{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 0.3, "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": 1.8748948574066162, "learning_rate": 1.8e-07, "loss": 1.5655, "step": 10 }, { "grad_norm": 2.1187195777893066, "learning_rate": 3.8e-07, "loss": 1.5725, "step": 20 }, { "grad_norm": 1.7767397165298462, "learning_rate": 5.8e-07, "loss": 1.5751, "step": 30 }, { "grad_norm": 1.4611892700195312, "learning_rate": 7.8e-07, "loss": 1.5629, "step": 40 }, { "grad_norm": 1.380027413368225, "learning_rate": 9.8e-07, "loss": 1.5549, "step": 50 }, { "grad_norm": 0.9406078457832336, "learning_rate": 1.18e-06, "loss": 1.5307, "step": 60 }, { "grad_norm": 0.8636054396629333, "learning_rate": 1.3800000000000001e-06, "loss": 1.5177, "step": 70 }, { "grad_norm": 0.559364378452301, "learning_rate": 1.5800000000000003e-06, "loss": 1.5004, "step": 80 }, { "grad_norm": 0.6128845810890198, "learning_rate": 1.7800000000000001e-06, "loss": 1.476, "step": 90 }, { "grad_norm": 0.6962971091270447, "learning_rate": 1.98e-06, "loss": 1.4565, "step": 100 }, { "grad_norm": 0.6421430110931396, "learning_rate": 2.1800000000000003e-06, "loss": 1.4303, "step": 110 }, { "grad_norm": 1.1678168773651123, "learning_rate": 2.38e-06, "loss": 1.3865, "step": 120 }, { "grad_norm": 3.637817621231079, "learning_rate": 2.5800000000000003e-06, "loss": 1.3426, "step": 130 }, { "grad_norm": 1.220801591873169, "learning_rate": 2.78e-06, "loss": 1.2789, "step": 140 }, { "grad_norm": 5.827583312988281, "learning_rate": 2.9800000000000003e-06, "loss": 1.2487, "step": 150 }, { "grad_norm": 0.950404942035675, "learning_rate": 3.1800000000000005e-06, "loss": 1.1934, "step": 160 }, { "grad_norm": 3.9640703201293945, "learning_rate": 3.38e-06, "loss": 1.1882, "step": 170 }, { "grad_norm": 3.234853744506836, "learning_rate": 3.58e-06, "loss": 1.1551, "step": 180 }, { "grad_norm": 2.0852625370025635, "learning_rate": 3.7800000000000002e-06, "loss": 1.1441, "step": 190 }, { "grad_norm": 2.2266321182250977, "learning_rate": 3.98e-06, "loss": 1.0989, "step": 200 }, { "grad_norm": 12.800365447998047, "learning_rate": 4.18e-06, "loss": 1.1333, "step": 210 }, { "grad_norm": 1.9679336547851562, "learning_rate": 4.38e-06, "loss": 1.0816, "step": 220 }, { "grad_norm": 3.1019535064697266, "learning_rate": 4.58e-06, "loss": 1.094, "step": 230 }, { "grad_norm": 2.8569350242614746, "learning_rate": 4.780000000000001e-06, "loss": 1.1288, "step": 240 }, { "grad_norm": 4.602937698364258, "learning_rate": 4.98e-06, "loss": 1.1011, "step": 250 }, { "grad_norm": 1.157566785812378, "learning_rate": 5.18e-06, "loss": 1.0946, "step": 260 }, { "grad_norm": 1.3599668741226196, "learning_rate": 5.38e-06, "loss": 1.1167, "step": 270 }, { "grad_norm": 0.6567723155021667, "learning_rate": 5.580000000000001e-06, "loss": 1.0788, "step": 280 }, { "grad_norm": 4.980498313903809, "learning_rate": 5.78e-06, "loss": 1.1015, "step": 290 }, { "grad_norm": 1.3965166807174683, "learning_rate": 5.98e-06, "loss": 1.065, "step": 300 }, { "grad_norm": 0.9805539846420288, "learning_rate": 6.18e-06, "loss": 1.0728, "step": 310 }, { "grad_norm": 1.4055203199386597, "learning_rate": 6.38e-06, "loss": 1.0769, "step": 320 }, { "grad_norm": 2.513490915298462, "learning_rate": 6.58e-06, "loss": 1.0673, "step": 330 }, { "grad_norm": 0.7278319597244263, "learning_rate": 6.78e-06, "loss": 1.0448, "step": 340 }, { "grad_norm": 1.1559531688690186, "learning_rate": 6.98e-06, "loss": 1.0524, "step": 350 }, { "grad_norm": 0.928160548210144, "learning_rate": 7.180000000000001e-06, "loss": 1.0668, "step": 360 }, { "grad_norm": 1.9730526208877563, "learning_rate": 7.3800000000000005e-06, "loss": 1.0741, "step": 370 }, { "grad_norm": 1.1189223527908325, "learning_rate": 7.580000000000001e-06, "loss": 1.0507, "step": 380 }, { "grad_norm": 0.9704246520996094, "learning_rate": 7.78e-06, "loss": 1.0616, "step": 390 }, { "grad_norm": 1.240396499633789, "learning_rate": 7.98e-06, "loss": 1.0608, "step": 400 }, { "grad_norm": 1.9834959506988525, "learning_rate": 8.18e-06, "loss": 1.0603, "step": 410 }, { "grad_norm": 0.7324520945549011, "learning_rate": 8.380000000000001e-06, "loss": 1.0523, "step": 420 }, { "grad_norm": 4.127306938171387, "learning_rate": 8.580000000000001e-06, "loss": 1.0565, "step": 430 }, { "grad_norm": 0.8520349264144897, "learning_rate": 8.78e-06, "loss": 1.0515, "step": 440 }, { "grad_norm": 0.7327303290367126, "learning_rate": 8.98e-06, "loss": 1.0495, "step": 450 }, { "grad_norm": 0.6935925483703613, "learning_rate": 9.180000000000002e-06, "loss": 1.047, "step": 460 }, { "grad_norm": 1.23868989944458, "learning_rate": 9.38e-06, "loss": 1.0518, "step": 470 }, { "grad_norm": 5.984791278839111, "learning_rate": 9.58e-06, "loss": 1.0373, "step": 480 }, { "grad_norm": 0.8798587322235107, "learning_rate": 9.78e-06, "loss": 1.0216, "step": 490 }, { "grad_norm": 1.0345003604888916, "learning_rate": 9.980000000000001e-06, "loss": 1.0307, "step": 500 }, { "grad_norm": 2.95879864692688, "learning_rate": 1.018e-05, "loss": 1.0241, "step": 510 }, { "grad_norm": 0.8421598672866821, "learning_rate": 1.038e-05, "loss": 1.018, "step": 520 }, { "grad_norm": 1.003597617149353, "learning_rate": 1.058e-05, "loss": 1.0272, "step": 530 }, { "grad_norm": 1.6245428323745728, "learning_rate": 1.0780000000000002e-05, "loss": 1.0094, "step": 540 }, { "grad_norm": 1.8128046989440918, "learning_rate": 1.098e-05, "loss": 0.9926, "step": 550 }, { "grad_norm": 1.0019460916519165, "learning_rate": 1.118e-05, "loss": 0.9902, "step": 560 }, { "grad_norm": 2.670926809310913, "learning_rate": 1.1380000000000001e-05, "loss": 0.994, "step": 570 }, { "grad_norm": 0.8104659914970398, "learning_rate": 1.1580000000000001e-05, "loss": 0.9795, "step": 580 }, { "grad_norm": 0.8928333520889282, "learning_rate": 1.178e-05, "loss": 0.9668, "step": 590 }, { "grad_norm": 0.8439814448356628, "learning_rate": 1.198e-05, "loss": 0.9715, "step": 600 }, { "grad_norm": 1.201135277748108, "learning_rate": 1.2180000000000002e-05, "loss": 0.9671, "step": 610 }, { "grad_norm": 1.9426558017730713, "learning_rate": 1.238e-05, "loss": 0.9346, "step": 620 }, { "grad_norm": 1.365238904953003, "learning_rate": 1.258e-05, "loss": 0.9362, "step": 630 }, { "grad_norm": 1.0579427480697632, "learning_rate": 1.278e-05, "loss": 0.9213, "step": 640 }, { "grad_norm": 2.6099791526794434, "learning_rate": 1.2980000000000001e-05, "loss": 0.9025, "step": 650 }, { "grad_norm": 0.9011796712875366, "learning_rate": 1.3180000000000001e-05, "loss": 0.8832, "step": 660 }, { "grad_norm": 1.4555141925811768, "learning_rate": 1.338e-05, "loss": 0.8763, "step": 670 }, { "grad_norm": 0.9403913617134094, "learning_rate": 1.358e-05, "loss": 0.8574, "step": 680 }, { "grad_norm": 2.6945314407348633, "learning_rate": 1.3780000000000002e-05, "loss": 0.847, "step": 690 }, { "grad_norm": 0.9131345748901367, "learning_rate": 1.3980000000000002e-05, "loss": 0.839, "step": 700 }, { "grad_norm": 1.7176071405410767, "learning_rate": 1.4180000000000001e-05, "loss": 0.8298, "step": 710 }, { "grad_norm": 1.593785285949707, "learning_rate": 1.4380000000000001e-05, "loss": 0.8125, "step": 720 }, { "grad_norm": 1.041443943977356, "learning_rate": 1.4580000000000003e-05, "loss": 0.7978, "step": 730 }, { "grad_norm": 3.8084022998809814, "learning_rate": 1.4779999999999999e-05, "loss": 0.8011, "step": 740 }, { "grad_norm": 2.5635263919830322, "learning_rate": 1.4979999999999999e-05, "loss": 0.7756, "step": 750 }, { "grad_norm": 1.4510605335235596, "learning_rate": 1.518e-05, "loss": 0.7622, "step": 760 }, { "grad_norm": 1.217301368713379, "learning_rate": 1.538e-05, "loss": 0.7482, "step": 770 }, { "grad_norm": 1.2995609045028687, "learning_rate": 1.558e-05, "loss": 0.7479, "step": 780 }, { "grad_norm": 2.4616358280181885, "learning_rate": 1.578e-05, "loss": 0.7163, "step": 790 }, { "grad_norm": 1.2900861501693726, "learning_rate": 1.598e-05, "loss": 0.7143, "step": 800 }, { "grad_norm": 1.5790479183197021, "learning_rate": 1.618e-05, "loss": 0.7092, "step": 810 }, { "grad_norm": 1.2235549688339233, "learning_rate": 1.6380000000000002e-05, "loss": 0.6865, "step": 820 }, { "grad_norm": 2.3063037395477295, "learning_rate": 1.658e-05, "loss": 0.6987, "step": 830 }, { "grad_norm": 0.9557511210441589, "learning_rate": 1.6780000000000002e-05, "loss": 0.6777, "step": 840 }, { "grad_norm": 2.530147075653076, "learning_rate": 1.698e-05, "loss": 0.681, "step": 850 }, { "grad_norm": 1.340712308883667, "learning_rate": 1.718e-05, "loss": 0.6505, "step": 860 }, { "grad_norm": 1.330559253692627, "learning_rate": 1.7380000000000003e-05, "loss": 0.6414, "step": 870 }, { "grad_norm": 1.8326561450958252, "learning_rate": 1.758e-05, "loss": 0.6219, "step": 880 }, { "grad_norm": 1.3623476028442383, "learning_rate": 1.7780000000000003e-05, "loss": 0.6025, "step": 890 }, { "grad_norm": 1.5193125009536743, "learning_rate": 1.798e-05, "loss": 0.6158, "step": 900 }, { "grad_norm": 1.45945143699646, "learning_rate": 1.818e-05, "loss": 0.5826, "step": 910 }, { "grad_norm": 1.315100908279419, "learning_rate": 1.838e-05, "loss": 0.586, "step": 920 }, { "grad_norm": 1.1674245595932007, "learning_rate": 1.858e-05, "loss": 0.5599, "step": 930 }, { "grad_norm": 1.031245231628418, "learning_rate": 1.878e-05, "loss": 0.5587, "step": 940 }, { "grad_norm": 1.1666947603225708, "learning_rate": 1.898e-05, "loss": 0.5533, "step": 950 }, { "grad_norm": 1.4355655908584595, "learning_rate": 1.918e-05, "loss": 0.5228, "step": 960 }, { "grad_norm": 1.5651236772537231, "learning_rate": 1.938e-05, "loss": 0.5246, "step": 970 }, { "grad_norm": 1.1883115768432617, "learning_rate": 1.9580000000000002e-05, "loss": 0.5199, "step": 980 }, { "grad_norm": 1.5693440437316895, "learning_rate": 1.978e-05, "loss": 0.5011, "step": 990 }, { "grad_norm": 1.5708034038543701, "learning_rate": 1.9980000000000002e-05, "loss": 0.5092, "step": 1000 }, { "grad_norm": 2.426574468612671, "learning_rate": 2.0180000000000003e-05, "loss": 0.5112, "step": 1010 }, { "grad_norm": 1.1455503702163696, "learning_rate": 2.038e-05, "loss": 0.485, "step": 1020 }, { "grad_norm": 1.4592437744140625, "learning_rate": 2.0580000000000003e-05, "loss": 0.4693, "step": 1030 }, { "grad_norm": 1.3149504661560059, "learning_rate": 2.078e-05, "loss": 0.4896, "step": 1040 }, { "grad_norm": 1.2561594247817993, "learning_rate": 2.098e-05, "loss": 0.4617, "step": 1050 }, { "grad_norm": 1.54478919506073, "learning_rate": 2.118e-05, "loss": 0.4673, "step": 1060 }, { "grad_norm": 1.3093321323394775, "learning_rate": 2.138e-05, "loss": 0.4435, "step": 1070 }, { "grad_norm": 1.1475777626037598, "learning_rate": 2.158e-05, "loss": 0.4314, "step": 1080 }, { "grad_norm": 1.414124608039856, "learning_rate": 2.178e-05, "loss": 0.4055, "step": 1090 }, { "grad_norm": 1.1952828168869019, "learning_rate": 2.198e-05, "loss": 0.4133, "step": 1100 }, { "grad_norm": 1.3651411533355713, "learning_rate": 2.218e-05, "loss": 0.4108, "step": 1110 }, { "grad_norm": 1.6452419757843018, "learning_rate": 2.2380000000000003e-05, "loss": 0.4091, "step": 1120 }, { "grad_norm": 1.48700749874115, "learning_rate": 2.258e-05, "loss": 0.4044, "step": 1130 }, { "grad_norm": 1.519075870513916, "learning_rate": 2.2780000000000002e-05, "loss": 0.3855, "step": 1140 }, { "grad_norm": 1.7138876914978027, "learning_rate": 2.298e-05, "loss": 0.3773, "step": 1150 }, { "grad_norm": 1.4698295593261719, "learning_rate": 2.318e-05, "loss": 0.3727, "step": 1160 }, { "grad_norm": 1.5901681184768677, "learning_rate": 2.3380000000000003e-05, "loss": 0.3661, "step": 1170 }, { "grad_norm": 1.2559553384780884, "learning_rate": 2.358e-05, "loss": 0.3501, "step": 1180 }, { "grad_norm": 1.6344045400619507, "learning_rate": 2.3780000000000003e-05, "loss": 0.3538, "step": 1190 }, { "grad_norm": 1.4126002788543701, "learning_rate": 2.398e-05, "loss": 0.3488, "step": 1200 }, { "grad_norm": 1.3249131441116333, "learning_rate": 2.418e-05, "loss": 0.3369, "step": 1210 }, { "grad_norm": 1.8253049850463867, "learning_rate": 2.438e-05, "loss": 0.32, "step": 1220 }, { "grad_norm": 1.4511300325393677, "learning_rate": 2.4580000000000002e-05, "loss": 0.3113, "step": 1230 }, { "grad_norm": 1.657767653465271, "learning_rate": 2.478e-05, "loss": 0.3047, "step": 1240 }, { "grad_norm": 1.3572914600372314, "learning_rate": 2.498e-05, "loss": 0.3087, "step": 1250 }, { "grad_norm": 1.2990957498550415, "learning_rate": 2.5180000000000003e-05, "loss": 0.2836, "step": 1260 }, { "grad_norm": 1.9438824653625488, "learning_rate": 2.5380000000000004e-05, "loss": 0.2949, "step": 1270 }, { "grad_norm": 1.4552688598632812, "learning_rate": 2.5580000000000002e-05, "loss": 0.2886, "step": 1280 }, { "grad_norm": 1.5942091941833496, "learning_rate": 2.5779999999999997e-05, "loss": 0.28, "step": 1290 }, { "grad_norm": 1.5174213647842407, "learning_rate": 2.598e-05, "loss": 0.2783, "step": 1300 }, { "grad_norm": 1.8981564044952393, "learning_rate": 2.618e-05, "loss": 0.2616, "step": 1310 }, { "grad_norm": 1.5345311164855957, "learning_rate": 2.6379999999999998e-05, "loss": 0.2861, "step": 1320 }, { "grad_norm": 1.8726272583007812, "learning_rate": 2.658e-05, "loss": 0.2441, "step": 1330 }, { "grad_norm": 1.387172818183899, "learning_rate": 2.678e-05, "loss": 0.2535, "step": 1340 }, { "grad_norm": 1.310351848602295, "learning_rate": 2.698e-05, "loss": 0.229, "step": 1350 }, { "grad_norm": 1.4305953979492188, "learning_rate": 2.718e-05, "loss": 0.2313, "step": 1360 }, { "grad_norm": 1.7435038089752197, "learning_rate": 2.738e-05, "loss": 0.2176, "step": 1370 }, { "grad_norm": 1.6080414056777954, "learning_rate": 2.758e-05, "loss": 0.2196, "step": 1380 }, { "grad_norm": 1.2151384353637695, "learning_rate": 2.778e-05, "loss": 0.207, "step": 1390 }, { "grad_norm": 1.444219946861267, "learning_rate": 2.798e-05, "loss": 0.2118, "step": 1400 }, { "grad_norm": 1.6158922910690308, "learning_rate": 2.818e-05, "loss": 0.1972, "step": 1410 }, { "grad_norm": 1.524274468421936, "learning_rate": 2.8380000000000003e-05, "loss": 0.2069, "step": 1420 }, { "grad_norm": 1.608817458152771, "learning_rate": 2.858e-05, "loss": 0.1922, "step": 1430 }, { "grad_norm": 1.763902187347412, "learning_rate": 2.8780000000000002e-05, "loss": 0.1982, "step": 1440 }, { "grad_norm": 2.1931698322296143, "learning_rate": 2.898e-05, "loss": 0.1946, "step": 1450 }, { "grad_norm": 1.7323309183120728, "learning_rate": 2.9180000000000002e-05, "loss": 0.1803, "step": 1460 }, { "grad_norm": 1.2583523988723755, "learning_rate": 2.9380000000000003e-05, "loss": 0.1817, "step": 1470 }, { "grad_norm": 1.5437161922454834, "learning_rate": 2.958e-05, "loss": 0.1773, "step": 1480 }, { "grad_norm": 1.7676451206207275, "learning_rate": 2.9780000000000003e-05, "loss": 0.1709, "step": 1490 }, { "grad_norm": 2.115121841430664, "learning_rate": 2.998e-05, "loss": 0.1894, "step": 1500 }, { "grad_norm": 1.4707292318344116, "learning_rate": 3.0180000000000002e-05, "loss": 0.165, "step": 1510 }, { "grad_norm": 1.7382837533950806, "learning_rate": 3.0380000000000004e-05, "loss": 0.1644, "step": 1520 }, { "grad_norm": 1.4237769842147827, "learning_rate": 3.058e-05, "loss": 0.1499, "step": 1530 }, { "grad_norm": 1.5430965423583984, "learning_rate": 3.078e-05, "loss": 0.1627, "step": 1540 }, { "grad_norm": 1.5984359979629517, "learning_rate": 3.0980000000000005e-05, "loss": 0.1489, "step": 1550 }, { "grad_norm": 1.4609631299972534, "learning_rate": 3.118e-05, "loss": 0.1458, "step": 1560 }, { "grad_norm": 1.4382436275482178, "learning_rate": 3.138e-05, "loss": 0.1416, "step": 1570 }, { "grad_norm": 1.4270118474960327, "learning_rate": 3.1580000000000006e-05, "loss": 0.1404, "step": 1580 }, { "grad_norm": 1.4451384544372559, "learning_rate": 3.1780000000000004e-05, "loss": 0.1378, "step": 1590 }, { "grad_norm": 1.6566460132598877, "learning_rate": 3.198e-05, "loss": 0.1522, "step": 1600 }, { "grad_norm": 1.3700045347213745, "learning_rate": 3.218e-05, "loss": 0.1357, "step": 1610 }, { "grad_norm": 1.1937892436981201, "learning_rate": 3.238e-05, "loss": 0.1385, "step": 1620 }, { "grad_norm": 2.0586397647857666, "learning_rate": 3.2579999999999996e-05, "loss": 0.1522, "step": 1630 }, { "grad_norm": 1.4003300666809082, "learning_rate": 3.278e-05, "loss": 0.1498, "step": 1640 }, { "grad_norm": 1.4799567461013794, "learning_rate": 3.298e-05, "loss": 0.1466, "step": 1650 }, { "grad_norm": 1.336797833442688, "learning_rate": 3.318e-05, "loss": 0.1365, "step": 1660 }, { "grad_norm": 1.907181739807129, "learning_rate": 3.338e-05, "loss": 0.1582, "step": 1670 }, { "grad_norm": 1.5125823020935059, "learning_rate": 3.358e-05, "loss": 0.1479, "step": 1680 }, { "grad_norm": 1.467471718788147, "learning_rate": 3.378e-05, "loss": 0.1395, "step": 1690 }, { "grad_norm": 2.3545548915863037, "learning_rate": 3.398e-05, "loss": 0.1417, "step": 1700 }, { "grad_norm": 1.3996036052703857, "learning_rate": 3.418e-05, "loss": 0.141, "step": 1710 }, { "grad_norm": 1.3959155082702637, "learning_rate": 3.438e-05, "loss": 0.158, "step": 1720 }, { "grad_norm": 1.4140926599502563, "learning_rate": 3.4580000000000004e-05, "loss": 0.1395, "step": 1730 }, { "grad_norm": 1.0964840650558472, "learning_rate": 3.478e-05, "loss": 0.1325, "step": 1740 }, { "grad_norm": 1.3664647340774536, "learning_rate": 3.498e-05, "loss": 0.1345, "step": 1750 }, { "grad_norm": 1.1473278999328613, "learning_rate": 3.518e-05, "loss": 0.1359, "step": 1760 }, { "grad_norm": 1.5458581447601318, "learning_rate": 3.5380000000000003e-05, "loss": 0.1273, "step": 1770 }, { "grad_norm": 1.4155246019363403, "learning_rate": 3.558e-05, "loss": 0.1446, "step": 1780 }, { "grad_norm": 1.5580581426620483, "learning_rate": 3.578e-05, "loss": 0.1611, "step": 1790 }, { "grad_norm": 1.4860072135925293, "learning_rate": 3.5980000000000004e-05, "loss": 0.1428, "step": 1800 }, { "grad_norm": 2.061711072921753, "learning_rate": 3.618e-05, "loss": 0.1435, "step": 1810 }, { "grad_norm": 1.1988595724105835, "learning_rate": 3.638e-05, "loss": 0.1351, "step": 1820 }, { "grad_norm": 1.8205078840255737, "learning_rate": 3.6580000000000006e-05, "loss": 0.1424, "step": 1830 }, { "grad_norm": 1.2913676500320435, "learning_rate": 3.6780000000000004e-05, "loss": 0.1421, "step": 1840 }, { "grad_norm": 1.2723727226257324, "learning_rate": 3.698e-05, "loss": 0.1401, "step": 1850 }, { "grad_norm": 1.2577519416809082, "learning_rate": 3.7180000000000007e-05, "loss": 0.1392, "step": 1860 }, { "grad_norm": 1.7551569938659668, "learning_rate": 3.7380000000000005e-05, "loss": 0.1558, "step": 1870 }, { "grad_norm": 1.5851534605026245, "learning_rate": 3.758e-05, "loss": 0.133, "step": 1880 }, { "grad_norm": 1.1263703107833862, "learning_rate": 3.778000000000001e-05, "loss": 0.1235, "step": 1890 }, { "grad_norm": 1.109619140625, "learning_rate": 3.7980000000000006e-05, "loss": 0.1345, "step": 1900 }, { "grad_norm": 1.7170437574386597, "learning_rate": 3.818e-05, "loss": 0.1241, "step": 1910 }, { "grad_norm": 1.1947234869003296, "learning_rate": 3.838e-05, "loss": 0.1212, "step": 1920 }, { "grad_norm": 1.4559831619262695, "learning_rate": 3.858e-05, "loss": 0.1213, "step": 1930 }, { "grad_norm": 1.3905948400497437, "learning_rate": 3.878e-05, "loss": 0.1151, "step": 1940 }, { "grad_norm": 1.5020700693130493, "learning_rate": 3.898e-05, "loss": 0.1077, "step": 1950 }, { "grad_norm": 1.0593889951705933, "learning_rate": 3.918e-05, "loss": 0.1013, "step": 1960 }, { "grad_norm": 0.9968692660331726, "learning_rate": 3.938e-05, "loss": 0.1149, "step": 1970 }, { "grad_norm": 1.207639455795288, "learning_rate": 3.958e-05, "loss": 0.1177, "step": 1980 }, { "grad_norm": 1.322926640510559, "learning_rate": 3.978e-05, "loss": 0.0987, "step": 1990 }, { "grad_norm": 1.8420583009719849, "learning_rate": 3.998e-05, "loss": 0.0996, "step": 2000 }, { "grad_norm": 1.145195722579956, "learning_rate": 4.018e-05, "loss": 0.0882, "step": 2010 }, { "grad_norm": 1.4924179315567017, "learning_rate": 4.038e-05, "loss": 0.1014, "step": 2020 }, { "grad_norm": 1.3658872842788696, "learning_rate": 4.058e-05, "loss": 0.1007, "step": 2030 }, { "grad_norm": 1.5268629789352417, "learning_rate": 4.078e-05, "loss": 0.1161, "step": 2040 }, { "grad_norm": 2.686342477798462, "learning_rate": 4.0980000000000004e-05, "loss": 0.118, "step": 2050 }, { "grad_norm": 1.2094306945800781, "learning_rate": 4.118e-05, "loss": 0.0999, "step": 2060 }, { "grad_norm": 1.805381178855896, "learning_rate": 4.138e-05, "loss": 0.0921, "step": 2070 }, { "grad_norm": 1.464465856552124, "learning_rate": 4.1580000000000005e-05, "loss": 0.0882, "step": 2080 }, { "grad_norm": 0.9803165197372437, "learning_rate": 4.178e-05, "loss": 0.0893, "step": 2090 }, { "grad_norm": 1.1378520727157593, "learning_rate": 4.198e-05, "loss": 0.0912, "step": 2100 }, { "grad_norm": 1.046370267868042, "learning_rate": 4.2180000000000006e-05, "loss": 0.0941, "step": 2110 }, { "grad_norm": 1.425757884979248, "learning_rate": 4.2380000000000004e-05, "loss": 0.0924, "step": 2120 }, { "grad_norm": 1.1320098638534546, "learning_rate": 4.258e-05, "loss": 0.0862, "step": 2130 }, { "grad_norm": 1.247001051902771, "learning_rate": 4.278e-05, "loss": 0.0933, "step": 2140 }, { "grad_norm": 0.9248917102813721, "learning_rate": 4.2980000000000005e-05, "loss": 0.0944, "step": 2150 }, { "grad_norm": 1.3971987962722778, "learning_rate": 4.318e-05, "loss": 0.0997, "step": 2160 }, { "grad_norm": 1.2635126113891602, "learning_rate": 4.338e-05, "loss": 0.0926, "step": 2170 }, { "grad_norm": 1.5223133563995361, "learning_rate": 4.3580000000000006e-05, "loss": 0.0963, "step": 2180 }, { "grad_norm": 1.2234301567077637, "learning_rate": 4.3780000000000004e-05, "loss": 0.0874, "step": 2190 }, { "grad_norm": 0.9697202444076538, "learning_rate": 4.398e-05, "loss": 0.0819, "step": 2200 }, { "grad_norm": 1.1768534183502197, "learning_rate": 4.418000000000001e-05, "loss": 0.0916, "step": 2210 }, { "grad_norm": 1.3359019756317139, "learning_rate": 4.438e-05, "loss": 0.0934, "step": 2220 }, { "grad_norm": 1.2703652381896973, "learning_rate": 4.458e-05, "loss": 0.0991, "step": 2230 }, { "grad_norm": 1.1888983249664307, "learning_rate": 4.478e-05, "loss": 0.0943, "step": 2240 }, { "grad_norm": 1.0050462484359741, "learning_rate": 4.498e-05, "loss": 0.1082, "step": 2250 }, { "grad_norm": 1.6533628702163696, "learning_rate": 4.518e-05, "loss": 0.1032, "step": 2260 }, { "grad_norm": 1.007835030555725, "learning_rate": 4.538e-05, "loss": 0.0881, "step": 2270 }, { "grad_norm": 1.0498948097229004, "learning_rate": 4.558e-05, "loss": 0.0948, "step": 2280 }, { "grad_norm": 0.963283121585846, "learning_rate": 4.578e-05, "loss": 0.0841, "step": 2290 }, { "grad_norm": 1.1583718061447144, "learning_rate": 4.5980000000000004e-05, "loss": 0.0768, "step": 2300 }, { "grad_norm": 0.9706757664680481, "learning_rate": 4.618e-05, "loss": 0.1014, "step": 2310 }, { "grad_norm": 1.0057862997055054, "learning_rate": 4.638e-05, "loss": 0.0831, "step": 2320 }, { "grad_norm": 1.2952053546905518, "learning_rate": 4.6580000000000005e-05, "loss": 0.1028, "step": 2330 }, { "grad_norm": 1.0748975276947021, "learning_rate": 4.678e-05, "loss": 0.0951, "step": 2340 }, { "grad_norm": 1.339827299118042, "learning_rate": 4.698e-05, "loss": 0.0914, "step": 2350 }, { "grad_norm": 1.6751843690872192, "learning_rate": 4.718e-05, "loss": 0.0911, "step": 2360 }, { "grad_norm": 1.2442045211791992, "learning_rate": 4.7380000000000004e-05, "loss": 0.0883, "step": 2370 }, { "grad_norm": 2.1004133224487305, "learning_rate": 4.758e-05, "loss": 0.0879, "step": 2380 }, { "grad_norm": 1.161981463432312, "learning_rate": 4.778e-05, "loss": 0.1086, "step": 2390 }, { "grad_norm": 1.316776156425476, "learning_rate": 4.7980000000000005e-05, "loss": 0.1083, "step": 2400 }, { "grad_norm": 1.8611112833023071, "learning_rate": 4.818e-05, "loss": 0.1209, "step": 2410 }, { "grad_norm": 1.2899459600448608, "learning_rate": 4.838e-05, "loss": 0.1103, "step": 2420 }, { "grad_norm": 1.369541049003601, "learning_rate": 4.8580000000000006e-05, "loss": 0.1085, "step": 2430 }, { "grad_norm": 1.3191884756088257, "learning_rate": 4.8780000000000004e-05, "loss": 0.1081, "step": 2440 }, { "grad_norm": 1.6112862825393677, "learning_rate": 4.898e-05, "loss": 0.0962, "step": 2450 }, { "grad_norm": 1.0228184461593628, "learning_rate": 4.918000000000001e-05, "loss": 0.0974, "step": 2460 }, { "grad_norm": 1.6294655799865723, "learning_rate": 4.9380000000000005e-05, "loss": 0.0878, "step": 2470 }, { "grad_norm": 1.076183795928955, "learning_rate": 4.958e-05, "loss": 0.0907, "step": 2480 }, { "grad_norm": 1.1367418766021729, "learning_rate": 4.978e-05, "loss": 0.0845, "step": 2490 }, { "grad_norm": 1.0144062042236328, "learning_rate": 4.9980000000000006e-05, "loss": 0.0898, "step": 2500 }, { "grad_norm": 0.9943013191223145, "learning_rate": 5.0180000000000004e-05, "loss": 0.0889, "step": 2510 }, { "grad_norm": 0.8168659806251526, "learning_rate": 5.038e-05, "loss": 0.0916, "step": 2520 }, { "grad_norm": 1.3381121158599854, "learning_rate": 5.058000000000001e-05, "loss": 0.0831, "step": 2530 }, { "grad_norm": 1.4005751609802246, "learning_rate": 5.0780000000000005e-05, "loss": 0.1036, "step": 2540 }, { "grad_norm": 1.025732398033142, "learning_rate": 5.098e-05, "loss": 0.0982, "step": 2550 }, { "grad_norm": 1.1671310663223267, "learning_rate": 5.118000000000001e-05, "loss": 0.0983, "step": 2560 }, { "grad_norm": 1.1943354606628418, "learning_rate": 5.1380000000000006e-05, "loss": 0.09, "step": 2570 }, { "grad_norm": 0.9647271037101746, "learning_rate": 5.1580000000000004e-05, "loss": 0.0934, "step": 2580 }, { "grad_norm": 0.9960209131240845, "learning_rate": 5.178000000000001e-05, "loss": 0.0909, "step": 2590 }, { "grad_norm": 1.2893314361572266, "learning_rate": 5.198000000000001e-05, "loss": 0.1059, "step": 2600 }, { "grad_norm": 1.1174474954605103, "learning_rate": 5.2180000000000005e-05, "loss": 0.0895, "step": 2610 }, { "grad_norm": 0.8401209712028503, "learning_rate": 5.238000000000001e-05, "loss": 0.0909, "step": 2620 }, { "grad_norm": 0.8056483268737793, "learning_rate": 5.258000000000001e-05, "loss": 0.0895, "step": 2630 }, { "grad_norm": 1.0440040826797485, "learning_rate": 5.2780000000000006e-05, "loss": 0.0871, "step": 2640 }, { "grad_norm": 0.7858893871307373, "learning_rate": 5.2980000000000004e-05, "loss": 0.0889, "step": 2650 }, { "grad_norm": 1.2451766729354858, "learning_rate": 5.318000000000001e-05, "loss": 0.099, "step": 2660 }, { "grad_norm": 0.9057985544204712, "learning_rate": 5.338000000000001e-05, "loss": 0.0871, "step": 2670 }, { "grad_norm": 1.0599905252456665, "learning_rate": 5.3580000000000005e-05, "loss": 0.0851, "step": 2680 }, { "grad_norm": 1.044230341911316, "learning_rate": 5.378e-05, "loss": 0.0811, "step": 2690 }, { "grad_norm": 1.134655237197876, "learning_rate": 5.3979999999999995e-05, "loss": 0.0875, "step": 2700 }, { "grad_norm": 0.9624427556991577, "learning_rate": 5.418e-05, "loss": 0.0805, "step": 2710 }, { "grad_norm": 0.9350464344024658, "learning_rate": 5.438e-05, "loss": 0.0854, "step": 2720 }, { "grad_norm": 1.0496246814727783, "learning_rate": 5.4579999999999996e-05, "loss": 0.0912, "step": 2730 }, { "grad_norm": 1.882257103919983, "learning_rate": 5.478e-05, "loss": 0.0902, "step": 2740 }, { "grad_norm": 1.090551495552063, "learning_rate": 5.498e-05, "loss": 0.0743, "step": 2750 }, { "grad_norm": 1.1900626420974731, "learning_rate": 5.518e-05, "loss": 0.085, "step": 2760 }, { "grad_norm": 0.8351054787635803, "learning_rate": 5.538e-05, "loss": 0.0812, "step": 2770 }, { "grad_norm": 0.8111416101455688, "learning_rate": 5.558e-05, "loss": 0.0765, "step": 2780 }, { "grad_norm": 0.9293792843818665, "learning_rate": 5.578e-05, "loss": 0.0716, "step": 2790 }, { "grad_norm": 0.9099953770637512, "learning_rate": 5.5979999999999996e-05, "loss": 0.097, "step": 2800 }, { "grad_norm": 1.0297311544418335, "learning_rate": 5.618e-05, "loss": 0.0755, "step": 2810 }, { "grad_norm": 1.0780147314071655, "learning_rate": 5.638e-05, "loss": 0.0767, "step": 2820 }, { "grad_norm": 0.8332295417785645, "learning_rate": 5.658e-05, "loss": 0.072, "step": 2830 }, { "grad_norm": 0.8472347855567932, "learning_rate": 5.678e-05, "loss": 0.0728, "step": 2840 }, { "grad_norm": 1.113635540008545, "learning_rate": 5.698e-05, "loss": 0.0823, "step": 2850 }, { "grad_norm": 1.1431188583374023, "learning_rate": 5.718e-05, "loss": 0.082, "step": 2860 }, { "grad_norm": 0.7382771968841553, "learning_rate": 5.738e-05, "loss": 0.0757, "step": 2870 }, { "grad_norm": 1.1114126443862915, "learning_rate": 5.758e-05, "loss": 0.0792, "step": 2880 }, { "grad_norm": 1.1058228015899658, "learning_rate": 5.778e-05, "loss": 0.0809, "step": 2890 }, { "grad_norm": 1.086748719215393, "learning_rate": 5.7980000000000004e-05, "loss": 0.0741, "step": 2900 }, { "grad_norm": 0.9133238792419434, "learning_rate": 5.818e-05, "loss": 0.0933, "step": 2910 }, { "grad_norm": 0.8397288918495178, "learning_rate": 5.838e-05, "loss": 0.0716, "step": 2920 }, { "grad_norm": 0.8847182393074036, "learning_rate": 5.858e-05, "loss": 0.0674, "step": 2930 }, { "grad_norm": 1.1472963094711304, "learning_rate": 5.878e-05, "loss": 0.0735, "step": 2940 }, { "grad_norm": 1.544686198234558, "learning_rate": 5.898e-05, "loss": 0.0833, "step": 2950 }, { "grad_norm": 1.0344477891921997, "learning_rate": 5.918e-05, "loss": 0.0861, "step": 2960 }, { "grad_norm": 1.2067776918411255, "learning_rate": 5.9380000000000004e-05, "loss": 0.0861, "step": 2970 }, { "grad_norm": 0.7481675148010254, "learning_rate": 5.958e-05, "loss": 0.0822, "step": 2980 }, { "grad_norm": 0.9611971378326416, "learning_rate": 5.978e-05, "loss": 0.0667, "step": 2990 }, { "grad_norm": 0.9111663699150085, "learning_rate": 5.9980000000000005e-05, "loss": 0.076, "step": 3000 }, { "grad_norm": 0.9374921321868896, "learning_rate": 6.018e-05, "loss": 0.0764, "step": 3010 }, { "grad_norm": 0.9026698470115662, "learning_rate": 6.038e-05, "loss": 0.0781, "step": 3020 }, { "grad_norm": 1.125892996788025, "learning_rate": 6.0580000000000006e-05, "loss": 0.0751, "step": 3030 }, { "grad_norm": 1.009718418121338, "learning_rate": 6.0780000000000004e-05, "loss": 0.0739, "step": 3040 }, { "grad_norm": 0.8881495594978333, "learning_rate": 6.098e-05, "loss": 0.0822, "step": 3050 }, { "grad_norm": 0.7472686171531677, "learning_rate": 6.118000000000001e-05, "loss": 0.0789, "step": 3060 }, { "grad_norm": 1.0667312145233154, "learning_rate": 6.138e-05, "loss": 0.0864, "step": 3070 }, { "grad_norm": 0.8416674137115479, "learning_rate": 6.158e-05, "loss": 0.0771, "step": 3080 }, { "grad_norm": 0.8106931447982788, "learning_rate": 6.178000000000001e-05, "loss": 0.0864, "step": 3090 }, { "grad_norm": 1.4251898527145386, "learning_rate": 6.198e-05, "loss": 0.0968, "step": 3100 }, { "grad_norm": 0.8575820326805115, "learning_rate": 6.218e-05, "loss": 0.0985, "step": 3110 }, { "grad_norm": 1.840651273727417, "learning_rate": 6.238000000000001e-05, "loss": 0.0975, "step": 3120 }, { "grad_norm": 1.0593422651290894, "learning_rate": 6.258e-05, "loss": 0.0964, "step": 3130 }, { "grad_norm": 0.9229077100753784, "learning_rate": 6.278e-05, "loss": 0.0884, "step": 3140 }, { "grad_norm": 1.2586861848831177, "learning_rate": 6.298000000000001e-05, "loss": 0.0923, "step": 3150 }, { "grad_norm": 0.9044884443283081, "learning_rate": 6.318e-05, "loss": 0.0861, "step": 3160 }, { "grad_norm": 0.7890645861625671, "learning_rate": 6.338e-05, "loss": 0.0885, "step": 3170 }, { "grad_norm": 0.8676195740699768, "learning_rate": 6.358000000000001e-05, "loss": 0.079, "step": 3180 }, { "grad_norm": 0.974872887134552, "learning_rate": 6.378e-05, "loss": 0.0752, "step": 3190 }, { "grad_norm": 0.8163201212882996, "learning_rate": 6.398000000000001e-05, "loss": 0.0775, "step": 3200 }, { "grad_norm": 0.8499476909637451, "learning_rate": 6.418000000000001e-05, "loss": 0.0728, "step": 3210 }, { "grad_norm": 0.9375150799751282, "learning_rate": 6.438e-05, "loss": 0.0673, "step": 3220 }, { "grad_norm": 0.5676933526992798, "learning_rate": 6.458000000000001e-05, "loss": 0.0777, "step": 3230 }, { "grad_norm": 0.7984086871147156, "learning_rate": 6.478000000000001e-05, "loss": 0.073, "step": 3240 }, { "grad_norm": 0.8062030076980591, "learning_rate": 6.498e-05, "loss": 0.0688, "step": 3250 }, { "grad_norm": 0.7415264248847961, "learning_rate": 6.518000000000001e-05, "loss": 0.0863, "step": 3260 }, { "grad_norm": 0.7692404389381409, "learning_rate": 6.538000000000001e-05, "loss": 0.0744, "step": 3270 }, { "grad_norm": 0.7777807116508484, "learning_rate": 6.558e-05, "loss": 0.0759, "step": 3280 }, { "grad_norm": 0.9378018379211426, "learning_rate": 6.578000000000001e-05, "loss": 0.0724, "step": 3290 }, { "grad_norm": 0.9101012945175171, "learning_rate": 6.598e-05, "loss": 0.07, "step": 3300 }, { "grad_norm": 0.72568279504776, "learning_rate": 6.618e-05, "loss": 0.0712, "step": 3310 }, { "grad_norm": 0.7288804054260254, "learning_rate": 6.638e-05, "loss": 0.0649, "step": 3320 }, { "grad_norm": 0.744612991809845, "learning_rate": 6.658e-05, "loss": 0.065, "step": 3330 }, { "grad_norm": 0.8648169636726379, "learning_rate": 6.678e-05, "loss": 0.0693, "step": 3340 }, { "grad_norm": 0.9377201199531555, "learning_rate": 6.698e-05, "loss": 0.072, "step": 3350 }, { "grad_norm": 0.8420950770378113, "learning_rate": 6.718e-05, "loss": 0.0682, "step": 3360 }, { "grad_norm": 0.7668231129646301, "learning_rate": 6.738e-05, "loss": 0.0777, "step": 3370 }, { "grad_norm": 0.6314890384674072, "learning_rate": 6.758e-05, "loss": 0.0701, "step": 3380 }, { "grad_norm": 0.857178807258606, "learning_rate": 6.778e-05, "loss": 0.0762, "step": 3390 }, { "grad_norm": 0.7089142799377441, "learning_rate": 6.798e-05, "loss": 0.0776, "step": 3400 }, { "grad_norm": 0.7062733173370361, "learning_rate": 6.818e-05, "loss": 0.0703, "step": 3410 }, { "grad_norm": 0.6975285410881042, "learning_rate": 6.838e-05, "loss": 0.0683, "step": 3420 }, { "grad_norm": 0.6709898114204407, "learning_rate": 6.858e-05, "loss": 0.0631, "step": 3430 }, { "grad_norm": 0.8096795082092285, "learning_rate": 6.878e-05, "loss": 0.0673, "step": 3440 }, { "grad_norm": 0.6939607262611389, "learning_rate": 6.898e-05, "loss": 0.0629, "step": 3450 }, { "grad_norm": 0.8013080954551697, "learning_rate": 6.918e-05, "loss": 0.0716, "step": 3460 }, { "grad_norm": 0.859581470489502, "learning_rate": 6.938e-05, "loss": 0.0661, "step": 3470 }, { "grad_norm": 0.755018949508667, "learning_rate": 6.958e-05, "loss": 0.0733, "step": 3480 }, { "grad_norm": 1.074064016342163, "learning_rate": 6.978e-05, "loss": 0.0777, "step": 3490 }, { "grad_norm": 0.7277959585189819, "learning_rate": 6.998e-05, "loss": 0.0768, "step": 3500 }, { "grad_norm": 0.7167660593986511, "learning_rate": 7.018e-05, "loss": 0.0741, "step": 3510 }, { "grad_norm": 1.158371925354004, "learning_rate": 7.038e-05, "loss": 0.0721, "step": 3520 }, { "grad_norm": 0.7432733178138733, "learning_rate": 7.058e-05, "loss": 0.0673, "step": 3530 }, { "grad_norm": 0.6698697805404663, "learning_rate": 7.078e-05, "loss": 0.0764, "step": 3540 }, { "grad_norm": 0.7771131992340088, "learning_rate": 7.098e-05, "loss": 0.0688, "step": 3550 }, { "grad_norm": 1.0553028583526611, "learning_rate": 7.118e-05, "loss": 0.0738, "step": 3560 }, { "grad_norm": 0.7653899192810059, "learning_rate": 7.138e-05, "loss": 0.0718, "step": 3570 }, { "grad_norm": 0.959377110004425, "learning_rate": 7.158e-05, "loss": 0.0714, "step": 3580 }, { "grad_norm": 0.8016287088394165, "learning_rate": 7.178000000000001e-05, "loss": 0.0697, "step": 3590 }, { "grad_norm": 0.8576526641845703, "learning_rate": 7.198e-05, "loss": 0.0771, "step": 3600 }, { "grad_norm": 0.7468317747116089, "learning_rate": 7.218e-05, "loss": 0.0614, "step": 3610 }, { "grad_norm": 0.7670191526412964, "learning_rate": 7.238000000000001e-05, "loss": 0.0649, "step": 3620 }, { "grad_norm": 1.8396720886230469, "learning_rate": 7.258e-05, "loss": 0.0708, "step": 3630 }, { "grad_norm": 0.7263420820236206, "learning_rate": 7.278e-05, "loss": 0.0703, "step": 3640 }, { "grad_norm": 0.7907016277313232, "learning_rate": 7.298000000000001e-05, "loss": 0.0762, "step": 3650 }, { "grad_norm": 0.6840285658836365, "learning_rate": 7.318e-05, "loss": 0.0682, "step": 3660 }, { "grad_norm": 0.9827133417129517, "learning_rate": 7.338e-05, "loss": 0.0686, "step": 3670 }, { "grad_norm": 0.8516344428062439, "learning_rate": 7.358000000000001e-05, "loss": 0.0815, "step": 3680 }, { "grad_norm": 0.8703101277351379, "learning_rate": 7.378e-05, "loss": 0.0892, "step": 3690 }, { "grad_norm": 0.9205076098442078, "learning_rate": 7.398e-05, "loss": 0.0855, "step": 3700 }, { "grad_norm": 0.8788275718688965, "learning_rate": 7.418000000000001e-05, "loss": 0.0985, "step": 3710 }, { "grad_norm": 0.7111381888389587, "learning_rate": 7.438e-05, "loss": 0.0757, "step": 3720 }, { "grad_norm": 0.9943393468856812, "learning_rate": 7.458000000000001e-05, "loss": 0.0678, "step": 3730 }, { "grad_norm": 1.689884901046753, "learning_rate": 7.478e-05, "loss": 0.0809, "step": 3740 }, { "grad_norm": 0.8596060276031494, "learning_rate": 7.498e-05, "loss": 0.0839, "step": 3750 }, { "grad_norm": 0.692905068397522, "learning_rate": 7.518000000000001e-05, "loss": 0.0903, "step": 3760 }, { "grad_norm": 0.5886113047599792, "learning_rate": 7.538e-05, "loss": 0.0817, "step": 3770 }, { "grad_norm": 0.6571245789527893, "learning_rate": 7.558e-05, "loss": 0.0706, "step": 3780 }, { "grad_norm": 0.6555784940719604, "learning_rate": 7.578000000000001e-05, "loss": 0.0787, "step": 3790 }, { "grad_norm": 0.723996639251709, "learning_rate": 7.598e-05, "loss": 0.0771, "step": 3800 }, { "grad_norm": 0.6848304271697998, "learning_rate": 7.618e-05, "loss": 0.0753, "step": 3810 }, { "grad_norm": 1.0222318172454834, "learning_rate": 7.638000000000001e-05, "loss": 0.0904, "step": 3820 }, { "grad_norm": 0.6441411972045898, "learning_rate": 7.658e-05, "loss": 0.0865, "step": 3830 }, { "grad_norm": 0.5627588629722595, "learning_rate": 7.678000000000001e-05, "loss": 0.0721, "step": 3840 }, { "grad_norm": 0.5323032736778259, "learning_rate": 7.698000000000001e-05, "loss": 0.0717, "step": 3850 }, { "grad_norm": 0.8352326154708862, "learning_rate": 7.718e-05, "loss": 0.0801, "step": 3860 }, { "grad_norm": 0.8071228265762329, "learning_rate": 7.738000000000001e-05, "loss": 0.0708, "step": 3870 }, { "grad_norm": 0.8767066597938538, "learning_rate": 7.758000000000001e-05, "loss": 0.0807, "step": 3880 }, { "grad_norm": 0.6243783235549927, "learning_rate": 7.778e-05, "loss": 0.0821, "step": 3890 }, { "grad_norm": 1.0665560960769653, "learning_rate": 7.798000000000001e-05, "loss": 0.0771, "step": 3900 }, { "grad_norm": 0.8080918192863464, "learning_rate": 7.818000000000001e-05, "loss": 0.0796, "step": 3910 }, { "grad_norm": 1.0600082874298096, "learning_rate": 7.838e-05, "loss": 0.0812, "step": 3920 }, { "grad_norm": 0.5748053193092346, "learning_rate": 7.858000000000001e-05, "loss": 0.083, "step": 3930 }, { "grad_norm": 0.766463577747345, "learning_rate": 7.878e-05, "loss": 0.0837, "step": 3940 }, { "grad_norm": 1.0252717733383179, "learning_rate": 7.897999999999999e-05, "loss": 0.101, "step": 3950 }, { "grad_norm": 0.8172783851623535, "learning_rate": 7.918e-05, "loss": 0.0996, "step": 3960 }, { "grad_norm": 0.6269286870956421, "learning_rate": 7.938e-05, "loss": 0.0968, "step": 3970 }, { "grad_norm": 0.7549141049385071, "learning_rate": 7.958e-05, "loss": 0.0775, "step": 3980 }, { "grad_norm": 1.0100730657577515, "learning_rate": 7.978e-05, "loss": 0.0832, "step": 3990 }, { "grad_norm": 0.7964673042297363, "learning_rate": 7.998e-05, "loss": 0.0875, "step": 4000 }, { "grad_norm": 0.8003693222999573, "learning_rate": 8.018e-05, "loss": 0.084, "step": 4010 }, { "grad_norm": 0.6626913547515869, "learning_rate": 8.038e-05, "loss": 0.083, "step": 4020 }, { "grad_norm": 0.7479267716407776, "learning_rate": 8.058e-05, "loss": 0.0846, "step": 4030 }, { "grad_norm": 0.8496220707893372, "learning_rate": 8.078e-05, "loss": 0.0851, "step": 4040 }, { "grad_norm": 0.5884184241294861, "learning_rate": 8.098e-05, "loss": 0.0842, "step": 4050 }, { "grad_norm": 0.7999506592750549, "learning_rate": 8.118e-05, "loss": 0.0879, "step": 4060 }, { "grad_norm": 0.7684029936790466, "learning_rate": 8.138e-05, "loss": 0.0853, "step": 4070 }, { "grad_norm": 0.6202629208564758, "learning_rate": 8.158e-05, "loss": 0.0796, "step": 4080 }, { "grad_norm": 0.7234943509101868, "learning_rate": 8.178e-05, "loss": 0.0746, "step": 4090 }, { "grad_norm": 0.618684709072113, "learning_rate": 8.198e-05, "loss": 0.1004, "step": 4100 }, { "grad_norm": 0.8074678778648376, "learning_rate": 8.218e-05, "loss": 0.0823, "step": 4110 }, { "grad_norm": 0.6307827830314636, "learning_rate": 8.238000000000001e-05, "loss": 0.0796, "step": 4120 }, { "grad_norm": 0.734110951423645, "learning_rate": 8.258e-05, "loss": 0.0855, "step": 4130 }, { "grad_norm": 0.7869684100151062, "learning_rate": 8.278e-05, "loss": 0.0862, "step": 4140 }, { "grad_norm": 0.7381423115730286, "learning_rate": 8.298000000000001e-05, "loss": 0.0876, "step": 4150 }, { "grad_norm": 0.5736311674118042, "learning_rate": 8.318e-05, "loss": 0.0794, "step": 4160 }, { "grad_norm": 0.7018775343894958, "learning_rate": 8.338e-05, "loss": 0.0871, "step": 4170 }, { "grad_norm": 0.8223671913146973, "learning_rate": 8.358e-05, "loss": 0.087, "step": 4180 }, { "grad_norm": 0.6518516540527344, "learning_rate": 8.378e-05, "loss": 0.0784, "step": 4190 }, { "grad_norm": 0.5366777181625366, "learning_rate": 8.398e-05, "loss": 0.0707, "step": 4200 }, { "grad_norm": 0.5725388526916504, "learning_rate": 8.418e-05, "loss": 0.0625, "step": 4210 }, { "grad_norm": 0.6559984087944031, "learning_rate": 8.438e-05, "loss": 0.0677, "step": 4220 }, { "grad_norm": 0.7473193407058716, "learning_rate": 8.458e-05, "loss": 0.0709, "step": 4230 }, { "grad_norm": 0.5993654727935791, "learning_rate": 8.478e-05, "loss": 0.0822, "step": 4240 }, { "grad_norm": 0.7250844240188599, "learning_rate": 8.498e-05, "loss": 0.0817, "step": 4250 }, { "grad_norm": 0.822116494178772, "learning_rate": 8.518000000000001e-05, "loss": 0.0771, "step": 4260 }, { "grad_norm": 0.6014385223388672, "learning_rate": 8.538e-05, "loss": 0.0671, "step": 4270 }, { "grad_norm": 0.6654970049858093, "learning_rate": 8.558e-05, "loss": 0.067, "step": 4280 }, { "grad_norm": 0.5610861778259277, "learning_rate": 8.578000000000001e-05, "loss": 0.069, "step": 4290 }, { "grad_norm": 0.6707329750061035, "learning_rate": 8.598e-05, "loss": 0.0671, "step": 4300 }, { "grad_norm": 0.6166923642158508, "learning_rate": 8.618e-05, "loss": 0.0621, "step": 4310 }, { "grad_norm": 0.6901628375053406, "learning_rate": 8.638000000000001e-05, "loss": 0.0698, "step": 4320 }, { "grad_norm": 0.6217272281646729, "learning_rate": 8.658e-05, "loss": 0.0666, "step": 4330 }, { "grad_norm": 0.6511088013648987, "learning_rate": 8.678e-05, "loss": 0.0843, "step": 4340 }, { "grad_norm": 0.7196616530418396, "learning_rate": 8.698000000000001e-05, "loss": 0.0747, "step": 4350 }, { "grad_norm": 0.6930388808250427, "learning_rate": 8.718e-05, "loss": 0.0756, "step": 4360 }, { "grad_norm": 0.7001752853393555, "learning_rate": 8.738000000000001e-05, "loss": 0.0778, "step": 4370 }, { "grad_norm": 0.727545440196991, "learning_rate": 8.758000000000001e-05, "loss": 0.0771, "step": 4380 }, { "grad_norm": 0.5902413129806519, "learning_rate": 8.778e-05, "loss": 0.0802, "step": 4390 }, { "grad_norm": 0.8177023530006409, "learning_rate": 8.798000000000001e-05, "loss": 0.085, "step": 4400 }, { "grad_norm": 0.7973567247390747, "learning_rate": 8.818000000000001e-05, "loss": 0.0775, "step": 4410 }, { "grad_norm": 0.6509308218955994, "learning_rate": 8.838e-05, "loss": 0.0769, "step": 4420 }, { "grad_norm": 0.6320347785949707, "learning_rate": 8.858000000000001e-05, "loss": 0.0681, "step": 4430 }, { "grad_norm": 0.6690737009048462, "learning_rate": 8.878000000000001e-05, "loss": 0.0856, "step": 4440 }, { "grad_norm": 0.8330267667770386, "learning_rate": 8.898e-05, "loss": 0.0679, "step": 4450 }, { "grad_norm": 0.5187621116638184, "learning_rate": 8.918000000000001e-05, "loss": 0.0731, "step": 4460 }, { "grad_norm": 0.6841365694999695, "learning_rate": 8.938e-05, "loss": 0.0685, "step": 4470 }, { "grad_norm": 0.6965979933738708, "learning_rate": 8.958e-05, "loss": 0.0737, "step": 4480 }, { "grad_norm": 0.5879324674606323, "learning_rate": 8.978000000000001e-05, "loss": 0.084, "step": 4490 }, { "grad_norm": 1.0918134450912476, "learning_rate": 8.998e-05, "loss": 0.0845, "step": 4500 }, { "grad_norm": 0.8728201985359192, "learning_rate": 9.018000000000001e-05, "loss": 0.0883, "step": 4510 }, { "grad_norm": 0.5646225214004517, "learning_rate": 9.038000000000001e-05, "loss": 0.0713, "step": 4520 }, { "grad_norm": 0.6001912951469421, "learning_rate": 9.058e-05, "loss": 0.072, "step": 4530 }, { "grad_norm": 0.6944179534912109, "learning_rate": 9.078000000000001e-05, "loss": 0.0691, "step": 4540 }, { "grad_norm": 0.6753365993499756, "learning_rate": 9.098000000000001e-05, "loss": 0.0707, "step": 4550 }, { "grad_norm": 0.7272933721542358, "learning_rate": 9.118e-05, "loss": 0.0681, "step": 4560 }, { "grad_norm": 0.9513645768165588, "learning_rate": 9.138e-05, "loss": 0.0823, "step": 4570 }, { "grad_norm": 0.8516632914543152, "learning_rate": 9.158e-05, "loss": 0.0916, "step": 4580 }, { "grad_norm": 0.6485275626182556, "learning_rate": 9.178e-05, "loss": 0.0747, "step": 4590 }, { "grad_norm": 0.5775065422058105, "learning_rate": 9.198e-05, "loss": 0.0829, "step": 4600 }, { "grad_norm": 0.6644838452339172, "learning_rate": 9.218e-05, "loss": 0.0756, "step": 4610 }, { "grad_norm": 0.5204406976699829, "learning_rate": 9.238e-05, "loss": 0.0719, "step": 4620 }, { "grad_norm": 0.7422964572906494, "learning_rate": 9.258e-05, "loss": 0.0693, "step": 4630 }, { "grad_norm": 0.5937347412109375, "learning_rate": 9.278e-05, "loss": 0.0724, "step": 4640 }, { "grad_norm": 0.7790461182594299, "learning_rate": 9.298e-05, "loss": 0.0703, "step": 4650 }, { "grad_norm": 4.291955947875977, "learning_rate": 9.318e-05, "loss": 0.0856, "step": 4660 }, { "grad_norm": 0.8092467784881592, "learning_rate": 9.338e-05, "loss": 0.0757, "step": 4670 }, { "grad_norm": 0.49525904655456543, "learning_rate": 9.358e-05, "loss": 0.0693, "step": 4680 }, { "grad_norm": 0.556628406047821, "learning_rate": 9.378e-05, "loss": 0.0786, "step": 4690 }, { "grad_norm": 0.5875357389450073, "learning_rate": 9.398e-05, "loss": 0.0798, "step": 4700 }, { "grad_norm": 0.5389693975448608, "learning_rate": 9.418e-05, "loss": 0.0763, "step": 4710 }, { "grad_norm": 0.5631381869316101, "learning_rate": 9.438e-05, "loss": 0.0756, "step": 4720 }, { "grad_norm": 0.654769778251648, "learning_rate": 9.458e-05, "loss": 0.0763, "step": 4730 }, { "grad_norm": 0.7793031334877014, "learning_rate": 9.478e-05, "loss": 0.0667, "step": 4740 }, { "grad_norm": 0.5327534079551697, "learning_rate": 9.498e-05, "loss": 0.0676, "step": 4750 }, { "grad_norm": 0.660384476184845, "learning_rate": 9.518000000000001e-05, "loss": 0.0773, "step": 4760 }, { "grad_norm": 0.7066521644592285, "learning_rate": 9.538e-05, "loss": 0.0693, "step": 4770 }, { "grad_norm": 0.5442914962768555, "learning_rate": 9.558e-05, "loss": 0.0737, "step": 4780 }, { "grad_norm": 0.5376337170600891, "learning_rate": 9.578000000000001e-05, "loss": 0.0773, "step": 4790 }, { "grad_norm": 0.6000149846076965, "learning_rate": 9.598e-05, "loss": 0.071, "step": 4800 }, { "grad_norm": 0.6042788028717041, "learning_rate": 9.618e-05, "loss": 0.0733, "step": 4810 }, { "grad_norm": 0.7254698276519775, "learning_rate": 9.638000000000001e-05, "loss": 0.0676, "step": 4820 }, { "grad_norm": 0.726879358291626, "learning_rate": 9.658e-05, "loss": 0.0745, "step": 4830 }, { "grad_norm": 0.6029928922653198, "learning_rate": 9.678e-05, "loss": 0.0895, "step": 4840 }, { "grad_norm": 0.5960591435432434, "learning_rate": 9.698000000000001e-05, "loss": 0.0802, "step": 4850 }, { "grad_norm": 0.5669956207275391, "learning_rate": 9.718e-05, "loss": 0.0669, "step": 4860 }, { "grad_norm": 0.4990383982658386, "learning_rate": 9.738e-05, "loss": 0.0652, "step": 4870 }, { "grad_norm": 0.5145519971847534, "learning_rate": 9.758000000000001e-05, "loss": 0.0718, "step": 4880 }, { "grad_norm": 0.6727977991104126, "learning_rate": 9.778e-05, "loss": 0.0685, "step": 4890 }, { "grad_norm": 0.6597524285316467, "learning_rate": 9.798000000000001e-05, "loss": 0.0692, "step": 4900 }, { "grad_norm": 0.5371189117431641, "learning_rate": 9.818000000000001e-05, "loss": 0.0762, "step": 4910 }, { "grad_norm": 0.8699682950973511, "learning_rate": 9.838e-05, "loss": 0.0689, "step": 4920 }, { "grad_norm": 0.7110276818275452, "learning_rate": 9.858000000000001e-05, "loss": 0.0797, "step": 4930 }, { "grad_norm": 0.6560996770858765, "learning_rate": 9.878e-05, "loss": 0.0749, "step": 4940 }, { "grad_norm": 2.6286098957061768, "learning_rate": 9.898e-05, "loss": 0.09, "step": 4950 }, { "grad_norm": 0.7119970917701721, "learning_rate": 9.918000000000001e-05, "loss": 0.0835, "step": 4960 }, { "grad_norm": 0.7166153788566589, "learning_rate": 9.938e-05, "loss": 0.0887, "step": 4970 }, { "grad_norm": 0.5162918567657471, "learning_rate": 9.958e-05, "loss": 0.0732, "step": 4980 }, { "grad_norm": 0.6755915880203247, "learning_rate": 9.978000000000001e-05, "loss": 0.0682, "step": 4990 }, { "grad_norm": 0.5119858980178833, "learning_rate": 9.998e-05, "loss": 0.0762, "step": 5000 }, { "grad_norm": 0.609869658946991, "learning_rate": 9.999999778549045e-05, "loss": 0.0761, "step": 5010 }, { "grad_norm": 0.5722142457962036, "learning_rate": 9.999999013039593e-05, "loss": 0.0687, "step": 5020 }, { "grad_norm": 0.5477919578552246, "learning_rate": 9.999997700737766e-05, "loss": 0.0644, "step": 5030 }, { "grad_norm": 0.7401325702667236, "learning_rate": 9.999995841643709e-05, "loss": 0.0684, "step": 5040 }, { "grad_norm": 0.6018873453140259, "learning_rate": 9.999993435757623e-05, "loss": 0.0651, "step": 5050 }, { "grad_norm": 0.5853024125099182, "learning_rate": 9.999990483079773e-05, "loss": 0.0729, "step": 5060 }, { "grad_norm": 0.6063034534454346, "learning_rate": 9.999986983610481e-05, "loss": 0.0717, "step": 5070 }, { "grad_norm": 0.6106910109519958, "learning_rate": 9.99998293735013e-05, "loss": 0.0777, "step": 5080 }, { "grad_norm": 0.5333430171012878, "learning_rate": 9.999978344299161e-05, "loss": 0.0753, "step": 5090 }, { "grad_norm": 0.5054991245269775, "learning_rate": 9.99997320445808e-05, "loss": 0.0761, "step": 5100 }, { "grad_norm": 0.5279659032821655, "learning_rate": 9.999967517827444e-05, "loss": 0.0769, "step": 5110 }, { "grad_norm": 0.5388793349266052, "learning_rate": 9.999961284407879e-05, "loss": 0.0837, "step": 5120 }, { "grad_norm": 0.47587189078330994, "learning_rate": 9.999954504200067e-05, "loss": 0.0665, "step": 5130 }, { "grad_norm": 0.6007452011108398, "learning_rate": 9.999947177204744e-05, "loss": 0.0641, "step": 5140 }, { "grad_norm": 0.614435613155365, "learning_rate": 9.999939303422718e-05, "loss": 0.0678, "step": 5150 }, { "grad_norm": 0.5836154818534851, "learning_rate": 9.999930882854847e-05, "loss": 0.073, "step": 5160 }, { "grad_norm": 0.6675926446914673, "learning_rate": 9.999921915502051e-05, "loss": 0.0748, "step": 5170 }, { "grad_norm": 0.4953056573867798, "learning_rate": 9.99991240136531e-05, "loss": 0.0648, "step": 5180 }, { "grad_norm": 0.538882315158844, "learning_rate": 9.999902340445668e-05, "loss": 0.0712, "step": 5190 }, { "grad_norm": 0.8540242314338684, "learning_rate": 9.999891732744224e-05, "loss": 0.0882, "step": 5200 }, { "grad_norm": 0.6914522647857666, "learning_rate": 9.999880578262135e-05, "loss": 0.0732, "step": 5210 }, { "grad_norm": 0.71214359998703, "learning_rate": 9.999868877000624e-05, "loss": 0.0663, "step": 5220 }, { "grad_norm": 0.6181578040122986, "learning_rate": 9.99985662896097e-05, "loss": 0.0713, "step": 5230 }, { "grad_norm": 0.500919759273529, "learning_rate": 9.999843834144513e-05, "loss": 0.0668, "step": 5240 }, { "grad_norm": 0.7898532748222351, "learning_rate": 9.99983049255265e-05, "loss": 0.06, "step": 5250 }, { "grad_norm": 0.5859411954879761, "learning_rate": 9.999816604186843e-05, "loss": 0.0687, "step": 5260 }, { "grad_norm": 0.4204474687576294, "learning_rate": 9.999802169048609e-05, "loss": 0.0746, "step": 5270 }, { "grad_norm": 0.5493097305297852, "learning_rate": 9.999787187139527e-05, "loss": 0.0643, "step": 5280 }, { "grad_norm": 0.5703192353248596, "learning_rate": 9.999771658461234e-05, "loss": 0.0681, "step": 5290 }, { "grad_norm": 0.4632253050804138, "learning_rate": 9.999755583015431e-05, "loss": 0.062, "step": 5300 }, { "grad_norm": 0.4347795248031616, "learning_rate": 9.999738960803874e-05, "loss": 0.0671, "step": 5310 }, { "grad_norm": 0.9041308760643005, "learning_rate": 9.99972179182838e-05, "loss": 0.0681, "step": 5320 }, { "grad_norm": 0.6308367848396301, "learning_rate": 9.99970407609083e-05, "loss": 0.0611, "step": 5330 }, { "grad_norm": 0.6336970329284668, "learning_rate": 9.999685813593159e-05, "loss": 0.0755, "step": 5340 }, { "grad_norm": 0.6455582976341248, "learning_rate": 9.999667004337362e-05, "loss": 0.0699, "step": 5350 }, { "grad_norm": 0.6340748071670532, "learning_rate": 9.9996476483255e-05, "loss": 0.0732, "step": 5360 }, { "grad_norm": 0.558720588684082, "learning_rate": 9.999627745559688e-05, "loss": 0.0741, "step": 5370 }, { "grad_norm": 0.4758220613002777, "learning_rate": 9.999607296042101e-05, "loss": 0.0673, "step": 5380 }, { "grad_norm": 0.7028363347053528, "learning_rate": 9.99958629977498e-05, "loss": 0.0738, "step": 5390 }, { "grad_norm": 0.4454802870750427, "learning_rate": 9.999564756760615e-05, "loss": 0.0653, "step": 5400 }, { "grad_norm": 0.621001124382019, "learning_rate": 9.999542667001366e-05, "loss": 0.0745, "step": 5410 }, { "grad_norm": 0.5095284581184387, "learning_rate": 9.999520030499647e-05, "loss": 0.062, "step": 5420 }, { "grad_norm": 0.5917644500732422, "learning_rate": 9.999496847257936e-05, "loss": 0.0643, "step": 5430 }, { "grad_norm": 0.6522612571716309, "learning_rate": 9.999473117278764e-05, "loss": 0.0662, "step": 5440 }, { "grad_norm": 0.5897451639175415, "learning_rate": 9.999448840564731e-05, "loss": 0.0743, "step": 5450 }, { "grad_norm": 0.49095258116722107, "learning_rate": 9.999424017118488e-05, "loss": 0.0687, "step": 5460 }, { "grad_norm": 0.5559768080711365, "learning_rate": 9.999398646942751e-05, "loss": 0.0675, "step": 5470 }, { "grad_norm": 0.4750228822231293, "learning_rate": 9.999372730040296e-05, "loss": 0.0628, "step": 5480 }, { "grad_norm": 0.6646190285682678, "learning_rate": 9.999346266413953e-05, "loss": 0.0611, "step": 5490 }, { "grad_norm": 0.47952428460121155, "learning_rate": 9.99931925606662e-05, "loss": 0.0634, "step": 5500 }, { "grad_norm": 0.44590631127357483, "learning_rate": 9.99929169900125e-05, "loss": 0.0651, "step": 5510 }, { "grad_norm": 0.3889639973640442, "learning_rate": 9.999263595220855e-05, "loss": 0.0618, "step": 5520 }, { "grad_norm": 0.593500018119812, "learning_rate": 9.99923494472851e-05, "loss": 0.0678, "step": 5530 }, { "grad_norm": 0.4622524082660675, "learning_rate": 9.999205747527348e-05, "loss": 0.0586, "step": 5540 }, { "grad_norm": 0.5075958967208862, "learning_rate": 9.999176003620561e-05, "loss": 0.0566, "step": 5550 }, { "grad_norm": 0.528433620929718, "learning_rate": 9.999145713011405e-05, "loss": 0.0572, "step": 5560 }, { "grad_norm": 0.7422134280204773, "learning_rate": 9.999114875703186e-05, "loss": 0.0626, "step": 5570 }, { "grad_norm": 1.485779881477356, "learning_rate": 9.999083491699281e-05, "loss": 0.0663, "step": 5580 }, { "grad_norm": 0.588018000125885, "learning_rate": 9.999051561003123e-05, "loss": 0.0625, "step": 5590 }, { "grad_norm": 0.5562824606895447, "learning_rate": 9.999019083618202e-05, "loss": 0.0712, "step": 5600 }, { "grad_norm": 0.584502100944519, "learning_rate": 9.99898605954807e-05, "loss": 0.0671, "step": 5610 }, { "grad_norm": 0.4320460557937622, "learning_rate": 9.998952488796338e-05, "loss": 0.0665, "step": 5620 }, { "grad_norm": 0.8627905249595642, "learning_rate": 9.998918371366676e-05, "loss": 0.0673, "step": 5630 }, { "grad_norm": 0.43105629086494446, "learning_rate": 9.99888370726282e-05, "loss": 0.0688, "step": 5640 }, { "grad_norm": 0.6221110820770264, "learning_rate": 9.998848496488556e-05, "loss": 0.0604, "step": 5650 }, { "grad_norm": 0.6551891565322876, "learning_rate": 9.998812739047736e-05, "loss": 0.0681, "step": 5660 }, { "grad_norm": 0.5090816020965576, "learning_rate": 9.99877643494427e-05, "loss": 0.0645, "step": 5670 }, { "grad_norm": 0.5320875644683838, "learning_rate": 9.998739584182128e-05, "loss": 0.0605, "step": 5680 }, { "grad_norm": 0.7255314588546753, "learning_rate": 9.998702186765342e-05, "loss": 0.0679, "step": 5690 }, { "grad_norm": 0.8149005770683289, "learning_rate": 9.998664242698e-05, "loss": 0.0773, "step": 5700 }, { "grad_norm": 0.6248178482055664, "learning_rate": 9.998625751984251e-05, "loss": 0.0719, "step": 5710 }, { "grad_norm": 0.6295396685600281, "learning_rate": 9.998586714628307e-05, "loss": 0.0745, "step": 5720 }, { "grad_norm": 0.6936106085777283, "learning_rate": 9.998547130634432e-05, "loss": 0.0767, "step": 5730 }, { "grad_norm": 0.5515219569206238, "learning_rate": 9.99850700000696e-05, "loss": 0.0702, "step": 5740 }, { "grad_norm": 0.5036241412162781, "learning_rate": 9.998466322750278e-05, "loss": 0.0666, "step": 5750 }, { "grad_norm": 0.5595634579658508, "learning_rate": 9.998425098868834e-05, "loss": 0.0726, "step": 5760 }, { "grad_norm": 0.5258265733718872, "learning_rate": 9.998383328367136e-05, "loss": 0.0695, "step": 5770 }, { "grad_norm": 0.5448219776153564, "learning_rate": 9.99834101124975e-05, "loss": 0.0662, "step": 5780 }, { "grad_norm": 0.44855397939682007, "learning_rate": 9.998298147521309e-05, "loss": 0.0685, "step": 5790 }, { "grad_norm": 0.43952372670173645, "learning_rate": 9.998254737186496e-05, "loss": 0.062, "step": 5800 }, { "grad_norm": 0.5768082737922668, "learning_rate": 9.99821078025006e-05, "loss": 0.062, "step": 5810 }, { "grad_norm": 0.48369914293289185, "learning_rate": 9.998166276716807e-05, "loss": 0.0592, "step": 5820 }, { "grad_norm": 0.6481833457946777, "learning_rate": 9.998121226591606e-05, "loss": 0.067, "step": 5830 }, { "grad_norm": 0.5869381427764893, "learning_rate": 9.998075629879382e-05, "loss": 0.0591, "step": 5840 }, { "grad_norm": 0.5763795971870422, "learning_rate": 9.99802948658512e-05, "loss": 0.0724, "step": 5850 }, { "grad_norm": 0.6296831369400024, "learning_rate": 9.99798279671387e-05, "loss": 0.0662, "step": 5860 }, { "grad_norm": 0.5029813647270203, "learning_rate": 9.997935560270734e-05, "loss": 0.0597, "step": 5870 }, { "grad_norm": 0.621174693107605, "learning_rate": 9.997887777260879e-05, "loss": 0.0726, "step": 5880 }, { "grad_norm": 0.5138165950775146, "learning_rate": 9.997839447689532e-05, "loss": 0.0561, "step": 5890 }, { "grad_norm": 0.4761827886104584, "learning_rate": 9.997790571561978e-05, "loss": 0.0634, "step": 5900 }, { "grad_norm": 0.5086909532546997, "learning_rate": 9.99774114888356e-05, "loss": 0.0735, "step": 5910 }, { "grad_norm": 0.4430907666683197, "learning_rate": 9.997691179659684e-05, "loss": 0.0588, "step": 5920 }, { "grad_norm": 0.7795563340187073, "learning_rate": 9.997640663895815e-05, "loss": 0.0639, "step": 5930 }, { "grad_norm": 0.6077259182929993, "learning_rate": 9.997589601597477e-05, "loss": 0.0598, "step": 5940 }, { "grad_norm": 0.46236473321914673, "learning_rate": 9.997537992770252e-05, "loss": 0.0585, "step": 5950 }, { "grad_norm": 0.47923436760902405, "learning_rate": 9.997485837419788e-05, "loss": 0.0582, "step": 5960 }, { "grad_norm": 0.4316174387931824, "learning_rate": 9.997433135551786e-05, "loss": 0.0629, "step": 5970 }, { "grad_norm": 0.41290563344955444, "learning_rate": 9.997379887172009e-05, "loss": 0.0615, "step": 5980 }, { "grad_norm": 0.46471500396728516, "learning_rate": 9.997326092286281e-05, "loss": 0.0695, "step": 5990 }, { "grad_norm": 0.6234973669052124, "learning_rate": 9.997271750900486e-05, "loss": 0.0583, "step": 6000 }, { "grad_norm": 0.5594924092292786, "learning_rate": 9.997216863020565e-05, "loss": 0.0641, "step": 6010 }, { "grad_norm": 0.4894625246524811, "learning_rate": 9.99716142865252e-05, "loss": 0.0655, "step": 6020 }, { "grad_norm": 0.496518075466156, "learning_rate": 9.997105447802415e-05, "loss": 0.0549, "step": 6030 }, { "grad_norm": 0.7337004542350769, "learning_rate": 9.997048920476373e-05, "loss": 0.0677, "step": 6040 }, { "grad_norm": 0.5290573835372925, "learning_rate": 9.996991846680572e-05, "loss": 0.0506, "step": 6050 }, { "grad_norm": 1.8044705390930176, "learning_rate": 9.996934226421257e-05, "loss": 0.0736, "step": 6060 }, { "grad_norm": 0.561804473400116, "learning_rate": 9.996876059704726e-05, "loss": 0.0643, "step": 6070 }, { "grad_norm": 0.7560541033744812, "learning_rate": 9.996817346537343e-05, "loss": 0.0699, "step": 6080 }, { "grad_norm": 0.5173971056938171, "learning_rate": 9.996758086925526e-05, "loss": 0.0739, "step": 6090 }, { "grad_norm": 0.6652621626853943, "learning_rate": 9.996698280875759e-05, "loss": 0.0588, "step": 6100 }, { "grad_norm": 0.4801855981349945, "learning_rate": 9.99663792839458e-05, "loss": 0.0597, "step": 6110 }, { "grad_norm": 0.444047212600708, "learning_rate": 9.99657702948859e-05, "loss": 0.0644, "step": 6120 }, { "grad_norm": 0.6361960172653198, "learning_rate": 9.996515584164448e-05, "loss": 0.0575, "step": 6130 }, { "grad_norm": 0.797155499458313, "learning_rate": 9.996453592428873e-05, "loss": 0.058, "step": 6140 }, { "grad_norm": 0.8113923072814941, "learning_rate": 9.996391054288646e-05, "loss": 0.0677, "step": 6150 }, { "grad_norm": 0.506161630153656, "learning_rate": 9.996327969750605e-05, "loss": 0.0618, "step": 6160 }, { "grad_norm": 0.5890957713127136, "learning_rate": 9.996264338821649e-05, "loss": 0.0568, "step": 6170 }, { "grad_norm": 0.5615513920783997, "learning_rate": 9.996200161508735e-05, "loss": 0.0642, "step": 6180 }, { "grad_norm": 0.5444278717041016, "learning_rate": 9.996135437818885e-05, "loss": 0.0618, "step": 6190 }, { "grad_norm": 0.4824307858943939, "learning_rate": 9.996070167759175e-05, "loss": 0.0574, "step": 6200 }, { "grad_norm": 0.5568683743476868, "learning_rate": 9.996004351336743e-05, "loss": 0.0583, "step": 6210 }, { "grad_norm": 0.5001839399337769, "learning_rate": 9.995937988558785e-05, "loss": 0.0547, "step": 6220 }, { "grad_norm": 0.4671173095703125, "learning_rate": 9.995871079432561e-05, "loss": 0.0669, "step": 6230 }, { "grad_norm": 0.7019325494766235, "learning_rate": 9.995803623965389e-05, "loss": 0.0629, "step": 6240 }, { "grad_norm": 0.6616390347480774, "learning_rate": 9.995735622164641e-05, "loss": 0.0723, "step": 6250 }, { "grad_norm": 0.4620645046234131, "learning_rate": 9.995667074037758e-05, "loss": 0.0609, "step": 6260 }, { "grad_norm": 0.5550916194915771, "learning_rate": 9.995597979592232e-05, "loss": 0.062, "step": 6270 }, { "grad_norm": 0.6498542428016663, "learning_rate": 9.995528338835625e-05, "loss": 0.0583, "step": 6280 }, { "grad_norm": 0.579028844833374, "learning_rate": 9.995458151775547e-05, "loss": 0.059, "step": 6290 }, { "grad_norm": 0.6354548335075378, "learning_rate": 9.995387418419677e-05, "loss": 0.0555, "step": 6300 }, { "grad_norm": 0.5341694951057434, "learning_rate": 9.99531613877575e-05, "loss": 0.0556, "step": 6310 }, { "grad_norm": 0.4556196331977844, "learning_rate": 9.995244312851559e-05, "loss": 0.0575, "step": 6320 }, { "grad_norm": 0.4890003204345703, "learning_rate": 9.995171940654961e-05, "loss": 0.0608, "step": 6330 }, { "grad_norm": 0.4820818305015564, "learning_rate": 9.995099022193871e-05, "loss": 0.0531, "step": 6340 }, { "grad_norm": 0.3736449182033539, "learning_rate": 9.995025557476261e-05, "loss": 0.049, "step": 6350 }, { "grad_norm": 0.4243585169315338, "learning_rate": 9.994951546510165e-05, "loss": 0.0537, "step": 6360 }, { "grad_norm": 0.5355581045150757, "learning_rate": 9.994876989303679e-05, "loss": 0.059, "step": 6370 }, { "grad_norm": 0.4827471971511841, "learning_rate": 9.994801885864955e-05, "loss": 0.0597, "step": 6380 }, { "grad_norm": 0.5569455027580261, "learning_rate": 9.994726236202205e-05, "loss": 0.0565, "step": 6390 }, { "grad_norm": 0.6064746379852295, "learning_rate": 9.994650040323704e-05, "loss": 0.0506, "step": 6400 }, { "grad_norm": 0.5007423162460327, "learning_rate": 9.994573298237784e-05, "loss": 0.0624, "step": 6410 }, { "grad_norm": 0.6105954051017761, "learning_rate": 9.994496009952837e-05, "loss": 0.0612, "step": 6420 }, { "grad_norm": 0.4334472119808197, "learning_rate": 9.994418175477316e-05, "loss": 0.0586, "step": 6430 }, { "grad_norm": 0.5655759572982788, "learning_rate": 9.994339794819733e-05, "loss": 0.0634, "step": 6440 }, { "grad_norm": 0.5539496541023254, "learning_rate": 9.994260867988658e-05, "loss": 0.0575, "step": 6450 }, { "grad_norm": 0.7006757855415344, "learning_rate": 9.994181394992723e-05, "loss": 0.0584, "step": 6460 }, { "grad_norm": 0.5843648314476013, "learning_rate": 9.994101375840618e-05, "loss": 0.0654, "step": 6470 }, { "grad_norm": 0.5381225347518921, "learning_rate": 9.994020810541098e-05, "loss": 0.0585, "step": 6480 }, { "grad_norm": 0.5914531350135803, "learning_rate": 9.99393969910297e-05, "loss": 0.0638, "step": 6490 }, { "grad_norm": 0.6728056073188782, "learning_rate": 9.993858041535104e-05, "loss": 0.0616, "step": 6500 }, { "grad_norm": 0.608848512172699, "learning_rate": 9.99377583784643e-05, "loss": 0.059, "step": 6510 }, { "grad_norm": 0.6038638353347778, "learning_rate": 9.993693088045939e-05, "loss": 0.0695, "step": 6520 }, { "grad_norm": 0.520363986492157, "learning_rate": 9.99360979214268e-05, "loss": 0.0652, "step": 6530 }, { "grad_norm": 0.5160427689552307, "learning_rate": 9.99352595014576e-05, "loss": 0.059, "step": 6540 }, { "grad_norm": 0.4902753531932831, "learning_rate": 9.993441562064354e-05, "loss": 0.0562, "step": 6550 }, { "grad_norm": 0.5716758966445923, "learning_rate": 9.993356627907685e-05, "loss": 0.0671, "step": 6560 }, { "grad_norm": 0.4779585301876068, "learning_rate": 9.99327114768504e-05, "loss": 0.0699, "step": 6570 }, { "grad_norm": 0.6001152992248535, "learning_rate": 9.99318512140577e-05, "loss": 0.076, "step": 6580 }, { "grad_norm": 0.5404191017150879, "learning_rate": 9.993098549079284e-05, "loss": 0.06, "step": 6590 }, { "grad_norm": 0.4177423417568207, "learning_rate": 9.993011430715047e-05, "loss": 0.0617, "step": 6600 }, { "grad_norm": 0.4390844702720642, "learning_rate": 9.992923766322586e-05, "loss": 0.0582, "step": 6610 }, { "grad_norm": 0.46668753027915955, "learning_rate": 9.99283555591149e-05, "loss": 0.0639, "step": 6620 }, { "grad_norm": 0.6341900825500488, "learning_rate": 9.992746799491404e-05, "loss": 0.0666, "step": 6630 }, { "grad_norm": 0.42222270369529724, "learning_rate": 9.992657497072033e-05, "loss": 0.0671, "step": 6640 }, { "grad_norm": 0.4647522568702698, "learning_rate": 9.992567648663147e-05, "loss": 0.057, "step": 6650 }, { "grad_norm": 0.49319180846214294, "learning_rate": 9.992477254274568e-05, "loss": 0.0532, "step": 6660 }, { "grad_norm": 0.40273258090019226, "learning_rate": 9.992386313916183e-05, "loss": 0.0586, "step": 6670 }, { "grad_norm": 0.4587417244911194, "learning_rate": 9.992294827597934e-05, "loss": 0.0631, "step": 6680 }, { "grad_norm": 0.4416927695274353, "learning_rate": 9.992202795329831e-05, "loss": 0.0574, "step": 6690 }, { "grad_norm": 0.6062463521957397, "learning_rate": 9.992110217121936e-05, "loss": 0.0639, "step": 6700 }, { "grad_norm": 0.7306600213050842, "learning_rate": 9.992017092984372e-05, "loss": 0.055, "step": 6710 }, { "grad_norm": 0.45201268792152405, "learning_rate": 9.991923422927326e-05, "loss": 0.0549, "step": 6720 }, { "grad_norm": 0.5411205291748047, "learning_rate": 9.991829206961037e-05, "loss": 0.0572, "step": 6730 }, { "grad_norm": 0.5462625622749329, "learning_rate": 9.991734445095813e-05, "loss": 0.06, "step": 6740 }, { "grad_norm": 0.6583700776100159, "learning_rate": 9.991639137342015e-05, "loss": 0.0566, "step": 6750 }, { "grad_norm": 0.5785687565803528, "learning_rate": 9.991543283710064e-05, "loss": 0.056, "step": 6760 }, { "grad_norm": 0.5158916711807251, "learning_rate": 9.991446884210445e-05, "loss": 0.0548, "step": 6770 }, { "grad_norm": 0.6058857440948486, "learning_rate": 9.9913499388537e-05, "loss": 0.0574, "step": 6780 }, { "grad_norm": 0.5114157795906067, "learning_rate": 9.99125244765043e-05, "loss": 0.0545, "step": 6790 }, { "grad_norm": 0.5207837224006653, "learning_rate": 9.991154410611296e-05, "loss": 0.0609, "step": 6800 }, { "grad_norm": 0.8999580144882202, "learning_rate": 9.99105582774702e-05, "loss": 0.0607, "step": 6810 }, { "grad_norm": 0.43764498829841614, "learning_rate": 9.990956699068384e-05, "loss": 0.0577, "step": 6820 }, { "grad_norm": 0.3837473392486572, "learning_rate": 9.990857024586224e-05, "loss": 0.0523, "step": 6830 }, { "grad_norm": 0.4309999942779541, "learning_rate": 9.990756804311446e-05, "loss": 0.0535, "step": 6840 }, { "grad_norm": 0.5109042525291443, "learning_rate": 9.990656038255006e-05, "loss": 0.0535, "step": 6850 }, { "grad_norm": 0.4840172231197357, "learning_rate": 9.990554726427926e-05, "loss": 0.0641, "step": 6860 }, { "grad_norm": 0.49469462037086487, "learning_rate": 9.990452868841284e-05, "loss": 0.058, "step": 6870 }, { "grad_norm": 0.5143592357635498, "learning_rate": 9.99035046550622e-05, "loss": 0.0696, "step": 6880 }, { "grad_norm": 0.6566168069839478, "learning_rate": 9.99024751643393e-05, "loss": 0.0576, "step": 6890 }, { "grad_norm": 0.5135437846183777, "learning_rate": 9.990144021635677e-05, "loss": 0.0557, "step": 6900 }, { "grad_norm": 0.44709157943725586, "learning_rate": 9.990039981122775e-05, "loss": 0.0581, "step": 6910 }, { "grad_norm": 0.437539666891098, "learning_rate": 9.989935394906602e-05, "loss": 0.0621, "step": 6920 }, { "grad_norm": 0.4593367278575897, "learning_rate": 9.989830262998598e-05, "loss": 0.0511, "step": 6930 }, { "grad_norm": 0.5522081851959229, "learning_rate": 9.989724585410259e-05, "loss": 0.0491, "step": 6940 }, { "grad_norm": 0.5561962127685547, "learning_rate": 9.989618362153139e-05, "loss": 0.0588, "step": 6950 }, { "grad_norm": 0.5469204187393188, "learning_rate": 9.989511593238859e-05, "loss": 0.0567, "step": 6960 }, { "grad_norm": 0.4487089216709137, "learning_rate": 9.98940427867909e-05, "loss": 0.0542, "step": 6970 }, { "grad_norm": 0.5293835997581482, "learning_rate": 9.989296418485573e-05, "loss": 0.066, "step": 6980 }, { "grad_norm": 0.5739679336547852, "learning_rate": 9.989188012670101e-05, "loss": 0.0703, "step": 6990 }, { "grad_norm": 0.5499370694160461, "learning_rate": 9.989079061244528e-05, "loss": 0.059, "step": 7000 }, { "grad_norm": 0.540097713470459, "learning_rate": 9.988969564220769e-05, "loss": 0.0546, "step": 7010 }, { "grad_norm": 0.6283277273178101, "learning_rate": 9.988859521610801e-05, "loss": 0.0541, "step": 7020 }, { "grad_norm": 0.9863875508308411, "learning_rate": 9.988748933426656e-05, "loss": 0.06, "step": 7030 }, { "grad_norm": 0.4722917079925537, "learning_rate": 9.988637799680428e-05, "loss": 0.071, "step": 7040 }, { "grad_norm": 0.5056697726249695, "learning_rate": 9.98852612038427e-05, "loss": 0.0581, "step": 7050 }, { "grad_norm": 0.490815669298172, "learning_rate": 9.988413895550397e-05, "loss": 0.0521, "step": 7060 }, { "grad_norm": 0.4851094186306, "learning_rate": 9.98830112519108e-05, "loss": 0.0618, "step": 7070 }, { "grad_norm": 0.676944375038147, "learning_rate": 9.98818780931865e-05, "loss": 0.0547, "step": 7080 }, { "grad_norm": 0.5119056105613708, "learning_rate": 9.988073947945502e-05, "loss": 0.0547, "step": 7090 }, { "grad_norm": 0.5255060791969299, "learning_rate": 9.987959541084087e-05, "loss": 0.0584, "step": 7100 }, { "grad_norm": 0.48249751329421997, "learning_rate": 9.987844588746915e-05, "loss": 0.0673, "step": 7110 }, { "grad_norm": 0.5260187387466431, "learning_rate": 9.987729090946558e-05, "loss": 0.068, "step": 7120 }, { "grad_norm": 0.41458603739738464, "learning_rate": 9.987613047695647e-05, "loss": 0.0583, "step": 7130 }, { "grad_norm": 0.43881431221961975, "learning_rate": 9.987496459006871e-05, "loss": 0.0608, "step": 7140 }, { "grad_norm": 0.9551960229873657, "learning_rate": 9.987379324892982e-05, "loss": 0.0699, "step": 7150 }, { "grad_norm": 0.3941069543361664, "learning_rate": 9.987261645366788e-05, "loss": 0.0629, "step": 7160 }, { "grad_norm": 0.4158521592617035, "learning_rate": 9.987143420441158e-05, "loss": 0.0593, "step": 7170 }, { "grad_norm": 0.48521843552589417, "learning_rate": 9.987024650129022e-05, "loss": 0.0548, "step": 7180 }, { "grad_norm": 0.6117210388183594, "learning_rate": 9.986905334443368e-05, "loss": 0.063, "step": 7190 }, { "grad_norm": 0.6015996336936951, "learning_rate": 9.986785473397245e-05, "loss": 0.0714, "step": 7200 }, { "grad_norm": 0.5979967713356018, "learning_rate": 9.98666506700376e-05, "loss": 0.0502, "step": 7210 }, { "grad_norm": 0.6005450487136841, "learning_rate": 9.986544115276081e-05, "loss": 0.0585, "step": 7220 }, { "grad_norm": 0.5622270703315735, "learning_rate": 9.986422618227433e-05, "loss": 0.0542, "step": 7230 }, { "grad_norm": 0.4300582706928253, "learning_rate": 9.986300575871106e-05, "loss": 0.0569, "step": 7240 }, { "grad_norm": 0.5736968517303467, "learning_rate": 9.986177988220444e-05, "loss": 0.0521, "step": 7250 }, { "grad_norm": 1.1668131351470947, "learning_rate": 9.986054855288856e-05, "loss": 0.0653, "step": 7260 }, { "grad_norm": 0.5183162093162537, "learning_rate": 9.985931177089802e-05, "loss": 0.0596, "step": 7270 }, { "grad_norm": 0.49437350034713745, "learning_rate": 9.985806953636814e-05, "loss": 0.0544, "step": 7280 }, { "grad_norm": 0.6506608724594116, "learning_rate": 9.985682184943471e-05, "loss": 0.07, "step": 7290 }, { "grad_norm": 0.47832077741622925, "learning_rate": 9.98555687102342e-05, "loss": 0.0596, "step": 7300 }, { "grad_norm": 0.49826234579086304, "learning_rate": 9.985431011890367e-05, "loss": 0.0679, "step": 7310 }, { "grad_norm": 0.5331818461418152, "learning_rate": 9.985304607558075e-05, "loss": 0.0601, "step": 7320 }, { "grad_norm": 0.5291305184364319, "learning_rate": 9.985177658040364e-05, "loss": 0.0606, "step": 7330 }, { "grad_norm": 0.476814866065979, "learning_rate": 9.985050163351119e-05, "loss": 0.057, "step": 7340 }, { "grad_norm": 0.39895159006118774, "learning_rate": 9.984922123504286e-05, "loss": 0.0518, "step": 7350 }, { "grad_norm": 0.4516090750694275, "learning_rate": 9.984793538513862e-05, "loss": 0.0523, "step": 7360 }, { "grad_norm": 1.226648211479187, "learning_rate": 9.984664408393912e-05, "loss": 0.0513, "step": 7370 }, { "grad_norm": 0.43320024013519287, "learning_rate": 9.984534733158556e-05, "loss": 0.0597, "step": 7380 }, { "grad_norm": 0.8738548159599304, "learning_rate": 9.984404512821977e-05, "loss": 0.0592, "step": 7390 }, { "grad_norm": 0.5630285739898682, "learning_rate": 9.984273747398411e-05, "loss": 0.0621, "step": 7400 }, { "grad_norm": 0.5730205178260803, "learning_rate": 9.984142436902165e-05, "loss": 0.0557, "step": 7410 }, { "grad_norm": 0.4245639741420746, "learning_rate": 9.984010581347596e-05, "loss": 0.0583, "step": 7420 }, { "grad_norm": 0.4028121531009674, "learning_rate": 9.983878180749121e-05, "loss": 0.0623, "step": 7430 }, { "grad_norm": 0.5876731872558594, "learning_rate": 9.983745235121222e-05, "loss": 0.0657, "step": 7440 }, { "grad_norm": 0.4907892942428589, "learning_rate": 9.983611744478438e-05, "loss": 0.0557, "step": 7450 }, { "grad_norm": 0.5375849008560181, "learning_rate": 9.983477708835365e-05, "loss": 0.0603, "step": 7460 }, { "grad_norm": 0.4267846643924713, "learning_rate": 9.983343128206664e-05, "loss": 0.0562, "step": 7470 }, { "grad_norm": 0.49437904357910156, "learning_rate": 9.983208002607049e-05, "loss": 0.0534, "step": 7480 }, { "grad_norm": 0.36132749915122986, "learning_rate": 9.9830723320513e-05, "loss": 0.0675, "step": 7490 }, { "grad_norm": 0.5141250491142273, "learning_rate": 9.982936116554254e-05, "loss": 0.0515, "step": 7500 }, { "grad_norm": 0.4111727178096771, "learning_rate": 9.982799356130803e-05, "loss": 0.0491, "step": 7510 }, { "grad_norm": 0.33820241689682007, "learning_rate": 9.982662050795908e-05, "loss": 0.0524, "step": 7520 }, { "grad_norm": 0.5056033730506897, "learning_rate": 9.982524200564583e-05, "loss": 0.0609, "step": 7530 }, { "grad_norm": 0.47481420636177063, "learning_rate": 9.982385805451901e-05, "loss": 0.0541, "step": 7540 }, { "grad_norm": 0.38378357887268066, "learning_rate": 9.982246865472998e-05, "loss": 0.0505, "step": 7550 }, { "grad_norm": 0.4489680230617523, "learning_rate": 9.982107380643069e-05, "loss": 0.0473, "step": 7560 }, { "grad_norm": 0.42669254541397095, "learning_rate": 9.981967350977368e-05, "loss": 0.051, "step": 7570 }, { "grad_norm": 0.4300386607646942, "learning_rate": 9.981826776491208e-05, "loss": 0.0553, "step": 7580 }, { "grad_norm": 0.4989317059516907, "learning_rate": 9.98168565719996e-05, "loss": 0.0525, "step": 7590 }, { "grad_norm": 0.42669135332107544, "learning_rate": 9.98154399311906e-05, "loss": 0.0514, "step": 7600 }, { "grad_norm": 0.5824506878852844, "learning_rate": 9.981401784263997e-05, "loss": 0.0446, "step": 7610 }, { "grad_norm": 0.7348403930664062, "learning_rate": 9.981259030650326e-05, "loss": 0.0534, "step": 7620 }, { "grad_norm": 0.3774261772632599, "learning_rate": 9.981115732293655e-05, "loss": 0.0515, "step": 7630 }, { "grad_norm": 0.39006590843200684, "learning_rate": 9.980971889209659e-05, "loss": 0.0521, "step": 7640 }, { "grad_norm": 2.5797019004821777, "learning_rate": 9.980827501414064e-05, "loss": 0.054, "step": 7650 }, { "grad_norm": 0.37622103095054626, "learning_rate": 9.980682568922663e-05, "loss": 0.0493, "step": 7660 }, { "grad_norm": 0.457053005695343, "learning_rate": 9.980537091751304e-05, "loss": 0.0553, "step": 7670 }, { "grad_norm": 0.3955957293510437, "learning_rate": 9.980391069915897e-05, "loss": 0.0534, "step": 7680 }, { "grad_norm": 0.5126962065696716, "learning_rate": 9.98024450343241e-05, "loss": 0.0516, "step": 7690 }, { "grad_norm": 0.4757890999317169, "learning_rate": 9.980097392316872e-05, "loss": 0.0556, "step": 7700 }, { "grad_norm": 0.6406366229057312, "learning_rate": 9.97994973658537e-05, "loss": 0.0585, "step": 7710 }, { "grad_norm": 0.6605742573738098, "learning_rate": 9.979801536254054e-05, "loss": 0.0542, "step": 7720 }, { "grad_norm": 0.46581995487213135, "learning_rate": 9.979652791339127e-05, "loss": 0.0842, "step": 7730 }, { "grad_norm": 0.5839205384254456, "learning_rate": 9.97950350185686e-05, "loss": 0.0629, "step": 7740 }, { "grad_norm": 0.5231837630271912, "learning_rate": 9.979353667823574e-05, "loss": 0.0671, "step": 7750 }, { "grad_norm": 0.949876070022583, "learning_rate": 9.979203289255658e-05, "loss": 0.0642, "step": 7760 }, { "grad_norm": 1.1924245357513428, "learning_rate": 9.979052366169557e-05, "loss": 0.06, "step": 7770 }, { "grad_norm": 0.430793434381485, "learning_rate": 9.978900898581775e-05, "loss": 0.0613, "step": 7780 }, { "grad_norm": 0.3639037013053894, "learning_rate": 9.978748886508875e-05, "loss": 0.0592, "step": 7790 }, { "grad_norm": 0.4344567060470581, "learning_rate": 9.978596329967484e-05, "loss": 0.0636, "step": 7800 }, { "grad_norm": 0.5066435933113098, "learning_rate": 9.978443228974284e-05, "loss": 0.0565, "step": 7810 }, { "grad_norm": 0.594862163066864, "learning_rate": 9.978289583546015e-05, "loss": 0.0665, "step": 7820 }, { "grad_norm": 0.3314914107322693, "learning_rate": 9.978135393699484e-05, "loss": 0.0611, "step": 7830 }, { "grad_norm": 0.44268137216567993, "learning_rate": 9.977980659451548e-05, "loss": 0.0474, "step": 7840 }, { "grad_norm": 0.46142610907554626, "learning_rate": 9.977825380819135e-05, "loss": 0.0566, "step": 7850 }, { "grad_norm": 0.4126046597957611, "learning_rate": 9.97766955781922e-05, "loss": 0.0638, "step": 7860 }, { "grad_norm": 0.44081223011016846, "learning_rate": 9.977513190468848e-05, "loss": 0.0572, "step": 7870 }, { "grad_norm": 0.4325331747531891, "learning_rate": 9.977356278785116e-05, "loss": 0.0545, "step": 7880 }, { "grad_norm": 0.6284185647964478, "learning_rate": 9.977198822785184e-05, "loss": 0.0514, "step": 7890 }, { "grad_norm": 0.48464301228523254, "learning_rate": 9.977040822486273e-05, "loss": 0.057, "step": 7900 }, { "grad_norm": 0.8892011046409607, "learning_rate": 9.97688227790566e-05, "loss": 0.0647, "step": 7910 }, { "grad_norm": 0.4517996907234192, "learning_rate": 9.976723189060684e-05, "loss": 0.054, "step": 7920 }, { "grad_norm": 0.579922616481781, "learning_rate": 9.976563555968742e-05, "loss": 0.0538, "step": 7930 }, { "grad_norm": 0.6101276278495789, "learning_rate": 9.976403378647292e-05, "loss": 0.0528, "step": 7940 }, { "grad_norm": 0.4213288128376007, "learning_rate": 9.97624265711385e-05, "loss": 0.054, "step": 7950 }, { "grad_norm": 0.3808591067790985, "learning_rate": 9.976081391385993e-05, "loss": 0.0543, "step": 7960 }, { "grad_norm": 0.6334872245788574, "learning_rate": 9.975919581481356e-05, "loss": 0.0518, "step": 7970 }, { "grad_norm": 0.5221696496009827, "learning_rate": 9.975757227417634e-05, "loss": 0.0617, "step": 7980 }, { "grad_norm": 0.5447481870651245, "learning_rate": 9.975594329212586e-05, "loss": 0.0485, "step": 7990 }, { "grad_norm": 0.49555233120918274, "learning_rate": 9.97543088688402e-05, "loss": 0.048, "step": 8000 }, { "grad_norm": 0.39010149240493774, "learning_rate": 9.975266900449814e-05, "loss": 0.0574, "step": 8010 }, { "grad_norm": 0.45987609028816223, "learning_rate": 9.975102369927898e-05, "loss": 0.0551, "step": 8020 }, { "grad_norm": 0.7003055810928345, "learning_rate": 9.974937295336269e-05, "loss": 0.0528, "step": 8030 }, { "grad_norm": 0.38044095039367676, "learning_rate": 9.974771676692975e-05, "loss": 0.0484, "step": 8040 }, { "grad_norm": 0.44663384556770325, "learning_rate": 9.974605514016131e-05, "loss": 0.0485, "step": 8050 }, { "grad_norm": 0.5800353288650513, "learning_rate": 9.974438807323907e-05, "loss": 0.0574, "step": 8060 }, { "grad_norm": 0.6199719309806824, "learning_rate": 9.974271556634535e-05, "loss": 0.0584, "step": 8070 }, { "grad_norm": 0.4199407398700714, "learning_rate": 9.974103761966302e-05, "loss": 0.0502, "step": 8080 }, { "grad_norm": 0.535765528678894, "learning_rate": 9.973935423337563e-05, "loss": 0.0427, "step": 8090 }, { "grad_norm": 0.5332664251327515, "learning_rate": 9.973766540766722e-05, "loss": 0.0502, "step": 8100 }, { "grad_norm": 0.6096673011779785, "learning_rate": 9.97359711427225e-05, "loss": 0.0518, "step": 8110 }, { "grad_norm": 0.5382151007652283, "learning_rate": 9.973427143872677e-05, "loss": 0.0551, "step": 8120 }, { "grad_norm": 0.3814971148967743, "learning_rate": 9.973256629586589e-05, "loss": 0.043, "step": 8130 }, { "grad_norm": 0.4498235583305359, "learning_rate": 9.973085571432632e-05, "loss": 0.0484, "step": 8140 }, { "grad_norm": 0.35552120208740234, "learning_rate": 9.972913969429513e-05, "loss": 0.0487, "step": 8150 }, { "grad_norm": 0.43856772780418396, "learning_rate": 9.972741823596e-05, "loss": 0.0441, "step": 8160 }, { "grad_norm": 0.5136361718177795, "learning_rate": 9.972569133950917e-05, "loss": 0.0429, "step": 8170 }, { "grad_norm": 0.49025586247444153, "learning_rate": 9.972395900513151e-05, "loss": 0.0508, "step": 8180 }, { "grad_norm": 0.4564492404460907, "learning_rate": 9.972222123301645e-05, "loss": 0.0464, "step": 8190 }, { "grad_norm": 0.5406316518783569, "learning_rate": 9.972047802335403e-05, "loss": 0.0455, "step": 8200 }, { "grad_norm": 0.41685816645622253, "learning_rate": 9.971872937633488e-05, "loss": 0.0496, "step": 8210 }, { "grad_norm": 0.4708992838859558, "learning_rate": 9.971697529215024e-05, "loss": 0.055, "step": 8220 }, { "grad_norm": 0.47125986218452454, "learning_rate": 9.971521577099192e-05, "loss": 0.0576, "step": 8230 }, { "grad_norm": 0.3826236128807068, "learning_rate": 9.971345081305236e-05, "loss": 0.0575, "step": 8240 }, { "grad_norm": 0.38222065567970276, "learning_rate": 9.971168041852456e-05, "loss": 0.0544, "step": 8250 }, { "grad_norm": 0.39231643080711365, "learning_rate": 9.970990458760215e-05, "loss": 0.0514, "step": 8260 }, { "grad_norm": 0.5133360028266907, "learning_rate": 9.970812332047929e-05, "loss": 0.0611, "step": 8270 }, { "grad_norm": 0.4220958650112152, "learning_rate": 9.97063366173508e-05, "loss": 0.045, "step": 8280 }, { "grad_norm": 0.5427730679512024, "learning_rate": 9.970454447841207e-05, "loss": 0.0482, "step": 8290 }, { "grad_norm": 0.3539881110191345, "learning_rate": 9.970274690385909e-05, "loss": 0.0486, "step": 8300 }, { "grad_norm": 0.6563968658447266, "learning_rate": 9.970094389388844e-05, "loss": 0.0522, "step": 8310 }, { "grad_norm": 0.6251498460769653, "learning_rate": 9.969913544869728e-05, "loss": 0.0478, "step": 8320 }, { "grad_norm": 2.5919947624206543, "learning_rate": 9.96973215684834e-05, "loss": 0.0553, "step": 8330 }, { "grad_norm": 0.4726683497428894, "learning_rate": 9.969550225344513e-05, "loss": 0.0451, "step": 8340 }, { "grad_norm": 0.42824530601501465, "learning_rate": 9.969367750378147e-05, "loss": 0.0523, "step": 8350 }, { "grad_norm": 0.4461694061756134, "learning_rate": 9.969184731969194e-05, "loss": 0.0427, "step": 8360 }, { "grad_norm": 0.41650262475013733, "learning_rate": 9.96900117013767e-05, "loss": 0.0424, "step": 8370 }, { "grad_norm": 0.4806586503982544, "learning_rate": 9.96881706490365e-05, "loss": 0.0427, "step": 8380 }, { "grad_norm": 0.37374037504196167, "learning_rate": 9.968632416287265e-05, "loss": 0.0427, "step": 8390 }, { "grad_norm": 0.33775007724761963, "learning_rate": 9.96844722430871e-05, "loss": 0.0449, "step": 8400 }, { "grad_norm": 0.30657219886779785, "learning_rate": 9.968261488988235e-05, "loss": 0.0409, "step": 8410 }, { "grad_norm": 0.6258964538574219, "learning_rate": 9.968075210346155e-05, "loss": 0.0513, "step": 8420 }, { "grad_norm": 0.45648840069770813, "learning_rate": 9.967888388402839e-05, "loss": 0.0408, "step": 8430 }, { "grad_norm": 0.3789416551589966, "learning_rate": 9.967701023178717e-05, "loss": 0.0392, "step": 8440 }, { "grad_norm": 0.6197512745857239, "learning_rate": 9.967513114694282e-05, "loss": 0.045, "step": 8450 }, { "grad_norm": 0.46207568049430847, "learning_rate": 9.967324662970079e-05, "loss": 0.0392, "step": 8460 }, { "grad_norm": 0.42255252599716187, "learning_rate": 9.96713566802672e-05, "loss": 0.0445, "step": 8470 }, { "grad_norm": 0.4404872953891754, "learning_rate": 9.966946129884873e-05, "loss": 0.0499, "step": 8480 }, { "grad_norm": 0.3843977153301239, "learning_rate": 9.966756048565265e-05, "loss": 0.0468, "step": 8490 }, { "grad_norm": 0.34840327501296997, "learning_rate": 9.966565424088681e-05, "loss": 0.043, "step": 8500 }, { "grad_norm": 0.38875988125801086, "learning_rate": 9.96637425647597e-05, "loss": 0.0456, "step": 8510 }, { "grad_norm": 0.4171721041202545, "learning_rate": 9.966182545748038e-05, "loss": 0.0399, "step": 8520 }, { "grad_norm": 0.46636903285980225, "learning_rate": 9.96599029192585e-05, "loss": 0.0437, "step": 8530 }, { "grad_norm": 0.9473591446876526, "learning_rate": 9.965797495030428e-05, "loss": 0.0455, "step": 8540 }, { "grad_norm": 0.5397949814796448, "learning_rate": 9.96560415508286e-05, "loss": 0.048, "step": 8550 }, { "grad_norm": 1.0774827003479004, "learning_rate": 9.965410272104286e-05, "loss": 0.0473, "step": 8560 }, { "grad_norm": 0.6427550911903381, "learning_rate": 9.96521584611591e-05, "loss": 0.0522, "step": 8570 }, { "grad_norm": 0.42389607429504395, "learning_rate": 9.965020877138994e-05, "loss": 0.0455, "step": 8580 }, { "grad_norm": 0.9097250699996948, "learning_rate": 9.964825365194861e-05, "loss": 0.0441, "step": 8590 }, { "grad_norm": 0.4613935649394989, "learning_rate": 9.96462931030489e-05, "loss": 0.0501, "step": 8600 }, { "grad_norm": 0.5302568078041077, "learning_rate": 9.96443271249052e-05, "loss": 0.0534, "step": 8610 }, { "grad_norm": 0.3909236192703247, "learning_rate": 9.964235571773255e-05, "loss": 0.0482, "step": 8620 }, { "grad_norm": 0.470441997051239, "learning_rate": 9.96403788817465e-05, "loss": 0.0515, "step": 8630 }, { "grad_norm": 0.6424553990364075, "learning_rate": 9.963839661716325e-05, "loss": 0.052, "step": 8640 }, { "grad_norm": 0.5089688301086426, "learning_rate": 9.963640892419958e-05, "loss": 0.0585, "step": 8650 }, { "grad_norm": 0.6183431148529053, "learning_rate": 9.963441580307286e-05, "loss": 0.0579, "step": 8660 }, { "grad_norm": 0.43516698479652405, "learning_rate": 9.963241725400104e-05, "loss": 0.0534, "step": 8670 }, { "grad_norm": 0.6103042960166931, "learning_rate": 9.963041327720271e-05, "loss": 0.0633, "step": 8680 }, { "grad_norm": 0.5465630888938904, "learning_rate": 9.962840387289697e-05, "loss": 0.0557, "step": 8690 }, { "grad_norm": 0.6141301989555359, "learning_rate": 9.962638904130363e-05, "loss": 0.0536, "step": 8700 }, { "grad_norm": 0.41912978887557983, "learning_rate": 9.962436878264298e-05, "loss": 0.0487, "step": 8710 }, { "grad_norm": 0.6041111350059509, "learning_rate": 9.962234309713598e-05, "loss": 0.056, "step": 8720 }, { "grad_norm": 0.47597089409828186, "learning_rate": 9.962031198500414e-05, "loss": 0.0533, "step": 8730 }, { "grad_norm": 0.47492969036102295, "learning_rate": 9.961827544646958e-05, "loss": 0.0495, "step": 8740 }, { "grad_norm": 0.3994531035423279, "learning_rate": 9.961623348175501e-05, "loss": 0.0534, "step": 8750 }, { "grad_norm": 0.4427611231803894, "learning_rate": 9.961418609108377e-05, "loss": 0.0566, "step": 8760 }, { "grad_norm": 0.43315884470939636, "learning_rate": 9.961213327467971e-05, "loss": 0.0501, "step": 8770 }, { "grad_norm": 0.49460458755493164, "learning_rate": 9.961007503276736e-05, "loss": 0.0517, "step": 8780 }, { "grad_norm": 0.44636285305023193, "learning_rate": 9.960801136557179e-05, "loss": 0.0503, "step": 8790 }, { "grad_norm": 0.40298500657081604, "learning_rate": 9.960594227331866e-05, "loss": 0.0487, "step": 8800 }, { "grad_norm": 0.4753078520298004, "learning_rate": 9.960386775623429e-05, "loss": 0.0561, "step": 8810 }, { "grad_norm": 0.46258002519607544, "learning_rate": 9.96017878145455e-05, "loss": 0.053, "step": 8820 }, { "grad_norm": 0.4538119435310364, "learning_rate": 9.959970244847977e-05, "loss": 0.0504, "step": 8830 }, { "grad_norm": 0.4339433014392853, "learning_rate": 9.959761165826518e-05, "loss": 0.0495, "step": 8840 }, { "grad_norm": 0.45652973651885986, "learning_rate": 9.959551544413033e-05, "loss": 0.0497, "step": 8850 }, { "grad_norm": 0.47318604588508606, "learning_rate": 9.959341380630448e-05, "loss": 0.0498, "step": 8860 }, { "grad_norm": 0.5372815728187561, "learning_rate": 9.959130674501746e-05, "loss": 0.0452, "step": 8870 }, { "grad_norm": 0.3914898931980133, "learning_rate": 9.958919426049968e-05, "loss": 0.0493, "step": 8880 }, { "grad_norm": 0.3784298002719879, "learning_rate": 9.958707635298219e-05, "loss": 0.0419, "step": 8890 }, { "grad_norm": 0.3570441007614136, "learning_rate": 9.958495302269657e-05, "loss": 0.0458, "step": 8900 }, { "grad_norm": 0.38492515683174133, "learning_rate": 9.958282426987503e-05, "loss": 0.0439, "step": 8910 }, { "grad_norm": 0.5066278576850891, "learning_rate": 9.95806900947504e-05, "loss": 0.0423, "step": 8920 }, { "grad_norm": 0.3536410629749298, "learning_rate": 9.957855049755604e-05, "loss": 0.0419, "step": 8930 }, { "grad_norm": 0.43232476711273193, "learning_rate": 9.957640547852593e-05, "loss": 0.0414, "step": 8940 }, { "grad_norm": 0.5929663181304932, "learning_rate": 9.957425503789466e-05, "loss": 0.041, "step": 8950 }, { "grad_norm": 0.33521702885627747, "learning_rate": 9.957209917589738e-05, "loss": 0.0448, "step": 8960 }, { "grad_norm": 0.83194500207901, "learning_rate": 9.956993789276987e-05, "loss": 0.0392, "step": 8970 }, { "grad_norm": 0.46756789088249207, "learning_rate": 9.956777118874847e-05, "loss": 0.0451, "step": 8980 }, { "grad_norm": 0.5201274752616882, "learning_rate": 9.956559906407016e-05, "loss": 0.0405, "step": 8990 }, { "grad_norm": 0.40706610679626465, "learning_rate": 9.956342151897245e-05, "loss": 0.058, "step": 9000 }, { "grad_norm": 2.4973959922790527, "learning_rate": 9.956123855369346e-05, "loss": 0.0477, "step": 9010 }, { "grad_norm": 0.3724578619003296, "learning_rate": 9.955905016847196e-05, "loss": 0.0389, "step": 9020 }, { "grad_norm": 0.4336690604686737, "learning_rate": 9.955685636354723e-05, "loss": 0.0453, "step": 9030 }, { "grad_norm": 0.35476231575012207, "learning_rate": 9.95546571391592e-05, "loss": 0.0459, "step": 9040 }, { "grad_norm": 0.3565969169139862, "learning_rate": 9.955245249554837e-05, "loss": 0.043, "step": 9050 }, { "grad_norm": 0.40900322794914246, "learning_rate": 9.955024243295582e-05, "loss": 0.0451, "step": 9060 }, { "grad_norm": 0.5235406160354614, "learning_rate": 9.954802695162328e-05, "loss": 0.0531, "step": 9070 }, { "grad_norm": 0.43600162863731384, "learning_rate": 9.954580605179302e-05, "loss": 0.0465, "step": 9080 }, { "grad_norm": 0.3779582977294922, "learning_rate": 9.954357973370788e-05, "loss": 0.0409, "step": 9090 }, { "grad_norm": 1.6371595859527588, "learning_rate": 9.954134799761135e-05, "loss": 0.044, "step": 9100 }, { "grad_norm": 0.542080819606781, "learning_rate": 9.953911084374748e-05, "loss": 0.048, "step": 9110 }, { "grad_norm": 0.38668009638786316, "learning_rate": 9.953686827236093e-05, "loss": 0.0504, "step": 9120 }, { "grad_norm": 0.9854800701141357, "learning_rate": 9.953462028369695e-05, "loss": 0.046, "step": 9130 }, { "grad_norm": 0.586185097694397, "learning_rate": 9.953236687800136e-05, "loss": 0.0597, "step": 9140 }, { "grad_norm": 0.4100928008556366, "learning_rate": 9.95301080555206e-05, "loss": 0.0481, "step": 9150 }, { "grad_norm": 0.3597351908683777, "learning_rate": 9.952784381650171e-05, "loss": 0.0521, "step": 9160 }, { "grad_norm": 0.8544567227363586, "learning_rate": 9.952557416119226e-05, "loss": 0.0487, "step": 9170 }, { "grad_norm": 0.3657475709915161, "learning_rate": 9.95232990898405e-05, "loss": 0.0446, "step": 9180 }, { "grad_norm": 0.2907901108264923, "learning_rate": 9.95210186026952e-05, "loss": 0.0379, "step": 9190 }, { "grad_norm": 0.49709370732307434, "learning_rate": 9.951873270000576e-05, "loss": 0.0444, "step": 9200 }, { "grad_norm": 0.31667590141296387, "learning_rate": 9.951644138202216e-05, "loss": 0.0482, "step": 9210 }, { "grad_norm": 0.3144932687282562, "learning_rate": 9.951414464899498e-05, "loss": 0.0513, "step": 9220 }, { "grad_norm": 0.5102889537811279, "learning_rate": 9.951184250117538e-05, "loss": 0.0506, "step": 9230 }, { "grad_norm": 0.36760976910591125, "learning_rate": 9.950953493881513e-05, "loss": 0.048, "step": 9240 }, { "grad_norm": 0.42952761054039, "learning_rate": 9.950722196216658e-05, "loss": 0.0494, "step": 9250 }, { "grad_norm": 0.4793704152107239, "learning_rate": 9.950490357148265e-05, "loss": 0.0476, "step": 9260 }, { "grad_norm": 0.4509572982788086, "learning_rate": 9.950257976701692e-05, "loss": 0.0469, "step": 9270 }, { "grad_norm": 0.42516887187957764, "learning_rate": 9.950025054902348e-05, "loss": 0.0474, "step": 9280 }, { "grad_norm": 0.48022642731666565, "learning_rate": 9.949791591775706e-05, "loss": 0.0513, "step": 9290 }, { "grad_norm": 0.4205854833126068, "learning_rate": 9.949557587347298e-05, "loss": 0.0492, "step": 9300 }, { "grad_norm": 0.45530200004577637, "learning_rate": 9.949323041642713e-05, "loss": 0.054, "step": 9310 }, { "grad_norm": 0.4792146682739258, "learning_rate": 9.949087954687602e-05, "loss": 0.0521, "step": 9320 }, { "grad_norm": 0.37379899621009827, "learning_rate": 9.948852326507672e-05, "loss": 0.051, "step": 9330 }, { "grad_norm": 0.4313650131225586, "learning_rate": 9.948616157128694e-05, "loss": 0.0489, "step": 9340 }, { "grad_norm": 0.4481431245803833, "learning_rate": 9.948379446576493e-05, "loss": 0.0476, "step": 9350 }, { "grad_norm": 0.34844905138015747, "learning_rate": 9.948142194876952e-05, "loss": 0.0446, "step": 9360 }, { "grad_norm": 1.3084111213684082, "learning_rate": 9.947904402056024e-05, "loss": 0.0538, "step": 9370 }, { "grad_norm": 0.30929216742515564, "learning_rate": 9.947666068139708e-05, "loss": 0.047, "step": 9380 }, { "grad_norm": 0.3543936610221863, "learning_rate": 9.947427193154071e-05, "loss": 0.0531, "step": 9390 }, { "grad_norm": 0.3135271668434143, "learning_rate": 9.947187777125233e-05, "loss": 0.048, "step": 9400 }, { "grad_norm": 0.36279356479644775, "learning_rate": 9.946947820079377e-05, "loss": 0.0466, "step": 9410 }, { "grad_norm": 0.44936349987983704, "learning_rate": 9.946707322042747e-05, "loss": 0.0479, "step": 9420 }, { "grad_norm": 0.7782585620880127, "learning_rate": 9.94646628304164e-05, "loss": 0.0454, "step": 9430 }, { "grad_norm": 0.3531700670719147, "learning_rate": 9.946224703102418e-05, "loss": 0.0478, "step": 9440 }, { "grad_norm": 0.4548860490322113, "learning_rate": 9.945982582251498e-05, "loss": 0.0468, "step": 9450 }, { "grad_norm": 0.4364585876464844, "learning_rate": 9.94573992051536e-05, "loss": 0.0479, "step": 9460 }, { "grad_norm": 0.550382673740387, "learning_rate": 9.94549671792054e-05, "loss": 0.0481, "step": 9470 }, { "grad_norm": 0.3582444190979004, "learning_rate": 9.945252974493635e-05, "loss": 0.0429, "step": 9480 }, { "grad_norm": 0.48598676919937134, "learning_rate": 9.9450086902613e-05, "loss": 0.0511, "step": 9490 }, { "grad_norm": 0.5367898344993591, "learning_rate": 9.944763865250248e-05, "loss": 0.0541, "step": 9500 }, { "grad_norm": 0.4475858211517334, "learning_rate": 9.944518499487254e-05, "loss": 0.0448, "step": 9510 }, { "grad_norm": 0.46784722805023193, "learning_rate": 9.944272592999151e-05, "loss": 0.0523, "step": 9520 }, { "grad_norm": 0.5221381187438965, "learning_rate": 9.94402614581283e-05, "loss": 0.0504, "step": 9530 }, { "grad_norm": 0.4996478855609894, "learning_rate": 9.943779157955244e-05, "loss": 0.0523, "step": 9540 }, { "grad_norm": 0.44015178084373474, "learning_rate": 9.943531629453403e-05, "loss": 0.0514, "step": 9550 }, { "grad_norm": 0.38468363881111145, "learning_rate": 9.943283560334375e-05, "loss": 0.0499, "step": 9560 }, { "grad_norm": 0.5547265410423279, "learning_rate": 9.943034950625288e-05, "loss": 0.0581, "step": 9570 }, { "grad_norm": 0.4739452302455902, "learning_rate": 9.942785800353332e-05, "loss": 0.064, "step": 9580 }, { "grad_norm": 0.6062375903129578, "learning_rate": 9.942536109545751e-05, "loss": 0.0493, "step": 9590 }, { "grad_norm": 0.3129434883594513, "learning_rate": 9.942285878229853e-05, "loss": 0.0467, "step": 9600 }, { "grad_norm": 0.47958388924598694, "learning_rate": 9.942035106433001e-05, "loss": 0.0562, "step": 9610 }, { "grad_norm": 0.4665764272212982, "learning_rate": 9.94178379418262e-05, "loss": 0.0488, "step": 9620 }, { "grad_norm": 0.5193402767181396, "learning_rate": 9.941531941506194e-05, "loss": 0.0487, "step": 9630 }, { "grad_norm": 0.43666744232177734, "learning_rate": 9.941279548431263e-05, "loss": 0.0501, "step": 9640 }, { "grad_norm": 0.39590010046958923, "learning_rate": 9.941026614985431e-05, "loss": 0.0501, "step": 9650 }, { "grad_norm": 0.624820351600647, "learning_rate": 9.940773141196357e-05, "loss": 0.0509, "step": 9660 }, { "grad_norm": 0.8073069453239441, "learning_rate": 9.94051912709176e-05, "loss": 0.0473, "step": 9670 }, { "grad_norm": 0.3416990637779236, "learning_rate": 9.940264572699421e-05, "loss": 0.0505, "step": 9680 }, { "grad_norm": 0.48800137639045715, "learning_rate": 9.940009478047174e-05, "loss": 0.0468, "step": 9690 }, { "grad_norm": 0.404014527797699, "learning_rate": 9.939753843162918e-05, "loss": 0.0484, "step": 9700 }, { "grad_norm": 0.9796441197395325, "learning_rate": 9.939497668074609e-05, "loss": 0.0503, "step": 9710 }, { "grad_norm": 0.44421103596687317, "learning_rate": 9.93924095281026e-05, "loss": 0.0572, "step": 9720 }, { "grad_norm": 0.458255797624588, "learning_rate": 9.938983697397948e-05, "loss": 0.0561, "step": 9730 }, { "grad_norm": 0.40066519379615784, "learning_rate": 9.938725901865805e-05, "loss": 0.0425, "step": 9740 }, { "grad_norm": 0.3824402093887329, "learning_rate": 9.93846756624202e-05, "loss": 0.0412, "step": 9750 }, { "grad_norm": 0.3348655104637146, "learning_rate": 9.938208690554849e-05, "loss": 0.0422, "step": 9760 }, { "grad_norm": 0.49581095576286316, "learning_rate": 9.9379492748326e-05, "loss": 0.0413, "step": 9770 }, { "grad_norm": 0.4103480279445648, "learning_rate": 9.937689319103641e-05, "loss": 0.0441, "step": 9780 }, { "grad_norm": 0.5735978484153748, "learning_rate": 9.937428823396404e-05, "loss": 0.0468, "step": 9790 }, { "grad_norm": 0.48176702857017517, "learning_rate": 9.937167787739372e-05, "loss": 0.0453, "step": 9800 }, { "grad_norm": 0.5894036293029785, "learning_rate": 9.936906212161095e-05, "loss": 0.0497, "step": 9810 }, { "grad_norm": 0.3627852201461792, "learning_rate": 9.936644096690176e-05, "loss": 0.0412, "step": 9820 }, { "grad_norm": 0.5372414588928223, "learning_rate": 9.936381441355282e-05, "loss": 0.0439, "step": 9830 }, { "grad_norm": 0.5121433138847351, "learning_rate": 9.936118246185136e-05, "loss": 0.0482, "step": 9840 }, { "grad_norm": 0.6527824997901917, "learning_rate": 9.935854511208518e-05, "loss": 0.0465, "step": 9850 }, { "grad_norm": 0.39444154500961304, "learning_rate": 9.935590236454272e-05, "loss": 0.0454, "step": 9860 }, { "grad_norm": 0.5493718385696411, "learning_rate": 9.935325421951298e-05, "loss": 0.043, "step": 9870 }, { "grad_norm": 0.5779134631156921, "learning_rate": 9.935060067728557e-05, "loss": 0.0515, "step": 9880 }, { "grad_norm": 0.5753738284111023, "learning_rate": 9.934794173815067e-05, "loss": 0.051, "step": 9890 }, { "grad_norm": 0.40682974457740784, "learning_rate": 9.934527740239906e-05, "loss": 0.047, "step": 9900 }, { "grad_norm": 0.5568611025810242, "learning_rate": 9.934260767032209e-05, "loss": 0.0472, "step": 9910 }, { "grad_norm": 0.4147585332393646, "learning_rate": 9.933993254221172e-05, "loss": 0.0461, "step": 9920 }, { "grad_norm": 0.4066184461116791, "learning_rate": 9.933725201836053e-05, "loss": 0.0384, "step": 9930 }, { "grad_norm": 0.4245709776878357, "learning_rate": 9.933456609906162e-05, "loss": 0.0495, "step": 9940 }, { "grad_norm": 0.4453997313976288, "learning_rate": 9.933187478460875e-05, "loss": 0.0475, "step": 9950 }, { "grad_norm": 0.40411266684532166, "learning_rate": 9.93291780752962e-05, "loss": 0.042, "step": 9960 }, { "grad_norm": 0.4879953861236572, "learning_rate": 9.932647597141893e-05, "loss": 0.046, "step": 9970 }, { "grad_norm": 0.375728040933609, "learning_rate": 9.932376847327239e-05, "loss": 0.043, "step": 9980 }, { "grad_norm": 0.4541049003601074, "learning_rate": 9.932105558115268e-05, "loss": 0.0442, "step": 9990 }, { "grad_norm": 0.3880147635936737, "learning_rate": 9.931833729535651e-05, "loss": 0.0472, "step": 10000 }, { "grad_norm": 0.41001608967781067, "learning_rate": 9.931561361618111e-05, "loss": 0.0433, "step": 10010 }, { "grad_norm": 0.3825324773788452, "learning_rate": 9.931288454392435e-05, "loss": 0.038, "step": 10020 }, { "grad_norm": 0.4072825014591217, "learning_rate": 9.931015007888467e-05, "loss": 0.0382, "step": 10030 }, { "grad_norm": 0.47972753643989563, "learning_rate": 9.930741022136112e-05, "loss": 0.0497, "step": 10040 }, { "grad_norm": 0.4792727828025818, "learning_rate": 9.930466497165333e-05, "loss": 0.0387, "step": 10050 }, { "grad_norm": 0.5192506313323975, "learning_rate": 9.93019143300615e-05, "loss": 0.0455, "step": 10060 }, { "grad_norm": 0.42360541224479675, "learning_rate": 9.929915829688644e-05, "loss": 0.047, "step": 10070 }, { "grad_norm": 0.35522958636283875, "learning_rate": 9.929639687242955e-05, "loss": 0.0383, "step": 10080 }, { "grad_norm": 0.8161829710006714, "learning_rate": 9.929363005699281e-05, "loss": 0.0392, "step": 10090 }, { "grad_norm": 0.4080018401145935, "learning_rate": 9.92908578508788e-05, "loss": 0.038, "step": 10100 }, { "grad_norm": 0.3996659517288208, "learning_rate": 9.928808025439069e-05, "loss": 0.038, "step": 10110 }, { "grad_norm": 0.38783514499664307, "learning_rate": 9.928529726783223e-05, "loss": 0.0427, "step": 10120 }, { "grad_norm": 0.35074904561042786, "learning_rate": 9.928250889150774e-05, "loss": 0.038, "step": 10130 }, { "grad_norm": 0.7045159339904785, "learning_rate": 9.92797151257222e-05, "loss": 0.0357, "step": 10140 }, { "grad_norm": 0.32784709334373474, "learning_rate": 9.927691597078108e-05, "loss": 0.0397, "step": 10150 }, { "grad_norm": 0.5771728157997131, "learning_rate": 9.927411142699053e-05, "loss": 0.0424, "step": 10160 }, { "grad_norm": 0.42653265595436096, "learning_rate": 9.927130149465725e-05, "loss": 0.0541, "step": 10170 }, { "grad_norm": 0.48524171113967896, "learning_rate": 9.92684861740885e-05, "loss": 0.0534, "step": 10180 }, { "grad_norm": 0.5034664869308472, "learning_rate": 9.926566546559217e-05, "loss": 0.0514, "step": 10190 }, { "grad_norm": 0.48989927768707275, "learning_rate": 9.926283936947673e-05, "loss": 0.0497, "step": 10200 }, { "grad_norm": 0.34274721145629883, "learning_rate": 9.926000788605126e-05, "loss": 0.0505, "step": 10210 }, { "grad_norm": 0.5400888323783875, "learning_rate": 9.92571710156254e-05, "loss": 0.049, "step": 10220 }, { "grad_norm": 0.3985130488872528, "learning_rate": 9.925432875850936e-05, "loss": 0.0477, "step": 10230 }, { "grad_norm": 0.49928349256515503, "learning_rate": 9.925148111501396e-05, "loss": 0.0444, "step": 10240 }, { "grad_norm": 0.4774416387081146, "learning_rate": 9.924862808545066e-05, "loss": 0.0413, "step": 10250 }, { "grad_norm": 0.3708741068840027, "learning_rate": 9.924576967013141e-05, "loss": 0.0462, "step": 10260 }, { "grad_norm": 2.802293300628662, "learning_rate": 9.924290586936887e-05, "loss": 0.0591, "step": 10270 }, { "grad_norm": 0.4758142828941345, "learning_rate": 9.924003668347614e-05, "loss": 0.0548, "step": 10280 }, { "grad_norm": 0.5156529545783997, "learning_rate": 9.923716211276704e-05, "loss": 0.0532, "step": 10290 }, { "grad_norm": 0.4080441892147064, "learning_rate": 9.923428215755594e-05, "loss": 0.0485, "step": 10300 }, { "grad_norm": 0.5022530555725098, "learning_rate": 9.923139681815775e-05, "loss": 0.0583, "step": 10310 }, { "grad_norm": 0.48110923171043396, "learning_rate": 9.922850609488801e-05, "loss": 0.0567, "step": 10320 }, { "grad_norm": 0.5404818058013916, "learning_rate": 9.922560998806287e-05, "loss": 0.0474, "step": 10330 }, { "grad_norm": 0.37352436780929565, "learning_rate": 9.922270849799905e-05, "loss": 0.0482, "step": 10340 }, { "grad_norm": 0.41054970026016235, "learning_rate": 9.92198016250138e-05, "loss": 0.051, "step": 10350 }, { "grad_norm": 0.49178558588027954, "learning_rate": 9.921688936942506e-05, "loss": 0.0446, "step": 10360 }, { "grad_norm": 0.4849238693714142, "learning_rate": 9.921397173155129e-05, "loss": 0.048, "step": 10370 }, { "grad_norm": 0.45968034863471985, "learning_rate": 9.921104871171157e-05, "loss": 0.0474, "step": 10380 }, { "grad_norm": 0.4358666241168976, "learning_rate": 9.920812031022554e-05, "loss": 0.0474, "step": 10390 }, { "grad_norm": 0.35827475786209106, "learning_rate": 9.920518652741348e-05, "loss": 0.046, "step": 10400 }, { "grad_norm": 0.5532656908035278, "learning_rate": 9.920224736359618e-05, "loss": 0.0502, "step": 10410 }, { "grad_norm": 0.3390635848045349, "learning_rate": 9.91993028190951e-05, "loss": 0.0433, "step": 10420 }, { "grad_norm": 0.44298893213272095, "learning_rate": 9.919635289423222e-05, "loss": 0.0559, "step": 10430 }, { "grad_norm": 5.9259934425354, "learning_rate": 9.919339758933015e-05, "loss": 0.0523, "step": 10440 }, { "grad_norm": 0.30493173003196716, "learning_rate": 9.919043690471209e-05, "loss": 0.048, "step": 10450 }, { "grad_norm": 0.3353487253189087, "learning_rate": 9.91874708407018e-05, "loss": 0.039, "step": 10460 }, { "grad_norm": 1.3167290687561035, "learning_rate": 9.918449939762367e-05, "loss": 0.0483, "step": 10470 }, { "grad_norm": 0.3567793071269989, "learning_rate": 9.91815225758026e-05, "loss": 0.0445, "step": 10480 }, { "grad_norm": 0.37306448817253113, "learning_rate": 9.917854037556419e-05, "loss": 0.05, "step": 10490 }, { "grad_norm": 0.87393718957901, "learning_rate": 9.917555279723454e-05, "loss": 0.0519, "step": 10500 }, { "grad_norm": 0.2930518388748169, "learning_rate": 9.917255984114036e-05, "loss": 0.0467, "step": 10510 }, { "grad_norm": 0.42179468274116516, "learning_rate": 9.916956150760896e-05, "loss": 0.0501, "step": 10520 }, { "grad_norm": 0.38239458203315735, "learning_rate": 9.916655779696826e-05, "loss": 0.0554, "step": 10530 }, { "grad_norm": 0.38811641931533813, "learning_rate": 9.916354870954671e-05, "loss": 0.0499, "step": 10540 }, { "grad_norm": 0.5315424799919128, "learning_rate": 9.91605342456734e-05, "loss": 0.0425, "step": 10550 }, { "grad_norm": 0.3190740942955017, "learning_rate": 9.915751440567795e-05, "loss": 0.0459, "step": 10560 }, { "grad_norm": 0.3855384290218353, "learning_rate": 9.915448918989066e-05, "loss": 0.0486, "step": 10570 }, { "grad_norm": 0.4042559564113617, "learning_rate": 9.915145859864232e-05, "loss": 0.0451, "step": 10580 }, { "grad_norm": 0.573398232460022, "learning_rate": 9.914842263226437e-05, "loss": 0.0454, "step": 10590 }, { "grad_norm": 0.3608562648296356, "learning_rate": 9.914538129108882e-05, "loss": 0.0446, "step": 10600 }, { "grad_norm": 0.8913280963897705, "learning_rate": 9.914233457544825e-05, "loss": 0.0475, "step": 10610 }, { "grad_norm": 0.24640516936779022, "learning_rate": 9.913928248567586e-05, "loss": 0.0441, "step": 10620 }, { "grad_norm": 0.7280918955802917, "learning_rate": 9.913622502210542e-05, "loss": 0.0449, "step": 10630 }, { "grad_norm": 0.5171113014221191, "learning_rate": 9.913316218507128e-05, "loss": 0.0416, "step": 10640 }, { "grad_norm": 0.4664342999458313, "learning_rate": 9.91300939749084e-05, "loss": 0.0418, "step": 10650 }, { "grad_norm": 0.441030889749527, "learning_rate": 9.91270203919523e-05, "loss": 0.0486, "step": 10660 }, { "grad_norm": 0.3960852324962616, "learning_rate": 9.912394143653912e-05, "loss": 0.0398, "step": 10670 }, { "grad_norm": 0.4535740911960602, "learning_rate": 9.912085710900555e-05, "loss": 0.0434, "step": 10680 }, { "grad_norm": 0.35190504789352417, "learning_rate": 9.911776740968892e-05, "loss": 0.0385, "step": 10690 }, { "grad_norm": 0.9080820083618164, "learning_rate": 9.911467233892709e-05, "loss": 0.0489, "step": 10700 }, { "grad_norm": 0.41224491596221924, "learning_rate": 9.911157189705853e-05, "loss": 0.0444, "step": 10710 }, { "grad_norm": 0.3600906431674957, "learning_rate": 9.910846608442229e-05, "loss": 0.0491, "step": 10720 }, { "grad_norm": 0.3066290616989136, "learning_rate": 9.910535490135805e-05, "loss": 0.0431, "step": 10730 }, { "grad_norm": 0.5505589246749878, "learning_rate": 9.910223834820603e-05, "loss": 0.0478, "step": 10740 }, { "grad_norm": 0.6013280153274536, "learning_rate": 9.909911642530703e-05, "loss": 0.0444, "step": 10750 }, { "grad_norm": 1.6639430522918701, "learning_rate": 9.909598913300249e-05, "loss": 0.0464, "step": 10760 }, { "grad_norm": 0.35945358872413635, "learning_rate": 9.909285647163438e-05, "loss": 0.0458, "step": 10770 }, { "grad_norm": 0.357055127620697, "learning_rate": 9.908971844154531e-05, "loss": 0.0454, "step": 10780 }, { "grad_norm": 0.3428214490413666, "learning_rate": 9.908657504307843e-05, "loss": 0.0412, "step": 10790 }, { "grad_norm": 0.47462695837020874, "learning_rate": 9.908342627657751e-05, "loss": 0.0474, "step": 10800 }, { "grad_norm": 0.3124297261238098, "learning_rate": 9.908027214238689e-05, "loss": 0.0427, "step": 10810 }, { "grad_norm": 0.32140323519706726, "learning_rate": 9.90771126408515e-05, "loss": 0.0468, "step": 10820 }, { "grad_norm": 0.3883281648159027, "learning_rate": 9.907394777231685e-05, "loss": 0.0425, "step": 10830 }, { "grad_norm": 0.37568792700767517, "learning_rate": 9.907077753712905e-05, "loss": 0.0411, "step": 10840 }, { "grad_norm": 0.37959951162338257, "learning_rate": 9.906760193563482e-05, "loss": 0.0396, "step": 10850 }, { "grad_norm": 0.3609012961387634, "learning_rate": 9.906442096818139e-05, "loss": 0.0406, "step": 10860 }, { "grad_norm": 0.33586907386779785, "learning_rate": 9.906123463511665e-05, "loss": 0.0374, "step": 10870 }, { "grad_norm": 0.3181557059288025, "learning_rate": 9.905804293678907e-05, "loss": 0.0404, "step": 10880 }, { "grad_norm": 0.3115096092224121, "learning_rate": 9.905484587354766e-05, "loss": 0.0386, "step": 10890 }, { "grad_norm": 0.2525269091129303, "learning_rate": 9.905164344574205e-05, "loss": 0.0351, "step": 10900 }, { "grad_norm": 0.29902365803718567, "learning_rate": 9.904843565372248e-05, "loss": 0.0356, "step": 10910 }, { "grad_norm": 0.33469054102897644, "learning_rate": 9.904522249783972e-05, "loss": 0.0361, "step": 10920 }, { "grad_norm": 0.28218498826026917, "learning_rate": 9.904200397844517e-05, "loss": 0.0398, "step": 10930 }, { "grad_norm": 0.3492736518383026, "learning_rate": 9.903878009589078e-05, "loss": 0.0359, "step": 10940 }, { "grad_norm": 0.5671636462211609, "learning_rate": 9.903555085052915e-05, "loss": 0.044, "step": 10950 }, { "grad_norm": 0.3495110273361206, "learning_rate": 9.903231624271338e-05, "loss": 0.0464, "step": 10960 }, { "grad_norm": 0.4700494110584259, "learning_rate": 9.902907627279724e-05, "loss": 0.04, "step": 10970 }, { "grad_norm": 0.4592263102531433, "learning_rate": 9.902583094113504e-05, "loss": 0.044, "step": 10980 }, { "grad_norm": 0.289448082447052, "learning_rate": 9.902258024808168e-05, "loss": 0.0408, "step": 10990 }, { "grad_norm": 0.31805455684661865, "learning_rate": 9.901932419399264e-05, "loss": 0.043, "step": 11000 }, { "grad_norm": 0.34543895721435547, "learning_rate": 9.9016062779224e-05, "loss": 0.0357, "step": 11010 }, { "grad_norm": 0.5194928646087646, "learning_rate": 9.901279600413242e-05, "loss": 0.039, "step": 11020 }, { "grad_norm": 0.4734278917312622, "learning_rate": 9.900952386907518e-05, "loss": 0.0493, "step": 11030 }, { "grad_norm": 0.42262619733810425, "learning_rate": 9.90062463744101e-05, "loss": 0.0382, "step": 11040 }, { "grad_norm": 0.7191244959831238, "learning_rate": 9.900296352049558e-05, "loss": 0.0466, "step": 11050 }, { "grad_norm": 0.4115765392780304, "learning_rate": 9.899967530769065e-05, "loss": 0.0421, "step": 11060 }, { "grad_norm": 0.4173801839351654, "learning_rate": 9.899638173635489e-05, "loss": 0.0533, "step": 11070 }, { "grad_norm": 0.47601160407066345, "learning_rate": 9.899308280684849e-05, "loss": 0.0482, "step": 11080 }, { "grad_norm": 1.1711596250534058, "learning_rate": 9.898977851953222e-05, "loss": 0.0525, "step": 11090 }, { "grad_norm": 0.42582064867019653, "learning_rate": 9.898646887476741e-05, "loss": 0.0464, "step": 11100 }, { "grad_norm": 0.42991185188293457, "learning_rate": 9.898315387291603e-05, "loss": 0.0491, "step": 11110 }, { "grad_norm": 0.4147515296936035, "learning_rate": 9.89798335143406e-05, "loss": 0.0462, "step": 11120 }, { "grad_norm": 0.38422876596450806, "learning_rate": 9.897650779940419e-05, "loss": 0.0453, "step": 11130 }, { "grad_norm": 0.30972784757614136, "learning_rate": 9.897317672847054e-05, "loss": 0.0432, "step": 11140 }, { "grad_norm": 0.38335004448890686, "learning_rate": 9.89698403019039e-05, "loss": 0.0474, "step": 11150 }, { "grad_norm": 0.38760992884635925, "learning_rate": 9.896649852006917e-05, "loss": 0.04, "step": 11160 }, { "grad_norm": 0.2687951326370239, "learning_rate": 9.896315138333177e-05, "loss": 0.0413, "step": 11170 }, { "grad_norm": 0.3548893332481384, "learning_rate": 9.895979889205774e-05, "loss": 0.045, "step": 11180 }, { "grad_norm": 0.3343964219093323, "learning_rate": 9.895644104661372e-05, "loss": 0.0452, "step": 11190 }, { "grad_norm": 0.3044091463088989, "learning_rate": 9.895307784736691e-05, "loss": 0.0432, "step": 11200 }, { "grad_norm": 0.2403644323348999, "learning_rate": 9.894970929468512e-05, "loss": 0.0472, "step": 11210 }, { "grad_norm": 0.3320629298686981, "learning_rate": 9.89463353889367e-05, "loss": 0.0417, "step": 11220 }, { "grad_norm": 0.3260575830936432, "learning_rate": 9.894295613049065e-05, "loss": 0.0437, "step": 11230 }, { "grad_norm": 0.6957546472549438, "learning_rate": 9.893957151971649e-05, "loss": 0.0386, "step": 11240 }, { "grad_norm": 0.3326488137245178, "learning_rate": 9.893618155698436e-05, "loss": 0.043, "step": 11250 }, { "grad_norm": 0.3494442105293274, "learning_rate": 9.8932786242665e-05, "loss": 0.0376, "step": 11260 }, { "grad_norm": 0.42507919669151306, "learning_rate": 9.89293855771297e-05, "loss": 0.0422, "step": 11270 }, { "grad_norm": 0.31131818890571594, "learning_rate": 9.892597956075036e-05, "loss": 0.0379, "step": 11280 }, { "grad_norm": 0.38325148820877075, "learning_rate": 9.892256819389947e-05, "loss": 0.0451, "step": 11290 }, { "grad_norm": 0.3089788556098938, "learning_rate": 9.891915147695006e-05, "loss": 0.042, "step": 11300 }, { "grad_norm": 0.3957148492336273, "learning_rate": 9.891572941027577e-05, "loss": 0.0409, "step": 11310 }, { "grad_norm": 1.8618247509002686, "learning_rate": 9.89123019942509e-05, "loss": 0.0396, "step": 11320 }, { "grad_norm": 0.38946500420570374, "learning_rate": 9.89088692292502e-05, "loss": 0.0478, "step": 11330 }, { "grad_norm": 0.39324021339416504, "learning_rate": 9.89054311156491e-05, "loss": 0.0467, "step": 11340 }, { "grad_norm": 0.4224552512168884, "learning_rate": 9.890198765382357e-05, "loss": 0.0459, "step": 11350 }, { "grad_norm": 0.3890160918235779, "learning_rate": 9.889853884415021e-05, "loss": 0.0418, "step": 11360 }, { "grad_norm": 0.34982168674468994, "learning_rate": 9.889508468700614e-05, "loss": 0.043, "step": 11370 }, { "grad_norm": 0.3818473517894745, "learning_rate": 9.889162518276915e-05, "loss": 0.0349, "step": 11380 }, { "grad_norm": 0.37020012736320496, "learning_rate": 9.888816033181752e-05, "loss": 0.0381, "step": 11390 }, { "grad_norm": 0.39548906683921814, "learning_rate": 9.888469013453018e-05, "loss": 0.0394, "step": 11400 }, { "grad_norm": 0.41065356135368347, "learning_rate": 9.888121459128663e-05, "loss": 0.0391, "step": 11410 }, { "grad_norm": 0.4419429302215576, "learning_rate": 9.887773370246693e-05, "loss": 0.0454, "step": 11420 }, { "grad_norm": 0.36775991320610046, "learning_rate": 9.887424746845177e-05, "loss": 0.0443, "step": 11430 }, { "grad_norm": 0.3948388695716858, "learning_rate": 9.887075588962239e-05, "loss": 0.0447, "step": 11440 }, { "grad_norm": 0.3822781443595886, "learning_rate": 9.88672589663606e-05, "loss": 0.0412, "step": 11450 }, { "grad_norm": 0.5574654936790466, "learning_rate": 9.886375669904886e-05, "loss": 0.049, "step": 11460 }, { "grad_norm": 0.48014774918556213, "learning_rate": 9.886024908807014e-05, "loss": 0.0495, "step": 11470 }, { "grad_norm": 0.4084067642688751, "learning_rate": 9.885673613380806e-05, "loss": 0.0418, "step": 11480 }, { "grad_norm": 0.45463699102401733, "learning_rate": 9.885321783664676e-05, "loss": 0.0394, "step": 11490 }, { "grad_norm": 0.35454389452934265, "learning_rate": 9.884969419697101e-05, "loss": 0.0449, "step": 11500 }, { "grad_norm": 0.40614399313926697, "learning_rate": 9.884616521516614e-05, "loss": 0.0392, "step": 11510 }, { "grad_norm": 0.6325273513793945, "learning_rate": 9.88426308916181e-05, "loss": 0.0399, "step": 11520 }, { "grad_norm": 0.4426933526992798, "learning_rate": 9.883909122671335e-05, "loss": 0.0495, "step": 11530 }, { "grad_norm": 0.41742992401123047, "learning_rate": 9.883554622083904e-05, "loss": 0.0372, "step": 11540 }, { "grad_norm": 0.3226037323474884, "learning_rate": 9.88319958743828e-05, "loss": 0.0432, "step": 11550 }, { "grad_norm": 0.4160597026348114, "learning_rate": 9.882844018773291e-05, "loss": 0.0388, "step": 11560 }, { "grad_norm": 0.7314870953559875, "learning_rate": 9.882487916127823e-05, "loss": 0.0358, "step": 11570 }, { "grad_norm": 0.3325332999229431, "learning_rate": 9.882131279540815e-05, "loss": 0.0317, "step": 11580 }, { "grad_norm": 0.3366764187812805, "learning_rate": 9.881774109051271e-05, "loss": 0.0437, "step": 11590 }, { "grad_norm": 0.9265457391738892, "learning_rate": 9.881416404698252e-05, "loss": 0.0389, "step": 11600 }, { "grad_norm": 0.36386287212371826, "learning_rate": 9.881058166520873e-05, "loss": 0.0401, "step": 11610 }, { "grad_norm": 0.32561710476875305, "learning_rate": 9.880699394558311e-05, "loss": 0.0409, "step": 11620 }, { "grad_norm": 0.30487537384033203, "learning_rate": 9.880340088849801e-05, "loss": 0.0412, "step": 11630 }, { "grad_norm": 0.32587745785713196, "learning_rate": 9.879980249434637e-05, "loss": 0.0411, "step": 11640 }, { "grad_norm": 0.344536155462265, "learning_rate": 9.879619876352168e-05, "loss": 0.0398, "step": 11650 }, { "grad_norm": 0.3813589811325073, "learning_rate": 9.879258969641809e-05, "loss": 0.0426, "step": 11660 }, { "grad_norm": 0.35393431782722473, "learning_rate": 9.878897529343023e-05, "loss": 0.0412, "step": 11670 }, { "grad_norm": 0.4298749268054962, "learning_rate": 9.878535555495338e-05, "loss": 0.0342, "step": 11680 }, { "grad_norm": 0.3777107894420624, "learning_rate": 9.87817304813834e-05, "loss": 0.0438, "step": 11690 }, { "grad_norm": 0.6569623947143555, "learning_rate": 9.877810007311671e-05, "loss": 0.0454, "step": 11700 }, { "grad_norm": 0.4120385944843292, "learning_rate": 9.877446433055035e-05, "loss": 0.0513, "step": 11710 }, { "grad_norm": 0.4460645020008087, "learning_rate": 9.877082325408191e-05, "loss": 0.0425, "step": 11720 }, { "grad_norm": 0.3482448160648346, "learning_rate": 9.876717684410954e-05, "loss": 0.0426, "step": 11730 }, { "grad_norm": 0.3618328869342804, "learning_rate": 9.876352510103204e-05, "loss": 0.0411, "step": 11740 }, { "grad_norm": 0.3794630765914917, "learning_rate": 9.875986802524875e-05, "loss": 0.0453, "step": 11750 }, { "grad_norm": 0.425236314535141, "learning_rate": 9.87562056171596e-05, "loss": 0.0461, "step": 11760 }, { "grad_norm": 0.6993487477302551, "learning_rate": 9.875253787716511e-05, "loss": 0.0366, "step": 11770 }, { "grad_norm": 0.3780994415283203, "learning_rate": 9.874886480566637e-05, "loss": 0.0422, "step": 11780 }, { "grad_norm": 0.4619484841823578, "learning_rate": 9.874518640306507e-05, "loss": 0.0514, "step": 11790 }, { "grad_norm": 0.3837403953075409, "learning_rate": 9.874150266976347e-05, "loss": 0.0471, "step": 11800 }, { "grad_norm": 0.4288999140262604, "learning_rate": 9.873781360616443e-05, "loss": 0.0454, "step": 11810 }, { "grad_norm": 0.5359160304069519, "learning_rate": 9.873411921267137e-05, "loss": 0.052, "step": 11820 }, { "grad_norm": 0.33744263648986816, "learning_rate": 9.873041948968829e-05, "loss": 0.0398, "step": 11830 }, { "grad_norm": 0.38088124990463257, "learning_rate": 9.872671443761981e-05, "loss": 0.0413, "step": 11840 }, { "grad_norm": 0.44417235255241394, "learning_rate": 9.872300405687109e-05, "loss": 0.0442, "step": 11850 }, { "grad_norm": 0.3089680075645447, "learning_rate": 9.871928834784792e-05, "loss": 0.0491, "step": 11860 }, { "grad_norm": 0.29265496134757996, "learning_rate": 9.871556731095661e-05, "loss": 0.0426, "step": 11870 }, { "grad_norm": 0.3239617347717285, "learning_rate": 9.871184094660411e-05, "loss": 0.0317, "step": 11880 }, { "grad_norm": 0.32439562678337097, "learning_rate": 9.870810925519791e-05, "loss": 0.0417, "step": 11890 }, { "grad_norm": 0.33126217126846313, "learning_rate": 9.870437223714612e-05, "loss": 0.0437, "step": 11900 }, { "grad_norm": 0.362852543592453, "learning_rate": 9.87006298928574e-05, "loss": 0.0477, "step": 11910 }, { "grad_norm": 0.29625049233436584, "learning_rate": 9.869688222274103e-05, "loss": 0.0402, "step": 11920 }, { "grad_norm": 0.28488287329673767, "learning_rate": 9.869312922720681e-05, "loss": 0.0322, "step": 11930 }, { "grad_norm": 0.33785027265548706, "learning_rate": 9.868937090666521e-05, "loss": 0.0384, "step": 11940 }, { "grad_norm": 0.3040868043899536, "learning_rate": 9.86856072615272e-05, "loss": 0.0366, "step": 11950 }, { "grad_norm": 0.4872821867465973, "learning_rate": 9.868183829220438e-05, "loss": 0.0467, "step": 11960 }, { "grad_norm": 0.34737110137939453, "learning_rate": 9.867806399910893e-05, "loss": 0.0468, "step": 11970 }, { "grad_norm": 0.8370829820632935, "learning_rate": 9.867428438265356e-05, "loss": 0.0452, "step": 11980 }, { "grad_norm": 0.3416152894496918, "learning_rate": 9.867049944325165e-05, "loss": 0.042, "step": 11990 }, { "grad_norm": 0.5381631255149841, "learning_rate": 9.86667091813171e-05, "loss": 0.0437, "step": 12000 }, { "grad_norm": 0.3521658480167389, "learning_rate": 9.866291359726438e-05, "loss": 0.0404, "step": 12010 }, { "grad_norm": 0.4846561849117279, "learning_rate": 9.865911269150861e-05, "loss": 0.044, "step": 12020 }, { "grad_norm": 0.30683648586273193, "learning_rate": 9.865530646446544e-05, "loss": 0.037, "step": 12030 }, { "grad_norm": 0.5036363005638123, "learning_rate": 9.86514949165511e-05, "loss": 0.0349, "step": 12040 }, { "grad_norm": 0.38015833497047424, "learning_rate": 9.864767804818243e-05, "loss": 0.0393, "step": 12050 }, { "grad_norm": 0.5206731557846069, "learning_rate": 9.86438558597768e-05, "loss": 0.0475, "step": 12060 }, { "grad_norm": 0.33338412642478943, "learning_rate": 9.864002835175225e-05, "loss": 0.043, "step": 12070 }, { "grad_norm": 0.41551706194877625, "learning_rate": 9.863619552452734e-05, "loss": 0.0432, "step": 12080 }, { "grad_norm": 0.32007306814193726, "learning_rate": 9.863235737852119e-05, "loss": 0.0406, "step": 12090 }, { "grad_norm": 0.3952377438545227, "learning_rate": 9.862851391415356e-05, "loss": 0.0404, "step": 12100 }, { "grad_norm": 0.2606843411922455, "learning_rate": 9.862466513184477e-05, "loss": 0.0396, "step": 12110 }, { "grad_norm": 0.4656042158603668, "learning_rate": 9.86208110320157e-05, "loss": 0.0407, "step": 12120 }, { "grad_norm": 0.2739422917366028, "learning_rate": 9.861695161508784e-05, "loss": 0.0377, "step": 12130 }, { "grad_norm": 0.4070933163166046, "learning_rate": 9.861308688148324e-05, "loss": 0.0397, "step": 12140 }, { "grad_norm": 0.31539520621299744, "learning_rate": 9.860921683162455e-05, "loss": 0.0347, "step": 12150 }, { "grad_norm": 0.7192708849906921, "learning_rate": 9.860534146593499e-05, "loss": 0.044, "step": 12160 }, { "grad_norm": 0.3282639980316162, "learning_rate": 9.860146078483836e-05, "loss": 0.0379, "step": 12170 }, { "grad_norm": 0.30749720335006714, "learning_rate": 9.859757478875905e-05, "loss": 0.0368, "step": 12180 }, { "grad_norm": 0.3352189362049103, "learning_rate": 9.859368347812204e-05, "loss": 0.0426, "step": 12190 }, { "grad_norm": 0.35666531324386597, "learning_rate": 9.858978685335285e-05, "loss": 0.031, "step": 12200 }, { "grad_norm": 0.38723650574684143, "learning_rate": 9.858588491487763e-05, "loss": 0.0413, "step": 12210 }, { "grad_norm": 0.4743836522102356, "learning_rate": 9.858197766312308e-05, "loss": 0.0425, "step": 12220 }, { "grad_norm": 0.3383595049381256, "learning_rate": 9.857806509851649e-05, "loss": 0.0397, "step": 12230 }, { "grad_norm": 0.40302881598472595, "learning_rate": 9.857414722148574e-05, "loss": 0.0411, "step": 12240 }, { "grad_norm": 0.2538917362689972, "learning_rate": 9.857022403245928e-05, "loss": 0.034, "step": 12250 }, { "grad_norm": 0.28748321533203125, "learning_rate": 9.856629553186615e-05, "loss": 0.0409, "step": 12260 }, { "grad_norm": 0.25438183546066284, "learning_rate": 9.856236172013595e-05, "loss": 0.0303, "step": 12270 }, { "grad_norm": 0.35878515243530273, "learning_rate": 9.85584225976989e-05, "loss": 0.036, "step": 12280 }, { "grad_norm": 0.7302705645561218, "learning_rate": 9.855447816498575e-05, "loss": 0.0328, "step": 12290 }, { "grad_norm": 0.25745344161987305, "learning_rate": 9.855052842242787e-05, "loss": 0.0315, "step": 12300 }, { "grad_norm": 0.7614349722862244, "learning_rate": 9.85465733704572e-05, "loss": 0.0373, "step": 12310 }, { "grad_norm": 0.32151147723197937, "learning_rate": 9.854261300950624e-05, "loss": 0.0343, "step": 12320 }, { "grad_norm": 0.36922672390937805, "learning_rate": 9.853864734000813e-05, "loss": 0.0411, "step": 12330 }, { "grad_norm": 0.3581951856613159, "learning_rate": 9.85346763623965e-05, "loss": 0.0303, "step": 12340 }, { "grad_norm": 0.3189183473587036, "learning_rate": 9.853070007710564e-05, "loss": 0.0343, "step": 12350 }, { "grad_norm": 0.33297401666641235, "learning_rate": 9.85267184845704e-05, "loss": 0.0291, "step": 12360 }, { "grad_norm": 0.3819049596786499, "learning_rate": 9.852273158522616e-05, "loss": 0.0318, "step": 12370 }, { "grad_norm": 0.4257146120071411, "learning_rate": 9.851873937950896e-05, "loss": 0.0457, "step": 12380 }, { "grad_norm": 0.29152974486351013, "learning_rate": 9.851474186785537e-05, "loss": 0.0308, "step": 12390 }, { "grad_norm": 0.2722107470035553, "learning_rate": 9.851073905070254e-05, "loss": 0.0329, "step": 12400 }, { "grad_norm": 0.4128510057926178, "learning_rate": 9.850673092848824e-05, "loss": 0.0346, "step": 12410 }, { "grad_norm": 0.46034523844718933, "learning_rate": 9.850271750165077e-05, "loss": 0.0336, "step": 12420 }, { "grad_norm": 0.3475636839866638, "learning_rate": 9.849869877062902e-05, "loss": 0.0324, "step": 12430 }, { "grad_norm": 0.314775288105011, "learning_rate": 9.849467473586252e-05, "loss": 0.0319, "step": 12440 }, { "grad_norm": 0.337896466255188, "learning_rate": 9.849064539779127e-05, "loss": 0.0327, "step": 12450 }, { "grad_norm": 0.3643074035644531, "learning_rate": 9.848661075685594e-05, "loss": 0.0345, "step": 12460 }, { "grad_norm": 0.6918458342552185, "learning_rate": 9.848257081349778e-05, "loss": 0.037, "step": 12470 }, { "grad_norm": 0.3660086989402771, "learning_rate": 9.847852556815856e-05, "loss": 0.0442, "step": 12480 }, { "grad_norm": 0.3973596692085266, "learning_rate": 9.847447502128067e-05, "loss": 0.0391, "step": 12490 }, { "grad_norm": 0.42662933468818665, "learning_rate": 9.847041917330708e-05, "loss": 0.0352, "step": 12500 }, { "grad_norm": 0.342841237783432, "learning_rate": 9.846635802468132e-05, "loss": 0.0313, "step": 12510 }, { "grad_norm": 0.36025747656822205, "learning_rate": 9.84622915758475e-05, "loss": 0.0357, "step": 12520 }, { "grad_norm": 0.2899133265018463, "learning_rate": 9.845821982725034e-05, "loss": 0.0299, "step": 12530 }, { "grad_norm": 0.3623582422733307, "learning_rate": 9.845414277933514e-05, "loss": 0.0323, "step": 12540 }, { "grad_norm": 0.33604639768600464, "learning_rate": 9.845006043254771e-05, "loss": 0.0323, "step": 12550 }, { "grad_norm": 0.3225909471511841, "learning_rate": 9.844597278733451e-05, "loss": 0.0353, "step": 12560 }, { "grad_norm": 0.263520747423172, "learning_rate": 9.844187984414259e-05, "loss": 0.0369, "step": 12570 }, { "grad_norm": 0.38938769698143005, "learning_rate": 9.84377816034195e-05, "loss": 0.0317, "step": 12580 }, { "grad_norm": 0.30623146891593933, "learning_rate": 9.843367806561345e-05, "loss": 0.0345, "step": 12590 }, { "grad_norm": 0.5779309272766113, "learning_rate": 9.842956923117317e-05, "loss": 0.0434, "step": 12600 }, { "grad_norm": 0.4781619608402252, "learning_rate": 9.842545510054802e-05, "loss": 0.0358, "step": 12610 }, { "grad_norm": 0.4956708252429962, "learning_rate": 9.842133567418792e-05, "loss": 0.036, "step": 12620 }, { "grad_norm": 0.2742350399494171, "learning_rate": 9.841721095254333e-05, "loss": 0.0365, "step": 12630 }, { "grad_norm": 0.4241258203983307, "learning_rate": 9.841308093606537e-05, "loss": 0.0347, "step": 12640 }, { "grad_norm": 0.2824845016002655, "learning_rate": 9.840894562520565e-05, "loss": 0.0323, "step": 12650 }, { "grad_norm": 0.34842753410339355, "learning_rate": 9.840480502041642e-05, "loss": 0.0349, "step": 12660 }, { "grad_norm": 0.38000521063804626, "learning_rate": 9.840065912215049e-05, "loss": 0.0499, "step": 12670 }, { "grad_norm": 0.36149370670318604, "learning_rate": 9.839650793086124e-05, "loss": 0.0405, "step": 12680 }, { "grad_norm": 0.3315519094467163, "learning_rate": 9.839235144700265e-05, "loss": 0.0413, "step": 12690 }, { "grad_norm": 0.30617454648017883, "learning_rate": 9.838818967102926e-05, "loss": 0.0392, "step": 12700 }, { "grad_norm": 0.845130980014801, "learning_rate": 9.83840226033962e-05, "loss": 0.0355, "step": 12710 }, { "grad_norm": 0.47882339358329773, "learning_rate": 9.837985024455918e-05, "loss": 0.0386, "step": 12720 }, { "grad_norm": 0.5742230415344238, "learning_rate": 9.837567259497447e-05, "loss": 0.0348, "step": 12730 }, { "grad_norm": 0.3155333697795868, "learning_rate": 9.837148965509894e-05, "loss": 0.0396, "step": 12740 }, { "grad_norm": 0.2564821243286133, "learning_rate": 9.836730142539001e-05, "loss": 0.0324, "step": 12750 }, { "grad_norm": 0.37985938787460327, "learning_rate": 9.836310790630574e-05, "loss": 0.0347, "step": 12760 }, { "grad_norm": 0.43542590737342834, "learning_rate": 9.83589090983047e-05, "loss": 0.0301, "step": 12770 }, { "grad_norm": 0.37803468108177185, "learning_rate": 9.835470500184605e-05, "loss": 0.0433, "step": 12780 }, { "grad_norm": 0.46378740668296814, "learning_rate": 9.835049561738957e-05, "loss": 0.0404, "step": 12790 }, { "grad_norm": 0.3863569498062134, "learning_rate": 9.834628094539558e-05, "loss": 0.0397, "step": 12800 }, { "grad_norm": 0.6249016523361206, "learning_rate": 9.834206098632499e-05, "loss": 0.0408, "step": 12810 }, { "grad_norm": 0.40804195404052734, "learning_rate": 9.833783574063931e-05, "loss": 0.0399, "step": 12820 }, { "grad_norm": 0.3580995500087738, "learning_rate": 9.833360520880058e-05, "loss": 0.0383, "step": 12830 }, { "grad_norm": 0.4365116059780121, "learning_rate": 9.832936939127144e-05, "loss": 0.0382, "step": 12840 }, { "grad_norm": 0.6366815567016602, "learning_rate": 9.832512828851515e-05, "loss": 0.0319, "step": 12850 }, { "grad_norm": 0.4187876284122467, "learning_rate": 9.832088190099546e-05, "loss": 0.0322, "step": 12860 }, { "grad_norm": 0.420536607503891, "learning_rate": 9.831663022917679e-05, "loss": 0.0337, "step": 12870 }, { "grad_norm": 1.6964929103851318, "learning_rate": 9.831237327352407e-05, "loss": 0.0373, "step": 12880 }, { "grad_norm": 0.4748116135597229, "learning_rate": 9.830811103450286e-05, "loss": 0.0424, "step": 12890 }, { "grad_norm": 1.6416715383529663, "learning_rate": 9.830384351257924e-05, "loss": 0.0416, "step": 12900 }, { "grad_norm": 2.590372085571289, "learning_rate": 9.829957070821993e-05, "loss": 0.0434, "step": 12910 }, { "grad_norm": 0.43357834219932556, "learning_rate": 9.829529262189218e-05, "loss": 0.0441, "step": 12920 }, { "grad_norm": 0.4055343270301819, "learning_rate": 9.829100925406385e-05, "loss": 0.0428, "step": 12930 }, { "grad_norm": 0.39213043451309204, "learning_rate": 9.828672060520333e-05, "loss": 0.0388, "step": 12940 }, { "grad_norm": 0.46822917461395264, "learning_rate": 9.828242667577966e-05, "loss": 0.0381, "step": 12950 }, { "grad_norm": 0.4218074381351471, "learning_rate": 9.82781274662624e-05, "loss": 0.0429, "step": 12960 }, { "grad_norm": 0.6366870999336243, "learning_rate": 9.82738229771217e-05, "loss": 0.0482, "step": 12970 }, { "grad_norm": 0.4826827049255371, "learning_rate": 9.826951320882829e-05, "loss": 0.0439, "step": 12980 }, { "grad_norm": 0.25711700320243835, "learning_rate": 9.826519816185351e-05, "loss": 0.0427, "step": 12990 }, { "grad_norm": 0.33314386010169983, "learning_rate": 9.826087783666921e-05, "loss": 0.0484, "step": 13000 }, { "grad_norm": 0.28592637181282043, "learning_rate": 9.825655223374787e-05, "loss": 0.0409, "step": 13010 }, { "grad_norm": 0.30010443925857544, "learning_rate": 9.825222135356253e-05, "loss": 0.0387, "step": 13020 }, { "grad_norm": 0.346475213766098, "learning_rate": 9.82478851965868e-05, "loss": 0.0338, "step": 13030 }, { "grad_norm": 0.24638640880584717, "learning_rate": 9.82435437632949e-05, "loss": 0.0279, "step": 13040 }, { "grad_norm": 0.37146392464637756, "learning_rate": 9.823919705416158e-05, "loss": 0.0325, "step": 13050 }, { "grad_norm": 0.31873708963394165, "learning_rate": 9.82348450696622e-05, "loss": 0.0289, "step": 13060 }, { "grad_norm": 0.3173167407512665, "learning_rate": 9.823048781027268e-05, "loss": 0.0293, "step": 13070 }, { "grad_norm": 0.31511619687080383, "learning_rate": 9.822612527646953e-05, "loss": 0.0297, "step": 13080 }, { "grad_norm": 0.2894497215747833, "learning_rate": 9.822175746872984e-05, "loss": 0.0303, "step": 13090 }, { "grad_norm": 0.2954093813896179, "learning_rate": 9.821738438753123e-05, "loss": 0.0295, "step": 13100 }, { "grad_norm": 0.2881334722042084, "learning_rate": 9.821300603335196e-05, "loss": 0.0267, "step": 13110 }, { "grad_norm": 2.1423983573913574, "learning_rate": 9.820862240667085e-05, "loss": 0.0326, "step": 13120 }, { "grad_norm": 0.7998659610748291, "learning_rate": 9.820423350796726e-05, "loss": 0.0315, "step": 13130 }, { "grad_norm": 0.315233439207077, "learning_rate": 9.819983933772118e-05, "loss": 0.0299, "step": 13140 }, { "grad_norm": 0.20786094665527344, "learning_rate": 9.819543989641314e-05, "loss": 0.0336, "step": 13150 }, { "grad_norm": 0.29925018548965454, "learning_rate": 9.819103518452423e-05, "loss": 0.026, "step": 13160 }, { "grad_norm": 0.3295865058898926, "learning_rate": 9.818662520253618e-05, "loss": 0.0302, "step": 13170 }, { "grad_norm": 0.4245566725730896, "learning_rate": 9.818220995093126e-05, "loss": 0.0358, "step": 13180 }, { "grad_norm": 0.3194658160209656, "learning_rate": 9.817778943019228e-05, "loss": 0.0272, "step": 13190 }, { "grad_norm": 0.3238598108291626, "learning_rate": 9.81733636408027e-05, "loss": 0.0341, "step": 13200 }, { "grad_norm": 0.48525843024253845, "learning_rate": 9.816893258324649e-05, "loss": 0.0314, "step": 13210 }, { "grad_norm": 0.33459874987602234, "learning_rate": 9.816449625800823e-05, "loss": 0.0338, "step": 13220 }, { "grad_norm": 0.4578686058521271, "learning_rate": 9.816005466557308e-05, "loss": 0.0344, "step": 13230 }, { "grad_norm": 0.36488470435142517, "learning_rate": 9.815560780642674e-05, "loss": 0.0312, "step": 13240 }, { "grad_norm": 0.42457717657089233, "learning_rate": 9.815115568105555e-05, "loss": 0.0337, "step": 13250 }, { "grad_norm": 0.2385275810956955, "learning_rate": 9.814669828994638e-05, "loss": 0.0343, "step": 13260 }, { "grad_norm": 0.3174508213996887, "learning_rate": 9.814223563358665e-05, "loss": 0.0372, "step": 13270 }, { "grad_norm": 0.3011401295661926, "learning_rate": 9.813776771246443e-05, "loss": 0.0305, "step": 13280 }, { "grad_norm": 0.2714065611362457, "learning_rate": 9.813329452706829e-05, "loss": 0.033, "step": 13290 }, { "grad_norm": 0.9950510263442993, "learning_rate": 9.812881607788744e-05, "loss": 0.0334, "step": 13300 }, { "grad_norm": 0.27672311663627625, "learning_rate": 9.812433236541163e-05, "loss": 0.03, "step": 13310 }, { "grad_norm": 0.26600322127342224, "learning_rate": 9.811984339013116e-05, "loss": 0.0318, "step": 13320 }, { "grad_norm": 0.31336942315101624, "learning_rate": 9.811534915253698e-05, "loss": 0.0411, "step": 13330 }, { "grad_norm": 0.39343616366386414, "learning_rate": 9.811084965312056e-05, "loss": 0.0306, "step": 13340 }, { "grad_norm": 0.37258899211883545, "learning_rate": 9.810634489237396e-05, "loss": 0.0284, "step": 13350 }, { "grad_norm": 0.28902480006217957, "learning_rate": 9.81018348707898e-05, "loss": 0.0327, "step": 13360 }, { "grad_norm": 0.22776129841804504, "learning_rate": 9.809731958886131e-05, "loss": 0.0279, "step": 13370 }, { "grad_norm": 0.33749738335609436, "learning_rate": 9.809279904708224e-05, "loss": 0.0293, "step": 13380 }, { "grad_norm": 0.41091954708099365, "learning_rate": 9.808827324594699e-05, "loss": 0.0366, "step": 13390 }, { "grad_norm": 0.3338131010532379, "learning_rate": 9.808374218595046e-05, "loss": 0.0322, "step": 13400 }, { "grad_norm": 0.3429839313030243, "learning_rate": 9.80792058675882e-05, "loss": 0.0307, "step": 13410 }, { "grad_norm": 0.4394823908805847, "learning_rate": 9.807466429135627e-05, "loss": 0.0331, "step": 13420 }, { "grad_norm": 0.2907661199569702, "learning_rate": 9.807011745775132e-05, "loss": 0.035, "step": 13430 }, { "grad_norm": 0.2547079026699066, "learning_rate": 9.806556536727061e-05, "loss": 0.0362, "step": 13440 }, { "grad_norm": 0.32820144295692444, "learning_rate": 9.806100802041193e-05, "loss": 0.0266, "step": 13450 }, { "grad_norm": 0.4875227212905884, "learning_rate": 9.805644541767368e-05, "loss": 0.0396, "step": 13460 }, { "grad_norm": 0.46049538254737854, "learning_rate": 9.805187755955478e-05, "loss": 0.0354, "step": 13470 }, { "grad_norm": 0.28768378496170044, "learning_rate": 9.804730444655483e-05, "loss": 0.0303, "step": 13480 }, { "grad_norm": 0.2867407202720642, "learning_rate": 9.804272607917388e-05, "loss": 0.0362, "step": 13490 }, { "grad_norm": 0.24716562032699585, "learning_rate": 9.803814245791265e-05, "loss": 0.0335, "step": 13500 }, { "grad_norm": 0.513874888420105, "learning_rate": 9.803355358327239e-05, "loss": 0.0342, "step": 13510 }, { "grad_norm": 0.2843143045902252, "learning_rate": 9.802895945575492e-05, "loss": 0.0265, "step": 13520 }, { "grad_norm": 0.28389474749565125, "learning_rate": 9.802436007586266e-05, "loss": 0.0241, "step": 13530 }, { "grad_norm": 0.9204317927360535, "learning_rate": 9.801975544409858e-05, "loss": 0.0338, "step": 13540 }, { "grad_norm": 0.35618487000465393, "learning_rate": 9.801514556096625e-05, "loss": 0.0306, "step": 13550 }, { "grad_norm": 0.36735448241233826, "learning_rate": 9.801053042696977e-05, "loss": 0.0365, "step": 13560 }, { "grad_norm": 0.35317206382751465, "learning_rate": 9.800591004261388e-05, "loss": 0.0254, "step": 13570 }, { "grad_norm": 0.3270695209503174, "learning_rate": 9.800128440840385e-05, "loss": 0.0366, "step": 13580 }, { "grad_norm": 0.44179993867874146, "learning_rate": 9.799665352484552e-05, "loss": 0.0335, "step": 13590 }, { "grad_norm": 0.43241795897483826, "learning_rate": 9.799201739244532e-05, "loss": 0.034, "step": 13600 }, { "grad_norm": 0.3950100243091583, "learning_rate": 9.798737601171025e-05, "loss": 0.0385, "step": 13610 }, { "grad_norm": 0.37080687284469604, "learning_rate": 9.79827293831479e-05, "loss": 0.0319, "step": 13620 }, { "grad_norm": 0.49937498569488525, "learning_rate": 9.797807750726638e-05, "loss": 0.0348, "step": 13630 }, { "grad_norm": 0.493335098028183, "learning_rate": 9.797342038457446e-05, "loss": 0.0368, "step": 13640 }, { "grad_norm": 0.35389986634254456, "learning_rate": 9.796875801558141e-05, "loss": 0.0304, "step": 13650 }, { "grad_norm": 0.35469210147857666, "learning_rate": 9.79640904007971e-05, "loss": 0.0321, "step": 13660 }, { "grad_norm": 0.5496077537536621, "learning_rate": 9.795941754073199e-05, "loss": 0.0337, "step": 13670 }, { "grad_norm": 0.406946063041687, "learning_rate": 9.795473943589705e-05, "loss": 0.034, "step": 13680 }, { "grad_norm": 0.3655180335044861, "learning_rate": 9.795005608680394e-05, "loss": 0.0316, "step": 13690 }, { "grad_norm": 0.45395946502685547, "learning_rate": 9.794536749396477e-05, "loss": 0.0344, "step": 13700 }, { "grad_norm": 0.4392809569835663, "learning_rate": 9.79406736578923e-05, "loss": 0.0373, "step": 13710 }, { "grad_norm": 1.5104910135269165, "learning_rate": 9.793597457909984e-05, "loss": 0.0381, "step": 13720 }, { "grad_norm": 0.5755910277366638, "learning_rate": 9.793127025810127e-05, "loss": 0.043, "step": 13730 }, { "grad_norm": 0.4747674763202667, "learning_rate": 9.792656069541104e-05, "loss": 0.0393, "step": 13740 }, { "grad_norm": 0.46654486656188965, "learning_rate": 9.79218458915442e-05, "loss": 0.0372, "step": 13750 }, { "grad_norm": 0.41333138942718506, "learning_rate": 9.791712584701634e-05, "loss": 0.0439, "step": 13760 }, { "grad_norm": 0.3293844759464264, "learning_rate": 9.791240056234364e-05, "loss": 0.0338, "step": 13770 }, { "grad_norm": 0.3495325744152069, "learning_rate": 9.790767003804283e-05, "loss": 0.0297, "step": 13780 }, { "grad_norm": 0.4118086099624634, "learning_rate": 9.790293427463126e-05, "loss": 0.0381, "step": 13790 }, { "grad_norm": 0.406476229429245, "learning_rate": 9.789819327262684e-05, "loss": 0.0366, "step": 13800 }, { "grad_norm": 0.42148929834365845, "learning_rate": 9.7893447032548e-05, "loss": 0.0444, "step": 13810 }, { "grad_norm": 0.38623765110969543, "learning_rate": 9.78886955549138e-05, "loss": 0.0328, "step": 13820 }, { "grad_norm": 0.3787703514099121, "learning_rate": 9.788393884024387e-05, "loss": 0.033, "step": 13830 }, { "grad_norm": 0.39868855476379395, "learning_rate": 9.787917688905836e-05, "loss": 0.0319, "step": 13840 }, { "grad_norm": 0.33632326126098633, "learning_rate": 9.787440970187807e-05, "loss": 0.0299, "step": 13850 }, { "grad_norm": 0.3922105133533478, "learning_rate": 9.786963727922429e-05, "loss": 0.0366, "step": 13860 }, { "grad_norm": 0.3569231331348419, "learning_rate": 9.786485962161897e-05, "loss": 0.0326, "step": 13870 }, { "grad_norm": 0.34598734974861145, "learning_rate": 9.786007672958455e-05, "loss": 0.0367, "step": 13880 }, { "grad_norm": 0.5754351019859314, "learning_rate": 9.78552886036441e-05, "loss": 0.0342, "step": 13890 }, { "grad_norm": 0.3346961438655853, "learning_rate": 9.785049524432124e-05, "loss": 0.0344, "step": 13900 }, { "grad_norm": 0.42309996485710144, "learning_rate": 9.784569665214016e-05, "loss": 0.0298, "step": 13910 }, { "grad_norm": 0.4125220775604248, "learning_rate": 9.784089282762563e-05, "loss": 0.0335, "step": 13920 }, { "grad_norm": 0.49755045771598816, "learning_rate": 9.7836083771303e-05, "loss": 0.0287, "step": 13930 }, { "grad_norm": 0.35542014241218567, "learning_rate": 9.783126948369817e-05, "loss": 0.0323, "step": 13940 }, { "grad_norm": 0.654690146446228, "learning_rate": 9.78264499653376e-05, "loss": 0.0416, "step": 13950 }, { "grad_norm": 0.3680904805660248, "learning_rate": 9.782162521674838e-05, "loss": 0.0341, "step": 13960 }, { "grad_norm": 0.40712907910346985, "learning_rate": 9.781679523845812e-05, "loss": 0.0296, "step": 13970 }, { "grad_norm": 0.35209301114082336, "learning_rate": 9.781196003099502e-05, "loss": 0.0272, "step": 13980 }, { "grad_norm": 0.3769875168800354, "learning_rate": 9.780711959488786e-05, "loss": 0.0275, "step": 13990 }, { "grad_norm": 0.270731657743454, "learning_rate": 9.780227393066599e-05, "loss": 0.0285, "step": 14000 }, { "grad_norm": 0.2898511290550232, "learning_rate": 9.77974230388593e-05, "loss": 0.0318, "step": 14010 }, { "grad_norm": 0.32239916920661926, "learning_rate": 9.779256691999829e-05, "loss": 0.0395, "step": 14020 }, { "grad_norm": 0.36103010177612305, "learning_rate": 9.778770557461403e-05, "loss": 0.0278, "step": 14030 }, { "grad_norm": 0.31281575560569763, "learning_rate": 9.778283900323812e-05, "loss": 0.0273, "step": 14040 }, { "grad_norm": 0.24616070091724396, "learning_rate": 9.777796720640277e-05, "loss": 0.0265, "step": 14050 }, { "grad_norm": 0.3254846930503845, "learning_rate": 9.777309018464078e-05, "loss": 0.0253, "step": 14060 }, { "grad_norm": 0.22312279045581818, "learning_rate": 9.776820793848547e-05, "loss": 0.0259, "step": 14070 }, { "grad_norm": 0.2545729875564575, "learning_rate": 9.776332046847075e-05, "loss": 0.0307, "step": 14080 }, { "grad_norm": 2.028982162475586, "learning_rate": 9.775842777513111e-05, "loss": 0.0329, "step": 14090 }, { "grad_norm": 0.2944822311401367, "learning_rate": 9.775352985900163e-05, "loss": 0.0298, "step": 14100 }, { "grad_norm": 0.22328904271125793, "learning_rate": 9.774862672061791e-05, "loss": 0.0244, "step": 14110 }, { "grad_norm": 0.33873772621154785, "learning_rate": 9.774371836051616e-05, "loss": 0.0301, "step": 14120 }, { "grad_norm": 0.4131632447242737, "learning_rate": 9.773880477923315e-05, "loss": 0.0385, "step": 14130 }, { "grad_norm": 0.29833778738975525, "learning_rate": 9.773388597730623e-05, "loss": 0.035, "step": 14140 }, { "grad_norm": 0.3255642056465149, "learning_rate": 9.77289619552733e-05, "loss": 0.0294, "step": 14150 }, { "grad_norm": 0.2699546813964844, "learning_rate": 9.772403271367285e-05, "loss": 0.032, "step": 14160 }, { "grad_norm": 0.23987895250320435, "learning_rate": 9.771909825304396e-05, "loss": 0.0351, "step": 14170 }, { "grad_norm": 0.2833257019519806, "learning_rate": 9.771415857392619e-05, "loss": 0.0319, "step": 14180 }, { "grad_norm": 0.30653274059295654, "learning_rate": 9.770921367685978e-05, "loss": 0.0247, "step": 14190 }, { "grad_norm": 0.7952741980552673, "learning_rate": 9.770426356238551e-05, "loss": 0.0299, "step": 14200 }, { "grad_norm": 0.40142735838890076, "learning_rate": 9.769930823104469e-05, "loss": 0.0286, "step": 14210 }, { "grad_norm": 0.36678972840309143, "learning_rate": 9.769434768337926e-05, "loss": 0.0399, "step": 14220 }, { "grad_norm": 0.33332061767578125, "learning_rate": 9.768938191993164e-05, "loss": 0.0309, "step": 14230 }, { "grad_norm": 0.3119073808193207, "learning_rate": 9.768441094124494e-05, "loss": 0.0261, "step": 14240 }, { "grad_norm": 0.38035792112350464, "learning_rate": 9.767943474786275e-05, "loss": 0.0421, "step": 14250 }, { "grad_norm": 0.26549357175827026, "learning_rate": 9.767445334032923e-05, "loss": 0.0363, "step": 14260 }, { "grad_norm": 0.329635888338089, "learning_rate": 9.766946671918919e-05, "loss": 0.0265, "step": 14270 }, { "grad_norm": 0.292977511882782, "learning_rate": 9.766447488498796e-05, "loss": 0.0316, "step": 14280 }, { "grad_norm": 0.25534775853157043, "learning_rate": 9.765947783827139e-05, "loss": 0.0248, "step": 14290 }, { "grad_norm": 0.448535293340683, "learning_rate": 9.765447557958599e-05, "loss": 0.0238, "step": 14300 }, { "grad_norm": 0.3085227608680725, "learning_rate": 9.764946810947879e-05, "loss": 0.0218, "step": 14310 }, { "grad_norm": 0.26347577571868896, "learning_rate": 9.764445542849738e-05, "loss": 0.0305, "step": 14320 }, { "grad_norm": 0.38117584586143494, "learning_rate": 9.763943753718998e-05, "loss": 0.0279, "step": 14330 }, { "grad_norm": 0.3356933891773224, "learning_rate": 9.76344144361053e-05, "loss": 0.0301, "step": 14340 }, { "grad_norm": 0.34477370977401733, "learning_rate": 9.762938612579269e-05, "loss": 0.0258, "step": 14350 }, { "grad_norm": 0.3984653949737549, "learning_rate": 9.762435260680202e-05, "loss": 0.0332, "step": 14360 }, { "grad_norm": 0.47709882259368896, "learning_rate": 9.761931387968373e-05, "loss": 0.0327, "step": 14370 }, { "grad_norm": 0.633571445941925, "learning_rate": 9.76142699449889e-05, "loss": 0.0341, "step": 14380 }, { "grad_norm": 0.7512354850769043, "learning_rate": 9.760922080326908e-05, "loss": 0.0343, "step": 14390 }, { "grad_norm": 0.3994940519332886, "learning_rate": 9.760416645507644e-05, "loss": 0.0296, "step": 14400 }, { "grad_norm": 0.3102820813655853, "learning_rate": 9.759910690096375e-05, "loss": 0.0324, "step": 14410 }, { "grad_norm": 0.26328223943710327, "learning_rate": 9.759404214148429e-05, "loss": 0.0275, "step": 14420 }, { "grad_norm": 0.3392398953437805, "learning_rate": 9.758897217719191e-05, "loss": 0.03, "step": 14430 }, { "grad_norm": 0.3158876597881317, "learning_rate": 9.758389700864113e-05, "loss": 0.0266, "step": 14440 }, { "grad_norm": 0.33082884550094604, "learning_rate": 9.757881663638688e-05, "loss": 0.0308, "step": 14450 }, { "grad_norm": 1.3602300882339478, "learning_rate": 9.757373106098478e-05, "loss": 0.0307, "step": 14460 }, { "grad_norm": 0.38357943296432495, "learning_rate": 9.756864028299097e-05, "loss": 0.0339, "step": 14470 }, { "grad_norm": 0.461823433637619, "learning_rate": 9.75635443029622e-05, "loss": 0.0413, "step": 14480 }, { "grad_norm": 0.2621620297431946, "learning_rate": 9.755844312145572e-05, "loss": 0.0312, "step": 14490 }, { "grad_norm": 0.2953386604785919, "learning_rate": 9.755333673902941e-05, "loss": 0.0291, "step": 14500 }, { "grad_norm": 0.4392470717430115, "learning_rate": 9.75482251562417e-05, "loss": 0.0277, "step": 14510 }, { "grad_norm": 0.3073704242706299, "learning_rate": 9.754310837365155e-05, "loss": 0.0252, "step": 14520 }, { "grad_norm": 1.1148661375045776, "learning_rate": 9.753798639181856e-05, "loss": 0.0281, "step": 14530 }, { "grad_norm": 0.36815324425697327, "learning_rate": 9.753285921130286e-05, "loss": 0.0344, "step": 14540 }, { "grad_norm": 0.3988829255104065, "learning_rate": 9.752772683266512e-05, "loss": 0.0366, "step": 14550 }, { "grad_norm": 0.2791203558444977, "learning_rate": 9.752258925646665e-05, "loss": 0.0299, "step": 14560 }, { "grad_norm": 0.28884661197662354, "learning_rate": 9.751744648326926e-05, "loss": 0.0359, "step": 14570 }, { "grad_norm": 0.43904635310173035, "learning_rate": 9.751229851363536e-05, "loss": 0.036, "step": 14580 }, { "grad_norm": 0.3026231825351715, "learning_rate": 9.750714534812793e-05, "loss": 0.0271, "step": 14590 }, { "grad_norm": 0.29989638924598694, "learning_rate": 9.750198698731053e-05, "loss": 0.0298, "step": 14600 }, { "grad_norm": 0.28688952326774597, "learning_rate": 9.749682343174722e-05, "loss": 0.0304, "step": 14610 }, { "grad_norm": 0.2878247797489166, "learning_rate": 9.749165468200272e-05, "loss": 0.023, "step": 14620 }, { "grad_norm": 0.6857254505157471, "learning_rate": 9.748648073864229e-05, "loss": 0.037, "step": 14630 }, { "grad_norm": 0.43341490626335144, "learning_rate": 9.748130160223168e-05, "loss": 0.0312, "step": 14640 }, { "grad_norm": 0.42723479866981506, "learning_rate": 9.747611727333734e-05, "loss": 0.0368, "step": 14650 }, { "grad_norm": 0.29862523078918457, "learning_rate": 9.74709277525262e-05, "loss": 0.034, "step": 14660 }, { "grad_norm": 0.4284236431121826, "learning_rate": 9.746573304036576e-05, "loss": 0.0296, "step": 14670 }, { "grad_norm": 0.24557600915431976, "learning_rate": 9.746053313742412e-05, "loss": 0.0299, "step": 14680 }, { "grad_norm": 0.2840944826602936, "learning_rate": 9.745532804426994e-05, "loss": 0.03, "step": 14690 }, { "grad_norm": 0.323404461145401, "learning_rate": 9.745011776147242e-05, "loss": 0.0304, "step": 14700 }, { "grad_norm": 0.7330734133720398, "learning_rate": 9.744490228960138e-05, "loss": 0.0544, "step": 14710 }, { "grad_norm": 0.3938566744327545, "learning_rate": 9.743968162922713e-05, "loss": 0.0376, "step": 14720 }, { "grad_norm": 0.7954387664794922, "learning_rate": 9.743445578092064e-05, "loss": 0.0401, "step": 14730 }, { "grad_norm": 0.45670410990715027, "learning_rate": 9.742922474525338e-05, "loss": 0.0419, "step": 14740 }, { "grad_norm": 0.43654221296310425, "learning_rate": 9.742398852279741e-05, "loss": 0.0331, "step": 14750 }, { "grad_norm": 0.2894115149974823, "learning_rate": 9.741874711412535e-05, "loss": 0.0371, "step": 14760 }, { "grad_norm": 0.7859236598014832, "learning_rate": 9.741350051981042e-05, "loss": 0.0397, "step": 14770 }, { "grad_norm": 0.5208998918533325, "learning_rate": 9.740824874042633e-05, "loss": 0.0457, "step": 14780 }, { "grad_norm": 0.3457801938056946, "learning_rate": 9.740299177654746e-05, "loss": 0.0373, "step": 14790 }, { "grad_norm": 0.30543872714042664, "learning_rate": 9.739772962874867e-05, "loss": 0.0261, "step": 14800 }, { "grad_norm": 0.43183934688568115, "learning_rate": 9.739246229760541e-05, "loss": 0.0331, "step": 14810 }, { "grad_norm": 0.4181123971939087, "learning_rate": 9.738718978369376e-05, "loss": 0.0353, "step": 14820 }, { "grad_norm": 0.3170887529850006, "learning_rate": 9.738191208759025e-05, "loss": 0.0352, "step": 14830 }, { "grad_norm": 0.37558266520500183, "learning_rate": 9.73766292098721e-05, "loss": 0.0293, "step": 14840 }, { "grad_norm": 1.0251072645187378, "learning_rate": 9.737134115111699e-05, "loss": 0.0341, "step": 14850 }, { "grad_norm": 0.3889578580856323, "learning_rate": 9.736604791190323e-05, "loss": 0.0344, "step": 14860 }, { "grad_norm": 0.322632372379303, "learning_rate": 9.73607494928097e-05, "loss": 0.0368, "step": 14870 }, { "grad_norm": 0.3868999779224396, "learning_rate": 9.735544589441581e-05, "loss": 0.0309, "step": 14880 }, { "grad_norm": 0.33648794889450073, "learning_rate": 9.735013711730154e-05, "loss": 0.0344, "step": 14890 }, { "grad_norm": 0.3726816177368164, "learning_rate": 9.734482316204747e-05, "loss": 0.0356, "step": 14900 }, { "grad_norm": 0.2653059959411621, "learning_rate": 9.733950402923473e-05, "loss": 0.028, "step": 14910 }, { "grad_norm": 0.3436557352542877, "learning_rate": 9.7334179719445e-05, "loss": 0.0313, "step": 14920 }, { "grad_norm": 0.5240812301635742, "learning_rate": 9.732885023326053e-05, "loss": 0.0256, "step": 14930 }, { "grad_norm": 0.37742316722869873, "learning_rate": 9.732351557126418e-05, "loss": 0.0321, "step": 14940 }, { "grad_norm": 0.28890568017959595, "learning_rate": 9.731817573403929e-05, "loss": 0.025, "step": 14950 }, { "grad_norm": 2.7037248611450195, "learning_rate": 9.731283072216985e-05, "loss": 0.0272, "step": 14960 }, { "grad_norm": 0.2655199468135834, "learning_rate": 9.730748053624039e-05, "loss": 0.0207, "step": 14970 }, { "grad_norm": 0.24932588636875153, "learning_rate": 9.730212517683598e-05, "loss": 0.0233, "step": 14980 }, { "grad_norm": 0.32305437326431274, "learning_rate": 9.729676464454228e-05, "loss": 0.0335, "step": 14990 }, { "grad_norm": 0.3086260259151459, "learning_rate": 9.72913989399455e-05, "loss": 0.0321, "step": 15000 }, { "grad_norm": 0.5432857275009155, "learning_rate": 9.728602806363242e-05, "loss": 0.0284, "step": 15010 }, { "grad_norm": 0.3656768202781677, "learning_rate": 9.728065201619043e-05, "loss": 0.0281, "step": 15020 }, { "grad_norm": 0.5326757431030273, "learning_rate": 9.727527079820742e-05, "loss": 0.0306, "step": 15030 }, { "grad_norm": 0.3715634047985077, "learning_rate": 9.726988441027186e-05, "loss": 0.0283, "step": 15040 }, { "grad_norm": 0.2983323633670807, "learning_rate": 9.726449285297281e-05, "loss": 0.0298, "step": 15050 }, { "grad_norm": 0.3230816423892975, "learning_rate": 9.72590961268999e-05, "loss": 0.0362, "step": 15060 }, { "grad_norm": 0.4288313090801239, "learning_rate": 9.725369423264328e-05, "loss": 0.0312, "step": 15070 }, { "grad_norm": 0.43540236353874207, "learning_rate": 9.72482871707937e-05, "loss": 0.0327, "step": 15080 }, { "grad_norm": 0.37690600752830505, "learning_rate": 9.724287494194247e-05, "loss": 0.0279, "step": 15090 }, { "grad_norm": 0.3483109474182129, "learning_rate": 9.723745754668147e-05, "loss": 0.0346, "step": 15100 }, { "grad_norm": 0.40049415826797485, "learning_rate": 9.723203498560313e-05, "loss": 0.0346, "step": 15110 }, { "grad_norm": 0.3689298927783966, "learning_rate": 9.722660725930046e-05, "loss": 0.0304, "step": 15120 }, { "grad_norm": 0.38356781005859375, "learning_rate": 9.722117436836702e-05, "loss": 0.0289, "step": 15130 }, { "grad_norm": 0.33033275604248047, "learning_rate": 9.721573631339696e-05, "loss": 0.0283, "step": 15140 }, { "grad_norm": 0.3776438534259796, "learning_rate": 9.721029309498494e-05, "loss": 0.03, "step": 15150 }, { "grad_norm": 0.2781200706958771, "learning_rate": 9.720484471372627e-05, "loss": 0.0241, "step": 15160 }, { "grad_norm": 0.26574623584747314, "learning_rate": 9.719939117021673e-05, "loss": 0.0223, "step": 15170 }, { "grad_norm": 0.3491028845310211, "learning_rate": 9.719393246505275e-05, "loss": 0.0279, "step": 15180 }, { "grad_norm": 0.22248195111751556, "learning_rate": 9.718846859883128e-05, "loss": 0.0227, "step": 15190 }, { "grad_norm": 0.3118223249912262, "learning_rate": 9.718299957214982e-05, "loss": 0.0271, "step": 15200 }, { "grad_norm": 0.2355257272720337, "learning_rate": 9.717752538560646e-05, "loss": 0.0215, "step": 15210 }, { "grad_norm": 0.22786621749401093, "learning_rate": 9.717204603979986e-05, "loss": 0.0212, "step": 15220 }, { "grad_norm": 0.22561262547969818, "learning_rate": 9.716656153532922e-05, "loss": 0.0235, "step": 15230 }, { "grad_norm": 0.37545230984687805, "learning_rate": 9.716107187279434e-05, "loss": 0.0277, "step": 15240 }, { "grad_norm": 0.3393203318119049, "learning_rate": 9.715557705279555e-05, "loss": 0.0302, "step": 15250 }, { "grad_norm": 0.2898181676864624, "learning_rate": 9.715007707593372e-05, "loss": 0.0244, "step": 15260 }, { "grad_norm": 0.4273955225944519, "learning_rate": 9.714457194281036e-05, "loss": 0.0275, "step": 15270 }, { "grad_norm": 0.4445452094078064, "learning_rate": 9.713906165402751e-05, "loss": 0.0281, "step": 15280 }, { "grad_norm": 0.8186404705047607, "learning_rate": 9.713354621018774e-05, "loss": 0.0344, "step": 15290 }, { "grad_norm": 0.31966885924339294, "learning_rate": 9.712802561189422e-05, "loss": 0.0259, "step": 15300 }, { "grad_norm": 0.417953222990036, "learning_rate": 9.712249985975069e-05, "loss": 0.0278, "step": 15310 }, { "grad_norm": 0.4864417016506195, "learning_rate": 9.71169689543614e-05, "loss": 0.0344, "step": 15320 }, { "grad_norm": 0.3469219207763672, "learning_rate": 9.711143289633123e-05, "loss": 0.0312, "step": 15330 }, { "grad_norm": 1.0818543434143066, "learning_rate": 9.710589168626561e-05, "loss": 0.0269, "step": 15340 }, { "grad_norm": 0.39087435603141785, "learning_rate": 9.710034532477048e-05, "loss": 0.0307, "step": 15350 }, { "grad_norm": 0.3670039772987366, "learning_rate": 9.709479381245239e-05, "loss": 0.0277, "step": 15360 }, { "grad_norm": 3.2225589752197266, "learning_rate": 9.708923714991847e-05, "loss": 0.035, "step": 15370 }, { "grad_norm": 0.2666358947753906, "learning_rate": 9.708367533777638e-05, "loss": 0.0308, "step": 15380 }, { "grad_norm": 0.33413419127464294, "learning_rate": 9.707810837663431e-05, "loss": 0.0293, "step": 15390 }, { "grad_norm": 0.291323721408844, "learning_rate": 9.707253626710113e-05, "loss": 0.027, "step": 15400 }, { "grad_norm": 0.2816231846809387, "learning_rate": 9.706695900978613e-05, "loss": 0.0236, "step": 15410 }, { "grad_norm": 0.2551313042640686, "learning_rate": 9.706137660529926e-05, "loss": 0.0276, "step": 15420 }, { "grad_norm": 0.23845326900482178, "learning_rate": 9.705578905425101e-05, "loss": 0.0205, "step": 15430 }, { "grad_norm": 0.3021365702152252, "learning_rate": 9.705019635725241e-05, "loss": 0.0247, "step": 15440 }, { "grad_norm": 0.2720939815044403, "learning_rate": 9.704459851491508e-05, "loss": 0.0279, "step": 15450 }, { "grad_norm": 0.23524396121501923, "learning_rate": 9.703899552785118e-05, "loss": 0.0215, "step": 15460 }, { "grad_norm": 0.35291966795921326, "learning_rate": 9.703338739667346e-05, "loss": 0.0308, "step": 15470 }, { "grad_norm": 0.24137084186077118, "learning_rate": 9.70277741219952e-05, "loss": 0.0247, "step": 15480 }, { "grad_norm": 0.33361610770225525, "learning_rate": 9.702215570443027e-05, "loss": 0.0254, "step": 15490 }, { "grad_norm": 0.28297385573387146, "learning_rate": 9.701653214459309e-05, "loss": 0.0238, "step": 15500 }, { "grad_norm": 0.2752216160297394, "learning_rate": 9.701090344309865e-05, "loss": 0.0224, "step": 15510 }, { "grad_norm": 0.27069777250289917, "learning_rate": 9.700526960056247e-05, "loss": 0.0213, "step": 15520 }, { "grad_norm": 0.4885970652103424, "learning_rate": 9.699963061760068e-05, "loss": 0.0264, "step": 15530 }, { "grad_norm": 0.29934003949165344, "learning_rate": 9.699398649482997e-05, "loss": 0.0288, "step": 15540 }, { "grad_norm": 0.2961845099925995, "learning_rate": 9.698833723286753e-05, "loss": 0.0227, "step": 15550 }, { "grad_norm": 0.3084588944911957, "learning_rate": 9.698268283233118e-05, "loss": 0.0275, "step": 15560 }, { "grad_norm": 0.291008323431015, "learning_rate": 9.697702329383929e-05, "loss": 0.0229, "step": 15570 }, { "grad_norm": 0.3796086311340332, "learning_rate": 9.697135861801074e-05, "loss": 0.024, "step": 15580 }, { "grad_norm": 0.36722347140312195, "learning_rate": 9.696568880546505e-05, "loss": 0.0243, "step": 15590 }, { "grad_norm": 0.22348898649215698, "learning_rate": 9.696001385682223e-05, "loss": 0.0216, "step": 15600 }, { "grad_norm": 0.2872260510921478, "learning_rate": 9.695433377270291e-05, "loss": 0.0217, "step": 15610 }, { "grad_norm": 0.2614717185497284, "learning_rate": 9.694864855372824e-05, "loss": 0.0215, "step": 15620 }, { "grad_norm": 0.25405097007751465, "learning_rate": 9.694295820051995e-05, "loss": 0.0234, "step": 15630 }, { "grad_norm": 0.3544996380805969, "learning_rate": 9.693726271370032e-05, "loss": 0.0221, "step": 15640 }, { "grad_norm": 0.24845857918262482, "learning_rate": 9.693156209389221e-05, "loss": 0.0314, "step": 15650 }, { "grad_norm": 0.31302401423454285, "learning_rate": 9.692585634171905e-05, "loss": 0.0208, "step": 15660 }, { "grad_norm": 0.28775930404663086, "learning_rate": 9.692014545780476e-05, "loss": 0.0225, "step": 15670 }, { "grad_norm": 0.27925851941108704, "learning_rate": 9.691442944277393e-05, "loss": 0.0221, "step": 15680 }, { "grad_norm": 0.2703622579574585, "learning_rate": 9.690870829725162e-05, "loss": 0.0274, "step": 15690 }, { "grad_norm": 0.30078786611557007, "learning_rate": 9.69029820218635e-05, "loss": 0.0397, "step": 15700 }, { "grad_norm": 0.19706454873085022, "learning_rate": 9.689725061723579e-05, "loss": 0.0277, "step": 15710 }, { "grad_norm": 0.4947742223739624, "learning_rate": 9.689151408399527e-05, "loss": 0.0263, "step": 15720 }, { "grad_norm": 0.2928309142589569, "learning_rate": 9.688577242276924e-05, "loss": 0.0318, "step": 15730 }, { "grad_norm": 0.6639355421066284, "learning_rate": 9.688002563418566e-05, "loss": 0.0286, "step": 15740 }, { "grad_norm": 1.2795798778533936, "learning_rate": 9.687427371887293e-05, "loss": 0.0266, "step": 15750 }, { "grad_norm": 0.2554658055305481, "learning_rate": 9.686851667746012e-05, "loss": 0.0272, "step": 15760 }, { "grad_norm": 0.29488635063171387, "learning_rate": 9.686275451057677e-05, "loss": 0.0266, "step": 15770 }, { "grad_norm": 0.3587301969528198, "learning_rate": 9.685698721885308e-05, "loss": 0.0234, "step": 15780 }, { "grad_norm": 0.3477267026901245, "learning_rate": 9.68512148029197e-05, "loss": 0.0241, "step": 15790 }, { "grad_norm": 0.8387120366096497, "learning_rate": 9.684543726340791e-05, "loss": 0.0246, "step": 15800 }, { "grad_norm": 0.2628493309020996, "learning_rate": 9.683965460094952e-05, "loss": 0.0262, "step": 15810 }, { "grad_norm": 0.43639278411865234, "learning_rate": 9.683386681617694e-05, "loss": 0.031, "step": 15820 }, { "grad_norm": 0.29864758253097534, "learning_rate": 9.68280739097231e-05, "loss": 0.0279, "step": 15830 }, { "grad_norm": 0.26883214712142944, "learning_rate": 9.682227588222148e-05, "loss": 0.0305, "step": 15840 }, { "grad_norm": 0.3300938308238983, "learning_rate": 9.681647273430618e-05, "loss": 0.0287, "step": 15850 }, { "grad_norm": 0.362908273935318, "learning_rate": 9.681066446661182e-05, "loss": 0.0217, "step": 15860 }, { "grad_norm": 0.3507244288921356, "learning_rate": 9.680485107977357e-05, "loss": 0.0259, "step": 15870 }, { "grad_norm": 0.8867154717445374, "learning_rate": 9.679903257442716e-05, "loss": 0.03, "step": 15880 }, { "grad_norm": 0.3333189785480499, "learning_rate": 9.679320895120891e-05, "loss": 0.0268, "step": 15890 }, { "grad_norm": 0.4115130305290222, "learning_rate": 9.67873802107557e-05, "loss": 0.0331, "step": 15900 }, { "grad_norm": 4.72317361831665, "learning_rate": 9.67815463537049e-05, "loss": 0.034, "step": 15910 }, { "grad_norm": 0.3038223385810852, "learning_rate": 9.677570738069457e-05, "loss": 0.0317, "step": 15920 }, { "grad_norm": 0.3687269389629364, "learning_rate": 9.676986329236318e-05, "loss": 0.0375, "step": 15930 }, { "grad_norm": 0.2928329110145569, "learning_rate": 9.676401408934987e-05, "loss": 0.0271, "step": 15940 }, { "grad_norm": 0.21710725128650665, "learning_rate": 9.675815977229428e-05, "loss": 0.0279, "step": 15950 }, { "grad_norm": 0.2988209128379822, "learning_rate": 9.675230034183664e-05, "loss": 0.0304, "step": 15960 }, { "grad_norm": 0.26576176285743713, "learning_rate": 9.674643579861773e-05, "loss": 0.026, "step": 15970 }, { "grad_norm": 0.25522953271865845, "learning_rate": 9.674056614327886e-05, "loss": 0.0261, "step": 15980 }, { "grad_norm": 0.256350040435791, "learning_rate": 9.673469137646198e-05, "loss": 0.0204, "step": 15990 }, { "grad_norm": 0.23507428169250488, "learning_rate": 9.67288114988095e-05, "loss": 0.0187, "step": 16000 }, { "grad_norm": 0.27579444646835327, "learning_rate": 9.672292651096447e-05, "loss": 0.0227, "step": 16010 }, { "grad_norm": 0.31773465871810913, "learning_rate": 9.671703641357042e-05, "loss": 0.0219, "step": 16020 }, { "grad_norm": 0.4271884262561798, "learning_rate": 9.67111412072715e-05, "loss": 0.0212, "step": 16030 }, { "grad_norm": 1.1359431743621826, "learning_rate": 9.670524089271242e-05, "loss": 0.0359, "step": 16040 }, { "grad_norm": 0.3808364272117615, "learning_rate": 9.669933547053842e-05, "loss": 0.029, "step": 16050 }, { "grad_norm": 0.3993522524833679, "learning_rate": 9.669342494139531e-05, "loss": 0.0262, "step": 16060 }, { "grad_norm": 0.2568957805633545, "learning_rate": 9.668750930592943e-05, "loss": 0.0201, "step": 16070 }, { "grad_norm": 0.3763359487056732, "learning_rate": 9.668158856478775e-05, "loss": 0.0261, "step": 16080 }, { "grad_norm": 0.30905869603157043, "learning_rate": 9.66756627186177e-05, "loss": 0.024, "step": 16090 }, { "grad_norm": 0.29408374428749084, "learning_rate": 9.666973176806737e-05, "loss": 0.0279, "step": 16100 }, { "grad_norm": 0.3388415575027466, "learning_rate": 9.666379571378534e-05, "loss": 0.0222, "step": 16110 }, { "grad_norm": 0.3603639602661133, "learning_rate": 9.665785455642076e-05, "loss": 0.0236, "step": 16120 }, { "grad_norm": 0.23763497173786163, "learning_rate": 9.665190829662337e-05, "loss": 0.0225, "step": 16130 }, { "grad_norm": 0.9277858138084412, "learning_rate": 9.664595693504342e-05, "loss": 0.0316, "step": 16140 }, { "grad_norm": 1.2661608457565308, "learning_rate": 9.664000047233175e-05, "loss": 0.0314, "step": 16150 }, { "grad_norm": 0.8028908967971802, "learning_rate": 9.663403890913976e-05, "loss": 0.0347, "step": 16160 }, { "grad_norm": 0.2976851165294647, "learning_rate": 9.662807224611938e-05, "loss": 0.0238, "step": 16170 }, { "grad_norm": 0.37158340215682983, "learning_rate": 9.662210048392311e-05, "loss": 0.0295, "step": 16180 }, { "grad_norm": 0.3612216114997864, "learning_rate": 9.661612362320405e-05, "loss": 0.0268, "step": 16190 }, { "grad_norm": 1.3044230937957764, "learning_rate": 9.661014166461579e-05, "loss": 0.0381, "step": 16200 }, { "grad_norm": 0.7124685049057007, "learning_rate": 9.66041546088125e-05, "loss": 0.0283, "step": 16210 }, { "grad_norm": 0.34938502311706543, "learning_rate": 9.659816245644895e-05, "loss": 0.0304, "step": 16220 }, { "grad_norm": 0.3942773640155792, "learning_rate": 9.65921652081804e-05, "loss": 0.0288, "step": 16230 }, { "grad_norm": 0.5756297707557678, "learning_rate": 9.658616286466271e-05, "loss": 0.0325, "step": 16240 }, { "grad_norm": 0.3941062390804291, "learning_rate": 9.65801554265523e-05, "loss": 0.0336, "step": 16250 }, { "grad_norm": 0.306449294090271, "learning_rate": 9.657414289450612e-05, "loss": 0.0381, "step": 16260 }, { "grad_norm": 0.2832280993461609, "learning_rate": 9.656812526918171e-05, "loss": 0.0231, "step": 16270 }, { "grad_norm": 1.1828715801239014, "learning_rate": 9.656210255123712e-05, "loss": 0.0303, "step": 16280 }, { "grad_norm": 0.49015623331069946, "learning_rate": 9.6556074741331e-05, "loss": 0.0323, "step": 16290 }, { "grad_norm": 0.39608487486839294, "learning_rate": 9.655004184012256e-05, "loss": 0.0313, "step": 16300 }, { "grad_norm": 0.28935328125953674, "learning_rate": 9.654400384827152e-05, "loss": 0.0299, "step": 16310 }, { "grad_norm": 0.2531849145889282, "learning_rate": 9.653796076643818e-05, "loss": 0.0315, "step": 16320 }, { "grad_norm": 0.33224064111709595, "learning_rate": 9.653191259528344e-05, "loss": 0.0366, "step": 16330 }, { "grad_norm": 5.275545597076416, "learning_rate": 9.65258593354687e-05, "loss": 0.031, "step": 16340 }, { "grad_norm": 0.2891685962677002, "learning_rate": 9.651980098765591e-05, "loss": 0.0241, "step": 16350 }, { "grad_norm": 0.2782089114189148, "learning_rate": 9.651373755250765e-05, "loss": 0.0242, "step": 16360 }, { "grad_norm": 0.49873119592666626, "learning_rate": 9.650766903068697e-05, "loss": 0.0253, "step": 16370 }, { "grad_norm": 0.3453041613101959, "learning_rate": 9.650159542285753e-05, "loss": 0.0273, "step": 16380 }, { "grad_norm": 0.26636937260627747, "learning_rate": 9.649551672968353e-05, "loss": 0.0239, "step": 16390 }, { "grad_norm": 0.3260734975337982, "learning_rate": 9.648943295182973e-05, "loss": 0.0229, "step": 16400 }, { "grad_norm": 0.34048619866371155, "learning_rate": 9.648334408996144e-05, "loss": 0.0279, "step": 16410 }, { "grad_norm": 0.37516751885414124, "learning_rate": 9.647725014474452e-05, "loss": 0.0215, "step": 16420 }, { "grad_norm": 0.20594127476215363, "learning_rate": 9.64711511168454e-05, "loss": 0.0295, "step": 16430 }, { "grad_norm": 0.7700263261795044, "learning_rate": 9.646504700693108e-05, "loss": 0.0231, "step": 16440 }, { "grad_norm": 0.2647849917411804, "learning_rate": 9.645893781566907e-05, "loss": 0.0201, "step": 16450 }, { "grad_norm": 0.45899537205696106, "learning_rate": 9.645282354372744e-05, "loss": 0.0363, "step": 16460 }, { "grad_norm": 0.4135955572128296, "learning_rate": 9.644670419177491e-05, "loss": 0.0263, "step": 16470 }, { "grad_norm": 0.2655588686466217, "learning_rate": 9.644057976048062e-05, "loss": 0.0359, "step": 16480 }, { "grad_norm": 0.31168463826179504, "learning_rate": 9.643445025051435e-05, "loss": 0.0289, "step": 16490 }, { "grad_norm": 0.22963719069957733, "learning_rate": 9.642831566254641e-05, "loss": 0.0302, "step": 16500 }, { "grad_norm": 0.24751153588294983, "learning_rate": 9.642217599724769e-05, "loss": 0.0212, "step": 16510 }, { "grad_norm": 0.31521815061569214, "learning_rate": 9.64160312552896e-05, "loss": 0.0249, "step": 16520 }, { "grad_norm": 0.36146438121795654, "learning_rate": 9.64098814373441e-05, "loss": 0.0252, "step": 16530 }, { "grad_norm": 1.0244265794754028, "learning_rate": 9.640372654408374e-05, "loss": 0.0225, "step": 16540 }, { "grad_norm": 0.25911641120910645, "learning_rate": 9.639756657618162e-05, "loss": 0.026, "step": 16550 }, { "grad_norm": 0.24471645057201385, "learning_rate": 9.639140153431138e-05, "loss": 0.0205, "step": 16560 }, { "grad_norm": 0.23258718848228455, "learning_rate": 9.638523141914721e-05, "loss": 0.0187, "step": 16570 }, { "grad_norm": 1.932033896446228, "learning_rate": 9.637905623136388e-05, "loss": 0.0256, "step": 16580 }, { "grad_norm": 0.26685962080955505, "learning_rate": 9.637287597163669e-05, "loss": 0.0218, "step": 16590 }, { "grad_norm": 0.30464813113212585, "learning_rate": 9.63666906406415e-05, "loss": 0.018, "step": 16600 }, { "grad_norm": 0.25194573402404785, "learning_rate": 9.636050023905473e-05, "loss": 0.0215, "step": 16610 }, { "grad_norm": 0.3350650370121002, "learning_rate": 9.635430476755336e-05, "loss": 0.021, "step": 16620 }, { "grad_norm": 0.2280234545469284, "learning_rate": 9.63481042268149e-05, "loss": 0.0198, "step": 16630 }, { "grad_norm": 0.2562885284423828, "learning_rate": 9.634189861751745e-05, "loss": 0.0193, "step": 16640 }, { "grad_norm": 0.24804653227329254, "learning_rate": 9.633568794033967e-05, "loss": 0.0224, "step": 16650 }, { "grad_norm": 1.8840179443359375, "learning_rate": 9.63294721959607e-05, "loss": 0.0198, "step": 16660 }, { "grad_norm": 0.30624401569366455, "learning_rate": 9.63232513850603e-05, "loss": 0.0204, "step": 16670 }, { "grad_norm": 0.21969784796237946, "learning_rate": 9.631702550831878e-05, "loss": 0.0207, "step": 16680 }, { "grad_norm": 0.2831776738166809, "learning_rate": 9.631079456641698e-05, "loss": 0.0248, "step": 16690 }, { "grad_norm": 0.34553900361061096, "learning_rate": 9.630455856003632e-05, "loss": 0.022, "step": 16700 }, { "grad_norm": 0.27578169107437134, "learning_rate": 9.629831748985876e-05, "loss": 0.0172, "step": 16710 }, { "grad_norm": 0.4354521334171295, "learning_rate": 9.629207135656679e-05, "loss": 0.0242, "step": 16720 }, { "grad_norm": 0.24992287158966064, "learning_rate": 9.628582016084353e-05, "loss": 0.0241, "step": 16730 }, { "grad_norm": 0.2008163332939148, "learning_rate": 9.627956390337254e-05, "loss": 0.0206, "step": 16740 }, { "grad_norm": 0.26739582419395447, "learning_rate": 9.627330258483802e-05, "loss": 0.0183, "step": 16750 }, { "grad_norm": 0.29411783814430237, "learning_rate": 9.62670362059247e-05, "loss": 0.0197, "step": 16760 }, { "grad_norm": 0.3139793574810028, "learning_rate": 9.626076476731786e-05, "loss": 0.0227, "step": 16770 }, { "grad_norm": 0.23665973544120789, "learning_rate": 9.625448826970336e-05, "loss": 0.0202, "step": 16780 }, { "grad_norm": 0.5765668749809265, "learning_rate": 9.624820671376755e-05, "loss": 0.0209, "step": 16790 }, { "grad_norm": 0.35276323556900024, "learning_rate": 9.62419201001974e-05, "loss": 0.0255, "step": 16800 }, { "grad_norm": 0.27548304200172424, "learning_rate": 9.623562842968037e-05, "loss": 0.0283, "step": 16810 }, { "grad_norm": 1.0914605855941772, "learning_rate": 9.622933170290454e-05, "loss": 0.0234, "step": 16820 }, { "grad_norm": 0.28097793459892273, "learning_rate": 9.622302992055849e-05, "loss": 0.0271, "step": 16830 }, { "grad_norm": 0.2844332456588745, "learning_rate": 9.62167230833314e-05, "loss": 0.0241, "step": 16840 }, { "grad_norm": 0.2729126214981079, "learning_rate": 9.621041119191295e-05, "loss": 0.0236, "step": 16850 }, { "grad_norm": 0.23929959535598755, "learning_rate": 9.620409424699342e-05, "loss": 0.0249, "step": 16860 }, { "grad_norm": 0.23889949917793274, "learning_rate": 9.619777224926359e-05, "loss": 0.0292, "step": 16870 }, { "grad_norm": 0.31957393884658813, "learning_rate": 9.619144519941485e-05, "loss": 0.0279, "step": 16880 }, { "grad_norm": 0.31436270475387573, "learning_rate": 9.618511309813912e-05, "loss": 0.0271, "step": 16890 }, { "grad_norm": 0.3298420011997223, "learning_rate": 9.617877594612886e-05, "loss": 0.0274, "step": 16900 }, { "grad_norm": 0.2879877984523773, "learning_rate": 9.617243374407707e-05, "loss": 0.0255, "step": 16910 }, { "grad_norm": 0.310416579246521, "learning_rate": 9.616608649267736e-05, "loss": 0.0209, "step": 16920 }, { "grad_norm": 0.27150702476501465, "learning_rate": 9.615973419262385e-05, "loss": 0.0211, "step": 16930 }, { "grad_norm": 0.27922967076301575, "learning_rate": 9.615337684461119e-05, "loss": 0.023, "step": 16940 }, { "grad_norm": 0.27990368008613586, "learning_rate": 9.614701444933465e-05, "loss": 0.0202, "step": 16950 }, { "grad_norm": 0.3294745981693268, "learning_rate": 9.614064700748997e-05, "loss": 0.0243, "step": 16960 }, { "grad_norm": 0.34377408027648926, "learning_rate": 9.613427451977352e-05, "loss": 0.0299, "step": 16970 }, { "grad_norm": 0.39955857396125793, "learning_rate": 9.612789698688216e-05, "loss": 0.0265, "step": 16980 }, { "grad_norm": 0.4865230321884155, "learning_rate": 9.612151440951334e-05, "loss": 0.0318, "step": 16990 }, { "grad_norm": 0.30490759015083313, "learning_rate": 9.611512678836506e-05, "loss": 0.0289, "step": 17000 }, { "grad_norm": 0.30359116196632385, "learning_rate": 9.610873412413584e-05, "loss": 0.0231, "step": 17010 }, { "grad_norm": 0.3520066440105438, "learning_rate": 9.610233641752476e-05, "loss": 0.0273, "step": 17020 }, { "grad_norm": 0.3952033221721649, "learning_rate": 9.609593366923151e-05, "loss": 0.0275, "step": 17030 }, { "grad_norm": 0.541163444519043, "learning_rate": 9.608952587995625e-05, "loss": 0.0265, "step": 17040 }, { "grad_norm": 0.39672306180000305, "learning_rate": 9.608311305039972e-05, "loss": 0.0248, "step": 17050 }, { "grad_norm": 0.6942551732063293, "learning_rate": 9.607669518126326e-05, "loss": 0.0192, "step": 17060 }, { "grad_norm": 0.30272844433784485, "learning_rate": 9.607027227324866e-05, "loss": 0.0222, "step": 17070 }, { "grad_norm": 0.43073970079421997, "learning_rate": 9.606384432705837e-05, "loss": 0.0265, "step": 17080 }, { "grad_norm": 0.23466549813747406, "learning_rate": 9.60574113433953e-05, "loss": 0.0208, "step": 17090 }, { "grad_norm": 0.21024467051029205, "learning_rate": 9.6050973322963e-05, "loss": 0.0219, "step": 17100 }, { "grad_norm": 0.3898206055164337, "learning_rate": 9.604453026646547e-05, "loss": 0.025, "step": 17110 }, { "grad_norm": 0.3978841006755829, "learning_rate": 9.603808217460735e-05, "loss": 0.0225, "step": 17120 }, { "grad_norm": 0.22569197416305542, "learning_rate": 9.603162904809377e-05, "loss": 0.0232, "step": 17130 }, { "grad_norm": 0.21507467329502106, "learning_rate": 9.602517088763045e-05, "loss": 0.0243, "step": 17140 }, { "grad_norm": 0.2805921733379364, "learning_rate": 9.601870769392365e-05, "loss": 0.0195, "step": 17150 }, { "grad_norm": 0.578444242477417, "learning_rate": 9.601223946768017e-05, "loss": 0.0208, "step": 17160 }, { "grad_norm": 0.38047873973846436, "learning_rate": 9.600576620960734e-05, "loss": 0.0214, "step": 17170 }, { "grad_norm": 0.30812233686447144, "learning_rate": 9.599928792041308e-05, "loss": 0.0252, "step": 17180 }, { "grad_norm": 0.6801915764808655, "learning_rate": 9.599280460080587e-05, "loss": 0.023, "step": 17190 }, { "grad_norm": 0.5400505661964417, "learning_rate": 9.59863162514947e-05, "loss": 0.0232, "step": 17200 }, { "grad_norm": 0.4090701639652252, "learning_rate": 9.597982287318911e-05, "loss": 0.028, "step": 17210 }, { "grad_norm": 0.2774084806442261, "learning_rate": 9.597332446659923e-05, "loss": 0.0272, "step": 17220 }, { "grad_norm": 0.2578655183315277, "learning_rate": 9.59668210324357e-05, "loss": 0.0288, "step": 17230 }, { "grad_norm": 0.39721009135246277, "learning_rate": 9.596031257140974e-05, "loss": 0.0277, "step": 17240 }, { "grad_norm": 0.31114089488983154, "learning_rate": 9.59537990842331e-05, "loss": 0.0253, "step": 17250 }, { "grad_norm": 0.291720449924469, "learning_rate": 9.594728057161806e-05, "loss": 0.0241, "step": 17260 }, { "grad_norm": 0.2671871781349182, "learning_rate": 9.594075703427752e-05, "loss": 0.0219, "step": 17270 }, { "grad_norm": 0.2204665094614029, "learning_rate": 9.593422847292486e-05, "loss": 0.0212, "step": 17280 }, { "grad_norm": 0.24244993925094604, "learning_rate": 9.592769488827402e-05, "loss": 0.0198, "step": 17290 }, { "grad_norm": 0.21807466447353363, "learning_rate": 9.592115628103952e-05, "loss": 0.0234, "step": 17300 }, { "grad_norm": 0.3316693603992462, "learning_rate": 9.591461265193643e-05, "loss": 0.0269, "step": 17310 }, { "grad_norm": 0.22565330564975739, "learning_rate": 9.590806400168032e-05, "loss": 0.0188, "step": 17320 }, { "grad_norm": 0.26623600721359253, "learning_rate": 9.590151033098735e-05, "loss": 0.0309, "step": 17330 }, { "grad_norm": 0.5869492292404175, "learning_rate": 9.589495164057423e-05, "loss": 0.03, "step": 17340 }, { "grad_norm": 0.289720743894577, "learning_rate": 9.58883879311582e-05, "loss": 0.0224, "step": 17350 }, { "grad_norm": 0.20555968582630157, "learning_rate": 9.588181920345705e-05, "loss": 0.0245, "step": 17360 }, { "grad_norm": 0.20359717309474945, "learning_rate": 9.587524545818913e-05, "loss": 0.019, "step": 17370 }, { "grad_norm": 0.4275858700275421, "learning_rate": 9.586866669607335e-05, "loss": 0.0207, "step": 17380 }, { "grad_norm": 0.2650333642959595, "learning_rate": 9.586208291782915e-05, "loss": 0.0209, "step": 17390 }, { "grad_norm": 0.2586217522621155, "learning_rate": 9.58554941241765e-05, "loss": 0.018, "step": 17400 }, { "grad_norm": 0.2902916967868805, "learning_rate": 9.584890031583596e-05, "loss": 0.0226, "step": 17410 }, { "grad_norm": 0.370688796043396, "learning_rate": 9.584230149352861e-05, "loss": 0.0208, "step": 17420 }, { "grad_norm": 0.32889971137046814, "learning_rate": 9.58356976579761e-05, "loss": 0.0251, "step": 17430 }, { "grad_norm": 0.27152588963508606, "learning_rate": 9.58290888099006e-05, "loss": 0.0214, "step": 17440 }, { "grad_norm": 0.2943321466445923, "learning_rate": 9.582247495002486e-05, "loss": 0.0193, "step": 17450 }, { "grad_norm": 0.2823135554790497, "learning_rate": 9.581585607907214e-05, "loss": 0.0216, "step": 17460 }, { "grad_norm": 0.3593665063381195, "learning_rate": 9.580923219776628e-05, "loss": 0.0251, "step": 17470 }, { "grad_norm": 0.2802600860595703, "learning_rate": 9.580260330683167e-05, "loss": 0.0211, "step": 17480 }, { "grad_norm": 0.3910055458545685, "learning_rate": 9.579596940699322e-05, "loss": 0.0209, "step": 17490 }, { "grad_norm": 0.2125251740217209, "learning_rate": 9.578933049897643e-05, "loss": 0.0311, "step": 17500 }, { "grad_norm": 0.22639618813991547, "learning_rate": 9.578268658350728e-05, "loss": 0.0239, "step": 17510 }, { "grad_norm": 0.24069362878799438, "learning_rate": 9.577603766131235e-05, "loss": 0.0202, "step": 17520 }, { "grad_norm": 0.4093543589115143, "learning_rate": 9.576938373311878e-05, "loss": 0.0242, "step": 17530 }, { "grad_norm": 0.30122438073158264, "learning_rate": 9.576272479965421e-05, "loss": 0.0281, "step": 17540 }, { "grad_norm": 0.2869948446750641, "learning_rate": 9.575606086164687e-05, "loss": 0.0254, "step": 17550 }, { "grad_norm": 0.3192873001098633, "learning_rate": 9.57493919198255e-05, "loss": 0.0273, "step": 17560 }, { "grad_norm": 0.3216192126274109, "learning_rate": 9.57427179749194e-05, "loss": 0.027, "step": 17570 }, { "grad_norm": 0.6479997634887695, "learning_rate": 9.573603902765846e-05, "loss": 0.0279, "step": 17580 }, { "grad_norm": 0.3436654806137085, "learning_rate": 9.572935507877304e-05, "loss": 0.0314, "step": 17590 }, { "grad_norm": 0.38634687662124634, "learning_rate": 9.57226661289941e-05, "loss": 0.0251, "step": 17600 }, { "grad_norm": 0.30729609727859497, "learning_rate": 9.571597217905315e-05, "loss": 0.0246, "step": 17610 }, { "grad_norm": 0.47010549902915955, "learning_rate": 9.57092732296822e-05, "loss": 0.0304, "step": 17620 }, { "grad_norm": 0.30494067072868347, "learning_rate": 9.570256928161385e-05, "loss": 0.0236, "step": 17630 }, { "grad_norm": 2.7865164279937744, "learning_rate": 9.569586033558126e-05, "loss": 0.024, "step": 17640 }, { "grad_norm": 1.1409285068511963, "learning_rate": 9.568914639231807e-05, "loss": 0.0236, "step": 17650 }, { "grad_norm": 0.349328875541687, "learning_rate": 9.568242745255852e-05, "loss": 0.0358, "step": 17660 }, { "grad_norm": 0.3215290606021881, "learning_rate": 9.567570351703739e-05, "loss": 0.0291, "step": 17670 }, { "grad_norm": 0.2957064211368561, "learning_rate": 9.566897458649001e-05, "loss": 0.028, "step": 17680 }, { "grad_norm": 0.26425251364707947, "learning_rate": 9.566224066165221e-05, "loss": 0.0238, "step": 17690 }, { "grad_norm": 0.3267761468887329, "learning_rate": 9.565550174326043e-05, "loss": 0.0279, "step": 17700 }, { "grad_norm": 0.285247802734375, "learning_rate": 9.564875783205162e-05, "loss": 0.0261, "step": 17710 }, { "grad_norm": 0.42421749234199524, "learning_rate": 9.564200892876328e-05, "loss": 0.0226, "step": 17720 }, { "grad_norm": 0.2371349185705185, "learning_rate": 9.563525503413348e-05, "loss": 0.0242, "step": 17730 }, { "grad_norm": 0.9543958306312561, "learning_rate": 9.562849614890079e-05, "loss": 0.0319, "step": 17740 }, { "grad_norm": 0.35816678404808044, "learning_rate": 9.562173227380436e-05, "loss": 0.0248, "step": 17750 }, { "grad_norm": 0.37372294068336487, "learning_rate": 9.561496340958389e-05, "loss": 0.0356, "step": 17760 }, { "grad_norm": 0.25947830080986023, "learning_rate": 9.560818955697959e-05, "loss": 0.0315, "step": 17770 }, { "grad_norm": 0.6574642658233643, "learning_rate": 9.560141071673228e-05, "loss": 0.0279, "step": 17780 }, { "grad_norm": 0.30652105808258057, "learning_rate": 9.559462688958323e-05, "loss": 0.0255, "step": 17790 }, { "grad_norm": 0.4481315016746521, "learning_rate": 9.558783807627434e-05, "loss": 0.0311, "step": 17800 }, { "grad_norm": 1.9269683361053467, "learning_rate": 9.558104427754801e-05, "loss": 0.0241, "step": 17810 }, { "grad_norm": 0.4064837098121643, "learning_rate": 9.557424549414722e-05, "loss": 0.0275, "step": 17820 }, { "grad_norm": 0.3115449845790863, "learning_rate": 9.556744172681546e-05, "loss": 0.0255, "step": 17830 }, { "grad_norm": 0.39283114671707153, "learning_rate": 9.556063297629677e-05, "loss": 0.0305, "step": 17840 }, { "grad_norm": 0.25051650404930115, "learning_rate": 9.555381924333578e-05, "loss": 0.0286, "step": 17850 }, { "grad_norm": 0.33087247610092163, "learning_rate": 9.554700052867758e-05, "loss": 0.0226, "step": 17860 }, { "grad_norm": 2.48012113571167, "learning_rate": 9.554017683306789e-05, "loss": 0.0229, "step": 17870 }, { "grad_norm": 0.33411920070648193, "learning_rate": 9.553334815725294e-05, "loss": 0.0241, "step": 17880 }, { "grad_norm": 0.255450963973999, "learning_rate": 9.552651450197949e-05, "loss": 0.0205, "step": 17890 }, { "grad_norm": 0.31666040420532227, "learning_rate": 9.551967586799486e-05, "loss": 0.0235, "step": 17900 }, { "grad_norm": 0.3213212490081787, "learning_rate": 9.551283225604692e-05, "loss": 0.0234, "step": 17910 }, { "grad_norm": 0.4116206765174866, "learning_rate": 9.550598366688406e-05, "loss": 0.026, "step": 17920 }, { "grad_norm": 0.20597022771835327, "learning_rate": 9.549913010125526e-05, "loss": 0.0248, "step": 17930 }, { "grad_norm": 0.3444312810897827, "learning_rate": 9.549227155990999e-05, "loss": 0.0282, "step": 17940 }, { "grad_norm": 0.2651103436946869, "learning_rate": 9.548540804359828e-05, "loss": 0.0216, "step": 17950 }, { "grad_norm": 0.4452589154243469, "learning_rate": 9.547853955307077e-05, "loss": 0.0275, "step": 17960 }, { "grad_norm": 0.4534446597099304, "learning_rate": 9.547166608907853e-05, "loss": 0.0315, "step": 17970 }, { "grad_norm": 0.4417286217212677, "learning_rate": 9.546478765237326e-05, "loss": 0.0329, "step": 17980 }, { "grad_norm": 0.27157649397850037, "learning_rate": 9.545790424370715e-05, "loss": 0.0262, "step": 17990 }, { "grad_norm": 0.3216375410556793, "learning_rate": 9.5451015863833e-05, "loss": 0.0423, "step": 18000 }, { "grad_norm": 0.7235367298126221, "learning_rate": 9.544412251350408e-05, "loss": 0.0335, "step": 18010 }, { "grad_norm": 0.2746730446815491, "learning_rate": 9.543722419347422e-05, "loss": 0.0302, "step": 18020 }, { "grad_norm": 0.24195393919944763, "learning_rate": 9.543032090449788e-05, "loss": 0.0263, "step": 18030 }, { "grad_norm": 0.330333411693573, "learning_rate": 9.542341264732992e-05, "loss": 0.0254, "step": 18040 }, { "grad_norm": 0.21633045375347137, "learning_rate": 9.541649942272585e-05, "loss": 0.0234, "step": 18050 }, { "grad_norm": 0.3061383068561554, "learning_rate": 9.54095812314417e-05, "loss": 0.0263, "step": 18060 }, { "grad_norm": 0.4142245948314667, "learning_rate": 9.540265807423401e-05, "loss": 0.0274, "step": 18070 }, { "grad_norm": 0.29460838437080383, "learning_rate": 9.53957299518599e-05, "loss": 0.0229, "step": 18080 }, { "grad_norm": 0.2508867084980011, "learning_rate": 9.5388796865077e-05, "loss": 0.0231, "step": 18090 }, { "grad_norm": 0.2631058394908905, "learning_rate": 9.538185881464353e-05, "loss": 0.0266, "step": 18100 }, { "grad_norm": 0.2352064698934555, "learning_rate": 9.537491580131821e-05, "loss": 0.0208, "step": 18110 }, { "grad_norm": 0.2634128928184509, "learning_rate": 9.53679678258603e-05, "loss": 0.027, "step": 18120 }, { "grad_norm": 0.6607298851013184, "learning_rate": 9.536101488902966e-05, "loss": 0.0243, "step": 18130 }, { "grad_norm": 0.48102396726608276, "learning_rate": 9.535405699158663e-05, "loss": 0.0235, "step": 18140 }, { "grad_norm": 0.29422837495803833, "learning_rate": 9.53470941342921e-05, "loss": 0.0262, "step": 18150 }, { "grad_norm": 0.38031867146492004, "learning_rate": 9.534012631790756e-05, "loss": 0.0247, "step": 18160 }, { "grad_norm": 0.2841142416000366, "learning_rate": 9.533315354319494e-05, "loss": 0.0263, "step": 18170 }, { "grad_norm": 0.44450464844703674, "learning_rate": 9.532617581091682e-05, "loss": 0.0341, "step": 18180 }, { "grad_norm": 0.3156808018684387, "learning_rate": 9.531919312183629e-05, "loss": 0.0221, "step": 18190 }, { "grad_norm": 0.2774066627025604, "learning_rate": 9.531220547671688e-05, "loss": 0.0248, "step": 18200 }, { "grad_norm": 0.25578463077545166, "learning_rate": 9.530521287632285e-05, "loss": 0.0179, "step": 18210 }, { "grad_norm": 0.4101707935333252, "learning_rate": 9.529821532141884e-05, "loss": 0.0253, "step": 18220 }, { "grad_norm": 0.28938764333724976, "learning_rate": 9.52912128127701e-05, "loss": 0.0318, "step": 18230 }, { "grad_norm": 0.2900889813899994, "learning_rate": 9.528420535114244e-05, "loss": 0.0241, "step": 18240 }, { "grad_norm": 0.33911648392677307, "learning_rate": 9.527719293730215e-05, "loss": 0.0214, "step": 18250 }, { "grad_norm": 0.23854020237922668, "learning_rate": 9.527017557201611e-05, "loss": 0.0255, "step": 18260 }, { "grad_norm": 0.23821808397769928, "learning_rate": 9.526315325605176e-05, "loss": 0.0209, "step": 18270 }, { "grad_norm": 0.2454766184091568, "learning_rate": 9.525612599017699e-05, "loss": 0.0226, "step": 18280 }, { "grad_norm": 0.35074958205223083, "learning_rate": 9.524909377516033e-05, "loss": 0.0248, "step": 18290 }, { "grad_norm": 0.3226265609264374, "learning_rate": 9.524205661177081e-05, "loss": 0.0244, "step": 18300 }, { "grad_norm": 0.26965129375457764, "learning_rate": 9.523501450077801e-05, "loss": 0.0207, "step": 18310 }, { "grad_norm": 0.26851651072502136, "learning_rate": 9.522796744295202e-05, "loss": 0.0154, "step": 18320 }, { "grad_norm": 0.24482351541519165, "learning_rate": 9.522091543906352e-05, "loss": 0.0226, "step": 18330 }, { "grad_norm": 0.2850465178489685, "learning_rate": 9.521385848988369e-05, "loss": 0.019, "step": 18340 }, { "grad_norm": 0.2919328212738037, "learning_rate": 9.520679659618428e-05, "loss": 0.0182, "step": 18350 }, { "grad_norm": 0.296440988779068, "learning_rate": 9.519972975873754e-05, "loss": 0.0197, "step": 18360 }, { "grad_norm": 1.2537482976913452, "learning_rate": 9.519265797831633e-05, "loss": 0.026, "step": 18370 }, { "grad_norm": 0.34484443068504333, "learning_rate": 9.518558125569399e-05, "loss": 0.0219, "step": 18380 }, { "grad_norm": 0.8234977126121521, "learning_rate": 9.517849959164442e-05, "loss": 0.0267, "step": 18390 }, { "grad_norm": 0.24787335097789764, "learning_rate": 9.517141298694205e-05, "loss": 0.0269, "step": 18400 }, { "grad_norm": 0.25152137875556946, "learning_rate": 9.516432144236188e-05, "loss": 0.0259, "step": 18410 }, { "grad_norm": 0.31702694296836853, "learning_rate": 9.515722495867941e-05, "loss": 0.0246, "step": 18420 }, { "grad_norm": 0.3048168420791626, "learning_rate": 9.515012353667072e-05, "loss": 0.0225, "step": 18430 }, { "grad_norm": 0.32746171951293945, "learning_rate": 9.51430171771124e-05, "loss": 0.025, "step": 18440 }, { "grad_norm": 0.27987951040267944, "learning_rate": 9.513590588078159e-05, "loss": 0.028, "step": 18450 }, { "grad_norm": 0.28585535287857056, "learning_rate": 9.512878964845597e-05, "loss": 0.0315, "step": 18460 }, { "grad_norm": 0.3153407573699951, "learning_rate": 9.512166848091377e-05, "loss": 0.0252, "step": 18470 }, { "grad_norm": 0.3161531984806061, "learning_rate": 9.511454237893376e-05, "loss": 0.0208, "step": 18480 }, { "grad_norm": 0.33237388730049133, "learning_rate": 9.51074113432952e-05, "loss": 0.0282, "step": 18490 }, { "grad_norm": 0.30680349469184875, "learning_rate": 9.510027537477797e-05, "loss": 0.0264, "step": 18500 }, { "grad_norm": 0.2950713038444519, "learning_rate": 9.509313447416242e-05, "loss": 0.0274, "step": 18510 }, { "grad_norm": 0.31435173749923706, "learning_rate": 9.508598864222949e-05, "loss": 0.0269, "step": 18520 }, { "grad_norm": 0.2659701406955719, "learning_rate": 9.507883787976062e-05, "loss": 0.0263, "step": 18530 }, { "grad_norm": 0.2927135229110718, "learning_rate": 9.507168218753781e-05, "loss": 0.0248, "step": 18540 }, { "grad_norm": 0.2939007580280304, "learning_rate": 9.506452156634362e-05, "loss": 0.0254, "step": 18550 }, { "grad_norm": 0.36583027243614197, "learning_rate": 9.505735601696109e-05, "loss": 0.0253, "step": 18560 }, { "grad_norm": 0.30402421951293945, "learning_rate": 9.505018554017385e-05, "loss": 0.0237, "step": 18570 }, { "grad_norm": 0.4066348671913147, "learning_rate": 9.504301013676604e-05, "loss": 0.0338, "step": 18580 }, { "grad_norm": 0.2738422453403473, "learning_rate": 9.503582980752238e-05, "loss": 0.0282, "step": 18590 }, { "grad_norm": 0.3853197991847992, "learning_rate": 9.502864455322809e-05, "loss": 0.0266, "step": 18600 }, { "grad_norm": 0.33863067626953125, "learning_rate": 9.502145437466891e-05, "loss": 0.0273, "step": 18610 }, { "grad_norm": 0.3418119251728058, "learning_rate": 9.501425927263116e-05, "loss": 0.0235, "step": 18620 }, { "grad_norm": 0.29215773940086365, "learning_rate": 9.500705924790172e-05, "loss": 0.0216, "step": 18630 }, { "grad_norm": 0.29165562987327576, "learning_rate": 9.499985430126794e-05, "loss": 0.0242, "step": 18640 }, { "grad_norm": 0.28203603625297546, "learning_rate": 9.499264443351775e-05, "loss": 0.0263, "step": 18650 }, { "grad_norm": 0.33807188272476196, "learning_rate": 9.498542964543961e-05, "loss": 0.0265, "step": 18660 }, { "grad_norm": 0.37643176317214966, "learning_rate": 9.497820993782252e-05, "loss": 0.042, "step": 18670 }, { "grad_norm": 1.163538932800293, "learning_rate": 9.497098531145601e-05, "loss": 0.0295, "step": 18680 }, { "grad_norm": 0.8247238397598267, "learning_rate": 9.496375576713017e-05, "loss": 0.0277, "step": 18690 }, { "grad_norm": 0.29493260383605957, "learning_rate": 9.49565213056356e-05, "loss": 0.0286, "step": 18700 }, { "grad_norm": 0.2855359613895416, "learning_rate": 9.494928192776342e-05, "loss": 0.0225, "step": 18710 }, { "grad_norm": 0.6115129590034485, "learning_rate": 9.494203763430538e-05, "loss": 0.0229, "step": 18720 }, { "grad_norm": 3.458172082901001, "learning_rate": 9.493478842605366e-05, "loss": 0.0252, "step": 18730 }, { "grad_norm": 0.26802611351013184, "learning_rate": 9.492753430380105e-05, "loss": 0.0228, "step": 18740 }, { "grad_norm": 0.234890878200531, "learning_rate": 9.492027526834083e-05, "loss": 0.0231, "step": 18750 }, { "grad_norm": 1.1123547554016113, "learning_rate": 9.491301132046684e-05, "loss": 0.0226, "step": 18760 }, { "grad_norm": 0.32461872696876526, "learning_rate": 9.490574246097345e-05, "loss": 0.0227, "step": 18770 }, { "grad_norm": 0.4877738058567047, "learning_rate": 9.48984686906556e-05, "loss": 0.0261, "step": 18780 }, { "grad_norm": 0.2534368336200714, "learning_rate": 9.489119001030871e-05, "loss": 0.0239, "step": 18790 }, { "grad_norm": 0.2989027202129364, "learning_rate": 9.488390642072878e-05, "loss": 0.0219, "step": 18800 }, { "grad_norm": 0.36575138568878174, "learning_rate": 9.48766179227123e-05, "loss": 0.0218, "step": 18810 }, { "grad_norm": 0.18737328052520752, "learning_rate": 9.486932451705636e-05, "loss": 0.0199, "step": 18820 }, { "grad_norm": 0.2078290432691574, "learning_rate": 9.486202620455857e-05, "loss": 0.0208, "step": 18830 }, { "grad_norm": 0.24152898788452148, "learning_rate": 9.485472298601704e-05, "loss": 0.0165, "step": 18840 }, { "grad_norm": 0.3209684193134308, "learning_rate": 9.484741486223043e-05, "loss": 0.0255, "step": 18850 }, { "grad_norm": 0.3225058615207672, "learning_rate": 9.484010183399797e-05, "loss": 0.0198, "step": 18860 }, { "grad_norm": 0.25418123602867126, "learning_rate": 9.483278390211938e-05, "loss": 0.0244, "step": 18870 }, { "grad_norm": 0.28569602966308594, "learning_rate": 9.482546106739496e-05, "loss": 0.0213, "step": 18880 }, { "grad_norm": 0.37583237886428833, "learning_rate": 9.48181333306255e-05, "loss": 0.0187, "step": 18890 }, { "grad_norm": 0.27763429284095764, "learning_rate": 9.481080069261237e-05, "loss": 0.0206, "step": 18900 }, { "grad_norm": 0.2886691987514496, "learning_rate": 9.480346315415745e-05, "loss": 0.0177, "step": 18910 }, { "grad_norm": 0.42735448479652405, "learning_rate": 9.479612071606314e-05, "loss": 0.0265, "step": 18920 }, { "grad_norm": 0.28576043248176575, "learning_rate": 9.478877337913244e-05, "loss": 0.0275, "step": 18930 }, { "grad_norm": 0.2698460817337036, "learning_rate": 9.478142114416881e-05, "loss": 0.0235, "step": 18940 }, { "grad_norm": 0.3371453285217285, "learning_rate": 9.47740640119763e-05, "loss": 0.0221, "step": 18950 }, { "grad_norm": 0.29136544466018677, "learning_rate": 9.476670198335947e-05, "loss": 0.0246, "step": 18960 }, { "grad_norm": 0.27292293310165405, "learning_rate": 9.47593350591234e-05, "loss": 0.0232, "step": 18970 }, { "grad_norm": 0.33777809143066406, "learning_rate": 9.475196324007376e-05, "loss": 0.0209, "step": 18980 }, { "grad_norm": 0.33522599935531616, "learning_rate": 9.474458652701669e-05, "loss": 0.0229, "step": 18990 }, { "grad_norm": 0.38738396763801575, "learning_rate": 9.473720492075892e-05, "loss": 0.0232, "step": 19000 }, { "grad_norm": 0.33137720823287964, "learning_rate": 9.472981842210768e-05, "loss": 0.0209, "step": 19010 }, { "grad_norm": 0.2563808858394623, "learning_rate": 9.472242703187074e-05, "loss": 0.0275, "step": 19020 }, { "grad_norm": 0.2902446985244751, "learning_rate": 9.471503075085643e-05, "loss": 0.0253, "step": 19030 }, { "grad_norm": 0.2823806405067444, "learning_rate": 9.470762957987359e-05, "loss": 0.0274, "step": 19040 }, { "grad_norm": 1.3610204458236694, "learning_rate": 9.470022351973158e-05, "loss": 0.0236, "step": 19050 }, { "grad_norm": 0.49005386233329773, "learning_rate": 9.469281257124034e-05, "loss": 0.024, "step": 19060 }, { "grad_norm": 0.22101213037967682, "learning_rate": 9.46853967352103e-05, "loss": 0.0274, "step": 19070 }, { "grad_norm": 0.337094783782959, "learning_rate": 9.467797601245246e-05, "loss": 0.0226, "step": 19080 }, { "grad_norm": 0.26385074853897095, "learning_rate": 9.467055040377834e-05, "loss": 0.0206, "step": 19090 }, { "grad_norm": 0.2354019433259964, "learning_rate": 9.466311990999999e-05, "loss": 0.0216, "step": 19100 }, { "grad_norm": 0.9434016942977905, "learning_rate": 9.465568453193e-05, "loss": 0.0237, "step": 19110 }, { "grad_norm": 0.3695833086967468, "learning_rate": 9.464824427038148e-05, "loss": 0.0212, "step": 19120 }, { "grad_norm": 0.35787227749824524, "learning_rate": 9.46407991261681e-05, "loss": 0.0234, "step": 19130 }, { "grad_norm": 0.27727755904197693, "learning_rate": 9.463334910010404e-05, "loss": 0.0222, "step": 19140 }, { "grad_norm": 0.28712737560272217, "learning_rate": 9.462589419300403e-05, "loss": 0.0179, "step": 19150 }, { "grad_norm": 0.6693757176399231, "learning_rate": 9.461843440568333e-05, "loss": 0.0272, "step": 19160 }, { "grad_norm": 0.23109100759029388, "learning_rate": 9.461096973895773e-05, "loss": 0.0228, "step": 19170 }, { "grad_norm": 0.3509211242198944, "learning_rate": 9.460350019364355e-05, "loss": 0.0198, "step": 19180 }, { "grad_norm": 0.2583344578742981, "learning_rate": 9.459602577055764e-05, "loss": 0.0261, "step": 19190 }, { "grad_norm": 0.3638736307621002, "learning_rate": 9.45885464705174e-05, "loss": 0.0234, "step": 19200 }, { "grad_norm": 0.40046340227127075, "learning_rate": 9.458106229434076e-05, "loss": 0.0282, "step": 19210 }, { "grad_norm": 0.3713398575782776, "learning_rate": 9.457357324284617e-05, "loss": 0.0304, "step": 19220 }, { "grad_norm": 0.9457757472991943, "learning_rate": 9.456607931685262e-05, "loss": 0.0271, "step": 19230 }, { "grad_norm": 0.26367148756980896, "learning_rate": 9.455858051717965e-05, "loss": 0.0253, "step": 19240 }, { "grad_norm": 0.24603374302387238, "learning_rate": 9.45510768446473e-05, "loss": 0.0248, "step": 19250 }, { "grad_norm": 0.33894118666648865, "learning_rate": 9.454356830007618e-05, "loss": 0.0251, "step": 19260 }, { "grad_norm": 0.29482635855674744, "learning_rate": 9.45360548842874e-05, "loss": 0.0224, "step": 19270 }, { "grad_norm": 0.25487810373306274, "learning_rate": 9.452853659810261e-05, "loss": 0.0207, "step": 19280 }, { "grad_norm": 0.2575036883354187, "learning_rate": 9.452101344234401e-05, "loss": 0.0205, "step": 19290 }, { "grad_norm": 0.21558618545532227, "learning_rate": 9.451348541783431e-05, "loss": 0.0218, "step": 19300 }, { "grad_norm": 0.23899617791175842, "learning_rate": 9.450595252539678e-05, "loss": 0.0182, "step": 19310 }, { "grad_norm": 0.2851795554161072, "learning_rate": 9.449841476585518e-05, "loss": 0.0168, "step": 19320 }, { "grad_norm": 0.5252262949943542, "learning_rate": 9.449087214003384e-05, "loss": 0.0201, "step": 19330 }, { "grad_norm": 0.24450582265853882, "learning_rate": 9.448332464875765e-05, "loss": 0.0209, "step": 19340 }, { "grad_norm": 0.23984456062316895, "learning_rate": 9.447577229285192e-05, "loss": 0.0275, "step": 19350 }, { "grad_norm": 0.34670892357826233, "learning_rate": 9.446821507314261e-05, "loss": 0.0287, "step": 19360 }, { "grad_norm": 0.3045893907546997, "learning_rate": 9.446065299045617e-05, "loss": 0.0237, "step": 19370 }, { "grad_norm": 0.34041064977645874, "learning_rate": 9.445308604561955e-05, "loss": 0.0215, "step": 19380 }, { "grad_norm": 0.21406178176403046, "learning_rate": 9.444551423946028e-05, "loss": 0.0243, "step": 19390 }, { "grad_norm": 0.24600757658481598, "learning_rate": 9.443793757280638e-05, "loss": 0.0229, "step": 19400 }, { "grad_norm": 0.27818697690963745, "learning_rate": 9.443035604648646e-05, "loss": 0.0314, "step": 19410 }, { "grad_norm": 0.27389782667160034, "learning_rate": 9.44227696613296e-05, "loss": 0.0212, "step": 19420 }, { "grad_norm": 0.22735479474067688, "learning_rate": 9.441517841816542e-05, "loss": 0.0214, "step": 19430 }, { "grad_norm": 0.22946764528751373, "learning_rate": 9.440758231782413e-05, "loss": 0.0225, "step": 19440 }, { "grad_norm": 0.18549834191799164, "learning_rate": 9.439998136113639e-05, "loss": 0.0192, "step": 19450 }, { "grad_norm": 0.22473646700382233, "learning_rate": 9.439237554893344e-05, "loss": 0.0196, "step": 19460 }, { "grad_norm": 0.2336454540491104, "learning_rate": 9.438476488204705e-05, "loss": 0.0196, "step": 19470 }, { "grad_norm": 0.32243359088897705, "learning_rate": 9.43771493613095e-05, "loss": 0.0276, "step": 19480 }, { "grad_norm": 0.20957736670970917, "learning_rate": 9.436952898755362e-05, "loss": 0.023, "step": 19490 }, { "grad_norm": 0.28684431314468384, "learning_rate": 9.436190376161276e-05, "loss": 0.0226, "step": 19500 }, { "grad_norm": 0.31248798966407776, "learning_rate": 9.43542736843208e-05, "loss": 0.028, "step": 19510 }, { "grad_norm": 0.4187794029712677, "learning_rate": 9.434663875651216e-05, "loss": 0.0238, "step": 19520 }, { "grad_norm": 0.2710784375667572, "learning_rate": 9.433899897902177e-05, "loss": 0.0296, "step": 19530 }, { "grad_norm": 0.19598041474819183, "learning_rate": 9.433135435268511e-05, "loss": 0.0193, "step": 19540 }, { "grad_norm": 0.34209442138671875, "learning_rate": 9.432370487833819e-05, "loss": 0.0263, "step": 19550 }, { "grad_norm": 0.2559722661972046, "learning_rate": 9.431605055681756e-05, "loss": 0.0221, "step": 19560 }, { "grad_norm": 0.2896353304386139, "learning_rate": 9.430839138896026e-05, "loss": 0.0267, "step": 19570 }, { "grad_norm": 0.4183095693588257, "learning_rate": 9.43007273756039e-05, "loss": 0.0216, "step": 19580 }, { "grad_norm": 0.2934457063674927, "learning_rate": 9.429305851758658e-05, "loss": 0.018, "step": 19590 }, { "grad_norm": 0.7340368628501892, "learning_rate": 9.428538481574699e-05, "loss": 0.0248, "step": 19600 }, { "grad_norm": 0.2995479702949524, "learning_rate": 9.42777062709243e-05, "loss": 0.0286, "step": 19610 }, { "grad_norm": 0.22973811626434326, "learning_rate": 9.427002288395821e-05, "loss": 0.0212, "step": 19620 }, { "grad_norm": 0.2708793878555298, "learning_rate": 9.426233465568898e-05, "loss": 0.0242, "step": 19630 }, { "grad_norm": 0.2674125134944916, "learning_rate": 9.42546415869574e-05, "loss": 0.0209, "step": 19640 }, { "grad_norm": 0.31163492798805237, "learning_rate": 9.424694367860473e-05, "loss": 0.0205, "step": 19650 }, { "grad_norm": 0.2961372137069702, "learning_rate": 9.423924093147284e-05, "loss": 0.0245, "step": 19660 }, { "grad_norm": 0.273637980222702, "learning_rate": 9.423153334640407e-05, "loss": 0.0247, "step": 19670 }, { "grad_norm": 0.25926896929740906, "learning_rate": 9.42238209242413e-05, "loss": 0.0267, "step": 19680 }, { "grad_norm": 0.34164056181907654, "learning_rate": 9.421610366582798e-05, "loss": 0.0252, "step": 19690 }, { "grad_norm": 0.33916351199150085, "learning_rate": 9.420838157200803e-05, "loss": 0.028, "step": 19700 }, { "grad_norm": 0.22663958370685577, "learning_rate": 9.420065464362594e-05, "loss": 0.0222, "step": 19710 }, { "grad_norm": 0.49451643228530884, "learning_rate": 9.419292288152673e-05, "loss": 0.0266, "step": 19720 }, { "grad_norm": 0.25883132219314575, "learning_rate": 9.418518628655588e-05, "loss": 0.0213, "step": 19730 }, { "grad_norm": 0.25890612602233887, "learning_rate": 9.417744485955951e-05, "loss": 0.0185, "step": 19740 }, { "grad_norm": 0.20565538108348846, "learning_rate": 9.41696986013842e-05, "loss": 0.0198, "step": 19750 }, { "grad_norm": 0.1927885264158249, "learning_rate": 9.416194751287705e-05, "loss": 0.0202, "step": 19760 }, { "grad_norm": 0.23130841553211212, "learning_rate": 9.415419159488572e-05, "loss": 0.0206, "step": 19770 }, { "grad_norm": 0.21814322471618652, "learning_rate": 9.414643084825837e-05, "loss": 0.0193, "step": 19780 }, { "grad_norm": 0.20410816371440887, "learning_rate": 9.413866527384372e-05, "loss": 0.0226, "step": 19790 }, { "grad_norm": 0.4869363009929657, "learning_rate": 9.4130894872491e-05, "loss": 0.0216, "step": 19800 }, { "grad_norm": 0.3355405032634735, "learning_rate": 9.412311964504998e-05, "loss": 0.023, "step": 19810 }, { "grad_norm": 0.215331569314003, "learning_rate": 9.411533959237091e-05, "loss": 0.0193, "step": 19820 }, { "grad_norm": 0.2636283040046692, "learning_rate": 9.410755471530464e-05, "loss": 0.021, "step": 19830 }, { "grad_norm": 0.3009485900402069, "learning_rate": 9.40997650147025e-05, "loss": 0.0248, "step": 19840 }, { "grad_norm": 0.23115243017673492, "learning_rate": 9.409197049141637e-05, "loss": 0.0251, "step": 19850 }, { "grad_norm": 0.240243062376976, "learning_rate": 9.408417114629863e-05, "loss": 0.0263, "step": 19860 }, { "grad_norm": 0.2879892587661743, "learning_rate": 9.40763669802022e-05, "loss": 0.0212, "step": 19870 }, { "grad_norm": 0.23390080034732819, "learning_rate": 9.406855799398056e-05, "loss": 0.0247, "step": 19880 }, { "grad_norm": 0.30471089482307434, "learning_rate": 9.406074418848767e-05, "loss": 0.0218, "step": 19890 }, { "grad_norm": 0.2129017859697342, "learning_rate": 9.405292556457805e-05, "loss": 0.0197, "step": 19900 }, { "grad_norm": 0.21858228743076324, "learning_rate": 9.404510212310671e-05, "loss": 0.0184, "step": 19910 }, { "grad_norm": 0.2653626799583435, "learning_rate": 9.403727386492924e-05, "loss": 0.0242, "step": 19920 }, { "grad_norm": 0.42525750398635864, "learning_rate": 9.40294407909017e-05, "loss": 0.0214, "step": 19930 }, { "grad_norm": 0.25080573558807373, "learning_rate": 9.40216029018807e-05, "loss": 0.0185, "step": 19940 }, { "grad_norm": 0.27139368653297424, "learning_rate": 9.401376019872338e-05, "loss": 0.0211, "step": 19950 }, { "grad_norm": 0.28134047985076904, "learning_rate": 9.400591268228746e-05, "loss": 0.0244, "step": 19960 }, { "grad_norm": 0.42171046137809753, "learning_rate": 9.399806035343106e-05, "loss": 0.0227, "step": 19970 }, { "grad_norm": 0.9932395219802856, "learning_rate": 9.399020321301294e-05, "loss": 0.0195, "step": 19980 }, { "grad_norm": 0.44251498579978943, "learning_rate": 9.398234126189234e-05, "loss": 0.0259, "step": 19990 }, { "grad_norm": 0.3916749358177185, "learning_rate": 9.397447450092902e-05, "loss": 0.0244, "step": 20000 }, { "grad_norm": 0.40594425797462463, "learning_rate": 9.39666029309833e-05, "loss": 0.0237, "step": 20010 }, { "grad_norm": 0.49174198508262634, "learning_rate": 9.395872655291596e-05, "loss": 0.0266, "step": 20020 }, { "grad_norm": 0.354988694190979, "learning_rate": 9.395084536758838e-05, "loss": 0.0245, "step": 20030 }, { "grad_norm": 0.2937333285808563, "learning_rate": 9.394295937586243e-05, "loss": 0.0225, "step": 20040 }, { "grad_norm": 0.40189456939697266, "learning_rate": 9.393506857860052e-05, "loss": 0.0246, "step": 20050 }, { "grad_norm": 0.2871283292770386, "learning_rate": 9.392717297666555e-05, "loss": 0.0208, "step": 20060 }, { "grad_norm": 0.27989163994789124, "learning_rate": 9.391927257092101e-05, "loss": 0.0258, "step": 20070 }, { "grad_norm": 0.2691628634929657, "learning_rate": 9.391136736223085e-05, "loss": 0.0233, "step": 20080 }, { "grad_norm": 0.26857030391693115, "learning_rate": 9.390345735145956e-05, "loss": 0.0222, "step": 20090 }, { "grad_norm": 0.21027307212352753, "learning_rate": 9.389554253947219e-05, "loss": 0.0185, "step": 20100 }, { "grad_norm": 0.22335326671600342, "learning_rate": 9.388762292713428e-05, "loss": 0.0246, "step": 20110 }, { "grad_norm": 0.2795960307121277, "learning_rate": 9.38796985153119e-05, "loss": 0.0244, "step": 20120 }, { "grad_norm": 0.32808077335357666, "learning_rate": 9.387176930487169e-05, "loss": 0.0237, "step": 20130 }, { "grad_norm": 0.36685389280319214, "learning_rate": 9.386383529668072e-05, "loss": 0.0215, "step": 20140 }, { "grad_norm": 0.25723016262054443, "learning_rate": 9.385589649160669e-05, "loss": 0.0263, "step": 20150 }, { "grad_norm": 0.3454265594482422, "learning_rate": 9.384795289051775e-05, "loss": 0.0236, "step": 20160 }, { "grad_norm": 0.2752419412136078, "learning_rate": 9.384000449428261e-05, "loss": 0.0266, "step": 20170 }, { "grad_norm": 0.29808303713798523, "learning_rate": 9.383205130377048e-05, "loss": 0.0243, "step": 20180 }, { "grad_norm": 0.2890200912952423, "learning_rate": 9.382409331985114e-05, "loss": 0.0252, "step": 20190 }, { "grad_norm": 0.24967168271541595, "learning_rate": 9.381613054339482e-05, "loss": 0.0259, "step": 20200 }, { "grad_norm": 0.24733322858810425, "learning_rate": 9.380816297527235e-05, "loss": 0.0222, "step": 20210 }, { "grad_norm": 0.28693512082099915, "learning_rate": 9.380019061635506e-05, "loss": 0.0214, "step": 20220 }, { "grad_norm": 0.296246737241745, "learning_rate": 9.379221346751474e-05, "loss": 0.0198, "step": 20230 }, { "grad_norm": 0.3003481328487396, "learning_rate": 9.378423152962382e-05, "loss": 0.0211, "step": 20240 }, { "grad_norm": 0.2340313196182251, "learning_rate": 9.377624480355517e-05, "loss": 0.0168, "step": 20250 }, { "grad_norm": 0.18052607774734497, "learning_rate": 9.376825329018219e-05, "loss": 0.0191, "step": 20260 }, { "grad_norm": 0.2916330397129059, "learning_rate": 9.376025699037884e-05, "loss": 0.0173, "step": 20270 }, { "grad_norm": 0.206673264503479, "learning_rate": 9.37522559050196e-05, "loss": 0.0198, "step": 20280 }, { "grad_norm": 0.3184839189052582, "learning_rate": 9.37442500349794e-05, "loss": 0.0171, "step": 20290 }, { "grad_norm": 0.2387523353099823, "learning_rate": 9.373623938113381e-05, "loss": 0.0197, "step": 20300 }, { "grad_norm": 0.20809704065322876, "learning_rate": 9.372822394435883e-05, "loss": 0.0188, "step": 20310 }, { "grad_norm": 0.25213518738746643, "learning_rate": 9.372020372553102e-05, "loss": 0.0151, "step": 20320 }, { "grad_norm": 0.28349822759628296, "learning_rate": 9.371217872552746e-05, "loss": 0.024, "step": 20330 }, { "grad_norm": 0.2729646861553192, "learning_rate": 9.370414894522576e-05, "loss": 0.0222, "step": 20340 }, { "grad_norm": 0.291485995054245, "learning_rate": 9.369611438550406e-05, "loss": 0.0189, "step": 20350 }, { "grad_norm": 0.23573951423168182, "learning_rate": 9.368807504724095e-05, "loss": 0.0286, "step": 20360 }, { "grad_norm": 1.0631096363067627, "learning_rate": 9.368003093131565e-05, "loss": 0.0218, "step": 20370 }, { "grad_norm": 0.2284475415945053, "learning_rate": 9.367198203860785e-05, "loss": 0.0168, "step": 20380 }, { "grad_norm": 0.25585225224494934, "learning_rate": 9.366392836999774e-05, "loss": 0.0231, "step": 20390 }, { "grad_norm": 0.22718989849090576, "learning_rate": 9.365586992636607e-05, "loss": 0.0203, "step": 20400 }, { "grad_norm": 0.2963326871395111, "learning_rate": 9.364780670859412e-05, "loss": 0.0218, "step": 20410 }, { "grad_norm": 0.19773739576339722, "learning_rate": 9.363973871756364e-05, "loss": 0.0187, "step": 20420 }, { "grad_norm": 0.23846706748008728, "learning_rate": 9.363166595415696e-05, "loss": 0.0246, "step": 20430 }, { "grad_norm": 0.26643574237823486, "learning_rate": 9.362358841925686e-05, "loss": 0.0233, "step": 20440 }, { "grad_norm": 0.23975525796413422, "learning_rate": 9.361550611374674e-05, "loss": 0.0165, "step": 20450 }, { "grad_norm": 0.22405660152435303, "learning_rate": 9.360741903851043e-05, "loss": 0.0207, "step": 20460 }, { "grad_norm": 0.23935213685035706, "learning_rate": 9.359932719443236e-05, "loss": 0.02, "step": 20470 }, { "grad_norm": 0.3944995403289795, "learning_rate": 9.35912305823974e-05, "loss": 0.0215, "step": 20480 }, { "grad_norm": 0.19331461191177368, "learning_rate": 9.358312920329101e-05, "loss": 0.0177, "step": 20490 }, { "grad_norm": 0.23351730406284332, "learning_rate": 9.357502305799914e-05, "loss": 0.0196, "step": 20500 }, { "grad_norm": 0.2122863382101059, "learning_rate": 9.356691214740824e-05, "loss": 0.0254, "step": 20510 }, { "grad_norm": 0.333894819021225, "learning_rate": 9.355879647240535e-05, "loss": 0.0272, "step": 20520 }, { "grad_norm": 0.2343405783176422, "learning_rate": 9.355067603387798e-05, "loss": 0.0176, "step": 20530 }, { "grad_norm": 0.18722057342529297, "learning_rate": 9.354255083271412e-05, "loss": 0.026, "step": 20540 }, { "grad_norm": 0.18035587668418884, "learning_rate": 9.353442086980239e-05, "loss": 0.0178, "step": 20550 }, { "grad_norm": 0.3446146547794342, "learning_rate": 9.352628614603185e-05, "loss": 0.0198, "step": 20560 }, { "grad_norm": 0.21457001566886902, "learning_rate": 9.351814666229209e-05, "loss": 0.0206, "step": 20570 }, { "grad_norm": 0.20254774391651154, "learning_rate": 9.351000241947324e-05, "loss": 0.0237, "step": 20580 }, { "grad_norm": 0.317060649394989, "learning_rate": 9.350185341846594e-05, "loss": 0.0193, "step": 20590 }, { "grad_norm": 0.26500794291496277, "learning_rate": 9.349369966016134e-05, "loss": 0.0215, "step": 20600 }, { "grad_norm": 0.2667083442211151, "learning_rate": 9.348554114545117e-05, "loss": 0.022, "step": 20610 }, { "grad_norm": 0.22264866530895233, "learning_rate": 9.347737787522758e-05, "loss": 0.0203, "step": 20620 }, { "grad_norm": 0.34156978130340576, "learning_rate": 9.346920985038332e-05, "loss": 0.026, "step": 20630 }, { "grad_norm": 0.4331700801849365, "learning_rate": 9.346103707181162e-05, "loss": 0.0226, "step": 20640 }, { "grad_norm": 0.30137965083122253, "learning_rate": 9.345285954040626e-05, "loss": 0.0195, "step": 20650 }, { "grad_norm": 0.29532894492149353, "learning_rate": 9.34446772570615e-05, "loss": 0.0277, "step": 20660 }, { "grad_norm": 0.28755906224250793, "learning_rate": 9.343649022267214e-05, "loss": 0.0257, "step": 20670 }, { "grad_norm": 0.2875427007675171, "learning_rate": 9.342829843813353e-05, "loss": 0.0201, "step": 20680 }, { "grad_norm": 0.19061510264873505, "learning_rate": 9.342010190434149e-05, "loss": 0.0279, "step": 20690 }, { "grad_norm": 0.3229869306087494, "learning_rate": 9.34119006221924e-05, "loss": 0.019, "step": 20700 }, { "grad_norm": 0.21197789907455444, "learning_rate": 9.340369459258313e-05, "loss": 0.0169, "step": 20710 }, { "grad_norm": 0.2507754862308502, "learning_rate": 9.339548381641106e-05, "loss": 0.0183, "step": 20720 }, { "grad_norm": 0.26607027649879456, "learning_rate": 9.338726829457413e-05, "loss": 0.0251, "step": 20730 }, { "grad_norm": 0.19992344081401825, "learning_rate": 9.337904802797078e-05, "loss": 0.0248, "step": 20740 }, { "grad_norm": 0.583223819732666, "learning_rate": 9.337082301749993e-05, "loss": 0.0171, "step": 20750 }, { "grad_norm": 0.30427971482276917, "learning_rate": 9.336259326406109e-05, "loss": 0.0186, "step": 20760 }, { "grad_norm": 0.2920789420604706, "learning_rate": 9.335435876855427e-05, "loss": 0.0221, "step": 20770 }, { "grad_norm": 0.35231757164001465, "learning_rate": 9.334611953187994e-05, "loss": 0.0227, "step": 20780 }, { "grad_norm": 0.7348176836967468, "learning_rate": 9.333787555493914e-05, "loss": 0.0232, "step": 20790 }, { "grad_norm": 0.39360418915748596, "learning_rate": 9.332962683863345e-05, "loss": 0.0221, "step": 20800 }, { "grad_norm": 0.291094034910202, "learning_rate": 9.332137338386489e-05, "loss": 0.0305, "step": 20810 }, { "grad_norm": 0.27588415145874023, "learning_rate": 9.33131151915361e-05, "loss": 0.025, "step": 20820 }, { "grad_norm": 0.3856353163719177, "learning_rate": 9.330485226255012e-05, "loss": 0.0257, "step": 20830 }, { "grad_norm": 0.20501327514648438, "learning_rate": 9.329658459781061e-05, "loss": 0.0215, "step": 20840 }, { "grad_norm": 0.8659300804138184, "learning_rate": 9.328831219822172e-05, "loss": 0.0231, "step": 20850 }, { "grad_norm": 0.382781445980072, "learning_rate": 9.328003506468808e-05, "loss": 0.0223, "step": 20860 }, { "grad_norm": 0.19384577870368958, "learning_rate": 9.327175319811488e-05, "loss": 0.0219, "step": 20870 }, { "grad_norm": 0.23919856548309326, "learning_rate": 9.326346659940781e-05, "loss": 0.0182, "step": 20880 }, { "grad_norm": 0.2674015164375305, "learning_rate": 9.325517526947308e-05, "loss": 0.0209, "step": 20890 }, { "grad_norm": 0.7221962213516235, "learning_rate": 9.32468792092174e-05, "loss": 0.0258, "step": 20900 }, { "grad_norm": 0.20922893285751343, "learning_rate": 9.323857841954803e-05, "loss": 0.021, "step": 20910 }, { "grad_norm": 0.4585886299610138, "learning_rate": 9.323027290137276e-05, "loss": 0.0255, "step": 20920 }, { "grad_norm": 0.2749691903591156, "learning_rate": 9.322196265559981e-05, "loss": 0.0257, "step": 20930 }, { "grad_norm": 0.3102920353412628, "learning_rate": 9.321364768313803e-05, "loss": 0.0247, "step": 20940 }, { "grad_norm": 0.1623324304819107, "learning_rate": 9.32053279848967e-05, "loss": 0.0248, "step": 20950 }, { "grad_norm": 0.21491128206253052, "learning_rate": 9.319700356178567e-05, "loss": 0.0182, "step": 20960 }, { "grad_norm": 0.24995197355747223, "learning_rate": 9.318867441471527e-05, "loss": 0.0199, "step": 20970 }, { "grad_norm": 0.22133271396160126, "learning_rate": 9.318034054459637e-05, "loss": 0.018, "step": 20980 }, { "grad_norm": 0.24403619766235352, "learning_rate": 9.317200195234034e-05, "loss": 0.0187, "step": 20990 }, { "grad_norm": 0.3374694287776947, "learning_rate": 9.316365863885909e-05, "loss": 0.0223, "step": 21000 }, { "grad_norm": 0.3101835250854492, "learning_rate": 9.315531060506502e-05, "loss": 0.021, "step": 21010 }, { "grad_norm": 0.617275595664978, "learning_rate": 9.314695785187108e-05, "loss": 0.0233, "step": 21020 }, { "grad_norm": 0.19602423906326294, "learning_rate": 9.313860038019069e-05, "loss": 0.0233, "step": 21030 }, { "grad_norm": 0.3471735715866089, "learning_rate": 9.313023819093782e-05, "loss": 0.0293, "step": 21040 }, { "grad_norm": 0.36022815108299255, "learning_rate": 9.312187128502695e-05, "loss": 0.0258, "step": 21050 }, { "grad_norm": 0.4689730405807495, "learning_rate": 9.311349966337307e-05, "loss": 0.0269, "step": 21060 }, { "grad_norm": 0.27229034900665283, "learning_rate": 9.310512332689169e-05, "loss": 0.0208, "step": 21070 }, { "grad_norm": 0.4463978707790375, "learning_rate": 9.309674227649883e-05, "loss": 0.0336, "step": 21080 }, { "grad_norm": 0.35268884897232056, "learning_rate": 9.308835651311103e-05, "loss": 0.0245, "step": 21090 }, { "grad_norm": 0.47654733061790466, "learning_rate": 9.307996603764533e-05, "loss": 0.0269, "step": 21100 }, { "grad_norm": 0.25702494382858276, "learning_rate": 9.307157085101932e-05, "loss": 0.0373, "step": 21110 }, { "grad_norm": 0.38990718126296997, "learning_rate": 9.306317095415109e-05, "loss": 0.0282, "step": 21120 }, { "grad_norm": 0.32317131757736206, "learning_rate": 9.305476634795922e-05, "loss": 0.0298, "step": 21130 }, { "grad_norm": 0.4385737180709839, "learning_rate": 9.304635703336284e-05, "loss": 0.0327, "step": 21140 }, { "grad_norm": 0.2753588855266571, "learning_rate": 9.303794301128157e-05, "loss": 0.0232, "step": 21150 }, { "grad_norm": 0.25216010212898254, "learning_rate": 9.302952428263555e-05, "loss": 0.0229, "step": 21160 }, { "grad_norm": 0.268305242061615, "learning_rate": 9.302110084834545e-05, "loss": 0.0258, "step": 21170 }, { "grad_norm": 0.26917847990989685, "learning_rate": 9.301267270933245e-05, "loss": 0.022, "step": 21180 }, { "grad_norm": 0.20655614137649536, "learning_rate": 9.300423986651823e-05, "loss": 0.0189, "step": 21190 }, { "grad_norm": 0.253476083278656, "learning_rate": 9.299580232082501e-05, "loss": 0.0187, "step": 21200 }, { "grad_norm": 0.23926246166229248, "learning_rate": 9.298736007317547e-05, "loss": 0.0197, "step": 21210 }, { "grad_norm": 0.27405160665512085, "learning_rate": 9.297891312449288e-05, "loss": 0.0169, "step": 21220 }, { "grad_norm": 0.24649891257286072, "learning_rate": 9.297046147570094e-05, "loss": 0.0231, "step": 21230 }, { "grad_norm": 0.2507282793521881, "learning_rate": 9.296200512772396e-05, "loss": 0.0188, "step": 21240 }, { "grad_norm": 0.3160552680492401, "learning_rate": 9.295354408148668e-05, "loss": 0.0197, "step": 21250 }, { "grad_norm": 0.22031867504119873, "learning_rate": 9.294507833791441e-05, "loss": 0.027, "step": 21260 }, { "grad_norm": 0.17644838988780975, "learning_rate": 9.293660789793295e-05, "loss": 0.0201, "step": 21270 }, { "grad_norm": 0.251954585313797, "learning_rate": 9.292813276246858e-05, "loss": 0.0238, "step": 21280 }, { "grad_norm": 0.2550179660320282, "learning_rate": 9.291965293244816e-05, "loss": 0.0196, "step": 21290 }, { "grad_norm": 0.2952556312084198, "learning_rate": 9.291116840879904e-05, "loss": 0.0227, "step": 21300 }, { "grad_norm": 0.23535941541194916, "learning_rate": 9.290267919244904e-05, "loss": 0.0209, "step": 21310 }, { "grad_norm": 0.31325680017471313, "learning_rate": 9.289418528432655e-05, "loss": 0.0219, "step": 21320 }, { "grad_norm": 0.2266864776611328, "learning_rate": 9.288568668536045e-05, "loss": 0.0193, "step": 21330 }, { "grad_norm": 0.3140270411968231, "learning_rate": 9.287718339648013e-05, "loss": 0.0162, "step": 21340 }, { "grad_norm": 0.2025868147611618, "learning_rate": 9.28686754186155e-05, "loss": 0.0193, "step": 21350 }, { "grad_norm": 0.3253897428512573, "learning_rate": 9.286016275269698e-05, "loss": 0.0249, "step": 21360 }, { "grad_norm": 0.260183185338974, "learning_rate": 9.285164539965551e-05, "loss": 0.0239, "step": 21370 }, { "grad_norm": 0.25673237442970276, "learning_rate": 9.284312336042251e-05, "loss": 0.02, "step": 21380 }, { "grad_norm": 0.23196786642074585, "learning_rate": 9.283459663592996e-05, "loss": 0.021, "step": 21390 }, { "grad_norm": 0.21147368848323822, "learning_rate": 9.282606522711033e-05, "loss": 0.0218, "step": 21400 }, { "grad_norm": 0.2911405861377716, "learning_rate": 9.281752913489657e-05, "loss": 0.0179, "step": 21410 }, { "grad_norm": 0.18618975579738617, "learning_rate": 9.280898836022222e-05, "loss": 0.0162, "step": 21420 }, { "grad_norm": 0.23749884963035583, "learning_rate": 9.280044290402126e-05, "loss": 0.0215, "step": 21430 }, { "grad_norm": 0.28505975008010864, "learning_rate": 9.279189276722821e-05, "loss": 0.0208, "step": 21440 }, { "grad_norm": 0.35766178369522095, "learning_rate": 9.278333795077812e-05, "loss": 0.02, "step": 21450 }, { "grad_norm": 0.19533583521842957, "learning_rate": 9.27747784556065e-05, "loss": 0.0173, "step": 21460 }, { "grad_norm": 1.2006808519363403, "learning_rate": 9.276621428264942e-05, "loss": 0.0225, "step": 21470 }, { "grad_norm": 0.21626698970794678, "learning_rate": 9.275764543284345e-05, "loss": 0.0161, "step": 21480 }, { "grad_norm": 0.2728550434112549, "learning_rate": 9.274907190712566e-05, "loss": 0.0207, "step": 21490 }, { "grad_norm": 0.5148512721061707, "learning_rate": 9.274049370643363e-05, "loss": 0.0174, "step": 21500 }, { "grad_norm": 0.21053196489810944, "learning_rate": 9.273191083170547e-05, "loss": 0.019, "step": 21510 }, { "grad_norm": 0.23221050202846527, "learning_rate": 9.27233232838798e-05, "loss": 0.02, "step": 21520 }, { "grad_norm": 0.27040886878967285, "learning_rate": 9.27147310638957e-05, "loss": 0.0242, "step": 21530 }, { "grad_norm": 0.2757311165332794, "learning_rate": 9.270613417269286e-05, "loss": 0.0177, "step": 21540 }, { "grad_norm": 0.3519951403141022, "learning_rate": 9.269753261121138e-05, "loss": 0.016, "step": 21550 }, { "grad_norm": 0.2035312056541443, "learning_rate": 9.268892638039194e-05, "loss": 0.015, "step": 21560 }, { "grad_norm": 0.26431459188461304, "learning_rate": 9.268031548117569e-05, "loss": 0.0163, "step": 21570 }, { "grad_norm": 0.4023219645023346, "learning_rate": 9.26716999145043e-05, "loss": 0.0176, "step": 21580 }, { "grad_norm": 0.214189812541008, "learning_rate": 9.266307968131998e-05, "loss": 0.022, "step": 21590 }, { "grad_norm": 0.34648725390434265, "learning_rate": 9.26544547825654e-05, "loss": 0.0206, "step": 21600 }, { "grad_norm": 0.2750266492366791, "learning_rate": 9.264582521918376e-05, "loss": 0.022, "step": 21610 }, { "grad_norm": 0.4048900008201599, "learning_rate": 9.263719099211881e-05, "loss": 0.0242, "step": 21620 }, { "grad_norm": 0.3402339518070221, "learning_rate": 9.262855210231476e-05, "loss": 0.0221, "step": 21630 }, { "grad_norm": 0.25042495131492615, "learning_rate": 9.261990855071633e-05, "loss": 0.0173, "step": 21640 }, { "grad_norm": 0.22340171039104462, "learning_rate": 9.261126033826878e-05, "loss": 0.0153, "step": 21650 }, { "grad_norm": 0.22981397807598114, "learning_rate": 9.260260746591786e-05, "loss": 0.0179, "step": 21660 }, { "grad_norm": 0.21550528705120087, "learning_rate": 9.259394993460985e-05, "loss": 0.0177, "step": 21670 }, { "grad_norm": 0.1690243035554886, "learning_rate": 9.258528774529151e-05, "loss": 0.0156, "step": 21680 }, { "grad_norm": 0.16472159326076508, "learning_rate": 9.257662089891013e-05, "loss": 0.0162, "step": 21690 }, { "grad_norm": 0.22735968232154846, "learning_rate": 9.25679493964135e-05, "loss": 0.0171, "step": 21700 }, { "grad_norm": 0.2253212183713913, "learning_rate": 9.255927323874994e-05, "loss": 0.0167, "step": 21710 }, { "grad_norm": 0.21449650824069977, "learning_rate": 9.255059242686822e-05, "loss": 0.0168, "step": 21720 }, { "grad_norm": 0.2070111334323883, "learning_rate": 9.254190696171769e-05, "loss": 0.0185, "step": 21730 }, { "grad_norm": 0.19534139335155487, "learning_rate": 9.25332168442482e-05, "loss": 0.0157, "step": 21740 }, { "grad_norm": 0.27708154916763306, "learning_rate": 9.252452207541004e-05, "loss": 0.0181, "step": 21750 }, { "grad_norm": 0.2645537257194519, "learning_rate": 9.251582265615409e-05, "loss": 0.0193, "step": 21760 }, { "grad_norm": 0.2100069522857666, "learning_rate": 9.250711858743169e-05, "loss": 0.0141, "step": 21770 }, { "grad_norm": 0.19161155819892883, "learning_rate": 9.24984098701947e-05, "loss": 0.0179, "step": 21780 }, { "grad_norm": 0.22644297778606415, "learning_rate": 9.248969650539552e-05, "loss": 0.0182, "step": 21790 }, { "grad_norm": 0.5675626397132874, "learning_rate": 9.2480978493987e-05, "loss": 0.0243, "step": 21800 }, { "grad_norm": 0.31623542308807373, "learning_rate": 9.247225583692256e-05, "loss": 0.0258, "step": 21810 }, { "grad_norm": 0.28908950090408325, "learning_rate": 9.246352853515607e-05, "loss": 0.0166, "step": 21820 }, { "grad_norm": 0.2600690722465515, "learning_rate": 9.245479658964194e-05, "loss": 0.0209, "step": 21830 }, { "grad_norm": 0.25609397888183594, "learning_rate": 9.244606000133507e-05, "loss": 0.0212, "step": 21840 }, { "grad_norm": 0.25856441259384155, "learning_rate": 9.24373187711909e-05, "loss": 0.0197, "step": 21850 }, { "grad_norm": 0.2946673035621643, "learning_rate": 9.242857290016537e-05, "loss": 0.0329, "step": 21860 }, { "grad_norm": 0.2593114972114563, "learning_rate": 9.241982238921488e-05, "loss": 0.0265, "step": 21870 }, { "grad_norm": 0.42808735370635986, "learning_rate": 9.24110672392964e-05, "loss": 0.0253, "step": 21880 }, { "grad_norm": 0.3685460686683655, "learning_rate": 9.240230745136737e-05, "loss": 0.0278, "step": 21890 }, { "grad_norm": 0.3015224039554596, "learning_rate": 9.239354302638575e-05, "loss": 0.0243, "step": 21900 }, { "grad_norm": 0.28689754009246826, "learning_rate": 9.238477396531e-05, "loss": 0.0255, "step": 21910 }, { "grad_norm": 0.276462197303772, "learning_rate": 9.23760002690991e-05, "loss": 0.0238, "step": 21920 }, { "grad_norm": 0.30265313386917114, "learning_rate": 9.236722193871252e-05, "loss": 0.0201, "step": 21930 }, { "grad_norm": 0.30846574902534485, "learning_rate": 9.235843897511023e-05, "loss": 0.0202, "step": 21940 }, { "grad_norm": 0.23068055510520935, "learning_rate": 9.234965137925276e-05, "loss": 0.0227, "step": 21950 }, { "grad_norm": 0.3020731806755066, "learning_rate": 9.234085915210108e-05, "loss": 0.0219, "step": 21960 }, { "grad_norm": 0.31188884377479553, "learning_rate": 9.23320622946167e-05, "loss": 0.0222, "step": 21970 }, { "grad_norm": 0.392160028219223, "learning_rate": 9.232326080776163e-05, "loss": 0.0279, "step": 21980 }, { "grad_norm": 0.2523057460784912, "learning_rate": 9.23144546924984e-05, "loss": 0.0221, "step": 21990 }, { "grad_norm": 0.37285733222961426, "learning_rate": 9.230564394979e-05, "loss": 0.0226, "step": 22000 }, { "grad_norm": 0.36007726192474365, "learning_rate": 9.22968285806e-05, "loss": 0.0233, "step": 22010 }, { "grad_norm": 0.34835460782051086, "learning_rate": 9.228800858589242e-05, "loss": 0.0231, "step": 22020 }, { "grad_norm": 0.20525479316711426, "learning_rate": 9.227918396663179e-05, "loss": 0.0208, "step": 22030 }, { "grad_norm": 0.3232020139694214, "learning_rate": 9.227035472378319e-05, "loss": 0.0217, "step": 22040 }, { "grad_norm": 0.259125679731369, "learning_rate": 9.226152085831213e-05, "loss": 0.0202, "step": 22050 }, { "grad_norm": 0.20610728859901428, "learning_rate": 9.22526823711847e-05, "loss": 0.0198, "step": 22060 }, { "grad_norm": 0.23439937829971313, "learning_rate": 9.224383926336745e-05, "loss": 0.0188, "step": 22070 }, { "grad_norm": 0.3169318437576294, "learning_rate": 9.223499153582744e-05, "loss": 0.0193, "step": 22080 }, { "grad_norm": 0.268914133310318, "learning_rate": 9.222613918953226e-05, "loss": 0.0198, "step": 22090 }, { "grad_norm": 0.27542075514793396, "learning_rate": 9.221728222544999e-05, "loss": 0.0235, "step": 22100 }, { "grad_norm": 0.3582070469856262, "learning_rate": 9.22084206445492e-05, "loss": 0.0212, "step": 22110 }, { "grad_norm": 0.21545951068401337, "learning_rate": 9.2199554447799e-05, "loss": 0.0246, "step": 22120 }, { "grad_norm": 0.2084716409444809, "learning_rate": 9.219068363616897e-05, "loss": 0.02, "step": 22130 }, { "grad_norm": 0.284559041261673, "learning_rate": 9.218180821062919e-05, "loss": 0.0175, "step": 22140 }, { "grad_norm": 0.27954626083374023, "learning_rate": 9.21729281721503e-05, "loss": 0.0182, "step": 22150 }, { "grad_norm": 0.29645034670829773, "learning_rate": 9.216404352170339e-05, "loss": 0.0187, "step": 22160 }, { "grad_norm": 0.6130965352058411, "learning_rate": 9.215515426026007e-05, "loss": 0.0209, "step": 22170 }, { "grad_norm": 0.27579328417778015, "learning_rate": 9.214626038879246e-05, "loss": 0.0201, "step": 22180 }, { "grad_norm": 0.2540324330329895, "learning_rate": 9.21373619082732e-05, "loss": 0.0356, "step": 22190 }, { "grad_norm": 0.2912975549697876, "learning_rate": 9.212845881967535e-05, "loss": 0.0253, "step": 22200 }, { "grad_norm": 0.23322436213493347, "learning_rate": 9.211955112397262e-05, "loss": 0.0215, "step": 22210 }, { "grad_norm": 0.2627612352371216, "learning_rate": 9.211063882213909e-05, "loss": 0.0244, "step": 22220 }, { "grad_norm": 0.2628291845321655, "learning_rate": 9.210172191514942e-05, "loss": 0.0236, "step": 22230 }, { "grad_norm": 0.6609616875648499, "learning_rate": 9.209280040397874e-05, "loss": 0.026, "step": 22240 }, { "grad_norm": 0.28516626358032227, "learning_rate": 9.208387428960268e-05, "loss": 0.0221, "step": 22250 }, { "grad_norm": 0.27914702892303467, "learning_rate": 9.20749435729974e-05, "loss": 0.0205, "step": 22260 }, { "grad_norm": 0.3033660352230072, "learning_rate": 9.206600825513957e-05, "loss": 0.0213, "step": 22270 }, { "grad_norm": 0.6195708513259888, "learning_rate": 9.20570683370063e-05, "loss": 0.0272, "step": 22280 }, { "grad_norm": 0.18154218792915344, "learning_rate": 9.204812381957528e-05, "loss": 0.0206, "step": 22290 }, { "grad_norm": 0.2787972092628479, "learning_rate": 9.203917470382465e-05, "loss": 0.0231, "step": 22300 }, { "grad_norm": 0.2808852791786194, "learning_rate": 9.203022099073309e-05, "loss": 0.0195, "step": 22310 }, { "grad_norm": 0.28487569093704224, "learning_rate": 9.202126268127976e-05, "loss": 0.0248, "step": 22320 }, { "grad_norm": 0.31485897302627563, "learning_rate": 9.20122997764443e-05, "loss": 0.0225, "step": 22330 }, { "grad_norm": 0.38694700598716736, "learning_rate": 9.200333227720692e-05, "loss": 0.0323, "step": 22340 }, { "grad_norm": 0.2692127823829651, "learning_rate": 9.199436018454826e-05, "loss": 0.0203, "step": 22350 }, { "grad_norm": 0.32322224974632263, "learning_rate": 9.198538349944952e-05, "loss": 0.0239, "step": 22360 }, { "grad_norm": 0.5114679932594299, "learning_rate": 9.197640222289234e-05, "loss": 0.0189, "step": 22370 }, { "grad_norm": 0.20521776378154755, "learning_rate": 9.196741635585895e-05, "loss": 0.019, "step": 22380 }, { "grad_norm": 0.217690110206604, "learning_rate": 9.195842589933199e-05, "loss": 0.0188, "step": 22390 }, { "grad_norm": 0.28319403529167175, "learning_rate": 9.194943085429466e-05, "loss": 0.0227, "step": 22400 }, { "grad_norm": 0.36305105686187744, "learning_rate": 9.194043122173065e-05, "loss": 0.0312, "step": 22410 }, { "grad_norm": 0.592848539352417, "learning_rate": 9.193142700262413e-05, "loss": 0.0216, "step": 22420 }, { "grad_norm": 0.21224194765090942, "learning_rate": 9.192241819795979e-05, "loss": 0.02, "step": 22430 }, { "grad_norm": 0.4931847155094147, "learning_rate": 9.191340480872284e-05, "loss": 0.0203, "step": 22440 }, { "grad_norm": 0.2071005254983902, "learning_rate": 9.190438683589895e-05, "loss": 0.0255, "step": 22450 }, { "grad_norm": 0.290958970785141, "learning_rate": 9.189536428047432e-05, "loss": 0.0276, "step": 22460 }, { "grad_norm": 0.21945494413375854, "learning_rate": 9.188633714343564e-05, "loss": 0.0167, "step": 22470 }, { "grad_norm": 0.26779747009277344, "learning_rate": 9.18773054257701e-05, "loss": 0.0183, "step": 22480 }, { "grad_norm": 0.24644342064857483, "learning_rate": 9.18682691284654e-05, "loss": 0.0189, "step": 22490 }, { "grad_norm": 0.21734024584293365, "learning_rate": 9.185922825250974e-05, "loss": 0.0185, "step": 22500 }, { "grad_norm": 0.20049931108951569, "learning_rate": 9.185018279889181e-05, "loss": 0.0158, "step": 22510 }, { "grad_norm": 0.31715553998947144, "learning_rate": 9.184113276860082e-05, "loss": 0.0191, "step": 22520 }, { "grad_norm": 0.6275793313980103, "learning_rate": 9.183207816262645e-05, "loss": 0.0202, "step": 22530 }, { "grad_norm": 0.30253419280052185, "learning_rate": 9.182301898195891e-05, "loss": 0.0178, "step": 22540 }, { "grad_norm": 0.1807330995798111, "learning_rate": 9.181395522758889e-05, "loss": 0.0144, "step": 22550 }, { "grad_norm": 0.2415606677532196, "learning_rate": 9.180488690050759e-05, "loss": 0.016, "step": 22560 }, { "grad_norm": 0.28887224197387695, "learning_rate": 9.179581400170671e-05, "loss": 0.0193, "step": 22570 }, { "grad_norm": 0.19661515951156616, "learning_rate": 9.178673653217845e-05, "loss": 0.0143, "step": 22580 }, { "grad_norm": 0.2877221405506134, "learning_rate": 9.177765449291551e-05, "loss": 0.0154, "step": 22590 }, { "grad_norm": 0.2810191214084625, "learning_rate": 9.176856788491109e-05, "loss": 0.0181, "step": 22600 }, { "grad_norm": 0.20389051735401154, "learning_rate": 9.175947670915887e-05, "loss": 0.0215, "step": 22610 }, { "grad_norm": 0.20803508162498474, "learning_rate": 9.175038096665309e-05, "loss": 0.019, "step": 22620 }, { "grad_norm": 0.2616812586784363, "learning_rate": 9.17412806583884e-05, "loss": 0.0238, "step": 22630 }, { "grad_norm": 0.28149595856666565, "learning_rate": 9.173217578536002e-05, "loss": 0.0172, "step": 22640 }, { "grad_norm": 0.30174142122268677, "learning_rate": 9.172306634856362e-05, "loss": 0.0195, "step": 22650 }, { "grad_norm": 0.23011693358421326, "learning_rate": 9.171395234899545e-05, "loss": 0.0172, "step": 22660 }, { "grad_norm": 0.2384185492992401, "learning_rate": 9.170483378765214e-05, "loss": 0.026, "step": 22670 }, { "grad_norm": 0.2795388102531433, "learning_rate": 9.169571066553091e-05, "loss": 0.0202, "step": 22680 }, { "grad_norm": 0.36797836422920227, "learning_rate": 9.168658298362946e-05, "loss": 0.024, "step": 22690 }, { "grad_norm": 0.25747230648994446, "learning_rate": 9.167745074294598e-05, "loss": 0.0239, "step": 22700 }, { "grad_norm": 0.22716304659843445, "learning_rate": 9.166831394447913e-05, "loss": 0.021, "step": 22710 }, { "grad_norm": 0.20912888646125793, "learning_rate": 9.165917258922812e-05, "loss": 0.0259, "step": 22720 }, { "grad_norm": 0.2401341050863266, "learning_rate": 9.165002667819262e-05, "loss": 0.0172, "step": 22730 }, { "grad_norm": 0.19844630360603333, "learning_rate": 9.164087621237282e-05, "loss": 0.018, "step": 22740 }, { "grad_norm": 0.19280469417572021, "learning_rate": 9.163172119276942e-05, "loss": 0.0175, "step": 22750 }, { "grad_norm": 0.2911529541015625, "learning_rate": 9.162256162038358e-05, "loss": 0.022, "step": 22760 }, { "grad_norm": 0.3418218195438385, "learning_rate": 9.161339749621698e-05, "loss": 0.0227, "step": 22770 }, { "grad_norm": 0.22829879820346832, "learning_rate": 9.160422882127177e-05, "loss": 0.0228, "step": 22780 }, { "grad_norm": 0.2430768758058548, "learning_rate": 9.159505559655069e-05, "loss": 0.0202, "step": 22790 }, { "grad_norm": 0.2218742072582245, "learning_rate": 9.158587782305684e-05, "loss": 0.0191, "step": 22800 }, { "grad_norm": 0.2883113622665405, "learning_rate": 9.157669550179391e-05, "loss": 0.0191, "step": 22810 }, { "grad_norm": 0.2160344123840332, "learning_rate": 9.156750863376609e-05, "loss": 0.0164, "step": 22820 }, { "grad_norm": 0.17824067175388336, "learning_rate": 9.155831721997801e-05, "loss": 0.0156, "step": 22830 }, { "grad_norm": 0.226395383477211, "learning_rate": 9.154912126143484e-05, "loss": 0.0174, "step": 22840 }, { "grad_norm": 0.16191789507865906, "learning_rate": 9.153992075914224e-05, "loss": 0.0132, "step": 22850 }, { "grad_norm": 0.23423299193382263, "learning_rate": 9.153071571410635e-05, "loss": 0.0168, "step": 22860 }, { "grad_norm": 0.45316168665885925, "learning_rate": 9.152150612733384e-05, "loss": 0.0189, "step": 22870 }, { "grad_norm": 1.0579135417938232, "learning_rate": 9.151229199983184e-05, "loss": 0.0183, "step": 22880 }, { "grad_norm": 0.17805105447769165, "learning_rate": 9.150307333260802e-05, "loss": 0.016, "step": 22890 }, { "grad_norm": 0.41600435972213745, "learning_rate": 9.149385012667048e-05, "loss": 0.018, "step": 22900 }, { "grad_norm": 0.29298070073127747, "learning_rate": 9.148462238302788e-05, "loss": 0.0186, "step": 22910 }, { "grad_norm": 0.39259710907936096, "learning_rate": 9.147539010268936e-05, "loss": 0.0225, "step": 22920 }, { "grad_norm": 0.26199811697006226, "learning_rate": 9.14661532866645e-05, "loss": 0.0218, "step": 22930 }, { "grad_norm": 0.31589174270629883, "learning_rate": 9.145691193596348e-05, "loss": 0.0197, "step": 22940 }, { "grad_norm": 0.24749232828617096, "learning_rate": 9.144766605159691e-05, "loss": 0.0205, "step": 22950 }, { "grad_norm": 0.32179126143455505, "learning_rate": 9.14384156345759e-05, "loss": 0.0228, "step": 22960 }, { "grad_norm": 0.5423637628555298, "learning_rate": 9.142916068591204e-05, "loss": 0.0324, "step": 22970 }, { "grad_norm": 0.5325278043746948, "learning_rate": 9.141990120661746e-05, "loss": 0.025, "step": 22980 }, { "grad_norm": 0.21991290152072906, "learning_rate": 9.141063719770475e-05, "loss": 0.0187, "step": 22990 }, { "grad_norm": 0.33177444338798523, "learning_rate": 9.140136866018704e-05, "loss": 0.0186, "step": 23000 }, { "grad_norm": 0.1953316032886505, "learning_rate": 9.139209559507788e-05, "loss": 0.0205, "step": 23010 }, { "grad_norm": 0.23114944994449615, "learning_rate": 9.13828180033914e-05, "loss": 0.0199, "step": 23020 }, { "grad_norm": 0.23521730303764343, "learning_rate": 9.137353588614212e-05, "loss": 0.0187, "step": 23030 }, { "grad_norm": 0.27467867732048035, "learning_rate": 9.136424924434519e-05, "loss": 0.0171, "step": 23040 }, { "grad_norm": 0.3177742063999176, "learning_rate": 9.135495807901615e-05, "loss": 0.0293, "step": 23050 }, { "grad_norm": 0.6091543436050415, "learning_rate": 9.134566239117108e-05, "loss": 0.0205, "step": 23060 }, { "grad_norm": 0.32055896520614624, "learning_rate": 9.13363621818265e-05, "loss": 0.0229, "step": 23070 }, { "grad_norm": 0.5158346891403198, "learning_rate": 9.132705745199953e-05, "loss": 0.0245, "step": 23080 }, { "grad_norm": 0.2543289065361023, "learning_rate": 9.131774820270768e-05, "loss": 0.0211, "step": 23090 }, { "grad_norm": 0.8393169045448303, "learning_rate": 9.130843443496901e-05, "loss": 0.0277, "step": 23100 }, { "grad_norm": 0.28836876153945923, "learning_rate": 9.129911614980206e-05, "loss": 0.024, "step": 23110 }, { "grad_norm": 0.2713874876499176, "learning_rate": 9.128979334822584e-05, "loss": 0.0221, "step": 23120 }, { "grad_norm": 0.4567454755306244, "learning_rate": 9.128046603125992e-05, "loss": 0.019, "step": 23130 }, { "grad_norm": 0.21424582600593567, "learning_rate": 9.12711341999243e-05, "loss": 0.0184, "step": 23140 }, { "grad_norm": 0.3091282248497009, "learning_rate": 9.12617978552395e-05, "loss": 0.0224, "step": 23150 }, { "grad_norm": 0.6709915399551392, "learning_rate": 9.12524569982265e-05, "loss": 0.0244, "step": 23160 }, { "grad_norm": 0.28813445568084717, "learning_rate": 9.124311162990684e-05, "loss": 0.0186, "step": 23170 }, { "grad_norm": 0.16360481083393097, "learning_rate": 9.12337617513025e-05, "loss": 0.0196, "step": 23180 }, { "grad_norm": 0.18448759615421295, "learning_rate": 9.122440736343596e-05, "loss": 0.0162, "step": 23190 }, { "grad_norm": 0.14182554185390472, "learning_rate": 9.12150484673302e-05, "loss": 0.0175, "step": 23200 }, { "grad_norm": 0.19716574251651764, "learning_rate": 9.120568506400873e-05, "loss": 0.0274, "step": 23210 }, { "grad_norm": 0.19353263080120087, "learning_rate": 9.119631715449548e-05, "loss": 0.0177, "step": 23220 }, { "grad_norm": 0.24630887806415558, "learning_rate": 9.118694473981493e-05, "loss": 0.0235, "step": 23230 }, { "grad_norm": 0.2748516798019409, "learning_rate": 9.117756782099203e-05, "loss": 0.0148, "step": 23240 }, { "grad_norm": 0.2473844736814499, "learning_rate": 9.11681863990522e-05, "loss": 0.0229, "step": 23250 }, { "grad_norm": 0.27575933933258057, "learning_rate": 9.115880047502142e-05, "loss": 0.0216, "step": 23260 }, { "grad_norm": 0.36843931674957275, "learning_rate": 9.114941004992609e-05, "loss": 0.0166, "step": 23270 }, { "grad_norm": 0.24442370235919952, "learning_rate": 9.114001512479317e-05, "loss": 0.0206, "step": 23280 }, { "grad_norm": 0.2747739255428314, "learning_rate": 9.113061570065003e-05, "loss": 0.0221, "step": 23290 }, { "grad_norm": 0.27011168003082275, "learning_rate": 9.112121177852459e-05, "loss": 0.0221, "step": 23300 }, { "grad_norm": 0.3184165060520172, "learning_rate": 9.111180335944527e-05, "loss": 0.0178, "step": 23310 }, { "grad_norm": 0.674570620059967, "learning_rate": 9.110239044444093e-05, "loss": 0.0195, "step": 23320 }, { "grad_norm": 0.2572823762893677, "learning_rate": 9.109297303454099e-05, "loss": 0.0187, "step": 23330 }, { "grad_norm": 0.26490822434425354, "learning_rate": 9.108355113077526e-05, "loss": 0.0167, "step": 23340 }, { "grad_norm": 0.3670174777507782, "learning_rate": 9.107412473417419e-05, "loss": 0.0259, "step": 23350 }, { "grad_norm": 0.24451079964637756, "learning_rate": 9.106469384576858e-05, "loss": 0.0177, "step": 23360 }, { "grad_norm": 0.2689421474933624, "learning_rate": 9.105525846658978e-05, "loss": 0.0213, "step": 23370 }, { "grad_norm": 0.228340283036232, "learning_rate": 9.104581859766965e-05, "loss": 0.0202, "step": 23380 }, { "grad_norm": 0.3908279240131378, "learning_rate": 9.10363742400405e-05, "loss": 0.0257, "step": 23390 }, { "grad_norm": 0.24830448627471924, "learning_rate": 9.102692539473518e-05, "loss": 0.0177, "step": 23400 }, { "grad_norm": 0.20695652067661285, "learning_rate": 9.101747206278697e-05, "loss": 0.0254, "step": 23410 }, { "grad_norm": 0.22963964939117432, "learning_rate": 9.100801424522968e-05, "loss": 0.0165, "step": 23420 }, { "grad_norm": 0.2977107763290405, "learning_rate": 9.099855194309762e-05, "loss": 0.0226, "step": 23430 }, { "grad_norm": 0.33069348335266113, "learning_rate": 9.098908515742554e-05, "loss": 0.0199, "step": 23440 }, { "grad_norm": 0.43307286500930786, "learning_rate": 9.097961388924873e-05, "loss": 0.0187, "step": 23450 }, { "grad_norm": 0.205771803855896, "learning_rate": 9.097013813960298e-05, "loss": 0.0219, "step": 23460 }, { "grad_norm": 0.24840746819972992, "learning_rate": 9.09606579095245e-05, "loss": 0.018, "step": 23470 }, { "grad_norm": 0.2139492630958557, "learning_rate": 9.095117320005008e-05, "loss": 0.0177, "step": 23480 }, { "grad_norm": 0.21040402352809906, "learning_rate": 9.094168401221691e-05, "loss": 0.0175, "step": 23490 }, { "grad_norm": 0.21502991020679474, "learning_rate": 9.093219034706273e-05, "loss": 0.0169, "step": 23500 }, { "grad_norm": 0.2551472783088684, "learning_rate": 9.092269220562577e-05, "loss": 0.0159, "step": 23510 }, { "grad_norm": 0.2823929488658905, "learning_rate": 9.09131895889447e-05, "loss": 0.0161, "step": 23520 }, { "grad_norm": 0.710692286491394, "learning_rate": 9.090368249805873e-05, "loss": 0.0245, "step": 23530 }, { "grad_norm": 0.26280102133750916, "learning_rate": 9.089417093400754e-05, "loss": 0.0195, "step": 23540 }, { "grad_norm": 0.24499653279781342, "learning_rate": 9.088465489783131e-05, "loss": 0.0198, "step": 23550 }, { "grad_norm": 0.22973202168941498, "learning_rate": 9.087513439057068e-05, "loss": 0.0184, "step": 23560 }, { "grad_norm": 0.22274737060070038, "learning_rate": 9.08656094132668e-05, "loss": 0.0237, "step": 23570 }, { "grad_norm": 0.22812782227993011, "learning_rate": 9.085607996696134e-05, "loss": 0.0189, "step": 23580 }, { "grad_norm": 0.3129557967185974, "learning_rate": 9.084654605269639e-05, "loss": 0.0206, "step": 23590 }, { "grad_norm": 0.28310316801071167, "learning_rate": 9.083700767151457e-05, "loss": 0.0184, "step": 23600 }, { "grad_norm": 0.23870491981506348, "learning_rate": 9.082746482445898e-05, "loss": 0.0169, "step": 23610 }, { "grad_norm": 0.32473164796829224, "learning_rate": 9.081791751257325e-05, "loss": 0.0169, "step": 23620 }, { "grad_norm": 0.2782258987426758, "learning_rate": 9.080836573690142e-05, "loss": 0.0154, "step": 23630 }, { "grad_norm": 0.26473528146743774, "learning_rate": 9.079880949848805e-05, "loss": 0.022, "step": 23640 }, { "grad_norm": 0.26748955249786377, "learning_rate": 9.078924879837822e-05, "loss": 0.0251, "step": 23650 }, { "grad_norm": 0.23124520480632782, "learning_rate": 9.077968363761747e-05, "loss": 0.0195, "step": 23660 }, { "grad_norm": 0.21552854776382446, "learning_rate": 9.077011401725182e-05, "loss": 0.0196, "step": 23670 }, { "grad_norm": 0.28524571657180786, "learning_rate": 9.07605399383278e-05, "loss": 0.017, "step": 23680 }, { "grad_norm": 0.28382691740989685, "learning_rate": 9.075096140189243e-05, "loss": 0.0177, "step": 23690 }, { "grad_norm": 0.25789469480514526, "learning_rate": 9.074137840899318e-05, "loss": 0.0181, "step": 23700 }, { "grad_norm": 0.2052384465932846, "learning_rate": 9.073179096067804e-05, "loss": 0.018, "step": 23710 }, { "grad_norm": 0.30945855379104614, "learning_rate": 9.072219905799549e-05, "loss": 0.023, "step": 23720 }, { "grad_norm": 0.21366892755031586, "learning_rate": 9.071260270199447e-05, "loss": 0.0284, "step": 23730 }, { "grad_norm": 0.2466370016336441, "learning_rate": 9.070300189372441e-05, "loss": 0.0206, "step": 23740 }, { "grad_norm": 0.2732536196708679, "learning_rate": 9.069339663423528e-05, "loss": 0.025, "step": 23750 }, { "grad_norm": 0.3006916046142578, "learning_rate": 9.068378692457747e-05, "loss": 0.0198, "step": 23760 }, { "grad_norm": 0.2127247452735901, "learning_rate": 9.067417276580189e-05, "loss": 0.0187, "step": 23770 }, { "grad_norm": 0.38126009702682495, "learning_rate": 9.066455415895993e-05, "loss": 0.0201, "step": 23780 }, { "grad_norm": 0.23765406012535095, "learning_rate": 9.065493110510346e-05, "loss": 0.023, "step": 23790 }, { "grad_norm": 0.4051819145679474, "learning_rate": 9.064530360528484e-05, "loss": 0.0211, "step": 23800 }, { "grad_norm": 0.20551727712154388, "learning_rate": 9.063567166055695e-05, "loss": 0.0213, "step": 23810 }, { "grad_norm": 0.22797903418540955, "learning_rate": 9.062603527197308e-05, "loss": 0.0161, "step": 23820 }, { "grad_norm": 0.17157123982906342, "learning_rate": 9.06163944405871e-05, "loss": 0.0174, "step": 23830 }, { "grad_norm": 0.27017056941986084, "learning_rate": 9.060674916745327e-05, "loss": 0.0163, "step": 23840 }, { "grad_norm": 0.20459602773189545, "learning_rate": 9.05970994536264e-05, "loss": 0.0233, "step": 23850 }, { "grad_norm": 6.89529275894165, "learning_rate": 9.05874453001618e-05, "loss": 0.0242, "step": 23860 }, { "grad_norm": 0.4499189257621765, "learning_rate": 9.057778670811517e-05, "loss": 0.0255, "step": 23870 }, { "grad_norm": 0.2770937979221344, "learning_rate": 9.056812367854281e-05, "loss": 0.0248, "step": 23880 }, { "grad_norm": 0.2693723738193512, "learning_rate": 9.055845621250143e-05, "loss": 0.0218, "step": 23890 }, { "grad_norm": 0.5443383455276489, "learning_rate": 9.054878431104825e-05, "loss": 0.0174, "step": 23900 }, { "grad_norm": 0.309083491563797, "learning_rate": 9.0539107975241e-05, "loss": 0.0219, "step": 23910 }, { "grad_norm": 0.2757817804813385, "learning_rate": 9.052942720613784e-05, "loss": 0.0273, "step": 23920 }, { "grad_norm": 0.18415193259716034, "learning_rate": 9.051974200479745e-05, "loss": 0.0159, "step": 23930 }, { "grad_norm": 0.2664022445678711, "learning_rate": 9.051005237227901e-05, "loss": 0.0183, "step": 23940 }, { "grad_norm": 0.24871709942817688, "learning_rate": 9.050035830964215e-05, "loss": 0.0157, "step": 23950 }, { "grad_norm": 0.2650960087776184, "learning_rate": 9.049065981794698e-05, "loss": 0.0157, "step": 23960 }, { "grad_norm": 0.25774431228637695, "learning_rate": 9.048095689825414e-05, "loss": 0.0139, "step": 23970 }, { "grad_norm": 0.26602107286453247, "learning_rate": 9.047124955162472e-05, "loss": 0.0277, "step": 23980 }, { "grad_norm": 0.28781598806381226, "learning_rate": 9.046153777912028e-05, "loss": 0.0273, "step": 23990 }, { "grad_norm": 0.31784147024154663, "learning_rate": 9.045182158180292e-05, "loss": 0.0295, "step": 24000 }, { "grad_norm": 0.19529899954795837, "learning_rate": 9.044210096073516e-05, "loss": 0.0193, "step": 24010 }, { "grad_norm": 0.24138545989990234, "learning_rate": 9.043237591698004e-05, "loss": 0.0207, "step": 24020 }, { "grad_norm": 0.2889857590198517, "learning_rate": 9.04226464516011e-05, "loss": 0.0163, "step": 24030 }, { "grad_norm": 0.24966497719287872, "learning_rate": 9.041291256566229e-05, "loss": 0.0163, "step": 24040 }, { "grad_norm": 0.14166219532489777, "learning_rate": 9.040317426022814e-05, "loss": 0.0134, "step": 24050 }, { "grad_norm": 0.1746325045824051, "learning_rate": 9.03934315363636e-05, "loss": 0.0135, "step": 24060 }, { "grad_norm": 0.14825233817100525, "learning_rate": 9.038368439513409e-05, "loss": 0.0126, "step": 24070 }, { "grad_norm": 0.2894132435321808, "learning_rate": 9.03739328376056e-05, "loss": 0.0138, "step": 24080 }, { "grad_norm": 0.2686930298805237, "learning_rate": 9.036417686484451e-05, "loss": 0.0141, "step": 24090 }, { "grad_norm": 0.45069777965545654, "learning_rate": 9.035441647791773e-05, "loss": 0.0121, "step": 24100 }, { "grad_norm": 0.20367003977298737, "learning_rate": 9.034465167789263e-05, "loss": 0.0131, "step": 24110 }, { "grad_norm": 0.21050019562244415, "learning_rate": 9.033488246583706e-05, "loss": 0.0147, "step": 24120 }, { "grad_norm": 0.2066519856452942, "learning_rate": 9.032510884281941e-05, "loss": 0.0124, "step": 24130 }, { "grad_norm": 0.14681760966777802, "learning_rate": 9.031533080990848e-05, "loss": 0.0163, "step": 24140 }, { "grad_norm": 0.20650209486484528, "learning_rate": 9.030554836817358e-05, "loss": 0.0143, "step": 24150 }, { "grad_norm": 0.1915746033191681, "learning_rate": 9.029576151868451e-05, "loss": 0.0142, "step": 24160 }, { "grad_norm": 0.17203529179096222, "learning_rate": 9.028597026251155e-05, "loss": 0.0145, "step": 24170 }, { "grad_norm": 0.19277405738830566, "learning_rate": 9.027617460072547e-05, "loss": 0.0141, "step": 24180 }, { "grad_norm": 0.21428155899047852, "learning_rate": 9.026637453439745e-05, "loss": 0.019, "step": 24190 }, { "grad_norm": 0.36146822571754456, "learning_rate": 9.025657006459927e-05, "loss": 0.0173, "step": 24200 }, { "grad_norm": 0.24829474091529846, "learning_rate": 9.024676119240311e-05, "loss": 0.0196, "step": 24210 }, { "grad_norm": 0.2019103467464447, "learning_rate": 9.023694791888166e-05, "loss": 0.0209, "step": 24220 }, { "grad_norm": 0.20815663039684296, "learning_rate": 9.022713024510808e-05, "loss": 0.0213, "step": 24230 }, { "grad_norm": 0.2705768346786499, "learning_rate": 9.021730817215601e-05, "loss": 0.0206, "step": 24240 }, { "grad_norm": 0.4143127202987671, "learning_rate": 9.02074817010996e-05, "loss": 0.0222, "step": 24250 }, { "grad_norm": 0.33608752489089966, "learning_rate": 9.019765083301342e-05, "loss": 0.0278, "step": 24260 }, { "grad_norm": 0.7163955569267273, "learning_rate": 9.01878155689726e-05, "loss": 0.0247, "step": 24270 }, { "grad_norm": 0.20514234900474548, "learning_rate": 9.017797591005268e-05, "loss": 0.0237, "step": 24280 }, { "grad_norm": 0.4339626133441925, "learning_rate": 9.016813185732972e-05, "loss": 0.0197, "step": 24290 }, { "grad_norm": 0.32475224137306213, "learning_rate": 9.015828341188027e-05, "loss": 0.018, "step": 24300 }, { "grad_norm": 0.15026463568210602, "learning_rate": 9.01484305747813e-05, "loss": 0.0166, "step": 24310 }, { "grad_norm": 0.20271052420139313, "learning_rate": 9.013857334711033e-05, "loss": 0.0268, "step": 24320 }, { "grad_norm": 0.20027481019496918, "learning_rate": 9.012871172994534e-05, "loss": 0.0192, "step": 24330 }, { "grad_norm": 0.2654629349708557, "learning_rate": 9.011884572436476e-05, "loss": 0.0221, "step": 24340 }, { "grad_norm": 0.2182600051164627, "learning_rate": 9.010897533144754e-05, "loss": 0.023, "step": 24350 }, { "grad_norm": 0.23860840499401093, "learning_rate": 9.009910055227306e-05, "loss": 0.0195, "step": 24360 }, { "grad_norm": 0.20513533055782318, "learning_rate": 9.008922138792124e-05, "loss": 0.0147, "step": 24370 }, { "grad_norm": 0.28189051151275635, "learning_rate": 9.007933783947244e-05, "loss": 0.0179, "step": 24380 }, { "grad_norm": 0.23342014849185944, "learning_rate": 9.006944990800752e-05, "loss": 0.0269, "step": 24390 }, { "grad_norm": 0.23631596565246582, "learning_rate": 9.005955759460779e-05, "loss": 0.0192, "step": 24400 }, { "grad_norm": 0.24846786260604858, "learning_rate": 9.004966090035508e-05, "loss": 0.0169, "step": 24410 }, { "grad_norm": 0.24962809681892395, "learning_rate": 9.003975982633166e-05, "loss": 0.0193, "step": 24420 }, { "grad_norm": 0.23601530492305756, "learning_rate": 9.00298543736203e-05, "loss": 0.018, "step": 24430 }, { "grad_norm": 0.2011566460132599, "learning_rate": 9.001994454330427e-05, "loss": 0.0182, "step": 24440 }, { "grad_norm": 0.2605632543563843, "learning_rate": 9.001003033646727e-05, "loss": 0.0193, "step": 24450 }, { "grad_norm": 0.37375950813293457, "learning_rate": 9.00001117541935e-05, "loss": 0.0195, "step": 24460 }, { "grad_norm": 0.18244829773902893, "learning_rate": 8.999018879756764e-05, "loss": 0.0195, "step": 24470 }, { "grad_norm": 1.1365798711776733, "learning_rate": 8.998026146767487e-05, "loss": 0.0177, "step": 24480 }, { "grad_norm": 0.2288869321346283, "learning_rate": 8.99703297656008e-05, "loss": 0.0189, "step": 24490 }, { "grad_norm": 0.2007974237203598, "learning_rate": 8.996039369243156e-05, "loss": 0.0158, "step": 24500 }, { "grad_norm": 0.27232202887535095, "learning_rate": 8.995045324925378e-05, "loss": 0.014, "step": 24510 }, { "grad_norm": 0.23424501717090607, "learning_rate": 8.994050843715448e-05, "loss": 0.0207, "step": 24520 }, { "grad_norm": 0.1841011345386505, "learning_rate": 8.993055925722121e-05, "loss": 0.0187, "step": 24530 }, { "grad_norm": 0.3130970001220703, "learning_rate": 8.992060571054202e-05, "loss": 0.0236, "step": 24540 }, { "grad_norm": 0.23781345784664154, "learning_rate": 8.991064779820542e-05, "loss": 0.0161, "step": 24550 }, { "grad_norm": 0.30447885394096375, "learning_rate": 8.990068552130036e-05, "loss": 0.0183, "step": 24560 }, { "grad_norm": 0.23506107926368713, "learning_rate": 8.989071888091634e-05, "loss": 0.0148, "step": 24570 }, { "grad_norm": 0.24427463114261627, "learning_rate": 8.988074787814329e-05, "loss": 0.0167, "step": 24580 }, { "grad_norm": 0.21162033081054688, "learning_rate": 8.987077251407158e-05, "loss": 0.0174, "step": 24590 }, { "grad_norm": 0.24398614466190338, "learning_rate": 8.986079278979216e-05, "loss": 0.0205, "step": 24600 }, { "grad_norm": 0.2267903834581375, "learning_rate": 8.985080870639635e-05, "loss": 0.0201, "step": 24610 }, { "grad_norm": 0.2234274446964264, "learning_rate": 8.984082026497603e-05, "loss": 0.0173, "step": 24620 }, { "grad_norm": 0.22193588316440582, "learning_rate": 8.98308274666235e-05, "loss": 0.017, "step": 24630 }, { "grad_norm": 0.24326927959918976, "learning_rate": 8.982083031243155e-05, "loss": 0.0162, "step": 24640 }, { "grad_norm": 0.46380648016929626, "learning_rate": 8.98108288034935e-05, "loss": 0.0142, "step": 24650 }, { "grad_norm": 0.21847295761108398, "learning_rate": 8.980082294090305e-05, "loss": 0.0221, "step": 24660 }, { "grad_norm": 0.1795979142189026, "learning_rate": 8.979081272575443e-05, "loss": 0.0158, "step": 24670 }, { "grad_norm": 0.18434005975723267, "learning_rate": 8.978079815914236e-05, "loss": 0.017, "step": 24680 }, { "grad_norm": 0.22577717900276184, "learning_rate": 8.977077924216202e-05, "loss": 0.0186, "step": 24690 }, { "grad_norm": 0.830283522605896, "learning_rate": 8.976075597590905e-05, "loss": 0.0201, "step": 24700 }, { "grad_norm": 0.22905708849430084, "learning_rate": 8.975072836147958e-05, "loss": 0.0185, "step": 24710 }, { "grad_norm": 0.19114725291728973, "learning_rate": 8.974069639997025e-05, "loss": 0.0249, "step": 24720 }, { "grad_norm": 0.18992318212985992, "learning_rate": 8.973066009247808e-05, "loss": 0.015, "step": 24730 }, { "grad_norm": 0.23831212520599365, "learning_rate": 8.972061944010066e-05, "loss": 0.0208, "step": 24740 }, { "grad_norm": 0.19280293583869934, "learning_rate": 8.971057444393603e-05, "loss": 0.0247, "step": 24750 }, { "grad_norm": 0.2099325954914093, "learning_rate": 8.970052510508268e-05, "loss": 0.0145, "step": 24760 }, { "grad_norm": 0.28043314814567566, "learning_rate": 8.969047142463959e-05, "loss": 0.0217, "step": 24770 }, { "grad_norm": 0.31689006090164185, "learning_rate": 8.968041340370621e-05, "loss": 0.0167, "step": 24780 }, { "grad_norm": 0.6600968837738037, "learning_rate": 8.96703510433825e-05, "loss": 0.0191, "step": 24790 }, { "grad_norm": 0.1976410448551178, "learning_rate": 8.966028434476883e-05, "loss": 0.0201, "step": 24800 }, { "grad_norm": 0.15214480459690094, "learning_rate": 8.96502133089661e-05, "loss": 0.0137, "step": 24810 }, { "grad_norm": 0.1902557909488678, "learning_rate": 8.964013793707564e-05, "loss": 0.0155, "step": 24820 }, { "grad_norm": 0.30783239006996155, "learning_rate": 8.963005823019932e-05, "loss": 0.0191, "step": 24830 }, { "grad_norm": 0.2708354890346527, "learning_rate": 8.961997418943939e-05, "loss": 0.0196, "step": 24840 }, { "grad_norm": 0.27264291048049927, "learning_rate": 8.960988581589865e-05, "loss": 0.0162, "step": 24850 }, { "grad_norm": 0.37427249550819397, "learning_rate": 8.959979311068037e-05, "loss": 0.0224, "step": 24860 }, { "grad_norm": 0.22117125988006592, "learning_rate": 8.958969607488823e-05, "loss": 0.0151, "step": 24870 }, { "grad_norm": 0.22502581775188446, "learning_rate": 8.957959470962647e-05, "loss": 0.0157, "step": 24880 }, { "grad_norm": 0.29376330971717834, "learning_rate": 8.956948901599971e-05, "loss": 0.0173, "step": 24890 }, { "grad_norm": 0.21139588952064514, "learning_rate": 8.955937899511315e-05, "loss": 0.0122, "step": 24900 }, { "grad_norm": 0.37818604707717896, "learning_rate": 8.954926464807238e-05, "loss": 0.0177, "step": 24910 }, { "grad_norm": 0.21154113113880157, "learning_rate": 8.953914597598347e-05, "loss": 0.0141, "step": 24920 }, { "grad_norm": 0.27206218242645264, "learning_rate": 8.952902297995303e-05, "loss": 0.0177, "step": 24930 }, { "grad_norm": 0.24604225158691406, "learning_rate": 8.951889566108804e-05, "loss": 0.0166, "step": 24940 }, { "grad_norm": 0.438059538602829, "learning_rate": 8.950876402049606e-05, "loss": 0.0182, "step": 24950 }, { "grad_norm": 1.9703102111816406, "learning_rate": 8.949862805928504e-05, "loss": 0.0163, "step": 24960 }, { "grad_norm": 0.2564370632171631, "learning_rate": 8.948848777856343e-05, "loss": 0.0246, "step": 24970 }, { "grad_norm": 0.28269097208976746, "learning_rate": 8.947834317944017e-05, "loss": 0.0171, "step": 24980 }, { "grad_norm": 0.37344399094581604, "learning_rate": 8.946819426302466e-05, "loss": 0.0181, "step": 24990 }, { "grad_norm": 0.23204509913921356, "learning_rate": 8.945804103042676e-05, "loss": 0.0165, "step": 25000 }, { "grad_norm": 0.29343435168266296, "learning_rate": 8.944788348275681e-05, "loss": 0.0222, "step": 25010 }, { "grad_norm": 0.23140743374824524, "learning_rate": 8.943772162112565e-05, "loss": 0.0146, "step": 25020 }, { "grad_norm": 0.22610029578208923, "learning_rate": 8.942755544664454e-05, "loss": 0.0179, "step": 25030 }, { "grad_norm": 0.15579436719417572, "learning_rate": 8.941738496042525e-05, "loss": 0.0178, "step": 25040 }, { "grad_norm": 0.22019284963607788, "learning_rate": 8.940721016357999e-05, "loss": 0.0157, "step": 25050 }, { "grad_norm": 0.23117369413375854, "learning_rate": 8.939703105722148e-05, "loss": 0.0186, "step": 25060 }, { "grad_norm": 0.248391792178154, "learning_rate": 8.93868476424629e-05, "loss": 0.0154, "step": 25070 }, { "grad_norm": 0.20556780695915222, "learning_rate": 8.937665992041786e-05, "loss": 0.0142, "step": 25080 }, { "grad_norm": 0.2860621213912964, "learning_rate": 8.93664678922005e-05, "loss": 0.0163, "step": 25090 }, { "grad_norm": 0.21333731710910797, "learning_rate": 8.93562715589254e-05, "loss": 0.0164, "step": 25100 }, { "grad_norm": 1.6042940616607666, "learning_rate": 8.934607092170762e-05, "loss": 0.0178, "step": 25110 }, { "grad_norm": 0.21113592386245728, "learning_rate": 8.933586598166266e-05, "loss": 0.0173, "step": 25120 }, { "grad_norm": 0.301507830619812, "learning_rate": 8.932565673990655e-05, "loss": 0.0156, "step": 25130 }, { "grad_norm": 0.35658517479896545, "learning_rate": 8.931544319755574e-05, "loss": 0.0201, "step": 25140 }, { "grad_norm": 0.21087488532066345, "learning_rate": 8.930522535572718e-05, "loss": 0.019, "step": 25150 }, { "grad_norm": 0.23003078997135162, "learning_rate": 8.929500321553826e-05, "loss": 0.0181, "step": 25160 }, { "grad_norm": 0.203372985124588, "learning_rate": 8.928477677810686e-05, "loss": 0.014, "step": 25170 }, { "grad_norm": 0.2348824441432953, "learning_rate": 8.927454604455137e-05, "loss": 0.0143, "step": 25180 }, { "grad_norm": 0.2455693483352661, "learning_rate": 8.926431101599053e-05, "loss": 0.0142, "step": 25190 }, { "grad_norm": 0.3195975422859192, "learning_rate": 8.925407169354369e-05, "loss": 0.0184, "step": 25200 }, { "grad_norm": 0.2193295806646347, "learning_rate": 8.92438280783306e-05, "loss": 0.0174, "step": 25210 }, { "grad_norm": 0.2849385142326355, "learning_rate": 8.923358017147146e-05, "loss": 0.0207, "step": 25220 }, { "grad_norm": 0.3144809901714325, "learning_rate": 8.922332797408697e-05, "loss": 0.0146, "step": 25230 }, { "grad_norm": 0.34379392862319946, "learning_rate": 8.921307148729831e-05, "loss": 0.0147, "step": 25240 }, { "grad_norm": 0.23508964478969574, "learning_rate": 8.920281071222712e-05, "loss": 0.0179, "step": 25250 }, { "grad_norm": 0.5161640644073486, "learning_rate": 8.919254564999548e-05, "loss": 0.0148, "step": 25260 }, { "grad_norm": 0.27676671743392944, "learning_rate": 8.918227630172598e-05, "loss": 0.0183, "step": 25270 }, { "grad_norm": 0.5270224213600159, "learning_rate": 8.917200266854165e-05, "loss": 0.0174, "step": 25280 }, { "grad_norm": 0.21475711464881897, "learning_rate": 8.9161724751566e-05, "loss": 0.0186, "step": 25290 }, { "grad_norm": 0.36474868655204773, "learning_rate": 8.915144255192302e-05, "loss": 0.0211, "step": 25300 }, { "grad_norm": 0.572040855884552, "learning_rate": 8.914115607073714e-05, "loss": 0.0163, "step": 25310 }, { "grad_norm": 0.1731254607439041, "learning_rate": 8.913086530913327e-05, "loss": 0.0169, "step": 25320 }, { "grad_norm": 0.2264804095029831, "learning_rate": 8.912057026823681e-05, "loss": 0.0156, "step": 25330 }, { "grad_norm": 0.21882478892803192, "learning_rate": 8.91102709491736e-05, "loss": 0.0149, "step": 25340 }, { "grad_norm": 0.2624695897102356, "learning_rate": 8.909996735306996e-05, "loss": 0.0162, "step": 25350 }, { "grad_norm": 0.22158470749855042, "learning_rate": 8.908965948105268e-05, "loss": 0.0134, "step": 25360 }, { "grad_norm": 0.28008148074150085, "learning_rate": 8.907934733424901e-05, "loss": 0.0203, "step": 25370 }, { "grad_norm": 0.35129353404045105, "learning_rate": 8.906903091378666e-05, "loss": 0.0171, "step": 25380 }, { "grad_norm": 0.2660105228424072, "learning_rate": 8.905871022079384e-05, "loss": 0.0162, "step": 25390 }, { "grad_norm": 0.22555577754974365, "learning_rate": 8.90483852563992e-05, "loss": 0.013, "step": 25400 }, { "grad_norm": 0.2961426377296448, "learning_rate": 8.903805602173185e-05, "loss": 0.0168, "step": 25410 }, { "grad_norm": 0.2951053977012634, "learning_rate": 8.902772251792137e-05, "loss": 0.0205, "step": 25420 }, { "grad_norm": 0.27905890345573425, "learning_rate": 8.901738474609786e-05, "loss": 0.0174, "step": 25430 }, { "grad_norm": 0.3871559500694275, "learning_rate": 8.900704270739179e-05, "loss": 0.0263, "step": 25440 }, { "grad_norm": 0.35595715045928955, "learning_rate": 8.89966964029342e-05, "loss": 0.0227, "step": 25450 }, { "grad_norm": 0.27089232206344604, "learning_rate": 8.898634583385652e-05, "loss": 0.0205, "step": 25460 }, { "grad_norm": 0.2426738440990448, "learning_rate": 8.897599100129065e-05, "loss": 0.0149, "step": 25470 }, { "grad_norm": 0.3213163912296295, "learning_rate": 8.896563190636903e-05, "loss": 0.0179, "step": 25480 }, { "grad_norm": 0.30047839879989624, "learning_rate": 8.895526855022448e-05, "loss": 0.0203, "step": 25490 }, { "grad_norm": 0.4241553544998169, "learning_rate": 8.894490093399033e-05, "loss": 0.0213, "step": 25500 }, { "grad_norm": 0.23172898590564728, "learning_rate": 8.893452905880035e-05, "loss": 0.0216, "step": 25510 }, { "grad_norm": 0.2798255980014801, "learning_rate": 8.892415292578883e-05, "loss": 0.0207, "step": 25520 }, { "grad_norm": 0.28134089708328247, "learning_rate": 8.891377253609046e-05, "loss": 0.0236, "step": 25530 }, { "grad_norm": 0.33176568150520325, "learning_rate": 8.890338789084043e-05, "loss": 0.0167, "step": 25540 }, { "grad_norm": 0.4901922345161438, "learning_rate": 8.88929989911744e-05, "loss": 0.0199, "step": 25550 }, { "grad_norm": 0.24114374816417694, "learning_rate": 8.888260583822847e-05, "loss": 0.0173, "step": 25560 }, { "grad_norm": 0.2101571261882782, "learning_rate": 8.887220843313921e-05, "loss": 0.016, "step": 25570 }, { "grad_norm": 0.2607964873313904, "learning_rate": 8.88618067770437e-05, "loss": 0.0174, "step": 25580 }, { "grad_norm": 0.46448349952697754, "learning_rate": 8.885140087107942e-05, "loss": 0.017, "step": 25590 }, { "grad_norm": 0.9917487502098083, "learning_rate": 8.884099071638436e-05, "loss": 0.0166, "step": 25600 }, { "grad_norm": 0.230671688914299, "learning_rate": 8.883057631409695e-05, "loss": 0.0146, "step": 25610 }, { "grad_norm": 0.3295581042766571, "learning_rate": 8.882015766535608e-05, "loss": 0.0147, "step": 25620 }, { "grad_norm": 0.22446171939373016, "learning_rate": 8.880973477130115e-05, "loss": 0.017, "step": 25630 }, { "grad_norm": 0.7358022332191467, "learning_rate": 8.879930763307197e-05, "loss": 0.0159, "step": 25640 }, { "grad_norm": 0.1751144975423813, "learning_rate": 8.878887625180884e-05, "loss": 0.0215, "step": 25650 }, { "grad_norm": 0.24536773562431335, "learning_rate": 8.877844062865253e-05, "loss": 0.0153, "step": 25660 }, { "grad_norm": 0.21858851611614227, "learning_rate": 8.876800076474424e-05, "loss": 0.0158, "step": 25670 }, { "grad_norm": 1.120990514755249, "learning_rate": 8.875755666122568e-05, "loss": 0.0158, "step": 25680 }, { "grad_norm": 0.16231669485569, "learning_rate": 8.8747108319239e-05, "loss": 0.0146, "step": 25690 }, { "grad_norm": 0.48482745885849, "learning_rate": 8.87366557399268e-05, "loss": 0.0242, "step": 25700 }, { "grad_norm": 0.2172406166791916, "learning_rate": 8.872619892443217e-05, "loss": 0.0151, "step": 25710 }, { "grad_norm": 1.0041569471359253, "learning_rate": 8.871573787389865e-05, "loss": 0.0218, "step": 25720 }, { "grad_norm": 0.9552342295646667, "learning_rate": 8.870527258947024e-05, "loss": 0.0171, "step": 25730 }, { "grad_norm": 0.2811117172241211, "learning_rate": 8.869480307229143e-05, "loss": 0.0205, "step": 25740 }, { "grad_norm": 0.27286821603775024, "learning_rate": 8.868432932350712e-05, "loss": 0.0193, "step": 25750 }, { "grad_norm": 0.2397620975971222, "learning_rate": 8.867385134426272e-05, "loss": 0.0192, "step": 25760 }, { "grad_norm": 0.20329155027866364, "learning_rate": 8.866336913570407e-05, "loss": 0.0156, "step": 25770 }, { "grad_norm": 0.24242474138736725, "learning_rate": 8.865288269897751e-05, "loss": 0.0227, "step": 25780 }, { "grad_norm": 0.2049749791622162, "learning_rate": 8.864239203522981e-05, "loss": 0.0154, "step": 25790 }, { "grad_norm": 0.20621436834335327, "learning_rate": 8.863189714560822e-05, "loss": 0.0177, "step": 25800 }, { "grad_norm": 0.267221599817276, "learning_rate": 8.862139803126043e-05, "loss": 0.0169, "step": 25810 }, { "grad_norm": 0.22629013657569885, "learning_rate": 8.861089469333463e-05, "loss": 0.0221, "step": 25820 }, { "grad_norm": 1.1954926252365112, "learning_rate": 8.860038713297944e-05, "loss": 0.0206, "step": 25830 }, { "grad_norm": 0.27728697657585144, "learning_rate": 8.858987535134394e-05, "loss": 0.0204, "step": 25840 }, { "grad_norm": 0.24234537780284882, "learning_rate": 8.857935934957769e-05, "loss": 0.0181, "step": 25850 }, { "grad_norm": 0.23012088239192963, "learning_rate": 8.856883912883071e-05, "loss": 0.0175, "step": 25860 }, { "grad_norm": 0.5957924723625183, "learning_rate": 8.855831469025346e-05, "loss": 0.0187, "step": 25870 }, { "grad_norm": 0.24835412204265594, "learning_rate": 8.854778603499689e-05, "loss": 0.0179, "step": 25880 }, { "grad_norm": 0.1650262326002121, "learning_rate": 8.85372531642124e-05, "loss": 0.0202, "step": 25890 }, { "grad_norm": 0.31785672903060913, "learning_rate": 8.852671607905185e-05, "loss": 0.0196, "step": 25900 }, { "grad_norm": 0.1627187579870224, "learning_rate": 8.851617478066754e-05, "loss": 0.0158, "step": 25910 }, { "grad_norm": 0.1639811247587204, "learning_rate": 8.850562927021227e-05, "loss": 0.013, "step": 25920 }, { "grad_norm": 0.18394315242767334, "learning_rate": 8.849507954883928e-05, "loss": 0.0182, "step": 25930 }, { "grad_norm": 0.14771004021167755, "learning_rate": 8.848452561770226e-05, "loss": 0.0152, "step": 25940 }, { "grad_norm": 0.5586133599281311, "learning_rate": 8.847396747795538e-05, "loss": 0.0154, "step": 25950 }, { "grad_norm": 0.1931462436914444, "learning_rate": 8.846340513075327e-05, "loss": 0.0163, "step": 25960 }, { "grad_norm": 0.2736116647720337, "learning_rate": 8.845283857725099e-05, "loss": 0.0205, "step": 25970 }, { "grad_norm": 0.1575614959001541, "learning_rate": 8.844226781860409e-05, "loss": 0.015, "step": 25980 }, { "grad_norm": 0.7883267402648926, "learning_rate": 8.84316928559686e-05, "loss": 0.0218, "step": 25990 }, { "grad_norm": 0.24075224995613098, "learning_rate": 8.842111369050094e-05, "loss": 0.0176, "step": 26000 }, { "grad_norm": 0.5103625059127808, "learning_rate": 8.841053032335808e-05, "loss": 0.0194, "step": 26010 }, { "grad_norm": 0.29610103368759155, "learning_rate": 8.839994275569735e-05, "loss": 0.0197, "step": 26020 }, { "grad_norm": 0.2788083255290985, "learning_rate": 8.838935098867662e-05, "loss": 0.0214, "step": 26030 }, { "grad_norm": 0.7803364992141724, "learning_rate": 8.837875502345418e-05, "loss": 0.0175, "step": 26040 }, { "grad_norm": 0.2156207263469696, "learning_rate": 8.83681548611888e-05, "loss": 0.0136, "step": 26050 }, { "grad_norm": 0.21690510213375092, "learning_rate": 8.835755050303969e-05, "loss": 0.0182, "step": 26060 }, { "grad_norm": 0.2076306939125061, "learning_rate": 8.834694195016653e-05, "loss": 0.0161, "step": 26070 }, { "grad_norm": 0.22071154415607452, "learning_rate": 8.833632920372942e-05, "loss": 0.0181, "step": 26080 }, { "grad_norm": 0.26014837622642517, "learning_rate": 8.832571226488903e-05, "loss": 0.0183, "step": 26090 }, { "grad_norm": 0.19362524151802063, "learning_rate": 8.831509113480634e-05, "loss": 0.0213, "step": 26100 }, { "grad_norm": 0.32247161865234375, "learning_rate": 8.83044658146429e-05, "loss": 0.0178, "step": 26110 }, { "grad_norm": 0.2321750968694687, "learning_rate": 8.829383630556067e-05, "loss": 0.0151, "step": 26120 }, { "grad_norm": 0.3079359531402588, "learning_rate": 8.828320260872207e-05, "loss": 0.0132, "step": 26130 }, { "grad_norm": 0.17463207244873047, "learning_rate": 8.827256472529e-05, "loss": 0.0133, "step": 26140 }, { "grad_norm": 0.28256186842918396, "learning_rate": 8.826192265642778e-05, "loss": 0.0157, "step": 26150 }, { "grad_norm": 0.1940022110939026, "learning_rate": 8.825127640329923e-05, "loss": 0.0195, "step": 26160 }, { "grad_norm": 0.23914456367492676, "learning_rate": 8.824062596706861e-05, "loss": 0.0171, "step": 26170 }, { "grad_norm": 0.228399395942688, "learning_rate": 8.822997134890062e-05, "loss": 0.0209, "step": 26180 }, { "grad_norm": 0.17689849436283112, "learning_rate": 8.821931254996044e-05, "loss": 0.0152, "step": 26190 }, { "grad_norm": 0.19622866809368134, "learning_rate": 8.82086495714137e-05, "loss": 0.0163, "step": 26200 }, { "grad_norm": 0.17262694239616394, "learning_rate": 8.81979824144265e-05, "loss": 0.0203, "step": 26210 }, { "grad_norm": 0.20143386721611023, "learning_rate": 8.818731108016536e-05, "loss": 0.0182, "step": 26220 }, { "grad_norm": 0.2570212781429291, "learning_rate": 8.81766355697973e-05, "loss": 0.0181, "step": 26230 }, { "grad_norm": 0.19531649351119995, "learning_rate": 8.816595588448977e-05, "loss": 0.015, "step": 26240 }, { "grad_norm": 0.3083621859550476, "learning_rate": 8.81552720254107e-05, "loss": 0.017, "step": 26250 }, { "grad_norm": 0.2407589554786682, "learning_rate": 8.814458399372842e-05, "loss": 0.0154, "step": 26260 }, { "grad_norm": 0.39793097972869873, "learning_rate": 8.813389179061181e-05, "loss": 0.0204, "step": 26270 }, { "grad_norm": 0.16185332834720612, "learning_rate": 8.812319541723012e-05, "loss": 0.0139, "step": 26280 }, { "grad_norm": 0.25193625688552856, "learning_rate": 8.811249487475309e-05, "loss": 0.0175, "step": 26290 }, { "grad_norm": 0.1919015347957611, "learning_rate": 8.810179016435092e-05, "loss": 0.0143, "step": 26300 }, { "grad_norm": 0.23852364718914032, "learning_rate": 8.809108128719428e-05, "loss": 0.0176, "step": 26310 }, { "grad_norm": 0.2047724425792694, "learning_rate": 8.808036824445424e-05, "loss": 0.0145, "step": 26320 }, { "grad_norm": 0.32164907455444336, "learning_rate": 8.806965103730238e-05, "loss": 0.0194, "step": 26330 }, { "grad_norm": 0.19646640121936798, "learning_rate": 8.805892966691074e-05, "loss": 0.0157, "step": 26340 }, { "grad_norm": 0.24989834427833557, "learning_rate": 8.804820413445175e-05, "loss": 0.0206, "step": 26350 }, { "grad_norm": 0.2496786266565323, "learning_rate": 8.803747444109837e-05, "loss": 0.018, "step": 26360 }, { "grad_norm": 0.32112792134284973, "learning_rate": 8.802674058802399e-05, "loss": 0.0173, "step": 26370 }, { "grad_norm": 0.2567668557167053, "learning_rate": 8.801600257640241e-05, "loss": 0.0185, "step": 26380 }, { "grad_norm": 0.24440079927444458, "learning_rate": 8.800526040740795e-05, "loss": 0.0171, "step": 26390 }, { "grad_norm": 0.2211829572916031, "learning_rate": 8.799451408221535e-05, "loss": 0.0163, "step": 26400 }, { "grad_norm": 4.540335655212402, "learning_rate": 8.798376360199982e-05, "loss": 0.0193, "step": 26410 }, { "grad_norm": 0.18100588023662567, "learning_rate": 8.797300896793701e-05, "loss": 0.0147, "step": 26420 }, { "grad_norm": 0.2570970356464386, "learning_rate": 8.796225018120302e-05, "loss": 0.0128, "step": 26430 }, { "grad_norm": 0.21032287180423737, "learning_rate": 8.795148724297444e-05, "loss": 0.0167, "step": 26440 }, { "grad_norm": 0.6322620511054993, "learning_rate": 8.794072015442825e-05, "loss": 0.0188, "step": 26450 }, { "grad_norm": 0.170518159866333, "learning_rate": 8.792994891674198e-05, "loss": 0.017, "step": 26460 }, { "grad_norm": 0.5769233703613281, "learning_rate": 8.79191735310935e-05, "loss": 0.0151, "step": 26470 }, { "grad_norm": 0.19738104939460754, "learning_rate": 8.790839399866122e-05, "loss": 0.0145, "step": 26480 }, { "grad_norm": 0.18298426270484924, "learning_rate": 8.789761032062397e-05, "loss": 0.0177, "step": 26490 }, { "grad_norm": 0.2663392424583435, "learning_rate": 8.788682249816103e-05, "loss": 0.0188, "step": 26500 }, { "grad_norm": 0.7127411365509033, "learning_rate": 8.787603053245215e-05, "loss": 0.0175, "step": 26510 }, { "grad_norm": 0.17314037680625916, "learning_rate": 8.78652344246775e-05, "loss": 0.0177, "step": 26520 }, { "grad_norm": 0.236983060836792, "learning_rate": 8.785443417601776e-05, "loss": 0.0216, "step": 26530 }, { "grad_norm": 0.4046250283718109, "learning_rate": 8.784362978765401e-05, "loss": 0.0207, "step": 26540 }, { "grad_norm": 0.222768634557724, "learning_rate": 8.783282126076779e-05, "loss": 0.0186, "step": 26550 }, { "grad_norm": 0.17607338726520538, "learning_rate": 8.782200859654112e-05, "loss": 0.0174, "step": 26560 }, { "grad_norm": 0.2171792984008789, "learning_rate": 8.781119179615646e-05, "loss": 0.0157, "step": 26570 }, { "grad_norm": 0.23903320729732513, "learning_rate": 8.780037086079674e-05, "loss": 0.02, "step": 26580 }, { "grad_norm": 0.15525610744953156, "learning_rate": 8.778954579164527e-05, "loss": 0.0189, "step": 26590 }, { "grad_norm": 0.24895235896110535, "learning_rate": 8.777871658988588e-05, "loss": 0.0178, "step": 26600 }, { "grad_norm": 0.26664435863494873, "learning_rate": 8.776788325670285e-05, "loss": 0.0172, "step": 26610 }, { "grad_norm": 0.1641007363796234, "learning_rate": 8.775704579328089e-05, "loss": 0.017, "step": 26620 }, { "grad_norm": 0.572363018989563, "learning_rate": 8.774620420080517e-05, "loss": 0.0184, "step": 26630 }, { "grad_norm": 0.26257628202438354, "learning_rate": 8.773535848046131e-05, "loss": 0.0191, "step": 26640 }, { "grad_norm": 0.5463648438453674, "learning_rate": 8.772450863343538e-05, "loss": 0.0188, "step": 26650 }, { "grad_norm": 0.14581914246082306, "learning_rate": 8.77136546609139e-05, "loss": 0.0161, "step": 26660 }, { "grad_norm": 0.27773404121398926, "learning_rate": 8.770279656408385e-05, "loss": 0.0168, "step": 26670 }, { "grad_norm": 0.21598254144191742, "learning_rate": 8.769193434413265e-05, "loss": 0.018, "step": 26680 }, { "grad_norm": 0.3081910014152527, "learning_rate": 8.76810680022482e-05, "loss": 0.017, "step": 26690 }, { "grad_norm": 0.26531580090522766, "learning_rate": 8.767019753961878e-05, "loss": 0.0188, "step": 26700 }, { "grad_norm": 0.6660181283950806, "learning_rate": 8.765932295743321e-05, "loss": 0.0173, "step": 26710 }, { "grad_norm": 0.20751728117465973, "learning_rate": 8.764844425688068e-05, "loss": 0.0194, "step": 26720 }, { "grad_norm": 0.552416980266571, "learning_rate": 8.763756143915092e-05, "loss": 0.0185, "step": 26730 }, { "grad_norm": 0.22027245163917542, "learning_rate": 8.7626674505434e-05, "loss": 0.0183, "step": 26740 }, { "grad_norm": 0.3139486014842987, "learning_rate": 8.761578345692053e-05, "loss": 0.0195, "step": 26750 }, { "grad_norm": 0.2805674076080322, "learning_rate": 8.760488829480156e-05, "loss": 0.0158, "step": 26760 }, { "grad_norm": 0.2987231910228729, "learning_rate": 8.759398902026854e-05, "loss": 0.0191, "step": 26770 }, { "grad_norm": 0.3513914942741394, "learning_rate": 8.758308563451339e-05, "loss": 0.0163, "step": 26780 }, { "grad_norm": 0.2572624385356903, "learning_rate": 8.75721781387285e-05, "loss": 0.014, "step": 26790 }, { "grad_norm": 0.25147905945777893, "learning_rate": 8.75612665341067e-05, "loss": 0.0192, "step": 26800 }, { "grad_norm": 0.24754558503627777, "learning_rate": 8.755035082184126e-05, "loss": 0.0169, "step": 26810 }, { "grad_norm": 0.2319660484790802, "learning_rate": 8.753943100312592e-05, "loss": 0.023, "step": 26820 }, { "grad_norm": 0.3420984745025635, "learning_rate": 8.752850707915484e-05, "loss": 0.0157, "step": 26830 }, { "grad_norm": 0.20867864787578583, "learning_rate": 8.751757905112264e-05, "loss": 0.0149, "step": 26840 }, { "grad_norm": 0.2706315517425537, "learning_rate": 8.75066469202244e-05, "loss": 0.0189, "step": 26850 }, { "grad_norm": 0.18844284117221832, "learning_rate": 8.749571068765567e-05, "loss": 0.0185, "step": 26860 }, { "grad_norm": 0.2364128828048706, "learning_rate": 8.748477035461238e-05, "loss": 0.0189, "step": 26870 }, { "grad_norm": 0.1910218894481659, "learning_rate": 8.747382592229095e-05, "loss": 0.0152, "step": 26880 }, { "grad_norm": 0.42948052287101746, "learning_rate": 8.746287739188828e-05, "loss": 0.0183, "step": 26890 }, { "grad_norm": 0.1957540065050125, "learning_rate": 8.745192476460165e-05, "loss": 0.0146, "step": 26900 }, { "grad_norm": 0.17157039046287537, "learning_rate": 8.744096804162882e-05, "loss": 0.0144, "step": 26910 }, { "grad_norm": 0.4120282828807831, "learning_rate": 8.743000722416804e-05, "loss": 0.0159, "step": 26920 }, { "grad_norm": 0.17348866164684296, "learning_rate": 8.741904231341793e-05, "loss": 0.0154, "step": 26930 }, { "grad_norm": 0.1860532909631729, "learning_rate": 8.740807331057762e-05, "loss": 0.02, "step": 26940 }, { "grad_norm": 0.2508503794670105, "learning_rate": 8.739710021684667e-05, "loss": 0.017, "step": 26950 }, { "grad_norm": 0.17340198159217834, "learning_rate": 8.738612303342503e-05, "loss": 0.0145, "step": 26960 }, { "grad_norm": 0.20417971909046173, "learning_rate": 8.73751417615132e-05, "loss": 0.0133, "step": 26970 }, { "grad_norm": 0.1802855134010315, "learning_rate": 8.736415640231208e-05, "loss": 0.0139, "step": 26980 }, { "grad_norm": 0.20477449893951416, "learning_rate": 8.735316695702297e-05, "loss": 0.0197, "step": 26990 }, { "grad_norm": 0.22735147178173065, "learning_rate": 8.734217342684769e-05, "loss": 0.0146, "step": 27000 }, { "grad_norm": 0.24868841469287872, "learning_rate": 8.733117581298847e-05, "loss": 0.0151, "step": 27010 }, { "grad_norm": 0.24879798293113708, "learning_rate": 8.732017411664796e-05, "loss": 0.0144, "step": 27020 }, { "grad_norm": 0.23741185665130615, "learning_rate": 8.730916833902936e-05, "loss": 0.0186, "step": 27030 }, { "grad_norm": 0.22353214025497437, "learning_rate": 8.729815848133618e-05, "loss": 0.0148, "step": 27040 }, { "grad_norm": 0.22970329225063324, "learning_rate": 8.728714454477247e-05, "loss": 0.0163, "step": 27050 }, { "grad_norm": 0.2419358640909195, "learning_rate": 8.727612653054269e-05, "loss": 0.0138, "step": 27060 }, { "grad_norm": 0.18681217730045319, "learning_rate": 8.726510443985176e-05, "loss": 0.0127, "step": 27070 }, { "grad_norm": 0.2220325618982315, "learning_rate": 8.725407827390503e-05, "loss": 0.0148, "step": 27080 }, { "grad_norm": 0.24459727108478546, "learning_rate": 8.724304803390833e-05, "loss": 0.0175, "step": 27090 }, { "grad_norm": 0.16972975432872772, "learning_rate": 8.723201372106788e-05, "loss": 0.017, "step": 27100 }, { "grad_norm": 0.3314785361289978, "learning_rate": 8.722097533659038e-05, "loss": 0.0191, "step": 27110 }, { "grad_norm": 0.22696885466575623, "learning_rate": 8.720993288168299e-05, "loss": 0.0157, "step": 27120 }, { "grad_norm": 0.24815303087234497, "learning_rate": 8.719888635755327e-05, "loss": 0.0156, "step": 27130 }, { "grad_norm": 0.31689512729644775, "learning_rate": 8.718783576540928e-05, "loss": 0.0172, "step": 27140 }, { "grad_norm": 0.27562031149864197, "learning_rate": 8.717678110645948e-05, "loss": 0.0214, "step": 27150 }, { "grad_norm": 0.2038087099790573, "learning_rate": 8.716572238191279e-05, "loss": 0.0191, "step": 27160 }, { "grad_norm": 0.25590986013412476, "learning_rate": 8.715465959297857e-05, "loss": 0.0162, "step": 27170 }, { "grad_norm": 0.28000274300575256, "learning_rate": 8.714359274086665e-05, "loss": 0.0189, "step": 27180 }, { "grad_norm": 0.2908867299556732, "learning_rate": 8.713252182678726e-05, "loss": 0.0213, "step": 27190 }, { "grad_norm": 0.2744143009185791, "learning_rate": 8.712144685195112e-05, "loss": 0.0198, "step": 27200 }, { "grad_norm": 0.2944989800453186, "learning_rate": 8.711036781756936e-05, "loss": 0.0189, "step": 27210 }, { "grad_norm": 0.28165942430496216, "learning_rate": 8.709928472485357e-05, "loss": 0.0162, "step": 27220 }, { "grad_norm": 0.19956624507904053, "learning_rate": 8.708819757501579e-05, "loss": 0.0154, "step": 27230 }, { "grad_norm": 0.170726478099823, "learning_rate": 8.707710636926846e-05, "loss": 0.0153, "step": 27240 }, { "grad_norm": 0.5496135950088501, "learning_rate": 8.706601110882455e-05, "loss": 0.0218, "step": 27250 }, { "grad_norm": 0.18973103165626526, "learning_rate": 8.705491179489738e-05, "loss": 0.0161, "step": 27260 }, { "grad_norm": 0.16648128628730774, "learning_rate": 8.704380842870077e-05, "loss": 0.0167, "step": 27270 }, { "grad_norm": 0.18762065470218658, "learning_rate": 8.703270101144895e-05, "loss": 0.0146, "step": 27280 }, { "grad_norm": 0.20639857649803162, "learning_rate": 8.702158954435664e-05, "loss": 0.0187, "step": 27290 }, { "grad_norm": 0.2655442953109741, "learning_rate": 8.701047402863896e-05, "loss": 0.0176, "step": 27300 }, { "grad_norm": 0.34732532501220703, "learning_rate": 8.699935446551148e-05, "loss": 0.0171, "step": 27310 }, { "grad_norm": 0.24545355141162872, "learning_rate": 8.698823085619022e-05, "loss": 0.0172, "step": 27320 }, { "grad_norm": 0.19243092834949493, "learning_rate": 8.697710320189166e-05, "loss": 0.0221, "step": 27330 }, { "grad_norm": 0.19757574796676636, "learning_rate": 8.696597150383268e-05, "loss": 0.015, "step": 27340 }, { "grad_norm": 0.25426802039146423, "learning_rate": 8.695483576323063e-05, "loss": 0.0159, "step": 27350 }, { "grad_norm": 0.1915292888879776, "learning_rate": 8.69436959813033e-05, "loss": 0.0178, "step": 27360 }, { "grad_norm": 0.2147328108549118, "learning_rate": 8.693255215926892e-05, "loss": 0.0156, "step": 27370 }, { "grad_norm": 0.23693567514419556, "learning_rate": 8.692140429834617e-05, "loss": 0.0214, "step": 27380 }, { "grad_norm": 0.2710728347301483, "learning_rate": 8.691025239975415e-05, "loss": 0.0165, "step": 27390 }, { "grad_norm": 0.2123018354177475, "learning_rate": 8.689909646471243e-05, "loss": 0.0189, "step": 27400 }, { "grad_norm": 0.27090489864349365, "learning_rate": 8.688793649444099e-05, "loss": 0.0156, "step": 27410 }, { "grad_norm": 0.16030971705913544, "learning_rate": 8.687677249016029e-05, "loss": 0.0184, "step": 27420 }, { "grad_norm": 0.5452331304550171, "learning_rate": 8.686560445309118e-05, "loss": 0.0151, "step": 27430 }, { "grad_norm": 0.19270393252372742, "learning_rate": 8.685443238445499e-05, "loss": 0.0253, "step": 27440 }, { "grad_norm": 0.2617785334587097, "learning_rate": 8.68432562854735e-05, "loss": 0.0204, "step": 27450 }, { "grad_norm": 0.2404518872499466, "learning_rate": 8.683207615736887e-05, "loss": 0.0191, "step": 27460 }, { "grad_norm": 0.2965605854988098, "learning_rate": 8.682089200136379e-05, "loss": 0.0151, "step": 27470 }, { "grad_norm": 0.4508267939090729, "learning_rate": 8.680970381868132e-05, "loss": 0.0156, "step": 27480 }, { "grad_norm": 0.2061055600643158, "learning_rate": 8.679851161054498e-05, "loss": 0.0199, "step": 27490 }, { "grad_norm": 0.34955719113349915, "learning_rate": 8.678731537817873e-05, "loss": 0.0194, "step": 27500 }, { "grad_norm": 0.21000608801841736, "learning_rate": 8.677611512280697e-05, "loss": 0.0168, "step": 27510 }, { "grad_norm": 0.214151069521904, "learning_rate": 8.676491084565457e-05, "loss": 0.0153, "step": 27520 }, { "grad_norm": 0.1896543651819229, "learning_rate": 8.675370254794678e-05, "loss": 0.0164, "step": 27530 }, { "grad_norm": 0.3821973502635956, "learning_rate": 8.674249023090935e-05, "loss": 0.018, "step": 27540 }, { "grad_norm": 0.2197140008211136, "learning_rate": 8.673127389576843e-05, "loss": 0.0124, "step": 27550 }, { "grad_norm": 0.17719154059886932, "learning_rate": 8.67200535437506e-05, "loss": 0.0141, "step": 27560 }, { "grad_norm": 0.23627321422100067, "learning_rate": 8.670882917608296e-05, "loss": 0.0158, "step": 27570 }, { "grad_norm": 0.2269435077905655, "learning_rate": 8.669760079399292e-05, "loss": 0.015, "step": 27580 }, { "grad_norm": 0.20929458737373352, "learning_rate": 8.668636839870845e-05, "loss": 0.0115, "step": 27590 }, { "grad_norm": 0.5751885771751404, "learning_rate": 8.667513199145789e-05, "loss": 0.0179, "step": 27600 }, { "grad_norm": 0.21920359134674072, "learning_rate": 8.666389157347002e-05, "loss": 0.0129, "step": 27610 }, { "grad_norm": 0.18097993731498718, "learning_rate": 8.66526471459741e-05, "loss": 0.0133, "step": 27620 }, { "grad_norm": 0.2223694920539856, "learning_rate": 8.66413987101998e-05, "loss": 0.0198, "step": 27630 }, { "grad_norm": 0.23119370639324188, "learning_rate": 8.663014626737723e-05, "loss": 0.0174, "step": 27640 }, { "grad_norm": 0.2819213271141052, "learning_rate": 8.661888981873691e-05, "loss": 0.0205, "step": 27650 }, { "grad_norm": 0.2137576937675476, "learning_rate": 8.660762936550988e-05, "loss": 0.017, "step": 27660 }, { "grad_norm": 0.2824553847312927, "learning_rate": 8.659636490892753e-05, "loss": 0.0171, "step": 27670 }, { "grad_norm": 0.23890897631645203, "learning_rate": 8.658509645022174e-05, "loss": 0.0199, "step": 27680 }, { "grad_norm": 0.1936948001384735, "learning_rate": 8.657382399062481e-05, "loss": 0.0142, "step": 27690 }, { "grad_norm": 0.24868527054786682, "learning_rate": 8.656254753136946e-05, "loss": 0.017, "step": 27700 }, { "grad_norm": 0.9867449998855591, "learning_rate": 8.655126707368891e-05, "loss": 0.0147, "step": 27710 }, { "grad_norm": 0.17128247022628784, "learning_rate": 8.653998261881672e-05, "loss": 0.0137, "step": 27720 }, { "grad_norm": 0.289130300283432, "learning_rate": 8.652869416798699e-05, "loss": 0.0233, "step": 27730 }, { "grad_norm": 0.2943936586380005, "learning_rate": 8.651740172243417e-05, "loss": 0.0204, "step": 27740 }, { "grad_norm": 0.21591918170452118, "learning_rate": 8.65061052833932e-05, "loss": 0.0275, "step": 27750 }, { "grad_norm": 0.2674432098865509, "learning_rate": 8.649480485209945e-05, "loss": 0.0187, "step": 27760 }, { "grad_norm": 0.2526676654815674, "learning_rate": 8.64835004297887e-05, "loss": 0.0195, "step": 27770 }, { "grad_norm": 0.3306019902229309, "learning_rate": 8.64721920176972e-05, "loss": 0.0191, "step": 27780 }, { "grad_norm": 0.20298177003860474, "learning_rate": 8.646087961706164e-05, "loss": 0.0169, "step": 27790 }, { "grad_norm": 0.32579702138900757, "learning_rate": 8.644956322911908e-05, "loss": 0.0146, "step": 27800 }, { "grad_norm": 0.20302732288837433, "learning_rate": 8.643824285510709e-05, "loss": 0.0165, "step": 27810 }, { "grad_norm": 0.28462010622024536, "learning_rate": 8.642691849626364e-05, "loss": 0.0142, "step": 27820 }, { "grad_norm": 0.2731457054615021, "learning_rate": 8.641559015382717e-05, "loss": 0.0193, "step": 27830 }, { "grad_norm": 0.2590716481208801, "learning_rate": 8.640425782903649e-05, "loss": 0.02, "step": 27840 }, { "grad_norm": 1.0802949666976929, "learning_rate": 8.639292152313091e-05, "loss": 0.0212, "step": 27850 }, { "grad_norm": 0.2767835259437561, "learning_rate": 8.638158123735015e-05, "loss": 0.0199, "step": 27860 }, { "grad_norm": 0.26460975408554077, "learning_rate": 8.637023697293436e-05, "loss": 0.0191, "step": 27870 }, { "grad_norm": 0.2279714047908783, "learning_rate": 8.635888873112414e-05, "loss": 0.0201, "step": 27880 }, { "grad_norm": 0.21042633056640625, "learning_rate": 8.634753651316052e-05, "loss": 0.0212, "step": 27890 }, { "grad_norm": 0.16314709186553955, "learning_rate": 8.633618032028496e-05, "loss": 0.0151, "step": 27900 }, { "grad_norm": 0.29389315843582153, "learning_rate": 8.632482015373934e-05, "loss": 0.0192, "step": 27910 }, { "grad_norm": 0.27636945247650146, "learning_rate": 8.6313456014766e-05, "loss": 0.0173, "step": 27920 }, { "grad_norm": 0.27490881085395813, "learning_rate": 8.630208790460771e-05, "loss": 0.0249, "step": 27930 }, { "grad_norm": 0.23662303388118744, "learning_rate": 8.629071582450768e-05, "loss": 0.0249, "step": 27940 }, { "grad_norm": 0.4706205725669861, "learning_rate": 8.62793397757095e-05, "loss": 0.0238, "step": 27950 }, { "grad_norm": 1.0445564985275269, "learning_rate": 8.626795975945729e-05, "loss": 0.0216, "step": 27960 }, { "grad_norm": 0.24586069583892822, "learning_rate": 8.625657577699551e-05, "loss": 0.0327, "step": 27970 }, { "grad_norm": 0.2973845899105072, "learning_rate": 8.624518782956914e-05, "loss": 0.0215, "step": 27980 }, { "grad_norm": 0.20961931347846985, "learning_rate": 8.62337959184235e-05, "loss": 0.0185, "step": 27990 }, { "grad_norm": 0.2792564332485199, "learning_rate": 8.622240004480441e-05, "loss": 0.0214, "step": 28000 }, { "grad_norm": 0.19305087625980377, "learning_rate": 8.621100020995814e-05, "loss": 0.0171, "step": 28010 }, { "grad_norm": 0.21323555707931519, "learning_rate": 8.619959641513132e-05, "loss": 0.0224, "step": 28020 }, { "grad_norm": 0.20345455408096313, "learning_rate": 8.618818866157105e-05, "loss": 0.0175, "step": 28030 }, { "grad_norm": 0.3533637523651123, "learning_rate": 8.617677695052487e-05, "loss": 0.0239, "step": 28040 }, { "grad_norm": 0.2377711981534958, "learning_rate": 8.616536128324078e-05, "loss": 0.0161, "step": 28050 }, { "grad_norm": 0.1715957075357437, "learning_rate": 8.615394166096712e-05, "loss": 0.0159, "step": 28060 }, { "grad_norm": 0.16459110379219055, "learning_rate": 8.614251808495279e-05, "loss": 0.0194, "step": 28070 }, { "grad_norm": 0.2418707311153412, "learning_rate": 8.6131090556447e-05, "loss": 0.0191, "step": 28080 }, { "grad_norm": 0.19714389741420746, "learning_rate": 8.611965907669947e-05, "loss": 0.017, "step": 28090 }, { "grad_norm": 0.22575686872005463, "learning_rate": 8.610822364696034e-05, "loss": 0.0166, "step": 28100 }, { "grad_norm": 0.14541777968406677, "learning_rate": 8.609678426848015e-05, "loss": 0.0135, "step": 28110 }, { "grad_norm": 0.1889914870262146, "learning_rate": 8.60853409425099e-05, "loss": 0.0145, "step": 28120 }, { "grad_norm": 0.2317986637353897, "learning_rate": 8.607389367030104e-05, "loss": 0.0178, "step": 28130 }, { "grad_norm": 0.23450319468975067, "learning_rate": 8.606244245310538e-05, "loss": 0.0152, "step": 28140 }, { "grad_norm": 0.23311364650726318, "learning_rate": 8.605098729217525e-05, "loss": 0.0153, "step": 28150 }, { "grad_norm": 0.23875688016414642, "learning_rate": 8.603952818876335e-05, "loss": 0.0166, "step": 28160 }, { "grad_norm": 1.520690679550171, "learning_rate": 8.602806514412281e-05, "loss": 0.0197, "step": 28170 }, { "grad_norm": 0.30878981947898865, "learning_rate": 8.601659815950726e-05, "loss": 0.0197, "step": 28180 }, { "grad_norm": 0.147846058011055, "learning_rate": 8.600512723617067e-05, "loss": 0.0154, "step": 28190 }, { "grad_norm": 0.2038206309080124, "learning_rate": 8.59936523753675e-05, "loss": 0.0151, "step": 28200 }, { "grad_norm": 0.19547154009342194, "learning_rate": 8.598217357835264e-05, "loss": 0.0155, "step": 28210 }, { "grad_norm": 0.3645862638950348, "learning_rate": 8.597069084638135e-05, "loss": 0.0165, "step": 28220 }, { "grad_norm": 0.24217122793197632, "learning_rate": 8.595920418070939e-05, "loss": 0.0167, "step": 28230 }, { "grad_norm": 0.33112454414367676, "learning_rate": 8.594771358259295e-05, "loss": 0.017, "step": 28240 }, { "grad_norm": 0.20558853447437286, "learning_rate": 8.593621905328858e-05, "loss": 0.0177, "step": 28250 }, { "grad_norm": 0.33176714181900024, "learning_rate": 8.592472059405333e-05, "loss": 0.0167, "step": 28260 }, { "grad_norm": 0.20160402357578278, "learning_rate": 8.591321820614464e-05, "loss": 0.013, "step": 28270 }, { "grad_norm": 0.22486324608325958, "learning_rate": 8.590171189082041e-05, "loss": 0.0134, "step": 28280 }, { "grad_norm": 0.19557125866413116, "learning_rate": 8.589020164933894e-05, "loss": 0.0121, "step": 28290 }, { "grad_norm": 0.17389525473117828, "learning_rate": 8.587868748295898e-05, "loss": 0.012, "step": 28300 }, { "grad_norm": 0.20234176516532898, "learning_rate": 8.586716939293971e-05, "loss": 0.0155, "step": 28310 }, { "grad_norm": 0.22807739675045013, "learning_rate": 8.58556473805407e-05, "loss": 0.0176, "step": 28320 }, { "grad_norm": 0.19655290246009827, "learning_rate": 8.584412144702202e-05, "loss": 0.0185, "step": 28330 }, { "grad_norm": 0.2722073793411255, "learning_rate": 8.58325915936441e-05, "loss": 0.0154, "step": 28340 }, { "grad_norm": 0.18727880716323853, "learning_rate": 8.582105782166783e-05, "loss": 0.0153, "step": 28350 }, { "grad_norm": 0.25851869583129883, "learning_rate": 8.580952013235455e-05, "loss": 0.0177, "step": 28360 }, { "grad_norm": 0.17474231123924255, "learning_rate": 8.579797852696596e-05, "loss": 0.0131, "step": 28370 }, { "grad_norm": 0.30704861879348755, "learning_rate": 8.578643300676428e-05, "loss": 0.0142, "step": 28380 }, { "grad_norm": 0.26514655351638794, "learning_rate": 8.577488357301209e-05, "loss": 0.0136, "step": 28390 }, { "grad_norm": 0.19340692460536957, "learning_rate": 8.576333022697242e-05, "loss": 0.0357, "step": 28400 }, { "grad_norm": 0.23866458237171173, "learning_rate": 8.575177296990873e-05, "loss": 0.0208, "step": 28410 }, { "grad_norm": 0.19754193723201752, "learning_rate": 8.574021180308489e-05, "loss": 0.0176, "step": 28420 }, { "grad_norm": 0.732576310634613, "learning_rate": 8.572864672776523e-05, "loss": 0.0194, "step": 28430 }, { "grad_norm": 0.17303860187530518, "learning_rate": 8.571707774521447e-05, "loss": 0.0203, "step": 28440 }, { "grad_norm": 2.5521905422210693, "learning_rate": 8.57055048566978e-05, "loss": 0.0167, "step": 28450 }, { "grad_norm": 0.25565260648727417, "learning_rate": 8.569392806348078e-05, "loss": 0.0193, "step": 28460 }, { "grad_norm": 0.21182477474212646, "learning_rate": 8.568234736682947e-05, "loss": 0.0159, "step": 28470 }, { "grad_norm": 0.21262168884277344, "learning_rate": 8.567076276801029e-05, "loss": 0.0165, "step": 28480 }, { "grad_norm": 0.3260476887226105, "learning_rate": 8.565917426829013e-05, "loss": 0.016, "step": 28490 }, { "grad_norm": 0.3179270327091217, "learning_rate": 8.564758186893628e-05, "loss": 0.017, "step": 28500 }, { "grad_norm": 0.43567314743995667, "learning_rate": 8.563598557121649e-05, "loss": 0.018, "step": 28510 }, { "grad_norm": 0.20220236480236053, "learning_rate": 8.562438537639888e-05, "loss": 0.0143, "step": 28520 }, { "grad_norm": 0.25577226281166077, "learning_rate": 8.561278128575206e-05, "loss": 0.0155, "step": 28530 }, { "grad_norm": 0.2620818614959717, "learning_rate": 8.5601173300545e-05, "loss": 0.0159, "step": 28540 }, { "grad_norm": 0.25546994805336, "learning_rate": 8.558956142204717e-05, "loss": 0.0129, "step": 28550 }, { "grad_norm": 0.2498588263988495, "learning_rate": 8.55779456515284e-05, "loss": 0.0144, "step": 28560 }, { "grad_norm": 0.1527666300535202, "learning_rate": 8.556632599025898e-05, "loss": 0.0125, "step": 28570 }, { "grad_norm": 0.1957472413778305, "learning_rate": 8.555470243950964e-05, "loss": 0.0164, "step": 28580 }, { "grad_norm": 0.17098833620548248, "learning_rate": 8.554307500055148e-05, "loss": 0.0143, "step": 28590 }, { "grad_norm": 0.33966493606567383, "learning_rate": 8.553144367465609e-05, "loss": 0.0138, "step": 28600 }, { "grad_norm": 0.22649921476840973, "learning_rate": 8.551980846309544e-05, "loss": 0.0139, "step": 28610 }, { "grad_norm": 0.21375255286693573, "learning_rate": 8.550816936714193e-05, "loss": 0.0211, "step": 28620 }, { "grad_norm": 0.1970326453447342, "learning_rate": 8.549652638806841e-05, "loss": 0.0157, "step": 28630 }, { "grad_norm": 0.20748870074748993, "learning_rate": 8.548487952714812e-05, "loss": 0.0136, "step": 28640 }, { "grad_norm": 0.2071775197982788, "learning_rate": 8.547322878565478e-05, "loss": 0.0152, "step": 28650 }, { "grad_norm": 0.16380493342876434, "learning_rate": 8.546157416486245e-05, "loss": 0.0146, "step": 28660 }, { "grad_norm": 0.21250544488430023, "learning_rate": 8.54499156660457e-05, "loss": 0.0167, "step": 28670 }, { "grad_norm": 0.4599718153476715, "learning_rate": 8.543825329047947e-05, "loss": 0.0146, "step": 28680 }, { "grad_norm": 0.21118006110191345, "learning_rate": 8.542658703943913e-05, "loss": 0.0166, "step": 28690 }, { "grad_norm": 0.23553454875946045, "learning_rate": 8.541491691420051e-05, "loss": 0.0153, "step": 28700 }, { "grad_norm": 0.21724621951580048, "learning_rate": 8.54032429160398e-05, "loss": 0.0144, "step": 28710 }, { "grad_norm": 0.2872966229915619, "learning_rate": 8.539156504623369e-05, "loss": 0.0151, "step": 28720 }, { "grad_norm": 0.268050879240036, "learning_rate": 8.537988330605923e-05, "loss": 0.0161, "step": 28730 }, { "grad_norm": 0.23306116461753845, "learning_rate": 8.536819769679393e-05, "loss": 0.0176, "step": 28740 }, { "grad_norm": 0.18655316531658173, "learning_rate": 8.53565082197157e-05, "loss": 0.0155, "step": 28750 }, { "grad_norm": 0.22965700924396515, "learning_rate": 8.534481487610289e-05, "loss": 0.0148, "step": 28760 }, { "grad_norm": 1.008207082748413, "learning_rate": 8.533311766723428e-05, "loss": 0.0145, "step": 28770 }, { "grad_norm": 0.28779157996177673, "learning_rate": 8.532141659438901e-05, "loss": 0.0167, "step": 28780 }, { "grad_norm": 0.31396350264549255, "learning_rate": 8.530971165884675e-05, "loss": 0.0176, "step": 28790 }, { "grad_norm": 0.30842074751853943, "learning_rate": 8.529800286188752e-05, "loss": 0.023, "step": 28800 }, { "grad_norm": 0.24207760393619537, "learning_rate": 8.528629020479175e-05, "loss": 0.0165, "step": 28810 }, { "grad_norm": 0.22265887260437012, "learning_rate": 8.527457368884033e-05, "loss": 0.0156, "step": 28820 }, { "grad_norm": 0.223166361451149, "learning_rate": 8.526285331531458e-05, "loss": 0.0151, "step": 28830 }, { "grad_norm": 0.36632010340690613, "learning_rate": 8.525112908549621e-05, "loss": 0.0171, "step": 28840 }, { "grad_norm": 0.28074073791503906, "learning_rate": 8.523940100066735e-05, "loss": 0.0184, "step": 28850 }, { "grad_norm": 0.21858909726142883, "learning_rate": 8.52276690621106e-05, "loss": 0.0196, "step": 28860 }, { "grad_norm": 0.22768527269363403, "learning_rate": 8.521593327110889e-05, "loss": 0.0178, "step": 28870 }, { "grad_norm": 0.24759435653686523, "learning_rate": 8.520419362894569e-05, "loss": 0.0168, "step": 28880 }, { "grad_norm": 0.34604620933532715, "learning_rate": 8.51924501369048e-05, "loss": 0.0182, "step": 28890 }, { "grad_norm": 0.25291699171066284, "learning_rate": 8.518070279627047e-05, "loss": 0.017, "step": 28900 }, { "grad_norm": 0.29387104511260986, "learning_rate": 8.516895160832737e-05, "loss": 0.0158, "step": 28910 }, { "grad_norm": 0.3454042077064514, "learning_rate": 8.515719657436061e-05, "loss": 0.0165, "step": 28920 }, { "grad_norm": 0.21971404552459717, "learning_rate": 8.514543769565568e-05, "loss": 0.0182, "step": 28930 }, { "grad_norm": 0.21687771379947662, "learning_rate": 8.513367497349853e-05, "loss": 0.0187, "step": 28940 }, { "grad_norm": 0.2622174024581909, "learning_rate": 8.51219084091755e-05, "loss": 0.0234, "step": 28950 }, { "grad_norm": 0.3467388153076172, "learning_rate": 8.511013800397338e-05, "loss": 0.017, "step": 28960 }, { "grad_norm": 0.25773730874061584, "learning_rate": 8.509836375917937e-05, "loss": 0.0184, "step": 28970 }, { "grad_norm": 0.506514847278595, "learning_rate": 8.508658567608104e-05, "loss": 0.017, "step": 28980 }, { "grad_norm": 0.2833511233329773, "learning_rate": 8.507480375596647e-05, "loss": 0.0187, "step": 28990 }, { "grad_norm": 0.32569968700408936, "learning_rate": 8.506301800012408e-05, "loss": 0.0262, "step": 29000 }, { "grad_norm": 0.24448135495185852, "learning_rate": 8.505122840984278e-05, "loss": 0.0213, "step": 29010 }, { "grad_norm": 0.29854583740234375, "learning_rate": 8.503943498641182e-05, "loss": 0.0208, "step": 29020 }, { "grad_norm": 0.3103184103965759, "learning_rate": 8.502763773112095e-05, "loss": 0.015, "step": 29030 }, { "grad_norm": 0.30390501022338867, "learning_rate": 8.501583664526026e-05, "loss": 0.0212, "step": 29040 }, { "grad_norm": 0.3059386909008026, "learning_rate": 8.500403173012032e-05, "loss": 0.0156, "step": 29050 }, { "grad_norm": 0.23463577032089233, "learning_rate": 8.499222298699211e-05, "loss": 0.0191, "step": 29060 }, { "grad_norm": 0.19026358425617218, "learning_rate": 8.498041041716701e-05, "loss": 0.019, "step": 29070 }, { "grad_norm": 0.3244224190711975, "learning_rate": 8.496859402193681e-05, "loss": 0.0219, "step": 29080 }, { "grad_norm": 0.31104961037635803, "learning_rate": 8.495677380259374e-05, "loss": 0.0173, "step": 29090 }, { "grad_norm": 0.2646641135215759, "learning_rate": 8.494494976043045e-05, "loss": 0.0185, "step": 29100 }, { "grad_norm": 0.16951651871204376, "learning_rate": 8.493312189673998e-05, "loss": 0.0198, "step": 29110 }, { "grad_norm": 0.23680131137371063, "learning_rate": 8.492129021281584e-05, "loss": 0.0149, "step": 29120 }, { "grad_norm": 0.2287280261516571, "learning_rate": 8.490945470995188e-05, "loss": 0.0153, "step": 29130 }, { "grad_norm": 0.2629561722278595, "learning_rate": 8.489761538944247e-05, "loss": 0.0154, "step": 29140 }, { "grad_norm": 0.39707180857658386, "learning_rate": 8.48857722525823e-05, "loss": 0.0157, "step": 29150 }, { "grad_norm": 0.6233974099159241, "learning_rate": 8.487392530066652e-05, "loss": 0.0176, "step": 29160 }, { "grad_norm": 0.14747372269630432, "learning_rate": 8.486207453499069e-05, "loss": 0.0148, "step": 29170 }, { "grad_norm": 0.2522071301937103, "learning_rate": 8.485021995685082e-05, "loss": 0.0157, "step": 29180 }, { "grad_norm": 0.1653725951910019, "learning_rate": 8.483836156754328e-05, "loss": 0.0161, "step": 29190 }, { "grad_norm": 0.19864250719547272, "learning_rate": 8.482649936836491e-05, "loss": 0.015, "step": 29200 }, { "grad_norm": 0.22901605069637299, "learning_rate": 8.481463336061293e-05, "loss": 0.0174, "step": 29210 }, { "grad_norm": 0.291579008102417, "learning_rate": 8.480276354558496e-05, "loss": 0.0193, "step": 29220 }, { "grad_norm": 0.2231159210205078, "learning_rate": 8.479088992457913e-05, "loss": 0.0163, "step": 29230 }, { "grad_norm": 0.18028676509857178, "learning_rate": 8.477901249889387e-05, "loss": 0.0152, "step": 29240 }, { "grad_norm": 0.19235104322433472, "learning_rate": 8.47671312698281e-05, "loss": 0.0148, "step": 29250 }, { "grad_norm": 0.1984068900346756, "learning_rate": 8.475524623868112e-05, "loss": 0.0179, "step": 29260 }, { "grad_norm": 0.20253169536590576, "learning_rate": 8.474335740675266e-05, "loss": 0.0146, "step": 29270 }, { "grad_norm": 0.14634306728839874, "learning_rate": 8.473146477534289e-05, "loss": 0.012, "step": 29280 }, { "grad_norm": 0.1992741823196411, "learning_rate": 8.471956834575232e-05, "loss": 0.0144, "step": 29290 }, { "grad_norm": 0.24093016982078552, "learning_rate": 8.470766811928197e-05, "loss": 0.0123, "step": 29300 }, { "grad_norm": 0.19171065092086792, "learning_rate": 8.469576409723323e-05, "loss": 0.0131, "step": 29310 }, { "grad_norm": 0.34409186244010925, "learning_rate": 8.468385628090788e-05, "loss": 0.0119, "step": 29320 }, { "grad_norm": 0.20098935067653656, "learning_rate": 8.467194467160815e-05, "loss": 0.0168, "step": 29330 }, { "grad_norm": 0.23058725893497467, "learning_rate": 8.466002927063667e-05, "loss": 0.0135, "step": 29340 }, { "grad_norm": 0.2240673452615738, "learning_rate": 8.464811007929651e-05, "loss": 0.0153, "step": 29350 }, { "grad_norm": 0.1966143399477005, "learning_rate": 8.463618709889114e-05, "loss": 0.0127, "step": 29360 }, { "grad_norm": 0.21869619190692902, "learning_rate": 8.462426033072442e-05, "loss": 0.0197, "step": 29370 }, { "grad_norm": 0.21828365325927734, "learning_rate": 8.461232977610061e-05, "loss": 0.0156, "step": 29380 }, { "grad_norm": 0.20918001234531403, "learning_rate": 8.46003954363245e-05, "loss": 0.015, "step": 29390 }, { "grad_norm": 0.22401376068592072, "learning_rate": 8.458845731270115e-05, "loss": 0.0134, "step": 29400 }, { "grad_norm": 6.211094856262207, "learning_rate": 8.45765154065361e-05, "loss": 0.0209, "step": 29410 }, { "grad_norm": 2.179225444793701, "learning_rate": 8.456456971913532e-05, "loss": 0.019, "step": 29420 }, { "grad_norm": 0.16014578938484192, "learning_rate": 8.455262025180517e-05, "loss": 0.0176, "step": 29430 }, { "grad_norm": 0.2254047840833664, "learning_rate": 8.454066700585242e-05, "loss": 0.0132, "step": 29440 }, { "grad_norm": 0.31218641996383667, "learning_rate": 8.452870998258423e-05, "loss": 0.0165, "step": 29450 }, { "grad_norm": 0.22419056296348572, "learning_rate": 8.451674918330825e-05, "loss": 0.0157, "step": 29460 }, { "grad_norm": 0.1977427452802658, "learning_rate": 8.450478460933246e-05, "loss": 0.0173, "step": 29470 }, { "grad_norm": 0.21389950811862946, "learning_rate": 8.449281626196532e-05, "loss": 0.0126, "step": 29480 }, { "grad_norm": 0.44721049070358276, "learning_rate": 8.448084414251564e-05, "loss": 0.0199, "step": 29490 }, { "grad_norm": 0.18673908710479736, "learning_rate": 8.446886825229271e-05, "loss": 0.0152, "step": 29500 }, { "grad_norm": 0.27183783054351807, "learning_rate": 8.445688859260615e-05, "loss": 0.0203, "step": 29510 }, { "grad_norm": 0.2270420789718628, "learning_rate": 8.444490516476606e-05, "loss": 0.0178, "step": 29520 }, { "grad_norm": 0.20182427763938904, "learning_rate": 8.443291797008293e-05, "loss": 0.0136, "step": 29530 }, { "grad_norm": 0.2498725950717926, "learning_rate": 8.442092700986765e-05, "loss": 0.0347, "step": 29540 }, { "grad_norm": 0.5870484709739685, "learning_rate": 8.440893228543156e-05, "loss": 0.017, "step": 29550 }, { "grad_norm": 0.2418125420808792, "learning_rate": 8.439693379808638e-05, "loss": 0.0172, "step": 29560 }, { "grad_norm": 0.9494852423667908, "learning_rate": 8.43849315491442e-05, "loss": 0.0138, "step": 29570 }, { "grad_norm": 0.18260110914707184, "learning_rate": 8.437292553991763e-05, "loss": 0.0152, "step": 29580 }, { "grad_norm": 0.19409584999084473, "learning_rate": 8.436091577171959e-05, "loss": 0.0169, "step": 29590 }, { "grad_norm": 0.38464999198913574, "learning_rate": 8.434890224586347e-05, "loss": 0.021, "step": 29600 }, { "grad_norm": 0.8843721747398376, "learning_rate": 8.433688496366303e-05, "loss": 0.0192, "step": 29610 }, { "grad_norm": 0.20495019853115082, "learning_rate": 8.432486392643248e-05, "loss": 0.0122, "step": 29620 }, { "grad_norm": 0.28233465552330017, "learning_rate": 8.431283913548643e-05, "loss": 0.0186, "step": 29630 }, { "grad_norm": 0.19670222699642181, "learning_rate": 8.430081059213985e-05, "loss": 0.0121, "step": 29640 }, { "grad_norm": 0.25074419379234314, "learning_rate": 8.428877829770823e-05, "loss": 0.0151, "step": 29650 }, { "grad_norm": 0.21488404273986816, "learning_rate": 8.427674225350735e-05, "loss": 0.0157, "step": 29660 }, { "grad_norm": 0.22501088678836823, "learning_rate": 8.426470246085347e-05, "loss": 0.0141, "step": 29670 }, { "grad_norm": 0.1662495732307434, "learning_rate": 8.425265892106324e-05, "loss": 0.0141, "step": 29680 }, { "grad_norm": 0.19494573771953583, "learning_rate": 8.424061163545374e-05, "loss": 0.0134, "step": 29690 }, { "grad_norm": 0.19449029862880707, "learning_rate": 8.422856060534243e-05, "loss": 0.0154, "step": 29700 }, { "grad_norm": 0.18995535373687744, "learning_rate": 8.421650583204718e-05, "loss": 0.0146, "step": 29710 }, { "grad_norm": 0.2092640995979309, "learning_rate": 8.420444731688633e-05, "loss": 0.0146, "step": 29720 }, { "grad_norm": 0.19483324885368347, "learning_rate": 8.419238506117852e-05, "loss": 0.0168, "step": 29730 }, { "grad_norm": 0.24443331360816956, "learning_rate": 8.418031906624289e-05, "loss": 0.017, "step": 29740 }, { "grad_norm": 0.15631121397018433, "learning_rate": 8.416824933339898e-05, "loss": 0.0159, "step": 29750 }, { "grad_norm": 0.4151070713996887, "learning_rate": 8.415617586396667e-05, "loss": 0.0217, "step": 29760 }, { "grad_norm": 0.17439235746860504, "learning_rate": 8.414409865926632e-05, "loss": 0.0142, "step": 29770 }, { "grad_norm": 0.20818893611431122, "learning_rate": 8.413201772061867e-05, "loss": 0.0148, "step": 29780 }, { "grad_norm": 0.20079176127910614, "learning_rate": 8.411993304934488e-05, "loss": 0.0139, "step": 29790 }, { "grad_norm": 0.6818634867668152, "learning_rate": 8.410784464676654e-05, "loss": 0.0176, "step": 29800 }, { "grad_norm": 0.21495500206947327, "learning_rate": 8.409575251420556e-05, "loss": 0.0175, "step": 29810 }, { "grad_norm": 0.21028795838356018, "learning_rate": 8.408365665298435e-05, "loss": 0.0162, "step": 29820 }, { "grad_norm": 0.2335408329963684, "learning_rate": 8.40715570644257e-05, "loss": 0.0134, "step": 29830 }, { "grad_norm": 0.2520395815372467, "learning_rate": 8.40594537498528e-05, "loss": 0.015, "step": 29840 }, { "grad_norm": 0.2745942771434784, "learning_rate": 8.404734671058924e-05, "loss": 0.0178, "step": 29850 }, { "grad_norm": 0.2680374085903168, "learning_rate": 8.403523594795902e-05, "loss": 0.0202, "step": 29860 }, { "grad_norm": 0.3923804461956024, "learning_rate": 8.402312146328659e-05, "loss": 0.0262, "step": 29870 }, { "grad_norm": 0.40010496973991394, "learning_rate": 8.401100325789675e-05, "loss": 0.017, "step": 29880 }, { "grad_norm": 0.1738666146993637, "learning_rate": 8.399888133311472e-05, "loss": 0.0135, "step": 29890 }, { "grad_norm": 0.19779323041439056, "learning_rate": 8.398675569026613e-05, "loss": 0.0213, "step": 29900 }, { "grad_norm": 0.20626622438430786, "learning_rate": 8.397462633067705e-05, "loss": 0.0157, "step": 29910 }, { "grad_norm": 0.3031200170516968, "learning_rate": 8.396249325567392e-05, "loss": 0.0131, "step": 29920 }, { "grad_norm": 0.1924838274717331, "learning_rate": 8.395035646658357e-05, "loss": 0.0178, "step": 29930 }, { "grad_norm": 0.22524407505989075, "learning_rate": 8.39382159647333e-05, "loss": 0.0189, "step": 29940 }, { "grad_norm": 0.21840332448482513, "learning_rate": 8.392607175145075e-05, "loss": 0.0162, "step": 29950 }, { "grad_norm": 0.1722864955663681, "learning_rate": 8.3913923828064e-05, "loss": 0.0246, "step": 29960 }, { "grad_norm": 0.2729056775569916, "learning_rate": 8.390177219590152e-05, "loss": 0.0175, "step": 29970 }, { "grad_norm": 0.200581356883049, "learning_rate": 8.388961685629222e-05, "loss": 0.0208, "step": 29980 }, { "grad_norm": 0.2928372919559479, "learning_rate": 8.387745781056536e-05, "loss": 0.0155, "step": 29990 }, { "grad_norm": 0.2989796996116638, "learning_rate": 8.386529506005065e-05, "loss": 0.0174, "step": 30000 }, { "grad_norm": 0.20322634279727936, "learning_rate": 8.38531286060782e-05, "loss": 0.0124, "step": 30010 }, { "grad_norm": 0.1764676719903946, "learning_rate": 8.384095844997849e-05, "loss": 0.0214, "step": 30020 }, { "grad_norm": 0.2015557587146759, "learning_rate": 8.382878459308245e-05, "loss": 0.015, "step": 30030 }, { "grad_norm": 0.2454947680234909, "learning_rate": 8.381660703672138e-05, "loss": 0.0155, "step": 30040 }, { "grad_norm": 0.20800457894802094, "learning_rate": 8.380442578222702e-05, "loss": 0.0136, "step": 30050 }, { "grad_norm": 0.21972806751728058, "learning_rate": 8.379224083093146e-05, "loss": 0.0124, "step": 30060 }, { "grad_norm": 0.43533626198768616, "learning_rate": 8.378005218416727e-05, "loss": 0.0117, "step": 30070 }, { "grad_norm": 0.23992159962654114, "learning_rate": 8.376785984326735e-05, "loss": 0.0105, "step": 30080 }, { "grad_norm": 0.6927849054336548, "learning_rate": 8.375566380956506e-05, "loss": 0.019, "step": 30090 }, { "grad_norm": 0.2558346390724182, "learning_rate": 8.374346408439411e-05, "loss": 0.0147, "step": 30100 }, { "grad_norm": 0.24107904732227325, "learning_rate": 8.373126066908868e-05, "loss": 0.0147, "step": 30110 }, { "grad_norm": 0.17815470695495605, "learning_rate": 8.371905356498326e-05, "loss": 0.0138, "step": 30120 }, { "grad_norm": 0.3538508415222168, "learning_rate": 8.370684277341288e-05, "loss": 0.0189, "step": 30130 }, { "grad_norm": 0.20786935091018677, "learning_rate": 8.369462829571282e-05, "loss": 0.0152, "step": 30140 }, { "grad_norm": 0.17652468383312225, "learning_rate": 8.36824101332189e-05, "loss": 0.0184, "step": 30150 }, { "grad_norm": 0.21800057590007782, "learning_rate": 8.367018828726721e-05, "loss": 0.0132, "step": 30160 }, { "grad_norm": 0.1910402923822403, "learning_rate": 8.365796275919438e-05, "loss": 0.0161, "step": 30170 }, { "grad_norm": 0.20340655744075775, "learning_rate": 8.364573355033734e-05, "loss": 0.0176, "step": 30180 }, { "grad_norm": 0.16833651065826416, "learning_rate": 8.363350066203346e-05, "loss": 0.0139, "step": 30190 }, { "grad_norm": 0.2386513501405716, "learning_rate": 8.362126409562053e-05, "loss": 0.0146, "step": 30200 }, { "grad_norm": 0.20931284129619598, "learning_rate": 8.360902385243667e-05, "loss": 0.0118, "step": 30210 }, { "grad_norm": 0.2541525661945343, "learning_rate": 8.359677993382052e-05, "loss": 0.016, "step": 30220 }, { "grad_norm": 0.2527700960636139, "learning_rate": 8.358453234111103e-05, "loss": 0.0151, "step": 30230 }, { "grad_norm": 0.2504188120365143, "learning_rate": 8.357228107564756e-05, "loss": 0.012, "step": 30240 }, { "grad_norm": 1.0162177085876465, "learning_rate": 8.356002613876993e-05, "loss": 0.009, "step": 30250 }, { "grad_norm": 0.169404998421669, "learning_rate": 8.35477675318183e-05, "loss": 0.0143, "step": 30260 }, { "grad_norm": 0.14518386125564575, "learning_rate": 8.353550525613323e-05, "loss": 0.0131, "step": 30270 }, { "grad_norm": 0.18430931866168976, "learning_rate": 8.352323931305572e-05, "loss": 0.012, "step": 30280 }, { "grad_norm": 0.2847309112548828, "learning_rate": 8.351096970392717e-05, "loss": 0.0124, "step": 30290 }, { "grad_norm": 0.6718608736991882, "learning_rate": 8.349869643008937e-05, "loss": 0.0168, "step": 30300 }, { "grad_norm": 0.5516356229782104, "learning_rate": 8.348641949288449e-05, "loss": 0.0259, "step": 30310 }, { "grad_norm": 0.18305489420890808, "learning_rate": 8.34741388936551e-05, "loss": 0.0145, "step": 30320 }, { "grad_norm": 0.21509553492069244, "learning_rate": 8.346185463374423e-05, "loss": 0.016, "step": 30330 }, { "grad_norm": 0.1791379600763321, "learning_rate": 8.344956671449524e-05, "loss": 0.0161, "step": 30340 }, { "grad_norm": 0.14676611125469208, "learning_rate": 8.343727513725192e-05, "loss": 0.0146, "step": 30350 }, { "grad_norm": 0.16838271915912628, "learning_rate": 8.342497990335847e-05, "loss": 0.0149, "step": 30360 }, { "grad_norm": 0.1691267490386963, "learning_rate": 8.341268101415946e-05, "loss": 0.0207, "step": 30370 }, { "grad_norm": 0.275440514087677, "learning_rate": 8.34003784709999e-05, "loss": 0.0236, "step": 30380 }, { "grad_norm": 0.2442091405391693, "learning_rate": 8.338807227522516e-05, "loss": 0.0178, "step": 30390 }, { "grad_norm": 0.15909931063652039, "learning_rate": 8.337576242818103e-05, "loss": 0.0151, "step": 30400 }, { "grad_norm": 0.2795294225215912, "learning_rate": 8.336344893121372e-05, "loss": 0.0179, "step": 30410 }, { "grad_norm": 0.253557026386261, "learning_rate": 8.335113178566977e-05, "loss": 0.0197, "step": 30420 }, { "grad_norm": 0.19368989765644073, "learning_rate": 8.33388109928962e-05, "loss": 0.0151, "step": 30430 }, { "grad_norm": 0.18421204388141632, "learning_rate": 8.33264865542404e-05, "loss": 0.0157, "step": 30440 }, { "grad_norm": 0.4878624975681305, "learning_rate": 8.331415847105013e-05, "loss": 0.0204, "step": 30450 }, { "grad_norm": 0.17415602505207062, "learning_rate": 8.330182674467357e-05, "loss": 0.0212, "step": 30460 }, { "grad_norm": 0.22052700817584991, "learning_rate": 8.32894913764593e-05, "loss": 0.0158, "step": 30470 }, { "grad_norm": 0.2582269310951233, "learning_rate": 8.327715236775633e-05, "loss": 0.0211, "step": 30480 }, { "grad_norm": 0.42515891790390015, "learning_rate": 8.326480971991398e-05, "loss": 0.0219, "step": 30490 }, { "grad_norm": 0.2495623379945755, "learning_rate": 8.325246343428206e-05, "loss": 0.015, "step": 30500 }, { "grad_norm": 0.23171241581439972, "learning_rate": 8.324011351221072e-05, "loss": 0.0202, "step": 30510 }, { "grad_norm": 0.2581500709056854, "learning_rate": 8.322775995505057e-05, "loss": 0.0185, "step": 30520 }, { "grad_norm": 0.223882794380188, "learning_rate": 8.321540276415254e-05, "loss": 0.0273, "step": 30530 }, { "grad_norm": 0.2591458559036255, "learning_rate": 8.320304194086798e-05, "loss": 0.0174, "step": 30540 }, { "grad_norm": 0.3024096190929413, "learning_rate": 8.31906774865487e-05, "loss": 0.0174, "step": 30550 }, { "grad_norm": 0.368361234664917, "learning_rate": 8.317830940254682e-05, "loss": 0.022, "step": 30560 }, { "grad_norm": 0.27011561393737793, "learning_rate": 8.316593769021491e-05, "loss": 0.0211, "step": 30570 }, { "grad_norm": 0.24208246171474457, "learning_rate": 8.315356235090592e-05, "loss": 0.019, "step": 30580 }, { "grad_norm": 0.24897857010364532, "learning_rate": 8.314118338597319e-05, "loss": 0.0203, "step": 30590 }, { "grad_norm": 0.28901538252830505, "learning_rate": 8.312880079677048e-05, "loss": 0.0217, "step": 30600 }, { "grad_norm": 1.423928141593933, "learning_rate": 8.311641458465191e-05, "loss": 0.0161, "step": 30610 }, { "grad_norm": 0.22467470169067383, "learning_rate": 8.310402475097205e-05, "loss": 0.0288, "step": 30620 }, { "grad_norm": 0.23671811819076538, "learning_rate": 8.309163129708581e-05, "loss": 0.0241, "step": 30630 }, { "grad_norm": 0.33326953649520874, "learning_rate": 8.307923422434852e-05, "loss": 0.0171, "step": 30640 }, { "grad_norm": 0.21149228513240814, "learning_rate": 8.30668335341159e-05, "loss": 0.0159, "step": 30650 }, { "grad_norm": 0.31640058755874634, "learning_rate": 8.305442922774408e-05, "loss": 0.0174, "step": 30660 }, { "grad_norm": 0.1974901407957077, "learning_rate": 8.304202130658959e-05, "loss": 0.0189, "step": 30670 }, { "grad_norm": 0.3450503945350647, "learning_rate": 8.302960977200931e-05, "loss": 0.0236, "step": 30680 }, { "grad_norm": 0.2417769879102707, "learning_rate": 8.301719462536058e-05, "loss": 0.0147, "step": 30690 }, { "grad_norm": 0.2860408127307892, "learning_rate": 8.300477586800108e-05, "loss": 0.0136, "step": 30700 }, { "grad_norm": 0.2005816251039505, "learning_rate": 8.299235350128892e-05, "loss": 0.0171, "step": 30710 }, { "grad_norm": 0.7553102374076843, "learning_rate": 8.297992752658258e-05, "loss": 0.0182, "step": 30720 }, { "grad_norm": 0.23208169639110565, "learning_rate": 8.296749794524094e-05, "loss": 0.0197, "step": 30730 }, { "grad_norm": 0.1403857171535492, "learning_rate": 8.295506475862332e-05, "loss": 0.0188, "step": 30740 }, { "grad_norm": 0.1631281077861786, "learning_rate": 8.294262796808933e-05, "loss": 0.0161, "step": 30750 }, { "grad_norm": 0.2952505350112915, "learning_rate": 8.293018757499909e-05, "loss": 0.0158, "step": 30760 }, { "grad_norm": 0.24250862002372742, "learning_rate": 8.291774358071305e-05, "loss": 0.0178, "step": 30770 }, { "grad_norm": 0.17778687179088593, "learning_rate": 8.290529598659205e-05, "loss": 0.0138, "step": 30780 }, { "grad_norm": 0.17640505731105804, "learning_rate": 8.289284479399738e-05, "loss": 0.0151, "step": 30790 }, { "grad_norm": 0.2115875780582428, "learning_rate": 8.288039000429064e-05, "loss": 0.0116, "step": 30800 }, { "grad_norm": 0.16881588101387024, "learning_rate": 8.286793161883388e-05, "loss": 0.0109, "step": 30810 }, { "grad_norm": 0.46461451053619385, "learning_rate": 8.285546963898954e-05, "loss": 0.0122, "step": 30820 }, { "grad_norm": 0.18491970002651215, "learning_rate": 8.284300406612044e-05, "loss": 0.0138, "step": 30830 }, { "grad_norm": 0.17686490714550018, "learning_rate": 8.283053490158978e-05, "loss": 0.0167, "step": 30840 }, { "grad_norm": 0.20564047992229462, "learning_rate": 8.28180621467612e-05, "loss": 0.0158, "step": 30850 }, { "grad_norm": 0.1923818290233612, "learning_rate": 8.280558580299867e-05, "loss": 0.0128, "step": 30860 }, { "grad_norm": 0.22404572367668152, "learning_rate": 8.279310587166661e-05, "loss": 0.0145, "step": 30870 }, { "grad_norm": 0.18815016746520996, "learning_rate": 8.278062235412978e-05, "loss": 0.0146, "step": 30880 }, { "grad_norm": 0.18534724414348602, "learning_rate": 8.276813525175339e-05, "loss": 0.0108, "step": 30890 }, { "grad_norm": 0.2188030183315277, "learning_rate": 8.2755644565903e-05, "loss": 0.0106, "step": 30900 }, { "grad_norm": 0.23333023488521576, "learning_rate": 8.274315029794454e-05, "loss": 0.0164, "step": 30910 }, { "grad_norm": 0.21634414792060852, "learning_rate": 8.273065244924443e-05, "loss": 0.0121, "step": 30920 }, { "grad_norm": 0.19908958673477173, "learning_rate": 8.271815102116936e-05, "loss": 0.0109, "step": 30930 }, { "grad_norm": 0.2875290513038635, "learning_rate": 8.270564601508648e-05, "loss": 0.0126, "step": 30940 }, { "grad_norm": 2.1272799968719482, "learning_rate": 8.269313743236333e-05, "loss": 0.0166, "step": 30950 }, { "grad_norm": 0.18275929987430573, "learning_rate": 8.268062527436783e-05, "loss": 0.0143, "step": 30960 }, { "grad_norm": 0.18650013208389282, "learning_rate": 8.266810954246828e-05, "loss": 0.0118, "step": 30970 }, { "grad_norm": 0.18730536103248596, "learning_rate": 8.265559023803338e-05, "loss": 0.0138, "step": 30980 }, { "grad_norm": 0.23166662454605103, "learning_rate": 8.264306736243223e-05, "loss": 0.0155, "step": 30990 }, { "grad_norm": 0.2058776319026947, "learning_rate": 8.263054091703432e-05, "loss": 0.0168, "step": 31000 }, { "grad_norm": 0.19625656306743622, "learning_rate": 8.26180109032095e-05, "loss": 0.0133, "step": 31010 }, { "grad_norm": 0.18734320998191833, "learning_rate": 8.260547732232807e-05, "loss": 0.0143, "step": 31020 }, { "grad_norm": 0.26671671867370605, "learning_rate": 8.259294017576063e-05, "loss": 0.0158, "step": 31030 }, { "grad_norm": 0.27152425050735474, "learning_rate": 8.258039946487828e-05, "loss": 0.0146, "step": 31040 }, { "grad_norm": 0.30821776390075684, "learning_rate": 8.256785519105241e-05, "loss": 0.0162, "step": 31050 }, { "grad_norm": 0.24520686268806458, "learning_rate": 8.255530735565488e-05, "loss": 0.0185, "step": 31060 }, { "grad_norm": 0.30368298292160034, "learning_rate": 8.254275596005787e-05, "loss": 0.0155, "step": 31070 }, { "grad_norm": 0.32754603028297424, "learning_rate": 8.2530201005634e-05, "loss": 0.0148, "step": 31080 }, { "grad_norm": 0.27335044741630554, "learning_rate": 8.251764249375626e-05, "loss": 0.0154, "step": 31090 }, { "grad_norm": 0.23197129368782043, "learning_rate": 8.250508042579803e-05, "loss": 0.0172, "step": 31100 }, { "grad_norm": 0.2622259259223938, "learning_rate": 8.249251480313307e-05, "loss": 0.0194, "step": 31110 }, { "grad_norm": 0.30041563510894775, "learning_rate": 8.247994562713555e-05, "loss": 0.0145, "step": 31120 }, { "grad_norm": 0.24500754475593567, "learning_rate": 8.246737289917998e-05, "loss": 0.0192, "step": 31130 }, { "grad_norm": 0.2789945602416992, "learning_rate": 8.245479662064135e-05, "loss": 0.0145, "step": 31140 }, { "grad_norm": 2.7837908267974854, "learning_rate": 8.244221679289496e-05, "loss": 0.0228, "step": 31150 }, { "grad_norm": 0.19818373024463654, "learning_rate": 8.242963341731652e-05, "loss": 0.0177, "step": 31160 }, { "grad_norm": 0.18859194219112396, "learning_rate": 8.24170464952821e-05, "loss": 0.0192, "step": 31170 }, { "grad_norm": 0.1787056177854538, "learning_rate": 8.240445602816824e-05, "loss": 0.0114, "step": 31180 }, { "grad_norm": 0.20833879709243774, "learning_rate": 8.239186201735179e-05, "loss": 0.0165, "step": 31190 }, { "grad_norm": 0.2604990005493164, "learning_rate": 8.237926446420998e-05, "loss": 0.0125, "step": 31200 }, { "grad_norm": 0.19431549310684204, "learning_rate": 8.236666337012052e-05, "loss": 0.0143, "step": 31210 }, { "grad_norm": 0.22422926127910614, "learning_rate": 8.23540587364614e-05, "loss": 0.0143, "step": 31220 }, { "grad_norm": 0.23146946728229523, "learning_rate": 8.234145056461107e-05, "loss": 0.0132, "step": 31230 }, { "grad_norm": 0.28589117527008057, "learning_rate": 8.232883885594832e-05, "loss": 0.0143, "step": 31240 }, { "grad_norm": 0.5519518256187439, "learning_rate": 8.231622361185236e-05, "loss": 0.0127, "step": 31250 }, { "grad_norm": 0.14615176618099213, "learning_rate": 8.230360483370278e-05, "loss": 0.0113, "step": 31260 }, { "grad_norm": 0.23232829570770264, "learning_rate": 8.229098252287953e-05, "loss": 0.014, "step": 31270 }, { "grad_norm": 0.19340771436691284, "learning_rate": 8.2278356680763e-05, "loss": 0.0166, "step": 31280 }, { "grad_norm": 0.2572993040084839, "learning_rate": 8.22657273087339e-05, "loss": 0.0194, "step": 31290 }, { "grad_norm": 0.23441749811172485, "learning_rate": 8.225309440817336e-05, "loss": 0.022, "step": 31300 }, { "grad_norm": 0.28689342737197876, "learning_rate": 8.224045798046293e-05, "loss": 0.0185, "step": 31310 }, { "grad_norm": 0.34470102190971375, "learning_rate": 8.22278180269845e-05, "loss": 0.0181, "step": 31320 }, { "grad_norm": 0.3311502933502197, "learning_rate": 8.221517454912031e-05, "loss": 0.0183, "step": 31330 }, { "grad_norm": 0.27815642952919006, "learning_rate": 8.220252754825308e-05, "loss": 0.0192, "step": 31340 }, { "grad_norm": 0.24339626729488373, "learning_rate": 8.218987702576586e-05, "loss": 0.0146, "step": 31350 }, { "grad_norm": 0.2419242411851883, "learning_rate": 8.217722298304207e-05, "loss": 0.0213, "step": 31360 }, { "grad_norm": 0.31630468368530273, "learning_rate": 8.216456542146557e-05, "loss": 0.0183, "step": 31370 }, { "grad_norm": 0.23777183890342712, "learning_rate": 8.215190434242055e-05, "loss": 0.0134, "step": 31380 }, { "grad_norm": 0.3053566515445709, "learning_rate": 8.213923974729161e-05, "loss": 0.0148, "step": 31390 }, { "grad_norm": 0.20240004360675812, "learning_rate": 8.212657163746373e-05, "loss": 0.0159, "step": 31400 }, { "grad_norm": 0.27489233016967773, "learning_rate": 8.211390001432227e-05, "loss": 0.0183, "step": 31410 }, { "grad_norm": 0.25865471363067627, "learning_rate": 8.210122487925297e-05, "loss": 0.0155, "step": 31420 }, { "grad_norm": 0.22248683869838715, "learning_rate": 8.208854623364202e-05, "loss": 0.0148, "step": 31430 }, { "grad_norm": 0.16842548549175262, "learning_rate": 8.207586407887589e-05, "loss": 0.0147, "step": 31440 }, { "grad_norm": 3.531489849090576, "learning_rate": 8.206317841634148e-05, "loss": 0.0229, "step": 31450 }, { "grad_norm": 0.18593086302280426, "learning_rate": 8.205048924742609e-05, "loss": 0.0144, "step": 31460 }, { "grad_norm": 0.44942039251327515, "learning_rate": 8.203779657351738e-05, "loss": 0.0176, "step": 31470 }, { "grad_norm": 0.17705421149730682, "learning_rate": 8.20251003960034e-05, "loss": 0.0143, "step": 31480 }, { "grad_norm": 0.33896127343177795, "learning_rate": 8.201240071627258e-05, "loss": 0.0214, "step": 31490 }, { "grad_norm": 0.23509350419044495, "learning_rate": 8.199969753571377e-05, "loss": 0.0156, "step": 31500 }, { "grad_norm": 0.24028635025024414, "learning_rate": 8.198699085571615e-05, "loss": 0.0177, "step": 31510 }, { "grad_norm": 0.19599203765392303, "learning_rate": 8.197428067766928e-05, "loss": 0.0135, "step": 31520 }, { "grad_norm": 0.19862323999404907, "learning_rate": 8.196156700296316e-05, "loss": 0.0139, "step": 31530 }, { "grad_norm": 0.1429218351840973, "learning_rate": 8.194884983298814e-05, "loss": 0.0125, "step": 31540 }, { "grad_norm": 1.2072792053222656, "learning_rate": 8.193612916913491e-05, "loss": 0.0123, "step": 31550 }, { "grad_norm": 0.14136146008968353, "learning_rate": 8.192340501279463e-05, "loss": 0.0135, "step": 31560 }, { "grad_norm": 1.5518473386764526, "learning_rate": 8.191067736535876e-05, "loss": 0.0188, "step": 31570 }, { "grad_norm": 0.15041841566562653, "learning_rate": 8.18979462282192e-05, "loss": 0.0153, "step": 31580 }, { "grad_norm": 0.21698561310768127, "learning_rate": 8.188521160276819e-05, "loss": 0.0193, "step": 31590 }, { "grad_norm": 0.18541069328784943, "learning_rate": 8.187247349039837e-05, "loss": 0.0147, "step": 31600 }, { "grad_norm": 0.2779867947101593, "learning_rate": 8.185973189250278e-05, "loss": 0.0131, "step": 31610 }, { "grad_norm": 0.13451644778251648, "learning_rate": 8.184698681047482e-05, "loss": 0.0163, "step": 31620 }, { "grad_norm": 0.21580663323402405, "learning_rate": 8.183423824570827e-05, "loss": 0.0124, "step": 31630 }, { "grad_norm": 0.18815988302230835, "learning_rate": 8.182148619959725e-05, "loss": 0.0135, "step": 31640 }, { "grad_norm": 0.6868282556533813, "learning_rate": 8.180873067353636e-05, "loss": 0.0182, "step": 31650 }, { "grad_norm": 0.20672006905078888, "learning_rate": 8.179597166892052e-05, "loss": 0.015, "step": 31660 }, { "grad_norm": 0.3136695921421051, "learning_rate": 8.178320918714501e-05, "loss": 0.0159, "step": 31670 }, { "grad_norm": 0.7020270824432373, "learning_rate": 8.177044322960554e-05, "loss": 0.0148, "step": 31680 }, { "grad_norm": 0.19034181535243988, "learning_rate": 8.175767379769816e-05, "loss": 0.0141, "step": 31690 }, { "grad_norm": 0.16890797019004822, "learning_rate": 8.174490089281932e-05, "loss": 0.0123, "step": 31700 }, { "grad_norm": 0.22001928091049194, "learning_rate": 8.173212451636584e-05, "loss": 0.0159, "step": 31710 }, { "grad_norm": 0.21832695603370667, "learning_rate": 8.171934466973493e-05, "loss": 0.0148, "step": 31720 }, { "grad_norm": 0.18793156743049622, "learning_rate": 8.170656135432418e-05, "loss": 0.0139, "step": 31730 }, { "grad_norm": 0.23191937804222107, "learning_rate": 8.169377457153155e-05, "loss": 0.0152, "step": 31740 }, { "grad_norm": 1.403316617012024, "learning_rate": 8.168098432275539e-05, "loss": 0.0205, "step": 31750 }, { "grad_norm": 0.17944346368312836, "learning_rate": 8.166819060939442e-05, "loss": 0.0152, "step": 31760 }, { "grad_norm": 0.16316302120685577, "learning_rate": 8.165539343284772e-05, "loss": 0.0153, "step": 31770 }, { "grad_norm": 0.37409791350364685, "learning_rate": 8.16425927945148e-05, "loss": 0.0165, "step": 31780 }, { "grad_norm": 0.2059422731399536, "learning_rate": 8.162978869579551e-05, "loss": 0.0111, "step": 31790 }, { "grad_norm": 0.19856373965740204, "learning_rate": 8.161698113809007e-05, "loss": 0.0134, "step": 31800 }, { "grad_norm": 0.20395247638225555, "learning_rate": 8.160417012279911e-05, "loss": 0.0126, "step": 31810 }, { "grad_norm": 0.2873542308807373, "learning_rate": 8.159135565132363e-05, "loss": 0.0131, "step": 31820 }, { "grad_norm": 0.2061079740524292, "learning_rate": 8.157853772506498e-05, "loss": 0.0134, "step": 31830 }, { "grad_norm": 0.2600100040435791, "learning_rate": 8.156571634542494e-05, "loss": 0.0145, "step": 31840 }, { "grad_norm": 0.2580675482749939, "learning_rate": 8.15528915138056e-05, "loss": 0.0154, "step": 31850 }, { "grad_norm": 0.21623341739177704, "learning_rate": 8.154006323160949e-05, "loss": 0.0176, "step": 31860 }, { "grad_norm": 0.3818117678165436, "learning_rate": 8.152723150023949e-05, "loss": 0.0196, "step": 31870 }, { "grad_norm": 0.24694573879241943, "learning_rate": 8.151439632109886e-05, "loss": 0.0143, "step": 31880 }, { "grad_norm": 0.20483095943927765, "learning_rate": 8.150155769559122e-05, "loss": 0.0125, "step": 31890 }, { "grad_norm": 0.28026050329208374, "learning_rate": 8.148871562512058e-05, "loss": 0.0206, "step": 31900 }, { "grad_norm": 0.21136441826820374, "learning_rate": 8.147587011109136e-05, "loss": 0.0131, "step": 31910 }, { "grad_norm": 0.39038586616516113, "learning_rate": 8.14630211549083e-05, "loss": 0.0172, "step": 31920 }, { "grad_norm": 0.24434950947761536, "learning_rate": 8.145016875797655e-05, "loss": 0.0128, "step": 31930 }, { "grad_norm": 0.2475896030664444, "learning_rate": 8.143731292170164e-05, "loss": 0.0163, "step": 31940 }, { "grad_norm": 0.18665693700313568, "learning_rate": 8.142445364748944e-05, "loss": 0.0118, "step": 31950 }, { "grad_norm": 0.22596991062164307, "learning_rate": 8.141159093674624e-05, "loss": 0.0163, "step": 31960 }, { "grad_norm": 0.190566748380661, "learning_rate": 8.139872479087869e-05, "loss": 0.0108, "step": 31970 }, { "grad_norm": 0.24559757113456726, "learning_rate": 8.13858552112938e-05, "loss": 0.0156, "step": 31980 }, { "grad_norm": 0.2838849127292633, "learning_rate": 8.137298219939895e-05, "loss": 0.0194, "step": 31990 }, { "grad_norm": 0.3207560181617737, "learning_rate": 8.136010575660196e-05, "loss": 0.0151, "step": 32000 }, { "grad_norm": 0.18475770950317383, "learning_rate": 8.134722588431095e-05, "loss": 0.0137, "step": 32010 }, { "grad_norm": 0.25028368830680847, "learning_rate": 8.133434258393444e-05, "loss": 0.0136, "step": 32020 }, { "grad_norm": 0.2084513008594513, "learning_rate": 8.132145585688134e-05, "loss": 0.0157, "step": 32030 }, { "grad_norm": 0.23763693869113922, "learning_rate": 8.130856570456093e-05, "loss": 0.0165, "step": 32040 }, { "grad_norm": 0.17786331474781036, "learning_rate": 8.129567212838283e-05, "loss": 0.0132, "step": 32050 }, { "grad_norm": 0.8275978565216064, "learning_rate": 8.128277512975708e-05, "loss": 0.0168, "step": 32060 }, { "grad_norm": 0.141961470246315, "learning_rate": 8.126987471009408e-05, "loss": 0.0118, "step": 32070 }, { "grad_norm": 0.16032914817333221, "learning_rate": 8.125697087080459e-05, "loss": 0.0175, "step": 32080 }, { "grad_norm": 0.16904258728027344, "learning_rate": 8.124406361329976e-05, "loss": 0.0164, "step": 32090 }, { "grad_norm": 0.2785748243331909, "learning_rate": 8.123115293899111e-05, "loss": 0.0145, "step": 32100 }, { "grad_norm": 0.2040366679430008, "learning_rate": 8.121823884929055e-05, "loss": 0.0125, "step": 32110 }, { "grad_norm": 0.20028819143772125, "learning_rate": 8.12053213456103e-05, "loss": 0.0171, "step": 32120 }, { "grad_norm": 0.24084313213825226, "learning_rate": 8.119240042936303e-05, "loss": 0.0121, "step": 32130 }, { "grad_norm": 0.32353365421295166, "learning_rate": 8.117947610196175e-05, "loss": 0.0168, "step": 32140 }, { "grad_norm": 0.19796812534332275, "learning_rate": 8.116654836481982e-05, "loss": 0.0155, "step": 32150 }, { "grad_norm": 0.24203310906887054, "learning_rate": 8.115361721935106e-05, "loss": 0.0214, "step": 32160 }, { "grad_norm": 0.23155421018600464, "learning_rate": 8.114068266696953e-05, "loss": 0.0227, "step": 32170 }, { "grad_norm": 0.3895174264907837, "learning_rate": 8.112774470908978e-05, "loss": 0.0225, "step": 32180 }, { "grad_norm": 0.20370805263519287, "learning_rate": 8.111480334712665e-05, "loss": 0.0154, "step": 32190 }, { "grad_norm": 0.24029336869716644, "learning_rate": 8.110185858249542e-05, "loss": 0.018, "step": 32200 }, { "grad_norm": 0.2838888466358185, "learning_rate": 8.108891041661168e-05, "loss": 0.0144, "step": 32210 }, { "grad_norm": 0.26261866092681885, "learning_rate": 8.107595885089146e-05, "loss": 0.0153, "step": 32220 }, { "grad_norm": 0.24094121158123016, "learning_rate": 8.106300388675108e-05, "loss": 0.0164, "step": 32230 }, { "grad_norm": 0.32782819867134094, "learning_rate": 8.105004552560731e-05, "loss": 0.0147, "step": 32240 }, { "grad_norm": 0.19657382369041443, "learning_rate": 8.103708376887724e-05, "loss": 0.0141, "step": 32250 }, { "grad_norm": 0.7200812697410583, "learning_rate": 8.102411861797836e-05, "loss": 0.0122, "step": 32260 }, { "grad_norm": 0.5558229088783264, "learning_rate": 8.101115007432851e-05, "loss": 0.023, "step": 32270 }, { "grad_norm": 0.19357213377952576, "learning_rate": 8.09981781393459e-05, "loss": 0.0161, "step": 32280 }, { "grad_norm": 0.1844843327999115, "learning_rate": 8.098520281444915e-05, "loss": 0.0169, "step": 32290 }, { "grad_norm": 0.27293285727500916, "learning_rate": 8.09722241010572e-05, "loss": 0.0166, "step": 32300 }, { "grad_norm": 0.21316654980182648, "learning_rate": 8.095924200058939e-05, "loss": 0.0132, "step": 32310 }, { "grad_norm": 0.17197437584400177, "learning_rate": 8.094625651446541e-05, "loss": 0.0188, "step": 32320 }, { "grad_norm": 0.2949024438858032, "learning_rate": 8.093326764410536e-05, "loss": 0.022, "step": 32330 }, { "grad_norm": 0.2561020255088806, "learning_rate": 8.092027539092966e-05, "loss": 0.0242, "step": 32340 }, { "grad_norm": 0.3020962178707123, "learning_rate": 8.090727975635913e-05, "loss": 0.0201, "step": 32350 }, { "grad_norm": 0.3350811302661896, "learning_rate": 8.089428074181497e-05, "loss": 0.0252, "step": 32360 }, { "grad_norm": 0.288692831993103, "learning_rate": 8.088127834871871e-05, "loss": 0.0206, "step": 32370 }, { "grad_norm": 0.26772332191467285, "learning_rate": 8.086827257849226e-05, "loss": 0.0191, "step": 32380 }, { "grad_norm": 0.2478875070810318, "learning_rate": 8.085526343255795e-05, "loss": 0.0237, "step": 32390 }, { "grad_norm": 0.2568882405757904, "learning_rate": 8.084225091233842e-05, "loss": 0.0174, "step": 32400 }, { "grad_norm": 0.16425621509552002, "learning_rate": 8.082923501925668e-05, "loss": 0.0135, "step": 32410 }, { "grad_norm": 0.15802748501300812, "learning_rate": 8.081621575473617e-05, "loss": 0.0174, "step": 32420 }, { "grad_norm": 0.21019965410232544, "learning_rate": 8.080319312020064e-05, "loss": 0.0196, "step": 32430 }, { "grad_norm": 0.34197279810905457, "learning_rate": 8.079016711707421e-05, "loss": 0.0153, "step": 32440 }, { "grad_norm": 0.3888288140296936, "learning_rate": 8.077713774678139e-05, "loss": 0.0174, "step": 32450 }, { "grad_norm": 0.24329829216003418, "learning_rate": 8.076410501074707e-05, "loss": 0.0137, "step": 32460 }, { "grad_norm": 0.2278023511171341, "learning_rate": 8.075106891039647e-05, "loss": 0.0157, "step": 32470 }, { "grad_norm": 0.5086213946342468, "learning_rate": 8.073802944715523e-05, "loss": 0.0143, "step": 32480 }, { "grad_norm": 0.21448557078838348, "learning_rate": 8.072498662244929e-05, "loss": 0.0163, "step": 32490 }, { "grad_norm": 0.20382757484912872, "learning_rate": 8.0711940437705e-05, "loss": 0.017, "step": 32500 }, { "grad_norm": 0.1491343230009079, "learning_rate": 8.06988908943491e-05, "loss": 0.0137, "step": 32510 }, { "grad_norm": 0.857982873916626, "learning_rate": 8.068583799380863e-05, "loss": 0.0184, "step": 32520 }, { "grad_norm": 1.2713159322738647, "learning_rate": 8.067278173751104e-05, "loss": 0.0213, "step": 32530 }, { "grad_norm": 0.28902295231819153, "learning_rate": 8.065972212688417e-05, "loss": 0.02, "step": 32540 }, { "grad_norm": 0.2905252277851105, "learning_rate": 8.064665916335618e-05, "loss": 0.0142, "step": 32550 }, { "grad_norm": 0.3457854092121124, "learning_rate": 8.063359284835564e-05, "loss": 0.0189, "step": 32560 }, { "grad_norm": 0.3853873610496521, "learning_rate": 8.062052318331142e-05, "loss": 0.0167, "step": 32570 }, { "grad_norm": 0.29614952206611633, "learning_rate": 8.060745016965283e-05, "loss": 0.025, "step": 32580 }, { "grad_norm": 0.2369009554386139, "learning_rate": 8.059437380880952e-05, "loss": 0.0196, "step": 32590 }, { "grad_norm": 0.2859675884246826, "learning_rate": 8.058129410221146e-05, "loss": 0.0173, "step": 32600 }, { "grad_norm": 2.406097650527954, "learning_rate": 8.056821105128908e-05, "loss": 0.0206, "step": 32610 }, { "grad_norm": 0.15962669253349304, "learning_rate": 8.05551246574731e-05, "loss": 0.0256, "step": 32620 }, { "grad_norm": 0.2816001772880554, "learning_rate": 8.05420349221946e-05, "loss": 0.0191, "step": 32630 }, { "grad_norm": 0.26398390531539917, "learning_rate": 8.05289418468851e-05, "loss": 0.0176, "step": 32640 }, { "grad_norm": 0.15645337104797363, "learning_rate": 8.051584543297642e-05, "loss": 0.0176, "step": 32650 }, { "grad_norm": 0.18632249534130096, "learning_rate": 8.050274568190074e-05, "loss": 0.0181, "step": 32660 }, { "grad_norm": 0.23844781517982483, "learning_rate": 8.048964259509067e-05, "loss": 0.0174, "step": 32670 }, { "grad_norm": 0.2884618043899536, "learning_rate": 8.047653617397914e-05, "loss": 0.016, "step": 32680 }, { "grad_norm": 0.17524665594100952, "learning_rate": 8.046342641999941e-05, "loss": 0.012, "step": 32690 }, { "grad_norm": 0.21931545436382294, "learning_rate": 8.045031333458517e-05, "loss": 0.0158, "step": 32700 }, { "grad_norm": 0.21437925100326538, "learning_rate": 8.043719691917047e-05, "loss": 0.0146, "step": 32710 }, { "grad_norm": 0.24542133510112762, "learning_rate": 8.042407717518966e-05, "loss": 0.0157, "step": 32720 }, { "grad_norm": 0.21579763293266296, "learning_rate": 8.041095410407751e-05, "loss": 0.0131, "step": 32730 }, { "grad_norm": 0.19122754037380219, "learning_rate": 8.039782770726913e-05, "loss": 0.0166, "step": 32740 }, { "grad_norm": 0.18842540681362152, "learning_rate": 8.038469798620004e-05, "loss": 0.015, "step": 32750 }, { "grad_norm": 1.1067546606063843, "learning_rate": 8.037156494230604e-05, "loss": 0.015, "step": 32760 }, { "grad_norm": 0.21988725662231445, "learning_rate": 8.035842857702338e-05, "loss": 0.0195, "step": 32770 }, { "grad_norm": 0.22462314367294312, "learning_rate": 8.03452888917886e-05, "loss": 0.0177, "step": 32780 }, { "grad_norm": 0.3910283148288727, "learning_rate": 8.033214588803866e-05, "loss": 0.0123, "step": 32790 }, { "grad_norm": 0.1593242585659027, "learning_rate": 8.031899956721083e-05, "loss": 0.0191, "step": 32800 }, { "grad_norm": 0.16123484075069427, "learning_rate": 8.030584993074282e-05, "loss": 0.0144, "step": 32810 }, { "grad_norm": 0.19678105413913727, "learning_rate": 8.02926969800726e-05, "loss": 0.0151, "step": 32820 }, { "grad_norm": 0.17564712464809418, "learning_rate": 8.027954071663859e-05, "loss": 0.0147, "step": 32830 }, { "grad_norm": 0.2218686044216156, "learning_rate": 8.026638114187954e-05, "loss": 0.0263, "step": 32840 }, { "grad_norm": 0.17025932669639587, "learning_rate": 8.025321825723456e-05, "loss": 0.0172, "step": 32850 }, { "grad_norm": 0.2215489149093628, "learning_rate": 8.02400520641431e-05, "loss": 0.0139, "step": 32860 }, { "grad_norm": 0.2013855278491974, "learning_rate": 8.022688256404501e-05, "loss": 0.012, "step": 32870 }, { "grad_norm": 0.19604600965976715, "learning_rate": 8.02137097583805e-05, "loss": 0.0157, "step": 32880 }, { "grad_norm": 0.1626562923192978, "learning_rate": 8.02005336485901e-05, "loss": 0.0167, "step": 32890 }, { "grad_norm": 0.18765375018119812, "learning_rate": 8.018735423611476e-05, "loss": 0.0135, "step": 32900 }, { "grad_norm": 0.1625117063522339, "learning_rate": 8.017417152239574e-05, "loss": 0.0191, "step": 32910 }, { "grad_norm": 0.2119964361190796, "learning_rate": 8.01609855088747e-05, "loss": 0.0134, "step": 32920 }, { "grad_norm": 0.8468178510665894, "learning_rate": 8.014779619699362e-05, "loss": 0.0174, "step": 32930 }, { "grad_norm": 0.22432513535022736, "learning_rate": 8.013460358819489e-05, "loss": 0.0173, "step": 32940 }, { "grad_norm": 0.17176057398319244, "learning_rate": 8.01214076839212e-05, "loss": 0.0182, "step": 32950 }, { "grad_norm": 0.19863978028297424, "learning_rate": 8.010820848561565e-05, "loss": 0.0192, "step": 32960 }, { "grad_norm": 0.25148236751556396, "learning_rate": 8.009500599472171e-05, "loss": 0.0174, "step": 32970 }, { "grad_norm": 0.21389299631118774, "learning_rate": 8.008180021268314e-05, "loss": 0.0153, "step": 32980 }, { "grad_norm": 0.15858763456344604, "learning_rate": 8.006859114094414e-05, "loss": 0.0144, "step": 32990 }, { "grad_norm": 0.21117378771305084, "learning_rate": 8.005537878094921e-05, "loss": 0.0151, "step": 33000 }, { "grad_norm": 0.13685040175914764, "learning_rate": 8.004216313414323e-05, "loss": 0.0154, "step": 33010 }, { "grad_norm": 0.6403921246528625, "learning_rate": 8.002894420197149e-05, "loss": 0.0127, "step": 33020 }, { "grad_norm": 0.1624477505683899, "learning_rate": 8.001572198587954e-05, "loss": 0.0132, "step": 33030 }, { "grad_norm": 0.13069719076156616, "learning_rate": 8.000249648731338e-05, "loss": 0.0116, "step": 33040 }, { "grad_norm": 0.18330782651901245, "learning_rate": 7.998926770771928e-05, "loss": 0.018, "step": 33050 }, { "grad_norm": 0.21391884982585907, "learning_rate": 7.997603564854397e-05, "loss": 0.0122, "step": 33060 }, { "grad_norm": 0.17820744216442108, "learning_rate": 7.996280031123448e-05, "loss": 0.0126, "step": 33070 }, { "grad_norm": 0.21607106924057007, "learning_rate": 7.994956169723818e-05, "loss": 0.0159, "step": 33080 }, { "grad_norm": 0.16685181856155396, "learning_rate": 7.993631980800285e-05, "loss": 0.0116, "step": 33090 }, { "grad_norm": 0.18241004645824432, "learning_rate": 7.992307464497659e-05, "loss": 0.0159, "step": 33100 }, { "grad_norm": 0.18104329705238342, "learning_rate": 7.990982620960787e-05, "loss": 0.0156, "step": 33110 }, { "grad_norm": 0.14306676387786865, "learning_rate": 7.989657450334554e-05, "loss": 0.0203, "step": 33120 }, { "grad_norm": 0.17902213335037231, "learning_rate": 7.988331952763877e-05, "loss": 0.0161, "step": 33130 }, { "grad_norm": 0.276770681142807, "learning_rate": 7.987006128393709e-05, "loss": 0.0146, "step": 33140 }, { "grad_norm": 0.21113979816436768, "learning_rate": 7.985679977369043e-05, "loss": 0.0145, "step": 33150 }, { "grad_norm": 0.14346911013126373, "learning_rate": 7.984353499834902e-05, "loss": 0.0111, "step": 33160 }, { "grad_norm": 0.2056216150522232, "learning_rate": 7.983026695936351e-05, "loss": 0.0131, "step": 33170 }, { "grad_norm": 0.21252721548080444, "learning_rate": 7.981699565818486e-05, "loss": 0.0152, "step": 33180 }, { "grad_norm": 0.2690648138523102, "learning_rate": 7.980372109626437e-05, "loss": 0.0154, "step": 33190 }, { "grad_norm": 0.17343349754810333, "learning_rate": 7.979044327505375e-05, "loss": 0.0134, "step": 33200 }, { "grad_norm": 0.17882075905799866, "learning_rate": 7.977716219600506e-05, "loss": 0.0153, "step": 33210 }, { "grad_norm": 0.20083186030387878, "learning_rate": 7.97638778605707e-05, "loss": 0.0132, "step": 33220 }, { "grad_norm": 0.23865525424480438, "learning_rate": 7.975059027020338e-05, "loss": 0.0145, "step": 33230 }, { "grad_norm": 0.1926245093345642, "learning_rate": 7.973729942635623e-05, "loss": 0.0138, "step": 33240 }, { "grad_norm": 0.3102167546749115, "learning_rate": 7.972400533048273e-05, "loss": 0.0171, "step": 33250 }, { "grad_norm": 0.29023998975753784, "learning_rate": 7.97107079840367e-05, "loss": 0.0135, "step": 33260 }, { "grad_norm": 0.35996779799461365, "learning_rate": 7.969740738847231e-05, "loss": 0.0199, "step": 33270 }, { "grad_norm": 0.34920603036880493, "learning_rate": 7.968410354524411e-05, "loss": 0.015, "step": 33280 }, { "grad_norm": 0.23830650746822357, "learning_rate": 7.967079645580697e-05, "loss": 0.013, "step": 33290 }, { "grad_norm": 0.30102792382240295, "learning_rate": 7.965748612161612e-05, "loss": 0.0172, "step": 33300 }, { "grad_norm": 0.20466962456703186, "learning_rate": 7.96441725441272e-05, "loss": 0.0167, "step": 33310 }, { "grad_norm": 0.21183954179286957, "learning_rate": 7.963085572479614e-05, "loss": 0.0194, "step": 33320 }, { "grad_norm": 0.37279319763183594, "learning_rate": 7.961753566507924e-05, "loss": 0.0182, "step": 33330 }, { "grad_norm": 0.29413679242134094, "learning_rate": 7.960421236643316e-05, "loss": 0.0187, "step": 33340 }, { "grad_norm": 0.29855549335479736, "learning_rate": 7.959088583031496e-05, "loss": 0.0183, "step": 33350 }, { "grad_norm": 0.2846105098724365, "learning_rate": 7.957755605818194e-05, "loss": 0.0219, "step": 33360 }, { "grad_norm": 0.32807469367980957, "learning_rate": 7.956422305149185e-05, "loss": 0.0228, "step": 33370 }, { "grad_norm": 0.43918463587760925, "learning_rate": 7.95508868117028e-05, "loss": 0.0227, "step": 33380 }, { "grad_norm": 0.2922639846801758, "learning_rate": 7.953754734027318e-05, "loss": 0.0227, "step": 33390 }, { "grad_norm": 0.25406551361083984, "learning_rate": 7.952420463866182e-05, "loss": 0.0177, "step": 33400 }, { "grad_norm": 0.3688148260116577, "learning_rate": 7.951085870832782e-05, "loss": 0.0183, "step": 33410 }, { "grad_norm": 0.2515830099582672, "learning_rate": 7.949750955073067e-05, "loss": 0.0158, "step": 33420 }, { "grad_norm": 0.17578986287117004, "learning_rate": 7.948415716733022e-05, "loss": 0.0162, "step": 33430 }, { "grad_norm": 0.16621534526348114, "learning_rate": 7.947080155958669e-05, "loss": 0.0122, "step": 33440 }, { "grad_norm": 0.21519677340984344, "learning_rate": 7.94574427289606e-05, "loss": 0.018, "step": 33450 }, { "grad_norm": 0.19597811996936798, "learning_rate": 7.944408067691284e-05, "loss": 0.0164, "step": 33460 }, { "grad_norm": 0.17837683856487274, "learning_rate": 7.943071540490473e-05, "loss": 0.0159, "step": 33470 }, { "grad_norm": 0.17684900760650635, "learning_rate": 7.94173469143978e-05, "loss": 0.0149, "step": 33480 }, { "grad_norm": 0.22750915586948395, "learning_rate": 7.940397520685406e-05, "loss": 0.0163, "step": 33490 }, { "grad_norm": 0.22967833280563354, "learning_rate": 7.939060028373577e-05, "loss": 0.015, "step": 33500 }, { "grad_norm": 0.17098014056682587, "learning_rate": 7.937722214650565e-05, "loss": 0.0174, "step": 33510 }, { "grad_norm": 0.2134232223033905, "learning_rate": 7.936384079662666e-05, "loss": 0.0136, "step": 33520 }, { "grad_norm": 0.3335719406604767, "learning_rate": 7.93504562355622e-05, "loss": 0.0193, "step": 33530 }, { "grad_norm": 0.21063855290412903, "learning_rate": 7.933706846477599e-05, "loss": 0.0136, "step": 33540 }, { "grad_norm": 0.18708236515522003, "learning_rate": 7.932367748573206e-05, "loss": 0.018, "step": 33550 }, { "grad_norm": 0.27419543266296387, "learning_rate": 7.931028329989485e-05, "loss": 0.0137, "step": 33560 }, { "grad_norm": 0.2983235716819763, "learning_rate": 7.929688590872913e-05, "loss": 0.0135, "step": 33570 }, { "grad_norm": 0.20831014215946198, "learning_rate": 7.928348531370003e-05, "loss": 0.0151, "step": 33580 }, { "grad_norm": 0.3028661012649536, "learning_rate": 7.927008151627297e-05, "loss": 0.0156, "step": 33590 }, { "grad_norm": 0.15365152060985565, "learning_rate": 7.925667451791383e-05, "loss": 0.0156, "step": 33600 }, { "grad_norm": 0.19565320014953613, "learning_rate": 7.924326432008874e-05, "loss": 0.0119, "step": 33610 }, { "grad_norm": 0.5616907477378845, "learning_rate": 7.922985092426422e-05, "loss": 0.018, "step": 33620 }, { "grad_norm": 0.16739338636398315, "learning_rate": 7.921643433190717e-05, "loss": 0.0153, "step": 33630 }, { "grad_norm": 0.17178307473659515, "learning_rate": 7.920301454448478e-05, "loss": 0.0153, "step": 33640 }, { "grad_norm": 0.1362619549036026, "learning_rate": 7.918959156346461e-05, "loss": 0.0151, "step": 33650 }, { "grad_norm": 0.2874087393283844, "learning_rate": 7.91761653903146e-05, "loss": 0.0184, "step": 33660 }, { "grad_norm": 0.8417396545410156, "learning_rate": 7.916273602650302e-05, "loss": 0.0177, "step": 33670 }, { "grad_norm": 0.22480565309524536, "learning_rate": 7.914930347349847e-05, "loss": 0.011, "step": 33680 }, { "grad_norm": 0.35921695828437805, "learning_rate": 7.913586773276992e-05, "loss": 0.0191, "step": 33690 }, { "grad_norm": 0.7933383584022522, "learning_rate": 7.912242880578667e-05, "loss": 0.0191, "step": 33700 }, { "grad_norm": 0.7703676819801331, "learning_rate": 7.910898669401839e-05, "loss": 0.0211, "step": 33710 }, { "grad_norm": 0.1814373880624771, "learning_rate": 7.909554139893511e-05, "loss": 0.0145, "step": 33720 }, { "grad_norm": 0.15035194158554077, "learning_rate": 7.908209292200715e-05, "loss": 0.0163, "step": 33730 }, { "grad_norm": 0.23261956870555878, "learning_rate": 7.906864126470523e-05, "loss": 0.0167, "step": 33740 }, { "grad_norm": 0.13876153528690338, "learning_rate": 7.905518642850041e-05, "loss": 0.0159, "step": 33750 }, { "grad_norm": 0.14593711495399475, "learning_rate": 7.904172841486409e-05, "loss": 0.0168, "step": 33760 }, { "grad_norm": 0.208529993891716, "learning_rate": 7.902826722526801e-05, "loss": 0.0159, "step": 33770 }, { "grad_norm": 0.18111248314380646, "learning_rate": 7.901480286118427e-05, "loss": 0.0138, "step": 33780 }, { "grad_norm": 0.10901843756437302, "learning_rate": 7.900133532408531e-05, "loss": 0.0127, "step": 33790 }, { "grad_norm": 0.17100635170936584, "learning_rate": 7.898786461544395e-05, "loss": 0.0154, "step": 33800 }, { "grad_norm": 0.18696509301662445, "learning_rate": 7.897439073673325e-05, "loss": 0.0151, "step": 33810 }, { "grad_norm": 0.17535577714443207, "learning_rate": 7.896091368942677e-05, "loss": 0.0142, "step": 33820 }, { "grad_norm": 0.20169706642627716, "learning_rate": 7.894743347499832e-05, "loss": 0.0117, "step": 33830 }, { "grad_norm": 0.20957276225090027, "learning_rate": 7.893395009492203e-05, "loss": 0.0154, "step": 33840 }, { "grad_norm": 0.1930110901594162, "learning_rate": 7.892046355067248e-05, "loss": 0.0194, "step": 33850 }, { "grad_norm": 0.19044263660907745, "learning_rate": 7.890697384372451e-05, "loss": 0.0135, "step": 33860 }, { "grad_norm": 0.1870979368686676, "learning_rate": 7.889348097555336e-05, "loss": 0.0123, "step": 33870 }, { "grad_norm": 0.24091500043869019, "learning_rate": 7.887998494763455e-05, "loss": 0.0122, "step": 33880 }, { "grad_norm": 0.21949413418769836, "learning_rate": 7.886648576144404e-05, "loss": 0.0143, "step": 33890 }, { "grad_norm": 0.2649540901184082, "learning_rate": 7.885298341845802e-05, "loss": 0.0156, "step": 33900 }, { "grad_norm": 0.2549557685852051, "learning_rate": 7.883947792015311e-05, "loss": 0.0131, "step": 33910 }, { "grad_norm": 0.3715779483318329, "learning_rate": 7.882596926800628e-05, "loss": 0.0166, "step": 33920 }, { "grad_norm": 0.15041793882846832, "learning_rate": 7.881245746349477e-05, "loss": 0.0126, "step": 33930 }, { "grad_norm": 0.16618433594703674, "learning_rate": 7.879894250809623e-05, "loss": 0.0107, "step": 33940 }, { "grad_norm": 0.2828567624092102, "learning_rate": 7.878542440328865e-05, "loss": 0.0151, "step": 33950 }, { "grad_norm": 0.17726734280586243, "learning_rate": 7.877190315055031e-05, "loss": 0.0195, "step": 33960 }, { "grad_norm": 0.18011057376861572, "learning_rate": 7.875837875135991e-05, "loss": 0.0212, "step": 33970 }, { "grad_norm": 0.19672009348869324, "learning_rate": 7.874485120719646e-05, "loss": 0.0134, "step": 33980 }, { "grad_norm": 0.17670570313930511, "learning_rate": 7.873132051953928e-05, "loss": 0.0139, "step": 33990 }, { "grad_norm": 0.22574447095394135, "learning_rate": 7.87177866898681e-05, "loss": 0.0157, "step": 34000 }, { "grad_norm": 0.3018534779548645, "learning_rate": 7.870424971966294e-05, "loss": 0.0158, "step": 34010 }, { "grad_norm": 0.23805473744869232, "learning_rate": 7.869070961040419e-05, "loss": 0.0135, "step": 34020 }, { "grad_norm": 0.24881738424301147, "learning_rate": 7.867716636357257e-05, "loss": 0.0193, "step": 34030 }, { "grad_norm": 0.1902012825012207, "learning_rate": 7.866361998064915e-05, "loss": 0.0172, "step": 34040 }, { "grad_norm": 0.19462703168392181, "learning_rate": 7.865007046311534e-05, "loss": 0.0132, "step": 34050 }, { "grad_norm": 0.4792376756668091, "learning_rate": 7.86365178124529e-05, "loss": 0.0157, "step": 34060 }, { "grad_norm": 0.31559208035469055, "learning_rate": 7.862296203014394e-05, "loss": 0.019, "step": 34070 }, { "grad_norm": 0.24239479005336761, "learning_rate": 7.860940311767088e-05, "loss": 0.0134, "step": 34080 }, { "grad_norm": 0.1398112177848816, "learning_rate": 7.85958410765165e-05, "loss": 0.0094, "step": 34090 }, { "grad_norm": 0.1484067440032959, "learning_rate": 7.858227590816394e-05, "loss": 0.013, "step": 34100 }, { "grad_norm": 0.25094208121299744, "learning_rate": 7.856870761409664e-05, "loss": 0.0161, "step": 34110 }, { "grad_norm": 0.1683342307806015, "learning_rate": 7.855513619579846e-05, "loss": 0.0162, "step": 34120 }, { "grad_norm": 0.14141422510147095, "learning_rate": 7.85415616547535e-05, "loss": 0.0097, "step": 34130 }, { "grad_norm": 0.18363718688488007, "learning_rate": 7.852798399244627e-05, "loss": 0.0147, "step": 34140 }, { "grad_norm": 0.14214400947093964, "learning_rate": 7.851440321036161e-05, "loss": 0.0162, "step": 34150 }, { "grad_norm": 0.2815537750720978, "learning_rate": 7.850081930998468e-05, "loss": 0.0141, "step": 34160 }, { "grad_norm": 0.13395747542381287, "learning_rate": 7.8487232292801e-05, "loss": 0.0122, "step": 34170 }, { "grad_norm": 0.19620954990386963, "learning_rate": 7.847364216029642e-05, "loss": 0.0127, "step": 34180 }, { "grad_norm": 0.14928565919399261, "learning_rate": 7.846004891395716e-05, "loss": 0.0133, "step": 34190 }, { "grad_norm": 0.2122565656900406, "learning_rate": 7.844645255526972e-05, "loss": 0.0182, "step": 34200 }, { "grad_norm": 0.2198401540517807, "learning_rate": 7.843285308572101e-05, "loss": 0.0165, "step": 34210 }, { "grad_norm": 0.27551448345184326, "learning_rate": 7.841925050679823e-05, "loss": 0.016, "step": 34220 }, { "grad_norm": 0.2514117360115051, "learning_rate": 7.840564481998895e-05, "loss": 0.0159, "step": 34230 }, { "grad_norm": 0.3770003914833069, "learning_rate": 7.839203602678105e-05, "loss": 0.0162, "step": 34240 }, { "grad_norm": 0.21251584589481354, "learning_rate": 7.837842412866279e-05, "loss": 0.0146, "step": 34250 }, { "grad_norm": 0.18825098872184753, "learning_rate": 7.836480912712272e-05, "loss": 0.0157, "step": 34260 }, { "grad_norm": 0.18316030502319336, "learning_rate": 7.835119102364976e-05, "loss": 0.0145, "step": 34270 }, { "grad_norm": 0.1844855695962906, "learning_rate": 7.83375698197332e-05, "loss": 0.0162, "step": 34280 }, { "grad_norm": 0.42917466163635254, "learning_rate": 7.83239455168626e-05, "loss": 0.0197, "step": 34290 }, { "grad_norm": 0.22217626869678497, "learning_rate": 7.83103181165279e-05, "loss": 0.0236, "step": 34300 }, { "grad_norm": 0.21674033999443054, "learning_rate": 7.829668762021937e-05, "loss": 0.0214, "step": 34310 }, { "grad_norm": 0.24928459525108337, "learning_rate": 7.828305402942764e-05, "loss": 0.0165, "step": 34320 }, { "grad_norm": 0.4894373118877411, "learning_rate": 7.826941734564363e-05, "loss": 0.0153, "step": 34330 }, { "grad_norm": 0.18807242810726166, "learning_rate": 7.825577757035865e-05, "loss": 0.0111, "step": 34340 }, { "grad_norm": 0.1575649082660675, "learning_rate": 7.824213470506431e-05, "loss": 0.0141, "step": 34350 }, { "grad_norm": 0.21236565709114075, "learning_rate": 7.822848875125257e-05, "loss": 0.0159, "step": 34360 }, { "grad_norm": 0.1573205590248108, "learning_rate": 7.821483971041576e-05, "loss": 0.0127, "step": 34370 }, { "grad_norm": 0.3634432256221771, "learning_rate": 7.820118758404649e-05, "loss": 0.0149, "step": 34380 }, { "grad_norm": 0.16696135699748993, "learning_rate": 7.818753237363776e-05, "loss": 0.014, "step": 34390 }, { "grad_norm": 0.2379564791917801, "learning_rate": 7.817387408068286e-05, "loss": 0.0144, "step": 34400 }, { "grad_norm": 0.2517138421535492, "learning_rate": 7.816021270667544e-05, "loss": 0.0156, "step": 34410 }, { "grad_norm": 0.17902030050754547, "learning_rate": 7.81465482531095e-05, "loss": 0.0137, "step": 34420 }, { "grad_norm": 0.3197586238384247, "learning_rate": 7.813288072147938e-05, "loss": 0.0255, "step": 34430 }, { "grad_norm": 0.23772548139095306, "learning_rate": 7.811921011327972e-05, "loss": 0.018, "step": 34440 }, { "grad_norm": 0.3430637717247009, "learning_rate": 7.810553643000549e-05, "loss": 0.0175, "step": 34450 }, { "grad_norm": 0.2667028307914734, "learning_rate": 7.809185967315206e-05, "loss": 0.0151, "step": 34460 }, { "grad_norm": 0.2903631925582886, "learning_rate": 7.80781798442151e-05, "loss": 0.0168, "step": 34470 }, { "grad_norm": 0.25114813446998596, "learning_rate": 7.806449694469061e-05, "loss": 0.0148, "step": 34480 }, { "grad_norm": 0.20875103771686554, "learning_rate": 7.805081097607492e-05, "loss": 0.0161, "step": 34490 }, { "grad_norm": 0.545311689376831, "learning_rate": 7.803712193986474e-05, "loss": 0.014, "step": 34500 }, { "grad_norm": 0.25582659244537354, "learning_rate": 7.802342983755702e-05, "loss": 0.0154, "step": 34510 }, { "grad_norm": 0.20260092616081238, "learning_rate": 7.800973467064918e-05, "loss": 0.0155, "step": 34520 }, { "grad_norm": 0.275881290435791, "learning_rate": 7.799603644063884e-05, "loss": 0.0134, "step": 34530 }, { "grad_norm": 0.24596859514713287, "learning_rate": 7.798233514902405e-05, "loss": 0.0121, "step": 34540 }, { "grad_norm": 0.27046316862106323, "learning_rate": 7.796863079730318e-05, "loss": 0.0145, "step": 34550 }, { "grad_norm": 0.260335236787796, "learning_rate": 7.795492338697488e-05, "loss": 0.0137, "step": 34560 }, { "grad_norm": 0.2238619327545166, "learning_rate": 7.794121291953819e-05, "loss": 0.018, "step": 34570 }, { "grad_norm": 0.18978077173233032, "learning_rate": 7.792749939649246e-05, "loss": 0.0183, "step": 34580 }, { "grad_norm": 0.2024185061454773, "learning_rate": 7.79137828193374e-05, "loss": 0.0137, "step": 34590 }, { "grad_norm": 0.21248073875904083, "learning_rate": 7.790006318957301e-05, "loss": 0.016, "step": 34600 }, { "grad_norm": 0.17418280243873596, "learning_rate": 7.788634050869965e-05, "loss": 0.0121, "step": 34610 }, { "grad_norm": 0.22580792009830475, "learning_rate": 7.787261477821803e-05, "loss": 0.0168, "step": 34620 }, { "grad_norm": 0.1358105093240738, "learning_rate": 7.785888599962916e-05, "loss": 0.0122, "step": 34630 }, { "grad_norm": 0.17848126590251923, "learning_rate": 7.784515417443439e-05, "loss": 0.0123, "step": 34640 }, { "grad_norm": 0.17793627083301544, "learning_rate": 7.783141930413545e-05, "loss": 0.0159, "step": 34650 }, { "grad_norm": 0.3336009979248047, "learning_rate": 7.78176813902343e-05, "loss": 0.0147, "step": 34660 }, { "grad_norm": 0.20606908202171326, "learning_rate": 7.780394043423336e-05, "loss": 0.0169, "step": 34670 }, { "grad_norm": 0.20357085764408112, "learning_rate": 7.77901964376353e-05, "loss": 0.0125, "step": 34680 }, { "grad_norm": 0.19074852764606476, "learning_rate": 7.777644940194316e-05, "loss": 0.014, "step": 34690 }, { "grad_norm": 0.2683101296424866, "learning_rate": 7.776269932866023e-05, "loss": 0.0111, "step": 34700 }, { "grad_norm": 0.17844077944755554, "learning_rate": 7.774894621929026e-05, "loss": 0.0109, "step": 34710 }, { "grad_norm": 0.20158594846725464, "learning_rate": 7.773519007533725e-05, "loss": 0.0152, "step": 34720 }, { "grad_norm": 0.36133110523223877, "learning_rate": 7.772143089830556e-05, "loss": 0.0169, "step": 34730 }, { "grad_norm": 0.26438087224960327, "learning_rate": 7.770766868969985e-05, "loss": 0.018, "step": 34740 }, { "grad_norm": 0.16115646064281464, "learning_rate": 7.769390345102518e-05, "loss": 0.0164, "step": 34750 }, { "grad_norm": 0.24118772149085999, "learning_rate": 7.768013518378683e-05, "loss": 0.0116, "step": 34760 }, { "grad_norm": 0.1696653962135315, "learning_rate": 7.766636388949053e-05, "loss": 0.0175, "step": 34770 }, { "grad_norm": 0.2610183358192444, "learning_rate": 7.765258956964229e-05, "loss": 0.0165, "step": 34780 }, { "grad_norm": 0.2434615045785904, "learning_rate": 7.76388122257484e-05, "loss": 0.0135, "step": 34790 }, { "grad_norm": 0.20108002424240112, "learning_rate": 7.762503185931558e-05, "loss": 0.018, "step": 34800 }, { "grad_norm": 0.21645452082157135, "learning_rate": 7.76112484718508e-05, "loss": 0.0143, "step": 34810 }, { "grad_norm": 0.16297321021556854, "learning_rate": 7.75974620648614e-05, "loss": 0.0127, "step": 34820 }, { "grad_norm": 0.2601620554924011, "learning_rate": 7.758367263985503e-05, "loss": 0.0163, "step": 34830 }, { "grad_norm": 0.1773247867822647, "learning_rate": 7.75698801983397e-05, "loss": 0.0115, "step": 34840 }, { "grad_norm": 0.6598048806190491, "learning_rate": 7.755608474182372e-05, "loss": 0.0139, "step": 34850 }, { "grad_norm": 0.21883662045001984, "learning_rate": 7.754228627181574e-05, "loss": 0.0159, "step": 34860 }, { "grad_norm": 0.17920978367328644, "learning_rate": 7.752848478982476e-05, "loss": 0.0161, "step": 34870 }, { "grad_norm": 0.21511033177375793, "learning_rate": 7.751468029736006e-05, "loss": 0.0176, "step": 34880 }, { "grad_norm": 0.38114655017852783, "learning_rate": 7.750087279593129e-05, "loss": 0.018, "step": 34890 }, { "grad_norm": 0.22108449041843414, "learning_rate": 7.748706228704843e-05, "loss": 0.0194, "step": 34900 }, { "grad_norm": 0.5060257315635681, "learning_rate": 7.747324877222176e-05, "loss": 0.0194, "step": 34910 }, { "grad_norm": 0.21269434690475464, "learning_rate": 7.745943225296188e-05, "loss": 0.0206, "step": 34920 }, { "grad_norm": 0.18430952727794647, "learning_rate": 7.744561273077981e-05, "loss": 0.0189, "step": 34930 }, { "grad_norm": 0.16985708475112915, "learning_rate": 7.743179020718678e-05, "loss": 0.014, "step": 34940 }, { "grad_norm": 0.20913857221603394, "learning_rate": 7.741796468369443e-05, "loss": 0.0155, "step": 34950 }, { "grad_norm": 0.17756158113479614, "learning_rate": 7.740413616181466e-05, "loss": 0.0182, "step": 34960 }, { "grad_norm": 0.17767798900604248, "learning_rate": 7.739030464305978e-05, "loss": 0.0121, "step": 34970 }, { "grad_norm": 0.16449391841888428, "learning_rate": 7.737647012894235e-05, "loss": 0.0184, "step": 34980 }, { "grad_norm": 0.20625242590904236, "learning_rate": 7.736263262097532e-05, "loss": 0.0111, "step": 34990 }, { "grad_norm": 0.12702372670173645, "learning_rate": 7.734879212067192e-05, "loss": 0.0131, "step": 35000 }, { "grad_norm": 0.2036971002817154, "learning_rate": 7.733494862954573e-05, "loss": 0.011, "step": 35010 }, { "grad_norm": 0.1647939383983612, "learning_rate": 7.732110214911066e-05, "loss": 0.0111, "step": 35020 }, { "grad_norm": 0.19639092683792114, "learning_rate": 7.730725268088094e-05, "loss": 0.0138, "step": 35030 }, { "grad_norm": 0.20925834774971008, "learning_rate": 7.729340022637112e-05, "loss": 0.0179, "step": 35040 }, { "grad_norm": 0.24478290975093842, "learning_rate": 7.727954478709607e-05, "loss": 0.0167, "step": 35050 }, { "grad_norm": 0.17600172758102417, "learning_rate": 7.726568636457103e-05, "loss": 0.018, "step": 35060 }, { "grad_norm": 0.19959942996501923, "learning_rate": 7.725182496031153e-05, "loss": 0.0217, "step": 35070 }, { "grad_norm": 0.38489505648612976, "learning_rate": 7.723796057583341e-05, "loss": 0.0198, "step": 35080 }, { "grad_norm": 0.290696918964386, "learning_rate": 7.722409321265287e-05, "loss": 0.0142, "step": 35090 }, { "grad_norm": 0.2071215808391571, "learning_rate": 7.721022287228645e-05, "loss": 0.014, "step": 35100 }, { "grad_norm": 0.14033816754817963, "learning_rate": 7.719634955625094e-05, "loss": 0.0206, "step": 35110 }, { "grad_norm": 0.1349952667951584, "learning_rate": 7.718247326606355e-05, "loss": 0.0145, "step": 35120 }, { "grad_norm": 0.18787169456481934, "learning_rate": 7.716859400324175e-05, "loss": 0.0166, "step": 35130 }, { "grad_norm": 0.13826146721839905, "learning_rate": 7.715471176930335e-05, "loss": 0.0135, "step": 35140 }, { "grad_norm": 0.17301510274410248, "learning_rate": 7.714082656576651e-05, "loss": 0.0193, "step": 35150 }, { "grad_norm": 0.2474849671125412, "learning_rate": 7.712693839414968e-05, "loss": 0.0146, "step": 35160 }, { "grad_norm": 0.3150117099285126, "learning_rate": 7.711304725597164e-05, "loss": 0.0155, "step": 35170 }, { "grad_norm": 0.2744595408439636, "learning_rate": 7.709915315275151e-05, "loss": 0.0131, "step": 35180 }, { "grad_norm": 0.21809105575084686, "learning_rate": 7.708525608600876e-05, "loss": 0.0185, "step": 35190 }, { "grad_norm": 0.24881349503993988, "learning_rate": 7.707135605726311e-05, "loss": 0.0191, "step": 35200 }, { "grad_norm": 0.22033949196338654, "learning_rate": 7.705745306803466e-05, "loss": 0.0129, "step": 35210 }, { "grad_norm": 0.3075407147407532, "learning_rate": 7.704354711984383e-05, "loss": 0.0167, "step": 35220 }, { "grad_norm": 0.2238188534975052, "learning_rate": 7.702963821421133e-05, "loss": 0.0138, "step": 35230 }, { "grad_norm": 0.2449130415916443, "learning_rate": 7.701572635265826e-05, "loss": 0.0132, "step": 35240 }, { "grad_norm": 0.27397534251213074, "learning_rate": 7.700181153670596e-05, "loss": 0.0155, "step": 35250 }, { "grad_norm": 0.17159008979797363, "learning_rate": 7.698789376787616e-05, "loss": 0.0125, "step": 35260 }, { "grad_norm": 0.16108748316764832, "learning_rate": 7.697397304769085e-05, "loss": 0.0159, "step": 35270 }, { "grad_norm": 0.20510053634643555, "learning_rate": 7.696004937767241e-05, "loss": 0.0154, "step": 35280 }, { "grad_norm": 0.21739906072616577, "learning_rate": 7.694612275934352e-05, "loss": 0.017, "step": 35290 }, { "grad_norm": 0.201621413230896, "learning_rate": 7.693219319422714e-05, "loss": 0.0152, "step": 35300 }, { "grad_norm": 0.28878912329673767, "learning_rate": 7.69182606838466e-05, "loss": 0.0156, "step": 35310 }, { "grad_norm": 0.2232048213481903, "learning_rate": 7.690432522972558e-05, "loss": 0.0164, "step": 35320 }, { "grad_norm": 0.12848345935344696, "learning_rate": 7.689038683338796e-05, "loss": 0.0156, "step": 35330 }, { "grad_norm": 0.13306978344917297, "learning_rate": 7.687644549635808e-05, "loss": 0.0116, "step": 35340 }, { "grad_norm": 0.20725862681865692, "learning_rate": 7.686250122016053e-05, "loss": 0.0112, "step": 35350 }, { "grad_norm": 0.16249017417430878, "learning_rate": 7.684855400632022e-05, "loss": 0.0137, "step": 35360 }, { "grad_norm": 0.19455328583717346, "learning_rate": 7.683460385636243e-05, "loss": 0.0181, "step": 35370 }, { "grad_norm": 0.11465544253587723, "learning_rate": 7.68206507718127e-05, "loss": 0.0145, "step": 35380 }, { "grad_norm": 0.2377612441778183, "learning_rate": 7.680669475419693e-05, "loss": 0.015, "step": 35390 }, { "grad_norm": 0.17409268021583557, "learning_rate": 7.679273580504132e-05, "loss": 0.0201, "step": 35400 }, { "grad_norm": 0.2297186553478241, "learning_rate": 7.67787739258724e-05, "loss": 0.0123, "step": 35410 }, { "grad_norm": 0.24091790616512299, "learning_rate": 7.676480911821705e-05, "loss": 0.0148, "step": 35420 }, { "grad_norm": 1.6660518646240234, "learning_rate": 7.675084138360238e-05, "loss": 0.0135, "step": 35430 }, { "grad_norm": 0.287748247385025, "learning_rate": 7.673687072355592e-05, "loss": 0.0147, "step": 35440 }, { "grad_norm": 0.2347988337278366, "learning_rate": 7.67228971396055e-05, "loss": 0.0148, "step": 35450 }, { "grad_norm": 0.31645166873931885, "learning_rate": 7.670892063327922e-05, "loss": 0.015, "step": 35460 }, { "grad_norm": 0.21443046629428864, "learning_rate": 7.669494120610552e-05, "loss": 0.0174, "step": 35470 }, { "grad_norm": 0.2831992208957672, "learning_rate": 7.66809588596132e-05, "loss": 0.0171, "step": 35480 }, { "grad_norm": 0.17991897463798523, "learning_rate": 7.666697359533132e-05, "loss": 0.0149, "step": 35490 }, { "grad_norm": 0.381980836391449, "learning_rate": 7.665298541478932e-05, "loss": 0.0164, "step": 35500 }, { "grad_norm": 0.23690862953662872, "learning_rate": 7.66389943195169e-05, "loss": 0.0164, "step": 35510 }, { "grad_norm": 0.3149346709251404, "learning_rate": 7.66250003110441e-05, "loss": 0.0166, "step": 35520 }, { "grad_norm": 0.28655776381492615, "learning_rate": 7.66110033909013e-05, "loss": 0.0187, "step": 35530 }, { "grad_norm": 0.2516115605831146, "learning_rate": 7.659700356061918e-05, "loss": 0.0173, "step": 35540 }, { "grad_norm": 0.18315090239048004, "learning_rate": 7.658300082172875e-05, "loss": 0.0164, "step": 35550 }, { "grad_norm": 0.2644847631454468, "learning_rate": 7.65689951757613e-05, "loss": 0.0137, "step": 35560 }, { "grad_norm": 0.1873132884502411, "learning_rate": 7.655498662424849e-05, "loss": 0.0124, "step": 35570 }, { "grad_norm": 0.20150326192378998, "learning_rate": 7.654097516872227e-05, "loss": 0.0163, "step": 35580 }, { "grad_norm": 0.2136354297399521, "learning_rate": 7.65269608107149e-05, "loss": 0.0137, "step": 35590 }, { "grad_norm": 0.2258763462305069, "learning_rate": 7.6512943551759e-05, "loss": 0.0148, "step": 35600 }, { "grad_norm": 0.2148067057132721, "learning_rate": 7.649892339338743e-05, "loss": 0.0198, "step": 35610 }, { "grad_norm": 0.17432500422000885, "learning_rate": 7.648490033713344e-05, "loss": 0.017, "step": 35620 }, { "grad_norm": 0.16675768792629242, "learning_rate": 7.647087438453058e-05, "loss": 0.0131, "step": 35630 }, { "grad_norm": 0.20917582511901855, "learning_rate": 7.64568455371127e-05, "loss": 0.0127, "step": 35640 }, { "grad_norm": 0.28380465507507324, "learning_rate": 7.644281379641396e-05, "loss": 0.0159, "step": 35650 }, { "grad_norm": 0.15073880553245544, "learning_rate": 7.642877916396887e-05, "loss": 0.0141, "step": 35660 }, { "grad_norm": 0.21565425395965576, "learning_rate": 7.641474164131221e-05, "loss": 0.0124, "step": 35670 }, { "grad_norm": 0.1755422204732895, "learning_rate": 7.640070122997913e-05, "loss": 0.0121, "step": 35680 }, { "grad_norm": 0.7366175651550293, "learning_rate": 7.638665793150505e-05, "loss": 0.0134, "step": 35690 }, { "grad_norm": 0.23708203434944153, "learning_rate": 7.637261174742574e-05, "loss": 0.0136, "step": 35700 }, { "grad_norm": 1.1451170444488525, "learning_rate": 7.635856267927725e-05, "loss": 0.0166, "step": 35710 }, { "grad_norm": 0.5997752547264099, "learning_rate": 7.634451072859597e-05, "loss": 0.0156, "step": 35720 }, { "grad_norm": 0.1968393623828888, "learning_rate": 7.633045589691863e-05, "loss": 0.0137, "step": 35730 }, { "grad_norm": 0.1672414392232895, "learning_rate": 7.63163981857822e-05, "loss": 0.0146, "step": 35740 }, { "grad_norm": 0.1764969676733017, "learning_rate": 7.630233759672403e-05, "loss": 0.0145, "step": 35750 }, { "grad_norm": 0.1397438645362854, "learning_rate": 7.628827413128177e-05, "loss": 0.0132, "step": 35760 }, { "grad_norm": 0.12774911522865295, "learning_rate": 7.627420779099338e-05, "loss": 0.0125, "step": 35770 }, { "grad_norm": 0.1534450352191925, "learning_rate": 7.626013857739711e-05, "loss": 0.0109, "step": 35780 }, { "grad_norm": 0.2133760303258896, "learning_rate": 7.624606649203158e-05, "loss": 0.0105, "step": 35790 }, { "grad_norm": 0.5734463930130005, "learning_rate": 7.623199153643569e-05, "loss": 0.0142, "step": 35800 }, { "grad_norm": 0.1654619574546814, "learning_rate": 7.621791371214863e-05, "loss": 0.013, "step": 35810 }, { "grad_norm": 0.11755799502134323, "learning_rate": 7.620383302070995e-05, "loss": 0.0107, "step": 35820 }, { "grad_norm": 0.15532010793685913, "learning_rate": 7.61897494636595e-05, "loss": 0.0125, "step": 35830 }, { "grad_norm": 0.21597686409950256, "learning_rate": 7.617566304253739e-05, "loss": 0.0113, "step": 35840 }, { "grad_norm": 0.1632712334394455, "learning_rate": 7.616157375888416e-05, "loss": 0.0116, "step": 35850 }, { "grad_norm": 0.13338914513587952, "learning_rate": 7.614748161424053e-05, "loss": 0.0114, "step": 35860 }, { "grad_norm": 0.2055957317352295, "learning_rate": 7.613338661014763e-05, "loss": 0.0141, "step": 35870 }, { "grad_norm": 0.17925402522087097, "learning_rate": 7.611928874814686e-05, "loss": 0.0126, "step": 35880 }, { "grad_norm": 0.1871696412563324, "learning_rate": 7.610518802977996e-05, "loss": 0.0163, "step": 35890 }, { "grad_norm": 0.2007332146167755, "learning_rate": 7.609108445658893e-05, "loss": 0.0125, "step": 35900 }, { "grad_norm": 0.20962592959403992, "learning_rate": 7.607697803011612e-05, "loss": 0.0191, "step": 35910 }, { "grad_norm": 0.4684516489505768, "learning_rate": 7.606286875190421e-05, "loss": 0.0138, "step": 35920 }, { "grad_norm": 0.15597987174987793, "learning_rate": 7.604875662349617e-05, "loss": 0.0115, "step": 35930 }, { "grad_norm": 0.1804436594247818, "learning_rate": 7.603464164643525e-05, "loss": 0.01, "step": 35940 }, { "grad_norm": 0.14506685733795166, "learning_rate": 7.602052382226507e-05, "loss": 0.0106, "step": 35950 }, { "grad_norm": 0.13644982874393463, "learning_rate": 7.600640315252954e-05, "loss": 0.0103, "step": 35960 }, { "grad_norm": 0.20544464886188507, "learning_rate": 7.599227963877284e-05, "loss": 0.0106, "step": 35970 }, { "grad_norm": 0.1622670292854309, "learning_rate": 7.597815328253954e-05, "loss": 0.0141, "step": 35980 }, { "grad_norm": 0.1052095890045166, "learning_rate": 7.596402408537443e-05, "loss": 0.0111, "step": 35990 }, { "grad_norm": 0.15196824073791504, "learning_rate": 7.59498920488227e-05, "loss": 0.0175, "step": 36000 }, { "grad_norm": 0.15474893152713776, "learning_rate": 7.593575717442979e-05, "loss": 0.0106, "step": 36010 }, { "grad_norm": 0.14508061110973358, "learning_rate": 7.592161946374147e-05, "loss": 0.0122, "step": 36020 }, { "grad_norm": 0.21698714792728424, "learning_rate": 7.590747891830381e-05, "loss": 0.0188, "step": 36030 }, { "grad_norm": 0.20474295318126678, "learning_rate": 7.58933355396632e-05, "loss": 0.011, "step": 36040 }, { "grad_norm": 0.15310834348201752, "learning_rate": 7.587918932936636e-05, "loss": 0.0142, "step": 36050 }, { "grad_norm": 0.22179001569747925, "learning_rate": 7.586504028896028e-05, "loss": 0.0148, "step": 36060 }, { "grad_norm": 0.15422070026397705, "learning_rate": 7.585088841999228e-05, "loss": 0.0151, "step": 36070 }, { "grad_norm": 0.23115523159503937, "learning_rate": 7.583673372400999e-05, "loss": 0.0191, "step": 36080 }, { "grad_norm": 0.1591062992811203, "learning_rate": 7.582257620256134e-05, "loss": 0.0149, "step": 36090 }, { "grad_norm": 0.1630774438381195, "learning_rate": 7.580841585719458e-05, "loss": 0.0116, "step": 36100 }, { "grad_norm": 0.22245685756206512, "learning_rate": 7.579425268945825e-05, "loss": 0.0136, "step": 36110 }, { "grad_norm": 0.20755542814731598, "learning_rate": 7.578008670090127e-05, "loss": 0.0133, "step": 36120 }, { "grad_norm": 0.19762887060642242, "learning_rate": 7.576591789307272e-05, "loss": 0.0195, "step": 36130 }, { "grad_norm": 0.8163414001464844, "learning_rate": 7.575174626752216e-05, "loss": 0.015, "step": 36140 }, { "grad_norm": 0.17620103061199188, "learning_rate": 7.573757182579934e-05, "loss": 0.0195, "step": 36150 }, { "grad_norm": 0.2306799590587616, "learning_rate": 7.572339456945435e-05, "loss": 0.0149, "step": 36160 }, { "grad_norm": 0.18109607696533203, "learning_rate": 7.570921450003762e-05, "loss": 0.0152, "step": 36170 }, { "grad_norm": 0.19682598114013672, "learning_rate": 7.569503161909983e-05, "loss": 0.0129, "step": 36180 }, { "grad_norm": 0.16118818521499634, "learning_rate": 7.568084592819202e-05, "loss": 0.013, "step": 36190 }, { "grad_norm": 0.22236187756061554, "learning_rate": 7.566665742886551e-05, "loss": 0.0173, "step": 36200 }, { "grad_norm": 0.41741806268692017, "learning_rate": 7.565246612267194e-05, "loss": 0.0124, "step": 36210 }, { "grad_norm": 0.24017734825611115, "learning_rate": 7.563827201116325e-05, "loss": 0.0133, "step": 36220 }, { "grad_norm": 0.1949363797903061, "learning_rate": 7.562407509589164e-05, "loss": 0.0139, "step": 36230 }, { "grad_norm": 0.20916052162647247, "learning_rate": 7.560987537840973e-05, "loss": 0.014, "step": 36240 }, { "grad_norm": 0.1308300495147705, "learning_rate": 7.559567286027036e-05, "loss": 0.0085, "step": 36250 }, { "grad_norm": 0.19994954764842987, "learning_rate": 7.558146754302668e-05, "loss": 0.0165, "step": 36260 }, { "grad_norm": 0.1996767669916153, "learning_rate": 7.556725942823217e-05, "loss": 0.0157, "step": 36270 }, { "grad_norm": 0.1975906640291214, "learning_rate": 7.555304851744061e-05, "loss": 0.0162, "step": 36280 }, { "grad_norm": 0.18548579514026642, "learning_rate": 7.553883481220608e-05, "loss": 0.0108, "step": 36290 }, { "grad_norm": 0.15384981036186218, "learning_rate": 7.552461831408298e-05, "loss": 0.0103, "step": 36300 }, { "grad_norm": 0.13988924026489258, "learning_rate": 7.551039902462599e-05, "loss": 0.0121, "step": 36310 }, { "grad_norm": 0.21561188995838165, "learning_rate": 7.549617694539011e-05, "loss": 0.0161, "step": 36320 }, { "grad_norm": 0.21402856707572937, "learning_rate": 7.548195207793066e-05, "loss": 0.0158, "step": 36330 }, { "grad_norm": 0.31121131777763367, "learning_rate": 7.546772442380323e-05, "loss": 0.0155, "step": 36340 }, { "grad_norm": 0.22836560010910034, "learning_rate": 7.545349398456376e-05, "loss": 0.0127, "step": 36350 }, { "grad_norm": 0.20202170312404633, "learning_rate": 7.543926076176845e-05, "loss": 0.0144, "step": 36360 }, { "grad_norm": 0.16368220746517181, "learning_rate": 7.542502475697385e-05, "loss": 0.0132, "step": 36370 }, { "grad_norm": 0.19764344394207, "learning_rate": 7.541078597173675e-05, "loss": 0.0127, "step": 36380 }, { "grad_norm": 0.21922507882118225, "learning_rate": 7.539654440761431e-05, "loss": 0.0136, "step": 36390 }, { "grad_norm": 0.18856991827487946, "learning_rate": 7.538230006616395e-05, "loss": 0.0148, "step": 36400 }, { "grad_norm": 0.40421706438064575, "learning_rate": 7.536805294894342e-05, "loss": 0.0167, "step": 36410 }, { "grad_norm": 0.2181703895330429, "learning_rate": 7.535380305751076e-05, "loss": 0.0144, "step": 36420 }, { "grad_norm": 0.2172774076461792, "learning_rate": 7.533955039342431e-05, "loss": 0.0194, "step": 36430 }, { "grad_norm": 0.20124267041683197, "learning_rate": 7.532529495824274e-05, "loss": 0.0124, "step": 36440 }, { "grad_norm": 0.17873577773571014, "learning_rate": 7.531103675352497e-05, "loss": 0.0142, "step": 36450 }, { "grad_norm": 0.154933899641037, "learning_rate": 7.529677578083029e-05, "loss": 0.0137, "step": 36460 }, { "grad_norm": 0.22946198284626007, "learning_rate": 7.528251204171823e-05, "loss": 0.0166, "step": 36470 }, { "grad_norm": 0.17934226989746094, "learning_rate": 7.526824553774867e-05, "loss": 0.0154, "step": 36480 }, { "grad_norm": 0.355411559343338, "learning_rate": 7.525397627048178e-05, "loss": 0.0162, "step": 36490 }, { "grad_norm": 0.19693516194820404, "learning_rate": 7.523970424147802e-05, "loss": 0.0163, "step": 36500 }, { "grad_norm": 0.24174799025058746, "learning_rate": 7.522542945229813e-05, "loss": 0.0213, "step": 36510 }, { "grad_norm": 0.21241603791713715, "learning_rate": 7.521115190450321e-05, "loss": 0.017, "step": 36520 }, { "grad_norm": 0.1838991343975067, "learning_rate": 7.519687159965462e-05, "loss": 0.0138, "step": 36530 }, { "grad_norm": 0.14705047011375427, "learning_rate": 7.518258853931403e-05, "loss": 0.0162, "step": 36540 }, { "grad_norm": 0.1783735752105713, "learning_rate": 7.516830272504342e-05, "loss": 0.0161, "step": 36550 }, { "grad_norm": 0.5429854989051819, "learning_rate": 7.515401415840509e-05, "loss": 0.014, "step": 36560 }, { "grad_norm": 0.3218836486339569, "learning_rate": 7.513972284096155e-05, "loss": 0.0115, "step": 36570 }, { "grad_norm": 0.15145520865917206, "learning_rate": 7.512542877427574e-05, "loss": 0.013, "step": 36580 }, { "grad_norm": 0.24491740763187408, "learning_rate": 7.51111319599108e-05, "loss": 0.016, "step": 36590 }, { "grad_norm": 0.20912446081638336, "learning_rate": 7.509683239943024e-05, "loss": 0.0122, "step": 36600 }, { "grad_norm": 0.16172091662883759, "learning_rate": 7.50825300943978e-05, "loss": 0.0098, "step": 36610 }, { "grad_norm": 0.3744473457336426, "learning_rate": 7.50682250463776e-05, "loss": 0.0156, "step": 36620 }, { "grad_norm": 0.2027370184659958, "learning_rate": 7.5053917256934e-05, "loss": 0.0112, "step": 36630 }, { "grad_norm": 0.1706119328737259, "learning_rate": 7.503960672763166e-05, "loss": 0.0132, "step": 36640 }, { "grad_norm": 0.2083195000886917, "learning_rate": 7.502529346003559e-05, "loss": 0.013, "step": 36650 }, { "grad_norm": 0.1802518516778946, "learning_rate": 7.501097745571106e-05, "loss": 0.015, "step": 36660 }, { "grad_norm": 0.1384134143590927, "learning_rate": 7.499665871622361e-05, "loss": 0.01, "step": 36670 }, { "grad_norm": 0.1519429087638855, "learning_rate": 7.498233724313917e-05, "loss": 0.0189, "step": 36680 }, { "grad_norm": 0.13757580518722534, "learning_rate": 7.496801303802387e-05, "loss": 0.0128, "step": 36690 }, { "grad_norm": 0.1492311805486679, "learning_rate": 7.495368610244423e-05, "loss": 0.0165, "step": 36700 }, { "grad_norm": 0.15557745099067688, "learning_rate": 7.493935643796697e-05, "loss": 0.0167, "step": 36710 }, { "grad_norm": 0.1305299699306488, "learning_rate": 7.49250240461592e-05, "loss": 0.0093, "step": 36720 }, { "grad_norm": 0.17843115329742432, "learning_rate": 7.491068892858826e-05, "loss": 0.016, "step": 36730 }, { "grad_norm": 0.10773951560258865, "learning_rate": 7.489635108682184e-05, "loss": 0.0098, "step": 36740 }, { "grad_norm": 0.19628311693668365, "learning_rate": 7.48820105224279e-05, "loss": 0.0163, "step": 36750 }, { "grad_norm": 0.6277165412902832, "learning_rate": 7.48676672369747e-05, "loss": 0.0145, "step": 36760 }, { "grad_norm": 0.17831717431545258, "learning_rate": 7.485332123203079e-05, "loss": 0.0087, "step": 36770 }, { "grad_norm": 0.13793550431728363, "learning_rate": 7.483897250916504e-05, "loss": 0.0132, "step": 36780 }, { "grad_norm": 0.17562903463840485, "learning_rate": 7.48246210699466e-05, "loss": 0.0166, "step": 36790 }, { "grad_norm": 0.29713496565818787, "learning_rate": 7.481026691594492e-05, "loss": 0.0133, "step": 36800 }, { "grad_norm": 1.1823060512542725, "learning_rate": 7.479591004872975e-05, "loss": 0.0181, "step": 36810 }, { "grad_norm": 0.25950148701667786, "learning_rate": 7.478155046987114e-05, "loss": 0.0178, "step": 36820 }, { "grad_norm": 0.2742452621459961, "learning_rate": 7.476718818093944e-05, "loss": 0.0124, "step": 36830 }, { "grad_norm": 0.25404685735702515, "learning_rate": 7.475282318350527e-05, "loss": 0.0155, "step": 36840 }, { "grad_norm": 0.283310204744339, "learning_rate": 7.473845547913957e-05, "loss": 0.0161, "step": 36850 }, { "grad_norm": 0.25743478536605835, "learning_rate": 7.472408506941357e-05, "loss": 0.0194, "step": 36860 }, { "grad_norm": 0.2740119993686676, "learning_rate": 7.470971195589881e-05, "loss": 0.014, "step": 36870 }, { "grad_norm": 0.2313365638256073, "learning_rate": 7.46953361401671e-05, "loss": 0.0135, "step": 36880 }, { "grad_norm": 0.21266715228557587, "learning_rate": 7.468095762379055e-05, "loss": 0.0177, "step": 36890 }, { "grad_norm": 0.20909801125526428, "learning_rate": 7.466657640834158e-05, "loss": 0.0122, "step": 36900 }, { "grad_norm": 0.18549853563308716, "learning_rate": 7.46521924953929e-05, "loss": 0.0122, "step": 36910 }, { "grad_norm": 0.293023943901062, "learning_rate": 7.463780588651752e-05, "loss": 0.0132, "step": 36920 }, { "grad_norm": 0.18460430204868317, "learning_rate": 7.462341658328873e-05, "loss": 0.0187, "step": 36930 }, { "grad_norm": 0.19928863644599915, "learning_rate": 7.460902458728012e-05, "loss": 0.0097, "step": 36940 }, { "grad_norm": 0.21762607991695404, "learning_rate": 7.459462990006558e-05, "loss": 0.0122, "step": 36950 }, { "grad_norm": 0.19736160337924957, "learning_rate": 7.458023252321929e-05, "loss": 0.0121, "step": 36960 }, { "grad_norm": 0.3058517575263977, "learning_rate": 7.456583245831573e-05, "loss": 0.015, "step": 36970 }, { "grad_norm": 0.17902906239032745, "learning_rate": 7.455142970692968e-05, "loss": 0.0177, "step": 36980 }, { "grad_norm": 0.5196970701217651, "learning_rate": 7.453702427063615e-05, "loss": 0.0176, "step": 36990 }, { "grad_norm": 0.1318819671869278, "learning_rate": 7.452261615101057e-05, "loss": 0.012, "step": 37000 }, { "grad_norm": 0.13360996544361115, "learning_rate": 7.450820534962856e-05, "loss": 0.0134, "step": 37010 }, { "grad_norm": 0.32330387830734253, "learning_rate": 7.449379186806604e-05, "loss": 0.0115, "step": 37020 }, { "grad_norm": 0.8281594514846802, "learning_rate": 7.44793757078993e-05, "loss": 0.0141, "step": 37030 }, { "grad_norm": 0.19950586557388306, "learning_rate": 7.44649568707048e-05, "loss": 0.0161, "step": 37040 }, { "grad_norm": 0.30215364694595337, "learning_rate": 7.445053535805942e-05, "loss": 0.0144, "step": 37050 }, { "grad_norm": 0.32432812452316284, "learning_rate": 7.443611117154024e-05, "loss": 0.023, "step": 37060 }, { "grad_norm": 0.20921072363853455, "learning_rate": 7.442168431272469e-05, "loss": 0.0134, "step": 37070 }, { "grad_norm": 0.19847238063812256, "learning_rate": 7.440725478319044e-05, "loss": 0.0136, "step": 37080 }, { "grad_norm": 0.6145092248916626, "learning_rate": 7.439282258451553e-05, "loss": 0.0156, "step": 37090 }, { "grad_norm": 0.18606150150299072, "learning_rate": 7.43783877182782e-05, "loss": 0.0131, "step": 37100 }, { "grad_norm": 0.321551114320755, "learning_rate": 7.436395018605707e-05, "loss": 0.0214, "step": 37110 }, { "grad_norm": 0.1574763059616089, "learning_rate": 7.434950998943094e-05, "loss": 0.0128, "step": 37120 }, { "grad_norm": 0.1523878127336502, "learning_rate": 7.433506712997902e-05, "loss": 0.0148, "step": 37130 }, { "grad_norm": 0.36829251050949097, "learning_rate": 7.432062160928077e-05, "loss": 0.0164, "step": 37140 }, { "grad_norm": 0.18524198234081268, "learning_rate": 7.430617342891588e-05, "loss": 0.0136, "step": 37150 }, { "grad_norm": 0.1749986708164215, "learning_rate": 7.429172259046442e-05, "loss": 0.0117, "step": 37160 }, { "grad_norm": 0.21089819073677063, "learning_rate": 7.42772690955067e-05, "loss": 0.0163, "step": 37170 }, { "grad_norm": 0.20497873425483704, "learning_rate": 7.426281294562332e-05, "loss": 0.0157, "step": 37180 }, { "grad_norm": 0.18213658034801483, "learning_rate": 7.424835414239519e-05, "loss": 0.0129, "step": 37190 }, { "grad_norm": 0.1609335094690323, "learning_rate": 7.423389268740352e-05, "loss": 0.0111, "step": 37200 }, { "grad_norm": 2.360750198364258, "learning_rate": 7.42194285822298e-05, "loss": 0.018, "step": 37210 }, { "grad_norm": 2.4654645919799805, "learning_rate": 7.420496182845576e-05, "loss": 0.0128, "step": 37220 }, { "grad_norm": 0.19496780633926392, "learning_rate": 7.419049242766348e-05, "loss": 0.012, "step": 37230 }, { "grad_norm": 0.22453638911247253, "learning_rate": 7.417602038143533e-05, "loss": 0.0111, "step": 37240 }, { "grad_norm": 0.1953367292881012, "learning_rate": 7.416154569135393e-05, "loss": 0.0131, "step": 37250 }, { "grad_norm": 0.20630647242069244, "learning_rate": 7.414706835900222e-05, "loss": 0.0125, "step": 37260 }, { "grad_norm": 0.16298116743564606, "learning_rate": 7.413258838596345e-05, "loss": 0.0121, "step": 37270 }, { "grad_norm": 0.14625169336795807, "learning_rate": 7.411810577382106e-05, "loss": 0.0141, "step": 37280 }, { "grad_norm": 0.18399274349212646, "learning_rate": 7.410362052415889e-05, "loss": 0.0092, "step": 37290 }, { "grad_norm": 0.34104207158088684, "learning_rate": 7.408913263856102e-05, "loss": 0.012, "step": 37300 }, { "grad_norm": 0.145651176571846, "learning_rate": 7.407464211861183e-05, "loss": 0.0159, "step": 37310 }, { "grad_norm": 0.2782619893550873, "learning_rate": 7.406014896589597e-05, "loss": 0.0109, "step": 37320 }, { "grad_norm": 0.22878910601139069, "learning_rate": 7.404565318199838e-05, "loss": 0.0099, "step": 37330 }, { "grad_norm": 0.20374581217765808, "learning_rate": 7.403115476850432e-05, "loss": 0.0138, "step": 37340 }, { "grad_norm": 0.3387020230293274, "learning_rate": 7.401665372699932e-05, "loss": 0.0131, "step": 37350 }, { "grad_norm": 0.15643367171287537, "learning_rate": 7.400215005906916e-05, "loss": 0.014, "step": 37360 }, { "grad_norm": 0.17817836999893188, "learning_rate": 7.398764376629996e-05, "loss": 0.0101, "step": 37370 }, { "grad_norm": 0.24269597232341766, "learning_rate": 7.39731348502781e-05, "loss": 0.0155, "step": 37380 }, { "grad_norm": 0.14999791979789734, "learning_rate": 7.395862331259028e-05, "loss": 0.0136, "step": 37390 }, { "grad_norm": 0.16184599697589874, "learning_rate": 7.394410915482341e-05, "loss": 0.0116, "step": 37400 }, { "grad_norm": 0.19166862964630127, "learning_rate": 7.392959237856476e-05, "loss": 0.0156, "step": 37410 }, { "grad_norm": 0.18082033097743988, "learning_rate": 7.391507298540187e-05, "loss": 0.0161, "step": 37420 }, { "grad_norm": 0.21484805643558502, "learning_rate": 7.390055097692257e-05, "loss": 0.0163, "step": 37430 }, { "grad_norm": 0.9991798400878906, "learning_rate": 7.388602635471494e-05, "loss": 0.0159, "step": 37440 }, { "grad_norm": 0.19711032509803772, "learning_rate": 7.387149912036738e-05, "loss": 0.0156, "step": 37450 }, { "grad_norm": 0.17150244116783142, "learning_rate": 7.385696927546854e-05, "loss": 0.0134, "step": 37460 }, { "grad_norm": 0.211372509598732, "learning_rate": 7.384243682160743e-05, "loss": 0.0165, "step": 37470 }, { "grad_norm": 0.21242110431194305, "learning_rate": 7.382790176037329e-05, "loss": 0.0107, "step": 37480 }, { "grad_norm": 0.15038520097732544, "learning_rate": 7.381336409335562e-05, "loss": 0.0154, "step": 37490 }, { "grad_norm": 0.17696407437324524, "learning_rate": 7.379882382214426e-05, "loss": 0.0164, "step": 37500 }, { "grad_norm": 0.12733188271522522, "learning_rate": 7.378428094832931e-05, "loss": 0.0096, "step": 37510 }, { "grad_norm": 0.17063279449939728, "learning_rate": 7.376973547350114e-05, "loss": 0.0107, "step": 37520 }, { "grad_norm": 0.20232105255126953, "learning_rate": 7.375518739925046e-05, "loss": 0.0165, "step": 37530 }, { "grad_norm": 0.1707378774881363, "learning_rate": 7.374063672716818e-05, "loss": 0.0125, "step": 37540 }, { "grad_norm": 0.1368248611688614, "learning_rate": 7.372608345884558e-05, "loss": 0.0121, "step": 37550 }, { "grad_norm": 0.1895175278186798, "learning_rate": 7.371152759587415e-05, "loss": 0.0115, "step": 37560 }, { "grad_norm": 0.18688499927520752, "learning_rate": 7.369696913984572e-05, "loss": 0.0141, "step": 37570 }, { "grad_norm": 0.3667033910751343, "learning_rate": 7.368240809235237e-05, "loss": 0.0118, "step": 37580 }, { "grad_norm": 0.23441003262996674, "learning_rate": 7.366784445498649e-05, "loss": 0.0156, "step": 37590 }, { "grad_norm": 0.26481345295906067, "learning_rate": 7.36532782293407e-05, "loss": 0.0163, "step": 37600 }, { "grad_norm": 0.3133242130279541, "learning_rate": 7.363870941700797e-05, "loss": 0.0174, "step": 37610 }, { "grad_norm": 0.1447845697402954, "learning_rate": 7.362413801958152e-05, "loss": 0.0127, "step": 37620 }, { "grad_norm": 0.21044287085533142, "learning_rate": 7.360956403865486e-05, "loss": 0.0131, "step": 37630 }, { "grad_norm": 0.3683547079563141, "learning_rate": 7.359498747582177e-05, "loss": 0.0199, "step": 37640 }, { "grad_norm": 0.3923424482345581, "learning_rate": 7.358040833267634e-05, "loss": 0.0154, "step": 37650 }, { "grad_norm": 0.24099257588386536, "learning_rate": 7.356582661081288e-05, "loss": 0.0208, "step": 37660 }, { "grad_norm": 0.5287302136421204, "learning_rate": 7.355124231182607e-05, "loss": 0.0141, "step": 37670 }, { "grad_norm": 0.19801880419254303, "learning_rate": 7.353665543731083e-05, "loss": 0.0146, "step": 37680 }, { "grad_norm": 0.2828977406024933, "learning_rate": 7.352206598886229e-05, "loss": 0.0153, "step": 37690 }, { "grad_norm": 0.24515609443187714, "learning_rate": 7.350747396807601e-05, "loss": 0.0165, "step": 37700 }, { "grad_norm": 0.2180965691804886, "learning_rate": 7.349287937654772e-05, "loss": 0.0165, "step": 37710 }, { "grad_norm": 0.21213854849338531, "learning_rate": 7.347828221587345e-05, "loss": 0.0155, "step": 37720 }, { "grad_norm": 0.24474821984767914, "learning_rate": 7.346368248764955e-05, "loss": 0.0131, "step": 37730 }, { "grad_norm": 0.16326627135276794, "learning_rate": 7.34490801934726e-05, "loss": 0.0135, "step": 37740 }, { "grad_norm": 0.18438541889190674, "learning_rate": 7.343447533493947e-05, "loss": 0.0132, "step": 37750 }, { "grad_norm": 0.2259000986814499, "learning_rate": 7.341986791364739e-05, "loss": 0.0165, "step": 37760 }, { "grad_norm": 0.17774759232997894, "learning_rate": 7.340525793119374e-05, "loss": 0.0154, "step": 37770 }, { "grad_norm": 0.20207327604293823, "learning_rate": 7.33906453891763e-05, "loss": 0.0117, "step": 37780 }, { "grad_norm": 0.22862422466278076, "learning_rate": 7.337603028919303e-05, "loss": 0.0153, "step": 37790 }, { "grad_norm": 0.26722440123558044, "learning_rate": 7.336141263284225e-05, "loss": 0.0158, "step": 37800 }, { "grad_norm": 0.201041117310524, "learning_rate": 7.334679242172249e-05, "loss": 0.0112, "step": 37810 }, { "grad_norm": 0.18582893908023834, "learning_rate": 7.33321696574326e-05, "loss": 0.0141, "step": 37820 }, { "grad_norm": 0.22530658543109894, "learning_rate": 7.331754434157174e-05, "loss": 0.019, "step": 37830 }, { "grad_norm": 0.19772638380527496, "learning_rate": 7.330291647573928e-05, "loss": 0.0148, "step": 37840 }, { "grad_norm": 0.18995992839336395, "learning_rate": 7.32882860615349e-05, "loss": 0.012, "step": 37850 }, { "grad_norm": 0.48413360118865967, "learning_rate": 7.327365310055858e-05, "loss": 0.0108, "step": 37860 }, { "grad_norm": 0.26702573895454407, "learning_rate": 7.325901759441055e-05, "loss": 0.0167, "step": 37870 }, { "grad_norm": 1.0351017713546753, "learning_rate": 7.324437954469132e-05, "loss": 0.0136, "step": 37880 }, { "grad_norm": 0.4049907326698303, "learning_rate": 7.32297389530017e-05, "loss": 0.0122, "step": 37890 }, { "grad_norm": 0.1272510290145874, "learning_rate": 7.321509582094274e-05, "loss": 0.0134, "step": 37900 }, { "grad_norm": 0.15222011506557465, "learning_rate": 7.320045015011582e-05, "loss": 0.0117, "step": 37910 }, { "grad_norm": 0.6529155373573303, "learning_rate": 7.318580194212253e-05, "loss": 0.0155, "step": 37920 }, { "grad_norm": 0.3067898750305176, "learning_rate": 7.317115119856483e-05, "loss": 0.0165, "step": 37930 }, { "grad_norm": 0.14864036440849304, "learning_rate": 7.315649792104485e-05, "loss": 0.0143, "step": 37940 }, { "grad_norm": 0.1672612875699997, "learning_rate": 7.314184211116507e-05, "loss": 0.0109, "step": 37950 }, { "grad_norm": 0.1861993819475174, "learning_rate": 7.312718377052823e-05, "loss": 0.014, "step": 37960 }, { "grad_norm": 0.15844102203845978, "learning_rate": 7.311252290073737e-05, "loss": 0.0114, "step": 37970 }, { "grad_norm": 0.22803178429603577, "learning_rate": 7.309785950339572e-05, "loss": 0.0117, "step": 37980 }, { "grad_norm": 0.23203347623348236, "learning_rate": 7.308319358010692e-05, "loss": 0.012, "step": 37990 }, { "grad_norm": 0.19305869936943054, "learning_rate": 7.306852513247474e-05, "loss": 0.0115, "step": 38000 }, { "grad_norm": 0.22097037732601166, "learning_rate": 7.305385416210336e-05, "loss": 0.0178, "step": 38010 }, { "grad_norm": 0.28856486082077026, "learning_rate": 7.303918067059711e-05, "loss": 0.0114, "step": 38020 }, { "grad_norm": 0.17868445813655853, "learning_rate": 7.302450465956075e-05, "loss": 0.0146, "step": 38030 }, { "grad_norm": 0.1929292380809784, "learning_rate": 7.300982613059914e-05, "loss": 0.0126, "step": 38040 }, { "grad_norm": 0.14298094809055328, "learning_rate": 7.299514508531757e-05, "loss": 0.0103, "step": 38050 }, { "grad_norm": 0.14197984337806702, "learning_rate": 7.298046152532148e-05, "loss": 0.0156, "step": 38060 }, { "grad_norm": 0.09653355181217194, "learning_rate": 7.296577545221668e-05, "loss": 0.0128, "step": 38070 }, { "grad_norm": 0.2149621695280075, "learning_rate": 7.29510868676092e-05, "loss": 0.0132, "step": 38080 }, { "grad_norm": 0.252555251121521, "learning_rate": 7.293639577310538e-05, "loss": 0.0144, "step": 38090 }, { "grad_norm": 0.23343652486801147, "learning_rate": 7.292170217031179e-05, "loss": 0.0133, "step": 38100 }, { "grad_norm": 0.14214925467967987, "learning_rate": 7.290700606083532e-05, "loss": 0.0156, "step": 38110 }, { "grad_norm": 0.21279840171337128, "learning_rate": 7.28923074462831e-05, "loss": 0.0139, "step": 38120 }, { "grad_norm": 0.4178216755390167, "learning_rate": 7.287760632826255e-05, "loss": 0.0165, "step": 38130 }, { "grad_norm": 0.25542891025543213, "learning_rate": 7.286290270838139e-05, "loss": 0.0122, "step": 38140 }, { "grad_norm": 0.2132958471775055, "learning_rate": 7.284819658824756e-05, "loss": 0.0174, "step": 38150 }, { "grad_norm": 0.17747029662132263, "learning_rate": 7.283348796946929e-05, "loss": 0.0106, "step": 38160 }, { "grad_norm": 0.1912110298871994, "learning_rate": 7.28187768536551e-05, "loss": 0.0137, "step": 38170 }, { "grad_norm": 0.24121159315109253, "learning_rate": 7.280406324241379e-05, "loss": 0.0118, "step": 38180 }, { "grad_norm": 0.2617289125919342, "learning_rate": 7.278934713735441e-05, "loss": 0.0145, "step": 38190 }, { "grad_norm": 0.26878026127815247, "learning_rate": 7.277462854008629e-05, "loss": 0.0178, "step": 38200 }, { "grad_norm": 0.21157042682170868, "learning_rate": 7.275990745221904e-05, "loss": 0.0174, "step": 38210 }, { "grad_norm": 0.19547943770885468, "learning_rate": 7.274518387536252e-05, "loss": 0.0156, "step": 38220 }, { "grad_norm": 0.16195477545261383, "learning_rate": 7.273045781112689e-05, "loss": 0.0134, "step": 38230 }, { "grad_norm": 0.1696745604276657, "learning_rate": 7.271572926112257e-05, "loss": 0.0125, "step": 38240 }, { "grad_norm": 0.2971963882446289, "learning_rate": 7.270099822696024e-05, "loss": 0.0116, "step": 38250 }, { "grad_norm": 0.18408958613872528, "learning_rate": 7.26862647102509e-05, "loss": 0.0144, "step": 38260 }, { "grad_norm": 0.1792343705892563, "learning_rate": 7.267152871260573e-05, "loss": 0.0125, "step": 38270 }, { "grad_norm": 0.1790303736925125, "learning_rate": 7.265679023563629e-05, "loss": 0.0166, "step": 38280 }, { "grad_norm": 0.23047664761543274, "learning_rate": 7.264204928095433e-05, "loss": 0.0181, "step": 38290 }, { "grad_norm": 0.18320105969905853, "learning_rate": 7.262730585017188e-05, "loss": 0.0114, "step": 38300 }, { "grad_norm": 0.20131297409534454, "learning_rate": 7.261255994490131e-05, "loss": 0.0135, "step": 38310 }, { "grad_norm": 0.29007408022880554, "learning_rate": 7.259781156675517e-05, "loss": 0.0108, "step": 38320 }, { "grad_norm": 0.1498374342918396, "learning_rate": 7.258306071734633e-05, "loss": 0.0083, "step": 38330 }, { "grad_norm": 0.11734296381473541, "learning_rate": 7.256830739828792e-05, "loss": 0.0108, "step": 38340 }, { "grad_norm": 0.23280295729637146, "learning_rate": 7.255355161119336e-05, "loss": 0.0139, "step": 38350 }, { "grad_norm": 0.23232676088809967, "learning_rate": 7.253879335767628e-05, "loss": 0.0106, "step": 38360 }, { "grad_norm": 0.7243506908416748, "learning_rate": 7.252403263935067e-05, "loss": 0.0133, "step": 38370 }, { "grad_norm": 0.24795354902744293, "learning_rate": 7.25092694578307e-05, "loss": 0.0128, "step": 38380 }, { "grad_norm": 0.20168353617191315, "learning_rate": 7.249450381473087e-05, "loss": 0.0141, "step": 38390 }, { "grad_norm": 0.15784397721290588, "learning_rate": 7.247973571166593e-05, "loss": 0.0108, "step": 38400 }, { "grad_norm": 0.14725418388843536, "learning_rate": 7.246496515025089e-05, "loss": 0.0139, "step": 38410 }, { "grad_norm": 0.14247475564479828, "learning_rate": 7.245019213210101e-05, "loss": 0.0114, "step": 38420 }, { "grad_norm": 0.21830318868160248, "learning_rate": 7.24354166588319e-05, "loss": 0.0125, "step": 38430 }, { "grad_norm": 0.1746465414762497, "learning_rate": 7.242063873205935e-05, "loss": 0.0109, "step": 38440 }, { "grad_norm": 0.21320904791355133, "learning_rate": 7.240585835339946e-05, "loss": 0.0115, "step": 38450 }, { "grad_norm": 0.2390453964471817, "learning_rate": 7.239107552446857e-05, "loss": 0.0134, "step": 38460 }, { "grad_norm": 0.21489211916923523, "learning_rate": 7.237629024688334e-05, "loss": 0.0147, "step": 38470 }, { "grad_norm": 0.21896202862262726, "learning_rate": 7.236150252226064e-05, "loss": 0.0123, "step": 38480 }, { "grad_norm": 0.19880259037017822, "learning_rate": 7.234671235221765e-05, "loss": 0.0129, "step": 38490 }, { "grad_norm": 0.21114131808280945, "learning_rate": 7.233191973837179e-05, "loss": 0.0166, "step": 38500 }, { "grad_norm": 0.2136686146259308, "learning_rate": 7.231712468234078e-05, "loss": 0.0131, "step": 38510 }, { "grad_norm": 0.21824023127555847, "learning_rate": 7.230232718574254e-05, "loss": 0.0154, "step": 38520 }, { "grad_norm": 0.1629984825849533, "learning_rate": 7.228752725019535e-05, "loss": 0.0111, "step": 38530 }, { "grad_norm": 0.14223521947860718, "learning_rate": 7.227272487731769e-05, "loss": 0.0116, "step": 38540 }, { "grad_norm": 0.25798875093460083, "learning_rate": 7.225792006872831e-05, "loss": 0.0131, "step": 38550 }, { "grad_norm": 0.20504291355609894, "learning_rate": 7.224311282604628e-05, "loss": 0.0125, "step": 38560 }, { "grad_norm": 0.22784565389156342, "learning_rate": 7.222830315089085e-05, "loss": 0.0164, "step": 38570 }, { "grad_norm": 0.15298813581466675, "learning_rate": 7.22134910448816e-05, "loss": 0.0133, "step": 38580 }, { "grad_norm": 0.20936618745326996, "learning_rate": 7.219867650963839e-05, "loss": 0.0139, "step": 38590 }, { "grad_norm": 0.21516789495944977, "learning_rate": 7.21838595467813e-05, "loss": 0.016, "step": 38600 }, { "grad_norm": 0.23857903480529785, "learning_rate": 7.216904015793067e-05, "loss": 0.0142, "step": 38610 }, { "grad_norm": 0.19746200740337372, "learning_rate": 7.215421834470713e-05, "loss": 0.0105, "step": 38620 }, { "grad_norm": 0.16289106011390686, "learning_rate": 7.213939410873158e-05, "loss": 0.0158, "step": 38630 }, { "grad_norm": 0.22067436575889587, "learning_rate": 7.212456745162519e-05, "loss": 0.0142, "step": 38640 }, { "grad_norm": 0.1985703706741333, "learning_rate": 7.210973837500937e-05, "loss": 0.0113, "step": 38650 }, { "grad_norm": 0.1928674876689911, "learning_rate": 7.209490688050578e-05, "loss": 0.0108, "step": 38660 }, { "grad_norm": 0.1382119208574295, "learning_rate": 7.208007296973641e-05, "loss": 0.0157, "step": 38670 }, { "grad_norm": 1.3751734495162964, "learning_rate": 7.206523664432345e-05, "loss": 0.0195, "step": 38680 }, { "grad_norm": 0.6276775002479553, "learning_rate": 7.205039790588939e-05, "loss": 0.0125, "step": 38690 }, { "grad_norm": 0.20393437147140503, "learning_rate": 7.203555675605697e-05, "loss": 0.013, "step": 38700 }, { "grad_norm": 0.17993563413619995, "learning_rate": 7.202071319644917e-05, "loss": 0.0144, "step": 38710 }, { "grad_norm": 0.20724298059940338, "learning_rate": 7.200586722868932e-05, "loss": 0.0142, "step": 38720 }, { "grad_norm": 0.25338536500930786, "learning_rate": 7.199101885440088e-05, "loss": 0.0142, "step": 38730 }, { "grad_norm": 0.45282432436943054, "learning_rate": 7.197616807520771e-05, "loss": 0.0218, "step": 38740 }, { "grad_norm": 0.192453071475029, "learning_rate": 7.196131489273381e-05, "loss": 0.0156, "step": 38750 }, { "grad_norm": 0.20519232749938965, "learning_rate": 7.194645930860355e-05, "loss": 0.0148, "step": 38760 }, { "grad_norm": 0.19843780994415283, "learning_rate": 7.19316013244415e-05, "loss": 0.0132, "step": 38770 }, { "grad_norm": 0.25642240047454834, "learning_rate": 7.191674094187248e-05, "loss": 0.0146, "step": 38780 }, { "grad_norm": 1.95575749874115, "learning_rate": 7.190187816252165e-05, "loss": 0.0205, "step": 38790 }, { "grad_norm": 0.266013503074646, "learning_rate": 7.188701298801435e-05, "loss": 0.0115, "step": 38800 }, { "grad_norm": 0.22917698323726654, "learning_rate": 7.18721454199762e-05, "loss": 0.0174, "step": 38810 }, { "grad_norm": 0.9450730085372925, "learning_rate": 7.185727546003312e-05, "loss": 0.0154, "step": 38820 }, { "grad_norm": 0.21034494042396545, "learning_rate": 7.184240310981126e-05, "loss": 0.0138, "step": 38830 }, { "grad_norm": 0.17330439388751984, "learning_rate": 7.182752837093704e-05, "loss": 0.0134, "step": 38840 }, { "grad_norm": 0.21782124042510986, "learning_rate": 7.181265124503711e-05, "loss": 0.0141, "step": 38850 }, { "grad_norm": 0.23000679910182953, "learning_rate": 7.179777173373847e-05, "loss": 0.0116, "step": 38860 }, { "grad_norm": 0.23756244778633118, "learning_rate": 7.178288983866826e-05, "loss": 0.0211, "step": 38870 }, { "grad_norm": 0.18697330355644226, "learning_rate": 7.176800556145397e-05, "loss": 0.0167, "step": 38880 }, { "grad_norm": 0.18299905955791473, "learning_rate": 7.175311890372334e-05, "loss": 0.0179, "step": 38890 }, { "grad_norm": 0.1497686803340912, "learning_rate": 7.17382298671043e-05, "loss": 0.0164, "step": 38900 }, { "grad_norm": 0.31135424971580505, "learning_rate": 7.172333845322515e-05, "loss": 0.0113, "step": 38910 }, { "grad_norm": 0.2664783000946045, "learning_rate": 7.170844466371436e-05, "loss": 0.0121, "step": 38920 }, { "grad_norm": 0.22176434099674225, "learning_rate": 7.16935485002007e-05, "loss": 0.0116, "step": 38930 }, { "grad_norm": 0.1845446228981018, "learning_rate": 7.167864996431319e-05, "loss": 0.0116, "step": 38940 }, { "grad_norm": 0.23168204724788666, "learning_rate": 7.166374905768111e-05, "loss": 0.012, "step": 38950 }, { "grad_norm": 0.18431350588798523, "learning_rate": 7.164884578193401e-05, "loss": 0.0135, "step": 38960 }, { "grad_norm": 0.20097430050373077, "learning_rate": 7.163394013870168e-05, "loss": 0.0141, "step": 38970 }, { "grad_norm": 0.14887410402297974, "learning_rate": 7.161903212961419e-05, "loss": 0.0109, "step": 38980 }, { "grad_norm": 0.20731870830059052, "learning_rate": 7.160412175630183e-05, "loss": 0.009, "step": 38990 }, { "grad_norm": 0.1718122810125351, "learning_rate": 7.158920902039521e-05, "loss": 0.0091, "step": 39000 }, { "grad_norm": 0.23842498660087585, "learning_rate": 7.157429392352514e-05, "loss": 0.0135, "step": 39010 }, { "grad_norm": 0.23385702073574066, "learning_rate": 7.155937646732275e-05, "loss": 0.0146, "step": 39020 }, { "grad_norm": 0.16499415040016174, "learning_rate": 7.154445665341933e-05, "loss": 0.0118, "step": 39030 }, { "grad_norm": 0.2271498441696167, "learning_rate": 7.152953448344654e-05, "loss": 0.0137, "step": 39040 }, { "grad_norm": 0.1471729874610901, "learning_rate": 7.151460995903624e-05, "loss": 0.0134, "step": 39050 }, { "grad_norm": 0.23021920025348663, "learning_rate": 7.149968308182052e-05, "loss": 0.0103, "step": 39060 }, { "grad_norm": 0.13004253804683685, "learning_rate": 7.14847538534318e-05, "loss": 0.0103, "step": 39070 }, { "grad_norm": 0.30811136960983276, "learning_rate": 7.14698222755027e-05, "loss": 0.0092, "step": 39080 }, { "grad_norm": 0.20774966478347778, "learning_rate": 7.14548883496661e-05, "loss": 0.0106, "step": 39090 }, { "grad_norm": 0.9769132733345032, "learning_rate": 7.143995207755517e-05, "loss": 0.012, "step": 39100 }, { "grad_norm": 0.2142520546913147, "learning_rate": 7.142501346080333e-05, "loss": 0.0132, "step": 39110 }, { "grad_norm": 0.31939274072647095, "learning_rate": 7.141007250104421e-05, "loss": 0.0114, "step": 39120 }, { "grad_norm": 0.21999108791351318, "learning_rate": 7.139512919991176e-05, "loss": 0.0102, "step": 39130 }, { "grad_norm": 0.15682430565357208, "learning_rate": 7.138018355904014e-05, "loss": 0.011, "step": 39140 }, { "grad_norm": 0.16187718510627747, "learning_rate": 7.13652355800638e-05, "loss": 0.0099, "step": 39150 }, { "grad_norm": 0.21725896000862122, "learning_rate": 7.135028526461743e-05, "loss": 0.0152, "step": 39160 }, { "grad_norm": 0.17739680409431458, "learning_rate": 7.133533261433594e-05, "loss": 0.0126, "step": 39170 }, { "grad_norm": 0.1877768635749817, "learning_rate": 7.132037763085457e-05, "loss": 0.0087, "step": 39180 }, { "grad_norm": 0.25214460492134094, "learning_rate": 7.130542031580875e-05, "loss": 0.0112, "step": 39190 }, { "grad_norm": 0.6127287745475769, "learning_rate": 7.12904606708342e-05, "loss": 0.0174, "step": 39200 }, { "grad_norm": 0.16711795330047607, "learning_rate": 7.127549869756687e-05, "loss": 0.0158, "step": 39210 }, { "grad_norm": 0.15935449302196503, "learning_rate": 7.126053439764299e-05, "loss": 0.0122, "step": 39220 }, { "grad_norm": 0.1770862638950348, "learning_rate": 7.124556777269904e-05, "loss": 0.0126, "step": 39230 }, { "grad_norm": 0.1926182359457016, "learning_rate": 7.123059882437174e-05, "loss": 0.0122, "step": 39240 }, { "grad_norm": 0.22994163632392883, "learning_rate": 7.121562755429807e-05, "loss": 0.0151, "step": 39250 }, { "grad_norm": 0.3312722444534302, "learning_rate": 7.120065396411528e-05, "loss": 0.0117, "step": 39260 }, { "grad_norm": 0.166741281747818, "learning_rate": 7.118567805546084e-05, "loss": 0.01, "step": 39270 }, { "grad_norm": 0.1806480586528778, "learning_rate": 7.117069982997248e-05, "loss": 0.0122, "step": 39280 }, { "grad_norm": 0.34510236978530884, "learning_rate": 7.115571928928825e-05, "loss": 0.0119, "step": 39290 }, { "grad_norm": 0.19001410901546478, "learning_rate": 7.114073643504635e-05, "loss": 0.0116, "step": 39300 }, { "grad_norm": 0.1427428424358368, "learning_rate": 7.11257512688853e-05, "loss": 0.0145, "step": 39310 }, { "grad_norm": 0.16664725542068481, "learning_rate": 7.111076379244384e-05, "loss": 0.0129, "step": 39320 }, { "grad_norm": 0.13702663779258728, "learning_rate": 7.109577400736101e-05, "loss": 0.0113, "step": 39330 }, { "grad_norm": 0.15539079904556274, "learning_rate": 7.108078191527605e-05, "loss": 0.0186, "step": 39340 }, { "grad_norm": 0.17018328607082367, "learning_rate": 7.106578751782847e-05, "loss": 0.0101, "step": 39350 }, { "grad_norm": 0.15210090577602386, "learning_rate": 7.105079081665803e-05, "loss": 0.0111, "step": 39360 }, { "grad_norm": 0.11795610934495926, "learning_rate": 7.103579181340476e-05, "loss": 0.0114, "step": 39370 }, { "grad_norm": 0.16429148614406586, "learning_rate": 7.102079050970893e-05, "loss": 0.0148, "step": 39380 }, { "grad_norm": 0.20360137522220612, "learning_rate": 7.100578690721105e-05, "loss": 0.0108, "step": 39390 }, { "grad_norm": 0.2622520625591278, "learning_rate": 7.09907810075519e-05, "loss": 0.0124, "step": 39400 }, { "grad_norm": 0.16284716129302979, "learning_rate": 7.097577281237249e-05, "loss": 0.0102, "step": 39410 }, { "grad_norm": 0.16823206841945648, "learning_rate": 7.09607623233141e-05, "loss": 0.0118, "step": 39420 }, { "grad_norm": 0.8444788455963135, "learning_rate": 7.094574954201827e-05, "loss": 0.0133, "step": 39430 }, { "grad_norm": 0.17488647997379303, "learning_rate": 7.093073447012675e-05, "loss": 0.0149, "step": 39440 }, { "grad_norm": 0.1801375299692154, "learning_rate": 7.09157171092816e-05, "loss": 0.0127, "step": 39450 }, { "grad_norm": 0.16596993803977966, "learning_rate": 7.090069746112504e-05, "loss": 0.0116, "step": 39460 }, { "grad_norm": 0.22569990158081055, "learning_rate": 7.088567552729965e-05, "loss": 0.0132, "step": 39470 }, { "grad_norm": 0.17099802196025848, "learning_rate": 7.087065130944818e-05, "loss": 0.0144, "step": 39480 }, { "grad_norm": 0.2679985463619232, "learning_rate": 7.085562480921366e-05, "loss": 0.0143, "step": 39490 }, { "grad_norm": 0.6611364483833313, "learning_rate": 7.084059602823937e-05, "loss": 0.0168, "step": 39500 }, { "grad_norm": 0.17532075941562653, "learning_rate": 7.082556496816882e-05, "loss": 0.0133, "step": 39510 }, { "grad_norm": 0.17750515043735504, "learning_rate": 7.081053163064582e-05, "loss": 0.0214, "step": 39520 }, { "grad_norm": 0.20983335375785828, "learning_rate": 7.079549601731434e-05, "loss": 0.012, "step": 39530 }, { "grad_norm": 0.15611539781093597, "learning_rate": 7.07804581298187e-05, "loss": 0.026, "step": 39540 }, { "grad_norm": 0.16808965802192688, "learning_rate": 7.07654179698034e-05, "loss": 0.0119, "step": 39550 }, { "grad_norm": 0.18686121702194214, "learning_rate": 7.075037553891321e-05, "loss": 0.0156, "step": 39560 }, { "grad_norm": 0.19828994572162628, "learning_rate": 7.073533083879315e-05, "loss": 0.0146, "step": 39570 }, { "grad_norm": 0.12464068084955215, "learning_rate": 7.072028387108849e-05, "loss": 0.0123, "step": 39580 }, { "grad_norm": 0.1560060828924179, "learning_rate": 7.070523463744472e-05, "loss": 0.0128, "step": 39590 }, { "grad_norm": 0.1603139191865921, "learning_rate": 7.069018313950763e-05, "loss": 0.0106, "step": 39600 }, { "grad_norm": 0.2416064590215683, "learning_rate": 7.067512937892322e-05, "loss": 0.0127, "step": 39610 }, { "grad_norm": 0.1642284095287323, "learning_rate": 7.066007335733775e-05, "loss": 0.0101, "step": 39620 }, { "grad_norm": 0.135154128074646, "learning_rate": 7.064501507639772e-05, "loss": 0.0082, "step": 39630 }, { "grad_norm": 0.15315014123916626, "learning_rate": 7.062995453774987e-05, "loss": 0.0108, "step": 39640 }, { "grad_norm": 0.1264643520116806, "learning_rate": 7.061489174304121e-05, "loss": 0.0149, "step": 39650 }, { "grad_norm": 0.17334838211536407, "learning_rate": 7.0599826693919e-05, "loss": 0.0128, "step": 39660 }, { "grad_norm": 0.13732975721359253, "learning_rate": 7.05847593920307e-05, "loss": 0.0131, "step": 39670 }, { "grad_norm": 0.12018121033906937, "learning_rate": 7.056968983902406e-05, "loss": 0.0114, "step": 39680 }, { "grad_norm": 0.13063944876194, "learning_rate": 7.055461803654706e-05, "loss": 0.0129, "step": 39690 }, { "grad_norm": 0.19898098707199097, "learning_rate": 7.053954398624794e-05, "loss": 0.0106, "step": 39700 }, { "grad_norm": 0.20495769381523132, "learning_rate": 7.052446768977518e-05, "loss": 0.0104, "step": 39710 }, { "grad_norm": 0.26791107654571533, "learning_rate": 7.050938914877748e-05, "loss": 0.0122, "step": 39720 }, { "grad_norm": 0.13768762350082397, "learning_rate": 7.049430836490382e-05, "loss": 0.0111, "step": 39730 }, { "grad_norm": 0.21159842610359192, "learning_rate": 7.047922533980342e-05, "loss": 0.0131, "step": 39740 }, { "grad_norm": 0.19854843616485596, "learning_rate": 7.046414007512571e-05, "loss": 0.0133, "step": 39750 }, { "grad_norm": 0.2313942313194275, "learning_rate": 7.044905257252042e-05, "loss": 0.0156, "step": 39760 }, { "grad_norm": 0.15610966086387634, "learning_rate": 7.04339628336375e-05, "loss": 0.0088, "step": 39770 }, { "grad_norm": 0.14412148296833038, "learning_rate": 7.041887086012711e-05, "loss": 0.0106, "step": 39780 }, { "grad_norm": 0.21773171424865723, "learning_rate": 7.040377665363969e-05, "loss": 0.0155, "step": 39790 }, { "grad_norm": 0.19093084335327148, "learning_rate": 7.038868021582594e-05, "loss": 0.0101, "step": 39800 }, { "grad_norm": 0.211310476064682, "learning_rate": 7.037358154833679e-05, "loss": 0.0125, "step": 39810 }, { "grad_norm": 0.2199700027704239, "learning_rate": 7.035848065282339e-05, "loss": 0.0131, "step": 39820 }, { "grad_norm": 0.1676655411720276, "learning_rate": 7.034337753093714e-05, "loss": 0.0116, "step": 39830 }, { "grad_norm": 0.31636542081832886, "learning_rate": 7.032827218432972e-05, "loss": 0.0145, "step": 39840 }, { "grad_norm": 0.19197919964790344, "learning_rate": 7.031316461465302e-05, "loss": 0.0143, "step": 39850 }, { "grad_norm": 0.4070257544517517, "learning_rate": 7.029805482355915e-05, "loss": 0.0147, "step": 39860 }, { "grad_norm": 0.18927977979183197, "learning_rate": 7.028294281270055e-05, "loss": 0.012, "step": 39870 }, { "grad_norm": 0.2212672382593155, "learning_rate": 7.026782858372982e-05, "loss": 0.0107, "step": 39880 }, { "grad_norm": 0.16890056431293488, "learning_rate": 7.025271213829982e-05, "loss": 0.0121, "step": 39890 }, { "grad_norm": 0.1838514655828476, "learning_rate": 7.023759347806366e-05, "loss": 0.0115, "step": 39900 }, { "grad_norm": 0.6263872981071472, "learning_rate": 7.022247260467472e-05, "loss": 0.0127, "step": 39910 }, { "grad_norm": 1.9299036264419556, "learning_rate": 7.020734951978658e-05, "loss": 0.0165, "step": 39920 }, { "grad_norm": 0.13483552634716034, "learning_rate": 7.019222422505307e-05, "loss": 0.0091, "step": 39930 }, { "grad_norm": 0.18513205647468567, "learning_rate": 7.017709672212827e-05, "loss": 0.0108, "step": 39940 }, { "grad_norm": 0.29351237416267395, "learning_rate": 7.016196701266652e-05, "loss": 0.0128, "step": 39950 }, { "grad_norm": 0.23679369688034058, "learning_rate": 7.014683509832238e-05, "loss": 0.0137, "step": 39960 }, { "grad_norm": 0.30866488814353943, "learning_rate": 7.013170098075063e-05, "loss": 0.0161, "step": 39970 }, { "grad_norm": 0.23784758150577545, "learning_rate": 7.011656466160632e-05, "loss": 0.0119, "step": 39980 }, { "grad_norm": 0.1279408037662506, "learning_rate": 7.010142614254475e-05, "loss": 0.0104, "step": 39990 }, { "grad_norm": 0.35845619440078735, "learning_rate": 7.008628542522147e-05, "loss": 0.0131, "step": 40000 }, { "grad_norm": 0.20251379907131195, "learning_rate": 7.007114251129218e-05, "loss": 0.0106, "step": 40010 }, { "grad_norm": 0.42941048741340637, "learning_rate": 7.005599740241293e-05, "loss": 0.0105, "step": 40020 }, { "grad_norm": 0.1369282305240631, "learning_rate": 7.004085010023996e-05, "loss": 0.0122, "step": 40030 }, { "grad_norm": 0.20080703496932983, "learning_rate": 7.002570060642976e-05, "loss": 0.0091, "step": 40040 }, { "grad_norm": 0.17517542839050293, "learning_rate": 7.001054892263903e-05, "loss": 0.0111, "step": 40050 }, { "grad_norm": 0.7137532830238342, "learning_rate": 6.999539505052477e-05, "loss": 0.016, "step": 40060 }, { "grad_norm": 0.20585952699184418, "learning_rate": 6.998023899174418e-05, "loss": 0.0118, "step": 40070 }, { "grad_norm": 0.34259292483329773, "learning_rate": 6.996508074795467e-05, "loss": 0.0144, "step": 40080 }, { "grad_norm": 0.1949521154165268, "learning_rate": 6.994992032081396e-05, "loss": 0.0129, "step": 40090 }, { "grad_norm": 0.3533462584018707, "learning_rate": 6.993475771197995e-05, "loss": 0.0148, "step": 40100 }, { "grad_norm": 0.12887944281101227, "learning_rate": 6.991959292311082e-05, "loss": 0.0107, "step": 40110 }, { "grad_norm": 0.16329272091388702, "learning_rate": 6.990442595586495e-05, "loss": 0.0122, "step": 40120 }, { "grad_norm": 0.25379711389541626, "learning_rate": 6.988925681190098e-05, "loss": 0.0108, "step": 40130 }, { "grad_norm": 0.1454923152923584, "learning_rate": 6.987408549287778e-05, "loss": 0.0105, "step": 40140 }, { "grad_norm": 0.13419151306152344, "learning_rate": 6.985891200045449e-05, "loss": 0.0094, "step": 40150 }, { "grad_norm": 0.13310188055038452, "learning_rate": 6.984373633629045e-05, "loss": 0.0109, "step": 40160 }, { "grad_norm": 0.1612207144498825, "learning_rate": 6.98285585020452e-05, "loss": 0.0125, "step": 40170 }, { "grad_norm": 0.14205028116703033, "learning_rate": 6.981337849937864e-05, "loss": 0.0086, "step": 40180 }, { "grad_norm": 0.12012888491153717, "learning_rate": 6.979819632995078e-05, "loss": 0.0084, "step": 40190 }, { "grad_norm": 0.3471764326095581, "learning_rate": 6.978301199542193e-05, "loss": 0.0117, "step": 40200 }, { "grad_norm": 0.16256652772426605, "learning_rate": 6.976782549745263e-05, "loss": 0.0124, "step": 40210 }, { "grad_norm": 0.1352369338274002, "learning_rate": 6.975263683770367e-05, "loss": 0.012, "step": 40220 }, { "grad_norm": 0.4237159192562103, "learning_rate": 6.973744601783604e-05, "loss": 0.0114, "step": 40230 }, { "grad_norm": 0.11503138393163681, "learning_rate": 6.972225303951097e-05, "loss": 0.012, "step": 40240 }, { "grad_norm": 0.9263916611671448, "learning_rate": 6.970705790438998e-05, "loss": 0.0193, "step": 40250 }, { "grad_norm": 0.16762100160121918, "learning_rate": 6.969186061413477e-05, "loss": 0.0137, "step": 40260 }, { "grad_norm": 0.1404484212398529, "learning_rate": 6.967666117040727e-05, "loss": 0.0099, "step": 40270 }, { "grad_norm": 0.18946251273155212, "learning_rate": 6.966145957486972e-05, "loss": 0.0096, "step": 40280 }, { "grad_norm": 0.1529417335987091, "learning_rate": 6.964625582918449e-05, "loss": 0.0123, "step": 40290 }, { "grad_norm": 0.16242581605911255, "learning_rate": 6.963104993501425e-05, "loss": 0.0127, "step": 40300 }, { "grad_norm": 0.1928403228521347, "learning_rate": 6.961584189402192e-05, "loss": 0.0142, "step": 40310 }, { "grad_norm": 0.2381058931350708, "learning_rate": 6.96006317078706e-05, "loss": 0.014, "step": 40320 }, { "grad_norm": 0.23436757922172546, "learning_rate": 6.958541937822367e-05, "loss": 0.0137, "step": 40330 }, { "grad_norm": 0.17017002403736115, "learning_rate": 6.957020490674473e-05, "loss": 0.0126, "step": 40340 }, { "grad_norm": 0.14884032309055328, "learning_rate": 6.95549882950976e-05, "loss": 0.0111, "step": 40350 }, { "grad_norm": 0.18268407881259918, "learning_rate": 6.953976954494635e-05, "loss": 0.0107, "step": 40360 }, { "grad_norm": 0.18374773859977722, "learning_rate": 6.952454865795528e-05, "loss": 0.0121, "step": 40370 }, { "grad_norm": 0.1570083349943161, "learning_rate": 6.950932563578892e-05, "loss": 0.0094, "step": 40380 }, { "grad_norm": 0.1853504478931427, "learning_rate": 6.949410048011206e-05, "loss": 0.0107, "step": 40390 }, { "grad_norm": 0.15984782576560974, "learning_rate": 6.947887319258966e-05, "loss": 0.0093, "step": 40400 }, { "grad_norm": 0.14070510864257812, "learning_rate": 6.946364377488696e-05, "loss": 0.0138, "step": 40410 }, { "grad_norm": 0.20940130949020386, "learning_rate": 6.944841222866947e-05, "loss": 0.017, "step": 40420 }, { "grad_norm": 0.15841764211654663, "learning_rate": 6.943317855560284e-05, "loss": 0.0113, "step": 40430 }, { "grad_norm": 0.17301099002361298, "learning_rate": 6.941794275735302e-05, "loss": 0.0138, "step": 40440 }, { "grad_norm": 0.15685543417930603, "learning_rate": 6.94027048355862e-05, "loss": 0.0112, "step": 40450 }, { "grad_norm": 0.21717718243598938, "learning_rate": 6.938746479196871e-05, "loss": 0.0121, "step": 40460 }, { "grad_norm": 0.22722549736499786, "learning_rate": 6.937222262816724e-05, "loss": 0.0152, "step": 40470 }, { "grad_norm": 0.18531250953674316, "learning_rate": 6.935697834584865e-05, "loss": 0.0105, "step": 40480 }, { "grad_norm": 0.13590534031391144, "learning_rate": 6.934173194667998e-05, "loss": 0.0105, "step": 40490 }, { "grad_norm": 0.18188147246837616, "learning_rate": 6.93264834323286e-05, "loss": 0.0101, "step": 40500 }, { "grad_norm": 0.1401703655719757, "learning_rate": 6.931123280446206e-05, "loss": 0.0098, "step": 40510 }, { "grad_norm": 0.18940697610378265, "learning_rate": 6.929598006474811e-05, "loss": 0.0084, "step": 40520 }, { "grad_norm": 0.1538427770137787, "learning_rate": 6.92807252148548e-05, "loss": 0.0095, "step": 40530 }, { "grad_norm": 0.1884579211473465, "learning_rate": 6.926546825645037e-05, "loss": 0.0101, "step": 40540 }, { "grad_norm": 0.154231995344162, "learning_rate": 6.92502091912033e-05, "loss": 0.0087, "step": 40550 }, { "grad_norm": 0.1594664305448532, "learning_rate": 6.923494802078229e-05, "loss": 0.0099, "step": 40560 }, { "grad_norm": 0.17892447113990784, "learning_rate": 6.921968474685628e-05, "loss": 0.0096, "step": 40570 }, { "grad_norm": 0.18556934595108032, "learning_rate": 6.920441937109445e-05, "loss": 0.0108, "step": 40580 }, { "grad_norm": 0.1726130247116089, "learning_rate": 6.91891518951662e-05, "loss": 0.0111, "step": 40590 }, { "grad_norm": 0.1727713793516159, "learning_rate": 6.917388232074114e-05, "loss": 0.0116, "step": 40600 }, { "grad_norm": 0.18892867863178253, "learning_rate": 6.915861064948914e-05, "loss": 0.0095, "step": 40610 }, { "grad_norm": 0.17229560017585754, "learning_rate": 6.914333688308025e-05, "loss": 0.0123, "step": 40620 }, { "grad_norm": 0.18788297474384308, "learning_rate": 6.912806102318487e-05, "loss": 0.0074, "step": 40630 }, { "grad_norm": 0.8987446427345276, "learning_rate": 6.911278307147347e-05, "loss": 0.0102, "step": 40640 }, { "grad_norm": 0.1498294621706009, "learning_rate": 6.909750302961684e-05, "loss": 0.0135, "step": 40650 }, { "grad_norm": 0.1559859663248062, "learning_rate": 6.9082220899286e-05, "loss": 0.008, "step": 40660 }, { "grad_norm": 0.19144712388515472, "learning_rate": 6.906693668215219e-05, "loss": 0.0108, "step": 40670 }, { "grad_norm": 0.18755163252353668, "learning_rate": 6.905165037988683e-05, "loss": 0.0123, "step": 40680 }, { "grad_norm": 0.17957574129104614, "learning_rate": 6.903636199416164e-05, "loss": 0.0096, "step": 40690 }, { "grad_norm": 0.21536992490291595, "learning_rate": 6.902107152664851e-05, "loss": 0.0095, "step": 40700 }, { "grad_norm": 0.10412172973155975, "learning_rate": 6.90057789790196e-05, "loss": 0.0104, "step": 40710 }, { "grad_norm": 0.605844259262085, "learning_rate": 6.899048435294728e-05, "loss": 0.0149, "step": 40720 }, { "grad_norm": 0.262114018201828, "learning_rate": 6.897518765010415e-05, "loss": 0.0191, "step": 40730 }, { "grad_norm": 0.28648555278778076, "learning_rate": 6.895988887216302e-05, "loss": 0.0113, "step": 40740 }, { "grad_norm": 0.24085186421871185, "learning_rate": 6.894458802079694e-05, "loss": 0.0124, "step": 40750 }, { "grad_norm": 0.48180633783340454, "learning_rate": 6.892928509767922e-05, "loss": 0.0173, "step": 40760 }, { "grad_norm": 0.2031354308128357, "learning_rate": 6.891398010448333e-05, "loss": 0.0094, "step": 40770 }, { "grad_norm": 0.1803450733423233, "learning_rate": 6.8898673042883e-05, "loss": 0.0111, "step": 40780 }, { "grad_norm": 0.1564648151397705, "learning_rate": 6.888336391455222e-05, "loss": 0.0131, "step": 40790 }, { "grad_norm": 0.18336759507656097, "learning_rate": 6.886805272116513e-05, "loss": 0.0108, "step": 40800 }, { "grad_norm": 0.1152130663394928, "learning_rate": 6.885273946439617e-05, "loss": 0.009, "step": 40810 }, { "grad_norm": 0.19713279604911804, "learning_rate": 6.883742414591998e-05, "loss": 0.0126, "step": 40820 }, { "grad_norm": 0.3653291165828705, "learning_rate": 6.88221067674114e-05, "loss": 0.011, "step": 40830 }, { "grad_norm": 0.1716330498456955, "learning_rate": 6.88067873305455e-05, "loss": 0.0132, "step": 40840 }, { "grad_norm": 0.17223429679870605, "learning_rate": 6.879146583699765e-05, "loss": 0.0093, "step": 40850 }, { "grad_norm": 0.19970430433750153, "learning_rate": 6.877614228844334e-05, "loss": 0.0123, "step": 40860 }, { "grad_norm": 0.6453712582588196, "learning_rate": 6.876081668655832e-05, "loss": 0.0139, "step": 40870 }, { "grad_norm": 0.16329237818717957, "learning_rate": 6.874548903301861e-05, "loss": 0.0119, "step": 40880 }, { "grad_norm": 0.202996164560318, "learning_rate": 6.873015932950039e-05, "loss": 0.0156, "step": 40890 }, { "grad_norm": 1.4616838693618774, "learning_rate": 6.871482757768012e-05, "loss": 0.0117, "step": 40900 }, { "grad_norm": 0.15434494614601135, "learning_rate": 6.869949377923442e-05, "loss": 0.0122, "step": 40910 }, { "grad_norm": 0.1455463469028473, "learning_rate": 6.868415793584022e-05, "loss": 0.0113, "step": 40920 }, { "grad_norm": 0.1835288256406784, "learning_rate": 6.86688200491746e-05, "loss": 0.0119, "step": 40930 }, { "grad_norm": 0.13993430137634277, "learning_rate": 6.865348012091488e-05, "loss": 0.0086, "step": 40940 }, { "grad_norm": 0.12757153809070587, "learning_rate": 6.86381381527386e-05, "loss": 0.0119, "step": 40950 }, { "grad_norm": 0.17446081340312958, "learning_rate": 6.862279414632359e-05, "loss": 0.0139, "step": 40960 }, { "grad_norm": 0.2026970535516739, "learning_rate": 6.860744810334778e-05, "loss": 0.0104, "step": 40970 }, { "grad_norm": 0.18340946733951569, "learning_rate": 6.859210002548943e-05, "loss": 0.011, "step": 40980 }, { "grad_norm": 0.148909792304039, "learning_rate": 6.8576749914427e-05, "loss": 0.0102, "step": 40990 }, { "grad_norm": 0.23247255384922028, "learning_rate": 6.85613977718391e-05, "loss": 0.0093, "step": 41000 }, { "grad_norm": 0.1468573659658432, "learning_rate": 6.854604359940464e-05, "loss": 0.0089, "step": 41010 }, { "grad_norm": 0.13509127497673035, "learning_rate": 6.853068739880276e-05, "loss": 0.0099, "step": 41020 }, { "grad_norm": 0.11827348917722702, "learning_rate": 6.851532917171276e-05, "loss": 0.0132, "step": 41030 }, { "grad_norm": 0.13133175671100616, "learning_rate": 6.849996891981418e-05, "loss": 0.0106, "step": 41040 }, { "grad_norm": 0.17503400146961212, "learning_rate": 6.848460664478684e-05, "loss": 0.0117, "step": 41050 }, { "grad_norm": 0.16050240397453308, "learning_rate": 6.846924234831069e-05, "loss": 0.0115, "step": 41060 }, { "grad_norm": 0.18170174956321716, "learning_rate": 6.845387603206597e-05, "loss": 0.02, "step": 41070 }, { "grad_norm": 0.2866954207420349, "learning_rate": 6.843850769773312e-05, "loss": 0.0109, "step": 41080 }, { "grad_norm": 0.15575100481510162, "learning_rate": 6.842313734699278e-05, "loss": 0.0164, "step": 41090 }, { "grad_norm": 0.239264115691185, "learning_rate": 6.840776498152584e-05, "loss": 0.0121, "step": 41100 }, { "grad_norm": 1.2519124746322632, "learning_rate": 6.83923906030134e-05, "loss": 0.0152, "step": 41110 }, { "grad_norm": 0.14729143679141998, "learning_rate": 6.837701421313677e-05, "loss": 0.0106, "step": 41120 }, { "grad_norm": 0.5006089806556702, "learning_rate": 6.83616358135775e-05, "loss": 0.0115, "step": 41130 }, { "grad_norm": 0.371076762676239, "learning_rate": 6.834625540601734e-05, "loss": 0.0124, "step": 41140 }, { "grad_norm": 0.3262348473072052, "learning_rate": 6.833087299213829e-05, "loss": 0.0137, "step": 41150 }, { "grad_norm": 0.23944075405597687, "learning_rate": 6.83154885736225e-05, "loss": 0.0129, "step": 41160 }, { "grad_norm": 0.3129112720489502, "learning_rate": 6.830010215215244e-05, "loss": 0.0143, "step": 41170 }, { "grad_norm": 0.20046597719192505, "learning_rate": 6.828471372941072e-05, "loss": 0.0125, "step": 41180 }, { "grad_norm": 0.1513710618019104, "learning_rate": 6.826932330708018e-05, "loss": 0.0132, "step": 41190 }, { "grad_norm": 0.2401137799024582, "learning_rate": 6.825393088684393e-05, "loss": 0.0128, "step": 41200 }, { "grad_norm": 0.2034979909658432, "learning_rate": 6.823853647038524e-05, "loss": 0.0102, "step": 41210 }, { "grad_norm": 0.19763627648353577, "learning_rate": 6.822314005938762e-05, "loss": 0.0114, "step": 41220 }, { "grad_norm": 0.238135427236557, "learning_rate": 6.820774165553479e-05, "loss": 0.0115, "step": 41230 }, { "grad_norm": 0.2563336491584778, "learning_rate": 6.819234126051072e-05, "loss": 0.0134, "step": 41240 }, { "grad_norm": 0.22809691727161407, "learning_rate": 6.817693887599956e-05, "loss": 0.0182, "step": 41250 }, { "grad_norm": 0.2199365198612213, "learning_rate": 6.816153450368568e-05, "loss": 0.0108, "step": 41260 }, { "grad_norm": 0.16532002389431, "learning_rate": 6.814612814525372e-05, "loss": 0.0094, "step": 41270 }, { "grad_norm": 0.28723934292793274, "learning_rate": 6.813071980238846e-05, "loss": 0.0144, "step": 41280 }, { "grad_norm": 0.24613900482654572, "learning_rate": 6.811530947677492e-05, "loss": 0.0115, "step": 41290 }, { "grad_norm": 0.15168264508247375, "learning_rate": 6.809989717009839e-05, "loss": 0.0113, "step": 41300 }, { "grad_norm": 0.19557209312915802, "learning_rate": 6.808448288404431e-05, "loss": 0.0094, "step": 41310 }, { "grad_norm": 0.5041817426681519, "learning_rate": 6.806906662029838e-05, "loss": 0.0105, "step": 41320 }, { "grad_norm": 0.14006425440311432, "learning_rate": 6.805364838054647e-05, "loss": 0.0157, "step": 41330 }, { "grad_norm": 0.584418535232544, "learning_rate": 6.803822816647474e-05, "loss": 0.013, "step": 41340 }, { "grad_norm": 0.1259288489818573, "learning_rate": 6.802280597976949e-05, "loss": 0.012, "step": 41350 }, { "grad_norm": 0.2084011286497116, "learning_rate": 6.800738182211727e-05, "loss": 0.0117, "step": 41360 }, { "grad_norm": 0.2613915205001831, "learning_rate": 6.799195569520485e-05, "loss": 0.0106, "step": 41370 }, { "grad_norm": 0.4741964638233185, "learning_rate": 6.79765276007192e-05, "loss": 0.017, "step": 41380 }, { "grad_norm": 0.16077305376529694, "learning_rate": 6.796109754034753e-05, "loss": 0.0153, "step": 41390 }, { "grad_norm": 0.13709504902362823, "learning_rate": 6.794566551577724e-05, "loss": 0.01, "step": 41400 }, { "grad_norm": 0.2791152000427246, "learning_rate": 6.793023152869594e-05, "loss": 0.0138, "step": 41410 }, { "grad_norm": 0.16079135239124298, "learning_rate": 6.791479558079147e-05, "loss": 0.0097, "step": 41420 }, { "grad_norm": 1.034993290901184, "learning_rate": 6.789935767375191e-05, "loss": 0.013, "step": 41430 }, { "grad_norm": 0.17327597737312317, "learning_rate": 6.78839178092655e-05, "loss": 0.0138, "step": 41440 }, { "grad_norm": 0.18402357399463654, "learning_rate": 6.786847598902072e-05, "loss": 0.0122, "step": 41450 }, { "grad_norm": 0.16861915588378906, "learning_rate": 6.785303221470627e-05, "loss": 0.012, "step": 41460 }, { "grad_norm": 0.21831996738910675, "learning_rate": 6.783758648801108e-05, "loss": 0.0099, "step": 41470 }, { "grad_norm": 0.2035500556230545, "learning_rate": 6.782213881062422e-05, "loss": 0.0149, "step": 41480 }, { "grad_norm": 0.17293818295001984, "learning_rate": 6.780668918423508e-05, "loss": 0.0113, "step": 41490 }, { "grad_norm": 0.19973871111869812, "learning_rate": 6.779123761053317e-05, "loss": 0.0122, "step": 41500 }, { "grad_norm": 0.1835593730211258, "learning_rate": 6.777578409120826e-05, "loss": 0.0102, "step": 41510 }, { "grad_norm": 0.16285018622875214, "learning_rate": 6.776032862795033e-05, "loss": 0.0112, "step": 41520 }, { "grad_norm": 0.16128630936145782, "learning_rate": 6.774487122244956e-05, "loss": 0.0119, "step": 41530 }, { "grad_norm": 0.2062903642654419, "learning_rate": 6.772941187639637e-05, "loss": 0.0089, "step": 41540 }, { "grad_norm": 0.1673114150762558, "learning_rate": 6.771395059148134e-05, "loss": 0.0111, "step": 41550 }, { "grad_norm": 0.11910336464643478, "learning_rate": 6.76984873693953e-05, "loss": 0.0082, "step": 41560 }, { "grad_norm": 0.11779454350471497, "learning_rate": 6.768302221182931e-05, "loss": 0.0107, "step": 41570 }, { "grad_norm": 0.17524106800556183, "learning_rate": 6.766755512047457e-05, "loss": 0.008, "step": 41580 }, { "grad_norm": 0.18800967931747437, "learning_rate": 6.765208609702259e-05, "loss": 0.0081, "step": 41590 }, { "grad_norm": 0.13115747272968292, "learning_rate": 6.763661514316499e-05, "loss": 0.0097, "step": 41600 }, { "grad_norm": 0.1470697522163391, "learning_rate": 6.76211422605937e-05, "loss": 0.0106, "step": 41610 }, { "grad_norm": 0.1357196718454361, "learning_rate": 6.760566745100076e-05, "loss": 0.0105, "step": 41620 }, { "grad_norm": 0.16434448957443237, "learning_rate": 6.75901907160785e-05, "loss": 0.008, "step": 41630 }, { "grad_norm": 0.24600259959697723, "learning_rate": 6.757471205751943e-05, "loss": 0.0117, "step": 41640 }, { "grad_norm": 0.545376181602478, "learning_rate": 6.75592314770163e-05, "loss": 0.0108, "step": 41650 }, { "grad_norm": 0.16782385110855103, "learning_rate": 6.7543748976262e-05, "loss": 0.0095, "step": 41660 }, { "grad_norm": 0.5229715704917908, "learning_rate": 6.752826455694968e-05, "loss": 0.0095, "step": 41670 }, { "grad_norm": 0.11559616774320602, "learning_rate": 6.751277822077271e-05, "loss": 0.0084, "step": 41680 }, { "grad_norm": 0.15937639772891998, "learning_rate": 6.749728996942465e-05, "loss": 0.012, "step": 41690 }, { "grad_norm": 0.20290686190128326, "learning_rate": 6.748179980459924e-05, "loss": 0.0094, "step": 41700 }, { "grad_norm": 0.14501696825027466, "learning_rate": 6.746630772799052e-05, "loss": 0.009, "step": 41710 }, { "grad_norm": 0.11398622393608093, "learning_rate": 6.745081374129262e-05, "loss": 0.0085, "step": 41720 }, { "grad_norm": 0.24145610630512238, "learning_rate": 6.743531784619998e-05, "loss": 0.0109, "step": 41730 }, { "grad_norm": 0.1266811192035675, "learning_rate": 6.741982004440719e-05, "loss": 0.0077, "step": 41740 }, { "grad_norm": 0.1422736793756485, "learning_rate": 6.740432033760907e-05, "loss": 0.0131, "step": 41750 }, { "grad_norm": 0.19582542777061462, "learning_rate": 6.738881872750066e-05, "loss": 0.0066, "step": 41760 }, { "grad_norm": 0.2640018165111542, "learning_rate": 6.737331521577715e-05, "loss": 0.0134, "step": 41770 }, { "grad_norm": 0.25265640020370483, "learning_rate": 6.735780980413403e-05, "loss": 0.0107, "step": 41780 }, { "grad_norm": 0.13993635773658752, "learning_rate": 6.734230249426692e-05, "loss": 0.0084, "step": 41790 }, { "grad_norm": 0.16988906264305115, "learning_rate": 6.732679328787168e-05, "loss": 0.011, "step": 41800 }, { "grad_norm": 0.27178335189819336, "learning_rate": 6.731128218664438e-05, "loss": 0.0141, "step": 41810 }, { "grad_norm": 0.17532067000865936, "learning_rate": 6.729576919228129e-05, "loss": 0.0097, "step": 41820 }, { "grad_norm": 0.1699039340019226, "learning_rate": 6.728025430647888e-05, "loss": 0.0128, "step": 41830 }, { "grad_norm": 0.2567572593688965, "learning_rate": 6.726473753093383e-05, "loss": 0.0116, "step": 41840 }, { "grad_norm": 0.13340607285499573, "learning_rate": 6.724921886734305e-05, "loss": 0.0097, "step": 41850 }, { "grad_norm": 0.15688671171665192, "learning_rate": 6.723369831740363e-05, "loss": 0.0119, "step": 41860 }, { "grad_norm": 0.15027892589569092, "learning_rate": 6.721817588281288e-05, "loss": 0.015, "step": 41870 }, { "grad_norm": 0.3671993017196655, "learning_rate": 6.720265156526828e-05, "loss": 0.0117, "step": 41880 }, { "grad_norm": 0.18232254683971405, "learning_rate": 6.718712536646758e-05, "loss": 0.0134, "step": 41890 }, { "grad_norm": 0.14251220226287842, "learning_rate": 6.71715972881087e-05, "loss": 0.0151, "step": 41900 }, { "grad_norm": 0.14359715580940247, "learning_rate": 6.715606733188974e-05, "loss": 0.0101, "step": 41910 }, { "grad_norm": 0.17910711467266083, "learning_rate": 6.714053549950909e-05, "loss": 0.0136, "step": 41920 }, { "grad_norm": 0.1292029321193695, "learning_rate": 6.71250017926652e-05, "loss": 0.0102, "step": 41930 }, { "grad_norm": 0.4272899329662323, "learning_rate": 6.71094662130569e-05, "loss": 0.0152, "step": 41940 }, { "grad_norm": 0.13772110641002655, "learning_rate": 6.709392876238307e-05, "loss": 0.0108, "step": 41950 }, { "grad_norm": 2.7178478240966797, "learning_rate": 6.707838944234289e-05, "loss": 0.0162, "step": 41960 }, { "grad_norm": 0.20691877603530884, "learning_rate": 6.706284825463574e-05, "loss": 0.0111, "step": 41970 }, { "grad_norm": 0.15829356014728546, "learning_rate": 6.704730520096114e-05, "loss": 0.0113, "step": 41980 }, { "grad_norm": 0.30467715859413147, "learning_rate": 6.703176028301888e-05, "loss": 0.012, "step": 41990 }, { "grad_norm": 0.1773892045021057, "learning_rate": 6.701621350250892e-05, "loss": 0.0129, "step": 42000 }, { "grad_norm": 0.15413977205753326, "learning_rate": 6.700066486113144e-05, "loss": 0.0106, "step": 42010 }, { "grad_norm": 0.22500592470169067, "learning_rate": 6.69851143605868e-05, "loss": 0.013, "step": 42020 }, { "grad_norm": 0.13076156377792358, "learning_rate": 6.696956200257561e-05, "loss": 0.0119, "step": 42030 }, { "grad_norm": 0.18389630317687988, "learning_rate": 6.695400778879863e-05, "loss": 0.0151, "step": 42040 }, { "grad_norm": 0.1714172512292862, "learning_rate": 6.693845172095683e-05, "loss": 0.0123, "step": 42050 }, { "grad_norm": 0.2631380259990692, "learning_rate": 6.692289380075142e-05, "loss": 0.0144, "step": 42060 }, { "grad_norm": 0.16054652631282806, "learning_rate": 6.690733402988379e-05, "loss": 0.0148, "step": 42070 }, { "grad_norm": 0.23351167142391205, "learning_rate": 6.689177241005553e-05, "loss": 0.0122, "step": 42080 }, { "grad_norm": 0.25296300649642944, "learning_rate": 6.687620894296844e-05, "loss": 0.0149, "step": 42090 }, { "grad_norm": 0.27076366543769836, "learning_rate": 6.686064363032451e-05, "loss": 0.0128, "step": 42100 }, { "grad_norm": 0.14453056454658508, "learning_rate": 6.684507647382596e-05, "loss": 0.0112, "step": 42110 }, { "grad_norm": 0.20485877990722656, "learning_rate": 6.682950747517513e-05, "loss": 0.01, "step": 42120 }, { "grad_norm": 0.1534576267004013, "learning_rate": 6.68139366360747e-05, "loss": 0.0087, "step": 42130 }, { "grad_norm": 0.19075807929039001, "learning_rate": 6.679836395822744e-05, "loss": 0.0109, "step": 42140 }, { "grad_norm": 0.1832122653722763, "learning_rate": 6.678278944333633e-05, "loss": 0.0141, "step": 42150 }, { "grad_norm": 0.16047105193138123, "learning_rate": 6.676721309310462e-05, "loss": 0.0124, "step": 42160 }, { "grad_norm": 0.8794337511062622, "learning_rate": 6.67516349092357e-05, "loss": 0.0108, "step": 42170 }, { "grad_norm": 0.1502603143453598, "learning_rate": 6.673605489343315e-05, "loss": 0.0083, "step": 42180 }, { "grad_norm": 0.5144222378730774, "learning_rate": 6.672047304740081e-05, "loss": 0.0107, "step": 42190 }, { "grad_norm": 0.17127662897109985, "learning_rate": 6.670488937284268e-05, "loss": 0.0096, "step": 42200 }, { "grad_norm": 1.6333357095718384, "learning_rate": 6.668930387146294e-05, "loss": 0.0124, "step": 42210 }, { "grad_norm": 0.2042832225561142, "learning_rate": 6.667371654496605e-05, "loss": 0.0159, "step": 42220 }, { "grad_norm": 0.15165214240550995, "learning_rate": 6.665812739505659e-05, "loss": 0.0084, "step": 42230 }, { "grad_norm": 0.16348710656166077, "learning_rate": 6.664253642343935e-05, "loss": 0.0157, "step": 42240 }, { "grad_norm": 0.14933957159519196, "learning_rate": 6.662694363181935e-05, "loss": 0.0119, "step": 42250 }, { "grad_norm": 0.13367538154125214, "learning_rate": 6.66113490219018e-05, "loss": 0.0112, "step": 42260 }, { "grad_norm": 0.38673675060272217, "learning_rate": 6.659575259539207e-05, "loss": 0.0103, "step": 42270 }, { "grad_norm": 0.1954668164253235, "learning_rate": 6.65801543539958e-05, "loss": 0.0105, "step": 42280 }, { "grad_norm": 0.21369025111198425, "learning_rate": 6.65645542994188e-05, "loss": 0.0116, "step": 42290 }, { "grad_norm": 0.1744960993528366, "learning_rate": 6.654895243336702e-05, "loss": 0.0115, "step": 42300 }, { "grad_norm": 0.15343739092350006, "learning_rate": 6.65333487575467e-05, "loss": 0.0145, "step": 42310 }, { "grad_norm": 1.4745889902114868, "learning_rate": 6.651774327366422e-05, "loss": 0.013, "step": 42320 }, { "grad_norm": 0.8849655389785767, "learning_rate": 6.650213598342613e-05, "loss": 0.0126, "step": 42330 }, { "grad_norm": 0.14685054123401642, "learning_rate": 6.648652688853931e-05, "loss": 0.0113, "step": 42340 }, { "grad_norm": 0.20213833451271057, "learning_rate": 6.647091599071066e-05, "loss": 0.0146, "step": 42350 }, { "grad_norm": 0.14748890697956085, "learning_rate": 6.645530329164743e-05, "loss": 0.0126, "step": 42360 }, { "grad_norm": 0.17264136672019958, "learning_rate": 6.643968879305694e-05, "loss": 0.0114, "step": 42370 }, { "grad_norm": 1.0991542339324951, "learning_rate": 6.642407249664683e-05, "loss": 0.016, "step": 42380 }, { "grad_norm": 0.11934586614370346, "learning_rate": 6.640845440412483e-05, "loss": 0.0117, "step": 42390 }, { "grad_norm": 0.18915696442127228, "learning_rate": 6.639283451719893e-05, "loss": 0.0126, "step": 42400 }, { "grad_norm": 0.22838032245635986, "learning_rate": 6.637721283757729e-05, "loss": 0.0124, "step": 42410 }, { "grad_norm": 0.261807382106781, "learning_rate": 6.636158936696828e-05, "loss": 0.0127, "step": 42420 }, { "grad_norm": 0.3321356177330017, "learning_rate": 6.634596410708047e-05, "loss": 0.0138, "step": 42430 }, { "grad_norm": 0.19519007205963135, "learning_rate": 6.633033705962257e-05, "loss": 0.0122, "step": 42440 }, { "grad_norm": 0.20316529273986816, "learning_rate": 6.631470822630359e-05, "loss": 0.017, "step": 42450 }, { "grad_norm": 0.1845511496067047, "learning_rate": 6.629907760883263e-05, "loss": 0.0178, "step": 42460 }, { "grad_norm": 0.5128469467163086, "learning_rate": 6.628344520891905e-05, "loss": 0.0114, "step": 42470 }, { "grad_norm": 0.16065455973148346, "learning_rate": 6.62678110282724e-05, "loss": 0.0155, "step": 42480 }, { "grad_norm": 0.1456352323293686, "learning_rate": 6.625217506860237e-05, "loss": 0.0141, "step": 42490 }, { "grad_norm": 0.14928945899009705, "learning_rate": 6.623653733161892e-05, "loss": 0.0115, "step": 42500 }, { "grad_norm": 0.17960414290428162, "learning_rate": 6.622089781903217e-05, "loss": 0.013, "step": 42510 }, { "grad_norm": 0.12448827177286148, "learning_rate": 6.620525653255241e-05, "loss": 0.0098, "step": 42520 }, { "grad_norm": 0.17442697286605835, "learning_rate": 6.618961347389015e-05, "loss": 0.0115, "step": 42530 }, { "grad_norm": 0.16666483879089355, "learning_rate": 6.617396864475613e-05, "loss": 0.0123, "step": 42540 }, { "grad_norm": 0.18783718347549438, "learning_rate": 6.61583220468612e-05, "loss": 0.0128, "step": 42550 }, { "grad_norm": 0.2329651415348053, "learning_rate": 6.614267368191645e-05, "loss": 0.0115, "step": 42560 }, { "grad_norm": 0.2980003356933594, "learning_rate": 6.61270235516332e-05, "loss": 0.0129, "step": 42570 }, { "grad_norm": 0.1835963875055313, "learning_rate": 6.611137165772287e-05, "loss": 0.012, "step": 42580 }, { "grad_norm": 0.14025723934173584, "learning_rate": 6.60957180018972e-05, "loss": 0.0118, "step": 42590 }, { "grad_norm": 0.24994084239006042, "learning_rate": 6.608006258586797e-05, "loss": 0.0108, "step": 42600 }, { "grad_norm": 0.6411197781562805, "learning_rate": 6.606440541134731e-05, "loss": 0.0154, "step": 42610 }, { "grad_norm": 0.15577034652233124, "learning_rate": 6.60487464800474e-05, "loss": 0.0126, "step": 42620 }, { "grad_norm": 0.17510436475276947, "learning_rate": 6.603308579368071e-05, "loss": 0.0115, "step": 42630 }, { "grad_norm": 0.19181093573570251, "learning_rate": 6.601742335395987e-05, "loss": 0.0145, "step": 42640 }, { "grad_norm": 0.18137700855731964, "learning_rate": 6.600175916259769e-05, "loss": 0.0137, "step": 42650 }, { "grad_norm": 0.15617944300174713, "learning_rate": 6.598609322130718e-05, "loss": 0.0099, "step": 42660 }, { "grad_norm": 0.2598097026348114, "learning_rate": 6.597042553180155e-05, "loss": 0.0112, "step": 42670 }, { "grad_norm": 0.14729857444763184, "learning_rate": 6.59547560957942e-05, "loss": 0.0098, "step": 42680 }, { "grad_norm": 0.29535531997680664, "learning_rate": 6.59390849149987e-05, "loss": 0.0145, "step": 42690 }, { "grad_norm": 0.2491978257894516, "learning_rate": 6.592341199112886e-05, "loss": 0.0135, "step": 42700 }, { "grad_norm": 0.12831570208072662, "learning_rate": 6.590773732589861e-05, "loss": 0.0106, "step": 42710 }, { "grad_norm": 0.2085951417684555, "learning_rate": 6.589206092102213e-05, "loss": 0.0122, "step": 42720 }, { "grad_norm": 0.17604002356529236, "learning_rate": 6.587638277821375e-05, "loss": 0.011, "step": 42730 }, { "grad_norm": 0.14530055224895477, "learning_rate": 6.586070289918803e-05, "loss": 0.0112, "step": 42740 }, { "grad_norm": 0.18607738614082336, "learning_rate": 6.584502128565968e-05, "loss": 0.0143, "step": 42750 }, { "grad_norm": 0.16607362031936646, "learning_rate": 6.582933793934363e-05, "loss": 0.0097, "step": 42760 }, { "grad_norm": 0.43715527653694153, "learning_rate": 6.581365286195498e-05, "loss": 0.0099, "step": 42770 }, { "grad_norm": 0.15358923375606537, "learning_rate": 6.579796605520903e-05, "loss": 0.0097, "step": 42780 }, { "grad_norm": 0.18782012164592743, "learning_rate": 6.578227752082127e-05, "loss": 0.0128, "step": 42790 }, { "grad_norm": 0.18952472507953644, "learning_rate": 6.576658726050735e-05, "loss": 0.0102, "step": 42800 }, { "grad_norm": 0.22799916565418243, "learning_rate": 6.575089527598316e-05, "loss": 0.0116, "step": 42810 }, { "grad_norm": 0.22045405209064484, "learning_rate": 6.573520156896476e-05, "loss": 0.0137, "step": 42820 }, { "grad_norm": 0.22566911578178406, "learning_rate": 6.571950614116835e-05, "loss": 0.0123, "step": 42830 }, { "grad_norm": 0.2619026005268097, "learning_rate": 6.570380899431039e-05, "loss": 0.0155, "step": 42840 }, { "grad_norm": 0.23173858225345612, "learning_rate": 6.568811013010749e-05, "loss": 0.0134, "step": 42850 }, { "grad_norm": 0.22090552747249603, "learning_rate": 6.567240955027645e-05, "loss": 0.0155, "step": 42860 }, { "grad_norm": 0.27229058742523193, "learning_rate": 6.565670725653427e-05, "loss": 0.014, "step": 42870 }, { "grad_norm": 0.2209639698266983, "learning_rate": 6.564100325059812e-05, "loss": 0.0143, "step": 42880 }, { "grad_norm": 0.20348870754241943, "learning_rate": 6.562529753418538e-05, "loss": 0.0113, "step": 42890 }, { "grad_norm": 0.18004311621189117, "learning_rate": 6.56095901090136e-05, "loss": 0.0132, "step": 42900 }, { "grad_norm": 0.276481956243515, "learning_rate": 6.55938809768005e-05, "loss": 0.0092, "step": 42910 }, { "grad_norm": 0.22283010184764862, "learning_rate": 6.557817013926402e-05, "loss": 0.0149, "step": 42920 }, { "grad_norm": 0.3285568952560425, "learning_rate": 6.556245759812227e-05, "loss": 0.0105, "step": 42930 }, { "grad_norm": 0.16411380469799042, "learning_rate": 6.554674335509357e-05, "loss": 0.0133, "step": 42940 }, { "grad_norm": 0.22339998185634613, "learning_rate": 6.553102741189638e-05, "loss": 0.0098, "step": 42950 }, { "grad_norm": 0.29430219531059265, "learning_rate": 6.55153097702494e-05, "loss": 0.0104, "step": 42960 }, { "grad_norm": 0.26381081342697144, "learning_rate": 6.549959043187146e-05, "loss": 0.0134, "step": 42970 }, { "grad_norm": 0.24281153082847595, "learning_rate": 6.54838693984816e-05, "loss": 0.0126, "step": 42980 }, { "grad_norm": 0.25482940673828125, "learning_rate": 6.546814667179909e-05, "loss": 0.0148, "step": 42990 }, { "grad_norm": 0.1381053626537323, "learning_rate": 6.545242225354328e-05, "loss": 0.0105, "step": 43000 }, { "grad_norm": 0.15775343775749207, "learning_rate": 6.543669614543383e-05, "loss": 0.0101, "step": 43010 }, { "grad_norm": 0.20930138230323792, "learning_rate": 6.542096834919049e-05, "loss": 0.0116, "step": 43020 }, { "grad_norm": 0.3160685896873474, "learning_rate": 6.540523886653324e-05, "loss": 0.0118, "step": 43030 }, { "grad_norm": 0.18442407250404358, "learning_rate": 6.538950769918222e-05, "loss": 0.0123, "step": 43040 }, { "grad_norm": 0.14836908876895905, "learning_rate": 6.537377484885779e-05, "loss": 0.0097, "step": 43050 }, { "grad_norm": 0.13515575230121613, "learning_rate": 6.535804031728045e-05, "loss": 0.0106, "step": 43060 }, { "grad_norm": 0.11439085751771927, "learning_rate": 6.534230410617092e-05, "loss": 0.0118, "step": 43070 }, { "grad_norm": 0.12451053410768509, "learning_rate": 6.532656621725007e-05, "loss": 0.0106, "step": 43080 }, { "grad_norm": 0.498340368270874, "learning_rate": 6.531082665223898e-05, "loss": 0.0127, "step": 43090 }, { "grad_norm": 0.1999969780445099, "learning_rate": 6.529508541285889e-05, "loss": 0.0096, "step": 43100 }, { "grad_norm": 0.16713391244411469, "learning_rate": 6.527934250083128e-05, "loss": 0.0142, "step": 43110 }, { "grad_norm": 0.13915541768074036, "learning_rate": 6.526359791787775e-05, "loss": 0.0121, "step": 43120 }, { "grad_norm": 0.13983526825904846, "learning_rate": 6.524785166572009e-05, "loss": 0.0105, "step": 43130 }, { "grad_norm": 0.24015262722969055, "learning_rate": 6.523210374608029e-05, "loss": 0.0128, "step": 43140 }, { "grad_norm": 0.19292478263378143, "learning_rate": 6.521635416068054e-05, "loss": 0.0155, "step": 43150 }, { "grad_norm": 0.1927485466003418, "learning_rate": 6.520060291124317e-05, "loss": 0.0145, "step": 43160 }, { "grad_norm": 0.16264189779758453, "learning_rate": 6.518484999949072e-05, "loss": 0.0131, "step": 43170 }, { "grad_norm": 0.2019888311624527, "learning_rate": 6.516909542714591e-05, "loss": 0.0135, "step": 43180 }, { "grad_norm": 0.20105092227458954, "learning_rate": 6.515333919593161e-05, "loss": 0.0107, "step": 43190 }, { "grad_norm": 0.12382736802101135, "learning_rate": 6.513758130757094e-05, "loss": 0.0149, "step": 43200 }, { "grad_norm": 0.13556408882141113, "learning_rate": 6.512182176378713e-05, "loss": 0.009, "step": 43210 }, { "grad_norm": 0.11886238306760788, "learning_rate": 6.510606056630362e-05, "loss": 0.0121, "step": 43220 }, { "grad_norm": 0.15665334463119507, "learning_rate": 6.509029771684405e-05, "loss": 0.0129, "step": 43230 }, { "grad_norm": 0.17098963260650635, "learning_rate": 6.50745332171322e-05, "loss": 0.0109, "step": 43240 }, { "grad_norm": 0.1487593650817871, "learning_rate": 6.505876706889207e-05, "loss": 0.0109, "step": 43250 }, { "grad_norm": 0.18903285264968872, "learning_rate": 6.504299927384781e-05, "loss": 0.0126, "step": 43260 }, { "grad_norm": 0.16453209519386292, "learning_rate": 6.502722983372378e-05, "loss": 0.0103, "step": 43270 }, { "grad_norm": 0.2900947630405426, "learning_rate": 6.501145875024446e-05, "loss": 0.0107, "step": 43280 }, { "grad_norm": 0.13705293834209442, "learning_rate": 6.499568602513459e-05, "loss": 0.012, "step": 43290 }, { "grad_norm": 0.154841810464859, "learning_rate": 6.497991166011903e-05, "loss": 0.0166, "step": 43300 }, { "grad_norm": 0.17480291426181793, "learning_rate": 6.49641356569229e-05, "loss": 0.0108, "step": 43310 }, { "grad_norm": 0.17301805317401886, "learning_rate": 6.494835801727135e-05, "loss": 0.0121, "step": 43320 }, { "grad_norm": 0.161381334066391, "learning_rate": 6.493257874288986e-05, "loss": 0.0093, "step": 43330 }, { "grad_norm": 0.21584948897361755, "learning_rate": 6.491679783550399e-05, "loss": 0.0098, "step": 43340 }, { "grad_norm": 0.2029428333044052, "learning_rate": 6.490101529683957e-05, "loss": 0.0114, "step": 43350 }, { "grad_norm": 0.22977715730667114, "learning_rate": 6.488523112862248e-05, "loss": 0.0102, "step": 43360 }, { "grad_norm": 0.1779511272907257, "learning_rate": 6.486944533257891e-05, "loss": 0.0115, "step": 43370 }, { "grad_norm": 0.18747031688690186, "learning_rate": 6.485365791043517e-05, "loss": 0.0139, "step": 43380 }, { "grad_norm": 0.17319077253341675, "learning_rate": 6.483786886391772e-05, "loss": 0.0118, "step": 43390 }, { "grad_norm": 0.19293901324272156, "learning_rate": 6.482207819475323e-05, "loss": 0.0107, "step": 43400 }, { "grad_norm": 0.2097543329000473, "learning_rate": 6.480628590466857e-05, "loss": 0.0126, "step": 43410 }, { "grad_norm": 0.16315370798110962, "learning_rate": 6.479049199539074e-05, "loss": 0.0115, "step": 43420 }, { "grad_norm": 0.1230928972363472, "learning_rate": 6.477469646864693e-05, "loss": 0.012, "step": 43430 }, { "grad_norm": 0.44373852014541626, "learning_rate": 6.475889932616454e-05, "loss": 0.0119, "step": 43440 }, { "grad_norm": 1.0520204305648804, "learning_rate": 6.474310056967111e-05, "loss": 0.0132, "step": 43450 }, { "grad_norm": 0.2048487663269043, "learning_rate": 6.472730020089437e-05, "loss": 0.0102, "step": 43460 }, { "grad_norm": 0.16741150617599487, "learning_rate": 6.471149822156222e-05, "loss": 0.0124, "step": 43470 }, { "grad_norm": 0.18437519669532776, "learning_rate": 6.469569463340275e-05, "loss": 0.0124, "step": 43480 }, { "grad_norm": 0.22992169857025146, "learning_rate": 6.467988943814418e-05, "loss": 0.013, "step": 43490 }, { "grad_norm": 0.19600573182106018, "learning_rate": 6.4664082637515e-05, "loss": 0.0134, "step": 43500 }, { "grad_norm": 0.1916801631450653, "learning_rate": 6.464827423324377e-05, "loss": 0.0108, "step": 43510 }, { "grad_norm": 0.25997188687324524, "learning_rate": 6.463246422705931e-05, "loss": 0.0127, "step": 43520 }, { "grad_norm": 0.17377105355262756, "learning_rate": 6.461665262069055e-05, "loss": 0.0228, "step": 43530 }, { "grad_norm": 0.14975157380104065, "learning_rate": 6.460083941586665e-05, "loss": 0.0158, "step": 43540 }, { "grad_norm": 0.2138051986694336, "learning_rate": 6.45850246143169e-05, "loss": 0.0173, "step": 43550 }, { "grad_norm": 0.17819061875343323, "learning_rate": 6.456920821777077e-05, "loss": 0.0142, "step": 43560 }, { "grad_norm": 0.18135212361812592, "learning_rate": 6.455339022795796e-05, "loss": 0.0105, "step": 43570 }, { "grad_norm": 0.6099959015846252, "learning_rate": 6.453757064660825e-05, "loss": 0.0122, "step": 43580 }, { "grad_norm": 0.7449890375137329, "learning_rate": 6.452174947545169e-05, "loss": 0.0102, "step": 43590 }, { "grad_norm": 0.3322823941707611, "learning_rate": 6.450592671621842e-05, "loss": 0.0116, "step": 43600 }, { "grad_norm": 0.1653304249048233, "learning_rate": 6.449010237063882e-05, "loss": 0.012, "step": 43610 }, { "grad_norm": 0.14139631390571594, "learning_rate": 6.447427644044343e-05, "loss": 0.0121, "step": 43620 }, { "grad_norm": 0.1304919719696045, "learning_rate": 6.44584489273629e-05, "loss": 0.0105, "step": 43630 }, { "grad_norm": 0.12144221365451813, "learning_rate": 6.444261983312817e-05, "loss": 0.0108, "step": 43640 }, { "grad_norm": 0.16898345947265625, "learning_rate": 6.442678915947023e-05, "loss": 0.0108, "step": 43650 }, { "grad_norm": 0.17124433815479279, "learning_rate": 6.441095690812032e-05, "loss": 0.0107, "step": 43660 }, { "grad_norm": 0.14324073493480682, "learning_rate": 6.439512308080982e-05, "loss": 0.0101, "step": 43670 }, { "grad_norm": 0.18053342401981354, "learning_rate": 6.43792876792703e-05, "loss": 0.0104, "step": 43680 }, { "grad_norm": 0.1443050354719162, "learning_rate": 6.43634507052335e-05, "loss": 0.0109, "step": 43690 }, { "grad_norm": 0.1858503818511963, "learning_rate": 6.434761216043135e-05, "loss": 0.0104, "step": 43700 }, { "grad_norm": 0.10615991055965424, "learning_rate": 6.433177204659589e-05, "loss": 0.0094, "step": 43710 }, { "grad_norm": 0.14204911887645721, "learning_rate": 6.431593036545938e-05, "loss": 0.0103, "step": 43720 }, { "grad_norm": 0.15439969301223755, "learning_rate": 6.430008711875426e-05, "loss": 0.0106, "step": 43730 }, { "grad_norm": 0.1786615401506424, "learning_rate": 6.42842423082131e-05, "loss": 0.0093, "step": 43740 }, { "grad_norm": 0.13694491982460022, "learning_rate": 6.426839593556868e-05, "loss": 0.0083, "step": 43750 }, { "grad_norm": 0.2737528681755066, "learning_rate": 6.425254800255392e-05, "loss": 0.0124, "step": 43760 }, { "grad_norm": 0.17262524366378784, "learning_rate": 6.423669851090196e-05, "loss": 0.0106, "step": 43770 }, { "grad_norm": 0.15198835730552673, "learning_rate": 6.422084746234604e-05, "loss": 0.0113, "step": 43780 }, { "grad_norm": 0.4476481080055237, "learning_rate": 6.420499485861965e-05, "loss": 0.0078, "step": 43790 }, { "grad_norm": 0.13634926080703735, "learning_rate": 6.418914070145636e-05, "loss": 0.0102, "step": 43800 }, { "grad_norm": 0.12113045156002045, "learning_rate": 6.417328499258996e-05, "loss": 0.0086, "step": 43810 }, { "grad_norm": 0.4700985550880432, "learning_rate": 6.415742773375446e-05, "loss": 0.0152, "step": 43820 }, { "grad_norm": 0.39514410495758057, "learning_rate": 6.414156892668393e-05, "loss": 0.0114, "step": 43830 }, { "grad_norm": 0.23775100708007812, "learning_rate": 6.412570857311267e-05, "loss": 0.0107, "step": 43840 }, { "grad_norm": 0.20372740924358368, "learning_rate": 6.410984667477518e-05, "loss": 0.0163, "step": 43850 }, { "grad_norm": 0.177334725856781, "learning_rate": 6.409398323340607e-05, "loss": 0.0127, "step": 43860 }, { "grad_norm": 0.15112468600273132, "learning_rate": 6.407811825074012e-05, "loss": 0.0134, "step": 43870 }, { "grad_norm": 0.2437109351158142, "learning_rate": 6.406225172851234e-05, "loss": 0.013, "step": 43880 }, { "grad_norm": 0.18862846493721008, "learning_rate": 6.404638366845786e-05, "loss": 0.0103, "step": 43890 }, { "grad_norm": 0.1299038678407669, "learning_rate": 6.403051407231196e-05, "loss": 0.0124, "step": 43900 }, { "grad_norm": 0.13684459030628204, "learning_rate": 6.401464294181016e-05, "loss": 0.0145, "step": 43910 }, { "grad_norm": 0.15998415648937225, "learning_rate": 6.399877027868808e-05, "loss": 0.0167, "step": 43920 }, { "grad_norm": 0.15979014337062836, "learning_rate": 6.39828960846815e-05, "loss": 0.0106, "step": 43930 }, { "grad_norm": 0.13954924046993256, "learning_rate": 6.396702036152644e-05, "loss": 0.0129, "step": 43940 }, { "grad_norm": 0.20949603617191315, "learning_rate": 6.395114311095905e-05, "loss": 0.0118, "step": 43950 }, { "grad_norm": 0.14750643074512482, "learning_rate": 6.393526433471559e-05, "loss": 0.0104, "step": 43960 }, { "grad_norm": 0.14921016991138458, "learning_rate": 6.391938403453262e-05, "loss": 0.0121, "step": 43970 }, { "grad_norm": 0.16248121857643127, "learning_rate": 6.390350221214671e-05, "loss": 0.0107, "step": 43980 }, { "grad_norm": 0.13610486686229706, "learning_rate": 6.38876188692947e-05, "loss": 0.0117, "step": 43990 }, { "grad_norm": 0.18003958463668823, "learning_rate": 6.387173400771358e-05, "loss": 0.0114, "step": 44000 }, { "grad_norm": 0.15929284691810608, "learning_rate": 6.385584762914047e-05, "loss": 0.0074, "step": 44010 }, { "grad_norm": 0.13427776098251343, "learning_rate": 6.38399597353127e-05, "loss": 0.0126, "step": 44020 }, { "grad_norm": 0.16508081555366516, "learning_rate": 6.382407032796775e-05, "loss": 0.0088, "step": 44030 }, { "grad_norm": 0.23622113466262817, "learning_rate": 6.380817940884325e-05, "loss": 0.0107, "step": 44040 }, { "grad_norm": 0.15466175973415375, "learning_rate": 6.379228697967702e-05, "loss": 0.0098, "step": 44050 }, { "grad_norm": 0.17820259928703308, "learning_rate": 6.3776393042207e-05, "loss": 0.0115, "step": 44060 }, { "grad_norm": 0.13188831508159637, "learning_rate": 6.376049759817137e-05, "loss": 0.0122, "step": 44070 }, { "grad_norm": 0.2318289577960968, "learning_rate": 6.37446006493084e-05, "loss": 0.0086, "step": 44080 }, { "grad_norm": 0.20087279379367828, "learning_rate": 6.372870219735656e-05, "loss": 0.0117, "step": 44090 }, { "grad_norm": 0.3018956184387207, "learning_rate": 6.371280224405452e-05, "loss": 0.0127, "step": 44100 }, { "grad_norm": 0.20812708139419556, "learning_rate": 6.369690079114101e-05, "loss": 0.0112, "step": 44110 }, { "grad_norm": 0.22161979973316193, "learning_rate": 6.368099784035504e-05, "loss": 0.0118, "step": 44120 }, { "grad_norm": 0.26043590903282166, "learning_rate": 6.366509339343572e-05, "loss": 0.012, "step": 44130 }, { "grad_norm": 0.1917296200990677, "learning_rate": 6.364918745212232e-05, "loss": 0.0113, "step": 44140 }, { "grad_norm": 0.17508122324943542, "learning_rate": 6.363328001815431e-05, "loss": 0.0121, "step": 44150 }, { "grad_norm": 0.20409220457077026, "learning_rate": 6.361737109327128e-05, "loss": 0.0158, "step": 44160 }, { "grad_norm": 0.628098726272583, "learning_rate": 6.360146067921301e-05, "loss": 0.0126, "step": 44170 }, { "grad_norm": 0.19966046512126923, "learning_rate": 6.35855487777195e-05, "loss": 0.0153, "step": 44180 }, { "grad_norm": 0.21009297668933868, "learning_rate": 6.356963539053076e-05, "loss": 0.0155, "step": 44190 }, { "grad_norm": 0.1484273225069046, "learning_rate": 6.355372051938711e-05, "loss": 0.009, "step": 44200 }, { "grad_norm": 0.20071633160114288, "learning_rate": 6.353780416602894e-05, "loss": 0.0098, "step": 44210 }, { "grad_norm": 0.3142208755016327, "learning_rate": 6.352188633219689e-05, "loss": 0.0128, "step": 44220 }, { "grad_norm": 0.2370382696390152, "learning_rate": 6.350596701963166e-05, "loss": 0.0116, "step": 44230 }, { "grad_norm": 0.1522538810968399, "learning_rate": 6.349004623007419e-05, "loss": 0.0124, "step": 44240 }, { "grad_norm": 0.2392958253622055, "learning_rate": 6.347412396526555e-05, "loss": 0.0133, "step": 44250 }, { "grad_norm": 0.3404901921749115, "learning_rate": 6.345820022694696e-05, "loss": 0.0153, "step": 44260 }, { "grad_norm": 0.1959448605775833, "learning_rate": 6.344227501685984e-05, "loss": 0.0139, "step": 44270 }, { "grad_norm": 0.3123486638069153, "learning_rate": 6.342634833674572e-05, "loss": 0.0128, "step": 44280 }, { "grad_norm": 0.21371692419052124, "learning_rate": 6.341042018834635e-05, "loss": 0.0125, "step": 44290 }, { "grad_norm": 0.22619332373142242, "learning_rate": 6.339449057340359e-05, "loss": 0.014, "step": 44300 }, { "grad_norm": 0.24378995597362518, "learning_rate": 6.337855949365945e-05, "loss": 0.015, "step": 44310 }, { "grad_norm": 0.18822520971298218, "learning_rate": 6.336262695085619e-05, "loss": 0.0104, "step": 44320 }, { "grad_norm": 0.2664892077445984, "learning_rate": 6.334669294673612e-05, "loss": 0.0133, "step": 44330 }, { "grad_norm": 0.16981291770935059, "learning_rate": 6.333075748304179e-05, "loss": 0.0129, "step": 44340 }, { "grad_norm": 0.19478729367256165, "learning_rate": 6.331482056151585e-05, "loss": 0.0118, "step": 44350 }, { "grad_norm": 0.35242965817451477, "learning_rate": 6.329888218390117e-05, "loss": 0.0117, "step": 44360 }, { "grad_norm": 0.11152566224336624, "learning_rate": 6.328294235194072e-05, "loss": 0.014, "step": 44370 }, { "grad_norm": 0.15617190301418304, "learning_rate": 6.326700106737765e-05, "loss": 0.0086, "step": 44380 }, { "grad_norm": 0.25351813435554504, "learning_rate": 6.325105833195531e-05, "loss": 0.0091, "step": 44390 }, { "grad_norm": 0.25821563601493835, "learning_rate": 6.323511414741715e-05, "loss": 0.0103, "step": 44400 }, { "grad_norm": 0.19787420332431793, "learning_rate": 6.321916851550678e-05, "loss": 0.0108, "step": 44410 }, { "grad_norm": 0.16020305454730988, "learning_rate": 6.320322143796806e-05, "loss": 0.015, "step": 44420 }, { "grad_norm": 0.28660857677459717, "learning_rate": 6.318727291654488e-05, "loss": 0.0111, "step": 44430 }, { "grad_norm": 0.17128050327301025, "learning_rate": 6.317132295298134e-05, "loss": 0.0088, "step": 44440 }, { "grad_norm": 0.1691237986087799, "learning_rate": 6.315537154902173e-05, "loss": 0.011, "step": 44450 }, { "grad_norm": 0.14854437112808228, "learning_rate": 6.313941870641048e-05, "loss": 0.0076, "step": 44460 }, { "grad_norm": 0.20780234038829803, "learning_rate": 6.312346442689214e-05, "loss": 0.009, "step": 44470 }, { "grad_norm": 0.1935044676065445, "learning_rate": 6.310750871221147e-05, "loss": 0.0088, "step": 44480 }, { "grad_norm": 0.14941562712192535, "learning_rate": 6.309155156411335e-05, "loss": 0.0096, "step": 44490 }, { "grad_norm": 0.7267181277275085, "learning_rate": 6.307559298434284e-05, "loss": 0.0107, "step": 44500 }, { "grad_norm": 0.16662254929542542, "learning_rate": 6.305963297464512e-05, "loss": 0.0092, "step": 44510 }, { "grad_norm": 0.17597435414791107, "learning_rate": 6.304367153676561e-05, "loss": 0.0108, "step": 44520 }, { "grad_norm": 0.17115455865859985, "learning_rate": 6.302770867244978e-05, "loss": 0.013, "step": 44530 }, { "grad_norm": 0.16551414132118225, "learning_rate": 6.301174438344328e-05, "loss": 0.0119, "step": 44540 }, { "grad_norm": 0.15711461007595062, "learning_rate": 6.299577867149202e-05, "loss": 0.0122, "step": 44550 }, { "grad_norm": 0.7513746023178101, "learning_rate": 6.297981153834192e-05, "loss": 0.015, "step": 44560 }, { "grad_norm": 0.16312170028686523, "learning_rate": 6.296384298573916e-05, "loss": 0.0143, "step": 44570 }, { "grad_norm": 0.1967903971672058, "learning_rate": 6.294787301543001e-05, "loss": 0.01, "step": 44580 }, { "grad_norm": 0.123384989798069, "learning_rate": 6.293190162916095e-05, "loss": 0.0124, "step": 44590 }, { "grad_norm": 0.3534413278102875, "learning_rate": 6.291592882867855e-05, "loss": 0.0125, "step": 44600 }, { "grad_norm": 0.23965205252170563, "learning_rate": 6.28999546157296e-05, "loss": 0.0121, "step": 44610 }, { "grad_norm": 0.2340671867132187, "learning_rate": 6.288397899206102e-05, "loss": 0.0124, "step": 44620 }, { "grad_norm": 0.160399928689003, "learning_rate": 6.286800195941984e-05, "loss": 0.0097, "step": 44630 }, { "grad_norm": 0.13932503759860992, "learning_rate": 6.285202351955334e-05, "loss": 0.0101, "step": 44640 }, { "grad_norm": 0.17237259447574615, "learning_rate": 6.283604367420887e-05, "loss": 0.0126, "step": 44650 }, { "grad_norm": 0.2597403824329376, "learning_rate": 6.282006242513394e-05, "loss": 0.0136, "step": 44660 }, { "grad_norm": 0.3030804395675659, "learning_rate": 6.280407977407628e-05, "loss": 0.0132, "step": 44670 }, { "grad_norm": 0.18838974833488464, "learning_rate": 6.27880957227837e-05, "loss": 0.0107, "step": 44680 }, { "grad_norm": 0.2924693524837494, "learning_rate": 6.27721102730042e-05, "loss": 0.0112, "step": 44690 }, { "grad_norm": 0.2096432000398636, "learning_rate": 6.275612342648592e-05, "loss": 0.0198, "step": 44700 }, { "grad_norm": 0.19784486293792725, "learning_rate": 6.274013518497716e-05, "loss": 0.0115, "step": 44710 }, { "grad_norm": 0.20021125674247742, "learning_rate": 6.272414555022636e-05, "loss": 0.0104, "step": 44720 }, { "grad_norm": 0.15786851942539215, "learning_rate": 6.270815452398215e-05, "loss": 0.012, "step": 44730 }, { "grad_norm": 0.14132723212242126, "learning_rate": 6.269216210799326e-05, "loss": 0.0118, "step": 44740 }, { "grad_norm": 0.20049655437469482, "learning_rate": 6.26761683040086e-05, "loss": 0.0117, "step": 44750 }, { "grad_norm": 0.2582985758781433, "learning_rate": 6.266017311377723e-05, "loss": 0.0104, "step": 44760 }, { "grad_norm": 2.053363561630249, "learning_rate": 6.264417653904839e-05, "loss": 0.0114, "step": 44770 }, { "grad_norm": 0.15742816030979156, "learning_rate": 6.262817858157139e-05, "loss": 0.0117, "step": 44780 }, { "grad_norm": 0.1859554946422577, "learning_rate": 6.261217924309576e-05, "loss": 0.0109, "step": 44790 }, { "grad_norm": 0.1855539083480835, "learning_rate": 6.259617852537118e-05, "loss": 0.0097, "step": 44800 }, { "grad_norm": 0.12002991139888763, "learning_rate": 6.258017643014747e-05, "loss": 0.0103, "step": 44810 }, { "grad_norm": 0.24273765087127686, "learning_rate": 6.256417295917456e-05, "loss": 0.0117, "step": 44820 }, { "grad_norm": 0.12003897875547409, "learning_rate": 6.254816811420258e-05, "loss": 0.0113, "step": 44830 }, { "grad_norm": 0.12210544943809509, "learning_rate": 6.253216189698183e-05, "loss": 0.0094, "step": 44840 }, { "grad_norm": 0.12846863269805908, "learning_rate": 6.251615430926267e-05, "loss": 0.0127, "step": 44850 }, { "grad_norm": 0.14341381192207336, "learning_rate": 6.25001453527957e-05, "loss": 0.0131, "step": 44860 }, { "grad_norm": 0.18763737380504608, "learning_rate": 6.248413502933164e-05, "loss": 0.0173, "step": 44870 }, { "grad_norm": 0.2033100724220276, "learning_rate": 6.246812334062133e-05, "loss": 0.0091, "step": 44880 }, { "grad_norm": 0.2540738582611084, "learning_rate": 6.245211028841579e-05, "loss": 0.014, "step": 44890 }, { "grad_norm": 0.15580126643180847, "learning_rate": 6.24360958744662e-05, "loss": 0.0125, "step": 44900 }, { "grad_norm": 0.23786434531211853, "learning_rate": 6.242008010052387e-05, "loss": 0.0124, "step": 44910 }, { "grad_norm": 0.6198393702507019, "learning_rate": 6.240406296834024e-05, "loss": 0.0121, "step": 44920 }, { "grad_norm": 0.5036019682884216, "learning_rate": 6.238804447966694e-05, "loss": 0.0102, "step": 44930 }, { "grad_norm": 0.408689945936203, "learning_rate": 6.237202463625573e-05, "loss": 0.0145, "step": 44940 }, { "grad_norm": 0.18683302402496338, "learning_rate": 6.235600343985848e-05, "loss": 0.0123, "step": 44950 }, { "grad_norm": 0.22400707006454468, "learning_rate": 6.233998089222729e-05, "loss": 0.0111, "step": 44960 }, { "grad_norm": 0.2810512185096741, "learning_rate": 6.232395699511433e-05, "loss": 0.0097, "step": 44970 }, { "grad_norm": 0.17191337049007416, "learning_rate": 6.230793175027199e-05, "loss": 0.012, "step": 44980 }, { "grad_norm": 0.19397884607315063, "learning_rate": 6.22919051594527e-05, "loss": 0.0147, "step": 44990 }, { "grad_norm": 0.20046299695968628, "learning_rate": 6.227587722440917e-05, "loss": 0.0134, "step": 45000 }, { "grad_norm": 0.17177703976631165, "learning_rate": 6.225984794689414e-05, "loss": 0.0119, "step": 45010 }, { "grad_norm": 0.13398945331573486, "learning_rate": 6.22438173286606e-05, "loss": 0.0146, "step": 45020 }, { "grad_norm": 0.19898122549057007, "learning_rate": 6.222778537146159e-05, "loss": 0.0122, "step": 45030 }, { "grad_norm": 0.19438359141349792, "learning_rate": 6.221175207705037e-05, "loss": 0.0117, "step": 45040 }, { "grad_norm": 0.12068232893943787, "learning_rate": 6.219571744718027e-05, "loss": 0.0104, "step": 45050 }, { "grad_norm": 0.11462627351284027, "learning_rate": 6.217968148360487e-05, "loss": 0.0111, "step": 45060 }, { "grad_norm": 0.6969968676567078, "learning_rate": 6.216364418807782e-05, "loss": 0.0115, "step": 45070 }, { "grad_norm": 0.5585520267486572, "learning_rate": 6.214760556235292e-05, "loss": 0.0122, "step": 45080 }, { "grad_norm": 0.15498106181621552, "learning_rate": 6.213156560818414e-05, "loss": 0.015, "step": 45090 }, { "grad_norm": 0.14791356027126312, "learning_rate": 6.211552432732559e-05, "loss": 0.0092, "step": 45100 }, { "grad_norm": 0.18388107419013977, "learning_rate": 6.20994817215315e-05, "loss": 0.0115, "step": 45110 }, { "grad_norm": 0.15451963245868683, "learning_rate": 6.208343779255628e-05, "loss": 0.0103, "step": 45120 }, { "grad_norm": 0.13717100024223328, "learning_rate": 6.206739254215449e-05, "loss": 0.0109, "step": 45130 }, { "grad_norm": 0.15180975198745728, "learning_rate": 6.205134597208077e-05, "loss": 0.009, "step": 45140 }, { "grad_norm": 0.1572180688381195, "learning_rate": 6.203529808409e-05, "loss": 0.0109, "step": 45150 }, { "grad_norm": 0.16415655612945557, "learning_rate": 6.201924887993712e-05, "loss": 0.0108, "step": 45160 }, { "grad_norm": 0.18743830919265747, "learning_rate": 6.200319836137725e-05, "loss": 0.0116, "step": 45170 }, { "grad_norm": 0.15471431612968445, "learning_rate": 6.198714653016565e-05, "loss": 0.0135, "step": 45180 }, { "grad_norm": 0.2172362506389618, "learning_rate": 6.197109338805774e-05, "loss": 0.0115, "step": 45190 }, { "grad_norm": 0.14691387116909027, "learning_rate": 6.195503893680903e-05, "loss": 0.0102, "step": 45200 }, { "grad_norm": 0.15891650319099426, "learning_rate": 6.193898317817524e-05, "loss": 0.0136, "step": 45210 }, { "grad_norm": 0.1717831790447235, "learning_rate": 6.192292611391221e-05, "loss": 0.0111, "step": 45220 }, { "grad_norm": 0.15237946808338165, "learning_rate": 6.190686774577591e-05, "loss": 0.0116, "step": 45230 }, { "grad_norm": 0.1880779266357422, "learning_rate": 6.189080807552245e-05, "loss": 0.011, "step": 45240 }, { "grad_norm": 0.1736827939748764, "learning_rate": 6.187474710490809e-05, "loss": 0.0108, "step": 45250 }, { "grad_norm": 0.24055151641368866, "learning_rate": 6.185868483568926e-05, "loss": 0.0097, "step": 45260 }, { "grad_norm": 0.14071498811244965, "learning_rate": 6.184262126962245e-05, "loss": 0.0111, "step": 45270 }, { "grad_norm": 0.36749935150146484, "learning_rate": 6.182655640846442e-05, "loss": 0.0133, "step": 45280 }, { "grad_norm": 0.1698015034198761, "learning_rate": 6.181049025397196e-05, "loss": 0.0126, "step": 45290 }, { "grad_norm": 0.16571389138698578, "learning_rate": 6.179442280790202e-05, "loss": 0.0121, "step": 45300 }, { "grad_norm": 0.1592601239681244, "learning_rate": 6.177835407201174e-05, "loss": 0.0125, "step": 45310 }, { "grad_norm": 0.19774173200130463, "learning_rate": 6.176228404805839e-05, "loss": 0.0108, "step": 45320 }, { "grad_norm": 0.19336813688278198, "learning_rate": 6.174621273779932e-05, "loss": 0.0133, "step": 45330 }, { "grad_norm": 0.14175847172737122, "learning_rate": 6.17301401429921e-05, "loss": 0.0112, "step": 45340 }, { "grad_norm": 0.16024702787399292, "learning_rate": 6.171406626539436e-05, "loss": 0.0104, "step": 45350 }, { "grad_norm": 0.86105877161026, "learning_rate": 6.169799110676398e-05, "loss": 0.0145, "step": 45360 }, { "grad_norm": 0.3029477596282959, "learning_rate": 6.168191466885885e-05, "loss": 0.011, "step": 45370 }, { "grad_norm": 0.16922660171985626, "learning_rate": 6.166583695343711e-05, "loss": 0.0134, "step": 45380 }, { "grad_norm": 0.2156543731689453, "learning_rate": 6.164975796225698e-05, "loss": 0.0121, "step": 45390 }, { "grad_norm": 0.1485009640455246, "learning_rate": 6.163367769707683e-05, "loss": 0.0133, "step": 45400 }, { "grad_norm": 0.27913230657577515, "learning_rate": 6.161759615965519e-05, "loss": 0.014, "step": 45410 }, { "grad_norm": 0.18668705224990845, "learning_rate": 6.16015133517507e-05, "loss": 0.0142, "step": 45420 }, { "grad_norm": 0.11551682651042938, "learning_rate": 6.158542927512214e-05, "loss": 0.0153, "step": 45430 }, { "grad_norm": 1.0563642978668213, "learning_rate": 6.156934393152846e-05, "loss": 0.0104, "step": 45440 }, { "grad_norm": 0.2104872316122055, "learning_rate": 6.15532573227287e-05, "loss": 0.0152, "step": 45450 }, { "grad_norm": 0.13016334176063538, "learning_rate": 6.153716945048212e-05, "loss": 0.0123, "step": 45460 }, { "grad_norm": 0.15394268929958344, "learning_rate": 6.152108031654802e-05, "loss": 0.0098, "step": 45470 }, { "grad_norm": 0.341384619474411, "learning_rate": 6.15049899226859e-05, "loss": 0.0109, "step": 45480 }, { "grad_norm": 0.14171601831912994, "learning_rate": 6.148889827065537e-05, "loss": 0.0097, "step": 45490 }, { "grad_norm": 1.522369384765625, "learning_rate": 6.147280536221622e-05, "loss": 0.0142, "step": 45500 }, { "grad_norm": 0.23321567475795746, "learning_rate": 6.145671119912832e-05, "loss": 0.0113, "step": 45510 }, { "grad_norm": 0.1753268837928772, "learning_rate": 6.144061578315169e-05, "loss": 0.0113, "step": 45520 }, { "grad_norm": 0.2202361524105072, "learning_rate": 6.142451911604654e-05, "loss": 0.0122, "step": 45530 }, { "grad_norm": 0.13022449612617493, "learning_rate": 6.140842119957315e-05, "loss": 0.0091, "step": 45540 }, { "grad_norm": 0.159070685505867, "learning_rate": 6.139232203549197e-05, "loss": 0.0128, "step": 45550 }, { "grad_norm": 0.15522053837776184, "learning_rate": 6.137622162556357e-05, "loss": 0.0103, "step": 45560 }, { "grad_norm": 0.2720797061920166, "learning_rate": 6.13601199715487e-05, "loss": 0.0136, "step": 45570 }, { "grad_norm": 0.20489837229251862, "learning_rate": 6.134401707520816e-05, "loss": 0.0142, "step": 45580 }, { "grad_norm": 0.19888371229171753, "learning_rate": 6.132791293830298e-05, "loss": 0.0086, "step": 45590 }, { "grad_norm": 0.21578346192836761, "learning_rate": 6.131180756259428e-05, "loss": 0.011, "step": 45600 }, { "grad_norm": 0.1693308800458908, "learning_rate": 6.129570094984331e-05, "loss": 0.013, "step": 45610 }, { "grad_norm": 0.1434796005487442, "learning_rate": 6.127959310181145e-05, "loss": 0.0106, "step": 45620 }, { "grad_norm": 0.17316582798957825, "learning_rate": 6.126348402026026e-05, "loss": 0.0107, "step": 45630 }, { "grad_norm": 0.13239452242851257, "learning_rate": 6.12473737069514e-05, "loss": 0.0082, "step": 45640 }, { "grad_norm": 0.2857217788696289, "learning_rate": 6.123126216364665e-05, "loss": 0.0074, "step": 45650 }, { "grad_norm": 0.3552144765853882, "learning_rate": 6.121514939210797e-05, "loss": 0.0077, "step": 45660 }, { "grad_norm": 0.1709095984697342, "learning_rate": 6.119903539409741e-05, "loss": 0.012, "step": 45670 }, { "grad_norm": 0.14171813428401947, "learning_rate": 6.118292017137716e-05, "loss": 0.0079, "step": 45680 }, { "grad_norm": 0.18488556146621704, "learning_rate": 6.116680372570959e-05, "loss": 0.0084, "step": 45690 }, { "grad_norm": 0.18789274990558624, "learning_rate": 6.115068605885713e-05, "loss": 0.0129, "step": 45700 }, { "grad_norm": 0.16347330808639526, "learning_rate": 6.113456717258243e-05, "loss": 0.0121, "step": 45710 }, { "grad_norm": 0.45543843507766724, "learning_rate": 6.11184470686482e-05, "loss": 0.0092, "step": 45720 }, { "grad_norm": 0.13188901543617249, "learning_rate": 6.110232574881731e-05, "loss": 0.0085, "step": 45730 }, { "grad_norm": 0.17999348044395447, "learning_rate": 6.108620321485277e-05, "loss": 0.0099, "step": 45740 }, { "grad_norm": 0.1672390103340149, "learning_rate": 6.107007946851773e-05, "loss": 0.0088, "step": 45750 }, { "grad_norm": 0.13423052430152893, "learning_rate": 6.105395451157542e-05, "loss": 0.0079, "step": 45760 }, { "grad_norm": 0.127106711268425, "learning_rate": 6.103782834578928e-05, "loss": 0.0083, "step": 45770 }, { "grad_norm": 0.3091249465942383, "learning_rate": 6.102170097292281e-05, "loss": 0.0106, "step": 45780 }, { "grad_norm": 0.13555210828781128, "learning_rate": 6.1005572394739716e-05, "loss": 0.0093, "step": 45790 }, { "grad_norm": 0.1307869553565979, "learning_rate": 6.0989442613003765e-05, "loss": 0.0115, "step": 45800 }, { "grad_norm": 0.1184900626540184, "learning_rate": 6.097331162947888e-05, "loss": 0.011, "step": 45810 }, { "grad_norm": 0.13002166152000427, "learning_rate": 6.095717944592914e-05, "loss": 0.009, "step": 45820 }, { "grad_norm": 0.1927500069141388, "learning_rate": 6.094104606411873e-05, "loss": 0.009, "step": 45830 }, { "grad_norm": 0.12969888746738434, "learning_rate": 6.0924911485811966e-05, "loss": 0.011, "step": 45840 }, { "grad_norm": 0.18420198559761047, "learning_rate": 6.090877571277331e-05, "loss": 0.0102, "step": 45850 }, { "grad_norm": 0.1249442994594574, "learning_rate": 6.089263874676736e-05, "loss": 0.0123, "step": 45860 }, { "grad_norm": 0.18532618880271912, "learning_rate": 6.0876500589558796e-05, "loss": 0.0125, "step": 45870 }, { "grad_norm": 0.36608609557151794, "learning_rate": 6.086036124291248e-05, "loss": 0.0089, "step": 45880 }, { "grad_norm": 0.14166681468486786, "learning_rate": 6.084422070859339e-05, "loss": 0.0117, "step": 45890 }, { "grad_norm": 0.11297009885311127, "learning_rate": 6.082807898836663e-05, "loss": 0.0103, "step": 45900 }, { "grad_norm": 0.1486792117357254, "learning_rate": 6.081193608399742e-05, "loss": 0.0098, "step": 45910 }, { "grad_norm": 0.13021573424339294, "learning_rate": 6.0795791997251164e-05, "loss": 0.0144, "step": 45920 }, { "grad_norm": 0.12789283692836761, "learning_rate": 6.0779646729893294e-05, "loss": 0.0081, "step": 45930 }, { "grad_norm": 0.13830947875976562, "learning_rate": 6.0763500283689476e-05, "loss": 0.0115, "step": 45940 }, { "grad_norm": 0.20002508163452148, "learning_rate": 6.0747352660405455e-05, "loss": 0.0111, "step": 45950 }, { "grad_norm": 0.22381719946861267, "learning_rate": 6.073120386180709e-05, "loss": 0.0139, "step": 45960 }, { "grad_norm": 0.15408113598823547, "learning_rate": 6.0715053889660425e-05, "loss": 0.0117, "step": 45970 }, { "grad_norm": 0.19579239189624786, "learning_rate": 6.069890274573157e-05, "loss": 0.0099, "step": 45980 }, { "grad_norm": 0.1948324739933014, "learning_rate": 6.068275043178679e-05, "loss": 0.013, "step": 45990 }, { "grad_norm": 0.1726461946964264, "learning_rate": 6.066659694959248e-05, "loss": 0.0087, "step": 46000 }, { "grad_norm": 0.15454798936843872, "learning_rate": 6.0650442300915176e-05, "loss": 0.0123, "step": 46010 }, { "grad_norm": 0.20182594656944275, "learning_rate": 6.063428648752152e-05, "loss": 0.0101, "step": 46020 }, { "grad_norm": 0.1898653656244278, "learning_rate": 6.0618129511178266e-05, "loss": 0.0114, "step": 46030 }, { "grad_norm": 0.1266120821237564, "learning_rate": 6.060197137365234e-05, "loss": 0.0103, "step": 46040 }, { "grad_norm": 0.2749675214290619, "learning_rate": 6.058581207671077e-05, "loss": 0.011, "step": 46050 }, { "grad_norm": 0.11559759825468063, "learning_rate": 6.056965162212072e-05, "loss": 0.0129, "step": 46060 }, { "grad_norm": 0.1605621874332428, "learning_rate": 6.055349001164943e-05, "loss": 0.0105, "step": 46070 }, { "grad_norm": 0.16100478172302246, "learning_rate": 6.0537327247064347e-05, "loss": 0.0112, "step": 46080 }, { "grad_norm": 0.2505899667739868, "learning_rate": 6.0521163330133014e-05, "loss": 0.0113, "step": 46090 }, { "grad_norm": 0.15594547986984253, "learning_rate": 6.050499826262306e-05, "loss": 0.0107, "step": 46100 }, { "grad_norm": 0.17735056579113007, "learning_rate": 6.0488832046302294e-05, "loss": 0.0102, "step": 46110 }, { "grad_norm": 0.16444432735443115, "learning_rate": 6.0472664682938626e-05, "loss": 0.0079, "step": 46120 }, { "grad_norm": 0.16216036677360535, "learning_rate": 6.045649617430009e-05, "loss": 0.0112, "step": 46130 }, { "grad_norm": 0.38000673055648804, "learning_rate": 6.0440326522154866e-05, "loss": 0.0114, "step": 46140 }, { "grad_norm": 0.1688133329153061, "learning_rate": 6.0424155728271224e-05, "loss": 0.0089, "step": 46150 }, { "grad_norm": 0.6408368945121765, "learning_rate": 6.040798379441758e-05, "loss": 0.0095, "step": 46160 }, { "grad_norm": 0.17611081898212433, "learning_rate": 6.0391810722362485e-05, "loss": 0.0146, "step": 46170 }, { "grad_norm": 0.545668363571167, "learning_rate": 6.037563651387458e-05, "loss": 0.0122, "step": 46180 }, { "grad_norm": 0.16605405509471893, "learning_rate": 6.0359461170722666e-05, "loss": 0.0103, "step": 46190 }, { "grad_norm": 0.21778149902820587, "learning_rate": 6.034328469467566e-05, "loss": 0.0125, "step": 46200 }, { "grad_norm": 0.2821521759033203, "learning_rate": 6.0327107087502596e-05, "loss": 0.0149, "step": 46210 }, { "grad_norm": 0.20166504383087158, "learning_rate": 6.031092835097262e-05, "loss": 0.0118, "step": 46220 }, { "grad_norm": 0.26503413915634155, "learning_rate": 6.0294748486855024e-05, "loss": 0.0112, "step": 46230 }, { "grad_norm": 0.28012216091156006, "learning_rate": 6.0278567496919216e-05, "loss": 0.0094, "step": 46240 }, { "grad_norm": 0.27148354053497314, "learning_rate": 6.026238538293472e-05, "loss": 0.0115, "step": 46250 }, { "grad_norm": 0.13765886425971985, "learning_rate": 6.024620214667118e-05, "loss": 0.0089, "step": 46260 }, { "grad_norm": 0.13095858693122864, "learning_rate": 6.0230017789898384e-05, "loss": 0.0102, "step": 46270 }, { "grad_norm": 0.11555904150009155, "learning_rate": 6.021383231438622e-05, "loss": 0.0096, "step": 46280 }, { "grad_norm": 0.16692902147769928, "learning_rate": 6.0197645721904704e-05, "loss": 0.0112, "step": 46290 }, { "grad_norm": 0.13095958530902863, "learning_rate": 6.0181458014224e-05, "loss": 0.0118, "step": 46300 }, { "grad_norm": 1.2112534046173096, "learning_rate": 6.016526919311435e-05, "loss": 0.0114, "step": 46310 }, { "grad_norm": 0.1710587590932846, "learning_rate": 6.014907926034613e-05, "loss": 0.0099, "step": 46320 }, { "grad_norm": 0.16245950758457184, "learning_rate": 6.0132888217689875e-05, "loss": 0.0101, "step": 46330 }, { "grad_norm": 0.12882408499717712, "learning_rate": 6.01166960669162e-05, "loss": 0.0114, "step": 46340 }, { "grad_norm": 0.18423309922218323, "learning_rate": 6.0100502809795845e-05, "loss": 0.0089, "step": 46350 }, { "grad_norm": 0.1935814768075943, "learning_rate": 6.00843084480997e-05, "loss": 0.0134, "step": 46360 }, { "grad_norm": 0.14638860523700714, "learning_rate": 6.0068112983598736e-05, "loss": 0.0103, "step": 46370 }, { "grad_norm": 0.17270435392856598, "learning_rate": 6.005191641806407e-05, "loss": 0.0094, "step": 46380 }, { "grad_norm": 0.19051885604858398, "learning_rate": 6.003571875326694e-05, "loss": 0.0124, "step": 46390 }, { "grad_norm": 0.12929797172546387, "learning_rate": 6.00195199909787e-05, "loss": 0.012, "step": 46400 }, { "grad_norm": 0.15447598695755005, "learning_rate": 6.000332013297082e-05, "loss": 0.0134, "step": 46410 }, { "grad_norm": 0.13802888989448547, "learning_rate": 5.998711918101487e-05, "loss": 0.0133, "step": 46420 }, { "grad_norm": 0.13384144008159637, "learning_rate": 5.997091713688261e-05, "loss": 0.0097, "step": 46430 }, { "grad_norm": 0.1661861389875412, "learning_rate": 5.9954714002345836e-05, "loss": 0.0127, "step": 46440 }, { "grad_norm": 0.8153813481330872, "learning_rate": 5.993850977917649e-05, "loss": 0.0106, "step": 46450 }, { "grad_norm": 0.15524989366531372, "learning_rate": 5.992230446914667e-05, "loss": 0.011, "step": 46460 }, { "grad_norm": 0.12919095158576965, "learning_rate": 5.990609807402855e-05, "loss": 0.01, "step": 46470 }, { "grad_norm": 0.1587669998407364, "learning_rate": 5.988989059559443e-05, "loss": 0.0114, "step": 46480 }, { "grad_norm": 0.17087037861347198, "learning_rate": 5.987368203561675e-05, "loss": 0.0119, "step": 46490 }, { "grad_norm": 0.1330178827047348, "learning_rate": 5.9857472395868055e-05, "loss": 0.0097, "step": 46500 }, { "grad_norm": 0.17555201053619385, "learning_rate": 5.9841261678120983e-05, "loss": 0.01, "step": 46510 }, { "grad_norm": 0.4010428488254547, "learning_rate": 5.982504988414834e-05, "loss": 0.0149, "step": 46520 }, { "grad_norm": 1.6820999383926392, "learning_rate": 5.9808837015723015e-05, "loss": 0.0118, "step": 46530 }, { "grad_norm": 0.7898386120796204, "learning_rate": 5.9792623074618016e-05, "loss": 0.0114, "step": 46540 }, { "grad_norm": 0.18820853531360626, "learning_rate": 5.977640806260648e-05, "loss": 0.0107, "step": 46550 }, { "grad_norm": 0.3586837649345398, "learning_rate": 5.976019198146166e-05, "loss": 0.0134, "step": 46560 }, { "grad_norm": 0.13068333268165588, "learning_rate": 5.974397483295692e-05, "loss": 0.0104, "step": 46570 }, { "grad_norm": 0.15415804088115692, "learning_rate": 5.972775661886575e-05, "loss": 0.0101, "step": 46580 }, { "grad_norm": 0.11610721796751022, "learning_rate": 5.971153734096173e-05, "loss": 0.0097, "step": 46590 }, { "grad_norm": 0.10430644452571869, "learning_rate": 5.969531700101859e-05, "loss": 0.0115, "step": 46600 }, { "grad_norm": 0.4553346335887909, "learning_rate": 5.9679095600810155e-05, "loss": 0.0141, "step": 46610 }, { "grad_norm": 0.21384131908416748, "learning_rate": 5.9662873142110384e-05, "loss": 0.0121, "step": 46620 }, { "grad_norm": 0.1959184855222702, "learning_rate": 5.964664962669333e-05, "loss": 0.0099, "step": 46630 }, { "grad_norm": 0.15202316641807556, "learning_rate": 5.9630425056333186e-05, "loss": 0.0125, "step": 46640 }, { "grad_norm": 0.21176597476005554, "learning_rate": 5.961419943280422e-05, "loss": 0.0117, "step": 46650 }, { "grad_norm": 0.19466610252857208, "learning_rate": 5.959797275788087e-05, "loss": 0.01, "step": 46660 }, { "grad_norm": 0.18289832770824432, "learning_rate": 5.958174503333765e-05, "loss": 0.0122, "step": 46670 }, { "grad_norm": 0.16750289499759674, "learning_rate": 5.9565516260949195e-05, "loss": 0.0103, "step": 46680 }, { "grad_norm": 0.13478174805641174, "learning_rate": 5.9549286442490273e-05, "loss": 0.0073, "step": 46690 }, { "grad_norm": 0.10846947133541107, "learning_rate": 5.953305557973572e-05, "loss": 0.0082, "step": 46700 }, { "grad_norm": 0.1562773436307907, "learning_rate": 5.9516823674460564e-05, "loss": 0.0127, "step": 46710 }, { "grad_norm": 0.12406516075134277, "learning_rate": 5.9500590728439875e-05, "loss": 0.0101, "step": 46720 }, { "grad_norm": 0.15856032073497772, "learning_rate": 5.948435674344886e-05, "loss": 0.0093, "step": 46730 }, { "grad_norm": 0.1376519352197647, "learning_rate": 5.946812172126285e-05, "loss": 0.0115, "step": 46740 }, { "grad_norm": 0.1916225254535675, "learning_rate": 5.945188566365729e-05, "loss": 0.0117, "step": 46750 }, { "grad_norm": 0.12127090245485306, "learning_rate": 5.943564857240773e-05, "loss": 0.0112, "step": 46760 }, { "grad_norm": 0.5111290216445923, "learning_rate": 5.941941044928981e-05, "loss": 0.0127, "step": 46770 }, { "grad_norm": 0.13261426985263824, "learning_rate": 5.940317129607935e-05, "loss": 0.0093, "step": 46780 }, { "grad_norm": 0.16879355907440186, "learning_rate": 5.9386931114552204e-05, "loss": 0.0081, "step": 46790 }, { "grad_norm": 0.20328417420387268, "learning_rate": 5.937068990648438e-05, "loss": 0.0115, "step": 46800 }, { "grad_norm": 0.1637757420539856, "learning_rate": 5.935444767365199e-05, "loss": 0.0093, "step": 46810 }, { "grad_norm": 0.1571579873561859, "learning_rate": 5.933820441783129e-05, "loss": 0.0107, "step": 46820 }, { "grad_norm": 0.1923513561487198, "learning_rate": 5.932196014079857e-05, "loss": 0.0106, "step": 46830 }, { "grad_norm": 0.17463268339633942, "learning_rate": 5.930571484433032e-05, "loss": 0.0085, "step": 46840 }, { "grad_norm": 0.1740446537733078, "learning_rate": 5.928946853020309e-05, "loss": 0.014, "step": 46850 }, { "grad_norm": 0.18397490680217743, "learning_rate": 5.927322120019352e-05, "loss": 0.0108, "step": 46860 }, { "grad_norm": 0.1220329999923706, "learning_rate": 5.925697285607844e-05, "loss": 0.011, "step": 46870 }, { "grad_norm": 0.14345146715641022, "learning_rate": 5.924072349963472e-05, "loss": 0.008, "step": 46880 }, { "grad_norm": 0.14101649820804596, "learning_rate": 5.922447313263938e-05, "loss": 0.0074, "step": 46890 }, { "grad_norm": 0.41095009446144104, "learning_rate": 5.920822175686952e-05, "loss": 0.0095, "step": 46900 }, { "grad_norm": 0.15000396966934204, "learning_rate": 5.919196937410239e-05, "loss": 0.0071, "step": 46910 }, { "grad_norm": 0.11356857419013977, "learning_rate": 5.917571598611529e-05, "loss": 0.0101, "step": 46920 }, { "grad_norm": 0.14905859529972076, "learning_rate": 5.9159461594685686e-05, "loss": 0.0139, "step": 46930 }, { "grad_norm": 0.15557636320590973, "learning_rate": 5.914320620159114e-05, "loss": 0.0108, "step": 46940 }, { "grad_norm": 0.19455529749393463, "learning_rate": 5.912694980860931e-05, "loss": 0.0106, "step": 46950 }, { "grad_norm": 0.13179895281791687, "learning_rate": 5.9110692417517964e-05, "loss": 0.0095, "step": 46960 }, { "grad_norm": 0.11340802907943726, "learning_rate": 5.9094434030095e-05, "loss": 0.0103, "step": 46970 }, { "grad_norm": 0.1835087239742279, "learning_rate": 5.907817464811841e-05, "loss": 0.0098, "step": 46980 }, { "grad_norm": 0.1354680359363556, "learning_rate": 5.906191427336627e-05, "loss": 0.009, "step": 46990 }, { "grad_norm": 0.29390332102775574, "learning_rate": 5.9045652907616835e-05, "loss": 0.011, "step": 47000 }, { "grad_norm": 0.12741824984550476, "learning_rate": 5.902939055264838e-05, "loss": 0.0123, "step": 47010 }, { "grad_norm": 0.21595455706119537, "learning_rate": 5.901312721023935e-05, "loss": 0.0109, "step": 47020 }, { "grad_norm": 0.2431575506925583, "learning_rate": 5.8996862882168294e-05, "loss": 0.0106, "step": 47030 }, { "grad_norm": 0.17017924785614014, "learning_rate": 5.8980597570213826e-05, "loss": 0.0105, "step": 47040 }, { "grad_norm": 0.16164430975914001, "learning_rate": 5.896433127615471e-05, "loss": 0.0141, "step": 47050 }, { "grad_norm": 0.17759352922439575, "learning_rate": 5.894806400176981e-05, "loss": 0.0101, "step": 47060 }, { "grad_norm": 0.1993110477924347, "learning_rate": 5.893179574883808e-05, "loss": 0.0123, "step": 47070 }, { "grad_norm": 0.4228001534938812, "learning_rate": 5.8915526519138585e-05, "loss": 0.0136, "step": 47080 }, { "grad_norm": 0.16345259547233582, "learning_rate": 5.889925631445053e-05, "loss": 0.0128, "step": 47090 }, { "grad_norm": 0.12886708974838257, "learning_rate": 5.888298513655318e-05, "loss": 0.0084, "step": 47100 }, { "grad_norm": 0.20676855742931366, "learning_rate": 5.886671298722592e-05, "loss": 0.0167, "step": 47110 }, { "grad_norm": 0.21896444261074066, "learning_rate": 5.885043986824828e-05, "loss": 0.0123, "step": 47120 }, { "grad_norm": 0.12579920887947083, "learning_rate": 5.883416578139982e-05, "loss": 0.0123, "step": 47130 }, { "grad_norm": 0.21946071088314056, "learning_rate": 5.881789072846029e-05, "loss": 0.0133, "step": 47140 }, { "grad_norm": 0.1974419802427292, "learning_rate": 5.8801614711209474e-05, "loss": 0.0093, "step": 47150 }, { "grad_norm": 0.1946006566286087, "learning_rate": 5.878533773142732e-05, "loss": 0.0108, "step": 47160 }, { "grad_norm": 0.12169280648231506, "learning_rate": 5.8769059790893834e-05, "loss": 0.0128, "step": 47170 }, { "grad_norm": 0.13861551880836487, "learning_rate": 5.8752780891389145e-05, "loss": 0.0092, "step": 47180 }, { "grad_norm": 0.16648460924625397, "learning_rate": 5.873650103469351e-05, "loss": 0.0129, "step": 47190 }, { "grad_norm": 0.14479364454746246, "learning_rate": 5.872022022258725e-05, "loss": 0.0086, "step": 47200 }, { "grad_norm": 0.13037800788879395, "learning_rate": 5.8703938456850814e-05, "loss": 0.0099, "step": 47210 }, { "grad_norm": 0.11076918244361877, "learning_rate": 5.868765573926477e-05, "loss": 0.0076, "step": 47220 }, { "grad_norm": 0.17890474200248718, "learning_rate": 5.867137207160974e-05, "loss": 0.0118, "step": 47230 }, { "grad_norm": 0.15564192831516266, "learning_rate": 5.86550874556665e-05, "loss": 0.0095, "step": 47240 }, { "grad_norm": 0.1373807191848755, "learning_rate": 5.863880189321592e-05, "loss": 0.0114, "step": 47250 }, { "grad_norm": 0.2972828149795532, "learning_rate": 5.862251538603894e-05, "loss": 0.012, "step": 47260 }, { "grad_norm": 0.15078450739383698, "learning_rate": 5.8606227935916656e-05, "loss": 0.0121, "step": 47270 }, { "grad_norm": 0.129394993185997, "learning_rate": 5.858993954463021e-05, "loss": 0.0082, "step": 47280 }, { "grad_norm": 0.14082224667072296, "learning_rate": 5.8573650213960896e-05, "loss": 0.0083, "step": 47290 }, { "grad_norm": 0.18357595801353455, "learning_rate": 5.8557359945690084e-05, "loss": 0.0099, "step": 47300 }, { "grad_norm": 0.11524554342031479, "learning_rate": 5.8541068741599246e-05, "loss": 0.0122, "step": 47310 }, { "grad_norm": 0.14858479797840118, "learning_rate": 5.8524776603469985e-05, "loss": 0.0104, "step": 47320 }, { "grad_norm": 0.26938357949256897, "learning_rate": 5.850848353308397e-05, "loss": 0.0102, "step": 47330 }, { "grad_norm": 0.19936859607696533, "learning_rate": 5.849218953222297e-05, "loss": 0.0099, "step": 47340 }, { "grad_norm": 0.33138319849967957, "learning_rate": 5.847589460266891e-05, "loss": 0.0161, "step": 47350 }, { "grad_norm": 0.19062566757202148, "learning_rate": 5.845959874620375e-05, "loss": 0.0143, "step": 47360 }, { "grad_norm": 0.18785883486270905, "learning_rate": 5.84433019646096e-05, "loss": 0.0121, "step": 47370 }, { "grad_norm": 0.12579593062400818, "learning_rate": 5.842700425966863e-05, "loss": 0.0094, "step": 47380 }, { "grad_norm": 0.22288304567337036, "learning_rate": 5.841070563316315e-05, "loss": 0.0105, "step": 47390 }, { "grad_norm": 0.1304945945739746, "learning_rate": 5.839440608687554e-05, "loss": 0.0096, "step": 47400 }, { "grad_norm": 0.09797977656126022, "learning_rate": 5.837810562258831e-05, "loss": 0.0077, "step": 47410 }, { "grad_norm": 0.2736712694168091, "learning_rate": 5.836180424208405e-05, "loss": 0.0123, "step": 47420 }, { "grad_norm": 0.17834874987602234, "learning_rate": 5.834550194714543e-05, "loss": 0.0131, "step": 47430 }, { "grad_norm": 0.10747207701206207, "learning_rate": 5.832919873955528e-05, "loss": 0.0073, "step": 47440 }, { "grad_norm": 0.09407423436641693, "learning_rate": 5.831289462109648e-05, "loss": 0.0086, "step": 47450 }, { "grad_norm": 0.13964618742465973, "learning_rate": 5.829658959355202e-05, "loss": 0.0118, "step": 47460 }, { "grad_norm": 0.1685425490140915, "learning_rate": 5.8280283658704995e-05, "loss": 0.0103, "step": 47470 }, { "grad_norm": 0.21621951460838318, "learning_rate": 5.826397681833859e-05, "loss": 0.0108, "step": 47480 }, { "grad_norm": 0.17448469996452332, "learning_rate": 5.824766907423612e-05, "loss": 0.0113, "step": 47490 }, { "grad_norm": 0.1181330606341362, "learning_rate": 5.8231360428180945e-05, "loss": 0.0076, "step": 47500 }, { "grad_norm": 0.9059327244758606, "learning_rate": 5.821505088195658e-05, "loss": 0.0126, "step": 47510 }, { "grad_norm": 0.17430223524570465, "learning_rate": 5.819874043734661e-05, "loss": 0.0103, "step": 47520 }, { "grad_norm": 0.13735392689704895, "learning_rate": 5.8182429096134695e-05, "loss": 0.0072, "step": 47530 }, { "grad_norm": 0.16917617619037628, "learning_rate": 5.816611686010465e-05, "loss": 0.0092, "step": 47540 }, { "grad_norm": 0.15341031551361084, "learning_rate": 5.814980373104033e-05, "loss": 0.0099, "step": 47550 }, { "grad_norm": 0.1642134189605713, "learning_rate": 5.813348971072572e-05, "loss": 0.0093, "step": 47560 }, { "grad_norm": 0.12060987204313278, "learning_rate": 5.811717480094492e-05, "loss": 0.0071, "step": 47570 }, { "grad_norm": 0.12685967981815338, "learning_rate": 5.810085900348209e-05, "loss": 0.0064, "step": 47580 }, { "grad_norm": 0.1527204066514969, "learning_rate": 5.8084542320121483e-05, "loss": 0.0081, "step": 47590 }, { "grad_norm": 0.08994235098361969, "learning_rate": 5.8068224752647497e-05, "loss": 0.0105, "step": 47600 }, { "grad_norm": 0.10621114075183868, "learning_rate": 5.805190630284457e-05, "loss": 0.0081, "step": 47610 }, { "grad_norm": 0.10294607281684875, "learning_rate": 5.803558697249729e-05, "loss": 0.0087, "step": 47620 }, { "grad_norm": 1.1281260251998901, "learning_rate": 5.8019266763390276e-05, "loss": 0.0123, "step": 47630 }, { "grad_norm": 0.147066131234169, "learning_rate": 5.800294567730833e-05, "loss": 0.0099, "step": 47640 }, { "grad_norm": 0.23283085227012634, "learning_rate": 5.798662371603628e-05, "loss": 0.0172, "step": 47650 }, { "grad_norm": 0.19509257376194, "learning_rate": 5.797030088135904e-05, "loss": 0.0114, "step": 47660 }, { "grad_norm": 0.2273392528295517, "learning_rate": 5.7953977175061705e-05, "loss": 0.011, "step": 47670 }, { "grad_norm": 0.1476437747478485, "learning_rate": 5.793765259892938e-05, "loss": 0.0106, "step": 47680 }, { "grad_norm": 5.023258209228516, "learning_rate": 5.792132715474729e-05, "loss": 0.0118, "step": 47690 }, { "grad_norm": 0.19369453191757202, "learning_rate": 5.790500084430078e-05, "loss": 0.0107, "step": 47700 }, { "grad_norm": 0.18673379719257355, "learning_rate": 5.7888673669375274e-05, "loss": 0.0105, "step": 47710 }, { "grad_norm": 0.2048833668231964, "learning_rate": 5.787234563175625e-05, "loss": 0.0122, "step": 47720 }, { "grad_norm": 0.1312127262353897, "learning_rate": 5.7856016733229355e-05, "loss": 0.008, "step": 47730 }, { "grad_norm": 0.12386781722307205, "learning_rate": 5.7839686975580297e-05, "loss": 0.0082, "step": 47740 }, { "grad_norm": 0.21844065189361572, "learning_rate": 5.782335636059484e-05, "loss": 0.0114, "step": 47750 }, { "grad_norm": 0.16235984861850739, "learning_rate": 5.780702489005889e-05, "loss": 0.0097, "step": 47760 }, { "grad_norm": 0.1987197995185852, "learning_rate": 5.779069256575845e-05, "loss": 0.0105, "step": 47770 }, { "grad_norm": 0.14372412860393524, "learning_rate": 5.7774359389479574e-05, "loss": 0.0121, "step": 47780 }, { "grad_norm": 0.21894730627536774, "learning_rate": 5.775802536300845e-05, "loss": 0.0088, "step": 47790 }, { "grad_norm": 0.17878243327140808, "learning_rate": 5.774169048813134e-05, "loss": 0.0163, "step": 47800 }, { "grad_norm": 0.16976141929626465, "learning_rate": 5.77253547666346e-05, "loss": 0.0112, "step": 47810 }, { "grad_norm": 0.16076715290546417, "learning_rate": 5.770901820030465e-05, "loss": 0.0106, "step": 47820 }, { "grad_norm": 0.18152958154678345, "learning_rate": 5.769268079092809e-05, "loss": 0.0079, "step": 47830 }, { "grad_norm": 0.1388883888721466, "learning_rate": 5.767634254029151e-05, "loss": 0.0086, "step": 47840 }, { "grad_norm": 0.15785756707191467, "learning_rate": 5.7660003450181655e-05, "loss": 0.0081, "step": 47850 }, { "grad_norm": 0.15917226672172546, "learning_rate": 5.764366352238534e-05, "loss": 0.0138, "step": 47860 }, { "grad_norm": 0.11033795773983002, "learning_rate": 5.7627322758689474e-05, "loss": 0.0082, "step": 47870 }, { "grad_norm": 0.4690685570240021, "learning_rate": 5.761098116088105e-05, "loss": 0.0185, "step": 47880 }, { "grad_norm": 0.14164577424526215, "learning_rate": 5.759463873074717e-05, "loss": 0.0104, "step": 47890 }, { "grad_norm": 0.20959234237670898, "learning_rate": 5.757829547007504e-05, "loss": 0.0112, "step": 47900 }, { "grad_norm": 0.21723152697086334, "learning_rate": 5.756195138065189e-05, "loss": 0.0148, "step": 47910 }, { "grad_norm": 0.49943727254867554, "learning_rate": 5.754560646426511e-05, "loss": 0.013, "step": 47920 }, { "grad_norm": 0.2526915967464447, "learning_rate": 5.752926072270216e-05, "loss": 0.012, "step": 47930 }, { "grad_norm": 1.6725521087646484, "learning_rate": 5.7512914157750563e-05, "loss": 0.0122, "step": 47940 }, { "grad_norm": 0.1947266310453415, "learning_rate": 5.749656677119798e-05, "loss": 0.0125, "step": 47950 }, { "grad_norm": 0.14712809026241302, "learning_rate": 5.7480218564832125e-05, "loss": 0.0111, "step": 47960 }, { "grad_norm": 0.3751247227191925, "learning_rate": 5.746386954044082e-05, "loss": 0.0134, "step": 47970 }, { "grad_norm": 0.14604127407073975, "learning_rate": 5.744751969981195e-05, "loss": 0.0105, "step": 47980 }, { "grad_norm": 0.2101837545633316, "learning_rate": 5.7431169044733526e-05, "loss": 0.0118, "step": 47990 }, { "grad_norm": 0.16075950860977173, "learning_rate": 5.741481757699364e-05, "loss": 0.0111, "step": 48000 }, { "grad_norm": 0.19091658294200897, "learning_rate": 5.7398465298380434e-05, "loss": 0.009, "step": 48010 }, { "grad_norm": 0.6088058352470398, "learning_rate": 5.7382112210682193e-05, "loss": 0.0107, "step": 48020 }, { "grad_norm": 0.17268340289592743, "learning_rate": 5.7365758315687266e-05, "loss": 0.0084, "step": 48030 }, { "grad_norm": 0.21552643179893494, "learning_rate": 5.734940361518407e-05, "loss": 0.0101, "step": 48040 }, { "grad_norm": 0.23790930211544037, "learning_rate": 5.733304811096116e-05, "loss": 0.0095, "step": 48050 }, { "grad_norm": 0.18163004517555237, "learning_rate": 5.731669180480713e-05, "loss": 0.0085, "step": 48060 }, { "grad_norm": 0.1722899079322815, "learning_rate": 5.730033469851067e-05, "loss": 0.0074, "step": 48070 }, { "grad_norm": 0.14702337980270386, "learning_rate": 5.7283976793860607e-05, "loss": 0.0079, "step": 48080 }, { "grad_norm": 0.1307867169380188, "learning_rate": 5.726761809264577e-05, "loss": 0.0099, "step": 48090 }, { "grad_norm": 0.12369723618030548, "learning_rate": 5.7251258596655155e-05, "loss": 0.0138, "step": 48100 }, { "grad_norm": 0.16074001789093018, "learning_rate": 5.72348983076778e-05, "loss": 0.01, "step": 48110 }, { "grad_norm": 0.24654719233512878, "learning_rate": 5.7218537227502854e-05, "loss": 0.0095, "step": 48120 }, { "grad_norm": 0.14892703294754028, "learning_rate": 5.720217535791951e-05, "loss": 0.0094, "step": 48130 }, { "grad_norm": 0.17118020355701447, "learning_rate": 5.718581270071711e-05, "loss": 0.0099, "step": 48140 }, { "grad_norm": 0.20105762779712677, "learning_rate": 5.716944925768505e-05, "loss": 0.0108, "step": 48150 }, { "grad_norm": 0.13982659578323364, "learning_rate": 5.7153085030612786e-05, "loss": 0.0098, "step": 48160 }, { "grad_norm": 0.16832995414733887, "learning_rate": 5.713672002128989e-05, "loss": 0.011, "step": 48170 }, { "grad_norm": 0.10633637756109238, "learning_rate": 5.712035423150602e-05, "loss": 0.0096, "step": 48180 }, { "grad_norm": 0.15069891512393951, "learning_rate": 5.710398766305094e-05, "loss": 0.0096, "step": 48190 }, { "grad_norm": 0.17340129613876343, "learning_rate": 5.708762031771442e-05, "loss": 0.0108, "step": 48200 }, { "grad_norm": 0.12478967756032944, "learning_rate": 5.707125219728642e-05, "loss": 0.0076, "step": 48210 }, { "grad_norm": 0.4464173913002014, "learning_rate": 5.7054883303556905e-05, "loss": 0.0138, "step": 48220 }, { "grad_norm": 0.14927437901496887, "learning_rate": 5.703851363831595e-05, "loss": 0.0085, "step": 48230 }, { "grad_norm": 0.14272776246070862, "learning_rate": 5.702214320335374e-05, "loss": 0.0097, "step": 48240 }, { "grad_norm": 0.10337880998849869, "learning_rate": 5.7005772000460514e-05, "loss": 0.009, "step": 48250 }, { "grad_norm": 0.18248993158340454, "learning_rate": 5.698940003142656e-05, "loss": 0.0096, "step": 48260 }, { "grad_norm": 0.14538097381591797, "learning_rate": 5.697302729804236e-05, "loss": 0.013, "step": 48270 }, { "grad_norm": 0.11529330909252167, "learning_rate": 5.695665380209837e-05, "loss": 0.0099, "step": 48280 }, { "grad_norm": 0.1902615875005722, "learning_rate": 5.6940279545385165e-05, "loss": 0.0076, "step": 48290 }, { "grad_norm": 0.1638169139623642, "learning_rate": 5.692390452969344e-05, "loss": 0.0111, "step": 48300 }, { "grad_norm": 0.20697256922721863, "learning_rate": 5.690752875681392e-05, "loss": 0.015, "step": 48310 }, { "grad_norm": 0.18406455218791962, "learning_rate": 5.6891152228537435e-05, "loss": 0.0134, "step": 48320 }, { "grad_norm": 0.17622047662734985, "learning_rate": 5.687477494665492e-05, "loss": 0.01, "step": 48330 }, { "grad_norm": 0.15538884699344635, "learning_rate": 5.685839691295733e-05, "loss": 0.013, "step": 48340 }, { "grad_norm": 0.13599270582199097, "learning_rate": 5.6842018129235786e-05, "loss": 0.0093, "step": 48350 }, { "grad_norm": 0.5300719738006592, "learning_rate": 5.6825638597281404e-05, "loss": 0.0123, "step": 48360 }, { "grad_norm": 0.09862959384918213, "learning_rate": 5.680925831888546e-05, "loss": 0.0085, "step": 48370 }, { "grad_norm": 0.1745425909757614, "learning_rate": 5.6792877295839274e-05, "loss": 0.0088, "step": 48380 }, { "grad_norm": 0.1742967963218689, "learning_rate": 5.6776495529934224e-05, "loss": 0.0129, "step": 48390 }, { "grad_norm": 0.11300253123044968, "learning_rate": 5.6760113022961824e-05, "loss": 0.0105, "step": 48400 }, { "grad_norm": 0.12772586941719055, "learning_rate": 5.6743729776713617e-05, "loss": 0.0084, "step": 48410 }, { "grad_norm": 0.15714110434055328, "learning_rate": 5.672734579298126e-05, "loss": 0.0124, "step": 48420 }, { "grad_norm": 0.12147945910692215, "learning_rate": 5.671096107355649e-05, "loss": 0.0077, "step": 48430 }, { "grad_norm": 0.5482985973358154, "learning_rate": 5.66945756202311e-05, "loss": 0.011, "step": 48440 }, { "grad_norm": 0.1541915088891983, "learning_rate": 5.667818943479699e-05, "loss": 0.0107, "step": 48450 }, { "grad_norm": 0.8257651329040527, "learning_rate": 5.666180251904612e-05, "loss": 0.0104, "step": 48460 }, { "grad_norm": 0.41817793250083923, "learning_rate": 5.6645414874770555e-05, "loss": 0.0083, "step": 48470 }, { "grad_norm": 0.11852550506591797, "learning_rate": 5.66290265037624e-05, "loss": 0.0092, "step": 48480 }, { "grad_norm": 0.15172182023525238, "learning_rate": 5.661263740781386e-05, "loss": 0.0081, "step": 48490 }, { "grad_norm": 0.26499006152153015, "learning_rate": 5.6596247588717254e-05, "loss": 0.0107, "step": 48500 }, { "grad_norm": 0.13905267417430878, "learning_rate": 5.6579857048264926e-05, "loss": 0.0098, "step": 48510 }, { "grad_norm": 0.21780313551425934, "learning_rate": 5.6563465788249314e-05, "loss": 0.0088, "step": 48520 }, { "grad_norm": 0.1544366478919983, "learning_rate": 5.6547073810462956e-05, "loss": 0.0108, "step": 48530 }, { "grad_norm": 0.3815976083278656, "learning_rate": 5.653068111669846e-05, "loss": 0.01, "step": 48540 }, { "grad_norm": 0.13445249199867249, "learning_rate": 5.651428770874848e-05, "loss": 0.0103, "step": 48550 }, { "grad_norm": 0.3418910503387451, "learning_rate": 5.64978935884058e-05, "loss": 0.0105, "step": 48560 }, { "grad_norm": 0.14962640404701233, "learning_rate": 5.6481498757463244e-05, "loss": 0.0087, "step": 48570 }, { "grad_norm": 0.14833304286003113, "learning_rate": 5.646510321771373e-05, "loss": 0.0083, "step": 48580 }, { "grad_norm": 0.29616647958755493, "learning_rate": 5.644870697095024e-05, "loss": 0.0141, "step": 48590 }, { "grad_norm": 0.2381250262260437, "learning_rate": 5.643231001896586e-05, "loss": 0.009, "step": 48600 }, { "grad_norm": 0.1458546668291092, "learning_rate": 5.6415912363553726e-05, "loss": 0.0102, "step": 48610 }, { "grad_norm": 0.16828331351280212, "learning_rate": 5.639951400650706e-05, "loss": 0.0084, "step": 48620 }, { "grad_norm": 0.14033567905426025, "learning_rate": 5.6383114949619165e-05, "loss": 0.0137, "step": 48630 }, { "grad_norm": 0.3237532079219818, "learning_rate": 5.636671519468342e-05, "loss": 0.0119, "step": 48640 }, { "grad_norm": 0.1475549191236496, "learning_rate": 5.635031474349327e-05, "loss": 0.0088, "step": 48650 }, { "grad_norm": 0.1357271671295166, "learning_rate": 5.6333913597842246e-05, "loss": 0.0081, "step": 48660 }, { "grad_norm": 0.15989139676094055, "learning_rate": 5.6317511759523955e-05, "loss": 0.0118, "step": 48670 }, { "grad_norm": 0.4138524830341339, "learning_rate": 5.630110923033207e-05, "loss": 0.0105, "step": 48680 }, { "grad_norm": 0.25262972712516785, "learning_rate": 5.628470601206036e-05, "loss": 0.0087, "step": 48690 }, { "grad_norm": 0.14551030099391937, "learning_rate": 5.626830210650266e-05, "loss": 0.0115, "step": 48700 }, { "grad_norm": 0.1269214004278183, "learning_rate": 5.625189751545285e-05, "loss": 0.0155, "step": 48710 }, { "grad_norm": 0.1748923510313034, "learning_rate": 5.6235492240704936e-05, "loss": 0.0117, "step": 48720 }, { "grad_norm": 0.1538962572813034, "learning_rate": 5.621908628405296e-05, "loss": 0.0078, "step": 48730 }, { "grad_norm": 0.15135465562343597, "learning_rate": 5.620267964729106e-05, "loss": 0.0077, "step": 48740 }, { "grad_norm": 0.1930686980485916, "learning_rate": 5.618627233221344e-05, "loss": 0.0073, "step": 48750 }, { "grad_norm": 0.1628979593515396, "learning_rate": 5.616986434061438e-05, "loss": 0.0082, "step": 48760 }, { "grad_norm": 0.16209132969379425, "learning_rate": 5.615345567428822e-05, "loss": 0.008, "step": 48770 }, { "grad_norm": 0.24185149371623993, "learning_rate": 5.613704633502941e-05, "loss": 0.0111, "step": 48780 }, { "grad_norm": 0.15103711187839508, "learning_rate": 5.612063632463245e-05, "loss": 0.0107, "step": 48790 }, { "grad_norm": 0.35170066356658936, "learning_rate": 5.610422564489188e-05, "loss": 0.0117, "step": 48800 }, { "grad_norm": 0.2503843605518341, "learning_rate": 5.608781429760239e-05, "loss": 0.013, "step": 48810 }, { "grad_norm": 0.20095472037792206, "learning_rate": 5.607140228455866e-05, "loss": 0.0108, "step": 48820 }, { "grad_norm": 0.2865165174007416, "learning_rate": 5.605498960755553e-05, "loss": 0.0086, "step": 48830 }, { "grad_norm": 0.39921262860298157, "learning_rate": 5.603857626838782e-05, "loss": 0.0119, "step": 48840 }, { "grad_norm": 0.14941540360450745, "learning_rate": 5.60221622688505e-05, "loss": 0.0103, "step": 48850 }, { "grad_norm": 0.49425506591796875, "learning_rate": 5.6005747610738565e-05, "loss": 0.0121, "step": 48860 }, { "grad_norm": 0.1853538155555725, "learning_rate": 5.59893322958471e-05, "loss": 0.0101, "step": 48870 }, { "grad_norm": 0.1701098084449768, "learning_rate": 5.5972916325971256e-05, "loss": 0.0124, "step": 48880 }, { "grad_norm": 0.18367817997932434, "learning_rate": 5.595649970290628e-05, "loss": 0.015, "step": 48890 }, { "grad_norm": 1.4597946405410767, "learning_rate": 5.5940082428447426e-05, "loss": 0.0131, "step": 48900 }, { "grad_norm": 0.12182057648897171, "learning_rate": 5.592366450439012e-05, "loss": 0.0072, "step": 48910 }, { "grad_norm": 0.19100473821163177, "learning_rate": 5.590724593252975e-05, "loss": 0.0096, "step": 48920 }, { "grad_norm": 0.30460116267204285, "learning_rate": 5.589082671466184e-05, "loss": 0.0081, "step": 48930 }, { "grad_norm": 0.08632616698741913, "learning_rate": 5.587440685258199e-05, "loss": 0.0098, "step": 48940 }, { "grad_norm": 0.2132650464773178, "learning_rate": 5.585798634808583e-05, "loss": 0.0104, "step": 48950 }, { "grad_norm": 0.12310122698545456, "learning_rate": 5.584156520296909e-05, "loss": 0.0088, "step": 48960 }, { "grad_norm": 0.14440467953681946, "learning_rate": 5.582514341902757e-05, "loss": 0.0084, "step": 48970 }, { "grad_norm": 0.14777636528015137, "learning_rate": 5.580872099805713e-05, "loss": 0.0081, "step": 48980 }, { "grad_norm": 0.1590174436569214, "learning_rate": 5.5792297941853674e-05, "loss": 0.0109, "step": 48990 }, { "grad_norm": 0.17826615273952484, "learning_rate": 5.5775874252213247e-05, "loss": 0.0101, "step": 49000 }, { "grad_norm": 0.22985339164733887, "learning_rate": 5.575944993093189e-05, "loss": 0.0105, "step": 49010 }, { "grad_norm": 0.2616671621799469, "learning_rate": 5.574302497980574e-05, "loss": 0.0138, "step": 49020 }, { "grad_norm": 0.13388164341449738, "learning_rate": 5.5726599400631005e-05, "loss": 0.0101, "step": 49030 }, { "grad_norm": 0.18011003732681274, "learning_rate": 5.571017319520399e-05, "loss": 0.0087, "step": 49040 }, { "grad_norm": 0.19027984142303467, "learning_rate": 5.5693746365320984e-05, "loss": 0.0093, "step": 49050 }, { "grad_norm": 0.26935648918151855, "learning_rate": 5.567731891277848e-05, "loss": 0.0083, "step": 49060 }, { "grad_norm": 0.14933079481124878, "learning_rate": 5.566089083937288e-05, "loss": 0.0073, "step": 49070 }, { "grad_norm": 0.1838313788175583, "learning_rate": 5.564446214690079e-05, "loss": 0.008, "step": 49080 }, { "grad_norm": 0.1309918314218521, "learning_rate": 5.56280328371588e-05, "loss": 0.0083, "step": 49090 }, { "grad_norm": 0.09483128786087036, "learning_rate": 5.5611602911943596e-05, "loss": 0.0101, "step": 49100 }, { "grad_norm": 0.13160203397274017, "learning_rate": 5.559517237305194e-05, "loss": 0.0087, "step": 49110 }, { "grad_norm": 0.17905347049236298, "learning_rate": 5.557874122228064e-05, "loss": 0.0085, "step": 49120 }, { "grad_norm": 0.14050854742527008, "learning_rate": 5.55623094614266e-05, "loss": 0.0099, "step": 49130 }, { "grad_norm": 0.1364348828792572, "learning_rate": 5.5545877092286743e-05, "loss": 0.009, "step": 49140 }, { "grad_norm": 0.1803288757801056, "learning_rate": 5.552944411665811e-05, "loss": 0.0095, "step": 49150 }, { "grad_norm": 0.2754315137863159, "learning_rate": 5.5513010536337795e-05, "loss": 0.0089, "step": 49160 }, { "grad_norm": 0.21490731835365295, "learning_rate": 5.5496576353122934e-05, "loss": 0.0099, "step": 49170 }, { "grad_norm": 0.16391237080097198, "learning_rate": 5.548014156881074e-05, "loss": 0.0094, "step": 49180 }, { "grad_norm": 0.17493443191051483, "learning_rate": 5.5463706185198494e-05, "loss": 0.0096, "step": 49190 }, { "grad_norm": 0.12817882001399994, "learning_rate": 5.544727020408358e-05, "loss": 0.0082, "step": 49200 }, { "grad_norm": 0.1626640260219574, "learning_rate": 5.5430833627263366e-05, "loss": 0.0092, "step": 49210 }, { "grad_norm": 0.12167546898126602, "learning_rate": 5.5414396456535354e-05, "loss": 0.0079, "step": 49220 }, { "grad_norm": 0.30042847990989685, "learning_rate": 5.539795869369709e-05, "loss": 0.0097, "step": 49230 }, { "grad_norm": 0.1521311104297638, "learning_rate": 5.538152034054618e-05, "loss": 0.0087, "step": 49240 }, { "grad_norm": 0.20310629904270172, "learning_rate": 5.536508139888028e-05, "loss": 0.0104, "step": 49250 }, { "grad_norm": 0.16050760447978973, "learning_rate": 5.534864187049716e-05, "loss": 0.0089, "step": 49260 }, { "grad_norm": 0.1988075226545334, "learning_rate": 5.5332201757194615e-05, "loss": 0.008, "step": 49270 }, { "grad_norm": 0.1574789583683014, "learning_rate": 5.531576106077049e-05, "loss": 0.009, "step": 49280 }, { "grad_norm": 0.2121402621269226, "learning_rate": 5.529931978302272e-05, "loss": 0.0073, "step": 49290 }, { "grad_norm": 0.12175214290618896, "learning_rate": 5.528287792574932e-05, "loss": 0.0083, "step": 49300 }, { "grad_norm": 0.2537054717540741, "learning_rate": 5.5266435490748324e-05, "loss": 0.0132, "step": 49310 }, { "grad_norm": 0.12159081548452377, "learning_rate": 5.524999247981787e-05, "loss": 0.007, "step": 49320 }, { "grad_norm": 0.14517661929130554, "learning_rate": 5.523354889475613e-05, "loss": 0.0087, "step": 49330 }, { "grad_norm": 0.16901762783527374, "learning_rate": 5.521710473736134e-05, "loss": 0.0088, "step": 49340 }, { "grad_norm": 0.13201956450939178, "learning_rate": 5.5200660009431826e-05, "loss": 0.0075, "step": 49350 }, { "grad_norm": 0.1664828360080719, "learning_rate": 5.518421471276596e-05, "loss": 0.0062, "step": 49360 }, { "grad_norm": 0.184669628739357, "learning_rate": 5.5167768849162174e-05, "loss": 0.007, "step": 49370 }, { "grad_norm": 0.21103236079216003, "learning_rate": 5.515132242041893e-05, "loss": 0.0119, "step": 49380 }, { "grad_norm": 0.1173931360244751, "learning_rate": 5.513487542833483e-05, "loss": 0.0077, "step": 49390 }, { "grad_norm": 0.21020576357841492, "learning_rate": 5.5118427874708466e-05, "loss": 0.0108, "step": 49400 }, { "grad_norm": 0.1769452542066574, "learning_rate": 5.510197976133853e-05, "loss": 0.0138, "step": 49410 }, { "grad_norm": 0.44134455919265747, "learning_rate": 5.508553109002376e-05, "loss": 0.0119, "step": 49420 }, { "grad_norm": 0.1656429022550583, "learning_rate": 5.5069081862562957e-05, "loss": 0.0152, "step": 49430 }, { "grad_norm": 0.46931275725364685, "learning_rate": 5.5052632080754965e-05, "loss": 0.0104, "step": 49440 }, { "grad_norm": 0.16179081797599792, "learning_rate": 5.503618174639874e-05, "loss": 0.0082, "step": 49450 }, { "grad_norm": 0.23983146250247955, "learning_rate": 5.501973086129325e-05, "loss": 0.0087, "step": 49460 }, { "grad_norm": 0.21064318716526031, "learning_rate": 5.500327942723753e-05, "loss": 0.0146, "step": 49470 }, { "grad_norm": 2.3581135272979736, "learning_rate": 5.498682744603071e-05, "loss": 0.0157, "step": 49480 }, { "grad_norm": 0.12948451936244965, "learning_rate": 5.497037491947194e-05, "loss": 0.0116, "step": 49490 }, { "grad_norm": 0.15251468122005463, "learning_rate": 5.4953921849360424e-05, "loss": 0.0114, "step": 49500 }, { "grad_norm": 0.16461485624313354, "learning_rate": 5.493746823749547e-05, "loss": 0.0106, "step": 49510 }, { "grad_norm": 0.1742333173751831, "learning_rate": 5.49210140856764e-05, "loss": 0.0089, "step": 49520 }, { "grad_norm": 0.13660134375095367, "learning_rate": 5.4904559395702647e-05, "loss": 0.0088, "step": 49530 }, { "grad_norm": 0.2146114856004715, "learning_rate": 5.488810416937364e-05, "loss": 0.0167, "step": 49540 }, { "grad_norm": 0.16552318632602692, "learning_rate": 5.4871648408488926e-05, "loss": 0.012, "step": 49550 }, { "grad_norm": 0.22573010623455048, "learning_rate": 5.485519211484807e-05, "loss": 0.0195, "step": 49560 }, { "grad_norm": 0.12226787209510803, "learning_rate": 5.4838735290250676e-05, "loss": 0.0162, "step": 49570 }, { "grad_norm": 0.1774284690618515, "learning_rate": 5.482227793649649e-05, "loss": 0.0181, "step": 49580 }, { "grad_norm": 0.20314253866672516, "learning_rate": 5.480582005538524e-05, "loss": 0.0085, "step": 49590 }, { "grad_norm": 0.17916829884052277, "learning_rate": 5.478936164871671e-05, "loss": 0.0126, "step": 49600 }, { "grad_norm": 0.13456758856773376, "learning_rate": 5.477290271829081e-05, "loss": 0.0106, "step": 49610 }, { "grad_norm": 0.12371963262557983, "learning_rate": 5.475644326590744e-05, "loss": 0.0104, "step": 49620 }, { "grad_norm": 0.13332070410251617, "learning_rate": 5.473998329336658e-05, "loss": 0.0089, "step": 49630 }, { "grad_norm": 0.13955122232437134, "learning_rate": 5.4723522802468286e-05, "loss": 0.0125, "step": 49640 }, { "grad_norm": 0.15049505233764648, "learning_rate": 5.470706179501264e-05, "loss": 0.0111, "step": 49650 }, { "grad_norm": 0.162215918302536, "learning_rate": 5.4690600272799776e-05, "loss": 0.0109, "step": 49660 }, { "grad_norm": 0.14119839668273926, "learning_rate": 5.467413823762993e-05, "loss": 0.0085, "step": 49670 }, { "grad_norm": 0.17300108075141907, "learning_rate": 5.465767569130335e-05, "loss": 0.0125, "step": 49680 }, { "grad_norm": 0.3717849850654602, "learning_rate": 5.464121263562036e-05, "loss": 0.0084, "step": 49690 }, { "grad_norm": 0.15008851885795593, "learning_rate": 5.4624749072381343e-05, "loss": 0.0096, "step": 49700 }, { "grad_norm": 0.20111767947673798, "learning_rate": 5.460828500338672e-05, "loss": 0.0118, "step": 49710 }, { "grad_norm": 0.199178084731102, "learning_rate": 5.459182043043698e-05, "loss": 0.0094, "step": 49720 }, { "grad_norm": 0.12386877834796906, "learning_rate": 5.457535535533265e-05, "loss": 0.0094, "step": 49730 }, { "grad_norm": 0.1659177988767624, "learning_rate": 5.4558889779874334e-05, "loss": 0.0109, "step": 49740 }, { "grad_norm": 0.11692991852760315, "learning_rate": 5.454242370586269e-05, "loss": 0.0096, "step": 49750 }, { "grad_norm": 0.19141745567321777, "learning_rate": 5.452595713509843e-05, "loss": 0.0103, "step": 49760 }, { "grad_norm": 0.5322921276092529, "learning_rate": 5.4509490069382274e-05, "loss": 0.011, "step": 49770 }, { "grad_norm": 0.12263180315494537, "learning_rate": 5.449302251051509e-05, "loss": 0.0088, "step": 49780 }, { "grad_norm": 0.2105763852596283, "learning_rate": 5.44765544602977e-05, "loss": 0.0077, "step": 49790 }, { "grad_norm": 0.15699279308319092, "learning_rate": 5.446008592053107e-05, "loss": 0.0104, "step": 49800 }, { "grad_norm": 0.12877367436885834, "learning_rate": 5.4443616893016135e-05, "loss": 0.0111, "step": 49810 }, { "grad_norm": 0.20298928022384644, "learning_rate": 5.4427147379553934e-05, "loss": 0.0109, "step": 49820 }, { "grad_norm": 0.15216177701950073, "learning_rate": 5.441067738194556e-05, "loss": 0.0079, "step": 49830 }, { "grad_norm": 0.20488503575325012, "learning_rate": 5.439420690199214e-05, "loss": 0.0084, "step": 49840 }, { "grad_norm": 0.14478321373462677, "learning_rate": 5.437773594149484e-05, "loss": 0.01, "step": 49850 }, { "grad_norm": 0.19834095239639282, "learning_rate": 5.436126450225495e-05, "loss": 0.0104, "step": 49860 }, { "grad_norm": 0.19047293066978455, "learning_rate": 5.434479258607371e-05, "loss": 0.0107, "step": 49870 }, { "grad_norm": 0.3026626706123352, "learning_rate": 5.43283201947525e-05, "loss": 0.0128, "step": 49880 }, { "grad_norm": 0.19818361103534698, "learning_rate": 5.431184733009268e-05, "loss": 0.0107, "step": 49890 }, { "grad_norm": 0.26595789194107056, "learning_rate": 5.4295373993895736e-05, "loss": 0.0107, "step": 49900 }, { "grad_norm": 0.17218677699565887, "learning_rate": 5.4278900187963157e-05, "loss": 0.0141, "step": 49910 }, { "grad_norm": 0.2260938584804535, "learning_rate": 5.426242591409646e-05, "loss": 0.0108, "step": 49920 }, { "grad_norm": 0.16870900988578796, "learning_rate": 5.42459511740973e-05, "loss": 0.0095, "step": 49930 }, { "grad_norm": 0.25692060589790344, "learning_rate": 5.422947596976729e-05, "loss": 0.011, "step": 49940 }, { "grad_norm": 0.13450093567371368, "learning_rate": 5.4213000302908134e-05, "loss": 0.0095, "step": 49950 }, { "grad_norm": 0.17325150966644287, "learning_rate": 5.419652417532162e-05, "loss": 0.0117, "step": 49960 }, { "grad_norm": 0.16780969500541687, "learning_rate": 5.4180047588809534e-05, "loss": 0.0102, "step": 49970 }, { "grad_norm": 0.15344150364398956, "learning_rate": 5.4163570545173704e-05, "loss": 0.0112, "step": 49980 }, { "grad_norm": 0.17918898165225983, "learning_rate": 5.414709304621608e-05, "loss": 0.0113, "step": 49990 }, { "grad_norm": 0.2392815202474594, "learning_rate": 5.413061509373858e-05, "loss": 0.0098, "step": 50000 }, { "grad_norm": 0.19956547021865845, "learning_rate": 5.411413668954324e-05, "loss": 0.0125, "step": 50010 }, { "grad_norm": 0.23662850260734558, "learning_rate": 5.4097657835432083e-05, "loss": 0.008, "step": 50020 }, { "grad_norm": 0.14588432013988495, "learning_rate": 5.408117853320723e-05, "loss": 0.0097, "step": 50030 }, { "grad_norm": 0.4190773069858551, "learning_rate": 5.406469878467084e-05, "loss": 0.009, "step": 50040 }, { "grad_norm": 0.13428814709186554, "learning_rate": 5.404821859162509e-05, "loss": 0.0105, "step": 50050 }, { "grad_norm": 0.3894754946231842, "learning_rate": 5.403173795587225e-05, "loss": 0.0165, "step": 50060 }, { "grad_norm": 0.14323456585407257, "learning_rate": 5.4015256879214606e-05, "loss": 0.0091, "step": 50070 }, { "grad_norm": 0.1701110601425171, "learning_rate": 5.39987753634545e-05, "loss": 0.0116, "step": 50080 }, { "grad_norm": 0.13612478971481323, "learning_rate": 5.3982293410394335e-05, "loss": 0.0111, "step": 50090 }, { "grad_norm": 0.33951494097709656, "learning_rate": 5.396581102183655e-05, "loss": 0.0142, "step": 50100 }, { "grad_norm": 0.1572892665863037, "learning_rate": 5.3949328199583634e-05, "loss": 0.0137, "step": 50110 }, { "grad_norm": 0.1594412922859192, "learning_rate": 5.393284494543813e-05, "loss": 0.0105, "step": 50120 }, { "grad_norm": 0.20292548835277557, "learning_rate": 5.391636126120262e-05, "loss": 0.0091, "step": 50130 }, { "grad_norm": 0.14421086013317108, "learning_rate": 5.389987714867971e-05, "loss": 0.0096, "step": 50140 }, { "grad_norm": 0.11417770385742188, "learning_rate": 5.3883392609672114e-05, "loss": 0.0114, "step": 50150 }, { "grad_norm": 0.10767723619937897, "learning_rate": 5.3866907645982545e-05, "loss": 0.0122, "step": 50160 }, { "grad_norm": 0.11643616110086441, "learning_rate": 5.385042225941375e-05, "loss": 0.0085, "step": 50170 }, { "grad_norm": 0.19245360791683197, "learning_rate": 5.3833936451768585e-05, "loss": 0.0182, "step": 50180 }, { "grad_norm": 0.42345404624938965, "learning_rate": 5.3817450224849884e-05, "loss": 0.0107, "step": 50190 }, { "grad_norm": 0.132253959774971, "learning_rate": 5.380096358046056e-05, "loss": 0.009, "step": 50200 }, { "grad_norm": 0.17996396124362946, "learning_rate": 5.378447652040359e-05, "loss": 0.0118, "step": 50210 }, { "grad_norm": 0.2010977566242218, "learning_rate": 5.376798904648195e-05, "loss": 0.0082, "step": 50220 }, { "grad_norm": 0.11685679107904434, "learning_rate": 5.375150116049869e-05, "loss": 0.0077, "step": 50230 }, { "grad_norm": 0.36754700541496277, "learning_rate": 5.373501286425691e-05, "loss": 0.0096, "step": 50240 }, { "grad_norm": 0.1444748342037201, "learning_rate": 5.3718524159559726e-05, "loss": 0.0097, "step": 50250 }, { "grad_norm": 0.1273660808801651, "learning_rate": 5.370203504821034e-05, "loss": 0.0122, "step": 50260 }, { "grad_norm": 0.15192635357379913, "learning_rate": 5.368554553201196e-05, "loss": 0.0098, "step": 50270 }, { "grad_norm": 0.12070261687040329, "learning_rate": 5.3669055612767874e-05, "loss": 0.0079, "step": 50280 }, { "grad_norm": 0.148673415184021, "learning_rate": 5.3652565292281376e-05, "loss": 0.0121, "step": 50290 }, { "grad_norm": 0.11152976751327515, "learning_rate": 5.3636074572355824e-05, "loss": 0.0057, "step": 50300 }, { "grad_norm": 0.12584014236927032, "learning_rate": 5.361958345479463e-05, "loss": 0.0087, "step": 50310 }, { "grad_norm": 0.5885133743286133, "learning_rate": 5.360309194140123e-05, "loss": 0.0135, "step": 50320 }, { "grad_norm": 0.13018111884593964, "learning_rate": 5.3586600033979106e-05, "loss": 0.0072, "step": 50330 }, { "grad_norm": 0.23788882791996002, "learning_rate": 5.357010773433181e-05, "loss": 0.0089, "step": 50340 }, { "grad_norm": 0.14376965165138245, "learning_rate": 5.3553615044262895e-05, "loss": 0.0096, "step": 50350 }, { "grad_norm": 0.1629733145236969, "learning_rate": 5.3537121965575964e-05, "loss": 0.0115, "step": 50360 }, { "grad_norm": 0.14971725642681122, "learning_rate": 5.352062850007471e-05, "loss": 0.0082, "step": 50370 }, { "grad_norm": 0.10451975464820862, "learning_rate": 5.350413464956282e-05, "loss": 0.0073, "step": 50380 }, { "grad_norm": 0.10679508000612259, "learning_rate": 5.348764041584403e-05, "loss": 0.0084, "step": 50390 }, { "grad_norm": 0.09815193712711334, "learning_rate": 5.347114580072212e-05, "loss": 0.0066, "step": 50400 }, { "grad_norm": 0.13848121464252472, "learning_rate": 5.345465080600093e-05, "loss": 0.0145, "step": 50410 }, { "grad_norm": 0.12551434338092804, "learning_rate": 5.3438155433484326e-05, "loss": 0.008, "step": 50420 }, { "grad_norm": 0.16348876059055328, "learning_rate": 5.3421659684976197e-05, "loss": 0.0128, "step": 50430 }, { "grad_norm": 0.13049724698066711, "learning_rate": 5.340516356228052e-05, "loss": 0.0132, "step": 50440 }, { "grad_norm": 0.19085077941417694, "learning_rate": 5.338866706720128e-05, "loss": 0.0093, "step": 50450 }, { "grad_norm": 0.38673993945121765, "learning_rate": 5.337217020154249e-05, "loss": 0.0115, "step": 50460 }, { "grad_norm": 0.28881344199180603, "learning_rate": 5.335567296710825e-05, "loss": 0.0094, "step": 50470 }, { "grad_norm": 0.14110024273395538, "learning_rate": 5.333917536570265e-05, "loss": 0.0103, "step": 50480 }, { "grad_norm": 0.17390428483486176, "learning_rate": 5.332267739912986e-05, "loss": 0.0091, "step": 50490 }, { "grad_norm": 0.17368319630622864, "learning_rate": 5.330617906919405e-05, "loss": 0.0098, "step": 50500 }, { "grad_norm": 0.19625309109687805, "learning_rate": 5.328968037769949e-05, "loss": 0.0091, "step": 50510 }, { "grad_norm": 0.1866019070148468, "learning_rate": 5.327318132645042e-05, "loss": 0.0077, "step": 50520 }, { "grad_norm": 0.1446562111377716, "learning_rate": 5.325668191725116e-05, "loss": 0.0086, "step": 50530 }, { "grad_norm": 0.30794528126716614, "learning_rate": 5.324018215190606e-05, "loss": 0.0106, "step": 50540 }, { "grad_norm": 0.12356224656105042, "learning_rate": 5.3223682032219515e-05, "loss": 0.0069, "step": 50550 }, { "grad_norm": 0.14277470111846924, "learning_rate": 5.320718155999595e-05, "loss": 0.0086, "step": 50560 }, { "grad_norm": 0.16032983362674713, "learning_rate": 5.3190680737039835e-05, "loss": 0.0087, "step": 50570 }, { "grad_norm": 0.34786364436149597, "learning_rate": 5.317417956515567e-05, "loss": 0.0107, "step": 50580 }, { "grad_norm": 0.1479780375957489, "learning_rate": 5.315767804614798e-05, "loss": 0.0091, "step": 50590 }, { "grad_norm": 0.1266741007566452, "learning_rate": 5.3141176181821395e-05, "loss": 0.0077, "step": 50600 }, { "grad_norm": 0.154609814286232, "learning_rate": 5.31246739739805e-05, "loss": 0.01, "step": 50610 }, { "grad_norm": 0.11978095024824142, "learning_rate": 5.310817142442994e-05, "loss": 0.0096, "step": 50620 }, { "grad_norm": 0.20176784694194794, "learning_rate": 5.309166853497445e-05, "loss": 0.0075, "step": 50630 }, { "grad_norm": 0.14303675293922424, "learning_rate": 5.307516530741873e-05, "loss": 0.0089, "step": 50640 }, { "grad_norm": 0.7298737168312073, "learning_rate": 5.305866174356754e-05, "loss": 0.0097, "step": 50650 }, { "grad_norm": 0.2441675364971161, "learning_rate": 5.304215784522571e-05, "loss": 0.0086, "step": 50660 }, { "grad_norm": 0.13653133809566498, "learning_rate": 5.302565361419808e-05, "loss": 0.0084, "step": 50670 }, { "grad_norm": 0.33993756771087646, "learning_rate": 5.3009149052289507e-05, "loss": 0.0124, "step": 50680 }, { "grad_norm": 0.17872686684131622, "learning_rate": 5.299264416130493e-05, "loss": 0.0128, "step": 50690 }, { "grad_norm": 0.14493384957313538, "learning_rate": 5.297613894304928e-05, "loss": 0.0122, "step": 50700 }, { "grad_norm": 0.15277914702892303, "learning_rate": 5.2959633399327534e-05, "loss": 0.0102, "step": 50710 }, { "grad_norm": 0.12036523967981339, "learning_rate": 5.294312753194476e-05, "loss": 0.0085, "step": 50720 }, { "grad_norm": 0.5381126403808594, "learning_rate": 5.292662134270596e-05, "loss": 0.0089, "step": 50730 }, { "grad_norm": 0.3497181534767151, "learning_rate": 5.291011483341626e-05, "loss": 0.0117, "step": 50740 }, { "grad_norm": 0.2224530577659607, "learning_rate": 5.2893608005880767e-05, "loss": 0.0091, "step": 50750 }, { "grad_norm": 0.11887647211551666, "learning_rate": 5.287710086190467e-05, "loss": 0.007, "step": 50760 }, { "grad_norm": 0.1727105975151062, "learning_rate": 5.286059340329314e-05, "loss": 0.0078, "step": 50770 }, { "grad_norm": 0.0929587110877037, "learning_rate": 5.284408563185141e-05, "loss": 0.0096, "step": 50780 }, { "grad_norm": 0.15861578285694122, "learning_rate": 5.282757754938476e-05, "loss": 0.0082, "step": 50790 }, { "grad_norm": 0.11760067194700241, "learning_rate": 5.281106915769849e-05, "loss": 0.0106, "step": 50800 }, { "grad_norm": 0.18528422713279724, "learning_rate": 5.2794560458597897e-05, "loss": 0.01, "step": 50810 }, { "grad_norm": 0.2265310287475586, "learning_rate": 5.27780514538884e-05, "loss": 0.0098, "step": 50820 }, { "grad_norm": 0.2061576396226883, "learning_rate": 5.2761542145375365e-05, "loss": 0.0082, "step": 50830 }, { "grad_norm": 0.22451996803283691, "learning_rate": 5.274503253486421e-05, "loss": 0.0103, "step": 50840 }, { "grad_norm": 0.2581171989440918, "learning_rate": 5.272852262416046e-05, "loss": 0.0122, "step": 50850 }, { "grad_norm": 0.23697037994861603, "learning_rate": 5.2712012415069555e-05, "loss": 0.0109, "step": 50860 }, { "grad_norm": 0.14406539499759674, "learning_rate": 5.269550190939705e-05, "loss": 0.0083, "step": 50870 }, { "grad_norm": 0.13273875415325165, "learning_rate": 5.267899110894852e-05, "loss": 0.0103, "step": 50880 }, { "grad_norm": 0.17651745676994324, "learning_rate": 5.266248001552955e-05, "loss": 0.0112, "step": 50890 }, { "grad_norm": 0.179323211312294, "learning_rate": 5.264596863094575e-05, "loss": 0.0108, "step": 50900 }, { "grad_norm": 0.12233174592256546, "learning_rate": 5.262945695700282e-05, "loss": 0.0099, "step": 50910 }, { "grad_norm": 0.1436372995376587, "learning_rate": 5.261294499550643e-05, "loss": 0.0109, "step": 50920 }, { "grad_norm": 0.13487203419208527, "learning_rate": 5.25964327482623e-05, "loss": 0.009, "step": 50930 }, { "grad_norm": 2.0303428173065186, "learning_rate": 5.257992021707617e-05, "loss": 0.0109, "step": 50940 }, { "grad_norm": 0.1407209038734436, "learning_rate": 5.256340740375387e-05, "loss": 0.0109, "step": 50950 }, { "grad_norm": 0.1610199362039566, "learning_rate": 5.254689431010117e-05, "loss": 0.0108, "step": 50960 }, { "grad_norm": 0.16889052093029022, "learning_rate": 5.253038093792395e-05, "loss": 0.0106, "step": 50970 }, { "grad_norm": 0.14668448269367218, "learning_rate": 5.251386728902806e-05, "loss": 0.0094, "step": 50980 }, { "grad_norm": 0.1617175191640854, "learning_rate": 5.2497353365219446e-05, "loss": 0.0095, "step": 50990 }, { "grad_norm": 0.14223189651966095, "learning_rate": 5.2480839168304e-05, "loss": 0.012, "step": 51000 }, { "grad_norm": 0.2175028920173645, "learning_rate": 5.246432470008773e-05, "loss": 0.0074, "step": 51010 }, { "grad_norm": 0.15514250099658966, "learning_rate": 5.2447809962376616e-05, "loss": 0.0108, "step": 51020 }, { "grad_norm": 0.13249224424362183, "learning_rate": 5.243129495697666e-05, "loss": 0.0084, "step": 51030 }, { "grad_norm": 0.2116071581840515, "learning_rate": 5.241477968569397e-05, "loss": 0.01, "step": 51040 }, { "grad_norm": 0.6013069748878479, "learning_rate": 5.23982641503346e-05, "loss": 0.0177, "step": 51050 }, { "grad_norm": 0.1571589857339859, "learning_rate": 5.238174835270465e-05, "loss": 0.0141, "step": 51060 }, { "grad_norm": 0.20407646894454956, "learning_rate": 5.23652322946103e-05, "loss": 0.0136, "step": 51070 }, { "grad_norm": 0.17827370762825012, "learning_rate": 5.234871597785771e-05, "loss": 0.0111, "step": 51080 }, { "grad_norm": 0.1403815895318985, "learning_rate": 5.233219940425306e-05, "loss": 0.0121, "step": 51090 }, { "grad_norm": 0.09984592348337173, "learning_rate": 5.2315682575602576e-05, "loss": 0.0101, "step": 51100 }, { "grad_norm": 0.30102041363716125, "learning_rate": 5.229916549371254e-05, "loss": 0.011, "step": 51110 }, { "grad_norm": 0.2444184273481369, "learning_rate": 5.228264816038924e-05, "loss": 0.0086, "step": 51120 }, { "grad_norm": 0.7774367332458496, "learning_rate": 5.2266130577438935e-05, "loss": 0.0128, "step": 51130 }, { "grad_norm": 0.31449171900749207, "learning_rate": 5.224961274666802e-05, "loss": 0.0107, "step": 51140 }, { "grad_norm": 0.33040663599967957, "learning_rate": 5.223309466988283e-05, "loss": 0.0087, "step": 51150 }, { "grad_norm": 0.32774004340171814, "learning_rate": 5.221657634888976e-05, "loss": 0.0099, "step": 51160 }, { "grad_norm": 0.16605497896671295, "learning_rate": 5.2200057785495235e-05, "loss": 0.0122, "step": 51170 }, { "grad_norm": 0.21236591041088104, "learning_rate": 5.2183538981505706e-05, "loss": 0.0112, "step": 51180 }, { "grad_norm": 0.11137451976537704, "learning_rate": 5.216701993872762e-05, "loss": 0.0103, "step": 51190 }, { "grad_norm": 0.14421764016151428, "learning_rate": 5.215050065896752e-05, "loss": 0.0097, "step": 51200 }, { "grad_norm": 0.14865435659885406, "learning_rate": 5.2133981144031875e-05, "loss": 0.0096, "step": 51210 }, { "grad_norm": 0.14888723194599152, "learning_rate": 5.211746139572727e-05, "loss": 0.0086, "step": 51220 }, { "grad_norm": 0.16383984684944153, "learning_rate": 5.210094141586026e-05, "loss": 0.0099, "step": 51230 }, { "grad_norm": 0.12896858155727386, "learning_rate": 5.2084421206237476e-05, "loss": 0.0106, "step": 51240 }, { "grad_norm": 0.14134624600410461, "learning_rate": 5.20679007686655e-05, "loss": 0.0113, "step": 51250 }, { "grad_norm": 0.24483422935009003, "learning_rate": 5.205138010495102e-05, "loss": 0.0084, "step": 51260 }, { "grad_norm": 1.0810753107070923, "learning_rate": 5.20348592169007e-05, "loss": 0.0077, "step": 51270 }, { "grad_norm": 0.13400983810424805, "learning_rate": 5.201833810632123e-05, "loss": 0.0081, "step": 51280 }, { "grad_norm": 0.5300341248512268, "learning_rate": 5.200181677501933e-05, "loss": 0.0085, "step": 51290 }, { "grad_norm": 0.15537728369235992, "learning_rate": 5.1985295224801777e-05, "loss": 0.0094, "step": 51300 }, { "grad_norm": 0.13514076173305511, "learning_rate": 5.196877345747531e-05, "loss": 0.0075, "step": 51310 }, { "grad_norm": 0.13664263486862183, "learning_rate": 5.195225147484674e-05, "loss": 0.0072, "step": 51320 }, { "grad_norm": 0.1620548665523529, "learning_rate": 5.193572927872288e-05, "loss": 0.0078, "step": 51330 }, { "grad_norm": 0.14119279384613037, "learning_rate": 5.191920687091059e-05, "loss": 0.008, "step": 51340 }, { "grad_norm": 0.37747278809547424, "learning_rate": 5.1902684253216695e-05, "loss": 0.0078, "step": 51350 }, { "grad_norm": 0.17375445365905762, "learning_rate": 5.1886161427448134e-05, "loss": 0.0125, "step": 51360 }, { "grad_norm": 0.17914360761642456, "learning_rate": 5.186963839541179e-05, "loss": 0.0112, "step": 51370 }, { "grad_norm": 0.13551488518714905, "learning_rate": 5.185311515891459e-05, "loss": 0.0093, "step": 51380 }, { "grad_norm": 0.22706249356269836, "learning_rate": 5.1836591719763516e-05, "loss": 0.0095, "step": 51390 }, { "grad_norm": 0.17805175483226776, "learning_rate": 5.182006807976554e-05, "loss": 0.0082, "step": 51400 }, { "grad_norm": 0.12497159093618393, "learning_rate": 5.180354424072763e-05, "loss": 0.0105, "step": 51410 }, { "grad_norm": 0.18491971492767334, "learning_rate": 5.178702020445686e-05, "loss": 0.0113, "step": 51420 }, { "grad_norm": 0.1979002207517624, "learning_rate": 5.177049597276024e-05, "loss": 0.0086, "step": 51430 }, { "grad_norm": 0.5238165259361267, "learning_rate": 5.175397154744482e-05, "loss": 0.0091, "step": 51440 }, { "grad_norm": 0.1333279311656952, "learning_rate": 5.1737446930317736e-05, "loss": 0.0093, "step": 51450 }, { "grad_norm": 0.1375017911195755, "learning_rate": 5.172092212318604e-05, "loss": 0.013, "step": 51460 }, { "grad_norm": 0.612653374671936, "learning_rate": 5.170439712785692e-05, "loss": 0.0087, "step": 51470 }, { "grad_norm": 0.14449521899223328, "learning_rate": 5.1687871946137467e-05, "loss": 0.0084, "step": 51480 }, { "grad_norm": 0.25469037890434265, "learning_rate": 5.16713465798349e-05, "loss": 0.0093, "step": 51490 }, { "grad_norm": 0.28329944610595703, "learning_rate": 5.165482103075637e-05, "loss": 0.0103, "step": 51500 }, { "grad_norm": 0.12827226519584656, "learning_rate": 5.163829530070909e-05, "loss": 0.0079, "step": 51510 }, { "grad_norm": 0.14178234338760376, "learning_rate": 5.162176939150032e-05, "loss": 0.0128, "step": 51520 }, { "grad_norm": 0.14171044528484344, "learning_rate": 5.160524330493728e-05, "loss": 0.0148, "step": 51530 }, { "grad_norm": 0.9221852421760559, "learning_rate": 5.158871704282724e-05, "loss": 0.0116, "step": 51540 }, { "grad_norm": 0.15153531730175018, "learning_rate": 5.15721906069775e-05, "loss": 0.01, "step": 51550 }, { "grad_norm": 0.12227185815572739, "learning_rate": 5.155566399919536e-05, "loss": 0.0087, "step": 51560 }, { "grad_norm": 0.1632934808731079, "learning_rate": 5.153913722128813e-05, "loss": 0.0133, "step": 51570 }, { "grad_norm": 0.19825497269630432, "learning_rate": 5.1522610275063176e-05, "loss": 0.0118, "step": 51580 }, { "grad_norm": 0.09582898765802383, "learning_rate": 5.150608316232786e-05, "loss": 0.0079, "step": 51590 }, { "grad_norm": 0.38674837350845337, "learning_rate": 5.148955588488955e-05, "loss": 0.0104, "step": 51600 }, { "grad_norm": 0.15050454437732697, "learning_rate": 5.147302844455564e-05, "loss": 0.0083, "step": 51610 }, { "grad_norm": 0.09235630929470062, "learning_rate": 5.145650084313356e-05, "loss": 0.0094, "step": 51620 }, { "grad_norm": 0.11412666738033295, "learning_rate": 5.1439973082430734e-05, "loss": 0.0074, "step": 51630 }, { "grad_norm": 0.9641023874282837, "learning_rate": 5.142344516425461e-05, "loss": 0.0086, "step": 51640 }, { "grad_norm": 0.10946469008922577, "learning_rate": 5.1406917090412673e-05, "loss": 0.0063, "step": 51650 }, { "grad_norm": 0.12120849639177322, "learning_rate": 5.1390388862712404e-05, "loss": 0.0098, "step": 51660 }, { "grad_norm": 0.3964587152004242, "learning_rate": 5.137386048296129e-05, "loss": 0.0087, "step": 51670 }, { "grad_norm": 0.10861481726169586, "learning_rate": 5.135733195296686e-05, "loss": 0.0118, "step": 51680 }, { "grad_norm": 0.11668484658002853, "learning_rate": 5.134080327453664e-05, "loss": 0.0077, "step": 51690 }, { "grad_norm": 0.2364623099565506, "learning_rate": 5.1324274449478204e-05, "loss": 0.008, "step": 51700 }, { "grad_norm": 0.16013848781585693, "learning_rate": 5.130774547959909e-05, "loss": 0.0075, "step": 51710 }, { "grad_norm": 0.10063917934894562, "learning_rate": 5.129121636670692e-05, "loss": 0.0082, "step": 51720 }, { "grad_norm": 0.10958848893642426, "learning_rate": 5.1274687112609246e-05, "loss": 0.0069, "step": 51730 }, { "grad_norm": 0.22498203814029694, "learning_rate": 5.125815771911372e-05, "loss": 0.0114, "step": 51740 }, { "grad_norm": 0.13577629625797272, "learning_rate": 5.1241628188027956e-05, "loss": 0.0076, "step": 51750 }, { "grad_norm": 0.11858911067247391, "learning_rate": 5.1225098521159585e-05, "loss": 0.0082, "step": 51760 }, { "grad_norm": 0.14087960124015808, "learning_rate": 5.12085687203163e-05, "loss": 0.0083, "step": 51770 }, { "grad_norm": 0.1418423056602478, "learning_rate": 5.119203878730575e-05, "loss": 0.0096, "step": 51780 }, { "grad_norm": 0.09555013477802277, "learning_rate": 5.117550872393564e-05, "loss": 0.0085, "step": 51790 }, { "grad_norm": 0.13228781521320343, "learning_rate": 5.1158978532013645e-05, "loss": 0.0082, "step": 51800 }, { "grad_norm": 0.1776932328939438, "learning_rate": 5.114244821334751e-05, "loss": 0.0075, "step": 51810 }, { "grad_norm": 0.13097722828388214, "learning_rate": 5.112591776974497e-05, "loss": 0.0095, "step": 51820 }, { "grad_norm": 0.11693930625915527, "learning_rate": 5.110938720301374e-05, "loss": 0.0063, "step": 51830 }, { "grad_norm": 0.15419550240039825, "learning_rate": 5.1092856514961595e-05, "loss": 0.0083, "step": 51840 }, { "grad_norm": 0.19528278708457947, "learning_rate": 5.107632570739631e-05, "loss": 0.0083, "step": 51850 }, { "grad_norm": 0.139987051486969, "learning_rate": 5.1059794782125645e-05, "loss": 0.008, "step": 51860 }, { "grad_norm": 0.19654987752437592, "learning_rate": 5.104326374095744e-05, "loss": 0.0105, "step": 51870 }, { "grad_norm": 0.14822609722614288, "learning_rate": 5.102673258569947e-05, "loss": 0.0102, "step": 51880 }, { "grad_norm": 0.32907530665397644, "learning_rate": 5.101020131815956e-05, "loss": 0.0089, "step": 51890 }, { "grad_norm": 0.2778848707675934, "learning_rate": 5.099366994014556e-05, "loss": 0.0073, "step": 51900 }, { "grad_norm": 0.12537695467472076, "learning_rate": 5.097713845346532e-05, "loss": 0.0112, "step": 51910 }, { "grad_norm": 0.14738067984580994, "learning_rate": 5.096060685992668e-05, "loss": 0.0081, "step": 51920 }, { "grad_norm": 0.1817896068096161, "learning_rate": 5.0944075161337514e-05, "loss": 0.0086, "step": 51930 }, { "grad_norm": 0.1387162208557129, "learning_rate": 5.09275433595057e-05, "loss": 0.0079, "step": 51940 }, { "grad_norm": 0.1324559450149536, "learning_rate": 5.0911011456239157e-05, "loss": 0.0103, "step": 51950 }, { "grad_norm": 0.16269883513450623, "learning_rate": 5.089447945334574e-05, "loss": 0.0086, "step": 51960 }, { "grad_norm": 0.13033908605575562, "learning_rate": 5.087794735263343e-05, "loss": 0.0136, "step": 51970 }, { "grad_norm": 0.09932519495487213, "learning_rate": 5.086141515591011e-05, "loss": 0.0098, "step": 51980 }, { "grad_norm": 0.24506299197673798, "learning_rate": 5.084488286498371e-05, "loss": 0.0104, "step": 51990 }, { "grad_norm": 0.1815008819103241, "learning_rate": 5.08283504816622e-05, "loss": 0.0096, "step": 52000 }, { "grad_norm": 0.1363763064146042, "learning_rate": 5.081181800775353e-05, "loss": 0.012, "step": 52010 }, { "grad_norm": 0.5293045043945312, "learning_rate": 5.079528544506565e-05, "loss": 0.0121, "step": 52020 }, { "grad_norm": 0.16130350530147552, "learning_rate": 5.077875279540657e-05, "loss": 0.0091, "step": 52030 }, { "grad_norm": 0.14164741337299347, "learning_rate": 5.076222006058424e-05, "loss": 0.0089, "step": 52040 }, { "grad_norm": 0.14158639311790466, "learning_rate": 5.074568724240667e-05, "loss": 0.0117, "step": 52050 }, { "grad_norm": 0.11197325587272644, "learning_rate": 5.072915434268187e-05, "loss": 0.0084, "step": 52060 }, { "grad_norm": 0.15699222683906555, "learning_rate": 5.071262136321785e-05, "loss": 0.0096, "step": 52070 }, { "grad_norm": 0.21750065684318542, "learning_rate": 5.069608830582262e-05, "loss": 0.0094, "step": 52080 }, { "grad_norm": 0.13699038326740265, "learning_rate": 5.067955517230423e-05, "loss": 0.0106, "step": 52090 }, { "grad_norm": 0.16609714925289154, "learning_rate": 5.06630219644707e-05, "loss": 0.009, "step": 52100 }, { "grad_norm": 0.8193879723548889, "learning_rate": 5.064648868413008e-05, "loss": 0.0102, "step": 52110 }, { "grad_norm": 0.16281796991825104, "learning_rate": 5.062995533309044e-05, "loss": 0.01, "step": 52120 }, { "grad_norm": 0.1760789453983307, "learning_rate": 5.061342191315983e-05, "loss": 0.0103, "step": 52130 }, { "grad_norm": 0.16317786276340485, "learning_rate": 5.0596888426146325e-05, "loss": 0.0098, "step": 52140 }, { "grad_norm": 0.20305223762989044, "learning_rate": 5.058035487385798e-05, "loss": 0.009, "step": 52150 }, { "grad_norm": 0.1351558417081833, "learning_rate": 5.0563821258102915e-05, "loss": 0.0076, "step": 52160 }, { "grad_norm": 0.22384275496006012, "learning_rate": 5.05472875806892e-05, "loss": 0.0097, "step": 52170 }, { "grad_norm": 0.14636948704719543, "learning_rate": 5.053075384342494e-05, "loss": 0.0099, "step": 52180 }, { "grad_norm": 0.11984230577945709, "learning_rate": 5.0514220048118234e-05, "loss": 0.0058, "step": 52190 }, { "grad_norm": 0.09789709001779556, "learning_rate": 5.049768619657721e-05, "loss": 0.0082, "step": 52200 }, { "grad_norm": 0.19012139737606049, "learning_rate": 5.048115229060996e-05, "loss": 0.0088, "step": 52210 }, { "grad_norm": 0.3522137701511383, "learning_rate": 5.046461833202462e-05, "loss": 0.0088, "step": 52220 }, { "grad_norm": 0.1498325765132904, "learning_rate": 5.044808432262932e-05, "loss": 0.0073, "step": 52230 }, { "grad_norm": 0.15748406946659088, "learning_rate": 5.043155026423219e-05, "loss": 0.0092, "step": 52240 }, { "grad_norm": 0.147593691945076, "learning_rate": 5.041501615864137e-05, "loss": 0.0063, "step": 52250 }, { "grad_norm": 0.13507044315338135, "learning_rate": 5.039848200766502e-05, "loss": 0.0088, "step": 52260 }, { "grad_norm": 0.1375115066766739, "learning_rate": 5.038194781311125e-05, "loss": 0.0104, "step": 52270 }, { "grad_norm": 0.12521898746490479, "learning_rate": 5.0365413576788264e-05, "loss": 0.0082, "step": 52280 }, { "grad_norm": 0.11332489550113678, "learning_rate": 5.0348879300504194e-05, "loss": 0.0072, "step": 52290 }, { "grad_norm": 0.17691394686698914, "learning_rate": 5.033234498606721e-05, "loss": 0.0111, "step": 52300 }, { "grad_norm": 0.08740447461605072, "learning_rate": 5.0315810635285476e-05, "loss": 0.0081, "step": 52310 }, { "grad_norm": 0.3086510896682739, "learning_rate": 5.0299276249967164e-05, "loss": 0.0076, "step": 52320 }, { "grad_norm": 0.1034131646156311, "learning_rate": 5.0282741831920454e-05, "loss": 0.0088, "step": 52330 }, { "grad_norm": 0.34795957803726196, "learning_rate": 5.026620738295351e-05, "loss": 0.0102, "step": 52340 }, { "grad_norm": 0.26575762033462524, "learning_rate": 5.024967290487455e-05, "loss": 0.0063, "step": 52350 }, { "grad_norm": 0.10869703441858292, "learning_rate": 5.023313839949172e-05, "loss": 0.007, "step": 52360 }, { "grad_norm": 0.12506167590618134, "learning_rate": 5.0216603868613224e-05, "loss": 0.0064, "step": 52370 }, { "grad_norm": 0.3238191604614258, "learning_rate": 5.020006931404726e-05, "loss": 0.0085, "step": 52380 }, { "grad_norm": 0.12599706649780273, "learning_rate": 5.0183534737602014e-05, "loss": 0.0081, "step": 52390 }, { "grad_norm": 0.1488192230463028, "learning_rate": 5.0167000141085686e-05, "loss": 0.008, "step": 52400 }, { "grad_norm": 0.12374897301197052, "learning_rate": 5.015046552630647e-05, "loss": 0.0073, "step": 52410 }, { "grad_norm": 0.13288839161396027, "learning_rate": 5.013393089507258e-05, "loss": 0.0066, "step": 52420 }, { "grad_norm": 0.12424556910991669, "learning_rate": 5.011739624919219e-05, "loss": 0.0063, "step": 52430 }, { "grad_norm": 0.1172964945435524, "learning_rate": 5.0100861590473545e-05, "loss": 0.0075, "step": 52440 }, { "grad_norm": 0.1473148763179779, "learning_rate": 5.008432692072482e-05, "loss": 0.007, "step": 52450 }, { "grad_norm": 0.13203993439674377, "learning_rate": 5.006779224175423e-05, "loss": 0.0075, "step": 52460 }, { "grad_norm": 0.12001419067382812, "learning_rate": 5.005125755536998e-05, "loss": 0.0079, "step": 52470 }, { "grad_norm": 0.10780492424964905, "learning_rate": 5.003472286338029e-05, "loss": 0.0088, "step": 52480 }, { "grad_norm": 0.11992159485816956, "learning_rate": 5.001818816759335e-05, "loss": 0.0076, "step": 52490 }, { "grad_norm": 0.387404203414917, "learning_rate": 5.000165346981738e-05, "loss": 0.0088, "step": 52500 }, { "grad_norm": 0.13329961895942688, "learning_rate": 4.9985118771860595e-05, "loss": 0.0122, "step": 52510 }, { "grad_norm": 0.10279656946659088, "learning_rate": 4.996858407553119e-05, "loss": 0.0096, "step": 52520 }, { "grad_norm": 0.13818448781967163, "learning_rate": 4.99520493826374e-05, "loss": 0.0067, "step": 52530 }, { "grad_norm": 0.2036019265651703, "learning_rate": 4.9935514694987384e-05, "loss": 0.008, "step": 52540 }, { "grad_norm": 0.10810697823762894, "learning_rate": 4.991898001438941e-05, "loss": 0.009, "step": 52550 }, { "grad_norm": 0.19215114414691925, "learning_rate": 4.990244534265166e-05, "loss": 0.0067, "step": 52560 }, { "grad_norm": 0.20776651799678802, "learning_rate": 4.988591068158234e-05, "loss": 0.0089, "step": 52570 }, { "grad_norm": 0.6906319856643677, "learning_rate": 4.986937603298965e-05, "loss": 0.0082, "step": 52580 }, { "grad_norm": 0.15606363117694855, "learning_rate": 4.985284139868181e-05, "loss": 0.0077, "step": 52590 }, { "grad_norm": 0.14239871501922607, "learning_rate": 4.9836306780467e-05, "loss": 0.0083, "step": 52600 }, { "grad_norm": 0.16087715327739716, "learning_rate": 4.981977218015343e-05, "loss": 0.0137, "step": 52610 }, { "grad_norm": 0.14035692811012268, "learning_rate": 4.980323759954932e-05, "loss": 0.0119, "step": 52620 }, { "grad_norm": 0.2483954131603241, "learning_rate": 4.978670304046284e-05, "loss": 0.0109, "step": 52630 }, { "grad_norm": 0.21684809029102325, "learning_rate": 4.9770168504702195e-05, "loss": 0.0107, "step": 52640 }, { "grad_norm": 0.11850225925445557, "learning_rate": 4.975363399407559e-05, "loss": 0.0098, "step": 52650 }, { "grad_norm": 0.1088569313287735, "learning_rate": 4.9737099510391194e-05, "loss": 0.0125, "step": 52660 }, { "grad_norm": 0.17336304485797882, "learning_rate": 4.972056505545719e-05, "loss": 0.0163, "step": 52670 }, { "grad_norm": 0.1812116801738739, "learning_rate": 4.9704030631081775e-05, "loss": 0.0107, "step": 52680 }, { "grad_norm": 0.17232482135295868, "learning_rate": 4.968749623907313e-05, "loss": 0.0103, "step": 52690 }, { "grad_norm": 0.18366900086402893, "learning_rate": 4.967096188123943e-05, "loss": 0.0098, "step": 52700 }, { "grad_norm": 0.16359908878803253, "learning_rate": 4.965442755938884e-05, "loss": 0.0075, "step": 52710 }, { "grad_norm": 0.14736415445804596, "learning_rate": 4.963789327532954e-05, "loss": 0.0096, "step": 52720 }, { "grad_norm": 0.4472680985927582, "learning_rate": 4.962135903086968e-05, "loss": 0.0148, "step": 52730 }, { "grad_norm": 0.2251555174589157, "learning_rate": 4.960482482781741e-05, "loss": 0.0109, "step": 52740 }, { "grad_norm": 0.1533697098493576, "learning_rate": 4.958829066798092e-05, "loss": 0.0093, "step": 52750 }, { "grad_norm": 0.2162429690361023, "learning_rate": 4.957175655316835e-05, "loss": 0.0104, "step": 52760 }, { "grad_norm": 0.1465611308813095, "learning_rate": 4.9555222485187834e-05, "loss": 0.0105, "step": 52770 }, { "grad_norm": 0.1940641701221466, "learning_rate": 4.953868846584752e-05, "loss": 0.0125, "step": 52780 }, { "grad_norm": 0.1450962871313095, "learning_rate": 4.9522154496955544e-05, "loss": 0.0098, "step": 52790 }, { "grad_norm": 0.12885645031929016, "learning_rate": 4.950562058032001e-05, "loss": 0.0136, "step": 52800 }, { "grad_norm": 0.1158093810081482, "learning_rate": 4.948908671774909e-05, "loss": 0.0086, "step": 52810 }, { "grad_norm": 0.13248911499977112, "learning_rate": 4.947255291105087e-05, "loss": 0.0097, "step": 52820 }, { "grad_norm": 0.1441650390625, "learning_rate": 4.945601916203348e-05, "loss": 0.0096, "step": 52830 }, { "grad_norm": 0.20267100632190704, "learning_rate": 4.9439485472505016e-05, "loss": 0.0146, "step": 52840 }, { "grad_norm": 0.11894859373569489, "learning_rate": 4.942295184427356e-05, "loss": 0.0105, "step": 52850 }, { "grad_norm": 0.1359253227710724, "learning_rate": 4.9406418279147234e-05, "loss": 0.0104, "step": 52860 }, { "grad_norm": 0.16021551191806793, "learning_rate": 4.9389884778934116e-05, "loss": 0.0128, "step": 52870 }, { "grad_norm": 0.15995268523693085, "learning_rate": 4.937335134544228e-05, "loss": 0.0115, "step": 52880 }, { "grad_norm": 0.15148447453975677, "learning_rate": 4.935681798047979e-05, "loss": 0.0082, "step": 52890 }, { "grad_norm": 0.1284741312265396, "learning_rate": 4.934028468585473e-05, "loss": 0.0066, "step": 52900 }, { "grad_norm": 0.13894721865653992, "learning_rate": 4.932375146337514e-05, "loss": 0.0083, "step": 52910 }, { "grad_norm": 0.14318683743476868, "learning_rate": 4.930721831484908e-05, "loss": 0.0075, "step": 52920 }, { "grad_norm": 0.1426883190870285, "learning_rate": 4.929068524208456e-05, "loss": 0.0086, "step": 52930 }, { "grad_norm": 0.1659451574087143, "learning_rate": 4.927415224688965e-05, "loss": 0.0103, "step": 52940 }, { "grad_norm": 0.15911537408828735, "learning_rate": 4.9257619331072375e-05, "loss": 0.0121, "step": 52950 }, { "grad_norm": 0.1073007732629776, "learning_rate": 4.924108649644072e-05, "loss": 0.0086, "step": 52960 }, { "grad_norm": 0.11447753757238388, "learning_rate": 4.922455374480271e-05, "loss": 0.0131, "step": 52970 }, { "grad_norm": 0.1957174688577652, "learning_rate": 4.920802107796634e-05, "loss": 0.0106, "step": 52980 }, { "grad_norm": 0.2542547583580017, "learning_rate": 4.9191488497739564e-05, "loss": 0.0094, "step": 52990 }, { "grad_norm": 0.24425625801086426, "learning_rate": 4.917495600593042e-05, "loss": 0.0123, "step": 53000 }, { "grad_norm": 0.23386435210704803, "learning_rate": 4.915842360434684e-05, "loss": 0.0133, "step": 53010 }, { "grad_norm": 0.20801697671413422, "learning_rate": 4.914189129479679e-05, "loss": 0.0124, "step": 53020 }, { "grad_norm": 0.17933310568332672, "learning_rate": 4.912535907908822e-05, "loss": 0.0131, "step": 53030 }, { "grad_norm": 0.13017050921916962, "learning_rate": 4.910882695902907e-05, "loss": 0.0117, "step": 53040 }, { "grad_norm": 0.12763024866580963, "learning_rate": 4.9092294936427234e-05, "loss": 0.0123, "step": 53050 }, { "grad_norm": 0.19199015200138092, "learning_rate": 4.9075763013090685e-05, "loss": 0.0094, "step": 53060 }, { "grad_norm": 0.13988588750362396, "learning_rate": 4.9059231190827305e-05, "loss": 0.0069, "step": 53070 }, { "grad_norm": 0.16891467571258545, "learning_rate": 4.904269947144499e-05, "loss": 0.0092, "step": 53080 }, { "grad_norm": 0.12128855288028717, "learning_rate": 4.902616785675161e-05, "loss": 0.0074, "step": 53090 }, { "grad_norm": 0.19383828341960907, "learning_rate": 4.900963634855505e-05, "loss": 0.009, "step": 53100 }, { "grad_norm": 0.11553213000297546, "learning_rate": 4.8993104948663173e-05, "loss": 0.0109, "step": 53110 }, { "grad_norm": 0.1822350174188614, "learning_rate": 4.8976573658883823e-05, "loss": 0.0075, "step": 53120 }, { "grad_norm": 0.1688367873430252, "learning_rate": 4.896004248102484e-05, "loss": 0.0098, "step": 53130 }, { "grad_norm": 0.1946473866701126, "learning_rate": 4.8943511416894045e-05, "loss": 0.0085, "step": 53140 }, { "grad_norm": 0.14290350675582886, "learning_rate": 4.892698046829926e-05, "loss": 0.0112, "step": 53150 }, { "grad_norm": 0.14505186676979065, "learning_rate": 4.8910449637048275e-05, "loss": 0.0103, "step": 53160 }, { "grad_norm": 0.8525344133377075, "learning_rate": 4.8893918924948884e-05, "loss": 0.009, "step": 53170 }, { "grad_norm": 0.16962502896785736, "learning_rate": 4.8877388333808835e-05, "loss": 0.0142, "step": 53180 }, { "grad_norm": 0.1963646560907364, "learning_rate": 4.8860857865435925e-05, "loss": 0.0152, "step": 53190 }, { "grad_norm": 0.2994474172592163, "learning_rate": 4.884432752163789e-05, "loss": 0.0107, "step": 53200 }, { "grad_norm": 0.21839690208435059, "learning_rate": 4.8827797304222466e-05, "loss": 0.0125, "step": 53210 }, { "grad_norm": 0.12381619215011597, "learning_rate": 4.8811267214997366e-05, "loss": 0.0074, "step": 53220 }, { "grad_norm": 0.1764136105775833, "learning_rate": 4.879473725577029e-05, "loss": 0.0066, "step": 53230 }, { "grad_norm": 0.13873684406280518, "learning_rate": 4.877820742834891e-05, "loss": 0.0091, "step": 53240 }, { "grad_norm": 0.11736984550952911, "learning_rate": 4.8761677734540956e-05, "loss": 0.0082, "step": 53250 }, { "grad_norm": 0.14454810321331024, "learning_rate": 4.874514817615407e-05, "loss": 0.0115, "step": 53260 }, { "grad_norm": 0.17779454588890076, "learning_rate": 4.872861875499588e-05, "loss": 0.0098, "step": 53270 }, { "grad_norm": 0.1483507603406906, "learning_rate": 4.871208947287404e-05, "loss": 0.0071, "step": 53280 }, { "grad_norm": 0.2119586020708084, "learning_rate": 4.869556033159615e-05, "loss": 0.0092, "step": 53290 }, { "grad_norm": 0.10971515625715256, "learning_rate": 4.8679031332969835e-05, "loss": 0.0087, "step": 53300 }, { "grad_norm": 0.11054138094186783, "learning_rate": 4.8662502478802635e-05, "loss": 0.0089, "step": 53310 }, { "grad_norm": 1.5488088130950928, "learning_rate": 4.8645973770902176e-05, "loss": 0.0147, "step": 53320 }, { "grad_norm": 0.12317867577075958, "learning_rate": 4.862944521107599e-05, "loss": 0.0112, "step": 53330 }, { "grad_norm": 0.08678693324327469, "learning_rate": 4.861291680113162e-05, "loss": 0.0061, "step": 53340 }, { "grad_norm": 0.3644050359725952, "learning_rate": 4.8596388542876563e-05, "loss": 0.0103, "step": 53350 }, { "grad_norm": 0.1013060137629509, "learning_rate": 4.857986043811836e-05, "loss": 0.0067, "step": 53360 }, { "grad_norm": 0.2803119421005249, "learning_rate": 4.856333248866448e-05, "loss": 0.0084, "step": 53370 }, { "grad_norm": 0.11685469001531601, "learning_rate": 4.85468046963224e-05, "loss": 0.0085, "step": 53380 }, { "grad_norm": 0.0798318162560463, "learning_rate": 4.853027706289957e-05, "loss": 0.0099, "step": 53390 }, { "grad_norm": 0.17368222773075104, "learning_rate": 4.851374959020344e-05, "loss": 0.009, "step": 53400 }, { "grad_norm": 0.134990856051445, "learning_rate": 4.849722228004141e-05, "loss": 0.0068, "step": 53410 }, { "grad_norm": 0.19066402316093445, "learning_rate": 4.848069513422091e-05, "loss": 0.0096, "step": 53420 }, { "grad_norm": 0.10496452450752258, "learning_rate": 4.8464168154549294e-05, "loss": 0.0083, "step": 53430 }, { "grad_norm": 0.14179711043834686, "learning_rate": 4.844764134283392e-05, "loss": 0.0084, "step": 53440 }, { "grad_norm": 0.09851684421300888, "learning_rate": 4.843111470088218e-05, "loss": 0.0097, "step": 53450 }, { "grad_norm": 0.1887170672416687, "learning_rate": 4.841458823050137e-05, "loss": 0.0111, "step": 53460 }, { "grad_norm": 0.10322446376085281, "learning_rate": 4.8398061933498816e-05, "loss": 0.0086, "step": 53470 }, { "grad_norm": 0.213486909866333, "learning_rate": 4.83815358116818e-05, "loss": 0.0081, "step": 53480 }, { "grad_norm": 0.2005520462989807, "learning_rate": 4.8365009866857584e-05, "loss": 0.011, "step": 53490 }, { "grad_norm": 0.12852421402931213, "learning_rate": 4.834848410083342e-05, "loss": 0.0071, "step": 53500 }, { "grad_norm": 0.13813620805740356, "learning_rate": 4.833195851541657e-05, "loss": 0.0088, "step": 53510 }, { "grad_norm": 0.16056737303733826, "learning_rate": 4.831543311241423e-05, "loss": 0.0094, "step": 53520 }, { "grad_norm": 0.2245536595582962, "learning_rate": 4.829890789363359e-05, "loss": 0.0128, "step": 53530 }, { "grad_norm": 0.12855733931064606, "learning_rate": 4.828238286088182e-05, "loss": 0.0079, "step": 53540 }, { "grad_norm": 0.15536893904209137, "learning_rate": 4.826585801596609e-05, "loss": 0.0161, "step": 53550 }, { "grad_norm": 0.11740891635417938, "learning_rate": 4.8249333360693486e-05, "loss": 0.0071, "step": 53560 }, { "grad_norm": 0.28493165969848633, "learning_rate": 4.8232808896871185e-05, "loss": 0.0125, "step": 53570 }, { "grad_norm": 0.12289149314165115, "learning_rate": 4.821628462630624e-05, "loss": 0.0081, "step": 53580 }, { "grad_norm": 0.33140331506729126, "learning_rate": 4.8199760550805735e-05, "loss": 0.0101, "step": 53590 }, { "grad_norm": 0.19302918016910553, "learning_rate": 4.818323667217669e-05, "loss": 0.011, "step": 53600 }, { "grad_norm": 0.15809744596481323, "learning_rate": 4.8166712992226165e-05, "loss": 0.011, "step": 53610 }, { "grad_norm": 0.146834135055542, "learning_rate": 4.8150189512761154e-05, "loss": 0.0075, "step": 53620 }, { "grad_norm": 0.12380270659923553, "learning_rate": 4.813366623558862e-05, "loss": 0.0071, "step": 53630 }, { "grad_norm": 0.13932767510414124, "learning_rate": 4.811714316251554e-05, "loss": 0.007, "step": 53640 }, { "grad_norm": 0.2153005450963974, "learning_rate": 4.810062029534886e-05, "loss": 0.0091, "step": 53650 }, { "grad_norm": 0.14404620230197906, "learning_rate": 4.8084097635895505e-05, "loss": 0.0096, "step": 53660 }, { "grad_norm": 0.12136553972959518, "learning_rate": 4.8067575185962335e-05, "loss": 0.0059, "step": 53670 }, { "grad_norm": 0.13308273255825043, "learning_rate": 4.805105294735625e-05, "loss": 0.0096, "step": 53680 }, { "grad_norm": 0.15429368615150452, "learning_rate": 4.803453092188406e-05, "loss": 0.0079, "step": 53690 }, { "grad_norm": 0.12847965955734253, "learning_rate": 4.801800911135263e-05, "loss": 0.0095, "step": 53700 }, { "grad_norm": 0.11348666250705719, "learning_rate": 4.8001487517568744e-05, "loss": 0.0086, "step": 53710 }, { "grad_norm": 0.12028060108423233, "learning_rate": 4.798496614233918e-05, "loss": 0.0083, "step": 53720 }, { "grad_norm": 0.11201376467943192, "learning_rate": 4.796844498747069e-05, "loss": 0.0074, "step": 53730 }, { "grad_norm": 0.12052884697914124, "learning_rate": 4.795192405477e-05, "loss": 0.0075, "step": 53740 }, { "grad_norm": 0.25218915939331055, "learning_rate": 4.79354033460438e-05, "loss": 0.0067, "step": 53750 }, { "grad_norm": 0.18566612899303436, "learning_rate": 4.79188828630988e-05, "loss": 0.0094, "step": 53760 }, { "grad_norm": 0.1659799963235855, "learning_rate": 4.790236260774165e-05, "loss": 0.0104, "step": 53770 }, { "grad_norm": 0.1708069145679474, "learning_rate": 4.788584258177896e-05, "loss": 0.0062, "step": 53780 }, { "grad_norm": 0.15360595285892487, "learning_rate": 4.7869322787017356e-05, "loss": 0.0062, "step": 53790 }, { "grad_norm": 0.09970257431268692, "learning_rate": 4.78528032252634e-05, "loss": 0.0087, "step": 53800 }, { "grad_norm": 0.13756363093852997, "learning_rate": 4.783628389832366e-05, "loss": 0.0095, "step": 53810 }, { "grad_norm": 0.1472741663455963, "learning_rate": 4.7819764808004644e-05, "loss": 0.0072, "step": 53820 }, { "grad_norm": 0.1398884505033493, "learning_rate": 4.780324595611289e-05, "loss": 0.0068, "step": 53830 }, { "grad_norm": 0.1612146943807602, "learning_rate": 4.7786727344454854e-05, "loss": 0.0066, "step": 53840 }, { "grad_norm": 0.14434367418289185, "learning_rate": 4.777020897483697e-05, "loss": 0.0072, "step": 53850 }, { "grad_norm": 0.13504284620285034, "learning_rate": 4.775369084906569e-05, "loss": 0.0093, "step": 53860 }, { "grad_norm": 0.15264204144477844, "learning_rate": 4.773717296894742e-05, "loss": 0.0067, "step": 53870 }, { "grad_norm": 0.12948577105998993, "learning_rate": 4.772065533628848e-05, "loss": 0.0057, "step": 53880 }, { "grad_norm": 0.10556609183549881, "learning_rate": 4.7704137952895265e-05, "loss": 0.0093, "step": 53890 }, { "grad_norm": 0.13905231654644012, "learning_rate": 4.768762082057407e-05, "loss": 0.0077, "step": 53900 }, { "grad_norm": 0.12068002671003342, "learning_rate": 4.7671103941131186e-05, "loss": 0.0073, "step": 53910 }, { "grad_norm": 0.16428668797016144, "learning_rate": 4.765458731637287e-05, "loss": 0.0074, "step": 53920 }, { "grad_norm": 0.19638536870479584, "learning_rate": 4.763807094810537e-05, "loss": 0.0071, "step": 53930 }, { "grad_norm": 0.13112585246562958, "learning_rate": 4.762155483813487e-05, "loss": 0.0092, "step": 53940 }, { "grad_norm": 0.12094597518444061, "learning_rate": 4.760503898826754e-05, "loss": 0.016, "step": 53950 }, { "grad_norm": 0.12298005819320679, "learning_rate": 4.758852340030957e-05, "loss": 0.0103, "step": 53960 }, { "grad_norm": 0.1682283878326416, "learning_rate": 4.757200807606705e-05, "loss": 0.0092, "step": 53970 }, { "grad_norm": 0.17910797894001007, "learning_rate": 4.755549301734607e-05, "loss": 0.0106, "step": 53980 }, { "grad_norm": 0.21587681770324707, "learning_rate": 4.753897822595271e-05, "loss": 0.0078, "step": 53990 }, { "grad_norm": 0.5111157298088074, "learning_rate": 4.7522463703692994e-05, "loss": 0.0115, "step": 54000 }, { "grad_norm": 0.1262611597776413, "learning_rate": 4.75059494523729e-05, "loss": 0.0073, "step": 54010 }, { "grad_norm": 0.15405254065990448, "learning_rate": 4.748943547379843e-05, "loss": 0.0073, "step": 54020 }, { "grad_norm": 0.14600373804569244, "learning_rate": 4.7472921769775535e-05, "loss": 0.007, "step": 54030 }, { "grad_norm": 0.20782074332237244, "learning_rate": 4.745640834211011e-05, "loss": 0.0098, "step": 54040 }, { "grad_norm": 0.30281174182891846, "learning_rate": 4.7439895192608054e-05, "loss": 0.0096, "step": 54050 }, { "grad_norm": 0.13162589073181152, "learning_rate": 4.7423382323075186e-05, "loss": 0.0114, "step": 54060 }, { "grad_norm": 0.13016241788864136, "learning_rate": 4.740686973531737e-05, "loss": 0.0097, "step": 54070 }, { "grad_norm": 0.16484013199806213, "learning_rate": 4.7390357431140375e-05, "loss": 0.0111, "step": 54080 }, { "grad_norm": 0.1488296091556549, "learning_rate": 4.737384541234998e-05, "loss": 0.0091, "step": 54090 }, { "grad_norm": 0.09556151926517487, "learning_rate": 4.735733368075189e-05, "loss": 0.0083, "step": 54100 }, { "grad_norm": 0.14663474261760712, "learning_rate": 4.734082223815182e-05, "loss": 0.0104, "step": 54110 }, { "grad_norm": 0.16243889927864075, "learning_rate": 4.7324311086355436e-05, "loss": 0.0085, "step": 54120 }, { "grad_norm": 0.18224555253982544, "learning_rate": 4.730780022716836e-05, "loss": 0.0093, "step": 54130 }, { "grad_norm": 0.28979265689849854, "learning_rate": 4.729128966239619e-05, "loss": 0.0115, "step": 54140 }, { "grad_norm": 0.4491150379180908, "learning_rate": 4.7274779393844536e-05, "loss": 0.0092, "step": 54150 }, { "grad_norm": 0.1365322321653366, "learning_rate": 4.725826942331891e-05, "loss": 0.0089, "step": 54160 }, { "grad_norm": 0.11007527261972427, "learning_rate": 4.724175975262481e-05, "loss": 0.0094, "step": 54170 }, { "grad_norm": 0.11544694006443024, "learning_rate": 4.722525038356772e-05, "loss": 0.0092, "step": 54180 }, { "grad_norm": 0.08685485273599625, "learning_rate": 4.7208741317953075e-05, "loss": 0.0072, "step": 54190 }, { "grad_norm": 0.12678438425064087, "learning_rate": 4.7192232557586264e-05, "loss": 0.0081, "step": 54200 }, { "grad_norm": 0.10236048698425293, "learning_rate": 4.717572410427269e-05, "loss": 0.0069, "step": 54210 }, { "grad_norm": 0.22420050203800201, "learning_rate": 4.71592159598177e-05, "loss": 0.008, "step": 54220 }, { "grad_norm": 0.13720344007015228, "learning_rate": 4.714270812602657e-05, "loss": 0.0102, "step": 54230 }, { "grad_norm": 0.10112904012203217, "learning_rate": 4.712620060470458e-05, "loss": 0.0126, "step": 54240 }, { "grad_norm": 0.12698031961917877, "learning_rate": 4.7109693397656986e-05, "loss": 0.011, "step": 54250 }, { "grad_norm": 0.14067301154136658, "learning_rate": 4.709318650668894e-05, "loss": 0.0077, "step": 54260 }, { "grad_norm": 2.669010639190674, "learning_rate": 4.7076679933605676e-05, "loss": 0.0116, "step": 54270 }, { "grad_norm": 0.10848080366849899, "learning_rate": 4.70601736802123e-05, "loss": 0.0108, "step": 54280 }, { "grad_norm": 0.1228441596031189, "learning_rate": 4.704366774831391e-05, "loss": 0.0074, "step": 54290 }, { "grad_norm": 0.3009294867515564, "learning_rate": 4.702716213971557e-05, "loss": 0.0091, "step": 54300 }, { "grad_norm": 0.2039751410484314, "learning_rate": 4.70106568562223e-05, "loss": 0.0089, "step": 54310 }, { "grad_norm": 0.1580435186624527, "learning_rate": 4.6994151899639105e-05, "loss": 0.0119, "step": 54320 }, { "grad_norm": 0.17874591052532196, "learning_rate": 4.697764727177093e-05, "loss": 0.0094, "step": 54330 }, { "grad_norm": 0.19161133468151093, "learning_rate": 4.696114297442272e-05, "loss": 0.0106, "step": 54340 }, { "grad_norm": 0.36453041434288025, "learning_rate": 4.694463900939933e-05, "loss": 0.0132, "step": 54350 }, { "grad_norm": 0.18292096257209778, "learning_rate": 4.692813537850564e-05, "loss": 0.0124, "step": 54360 }, { "grad_norm": 0.16461718082427979, "learning_rate": 4.6911632083546445e-05, "loss": 0.0118, "step": 54370 }, { "grad_norm": 0.1964336484670639, "learning_rate": 4.6895129126326525e-05, "loss": 0.0083, "step": 54380 }, { "grad_norm": 0.16256386041641235, "learning_rate": 4.687862650865059e-05, "loss": 0.0086, "step": 54390 }, { "grad_norm": 0.13097865879535675, "learning_rate": 4.686212423232339e-05, "loss": 0.0068, "step": 54400 }, { "grad_norm": 0.15809594094753265, "learning_rate": 4.6845622299149565e-05, "loss": 0.0064, "step": 54410 }, { "grad_norm": 0.16247262060642242, "learning_rate": 4.682912071093374e-05, "loss": 0.0116, "step": 54420 }, { "grad_norm": 0.20050443708896637, "learning_rate": 4.681261946948052e-05, "loss": 0.0076, "step": 54430 }, { "grad_norm": 0.22409555315971375, "learning_rate": 4.679611857659443e-05, "loss": 0.0074, "step": 54440 }, { "grad_norm": 0.1533021628856659, "learning_rate": 4.6779618034079975e-05, "loss": 0.009, "step": 54450 }, { "grad_norm": 0.1278146356344223, "learning_rate": 4.6763117843741675e-05, "loss": 0.0085, "step": 54460 }, { "grad_norm": 0.2704311013221741, "learning_rate": 4.674661800738394e-05, "loss": 0.008, "step": 54470 }, { "grad_norm": 0.13204757869243622, "learning_rate": 4.673011852681117e-05, "loss": 0.0092, "step": 54480 }, { "grad_norm": 0.1921187937259674, "learning_rate": 4.671361940382772e-05, "loss": 0.0105, "step": 54490 }, { "grad_norm": 0.1510177105665207, "learning_rate": 4.669712064023791e-05, "loss": 0.0094, "step": 54500 }, { "grad_norm": 0.12585248053073883, "learning_rate": 4.668062223784602e-05, "loss": 0.0087, "step": 54510 }, { "grad_norm": 0.1726580411195755, "learning_rate": 4.6664124198456274e-05, "loss": 0.0067, "step": 54520 }, { "grad_norm": 0.17375028133392334, "learning_rate": 4.6647626523872905e-05, "loss": 0.0082, "step": 54530 }, { "grad_norm": 0.163260817527771, "learning_rate": 4.6631129215900066e-05, "loss": 0.0091, "step": 54540 }, { "grad_norm": 0.13826943933963776, "learning_rate": 4.661463227634187e-05, "loss": 0.0116, "step": 54550 }, { "grad_norm": 0.14011012017726898, "learning_rate": 4.659813570700238e-05, "loss": 0.0098, "step": 54560 }, { "grad_norm": 0.09074294567108154, "learning_rate": 4.658163950968568e-05, "loss": 0.0093, "step": 54570 }, { "grad_norm": 0.16349656879901886, "learning_rate": 4.6565143686195726e-05, "loss": 0.0081, "step": 54580 }, { "grad_norm": 0.45881542563438416, "learning_rate": 4.654864823833649e-05, "loss": 0.0098, "step": 54590 }, { "grad_norm": 0.1577688455581665, "learning_rate": 4.65321531679119e-05, "loss": 0.0067, "step": 54600 }, { "grad_norm": 0.15306907892227173, "learning_rate": 4.6515658476725834e-05, "loss": 0.0071, "step": 54610 }, { "grad_norm": 0.119349904358387, "learning_rate": 4.649916416658212e-05, "loss": 0.0078, "step": 54620 }, { "grad_norm": 0.17419715225696564, "learning_rate": 4.648267023928455e-05, "loss": 0.0074, "step": 54630 }, { "grad_norm": 0.16241343319416046, "learning_rate": 4.646617669663687e-05, "loss": 0.0105, "step": 54640 }, { "grad_norm": 0.1611316055059433, "learning_rate": 4.644968354044277e-05, "loss": 0.0084, "step": 54650 }, { "grad_norm": 0.22330088913440704, "learning_rate": 4.643319077250596e-05, "loss": 0.0077, "step": 54660 }, { "grad_norm": 0.16603797674179077, "learning_rate": 4.6416698394630055e-05, "loss": 0.0076, "step": 54670 }, { "grad_norm": 0.14540192484855652, "learning_rate": 4.640020640861862e-05, "loss": 0.0087, "step": 54680 }, { "grad_norm": 0.12655580043792725, "learning_rate": 4.6383714816275195e-05, "loss": 0.0096, "step": 54690 }, { "grad_norm": 0.13486099243164062, "learning_rate": 4.636722361940329e-05, "loss": 0.0082, "step": 54700 }, { "grad_norm": 0.21174941956996918, "learning_rate": 4.635073281980632e-05, "loss": 0.0069, "step": 54710 }, { "grad_norm": 0.12044665962457657, "learning_rate": 4.633424241928775e-05, "loss": 0.0088, "step": 54720 }, { "grad_norm": 0.10428781062364578, "learning_rate": 4.6317752419650914e-05, "loss": 0.0094, "step": 54730 }, { "grad_norm": 0.14368924498558044, "learning_rate": 4.630126282269914e-05, "loss": 0.0093, "step": 54740 }, { "grad_norm": 0.1663886159658432, "learning_rate": 4.628477363023569e-05, "loss": 0.009, "step": 54750 }, { "grad_norm": 0.1747131049633026, "learning_rate": 4.626828484406383e-05, "loss": 0.0134, "step": 54760 }, { "grad_norm": 0.17067931592464447, "learning_rate": 4.625179646598669e-05, "loss": 0.0072, "step": 54770 }, { "grad_norm": 0.12857544422149658, "learning_rate": 4.6235308497807487e-05, "loss": 0.0058, "step": 54780 }, { "grad_norm": 0.08840381354093552, "learning_rate": 4.6218820941329276e-05, "loss": 0.0077, "step": 54790 }, { "grad_norm": 0.12155678123235703, "learning_rate": 4.620233379835513e-05, "loss": 0.0094, "step": 54800 }, { "grad_norm": 0.20808973908424377, "learning_rate": 4.618584707068804e-05, "loss": 0.0107, "step": 54810 }, { "grad_norm": 0.17756694555282593, "learning_rate": 4.616936076013099e-05, "loss": 0.0091, "step": 54820 }, { "grad_norm": 0.1437169462442398, "learning_rate": 4.615287486848688e-05, "loss": 0.0087, "step": 54830 }, { "grad_norm": 0.1302165538072586, "learning_rate": 4.6136389397558584e-05, "loss": 0.0079, "step": 54840 }, { "grad_norm": 0.11825666576623917, "learning_rate": 4.6119904349148935e-05, "loss": 0.0099, "step": 54850 }, { "grad_norm": 0.16166427731513977, "learning_rate": 4.6103419725060726e-05, "loss": 0.0087, "step": 54860 }, { "grad_norm": 0.13464875519275665, "learning_rate": 4.608693552709667e-05, "loss": 0.0103, "step": 54870 }, { "grad_norm": 0.1260603815317154, "learning_rate": 4.607045175705947e-05, "loss": 0.0093, "step": 54880 }, { "grad_norm": 0.42770543694496155, "learning_rate": 4.605396841675176e-05, "loss": 0.0123, "step": 54890 }, { "grad_norm": 0.1318567395210266, "learning_rate": 4.6037485507976094e-05, "loss": 0.0073, "step": 54900 }, { "grad_norm": 0.11779389530420303, "learning_rate": 4.602100303253509e-05, "loss": 0.0105, "step": 54910 }, { "grad_norm": 0.1444023698568344, "learning_rate": 4.600452099223122e-05, "loss": 0.0115, "step": 54920 }, { "grad_norm": 0.41361695528030396, "learning_rate": 4.598803938886691e-05, "loss": 0.0103, "step": 54930 }, { "grad_norm": 0.12439300864934921, "learning_rate": 4.597155822424459e-05, "loss": 0.0083, "step": 54940 }, { "grad_norm": 0.11025860160589218, "learning_rate": 4.59550775001666e-05, "loss": 0.007, "step": 54950 }, { "grad_norm": 0.17875924706459045, "learning_rate": 4.5938597218435234e-05, "loss": 0.0077, "step": 54960 }, { "grad_norm": 0.15065059065818787, "learning_rate": 4.592211738085279e-05, "loss": 0.0082, "step": 54970 }, { "grad_norm": 1.0731678009033203, "learning_rate": 4.590563798922146e-05, "loss": 0.0101, "step": 54980 }, { "grad_norm": 0.5121635794639587, "learning_rate": 4.5889159045343404e-05, "loss": 0.0089, "step": 54990 }, { "grad_norm": 0.14470364153385162, "learning_rate": 4.5872680551020734e-05, "loss": 0.0079, "step": 55000 }, { "grad_norm": 0.09176689386367798, "learning_rate": 4.585620250805552e-05, "loss": 0.0076, "step": 55010 }, { "grad_norm": 0.0732535719871521, "learning_rate": 4.583972491824975e-05, "loss": 0.006, "step": 55020 }, { "grad_norm": 0.3849174380302429, "learning_rate": 4.582324778340541e-05, "loss": 0.0094, "step": 55030 }, { "grad_norm": 0.19576884806156158, "learning_rate": 4.5806771105324424e-05, "loss": 0.011, "step": 55040 }, { "grad_norm": 0.10936469584703445, "learning_rate": 4.579029488580865e-05, "loss": 0.0071, "step": 55050 }, { "grad_norm": 0.10097318142652512, "learning_rate": 4.5773819126659876e-05, "loss": 0.0084, "step": 55060 }, { "grad_norm": 0.13838034868240356, "learning_rate": 4.575734382967991e-05, "loss": 0.0061, "step": 55070 }, { "grad_norm": 0.16823090612888336, "learning_rate": 4.574086899667044e-05, "loss": 0.0101, "step": 55080 }, { "grad_norm": 0.1244187280535698, "learning_rate": 4.5724394629433106e-05, "loss": 0.0073, "step": 55090 }, { "grad_norm": 0.17874208092689514, "learning_rate": 4.570792072976958e-05, "loss": 0.0075, "step": 55100 }, { "grad_norm": 0.1543922871351242, "learning_rate": 4.569144729948138e-05, "loss": 0.0089, "step": 55110 }, { "grad_norm": 0.1349722146987915, "learning_rate": 4.5674974340370033e-05, "loss": 0.0105, "step": 55120 }, { "grad_norm": 0.7475873231887817, "learning_rate": 4.565850185423699e-05, "loss": 0.0079, "step": 55130 }, { "grad_norm": 0.12029337882995605, "learning_rate": 4.564202984288365e-05, "loss": 0.0087, "step": 55140 }, { "grad_norm": 0.10128523409366608, "learning_rate": 4.5625558308111356e-05, "loss": 0.0071, "step": 55150 }, { "grad_norm": 0.132478266954422, "learning_rate": 4.560908725172145e-05, "loss": 0.0077, "step": 55160 }, { "grad_norm": 0.1643057018518448, "learning_rate": 4.559261667551516e-05, "loss": 0.0067, "step": 55170 }, { "grad_norm": 0.1185729131102562, "learning_rate": 4.5576146581293685e-05, "loss": 0.0099, "step": 55180 }, { "grad_norm": 0.152525395154953, "learning_rate": 4.5559676970858164e-05, "loss": 0.0079, "step": 55190 }, { "grad_norm": 0.6169993877410889, "learning_rate": 4.5543207846009705e-05, "loss": 0.015, "step": 55200 }, { "grad_norm": 0.29591232538223267, "learning_rate": 4.552673920854933e-05, "loss": 0.011, "step": 55210 }, { "grad_norm": 0.16646847128868103, "learning_rate": 4.5510271060278006e-05, "loss": 0.0091, "step": 55220 }, { "grad_norm": 0.16499698162078857, "learning_rate": 4.5493803402996704e-05, "loss": 0.0093, "step": 55230 }, { "grad_norm": 0.1603911966085434, "learning_rate": 4.5477336238506295e-05, "loss": 0.0088, "step": 55240 }, { "grad_norm": 0.16457781195640564, "learning_rate": 4.546086956860759e-05, "loss": 0.0086, "step": 55250 }, { "grad_norm": 0.1435052454471588, "learning_rate": 4.5444403395101374e-05, "loss": 0.0091, "step": 55260 }, { "grad_norm": 0.21454717218875885, "learning_rate": 4.542793771978834e-05, "loss": 0.008, "step": 55270 }, { "grad_norm": 0.16688868403434753, "learning_rate": 4.541147254446916e-05, "loss": 0.0087, "step": 55280 }, { "grad_norm": 0.12439963221549988, "learning_rate": 4.539500787094445e-05, "loss": 0.0068, "step": 55290 }, { "grad_norm": 0.1320614367723465, "learning_rate": 4.537854370101475e-05, "loss": 0.0082, "step": 55300 }, { "grad_norm": 0.16235867142677307, "learning_rate": 4.5362080036480566e-05, "loss": 0.0093, "step": 55310 }, { "grad_norm": 0.11677516996860504, "learning_rate": 4.534561687914235e-05, "loss": 0.0101, "step": 55320 }, { "grad_norm": 0.11516277492046356, "learning_rate": 4.5329154230800466e-05, "loss": 0.0069, "step": 55330 }, { "grad_norm": 0.17457711696624756, "learning_rate": 4.531269209325525e-05, "loss": 0.0079, "step": 55340 }, { "grad_norm": 0.12947601079940796, "learning_rate": 4.529623046830696e-05, "loss": 0.0111, "step": 55350 }, { "grad_norm": 0.1279706060886383, "learning_rate": 4.5279769357755866e-05, "loss": 0.0074, "step": 55360 }, { "grad_norm": 0.13573338091373444, "learning_rate": 4.5263308763402084e-05, "loss": 0.0106, "step": 55370 }, { "grad_norm": 0.117520272731781, "learning_rate": 4.524684868704574e-05, "loss": 0.0068, "step": 55380 }, { "grad_norm": 0.15912273526191711, "learning_rate": 4.523038913048688e-05, "loss": 0.0078, "step": 55390 }, { "grad_norm": 0.18689744174480438, "learning_rate": 4.521393009552549e-05, "loss": 0.0109, "step": 55400 }, { "grad_norm": 0.16099953651428223, "learning_rate": 4.5197471583961484e-05, "loss": 0.0096, "step": 55410 }, { "grad_norm": 0.13734027743339539, "learning_rate": 4.51810135975948e-05, "loss": 0.0107, "step": 55420 }, { "grad_norm": 0.15927690267562866, "learning_rate": 4.516455613822521e-05, "loss": 0.0124, "step": 55430 }, { "grad_norm": 0.17601333558559418, "learning_rate": 4.514809920765249e-05, "loss": 0.0101, "step": 55440 }, { "grad_norm": 1.3264374732971191, "learning_rate": 4.5131642807676356e-05, "loss": 0.0084, "step": 55450 }, { "grad_norm": 0.3955506384372711, "learning_rate": 4.511518694009644e-05, "loss": 0.0122, "step": 55460 }, { "grad_norm": 0.12009409815073013, "learning_rate": 4.509873160671231e-05, "loss": 0.0099, "step": 55470 }, { "grad_norm": 0.14535152912139893, "learning_rate": 4.508227680932356e-05, "loss": 0.007, "step": 55480 }, { "grad_norm": 0.17169909179210663, "learning_rate": 4.5065822549729605e-05, "loss": 0.0083, "step": 55490 }, { "grad_norm": 0.2436017394065857, "learning_rate": 4.504936882972989e-05, "loss": 0.008, "step": 55500 }, { "grad_norm": 0.13914090394973755, "learning_rate": 4.503291565112374e-05, "loss": 0.0076, "step": 55510 }, { "grad_norm": 0.2803422808647156, "learning_rate": 4.501646301571048e-05, "loss": 0.0086, "step": 55520 }, { "grad_norm": 0.10333769023418427, "learning_rate": 4.500001092528933e-05, "loss": 0.0063, "step": 55530 }, { "grad_norm": 0.11732829362154007, "learning_rate": 4.4983559381659455e-05, "loss": 0.0096, "step": 55540 }, { "grad_norm": 0.19492973387241364, "learning_rate": 4.496710838661999e-05, "loss": 0.0128, "step": 55550 }, { "grad_norm": 0.4801163971424103, "learning_rate": 4.495065794196999e-05, "loss": 0.0075, "step": 55560 }, { "grad_norm": 0.09895288199186325, "learning_rate": 4.4934208049508455e-05, "loss": 0.0068, "step": 55570 }, { "grad_norm": 0.2007027566432953, "learning_rate": 4.4917758711034303e-05, "loss": 0.0078, "step": 55580 }, { "grad_norm": 0.11366873234510422, "learning_rate": 4.490130992834642e-05, "loss": 0.0138, "step": 55590 }, { "grad_norm": 0.12016113847494125, "learning_rate": 4.488486170324359e-05, "loss": 0.0077, "step": 55600 }, { "grad_norm": 0.1315346211194992, "learning_rate": 4.486841403752463e-05, "loss": 0.0095, "step": 55610 }, { "grad_norm": 0.15750175714492798, "learning_rate": 4.4851966932988174e-05, "loss": 0.0089, "step": 55620 }, { "grad_norm": 0.2800150513648987, "learning_rate": 4.4835520391432894e-05, "loss": 0.0103, "step": 55630 }, { "grad_norm": 0.12651921808719635, "learning_rate": 4.481907441465733e-05, "loss": 0.009, "step": 55640 }, { "grad_norm": 0.20327460765838623, "learning_rate": 4.480262900446e-05, "loss": 0.0111, "step": 55650 }, { "grad_norm": 0.40440499782562256, "learning_rate": 4.478618416263933e-05, "loss": 0.0083, "step": 55660 }, { "grad_norm": 0.10982658714056015, "learning_rate": 4.476973989099373e-05, "loss": 0.007, "step": 55670 }, { "grad_norm": 0.13505057990550995, "learning_rate": 4.475329619132153e-05, "loss": 0.0066, "step": 55680 }, { "grad_norm": 0.1627795696258545, "learning_rate": 4.4736853065420956e-05, "loss": 0.0065, "step": 55690 }, { "grad_norm": 0.13435187935829163, "learning_rate": 4.472041051509023e-05, "loss": 0.0076, "step": 55700 }, { "grad_norm": 0.10883857309818268, "learning_rate": 4.470396854212746e-05, "loss": 0.0069, "step": 55710 }, { "grad_norm": 0.1687563806772232, "learning_rate": 4.468752714833072e-05, "loss": 0.008, "step": 55720 }, { "grad_norm": 0.3326200842857361, "learning_rate": 4.467108633549803e-05, "loss": 0.0076, "step": 55730 }, { "grad_norm": 0.4012129604816437, "learning_rate": 4.4654646105427335e-05, "loss": 0.0106, "step": 55740 }, { "grad_norm": 0.13360853493213654, "learning_rate": 4.463820645991651e-05, "loss": 0.0082, "step": 55750 }, { "grad_norm": 0.15518051385879517, "learning_rate": 4.462176740076334e-05, "loss": 0.0101, "step": 55760 }, { "grad_norm": 0.10639043897390366, "learning_rate": 4.4605328929765624e-05, "loss": 0.0079, "step": 55770 }, { "grad_norm": 0.1946135312318802, "learning_rate": 4.458889104872102e-05, "loss": 0.012, "step": 55780 }, { "grad_norm": 0.15207263827323914, "learning_rate": 4.457245375942715e-05, "loss": 0.0093, "step": 55790 }, { "grad_norm": 1.4075922966003418, "learning_rate": 4.455601706368157e-05, "loss": 0.009, "step": 55800 }, { "grad_norm": 0.10375825315713882, "learning_rate": 4.453958096328179e-05, "loss": 0.0075, "step": 55810 }, { "grad_norm": 0.09678203612565994, "learning_rate": 4.452314546002523e-05, "loss": 0.0119, "step": 55820 }, { "grad_norm": 0.16473747789859772, "learning_rate": 4.4506710555709246e-05, "loss": 0.0086, "step": 55830 }, { "grad_norm": 0.20334690809249878, "learning_rate": 4.449027625213114e-05, "loss": 0.0136, "step": 55840 }, { "grad_norm": 0.2631480097770691, "learning_rate": 4.4473842551088135e-05, "loss": 0.0078, "step": 55850 }, { "grad_norm": 0.16774047911167145, "learning_rate": 4.445740945437738e-05, "loss": 0.0069, "step": 55860 }, { "grad_norm": 0.4085122346878052, "learning_rate": 4.444097696379602e-05, "loss": 0.0078, "step": 55870 }, { "grad_norm": 0.16787847876548767, "learning_rate": 4.442454508114106e-05, "loss": 0.0107, "step": 55880 }, { "grad_norm": 0.1472003310918808, "learning_rate": 4.4408113808209465e-05, "loss": 0.012, "step": 55890 }, { "grad_norm": 0.20450778305530548, "learning_rate": 4.439168314679813e-05, "loss": 0.0065, "step": 55900 }, { "grad_norm": 0.18691684305667877, "learning_rate": 4.43752530987039e-05, "loss": 0.0123, "step": 55910 }, { "grad_norm": 0.1694447249174118, "learning_rate": 4.4358823665723515e-05, "loss": 0.0127, "step": 55920 }, { "grad_norm": 0.20129606127738953, "learning_rate": 4.434239484965371e-05, "loss": 0.0102, "step": 55930 }, { "grad_norm": 0.21143989264965057, "learning_rate": 4.4325966652291103e-05, "loss": 0.0106, "step": 55940 }, { "grad_norm": 0.1909371316432953, "learning_rate": 4.430953907543225e-05, "loss": 0.0113, "step": 55950 }, { "grad_norm": 0.09925603121519089, "learning_rate": 4.4293112120873645e-05, "loss": 0.0081, "step": 55960 }, { "grad_norm": 0.22539372742176056, "learning_rate": 4.427668579041172e-05, "loss": 0.0111, "step": 55970 }, { "grad_norm": 0.3591972887516022, "learning_rate": 4.4260260085842816e-05, "loss": 0.0173, "step": 55980 }, { "grad_norm": 0.13143296539783478, "learning_rate": 4.4243835008963256e-05, "loss": 0.011, "step": 55990 }, { "grad_norm": 0.15353116393089294, "learning_rate": 4.422741056156925e-05, "loss": 0.0092, "step": 56000 }, { "grad_norm": 0.24498754739761353, "learning_rate": 4.421098674545693e-05, "loss": 0.01, "step": 56010 }, { "grad_norm": 0.309546560049057, "learning_rate": 4.419456356242241e-05, "loss": 0.0089, "step": 56020 }, { "grad_norm": 0.15136973559856415, "learning_rate": 4.4178141014261685e-05, "loss": 0.0091, "step": 56030 }, { "grad_norm": 0.08149728924036026, "learning_rate": 4.416171910277071e-05, "loss": 0.0096, "step": 56040 }, { "grad_norm": 0.09519147127866745, "learning_rate": 4.414529782974533e-05, "loss": 0.0081, "step": 56050 }, { "grad_norm": 0.08105017989873886, "learning_rate": 4.41288771969814e-05, "loss": 0.0061, "step": 56060 }, { "grad_norm": 0.5995236039161682, "learning_rate": 4.411245720627464e-05, "loss": 0.0077, "step": 56070 }, { "grad_norm": 0.09378941357135773, "learning_rate": 4.40960378594207e-05, "loss": 0.0084, "step": 56080 }, { "grad_norm": 0.11922793835401535, "learning_rate": 4.407961915821519e-05, "loss": 0.0077, "step": 56090 }, { "grad_norm": 0.08367888629436493, "learning_rate": 4.406320110445362e-05, "loss": 0.0096, "step": 56100 }, { "grad_norm": 0.15362995862960815, "learning_rate": 4.404678369993144e-05, "loss": 0.0112, "step": 56110 }, { "grad_norm": 0.10924874246120453, "learning_rate": 4.403036694644406e-05, "loss": 0.0099, "step": 56120 }, { "grad_norm": 0.11960777640342712, "learning_rate": 4.4013950845786764e-05, "loss": 0.0063, "step": 56130 }, { "grad_norm": 0.21750445663928986, "learning_rate": 4.399753539975482e-05, "loss": 0.0081, "step": 56140 }, { "grad_norm": 0.08122721314430237, "learning_rate": 4.398112061014337e-05, "loss": 0.0107, "step": 56150 }, { "grad_norm": 0.12432977557182312, "learning_rate": 4.396470647874753e-05, "loss": 0.0065, "step": 56160 }, { "grad_norm": 0.10687609761953354, "learning_rate": 4.394829300736229e-05, "loss": 0.0091, "step": 56170 }, { "grad_norm": 0.1837281435728073, "learning_rate": 4.393188019778265e-05, "loss": 0.0113, "step": 56180 }, { "grad_norm": 0.13109025359153748, "learning_rate": 4.391546805180347e-05, "loss": 0.0082, "step": 56190 }, { "grad_norm": 0.13690069317817688, "learning_rate": 4.389905657121955e-05, "loss": 0.0069, "step": 56200 }, { "grad_norm": 0.12502029538154602, "learning_rate": 4.3882645757825635e-05, "loss": 0.0105, "step": 56210 }, { "grad_norm": 0.3059963285923004, "learning_rate": 4.386623561341637e-05, "loss": 0.0084, "step": 56220 }, { "grad_norm": 0.09245708584785461, "learning_rate": 4.384982613978637e-05, "loss": 0.0091, "step": 56230 }, { "grad_norm": 0.13053569197654724, "learning_rate": 4.383341733873012e-05, "loss": 0.0088, "step": 56240 }, { "grad_norm": 0.8119770884513855, "learning_rate": 4.381700921204209e-05, "loss": 0.0092, "step": 56250 }, { "grad_norm": 0.1109800860285759, "learning_rate": 4.380060176151663e-05, "loss": 0.0081, "step": 56260 }, { "grad_norm": 0.19461648166179657, "learning_rate": 4.378419498894805e-05, "loss": 0.0099, "step": 56270 }, { "grad_norm": 0.3501560091972351, "learning_rate": 4.376778889613056e-05, "loss": 0.0101, "step": 56280 }, { "grad_norm": 0.14601148664951324, "learning_rate": 4.375138348485831e-05, "loss": 0.0133, "step": 56290 }, { "grad_norm": 0.22543665766716003, "learning_rate": 4.3734978756925346e-05, "loss": 0.0097, "step": 56300 }, { "grad_norm": 0.1093212217092514, "learning_rate": 4.3718574714125706e-05, "loss": 0.0073, "step": 56310 }, { "grad_norm": 0.10678073018789291, "learning_rate": 4.370217135825329e-05, "loss": 0.0109, "step": 56320 }, { "grad_norm": 0.17211630940437317, "learning_rate": 4.368576869110194e-05, "loss": 0.0071, "step": 56330 }, { "grad_norm": 0.13112498819828033, "learning_rate": 4.366936671446544e-05, "loss": 0.0082, "step": 56340 }, { "grad_norm": 0.13296973705291748, "learning_rate": 4.3652965430137474e-05, "loss": 0.0098, "step": 56350 }, { "grad_norm": 0.14802797138690948, "learning_rate": 4.3636564839911646e-05, "loss": 0.0092, "step": 56360 }, { "grad_norm": 0.13695889711380005, "learning_rate": 4.3620164945581545e-05, "loss": 0.0088, "step": 56370 }, { "grad_norm": 0.14536458253860474, "learning_rate": 4.360376574894061e-05, "loss": 0.0067, "step": 56380 }, { "grad_norm": 0.24227651953697205, "learning_rate": 4.358736725178224e-05, "loss": 0.0098, "step": 56390 }, { "grad_norm": 0.14734648168087006, "learning_rate": 4.357096945589974e-05, "loss": 0.01, "step": 56400 }, { "grad_norm": 0.1324617862701416, "learning_rate": 4.355457236308636e-05, "loss": 0.0058, "step": 56410 }, { "grad_norm": 0.1288238763809204, "learning_rate": 4.353817597513526e-05, "loss": 0.0056, "step": 56420 }, { "grad_norm": 0.10489191114902496, "learning_rate": 4.352178029383948e-05, "loss": 0.0075, "step": 56430 }, { "grad_norm": 0.16921091079711914, "learning_rate": 4.3505385320992105e-05, "loss": 0.0085, "step": 56440 }, { "grad_norm": 0.17024162411689758, "learning_rate": 4.348899105838602e-05, "loss": 0.0075, "step": 56450 }, { "grad_norm": 0.12448390573263168, "learning_rate": 4.3472597507814087e-05, "loss": 0.0066, "step": 56460 }, { "grad_norm": 0.11580658704042435, "learning_rate": 4.3456204671069066e-05, "loss": 0.0088, "step": 56470 }, { "grad_norm": 0.13394588232040405, "learning_rate": 4.343981254994367e-05, "loss": 0.0078, "step": 56480 }, { "grad_norm": 0.09383895993232727, "learning_rate": 4.34234211462305e-05, "loss": 0.0089, "step": 56490 }, { "grad_norm": 0.2554352581501007, "learning_rate": 4.340703046172213e-05, "loss": 0.0082, "step": 56500 }, { "grad_norm": 0.14612944424152374, "learning_rate": 4.339064049821097e-05, "loss": 0.0082, "step": 56510 }, { "grad_norm": 0.4032727777957916, "learning_rate": 4.3374251257489446e-05, "loss": 0.0111, "step": 56520 }, { "grad_norm": 0.16887664794921875, "learning_rate": 4.335786274134984e-05, "loss": 0.0081, "step": 56530 }, { "grad_norm": 0.2181091457605362, "learning_rate": 4.334147495158438e-05, "loss": 0.0091, "step": 56540 }, { "grad_norm": 0.09377556294202805, "learning_rate": 4.33250878899852e-05, "loss": 0.0069, "step": 56550 }, { "grad_norm": 0.08814484626054764, "learning_rate": 4.330870155834437e-05, "loss": 0.0079, "step": 56560 }, { "grad_norm": 0.16272152960300446, "learning_rate": 4.3292315958453884e-05, "loss": 0.0086, "step": 56570 }, { "grad_norm": 0.0996633917093277, "learning_rate": 4.3275931092105646e-05, "loss": 0.0049, "step": 56580 }, { "grad_norm": 0.1984478235244751, "learning_rate": 4.325954696109147e-05, "loss": 0.007, "step": 56590 }, { "grad_norm": 0.12177035212516785, "learning_rate": 4.32431635672031e-05, "loss": 0.0098, "step": 56600 }, { "grad_norm": 0.15084096789360046, "learning_rate": 4.3226780912232214e-05, "loss": 0.0081, "step": 56610 }, { "grad_norm": 0.12570804357528687, "learning_rate": 4.321039899797036e-05, "loss": 0.0071, "step": 56620 }, { "grad_norm": 0.19780486822128296, "learning_rate": 4.319401782620908e-05, "loss": 0.01, "step": 56630 }, { "grad_norm": 0.1266993284225464, "learning_rate": 4.317763739873978e-05, "loss": 0.0065, "step": 56640 }, { "grad_norm": 0.3654629588127136, "learning_rate": 4.31612577173538e-05, "loss": 0.0071, "step": 56650 }, { "grad_norm": 0.18371815979480743, "learning_rate": 4.314487878384239e-05, "loss": 0.0091, "step": 56660 }, { "grad_norm": 0.2520676553249359, "learning_rate": 4.3128500599996736e-05, "loss": 0.0081, "step": 56670 }, { "grad_norm": 0.14902162551879883, "learning_rate": 4.31121231676079e-05, "loss": 0.0069, "step": 56680 }, { "grad_norm": 0.2530784606933594, "learning_rate": 4.309574648846694e-05, "loss": 0.009, "step": 56690 }, { "grad_norm": 0.19203433394432068, "learning_rate": 4.3079370564364755e-05, "loss": 0.008, "step": 56700 }, { "grad_norm": 0.12298604846000671, "learning_rate": 4.306299539709221e-05, "loss": 0.0088, "step": 56710 }, { "grad_norm": 0.1640535145998001, "learning_rate": 4.3046620988440046e-05, "loss": 0.0093, "step": 56720 }, { "grad_norm": 0.15716615319252014, "learning_rate": 4.303024734019897e-05, "loss": 0.0068, "step": 56730 }, { "grad_norm": 0.10513154417276382, "learning_rate": 4.3013874454159565e-05, "loss": 0.007, "step": 56740 }, { "grad_norm": 1.3631432056427002, "learning_rate": 4.299750233211233e-05, "loss": 0.0094, "step": 56750 }, { "grad_norm": 0.14341987669467926, "learning_rate": 4.2981130975847715e-05, "loss": 0.0076, "step": 56760 }, { "grad_norm": 0.21853867173194885, "learning_rate": 4.2964760387156076e-05, "loss": 0.0123, "step": 56770 }, { "grad_norm": 0.17629195749759674, "learning_rate": 4.294839056782767e-05, "loss": 0.0082, "step": 56780 }, { "grad_norm": 0.14471496641635895, "learning_rate": 4.293202151965266e-05, "loss": 0.008, "step": 56790 }, { "grad_norm": 0.14702418446540833, "learning_rate": 4.291565324442115e-05, "loss": 0.0085, "step": 56800 }, { "grad_norm": 0.16848579049110413, "learning_rate": 4.2899285743923135e-05, "loss": 0.0075, "step": 56810 }, { "grad_norm": 0.116803377866745, "learning_rate": 4.288291901994857e-05, "loss": 0.0111, "step": 56820 }, { "grad_norm": 0.14411136507987976, "learning_rate": 4.286655307428729e-05, "loss": 0.0089, "step": 56830 }, { "grad_norm": 0.2244618535041809, "learning_rate": 4.285018790872903e-05, "loss": 0.008, "step": 56840 }, { "grad_norm": 0.14027740061283112, "learning_rate": 4.283382352506347e-05, "loss": 0.0056, "step": 56850 }, { "grad_norm": 0.11642058938741684, "learning_rate": 4.2817459925080205e-05, "loss": 0.0078, "step": 56860 }, { "grad_norm": 0.10345840454101562, "learning_rate": 4.28010971105687e-05, "loss": 0.0077, "step": 56870 }, { "grad_norm": 0.14303511381149292, "learning_rate": 4.278473508331841e-05, "loss": 0.0073, "step": 56880 }, { "grad_norm": 0.12785401940345764, "learning_rate": 4.276837384511864e-05, "loss": 0.0078, "step": 56890 }, { "grad_norm": 0.10313466936349869, "learning_rate": 4.275201339775864e-05, "loss": 0.0131, "step": 56900 }, { "grad_norm": 0.15214285254478455, "learning_rate": 4.273565374302756e-05, "loss": 0.0072, "step": 56910 }, { "grad_norm": 0.1313152313232422, "learning_rate": 4.2719294882714454e-05, "loss": 0.0047, "step": 56920 }, { "grad_norm": 0.08695296943187714, "learning_rate": 4.270293681860831e-05, "loss": 0.0094, "step": 56930 }, { "grad_norm": 0.14658764004707336, "learning_rate": 4.268657955249802e-05, "loss": 0.008, "step": 56940 }, { "grad_norm": 0.14576098322868347, "learning_rate": 4.267022308617241e-05, "loss": 0.0083, "step": 56950 }, { "grad_norm": 0.128310427069664, "learning_rate": 4.265386742142018e-05, "loss": 0.0083, "step": 56960 }, { "grad_norm": 0.21275538206100464, "learning_rate": 4.263751256002996e-05, "loss": 0.0072, "step": 56970 }, { "grad_norm": 0.17105799913406372, "learning_rate": 4.26211585037903e-05, "loss": 0.0073, "step": 56980 }, { "grad_norm": 0.15840326249599457, "learning_rate": 4.2604805254489656e-05, "loss": 0.0096, "step": 56990 }, { "grad_norm": 0.1486247181892395, "learning_rate": 4.258845281391638e-05, "loss": 0.0073, "step": 57000 }, { "grad_norm": 0.1466895490884781, "learning_rate": 4.257210118385876e-05, "loss": 0.0068, "step": 57010 }, { "grad_norm": 0.11399376392364502, "learning_rate": 4.2555750366104994e-05, "loss": 0.0067, "step": 57020 }, { "grad_norm": 0.11313561350107193, "learning_rate": 4.2539400362443184e-05, "loss": 0.0082, "step": 57030 }, { "grad_norm": 0.12969498336315155, "learning_rate": 4.252305117466133e-05, "loss": 0.0067, "step": 57040 }, { "grad_norm": 0.1140051782131195, "learning_rate": 4.2506702804547356e-05, "loss": 0.0072, "step": 57050 }, { "grad_norm": 0.1046958714723587, "learning_rate": 4.249035525388907e-05, "loss": 0.0077, "step": 57060 }, { "grad_norm": 0.12210997194051743, "learning_rate": 4.247400852447427e-05, "loss": 0.0058, "step": 57070 }, { "grad_norm": 0.11167602986097336, "learning_rate": 4.245766261809059e-05, "loss": 0.0073, "step": 57080 }, { "grad_norm": 0.11451682448387146, "learning_rate": 4.244131753652558e-05, "loss": 0.0067, "step": 57090 }, { "grad_norm": 0.10310234874486923, "learning_rate": 4.242497328156672e-05, "loss": 0.0047, "step": 57100 }, { "grad_norm": 0.09933371841907501, "learning_rate": 4.240862985500139e-05, "loss": 0.0084, "step": 57110 }, { "grad_norm": 0.11573042720556259, "learning_rate": 4.2392287258616894e-05, "loss": 0.0071, "step": 57120 }, { "grad_norm": 0.14270669221878052, "learning_rate": 4.2375945494200396e-05, "loss": 0.0086, "step": 57130 }, { "grad_norm": 0.27209144830703735, "learning_rate": 4.2359604563539054e-05, "loss": 0.0068, "step": 57140 }, { "grad_norm": 0.23231711983680725, "learning_rate": 4.2343264468419885e-05, "loss": 0.0086, "step": 57150 }, { "grad_norm": 0.23429395258426666, "learning_rate": 4.232692521062979e-05, "loss": 0.0095, "step": 57160 }, { "grad_norm": 0.16969451308250427, "learning_rate": 4.231058679195562e-05, "loss": 0.0057, "step": 57170 }, { "grad_norm": 0.2752041518688202, "learning_rate": 4.229424921418411e-05, "loss": 0.0078, "step": 57180 }, { "grad_norm": 0.12716369330883026, "learning_rate": 4.2277912479101916e-05, "loss": 0.0081, "step": 57190 }, { "grad_norm": 0.17329995334148407, "learning_rate": 4.226157658849562e-05, "loss": 0.0137, "step": 57200 }, { "grad_norm": 0.8078318238258362, "learning_rate": 4.2245241544151664e-05, "loss": 0.0076, "step": 57210 }, { "grad_norm": 0.08605687320232391, "learning_rate": 4.2228907347856425e-05, "loss": 0.0069, "step": 57220 }, { "grad_norm": 0.19316242635250092, "learning_rate": 4.2212574001396214e-05, "loss": 0.008, "step": 57230 }, { "grad_norm": 0.1649254709482193, "learning_rate": 4.21962415065572e-05, "loss": 0.0084, "step": 57240 }, { "grad_norm": 0.13491179049015045, "learning_rate": 4.217990986512548e-05, "loss": 0.0127, "step": 57250 }, { "grad_norm": 0.11973950266838074, "learning_rate": 4.216357907888704e-05, "loss": 0.0068, "step": 57260 }, { "grad_norm": 0.140670508146286, "learning_rate": 4.2147249149627824e-05, "loss": 0.0069, "step": 57270 }, { "grad_norm": 0.16179601848125458, "learning_rate": 4.213092007913364e-05, "loss": 0.008, "step": 57280 }, { "grad_norm": 0.09794832766056061, "learning_rate": 4.2114591869190205e-05, "loss": 0.0071, "step": 57290 }, { "grad_norm": 1.1047126054763794, "learning_rate": 4.209826452158315e-05, "loss": 0.0056, "step": 57300 }, { "grad_norm": 0.0986659824848175, "learning_rate": 4.2081938038098e-05, "loss": 0.0085, "step": 57310 }, { "grad_norm": 0.21679265797138214, "learning_rate": 4.2065612420520185e-05, "loss": 0.0067, "step": 57320 }, { "grad_norm": 0.11107142269611359, "learning_rate": 4.204928767063509e-05, "loss": 0.0099, "step": 57330 }, { "grad_norm": 0.18181708455085754, "learning_rate": 4.2032963790227944e-05, "loss": 0.0092, "step": 57340 }, { "grad_norm": 0.13281689584255219, "learning_rate": 4.2016640781083887e-05, "loss": 0.0088, "step": 57350 }, { "grad_norm": 0.11259633302688599, "learning_rate": 4.2000318644988004e-05, "loss": 0.0067, "step": 57360 }, { "grad_norm": 0.14761847257614136, "learning_rate": 4.198399738372524e-05, "loss": 0.0084, "step": 57370 }, { "grad_norm": 0.22543805837631226, "learning_rate": 4.196767699908045e-05, "loss": 0.007, "step": 57380 }, { "grad_norm": 0.12465398758649826, "learning_rate": 4.195135749283845e-05, "loss": 0.0077, "step": 57390 }, { "grad_norm": 0.16919326782226562, "learning_rate": 4.1935038866783894e-05, "loss": 0.007, "step": 57400 }, { "grad_norm": 0.15420807898044586, "learning_rate": 4.1918721122701355e-05, "loss": 0.0076, "step": 57410 }, { "grad_norm": 0.13293106853961945, "learning_rate": 4.190240426237533e-05, "loss": 0.0113, "step": 57420 }, { "grad_norm": 0.11707668751478195, "learning_rate": 4.188608828759019e-05, "loss": 0.0087, "step": 57430 }, { "grad_norm": 0.1389446258544922, "learning_rate": 4.1869773200130246e-05, "loss": 0.0087, "step": 57440 }, { "grad_norm": 0.12871414422988892, "learning_rate": 4.185345900177967e-05, "loss": 0.0108, "step": 57450 }, { "grad_norm": 0.11352963000535965, "learning_rate": 4.183714569432258e-05, "loss": 0.0066, "step": 57460 }, { "grad_norm": 0.1510687619447708, "learning_rate": 4.182083327954295e-05, "loss": 0.0093, "step": 57470 }, { "grad_norm": 0.15336136519908905, "learning_rate": 4.1804521759224714e-05, "loss": 0.0072, "step": 57480 }, { "grad_norm": 0.19930128753185272, "learning_rate": 4.178821113515165e-05, "loss": 0.0081, "step": 57490 }, { "grad_norm": 0.15893840789794922, "learning_rate": 4.1771901409107476e-05, "loss": 0.0123, "step": 57500 }, { "grad_norm": 0.12853410840034485, "learning_rate": 4.175559258287577e-05, "loss": 0.0115, "step": 57510 }, { "grad_norm": 0.8798114657402039, "learning_rate": 4.173928465824008e-05, "loss": 0.0101, "step": 57520 }, { "grad_norm": 0.1357572376728058, "learning_rate": 4.17229776369838e-05, "loss": 0.0082, "step": 57530 }, { "grad_norm": 0.20101375877857208, "learning_rate": 4.170667152089025e-05, "loss": 0.0085, "step": 57540 }, { "grad_norm": 0.22616247832775116, "learning_rate": 4.1690366311742626e-05, "loss": 0.0118, "step": 57550 }, { "grad_norm": 0.15517976880073547, "learning_rate": 4.1674062011324055e-05, "loss": 0.0083, "step": 57560 }, { "grad_norm": 0.1824551671743393, "learning_rate": 4.165775862141752e-05, "loss": 0.0126, "step": 57570 }, { "grad_norm": 0.13701413571834564, "learning_rate": 4.164145614380598e-05, "loss": 0.0078, "step": 57580 }, { "grad_norm": 0.13642530143260956, "learning_rate": 4.162515458027223e-05, "loss": 0.0081, "step": 57590 }, { "grad_norm": 0.15846586227416992, "learning_rate": 4.160885393259899e-05, "loss": 0.0096, "step": 57600 }, { "grad_norm": 0.1737145036458969, "learning_rate": 4.159255420256886e-05, "loss": 0.0096, "step": 57610 }, { "grad_norm": 0.12309322506189346, "learning_rate": 4.157625539196436e-05, "loss": 0.0083, "step": 57620 }, { "grad_norm": 0.1604074090719223, "learning_rate": 4.155995750256791e-05, "loss": 0.0105, "step": 57630 }, { "grad_norm": 0.11664202064275742, "learning_rate": 4.15436605361618e-05, "loss": 0.0078, "step": 57640 }, { "grad_norm": 0.4649597108364105, "learning_rate": 4.152736449452827e-05, "loss": 0.0106, "step": 57650 }, { "grad_norm": 0.12477774173021317, "learning_rate": 4.151106937944943e-05, "loss": 0.0063, "step": 57660 }, { "grad_norm": 0.17524532973766327, "learning_rate": 4.149477519270727e-05, "loss": 0.0094, "step": 57670 }, { "grad_norm": 0.1472548246383667, "learning_rate": 4.147848193608371e-05, "loss": 0.0081, "step": 57680 }, { "grad_norm": 0.08445919305086136, "learning_rate": 4.146218961136056e-05, "loss": 0.0075, "step": 57690 }, { "grad_norm": 0.2399773895740509, "learning_rate": 4.14458982203195e-05, "loss": 0.0091, "step": 57700 }, { "grad_norm": 0.12264037877321243, "learning_rate": 4.1429607764742165e-05, "loss": 0.0089, "step": 57710 }, { "grad_norm": 0.657680869102478, "learning_rate": 4.141331824641003e-05, "loss": 0.0092, "step": 57720 }, { "grad_norm": 0.09306024760007858, "learning_rate": 4.139702966710452e-05, "loss": 0.0086, "step": 57730 }, { "grad_norm": 0.13913172483444214, "learning_rate": 4.13807420286069e-05, "loss": 0.0079, "step": 57740 }, { "grad_norm": 0.5907408595085144, "learning_rate": 4.136445533269837e-05, "loss": 0.0138, "step": 57750 }, { "grad_norm": 0.1779555231332779, "learning_rate": 4.134816958116003e-05, "loss": 0.008, "step": 57760 }, { "grad_norm": 0.13021770119667053, "learning_rate": 4.133188477577283e-05, "loss": 0.0063, "step": 57770 }, { "grad_norm": 0.13915465772151947, "learning_rate": 4.131560091831771e-05, "loss": 0.0098, "step": 57780 }, { "grad_norm": 0.15427593886852264, "learning_rate": 4.1299318010575405e-05, "loss": 0.0064, "step": 57790 }, { "grad_norm": 0.1206718236207962, "learning_rate": 4.128303605432661e-05, "loss": 0.0083, "step": 57800 }, { "grad_norm": 0.13017597794532776, "learning_rate": 4.126675505135188e-05, "loss": 0.0083, "step": 57810 }, { "grad_norm": 0.13671989738941193, "learning_rate": 4.125047500343169e-05, "loss": 0.0094, "step": 57820 }, { "grad_norm": 0.3449282646179199, "learning_rate": 4.1234195912346376e-05, "loss": 0.0085, "step": 57830 }, { "grad_norm": 0.13767372071743011, "learning_rate": 4.1217917779876235e-05, "loss": 0.007, "step": 57840 }, { "grad_norm": 0.6054291725158691, "learning_rate": 4.120164060780141e-05, "loss": 0.0091, "step": 57850 }, { "grad_norm": 0.1519489288330078, "learning_rate": 4.1185364397901935e-05, "loss": 0.0105, "step": 57860 }, { "grad_norm": 0.3419247567653656, "learning_rate": 4.116908915195776e-05, "loss": 0.0081, "step": 57870 }, { "grad_norm": 0.24631336331367493, "learning_rate": 4.115281487174872e-05, "loss": 0.007, "step": 57880 }, { "grad_norm": 0.1255791336297989, "learning_rate": 4.113654155905452e-05, "loss": 0.0079, "step": 57890 }, { "grad_norm": 0.663740873336792, "learning_rate": 4.112026921565483e-05, "loss": 0.0127, "step": 57900 }, { "grad_norm": 0.23793165385723114, "learning_rate": 4.1103997843329147e-05, "loss": 0.0096, "step": 57910 }, { "grad_norm": 0.2069811224937439, "learning_rate": 4.108772744385689e-05, "loss": 0.0097, "step": 57920 }, { "grad_norm": 0.15189246833324432, "learning_rate": 4.1071458019017346e-05, "loss": 0.0072, "step": 57930 }, { "grad_norm": 0.13666124641895294, "learning_rate": 4.1055189570589734e-05, "loss": 0.007, "step": 57940 }, { "grad_norm": 0.3025372326374054, "learning_rate": 4.103892210035315e-05, "loss": 0.0088, "step": 57950 }, { "grad_norm": 0.0901283472776413, "learning_rate": 4.102265561008656e-05, "loss": 0.0068, "step": 57960 }, { "grad_norm": 0.12143513560295105, "learning_rate": 4.100639010156885e-05, "loss": 0.0096, "step": 57970 }, { "grad_norm": 0.13114868104457855, "learning_rate": 4.099012557657882e-05, "loss": 0.008, "step": 57980 }, { "grad_norm": 0.18415169417858124, "learning_rate": 4.09738620368951e-05, "loss": 0.0074, "step": 57990 }, { "grad_norm": 0.10994161665439606, "learning_rate": 4.095759948429626e-05, "loss": 0.0066, "step": 58000 }, { "grad_norm": 0.17779192328453064, "learning_rate": 4.094133792056075e-05, "loss": 0.0081, "step": 58010 }, { "grad_norm": 0.15601876378059387, "learning_rate": 4.092507734746688e-05, "loss": 0.0072, "step": 58020 }, { "grad_norm": 0.14950722455978394, "learning_rate": 4.090881776679293e-05, "loss": 0.0067, "step": 58030 }, { "grad_norm": 0.1067788153886795, "learning_rate": 4.0892559180316994e-05, "loss": 0.0065, "step": 58040 }, { "grad_norm": 0.7032128572463989, "learning_rate": 4.08763015898171e-05, "loss": 0.0077, "step": 58050 }, { "grad_norm": 0.1315583884716034, "learning_rate": 4.0860044997071145e-05, "loss": 0.0068, "step": 58060 }, { "grad_norm": 0.11562436819076538, "learning_rate": 4.084378940385693e-05, "loss": 0.0065, "step": 58070 }, { "grad_norm": 0.12926135957241058, "learning_rate": 4.082753481195211e-05, "loss": 0.0077, "step": 58080 }, { "grad_norm": 0.08109541982412338, "learning_rate": 4.0811281223134324e-05, "loss": 0.0067, "step": 58090 }, { "grad_norm": 0.14921891689300537, "learning_rate": 4.0795028639181e-05, "loss": 0.0093, "step": 58100 }, { "grad_norm": 0.3243347406387329, "learning_rate": 4.077877706186951e-05, "loss": 0.0096, "step": 58110 }, { "grad_norm": 0.18112090229988098, "learning_rate": 4.076252649297709e-05, "loss": 0.0105, "step": 58120 }, { "grad_norm": 1.036967396736145, "learning_rate": 4.07462769342809e-05, "loss": 0.0095, "step": 58130 }, { "grad_norm": 0.13443057239055634, "learning_rate": 4.073002838755793e-05, "loss": 0.0087, "step": 58140 }, { "grad_norm": 0.10534494370222092, "learning_rate": 4.0713780854585115e-05, "loss": 0.0096, "step": 58150 }, { "grad_norm": 0.15453626215457916, "learning_rate": 4.069753433713927e-05, "loss": 0.011, "step": 58160 }, { "grad_norm": 0.2841566801071167, "learning_rate": 4.068128883699709e-05, "loss": 0.0078, "step": 58170 }, { "grad_norm": 0.2269725650548935, "learning_rate": 4.0665044355935155e-05, "loss": 0.0079, "step": 58180 }, { "grad_norm": 0.1696477234363556, "learning_rate": 4.064880089572993e-05, "loss": 0.0079, "step": 58190 }, { "grad_norm": 0.17399781942367554, "learning_rate": 4.063255845815779e-05, "loss": 0.0085, "step": 58200 }, { "grad_norm": 0.19498972594738007, "learning_rate": 4.0616317044994945e-05, "loss": 0.0063, "step": 58210 }, { "grad_norm": 0.13893739879131317, "learning_rate": 4.0600076658017585e-05, "loss": 0.0096, "step": 58220 }, { "grad_norm": 0.12043306231498718, "learning_rate": 4.0583837299001706e-05, "loss": 0.0084, "step": 58230 }, { "grad_norm": 0.17156566679477692, "learning_rate": 4.056759896972322e-05, "loss": 0.0077, "step": 58240 }, { "grad_norm": 0.1769760251045227, "learning_rate": 4.055136167195793e-05, "loss": 0.0098, "step": 58250 }, { "grad_norm": 0.11039112508296967, "learning_rate": 4.053512540748152e-05, "loss": 0.0082, "step": 58260 }, { "grad_norm": 0.10409170389175415, "learning_rate": 4.051889017806955e-05, "loss": 0.0074, "step": 58270 }, { "grad_norm": 0.3711540400981903, "learning_rate": 4.0502655985497516e-05, "loss": 0.0088, "step": 58280 }, { "grad_norm": 0.17517820000648499, "learning_rate": 4.048642283154074e-05, "loss": 0.01, "step": 58290 }, { "grad_norm": 0.2560346722602844, "learning_rate": 4.0470190717974464e-05, "loss": 0.0084, "step": 58300 }, { "grad_norm": 0.13574840128421783, "learning_rate": 4.045395964657381e-05, "loss": 0.0087, "step": 58310 }, { "grad_norm": 0.2083165943622589, "learning_rate": 4.043772961911376e-05, "loss": 0.0087, "step": 58320 }, { "grad_norm": 0.1940324604511261, "learning_rate": 4.042150063736924e-05, "loss": 0.007, "step": 58330 }, { "grad_norm": 0.0855185016989708, "learning_rate": 4.0405272703114976e-05, "loss": 0.0082, "step": 58340 }, { "grad_norm": 0.15626858174800873, "learning_rate": 4.038904581812569e-05, "loss": 0.0079, "step": 58350 }, { "grad_norm": 0.40605539083480835, "learning_rate": 4.03728199841759e-05, "loss": 0.0075, "step": 58360 }, { "grad_norm": 0.2451140135526657, "learning_rate": 4.0356595203040044e-05, "loss": 0.0083, "step": 58370 }, { "grad_norm": 0.12073425948619843, "learning_rate": 4.0340371476492426e-05, "loss": 0.0078, "step": 58380 }, { "grad_norm": 0.15974508225917816, "learning_rate": 4.032414880630727e-05, "loss": 0.0087, "step": 58390 }, { "grad_norm": 0.3680281937122345, "learning_rate": 4.0307927194258646e-05, "loss": 0.0076, "step": 58400 }, { "grad_norm": 0.09506340324878693, "learning_rate": 4.029170664212054e-05, "loss": 0.0109, "step": 58410 }, { "grad_norm": 0.11142800748348236, "learning_rate": 4.0275487151666794e-05, "loss": 0.0095, "step": 58420 }, { "grad_norm": 0.13063420355319977, "learning_rate": 4.025926872467116e-05, "loss": 0.0112, "step": 58430 }, { "grad_norm": 0.11587173491716385, "learning_rate": 4.024305136290725e-05, "loss": 0.0068, "step": 58440 }, { "grad_norm": 0.26262181997299194, "learning_rate": 4.022683506814858e-05, "loss": 0.0081, "step": 58450 }, { "grad_norm": 0.12907743453979492, "learning_rate": 4.021061984216853e-05, "loss": 0.0091, "step": 58460 }, { "grad_norm": 0.1799430549144745, "learning_rate": 4.019440568674035e-05, "loss": 0.0068, "step": 58470 }, { "grad_norm": 0.18542031943798065, "learning_rate": 4.017819260363724e-05, "loss": 0.0118, "step": 58480 }, { "grad_norm": 0.15074469149112701, "learning_rate": 4.016198059463223e-05, "loss": 0.0094, "step": 58490 }, { "grad_norm": 0.12200883775949478, "learning_rate": 4.0145769661498233e-05, "loss": 0.0075, "step": 58500 }, { "grad_norm": 0.11549049615859985, "learning_rate": 4.012955980600804e-05, "loss": 0.01, "step": 58510 }, { "grad_norm": 0.2284967452287674, "learning_rate": 4.0113351029934355e-05, "loss": 0.0084, "step": 58520 }, { "grad_norm": 0.5146021842956543, "learning_rate": 4.0097143335049714e-05, "loss": 0.0084, "step": 58530 }, { "grad_norm": 0.14211811125278473, "learning_rate": 4.0080936723126606e-05, "loss": 0.009, "step": 58540 }, { "grad_norm": 0.2898949682712555, "learning_rate": 4.006473119593735e-05, "loss": 0.0113, "step": 58550 }, { "grad_norm": 0.17122243344783783, "learning_rate": 4.004852675525415e-05, "loss": 0.0092, "step": 58560 }, { "grad_norm": 0.13935810327529907, "learning_rate": 4.00323234028491e-05, "loss": 0.0069, "step": 58570 }, { "grad_norm": 0.165359228849411, "learning_rate": 4.0016121140494186e-05, "loss": 0.0078, "step": 58580 }, { "grad_norm": 0.11959883570671082, "learning_rate": 3.999991996996122e-05, "loss": 0.0109, "step": 58590 }, { "grad_norm": 0.19419501721858978, "learning_rate": 3.9983719893022e-05, "loss": 0.008, "step": 58600 }, { "grad_norm": 0.225179985165596, "learning_rate": 3.996752091144812e-05, "loss": 0.0076, "step": 58610 }, { "grad_norm": 0.1695568710565567, "learning_rate": 3.9951323027011055e-05, "loss": 0.0072, "step": 58620 }, { "grad_norm": 1.0335062742233276, "learning_rate": 3.993512624148219e-05, "loss": 0.0102, "step": 58630 }, { "grad_norm": 0.13101747632026672, "learning_rate": 3.99189305566328e-05, "loss": 0.0063, "step": 58640 }, { "grad_norm": 0.09392668306827545, "learning_rate": 3.990273597423401e-05, "loss": 0.0095, "step": 58650 }, { "grad_norm": 0.07227963954210281, "learning_rate": 3.988654249605681e-05, "loss": 0.007, "step": 58660 }, { "grad_norm": 0.10919888317584991, "learning_rate": 3.987035012387213e-05, "loss": 0.0066, "step": 58670 }, { "grad_norm": 0.1517353355884552, "learning_rate": 3.9854158859450735e-05, "loss": 0.008, "step": 58680 }, { "grad_norm": 0.18754442036151886, "learning_rate": 3.983796870456328e-05, "loss": 0.0088, "step": 58690 }, { "grad_norm": 0.12072113156318665, "learning_rate": 3.982177966098028e-05, "loss": 0.0066, "step": 58700 }, { "grad_norm": 0.2155497968196869, "learning_rate": 3.980559173047216e-05, "loss": 0.0094, "step": 58710 }, { "grad_norm": 0.1567120999097824, "learning_rate": 3.9789404914809176e-05, "loss": 0.0129, "step": 58720 }, { "grad_norm": 0.19593091309070587, "learning_rate": 3.9773219215761544e-05, "loss": 0.009, "step": 58730 }, { "grad_norm": 0.17537431418895721, "learning_rate": 3.975703463509928e-05, "loss": 0.0091, "step": 58740 }, { "grad_norm": 0.10440371930599213, "learning_rate": 3.974085117459231e-05, "loss": 0.0093, "step": 58750 }, { "grad_norm": 0.1779242604970932, "learning_rate": 3.972466883601043e-05, "loss": 0.0052, "step": 58760 }, { "grad_norm": 3.6941897869110107, "learning_rate": 3.9708487621123316e-05, "loss": 0.0087, "step": 58770 }, { "grad_norm": 0.17620837688446045, "learning_rate": 3.969230753170051e-05, "loss": 0.0088, "step": 58780 }, { "grad_norm": 0.11530786752700806, "learning_rate": 3.967612856951146e-05, "loss": 0.0076, "step": 58790 }, { "grad_norm": 0.15974636375904083, "learning_rate": 3.9659950736325476e-05, "loss": 0.0083, "step": 58800 }, { "grad_norm": 0.13019487261772156, "learning_rate": 3.964377403391174e-05, "loss": 0.0098, "step": 58810 }, { "grad_norm": 0.13627082109451294, "learning_rate": 3.9627598464039303e-05, "loss": 0.0091, "step": 58820 }, { "grad_norm": 0.10101525485515594, "learning_rate": 3.9611424028477096e-05, "loss": 0.0057, "step": 58830 }, { "grad_norm": 0.1380540430545807, "learning_rate": 3.959525072899394e-05, "loss": 0.0106, "step": 58840 }, { "grad_norm": 0.0993976816534996, "learning_rate": 3.9579078567358506e-05, "loss": 0.0073, "step": 58850 }, { "grad_norm": 0.30888912081718445, "learning_rate": 3.956290754533939e-05, "loss": 0.014, "step": 58860 }, { "grad_norm": 0.14182381331920624, "learning_rate": 3.954673766470501e-05, "loss": 0.0071, "step": 58870 }, { "grad_norm": 0.22366149723529816, "learning_rate": 3.9530568927223674e-05, "loss": 0.0074, "step": 58880 }, { "grad_norm": 0.1795283406972885, "learning_rate": 3.95144013346636e-05, "loss": 0.009, "step": 58890 }, { "grad_norm": 0.20565265417099, "learning_rate": 3.949823488879282e-05, "loss": 0.0104, "step": 58900 }, { "grad_norm": 0.1410669982433319, "learning_rate": 3.9482069591379276e-05, "loss": 0.0077, "step": 58910 }, { "grad_norm": 0.21557694673538208, "learning_rate": 3.9465905444190784e-05, "loss": 0.01, "step": 58920 }, { "grad_norm": 0.1039988100528717, "learning_rate": 3.944974244899505e-05, "loss": 0.0073, "step": 58930 }, { "grad_norm": 0.09771806001663208, "learning_rate": 3.9433580607559616e-05, "loss": 0.0103, "step": 58940 }, { "grad_norm": 0.16665619611740112, "learning_rate": 3.941741992165193e-05, "loss": 0.0081, "step": 58950 }, { "grad_norm": 0.11047559976577759, "learning_rate": 3.940126039303928e-05, "loss": 0.0074, "step": 58960 }, { "grad_norm": 0.20317213237285614, "learning_rate": 3.938510202348886e-05, "loss": 0.0079, "step": 58970 }, { "grad_norm": 0.2291172444820404, "learning_rate": 3.93689448147677e-05, "loss": 0.0073, "step": 58980 }, { "grad_norm": 0.11538688838481903, "learning_rate": 3.935278876864278e-05, "loss": 0.0101, "step": 58990 }, { "grad_norm": 0.11835937201976776, "learning_rate": 3.933663388688087e-05, "loss": 0.0078, "step": 59000 }, { "grad_norm": 0.18890032172203064, "learning_rate": 3.9320480171248645e-05, "loss": 0.0084, "step": 59010 }, { "grad_norm": 0.1485418975353241, "learning_rate": 3.930432762351265e-05, "loss": 0.0094, "step": 59020 }, { "grad_norm": 0.13613058626651764, "learning_rate": 3.928817624543931e-05, "loss": 0.0084, "step": 59030 }, { "grad_norm": 0.1097785010933876, "learning_rate": 3.9272026038794885e-05, "loss": 0.0102, "step": 59040 }, { "grad_norm": 0.13876953721046448, "learning_rate": 3.9255877005345586e-05, "loss": 0.0082, "step": 59050 }, { "grad_norm": 0.12809127569198608, "learning_rate": 3.923972914685743e-05, "loss": 0.0063, "step": 59060 }, { "grad_norm": 0.10024664551019669, "learning_rate": 3.922358246509631e-05, "loss": 0.0074, "step": 59070 }, { "grad_norm": 1.0633243322372437, "learning_rate": 3.920743696182801e-05, "loss": 0.0071, "step": 59080 }, { "grad_norm": 0.5724990367889404, "learning_rate": 3.9191292638818164e-05, "loss": 0.0113, "step": 59090 }, { "grad_norm": 0.12947829067707062, "learning_rate": 3.9175149497832316e-05, "loss": 0.0086, "step": 59100 }, { "grad_norm": 0.18787267804145813, "learning_rate": 3.915900754063584e-05, "loss": 0.0082, "step": 59110 }, { "grad_norm": 0.25751054286956787, "learning_rate": 3.9142866768994014e-05, "loss": 0.0088, "step": 59120 }, { "grad_norm": 0.11925101280212402, "learning_rate": 3.912672718467193e-05, "loss": 0.0064, "step": 59130 }, { "grad_norm": 0.8339115381240845, "learning_rate": 3.911058878943463e-05, "loss": 0.0064, "step": 59140 }, { "grad_norm": 0.08333232998847961, "learning_rate": 3.9094451585046974e-05, "loss": 0.0082, "step": 59150 }, { "grad_norm": 0.15778033435344696, "learning_rate": 3.90783155732737e-05, "loss": 0.0095, "step": 59160 }, { "grad_norm": 0.09205187857151031, "learning_rate": 3.9062180755879373e-05, "loss": 0.0135, "step": 59170 }, { "grad_norm": 0.11105674505233765, "learning_rate": 3.904604713462855e-05, "loss": 0.0082, "step": 59180 }, { "grad_norm": 0.15366171300411224, "learning_rate": 3.902991471128554e-05, "loss": 0.008, "step": 59190 }, { "grad_norm": 0.12105274200439453, "learning_rate": 3.901378348761456e-05, "loss": 0.0096, "step": 59200 }, { "grad_norm": 0.2480236142873764, "learning_rate": 3.899765346537971e-05, "loss": 0.0066, "step": 59210 }, { "grad_norm": 0.1246001273393631, "learning_rate": 3.898152464634493e-05, "loss": 0.0077, "step": 59220 }, { "grad_norm": 0.13694147765636444, "learning_rate": 3.896539703227402e-05, "loss": 0.0103, "step": 59230 }, { "grad_norm": 0.32576099038124084, "learning_rate": 3.894927062493073e-05, "loss": 0.0126, "step": 59240 }, { "grad_norm": 0.0923779308795929, "learning_rate": 3.893314542607859e-05, "loss": 0.0085, "step": 59250 }, { "grad_norm": 0.19004002213478088, "learning_rate": 3.891702143748103e-05, "loss": 0.0114, "step": 59260 }, { "grad_norm": 0.09314586222171783, "learning_rate": 3.890089866090133e-05, "loss": 0.0074, "step": 59270 }, { "grad_norm": 0.14539989829063416, "learning_rate": 3.888477709810267e-05, "loss": 0.0113, "step": 59280 }, { "grad_norm": 0.10264234244823456, "learning_rate": 3.886865675084806e-05, "loss": 0.0089, "step": 59290 }, { "grad_norm": 0.21644391119480133, "learning_rate": 3.885253762090043e-05, "loss": 0.0086, "step": 59300 }, { "grad_norm": 0.14975430071353912, "learning_rate": 3.8836419710022534e-05, "loss": 0.0083, "step": 59310 }, { "grad_norm": 0.14437663555145264, "learning_rate": 3.882030301997698e-05, "loss": 0.0067, "step": 59320 }, { "grad_norm": 0.1832246035337448, "learning_rate": 3.880418755252629e-05, "loss": 0.0062, "step": 59330 }, { "grad_norm": 0.1887161284685135, "learning_rate": 3.87880733094328e-05, "loss": 0.0066, "step": 59340 }, { "grad_norm": 0.1463567465543747, "learning_rate": 3.877196029245877e-05, "loss": 0.0057, "step": 59350 }, { "grad_norm": 0.1381448358297348, "learning_rate": 3.875584850336627e-05, "loss": 0.0077, "step": 59360 }, { "grad_norm": 0.1768748164176941, "learning_rate": 3.873973794391728e-05, "loss": 0.0084, "step": 59370 }, { "grad_norm": 0.0856093019247055, "learning_rate": 3.872362861587361e-05, "loss": 0.006, "step": 59380 }, { "grad_norm": 1.8255677223205566, "learning_rate": 3.870752052099698e-05, "loss": 0.0118, "step": 59390 }, { "grad_norm": 0.1447019875049591, "learning_rate": 3.8691413661048916e-05, "loss": 0.0108, "step": 59400 }, { "grad_norm": 0.29105889797210693, "learning_rate": 3.867530803779085e-05, "loss": 0.0084, "step": 59410 }, { "grad_norm": 0.12304434180259705, "learning_rate": 3.8659203652984054e-05, "loss": 0.0048, "step": 59420 }, { "grad_norm": 0.11698009818792343, "learning_rate": 3.864310050838972e-05, "loss": 0.0061, "step": 59430 }, { "grad_norm": 0.18521839380264282, "learning_rate": 3.8626998605768835e-05, "loss": 0.0063, "step": 59440 }, { "grad_norm": 0.1548963189125061, "learning_rate": 3.8610897946882294e-05, "loss": 0.0062, "step": 59450 }, { "grad_norm": 0.11839081346988678, "learning_rate": 3.859479853349082e-05, "loss": 0.006, "step": 59460 }, { "grad_norm": 0.11829276382923126, "learning_rate": 3.857870036735503e-05, "loss": 0.0065, "step": 59470 }, { "grad_norm": 0.19410961866378784, "learning_rate": 3.856260345023538e-05, "loss": 0.0077, "step": 59480 }, { "grad_norm": 0.10839658230543137, "learning_rate": 3.854650778389224e-05, "loss": 0.0052, "step": 59490 }, { "grad_norm": 0.2534045875072479, "learning_rate": 3.8530413370085786e-05, "loss": 0.0088, "step": 59500 }, { "grad_norm": 0.14399659633636475, "learning_rate": 3.851432021057608e-05, "loss": 0.007, "step": 59510 }, { "grad_norm": 0.1171506717801094, "learning_rate": 3.849822830712306e-05, "loss": 0.0058, "step": 59520 }, { "grad_norm": 0.13214372098445892, "learning_rate": 3.848213766148649e-05, "loss": 0.0077, "step": 59530 }, { "grad_norm": 0.2606731951236725, "learning_rate": 3.846604827542602e-05, "loss": 0.0077, "step": 59540 }, { "grad_norm": 0.12443788349628448, "learning_rate": 3.844996015070117e-05, "loss": 0.0086, "step": 59550 }, { "grad_norm": 0.1871722787618637, "learning_rate": 3.843387328907132e-05, "loss": 0.0088, "step": 59560 }, { "grad_norm": 0.20393267273902893, "learning_rate": 3.8417787692295696e-05, "loss": 0.0059, "step": 59570 }, { "grad_norm": 0.155174121260643, "learning_rate": 3.84017033621334e-05, "loss": 0.0064, "step": 59580 }, { "grad_norm": 0.1420014500617981, "learning_rate": 3.838562030034338e-05, "loss": 0.0083, "step": 59590 }, { "grad_norm": 0.21539299190044403, "learning_rate": 3.836953850868447e-05, "loss": 0.0084, "step": 59600 }, { "grad_norm": 0.10199601203203201, "learning_rate": 3.835345798891533e-05, "loss": 0.0114, "step": 59610 }, { "grad_norm": 0.13325625658035278, "learning_rate": 3.833737874279452e-05, "loss": 0.007, "step": 59620 }, { "grad_norm": 0.1580730527639389, "learning_rate": 3.8321300772080433e-05, "loss": 0.0077, "step": 59630 }, { "grad_norm": 0.1844644844532013, "learning_rate": 3.8305224078531345e-05, "loss": 0.0077, "step": 59640 }, { "grad_norm": 0.1173093318939209, "learning_rate": 3.828914866390537e-05, "loss": 0.0078, "step": 59650 }, { "grad_norm": 0.09478757530450821, "learning_rate": 3.827307452996048e-05, "loss": 0.0069, "step": 59660 }, { "grad_norm": 0.09613043814897537, "learning_rate": 3.825700167845454e-05, "loss": 0.0081, "step": 59670 }, { "grad_norm": 0.14967182278633118, "learning_rate": 3.8240930111145214e-05, "loss": 0.0084, "step": 59680 }, { "grad_norm": 0.17240659892559052, "learning_rate": 3.8224859829790116e-05, "loss": 0.0121, "step": 59690 }, { "grad_norm": 0.12200433015823364, "learning_rate": 3.820879083614664e-05, "loss": 0.0073, "step": 59700 }, { "grad_norm": 0.12114177644252777, "learning_rate": 3.819272313197208e-05, "loss": 0.0071, "step": 59710 }, { "grad_norm": 0.15760622918605804, "learning_rate": 3.817665671902355e-05, "loss": 0.0084, "step": 59720 }, { "grad_norm": 0.17604152858257294, "learning_rate": 3.816059159905808e-05, "loss": 0.0089, "step": 59730 }, { "grad_norm": 0.12794749438762665, "learning_rate": 3.8144527773832476e-05, "loss": 0.0083, "step": 59740 }, { "grad_norm": 0.12912485003471375, "learning_rate": 3.812846524510352e-05, "loss": 0.0094, "step": 59750 }, { "grad_norm": 0.1824774146080017, "learning_rate": 3.811240401462775e-05, "loss": 0.0092, "step": 59760 }, { "grad_norm": 0.11894414573907852, "learning_rate": 3.809634408416162e-05, "loss": 0.0083, "step": 59770 }, { "grad_norm": 0.18875785171985626, "learning_rate": 3.8080285455461394e-05, "loss": 0.0091, "step": 59780 }, { "grad_norm": 0.119344063103199, "learning_rate": 3.806422813028323e-05, "loss": 0.0073, "step": 59790 }, { "grad_norm": 0.12741363048553467, "learning_rate": 3.8048172110383114e-05, "loss": 0.0071, "step": 59800 }, { "grad_norm": 0.12864233553409576, "learning_rate": 3.8032117397516944e-05, "loss": 0.0064, "step": 59810 }, { "grad_norm": 0.14593686163425446, "learning_rate": 3.801606399344042e-05, "loss": 0.0068, "step": 59820 }, { "grad_norm": 0.2057190239429474, "learning_rate": 3.800001189990913e-05, "loss": 0.0116, "step": 59830 }, { "grad_norm": 0.16613546013832092, "learning_rate": 3.798396111867847e-05, "loss": 0.0098, "step": 59840 }, { "grad_norm": 0.18168939650058746, "learning_rate": 3.796791165150378e-05, "loss": 0.0107, "step": 59850 }, { "grad_norm": 0.2936127781867981, "learning_rate": 3.795186350014018e-05, "loss": 0.0108, "step": 59860 }, { "grad_norm": 0.14794641733169556, "learning_rate": 3.793581666634266e-05, "loss": 0.0076, "step": 59870 }, { "grad_norm": 0.12746822834014893, "learning_rate": 3.791977115186609e-05, "loss": 0.0081, "step": 59880 }, { "grad_norm": 0.13938008248806, "learning_rate": 3.79037269584652e-05, "loss": 0.008, "step": 59890 }, { "grad_norm": 0.11500798910856247, "learning_rate": 3.7887684087894545e-05, "loss": 0.0079, "step": 59900 }, { "grad_norm": 0.23481790721416473, "learning_rate": 3.787164254190854e-05, "loss": 0.0073, "step": 59910 }, { "grad_norm": 0.06753039360046387, "learning_rate": 3.7855602322261473e-05, "loss": 0.0076, "step": 59920 }, { "grad_norm": 0.1931367963552475, "learning_rate": 3.783956343070746e-05, "loss": 0.0069, "step": 59930 }, { "grad_norm": 0.25579532980918884, "learning_rate": 3.7823525869000526e-05, "loss": 0.0062, "step": 59940 }, { "grad_norm": 0.15796516835689545, "learning_rate": 3.78074896388945e-05, "loss": 0.0056, "step": 59950 }, { "grad_norm": 0.1472947895526886, "learning_rate": 3.7791454742143075e-05, "loss": 0.006, "step": 59960 }, { "grad_norm": 2.215386152267456, "learning_rate": 3.7775421180499795e-05, "loss": 0.008, "step": 59970 }, { "grad_norm": 0.1275779753923416, "learning_rate": 3.7759388955718086e-05, "loss": 0.0075, "step": 59980 }, { "grad_norm": 0.342710942029953, "learning_rate": 3.774335806955117e-05, "loss": 0.0087, "step": 59990 }, { "grad_norm": 0.12884554266929626, "learning_rate": 3.7727328523752215e-05, "loss": 0.0047, "step": 60000 } ], "logging_steps": 10, "max_steps": 100000, "num_input_tokens_seen": 0, "num_train_epochs": 9223372036854775807, "save_steps": 10000, "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": 32, "trial_name": null, "trial_params": null }