05, "loss": 0.0397, "step": 50680 }, { "grad_norm": 0.3645227551460266, "learning_rate": 5.299264416130493e-05, "loss": 0.0399, "step": 50690 }, { "grad_norm": 0.42876872420310974, "learning_rate": 5.297613894304928e-05, "loss": 0.0233, "step": 50700 }, { "grad_norm": 0.34405967593193054, "learning_rate": 5.2959633399327534e-05, "loss": 0.0274, "step": 50710 }, { "grad_norm": 0.3509049713611603, "learning_rate": 5.294312753194476e-05, "loss": 0.0267, "step": 50720 }, { "grad_norm": 0.276118665933609, "learning_rate": 5.292662134270596e-05, "loss": 0.0267, "step": 50730 }, { "grad_norm": 0.3628460764884949, "learning_rate": 5.291011483341626e-05, "loss": 0.0345, "step": 50740 }, { "grad_norm": 0.32360562682151794, "learning_rate": 5.2893608005880767e-05, "loss": 0.0295, "step": 50750 }, { "grad_norm": 0.35989293456077576, "learning_rate": 5.287710086190467e-05, "loss": 0.0368, "step": 50760 }, { "grad_norm": 0.22679133713245392, "learning_rate": 5.286059340329314e-05, "loss": 0.0372, "step": 50770 }, { "grad_norm": 0.3006587624549866, "learning_rate": 5.284408563185141e-05, "loss": 0.0367, "step": 50780 }, { "grad_norm": 0.48800426721572876, "learning_rate": 5.282757754938476e-05, "loss": 0.0366, "step": 50790 }, { "grad_norm": 0.3498825430870056, "learning_rate": 5.281106915769849e-05, "loss": 0.0328, "step": 50800 }, { "grad_norm": 0.31928572058677673, "learning_rate": 5.2794560458597897e-05, "loss": 0.0278, "step": 50810 }, { "grad_norm": 0.541029691696167, "learning_rate": 5.27780514538884e-05, "loss": 0.0268, "step": 50820 }, { "grad_norm": 0.21769817173480988, "learning_rate": 5.2761542145375365e-05, "loss": 0.0279, "step": 50830 }, { "grad_norm": 0.32266244292259216, "learning_rate": 5.274503253486421e-05, "loss": 0.0257, "step": 50840 }, { "grad_norm": 0.3344782292842865, "learning_rate": 5.272852262416046e-05, "loss": 0.0267, "step": 50850 }, { "grad_norm": 0.5177004337310791, "learning_rate": 5.2712012415069555e-05, "loss": 0.0306, "step": 50860 }, { "grad_norm": 0.5792094469070435, "learning_rate": 5.269550190939705e-05, "loss": 0.0402, "step": 50870 }, { "grad_norm": 0.32876715064048767, "learning_rate": 5.267899110894852e-05, "loss": 0.0336, "step": 50880 }, { "grad_norm": 0.3234984576702118, "learning_rate": 5.266248001552955e-05, "loss": 0.027, "step": 50890 }, { "grad_norm": 0.23161102831363678, "learning_rate": 5.264596863094575e-05, "loss": 0.0218, "step": 50900 }, { "grad_norm": 0.26960253715515137, "learning_rate": 5.262945695700282e-05, "loss": 0.0289, "step": 50910 }, { "grad_norm": 0.3359469175338745, "learning_rate": 5.261294499550643e-05, "loss": 0.0241, "step": 50920 }, { "grad_norm": 0.3035518229007721, "learning_rate": 5.25964327482623e-05, "loss": 0.0234, "step": 50930 }, { "grad_norm": 0.2958924174308777, "learning_rate": 5.257992021707617e-05, "loss": 0.0361, "step": 50940 }, { "grad_norm": 0.3062402307987213, "learning_rate": 5.256340740375387e-05, "loss": 0.0356, "step": 50950 }, { "grad_norm": 0.3348551392555237, "learning_rate": 5.254689431010117e-05, "loss": 0.0343, "step": 50960 }, { "grad_norm": 0.5003553628921509, "learning_rate": 5.253038093792395e-05, "loss": 0.0255, "step": 50970 }, { "grad_norm": 0.4943641722202301, "learning_rate": 5.251386728902806e-05, "loss": 0.0323, "step": 50980 }, { "grad_norm": 0.34492579102516174, "learning_rate": 5.2497353365219446e-05, "loss": 0.028, "step": 50990 }, { "grad_norm": 0.29556915163993835, "learning_rate": 5.2480839168304e-05, "loss": 0.0302, "step": 51000 }, { "grad_norm": 0.4862864911556244, "learning_rate": 5.246432470008773e-05, "loss": 0.0272, "step": 51010 }, { "grad_norm": 0.43924281001091003, "learning_rate": 5.2447809962376616e-05, "loss": 0.0296, "step": 51020 }, { "grad_norm": 0.4462554156780243, "learning_rate": 5.243129495697666e-05, "loss": 0.0255, "step": 51030 }, { "grad_norm": 0.346802681684494, "learning_rate": 5.241477968569397e-05, "loss": 0.0313, "step": 51040 }, { "grad_norm": 0.4626866281032562, "learning_rate": 5.23982641503346e-05, "loss": 0.0308, "step": 51050 }, { "grad_norm": 0.6749685406684875, "learning_rate": 5.238174835270465e-05, "loss": 0.0353, "step": 51060 }, { "grad_norm": 0.5794435739517212, "learning_rate": 5.23652322946103e-05, "loss": 0.0304, "step": 51070 }, { "grad_norm": 0.5278174877166748, "learning_rate": 5.234871597785771e-05, "loss": 0.0303, "step": 51080 }, { "grad_norm": 0.5461829304695129, "learning_rate": 5.233219940425306e-05, "loss": 0.0346, "step": 51090 }, { "grad_norm": 0.5750077366828918, "learning_rate": 5.2315682575602576e-05, "loss": 0.0488, "step": 51100 }, { "grad_norm": 0.5272740125656128, "learning_rate": 5.229916549371254e-05, "loss": 0.0326, "step": 51110 }, { "grad_norm": 0.35555335879325867, "learning_rate": 5.228264816038924e-05, "loss": 0.0319, "step": 51120 }, { "grad_norm": 0.298885315656662, "learning_rate": 5.2266130577438935e-05, "loss": 0.0282, "step": 51130 }, { "grad_norm": 0.29717057943344116, "learning_rate": 5.224961274666802e-05, "loss": 0.0272, "step": 51140 }, { "grad_norm": 0.25632116198539734, "learning_rate": 5.223309466988283e-05, "loss": 0.034, "step": 51150 }, { "grad_norm": 0.376380980014801, "learning_rate": 5.221657634888976e-05, "loss": 0.0455, "step": 51160 }, { "grad_norm": 0.34470728039741516, "learning_rate": 5.2200057785495235e-05, "loss": 0.0374, "step": 51170 }, { "grad_norm": 0.4148538410663605, "learning_rate": 5.2183538981505706e-05, "loss": 0.0296, "step": 51180 }, { "grad_norm": 0.2640397846698761, "learning_rate": 5.216701993872762e-05, "loss": 0.0237, "step": 51190 }, { "grad_norm": 0.3213750720024109, "learning_rate": 5.215050065896752e-05, "loss": 0.0349, "step": 51200 }, { "grad_norm": 0.34417852759361267, "learning_rate": 5.2133981144031875e-05, "loss": 0.0441, "step": 51210 }, { "grad_norm": 0.2687736749649048, "learning_rate": 5.211746139572727e-05, "loss": 0.0394, "step": 51220 }, { "grad_norm": 0.34690433740615845, "learning_rate": 5.210094141586026e-05, "loss": 0.0322, "step": 51230 }, { "grad_norm": 0.38761213421821594, "learning_rate": 5.2084421206237476e-05, "loss": 0.0341, "step": 51240 }, { "grad_norm": 0.36616936326026917, "learning_rate": 5.20679007686655e-05, "loss": 0.0284, "step": 51250 }, { "grad_norm": 0.31473296880722046, "learning_rate": 5.205138010495102e-05, "loss": 0.0299, "step": 51260 }, { "grad_norm": 0.3315679132938385, "learning_rate": 5.20348592169007e-05, "loss": 0.0285, "step": 51270 }, { "grad_norm": 0.3549755811691284, "learning_rate": 5.201833810632123e-05, "loss": 0.035, "step": 51280 }, { "grad_norm": 0.32779455184936523, "learning_rate": 5.200181677501933e-05, "loss": 0.026, "step": 51290 }, { "grad_norm": 0.3447127044200897, "learning_rate": 5.1985295224801777e-05, "loss": 0.0323, "step": 51300 }, { "grad_norm": 0.30664750933647156, "learning_rate": 5.196877345747531e-05, "loss": 0.0276, "step": 51310 }, { "grad_norm": 0.345572292804718, "learning_rate": 5.195225147484674e-05, "loss": 0.033, "step": 51320 }, { "grad_norm": 0.40851545333862305, "learning_rate": 5.193572927872288e-05, "loss": 0.0361, "step": 51330 }, { "grad_norm": 0.2690412700176239, "learning_rate": 5.191920687091059e-05, "loss": 0.0318, "step": 51340 }, { "grad_norm": 0.3629217743873596, "learning_rate": 5.1902684253216695e-05, "loss": 0.0243, "step": 51350 }, { "grad_norm": 1.1057788133621216, "learning_rate": 5.1886161427448134e-05, "loss": 0.0289, "step": 51360 }, { "grad_norm": 0.30076393485069275, "learning_rate": 5.186963839541179e-05, "loss": 0.0267, "step": 51370 }, { "grad_norm": 0.3163772523403168, "learning_rate": 5.185311515891459e-05, "loss": 0.0218, "step": 51380 }, { "grad_norm": 0.3408636152744293, "learning_rate": 5.1836591719763516e-05, "loss": 0.0268, "step": 51390 }, { "grad_norm": 0.27679088711738586, "learning_rate": 5.182006807976554e-05, "loss": 0.026, "step": 51400 }, { "grad_norm": 0.40725746750831604, "learning_rate": 5.180354424072763e-05, "loss": 0.0227, "step": 51410 }, { "grad_norm": 1.8462779521942139, "learning_rate": 5.178702020445686e-05, "loss": 0.028, "step": 51420 }, { "grad_norm": 0.9402227997779846, "learning_rate": 5.177049597276024e-05, "loss": 0.0335, "step": 51430 }, { "grad_norm": 0.4140399396419525, "learning_rate": 5.175397154744482e-05, "loss": 0.0284, "step": 51440 }, { "grad_norm": 0.20881246030330658, "learning_rate": 5.1737446930317736e-05, "loss": 0.0186, "step": 51450 }, { "grad_norm": 0.2301422655582428, "learning_rate": 5.172092212318604e-05, "loss": 0.0291, "step": 51460 }, { "grad_norm": 0.3615567684173584, "learning_rate": 5.170439712785692e-05, "loss": 0.0319, "step": 51470 }, { "grad_norm": 0.40429630875587463, "learning_rate": 5.1687871946137467e-05, "loss": 0.0278, "step": 51480 }, { "grad_norm": 0.20526129007339478, "learning_rate": 5.16713465798349e-05, "loss": 0.0272, "step": 51490 }, { "grad_norm": 0.38944268226623535, "learning_rate": 5.165482103075637e-05, "loss": 0.0365, "step": 51500 }, { "grad_norm": 0.24402476847171783, "learning_rate": 5.163829530070909e-05, "loss": 0.0297, "step": 51510 }, { "grad_norm": 0.35745158791542053, "learning_rate": 5.162176939150032e-05, "loss": 0.0361, "step": 51520 }, { "grad_norm": 0.5093696117401123, "learning_rate": 5.160524330493728e-05, "loss": 0.0281, "step": 51530 }, { "grad_norm": 0.38268938660621643, "learning_rate": 5.158871704282724e-05, "loss": 0.03, "step": 51540 }, { "grad_norm": 0.23032692074775696, "learning_rate": 5.15721906069775e-05, "loss": 0.0324, "step": 51550 }, { "grad_norm": 0.2430773228406906, "learning_rate": 5.155566399919536e-05, "loss": 0.0241, "step": 51560 }, { "grad_norm": 0.2769533097743988, "learning_rate": 5.153913722128813e-05, "loss": 0.0257, "step": 51570 }, { "grad_norm": 0.24662552773952484, "learning_rate": 5.1522610275063176e-05, "loss": 0.0221, "step": 51580 }, { "grad_norm": 0.30245324969291687, "learning_rate": 5.150608316232786e-05, "loss": 0.0231, "step": 51590 }, { "grad_norm": 0.24379387497901917, "learning_rate": 5.148955588488955e-05, "loss": 0.0247, "step": 51600 }, { "grad_norm": 0.4330059885978699, "learning_rate": 5.147302844455564e-05, "loss": 0.0372, "step": 51610 }, { "grad_norm": 0.264771968126297, "learning_rate": 5.145650084313356e-05, "loss": 0.032, "step": 51620 }, { "grad_norm": 0.45916318893432617, "learning_rate": 5.1439973082430734e-05, "loss": 0.0323, "step": 51630 }, { "grad_norm": 0.33252209424972534, "learning_rate": 5.142344516425461e-05, "loss": 0.024, "step": 51640 }, { "grad_norm": 0.2509540915489197, "learning_rate": 5.1406917090412673e-05, "loss": 0.0302, "step": 51650 }, { "grad_norm": 0.4192622900009155, "learning_rate": 5.1390388862712404e-05, "loss": 0.0264, "step": 51660 }, { "grad_norm": 0.33821848034858704, "learning_rate": 5.137386048296129e-05, "loss": 0.0255, "step": 51670 }, { "grad_norm": 0.42575201392173767, "learning_rate": 5.135733195296686e-05, "loss": 0.0389, "step": 51680 }, { "grad_norm": 0.35473933815956116, "learning_rate": 5.134080327453664e-05, "loss": 0.0242, "step": 51690 }, { "grad_norm": 0.31233662366867065, "learning_rate": 5.1324274449478204e-05, "loss": 0.0293, "step": 51700 }, { "grad_norm": 0.6825834512710571, "learning_rate": 5.130774547959909e-05, "loss": 0.0344, "step": 51710 }, { "grad_norm": 0.585064172744751, "learning_rate": 5.129121636670692e-05, "loss": 0.0273, "step": 51720 }, { "grad_norm": 0.20392712950706482, "learning_rate": 5.1274687112609246e-05, "loss": 0.0211, "step": 51730 }, { "grad_norm": 0.35153117775917053, "learning_rate": 5.125815771911372e-05, "loss": 0.0283, "step": 51740 }, { "grad_norm": 0.24078796803951263, "learning_rate": 5.1241628188027956e-05, "loss": 0.0254, "step": 51750 }, { "grad_norm": 0.2389001101255417, "learning_rate": 5.1225098521159585e-05, "loss": 0.0225, "step": 51760 }, { "grad_norm": 0.4195527732372284, "learning_rate": 5.12085687203163e-05, "loss": 0.0276, "step": 51770 }, { "grad_norm": 0.2827502191066742, "learning_rate": 5.119203878730575e-05, "loss": 0.026, "step": 51780 }, { "grad_norm": 0.3922252953052521, "learning_rate": 5.117550872393564e-05, "loss": 0.0278, "step": 51790 }, { "grad_norm": 0.6955768465995789, "learning_rate": 5.1158978532013645e-05, "loss": 0.0284, "step": 51800 }, { "grad_norm": 0.22279419004917145, "learning_rate": 5.114244821334751e-05, "loss": 0.0297, "step": 51810 }, { "grad_norm": 0.37805458903312683, "learning_rate": 5.112591776974497e-05, "loss": 0.0273, "step": 51820 }, { "grad_norm": 0.17291679978370667, "learning_rate": 5.110938720301374e-05, "loss": 0.0285, "step": 51830 }, { "grad_norm": 0.4653705656528473, "learning_rate": 5.1092856514961595e-05, "loss": 0.0292, "step": 51840 }, { "grad_norm": 0.4551083445549011, "learning_rate": 5.107632570739631e-05, "loss": 0.0337, "step": 51850 }, { "grad_norm": 0.3224749267101288, "learning_rate": 5.1059794782125645e-05, "loss": 0.0272, "step": 51860 }, { "grad_norm": 0.39537060260772705, "learning_rate": 5.104326374095744e-05, "loss": 0.0381, "step": 51870 }, { "grad_norm": 0.33915975689888, "learning_rate": 5.102673258569947e-05, "loss": 0.026, "step": 51880 }, { "grad_norm": 0.41152361035346985, "learning_rate": 5.101020131815956e-05, "loss": 0.0404, "step": 51890 }, { "grad_norm": 0.3114345669746399, "learning_rate": 5.099366994014556e-05, "loss": 0.0284, "step": 51900 }, { "grad_norm": 0.3859998285770416, "learning_rate": 5.097713845346532e-05, "loss": 0.0379, "step": 51910 }, { "grad_norm": 0.48857566714286804, "learning_rate": 5.096060685992668e-05, "loss": 0.0287, "step": 51920 }, { "grad_norm": 0.38799670338630676, "learning_rate": 5.0944075161337514e-05, "loss": 0.0342, "step": 51930 }, { "grad_norm": 0.48097869753837585, "learning_rate": 5.09275433595057e-05, "loss": 0.0281, "step": 51940 }, { "grad_norm": 0.5175624489784241, "learning_rate": 5.0911011456239157e-05, "loss": 0.0305, "step": 51950 }, { "grad_norm": 0.4325268566608429, "learning_rate": 5.089447945334574e-05, "loss": 0.0403, "step": 51960 }, { "grad_norm": 0.846971869468689, "learning_rate": 5.087794735263343e-05, "loss": 0.028, "step": 51970 }, { "grad_norm": 0.4392889738082886, "learning_rate": 5.086141515591011e-05, "loss": 0.0313, "step": 51980 }, { "grad_norm": 0.46305230259895325, "learning_rate": 5.084488286498371e-05, "loss": 0.0265, "step": 51990 }, { "grad_norm": 0.5204257965087891, "learning_rate": 5.08283504816622e-05, "loss": 0.0378, "step": 52000 }, { "grad_norm": 0.8133208751678467, "learning_rate": 5.081181800775353e-05, "loss": 0.0348, "step": 52010 }, { "grad_norm": 0.25836414098739624, "learning_rate": 5.079528544506565e-05, "loss": 0.0275, "step": 52020 }, { "grad_norm": 0.2666800618171692, "learning_rate": 5.077875279540657e-05, "loss": 0.0214, "step": 52030 }, { "grad_norm": 0.3019461929798126, "learning_rate": 5.076222006058424e-05, "loss": 0.0395, "step": 52040 }, { "grad_norm": 0.37736940383911133, "learning_rate": 5.074568724240667e-05, "loss": 0.0304, "step": 52050 }, { "grad_norm": 0.5859869122505188, "learning_rate": 5.072915434268187e-05, "loss": 0.0291, "step": 52060 }, { "grad_norm": 0.6148273944854736, "learning_rate": 5.071262136321785e-05, "loss": 0.0333, "step": 52070 }, { "grad_norm": 0.3598586916923523, "learning_rate": 5.069608830582262e-05, "loss": 0.0376, "step": 52080 }, { "grad_norm": 0.2905736565589905, "learning_rate": 5.067955517230423e-05, "loss": 0.0302, "step": 52090 }, { "grad_norm": 0.34741106629371643, "learning_rate": 5.06630219644707e-05, "loss": 0.0338, "step": 52100 }, { "grad_norm": 0.3324088454246521, "learning_rate": 5.064648868413008e-05, "loss": 0.0306, "step": 52110 }, { "grad_norm": 0.35944029688835144, "learning_rate": 5.062995533309044e-05, "loss": 0.0274, "step": 52120 }, { "grad_norm": 0.32104089856147766, "learning_rate": 5.061342191315983e-05, "loss": 0.035, "step": 52130 }, { "grad_norm": 0.46122729778289795, "learning_rate": 5.0596888426146325e-05, "loss": 0.0247, "step": 52140 }, { "grad_norm": 0.22782915830612183, "learning_rate": 5.058035487385798e-05, "loss": 0.0192, "step": 52150 }, { "grad_norm": 0.1585128754377365, "learning_rate": 5.0563821258102915e-05, "loss": 0.0186, "step": 52160 }, { "grad_norm": 0.28189101815223694, "learning_rate": 5.05472875806892e-05, "loss": 0.0228, "step": 52170 }, { "grad_norm": 0.3881937563419342, "learning_rate": 5.053075384342494e-05, "loss": 0.0265, "step": 52180 }, { "grad_norm": 0.24970926344394684, "learning_rate": 5.0514220048118234e-05, "loss": 0.0278, "step": 52190 }, { "grad_norm": 0.6409633755683899, "learning_rate": 5.049768619657721e-05, "loss": 0.027, "step": 52200 }, { "grad_norm": 0.20097792148590088, "learning_rate": 5.048115229060996e-05, "loss": 0.0255, "step": 52210 }, { "grad_norm": 0.3083018660545349, "learning_rate": 5.046461833202462e-05, "loss": 0.0284, "step": 52220 }, { "grad_norm": 0.34600961208343506, "learning_rate": 5.044808432262932e-05, "loss": 0.0305, "step": 52230 }, { "grad_norm": 0.2771706283092499, "learning_rate": 5.043155026423219e-05, "loss": 0.031, "step": 52240 }, { "grad_norm": 0.3887816369533539, "learning_rate": 5.041501615864137e-05, "loss": 0.0251, "step": 52250 }, { "grad_norm": 0.267182320356369, "learning_rate": 5.039848200766502e-05, "loss": 0.0196, "step": 52260 }, { "grad_norm": 0.39083534479141235, "learning_rate": 5.038194781311125e-05, "loss": 0.0203, "step": 52270 }, { "grad_norm": 0.34880250692367554, "learning_rate": 5.0365413576788264e-05, "loss": 0.0281, "step": 52280 }, { "grad_norm": 0.3998148441314697, "learning_rate": 5.0348879300504194e-05, "loss": 0.0356, "step": 52290 }, { "grad_norm": 0.25125136971473694, "learning_rate": 5.033234498606721e-05, "loss": 0.0272, "step": 52300 }, { "grad_norm": 0.2305493950843811, "learning_rate": 5.0315810635285476e-05, "loss": 0.0284, "step": 52310 }, { "grad_norm": 0.5433259010314941, "learning_rate": 5.0299276249967164e-05, "loss": 0.0286, "step": 52320 }, { "grad_norm": 0.32730385661125183, "learning_rate": 5.0282741831920454e-05, "loss": 0.0376, "step": 52330 }, { "grad_norm": 0.34256380796432495, "learning_rate": 5.026620738295351e-05, "loss": 0.0343, "step": 52340 }, { "grad_norm": 0.428774893283844, "learning_rate": 5.024967290487455e-05, "loss": 0.0258, "step": 52350 }, { "grad_norm": 0.30434080958366394, "learning_rate": 5.023313839949172e-05, "loss": 0.0258, "step": 52360 }, { "grad_norm": 0.31601300835609436, "learning_rate": 5.0216603868613224e-05, "loss": 0.0278, "step": 52370 }, { "grad_norm": 0.5507542490959167, "learning_rate": 5.020006931404726e-05, "loss": 0.0379, "step": 52380 }, { "grad_norm": 0.21967649459838867, "learning_rate": 5.0183534737602014e-05, "loss": 0.0269, "step": 52390 }, { "grad_norm": 0.22468428313732147, "learning_rate": 5.0167000141085686e-05, "loss": 0.0267, "step": 52400 }, { "grad_norm": 0.2983384430408478, "learning_rate": 5.015046552630647e-05, "loss": 0.0259, "step": 52410 }, { "grad_norm": 0.5002513527870178, "learning_rate": 5.013393089507258e-05, "loss": 0.0364, "step": 52420 }, { "grad_norm": 0.26650863885879517, "learning_rate": 5.011739624919219e-05, "loss": 0.0238, "step": 52430 }, { "grad_norm": 0.37762725353240967, "learning_rate": 5.0100861590473545e-05, "loss": 0.0234, "step": 52440 }, { "grad_norm": 0.3825562000274658, "learning_rate": 5.008432692072482e-05, "loss": 0.0274, "step": 52450 }, { "grad_norm": 0.2842496931552887, "learning_rate": 5.006779224175423e-05, "loss": 0.0229, "step": 52460 }, { "grad_norm": 0.34786105155944824, "learning_rate": 5.005125755536998e-05, "loss": 0.0299, "step": 52470 }, { "grad_norm": 0.23801060020923615, "learning_rate": 5.003472286338029e-05, "loss": 0.0273, "step": 52480 }, { "grad_norm": 0.26159173250198364, "learning_rate": 5.001818816759335e-05, "loss": 0.0375, "step": 52490 }, { "grad_norm": 0.4542264938354492, "learning_rate": 5.000165346981738e-05, "loss": 0.0301, "step": 52500 }, { "grad_norm": 0.26013413071632385, "learning_rate": 4.9985118771860595e-05, "loss": 0.035, "step": 52510 }, { "grad_norm": 0.2235746830701828, "learning_rate": 4.996858407553119e-05, "loss": 0.026, "step": 52520 }, { "grad_norm": 0.2655770778656006, "learning_rate": 4.99520493826374e-05, "loss": 0.03, "step": 52530 }, { "grad_norm": 0.4080054759979248, "learning_rate": 4.9935514694987384e-05, "loss": 0.0422, "step": 52540 }, { "grad_norm": 0.3504994809627533, "learning_rate": 4.991898001438941e-05, "loss": 0.0319, "step": 52550 }, { "grad_norm": 0.2554074823856354, "learning_rate": 4.990244534265166e-05, "loss": 0.0226, "step": 52560 }, { "grad_norm": 0.2776630222797394, "learning_rate": 4.988591068158234e-05, "loss": 0.0266, "step": 52570 }, { "grad_norm": 0.441398948431015, "learning_rate": 4.986937603298965e-05, "loss": 0.0318, "step": 52580 }, { "grad_norm": 0.26329395174980164, "learning_rate": 4.985284139868181e-05, "loss": 0.0233, "step": 52590 }, { "grad_norm": 0.25635474920272827, "learning_rate": 4.9836306780467e-05, "loss": 0.0298, "step": 52600 }, { "grad_norm": 0.3350590765476227, "learning_rate": 4.981977218015343e-05, "loss": 0.0238, "step": 52610 }, { "grad_norm": 1.0804805755615234, "learning_rate": 4.980323759954932e-05, "loss": 0.0347, "step": 52620 }, { "grad_norm": 0.23199212551116943, "learning_rate": 4.978670304046284e-05, "loss": 0.0224, "step": 52630 }, { "grad_norm": 1.1181426048278809, "learning_rate": 4.9770168504702195e-05, "loss": 0.0287, "step": 52640 }, { "grad_norm": 0.3854381740093231, "learning_rate": 4.975363399407559e-05, "loss": 0.0382, "step": 52650 }, { "grad_norm": 0.2977476119995117, "learning_rate": 4.9737099510391194e-05, "loss": 0.0338, "step": 52660 }, { "grad_norm": 0.3403710722923279, "learning_rate": 4.972056505545719e-05, "loss": 0.0306, "step": 52670 }, { "grad_norm": 0.34063032269477844, "learning_rate": 4.9704030631081775e-05, "loss": 0.0291, "step": 52680 }, { "grad_norm": 0.20553049445152283, "learning_rate": 4.968749623907313e-05, "loss": 0.0216, "step": 52690 }, { "grad_norm": 0.2879527807235718, "learning_rate": 4.967096188123943e-05, "loss": 0.035, "step": 52700 }, { "grad_norm": 0.29595598578453064, "learning_rate": 4.965442755938884e-05, "loss": 0.0356, "step": 52710 }, { "grad_norm": 0.157038152217865, "learning_rate": 4.963789327532954e-05, "loss": 0.0283, "step": 52720 }, { "grad_norm": 0.2776145339012146, "learning_rate": 4.962135903086968e-05, "loss": 0.0236, "step": 52730 }, { "grad_norm": 0.35752084851264954, "learning_rate": 4.960482482781741e-05, "loss": 0.0241, "step": 52740 }, { "grad_norm": 0.2698407471179962, "learning_rate": 4.958829066798092e-05, "loss": 0.0236, "step": 52750 }, { "grad_norm": 0.5500395894050598, "learning_rate": 4.957175655316835e-05, "loss": 0.0287, "step": 52760 }, { "grad_norm": 0.33694741129875183, "learning_rate": 4.9555222485187834e-05, "loss": 0.038, "step": 52770 }, { "grad_norm": 0.3077342212200165, "learning_rate": 4.953868846584752e-05, "loss": 0.0295, "step": 52780 }, { "grad_norm": 0.23254507780075073, "learning_rate": 4.9522154496955544e-05, "loss": 0.0422, "step": 52790 }, { "grad_norm": 0.36538761854171753, "learning_rate": 4.950562058032001e-05, "loss": 0.0325, "step": 52800 }, { "grad_norm": 0.6737560033798218, "learning_rate": 4.948908671774909e-05, "loss": 0.0332, "step": 52810 }, { "grad_norm": 0.7088841795921326, "learning_rate": 4.947255291105087e-05, "loss": 0.0221, "step": 52820 }, { "grad_norm": 0.36995840072631836, "learning_rate": 4.945601916203348e-05, "loss": 0.0316, "step": 52830 }, { "grad_norm": 0.29770052433013916, "learning_rate": 4.9439485472505016e-05, "loss": 0.0294, "step": 52840 }, { "grad_norm": 0.49156996607780457, "learning_rate": 4.942295184427356e-05, "loss": 0.0316, "step": 52850 }, { "grad_norm": 0.37941285967826843, "learning_rate": 4.9406418279147234e-05, "loss": 0.0342, "step": 52860 }, { "grad_norm": 0.44255152344703674, "learning_rate": 4.9389884778934116e-05, "loss": 0.0334, "step": 52870 }, { "grad_norm": 0.34804242849349976, "learning_rate": 4.937335134544228e-05, "loss": 0.0315, "step": 52880 }, { "grad_norm": 0.3515697121620178, "learning_rate": 4.935681798047979e-05, "loss": 0.0327, "step": 52890 }, { "grad_norm": 0.38687631487846375, "learning_rate": 4.934028468585473e-05, "loss": 0.0305, "step": 52900 }, { "grad_norm": 0.24284569919109344, "learning_rate": 4.932375146337514e-05, "loss": 0.0245, "step": 52910 }, { "grad_norm": 0.2289842963218689, "learning_rate": 4.930721831484908e-05, "loss": 0.0275, "step": 52920 }, { "grad_norm": 0.26434820890426636, "learning_rate": 4.929068524208456e-05, "loss": 0.029, "step": 52930 }, { "grad_norm": 0.3281991183757782, "learning_rate": 4.927415224688965e-05, "loss": 0.0401, "step": 52940 }, { "grad_norm": 0.2865818440914154, "learning_rate": 4.9257619331072375e-05, "loss": 0.0377, "step": 52950 }, { "grad_norm": 0.3179178535938263, "learning_rate": 4.924108649644072e-05, "loss": 0.0261, "step": 52960 }, { "grad_norm": 0.37451696395874023, "learning_rate": 4.922455374480271e-05, "loss": 0.0331, "step": 52970 }, { "grad_norm": 0.2831512987613678, "learning_rate": 4.920802107796634e-05, "loss": 0.0321, "step": 52980 }, { "grad_norm": 0.39784637093544006, "learning_rate": 4.9191488497739564e-05, "loss": 0.0229, "step": 52990 }, { "grad_norm": 0.21460629999637604, "learning_rate": 4.917495600593042e-05, "loss": 0.0227, "step": 53000 }, { "grad_norm": 0.27619069814682007, "learning_rate": 4.915842360434684e-05, "loss": 0.0306, "step": 53010 }, { "grad_norm": 0.4455167353153229, "learning_rate": 4.914189129479679e-05, "loss": 0.0336, "step": 53020 }, { "grad_norm": 0.4914616346359253, "learning_rate": 4.912535907908822e-05, "loss": 0.0264, "step": 53030 }, { "grad_norm": 0.4117039740085602, "learning_rate": 4.910882695902907e-05, "loss": 0.0336, "step": 53040 }, { "grad_norm": 0.21064288914203644, "learning_rate": 4.9092294936427234e-05, "loss": 0.035, "step": 53050 }, { "grad_norm": 0.7510010600090027, "learning_rate": 4.9075763013090685e-05, "loss": 0.0366, "step": 53060 }, { "grad_norm": 0.5405364036560059, "learning_rate": 4.9059231190827305e-05, "loss": 0.0354, "step": 53070 }, { "grad_norm": 0.3396366536617279, "learning_rate": 4.904269947144499e-05, "loss": 0.0453, "step": 53080 }, { "grad_norm": 0.3323848843574524, "learning_rate": 4.902616785675161e-05, "loss": 0.0305, "step": 53090 }, { "grad_norm": 0.4701108932495117, "learning_rate": 4.900963634855505e-05, "loss": 0.0288, "step": 53100 }, { "grad_norm": 0.2741495668888092, "learning_rate": 4.8993104948663173e-05, "loss": 0.0258, "step": 53110 }, { "grad_norm": 0.3104530870914459, "learning_rate": 4.8976573658883823e-05, "loss": 0.0296, "step": 53120 }, { "grad_norm": 0.310642272233963, "learning_rate": 4.896004248102484e-05, "loss": 0.0275, "step": 53130 }, { "grad_norm": 0.3409620523452759, "learning_rate": 4.8943511416894045e-05, "loss": 0.0203, "step": 53140 }, { "grad_norm": 0.4269859492778778, "learning_rate": 4.892698046829926e-05, "loss": 0.0298, "step": 53150 }, { "grad_norm": 0.3300396800041199, "learning_rate": 4.8910449637048275e-05, "loss": 0.0343, "step": 53160 }, { "grad_norm": 0.5225651264190674, "learning_rate": 4.8893918924948884e-05, "loss": 0.0363, "step": 53170 }, { "grad_norm": 0.41589462757110596, "learning_rate": 4.8877388333808835e-05, "loss": 0.0354, "step": 53180 }, { "grad_norm": 0.47303706407546997, "learning_rate": 4.8860857865435925e-05, "loss": 0.0343, "step": 53190 }, { "grad_norm": 0.4157014787197113, "learning_rate": 4.884432752163789e-05, "loss": 0.0338, "step": 53200 }, { "grad_norm": 0.4337617754936218, "learning_rate": 4.8827797304222466e-05, "loss": 0.0309, "step": 53210 }, { "grad_norm": 0.32623401284217834, "learning_rate": 4.8811267214997366e-05, "loss": 0.0245, "step": 53220 }, { "grad_norm": 0.41897913813591003, "learning_rate": 4.879473725577029e-05, "loss": 0.028, "step": 53230 }, { "grad_norm": 0.33609360456466675, "learning_rate": 4.877820742834891e-05, "loss": 0.0259, "step": 53240 }, { "grad_norm": 0.38372334837913513, "learning_rate": 4.8761677734540956e-05, "loss": 0.0246, "step": 53250 }, { "grad_norm": 0.43324610590934753, "learning_rate": 4.874514817615407e-05, "loss": 0.025, "step": 53260 }, { "grad_norm": 0.4343586564064026, "learning_rate": 4.872861875499588e-05, "loss": 0.0233, "step": 53270 }, { "grad_norm": 0.3728191554546356, "learning_rate": 4.871208947287404e-05, "loss": 0.0279, "step": 53280 }, { "grad_norm": 0.29862335324287415, "learning_rate": 4.869556033159615e-05, "loss": 0.0249, "step": 53290 }, { "grad_norm": 0.372236043214798, "learning_rate": 4.8679031332969835e-05, "loss": 0.0296, "step": 53300 }, { "grad_norm": 0.30127325654029846, "learning_rate": 4.8662502478802635e-05, "loss": 0.0323, "step": 53310 }, { "grad_norm": 0.31000444293022156, "learning_rate": 4.8645973770902176e-05, "loss": 0.0265, "step": 53320 }, { "grad_norm": 0.29888916015625, "learning_rate": 4.862944521107599e-05, "loss": 0.0284, "step": 53330 }, { "grad_norm": 0.4507989287376404, "learning_rate": 4.861291680113162e-05, "loss": 0.0307, "step": 53340 }, { "grad_norm": 0.3002409338951111, "learning_rate": 4.8596388542876563e-05, "loss": 0.0275, "step": 53350 }, { "grad_norm": 0.2684943377971649, "learning_rate": 4.857986043811836e-05, "loss": 0.0347, "step": 53360 }, { "grad_norm": 0.3085690438747406, "learning_rate": 4.856333248866448e-05, "loss": 0.0272, "step": 53370 }, { "grad_norm": 0.26381680369377136, "learning_rate": 4.85468046963224e-05, "loss": 0.0264, "step": 53380 }, { "grad_norm": 0.4475303888320923, "learning_rate": 4.853027706289957e-05, "loss": 0.0308, "step": 53390 }, { "grad_norm": 0.37400373816490173, "learning_rate": 4.851374959020344e-05, "loss": 0.0215, "step": 53400 }, { "grad_norm": 0.45199349522590637, "learning_rate": 4.849722228004141e-05, "loss": 0.0325, "step": 53410 }, { "grad_norm": 0.32667794823646545, "learning_rate": 4.848069513422091e-05, "loss": 0.0223, "step": 53420 }, { "grad_norm": 0.3511837124824524, "learning_rate": 4.8464168154549294e-05, "loss": 0.0276, "step": 53430 }, { "grad_norm": 0.2688120901584625, "learning_rate": 4.844764134283392e-05, "loss": 0.0319, "step": 53440 }, { "grad_norm": 0.2619410455226898, "learning_rate": 4.843111470088218e-05, "loss": 0.0251, "step": 53450 }, { "grad_norm": 0.26017218828201294, "learning_rate": 4.841458823050137e-05, "loss": 0.0338, "step": 53460 }, { "grad_norm": 0.36478540301322937, "learning_rate": 4.8398061933498816e-05, "loss": 0.0233, "step": 53470 }, { "grad_norm": 0.32084763050079346, "learning_rate": 4.83815358116818e-05, "loss": 0.024, "step": 53480 }, { "grad_norm": 0.3249172568321228, "learning_rate": 4.8365009866857584e-05, "loss": 0.0329, "step": 53490 }, { "grad_norm": 0.22486627101898193, "learning_rate": 4.834848410083342e-05, "loss": 0.0265, "step": 53500 }, { "grad_norm": 0.496351957321167, "learning_rate": 4.833195851541657e-05, "loss": 0.0395, "step": 53510 }, { "grad_norm": 0.41327157616615295, "learning_rate": 4.831543311241423e-05, "loss": 0.0282, "step": 53520 }, { "grad_norm": 0.6069287657737732, "learning_rate": 4.829890789363359e-05, "loss": 0.0325, "step": 53530 }, { "grad_norm": 0.42001816630363464, "learning_rate": 4.828238286088182e-05, "loss": 0.0238, "step": 53540 }, { "grad_norm": 0.3098175525665283, "learning_rate": 4.826585801596609e-05, "loss": 0.0362, "step": 53550 }, { "grad_norm": 0.5693561434745789, "learning_rate": 4.8249333360693486e-05, "loss": 0.0268, "step": 53560 }, { "grad_norm": 0.34483271837234497, "learning_rate": 4.8232808896871185e-05, "loss": 0.0372, "step": 53570 }, { "grad_norm": 0.5204162001609802, "learning_rate": 4.821628462630624e-05, "loss": 0.0331, "step": 53580 }, { "grad_norm": 0.4094139039516449, "learning_rate": 4.8199760550805735e-05, "loss": 0.032, "step": 53590 }, { "grad_norm": 0.46685314178466797, "learning_rate": 4.818323667217669e-05, "loss": 0.0329, "step": 53600 }, { "grad_norm": 0.36648625135421753, "learning_rate": 4.8166712992226165e-05, "loss": 0.0285, "step": 53610 }, { "grad_norm": 0.3136015832424164, "learning_rate": 4.8150189512761154e-05, "loss": 0.0276, "step": 53620 }, { "grad_norm": 0.5329737067222595, "learning_rate": 4.813366623558862e-05, "loss": 0.0277, "step": 53630 }, { "grad_norm": 0.4552839696407318, "learning_rate": 4.811714316251554e-05, "loss": 0.028, "step": 53640 }, { "grad_norm": 0.46989938616752625, "learning_rate": 4.810062029534886e-05, "loss": 0.0351, "step": 53650 }, { "grad_norm": 0.36841318011283875, "learning_rate": 4.8084097635895505e-05, "loss": 0.0339, "step": 53660 }, { "grad_norm": 0.34072983264923096, "learning_rate": 4.8067575185962335e-05, "loss": 0.0296, "step": 53670 }, { "grad_norm": 0.313167929649353, "learning_rate": 4.805105294735625e-05, "loss": 0.0297, "step": 53680 }, { "grad_norm": 0.20722773671150208, "learning_rate": 4.803453092188406e-05, "loss": 0.0287, "step": 53690 }, { "grad_norm": 0.5080579519271851, "learning_rate": 4.801800911135263e-05, "loss": 0.0242, "step": 53700 }, { "grad_norm": 0.28931334614753723, "learning_rate": 4.8001487517568744e-05, "loss": 0.0255, "step": 53710 }, { "grad_norm": 0.5337929129600525, "learning_rate": 4.798496614233918e-05, "loss": 0.0308, "step": 53720 }, { "grad_norm": 0.40024900436401367, "learning_rate": 4.796844498747069e-05, "loss": 0.0368, "step": 53730 }, { "grad_norm": 0.2631971836090088, "learning_rate": 4.795192405477e-05, "loss": 0.0257, "step": 53740 }, { "grad_norm": 0.4212474226951599, "learning_rate": 4.79354033460438e-05, "loss": 0.0249, "step": 53750 }, { "grad_norm": 0.35207951068878174, "learning_rate": 4.79188828630988e-05, "loss": 0.0348, "step": 53760 }, { "grad_norm": 0.3568168580532074, "learning_rate": 4.790236260774165e-05, "loss": 0.0221, "step": 53770 }, { "grad_norm": 0.496165007352829, "learning_rate": 4.788584258177896e-05, "loss": 0.0221, "step": 53780 }, { "grad_norm": 0.30478984117507935, "learning_rate": 4.7869322787017356e-05, "loss": 0.0309, "step": 53790 }, { "grad_norm": 0.3392786979675293, "learning_rate": 4.78528032252634e-05, "loss": 0.02, "step": 53800 }, { "grad_norm": 0.2901761829853058, "learning_rate": 4.783628389832366e-05, "loss": 0.0285, "step": 53810 }, { "grad_norm": 0.2384856790304184, "learning_rate": 4.7819764808004644e-05, "loss": 0.0247, "step": 53820 }, { "grad_norm": 0.36793428659439087, "learning_rate": 4.780324595611289e-05, "loss": 0.0228, "step": 53830 }, { "grad_norm": 0.3225470781326294, "learning_rate": 4.7786727344454854e-05, "loss": 0.0356, "step": 53840 }, { "grad_norm": 0.2505476474761963, "learning_rate": 4.777020897483697e-05, "loss": 0.0337, "step": 53850 }, { "grad_norm": 0.7983855605125427, "learning_rate": 4.775369084906569e-05, "loss": 0.0314, "step": 53860 }, { "grad_norm": 0.34481996297836304, "learning_rate": 4.773717296894742e-05, "loss": 0.036, "step": 53870 }, { "grad_norm": 0.25738754868507385, "learning_rate": 4.772065533628848e-05, "loss": 0.0274, "step": 53880 }, { "grad_norm": 0.3449562191963196, "learning_rate": 4.7704137952895265e-05, "loss": 0.0248, "step": 53890 }, { "grad_norm": 0.4035487473011017, "learning_rate": 4.768762082057407e-05, "loss": 0.0257, "step": 53900 }, { "grad_norm": 0.6028934121131897, "learning_rate": 4.7671103941131186e-05, "loss": 0.0352, "step": 53910 }, { "grad_norm": 0.32780683040618896, "learning_rate": 4.765458731637287e-05, "loss": 0.0275, "step": 53920 }, { "grad_norm": 0.4051375985145569, "learning_rate": 4.763807094810537e-05, "loss": 0.0411, "step": 53930 }, { "grad_norm": 0.3228916525840759, "learning_rate": 4.762155483813487e-05, "loss": 0.0365, "step": 53940 }, { "grad_norm": 0.6029555201530457, "learning_rate": 4.760503898826754e-05, "loss": 0.0373, "step": 53950 }, { "grad_norm": 0.32774919271469116, "learning_rate": 4.758852340030957e-05, "loss": 0.0376, "step": 53960 }, { "grad_norm": 0.2845081388950348, "learning_rate": 4.757200807606705e-05, "loss": 0.031, "step": 53970 }, { "grad_norm": 0.36594563722610474, "learning_rate": 4.755549301734607e-05, "loss": 0.0254, "step": 53980 }, { "grad_norm": 0.43182194232940674, "learning_rate": 4.753897822595271e-05, "loss": 0.0337, "step": 53990 }, { "grad_norm": 0.47035619616508484, "learning_rate": 4.7522463703692994e-05, "loss": 0.0264, "step": 54000 }, { "grad_norm": 0.3785128891468048, "learning_rate": 4.75059494523729e-05, "loss": 0.0305, "step": 54010 }, { "grad_norm": 0.34642431139945984, "learning_rate": 4.748943547379843e-05, "loss": 0.03, "step": 54020 }, { "grad_norm": 0.43082913756370544, "learning_rate": 4.7472921769775535e-05, "loss": 0.0451, "step": 54030 }, { "grad_norm": 0.37985119223594666, "learning_rate": 4.745640834211011e-05, "loss": 0.0336, "step": 54040 }, { "grad_norm": 0.27916210889816284, "learning_rate": 4.7439895192608054e-05, "loss": 0.0298, "step": 54050 }, { "grad_norm": 0.48753559589385986, "learning_rate": 4.7423382323075186e-05, "loss": 0.0313, "step": 54060 }, { "grad_norm": 0.3633037507534027, "learning_rate": 4.740686973531737e-05, "loss": 0.027, "step": 54070 }, { "grad_norm": 0.4080250561237335, "learning_rate": 4.7390357431140375e-05, "loss": 0.0213, "step": 54080 }, { "grad_norm": 0.2797335386276245, "learning_rate": 4.737384541234998e-05, "loss": 0.0261, "step": 54090 }, { "grad_norm": 0.31495487689971924, "learning_rate": 4.735733368075189e-05, "loss": 0.0236, "step": 54100 }, { "grad_norm": 0.303155779838562, "learning_rate": 4.734082223815182e-05, "loss": 0.0313, "step": 54110 }, { "grad_norm": 0.24128273129463196, "learning_rate": 4.7324311086355436e-05, "loss": 0.0246, "step": 54120 }, { "grad_norm": 0.42299845814704895, "learning_rate": 4.730780022716836e-05, "loss": 0.0366, "step": 54130 }, { "grad_norm": 0.38666048645973206, "learning_rate": 4.729128966239619e-05, "loss": 0.026, "step": 54140 }, { "grad_norm": 0.3009757697582245, "learning_rate": 4.7274779393844536e-05, "loss": 0.028, "step": 54150 }, { "grad_norm": 0.576839029788971, "learning_rate": 4.725826942331891e-05, "loss": 0.0459, "step": 54160 }, { "grad_norm": 0.4731467068195343, "learning_rate": 4.724175975262481e-05, "loss": 0.0299, "step": 54170 }, { "grad_norm": 0.25073346495628357, "learning_rate": 4.722525038356772e-05, "loss": 0.0283, "step": 54180 }, { "grad_norm": 0.3514212667942047, "learning_rate": 4.7208741317953075e-05, "loss": 0.0246, "step": 54190 }, { "grad_norm": 0.2582280933856964, "learning_rate": 4.7192232557586264e-05, "loss": 0.0337, "step": 54200 }, { "grad_norm": 0.5927706956863403, "learning_rate": 4.717572410427269e-05, "loss": 0.0374, "step": 54210 }, { "grad_norm": 0.2973622977733612, "learning_rate": 4.71592159598177e-05, "loss": 0.0312, "step": 54220 }, { "grad_norm": 0.357306569814682, "learning_rate": 4.714270812602657e-05, "loss": 0.0279, "step": 54230 }, { "grad_norm": 0.36562401056289673, "learning_rate": 4.712620060470458e-05, "loss": 0.0355, "step": 54240 }, { "grad_norm": 0.40497249364852905, "learning_rate": 4.7109693397656986e-05, "loss": 0.0267, "step": 54250 }, { "grad_norm": 1.039495825767517, "learning_rate": 4.709318650668894e-05, "loss": 0.0261, "step": 54260 }, { "grad_norm": 0.2229836881160736, "learning_rate": 4.7076679933605676e-05, "loss": 0.0289, "step": 54270 }, { "grad_norm": 0.3240336775779724, "learning_rate": 4.70601736802123e-05, "loss": 0.0303, "step": 54280 }, { "grad_norm": 0.2907140254974365, "learning_rate": 4.704366774831391e-05, "loss": 0.0382, "step": 54290 }, { "grad_norm": 0.4320383369922638, "learning_rate": 4.702716213971557e-05, "loss": 0.0303, "step": 54300 }, { "grad_norm": 0.6947411894798279, "learning_rate": 4.70106568562223e-05, "loss": 0.0345, "step": 54310 }, { "grad_norm": 0.2743664085865021, "learning_rate": 4.6994151899639105e-05, "loss": 0.0306, "step": 54320 }, { "grad_norm": 0.29208648204803467, "learning_rate": 4.697764727177093e-05, "loss": 0.0319, "step": 54330 }, { "grad_norm": 0.30564799904823303, "learning_rate": 4.696114297442272e-05, "loss": 0.0256, "step": 54340 }, { "grad_norm": 0.27260756492614746, "learning_rate": 4.694463900939933e-05, "loss": 0.0268, "step": 54350 }, { "grad_norm": 0.2960275411605835, "learning_rate": 4.692813537850564e-05, "loss": 0.0316, "step": 54360 }, { "grad_norm": 0.5757256746292114, "learning_rate": 4.6911632083546445e-05, "loss": 0.0273, "step": 54370 }, { "grad_norm": 0.33070826530456543, "learning_rate": 4.6895129126326525e-05, "loss": 0.0402, "step": 54380 }, { "grad_norm": 0.8919933438301086, "learning_rate": 4.687862650865059e-05, "loss": 0.0297, "step": 54390 }, { "grad_norm": 0.27049294114112854, "learning_rate": 4.686212423232339e-05, "loss": 0.0298, "step": 54400 }, { "grad_norm": 0.405985563993454, "learning_rate": 4.6845622299149565e-05, "loss": 0.0246, "step": 54410 }, { "grad_norm": 0.3516412675380707, "learning_rate": 4.682912071093374e-05, "loss": 0.0277, "step": 54420 }, { "grad_norm": 0.4103486239910126, "learning_rate": 4.681261946948052e-05, "loss": 0.0292, "step": 54430 }, { "grad_norm": 0.13284172117710114, "learning_rate": 4.679611857659443e-05, "loss": 0.0363, "step": 54440 }, { "grad_norm": 0.2286580353975296, "learning_rate": 4.6779618034079975e-05, "loss": 0.0333, "step": 54450 }, { "grad_norm": 0.2923797070980072, "learning_rate": 4.6763117843741675e-05, "loss": 0.0321, "step": 54460 }, { "grad_norm": 0.3819100856781006, "learning_rate": 4.674661800738394e-05, "loss": 0.0215, "step": 54470 }, { "grad_norm": 0.3868652284145355, "learning_rate": 4.673011852681117e-05, "loss": 0.026, "step": 54480 }, { "grad_norm": 0.2139601707458496, "learning_rate": 4.671361940382772e-05, "loss": 0.0293, "step": 54490 }, { "grad_norm": 0.34609219431877136, "learning_rate": 4.669712064023791e-05, "loss": 0.0261, "step": 54500 }, { "grad_norm": 0.427232950925827, "learning_rate": 4.668062223784602e-05, "loss": 0.0278, "step": 54510 }, { "grad_norm": 0.3935176134109497, "learning_rate": 4.6664124198456274e-05, "loss": 0.0268, "step": 54520 }, { "grad_norm": 0.2909163236618042, "learning_rate": 4.6647626523872905e-05, "loss": 0.0266, "step": 54530 }, { "grad_norm": 0.34692448377609253, "learning_rate": 4.6631129215900066e-05, "loss": 0.0289, "step": 54540 }, { "grad_norm": 0.4690035581588745, "learning_rate": 4.661463227634187e-05, "loss": 0.0385, "step": 54550 }, { "grad_norm": 0.5446712374687195, "learning_rate": 4.659813570700238e-05, "loss": 0.0262, "step": 54560 }, { "grad_norm": 0.45751649141311646, "learning_rate": 4.658163950968568e-05, "loss": 0.035, "step": 54570 }, { "grad_norm": 0.44779127836227417, "learning_rate": 4.6565143686195726e-05, "loss": 0.0286, "step": 54580 }, { "grad_norm": 0.3467075526714325, "learning_rate": 4.654864823833649e-05, "loss": 0.0261, "step": 54590 }, { "grad_norm": 0.3304745852947235, "learning_rate": 4.65321531679119e-05, "loss": 0.0254, "step": 54600 }, { "grad_norm": 0.2580978274345398, "learning_rate": 4.6515658476725834e-05, "loss": 0.0384, "step": 54610 }, { "grad_norm": 0.36343586444854736, "learning_rate": 4.649916416658212e-05, "loss": 0.0342, "step": 54620 }, { "grad_norm": 0.23835106194019318, "learning_rate": 4.648267023928455e-05, "loss": 0.03, "step": 54630 }, { "grad_norm": 0.38779833912849426, "learning_rate": 4.646617669663687e-05, "loss": 0.0236, "step": 54640 }, { "grad_norm": 0.3485116958618164, "learning_rate": 4.644968354044277e-05, "loss": 0.0255, "step": 54650 }, { "grad_norm": 0.5946650505065918, "learning_rate": 4.643319077250596e-05, "loss": 0.0328, "step": 54660 }, { "grad_norm": 0.2894023358821869, "learning_rate": 4.6416698394630055e-05, "loss": 0.0301, "step": 54670 }, { "grad_norm": 0.3310626745223999, "learning_rate": 4.640020640861862e-05, "loss": 0.0221, "step": 54680 }, { "grad_norm": 0.33898890018463135, "learning_rate": 4.6383714816275195e-05, "loss": 0.033, "step": 54690 }, { "grad_norm": 0.3526555895805359, "learning_rate": 4.636722361940329e-05, "loss": 0.0244, "step": 54700 }, { "grad_norm": 0.5452800989151001, "learning_rate": 4.635073281980632e-05, "loss": 0.027, "step": 54710 }, { "grad_norm": 0.30170780420303345, "learning_rate": 4.633424241928775e-05, "loss": 0.0364, "step": 54720 }, { "grad_norm": 0.37543877959251404, "learning_rate": 4.6317752419650914e-05, "loss": 0.0247, "step": 54730 }, { "grad_norm": 0.2745825946331024, "learning_rate": 4.630126282269914e-05, "loss": 0.0331, "step": 54740 }, { "grad_norm": 0.2867138087749481, "learning_rate": 4.628477363023569e-05, "loss": 0.023, "step": 54750 }, { "grad_norm": 0.4377797544002533, "learning_rate": 4.626828484406383e-05, "loss": 0.0247, "step": 54760 }, { "grad_norm": 0.2944487929344177, "learning_rate": 4.625179646598669e-05, "loss": 0.0319, "step": 54770 }, { "grad_norm": 0.2723098695278168, "learning_rate": 4.6235308497807487e-05, "loss": 0.0205, "step": 54780 }, { "grad_norm": 0.6713022589683533, "learning_rate": 4.6218820941329276e-05, "loss": 0.0285, "step": 54790 }, { "grad_norm": 0.2162945717573166, "learning_rate": 4.620233379835513e-05, "loss": 0.0215, "step": 54800 }, { "grad_norm": 0.25302207469940186, "learning_rate": 4.618584707068804e-05, "loss": 0.0328, "step": 54810 }, { "grad_norm": 0.39613762497901917, "learning_rate": 4.616936076013099e-05, "loss": 0.0301, "step": 54820 }, { "grad_norm": 0.2497977465391159, "learning_rate": 4.615287486848688e-05, "loss": 0.0252, "step": 54830 }, { "grad_norm": 0.3850354254245758, "learning_rate": 4.6136389397558584e-05, "loss": 0.0259, "step": 54840 }, { "grad_norm": 0.368772953748703, "learning_rate": 4.6119904349148935e-05, "loss": 0.0203, "step": 54850 }, { "grad_norm": 0.3455705940723419, "learning_rate": 4.6103419725060726e-05, "loss": 0.0278, "step": 54860 }, { "grad_norm": 0.32496368885040283, "learning_rate": 4.608693552709667e-05, "loss": 0.0261, "step": 54870 }, { "grad_norm": 0.4821285009384155, "learning_rate": 4.607045175705947e-05, "loss": 0.0363, "step": 54880 }, { "grad_norm": 0.5236911773681641, "learning_rate": 4.605396841675176e-05, "loss": 0.0354, "step": 54890 }, { "grad_norm": 0.3509922921657562, "learning_rate": 4.6037485507976094e-05, "loss": 0.0298, "step": 54900 }, { "grad_norm": 0.3059075176715851, "learning_rate": 4.602100303253509e-05, "loss": 0.0172, "step": 54910 }, { "grad_norm": 0.17772497236728668, "learning_rate": 4.600452099223122e-05, "loss": 0.0358, "step": 54920 }, { "grad_norm": 0.29905420541763306, "learning_rate": 4.598803938886691e-05, "loss": 0.018, "step": 54930 }, { "grad_norm": 0.2972230017185211, "learning_rate": 4.597155822424459e-05, "loss": 0.0207, "step": 54940 }, { "grad_norm": 0.3105389475822449, "learning_rate": 4.59550775001666e-05, "loss": 0.0292, "step": 54950 }, { "grad_norm": 0.23620566725730896, "learning_rate": 4.5938597218435234e-05, "loss": 0.0273, "step": 54960 }, { "grad_norm": 0.3455877900123596, "learning_rate": 4.592211738085279e-05, "loss": 0.0345, "step": 54970 }, { "grad_norm": 0.22170642018318176, "learning_rate": 4.590563798922146e-05, "loss": 0.0284, "step": 54980 }, { "grad_norm": 0.3607228398323059, "learning_rate": 4.5889159045343404e-05, "loss": 0.0265, "step": 54990 }, { "grad_norm": 0.31893134117126465, "learning_rate": 4.5872680551020734e-05, "loss": 0.0296, "step": 55000 }, { "grad_norm": 0.42681676149368286, "learning_rate": 4.585620250805552e-05, "loss": 0.0253, "step": 55010 }, { "grad_norm": 0.24539832770824432, "learning_rate": 4.583972491824975e-05, "loss": 0.0208, "step": 55020 }, { "grad_norm": 0.3384834825992584, "learning_rate": 4.582324778340541e-05, "loss": 0.025, "step": 55030 }, { "grad_norm": 0.29242196679115295, "learning_rate": 4.5806771105324424e-05, "loss": 0.0406, "step": 55040 }, { "grad_norm": 0.37654346227645874, "learning_rate": 4.579029488580865e-05, "loss": 0.0307, "step": 55050 }, { "grad_norm": 0.28195008635520935, "learning_rate": 4.5773819126659876e-05, "loss": 0.0314, "step": 55060 }, { "grad_norm": 0.3108115792274475, "learning_rate": 4.575734382967991e-05, "loss": 0.0235, "step": 55070 }, { "grad_norm": 0.3957006335258484, "learning_rate": 4.574086899667044e-05, "loss": 0.033, "step": 55080 }, { "grad_norm": 0.4271083474159241, "learning_rate": 4.5724394629433106e-05, "loss": 0.0236, "step": 55090 }, { "grad_norm": 0.3140410780906677, "learning_rate": 4.570792072976958e-05, "loss": 0.0326, "step": 55100 }, { "grad_norm": 0.18151327967643738, "learning_rate": 4.569144729948138e-05, "loss": 0.0201, "step": 55110 }, { "grad_norm": 0.44240692257881165, "learning_rate": 4.5674974340370033e-05, "loss": 0.0256, "step": 55120 }, { "grad_norm": 0.4364030957221985, "learning_rate": 4.565850185423699e-05, "loss": 0.0348, "step": 55130 }, { "grad_norm": 0.39406219124794006, "learning_rate": 4.564202984288365e-05, "loss": 0.0371, "step": 55140 }, { "grad_norm": 0.35844719409942627, "learning_rate": 4.5625558308111356e-05, "loss": 0.0323, "step": 55150 }, { "grad_norm": 0.42905619740486145, "learning_rate": 4.560908725172145e-05, "loss": 0.0306, "step": 55160 }, { "grad_norm": 0.43358418345451355, "learning_rate": 4.559261667551516e-05, "loss": 0.0303, "step": 55170 }, { "grad_norm": 0.3373488485813141, "learning_rate": 4.5576146581293685e-05, "loss": 0.0337, "step": 55180 }, { "grad_norm": 0.4309963285923004, "learning_rate": 4.5559676970858164e-05, "loss": 0.0344, "step": 55190 }, { "grad_norm": 0.486264705657959, "learning_rate": 4.5543207846009705e-05, "loss": 0.0312, "step": 55200 }, { "grad_norm": 0.3204720616340637, "learning_rate": 4.552673920854933e-05, "loss": 0.0314, "step": 55210 }, { "grad_norm": 0.271463006734848, "learning_rate": 4.5510271060278006e-05, "loss": 0.0324, "step": 55220 }, { "grad_norm": 0.3016945421695709, "learning_rate": 4.5493803402996704e-05, "loss": 0.0392, "step": 55230 }, { "grad_norm": 0.40072259306907654, "learning_rate": 4.5477336238506295e-05, "loss": 0.0226, "step": 55240 }, { "grad_norm": 0.2660433053970337, "learning_rate": 4.546086956860759e-05, "loss": 0.0236, "step": 55250 }, { "grad_norm": 0.4021604359149933, "learning_rate": 4.5444403395101374e-05, "loss": 0.0193, "step": 55260 }, { "grad_norm": 0.35159096121788025, "learning_rate": 4.542793771978834e-05, "loss": 0.0308, "step": 55270 }, { "grad_norm": 0.652253270149231, "learning_rate": 4.541147254446916e-05, "loss": 0.035, "step": 55280 }, { "grad_norm": 0.32222050428390503, "learning_rate": 4.539500787094445e-05, "loss": 0.0408, "step": 55290 }, { "grad_norm": 0.408311665058136, "learning_rate": 4.537854370101475e-05, "loss": 0.0329, "step": 55300 }, { "grad_norm": 0.7433903813362122, "learning_rate": 4.5362080036480566e-05, "loss": 0.0324, "step": 55310 }, { "grad_norm": 0.291818231344223, "learning_rate": 4.534561687914235e-05, "loss": 0.0284, "step": 55320 }, { "grad_norm": 0.7100613117218018, "learning_rate": 4.5329154230800466e-05, "loss": 0.0328, "step": 55330 }, { "grad_norm": 0.36080655455589294, "learning_rate": 4.531269209325525e-05, "loss": 0.041, "step": 55340 }, { "grad_norm": 0.3695662319660187, "learning_rate": 4.529623046830696e-05, "loss": 0.0244, "step": 55350 }, { "grad_norm": 0.31511402130126953, "learning_rate": 4.5279769357755866e-05, "loss": 0.0276, "step": 55360 }, { "grad_norm": 0.659506618976593, "learning_rate": 4.5263308763402084e-05, "loss": 0.0364, "step": 55370 }, { "grad_norm": 0.40202614665031433, "learning_rate": 4.524684868704574e-05, "loss": 0.0302, "step": 55380 }, { "grad_norm": 0.33482682704925537, "learning_rate": 4.523038913048688e-05, "loss": 0.0284, "step": 55390 }, { "grad_norm": 0.3753792643547058, "learning_rate": 4.521393009552549e-05, "loss": 0.0301, "step": 55400 }, { "grad_norm": 0.3660541772842407, "learning_rate": 4.5197471583961484e-05, "loss": 0.0365, "step": 55410 }, { "grad_norm": 0.4033575654029846, "learning_rate": 4.51810135975948e-05, "loss": 0.0316, "step": 55420 }, { "grad_norm": 0.205950528383255, "learning_rate": 4.516455613822521e-05, "loss": 0.0254, "step": 55430 }, { "grad_norm": 0.22411119937896729, "learning_rate": 4.514809920765249e-05, "loss": 0.024, "step": 55440 }, { "grad_norm": 0.3215433657169342, "learning_rate": 4.5131642807676356e-05, "loss": 0.0229, "step": 55450 }, { "grad_norm": 0.29029595851898193, "learning_rate": 4.511518694009644e-05, "loss": 0.0241, "step": 55460 }, { "grad_norm": 0.2040344774723053, "learning_rate": 4.509873160671231e-05, "loss": 0.0208, "step": 55470 }, { "grad_norm": 0.4600280821323395, "learning_rate": 4.508227680932356e-05, "loss": 0.0195, "step": 55480 }, { "grad_norm": 0.2875891327857971, "learning_rate": 4.5065822549729605e-05, "loss": 0.0273, "step": 55490 }, { "grad_norm": 0.5517104864120483, "learning_rate": 4.504936882972989e-05, "loss": 0.027, "step": 55500 }, { "grad_norm": 0.2850337326526642, "learning_rate": 4.503291565112374e-05, "loss": 0.0225, "step": 55510 }, { "grad_norm": 0.4359146058559418, "learning_rate": 4.501646301571048e-05, "loss": 0.0271, "step": 55520 }, { "grad_norm": 0.25385481119155884, "learning_rate": 4.500001092528933e-05, "loss": 0.0268, "step": 55530 }, { "grad_norm": 0.27270814776420593, "learning_rate": 4.4983559381659455e-05, "loss": 0.0233, "step": 55540 }, { "grad_norm": 0.321444571018219, "learning_rate": 4.496710838661999e-05, "loss": 0.0288, "step": 55550 }, { "grad_norm": 0.6731830835342407, "learning_rate": 4.495065794196999e-05, "loss": 0.0342, "step": 55560 }, { "grad_norm": 0.3915274739265442, "learning_rate": 4.4934208049508455e-05, "loss": 0.0222, "step": 55570 }, { "grad_norm": 0.5561997890472412, "learning_rate": 4.4917758711034303e-05, "loss": 0.0326, "step": 55580 }, { "grad_norm": 0.22641098499298096, "learning_rate": 4.490130992834642e-05, "loss": 0.0286, "step": 55590 }, { "grad_norm": 0.43921971321105957, "learning_rate": 4.488486170324359e-05, "loss": 0.0382, "step": 55600 }, { "grad_norm": 0.24865572154521942, "learning_rate": 4.486841403752463e-05, "loss": 0.0252, "step": 55610 }, { "grad_norm": 0.31190595030784607, "learning_rate": 4.4851966932988174e-05, "loss": 0.029, "step": 55620 }, { "grad_norm": 0.4484034478664398, "learning_rate": 4.4835520391432894e-05, "loss": 0.0264, "step": 55630 }, { "grad_norm": 0.47080275416374207, "learning_rate": 4.481907441465733e-05, "loss": 0.0316, "step": 55640 }, { "grad_norm": 0.3482804000377655, "learning_rate": 4.480262900446e-05, "loss": 0.0238, "step": 55650 }, { "grad_norm": 0.3563755750656128, "learning_rate": 4.478618416263933e-05, "loss": 0.0263, "step": 55660 }, { "grad_norm": 0.2941230833530426, "learning_rate": 4.476973989099373e-05, "loss": 0.0228, "step": 55670 }, { "grad_norm": 0.7403483986854553, "learning_rate": 4.475329619132153e-05, "loss": 0.0334, "step": 55680 }, { "grad_norm": 0.43252235651016235, "learning_rate": 4.4736853065420956e-05, "loss": 0.0332, "step": 55690 }, { "grad_norm": 0.3519996106624603, "learning_rate": 4.472041051509023e-05, "loss": 0.0236, "step": 55700 }, { "grad_norm": 0.3352924883365631, "learning_rate": 4.470396854212746e-05, "loss": 0.0329, "step": 55710 }, { "grad_norm": 0.23660823702812195, "learning_rate": 4.468752714833072e-05, "loss": 0.026, "step": 55720 }, { "grad_norm": 0.21604672074317932, "learning_rate": 4.467108633549803e-05, "loss": 0.0229, "step": 55730 }, { "grad_norm": 0.24958287179470062, "learning_rate": 4.4654646105427335e-05, "loss": 0.0366, "step": 55740 }, { "grad_norm": 0.2796739339828491, "learning_rate": 4.463820645991651e-05, "loss": 0.0226, "step": 55750 }, { "grad_norm": 0.3717007637023926, "learning_rate": 4.462176740076334e-05, "loss": 0.0261, "step": 55760 }, { "grad_norm": 0.6308490633964539, "learning_rate": 4.4605328929765624e-05, "loss": 0.0313, "step": 55770 }, { "grad_norm": 0.27780798077583313, "learning_rate": 4.458889104872102e-05, "loss": 0.024, "step": 55780 }, { "grad_norm": 0.42717841267585754, "learning_rate": 4.457245375942715e-05, "loss": 0.0339, "step": 55790 }, { "grad_norm": 0.2381078153848648, "learning_rate": 4.455601706368157e-05, "loss": 0.0282, "step": 55800 }, { "grad_norm": 0.4644908905029297, "learning_rate": 4.453958096328179e-05, "loss": 0.0231, "step": 55810 }, { "grad_norm": 0.386539101600647, "learning_rate": 4.452314546002523e-05, "loss": 0.0264, "step": 55820 }, { "grad_norm": 0.38872963190078735, "learning_rate": 4.4506710555709246e-05, "loss": 0.0241, "step": 55830 }, { "grad_norm": 0.2696382999420166, "learning_rate": 4.449027625213114e-05, "loss": 0.0299, "step": 55840 }, { "grad_norm": 0.6942484974861145, "learning_rate": 4.4473842551088135e-05, "loss": 0.0304, "step": 55850 }, { "grad_norm": 0.24949494004249573, "learning_rate": 4.445740945437738e-05, "loss": 0.0262, "step": 55860 }, { "grad_norm": 0.41756105422973633, "learning_rate": 4.444097696379602e-05, "loss": 0.0416, "step": 55870 }, { "grad_norm": 0.34267035126686096, "learning_rate": 4.442454508114106e-05, "loss": 0.026, "step": 55880 }, { "grad_norm": 0.24032269418239594, "learning_rate": 4.4408113808209465e-05, "loss": 0.0294, "step": 55890 }, { "grad_norm": 0.3009931445121765, "learning_rate": 4.439168314679813e-05, "loss": 0.0304, "step": 55900 }, { "grad_norm": 0.33407774567604065, "learning_rate": 4.43752530987039e-05, "loss": 0.03, "step": 55910 }, { "grad_norm": 0.19144165515899658, "learning_rate": 4.4358823665723515e-05, "loss": 0.0191, "step": 55920 }, { "grad_norm": 0.3977154493331909, "learning_rate": 4.434239484965371e-05, "loss": 0.0233, "step": 55930 }, { "grad_norm": 0.18888695538043976, "learning_rate": 4.4325966652291103e-05, "loss": 0.0274, "step": 55940 }, { "grad_norm": 0.35874685645103455, "learning_rate": 4.430953907543225e-05, "loss": 0.0235, "step": 55950 }, { "grad_norm": 0.27937161922454834, "learning_rate": 4.4293112120873645e-05, "loss": 0.0213, "step": 55960 }, { "grad_norm": 0.42662471532821655, "learning_rate": 4.427668579041172e-05, "loss": 0.0202, "step": 55970 }, { "grad_norm": 0.34605905413627625, "learning_rate": 4.4260260085842816e-05, "loss": 0.0262, "step": 55980 }, { "grad_norm": 0.5672028660774231, "learning_rate": 4.4243835008963256e-05, "loss": 0.0316, "step": 55990 }, { "grad_norm": 0.32139265537261963, "learning_rate": 4.422741056156925e-05, "loss": 0.0254, "step": 56000 }, { "grad_norm": 0.3130607008934021, "learning_rate": 4.421098674545693e-05, "loss": 0.0288, "step": 56010 }, { "grad_norm": 0.29535990953445435, "learning_rate": 4.419456356242241e-05, "loss": 0.0349, "step": 56020 }, { "grad_norm": 0.20476093888282776, "learning_rate": 4.4178141014261685e-05, "loss": 0.0275, "step": 56030 }, { "grad_norm": 0.4162808358669281, "learning_rate": 4.416171910277071e-05, "loss": 0.0376, "step": 56040 }, { "grad_norm": 0.42994847893714905, "learning_rate": 4.414529782974533e-05, "loss": 0.0286, "step": 56050 }, { "grad_norm": 0.253328800201416, "learning_rate": 4.41288771969814e-05, "loss": 0.0364, "step": 56060 }, { "grad_norm": 0.6027291417121887, "learning_rate": 4.411245720627464e-05, "loss": 0.0485, "step": 56070 }, { "grad_norm": 0.31527113914489746, "learning_rate": 4.40960378594207e-05, "loss": 0.0278, "step": 56080 }, { "grad_norm": 0.21747228503227234, "learning_rate": 4.407961915821519e-05, "loss": 0.0375, "step": 56090 }, { "grad_norm": 0.4241228997707367, "learning_rate": 4.406320110445362e-05, "loss": 0.0278, "step": 56100 }, { "grad_norm": 0.21945205330848694, "learning_rate": 4.404678369993144e-05, "loss": 0.0232, "step": 56110 }, { "grad_norm": 0.17017710208892822, "learning_rate": 4.403036694644406e-05, "loss": 0.0284, "step": 56120 }, { "grad_norm": 0.25242841243743896, "learning_rate": 4.4013950845786764e-05, "loss": 0.0239, "step": 56130 }, { "grad_norm": 0.433206170797348, "learning_rate": 4.399753539975482e-05, "loss": 0.0272, "step": 56140 }, { "grad_norm": 0.3057687282562256, "learning_rate": 4.398112061014337e-05, "loss": 0.0278, "step": 56150 }, { "grad_norm": 0.33693164587020874, "learning_rate": 4.396470647874753e-05, "loss": 0.0285, "step": 56160 }, { "grad_norm": 0.29110345244407654, "learning_rate": 4.394829300736229e-05, "loss": 0.0195, "step": 56170 }, { "grad_norm": 0.32338613271713257, "learning_rate": 4.393188019778265e-05, "loss": 0.026, "step": 56180 }, { "grad_norm": 0.2122633457183838, "learning_rate": 4.391546805180347e-05, "loss": 0.0224, "step": 56190 }, { "grad_norm": 0.2673439085483551, "learning_rate": 4.389905657121955e-05, "loss": 0.0182, "step": 56200 }, { "grad_norm": 0.3054761290550232, "learning_rate": 4.3882645757825635e-05, "loss": 0.0323, "step": 56210 }, { "grad_norm": 0.3931015729904175, "learning_rate": 4.386623561341637e-05, "loss": 0.0254, "step": 56220 }, { "grad_norm": 0.30605053901672363, "learning_rate": 4.384982613978637e-05, "loss": 0.0251, "step": 56230 }, { "grad_norm": 0.40435662865638733, "learning_rate": 4.383341733873012e-05, "loss": 0.025, "step": 56240 }, { "grad_norm": 0.47601446509361267, "learning_rate": 4.381700921204209e-05, "loss": 0.0252, "step": 56250 }, { "grad_norm": 0.3284130096435547, "learning_rate": 4.380060176151663e-05, "loss": 0.03, "step": 56260 }, { "grad_norm": 0.3691859841346741, "learning_rate": 4.378419498894805e-05, "loss": 0.0287, "step": 56270 }, { "grad_norm": 0.7901293635368347, "learning_rate": 4.376778889613056e-05, "loss": 0.0291, "step": 56280 }, { "grad_norm": 0.5506255030632019, "learning_rate": 4.375138348485831e-05, "loss": 0.0345, "step": 56290 }, { "grad_norm": 0.3570161461830139, "learning_rate": 4.3734978756925346e-05, "loss": 0.0253, "step": 56300 }, { "grad_norm": 0.34320104122161865, "learning_rate": 4.3718574714125706e-05, "loss": 0.0369, "step": 56310 }, { "grad_norm": 0.34419411420822144, "learning_rate": 4.370217135825329e-05, "loss": 0.0258, "step": 56320 }, { "grad_norm": 0.2292080968618393, "learning_rate": 4.368576869110194e-05, "loss": 0.0209, "step": 56330 }, { "grad_norm": 0.30844080448150635, "learning_rate": 4.366936671446544e-05, "loss": 0.0215, "step": 56340 }, { "grad_norm": 0.40315982699394226, "learning_rate": 4.3652965430137474e-05, "loss": 0.0349, "step": 56350 }, { "grad_norm": 0.3425150513648987, "learning_rate": 4.3636564839911646e-05, "loss": 0.0271, "step": 56360 }, { "grad_norm": 0.28812175989151, "learning_rate": 4.3620164945581545e-05, "loss": 0.021, "step": 56370 }, { "grad_norm": 0.22143582999706268, "learning_rate": 4.360376574894061e-05, "loss": 0.0267, "step": 56380 }, { "grad_norm": 0.20303232967853546, "learning_rate": 4.358736725178224e-05, "loss": 0.0214, "step": 56390 }, { "grad_norm": 0.6644174456596375, "learning_rate": 4.357096945589974e-05, "loss": 0.0294, "step": 56400 }, { "grad_norm": 0.5118217468261719, "learning_rate": 4.355457236308636e-05, "loss": 0.026, "step": 56410 }, { "grad_norm": 0.24673353135585785, "learning_rate": 4.353817597513526e-05, "loss": 0.0232, "step": 56420 }, { "grad_norm": 0.20858830213546753, "learning_rate": 4.352178029383948e-05, "loss": 0.0211, "step": 56430 }, { "grad_norm": 0.32197967171669006, "learning_rate": 4.3505385320992105e-05, "loss": 0.0271, "step": 56440 }, { "grad_norm": 0.28101086616516113, "learning_rate": 4.348899105838602e-05, "loss": 0.0248, "step": 56450 }, { "grad_norm": 0.2451341599225998, "learning_rate": 4.3472597507814087e-05, "loss": 0.0219, "step": 56460 }, { "grad_norm": 0.28156739473342896, "learning_rate": 4.3456204671069066e-05, "loss": 0.0294, "step": 56470 }, { "grad_norm": 0.1752633899450302, "learning_rate": 4.343981254994367e-05, "loss": 0.0238, "step": 56480 }, { "grad_norm": 0.20002657175064087, "learning_rate": 4.34234211462305e-05, "loss": 0.0284, "step": 56490 }, { "grad_norm": 0.31806594133377075, "learning_rate": 4.340703046172213e-05, "loss": 0.0269, "step": 56500 }, { "grad_norm": 0.21828976273536682, "learning_rate": 4.339064049821097e-05, "loss": 0.0302, "step": 56510 }, { "grad_norm": 0.4150194823741913, "learning_rate": 4.3374251257489446e-05, "loss": 0.0276, "step": 56520 }, { "grad_norm": 0.32625439763069153, "learning_rate": 4.335786274134984e-05, "loss": 0.0214, "step": 56530 }, { "grad_norm": 0.5379064679145813, "learning_rate": 4.334147495158438e-05, "loss": 0.0385, "step": 56540 }, { "grad_norm": 0.4074461758136749, "learning_rate": 4.33250878899852e-05, "loss": 0.0264, "step": 56550 }, { "grad_norm": 0.3951420783996582, "learning_rate": 4.330870155834437e-05, "loss": 0.0309, "step": 56560 }, { "grad_norm": 0.3526398539543152, "learning_rate": 4.3292315958453884e-05, "loss": 0.02, "step": 56570 }, { "grad_norm": 0.5703907608985901, "learning_rate": 4.3275931092105646e-05, "loss": 0.0234, "step": 56580 }, { "grad_norm": 0.4320262670516968, "learning_rate": 4.325954696109147e-05, "loss": 0.024, "step": 56590 }, { "grad_norm": 0.3028514087200165, "learning_rate": 4.32431635672031e-05, "loss": 0.0308, "step": 56600 }, { "grad_norm": 0.40479594469070435, "learning_rate": 4.3226780912232214e-05, "loss": 0.034, "step": 56610 }, { "grad_norm": 0.31764015555381775, "learning_rate": 4.321039899797036e-05, "loss": 0.0247, "step": 56620 }, { "grad_norm": 0.2283325344324112, "learning_rate": 4.319401782620908e-05, "loss": 0.0249, "step": 56630 }, { "grad_norm": 0.25177547335624695, "learning_rate": 4.317763739873978e-05, "loss": 0.0214, "step": 56640 }, { "grad_norm": 0.33795827627182007, "learning_rate": 4.31612577173538e-05, "loss": 0.0305, "step": 56650 }, { "grad_norm": 0.2842675447463989, "learning_rate": 4.314487878384239e-05, "loss": 0.0374, "step": 56660 }, { "grad_norm": 0.44195789098739624, "learning_rate": 4.3128500599996736e-05, "loss": 0.0215, "step": 56670 }, { "grad_norm": 0.3177262544631958, "learning_rate": 4.31121231676079e-05, "loss": 0.0186, "step": 56680 }, { "grad_norm": 0.3186090588569641, "learning_rate": 4.309574648846694e-05, "loss": 0.0277, "step": 56690 }, { "grad_norm": 0.3455342650413513, "learning_rate": 4.3079370564364755e-05, "loss": 0.0288, "step": 56700 }, { "grad_norm": 0.3151859641075134, "learning_rate": 4.306299539709221e-05, "loss": 0.0248, "step": 56710 }, { "grad_norm": 0.29458853602409363, "learning_rate": 4.3046620988440046e-05, "loss": 0.0251, "step": 56720 }, { "grad_norm": 0.30187809467315674, "learning_rate": 4.303024734019897e-05, "loss": 0.0202, "step": 56730 }, { "grad_norm": 0.5808538794517517, "learning_rate": 4.3013874454159565e-05, "loss": 0.0241, "step": 56740 }, { "grad_norm": 0.2535826861858368, "learning_rate": 4.299750233211233e-05, "loss": 0.0333, "step": 56750 }, { "grad_norm": 0.33272460103034973, "learning_rate": 4.2981130975847715e-05, "loss": 0.0213, "step": 56760 }, { "grad_norm": 0.31041714549064636, "learning_rate": 4.2964760387156076e-05, "loss": 0.0239, "step": 56770 }, { "grad_norm": 0.3229456841945648, "learning_rate": 4.294839056782767e-05, "loss": 0.0216, "step": 56780 }, { "grad_norm": 0.49387073516845703, "learning_rate": 4.293202151965266e-05, "loss": 0.0239, "step": 56790 }, { "grad_norm": 0.35890889167785645, "learning_rate": 4.291565324442115e-05, "loss": 0.0241, "step": 56800 }, { "grad_norm": 0.4171217679977417, "learning_rate": 4.2899285743923135e-05, "loss": 0.0327, "step": 56810 }, { "grad_norm": 0.3480742275714874, "learning_rate": 4.288291901994857e-05, "loss": 0.0291, "step": 56820 }, { "grad_norm": 0.22872145473957062, "learning_rate": 4.286655307428729e-05, "loss": 0.0241, "step": 56830 }, { "grad_norm": 0.28356418013572693, "learning_rate": 4.285018790872903e-05, "loss": 0.0281, "step": 56840 }, { "grad_norm": 0.3988681137561798, "learning_rate": 4.283382352506347e-05, "loss": 0.0275, "step": 56850 }, { "grad_norm": 0.3573107421398163, "learning_rate": 4.2817459925080205e-05, "loss": 0.0313, "step": 56860 }, { "grad_norm": 0.2668134570121765, "learning_rate": 4.28010971105687e-05, "loss": 0.0293, "step": 56870 }, { "grad_norm": 0.5500979423522949, "learning_rate": 4.278473508331841e-05, "loss": 0.0272, "step": 56880 }, { "grad_norm": 0.516118049621582, "learning_rate": 4.276837384511864e-05, "loss": 0.0385, "step": 56890 }, { "grad_norm": 0.31070688366889954, "learning_rate": 4.275201339775864e-05, "loss": 0.0245, "step": 56900 }, { "grad_norm": 0.3887755572795868, "learning_rate": 4.273565374302756e-05, "loss": 0.0346, "step": 56910 }, { "grad_norm": 0.442964643239975, "learning_rate": 4.2719294882714454e-05, "loss": 0.024, "step": 56920 }, { "grad_norm": 0.37017810344696045, "learning_rate": 4.270293681860831e-05, "loss": 0.0265, "step": 56930 }, { "grad_norm": 0.3487839698791504, "learning_rate": 4.268657955249802e-05, "loss": 0.0209, "step": 56940 }, { "grad_norm": 0.3357986807823181, "learning_rate": 4.267022308617241e-05, "loss": 0.0236, "step": 56950 }, { "grad_norm": 0.4178588092327118, "learning_rate": 4.265386742142018e-05, "loss": 0.0249, "step": 56960 }, { "grad_norm": 0.31230518221855164, "learning_rate": 4.263751256002996e-05, "loss": 0.0279, "step": 56970 }, { "grad_norm": 0.2651432752609253, "learning_rate": 4.26211585037903e-05, "loss": 0.0209, "step": 56980 }, { "grad_norm": 0.2494339942932129, "learning_rate": 4.2604805254489656e-05, "loss": 0.0261, "step": 56990 }, { "grad_norm": 0.28767991065979004, "learning_rate": 4.258845281391638e-05, "loss": 0.0439, "step": 57000 }, { "grad_norm": 0.3653326630592346, "learning_rate": 4.257210118385876e-05, "loss": 0.0379, "step": 57010 }, { "grad_norm": 0.35478025674819946, "learning_rate": 4.2555750366104994e-05, "loss": 0.0263, "step": 57020 }, { "grad_norm": 0.4264744520187378, "learning_rate": 4.2539400362443184e-05, "loss": 0.0279, "step": 57030 }, { "grad_norm": 0.36078834533691406, "learning_rate": 4.252305117466133e-05, "loss": 0.0279, "step": 57040 }, { "grad_norm": 0.3186427056789398, "learning_rate": 4.2506702804547356e-05, "loss": 0.0332, "step": 57050 }, { "grad_norm": 0.2699580192565918, "learning_rate": 4.249035525388907e-05, "loss": 0.0225, "step": 57060 }, { "grad_norm": 0.30766406655311584, "learning_rate": 4.247400852447427e-05, "loss": 0.019, "step": 57070 }, { "grad_norm": 0.313906729221344, "learning_rate": 4.245766261809059e-05, "loss": 0.0245, "step": 57080 }, { "grad_norm": 0.30275171995162964, "learning_rate": 4.244131753652558e-05, "loss": 0.0298, "step": 57090 }, { "grad_norm": 0.47033005952835083, "learning_rate": 4.242497328156672e-05, "loss": 0.0195, "step": 57100 }, { "grad_norm": 0.40499261021614075, "learning_rate": 4.240862985500139e-05, "loss": 0.0304, "step": 57110 }, { "grad_norm": 0.25997790694236755, "learning_rate": 4.2392287258616894e-05, "loss": 0.0247, "step": 57120 }, { "grad_norm": 0.26879948377609253, "learning_rate": 4.2375945494200396e-05, "loss": 0.0245, "step": 57130 }, { "grad_norm": 0.29955682158470154, "learning_rate": 4.2359604563539054e-05, "loss": 0.0252, "step": 57140 }, { "grad_norm": 0.24897591769695282, "learning_rate": 4.2343264468419885e-05, "loss": 0.0226, "step": 57150 }, { "grad_norm": 0.33538520336151123, "learning_rate": 4.232692521062979e-05, "loss": 0.0202, "step": 57160 }, { "grad_norm": 0.19769196212291718, "learning_rate": 4.231058679195562e-05, "loss": 0.0251, "step": 57170 }, { "grad_norm": 0.326953649520874, "learning_rate": 4.229424921418411e-05, "loss": 0.0236, "step": 57180 }, { "grad_norm": 0.2612185478210449, "learning_rate": 4.2277912479101916e-05, "loss": 0.027, "step": 57190 }, { "grad_norm": 0.37738192081451416, "learning_rate": 4.226157658849562e-05, "loss": 0.0303, "step": 57200 }, { "grad_norm": 0.3738168478012085, "learning_rate": 4.2245241544151664e-05, "loss": 0.0279, "step": 57210 }, { "grad_norm": 0.436472088098526, "learning_rate": 4.2228907347856425e-05, "loss": 0.0266, "step": 57220 }, { "grad_norm": 0.2398497611284256, "learning_rate": 4.2212574001396214e-05, "loss": 0.0252, "step": 57230 }, { "grad_norm": 0.40751850605010986, "learning_rate": 4.21962415065572e-05, "loss": 0.0349, "step": 57240 }, { "grad_norm": 0.3263111114501953, "learning_rate": 4.217990986512548e-05, "loss": 0.0273, "step": 57250 }, { "grad_norm": 0.3101446330547333, "learning_rate": 4.216357907888704e-05, "loss": 0.0357, "step": 57260 }, { "grad_norm": 0.3966955542564392, "learning_rate": 4.2147249149627824e-05, "loss": 0.03, "step": 57270 }, { "grad_norm": 0.31491291522979736, "learning_rate": 4.213092007913364e-05, "loss": 0.0311, "step": 57280 }, { "grad_norm": 0.4144940674304962, "learning_rate": 4.2114591869190205e-05, "loss": 0.0281, "step": 57290 }, { "grad_norm": 0.2562287151813507, "learning_rate": 4.209826452158315e-05, "loss": 0.0283, "step": 57300 }, { "grad_norm": 0.499380499124527, "learning_rate": 4.2081938038098e-05, "loss": 0.0248, "step": 57310 }, { "grad_norm": 0.2818598747253418, "learning_rate": 4.2065612420520185e-05, "loss": 0.0202, "step": 57320 }, { "grad_norm": 0.2388124167919159, "learning_rate": 4.204928767063509e-05, "loss": 0.0219, "step": 57330 }, { "grad_norm": 0.37422052025794983, "learning_rate": 4.2032963790227944e-05, "loss": 0.028, "step": 57340 }, { "grad_norm": 0.22201132774353027, "learning_rate": 4.2016640781083887e-05, "loss": 0.0216, "step": 57350 }, { "grad_norm": 0.4371873140335083, "learning_rate": 4.2000318644988004e-05, "loss": 0.0453, "step": 57360 }, { "grad_norm": 0.24007360637187958, "learning_rate": 4.198399738372524e-05, "loss": 0.0273, "step": 57370 }, { "grad_norm": 0.311011403799057, "learning_rate": 4.196767699908045e-05, "loss": 0.0233, "step": 57380 }, { "grad_norm": 0.3679496943950653, "learning_rate": 4.195135749283845e-05, "loss": 0.0322, "step": 57390 }, { "grad_norm": 0.40693506598472595, "learning_rate": 4.1935038866783894e-05, "loss": 0.0262, "step": 57400 }, { "grad_norm": 0.5427535176277161, "learning_rate": 4.1918721122701355e-05, "loss": 0.0279, "step": 57410 }, { "grad_norm": 0.5182141661643982, "learning_rate": 4.190240426237533e-05, "loss": 0.0275, "step": 57420 }, { "grad_norm": 0.7269392609596252, "learning_rate": 4.188608828759019e-05, "loss": 0.0253, "step": 57430 }, { "grad_norm": 0.26761749386787415, "learning_rate": 4.1869773200130246e-05, "loss": 0.0319, "step": 57440 }, { "grad_norm": 0.9864575862884521, "learning_rate": 4.185345900177967e-05, "loss": 0.034, "step": 57450 }, { "grad_norm": 0.3336949944496155, "learning_rate": 4.183714569432258e-05, "loss": 0.0278, "step": 57460 }, { "grad_norm": 0.19595448672771454, "learning_rate": 4.182083327954295e-05, "loss": 0.0237, "step": 57470 }, { "grad_norm": 0.23220233619213104, "learning_rate": 4.1804521759224714e-05, "loss": 0.0319, "step": 57480 }, { "grad_norm": 0.32403719425201416, "learning_rate": 4.178821113515165e-05, "loss": 0.0209, "step": 57490 }, { "grad_norm": 0.3177710771560669, "learning_rate": 4.1771901409107476e-05, "loss": 0.0238, "step": 57500 }, { "grad_norm": 0.3572424054145813, "learning_rate": 4.175559258287577e-05, "loss": 0.0271, "step": 57510 }, { "grad_norm": 0.2855249047279358, "learning_rate": 4.173928465824008e-05, "loss": 0.0259, "step": 57520 }, { "grad_norm": 0.355316698551178, "learning_rate": 4.17229776369838e-05, "loss": 0.0276, "step": 57530 }, { "grad_norm": 0.2552783489227295, "learning_rate": 4.170667152089025e-05, "loss": 0.0294, "step": 57540 }, { "grad_norm": 0.16990993916988373, "learning_rate": 4.1690366311742626e-05, "loss": 0.018, "step": 57550 }, { "grad_norm": 0.42807963490486145, "learning_rate": 4.1674062011324055e-05, "loss": 0.0257, "step": 57560 }, { "grad_norm": 0.2696121633052826, "learning_rate": 4.165775862141752e-05, "loss": 0.0234, "step": 57570 }, { "grad_norm": 0.2869868576526642, "learning_rate": 4.164145614380598e-05, "loss": 0.0245, "step": 57580 }, { "grad_norm": 0.2912091314792633, "learning_rate": 4.162515458027223e-05, "loss": 0.0304, "step": 57590 }, { "grad_norm": 0.3836189806461334, "learning_rate": 4.160885393259899e-05, "loss": 0.0316, "step": 57600 }, { "grad_norm": 0.4748143255710602, "learning_rate": 4.159255420256886e-05, "loss": 0.0267, "step": 57610 }, { "grad_norm": 0.40141499042510986, "learning_rate": 4.157625539196436e-05, "loss": 0.027, "step": 57620 }, { "grad_norm": 0.2533206641674042, "learning_rate": 4.155995750256791e-05, "loss": 0.0448, "step": 57630 }, { "grad_norm": 0.3425056040287018, "learning_rate": 4.15436605361618e-05, "loss": 0.03, "step": 57640 }, { "grad_norm": 0.34874752163887024, "learning_rate": 4.152736449452827e-05, "loss": 0.0327, "step": 57650 }, { "grad_norm": 0.47456350922584534, "learning_rate": 4.151106937944943e-05, "loss": 0.0282, "step": 57660 }, { "grad_norm": 0.25782859325408936, "learning_rate": 4.149477519270727e-05, "loss": 0.0277, "step": 57670 }, { "grad_norm": 0.3509162962436676, "learning_rate": 4.147848193608371e-05, "loss": 0.0183, "step": 57680 }, { "grad_norm": 0.6710944175720215, "learning_rate": 4.146218961136056e-05, "loss": 0.023, "step": 57690 }, { "grad_norm": 0.1739538013935089, "learning_rate": 4.14458982203195e-05, "loss": 0.022, "step": 57700 }, { "grad_norm": 0.33625319600105286, "learning_rate": 4.1429607764742165e-05, "loss": 0.0284, "step": 57710 }, { "grad_norm": 0.2011415809392929, "learning_rate": 4.141331824641003e-05, "loss": 0.0293, "step": 57720 }, { "grad_norm": 0.3864832818508148, "learning_rate": 4.139702966710452e-05, "loss": 0.0236, "step": 57730 }, { "grad_norm": 0.3327973783016205, "learning_rate": 4.13807420286069e-05, "loss": 0.02, "step": 57740 }, { "grad_norm": 0.5520060062408447, "learning_rate": 4.136445533269837e-05, "loss": 0.0267, "step": 57750 }, { "grad_norm": 0.33909133076667786, "learning_rate": 4.134816958116003e-05, "loss": 0.0232, "step": 57760 }, { "grad_norm": 0.4262330234050751, "learning_rate": 4.133188477577283e-05, "loss": 0.0215, "step": 57770 }, { "grad_norm": 0.4744054675102234, "learning_rate": 4.131560091831771e-05, "loss": 0.0274, "step": 57780 }, { "grad_norm": 0.4807890057563782, "learning_rate": 4.1299318010575405e-05, "loss": 0.0275, "step": 57790 }, { "grad_norm": 0.3433205485343933, "learning_rate": 4.128303605432661e-05, "loss": 0.0372, "step": 57800 }, { "grad_norm": 0.4076528549194336, "learning_rate": 4.126675505135188e-05, "loss": 0.029, "step": 57810 }, { "grad_norm": 0.598518967628479, "learning_rate": 4.125047500343169e-05, "loss": 0.0341, "step": 57820 }, { "grad_norm": 0.40097224712371826, "learning_rate": 4.1234195912346376e-05, "loss": 0.032, "step": 57830 }, { "grad_norm": 0.33018866181373596, "learning_rate": 4.1217917779876235e-05, "loss": 0.0222, "step": 57840 }, { "grad_norm": 0.30442947149276733, "learning_rate": 4.120164060780141e-05, "loss": 0.0312, "step": 57850 }, { "grad_norm": 0.34925565123558044, "learning_rate": 4.1185364397901935e-05, "loss": 0.0318, "step": 57860 }, { "grad_norm": 0.5565095543861389, "learning_rate": 4.116908915195776e-05, "loss": 0.0225, "step": 57870 }, { "grad_norm": 0.36239227652549744, "learning_rate": 4.115281487174872e-05, "loss": 0.0329, "step": 57880 }, { "grad_norm": 0.39606407284736633, "learning_rate": 4.113654155905452e-05, "loss": 0.0297, "step": 57890 }, { "grad_norm": 0.28931576013565063, "learning_rate": 4.112026921565483e-05, "loss": 0.0296, "step": 57900 }, { "grad_norm": 0.28570470213890076, "learning_rate": 4.1103997843329147e-05, "loss": 0.0244, "step": 57910 }, { "grad_norm": 0.3128548860549927, "learning_rate": 4.108772744385689e-05, "loss": 0.0377, "step": 57920 }, { "grad_norm": 0.2893100678920746, "learning_rate": 4.1071458019017346e-05, "loss": 0.0279, "step": 57930 }, { "grad_norm": 0.27389535307884216, "learning_rate": 4.1055189570589734e-05, "loss": 0.0195, "step": 57940 }, { "grad_norm": 0.43876999616622925, "learning_rate": 4.103892210035315e-05, "loss": 0.0309, "step": 57950 }, { "grad_norm": 0.3659018576145172, "learning_rate": 4.102265561008656e-05, "loss": 0.0281, "step": 57960 }, { "grad_norm": 0.5375710725784302, "learning_rate": 4.100639010156885e-05, "loss": 0.0236, "step": 57970 }, { "grad_norm": 0.23661106824874878, "learning_rate": 4.099012557657882e-05, "loss": 0.0211, "step": 57980 }, { "grad_norm": 0.27771279215812683, "learning_rate": 4.09738620368951e-05, "loss": 0.0242, "step": 57990 }, { "grad_norm": 0.25620341300964355, "learning_rate": 4.095759948429626e-05, "loss": 0.0246, "step": 58000 }, { "grad_norm": 0.2590731084346771, "learning_rate": 4.094133792056075e-05, "loss": 0.023, "step": 58010 }, { "grad_norm": 0.19457358121871948, "learning_rate": 4.092507734746688e-05, "loss": 0.0272, "step": 58020 }, { "grad_norm": 0.3917819857597351, "learning_rate": 4.090881776679293e-05, "loss": 0.0311, "step": 58030 }, { "grad_norm": 0.2703668475151062, "learning_rate": 4.0892559180316994e-05, "loss": 0.0245, "step": 58040 }, { "grad_norm": 4.271255970001221, "learning_rate": 4.08763015898171e-05, "loss": 0.0286, "step": 58050 }, { "grad_norm": 0.37000927329063416, "learning_rate": 4.0860044997071145e-05, "loss": 0.0269, "step": 58060 }, { "grad_norm": 0.32007744908332825, "learning_rate": 4.084378940385693e-05, "loss": 0.0292, "step": 58070 }, { "grad_norm": 0.4968891143798828, "learning_rate": 4.082753481195211e-05, "loss": 0.0207, "step": 58080 }, { "grad_norm": 0.31524908542633057, "learning_rate": 4.0811281223134324e-05, "loss": 0.0222, "step": 58090 }, { "grad_norm": 0.33349496126174927, "learning_rate": 4.0795028639181e-05, "loss": 0.0363, "step": 58100 }, { "grad_norm": 0.4067779779434204, "learning_rate": 4.077877706186951e-05, "loss": 0.0293, "step": 58110 }, { "grad_norm": 0.41383907198905945, "learning_rate": 4.076252649297709e-05, "loss": 0.0279, "step": 58120 }, { "grad_norm": 0.29896920919418335, "learning_rate": 4.07462769342809e-05, "loss": 0.0294, "step": 58130 }, { "grad_norm": 0.36186593770980835, "learning_rate": 4.073002838755793e-05, "loss": 0.0291, "step": 58140 }, { "grad_norm": 0.34282970428466797, "learning_rate": 4.0713780854585115e-05, "loss": 0.023, "step": 58150 }, { "grad_norm": 0.25975537300109863, "learning_rate": 4.069753433713927e-05, "loss": 0.023, "step": 58160 }, { "grad_norm": 0.2561874985694885, "learning_rate": 4.068128883699709e-05, "loss": 0.0199, "step": 58170 }, { "grad_norm": 0.4969034790992737, "learning_rate": 4.0665044355935155e-05, "loss": 0.0331, "step": 58180 }, { "grad_norm": 0.3156176209449768, "learning_rate": 4.064880089572993e-05, "loss": 0.0215, "step": 58190 }, { "grad_norm": 0.17576462030410767, "learning_rate": 4.063255845815779e-05, "loss": 0.0309, "step": 58200 }, { "grad_norm": 0.3010769486427307, "learning_rate": 4.0616317044994945e-05, "loss": 0.0179, "step": 58210 }, { "grad_norm": 0.3292473256587982, "learning_rate": 4.0600076658017585e-05, "loss": 0.0251, "step": 58220 }, { "grad_norm": 0.2582392990589142, "learning_rate": 4.0583837299001706e-05, "loss": 0.0293, "step": 58230 }, { "grad_norm": 0.603661835193634, "learning_rate": 4.056759896972322e-05, "loss": 0.0226, "step": 58240 }, { "grad_norm": 0.35760781168937683, "learning_rate": 4.055136167195793e-05, "loss": 0.0308, "step": 58250 }, { "grad_norm": 0.30045124888420105, "learning_rate": 4.053512540748152e-05, "loss": 0.033, "step": 58260 }, { "grad_norm": 0.3124827742576599, "learning_rate": 4.051889017806955e-05, "loss": 0.0276, "step": 58270 }, { "grad_norm": 0.25796863436698914, "learning_rate": 4.0502655985497516e-05, "loss": 0.0199, "step": 58280 }, { "grad_norm": 0.41379284858703613, "learning_rate": 4.048642283154074e-05, "loss": 0.0245, "step": 58290 }, { "grad_norm": 0.28452712297439575, "learning_rate": 4.0470190717974464e-05, "loss": 0.0251, "step": 58300 }, { "grad_norm": 0.26861095428466797, "learning_rate": 4.045395964657381e-05, "loss": 0.0272, "step": 58310 }, { "grad_norm": 0.34252026677131653, "learning_rate": 4.043772961911376e-05, "loss": 0.0249, "step": 58320 }, { "grad_norm": 0.3457999527454376, "learning_rate": 4.042150063736924e-05, "loss": 0.0224, "step": 58330 }, { "grad_norm": 0.29102107882499695, "learning_rate": 4.0405272703114976e-05, "loss": 0.0284, "step": 58340 }, { "grad_norm": 0.30370602011680603, "learning_rate": 4.038904581812569e-05, "loss": 0.0249, "step": 58350 }, { "grad_norm": 0.32573068141937256, "learning_rate": 4.03728199841759e-05, "loss": 0.0501, "step": 58360 }, { "grad_norm": 0.286594420671463, "learning_rate": 4.0356595203040044e-05, "loss": 0.0315, "step": 58370 }, { "grad_norm": 0.4645967483520508, "learning_rate": 4.0340371476492426e-05, "loss": 0.0311, "step": 58380 }, { "grad_norm": 0.4559743106365204, "learning_rate": 4.032414880630727e-05, "loss": 0.0309, "step": 58390 }, { "grad_norm": 0.41867512464523315, "learning_rate": 4.0307927194258646e-05, "loss": 0.0265, "step": 58400 }, { "grad_norm": 0.8974992036819458, "learning_rate": 4.029170664212054e-05, "loss": 0.021, "step": 58410 }, { "grad_norm": 0.344905287027359, "learning_rate": 4.0275487151666794e-05, "loss": 0.0315, "step": 58420 }, { "grad_norm": 0.28486356139183044, "learning_rate": 4.025926872467116e-05, "loss": 0.0256, "step": 58430 }, { "grad_norm": 0.5482534170150757, "learning_rate": 4.024305136290725e-05, "loss": 0.0271, "step": 58440 }, { "grad_norm": 0.25474727153778076, "learning_rate": 4.022683506814858e-05, "loss": 0.0238, "step": 58450 }, { "grad_norm": 0.3390779495239258, "learning_rate": 4.021061984216853e-05, "loss": 0.0271, "step": 58460 }, { "grad_norm": 0.2951248586177826, "learning_rate": 4.019440568674035e-05, "loss": 0.0248, "step": 58470 }, { "grad_norm": 0.3405727744102478, "learning_rate": 4.017819260363724e-05, "loss": 0.0227, "step": 58480 }, { "grad_norm": 0.2540043890476227, "learning_rate": 4.016198059463223e-05, "loss": 0.0318, "step": 58490 }, { "grad_norm": 0.45482638478279114, "learning_rate": 4.0145769661498233e-05, "loss": 0.0324, "step": 58500 }, { "grad_norm": 0.2829696834087372, "learning_rate": 4.012955980600804e-05, "loss": 0.0232, "step": 58510 }, { "grad_norm": 0.5306842923164368, "learning_rate": 4.0113351029934355e-05, "loss": 0.0279, "step": 58520 }, { "grad_norm": 0.34142062067985535, "learning_rate": 4.0097143335049714e-05, "loss": 0.0251, "step": 58530 }, { "grad_norm": 0.4411470293998718, "learning_rate": 4.0080936723126606e-05, "loss": 0.0243, "step": 58540 }, { "grad_norm": 0.28822487592697144, "learning_rate": 4.006473119593735e-05, "loss": 0.0248, "step": 58550 }, { "grad_norm": 0.3369765281677246, "learning_rate": 4.004852675525415e-05, "loss": 0.0323, "step": 58560 }, { "grad_norm": 0.3206196129322052, "learning_rate": 4.00323234028491e-05, "loss": 0.0244, "step": 58570 }, { "grad_norm": 0.41291141510009766, "learning_rate": 4.0016121140494186e-05, "loss": 0.0296, "step": 58580 }, { "grad_norm": 0.31337928771972656, "learning_rate": 3.999991996996122e-05, "loss": 0.0254, "step": 58590 }, { "grad_norm": 0.2690179944038391, "learning_rate": 3.9983719893022e-05, "loss": 0.0201, "step": 58600 }, { "grad_norm": 0.3712173402309418, "learning_rate": 3.996752091144812e-05, "loss": 0.0237, "step": 58610 }, { "grad_norm": 0.478903591632843, "learning_rate": 3.9951323027011055e-05, "loss": 0.0292, "step": 58620 }, { "grad_norm": 0.2466793656349182, "learning_rate": 3.993512624148219e-05, "loss": 0.0309, "step": 58630 }, { "grad_norm": 0.416384756565094, "learning_rate": 3.99189305566328e-05, "loss": 0.0206, "step": 58640 }, { "grad_norm": 0.3192440867424011, "learning_rate": 3.990273597423401e-05, "loss": 0.0239, "step": 58650 }, { "grad_norm": 0.4751407504081726, "learning_rate": 3.988654249605681e-05, "loss": 0.0253, "step": 58660 }, { "grad_norm": 0.3790518045425415, "learning_rate": 3.987035012387213e-05, "loss": 0.023, "step": 58670 }, { "grad_norm": 0.31920501589775085, "learning_rate": 3.9854158859450735e-05, "loss": 0.0209, "step": 58680 }, { "grad_norm": 0.4022061824798584, "learning_rate": 3.983796870456328e-05, "loss": 0.0286, "step": 58690 }, { "grad_norm": 0.3185911774635315, "learning_rate": 3.982177966098028e-05, "loss": 0.0177, "step": 58700 }, { "grad_norm": 0.4002557396888733, "learning_rate": 3.980559173047216e-05, "loss": 0.0216, "step": 58710 }, { "grad_norm": 0.5107930302619934, "learning_rate": 3.9789404914809176e-05, "loss": 0.0166, "step": 58720 }, { "grad_norm": 0.17321091890335083, "learning_rate": 3.9773219215761544e-05, "loss": 0.026, "step": 58730 }, { "grad_norm": 0.1933107078075409, "learning_rate": 3.975703463509928e-05, "loss": 0.0167, "step": 58740 }, { "grad_norm": 0.37411805987358093, "learning_rate": 3.974085117459231e-05, "loss": 0.0224, "step": 58750 }, { "grad_norm": 0.37847572565078735, "learning_rate": 3.972466883601043e-05, "loss": 0.0273, "step": 58760 }, { "grad_norm": 0.26686766743659973, "learning_rate": 3.9708487621123316e-05, "loss": 0.0232, "step": 58770 }, { "grad_norm": 0.3422321379184723, "learning_rate": 3.969230753170051e-05, "loss": 0.0267, "step": 58780 }, { "grad_norm": 0.9154071807861328, "learning_rate": 3.967612856951146e-05, "loss": 0.0268, "step": 58790 }, { "grad_norm": 0.2639353275299072, "learning_rate": 3.9659950736325476e-05, "loss": 0.0246, "step": 58800 }, { "grad_norm": 0.3442087173461914, "learning_rate": 3.964377403391174e-05, "loss": 0.052, "step": 58810 }, { "grad_norm": 0.43547379970550537, "learning_rate": 3.9627598464039303e-05, "loss": 0.0379, "step": 58820 }, { "grad_norm": 0.37755370140075684, "learning_rate": 3.9611424028477096e-05, "loss": 0.0228, "step": 58830 }, { "grad_norm": 0.20900797843933105, "learning_rate": 3.959525072899394e-05, "loss": 0.0359, "step": 58840 }, { "grad_norm": 0.2673790454864502, "learning_rate": 3.9579078567358506e-05, "loss": 0.0207, "step": 58850 }, { "grad_norm": 0.2594112157821655, "learning_rate": 3.956290754533939e-05, "loss": 0.0269, "step": 58860 }, { "grad_norm": 0.2851230502128601, "learning_rate": 3.954673766470501e-05, "loss": 0.0298, "step": 58870 }, { "grad_norm": 0.4207901656627655, "learning_rate": 3.9530568927223674e-05, "loss": 0.0222, "step": 58880 }, { "grad_norm": 0.23761330544948578, "learning_rate": 3.95144013346636e-05, "loss": 0.02, "step": 58890 }, { "grad_norm": 0.24464039504528046, "learning_rate": 3.949823488879282e-05, "loss": 0.0207, "step": 58900 }, { "grad_norm": 0.31503769755363464, "learning_rate": 3.9482069591379276e-05, "loss": 0.0397, "step": 58910 }, { "grad_norm": 0.3429771661758423, "learning_rate": 3.9465905444190784e-05, "loss": 0.0263, "step": 58920 }, { "grad_norm": 0.2547242343425751, "learning_rate": 3.944974244899505e-05, "loss": 0.0268, "step": 58930 }, { "grad_norm": 0.47508856654167175, "learning_rate": 3.9433580607559616e-05, "loss": 0.0297, "step": 58940 }, { "grad_norm": 0.43942153453826904, "learning_rate": 3.941741992165193e-05, "loss": 0.035, "step": 58950 }, { "grad_norm": 0.47895461320877075, "learning_rate": 3.940126039303928e-05, "loss": 0.0266, "step": 58960 }, { "grad_norm": 0.27836930751800537, "learning_rate": 3.938510202348886e-05, "loss": 0.0276, "step": 58970 }, { "grad_norm": 0.17320676147937775, "learning_rate": 3.93689448147677e-05, "loss": 0.0202, "step": 58980 }, { "grad_norm": 0.42220941185951233, "learning_rate": 3.935278876864278e-05, "loss": 0.033, "step": 58990 }, { "grad_norm": 0.4102512001991272, "learning_rate": 3.933663388688087e-05, "loss": 0.0313, "step": 59000 }, { "grad_norm": 0.595229983329773, "learning_rate": 3.9320480171248645e-05, "loss": 0.0277, "step": 59010 }, { "grad_norm": 0.5553661584854126, "learning_rate": 3.930432762351265e-05, "loss": 0.0401, "step": 59020 }, { "grad_norm": 0.30221500992774963, "learning_rate": 3.928817624543931e-05, "loss": 0.0328, "step": 59030 }, { "grad_norm": 0.1600813865661621, "learning_rate": 3.9272026038794885e-05, "loss": 0.0225, "step": 59040 }, { "grad_norm": 0.5477560758590698, "learning_rate": 3.9255877005345586e-05, "loss": 0.0319, "step": 59050 }, { "grad_norm": 0.18177026510238647, "learning_rate": 3.923972914685743e-05, "loss": 0.0251, "step": 59060 }, { "grad_norm": 0.2334195077419281, "learning_rate": 3.922358246509631e-05, "loss": 0.0246, "step": 59070 }, { "grad_norm": 0.28794965147972107, "learning_rate": 3.920743696182801e-05, "loss": 0.0284, "step": 59080 }, { "grad_norm": 0.3269367516040802, "learning_rate": 3.9191292638818164e-05, "loss": 0.0217, "step": 59090 }, { "grad_norm": 0.2996080815792084, "learning_rate": 3.9175149497832316e-05, "loss": 0.0236, "step": 59100 }, { "grad_norm": 0.49428197741508484, "learning_rate": 3.915900754063584e-05, "loss": 0.0223, "step": 59110 }, { "grad_norm": 0.2960159480571747, "learning_rate": 3.9142866768994014e-05, "loss": 0.021, "step": 59120 }, { "grad_norm": 0.4034072458744049, "learning_rate": 3.912672718467193e-05, "loss": 0.0215, "step": 59130 }, { "grad_norm": 0.5619720816612244, "learning_rate": 3.911058878943463e-05, "loss": 0.0195, "step": 59140 }, { "grad_norm": 0.24903494119644165, "learning_rate": 3.9094451585046974e-05, "loss": 0.0224, "step": 59150 }, { "grad_norm": 0.8679423332214355, "learning_rate": 3.90783155732737e-05, "loss": 0.0294, "step": 59160 }, { "grad_norm": 0.36364805698394775, "learning_rate": 3.9062180755879373e-05, "loss": 0.0272, "step": 59170 }, { "grad_norm": 0.37452536821365356, "learning_rate": 3.904604713462855e-05, "loss": 0.0227, "step": 59180 }, { "grad_norm": 1.6640256643295288, "learning_rate": 3.902991471128554e-05, "loss": 0.0188, "step": 59190 }, { "grad_norm": 0.23287741839885712, "learning_rate": 3.901378348761456e-05, "loss": 0.0176, "step": 59200 }, { "grad_norm": 0.27869313955307007, "learning_rate": 3.899765346537971e-05, "loss": 0.0291, "step": 59210 }, { "grad_norm": 0.3308163583278656, "learning_rate": 3.898152464634493e-05, "loss": 0.0298, "step": 59220 }, { "grad_norm": 0.23890846967697144, "learning_rate": 3.896539703227402e-05, "loss": 0.028, "step": 59230 }, { "grad_norm": 0.36613136529922485, "learning_rate": 3.894927062493073e-05, "loss": 0.0215, "step": 59240 }, { "grad_norm": 0.31980493664741516, "learning_rate": 3.893314542607859e-05, "loss": 0.0215, "step": 59250 }, { "grad_norm": 0.23348458111286163, "learning_rate": 3.891702143748103e-05, "loss": 0.0206, "step": 59260 }, { "grad_norm": 0.25635677576065063, "learning_rate": 3.890089866090133e-05, "loss": 0.0231, "step": 59270 }, { "grad_norm": 0.3384808897972107, "learning_rate": 3.888477709810267e-05, "loss": 0.0297, "step": 59280 }, { "grad_norm": 0.30128732323646545, "learning_rate": 3.886865675084806e-05, "loss": 0.0214, "step": 59290 }, { "grad_norm": 0.1985160857439041, "learning_rate": 3.885253762090043e-05, "loss": 0.0236, "step": 59300 }, { "grad_norm": 0.2272648960351944, "learning_rate": 3.8836419710022534e-05, "loss": 0.0272, "step": 59310 }, { "grad_norm": 0.7080600261688232, "learning_rate": 3.882030301997698e-05, "loss": 0.0253, "step": 59320 }, { "grad_norm": 0.3705931305885315, "learning_rate": 3.880418755252629e-05, "loss": 0.0253, "step": 59330 }, { "grad_norm": 0.4493604898452759, "learning_rate": 3.87880733094328e-05, "loss": 0.0281, "step": 59340 }, { "grad_norm": 0.2520028054714203, "learning_rate": 3.877196029245877e-05, "loss": 0.0176, "step": 59350 }, { "grad_norm": 0.5553188323974609, "learning_rate": 3.875584850336627e-05, "loss": 0.0283, "step": 59360 }, { "grad_norm": 0.380087673664093, "learning_rate": 3.873973794391728e-05, "loss": 0.029, "step": 59370 }, { "grad_norm": 0.17789943516254425, "learning_rate": 3.872362861587361e-05, "loss": 0.0228, "step": 59380 }, { "grad_norm": 0.44695335626602173, "learning_rate": 3.870752052099698e-05, "loss": 0.0226, "step": 59390 }, { "grad_norm": 0.32939133048057556, "learning_rate": 3.8691413661048916e-05, "loss": 0.0216, "step": 59400 }, { "grad_norm": 0.34373900294303894, "learning_rate": 3.867530803779085e-05, "loss": 0.0282, "step": 59410 }, { "grad_norm": 0.3478907346725464, "learning_rate": 3.8659203652984054e-05, "loss": 0.0236, "step": 59420 }, { "grad_norm": 0.40529918670654297, "learning_rate": 3.864310050838972e-05, "loss": 0.0244, "step": 59430 }, { "grad_norm": 0.1434820145368576, "learning_rate": 3.8626998605768835e-05, "loss": 0.018, "step": 59440 }, { "grad_norm": 0.37598371505737305, "learning_rate": 3.8610897946882294e-05, "loss": 0.028, "step": 59450 }, { "grad_norm": 0.42417383193969727, "learning_rate": 3.859479853349082e-05, "loss": 0.0286, "step": 59460 }, { "grad_norm": 0.28685489296913147, "learning_rate": 3.857870036735503e-05, "loss": 0.0299, "step": 59470 }, { "grad_norm": 0.2963792383670807, "learning_rate": 3.856260345023538e-05, "loss": 0.0254, "step": 59480 }, { "grad_norm": 0.33171024918556213, "learning_rate": 3.854650778389224e-05, "loss": 0.0236, "step": 59490 }, { "grad_norm": 0.26511842012405396, "learning_rate": 3.8530413370085786e-05, "loss": 0.0309, "step": 59500 }, { "grad_norm": 0.26514917612075806, "learning_rate": 3.851432021057608e-05, "loss": 0.019, "step": 59510 }, { "grad_norm": 0.4658913314342499, "learning_rate": 3.849822830712306e-05, "loss": 0.0242, "step": 59520 }, { "grad_norm": 1.3090038299560547, "learning_rate": 3.848213766148649e-05, "loss": 0.0172, "step": 59530 }, { "grad_norm": 0.3377842307090759, "learning_rate": 3.846604827542602e-05, "loss": 0.0298, "step": 59540 }, { "grad_norm": 0.29750508069992065, "learning_rate": 3.844996015070117e-05, "loss": 0.0337, "step": 59550 }, { "grad_norm": 0.24478307366371155, "learning_rate": 3.843387328907132e-05, "loss": 0.0207, "step": 59560 }, { "grad_norm": 0.3251774311065674, "learning_rate": 3.8417787692295696e-05, "loss": 0.0323, "step": 59570 }, { "grad_norm": 0.18562060594558716, "learning_rate": 3.84017033621334e-05, "loss": 0.0215, "step": 59580 }, { "grad_norm": 0.3603655695915222, "learning_rate": 3.838562030034338e-05, "loss": 0.0238, "step": 59590 }, { "grad_norm": 0.24101832509040833, "learning_rate": 3.836953850868447e-05, "loss": 0.0269, "step": 59600 }, { "grad_norm": 0.21885350346565247, "learning_rate": 3.835345798891533e-05, "loss": 0.0177, "step": 59610 }, { "grad_norm": 0.30942457914352417, "learning_rate": 3.833737874279452e-05, "loss": 0.0242, "step": 59620 }, { "grad_norm": 0.36648690700531006, "learning_rate": 3.8321300772080433e-05, "loss": 0.0206, "step": 59630 }, { "grad_norm": 0.20762930810451508, "learning_rate": 3.8305224078531345e-05, "loss": 0.0204, "step": 59640 }, { "grad_norm": 0.4083695411682129, "learning_rate": 3.828914866390537e-05, "loss": 0.0252, "step": 59650 }, { "grad_norm": 0.347578227519989, "learning_rate": 3.827307452996048e-05, "loss": 0.0248, "step": 59660 }, { "grad_norm": 0.24485713243484497, "learning_rate": 3.825700167845454e-05, "loss": 0.0332, "step": 59670 }, { "grad_norm": 0.5174184441566467, "learning_rate": 3.8240930111145214e-05, "loss": 0.0311, "step": 59680 }, { "grad_norm": 0.5127410292625427, "learning_rate": 3.8224859829790116e-05, "loss": 0.0215, "step": 59690 }, { "grad_norm": 0.22170399129390717, "learning_rate": 3.820879083614664e-05, "loss": 0.0265, "step": 59700 }, { "grad_norm": 0.4423191249370575, "learning_rate": 3.819272313197208e-05, "loss": 0.0284, "step": 59710 }, { "grad_norm": 0.2567908465862274, "learning_rate": 3.817665671902355e-05, "loss": 0.0296, "step": 59720 }, { "grad_norm": 0.22612062096595764, "learning_rate": 3.816059159905808e-05, "loss": 0.0208, "step": 59730 }, { "grad_norm": 0.2572367191314697, "learning_rate": 3.8144527773832476e-05, "loss": 0.0301, "step": 59740 }, { "grad_norm": 0.2962067127227783, "learning_rate": 3.812846524510352e-05, "loss": 0.0199, "step": 59750 }, { "grad_norm": 0.42394351959228516, "learning_rate": 3.811240401462775e-05, "loss": 0.0278, "step": 59760 }, { "grad_norm": 0.4352007508277893, "learning_rate": 3.809634408416162e-05, "loss": 0.0227, "step": 59770 }, { "grad_norm": 0.32050302624702454, "learning_rate": 3.8080285455461394e-05, "loss": 0.0258, "step": 59780 }, { "grad_norm": 0.262387216091156, "learning_rate": 3.806422813028323e-05, "loss": 0.0247, "step": 59790 }, { "grad_norm": 0.3290961980819702, "learning_rate": 3.8048172110383114e-05, "loss": 0.0238, "step": 59800 }, { "grad_norm": 0.3069603443145752, "learning_rate": 3.8032117397516944e-05, "loss": 0.0219, "step": 59810 }, { "grad_norm": 0.2535039484500885, "learning_rate": 3.801606399344042e-05, "loss": 0.0198, "step": 59820 }, { "grad_norm": 0.3556448817253113, "learning_rate": 3.800001189990913e-05, "loss": 0.0265, "step": 59830 }, { "grad_norm": 0.2859640419483185, "learning_rate": 3.798396111867847e-05, "loss": 0.0233, "step": 59840 }, { "grad_norm": 0.20364101231098175, "learning_rate": 3.796791165150378e-05, "loss": 0.0185, "step": 59850 }, { "grad_norm": 0.3288993835449219, "learning_rate": 3.795186350014018e-05, "loss": 0.025, "step": 59860 }, { "grad_norm": 0.6271167993545532, "learning_rate": 3.793581666634266e-05, "loss": 0.0263, "step": 59870 }, { "grad_norm": 0.30976298451423645, "learning_rate": 3.791977115186609e-05, "loss": 0.0215, "step": 59880 }, { "grad_norm": 0.3344740569591522, "learning_rate": 3.79037269584652e-05, "loss": 0.0194, "step": 59890 }, { "grad_norm": 0.2501579821109772, "learning_rate": 3.7887684087894545e-05, "loss": 0.0259, "step": 59900 }, { "grad_norm": 0.38892897963523865, "learning_rate": 3.787164254190854e-05, "loss": 0.0212, "step": 59910 }, { "grad_norm": 0.24608084559440613, "learning_rate": 3.7855602322261473e-05, "loss": 0.0208, "step": 59920 }, { "grad_norm": 0.33329975605010986, "learning_rate": 3.783956343070746e-05, "loss": 0.035, "step": 59930 }, { "grad_norm": 0.25425639748573303, "learning_rate": 3.7823525869000526e-05, "loss": 0.0213, "step": 59940 }, { "grad_norm": 0.3918655514717102, "learning_rate": 3.78074896388945e-05, "loss": 0.0235, "step": 59950 }, { "grad_norm": 0.3233579099178314, "learning_rate": 3.7791454742143075e-05, "loss": 0.0261, "step": 59960 }, { "grad_norm": 0.213889017701149, "learning_rate": 3.7775421180499795e-05, "loss": 0.0163, "step": 59970 }, { "grad_norm": 0.43472760915756226, "learning_rate": 3.7759388955718086e-05, "loss": 0.0228, "step": 59980 }, { "grad_norm": 0.2733432948589325, "learning_rate": 3.774335806955117e-05, "loss": 0.0235, "step": 59990 }, { "grad_norm": 0.27368810772895813, "learning_rate": 3.7727328523752215e-05, "loss": 0.0215, "step": 60000 } ], "logging_steps": 10, "max_steps": 100000, "num_input_tokens_seen": 0, "num_train_epochs": 9223372036854775807, "save_steps": 20000, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": false }, "attributes": {} } }, "total_flos": 0.0, "train_batch_size": 8, "trial_name": null, "trial_params": null }