{ "best_global_step": null, "best_metric": null, "best_model_checkpoint": null, "epoch": 1.0, "eval_steps": 500, "global_step": 100000, "is_hyper_param_search": false, "is_local_process_zero": true, "is_world_process_zero": true, "log_history": [ { "grad_norm": 0.9459261894226074, "learning_rate": 1.8e-07, "loss": 1.4258, "step": 10 }, { "grad_norm": 0.8486508727073669, "learning_rate": 3.8e-07, "loss": 1.4201, "step": 20 }, { "grad_norm": 0.990315854549408, "learning_rate": 5.8e-07, "loss": 1.4073, "step": 30 }, { "grad_norm": 0.9264404773712158, "learning_rate": 7.8e-07, "loss": 1.3987, "step": 40 }, { "grad_norm": 1.626911997795105, "learning_rate": 9.8e-07, "loss": 1.3604, "step": 50 }, { "grad_norm": 2.0519943237304688, "learning_rate": 1.18e-06, "loss": 1.3397, "step": 60 }, { "grad_norm": 1.7580655813217163, "learning_rate": 1.3800000000000001e-06, "loss": 1.3016, "step": 70 }, { "grad_norm": 1.079563021659851, "learning_rate": 1.5800000000000003e-06, "loss": 1.2558, "step": 80 }, { "grad_norm": 1.4976706504821777, "learning_rate": 1.7800000000000001e-06, "loss": 1.2876, "step": 90 }, { "grad_norm": 1.1991931200027466, "learning_rate": 1.98e-06, "loss": 1.2215, "step": 100 }, { "grad_norm": 1.8186744451522827, "learning_rate": 2.1800000000000003e-06, "loss": 1.2668, "step": 110 }, { "grad_norm": 1.459107756614685, "learning_rate": 2.38e-06, "loss": 1.2356, "step": 120 }, { "grad_norm": 1.3902907371520996, "learning_rate": 2.5800000000000003e-06, "loss": 1.2602, "step": 130 }, { "grad_norm": 1.7752374410629272, "learning_rate": 2.78e-06, "loss": 1.2213, "step": 140 }, { "grad_norm": 1.332119345664978, "learning_rate": 2.9800000000000003e-06, "loss": 1.19, "step": 150 }, { "grad_norm": 1.31509530544281, "learning_rate": 3.1800000000000005e-06, "loss": 1.2049, "step": 160 }, { "grad_norm": 2.379903793334961, "learning_rate": 3.38e-06, "loss": 1.1761, "step": 170 }, { "grad_norm": 1.784712553024292, "learning_rate": 3.58e-06, "loss": 1.1634, "step": 180 }, { "grad_norm": 2.008405923843384, "learning_rate": 3.7800000000000002e-06, "loss": 1.1407, "step": 190 }, { "grad_norm": 2.7953391075134277, "learning_rate": 3.98e-06, "loss": 1.1083, "step": 200 }, { "grad_norm": 3.6884050369262695, "learning_rate": 4.18e-06, "loss": 1.079, "step": 210 }, { "grad_norm": 1.3087931871414185, "learning_rate": 4.38e-06, "loss": 1.1455, "step": 220 }, { "grad_norm": 1.4400829076766968, "learning_rate": 4.58e-06, "loss": 1.0932, "step": 230 }, { "grad_norm": 1.9079021215438843, "learning_rate": 4.780000000000001e-06, "loss": 1.089, "step": 240 }, { "grad_norm": 1.774671196937561, "learning_rate": 4.98e-06, "loss": 1.1006, "step": 250 }, { "grad_norm": 1.650389313697815, "learning_rate": 5.18e-06, "loss": 1.0865, "step": 260 }, { "grad_norm": 1.473334550857544, "learning_rate": 5.38e-06, "loss": 1.1002, "step": 270 }, { "grad_norm": 1.770915150642395, "learning_rate": 5.580000000000001e-06, "loss": 1.0768, "step": 280 }, { "grad_norm": 2.0541555881500244, "learning_rate": 5.78e-06, "loss": 1.0633, "step": 290 }, { "grad_norm": 1.5642814636230469, "learning_rate": 5.98e-06, "loss": 1.0733, "step": 300 }, { "grad_norm": 1.665672779083252, "learning_rate": 6.18e-06, "loss": 1.0978, "step": 310 }, { "grad_norm": 1.254653811454773, "learning_rate": 6.38e-06, "loss": 1.0753, "step": 320 }, { "grad_norm": 2.1704907417297363, "learning_rate": 6.58e-06, "loss": 1.0813, "step": 330 }, { "grad_norm": 1.4272077083587646, "learning_rate": 6.78e-06, "loss": 1.0699, "step": 340 }, { "grad_norm": 1.1982589960098267, "learning_rate": 6.98e-06, "loss": 1.0642, "step": 350 }, { "grad_norm": 1.7140731811523438, "learning_rate": 7.180000000000001e-06, "loss": 1.0769, "step": 360 }, { "grad_norm": 1.5882264375686646, "learning_rate": 7.3800000000000005e-06, "loss": 1.0722, "step": 370 }, { "grad_norm": 1.3987808227539062, "learning_rate": 7.580000000000001e-06, "loss": 1.0607, "step": 380 }, { "grad_norm": 0.979117751121521, "learning_rate": 7.78e-06, "loss": 1.0589, "step": 390 }, { "grad_norm": 1.3536312580108643, "learning_rate": 7.98e-06, "loss": 1.0751, "step": 400 }, { "grad_norm": 1.5076813697814941, "learning_rate": 8.18e-06, "loss": 1.0693, "step": 410 }, { "grad_norm": 1.938186526298523, "learning_rate": 8.380000000000001e-06, "loss": 1.0531, "step": 420 }, { "grad_norm": 1.7611913681030273, "learning_rate": 8.580000000000001e-06, "loss": 1.0488, "step": 430 }, { "grad_norm": 2.259263515472412, "learning_rate": 8.78e-06, "loss": 1.0534, "step": 440 }, { "grad_norm": 1.5667383670806885, "learning_rate": 8.98e-06, "loss": 0.9934, "step": 450 }, { "grad_norm": 1.945088267326355, "learning_rate": 9.180000000000002e-06, "loss": 1.0188, "step": 460 }, { "grad_norm": 2.1142642498016357, "learning_rate": 9.38e-06, "loss": 0.9901, "step": 470 }, { "grad_norm": 1.660430908203125, "learning_rate": 9.58e-06, "loss": 0.9784, "step": 480 }, { "grad_norm": 1.9705544710159302, "learning_rate": 9.78e-06, "loss": 0.9529, "step": 490 }, { "grad_norm": 2.5003745555877686, "learning_rate": 9.980000000000001e-06, "loss": 0.909, "step": 500 }, { "grad_norm": 2.479682683944702, "learning_rate": 1.018e-05, "loss": 0.8907, "step": 510 }, { "grad_norm": 3.2883267402648926, "learning_rate": 1.038e-05, "loss": 0.8653, "step": 520 }, { "grad_norm": 2.0643606185913086, "learning_rate": 1.058e-05, "loss": 0.8364, "step": 530 }, { "grad_norm": 1.9223861694335938, "learning_rate": 1.0780000000000002e-05, "loss": 0.7964, "step": 540 }, { "grad_norm": 2.528588056564331, "learning_rate": 1.098e-05, "loss": 0.8047, "step": 550 }, { "grad_norm": 2.390781879425049, "learning_rate": 1.118e-05, "loss": 0.7802, "step": 560 }, { "grad_norm": 2.3060264587402344, "learning_rate": 1.1380000000000001e-05, "loss": 0.7292, "step": 570 }, { "grad_norm": 2.488738775253296, "learning_rate": 1.1580000000000001e-05, "loss": 0.7082, "step": 580 }, { "grad_norm": 2.5440170764923096, "learning_rate": 1.178e-05, "loss": 0.7113, "step": 590 }, { "grad_norm": 2.2694947719573975, "learning_rate": 1.198e-05, "loss": 0.6693, "step": 600 }, { "grad_norm": 2.781588315963745, "learning_rate": 1.2180000000000002e-05, "loss": 0.6464, "step": 610 }, { "grad_norm": 3.5640134811401367, "learning_rate": 1.238e-05, "loss": 0.6297, "step": 620 }, { "grad_norm": 3.0702695846557617, "learning_rate": 1.258e-05, "loss": 0.6096, "step": 630 }, { "grad_norm": 3.237060070037842, "learning_rate": 1.278e-05, "loss": 0.5874, "step": 640 }, { "grad_norm": 2.9110267162323, "learning_rate": 1.2980000000000001e-05, "loss": 0.5384, "step": 650 }, { "grad_norm": 3.4519009590148926, "learning_rate": 1.3180000000000001e-05, "loss": 0.5221, "step": 660 }, { "grad_norm": 2.483438491821289, "learning_rate": 1.338e-05, "loss": 0.4612, "step": 670 }, { "grad_norm": 3.5535728931427, "learning_rate": 1.358e-05, "loss": 0.48, "step": 680 }, { "grad_norm": 2.503380298614502, "learning_rate": 1.3780000000000002e-05, "loss": 0.4616, "step": 690 }, { "grad_norm": 3.3134467601776123, "learning_rate": 1.3980000000000002e-05, "loss": 0.4722, "step": 700 }, { "grad_norm": 4.015773773193359, "learning_rate": 1.4180000000000001e-05, "loss": 0.4077, "step": 710 }, { "grad_norm": 3.504662275314331, "learning_rate": 1.4380000000000001e-05, "loss": 0.4267, "step": 720 }, { "grad_norm": 2.663461685180664, "learning_rate": 1.4580000000000003e-05, "loss": 0.3911, "step": 730 }, { "grad_norm": 3.144181489944458, "learning_rate": 1.4779999999999999e-05, "loss": 0.3425, "step": 740 }, { "grad_norm": 3.0211732387542725, "learning_rate": 1.4979999999999999e-05, "loss": 0.3676, "step": 750 }, { "grad_norm": 3.003472089767456, "learning_rate": 1.518e-05, "loss": 0.3598, "step": 760 }, { "grad_norm": 2.5838558673858643, "learning_rate": 1.538e-05, "loss": 0.3358, "step": 770 }, { "grad_norm": 2.9738385677337646, "learning_rate": 1.558e-05, "loss": 0.3537, "step": 780 }, { "grad_norm": 2.7872018814086914, "learning_rate": 1.578e-05, "loss": 0.2898, "step": 790 }, { "grad_norm": 4.099661827087402, "learning_rate": 1.598e-05, "loss": 0.2772, "step": 800 }, { "grad_norm": 3.068772315979004, "learning_rate": 1.618e-05, "loss": 0.2924, "step": 810 }, { "grad_norm": 3.160888910293579, "learning_rate": 1.6380000000000002e-05, "loss": 0.2976, "step": 820 }, { "grad_norm": 2.149221420288086, "learning_rate": 1.658e-05, "loss": 0.3043, "step": 830 }, { "grad_norm": 3.0473878383636475, "learning_rate": 1.6780000000000002e-05, "loss": 0.3007, "step": 840 }, { "grad_norm": 2.8950283527374268, "learning_rate": 1.698e-05, "loss": 0.2753, "step": 850 }, { "grad_norm": 3.2187559604644775, "learning_rate": 1.718e-05, "loss": 0.265, "step": 860 }, { "grad_norm": 2.5031700134277344, "learning_rate": 1.7380000000000003e-05, "loss": 0.2829, "step": 870 }, { "grad_norm": 2.7616262435913086, "learning_rate": 1.758e-05, "loss": 0.2576, "step": 880 }, { "grad_norm": 2.896294116973877, "learning_rate": 1.7780000000000003e-05, "loss": 0.2664, "step": 890 }, { "grad_norm": 2.0850021839141846, "learning_rate": 1.798e-05, "loss": 0.2541, "step": 900 }, { "grad_norm": 1.9314359426498413, "learning_rate": 1.818e-05, "loss": 0.2198, "step": 910 }, { "grad_norm": 1.8471407890319824, "learning_rate": 1.838e-05, "loss": 0.1929, "step": 920 }, { "grad_norm": 2.1967244148254395, "learning_rate": 1.858e-05, "loss": 0.2071, "step": 930 }, { "grad_norm": 3.4674909114837646, "learning_rate": 1.878e-05, "loss": 0.2084, "step": 940 }, { "grad_norm": 1.6144940853118896, "learning_rate": 1.898e-05, "loss": 0.1847, "step": 950 }, { "grad_norm": 2.1287527084350586, "learning_rate": 1.918e-05, "loss": 0.1972, "step": 960 }, { "grad_norm": 2.7032134532928467, "learning_rate": 1.938e-05, "loss": 0.2266, "step": 970 }, { "grad_norm": 1.987121343612671, "learning_rate": 1.9580000000000002e-05, "loss": 0.1797, "step": 980 }, { "grad_norm": 3.270242929458618, "learning_rate": 1.978e-05, "loss": 0.1701, "step": 990 }, { "grad_norm": 2.9609570503234863, "learning_rate": 1.9980000000000002e-05, "loss": 0.1834, "step": 1000 }, { "grad_norm": 1.8890717029571533, "learning_rate": 2.0180000000000003e-05, "loss": 0.1648, "step": 1010 }, { "grad_norm": 2.5171937942504883, "learning_rate": 2.038e-05, "loss": 0.1846, "step": 1020 }, { "grad_norm": 1.980997085571289, "learning_rate": 2.0580000000000003e-05, "loss": 0.1723, "step": 1030 }, { "grad_norm": 2.443849563598633, "learning_rate": 2.078e-05, "loss": 0.1661, "step": 1040 }, { "grad_norm": 2.3949363231658936, "learning_rate": 2.098e-05, "loss": 0.1821, "step": 1050 }, { "grad_norm": 2.3386902809143066, "learning_rate": 2.118e-05, "loss": 0.2041, "step": 1060 }, { "grad_norm": 2.5428848266601562, "learning_rate": 2.138e-05, "loss": 0.1965, "step": 1070 }, { "grad_norm": 2.5370726585388184, "learning_rate": 2.158e-05, "loss": 0.1969, "step": 1080 }, { "grad_norm": 2.0185234546661377, "learning_rate": 2.178e-05, "loss": 0.1764, "step": 1090 }, { "grad_norm": 1.9558322429656982, "learning_rate": 2.198e-05, "loss": 0.1452, "step": 1100 }, { "grad_norm": 1.951897144317627, "learning_rate": 2.218e-05, "loss": 0.2115, "step": 1110 }, { "grad_norm": 1.5519278049468994, "learning_rate": 2.2380000000000003e-05, "loss": 0.142, "step": 1120 }, { "grad_norm": 1.760968565940857, "learning_rate": 2.258e-05, "loss": 0.1635, "step": 1130 }, { "grad_norm": 3.4402849674224854, "learning_rate": 2.2780000000000002e-05, "loss": 0.2009, "step": 1140 }, { "grad_norm": 2.6486635208129883, "learning_rate": 2.298e-05, "loss": 0.1869, "step": 1150 }, { "grad_norm": 2.259247064590454, "learning_rate": 2.318e-05, "loss": 0.172, "step": 1160 }, { "grad_norm": 2.2930827140808105, "learning_rate": 2.3380000000000003e-05, "loss": 0.1614, "step": 1170 }, { "grad_norm": 1.8407241106033325, "learning_rate": 2.358e-05, "loss": 0.144, "step": 1180 }, { "grad_norm": 2.537707805633545, "learning_rate": 2.3780000000000003e-05, "loss": 0.1552, "step": 1190 }, { "grad_norm": 3.5114846229553223, "learning_rate": 2.398e-05, "loss": 0.173, "step": 1200 }, { "grad_norm": 3.5085384845733643, "learning_rate": 2.418e-05, "loss": 0.19, "step": 1210 }, { "grad_norm": 1.4947302341461182, "learning_rate": 2.438e-05, "loss": 0.152, "step": 1220 }, { "grad_norm": 2.236494302749634, "learning_rate": 2.4580000000000002e-05, "loss": 0.1599, "step": 1230 }, { "grad_norm": 1.3809685707092285, "learning_rate": 2.478e-05, "loss": 0.1641, "step": 1240 }, { "grad_norm": 2.3399391174316406, "learning_rate": 2.498e-05, "loss": 0.1437, "step": 1250 }, { "grad_norm": 2.3176801204681396, "learning_rate": 2.5180000000000003e-05, "loss": 0.1668, "step": 1260 }, { "grad_norm": 1.9638986587524414, "learning_rate": 2.5380000000000004e-05, "loss": 0.1502, "step": 1270 }, { "grad_norm": 1.768080234527588, "learning_rate": 2.5580000000000002e-05, "loss": 0.1655, "step": 1280 }, { "grad_norm": 1.5898890495300293, "learning_rate": 2.5779999999999997e-05, "loss": 0.1437, "step": 1290 }, { "grad_norm": 1.7553728818893433, "learning_rate": 2.598e-05, "loss": 0.1492, "step": 1300 }, { "grad_norm": 1.749345064163208, "learning_rate": 2.618e-05, "loss": 0.1422, "step": 1310 }, { "grad_norm": 1.9791946411132812, "learning_rate": 2.6379999999999998e-05, "loss": 0.1575, "step": 1320 }, { "grad_norm": 2.140104293823242, "learning_rate": 2.658e-05, "loss": 0.1308, "step": 1330 }, { "grad_norm": 2.273991584777832, "learning_rate": 2.678e-05, "loss": 0.1446, "step": 1340 }, { "grad_norm": 1.765952467918396, "learning_rate": 2.698e-05, "loss": 0.1603, "step": 1350 }, { "grad_norm": 2.424708843231201, "learning_rate": 2.718e-05, "loss": 0.1624, "step": 1360 }, { "grad_norm": 2.2968926429748535, "learning_rate": 2.738e-05, "loss": 0.1224, "step": 1370 }, { "grad_norm": 1.669359803199768, "learning_rate": 2.758e-05, "loss": 0.1253, "step": 1380 }, { "grad_norm": 1.961016058921814, "learning_rate": 2.778e-05, "loss": 0.1487, "step": 1390 }, { "grad_norm": 1.805829405784607, "learning_rate": 2.798e-05, "loss": 0.1323, "step": 1400 }, { "grad_norm": 1.6038990020751953, "learning_rate": 2.818e-05, "loss": 0.1413, "step": 1410 }, { "grad_norm": 2.0714316368103027, "learning_rate": 2.8380000000000003e-05, "loss": 0.1551, "step": 1420 }, { "grad_norm": 1.7500633001327515, "learning_rate": 2.858e-05, "loss": 0.1277, "step": 1430 }, { "grad_norm": 2.1483771800994873, "learning_rate": 2.8780000000000002e-05, "loss": 0.1721, "step": 1440 }, { "grad_norm": 1.7815011739730835, "learning_rate": 2.898e-05, "loss": 0.1467, "step": 1450 }, { "grad_norm": 1.3851617574691772, "learning_rate": 2.9180000000000002e-05, "loss": 0.144, "step": 1460 }, { "grad_norm": 1.3885630369186401, "learning_rate": 2.9380000000000003e-05, "loss": 0.1406, "step": 1470 }, { "grad_norm": 2.0709519386291504, "learning_rate": 2.958e-05, "loss": 0.1372, "step": 1480 }, { "grad_norm": 1.9330464601516724, "learning_rate": 2.9780000000000003e-05, "loss": 0.1342, "step": 1490 }, { "grad_norm": 1.3637821674346924, "learning_rate": 2.998e-05, "loss": 0.1457, "step": 1500 }, { "grad_norm": 1.3421761989593506, "learning_rate": 3.0180000000000002e-05, "loss": 0.1239, "step": 1510 }, { "grad_norm": 1.9732586145401, "learning_rate": 3.0380000000000004e-05, "loss": 0.1472, "step": 1520 }, { "grad_norm": 1.4767309427261353, "learning_rate": 3.058e-05, "loss": 0.1488, "step": 1530 }, { "grad_norm": 2.2870492935180664, "learning_rate": 3.078e-05, "loss": 0.1464, "step": 1540 }, { "grad_norm": 1.6560474634170532, "learning_rate": 3.0980000000000005e-05, "loss": 0.134, "step": 1550 }, { "grad_norm": 1.8410605192184448, "learning_rate": 3.118e-05, "loss": 0.1635, "step": 1560 }, { "grad_norm": 1.2403407096862793, "learning_rate": 3.138e-05, "loss": 0.1347, "step": 1570 }, { "grad_norm": 2.1829280853271484, "learning_rate": 3.1580000000000006e-05, "loss": 0.1557, "step": 1580 }, { "grad_norm": 1.91374671459198, "learning_rate": 3.1780000000000004e-05, "loss": 0.1482, "step": 1590 }, { "grad_norm": 1.4037145376205444, "learning_rate": 3.198e-05, "loss": 0.1352, "step": 1600 }, { "grad_norm": 1.294970154762268, "learning_rate": 3.218e-05, "loss": 0.1188, "step": 1610 }, { "grad_norm": 1.648858904838562, "learning_rate": 3.238e-05, "loss": 0.1603, "step": 1620 }, { "grad_norm": 1.5366950035095215, "learning_rate": 3.2579999999999996e-05, "loss": 0.1708, "step": 1630 }, { "grad_norm": 1.2279011011123657, "learning_rate": 3.278e-05, "loss": 0.1405, "step": 1640 }, { "grad_norm": 1.5147924423217773, "learning_rate": 3.298e-05, "loss": 0.1501, "step": 1650 }, { "grad_norm": 2.436988115310669, "learning_rate": 3.318e-05, "loss": 0.1454, "step": 1660 }, { "grad_norm": 1.3503872156143188, "learning_rate": 3.338e-05, "loss": 0.1341, "step": 1670 }, { "grad_norm": 1.7752535343170166, "learning_rate": 3.358e-05, "loss": 0.1348, "step": 1680 }, { "grad_norm": 0.8828756213188171, "learning_rate": 3.378e-05, "loss": 0.1319, "step": 1690 }, { "grad_norm": 2.5315651893615723, "learning_rate": 3.398e-05, "loss": 0.1539, "step": 1700 }, { "grad_norm": 1.9102461338043213, "learning_rate": 3.418e-05, "loss": 0.1556, "step": 1710 }, { "grad_norm": 1.3507895469665527, "learning_rate": 3.438e-05, "loss": 0.1476, "step": 1720 }, { "grad_norm": 1.5133870840072632, "learning_rate": 3.4580000000000004e-05, "loss": 0.1366, "step": 1730 }, { "grad_norm": 2.0087478160858154, "learning_rate": 3.478e-05, "loss": 0.1324, "step": 1740 }, { "grad_norm": 1.5120078325271606, "learning_rate": 3.498e-05, "loss": 0.1381, "step": 1750 }, { "grad_norm": 2.21498703956604, "learning_rate": 3.518e-05, "loss": 0.1384, "step": 1760 }, { "grad_norm": 1.48771333694458, "learning_rate": 3.5380000000000003e-05, "loss": 0.1308, "step": 1770 }, { "grad_norm": 1.2768933773040771, "learning_rate": 3.558e-05, "loss": 0.1554, "step": 1780 }, { "grad_norm": 1.7277313470840454, "learning_rate": 3.578e-05, "loss": 0.1306, "step": 1790 }, { "grad_norm": 1.437050223350525, "learning_rate": 3.5980000000000004e-05, "loss": 0.1343, "step": 1800 }, { "grad_norm": 1.2504438161849976, "learning_rate": 3.618e-05, "loss": 0.1247, "step": 1810 }, { "grad_norm": 1.7310789823532104, "learning_rate": 3.638e-05, "loss": 0.1326, "step": 1820 }, { "grad_norm": 1.996170163154602, "learning_rate": 3.6580000000000006e-05, "loss": 0.1306, "step": 1830 }, { "grad_norm": 1.3490618467330933, "learning_rate": 3.6780000000000004e-05, "loss": 0.1371, "step": 1840 }, { "grad_norm": 1.1523011922836304, "learning_rate": 3.698e-05, "loss": 0.1657, "step": 1850 }, { "grad_norm": 1.4500136375427246, "learning_rate": 3.7180000000000007e-05, "loss": 0.1588, "step": 1860 }, { "grad_norm": 1.1105835437774658, "learning_rate": 3.7380000000000005e-05, "loss": 0.1329, "step": 1870 }, { "grad_norm": 1.47608482837677, "learning_rate": 3.758e-05, "loss": 0.1121, "step": 1880 }, { "grad_norm": 1.731147050857544, "learning_rate": 3.778000000000001e-05, "loss": 0.1305, "step": 1890 }, { "grad_norm": 1.194965124130249, "learning_rate": 3.7980000000000006e-05, "loss": 0.1357, "step": 1900 }, { "grad_norm": 1.4782819747924805, "learning_rate": 3.818e-05, "loss": 0.1539, "step": 1910 }, { "grad_norm": 1.0729936361312866, "learning_rate": 3.838e-05, "loss": 0.1425, "step": 1920 }, { "grad_norm": 1.0189318656921387, "learning_rate": 3.858e-05, "loss": 0.1314, "step": 1930 }, { "grad_norm": 1.479162573814392, "learning_rate": 3.878e-05, "loss": 0.1268, "step": 1940 }, { "grad_norm": 1.2203335762023926, "learning_rate": 3.898e-05, "loss": 0.1221, "step": 1950 }, { "grad_norm": 1.552198052406311, "learning_rate": 3.918e-05, "loss": 0.149, "step": 1960 }, { "grad_norm": 1.6309545040130615, "learning_rate": 3.938e-05, "loss": 0.1191, "step": 1970 }, { "grad_norm": 1.2312177419662476, "learning_rate": 3.958e-05, "loss": 0.1355, "step": 1980 }, { "grad_norm": 1.311187505722046, "learning_rate": 3.978e-05, "loss": 0.1188, "step": 1990 }, { "grad_norm": 1.9581589698791504, "learning_rate": 3.998e-05, "loss": 0.1267, "step": 2000 }, { "grad_norm": 1.164931058883667, "learning_rate": 4.018e-05, "loss": 0.1596, "step": 2010 }, { "grad_norm": 1.4680120944976807, "learning_rate": 4.038e-05, "loss": 0.1294, "step": 2020 }, { "grad_norm": 1.9081614017486572, "learning_rate": 4.058e-05, "loss": 0.1147, "step": 2030 }, { "grad_norm": 0.7715723514556885, "learning_rate": 4.078e-05, "loss": 0.1238, "step": 2040 }, { "grad_norm": 1.3621255159378052, "learning_rate": 4.0980000000000004e-05, "loss": 0.1316, "step": 2050 }, { "grad_norm": 1.8249250650405884, "learning_rate": 4.118e-05, "loss": 0.1191, "step": 2060 }, { "grad_norm": 1.7425193786621094, "learning_rate": 4.138e-05, "loss": 0.1035, "step": 2070 }, { "grad_norm": 1.6592049598693848, "learning_rate": 4.1580000000000005e-05, "loss": 0.1327, "step": 2080 }, { "grad_norm": 1.4097089767456055, "learning_rate": 4.178e-05, "loss": 0.1343, "step": 2090 }, { "grad_norm": 0.8993145823478699, "learning_rate": 4.198e-05, "loss": 0.1199, "step": 2100 }, { "grad_norm": 1.2581959962844849, "learning_rate": 4.2180000000000006e-05, "loss": 0.1573, "step": 2110 }, { "grad_norm": 1.3765215873718262, "learning_rate": 4.2380000000000004e-05, "loss": 0.1327, "step": 2120 }, { "grad_norm": 1.080201506614685, "learning_rate": 4.258e-05, "loss": 0.1201, "step": 2130 }, { "grad_norm": 1.182442545890808, "learning_rate": 4.278e-05, "loss": 0.1008, "step": 2140 }, { "grad_norm": 0.9688419103622437, "learning_rate": 4.2980000000000005e-05, "loss": 0.1307, "step": 2150 }, { "grad_norm": 1.2984864711761475, "learning_rate": 4.318e-05, "loss": 0.1266, "step": 2160 }, { "grad_norm": 1.4137212038040161, "learning_rate": 4.338e-05, "loss": 0.1182, "step": 2170 }, { "grad_norm": 0.8451075553894043, "learning_rate": 4.3580000000000006e-05, "loss": 0.1434, "step": 2180 }, { "grad_norm": 1.167461633682251, "learning_rate": 4.3780000000000004e-05, "loss": 0.1164, "step": 2190 }, { "grad_norm": 1.8233699798583984, "learning_rate": 4.398e-05, "loss": 0.1262, "step": 2200 }, { "grad_norm": 0.8227609992027283, "learning_rate": 4.418000000000001e-05, "loss": 0.1215, "step": 2210 }, { "grad_norm": 1.195960521697998, "learning_rate": 4.438e-05, "loss": 0.1089, "step": 2220 }, { "grad_norm": 1.2643051147460938, "learning_rate": 4.458e-05, "loss": 0.1249, "step": 2230 }, { "grad_norm": 0.9935944676399231, "learning_rate": 4.478e-05, "loss": 0.1193, "step": 2240 }, { "grad_norm": 1.0054552555084229, "learning_rate": 4.498e-05, "loss": 0.1288, "step": 2250 }, { "grad_norm": 1.0699927806854248, "learning_rate": 4.518e-05, "loss": 0.1078, "step": 2260 }, { "grad_norm": 0.9857912063598633, "learning_rate": 4.538e-05, "loss": 0.1046, "step": 2270 }, { "grad_norm": 1.1970878839492798, "learning_rate": 4.558e-05, "loss": 0.1122, "step": 2280 }, { "grad_norm": 1.6302025318145752, "learning_rate": 4.578e-05, "loss": 0.1503, "step": 2290 }, { "grad_norm": 0.7918042540550232, "learning_rate": 4.5980000000000004e-05, "loss": 0.101, "step": 2300 }, { "grad_norm": 1.0914539098739624, "learning_rate": 4.618e-05, "loss": 0.1245, "step": 2310 }, { "grad_norm": 1.1831012964248657, "learning_rate": 4.638e-05, "loss": 0.1117, "step": 2320 }, { "grad_norm": 1.3333849906921387, "learning_rate": 4.6580000000000005e-05, "loss": 0.1439, "step": 2330 }, { "grad_norm": 1.160601258277893, "learning_rate": 4.678e-05, "loss": 0.1231, "step": 2340 }, { "grad_norm": 0.7741008996963501, "learning_rate": 4.698e-05, "loss": 0.1194, "step": 2350 }, { "grad_norm": 1.5524438619613647, "learning_rate": 4.718e-05, "loss": 0.1261, "step": 2360 }, { "grad_norm": 0.9362724423408508, "learning_rate": 4.7380000000000004e-05, "loss": 0.1376, "step": 2370 }, { "grad_norm": 0.9975263476371765, "learning_rate": 4.758e-05, "loss": 0.1229, "step": 2380 }, { "grad_norm": 0.8787640333175659, "learning_rate": 4.778e-05, "loss": 0.1309, "step": 2390 }, { "grad_norm": 1.1039294004440308, "learning_rate": 4.7980000000000005e-05, "loss": 0.116, "step": 2400 }, { "grad_norm": 1.1863065958023071, "learning_rate": 4.818e-05, "loss": 0.1228, "step": 2410 }, { "grad_norm": 1.4402559995651245, "learning_rate": 4.838e-05, "loss": 0.1738, "step": 2420 }, { "grad_norm": 1.3415783643722534, "learning_rate": 4.8580000000000006e-05, "loss": 0.1333, "step": 2430 }, { "grad_norm": 1.3708784580230713, "learning_rate": 4.8780000000000004e-05, "loss": 0.1343, "step": 2440 }, { "grad_norm": 1.305099606513977, "learning_rate": 4.898e-05, "loss": 0.1139, "step": 2450 }, { "grad_norm": 0.991532564163208, "learning_rate": 4.918000000000001e-05, "loss": 0.0993, "step": 2460 }, { "grad_norm": 1.1411195993423462, "learning_rate": 4.9380000000000005e-05, "loss": 0.0995, "step": 2470 }, { "grad_norm": 0.9489014148712158, "learning_rate": 4.958e-05, "loss": 0.1335, "step": 2480 }, { "grad_norm": 1.1831046342849731, "learning_rate": 4.978e-05, "loss": 0.103, "step": 2490 }, { "grad_norm": 0.8711024522781372, "learning_rate": 4.9980000000000006e-05, "loss": 0.1147, "step": 2500 }, { "grad_norm": 1.2718356847763062, "learning_rate": 5.0180000000000004e-05, "loss": 0.1132, "step": 2510 }, { "grad_norm": 0.7362115383148193, "learning_rate": 5.038e-05, "loss": 0.1213, "step": 2520 }, { "grad_norm": 0.9108414053916931, "learning_rate": 5.058000000000001e-05, "loss": 0.1074, "step": 2530 }, { "grad_norm": 0.9626086950302124, "learning_rate": 5.0780000000000005e-05, "loss": 0.1025, "step": 2540 }, { "grad_norm": 0.7758634090423584, "learning_rate": 5.098e-05, "loss": 0.1117, "step": 2550 }, { "grad_norm": 1.2039620876312256, "learning_rate": 5.118000000000001e-05, "loss": 0.1256, "step": 2560 }, { "grad_norm": 1.1154561042785645, "learning_rate": 5.1380000000000006e-05, "loss": 0.1114, "step": 2570 }, { "grad_norm": 0.7883026003837585, "learning_rate": 5.1580000000000004e-05, "loss": 0.1187, "step": 2580 }, { "grad_norm": 0.7832427024841309, "learning_rate": 5.178000000000001e-05, "loss": 0.1104, "step": 2590 }, { "grad_norm": 0.9387140870094299, "learning_rate": 5.198000000000001e-05, "loss": 0.1205, "step": 2600 }, { "grad_norm": 1.149295687675476, "learning_rate": 5.2180000000000005e-05, "loss": 0.1172, "step": 2610 }, { "grad_norm": 0.988098680973053, "learning_rate": 5.238000000000001e-05, "loss": 0.1109, "step": 2620 }, { "grad_norm": 0.856472373008728, "learning_rate": 5.258000000000001e-05, "loss": 0.1075, "step": 2630 }, { "grad_norm": 1.0677911043167114, "learning_rate": 5.2780000000000006e-05, "loss": 0.1357, "step": 2640 }, { "grad_norm": 1.002280831336975, "learning_rate": 5.2980000000000004e-05, "loss": 0.1195, "step": 2650 }, { "grad_norm": 1.15299654006958, "learning_rate": 5.318000000000001e-05, "loss": 0.1284, "step": 2660 }, { "grad_norm": 0.8538424372673035, "learning_rate": 5.338000000000001e-05, "loss": 0.1381, "step": 2670 }, { "grad_norm": 1.1826976537704468, "learning_rate": 5.3580000000000005e-05, "loss": 0.133, "step": 2680 }, { "grad_norm": 1.0323041677474976, "learning_rate": 5.378e-05, "loss": 0.1088, "step": 2690 }, { "grad_norm": 0.9892850518226624, "learning_rate": 5.3979999999999995e-05, "loss": 0.1041, "step": 2700 }, { "grad_norm": 0.9016466736793518, "learning_rate": 5.418e-05, "loss": 0.1064, "step": 2710 }, { "grad_norm": 0.7684451937675476, "learning_rate": 5.438e-05, "loss": 0.1237, "step": 2720 }, { "grad_norm": 0.8715349435806274, "learning_rate": 5.4579999999999996e-05, "loss": 0.1039, "step": 2730 }, { "grad_norm": 1.3136632442474365, "learning_rate": 5.478e-05, "loss": 0.1105, "step": 2740 }, { "grad_norm": 0.9536234140396118, "learning_rate": 5.498e-05, "loss": 0.1105, "step": 2750 }, { "grad_norm": 1.0957611799240112, "learning_rate": 5.518e-05, "loss": 0.0927, "step": 2760 }, { "grad_norm": 1.0500699281692505, "learning_rate": 5.538e-05, "loss": 0.1124, "step": 2770 }, { "grad_norm": 0.8099943399429321, "learning_rate": 5.558e-05, "loss": 0.1248, "step": 2780 }, { "grad_norm": 0.8438727855682373, "learning_rate": 5.578e-05, "loss": 0.1103, "step": 2790 }, { "grad_norm": 0.9357572197914124, "learning_rate": 5.5979999999999996e-05, "loss": 0.1222, "step": 2800 }, { "grad_norm": 1.2773019075393677, "learning_rate": 5.618e-05, "loss": 0.1527, "step": 2810 }, { "grad_norm": 1.181683897972107, "learning_rate": 5.638e-05, "loss": 0.1262, "step": 2820 }, { "grad_norm": 0.7286288738250732, "learning_rate": 5.658e-05, "loss": 0.1052, "step": 2830 }, { "grad_norm": 0.8067030906677246, "learning_rate": 5.678e-05, "loss": 0.1351, "step": 2840 }, { "grad_norm": 0.9999143481254578, "learning_rate": 5.698e-05, "loss": 0.1224, "step": 2850 }, { "grad_norm": 0.9250352382659912, "learning_rate": 5.718e-05, "loss": 0.1142, "step": 2860 }, { "grad_norm": 0.7685880661010742, "learning_rate": 5.738e-05, "loss": 0.0984, "step": 2870 }, { "grad_norm": 0.856737494468689, "learning_rate": 5.758e-05, "loss": 0.1062, "step": 2880 }, { "grad_norm": 0.8719803690910339, "learning_rate": 5.778e-05, "loss": 0.1175, "step": 2890 }, { "grad_norm": 0.7429449558258057, "learning_rate": 5.7980000000000004e-05, "loss": 0.1264, "step": 2900 }, { "grad_norm": 1.1377439498901367, "learning_rate": 5.818e-05, "loss": 0.113, "step": 2910 }, { "grad_norm": 1.105124831199646, "learning_rate": 5.838e-05, "loss": 0.1205, "step": 2920 }, { "grad_norm": 0.8858707547187805, "learning_rate": 5.858e-05, "loss": 0.1012, "step": 2930 }, { "grad_norm": 1.0175861120224, "learning_rate": 5.878e-05, "loss": 0.1016, "step": 2940 }, { "grad_norm": 0.8367810845375061, "learning_rate": 5.898e-05, "loss": 0.0922, "step": 2950 }, { "grad_norm": 0.8710784316062927, "learning_rate": 5.918e-05, "loss": 0.1025, "step": 2960 }, { "grad_norm": 1.0402953624725342, "learning_rate": 5.9380000000000004e-05, "loss": 0.1006, "step": 2970 }, { "grad_norm": 0.8763878345489502, "learning_rate": 5.958e-05, "loss": 0.1062, "step": 2980 }, { "grad_norm": 1.6982557773590088, "learning_rate": 5.978e-05, "loss": 0.1189, "step": 2990 }, { "grad_norm": 1.1766700744628906, "learning_rate": 5.9980000000000005e-05, "loss": 0.1169, "step": 3000 }, { "grad_norm": 1.088989019393921, "learning_rate": 6.018e-05, "loss": 0.0975, "step": 3010 }, { "grad_norm": 0.9652125239372253, "learning_rate": 6.038e-05, "loss": 0.1122, "step": 3020 }, { "grad_norm": 0.8620757460594177, "learning_rate": 6.0580000000000006e-05, "loss": 0.0973, "step": 3030 }, { "grad_norm": 1.0115536451339722, "learning_rate": 6.0780000000000004e-05, "loss": 0.1048, "step": 3040 }, { "grad_norm": 1.1484942436218262, "learning_rate": 6.098e-05, "loss": 0.0987, "step": 3050 }, { "grad_norm": 0.9232482314109802, "learning_rate": 6.118000000000001e-05, "loss": 0.1144, "step": 3060 }, { "grad_norm": 0.7999034523963928, "learning_rate": 6.138e-05, "loss": 0.0995, "step": 3070 }, { "grad_norm": 0.8097646832466125, "learning_rate": 6.158e-05, "loss": 0.1036, "step": 3080 }, { "grad_norm": 1.1258857250213623, "learning_rate": 6.178000000000001e-05, "loss": 0.1268, "step": 3090 }, { "grad_norm": 1.8982599973678589, "learning_rate": 6.198e-05, "loss": 0.1171, "step": 3100 }, { "grad_norm": 0.6318702697753906, "learning_rate": 6.218e-05, "loss": 0.0955, "step": 3110 }, { "grad_norm": 0.8319879770278931, "learning_rate": 6.238000000000001e-05, "loss": 0.0824, "step": 3120 }, { "grad_norm": 1.0139930248260498, "learning_rate": 6.258e-05, "loss": 0.1005, "step": 3130 }, { "grad_norm": 0.831477701663971, "learning_rate": 6.278e-05, "loss": 0.1085, "step": 3140 }, { "grad_norm": 0.7283073663711548, "learning_rate": 6.298000000000001e-05, "loss": 0.1183, "step": 3150 }, { "grad_norm": 0.8643047213554382, "learning_rate": 6.318e-05, "loss": 0.1061, "step": 3160 }, { "grad_norm": 0.9654467105865479, "learning_rate": 6.338e-05, "loss": 0.1027, "step": 3170 }, { "grad_norm": 1.1763206720352173, "learning_rate": 6.358000000000001e-05, "loss": 0.1351, "step": 3180 }, { "grad_norm": 0.6956610083580017, "learning_rate": 6.378e-05, "loss": 0.1205, "step": 3190 }, { "grad_norm": 0.8414002656936646, "learning_rate": 6.398000000000001e-05, "loss": 0.1082, "step": 3200 }, { "grad_norm": 0.7689759135246277, "learning_rate": 6.418000000000001e-05, "loss": 0.1126, "step": 3210 }, { "grad_norm": 1.7665131092071533, "learning_rate": 6.438e-05, "loss": 0.1448, "step": 3220 }, { "grad_norm": 0.8510509729385376, "learning_rate": 6.458000000000001e-05, "loss": 0.1309, "step": 3230 }, { "grad_norm": 0.8111070990562439, "learning_rate": 6.478000000000001e-05, "loss": 0.0948, "step": 3240 }, { "grad_norm": 0.726142406463623, "learning_rate": 6.498e-05, "loss": 0.114, "step": 3250 }, { "grad_norm": 1.491731882095337, "learning_rate": 6.518000000000001e-05, "loss": 0.1271, "step": 3260 }, { "grad_norm": 1.024443507194519, "learning_rate": 6.538000000000001e-05, "loss": 0.1099, "step": 3270 }, { "grad_norm": 0.9041016697883606, "learning_rate": 6.558e-05, "loss": 0.0915, "step": 3280 }, { "grad_norm": 0.8089872598648071, "learning_rate": 6.578000000000001e-05, "loss": 0.1005, "step": 3290 }, { "grad_norm": 0.9768199920654297, "learning_rate": 6.598e-05, "loss": 0.1148, "step": 3300 }, { "grad_norm": 1.1090372800827026, "learning_rate": 6.618e-05, "loss": 0.1138, "step": 3310 }, { "grad_norm": 0.6871408224105835, "learning_rate": 6.638e-05, "loss": 0.1253, "step": 3320 }, { "grad_norm": 0.723433792591095, "learning_rate": 6.658e-05, "loss": 0.0949, "step": 3330 }, { "grad_norm": 0.9317322969436646, "learning_rate": 6.678e-05, "loss": 0.1068, "step": 3340 }, { "grad_norm": 0.6566445827484131, "learning_rate": 6.698e-05, "loss": 0.1091, "step": 3350 }, { "grad_norm": 1.1063474416732788, "learning_rate": 6.718e-05, "loss": 0.1012, "step": 3360 }, { "grad_norm": 0.689691424369812, "learning_rate": 6.738e-05, "loss": 0.1024, "step": 3370 }, { "grad_norm": 0.8004556894302368, "learning_rate": 6.758e-05, "loss": 0.1446, "step": 3380 }, { "grad_norm": 0.7232429385185242, "learning_rate": 6.778e-05, "loss": 0.1109, "step": 3390 }, { "grad_norm": 0.8063846826553345, "learning_rate": 6.798e-05, "loss": 0.113, "step": 3400 }, { "grad_norm": 1.0359282493591309, "learning_rate": 6.818e-05, "loss": 0.1079, "step": 3410 }, { "grad_norm": 1.002705454826355, "learning_rate": 6.838e-05, "loss": 0.0941, "step": 3420 }, { "grad_norm": 0.9362946152687073, "learning_rate": 6.858e-05, "loss": 0.0957, "step": 3430 }, { "grad_norm": 0.8080301284790039, "learning_rate": 6.878e-05, "loss": 0.112, "step": 3440 }, { "grad_norm": 0.791155219078064, "learning_rate": 6.898e-05, "loss": 0.1016, "step": 3450 }, { "grad_norm": 1.070873737335205, "learning_rate": 6.918e-05, "loss": 0.1067, "step": 3460 }, { "grad_norm": 0.8716144561767578, "learning_rate": 6.938e-05, "loss": 0.1097, "step": 3470 }, { "grad_norm": 0.8511151075363159, "learning_rate": 6.958e-05, "loss": 0.1144, "step": 3480 }, { "grad_norm": 0.7660459280014038, "learning_rate": 6.978e-05, "loss": 0.1004, "step": 3490 }, { "grad_norm": 0.7956763505935669, "learning_rate": 6.998e-05, "loss": 0.0855, "step": 3500 }, { "grad_norm": 0.9139102101325989, "learning_rate": 7.018e-05, "loss": 0.1069, "step": 3510 }, { "grad_norm": 0.7509374618530273, "learning_rate": 7.038e-05, "loss": 0.098, "step": 3520 }, { "grad_norm": 0.7979635000228882, "learning_rate": 7.058e-05, "loss": 0.0981, "step": 3530 }, { "grad_norm": 0.7399846911430359, "learning_rate": 7.078e-05, "loss": 0.0891, "step": 3540 }, { "grad_norm": 0.8520128726959229, "learning_rate": 7.098e-05, "loss": 0.1081, "step": 3550 }, { "grad_norm": 1.2763704061508179, "learning_rate": 7.118e-05, "loss": 0.0802, "step": 3560 }, { "grad_norm": 0.8150895237922668, "learning_rate": 7.138e-05, "loss": 0.1187, "step": 3570 }, { "grad_norm": 0.8125203251838684, "learning_rate": 7.158e-05, "loss": 0.1206, "step": 3580 }, { "grad_norm": 1.0183290243148804, "learning_rate": 7.178000000000001e-05, "loss": 0.1134, "step": 3590 }, { "grad_norm": 0.8479904532432556, "learning_rate": 7.198e-05, "loss": 0.1143, "step": 3600 }, { "grad_norm": 0.8950806260108948, "learning_rate": 7.218e-05, "loss": 0.1004, "step": 3610 }, { "grad_norm": 0.6813952922821045, "learning_rate": 7.238000000000001e-05, "loss": 0.0982, "step": 3620 }, { "grad_norm": 0.7372170090675354, "learning_rate": 7.258e-05, "loss": 0.0941, "step": 3630 }, { "grad_norm": 1.2499850988388062, "learning_rate": 7.278e-05, "loss": 0.1085, "step": 3640 }, { "grad_norm": 1.0892032384872437, "learning_rate": 7.298000000000001e-05, "loss": 0.115, "step": 3650 }, { "grad_norm": 0.7026942372322083, "learning_rate": 7.318e-05, "loss": 0.1091, "step": 3660 }, { "grad_norm": 0.8356026411056519, "learning_rate": 7.338e-05, "loss": 0.1133, "step": 3670 }, { "grad_norm": 0.9460915327072144, "learning_rate": 7.358000000000001e-05, "loss": 0.1133, "step": 3680 }, { "grad_norm": 0.7990611791610718, "learning_rate": 7.378e-05, "loss": 0.0997, "step": 3690 }, { "grad_norm": 0.7230422496795654, "learning_rate": 7.398e-05, "loss": 0.1065, "step": 3700 }, { "grad_norm": 0.8864678144454956, "learning_rate": 7.418000000000001e-05, "loss": 0.1102, "step": 3710 }, { "grad_norm": 0.5993984341621399, "learning_rate": 7.438e-05, "loss": 0.1091, "step": 3720 }, { "grad_norm": 0.8070017099380493, "learning_rate": 7.458000000000001e-05, "loss": 0.0969, "step": 3730 }, { "grad_norm": 0.676444411277771, "learning_rate": 7.478e-05, "loss": 0.0935, "step": 3740 }, { "grad_norm": 0.8065376281738281, "learning_rate": 7.498e-05, "loss": 0.1105, "step": 3750 }, { "grad_norm": 0.5435248613357544, "learning_rate": 7.518000000000001e-05, "loss": 0.093, "step": 3760 }, { "grad_norm": 0.8298556804656982, "learning_rate": 7.538e-05, "loss": 0.1109, "step": 3770 }, { "grad_norm": 0.8638710975646973, "learning_rate": 7.558e-05, "loss": 0.1069, "step": 3780 }, { "grad_norm": 0.933175802230835, "learning_rate": 7.578000000000001e-05, "loss": 0.1084, "step": 3790 }, { "grad_norm": 0.8879765868186951, "learning_rate": 7.598e-05, "loss": 0.1333, "step": 3800 }, { "grad_norm": 0.48486754298210144, "learning_rate": 7.618e-05, "loss": 0.0986, "step": 3810 }, { "grad_norm": 0.7761489152908325, "learning_rate": 7.638000000000001e-05, "loss": 0.107, "step": 3820 }, { "grad_norm": 0.6006704568862915, "learning_rate": 7.658e-05, "loss": 0.0916, "step": 3830 }, { "grad_norm": 0.702650249004364, "learning_rate": 7.678000000000001e-05, "loss": 0.0936, "step": 3840 }, { "grad_norm": 0.7310730814933777, "learning_rate": 7.698000000000001e-05, "loss": 0.1032, "step": 3850 }, { "grad_norm": 0.8775366544723511, "learning_rate": 7.718e-05, "loss": 0.1029, "step": 3860 }, { "grad_norm": 0.7558834552764893, "learning_rate": 7.738000000000001e-05, "loss": 0.092, "step": 3870 }, { "grad_norm": 0.6840866208076477, "learning_rate": 7.758000000000001e-05, "loss": 0.0849, "step": 3880 }, { "grad_norm": 1.0062370300292969, "learning_rate": 7.778e-05, "loss": 0.1101, "step": 3890 }, { "grad_norm": 0.8000427484512329, "learning_rate": 7.798000000000001e-05, "loss": 0.0986, "step": 3900 }, { "grad_norm": 1.071648359298706, "learning_rate": 7.818000000000001e-05, "loss": 0.0956, "step": 3910 }, { "grad_norm": 1.2869763374328613, "learning_rate": 7.838e-05, "loss": 0.1175, "step": 3920 }, { "grad_norm": 0.6356104612350464, "learning_rate": 7.858000000000001e-05, "loss": 0.1018, "step": 3930 }, { "grad_norm": 1.0901991128921509, "learning_rate": 7.878e-05, "loss": 0.1026, "step": 3940 }, { "grad_norm": 0.7468140125274658, "learning_rate": 7.897999999999999e-05, "loss": 0.1011, "step": 3950 }, { "grad_norm": 0.816249430179596, "learning_rate": 7.918e-05, "loss": 0.1107, "step": 3960 }, { "grad_norm": 0.5884287357330322, "learning_rate": 7.938e-05, "loss": 0.0782, "step": 3970 }, { "grad_norm": 0.6373471617698669, "learning_rate": 7.958e-05, "loss": 0.1131, "step": 3980 }, { "grad_norm": 0.7779890298843384, "learning_rate": 7.978e-05, "loss": 0.1018, "step": 3990 }, { "grad_norm": 0.8450542092323303, "learning_rate": 7.998e-05, "loss": 0.0998, "step": 4000 }, { "grad_norm": 0.6777880191802979, "learning_rate": 8.018e-05, "loss": 0.0935, "step": 4010 }, { "grad_norm": 0.6117426156997681, "learning_rate": 8.038e-05, "loss": 0.0972, "step": 4020 }, { "grad_norm": 0.6589750647544861, "learning_rate": 8.058e-05, "loss": 0.098, "step": 4030 }, { "grad_norm": 0.7848002910614014, "learning_rate": 8.078e-05, "loss": 0.0888, "step": 4040 }, { "grad_norm": 0.6850364804267883, "learning_rate": 8.098e-05, "loss": 0.1076, "step": 4050 }, { "grad_norm": 0.8841881155967712, "learning_rate": 8.118e-05, "loss": 0.1005, "step": 4060 }, { "grad_norm": 0.6852495670318604, "learning_rate": 8.138e-05, "loss": 0.0878, "step": 4070 }, { "grad_norm": 0.6859447360038757, "learning_rate": 8.158e-05, "loss": 0.1044, "step": 4080 }, { "grad_norm": 0.6326321959495544, "learning_rate": 8.178e-05, "loss": 0.0917, "step": 4090 }, { "grad_norm": 0.8162169456481934, "learning_rate": 8.198e-05, "loss": 0.0787, "step": 4100 }, { "grad_norm": 1.0248223543167114, "learning_rate": 8.218e-05, "loss": 0.1017, "step": 4110 }, { "grad_norm": 0.9729684591293335, "learning_rate": 8.238000000000001e-05, "loss": 0.1205, "step": 4120 }, { "grad_norm": 0.604470431804657, "learning_rate": 8.258e-05, "loss": 0.087, "step": 4130 }, { "grad_norm": 0.6283584237098694, "learning_rate": 8.278e-05, "loss": 0.0961, "step": 4140 }, { "grad_norm": 1.092703938484192, "learning_rate": 8.298000000000001e-05, "loss": 0.1222, "step": 4150 }, { "grad_norm": 0.6031334400177002, "learning_rate": 8.318e-05, "loss": 0.0776, "step": 4160 }, { "grad_norm": 0.6195088624954224, "learning_rate": 8.338e-05, "loss": 0.0934, "step": 4170 }, { "grad_norm": 0.6271294355392456, "learning_rate": 8.358e-05, "loss": 0.0937, "step": 4180 }, { "grad_norm": 1.0060455799102783, "learning_rate": 8.378e-05, "loss": 0.104, "step": 4190 }, { "grad_norm": 0.5826853513717651, "learning_rate": 8.398e-05, "loss": 0.0938, "step": 4200 }, { "grad_norm": 0.7909671664237976, "learning_rate": 8.418e-05, "loss": 0.1092, "step": 4210 }, { "grad_norm": 0.6888346076011658, "learning_rate": 8.438e-05, "loss": 0.1197, "step": 4220 }, { "grad_norm": 0.9161320328712463, "learning_rate": 8.458e-05, "loss": 0.1208, "step": 4230 }, { "grad_norm": 0.7674065828323364, "learning_rate": 8.478e-05, "loss": 0.1101, "step": 4240 }, { "grad_norm": 0.722567617893219, "learning_rate": 8.498e-05, "loss": 0.1108, "step": 4250 }, { "grad_norm": 0.9003739953041077, "learning_rate": 8.518000000000001e-05, "loss": 0.1137, "step": 4260 }, { "grad_norm": 1.2488899230957031, "learning_rate": 8.538e-05, "loss": 0.1065, "step": 4270 }, { "grad_norm": 0.9554003477096558, "learning_rate": 8.558e-05, "loss": 0.1084, "step": 4280 }, { "grad_norm": 0.7637826800346375, "learning_rate": 8.578000000000001e-05, "loss": 0.1008, "step": 4290 }, { "grad_norm": 0.7809488773345947, "learning_rate": 8.598e-05, "loss": 0.1045, "step": 4300 }, { "grad_norm": 0.8199631571769714, "learning_rate": 8.618e-05, "loss": 0.1254, "step": 4310 }, { "grad_norm": 0.8796407580375671, "learning_rate": 8.638000000000001e-05, "loss": 0.153, "step": 4320 }, { "grad_norm": 1.0208702087402344, "learning_rate": 8.658e-05, "loss": 0.1044, "step": 4330 }, { "grad_norm": 0.6230844259262085, "learning_rate": 8.678e-05, "loss": 0.0912, "step": 4340 }, { "grad_norm": 0.6172550916671753, "learning_rate": 8.698000000000001e-05, "loss": 0.0765, "step": 4350 }, { "grad_norm": 0.7685699462890625, "learning_rate": 8.718e-05, "loss": 0.0999, "step": 4360 }, { "grad_norm": 0.850823700428009, "learning_rate": 8.738000000000001e-05, "loss": 0.1038, "step": 4370 }, { "grad_norm": 0.923062264919281, "learning_rate": 8.758000000000001e-05, "loss": 0.1088, "step": 4380 }, { "grad_norm": 0.8027415871620178, "learning_rate": 8.778e-05, "loss": 0.0989, "step": 4390 }, { "grad_norm": 0.6547898054122925, "learning_rate": 8.798000000000001e-05, "loss": 0.1254, "step": 4400 }, { "grad_norm": 0.8299077153205872, "learning_rate": 8.818000000000001e-05, "loss": 0.1106, "step": 4410 }, { "grad_norm": 0.9025068879127502, "learning_rate": 8.838e-05, "loss": 0.1111, "step": 4420 }, { "grad_norm": 0.5788818597793579, "learning_rate": 8.858000000000001e-05, "loss": 0.106, "step": 4430 }, { "grad_norm": 0.8205111622810364, "learning_rate": 8.878000000000001e-05, "loss": 0.099, "step": 4440 }, { "grad_norm": 0.7537466287612915, "learning_rate": 8.898e-05, "loss": 0.1219, "step": 4450 }, { "grad_norm": 0.8846111297607422, "learning_rate": 8.918000000000001e-05, "loss": 0.1079, "step": 4460 }, { "grad_norm": 0.763796329498291, "learning_rate": 8.938e-05, "loss": 0.114, "step": 4470 }, { "grad_norm": 0.7674116492271423, "learning_rate": 8.958e-05, "loss": 0.0983, "step": 4480 }, { "grad_norm": 0.8815690279006958, "learning_rate": 8.978000000000001e-05, "loss": 0.1128, "step": 4490 }, { "grad_norm": 0.6880037784576416, "learning_rate": 8.998e-05, "loss": 0.102, "step": 4500 }, { "grad_norm": 0.7254507541656494, "learning_rate": 9.018000000000001e-05, "loss": 0.1245, "step": 4510 }, { "grad_norm": 1.0345450639724731, "learning_rate": 9.038000000000001e-05, "loss": 0.1123, "step": 4520 }, { "grad_norm": 0.8974489569664001, "learning_rate": 9.058e-05, "loss": 0.1324, "step": 4530 }, { "grad_norm": 0.8984237909317017, "learning_rate": 9.078000000000001e-05, "loss": 0.1348, "step": 4540 }, { "grad_norm": 0.7160595059394836, "learning_rate": 9.098000000000001e-05, "loss": 0.1265, "step": 4550 }, { "grad_norm": 0.6746520400047302, "learning_rate": 9.118e-05, "loss": 0.1012, "step": 4560 }, { "grad_norm": 0.6093479990959167, "learning_rate": 9.138e-05, "loss": 0.1231, "step": 4570 }, { "grad_norm": 1.1615568399429321, "learning_rate": 9.158e-05, "loss": 0.1117, "step": 4580 }, { "grad_norm": 0.7336333990097046, "learning_rate": 9.178e-05, "loss": 0.1224, "step": 4590 }, { "grad_norm": 0.7883663177490234, "learning_rate": 9.198e-05, "loss": 0.108, "step": 4600 }, { "grad_norm": 0.7641025185585022, "learning_rate": 9.218e-05, "loss": 0.1111, "step": 4610 }, { "grad_norm": 0.6439140439033508, "learning_rate": 9.238e-05, "loss": 0.0947, "step": 4620 }, { "grad_norm": 1.2768187522888184, "learning_rate": 9.258e-05, "loss": 0.1099, "step": 4630 }, { "grad_norm": 0.8640211820602417, "learning_rate": 9.278e-05, "loss": 0.1193, "step": 4640 }, { "grad_norm": 0.8575178980827332, "learning_rate": 9.298e-05, "loss": 0.1048, "step": 4650 }, { "grad_norm": 0.7345477938652039, "learning_rate": 9.318e-05, "loss": 0.1082, "step": 4660 }, { "grad_norm": 0.9878590106964111, "learning_rate": 9.338e-05, "loss": 0.092, "step": 4670 }, { "grad_norm": 0.6033908724784851, "learning_rate": 9.358e-05, "loss": 0.0987, "step": 4680 }, { "grad_norm": 0.5749732255935669, "learning_rate": 9.378e-05, "loss": 0.0982, "step": 4690 }, { "grad_norm": 0.6847135424613953, "learning_rate": 9.398e-05, "loss": 0.1043, "step": 4700 }, { "grad_norm": 0.5687528848648071, "learning_rate": 9.418e-05, "loss": 0.1103, "step": 4710 }, { "grad_norm": 0.6227893233299255, "learning_rate": 9.438e-05, "loss": 0.1069, "step": 4720 }, { "grad_norm": 0.6520886421203613, "learning_rate": 9.458e-05, "loss": 0.107, "step": 4730 }, { "grad_norm": 0.8329691290855408, "learning_rate": 9.478e-05, "loss": 0.0986, "step": 4740 }, { "grad_norm": 1.4543428421020508, "learning_rate": 9.498e-05, "loss": 0.1016, "step": 4750 }, { "grad_norm": 1.048619270324707, "learning_rate": 9.518000000000001e-05, "loss": 0.1099, "step": 4760 }, { "grad_norm": 0.9013441205024719, "learning_rate": 9.538e-05, "loss": 0.114, "step": 4770 }, { "grad_norm": 1.055310606956482, "learning_rate": 9.558e-05, "loss": 0.1136, "step": 4780 }, { "grad_norm": 0.6835809946060181, "learning_rate": 9.578000000000001e-05, "loss": 0.0979, "step": 4790 }, { "grad_norm": 0.6577360033988953, "learning_rate": 9.598e-05, "loss": 0.1052, "step": 4800 }, { "grad_norm": 0.928693950176239, "learning_rate": 9.618e-05, "loss": 0.1009, "step": 4810 }, { "grad_norm": 0.6204549074172974, "learning_rate": 9.638000000000001e-05, "loss": 0.0992, "step": 4820 }, { "grad_norm": 0.7348156571388245, "learning_rate": 9.658e-05, "loss": 0.1097, "step": 4830 }, { "grad_norm": 0.6143003106117249, "learning_rate": 9.678e-05, "loss": 0.1019, "step": 4840 }, { "grad_norm": 0.8413380980491638, "learning_rate": 9.698000000000001e-05, "loss": 0.1255, "step": 4850 }, { "grad_norm": 0.6700335741043091, "learning_rate": 9.718e-05, "loss": 0.0993, "step": 4860 }, { "grad_norm": 0.7454223036766052, "learning_rate": 9.738e-05, "loss": 0.1115, "step": 4870 }, { "grad_norm": 0.8814431428909302, "learning_rate": 9.758000000000001e-05, "loss": 0.1002, "step": 4880 }, { "grad_norm": 0.8362065553665161, "learning_rate": 9.778e-05, "loss": 0.1049, "step": 4890 }, { "grad_norm": 0.6192665100097656, "learning_rate": 9.798000000000001e-05, "loss": 0.1125, "step": 4900 }, { "grad_norm": 0.7068849802017212, "learning_rate": 9.818000000000001e-05, "loss": 0.1179, "step": 4910 }, { "grad_norm": 0.7047024965286255, "learning_rate": 9.838e-05, "loss": 0.1259, "step": 4920 }, { "grad_norm": 1.034216284751892, "learning_rate": 9.858000000000001e-05, "loss": 0.1188, "step": 4930 }, { "grad_norm": 0.6781094074249268, "learning_rate": 9.878e-05, "loss": 0.114, "step": 4940 }, { "grad_norm": 0.6725171208381653, "learning_rate": 9.898e-05, "loss": 0.1163, "step": 4950 }, { "grad_norm": 0.5920510292053223, "learning_rate": 9.918000000000001e-05, "loss": 0.1022, "step": 4960 }, { "grad_norm": 0.8882537484169006, "learning_rate": 9.938e-05, "loss": 0.1269, "step": 4970 }, { "grad_norm": 0.7428448796272278, "learning_rate": 9.958e-05, "loss": 0.1187, "step": 4980 }, { "grad_norm": 0.6157662272453308, "learning_rate": 9.978000000000001e-05, "loss": 0.109, "step": 4990 }, { "grad_norm": 0.7013158202171326, "learning_rate": 9.998e-05, "loss": 0.1186, "step": 5000 }, { "grad_norm": 0.7433786392211914, "learning_rate": 9.999999778549045e-05, "loss": 0.1062, "step": 5010 }, { "grad_norm": 0.7640990614891052, "learning_rate": 9.999999013039593e-05, "loss": 0.1023, "step": 5020 }, { "grad_norm": 0.6293370127677917, "learning_rate": 9.999997700737766e-05, "loss": 0.0905, "step": 5030 }, { "grad_norm": 0.6147626042366028, "learning_rate": 9.999995841643709e-05, "loss": 0.0797, "step": 5040 }, { "grad_norm": 0.5667362213134766, "learning_rate": 9.999993435757623e-05, "loss": 0.1048, "step": 5050 }, { "grad_norm": 0.513820230960846, "learning_rate": 9.999990483079773e-05, "loss": 0.0739, "step": 5060 }, { "grad_norm": 0.5506157875061035, "learning_rate": 9.999986983610481e-05, "loss": 0.1317, "step": 5070 }, { "grad_norm": 0.6240319013595581, "learning_rate": 9.99998293735013e-05, "loss": 0.1013, "step": 5080 }, { "grad_norm": 0.5468055009841919, "learning_rate": 9.999978344299161e-05, "loss": 0.0927, "step": 5090 }, { "grad_norm": 0.5403544306755066, "learning_rate": 9.99997320445808e-05, "loss": 0.0881, "step": 5100 }, { "grad_norm": 0.9411703944206238, "learning_rate": 9.999967517827444e-05, "loss": 0.119, "step": 5110 }, { "grad_norm": 0.8228616714477539, "learning_rate": 9.999961284407879e-05, "loss": 0.1188, "step": 5120 }, { "grad_norm": 0.8880684971809387, "learning_rate": 9.999954504200067e-05, "loss": 0.1212, "step": 5130 }, { "grad_norm": 0.6813204884529114, "learning_rate": 9.999947177204744e-05, "loss": 0.1135, "step": 5140 }, { "grad_norm": 0.5709890127182007, "learning_rate": 9.999939303422718e-05, "loss": 0.0991, "step": 5150 }, { "grad_norm": 0.5274611711502075, "learning_rate": 9.999930882854847e-05, "loss": 0.1219, "step": 5160 }, { "grad_norm": 0.7945351600646973, "learning_rate": 9.999921915502051e-05, "loss": 0.1035, "step": 5170 }, { "grad_norm": 0.655016303062439, "learning_rate": 9.99991240136531e-05, "loss": 0.0827, "step": 5180 }, { "grad_norm": 0.811129093170166, "learning_rate": 9.999902340445668e-05, "loss": 0.1179, "step": 5190 }, { "grad_norm": 0.6880759596824646, "learning_rate": 9.999891732744224e-05, "loss": 0.086, "step": 5200 }, { "grad_norm": 0.669384241104126, "learning_rate": 9.999880578262135e-05, "loss": 0.1, "step": 5210 }, { "grad_norm": 0.9075090885162354, "learning_rate": 9.999868877000624e-05, "loss": 0.099, "step": 5220 }, { "grad_norm": 0.743440568447113, "learning_rate": 9.99985662896097e-05, "loss": 0.0926, "step": 5230 }, { "grad_norm": 0.6647671461105347, "learning_rate": 9.999843834144513e-05, "loss": 0.1052, "step": 5240 }, { "grad_norm": 0.819734513759613, "learning_rate": 9.99983049255265e-05, "loss": 0.117, "step": 5250 }, { "grad_norm": 0.8658069372177124, "learning_rate": 9.999816604186843e-05, "loss": 0.0906, "step": 5260 }, { "grad_norm": 0.7474134564399719, "learning_rate": 9.999802169048609e-05, "loss": 0.0963, "step": 5270 }, { "grad_norm": 0.5253500938415527, "learning_rate": 9.999787187139527e-05, "loss": 0.1059, "step": 5280 }, { "grad_norm": 0.7131204009056091, "learning_rate": 9.999771658461234e-05, "loss": 0.1063, "step": 5290 }, { "grad_norm": 0.4798443019390106, "learning_rate": 9.999755583015431e-05, "loss": 0.0952, "step": 5300 }, { "grad_norm": 0.5963921546936035, "learning_rate": 9.999738960803874e-05, "loss": 0.0806, "step": 5310 }, { "grad_norm": 0.9452157020568848, "learning_rate": 9.99972179182838e-05, "loss": 0.1012, "step": 5320 }, { "grad_norm": 0.7833077907562256, "learning_rate": 9.99970407609083e-05, "loss": 0.1092, "step": 5330 }, { "grad_norm": 0.6371163129806519, "learning_rate": 9.999685813593159e-05, "loss": 0.0868, "step": 5340 }, { "grad_norm": 0.4741222858428955, "learning_rate": 9.999667004337362e-05, "loss": 0.0828, "step": 5350 }, { "grad_norm": 0.8321903944015503, "learning_rate": 9.9996476483255e-05, "loss": 0.0998, "step": 5360 }, { "grad_norm": 0.6570354700088501, "learning_rate": 9.999627745559688e-05, "loss": 0.0877, "step": 5370 }, { "grad_norm": 0.7659097909927368, "learning_rate": 9.999607296042101e-05, "loss": 0.1097, "step": 5380 }, { "grad_norm": 0.6122344136238098, "learning_rate": 9.99958629977498e-05, "loss": 0.0826, "step": 5390 }, { "grad_norm": 0.9396655559539795, "learning_rate": 9.999564756760615e-05, "loss": 0.1037, "step": 5400 }, { "grad_norm": 0.5437023043632507, "learning_rate": 9.999542667001366e-05, "loss": 0.0925, "step": 5410 }, { "grad_norm": 0.7191354632377625, "learning_rate": 9.999520030499647e-05, "loss": 0.1129, "step": 5420 }, { "grad_norm": 0.7278963923454285, "learning_rate": 9.999496847257936e-05, "loss": 0.1034, "step": 5430 }, { "grad_norm": 0.6370688080787659, "learning_rate": 9.999473117278764e-05, "loss": 0.1213, "step": 5440 }, { "grad_norm": 0.7073980569839478, "learning_rate": 9.999448840564731e-05, "loss": 0.0856, "step": 5450 }, { "grad_norm": 0.9785199165344238, "learning_rate": 9.999424017118488e-05, "loss": 0.0876, "step": 5460 }, { "grad_norm": 0.7358378767967224, "learning_rate": 9.999398646942751e-05, "loss": 0.0963, "step": 5470 }, { "grad_norm": 0.5803406238555908, "learning_rate": 9.999372730040296e-05, "loss": 0.0953, "step": 5480 }, { "grad_norm": 0.6147920489311218, "learning_rate": 9.999346266413953e-05, "loss": 0.0979, "step": 5490 }, { "grad_norm": 0.6988048553466797, "learning_rate": 9.99931925606662e-05, "loss": 0.1059, "step": 5500 }, { "grad_norm": 0.5387954115867615, "learning_rate": 9.99929169900125e-05, "loss": 0.1036, "step": 5510 }, { "grad_norm": 0.5938613414764404, "learning_rate": 9.999263595220855e-05, "loss": 0.0785, "step": 5520 }, { "grad_norm": 0.6445498466491699, "learning_rate": 9.99923494472851e-05, "loss": 0.1015, "step": 5530 }, { "grad_norm": 0.6709776520729065, "learning_rate": 9.999205747527348e-05, "loss": 0.0913, "step": 5540 }, { "grad_norm": 0.5468908548355103, "learning_rate": 9.999176003620561e-05, "loss": 0.0926, "step": 5550 }, { "grad_norm": 0.6085084080696106, "learning_rate": 9.999145713011405e-05, "loss": 0.0935, "step": 5560 }, { "grad_norm": 0.729465901851654, "learning_rate": 9.999114875703186e-05, "loss": 0.0964, "step": 5570 }, { "grad_norm": 0.6749862432479858, "learning_rate": 9.999083491699281e-05, "loss": 0.113, "step": 5580 }, { "grad_norm": 0.6825227737426758, "learning_rate": 9.999051561003123e-05, "loss": 0.1001, "step": 5590 }, { "grad_norm": 0.712219774723053, "learning_rate": 9.999019083618202e-05, "loss": 0.0944, "step": 5600 }, { "grad_norm": 0.7717708945274353, "learning_rate": 9.99898605954807e-05, "loss": 0.1028, "step": 5610 }, { "grad_norm": 0.7421357035636902, "learning_rate": 9.998952488796338e-05, "loss": 0.0906, "step": 5620 }, { "grad_norm": 0.43797755241394043, "learning_rate": 9.998918371366676e-05, "loss": 0.0897, "step": 5630 }, { "grad_norm": 0.592625081539154, "learning_rate": 9.99888370726282e-05, "loss": 0.0887, "step": 5640 }, { "grad_norm": 0.5333631038665771, "learning_rate": 9.998848496488556e-05, "loss": 0.1125, "step": 5650 }, { "grad_norm": 0.7049597501754761, "learning_rate": 9.998812739047736e-05, "loss": 0.0811, "step": 5660 }, { "grad_norm": 0.7127244472503662, "learning_rate": 9.99877643494427e-05, "loss": 0.1117, "step": 5670 }, { "grad_norm": 0.8044116497039795, "learning_rate": 9.998739584182128e-05, "loss": 0.0968, "step": 5680 }, { "grad_norm": 0.7221351861953735, "learning_rate": 9.998702186765342e-05, "loss": 0.1122, "step": 5690 }, { "grad_norm": 0.4842182397842407, "learning_rate": 9.998664242698e-05, "loss": 0.0947, "step": 5700 }, { "grad_norm": 0.632375180721283, "learning_rate": 9.998625751984251e-05, "loss": 0.0931, "step": 5710 }, { "grad_norm": 0.5530588626861572, "learning_rate": 9.998586714628307e-05, "loss": 0.0902, "step": 5720 }, { "grad_norm": 0.7459558248519897, "learning_rate": 9.998547130634432e-05, "loss": 0.0982, "step": 5730 }, { "grad_norm": 0.9216886162757874, "learning_rate": 9.99850700000696e-05, "loss": 0.1129, "step": 5740 }, { "grad_norm": 0.5519036054611206, "learning_rate": 9.998466322750278e-05, "loss": 0.0803, "step": 5750 }, { "grad_norm": 0.7472763061523438, "learning_rate": 9.998425098868834e-05, "loss": 0.0942, "step": 5760 }, { "grad_norm": 0.5048765540122986, "learning_rate": 9.998383328367136e-05, "loss": 0.0828, "step": 5770 }, { "grad_norm": 0.8838035464286804, "learning_rate": 9.99834101124975e-05, "loss": 0.1102, "step": 5780 }, { "grad_norm": 0.9299794435501099, "learning_rate": 9.998298147521309e-05, "loss": 0.0904, "step": 5790 }, { "grad_norm": 0.7688820958137512, "learning_rate": 9.998254737186496e-05, "loss": 0.117, "step": 5800 }, { "grad_norm": 0.5672100186347961, "learning_rate": 9.99821078025006e-05, "loss": 0.0948, "step": 5810 }, { "grad_norm": 0.805818498134613, "learning_rate": 9.998166276716807e-05, "loss": 0.098, "step": 5820 }, { "grad_norm": 0.6378475427627563, "learning_rate": 9.998121226591606e-05, "loss": 0.0885, "step": 5830 }, { "grad_norm": 0.5852627158164978, "learning_rate": 9.998075629879382e-05, "loss": 0.0912, "step": 5840 }, { "grad_norm": 0.662926435470581, "learning_rate": 9.99802948658512e-05, "loss": 0.0775, "step": 5850 }, { "grad_norm": 0.7214818000793457, "learning_rate": 9.99798279671387e-05, "loss": 0.0959, "step": 5860 }, { "grad_norm": 0.5233770608901978, "learning_rate": 9.997935560270734e-05, "loss": 0.0948, "step": 5870 }, { "grad_norm": 0.6718778014183044, "learning_rate": 9.997887777260879e-05, "loss": 0.1025, "step": 5880 }, { "grad_norm": 0.569084107875824, "learning_rate": 9.997839447689532e-05, "loss": 0.0952, "step": 5890 }, { "grad_norm": 0.594382643699646, "learning_rate": 9.997790571561978e-05, "loss": 0.1005, "step": 5900 }, { "grad_norm": 0.5452378988265991, "learning_rate": 9.99774114888356e-05, "loss": 0.1098, "step": 5910 }, { "grad_norm": 0.7712336778640747, "learning_rate": 9.997691179659684e-05, "loss": 0.0834, "step": 5920 }, { "grad_norm": 0.5193372368812561, "learning_rate": 9.997640663895815e-05, "loss": 0.0846, "step": 5930 }, { "grad_norm": 0.6011826992034912, "learning_rate": 9.997589601597477e-05, "loss": 0.0933, "step": 5940 }, { "grad_norm": 0.63763427734375, "learning_rate": 9.997537992770252e-05, "loss": 0.0948, "step": 5950 }, { "grad_norm": 1.0243967771530151, "learning_rate": 9.997485837419788e-05, "loss": 0.0944, "step": 5960 }, { "grad_norm": 0.7639005780220032, "learning_rate": 9.997433135551786e-05, "loss": 0.1045, "step": 5970 }, { "grad_norm": 0.433597207069397, "learning_rate": 9.997379887172009e-05, "loss": 0.089, "step": 5980 }, { "grad_norm": 0.6416061520576477, "learning_rate": 9.997326092286281e-05, "loss": 0.1005, "step": 5990 }, { "grad_norm": 0.6202622056007385, "learning_rate": 9.997271750900486e-05, "loss": 0.1095, "step": 6000 }, { "grad_norm": 0.6560165882110596, "learning_rate": 9.997216863020565e-05, "loss": 0.0894, "step": 6010 }, { "grad_norm": 0.5661751627922058, "learning_rate": 9.99716142865252e-05, "loss": 0.095, "step": 6020 }, { "grad_norm": 0.5906537771224976, "learning_rate": 9.997105447802415e-05, "loss": 0.0836, "step": 6030 }, { "grad_norm": 0.5525203347206116, "learning_rate": 9.997048920476373e-05, "loss": 0.0915, "step": 6040 }, { "grad_norm": 0.562840461730957, "learning_rate": 9.996991846680572e-05, "loss": 0.0971, "step": 6050 }, { "grad_norm": 0.48199376463890076, "learning_rate": 9.996934226421257e-05, "loss": 0.0804, "step": 6060 }, { "grad_norm": 0.5969526767730713, "learning_rate": 9.996876059704726e-05, "loss": 0.0854, "step": 6070 }, { "grad_norm": 0.6692765951156616, "learning_rate": 9.996817346537343e-05, "loss": 0.0754, "step": 6080 }, { "grad_norm": 0.5895896553993225, "learning_rate": 9.996758086925526e-05, "loss": 0.0982, "step": 6090 }, { "grad_norm": 0.5111535787582397, "learning_rate": 9.996698280875759e-05, "loss": 0.1035, "step": 6100 }, { "grad_norm": 0.548424243927002, "learning_rate": 9.99663792839458e-05, "loss": 0.0784, "step": 6110 }, { "grad_norm": 0.8046601414680481, "learning_rate": 9.99657702948859e-05, "loss": 0.0982, "step": 6120 }, { "grad_norm": 0.5645396113395691, "learning_rate": 9.996515584164448e-05, "loss": 0.075, "step": 6130 }, { "grad_norm": 0.8996025323867798, "learning_rate": 9.996453592428873e-05, "loss": 0.0804, "step": 6140 }, { "grad_norm": 0.6267890334129333, "learning_rate": 9.996391054288646e-05, "loss": 0.1126, "step": 6150 }, { "grad_norm": 0.633830726146698, "learning_rate": 9.996327969750605e-05, "loss": 0.1169, "step": 6160 }, { "grad_norm": 0.7145974040031433, "learning_rate": 9.996264338821649e-05, "loss": 0.1054, "step": 6170 }, { "grad_norm": 0.4930194020271301, "learning_rate": 9.996200161508735e-05, "loss": 0.0744, "step": 6180 }, { "grad_norm": 0.48952555656433105, "learning_rate": 9.996135437818885e-05, "loss": 0.0936, "step": 6190 }, { "grad_norm": 0.6348698139190674, "learning_rate": 9.996070167759175e-05, "loss": 0.0929, "step": 6200 }, { "grad_norm": 0.9334197044372559, "learning_rate": 9.996004351336743e-05, "loss": 0.1014, "step": 6210 }, { "grad_norm": 0.7853072881698608, "learning_rate": 9.995937988558785e-05, "loss": 0.0903, "step": 6220 }, { "grad_norm": 0.6397085785865784, "learning_rate": 9.995871079432561e-05, "loss": 0.0987, "step": 6230 }, { "grad_norm": 0.5897607207298279, "learning_rate": 9.995803623965389e-05, "loss": 0.0921, "step": 6240 }, { "grad_norm": 0.9197136163711548, "learning_rate": 9.995735622164641e-05, "loss": 0.098, "step": 6250 }, { "grad_norm": 0.6355026960372925, "learning_rate": 9.995667074037758e-05, "loss": 0.0762, "step": 6260 }, { "grad_norm": 0.6083581447601318, "learning_rate": 9.995597979592232e-05, "loss": 0.0808, "step": 6270 }, { "grad_norm": 0.5697099566459656, "learning_rate": 9.995528338835625e-05, "loss": 0.0871, "step": 6280 }, { "grad_norm": 0.539033055305481, "learning_rate": 9.995458151775547e-05, "loss": 0.0873, "step": 6290 }, { "grad_norm": 0.6104947924613953, "learning_rate": 9.995387418419677e-05, "loss": 0.0865, "step": 6300 }, { "grad_norm": 0.5016438364982605, "learning_rate": 9.99531613877575e-05, "loss": 0.0697, "step": 6310 }, { "grad_norm": 0.5295462608337402, "learning_rate": 9.995244312851559e-05, "loss": 0.086, "step": 6320 }, { "grad_norm": 0.605601966381073, "learning_rate": 9.995171940654961e-05, "loss": 0.0839, "step": 6330 }, { "grad_norm": 0.9326311945915222, "learning_rate": 9.995099022193871e-05, "loss": 0.0944, "step": 6340 }, { "grad_norm": 0.4562109708786011, "learning_rate": 9.995025557476261e-05, "loss": 0.0805, "step": 6350 }, { "grad_norm": 0.845052182674408, "learning_rate": 9.994951546510165e-05, "loss": 0.0861, "step": 6360 }, { "grad_norm": 0.9572100639343262, "learning_rate": 9.994876989303679e-05, "loss": 0.1002, "step": 6370 }, { "grad_norm": 0.7859735488891602, "learning_rate": 9.994801885864955e-05, "loss": 0.0895, "step": 6380 }, { "grad_norm": 0.7839677929878235, "learning_rate": 9.994726236202205e-05, "loss": 0.1029, "step": 6390 }, { "grad_norm": 0.563023030757904, "learning_rate": 9.994650040323704e-05, "loss": 0.0967, "step": 6400 }, { "grad_norm": 0.4636004567146301, "learning_rate": 9.994573298237784e-05, "loss": 0.0697, "step": 6410 }, { "grad_norm": 0.658502459526062, "learning_rate": 9.994496009952837e-05, "loss": 0.0915, "step": 6420 }, { "grad_norm": 0.47311386466026306, "learning_rate": 9.994418175477316e-05, "loss": 0.0848, "step": 6430 }, { "grad_norm": 0.5775735378265381, "learning_rate": 9.994339794819733e-05, "loss": 0.0897, "step": 6440 }, { "grad_norm": 0.5808118581771851, "learning_rate": 9.994260867988658e-05, "loss": 0.0874, "step": 6450 }, { "grad_norm": 0.5246264934539795, "learning_rate": 9.994181394992723e-05, "loss": 0.0801, "step": 6460 }, { "grad_norm": 0.6369976997375488, "learning_rate": 9.994101375840618e-05, "loss": 0.08, "step": 6470 }, { "grad_norm": 0.6759586334228516, "learning_rate": 9.994020810541098e-05, "loss": 0.0773, "step": 6480 }, { "grad_norm": 0.8901851177215576, "learning_rate": 9.99393969910297e-05, "loss": 0.0987, "step": 6490 }, { "grad_norm": 0.5735053420066833, "learning_rate": 9.993858041535104e-05, "loss": 0.0785, "step": 6500 }, { "grad_norm": 0.717368483543396, "learning_rate": 9.99377583784643e-05, "loss": 0.0783, "step": 6510 }, { "grad_norm": 0.761756420135498, "learning_rate": 9.993693088045939e-05, "loss": 0.0738, "step": 6520 }, { "grad_norm": 0.5912415981292725, "learning_rate": 9.99360979214268e-05, "loss": 0.0786, "step": 6530 }, { "grad_norm": 0.6221649646759033, "learning_rate": 9.99352595014576e-05, "loss": 0.1034, "step": 6540 }, { "grad_norm": 0.7175846695899963, "learning_rate": 9.993441562064354e-05, "loss": 0.0998, "step": 6550 }, { "grad_norm": 0.5381503105163574, "learning_rate": 9.993356627907685e-05, "loss": 0.095, "step": 6560 }, { "grad_norm": 0.5963038206100464, "learning_rate": 9.99327114768504e-05, "loss": 0.0937, "step": 6570 }, { "grad_norm": 0.5651427507400513, "learning_rate": 9.99318512140577e-05, "loss": 0.0767, "step": 6580 }, { "grad_norm": 0.5673907995223999, "learning_rate": 9.993098549079284e-05, "loss": 0.1087, "step": 6590 }, { "grad_norm": 0.7539456486701965, "learning_rate": 9.993011430715047e-05, "loss": 0.1075, "step": 6600 }, { "grad_norm": 0.7137300968170166, "learning_rate": 9.992923766322586e-05, "loss": 0.076, "step": 6610 }, { "grad_norm": 0.4801902174949646, "learning_rate": 9.99283555591149e-05, "loss": 0.0853, "step": 6620 }, { "grad_norm": 0.6104915738105774, "learning_rate": 9.992746799491404e-05, "loss": 0.0852, "step": 6630 }, { "grad_norm": 0.7295399904251099, "learning_rate": 9.992657497072033e-05, "loss": 0.0776, "step": 6640 }, { "grad_norm": 0.7991737127304077, "learning_rate": 9.992567648663147e-05, "loss": 0.1059, "step": 6650 }, { "grad_norm": 0.7366912364959717, "learning_rate": 9.992477254274568e-05, "loss": 0.0757, "step": 6660 }, { "grad_norm": 0.6850689649581909, "learning_rate": 9.992386313916183e-05, "loss": 0.0783, "step": 6670 }, { "grad_norm": 0.7021750211715698, "learning_rate": 9.992294827597934e-05, "loss": 0.0875, "step": 6680 }, { "grad_norm": 0.5327901840209961, "learning_rate": 9.992202795329831e-05, "loss": 0.0768, "step": 6690 }, { "grad_norm": 0.4543144702911377, "learning_rate": 9.992110217121936e-05, "loss": 0.0919, "step": 6700 }, { "grad_norm": 0.7270269393920898, "learning_rate": 9.992017092984372e-05, "loss": 0.0886, "step": 6710 }, { "grad_norm": 0.701560378074646, "learning_rate": 9.991923422927326e-05, "loss": 0.1014, "step": 6720 }, { "grad_norm": 0.5398039221763611, "learning_rate": 9.991829206961037e-05, "loss": 0.0928, "step": 6730 }, { "grad_norm": 0.6080025434494019, "learning_rate": 9.991734445095813e-05, "loss": 0.1102, "step": 6740 }, { "grad_norm": 0.5576053857803345, "learning_rate": 9.991639137342015e-05, "loss": 0.0833, "step": 6750 }, { "grad_norm": 0.687573254108429, "learning_rate": 9.991543283710064e-05, "loss": 0.0969, "step": 6760 }, { "grad_norm": 0.5963261127471924, "learning_rate": 9.991446884210445e-05, "loss": 0.0751, "step": 6770 }, { "grad_norm": 0.5866814851760864, "learning_rate": 9.9913499388537e-05, "loss": 0.0863, "step": 6780 }, { "grad_norm": 0.8024929165840149, "learning_rate": 9.99125244765043e-05, "loss": 0.0804, "step": 6790 }, { "grad_norm": 0.5599513053894043, "learning_rate": 9.991154410611296e-05, "loss": 0.0826, "step": 6800 }, { "grad_norm": 0.8723310828208923, "learning_rate": 9.99105582774702e-05, "loss": 0.0808, "step": 6810 }, { "grad_norm": 0.8503301739692688, "learning_rate": 9.990956699068384e-05, "loss": 0.0945, "step": 6820 }, { "grad_norm": 0.6770226359367371, "learning_rate": 9.990857024586224e-05, "loss": 0.0891, "step": 6830 }, { "grad_norm": 0.5987378358840942, "learning_rate": 9.990756804311446e-05, "loss": 0.0947, "step": 6840 }, { "grad_norm": 0.524945080280304, "learning_rate": 9.990656038255006e-05, "loss": 0.0814, "step": 6850 }, { "grad_norm": 0.5670360326766968, "learning_rate": 9.990554726427926e-05, "loss": 0.0913, "step": 6860 }, { "grad_norm": 0.7352789044380188, "learning_rate": 9.990452868841284e-05, "loss": 0.101, "step": 6870 }, { "grad_norm": 1.0883811712265015, "learning_rate": 9.99035046550622e-05, "loss": 0.1082, "step": 6880 }, { "grad_norm": 0.6105899214744568, "learning_rate": 9.99024751643393e-05, "loss": 0.094, "step": 6890 }, { "grad_norm": 0.7231579422950745, "learning_rate": 9.990144021635677e-05, "loss": 0.0801, "step": 6900 }, { "grad_norm": 0.6082424521446228, "learning_rate": 9.990039981122775e-05, "loss": 0.0816, "step": 6910 }, { "grad_norm": 0.7271661162376404, "learning_rate": 9.989935394906602e-05, "loss": 0.0823, "step": 6920 }, { "grad_norm": 0.5139390230178833, "learning_rate": 9.989830262998598e-05, "loss": 0.0955, "step": 6930 }, { "grad_norm": 0.5900577306747437, "learning_rate": 9.989724585410259e-05, "loss": 0.0705, "step": 6940 }, { "grad_norm": 0.5457778573036194, "learning_rate": 9.989618362153139e-05, "loss": 0.074, "step": 6950 }, { "grad_norm": 0.5816399455070496, "learning_rate": 9.989511593238859e-05, "loss": 0.0863, "step": 6960 }, { "grad_norm": 0.522474467754364, "learning_rate": 9.98940427867909e-05, "loss": 0.0966, "step": 6970 }, { "grad_norm": 1.093363642692566, "learning_rate": 9.989296418485573e-05, "loss": 0.0809, "step": 6980 }, { "grad_norm": 0.5803808569908142, "learning_rate": 9.989188012670101e-05, "loss": 0.0925, "step": 6990 }, { "grad_norm": 0.6360449194908142, "learning_rate": 9.989079061244528e-05, "loss": 0.1068, "step": 7000 }, { "grad_norm": 0.5718036890029907, "learning_rate": 9.988969564220769e-05, "loss": 0.1139, "step": 7010 }, { "grad_norm": 0.5471646189689636, "learning_rate": 9.988859521610801e-05, "loss": 0.0944, "step": 7020 }, { "grad_norm": 0.719774067401886, "learning_rate": 9.988748933426656e-05, "loss": 0.1127, "step": 7030 }, { "grad_norm": 0.5599322319030762, "learning_rate": 9.988637799680428e-05, "loss": 0.0942, "step": 7040 }, { "grad_norm": 0.6722057461738586, "learning_rate": 9.98852612038427e-05, "loss": 0.0863, "step": 7050 }, { "grad_norm": 0.9085827469825745, "learning_rate": 9.988413895550397e-05, "loss": 0.0865, "step": 7060 }, { "grad_norm": 0.6703898906707764, "learning_rate": 9.98830112519108e-05, "loss": 0.0813, "step": 7070 }, { "grad_norm": 0.6131898164749146, "learning_rate": 9.98818780931865e-05, "loss": 0.114, "step": 7080 }, { "grad_norm": 0.8132093548774719, "learning_rate": 9.988073947945502e-05, "loss": 0.105, "step": 7090 }, { "grad_norm": 0.7145311832427979, "learning_rate": 9.987959541084087e-05, "loss": 0.0942, "step": 7100 }, { "grad_norm": 0.5373666882514954, "learning_rate": 9.987844588746915e-05, "loss": 0.0926, "step": 7110 }, { "grad_norm": 0.703448474407196, "learning_rate": 9.987729090946558e-05, "loss": 0.0805, "step": 7120 }, { "grad_norm": 0.5771141648292542, "learning_rate": 9.987613047695647e-05, "loss": 0.0825, "step": 7130 }, { "grad_norm": 0.5777058601379395, "learning_rate": 9.987496459006871e-05, "loss": 0.098, "step": 7140 }, { "grad_norm": 0.5376915335655212, "learning_rate": 9.987379324892982e-05, "loss": 0.08, "step": 7150 }, { "grad_norm": 0.7465864419937134, "learning_rate": 9.987261645366788e-05, "loss": 0.0968, "step": 7160 }, { "grad_norm": 0.6211589574813843, "learning_rate": 9.987143420441158e-05, "loss": 0.0939, "step": 7170 }, { "grad_norm": 0.6650754809379578, "learning_rate": 9.987024650129022e-05, "loss": 0.0782, "step": 7180 }, { "grad_norm": 0.600862443447113, "learning_rate": 9.986905334443368e-05, "loss": 0.0908, "step": 7190 }, { "grad_norm": 0.7085986137390137, "learning_rate": 9.986785473397245e-05, "loss": 0.0739, "step": 7200 }, { "grad_norm": 0.5728276371955872, "learning_rate": 9.98666506700376e-05, "loss": 0.0919, "step": 7210 }, { "grad_norm": 0.6316154599189758, "learning_rate": 9.986544115276081e-05, "loss": 0.0792, "step": 7220 }, { "grad_norm": 0.7218353748321533, "learning_rate": 9.986422618227433e-05, "loss": 0.0816, "step": 7230 }, { "grad_norm": 0.8177704215049744, "learning_rate": 9.986300575871106e-05, "loss": 0.0841, "step": 7240 }, { "grad_norm": 0.6810714602470398, "learning_rate": 9.986177988220444e-05, "loss": 0.0893, "step": 7250 }, { "grad_norm": 0.5165241360664368, "learning_rate": 9.986054855288856e-05, "loss": 0.0867, "step": 7260 }, { "grad_norm": 0.6972900032997131, "learning_rate": 9.985931177089802e-05, "loss": 0.0875, "step": 7270 }, { "grad_norm": 0.5534729957580566, "learning_rate": 9.985806953636814e-05, "loss": 0.0832, "step": 7280 }, { "grad_norm": 0.7249010801315308, "learning_rate": 9.985682184943471e-05, "loss": 0.0882, "step": 7290 }, { "grad_norm": 0.5354682207107544, "learning_rate": 9.98555687102342e-05, "loss": 0.0881, "step": 7300 }, { "grad_norm": 1.0035754442214966, "learning_rate": 9.985431011890367e-05, "loss": 0.0973, "step": 7310 }, { "grad_norm": 0.606238067150116, "learning_rate": 9.985304607558075e-05, "loss": 0.0709, "step": 7320 }, { "grad_norm": 0.37214821577072144, "learning_rate": 9.985177658040364e-05, "loss": 0.0778, "step": 7330 }, { "grad_norm": 0.4664987325668335, "learning_rate": 9.985050163351119e-05, "loss": 0.0836, "step": 7340 }, { "grad_norm": 0.6462200880050659, "learning_rate": 9.984922123504286e-05, "loss": 0.0783, "step": 7350 }, { "grad_norm": 0.5300299525260925, "learning_rate": 9.984793538513862e-05, "loss": 0.0701, "step": 7360 }, { "grad_norm": 0.8233498334884644, "learning_rate": 9.984664408393912e-05, "loss": 0.0939, "step": 7370 }, { "grad_norm": 0.542137086391449, "learning_rate": 9.984534733158556e-05, "loss": 0.0789, "step": 7380 }, { "grad_norm": 0.7838888764381409, "learning_rate": 9.984404512821977e-05, "loss": 0.0869, "step": 7390 }, { "grad_norm": 0.5848798751831055, "learning_rate": 9.984273747398411e-05, "loss": 0.0875, "step": 7400 }, { "grad_norm": 0.5880289077758789, "learning_rate": 9.984142436902165e-05, "loss": 0.0941, "step": 7410 }, { "grad_norm": 0.6940523982048035, "learning_rate": 9.984010581347596e-05, "loss": 0.1126, "step": 7420 }, { "grad_norm": 0.6768618822097778, "learning_rate": 9.983878180749121e-05, "loss": 0.1066, "step": 7430 }, { "grad_norm": 0.5192753672599792, "learning_rate": 9.983745235121222e-05, "loss": 0.0962, "step": 7440 }, { "grad_norm": 0.5217660069465637, "learning_rate": 9.983611744478438e-05, "loss": 0.0836, "step": 7450 }, { "grad_norm": 0.482240229845047, "learning_rate": 9.983477708835365e-05, "loss": 0.084, "step": 7460 }, { "grad_norm": 0.6920139193534851, "learning_rate": 9.983343128206664e-05, "loss": 0.1043, "step": 7470 }, { "grad_norm": 0.6599873304367065, "learning_rate": 9.983208002607049e-05, "loss": 0.0905, "step": 7480 }, { "grad_norm": 0.6043674349784851, "learning_rate": 9.9830723320513e-05, "loss": 0.0796, "step": 7490 }, { "grad_norm": 0.5614514350891113, "learning_rate": 9.982936116554254e-05, "loss": 0.0733, "step": 7500 }, { "grad_norm": 0.6991101503372192, "learning_rate": 9.982799356130803e-05, "loss": 0.0711, "step": 7510 }, { "grad_norm": 0.4636584520339966, "learning_rate": 9.982662050795908e-05, "loss": 0.076, "step": 7520 }, { "grad_norm": 0.45654284954071045, "learning_rate": 9.982524200564583e-05, "loss": 0.0744, "step": 7530 }, { "grad_norm": 0.5641209483146667, "learning_rate": 9.982385805451901e-05, "loss": 0.0865, "step": 7540 }, { "grad_norm": 0.5123664736747742, "learning_rate": 9.982246865472998e-05, "loss": 0.0939, "step": 7550 }, { "grad_norm": 0.4878160357475281, "learning_rate": 9.982107380643069e-05, "loss": 0.0809, "step": 7560 }, { "grad_norm": 0.5616593956947327, "learning_rate": 9.981967350977368e-05, "loss": 0.0984, "step": 7570 }, { "grad_norm": 0.6464055180549622, "learning_rate": 9.981826776491208e-05, "loss": 0.0882, "step": 7580 }, { "grad_norm": 0.4690965712070465, "learning_rate": 9.98168565719996e-05, "loss": 0.0854, "step": 7590 }, { "grad_norm": 0.5813785791397095, "learning_rate": 9.98154399311906e-05, "loss": 0.0914, "step": 7600 }, { "grad_norm": 0.6235758662223816, "learning_rate": 9.981401784263997e-05, "loss": 0.0906, "step": 7610 }, { "grad_norm": 0.6333611607551575, "learning_rate": 9.981259030650326e-05, "loss": 0.0846, "step": 7620 }, { "grad_norm": 0.6168288588523865, "learning_rate": 9.981115732293655e-05, "loss": 0.0741, "step": 7630 }, { "grad_norm": 0.569696307182312, "learning_rate": 9.980971889209659e-05, "loss": 0.0738, "step": 7640 }, { "grad_norm": 0.6999420523643494, "learning_rate": 9.980827501414064e-05, "loss": 0.0652, "step": 7650 }, { "grad_norm": 0.8700419068336487, "learning_rate": 9.980682568922663e-05, "loss": 0.0967, "step": 7660 }, { "grad_norm": 0.6429250836372375, "learning_rate": 9.980537091751304e-05, "loss": 0.0891, "step": 7670 }, { "grad_norm": 0.5313330888748169, "learning_rate": 9.980391069915897e-05, "loss": 0.1088, "step": 7680 }, { "grad_norm": 0.6169960498809814, "learning_rate": 9.98024450343241e-05, "loss": 0.0768, "step": 7690 }, { "grad_norm": 0.5175290107727051, "learning_rate": 9.980097392316872e-05, "loss": 0.0641, "step": 7700 }, { "grad_norm": 0.5751611590385437, "learning_rate": 9.97994973658537e-05, "loss": 0.0885, "step": 7710 }, { "grad_norm": 0.5692751407623291, "learning_rate": 9.979801536254054e-05, "loss": 0.0887, "step": 7720 }, { "grad_norm": 0.6373639702796936, "learning_rate": 9.979652791339127e-05, "loss": 0.0975, "step": 7730 }, { "grad_norm": 0.9422372579574585, "learning_rate": 9.97950350185686e-05, "loss": 0.0904, "step": 7740 }, { "grad_norm": 0.386074960231781, "learning_rate": 9.979353667823574e-05, "loss": 0.0705, "step": 7750 }, { "grad_norm": 0.5221442580223083, "learning_rate": 9.979203289255658e-05, "loss": 0.0827, "step": 7760 }, { "grad_norm": 0.6128689050674438, "learning_rate": 9.979052366169557e-05, "loss": 0.0788, "step": 7770 }, { "grad_norm": 0.5114107728004456, "learning_rate": 9.978900898581775e-05, "loss": 0.0869, "step": 7780 }, { "grad_norm": 0.6729180216789246, "learning_rate": 9.978748886508875e-05, "loss": 0.0809, "step": 7790 }, { "grad_norm": 0.702940046787262, "learning_rate": 9.978596329967484e-05, "loss": 0.0855, "step": 7800 }, { "grad_norm": 0.5275229215621948, "learning_rate": 9.978443228974284e-05, "loss": 0.0809, "step": 7810 }, { "grad_norm": 0.5798386931419373, "learning_rate": 9.978289583546015e-05, "loss": 0.079, "step": 7820 }, { "grad_norm": 0.5466773509979248, "learning_rate": 9.978135393699484e-05, "loss": 0.0829, "step": 7830 }, { "grad_norm": 0.9479439854621887, "learning_rate": 9.977980659451548e-05, "loss": 0.0705, "step": 7840 }, { "grad_norm": 0.8529976010322571, "learning_rate": 9.977825380819135e-05, "loss": 0.0777, "step": 7850 }, { "grad_norm": 0.7422564625740051, "learning_rate": 9.97766955781922e-05, "loss": 0.0861, "step": 7860 }, { "grad_norm": 0.5492268204689026, "learning_rate": 9.977513190468848e-05, "loss": 0.0833, "step": 7870 }, { "grad_norm": 0.5253100991249084, "learning_rate": 9.977356278785116e-05, "loss": 0.0905, "step": 7880 }, { "grad_norm": 0.5318015217781067, "learning_rate": 9.977198822785184e-05, "loss": 0.0768, "step": 7890 }, { "grad_norm": 0.7309349179267883, "learning_rate": 9.977040822486273e-05, "loss": 0.0962, "step": 7900 }, { "grad_norm": 0.40528103709220886, "learning_rate": 9.97688227790566e-05, "loss": 0.0845, "step": 7910 }, { "grad_norm": 0.5530934929847717, "learning_rate": 9.976723189060684e-05, "loss": 0.1085, "step": 7920 }, { "grad_norm": 0.6758176684379578, "learning_rate": 9.976563555968742e-05, "loss": 0.0833, "step": 7930 }, { "grad_norm": 0.6564778685569763, "learning_rate": 9.976403378647292e-05, "loss": 0.0946, "step": 7940 }, { "grad_norm": 0.5073357224464417, "learning_rate": 9.97624265711385e-05, "loss": 0.0815, "step": 7950 }, { "grad_norm": 0.6293243765830994, "learning_rate": 9.976081391385993e-05, "loss": 0.0864, "step": 7960 }, { "grad_norm": 0.5289585590362549, "learning_rate": 9.975919581481356e-05, "loss": 0.0807, "step": 7970 }, { "grad_norm": 0.5023589134216309, "learning_rate": 9.975757227417634e-05, "loss": 0.0898, "step": 7980 }, { "grad_norm": 0.5307055115699768, "learning_rate": 9.975594329212586e-05, "loss": 0.0855, "step": 7990 }, { "grad_norm": 0.49068760871887207, "learning_rate": 9.97543088688402e-05, "loss": 0.088, "step": 8000 }, { "grad_norm": 0.5661093592643738, "learning_rate": 9.975266900449814e-05, "loss": 0.0747, "step": 8010 }, { "grad_norm": 0.5814194083213806, "learning_rate": 9.975102369927898e-05, "loss": 0.0735, "step": 8020 }, { "grad_norm": 0.37768539786338806, "learning_rate": 9.974937295336269e-05, "loss": 0.0667, "step": 8030 }, { "grad_norm": 0.6323707103729248, "learning_rate": 9.974771676692975e-05, "loss": 0.0835, "step": 8040 }, { "grad_norm": 0.5180512070655823, "learning_rate": 9.974605514016131e-05, "loss": 0.0752, "step": 8050 }, { "grad_norm": 0.45860758423805237, "learning_rate": 9.974438807323907e-05, "loss": 0.0808, "step": 8060 }, { "grad_norm": 0.5177556872367859, "learning_rate": 9.974271556634535e-05, "loss": 0.0819, "step": 8070 }, { "grad_norm": 0.7754493951797485, "learning_rate": 9.974103761966302e-05, "loss": 0.0984, "step": 8080 }, { "grad_norm": 0.41412046551704407, "learning_rate": 9.973935423337563e-05, "loss": 0.0811, "step": 8090 }, { "grad_norm": 1.342024803161621, "learning_rate": 9.973766540766722e-05, "loss": 0.095, "step": 8100 }, { "grad_norm": 0.7508848309516907, "learning_rate": 9.97359711427225e-05, "loss": 0.0944, "step": 8110 }, { "grad_norm": 0.49602219462394714, "learning_rate": 9.973427143872677e-05, "loss": 0.0942, "step": 8120 }, { "grad_norm": 0.9728596210479736, "learning_rate": 9.973256629586589e-05, "loss": 0.1137, "step": 8130 }, { "grad_norm": 0.7031168341636658, "learning_rate": 9.973085571432632e-05, "loss": 0.0881, "step": 8140 }, { "grad_norm": 0.6648826003074646, "learning_rate": 9.972913969429513e-05, "loss": 0.0765, "step": 8150 }, { "grad_norm": 0.6257417798042297, "learning_rate": 9.972741823596e-05, "loss": 0.0968, "step": 8160 }, { "grad_norm": 0.6973499059677124, "learning_rate": 9.972569133950917e-05, "loss": 0.084, "step": 8170 }, { "grad_norm": 0.48094266653060913, "learning_rate": 9.972395900513151e-05, "loss": 0.0767, "step": 8180 }, { "grad_norm": 0.7128446102142334, "learning_rate": 9.972222123301645e-05, "loss": 0.0805, "step": 8190 }, { "grad_norm": 0.5772319436073303, "learning_rate": 9.972047802335403e-05, "loss": 0.0848, "step": 8200 }, { "grad_norm": 0.3920798599720001, "learning_rate": 9.971872937633488e-05, "loss": 0.0694, "step": 8210 }, { "grad_norm": 0.8871675133705139, "learning_rate": 9.971697529215024e-05, "loss": 0.0883, "step": 8220 }, { "grad_norm": 0.6105732917785645, "learning_rate": 9.971521577099192e-05, "loss": 0.0902, "step": 8230 }, { "grad_norm": 0.4677106738090515, "learning_rate": 9.971345081305236e-05, "loss": 0.0811, "step": 8240 }, { "grad_norm": 0.4280383884906769, "learning_rate": 9.971168041852456e-05, "loss": 0.0783, "step": 8250 }, { "grad_norm": 0.6190682053565979, "learning_rate": 9.970990458760215e-05, "loss": 0.074, "step": 8260 }, { "grad_norm": 0.5114795565605164, "learning_rate": 9.970812332047929e-05, "loss": 0.0747, "step": 8270 }, { "grad_norm": 0.4845590889453888, "learning_rate": 9.97063366173508e-05, "loss": 0.0774, "step": 8280 }, { "grad_norm": 0.5193896889686584, "learning_rate": 9.970454447841207e-05, "loss": 0.0737, "step": 8290 }, { "grad_norm": 0.5174989104270935, "learning_rate": 9.970274690385909e-05, "loss": 0.0794, "step": 8300 }, { "grad_norm": 0.7027796506881714, "learning_rate": 9.970094389388844e-05, "loss": 0.0851, "step": 8310 }, { "grad_norm": 0.7299913763999939, "learning_rate": 9.969913544869728e-05, "loss": 0.0684, "step": 8320 }, { "grad_norm": 0.7921730279922485, "learning_rate": 9.96973215684834e-05, "loss": 0.0906, "step": 8330 }, { "grad_norm": 0.7204760909080505, "learning_rate": 9.969550225344513e-05, "loss": 0.0826, "step": 8340 }, { "grad_norm": 0.869807243347168, "learning_rate": 9.969367750378147e-05, "loss": 0.0874, "step": 8350 }, { "grad_norm": 0.5952873826026917, "learning_rate": 9.969184731969194e-05, "loss": 0.0941, "step": 8360 }, { "grad_norm": 0.7064881324768066, "learning_rate": 9.96900117013767e-05, "loss": 0.1059, "step": 8370 }, { "grad_norm": 0.7052238583564758, "learning_rate": 9.96881706490365e-05, "loss": 0.0948, "step": 8380 }, { "grad_norm": 0.5364582538604736, "learning_rate": 9.968632416287265e-05, "loss": 0.0889, "step": 8390 }, { "grad_norm": 0.9186429381370544, "learning_rate": 9.96844722430871e-05, "loss": 0.0924, "step": 8400 }, { "grad_norm": 0.6793975234031677, "learning_rate": 9.968261488988235e-05, "loss": 0.0861, "step": 8410 }, { "grad_norm": 0.6501414179801941, "learning_rate": 9.968075210346155e-05, "loss": 0.0841, "step": 8420 }, { "grad_norm": 0.5936107635498047, "learning_rate": 9.967888388402839e-05, "loss": 0.0896, "step": 8430 }, { "grad_norm": 0.5107994079589844, "learning_rate": 9.967701023178717e-05, "loss": 0.0749, "step": 8440 }, { "grad_norm": 0.6124663352966309, "learning_rate": 9.967513114694282e-05, "loss": 0.0757, "step": 8450 }, { "grad_norm": 0.6391571164131165, "learning_rate": 9.967324662970079e-05, "loss": 0.0873, "step": 8460 }, { "grad_norm": 0.5402523875236511, "learning_rate": 9.96713566802672e-05, "loss": 0.076, "step": 8470 }, { "grad_norm": 0.4711114764213562, "learning_rate": 9.966946129884873e-05, "loss": 0.0854, "step": 8480 }, { "grad_norm": 0.7220969796180725, "learning_rate": 9.966756048565265e-05, "loss": 0.0865, "step": 8490 }, { "grad_norm": 0.6544954180717468, "learning_rate": 9.966565424088681e-05, "loss": 0.0937, "step": 8500 }, { "grad_norm": 0.6232096552848816, "learning_rate": 9.96637425647597e-05, "loss": 0.0728, "step": 8510 }, { "grad_norm": 0.6052765250205994, "learning_rate": 9.966182545748038e-05, "loss": 0.0727, "step": 8520 }, { "grad_norm": 0.503201961517334, "learning_rate": 9.96599029192585e-05, "loss": 0.0837, "step": 8530 }, { "grad_norm": 0.5877857208251953, "learning_rate": 9.965797495030428e-05, "loss": 0.0753, "step": 8540 }, { "grad_norm": 0.4832262098789215, "learning_rate": 9.96560415508286e-05, "loss": 0.0772, "step": 8550 }, { "grad_norm": 0.42622971534729004, "learning_rate": 9.965410272104286e-05, "loss": 0.073, "step": 8560 }, { "grad_norm": 0.5719324350357056, "learning_rate": 9.96521584611591e-05, "loss": 0.1192, "step": 8570 }, { "grad_norm": 0.5796010494232178, "learning_rate": 9.965020877138994e-05, "loss": 0.0785, "step": 8580 }, { "grad_norm": 0.4964612126350403, "learning_rate": 9.964825365194861e-05, "loss": 0.0987, "step": 8590 }, { "grad_norm": 0.5007039308547974, "learning_rate": 9.96462931030489e-05, "loss": 0.0732, "step": 8600 }, { "grad_norm": 0.574251651763916, "learning_rate": 9.96443271249052e-05, "loss": 0.0838, "step": 8610 }, { "grad_norm": 0.5435082912445068, "learning_rate": 9.964235571773255e-05, "loss": 0.0786, "step": 8620 }, { "grad_norm": 0.4938359558582306, "learning_rate": 9.96403788817465e-05, "loss": 0.0726, "step": 8630 }, { "grad_norm": 0.7026623487472534, "learning_rate": 9.963839661716325e-05, "loss": 0.0882, "step": 8640 }, { "grad_norm": 0.5690603256225586, "learning_rate": 9.963640892419958e-05, "loss": 0.0792, "step": 8650 }, { "grad_norm": 0.6216589212417603, "learning_rate": 9.963441580307286e-05, "loss": 0.0907, "step": 8660 }, { "grad_norm": 0.43652382493019104, "learning_rate": 9.963241725400104e-05, "loss": 0.0912, "step": 8670 }, { "grad_norm": 0.43851688504219055, "learning_rate": 9.963041327720271e-05, "loss": 0.0633, "step": 8680 }, { "grad_norm": 0.6204845309257507, "learning_rate": 9.962840387289697e-05, "loss": 0.0785, "step": 8690 }, { "grad_norm": 0.44881460070610046, "learning_rate": 9.962638904130363e-05, "loss": 0.0823, "step": 8700 }, { "grad_norm": 0.4748745262622833, "learning_rate": 9.962436878264298e-05, "loss": 0.0819, "step": 8710 }, { "grad_norm": 0.6746164560317993, "learning_rate": 9.962234309713598e-05, "loss": 0.0971, "step": 8720 }, { "grad_norm": 0.64650559425354, "learning_rate": 9.962031198500414e-05, "loss": 0.0746, "step": 8730 }, { "grad_norm": 0.5387746691703796, "learning_rate": 9.961827544646958e-05, "loss": 0.0713, "step": 8740 }, { "grad_norm": 0.6068204045295715, "learning_rate": 9.961623348175501e-05, "loss": 0.0968, "step": 8750 }, { "grad_norm": 0.6199666261672974, "learning_rate": 9.961418609108377e-05, "loss": 0.0783, "step": 8760 }, { "grad_norm": 0.4947487413883209, "learning_rate": 9.961213327467971e-05, "loss": 0.0811, "step": 8770 }, { "grad_norm": 0.5738400816917419, "learning_rate": 9.961007503276736e-05, "loss": 0.0861, "step": 8780 }, { "grad_norm": 1.0082316398620605, "learning_rate": 9.960801136557179e-05, "loss": 0.0822, "step": 8790 }, { "grad_norm": 0.5992489457130432, "learning_rate": 9.960594227331866e-05, "loss": 0.0724, "step": 8800 }, { "grad_norm": 0.5043293237686157, "learning_rate": 9.960386775623429e-05, "loss": 0.0869, "step": 8810 }, { "grad_norm": 0.5649120807647705, "learning_rate": 9.96017878145455e-05, "loss": 0.0815, "step": 8820 }, { "grad_norm": 0.6691704392433167, "learning_rate": 9.959970244847977e-05, "loss": 0.074, "step": 8830 }, { "grad_norm": 0.7095054984092712, "learning_rate": 9.959761165826518e-05, "loss": 0.0709, "step": 8840 }, { "grad_norm": 0.5080227255821228, "learning_rate": 9.959551544413033e-05, "loss": 0.076, "step": 8850 }, { "grad_norm": 0.4972730278968811, "learning_rate": 9.959341380630448e-05, "loss": 0.0688, "step": 8860 }, { "grad_norm": 0.7088400721549988, "learning_rate": 9.959130674501746e-05, "loss": 0.0715, "step": 8870 }, { "grad_norm": 0.5337397456169128, "learning_rate": 9.958919426049968e-05, "loss": 0.0772, "step": 8880 }, { "grad_norm": 0.528353750705719, "learning_rate": 9.958707635298219e-05, "loss": 0.0735, "step": 8890 }, { "grad_norm": 0.7903023362159729, "learning_rate": 9.958495302269657e-05, "loss": 0.0837, "step": 8900 }, { "grad_norm": 0.6325145363807678, "learning_rate": 9.958282426987503e-05, "loss": 0.0832, "step": 8910 }, { "grad_norm": 0.765367329120636, "learning_rate": 9.95806900947504e-05, "loss": 0.0871, "step": 8920 }, { "grad_norm": 0.6772627830505371, "learning_rate": 9.957855049755604e-05, "loss": 0.0664, "step": 8930 }, { "grad_norm": 0.5488401055335999, "learning_rate": 9.957640547852593e-05, "loss": 0.0814, "step": 8940 }, { "grad_norm": 0.4480608105659485, "learning_rate": 9.957425503789466e-05, "loss": 0.0715, "step": 8950 }, { "grad_norm": 0.5295181274414062, "learning_rate": 9.957209917589738e-05, "loss": 0.0654, "step": 8960 }, { "grad_norm": 0.46134236454963684, "learning_rate": 9.956993789276987e-05, "loss": 0.0696, "step": 8970 }, { "grad_norm": 0.5835365056991577, "learning_rate": 9.956777118874847e-05, "loss": 0.0802, "step": 8980 }, { "grad_norm": 0.980371356010437, "learning_rate": 9.956559906407016e-05, "loss": 0.0976, "step": 8990 }, { "grad_norm": 0.5971601605415344, "learning_rate": 9.956342151897245e-05, "loss": 0.0846, "step": 9000 }, { "grad_norm": 0.9312708377838135, "learning_rate": 9.956123855369346e-05, "loss": 0.0753, "step": 9010 }, { "grad_norm": 0.9828181862831116, "learning_rate": 9.955905016847196e-05, "loss": 0.0812, "step": 9020 }, { "grad_norm": 0.6845024228096008, "learning_rate": 9.955685636354723e-05, "loss": 0.0915, "step": 9030 }, { "grad_norm": 0.6630733013153076, "learning_rate": 9.95546571391592e-05, "loss": 0.0894, "step": 9040 }, { "grad_norm": 0.6680256724357605, "learning_rate": 9.955245249554837e-05, "loss": 0.0639, "step": 9050 }, { "grad_norm": 0.7093633413314819, "learning_rate": 9.955024243295582e-05, "loss": 0.1058, "step": 9060 }, { "grad_norm": 0.5042134523391724, "learning_rate": 9.954802695162328e-05, "loss": 0.0776, "step": 9070 }, { "grad_norm": 0.598629355430603, "learning_rate": 9.954580605179302e-05, "loss": 0.0823, "step": 9080 }, { "grad_norm": 0.5118315815925598, "learning_rate": 9.954357973370788e-05, "loss": 0.0841, "step": 9090 }, { "grad_norm": 0.4762974679470062, "learning_rate": 9.954134799761135e-05, "loss": 0.0951, "step": 9100 }, { "grad_norm": 0.7583910226821899, "learning_rate": 9.953911084374748e-05, "loss": 0.0809, "step": 9110 }, { "grad_norm": 0.5328139066696167, "learning_rate": 9.953686827236093e-05, "loss": 0.0923, "step": 9120 }, { "grad_norm": 0.5710254907608032, "learning_rate": 9.953462028369695e-05, "loss": 0.0813, "step": 9130 }, { "grad_norm": 0.6575110554695129, "learning_rate": 9.953236687800136e-05, "loss": 0.0714, "step": 9140 }, { "grad_norm": 0.5444419384002686, "learning_rate": 9.95301080555206e-05, "loss": 0.0833, "step": 9150 }, { "grad_norm": 0.5787918567657471, "learning_rate": 9.952784381650171e-05, "loss": 0.0726, "step": 9160 }, { "grad_norm": 0.628159761428833, "learning_rate": 9.952557416119226e-05, "loss": 0.0774, "step": 9170 }, { "grad_norm": 0.4763994812965393, "learning_rate": 9.95232990898405e-05, "loss": 0.0715, "step": 9180 }, { "grad_norm": 0.5038716793060303, "learning_rate": 9.95210186026952e-05, "loss": 0.0613, "step": 9190 }, { "grad_norm": 0.5693509578704834, "learning_rate": 9.951873270000576e-05, "loss": 0.076, "step": 9200 }, { "grad_norm": 0.49097299575805664, "learning_rate": 9.951644138202216e-05, "loss": 0.093, "step": 9210 }, { "grad_norm": 0.6195060610771179, "learning_rate": 9.951414464899498e-05, "loss": 0.0874, "step": 9220 }, { "grad_norm": 0.6142590045928955, "learning_rate": 9.951184250117538e-05, "loss": 0.0895, "step": 9230 }, { "grad_norm": 0.5826259255409241, "learning_rate": 9.950953493881513e-05, "loss": 0.0855, "step": 9240 }, { "grad_norm": 0.53635174036026, "learning_rate": 9.950722196216658e-05, "loss": 0.0696, "step": 9250 }, { "grad_norm": 0.5262838006019592, "learning_rate": 9.950490357148265e-05, "loss": 0.0726, "step": 9260 }, { "grad_norm": 0.45539432764053345, "learning_rate": 9.950257976701692e-05, "loss": 0.0807, "step": 9270 }, { "grad_norm": 0.5284162759780884, "learning_rate": 9.950025054902348e-05, "loss": 0.0798, "step": 9280 }, { "grad_norm": 0.8632228374481201, "learning_rate": 9.949791591775706e-05, "loss": 0.0761, "step": 9290 }, { "grad_norm": 0.544039249420166, "learning_rate": 9.949557587347298e-05, "loss": 0.0755, "step": 9300 }, { "grad_norm": 0.4411362111568451, "learning_rate": 9.949323041642713e-05, "loss": 0.0694, "step": 9310 }, { "grad_norm": 0.5931519865989685, "learning_rate": 9.949087954687602e-05, "loss": 0.0949, "step": 9320 }, { "grad_norm": 0.6235281229019165, "learning_rate": 9.948852326507672e-05, "loss": 0.0851, "step": 9330 }, { "grad_norm": 0.44561249017715454, "learning_rate": 9.948616157128694e-05, "loss": 0.0805, "step": 9340 }, { "grad_norm": 0.3663666546344757, "learning_rate": 9.948379446576493e-05, "loss": 0.0643, "step": 9350 }, { "grad_norm": 0.4393094778060913, "learning_rate": 9.948142194876952e-05, "loss": 0.0781, "step": 9360 }, { "grad_norm": 0.5439987778663635, "learning_rate": 9.947904402056024e-05, "loss": 0.0839, "step": 9370 }, { "grad_norm": 0.38433268666267395, "learning_rate": 9.947666068139708e-05, "loss": 0.0693, "step": 9380 }, { "grad_norm": 0.5645131468772888, "learning_rate": 9.947427193154071e-05, "loss": 0.0683, "step": 9390 }, { "grad_norm": 0.76106858253479, "learning_rate": 9.947187777125233e-05, "loss": 0.0872, "step": 9400 }, { "grad_norm": 0.4901990592479706, "learning_rate": 9.946947820079377e-05, "loss": 0.084, "step": 9410 }, { "grad_norm": 0.8077124357223511, "learning_rate": 9.946707322042747e-05, "loss": 0.0788, "step": 9420 }, { "grad_norm": 0.5286504030227661, "learning_rate": 9.94646628304164e-05, "loss": 0.0758, "step": 9430 }, { "grad_norm": 0.7377225756645203, "learning_rate": 9.946224703102418e-05, "loss": 0.0934, "step": 9440 }, { "grad_norm": 0.9042193293571472, "learning_rate": 9.945982582251498e-05, "loss": 0.0827, "step": 9450 }, { "grad_norm": 0.8793390393257141, "learning_rate": 9.94573992051536e-05, "loss": 0.0821, "step": 9460 }, { "grad_norm": 0.5569691061973572, "learning_rate": 9.94549671792054e-05, "loss": 0.1091, "step": 9470 }, { "grad_norm": 0.6458882689476013, "learning_rate": 9.945252974493635e-05, "loss": 0.0808, "step": 9480 }, { "grad_norm": 0.5983973145484924, "learning_rate": 9.9450086902613e-05, "loss": 0.0832, "step": 9490 }, { "grad_norm": 0.6902620792388916, "learning_rate": 9.944763865250248e-05, "loss": 0.071, "step": 9500 }, { "grad_norm": 0.9537035822868347, "learning_rate": 9.944518499487254e-05, "loss": 0.067, "step": 9510 }, { "grad_norm": 0.6053117513656616, "learning_rate": 9.944272592999151e-05, "loss": 0.0806, "step": 9520 }, { "grad_norm": 0.7298056483268738, "learning_rate": 9.94402614581283e-05, "loss": 0.0865, "step": 9530 }, { "grad_norm": 0.5816913843154907, "learning_rate": 9.943779157955244e-05, "loss": 0.0791, "step": 9540 }, { "grad_norm": 0.45229604840278625, "learning_rate": 9.943531629453403e-05, "loss": 0.0722, "step": 9550 }, { "grad_norm": 0.5710111260414124, "learning_rate": 9.943283560334375e-05, "loss": 0.0785, "step": 9560 }, { "grad_norm": 0.5497694611549377, "learning_rate": 9.943034950625288e-05, "loss": 0.0931, "step": 9570 }, { "grad_norm": 1.5357106924057007, "learning_rate": 9.942785800353332e-05, "loss": 0.086, "step": 9580 }, { "grad_norm": 0.5444272756576538, "learning_rate": 9.942536109545751e-05, "loss": 0.0876, "step": 9590 }, { "grad_norm": 0.597730815410614, "learning_rate": 9.942285878229853e-05, "loss": 0.0688, "step": 9600 }, { "grad_norm": 0.5809599757194519, "learning_rate": 9.942035106433001e-05, "loss": 0.0868, "step": 9610 }, { "grad_norm": 0.6411694288253784, "learning_rate": 9.94178379418262e-05, "loss": 0.0881, "step": 9620 }, { "grad_norm": 0.4122018814086914, "learning_rate": 9.941531941506194e-05, "loss": 0.0802, "step": 9630 }, { "grad_norm": 0.5429744720458984, "learning_rate": 9.941279548431263e-05, "loss": 0.0724, "step": 9640 }, { "grad_norm": 0.5875747799873352, "learning_rate": 9.941026614985431e-05, "loss": 0.0756, "step": 9650 }, { "grad_norm": 0.38231196999549866, "learning_rate": 9.940773141196357e-05, "loss": 0.0702, "step": 9660 }, { "grad_norm": 0.6498134136199951, "learning_rate": 9.94051912709176e-05, "loss": 0.0805, "step": 9670 }, { "grad_norm": 0.591346263885498, "learning_rate": 9.940264572699421e-05, "loss": 0.0773, "step": 9680 }, { "grad_norm": 0.677375853061676, "learning_rate": 9.940009478047174e-05, "loss": 0.0648, "step": 9690 }, { "grad_norm": 0.6143398880958557, "learning_rate": 9.939753843162918e-05, "loss": 0.0672, "step": 9700 }, { "grad_norm": 0.6572757959365845, "learning_rate": 9.939497668074609e-05, "loss": 0.0768, "step": 9710 }, { "grad_norm": 0.7134842872619629, "learning_rate": 9.93924095281026e-05, "loss": 0.0895, "step": 9720 }, { "grad_norm": 0.6101763844490051, "learning_rate": 9.938983697397948e-05, "loss": 0.0838, "step": 9730 }, { "grad_norm": 0.5120763778686523, "learning_rate": 9.938725901865805e-05, "loss": 0.0734, "step": 9740 }, { "grad_norm": 0.5229902267456055, "learning_rate": 9.93846756624202e-05, "loss": 0.0831, "step": 9750 }, { "grad_norm": 0.58202064037323, "learning_rate": 9.938208690554849e-05, "loss": 0.0719, "step": 9760 }, { "grad_norm": 0.45823922753334045, "learning_rate": 9.9379492748326e-05, "loss": 0.0712, "step": 9770 }, { "grad_norm": 0.5908554196357727, "learning_rate": 9.937689319103641e-05, "loss": 0.0722, "step": 9780 }, { "grad_norm": 0.6159128546714783, "learning_rate": 9.937428823396404e-05, "loss": 0.0772, "step": 9790 }, { "grad_norm": 0.7382659912109375, "learning_rate": 9.937167787739372e-05, "loss": 0.0788, "step": 9800 }, { "grad_norm": 0.5711002349853516, "learning_rate": 9.936906212161095e-05, "loss": 0.0846, "step": 9810 }, { "grad_norm": 0.47720468044281006, "learning_rate": 9.936644096690176e-05, "loss": 0.0967, "step": 9820 }, { "grad_norm": 0.4519250988960266, "learning_rate": 9.936381441355282e-05, "loss": 0.0803, "step": 9830 }, { "grad_norm": 0.511817216873169, "learning_rate": 9.936118246185136e-05, "loss": 0.0881, "step": 9840 }, { "grad_norm": 0.4965539276599884, "learning_rate": 9.935854511208518e-05, "loss": 0.0861, "step": 9850 }, { "grad_norm": 0.4033333659172058, "learning_rate": 9.935590236454272e-05, "loss": 0.0711, "step": 9860 }, { "grad_norm": 0.5841665863990784, "learning_rate": 9.935325421951298e-05, "loss": 0.0724, "step": 9870 }, { "grad_norm": 0.6452985405921936, "learning_rate": 9.935060067728557e-05, "loss": 0.0689, "step": 9880 }, { "grad_norm": 0.4314008951187134, "learning_rate": 9.934794173815067e-05, "loss": 0.0592, "step": 9890 }, { "grad_norm": 0.7929680347442627, "learning_rate": 9.934527740239906e-05, "loss": 0.0803, "step": 9900 }, { "grad_norm": 0.46681472659111023, "learning_rate": 9.934260767032209e-05, "loss": 0.0856, "step": 9910 }, { "grad_norm": 0.6425908803939819, "learning_rate": 9.933993254221172e-05, "loss": 0.0845, "step": 9920 }, { "grad_norm": 0.6164923906326294, "learning_rate": 9.933725201836053e-05, "loss": 0.0772, "step": 9930 }, { "grad_norm": 0.5286171436309814, "learning_rate": 9.933456609906162e-05, "loss": 0.0736, "step": 9940 }, { "grad_norm": 0.5377983450889587, "learning_rate": 9.933187478460875e-05, "loss": 0.0783, "step": 9950 }, { "grad_norm": 0.7759293913841248, "learning_rate": 9.93291780752962e-05, "loss": 0.0834, "step": 9960 }, { "grad_norm": 0.5159414410591125, "learning_rate": 9.932647597141893e-05, "loss": 0.1107, "step": 9970 }, { "grad_norm": 0.6012297868728638, "learning_rate": 9.932376847327239e-05, "loss": 0.0874, "step": 9980 }, { "grad_norm": 0.5643316507339478, "learning_rate": 9.932105558115268e-05, "loss": 0.0816, "step": 9990 }, { "grad_norm": 0.7654195427894592, "learning_rate": 9.931833729535651e-05, "loss": 0.0933, "step": 10000 }, { "grad_norm": 0.6427374482154846, "learning_rate": 9.931561361618111e-05, "loss": 0.0837, "step": 10010 }, { "grad_norm": 0.5913988351821899, "learning_rate": 9.931288454392435e-05, "loss": 0.0918, "step": 10020 }, { "grad_norm": 0.5037696957588196, "learning_rate": 9.931015007888467e-05, "loss": 0.0706, "step": 10030 }, { "grad_norm": 0.6654314398765564, "learning_rate": 9.930741022136112e-05, "loss": 0.0784, "step": 10040 }, { "grad_norm": 0.6026290059089661, "learning_rate": 9.930466497165333e-05, "loss": 0.0733, "step": 10050 }, { "grad_norm": 0.5643529891967773, "learning_rate": 9.93019143300615e-05, "loss": 0.0728, "step": 10060 }, { "grad_norm": 0.48484480381011963, "learning_rate": 9.929915829688644e-05, "loss": 0.0577, "step": 10070 }, { "grad_norm": 0.502607524394989, "learning_rate": 9.929639687242955e-05, "loss": 0.0718, "step": 10080 }, { "grad_norm": 0.5465019345283508, "learning_rate": 9.929363005699281e-05, "loss": 0.0719, "step": 10090 }, { "grad_norm": 0.6133986711502075, "learning_rate": 9.92908578508788e-05, "loss": 0.0767, "step": 10100 }, { "grad_norm": 0.5725678205490112, "learning_rate": 9.928808025439069e-05, "loss": 0.0667, "step": 10110 }, { "grad_norm": 0.5427839159965515, "learning_rate": 9.928529726783223e-05, "loss": 0.0759, "step": 10120 }, { "grad_norm": 0.8249152898788452, "learning_rate": 9.928250889150774e-05, "loss": 0.096, "step": 10130 }, { "grad_norm": 0.599432647228241, "learning_rate": 9.92797151257222e-05, "loss": 0.0807, "step": 10140 }, { "grad_norm": 0.5824148654937744, "learning_rate": 9.927691597078108e-05, "loss": 0.0875, "step": 10150 }, { "grad_norm": 0.674490749835968, "learning_rate": 9.927411142699053e-05, "loss": 0.0906, "step": 10160 }, { "grad_norm": 0.677455723285675, "learning_rate": 9.927130149465725e-05, "loss": 0.0796, "step": 10170 }, { "grad_norm": 0.789050817489624, "learning_rate": 9.92684861740885e-05, "loss": 0.0938, "step": 10180 }, { "grad_norm": 0.674507200717926, "learning_rate": 9.926566546559217e-05, "loss": 0.0868, "step": 10190 }, { "grad_norm": 0.5477504730224609, "learning_rate": 9.926283936947673e-05, "loss": 0.0747, "step": 10200 }, { "grad_norm": 0.47681593894958496, "learning_rate": 9.926000788605126e-05, "loss": 0.08, "step": 10210 }, { "grad_norm": 0.40470096468925476, "learning_rate": 9.92571710156254e-05, "loss": 0.0855, "step": 10220 }, { "grad_norm": 0.624276340007782, "learning_rate": 9.925432875850936e-05, "loss": 0.0835, "step": 10230 }, { "grad_norm": 0.5507956743240356, "learning_rate": 9.925148111501396e-05, "loss": 0.0887, "step": 10240 }, { "grad_norm": 0.8159836530685425, "learning_rate": 9.924862808545066e-05, "loss": 0.0819, "step": 10250 }, { "grad_norm": 0.3343769311904907, "learning_rate": 9.924576967013141e-05, "loss": 0.065, "step": 10260 }, { "grad_norm": 0.691641628742218, "learning_rate": 9.924290586936887e-05, "loss": 0.082, "step": 10270 }, { "grad_norm": 0.49565091729164124, "learning_rate": 9.924003668347614e-05, "loss": 0.0758, "step": 10280 }, { "grad_norm": 0.6189122200012207, "learning_rate": 9.923716211276704e-05, "loss": 0.0839, "step": 10290 }, { "grad_norm": 0.5425558090209961, "learning_rate": 9.923428215755594e-05, "loss": 0.0766, "step": 10300 }, { "grad_norm": 0.731164813041687, "learning_rate": 9.923139681815775e-05, "loss": 0.0657, "step": 10310 }, { "grad_norm": 0.6877502202987671, "learning_rate": 9.922850609488801e-05, "loss": 0.0766, "step": 10320 }, { "grad_norm": 0.5508130788803101, "learning_rate": 9.922560998806287e-05, "loss": 0.0692, "step": 10330 }, { "grad_norm": 0.5991158485412598, "learning_rate": 9.922270849799905e-05, "loss": 0.0847, "step": 10340 }, { "grad_norm": 0.7190307378768921, "learning_rate": 9.92198016250138e-05, "loss": 0.0814, "step": 10350 }, { "grad_norm": 0.2959945499897003, "learning_rate": 9.921688936942506e-05, "loss": 0.0705, "step": 10360 }, { "grad_norm": 0.5507174730300903, "learning_rate": 9.921397173155129e-05, "loss": 0.0646, "step": 10370 }, { "grad_norm": 0.7134098410606384, "learning_rate": 9.921104871171157e-05, "loss": 0.0636, "step": 10380 }, { "grad_norm": 0.46504876017570496, "learning_rate": 9.920812031022554e-05, "loss": 0.0633, "step": 10390 }, { "grad_norm": 0.5362626314163208, "learning_rate": 9.920518652741348e-05, "loss": 0.0654, "step": 10400 }, { "grad_norm": 0.47509774565696716, "learning_rate": 9.920224736359618e-05, "loss": 0.0598, "step": 10410 }, { "grad_norm": 0.4741125702857971, "learning_rate": 9.91993028190951e-05, "loss": 0.0814, "step": 10420 }, { "grad_norm": 0.5852957367897034, "learning_rate": 9.919635289423222e-05, "loss": 0.0593, "step": 10430 }, { "grad_norm": 0.6544840931892395, "learning_rate": 9.919339758933015e-05, "loss": 0.0607, "step": 10440 }, { "grad_norm": 0.7947887778282166, "learning_rate": 9.919043690471209e-05, "loss": 0.0923, "step": 10450 }, { "grad_norm": 0.5865153074264526, "learning_rate": 9.91874708407018e-05, "loss": 0.0757, "step": 10460 }, { "grad_norm": 0.5049378871917725, "learning_rate": 9.918449939762367e-05, "loss": 0.0772, "step": 10470 }, { "grad_norm": 0.585197925567627, "learning_rate": 9.91815225758026e-05, "loss": 0.0722, "step": 10480 }, { "grad_norm": 0.5447506904602051, "learning_rate": 9.917854037556419e-05, "loss": 0.0781, "step": 10490 }, { "grad_norm": 0.5662792921066284, "learning_rate": 9.917555279723454e-05, "loss": 0.0611, "step": 10500 }, { "grad_norm": 0.5425170063972473, "learning_rate": 9.917255984114036e-05, "loss": 0.0899, "step": 10510 }, { "grad_norm": 0.4015394151210785, "learning_rate": 9.916956150760896e-05, "loss": 0.063, "step": 10520 }, { "grad_norm": 0.550714910030365, "learning_rate": 9.916655779696826e-05, "loss": 0.0865, "step": 10530 }, { "grad_norm": 0.6406130194664001, "learning_rate": 9.916354870954671e-05, "loss": 0.0884, "step": 10540 }, { "grad_norm": 0.6337080597877502, "learning_rate": 9.91605342456734e-05, "loss": 0.0798, "step": 10550 }, { "grad_norm": 0.5285753607749939, "learning_rate": 9.915751440567795e-05, "loss": 0.0701, "step": 10560 }, { "grad_norm": 0.6013398766517639, "learning_rate": 9.915448918989066e-05, "loss": 0.0638, "step": 10570 }, { "grad_norm": 0.6160750389099121, "learning_rate": 9.915145859864232e-05, "loss": 0.0808, "step": 10580 }, { "grad_norm": 0.729401171207428, "learning_rate": 9.914842263226437e-05, "loss": 0.0879, "step": 10590 }, { "grad_norm": 0.6127458214759827, "learning_rate": 9.914538129108882e-05, "loss": 0.087, "step": 10600 }, { "grad_norm": 0.6738700270652771, "learning_rate": 9.914233457544825e-05, "loss": 0.0962, "step": 10610 }, { "grad_norm": 0.5540087819099426, "learning_rate": 9.913928248567586e-05, "loss": 0.086, "step": 10620 }, { "grad_norm": 0.546323835849762, "learning_rate": 9.913622502210542e-05, "loss": 0.0734, "step": 10630 }, { "grad_norm": 0.5208922028541565, "learning_rate": 9.913316218507128e-05, "loss": 0.0853, "step": 10640 }, { "grad_norm": 0.35111531615257263, "learning_rate": 9.91300939749084e-05, "loss": 0.0708, "step": 10650 }, { "grad_norm": 0.6765012741088867, "learning_rate": 9.91270203919523e-05, "loss": 0.0801, "step": 10660 }, { "grad_norm": 0.6492353081703186, "learning_rate": 9.912394143653912e-05, "loss": 0.0782, "step": 10670 }, { "grad_norm": 0.44306424260139465, "learning_rate": 9.912085710900555e-05, "loss": 0.0671, "step": 10680 }, { "grad_norm": 0.6503391265869141, "learning_rate": 9.911776740968892e-05, "loss": 0.0699, "step": 10690 }, { "grad_norm": 0.5294598340988159, "learning_rate": 9.911467233892709e-05, "loss": 0.0963, "step": 10700 }, { "grad_norm": 0.6528551578521729, "learning_rate": 9.911157189705853e-05, "loss": 0.0752, "step": 10710 }, { "grad_norm": 0.7502562403678894, "learning_rate": 9.910846608442229e-05, "loss": 0.0707, "step": 10720 }, { "grad_norm": 0.5683915615081787, "learning_rate": 9.910535490135805e-05, "loss": 0.077, "step": 10730 }, { "grad_norm": 0.5196072459220886, "learning_rate": 9.910223834820603e-05, "loss": 0.0622, "step": 10740 }, { "grad_norm": 0.49101942777633667, "learning_rate": 9.909911642530703e-05, "loss": 0.0752, "step": 10750 }, { "grad_norm": 0.7154257893562317, "learning_rate": 9.909598913300249e-05, "loss": 0.0843, "step": 10760 }, { "grad_norm": 0.5089460015296936, "learning_rate": 9.909285647163438e-05, "loss": 0.0781, "step": 10770 }, { "grad_norm": 0.5704904794692993, "learning_rate": 9.908971844154531e-05, "loss": 0.0788, "step": 10780 }, { "grad_norm": 0.5844739079475403, "learning_rate": 9.908657504307843e-05, "loss": 0.0678, "step": 10790 }, { "grad_norm": 0.4463491439819336, "learning_rate": 9.908342627657751e-05, "loss": 0.0799, "step": 10800 }, { "grad_norm": 0.6713463068008423, "learning_rate": 9.908027214238689e-05, "loss": 0.0708, "step": 10810 }, { "grad_norm": 0.41794052720069885, "learning_rate": 9.90771126408515e-05, "loss": 0.0879, "step": 10820 }, { "grad_norm": 0.6867601871490479, "learning_rate": 9.907394777231685e-05, "loss": 0.0803, "step": 10830 }, { "grad_norm": 0.540793240070343, "learning_rate": 9.907077753712905e-05, "loss": 0.0908, "step": 10840 }, { "grad_norm": 0.7088074684143066, "learning_rate": 9.906760193563482e-05, "loss": 0.0782, "step": 10850 }, { "grad_norm": 0.8092772364616394, "learning_rate": 9.906442096818139e-05, "loss": 0.0744, "step": 10860 }, { "grad_norm": 0.42569851875305176, "learning_rate": 9.906123463511665e-05, "loss": 0.0674, "step": 10870 }, { "grad_norm": 0.4309368133544922, "learning_rate": 9.905804293678907e-05, "loss": 0.0817, "step": 10880 }, { "grad_norm": 0.421725869178772, "learning_rate": 9.905484587354766e-05, "loss": 0.068, "step": 10890 }, { "grad_norm": 0.4349479377269745, "learning_rate": 9.905164344574205e-05, "loss": 0.0789, "step": 10900 }, { "grad_norm": 0.48771533370018005, "learning_rate": 9.904843565372248e-05, "loss": 0.0736, "step": 10910 }, { "grad_norm": 0.5985084176063538, "learning_rate": 9.904522249783972e-05, "loss": 0.0758, "step": 10920 }, { "grad_norm": 0.4823741316795349, "learning_rate": 9.904200397844517e-05, "loss": 0.0895, "step": 10930 }, { "grad_norm": 0.5277074575424194, "learning_rate": 9.903878009589078e-05, "loss": 0.0791, "step": 10940 }, { "grad_norm": 0.6420908570289612, "learning_rate": 9.903555085052915e-05, "loss": 0.0648, "step": 10950 }, { "grad_norm": 0.7760536074638367, "learning_rate": 9.903231624271338e-05, "loss": 0.0565, "step": 10960 }, { "grad_norm": 0.4667903482913971, "learning_rate": 9.902907627279724e-05, "loss": 0.0666, "step": 10970 }, { "grad_norm": 0.7290205955505371, "learning_rate": 9.902583094113504e-05, "loss": 0.0758, "step": 10980 }, { "grad_norm": 0.6758085489273071, "learning_rate": 9.902258024808168e-05, "loss": 0.0693, "step": 10990 }, { "grad_norm": 0.43658801913261414, "learning_rate": 9.901932419399264e-05, "loss": 0.0702, "step": 11000 }, { "grad_norm": 0.6106215715408325, "learning_rate": 9.9016062779224e-05, "loss": 0.0736, "step": 11010 }, { "grad_norm": 1.1419212818145752, "learning_rate": 9.901279600413242e-05, "loss": 0.0809, "step": 11020 }, { "grad_norm": 0.6537956595420837, "learning_rate": 9.900952386907518e-05, "loss": 0.077, "step": 11030 }, { "grad_norm": 0.590052604675293, "learning_rate": 9.90062463744101e-05, "loss": 0.083, "step": 11040 }, { "grad_norm": 0.6480523347854614, "learning_rate": 9.900296352049558e-05, "loss": 0.0642, "step": 11050 }, { "grad_norm": 0.6845698356628418, "learning_rate": 9.899967530769065e-05, "loss": 0.083, "step": 11060 }, { "grad_norm": 0.7968766689300537, "learning_rate": 9.899638173635489e-05, "loss": 0.0738, "step": 11070 }, { "grad_norm": 0.7178570628166199, "learning_rate": 9.899308280684849e-05, "loss": 0.077, "step": 11080 }, { "grad_norm": 0.5889654159545898, "learning_rate": 9.898977851953222e-05, "loss": 0.0787, "step": 11090 }, { "grad_norm": 0.4368932247161865, "learning_rate": 9.898646887476741e-05, "loss": 0.0794, "step": 11100 }, { "grad_norm": 0.43081390857696533, "learning_rate": 9.898315387291603e-05, "loss": 0.0813, "step": 11110 }, { "grad_norm": 0.49789202213287354, "learning_rate": 9.89798335143406e-05, "loss": 0.0672, "step": 11120 }, { "grad_norm": 0.7046162486076355, "learning_rate": 9.897650779940419e-05, "loss": 0.0768, "step": 11130 }, { "grad_norm": 0.5344384908676147, "learning_rate": 9.897317672847054e-05, "loss": 0.0672, "step": 11140 }, { "grad_norm": 0.6097689270973206, "learning_rate": 9.89698403019039e-05, "loss": 0.0639, "step": 11150 }, { "grad_norm": 0.7525168657302856, "learning_rate": 9.896649852006917e-05, "loss": 0.0732, "step": 11160 }, { "grad_norm": 0.5794386863708496, "learning_rate": 9.896315138333177e-05, "loss": 0.0709, "step": 11170 }, { "grad_norm": 0.6384442448616028, "learning_rate": 9.895979889205774e-05, "loss": 0.0752, "step": 11180 }, { "grad_norm": 0.5476356744766235, "learning_rate": 9.895644104661372e-05, "loss": 0.0627, "step": 11190 }, { "grad_norm": 0.6137669682502747, "learning_rate": 9.895307784736691e-05, "loss": 0.0695, "step": 11200 }, { "grad_norm": 0.6862911581993103, "learning_rate": 9.894970929468512e-05, "loss": 0.0771, "step": 11210 }, { "grad_norm": 0.3684628903865814, "learning_rate": 9.89463353889367e-05, "loss": 0.0681, "step": 11220 }, { "grad_norm": 0.43850937485694885, "learning_rate": 9.894295613049065e-05, "loss": 0.0779, "step": 11230 }, { "grad_norm": 0.5790863633155823, "learning_rate": 9.893957151971649e-05, "loss": 0.0718, "step": 11240 }, { "grad_norm": 0.5577159523963928, "learning_rate": 9.893618155698436e-05, "loss": 0.0724, "step": 11250 }, { "grad_norm": 0.5518888235092163, "learning_rate": 9.8932786242665e-05, "loss": 0.0666, "step": 11260 }, { "grad_norm": 0.5102993249893188, "learning_rate": 9.89293855771297e-05, "loss": 0.066, "step": 11270 }, { "grad_norm": 0.5883191823959351, "learning_rate": 9.892597956075036e-05, "loss": 0.0735, "step": 11280 }, { "grad_norm": 0.7495232224464417, "learning_rate": 9.892256819389947e-05, "loss": 0.0753, "step": 11290 }, { "grad_norm": 0.5165077447891235, "learning_rate": 9.891915147695006e-05, "loss": 0.0793, "step": 11300 }, { "grad_norm": 0.7923174500465393, "learning_rate": 9.891572941027577e-05, "loss": 0.0714, "step": 11310 }, { "grad_norm": 0.47077614068984985, "learning_rate": 9.89123019942509e-05, "loss": 0.0725, "step": 11320 }, { "grad_norm": 0.5746206641197205, "learning_rate": 9.89088692292502e-05, "loss": 0.0621, "step": 11330 }, { "grad_norm": 0.53270024061203, "learning_rate": 9.89054311156491e-05, "loss": 0.0567, "step": 11340 }, { "grad_norm": 0.5039722919464111, "learning_rate": 9.890198765382357e-05, "loss": 0.0882, "step": 11350 }, { "grad_norm": 0.4344746172428131, "learning_rate": 9.889853884415021e-05, "loss": 0.0619, "step": 11360 }, { "grad_norm": 0.3452312648296356, "learning_rate": 9.889508468700614e-05, "loss": 0.0637, "step": 11370 }, { "grad_norm": 0.48165228962898254, "learning_rate": 9.889162518276915e-05, "loss": 0.0574, "step": 11380 }, { "grad_norm": 0.5680296421051025, "learning_rate": 9.888816033181752e-05, "loss": 0.0752, "step": 11390 }, { "grad_norm": 0.4580105245113373, "learning_rate": 9.888469013453018e-05, "loss": 0.067, "step": 11400 }, { "grad_norm": 0.7285307049751282, "learning_rate": 9.888121459128663e-05, "loss": 0.0685, "step": 11410 }, { "grad_norm": 0.5389612913131714, "learning_rate": 9.887773370246693e-05, "loss": 0.0815, "step": 11420 }, { "grad_norm": 0.5608277916908264, "learning_rate": 9.887424746845177e-05, "loss": 0.0728, "step": 11430 }, { "grad_norm": 0.5034499764442444, "learning_rate": 9.887075588962239e-05, "loss": 0.0663, "step": 11440 }, { "grad_norm": 0.5840266942977905, "learning_rate": 9.88672589663606e-05, "loss": 0.0638, "step": 11450 }, { "grad_norm": 0.4033248722553253, "learning_rate": 9.886375669904886e-05, "loss": 0.0822, "step": 11460 }, { "grad_norm": 0.45468106865882874, "learning_rate": 9.886024908807014e-05, "loss": 0.0567, "step": 11470 }, { "grad_norm": 0.527521550655365, "learning_rate": 9.885673613380806e-05, "loss": 0.0598, "step": 11480 }, { "grad_norm": 0.4747599959373474, "learning_rate": 9.885321783664676e-05, "loss": 0.071, "step": 11490 }, { "grad_norm": 0.47813647985458374, "learning_rate": 9.884969419697101e-05, "loss": 0.073, "step": 11500 }, { "grad_norm": 0.5118080377578735, "learning_rate": 9.884616521516614e-05, "loss": 0.0662, "step": 11510 }, { "grad_norm": 0.429801881313324, "learning_rate": 9.88426308916181e-05, "loss": 0.0736, "step": 11520 }, { "grad_norm": 0.6071243286132812, "learning_rate": 9.883909122671335e-05, "loss": 0.0618, "step": 11530 }, { "grad_norm": 0.5082982778549194, "learning_rate": 9.883554622083904e-05, "loss": 0.0799, "step": 11540 }, { "grad_norm": 0.5397990345954895, "learning_rate": 9.88319958743828e-05, "loss": 0.0824, "step": 11550 }, { "grad_norm": 0.6101346611976624, "learning_rate": 9.882844018773291e-05, "loss": 0.0924, "step": 11560 }, { "grad_norm": 0.6123495101928711, "learning_rate": 9.882487916127823e-05, "loss": 0.0774, "step": 11570 }, { "grad_norm": 0.5882100462913513, "learning_rate": 9.882131279540815e-05, "loss": 0.091, "step": 11580 }, { "grad_norm": 0.6352827548980713, "learning_rate": 9.881774109051271e-05, "loss": 0.075, "step": 11590 }, { "grad_norm": 0.47534605860710144, "learning_rate": 9.881416404698252e-05, "loss": 0.062, "step": 11600 }, { "grad_norm": 0.48596468567848206, "learning_rate": 9.881058166520873e-05, "loss": 0.0934, "step": 11610 }, { "grad_norm": 0.6736618280410767, "learning_rate": 9.880699394558311e-05, "loss": 0.0803, "step": 11620 }, { "grad_norm": 0.41677263379096985, "learning_rate": 9.880340088849801e-05, "loss": 0.0871, "step": 11630 }, { "grad_norm": 0.6256076693534851, "learning_rate": 9.879980249434637e-05, "loss": 0.0673, "step": 11640 }, { "grad_norm": 0.5016670823097229, "learning_rate": 9.879619876352168e-05, "loss": 0.0635, "step": 11650 }, { "grad_norm": 0.4704257547855377, "learning_rate": 9.879258969641809e-05, "loss": 0.0837, "step": 11660 }, { "grad_norm": 0.3535700738430023, "learning_rate": 9.878897529343023e-05, "loss": 0.0568, "step": 11670 }, { "grad_norm": 0.5980286002159119, "learning_rate": 9.878535555495338e-05, "loss": 0.0749, "step": 11680 }, { "grad_norm": 0.6338942050933838, "learning_rate": 9.87817304813834e-05, "loss": 0.0826, "step": 11690 }, { "grad_norm": 0.4885433316230774, "learning_rate": 9.877810007311671e-05, "loss": 0.0793, "step": 11700 }, { "grad_norm": 0.45223480463027954, "learning_rate": 9.877446433055035e-05, "loss": 0.072, "step": 11710 }, { "grad_norm": 0.7132576704025269, "learning_rate": 9.877082325408191e-05, "loss": 0.0783, "step": 11720 }, { "grad_norm": 0.5287561416625977, "learning_rate": 9.876717684410954e-05, "loss": 0.0574, "step": 11730 }, { "grad_norm": 0.5997157692909241, "learning_rate": 9.876352510103204e-05, "loss": 0.0756, "step": 11740 }, { "grad_norm": 0.6251075863838196, "learning_rate": 9.875986802524875e-05, "loss": 0.0753, "step": 11750 }, { "grad_norm": 0.7383955121040344, "learning_rate": 9.87562056171596e-05, "loss": 0.0708, "step": 11760 }, { "grad_norm": 0.7504293322563171, "learning_rate": 9.875253787716511e-05, "loss": 0.0922, "step": 11770 }, { "grad_norm": 0.5392804145812988, "learning_rate": 9.874886480566637e-05, "loss": 0.0692, "step": 11780 }, { "grad_norm": 0.3815072774887085, "learning_rate": 9.874518640306507e-05, "loss": 0.0665, "step": 11790 }, { "grad_norm": 0.42273345589637756, "learning_rate": 9.874150266976347e-05, "loss": 0.0577, "step": 11800 }, { "grad_norm": 0.478396475315094, "learning_rate": 9.873781360616443e-05, "loss": 0.0683, "step": 11810 }, { "grad_norm": 0.34927183389663696, "learning_rate": 9.873411921267137e-05, "loss": 0.0635, "step": 11820 }, { "grad_norm": 0.592807948589325, "learning_rate": 9.873041948968829e-05, "loss": 0.084, "step": 11830 }, { "grad_norm": 0.5133899450302124, "learning_rate": 9.872671443761981e-05, "loss": 0.0656, "step": 11840 }, { "grad_norm": 0.8440218567848206, "learning_rate": 9.872300405687109e-05, "loss": 0.0615, "step": 11850 }, { "grad_norm": 0.6929670572280884, "learning_rate": 9.871928834784792e-05, "loss": 0.076, "step": 11860 }, { "grad_norm": 0.40489068627357483, "learning_rate": 9.871556731095661e-05, "loss": 0.0705, "step": 11870 }, { "grad_norm": 0.5088246464729309, "learning_rate": 9.871184094660411e-05, "loss": 0.0791, "step": 11880 }, { "grad_norm": 0.462959349155426, "learning_rate": 9.870810925519791e-05, "loss": 0.0679, "step": 11890 }, { "grad_norm": 0.504414975643158, "learning_rate": 9.870437223714612e-05, "loss": 0.0652, "step": 11900 }, { "grad_norm": 0.5907100439071655, "learning_rate": 9.87006298928574e-05, "loss": 0.0854, "step": 11910 }, { "grad_norm": 0.39538681507110596, "learning_rate": 9.869688222274103e-05, "loss": 0.0666, "step": 11920 }, { "grad_norm": 0.44356659054756165, "learning_rate": 9.869312922720681e-05, "loss": 0.0671, "step": 11930 }, { "grad_norm": 0.4872790575027466, "learning_rate": 9.868937090666521e-05, "loss": 0.0678, "step": 11940 }, { "grad_norm": 0.6944438815116882, "learning_rate": 9.86856072615272e-05, "loss": 0.0678, "step": 11950 }, { "grad_norm": 0.7726845741271973, "learning_rate": 9.868183829220438e-05, "loss": 0.069, "step": 11960 }, { "grad_norm": 0.49922698736190796, "learning_rate": 9.867806399910893e-05, "loss": 0.0684, "step": 11970 }, { "grad_norm": 0.586463451385498, "learning_rate": 9.867428438265356e-05, "loss": 0.0776, "step": 11980 }, { "grad_norm": 0.49765777587890625, "learning_rate": 9.867049944325165e-05, "loss": 0.0742, "step": 11990 }, { "grad_norm": 0.5146511793136597, "learning_rate": 9.86667091813171e-05, "loss": 0.0655, "step": 12000 }, { "grad_norm": 0.501592218875885, "learning_rate": 9.866291359726438e-05, "loss": 0.0605, "step": 12010 }, { "grad_norm": 0.7261500954627991, "learning_rate": 9.865911269150861e-05, "loss": 0.0629, "step": 12020 }, { "grad_norm": 0.832638144493103, "learning_rate": 9.865530646446544e-05, "loss": 0.0746, "step": 12030 }, { "grad_norm": 0.7442018389701843, "learning_rate": 9.86514949165511e-05, "loss": 0.0757, "step": 12040 }, { "grad_norm": 0.6052802205085754, "learning_rate": 9.864767804818243e-05, "loss": 0.0853, "step": 12050 }, { "grad_norm": 0.4690454602241516, "learning_rate": 9.86438558597768e-05, "loss": 0.0697, "step": 12060 }, { "grad_norm": 0.5457866787910461, "learning_rate": 9.864002835175225e-05, "loss": 0.0933, "step": 12070 }, { "grad_norm": 0.5055340528488159, "learning_rate": 9.863619552452734e-05, "loss": 0.0752, "step": 12080 }, { "grad_norm": 0.44193512201309204, "learning_rate": 9.863235737852119e-05, "loss": 0.057, "step": 12090 }, { "grad_norm": 0.5447049140930176, "learning_rate": 9.862851391415356e-05, "loss": 0.0674, "step": 12100 }, { "grad_norm": 0.414071261882782, "learning_rate": 9.862466513184477e-05, "loss": 0.0721, "step": 12110 }, { "grad_norm": 0.6266319155693054, "learning_rate": 9.86208110320157e-05, "loss": 0.0599, "step": 12120 }, { "grad_norm": 0.4596345126628876, "learning_rate": 9.861695161508784e-05, "loss": 0.0726, "step": 12130 }, { "grad_norm": 0.4168185591697693, "learning_rate": 9.861308688148324e-05, "loss": 0.0593, "step": 12140 }, { "grad_norm": 0.5578635334968567, "learning_rate": 9.860921683162455e-05, "loss": 0.0656, "step": 12150 }, { "grad_norm": 0.4913216233253479, "learning_rate": 9.860534146593499e-05, "loss": 0.062, "step": 12160 }, { "grad_norm": 0.5117233395576477, "learning_rate": 9.860146078483836e-05, "loss": 0.0629, "step": 12170 }, { "grad_norm": 0.5830202102661133, "learning_rate": 9.859757478875905e-05, "loss": 0.0702, "step": 12180 }, { "grad_norm": 0.578267514705658, "learning_rate": 9.859368347812204e-05, "loss": 0.0638, "step": 12190 }, { "grad_norm": 0.42359548807144165, "learning_rate": 9.858978685335285e-05, "loss": 0.0734, "step": 12200 }, { "grad_norm": 0.7946180701255798, "learning_rate": 9.858588491487763e-05, "loss": 0.0914, "step": 12210 }, { "grad_norm": 0.6070917844772339, "learning_rate": 9.858197766312308e-05, "loss": 0.0703, "step": 12220 }, { "grad_norm": 0.42752259969711304, "learning_rate": 9.857806509851649e-05, "loss": 0.0673, "step": 12230 }, { "grad_norm": 0.5968979597091675, "learning_rate": 9.857414722148574e-05, "loss": 0.0713, "step": 12240 }, { "grad_norm": 0.4393758773803711, "learning_rate": 9.857022403245928e-05, "loss": 0.0759, "step": 12250 }, { "grad_norm": 0.7712079882621765, "learning_rate": 9.856629553186615e-05, "loss": 0.0884, "step": 12260 }, { "grad_norm": 0.7675121426582336, "learning_rate": 9.856236172013595e-05, "loss": 0.0752, "step": 12270 }, { "grad_norm": 0.49950793385505676, "learning_rate": 9.85584225976989e-05, "loss": 0.0686, "step": 12280 }, { "grad_norm": 0.5969711542129517, "learning_rate": 9.855447816498575e-05, "loss": 0.0721, "step": 12290 }, { "grad_norm": 0.7397503852844238, "learning_rate": 9.855052842242787e-05, "loss": 0.0728, "step": 12300 }, { "grad_norm": 0.5287004113197327, "learning_rate": 9.85465733704572e-05, "loss": 0.065, "step": 12310 }, { "grad_norm": 0.759196400642395, "learning_rate": 9.854261300950624e-05, "loss": 0.0696, "step": 12320 }, { "grad_norm": 0.7621052265167236, "learning_rate": 9.853864734000813e-05, "loss": 0.0829, "step": 12330 }, { "grad_norm": 0.6193182468414307, "learning_rate": 9.85346763623965e-05, "loss": 0.0769, "step": 12340 }, { "grad_norm": 0.46172547340393066, "learning_rate": 9.853070007710564e-05, "loss": 0.0731, "step": 12350 }, { "grad_norm": 0.433502733707428, "learning_rate": 9.85267184845704e-05, "loss": 0.0684, "step": 12360 }, { "grad_norm": 0.5461500883102417, "learning_rate": 9.852273158522616e-05, "loss": 0.0594, "step": 12370 }, { "grad_norm": 0.5688590407371521, "learning_rate": 9.851873937950896e-05, "loss": 0.0751, "step": 12380 }, { "grad_norm": 0.5662327408790588, "learning_rate": 9.851474186785537e-05, "loss": 0.073, "step": 12390 }, { "grad_norm": 0.5819305181503296, "learning_rate": 9.851073905070254e-05, "loss": 0.0841, "step": 12400 }, { "grad_norm": 0.6639615297317505, "learning_rate": 9.850673092848824e-05, "loss": 0.0787, "step": 12410 }, { "grad_norm": 0.6016116142272949, "learning_rate": 9.850271750165077e-05, "loss": 0.0818, "step": 12420 }, { "grad_norm": 0.6345705389976501, "learning_rate": 9.849869877062902e-05, "loss": 0.0859, "step": 12430 }, { "grad_norm": 0.6082649827003479, "learning_rate": 9.849467473586252e-05, "loss": 0.0913, "step": 12440 }, { "grad_norm": 0.6596473455429077, "learning_rate": 9.849064539779127e-05, "loss": 0.0716, "step": 12450 }, { "grad_norm": 0.44764652848243713, "learning_rate": 9.848661075685594e-05, "loss": 0.0928, "step": 12460 }, { "grad_norm": 0.5074753761291504, "learning_rate": 9.848257081349778e-05, "loss": 0.0784, "step": 12470 }, { "grad_norm": 0.5533052682876587, "learning_rate": 9.847852556815856e-05, "loss": 0.0544, "step": 12480 }, { "grad_norm": 0.5885841250419617, "learning_rate": 9.847447502128067e-05, "loss": 0.0773, "step": 12490 }, { "grad_norm": 0.41210606694221497, "learning_rate": 9.847041917330708e-05, "loss": 0.0797, "step": 12500 }, { "grad_norm": 0.5636516809463501, "learning_rate": 9.846635802468132e-05, "loss": 0.0853, "step": 12510 }, { "grad_norm": 0.39357396960258484, "learning_rate": 9.84622915758475e-05, "loss": 0.065, "step": 12520 }, { "grad_norm": 0.7704490423202515, "learning_rate": 9.845821982725034e-05, "loss": 0.0663, "step": 12530 }, { "grad_norm": 0.48410916328430176, "learning_rate": 9.845414277933514e-05, "loss": 0.0696, "step": 12540 }, { "grad_norm": 0.5838156342506409, "learning_rate": 9.845006043254771e-05, "loss": 0.066, "step": 12550 }, { "grad_norm": 0.7510056495666504, "learning_rate": 9.844597278733451e-05, "loss": 0.0635, "step": 12560 }, { "grad_norm": 0.5712710618972778, "learning_rate": 9.844187984414259e-05, "loss": 0.0757, "step": 12570 }, { "grad_norm": 0.41844016313552856, "learning_rate": 9.84377816034195e-05, "loss": 0.0632, "step": 12580 }, { "grad_norm": 0.5220766663551331, "learning_rate": 9.843367806561345e-05, "loss": 0.08, "step": 12590 }, { "grad_norm": 0.5154915452003479, "learning_rate": 9.842956923117317e-05, "loss": 0.0782, "step": 12600 }, { "grad_norm": 0.4979873597621918, "learning_rate": 9.842545510054802e-05, "loss": 0.0593, "step": 12610 }, { "grad_norm": 0.513542115688324, "learning_rate": 9.842133567418792e-05, "loss": 0.0637, "step": 12620 }, { "grad_norm": 0.5202860236167908, "learning_rate": 9.841721095254333e-05, "loss": 0.0631, "step": 12630 }, { "grad_norm": 0.5741661787033081, "learning_rate": 9.841308093606537e-05, "loss": 0.073, "step": 12640 }, { "grad_norm": 0.4571308493614197, "learning_rate": 9.840894562520565e-05, "loss": 0.0764, "step": 12650 }, { "grad_norm": 0.637315571308136, "learning_rate": 9.840480502041642e-05, "loss": 0.0835, "step": 12660 }, { "grad_norm": 0.5991071462631226, "learning_rate": 9.840065912215049e-05, "loss": 0.0735, "step": 12670 }, { "grad_norm": 0.35266491770744324, "learning_rate": 9.839650793086124e-05, "loss": 0.0787, "step": 12680 }, { "grad_norm": 0.5887558460235596, "learning_rate": 9.839235144700265e-05, "loss": 0.0629, "step": 12690 }, { "grad_norm": 0.4725438058376312, "learning_rate": 9.838818967102926e-05, "loss": 0.0614, "step": 12700 }, { "grad_norm": 0.5964264273643494, "learning_rate": 9.83840226033962e-05, "loss": 0.0636, "step": 12710 }, { "grad_norm": 0.904531717300415, "learning_rate": 9.837985024455918e-05, "loss": 0.0741, "step": 12720 }, { "grad_norm": 0.4650713801383972, "learning_rate": 9.837567259497447e-05, "loss": 0.074, "step": 12730 }, { "grad_norm": 0.42584091424942017, "learning_rate": 9.837148965509894e-05, "loss": 0.076, "step": 12740 }, { "grad_norm": 0.6133404970169067, "learning_rate": 9.836730142539001e-05, "loss": 0.0683, "step": 12750 }, { "grad_norm": 0.7446176409721375, "learning_rate": 9.836310790630574e-05, "loss": 0.0713, "step": 12760 }, { "grad_norm": 0.8197039365768433, "learning_rate": 9.83589090983047e-05, "loss": 0.0791, "step": 12770 }, { "grad_norm": 0.6010059714317322, "learning_rate": 9.835470500184605e-05, "loss": 0.0739, "step": 12780 }, { "grad_norm": 0.3399631083011627, "learning_rate": 9.835049561738957e-05, "loss": 0.0653, "step": 12790 }, { "grad_norm": 0.5061793923377991, "learning_rate": 9.834628094539558e-05, "loss": 0.0647, "step": 12800 }, { "grad_norm": 0.5026017427444458, "learning_rate": 9.834206098632499e-05, "loss": 0.0621, "step": 12810 }, { "grad_norm": 0.4809717833995819, "learning_rate": 9.833783574063931e-05, "loss": 0.0578, "step": 12820 }, { "grad_norm": 0.7253513336181641, "learning_rate": 9.833360520880058e-05, "loss": 0.0887, "step": 12830 }, { "grad_norm": 0.367409348487854, "learning_rate": 9.832936939127144e-05, "loss": 0.0746, "step": 12840 }, { "grad_norm": 0.8751614689826965, "learning_rate": 9.832512828851515e-05, "loss": 0.0633, "step": 12850 }, { "grad_norm": 0.5153828263282776, "learning_rate": 9.832088190099546e-05, "loss": 0.0624, "step": 12860 }, { "grad_norm": 0.5172762870788574, "learning_rate": 9.831663022917679e-05, "loss": 0.068, "step": 12870 }, { "grad_norm": 0.529454231262207, "learning_rate": 9.831237327352407e-05, "loss": 0.0651, "step": 12880 }, { "grad_norm": 0.8044376969337463, "learning_rate": 9.830811103450286e-05, "loss": 0.0714, "step": 12890 }, { "grad_norm": 0.5174671411514282, "learning_rate": 9.830384351257924e-05, "loss": 0.0785, "step": 12900 }, { "grad_norm": 0.4999610185623169, "learning_rate": 9.829957070821993e-05, "loss": 0.0642, "step": 12910 }, { "grad_norm": 0.543755054473877, "learning_rate": 9.829529262189218e-05, "loss": 0.0661, "step": 12920 }, { "grad_norm": 0.7964004278182983, "learning_rate": 9.829100925406385e-05, "loss": 0.0731, "step": 12930 }, { "grad_norm": 0.5573684573173523, "learning_rate": 9.828672060520333e-05, "loss": 0.0791, "step": 12940 }, { "grad_norm": 0.42689749598503113, "learning_rate": 9.828242667577966e-05, "loss": 0.069, "step": 12950 }, { "grad_norm": 0.4449370205402374, "learning_rate": 9.82781274662624e-05, "loss": 0.0614, "step": 12960 }, { "grad_norm": 0.5911786556243896, "learning_rate": 9.82738229771217e-05, "loss": 0.0613, "step": 12970 }, { "grad_norm": 0.4373347759246826, "learning_rate": 9.826951320882829e-05, "loss": 0.0565, "step": 12980 }, { "grad_norm": 0.3101674020290375, "learning_rate": 9.826519816185351e-05, "loss": 0.0565, "step": 12990 }, { "grad_norm": 0.7042846083641052, "learning_rate": 9.826087783666921e-05, "loss": 0.0809, "step": 13000 }, { "grad_norm": 0.4248759150505066, "learning_rate": 9.825655223374787e-05, "loss": 0.0884, "step": 13010 }, { "grad_norm": 0.5952534675598145, "learning_rate": 9.825222135356253e-05, "loss": 0.0681, "step": 13020 }, { "grad_norm": 0.5030083656311035, "learning_rate": 9.82478851965868e-05, "loss": 0.0784, "step": 13030 }, { "grad_norm": 0.5082656145095825, "learning_rate": 9.82435437632949e-05, "loss": 0.0631, "step": 13040 }, { "grad_norm": 0.3810165524482727, "learning_rate": 9.823919705416158e-05, "loss": 0.0553, "step": 13050 }, { "grad_norm": 0.46733832359313965, "learning_rate": 9.82348450696622e-05, "loss": 0.0627, "step": 13060 }, { "grad_norm": 0.7237235903739929, "learning_rate": 9.823048781027268e-05, "loss": 0.0764, "step": 13070 }, { "grad_norm": 0.4902152717113495, "learning_rate": 9.822612527646953e-05, "loss": 0.0725, "step": 13080 }, { "grad_norm": 0.863141655921936, "learning_rate": 9.822175746872984e-05, "loss": 0.0797, "step": 13090 }, { "grad_norm": 0.7832042574882507, "learning_rate": 9.821738438753123e-05, "loss": 0.0749, "step": 13100 }, { "grad_norm": 0.498169481754303, "learning_rate": 9.821300603335196e-05, "loss": 0.0657, "step": 13110 }, { "grad_norm": 0.5696436166763306, "learning_rate": 9.820862240667085e-05, "loss": 0.0722, "step": 13120 }, { "grad_norm": 0.7882247567176819, "learning_rate": 9.820423350796726e-05, "loss": 0.0739, "step": 13130 }, { "grad_norm": 0.41701075434684753, "learning_rate": 9.819983933772118e-05, "loss": 0.0639, "step": 13140 }, { "grad_norm": 0.650445282459259, "learning_rate": 9.819543989641314e-05, "loss": 0.0713, "step": 13150 }, { "grad_norm": 0.4793495833873749, "learning_rate": 9.819103518452423e-05, "loss": 0.0627, "step": 13160 }, { "grad_norm": 0.4247313439846039, "learning_rate": 9.818662520253618e-05, "loss": 0.0719, "step": 13170 }, { "grad_norm": 0.5854836702346802, "learning_rate": 9.818220995093126e-05, "loss": 0.0672, "step": 13180 }, { "grad_norm": 0.479922354221344, "learning_rate": 9.817778943019228e-05, "loss": 0.07, "step": 13190 }, { "grad_norm": 0.5279218554496765, "learning_rate": 9.81733636408027e-05, "loss": 0.0775, "step": 13200 }, { "grad_norm": 0.3863547444343567, "learning_rate": 9.816893258324649e-05, "loss": 0.0626, "step": 13210 }, { "grad_norm": 0.6885227560997009, "learning_rate": 9.816449625800823e-05, "loss": 0.0625, "step": 13220 }, { "grad_norm": 0.5111488699913025, "learning_rate": 9.816005466557308e-05, "loss": 0.0635, "step": 13230 }, { "grad_norm": 0.5683909058570862, "learning_rate": 9.815560780642674e-05, "loss": 0.079, "step": 13240 }, { "grad_norm": 0.5828752517700195, "learning_rate": 9.815115568105555e-05, "loss": 0.0702, "step": 13250 }, { "grad_norm": 0.6883361339569092, "learning_rate": 9.814669828994638e-05, "loss": 0.0803, "step": 13260 }, { "grad_norm": 0.8285412192344666, "learning_rate": 9.814223563358665e-05, "loss": 0.0729, "step": 13270 }, { "grad_norm": 0.5934715867042542, "learning_rate": 9.813776771246443e-05, "loss": 0.0647, "step": 13280 }, { "grad_norm": 0.4591161608695984, "learning_rate": 9.813329452706829e-05, "loss": 0.0663, "step": 13290 }, { "grad_norm": 1.1452804803848267, "learning_rate": 9.812881607788744e-05, "loss": 0.0668, "step": 13300 }, { "grad_norm": 0.6690175533294678, "learning_rate": 9.812433236541163e-05, "loss": 0.0763, "step": 13310 }, { "grad_norm": 0.6589303016662598, "learning_rate": 9.811984339013116e-05, "loss": 0.086, "step": 13320 }, { "grad_norm": 0.624547004699707, "learning_rate": 9.811534915253698e-05, "loss": 0.0914, "step": 13330 }, { "grad_norm": 0.5113986730575562, "learning_rate": 9.811084965312056e-05, "loss": 0.0719, "step": 13340 }, { "grad_norm": 0.5587060451507568, "learning_rate": 9.810634489237396e-05, "loss": 0.0636, "step": 13350 }, { "grad_norm": 0.43832629919052124, "learning_rate": 9.81018348707898e-05, "loss": 0.0682, "step": 13360 }, { "grad_norm": 0.40668049454689026, "learning_rate": 9.809731958886131e-05, "loss": 0.0759, "step": 13370 }, { "grad_norm": 0.620112955570221, "learning_rate": 9.809279904708224e-05, "loss": 0.06, "step": 13380 }, { "grad_norm": 0.7530269026756287, "learning_rate": 9.808827324594699e-05, "loss": 0.0724, "step": 13390 }, { "grad_norm": 0.5538049340248108, "learning_rate": 9.808374218595046e-05, "loss": 0.071, "step": 13400 }, { "grad_norm": 0.5316440463066101, "learning_rate": 9.80792058675882e-05, "loss": 0.0736, "step": 13410 }, { "grad_norm": 0.4782906174659729, "learning_rate": 9.807466429135627e-05, "loss": 0.0639, "step": 13420 }, { "grad_norm": 0.49148738384246826, "learning_rate": 9.807011745775132e-05, "loss": 0.0649, "step": 13430 }, { "grad_norm": 0.5392848253250122, "learning_rate": 9.806556536727061e-05, "loss": 0.0544, "step": 13440 }, { "grad_norm": 0.6316900253295898, "learning_rate": 9.806100802041193e-05, "loss": 0.0737, "step": 13450 }, { "grad_norm": 0.46576064825057983, "learning_rate": 9.805644541767368e-05, "loss": 0.0766, "step": 13460 }, { "grad_norm": 0.48292088508605957, "learning_rate": 9.805187755955478e-05, "loss": 0.0819, "step": 13470 }, { "grad_norm": 0.4719443619251251, "learning_rate": 9.804730444655483e-05, "loss": 0.0643, "step": 13480 }, { "grad_norm": 0.42616453766822815, "learning_rate": 9.804272607917388e-05, "loss": 0.078, "step": 13490 }, { "grad_norm": 0.4672476351261139, "learning_rate": 9.803814245791265e-05, "loss": 0.0612, "step": 13500 }, { "grad_norm": 0.5074482560157776, "learning_rate": 9.803355358327239e-05, "loss": 0.0746, "step": 13510 }, { "grad_norm": 0.45241639018058777, "learning_rate": 9.802895945575492e-05, "loss": 0.0759, "step": 13520 }, { "grad_norm": 0.6042999625205994, "learning_rate": 9.802436007586266e-05, "loss": 0.075, "step": 13530 }, { "grad_norm": 0.5399840474128723, "learning_rate": 9.801975544409858e-05, "loss": 0.0762, "step": 13540 }, { "grad_norm": 0.551956832408905, "learning_rate": 9.801514556096625e-05, "loss": 0.0642, "step": 13550 }, { "grad_norm": 0.4691571593284607, "learning_rate": 9.801053042696977e-05, "loss": 0.0669, "step": 13560 }, { "grad_norm": 0.569829523563385, "learning_rate": 9.800591004261388e-05, "loss": 0.062, "step": 13570 }, { "grad_norm": 0.5322523713111877, "learning_rate": 9.800128440840385e-05, "loss": 0.06, "step": 13580 }, { "grad_norm": 0.3659668266773224, "learning_rate": 9.799665352484552e-05, "loss": 0.0669, "step": 13590 }, { "grad_norm": 0.4201550781726837, "learning_rate": 9.799201739244532e-05, "loss": 0.0889, "step": 13600 }, { "grad_norm": 0.42288467288017273, "learning_rate": 9.798737601171025e-05, "loss": 0.057, "step": 13610 }, { "grad_norm": 0.4302201569080353, "learning_rate": 9.79827293831479e-05, "loss": 0.0702, "step": 13620 }, { "grad_norm": 0.5343929529190063, "learning_rate": 9.797807750726638e-05, "loss": 0.0558, "step": 13630 }, { "grad_norm": 0.7997658848762512, "learning_rate": 9.797342038457446e-05, "loss": 0.0695, "step": 13640 }, { "grad_norm": 0.49725982546806335, "learning_rate": 9.796875801558141e-05, "loss": 0.0672, "step": 13650 }, { "grad_norm": 0.5965332984924316, "learning_rate": 9.79640904007971e-05, "loss": 0.0646, "step": 13660 }, { "grad_norm": 0.4266302287578583, "learning_rate": 9.795941754073199e-05, "loss": 0.0712, "step": 13670 }, { "grad_norm": 0.8706985116004944, "learning_rate": 9.795473943589705e-05, "loss": 0.0614, "step": 13680 }, { "grad_norm": 0.6896582245826721, "learning_rate": 9.795005608680394e-05, "loss": 0.0731, "step": 13690 }, { "grad_norm": 0.521728515625, "learning_rate": 9.794536749396477e-05, "loss": 0.0617, "step": 13700 }, { "grad_norm": 0.42736414074897766, "learning_rate": 9.79406736578923e-05, "loss": 0.064, "step": 13710 }, { "grad_norm": 0.5469321608543396, "learning_rate": 9.793597457909984e-05, "loss": 0.0562, "step": 13720 }, { "grad_norm": 0.47165441513061523, "learning_rate": 9.793127025810127e-05, "loss": 0.0609, "step": 13730 }, { "grad_norm": 0.7437839508056641, "learning_rate": 9.792656069541104e-05, "loss": 0.0741, "step": 13740 }, { "grad_norm": 0.5189257860183716, "learning_rate": 9.79218458915442e-05, "loss": 0.0651, "step": 13750 }, { "grad_norm": 1.1942404508590698, "learning_rate": 9.791712584701634e-05, "loss": 0.0883, "step": 13760 }, { "grad_norm": 0.5302950143814087, "learning_rate": 9.791240056234364e-05, "loss": 0.0586, "step": 13770 }, { "grad_norm": 0.6368688941001892, "learning_rate": 9.790767003804283e-05, "loss": 0.0602, "step": 13780 }, { "grad_norm": 0.47332489490509033, "learning_rate": 9.790293427463126e-05, "loss": 0.071, "step": 13790 }, { "grad_norm": 0.5212303400039673, "learning_rate": 9.789819327262684e-05, "loss": 0.0785, "step": 13800 }, { "grad_norm": 0.7147567868232727, "learning_rate": 9.7893447032548e-05, "loss": 0.0847, "step": 13810 }, { "grad_norm": 0.5030625462532043, "learning_rate": 9.78886955549138e-05, "loss": 0.08, "step": 13820 }, { "grad_norm": 0.965728223323822, "learning_rate": 9.788393884024387e-05, "loss": 0.0692, "step": 13830 }, { "grad_norm": 0.5742402076721191, "learning_rate": 9.787917688905836e-05, "loss": 0.0651, "step": 13840 }, { "grad_norm": 0.611840546131134, "learning_rate": 9.787440970187807e-05, "loss": 0.0619, "step": 13850 }, { "grad_norm": 0.5815209746360779, "learning_rate": 9.786963727922429e-05, "loss": 0.0671, "step": 13860 }, { "grad_norm": 0.5347383618354797, "learning_rate": 9.786485962161897e-05, "loss": 0.0815, "step": 13870 }, { "grad_norm": 0.46453970670700073, "learning_rate": 9.786007672958455e-05, "loss": 0.0494, "step": 13880 }, { "grad_norm": 0.580061674118042, "learning_rate": 9.78552886036441e-05, "loss": 0.0651, "step": 13890 }, { "grad_norm": 0.37128162384033203, "learning_rate": 9.785049524432124e-05, "loss": 0.0501, "step": 13900 }, { "grad_norm": 0.5777103900909424, "learning_rate": 9.784569665214016e-05, "loss": 0.0641, "step": 13910 }, { "grad_norm": 0.39552032947540283, "learning_rate": 9.784089282762563e-05, "loss": 0.0644, "step": 13920 }, { "grad_norm": 0.4344431757926941, "learning_rate": 9.7836083771303e-05, "loss": 0.0573, "step": 13930 }, { "grad_norm": 0.47207462787628174, "learning_rate": 9.783126948369817e-05, "loss": 0.0651, "step": 13940 }, { "grad_norm": 0.4290911555290222, "learning_rate": 9.78264499653376e-05, "loss": 0.0638, "step": 13950 }, { "grad_norm": 0.40351781249046326, "learning_rate": 9.782162521674838e-05, "loss": 0.0675, "step": 13960 }, { "grad_norm": 0.3902740776538849, "learning_rate": 9.781679523845812e-05, "loss": 0.0773, "step": 13970 }, { "grad_norm": 0.5706748962402344, "learning_rate": 9.781196003099502e-05, "loss": 0.0669, "step": 13980 }, { "grad_norm": 0.4223664700984955, "learning_rate": 9.780711959488786e-05, "loss": 0.0742, "step": 13990 }, { "grad_norm": 0.43105730414390564, "learning_rate": 9.780227393066599e-05, "loss": 0.0626, "step": 14000 }, { "grad_norm": 0.4419293999671936, "learning_rate": 9.77974230388593e-05, "loss": 0.0697, "step": 14010 }, { "grad_norm": 0.7239528894424438, "learning_rate": 9.779256691999829e-05, "loss": 0.0736, "step": 14020 }, { "grad_norm": 0.8068428635597229, "learning_rate": 9.778770557461403e-05, "loss": 0.0756, "step": 14030 }, { "grad_norm": 0.7830337285995483, "learning_rate": 9.778283900323812e-05, "loss": 0.0746, "step": 14040 }, { "grad_norm": 0.5179188847541809, "learning_rate": 9.777796720640277e-05, "loss": 0.0607, "step": 14050 }, { "grad_norm": 1.1685646772384644, "learning_rate": 9.777309018464078e-05, "loss": 0.0674, "step": 14060 }, { "grad_norm": 0.937086284160614, "learning_rate": 9.776820793848547e-05, "loss": 0.0672, "step": 14070 }, { "grad_norm": 0.638268232345581, "learning_rate": 9.776332046847075e-05, "loss": 0.0723, "step": 14080 }, { "grad_norm": 0.4416545033454895, "learning_rate": 9.775842777513111e-05, "loss": 0.0707, "step": 14090 }, { "grad_norm": 0.7976468205451965, "learning_rate": 9.775352985900163e-05, "loss": 0.0849, "step": 14100 }, { "grad_norm": 0.7249411940574646, "learning_rate": 9.774862672061791e-05, "loss": 0.0849, "step": 14110 }, { "grad_norm": 0.6748083233833313, "learning_rate": 9.774371836051616e-05, "loss": 0.0765, "step": 14120 }, { "grad_norm": 0.6037312150001526, "learning_rate": 9.773880477923315e-05, "loss": 0.0768, "step": 14130 }, { "grad_norm": 0.5690031051635742, "learning_rate": 9.773388597730623e-05, "loss": 0.0703, "step": 14140 }, { "grad_norm": 0.5851536393165588, "learning_rate": 9.77289619552733e-05, "loss": 0.0686, "step": 14150 }, { "grad_norm": 0.8313645124435425, "learning_rate": 9.772403271367285e-05, "loss": 0.0744, "step": 14160 }, { "grad_norm": 0.6811233162879944, "learning_rate": 9.771909825304396e-05, "loss": 0.07, "step": 14170 }, { "grad_norm": 0.5600734353065491, "learning_rate": 9.771415857392619e-05, "loss": 0.0858, "step": 14180 }, { "grad_norm": 0.5699511170387268, "learning_rate": 9.770921367685978e-05, "loss": 0.0777, "step": 14190 }, { "grad_norm": 0.49200549721717834, "learning_rate": 9.770426356238551e-05, "loss": 0.0709, "step": 14200 }, { "grad_norm": 0.4905511140823364, "learning_rate": 9.769930823104469e-05, "loss": 0.0751, "step": 14210 }, { "grad_norm": 0.47793442010879517, "learning_rate": 9.769434768337926e-05, "loss": 0.0646, "step": 14220 }, { "grad_norm": 0.45386338233947754, "learning_rate": 9.768938191993164e-05, "loss": 0.061, "step": 14230 }, { "grad_norm": 0.49353793263435364, "learning_rate": 9.768441094124494e-05, "loss": 0.0605, "step": 14240 }, { "grad_norm": 0.4256226420402527, "learning_rate": 9.767943474786275e-05, "loss": 0.0682, "step": 14250 }, { "grad_norm": 0.35099565982818604, "learning_rate": 9.767445334032923e-05, "loss": 0.0618, "step": 14260 }, { "grad_norm": 0.42681393027305603, "learning_rate": 9.766946671918919e-05, "loss": 0.0685, "step": 14270 }, { "grad_norm": 0.5689528584480286, "learning_rate": 9.766447488498796e-05, "loss": 0.0603, "step": 14280 }, { "grad_norm": 0.5375341773033142, "learning_rate": 9.765947783827139e-05, "loss": 0.0647, "step": 14290 }, { "grad_norm": 0.9000968337059021, "learning_rate": 9.765447557958599e-05, "loss": 0.0769, "step": 14300 }, { "grad_norm": 0.6363470554351807, "learning_rate": 9.764946810947879e-05, "loss": 0.063, "step": 14310 }, { "grad_norm": 0.5748863220214844, "learning_rate": 9.764445542849738e-05, "loss": 0.0766, "step": 14320 }, { "grad_norm": 0.49726080894470215, "learning_rate": 9.763943753718998e-05, "loss": 0.0666, "step": 14330 }, { "grad_norm": 0.39607885479927063, "learning_rate": 9.76344144361053e-05, "loss": 0.0737, "step": 14340 }, { "grad_norm": 0.530441403388977, "learning_rate": 9.762938612579269e-05, "loss": 0.0582, "step": 14350 }, { "grad_norm": 0.4575827419757843, "learning_rate": 9.762435260680202e-05, "loss": 0.0537, "step": 14360 }, { "grad_norm": 0.4724288284778595, "learning_rate": 9.761931387968373e-05, "loss": 0.0622, "step": 14370 }, { "grad_norm": 0.33677881956100464, "learning_rate": 9.76142699449889e-05, "loss": 0.0624, "step": 14380 }, { "grad_norm": 0.45723286271095276, "learning_rate": 9.760922080326908e-05, "loss": 0.0781, "step": 14390 }, { "grad_norm": 0.9069262146949768, "learning_rate": 9.760416645507644e-05, "loss": 0.0621, "step": 14400 }, { "grad_norm": 0.5313320755958557, "learning_rate": 9.759910690096375e-05, "loss": 0.0664, "step": 14410 }, { "grad_norm": 0.42108917236328125, "learning_rate": 9.759404214148429e-05, "loss": 0.0665, "step": 14420 }, { "grad_norm": 0.5658096671104431, "learning_rate": 9.758897217719191e-05, "loss": 0.0755, "step": 14430 }, { "grad_norm": 0.5782626271247864, "learning_rate": 9.758389700864113e-05, "loss": 0.0634, "step": 14440 }, { "grad_norm": 0.5939950942993164, "learning_rate": 9.757881663638688e-05, "loss": 0.0726, "step": 14450 }, { "grad_norm": 0.5222119688987732, "learning_rate": 9.757373106098478e-05, "loss": 0.0711, "step": 14460 }, { "grad_norm": 0.45096611976623535, "learning_rate": 9.756864028299097e-05, "loss": 0.087, "step": 14470 }, { "grad_norm": 0.6924872994422913, "learning_rate": 9.75635443029622e-05, "loss": 0.0673, "step": 14480 }, { "grad_norm": 0.7884241938591003, "learning_rate": 9.755844312145572e-05, "loss": 0.0781, "step": 14490 }, { "grad_norm": 0.5945374369621277, "learning_rate": 9.755333673902941e-05, "loss": 0.0825, "step": 14500 }, { "grad_norm": 0.4254246950149536, "learning_rate": 9.75482251562417e-05, "loss": 0.0704, "step": 14510 }, { "grad_norm": 0.6052230000495911, "learning_rate": 9.754310837365155e-05, "loss": 0.0597, "step": 14520 }, { "grad_norm": 0.5188802480697632, "learning_rate": 9.753798639181856e-05, "loss": 0.0786, "step": 14530 }, { "grad_norm": 0.49321112036705017, "learning_rate": 9.753285921130286e-05, "loss": 0.0574, "step": 14540 }, { "grad_norm": 0.6275296807289124, "learning_rate": 9.752772683266512e-05, "loss": 0.0633, "step": 14550 }, { "grad_norm": 0.47505107522010803, "learning_rate": 9.752258925646665e-05, "loss": 0.058, "step": 14560 }, { "grad_norm": 0.6074385046958923, "learning_rate": 9.751744648326926e-05, "loss": 0.0701, "step": 14570 }, { "grad_norm": 0.5511032938957214, "learning_rate": 9.751229851363536e-05, "loss": 0.0676, "step": 14580 }, { "grad_norm": 0.6039498448371887, "learning_rate": 9.750714534812793e-05, "loss": 0.0828, "step": 14590 }, { "grad_norm": 0.4910600781440735, "learning_rate": 9.750198698731053e-05, "loss": 0.0718, "step": 14600 }, { "grad_norm": 0.5692631602287292, "learning_rate": 9.749682343174722e-05, "loss": 0.0586, "step": 14610 }, { "grad_norm": 0.41420018672943115, "learning_rate": 9.749165468200272e-05, "loss": 0.0609, "step": 14620 }, { "grad_norm": 0.6288201212882996, "learning_rate": 9.748648073864229e-05, "loss": 0.0769, "step": 14630 }, { "grad_norm": 0.35637035965919495, "learning_rate": 9.748130160223168e-05, "loss": 0.0724, "step": 14640 }, { "grad_norm": 0.5064817667007446, "learning_rate": 9.747611727333734e-05, "loss": 0.08, "step": 14650 }, { "grad_norm": 0.460708886384964, "learning_rate": 9.74709277525262e-05, "loss": 0.0711, "step": 14660 }, { "grad_norm": 0.5748423933982849, "learning_rate": 9.746573304036576e-05, "loss": 0.0671, "step": 14670 }, { "grad_norm": 0.424415647983551, "learning_rate": 9.746053313742412e-05, "loss": 0.0664, "step": 14680 }, { "grad_norm": 0.5766632556915283, "learning_rate": 9.745532804426994e-05, "loss": 0.0654, "step": 14690 }, { "grad_norm": 0.3092271685600281, "learning_rate": 9.745011776147242e-05, "loss": 0.0633, "step": 14700 }, { "grad_norm": 0.6697087287902832, "learning_rate": 9.744490228960138e-05, "loss": 0.0791, "step": 14710 }, { "grad_norm": 0.5717688202857971, "learning_rate": 9.743968162922713e-05, "loss": 0.0681, "step": 14720 }, { "grad_norm": 0.43901440501213074, "learning_rate": 9.743445578092064e-05, "loss": 0.0739, "step": 14730 }, { "grad_norm": 0.8077151775360107, "learning_rate": 9.742922474525338e-05, "loss": 0.0692, "step": 14740 }, { "grad_norm": 0.7857585549354553, "learning_rate": 9.742398852279741e-05, "loss": 0.0676, "step": 14750 }, { "grad_norm": 0.508569061756134, "learning_rate": 9.741874711412535e-05, "loss": 0.0672, "step": 14760 }, { "grad_norm": 0.4247857928276062, "learning_rate": 9.741350051981042e-05, "loss": 0.0644, "step": 14770 }, { "grad_norm": 0.6478984951972961, "learning_rate": 9.740824874042633e-05, "loss": 0.0588, "step": 14780 }, { "grad_norm": 0.33982226252555847, "learning_rate": 9.740299177654746e-05, "loss": 0.049, "step": 14790 }, { "grad_norm": 0.48487719893455505, "learning_rate": 9.739772962874867e-05, "loss": 0.0576, "step": 14800 }, { "grad_norm": 0.4571469724178314, "learning_rate": 9.739246229760541e-05, "loss": 0.0603, "step": 14810 }, { "grad_norm": 0.6086424589157104, "learning_rate": 9.738718978369376e-05, "loss": 0.0708, "step": 14820 }, { "grad_norm": 0.6537250876426697, "learning_rate": 9.738191208759025e-05, "loss": 0.0818, "step": 14830 }, { "grad_norm": 0.49533984065055847, "learning_rate": 9.73766292098721e-05, "loss": 0.0767, "step": 14840 }, { "grad_norm": 0.47976621985435486, "learning_rate": 9.737134115111699e-05, "loss": 0.079, "step": 14850 }, { "grad_norm": 0.6199462413787842, "learning_rate": 9.736604791190323e-05, "loss": 0.0617, "step": 14860 }, { "grad_norm": 0.39450111985206604, "learning_rate": 9.73607494928097e-05, "loss": 0.062, "step": 14870 }, { "grad_norm": 0.45209673047065735, "learning_rate": 9.735544589441581e-05, "loss": 0.0636, "step": 14880 }, { "grad_norm": 0.46607068181037903, "learning_rate": 9.735013711730154e-05, "loss": 0.0617, "step": 14890 }, { "grad_norm": 0.48987075686454773, "learning_rate": 9.734482316204747e-05, "loss": 0.0499, "step": 14900 }, { "grad_norm": 0.4969205856323242, "learning_rate": 9.733950402923473e-05, "loss": 0.0647, "step": 14910 }, { "grad_norm": 0.5465242862701416, "learning_rate": 9.7334179719445e-05, "loss": 0.0507, "step": 14920 }, { "grad_norm": 0.5246052742004395, "learning_rate": 9.732885023326053e-05, "loss": 0.0541, "step": 14930 }, { "grad_norm": 0.5849772691726685, "learning_rate": 9.732351557126418e-05, "loss": 0.068, "step": 14940 }, { "grad_norm": 0.626259446144104, "learning_rate": 9.731817573403929e-05, "loss": 0.0738, "step": 14950 }, { "grad_norm": 0.5273138284683228, "learning_rate": 9.731283072216985e-05, "loss": 0.0577, "step": 14960 }, { "grad_norm": 0.6684461236000061, "learning_rate": 9.730748053624039e-05, "loss": 0.0568, "step": 14970 }, { "grad_norm": 0.43480733036994934, "learning_rate": 9.730212517683598e-05, "loss": 0.0587, "step": 14980 }, { "grad_norm": 0.414713978767395, "learning_rate": 9.729676464454228e-05, "loss": 0.047, "step": 14990 }, { "grad_norm": 0.4042268395423889, "learning_rate": 9.72913989399455e-05, "loss": 0.0578, "step": 15000 }, { "grad_norm": 0.5189514756202698, "learning_rate": 9.728602806363242e-05, "loss": 0.0531, "step": 15010 }, { "grad_norm": 0.5207912921905518, "learning_rate": 9.728065201619043e-05, "loss": 0.0666, "step": 15020 }, { "grad_norm": 0.46483662724494934, "learning_rate": 9.727527079820742e-05, "loss": 0.0548, "step": 15030 }, { "grad_norm": 0.3647286593914032, "learning_rate": 9.726988441027186e-05, "loss": 0.0638, "step": 15040 }, { "grad_norm": 0.5733749270439148, "learning_rate": 9.726449285297281e-05, "loss": 0.0753, "step": 15050 }, { "grad_norm": 0.4163213074207306, "learning_rate": 9.72590961268999e-05, "loss": 0.0627, "step": 15060 }, { "grad_norm": 0.46134331822395325, "learning_rate": 9.725369423264328e-05, "loss": 0.0636, "step": 15070 }, { "grad_norm": 0.4447627365589142, "learning_rate": 9.72482871707937e-05, "loss": 0.0564, "step": 15080 }, { "grad_norm": 0.5280786156654358, "learning_rate": 9.724287494194247e-05, "loss": 0.0701, "step": 15090 }, { "grad_norm": 0.6419306397438049, "learning_rate": 9.723745754668147e-05, "loss": 0.0611, "step": 15100 }, { "grad_norm": 0.45905572175979614, "learning_rate": 9.723203498560313e-05, "loss": 0.061, "step": 15110 }, { "grad_norm": 0.33454522490501404, "learning_rate": 9.722660725930046e-05, "loss": 0.052, "step": 15120 }, { "grad_norm": 0.40616774559020996, "learning_rate": 9.722117436836702e-05, "loss": 0.0503, "step": 15130 }, { "grad_norm": 0.5263128280639648, "learning_rate": 9.721573631339696e-05, "loss": 0.0412, "step": 15140 }, { "grad_norm": 0.49973055720329285, "learning_rate": 9.721029309498494e-05, "loss": 0.0698, "step": 15150 }, { "grad_norm": 0.49245572090148926, "learning_rate": 9.720484471372627e-05, "loss": 0.0575, "step": 15160 }, { "grad_norm": 0.5904422402381897, "learning_rate": 9.719939117021673e-05, "loss": 0.0643, "step": 15170 }, { "grad_norm": 0.5249890089035034, "learning_rate": 9.719393246505275e-05, "loss": 0.0586, "step": 15180 }, { "grad_norm": 0.5749672651290894, "learning_rate": 9.718846859883128e-05, "loss": 0.0618, "step": 15190 }, { "grad_norm": 0.5862212777137756, "learning_rate": 9.718299957214982e-05, "loss": 0.0605, "step": 15200 }, { "grad_norm": 0.5593486428260803, "learning_rate": 9.717752538560646e-05, "loss": 0.0591, "step": 15210 }, { "grad_norm": 0.5890950560569763, "learning_rate": 9.717204603979986e-05, "loss": 0.0576, "step": 15220 }, { "grad_norm": 0.4671754240989685, "learning_rate": 9.716656153532922e-05, "loss": 0.0611, "step": 15230 }, { "grad_norm": 0.3936401307582855, "learning_rate": 9.716107187279434e-05, "loss": 0.0565, "step": 15240 }, { "grad_norm": 0.5559470057487488, "learning_rate": 9.715557705279555e-05, "loss": 0.0739, "step": 15250 }, { "grad_norm": 0.6405946016311646, "learning_rate": 9.715007707593372e-05, "loss": 0.0743, "step": 15260 }, { "grad_norm": 0.5666772127151489, "learning_rate": 9.714457194281036e-05, "loss": 0.0584, "step": 15270 }, { "grad_norm": 0.37420913577079773, "learning_rate": 9.713906165402751e-05, "loss": 0.0502, "step": 15280 }, { "grad_norm": 0.6891595125198364, "learning_rate": 9.713354621018774e-05, "loss": 0.0713, "step": 15290 }, { "grad_norm": 0.7179263234138489, "learning_rate": 9.712802561189422e-05, "loss": 0.077, "step": 15300 }, { "grad_norm": 0.4890364110469818, "learning_rate": 9.712249985975069e-05, "loss": 0.066, "step": 15310 }, { "grad_norm": 0.3502349555492401, "learning_rate": 9.71169689543614e-05, "loss": 0.055, "step": 15320 }, { "grad_norm": 0.5238310694694519, "learning_rate": 9.711143289633123e-05, "loss": 0.0536, "step": 15330 }, { "grad_norm": 0.5148079991340637, "learning_rate": 9.710589168626561e-05, "loss": 0.069, "step": 15340 }, { "grad_norm": 0.5885483622550964, "learning_rate": 9.710034532477048e-05, "loss": 0.07, "step": 15350 }, { "grad_norm": 0.4604142904281616, "learning_rate": 9.709479381245239e-05, "loss": 0.0663, "step": 15360 }, { "grad_norm": 0.40592172741889954, "learning_rate": 9.708923714991847e-05, "loss": 0.0639, "step": 15370 }, { "grad_norm": 0.4161047041416168, "learning_rate": 9.708367533777638e-05, "loss": 0.0704, "step": 15380 }, { "grad_norm": 0.6188260316848755, "learning_rate": 9.707810837663431e-05, "loss": 0.064, "step": 15390 }, { "grad_norm": 0.5964534878730774, "learning_rate": 9.707253626710113e-05, "loss": 0.0619, "step": 15400 }, { "grad_norm": 0.47295045852661133, "learning_rate": 9.706695900978613e-05, "loss": 0.0601, "step": 15410 }, { "grad_norm": 0.8048691749572754, "learning_rate": 9.706137660529926e-05, "loss": 0.0688, "step": 15420 }, { "grad_norm": 0.52843177318573, "learning_rate": 9.705578905425101e-05, "loss": 0.0794, "step": 15430 }, { "grad_norm": 0.3970028758049011, "learning_rate": 9.705019635725241e-05, "loss": 0.0642, "step": 15440 }, { "grad_norm": 0.47371706366539, "learning_rate": 9.704459851491508e-05, "loss": 0.0719, "step": 15450 }, { "grad_norm": 0.49463504552841187, "learning_rate": 9.703899552785118e-05, "loss": 0.0642, "step": 15460 }, { "grad_norm": 0.4815402626991272, "learning_rate": 9.703338739667346e-05, "loss": 0.0692, "step": 15470 }, { "grad_norm": 0.4611077904701233, "learning_rate": 9.70277741219952e-05, "loss": 0.0702, "step": 15480 }, { "grad_norm": 0.8253237009048462, "learning_rate": 9.702215570443027e-05, "loss": 0.0673, "step": 15490 }, { "grad_norm": 0.5338916778564453, "learning_rate": 9.701653214459309e-05, "loss": 0.0714, "step": 15500 }, { "grad_norm": 0.3910163342952728, "learning_rate": 9.701090344309865e-05, "loss": 0.0737, "step": 15510 }, { "grad_norm": 0.5199456810951233, "learning_rate": 9.700526960056247e-05, "loss": 0.0568, "step": 15520 }, { "grad_norm": 0.33421647548675537, "learning_rate": 9.699963061760068e-05, "loss": 0.0524, "step": 15530 }, { "grad_norm": 0.4987484812736511, "learning_rate": 9.699398649482997e-05, "loss": 0.0595, "step": 15540 }, { "grad_norm": 0.3987712264060974, "learning_rate": 9.698833723286753e-05, "loss": 0.0482, "step": 15550 }, { "grad_norm": 0.4790008068084717, "learning_rate": 9.698268283233118e-05, "loss": 0.0638, "step": 15560 }, { "grad_norm": 0.39006078243255615, "learning_rate": 9.697702329383929e-05, "loss": 0.0682, "step": 15570 }, { "grad_norm": 0.4455643594264984, "learning_rate": 9.697135861801074e-05, "loss": 0.0737, "step": 15580 }, { "grad_norm": 0.6299931406974792, "learning_rate": 9.696568880546505e-05, "loss": 0.0629, "step": 15590 }, { "grad_norm": 0.4983065128326416, "learning_rate": 9.696001385682223e-05, "loss": 0.0527, "step": 15600 }, { "grad_norm": 0.4776986241340637, "learning_rate": 9.695433377270291e-05, "loss": 0.0869, "step": 15610 }, { "grad_norm": 0.6352397799491882, "learning_rate": 9.694864855372824e-05, "loss": 0.0635, "step": 15620 }, { "grad_norm": 0.5125072002410889, "learning_rate": 9.694295820051995e-05, "loss": 0.0609, "step": 15630 }, { "grad_norm": 0.49613043665885925, "learning_rate": 9.693726271370032e-05, "loss": 0.0504, "step": 15640 }, { "grad_norm": 0.48086488246917725, "learning_rate": 9.693156209389221e-05, "loss": 0.0523, "step": 15650 }, { "grad_norm": 0.44909998774528503, "learning_rate": 9.692585634171905e-05, "loss": 0.0683, "step": 15660 }, { "grad_norm": 0.5778517723083496, "learning_rate": 9.692014545780476e-05, "loss": 0.0679, "step": 15670 }, { "grad_norm": 0.4440891146659851, "learning_rate": 9.691442944277393e-05, "loss": 0.0638, "step": 15680 }, { "grad_norm": 0.5055632591247559, "learning_rate": 9.690870829725162e-05, "loss": 0.0685, "step": 15690 }, { "grad_norm": 0.8987529873847961, "learning_rate": 9.69029820218635e-05, "loss": 0.066, "step": 15700 }, { "grad_norm": 0.47741448879241943, "learning_rate": 9.689725061723579e-05, "loss": 0.0681, "step": 15710 }, { "grad_norm": 0.49236512184143066, "learning_rate": 9.689151408399527e-05, "loss": 0.0643, "step": 15720 }, { "grad_norm": 0.5237130522727966, "learning_rate": 9.688577242276924e-05, "loss": 0.0814, "step": 15730 }, { "grad_norm": 0.4103392958641052, "learning_rate": 9.688002563418566e-05, "loss": 0.0626, "step": 15740 }, { "grad_norm": 0.5047990679740906, "learning_rate": 9.687427371887293e-05, "loss": 0.0667, "step": 15750 }, { "grad_norm": 0.5387030243873596, "learning_rate": 9.686851667746012e-05, "loss": 0.0603, "step": 15760 }, { "grad_norm": 0.6982828974723816, "learning_rate": 9.686275451057677e-05, "loss": 0.0686, "step": 15770 }, { "grad_norm": 0.37987709045410156, "learning_rate": 9.685698721885308e-05, "loss": 0.0593, "step": 15780 }, { "grad_norm": 0.6230384111404419, "learning_rate": 9.68512148029197e-05, "loss": 0.0753, "step": 15790 }, { "grad_norm": 0.3923095166683197, "learning_rate": 9.684543726340791e-05, "loss": 0.0522, "step": 15800 }, { "grad_norm": 0.5263993144035339, "learning_rate": 9.683965460094952e-05, "loss": 0.0669, "step": 15810 }, { "grad_norm": 0.8090487122535706, "learning_rate": 9.683386681617694e-05, "loss": 0.0688, "step": 15820 }, { "grad_norm": 0.3612024784088135, "learning_rate": 9.68280739097231e-05, "loss": 0.0585, "step": 15830 }, { "grad_norm": 0.4162665903568268, "learning_rate": 9.682227588222148e-05, "loss": 0.0607, "step": 15840 }, { "grad_norm": 0.47654011845588684, "learning_rate": 9.681647273430618e-05, "loss": 0.0579, "step": 15850 }, { "grad_norm": 0.4620693624019623, "learning_rate": 9.681066446661182e-05, "loss": 0.0619, "step": 15860 }, { "grad_norm": 0.5930160880088806, "learning_rate": 9.680485107977357e-05, "loss": 0.0665, "step": 15870 }, { "grad_norm": 0.49024128913879395, "learning_rate": 9.679903257442716e-05, "loss": 0.0774, "step": 15880 }, { "grad_norm": 0.4950331151485443, "learning_rate": 9.679320895120891e-05, "loss": 0.0661, "step": 15890 }, { "grad_norm": 0.38914021849632263, "learning_rate": 9.67873802107557e-05, "loss": 0.0562, "step": 15900 }, { "grad_norm": 0.5436516404151917, "learning_rate": 9.67815463537049e-05, "loss": 0.0639, "step": 15910 }, { "grad_norm": 0.44065940380096436, "learning_rate": 9.677570738069457e-05, "loss": 0.0689, "step": 15920 }, { "grad_norm": 0.488126665353775, "learning_rate": 9.676986329236318e-05, "loss": 0.0597, "step": 15930 }, { "grad_norm": 0.45541810989379883, "learning_rate": 9.676401408934987e-05, "loss": 0.0522, "step": 15940 }, { "grad_norm": 0.5430143475532532, "learning_rate": 9.675815977229428e-05, "loss": 0.0549, "step": 15950 }, { "grad_norm": 0.775745689868927, "learning_rate": 9.675230034183664e-05, "loss": 0.0523, "step": 15960 }, { "grad_norm": 0.6754480004310608, "learning_rate": 9.674643579861773e-05, "loss": 0.077, "step": 15970 }, { "grad_norm": 0.692842960357666, "learning_rate": 9.674056614327886e-05, "loss": 0.0506, "step": 15980 }, { "grad_norm": 0.6590219140052795, "learning_rate": 9.673469137646198e-05, "loss": 0.078, "step": 15990 }, { "grad_norm": 0.4245206415653229, "learning_rate": 9.67288114988095e-05, "loss": 0.0543, "step": 16000 }, { "grad_norm": 1.0911564826965332, "learning_rate": 9.672292651096447e-05, "loss": 0.0492, "step": 16010 }, { "grad_norm": 0.4782007932662964, "learning_rate": 9.671703641357042e-05, "loss": 0.0731, "step": 16020 }, { "grad_norm": 0.7633199691772461, "learning_rate": 9.67111412072715e-05, "loss": 0.0682, "step": 16030 }, { "grad_norm": 0.5903179049491882, "learning_rate": 9.670524089271242e-05, "loss": 0.0605, "step": 16040 }, { "grad_norm": 0.46060165762901306, "learning_rate": 9.669933547053842e-05, "loss": 0.0626, "step": 16050 }, { "grad_norm": 0.4884874224662781, "learning_rate": 9.669342494139531e-05, "loss": 0.0577, "step": 16060 }, { "grad_norm": 0.467093288898468, "learning_rate": 9.668750930592943e-05, "loss": 0.0505, "step": 16070 }, { "grad_norm": 0.42434412240982056, "learning_rate": 9.668158856478775e-05, "loss": 0.0648, "step": 16080 }, { "grad_norm": 0.566001296043396, "learning_rate": 9.66756627186177e-05, "loss": 0.065, "step": 16090 }, { "grad_norm": 0.9567698836326599, "learning_rate": 9.666973176806737e-05, "loss": 0.066, "step": 16100 }, { "grad_norm": 0.8696244955062866, "learning_rate": 9.666379571378534e-05, "loss": 0.0721, "step": 16110 }, { "grad_norm": 0.7754588723182678, "learning_rate": 9.665785455642076e-05, "loss": 0.0737, "step": 16120 }, { "grad_norm": 0.651322066783905, "learning_rate": 9.665190829662337e-05, "loss": 0.0747, "step": 16130 }, { "grad_norm": 0.509327232837677, "learning_rate": 9.664595693504342e-05, "loss": 0.0757, "step": 16140 }, { "grad_norm": 0.5440751314163208, "learning_rate": 9.664000047233175e-05, "loss": 0.0629, "step": 16150 }, { "grad_norm": 0.515956461429596, "learning_rate": 9.663403890913976e-05, "loss": 0.061, "step": 16160 }, { "grad_norm": 0.5232966542243958, "learning_rate": 9.662807224611938e-05, "loss": 0.0618, "step": 16170 }, { "grad_norm": 0.4142526388168335, "learning_rate": 9.662210048392311e-05, "loss": 0.0571, "step": 16180 }, { "grad_norm": 0.7143855094909668, "learning_rate": 9.661612362320405e-05, "loss": 0.0704, "step": 16190 }, { "grad_norm": 0.4680284261703491, "learning_rate": 9.661014166461579e-05, "loss": 0.0583, "step": 16200 }, { "grad_norm": 0.3655322790145874, "learning_rate": 9.66041546088125e-05, "loss": 0.0721, "step": 16210 }, { "grad_norm": 0.5988216400146484, "learning_rate": 9.659816245644895e-05, "loss": 0.1059, "step": 16220 }, { "grad_norm": 0.49666136503219604, "learning_rate": 9.65921652081804e-05, "loss": 0.0767, "step": 16230 }, { "grad_norm": 0.6446865200996399, "learning_rate": 9.658616286466271e-05, "loss": 0.0723, "step": 16240 }, { "grad_norm": 0.524290144443512, "learning_rate": 9.65801554265523e-05, "loss": 0.0653, "step": 16250 }, { "grad_norm": 0.7550839185714722, "learning_rate": 9.657414289450612e-05, "loss": 0.0625, "step": 16260 }, { "grad_norm": 0.744278073310852, "learning_rate": 9.656812526918171e-05, "loss": 0.0537, "step": 16270 }, { "grad_norm": 0.49611398577690125, "learning_rate": 9.656210255123712e-05, "loss": 0.066, "step": 16280 }, { "grad_norm": 0.3925868272781372, "learning_rate": 9.6556074741331e-05, "loss": 0.0611, "step": 16290 }, { "grad_norm": 0.44819843769073486, "learning_rate": 9.655004184012256e-05, "loss": 0.0603, "step": 16300 }, { "grad_norm": 0.7326367497444153, "learning_rate": 9.654400384827152e-05, "loss": 0.0719, "step": 16310 }, { "grad_norm": 0.5178021788597107, "learning_rate": 9.653796076643818e-05, "loss": 0.0607, "step": 16320 }, { "grad_norm": 0.5047869682312012, "learning_rate": 9.653191259528344e-05, "loss": 0.0623, "step": 16330 }, { "grad_norm": 0.39463141560554504, "learning_rate": 9.65258593354687e-05, "loss": 0.0512, "step": 16340 }, { "grad_norm": 0.39428073167800903, "learning_rate": 9.651980098765591e-05, "loss": 0.0561, "step": 16350 }, { "grad_norm": 0.5445032715797424, "learning_rate": 9.651373755250765e-05, "loss": 0.0629, "step": 16360 }, { "grad_norm": 0.5957947373390198, "learning_rate": 9.650766903068697e-05, "loss": 0.0666, "step": 16370 }, { "grad_norm": 0.41457054018974304, "learning_rate": 9.650159542285753e-05, "loss": 0.0544, "step": 16380 }, { "grad_norm": 0.5238253474235535, "learning_rate": 9.649551672968353e-05, "loss": 0.067, "step": 16390 }, { "grad_norm": 0.49354028701782227, "learning_rate": 9.648943295182973e-05, "loss": 0.0546, "step": 16400 }, { "grad_norm": 0.5727413892745972, "learning_rate": 9.648334408996144e-05, "loss": 0.0545, "step": 16410 }, { "grad_norm": 0.5527509450912476, "learning_rate": 9.647725014474452e-05, "loss": 0.0579, "step": 16420 }, { "grad_norm": 0.4998782277107239, "learning_rate": 9.64711511168454e-05, "loss": 0.0703, "step": 16430 }, { "grad_norm": 0.5267683267593384, "learning_rate": 9.646504700693108e-05, "loss": 0.06, "step": 16440 }, { "grad_norm": 0.6498410105705261, "learning_rate": 9.645893781566907e-05, "loss": 0.0579, "step": 16450 }, { "grad_norm": 0.7534765005111694, "learning_rate": 9.645282354372744e-05, "loss": 0.057, "step": 16460 }, { "grad_norm": 1.1248409748077393, "learning_rate": 9.644670419177491e-05, "loss": 0.056, "step": 16470 }, { "grad_norm": 0.5870000123977661, "learning_rate": 9.644057976048062e-05, "loss": 0.0734, "step": 16480 }, { "grad_norm": 0.5936048626899719, "learning_rate": 9.643445025051435e-05, "loss": 0.0661, "step": 16490 }, { "grad_norm": 0.6110405325889587, "learning_rate": 9.642831566254641e-05, "loss": 0.0579, "step": 16500 }, { "grad_norm": 0.46253448724746704, "learning_rate": 9.642217599724769e-05, "loss": 0.0629, "step": 16510 }, { "grad_norm": 0.6538833379745483, "learning_rate": 9.64160312552896e-05, "loss": 0.0772, "step": 16520 }, { "grad_norm": 0.5285059213638306, "learning_rate": 9.64098814373441e-05, "loss": 0.0659, "step": 16530 }, { "grad_norm": 0.489730566740036, "learning_rate": 9.640372654408374e-05, "loss": 0.0699, "step": 16540 }, { "grad_norm": 0.5300391316413879, "learning_rate": 9.639756657618162e-05, "loss": 0.0621, "step": 16550 }, { "grad_norm": 0.5581291317939758, "learning_rate": 9.639140153431138e-05, "loss": 0.0739, "step": 16560 }, { "grad_norm": 0.3825441002845764, "learning_rate": 9.638523141914721e-05, "loss": 0.0685, "step": 16570 }, { "grad_norm": 0.4214017391204834, "learning_rate": 9.637905623136388e-05, "loss": 0.0664, "step": 16580 }, { "grad_norm": 0.4311758875846863, "learning_rate": 9.637287597163669e-05, "loss": 0.0652, "step": 16590 }, { "grad_norm": 0.3528863787651062, "learning_rate": 9.63666906406415e-05, "loss": 0.0454, "step": 16600 }, { "grad_norm": 0.5038178563117981, "learning_rate": 9.636050023905473e-05, "loss": 0.0678, "step": 16610 }, { "grad_norm": 0.5317156910896301, "learning_rate": 9.635430476755336e-05, "loss": 0.0646, "step": 16620 }, { "grad_norm": 0.5402805209159851, "learning_rate": 9.63481042268149e-05, "loss": 0.0623, "step": 16630 }, { "grad_norm": 0.6388304233551025, "learning_rate": 9.634189861751745e-05, "loss": 0.0732, "step": 16640 }, { "grad_norm": 0.6610687375068665, "learning_rate": 9.633568794033967e-05, "loss": 0.0635, "step": 16650 }, { "grad_norm": 0.3834313452243805, "learning_rate": 9.63294721959607e-05, "loss": 0.0545, "step": 16660 }, { "grad_norm": 0.37931498885154724, "learning_rate": 9.63232513850603e-05, "loss": 0.0772, "step": 16670 }, { "grad_norm": 0.5409212112426758, "learning_rate": 9.631702550831878e-05, "loss": 0.0571, "step": 16680 }, { "grad_norm": 0.4866147041320801, "learning_rate": 9.631079456641698e-05, "loss": 0.0637, "step": 16690 }, { "grad_norm": 0.6573119163513184, "learning_rate": 9.630455856003632e-05, "loss": 0.0685, "step": 16700 }, { "grad_norm": 0.44761013984680176, "learning_rate": 9.629831748985876e-05, "loss": 0.0597, "step": 16710 }, { "grad_norm": 0.6456621885299683, "learning_rate": 9.629207135656679e-05, "loss": 0.0845, "step": 16720 }, { "grad_norm": 0.44935545325279236, "learning_rate": 9.628582016084353e-05, "loss": 0.0691, "step": 16730 }, { "grad_norm": 0.5192015767097473, "learning_rate": 9.627956390337254e-05, "loss": 0.047, "step": 16740 }, { "grad_norm": 0.48161745071411133, "learning_rate": 9.627330258483802e-05, "loss": 0.0655, "step": 16750 }, { "grad_norm": 0.5311198830604553, "learning_rate": 9.62670362059247e-05, "loss": 0.0536, "step": 16760 }, { "grad_norm": 0.38349032402038574, "learning_rate": 9.626076476731786e-05, "loss": 0.0477, "step": 16770 }, { "grad_norm": 0.45081573724746704, "learning_rate": 9.625448826970336e-05, "loss": 0.0564, "step": 16780 }, { "grad_norm": 0.4091010093688965, "learning_rate": 9.624820671376755e-05, "loss": 0.0503, "step": 16790 }, { "grad_norm": 0.4201938807964325, "learning_rate": 9.62419201001974e-05, "loss": 0.0558, "step": 16800 }, { "grad_norm": 0.5944404602050781, "learning_rate": 9.623562842968037e-05, "loss": 0.0553, "step": 16810 }, { "grad_norm": 0.5446292757987976, "learning_rate": 9.622933170290454e-05, "loss": 0.0521, "step": 16820 }, { "grad_norm": 0.7065128087997437, "learning_rate": 9.622302992055849e-05, "loss": 0.052, "step": 16830 }, { "grad_norm": 0.4605378210544586, "learning_rate": 9.62167230833314e-05, "loss": 0.0525, "step": 16840 }, { "grad_norm": 0.38121357560157776, "learning_rate": 9.621041119191295e-05, "loss": 0.0496, "step": 16850 }, { "grad_norm": 0.5375354886054993, "learning_rate": 9.620409424699342e-05, "loss": 0.0649, "step": 16860 }, { "grad_norm": 0.5846213102340698, "learning_rate": 9.619777224926359e-05, "loss": 0.0572, "step": 16870 }, { "grad_norm": 0.49664172530174255, "learning_rate": 9.619144519941485e-05, "loss": 0.0583, "step": 16880 }, { "grad_norm": 0.5472205281257629, "learning_rate": 9.618511309813912e-05, "loss": 0.0582, "step": 16890 }, { "grad_norm": 0.616416335105896, "learning_rate": 9.617877594612886e-05, "loss": 0.0457, "step": 16900 }, { "grad_norm": 0.557905912399292, "learning_rate": 9.617243374407707e-05, "loss": 0.0497, "step": 16910 }, { "grad_norm": 0.41771334409713745, "learning_rate": 9.616608649267736e-05, "loss": 0.0572, "step": 16920 }, { "grad_norm": 0.5202561616897583, "learning_rate": 9.615973419262385e-05, "loss": 0.0629, "step": 16930 }, { "grad_norm": 0.40661197900772095, "learning_rate": 9.615337684461119e-05, "loss": 0.0556, "step": 16940 }, { "grad_norm": 0.5836089849472046, "learning_rate": 9.614701444933465e-05, "loss": 0.0541, "step": 16950 }, { "grad_norm": 0.5746919512748718, "learning_rate": 9.614064700748997e-05, "loss": 0.0604, "step": 16960 }, { "grad_norm": 0.49866971373558044, "learning_rate": 9.613427451977352e-05, "loss": 0.0592, "step": 16970 }, { "grad_norm": 0.38405540585517883, "learning_rate": 9.612789698688216e-05, "loss": 0.0646, "step": 16980 }, { "grad_norm": 0.6108902096748352, "learning_rate": 9.612151440951334e-05, "loss": 0.0668, "step": 16990 }, { "grad_norm": 0.5524612665176392, "learning_rate": 9.611512678836506e-05, "loss": 0.0512, "step": 17000 }, { "grad_norm": 0.5688943862915039, "learning_rate": 9.610873412413584e-05, "loss": 0.0674, "step": 17010 }, { "grad_norm": 0.5144000053405762, "learning_rate": 9.610233641752476e-05, "loss": 0.0713, "step": 17020 }, { "grad_norm": 0.6713067889213562, "learning_rate": 9.609593366923151e-05, "loss": 0.0559, "step": 17030 }, { "grad_norm": 0.35204747319221497, "learning_rate": 9.608952587995625e-05, "loss": 0.0605, "step": 17040 }, { "grad_norm": 0.7004467248916626, "learning_rate": 9.608311305039972e-05, "loss": 0.0659, "step": 17050 }, { "grad_norm": 0.5487527251243591, "learning_rate": 9.607669518126326e-05, "loss": 0.0639, "step": 17060 }, { "grad_norm": 0.4154554605484009, "learning_rate": 9.607027227324866e-05, "loss": 0.0556, "step": 17070 }, { "grad_norm": 0.4598422646522522, "learning_rate": 9.606384432705837e-05, "loss": 0.054, "step": 17080 }, { "grad_norm": 0.3518884778022766, "learning_rate": 9.60574113433953e-05, "loss": 0.046, "step": 17090 }, { "grad_norm": 0.6780626773834229, "learning_rate": 9.6050973322963e-05, "loss": 0.055, "step": 17100 }, { "grad_norm": 0.6057045459747314, "learning_rate": 9.604453026646547e-05, "loss": 0.0653, "step": 17110 }, { "grad_norm": 0.5288112163543701, "learning_rate": 9.603808217460735e-05, "loss": 0.0594, "step": 17120 }, { "grad_norm": 0.5261109471321106, "learning_rate": 9.603162904809377e-05, "loss": 0.0576, "step": 17130 }, { "grad_norm": 0.43568408489227295, "learning_rate": 9.602517088763045e-05, "loss": 0.0448, "step": 17140 }, { "grad_norm": 0.41939496994018555, "learning_rate": 9.601870769392365e-05, "loss": 0.0614, "step": 17150 }, { "grad_norm": 0.591866135597229, "learning_rate": 9.601223946768017e-05, "loss": 0.0607, "step": 17160 }, { "grad_norm": 0.32941949367523193, "learning_rate": 9.600576620960734e-05, "loss": 0.0696, "step": 17170 }, { "grad_norm": 0.6295596957206726, "learning_rate": 9.599928792041308e-05, "loss": 0.0593, "step": 17180 }, { "grad_norm": 0.4894897937774658, "learning_rate": 9.599280460080587e-05, "loss": 0.0554, "step": 17190 }, { "grad_norm": 0.4724135994911194, "learning_rate": 9.59863162514947e-05, "loss": 0.0568, "step": 17200 }, { "grad_norm": 0.4502808451652527, "learning_rate": 9.597982287318911e-05, "loss": 0.0477, "step": 17210 }, { "grad_norm": 0.5424938201904297, "learning_rate": 9.597332446659923e-05, "loss": 0.0536, "step": 17220 }, { "grad_norm": 0.49244803190231323, "learning_rate": 9.59668210324357e-05, "loss": 0.0487, "step": 17230 }, { "grad_norm": 0.3725699186325073, "learning_rate": 9.596031257140974e-05, "loss": 0.0523, "step": 17240 }, { "grad_norm": 0.4227850139141083, "learning_rate": 9.59537990842331e-05, "loss": 0.0603, "step": 17250 }, { "grad_norm": 0.61527419090271, "learning_rate": 9.594728057161806e-05, "loss": 0.0589, "step": 17260 }, { "grad_norm": 0.34257790446281433, "learning_rate": 9.594075703427752e-05, "loss": 0.0542, "step": 17270 }, { "grad_norm": 0.3590562045574188, "learning_rate": 9.593422847292486e-05, "loss": 0.045, "step": 17280 }, { "grad_norm": 0.48943281173706055, "learning_rate": 9.592769488827402e-05, "loss": 0.0562, "step": 17290 }, { "grad_norm": 0.5050039291381836, "learning_rate": 9.592115628103952e-05, "loss": 0.0531, "step": 17300 }, { "grad_norm": 0.4891231954097748, "learning_rate": 9.591461265193643e-05, "loss": 0.0543, "step": 17310 }, { "grad_norm": 0.7886250019073486, "learning_rate": 9.590806400168032e-05, "loss": 0.0502, "step": 17320 }, { "grad_norm": 0.461408406496048, "learning_rate": 9.590151033098735e-05, "loss": 0.0599, "step": 17330 }, { "grad_norm": 0.6936858296394348, "learning_rate": 9.589495164057423e-05, "loss": 0.0596, "step": 17340 }, { "grad_norm": 0.4136667549610138, "learning_rate": 9.58883879311582e-05, "loss": 0.0557, "step": 17350 }, { "grad_norm": 0.38798415660858154, "learning_rate": 9.588181920345705e-05, "loss": 0.0606, "step": 17360 }, { "grad_norm": 0.42536625266075134, "learning_rate": 9.587524545818913e-05, "loss": 0.0562, "step": 17370 }, { "grad_norm": 0.6876450181007385, "learning_rate": 9.586866669607335e-05, "loss": 0.0555, "step": 17380 }, { "grad_norm": 0.5317545533180237, "learning_rate": 9.586208291782915e-05, "loss": 0.0609, "step": 17390 }, { "grad_norm": 0.5708737969398499, "learning_rate": 9.58554941241765e-05, "loss": 0.0582, "step": 17400 }, { "grad_norm": 0.44680821895599365, "learning_rate": 9.584890031583596e-05, "loss": 0.0438, "step": 17410 }, { "grad_norm": 0.5893821716308594, "learning_rate": 9.584230149352861e-05, "loss": 0.0564, "step": 17420 }, { "grad_norm": 0.46743589639663696, "learning_rate": 9.58356976579761e-05, "loss": 0.0544, "step": 17430 }, { "grad_norm": 0.5380210876464844, "learning_rate": 9.58290888099006e-05, "loss": 0.0744, "step": 17440 }, { "grad_norm": 0.5973815321922302, "learning_rate": 9.582247495002486e-05, "loss": 0.06, "step": 17450 }, { "grad_norm": 0.46017882227897644, "learning_rate": 9.581585607907214e-05, "loss": 0.055, "step": 17460 }, { "grad_norm": 0.3998187482357025, "learning_rate": 9.580923219776628e-05, "loss": 0.0447, "step": 17470 }, { "grad_norm": 0.3820275068283081, "learning_rate": 9.580260330683167e-05, "loss": 0.0457, "step": 17480 }, { "grad_norm": 0.6371897459030151, "learning_rate": 9.579596940699322e-05, "loss": 0.0481, "step": 17490 }, { "grad_norm": 0.28932303190231323, "learning_rate": 9.578933049897643e-05, "loss": 0.0438, "step": 17500 }, { "grad_norm": 0.3780129551887512, "learning_rate": 9.578268658350728e-05, "loss": 0.0605, "step": 17510 }, { "grad_norm": 0.5865238308906555, "learning_rate": 9.577603766131235e-05, "loss": 0.0525, "step": 17520 }, { "grad_norm": 0.3572582006454468, "learning_rate": 9.576938373311878e-05, "loss": 0.0499, "step": 17530 }, { "grad_norm": 0.5426837801933289, "learning_rate": 9.576272479965421e-05, "loss": 0.067, "step": 17540 }, { "grad_norm": 0.4315910041332245, "learning_rate": 9.575606086164687e-05, "loss": 0.0516, "step": 17550 }, { "grad_norm": 0.41710737347602844, "learning_rate": 9.57493919198255e-05, "loss": 0.0545, "step": 17560 }, { "grad_norm": 1.098666787147522, "learning_rate": 9.57427179749194e-05, "loss": 0.0577, "step": 17570 }, { "grad_norm": 0.5175349712371826, "learning_rate": 9.573603902765846e-05, "loss": 0.0539, "step": 17580 }, { "grad_norm": 0.47975045442581177, "learning_rate": 9.572935507877304e-05, "loss": 0.0659, "step": 17590 }, { "grad_norm": 0.54185950756073, "learning_rate": 9.57226661289941e-05, "loss": 0.0594, "step": 17600 }, { "grad_norm": 0.40437760949134827, "learning_rate": 9.571597217905315e-05, "loss": 0.0612, "step": 17610 }, { "grad_norm": 0.39808911085128784, "learning_rate": 9.57092732296822e-05, "loss": 0.0487, "step": 17620 }, { "grad_norm": 0.551297664642334, "learning_rate": 9.570256928161385e-05, "loss": 0.051, "step": 17630 }, { "grad_norm": 0.4640119969844818, "learning_rate": 9.569586033558126e-05, "loss": 0.0593, "step": 17640 }, { "grad_norm": 0.3348730206489563, "learning_rate": 9.568914639231807e-05, "loss": 0.0536, "step": 17650 }, { "grad_norm": 0.7946043014526367, "learning_rate": 9.568242745255852e-05, "loss": 0.0745, "step": 17660 }, { "grad_norm": 0.3725328743457794, "learning_rate": 9.567570351703739e-05, "loss": 0.0638, "step": 17670 }, { "grad_norm": 0.5565611720085144, "learning_rate": 9.566897458649001e-05, "loss": 0.0614, "step": 17680 }, { "grad_norm": 0.42570191621780396, "learning_rate": 9.566224066165221e-05, "loss": 0.0534, "step": 17690 }, { "grad_norm": 0.4624963700771332, "learning_rate": 9.565550174326043e-05, "loss": 0.0612, "step": 17700 }, { "grad_norm": 0.6752130389213562, "learning_rate": 9.564875783205162e-05, "loss": 0.0742, "step": 17710 }, { "grad_norm": 0.7593767642974854, "learning_rate": 9.564200892876328e-05, "loss": 0.0623, "step": 17720 }, { "grad_norm": 0.5668395161628723, "learning_rate": 9.563525503413348e-05, "loss": 0.0464, "step": 17730 }, { "grad_norm": 0.7737988233566284, "learning_rate": 9.562849614890079e-05, "loss": 0.0627, "step": 17740 }, { "grad_norm": 0.5730684399604797, "learning_rate": 9.562173227380436e-05, "loss": 0.0683, "step": 17750 }, { "grad_norm": 0.4357483386993408, "learning_rate": 9.561496340958389e-05, "loss": 0.0658, "step": 17760 }, { "grad_norm": 0.7904661297798157, "learning_rate": 9.560818955697959e-05, "loss": 0.0645, "step": 17770 }, { "grad_norm": 0.5644087791442871, "learning_rate": 9.560141071673228e-05, "loss": 0.0581, "step": 17780 }, { "grad_norm": 0.5032215118408203, "learning_rate": 9.559462688958323e-05, "loss": 0.0537, "step": 17790 }, { "grad_norm": 0.4478987753391266, "learning_rate": 9.558783807627434e-05, "loss": 0.0562, "step": 17800 }, { "grad_norm": 0.6434923410415649, "learning_rate": 9.558104427754801e-05, "loss": 0.0624, "step": 17810 }, { "grad_norm": 0.6241814494132996, "learning_rate": 9.557424549414722e-05, "loss": 0.0674, "step": 17820 }, { "grad_norm": 0.3435150980949402, "learning_rate": 9.556744172681546e-05, "loss": 0.0482, "step": 17830 }, { "grad_norm": 0.7999785542488098, "learning_rate": 9.556063297629677e-05, "loss": 0.0672, "step": 17840 }, { "grad_norm": 0.5029222369194031, "learning_rate": 9.555381924333578e-05, "loss": 0.0734, "step": 17850 }, { "grad_norm": 0.8770102262496948, "learning_rate": 9.554700052867758e-05, "loss": 0.0778, "step": 17860 }, { "grad_norm": 0.8175155520439148, "learning_rate": 9.554017683306789e-05, "loss": 0.0593, "step": 17870 }, { "grad_norm": 0.4502147436141968, "learning_rate": 9.553334815725294e-05, "loss": 0.0641, "step": 17880 }, { "grad_norm": 0.5152272582054138, "learning_rate": 9.552651450197949e-05, "loss": 0.0593, "step": 17890 }, { "grad_norm": 0.5766134858131409, "learning_rate": 9.551967586799486e-05, "loss": 0.0463, "step": 17900 }, { "grad_norm": 0.6077412366867065, "learning_rate": 9.551283225604692e-05, "loss": 0.0632, "step": 17910 }, { "grad_norm": 0.6079793572425842, "learning_rate": 9.550598366688406e-05, "loss": 0.0604, "step": 17920 }, { "grad_norm": 0.4866127371788025, "learning_rate": 9.549913010125526e-05, "loss": 0.0568, "step": 17930 }, { "grad_norm": 0.6879399418830872, "learning_rate": 9.549227155990999e-05, "loss": 0.0571, "step": 17940 }, { "grad_norm": 0.4346061050891876, "learning_rate": 9.548540804359828e-05, "loss": 0.06, "step": 17950 }, { "grad_norm": 0.6316641569137573, "learning_rate": 9.547853955307077e-05, "loss": 0.0666, "step": 17960 }, { "grad_norm": 0.5269069671630859, "learning_rate": 9.547166608907853e-05, "loss": 0.0614, "step": 17970 }, { "grad_norm": 0.41477906703948975, "learning_rate": 9.546478765237326e-05, "loss": 0.0629, "step": 17980 }, { "grad_norm": 0.44605880975723267, "learning_rate": 9.545790424370715e-05, "loss": 0.0457, "step": 17990 }, { "grad_norm": 0.46964171528816223, "learning_rate": 9.5451015863833e-05, "loss": 0.0651, "step": 18000 }, { "grad_norm": 0.39565974473953247, "learning_rate": 9.544412251350408e-05, "loss": 0.0693, "step": 18010 }, { "grad_norm": 0.3670148253440857, "learning_rate": 9.543722419347422e-05, "loss": 0.0602, "step": 18020 }, { "grad_norm": 0.5195263624191284, "learning_rate": 9.543032090449788e-05, "loss": 0.0699, "step": 18030 }, { "grad_norm": 0.49189791083335876, "learning_rate": 9.542341264732992e-05, "loss": 0.0644, "step": 18040 }, { "grad_norm": 0.5561041235923767, "learning_rate": 9.541649942272585e-05, "loss": 0.0658, "step": 18050 }, { "grad_norm": 0.4482741355895996, "learning_rate": 9.54095812314417e-05, "loss": 0.0608, "step": 18060 }, { "grad_norm": 0.6667436361312866, "learning_rate": 9.540265807423401e-05, "loss": 0.0617, "step": 18070 }, { "grad_norm": 0.8055638670921326, "learning_rate": 9.53957299518599e-05, "loss": 0.0571, "step": 18080 }, { "grad_norm": 0.34165462851524353, "learning_rate": 9.5388796865077e-05, "loss": 0.0506, "step": 18090 }, { "grad_norm": 0.4581902325153351, "learning_rate": 9.538185881464353e-05, "loss": 0.0606, "step": 18100 }, { "grad_norm": 0.5015438795089722, "learning_rate": 9.537491580131821e-05, "loss": 0.0519, "step": 18110 }, { "grad_norm": 0.6313819289207458, "learning_rate": 9.53679678258603e-05, "loss": 0.058, "step": 18120 }, { "grad_norm": 0.6888706684112549, "learning_rate": 9.536101488902966e-05, "loss": 0.061, "step": 18130 }, { "grad_norm": 0.3889033794403076, "learning_rate": 9.535405699158663e-05, "loss": 0.0514, "step": 18140 }, { "grad_norm": 0.4457211494445801, "learning_rate": 9.53470941342921e-05, "loss": 0.0603, "step": 18150 }, { "grad_norm": 0.3342069387435913, "learning_rate": 9.534012631790756e-05, "loss": 0.0597, "step": 18160 }, { "grad_norm": 0.3890748620033264, "learning_rate": 9.533315354319494e-05, "loss": 0.0633, "step": 18170 }, { "grad_norm": 0.4117501974105835, "learning_rate": 9.532617581091682e-05, "loss": 0.0511, "step": 18180 }, { "grad_norm": 0.5256150364875793, "learning_rate": 9.531919312183629e-05, "loss": 0.0595, "step": 18190 }, { "grad_norm": 0.450804203748703, "learning_rate": 9.531220547671688e-05, "loss": 0.0529, "step": 18200 }, { "grad_norm": 0.7829875349998474, "learning_rate": 9.530521287632285e-05, "loss": 0.053, "step": 18210 }, { "grad_norm": 0.4339303970336914, "learning_rate": 9.529821532141884e-05, "loss": 0.0495, "step": 18220 }, { "grad_norm": 0.300421804189682, "learning_rate": 9.52912128127701e-05, "loss": 0.0453, "step": 18230 }, { "grad_norm": 0.5763547420501709, "learning_rate": 9.528420535114244e-05, "loss": 0.0555, "step": 18240 }, { "grad_norm": 0.533665120601654, "learning_rate": 9.527719293730215e-05, "loss": 0.0541, "step": 18250 }, { "grad_norm": 0.5985175967216492, "learning_rate": 9.527017557201611e-05, "loss": 0.0563, "step": 18260 }, { "grad_norm": 0.4453882575035095, "learning_rate": 9.526315325605176e-05, "loss": 0.0565, "step": 18270 }, { "grad_norm": 0.4191587269306183, "learning_rate": 9.525612599017699e-05, "loss": 0.0462, "step": 18280 }, { "grad_norm": 0.37818443775177, "learning_rate": 9.524909377516033e-05, "loss": 0.0524, "step": 18290 }, { "grad_norm": 0.5281511545181274, "learning_rate": 9.524205661177081e-05, "loss": 0.0432, "step": 18300 }, { "grad_norm": 0.4105243384838104, "learning_rate": 9.523501450077801e-05, "loss": 0.0608, "step": 18310 }, { "grad_norm": 0.5429933667182922, "learning_rate": 9.522796744295202e-05, "loss": 0.0587, "step": 18320 }, { "grad_norm": 0.3577355146408081, "learning_rate": 9.522091543906352e-05, "loss": 0.0521, "step": 18330 }, { "grad_norm": 0.7839014530181885, "learning_rate": 9.521385848988369e-05, "loss": 0.0594, "step": 18340 }, { "grad_norm": 0.5361480712890625, "learning_rate": 9.520679659618428e-05, "loss": 0.0652, "step": 18350 }, { "grad_norm": 0.5128698945045471, "learning_rate": 9.519972975873754e-05, "loss": 0.0486, "step": 18360 }, { "grad_norm": 0.6037101149559021, "learning_rate": 9.519265797831633e-05, "loss": 0.05, "step": 18370 }, { "grad_norm": 0.40129920840263367, "learning_rate": 9.518558125569399e-05, "loss": 0.0443, "step": 18380 }, { "grad_norm": 0.8011060357093811, "learning_rate": 9.517849959164442e-05, "loss": 0.0505, "step": 18390 }, { "grad_norm": 0.38111525774002075, "learning_rate": 9.517141298694205e-05, "loss": 0.0605, "step": 18400 }, { "grad_norm": 0.5940389037132263, "learning_rate": 9.516432144236188e-05, "loss": 0.0691, "step": 18410 }, { "grad_norm": 0.4367603361606598, "learning_rate": 9.515722495867941e-05, "loss": 0.0631, "step": 18420 }, { "grad_norm": 0.4260905086994171, "learning_rate": 9.515012353667072e-05, "loss": 0.0578, "step": 18430 }, { "grad_norm": 0.44263753294944763, "learning_rate": 9.51430171771124e-05, "loss": 0.0475, "step": 18440 }, { "grad_norm": 0.43477240204811096, "learning_rate": 9.513590588078159e-05, "loss": 0.0721, "step": 18450 }, { "grad_norm": 0.7254799008369446, "learning_rate": 9.512878964845597e-05, "loss": 0.0604, "step": 18460 }, { "grad_norm": 0.5831712484359741, "learning_rate": 9.512166848091377e-05, "loss": 0.0588, "step": 18470 }, { "grad_norm": 0.4535139203071594, "learning_rate": 9.511454237893376e-05, "loss": 0.0582, "step": 18480 }, { "grad_norm": 0.46276775002479553, "learning_rate": 9.51074113432952e-05, "loss": 0.052, "step": 18490 }, { "grad_norm": 0.5095980167388916, "learning_rate": 9.510027537477797e-05, "loss": 0.0722, "step": 18500 }, { "grad_norm": 0.5851815342903137, "learning_rate": 9.509313447416242e-05, "loss": 0.0562, "step": 18510 }, { "grad_norm": 0.54465651512146, "learning_rate": 9.508598864222949e-05, "loss": 0.0516, "step": 18520 }, { "grad_norm": 0.43857651948928833, "learning_rate": 9.507883787976062e-05, "loss": 0.0579, "step": 18530 }, { "grad_norm": 0.3243032991886139, "learning_rate": 9.507168218753781e-05, "loss": 0.0548, "step": 18540 }, { "grad_norm": 0.453378289937973, "learning_rate": 9.506452156634362e-05, "loss": 0.0563, "step": 18550 }, { "grad_norm": 0.6192528605461121, "learning_rate": 9.505735601696109e-05, "loss": 0.0684, "step": 18560 }, { "grad_norm": 0.5487604737281799, "learning_rate": 9.505018554017385e-05, "loss": 0.0634, "step": 18570 }, { "grad_norm": 0.6359131336212158, "learning_rate": 9.504301013676604e-05, "loss": 0.0562, "step": 18580 }, { "grad_norm": 0.3434419333934784, "learning_rate": 9.503582980752238e-05, "loss": 0.0531, "step": 18590 }, { "grad_norm": 0.646706759929657, "learning_rate": 9.502864455322809e-05, "loss": 0.0718, "step": 18600 }, { "grad_norm": 0.643419086933136, "learning_rate": 9.502145437466891e-05, "loss": 0.0544, "step": 18610 }, { "grad_norm": 0.5972371101379395, "learning_rate": 9.501425927263116e-05, "loss": 0.0652, "step": 18620 }, { "grad_norm": 0.6330505609512329, "learning_rate": 9.500705924790172e-05, "loss": 0.0764, "step": 18630 }, { "grad_norm": 0.6028775572776794, "learning_rate": 9.499985430126794e-05, "loss": 0.0662, "step": 18640 }, { "grad_norm": 0.5279593467712402, "learning_rate": 9.499264443351775e-05, "loss": 0.0577, "step": 18650 }, { "grad_norm": 0.4251956343650818, "learning_rate": 9.498542964543961e-05, "loss": 0.0524, "step": 18660 }, { "grad_norm": 0.4060191214084625, "learning_rate": 9.497820993782252e-05, "loss": 0.0599, "step": 18670 }, { "grad_norm": 0.3972991406917572, "learning_rate": 9.497098531145601e-05, "loss": 0.0522, "step": 18680 }, { "grad_norm": 0.7553820610046387, "learning_rate": 9.496375576713017e-05, "loss": 0.0509, "step": 18690 }, { "grad_norm": 0.3645836114883423, "learning_rate": 9.49565213056356e-05, "loss": 0.0451, "step": 18700 }, { "grad_norm": 0.5231116414070129, "learning_rate": 9.494928192776342e-05, "loss": 0.0533, "step": 18710 }, { "grad_norm": 0.7050930261611938, "learning_rate": 9.494203763430538e-05, "loss": 0.0498, "step": 18720 }, { "grad_norm": 0.48336100578308105, "learning_rate": 9.493478842605366e-05, "loss": 0.0861, "step": 18730 }, { "grad_norm": 0.454843670129776, "learning_rate": 9.492753430380105e-05, "loss": 0.0606, "step": 18740 }, { "grad_norm": 0.4347245395183563, "learning_rate": 9.492027526834083e-05, "loss": 0.0496, "step": 18750 }, { "grad_norm": 0.5088160037994385, "learning_rate": 9.491301132046684e-05, "loss": 0.0565, "step": 18760 }, { "grad_norm": 0.29171645641326904, "learning_rate": 9.490574246097345e-05, "loss": 0.0684, "step": 18770 }, { "grad_norm": 0.6024371385574341, "learning_rate": 9.48984686906556e-05, "loss": 0.083, "step": 18780 }, { "grad_norm": 0.7470710277557373, "learning_rate": 9.489119001030871e-05, "loss": 0.0624, "step": 18790 }, { "grad_norm": 0.5350052118301392, "learning_rate": 9.488390642072878e-05, "loss": 0.0551, "step": 18800 }, { "grad_norm": 0.7640416622161865, "learning_rate": 9.48766179227123e-05, "loss": 0.054, "step": 18810 }, { "grad_norm": 0.5594049692153931, "learning_rate": 9.486932451705636e-05, "loss": 0.0713, "step": 18820 }, { "grad_norm": 0.3056856095790863, "learning_rate": 9.486202620455857e-05, "loss": 0.0555, "step": 18830 }, { "grad_norm": 0.4777296781539917, "learning_rate": 9.485472298601704e-05, "loss": 0.0539, "step": 18840 }, { "grad_norm": 0.515035092830658, "learning_rate": 9.484741486223043e-05, "loss": 0.0554, "step": 18850 }, { "grad_norm": 0.46483418345451355, "learning_rate": 9.484010183399797e-05, "loss": 0.0566, "step": 18860 }, { "grad_norm": 0.370230108499527, "learning_rate": 9.483278390211938e-05, "loss": 0.0633, "step": 18870 }, { "grad_norm": 1.1101012229919434, "learning_rate": 9.482546106739496e-05, "loss": 0.0669, "step": 18880 }, { "grad_norm": 0.6531482934951782, "learning_rate": 9.48181333306255e-05, "loss": 0.0594, "step": 18890 }, { "grad_norm": 0.4987500309944153, "learning_rate": 9.481080069261237e-05, "loss": 0.0776, "step": 18900 }, { "grad_norm": 0.5488490462303162, "learning_rate": 9.480346315415745e-05, "loss": 0.0639, "step": 18910 }, { "grad_norm": 0.4280111491680145, "learning_rate": 9.479612071606314e-05, "loss": 0.052, "step": 18920 }, { "grad_norm": 0.4268420934677124, "learning_rate": 9.478877337913244e-05, "loss": 0.0548, "step": 18930 }, { "grad_norm": 0.5606185793876648, "learning_rate": 9.478142114416881e-05, "loss": 0.0482, "step": 18940 }, { "grad_norm": 0.41155338287353516, "learning_rate": 9.47740640119763e-05, "loss": 0.0459, "step": 18950 }, { "grad_norm": 0.4573725163936615, "learning_rate": 9.476670198335947e-05, "loss": 0.0534, "step": 18960 }, { "grad_norm": 0.432553231716156, "learning_rate": 9.47593350591234e-05, "loss": 0.0522, "step": 18970 }, { "grad_norm": 0.47731417417526245, "learning_rate": 9.475196324007376e-05, "loss": 0.0518, "step": 18980 }, { "grad_norm": 0.5182185173034668, "learning_rate": 9.474458652701669e-05, "loss": 0.0509, "step": 18990 }, { "grad_norm": 0.43803882598876953, "learning_rate": 9.473720492075892e-05, "loss": 0.0581, "step": 19000 }, { "grad_norm": 0.6722908020019531, "learning_rate": 9.472981842210768e-05, "loss": 0.0528, "step": 19010 }, { "grad_norm": 0.44999516010284424, "learning_rate": 9.472242703187074e-05, "loss": 0.0554, "step": 19020 }, { "grad_norm": 0.46993112564086914, "learning_rate": 9.471503075085643e-05, "loss": 0.0579, "step": 19030 }, { "grad_norm": 0.4959166944026947, "learning_rate": 9.470762957987359e-05, "loss": 0.0589, "step": 19040 }, { "grad_norm": 0.42373308539390564, "learning_rate": 9.470022351973158e-05, "loss": 0.0617, "step": 19050 }, { "grad_norm": 0.6909779906272888, "learning_rate": 9.469281257124034e-05, "loss": 0.0613, "step": 19060 }, { "grad_norm": 0.283770889043808, "learning_rate": 9.46853967352103e-05, "loss": 0.0592, "step": 19070 }, { "grad_norm": 0.6125390529632568, "learning_rate": 9.467797601245246e-05, "loss": 0.0522, "step": 19080 }, { "grad_norm": 0.39175736904144287, "learning_rate": 9.467055040377834e-05, "loss": 0.0669, "step": 19090 }, { "grad_norm": 0.7035505771636963, "learning_rate": 9.466311990999999e-05, "loss": 0.0668, "step": 19100 }, { "grad_norm": 0.5114737749099731, "learning_rate": 9.465568453193e-05, "loss": 0.0557, "step": 19110 }, { "grad_norm": 0.40236836671829224, "learning_rate": 9.464824427038148e-05, "loss": 0.0449, "step": 19120 }, { "grad_norm": 0.5761513710021973, "learning_rate": 9.46407991261681e-05, "loss": 0.0622, "step": 19130 }, { "grad_norm": 0.710936963558197, "learning_rate": 9.463334910010404e-05, "loss": 0.0434, "step": 19140 }, { "grad_norm": 0.6365504264831543, "learning_rate": 9.462589419300403e-05, "loss": 0.0454, "step": 19150 }, { "grad_norm": 0.4376883804798126, "learning_rate": 9.461843440568333e-05, "loss": 0.0435, "step": 19160 }, { "grad_norm": 0.39708852767944336, "learning_rate": 9.461096973895773e-05, "loss": 0.0444, "step": 19170 }, { "grad_norm": 0.5251458883285522, "learning_rate": 9.460350019364355e-05, "loss": 0.0441, "step": 19180 }, { "grad_norm": 0.5385492444038391, "learning_rate": 9.459602577055764e-05, "loss": 0.0711, "step": 19190 }, { "grad_norm": 0.5127158761024475, "learning_rate": 9.45885464705174e-05, "loss": 0.0569, "step": 19200 }, { "grad_norm": 0.5317526459693909, "learning_rate": 9.458106229434076e-05, "loss": 0.0501, "step": 19210 }, { "grad_norm": 0.4156549572944641, "learning_rate": 9.457357324284617e-05, "loss": 0.0565, "step": 19220 }, { "grad_norm": 0.5545117855072021, "learning_rate": 9.456607931685262e-05, "loss": 0.0646, "step": 19230 }, { "grad_norm": 0.7075605392456055, "learning_rate": 9.455858051717965e-05, "loss": 0.0574, "step": 19240 }, { "grad_norm": 0.38685470819473267, "learning_rate": 9.45510768446473e-05, "loss": 0.0485, "step": 19250 }, { "grad_norm": 0.7506507635116577, "learning_rate": 9.454356830007618e-05, "loss": 0.0468, "step": 19260 }, { "grad_norm": 1.0178924798965454, "learning_rate": 9.45360548842874e-05, "loss": 0.0455, "step": 19270 }, { "grad_norm": 0.44286781549453735, "learning_rate": 9.452853659810261e-05, "loss": 0.0574, "step": 19280 }, { "grad_norm": 0.654329776763916, "learning_rate": 9.452101344234401e-05, "loss": 0.0599, "step": 19290 }, { "grad_norm": 0.520972490310669, "learning_rate": 9.451348541783431e-05, "loss": 0.0661, "step": 19300 }, { "grad_norm": 0.47179967164993286, "learning_rate": 9.450595252539678e-05, "loss": 0.0612, "step": 19310 }, { "grad_norm": 0.35711199045181274, "learning_rate": 9.449841476585518e-05, "loss": 0.0624, "step": 19320 }, { "grad_norm": 0.5294210314750671, "learning_rate": 9.449087214003384e-05, "loss": 0.0645, "step": 19330 }, { "grad_norm": 0.638617217540741, "learning_rate": 9.448332464875765e-05, "loss": 0.0773, "step": 19340 }, { "grad_norm": 0.4390103220939636, "learning_rate": 9.447577229285192e-05, "loss": 0.0622, "step": 19350 }, { "grad_norm": 0.44352057576179504, "learning_rate": 9.446821507314261e-05, "loss": 0.0521, "step": 19360 }, { "grad_norm": 0.375508576631546, "learning_rate": 9.446065299045617e-05, "loss": 0.0603, "step": 19370 }, { "grad_norm": 0.4548635482788086, "learning_rate": 9.445308604561955e-05, "loss": 0.0572, "step": 19380 }, { "grad_norm": 0.38278862833976746, "learning_rate": 9.444551423946028e-05, "loss": 0.062, "step": 19390 }, { "grad_norm": 0.472292423248291, "learning_rate": 9.443793757280638e-05, "loss": 0.0623, "step": 19400 }, { "grad_norm": 0.3636406660079956, "learning_rate": 9.443035604648646e-05, "loss": 0.0526, "step": 19410 }, { "grad_norm": 0.5384407639503479, "learning_rate": 9.44227696613296e-05, "loss": 0.0489, "step": 19420 }, { "grad_norm": 0.5064061284065247, "learning_rate": 9.441517841816542e-05, "loss": 0.0727, "step": 19430 }, { "grad_norm": 0.4595544636249542, "learning_rate": 9.440758231782413e-05, "loss": 0.0628, "step": 19440 }, { "grad_norm": 0.3942103683948517, "learning_rate": 9.439998136113639e-05, "loss": 0.0532, "step": 19450 }, { "grad_norm": 0.7301158905029297, "learning_rate": 9.439237554893344e-05, "loss": 0.0687, "step": 19460 }, { "grad_norm": 0.6318449974060059, "learning_rate": 9.438476488204705e-05, "loss": 0.0691, "step": 19470 }, { "grad_norm": 0.5775445699691772, "learning_rate": 9.43771493613095e-05, "loss": 0.053, "step": 19480 }, { "grad_norm": 0.6516630053520203, "learning_rate": 9.436952898755362e-05, "loss": 0.0785, "step": 19490 }, { "grad_norm": 0.5347544550895691, "learning_rate": 9.436190376161276e-05, "loss": 0.0564, "step": 19500 }, { "grad_norm": 0.6758641004562378, "learning_rate": 9.43542736843208e-05, "loss": 0.0731, "step": 19510 }, { "grad_norm": 0.3829323351383209, "learning_rate": 9.434663875651216e-05, "loss": 0.0447, "step": 19520 }, { "grad_norm": 0.509040117263794, "learning_rate": 9.433899897902177e-05, "loss": 0.0522, "step": 19530 }, { "grad_norm": 0.44475945830345154, "learning_rate": 9.433135435268511e-05, "loss": 0.0479, "step": 19540 }, { "grad_norm": 0.46189695596694946, "learning_rate": 9.432370487833819e-05, "loss": 0.0626, "step": 19550 }, { "grad_norm": 0.7815489768981934, "learning_rate": 9.431605055681756e-05, "loss": 0.0633, "step": 19560 }, { "grad_norm": 0.36428359150886536, "learning_rate": 9.430839138896026e-05, "loss": 0.0567, "step": 19570 }, { "grad_norm": 0.3421879708766937, "learning_rate": 9.43007273756039e-05, "loss": 0.0615, "step": 19580 }, { "grad_norm": 0.5070589780807495, "learning_rate": 9.429305851758658e-05, "loss": 0.0572, "step": 19590 }, { "grad_norm": 0.5866042375564575, "learning_rate": 9.428538481574699e-05, "loss": 0.0639, "step": 19600 }, { "grad_norm": 0.4565601944923401, "learning_rate": 9.42777062709243e-05, "loss": 0.0588, "step": 19610 }, { "grad_norm": 0.5931985378265381, "learning_rate": 9.427002288395821e-05, "loss": 0.0459, "step": 19620 }, { "grad_norm": 0.4487312138080597, "learning_rate": 9.426233465568898e-05, "loss": 0.0597, "step": 19630 }, { "grad_norm": 0.8921764492988586, "learning_rate": 9.42546415869574e-05, "loss": 0.0556, "step": 19640 }, { "grad_norm": 0.5196709036827087, "learning_rate": 9.424694367860473e-05, "loss": 0.0562, "step": 19650 }, { "grad_norm": 0.36224013566970825, "learning_rate": 9.423924093147284e-05, "loss": 0.0484, "step": 19660 }, { "grad_norm": 0.6987389326095581, "learning_rate": 9.423153334640407e-05, "loss": 0.0466, "step": 19670 }, { "grad_norm": 0.4513828456401825, "learning_rate": 9.42238209242413e-05, "loss": 0.0458, "step": 19680 }, { "grad_norm": 0.9941230416297913, "learning_rate": 9.421610366582798e-05, "loss": 0.051, "step": 19690 }, { "grad_norm": 0.5222848653793335, "learning_rate": 9.420838157200803e-05, "loss": 0.0527, "step": 19700 }, { "grad_norm": 0.47646358609199524, "learning_rate": 9.420065464362594e-05, "loss": 0.0542, "step": 19710 }, { "grad_norm": 0.5709507465362549, "learning_rate": 9.419292288152673e-05, "loss": 0.0562, "step": 19720 }, { "grad_norm": 0.4868936836719513, "learning_rate": 9.418518628655588e-05, "loss": 0.0546, "step": 19730 }, { "grad_norm": 0.5236824750900269, "learning_rate": 9.417744485955951e-05, "loss": 0.0505, "step": 19740 }, { "grad_norm": 0.8715664744377136, "learning_rate": 9.41696986013842e-05, "loss": 0.0553, "step": 19750 }, { "grad_norm": 0.48016852140426636, "learning_rate": 9.416194751287705e-05, "loss": 0.0622, "step": 19760 }, { "grad_norm": 1.0743039846420288, "learning_rate": 9.415419159488572e-05, "loss": 0.0545, "step": 19770 }, { "grad_norm": 0.5559020042419434, "learning_rate": 9.414643084825837e-05, "loss": 0.0667, "step": 19780 }, { "grad_norm": 0.43257731199264526, "learning_rate": 9.413866527384372e-05, "loss": 0.0646, "step": 19790 }, { "grad_norm": 0.4631665349006653, "learning_rate": 9.4130894872491e-05, "loss": 0.053, "step": 19800 }, { "grad_norm": 0.3767966330051422, "learning_rate": 9.412311964504998e-05, "loss": 0.054, "step": 19810 }, { "grad_norm": 0.38201576471328735, "learning_rate": 9.411533959237091e-05, "loss": 0.0609, "step": 19820 }, { "grad_norm": 0.5926958322525024, "learning_rate": 9.410755471530464e-05, "loss": 0.0396, "step": 19830 }, { "grad_norm": 0.7944653034210205, "learning_rate": 9.40997650147025e-05, "loss": 0.05, "step": 19840 }, { "grad_norm": 0.582400918006897, "learning_rate": 9.409197049141637e-05, "loss": 0.0621, "step": 19850 }, { "grad_norm": 0.4524059295654297, "learning_rate": 9.408417114629863e-05, "loss": 0.0599, "step": 19860 }, { "grad_norm": 0.4819801449775696, "learning_rate": 9.40763669802022e-05, "loss": 0.0448, "step": 19870 }, { "grad_norm": 0.6191035509109497, "learning_rate": 9.406855799398056e-05, "loss": 0.0602, "step": 19880 }, { "grad_norm": 0.4921814203262329, "learning_rate": 9.406074418848767e-05, "loss": 0.0523, "step": 19890 }, { "grad_norm": 0.5708860754966736, "learning_rate": 9.405292556457805e-05, "loss": 0.0665, "step": 19900 }, { "grad_norm": 0.5034879446029663, "learning_rate": 9.404510212310671e-05, "loss": 0.0502, "step": 19910 }, { "grad_norm": 1.401654839515686, "learning_rate": 9.403727386492924e-05, "loss": 0.0542, "step": 19920 }, { "grad_norm": 0.515886664390564, "learning_rate": 9.40294407909017e-05, "loss": 0.0559, "step": 19930 }, { "grad_norm": 0.5575571656227112, "learning_rate": 9.40216029018807e-05, "loss": 0.0595, "step": 19940 }, { "grad_norm": 0.3733218014240265, "learning_rate": 9.401376019872338e-05, "loss": 0.0533, "step": 19950 }, { "grad_norm": 0.49118340015411377, "learning_rate": 9.400591268228746e-05, "loss": 0.0582, "step": 19960 }, { "grad_norm": 0.5206121802330017, "learning_rate": 9.399806035343106e-05, "loss": 0.0633, "step": 19970 }, { "grad_norm": 0.6191509366035461, "learning_rate": 9.399020321301294e-05, "loss": 0.0646, "step": 19980 }, { "grad_norm": 0.7771605253219604, "learning_rate": 9.398234126189234e-05, "loss": 0.0499, "step": 19990 }, { "grad_norm": 0.4293709397315979, "learning_rate": 9.397447450092902e-05, "loss": 0.0564, "step": 20000 }, { "grad_norm": 0.4811517298221588, "learning_rate": 9.39666029309833e-05, "loss": 0.0541, "step": 20010 }, { "grad_norm": 0.42210617661476135, "learning_rate": 9.395872655291596e-05, "loss": 0.0641, "step": 20020 }, { "grad_norm": 0.40398380160331726, "learning_rate": 9.395084536758838e-05, "loss": 0.0507, "step": 20030 }, { "grad_norm": 0.4026342034339905, "learning_rate": 9.394295937586243e-05, "loss": 0.0671, "step": 20040 }, { "grad_norm": 0.5491907596588135, "learning_rate": 9.393506857860052e-05, "loss": 0.0533, "step": 20050 }, { "grad_norm": 0.6190213561058044, "learning_rate": 9.392717297666555e-05, "loss": 0.0449, "step": 20060 }, { "grad_norm": 0.6713501214981079, "learning_rate": 9.391927257092101e-05, "loss": 0.0582, "step": 20070 }, { "grad_norm": 0.47603639960289, "learning_rate": 9.391136736223085e-05, "loss": 0.0577, "step": 20080 }, { "grad_norm": 0.57696932554245, "learning_rate": 9.390345735145956e-05, "loss": 0.0647, "step": 20090 }, { "grad_norm": 0.4070863127708435, "learning_rate": 9.389554253947219e-05, "loss": 0.0534, "step": 20100 }, { "grad_norm": 1.8311028480529785, "learning_rate": 9.388762292713428e-05, "loss": 0.0605, "step": 20110 }, { "grad_norm": 0.3946326673030853, "learning_rate": 9.38796985153119e-05, "loss": 0.0606, "step": 20120 }, { "grad_norm": 0.4746977686882019, "learning_rate": 9.387176930487169e-05, "loss": 0.0595, "step": 20130 }, { "grad_norm": 0.39259862899780273, "learning_rate": 9.386383529668072e-05, "loss": 0.0462, "step": 20140 }, { "grad_norm": 0.5418513417243958, "learning_rate": 9.385589649160669e-05, "loss": 0.0559, "step": 20150 }, { "grad_norm": 0.7683694362640381, "learning_rate": 9.384795289051775e-05, "loss": 0.0593, "step": 20160 }, { "grad_norm": 0.5363640189170837, "learning_rate": 9.384000449428261e-05, "loss": 0.0545, "step": 20170 }, { "grad_norm": 0.4425567090511322, "learning_rate": 9.383205130377048e-05, "loss": 0.0524, "step": 20180 }, { "grad_norm": 0.4784507155418396, "learning_rate": 9.382409331985114e-05, "loss": 0.0489, "step": 20190 }, { "grad_norm": 0.5193387269973755, "learning_rate": 9.381613054339482e-05, "loss": 0.0428, "step": 20200 }, { "grad_norm": 0.4679228365421295, "learning_rate": 9.380816297527235e-05, "loss": 0.0519, "step": 20210 }, { "grad_norm": 0.5138952136039734, "learning_rate": 9.380019061635506e-05, "loss": 0.0553, "step": 20220 }, { "grad_norm": 0.4298314154148102, "learning_rate": 9.379221346751474e-05, "loss": 0.0634, "step": 20230 }, { "grad_norm": 0.47275879979133606, "learning_rate": 9.378423152962382e-05, "loss": 0.0494, "step": 20240 }, { "grad_norm": 0.34614676237106323, "learning_rate": 9.377624480355517e-05, "loss": 0.0489, "step": 20250 }, { "grad_norm": 0.6243699193000793, "learning_rate": 9.376825329018219e-05, "loss": 0.0436, "step": 20260 }, { "grad_norm": 0.38554075360298157, "learning_rate": 9.376025699037884e-05, "loss": 0.0466, "step": 20270 }, { "grad_norm": 0.42981040477752686, "learning_rate": 9.37522559050196e-05, "loss": 0.0426, "step": 20280 }, { "grad_norm": 0.3366658389568329, "learning_rate": 9.37442500349794e-05, "loss": 0.0554, "step": 20290 }, { "grad_norm": 0.40511196851730347, "learning_rate": 9.373623938113381e-05, "loss": 0.0472, "step": 20300 }, { "grad_norm": 0.3858378529548645, "learning_rate": 9.372822394435883e-05, "loss": 0.0614, "step": 20310 }, { "grad_norm": 0.5015723705291748, "learning_rate": 9.372020372553102e-05, "loss": 0.0644, "step": 20320 }, { "grad_norm": 0.47628673911094666, "learning_rate": 9.371217872552746e-05, "loss": 0.0699, "step": 20330 }, { "grad_norm": 0.5272146463394165, "learning_rate": 9.370414894522576e-05, "loss": 0.064, "step": 20340 }, { "grad_norm": 0.4326249063014984, "learning_rate": 9.369611438550406e-05, "loss": 0.0605, "step": 20350 }, { "grad_norm": 0.43426835536956787, "learning_rate": 9.368807504724095e-05, "loss": 0.0662, "step": 20360 }, { "grad_norm": 0.5686588883399963, "learning_rate": 9.368003093131565e-05, "loss": 0.0629, "step": 20370 }, { "grad_norm": 0.5262769460678101, "learning_rate": 9.367198203860785e-05, "loss": 0.0716, "step": 20380 }, { "grad_norm": 0.5444419980049133, "learning_rate": 9.366392836999774e-05, "loss": 0.0494, "step": 20390 }, { "grad_norm": 0.5575445294380188, "learning_rate": 9.365586992636607e-05, "loss": 0.0538, "step": 20400 }, { "grad_norm": 0.7223328351974487, "learning_rate": 9.364780670859412e-05, "loss": 0.0537, "step": 20410 }, { "grad_norm": 0.5716792941093445, "learning_rate": 9.363973871756364e-05, "loss": 0.0509, "step": 20420 }, { "grad_norm": 0.4320714771747589, "learning_rate": 9.363166595415696e-05, "loss": 0.0505, "step": 20430 }, { "grad_norm": 0.6124651432037354, "learning_rate": 9.362358841925686e-05, "loss": 0.0681, "step": 20440 }, { "grad_norm": 0.3119122385978699, "learning_rate": 9.361550611374674e-05, "loss": 0.0493, "step": 20450 }, { "grad_norm": 0.3185274600982666, "learning_rate": 9.360741903851043e-05, "loss": 0.0716, "step": 20460 }, { "grad_norm": 0.3493974506855011, "learning_rate": 9.359932719443236e-05, "loss": 0.0617, "step": 20470 }, { "grad_norm": 0.46068134903907776, "learning_rate": 9.35912305823974e-05, "loss": 0.0479, "step": 20480 }, { "grad_norm": 0.74717116355896, "learning_rate": 9.358312920329101e-05, "loss": 0.0522, "step": 20490 }, { "grad_norm": 0.3858349919319153, "learning_rate": 9.357502305799914e-05, "loss": 0.0427, "step": 20500 }, { "grad_norm": 0.49268677830696106, "learning_rate": 9.356691214740824e-05, "loss": 0.0522, "step": 20510 }, { "grad_norm": 0.3949563205242157, "learning_rate": 9.355879647240535e-05, "loss": 0.0471, "step": 20520 }, { "grad_norm": 0.38968420028686523, "learning_rate": 9.355067603387798e-05, "loss": 0.0555, "step": 20530 }, { "grad_norm": 0.4423165023326874, "learning_rate": 9.354255083271412e-05, "loss": 0.0453, "step": 20540 }, { "grad_norm": 0.4159826636314392, "learning_rate": 9.353442086980239e-05, "loss": 0.0513, "step": 20550 }, { "grad_norm": 0.6082319617271423, "learning_rate": 9.352628614603185e-05, "loss": 0.0523, "step": 20560 }, { "grad_norm": 0.45362046360969543, "learning_rate": 9.351814666229209e-05, "loss": 0.0424, "step": 20570 }, { "grad_norm": 0.6326914429664612, "learning_rate": 9.351000241947324e-05, "loss": 0.0488, "step": 20580 }, { "grad_norm": 0.7517845630645752, "learning_rate": 9.350185341846594e-05, "loss": 0.0577, "step": 20590 }, { "grad_norm": 0.6609624624252319, "learning_rate": 9.349369966016134e-05, "loss": 0.0587, "step": 20600 }, { "grad_norm": 0.568343460559845, "learning_rate": 9.348554114545117e-05, "loss": 0.0594, "step": 20610 }, { "grad_norm": 0.4993079602718353, "learning_rate": 9.347737787522758e-05, "loss": 0.0432, "step": 20620 }, { "grad_norm": 0.6619080305099487, "learning_rate": 9.346920985038332e-05, "loss": 0.0482, "step": 20630 }, { "grad_norm": 0.41815903782844543, "learning_rate": 9.346103707181162e-05, "loss": 0.0504, "step": 20640 }, { "grad_norm": 0.43335890769958496, "learning_rate": 9.345285954040626e-05, "loss": 0.0462, "step": 20650 }, { "grad_norm": 0.7904349565505981, "learning_rate": 9.34446772570615e-05, "loss": 0.0558, "step": 20660 }, { "grad_norm": 0.8203701972961426, "learning_rate": 9.343649022267214e-05, "loss": 0.0429, "step": 20670 }, { "grad_norm": 0.4623432755470276, "learning_rate": 9.342829843813353e-05, "loss": 0.0481, "step": 20680 }, { "grad_norm": 0.46647754311561584, "learning_rate": 9.342010190434149e-05, "loss": 0.0657, "step": 20690 }, { "grad_norm": 0.38582032918930054, "learning_rate": 9.34119006221924e-05, "loss": 0.06, "step": 20700 }, { "grad_norm": 0.4934563934803009, "learning_rate": 9.340369459258313e-05, "loss": 0.0613, "step": 20710 }, { "grad_norm": 0.46910712122917175, "learning_rate": 9.339548381641106e-05, "loss": 0.0601, "step": 20720 }, { "grad_norm": 0.46452009677886963, "learning_rate": 9.338726829457413e-05, "loss": 0.0607, "step": 20730 }, { "grad_norm": 0.5675802230834961, "learning_rate": 9.337904802797078e-05, "loss": 0.048, "step": 20740 }, { "grad_norm": 0.5029054880142212, "learning_rate": 9.337082301749993e-05, "loss": 0.0558, "step": 20750 }, { "grad_norm": 0.4820767641067505, "learning_rate": 9.336259326406109e-05, "loss": 0.0614, "step": 20760 }, { "grad_norm": 0.7513840794563293, "learning_rate": 9.335435876855427e-05, "loss": 0.0632, "step": 20770 }, { "grad_norm": 0.5867859721183777, "learning_rate": 9.334611953187994e-05, "loss": 0.0613, "step": 20780 }, { "grad_norm": 0.41021353006362915, "learning_rate": 9.333787555493914e-05, "loss": 0.0381, "step": 20790 }, { "grad_norm": 0.47146400809288025, "learning_rate": 9.332962683863345e-05, "loss": 0.0501, "step": 20800 }, { "grad_norm": 0.6076644062995911, "learning_rate": 9.332137338386489e-05, "loss": 0.049, "step": 20810 }, { "grad_norm": 0.7245010137557983, "learning_rate": 9.33131151915361e-05, "loss": 0.0527, "step": 20820 }, { "grad_norm": 0.38685330748558044, "learning_rate": 9.330485226255012e-05, "loss": 0.0453, "step": 20830 }, { "grad_norm": 0.47533470392227173, "learning_rate": 9.329658459781061e-05, "loss": 0.0513, "step": 20840 }, { "grad_norm": 0.3980832099914551, "learning_rate": 9.328831219822172e-05, "loss": 0.0523, "step": 20850 }, { "grad_norm": 0.5308375358581543, "learning_rate": 9.328003506468808e-05, "loss": 0.0515, "step": 20860 }, { "grad_norm": 0.4667956829071045, "learning_rate": 9.327175319811488e-05, "loss": 0.0564, "step": 20870 }, { "grad_norm": 0.4192279875278473, "learning_rate": 9.326346659940781e-05, "loss": 0.0571, "step": 20880 }, { "grad_norm": 0.6692657470703125, "learning_rate": 9.325517526947308e-05, "loss": 0.062, "step": 20890 }, { "grad_norm": 0.6008168458938599, "learning_rate": 9.32468792092174e-05, "loss": 0.0673, "step": 20900 }, { "grad_norm": 0.5325904488563538, "learning_rate": 9.323857841954803e-05, "loss": 0.0495, "step": 20910 }, { "grad_norm": 0.4224962592124939, "learning_rate": 9.323027290137276e-05, "loss": 0.0569, "step": 20920 }, { "grad_norm": 0.5443735718727112, "learning_rate": 9.322196265559981e-05, "loss": 0.0576, "step": 20930 }, { "grad_norm": 0.47143682837486267, "learning_rate": 9.321364768313803e-05, "loss": 0.061, "step": 20940 }, { "grad_norm": 0.5350097417831421, "learning_rate": 9.32053279848967e-05, "loss": 0.0504, "step": 20950 }, { "grad_norm": 0.3257359564304352, "learning_rate": 9.319700356178567e-05, "loss": 0.056, "step": 20960 }, { "grad_norm": 0.4280160367488861, "learning_rate": 9.318867441471527e-05, "loss": 0.0478, "step": 20970 }, { "grad_norm": 0.5626288056373596, "learning_rate": 9.318034054459637e-05, "loss": 0.0428, "step": 20980 }, { "grad_norm": 0.4925210177898407, "learning_rate": 9.317200195234034e-05, "loss": 0.0428, "step": 20990 }, { "grad_norm": 0.3900071084499359, "learning_rate": 9.316365863885909e-05, "loss": 0.0482, "step": 21000 }, { "grad_norm": 0.491330087184906, "learning_rate": 9.315531060506502e-05, "loss": 0.0559, "step": 21010 }, { "grad_norm": 0.7274550795555115, "learning_rate": 9.314695785187108e-05, "loss": 0.0537, "step": 21020 }, { "grad_norm": 0.3946344554424286, "learning_rate": 9.313860038019069e-05, "loss": 0.0563, "step": 21030 }, { "grad_norm": 0.8328801393508911, "learning_rate": 9.313023819093782e-05, "loss": 0.0487, "step": 21040 }, { "grad_norm": 0.6924277544021606, "learning_rate": 9.312187128502695e-05, "loss": 0.0641, "step": 21050 }, { "grad_norm": 0.4520007073879242, "learning_rate": 9.311349966337307e-05, "loss": 0.0555, "step": 21060 }, { "grad_norm": 0.3641352653503418, "learning_rate": 9.310512332689169e-05, "loss": 0.0525, "step": 21070 }, { "grad_norm": 0.32729461789131165, "learning_rate": 9.309674227649883e-05, "loss": 0.0455, "step": 21080 }, { "grad_norm": 0.6919976472854614, "learning_rate": 9.308835651311103e-05, "loss": 0.0559, "step": 21090 }, { "grad_norm": 0.41793838143348694, "learning_rate": 9.307996603764533e-05, "loss": 0.0585, "step": 21100 }, { "grad_norm": 0.47210171818733215, "learning_rate": 9.307157085101932e-05, "loss": 0.0487, "step": 21110 }, { "grad_norm": 0.5846587419509888, "learning_rate": 9.306317095415109e-05, "loss": 0.0534, "step": 21120 }, { "grad_norm": 0.548490047454834, "learning_rate": 9.305476634795922e-05, "loss": 0.0583, "step": 21130 }, { "grad_norm": 0.6784730553627014, "learning_rate": 9.304635703336284e-05, "loss": 0.0523, "step": 21140 }, { "grad_norm": 0.47863009572029114, "learning_rate": 9.303794301128157e-05, "loss": 0.0444, "step": 21150 }, { "grad_norm": 0.39150968194007874, "learning_rate": 9.302952428263555e-05, "loss": 0.048, "step": 21160 }, { "grad_norm": 0.4181963801383972, "learning_rate": 9.302110084834545e-05, "loss": 0.0423, "step": 21170 }, { "grad_norm": 0.5106458067893982, "learning_rate": 9.301267270933245e-05, "loss": 0.0459, "step": 21180 }, { "grad_norm": 0.5923036932945251, "learning_rate": 9.300423986651823e-05, "loss": 0.0479, "step": 21190 }, { "grad_norm": 0.6639412045478821, "learning_rate": 9.299580232082501e-05, "loss": 0.0514, "step": 21200 }, { "grad_norm": 0.5748220682144165, "learning_rate": 9.298736007317547e-05, "loss": 0.0557, "step": 21210 }, { "grad_norm": 0.587500274181366, "learning_rate": 9.297891312449288e-05, "loss": 0.0651, "step": 21220 }, { "grad_norm": 0.4702360928058624, "learning_rate": 9.297046147570094e-05, "loss": 0.0633, "step": 21230 }, { "grad_norm": 0.5268077850341797, "learning_rate": 9.296200512772396e-05, "loss": 0.0541, "step": 21240 }, { "grad_norm": 0.3826742172241211, "learning_rate": 9.295354408148668e-05, "loss": 0.0437, "step": 21250 }, { "grad_norm": 0.42525044083595276, "learning_rate": 9.294507833791441e-05, "loss": 0.0585, "step": 21260 }, { "grad_norm": 0.4046744704246521, "learning_rate": 9.293660789793295e-05, "loss": 0.0532, "step": 21270 }, { "grad_norm": 0.4592534005641937, "learning_rate": 9.292813276246858e-05, "loss": 0.0484, "step": 21280 }, { "grad_norm": 0.43012353777885437, "learning_rate": 9.291965293244816e-05, "loss": 0.0498, "step": 21290 }, { "grad_norm": 0.30014416575431824, "learning_rate": 9.291116840879904e-05, "loss": 0.0428, "step": 21300 }, { "grad_norm": 0.5113003849983215, "learning_rate": 9.290267919244904e-05, "loss": 0.0521, "step": 21310 }, { "grad_norm": 0.3654489815235138, "learning_rate": 9.289418528432655e-05, "loss": 0.0429, "step": 21320 }, { "grad_norm": 0.5724778175354004, "learning_rate": 9.288568668536045e-05, "loss": 0.0558, "step": 21330 }, { "grad_norm": 0.5983151793479919, "learning_rate": 9.287718339648013e-05, "loss": 0.0539, "step": 21340 }, { "grad_norm": 0.44103991985321045, "learning_rate": 9.28686754186155e-05, "loss": 0.0541, "step": 21350 }, { "grad_norm": 0.477556973695755, "learning_rate": 9.286016275269698e-05, "loss": 0.0587, "step": 21360 }, { "grad_norm": 0.6768593192100525, "learning_rate": 9.285164539965551e-05, "loss": 0.0545, "step": 21370 }, { "grad_norm": 0.572586715221405, "learning_rate": 9.284312336042251e-05, "loss": 0.0427, "step": 21380 }, { "grad_norm": 0.5999630093574524, "learning_rate": 9.283459663592996e-05, "loss": 0.0484, "step": 21390 }, { "grad_norm": 0.4119979441165924, "learning_rate": 9.282606522711033e-05, "loss": 0.0547, "step": 21400 }, { "grad_norm": 0.37442463636398315, "learning_rate": 9.281752913489657e-05, "loss": 0.0585, "step": 21410 }, { "grad_norm": 0.5383127331733704, "learning_rate": 9.280898836022222e-05, "loss": 0.0596, "step": 21420 }, { "grad_norm": 0.4425511062145233, "learning_rate": 9.280044290402126e-05, "loss": 0.0488, "step": 21430 }, { "grad_norm": 0.5804154276847839, "learning_rate": 9.279189276722821e-05, "loss": 0.0607, "step": 21440 }, { "grad_norm": 0.731231153011322, "learning_rate": 9.278333795077812e-05, "loss": 0.067, "step": 21450 }, { "grad_norm": 0.4890766441822052, "learning_rate": 9.27747784556065e-05, "loss": 0.0537, "step": 21460 }, { "grad_norm": 0.38840922713279724, "learning_rate": 9.276621428264942e-05, "loss": 0.0554, "step": 21470 }, { "grad_norm": 0.5984362363815308, "learning_rate": 9.275764543284345e-05, "loss": 0.0462, "step": 21480 }, { "grad_norm": 0.6427992582321167, "learning_rate": 9.274907190712566e-05, "loss": 0.0625, "step": 21490 }, { "grad_norm": 0.5300652980804443, "learning_rate": 9.274049370643363e-05, "loss": 0.0556, "step": 21500 }, { "grad_norm": 0.6288553476333618, "learning_rate": 9.273191083170547e-05, "loss": 0.0594, "step": 21510 }, { "grad_norm": 0.32101160287857056, "learning_rate": 9.27233232838798e-05, "loss": 0.0591, "step": 21520 }, { "grad_norm": 0.668433427810669, "learning_rate": 9.27147310638957e-05, "loss": 0.0521, "step": 21530 }, { "grad_norm": 0.41740682721138, "learning_rate": 9.270613417269286e-05, "loss": 0.0428, "step": 21540 }, { "grad_norm": 0.5659834742546082, "learning_rate": 9.269753261121138e-05, "loss": 0.0535, "step": 21550 }, { "grad_norm": 0.5552884340286255, "learning_rate": 9.268892638039194e-05, "loss": 0.0465, "step": 21560 }, { "grad_norm": 0.34924814105033875, "learning_rate": 9.268031548117569e-05, "loss": 0.0483, "step": 21570 }, { "grad_norm": 0.5680108666419983, "learning_rate": 9.26716999145043e-05, "loss": 0.0433, "step": 21580 }, { "grad_norm": 0.371097207069397, "learning_rate": 9.266307968131998e-05, "loss": 0.0464, "step": 21590 }, { "grad_norm": 0.49794989824295044, "learning_rate": 9.26544547825654e-05, "loss": 0.0529, "step": 21600 }, { "grad_norm": 0.6273518800735474, "learning_rate": 9.264582521918376e-05, "loss": 0.0622, "step": 21610 }, { "grad_norm": 0.350774884223938, "learning_rate": 9.263719099211881e-05, "loss": 0.0583, "step": 21620 }, { "grad_norm": 0.5075017809867859, "learning_rate": 9.262855210231476e-05, "loss": 0.0528, "step": 21630 }, { "grad_norm": 0.5543321967124939, "learning_rate": 9.261990855071633e-05, "loss": 0.0496, "step": 21640 }, { "grad_norm": 0.36247575283050537, "learning_rate": 9.261126033826878e-05, "loss": 0.0399, "step": 21650 }, { "grad_norm": 0.4323035776615143, "learning_rate": 9.260260746591786e-05, "loss": 0.0534, "step": 21660 }, { "grad_norm": 0.37324267625808716, "learning_rate": 9.259394993460985e-05, "loss": 0.0436, "step": 21670 }, { "grad_norm": 0.31500276923179626, "learning_rate": 9.258528774529151e-05, "loss": 0.0479, "step": 21680 }, { "grad_norm": 0.631481945514679, "learning_rate": 9.257662089891013e-05, "loss": 0.0455, "step": 21690 }, { "grad_norm": 0.5473718047142029, "learning_rate": 9.25679493964135e-05, "loss": 0.0592, "step": 21700 }, { "grad_norm": 0.3803761303424835, "learning_rate": 9.255927323874994e-05, "loss": 0.0487, "step": 21710 }, { "grad_norm": 0.3946908116340637, "learning_rate": 9.255059242686822e-05, "loss": 0.053, "step": 21720 }, { "grad_norm": 0.3539426326751709, "learning_rate": 9.254190696171769e-05, "loss": 0.0411, "step": 21730 }, { "grad_norm": 0.4266619086265564, "learning_rate": 9.25332168442482e-05, "loss": 0.0475, "step": 21740 }, { "grad_norm": 0.44441771507263184, "learning_rate": 9.252452207541004e-05, "loss": 0.0432, "step": 21750 }, { "grad_norm": 0.5168728828430176, "learning_rate": 9.251582265615409e-05, "loss": 0.0594, "step": 21760 }, { "grad_norm": 0.5186612606048584, "learning_rate": 9.250711858743169e-05, "loss": 0.0567, "step": 21770 }, { "grad_norm": 0.44857895374298096, "learning_rate": 9.24984098701947e-05, "loss": 0.066, "step": 21780 }, { "grad_norm": 0.5071711540222168, "learning_rate": 9.248969650539552e-05, "loss": 0.0531, "step": 21790 }, { "grad_norm": 0.6167186498641968, "learning_rate": 9.2480978493987e-05, "loss": 0.0556, "step": 21800 }, { "grad_norm": 0.5756961107254028, "learning_rate": 9.247225583692256e-05, "loss": 0.0449, "step": 21810 }, { "grad_norm": 0.2936396896839142, "learning_rate": 9.246352853515607e-05, "loss": 0.0501, "step": 21820 }, { "grad_norm": 0.46540015935897827, "learning_rate": 9.245479658964194e-05, "loss": 0.0565, "step": 21830 }, { "grad_norm": 0.6821207404136658, "learning_rate": 9.244606000133507e-05, "loss": 0.0433, "step": 21840 }, { "grad_norm": 0.6983935236930847, "learning_rate": 9.24373187711909e-05, "loss": 0.0479, "step": 21850 }, { "grad_norm": 0.3821820616722107, "learning_rate": 9.242857290016537e-05, "loss": 0.0537, "step": 21860 }, { "grad_norm": 0.7067357301712036, "learning_rate": 9.241982238921488e-05, "loss": 0.0438, "step": 21870 }, { "grad_norm": 0.558064877986908, "learning_rate": 9.24110672392964e-05, "loss": 0.0445, "step": 21880 }, { "grad_norm": 0.4300767481327057, "learning_rate": 9.240230745136737e-05, "loss": 0.0524, "step": 21890 }, { "grad_norm": 0.37658387422561646, "learning_rate": 9.239354302638575e-05, "loss": 0.0452, "step": 21900 }, { "grad_norm": 0.7590086460113525, "learning_rate": 9.238477396531e-05, "loss": 0.0579, "step": 21910 }, { "grad_norm": 0.49723321199417114, "learning_rate": 9.23760002690991e-05, "loss": 0.0507, "step": 21920 }, { "grad_norm": 0.36752891540527344, "learning_rate": 9.236722193871252e-05, "loss": 0.0481, "step": 21930 }, { "grad_norm": 0.6505226492881775, "learning_rate": 9.235843897511023e-05, "loss": 0.0461, "step": 21940 }, { "grad_norm": 0.5916743874549866, "learning_rate": 9.234965137925276e-05, "loss": 0.055, "step": 21950 }, { "grad_norm": 0.4186338186264038, "learning_rate": 9.234085915210108e-05, "loss": 0.0482, "step": 21960 }, { "grad_norm": 0.4339730739593506, "learning_rate": 9.23320622946167e-05, "loss": 0.0525, "step": 21970 }, { "grad_norm": 0.35412442684173584, "learning_rate": 9.232326080776163e-05, "loss": 0.0438, "step": 21980 }, { "grad_norm": 0.6717664003372192, "learning_rate": 9.23144546924984e-05, "loss": 0.0461, "step": 21990 }, { "grad_norm": 0.5786798596382141, "learning_rate": 9.230564394979e-05, "loss": 0.0639, "step": 22000 }, { "grad_norm": 0.42496418952941895, "learning_rate": 9.22968285806e-05, "loss": 0.0479, "step": 22010 }, { "grad_norm": 0.5270272493362427, "learning_rate": 9.228800858589242e-05, "loss": 0.0576, "step": 22020 }, { "grad_norm": 0.4291243553161621, "learning_rate": 9.227918396663179e-05, "loss": 0.051, "step": 22030 }, { "grad_norm": 0.8509631156921387, "learning_rate": 9.227035472378319e-05, "loss": 0.0567, "step": 22040 }, { "grad_norm": 0.6971508264541626, "learning_rate": 9.226152085831213e-05, "loss": 0.068, "step": 22050 }, { "grad_norm": 0.43718421459198, "learning_rate": 9.22526823711847e-05, "loss": 0.0568, "step": 22060 }, { "grad_norm": 0.48706313967704773, "learning_rate": 9.224383926336745e-05, "loss": 0.051, "step": 22070 }, { "grad_norm": 0.5099806189537048, "learning_rate": 9.223499153582744e-05, "loss": 0.0476, "step": 22080 }, { "grad_norm": 0.5011753439903259, "learning_rate": 9.222613918953226e-05, "loss": 0.0632, "step": 22090 }, { "grad_norm": 0.3192557692527771, "learning_rate": 9.221728222544999e-05, "loss": 0.0389, "step": 22100 }, { "grad_norm": 0.5668396949768066, "learning_rate": 9.22084206445492e-05, "loss": 0.0556, "step": 22110 }, { "grad_norm": 0.5958238840103149, "learning_rate": 9.2199554447799e-05, "loss": 0.0701, "step": 22120 }, { "grad_norm": 0.4058379828929901, "learning_rate": 9.219068363616897e-05, "loss": 0.0505, "step": 22130 }, { "grad_norm": 0.3169485330581665, "learning_rate": 9.218180821062919e-05, "loss": 0.0437, "step": 22140 }, { "grad_norm": 2.558947801589966, "learning_rate": 9.21729281721503e-05, "loss": 0.0512, "step": 22150 }, { "grad_norm": 0.5458787679672241, "learning_rate": 9.216404352170339e-05, "loss": 0.0559, "step": 22160 }, { "grad_norm": 0.6073387861251831, "learning_rate": 9.215515426026007e-05, "loss": 0.0509, "step": 22170 }, { "grad_norm": 0.47232815623283386, "learning_rate": 9.214626038879246e-05, "loss": 0.0559, "step": 22180 }, { "grad_norm": 0.42491358518600464, "learning_rate": 9.21373619082732e-05, "loss": 0.0583, "step": 22190 }, { "grad_norm": 0.49214276671409607, "learning_rate": 9.212845881967535e-05, "loss": 0.0584, "step": 22200 }, { "grad_norm": 0.5391765832901001, "learning_rate": 9.211955112397262e-05, "loss": 0.0634, "step": 22210 }, { "grad_norm": 0.2965449094772339, "learning_rate": 9.211063882213909e-05, "loss": 0.0469, "step": 22220 }, { "grad_norm": 0.4146352708339691, "learning_rate": 9.210172191514942e-05, "loss": 0.0422, "step": 22230 }, { "grad_norm": 0.45930179953575134, "learning_rate": 9.209280040397874e-05, "loss": 0.0499, "step": 22240 }, { "grad_norm": 0.41437214612960815, "learning_rate": 9.208387428960268e-05, "loss": 0.0431, "step": 22250 }, { "grad_norm": 0.4209330379962921, "learning_rate": 9.20749435729974e-05, "loss": 0.0478, "step": 22260 }, { "grad_norm": 0.3738850951194763, "learning_rate": 9.206600825513957e-05, "loss": 0.0472, "step": 22270 }, { "grad_norm": 0.420818954706192, "learning_rate": 9.20570683370063e-05, "loss": 0.0478, "step": 22280 }, { "grad_norm": 0.7731831073760986, "learning_rate": 9.204812381957528e-05, "loss": 0.0469, "step": 22290 }, { "grad_norm": 0.3649783730506897, "learning_rate": 9.203917470382465e-05, "loss": 0.0434, "step": 22300 }, { "grad_norm": 0.7883176207542419, "learning_rate": 9.203022099073309e-05, "loss": 0.0592, "step": 22310 }, { "grad_norm": 0.43873146176338196, "learning_rate": 9.202126268127976e-05, "loss": 0.0536, "step": 22320 }, { "grad_norm": 0.6572496891021729, "learning_rate": 9.20122997764443e-05, "loss": 0.0505, "step": 22330 }, { "grad_norm": 0.4579138457775116, "learning_rate": 9.200333227720692e-05, "loss": 0.059, "step": 22340 }, { "grad_norm": 0.6683023571968079, "learning_rate": 9.199436018454826e-05, "loss": 0.0597, "step": 22350 }, { "grad_norm": 0.3308870792388916, "learning_rate": 9.198538349944952e-05, "loss": 0.0469, "step": 22360 }, { "grad_norm": 0.37740558385849, "learning_rate": 9.197640222289234e-05, "loss": 0.0664, "step": 22370 }, { "grad_norm": 0.5956355929374695, "learning_rate": 9.196741635585895e-05, "loss": 0.0505, "step": 22380 }, { "grad_norm": 1.0719035863876343, "learning_rate": 9.195842589933199e-05, "loss": 0.0493, "step": 22390 }, { "grad_norm": 0.5363498330116272, "learning_rate": 9.194943085429466e-05, "loss": 0.0571, "step": 22400 }, { "grad_norm": 0.46925580501556396, "learning_rate": 9.194043122173065e-05, "loss": 0.075, "step": 22410 }, { "grad_norm": 0.5972708463668823, "learning_rate": 9.193142700262413e-05, "loss": 0.062, "step": 22420 }, { "grad_norm": 0.31660422682762146, "learning_rate": 9.192241819795979e-05, "loss": 0.046, "step": 22430 }, { "grad_norm": 0.41364336013793945, "learning_rate": 9.191340480872284e-05, "loss": 0.0429, "step": 22440 }, { "grad_norm": 0.3678838014602661, "learning_rate": 9.190438683589895e-05, "loss": 0.0462, "step": 22450 }, { "grad_norm": 0.5342692136764526, "learning_rate": 9.189536428047432e-05, "loss": 0.0541, "step": 22460 }, { "grad_norm": 0.5553489327430725, "learning_rate": 9.188633714343564e-05, "loss": 0.0482, "step": 22470 }, { "grad_norm": 0.6875166296958923, "learning_rate": 9.18773054257701e-05, "loss": 0.0549, "step": 22480 }, { "grad_norm": 0.6902441382408142, "learning_rate": 9.18682691284654e-05, "loss": 0.0457, "step": 22490 }, { "grad_norm": 0.3507857024669647, "learning_rate": 9.185922825250974e-05, "loss": 0.0496, "step": 22500 }, { "grad_norm": 0.6505299806594849, "learning_rate": 9.185018279889181e-05, "loss": 0.0512, "step": 22510 }, { "grad_norm": 1.1713340282440186, "learning_rate": 9.184113276860082e-05, "loss": 0.0579, "step": 22520 }, { "grad_norm": 0.5207945108413696, "learning_rate": 9.183207816262645e-05, "loss": 0.0538, "step": 22530 }, { "grad_norm": 0.3537262976169586, "learning_rate": 9.182301898195891e-05, "loss": 0.0435, "step": 22540 }, { "grad_norm": 0.7456658482551575, "learning_rate": 9.181395522758889e-05, "loss": 0.0501, "step": 22550 }, { "grad_norm": 0.7994046807289124, "learning_rate": 9.180488690050759e-05, "loss": 0.0748, "step": 22560 }, { "grad_norm": 0.4470898509025574, "learning_rate": 9.179581400170671e-05, "loss": 0.0571, "step": 22570 }, { "grad_norm": 0.2823653817176819, "learning_rate": 9.178673653217845e-05, "loss": 0.0447, "step": 22580 }, { "grad_norm": 0.5490701794624329, "learning_rate": 9.177765449291551e-05, "loss": 0.0499, "step": 22590 }, { "grad_norm": 0.4075702130794525, "learning_rate": 9.176856788491109e-05, "loss": 0.0548, "step": 22600 }, { "grad_norm": 0.5449076890945435, "learning_rate": 9.175947670915887e-05, "loss": 0.0501, "step": 22610 }, { "grad_norm": 0.423504501581192, "learning_rate": 9.175038096665309e-05, "loss": 0.0452, "step": 22620 }, { "grad_norm": 0.5431160926818848, "learning_rate": 9.17412806583884e-05, "loss": 0.0503, "step": 22630 }, { "grad_norm": 0.5970975756645203, "learning_rate": 9.173217578536002e-05, "loss": 0.0694, "step": 22640 }, { "grad_norm": 0.5112806558609009, "learning_rate": 9.172306634856362e-05, "loss": 0.0494, "step": 22650 }, { "grad_norm": 0.4844694435596466, "learning_rate": 9.171395234899545e-05, "loss": 0.0486, "step": 22660 }, { "grad_norm": 0.5725194215774536, "learning_rate": 9.170483378765214e-05, "loss": 0.0591, "step": 22670 }, { "grad_norm": 0.8955755233764648, "learning_rate": 9.169571066553091e-05, "loss": 0.0554, "step": 22680 }, { "grad_norm": 2.3442652225494385, "learning_rate": 9.168658298362946e-05, "loss": 0.0509, "step": 22690 }, { "grad_norm": 0.5205751061439514, "learning_rate": 9.167745074294598e-05, "loss": 0.0505, "step": 22700 }, { "grad_norm": 0.5802851319313049, "learning_rate": 9.166831394447913e-05, "loss": 0.0598, "step": 22710 }, { "grad_norm": 0.47388744354248047, "learning_rate": 9.165917258922812e-05, "loss": 0.0414, "step": 22720 }, { "grad_norm": 0.7102677822113037, "learning_rate": 9.165002667819262e-05, "loss": 0.0564, "step": 22730 }, { "grad_norm": 0.4345396161079407, "learning_rate": 9.164087621237282e-05, "loss": 0.0516, "step": 22740 }, { "grad_norm": 0.5483841896057129, "learning_rate": 9.163172119276942e-05, "loss": 0.0651, "step": 22750 }, { "grad_norm": 0.5739604830741882, "learning_rate": 9.162256162038358e-05, "loss": 0.0556, "step": 22760 }, { "grad_norm": 0.44833266735076904, "learning_rate": 9.161339749621698e-05, "loss": 0.0586, "step": 22770 }, { "grad_norm": 0.6263863444328308, "learning_rate": 9.160422882127177e-05, "loss": 0.0492, "step": 22780 }, { "grad_norm": 0.471983939409256, "learning_rate": 9.159505559655069e-05, "loss": 0.0514, "step": 22790 }, { "grad_norm": 0.39646434783935547, "learning_rate": 9.158587782305684e-05, "loss": 0.0477, "step": 22800 }, { "grad_norm": 0.5284303426742554, "learning_rate": 9.157669550179391e-05, "loss": 0.053, "step": 22810 }, { "grad_norm": 0.8260791897773743, "learning_rate": 9.156750863376609e-05, "loss": 0.0563, "step": 22820 }, { "grad_norm": 0.37159106135368347, "learning_rate": 9.155831721997801e-05, "loss": 0.0442, "step": 22830 }, { "grad_norm": 0.39166200160980225, "learning_rate": 9.154912126143484e-05, "loss": 0.0512, "step": 22840 }, { "grad_norm": 0.48090508580207825, "learning_rate": 9.153992075914224e-05, "loss": 0.0653, "step": 22850 }, { "grad_norm": 0.3329048156738281, "learning_rate": 9.153071571410635e-05, "loss": 0.0359, "step": 22860 }, { "grad_norm": 0.7373431324958801, "learning_rate": 9.152150612733384e-05, "loss": 0.0523, "step": 22870 }, { "grad_norm": 0.4703950881958008, "learning_rate": 9.151229199983184e-05, "loss": 0.0555, "step": 22880 }, { "grad_norm": 0.7543070912361145, "learning_rate": 9.150307333260802e-05, "loss": 0.0495, "step": 22890 }, { "grad_norm": 0.3650898337364197, "learning_rate": 9.149385012667048e-05, "loss": 0.0548, "step": 22900 }, { "grad_norm": 0.4061911106109619, "learning_rate": 9.148462238302788e-05, "loss": 0.0476, "step": 22910 }, { "grad_norm": 0.5749254822731018, "learning_rate": 9.147539010268936e-05, "loss": 0.0578, "step": 22920 }, { "grad_norm": 0.47148817777633667, "learning_rate": 9.14661532866645e-05, "loss": 0.0522, "step": 22930 }, { "grad_norm": 0.3680020272731781, "learning_rate": 9.145691193596348e-05, "loss": 0.0461, "step": 22940 }, { "grad_norm": 0.6482396721839905, "learning_rate": 9.144766605159691e-05, "loss": 0.0598, "step": 22950 }, { "grad_norm": 0.3336805999279022, "learning_rate": 9.14384156345759e-05, "loss": 0.0548, "step": 22960 }, { "grad_norm": 0.9241172075271606, "learning_rate": 9.142916068591204e-05, "loss": 0.0445, "step": 22970 }, { "grad_norm": 1.122294545173645, "learning_rate": 9.141990120661746e-05, "loss": 0.0587, "step": 22980 }, { "grad_norm": 0.5983197689056396, "learning_rate": 9.141063719770475e-05, "loss": 0.0471, "step": 22990 }, { "grad_norm": 0.5318648815155029, "learning_rate": 9.140136866018704e-05, "loss": 0.052, "step": 23000 }, { "grad_norm": 0.3242218494415283, "learning_rate": 9.139209559507788e-05, "loss": 0.0481, "step": 23010 }, { "grad_norm": 0.4408762753009796, "learning_rate": 9.13828180033914e-05, "loss": 0.0467, "step": 23020 }, { "grad_norm": 0.4095713198184967, "learning_rate": 9.137353588614212e-05, "loss": 0.0445, "step": 23030 }, { "grad_norm": 0.3614107370376587, "learning_rate": 9.136424924434519e-05, "loss": 0.0438, "step": 23040 }, { "grad_norm": 0.45477867126464844, "learning_rate": 9.135495807901615e-05, "loss": 0.0614, "step": 23050 }, { "grad_norm": 0.37865176796913147, "learning_rate": 9.134566239117108e-05, "loss": 0.0476, "step": 23060 }, { "grad_norm": 0.707849383354187, "learning_rate": 9.13363621818265e-05, "loss": 0.0534, "step": 23070 }, { "grad_norm": 0.5870724320411682, "learning_rate": 9.132705745199953e-05, "loss": 0.0582, "step": 23080 }, { "grad_norm": 0.41257068514823914, "learning_rate": 9.131774820270768e-05, "loss": 0.0501, "step": 23090 }, { "grad_norm": 0.46944889426231384, "learning_rate": 9.130843443496901e-05, "loss": 0.0637, "step": 23100 }, { "grad_norm": 0.7722094058990479, "learning_rate": 9.129911614980206e-05, "loss": 0.0655, "step": 23110 }, { "grad_norm": 0.5680814981460571, "learning_rate": 9.128979334822584e-05, "loss": 0.0523, "step": 23120 }, { "grad_norm": 1.1128933429718018, "learning_rate": 9.128046603125992e-05, "loss": 0.0516, "step": 23130 }, { "grad_norm": 0.4979366660118103, "learning_rate": 9.12711341999243e-05, "loss": 0.0482, "step": 23140 }, { "grad_norm": 0.6141552925109863, "learning_rate": 9.12617978552395e-05, "loss": 0.0793, "step": 23150 }, { "grad_norm": 0.4961397051811218, "learning_rate": 9.12524569982265e-05, "loss": 0.0558, "step": 23160 }, { "grad_norm": 0.5830203294754028, "learning_rate": 9.124311162990684e-05, "loss": 0.0493, "step": 23170 }, { "grad_norm": 0.5050665736198425, "learning_rate": 9.12337617513025e-05, "loss": 0.0445, "step": 23180 }, { "grad_norm": 0.4757634401321411, "learning_rate": 9.122440736343596e-05, "loss": 0.052, "step": 23190 }, { "grad_norm": 1.1444047689437866, "learning_rate": 9.12150484673302e-05, "loss": 0.059, "step": 23200 }, { "grad_norm": 0.3514430522918701, "learning_rate": 9.120568506400873e-05, "loss": 0.0444, "step": 23210 }, { "grad_norm": 0.5139452815055847, "learning_rate": 9.119631715449548e-05, "loss": 0.0558, "step": 23220 }, { "grad_norm": 0.43602535128593445, "learning_rate": 9.118694473981493e-05, "loss": 0.0486, "step": 23230 }, { "grad_norm": 0.3611375391483307, "learning_rate": 9.117756782099203e-05, "loss": 0.0446, "step": 23240 }, { "grad_norm": 0.5515069961547852, "learning_rate": 9.11681863990522e-05, "loss": 0.0512, "step": 23250 }, { "grad_norm": 0.5537831783294678, "learning_rate": 9.115880047502142e-05, "loss": 0.0499, "step": 23260 }, { "grad_norm": 0.6723597645759583, "learning_rate": 9.114941004992609e-05, "loss": 0.0535, "step": 23270 }, { "grad_norm": 0.35793885588645935, "learning_rate": 9.114001512479317e-05, "loss": 0.0438, "step": 23280 }, { "grad_norm": 0.4914464056491852, "learning_rate": 9.113061570065003e-05, "loss": 0.0721, "step": 23290 }, { "grad_norm": 0.46814030408859253, "learning_rate": 9.112121177852459e-05, "loss": 0.0565, "step": 23300 }, { "grad_norm": 0.4813085198402405, "learning_rate": 9.111180335944527e-05, "loss": 0.0558, "step": 23310 }, { "grad_norm": 0.6601448059082031, "learning_rate": 9.110239044444093e-05, "loss": 0.0491, "step": 23320 }, { "grad_norm": 0.41690152883529663, "learning_rate": 9.109297303454099e-05, "loss": 0.0551, "step": 23330 }, { "grad_norm": 0.3411155939102173, "learning_rate": 9.108355113077526e-05, "loss": 0.0459, "step": 23340 }, { "grad_norm": 0.3194023072719574, "learning_rate": 9.107412473417419e-05, "loss": 0.0389, "step": 23350 }, { "grad_norm": 0.3397815227508545, "learning_rate": 9.106469384576858e-05, "loss": 0.0449, "step": 23360 }, { "grad_norm": 0.42989715933799744, "learning_rate": 9.105525846658978e-05, "loss": 0.0463, "step": 23370 }, { "grad_norm": 0.42911139130592346, "learning_rate": 9.104581859766965e-05, "loss": 0.0518, "step": 23380 }, { "grad_norm": 0.38278183341026306, "learning_rate": 9.10363742400405e-05, "loss": 0.0498, "step": 23390 }, { "grad_norm": 0.47649648785591125, "learning_rate": 9.102692539473518e-05, "loss": 0.0539, "step": 23400 }, { "grad_norm": 0.49192509055137634, "learning_rate": 9.101747206278697e-05, "loss": 0.0548, "step": 23410 }, { "grad_norm": 0.6494051218032837, "learning_rate": 9.100801424522968e-05, "loss": 0.0444, "step": 23420 }, { "grad_norm": 0.4016605019569397, "learning_rate": 9.099855194309762e-05, "loss": 0.0467, "step": 23430 }, { "grad_norm": 0.3754083216190338, "learning_rate": 9.098908515742554e-05, "loss": 0.0498, "step": 23440 }, { "grad_norm": 0.5448657870292664, "learning_rate": 9.097961388924873e-05, "loss": 0.0405, "step": 23450 }, { "grad_norm": 0.38733533024787903, "learning_rate": 9.097013813960298e-05, "loss": 0.0423, "step": 23460 }, { "grad_norm": 0.6558971405029297, "learning_rate": 9.09606579095245e-05, "loss": 0.0581, "step": 23470 }, { "grad_norm": 0.3467830717563629, "learning_rate": 9.095117320005008e-05, "loss": 0.0539, "step": 23480 }, { "grad_norm": 0.6047874689102173, "learning_rate": 9.094168401221691e-05, "loss": 0.0528, "step": 23490 }, { "grad_norm": 0.31695371866226196, "learning_rate": 9.093219034706273e-05, "loss": 0.0575, "step": 23500 }, { "grad_norm": 0.5650436282157898, "learning_rate": 9.092269220562577e-05, "loss": 0.0546, "step": 23510 }, { "grad_norm": 0.38496407866477966, "learning_rate": 9.09131895889447e-05, "loss": 0.0754, "step": 23520 }, { "grad_norm": 0.5613968968391418, "learning_rate": 9.090368249805873e-05, "loss": 0.047, "step": 23530 }, { "grad_norm": 0.44108471274375916, "learning_rate": 9.089417093400754e-05, "loss": 0.052, "step": 23540 }, { "grad_norm": 0.3630210757255554, "learning_rate": 9.088465489783131e-05, "loss": 0.0522, "step": 23550 }, { "grad_norm": 0.35798779129981995, "learning_rate": 9.087513439057068e-05, "loss": 0.0526, "step": 23560 }, { "grad_norm": 0.3327755331993103, "learning_rate": 9.08656094132668e-05, "loss": 0.0553, "step": 23570 }, { "grad_norm": 0.41658976674079895, "learning_rate": 9.085607996696134e-05, "loss": 0.0465, "step": 23580 }, { "grad_norm": 0.3814767301082611, "learning_rate": 9.084654605269639e-05, "loss": 0.0512, "step": 23590 }, { "grad_norm": 0.43943214416503906, "learning_rate": 9.083700767151457e-05, "loss": 0.0476, "step": 23600 }, { "grad_norm": 0.58415687084198, "learning_rate": 9.082746482445898e-05, "loss": 0.0596, "step": 23610 }, { "grad_norm": 0.381387323141098, "learning_rate": 9.081791751257325e-05, "loss": 0.0516, "step": 23620 }, { "grad_norm": 0.6569604873657227, "learning_rate": 9.080836573690142e-05, "loss": 0.0518, "step": 23630 }, { "grad_norm": 0.4416153132915497, "learning_rate": 9.079880949848805e-05, "loss": 0.0631, "step": 23640 }, { "grad_norm": 0.4901922047138214, "learning_rate": 9.078924879837822e-05, "loss": 0.0484, "step": 23650 }, { "grad_norm": 0.6526432037353516, "learning_rate": 9.077968363761747e-05, "loss": 0.0494, "step": 23660 }, { "grad_norm": 0.3462112843990326, "learning_rate": 9.077011401725182e-05, "loss": 0.0454, "step": 23670 }, { "grad_norm": 0.3463423550128937, "learning_rate": 9.07605399383278e-05, "loss": 0.0469, "step": 23680 }, { "grad_norm": 0.5242695808410645, "learning_rate": 9.075096140189243e-05, "loss": 0.042, "step": 23690 }, { "grad_norm": 0.5250760912895203, "learning_rate": 9.074137840899318e-05, "loss": 0.0577, "step": 23700 }, { "grad_norm": 0.3156971335411072, "learning_rate": 9.073179096067804e-05, "loss": 0.0368, "step": 23710 }, { "grad_norm": 0.4991817772388458, "learning_rate": 9.072219905799549e-05, "loss": 0.0447, "step": 23720 }, { "grad_norm": 0.620134174823761, "learning_rate": 9.071260270199447e-05, "loss": 0.0557, "step": 23730 }, { "grad_norm": 0.8030580282211304, "learning_rate": 9.070300189372441e-05, "loss": 0.0479, "step": 23740 }, { "grad_norm": 0.6708568930625916, "learning_rate": 9.069339663423528e-05, "loss": 0.0406, "step": 23750 }, { "grad_norm": 0.299669086933136, "learning_rate": 9.068378692457747e-05, "loss": 0.0404, "step": 23760 }, { "grad_norm": 0.3732960522174835, "learning_rate": 9.067417276580189e-05, "loss": 0.045, "step": 23770 }, { "grad_norm": 0.39729392528533936, "learning_rate": 9.066455415895993e-05, "loss": 0.046, "step": 23780 }, { "grad_norm": 0.37325945496559143, "learning_rate": 9.065493110510346e-05, "loss": 0.041, "step": 23790 }, { "grad_norm": 0.4224688708782196, "learning_rate": 9.064530360528484e-05, "loss": 0.0493, "step": 23800 }, { "grad_norm": 0.5873833894729614, "learning_rate": 9.063567166055695e-05, "loss": 0.0508, "step": 23810 }, { "grad_norm": 0.5386972427368164, "learning_rate": 9.062603527197308e-05, "loss": 0.0415, "step": 23820 }, { "grad_norm": 0.5240578055381775, "learning_rate": 9.06163944405871e-05, "loss": 0.0475, "step": 23830 }, { "grad_norm": 0.32852426171302795, "learning_rate": 9.060674916745327e-05, "loss": 0.0538, "step": 23840 }, { "grad_norm": 0.4685870110988617, "learning_rate": 9.05970994536264e-05, "loss": 0.0445, "step": 23850 }, { "grad_norm": 1.5050666332244873, "learning_rate": 9.05874453001618e-05, "loss": 0.0513, "step": 23860 }, { "grad_norm": 0.4461354613304138, "learning_rate": 9.057778670811517e-05, "loss": 0.0492, "step": 23870 }, { "grad_norm": 0.46199265122413635, "learning_rate": 9.056812367854281e-05, "loss": 0.0547, "step": 23880 }, { "grad_norm": 0.5875240564346313, "learning_rate": 9.055845621250143e-05, "loss": 0.042, "step": 23890 }, { "grad_norm": 0.5690351128578186, "learning_rate": 9.054878431104825e-05, "loss": 0.0387, "step": 23900 }, { "grad_norm": 0.3876402974128723, "learning_rate": 9.0539107975241e-05, "loss": 0.0422, "step": 23910 }, { "grad_norm": 0.357354074716568, "learning_rate": 9.052942720613784e-05, "loss": 0.0543, "step": 23920 }, { "grad_norm": 0.5735469460487366, "learning_rate": 9.051974200479745e-05, "loss": 0.0395, "step": 23930 }, { "grad_norm": 0.604789674282074, "learning_rate": 9.051005237227901e-05, "loss": 0.0535, "step": 23940 }, { "grad_norm": 0.38870763778686523, "learning_rate": 9.050035830964215e-05, "loss": 0.0411, "step": 23950 }, { "grad_norm": 0.678630530834198, "learning_rate": 9.049065981794698e-05, "loss": 0.0569, "step": 23960 }, { "grad_norm": 0.5260201096534729, "learning_rate": 9.048095689825414e-05, "loss": 0.0548, "step": 23970 }, { "grad_norm": 0.33381521701812744, "learning_rate": 9.047124955162472e-05, "loss": 0.0421, "step": 23980 }, { "grad_norm": 0.5034218430519104, "learning_rate": 9.046153777912028e-05, "loss": 0.0565, "step": 23990 }, { "grad_norm": 0.53938227891922, "learning_rate": 9.045182158180292e-05, "loss": 0.0466, "step": 24000 }, { "grad_norm": 0.6318572759628296, "learning_rate": 9.044210096073516e-05, "loss": 0.0435, "step": 24010 }, { "grad_norm": 0.4129359722137451, "learning_rate": 9.043237591698004e-05, "loss": 0.066, "step": 24020 }, { "grad_norm": 0.5381425023078918, "learning_rate": 9.04226464516011e-05, "loss": 0.0571, "step": 24030 }, { "grad_norm": 0.5679595470428467, "learning_rate": 9.041291256566229e-05, "loss": 0.0608, "step": 24040 }, { "grad_norm": 0.6957515478134155, "learning_rate": 9.040317426022814e-05, "loss": 0.0578, "step": 24050 }, { "grad_norm": 0.39428645372390747, "learning_rate": 9.03934315363636e-05, "loss": 0.053, "step": 24060 }, { "grad_norm": 0.7223996520042419, "learning_rate": 9.038368439513409e-05, "loss": 0.0538, "step": 24070 }, { "grad_norm": 0.9450490474700928, "learning_rate": 9.03739328376056e-05, "loss": 0.0604, "step": 24080 }, { "grad_norm": 0.7665397524833679, "learning_rate": 9.036417686484451e-05, "loss": 0.0582, "step": 24090 }, { "grad_norm": 0.6978155970573425, "learning_rate": 9.035441647791773e-05, "loss": 0.061, "step": 24100 }, { "grad_norm": 0.43964019417762756, "learning_rate": 9.034465167789263e-05, "loss": 0.0518, "step": 24110 }, { "grad_norm": 0.3698612153530121, "learning_rate": 9.033488246583706e-05, "loss": 0.0476, "step": 24120 }, { "grad_norm": 0.4826624095439911, "learning_rate": 9.032510884281941e-05, "loss": 0.0717, "step": 24130 }, { "grad_norm": 0.529681384563446, "learning_rate": 9.031533080990848e-05, "loss": 0.0492, "step": 24140 }, { "grad_norm": 0.43023186922073364, "learning_rate": 9.030554836817358e-05, "loss": 0.0436, "step": 24150 }, { "grad_norm": 0.5680093765258789, "learning_rate": 9.029576151868451e-05, "loss": 0.0438, "step": 24160 }, { "grad_norm": 0.3928593397140503, "learning_rate": 9.028597026251155e-05, "loss": 0.0455, "step": 24170 }, { "grad_norm": 0.5175392627716064, "learning_rate": 9.027617460072547e-05, "loss": 0.0448, "step": 24180 }, { "grad_norm": 0.7136282324790955, "learning_rate": 9.026637453439745e-05, "loss": 0.0396, "step": 24190 }, { "grad_norm": 0.4047495424747467, "learning_rate": 9.025657006459927e-05, "loss": 0.0548, "step": 24200 }, { "grad_norm": 0.5948030948638916, "learning_rate": 9.024676119240311e-05, "loss": 0.0509, "step": 24210 }, { "grad_norm": 0.5576897859573364, "learning_rate": 9.023694791888166e-05, "loss": 0.0486, "step": 24220 }, { "grad_norm": 0.6353868246078491, "learning_rate": 9.022713024510808e-05, "loss": 0.0508, "step": 24230 }, { "grad_norm": 1.243534803390503, "learning_rate": 9.021730817215601e-05, "loss": 0.0484, "step": 24240 }, { "grad_norm": 0.37018096446990967, "learning_rate": 9.02074817010996e-05, "loss": 0.0372, "step": 24250 }, { "grad_norm": 0.5529917478561401, "learning_rate": 9.019765083301342e-05, "loss": 0.0486, "step": 24260 }, { "grad_norm": 0.46682220697402954, "learning_rate": 9.01878155689726e-05, "loss": 0.0605, "step": 24270 }, { "grad_norm": 0.4355440139770508, "learning_rate": 9.017797591005268e-05, "loss": 0.0576, "step": 24280 }, { "grad_norm": 0.3908690810203552, "learning_rate": 9.016813185732972e-05, "loss": 0.0516, "step": 24290 }, { "grad_norm": 0.5189673900604248, "learning_rate": 9.015828341188027e-05, "loss": 0.059, "step": 24300 }, { "grad_norm": 0.5573127865791321, "learning_rate": 9.01484305747813e-05, "loss": 0.0619, "step": 24310 }, { "grad_norm": 0.5192828178405762, "learning_rate": 9.013857334711033e-05, "loss": 0.044, "step": 24320 }, { "grad_norm": 0.7322401404380798, "learning_rate": 9.012871172994534e-05, "loss": 0.0474, "step": 24330 }, { "grad_norm": 0.5372781753540039, "learning_rate": 9.011884572436476e-05, "loss": 0.0488, "step": 24340 }, { "grad_norm": 0.4642522633075714, "learning_rate": 9.010897533144754e-05, "loss": 0.0456, "step": 24350 }, { "grad_norm": 0.36471351981163025, "learning_rate": 9.009910055227306e-05, "loss": 0.0483, "step": 24360 }, { "grad_norm": 0.6727672219276428, "learning_rate": 9.008922138792124e-05, "loss": 0.0491, "step": 24370 }, { "grad_norm": 0.5959544777870178, "learning_rate": 9.007933783947244e-05, "loss": 0.0486, "step": 24380 }, { "grad_norm": 0.7003111243247986, "learning_rate": 9.006944990800752e-05, "loss": 0.0402, "step": 24390 }, { "grad_norm": 0.5262545943260193, "learning_rate": 9.005955759460779e-05, "loss": 0.0484, "step": 24400 }, { "grad_norm": 0.3713342547416687, "learning_rate": 9.004966090035508e-05, "loss": 0.0395, "step": 24410 }, { "grad_norm": 0.44530004262924194, "learning_rate": 9.003975982633166e-05, "loss": 0.0443, "step": 24420 }, { "grad_norm": 0.49462491273880005, "learning_rate": 9.00298543736203e-05, "loss": 0.049, "step": 24430 }, { "grad_norm": 0.31982287764549255, "learning_rate": 9.001994454330427e-05, "loss": 0.0491, "step": 24440 }, { "grad_norm": 0.7430485486984253, "learning_rate": 9.001003033646727e-05, "loss": 0.0595, "step": 24450 }, { "grad_norm": 0.35857680439949036, "learning_rate": 9.00001117541935e-05, "loss": 0.0492, "step": 24460 }, { "grad_norm": 0.5138530135154724, "learning_rate": 8.999018879756764e-05, "loss": 0.0439, "step": 24470 }, { "grad_norm": 0.6403698325157166, "learning_rate": 8.998026146767487e-05, "loss": 0.0455, "step": 24480 }, { "grad_norm": 0.5421737432479858, "learning_rate": 8.99703297656008e-05, "loss": 0.0575, "step": 24490 }, { "grad_norm": 0.5969477891921997, "learning_rate": 8.996039369243156e-05, "loss": 0.065, "step": 24500 }, { "grad_norm": 0.777976930141449, "learning_rate": 8.995045324925378e-05, "loss": 0.044, "step": 24510 }, { "grad_norm": 0.523292064666748, "learning_rate": 8.994050843715448e-05, "loss": 0.0635, "step": 24520 }, { "grad_norm": 0.47275805473327637, "learning_rate": 8.993055925722121e-05, "loss": 0.0494, "step": 24530 }, { "grad_norm": 0.591200590133667, "learning_rate": 8.992060571054202e-05, "loss": 0.0482, "step": 24540 }, { "grad_norm": 0.723886251449585, "learning_rate": 8.991064779820542e-05, "loss": 0.047, "step": 24550 }, { "grad_norm": 0.5302926301956177, "learning_rate": 8.990068552130036e-05, "loss": 0.0339, "step": 24560 }, { "grad_norm": 0.37057653069496155, "learning_rate": 8.989071888091634e-05, "loss": 0.0507, "step": 24570 }, { "grad_norm": 0.5783964395523071, "learning_rate": 8.988074787814329e-05, "loss": 0.0517, "step": 24580 }, { "grad_norm": 0.44060617685317993, "learning_rate": 8.987077251407158e-05, "loss": 0.049, "step": 24590 }, { "grad_norm": 0.4870796501636505, "learning_rate": 8.986079278979216e-05, "loss": 0.0457, "step": 24600 }, { "grad_norm": 0.5245763063430786, "learning_rate": 8.985080870639635e-05, "loss": 0.0551, "step": 24610 }, { "grad_norm": 0.3833790719509125, "learning_rate": 8.984082026497603e-05, "loss": 0.041, "step": 24620 }, { "grad_norm": 0.29885420203208923, "learning_rate": 8.98308274666235e-05, "loss": 0.0529, "step": 24630 }, { "grad_norm": 0.4740760326385498, "learning_rate": 8.982083031243155e-05, "loss": 0.0454, "step": 24640 }, { "grad_norm": 0.37409546971321106, "learning_rate": 8.98108288034935e-05, "loss": 0.0428, "step": 24650 }, { "grad_norm": 0.5880039930343628, "learning_rate": 8.980082294090305e-05, "loss": 0.0446, "step": 24660 }, { "grad_norm": 0.4679422378540039, "learning_rate": 8.979081272575443e-05, "loss": 0.052, "step": 24670 }, { "grad_norm": 0.4626926779747009, "learning_rate": 8.978079815914236e-05, "loss": 0.0391, "step": 24680 }, { "grad_norm": 0.49229779839515686, "learning_rate": 8.977077924216202e-05, "loss": 0.044, "step": 24690 }, { "grad_norm": 0.7387017011642456, "learning_rate": 8.976075597590905e-05, "loss": 0.0581, "step": 24700 }, { "grad_norm": 0.40693214535713196, "learning_rate": 8.975072836147958e-05, "loss": 0.0503, "step": 24710 }, { "grad_norm": 0.40757107734680176, "learning_rate": 8.974069639997025e-05, "loss": 0.0468, "step": 24720 }, { "grad_norm": 0.6317933201789856, "learning_rate": 8.973066009247808e-05, "loss": 0.0507, "step": 24730 }, { "grad_norm": 0.5860241651535034, "learning_rate": 8.972061944010066e-05, "loss": 0.0472, "step": 24740 }, { "grad_norm": 0.5533055067062378, "learning_rate": 8.971057444393603e-05, "loss": 0.0517, "step": 24750 }, { "grad_norm": 0.45059865713119507, "learning_rate": 8.970052510508268e-05, "loss": 0.0477, "step": 24760 }, { "grad_norm": 0.4465303122997284, "learning_rate": 8.969047142463959e-05, "loss": 0.0376, "step": 24770 }, { "grad_norm": 1.3057479858398438, "learning_rate": 8.968041340370621e-05, "loss": 0.0529, "step": 24780 }, { "grad_norm": 0.6836382746696472, "learning_rate": 8.96703510433825e-05, "loss": 0.0339, "step": 24790 }, { "grad_norm": 0.8577631711959839, "learning_rate": 8.966028434476883e-05, "loss": 0.0512, "step": 24800 }, { "grad_norm": 1.1802271604537964, "learning_rate": 8.96502133089661e-05, "loss": 0.0502, "step": 24810 }, { "grad_norm": 0.6143561601638794, "learning_rate": 8.964013793707564e-05, "loss": 0.0646, "step": 24820 }, { "grad_norm": 0.44104504585266113, "learning_rate": 8.963005823019932e-05, "loss": 0.0473, "step": 24830 }, { "grad_norm": 0.6164815425872803, "learning_rate": 8.961997418943939e-05, "loss": 0.0376, "step": 24840 }, { "grad_norm": 0.43437516689300537, "learning_rate": 8.960988581589865e-05, "loss": 0.0443, "step": 24850 }, { "grad_norm": 0.4484965205192566, "learning_rate": 8.959979311068037e-05, "loss": 0.0462, "step": 24860 }, { "grad_norm": 0.47397616505622864, "learning_rate": 8.958969607488823e-05, "loss": 0.0585, "step": 24870 }, { "grad_norm": 0.5090229511260986, "learning_rate": 8.957959470962647e-05, "loss": 0.0441, "step": 24880 }, { "grad_norm": 0.4661056101322174, "learning_rate": 8.956948901599971e-05, "loss": 0.0441, "step": 24890 }, { "grad_norm": 0.423111230134964, "learning_rate": 8.955937899511315e-05, "loss": 0.0494, "step": 24900 }, { "grad_norm": 0.6220937371253967, "learning_rate": 8.954926464807238e-05, "loss": 0.0504, "step": 24910 }, { "grad_norm": 0.37263262271881104, "learning_rate": 8.953914597598347e-05, "loss": 0.0428, "step": 24920 }, { "grad_norm": 0.5419948697090149, "learning_rate": 8.952902297995303e-05, "loss": 0.0391, "step": 24930 }, { "grad_norm": 0.9188858866691589, "learning_rate": 8.951889566108804e-05, "loss": 0.0428, "step": 24940 }, { "grad_norm": 1.0625866651535034, "learning_rate": 8.950876402049606e-05, "loss": 0.0506, "step": 24950 }, { "grad_norm": 0.40825074911117554, "learning_rate": 8.949862805928504e-05, "loss": 0.0399, "step": 24960 }, { "grad_norm": 0.5075710415840149, "learning_rate": 8.948848777856343e-05, "loss": 0.0498, "step": 24970 }, { "grad_norm": 0.46731433272361755, "learning_rate": 8.947834317944017e-05, "loss": 0.0451, "step": 24980 }, { "grad_norm": 0.48739227652549744, "learning_rate": 8.946819426302466e-05, "loss": 0.0661, "step": 24990 }, { "grad_norm": 0.5782087445259094, "learning_rate": 8.945804103042676e-05, "loss": 0.0553, "step": 25000 }, { "grad_norm": 0.5149006247520447, "learning_rate": 8.944788348275681e-05, "loss": 0.0467, "step": 25010 }, { "grad_norm": 0.45863276720046997, "learning_rate": 8.943772162112565e-05, "loss": 0.0463, "step": 25020 }, { "grad_norm": 0.7231406569480896, "learning_rate": 8.942755544664454e-05, "loss": 0.0607, "step": 25030 }, { "grad_norm": 0.3373241424560547, "learning_rate": 8.941738496042525e-05, "loss": 0.0515, "step": 25040 }, { "grad_norm": 0.4707395136356354, "learning_rate": 8.940721016357999e-05, "loss": 0.0472, "step": 25050 }, { "grad_norm": 0.3234621286392212, "learning_rate": 8.939703105722148e-05, "loss": 0.0441, "step": 25060 }, { "grad_norm": 0.5556923747062683, "learning_rate": 8.93868476424629e-05, "loss": 0.0318, "step": 25070 }, { "grad_norm": 0.4999253749847412, "learning_rate": 8.937665992041786e-05, "loss": 0.0429, "step": 25080 }, { "grad_norm": 0.5775881409645081, "learning_rate": 8.93664678922005e-05, "loss": 0.052, "step": 25090 }, { "grad_norm": 0.3771820664405823, "learning_rate": 8.93562715589254e-05, "loss": 0.0489, "step": 25100 }, { "grad_norm": 0.7002109885215759, "learning_rate": 8.934607092170762e-05, "loss": 0.0544, "step": 25110 }, { "grad_norm": 0.5906303524971008, "learning_rate": 8.933586598166266e-05, "loss": 0.0493, "step": 25120 }, { "grad_norm": 0.43547868728637695, "learning_rate": 8.932565673990655e-05, "loss": 0.0597, "step": 25130 }, { "grad_norm": 0.4701801538467407, "learning_rate": 8.931544319755574e-05, "loss": 0.0427, "step": 25140 }, { "grad_norm": 0.28895264863967896, "learning_rate": 8.930522535572718e-05, "loss": 0.047, "step": 25150 }, { "grad_norm": 0.44159725308418274, "learning_rate": 8.929500321553826e-05, "loss": 0.0446, "step": 25160 }, { "grad_norm": 0.5244386792182922, "learning_rate": 8.928477677810686e-05, "loss": 0.0407, "step": 25170 }, { "grad_norm": 0.4186660349369049, "learning_rate": 8.927454604455137e-05, "loss": 0.0511, "step": 25180 }, { "grad_norm": 0.5031976699829102, "learning_rate": 8.926431101599053e-05, "loss": 0.057, "step": 25190 }, { "grad_norm": 0.41401737928390503, "learning_rate": 8.925407169354369e-05, "loss": 0.0429, "step": 25200 }, { "grad_norm": 0.5092952847480774, "learning_rate": 8.92438280783306e-05, "loss": 0.0557, "step": 25210 }, { "grad_norm": 0.4278392791748047, "learning_rate": 8.923358017147146e-05, "loss": 0.0515, "step": 25220 }, { "grad_norm": 0.40602177381515503, "learning_rate": 8.922332797408697e-05, "loss": 0.0542, "step": 25230 }, { "grad_norm": 0.3926575481891632, "learning_rate": 8.921307148729831e-05, "loss": 0.0372, "step": 25240 }, { "grad_norm": 0.5068452954292297, "learning_rate": 8.920281071222712e-05, "loss": 0.0467, "step": 25250 }, { "grad_norm": 0.691787600517273, "learning_rate": 8.919254564999548e-05, "loss": 0.0475, "step": 25260 }, { "grad_norm": 0.6025989055633545, "learning_rate": 8.918227630172598e-05, "loss": 0.0408, "step": 25270 }, { "grad_norm": 0.38390424847602844, "learning_rate": 8.917200266854165e-05, "loss": 0.0459, "step": 25280 }, { "grad_norm": 0.4868675768375397, "learning_rate": 8.9161724751566e-05, "loss": 0.0443, "step": 25290 }, { "grad_norm": 0.2771192789077759, "learning_rate": 8.915144255192302e-05, "loss": 0.0377, "step": 25300 }, { "grad_norm": 0.47067537903785706, "learning_rate": 8.914115607073714e-05, "loss": 0.0446, "step": 25310 }, { "grad_norm": 0.34645554423332214, "learning_rate": 8.913086530913327e-05, "loss": 0.0396, "step": 25320 }, { "grad_norm": 0.3806734085083008, "learning_rate": 8.912057026823681e-05, "loss": 0.0346, "step": 25330 }, { "grad_norm": 0.4676400423049927, "learning_rate": 8.91102709491736e-05, "loss": 0.0419, "step": 25340 }, { "grad_norm": 0.5100216865539551, "learning_rate": 8.909996735306996e-05, "loss": 0.053, "step": 25350 }, { "grad_norm": 0.38603439927101135, "learning_rate": 8.908965948105268e-05, "loss": 0.0362, "step": 25360 }, { "grad_norm": 0.4482831358909607, "learning_rate": 8.907934733424901e-05, "loss": 0.0495, "step": 25370 }, { "grad_norm": 0.7251182794570923, "learning_rate": 8.906903091378666e-05, "loss": 0.0418, "step": 25380 }, { "grad_norm": 0.629601001739502, "learning_rate": 8.905871022079384e-05, "loss": 0.0472, "step": 25390 }, { "grad_norm": 0.5213294625282288, "learning_rate": 8.90483852563992e-05, "loss": 0.0427, "step": 25400 }, { "grad_norm": 0.43732815980911255, "learning_rate": 8.903805602173185e-05, "loss": 0.0393, "step": 25410 }, { "grad_norm": 0.6323701739311218, "learning_rate": 8.902772251792137e-05, "loss": 0.043, "step": 25420 }, { "grad_norm": 0.6304997205734253, "learning_rate": 8.901738474609786e-05, "loss": 0.0525, "step": 25430 }, { "grad_norm": 0.6719222664833069, "learning_rate": 8.900704270739179e-05, "loss": 0.0399, "step": 25440 }, { "grad_norm": 0.43904128670692444, "learning_rate": 8.89966964029342e-05, "loss": 0.0529, "step": 25450 }, { "grad_norm": 0.4929118752479553, "learning_rate": 8.898634583385652e-05, "loss": 0.048, "step": 25460 }, { "grad_norm": 0.3406476080417633, "learning_rate": 8.897599100129065e-05, "loss": 0.0466, "step": 25470 }, { "grad_norm": 0.7704930901527405, "learning_rate": 8.896563190636903e-05, "loss": 0.048, "step": 25480 }, { "grad_norm": 0.4918549954891205, "learning_rate": 8.895526855022448e-05, "loss": 0.0475, "step": 25490 }, { "grad_norm": 0.4742252230644226, "learning_rate": 8.894490093399033e-05, "loss": 0.0553, "step": 25500 }, { "grad_norm": 0.2806852459907532, "learning_rate": 8.893452905880035e-05, "loss": 0.05, "step": 25510 }, { "grad_norm": 0.5117797255516052, "learning_rate": 8.892415292578883e-05, "loss": 0.0382, "step": 25520 }, { "grad_norm": 0.369126558303833, "learning_rate": 8.891377253609046e-05, "loss": 0.0451, "step": 25530 }, { "grad_norm": 0.4180552363395691, "learning_rate": 8.890338789084043e-05, "loss": 0.0412, "step": 25540 }, { "grad_norm": 0.4410805404186249, "learning_rate": 8.88929989911744e-05, "loss": 0.0487, "step": 25550 }, { "grad_norm": 0.5801993608474731, "learning_rate": 8.888260583822847e-05, "loss": 0.0535, "step": 25560 }, { "grad_norm": 0.49139419198036194, "learning_rate": 8.887220843313921e-05, "loss": 0.054, "step": 25570 }, { "grad_norm": 0.36647385358810425, "learning_rate": 8.88618067770437e-05, "loss": 0.052, "step": 25580 }, { "grad_norm": 0.38257259130477905, "learning_rate": 8.885140087107942e-05, "loss": 0.0381, "step": 25590 }, { "grad_norm": 0.39734983444213867, "learning_rate": 8.884099071638436e-05, "loss": 0.0448, "step": 25600 }, { "grad_norm": 0.385197252035141, "learning_rate": 8.883057631409695e-05, "loss": 0.0556, "step": 25610 }, { "grad_norm": 0.7324376702308655, "learning_rate": 8.882015766535608e-05, "loss": 0.05, "step": 25620 }, { "grad_norm": 0.587866485118866, "learning_rate": 8.880973477130115e-05, "loss": 0.0454, "step": 25630 }, { "grad_norm": 0.4271971881389618, "learning_rate": 8.879930763307197e-05, "loss": 0.0375, "step": 25640 }, { "grad_norm": 0.370094895362854, "learning_rate": 8.878887625180884e-05, "loss": 0.045, "step": 25650 }, { "grad_norm": 0.39268049597740173, "learning_rate": 8.877844062865253e-05, "loss": 0.0337, "step": 25660 }, { "grad_norm": 0.47035765647888184, "learning_rate": 8.876800076474424e-05, "loss": 0.0578, "step": 25670 }, { "grad_norm": 0.5848228335380554, "learning_rate": 8.875755666122568e-05, "loss": 0.0537, "step": 25680 }, { "grad_norm": 0.3687644302845001, "learning_rate": 8.8747108319239e-05, "loss": 0.0386, "step": 25690 }, { "grad_norm": 0.46587204933166504, "learning_rate": 8.87366557399268e-05, "loss": 0.0418, "step": 25700 }, { "grad_norm": 0.488018274307251, "learning_rate": 8.872619892443217e-05, "loss": 0.0451, "step": 25710 }, { "grad_norm": 0.4836300015449524, "learning_rate": 8.871573787389865e-05, "loss": 0.051, "step": 25720 }, { "grad_norm": 0.4278898537158966, "learning_rate": 8.870527258947024e-05, "loss": 0.0506, "step": 25730 }, { "grad_norm": 0.37532109022140503, "learning_rate": 8.869480307229143e-05, "loss": 0.0409, "step": 25740 }, { "grad_norm": 0.5887818932533264, "learning_rate": 8.868432932350712e-05, "loss": 0.0404, "step": 25750 }, { "grad_norm": 0.501945972442627, "learning_rate": 8.867385134426272e-05, "loss": 0.0363, "step": 25760 }, { "grad_norm": 0.4260500371456146, "learning_rate": 8.866336913570407e-05, "loss": 0.048, "step": 25770 }, { "grad_norm": 0.4665229618549347, "learning_rate": 8.865288269897751e-05, "loss": 0.0317, "step": 25780 }, { "grad_norm": 0.5004241466522217, "learning_rate": 8.864239203522981e-05, "loss": 0.0483, "step": 25790 }, { "grad_norm": 0.41586408019065857, "learning_rate": 8.863189714560822e-05, "loss": 0.0459, "step": 25800 }, { "grad_norm": 0.3256419897079468, "learning_rate": 8.862139803126043e-05, "loss": 0.0376, "step": 25810 }, { "grad_norm": 0.35025593638420105, "learning_rate": 8.861089469333463e-05, "loss": 0.0402, "step": 25820 }, { "grad_norm": 0.48353010416030884, "learning_rate": 8.860038713297944e-05, "loss": 0.0439, "step": 25830 }, { "grad_norm": 1.1535478830337524, "learning_rate": 8.858987535134394e-05, "loss": 0.0662, "step": 25840 }, { "grad_norm": 0.8368587493896484, "learning_rate": 8.857935934957769e-05, "loss": 0.0497, "step": 25850 }, { "grad_norm": 0.4684571921825409, "learning_rate": 8.856883912883071e-05, "loss": 0.0459, "step": 25860 }, { "grad_norm": 0.341118723154068, "learning_rate": 8.855831469025346e-05, "loss": 0.0459, "step": 25870 }, { "grad_norm": 0.41350090503692627, "learning_rate": 8.854778603499689e-05, "loss": 0.0577, "step": 25880 }, { "grad_norm": 0.4497312903404236, "learning_rate": 8.85372531642124e-05, "loss": 0.045, "step": 25890 }, { "grad_norm": 0.5344549417495728, "learning_rate": 8.852671607905185e-05, "loss": 0.0393, "step": 25900 }, { "grad_norm": 0.49999114871025085, "learning_rate": 8.851617478066754e-05, "loss": 0.0466, "step": 25910 }, { "grad_norm": 0.5056952834129333, "learning_rate": 8.850562927021227e-05, "loss": 0.0473, "step": 25920 }, { "grad_norm": 0.5951551198959351, "learning_rate": 8.849507954883928e-05, "loss": 0.0499, "step": 25930 }, { "grad_norm": 0.6042812466621399, "learning_rate": 8.848452561770226e-05, "loss": 0.0455, "step": 25940 }, { "grad_norm": 0.32100191712379456, "learning_rate": 8.847396747795538e-05, "loss": 0.0465, "step": 25950 }, { "grad_norm": 0.48406684398651123, "learning_rate": 8.846340513075327e-05, "loss": 0.0501, "step": 25960 }, { "grad_norm": 0.43512266874313354, "learning_rate": 8.845283857725099e-05, "loss": 0.0437, "step": 25970 }, { "grad_norm": 0.4502815008163452, "learning_rate": 8.844226781860409e-05, "loss": 0.0486, "step": 25980 }, { "grad_norm": 0.44893205165863037, "learning_rate": 8.84316928559686e-05, "loss": 0.0475, "step": 25990 }, { "grad_norm": 0.29303833842277527, "learning_rate": 8.842111369050094e-05, "loss": 0.0368, "step": 26000 }, { "grad_norm": 0.531441330909729, "learning_rate": 8.841053032335808e-05, "loss": 0.0416, "step": 26010 }, { "grad_norm": 0.4999118447303772, "learning_rate": 8.839994275569735e-05, "loss": 0.0389, "step": 26020 }, { "grad_norm": 0.5298131704330444, "learning_rate": 8.838935098867662e-05, "loss": 0.0499, "step": 26030 }, { "grad_norm": 0.5716720819473267, "learning_rate": 8.837875502345418e-05, "loss": 0.0587, "step": 26040 }, { "grad_norm": 0.4766958951950073, "learning_rate": 8.83681548611888e-05, "loss": 0.0459, "step": 26050 }, { "grad_norm": 0.34204304218292236, "learning_rate": 8.835755050303969e-05, "loss": 0.0459, "step": 26060 }, { "grad_norm": 0.4187299311161041, "learning_rate": 8.834694195016653e-05, "loss": 0.046, "step": 26070 }, { "grad_norm": 0.5572153329849243, "learning_rate": 8.833632920372942e-05, "loss": 0.0612, "step": 26080 }, { "grad_norm": 0.5572422742843628, "learning_rate": 8.832571226488903e-05, "loss": 0.0658, "step": 26090 }, { "grad_norm": 0.5319986939430237, "learning_rate": 8.831509113480634e-05, "loss": 0.0585, "step": 26100 }, { "grad_norm": 1.431232213973999, "learning_rate": 8.83044658146429e-05, "loss": 0.0461, "step": 26110 }, { "grad_norm": 0.4763292968273163, "learning_rate": 8.829383630556067e-05, "loss": 0.0444, "step": 26120 }, { "grad_norm": 0.4700663387775421, "learning_rate": 8.828320260872207e-05, "loss": 0.0478, "step": 26130 }, { "grad_norm": 0.5054323673248291, "learning_rate": 8.827256472529e-05, "loss": 0.0436, "step": 26140 }, { "grad_norm": 0.35899224877357483, "learning_rate": 8.826192265642778e-05, "loss": 0.0428, "step": 26150 }, { "grad_norm": 0.5303447842597961, "learning_rate": 8.825127640329923e-05, "loss": 0.0422, "step": 26160 }, { "grad_norm": 0.4225934147834778, "learning_rate": 8.824062596706861e-05, "loss": 0.0458, "step": 26170 }, { "grad_norm": 0.38148483633995056, "learning_rate": 8.822997134890062e-05, "loss": 0.0388, "step": 26180 }, { "grad_norm": 0.5177087783813477, "learning_rate": 8.821931254996044e-05, "loss": 0.0493, "step": 26190 }, { "grad_norm": 0.611568808555603, "learning_rate": 8.82086495714137e-05, "loss": 0.0492, "step": 26200 }, { "grad_norm": 0.5499799847602844, "learning_rate": 8.81979824144265e-05, "loss": 0.0446, "step": 26210 }, { "grad_norm": 0.7274911999702454, "learning_rate": 8.818731108016536e-05, "loss": 0.0498, "step": 26220 }, { "grad_norm": 0.40113502740859985, "learning_rate": 8.81766355697973e-05, "loss": 0.0511, "step": 26230 }, { "grad_norm": 0.48961764574050903, "learning_rate": 8.816595588448977e-05, "loss": 0.0497, "step": 26240 }, { "grad_norm": 0.6244826316833496, "learning_rate": 8.81552720254107e-05, "loss": 0.0571, "step": 26250 }, { "grad_norm": 0.4569128453731537, "learning_rate": 8.814458399372842e-05, "loss": 0.042, "step": 26260 }, { "grad_norm": 0.37220388650894165, "learning_rate": 8.813389179061181e-05, "loss": 0.0392, "step": 26270 }, { "grad_norm": 0.5164251327514648, "learning_rate": 8.812319541723012e-05, "loss": 0.0533, "step": 26280 }, { "grad_norm": 0.5139145851135254, "learning_rate": 8.811249487475309e-05, "loss": 0.065, "step": 26290 }, { "grad_norm": 0.5763834714889526, "learning_rate": 8.810179016435092e-05, "loss": 0.0567, "step": 26300 }, { "grad_norm": 0.5295284390449524, "learning_rate": 8.809108128719428e-05, "loss": 0.0753, "step": 26310 }, { "grad_norm": 0.7150000929832458, "learning_rate": 8.808036824445424e-05, "loss": 0.0498, "step": 26320 }, { "grad_norm": 0.44953247904777527, "learning_rate": 8.806965103730238e-05, "loss": 0.0536, "step": 26330 }, { "grad_norm": 0.3906083405017853, "learning_rate": 8.805892966691074e-05, "loss": 0.0464, "step": 26340 }, { "grad_norm": 0.2944808006286621, "learning_rate": 8.804820413445175e-05, "loss": 0.0421, "step": 26350 }, { "grad_norm": 0.3767523765563965, "learning_rate": 8.803747444109837e-05, "loss": 0.0513, "step": 26360 }, { "grad_norm": 0.49999281764030457, "learning_rate": 8.802674058802399e-05, "loss": 0.0393, "step": 26370 }, { "grad_norm": 0.39936885237693787, "learning_rate": 8.801600257640241e-05, "loss": 0.0401, "step": 26380 }, { "grad_norm": 0.37591496109962463, "learning_rate": 8.800526040740795e-05, "loss": 0.042, "step": 26390 }, { "grad_norm": 0.7211046814918518, "learning_rate": 8.799451408221535e-05, "loss": 0.0521, "step": 26400 }, { "grad_norm": 0.571694016456604, "learning_rate": 8.798376360199982e-05, "loss": 0.0611, "step": 26410 }, { "grad_norm": 0.5004188418388367, "learning_rate": 8.797300896793701e-05, "loss": 0.0417, "step": 26420 }, { "grad_norm": 0.7119616270065308, "learning_rate": 8.796225018120302e-05, "loss": 0.0512, "step": 26430 }, { "grad_norm": 0.28978231549263, "learning_rate": 8.795148724297444e-05, "loss": 0.0516, "step": 26440 }, { "grad_norm": 0.5139867067337036, "learning_rate": 8.794072015442825e-05, "loss": 0.0512, "step": 26450 }, { "grad_norm": 0.4705303907394409, "learning_rate": 8.792994891674198e-05, "loss": 0.0593, "step": 26460 }, { "grad_norm": 0.8884973526000977, "learning_rate": 8.79191735310935e-05, "loss": 0.0506, "step": 26470 }, { "grad_norm": 0.4323621690273285, "learning_rate": 8.790839399866122e-05, "loss": 0.0442, "step": 26480 }, { "grad_norm": 0.48924848437309265, "learning_rate": 8.789761032062397e-05, "loss": 0.0538, "step": 26490 }, { "grad_norm": 0.6849501132965088, "learning_rate": 8.788682249816103e-05, "loss": 0.0484, "step": 26500 }, { "grad_norm": 0.40282607078552246, "learning_rate": 8.787603053245215e-05, "loss": 0.0527, "step": 26510 }, { "grad_norm": 0.36406397819519043, "learning_rate": 8.78652344246775e-05, "loss": 0.0379, "step": 26520 }, { "grad_norm": 0.6290378570556641, "learning_rate": 8.785443417601776e-05, "loss": 0.048, "step": 26530 }, { "grad_norm": 0.47465139627456665, "learning_rate": 8.784362978765401e-05, "loss": 0.0575, "step": 26540 }, { "grad_norm": 0.48689982295036316, "learning_rate": 8.783282126076779e-05, "loss": 0.0432, "step": 26550 }, { "grad_norm": 0.3522404134273529, "learning_rate": 8.782200859654112e-05, "loss": 0.0481, "step": 26560 }, { "grad_norm": 1.912695288658142, "learning_rate": 8.781119179615646e-05, "loss": 0.0496, "step": 26570 }, { "grad_norm": 0.6084219217300415, "learning_rate": 8.780037086079674e-05, "loss": 0.05, "step": 26580 }, { "grad_norm": 0.6647035479545593, "learning_rate": 8.778954579164527e-05, "loss": 0.047, "step": 26590 }, { "grad_norm": 0.5399903059005737, "learning_rate": 8.777871658988588e-05, "loss": 0.0453, "step": 26600 }, { "grad_norm": 0.3967258036136627, "learning_rate": 8.776788325670285e-05, "loss": 0.0417, "step": 26610 }, { "grad_norm": 0.40191105008125305, "learning_rate": 8.775704579328089e-05, "loss": 0.0496, "step": 26620 }, { "grad_norm": 0.42371851205825806, "learning_rate": 8.774620420080517e-05, "loss": 0.0505, "step": 26630 }, { "grad_norm": 1.0007312297821045, "learning_rate": 8.773535848046131e-05, "loss": 0.0531, "step": 26640 }, { "grad_norm": 0.46668577194213867, "learning_rate": 8.772450863343538e-05, "loss": 0.0395, "step": 26650 }, { "grad_norm": 0.9328535795211792, "learning_rate": 8.77136546609139e-05, "loss": 0.0652, "step": 26660 }, { "grad_norm": 2.822657823562622, "learning_rate": 8.770279656408385e-05, "loss": 0.0414, "step": 26670 }, { "grad_norm": 0.4718461036682129, "learning_rate": 8.769193434413265e-05, "loss": 0.0587, "step": 26680 }, { "grad_norm": 0.4063887596130371, "learning_rate": 8.76810680022482e-05, "loss": 0.0538, "step": 26690 }, { "grad_norm": 0.8192008137702942, "learning_rate": 8.767019753961878e-05, "loss": 0.0662, "step": 26700 }, { "grad_norm": 0.3904474377632141, "learning_rate": 8.765932295743321e-05, "loss": 0.0497, "step": 26710 }, { "grad_norm": 0.28833097219467163, "learning_rate": 8.764844425688068e-05, "loss": 0.038, "step": 26720 }, { "grad_norm": 0.5233030915260315, "learning_rate": 8.763756143915092e-05, "loss": 0.0375, "step": 26730 }, { "grad_norm": 1.7609364986419678, "learning_rate": 8.7626674505434e-05, "loss": 0.0462, "step": 26740 }, { "grad_norm": 0.40483638644218445, "learning_rate": 8.761578345692053e-05, "loss": 0.0625, "step": 26750 }, { "grad_norm": 0.5350319743156433, "learning_rate": 8.760488829480156e-05, "loss": 0.0484, "step": 26760 }, { "grad_norm": 0.30753907561302185, "learning_rate": 8.759398902026854e-05, "loss": 0.0348, "step": 26770 }, { "grad_norm": 0.3707515597343445, "learning_rate": 8.758308563451339e-05, "loss": 0.0424, "step": 26780 }, { "grad_norm": 0.39103278517723083, "learning_rate": 8.75721781387285e-05, "loss": 0.0422, "step": 26790 }, { "grad_norm": 0.319448322057724, "learning_rate": 8.75612665341067e-05, "loss": 0.0381, "step": 26800 }, { "grad_norm": 0.27769094705581665, "learning_rate": 8.755035082184126e-05, "loss": 0.0345, "step": 26810 }, { "grad_norm": 0.3754071295261383, "learning_rate": 8.753943100312592e-05, "loss": 0.0349, "step": 26820 }, { "grad_norm": 0.4059769809246063, "learning_rate": 8.752850707915484e-05, "loss": 0.0348, "step": 26830 }, { "grad_norm": 0.7737779021263123, "learning_rate": 8.751757905112264e-05, "loss": 0.0622, "step": 26840 }, { "grad_norm": 0.44099554419517517, "learning_rate": 8.75066469202244e-05, "loss": 0.0382, "step": 26850 }, { "grad_norm": 0.37583598494529724, "learning_rate": 8.749571068765567e-05, "loss": 0.0494, "step": 26860 }, { "grad_norm": 0.5377492904663086, "learning_rate": 8.748477035461238e-05, "loss": 0.0458, "step": 26870 }, { "grad_norm": 0.6076768040657043, "learning_rate": 8.747382592229095e-05, "loss": 0.0484, "step": 26880 }, { "grad_norm": 0.5977025628089905, "learning_rate": 8.746287739188828e-05, "loss": 0.0529, "step": 26890 }, { "grad_norm": 0.3754562437534332, "learning_rate": 8.745192476460165e-05, "loss": 0.0465, "step": 26900 }, { "grad_norm": 0.46383073925971985, "learning_rate": 8.744096804162882e-05, "loss": 0.0498, "step": 26910 }, { "grad_norm": 0.5775179862976074, "learning_rate": 8.743000722416804e-05, "loss": 0.0543, "step": 26920 }, { "grad_norm": 0.3196276128292084, "learning_rate": 8.741904231341793e-05, "loss": 0.0447, "step": 26930 }, { "grad_norm": 0.7570444345474243, "learning_rate": 8.740807331057762e-05, "loss": 0.0612, "step": 26940 }, { "grad_norm": 0.6378828883171082, "learning_rate": 8.739710021684667e-05, "loss": 0.0516, "step": 26950 }, { "grad_norm": 0.6082201600074768, "learning_rate": 8.738612303342503e-05, "loss": 0.0442, "step": 26960 }, { "grad_norm": 0.34438756108283997, "learning_rate": 8.73751417615132e-05, "loss": 0.0506, "step": 26970 }, { "grad_norm": 1.4006565809249878, "learning_rate": 8.736415640231208e-05, "loss": 0.0497, "step": 26980 }, { "grad_norm": 0.5298316478729248, "learning_rate": 8.735316695702297e-05, "loss": 0.0499, "step": 26990 }, { "grad_norm": 0.4633185863494873, "learning_rate": 8.734217342684769e-05, "loss": 0.0441, "step": 27000 }, { "grad_norm": 0.5254681706428528, "learning_rate": 8.733117581298847e-05, "loss": 0.0388, "step": 27010 }, { "grad_norm": 1.419663906097412, "learning_rate": 8.732017411664796e-05, "loss": 0.0436, "step": 27020 }, { "grad_norm": 0.461226224899292, "learning_rate": 8.730916833902936e-05, "loss": 0.0492, "step": 27030 }, { "grad_norm": 0.5935158729553223, "learning_rate": 8.729815848133618e-05, "loss": 0.0525, "step": 27040 }, { "grad_norm": 0.3777996599674225, "learning_rate": 8.728714454477247e-05, "loss": 0.0421, "step": 27050 }, { "grad_norm": 0.4131649434566498, "learning_rate": 8.727612653054269e-05, "loss": 0.0457, "step": 27060 }, { "grad_norm": 0.380674809217453, "learning_rate": 8.726510443985176e-05, "loss": 0.049, "step": 27070 }, { "grad_norm": 0.3860926628112793, "learning_rate": 8.725407827390503e-05, "loss": 0.0443, "step": 27080 }, { "grad_norm": 0.5460848212242126, "learning_rate": 8.724304803390833e-05, "loss": 0.0513, "step": 27090 }, { "grad_norm": 0.5465627312660217, "learning_rate": 8.723201372106788e-05, "loss": 0.0443, "step": 27100 }, { "grad_norm": 0.6367135047912598, "learning_rate": 8.722097533659038e-05, "loss": 0.0515, "step": 27110 }, { "grad_norm": 0.6030886769294739, "learning_rate": 8.720993288168299e-05, "loss": 0.048, "step": 27120 }, { "grad_norm": 0.5986776351928711, "learning_rate": 8.719888635755327e-05, "loss": 0.0502, "step": 27130 }, { "grad_norm": 0.3418634533882141, "learning_rate": 8.718783576540928e-05, "loss": 0.0507, "step": 27140 }, { "grad_norm": 0.5240375995635986, "learning_rate": 8.717678110645948e-05, "loss": 0.0504, "step": 27150 }, { "grad_norm": 0.4299788177013397, "learning_rate": 8.716572238191279e-05, "loss": 0.0468, "step": 27160 }, { "grad_norm": 0.657691478729248, "learning_rate": 8.715465959297857e-05, "loss": 0.061, "step": 27170 }, { "grad_norm": 0.8894374370574951, "learning_rate": 8.714359274086665e-05, "loss": 0.0667, "step": 27180 }, { "grad_norm": 0.5226672887802124, "learning_rate": 8.713252182678726e-05, "loss": 0.0484, "step": 27190 }, { "grad_norm": 0.47503820061683655, "learning_rate": 8.712144685195112e-05, "loss": 0.0559, "step": 27200 }, { "grad_norm": 0.36669859290122986, "learning_rate": 8.711036781756936e-05, "loss": 0.0502, "step": 27210 }, { "grad_norm": 0.4028435945510864, "learning_rate": 8.709928472485357e-05, "loss": 0.0379, "step": 27220 }, { "grad_norm": 0.41120150685310364, "learning_rate": 8.708819757501579e-05, "loss": 0.0383, "step": 27230 }, { "grad_norm": 0.27098777890205383, "learning_rate": 8.707710636926846e-05, "loss": 0.0431, "step": 27240 }, { "grad_norm": 0.424160361289978, "learning_rate": 8.706601110882455e-05, "loss": 0.0563, "step": 27250 }, { "grad_norm": 0.43715396523475647, "learning_rate": 8.705491179489738e-05, "loss": 0.0424, "step": 27260 }, { "grad_norm": 0.6235365867614746, "learning_rate": 8.704380842870077e-05, "loss": 0.0474, "step": 27270 }, { "grad_norm": 0.440600723028183, "learning_rate": 8.703270101144895e-05, "loss": 0.0422, "step": 27280 }, { "grad_norm": 0.4090828001499176, "learning_rate": 8.702158954435664e-05, "loss": 0.0415, "step": 27290 }, { "grad_norm": 0.5119929313659668, "learning_rate": 8.701047402863896e-05, "loss": 0.0384, "step": 27300 }, { "grad_norm": 0.5941724181175232, "learning_rate": 8.699935446551148e-05, "loss": 0.046, "step": 27310 }, { "grad_norm": 0.3611413836479187, "learning_rate": 8.698823085619022e-05, "loss": 0.0405, "step": 27320 }, { "grad_norm": 0.3929372727870941, "learning_rate": 8.697710320189166e-05, "loss": 0.0445, "step": 27330 }, { "grad_norm": 0.5723827481269836, "learning_rate": 8.696597150383268e-05, "loss": 0.0448, "step": 27340 }, { "grad_norm": 0.5563238859176636, "learning_rate": 8.695483576323063e-05, "loss": 0.05, "step": 27350 }, { "grad_norm": 0.3852068781852722, "learning_rate": 8.69436959813033e-05, "loss": 0.0548, "step": 27360 }, { "grad_norm": 0.3242036998271942, "learning_rate": 8.693255215926892e-05, "loss": 0.0461, "step": 27370 }, { "grad_norm": 0.35730278491973877, "learning_rate": 8.692140429834617e-05, "loss": 0.0413, "step": 27380 }, { "grad_norm": 0.47895944118499756, "learning_rate": 8.691025239975415e-05, "loss": 0.0521, "step": 27390 }, { "grad_norm": 0.6044822931289673, "learning_rate": 8.689909646471243e-05, "loss": 0.0424, "step": 27400 }, { "grad_norm": 0.2830978333950043, "learning_rate": 8.688793649444099e-05, "loss": 0.0521, "step": 27410 }, { "grad_norm": 0.6194770336151123, "learning_rate": 8.687677249016029e-05, "loss": 0.0494, "step": 27420 }, { "grad_norm": 0.4750068783760071, "learning_rate": 8.686560445309118e-05, "loss": 0.0479, "step": 27430 }, { "grad_norm": 0.5452364087104797, "learning_rate": 8.685443238445499e-05, "loss": 0.0495, "step": 27440 }, { "grad_norm": 0.406810462474823, "learning_rate": 8.68432562854735e-05, "loss": 0.048, "step": 27450 }, { "grad_norm": 0.3962269425392151, "learning_rate": 8.683207615736887e-05, "loss": 0.0389, "step": 27460 }, { "grad_norm": 0.4759919345378876, "learning_rate": 8.682089200136379e-05, "loss": 0.0598, "step": 27470 }, { "grad_norm": 0.4369910657405853, "learning_rate": 8.680970381868132e-05, "loss": 0.0389, "step": 27480 }, { "grad_norm": 0.41898876428604126, "learning_rate": 8.679851161054498e-05, "loss": 0.05, "step": 27490 }, { "grad_norm": 0.5455673933029175, "learning_rate": 8.678731537817873e-05, "loss": 0.0405, "step": 27500 }, { "grad_norm": 0.5174407362937927, "learning_rate": 8.677611512280697e-05, "loss": 0.0402, "step": 27510 }, { "grad_norm": 0.29934370517730713, "learning_rate": 8.676491084565457e-05, "loss": 0.0489, "step": 27520 }, { "grad_norm": 0.4606130123138428, "learning_rate": 8.675370254794678e-05, "loss": 0.0562, "step": 27530 }, { "grad_norm": 0.4297407865524292, "learning_rate": 8.674249023090935e-05, "loss": 0.0513, "step": 27540 }, { "grad_norm": 0.5004755258560181, "learning_rate": 8.673127389576843e-05, "loss": 0.0486, "step": 27550 }, { "grad_norm": 0.5661733150482178, "learning_rate": 8.67200535437506e-05, "loss": 0.0427, "step": 27560 }, { "grad_norm": 0.5643385648727417, "learning_rate": 8.670882917608296e-05, "loss": 0.0561, "step": 27570 }, { "grad_norm": 0.7664742469787598, "learning_rate": 8.669760079399292e-05, "loss": 0.0449, "step": 27580 }, { "grad_norm": 0.6514461040496826, "learning_rate": 8.668636839870845e-05, "loss": 0.0402, "step": 27590 }, { "grad_norm": 0.41977038979530334, "learning_rate": 8.667513199145789e-05, "loss": 0.0499, "step": 27600 }, { "grad_norm": 0.38575154542922974, "learning_rate": 8.666389157347002e-05, "loss": 0.0477, "step": 27610 }, { "grad_norm": 0.6751924157142639, "learning_rate": 8.66526471459741e-05, "loss": 0.0419, "step": 27620 }, { "grad_norm": 0.6737375855445862, "learning_rate": 8.66413987101998e-05, "loss": 0.0406, "step": 27630 }, { "grad_norm": 0.34821170568466187, "learning_rate": 8.663014626737723e-05, "loss": 0.0396, "step": 27640 }, { "grad_norm": 0.4082823395729065, "learning_rate": 8.661888981873691e-05, "loss": 0.0472, "step": 27650 }, { "grad_norm": 0.6861693859100342, "learning_rate": 8.660762936550988e-05, "loss": 0.0524, "step": 27660 }, { "grad_norm": 0.5130075812339783, "learning_rate": 8.659636490892753e-05, "loss": 0.0502, "step": 27670 }, { "grad_norm": 0.4429989159107208, "learning_rate": 8.658509645022174e-05, "loss": 0.0496, "step": 27680 }, { "grad_norm": 0.505844235420227, "learning_rate": 8.657382399062481e-05, "loss": 0.0377, "step": 27690 }, { "grad_norm": 0.4676674008369446, "learning_rate": 8.656254753136946e-05, "loss": 0.0515, "step": 27700 }, { "grad_norm": 0.41031500697135925, "learning_rate": 8.655126707368891e-05, "loss": 0.0347, "step": 27710 }, { "grad_norm": 0.3989253342151642, "learning_rate": 8.653998261881672e-05, "loss": 0.042, "step": 27720 }, { "grad_norm": 0.39280402660369873, "learning_rate": 8.652869416798699e-05, "loss": 0.0481, "step": 27730 }, { "grad_norm": 0.455786794424057, "learning_rate": 8.651740172243417e-05, "loss": 0.0445, "step": 27740 }, { "grad_norm": 0.44789981842041016, "learning_rate": 8.65061052833932e-05, "loss": 0.0416, "step": 27750 }, { "grad_norm": 0.3439488708972931, "learning_rate": 8.649480485209945e-05, "loss": 0.0358, "step": 27760 }, { "grad_norm": 0.4393503665924072, "learning_rate": 8.64835004297887e-05, "loss": 0.0366, "step": 27770 }, { "grad_norm": 0.4955854117870331, "learning_rate": 8.64721920176972e-05, "loss": 0.0519, "step": 27780 }, { "grad_norm": 0.466109037399292, "learning_rate": 8.646087961706164e-05, "loss": 0.0547, "step": 27790 }, { "grad_norm": 0.44086819887161255, "learning_rate": 8.644956322911908e-05, "loss": 0.0445, "step": 27800 }, { "grad_norm": 0.5399383306503296, "learning_rate": 8.643824285510709e-05, "loss": 0.0513, "step": 27810 }, { "grad_norm": 0.304101824760437, "learning_rate": 8.642691849626364e-05, "loss": 0.0395, "step": 27820 }, { "grad_norm": 0.5261934399604797, "learning_rate": 8.641559015382717e-05, "loss": 0.0382, "step": 27830 }, { "grad_norm": 0.9961673617362976, "learning_rate": 8.640425782903649e-05, "loss": 0.0556, "step": 27840 }, { "grad_norm": 0.4032561182975769, "learning_rate": 8.639292152313091e-05, "loss": 0.0505, "step": 27850 }, { "grad_norm": 0.5034355521202087, "learning_rate": 8.638158123735015e-05, "loss": 0.0431, "step": 27860 }, { "grad_norm": 0.4398230314254761, "learning_rate": 8.637023697293436e-05, "loss": 0.0361, "step": 27870 }, { "grad_norm": 0.5875064134597778, "learning_rate": 8.635888873112414e-05, "loss": 0.0549, "step": 27880 }, { "grad_norm": 0.46117159724235535, "learning_rate": 8.634753651316052e-05, "loss": 0.0414, "step": 27890 }, { "grad_norm": 0.37604400515556335, "learning_rate": 8.633618032028496e-05, "loss": 0.0504, "step": 27900 }, { "grad_norm": 0.37418442964553833, "learning_rate": 8.632482015373934e-05, "loss": 0.0446, "step": 27910 }, { "grad_norm": 0.5990854501724243, "learning_rate": 8.6313456014766e-05, "loss": 0.0586, "step": 27920 }, { "grad_norm": 0.3829231560230255, "learning_rate": 8.630208790460771e-05, "loss": 0.0515, "step": 27930 }, { "grad_norm": 0.6035944819450378, "learning_rate": 8.629071582450768e-05, "loss": 0.0433, "step": 27940 }, { "grad_norm": 0.5549357533454895, "learning_rate": 8.62793397757095e-05, "loss": 0.0422, "step": 27950 }, { "grad_norm": 0.6605728268623352, "learning_rate": 8.626795975945729e-05, "loss": 0.058, "step": 27960 }, { "grad_norm": 0.47910264134407043, "learning_rate": 8.625657577699551e-05, "loss": 0.0402, "step": 27970 }, { "grad_norm": 0.5499493479728699, "learning_rate": 8.624518782956914e-05, "loss": 0.0399, "step": 27980 }, { "grad_norm": 0.30308032035827637, "learning_rate": 8.62337959184235e-05, "loss": 0.0382, "step": 27990 }, { "grad_norm": 0.523891031742096, "learning_rate": 8.622240004480441e-05, "loss": 0.0521, "step": 28000 }, { "grad_norm": 0.3403729796409607, "learning_rate": 8.621100020995814e-05, "loss": 0.0392, "step": 28010 }, { "grad_norm": 0.4384777545928955, "learning_rate": 8.619959641513132e-05, "loss": 0.0435, "step": 28020 }, { "grad_norm": 0.42329007387161255, "learning_rate": 8.618818866157105e-05, "loss": 0.0396, "step": 28030 }, { "grad_norm": 0.5878270268440247, "learning_rate": 8.617677695052487e-05, "loss": 0.0429, "step": 28040 }, { "grad_norm": 0.2817266583442688, "learning_rate": 8.616536128324078e-05, "loss": 0.0411, "step": 28050 }, { "grad_norm": 0.4237925112247467, "learning_rate": 8.615394166096712e-05, "loss": 0.033, "step": 28060 }, { "grad_norm": 0.7230992913246155, "learning_rate": 8.614251808495279e-05, "loss": 0.0481, "step": 28070 }, { "grad_norm": 0.3776262700557709, "learning_rate": 8.6131090556447e-05, "loss": 0.0443, "step": 28080 }, { "grad_norm": 0.4826192557811737, "learning_rate": 8.611965907669947e-05, "loss": 0.0441, "step": 28090 }, { "grad_norm": 0.5245863199234009, "learning_rate": 8.610822364696034e-05, "loss": 0.0472, "step": 28100 }, { "grad_norm": 0.5459203124046326, "learning_rate": 8.609678426848015e-05, "loss": 0.0505, "step": 28110 }, { "grad_norm": 0.5919508337974548, "learning_rate": 8.60853409425099e-05, "loss": 0.0479, "step": 28120 }, { "grad_norm": 1.187495470046997, "learning_rate": 8.607389367030104e-05, "loss": 0.0564, "step": 28130 }, { "grad_norm": 0.35676831007003784, "learning_rate": 8.606244245310538e-05, "loss": 0.0549, "step": 28140 }, { "grad_norm": 0.6324817538261414, "learning_rate": 8.605098729217525e-05, "loss": 0.0549, "step": 28150 }, { "grad_norm": 0.3781352937221527, "learning_rate": 8.603952818876335e-05, "loss": 0.0493, "step": 28160 }, { "grad_norm": 0.45784398913383484, "learning_rate": 8.602806514412281e-05, "loss": 0.0451, "step": 28170 }, { "grad_norm": 0.3803609311580658, "learning_rate": 8.601659815950726e-05, "loss": 0.0523, "step": 28180 }, { "grad_norm": 0.530269980430603, "learning_rate": 8.600512723617067e-05, "loss": 0.0586, "step": 28190 }, { "grad_norm": 0.5591385364532471, "learning_rate": 8.59936523753675e-05, "loss": 0.0507, "step": 28200 }, { "grad_norm": 0.396678626537323, "learning_rate": 8.598217357835264e-05, "loss": 0.0439, "step": 28210 }, { "grad_norm": 0.3805195689201355, "learning_rate": 8.597069084638135e-05, "loss": 0.039, "step": 28220 }, { "grad_norm": 0.42077916860580444, "learning_rate": 8.595920418070939e-05, "loss": 0.0371, "step": 28230 }, { "grad_norm": 0.4275539219379425, "learning_rate": 8.594771358259295e-05, "loss": 0.0301, "step": 28240 }, { "grad_norm": 0.44919613003730774, "learning_rate": 8.593621905328858e-05, "loss": 0.0581, "step": 28250 }, { "grad_norm": 0.3704002797603607, "learning_rate": 8.592472059405333e-05, "loss": 0.0436, "step": 28260 }, { "grad_norm": 0.6004627346992493, "learning_rate": 8.591321820614464e-05, "loss": 0.0345, "step": 28270 }, { "grad_norm": 0.29534032940864563, "learning_rate": 8.590171189082041e-05, "loss": 0.0402, "step": 28280 }, { "grad_norm": 0.4498395621776581, "learning_rate": 8.589020164933894e-05, "loss": 0.0481, "step": 28290 }, { "grad_norm": 0.34259605407714844, "learning_rate": 8.587868748295898e-05, "loss": 0.0465, "step": 28300 }, { "grad_norm": 0.38065865635871887, "learning_rate": 8.586716939293971e-05, "loss": 0.0474, "step": 28310 }, { "grad_norm": 0.5739007592201233, "learning_rate": 8.58556473805407e-05, "loss": 0.0506, "step": 28320 }, { "grad_norm": 0.3051367998123169, "learning_rate": 8.584412144702202e-05, "loss": 0.0464, "step": 28330 }, { "grad_norm": 0.5111303329467773, "learning_rate": 8.58325915936441e-05, "loss": 0.045, "step": 28340 }, { "grad_norm": 0.5259242653846741, "learning_rate": 8.582105782166783e-05, "loss": 0.0427, "step": 28350 }, { "grad_norm": 0.30136364698410034, "learning_rate": 8.580952013235455e-05, "loss": 0.0386, "step": 28360 }, { "grad_norm": 0.3579998314380646, "learning_rate": 8.579797852696596e-05, "loss": 0.0428, "step": 28370 }, { "grad_norm": 0.5084871053695679, "learning_rate": 8.578643300676428e-05, "loss": 0.0576, "step": 28380 }, { "grad_norm": 0.5482230186462402, "learning_rate": 8.577488357301209e-05, "loss": 0.0485, "step": 28390 }, { "grad_norm": 0.44982266426086426, "learning_rate": 8.576333022697242e-05, "loss": 0.0419, "step": 28400 }, { "grad_norm": 0.527641773223877, "learning_rate": 8.575177296990873e-05, "loss": 0.0377, "step": 28410 }, { "grad_norm": 0.6829428672790527, "learning_rate": 8.574021180308489e-05, "loss": 0.0497, "step": 28420 }, { "grad_norm": 0.5284566879272461, "learning_rate": 8.572864672776523e-05, "loss": 0.0379, "step": 28430 }, { "grad_norm": 0.45375901460647583, "learning_rate": 8.571707774521447e-05, "loss": 0.0338, "step": 28440 }, { "grad_norm": 0.26863348484039307, "learning_rate": 8.57055048566978e-05, "loss": 0.0369, "step": 28450 }, { "grad_norm": 0.5052701830863953, "learning_rate": 8.569392806348078e-05, "loss": 0.0409, "step": 28460 }, { "grad_norm": 0.3883516192436218, "learning_rate": 8.568234736682947e-05, "loss": 0.0384, "step": 28470 }, { "grad_norm": 0.5233781933784485, "learning_rate": 8.567076276801029e-05, "loss": 0.0578, "step": 28480 }, { "grad_norm": 0.577996551990509, "learning_rate": 8.565917426829013e-05, "loss": 0.045, "step": 28490 }, { "grad_norm": 0.37163567543029785, "learning_rate": 8.564758186893628e-05, "loss": 0.0442, "step": 28500 }, { "grad_norm": 0.5357582569122314, "learning_rate": 8.563598557121649e-05, "loss": 0.0459, "step": 28510 }, { "grad_norm": 0.41617029905319214, "learning_rate": 8.562438537639888e-05, "loss": 0.0429, "step": 28520 }, { "grad_norm": 0.45083069801330566, "learning_rate": 8.561278128575206e-05, "loss": 0.063, "step": 28530 }, { "grad_norm": 0.34460780024528503, "learning_rate": 8.5601173300545e-05, "loss": 0.036, "step": 28540 }, { "grad_norm": 0.8429279327392578, "learning_rate": 8.558956142204717e-05, "loss": 0.059, "step": 28550 }, { "grad_norm": 0.7274558544158936, "learning_rate": 8.55779456515284e-05, "loss": 0.0567, "step": 28560 }, { "grad_norm": 0.5377923846244812, "learning_rate": 8.556632599025898e-05, "loss": 0.0504, "step": 28570 }, { "grad_norm": 0.4748116433620453, "learning_rate": 8.555470243950964e-05, "loss": 0.0438, "step": 28580 }, { "grad_norm": 0.5533108711242676, "learning_rate": 8.554307500055148e-05, "loss": 0.0451, "step": 28590 }, { "grad_norm": 0.41432684659957886, "learning_rate": 8.553144367465609e-05, "loss": 0.0414, "step": 28600 }, { "grad_norm": 0.5581527352333069, "learning_rate": 8.551980846309544e-05, "loss": 0.0612, "step": 28610 }, { "grad_norm": 0.5656733512878418, "learning_rate": 8.550816936714193e-05, "loss": 0.0447, "step": 28620 }, { "grad_norm": 0.6470417976379395, "learning_rate": 8.549652638806841e-05, "loss": 0.0431, "step": 28630 }, { "grad_norm": 0.24643030762672424, "learning_rate": 8.548487952714812e-05, "loss": 0.0416, "step": 28640 }, { "grad_norm": 0.3148961365222931, "learning_rate": 8.547322878565478e-05, "loss": 0.0449, "step": 28650 }, { "grad_norm": 0.4061635434627533, "learning_rate": 8.546157416486245e-05, "loss": 0.0511, "step": 28660 }, { "grad_norm": 0.953089714050293, "learning_rate": 8.54499156660457e-05, "loss": 0.0346, "step": 28670 }, { "grad_norm": 0.31152504682540894, "learning_rate": 8.543825329047947e-05, "loss": 0.0526, "step": 28680 }, { "grad_norm": 0.5677733421325684, "learning_rate": 8.542658703943913e-05, "loss": 0.048, "step": 28690 }, { "grad_norm": 0.2885165214538574, "learning_rate": 8.541491691420051e-05, "loss": 0.0404, "step": 28700 }, { "grad_norm": 0.4194226861000061, "learning_rate": 8.54032429160398e-05, "loss": 0.0392, "step": 28710 }, { "grad_norm": 0.24521377682685852, "learning_rate": 8.539156504623369e-05, "loss": 0.0355, "step": 28720 }, { "grad_norm": 0.5512294173240662, "learning_rate": 8.537988330605923e-05, "loss": 0.0706, "step": 28730 }, { "grad_norm": 0.323910117149353, "learning_rate": 8.536819769679393e-05, "loss": 0.0405, "step": 28740 }, { "grad_norm": 0.5325085520744324, "learning_rate": 8.53565082197157e-05, "loss": 0.0525, "step": 28750 }, { "grad_norm": 0.6373860239982605, "learning_rate": 8.534481487610289e-05, "loss": 0.0468, "step": 28760 }, { "grad_norm": 0.30533596873283386, "learning_rate": 8.533311766723428e-05, "loss": 0.0372, "step": 28770 }, { "grad_norm": 0.4803581237792969, "learning_rate": 8.532141659438901e-05, "loss": 0.0474, "step": 28780 }, { "grad_norm": 0.2854142487049103, "learning_rate": 8.530971165884675e-05, "loss": 0.0422, "step": 28790 }, { "grad_norm": 0.41590234637260437, "learning_rate": 8.529800286188752e-05, "loss": 0.0407, "step": 28800 }, { "grad_norm": 0.39886313676834106, "learning_rate": 8.528629020479175e-05, "loss": 0.035, "step": 28810 }, { "grad_norm": 0.35015562176704407, "learning_rate": 8.527457368884033e-05, "loss": 0.0391, "step": 28820 }, { "grad_norm": 0.34103167057037354, "learning_rate": 8.526285331531458e-05, "loss": 0.0351, "step": 28830 }, { "grad_norm": 0.4851154088973999, "learning_rate": 8.525112908549621e-05, "loss": 0.0402, "step": 28840 }, { "grad_norm": 0.6451430320739746, "learning_rate": 8.523940100066735e-05, "loss": 0.0445, "step": 28850 }, { "grad_norm": 0.3390893042087555, "learning_rate": 8.52276690621106e-05, "loss": 0.0393, "step": 28860 }, { "grad_norm": 0.535434365272522, "learning_rate": 8.521593327110889e-05, "loss": 0.0521, "step": 28870 }, { "grad_norm": 0.4570687711238861, "learning_rate": 8.520419362894569e-05, "loss": 0.0548, "step": 28880 }, { "grad_norm": 0.4258609414100647, "learning_rate": 8.51924501369048e-05, "loss": 0.0496, "step": 28890 }, { "grad_norm": 0.7549803853034973, "learning_rate": 8.518070279627047e-05, "loss": 0.0441, "step": 28900 }, { "grad_norm": 0.3121202290058136, "learning_rate": 8.516895160832737e-05, "loss": 0.0438, "step": 28910 }, { "grad_norm": 0.4993830621242523, "learning_rate": 8.515719657436061e-05, "loss": 0.0434, "step": 28920 }, { "grad_norm": 0.5437288880348206, "learning_rate": 8.514543769565568e-05, "loss": 0.0478, "step": 28930 }, { "grad_norm": 0.512362539768219, "learning_rate": 8.513367497349853e-05, "loss": 0.0467, "step": 28940 }, { "grad_norm": 0.41498538851737976, "learning_rate": 8.51219084091755e-05, "loss": 0.0482, "step": 28950 }, { "grad_norm": 0.31411227583885193, "learning_rate": 8.511013800397338e-05, "loss": 0.0461, "step": 28960 }, { "grad_norm": 0.5016070604324341, "learning_rate": 8.509836375917937e-05, "loss": 0.0461, "step": 28970 }, { "grad_norm": 0.6502344608306885, "learning_rate": 8.508658567608104e-05, "loss": 0.048, "step": 28980 }, { "grad_norm": 0.5311003923416138, "learning_rate": 8.507480375596647e-05, "loss": 0.0468, "step": 28990 }, { "grad_norm": 0.5337628126144409, "learning_rate": 8.506301800012408e-05, "loss": 0.0475, "step": 29000 }, { "grad_norm": 0.4806654453277588, "learning_rate": 8.505122840984278e-05, "loss": 0.0545, "step": 29010 }, { "grad_norm": 0.4612351655960083, "learning_rate": 8.503943498641182e-05, "loss": 0.0532, "step": 29020 }, { "grad_norm": 0.47187694907188416, "learning_rate": 8.502763773112095e-05, "loss": 0.0386, "step": 29030 }, { "grad_norm": 0.5589115619659424, "learning_rate": 8.501583664526026e-05, "loss": 0.042, "step": 29040 }, { "grad_norm": 0.5162854194641113, "learning_rate": 8.500403173012032e-05, "loss": 0.048, "step": 29050 }, { "grad_norm": 0.5312401056289673, "learning_rate": 8.499222298699211e-05, "loss": 0.0462, "step": 29060 }, { "grad_norm": 0.4237351417541504, "learning_rate": 8.498041041716701e-05, "loss": 0.048, "step": 29070 }, { "grad_norm": 0.32368916273117065, "learning_rate": 8.496859402193681e-05, "loss": 0.0447, "step": 29080 }, { "grad_norm": 0.5608553290367126, "learning_rate": 8.495677380259374e-05, "loss": 0.0447, "step": 29090 }, { "grad_norm": 0.33293306827545166, "learning_rate": 8.494494976043045e-05, "loss": 0.0437, "step": 29100 }, { "grad_norm": 0.36543112993240356, "learning_rate": 8.493312189673998e-05, "loss": 0.0415, "step": 29110 }, { "grad_norm": 0.4238094687461853, "learning_rate": 8.492129021281584e-05, "loss": 0.0441, "step": 29120 }, { "grad_norm": 0.37774839997291565, "learning_rate": 8.490945470995188e-05, "loss": 0.0394, "step": 29130 }, { "grad_norm": 1.2565696239471436, "learning_rate": 8.489761538944247e-05, "loss": 0.0638, "step": 29140 }, { "grad_norm": 0.4438434839248657, "learning_rate": 8.48857722525823e-05, "loss": 0.0499, "step": 29150 }, { "grad_norm": 0.35741978883743286, "learning_rate": 8.487392530066652e-05, "loss": 0.0511, "step": 29160 }, { "grad_norm": 0.44870296120643616, "learning_rate": 8.486207453499069e-05, "loss": 0.0501, "step": 29170 }, { "grad_norm": 0.4450171887874603, "learning_rate": 8.485021995685082e-05, "loss": 0.0543, "step": 29180 }, { "grad_norm": 0.6550132632255554, "learning_rate": 8.483836156754328e-05, "loss": 0.0586, "step": 29190 }, { "grad_norm": 0.8556073307991028, "learning_rate": 8.482649936836491e-05, "loss": 0.0513, "step": 29200 }, { "grad_norm": 0.550433874130249, "learning_rate": 8.481463336061293e-05, "loss": 0.0504, "step": 29210 }, { "grad_norm": 0.6700381636619568, "learning_rate": 8.480276354558496e-05, "loss": 0.0521, "step": 29220 }, { "grad_norm": 0.35091298818588257, "learning_rate": 8.479088992457913e-05, "loss": 0.0413, "step": 29230 }, { "grad_norm": 0.4967332184314728, "learning_rate": 8.477901249889387e-05, "loss": 0.0569, "step": 29240 }, { "grad_norm": 0.3833649456501007, "learning_rate": 8.47671312698281e-05, "loss": 0.0499, "step": 29250 }, { "grad_norm": 0.4217756688594818, "learning_rate": 8.475524623868112e-05, "loss": 0.0394, "step": 29260 }, { "grad_norm": 0.3045164942741394, "learning_rate": 8.474335740675266e-05, "loss": 0.0442, "step": 29270 }, { "grad_norm": 1.0788886547088623, "learning_rate": 8.473146477534289e-05, "loss": 0.0431, "step": 29280 }, { "grad_norm": 0.5171101689338684, "learning_rate": 8.471956834575232e-05, "loss": 0.0457, "step": 29290 }, { "grad_norm": 0.4010612666606903, "learning_rate": 8.470766811928197e-05, "loss": 0.0377, "step": 29300 }, { "grad_norm": 0.4391600489616394, "learning_rate": 8.469576409723323e-05, "loss": 0.048, "step": 29310 }, { "grad_norm": 0.44120678305625916, "learning_rate": 8.468385628090788e-05, "loss": 0.0457, "step": 29320 }, { "grad_norm": 0.40657493472099304, "learning_rate": 8.467194467160815e-05, "loss": 0.0466, "step": 29330 }, { "grad_norm": 0.5337510704994202, "learning_rate": 8.466002927063667e-05, "loss": 0.0494, "step": 29340 }, { "grad_norm": 0.5886551737785339, "learning_rate": 8.464811007929651e-05, "loss": 0.0463, "step": 29350 }, { "grad_norm": 0.48299968242645264, "learning_rate": 8.463618709889114e-05, "loss": 0.0548, "step": 29360 }, { "grad_norm": 0.6088058948516846, "learning_rate": 8.462426033072442e-05, "loss": 0.0466, "step": 29370 }, { "grad_norm": 0.568727970123291, "learning_rate": 8.461232977610061e-05, "loss": 0.0372, "step": 29380 }, { "grad_norm": 0.4830573499202728, "learning_rate": 8.46003954363245e-05, "loss": 0.0395, "step": 29390 }, { "grad_norm": 0.45418575406074524, "learning_rate": 8.458845731270115e-05, "loss": 0.0491, "step": 29400 }, { "grad_norm": 0.8093655109405518, "learning_rate": 8.45765154065361e-05, "loss": 0.0512, "step": 29410 }, { "grad_norm": 0.41156595945358276, "learning_rate": 8.456456971913532e-05, "loss": 0.0362, "step": 29420 }, { "grad_norm": 1.1981477737426758, "learning_rate": 8.455262025180517e-05, "loss": 0.0452, "step": 29430 }, { "grad_norm": 0.3246181011199951, "learning_rate": 8.454066700585242e-05, "loss": 0.0457, "step": 29440 }, { "grad_norm": 0.5054764747619629, "learning_rate": 8.452870998258423e-05, "loss": 0.0446, "step": 29450 }, { "grad_norm": 0.3898976743221283, "learning_rate": 8.451674918330825e-05, "loss": 0.0467, "step": 29460 }, { "grad_norm": 0.4498261511325836, "learning_rate": 8.450478460933246e-05, "loss": 0.0527, "step": 29470 }, { "grad_norm": 0.47331273555755615, "learning_rate": 8.449281626196532e-05, "loss": 0.0477, "step": 29480 }, { "grad_norm": 0.6143174767494202, "learning_rate": 8.448084414251564e-05, "loss": 0.0507, "step": 29490 }, { "grad_norm": 0.4872731566429138, "learning_rate": 8.446886825229271e-05, "loss": 0.0547, "step": 29500 }, { "grad_norm": 0.43660780787467957, "learning_rate": 8.445688859260615e-05, "loss": 0.05, "step": 29510 }, { "grad_norm": 0.5831051468849182, "learning_rate": 8.444490516476606e-05, "loss": 0.0425, "step": 29520 }, { "grad_norm": 0.9705554246902466, "learning_rate": 8.443291797008293e-05, "loss": 0.0532, "step": 29530 }, { "grad_norm": 0.35977765917778015, "learning_rate": 8.442092700986765e-05, "loss": 0.0563, "step": 29540 }, { "grad_norm": 0.4187259376049042, "learning_rate": 8.440893228543156e-05, "loss": 0.0508, "step": 29550 }, { "grad_norm": 0.3657747209072113, "learning_rate": 8.439693379808638e-05, "loss": 0.0344, "step": 29560 }, { "grad_norm": 0.3728261888027191, "learning_rate": 8.43849315491442e-05, "loss": 0.0432, "step": 29570 }, { "grad_norm": 0.6894286274909973, "learning_rate": 8.437292553991763e-05, "loss": 0.0394, "step": 29580 }, { "grad_norm": 0.42397698760032654, "learning_rate": 8.436091577171959e-05, "loss": 0.0388, "step": 29590 }, { "grad_norm": 0.6145884990692139, "learning_rate": 8.434890224586347e-05, "loss": 0.0376, "step": 29600 }, { "grad_norm": 0.2966180741786957, "learning_rate": 8.433688496366303e-05, "loss": 0.033, "step": 29610 }, { "grad_norm": 0.6161452531814575, "learning_rate": 8.432486392643248e-05, "loss": 0.0474, "step": 29620 }, { "grad_norm": 0.5106310248374939, "learning_rate": 8.431283913548643e-05, "loss": 0.0475, "step": 29630 }, { "grad_norm": 0.4425097107887268, "learning_rate": 8.430081059213985e-05, "loss": 0.0468, "step": 29640 }, { "grad_norm": 0.8370267152786255, "learning_rate": 8.428877829770823e-05, "loss": 0.0569, "step": 29650 }, { "grad_norm": 0.4991222321987152, "learning_rate": 8.427674225350735e-05, "loss": 0.0504, "step": 29660 }, { "grad_norm": 0.5037361979484558, "learning_rate": 8.426470246085347e-05, "loss": 0.0489, "step": 29670 }, { "grad_norm": 0.2968091368675232, "learning_rate": 8.425265892106324e-05, "loss": 0.0395, "step": 29680 }, { "grad_norm": 0.4306069314479828, "learning_rate": 8.424061163545374e-05, "loss": 0.0449, "step": 29690 }, { "grad_norm": 0.4555571675300598, "learning_rate": 8.422856060534243e-05, "loss": 0.0377, "step": 29700 }, { "grad_norm": 0.5291247367858887, "learning_rate": 8.421650583204718e-05, "loss": 0.059, "step": 29710 }, { "grad_norm": 0.5782086253166199, "learning_rate": 8.420444731688633e-05, "loss": 0.0466, "step": 29720 }, { "grad_norm": 0.38665348291397095, "learning_rate": 8.419238506117852e-05, "loss": 0.0519, "step": 29730 }, { "grad_norm": 0.535662055015564, "learning_rate": 8.418031906624289e-05, "loss": 0.0431, "step": 29740 }, { "grad_norm": 0.5558139085769653, "learning_rate": 8.416824933339898e-05, "loss": 0.0566, "step": 29750 }, { "grad_norm": 0.626213014125824, "learning_rate": 8.415617586396667e-05, "loss": 0.0477, "step": 29760 }, { "grad_norm": 0.36177775263786316, "learning_rate": 8.414409865926632e-05, "loss": 0.0534, "step": 29770 }, { "grad_norm": 0.3876655101776123, "learning_rate": 8.413201772061867e-05, "loss": 0.0438, "step": 29780 }, { "grad_norm": 0.4869330823421478, "learning_rate": 8.411993304934488e-05, "loss": 0.0396, "step": 29790 }, { "grad_norm": 0.361195832490921, "learning_rate": 8.410784464676654e-05, "loss": 0.0418, "step": 29800 }, { "grad_norm": 0.3818972408771515, "learning_rate": 8.409575251420556e-05, "loss": 0.0363, "step": 29810 }, { "grad_norm": 0.395203560590744, "learning_rate": 8.408365665298435e-05, "loss": 0.0419, "step": 29820 }, { "grad_norm": 0.5051153302192688, "learning_rate": 8.40715570644257e-05, "loss": 0.0613, "step": 29830 }, { "grad_norm": 0.4187925159931183, "learning_rate": 8.40594537498528e-05, "loss": 0.055, "step": 29840 }, { "grad_norm": 0.5416282415390015, "learning_rate": 8.404734671058924e-05, "loss": 0.0475, "step": 29850 }, { "grad_norm": 0.38795140385627747, "learning_rate": 8.403523594795902e-05, "loss": 0.0393, "step": 29860 }, { "grad_norm": 0.5975579023361206, "learning_rate": 8.402312146328659e-05, "loss": 0.0493, "step": 29870 }, { "grad_norm": 0.3835308253765106, "learning_rate": 8.401100325789675e-05, "loss": 0.0402, "step": 29880 }, { "grad_norm": 0.32229846715927124, "learning_rate": 8.399888133311472e-05, "loss": 0.0424, "step": 29890 }, { "grad_norm": 0.44156908988952637, "learning_rate": 8.398675569026613e-05, "loss": 0.0518, "step": 29900 }, { "grad_norm": 0.49531376361846924, "learning_rate": 8.397462633067705e-05, "loss": 0.0402, "step": 29910 }, { "grad_norm": 0.3203009068965912, "learning_rate": 8.396249325567392e-05, "loss": 0.0409, "step": 29920 }, { "grad_norm": 0.4312969446182251, "learning_rate": 8.395035646658357e-05, "loss": 0.0436, "step": 29930 }, { "grad_norm": 0.37073764204978943, "learning_rate": 8.39382159647333e-05, "loss": 0.0421, "step": 29940 }, { "grad_norm": 0.4245702922344208, "learning_rate": 8.392607175145075e-05, "loss": 0.0382, "step": 29950 }, { "grad_norm": 0.35712459683418274, "learning_rate": 8.3913923828064e-05, "loss": 0.0484, "step": 29960 }, { "grad_norm": 0.5462590456008911, "learning_rate": 8.390177219590152e-05, "loss": 0.0409, "step": 29970 }, { "grad_norm": 0.493204802274704, "learning_rate": 8.388961685629222e-05, "loss": 0.0478, "step": 29980 }, { "grad_norm": 0.4250517785549164, "learning_rate": 8.387745781056536e-05, "loss": 0.0375, "step": 29990 }, { "grad_norm": 0.47008347511291504, "learning_rate": 8.386529506005065e-05, "loss": 0.0451, "step": 30000 }, { "grad_norm": 0.5943960547447205, "learning_rate": 8.38531286060782e-05, "loss": 0.0413, "step": 30010 }, { "grad_norm": 0.4030984938144684, "learning_rate": 8.384095844997849e-05, "loss": 0.0472, "step": 30020 }, { "grad_norm": 0.42597222328186035, "learning_rate": 8.382878459308245e-05, "loss": 0.0397, "step": 30030 }, { "grad_norm": 0.35812458395957947, "learning_rate": 8.381660703672138e-05, "loss": 0.0419, "step": 30040 }, { "grad_norm": 0.5155025124549866, "learning_rate": 8.380442578222702e-05, "loss": 0.0585, "step": 30050 }, { "grad_norm": 0.5187339186668396, "learning_rate": 8.379224083093146e-05, "loss": 0.0518, "step": 30060 }, { "grad_norm": 0.404449999332428, "learning_rate": 8.378005218416727e-05, "loss": 0.0377, "step": 30070 }, { "grad_norm": 1.2006282806396484, "learning_rate": 8.376785984326735e-05, "loss": 0.0444, "step": 30080 }, { "grad_norm": 0.5490921139717102, "learning_rate": 8.375566380956506e-05, "loss": 0.0436, "step": 30090 }, { "grad_norm": 0.5369464159011841, "learning_rate": 8.374346408439411e-05, "loss": 0.042, "step": 30100 }, { "grad_norm": 0.856392502784729, "learning_rate": 8.373126066908868e-05, "loss": 0.0529, "step": 30110 }, { "grad_norm": 0.35711339116096497, "learning_rate": 8.371905356498326e-05, "loss": 0.0396, "step": 30120 }, { "grad_norm": 0.4295373857021332, "learning_rate": 8.370684277341288e-05, "loss": 0.0434, "step": 30130 }, { "grad_norm": 0.36283594369888306, "learning_rate": 8.369462829571282e-05, "loss": 0.0453, "step": 30140 }, { "grad_norm": 0.3526519536972046, "learning_rate": 8.36824101332189e-05, "loss": 0.0396, "step": 30150 }, { "grad_norm": 0.47937890887260437, "learning_rate": 8.367018828726721e-05, "loss": 0.0372, "step": 30160 }, { "grad_norm": 0.4882480502128601, "learning_rate": 8.365796275919438e-05, "loss": 0.0529, "step": 30170 }, { "grad_norm": 0.254130095243454, "learning_rate": 8.364573355033734e-05, "loss": 0.0326, "step": 30180 }, { "grad_norm": 0.37624451518058777, "learning_rate": 8.363350066203346e-05, "loss": 0.0427, "step": 30190 }, { "grad_norm": 0.36128050088882446, "learning_rate": 8.362126409562053e-05, "loss": 0.0364, "step": 30200 }, { "grad_norm": 0.404781699180603, "learning_rate": 8.360902385243667e-05, "loss": 0.0392, "step": 30210 }, { "grad_norm": 0.4092870056629181, "learning_rate": 8.359677993382052e-05, "loss": 0.0408, "step": 30220 }, { "grad_norm": 0.4619481563568115, "learning_rate": 8.358453234111103e-05, "loss": 0.0429, "step": 30230 }, { "grad_norm": 0.3329189717769623, "learning_rate": 8.357228107564756e-05, "loss": 0.0467, "step": 30240 }, { "grad_norm": 0.46887487173080444, "learning_rate": 8.356002613876993e-05, "loss": 0.047, "step": 30250 }, { "grad_norm": 0.5666639804840088, "learning_rate": 8.35477675318183e-05, "loss": 0.0591, "step": 30260 }, { "grad_norm": 0.43999233841896057, "learning_rate": 8.353550525613323e-05, "loss": 0.0443, "step": 30270 }, { "grad_norm": 0.47558531165122986, "learning_rate": 8.352323931305572e-05, "loss": 0.0572, "step": 30280 }, { "grad_norm": 0.5498607754707336, "learning_rate": 8.351096970392717e-05, "loss": 0.0469, "step": 30290 }, { "grad_norm": 0.4621579349040985, "learning_rate": 8.349869643008937e-05, "loss": 0.0332, "step": 30300 }, { "grad_norm": 0.36722102761268616, "learning_rate": 8.348641949288449e-05, "loss": 0.0559, "step": 30310 }, { "grad_norm": 0.7738141417503357, "learning_rate": 8.34741388936551e-05, "loss": 0.0511, "step": 30320 }, { "grad_norm": 0.43734827637672424, "learning_rate": 8.346185463374423e-05, "loss": 0.0635, "step": 30330 }, { "grad_norm": 0.3111410140991211, "learning_rate": 8.344956671449524e-05, "loss": 0.0534, "step": 30340 }, { "grad_norm": 0.3538481295108795, "learning_rate": 8.343727513725192e-05, "loss": 0.0455, "step": 30350 }, { "grad_norm": 0.3785705268383026, "learning_rate": 8.342497990335847e-05, "loss": 0.0401, "step": 30360 }, { "grad_norm": 0.41724660992622375, "learning_rate": 8.341268101415946e-05, "loss": 0.0522, "step": 30370 }, { "grad_norm": 0.5363361835479736, "learning_rate": 8.34003784709999e-05, "loss": 0.0493, "step": 30380 }, { "grad_norm": 0.614670991897583, "learning_rate": 8.338807227522516e-05, "loss": 0.0547, "step": 30390 }, { "grad_norm": 0.5856308937072754, "learning_rate": 8.337576242818103e-05, "loss": 0.0515, "step": 30400 }, { "grad_norm": 0.3127804398536682, "learning_rate": 8.336344893121372e-05, "loss": 0.0344, "step": 30410 }, { "grad_norm": 0.939545750617981, "learning_rate": 8.335113178566977e-05, "loss": 0.0449, "step": 30420 }, { "grad_norm": 0.5447235703468323, "learning_rate": 8.33388109928962e-05, "loss": 0.044, "step": 30430 }, { "grad_norm": 0.4394667148590088, "learning_rate": 8.33264865542404e-05, "loss": 0.0462, "step": 30440 }, { "grad_norm": 0.3512713313102722, "learning_rate": 8.331415847105013e-05, "loss": 0.0457, "step": 30450 }, { "grad_norm": 0.34883448481559753, "learning_rate": 8.330182674467357e-05, "loss": 0.0427, "step": 30460 }, { "grad_norm": 0.6190721392631531, "learning_rate": 8.32894913764593e-05, "loss": 0.0384, "step": 30470 }, { "grad_norm": 0.3457951247692108, "learning_rate": 8.327715236775633e-05, "loss": 0.0448, "step": 30480 }, { "grad_norm": 0.48090434074401855, "learning_rate": 8.326480971991398e-05, "loss": 0.049, "step": 30490 }, { "grad_norm": 0.4223628640174866, "learning_rate": 8.325246343428206e-05, "loss": 0.046, "step": 30500 }, { "grad_norm": 0.37449923157691956, "learning_rate": 8.324011351221072e-05, "loss": 0.0395, "step": 30510 }, { "grad_norm": 0.4148957431316376, "learning_rate": 8.322775995505057e-05, "loss": 0.0406, "step": 30520 }, { "grad_norm": 0.4901459217071533, "learning_rate": 8.321540276415254e-05, "loss": 0.0451, "step": 30530 }, { "grad_norm": 0.5269688367843628, "learning_rate": 8.320304194086798e-05, "loss": 0.0439, "step": 30540 }, { "grad_norm": 0.3484671711921692, "learning_rate": 8.31906774865487e-05, "loss": 0.0366, "step": 30550 }, { "grad_norm": 0.49559593200683594, "learning_rate": 8.317830940254682e-05, "loss": 0.0451, "step": 30560 }, { "grad_norm": 0.44926297664642334, "learning_rate": 8.316593769021491e-05, "loss": 0.0455, "step": 30570 }, { "grad_norm": 0.3985862731933594, "learning_rate": 8.315356235090592e-05, "loss": 0.0522, "step": 30580 }, { "grad_norm": 0.3430386781692505, "learning_rate": 8.314118338597319e-05, "loss": 0.0398, "step": 30590 }, { "grad_norm": 0.5392090678215027, "learning_rate": 8.312880079677048e-05, "loss": 0.0377, "step": 30600 }, { "grad_norm": 0.39940178394317627, "learning_rate": 8.311641458465191e-05, "loss": 0.0453, "step": 30610 }, { "grad_norm": 0.4443845748901367, "learning_rate": 8.310402475097205e-05, "loss": 0.0467, "step": 30620 }, { "grad_norm": 0.5094236135482788, "learning_rate": 8.309163129708581e-05, "loss": 0.05, "step": 30630 }, { "grad_norm": 0.537760853767395, "learning_rate": 8.307923422434852e-05, "loss": 0.0462, "step": 30640 }, { "grad_norm": 0.6606910228729248, "learning_rate": 8.30668335341159e-05, "loss": 0.0462, "step": 30650 }, { "grad_norm": 0.5122088193893433, "learning_rate": 8.305442922774408e-05, "loss": 0.0451, "step": 30660 }, { "grad_norm": 0.3100892901420593, "learning_rate": 8.304202130658959e-05, "loss": 0.0397, "step": 30670 }, { "grad_norm": 0.41698774695396423, "learning_rate": 8.302960977200931e-05, "loss": 0.0426, "step": 30680 }, { "grad_norm": 0.5597530007362366, "learning_rate": 8.301719462536058e-05, "loss": 0.0318, "step": 30690 }, { "grad_norm": 0.46857914328575134, "learning_rate": 8.300477586800108e-05, "loss": 0.0423, "step": 30700 }, { "grad_norm": 2.0979440212249756, "learning_rate": 8.299235350128892e-05, "loss": 0.0468, "step": 30710 }, { "grad_norm": 0.7014055252075195, "learning_rate": 8.297992752658258e-05, "loss": 0.0473, "step": 30720 }, { "grad_norm": 0.41637057065963745, "learning_rate": 8.296749794524094e-05, "loss": 0.043, "step": 30730 }, { "grad_norm": 0.4932733178138733, "learning_rate": 8.295506475862332e-05, "loss": 0.0528, "step": 30740 }, { "grad_norm": 0.5621941685676575, "learning_rate": 8.294262796808933e-05, "loss": 0.0432, "step": 30750 }, { "grad_norm": 0.38349539041519165, "learning_rate": 8.293018757499909e-05, "loss": 0.0485, "step": 30760 }, { "grad_norm": 0.48172271251678467, "learning_rate": 8.291774358071305e-05, "loss": 0.0488, "step": 30770 }, { "grad_norm": 0.41163507103919983, "learning_rate": 8.290529598659205e-05, "loss": 0.042, "step": 30780 }, { "grad_norm": 0.4923691749572754, "learning_rate": 8.289284479399738e-05, "loss": 0.0484, "step": 30790 }, { "grad_norm": 0.3583037257194519, "learning_rate": 8.288039000429064e-05, "loss": 0.0407, "step": 30800 }, { "grad_norm": 0.46515247225761414, "learning_rate": 8.286793161883388e-05, "loss": 0.0451, "step": 30810 }, { "grad_norm": 0.5825690031051636, "learning_rate": 8.285546963898954e-05, "loss": 0.0433, "step": 30820 }, { "grad_norm": 0.5838467478752136, "learning_rate": 8.284300406612044e-05, "loss": 0.0478, "step": 30830 }, { "grad_norm": 0.4621959328651428, "learning_rate": 8.283053490158978e-05, "loss": 0.0418, "step": 30840 }, { "grad_norm": 0.2766452133655548, "learning_rate": 8.28180621467612e-05, "loss": 0.0394, "step": 30850 }, { "grad_norm": 0.4272378385066986, "learning_rate": 8.280558580299867e-05, "loss": 0.0507, "step": 30860 }, { "grad_norm": 0.3552391529083252, "learning_rate": 8.279310587166661e-05, "loss": 0.0397, "step": 30870 }, { "grad_norm": 0.33623436093330383, "learning_rate": 8.278062235412978e-05, "loss": 0.0372, "step": 30880 }, { "grad_norm": 0.4875522553920746, "learning_rate": 8.276813525175339e-05, "loss": 0.0467, "step": 30890 }, { "grad_norm": 0.3612241744995117, "learning_rate": 8.2755644565903e-05, "loss": 0.0417, "step": 30900 }, { "grad_norm": 0.6763073801994324, "learning_rate": 8.274315029794454e-05, "loss": 0.0472, "step": 30910 }, { "grad_norm": 0.2831462323665619, "learning_rate": 8.273065244924443e-05, "loss": 0.0466, "step": 30920 }, { "grad_norm": 0.4298490583896637, "learning_rate": 8.271815102116936e-05, "loss": 0.0339, "step": 30930 }, { "grad_norm": 0.4064736068248749, "learning_rate": 8.270564601508648e-05, "loss": 0.0537, "step": 30940 }, { "grad_norm": 0.2946764826774597, "learning_rate": 8.269313743236333e-05, "loss": 0.0435, "step": 30950 }, { "grad_norm": 0.6589621901512146, "learning_rate": 8.268062527436783e-05, "loss": 0.0404, "step": 30960 }, { "grad_norm": 0.705993115901947, "learning_rate": 8.266810954246828e-05, "loss": 0.0396, "step": 30970 }, { "grad_norm": 0.4779936969280243, "learning_rate": 8.265559023803338e-05, "loss": 0.0353, "step": 30980 }, { "grad_norm": 0.33704671263694763, "learning_rate": 8.264306736243223e-05, "loss": 0.0349, "step": 30990 }, { "grad_norm": 0.3492327630519867, "learning_rate": 8.263054091703432e-05, "loss": 0.0382, "step": 31000 }, { "grad_norm": 0.3157186210155487, "learning_rate": 8.26180109032095e-05, "loss": 0.0413, "step": 31010 }, { "grad_norm": 0.4078807830810547, "learning_rate": 8.260547732232807e-05, "loss": 0.0287, "step": 31020 }, { "grad_norm": 0.36994022130966187, "learning_rate": 8.259294017576063e-05, "loss": 0.0359, "step": 31030 }, { "grad_norm": 0.7777881026268005, "learning_rate": 8.258039946487828e-05, "loss": 0.0357, "step": 31040 }, { "grad_norm": 0.5498706102371216, "learning_rate": 8.256785519105241e-05, "loss": 0.034, "step": 31050 }, { "grad_norm": 0.40915167331695557, "learning_rate": 8.255530735565488e-05, "loss": 0.0377, "step": 31060 }, { "grad_norm": 0.28503140807151794, "learning_rate": 8.254275596005787e-05, "loss": 0.0398, "step": 31070 }, { "grad_norm": 0.5976194739341736, "learning_rate": 8.2530201005634e-05, "loss": 0.0426, "step": 31080 }, { "grad_norm": 0.43541035056114197, "learning_rate": 8.251764249375626e-05, "loss": 0.0484, "step": 31090 }, { "grad_norm": 0.4941195547580719, "learning_rate": 8.250508042579803e-05, "loss": 0.0446, "step": 31100 }, { "grad_norm": 0.5529252290725708, "learning_rate": 8.249251480313307e-05, "loss": 0.0418, "step": 31110 }, { "grad_norm": 0.34765923023223877, "learning_rate": 8.247994562713555e-05, "loss": 0.0411, "step": 31120 }, { "grad_norm": 0.4544772803783417, "learning_rate": 8.246737289917998e-05, "loss": 0.0361, "step": 31130 }, { "grad_norm": 0.40066996216773987, "learning_rate": 8.245479662064135e-05, "loss": 0.0355, "step": 31140 }, { "grad_norm": 0.6965216994285583, "learning_rate": 8.244221679289496e-05, "loss": 0.0416, "step": 31150 }, { "grad_norm": 0.3547831177711487, "learning_rate": 8.242963341731652e-05, "loss": 0.0352, "step": 31160 }, { "grad_norm": 0.47185012698173523, "learning_rate": 8.24170464952821e-05, "loss": 0.0419, "step": 31170 }, { "grad_norm": 0.35645991563796997, "learning_rate": 8.240445602816824e-05, "loss": 0.0475, "step": 31180 }, { "grad_norm": 0.3788387179374695, "learning_rate": 8.239186201735179e-05, "loss": 0.0305, "step": 31190 }, { "grad_norm": 0.40885472297668457, "learning_rate": 8.237926446420998e-05, "loss": 0.0306, "step": 31200 }, { "grad_norm": 0.4893163740634918, "learning_rate": 8.236666337012052e-05, "loss": 0.0379, "step": 31210 }, { "grad_norm": 0.6075450778007507, "learning_rate": 8.23540587364614e-05, "loss": 0.0423, "step": 31220 }, { "grad_norm": 0.4208647906780243, "learning_rate": 8.234145056461107e-05, "loss": 0.0393, "step": 31230 }, { "grad_norm": 0.4262140393257141, "learning_rate": 8.232883885594832e-05, "loss": 0.0467, "step": 31240 }, { "grad_norm": 0.4615202248096466, "learning_rate": 8.231622361185236e-05, "loss": 0.0407, "step": 31250 }, { "grad_norm": 0.6432441473007202, "learning_rate": 8.230360483370278e-05, "loss": 0.0702, "step": 31260 }, { "grad_norm": 0.7278408408164978, "learning_rate": 8.229098252287953e-05, "loss": 0.0382, "step": 31270 }, { "grad_norm": 0.34130263328552246, "learning_rate": 8.2278356680763e-05, "loss": 0.0497, "step": 31280 }, { "grad_norm": 0.5184794664382935, "learning_rate": 8.22657273087339e-05, "loss": 0.0407, "step": 31290 }, { "grad_norm": 0.39205488562583923, "learning_rate": 8.225309440817336e-05, "loss": 0.0461, "step": 31300 }, { "grad_norm": 0.7177898287773132, "learning_rate": 8.224045798046293e-05, "loss": 0.0678, "step": 31310 }, { "grad_norm": 0.4617948830127716, "learning_rate": 8.22278180269845e-05, "loss": 0.0389, "step": 31320 }, { "grad_norm": 0.35189834237098694, "learning_rate": 8.221517454912031e-05, "loss": 0.0302, "step": 31330 }, { "grad_norm": 0.7965182065963745, "learning_rate": 8.220252754825308e-05, "loss": 0.044, "step": 31340 }, { "grad_norm": 0.6264419555664062, "learning_rate": 8.218987702576586e-05, "loss": 0.0536, "step": 31350 }, { "grad_norm": 0.5022653341293335, "learning_rate": 8.217722298304207e-05, "loss": 0.0603, "step": 31360 }, { "grad_norm": 0.6284124851226807, "learning_rate": 8.216456542146557e-05, "loss": 0.054, "step": 31370 }, { "grad_norm": 0.4295307993888855, "learning_rate": 8.215190434242055e-05, "loss": 0.0421, "step": 31380 }, { "grad_norm": 0.29305121302604675, "learning_rate": 8.213923974729161e-05, "loss": 0.0377, "step": 31390 }, { "grad_norm": 0.8602461218833923, "learning_rate": 8.212657163746373e-05, "loss": 0.0481, "step": 31400 }, { "grad_norm": 0.4078429341316223, "learning_rate": 8.211390001432227e-05, "loss": 0.037, "step": 31410 }, { "grad_norm": 0.326193243265152, "learning_rate": 8.210122487925297e-05, "loss": 0.0397, "step": 31420 }, { "grad_norm": 0.39560237526893616, "learning_rate": 8.208854623364202e-05, "loss": 0.0442, "step": 31430 }, { "grad_norm": 0.5499683022499084, "learning_rate": 8.207586407887589e-05, "loss": 0.0464, "step": 31440 }, { "grad_norm": 0.38417747616767883, "learning_rate": 8.206317841634148e-05, "loss": 0.0399, "step": 31450 }, { "grad_norm": 0.5520546436309814, "learning_rate": 8.205048924742609e-05, "loss": 0.0451, "step": 31460 }, { "grad_norm": 0.40068355202674866, "learning_rate": 8.203779657351738e-05, "loss": 0.0394, "step": 31470 }, { "grad_norm": 0.352506548166275, "learning_rate": 8.20251003960034e-05, "loss": 0.0437, "step": 31480 }, { "grad_norm": 0.5315417051315308, "learning_rate": 8.201240071627258e-05, "loss": 0.0448, "step": 31490 }, { "grad_norm": 0.37410685420036316, "learning_rate": 8.199969753571377e-05, "loss": 0.0423, "step": 31500 }, { "grad_norm": 0.34289276599884033, "learning_rate": 8.198699085571615e-05, "loss": 0.0422, "step": 31510 }, { "grad_norm": 0.4871188700199127, "learning_rate": 8.197428067766928e-05, "loss": 0.0517, "step": 31520 }, { "grad_norm": 0.5503485202789307, "learning_rate": 8.196156700296316e-05, "loss": 0.0577, "step": 31530 }, { "grad_norm": 0.4052006006240845, "learning_rate": 8.194884983298814e-05, "loss": 0.0352, "step": 31540 }, { "grad_norm": 0.42859479784965515, "learning_rate": 8.193612916913491e-05, "loss": 0.0357, "step": 31550 }, { "grad_norm": 0.519128143787384, "learning_rate": 8.192340501279463e-05, "loss": 0.0426, "step": 31560 }, { "grad_norm": 0.43443024158477783, "learning_rate": 8.191067736535876e-05, "loss": 0.0506, "step": 31570 }, { "grad_norm": 0.4256979525089264, "learning_rate": 8.18979462282192e-05, "loss": 0.0459, "step": 31580 }, { "grad_norm": 0.36898109316825867, "learning_rate": 8.188521160276819e-05, "loss": 0.0406, "step": 31590 }, { "grad_norm": 0.3646933138370514, "learning_rate": 8.187247349039837e-05, "loss": 0.0373, "step": 31600 }, { "grad_norm": 0.43608358502388, "learning_rate": 8.185973189250278e-05, "loss": 0.0506, "step": 31610 }, { "grad_norm": 0.3558509647846222, "learning_rate": 8.184698681047482e-05, "loss": 0.0471, "step": 31620 }, { "grad_norm": 0.4882546365261078, "learning_rate": 8.183423824570827e-05, "loss": 0.0466, "step": 31630 }, { "grad_norm": 0.5132888555526733, "learning_rate": 8.182148619959725e-05, "loss": 0.0447, "step": 31640 }, { "grad_norm": 0.47029393911361694, "learning_rate": 8.180873067353636e-05, "loss": 0.0449, "step": 31650 }, { "grad_norm": 0.5744498372077942, "learning_rate": 8.179597166892052e-05, "loss": 0.0434, "step": 31660 }, { "grad_norm": 0.46799787878990173, "learning_rate": 8.178320918714501e-05, "loss": 0.0388, "step": 31670 }, { "grad_norm": 0.35897594690322876, "learning_rate": 8.177044322960554e-05, "loss": 0.0351, "step": 31680 }, { "grad_norm": 0.41447171568870544, "learning_rate": 8.175767379769816e-05, "loss": 0.0336, "step": 31690 }, { "grad_norm": 0.5876629948616028, "learning_rate": 8.174490089281932e-05, "loss": 0.0442, "step": 31700 }, { "grad_norm": 0.5378663539886475, "learning_rate": 8.173212451636584e-05, "loss": 0.0455, "step": 31710 }, { "grad_norm": 0.6620221734046936, "learning_rate": 8.171934466973493e-05, "loss": 0.054, "step": 31720 }, { "grad_norm": 0.4527760148048401, "learning_rate": 8.170656135432418e-05, "loss": 0.0376, "step": 31730 }, { "grad_norm": 0.4178939163684845, "learning_rate": 8.169377457153155e-05, "loss": 0.0459, "step": 31740 }, { "grad_norm": 0.36494606733322144, "learning_rate": 8.168098432275539e-05, "loss": 0.0334, "step": 31750 }, { "grad_norm": 0.4404921531677246, "learning_rate": 8.166819060939442e-05, "loss": 0.0383, "step": 31760 }, { "grad_norm": 0.5363618731498718, "learning_rate": 8.165539343284772e-05, "loss": 0.0605, "step": 31770 }, { "grad_norm": 0.4314001500606537, "learning_rate": 8.16425927945148e-05, "loss": 0.0436, "step": 31780 }, { "grad_norm": 0.5185925364494324, "learning_rate": 8.162978869579551e-05, "loss": 0.0609, "step": 31790 }, { "grad_norm": 0.41760948300361633, "learning_rate": 8.161698113809007e-05, "loss": 0.0458, "step": 31800 }, { "grad_norm": 0.3438664674758911, "learning_rate": 8.160417012279911e-05, "loss": 0.0678, "step": 31810 }, { "grad_norm": 0.3658570349216461, "learning_rate": 8.159135565132363e-05, "loss": 0.0476, "step": 31820 }, { "grad_norm": 0.3173803389072418, "learning_rate": 8.157853772506498e-05, "loss": 0.0413, "step": 31830 }, { "grad_norm": 0.7027190923690796, "learning_rate": 8.156571634542494e-05, "loss": 0.0585, "step": 31840 }, { "grad_norm": 0.5640349388122559, "learning_rate": 8.15528915138056e-05, "loss": 0.0472, "step": 31850 }, { "grad_norm": 0.357839435338974, "learning_rate": 8.154006323160949e-05, "loss": 0.0449, "step": 31860 }, { "grad_norm": 0.3619067668914795, "learning_rate": 8.152723150023949e-05, "loss": 0.0434, "step": 31870 }, { "grad_norm": 0.6625133752822876, "learning_rate": 8.151439632109886e-05, "loss": 0.0384, "step": 31880 }, { "grad_norm": 0.3943592607975006, "learning_rate": 8.150155769559122e-05, "loss": 0.0569, "step": 31890 }, { "grad_norm": 0.37442031502723694, "learning_rate": 8.148871562512058e-05, "loss": 0.0416, "step": 31900 }, { "grad_norm": 0.5273804664611816, "learning_rate": 8.147587011109136e-05, "loss": 0.052, "step": 31910 }, { "grad_norm": 0.4305369555950165, "learning_rate": 8.14630211549083e-05, "loss": 0.0542, "step": 31920 }, { "grad_norm": 0.2528999447822571, "learning_rate": 8.145016875797655e-05, "loss": 0.0315, "step": 31930 }, { "grad_norm": 0.321143239736557, "learning_rate": 8.143731292170164e-05, "loss": 0.0363, "step": 31940 }, { "grad_norm": 0.4524212181568146, "learning_rate": 8.142445364748944e-05, "loss": 0.0544, "step": 31950 }, { "grad_norm": 0.4260450601577759, "learning_rate": 8.141159093674624e-05, "loss": 0.048, "step": 31960 }, { "grad_norm": 0.7683606743812561, "learning_rate": 8.139872479087869e-05, "loss": 0.0399, "step": 31970 }, { "grad_norm": 0.34627145528793335, "learning_rate": 8.13858552112938e-05, "loss": 0.0333, "step": 31980 }, { "grad_norm": 0.43492525815963745, "learning_rate": 8.137298219939895e-05, "loss": 0.0512, "step": 31990 }, { "grad_norm": 0.4878270924091339, "learning_rate": 8.136010575660196e-05, "loss": 0.0384, "step": 32000 }, { "grad_norm": 0.4570207893848419, "learning_rate": 8.134722588431095e-05, "loss": 0.0396, "step": 32010 }, { "grad_norm": 0.29322579503059387, "learning_rate": 8.133434258393444e-05, "loss": 0.0442, "step": 32020 }, { "grad_norm": 0.4850902259349823, "learning_rate": 8.132145585688134e-05, "loss": 0.0437, "step": 32030 }, { "grad_norm": 0.4081157147884369, "learning_rate": 8.130856570456093e-05, "loss": 0.0399, "step": 32040 }, { "grad_norm": 0.36177995800971985, "learning_rate": 8.129567212838283e-05, "loss": 0.0399, "step": 32050 }, { "grad_norm": 0.5890785455703735, "learning_rate": 8.128277512975708e-05, "loss": 0.0496, "step": 32060 }, { "grad_norm": 0.55210942029953, "learning_rate": 8.126987471009408e-05, "loss": 0.0443, "step": 32070 }, { "grad_norm": 0.3612568974494934, "learning_rate": 8.125697087080459e-05, "loss": 0.031, "step": 32080 }, { "grad_norm": 0.4223926365375519, "learning_rate": 8.124406361329976e-05, "loss": 0.0396, "step": 32090 }, { "grad_norm": 0.5180970430374146, "learning_rate": 8.123115293899111e-05, "loss": 0.037, "step": 32100 }, { "grad_norm": 0.3838447630405426, "learning_rate": 8.121823884929055e-05, "loss": 0.0387, "step": 32110 }, { "grad_norm": 0.47433748841285706, "learning_rate": 8.12053213456103e-05, "loss": 0.0496, "step": 32120 }, { "grad_norm": 0.9237806797027588, "learning_rate": 8.119240042936303e-05, "loss": 0.0477, "step": 32130 }, { "grad_norm": 0.9465075135231018, "learning_rate": 8.117947610196175e-05, "loss": 0.048, "step": 32140 }, { "grad_norm": 0.5364922881126404, "learning_rate": 8.116654836481982e-05, "loss": 0.0317, "step": 32150 }, { "grad_norm": 0.41498124599456787, "learning_rate": 8.115361721935106e-05, "loss": 0.0401, "step": 32160 }, { "grad_norm": 0.33622199296951294, "learning_rate": 8.114068266696953e-05, "loss": 0.0438, "step": 32170 }, { "grad_norm": 0.49946948885917664, "learning_rate": 8.112774470908978e-05, "loss": 0.0421, "step": 32180 }, { "grad_norm": 0.48305755853652954, "learning_rate": 8.111480334712665e-05, "loss": 0.0466, "step": 32190 }, { "grad_norm": 0.31151875853538513, "learning_rate": 8.110185858249542e-05, "loss": 0.0363, "step": 32200 }, { "grad_norm": 0.44582077860832214, "learning_rate": 8.108891041661168e-05, "loss": 0.0473, "step": 32210 }, { "grad_norm": 0.4202009439468384, "learning_rate": 8.107595885089146e-05, "loss": 0.0567, "step": 32220 }, { "grad_norm": 1.1725714206695557, "learning_rate": 8.106300388675108e-05, "loss": 0.0412, "step": 32230 }, { "grad_norm": 0.5046017169952393, "learning_rate": 8.105004552560731e-05, "loss": 0.0516, "step": 32240 }, { "grad_norm": 0.6658942103385925, "learning_rate": 8.103708376887724e-05, "loss": 0.0439, "step": 32250 }, { "grad_norm": 0.49766430258750916, "learning_rate": 8.102411861797836e-05, "loss": 0.0479, "step": 32260 }, { "grad_norm": 0.6963645815849304, "learning_rate": 8.101115007432851e-05, "loss": 0.041, "step": 32270 }, { "grad_norm": 0.43533360958099365, "learning_rate": 8.09981781393459e-05, "loss": 0.0322, "step": 32280 }, { "grad_norm": 0.3846853971481323, "learning_rate": 8.098520281444915e-05, "loss": 0.0518, "step": 32290 }, { "grad_norm": 0.4638234078884125, "learning_rate": 8.09722241010572e-05, "loss": 0.0456, "step": 32300 }, { "grad_norm": 0.665976881980896, "learning_rate": 8.095924200058939e-05, "loss": 0.052, "step": 32310 }, { "grad_norm": 0.7014197707176208, "learning_rate": 8.094625651446541e-05, "loss": 0.0452, "step": 32320 }, { "grad_norm": 0.4746362268924713, "learning_rate": 8.093326764410536e-05, "loss": 0.0381, "step": 32330 }, { "grad_norm": 0.5534470081329346, "learning_rate": 8.092027539092966e-05, "loss": 0.0434, "step": 32340 }, { "grad_norm": 0.25343504548072815, "learning_rate": 8.090727975635913e-05, "loss": 0.0303, "step": 32350 }, { "grad_norm": 0.3258708119392395, "learning_rate": 8.089428074181497e-05, "loss": 0.0397, "step": 32360 }, { "grad_norm": 0.35905447602272034, "learning_rate": 8.088127834871871e-05, "loss": 0.0354, "step": 32370 }, { "grad_norm": 0.6122063994407654, "learning_rate": 8.086827257849226e-05, "loss": 0.0376, "step": 32380 }, { "grad_norm": 0.5304968953132629, "learning_rate": 8.085526343255795e-05, "loss": 0.0327, "step": 32390 }, { "grad_norm": 0.8744826912879944, "learning_rate": 8.084225091233842e-05, "loss": 0.0409, "step": 32400 }, { "grad_norm": 0.5288530588150024, "learning_rate": 8.082923501925668e-05, "loss": 0.0361, "step": 32410 }, { "grad_norm": 0.45131754875183105, "learning_rate": 8.081621575473617e-05, "loss": 0.0378, "step": 32420 }, { "grad_norm": 0.36339864134788513, "learning_rate": 8.080319312020064e-05, "loss": 0.0453, "step": 32430 }, { "grad_norm": 0.7559546828269958, "learning_rate": 8.079016711707421e-05, "loss": 0.0441, "step": 32440 }, { "grad_norm": 0.5391356348991394, "learning_rate": 8.077713774678139e-05, "loss": 0.0435, "step": 32450 }, { "grad_norm": 0.3625837564468384, "learning_rate": 8.076410501074707e-05, "loss": 0.0325, "step": 32460 }, { "grad_norm": 0.4466501772403717, "learning_rate": 8.075106891039647e-05, "loss": 0.0474, "step": 32470 }, { "grad_norm": 0.39971113204956055, "learning_rate": 8.073802944715523e-05, "loss": 0.0383, "step": 32480 }, { "grad_norm": 0.43468332290649414, "learning_rate": 8.072498662244929e-05, "loss": 0.0384, "step": 32490 }, { "grad_norm": 0.4158671498298645, "learning_rate": 8.0711940437705e-05, "loss": 0.0439, "step": 32500 }, { "grad_norm": 0.7242623567581177, "learning_rate": 8.06988908943491e-05, "loss": 0.0364, "step": 32510 }, { "grad_norm": 0.25087589025497437, "learning_rate": 8.068583799380863e-05, "loss": 0.0454, "step": 32520 }, { "grad_norm": 0.4115152060985565, "learning_rate": 8.067278173751104e-05, "loss": 0.0405, "step": 32530 }, { "grad_norm": 0.669282853603363, "learning_rate": 8.065972212688417e-05, "loss": 0.0431, "step": 32540 }, { "grad_norm": 0.3808533549308777, "learning_rate": 8.064665916335618e-05, "loss": 0.052, "step": 32550 }, { "grad_norm": 0.4230530858039856, "learning_rate": 8.063359284835564e-05, "loss": 0.0467, "step": 32560 }, { "grad_norm": 0.6758075952529907, "learning_rate": 8.062052318331142e-05, "loss": 0.05, "step": 32570 }, { "grad_norm": 0.6160431504249573, "learning_rate": 8.060745016965283e-05, "loss": 0.0399, "step": 32580 }, { "grad_norm": 0.4782426059246063, "learning_rate": 8.059437380880952e-05, "loss": 0.0427, "step": 32590 }, { "grad_norm": 0.4590315818786621, "learning_rate": 8.058129410221146e-05, "loss": 0.0488, "step": 32600 }, { "grad_norm": 0.5257874131202698, "learning_rate": 8.056821105128908e-05, "loss": 0.0451, "step": 32610 }, { "grad_norm": 0.32705098390579224, "learning_rate": 8.05551246574731e-05, "loss": 0.0377, "step": 32620 }, { "grad_norm": 0.4784913659095764, "learning_rate": 8.05420349221946e-05, "loss": 0.0437, "step": 32630 }, { "grad_norm": 0.8530839681625366, "learning_rate": 8.05289418468851e-05, "loss": 0.0401, "step": 32640 }, { "grad_norm": 0.362223356962204, "learning_rate": 8.051584543297642e-05, "loss": 0.0335, "step": 32650 }, { "grad_norm": 0.602949321269989, "learning_rate": 8.050274568190074e-05, "loss": 0.0346, "step": 32660 }, { "grad_norm": 0.41307926177978516, "learning_rate": 8.048964259509067e-05, "loss": 0.0403, "step": 32670 }, { "grad_norm": 0.4782715141773224, "learning_rate": 8.047653617397914e-05, "loss": 0.0351, "step": 32680 }, { "grad_norm": 0.5375819206237793, "learning_rate": 8.046342641999941e-05, "loss": 0.0482, "step": 32690 }, { "grad_norm": 0.5690346360206604, "learning_rate": 8.045031333458517e-05, "loss": 0.0512, "step": 32700 }, { "grad_norm": 0.9029014706611633, "learning_rate": 8.043719691917047e-05, "loss": 0.044, "step": 32710 }, { "grad_norm": 0.5117477178573608, "learning_rate": 8.042407717518966e-05, "loss": 0.0377, "step": 32720 }, { "grad_norm": 0.355373352766037, "learning_rate": 8.041095410407751e-05, "loss": 0.0385, "step": 32730 }, { "grad_norm": 0.2634587287902832, "learning_rate": 8.039782770726913e-05, "loss": 0.0426, "step": 32740 }, { "grad_norm": 0.3843840956687927, "learning_rate": 8.038469798620004e-05, "loss": 0.0409, "step": 32750 }, { "grad_norm": 0.8442336320877075, "learning_rate": 8.037156494230604e-05, "loss": 0.0487, "step": 32760 }, { "grad_norm": 0.319492906332016, "learning_rate": 8.035842857702338e-05, "loss": 0.0397, "step": 32770 }, { "grad_norm": 0.4252508878707886, "learning_rate": 8.03452888917886e-05, "loss": 0.0452, "step": 32780 }, { "grad_norm": 0.2341730296611786, "learning_rate": 8.033214588803866e-05, "loss": 0.0424, "step": 32790 }, { "grad_norm": 0.4146616458892822, "learning_rate": 8.031899956721083e-05, "loss": 0.0425, "step": 32800 }, { "grad_norm": 0.5834800004959106, "learning_rate": 8.030584993074282e-05, "loss": 0.0434, "step": 32810 }, { "grad_norm": 0.4650205671787262, "learning_rate": 8.02926969800726e-05, "loss": 0.05, "step": 32820 }, { "grad_norm": 0.4124797582626343, "learning_rate": 8.027954071663859e-05, "loss": 0.0481, "step": 32830 }, { "grad_norm": 0.4287990927696228, "learning_rate": 8.026638114187954e-05, "loss": 0.0439, "step": 32840 }, { "grad_norm": 0.48364782333374023, "learning_rate": 8.025321825723456e-05, "loss": 0.0481, "step": 32850 }, { "grad_norm": 0.44401636719703674, "learning_rate": 8.02400520641431e-05, "loss": 0.05, "step": 32860 }, { "grad_norm": 0.39034581184387207, "learning_rate": 8.022688256404501e-05, "loss": 0.0362, "step": 32870 }, { "grad_norm": 0.3304535150527954, "learning_rate": 8.02137097583805e-05, "loss": 0.0438, "step": 32880 }, { "grad_norm": 0.4455600082874298, "learning_rate": 8.02005336485901e-05, "loss": 0.0431, "step": 32890 }, { "grad_norm": 0.4247726500034332, "learning_rate": 8.018735423611476e-05, "loss": 0.042, "step": 32900 }, { "grad_norm": 0.41292667388916016, "learning_rate": 8.017417152239574e-05, "loss": 0.0413, "step": 32910 }, { "grad_norm": 0.544988214969635, "learning_rate": 8.01609855088747e-05, "loss": 0.047, "step": 32920 }, { "grad_norm": 0.6288827657699585, "learning_rate": 8.014779619699362e-05, "loss": 0.0355, "step": 32930 }, { "grad_norm": 1.130139946937561, "learning_rate": 8.013460358819489e-05, "loss": 0.0429, "step": 32940 }, { "grad_norm": 0.7226163148880005, "learning_rate": 8.01214076839212e-05, "loss": 0.0438, "step": 32950 }, { "grad_norm": 0.3200037181377411, "learning_rate": 8.010820848561565e-05, "loss": 0.0339, "step": 32960 }, { "grad_norm": 2.8544507026672363, "learning_rate": 8.009500599472171e-05, "loss": 0.0455, "step": 32970 }, { "grad_norm": 0.2940044701099396, "learning_rate": 8.008180021268314e-05, "loss": 0.0268, "step": 32980 }, { "grad_norm": 0.528751015663147, "learning_rate": 8.006859114094414e-05, "loss": 0.0365, "step": 32990 }, { "grad_norm": 0.5119894742965698, "learning_rate": 8.005537878094921e-05, "loss": 0.0391, "step": 33000 }, { "grad_norm": 0.30477699637413025, "learning_rate": 8.004216313414323e-05, "loss": 0.0355, "step": 33010 }, { "grad_norm": 0.3614968955516815, "learning_rate": 8.002894420197149e-05, "loss": 0.0316, "step": 33020 }, { "grad_norm": 0.42911913990974426, "learning_rate": 8.001572198587954e-05, "loss": 0.0386, "step": 33030 }, { "grad_norm": 0.5890842080116272, "learning_rate": 8.000249648731338e-05, "loss": 0.0352, "step": 33040 }, { "grad_norm": 0.33741435408592224, "learning_rate": 7.998926770771928e-05, "loss": 0.0556, "step": 33050 }, { "grad_norm": 1.0396050214767456, "learning_rate": 7.997603564854397e-05, "loss": 0.0383, "step": 33060 }, { "grad_norm": 0.34568721055984497, "learning_rate": 7.996280031123448e-05, "loss": 0.0335, "step": 33070 }, { "grad_norm": 0.34717729687690735, "learning_rate": 7.994956169723818e-05, "loss": 0.0597, "step": 33080 }, { "grad_norm": 0.40070000290870667, "learning_rate": 7.993631980800285e-05, "loss": 0.0326, "step": 33090 }, { "grad_norm": 0.32401493191719055, "learning_rate": 7.992307464497659e-05, "loss": 0.0414, "step": 33100 }, { "grad_norm": 0.34281718730926514, "learning_rate": 7.990982620960787e-05, "loss": 0.0428, "step": 33110 }, { "grad_norm": 0.3214978277683258, "learning_rate": 7.989657450334554e-05, "loss": 0.0383, "step": 33120 }, { "grad_norm": 1.4906597137451172, "learning_rate": 7.988331952763877e-05, "loss": 0.0425, "step": 33130 }, { "grad_norm": 0.39832445979118347, "learning_rate": 7.987006128393709e-05, "loss": 0.0501, "step": 33140 }, { "grad_norm": 0.353378564119339, "learning_rate": 7.985679977369043e-05, "loss": 0.0298, "step": 33150 }, { "grad_norm": 0.6360325813293457, "learning_rate": 7.984353499834902e-05, "loss": 0.0317, "step": 33160 }, { "grad_norm": 0.28246864676475525, "learning_rate": 7.983026695936351e-05, "loss": 0.0354, "step": 33170 }, { "grad_norm": 0.4076961576938629, "learning_rate": 7.981699565818486e-05, "loss": 0.0355, "step": 33180 }, { "grad_norm": 0.5091823935508728, "learning_rate": 7.980372109626437e-05, "loss": 0.0357, "step": 33190 }, { "grad_norm": 0.5357393026351929, "learning_rate": 7.979044327505375e-05, "loss": 0.0416, "step": 33200 }, { "grad_norm": 0.3418760895729065, "learning_rate": 7.977716219600506e-05, "loss": 0.0344, "step": 33210 }, { "grad_norm": 0.398892343044281, "learning_rate": 7.97638778605707e-05, "loss": 0.0523, "step": 33220 }, { "grad_norm": 1.1649560928344727, "learning_rate": 7.975059027020338e-05, "loss": 0.0648, "step": 33230 }, { "grad_norm": 0.5979032516479492, "learning_rate": 7.973729942635623e-05, "loss": 0.044, "step": 33240 }, { "grad_norm": 0.5704482197761536, "learning_rate": 7.972400533048273e-05, "loss": 0.0595, "step": 33250 }, { "grad_norm": 0.37778985500335693, "learning_rate": 7.97107079840367e-05, "loss": 0.0482, "step": 33260 }, { "grad_norm": 0.4403383731842041, "learning_rate": 7.969740738847231e-05, "loss": 0.038, "step": 33270 }, { "grad_norm": 0.3699200451374054, "learning_rate": 7.968410354524411e-05, "loss": 0.0413, "step": 33280 }, { "grad_norm": 0.43821802735328674, "learning_rate": 7.967079645580697e-05, "loss": 0.0421, "step": 33290 }, { "grad_norm": 0.4020671546459198, "learning_rate": 7.965748612161612e-05, "loss": 0.0357, "step": 33300 }, { "grad_norm": 0.6674149632453918, "learning_rate": 7.96441725441272e-05, "loss": 0.0449, "step": 33310 }, { "grad_norm": 0.498830109834671, "learning_rate": 7.963085572479614e-05, "loss": 0.0506, "step": 33320 }, { "grad_norm": 0.4764682352542877, "learning_rate": 7.961753566507924e-05, "loss": 0.0437, "step": 33330 }, { "grad_norm": 0.46973878145217896, "learning_rate": 7.960421236643316e-05, "loss": 0.0528, "step": 33340 }, { "grad_norm": 0.4798254370689392, "learning_rate": 7.959088583031496e-05, "loss": 0.0344, "step": 33350 }, { "grad_norm": 0.3470642864704132, "learning_rate": 7.957755605818194e-05, "loss": 0.0317, "step": 33360 }, { "grad_norm": 0.30891820788383484, "learning_rate": 7.956422305149185e-05, "loss": 0.0506, "step": 33370 }, { "grad_norm": 0.5878912806510925, "learning_rate": 7.95508868117028e-05, "loss": 0.0504, "step": 33380 }, { "grad_norm": 0.6042540073394775, "learning_rate": 7.953754734027318e-05, "loss": 0.0466, "step": 33390 }, { "grad_norm": 0.40953874588012695, "learning_rate": 7.952420463866182e-05, "loss": 0.0367, "step": 33400 }, { "grad_norm": 0.28918883204460144, "learning_rate": 7.951085870832782e-05, "loss": 0.0394, "step": 33410 }, { "grad_norm": 0.2645898163318634, "learning_rate": 7.949750955073067e-05, "loss": 0.0396, "step": 33420 }, { "grad_norm": 0.3435303866863251, "learning_rate": 7.948415716733022e-05, "loss": 0.0367, "step": 33430 }, { "grad_norm": 0.4496237337589264, "learning_rate": 7.947080155958669e-05, "loss": 0.0451, "step": 33440 }, { "grad_norm": 0.48213356733322144, "learning_rate": 7.94574427289606e-05, "loss": 0.0367, "step": 33450 }, { "grad_norm": 0.3716524839401245, "learning_rate": 7.944408067691284e-05, "loss": 0.0296, "step": 33460 }, { "grad_norm": 0.29801619052886963, "learning_rate": 7.943071540490473e-05, "loss": 0.0429, "step": 33470 }, { "grad_norm": 0.37627294659614563, "learning_rate": 7.94173469143978e-05, "loss": 0.0437, "step": 33480 }, { "grad_norm": 0.49420785903930664, "learning_rate": 7.940397520685406e-05, "loss": 0.0415, "step": 33490 }, { "grad_norm": 0.6022925972938538, "learning_rate": 7.939060028373577e-05, "loss": 0.0551, "step": 33500 }, { "grad_norm": 0.4479048550128937, "learning_rate": 7.937722214650565e-05, "loss": 0.0577, "step": 33510 }, { "grad_norm": 0.3716249167919159, "learning_rate": 7.936384079662666e-05, "loss": 0.0425, "step": 33520 }, { "grad_norm": 0.5881301760673523, "learning_rate": 7.93504562355622e-05, "loss": 0.0454, "step": 33530 }, { "grad_norm": 0.4491477906703949, "learning_rate": 7.933706846477599e-05, "loss": 0.0495, "step": 33540 }, { "grad_norm": 0.4020298719406128, "learning_rate": 7.932367748573206e-05, "loss": 0.0399, "step": 33550 }, { "grad_norm": 0.4998050034046173, "learning_rate": 7.931028329989485e-05, "loss": 0.0405, "step": 33560 }, { "grad_norm": 0.6972233057022095, "learning_rate": 7.929688590872913e-05, "loss": 0.045, "step": 33570 }, { "grad_norm": 0.39428195357322693, "learning_rate": 7.928348531370003e-05, "loss": 0.0396, "step": 33580 }, { "grad_norm": 0.4816937744617462, "learning_rate": 7.927008151627297e-05, "loss": 0.0427, "step": 33590 }, { "grad_norm": 0.8770269751548767, "learning_rate": 7.925667451791383e-05, "loss": 0.0419, "step": 33600 }, { "grad_norm": 0.6923696994781494, "learning_rate": 7.924326432008874e-05, "loss": 0.0407, "step": 33610 }, { "grad_norm": 0.4040279984474182, "learning_rate": 7.922985092426422e-05, "loss": 0.0444, "step": 33620 }, { "grad_norm": 0.34419891238212585, "learning_rate": 7.921643433190717e-05, "loss": 0.0472, "step": 33630 }, { "grad_norm": 0.4357231855392456, "learning_rate": 7.920301454448478e-05, "loss": 0.0419, "step": 33640 }, { "grad_norm": 0.3440767526626587, "learning_rate": 7.918959156346461e-05, "loss": 0.0363, "step": 33650 }, { "grad_norm": 0.38937997817993164, "learning_rate": 7.91761653903146e-05, "loss": 0.0341, "step": 33660 }, { "grad_norm": 0.5877862572669983, "learning_rate": 7.916273602650302e-05, "loss": 0.0423, "step": 33670 }, { "grad_norm": 0.4091138243675232, "learning_rate": 7.914930347349847e-05, "loss": 0.0352, "step": 33680 }, { "grad_norm": 0.40461063385009766, "learning_rate": 7.913586773276992e-05, "loss": 0.0355, "step": 33690 }, { "grad_norm": 0.4649813175201416, "learning_rate": 7.912242880578667e-05, "loss": 0.0319, "step": 33700 }, { "grad_norm": 0.43441128730773926, "learning_rate": 7.910898669401839e-05, "loss": 0.0413, "step": 33710 }, { "grad_norm": 0.39358046650886536, "learning_rate": 7.909554139893511e-05, "loss": 0.0365, "step": 33720 }, { "grad_norm": 0.35143858194351196, "learning_rate": 7.908209292200715e-05, "loss": 0.0336, "step": 33730 }, { "grad_norm": 0.3438955247402191, "learning_rate": 7.906864126470523e-05, "loss": 0.0315, "step": 33740 }, { "grad_norm": 0.4858776330947876, "learning_rate": 7.905518642850041e-05, "loss": 0.0347, "step": 33750 }, { "grad_norm": 0.4650120437145233, "learning_rate": 7.904172841486409e-05, "loss": 0.0438, "step": 33760 }, { "grad_norm": 0.6772975325584412, "learning_rate": 7.902826722526801e-05, "loss": 0.0395, "step": 33770 }, { "grad_norm": 0.3941211700439453, "learning_rate": 7.901480286118427e-05, "loss": 0.0332, "step": 33780 }, { "grad_norm": 0.4730449616909027, "learning_rate": 7.900133532408531e-05, "loss": 0.0333, "step": 33790 }, { "grad_norm": 0.6421687006950378, "learning_rate": 7.898786461544395e-05, "loss": 0.0391, "step": 33800 }, { "grad_norm": 0.575065553188324, "learning_rate": 7.897439073673325e-05, "loss": 0.0356, "step": 33810 }, { "grad_norm": 0.3668065071105957, "learning_rate": 7.896091368942677e-05, "loss": 0.0359, "step": 33820 }, { "grad_norm": 0.3744159936904907, "learning_rate": 7.894743347499832e-05, "loss": 0.0377, "step": 33830 }, { "grad_norm": 0.32046443223953247, "learning_rate": 7.893395009492203e-05, "loss": 0.0389, "step": 33840 }, { "grad_norm": 0.4035281836986542, "learning_rate": 7.892046355067248e-05, "loss": 0.033, "step": 33850 }, { "grad_norm": 0.35665270686149597, "learning_rate": 7.890697384372451e-05, "loss": 0.0365, "step": 33860 }, { "grad_norm": 0.860231876373291, "learning_rate": 7.889348097555336e-05, "loss": 0.0378, "step": 33870 }, { "grad_norm": 0.32906150817871094, "learning_rate": 7.887998494763455e-05, "loss": 0.0337, "step": 33880 }, { "grad_norm": 0.29543671011924744, "learning_rate": 7.886648576144404e-05, "loss": 0.0294, "step": 33890 }, { "grad_norm": 0.29869550466537476, "learning_rate": 7.885298341845802e-05, "loss": 0.0279, "step": 33900 }, { "grad_norm": 0.4052099287509918, "learning_rate": 7.883947792015311e-05, "loss": 0.0436, "step": 33910 }, { "grad_norm": 0.6500746011734009, "learning_rate": 7.882596926800628e-05, "loss": 0.0626, "step": 33920 }, { "grad_norm": 0.3955408036708832, "learning_rate": 7.881245746349477e-05, "loss": 0.0413, "step": 33930 }, { "grad_norm": 0.5164304375648499, "learning_rate": 7.879894250809623e-05, "loss": 0.0371, "step": 33940 }, { "grad_norm": 0.7296827435493469, "learning_rate": 7.878542440328865e-05, "loss": 0.0605, "step": 33950 }, { "grad_norm": 0.41486191749572754, "learning_rate": 7.877190315055031e-05, "loss": 0.0369, "step": 33960 }, { "grad_norm": 0.7898417115211487, "learning_rate": 7.875837875135991e-05, "loss": 0.0398, "step": 33970 }, { "grad_norm": 0.599697470664978, "learning_rate": 7.874485120719646e-05, "loss": 0.0412, "step": 33980 }, { "grad_norm": 0.3837394118309021, "learning_rate": 7.873132051953928e-05, "loss": 0.0333, "step": 33990 }, { "grad_norm": 0.48743700981140137, "learning_rate": 7.87177866898681e-05, "loss": 0.0492, "step": 34000 }, { "grad_norm": 0.5616918802261353, "learning_rate": 7.870424971966294e-05, "loss": 0.0348, "step": 34010 }, { "grad_norm": 0.27352777123451233, "learning_rate": 7.869070961040419e-05, "loss": 0.0405, "step": 34020 }, { "grad_norm": 0.6459752917289734, "learning_rate": 7.867716636357257e-05, "loss": 0.039, "step": 34030 }, { "grad_norm": 0.32534679770469666, "learning_rate": 7.866361998064915e-05, "loss": 0.043, "step": 34040 }, { "grad_norm": 0.6143462061882019, "learning_rate": 7.865007046311534e-05, "loss": 0.0426, "step": 34050 }, { "grad_norm": 0.35061731934547424, "learning_rate": 7.86365178124529e-05, "loss": 0.0402, "step": 34060 }, { "grad_norm": 0.5516136288642883, "learning_rate": 7.862296203014394e-05, "loss": 0.0426, "step": 34070 }, { "grad_norm": 0.4760338366031647, "learning_rate": 7.860940311767088e-05, "loss": 0.0484, "step": 34080 }, { "grad_norm": 0.38754889369010925, "learning_rate": 7.85958410765165e-05, "loss": 0.0564, "step": 34090 }, { "grad_norm": 0.3782710134983063, "learning_rate": 7.858227590816394e-05, "loss": 0.0425, "step": 34100 }, { "grad_norm": 0.6609526872634888, "learning_rate": 7.856870761409664e-05, "loss": 0.0507, "step": 34110 }, { "grad_norm": 0.35747140645980835, "learning_rate": 7.855513619579846e-05, "loss": 0.036, "step": 34120 }, { "grad_norm": 0.5310841798782349, "learning_rate": 7.85415616547535e-05, "loss": 0.0408, "step": 34130 }, { "grad_norm": 0.5232899188995361, "learning_rate": 7.852798399244627e-05, "loss": 0.0413, "step": 34140 }, { "grad_norm": 0.3935140073299408, "learning_rate": 7.851440321036161e-05, "loss": 0.0339, "step": 34150 }, { "grad_norm": 0.3551129996776581, "learning_rate": 7.850081930998468e-05, "loss": 0.0413, "step": 34160 }, { "grad_norm": 0.32026389241218567, "learning_rate": 7.8487232292801e-05, "loss": 0.045, "step": 34170 }, { "grad_norm": 0.2947847545146942, "learning_rate": 7.847364216029642e-05, "loss": 0.0459, "step": 34180 }, { "grad_norm": 0.40978071093559265, "learning_rate": 7.846004891395716e-05, "loss": 0.038, "step": 34190 }, { "grad_norm": 0.3418810963630676, "learning_rate": 7.844645255526972e-05, "loss": 0.0361, "step": 34200 }, { "grad_norm": 0.9622474908828735, "learning_rate": 7.843285308572101e-05, "loss": 0.0385, "step": 34210 }, { "grad_norm": 0.39285436272621155, "learning_rate": 7.841925050679823e-05, "loss": 0.039, "step": 34220 }, { "grad_norm": 0.581369161605835, "learning_rate": 7.840564481998895e-05, "loss": 0.0398, "step": 34230 }, { "grad_norm": 0.6356263160705566, "learning_rate": 7.839203602678105e-05, "loss": 0.041, "step": 34240 }, { "grad_norm": 0.6116909384727478, "learning_rate": 7.837842412866279e-05, "loss": 0.0392, "step": 34250 }, { "grad_norm": 0.5457800030708313, "learning_rate": 7.836480912712272e-05, "loss": 0.0455, "step": 34260 }, { "grad_norm": 0.5732704401016235, "learning_rate": 7.835119102364976e-05, "loss": 0.0409, "step": 34270 }, { "grad_norm": 0.46153122186660767, "learning_rate": 7.83375698197332e-05, "loss": 0.0416, "step": 34280 }, { "grad_norm": 0.4330441355705261, "learning_rate": 7.83239455168626e-05, "loss": 0.0401, "step": 34290 }, { "grad_norm": 0.4032882750034332, "learning_rate": 7.83103181165279e-05, "loss": 0.0367, "step": 34300 }, { "grad_norm": 0.44865137338638306, "learning_rate": 7.829668762021937e-05, "loss": 0.0384, "step": 34310 }, { "grad_norm": 0.49443063139915466, "learning_rate": 7.828305402942764e-05, "loss": 0.041, "step": 34320 }, { "grad_norm": 1.160690188407898, "learning_rate": 7.826941734564363e-05, "loss": 0.0344, "step": 34330 }, { "grad_norm": 0.5069708228111267, "learning_rate": 7.825577757035865e-05, "loss": 0.033, "step": 34340 }, { "grad_norm": 0.5979081392288208, "learning_rate": 7.824213470506431e-05, "loss": 0.0373, "step": 34350 }, { "grad_norm": 0.5257519483566284, "learning_rate": 7.822848875125257e-05, "loss": 0.0432, "step": 34360 }, { "grad_norm": 0.4312446713447571, "learning_rate": 7.821483971041576e-05, "loss": 0.0389, "step": 34370 }, { "grad_norm": 0.3076700270175934, "learning_rate": 7.820118758404649e-05, "loss": 0.0393, "step": 34380 }, { "grad_norm": 0.3010885417461395, "learning_rate": 7.818753237363776e-05, "loss": 0.0395, "step": 34390 }, { "grad_norm": 0.365468293428421, "learning_rate": 7.817387408068286e-05, "loss": 0.0348, "step": 34400 }, { "grad_norm": 0.40750300884246826, "learning_rate": 7.816021270667544e-05, "loss": 0.0459, "step": 34410 }, { "grad_norm": 0.7347429990768433, "learning_rate": 7.81465482531095e-05, "loss": 0.0441, "step": 34420 }, { "grad_norm": 0.470058411359787, "learning_rate": 7.813288072147938e-05, "loss": 0.0359, "step": 34430 }, { "grad_norm": 0.4708084166049957, "learning_rate": 7.811921011327972e-05, "loss": 0.0444, "step": 34440 }, { "grad_norm": 0.547634482383728, "learning_rate": 7.810553643000549e-05, "loss": 0.0518, "step": 34450 }, { "grad_norm": 0.539611279964447, "learning_rate": 7.809185967315206e-05, "loss": 0.0489, "step": 34460 }, { "grad_norm": 0.4013727903366089, "learning_rate": 7.80781798442151e-05, "loss": 0.0461, "step": 34470 }, { "grad_norm": 0.3741285800933838, "learning_rate": 7.806449694469061e-05, "loss": 0.0362, "step": 34480 }, { "grad_norm": 0.38440948724746704, "learning_rate": 7.805081097607492e-05, "loss": 0.0397, "step": 34490 }, { "grad_norm": 0.49074381589889526, "learning_rate": 7.803712193986474e-05, "loss": 0.0404, "step": 34500 }, { "grad_norm": 0.4097536504268646, "learning_rate": 7.802342983755702e-05, "loss": 0.0403, "step": 34510 }, { "grad_norm": 0.64665687084198, "learning_rate": 7.800973467064918e-05, "loss": 0.0483, "step": 34520 }, { "grad_norm": 0.5059831142425537, "learning_rate": 7.799603644063884e-05, "loss": 0.0513, "step": 34530 }, { "grad_norm": 0.42088955640792847, "learning_rate": 7.798233514902405e-05, "loss": 0.0481, "step": 34540 }, { "grad_norm": 0.7497363090515137, "learning_rate": 7.796863079730318e-05, "loss": 0.0487, "step": 34550 }, { "grad_norm": 0.42015740275382996, "learning_rate": 7.795492338697488e-05, "loss": 0.0422, "step": 34560 }, { "grad_norm": 0.32539165019989014, "learning_rate": 7.794121291953819e-05, "loss": 0.0461, "step": 34570 }, { "grad_norm": 0.8733068108558655, "learning_rate": 7.792749939649246e-05, "loss": 0.0498, "step": 34580 }, { "grad_norm": 0.43100258708000183, "learning_rate": 7.79137828193374e-05, "loss": 0.0355, "step": 34590 }, { "grad_norm": 0.4014357030391693, "learning_rate": 7.790006318957301e-05, "loss": 0.0454, "step": 34600 }, { "grad_norm": 0.4514414072036743, "learning_rate": 7.788634050869965e-05, "loss": 0.0386, "step": 34610 }, { "grad_norm": 0.23735518753528595, "learning_rate": 7.787261477821803e-05, "loss": 0.0392, "step": 34620 }, { "grad_norm": 0.4726032316684723, "learning_rate": 7.785888599962916e-05, "loss": 0.04, "step": 34630 }, { "grad_norm": 0.4838544726371765, "learning_rate": 7.784515417443439e-05, "loss": 0.0334, "step": 34640 }, { "grad_norm": 0.40349671244621277, "learning_rate": 7.783141930413545e-05, "loss": 0.0395, "step": 34650 }, { "grad_norm": 0.34406042098999023, "learning_rate": 7.78176813902343e-05, "loss": 0.0416, "step": 34660 }, { "grad_norm": 0.3506805896759033, "learning_rate": 7.780394043423336e-05, "loss": 0.0382, "step": 34670 }, { "grad_norm": 0.4515925943851471, "learning_rate": 7.77901964376353e-05, "loss": 0.0483, "step": 34680 }, { "grad_norm": 0.3606894612312317, "learning_rate": 7.777644940194316e-05, "loss": 0.0455, "step": 34690 }, { "grad_norm": 0.4302610158920288, "learning_rate": 7.776269932866023e-05, "loss": 0.0414, "step": 34700 }, { "grad_norm": 0.30824774503707886, "learning_rate": 7.774894621929026e-05, "loss": 0.0412, "step": 34710 }, { "grad_norm": 0.34618109464645386, "learning_rate": 7.773519007533725e-05, "loss": 0.0425, "step": 34720 }, { "grad_norm": 0.406938374042511, "learning_rate": 7.772143089830556e-05, "loss": 0.0423, "step": 34730 }, { "grad_norm": 0.47539591789245605, "learning_rate": 7.770766868969985e-05, "loss": 0.0478, "step": 34740 }, { "grad_norm": 0.978001058101654, "learning_rate": 7.769390345102518e-05, "loss": 0.0454, "step": 34750 }, { "grad_norm": 0.565405011177063, "learning_rate": 7.768013518378683e-05, "loss": 0.0565, "step": 34760 }, { "grad_norm": 0.49430614709854126, "learning_rate": 7.766636388949053e-05, "loss": 0.0346, "step": 34770 }, { "grad_norm": 1.0439467430114746, "learning_rate": 7.765258956964229e-05, "loss": 0.0559, "step": 34780 }, { "grad_norm": 0.5307043194770813, "learning_rate": 7.76388122257484e-05, "loss": 0.0496, "step": 34790 }, { "grad_norm": 0.4636370837688446, "learning_rate": 7.762503185931558e-05, "loss": 0.0354, "step": 34800 }, { "grad_norm": 0.7339120507240295, "learning_rate": 7.76112484718508e-05, "loss": 0.0375, "step": 34810 }, { "grad_norm": 0.6217840313911438, "learning_rate": 7.75974620648614e-05, "loss": 0.0476, "step": 34820 }, { "grad_norm": 0.4555719792842865, "learning_rate": 7.758367263985503e-05, "loss": 0.0384, "step": 34830 }, { "grad_norm": 0.5233906507492065, "learning_rate": 7.75698801983397e-05, "loss": 0.035, "step": 34840 }, { "grad_norm": 0.6840873956680298, "learning_rate": 7.755608474182372e-05, "loss": 0.0461, "step": 34850 }, { "grad_norm": 0.34349772334098816, "learning_rate": 7.754228627181574e-05, "loss": 0.0399, "step": 34860 }, { "grad_norm": 0.6851680874824524, "learning_rate": 7.752848478982476e-05, "loss": 0.0375, "step": 34870 }, { "grad_norm": 0.5211324691772461, "learning_rate": 7.751468029736006e-05, "loss": 0.0414, "step": 34880 }, { "grad_norm": 0.41524752974510193, "learning_rate": 7.750087279593129e-05, "loss": 0.0496, "step": 34890 }, { "grad_norm": 0.37424829602241516, "learning_rate": 7.748706228704843e-05, "loss": 0.0413, "step": 34900 }, { "grad_norm": 0.3020016551017761, "learning_rate": 7.747324877222176e-05, "loss": 0.045, "step": 34910 }, { "grad_norm": 0.311024010181427, "learning_rate": 7.745943225296188e-05, "loss": 0.0392, "step": 34920 }, { "grad_norm": 0.3480997085571289, "learning_rate": 7.744561273077981e-05, "loss": 0.0328, "step": 34930 }, { "grad_norm": 0.33123505115509033, "learning_rate": 7.743179020718678e-05, "loss": 0.0381, "step": 34940 }, { "grad_norm": 0.3815801739692688, "learning_rate": 7.741796468369443e-05, "loss": 0.0451, "step": 34950 }, { "grad_norm": 0.3838125467300415, "learning_rate": 7.740413616181466e-05, "loss": 0.0373, "step": 34960 }, { "grad_norm": 0.41653963923454285, "learning_rate": 7.739030464305978e-05, "loss": 0.0373, "step": 34970 }, { "grad_norm": 0.32695960998535156, "learning_rate": 7.737647012894235e-05, "loss": 0.0384, "step": 34980 }, { "grad_norm": 0.27098342776298523, "learning_rate": 7.736263262097532e-05, "loss": 0.0354, "step": 34990 }, { "grad_norm": 0.3940219283103943, "learning_rate": 7.734879212067192e-05, "loss": 0.0349, "step": 35000 }, { "grad_norm": 0.36091333627700806, "learning_rate": 7.733494862954573e-05, "loss": 0.0373, "step": 35010 }, { "grad_norm": 0.3512953817844391, "learning_rate": 7.732110214911066e-05, "loss": 0.0331, "step": 35020 }, { "grad_norm": 0.428838849067688, "learning_rate": 7.730725268088094e-05, "loss": 0.0412, "step": 35030 }, { "grad_norm": 0.3843468725681305, "learning_rate": 7.729340022637112e-05, "loss": 0.0385, "step": 35040 }, { "grad_norm": 0.5124309062957764, "learning_rate": 7.727954478709607e-05, "loss": 0.0286, "step": 35050 }, { "grad_norm": 0.24809011816978455, "learning_rate": 7.726568636457103e-05, "loss": 0.0319, "step": 35060 }, { "grad_norm": 0.623034656047821, "learning_rate": 7.725182496031153e-05, "loss": 0.0332, "step": 35070 }, { "grad_norm": 0.35204270482063293, "learning_rate": 7.723796057583341e-05, "loss": 0.0395, "step": 35080 }, { "grad_norm": 0.3685414493083954, "learning_rate": 7.722409321265287e-05, "loss": 0.0432, "step": 35090 }, { "grad_norm": 0.38538119196891785, "learning_rate": 7.721022287228645e-05, "loss": 0.0456, "step": 35100 }, { "grad_norm": 0.8998087644577026, "learning_rate": 7.719634955625094e-05, "loss": 0.043, "step": 35110 }, { "grad_norm": 0.4675561785697937, "learning_rate": 7.718247326606355e-05, "loss": 0.0312, "step": 35120 }, { "grad_norm": 0.4774327576160431, "learning_rate": 7.716859400324175e-05, "loss": 0.0336, "step": 35130 }, { "grad_norm": 0.365633487701416, "learning_rate": 7.715471176930335e-05, "loss": 0.0502, "step": 35140 }, { "grad_norm": 0.48322010040283203, "learning_rate": 7.714082656576651e-05, "loss": 0.0408, "step": 35150 }, { "grad_norm": 0.8181137442588806, "learning_rate": 7.712693839414968e-05, "loss": 0.039, "step": 35160 }, { "grad_norm": 0.29627832770347595, "learning_rate": 7.711304725597164e-05, "loss": 0.0383, "step": 35170 }, { "grad_norm": 0.33085566759109497, "learning_rate": 7.709915315275151e-05, "loss": 0.0437, "step": 35180 }, { "grad_norm": 0.47082626819610596, "learning_rate": 7.708525608600876e-05, "loss": 0.0497, "step": 35190 }, { "grad_norm": 0.3592323064804077, "learning_rate": 7.707135605726311e-05, "loss": 0.0379, "step": 35200 }, { "grad_norm": 0.3720396161079407, "learning_rate": 7.705745306803466e-05, "loss": 0.0552, "step": 35210 }, { "grad_norm": 0.3879997432231903, "learning_rate": 7.704354711984383e-05, "loss": 0.0376, "step": 35220 }, { "grad_norm": 0.2834174931049347, "learning_rate": 7.702963821421133e-05, "loss": 0.0346, "step": 35230 }, { "grad_norm": 0.29193785786628723, "learning_rate": 7.701572635265826e-05, "loss": 0.0399, "step": 35240 }, { "grad_norm": 0.36529305577278137, "learning_rate": 7.700181153670596e-05, "loss": 0.0392, "step": 35250 }, { "grad_norm": 0.3117501735687256, "learning_rate": 7.698789376787616e-05, "loss": 0.0444, "step": 35260 }, { "grad_norm": 0.4730157256126404, "learning_rate": 7.697397304769085e-05, "loss": 0.0456, "step": 35270 }, { "grad_norm": 0.4684808850288391, "learning_rate": 7.696004937767241e-05, "loss": 0.0488, "step": 35280 }, { "grad_norm": 0.5982465147972107, "learning_rate": 7.694612275934352e-05, "loss": 0.037, "step": 35290 }, { "grad_norm": 0.4418020248413086, "learning_rate": 7.693219319422714e-05, "loss": 0.0449, "step": 35300 }, { "grad_norm": 0.7729032039642334, "learning_rate": 7.69182606838466e-05, "loss": 0.0378, "step": 35310 }, { "grad_norm": 0.34382617473602295, "learning_rate": 7.690432522972558e-05, "loss": 0.0314, "step": 35320 }, { "grad_norm": 0.487794429063797, "learning_rate": 7.689038683338796e-05, "loss": 0.0367, "step": 35330 }, { "grad_norm": 0.5323545932769775, "learning_rate": 7.687644549635808e-05, "loss": 0.0289, "step": 35340 }, { "grad_norm": 0.4264780580997467, "learning_rate": 7.686250122016053e-05, "loss": 0.0348, "step": 35350 }, { "grad_norm": 0.6588394045829773, "learning_rate": 7.684855400632022e-05, "loss": 0.0359, "step": 35360 }, { "grad_norm": 0.5991427302360535, "learning_rate": 7.683460385636243e-05, "loss": 0.035, "step": 35370 }, { "grad_norm": 0.47762632369995117, "learning_rate": 7.68206507718127e-05, "loss": 0.0435, "step": 35380 }, { "grad_norm": 0.5180787444114685, "learning_rate": 7.680669475419693e-05, "loss": 0.0544, "step": 35390 }, { "grad_norm": 0.3794964849948883, "learning_rate": 7.679273580504132e-05, "loss": 0.0352, "step": 35400 }, { "grad_norm": 0.49176910519599915, "learning_rate": 7.67787739258724e-05, "loss": 0.0338, "step": 35410 }, { "grad_norm": 0.39843037724494934, "learning_rate": 7.676480911821705e-05, "loss": 0.0468, "step": 35420 }, { "grad_norm": 0.5115394592285156, "learning_rate": 7.675084138360238e-05, "loss": 0.0524, "step": 35430 }, { "grad_norm": 0.5436171293258667, "learning_rate": 7.673687072355592e-05, "loss": 0.0379, "step": 35440 }, { "grad_norm": 0.588992178440094, "learning_rate": 7.67228971396055e-05, "loss": 0.0359, "step": 35450 }, { "grad_norm": 0.7235599160194397, "learning_rate": 7.670892063327922e-05, "loss": 0.0433, "step": 35460 }, { "grad_norm": 0.2851807475090027, "learning_rate": 7.669494120610552e-05, "loss": 0.0406, "step": 35470 }, { "grad_norm": 0.30024200677871704, "learning_rate": 7.66809588596132e-05, "loss": 0.0493, "step": 35480 }, { "grad_norm": 0.40195947885513306, "learning_rate": 7.666697359533132e-05, "loss": 0.0447, "step": 35490 }, { "grad_norm": 0.7002694606781006, "learning_rate": 7.665298541478932e-05, "loss": 0.0532, "step": 35500 }, { "grad_norm": 0.4872800409793854, "learning_rate": 7.66389943195169e-05, "loss": 0.0487, "step": 35510 }, { "grad_norm": 0.31284642219543457, "learning_rate": 7.66250003110441e-05, "loss": 0.044, "step": 35520 }, { "grad_norm": 0.38482868671417236, "learning_rate": 7.66110033909013e-05, "loss": 0.0434, "step": 35530 }, { "grad_norm": 0.23220092058181763, "learning_rate": 7.659700356061918e-05, "loss": 0.0401, "step": 35540 }, { "grad_norm": 0.2621551752090454, "learning_rate": 7.658300082172875e-05, "loss": 0.0369, "step": 35550 }, { "grad_norm": 0.35735002160072327, "learning_rate": 7.65689951757613e-05, "loss": 0.0412, "step": 35560 }, { "grad_norm": 0.39448776841163635, "learning_rate": 7.655498662424849e-05, "loss": 0.0323, "step": 35570 }, { "grad_norm": 0.33115774393081665, "learning_rate": 7.654097516872227e-05, "loss": 0.0322, "step": 35580 }, { "grad_norm": 0.493624746799469, "learning_rate": 7.65269608107149e-05, "loss": 0.0434, "step": 35590 }, { "grad_norm": 0.32781165838241577, "learning_rate": 7.6512943551759e-05, "loss": 0.0518, "step": 35600 }, { "grad_norm": 0.39471280574798584, "learning_rate": 7.649892339338743e-05, "loss": 0.0324, "step": 35610 }, { "grad_norm": 0.79950350522995, "learning_rate": 7.648490033713344e-05, "loss": 0.039, "step": 35620 }, { "grad_norm": 0.8366092443466187, "learning_rate": 7.647087438453058e-05, "loss": 0.0309, "step": 35630 }, { "grad_norm": 0.45896610617637634, "learning_rate": 7.64568455371127e-05, "loss": 0.0367, "step": 35640 }, { "grad_norm": 0.5909751653671265, "learning_rate": 7.644281379641396e-05, "loss": 0.0351, "step": 35650 }, { "grad_norm": 0.36708950996398926, "learning_rate": 7.642877916396887e-05, "loss": 0.044, "step": 35660 }, { "grad_norm": 0.26329806447029114, "learning_rate": 7.641474164131221e-05, "loss": 0.0396, "step": 35670 }, { "grad_norm": 0.6580035090446472, "learning_rate": 7.640070122997913e-05, "loss": 0.0403, "step": 35680 }, { "grad_norm": 0.794134259223938, "learning_rate": 7.638665793150505e-05, "loss": 0.039, "step": 35690 }, { "grad_norm": 0.6033146977424622, "learning_rate": 7.637261174742574e-05, "loss": 0.0448, "step": 35700 }, { "grad_norm": 0.3220321834087372, "learning_rate": 7.635856267927725e-05, "loss": 0.0443, "step": 35710 }, { "grad_norm": 0.5217519402503967, "learning_rate": 7.634451072859597e-05, "loss": 0.0331, "step": 35720 }, { "grad_norm": 1.1322437524795532, "learning_rate": 7.633045589691863e-05, "loss": 0.0474, "step": 35730 }, { "grad_norm": 0.2896076738834381, "learning_rate": 7.63163981857822e-05, "loss": 0.0328, "step": 35740 }, { "grad_norm": 0.4890381693840027, "learning_rate": 7.630233759672403e-05, "loss": 0.0392, "step": 35750 }, { "grad_norm": 0.2819179594516754, "learning_rate": 7.628827413128177e-05, "loss": 0.0319, "step": 35760 }, { "grad_norm": 0.3703165352344513, "learning_rate": 7.627420779099338e-05, "loss": 0.0446, "step": 35770 }, { "grad_norm": 0.3570004999637604, "learning_rate": 7.626013857739711e-05, "loss": 0.0494, "step": 35780 }, { "grad_norm": 0.2365097999572754, "learning_rate": 7.624606649203158e-05, "loss": 0.0316, "step": 35790 }, { "grad_norm": 0.47649964690208435, "learning_rate": 7.623199153643569e-05, "loss": 0.0418, "step": 35800 }, { "grad_norm": 0.3640861511230469, "learning_rate": 7.621791371214863e-05, "loss": 0.0418, "step": 35810 }, { "grad_norm": 0.57331782579422, "learning_rate": 7.620383302070995e-05, "loss": 0.0397, "step": 35820 }, { "grad_norm": 0.47297805547714233, "learning_rate": 7.61897494636595e-05, "loss": 0.0386, "step": 35830 }, { "grad_norm": 0.4383718967437744, "learning_rate": 7.617566304253739e-05, "loss": 0.0341, "step": 35840 }, { "grad_norm": 0.48496365547180176, "learning_rate": 7.616157375888416e-05, "loss": 0.0386, "step": 35850 }, { "grad_norm": 1.0945638418197632, "learning_rate": 7.614748161424053e-05, "loss": 0.0389, "step": 35860 }, { "grad_norm": 0.3471943736076355, "learning_rate": 7.613338661014763e-05, "loss": 0.0396, "step": 35870 }, { "grad_norm": 0.47178736329078674, "learning_rate": 7.611928874814686e-05, "loss": 0.0374, "step": 35880 }, { "grad_norm": 0.4193592369556427, "learning_rate": 7.610518802977996e-05, "loss": 0.0298, "step": 35890 }, { "grad_norm": 0.29084351658821106, "learning_rate": 7.609108445658893e-05, "loss": 0.0399, "step": 35900 }, { "grad_norm": 0.2818479835987091, "learning_rate": 7.607697803011612e-05, "loss": 0.051, "step": 35910 }, { "grad_norm": 0.3911390006542206, "learning_rate": 7.606286875190421e-05, "loss": 0.0365, "step": 35920 }, { "grad_norm": 0.4851650893688202, "learning_rate": 7.604875662349617e-05, "loss": 0.0422, "step": 35930 }, { "grad_norm": 0.45234620571136475, "learning_rate": 7.603464164643525e-05, "loss": 0.0525, "step": 35940 }, { "grad_norm": 0.49013885855674744, "learning_rate": 7.602052382226507e-05, "loss": 0.033, "step": 35950 }, { "grad_norm": 0.6160802245140076, "learning_rate": 7.600640315252954e-05, "loss": 0.0418, "step": 35960 }, { "grad_norm": 0.6054615378379822, "learning_rate": 7.599227963877284e-05, "loss": 0.0494, "step": 35970 }, { "grad_norm": 0.9006298780441284, "learning_rate": 7.597815328253954e-05, "loss": 0.0437, "step": 35980 }, { "grad_norm": 0.5259411334991455, "learning_rate": 7.596402408537443e-05, "loss": 0.0413, "step": 35990 }, { "grad_norm": 0.5271950960159302, "learning_rate": 7.59498920488227e-05, "loss": 0.0508, "step": 36000 }, { "grad_norm": 0.5487973093986511, "learning_rate": 7.593575717442979e-05, "loss": 0.0384, "step": 36010 }, { "grad_norm": 0.6727715134620667, "learning_rate": 7.592161946374147e-05, "loss": 0.0449, "step": 36020 }, { "grad_norm": 0.4744799733161926, "learning_rate": 7.590747891830381e-05, "loss": 0.0365, "step": 36030 }, { "grad_norm": 0.506463885307312, "learning_rate": 7.58933355396632e-05, "loss": 0.0395, "step": 36040 }, { "grad_norm": 0.38506150245666504, "learning_rate": 7.587918932936636e-05, "loss": 0.0326, "step": 36050 }, { "grad_norm": 0.3308866620063782, "learning_rate": 7.586504028896028e-05, "loss": 0.0391, "step": 36060 }, { "grad_norm": 0.3760378956794739, "learning_rate": 7.585088841999228e-05, "loss": 0.0361, "step": 36070 }, { "grad_norm": 0.8899308443069458, "learning_rate": 7.583673372400999e-05, "loss": 0.0437, "step": 36080 }, { "grad_norm": 0.39729058742523193, "learning_rate": 7.582257620256134e-05, "loss": 0.0417, "step": 36090 }, { "grad_norm": 0.4834009110927582, "learning_rate": 7.580841585719458e-05, "loss": 0.0439, "step": 36100 }, { "grad_norm": 0.3580949306488037, "learning_rate": 7.579425268945825e-05, "loss": 0.0295, "step": 36110 }, { "grad_norm": 0.32383930683135986, "learning_rate": 7.578008670090127e-05, "loss": 0.0369, "step": 36120 }, { "grad_norm": 0.46606311202049255, "learning_rate": 7.576591789307272e-05, "loss": 0.0379, "step": 36130 }, { "grad_norm": 0.5259513258934021, "learning_rate": 7.575174626752216e-05, "loss": 0.0491, "step": 36140 }, { "grad_norm": 0.5438039302825928, "learning_rate": 7.573757182579934e-05, "loss": 0.0346, "step": 36150 }, { "grad_norm": 1.370609164237976, "learning_rate": 7.572339456945435e-05, "loss": 0.0427, "step": 36160 }, { "grad_norm": 0.599475085735321, "learning_rate": 7.570921450003762e-05, "loss": 0.0422, "step": 36170 }, { "grad_norm": 0.33169251680374146, "learning_rate": 7.569503161909983e-05, "loss": 0.0456, "step": 36180 }, { "grad_norm": 0.4023124873638153, "learning_rate": 7.568084592819202e-05, "loss": 0.0457, "step": 36190 }, { "grad_norm": 0.44447556138038635, "learning_rate": 7.566665742886551e-05, "loss": 0.0494, "step": 36200 }, { "grad_norm": 0.6382575035095215, "learning_rate": 7.565246612267194e-05, "loss": 0.0478, "step": 36210 }, { "grad_norm": 0.3962046802043915, "learning_rate": 7.563827201116325e-05, "loss": 0.0424, "step": 36220 }, { "grad_norm": 0.3785344958305359, "learning_rate": 7.562407509589164e-05, "loss": 0.0363, "step": 36230 }, { "grad_norm": 0.33538389205932617, "learning_rate": 7.560987537840973e-05, "loss": 0.0245, "step": 36240 }, { "grad_norm": 0.39738672971725464, "learning_rate": 7.559567286027036e-05, "loss": 0.0322, "step": 36250 }, { "grad_norm": 0.26799479126930237, "learning_rate": 7.558146754302668e-05, "loss": 0.0364, "step": 36260 }, { "grad_norm": 0.5101438760757446, "learning_rate": 7.556725942823217e-05, "loss": 0.0484, "step": 36270 }, { "grad_norm": 0.43125468492507935, "learning_rate": 7.555304851744061e-05, "loss": 0.0369, "step": 36280 }, { "grad_norm": 0.5013206601142883, "learning_rate": 7.553883481220608e-05, "loss": 0.0458, "step": 36290 }, { "grad_norm": 0.4368630647659302, "learning_rate": 7.552461831408298e-05, "loss": 0.0342, "step": 36300 }, { "grad_norm": 0.6029016375541687, "learning_rate": 7.551039902462599e-05, "loss": 0.0328, "step": 36310 }, { "grad_norm": 0.3916877508163452, "learning_rate": 7.549617694539011e-05, "loss": 0.0331, "step": 36320 }, { "grad_norm": 0.5272091031074524, "learning_rate": 7.548195207793066e-05, "loss": 0.0385, "step": 36330 }, { "grad_norm": 0.30479875206947327, "learning_rate": 7.546772442380323e-05, "loss": 0.0345, "step": 36340 }, { "grad_norm": 1.3094403743743896, "learning_rate": 7.545349398456376e-05, "loss": 0.0376, "step": 36350 }, { "grad_norm": 0.4504629373550415, "learning_rate": 7.543926076176845e-05, "loss": 0.0361, "step": 36360 }, { "grad_norm": 0.6727381348609924, "learning_rate": 7.542502475697385e-05, "loss": 0.0425, "step": 36370 }, { "grad_norm": 0.4495624005794525, "learning_rate": 7.541078597173675e-05, "loss": 0.0409, "step": 36380 }, { "grad_norm": 0.3283604681491852, "learning_rate": 7.539654440761431e-05, "loss": 0.0333, "step": 36390 }, { "grad_norm": 0.35669800639152527, "learning_rate": 7.538230006616395e-05, "loss": 0.0325, "step": 36400 }, { "grad_norm": 0.5584063529968262, "learning_rate": 7.536805294894342e-05, "loss": 0.0407, "step": 36410 }, { "grad_norm": 0.4010809063911438, "learning_rate": 7.535380305751076e-05, "loss": 0.0354, "step": 36420 }, { "grad_norm": 0.4412437081336975, "learning_rate": 7.533955039342431e-05, "loss": 0.0391, "step": 36430 }, { "grad_norm": 0.26737064123153687, "learning_rate": 7.532529495824274e-05, "loss": 0.0474, "step": 36440 }, { "grad_norm": 1.1855440139770508, "learning_rate": 7.531103675352497e-05, "loss": 0.0342, "step": 36450 }, { "grad_norm": 1.0021191835403442, "learning_rate": 7.529677578083029e-05, "loss": 0.0407, "step": 36460 }, { "grad_norm": 0.8436310887336731, "learning_rate": 7.528251204171823e-05, "loss": 0.0408, "step": 36470 }, { "grad_norm": 0.348018616437912, "learning_rate": 7.526824553774867e-05, "loss": 0.0445, "step": 36480 }, { "grad_norm": 0.30384618043899536, "learning_rate": 7.525397627048178e-05, "loss": 0.0434, "step": 36490 }, { "grad_norm": 0.6037603616714478, "learning_rate": 7.523970424147802e-05, "loss": 0.0354, "step": 36500 }, { "grad_norm": 0.5054721236228943, "learning_rate": 7.522542945229813e-05, "loss": 0.0438, "step": 36510 }, { "grad_norm": 0.4363487660884857, "learning_rate": 7.521115190450321e-05, "loss": 0.0285, "step": 36520 }, { "grad_norm": 0.5126101970672607, "learning_rate": 7.519687159965462e-05, "loss": 0.0338, "step": 36530 }, { "grad_norm": 0.2862281799316406, "learning_rate": 7.518258853931403e-05, "loss": 0.0322, "step": 36540 }, { "grad_norm": 0.3947157561779022, "learning_rate": 7.516830272504342e-05, "loss": 0.0323, "step": 36550 }, { "grad_norm": 0.6141855716705322, "learning_rate": 7.515401415840509e-05, "loss": 0.0387, "step": 36560 }, { "grad_norm": 0.48457103967666626, "learning_rate": 7.513972284096155e-05, "loss": 0.0332, "step": 36570 }, { "grad_norm": 0.3734654188156128, "learning_rate": 7.512542877427574e-05, "loss": 0.0454, "step": 36580 }, { "grad_norm": 0.7129757404327393, "learning_rate": 7.51111319599108e-05, "loss": 0.0503, "step": 36590 }, { "grad_norm": 0.3130362331867218, "learning_rate": 7.509683239943024e-05, "loss": 0.0417, "step": 36600 }, { "grad_norm": 0.5054207444190979, "learning_rate": 7.50825300943978e-05, "loss": 0.0421, "step": 36610 }, { "grad_norm": 0.35420486330986023, "learning_rate": 7.50682250463776e-05, "loss": 0.0349, "step": 36620 }, { "grad_norm": 0.4003522992134094, "learning_rate": 7.5053917256934e-05, "loss": 0.0468, "step": 36630 }, { "grad_norm": 0.5040212273597717, "learning_rate": 7.503960672763166e-05, "loss": 0.0385, "step": 36640 }, { "grad_norm": 0.3516501486301422, "learning_rate": 7.502529346003559e-05, "loss": 0.034, "step": 36650 }, { "grad_norm": 0.394055038690567, "learning_rate": 7.501097745571106e-05, "loss": 0.0311, "step": 36660 }, { "grad_norm": 0.519455075263977, "learning_rate": 7.499665871622361e-05, "loss": 0.0502, "step": 36670 }, { "grad_norm": 0.4528326988220215, "learning_rate": 7.498233724313917e-05, "loss": 0.0456, "step": 36680 }, { "grad_norm": 0.31232750415802, "learning_rate": 7.496801303802387e-05, "loss": 0.0344, "step": 36690 }, { "grad_norm": 0.39010778069496155, "learning_rate": 7.495368610244423e-05, "loss": 0.0267, "step": 36700 }, { "grad_norm": 0.7881788015365601, "learning_rate": 7.493935643796697e-05, "loss": 0.041, "step": 36710 }, { "grad_norm": 0.4272058606147766, "learning_rate": 7.49250240461592e-05, "loss": 0.0391, "step": 36720 }, { "grad_norm": 0.8232831954956055, "learning_rate": 7.491068892858826e-05, "loss": 0.0371, "step": 36730 }, { "grad_norm": 0.2940135598182678, "learning_rate": 7.489635108682184e-05, "loss": 0.0344, "step": 36740 }, { "grad_norm": 0.49448642134666443, "learning_rate": 7.48820105224279e-05, "loss": 0.0293, "step": 36750 }, { "grad_norm": 0.29342925548553467, "learning_rate": 7.48676672369747e-05, "loss": 0.0356, "step": 36760 }, { "grad_norm": 0.43540894985198975, "learning_rate": 7.485332123203079e-05, "loss": 0.0316, "step": 36770 }, { "grad_norm": 0.7001813054084778, "learning_rate": 7.483897250916504e-05, "loss": 0.0445, "step": 36780 }, { "grad_norm": 0.43263915181159973, "learning_rate": 7.48246210699466e-05, "loss": 0.0395, "step": 36790 }, { "grad_norm": 0.5276368260383606, "learning_rate": 7.481026691594492e-05, "loss": 0.0292, "step": 36800 }, { "grad_norm": 0.4238300323486328, "learning_rate": 7.479591004872975e-05, "loss": 0.0379, "step": 36810 }, { "grad_norm": 0.3517138659954071, "learning_rate": 7.478155046987114e-05, "loss": 0.0511, "step": 36820 }, { "grad_norm": 0.3380123972892761, "learning_rate": 7.476718818093944e-05, "loss": 0.0395, "step": 36830 }, { "grad_norm": 0.36194902658462524, "learning_rate": 7.475282318350527e-05, "loss": 0.0312, "step": 36840 }, { "grad_norm": 0.48230329155921936, "learning_rate": 7.473845547913957e-05, "loss": 0.038, "step": 36850 }, { "grad_norm": 0.37215960025787354, "learning_rate": 7.472408506941357e-05, "loss": 0.0453, "step": 36860 }, { "grad_norm": 1.0599876642227173, "learning_rate": 7.470971195589881e-05, "loss": 0.0414, "step": 36870 }, { "grad_norm": 0.35680627822875977, "learning_rate": 7.46953361401671e-05, "loss": 0.0418, "step": 36880 }, { "grad_norm": 0.4599854052066803, "learning_rate": 7.468095762379055e-05, "loss": 0.0429, "step": 36890 }, { "grad_norm": 0.4599834084510803, "learning_rate": 7.466657640834158e-05, "loss": 0.0433, "step": 36900 }, { "grad_norm": 0.5505711436271667, "learning_rate": 7.46521924953929e-05, "loss": 0.0507, "step": 36910 }, { "grad_norm": 0.3869633972644806, "learning_rate": 7.463780588651752e-05, "loss": 0.036, "step": 36920 }, { "grad_norm": 0.709862470626831, "learning_rate": 7.462341658328873e-05, "loss": 0.0432, "step": 36930 }, { "grad_norm": 0.5190634727478027, "learning_rate": 7.460902458728012e-05, "loss": 0.0438, "step": 36940 }, { "grad_norm": 0.6201058030128479, "learning_rate": 7.459462990006558e-05, "loss": 0.0357, "step": 36950 }, { "grad_norm": 0.4681072235107422, "learning_rate": 7.458023252321929e-05, "loss": 0.0399, "step": 36960 }, { "grad_norm": 0.37484362721443176, "learning_rate": 7.456583245831573e-05, "loss": 0.0326, "step": 36970 }, { "grad_norm": 0.38634106516838074, "learning_rate": 7.455142970692968e-05, "loss": 0.054, "step": 36980 }, { "grad_norm": 0.5684011578559875, "learning_rate": 7.453702427063615e-05, "loss": 0.0418, "step": 36990 }, { "grad_norm": 0.4436700940132141, "learning_rate": 7.452261615101057e-05, "loss": 0.0485, "step": 37000 }, { "grad_norm": 0.4910903573036194, "learning_rate": 7.450820534962856e-05, "loss": 0.0407, "step": 37010 }, { "grad_norm": 1.0589749813079834, "learning_rate": 7.449379186806604e-05, "loss": 0.0416, "step": 37020 }, { "grad_norm": 0.375535786151886, "learning_rate": 7.44793757078993e-05, "loss": 0.0361, "step": 37030 }, { "grad_norm": 0.37704145908355713, "learning_rate": 7.44649568707048e-05, "loss": 0.0413, "step": 37040 }, { "grad_norm": 0.4155043661594391, "learning_rate": 7.445053535805942e-05, "loss": 0.0525, "step": 37050 }, { "grad_norm": 0.37972700595855713, "learning_rate": 7.443611117154024e-05, "loss": 0.0405, "step": 37060 }, { "grad_norm": 0.8024187088012695, "learning_rate": 7.442168431272469e-05, "loss": 0.0401, "step": 37070 }, { "grad_norm": 0.2878971993923187, "learning_rate": 7.440725478319044e-05, "loss": 0.0293, "step": 37080 }, { "grad_norm": 0.38682985305786133, "learning_rate": 7.439282258451553e-05, "loss": 0.037, "step": 37090 }, { "grad_norm": 0.41808438301086426, "learning_rate": 7.43783877182782e-05, "loss": 0.0343, "step": 37100 }, { "grad_norm": 1.360660433769226, "learning_rate": 7.436395018605707e-05, "loss": 0.0359, "step": 37110 }, { "grad_norm": 0.3594885468482971, "learning_rate": 7.434950998943094e-05, "loss": 0.0375, "step": 37120 }, { "grad_norm": 0.3262229263782501, "learning_rate": 7.433506712997902e-05, "loss": 0.0415, "step": 37130 }, { "grad_norm": 0.3223366141319275, "learning_rate": 7.432062160928077e-05, "loss": 0.0425, "step": 37140 }, { "grad_norm": 0.42241373658180237, "learning_rate": 7.430617342891588e-05, "loss": 0.0294, "step": 37150 }, { "grad_norm": 0.308679461479187, "learning_rate": 7.429172259046442e-05, "loss": 0.0357, "step": 37160 }, { "grad_norm": 0.31039461493492126, "learning_rate": 7.42772690955067e-05, "loss": 0.0325, "step": 37170 }, { "grad_norm": 0.5082996487617493, "learning_rate": 7.426281294562332e-05, "loss": 0.0452, "step": 37180 }, { "grad_norm": 0.38240304589271545, "learning_rate": 7.424835414239519e-05, "loss": 0.0387, "step": 37190 }, { "grad_norm": 1.1042497158050537, "learning_rate": 7.423389268740352e-05, "loss": 0.0435, "step": 37200 }, { "grad_norm": 0.5302528738975525, "learning_rate": 7.42194285822298e-05, "loss": 0.0441, "step": 37210 }, { "grad_norm": 0.45132526755332947, "learning_rate": 7.420496182845576e-05, "loss": 0.0405, "step": 37220 }, { "grad_norm": 0.3944547474384308, "learning_rate": 7.419049242766348e-05, "loss": 0.0406, "step": 37230 }, { "grad_norm": 0.8200123310089111, "learning_rate": 7.417602038143533e-05, "loss": 0.0369, "step": 37240 }, { "grad_norm": 1.3682395219802856, "learning_rate": 7.416154569135393e-05, "loss": 0.0399, "step": 37250 }, { "grad_norm": 0.7201208472251892, "learning_rate": 7.414706835900222e-05, "loss": 0.0396, "step": 37260 }, { "grad_norm": 0.3162643313407898, "learning_rate": 7.413258838596345e-05, "loss": 0.0328, "step": 37270 }, { "grad_norm": 0.7926799654960632, "learning_rate": 7.411810577382106e-05, "loss": 0.0422, "step": 37280 }, { "grad_norm": 0.4778272211551666, "learning_rate": 7.410362052415889e-05, "loss": 0.0425, "step": 37290 }, { "grad_norm": 0.6702312231063843, "learning_rate": 7.408913263856102e-05, "loss": 0.0392, "step": 37300 }, { "grad_norm": 0.42305800318717957, "learning_rate": 7.407464211861183e-05, "loss": 0.0459, "step": 37310 }, { "grad_norm": 0.390199214220047, "learning_rate": 7.406014896589597e-05, "loss": 0.0482, "step": 37320 }, { "grad_norm": 0.4163516163825989, "learning_rate": 7.404565318199838e-05, "loss": 0.0373, "step": 37330 }, { "grad_norm": 0.36840829253196716, "learning_rate": 7.403115476850432e-05, "loss": 0.045, "step": 37340 }, { "grad_norm": 0.3285016417503357, "learning_rate": 7.401665372699932e-05, "loss": 0.0424, "step": 37350 }, { "grad_norm": 0.4095945656299591, "learning_rate": 7.400215005906916e-05, "loss": 0.0408, "step": 37360 }, { "grad_norm": 0.4097987413406372, "learning_rate": 7.398764376629996e-05, "loss": 0.0358, "step": 37370 }, { "grad_norm": 0.3204151690006256, "learning_rate": 7.39731348502781e-05, "loss": 0.0334, "step": 37380 }, { "grad_norm": 0.6203372478485107, "learning_rate": 7.395862331259028e-05, "loss": 0.0363, "step": 37390 }, { "grad_norm": 0.2855147123336792, "learning_rate": 7.394410915482341e-05, "loss": 0.0382, "step": 37400 }, { "grad_norm": 0.4837598502635956, "learning_rate": 7.392959237856476e-05, "loss": 0.0426, "step": 37410 }, { "grad_norm": 0.6289206147193909, "learning_rate": 7.391507298540187e-05, "loss": 0.0364, "step": 37420 }, { "grad_norm": 0.4613901972770691, "learning_rate": 7.390055097692257e-05, "loss": 0.0337, "step": 37430 }, { "grad_norm": 0.39842095971107483, "learning_rate": 7.388602635471494e-05, "loss": 0.0564, "step": 37440 }, { "grad_norm": 0.5318056344985962, "learning_rate": 7.387149912036738e-05, "loss": 0.0437, "step": 37450 }, { "grad_norm": 0.290126234292984, "learning_rate": 7.385696927546854e-05, "loss": 0.0352, "step": 37460 }, { "grad_norm": 0.3098663091659546, "learning_rate": 7.384243682160743e-05, "loss": 0.0493, "step": 37470 }, { "grad_norm": 0.3707784116268158, "learning_rate": 7.382790176037329e-05, "loss": 0.0366, "step": 37480 }, { "grad_norm": 0.37259355187416077, "learning_rate": 7.381336409335562e-05, "loss": 0.0318, "step": 37490 }, { "grad_norm": 0.36998578906059265, "learning_rate": 7.379882382214426e-05, "loss": 0.0404, "step": 37500 }, { "grad_norm": 0.2866964638233185, "learning_rate": 7.378428094832931e-05, "loss": 0.0421, "step": 37510 }, { "grad_norm": 0.48246148228645325, "learning_rate": 7.376973547350114e-05, "loss": 0.0542, "step": 37520 }, { "grad_norm": 0.559559166431427, "learning_rate": 7.375518739925046e-05, "loss": 0.0358, "step": 37530 }, { "grad_norm": 0.3429047167301178, "learning_rate": 7.374063672716818e-05, "loss": 0.0317, "step": 37540 }, { "grad_norm": 0.3080115020275116, "learning_rate": 7.372608345884558e-05, "loss": 0.0381, "step": 37550 }, { "grad_norm": 0.6332041621208191, "learning_rate": 7.371152759587415e-05, "loss": 0.0443, "step": 37560 }, { "grad_norm": 0.2797832190990448, "learning_rate": 7.369696913984572e-05, "loss": 0.0316, "step": 37570 }, { "grad_norm": 0.4222661256790161, "learning_rate": 7.368240809235237e-05, "loss": 0.0311, "step": 37580 }, { "grad_norm": 0.43292325735092163, "learning_rate": 7.366784445498649e-05, "loss": 0.0263, "step": 37590 }, { "grad_norm": 0.43491169810295105, "learning_rate": 7.36532782293407e-05, "loss": 0.0331, "step": 37600 }, { "grad_norm": 0.3777763843536377, "learning_rate": 7.363870941700797e-05, "loss": 0.032, "step": 37610 }, { "grad_norm": 0.5151264071464539, "learning_rate": 7.362413801958152e-05, "loss": 0.0414, "step": 37620 }, { "grad_norm": 0.6014176607131958, "learning_rate": 7.360956403865486e-05, "loss": 0.0425, "step": 37630 }, { "grad_norm": 0.4024471640586853, "learning_rate": 7.359498747582177e-05, "loss": 0.0359, "step": 37640 }, { "grad_norm": 0.28098663687705994, "learning_rate": 7.358040833267634e-05, "loss": 0.0407, "step": 37650 }, { "grad_norm": 0.4171486496925354, "learning_rate": 7.356582661081288e-05, "loss": 0.0441, "step": 37660 }, { "grad_norm": 0.8660756945610046, "learning_rate": 7.355124231182607e-05, "loss": 0.0401, "step": 37670 }, { "grad_norm": 0.47537562251091003, "learning_rate": 7.353665543731083e-05, "loss": 0.0501, "step": 37680 }, { "grad_norm": 0.2695032060146332, "learning_rate": 7.352206598886229e-05, "loss": 0.0396, "step": 37690 }, { "grad_norm": 0.35409852862358093, "learning_rate": 7.350747396807601e-05, "loss": 0.039, "step": 37700 }, { "grad_norm": 0.4524591267108917, "learning_rate": 7.349287937654772e-05, "loss": 0.0322, "step": 37710 }, { "grad_norm": 0.41524413228034973, "learning_rate": 7.347828221587345e-05, "loss": 0.0418, "step": 37720 }, { "grad_norm": 0.4017085134983063, "learning_rate": 7.346368248764955e-05, "loss": 0.0326, "step": 37730 }, { "grad_norm": 0.24911284446716309, "learning_rate": 7.34490801934726e-05, "loss": 0.0417, "step": 37740 }, { "grad_norm": 0.4365708529949188, "learning_rate": 7.343447533493947e-05, "loss": 0.0426, "step": 37750 }, { "grad_norm": 0.42345911264419556, "learning_rate": 7.341986791364739e-05, "loss": 0.054, "step": 37760 }, { "grad_norm": 0.31415092945098877, "learning_rate": 7.340525793119374e-05, "loss": 0.0443, "step": 37770 }, { "grad_norm": 0.5031371116638184, "learning_rate": 7.33906453891763e-05, "loss": 0.0407, "step": 37780 }, { "grad_norm": 0.3302220106124878, "learning_rate": 7.337603028919303e-05, "loss": 0.0387, "step": 37790 }, { "grad_norm": 0.5528837442398071, "learning_rate": 7.336141263284225e-05, "loss": 0.0354, "step": 37800 }, { "grad_norm": 0.2996564507484436, "learning_rate": 7.334679242172249e-05, "loss": 0.0375, "step": 37810 }, { "grad_norm": 0.38368120789527893, "learning_rate": 7.33321696574326e-05, "loss": 0.0408, "step": 37820 }, { "grad_norm": 0.28276169300079346, "learning_rate": 7.331754434157174e-05, "loss": 0.0354, "step": 37830 }, { "grad_norm": 0.38308998942375183, "learning_rate": 7.330291647573928e-05, "loss": 0.0357, "step": 37840 }, { "grad_norm": 0.5015305280685425, "learning_rate": 7.32882860615349e-05, "loss": 0.0361, "step": 37850 }, { "grad_norm": 0.578231930732727, "learning_rate": 7.327365310055858e-05, "loss": 0.0342, "step": 37860 }, { "grad_norm": 0.34371769428253174, "learning_rate": 7.325901759441055e-05, "loss": 0.0287, "step": 37870 }, { "grad_norm": 0.47811177372932434, "learning_rate": 7.324437954469132e-05, "loss": 0.052, "step": 37880 }, { "grad_norm": 0.5777804255485535, "learning_rate": 7.32297389530017e-05, "loss": 0.0439, "step": 37890 }, { "grad_norm": 0.39846959710121155, "learning_rate": 7.321509582094274e-05, "loss": 0.0293, "step": 37900 }, { "grad_norm": 0.27874916791915894, "learning_rate": 7.320045015011582e-05, "loss": 0.0403, "step": 37910 }, { "grad_norm": 0.28523656725883484, "learning_rate": 7.318580194212253e-05, "loss": 0.0276, "step": 37920 }, { "grad_norm": 0.46031713485717773, "learning_rate": 7.317115119856483e-05, "loss": 0.0379, "step": 37930 }, { "grad_norm": 0.418243944644928, "learning_rate": 7.315649792104485e-05, "loss": 0.0354, "step": 37940 }, { "grad_norm": 0.30038508772850037, "learning_rate": 7.314184211116507e-05, "loss": 0.0471, "step": 37950 }, { "grad_norm": 0.45764508843421936, "learning_rate": 7.312718377052823e-05, "loss": 0.0326, "step": 37960 }, { "grad_norm": 0.4336109459400177, "learning_rate": 7.311252290073737e-05, "loss": 0.0406, "step": 37970 }, { "grad_norm": 0.5148743987083435, "learning_rate": 7.309785950339572e-05, "loss": 0.035, "step": 37980 }, { "grad_norm": 0.621707022190094, "learning_rate": 7.308319358010692e-05, "loss": 0.0381, "step": 37990 }, { "grad_norm": 0.32057100534439087, "learning_rate": 7.306852513247474e-05, "loss": 0.0483, "step": 38000 }, { "grad_norm": 0.3497685194015503, "learning_rate": 7.305385416210336e-05, "loss": 0.0467, "step": 38010 }, { "grad_norm": 0.3824908137321472, "learning_rate": 7.303918067059711e-05, "loss": 0.0356, "step": 38020 }, { "grad_norm": 0.5089337229728699, "learning_rate": 7.302450465956075e-05, "loss": 0.0426, "step": 38030 }, { "grad_norm": 0.34597480297088623, "learning_rate": 7.300982613059914e-05, "loss": 0.0296, "step": 38040 }, { "grad_norm": 0.3271228075027466, "learning_rate": 7.299514508531757e-05, "loss": 0.0447, "step": 38050 }, { "grad_norm": 0.8207027316093445, "learning_rate": 7.298046152532148e-05, "loss": 0.0459, "step": 38060 }, { "grad_norm": 0.39293769001960754, "learning_rate": 7.296577545221668e-05, "loss": 0.027, "step": 38070 }, { "grad_norm": 0.5091556906700134, "learning_rate": 7.29510868676092e-05, "loss": 0.0512, "step": 38080 }, { "grad_norm": 0.3075079321861267, "learning_rate": 7.293639577310538e-05, "loss": 0.0355, "step": 38090 }, { "grad_norm": 0.38084548711776733, "learning_rate": 7.292170217031179e-05, "loss": 0.0365, "step": 38100 }, { "grad_norm": 0.5259070992469788, "learning_rate": 7.290700606083532e-05, "loss": 0.0362, "step": 38110 }, { "grad_norm": 0.4775256812572479, "learning_rate": 7.28923074462831e-05, "loss": 0.0383, "step": 38120 }, { "grad_norm": 0.333418607711792, "learning_rate": 7.287760632826255e-05, "loss": 0.0395, "step": 38130 }, { "grad_norm": 0.5956054329872131, "learning_rate": 7.286290270838139e-05, "loss": 0.0541, "step": 38140 }, { "grad_norm": 0.4506150186061859, "learning_rate": 7.284819658824756e-05, "loss": 0.0379, "step": 38150 }, { "grad_norm": 0.37035316228866577, "learning_rate": 7.283348796946929e-05, "loss": 0.0402, "step": 38160 }, { "grad_norm": 0.5370616912841797, "learning_rate": 7.28187768536551e-05, "loss": 0.0369, "step": 38170 }, { "grad_norm": 0.2968353033065796, "learning_rate": 7.280406324241379e-05, "loss": 0.0396, "step": 38180 }, { "grad_norm": 0.34657004475593567, "learning_rate": 7.278934713735441e-05, "loss": 0.0433, "step": 38190 }, { "grad_norm": 0.5524124503135681, "learning_rate": 7.277462854008629e-05, "loss": 0.0351, "step": 38200 }, { "grad_norm": 0.9136991500854492, "learning_rate": 7.275990745221904e-05, "loss": 0.0308, "step": 38210 }, { "grad_norm": 0.2780088782310486, "learning_rate": 7.274518387536252e-05, "loss": 0.0349, "step": 38220 }, { "grad_norm": 0.6230108737945557, "learning_rate": 7.273045781112689e-05, "loss": 0.0492, "step": 38230 }, { "grad_norm": 0.44167348742485046, "learning_rate": 7.271572926112257e-05, "loss": 0.0412, "step": 38240 }, { "grad_norm": 0.3509410619735718, "learning_rate": 7.270099822696024e-05, "loss": 0.0303, "step": 38250 }, { "grad_norm": 0.4648244380950928, "learning_rate": 7.26862647102509e-05, "loss": 0.0342, "step": 38260 }, { "grad_norm": 0.35027116537094116, "learning_rate": 7.267152871260573e-05, "loss": 0.039, "step": 38270 }, { "grad_norm": 0.5548690557479858, "learning_rate": 7.265679023563629e-05, "loss": 0.0404, "step": 38280 }, { "grad_norm": 0.37340331077575684, "learning_rate": 7.264204928095433e-05, "loss": 0.035, "step": 38290 }, { "grad_norm": 0.3003499507904053, "learning_rate": 7.262730585017188e-05, "loss": 0.0445, "step": 38300 }, { "grad_norm": 0.4286699593067169, "learning_rate": 7.261255994490131e-05, "loss": 0.0417, "step": 38310 }, { "grad_norm": 0.282181054353714, "learning_rate": 7.259781156675517e-05, "loss": 0.0396, "step": 38320 }, { "grad_norm": 0.4279148280620575, "learning_rate": 7.258306071734633e-05, "loss": 0.0437, "step": 38330 }, { "grad_norm": 0.32326409220695496, "learning_rate": 7.256830739828792e-05, "loss": 0.0414, "step": 38340 }, { "grad_norm": 0.3017865717411041, "learning_rate": 7.255355161119336e-05, "loss": 0.0303, "step": 38350 }, { "grad_norm": 0.384621262550354, "learning_rate": 7.253879335767628e-05, "loss": 0.0302, "step": 38360 }, { "grad_norm": 0.4232158064842224, "learning_rate": 7.252403263935067e-05, "loss": 0.0357, "step": 38370 }, { "grad_norm": 0.38410183787345886, "learning_rate": 7.25092694578307e-05, "loss": 0.0392, "step": 38380 }, { "grad_norm": 0.3997057378292084, "learning_rate": 7.249450381473087e-05, "loss": 0.0357, "step": 38390 }, { "grad_norm": 0.3434400260448456, "learning_rate": 7.247973571166593e-05, "loss": 0.0464, "step": 38400 }, { "grad_norm": 0.34040430188179016, "learning_rate": 7.246496515025089e-05, "loss": 0.0374, "step": 38410 }, { "grad_norm": 0.6994849443435669, "learning_rate": 7.245019213210101e-05, "loss": 0.0389, "step": 38420 }, { "grad_norm": 0.39717090129852295, "learning_rate": 7.24354166588319e-05, "loss": 0.0386, "step": 38430 }, { "grad_norm": 0.5063236951828003, "learning_rate": 7.242063873205935e-05, "loss": 0.0508, "step": 38440 }, { "grad_norm": 0.3839770257472992, "learning_rate": 7.240585835339946e-05, "loss": 0.037, "step": 38450 }, { "grad_norm": 0.40211403369903564, "learning_rate": 7.239107552446857e-05, "loss": 0.039, "step": 38460 }, { "grad_norm": 0.39493465423583984, "learning_rate": 7.237629024688334e-05, "loss": 0.034, "step": 38470 }, { "grad_norm": 0.2901324927806854, "learning_rate": 7.236150252226064e-05, "loss": 0.0351, "step": 38480 }, { "grad_norm": 0.7914524078369141, "learning_rate": 7.234671235221765e-05, "loss": 0.0627, "step": 38490 }, { "grad_norm": 0.3923160135746002, "learning_rate": 7.233191973837179e-05, "loss": 0.0382, "step": 38500 }, { "grad_norm": 0.4267316460609436, "learning_rate": 7.231712468234078e-05, "loss": 0.0334, "step": 38510 }, { "grad_norm": 0.3661896288394928, "learning_rate": 7.230232718574254e-05, "loss": 0.0498, "step": 38520 }, { "grad_norm": 0.5182576775550842, "learning_rate": 7.228752725019535e-05, "loss": 0.0394, "step": 38530 }, { "grad_norm": 0.42716795206069946, "learning_rate": 7.227272487731769e-05, "loss": 0.0385, "step": 38540 }, { "grad_norm": 0.4651821255683899, "learning_rate": 7.225792006872831e-05, "loss": 0.0355, "step": 38550 }, { "grad_norm": 0.6335281729698181, "learning_rate": 7.224311282604628e-05, "loss": 0.0406, "step": 38560 }, { "grad_norm": 1.5460994243621826, "learning_rate": 7.222830315089085e-05, "loss": 0.043, "step": 38570 }, { "grad_norm": 0.385213166475296, "learning_rate": 7.22134910448816e-05, "loss": 0.0381, "step": 38580 }, { "grad_norm": 0.4568043053150177, "learning_rate": 7.219867650963839e-05, "loss": 0.0477, "step": 38590 }, { "grad_norm": 0.6809835433959961, "learning_rate": 7.21838595467813e-05, "loss": 0.0477, "step": 38600 }, { "grad_norm": 0.3615412414073944, "learning_rate": 7.216904015793067e-05, "loss": 0.0365, "step": 38610 }, { "grad_norm": 0.36423495411872864, "learning_rate": 7.215421834470713e-05, "loss": 0.0365, "step": 38620 }, { "grad_norm": 0.4782085716724396, "learning_rate": 7.213939410873158e-05, "loss": 0.0434, "step": 38630 }, { "grad_norm": 0.386661171913147, "learning_rate": 7.212456745162519e-05, "loss": 0.029, "step": 38640 }, { "grad_norm": 0.45436424016952515, "learning_rate": 7.210973837500937e-05, "loss": 0.0386, "step": 38650 }, { "grad_norm": 0.23588375747203827, "learning_rate": 7.209490688050578e-05, "loss": 0.0351, "step": 38660 }, { "grad_norm": 0.4407164752483368, "learning_rate": 7.208007296973641e-05, "loss": 0.0345, "step": 38670 }, { "grad_norm": 1.6857028007507324, "learning_rate": 7.206523664432345e-05, "loss": 0.0579, "step": 38680 }, { "grad_norm": 0.44041961431503296, "learning_rate": 7.205039790588939e-05, "loss": 0.0273, "step": 38690 }, { "grad_norm": 0.4176045358181, "learning_rate": 7.203555675605697e-05, "loss": 0.0399, "step": 38700 }, { "grad_norm": 0.526624858379364, "learning_rate": 7.202071319644917e-05, "loss": 0.0479, "step": 38710 }, { "grad_norm": 0.4528214633464813, "learning_rate": 7.200586722868932e-05, "loss": 0.0398, "step": 38720 }, { "grad_norm": 0.5917013883590698, "learning_rate": 7.199101885440088e-05, "loss": 0.0326, "step": 38730 }, { "grad_norm": 0.26869121193885803, "learning_rate": 7.197616807520771e-05, "loss": 0.037, "step": 38740 }, { "grad_norm": 0.32581087946891785, "learning_rate": 7.196131489273381e-05, "loss": 0.0285, "step": 38750 }, { "grad_norm": 0.6504679918289185, "learning_rate": 7.194645930860355e-05, "loss": 0.0367, "step": 38760 }, { "grad_norm": 0.5121772289276123, "learning_rate": 7.19316013244415e-05, "loss": 0.0436, "step": 38770 }, { "grad_norm": 0.4803525507450104, "learning_rate": 7.191674094187248e-05, "loss": 0.0316, "step": 38780 }, { "grad_norm": 0.4237562417984009, "learning_rate": 7.190187816252165e-05, "loss": 0.0351, "step": 38790 }, { "grad_norm": 0.40110543370246887, "learning_rate": 7.188701298801435e-05, "loss": 0.0302, "step": 38800 }, { "grad_norm": 0.44631385803222656, "learning_rate": 7.18721454199762e-05, "loss": 0.039, "step": 38810 }, { "grad_norm": 1.2334145307540894, "learning_rate": 7.185727546003312e-05, "loss": 0.0395, "step": 38820 }, { "grad_norm": 0.6580168604850769, "learning_rate": 7.184240310981126e-05, "loss": 0.0462, "step": 38830 }, { "grad_norm": 0.42204785346984863, "learning_rate": 7.182752837093704e-05, "loss": 0.0363, "step": 38840 }, { "grad_norm": 0.583182692527771, "learning_rate": 7.181265124503711e-05, "loss": 0.0372, "step": 38850 }, { "grad_norm": 0.43733707070350647, "learning_rate": 7.179777173373847e-05, "loss": 0.0437, "step": 38860 }, { "grad_norm": 0.3774440288543701, "learning_rate": 7.178288983866826e-05, "loss": 0.0356, "step": 38870 }, { "grad_norm": 0.47780242562294006, "learning_rate": 7.176800556145397e-05, "loss": 0.0456, "step": 38880 }, { "grad_norm": 0.2601991891860962, "learning_rate": 7.175311890372334e-05, "loss": 0.0388, "step": 38890 }, { "grad_norm": 0.5105530023574829, "learning_rate": 7.17382298671043e-05, "loss": 0.0348, "step": 38900 }, { "grad_norm": 0.6389373540878296, "learning_rate": 7.172333845322515e-05, "loss": 0.0401, "step": 38910 }, { "grad_norm": 0.5470996499061584, "learning_rate": 7.170844466371436e-05, "loss": 0.0426, "step": 38920 }, { "grad_norm": 0.37701305747032166, "learning_rate": 7.16935485002007e-05, "loss": 0.0339, "step": 38930 }, { "grad_norm": 0.5104193091392517, "learning_rate": 7.167864996431319e-05, "loss": 0.0405, "step": 38940 }, { "grad_norm": 0.355042427778244, "learning_rate": 7.166374905768111e-05, "loss": 0.0331, "step": 38950 }, { "grad_norm": 1.4268176555633545, "learning_rate": 7.164884578193401e-05, "loss": 0.0329, "step": 38960 }, { "grad_norm": 0.8803805708885193, "learning_rate": 7.163394013870168e-05, "loss": 0.038, "step": 38970 }, { "grad_norm": 0.43810972571372986, "learning_rate": 7.161903212961419e-05, "loss": 0.0435, "step": 38980 }, { "grad_norm": 1.4156614542007446, "learning_rate": 7.160412175630183e-05, "loss": 0.0372, "step": 38990 }, { "grad_norm": 0.5517479181289673, "learning_rate": 7.158920902039521e-05, "loss": 0.0349, "step": 39000 }, { "grad_norm": 0.2727552354335785, "learning_rate": 7.157429392352514e-05, "loss": 0.0297, "step": 39010 }, { "grad_norm": 0.7129851579666138, "learning_rate": 7.155937646732275e-05, "loss": 0.0339, "step": 39020 }, { "grad_norm": 2.7432150840759277, "learning_rate": 7.154445665341933e-05, "loss": 0.0401, "step": 39030 }, { "grad_norm": 0.47429823875427246, "learning_rate": 7.152953448344654e-05, "loss": 0.0326, "step": 39040 }, { "grad_norm": 0.47989118099212646, "learning_rate": 7.151460995903624e-05, "loss": 0.0362, "step": 39050 }, { "grad_norm": 0.4301607012748718, "learning_rate": 7.149968308182052e-05, "loss": 0.0263, "step": 39060 }, { "grad_norm": 0.9813914895057678, "learning_rate": 7.14847538534318e-05, "loss": 0.0435, "step": 39070 }, { "grad_norm": 0.37249693274497986, "learning_rate": 7.14698222755027e-05, "loss": 0.0353, "step": 39080 }, { "grad_norm": 0.4277299642562866, "learning_rate": 7.14548883496661e-05, "loss": 0.0329, "step": 39090 }, { "grad_norm": 0.3646800220012665, "learning_rate": 7.143995207755517e-05, "loss": 0.0373, "step": 39100 }, { "grad_norm": 0.2587986886501312, "learning_rate": 7.142501346080333e-05, "loss": 0.0368, "step": 39110 }, { "grad_norm": 0.3339112102985382, "learning_rate": 7.141007250104421e-05, "loss": 0.0326, "step": 39120 }, { "grad_norm": 0.5946472883224487, "learning_rate": 7.139512919991176e-05, "loss": 0.0346, "step": 39130 }, { "grad_norm": 0.6295541524887085, "learning_rate": 7.138018355904014e-05, "loss": 0.0311, "step": 39140 }, { "grad_norm": 0.4541046917438507, "learning_rate": 7.13652355800638e-05, "loss": 0.0501, "step": 39150 }, { "grad_norm": 0.40214836597442627, "learning_rate": 7.135028526461743e-05, "loss": 0.0333, "step": 39160 }, { "grad_norm": 0.6346287727355957, "learning_rate": 7.133533261433594e-05, "loss": 0.0383, "step": 39170 }, { "grad_norm": 0.7918782234191895, "learning_rate": 7.132037763085457e-05, "loss": 0.0434, "step": 39180 }, { "grad_norm": 0.3601260781288147, "learning_rate": 7.130542031580875e-05, "loss": 0.0304, "step": 39190 }, { "grad_norm": 0.4629170596599579, "learning_rate": 7.12904606708342e-05, "loss": 0.0381, "step": 39200 }, { "grad_norm": 0.40740805864334106, "learning_rate": 7.127549869756687e-05, "loss": 0.0403, "step": 39210 }, { "grad_norm": 0.4791560769081116, "learning_rate": 7.126053439764299e-05, "loss": 0.0314, "step": 39220 }, { "grad_norm": 0.4456891715526581, "learning_rate": 7.124556777269904e-05, "loss": 0.0391, "step": 39230 }, { "grad_norm": 0.23568330705165863, "learning_rate": 7.123059882437174e-05, "loss": 0.0411, "step": 39240 }, { "grad_norm": 0.5383039116859436, "learning_rate": 7.121562755429807e-05, "loss": 0.0231, "step": 39250 }, { "grad_norm": 0.28707975149154663, "learning_rate": 7.120065396411528e-05, "loss": 0.0266, "step": 39260 }, { "grad_norm": 0.19947750866413116, "learning_rate": 7.118567805546084e-05, "loss": 0.0268, "step": 39270 }, { "grad_norm": 0.4814129173755646, "learning_rate": 7.117069982997248e-05, "loss": 0.0358, "step": 39280 }, { "grad_norm": 0.45651698112487793, "learning_rate": 7.115571928928825e-05, "loss": 0.047, "step": 39290 }, { "grad_norm": 0.4129495322704315, "learning_rate": 7.114073643504635e-05, "loss": 0.0323, "step": 39300 }, { "grad_norm": 0.3620850443840027, "learning_rate": 7.11257512688853e-05, "loss": 0.032, "step": 39310 }, { "grad_norm": 0.3355359137058258, "learning_rate": 7.111076379244384e-05, "loss": 0.0357, "step": 39320 }, { "grad_norm": 0.39237266778945923, "learning_rate": 7.109577400736101e-05, "loss": 0.0364, "step": 39330 }, { "grad_norm": 0.27185696363449097, "learning_rate": 7.108078191527605e-05, "loss": 0.0341, "step": 39340 }, { "grad_norm": 0.5348878502845764, "learning_rate": 7.106578751782847e-05, "loss": 0.0415, "step": 39350 }, { "grad_norm": 0.36867162585258484, "learning_rate": 7.105079081665803e-05, "loss": 0.0367, "step": 39360 }, { "grad_norm": 0.5039623379707336, "learning_rate": 7.103579181340476e-05, "loss": 0.0371, "step": 39370 }, { "grad_norm": 0.8449463844299316, "learning_rate": 7.102079050970893e-05, "loss": 0.0376, "step": 39380 }, { "grad_norm": 0.35176488757133484, "learning_rate": 7.100578690721105e-05, "loss": 0.0516, "step": 39390 }, { "grad_norm": 0.3498873710632324, "learning_rate": 7.09907810075519e-05, "loss": 0.0396, "step": 39400 }, { "grad_norm": 0.4386308193206787, "learning_rate": 7.097577281237249e-05, "loss": 0.0438, "step": 39410 }, { "grad_norm": 1.5931715965270996, "learning_rate": 7.09607623233141e-05, "loss": 0.04, "step": 39420 }, { "grad_norm": 0.6121813654899597, "learning_rate": 7.094574954201827e-05, "loss": 0.038, "step": 39430 }, { "grad_norm": 0.3552328050136566, "learning_rate": 7.093073447012675e-05, "loss": 0.0413, "step": 39440 }, { "grad_norm": 0.4693823456764221, "learning_rate": 7.09157171092816e-05, "loss": 0.0306, "step": 39450 }, { "grad_norm": 0.3099629282951355, "learning_rate": 7.090069746112504e-05, "loss": 0.0348, "step": 39460 }, { "grad_norm": 0.47229447960853577, "learning_rate": 7.088567552729965e-05, "loss": 0.0367, "step": 39470 }, { "grad_norm": 0.263786643743515, "learning_rate": 7.087065130944818e-05, "loss": 0.0472, "step": 39480 }, { "grad_norm": 0.4552741050720215, "learning_rate": 7.085562480921366e-05, "loss": 0.0382, "step": 39490 }, { "grad_norm": 0.9524546265602112, "learning_rate": 7.084059602823937e-05, "loss": 0.038, "step": 39500 }, { "grad_norm": 0.37267929315567017, "learning_rate": 7.082556496816882e-05, "loss": 0.0334, "step": 39510 }, { "grad_norm": 0.4213610887527466, "learning_rate": 7.081053163064582e-05, "loss": 0.0393, "step": 39520 }, { "grad_norm": 0.4861787259578705, "learning_rate": 7.079549601731434e-05, "loss": 0.0423, "step": 39530 }, { "grad_norm": 0.288184255361557, "learning_rate": 7.07804581298187e-05, "loss": 0.0267, "step": 39540 }, { "grad_norm": 0.3739631175994873, "learning_rate": 7.07654179698034e-05, "loss": 0.0493, "step": 39550 }, { "grad_norm": 0.42202436923980713, "learning_rate": 7.075037553891321e-05, "loss": 0.0462, "step": 39560 }, { "grad_norm": 0.7007333636283875, "learning_rate": 7.073533083879315e-05, "loss": 0.0399, "step": 39570 }, { "grad_norm": 0.5188355445861816, "learning_rate": 7.072028387108849e-05, "loss": 0.0397, "step": 39580 }, { "grad_norm": 0.5098081827163696, "learning_rate": 7.070523463744472e-05, "loss": 0.0463, "step": 39590 }, { "grad_norm": 0.3654913604259491, "learning_rate": 7.069018313950763e-05, "loss": 0.0329, "step": 39600 }, { "grad_norm": 0.3900180160999298, "learning_rate": 7.067512937892322e-05, "loss": 0.0294, "step": 39610 }, { "grad_norm": 0.5910829305648804, "learning_rate": 7.066007335733775e-05, "loss": 0.0324, "step": 39620 }, { "grad_norm": 0.38274866342544556, "learning_rate": 7.064501507639772e-05, "loss": 0.0329, "step": 39630 }, { "grad_norm": 1.9702190160751343, "learning_rate": 7.062995453774987e-05, "loss": 0.0583, "step": 39640 }, { "grad_norm": 0.6181513667106628, "learning_rate": 7.061489174304121e-05, "loss": 0.0556, "step": 39650 }, { "grad_norm": 0.24239329993724823, "learning_rate": 7.0599826693919e-05, "loss": 0.0346, "step": 39660 }, { "grad_norm": 0.7871586680412292, "learning_rate": 7.05847593920307e-05, "loss": 0.0414, "step": 39670 }, { "grad_norm": 0.6239188313484192, "learning_rate": 7.056968983902406e-05, "loss": 0.04, "step": 39680 }, { "grad_norm": 0.34933701157569885, "learning_rate": 7.055461803654706e-05, "loss": 0.0415, "step": 39690 }, { "grad_norm": 0.42508000135421753, "learning_rate": 7.053954398624794e-05, "loss": 0.0342, "step": 39700 }, { "grad_norm": 0.437115877866745, "learning_rate": 7.052446768977518e-05, "loss": 0.049, "step": 39710 }, { "grad_norm": 0.3633319139480591, "learning_rate": 7.050938914877748e-05, "loss": 0.0332, "step": 39720 }, { "grad_norm": 0.6255608201026917, "learning_rate": 7.049430836490382e-05, "loss": 0.0607, "step": 39730 }, { "grad_norm": 0.4528372883796692, "learning_rate": 7.047922533980342e-05, "loss": 0.0305, "step": 39740 }, { "grad_norm": 0.5460274815559387, "learning_rate": 7.046414007512571e-05, "loss": 0.0478, "step": 39750 }, { "grad_norm": 0.29443737864494324, "learning_rate": 7.044905257252042e-05, "loss": 0.0434, "step": 39760 }, { "grad_norm": 0.3049499988555908, "learning_rate": 7.04339628336375e-05, "loss": 0.0364, "step": 39770 }, { "grad_norm": 0.5774469971656799, "learning_rate": 7.041887086012711e-05, "loss": 0.0415, "step": 39780 }, { "grad_norm": 0.5198842883110046, "learning_rate": 7.040377665363969e-05, "loss": 0.0331, "step": 39790 }, { "grad_norm": 1.0593348741531372, "learning_rate": 7.038868021582594e-05, "loss": 0.0468, "step": 39800 }, { "grad_norm": 0.32777124643325806, "learning_rate": 7.037358154833679e-05, "loss": 0.0363, "step": 39810 }, { "grad_norm": 0.4120958745479584, "learning_rate": 7.035848065282339e-05, "loss": 0.0333, "step": 39820 }, { "grad_norm": 0.47988995909690857, "learning_rate": 7.034337753093714e-05, "loss": 0.034, "step": 39830 }, { "grad_norm": 0.25509124994277954, "learning_rate": 7.032827218432972e-05, "loss": 0.0404, "step": 39840 }, { "grad_norm": 0.4999595582485199, "learning_rate": 7.031316461465302e-05, "loss": 0.0446, "step": 39850 }, { "grad_norm": 0.5504226088523865, "learning_rate": 7.029805482355915e-05, "loss": 0.0326, "step": 39860 }, { "grad_norm": 0.2943902313709259, "learning_rate": 7.028294281270055e-05, "loss": 0.032, "step": 39870 }, { "grad_norm": 0.794964075088501, "learning_rate": 7.026782858372982e-05, "loss": 0.0417, "step": 39880 }, { "grad_norm": 0.9559164643287659, "learning_rate": 7.025271213829982e-05, "loss": 0.0332, "step": 39890 }, { "grad_norm": 0.39904269576072693, "learning_rate": 7.023759347806366e-05, "loss": 0.03, "step": 39900 }, { "grad_norm": 0.4118219316005707, "learning_rate": 7.022247260467472e-05, "loss": 0.0378, "step": 39910 }, { "grad_norm": 0.4729689657688141, "learning_rate": 7.020734951978658e-05, "loss": 0.0329, "step": 39920 }, { "grad_norm": 0.7594442367553711, "learning_rate": 7.019222422505307e-05, "loss": 0.0387, "step": 39930 }, { "grad_norm": 0.4152415990829468, "learning_rate": 7.017709672212827e-05, "loss": 0.0339, "step": 39940 }, { "grad_norm": 0.3989831209182739, "learning_rate": 7.016196701266652e-05, "loss": 0.0317, "step": 39950 }, { "grad_norm": 0.4104349911212921, "learning_rate": 7.014683509832238e-05, "loss": 0.0338, "step": 39960 }, { "grad_norm": 0.5389131307601929, "learning_rate": 7.013170098075063e-05, "loss": 0.0437, "step": 39970 }, { "grad_norm": 1.123255729675293, "learning_rate": 7.011656466160632e-05, "loss": 0.0462, "step": 39980 }, { "grad_norm": 0.3440265357494354, "learning_rate": 7.010142614254475e-05, "loss": 0.031, "step": 39990 }, { "grad_norm": 0.42494121193885803, "learning_rate": 7.008628542522147e-05, "loss": 0.0375, "step": 40000 }, { "grad_norm": 0.49281954765319824, "learning_rate": 7.007114251129218e-05, "loss": 0.0344, "step": 40010 }, { "grad_norm": 0.6798679232597351, "learning_rate": 7.005599740241293e-05, "loss": 0.0345, "step": 40020 }, { "grad_norm": 0.5010725259780884, "learning_rate": 7.004085010023996e-05, "loss": 0.0405, "step": 40030 }, { "grad_norm": 0.3912901282310486, "learning_rate": 7.002570060642976e-05, "loss": 0.0364, "step": 40040 }, { "grad_norm": 0.6178283095359802, "learning_rate": 7.001054892263903e-05, "loss": 0.0389, "step": 40050 }, { "grad_norm": 0.39458340406417847, "learning_rate": 6.999539505052477e-05, "loss": 0.0388, "step": 40060 }, { "grad_norm": 0.33163782954216003, "learning_rate": 6.998023899174418e-05, "loss": 0.0394, "step": 40070 }, { "grad_norm": 0.3299489915370941, "learning_rate": 6.996508074795467e-05, "loss": 0.0368, "step": 40080 }, { "grad_norm": 0.4369674026966095, "learning_rate": 6.994992032081396e-05, "loss": 0.0449, "step": 40090 }, { "grad_norm": 0.40856534242630005, "learning_rate": 6.993475771197995e-05, "loss": 0.0282, "step": 40100 }, { "grad_norm": 0.24447950720787048, "learning_rate": 6.991959292311082e-05, "loss": 0.0297, "step": 40110 }, { "grad_norm": 0.7184130549430847, "learning_rate": 6.990442595586495e-05, "loss": 0.0389, "step": 40120 }, { "grad_norm": 0.507819652557373, "learning_rate": 6.988925681190098e-05, "loss": 0.0395, "step": 40130 }, { "grad_norm": 0.8190689086914062, "learning_rate": 6.987408549287778e-05, "loss": 0.046, "step": 40140 }, { "grad_norm": 0.6298908591270447, "learning_rate": 6.985891200045449e-05, "loss": 0.0432, "step": 40150 }, { "grad_norm": 0.28257811069488525, "learning_rate": 6.984373633629045e-05, "loss": 0.0372, "step": 40160 }, { "grad_norm": 0.7001854777336121, "learning_rate": 6.98285585020452e-05, "loss": 0.0295, "step": 40170 }, { "grad_norm": 0.32081276178359985, "learning_rate": 6.981337849937864e-05, "loss": 0.0275, "step": 40180 }, { "grad_norm": 0.2827073633670807, "learning_rate": 6.979819632995078e-05, "loss": 0.0371, "step": 40190 }, { "grad_norm": 0.30913984775543213, "learning_rate": 6.978301199542193e-05, "loss": 0.026, "step": 40200 }, { "grad_norm": 0.4682695269584656, "learning_rate": 6.976782549745263e-05, "loss": 0.0302, "step": 40210 }, { "grad_norm": 0.5909544229507446, "learning_rate": 6.975263683770367e-05, "loss": 0.0385, "step": 40220 }, { "grad_norm": 0.4396134614944458, "learning_rate": 6.973744601783604e-05, "loss": 0.0334, "step": 40230 }, { "grad_norm": 0.474434494972229, "learning_rate": 6.972225303951097e-05, "loss": 0.0378, "step": 40240 }, { "grad_norm": 0.6887074112892151, "learning_rate": 6.970705790438998e-05, "loss": 0.0387, "step": 40250 }, { "grad_norm": 0.5885458588600159, "learning_rate": 6.969186061413477e-05, "loss": 0.0312, "step": 40260 }, { "grad_norm": 1.725856065750122, "learning_rate": 6.967666117040727e-05, "loss": 0.0405, "step": 40270 }, { "grad_norm": 0.23891162872314453, "learning_rate": 6.966145957486972e-05, "loss": 0.0332, "step": 40280 }, { "grad_norm": 0.37012913823127747, "learning_rate": 6.964625582918449e-05, "loss": 0.0366, "step": 40290 }, { "grad_norm": 0.388851135969162, "learning_rate": 6.963104993501425e-05, "loss": 0.031, "step": 40300 }, { "grad_norm": 0.6558749675750732, "learning_rate": 6.961584189402192e-05, "loss": 0.0401, "step": 40310 }, { "grad_norm": 0.2984124720096588, "learning_rate": 6.96006317078706e-05, "loss": 0.0384, "step": 40320 }, { "grad_norm": 0.2986827492713928, "learning_rate": 6.958541937822367e-05, "loss": 0.0413, "step": 40330 }, { "grad_norm": 0.31642645597457886, "learning_rate": 6.957020490674473e-05, "loss": 0.0346, "step": 40340 }, { "grad_norm": 0.28204479813575745, "learning_rate": 6.95549882950976e-05, "loss": 0.0378, "step": 40350 }, { "grad_norm": 0.5668260455131531, "learning_rate": 6.953976954494635e-05, "loss": 0.0455, "step": 40360 }, { "grad_norm": 0.331870973110199, "learning_rate": 6.952454865795528e-05, "loss": 0.0344, "step": 40370 }, { "grad_norm": 0.32583752274513245, "learning_rate": 6.950932563578892e-05, "loss": 0.0337, "step": 40380 }, { "grad_norm": 0.2962581515312195, "learning_rate": 6.949410048011206e-05, "loss": 0.0308, "step": 40390 }, { "grad_norm": 0.25240573287010193, "learning_rate": 6.947887319258966e-05, "loss": 0.0398, "step": 40400 }, { "grad_norm": 0.38080933690071106, "learning_rate": 6.946364377488696e-05, "loss": 0.0341, "step": 40410 }, { "grad_norm": 0.3693394064903259, "learning_rate": 6.944841222866947e-05, "loss": 0.0396, "step": 40420 }, { "grad_norm": 0.24908486008644104, "learning_rate": 6.943317855560284e-05, "loss": 0.0406, "step": 40430 }, { "grad_norm": 0.3358311057090759, "learning_rate": 6.941794275735302e-05, "loss": 0.0231, "step": 40440 }, { "grad_norm": 0.34895262122154236, "learning_rate": 6.94027048355862e-05, "loss": 0.0415, "step": 40450 }, { "grad_norm": 0.4150421619415283, "learning_rate": 6.938746479196871e-05, "loss": 0.0371, "step": 40460 }, { "grad_norm": 0.30746063590049744, "learning_rate": 6.937222262816724e-05, "loss": 0.0332, "step": 40470 }, { "grad_norm": 0.6053364872932434, "learning_rate": 6.935697834584865e-05, "loss": 0.0384, "step": 40480 }, { "grad_norm": 0.5379820466041565, "learning_rate": 6.934173194667998e-05, "loss": 0.0355, "step": 40490 }, { "grad_norm": 0.9543560147285461, "learning_rate": 6.93264834323286e-05, "loss": 0.0343, "step": 40500 }, { "grad_norm": 0.3225741386413574, "learning_rate": 6.931123280446206e-05, "loss": 0.0392, "step": 40510 }, { "grad_norm": 0.4472220540046692, "learning_rate": 6.929598006474811e-05, "loss": 0.0429, "step": 40520 }, { "grad_norm": 0.45663461089134216, "learning_rate": 6.92807252148548e-05, "loss": 0.0341, "step": 40530 }, { "grad_norm": 0.39584437012672424, "learning_rate": 6.926546825645037e-05, "loss": 0.0384, "step": 40540 }, { "grad_norm": 0.26102933287620544, "learning_rate": 6.92502091912033e-05, "loss": 0.0485, "step": 40550 }, { "grad_norm": 0.46497392654418945, "learning_rate": 6.923494802078229e-05, "loss": 0.0368, "step": 40560 }, { "grad_norm": 1.0848078727722168, "learning_rate": 6.921968474685628e-05, "loss": 0.039, "step": 40570 }, { "grad_norm": 0.5042684078216553, "learning_rate": 6.920441937109445e-05, "loss": 0.0394, "step": 40580 }, { "grad_norm": 0.5945279002189636, "learning_rate": 6.91891518951662e-05, "loss": 0.0464, "step": 40590 }, { "grad_norm": 0.3717667758464813, "learning_rate": 6.917388232074114e-05, "loss": 0.0462, "step": 40600 }, { "grad_norm": 0.3077390789985657, "learning_rate": 6.915861064948914e-05, "loss": 0.0304, "step": 40610 }, { "grad_norm": 0.4235301613807678, "learning_rate": 6.914333688308025e-05, "loss": 0.0331, "step": 40620 }, { "grad_norm": 0.9476445913314819, "learning_rate": 6.912806102318487e-05, "loss": 0.0265, "step": 40630 }, { "grad_norm": 0.3232414722442627, "learning_rate": 6.911278307147347e-05, "loss": 0.0299, "step": 40640 }, { "grad_norm": 0.27921268343925476, "learning_rate": 6.909750302961684e-05, "loss": 0.0266, "step": 40650 }, { "grad_norm": 0.5552327036857605, "learning_rate": 6.9082220899286e-05, "loss": 0.0477, "step": 40660 }, { "grad_norm": 0.786115288734436, "learning_rate": 6.906693668215219e-05, "loss": 0.0406, "step": 40670 }, { "grad_norm": 0.48528096079826355, "learning_rate": 6.905165037988683e-05, "loss": 0.0267, "step": 40680 }, { "grad_norm": 0.38459405303001404, "learning_rate": 6.903636199416164e-05, "loss": 0.0285, "step": 40690 }, { "grad_norm": 0.652760922908783, "learning_rate": 6.902107152664851e-05, "loss": 0.0346, "step": 40700 }, { "grad_norm": 0.9248937368392944, "learning_rate": 6.90057789790196e-05, "loss": 0.0466, "step": 40710 }, { "grad_norm": 0.44375476241111755, "learning_rate": 6.899048435294728e-05, "loss": 0.0479, "step": 40720 }, { "grad_norm": 0.8670215010643005, "learning_rate": 6.897518765010415e-05, "loss": 0.0388, "step": 40730 }, { "grad_norm": 0.6401398181915283, "learning_rate": 6.895988887216302e-05, "loss": 0.0366, "step": 40740 }, { "grad_norm": 0.5766590237617493, "learning_rate": 6.894458802079694e-05, "loss": 0.041, "step": 40750 }, { "grad_norm": 0.4371606111526489, "learning_rate": 6.892928509767922e-05, "loss": 0.032, "step": 40760 }, { "grad_norm": 0.6140408515930176, "learning_rate": 6.891398010448333e-05, "loss": 0.0348, "step": 40770 }, { "grad_norm": 0.632433295249939, "learning_rate": 6.8898673042883e-05, "loss": 0.0484, "step": 40780 }, { "grad_norm": 0.3676915466785431, "learning_rate": 6.888336391455222e-05, "loss": 0.0494, "step": 40790 }, { "grad_norm": 0.31532883644104004, "learning_rate": 6.886805272116513e-05, "loss": 0.0334, "step": 40800 }, { "grad_norm": 0.5004999041557312, "learning_rate": 6.885273946439617e-05, "loss": 0.0673, "step": 40810 }, { "grad_norm": 0.4511488676071167, "learning_rate": 6.883742414591998e-05, "loss": 0.0388, "step": 40820 }, { "grad_norm": 0.5046399831771851, "learning_rate": 6.88221067674114e-05, "loss": 0.0309, "step": 40830 }, { "grad_norm": 0.4538247585296631, "learning_rate": 6.88067873305455e-05, "loss": 0.0376, "step": 40840 }, { "grad_norm": 0.6423531174659729, "learning_rate": 6.879146583699765e-05, "loss": 0.0492, "step": 40850 }, { "grad_norm": 0.33164191246032715, "learning_rate": 6.877614228844334e-05, "loss": 0.034, "step": 40860 }, { "grad_norm": 0.2613283693790436, "learning_rate": 6.876081668655832e-05, "loss": 0.0372, "step": 40870 }, { "grad_norm": 0.4545220732688904, "learning_rate": 6.874548903301861e-05, "loss": 0.0387, "step": 40880 }, { "grad_norm": 0.4571930766105652, "learning_rate": 6.873015932950039e-05, "loss": 0.0443, "step": 40890 }, { "grad_norm": 0.3452457785606384, "learning_rate": 6.871482757768012e-05, "loss": 0.0352, "step": 40900 }, { "grad_norm": 0.8571106195449829, "learning_rate": 6.869949377923442e-05, "loss": 0.0403, "step": 40910 }, { "grad_norm": 0.33967840671539307, "learning_rate": 6.868415793584022e-05, "loss": 0.0269, "step": 40920 }, { "grad_norm": 0.2919897437095642, "learning_rate": 6.86688200491746e-05, "loss": 0.0336, "step": 40930 }, { "grad_norm": 0.8589796423912048, "learning_rate": 6.865348012091488e-05, "loss": 0.029, "step": 40940 }, { "grad_norm": 0.45878663659095764, "learning_rate": 6.86381381527386e-05, "loss": 0.0303, "step": 40950 }, { "grad_norm": 0.6019317507743835, "learning_rate": 6.862279414632359e-05, "loss": 0.031, "step": 40960 }, { "grad_norm": 0.5108235478401184, "learning_rate": 6.860744810334778e-05, "loss": 0.0378, "step": 40970 }, { "grad_norm": 0.3538920283317566, "learning_rate": 6.859210002548943e-05, "loss": 0.027, "step": 40980 }, { "grad_norm": 0.22866162657737732, "learning_rate": 6.8576749914427e-05, "loss": 0.0333, "step": 40990 }, { "grad_norm": 0.401348352432251, "learning_rate": 6.85613977718391e-05, "loss": 0.0289, "step": 41000 }, { "grad_norm": 0.3439684808254242, "learning_rate": 6.854604359940464e-05, "loss": 0.0265, "step": 41010 }, { "grad_norm": 0.5547568202018738, "learning_rate": 6.853068739880276e-05, "loss": 0.0283, "step": 41020 }, { "grad_norm": 0.5181288123130798, "learning_rate": 6.851532917171276e-05, "loss": 0.0427, "step": 41030 }, { "grad_norm": 0.6503776907920837, "learning_rate": 6.849996891981418e-05, "loss": 0.038, "step": 41040 }, { "grad_norm": 0.7203575968742371, "learning_rate": 6.848460664478684e-05, "loss": 0.0404, "step": 41050 }, { "grad_norm": 0.24683304131031036, "learning_rate": 6.846924234831069e-05, "loss": 0.0311, "step": 41060 }, { "grad_norm": 0.29751476645469666, "learning_rate": 6.845387603206597e-05, "loss": 0.0401, "step": 41070 }, { "grad_norm": 0.8007993698120117, "learning_rate": 6.843850769773312e-05, "loss": 0.0321, "step": 41080 }, { "grad_norm": 0.5372536182403564, "learning_rate": 6.842313734699278e-05, "loss": 0.0492, "step": 41090 }, { "grad_norm": 0.346329927444458, "learning_rate": 6.840776498152584e-05, "loss": 0.0355, "step": 41100 }, { "grad_norm": 1.15480637550354, "learning_rate": 6.83923906030134e-05, "loss": 0.0495, "step": 41110 }, { "grad_norm": 0.3967352509498596, "learning_rate": 6.837701421313677e-05, "loss": 0.0372, "step": 41120 }, { "grad_norm": 0.539177417755127, "learning_rate": 6.83616358135775e-05, "loss": 0.0402, "step": 41130 }, { "grad_norm": 0.511263370513916, "learning_rate": 6.834625540601734e-05, "loss": 0.0391, "step": 41140 }, { "grad_norm": 0.32602575421333313, "learning_rate": 6.833087299213829e-05, "loss": 0.0314, "step": 41150 }, { "grad_norm": 0.34366026520729065, "learning_rate": 6.83154885736225e-05, "loss": 0.0375, "step": 41160 }, { "grad_norm": 0.5090295672416687, "learning_rate": 6.830010215215244e-05, "loss": 0.051, "step": 41170 }, { "grad_norm": 0.3157753050327301, "learning_rate": 6.828471372941072e-05, "loss": 0.0368, "step": 41180 }, { "grad_norm": 0.4097294211387634, "learning_rate": 6.826932330708018e-05, "loss": 0.0399, "step": 41190 }, { "grad_norm": 0.23573961853981018, "learning_rate": 6.825393088684393e-05, "loss": 0.0351, "step": 41200 }, { "grad_norm": 0.605628252029419, "learning_rate": 6.823853647038524e-05, "loss": 0.0349, "step": 41210 }, { "grad_norm": 0.21993954479694366, "learning_rate": 6.822314005938762e-05, "loss": 0.0288, "step": 41220 }, { "grad_norm": 0.35156720876693726, "learning_rate": 6.820774165553479e-05, "loss": 0.0451, "step": 41230 }, { "grad_norm": 0.6780062317848206, "learning_rate": 6.819234126051072e-05, "loss": 0.034, "step": 41240 }, { "grad_norm": 1.463751196861267, "learning_rate": 6.817693887599956e-05, "loss": 0.0397, "step": 41250 }, { "grad_norm": 0.40088847279548645, "learning_rate": 6.816153450368568e-05, "loss": 0.0432, "step": 41260 }, { "grad_norm": 0.33928143978118896, "learning_rate": 6.814612814525372e-05, "loss": 0.0437, "step": 41270 }, { "grad_norm": 0.5135124921798706, "learning_rate": 6.813071980238846e-05, "loss": 0.0313, "step": 41280 }, { "grad_norm": 0.3752073049545288, "learning_rate": 6.811530947677492e-05, "loss": 0.0303, "step": 41290 }, { "grad_norm": 0.33358433842658997, "learning_rate": 6.809989717009839e-05, "loss": 0.0304, "step": 41300 }, { "grad_norm": 0.52106112241745, "learning_rate": 6.808448288404431e-05, "loss": 0.0447, "step": 41310 }, { "grad_norm": 1.0622190237045288, "learning_rate": 6.806906662029838e-05, "loss": 0.0309, "step": 41320 }, { "grad_norm": 0.3619193732738495, "learning_rate": 6.805364838054647e-05, "loss": 0.0342, "step": 41330 }, { "grad_norm": 0.4010327160358429, "learning_rate": 6.803822816647474e-05, "loss": 0.029, "step": 41340 }, { "grad_norm": 0.3959379196166992, "learning_rate": 6.802280597976949e-05, "loss": 0.0385, "step": 41350 }, { "grad_norm": 0.41695818305015564, "learning_rate": 6.800738182211727e-05, "loss": 0.0369, "step": 41360 }, { "grad_norm": 0.5354719758033752, "learning_rate": 6.799195569520485e-05, "loss": 0.0389, "step": 41370 }, { "grad_norm": 0.31100568175315857, "learning_rate": 6.79765276007192e-05, "loss": 0.0324, "step": 41380 }, { "grad_norm": 0.2895991802215576, "learning_rate": 6.796109754034753e-05, "loss": 0.0292, "step": 41390 }, { "grad_norm": 0.2964189648628235, "learning_rate": 6.794566551577724e-05, "loss": 0.0322, "step": 41400 }, { "grad_norm": 0.7126063108444214, "learning_rate": 6.793023152869594e-05, "loss": 0.0326, "step": 41410 }, { "grad_norm": 0.37103524804115295, "learning_rate": 6.791479558079147e-05, "loss": 0.0323, "step": 41420 }, { "grad_norm": 0.581584095954895, "learning_rate": 6.789935767375191e-05, "loss": 0.0367, "step": 41430 }, { "grad_norm": 0.5865527987480164, "learning_rate": 6.78839178092655e-05, "loss": 0.035, "step": 41440 }, { "grad_norm": 0.2726215124130249, "learning_rate": 6.786847598902072e-05, "loss": 0.0366, "step": 41450 }, { "grad_norm": 0.3800661563873291, "learning_rate": 6.785303221470627e-05, "loss": 0.0393, "step": 41460 }, { "grad_norm": 0.29270297288894653, "learning_rate": 6.783758648801108e-05, "loss": 0.0296, "step": 41470 }, { "grad_norm": 0.5385745763778687, "learning_rate": 6.782213881062422e-05, "loss": 0.0414, "step": 41480 }, { "grad_norm": 0.3086560070514679, "learning_rate": 6.780668918423508e-05, "loss": 0.0339, "step": 41490 }, { "grad_norm": 0.44623252749443054, "learning_rate": 6.779123761053317e-05, "loss": 0.03, "step": 41500 }, { "grad_norm": 0.4344629645347595, "learning_rate": 6.777578409120826e-05, "loss": 0.0368, "step": 41510 }, { "grad_norm": 0.4493427276611328, "learning_rate": 6.776032862795033e-05, "loss": 0.0327, "step": 41520 }, { "grad_norm": 0.35656309127807617, "learning_rate": 6.774487122244956e-05, "loss": 0.0341, "step": 41530 }, { "grad_norm": 0.3973935544490814, "learning_rate": 6.772941187639637e-05, "loss": 0.034, "step": 41540 }, { "grad_norm": 0.4536956548690796, "learning_rate": 6.771395059148134e-05, "loss": 0.0298, "step": 41550 }, { "grad_norm": 0.5067888498306274, "learning_rate": 6.76984873693953e-05, "loss": 0.0297, "step": 41560 }, { "grad_norm": 0.6301873922348022, "learning_rate": 6.768302221182931e-05, "loss": 0.0427, "step": 41570 }, { "grad_norm": 0.5547270178794861, "learning_rate": 6.766755512047457e-05, "loss": 0.0473, "step": 41580 }, { "grad_norm": 0.3437977731227875, "learning_rate": 6.765208609702259e-05, "loss": 0.0304, "step": 41590 }, { "grad_norm": 0.3705543875694275, "learning_rate": 6.763661514316499e-05, "loss": 0.0286, "step": 41600 }, { "grad_norm": 0.4481023848056793, "learning_rate": 6.76211422605937e-05, "loss": 0.0357, "step": 41610 }, { "grad_norm": 0.7079235911369324, "learning_rate": 6.760566745100076e-05, "loss": 0.0391, "step": 41620 }, { "grad_norm": 0.46686890721321106, "learning_rate": 6.75901907160785e-05, "loss": 0.0408, "step": 41630 }, { "grad_norm": 0.4561273455619812, "learning_rate": 6.757471205751943e-05, "loss": 0.0314, "step": 41640 }, { "grad_norm": 0.9493530988693237, "learning_rate": 6.75592314770163e-05, "loss": 0.0332, "step": 41650 }, { "grad_norm": 0.38116753101348877, "learning_rate": 6.7543748976262e-05, "loss": 0.0375, "step": 41660 }, { "grad_norm": 0.5556145310401917, "learning_rate": 6.752826455694968e-05, "loss": 0.0277, "step": 41670 }, { "grad_norm": 0.2770386338233948, "learning_rate": 6.751277822077271e-05, "loss": 0.0317, "step": 41680 }, { "grad_norm": 0.49925920367240906, "learning_rate": 6.749728996942465e-05, "loss": 0.0362, "step": 41690 }, { "grad_norm": 0.3150986135005951, "learning_rate": 6.748179980459924e-05, "loss": 0.0331, "step": 41700 }, { "grad_norm": 0.6732445955276489, "learning_rate": 6.746630772799052e-05, "loss": 0.0325, "step": 41710 }, { "grad_norm": 0.4522888660430908, "learning_rate": 6.745081374129262e-05, "loss": 0.0516, "step": 41720 }, { "grad_norm": 0.4986149072647095, "learning_rate": 6.743531784619998e-05, "loss": 0.0384, "step": 41730 }, { "grad_norm": 0.2577666938304901, "learning_rate": 6.741982004440719e-05, "loss": 0.0294, "step": 41740 }, { "grad_norm": 0.3571920692920685, "learning_rate": 6.740432033760907e-05, "loss": 0.0319, "step": 41750 }, { "grad_norm": 0.4308202266693115, "learning_rate": 6.738881872750066e-05, "loss": 0.0305, "step": 41760 }, { "grad_norm": 0.9513458013534546, "learning_rate": 6.737331521577715e-05, "loss": 0.0437, "step": 41770 }, { "grad_norm": 0.5287535786628723, "learning_rate": 6.735780980413403e-05, "loss": 0.0345, "step": 41780 }, { "grad_norm": 0.431191086769104, "learning_rate": 6.734230249426692e-05, "loss": 0.0449, "step": 41790 }, { "grad_norm": 0.36839795112609863, "learning_rate": 6.732679328787168e-05, "loss": 0.0373, "step": 41800 }, { "grad_norm": 0.6142537593841553, "learning_rate": 6.731128218664438e-05, "loss": 0.0354, "step": 41810 }, { "grad_norm": 0.4811866283416748, "learning_rate": 6.729576919228129e-05, "loss": 0.0368, "step": 41820 }, { "grad_norm": 0.4798731803894043, "learning_rate": 6.728025430647888e-05, "loss": 0.0318, "step": 41830 }, { "grad_norm": 0.4754186272621155, "learning_rate": 6.726473753093383e-05, "loss": 0.0391, "step": 41840 }, { "grad_norm": 0.455397367477417, "learning_rate": 6.724921886734305e-05, "loss": 0.0306, "step": 41850 }, { "grad_norm": 0.25119009613990784, "learning_rate": 6.723369831740363e-05, "loss": 0.0426, "step": 41860 }, { "grad_norm": 0.83645099401474, "learning_rate": 6.721817588281288e-05, "loss": 0.0325, "step": 41870 }, { "grad_norm": 0.505295991897583, "learning_rate": 6.720265156526828e-05, "loss": 0.033, "step": 41880 }, { "grad_norm": 0.3248940110206604, "learning_rate": 6.718712536646758e-05, "loss": 0.0297, "step": 41890 }, { "grad_norm": 0.3169110417366028, "learning_rate": 6.71715972881087e-05, "loss": 0.0524, "step": 41900 }, { "grad_norm": 0.587904691696167, "learning_rate": 6.715606733188974e-05, "loss": 0.0406, "step": 41910 }, { "grad_norm": 0.5132127404212952, "learning_rate": 6.714053549950909e-05, "loss": 0.0311, "step": 41920 }, { "grad_norm": 0.32337480783462524, "learning_rate": 6.71250017926652e-05, "loss": 0.0424, "step": 41930 }, { "grad_norm": 0.3933698832988739, "learning_rate": 6.71094662130569e-05, "loss": 0.0342, "step": 41940 }, { "grad_norm": 0.4345397353172302, "learning_rate": 6.709392876238307e-05, "loss": 0.0358, "step": 41950 }, { "grad_norm": 0.4693344831466675, "learning_rate": 6.707838944234289e-05, "loss": 0.0345, "step": 41960 }, { "grad_norm": 0.4038565754890442, "learning_rate": 6.706284825463574e-05, "loss": 0.0385, "step": 41970 }, { "grad_norm": 0.34038403630256653, "learning_rate": 6.704730520096114e-05, "loss": 0.0268, "step": 41980 }, { "grad_norm": 0.4454091489315033, "learning_rate": 6.703176028301888e-05, "loss": 0.0361, "step": 41990 }, { "grad_norm": 0.3808300495147705, "learning_rate": 6.701621350250892e-05, "loss": 0.0331, "step": 42000 }, { "grad_norm": 0.30531731247901917, "learning_rate": 6.700066486113144e-05, "loss": 0.0314, "step": 42010 }, { "grad_norm": 0.7475726008415222, "learning_rate": 6.69851143605868e-05, "loss": 0.036, "step": 42020 }, { "grad_norm": 0.41938334703445435, "learning_rate": 6.696956200257561e-05, "loss": 0.0381, "step": 42030 }, { "grad_norm": 0.23173880577087402, "learning_rate": 6.695400778879863e-05, "loss": 0.0404, "step": 42040 }, { "grad_norm": 0.47978782653808594, "learning_rate": 6.693845172095683e-05, "loss": 0.0475, "step": 42050 }, { "grad_norm": 0.4881722331047058, "learning_rate": 6.692289380075142e-05, "loss": 0.0349, "step": 42060 }, { "grad_norm": 0.20829379558563232, "learning_rate": 6.690733402988379e-05, "loss": 0.0298, "step": 42070 }, { "grad_norm": 0.35633519291877747, "learning_rate": 6.689177241005553e-05, "loss": 0.0401, "step": 42080 }, { "grad_norm": 0.3814816474914551, "learning_rate": 6.687620894296844e-05, "loss": 0.0385, "step": 42090 }, { "grad_norm": 0.4255255460739136, "learning_rate": 6.686064363032451e-05, "loss": 0.0341, "step": 42100 }, { "grad_norm": 0.8078674077987671, "learning_rate": 6.684507647382596e-05, "loss": 0.037, "step": 42110 }, { "grad_norm": 0.3626347482204437, "learning_rate": 6.682950747517513e-05, "loss": 0.0331, "step": 42120 }, { "grad_norm": 0.5286731719970703, "learning_rate": 6.68139366360747e-05, "loss": 0.0356, "step": 42130 }, { "grad_norm": 0.3067762851715088, "learning_rate": 6.679836395822744e-05, "loss": 0.0349, "step": 42140 }, { "grad_norm": 0.49907857179641724, "learning_rate": 6.678278944333633e-05, "loss": 0.0399, "step": 42150 }, { "grad_norm": 0.6020921468734741, "learning_rate": 6.676721309310462e-05, "loss": 0.0464, "step": 42160 }, { "grad_norm": 0.29768699407577515, "learning_rate": 6.67516349092357e-05, "loss": 0.0317, "step": 42170 }, { "grad_norm": 0.2692851126194, "learning_rate": 6.673605489343315e-05, "loss": 0.0333, "step": 42180 }, { "grad_norm": 0.294294536113739, "learning_rate": 6.672047304740081e-05, "loss": 0.035, "step": 42190 }, { "grad_norm": 0.5556336641311646, "learning_rate": 6.670488937284268e-05, "loss": 0.0347, "step": 42200 }, { "grad_norm": 0.3482843339443207, "learning_rate": 6.668930387146294e-05, "loss": 0.0341, "step": 42210 }, { "grad_norm": 0.5164897441864014, "learning_rate": 6.667371654496605e-05, "loss": 0.039, "step": 42220 }, { "grad_norm": 0.28883451223373413, "learning_rate": 6.665812739505659e-05, "loss": 0.0335, "step": 42230 }, { "grad_norm": 0.34586644172668457, "learning_rate": 6.664253642343935e-05, "loss": 0.0317, "step": 42240 }, { "grad_norm": 0.3428298532962799, "learning_rate": 6.662694363181935e-05, "loss": 0.038, "step": 42250 }, { "grad_norm": 0.31779223680496216, "learning_rate": 6.66113490219018e-05, "loss": 0.0342, "step": 42260 }, { "grad_norm": 0.43152880668640137, "learning_rate": 6.659575259539207e-05, "loss": 0.0456, "step": 42270 }, { "grad_norm": 1.5001204013824463, "learning_rate": 6.65801543539958e-05, "loss": 0.034, "step": 42280 }, { "grad_norm": 0.36033403873443604, "learning_rate": 6.65645542994188e-05, "loss": 0.042, "step": 42290 }, { "grad_norm": 0.31782764196395874, "learning_rate": 6.654895243336702e-05, "loss": 0.0317, "step": 42300 }, { "grad_norm": 0.4142986536026001, "learning_rate": 6.65333487575467e-05, "loss": 0.0301, "step": 42310 }, { "grad_norm": 0.5665441751480103, "learning_rate": 6.651774327366422e-05, "loss": 0.0315, "step": 42320 }, { "grad_norm": 1.668880581855774, "learning_rate": 6.650213598342613e-05, "loss": 0.0402, "step": 42330 }, { "grad_norm": 0.5057075619697571, "learning_rate": 6.648652688853931e-05, "loss": 0.0397, "step": 42340 }, { "grad_norm": 1.073456048965454, "learning_rate": 6.647091599071066e-05, "loss": 0.0407, "step": 42350 }, { "grad_norm": 0.3817110061645508, "learning_rate": 6.645530329164743e-05, "loss": 0.0354, "step": 42360 }, { "grad_norm": 0.5646499991416931, "learning_rate": 6.643968879305694e-05, "loss": 0.0384, "step": 42370 }, { "grad_norm": 0.26892438530921936, "learning_rate": 6.642407249664683e-05, "loss": 0.0262, "step": 42380 }, { "grad_norm": 0.4002719223499298, "learning_rate": 6.640845440412483e-05, "loss": 0.0384, "step": 42390 }, { "grad_norm": 0.32470929622650146, "learning_rate": 6.639283451719893e-05, "loss": 0.0297, "step": 42400 }, { "grad_norm": 0.41453567147254944, "learning_rate": 6.637721283757729e-05, "loss": 0.0284, "step": 42410 }, { "grad_norm": 0.758289098739624, "learning_rate": 6.636158936696828e-05, "loss": 0.031, "step": 42420 }, { "grad_norm": 0.4602064788341522, "learning_rate": 6.634596410708047e-05, "loss": 0.0421, "step": 42430 }, { "grad_norm": 0.601749837398529, "learning_rate": 6.633033705962257e-05, "loss": 0.0343, "step": 42440 }, { "grad_norm": 0.5784432888031006, "learning_rate": 6.631470822630359e-05, "loss": 0.0364, "step": 42450 }, { "grad_norm": 0.56002277135849, "learning_rate": 6.629907760883263e-05, "loss": 0.0436, "step": 42460 }, { "grad_norm": 0.7101974487304688, "learning_rate": 6.628344520891905e-05, "loss": 0.0319, "step": 42470 }, { "grad_norm": 0.3856634199619293, "learning_rate": 6.62678110282724e-05, "loss": 0.0334, "step": 42480 }, { "grad_norm": 0.4534544348716736, "learning_rate": 6.625217506860237e-05, "loss": 0.0362, "step": 42490 }, { "grad_norm": 0.4343525171279907, "learning_rate": 6.623653733161892e-05, "loss": 0.0337, "step": 42500 }, { "grad_norm": 0.31360793113708496, "learning_rate": 6.622089781903217e-05, "loss": 0.0371, "step": 42510 }, { "grad_norm": 0.3996891975402832, "learning_rate": 6.620525653255241e-05, "loss": 0.037, "step": 42520 }, { "grad_norm": 0.49879831075668335, "learning_rate": 6.618961347389015e-05, "loss": 0.0347, "step": 42530 }, { "grad_norm": 0.6512076258659363, "learning_rate": 6.617396864475613e-05, "loss": 0.032, "step": 42540 }, { "grad_norm": 0.39417511224746704, "learning_rate": 6.61583220468612e-05, "loss": 0.0245, "step": 42550 }, { "grad_norm": 0.31386494636535645, "learning_rate": 6.614267368191645e-05, "loss": 0.0368, "step": 42560 }, { "grad_norm": 0.4736877381801605, "learning_rate": 6.61270235516332e-05, "loss": 0.0356, "step": 42570 }, { "grad_norm": 0.36635664105415344, "learning_rate": 6.611137165772287e-05, "loss": 0.0321, "step": 42580 }, { "grad_norm": 0.3008686304092407, "learning_rate": 6.60957180018972e-05, "loss": 0.0368, "step": 42590 }, { "grad_norm": 0.5368792414665222, "learning_rate": 6.608006258586797e-05, "loss": 0.025, "step": 42600 }, { "grad_norm": 0.6786910891532898, "learning_rate": 6.606440541134731e-05, "loss": 0.037, "step": 42610 }, { "grad_norm": 0.19145521521568298, "learning_rate": 6.60487464800474e-05, "loss": 0.0489, "step": 42620 }, { "grad_norm": 0.298405259847641, "learning_rate": 6.603308579368071e-05, "loss": 0.0286, "step": 42630 }, { "grad_norm": 0.7722867727279663, "learning_rate": 6.601742335395987e-05, "loss": 0.038, "step": 42640 }, { "grad_norm": 0.5302162766456604, "learning_rate": 6.600175916259769e-05, "loss": 0.0353, "step": 42650 }, { "grad_norm": 0.2800544500350952, "learning_rate": 6.598609322130718e-05, "loss": 0.0356, "step": 42660 }, { "grad_norm": 0.3090458810329437, "learning_rate": 6.597042553180155e-05, "loss": 0.0309, "step": 42670 }, { "grad_norm": 0.4526606500148773, "learning_rate": 6.59547560957942e-05, "loss": 0.0342, "step": 42680 }, { "grad_norm": 0.849938154220581, "learning_rate": 6.59390849149987e-05, "loss": 0.0614, "step": 42690 }, { "grad_norm": 0.46605417132377625, "learning_rate": 6.592341199112886e-05, "loss": 0.0438, "step": 42700 }, { "grad_norm": 0.48959335684776306, "learning_rate": 6.590773732589861e-05, "loss": 0.0346, "step": 42710 }, { "grad_norm": 0.6598401665687561, "learning_rate": 6.589206092102213e-05, "loss": 0.0534, "step": 42720 }, { "grad_norm": 0.4180408716201782, "learning_rate": 6.587638277821375e-05, "loss": 0.041, "step": 42730 }, { "grad_norm": 0.4619709253311157, "learning_rate": 6.586070289918803e-05, "loss": 0.0411, "step": 42740 }, { "grad_norm": 0.42245355248451233, "learning_rate": 6.584502128565968e-05, "loss": 0.0362, "step": 42750 }, { "grad_norm": 0.3503273129463196, "learning_rate": 6.582933793934363e-05, "loss": 0.0325, "step": 42760 }, { "grad_norm": 0.26538336277008057, "learning_rate": 6.581365286195498e-05, "loss": 0.0349, "step": 42770 }, { "grad_norm": 0.43600451946258545, "learning_rate": 6.579796605520903e-05, "loss": 0.0329, "step": 42780 }, { "grad_norm": 0.5251896977424622, "learning_rate": 6.578227752082127e-05, "loss": 0.0312, "step": 42790 }, { "grad_norm": 0.6421582102775574, "learning_rate": 6.576658726050735e-05, "loss": 0.0325, "step": 42800 }, { "grad_norm": 0.3824252188205719, "learning_rate": 6.575089527598316e-05, "loss": 0.0279, "step": 42810 }, { "grad_norm": 0.6879721879959106, "learning_rate": 6.573520156896476e-05, "loss": 0.0327, "step": 42820 }, { "grad_norm": 0.4110564887523651, "learning_rate": 6.571950614116835e-05, "loss": 0.0331, "step": 42830 }, { "grad_norm": 0.8248251080513, "learning_rate": 6.570380899431039e-05, "loss": 0.0548, "step": 42840 }, { "grad_norm": 0.5084678530693054, "learning_rate": 6.568811013010749e-05, "loss": 0.0292, "step": 42850 }, { "grad_norm": 0.259429931640625, "learning_rate": 6.567240955027645e-05, "loss": 0.0317, "step": 42860 }, { "grad_norm": 0.4874325096607208, "learning_rate": 6.565670725653427e-05, "loss": 0.0273, "step": 42870 }, { "grad_norm": 0.3592901825904846, "learning_rate": 6.564100325059812e-05, "loss": 0.0334, "step": 42880 }, { "grad_norm": 0.41555172204971313, "learning_rate": 6.562529753418538e-05, "loss": 0.0264, "step": 42890 }, { "grad_norm": 0.4882093071937561, "learning_rate": 6.56095901090136e-05, "loss": 0.0343, "step": 42900 }, { "grad_norm": 0.29990077018737793, "learning_rate": 6.55938809768005e-05, "loss": 0.0291, "step": 42910 }, { "grad_norm": 0.6002741456031799, "learning_rate": 6.557817013926402e-05, "loss": 0.0303, "step": 42920 }, { "grad_norm": 0.24352063238620758, "learning_rate": 6.556245759812227e-05, "loss": 0.0373, "step": 42930 }, { "grad_norm": 0.424421101808548, "learning_rate": 6.554674335509357e-05, "loss": 0.0421, "step": 42940 }, { "grad_norm": 0.6564662456512451, "learning_rate": 6.553102741189638e-05, "loss": 0.0314, "step": 42950 }, { "grad_norm": 0.4244179129600525, "learning_rate": 6.55153097702494e-05, "loss": 0.0395, "step": 42960 }, { "grad_norm": 0.6195816397666931, "learning_rate": 6.549959043187146e-05, "loss": 0.0457, "step": 42970 }, { "grad_norm": 0.3078143000602722, "learning_rate": 6.54838693984816e-05, "loss": 0.0285, "step": 42980 }, { "grad_norm": 0.5805699825286865, "learning_rate": 6.546814667179909e-05, "loss": 0.0346, "step": 42990 }, { "grad_norm": 0.3724343776702881, "learning_rate": 6.545242225354328e-05, "loss": 0.0364, "step": 43000 }, { "grad_norm": 0.36688295006752014, "learning_rate": 6.543669614543383e-05, "loss": 0.0305, "step": 43010 }, { "grad_norm": 0.3674624264240265, "learning_rate": 6.542096834919049e-05, "loss": 0.0293, "step": 43020 }, { "grad_norm": 0.40538719296455383, "learning_rate": 6.540523886653324e-05, "loss": 0.0368, "step": 43030 }, { "grad_norm": 0.46050071716308594, "learning_rate": 6.538950769918222e-05, "loss": 0.0338, "step": 43040 }, { "grad_norm": 0.37863337993621826, "learning_rate": 6.537377484885779e-05, "loss": 0.0302, "step": 43050 }, { "grad_norm": 0.7922949194908142, "learning_rate": 6.535804031728045e-05, "loss": 0.0342, "step": 43060 }, { "grad_norm": 0.4254491627216339, "learning_rate": 6.534230410617092e-05, "loss": 0.037, "step": 43070 }, { "grad_norm": 0.2974119782447815, "learning_rate": 6.532656621725007e-05, "loss": 0.0419, "step": 43080 }, { "grad_norm": 0.7366470098495483, "learning_rate": 6.531082665223898e-05, "loss": 0.0355, "step": 43090 }, { "grad_norm": 0.44489821791648865, "learning_rate": 6.529508541285889e-05, "loss": 0.0282, "step": 43100 }, { "grad_norm": 0.49206554889678955, "learning_rate": 6.527934250083128e-05, "loss": 0.0403, "step": 43110 }, { "grad_norm": 0.4780605137348175, "learning_rate": 6.526359791787775e-05, "loss": 0.0407, "step": 43120 }, { "grad_norm": 0.5059494376182556, "learning_rate": 6.524785166572009e-05, "loss": 0.0425, "step": 43130 }, { "grad_norm": 0.5100350379943848, "learning_rate": 6.523210374608029e-05, "loss": 0.0361, "step": 43140 }, { "grad_norm": 0.3357236981391907, "learning_rate": 6.521635416068054e-05, "loss": 0.0328, "step": 43150 }, { "grad_norm": 0.4695849120616913, "learning_rate": 6.520060291124317e-05, "loss": 0.0485, "step": 43160 }, { "grad_norm": 0.7232396602630615, "learning_rate": 6.518484999949072e-05, "loss": 0.0337, "step": 43170 }, { "grad_norm": 0.43533560633659363, "learning_rate": 6.516909542714591e-05, "loss": 0.043, "step": 43180 }, { "grad_norm": 0.5306172966957092, "learning_rate": 6.515333919593161e-05, "loss": 0.0415, "step": 43190 }, { "grad_norm": 0.46157610416412354, "learning_rate": 6.513758130757094e-05, "loss": 0.0413, "step": 43200 }, { "grad_norm": 0.3863257169723511, "learning_rate": 6.512182176378713e-05, "loss": 0.0295, "step": 43210 }, { "grad_norm": 0.43067073822021484, "learning_rate": 6.510606056630362e-05, "loss": 0.0303, "step": 43220 }, { "grad_norm": 0.3806733787059784, "learning_rate": 6.509029771684405e-05, "loss": 0.0398, "step": 43230 }, { "grad_norm": 0.4199983477592468, "learning_rate": 6.50745332171322e-05, "loss": 0.034, "step": 43240 }, { "grad_norm": 0.800906777381897, "learning_rate": 6.505876706889207e-05, "loss": 0.0369, "step": 43250 }, { "grad_norm": 0.9791269302368164, "learning_rate": 6.504299927384781e-05, "loss": 0.04, "step": 43260 }, { "grad_norm": 0.48850151896476746, "learning_rate": 6.502722983372378e-05, "loss": 0.0322, "step": 43270 }, { "grad_norm": 0.5464194416999817, "learning_rate": 6.501145875024446e-05, "loss": 0.0447, "step": 43280 }, { "grad_norm": 0.32968375086784363, "learning_rate": 6.499568602513459e-05, "loss": 0.0389, "step": 43290 }, { "grad_norm": 0.46209874749183655, "learning_rate": 6.497991166011903e-05, "loss": 0.0398, "step": 43300 }, { "grad_norm": 0.5021122097969055, "learning_rate": 6.49641356569229e-05, "loss": 0.0337, "step": 43310 }, { "grad_norm": 0.4167625308036804, "learning_rate": 6.494835801727135e-05, "loss": 0.0377, "step": 43320 }, { "grad_norm": 0.7344597578048706, "learning_rate": 6.493257874288986e-05, "loss": 0.0358, "step": 43330 }, { "grad_norm": 0.3139162063598633, "learning_rate": 6.491679783550399e-05, "loss": 0.0341, "step": 43340 }, { "grad_norm": 0.5116096138954163, "learning_rate": 6.490101529683957e-05, "loss": 0.0273, "step": 43350 }, { "grad_norm": 0.3800768256187439, "learning_rate": 6.488523112862248e-05, "loss": 0.0443, "step": 43360 }, { "grad_norm": 0.33084744215011597, "learning_rate": 6.486944533257891e-05, "loss": 0.0501, "step": 43370 }, { "grad_norm": 0.48721885681152344, "learning_rate": 6.485365791043517e-05, "loss": 0.0395, "step": 43380 }, { "grad_norm": 0.3323051929473877, "learning_rate": 6.483786886391772e-05, "loss": 0.039, "step": 43390 }, { "grad_norm": 0.31253522634506226, "learning_rate": 6.482207819475323e-05, "loss": 0.0286, "step": 43400 }, { "grad_norm": 0.33519038558006287, "learning_rate": 6.480628590466857e-05, "loss": 0.0371, "step": 43410 }, { "grad_norm": 0.5769549012184143, "learning_rate": 6.479049199539074e-05, "loss": 0.0389, "step": 43420 }, { "grad_norm": 0.3355275094509125, "learning_rate": 6.477469646864693e-05, "loss": 0.035, "step": 43430 }, { "grad_norm": 0.4475671947002411, "learning_rate": 6.475889932616454e-05, "loss": 0.0318, "step": 43440 }, { "grad_norm": 0.39655598998069763, "learning_rate": 6.474310056967111e-05, "loss": 0.0344, "step": 43450 }, { "grad_norm": 0.5064233541488647, "learning_rate": 6.472730020089437e-05, "loss": 0.0269, "step": 43460 }, { "grad_norm": 0.5362906455993652, "learning_rate": 6.471149822156222e-05, "loss": 0.028, "step": 43470 }, { "grad_norm": 0.45754173398017883, "learning_rate": 6.469569463340275e-05, "loss": 0.0284, "step": 43480 }, { "grad_norm": 0.2486901879310608, "learning_rate": 6.467988943814418e-05, "loss": 0.0309, "step": 43490 }, { "grad_norm": 0.6527011394500732, "learning_rate": 6.4664082637515e-05, "loss": 0.0341, "step": 43500 }, { "grad_norm": 0.6592200994491577, "learning_rate": 6.464827423324377e-05, "loss": 0.0359, "step": 43510 }, { "grad_norm": 0.35172075033187866, "learning_rate": 6.463246422705931e-05, "loss": 0.0289, "step": 43520 }, { "grad_norm": 0.30094394087791443, "learning_rate": 6.461665262069055e-05, "loss": 0.0276, "step": 43530 }, { "grad_norm": 0.4115943908691406, "learning_rate": 6.460083941586665e-05, "loss": 0.0322, "step": 43540 }, { "grad_norm": 0.39339014887809753, "learning_rate": 6.45850246143169e-05, "loss": 0.0356, "step": 43550 }, { "grad_norm": 0.42495661973953247, "learning_rate": 6.456920821777077e-05, "loss": 0.035, "step": 43560 }, { "grad_norm": 0.3388688266277313, "learning_rate": 6.455339022795796e-05, "loss": 0.0269, "step": 43570 }, { "grad_norm": 0.4721132218837738, "learning_rate": 6.453757064660825e-05, "loss": 0.0301, "step": 43580 }, { "grad_norm": 0.6662044525146484, "learning_rate": 6.452174947545169e-05, "loss": 0.0298, "step": 43590 }, { "grad_norm": 0.6256831884384155, "learning_rate": 6.450592671621842e-05, "loss": 0.0305, "step": 43600 }, { "grad_norm": 0.321369469165802, "learning_rate": 6.449010237063882e-05, "loss": 0.0284, "step": 43610 }, { "grad_norm": 0.42143380641937256, "learning_rate": 6.447427644044343e-05, "loss": 0.0361, "step": 43620 }, { "grad_norm": 0.49187785387039185, "learning_rate": 6.44584489273629e-05, "loss": 0.0385, "step": 43630 }, { "grad_norm": 0.5091325044631958, "learning_rate": 6.444261983312817e-05, "loss": 0.0304, "step": 43640 }, { "grad_norm": 0.4563216269016266, "learning_rate": 6.442678915947023e-05, "loss": 0.0362, "step": 43650 }, { "grad_norm": 0.47661224007606506, "learning_rate": 6.441095690812032e-05, "loss": 0.0264, "step": 43660 }, { "grad_norm": 0.27554526925086975, "learning_rate": 6.439512308080982e-05, "loss": 0.0378, "step": 43670 }, { "grad_norm": 0.7045416831970215, "learning_rate": 6.43792876792703e-05, "loss": 0.0318, "step": 43680 }, { "grad_norm": 0.4549560248851776, "learning_rate": 6.43634507052335e-05, "loss": 0.0322, "step": 43690 }, { "grad_norm": 0.37789952754974365, "learning_rate": 6.434761216043135e-05, "loss": 0.0312, "step": 43700 }, { "grad_norm": 0.4276963174343109, "learning_rate": 6.433177204659589e-05, "loss": 0.0324, "step": 43710 }, { "grad_norm": 1.124252438545227, "learning_rate": 6.431593036545938e-05, "loss": 0.0326, "step": 43720 }, { "grad_norm": 0.4437938630580902, "learning_rate": 6.430008711875426e-05, "loss": 0.0343, "step": 43730 }, { "grad_norm": 0.42869868874549866, "learning_rate": 6.42842423082131e-05, "loss": 0.0292, "step": 43740 }, { "grad_norm": 0.38862481713294983, "learning_rate": 6.426839593556868e-05, "loss": 0.0321, "step": 43750 }, { "grad_norm": 0.327303022146225, "learning_rate": 6.425254800255392e-05, "loss": 0.033, "step": 43760 }, { "grad_norm": 0.465413898229599, "learning_rate": 6.423669851090196e-05, "loss": 0.0399, "step": 43770 }, { "grad_norm": 0.6889411211013794, "learning_rate": 6.422084746234604e-05, "loss": 0.0319, "step": 43780 }, { "grad_norm": 0.36255452036857605, "learning_rate": 6.420499485861965e-05, "loss": 0.0299, "step": 43790 }, { "grad_norm": 0.5853466391563416, "learning_rate": 6.418914070145636e-05, "loss": 0.0342, "step": 43800 }, { "grad_norm": 0.4835857152938843, "learning_rate": 6.417328499258996e-05, "loss": 0.0315, "step": 43810 }, { "grad_norm": 0.28239262104034424, "learning_rate": 6.415742773375446e-05, "loss": 0.0343, "step": 43820 }, { "grad_norm": 0.2754170596599579, "learning_rate": 6.414156892668393e-05, "loss": 0.0289, "step": 43830 }, { "grad_norm": 0.3069155514240265, "learning_rate": 6.412570857311267e-05, "loss": 0.0278, "step": 43840 }, { "grad_norm": 0.7759360074996948, "learning_rate": 6.410984667477518e-05, "loss": 0.0358, "step": 43850 }, { "grad_norm": 0.332821249961853, "learning_rate": 6.409398323340607e-05, "loss": 0.0282, "step": 43860 }, { "grad_norm": 0.3577270805835724, "learning_rate": 6.407811825074012e-05, "loss": 0.0257, "step": 43870 }, { "grad_norm": 0.2658315598964691, "learning_rate": 6.406225172851234e-05, "loss": 0.0381, "step": 43880 }, { "grad_norm": 0.37250396609306335, "learning_rate": 6.404638366845786e-05, "loss": 0.0285, "step": 43890 }, { "grad_norm": 0.5210283398628235, "learning_rate": 6.403051407231196e-05, "loss": 0.0332, "step": 43900 }, { "grad_norm": 0.4806433320045471, "learning_rate": 6.401464294181016e-05, "loss": 0.0371, "step": 43910 }, { "grad_norm": 0.415238618850708, "learning_rate": 6.399877027868808e-05, "loss": 0.0343, "step": 43920 }, { "grad_norm": 0.41847825050354004, "learning_rate": 6.39828960846815e-05, "loss": 0.0357, "step": 43930 }, { "grad_norm": 0.35516196489334106, "learning_rate": 6.396702036152644e-05, "loss": 0.0236, "step": 43940 }, { "grad_norm": 0.28015175461769104, "learning_rate": 6.395114311095905e-05, "loss": 0.0375, "step": 43950 }, { "grad_norm": 0.34101158380508423, "learning_rate": 6.393526433471559e-05, "loss": 0.0364, "step": 43960 }, { "grad_norm": 0.5303584337234497, "learning_rate": 6.391938403453262e-05, "loss": 0.046, "step": 43970 }, { "grad_norm": 1.5916262865066528, "learning_rate": 6.390350221214671e-05, "loss": 0.0355, "step": 43980 }, { "grad_norm": 0.31558266282081604, "learning_rate": 6.38876188692947e-05, "loss": 0.0464, "step": 43990 }, { "grad_norm": 0.4273056983947754, "learning_rate": 6.387173400771358e-05, "loss": 0.036, "step": 44000 }, { "grad_norm": 0.18191774189472198, "learning_rate": 6.385584762914047e-05, "loss": 0.0333, "step": 44010 }, { "grad_norm": 0.3787201941013336, "learning_rate": 6.38399597353127e-05, "loss": 0.0379, "step": 44020 }, { "grad_norm": 0.26170122623443604, "learning_rate": 6.382407032796775e-05, "loss": 0.0319, "step": 44030 }, { "grad_norm": 0.5681836605072021, "learning_rate": 6.380817940884325e-05, "loss": 0.0346, "step": 44040 }, { "grad_norm": 0.43683090806007385, "learning_rate": 6.379228697967702e-05, "loss": 0.0396, "step": 44050 }, { "grad_norm": 0.41223886609077454, "learning_rate": 6.3776393042207e-05, "loss": 0.0322, "step": 44060 }, { "grad_norm": 0.4029214382171631, "learning_rate": 6.376049759817137e-05, "loss": 0.0247, "step": 44070 }, { "grad_norm": 0.21883539855480194, "learning_rate": 6.37446006493084e-05, "loss": 0.0288, "step": 44080 }, { "grad_norm": 0.3464401960372925, "learning_rate": 6.372870219735656e-05, "loss": 0.0375, "step": 44090 }, { "grad_norm": 0.2649405300617218, "learning_rate": 6.371280224405452e-05, "loss": 0.0306, "step": 44100 }, { "grad_norm": 0.5479762554168701, "learning_rate": 6.369690079114101e-05, "loss": 0.0341, "step": 44110 }, { "grad_norm": 0.6844581365585327, "learning_rate": 6.368099784035504e-05, "loss": 0.0354, "step": 44120 }, { "grad_norm": 0.4325656294822693, "learning_rate": 6.366509339343572e-05, "loss": 0.0332, "step": 44130 }, { "grad_norm": 0.5157604217529297, "learning_rate": 6.364918745212232e-05, "loss": 0.0281, "step": 44140 }, { "grad_norm": 0.24829426407814026, "learning_rate": 6.363328001815431e-05, "loss": 0.0307, "step": 44150 }, { "grad_norm": 0.25100797414779663, "learning_rate": 6.361737109327128e-05, "loss": 0.0319, "step": 44160 }, { "grad_norm": 0.3723805844783783, "learning_rate": 6.360146067921301e-05, "loss": 0.0319, "step": 44170 }, { "grad_norm": 0.38939163088798523, "learning_rate": 6.35855487777195e-05, "loss": 0.0353, "step": 44180 }, { "grad_norm": 0.28558722138404846, "learning_rate": 6.356963539053076e-05, "loss": 0.0428, "step": 44190 }, { "grad_norm": 0.39899948239326477, "learning_rate": 6.355372051938711e-05, "loss": 0.0395, "step": 44200 }, { "grad_norm": 0.7559741735458374, "learning_rate": 6.353780416602894e-05, "loss": 0.0348, "step": 44210 }, { "grad_norm": 0.36093565821647644, "learning_rate": 6.352188633219689e-05, "loss": 0.0305, "step": 44220 }, { "grad_norm": 0.37561386823654175, "learning_rate": 6.350596701963166e-05, "loss": 0.0348, "step": 44230 }, { "grad_norm": 0.281663715839386, "learning_rate": 6.349004623007419e-05, "loss": 0.0306, "step": 44240 }, { "grad_norm": 0.5207428932189941, "learning_rate": 6.347412396526555e-05, "loss": 0.0369, "step": 44250 }, { "grad_norm": 0.3983456492424011, "learning_rate": 6.345820022694696e-05, "loss": 0.0319, "step": 44260 }, { "grad_norm": 0.5080649256706238, "learning_rate": 6.344227501685984e-05, "loss": 0.0283, "step": 44270 }, { "grad_norm": 0.3114300072193146, "learning_rate": 6.342634833674572e-05, "loss": 0.0284, "step": 44280 }, { "grad_norm": 0.43644067645072937, "learning_rate": 6.341042018834635e-05, "loss": 0.037, "step": 44290 }, { "grad_norm": 0.302146315574646, "learning_rate": 6.339449057340359e-05, "loss": 0.0345, "step": 44300 }, { "grad_norm": 0.29658836126327515, "learning_rate": 6.337855949365945e-05, "loss": 0.03, "step": 44310 }, { "grad_norm": 0.4006233215332031, "learning_rate": 6.336262695085619e-05, "loss": 0.0255, "step": 44320 }, { "grad_norm": 0.6887162923812866, "learning_rate": 6.334669294673612e-05, "loss": 0.0354, "step": 44330 }, { "grad_norm": 0.3895954191684723, "learning_rate": 6.333075748304179e-05, "loss": 0.0309, "step": 44340 }, { "grad_norm": 0.45061030983924866, "learning_rate": 6.331482056151585e-05, "loss": 0.0369, "step": 44350 }, { "grad_norm": 0.6249396204948425, "learning_rate": 6.329888218390117e-05, "loss": 0.0288, "step": 44360 }, { "grad_norm": 0.5316867828369141, "learning_rate": 6.328294235194072e-05, "loss": 0.0268, "step": 44370 }, { "grad_norm": 0.4380837082862854, "learning_rate": 6.326700106737765e-05, "loss": 0.0306, "step": 44380 }, { "grad_norm": 0.2844911515712738, "learning_rate": 6.325105833195531e-05, "loss": 0.0262, "step": 44390 }, { "grad_norm": 0.4112860858440399, "learning_rate": 6.323511414741715e-05, "loss": 0.0339, "step": 44400 }, { "grad_norm": 0.6496081352233887, "learning_rate": 6.321916851550678e-05, "loss": 0.0412, "step": 44410 }, { "grad_norm": 0.3840494155883789, "learning_rate": 6.320322143796806e-05, "loss": 0.0376, "step": 44420 }, { "grad_norm": 0.3368489146232605, "learning_rate": 6.318727291654488e-05, "loss": 0.0276, "step": 44430 }, { "grad_norm": 0.3295472264289856, "learning_rate": 6.317132295298134e-05, "loss": 0.0353, "step": 44440 }, { "grad_norm": 2.265596628189087, "learning_rate": 6.315537154902173e-05, "loss": 0.0348, "step": 44450 }, { "grad_norm": 0.6322470307350159, "learning_rate": 6.313941870641048e-05, "loss": 0.0364, "step": 44460 }, { "grad_norm": 0.4110014736652374, "learning_rate": 6.312346442689214e-05, "loss": 0.0427, "step": 44470 }, { "grad_norm": 0.5566586852073669, "learning_rate": 6.310750871221147e-05, "loss": 0.0407, "step": 44480 }, { "grad_norm": 0.5112354755401611, "learning_rate": 6.309155156411335e-05, "loss": 0.0303, "step": 44490 }, { "grad_norm": 0.930804431438446, "learning_rate": 6.307559298434284e-05, "loss": 0.0432, "step": 44500 }, { "grad_norm": 0.43076464533805847, "learning_rate": 6.305963297464512e-05, "loss": 0.034, "step": 44510 }, { "grad_norm": 0.4000369608402252, "learning_rate": 6.304367153676561e-05, "loss": 0.0351, "step": 44520 }, { "grad_norm": 0.28693872690200806, "learning_rate": 6.302770867244978e-05, "loss": 0.0351, "step": 44530 }, { "grad_norm": 0.2420736700296402, "learning_rate": 6.301174438344328e-05, "loss": 0.0341, "step": 44540 }, { "grad_norm": 0.32191208004951477, "learning_rate": 6.299577867149202e-05, "loss": 0.0414, "step": 44550 }, { "grad_norm": 0.4164857268333435, "learning_rate": 6.297981153834192e-05, "loss": 0.0353, "step": 44560 }, { "grad_norm": 0.5814740061759949, "learning_rate": 6.296384298573916e-05, "loss": 0.0399, "step": 44570 }, { "grad_norm": 0.43929216265678406, "learning_rate": 6.294787301543001e-05, "loss": 0.0492, "step": 44580 }, { "grad_norm": 0.43585824966430664, "learning_rate": 6.293190162916095e-05, "loss": 0.0294, "step": 44590 }, { "grad_norm": 1.1591870784759521, "learning_rate": 6.291592882867855e-05, "loss": 0.0367, "step": 44600 }, { "grad_norm": 0.6050161123275757, "learning_rate": 6.28999546157296e-05, "loss": 0.037, "step": 44610 }, { "grad_norm": 0.4712158441543579, "learning_rate": 6.288397899206102e-05, "loss": 0.0334, "step": 44620 }, { "grad_norm": 0.5422968864440918, "learning_rate": 6.286800195941984e-05, "loss": 0.0432, "step": 44630 }, { "grad_norm": 0.43888401985168457, "learning_rate": 6.285202351955334e-05, "loss": 0.0518, "step": 44640 }, { "grad_norm": 0.21428745985031128, "learning_rate": 6.283604367420887e-05, "loss": 0.0308, "step": 44650 }, { "grad_norm": 0.5184733867645264, "learning_rate": 6.282006242513394e-05, "loss": 0.0357, "step": 44660 }, { "grad_norm": 0.41958823800086975, "learning_rate": 6.280407977407628e-05, "loss": 0.0382, "step": 44670 }, { "grad_norm": 0.31030502915382385, "learning_rate": 6.27880957227837e-05, "loss": 0.0403, "step": 44680 }, { "grad_norm": 1.1419682502746582, "learning_rate": 6.27721102730042e-05, "loss": 0.0322, "step": 44690 }, { "grad_norm": 0.47511035203933716, "learning_rate": 6.275612342648592e-05, "loss": 0.0401, "step": 44700 }, { "grad_norm": 0.2115790992975235, "learning_rate": 6.274013518497716e-05, "loss": 0.0288, "step": 44710 }, { "grad_norm": 0.7105833292007446, "learning_rate": 6.272414555022636e-05, "loss": 0.0294, "step": 44720 }, { "grad_norm": 0.39752426743507385, "learning_rate": 6.270815452398215e-05, "loss": 0.0323, "step": 44730 }, { "grad_norm": 0.2827417254447937, "learning_rate": 6.269216210799326e-05, "loss": 0.038, "step": 44740 }, { "grad_norm": 0.32910385727882385, "learning_rate": 6.26761683040086e-05, "loss": 0.045, "step": 44750 }, { "grad_norm": 0.4726715087890625, "learning_rate": 6.266017311377723e-05, "loss": 0.0402, "step": 44760 }, { "grad_norm": 0.4371297061443329, "learning_rate": 6.264417653904839e-05, "loss": 0.0346, "step": 44770 }, { "grad_norm": 0.37366101145744324, "learning_rate": 6.262817858157139e-05, "loss": 0.0332, "step": 44780 }, { "grad_norm": 0.33235329389572144, "learning_rate": 6.261217924309576e-05, "loss": 0.0388, "step": 44790 }, { "grad_norm": 0.4859005808830261, "learning_rate": 6.259617852537118e-05, "loss": 0.0276, "step": 44800 }, { "grad_norm": 0.5492725968360901, "learning_rate": 6.258017643014747e-05, "loss": 0.0425, "step": 44810 }, { "grad_norm": 0.4831426739692688, "learning_rate": 6.256417295917456e-05, "loss": 0.0387, "step": 44820 }, { "grad_norm": 0.5057026147842407, "learning_rate": 6.254816811420258e-05, "loss": 0.0369, "step": 44830 }, { "grad_norm": 0.4347127676010132, "learning_rate": 6.253216189698183e-05, "loss": 0.0327, "step": 44840 }, { "grad_norm": 0.34099140763282776, "learning_rate": 6.251615430926267e-05, "loss": 0.0248, "step": 44850 }, { "grad_norm": 0.42293503880500793, "learning_rate": 6.25001453527957e-05, "loss": 0.033, "step": 44860 }, { "grad_norm": 0.3559451997280121, "learning_rate": 6.248413502933164e-05, "loss": 0.0371, "step": 44870 }, { "grad_norm": 0.4102655053138733, "learning_rate": 6.246812334062133e-05, "loss": 0.0415, "step": 44880 }, { "grad_norm": 0.7857890129089355, "learning_rate": 6.245211028841579e-05, "loss": 0.036, "step": 44890 }, { "grad_norm": 0.45596206188201904, "learning_rate": 6.24360958744662e-05, "loss": 0.0314, "step": 44900 }, { "grad_norm": 0.7182729840278625, "learning_rate": 6.242008010052387e-05, "loss": 0.0361, "step": 44910 }, { "grad_norm": 0.43875402212142944, "learning_rate": 6.240406296834024e-05, "loss": 0.0425, "step": 44920 }, { "grad_norm": 0.38113436102867126, "learning_rate": 6.238804447966694e-05, "loss": 0.0285, "step": 44930 }, { "grad_norm": 0.4908232092857361, "learning_rate": 6.237202463625573e-05, "loss": 0.0303, "step": 44940 }, { "grad_norm": 0.4427206218242645, "learning_rate": 6.235600343985848e-05, "loss": 0.0356, "step": 44950 }, { "grad_norm": 0.24516688287258148, "learning_rate": 6.233998089222729e-05, "loss": 0.0265, "step": 44960 }, { "grad_norm": 0.6581555008888245, "learning_rate": 6.232395699511433e-05, "loss": 0.0426, "step": 44970 }, { "grad_norm": 0.5040439367294312, "learning_rate": 6.230793175027199e-05, "loss": 0.0362, "step": 44980 }, { "grad_norm": 0.3674613833427429, "learning_rate": 6.22919051594527e-05, "loss": 0.0236, "step": 44990 }, { "grad_norm": 0.40573376417160034, "learning_rate": 6.227587722440917e-05, "loss": 0.0353, "step": 45000 }, { "grad_norm": 0.3867992162704468, "learning_rate": 6.225984794689414e-05, "loss": 0.0402, "step": 45010 }, { "grad_norm": 0.6152178645133972, "learning_rate": 6.22438173286606e-05, "loss": 0.041, "step": 45020 }, { "grad_norm": 0.5327362418174744, "learning_rate": 6.222778537146159e-05, "loss": 0.0353, "step": 45030 }, { "grad_norm": 0.7962354421615601, "learning_rate": 6.221175207705037e-05, "loss": 0.037, "step": 45040 }, { "grad_norm": 0.3142959475517273, "learning_rate": 6.219571744718027e-05, "loss": 0.033, "step": 45050 }, { "grad_norm": 0.4850214719772339, "learning_rate": 6.217968148360487e-05, "loss": 0.0376, "step": 45060 }, { "grad_norm": 0.3648717701435089, "learning_rate": 6.216364418807782e-05, "loss": 0.0318, "step": 45070 }, { "grad_norm": 0.5576516389846802, "learning_rate": 6.214760556235292e-05, "loss": 0.0296, "step": 45080 }, { "grad_norm": 1.1212629079818726, "learning_rate": 6.213156560818414e-05, "loss": 0.0349, "step": 45090 }, { "grad_norm": 0.378301739692688, "learning_rate": 6.211552432732559e-05, "loss": 0.0382, "step": 45100 }, { "grad_norm": 0.36754608154296875, "learning_rate": 6.20994817215315e-05, "loss": 0.0275, "step": 45110 }, { "grad_norm": 0.2803562879562378, "learning_rate": 6.208343779255628e-05, "loss": 0.0337, "step": 45120 }, { "grad_norm": 0.47587576508522034, "learning_rate": 6.206739254215449e-05, "loss": 0.0316, "step": 45130 }, { "grad_norm": 0.3043261170387268, "learning_rate": 6.205134597208077e-05, "loss": 0.0243, "step": 45140 }, { "grad_norm": 0.4934673309326172, "learning_rate": 6.203529808409e-05, "loss": 0.0329, "step": 45150 }, { "grad_norm": 0.46675905585289, "learning_rate": 6.201924887993712e-05, "loss": 0.0508, "step": 45160 }, { "grad_norm": 0.4800298511981964, "learning_rate": 6.200319836137725e-05, "loss": 0.0375, "step": 45170 }, { "grad_norm": 0.21463268995285034, "learning_rate": 6.198714653016565e-05, "loss": 0.0296, "step": 45180 }, { "grad_norm": 0.2835547626018524, "learning_rate": 6.197109338805774e-05, "loss": 0.0381, "step": 45190 }, { "grad_norm": 0.2751425802707672, "learning_rate": 6.195503893680903e-05, "loss": 0.0353, "step": 45200 }, { "grad_norm": 0.4135186970233917, "learning_rate": 6.193898317817524e-05, "loss": 0.0299, "step": 45210 }, { "grad_norm": 0.24977907538414001, "learning_rate": 6.192292611391221e-05, "loss": 0.0297, "step": 45220 }, { "grad_norm": 0.47629913687705994, "learning_rate": 6.190686774577591e-05, "loss": 0.0387, "step": 45230 }, { "grad_norm": 0.2266262024641037, "learning_rate": 6.189080807552245e-05, "loss": 0.0468, "step": 45240 }, { "grad_norm": 0.39411699771881104, "learning_rate": 6.187474710490809e-05, "loss": 0.0281, "step": 45250 }, { "grad_norm": 0.3355429768562317, "learning_rate": 6.185868483568926e-05, "loss": 0.0303, "step": 45260 }, { "grad_norm": 0.6392233371734619, "learning_rate": 6.184262126962245e-05, "loss": 0.0341, "step": 45270 }, { "grad_norm": 0.5016046166419983, "learning_rate": 6.182655640846442e-05, "loss": 0.0293, "step": 45280 }, { "grad_norm": 0.34056514501571655, "learning_rate": 6.181049025397196e-05, "loss": 0.0325, "step": 45290 }, { "grad_norm": 0.4134875535964966, "learning_rate": 6.179442280790202e-05, "loss": 0.0358, "step": 45300 }, { "grad_norm": 1.1611992120742798, "learning_rate": 6.177835407201174e-05, "loss": 0.0392, "step": 45310 }, { "grad_norm": 0.2578577399253845, "learning_rate": 6.176228404805839e-05, "loss": 0.0353, "step": 45320 }, { "grad_norm": 0.8216041326522827, "learning_rate": 6.174621273779932e-05, "loss": 0.0292, "step": 45330 }, { "grad_norm": 0.4004681706428528, "learning_rate": 6.17301401429921e-05, "loss": 0.0353, "step": 45340 }, { "grad_norm": 0.41665732860565186, "learning_rate": 6.171406626539436e-05, "loss": 0.0306, "step": 45350 }, { "grad_norm": 0.7721937298774719, "learning_rate": 6.169799110676398e-05, "loss": 0.0457, "step": 45360 }, { "grad_norm": 0.3618914783000946, "learning_rate": 6.168191466885885e-05, "loss": 0.0268, "step": 45370 }, { "grad_norm": 0.3692935109138489, "learning_rate": 6.166583695343711e-05, "loss": 0.0344, "step": 45380 }, { "grad_norm": 0.9388097524642944, "learning_rate": 6.164975796225698e-05, "loss": 0.0454, "step": 45390 }, { "grad_norm": 0.38621577620506287, "learning_rate": 6.163367769707683e-05, "loss": 0.025, "step": 45400 }, { "grad_norm": 0.34010571241378784, "learning_rate": 6.161759615965519e-05, "loss": 0.0405, "step": 45410 }, { "grad_norm": 0.3520328998565674, "learning_rate": 6.16015133517507e-05, "loss": 0.0338, "step": 45420 }, { "grad_norm": 0.47414007782936096, "learning_rate": 6.158542927512214e-05, "loss": 0.0394, "step": 45430 }, { "grad_norm": 0.38675999641418457, "learning_rate": 6.156934393152846e-05, "loss": 0.0255, "step": 45440 }, { "grad_norm": 0.38445204496383667, "learning_rate": 6.15532573227287e-05, "loss": 0.0398, "step": 45450 }, { "grad_norm": 0.47045525908470154, "learning_rate": 6.153716945048212e-05, "loss": 0.0453, "step": 45460 }, { "grad_norm": 0.46095114946365356, "learning_rate": 6.152108031654802e-05, "loss": 0.0446, "step": 45470 }, { "grad_norm": 1.349812388420105, "learning_rate": 6.15049899226859e-05, "loss": 0.0369, "step": 45480 }, { "grad_norm": 0.5535185933113098, "learning_rate": 6.148889827065537e-05, "loss": 0.0441, "step": 45490 }, { "grad_norm": 1.2769781351089478, "learning_rate": 6.147280536221622e-05, "loss": 0.0397, "step": 45500 }, { "grad_norm": 0.7088813781738281, "learning_rate": 6.145671119912832e-05, "loss": 0.0397, "step": 45510 }, { "grad_norm": 0.40538159012794495, "learning_rate": 6.144061578315169e-05, "loss": 0.0443, "step": 45520 }, { "grad_norm": 0.7184475064277649, "learning_rate": 6.142451911604654e-05, "loss": 0.0268, "step": 45530 }, { "grad_norm": 0.3477117121219635, "learning_rate": 6.140842119957315e-05, "loss": 0.0317, "step": 45540 }, { "grad_norm": 1.017311692237854, "learning_rate": 6.139232203549197e-05, "loss": 0.0378, "step": 45550 }, { "grad_norm": 0.4477108418941498, "learning_rate": 6.137622162556357e-05, "loss": 0.0389, "step": 45560 }, { "grad_norm": 0.29012495279312134, "learning_rate": 6.13601199715487e-05, "loss": 0.0357, "step": 45570 }, { "grad_norm": 0.2751449942588806, "learning_rate": 6.134401707520816e-05, "loss": 0.0305, "step": 45580 }, { "grad_norm": 0.331444650888443, "learning_rate": 6.132791293830298e-05, "loss": 0.0241, "step": 45590 }, { "grad_norm": 0.44302913546562195, "learning_rate": 6.131180756259428e-05, "loss": 0.0371, "step": 45600 }, { "grad_norm": 0.3917071521282196, "learning_rate": 6.129570094984331e-05, "loss": 0.0377, "step": 45610 }, { "grad_norm": 1.2838162183761597, "learning_rate": 6.127959310181145e-05, "loss": 0.0525, "step": 45620 }, { "grad_norm": 0.32109445333480835, "learning_rate": 6.126348402026026e-05, "loss": 0.034, "step": 45630 }, { "grad_norm": 0.3268768787384033, "learning_rate": 6.12473737069514e-05, "loss": 0.0341, "step": 45640 }, { "grad_norm": 0.28794950246810913, "learning_rate": 6.123126216364665e-05, "loss": 0.0326, "step": 45650 }, { "grad_norm": 0.44321298599243164, "learning_rate": 6.121514939210797e-05, "loss": 0.0317, "step": 45660 }, { "grad_norm": 0.4026380479335785, "learning_rate": 6.119903539409741e-05, "loss": 0.0327, "step": 45670 }, { "grad_norm": 0.3251964747905731, "learning_rate": 6.118292017137716e-05, "loss": 0.0418, "step": 45680 }, { "grad_norm": 0.41334131360054016, "learning_rate": 6.116680372570959e-05, "loss": 0.0298, "step": 45690 }, { "grad_norm": 0.7527034282684326, "learning_rate": 6.115068605885713e-05, "loss": 0.0302, "step": 45700 }, { "grad_norm": 0.28972408175468445, "learning_rate": 6.113456717258243e-05, "loss": 0.0341, "step": 45710 }, { "grad_norm": 0.4733351767063141, "learning_rate": 6.11184470686482e-05, "loss": 0.0476, "step": 45720 }, { "grad_norm": 0.31209665536880493, "learning_rate": 6.110232574881731e-05, "loss": 0.0227, "step": 45730 }, { "grad_norm": 0.32096564769744873, "learning_rate": 6.108620321485277e-05, "loss": 0.0313, "step": 45740 }, { "grad_norm": 0.3098985254764557, "learning_rate": 6.107007946851773e-05, "loss": 0.0382, "step": 45750 }, { "grad_norm": 0.4483199417591095, "learning_rate": 6.105395451157542e-05, "loss": 0.0408, "step": 45760 }, { "grad_norm": 0.5073318481445312, "learning_rate": 6.103782834578928e-05, "loss": 0.0376, "step": 45770 }, { "grad_norm": 0.4287470877170563, "learning_rate": 6.102170097292281e-05, "loss": 0.0315, "step": 45780 }, { "grad_norm": 0.4358597993850708, "learning_rate": 6.1005572394739716e-05, "loss": 0.0409, "step": 45790 }, { "grad_norm": 0.3117103576660156, "learning_rate": 6.0989442613003765e-05, "loss": 0.0408, "step": 45800 }, { "grad_norm": 0.48433759808540344, "learning_rate": 6.097331162947888e-05, "loss": 0.0402, "step": 45810 }, { "grad_norm": 0.38875219225883484, "learning_rate": 6.095717944592914e-05, "loss": 0.0343, "step": 45820 }, { "grad_norm": 0.28988179564476013, "learning_rate": 6.094104606411873e-05, "loss": 0.045, "step": 45830 }, { "grad_norm": 0.43239521980285645, "learning_rate": 6.0924911485811966e-05, "loss": 0.0483, "step": 45840 }, { "grad_norm": 1.0407754182815552, "learning_rate": 6.090877571277331e-05, "loss": 0.0325, "step": 45850 }, { "grad_norm": 0.5004366040229797, "learning_rate": 6.089263874676736e-05, "loss": 0.0308, "step": 45860 }, { "grad_norm": 0.5698603391647339, "learning_rate": 6.0876500589558796e-05, "loss": 0.0399, "step": 45870 }, { "grad_norm": 0.5114035606384277, "learning_rate": 6.086036124291248e-05, "loss": 0.03, "step": 45880 }, { "grad_norm": 0.3555012345314026, "learning_rate": 6.084422070859339e-05, "loss": 0.0387, "step": 45890 }, { "grad_norm": 0.34919261932373047, "learning_rate": 6.082807898836663e-05, "loss": 0.0305, "step": 45900 }, { "grad_norm": 0.43967926502227783, "learning_rate": 6.081193608399742e-05, "loss": 0.0221, "step": 45910 }, { "grad_norm": 0.35923632979393005, "learning_rate": 6.0795791997251164e-05, "loss": 0.0248, "step": 45920 }, { "grad_norm": 0.6295490264892578, "learning_rate": 6.0779646729893294e-05, "loss": 0.0314, "step": 45930 }, { "grad_norm": 0.27324479818344116, "learning_rate": 6.0763500283689476e-05, "loss": 0.0207, "step": 45940 }, { "grad_norm": 0.5575752854347229, "learning_rate": 6.0747352660405455e-05, "loss": 0.0335, "step": 45950 }, { "grad_norm": 0.25367721915245056, "learning_rate": 6.073120386180709e-05, "loss": 0.0282, "step": 45960 }, { "grad_norm": 0.44708216190338135, "learning_rate": 6.0715053889660425e-05, "loss": 0.0355, "step": 45970 }, { "grad_norm": 0.375409334897995, "learning_rate": 6.069890274573157e-05, "loss": 0.0293, "step": 45980 }, { "grad_norm": 0.7241041660308838, "learning_rate": 6.068275043178679e-05, "loss": 0.0385, "step": 45990 }, { "grad_norm": 0.5633907318115234, "learning_rate": 6.066659694959248e-05, "loss": 0.0421, "step": 46000 }, { "grad_norm": 0.28178828954696655, "learning_rate": 6.0650442300915176e-05, "loss": 0.0264, "step": 46010 }, { "grad_norm": 0.29884910583496094, "learning_rate": 6.063428648752152e-05, "loss": 0.0311, "step": 46020 }, { "grad_norm": 0.4560346007347107, "learning_rate": 6.0618129511178266e-05, "loss": 0.0318, "step": 46030 }, { "grad_norm": 0.7090746164321899, "learning_rate": 6.060197137365234e-05, "loss": 0.0328, "step": 46040 }, { "grad_norm": 0.39117616415023804, "learning_rate": 6.058581207671077e-05, "loss": 0.0281, "step": 46050 }, { "grad_norm": 0.2791864573955536, "learning_rate": 6.056965162212072e-05, "loss": 0.0294, "step": 46060 }, { "grad_norm": 0.29385116696357727, "learning_rate": 6.055349001164943e-05, "loss": 0.0333, "step": 46070 }, { "grad_norm": 0.3858928084373474, "learning_rate": 6.0537327247064347e-05, "loss": 0.0373, "step": 46080 }, { "grad_norm": 0.8188799023628235, "learning_rate": 6.0521163330133014e-05, "loss": 0.0459, "step": 46090 }, { "grad_norm": 0.5459723472595215, "learning_rate": 6.050499826262306e-05, "loss": 0.0364, "step": 46100 }, { "grad_norm": 0.4593563675880432, "learning_rate": 6.0488832046302294e-05, "loss": 0.0333, "step": 46110 }, { "grad_norm": 0.2690090537071228, "learning_rate": 6.0472664682938626e-05, "loss": 0.0331, "step": 46120 }, { "grad_norm": 0.2734754979610443, "learning_rate": 6.045649617430009e-05, "loss": 0.026, "step": 46130 }, { "grad_norm": 0.4821445345878601, "learning_rate": 6.0440326522154866e-05, "loss": 0.0268, "step": 46140 }, { "grad_norm": 0.9608132839202881, "learning_rate": 6.0424155728271224e-05, "loss": 0.0396, "step": 46150 }, { "grad_norm": 1.9722832441329956, "learning_rate": 6.040798379441758e-05, "loss": 0.0329, "step": 46160 }, { "grad_norm": 0.33975091576576233, "learning_rate": 6.0391810722362485e-05, "loss": 0.0372, "step": 46170 }, { "grad_norm": 0.5173523426055908, "learning_rate": 6.037563651387458e-05, "loss": 0.0336, "step": 46180 }, { "grad_norm": 0.5239368677139282, "learning_rate": 6.0359461170722666e-05, "loss": 0.0335, "step": 46190 }, { "grad_norm": 0.348030149936676, "learning_rate": 6.034328469467566e-05, "loss": 0.0382, "step": 46200 }, { "grad_norm": 0.40311041474342346, "learning_rate": 6.0327107087502596e-05, "loss": 0.0296, "step": 46210 }, { "grad_norm": 0.9829893708229065, "learning_rate": 6.031092835097262e-05, "loss": 0.0289, "step": 46220 }, { "grad_norm": 0.5072885155677795, "learning_rate": 6.0294748486855024e-05, "loss": 0.0326, "step": 46230 }, { "grad_norm": 0.4672289788722992, "learning_rate": 6.0278567496919216e-05, "loss": 0.0336, "step": 46240 }, { "grad_norm": 0.515905499458313, "learning_rate": 6.026238538293472e-05, "loss": 0.0324, "step": 46250 }, { "grad_norm": 0.6071465611457825, "learning_rate": 6.024620214667118e-05, "loss": 0.0309, "step": 46260 }, { "grad_norm": 0.2824912369251251, "learning_rate": 6.0230017789898384e-05, "loss": 0.0282, "step": 46270 }, { "grad_norm": 0.30794239044189453, "learning_rate": 6.021383231438622e-05, "loss": 0.0244, "step": 46280 }, { "grad_norm": 0.3334367871284485, "learning_rate": 6.0197645721904704e-05, "loss": 0.0275, "step": 46290 }, { "grad_norm": 0.5117560625076294, "learning_rate": 6.0181458014224e-05, "loss": 0.0346, "step": 46300 }, { "grad_norm": 0.7542412281036377, "learning_rate": 6.016526919311435e-05, "loss": 0.0321, "step": 46310 }, { "grad_norm": 0.48318779468536377, "learning_rate": 6.014907926034613e-05, "loss": 0.0307, "step": 46320 }, { "grad_norm": 0.4457336366176605, "learning_rate": 6.0132888217689875e-05, "loss": 0.0304, "step": 46330 }, { "grad_norm": 0.5176985263824463, "learning_rate": 6.01166960669162e-05, "loss": 0.0439, "step": 46340 }, { "grad_norm": 0.3286806046962738, "learning_rate": 6.0100502809795845e-05, "loss": 0.0331, "step": 46350 }, { "grad_norm": 0.3894435465335846, "learning_rate": 6.00843084480997e-05, "loss": 0.0352, "step": 46360 }, { "grad_norm": 0.5526867508888245, "learning_rate": 6.0068112983598736e-05, "loss": 0.029, "step": 46370 }, { "grad_norm": 0.33466455340385437, "learning_rate": 6.005191641806407e-05, "loss": 0.0292, "step": 46380 }, { "grad_norm": 0.600621223449707, "learning_rate": 6.003571875326694e-05, "loss": 0.0364, "step": 46390 }, { "grad_norm": 0.5413956642150879, "learning_rate": 6.00195199909787e-05, "loss": 0.0275, "step": 46400 }, { "grad_norm": 0.39794325828552246, "learning_rate": 6.000332013297082e-05, "loss": 0.0397, "step": 46410 }, { "grad_norm": 1.1632176637649536, "learning_rate": 5.998711918101487e-05, "loss": 0.0381, "step": 46420 }, { "grad_norm": 2.8395373821258545, "learning_rate": 5.997091713688261e-05, "loss": 0.0306, "step": 46430 }, { "grad_norm": 0.5231249928474426, "learning_rate": 5.9954714002345836e-05, "loss": 0.0363, "step": 46440 }, { "grad_norm": 0.430405855178833, "learning_rate": 5.993850977917649e-05, "loss": 0.0297, "step": 46450 }, { "grad_norm": 0.48899421095848083, "learning_rate": 5.992230446914667e-05, "loss": 0.0382, "step": 46460 }, { "grad_norm": 0.5202718377113342, "learning_rate": 5.990609807402855e-05, "loss": 0.0354, "step": 46470 }, { "grad_norm": 0.2937833070755005, "learning_rate": 5.988989059559443e-05, "loss": 0.0341, "step": 46480 }, { "grad_norm": 0.34503433108329773, "learning_rate": 5.987368203561675e-05, "loss": 0.0314, "step": 46490 }, { "grad_norm": 0.3265646994113922, "learning_rate": 5.9857472395868055e-05, "loss": 0.0257, "step": 46500 }, { "grad_norm": 0.32469311356544495, "learning_rate": 5.9841261678120983e-05, "loss": 0.0311, "step": 46510 }, { "grad_norm": 0.8559914827346802, "learning_rate": 5.982504988414834e-05, "loss": 0.0249, "step": 46520 }, { "grad_norm": 0.5104935169219971, "learning_rate": 5.9808837015723015e-05, "loss": 0.045, "step": 46530 }, { "grad_norm": 0.4963603615760803, "learning_rate": 5.9792623074618016e-05, "loss": 0.0313, "step": 46540 }, { "grad_norm": 0.30864521861076355, "learning_rate": 5.977640806260648e-05, "loss": 0.0279, "step": 46550 }, { "grad_norm": 0.5590896010398865, "learning_rate": 5.976019198146166e-05, "loss": 0.0344, "step": 46560 }, { "grad_norm": 0.5524641871452332, "learning_rate": 5.974397483295692e-05, "loss": 0.0224, "step": 46570 }, { "grad_norm": 0.399264931678772, "learning_rate": 5.972775661886575e-05, "loss": 0.0355, "step": 46580 }, { "grad_norm": 0.3071072995662689, "learning_rate": 5.971153734096173e-05, "loss": 0.0247, "step": 46590 }, { "grad_norm": 0.3078159987926483, "learning_rate": 5.969531700101859e-05, "loss": 0.0241, "step": 46600 }, { "grad_norm": 0.6086757183074951, "learning_rate": 5.9679095600810155e-05, "loss": 0.0349, "step": 46610 }, { "grad_norm": 0.4535888135433197, "learning_rate": 5.9662873142110384e-05, "loss": 0.0262, "step": 46620 }, { "grad_norm": 0.227308988571167, "learning_rate": 5.964664962669333e-05, "loss": 0.0288, "step": 46630 }, { "grad_norm": 0.23566626012325287, "learning_rate": 5.9630425056333186e-05, "loss": 0.0293, "step": 46640 }, { "grad_norm": 0.4949178695678711, "learning_rate": 5.961419943280422e-05, "loss": 0.0357, "step": 46650 }, { "grad_norm": 0.387265145778656, "learning_rate": 5.959797275788087e-05, "loss": 0.0309, "step": 46660 }, { "grad_norm": 0.3918225169181824, "learning_rate": 5.958174503333765e-05, "loss": 0.0398, "step": 46670 }, { "grad_norm": 0.30332422256469727, "learning_rate": 5.9565516260949195e-05, "loss": 0.0293, "step": 46680 }, { "grad_norm": 0.286986380815506, "learning_rate": 5.9549286442490273e-05, "loss": 0.0277, "step": 46690 }, { "grad_norm": 0.49663886427879333, "learning_rate": 5.953305557973572e-05, "loss": 0.0322, "step": 46700 }, { "grad_norm": 0.39360204339027405, "learning_rate": 5.9516823674460564e-05, "loss": 0.0318, "step": 46710 }, { "grad_norm": 0.48509296774864197, "learning_rate": 5.9500590728439875e-05, "loss": 0.0281, "step": 46720 }, { "grad_norm": 0.4635147452354431, "learning_rate": 5.948435674344886e-05, "loss": 0.0421, "step": 46730 }, { "grad_norm": 0.44711020588874817, "learning_rate": 5.946812172126285e-05, "loss": 0.0252, "step": 46740 }, { "grad_norm": 1.2286065816879272, "learning_rate": 5.945188566365729e-05, "loss": 0.0448, "step": 46750 }, { "grad_norm": 0.8204495906829834, "learning_rate": 5.943564857240773e-05, "loss": 0.0358, "step": 46760 }, { "grad_norm": 0.4798734486103058, "learning_rate": 5.941941044928981e-05, "loss": 0.0396, "step": 46770 }, { "grad_norm": 0.3066505789756775, "learning_rate": 5.940317129607935e-05, "loss": 0.0288, "step": 46780 }, { "grad_norm": 0.44854453206062317, "learning_rate": 5.9386931114552204e-05, "loss": 0.0253, "step": 46790 }, { "grad_norm": 0.32804247736930847, "learning_rate": 5.937068990648438e-05, "loss": 0.0294, "step": 46800 }, { "grad_norm": 0.23412200808525085, "learning_rate": 5.935444767365199e-05, "loss": 0.0416, "step": 46810 }, { "grad_norm": 0.4692685902118683, "learning_rate": 5.933820441783129e-05, "loss": 0.035, "step": 46820 }, { "grad_norm": 0.5193591117858887, "learning_rate": 5.932196014079857e-05, "loss": 0.0345, "step": 46830 }, { "grad_norm": 0.5122185945510864, "learning_rate": 5.930571484433032e-05, "loss": 0.0379, "step": 46840 }, { "grad_norm": 0.5793992280960083, "learning_rate": 5.928946853020309e-05, "loss": 0.0272, "step": 46850 }, { "grad_norm": 0.42376890778541565, "learning_rate": 5.927322120019352e-05, "loss": 0.0301, "step": 46860 }, { "grad_norm": 0.24628281593322754, "learning_rate": 5.925697285607844e-05, "loss": 0.0278, "step": 46870 }, { "grad_norm": 0.2844129502773285, "learning_rate": 5.924072349963472e-05, "loss": 0.0295, "step": 46880 }, { "grad_norm": 0.18620415031909943, "learning_rate": 5.922447313263938e-05, "loss": 0.0272, "step": 46890 }, { "grad_norm": 0.47777023911476135, "learning_rate": 5.920822175686952e-05, "loss": 0.0342, "step": 46900 }, { "grad_norm": 0.44970884919166565, "learning_rate": 5.919196937410239e-05, "loss": 0.0304, "step": 46910 }, { "grad_norm": 0.41479846835136414, "learning_rate": 5.917571598611529e-05, "loss": 0.0373, "step": 46920 }, { "grad_norm": 0.28074899315834045, "learning_rate": 5.9159461594685686e-05, "loss": 0.0316, "step": 46930 }, { "grad_norm": 0.3784785270690918, "learning_rate": 5.914320620159114e-05, "loss": 0.029, "step": 46940 }, { "grad_norm": 0.4104737937450409, "learning_rate": 5.912694980860931e-05, "loss": 0.0275, "step": 46950 }, { "grad_norm": 0.468567818403244, "learning_rate": 5.9110692417517964e-05, "loss": 0.0222, "step": 46960 }, { "grad_norm": 0.2473250925540924, "learning_rate": 5.9094434030095e-05, "loss": 0.0329, "step": 46970 }, { "grad_norm": 0.6593247056007385, "learning_rate": 5.907817464811841e-05, "loss": 0.0373, "step": 46980 }, { "grad_norm": 0.4499552845954895, "learning_rate": 5.906191427336627e-05, "loss": 0.0317, "step": 46990 }, { "grad_norm": 0.3740314841270447, "learning_rate": 5.9045652907616835e-05, "loss": 0.0265, "step": 47000 }, { "grad_norm": 0.29420575499534607, "learning_rate": 5.902939055264838e-05, "loss": 0.0299, "step": 47010 }, { "grad_norm": 0.31212127208709717, "learning_rate": 5.901312721023935e-05, "loss": 0.0323, "step": 47020 }, { "grad_norm": 0.318006306886673, "learning_rate": 5.8996862882168294e-05, "loss": 0.0288, "step": 47030 }, { "grad_norm": 0.3016330301761627, "learning_rate": 5.8980597570213826e-05, "loss": 0.0322, "step": 47040 }, { "grad_norm": 0.28406304121017456, "learning_rate": 5.896433127615471e-05, "loss": 0.0229, "step": 47050 }, { "grad_norm": 0.5584781765937805, "learning_rate": 5.894806400176981e-05, "loss": 0.0276, "step": 47060 }, { "grad_norm": 0.3841533362865448, "learning_rate": 5.893179574883808e-05, "loss": 0.0384, "step": 47070 }, { "grad_norm": 0.3427771031856537, "learning_rate": 5.8915526519138585e-05, "loss": 0.029, "step": 47080 }, { "grad_norm": 0.2909781038761139, "learning_rate": 5.889925631445053e-05, "loss": 0.0249, "step": 47090 }, { "grad_norm": 0.3684578537940979, "learning_rate": 5.888298513655318e-05, "loss": 0.0269, "step": 47100 }, { "grad_norm": 0.48697125911712646, "learning_rate": 5.886671298722592e-05, "loss": 0.0265, "step": 47110 }, { "grad_norm": 0.3142596483230591, "learning_rate": 5.885043986824828e-05, "loss": 0.0226, "step": 47120 }, { "grad_norm": 0.6416820287704468, "learning_rate": 5.883416578139982e-05, "loss": 0.0356, "step": 47130 }, { "grad_norm": 0.8698378801345825, "learning_rate": 5.881789072846029e-05, "loss": 0.0369, "step": 47140 }, { "grad_norm": 0.32490113377571106, "learning_rate": 5.8801614711209474e-05, "loss": 0.0218, "step": 47150 }, { "grad_norm": 0.49006274342536926, "learning_rate": 5.878533773142732e-05, "loss": 0.0344, "step": 47160 }, { "grad_norm": 0.36511602997779846, "learning_rate": 5.8769059790893834e-05, "loss": 0.0278, "step": 47170 }, { "grad_norm": 0.3455280065536499, "learning_rate": 5.8752780891389145e-05, "loss": 0.0337, "step": 47180 }, { "grad_norm": 0.36223113536834717, "learning_rate": 5.873650103469351e-05, "loss": 0.0318, "step": 47190 }, { "grad_norm": 0.2300545871257782, "learning_rate": 5.872022022258725e-05, "loss": 0.027, "step": 47200 }, { "grad_norm": 0.5437981486320496, "learning_rate": 5.8703938456850814e-05, "loss": 0.0324, "step": 47210 }, { "grad_norm": 0.34807154536247253, "learning_rate": 5.868765573926477e-05, "loss": 0.0381, "step": 47220 }, { "grad_norm": 0.5511731505393982, "learning_rate": 5.867137207160974e-05, "loss": 0.0389, "step": 47230 }, { "grad_norm": 0.3431013822555542, "learning_rate": 5.86550874556665e-05, "loss": 0.0317, "step": 47240 }, { "grad_norm": 0.2408144772052765, "learning_rate": 5.863880189321592e-05, "loss": 0.0324, "step": 47250 }, { "grad_norm": 0.6007189750671387, "learning_rate": 5.862251538603894e-05, "loss": 0.0435, "step": 47260 }, { "grad_norm": 0.5584697127342224, "learning_rate": 5.8606227935916656e-05, "loss": 0.0315, "step": 47270 }, { "grad_norm": 0.286432147026062, "learning_rate": 5.858993954463021e-05, "loss": 0.0268, "step": 47280 }, { "grad_norm": 0.40069136023521423, "learning_rate": 5.8573650213960896e-05, "loss": 0.0388, "step": 47290 }, { "grad_norm": 0.36031728982925415, "learning_rate": 5.8557359945690084e-05, "loss": 0.0283, "step": 47300 }, { "grad_norm": 0.4099176824092865, "learning_rate": 5.8541068741599246e-05, "loss": 0.0307, "step": 47310 }, { "grad_norm": 0.4240204691886902, "learning_rate": 5.8524776603469985e-05, "loss": 0.0259, "step": 47320 }, { "grad_norm": 0.43157684803009033, "learning_rate": 5.850848353308397e-05, "loss": 0.0237, "step": 47330 }, { "grad_norm": 0.9741335511207581, "learning_rate": 5.849218953222297e-05, "loss": 0.0316, "step": 47340 }, { "grad_norm": 0.4385406970977783, "learning_rate": 5.847589460266891e-05, "loss": 0.0383, "step": 47350 }, { "grad_norm": 0.45360732078552246, "learning_rate": 5.845959874620375e-05, "loss": 0.0331, "step": 47360 }, { "grad_norm": 0.2528863549232483, "learning_rate": 5.84433019646096e-05, "loss": 0.0308, "step": 47370 }, { "grad_norm": 0.3714728057384491, "learning_rate": 5.842700425966863e-05, "loss": 0.0353, "step": 47380 }, { "grad_norm": 0.18689608573913574, "learning_rate": 5.841070563316315e-05, "loss": 0.0242, "step": 47390 }, { "grad_norm": 0.5001953840255737, "learning_rate": 5.839440608687554e-05, "loss": 0.0365, "step": 47400 }, { "grad_norm": 0.38249480724334717, "learning_rate": 5.837810562258831e-05, "loss": 0.0255, "step": 47410 }, { "grad_norm": 0.3806323707103729, "learning_rate": 5.836180424208405e-05, "loss": 0.0406, "step": 47420 }, { "grad_norm": 0.5067275166511536, "learning_rate": 5.834550194714543e-05, "loss": 0.0459, "step": 47430 }, { "grad_norm": 0.3775635063648224, "learning_rate": 5.832919873955528e-05, "loss": 0.0386, "step": 47440 }, { "grad_norm": 0.33862578868865967, "learning_rate": 5.831289462109648e-05, "loss": 0.0299, "step": 47450 }, { "grad_norm": 0.6673754453659058, "learning_rate": 5.829658959355202e-05, "loss": 0.05, "step": 47460 }, { "grad_norm": 0.5858519673347473, "learning_rate": 5.8280283658704995e-05, "loss": 0.0386, "step": 47470 }, { "grad_norm": 0.4308066666126251, "learning_rate": 5.826397681833859e-05, "loss": 0.0292, "step": 47480 }, { "grad_norm": 0.44465336203575134, "learning_rate": 5.824766907423612e-05, "loss": 0.0245, "step": 47490 }, { "grad_norm": 0.3561103940010071, "learning_rate": 5.8231360428180945e-05, "loss": 0.0313, "step": 47500 }, { "grad_norm": 0.6437709331512451, "learning_rate": 5.821505088195658e-05, "loss": 0.0343, "step": 47510 }, { "grad_norm": 0.4612431824207306, "learning_rate": 5.819874043734661e-05, "loss": 0.0281, "step": 47520 }, { "grad_norm": 0.9869946837425232, "learning_rate": 5.8182429096134695e-05, "loss": 0.0243, "step": 47530 }, { "grad_norm": 0.42404070496559143, "learning_rate": 5.816611686010465e-05, "loss": 0.0309, "step": 47540 }, { "grad_norm": 0.36006006598472595, "learning_rate": 5.814980373104033e-05, "loss": 0.0325, "step": 47550 }, { "grad_norm": 0.5892615914344788, "learning_rate": 5.813348971072572e-05, "loss": 0.0398, "step": 47560 }, { "grad_norm": 0.33820900321006775, "learning_rate": 5.811717480094492e-05, "loss": 0.0294, "step": 47570 }, { "grad_norm": 0.37610888481140137, "learning_rate": 5.810085900348209e-05, "loss": 0.0337, "step": 47580 }, { "grad_norm": 0.5190565586090088, "learning_rate": 5.8084542320121483e-05, "loss": 0.0315, "step": 47590 }, { "grad_norm": 0.582615852355957, "learning_rate": 5.8068224752647497e-05, "loss": 0.0418, "step": 47600 }, { "grad_norm": 0.42900463938713074, "learning_rate": 5.805190630284457e-05, "loss": 0.0357, "step": 47610 }, { "grad_norm": 0.43146535754203796, "learning_rate": 5.803558697249729e-05, "loss": 0.0341, "step": 47620 }, { "grad_norm": 0.4057716727256775, "learning_rate": 5.8019266763390276e-05, "loss": 0.0416, "step": 47630 }, { "grad_norm": 0.24561259150505066, "learning_rate": 5.800294567730833e-05, "loss": 0.0284, "step": 47640 }, { "grad_norm": 0.31849929690361023, "learning_rate": 5.798662371603628e-05, "loss": 0.0307, "step": 47650 }, { "grad_norm": 0.28963330388069153, "learning_rate": 5.797030088135904e-05, "loss": 0.0303, "step": 47660 }, { "grad_norm": 0.3603985011577606, "learning_rate": 5.7953977175061705e-05, "loss": 0.0267, "step": 47670 }, { "grad_norm": 0.27421459555625916, "learning_rate": 5.793765259892938e-05, "loss": 0.0311, "step": 47680 }, { "grad_norm": 0.22921240329742432, "learning_rate": 5.792132715474729e-05, "loss": 0.0272, "step": 47690 }, { "grad_norm": 0.3759959042072296, "learning_rate": 5.790500084430078e-05, "loss": 0.0239, "step": 47700 }, { "grad_norm": 0.5109516382217407, "learning_rate": 5.7888673669375274e-05, "loss": 0.0279, "step": 47710 }, { "grad_norm": 0.2582080662250519, "learning_rate": 5.787234563175625e-05, "loss": 0.0303, "step": 47720 }, { "grad_norm": 0.43870341777801514, "learning_rate": 5.7856016733229355e-05, "loss": 0.0283, "step": 47730 }, { "grad_norm": 0.2750362455844879, "learning_rate": 5.7839686975580297e-05, "loss": 0.0346, "step": 47740 }, { "grad_norm": 0.524959921836853, "learning_rate": 5.782335636059484e-05, "loss": 0.0284, "step": 47750 }, { "grad_norm": 0.2732360363006592, "learning_rate": 5.780702489005889e-05, "loss": 0.0294, "step": 47760 }, { "grad_norm": 0.5865368843078613, "learning_rate": 5.779069256575845e-05, "loss": 0.0318, "step": 47770 }, { "grad_norm": 0.5461736917495728, "learning_rate": 5.7774359389479574e-05, "loss": 0.0421, "step": 47780 }, { "grad_norm": 0.37194007635116577, "learning_rate": 5.775802536300845e-05, "loss": 0.0272, "step": 47790 }, { "grad_norm": 0.3324282169342041, "learning_rate": 5.774169048813134e-05, "loss": 0.0297, "step": 47800 }, { "grad_norm": 0.2665850520133972, "learning_rate": 5.77253547666346e-05, "loss": 0.0324, "step": 47810 }, { "grad_norm": 0.6247084140777588, "learning_rate": 5.770901820030465e-05, "loss": 0.0341, "step": 47820 }, { "grad_norm": 0.4059923589229584, "learning_rate": 5.769268079092809e-05, "loss": 0.0324, "step": 47830 }, { "grad_norm": 0.5699458122253418, "learning_rate": 5.767634254029151e-05, "loss": 0.0302, "step": 47840 }, { "grad_norm": 0.355534166097641, "learning_rate": 5.7660003450181655e-05, "loss": 0.0402, "step": 47850 }, { "grad_norm": 0.3228330612182617, "learning_rate": 5.764366352238534e-05, "loss": 0.023, "step": 47860 }, { "grad_norm": 0.32499340176582336, "learning_rate": 5.7627322758689474e-05, "loss": 0.0234, "step": 47870 }, { "grad_norm": 0.36334431171417236, "learning_rate": 5.761098116088105e-05, "loss": 0.0289, "step": 47880 }, { "grad_norm": 1.4146640300750732, "learning_rate": 5.759463873074717e-05, "loss": 0.0304, "step": 47890 }, { "grad_norm": 0.4356500506401062, "learning_rate": 5.757829547007504e-05, "loss": 0.0325, "step": 47900 }, { "grad_norm": 0.5106011629104614, "learning_rate": 5.756195138065189e-05, "loss": 0.0198, "step": 47910 }, { "grad_norm": 0.3213779926300049, "learning_rate": 5.754560646426511e-05, "loss": 0.0335, "step": 47920 }, { "grad_norm": 0.315433531999588, "learning_rate": 5.752926072270216e-05, "loss": 0.0306, "step": 47930 }, { "grad_norm": 0.38668298721313477, "learning_rate": 5.7512914157750563e-05, "loss": 0.0274, "step": 47940 }, { "grad_norm": 1.3010655641555786, "learning_rate": 5.749656677119798e-05, "loss": 0.0327, "step": 47950 }, { "grad_norm": 0.4803418219089508, "learning_rate": 5.7480218564832125e-05, "loss": 0.0393, "step": 47960 }, { "grad_norm": 0.48199135065078735, "learning_rate": 5.746386954044082e-05, "loss": 0.0284, "step": 47970 }, { "grad_norm": 0.3145172595977783, "learning_rate": 5.744751969981195e-05, "loss": 0.0217, "step": 47980 }, { "grad_norm": 0.29158449172973633, "learning_rate": 5.7431169044733526e-05, "loss": 0.0352, "step": 47990 }, { "grad_norm": 0.23792070150375366, "learning_rate": 5.741481757699364e-05, "loss": 0.0286, "step": 48000 }, { "grad_norm": 0.6076698303222656, "learning_rate": 5.7398465298380434e-05, "loss": 0.0309, "step": 48010 }, { "grad_norm": 0.2361958920955658, "learning_rate": 5.7382112210682193e-05, "loss": 0.0448, "step": 48020 }, { "grad_norm": 1.1633813381195068, "learning_rate": 5.7365758315687266e-05, "loss": 0.0359, "step": 48030 }, { "grad_norm": 0.399671733379364, "learning_rate": 5.734940361518407e-05, "loss": 0.0363, "step": 48040 }, { "grad_norm": 0.44000622630119324, "learning_rate": 5.733304811096116e-05, "loss": 0.0271, "step": 48050 }, { "grad_norm": 0.2461535483598709, "learning_rate": 5.731669180480713e-05, "loss": 0.0369, "step": 48060 }, { "grad_norm": 0.45963555574417114, "learning_rate": 5.730033469851067e-05, "loss": 0.0428, "step": 48070 }, { "grad_norm": 1.3969687223434448, "learning_rate": 5.7283976793860607e-05, "loss": 0.0316, "step": 48080 }, { "grad_norm": 0.7385684251785278, "learning_rate": 5.726761809264577e-05, "loss": 0.0315, "step": 48090 }, { "grad_norm": 0.5658829808235168, "learning_rate": 5.7251258596655155e-05, "loss": 0.026, "step": 48100 }, { "grad_norm": 0.32430511713027954, "learning_rate": 5.72348983076778e-05, "loss": 0.024, "step": 48110 }, { "grad_norm": 0.42691320180892944, "learning_rate": 5.7218537227502854e-05, "loss": 0.0273, "step": 48120 }, { "grad_norm": 0.4395076036453247, "learning_rate": 5.720217535791951e-05, "loss": 0.0282, "step": 48130 }, { "grad_norm": 0.607099175453186, "learning_rate": 5.718581270071711e-05, "loss": 0.0327, "step": 48140 }, { "grad_norm": 0.6602964401245117, "learning_rate": 5.716944925768505e-05, "loss": 0.0339, "step": 48150 }, { "grad_norm": 0.5572214126586914, "learning_rate": 5.7153085030612786e-05, "loss": 0.033, "step": 48160 }, { "grad_norm": 0.34438633918762207, "learning_rate": 5.713672002128989e-05, "loss": 0.0274, "step": 48170 }, { "grad_norm": 0.5095617175102234, "learning_rate": 5.712035423150602e-05, "loss": 0.0312, "step": 48180 }, { "grad_norm": 0.29132211208343506, "learning_rate": 5.710398766305094e-05, "loss": 0.0281, "step": 48190 }, { "grad_norm": 0.4822615385055542, "learning_rate": 5.708762031771442e-05, "loss": 0.0381, "step": 48200 }, { "grad_norm": 0.6018019318580627, "learning_rate": 5.707125219728642e-05, "loss": 0.0239, "step": 48210 }, { "grad_norm": 0.35329434275627136, "learning_rate": 5.7054883303556905e-05, "loss": 0.0265, "step": 48220 }, { "grad_norm": 0.400761216878891, "learning_rate": 5.703851363831595e-05, "loss": 0.0294, "step": 48230 }, { "grad_norm": 0.5919767618179321, "learning_rate": 5.702214320335374e-05, "loss": 0.0468, "step": 48240 }, { "grad_norm": 0.4359937906265259, "learning_rate": 5.7005772000460514e-05, "loss": 0.0375, "step": 48250 }, { "grad_norm": 0.2581958770751953, "learning_rate": 5.698940003142656e-05, "loss": 0.0351, "step": 48260 }, { "grad_norm": 0.422160267829895, "learning_rate": 5.697302729804236e-05, "loss": 0.0329, "step": 48270 }, { "grad_norm": 0.3874298334121704, "learning_rate": 5.695665380209837e-05, "loss": 0.0301, "step": 48280 }, { "grad_norm": 0.30431458353996277, "learning_rate": 5.6940279545385165e-05, "loss": 0.0275, "step": 48290 }, { "grad_norm": 0.36748799681663513, "learning_rate": 5.692390452969344e-05, "loss": 0.0291, "step": 48300 }, { "grad_norm": 0.2506020963191986, "learning_rate": 5.690752875681392e-05, "loss": 0.0299, "step": 48310 }, { "grad_norm": 0.5387968420982361, "learning_rate": 5.6891152228537435e-05, "loss": 0.0306, "step": 48320 }, { "grad_norm": 0.5278222560882568, "learning_rate": 5.687477494665492e-05, "loss": 0.0266, "step": 48330 }, { "grad_norm": 0.3581477403640747, "learning_rate": 5.685839691295733e-05, "loss": 0.0295, "step": 48340 }, { "grad_norm": 0.4723329246044159, "learning_rate": 5.6842018129235786e-05, "loss": 0.0365, "step": 48350 }, { "grad_norm": 0.4954172670841217, "learning_rate": 5.6825638597281404e-05, "loss": 0.0328, "step": 48360 }, { "grad_norm": 0.45169293880462646, "learning_rate": 5.680925831888546e-05, "loss": 0.0352, "step": 48370 }, { "grad_norm": 0.31064361333847046, "learning_rate": 5.6792877295839274e-05, "loss": 0.0452, "step": 48380 }, { "grad_norm": 0.4834323823451996, "learning_rate": 5.6776495529934224e-05, "loss": 0.0359, "step": 48390 }, { "grad_norm": 0.3209650218486786, "learning_rate": 5.6760113022961824e-05, "loss": 0.0398, "step": 48400 }, { "grad_norm": 0.32249268889427185, "learning_rate": 5.6743729776713617e-05, "loss": 0.0221, "step": 48410 }, { "grad_norm": 0.23717202246189117, "learning_rate": 5.672734579298126e-05, "loss": 0.0265, "step": 48420 }, { "grad_norm": 0.431745707988739, "learning_rate": 5.671096107355649e-05, "loss": 0.0359, "step": 48430 }, { "grad_norm": 0.40215131640434265, "learning_rate": 5.66945756202311e-05, "loss": 0.0337, "step": 48440 }, { "grad_norm": 0.2539003789424896, "learning_rate": 5.667818943479699e-05, "loss": 0.0233, "step": 48450 }, { "grad_norm": 0.5903467535972595, "learning_rate": 5.666180251904612e-05, "loss": 0.0374, "step": 48460 }, { "grad_norm": 0.4567854404449463, "learning_rate": 5.6645414874770555e-05, "loss": 0.0322, "step": 48470 }, { "grad_norm": 0.4105947017669678, "learning_rate": 5.66290265037624e-05, "loss": 0.0286, "step": 48480 }, { "grad_norm": 0.4950917959213257, "learning_rate": 5.661263740781386e-05, "loss": 0.0337, "step": 48490 }, { "grad_norm": 0.5368427038192749, "learning_rate": 5.6596247588717254e-05, "loss": 0.0253, "step": 48500 }, { "grad_norm": 0.3652430772781372, "learning_rate": 5.6579857048264926e-05, "loss": 0.0298, "step": 48510 }, { "grad_norm": 0.4560070335865021, "learning_rate": 5.6563465788249314e-05, "loss": 0.0333, "step": 48520 }, { "grad_norm": 1.5754117965698242, "learning_rate": 5.6547073810462956e-05, "loss": 0.0276, "step": 48530 }, { "grad_norm": 0.2934548854827881, "learning_rate": 5.653068111669846e-05, "loss": 0.0263, "step": 48540 }, { "grad_norm": 0.3256727457046509, "learning_rate": 5.651428770874848e-05, "loss": 0.031, "step": 48550 }, { "grad_norm": 0.23314261436462402, "learning_rate": 5.64978935884058e-05, "loss": 0.0254, "step": 48560 }, { "grad_norm": 0.5229955911636353, "learning_rate": 5.6481498757463244e-05, "loss": 0.036, "step": 48570 }, { "grad_norm": 0.22668702900409698, "learning_rate": 5.646510321771373e-05, "loss": 0.03, "step": 48580 }, { "grad_norm": 0.2235344648361206, "learning_rate": 5.644870697095024e-05, "loss": 0.0347, "step": 48590 }, { "grad_norm": 0.469383180141449, "learning_rate": 5.643231001896586e-05, "loss": 0.034, "step": 48600 }, { "grad_norm": 0.43197885155677795, "learning_rate": 5.6415912363553726e-05, "loss": 0.0573, "step": 48610 }, { "grad_norm": 0.4474623203277588, "learning_rate": 5.639951400650706e-05, "loss": 0.0327, "step": 48620 }, { "grad_norm": 0.27796000242233276, "learning_rate": 5.6383114949619165e-05, "loss": 0.0319, "step": 48630 }, { "grad_norm": 0.31508976221084595, "learning_rate": 5.636671519468342e-05, "loss": 0.0318, "step": 48640 }, { "grad_norm": 0.3639717400074005, "learning_rate": 5.635031474349327e-05, "loss": 0.0377, "step": 48650 }, { "grad_norm": 0.5010038018226624, "learning_rate": 5.6333913597842246e-05, "loss": 0.0356, "step": 48660 }, { "grad_norm": 0.5697075128555298, "learning_rate": 5.6317511759523955e-05, "loss": 0.0385, "step": 48670 }, { "grad_norm": 0.28535720705986023, "learning_rate": 5.630110923033207e-05, "loss": 0.0296, "step": 48680 }, { "grad_norm": 0.4416556656360626, "learning_rate": 5.628470601206036e-05, "loss": 0.0431, "step": 48690 }, { "grad_norm": 0.6188839673995972, "learning_rate": 5.626830210650266e-05, "loss": 0.0255, "step": 48700 }, { "grad_norm": 0.4355860948562622, "learning_rate": 5.625189751545285e-05, "loss": 0.0324, "step": 48710 }, { "grad_norm": 0.3653810918331146, "learning_rate": 5.6235492240704936e-05, "loss": 0.0342, "step": 48720 }, { "grad_norm": 0.42276033759117126, "learning_rate": 5.621908628405296e-05, "loss": 0.0323, "step": 48730 }, { "grad_norm": 0.7485533356666565, "learning_rate": 5.620267964729106e-05, "loss": 0.0367, "step": 48740 }, { "grad_norm": 0.30803894996643066, "learning_rate": 5.618627233221344e-05, "loss": 0.0288, "step": 48750 }, { "grad_norm": 0.42239853739738464, "learning_rate": 5.616986434061438e-05, "loss": 0.0356, "step": 48760 }, { "grad_norm": 0.2902284264564514, "learning_rate": 5.615345567428822e-05, "loss": 0.0316, "step": 48770 }, { "grad_norm": 0.2792314291000366, "learning_rate": 5.613704633502941e-05, "loss": 0.0234, "step": 48780 }, { "grad_norm": 0.2677810788154602, "learning_rate": 5.612063632463245e-05, "loss": 0.022, "step": 48790 }, { "grad_norm": 0.2760339379310608, "learning_rate": 5.610422564489188e-05, "loss": 0.0283, "step": 48800 }, { "grad_norm": 0.46434131264686584, "learning_rate": 5.608781429760239e-05, "loss": 0.0304, "step": 48810 }, { "grad_norm": 0.580717921257019, "learning_rate": 5.607140228455866e-05, "loss": 0.0337, "step": 48820 }, { "grad_norm": 0.2852136194705963, "learning_rate": 5.605498960755553e-05, "loss": 0.0279, "step": 48830 }, { "grad_norm": 0.333911657333374, "learning_rate": 5.603857626838782e-05, "loss": 0.0332, "step": 48840 }, { "grad_norm": 0.8487313985824585, "learning_rate": 5.60221622688505e-05, "loss": 0.0348, "step": 48850 }, { "grad_norm": 0.6145350933074951, "learning_rate": 5.6005747610738565e-05, "loss": 0.034, "step": 48860 }, { "grad_norm": 0.42234787344932556, "learning_rate": 5.59893322958471e-05, "loss": 0.0246, "step": 48870 }, { "grad_norm": 0.2779742181301117, "learning_rate": 5.5972916325971256e-05, "loss": 0.04, "step": 48880 }, { "grad_norm": 0.7425851225852966, "learning_rate": 5.595649970290628e-05, "loss": 0.0368, "step": 48890 }, { "grad_norm": 0.3979838788509369, "learning_rate": 5.5940082428447426e-05, "loss": 0.034, "step": 48900 }, { "grad_norm": 0.4465990364551544, "learning_rate": 5.592366450439012e-05, "loss": 0.0457, "step": 48910 }, { "grad_norm": 0.48088744282722473, "learning_rate": 5.590724593252975e-05, "loss": 0.0317, "step": 48920 }, { "grad_norm": 0.6080944538116455, "learning_rate": 5.589082671466184e-05, "loss": 0.0301, "step": 48930 }, { "grad_norm": 0.25675052404403687, "learning_rate": 5.587440685258199e-05, "loss": 0.0271, "step": 48940 }, { "grad_norm": 0.29850271344184875, "learning_rate": 5.585798634808583e-05, "loss": 0.0221, "step": 48950 }, { "grad_norm": 0.5538147687911987, "learning_rate": 5.584156520296909e-05, "loss": 0.0276, "step": 48960 }, { "grad_norm": 0.4315246045589447, "learning_rate": 5.582514341902757e-05, "loss": 0.0237, "step": 48970 }, { "grad_norm": 0.182951882481575, "learning_rate": 5.580872099805713e-05, "loss": 0.0264, "step": 48980 }, { "grad_norm": 0.47477656602859497, "learning_rate": 5.5792297941853674e-05, "loss": 0.0319, "step": 48990 }, { "grad_norm": 0.30303069949150085, "learning_rate": 5.5775874252213247e-05, "loss": 0.032, "step": 49000 }, { "grad_norm": 0.8421832919120789, "learning_rate": 5.575944993093189e-05, "loss": 0.033, "step": 49010 }, { "grad_norm": 0.2396126091480255, "learning_rate": 5.574302497980574e-05, "loss": 0.029, "step": 49020 }, { "grad_norm": 0.524905800819397, "learning_rate": 5.5726599400631005e-05, "loss": 0.0328, "step": 49030 }, { "grad_norm": 0.3925304710865021, "learning_rate": 5.571017319520399e-05, "loss": 0.0266, "step": 49040 }, { "grad_norm": 0.6035753488540649, "learning_rate": 5.5693746365320984e-05, "loss": 0.0319, "step": 49050 }, { "grad_norm": 0.2829163372516632, "learning_rate": 5.567731891277848e-05, "loss": 0.0292, "step": 49060 }, { "grad_norm": 0.4756374657154083, "learning_rate": 5.566089083937288e-05, "loss": 0.0367, "step": 49070 }, { "grad_norm": 0.5403693914413452, "learning_rate": 5.564446214690079e-05, "loss": 0.041, "step": 49080 }, { "grad_norm": 0.27542436122894287, "learning_rate": 5.56280328371588e-05, "loss": 0.0288, "step": 49090 }, { "grad_norm": 0.3429904878139496, "learning_rate": 5.5611602911943596e-05, "loss": 0.03, "step": 49100 }, { "grad_norm": 0.43052855134010315, "learning_rate": 5.559517237305194e-05, "loss": 0.0399, "step": 49110 }, { "grad_norm": 0.4177148938179016, "learning_rate": 5.557874122228064e-05, "loss": 0.0327, "step": 49120 }, { "grad_norm": 0.5079766511917114, "learning_rate": 5.55623094614266e-05, "loss": 0.0386, "step": 49130 }, { "grad_norm": 0.30412670969963074, "learning_rate": 5.5545877092286743e-05, "loss": 0.0366, "step": 49140 }, { "grad_norm": 0.42085063457489014, "learning_rate": 5.552944411665811e-05, "loss": 0.0263, "step": 49150 }, { "grad_norm": 0.677233874797821, "learning_rate": 5.5513010536337795e-05, "loss": 0.0368, "step": 49160 }, { "grad_norm": 0.46972745656967163, "learning_rate": 5.5496576353122934e-05, "loss": 0.0348, "step": 49170 }, { "grad_norm": 0.30211955308914185, "learning_rate": 5.548014156881074e-05, "loss": 0.0255, "step": 49180 }, { "grad_norm": 0.28399622440338135, "learning_rate": 5.5463706185198494e-05, "loss": 0.0361, "step": 49190 }, { "grad_norm": 0.2784091830253601, "learning_rate": 5.544727020408358e-05, "loss": 0.0272, "step": 49200 }, { "grad_norm": 0.3470260798931122, "learning_rate": 5.5430833627263366e-05, "loss": 0.0332, "step": 49210 }, { "grad_norm": 0.37001949548721313, "learning_rate": 5.5414396456535354e-05, "loss": 0.0348, "step": 49220 }, { "grad_norm": 0.6378259658813477, "learning_rate": 5.539795869369709e-05, "loss": 0.0298, "step": 49230 }, { "grad_norm": 0.37470000982284546, "learning_rate": 5.538152034054618e-05, "loss": 0.0321, "step": 49240 }, { "grad_norm": 0.7317822575569153, "learning_rate": 5.536508139888028e-05, "loss": 0.033, "step": 49250 }, { "grad_norm": 0.30726516246795654, "learning_rate": 5.534864187049716e-05, "loss": 0.0289, "step": 49260 }, { "grad_norm": 0.9797294735908508, "learning_rate": 5.5332201757194615e-05, "loss": 0.0316, "step": 49270 }, { "grad_norm": 0.5028276443481445, "learning_rate": 5.531576106077049e-05, "loss": 0.0293, "step": 49280 }, { "grad_norm": 0.5004437565803528, "learning_rate": 5.529931978302272e-05, "loss": 0.0321, "step": 49290 }, { "grad_norm": 1.168291687965393, "learning_rate": 5.528287792574932e-05, "loss": 0.0406, "step": 49300 }, { "grad_norm": 0.3204805850982666, "learning_rate": 5.5266435490748324e-05, "loss": 0.0335, "step": 49310 }, { "grad_norm": 0.38987982273101807, "learning_rate": 5.524999247981787e-05, "loss": 0.032, "step": 49320 }, { "grad_norm": 0.43897542357444763, "learning_rate": 5.523354889475613e-05, "loss": 0.0335, "step": 49330 }, { "grad_norm": 0.3993476927280426, "learning_rate": 5.521710473736134e-05, "loss": 0.036, "step": 49340 }, { "grad_norm": 0.1820259988307953, "learning_rate": 5.5200660009431826e-05, "loss": 0.0265, "step": 49350 }, { "grad_norm": 0.5505014061927795, "learning_rate": 5.518421471276596e-05, "loss": 0.0273, "step": 49360 }, { "grad_norm": 0.869452953338623, "learning_rate": 5.5167768849162174e-05, "loss": 0.0333, "step": 49370 }, { "grad_norm": 0.6299329400062561, "learning_rate": 5.515132242041893e-05, "loss": 0.0362, "step": 49380 }, { "grad_norm": 0.402167946100235, "learning_rate": 5.513487542833483e-05, "loss": 0.0315, "step": 49390 }, { "grad_norm": 0.3906133770942688, "learning_rate": 5.5118427874708466e-05, "loss": 0.0299, "step": 49400 }, { "grad_norm": 0.37455639243125916, "learning_rate": 5.510197976133853e-05, "loss": 0.0275, "step": 49410 }, { "grad_norm": 0.5108087658882141, "learning_rate": 5.508553109002376e-05, "loss": 0.0286, "step": 49420 }, { "grad_norm": 0.3868239223957062, "learning_rate": 5.5069081862562957e-05, "loss": 0.0318, "step": 49430 }, { "grad_norm": 0.3509414494037628, "learning_rate": 5.5052632080754965e-05, "loss": 0.0286, "step": 49440 }, { "grad_norm": 0.4325297772884369, "learning_rate": 5.503618174639874e-05, "loss": 0.0349, "step": 49450 }, { "grad_norm": 0.5665913224220276, "learning_rate": 5.501973086129325e-05, "loss": 0.0326, "step": 49460 }, { "grad_norm": 0.681962251663208, "learning_rate": 5.500327942723753e-05, "loss": 0.037, "step": 49470 }, { "grad_norm": 0.6913833022117615, "learning_rate": 5.498682744603071e-05, "loss": 0.037, "step": 49480 }, { "grad_norm": 0.3790149688720703, "learning_rate": 5.497037491947194e-05, "loss": 0.0388, "step": 49490 }, { "grad_norm": 0.32635441422462463, "learning_rate": 5.4953921849360424e-05, "loss": 0.0387, "step": 49500 }, { "grad_norm": 0.5296892523765564, "learning_rate": 5.493746823749547e-05, "loss": 0.0376, "step": 49510 }, { "grad_norm": 0.30827173590660095, "learning_rate": 5.49210140856764e-05, "loss": 0.0361, "step": 49520 }, { "grad_norm": 0.29791611433029175, "learning_rate": 5.4904559395702647e-05, "loss": 0.0274, "step": 49530 }, { "grad_norm": 0.5446619987487793, "learning_rate": 5.488810416937364e-05, "loss": 0.0241, "step": 49540 }, { "grad_norm": 0.42017391324043274, "learning_rate": 5.4871648408488926e-05, "loss": 0.0315, "step": 49550 }, { "grad_norm": 0.5213726162910461, "learning_rate": 5.485519211484807e-05, "loss": 0.0305, "step": 49560 }, { "grad_norm": 0.4917627274990082, "learning_rate": 5.4838735290250676e-05, "loss": 0.0376, "step": 49570 }, { "grad_norm": 0.46964818239212036, "learning_rate": 5.482227793649649e-05, "loss": 0.0302, "step": 49580 }, { "grad_norm": 0.721048891544342, "learning_rate": 5.480582005538524e-05, "loss": 0.0345, "step": 49590 }, { "grad_norm": 0.4563334882259369, "learning_rate": 5.478936164871671e-05, "loss": 0.0279, "step": 49600 }, { "grad_norm": 0.6096211671829224, "learning_rate": 5.477290271829081e-05, "loss": 0.0417, "step": 49610 }, { "grad_norm": 0.4859163165092468, "learning_rate": 5.475644326590744e-05, "loss": 0.0306, "step": 49620 }, { "grad_norm": 0.311980664730072, "learning_rate": 5.473998329336658e-05, "loss": 0.0279, "step": 49630 }, { "grad_norm": 0.3655780553817749, "learning_rate": 5.4723522802468286e-05, "loss": 0.0251, "step": 49640 }, { "grad_norm": 0.2369106411933899, "learning_rate": 5.470706179501264e-05, "loss": 0.0294, "step": 49650 }, { "grad_norm": 0.2920266091823578, "learning_rate": 5.4690600272799776e-05, "loss": 0.0323, "step": 49660 }, { "grad_norm": 0.728143572807312, "learning_rate": 5.467413823762993e-05, "loss": 0.0292, "step": 49670 }, { "grad_norm": 0.43634161353111267, "learning_rate": 5.465767569130335e-05, "loss": 0.025, "step": 49680 }, { "grad_norm": 0.3314480185508728, "learning_rate": 5.464121263562036e-05, "loss": 0.0302, "step": 49690 }, { "grad_norm": 0.42329540848731995, "learning_rate": 5.4624749072381343e-05, "loss": 0.0318, "step": 49700 }, { "grad_norm": 0.24307842552661896, "learning_rate": 5.460828500338672e-05, "loss": 0.025, "step": 49710 }, { "grad_norm": 0.37494537234306335, "learning_rate": 5.459182043043698e-05, "loss": 0.0424, "step": 49720 }, { "grad_norm": 0.22862520813941956, "learning_rate": 5.457535535533265e-05, "loss": 0.0241, "step": 49730 }, { "grad_norm": 0.6065147519111633, "learning_rate": 5.4558889779874334e-05, "loss": 0.0394, "step": 49740 }, { "grad_norm": 0.27705469727516174, "learning_rate": 5.454242370586269e-05, "loss": 0.0306, "step": 49750 }, { "grad_norm": 0.3103538751602173, "learning_rate": 5.452595713509843e-05, "loss": 0.0325, "step": 49760 }, { "grad_norm": 0.895915687084198, "learning_rate": 5.4509490069382274e-05, "loss": 0.0282, "step": 49770 }, { "grad_norm": 0.3661503493785858, "learning_rate": 5.449302251051509e-05, "loss": 0.0305, "step": 49780 }, { "grad_norm": 0.3098697364330292, "learning_rate": 5.44765544602977e-05, "loss": 0.0382, "step": 49790 }, { "grad_norm": 0.84223872423172, "learning_rate": 5.446008592053107e-05, "loss": 0.031, "step": 49800 }, { "grad_norm": 0.3349115550518036, "learning_rate": 5.4443616893016135e-05, "loss": 0.0312, "step": 49810 }, { "grad_norm": 0.3317144215106964, "learning_rate": 5.4427147379553934e-05, "loss": 0.0277, "step": 49820 }, { "grad_norm": 0.32735475897789, "learning_rate": 5.441067738194556e-05, "loss": 0.0279, "step": 49830 }, { "grad_norm": 0.45184335112571716, "learning_rate": 5.439420690199214e-05, "loss": 0.0306, "step": 49840 }, { "grad_norm": 0.41968363523483276, "learning_rate": 5.437773594149484e-05, "loss": 0.0359, "step": 49850 }, { "grad_norm": 0.7803509831428528, "learning_rate": 5.436126450225495e-05, "loss": 0.0276, "step": 49860 }, { "grad_norm": 1.0771464109420776, "learning_rate": 5.434479258607371e-05, "loss": 0.0368, "step": 49870 }, { "grad_norm": 1.3151484727859497, "learning_rate": 5.43283201947525e-05, "loss": 0.0271, "step": 49880 }, { "grad_norm": 0.5253024697303772, "learning_rate": 5.431184733009268e-05, "loss": 0.0443, "step": 49890 }, { "grad_norm": 1.1382546424865723, "learning_rate": 5.4295373993895736e-05, "loss": 0.0284, "step": 49900 }, { "grad_norm": 0.4852745831012726, "learning_rate": 5.4278900187963157e-05, "loss": 0.0291, "step": 49910 }, { "grad_norm": 0.25211405754089355, "learning_rate": 5.426242591409646e-05, "loss": 0.0358, "step": 49920 }, { "grad_norm": 0.5186435580253601, "learning_rate": 5.42459511740973e-05, "loss": 0.0282, "step": 49930 }, { "grad_norm": 0.27852246165275574, "learning_rate": 5.422947596976729e-05, "loss": 0.0249, "step": 49940 }, { "grad_norm": 0.33345142006874084, "learning_rate": 5.4213000302908134e-05, "loss": 0.0306, "step": 49950 }, { "grad_norm": 0.3201608657836914, "learning_rate": 5.419652417532162e-05, "loss": 0.0225, "step": 49960 }, { "grad_norm": 0.3762829005718231, "learning_rate": 5.4180047588809534e-05, "loss": 0.0228, "step": 49970 }, { "grad_norm": 0.32412436604499817, "learning_rate": 5.4163570545173704e-05, "loss": 0.0245, "step": 49980 }, { "grad_norm": 0.5661022067070007, "learning_rate": 5.414709304621608e-05, "loss": 0.0354, "step": 49990 }, { "grad_norm": 0.8119369745254517, "learning_rate": 5.413061509373858e-05, "loss": 0.032, "step": 50000 }, { "grad_norm": 0.7022175788879395, "learning_rate": 5.411413668954324e-05, "loss": 0.034, "step": 50010 }, { "grad_norm": 0.34077826142311096, "learning_rate": 5.4097657835432083e-05, "loss": 0.0267, "step": 50020 }, { "grad_norm": 0.252015084028244, "learning_rate": 5.408117853320723e-05, "loss": 0.0263, "step": 50030 }, { "grad_norm": 0.5219821333885193, "learning_rate": 5.406469878467084e-05, "loss": 0.0288, "step": 50040 }, { "grad_norm": 0.21579720079898834, "learning_rate": 5.404821859162509e-05, "loss": 0.0351, "step": 50050 }, { "grad_norm": 0.4051905572414398, "learning_rate": 5.403173795587225e-05, "loss": 0.0274, "step": 50060 }, { "grad_norm": 0.5302080512046814, "learning_rate": 5.4015256879214606e-05, "loss": 0.0253, "step": 50070 }, { "grad_norm": 1.0168821811676025, "learning_rate": 5.39987753634545e-05, "loss": 0.0293, "step": 50080 }, { "grad_norm": 0.6739202737808228, "learning_rate": 5.3982293410394335e-05, "loss": 0.0358, "step": 50090 }, { "grad_norm": 0.5580827593803406, "learning_rate": 5.396581102183655e-05, "loss": 0.0337, "step": 50100 }, { "grad_norm": 0.69413161277771, "learning_rate": 5.3949328199583634e-05, "loss": 0.029, "step": 50110 }, { "grad_norm": 0.34979692101478577, "learning_rate": 5.393284494543813e-05, "loss": 0.0353, "step": 50120 }, { "grad_norm": 0.4663994014263153, "learning_rate": 5.391636126120262e-05, "loss": 0.0309, "step": 50130 }, { "grad_norm": 0.3703083097934723, "learning_rate": 5.389987714867971e-05, "loss": 0.0231, "step": 50140 }, { "grad_norm": 0.22758717834949493, "learning_rate": 5.3883392609672114e-05, "loss": 0.0259, "step": 50150 }, { "grad_norm": 0.32267019152641296, "learning_rate": 5.3866907645982545e-05, "loss": 0.0264, "step": 50160 }, { "grad_norm": 0.35858073830604553, "learning_rate": 5.385042225941375e-05, "loss": 0.0282, "step": 50170 }, { "grad_norm": 0.3648383319377899, "learning_rate": 5.3833936451768585e-05, "loss": 0.0302, "step": 50180 }, { "grad_norm": 0.42261067032814026, "learning_rate": 5.3817450224849884e-05, "loss": 0.0345, "step": 50190 }, { "grad_norm": 0.2579861283302307, "learning_rate": 5.380096358046056e-05, "loss": 0.0419, "step": 50200 }, { "grad_norm": 0.37808188796043396, "learning_rate": 5.378447652040359e-05, "loss": 0.0329, "step": 50210 }, { "grad_norm": 0.4559815227985382, "learning_rate": 5.376798904648195e-05, "loss": 0.0284, "step": 50220 }, { "grad_norm": 0.3127017021179199, "learning_rate": 5.375150116049869e-05, "loss": 0.0248, "step": 50230 }, { "grad_norm": 0.6172427535057068, "learning_rate": 5.373501286425691e-05, "loss": 0.0275, "step": 50240 }, { "grad_norm": 0.4373125433921814, "learning_rate": 5.3718524159559726e-05, "loss": 0.0283, "step": 50250 }, { "grad_norm": 0.5088798403739929, "learning_rate": 5.370203504821034e-05, "loss": 0.0328, "step": 50260 }, { "grad_norm": 0.29379144310951233, "learning_rate": 5.368554553201196e-05, "loss": 0.0304, "step": 50270 }, { "grad_norm": 0.7691614627838135, "learning_rate": 5.3669055612767874e-05, "loss": 0.0419, "step": 50280 }, { "grad_norm": 0.4629460275173187, "learning_rate": 5.3652565292281376e-05, "loss": 0.0329, "step": 50290 }, { "grad_norm": 0.919512927532196, "learning_rate": 5.3636074572355824e-05, "loss": 0.0347, "step": 50300 }, { "grad_norm": 0.21131785213947296, "learning_rate": 5.361958345479463e-05, "loss": 0.0279, "step": 50310 }, { "grad_norm": 0.3014439642429352, "learning_rate": 5.360309194140123e-05, "loss": 0.0369, "step": 50320 }, { "grad_norm": 0.36854633688926697, "learning_rate": 5.3586600033979106e-05, "loss": 0.0289, "step": 50330 }, { "grad_norm": 0.39299872517585754, "learning_rate": 5.357010773433181e-05, "loss": 0.038, "step": 50340 }, { "grad_norm": 0.5893083214759827, "learning_rate": 5.3553615044262895e-05, "loss": 0.0262, "step": 50350 }, { "grad_norm": 0.6591520309448242, "learning_rate": 5.3537121965575964e-05, "loss": 0.0231, "step": 50360 }, { "grad_norm": 0.5269178748130798, "learning_rate": 5.352062850007471e-05, "loss": 0.027, "step": 50370 }, { "grad_norm": 0.5055253505706787, "learning_rate": 5.350413464956282e-05, "loss": 0.0278, "step": 50380 }, { "grad_norm": 0.770679235458374, "learning_rate": 5.348764041584403e-05, "loss": 0.0325, "step": 50390 }, { "grad_norm": 0.9418997168540955, "learning_rate": 5.347114580072212e-05, "loss": 0.0258, "step": 50400 }, { "grad_norm": 0.38858675956726074, "learning_rate": 5.345465080600093e-05, "loss": 0.0345, "step": 50410 }, { "grad_norm": 0.6020720601081848, "learning_rate": 5.3438155433484326e-05, "loss": 0.0264, "step": 50420 }, { "grad_norm": 0.424446702003479, "learning_rate": 5.3421659684976197e-05, "loss": 0.0343, "step": 50430 }, { "grad_norm": 0.27501392364501953, "learning_rate": 5.340516356228052e-05, "loss": 0.029, "step": 50440 }, { "grad_norm": 0.5483273863792419, "learning_rate": 5.338866706720128e-05, "loss": 0.0299, "step": 50450 }, { "grad_norm": 0.2277235984802246, "learning_rate": 5.337217020154249e-05, "loss": 0.0286, "step": 50460 }, { "grad_norm": 0.34887057542800903, "learning_rate": 5.335567296710825e-05, "loss": 0.0306, "step": 50470 }, { "grad_norm": 0.49829956889152527, "learning_rate": 5.333917536570265e-05, "loss": 0.0232, "step": 50480 }, { "grad_norm": 0.2646428644657135, "learning_rate": 5.332267739912986e-05, "loss": 0.029, "step": 50490 }, { "grad_norm": 0.5564221739768982, "learning_rate": 5.330617906919405e-05, "loss": 0.0361, "step": 50500 }, { "grad_norm": 0.2563458979129791, "learning_rate": 5.328968037769949e-05, "loss": 0.033, "step": 50510 }, { "grad_norm": 0.31374964118003845, "learning_rate": 5.327318132645042e-05, "loss": 0.0282, "step": 50520 }, { "grad_norm": 0.4504620432853699, "learning_rate": 5.325668191725116e-05, "loss": 0.0298, "step": 50530 }, { "grad_norm": 0.7490648031234741, "learning_rate": 5.324018215190606e-05, "loss": 0.025, "step": 50540 }, { "grad_norm": 0.29008060693740845, "learning_rate": 5.3223682032219515e-05, "loss": 0.0275, "step": 50550 }, { "grad_norm": 0.20708203315734863, "learning_rate": 5.320718155999595e-05, "loss": 0.0269, "step": 50560 }, { "grad_norm": 0.35468316078186035, "learning_rate": 5.3190680737039835e-05, "loss": 0.0271, "step": 50570 }, { "grad_norm": 0.23807455599308014, "learning_rate": 5.317417956515567e-05, "loss": 0.0245, "step": 50580 }, { "grad_norm": 0.835609495639801, "learning_rate": 5.315767804614798e-05, "loss": 0.0328, "step": 50590 }, { "grad_norm": 0.3078923225402832, "learning_rate": 5.3141176181821395e-05, "loss": 0.0283, "step": 50600 }, { "grad_norm": 0.5041002631187439, "learning_rate": 5.31246739739805e-05, "loss": 0.0256, "step": 50610 }, { "grad_norm": 0.4078758656978607, "learning_rate": 5.310817142442994e-05, "loss": 0.0334, "step": 50620 }, { "grad_norm": 0.49576815962791443, "learning_rate": 5.309166853497445e-05, "loss": 0.0358, "step": 50630 }, { "grad_norm": 0.4912540018558502, "learning_rate": 5.307516530741873e-05, "loss": 0.0325, "step": 50640 }, { "grad_norm": 0.39253783226013184, "learning_rate": 5.305866174356754e-05, "loss": 0.0316, "step": 50650 }, { "grad_norm": 0.35989877581596375, "learning_rate": 5.304215784522571e-05, "loss": 0.0252, "step": 50660 }, { "grad_norm": 0.34256136417388916, "learning_rate": 5.302565361419808e-05, "loss": 0.0277, "step": 50670 }, { "grad_norm": 0.5072266459465027, "learning_rate": 5.3009149052289507e-05, "loss": 0.0395, "step": 50680 }, { "grad_norm": 0.43622031807899475, "learning_rate": 5.299264416130493e-05, "loss": 0.0389, "step": 50690 }, { "grad_norm": 0.382723867893219, "learning_rate": 5.297613894304928e-05, "loss": 0.0291, "step": 50700 }, { "grad_norm": 1.169074535369873, "learning_rate": 5.2959633399327534e-05, "loss": 0.0318, "step": 50710 }, { "grad_norm": 0.3906435966491699, "learning_rate": 5.294312753194476e-05, "loss": 0.0291, "step": 50720 }, { "grad_norm": 0.3163647949695587, "learning_rate": 5.292662134270596e-05, "loss": 0.0268, "step": 50730 }, { "grad_norm": 0.41180577874183655, "learning_rate": 5.291011483341626e-05, "loss": 0.0373, "step": 50740 }, { "grad_norm": 0.35085463523864746, "learning_rate": 5.2893608005880767e-05, "loss": 0.0264, "step": 50750 }, { "grad_norm": 0.588840663433075, "learning_rate": 5.287710086190467e-05, "loss": 0.0439, "step": 50760 }, { "grad_norm": 0.28632041811943054, "learning_rate": 5.286059340329314e-05, "loss": 0.0319, "step": 50770 }, { "grad_norm": 0.37105950713157654, "learning_rate": 5.284408563185141e-05, "loss": 0.0353, "step": 50780 }, { "grad_norm": 0.5131710171699524, "learning_rate": 5.282757754938476e-05, "loss": 0.0311, "step": 50790 }, { "grad_norm": 0.3904326260089874, "learning_rate": 5.281106915769849e-05, "loss": 0.0304, "step": 50800 }, { "grad_norm": 0.3025017976760864, "learning_rate": 5.2794560458597897e-05, "loss": 0.0228, "step": 50810 }, { "grad_norm": 0.3501259386539459, "learning_rate": 5.27780514538884e-05, "loss": 0.0245, "step": 50820 }, { "grad_norm": 0.21475578844547272, "learning_rate": 5.2761542145375365e-05, "loss": 0.0297, "step": 50830 }, { "grad_norm": 0.4536674916744232, "learning_rate": 5.274503253486421e-05, "loss": 0.024, "step": 50840 }, { "grad_norm": 0.35559889674186707, "learning_rate": 5.272852262416046e-05, "loss": 0.0237, "step": 50850 }, { "grad_norm": 0.721781849861145, "learning_rate": 5.2712012415069555e-05, "loss": 0.0298, "step": 50860 }, { "grad_norm": 0.7347685694694519, "learning_rate": 5.269550190939705e-05, "loss": 0.0456, "step": 50870 }, { "grad_norm": 0.39622730016708374, "learning_rate": 5.267899110894852e-05, "loss": 0.0322, "step": 50880 }, { "grad_norm": 0.2791634500026703, "learning_rate": 5.266248001552955e-05, "loss": 0.0253, "step": 50890 }, { "grad_norm": 0.31146034598350525, "learning_rate": 5.264596863094575e-05, "loss": 0.0236, "step": 50900 }, { "grad_norm": 0.3851374387741089, "learning_rate": 5.262945695700282e-05, "loss": 0.0397, "step": 50910 }, { "grad_norm": 0.35440531373023987, "learning_rate": 5.261294499550643e-05, "loss": 0.0235, "step": 50920 }, { "grad_norm": 0.4300096035003662, "learning_rate": 5.25964327482623e-05, "loss": 0.0244, "step": 50930 }, { "grad_norm": 0.30039405822753906, "learning_rate": 5.257992021707617e-05, "loss": 0.0393, "step": 50940 }, { "grad_norm": 0.6044318675994873, "learning_rate": 5.256340740375387e-05, "loss": 0.0355, "step": 50950 }, { "grad_norm": 0.3843453526496887, "learning_rate": 5.254689431010117e-05, "loss": 0.0347, "step": 50960 }, { "grad_norm": 0.772076427936554, "learning_rate": 5.253038093792395e-05, "loss": 0.0265, "step": 50970 }, { "grad_norm": 0.4266960322856903, "learning_rate": 5.251386728902806e-05, "loss": 0.036, "step": 50980 }, { "grad_norm": 0.3266679346561432, "learning_rate": 5.2497353365219446e-05, "loss": 0.0306, "step": 50990 }, { "grad_norm": 0.40788695216178894, "learning_rate": 5.2480839168304e-05, "loss": 0.0309, "step": 51000 }, { "grad_norm": 0.5195409059524536, "learning_rate": 5.246432470008773e-05, "loss": 0.0248, "step": 51010 }, { "grad_norm": 0.49133315682411194, "learning_rate": 5.2447809962376616e-05, "loss": 0.0317, "step": 51020 }, { "grad_norm": 0.5613606572151184, "learning_rate": 5.243129495697666e-05, "loss": 0.0277, "step": 51030 }, { "grad_norm": 0.4830307960510254, "learning_rate": 5.241477968569397e-05, "loss": 0.0355, "step": 51040 }, { "grad_norm": 0.4418702721595764, "learning_rate": 5.23982641503346e-05, "loss": 0.0372, "step": 51050 }, { "grad_norm": 0.566129207611084, "learning_rate": 5.238174835270465e-05, "loss": 0.0355, "step": 51060 }, { "grad_norm": 0.4751199185848236, "learning_rate": 5.23652322946103e-05, "loss": 0.0279, "step": 51070 }, { "grad_norm": 0.42787784337997437, "learning_rate": 5.234871597785771e-05, "loss": 0.0284, "step": 51080 }, { "grad_norm": 0.46624770760536194, "learning_rate": 5.233219940425306e-05, "loss": 0.0353, "step": 51090 }, { "grad_norm": 0.5503450036048889, "learning_rate": 5.2315682575602576e-05, "loss": 0.0455, "step": 51100 }, { "grad_norm": 0.2786264419555664, "learning_rate": 5.229916549371254e-05, "loss": 0.0321, "step": 51110 }, { "grad_norm": 0.3881898522377014, "learning_rate": 5.228264816038924e-05, "loss": 0.0253, "step": 51120 }, { "grad_norm": 0.23793506622314453, "learning_rate": 5.2266130577438935e-05, "loss": 0.0245, "step": 51130 }, { "grad_norm": 0.26750844717025757, "learning_rate": 5.224961274666802e-05, "loss": 0.0253, "step": 51140 }, { "grad_norm": 0.2907799184322357, "learning_rate": 5.223309466988283e-05, "loss": 0.034, "step": 51150 }, { "grad_norm": 0.5585856437683105, "learning_rate": 5.221657634888976e-05, "loss": 0.0451, "step": 51160 }, { "grad_norm": 0.4337571859359741, "learning_rate": 5.2200057785495235e-05, "loss": 0.0354, "step": 51170 }, { "grad_norm": 0.4064595401287079, "learning_rate": 5.2183538981505706e-05, "loss": 0.0304, "step": 51180 }, { "grad_norm": 0.30040428042411804, "learning_rate": 5.216701993872762e-05, "loss": 0.0232, "step": 51190 }, { "grad_norm": 0.44666019082069397, "learning_rate": 5.215050065896752e-05, "loss": 0.0304, "step": 51200 }, { "grad_norm": 0.4411696195602417, "learning_rate": 5.2133981144031875e-05, "loss": 0.0421, "step": 51210 }, { "grad_norm": 0.35753440856933594, "learning_rate": 5.211746139572727e-05, "loss": 0.0367, "step": 51220 }, { "grad_norm": 0.5836200714111328, "learning_rate": 5.210094141586026e-05, "loss": 0.0308, "step": 51230 }, { "grad_norm": 0.41065940260887146, "learning_rate": 5.2084421206237476e-05, "loss": 0.0311, "step": 51240 }, { "grad_norm": 0.30077898502349854, "learning_rate": 5.20679007686655e-05, "loss": 0.0256, "step": 51250 }, { "grad_norm": 0.39519748091697693, "learning_rate": 5.205138010495102e-05, "loss": 0.0269, "step": 51260 }, { "grad_norm": 0.2676931917667389, "learning_rate": 5.20348592169007e-05, "loss": 0.0316, "step": 51270 }, { "grad_norm": 0.4131457209587097, "learning_rate": 5.201833810632123e-05, "loss": 0.0356, "step": 51280 }, { "grad_norm": 0.51485675573349, "learning_rate": 5.200181677501933e-05, "loss": 0.0273, "step": 51290 }, { "grad_norm": 0.5168179869651794, "learning_rate": 5.1985295224801777e-05, "loss": 0.0317, "step": 51300 }, { "grad_norm": 0.3507746756076813, "learning_rate": 5.196877345747531e-05, "loss": 0.0265, "step": 51310 }, { "grad_norm": 0.3873862326145172, "learning_rate": 5.195225147484674e-05, "loss": 0.0314, "step": 51320 }, { "grad_norm": 0.6557160019874573, "learning_rate": 5.193572927872288e-05, "loss": 0.0385, "step": 51330 }, { "grad_norm": 0.3467380702495575, "learning_rate": 5.191920687091059e-05, "loss": 0.0341, "step": 51340 }, { "grad_norm": 0.5113701820373535, "learning_rate": 5.1902684253216695e-05, "loss": 0.0257, "step": 51350 }, { "grad_norm": 0.37601229548454285, "learning_rate": 5.1886161427448134e-05, "loss": 0.0311, "step": 51360 }, { "grad_norm": 0.4461612105369568, "learning_rate": 5.186963839541179e-05, "loss": 0.0274, "step": 51370 }, { "grad_norm": 0.32089853286743164, "learning_rate": 5.185311515891459e-05, "loss": 0.0248, "step": 51380 }, { "grad_norm": 0.4543326795101166, "learning_rate": 5.1836591719763516e-05, "loss": 0.0258, "step": 51390 }, { "grad_norm": 0.2063264548778534, "learning_rate": 5.182006807976554e-05, "loss": 0.0287, "step": 51400 }, { "grad_norm": 0.41948065161705017, "learning_rate": 5.180354424072763e-05, "loss": 0.0213, "step": 51410 }, { "grad_norm": 0.6786588430404663, "learning_rate": 5.178702020445686e-05, "loss": 0.0249, "step": 51420 }, { "grad_norm": 0.5645367503166199, "learning_rate": 5.177049597276024e-05, "loss": 0.0367, "step": 51430 }, { "grad_norm": 0.4611809253692627, "learning_rate": 5.175397154744482e-05, "loss": 0.027, "step": 51440 }, { "grad_norm": 0.43945860862731934, "learning_rate": 5.1737446930317736e-05, "loss": 0.0201, "step": 51450 }, { "grad_norm": 0.28276535868644714, "learning_rate": 5.172092212318604e-05, "loss": 0.0316, "step": 51460 }, { "grad_norm": 0.4273075461387634, "learning_rate": 5.170439712785692e-05, "loss": 0.0323, "step": 51470 }, { "grad_norm": 1.0564539432525635, "learning_rate": 5.1687871946137467e-05, "loss": 0.0279, "step": 51480 }, { "grad_norm": 0.3016927242279053, "learning_rate": 5.16713465798349e-05, "loss": 0.03, "step": 51490 }, { "grad_norm": 0.5591642260551453, "learning_rate": 5.165482103075637e-05, "loss": 0.0389, "step": 51500 }, { "grad_norm": 0.26482275128364563, "learning_rate": 5.163829530070909e-05, "loss": 0.0294, "step": 51510 }, { "grad_norm": 0.5402645468711853, "learning_rate": 5.162176939150032e-05, "loss": 0.04, "step": 51520 }, { "grad_norm": 0.5746093988418579, "learning_rate": 5.160524330493728e-05, "loss": 0.0288, "step": 51530 }, { "grad_norm": 0.507724940776825, "learning_rate": 5.158871704282724e-05, "loss": 0.0296, "step": 51540 }, { "grad_norm": 0.28149330615997314, "learning_rate": 5.15721906069775e-05, "loss": 0.0324, "step": 51550 }, { "grad_norm": 0.437618225812912, "learning_rate": 5.155566399919536e-05, "loss": 0.0265, "step": 51560 }, { "grad_norm": 0.3925884962081909, "learning_rate": 5.153913722128813e-05, "loss": 0.0278, "step": 51570 }, { "grad_norm": 0.302827924489975, "learning_rate": 5.1522610275063176e-05, "loss": 0.0246, "step": 51580 }, { "grad_norm": 0.34413090348243713, "learning_rate": 5.150608316232786e-05, "loss": 0.028, "step": 51590 }, { "grad_norm": 0.3028325140476227, "learning_rate": 5.148955588488955e-05, "loss": 0.0269, "step": 51600 }, { "grad_norm": 0.40912508964538574, "learning_rate": 5.147302844455564e-05, "loss": 0.0371, "step": 51610 }, { "grad_norm": 0.3261508345603943, "learning_rate": 5.145650084313356e-05, "loss": 0.0319, "step": 51620 }, { "grad_norm": 0.4384460151195526, "learning_rate": 5.1439973082430734e-05, "loss": 0.0269, "step": 51630 }, { "grad_norm": 0.34099745750427246, "learning_rate": 5.142344516425461e-05, "loss": 0.0237, "step": 51640 }, { "grad_norm": 0.385729044675827, "learning_rate": 5.1406917090412673e-05, "loss": 0.0295, "step": 51650 }, { "grad_norm": 0.46361491084098816, "learning_rate": 5.1390388862712404e-05, "loss": 0.0265, "step": 51660 }, { "grad_norm": 0.2985980212688446, "learning_rate": 5.137386048296129e-05, "loss": 0.0251, "step": 51670 }, { "grad_norm": 0.4484257102012634, "learning_rate": 5.135733195296686e-05, "loss": 0.0408, "step": 51680 }, { "grad_norm": 0.3773409128189087, "learning_rate": 5.134080327453664e-05, "loss": 0.0225, "step": 51690 }, { "grad_norm": 0.3432060182094574, "learning_rate": 5.1324274449478204e-05, "loss": 0.0303, "step": 51700 }, { "grad_norm": 0.6328067779541016, "learning_rate": 5.130774547959909e-05, "loss": 0.0333, "step": 51710 }, { "grad_norm": 0.7995730042457581, "learning_rate": 5.129121636670692e-05, "loss": 0.0291, "step": 51720 }, { "grad_norm": 0.1771034151315689, "learning_rate": 5.1274687112609246e-05, "loss": 0.0214, "step": 51730 }, { "grad_norm": 0.38883692026138306, "learning_rate": 5.125815771911372e-05, "loss": 0.0329, "step": 51740 }, { "grad_norm": 0.3526116609573364, "learning_rate": 5.1241628188027956e-05, "loss": 0.0236, "step": 51750 }, { "grad_norm": 0.28095099329948425, "learning_rate": 5.1225098521159585e-05, "loss": 0.023, "step": 51760 }, { "grad_norm": 0.5281974673271179, "learning_rate": 5.12085687203163e-05, "loss": 0.0344, "step": 51770 }, { "grad_norm": 0.3337174952030182, "learning_rate": 5.119203878730575e-05, "loss": 0.0298, "step": 51780 }, { "grad_norm": 0.43579283356666565, "learning_rate": 5.117550872393564e-05, "loss": 0.0332, "step": 51790 }, { "grad_norm": 0.24172383546829224, "learning_rate": 5.1158978532013645e-05, "loss": 0.0284, "step": 51800 }, { "grad_norm": 0.2511451542377472, "learning_rate": 5.114244821334751e-05, "loss": 0.0305, "step": 51810 }, { "grad_norm": 0.4405096173286438, "learning_rate": 5.112591776974497e-05, "loss": 0.0278, "step": 51820 }, { "grad_norm": 0.18510867655277252, "learning_rate": 5.110938720301374e-05, "loss": 0.0265, "step": 51830 }, { "grad_norm": 0.526923656463623, "learning_rate": 5.1092856514961595e-05, "loss": 0.0285, "step": 51840 }, { "grad_norm": 0.4754348397254944, "learning_rate": 5.107632570739631e-05, "loss": 0.0318, "step": 51850 }, { "grad_norm": 0.3710760474205017, "learning_rate": 5.1059794782125645e-05, "loss": 0.028, "step": 51860 }, { "grad_norm": 0.4125397205352783, "learning_rate": 5.104326374095744e-05, "loss": 0.0352, "step": 51870 }, { "grad_norm": 0.4652840197086334, "learning_rate": 5.102673258569947e-05, "loss": 0.0256, "step": 51880 }, { "grad_norm": 0.4991570711135864, "learning_rate": 5.101020131815956e-05, "loss": 0.0445, "step": 51890 }, { "grad_norm": 0.35196009278297424, "learning_rate": 5.099366994014556e-05, "loss": 0.028, "step": 51900 }, { "grad_norm": 0.4787026047706604, "learning_rate": 5.097713845346532e-05, "loss": 0.0336, "step": 51910 }, { "grad_norm": 0.39897656440734863, "learning_rate": 5.096060685992668e-05, "loss": 0.029, "step": 51920 }, { "grad_norm": 0.49503129720687866, "learning_rate": 5.0944075161337514e-05, "loss": 0.036, "step": 51930 }, { "grad_norm": 0.6985798478126526, "learning_rate": 5.09275433595057e-05, "loss": 0.0308, "step": 51940 }, { "grad_norm": 0.4127313494682312, "learning_rate": 5.0911011456239157e-05, "loss": 0.0294, "step": 51950 }, { "grad_norm": 0.42849597334861755, "learning_rate": 5.089447945334574e-05, "loss": 0.0421, "step": 51960 }, { "grad_norm": 0.9913869500160217, "learning_rate": 5.087794735263343e-05, "loss": 0.0255, "step": 51970 }, { "grad_norm": 0.7946903109550476, "learning_rate": 5.086141515591011e-05, "loss": 0.0327, "step": 51980 }, { "grad_norm": 0.38579896092414856, "learning_rate": 5.084488286498371e-05, "loss": 0.0249, "step": 51990 }, { "grad_norm": 0.4107874929904938, "learning_rate": 5.08283504816622e-05, "loss": 0.0347, "step": 52000 }, { "grad_norm": 0.8069729208946228, "learning_rate": 5.081181800775353e-05, "loss": 0.0341, "step": 52010 }, { "grad_norm": 0.3793395757675171, "learning_rate": 5.079528544506565e-05, "loss": 0.0269, "step": 52020 }, { "grad_norm": 0.27202603220939636, "learning_rate": 5.077875279540657e-05, "loss": 0.0223, "step": 52030 }, { "grad_norm": 0.36125317215919495, "learning_rate": 5.076222006058424e-05, "loss": 0.0345, "step": 52040 }, { "grad_norm": 0.3981855809688568, "learning_rate": 5.074568724240667e-05, "loss": 0.0338, "step": 52050 }, { "grad_norm": 0.4820345640182495, "learning_rate": 5.072915434268187e-05, "loss": 0.0288, "step": 52060 }, { "grad_norm": 0.5909572839736938, "learning_rate": 5.071262136321785e-05, "loss": 0.0328, "step": 52070 }, { "grad_norm": 0.35376831889152527, "learning_rate": 5.069608830582262e-05, "loss": 0.0322, "step": 52080 }, { "grad_norm": 0.3872132897377014, "learning_rate": 5.067955517230423e-05, "loss": 0.0293, "step": 52090 }, { "grad_norm": 0.5581766963005066, "learning_rate": 5.06630219644707e-05, "loss": 0.0373, "step": 52100 }, { "grad_norm": 0.48435312509536743, "learning_rate": 5.064648868413008e-05, "loss": 0.0301, "step": 52110 }, { "grad_norm": 0.2622029185295105, "learning_rate": 5.062995533309044e-05, "loss": 0.027, "step": 52120 }, { "grad_norm": 0.27880629897117615, "learning_rate": 5.061342191315983e-05, "loss": 0.0324, "step": 52130 }, { "grad_norm": 0.4513026177883148, "learning_rate": 5.0596888426146325e-05, "loss": 0.0261, "step": 52140 }, { "grad_norm": 0.17543306946754456, "learning_rate": 5.058035487385798e-05, "loss": 0.0206, "step": 52150 }, { "grad_norm": 0.21734070777893066, "learning_rate": 5.0563821258102915e-05, "loss": 0.0203, "step": 52160 }, { "grad_norm": 0.25779271125793457, "learning_rate": 5.05472875806892e-05, "loss": 0.0195, "step": 52170 }, { "grad_norm": 0.6274759769439697, "learning_rate": 5.053075384342494e-05, "loss": 0.0259, "step": 52180 }, { "grad_norm": 0.22694182395935059, "learning_rate": 5.0514220048118234e-05, "loss": 0.0251, "step": 52190 }, { "grad_norm": 0.5141454935073853, "learning_rate": 5.049768619657721e-05, "loss": 0.0279, "step": 52200 }, { "grad_norm": 0.2534477412700653, "learning_rate": 5.048115229060996e-05, "loss": 0.0227, "step": 52210 }, { "grad_norm": 0.43440955877304077, "learning_rate": 5.046461833202462e-05, "loss": 0.0282, "step": 52220 }, { "grad_norm": 0.754688560962677, "learning_rate": 5.044808432262932e-05, "loss": 0.0302, "step": 52230 }, { "grad_norm": 0.2947394847869873, "learning_rate": 5.043155026423219e-05, "loss": 0.0315, "step": 52240 }, { "grad_norm": 0.4091012179851532, "learning_rate": 5.041501615864137e-05, "loss": 0.0247, "step": 52250 }, { "grad_norm": 0.22328436374664307, "learning_rate": 5.039848200766502e-05, "loss": 0.0198, "step": 52260 }, { "grad_norm": 0.31636595726013184, "learning_rate": 5.038194781311125e-05, "loss": 0.0184, "step": 52270 }, { "grad_norm": 0.41804441809654236, "learning_rate": 5.0365413576788264e-05, "loss": 0.0303, "step": 52280 }, { "grad_norm": 0.39229243993759155, "learning_rate": 5.0348879300504194e-05, "loss": 0.0346, "step": 52290 }, { "grad_norm": 0.26440224051475525, "learning_rate": 5.033234498606721e-05, "loss": 0.0329, "step": 52300 }, { "grad_norm": 0.2776966094970703, "learning_rate": 5.0315810635285476e-05, "loss": 0.0249, "step": 52310 }, { "grad_norm": 0.8816989660263062, "learning_rate": 5.0299276249967164e-05, "loss": 0.0307, "step": 52320 }, { "grad_norm": 0.37572380900382996, "learning_rate": 5.0282741831920454e-05, "loss": 0.0353, "step": 52330 }, { "grad_norm": 0.48933374881744385, "learning_rate": 5.026620738295351e-05, "loss": 0.0344, "step": 52340 }, { "grad_norm": 0.5350673794746399, "learning_rate": 5.024967290487455e-05, "loss": 0.0269, "step": 52350 }, { "grad_norm": 0.17382967472076416, "learning_rate": 5.023313839949172e-05, "loss": 0.0269, "step": 52360 }, { "grad_norm": 0.33755356073379517, "learning_rate": 5.0216603868613224e-05, "loss": 0.0297, "step": 52370 }, { "grad_norm": 1.0759451389312744, "learning_rate": 5.020006931404726e-05, "loss": 0.0399, "step": 52380 }, { "grad_norm": 0.28762954473495483, "learning_rate": 5.0183534737602014e-05, "loss": 0.0278, "step": 52390 }, { "grad_norm": 0.33449485898017883, "learning_rate": 5.0167000141085686e-05, "loss": 0.0267, "step": 52400 }, { "grad_norm": 0.3733857572078705, "learning_rate": 5.015046552630647e-05, "loss": 0.0253, "step": 52410 }, { "grad_norm": 0.6525874137878418, "learning_rate": 5.013393089507258e-05, "loss": 0.0367, "step": 52420 }, { "grad_norm": 0.48515835404396057, "learning_rate": 5.011739624919219e-05, "loss": 0.0251, "step": 52430 }, { "grad_norm": 0.4316878020763397, "learning_rate": 5.0100861590473545e-05, "loss": 0.0239, "step": 52440 }, { "grad_norm": 0.37372326850891113, "learning_rate": 5.008432692072482e-05, "loss": 0.0301, "step": 52450 }, { "grad_norm": 0.2845245897769928, "learning_rate": 5.006779224175423e-05, "loss": 0.0269, "step": 52460 }, { "grad_norm": 0.4938459098339081, "learning_rate": 5.005125755536998e-05, "loss": 0.0315, "step": 52470 }, { "grad_norm": 0.21389314532279968, "learning_rate": 5.003472286338029e-05, "loss": 0.03, "step": 52480 }, { "grad_norm": 0.24352103471755981, "learning_rate": 5.001818816759335e-05, "loss": 0.0404, "step": 52490 }, { "grad_norm": 0.626021683216095, "learning_rate": 5.000165346981738e-05, "loss": 0.033, "step": 52500 }, { "grad_norm": 0.27423277497291565, "learning_rate": 4.9985118771860595e-05, "loss": 0.0347, "step": 52510 }, { "grad_norm": 0.22570829093456268, "learning_rate": 4.996858407553119e-05, "loss": 0.0277, "step": 52520 }, { "grad_norm": 0.3009047210216522, "learning_rate": 4.99520493826374e-05, "loss": 0.038, "step": 52530 }, { "grad_norm": 0.7081436514854431, "learning_rate": 4.9935514694987384e-05, "loss": 0.0418, "step": 52540 }, { "grad_norm": 0.9430995583534241, "learning_rate": 4.991898001438941e-05, "loss": 0.0336, "step": 52550 }, { "grad_norm": 0.21039225161075592, "learning_rate": 4.990244534265166e-05, "loss": 0.0231, "step": 52560 }, { "grad_norm": 0.25487658381462097, "learning_rate": 4.988591068158234e-05, "loss": 0.0254, "step": 52570 }, { "grad_norm": 0.3982323706150055, "learning_rate": 4.986937603298965e-05, "loss": 0.0327, "step": 52580 }, { "grad_norm": 0.3448614180088043, "learning_rate": 4.985284139868181e-05, "loss": 0.0253, "step": 52590 }, { "grad_norm": 0.22497259080410004, "learning_rate": 4.9836306780467e-05, "loss": 0.0291, "step": 52600 }, { "grad_norm": 0.5217025279998779, "learning_rate": 4.981977218015343e-05, "loss": 0.0261, "step": 52610 }, { "grad_norm": 1.5103286504745483, "learning_rate": 4.980323759954932e-05, "loss": 0.033, "step": 52620 }, { "grad_norm": 0.30256974697113037, "learning_rate": 4.978670304046284e-05, "loss": 0.0229, "step": 52630 }, { "grad_norm": 0.3650570213794708, "learning_rate": 4.9770168504702195e-05, "loss": 0.033, "step": 52640 }, { "grad_norm": 0.3782463073730469, "learning_rate": 4.975363399407559e-05, "loss": 0.0341, "step": 52650 }, { "grad_norm": 0.40875738859176636, "learning_rate": 4.9737099510391194e-05, "loss": 0.0378, "step": 52660 }, { "grad_norm": 0.27302470803260803, "learning_rate": 4.972056505545719e-05, "loss": 0.0313, "step": 52670 }, { "grad_norm": 0.3827291429042816, "learning_rate": 4.9704030631081775e-05, "loss": 0.028, "step": 52680 }, { "grad_norm": 0.21560479700565338, "learning_rate": 4.968749623907313e-05, "loss": 0.0206, "step": 52690 }, { "grad_norm": 0.37602394819259644, "learning_rate": 4.967096188123943e-05, "loss": 0.0343, "step": 52700 }, { "grad_norm": 0.29899007081985474, "learning_rate": 4.965442755938884e-05, "loss": 0.0357, "step": 52710 }, { "grad_norm": 0.2538154721260071, "learning_rate": 4.963789327532954e-05, "loss": 0.0276, "step": 52720 }, { "grad_norm": 0.3427145779132843, "learning_rate": 4.962135903086968e-05, "loss": 0.0254, "step": 52730 }, { "grad_norm": 0.29014572501182556, "learning_rate": 4.960482482781741e-05, "loss": 0.0236, "step": 52740 }, { "grad_norm": 0.3501792848110199, "learning_rate": 4.958829066798092e-05, "loss": 0.0252, "step": 52750 }, { "grad_norm": 0.6245918273925781, "learning_rate": 4.957175655316835e-05, "loss": 0.0375, "step": 52760 }, { "grad_norm": 0.3486614525318146, "learning_rate": 4.9555222485187834e-05, "loss": 0.0383, "step": 52770 }, { "grad_norm": 0.34157705307006836, "learning_rate": 4.953868846584752e-05, "loss": 0.0311, "step": 52780 }, { "grad_norm": 0.23039181530475616, "learning_rate": 4.9522154496955544e-05, "loss": 0.0409, "step": 52790 }, { "grad_norm": 1.2667008638381958, "learning_rate": 4.950562058032001e-05, "loss": 0.0305, "step": 52800 }, { "grad_norm": 0.7370600700378418, "learning_rate": 4.948908671774909e-05, "loss": 0.0344, "step": 52810 }, { "grad_norm": 0.43992137908935547, "learning_rate": 4.947255291105087e-05, "loss": 0.0249, "step": 52820 }, { "grad_norm": 0.32215285301208496, "learning_rate": 4.945601916203348e-05, "loss": 0.0301, "step": 52830 }, { "grad_norm": 0.36888694763183594, "learning_rate": 4.9439485472505016e-05, "loss": 0.032, "step": 52840 }, { "grad_norm": 0.6075964570045471, "learning_rate": 4.942295184427356e-05, "loss": 0.028, "step": 52850 }, { "grad_norm": 0.40018314123153687, "learning_rate": 4.9406418279147234e-05, "loss": 0.0283, "step": 52860 }, { "grad_norm": 1.3332315683364868, "learning_rate": 4.9389884778934116e-05, "loss": 0.0323, "step": 52870 }, { "grad_norm": 0.34473681449890137, "learning_rate": 4.937335134544228e-05, "loss": 0.028, "step": 52880 }, { "grad_norm": 0.43767204880714417, "learning_rate": 4.935681798047979e-05, "loss": 0.0323, "step": 52890 }, { "grad_norm": 0.6573571562767029, "learning_rate": 4.934028468585473e-05, "loss": 0.0291, "step": 52900 }, { "grad_norm": 0.25778910517692566, "learning_rate": 4.932375146337514e-05, "loss": 0.0225, "step": 52910 }, { "grad_norm": 0.26645195484161377, "learning_rate": 4.930721831484908e-05, "loss": 0.0263, "step": 52920 }, { "grad_norm": 0.31460535526275635, "learning_rate": 4.929068524208456e-05, "loss": 0.0289, "step": 52930 }, { "grad_norm": 0.33102163672447205, "learning_rate": 4.927415224688965e-05, "loss": 0.0435, "step": 52940 }, { "grad_norm": 0.3055643141269684, "learning_rate": 4.9257619331072375e-05, "loss": 0.0383, "step": 52950 }, { "grad_norm": 0.40921950340270996, "learning_rate": 4.924108649644072e-05, "loss": 0.0233, "step": 52960 }, { "grad_norm": 0.46464309096336365, "learning_rate": 4.922455374480271e-05, "loss": 0.0337, "step": 52970 }, { "grad_norm": 0.26587149500846863, "learning_rate": 4.920802107796634e-05, "loss": 0.0312, "step": 52980 }, { "grad_norm": 0.44910019636154175, "learning_rate": 4.9191488497739564e-05, "loss": 0.0227, "step": 52990 }, { "grad_norm": 0.2743675708770752, "learning_rate": 4.917495600593042e-05, "loss": 0.0229, "step": 53000 }, { "grad_norm": 0.31064996123313904, "learning_rate": 4.915842360434684e-05, "loss": 0.0267, "step": 53010 }, { "grad_norm": 0.7242295145988464, "learning_rate": 4.914189129479679e-05, "loss": 0.039, "step": 53020 }, { "grad_norm": 0.43311241269111633, "learning_rate": 4.912535907908822e-05, "loss": 0.0247, "step": 53030 }, { "grad_norm": 0.5018619298934937, "learning_rate": 4.910882695902907e-05, "loss": 0.0346, "step": 53040 }, { "grad_norm": 0.20388086140155792, "learning_rate": 4.9092294936427234e-05, "loss": 0.0343, "step": 53050 }, { "grad_norm": 0.5017306804656982, "learning_rate": 4.9075763013090685e-05, "loss": 0.0365, "step": 53060 }, { "grad_norm": 0.4541921317577362, "learning_rate": 4.9059231190827305e-05, "loss": 0.0376, "step": 53070 }, { "grad_norm": 0.27965760231018066, "learning_rate": 4.904269947144499e-05, "loss": 0.0438, "step": 53080 }, { "grad_norm": 0.38762086629867554, "learning_rate": 4.902616785675161e-05, "loss": 0.027, "step": 53090 }, { "grad_norm": 0.48353463411331177, "learning_rate": 4.900963634855505e-05, "loss": 0.0285, "step": 53100 }, { "grad_norm": 0.3804241418838501, "learning_rate": 4.8993104948663173e-05, "loss": 0.0291, "step": 53110 }, { "grad_norm": 0.25605204701423645, "learning_rate": 4.8976573658883823e-05, "loss": 0.0274, "step": 53120 }, { "grad_norm": 0.3110620379447937, "learning_rate": 4.896004248102484e-05, "loss": 0.0297, "step": 53130 }, { "grad_norm": 0.2822604477405548, "learning_rate": 4.8943511416894045e-05, "loss": 0.0191, "step": 53140 }, { "grad_norm": 1.0272753238677979, "learning_rate": 4.892698046829926e-05, "loss": 0.0321, "step": 53150 }, { "grad_norm": 0.2569963335990906, "learning_rate": 4.8910449637048275e-05, "loss": 0.0342, "step": 53160 }, { "grad_norm": 0.558353066444397, "learning_rate": 4.8893918924948884e-05, "loss": 0.0341, "step": 53170 }, { "grad_norm": 0.4896988868713379, "learning_rate": 4.8877388333808835e-05, "loss": 0.0328, "step": 53180 }, { "grad_norm": 0.51132732629776, "learning_rate": 4.8860857865435925e-05, "loss": 0.0305, "step": 53190 }, { "grad_norm": 0.57176274061203, "learning_rate": 4.884432752163789e-05, "loss": 0.0377, "step": 53200 }, { "grad_norm": 0.4660382866859436, "learning_rate": 4.8827797304222466e-05, "loss": 0.0266, "step": 53210 }, { "grad_norm": 0.6373213529586792, "learning_rate": 4.8811267214997366e-05, "loss": 0.026, "step": 53220 }, { "grad_norm": 0.4447363615036011, "learning_rate": 4.879473725577029e-05, "loss": 0.0272, "step": 53230 }, { "grad_norm": 0.3773077428340912, "learning_rate": 4.877820742834891e-05, "loss": 0.0264, "step": 53240 }, { "grad_norm": 0.39459148049354553, "learning_rate": 4.8761677734540956e-05, "loss": 0.0234, "step": 53250 }, { "grad_norm": 0.43905654549598694, "learning_rate": 4.874514817615407e-05, "loss": 0.0245, "step": 53260 }, { "grad_norm": 0.6925960183143616, "learning_rate": 4.872861875499588e-05, "loss": 0.0237, "step": 53270 }, { "grad_norm": 0.4214797019958496, "learning_rate": 4.871208947287404e-05, "loss": 0.0306, "step": 53280 }, { "grad_norm": 0.38336214423179626, "learning_rate": 4.869556033159615e-05, "loss": 0.0282, "step": 53290 }, { "grad_norm": 0.537473201751709, "learning_rate": 4.8679031332969835e-05, "loss": 0.0305, "step": 53300 }, { "grad_norm": 0.37207192182540894, "learning_rate": 4.8662502478802635e-05, "loss": 0.0338, "step": 53310 }, { "grad_norm": 0.44980230927467346, "learning_rate": 4.8645973770902176e-05, "loss": 0.0267, "step": 53320 }, { "grad_norm": 0.37259241938591003, "learning_rate": 4.862944521107599e-05, "loss": 0.0246, "step": 53330 }, { "grad_norm": 0.937040388584137, "learning_rate": 4.861291680113162e-05, "loss": 0.0304, "step": 53340 }, { "grad_norm": 0.45289963483810425, "learning_rate": 4.8596388542876563e-05, "loss": 0.0262, "step": 53350 }, { "grad_norm": 0.4396633505821228, "learning_rate": 4.857986043811836e-05, "loss": 0.0354, "step": 53360 }, { "grad_norm": 0.3412884771823883, "learning_rate": 4.856333248866448e-05, "loss": 0.0244, "step": 53370 }, { "grad_norm": 0.2970402240753174, "learning_rate": 4.85468046963224e-05, "loss": 0.0264, "step": 53380 }, { "grad_norm": 0.7093074917793274, "learning_rate": 4.853027706289957e-05, "loss": 0.0277, "step": 53390 }, { "grad_norm": 0.3523632287979126, "learning_rate": 4.851374959020344e-05, "loss": 0.023, "step": 53400 }, { "grad_norm": 0.481272429227829, "learning_rate": 4.849722228004141e-05, "loss": 0.0333, "step": 53410 }, { "grad_norm": 0.38179177045822144, "learning_rate": 4.848069513422091e-05, "loss": 0.0222, "step": 53420 }, { "grad_norm": 0.24766339361667633, "learning_rate": 4.8464168154549294e-05, "loss": 0.0238, "step": 53430 }, { "grad_norm": 0.27460137009620667, "learning_rate": 4.844764134283392e-05, "loss": 0.0312, "step": 53440 }, { "grad_norm": 0.2584744989871979, "learning_rate": 4.843111470088218e-05, "loss": 0.0272, "step": 53450 }, { "grad_norm": 0.29720956087112427, "learning_rate": 4.841458823050137e-05, "loss": 0.0369, "step": 53460 }, { "grad_norm": 0.3407325744628906, "learning_rate": 4.8398061933498816e-05, "loss": 0.0229, "step": 53470 }, { "grad_norm": 0.39972034096717834, "learning_rate": 4.83815358116818e-05, "loss": 0.0214, "step": 53480 }, { "grad_norm": 0.19527021050453186, "learning_rate": 4.8365009866857584e-05, "loss": 0.0294, "step": 53490 }, { "grad_norm": 0.2278662919998169, "learning_rate": 4.834848410083342e-05, "loss": 0.0259, "step": 53500 }, { "grad_norm": 0.6456232070922852, "learning_rate": 4.833195851541657e-05, "loss": 0.0383, "step": 53510 }, { "grad_norm": 0.7928399443626404, "learning_rate": 4.831543311241423e-05, "loss": 0.0243, "step": 53520 }, { "grad_norm": 0.6702359318733215, "learning_rate": 4.829890789363359e-05, "loss": 0.0339, "step": 53530 }, { "grad_norm": 0.39216551184654236, "learning_rate": 4.828238286088182e-05, "loss": 0.0213, "step": 53540 }, { "grad_norm": 0.30709272623062134, "learning_rate": 4.826585801596609e-05, "loss": 0.0328, "step": 53550 }, { "grad_norm": 0.509445309638977, "learning_rate": 4.8249333360693486e-05, "loss": 0.0251, "step": 53560 }, { "grad_norm": 0.4232875108718872, "learning_rate": 4.8232808896871185e-05, "loss": 0.0416, "step": 53570 }, { "grad_norm": 0.44188088178634644, "learning_rate": 4.821628462630624e-05, "loss": 0.0302, "step": 53580 }, { "grad_norm": 0.35204023122787476, "learning_rate": 4.8199760550805735e-05, "loss": 0.0321, "step": 53590 }, { "grad_norm": 0.5153380632400513, "learning_rate": 4.818323667217669e-05, "loss": 0.0329, "step": 53600 }, { "grad_norm": 0.4371042251586914, "learning_rate": 4.8166712992226165e-05, "loss": 0.0288, "step": 53610 }, { "grad_norm": 0.3444863259792328, "learning_rate": 4.8150189512761154e-05, "loss": 0.0267, "step": 53620 }, { "grad_norm": 0.6271577477455139, "learning_rate": 4.813366623558862e-05, "loss": 0.0306, "step": 53630 }, { "grad_norm": 0.39907607436180115, "learning_rate": 4.811714316251554e-05, "loss": 0.0278, "step": 53640 }, { "grad_norm": 0.8836687207221985, "learning_rate": 4.810062029534886e-05, "loss": 0.0321, "step": 53650 }, { "grad_norm": 0.441487580537796, "learning_rate": 4.8084097635895505e-05, "loss": 0.0341, "step": 53660 }, { "grad_norm": 0.3577110171318054, "learning_rate": 4.8067575185962335e-05, "loss": 0.0295, "step": 53670 }, { "grad_norm": 0.2937314510345459, "learning_rate": 4.805105294735625e-05, "loss": 0.0357, "step": 53680 }, { "grad_norm": 0.23495303094387054, "learning_rate": 4.803453092188406e-05, "loss": 0.0337, "step": 53690 }, { "grad_norm": 0.40736716985702515, "learning_rate": 4.801800911135263e-05, "loss": 0.0211, "step": 53700 }, { "grad_norm": 0.3462546467781067, "learning_rate": 4.8001487517568744e-05, "loss": 0.0273, "step": 53710 }, { "grad_norm": 0.5931761264801025, "learning_rate": 4.798496614233918e-05, "loss": 0.0345, "step": 53720 }, { "grad_norm": 0.5133602023124695, "learning_rate": 4.796844498747069e-05, "loss": 0.0382, "step": 53730 }, { "grad_norm": 0.3305034041404724, "learning_rate": 4.795192405477e-05, "loss": 0.0284, "step": 53740 }, { "grad_norm": 0.4511057138442993, "learning_rate": 4.79354033460438e-05, "loss": 0.0247, "step": 53750 }, { "grad_norm": 0.3935585021972656, "learning_rate": 4.79188828630988e-05, "loss": 0.0394, "step": 53760 }, { "grad_norm": 0.3819066286087036, "learning_rate": 4.790236260774165e-05, "loss": 0.0232, "step": 53770 }, { "grad_norm": 0.39719852805137634, "learning_rate": 4.788584258177896e-05, "loss": 0.0201, "step": 53780 }, { "grad_norm": 0.38283026218414307, "learning_rate": 4.7869322787017356e-05, "loss": 0.0296, "step": 53790 }, { "grad_norm": 0.2535897195339203, "learning_rate": 4.78528032252634e-05, "loss": 0.021, "step": 53800 }, { "grad_norm": 0.1772453635931015, "learning_rate": 4.783628389832366e-05, "loss": 0.0277, "step": 53810 }, { "grad_norm": 0.3165433704853058, "learning_rate": 4.7819764808004644e-05, "loss": 0.0282, "step": 53820 }, { "grad_norm": 0.3072853088378906, "learning_rate": 4.780324595611289e-05, "loss": 0.0237, "step": 53830 }, { "grad_norm": 0.5964232087135315, "learning_rate": 4.7786727344454854e-05, "loss": 0.0366, "step": 53840 }, { "grad_norm": 0.27821627259254456, "learning_rate": 4.777020897483697e-05, "loss": 0.0325, "step": 53850 }, { "grad_norm": 0.4493975341320038, "learning_rate": 4.775369084906569e-05, "loss": 0.0363, "step": 53860 }, { "grad_norm": 0.3944585621356964, "learning_rate": 4.773717296894742e-05, "loss": 0.039, "step": 53870 }, { "grad_norm": 0.2483978569507599, "learning_rate": 4.772065533628848e-05, "loss": 0.0285, "step": 53880 }, { "grad_norm": 0.3861068785190582, "learning_rate": 4.7704137952895265e-05, "loss": 0.0228, "step": 53890 }, { "grad_norm": 0.35330337285995483, "learning_rate": 4.768762082057407e-05, "loss": 0.0302, "step": 53900 }, { "grad_norm": 0.6343773007392883, "learning_rate": 4.7671103941131186e-05, "loss": 0.0325, "step": 53910 }, { "grad_norm": 0.40074288845062256, "learning_rate": 4.765458731637287e-05, "loss": 0.025, "step": 53920 }, { "grad_norm": 0.3778366148471832, "learning_rate": 4.763807094810537e-05, "loss": 0.042, "step": 53930 }, { "grad_norm": 0.29394277930259705, "learning_rate": 4.762155483813487e-05, "loss": 0.0349, "step": 53940 }, { "grad_norm": 0.7143939137458801, "learning_rate": 4.760503898826754e-05, "loss": 0.0379, "step": 53950 }, { "grad_norm": 0.267439067363739, "learning_rate": 4.758852340030957e-05, "loss": 0.0343, "step": 53960 }, { "grad_norm": 0.43076443672180176, "learning_rate": 4.757200807606705e-05, "loss": 0.0346, "step": 53970 }, { "grad_norm": 0.381516695022583, "learning_rate": 4.755549301734607e-05, "loss": 0.0278, "step": 53980 }, { "grad_norm": 0.47972866892814636, "learning_rate": 4.753897822595271e-05, "loss": 0.0409, "step": 53990 }, { "grad_norm": 0.5561927556991577, "learning_rate": 4.7522463703692994e-05, "loss": 0.026, "step": 54000 }, { "grad_norm": 0.35077592730522156, "learning_rate": 4.75059494523729e-05, "loss": 0.0286, "step": 54010 }, { "grad_norm": 0.5557777285575867, "learning_rate": 4.748943547379843e-05, "loss": 0.034, "step": 54020 }, { "grad_norm": 0.3745459020137787, "learning_rate": 4.7472921769775535e-05, "loss": 0.0485, "step": 54030 }, { "grad_norm": 0.46246984601020813, "learning_rate": 4.745640834211011e-05, "loss": 0.0334, "step": 54040 }, { "grad_norm": 0.3719843327999115, "learning_rate": 4.7439895192608054e-05, "loss": 0.027, "step": 54050 }, { "grad_norm": 0.5584675073623657, "learning_rate": 4.7423382323075186e-05, "loss": 0.0348, "step": 54060 }, { "grad_norm": 0.5345647931098938, "learning_rate": 4.740686973531737e-05, "loss": 0.0274, "step": 54070 }, { "grad_norm": 0.4231966435909271, "learning_rate": 4.7390357431140375e-05, "loss": 0.0185, "step": 54080 }, { "grad_norm": 0.4147888720035553, "learning_rate": 4.737384541234998e-05, "loss": 0.0269, "step": 54090 }, { "grad_norm": 0.3352668881416321, "learning_rate": 4.735733368075189e-05, "loss": 0.0217, "step": 54100 }, { "grad_norm": 0.26630181074142456, "learning_rate": 4.734082223815182e-05, "loss": 0.0263, "step": 54110 }, { "grad_norm": 0.24084784090518951, "learning_rate": 4.7324311086355436e-05, "loss": 0.0281, "step": 54120 }, { "grad_norm": 0.3687857985496521, "learning_rate": 4.730780022716836e-05, "loss": 0.038, "step": 54130 }, { "grad_norm": 0.37592747807502747, "learning_rate": 4.729128966239619e-05, "loss": 0.0245, "step": 54140 }, { "grad_norm": 0.711786687374115, "learning_rate": 4.7274779393844536e-05, "loss": 0.0272, "step": 54150 }, { "grad_norm": 0.6102073192596436, "learning_rate": 4.725826942331891e-05, "loss": 0.0458, "step": 54160 }, { "grad_norm": 0.68646240234375, "learning_rate": 4.724175975262481e-05, "loss": 0.0298, "step": 54170 }, { "grad_norm": 0.26369810104370117, "learning_rate": 4.722525038356772e-05, "loss": 0.0296, "step": 54180 }, { "grad_norm": 0.32259514927864075, "learning_rate": 4.7208741317953075e-05, "loss": 0.0291, "step": 54190 }, { "grad_norm": 0.30249035358428955, "learning_rate": 4.7192232557586264e-05, "loss": 0.0374, "step": 54200 }, { "grad_norm": 0.5980208516120911, "learning_rate": 4.717572410427269e-05, "loss": 0.0331, "step": 54210 }, { "grad_norm": 0.35591715574264526, "learning_rate": 4.71592159598177e-05, "loss": 0.0293, "step": 54220 }, { "grad_norm": 0.35764390230178833, "learning_rate": 4.714270812602657e-05, "loss": 0.031, "step": 54230 }, { "grad_norm": 0.43751615285873413, "learning_rate": 4.712620060470458e-05, "loss": 0.0348, "step": 54240 }, { "grad_norm": 0.651049792766571, "learning_rate": 4.7109693397656986e-05, "loss": 0.0294, "step": 54250 }, { "grad_norm": 0.3583420217037201, "learning_rate": 4.709318650668894e-05, "loss": 0.0224, "step": 54260 }, { "grad_norm": 0.3665618896484375, "learning_rate": 4.7076679933605676e-05, "loss": 0.0279, "step": 54270 }, { "grad_norm": 0.45712605118751526, "learning_rate": 4.70601736802123e-05, "loss": 0.0264, "step": 54280 }, { "grad_norm": 0.2735292613506317, "learning_rate": 4.704366774831391e-05, "loss": 0.0387, "step": 54290 }, { "grad_norm": 0.6054320335388184, "learning_rate": 4.702716213971557e-05, "loss": 0.0268, "step": 54300 }, { "grad_norm": 0.547607421875, "learning_rate": 4.70106568562223e-05, "loss": 0.0381, "step": 54310 }, { "grad_norm": 0.2702026069164276, "learning_rate": 4.6994151899639105e-05, "loss": 0.0304, "step": 54320 }, { "grad_norm": 0.1818232387304306, "learning_rate": 4.697764727177093e-05, "loss": 0.0333, "step": 54330 }, { "grad_norm": 0.45880651473999023, "learning_rate": 4.696114297442272e-05, "loss": 0.0215, "step": 54340 }, { "grad_norm": 0.27025482058525085, "learning_rate": 4.694463900939933e-05, "loss": 0.0224, "step": 54350 }, { "grad_norm": 0.326688677072525, "learning_rate": 4.692813537850564e-05, "loss": 0.0344, "step": 54360 }, { "grad_norm": 0.8867314457893372, "learning_rate": 4.6911632083546445e-05, "loss": 0.0302, "step": 54370 }, { "grad_norm": 0.38542863726615906, "learning_rate": 4.6895129126326525e-05, "loss": 0.0405, "step": 54380 }, { "grad_norm": 0.6239286661148071, "learning_rate": 4.687862650865059e-05, "loss": 0.0249, "step": 54390 }, { "grad_norm": 0.318293035030365, "learning_rate": 4.686212423232339e-05, "loss": 0.0323, "step": 54400 }, { "grad_norm": 0.30857688188552856, "learning_rate": 4.6845622299149565e-05, "loss": 0.0234, "step": 54410 }, { "grad_norm": 0.4092552065849304, "learning_rate": 4.682912071093374e-05, "loss": 0.0249, "step": 54420 }, { "grad_norm": 0.4393930435180664, "learning_rate": 4.681261946948052e-05, "loss": 0.0247, "step": 54430 }, { "grad_norm": 0.14411340653896332, "learning_rate": 4.679611857659443e-05, "loss": 0.0371, "step": 54440 }, { "grad_norm": 0.30927130579948425, "learning_rate": 4.6779618034079975e-05, "loss": 0.0352, "step": 54450 }, { "grad_norm": 0.3138860762119293, "learning_rate": 4.6763117843741675e-05, "loss": 0.0293, "step": 54460 }, { "grad_norm": 0.36083367466926575, "learning_rate": 4.674661800738394e-05, "loss": 0.0217, "step": 54470 }, { "grad_norm": 0.4856441617012024, "learning_rate": 4.673011852681117e-05, "loss": 0.0261, "step": 54480 }, { "grad_norm": 0.22082176804542542, "learning_rate": 4.671361940382772e-05, "loss": 0.0294, "step": 54490 }, { "grad_norm": 0.37781432271003723, "learning_rate": 4.669712064023791e-05, "loss": 0.0249, "step": 54500 }, { "grad_norm": 0.4703184962272644, "learning_rate": 4.668062223784602e-05, "loss": 0.0277, "step": 54510 }, { "grad_norm": 0.5765159726142883, "learning_rate": 4.6664124198456274e-05, "loss": 0.0279, "step": 54520 }, { "grad_norm": 0.41923853754997253, "learning_rate": 4.6647626523872905e-05, "loss": 0.03, "step": 54530 }, { "grad_norm": 0.5256839394569397, "learning_rate": 4.6631129215900066e-05, "loss": 0.0297, "step": 54540 }, { "grad_norm": 0.4816567897796631, "learning_rate": 4.661463227634187e-05, "loss": 0.037, "step": 54550 }, { "grad_norm": 0.2711130380630493, "learning_rate": 4.659813570700238e-05, "loss": 0.0251, "step": 54560 }, { "grad_norm": 0.46467435359954834, "learning_rate": 4.658163950968568e-05, "loss": 0.0353, "step": 54570 }, { "grad_norm": 0.5237452387809753, "learning_rate": 4.6565143686195726e-05, "loss": 0.0305, "step": 54580 }, { "grad_norm": 0.27883413434028625, "learning_rate": 4.654864823833649e-05, "loss": 0.0255, "step": 54590 }, { "grad_norm": 0.45590028166770935, "learning_rate": 4.65321531679119e-05, "loss": 0.0244, "step": 54600 }, { "grad_norm": 0.3285614550113678, "learning_rate": 4.6515658476725834e-05, "loss": 0.0394, "step": 54610 }, { "grad_norm": 0.3180742561817169, "learning_rate": 4.649916416658212e-05, "loss": 0.0348, "step": 54620 }, { "grad_norm": 0.28733471035957336, "learning_rate": 4.648267023928455e-05, "loss": 0.0315, "step": 54630 }, { "grad_norm": 0.3549782931804657, "learning_rate": 4.646617669663687e-05, "loss": 0.0238, "step": 54640 }, { "grad_norm": 0.43168750405311584, "learning_rate": 4.644968354044277e-05, "loss": 0.0273, "step": 54650 }, { "grad_norm": 1.1161586046218872, "learning_rate": 4.643319077250596e-05, "loss": 0.0323, "step": 54660 }, { "grad_norm": 0.5770831108093262, "learning_rate": 4.6416698394630055e-05, "loss": 0.0319, "step": 54670 }, { "grad_norm": 0.3363921642303467, "learning_rate": 4.640020640861862e-05, "loss": 0.0259, "step": 54680 }, { "grad_norm": 0.3795006275177002, "learning_rate": 4.6383714816275195e-05, "loss": 0.0306, "step": 54690 }, { "grad_norm": 0.3911392092704773, "learning_rate": 4.636722361940329e-05, "loss": 0.0243, "step": 54700 }, { "grad_norm": 0.6870846152305603, "learning_rate": 4.635073281980632e-05, "loss": 0.0263, "step": 54710 }, { "grad_norm": 0.3045141100883484, "learning_rate": 4.633424241928775e-05, "loss": 0.0339, "step": 54720 }, { "grad_norm": 0.3520439863204956, "learning_rate": 4.6317752419650914e-05, "loss": 0.0252, "step": 54730 }, { "grad_norm": 0.24685797095298767, "learning_rate": 4.630126282269914e-05, "loss": 0.0333, "step": 54740 }, { "grad_norm": 0.45447248220443726, "learning_rate": 4.628477363023569e-05, "loss": 0.0232, "step": 54750 }, { "grad_norm": 0.41704505681991577, "learning_rate": 4.626828484406383e-05, "loss": 0.023, "step": 54760 }, { "grad_norm": 0.4233444929122925, "learning_rate": 4.625179646598669e-05, "loss": 0.0319, "step": 54770 }, { "grad_norm": 0.3885844349861145, "learning_rate": 4.6235308497807487e-05, "loss": 0.0205, "step": 54780 }, { "grad_norm": 1.0077202320098877, "learning_rate": 4.6218820941329276e-05, "loss": 0.028, "step": 54790 }, { "grad_norm": 0.2637452483177185, "learning_rate": 4.620233379835513e-05, "loss": 0.0205, "step": 54800 }, { "grad_norm": 0.30315712094306946, "learning_rate": 4.618584707068804e-05, "loss": 0.0345, "step": 54810 }, { "grad_norm": 0.4273223876953125, "learning_rate": 4.616936076013099e-05, "loss": 0.0345, "step": 54820 }, { "grad_norm": 0.346890926361084, "learning_rate": 4.615287486848688e-05, "loss": 0.0283, "step": 54830 }, { "grad_norm": 0.43435734510421753, "learning_rate": 4.6136389397558584e-05, "loss": 0.029, "step": 54840 }, { "grad_norm": 0.2960573434829712, "learning_rate": 4.6119904349148935e-05, "loss": 0.0206, "step": 54850 }, { "grad_norm": 0.6504039168357849, "learning_rate": 4.6103419725060726e-05, "loss": 0.028, "step": 54860 }, { "grad_norm": 0.43345099687576294, "learning_rate": 4.608693552709667e-05, "loss": 0.0297, "step": 54870 }, { "grad_norm": 0.4360215961933136, "learning_rate": 4.607045175705947e-05, "loss": 0.0374, "step": 54880 }, { "grad_norm": 0.6269816756248474, "learning_rate": 4.605396841675176e-05, "loss": 0.0352, "step": 54890 }, { "grad_norm": 0.3005939722061157, "learning_rate": 4.6037485507976094e-05, "loss": 0.0307, "step": 54900 }, { "grad_norm": 0.4484265446662903, "learning_rate": 4.602100303253509e-05, "loss": 0.0186, "step": 54910 }, { "grad_norm": 0.21464911103248596, "learning_rate": 4.600452099223122e-05, "loss": 0.0385, "step": 54920 }, { "grad_norm": 0.3421194851398468, "learning_rate": 4.598803938886691e-05, "loss": 0.0212, "step": 54930 }, { "grad_norm": 0.25524601340293884, "learning_rate": 4.597155822424459e-05, "loss": 0.0244, "step": 54940 }, { "grad_norm": 0.3042580783367157, "learning_rate": 4.59550775001666e-05, "loss": 0.0291, "step": 54950 }, { "grad_norm": 0.3201062083244324, "learning_rate": 4.5938597218435234e-05, "loss": 0.0326, "step": 54960 }, { "grad_norm": 0.42699480056762695, "learning_rate": 4.592211738085279e-05, "loss": 0.0342, "step": 54970 }, { "grad_norm": 0.2390858381986618, "learning_rate": 4.590563798922146e-05, "loss": 0.0276, "step": 54980 }, { "grad_norm": 0.39970511198043823, "learning_rate": 4.5889159045343404e-05, "loss": 0.0291, "step": 54990 }, { "grad_norm": 0.29331812262535095, "learning_rate": 4.5872680551020734e-05, "loss": 0.0287, "step": 55000 }, { "grad_norm": 0.3552972674369812, "learning_rate": 4.585620250805552e-05, "loss": 0.0246, "step": 55010 }, { "grad_norm": 0.3204435110092163, "learning_rate": 4.583972491824975e-05, "loss": 0.0218, "step": 55020 }, { "grad_norm": 0.33223819732666016, "learning_rate": 4.582324778340541e-05, "loss": 0.0282, "step": 55030 }, { "grad_norm": 0.38727182149887085, "learning_rate": 4.5806771105324424e-05, "loss": 0.037, "step": 55040 }, { "grad_norm": 0.3469320237636566, "learning_rate": 4.579029488580865e-05, "loss": 0.0279, "step": 55050 }, { "grad_norm": 0.3237510919570923, "learning_rate": 4.5773819126659876e-05, "loss": 0.0332, "step": 55060 }, { "grad_norm": 0.28739678859710693, "learning_rate": 4.575734382967991e-05, "loss": 0.0237, "step": 55070 }, { "grad_norm": 0.5295168161392212, "learning_rate": 4.574086899667044e-05, "loss": 0.0291, "step": 55080 }, { "grad_norm": 0.5544783473014832, "learning_rate": 4.5724394629433106e-05, "loss": 0.0245, "step": 55090 }, { "grad_norm": 0.38323941826820374, "learning_rate": 4.570792072976958e-05, "loss": 0.0324, "step": 55100 }, { "grad_norm": 0.2511177062988281, "learning_rate": 4.569144729948138e-05, "loss": 0.0208, "step": 55110 }, { "grad_norm": 0.41753488779067993, "learning_rate": 4.5674974340370033e-05, "loss": 0.0297, "step": 55120 }, { "grad_norm": 0.5703511834144592, "learning_rate": 4.565850185423699e-05, "loss": 0.0342, "step": 55130 }, { "grad_norm": 0.3947920501232147, "learning_rate": 4.564202984288365e-05, "loss": 0.0366, "step": 55140 }, { "grad_norm": 0.5184618234634399, "learning_rate": 4.5625558308111356e-05, "loss": 0.0277, "step": 55150 }, { "grad_norm": 1.1450445652008057, "learning_rate": 4.560908725172145e-05, "loss": 0.029, "step": 55160 }, { "grad_norm": 0.8164766430854797, "learning_rate": 4.559261667551516e-05, "loss": 0.0245, "step": 55170 }, { "grad_norm": 0.35819146037101746, "learning_rate": 4.5576146581293685e-05, "loss": 0.0333, "step": 55180 }, { "grad_norm": 0.37319180369377136, "learning_rate": 4.5559676970858164e-05, "loss": 0.0296, "step": 55190 }, { "grad_norm": 0.5300320386886597, "learning_rate": 4.5543207846009705e-05, "loss": 0.0318, "step": 55200 }, { "grad_norm": 0.7572603821754456, "learning_rate": 4.552673920854933e-05, "loss": 0.0313, "step": 55210 }, { "grad_norm": 0.329375684261322, "learning_rate": 4.5510271060278006e-05, "loss": 0.0316, "step": 55220 }, { "grad_norm": 0.25627222657203674, "learning_rate": 4.5493803402996704e-05, "loss": 0.0395, "step": 55230 }, { "grad_norm": 0.48620420694351196, "learning_rate": 4.5477336238506295e-05, "loss": 0.0226, "step": 55240 }, { "grad_norm": 0.1923062652349472, "learning_rate": 4.546086956860759e-05, "loss": 0.0237, "step": 55250 }, { "grad_norm": 0.2469247728586197, "learning_rate": 4.5444403395101374e-05, "loss": 0.0186, "step": 55260 }, { "grad_norm": 0.4186932146549225, "learning_rate": 4.542793771978834e-05, "loss": 0.0297, "step": 55270 }, { "grad_norm": 0.6540754437446594, "learning_rate": 4.541147254446916e-05, "loss": 0.0321, "step": 55280 }, { "grad_norm": 0.28967416286468506, "learning_rate": 4.539500787094445e-05, "loss": 0.0358, "step": 55290 }, { "grad_norm": 0.3732459843158722, "learning_rate": 4.537854370101475e-05, "loss": 0.0317, "step": 55300 }, { "grad_norm": 1.0006945133209229, "learning_rate": 4.5362080036480566e-05, "loss": 0.0279, "step": 55310 }, { "grad_norm": 0.2768131196498871, "learning_rate": 4.534561687914235e-05, "loss": 0.0278, "step": 55320 }, { "grad_norm": 0.2403191477060318, "learning_rate": 4.5329154230800466e-05, "loss": 0.0275, "step": 55330 }, { "grad_norm": 0.37272149324417114, "learning_rate": 4.531269209325525e-05, "loss": 0.0369, "step": 55340 }, { "grad_norm": 0.45520204305648804, "learning_rate": 4.529623046830696e-05, "loss": 0.0223, "step": 55350 }, { "grad_norm": 0.3401035964488983, "learning_rate": 4.5279769357755866e-05, "loss": 0.0292, "step": 55360 }, { "grad_norm": 0.5913304686546326, "learning_rate": 4.5263308763402084e-05, "loss": 0.0363, "step": 55370 }, { "grad_norm": 0.41846561431884766, "learning_rate": 4.524684868704574e-05, "loss": 0.0314, "step": 55380 }, { "grad_norm": 0.44719409942626953, "learning_rate": 4.523038913048688e-05, "loss": 0.0367, "step": 55390 }, { "grad_norm": 0.3620338439941406, "learning_rate": 4.521393009552549e-05, "loss": 0.0293, "step": 55400 }, { "grad_norm": 0.4867200553417206, "learning_rate": 4.5197471583961484e-05, "loss": 0.0372, "step": 55410 }, { "grad_norm": 0.31389158964157104, "learning_rate": 4.51810135975948e-05, "loss": 0.0295, "step": 55420 }, { "grad_norm": 0.5406965613365173, "learning_rate": 4.516455613822521e-05, "loss": 0.0254, "step": 55430 }, { "grad_norm": 0.20204810798168182, "learning_rate": 4.514809920765249e-05, "loss": 0.0258, "step": 55440 }, { "grad_norm": 0.3971952795982361, "learning_rate": 4.5131642807676356e-05, "loss": 0.0223, "step": 55450 }, { "grad_norm": 0.4788724482059479, "learning_rate": 4.511518694009644e-05, "loss": 0.024, "step": 55460 }, { "grad_norm": 0.21460160613059998, "learning_rate": 4.509873160671231e-05, "loss": 0.0285, "step": 55470 }, { "grad_norm": 1.0113657712936401, "learning_rate": 4.508227680932356e-05, "loss": 0.0229, "step": 55480 }, { "grad_norm": 0.3362411856651306, "learning_rate": 4.5065822549729605e-05, "loss": 0.0296, "step": 55490 }, { "grad_norm": 0.5359392166137695, "learning_rate": 4.504936882972989e-05, "loss": 0.0251, "step": 55500 }, { "grad_norm": 0.273074746131897, "learning_rate": 4.503291565112374e-05, "loss": 0.0249, "step": 55510 }, { "grad_norm": 0.5161107182502747, "learning_rate": 4.501646301571048e-05, "loss": 0.0279, "step": 55520 }, { "grad_norm": 0.26667898893356323, "learning_rate": 4.500001092528933e-05, "loss": 0.0241, "step": 55530 }, { "grad_norm": 0.3148522675037384, "learning_rate": 4.4983559381659455e-05, "loss": 0.0233, "step": 55540 }, { "grad_norm": 0.3209737539291382, "learning_rate": 4.496710838661999e-05, "loss": 0.0258, "step": 55550 }, { "grad_norm": 0.5228027105331421, "learning_rate": 4.495065794196999e-05, "loss": 0.0357, "step": 55560 }, { "grad_norm": 0.381213903427124, "learning_rate": 4.4934208049508455e-05, "loss": 0.0235, "step": 55570 }, { "grad_norm": 0.47292497754096985, "learning_rate": 4.4917758711034303e-05, "loss": 0.0268, "step": 55580 }, { "grad_norm": 0.23092183470726013, "learning_rate": 4.490130992834642e-05, "loss": 0.0292, "step": 55590 }, { "grad_norm": 0.5410224795341492, "learning_rate": 4.488486170324359e-05, "loss": 0.0424, "step": 55600 }, { "grad_norm": 0.3667568266391754, "learning_rate": 4.486841403752463e-05, "loss": 0.0247, "step": 55610 }, { "grad_norm": 0.3975217938423157, "learning_rate": 4.4851966932988174e-05, "loss": 0.0314, "step": 55620 }, { "grad_norm": 0.500153124332428, "learning_rate": 4.4835520391432894e-05, "loss": 0.0246, "step": 55630 }, { "grad_norm": 0.42978131771087646, "learning_rate": 4.481907441465733e-05, "loss": 0.0334, "step": 55640 }, { "grad_norm": 0.30094635486602783, "learning_rate": 4.480262900446e-05, "loss": 0.0249, "step": 55650 }, { "grad_norm": 0.2560845911502838, "learning_rate": 4.478618416263933e-05, "loss": 0.0207, "step": 55660 }, { "grad_norm": 0.24554666876792908, "learning_rate": 4.476973989099373e-05, "loss": 0.022, "step": 55670 }, { "grad_norm": 0.6693117022514343, "learning_rate": 4.475329619132153e-05, "loss": 0.0395, "step": 55680 }, { "grad_norm": 1.9432169198989868, "learning_rate": 4.4736853065420956e-05, "loss": 0.0342, "step": 55690 }, { "grad_norm": 0.35966384410858154, "learning_rate": 4.472041051509023e-05, "loss": 0.0314, "step": 55700 }, { "grad_norm": 0.31234118342399597, "learning_rate": 4.470396854212746e-05, "loss": 0.0285, "step": 55710 }, { "grad_norm": 0.28294578194618225, "learning_rate": 4.468752714833072e-05, "loss": 0.024, "step": 55720 }, { "grad_norm": 0.24813541769981384, "learning_rate": 4.467108633549803e-05, "loss": 0.0216, "step": 55730 }, { "grad_norm": 0.2743713855743408, "learning_rate": 4.4654646105427335e-05, "loss": 0.035, "step": 55740 }, { "grad_norm": 0.31045010685920715, "learning_rate": 4.463820645991651e-05, "loss": 0.0255, "step": 55750 }, { "grad_norm": 0.3568105697631836, "learning_rate": 4.462176740076334e-05, "loss": 0.0226, "step": 55760 }, { "grad_norm": 0.6966755986213684, "learning_rate": 4.4605328929765624e-05, "loss": 0.0344, "step": 55770 }, { "grad_norm": 0.36944636702537537, "learning_rate": 4.458889104872102e-05, "loss": 0.0303, "step": 55780 }, { "grad_norm": 0.486734539270401, "learning_rate": 4.457245375942715e-05, "loss": 0.0336, "step": 55790 }, { "grad_norm": 0.7463470697402954, "learning_rate": 4.455601706368157e-05, "loss": 0.0296, "step": 55800 }, { "grad_norm": 0.38742342591285706, "learning_rate": 4.453958096328179e-05, "loss": 0.0238, "step": 55810 }, { "grad_norm": 0.43405091762542725, "learning_rate": 4.452314546002523e-05, "loss": 0.0252, "step": 55820 }, { "grad_norm": 0.3706890642642975, "learning_rate": 4.4506710555709246e-05, "loss": 0.0244, "step": 55830 }, { "grad_norm": 0.2749883532524109, "learning_rate": 4.449027625213114e-05, "loss": 0.0288, "step": 55840 }, { "grad_norm": 0.9955779314041138, "learning_rate": 4.4473842551088135e-05, "loss": 0.0297, "step": 55850 }, { "grad_norm": 0.24394814670085907, "learning_rate": 4.445740945437738e-05, "loss": 0.0213, "step": 55860 }, { "grad_norm": 0.38532140851020813, "learning_rate": 4.444097696379602e-05, "loss": 0.0332, "step": 55870 }, { "grad_norm": 0.27298688888549805, "learning_rate": 4.442454508114106e-05, "loss": 0.0236, "step": 55880 }, { "grad_norm": 0.2500867545604706, "learning_rate": 4.4408113808209465e-05, "loss": 0.0229, "step": 55890 }, { "grad_norm": 0.3778838813304901, "learning_rate": 4.439168314679813e-05, "loss": 0.0299, "step": 55900 }, { "grad_norm": 0.33067429065704346, "learning_rate": 4.43752530987039e-05, "loss": 0.0351, "step": 55910 }, { "grad_norm": 0.25673234462738037, "learning_rate": 4.4358823665723515e-05, "loss": 0.0197, "step": 55920 }, { "grad_norm": 0.2675057053565979, "learning_rate": 4.434239484965371e-05, "loss": 0.0232, "step": 55930 }, { "grad_norm": 0.1662369668483734, "learning_rate": 4.4325966652291103e-05, "loss": 0.0297, "step": 55940 }, { "grad_norm": 0.5593357682228088, "learning_rate": 4.430953907543225e-05, "loss": 0.0252, "step": 55950 }, { "grad_norm": 0.28898555040359497, "learning_rate": 4.4293112120873645e-05, "loss": 0.0226, "step": 55960 }, { "grad_norm": 0.515130341053009, "learning_rate": 4.427668579041172e-05, "loss": 0.0233, "step": 55970 }, { "grad_norm": 0.4315454065799713, "learning_rate": 4.4260260085842816e-05, "loss": 0.0252, "step": 55980 }, { "grad_norm": 0.71016925573349, "learning_rate": 4.4243835008963256e-05, "loss": 0.028, "step": 55990 }, { "grad_norm": 0.6546411514282227, "learning_rate": 4.422741056156925e-05, "loss": 0.0271, "step": 56000 }, { "grad_norm": 0.2307511270046234, "learning_rate": 4.421098674545693e-05, "loss": 0.0271, "step": 56010 }, { "grad_norm": 0.31354549527168274, "learning_rate": 4.419456356242241e-05, "loss": 0.0339, "step": 56020 }, { "grad_norm": 0.3832899332046509, "learning_rate": 4.4178141014261685e-05, "loss": 0.0287, "step": 56030 }, { "grad_norm": 0.2520264685153961, "learning_rate": 4.416171910277071e-05, "loss": 0.0371, "step": 56040 }, { "grad_norm": 0.5170483589172363, "learning_rate": 4.414529782974533e-05, "loss": 0.0258, "step": 56050 }, { "grad_norm": 0.27244362235069275, "learning_rate": 4.41288771969814e-05, "loss": 0.0338, "step": 56060 }, { "grad_norm": 0.4693492650985718, "learning_rate": 4.411245720627464e-05, "loss": 0.0453, "step": 56070 }, { "grad_norm": 0.17802362143993378, "learning_rate": 4.40960378594207e-05, "loss": 0.0272, "step": 56080 }, { "grad_norm": 0.23633331060409546, "learning_rate": 4.407961915821519e-05, "loss": 0.0371, "step": 56090 }, { "grad_norm": 0.7806608080863953, "learning_rate": 4.406320110445362e-05, "loss": 0.0247, "step": 56100 }, { "grad_norm": 0.2743542492389679, "learning_rate": 4.404678369993144e-05, "loss": 0.0208, "step": 56110 }, { "grad_norm": 0.19181206822395325, "learning_rate": 4.403036694644406e-05, "loss": 0.0286, "step": 56120 }, { "grad_norm": 0.30775389075279236, "learning_rate": 4.4013950845786764e-05, "loss": 0.0237, "step": 56130 }, { "grad_norm": 0.4386823773384094, "learning_rate": 4.399753539975482e-05, "loss": 0.0249, "step": 56140 }, { "grad_norm": 0.6098878383636475, "learning_rate": 4.398112061014337e-05, "loss": 0.0276, "step": 56150 }, { "grad_norm": 0.43676427006721497, "learning_rate": 4.396470647874753e-05, "loss": 0.0258, "step": 56160 }, { "grad_norm": 0.2477753460407257, "learning_rate": 4.394829300736229e-05, "loss": 0.0215, "step": 56170 }, { "grad_norm": 0.3146856129169464, "learning_rate": 4.393188019778265e-05, "loss": 0.0275, "step": 56180 }, { "grad_norm": 0.24056677520275116, "learning_rate": 4.391546805180347e-05, "loss": 0.0218, "step": 56190 }, { "grad_norm": 0.355823814868927, "learning_rate": 4.389905657121955e-05, "loss": 0.019, "step": 56200 }, { "grad_norm": 0.33265525102615356, "learning_rate": 4.3882645757825635e-05, "loss": 0.0325, "step": 56210 }, { "grad_norm": 0.3358009457588196, "learning_rate": 4.386623561341637e-05, "loss": 0.0277, "step": 56220 }, { "grad_norm": 0.7966499924659729, "learning_rate": 4.384982613978637e-05, "loss": 0.0266, "step": 56230 }, { "grad_norm": 0.8322200179100037, "learning_rate": 4.383341733873012e-05, "loss": 0.0253, "step": 56240 }, { "grad_norm": 0.47695139050483704, "learning_rate": 4.381700921204209e-05, "loss": 0.0261, "step": 56250 }, { "grad_norm": 0.4545392692089081, "learning_rate": 4.380060176151663e-05, "loss": 0.0304, "step": 56260 }, { "grad_norm": 0.6411017775535583, "learning_rate": 4.378419498894805e-05, "loss": 0.0273, "step": 56270 }, { "grad_norm": 0.817712664604187, "learning_rate": 4.376778889613056e-05, "loss": 0.0276, "step": 56280 }, { "grad_norm": 1.2391395568847656, "learning_rate": 4.375138348485831e-05, "loss": 0.0362, "step": 56290 }, { "grad_norm": 1.4294395446777344, "learning_rate": 4.3734978756925346e-05, "loss": 0.0243, "step": 56300 }, { "grad_norm": 0.3554597795009613, "learning_rate": 4.3718574714125706e-05, "loss": 0.0288, "step": 56310 }, { "grad_norm": 0.4506315588951111, "learning_rate": 4.370217135825329e-05, "loss": 0.0291, "step": 56320 }, { "grad_norm": 0.24969744682312012, "learning_rate": 4.368576869110194e-05, "loss": 0.0216, "step": 56330 }, { "grad_norm": 0.5066273212432861, "learning_rate": 4.366936671446544e-05, "loss": 0.0216, "step": 56340 }, { "grad_norm": 0.3889021575450897, "learning_rate": 4.3652965430137474e-05, "loss": 0.0327, "step": 56350 }, { "grad_norm": 0.25216326117515564, "learning_rate": 4.3636564839911646e-05, "loss": 0.0264, "step": 56360 }, { "grad_norm": 0.42121368646621704, "learning_rate": 4.3620164945581545e-05, "loss": 0.021, "step": 56370 }, { "grad_norm": 0.22010087966918945, "learning_rate": 4.360376574894061e-05, "loss": 0.0248, "step": 56380 }, { "grad_norm": 0.679410457611084, "learning_rate": 4.358736725178224e-05, "loss": 0.0242, "step": 56390 }, { "grad_norm": 0.5045202970504761, "learning_rate": 4.357096945589974e-05, "loss": 0.0317, "step": 56400 }, { "grad_norm": 0.7174645662307739, "learning_rate": 4.355457236308636e-05, "loss": 0.0253, "step": 56410 }, { "grad_norm": 0.217857226729393, "learning_rate": 4.353817597513526e-05, "loss": 0.0237, "step": 56420 }, { "grad_norm": 0.24457813799381256, "learning_rate": 4.352178029383948e-05, "loss": 0.0207, "step": 56430 }, { "grad_norm": 0.31130868196487427, "learning_rate": 4.3505385320992105e-05, "loss": 0.0274, "step": 56440 }, { "grad_norm": 0.22095033526420593, "learning_rate": 4.348899105838602e-05, "loss": 0.0213, "step": 56450 }, { "grad_norm": 0.197973370552063, "learning_rate": 4.3472597507814087e-05, "loss": 0.0243, "step": 56460 }, { "grad_norm": 0.35517242550849915, "learning_rate": 4.3456204671069066e-05, "loss": 0.0308, "step": 56470 }, { "grad_norm": 0.1495855748653412, "learning_rate": 4.343981254994367e-05, "loss": 0.0234, "step": 56480 }, { "grad_norm": 0.3847281038761139, "learning_rate": 4.34234211462305e-05, "loss": 0.0285, "step": 56490 }, { "grad_norm": 0.35415011644363403, "learning_rate": 4.340703046172213e-05, "loss": 0.0279, "step": 56500 }, { "grad_norm": 0.2614319324493408, "learning_rate": 4.339064049821097e-05, "loss": 0.028, "step": 56510 }, { "grad_norm": 0.3637849986553192, "learning_rate": 4.3374251257489446e-05, "loss": 0.0295, "step": 56520 }, { "grad_norm": 0.32834485173225403, "learning_rate": 4.335786274134984e-05, "loss": 0.0201, "step": 56530 }, { "grad_norm": 0.46596255898475647, "learning_rate": 4.334147495158438e-05, "loss": 0.0379, "step": 56540 }, { "grad_norm": 0.46123746037483215, "learning_rate": 4.33250878899852e-05, "loss": 0.0249, "step": 56550 }, { "grad_norm": 0.5172550678253174, "learning_rate": 4.330870155834437e-05, "loss": 0.0294, "step": 56560 }, { "grad_norm": 0.3886840343475342, "learning_rate": 4.3292315958453884e-05, "loss": 0.0204, "step": 56570 }, { "grad_norm": 0.47933149337768555, "learning_rate": 4.3275931092105646e-05, "loss": 0.0233, "step": 56580 }, { "grad_norm": 0.40490028262138367, "learning_rate": 4.325954696109147e-05, "loss": 0.0218, "step": 56590 }, { "grad_norm": 0.4815916121006012, "learning_rate": 4.32431635672031e-05, "loss": 0.0299, "step": 56600 }, { "grad_norm": 0.33633384108543396, "learning_rate": 4.3226780912232214e-05, "loss": 0.036, "step": 56610 }, { "grad_norm": 0.4248856008052826, "learning_rate": 4.321039899797036e-05, "loss": 0.026, "step": 56620 }, { "grad_norm": 0.2445598542690277, "learning_rate": 4.319401782620908e-05, "loss": 0.0257, "step": 56630 }, { "grad_norm": 0.34284570813179016, "learning_rate": 4.317763739873978e-05, "loss": 0.0222, "step": 56640 }, { "grad_norm": 0.4751773774623871, "learning_rate": 4.31612577173538e-05, "loss": 0.0273, "step": 56650 }, { "grad_norm": 0.27386003732681274, "learning_rate": 4.314487878384239e-05, "loss": 0.0345, "step": 56660 }, { "grad_norm": 0.46879491209983826, "learning_rate": 4.3128500599996736e-05, "loss": 0.0188, "step": 56670 }, { "grad_norm": 0.5247294306755066, "learning_rate": 4.31121231676079e-05, "loss": 0.0172, "step": 56680 }, { "grad_norm": 0.3510391414165497, "learning_rate": 4.309574648846694e-05, "loss": 0.0264, "step": 56690 }, { "grad_norm": 0.43392786383628845, "learning_rate": 4.3079370564364755e-05, "loss": 0.0278, "step": 56700 }, { "grad_norm": 0.663727343082428, "learning_rate": 4.306299539709221e-05, "loss": 0.0238, "step": 56710 }, { "grad_norm": 0.30800777673721313, "learning_rate": 4.3046620988440046e-05, "loss": 0.0255, "step": 56720 }, { "grad_norm": 0.3252830505371094, "learning_rate": 4.303024734019897e-05, "loss": 0.0205, "step": 56730 }, { "grad_norm": 0.5497232675552368, "learning_rate": 4.3013874454159565e-05, "loss": 0.0264, "step": 56740 }, { "grad_norm": 0.21691730618476868, "learning_rate": 4.299750233211233e-05, "loss": 0.0341, "step": 56750 }, { "grad_norm": 0.3034599721431732, "learning_rate": 4.2981130975847715e-05, "loss": 0.0214, "step": 56760 }, { "grad_norm": 0.3090207278728485, "learning_rate": 4.2964760387156076e-05, "loss": 0.0251, "step": 56770 }, { "grad_norm": 0.3031577169895172, "learning_rate": 4.294839056782767e-05, "loss": 0.0233, "step": 56780 }, { "grad_norm": 1.2132833003997803, "learning_rate": 4.293202151965266e-05, "loss": 0.0268, "step": 56790 }, { "grad_norm": 0.2843445837497711, "learning_rate": 4.291565324442115e-05, "loss": 0.0246, "step": 56800 }, { "grad_norm": 0.5267975926399231, "learning_rate": 4.2899285743923135e-05, "loss": 0.0324, "step": 56810 }, { "grad_norm": 0.4191127121448517, "learning_rate": 4.288291901994857e-05, "loss": 0.0207, "step": 56820 }, { "grad_norm": 0.28407129645347595, "learning_rate": 4.286655307428729e-05, "loss": 0.0244, "step": 56830 }, { "grad_norm": 0.3088545501232147, "learning_rate": 4.285018790872903e-05, "loss": 0.0258, "step": 56840 }, { "grad_norm": 0.43602532148361206, "learning_rate": 4.283382352506347e-05, "loss": 0.0331, "step": 56850 }, { "grad_norm": 0.43339666724205017, "learning_rate": 4.2817459925080205e-05, "loss": 0.0335, "step": 56860 }, { "grad_norm": 0.2745773196220398, "learning_rate": 4.28010971105687e-05, "loss": 0.0308, "step": 56870 }, { "grad_norm": 0.6349016427993774, "learning_rate": 4.278473508331841e-05, "loss": 0.0289, "step": 56880 }, { "grad_norm": 0.6458035111427307, "learning_rate": 4.276837384511864e-05, "loss": 0.0362, "step": 56890 }, { "grad_norm": 0.3207070231437683, "learning_rate": 4.275201339775864e-05, "loss": 0.0234, "step": 56900 }, { "grad_norm": 0.39885133504867554, "learning_rate": 4.273565374302756e-05, "loss": 0.0313, "step": 56910 }, { "grad_norm": 0.30277976393699646, "learning_rate": 4.2719294882714454e-05, "loss": 0.027, "step": 56920 }, { "grad_norm": 0.3268052637577057, "learning_rate": 4.270293681860831e-05, "loss": 0.0263, "step": 56930 }, { "grad_norm": 0.3362641930580139, "learning_rate": 4.268657955249802e-05, "loss": 0.0211, "step": 56940 }, { "grad_norm": 0.40416067838668823, "learning_rate": 4.267022308617241e-05, "loss": 0.0257, "step": 56950 }, { "grad_norm": 0.36016955971717834, "learning_rate": 4.265386742142018e-05, "loss": 0.0253, "step": 56960 }, { "grad_norm": 0.31345391273498535, "learning_rate": 4.263751256002996e-05, "loss": 0.0272, "step": 56970 }, { "grad_norm": 0.32288676500320435, "learning_rate": 4.26211585037903e-05, "loss": 0.0239, "step": 56980 }, { "grad_norm": 1.137452483177185, "learning_rate": 4.2604805254489656e-05, "loss": 0.0261, "step": 56990 }, { "grad_norm": 0.32411471009254456, "learning_rate": 4.258845281391638e-05, "loss": 0.0306, "step": 57000 }, { "grad_norm": 0.3477889597415924, "learning_rate": 4.257210118385876e-05, "loss": 0.0367, "step": 57010 }, { "grad_norm": 0.3461518883705139, "learning_rate": 4.2555750366104994e-05, "loss": 0.0252, "step": 57020 }, { "grad_norm": 0.3868712782859802, "learning_rate": 4.2539400362443184e-05, "loss": 0.0261, "step": 57030 }, { "grad_norm": 0.3723822832107544, "learning_rate": 4.252305117466133e-05, "loss": 0.0278, "step": 57040 }, { "grad_norm": 0.42905840277671814, "learning_rate": 4.2506702804547356e-05, "loss": 0.0324, "step": 57050 }, { "grad_norm": 0.34532099962234497, "learning_rate": 4.249035525388907e-05, "loss": 0.0212, "step": 57060 }, { "grad_norm": 0.5376811623573303, "learning_rate": 4.247400852447427e-05, "loss": 0.019, "step": 57070 }, { "grad_norm": 0.32184162735939026, "learning_rate": 4.245766261809059e-05, "loss": 0.0254, "step": 57080 }, { "grad_norm": 0.40617701411247253, "learning_rate": 4.244131753652558e-05, "loss": 0.0312, "step": 57090 }, { "grad_norm": 0.5924246907234192, "learning_rate": 4.242497328156672e-05, "loss": 0.023, "step": 57100 }, { "grad_norm": 0.304699569940567, "learning_rate": 4.240862985500139e-05, "loss": 0.0285, "step": 57110 }, { "grad_norm": 0.28114110231399536, "learning_rate": 4.2392287258616894e-05, "loss": 0.0249, "step": 57120 }, { "grad_norm": 0.3156323730945587, "learning_rate": 4.2375945494200396e-05, "loss": 0.024, "step": 57130 }, { "grad_norm": 0.31699031591415405, "learning_rate": 4.2359604563539054e-05, "loss": 0.023, "step": 57140 }, { "grad_norm": 0.3637801706790924, "learning_rate": 4.2343264468419885e-05, "loss": 0.023, "step": 57150 }, { "grad_norm": 0.3247225880622864, "learning_rate": 4.232692521062979e-05, "loss": 0.021, "step": 57160 }, { "grad_norm": 0.450773686170578, "learning_rate": 4.231058679195562e-05, "loss": 0.0256, "step": 57170 }, { "grad_norm": 0.36647576093673706, "learning_rate": 4.229424921418411e-05, "loss": 0.0218, "step": 57180 }, { "grad_norm": 0.32641130685806274, "learning_rate": 4.2277912479101916e-05, "loss": 0.0265, "step": 57190 }, { "grad_norm": 0.9566419720649719, "learning_rate": 4.226157658849562e-05, "loss": 0.0316, "step": 57200 }, { "grad_norm": 0.33637532591819763, "learning_rate": 4.2245241544151664e-05, "loss": 0.0305, "step": 57210 }, { "grad_norm": 0.4845874309539795, "learning_rate": 4.2228907347856425e-05, "loss": 0.026, "step": 57220 }, { "grad_norm": 0.23125576972961426, "learning_rate": 4.2212574001396214e-05, "loss": 0.0259, "step": 57230 }, { "grad_norm": 0.4842059910297394, "learning_rate": 4.21962415065572e-05, "loss": 0.0357, "step": 57240 }, { "grad_norm": 0.45922335982322693, "learning_rate": 4.217990986512548e-05, "loss": 0.029, "step": 57250 }, { "grad_norm": 0.27912428975105286, "learning_rate": 4.216357907888704e-05, "loss": 0.0334, "step": 57260 }, { "grad_norm": 0.5166895389556885, "learning_rate": 4.2147249149627824e-05, "loss": 0.0296, "step": 57270 }, { "grad_norm": 0.33903345465660095, "learning_rate": 4.213092007913364e-05, "loss": 0.0299, "step": 57280 }, { "grad_norm": 0.44289523363113403, "learning_rate": 4.2114591869190205e-05, "loss": 0.0272, "step": 57290 }, { "grad_norm": 0.420897513628006, "learning_rate": 4.209826452158315e-05, "loss": 0.0288, "step": 57300 }, { "grad_norm": 0.5195884108543396, "learning_rate": 4.2081938038098e-05, "loss": 0.0245, "step": 57310 }, { "grad_norm": 0.2502018213272095, "learning_rate": 4.2065612420520185e-05, "loss": 0.0181, "step": 57320 }, { "grad_norm": 0.2588784992694855, "learning_rate": 4.204928767063509e-05, "loss": 0.0231, "step": 57330 }, { "grad_norm": 0.351768434047699, "learning_rate": 4.2032963790227944e-05, "loss": 0.0291, "step": 57340 }, { "grad_norm": 0.26530152559280396, "learning_rate": 4.2016640781083887e-05, "loss": 0.0243, "step": 57350 }, { "grad_norm": 0.9421210289001465, "learning_rate": 4.2000318644988004e-05, "loss": 0.0435, "step": 57360 }, { "grad_norm": 0.23558907210826874, "learning_rate": 4.198399738372524e-05, "loss": 0.0247, "step": 57370 }, { "grad_norm": 0.322589635848999, "learning_rate": 4.196767699908045e-05, "loss": 0.0235, "step": 57380 }, { "grad_norm": 0.5114098787307739, "learning_rate": 4.195135749283845e-05, "loss": 0.0291, "step": 57390 }, { "grad_norm": 0.3863403797149658, "learning_rate": 4.1935038866783894e-05, "loss": 0.0274, "step": 57400 }, { "grad_norm": 0.5926691293716431, "learning_rate": 4.1918721122701355e-05, "loss": 0.0285, "step": 57410 }, { "grad_norm": 0.4553084373474121, "learning_rate": 4.190240426237533e-05, "loss": 0.026, "step": 57420 }, { "grad_norm": 0.7538052201271057, "learning_rate": 4.188608828759019e-05, "loss": 0.0294, "step": 57430 }, { "grad_norm": 0.26010990142822266, "learning_rate": 4.1869773200130246e-05, "loss": 0.0339, "step": 57440 }, { "grad_norm": 0.3433341383934021, "learning_rate": 4.185345900177967e-05, "loss": 0.0327, "step": 57450 }, { "grad_norm": 0.3819783926010132, "learning_rate": 4.183714569432258e-05, "loss": 0.0256, "step": 57460 }, { "grad_norm": 0.24308645725250244, "learning_rate": 4.182083327954295e-05, "loss": 0.0243, "step": 57470 }, { "grad_norm": 0.3020056188106537, "learning_rate": 4.1804521759224714e-05, "loss": 0.0296, "step": 57480 }, { "grad_norm": 0.40200603008270264, "learning_rate": 4.178821113515165e-05, "loss": 0.0224, "step": 57490 }, { "grad_norm": 0.2555781602859497, "learning_rate": 4.1771901409107476e-05, "loss": 0.0216, "step": 57500 }, { "grad_norm": 0.49784770607948303, "learning_rate": 4.175559258287577e-05, "loss": 0.0303, "step": 57510 }, { "grad_norm": 0.5393710136413574, "learning_rate": 4.173928465824008e-05, "loss": 0.0253, "step": 57520 }, { "grad_norm": 0.4095311760902405, "learning_rate": 4.17229776369838e-05, "loss": 0.0316, "step": 57530 }, { "grad_norm": 0.23480692505836487, "learning_rate": 4.170667152089025e-05, "loss": 0.0299, "step": 57540 }, { "grad_norm": 0.20640628039836884, "learning_rate": 4.1690366311742626e-05, "loss": 0.0178, "step": 57550 }, { "grad_norm": 0.4360080659389496, "learning_rate": 4.1674062011324055e-05, "loss": 0.025, "step": 57560 }, { "grad_norm": 0.32228097319602966, "learning_rate": 4.165775862141752e-05, "loss": 0.0213, "step": 57570 }, { "grad_norm": 0.2977370023727417, "learning_rate": 4.164145614380598e-05, "loss": 0.0275, "step": 57580 }, { "grad_norm": 0.28889137506484985, "learning_rate": 4.162515458027223e-05, "loss": 0.0308, "step": 57590 }, { "grad_norm": 0.39109575748443604, "learning_rate": 4.160885393259899e-05, "loss": 0.0339, "step": 57600 }, { "grad_norm": 0.34372207522392273, "learning_rate": 4.159255420256886e-05, "loss": 0.0269, "step": 57610 }, { "grad_norm": 0.4091742932796478, "learning_rate": 4.157625539196436e-05, "loss": 0.0291, "step": 57620 }, { "grad_norm": 0.9768387079238892, "learning_rate": 4.155995750256791e-05, "loss": 0.0509, "step": 57630 }, { "grad_norm": 0.41037502884864807, "learning_rate": 4.15436605361618e-05, "loss": 0.0284, "step": 57640 }, { "grad_norm": 0.37744611501693726, "learning_rate": 4.152736449452827e-05, "loss": 0.0297, "step": 57650 }, { "grad_norm": 0.7048109173774719, "learning_rate": 4.151106937944943e-05, "loss": 0.0287, "step": 57660 }, { "grad_norm": 0.20534588396549225, "learning_rate": 4.149477519270727e-05, "loss": 0.0275, "step": 57670 }, { "grad_norm": 0.5177218914031982, "learning_rate": 4.147848193608371e-05, "loss": 0.0187, "step": 57680 }, { "grad_norm": 0.3005872368812561, "learning_rate": 4.146218961136056e-05, "loss": 0.0237, "step": 57690 }, { "grad_norm": 0.21760860085487366, "learning_rate": 4.14458982203195e-05, "loss": 0.0232, "step": 57700 }, { "grad_norm": 0.35259175300598145, "learning_rate": 4.1429607764742165e-05, "loss": 0.0314, "step": 57710 }, { "grad_norm": 0.2198553830385208, "learning_rate": 4.141331824641003e-05, "loss": 0.0284, "step": 57720 }, { "grad_norm": 0.33240097761154175, "learning_rate": 4.139702966710452e-05, "loss": 0.0296, "step": 57730 }, { "grad_norm": 0.3373669981956482, "learning_rate": 4.13807420286069e-05, "loss": 0.0191, "step": 57740 }, { "grad_norm": 0.3612736761569977, "learning_rate": 4.136445533269837e-05, "loss": 0.0264, "step": 57750 }, { "grad_norm": 0.3509449362754822, "learning_rate": 4.134816958116003e-05, "loss": 0.026, "step": 57760 }, { "grad_norm": 0.933240532875061, "learning_rate": 4.133188477577283e-05, "loss": 0.0226, "step": 57770 }, { "grad_norm": 0.3700348436832428, "learning_rate": 4.131560091831771e-05, "loss": 0.0277, "step": 57780 }, { "grad_norm": 0.5694341063499451, "learning_rate": 4.1299318010575405e-05, "loss": 0.0269, "step": 57790 }, { "grad_norm": 0.264900267124176, "learning_rate": 4.128303605432661e-05, "loss": 0.0355, "step": 57800 }, { "grad_norm": 0.3853531777858734, "learning_rate": 4.126675505135188e-05, "loss": 0.0265, "step": 57810 }, { "grad_norm": 0.8060511350631714, "learning_rate": 4.125047500343169e-05, "loss": 0.0353, "step": 57820 }, { "grad_norm": 0.30937838554382324, "learning_rate": 4.1234195912346376e-05, "loss": 0.0347, "step": 57830 }, { "grad_norm": 0.5282112956047058, "learning_rate": 4.1217917779876235e-05, "loss": 0.0252, "step": 57840 }, { "grad_norm": 0.5748558640480042, "learning_rate": 4.120164060780141e-05, "loss": 0.0344, "step": 57850 }, { "grad_norm": 0.4478416442871094, "learning_rate": 4.1185364397901935e-05, "loss": 0.0339, "step": 57860 }, { "grad_norm": 0.8083096146583557, "learning_rate": 4.116908915195776e-05, "loss": 0.0276, "step": 57870 }, { "grad_norm": 0.35992544889450073, "learning_rate": 4.115281487174872e-05, "loss": 0.0301, "step": 57880 }, { "grad_norm": 0.3890489935874939, "learning_rate": 4.113654155905452e-05, "loss": 0.0285, "step": 57890 }, { "grad_norm": 0.2530921399593353, "learning_rate": 4.112026921565483e-05, "loss": 0.0315, "step": 57900 }, { "grad_norm": 0.22500891983509064, "learning_rate": 4.1103997843329147e-05, "loss": 0.0232, "step": 57910 }, { "grad_norm": 0.4319465160369873, "learning_rate": 4.108772744385689e-05, "loss": 0.0382, "step": 57920 }, { "grad_norm": 0.29766613245010376, "learning_rate": 4.1071458019017346e-05, "loss": 0.0296, "step": 57930 }, { "grad_norm": 0.36204832792282104, "learning_rate": 4.1055189570589734e-05, "loss": 0.0205, "step": 57940 }, { "grad_norm": 0.39527854323387146, "learning_rate": 4.103892210035315e-05, "loss": 0.0329, "step": 57950 }, { "grad_norm": 0.39159509539604187, "learning_rate": 4.102265561008656e-05, "loss": 0.0272, "step": 57960 }, { "grad_norm": 0.5468004941940308, "learning_rate": 4.100639010156885e-05, "loss": 0.0267, "step": 57970 }, { "grad_norm": 0.22748339176177979, "learning_rate": 4.099012557657882e-05, "loss": 0.0217, "step": 57980 }, { "grad_norm": 0.28986656665802, "learning_rate": 4.09738620368951e-05, "loss": 0.0231, "step": 57990 }, { "grad_norm": 0.3031329810619354, "learning_rate": 4.095759948429626e-05, "loss": 0.0266, "step": 58000 }, { "grad_norm": 0.31441763043403625, "learning_rate": 4.094133792056075e-05, "loss": 0.0222, "step": 58010 }, { "grad_norm": 0.2860593795776367, "learning_rate": 4.092507734746688e-05, "loss": 0.0268, "step": 58020 }, { "grad_norm": 0.4659053087234497, "learning_rate": 4.090881776679293e-05, "loss": 0.0315, "step": 58030 }, { "grad_norm": 0.31926801800727844, "learning_rate": 4.0892559180316994e-05, "loss": 0.0251, "step": 58040 }, { "grad_norm": 0.49971604347229004, "learning_rate": 4.08763015898171e-05, "loss": 0.0268, "step": 58050 }, { "grad_norm": 0.625282347202301, "learning_rate": 4.0860044997071145e-05, "loss": 0.0247, "step": 58060 }, { "grad_norm": 0.30764609575271606, "learning_rate": 4.084378940385693e-05, "loss": 0.0298, "step": 58070 }, { "grad_norm": 0.40909698605537415, "learning_rate": 4.082753481195211e-05, "loss": 0.0228, "step": 58080 }, { "grad_norm": 0.38371339440345764, "learning_rate": 4.0811281223134324e-05, "loss": 0.021, "step": 58090 }, { "grad_norm": 0.5245224237442017, "learning_rate": 4.0795028639181e-05, "loss": 0.0339, "step": 58100 }, { "grad_norm": 0.3998255729675293, "learning_rate": 4.077877706186951e-05, "loss": 0.0267, "step": 58110 }, { "grad_norm": 0.5648133754730225, "learning_rate": 4.076252649297709e-05, "loss": 0.0262, "step": 58120 }, { "grad_norm": 0.28875741362571716, "learning_rate": 4.07462769342809e-05, "loss": 0.0309, "step": 58130 }, { "grad_norm": 0.4064728319644928, "learning_rate": 4.073002838755793e-05, "loss": 0.0296, "step": 58140 }, { "grad_norm": 0.5276339650154114, "learning_rate": 4.0713780854585115e-05, "loss": 0.025, "step": 58150 }, { "grad_norm": 0.25706422328948975, "learning_rate": 4.069753433713927e-05, "loss": 0.0253, "step": 58160 }, { "grad_norm": 0.27904966473579407, "learning_rate": 4.068128883699709e-05, "loss": 0.0207, "step": 58170 }, { "grad_norm": 0.2956501543521881, "learning_rate": 4.0665044355935155e-05, "loss": 0.0249, "step": 58180 }, { "grad_norm": 0.4030340313911438, "learning_rate": 4.064880089572993e-05, "loss": 0.0218, "step": 58190 }, { "grad_norm": 0.2561085522174835, "learning_rate": 4.063255845815779e-05, "loss": 0.0316, "step": 58200 }, { "grad_norm": 0.4154582917690277, "learning_rate": 4.0616317044994945e-05, "loss": 0.0173, "step": 58210 }, { "grad_norm": 0.6479082107543945, "learning_rate": 4.0600076658017585e-05, "loss": 0.0251, "step": 58220 }, { "grad_norm": 0.3333488404750824, "learning_rate": 4.0583837299001706e-05, "loss": 0.0314, "step": 58230 }, { "grad_norm": 0.5774791836738586, "learning_rate": 4.056759896972322e-05, "loss": 0.0195, "step": 58240 }, { "grad_norm": 0.33193013072013855, "learning_rate": 4.055136167195793e-05, "loss": 0.0317, "step": 58250 }, { "grad_norm": 0.30565547943115234, "learning_rate": 4.053512540748152e-05, "loss": 0.0355, "step": 58260 }, { "grad_norm": 0.2719111442565918, "learning_rate": 4.051889017806955e-05, "loss": 0.0275, "step": 58270 }, { "grad_norm": 0.3102551996707916, "learning_rate": 4.0502655985497516e-05, "loss": 0.0212, "step": 58280 }, { "grad_norm": 0.6899694800376892, "learning_rate": 4.048642283154074e-05, "loss": 0.0259, "step": 58290 }, { "grad_norm": 0.5649167895317078, "learning_rate": 4.0470190717974464e-05, "loss": 0.0323, "step": 58300 }, { "grad_norm": 0.3755270838737488, "learning_rate": 4.045395964657381e-05, "loss": 0.0254, "step": 58310 }, { "grad_norm": 0.2974102199077606, "learning_rate": 4.043772961911376e-05, "loss": 0.0242, "step": 58320 }, { "grad_norm": 0.43260377645492554, "learning_rate": 4.042150063736924e-05, "loss": 0.0247, "step": 58330 }, { "grad_norm": 0.3459332585334778, "learning_rate": 4.0405272703114976e-05, "loss": 0.0292, "step": 58340 }, { "grad_norm": 0.33334439992904663, "learning_rate": 4.038904581812569e-05, "loss": 0.0243, "step": 58350 }, { "grad_norm": 0.3144301772117615, "learning_rate": 4.03728199841759e-05, "loss": 0.0346, "step": 58360 }, { "grad_norm": 0.2372327744960785, "learning_rate": 4.0356595203040044e-05, "loss": 0.0308, "step": 58370 }, { "grad_norm": 0.4341113865375519, "learning_rate": 4.0340371476492426e-05, "loss": 0.0275, "step": 58380 }, { "grad_norm": 0.3009819984436035, "learning_rate": 4.032414880630727e-05, "loss": 0.0346, "step": 58390 }, { "grad_norm": 0.504856288433075, "learning_rate": 4.0307927194258646e-05, "loss": 0.0252, "step": 58400 }, { "grad_norm": 0.34457066655158997, "learning_rate": 4.029170664212054e-05, "loss": 0.0202, "step": 58410 }, { "grad_norm": 0.3194849193096161, "learning_rate": 4.0275487151666794e-05, "loss": 0.0285, "step": 58420 }, { "grad_norm": 0.3522654175758362, "learning_rate": 4.025926872467116e-05, "loss": 0.024, "step": 58430 }, { "grad_norm": 0.5349761843681335, "learning_rate": 4.024305136290725e-05, "loss": 0.0264, "step": 58440 }, { "grad_norm": 0.2311253696680069, "learning_rate": 4.022683506814858e-05, "loss": 0.0227, "step": 58450 }, { "grad_norm": 0.35496604442596436, "learning_rate": 4.021061984216853e-05, "loss": 0.0265, "step": 58460 }, { "grad_norm": 0.32132649421691895, "learning_rate": 4.019440568674035e-05, "loss": 0.0271, "step": 58470 }, { "grad_norm": 0.3788212537765503, "learning_rate": 4.017819260363724e-05, "loss": 0.0221, "step": 58480 }, { "grad_norm": 0.23087237775325775, "learning_rate": 4.016198059463223e-05, "loss": 0.0337, "step": 58490 }, { "grad_norm": 0.9365787506103516, "learning_rate": 4.0145769661498233e-05, "loss": 0.03, "step": 58500 }, { "grad_norm": 0.8327637314796448, "learning_rate": 4.012955980600804e-05, "loss": 0.0233, "step": 58510 }, { "grad_norm": 0.46031442284584045, "learning_rate": 4.0113351029934355e-05, "loss": 0.0295, "step": 58520 }, { "grad_norm": 0.32709991931915283, "learning_rate": 4.0097143335049714e-05, "loss": 0.0259, "step": 58530 }, { "grad_norm": 0.4215274453163147, "learning_rate": 4.0080936723126606e-05, "loss": 0.0236, "step": 58540 }, { "grad_norm": 0.3057629466056824, "learning_rate": 4.006473119593735e-05, "loss": 0.0245, "step": 58550 }, { "grad_norm": 0.3110097646713257, "learning_rate": 4.004852675525415e-05, "loss": 0.0273, "step": 58560 }, { "grad_norm": 0.21897514164447784, "learning_rate": 4.00323234028491e-05, "loss": 0.024, "step": 58570 }, { "grad_norm": 0.37610751390457153, "learning_rate": 4.0016121140494186e-05, "loss": 0.0293, "step": 58580 }, { "grad_norm": 0.40913528203964233, "learning_rate": 3.999991996996122e-05, "loss": 0.03, "step": 58590 }, { "grad_norm": 0.23969732224941254, "learning_rate": 3.9983719893022e-05, "loss": 0.0211, "step": 58600 }, { "grad_norm": 0.40082985162734985, "learning_rate": 3.996752091144812e-05, "loss": 0.0243, "step": 58610 }, { "grad_norm": 0.42331230640411377, "learning_rate": 3.9951323027011055e-05, "loss": 0.0292, "step": 58620 }, { "grad_norm": 0.460370272397995, "learning_rate": 3.993512624148219e-05, "loss": 0.0265, "step": 58630 }, { "grad_norm": 0.3761483132839203, "learning_rate": 3.99189305566328e-05, "loss": 0.0198, "step": 58640 }, { "grad_norm": 0.20437318086624146, "learning_rate": 3.990273597423401e-05, "loss": 0.0203, "step": 58650 }, { "grad_norm": 0.5417617559432983, "learning_rate": 3.988654249605681e-05, "loss": 0.0253, "step": 58660 }, { "grad_norm": 0.35846298933029175, "learning_rate": 3.987035012387213e-05, "loss": 0.0214, "step": 58670 }, { "grad_norm": 0.33877235651016235, "learning_rate": 3.9854158859450735e-05, "loss": 0.0222, "step": 58680 }, { "grad_norm": 0.45915526151657104, "learning_rate": 3.983796870456328e-05, "loss": 0.0278, "step": 58690 }, { "grad_norm": 0.2192818969488144, "learning_rate": 3.982177966098028e-05, "loss": 0.0168, "step": 58700 }, { "grad_norm": 0.5148956179618835, "learning_rate": 3.980559173047216e-05, "loss": 0.0233, "step": 58710 }, { "grad_norm": 0.3858189284801483, "learning_rate": 3.9789404914809176e-05, "loss": 0.0163, "step": 58720 }, { "grad_norm": 0.21780920028686523, "learning_rate": 3.9773219215761544e-05, "loss": 0.026, "step": 58730 }, { "grad_norm": 0.29859158396720886, "learning_rate": 3.975703463509928e-05, "loss": 0.0168, "step": 58740 }, { "grad_norm": 0.5627537369728088, "learning_rate": 3.974085117459231e-05, "loss": 0.0234, "step": 58750 }, { "grad_norm": 0.47974249720573425, "learning_rate": 3.972466883601043e-05, "loss": 0.0264, "step": 58760 }, { "grad_norm": 0.30461204051971436, "learning_rate": 3.9708487621123316e-05, "loss": 0.0268, "step": 58770 }, { "grad_norm": 0.4244178533554077, "learning_rate": 3.969230753170051e-05, "loss": 0.0302, "step": 58780 }, { "grad_norm": 0.22339895367622375, "learning_rate": 3.967612856951146e-05, "loss": 0.0244, "step": 58790 }, { "grad_norm": 0.6085545420646667, "learning_rate": 3.9659950736325476e-05, "loss": 0.0255, "step": 58800 }, { "grad_norm": 0.28592649102211, "learning_rate": 3.964377403391174e-05, "loss": 0.0506, "step": 58810 }, { "grad_norm": 0.3837794065475464, "learning_rate": 3.9627598464039303e-05, "loss": 0.0378, "step": 58820 }, { "grad_norm": 0.31436780095100403, "learning_rate": 3.9611424028477096e-05, "loss": 0.0212, "step": 58830 }, { "grad_norm": 0.2354860007762909, "learning_rate": 3.959525072899394e-05, "loss": 0.0359, "step": 58840 }, { "grad_norm": 0.273749977350235, "learning_rate": 3.9579078567358506e-05, "loss": 0.0217, "step": 58850 }, { "grad_norm": 0.34773072600364685, "learning_rate": 3.956290754533939e-05, "loss": 0.0345, "step": 58860 }, { "grad_norm": 0.20536065101623535, "learning_rate": 3.954673766470501e-05, "loss": 0.0305, "step": 58870 }, { "grad_norm": 0.3913843333721161, "learning_rate": 3.9530568927223674e-05, "loss": 0.0234, "step": 58880 }, { "grad_norm": 0.2579593062400818, "learning_rate": 3.95144013346636e-05, "loss": 0.0202, "step": 58890 }, { "grad_norm": 0.32103103399276733, "learning_rate": 3.949823488879282e-05, "loss": 0.025, "step": 58900 }, { "grad_norm": 0.44225984811782837, "learning_rate": 3.9482069591379276e-05, "loss": 0.0401, "step": 58910 }, { "grad_norm": 0.33145278692245483, "learning_rate": 3.9465905444190784e-05, "loss": 0.0279, "step": 58920 }, { "grad_norm": 0.275631844997406, "learning_rate": 3.944974244899505e-05, "loss": 0.0263, "step": 58930 }, { "grad_norm": 0.3630845546722412, "learning_rate": 3.9433580607559616e-05, "loss": 0.0293, "step": 58940 }, { "grad_norm": 0.3675162196159363, "learning_rate": 3.941741992165193e-05, "loss": 0.0323, "step": 58950 }, { "grad_norm": 0.6501790881156921, "learning_rate": 3.940126039303928e-05, "loss": 0.0245, "step": 58960 }, { "grad_norm": 0.4090268015861511, "learning_rate": 3.938510202348886e-05, "loss": 0.03, "step": 58970 }, { "grad_norm": 0.1585734784603119, "learning_rate": 3.93689448147677e-05, "loss": 0.0201, "step": 58980 }, { "grad_norm": 0.45483294129371643, "learning_rate": 3.935278876864278e-05, "loss": 0.0296, "step": 58990 }, { "grad_norm": 0.41235411167144775, "learning_rate": 3.933663388688087e-05, "loss": 0.0291, "step": 59000 }, { "grad_norm": 0.6916952729225159, "learning_rate": 3.9320480171248645e-05, "loss": 0.0277, "step": 59010 }, { "grad_norm": 0.604472815990448, "learning_rate": 3.930432762351265e-05, "loss": 0.0389, "step": 59020 }, { "grad_norm": 0.36767256259918213, "learning_rate": 3.928817624543931e-05, "loss": 0.0334, "step": 59030 }, { "grad_norm": 0.17834676802158356, "learning_rate": 3.9272026038794885e-05, "loss": 0.0231, "step": 59040 }, { "grad_norm": 0.6845059394836426, "learning_rate": 3.9255877005345586e-05, "loss": 0.0296, "step": 59050 }, { "grad_norm": 0.28951534628868103, "learning_rate": 3.923972914685743e-05, "loss": 0.0265, "step": 59060 }, { "grad_norm": 0.33724167943000793, "learning_rate": 3.922358246509631e-05, "loss": 0.0253, "step": 59070 }, { "grad_norm": 0.21379072964191437, "learning_rate": 3.920743696182801e-05, "loss": 0.0267, "step": 59080 }, { "grad_norm": 0.23766294121742249, "learning_rate": 3.9191292638818164e-05, "loss": 0.0229, "step": 59090 }, { "grad_norm": 0.3066385090351105, "learning_rate": 3.9175149497832316e-05, "loss": 0.0287, "step": 59100 }, { "grad_norm": 0.910790205001831, "learning_rate": 3.915900754063584e-05, "loss": 0.0287, "step": 59110 }, { "grad_norm": 0.36438897252082825, "learning_rate": 3.9142866768994014e-05, "loss": 0.0199, "step": 59120 }, { "grad_norm": 0.38738059997558594, "learning_rate": 3.912672718467193e-05, "loss": 0.0239, "step": 59130 }, { "grad_norm": 0.32228291034698486, "learning_rate": 3.911058878943463e-05, "loss": 0.0192, "step": 59140 }, { "grad_norm": 0.1732323169708252, "learning_rate": 3.9094451585046974e-05, "loss": 0.0192, "step": 59150 }, { "grad_norm": 0.6437647342681885, "learning_rate": 3.90783155732737e-05, "loss": 0.0294, "step": 59160 }, { "grad_norm": 0.3426356315612793, "learning_rate": 3.9062180755879373e-05, "loss": 0.0266, "step": 59170 }, { "grad_norm": 0.36464688181877136, "learning_rate": 3.904604713462855e-05, "loss": 0.0243, "step": 59180 }, { "grad_norm": 0.37300166487693787, "learning_rate": 3.902991471128554e-05, "loss": 0.0176, "step": 59190 }, { "grad_norm": 0.2071942687034607, "learning_rate": 3.901378348761456e-05, "loss": 0.0179, "step": 59200 }, { "grad_norm": 0.3147166967391968, "learning_rate": 3.899765346537971e-05, "loss": 0.0307, "step": 59210 }, { "grad_norm": 0.31957098841667175, "learning_rate": 3.898152464634493e-05, "loss": 0.0278, "step": 59220 }, { "grad_norm": 0.4013623893260956, "learning_rate": 3.896539703227402e-05, "loss": 0.0272, "step": 59230 }, { "grad_norm": 0.3381158709526062, "learning_rate": 3.894927062493073e-05, "loss": 0.0215, "step": 59240 }, { "grad_norm": 0.3765850365161896, "learning_rate": 3.893314542607859e-05, "loss": 0.0198, "step": 59250 }, { "grad_norm": 0.22832024097442627, "learning_rate": 3.891702143748103e-05, "loss": 0.0188, "step": 59260 }, { "grad_norm": 0.2922133803367615, "learning_rate": 3.890089866090133e-05, "loss": 0.0233, "step": 59270 }, { "grad_norm": 0.4218256175518036, "learning_rate": 3.888477709810267e-05, "loss": 0.0302, "step": 59280 }, { "grad_norm": 0.3059867322444916, "learning_rate": 3.886865675084806e-05, "loss": 0.0179, "step": 59290 }, { "grad_norm": 0.23610931634902954, "learning_rate": 3.885253762090043e-05, "loss": 0.0223, "step": 59300 }, { "grad_norm": 0.28722119331359863, "learning_rate": 3.8836419710022534e-05, "loss": 0.0239, "step": 59310 }, { "grad_norm": 0.622157096862793, "learning_rate": 3.882030301997698e-05, "loss": 0.0273, "step": 59320 }, { "grad_norm": 0.22091692686080933, "learning_rate": 3.880418755252629e-05, "loss": 0.0236, "step": 59330 }, { "grad_norm": 0.41488584876060486, "learning_rate": 3.87880733094328e-05, "loss": 0.0246, "step": 59340 }, { "grad_norm": 0.3158981204032898, "learning_rate": 3.877196029245877e-05, "loss": 0.0182, "step": 59350 }, { "grad_norm": 0.2520454525947571, "learning_rate": 3.875584850336627e-05, "loss": 0.024, "step": 59360 }, { "grad_norm": 0.4603862464427948, "learning_rate": 3.873973794391728e-05, "loss": 0.0289, "step": 59370 }, { "grad_norm": 0.2287132441997528, "learning_rate": 3.872362861587361e-05, "loss": 0.0231, "step": 59380 }, { "grad_norm": 0.5553401708602905, "learning_rate": 3.870752052099698e-05, "loss": 0.0242, "step": 59390 }, { "grad_norm": 0.2525430917739868, "learning_rate": 3.8691413661048916e-05, "loss": 0.0195, "step": 59400 }, { "grad_norm": 0.3804602324962616, "learning_rate": 3.867530803779085e-05, "loss": 0.0261, "step": 59410 }, { "grad_norm": 0.5077958703041077, "learning_rate": 3.8659203652984054e-05, "loss": 0.0238, "step": 59420 }, { "grad_norm": 0.2472003698348999, "learning_rate": 3.864310050838972e-05, "loss": 0.0236, "step": 59430 }, { "grad_norm": 0.15174086391925812, "learning_rate": 3.8626998605768835e-05, "loss": 0.0193, "step": 59440 }, { "grad_norm": 0.3174666166305542, "learning_rate": 3.8610897946882294e-05, "loss": 0.0246, "step": 59450 }, { "grad_norm": 0.42065587639808655, "learning_rate": 3.859479853349082e-05, "loss": 0.0247, "step": 59460 }, { "grad_norm": 0.26974570751190186, "learning_rate": 3.857870036735503e-05, "loss": 0.0274, "step": 59470 }, { "grad_norm": 0.3030029833316803, "learning_rate": 3.856260345023538e-05, "loss": 0.0249, "step": 59480 }, { "grad_norm": 1.2188880443572998, "learning_rate": 3.854650778389224e-05, "loss": 0.022, "step": 59490 }, { "grad_norm": 0.33095231652259827, "learning_rate": 3.8530413370085786e-05, "loss": 0.0318, "step": 59500 }, { "grad_norm": 0.36901751160621643, "learning_rate": 3.851432021057608e-05, "loss": 0.0212, "step": 59510 }, { "grad_norm": 0.7669134736061096, "learning_rate": 3.849822830712306e-05, "loss": 0.0274, "step": 59520 }, { "grad_norm": 0.4712776243686676, "learning_rate": 3.848213766148649e-05, "loss": 0.0157, "step": 59530 }, { "grad_norm": 0.41223981976509094, "learning_rate": 3.846604827542602e-05, "loss": 0.0309, "step": 59540 }, { "grad_norm": 0.28724485635757446, "learning_rate": 3.844996015070117e-05, "loss": 0.0312, "step": 59550 }, { "grad_norm": 0.2830778956413269, "learning_rate": 3.843387328907132e-05, "loss": 0.0192, "step": 59560 }, { "grad_norm": 0.3448483347892761, "learning_rate": 3.8417787692295696e-05, "loss": 0.0302, "step": 59570 }, { "grad_norm": 0.12487959861755371, "learning_rate": 3.84017033621334e-05, "loss": 0.0161, "step": 59580 }, { "grad_norm": 0.3686720132827759, "learning_rate": 3.838562030034338e-05, "loss": 0.0222, "step": 59590 }, { "grad_norm": 0.29464730620384216, "learning_rate": 3.836953850868447e-05, "loss": 0.0257, "step": 59600 }, { "grad_norm": 0.249933660030365, "learning_rate": 3.835345798891533e-05, "loss": 0.0164, "step": 59610 }, { "grad_norm": 0.45508599281311035, "learning_rate": 3.833737874279452e-05, "loss": 0.0264, "step": 59620 }, { "grad_norm": 0.6317887902259827, "learning_rate": 3.8321300772080433e-05, "loss": 0.019, "step": 59630 }, { "grad_norm": 0.22205418348312378, "learning_rate": 3.8305224078531345e-05, "loss": 0.0202, "step": 59640 }, { "grad_norm": 0.4635394215583801, "learning_rate": 3.828914866390537e-05, "loss": 0.0248, "step": 59650 }, { "grad_norm": 0.38184767961502075, "learning_rate": 3.827307452996048e-05, "loss": 0.0317, "step": 59660 }, { "grad_norm": 0.2324819564819336, "learning_rate": 3.825700167845454e-05, "loss": 0.0333, "step": 59670 }, { "grad_norm": 0.6262980103492737, "learning_rate": 3.8240930111145214e-05, "loss": 0.0348, "step": 59680 }, { "grad_norm": 0.48125720024108887, "learning_rate": 3.8224859829790116e-05, "loss": 0.0249, "step": 59690 }, { "grad_norm": 0.19351449608802795, "learning_rate": 3.820879083614664e-05, "loss": 0.0279, "step": 59700 }, { "grad_norm": 0.5442586541175842, "learning_rate": 3.819272313197208e-05, "loss": 0.0284, "step": 59710 }, { "grad_norm": 0.36069631576538086, "learning_rate": 3.817665671902355e-05, "loss": 0.032, "step": 59720 }, { "grad_norm": 0.24588267505168915, "learning_rate": 3.816059159905808e-05, "loss": 0.024, "step": 59730 }, { "grad_norm": 0.32764044404029846, "learning_rate": 3.8144527773832476e-05, "loss": 0.0282, "step": 59740 }, { "grad_norm": 0.3152892291545868, "learning_rate": 3.812846524510352e-05, "loss": 0.0212, "step": 59750 }, { "grad_norm": 0.45896586775779724, "learning_rate": 3.811240401462775e-05, "loss": 0.0257, "step": 59760 }, { "grad_norm": 0.3871091604232788, "learning_rate": 3.809634408416162e-05, "loss": 0.0246, "step": 59770 }, { "grad_norm": 0.3284693658351898, "learning_rate": 3.8080285455461394e-05, "loss": 0.0219, "step": 59780 }, { "grad_norm": 0.2287173718214035, "learning_rate": 3.806422813028323e-05, "loss": 0.0246, "step": 59790 }, { "grad_norm": 0.3804458677768707, "learning_rate": 3.8048172110383114e-05, "loss": 0.0233, "step": 59800 }, { "grad_norm": 0.318350613117218, "learning_rate": 3.8032117397516944e-05, "loss": 0.023, "step": 59810 }, { "grad_norm": 0.25469112396240234, "learning_rate": 3.801606399344042e-05, "loss": 0.0188, "step": 59820 }, { "grad_norm": 0.3824383616447449, "learning_rate": 3.800001189990913e-05, "loss": 0.0249, "step": 59830 }, { "grad_norm": 0.3226718306541443, "learning_rate": 3.798396111867847e-05, "loss": 0.025, "step": 59840 }, { "grad_norm": 0.2506597936153412, "learning_rate": 3.796791165150378e-05, "loss": 0.0217, "step": 59850 }, { "grad_norm": 0.3784332573413849, "learning_rate": 3.795186350014018e-05, "loss": 0.0286, "step": 59860 }, { "grad_norm": 1.0301969051361084, "learning_rate": 3.793581666634266e-05, "loss": 0.0286, "step": 59870 }, { "grad_norm": 0.37204229831695557, "learning_rate": 3.791977115186609e-05, "loss": 0.0226, "step": 59880 }, { "grad_norm": 0.2975172698497772, "learning_rate": 3.79037269584652e-05, "loss": 0.0196, "step": 59890 }, { "grad_norm": 0.2549465298652649, "learning_rate": 3.7887684087894545e-05, "loss": 0.0325, "step": 59900 }, { "grad_norm": 0.46138709783554077, "learning_rate": 3.787164254190854e-05, "loss": 0.0222, "step": 59910 }, { "grad_norm": 0.2720067799091339, "learning_rate": 3.7855602322261473e-05, "loss": 0.0215, "step": 59920 }, { "grad_norm": 0.3689378798007965, "learning_rate": 3.783956343070746e-05, "loss": 0.0363, "step": 59930 }, { "grad_norm": 0.3278502821922302, "learning_rate": 3.7823525869000526e-05, "loss": 0.0215, "step": 59940 }, { "grad_norm": 0.2827146053314209, "learning_rate": 3.78074896388945e-05, "loss": 0.025, "step": 59950 }, { "grad_norm": 0.4215521812438965, "learning_rate": 3.7791454742143075e-05, "loss": 0.028, "step": 59960 }, { "grad_norm": 0.1561926305294037, "learning_rate": 3.7775421180499795e-05, "loss": 0.0168, "step": 59970 }, { "grad_norm": 0.4515402019023895, "learning_rate": 3.7759388955718086e-05, "loss": 0.021, "step": 59980 }, { "grad_norm": 0.29746213555336, "learning_rate": 3.774335806955117e-05, "loss": 0.0226, "step": 59990 }, { "grad_norm": 0.2635650932788849, "learning_rate": 3.7727328523752215e-05, "loss": 0.0215, "step": 60000 }, { "grad_norm": 0.22395558655261993, "learning_rate": 3.771130032007415e-05, "loss": 0.0226, "step": 60010 }, { "grad_norm": 0.42279505729675293, "learning_rate": 3.769527346026982e-05, "loss": 0.032, "step": 60020 }, { "grad_norm": 0.4856468439102173, "learning_rate": 3.767924794609188e-05, "loss": 0.0204, "step": 60030 }, { "grad_norm": 0.37404996156692505, "learning_rate": 3.7663223779292854e-05, "loss": 0.0348, "step": 60040 }, { "grad_norm": 0.24766261875629425, "learning_rate": 3.764720096162512e-05, "loss": 0.0228, "step": 60050 }, { "grad_norm": 0.4965220093727112, "learning_rate": 3.763117949484092e-05, "loss": 0.021, "step": 60060 }, { "grad_norm": 0.57993084192276, "learning_rate": 3.7615159380692344e-05, "loss": 0.0261, "step": 60070 }, { "grad_norm": 0.43586912751197815, "learning_rate": 3.759914062093132e-05, "loss": 0.0207, "step": 60080 }, { "grad_norm": 0.23000703752040863, "learning_rate": 3.7583123217309615e-05, "loss": 0.023, "step": 60090 }, { "grad_norm": 0.36954981088638306, "learning_rate": 3.7567107171578904e-05, "loss": 0.0289, "step": 60100 }, { "grad_norm": 0.510174036026001, "learning_rate": 3.755109248549066e-05, "loss": 0.0322, "step": 60110 }, { "grad_norm": 0.32326045632362366, "learning_rate": 3.75350791607962e-05, "loss": 0.0219, "step": 60120 }, { "grad_norm": 0.27654507756233215, "learning_rate": 3.751906719924676e-05, "loss": 0.0243, "step": 60130 }, { "grad_norm": 0.5664333701133728, "learning_rate": 3.750305660259337e-05, "loss": 0.0283, "step": 60140 }, { "grad_norm": 0.328734815120697, "learning_rate": 3.748704737258693e-05, "loss": 0.0279, "step": 60150 }, { "grad_norm": 0.3671199679374695, "learning_rate": 3.747103951097816e-05, "loss": 0.0313, "step": 60160 }, { "grad_norm": 0.601521909236908, "learning_rate": 3.745503301951768e-05, "loss": 0.0348, "step": 60170 }, { "grad_norm": 0.3471565842628479, "learning_rate": 3.743902789995592e-05, "loss": 0.0233, "step": 60180 }, { "grad_norm": 1.8496513366699219, "learning_rate": 3.7423024154043183e-05, "loss": 0.0309, "step": 60190 }, { "grad_norm": 0.3517126441001892, "learning_rate": 3.740702178352963e-05, "loss": 0.0255, "step": 60200 }, { "grad_norm": 0.2629753351211548, "learning_rate": 3.739102079016523e-05, "loss": 0.0234, "step": 60210 }, { "grad_norm": 0.2821153998374939, "learning_rate": 3.7375021175699846e-05, "loss": 0.02, "step": 60220 }, { "grad_norm": 0.3373783826828003, "learning_rate": 3.735902294188317e-05, "loss": 0.0257, "step": 60230 }, { "grad_norm": 0.43248268961906433, "learning_rate": 3.734302609046473e-05, "loss": 0.0228, "step": 60240 }, { "grad_norm": 0.4628937840461731, "learning_rate": 3.732703062319391e-05, "loss": 0.0298, "step": 60250 }, { "grad_norm": 0.41820839047431946, "learning_rate": 3.731103654181999e-05, "loss": 0.0251, "step": 60260 }, { "grad_norm": 0.5519158840179443, "learning_rate": 3.729504384809203e-05, "loss": 0.0241, "step": 60270 }, { "grad_norm": 0.3211219608783722, "learning_rate": 3.727905254375898e-05, "loss": 0.0313, "step": 60280 }, { "grad_norm": 0.27811092138290405, "learning_rate": 3.7263062630569614e-05, "loss": 0.0227, "step": 60290 }, { "grad_norm": 0.6429304480552673, "learning_rate": 3.724707411027256e-05, "loss": 0.0254, "step": 60300 }, { "grad_norm": 0.24474279582500458, "learning_rate": 3.723108698461631e-05, "loss": 0.0212, "step": 60310 }, { "grad_norm": 0.31849774718284607, "learning_rate": 3.72151012553492e-05, "loss": 0.022, "step": 60320 }, { "grad_norm": 0.3847303092479706, "learning_rate": 3.719911692421939e-05, "loss": 0.0192, "step": 60330 }, { "grad_norm": 0.7219948172569275, "learning_rate": 3.71831339929749e-05, "loss": 0.0402, "step": 60340 }, { "grad_norm": 0.8593580722808838, "learning_rate": 3.716715246336362e-05, "loss": 0.0302, "step": 60350 }, { "grad_norm": 0.5208174586296082, "learning_rate": 3.715117233713324e-05, "loss": 0.0215, "step": 60360 }, { "grad_norm": 0.311972051858902, "learning_rate": 3.713519361603135e-05, "loss": 0.0222, "step": 60370 }, { "grad_norm": 0.41682305932044983, "learning_rate": 3.711921630180532e-05, "loss": 0.0293, "step": 60380 }, { "grad_norm": 0.6584857702255249, "learning_rate": 3.710324039620245e-05, "loss": 0.0254, "step": 60390 }, { "grad_norm": 0.27669814229011536, "learning_rate": 3.708726590096982e-05, "loss": 0.0272, "step": 60400 }, { "grad_norm": 0.22942526638507843, "learning_rate": 3.707129281785437e-05, "loss": 0.0236, "step": 60410 }, { "grad_norm": 0.49416446685791016, "learning_rate": 3.705532114860291e-05, "loss": 0.0224, "step": 60420 }, { "grad_norm": 0.6176917552947998, "learning_rate": 3.703935089496207e-05, "loss": 0.0356, "step": 60430 }, { "grad_norm": 0.36084774136543274, "learning_rate": 3.702338205867829e-05, "loss": 0.0215, "step": 60440 }, { "grad_norm": 0.3804023861885071, "learning_rate": 3.7007414641497976e-05, "loss": 0.0242, "step": 60450 }, { "grad_norm": 0.25246661901474, "learning_rate": 3.699144864516726e-05, "loss": 0.0263, "step": 60460 }, { "grad_norm": 0.4047491252422333, "learning_rate": 3.6975484071432153e-05, "loss": 0.0248, "step": 60470 }, { "grad_norm": 0.5519661903381348, "learning_rate": 3.695952092203854e-05, "loss": 0.0242, "step": 60480 }, { "grad_norm": 0.4771899878978729, "learning_rate": 3.69435591987321e-05, "loss": 0.0389, "step": 60490 }, { "grad_norm": 0.2026410847902298, "learning_rate": 3.6927598903258374e-05, "loss": 0.0216, "step": 60500 }, { "grad_norm": 0.44001340866088867, "learning_rate": 3.691164003736281e-05, "loss": 0.0202, "step": 60510 }, { "grad_norm": 0.4195455014705658, "learning_rate": 3.689568260279061e-05, "loss": 0.0251, "step": 60520 }, { "grad_norm": 0.4270695745944977, "learning_rate": 3.687972660128686e-05, "loss": 0.0252, "step": 60530 }, { "grad_norm": 0.4284704029560089, "learning_rate": 3.686377203459648e-05, "loss": 0.0261, "step": 60540 }, { "grad_norm": 0.49642062187194824, "learning_rate": 3.6847818904464226e-05, "loss": 0.0317, "step": 60550 }, { "grad_norm": 0.3280431926250458, "learning_rate": 3.683186721263474e-05, "loss": 0.0262, "step": 60560 }, { "grad_norm": 0.28839969635009766, "learning_rate": 3.6815916960852435e-05, "loss": 0.0247, "step": 60570 }, { "grad_norm": 0.233305424451828, "learning_rate": 3.679996815086165e-05, "loss": 0.0252, "step": 60580 }, { "grad_norm": 0.45629116892814636, "learning_rate": 3.678402078440649e-05, "loss": 0.0297, "step": 60590 }, { "grad_norm": 0.250756174325943, "learning_rate": 3.676807486323096e-05, "loss": 0.0209, "step": 60600 }, { "grad_norm": 0.555660605430603, "learning_rate": 3.6752130389078864e-05, "loss": 0.0299, "step": 60610 }, { "grad_norm": 0.41596364974975586, "learning_rate": 3.673618736369388e-05, "loss": 0.0206, "step": 60620 }, { "grad_norm": 0.45318013429641724, "learning_rate": 3.6720245788819474e-05, "loss": 0.0236, "step": 60630 }, { "grad_norm": 0.38130179047584534, "learning_rate": 3.6704305666199044e-05, "loss": 0.0253, "step": 60640 }, { "grad_norm": 0.4286263585090637, "learning_rate": 3.668836699757576e-05, "loss": 0.0214, "step": 60650 }, { "grad_norm": 0.2915133833885193, "learning_rate": 3.667242978469265e-05, "loss": 0.0197, "step": 60660 }, { "grad_norm": 0.6728572249412537, "learning_rate": 3.6656494029292575e-05, "loss": 0.0253, "step": 60670 }, { "grad_norm": 0.2214326113462448, "learning_rate": 3.664055973311825e-05, "loss": 0.0208, "step": 60680 }, { "grad_norm": 0.23841027915477753, "learning_rate": 3.662462689791221e-05, "loss": 0.0213, "step": 60690 }, { "grad_norm": 0.18411032855510712, "learning_rate": 3.660869552541689e-05, "loss": 0.0204, "step": 60700 }, { "grad_norm": 0.4427669942378998, "learning_rate": 3.659276561737448e-05, "loss": 0.0234, "step": 60710 }, { "grad_norm": 0.3990647792816162, "learning_rate": 3.657683717552708e-05, "loss": 0.0206, "step": 60720 }, { "grad_norm": 0.3458646535873413, "learning_rate": 3.6560910201616574e-05, "loss": 0.021, "step": 60730 }, { "grad_norm": 0.3306885361671448, "learning_rate": 3.654498469738472e-05, "loss": 0.022, "step": 60740 }, { "grad_norm": 0.7015878558158875, "learning_rate": 3.6529060664573126e-05, "loss": 0.0359, "step": 60750 }, { "grad_norm": 0.23617014288902283, "learning_rate": 3.6513138104923176e-05, "loss": 0.0323, "step": 60760 }, { "grad_norm": 0.4920632243156433, "learning_rate": 3.6497217020176176e-05, "loss": 0.0207, "step": 60770 }, { "grad_norm": 0.3702542185783386, "learning_rate": 3.648129741207323e-05, "loss": 0.0272, "step": 60780 }, { "grad_norm": 0.33290529251098633, "learning_rate": 3.646537928235527e-05, "loss": 0.0316, "step": 60790 }, { "grad_norm": 0.32675567269325256, "learning_rate": 3.6449462632763075e-05, "loss": 0.0231, "step": 60800 }, { "grad_norm": 0.3409334719181061, "learning_rate": 3.643354746503729e-05, "loss": 0.0251, "step": 60810 }, { "grad_norm": 0.364276260137558, "learning_rate": 3.6417633780918345e-05, "loss": 0.0278, "step": 60820 }, { "grad_norm": 0.22573022544384003, "learning_rate": 3.640172158214655e-05, "loss": 0.0226, "step": 60830 }, { "grad_norm": 0.4772534668445587, "learning_rate": 3.6385810870462035e-05, "loss": 0.0229, "step": 60840 }, { "grad_norm": 0.7001233696937561, "learning_rate": 3.636990164760479e-05, "loss": 0.0274, "step": 60850 }, { "grad_norm": 0.37526047229766846, "learning_rate": 3.6353993915314596e-05, "loss": 0.0364, "step": 60860 }, { "grad_norm": 0.20642989873886108, "learning_rate": 3.6338087675331124e-05, "loss": 0.022, "step": 60870 }, { "grad_norm": 0.4136331379413605, "learning_rate": 3.632218292939383e-05, "loss": 0.0234, "step": 60880 }, { "grad_norm": 0.40971264243125916, "learning_rate": 3.6306279679242033e-05, "loss": 0.0247, "step": 60890 }, { "grad_norm": 0.27206164598464966, "learning_rate": 3.629037792661492e-05, "loss": 0.0243, "step": 60900 }, { "grad_norm": 0.5587205290794373, "learning_rate": 3.627447767325147e-05, "loss": 0.0283, "step": 60910 }, { "grad_norm": 0.2675841748714447, "learning_rate": 3.62585789208905e-05, "loss": 0.0197, "step": 60920 }, { "grad_norm": 0.29478028416633606, "learning_rate": 3.624268167127068e-05, "loss": 0.0341, "step": 60930 }, { "grad_norm": 0.36037737131118774, "learning_rate": 3.622678592613051e-05, "loss": 0.0303, "step": 60940 }, { "grad_norm": 0.40548980236053467, "learning_rate": 3.6210891687208304e-05, "loss": 0.0371, "step": 60950 }, { "grad_norm": 0.294820100069046, "learning_rate": 3.619499895624228e-05, "loss": 0.0213, "step": 60960 }, { "grad_norm": 0.5693414211273193, "learning_rate": 3.61791077349704e-05, "loss": 0.0314, "step": 60970 }, { "grad_norm": 0.638220489025116, "learning_rate": 3.616321802513053e-05, "loss": 0.0343, "step": 60980 }, { "grad_norm": 0.4913623631000519, "learning_rate": 3.6147329828460325e-05, "loss": 0.029, "step": 60990 }, { "grad_norm": 0.3578588664531708, "learning_rate": 3.6131443146697306e-05, "loss": 0.0305, "step": 61000 }, { "grad_norm": 0.30040666460990906, "learning_rate": 3.611555798157879e-05, "loss": 0.0288, "step": 61010 }, { "grad_norm": 0.22762379050254822, "learning_rate": 3.6099674334842e-05, "loss": 0.0216, "step": 61020 }, { "grad_norm": 0.4450261890888214, "learning_rate": 3.6083792208223934e-05, "loss": 0.0326, "step": 61030 }, { "grad_norm": 0.44485950469970703, "learning_rate": 3.606791160346142e-05, "loss": 0.0305, "step": 61040 }, { "grad_norm": 0.806799590587616, "learning_rate": 3.6052032522291134e-05, "loss": 0.0303, "step": 61050 }, { "grad_norm": 0.5032440423965454, "learning_rate": 3.603615496644962e-05, "loss": 0.0375, "step": 61060 }, { "grad_norm": 0.7556460499763489, "learning_rate": 3.60202789376732e-05, "loss": 0.0274, "step": 61070 }, { "grad_norm": 0.3291611969470978, "learning_rate": 3.6004404437698036e-05, "loss": 0.0246, "step": 61080 }, { "grad_norm": 0.5803008675575256, "learning_rate": 3.5988531468260184e-05, "loss": 0.0275, "step": 61090 }, { "grad_norm": 0.551982045173645, "learning_rate": 3.597266003109546e-05, "loss": 0.0343, "step": 61100 }, { "grad_norm": 0.34197747707366943, "learning_rate": 3.5956790127939557e-05, "loss": 0.0181, "step": 61110 }, { "grad_norm": 0.20616552233695984, "learning_rate": 3.594092176052796e-05, "loss": 0.0208, "step": 61120 }, { "grad_norm": 0.26061058044433594, "learning_rate": 3.592505493059603e-05, "loss": 0.0215, "step": 61130 }, { "grad_norm": 0.20737268030643463, "learning_rate": 3.5909189639878905e-05, "loss": 0.0342, "step": 61140 }, { "grad_norm": 1.426924705505371, "learning_rate": 3.5893325890111626e-05, "loss": 0.0268, "step": 61150 }, { "grad_norm": 0.2668014466762543, "learning_rate": 3.5877463683029034e-05, "loss": 0.0272, "step": 61160 }, { "grad_norm": 0.5432900786399841, "learning_rate": 3.586160302036578e-05, "loss": 0.0253, "step": 61170 }, { "grad_norm": 0.27680936455726624, "learning_rate": 3.584574390385636e-05, "loss": 0.0221, "step": 61180 }, { "grad_norm": 0.5070540904998779, "learning_rate": 3.582988633523511e-05, "loss": 0.0245, "step": 61190 }, { "grad_norm": 0.28770139813423157, "learning_rate": 3.581403031623616e-05, "loss": 0.0278, "step": 61200 }, { "grad_norm": 0.33585023880004883, "learning_rate": 3.579817584859354e-05, "loss": 0.0256, "step": 61210 }, { "grad_norm": 0.44144004583358765, "learning_rate": 3.578232293404107e-05, "loss": 0.0191, "step": 61220 }, { "grad_norm": 0.41288426518440247, "learning_rate": 3.5766471574312385e-05, "loss": 0.0244, "step": 61230 }, { "grad_norm": 0.23623301088809967, "learning_rate": 3.5750621771140964e-05, "loss": 0.0305, "step": 61240 }, { "grad_norm": 0.3200322687625885, "learning_rate": 3.5734773526260115e-05, "loss": 0.0223, "step": 61250 }, { "grad_norm": 0.5553191304206848, "learning_rate": 3.5718926841402996e-05, "loss": 0.0324, "step": 61260 }, { "grad_norm": 0.3892800807952881, "learning_rate": 3.5703081718302555e-05, "loss": 0.0248, "step": 61270 }, { "grad_norm": 0.35571804642677307, "learning_rate": 3.568723815869162e-05, "loss": 0.0299, "step": 61280 }, { "grad_norm": 0.40557992458343506, "learning_rate": 3.5671396164302777e-05, "loss": 0.0352, "step": 61290 }, { "grad_norm": 0.5044674277305603, "learning_rate": 3.565555573686853e-05, "loss": 0.0271, "step": 61300 }, { "grad_norm": 0.2535651922225952, "learning_rate": 3.563971687812113e-05, "loss": 0.0312, "step": 61310 }, { "grad_norm": 0.34087133407592773, "learning_rate": 3.5623879589792706e-05, "loss": 0.0261, "step": 61320 }, { "grad_norm": 0.38608697056770325, "learning_rate": 3.5608043873615185e-05, "loss": 0.0299, "step": 61330 }, { "grad_norm": 0.39329713582992554, "learning_rate": 3.559220973132035e-05, "loss": 0.019, "step": 61340 }, { "grad_norm": 0.3842011094093323, "learning_rate": 3.557637716463981e-05, "loss": 0.0215, "step": 61350 }, { "grad_norm": 0.30459970235824585, "learning_rate": 3.556054617530497e-05, "loss": 0.0213, "step": 61360 }, { "grad_norm": 0.2990573048591614, "learning_rate": 3.5544716765047106e-05, "loss": 0.0212, "step": 61370 }, { "grad_norm": 0.2721658945083618, "learning_rate": 3.5528888935597274e-05, "loss": 0.0258, "step": 61380 }, { "grad_norm": 1.3153995275497437, "learning_rate": 3.551306268868636e-05, "loss": 0.0257, "step": 61390 }, { "grad_norm": 0.7333866357803345, "learning_rate": 3.5497238026045166e-05, "loss": 0.0219, "step": 61400 }, { "grad_norm": 0.5186457633972168, "learning_rate": 3.548141494940421e-05, "loss": 0.0201, "step": 61410 }, { "grad_norm": 0.4243963956832886, "learning_rate": 3.546559346049388e-05, "loss": 0.0265, "step": 61420 }, { "grad_norm": 0.3198447525501251, "learning_rate": 3.544977356104441e-05, "loss": 0.0169, "step": 61430 }, { "grad_norm": 0.4700428247451782, "learning_rate": 3.543395525278582e-05, "loss": 0.0302, "step": 61440 }, { "grad_norm": 0.26399165391921997, "learning_rate": 3.5418138537447985e-05, "loss": 0.0219, "step": 61450 }, { "grad_norm": 0.46437135338783264, "learning_rate": 3.5402323416760574e-05, "loss": 0.0194, "step": 61460 }, { "grad_norm": 0.2847830653190613, "learning_rate": 3.538650989245316e-05, "loss": 0.0237, "step": 61470 }, { "grad_norm": 0.6081901788711548, "learning_rate": 3.537069796625504e-05, "loss": 0.0275, "step": 61480 }, { "grad_norm": 0.586835503578186, "learning_rate": 3.53548876398954e-05, "loss": 0.0212, "step": 61490 }, { "grad_norm": 0.32588568329811096, "learning_rate": 3.5339078915103214e-05, "loss": 0.0228, "step": 61500 }, { "grad_norm": 0.32229968905448914, "learning_rate": 3.532327179360733e-05, "loss": 0.0215, "step": 61510 }, { "grad_norm": 0.43439459800720215, "learning_rate": 3.530746627713636e-05, "loss": 0.0246, "step": 61520 }, { "grad_norm": 0.226981058716774, "learning_rate": 3.5291662367418805e-05, "loss": 0.0228, "step": 61530 }, { "grad_norm": 0.4200218617916107, "learning_rate": 3.527586006618293e-05, "loss": 0.0226, "step": 61540 }, { "grad_norm": 0.23390953242778778, "learning_rate": 3.526005937515687e-05, "loss": 0.0294, "step": 61550 }, { "grad_norm": 0.39200079441070557, "learning_rate": 3.524426029606856e-05, "loss": 0.0382, "step": 61560 }, { "grad_norm": 0.21723829209804535, "learning_rate": 3.5228462830645746e-05, "loss": 0.0224, "step": 61570 }, { "grad_norm": 0.5046965479850769, "learning_rate": 3.5212666980616036e-05, "loss": 0.031, "step": 61580 }, { "grad_norm": 0.37482911348342896, "learning_rate": 3.519687274770681e-05, "loss": 0.029, "step": 61590 }, { "grad_norm": 0.38189971446990967, "learning_rate": 3.518108013364535e-05, "loss": 0.0316, "step": 61600 }, { "grad_norm": 0.39856112003326416, "learning_rate": 3.516528914015868e-05, "loss": 0.0338, "step": 61610 }, { "grad_norm": 0.27646157145500183, "learning_rate": 3.5149499768973694e-05, "loss": 0.0246, "step": 61620 }, { "grad_norm": 0.22183826565742493, "learning_rate": 3.513371202181708e-05, "loss": 0.0219, "step": 61630 }, { "grad_norm": 0.330364465713501, "learning_rate": 3.511792590041537e-05, "loss": 0.0245, "step": 61640 }, { "grad_norm": 0.7520245909690857, "learning_rate": 3.510214140649488e-05, "loss": 0.0355, "step": 61650 }, { "grad_norm": 0.3082536458969116, "learning_rate": 3.508635854178184e-05, "loss": 0.0373, "step": 61660 }, { "grad_norm": 0.40939778089523315, "learning_rate": 3.50705773080022e-05, "loss": 0.0259, "step": 61670 }, { "grad_norm": 0.39141911268234253, "learning_rate": 3.505479770688178e-05, "loss": 0.0252, "step": 61680 }, { "grad_norm": 0.33377984166145325, "learning_rate": 3.503901974014622e-05, "loss": 0.0202, "step": 61690 }, { "grad_norm": 0.4250142276287079, "learning_rate": 3.502324340952095e-05, "loss": 0.0236, "step": 61700 }, { "grad_norm": 0.32116687297821045, "learning_rate": 3.500746871673126e-05, "loss": 0.0264, "step": 61710 }, { "grad_norm": 0.41089141368865967, "learning_rate": 3.499169566350227e-05, "loss": 0.0258, "step": 61720 }, { "grad_norm": 0.46036502718925476, "learning_rate": 3.4975924251558876e-05, "loss": 0.0281, "step": 61730 }, { "grad_norm": 0.2210351824760437, "learning_rate": 3.496015448262582e-05, "loss": 0.0249, "step": 61740 }, { "grad_norm": 0.44938308000564575, "learning_rate": 3.494438635842764e-05, "loss": 0.0176, "step": 61750 }, { "grad_norm": 0.1724987030029297, "learning_rate": 3.492861988068874e-05, "loss": 0.0252, "step": 61760 }, { "grad_norm": 0.3522700369358063, "learning_rate": 3.491285505113332e-05, "loss": 0.028, "step": 61770 }, { "grad_norm": 0.25800344347953796, "learning_rate": 3.4897091871485376e-05, "loss": 0.0283, "step": 61780 }, { "grad_norm": 0.6295478343963623, "learning_rate": 3.488133034346875e-05, "loss": 0.0248, "step": 61790 }, { "grad_norm": 0.4746167063713074, "learning_rate": 3.486557046880713e-05, "loss": 0.0203, "step": 61800 }, { "grad_norm": 0.5615503191947937, "learning_rate": 3.484981224922395e-05, "loss": 0.0253, "step": 61810 }, { "grad_norm": 0.28186145424842834, "learning_rate": 3.4834055686442544e-05, "loss": 0.0196, "step": 61820 }, { "grad_norm": 0.2493869662284851, "learning_rate": 3.4818300782186e-05, "loss": 0.0257, "step": 61830 }, { "grad_norm": 0.5579816699028015, "learning_rate": 3.4802547538177236e-05, "loss": 0.024, "step": 61840 }, { "grad_norm": 0.6237836480140686, "learning_rate": 3.478679595613905e-05, "loss": 0.033, "step": 61850 }, { "grad_norm": 0.6475377082824707, "learning_rate": 3.4771046037793975e-05, "loss": 0.0299, "step": 61860 }, { "grad_norm": 0.48019155859947205, "learning_rate": 3.4755297784864416e-05, "loss": 0.0274, "step": 61870 }, { "grad_norm": 0.48357146978378296, "learning_rate": 3.473955119907257e-05, "loss": 0.0242, "step": 61880 }, { "grad_norm": 0.17264531552791595, "learning_rate": 3.472380628214046e-05, "loss": 0.0201, "step": 61890 }, { "grad_norm": 0.22609731554985046, "learning_rate": 3.4708063035789906e-05, "loss": 0.0239, "step": 61900 }, { "grad_norm": 0.3924742043018341, "learning_rate": 3.4692321461742614e-05, "loss": 0.026, "step": 61910 }, { "grad_norm": 0.703045666217804, "learning_rate": 3.4676581561720023e-05, "loss": 0.02, "step": 61920 }, { "grad_norm": 0.5500778555870056, "learning_rate": 3.466084333744344e-05, "loss": 0.0283, "step": 61930 }, { "grad_norm": 0.3876364529132843, "learning_rate": 3.4645106790633964e-05, "loss": 0.0216, "step": 61940 }, { "grad_norm": 0.34632670879364014, "learning_rate": 3.4629371923012526e-05, "loss": 0.0249, "step": 61950 }, { "grad_norm": 0.2855032682418823, "learning_rate": 3.4613638736299856e-05, "loss": 0.0235, "step": 61960 }, { "grad_norm": 0.20373104512691498, "learning_rate": 3.459790723221652e-05, "loss": 0.0184, "step": 61970 }, { "grad_norm": 0.5651083588600159, "learning_rate": 3.45821774124829e-05, "loss": 0.029, "step": 61980 }, { "grad_norm": 0.32330504059791565, "learning_rate": 3.456644927881918e-05, "loss": 0.0173, "step": 61990 }, { "grad_norm": 0.41711536049842834, "learning_rate": 3.4550722832945354e-05, "loss": 0.0258, "step": 62000 }, { "grad_norm": 0.7114323377609253, "learning_rate": 3.453499807658126e-05, "loss": 0.0245, "step": 62010 }, { "grad_norm": 0.19260241091251373, "learning_rate": 3.451927501144653e-05, "loss": 0.0238, "step": 62020 }, { "grad_norm": 0.42947056889533997, "learning_rate": 3.450355363926059e-05, "loss": 0.0205, "step": 62030 }, { "grad_norm": 0.2630860209465027, "learning_rate": 3.448783396174273e-05, "loss": 0.0268, "step": 62040 }, { "grad_norm": 0.5228438377380371, "learning_rate": 3.4472115980612034e-05, "loss": 0.0243, "step": 62050 }, { "grad_norm": 0.45821669697761536, "learning_rate": 3.4456399697587396e-05, "loss": 0.0186, "step": 62060 }, { "grad_norm": 0.6600691676139832, "learning_rate": 3.4440685114387506e-05, "loss": 0.0209, "step": 62070 }, { "grad_norm": 0.2803230881690979, "learning_rate": 3.442497223273091e-05, "loss": 0.0277, "step": 62080 }, { "grad_norm": 1.2647570371627808, "learning_rate": 3.4409261054335926e-05, "loss": 0.0432, "step": 62090 }, { "grad_norm": 0.2671799063682556, "learning_rate": 3.439355158092069e-05, "loss": 0.032, "step": 62100 }, { "grad_norm": 0.34556615352630615, "learning_rate": 3.437784381420322e-05, "loss": 0.024, "step": 62110 }, { "grad_norm": 0.48567622900009155, "learning_rate": 3.436213775590126e-05, "loss": 0.0189, "step": 62120 }, { "grad_norm": 0.1779821664094925, "learning_rate": 3.4346433407732405e-05, "loss": 0.0262, "step": 62130 }, { "grad_norm": 0.19411689043045044, "learning_rate": 3.433073077141405e-05, "loss": 0.0237, "step": 62140 }, { "grad_norm": 0.2269805669784546, "learning_rate": 3.431502984866343e-05, "loss": 0.0305, "step": 62150 }, { "grad_norm": 0.32422006130218506, "learning_rate": 3.429933064119753e-05, "loss": 0.0283, "step": 62160 }, { "grad_norm": 0.18864639103412628, "learning_rate": 3.4283633150733255e-05, "loss": 0.0262, "step": 62170 }, { "grad_norm": 0.23653508722782135, "learning_rate": 3.4267937378987225e-05, "loss": 0.0275, "step": 62180 }, { "grad_norm": 0.49651163816452026, "learning_rate": 3.425224332767592e-05, "loss": 0.0201, "step": 62190 }, { "grad_norm": 0.1771884709596634, "learning_rate": 3.4236550998515595e-05, "loss": 0.0296, "step": 62200 }, { "grad_norm": 0.44451189041137695, "learning_rate": 3.422086039322235e-05, "loss": 0.0201, "step": 62210 }, { "grad_norm": 0.5079219341278076, "learning_rate": 3.420517151351209e-05, "loss": 0.0251, "step": 62220 }, { "grad_norm": 0.33236560225486755, "learning_rate": 3.418948436110053e-05, "loss": 0.027, "step": 62230 }, { "grad_norm": 0.20040656626224518, "learning_rate": 3.417379893770319e-05, "loss": 0.0237, "step": 62240 }, { "grad_norm": 0.27979302406311035, "learning_rate": 3.41581152450354e-05, "loss": 0.0287, "step": 62250 }, { "grad_norm": 0.5142138004302979, "learning_rate": 3.414243328481231e-05, "loss": 0.0217, "step": 62260 }, { "grad_norm": 0.45071732997894287, "learning_rate": 3.412675305874888e-05, "loss": 0.0275, "step": 62270 }, { "grad_norm": 0.2987522780895233, "learning_rate": 3.4111074568559876e-05, "loss": 0.0292, "step": 62280 }, { "grad_norm": 0.45906710624694824, "learning_rate": 3.4095397815959846e-05, "loss": 0.0243, "step": 62290 }, { "grad_norm": 0.4031469225883484, "learning_rate": 3.407972280266322e-05, "loss": 0.0247, "step": 62300 }, { "grad_norm": 0.33587589859962463, "learning_rate": 3.4064049530384166e-05, "loss": 0.0166, "step": 62310 }, { "grad_norm": 0.2827959656715393, "learning_rate": 3.40483780008367e-05, "loss": 0.0233, "step": 62320 }, { "grad_norm": 0.30133333802223206, "learning_rate": 3.4032708215734635e-05, "loss": 0.0232, "step": 62330 }, { "grad_norm": 0.3709263801574707, "learning_rate": 3.4017040176791584e-05, "loss": 0.0187, "step": 62340 }, { "grad_norm": 0.29243025183677673, "learning_rate": 3.4001373885720976e-05, "loss": 0.0244, "step": 62350 }, { "grad_norm": 0.2817918360233307, "learning_rate": 3.398570934423608e-05, "loss": 0.0243, "step": 62360 }, { "grad_norm": 0.3807034194469452, "learning_rate": 3.397004655404994e-05, "loss": 0.0228, "step": 62370 }, { "grad_norm": 0.45211541652679443, "learning_rate": 3.395438551687541e-05, "loss": 0.0267, "step": 62380 }, { "grad_norm": 0.20434550940990448, "learning_rate": 3.393872623442515e-05, "loss": 0.0184, "step": 62390 }, { "grad_norm": 0.36847829818725586, "learning_rate": 3.392306870841164e-05, "loss": 0.0229, "step": 62400 }, { "grad_norm": 0.2914714515209198, "learning_rate": 3.390741294054715e-05, "loss": 0.0204, "step": 62410 }, { "grad_norm": 0.6183090209960938, "learning_rate": 3.38917589325438e-05, "loss": 0.0225, "step": 62420 }, { "grad_norm": 0.3599075376987457, "learning_rate": 3.387610668611349e-05, "loss": 0.0201, "step": 62430 }, { "grad_norm": 0.5090565085411072, "learning_rate": 3.38604562029679e-05, "loss": 0.0261, "step": 62440 }, { "grad_norm": 0.3288043141365051, "learning_rate": 3.384480748481855e-05, "loss": 0.0237, "step": 62450 }, { "grad_norm": 0.16114675998687744, "learning_rate": 3.382916053337676e-05, "loss": 0.0279, "step": 62460 }, { "grad_norm": 0.3769691586494446, "learning_rate": 3.381351535035366e-05, "loss": 0.0305, "step": 62470 }, { "grad_norm": 0.32320743799209595, "learning_rate": 3.379787193746019e-05, "loss": 0.0197, "step": 62480 }, { "grad_norm": 0.3036877512931824, "learning_rate": 3.378223029640709e-05, "loss": 0.0203, "step": 62490 }, { "grad_norm": 0.3677041828632355, "learning_rate": 3.376659042890489e-05, "loss": 0.0199, "step": 62500 }, { "grad_norm": 0.235005185008049, "learning_rate": 3.375095233666397e-05, "loss": 0.0265, "step": 62510 }, { "grad_norm": 0.6854440569877625, "learning_rate": 3.373531602139447e-05, "loss": 0.0254, "step": 62520 }, { "grad_norm": 0.392139732837677, "learning_rate": 3.371968148480635e-05, "loss": 0.0241, "step": 62530 }, { "grad_norm": 0.3587420880794525, "learning_rate": 3.370404872860937e-05, "loss": 0.0311, "step": 62540 }, { "grad_norm": 0.26768290996551514, "learning_rate": 3.3688417754513145e-05, "loss": 0.0236, "step": 62550 }, { "grad_norm": 0.37522125244140625, "learning_rate": 3.367278856422703e-05, "loss": 0.023, "step": 62560 }, { "grad_norm": 0.2511769235134125, "learning_rate": 3.365716115946021e-05, "loss": 0.0239, "step": 62570 }, { "grad_norm": 0.34009629487991333, "learning_rate": 3.3641535541921666e-05, "loss": 0.036, "step": 62580 }, { "grad_norm": 0.5080908536911011, "learning_rate": 3.36259117133202e-05, "loss": 0.0209, "step": 62590 }, { "grad_norm": 0.15613095462322235, "learning_rate": 3.36102896753644e-05, "loss": 0.017, "step": 62600 }, { "grad_norm": 0.3096027672290802, "learning_rate": 3.35946694297627e-05, "loss": 0.0263, "step": 62610 }, { "grad_norm": 0.2699425220489502, "learning_rate": 3.357905097822328e-05, "loss": 0.0311, "step": 62620 }, { "grad_norm": 0.7718918323516846, "learning_rate": 3.356343432245416e-05, "loss": 0.0231, "step": 62630 }, { "grad_norm": 0.5597333908081055, "learning_rate": 3.354781946416315e-05, "loss": 0.0362, "step": 62640 }, { "grad_norm": 0.2532132565975189, "learning_rate": 3.3532206405057864e-05, "loss": 0.0214, "step": 62650 }, { "grad_norm": 0.46893510222435, "learning_rate": 3.351659514684573e-05, "loss": 0.0213, "step": 62660 }, { "grad_norm": 0.3027859926223755, "learning_rate": 3.350098569123395e-05, "loss": 0.0218, "step": 62670 }, { "grad_norm": 0.5851085186004639, "learning_rate": 3.348537803992958e-05, "loss": 0.0295, "step": 62680 }, { "grad_norm": 0.2799842059612274, "learning_rate": 3.346977219463945e-05, "loss": 0.023, "step": 62690 }, { "grad_norm": 0.253638356924057, "learning_rate": 3.345416815707017e-05, "loss": 0.0208, "step": 62700 }, { "grad_norm": 0.46372607350349426, "learning_rate": 3.3438565928928185e-05, "loss": 0.0286, "step": 62710 }, { "grad_norm": 0.43547576665878296, "learning_rate": 3.3422965511919736e-05, "loss": 0.0232, "step": 62720 }, { "grad_norm": 0.7519691586494446, "learning_rate": 3.3407366907750846e-05, "loss": 0.0187, "step": 62730 }, { "grad_norm": 0.22896675765514374, "learning_rate": 3.339177011812737e-05, "loss": 0.021, "step": 62740 }, { "grad_norm": 0.28713539242744446, "learning_rate": 3.3376175144754935e-05, "loss": 0.0222, "step": 62750 }, { "grad_norm": 0.2140561193227768, "learning_rate": 3.3360581989339004e-05, "loss": 0.0182, "step": 62760 }, { "grad_norm": 1.5377970933914185, "learning_rate": 3.3344990653584804e-05, "loss": 0.0247, "step": 62770 }, { "grad_norm": 0.4280342757701874, "learning_rate": 3.332940113919739e-05, "loss": 0.0264, "step": 62780 }, { "grad_norm": 0.2744985520839691, "learning_rate": 3.331381344788161e-05, "loss": 0.0254, "step": 62790 }, { "grad_norm": 0.22890302538871765, "learning_rate": 3.3298227581342076e-05, "loss": 0.0286, "step": 62800 }, { "grad_norm": 0.21129868924617767, "learning_rate": 3.328264354128329e-05, "loss": 0.0297, "step": 62810 }, { "grad_norm": 0.6244639754295349, "learning_rate": 3.3267061329409465e-05, "loss": 0.0254, "step": 62820 }, { "grad_norm": 0.20976203680038452, "learning_rate": 3.325148094742466e-05, "loss": 0.0263, "step": 62830 }, { "grad_norm": 0.3839382529258728, "learning_rate": 3.323590239703271e-05, "loss": 0.0198, "step": 62840 }, { "grad_norm": 0.4067752957344055, "learning_rate": 3.322032567993728e-05, "loss": 0.0281, "step": 62850 }, { "grad_norm": 0.4027675986289978, "learning_rate": 3.3204750797841774e-05, "loss": 0.0233, "step": 62860 }, { "grad_norm": 0.3309313952922821, "learning_rate": 3.31891777524495e-05, "loss": 0.0228, "step": 62870 }, { "grad_norm": 0.22336500883102417, "learning_rate": 3.317360654546347e-05, "loss": 0.0303, "step": 62880 }, { "grad_norm": 0.39958497881889343, "learning_rate": 3.3158037178586534e-05, "loss": 0.0254, "step": 62890 }, { "grad_norm": 0.33551427721977234, "learning_rate": 3.3142469653521316e-05, "loss": 0.0211, "step": 62900 }, { "grad_norm": 0.3424474895000458, "learning_rate": 3.3126903971970284e-05, "loss": 0.0228, "step": 62910 }, { "grad_norm": 0.49548572301864624, "learning_rate": 3.311134013563564e-05, "loss": 0.0243, "step": 62920 }, { "grad_norm": 0.36138394474983215, "learning_rate": 3.309577814621947e-05, "loss": 0.0237, "step": 62930 }, { "grad_norm": 0.23588630557060242, "learning_rate": 3.308021800542359e-05, "loss": 0.0217, "step": 62940 }, { "grad_norm": 0.6392054557800293, "learning_rate": 3.306465971494963e-05, "loss": 0.0273, "step": 62950 }, { "grad_norm": 0.2626854181289673, "learning_rate": 3.304910327649901e-05, "loss": 0.0217, "step": 62960 }, { "grad_norm": 0.6008992791175842, "learning_rate": 3.303354869177297e-05, "loss": 0.0227, "step": 62970 }, { "grad_norm": 0.44494500756263733, "learning_rate": 3.301799596247255e-05, "loss": 0.03, "step": 62980 }, { "grad_norm": 0.8803572654724121, "learning_rate": 3.300244509029854e-05, "loss": 0.031, "step": 62990 }, { "grad_norm": 0.3521803915500641, "learning_rate": 3.2986896076951585e-05, "loss": 0.0168, "step": 63000 }, { "grad_norm": 0.25514450669288635, "learning_rate": 3.297134892413211e-05, "loss": 0.0312, "step": 63010 }, { "grad_norm": 0.28857383131980896, "learning_rate": 3.295580363354031e-05, "loss": 0.0272, "step": 63020 }, { "grad_norm": 0.4060826003551483, "learning_rate": 3.2940260206876194e-05, "loss": 0.0182, "step": 63030 }, { "grad_norm": 0.36279162764549255, "learning_rate": 3.292471864583958e-05, "loss": 0.0207, "step": 63040 }, { "grad_norm": 0.7675402164459229, "learning_rate": 3.290917895213004e-05, "loss": 0.0218, "step": 63050 }, { "grad_norm": 0.26332637667655945, "learning_rate": 3.289364112744701e-05, "loss": 0.0187, "step": 63060 }, { "grad_norm": 0.47731098532676697, "learning_rate": 3.287810517348967e-05, "loss": 0.0278, "step": 63070 }, { "grad_norm": 0.39852428436279297, "learning_rate": 3.2862571091957014e-05, "loss": 0.0325, "step": 63080 }, { "grad_norm": 0.4391220808029175, "learning_rate": 3.284703888454781e-05, "loss": 0.0199, "step": 63090 }, { "grad_norm": 1.469948410987854, "learning_rate": 3.2831508552960646e-05, "loss": 0.0321, "step": 63100 }, { "grad_norm": 0.4305415749549866, "learning_rate": 3.281598009889387e-05, "loss": 0.0255, "step": 63110 }, { "grad_norm": 0.22319529950618744, "learning_rate": 3.28004535240457e-05, "loss": 0.0291, "step": 63120 }, { "grad_norm": 0.3342220187187195, "learning_rate": 3.278492883011407e-05, "loss": 0.0241, "step": 63130 }, { "grad_norm": 0.9109407067298889, "learning_rate": 3.276940601879675e-05, "loss": 0.0227, "step": 63140 }, { "grad_norm": 0.5146210193634033, "learning_rate": 3.275388509179127e-05, "loss": 0.0301, "step": 63150 }, { "grad_norm": 0.31629472970962524, "learning_rate": 3.2738366050794985e-05, "loss": 0.0314, "step": 63160 }, { "grad_norm": 0.3240079879760742, "learning_rate": 3.2722848897505034e-05, "loss": 0.0206, "step": 63170 }, { "grad_norm": 0.6657239198684692, "learning_rate": 3.2707333633618356e-05, "loss": 0.023, "step": 63180 }, { "grad_norm": 1.022111177444458, "learning_rate": 3.269182026083167e-05, "loss": 0.0225, "step": 63190 }, { "grad_norm": 0.22740387916564941, "learning_rate": 3.2676308780841516e-05, "loss": 0.0377, "step": 63200 }, { "grad_norm": 0.27061858773231506, "learning_rate": 3.266079919534416e-05, "loss": 0.0168, "step": 63210 }, { "grad_norm": 0.3592807948589325, "learning_rate": 3.264529150603575e-05, "loss": 0.0286, "step": 63220 }, { "grad_norm": 0.30270999670028687, "learning_rate": 3.262978571461217e-05, "loss": 0.0249, "step": 63230 }, { "grad_norm": 0.4300633668899536, "learning_rate": 3.26142818227691e-05, "loss": 0.0257, "step": 63240 }, { "grad_norm": 0.560641884803772, "learning_rate": 3.259877983220203e-05, "loss": 0.027, "step": 63250 }, { "grad_norm": 0.5786342024803162, "learning_rate": 3.2583279744606246e-05, "loss": 0.0304, "step": 63260 }, { "grad_norm": 0.21690164506435394, "learning_rate": 3.25677815616768e-05, "loss": 0.0189, "step": 63270 }, { "grad_norm": 0.4051227569580078, "learning_rate": 3.255228528510856e-05, "loss": 0.0187, "step": 63280 }, { "grad_norm": 0.2934202253818512, "learning_rate": 3.253679091659617e-05, "loss": 0.0187, "step": 63290 }, { "grad_norm": 0.3066652715206146, "learning_rate": 3.252129845783404e-05, "loss": 0.025, "step": 63300 }, { "grad_norm": 0.4491553008556366, "learning_rate": 3.2505807910516456e-05, "loss": 0.0256, "step": 63310 }, { "grad_norm": 0.2579970359802246, "learning_rate": 3.249031927633742e-05, "loss": 0.0266, "step": 63320 }, { "grad_norm": 0.6232632398605347, "learning_rate": 3.2474832556990737e-05, "loss": 0.0204, "step": 63330 }, { "grad_norm": 0.23528006672859192, "learning_rate": 3.245934775417002e-05, "loss": 0.0166, "step": 63340 }, { "grad_norm": 0.44955429434776306, "learning_rate": 3.244386486956866e-05, "loss": 0.0232, "step": 63350 }, { "grad_norm": 0.26345595717430115, "learning_rate": 3.2428383904879846e-05, "loss": 0.0235, "step": 63360 }, { "grad_norm": 0.4028763771057129, "learning_rate": 3.241290486179652e-05, "loss": 0.0365, "step": 63370 }, { "grad_norm": 0.3190040588378906, "learning_rate": 3.239742774201151e-05, "loss": 0.0248, "step": 63380 }, { "grad_norm": 0.27431708574295044, "learning_rate": 3.238195254721733e-05, "loss": 0.0208, "step": 63390 }, { "grad_norm": 0.32245615124702454, "learning_rate": 3.236647927910633e-05, "loss": 0.021, "step": 63400 }, { "grad_norm": 0.35012879967689514, "learning_rate": 3.2351007939370646e-05, "loss": 0.0216, "step": 63410 }, { "grad_norm": 0.3704194128513336, "learning_rate": 3.2335538529702184e-05, "loss": 0.0302, "step": 63420 }, { "grad_norm": 0.35088133811950684, "learning_rate": 3.232007105179268e-05, "loss": 0.0215, "step": 63430 }, { "grad_norm": 0.6077689528465271, "learning_rate": 3.230460550733363e-05, "loss": 0.0163, "step": 63440 }, { "grad_norm": 0.2946222424507141, "learning_rate": 3.2289141898016315e-05, "loss": 0.021, "step": 63450 }, { "grad_norm": 0.31813499331474304, "learning_rate": 3.227368022553181e-05, "loss": 0.0273, "step": 63460 }, { "grad_norm": 0.5997827649116516, "learning_rate": 3.225822049157099e-05, "loss": 0.0275, "step": 63470 }, { "grad_norm": 0.38379159569740295, "learning_rate": 3.2242762697824494e-05, "loss": 0.0184, "step": 63480 }, { "grad_norm": 0.41500359773635864, "learning_rate": 3.222730684598279e-05, "loss": 0.0309, "step": 63490 }, { "grad_norm": 0.38317644596099854, "learning_rate": 3.221185293773605e-05, "loss": 0.0347, "step": 63500 }, { "grad_norm": 0.32133355736732483, "learning_rate": 3.2196400974774354e-05, "loss": 0.0238, "step": 63510 }, { "grad_norm": 0.24447397887706757, "learning_rate": 3.2180950958787485e-05, "loss": 0.0396, "step": 63520 }, { "grad_norm": 0.48156923055648804, "learning_rate": 3.2165502891465024e-05, "loss": 0.0188, "step": 63530 }, { "grad_norm": 0.302298903465271, "learning_rate": 3.215005677449636e-05, "loss": 0.0189, "step": 63540 }, { "grad_norm": 0.2647283673286438, "learning_rate": 3.2134612609570644e-05, "loss": 0.0235, "step": 63550 }, { "grad_norm": 0.3168220520019531, "learning_rate": 3.211917039837682e-05, "loss": 0.0239, "step": 63560 }, { "grad_norm": 0.46038153767585754, "learning_rate": 3.210373014260365e-05, "loss": 0.0362, "step": 63570 }, { "grad_norm": 0.30288538336753845, "learning_rate": 3.208829184393965e-05, "loss": 0.0149, "step": 63580 }, { "grad_norm": 1.182807445526123, "learning_rate": 3.207285550407312e-05, "loss": 0.0346, "step": 63590 }, { "grad_norm": 0.23641051352024078, "learning_rate": 3.205742112469217e-05, "loss": 0.0236, "step": 63600 }, { "grad_norm": 0.5418184399604797, "learning_rate": 3.204198870748465e-05, "loss": 0.0303, "step": 63610 }, { "grad_norm": 0.24855445325374603, "learning_rate": 3.202655825413824e-05, "loss": 0.0257, "step": 63620 }, { "grad_norm": 0.33510062098503113, "learning_rate": 3.2011129766340416e-05, "loss": 0.024, "step": 63630 }, { "grad_norm": 0.2237163633108139, "learning_rate": 3.1995703245778384e-05, "loss": 0.0226, "step": 63640 }, { "grad_norm": 0.24429041147232056, "learning_rate": 3.1980278694139186e-05, "loss": 0.0198, "step": 63650 }, { "grad_norm": 0.26676100492477417, "learning_rate": 3.196485611310962e-05, "loss": 0.0199, "step": 63660 }, { "grad_norm": 0.3920333683490753, "learning_rate": 3.1949435504376257e-05, "loss": 0.0226, "step": 63670 }, { "grad_norm": 0.18664315342903137, "learning_rate": 3.1934016869625505e-05, "loss": 0.0216, "step": 63680 }, { "grad_norm": 0.6667233109474182, "learning_rate": 3.191860021054349e-05, "loss": 0.0342, "step": 63690 }, { "grad_norm": 0.34264814853668213, "learning_rate": 3.190318552881618e-05, "loss": 0.0279, "step": 63700 }, { "grad_norm": 0.37423959374427795, "learning_rate": 3.1887772826129285e-05, "loss": 0.0293, "step": 63710 }, { "grad_norm": 0.2411840856075287, "learning_rate": 3.1872362104168335e-05, "loss": 0.0207, "step": 63720 }, { "grad_norm": 0.3677484691143036, "learning_rate": 3.185695336461861e-05, "loss": 0.0254, "step": 63730 }, { "grad_norm": 0.35118427872657776, "learning_rate": 3.184154660916519e-05, "loss": 0.0149, "step": 63740 }, { "grad_norm": 0.2949081063270569, "learning_rate": 3.1826141839492906e-05, "loss": 0.0174, "step": 63750 }, { "grad_norm": 0.5912529826164246, "learning_rate": 3.181073905728644e-05, "loss": 0.0325, "step": 63760 }, { "grad_norm": 0.4626825451850891, "learning_rate": 3.1795338264230214e-05, "loss": 0.025, "step": 63770 }, { "grad_norm": 0.5177993178367615, "learning_rate": 3.1779939462008426e-05, "loss": 0.0338, "step": 63780 }, { "grad_norm": 0.6214395761489868, "learning_rate": 3.1764542652305066e-05, "loss": 0.0234, "step": 63790 }, { "grad_norm": 0.37287256121635437, "learning_rate": 3.17491478368039e-05, "loss": 0.0218, "step": 63800 }, { "grad_norm": 0.32884618639945984, "learning_rate": 3.173375501718847e-05, "loss": 0.0216, "step": 63810 }, { "grad_norm": 0.2105170041322708, "learning_rate": 3.171836419514215e-05, "loss": 0.0214, "step": 63820 }, { "grad_norm": 0.470283567905426, "learning_rate": 3.170297537234803e-05, "loss": 0.0293, "step": 63830 }, { "grad_norm": 0.2941291630268097, "learning_rate": 3.168758855048901e-05, "loss": 0.0246, "step": 63840 }, { "grad_norm": 0.20888078212738037, "learning_rate": 3.1672203731247785e-05, "loss": 0.028, "step": 63850 }, { "grad_norm": 0.4689333140850067, "learning_rate": 3.16568209163068e-05, "loss": 0.0189, "step": 63860 }, { "grad_norm": 0.4968782961368561, "learning_rate": 3.1641440107348295e-05, "loss": 0.0271, "step": 63870 }, { "grad_norm": 0.42544257640838623, "learning_rate": 3.162606130605429e-05, "loss": 0.0229, "step": 63880 }, { "grad_norm": 0.3042657673358917, "learning_rate": 3.16106845141066e-05, "loss": 0.0243, "step": 63890 }, { "grad_norm": 0.43306660652160645, "learning_rate": 3.159530973318682e-05, "loss": 0.0229, "step": 63900 }, { "grad_norm": 0.493266224861145, "learning_rate": 3.1579936964976286e-05, "loss": 0.0221, "step": 63910 }, { "grad_norm": 0.33932098746299744, "learning_rate": 3.156456621115615e-05, "loss": 0.0314, "step": 63920 }, { "grad_norm": 0.3421765863895416, "learning_rate": 3.154919747340734e-05, "loss": 0.0219, "step": 63930 }, { "grad_norm": 0.4556009769439697, "learning_rate": 3.153383075341054e-05, "loss": 0.017, "step": 63940 }, { "grad_norm": 0.9592500329017639, "learning_rate": 3.1518466052846264e-05, "loss": 0.0234, "step": 63950 }, { "grad_norm": 0.3458321988582611, "learning_rate": 3.150310337339474e-05, "loss": 0.0222, "step": 63960 }, { "grad_norm": 0.6129643321037292, "learning_rate": 3.148774271673603e-05, "loss": 0.0238, "step": 63970 }, { "grad_norm": 0.25269249081611633, "learning_rate": 3.147238408454995e-05, "loss": 0.0211, "step": 63980 }, { "grad_norm": 1.2054494619369507, "learning_rate": 3.1457027478516086e-05, "loss": 0.0202, "step": 63990 }, { "grad_norm": 0.321661114692688, "learning_rate": 3.144167290031382e-05, "loss": 0.0219, "step": 64000 }, { "grad_norm": 0.3273581564426422, "learning_rate": 3.142632035162227e-05, "loss": 0.0215, "step": 64010 }, { "grad_norm": 0.3011666536331177, "learning_rate": 3.141096983412044e-05, "loss": 0.0169, "step": 64020 }, { "grad_norm": 0.367443323135376, "learning_rate": 3.139562134948699e-05, "loss": 0.0216, "step": 64030 }, { "grad_norm": 0.8557170629501343, "learning_rate": 3.1380274899400424e-05, "loss": 0.0247, "step": 64040 }, { "grad_norm": 0.5853665471076965, "learning_rate": 3.1364930485539e-05, "loss": 0.0209, "step": 64050 }, { "grad_norm": 0.2761857807636261, "learning_rate": 3.134958810958076e-05, "loss": 0.0215, "step": 64060 }, { "grad_norm": 0.22265085577964783, "learning_rate": 3.13342477732035e-05, "loss": 0.0227, "step": 64070 }, { "grad_norm": 0.21904708445072174, "learning_rate": 3.131890947808487e-05, "loss": 0.0247, "step": 64080 }, { "grad_norm": 0.43813556432724, "learning_rate": 3.13035732259022e-05, "loss": 0.0211, "step": 64090 }, { "grad_norm": 0.4028652012348175, "learning_rate": 3.1288239018332656e-05, "loss": 0.0211, "step": 64100 }, { "grad_norm": 0.33749309182167053, "learning_rate": 3.1272906857053164e-05, "loss": 0.0254, "step": 64110 }, { "grad_norm": 0.2138042151927948, "learning_rate": 3.125757674374042e-05, "loss": 0.0258, "step": 64120 }, { "grad_norm": 0.3081760108470917, "learning_rate": 3.124224868007088e-05, "loss": 0.0221, "step": 64130 }, { "grad_norm": 0.4429628252983093, "learning_rate": 3.1226922667720854e-05, "loss": 0.0197, "step": 64140 }, { "grad_norm": 0.3836100399494171, "learning_rate": 3.121159870836633e-05, "loss": 0.0286, "step": 64150 }, { "grad_norm": 0.5313862562179565, "learning_rate": 3.119627680368312e-05, "loss": 0.0225, "step": 64160 }, { "grad_norm": 0.2577914297580719, "learning_rate": 3.1180956955346816e-05, "loss": 0.0245, "step": 64170 }, { "grad_norm": 0.23249304294586182, "learning_rate": 3.116563916503276e-05, "loss": 0.0241, "step": 64180 }, { "grad_norm": 0.2585941553115845, "learning_rate": 3.1150323434416095e-05, "loss": 0.0207, "step": 64190 }, { "grad_norm": 0.2511335015296936, "learning_rate": 3.1135009765171696e-05, "loss": 0.0157, "step": 64200 }, { "grad_norm": 0.38916537165641785, "learning_rate": 3.111969815897429e-05, "loss": 0.0213, "step": 64210 }, { "grad_norm": 0.3843024969100952, "learning_rate": 3.1104388617498295e-05, "loss": 0.0198, "step": 64220 }, { "grad_norm": 0.26518014073371887, "learning_rate": 3.1089081142417966e-05, "loss": 0.019, "step": 64230 }, { "grad_norm": 0.3403163552284241, "learning_rate": 3.1073775735407285e-05, "loss": 0.0227, "step": 64240 }, { "grad_norm": 0.29394620656967163, "learning_rate": 3.1058472398140035e-05, "loss": 0.0239, "step": 64250 }, { "grad_norm": 0.4276445209980011, "learning_rate": 3.1043171132289736e-05, "loss": 0.0189, "step": 64260 }, { "grad_norm": 0.37901410460472107, "learning_rate": 3.1027871939529774e-05, "loss": 0.0199, "step": 64270 }, { "grad_norm": 0.47905319929122925, "learning_rate": 3.1012574821533214e-05, "loss": 0.0257, "step": 64280 }, { "grad_norm": 0.4423626661300659, "learning_rate": 3.099727977997291e-05, "loss": 0.0255, "step": 64290 }, { "grad_norm": 0.16362284123897552, "learning_rate": 3.098198681652154e-05, "loss": 0.0257, "step": 64300 }, { "grad_norm": 0.38109466433525085, "learning_rate": 3.0966695932851475e-05, "loss": 0.0265, "step": 64310 }, { "grad_norm": 0.5473411083221436, "learning_rate": 3.0951407130634926e-05, "loss": 0.0309, "step": 64320 }, { "grad_norm": 0.25160732865333557, "learning_rate": 3.0936120411543864e-05, "loss": 0.0178, "step": 64330 }, { "grad_norm": 0.3827322721481323, "learning_rate": 3.0920835777250015e-05, "loss": 0.0232, "step": 64340 }, { "grad_norm": 0.31954532861709595, "learning_rate": 3.090555322942488e-05, "loss": 0.024, "step": 64350 }, { "grad_norm": 0.2563348710536957, "learning_rate": 3.089027276973973e-05, "loss": 0.0247, "step": 64360 }, { "grad_norm": 0.27418282628059387, "learning_rate": 3.0874994399865615e-05, "loss": 0.0187, "step": 64370 }, { "grad_norm": 0.43193602561950684, "learning_rate": 3.085971812147337e-05, "loss": 0.0305, "step": 64380 }, { "grad_norm": 0.21144099533557892, "learning_rate": 3.0844443936233557e-05, "loss": 0.0206, "step": 64390 }, { "grad_norm": 0.33032405376434326, "learning_rate": 3.082917184581657e-05, "loss": 0.0246, "step": 64400 }, { "grad_norm": 0.18944016098976135, "learning_rate": 3.0813901851892513e-05, "loss": 0.0192, "step": 64410 }, { "grad_norm": 0.3560140132904053, "learning_rate": 3.0798633956131316e-05, "loss": 0.03, "step": 64420 }, { "grad_norm": 0.22053204476833344, "learning_rate": 3.078336816020264e-05, "loss": 0.0237, "step": 64430 }, { "grad_norm": 0.22361372411251068, "learning_rate": 3.0768104465775935e-05, "loss": 0.0204, "step": 64440 }, { "grad_norm": 0.34203749895095825, "learning_rate": 3.0752842874520396e-05, "loss": 0.0201, "step": 64450 }, { "grad_norm": 0.5758580565452576, "learning_rate": 3.073758338810503e-05, "loss": 0.0246, "step": 64460 }, { "grad_norm": 0.5307965278625488, "learning_rate": 3.072232600819859e-05, "loss": 0.0209, "step": 64470 }, { "grad_norm": 1.0748661756515503, "learning_rate": 3.070707073646959e-05, "loss": 0.0298, "step": 64480 }, { "grad_norm": 0.6082956194877625, "learning_rate": 3.069181757458633e-05, "loss": 0.0177, "step": 64490 }, { "grad_norm": 0.46053558588027954, "learning_rate": 3.067656652421687e-05, "loss": 0.0208, "step": 64500 }, { "grad_norm": 0.2728267014026642, "learning_rate": 3.0661317587029025e-05, "loss": 0.0209, "step": 64510 }, { "grad_norm": 0.33512061834335327, "learning_rate": 3.064607076469043e-05, "loss": 0.0294, "step": 64520 }, { "grad_norm": 0.4129727780818939, "learning_rate": 3.063082605886845e-05, "loss": 0.0209, "step": 64530 }, { "grad_norm": 0.3122476637363434, "learning_rate": 3.061558347123019e-05, "loss": 0.0302, "step": 64540 }, { "grad_norm": 0.14358924329280853, "learning_rate": 3.060034300344259e-05, "loss": 0.0216, "step": 64550 }, { "grad_norm": 0.8075181245803833, "learning_rate": 3.058510465717231e-05, "loss": 0.0214, "step": 64560 }, { "grad_norm": 0.28967729210853577, "learning_rate": 3.0569868434085804e-05, "loss": 0.0309, "step": 64570 }, { "grad_norm": 0.8660441637039185, "learning_rate": 3.055463433584924e-05, "loss": 0.021, "step": 64580 }, { "grad_norm": 0.29025015234947205, "learning_rate": 3.0539402364128656e-05, "loss": 0.0198, "step": 64590 }, { "grad_norm": 2.036562919616699, "learning_rate": 3.052417252058977e-05, "loss": 0.022, "step": 64600 }, { "grad_norm": 0.35206031799316406, "learning_rate": 3.0508944806898092e-05, "loss": 0.0293, "step": 64610 }, { "grad_norm": 0.27433863282203674, "learning_rate": 3.04937192247189e-05, "loss": 0.0288, "step": 64620 }, { "grad_norm": 0.31796199083328247, "learning_rate": 3.047849577571725e-05, "loss": 0.0413, "step": 64630 }, { "grad_norm": 0.4888708293437958, "learning_rate": 3.046327446155795e-05, "loss": 0.0244, "step": 64640 }, { "grad_norm": 0.32222652435302734, "learning_rate": 3.0448055283905587e-05, "loss": 0.0194, "step": 64650 }, { "grad_norm": 0.3061160743236542, "learning_rate": 3.0432838244424494e-05, "loss": 0.027, "step": 64660 }, { "grad_norm": 0.46130624413490295, "learning_rate": 3.04176233447788e-05, "loss": 0.0209, "step": 64670 }, { "grad_norm": 0.4585167169570923, "learning_rate": 3.040241058663238e-05, "loss": 0.026, "step": 64680 }, { "grad_norm": 0.37663179636001587, "learning_rate": 3.038719997164887e-05, "loss": 0.0296, "step": 64690 }, { "grad_norm": 0.26578912138938904, "learning_rate": 3.0371991501491683e-05, "loss": 0.0301, "step": 64700 }, { "grad_norm": 0.40511009097099304, "learning_rate": 3.0356785177823975e-05, "loss": 0.021, "step": 64710 }, { "grad_norm": 0.3067811131477356, "learning_rate": 3.0341581002308726e-05, "loss": 0.0197, "step": 64720 }, { "grad_norm": 0.48034441471099854, "learning_rate": 3.0326378976608627e-05, "loss": 0.0221, "step": 64730 }, { "grad_norm": 0.5269960761070251, "learning_rate": 3.0311179102386146e-05, "loss": 0.0256, "step": 64740 }, { "grad_norm": 0.2839542329311371, "learning_rate": 3.0295981381303513e-05, "loss": 0.0203, "step": 64750 }, { "grad_norm": 0.4258054196834564, "learning_rate": 3.0280785815022727e-05, "loss": 0.0245, "step": 64760 }, { "grad_norm": 0.42495906352996826, "learning_rate": 3.0265592405205544e-05, "loss": 0.0215, "step": 64770 }, { "grad_norm": 0.3710996210575104, "learning_rate": 3.0250401153513517e-05, "loss": 0.0231, "step": 64780 }, { "grad_norm": 0.42102596163749695, "learning_rate": 3.0235212061607936e-05, "loss": 0.0259, "step": 64790 }, { "grad_norm": 0.2941407263278961, "learning_rate": 3.0220025131149842e-05, "loss": 0.0221, "step": 64800 }, { "grad_norm": 0.13928575813770294, "learning_rate": 3.020484036380007e-05, "loss": 0.0171, "step": 64810 }, { "grad_norm": 0.30198192596435547, "learning_rate": 3.0189657761219188e-05, "loss": 0.0232, "step": 64820 }, { "grad_norm": 0.2060975730419159, "learning_rate": 3.017447732506753e-05, "loss": 0.0238, "step": 64830 }, { "grad_norm": 0.2969757914543152, "learning_rate": 3.015929905700524e-05, "loss": 0.0222, "step": 64840 }, { "grad_norm": 0.12756524980068207, "learning_rate": 3.014412295869218e-05, "loss": 0.0174, "step": 64850 }, { "grad_norm": 0.30677923560142517, "learning_rate": 3.0128949031787977e-05, "loss": 0.0242, "step": 64860 }, { "grad_norm": 0.39194896817207336, "learning_rate": 3.0113777277952022e-05, "loss": 0.0232, "step": 64870 }, { "grad_norm": 0.27975761890411377, "learning_rate": 3.0098607698843496e-05, "loss": 0.0264, "step": 64880 }, { "grad_norm": 0.8268909454345703, "learning_rate": 3.0083440296121312e-05, "loss": 0.0321, "step": 64890 }, { "grad_norm": 0.5770745873451233, "learning_rate": 3.0068275071444142e-05, "loss": 0.0246, "step": 64900 }, { "grad_norm": 0.30032384395599365, "learning_rate": 3.0053112026470442e-05, "loss": 0.032, "step": 64910 }, { "grad_norm": 0.3440760672092438, "learning_rate": 3.0037951162858434e-05, "loss": 0.0243, "step": 64920 }, { "grad_norm": 0.3863745927810669, "learning_rate": 3.002279248226607e-05, "loss": 0.0161, "step": 64930 }, { "grad_norm": 0.24868367612361908, "learning_rate": 3.000763598635109e-05, "loss": 0.0228, "step": 64940 }, { "grad_norm": 0.5206490159034729, "learning_rate": 2.999248167677098e-05, "loss": 0.0221, "step": 64950 }, { "grad_norm": 0.4291365146636963, "learning_rate": 2.9977329555182976e-05, "loss": 0.0219, "step": 64960 }, { "grad_norm": 0.31781283020973206, "learning_rate": 2.9962179623244124e-05, "loss": 0.0246, "step": 64970 }, { "grad_norm": 0.22142191231250763, "learning_rate": 2.9947031882611175e-05, "loss": 0.0285, "step": 64980 }, { "grad_norm": 0.6307939887046814, "learning_rate": 2.9931886334940685e-05, "loss": 0.0223, "step": 64990 }, { "grad_norm": 0.2889912724494934, "learning_rate": 2.9916742981888923e-05, "loss": 0.0288, "step": 65000 }, { "grad_norm": 2.64505672454834, "learning_rate": 2.990160182511196e-05, "loss": 0.0362, "step": 65010 }, { "grad_norm": 0.2744062840938568, "learning_rate": 2.9886462866265574e-05, "loss": 0.0277, "step": 65020 }, { "grad_norm": 0.41929560899734497, "learning_rate": 2.9871326107005392e-05, "loss": 0.0198, "step": 65030 }, { "grad_norm": 0.4920194149017334, "learning_rate": 2.9856191548986735e-05, "loss": 0.0281, "step": 65040 }, { "grad_norm": 0.3084554672241211, "learning_rate": 2.9841059193864672e-05, "loss": 0.0197, "step": 65050 }, { "grad_norm": 0.33363014459609985, "learning_rate": 2.982592904329407e-05, "loss": 0.0267, "step": 65060 }, { "grad_norm": 0.28800103068351746, "learning_rate": 2.981080109892954e-05, "loss": 0.0305, "step": 65070 }, { "grad_norm": 0.5232083797454834, "learning_rate": 2.9795675362425434e-05, "loss": 0.0216, "step": 65080 }, { "grad_norm": 0.294028639793396, "learning_rate": 2.97805518354359e-05, "loss": 0.0246, "step": 65090 }, { "grad_norm": 0.4841442108154297, "learning_rate": 2.9765430519614825e-05, "loss": 0.0237, "step": 65100 }, { "grad_norm": 0.39770767092704773, "learning_rate": 2.975031141661584e-05, "loss": 0.0249, "step": 65110 }, { "grad_norm": 0.7971009612083435, "learning_rate": 2.9735194528092347e-05, "loss": 0.0281, "step": 65120 }, { "grad_norm": 0.4681215286254883, "learning_rate": 2.9720079855697525e-05, "loss": 0.0341, "step": 65130 }, { "grad_norm": 0.3630310297012329, "learning_rate": 2.970496740108427e-05, "loss": 0.0334, "step": 65140 }, { "grad_norm": 0.25188013911247253, "learning_rate": 2.968985716590527e-05, "loss": 0.0218, "step": 65150 }, { "grad_norm": 0.34881681203842163, "learning_rate": 2.9674749151812953e-05, "loss": 0.0192, "step": 65160 }, { "grad_norm": 0.2304219752550125, "learning_rate": 2.965964336045951e-05, "loss": 0.0226, "step": 65170 }, { "grad_norm": 0.9558779001235962, "learning_rate": 2.9644539793496894e-05, "loss": 0.0181, "step": 65180 }, { "grad_norm": 0.34271496534347534, "learning_rate": 2.9629438452576807e-05, "loss": 0.0239, "step": 65190 }, { "grad_norm": 0.3899625837802887, "learning_rate": 2.9614339339350698e-05, "loss": 0.0216, "step": 65200 }, { "grad_norm": 0.21409624814987183, "learning_rate": 2.959924245546979e-05, "loss": 0.0234, "step": 65210 }, { "grad_norm": 0.4188823401927948, "learning_rate": 2.9584147802585038e-05, "loss": 0.0249, "step": 65220 }, { "grad_norm": 0.2015790045261383, "learning_rate": 2.956905538234721e-05, "loss": 0.017, "step": 65230 }, { "grad_norm": 0.39671188592910767, "learning_rate": 2.9553965196406764e-05, "loss": 0.0257, "step": 65240 }, { "grad_norm": 0.49833059310913086, "learning_rate": 2.9538877246413943e-05, "loss": 0.0215, "step": 65250 }, { "grad_norm": 0.33946728706359863, "learning_rate": 2.952379153401874e-05, "loss": 0.0225, "step": 65260 }, { "grad_norm": 0.41037020087242126, "learning_rate": 2.9508708060870914e-05, "loss": 0.0234, "step": 65270 }, { "grad_norm": 1.1227058172225952, "learning_rate": 2.9493626828619948e-05, "loss": 0.0203, "step": 65280 }, { "grad_norm": 0.30909329652786255, "learning_rate": 2.9478547838915137e-05, "loss": 0.0241, "step": 65290 }, { "grad_norm": 0.2885662317276001, "learning_rate": 2.946347109340548e-05, "loss": 0.0221, "step": 65300 }, { "grad_norm": 0.33357518911361694, "learning_rate": 2.944839659373976e-05, "loss": 0.0211, "step": 65310 }, { "grad_norm": 0.34540924429893494, "learning_rate": 2.943332434156647e-05, "loss": 0.0253, "step": 65320 }, { "grad_norm": 0.36451011896133423, "learning_rate": 2.9418254338533912e-05, "loss": 0.026, "step": 65330 }, { "grad_norm": 0.22667793929576874, "learning_rate": 2.940318658629011e-05, "loss": 0.0245, "step": 65340 }, { "grad_norm": 0.41194456815719604, "learning_rate": 2.938812108648287e-05, "loss": 0.0245, "step": 65350 }, { "grad_norm": 0.2519077658653259, "learning_rate": 2.937305784075972e-05, "loss": 0.022, "step": 65360 }, { "grad_norm": 0.26007235050201416, "learning_rate": 2.9357996850767937e-05, "loss": 0.0183, "step": 65370 }, { "grad_norm": 0.37516796588897705, "learning_rate": 2.9342938118154595e-05, "loss": 0.0226, "step": 65380 }, { "grad_norm": 0.6113616228103638, "learning_rate": 2.9327881644566485e-05, "loss": 0.0265, "step": 65390 }, { "grad_norm": 0.5458047389984131, "learning_rate": 2.9312827431650157e-05, "loss": 0.0223, "step": 65400 }, { "grad_norm": 0.3034141957759857, "learning_rate": 2.92977754810519e-05, "loss": 0.0226, "step": 65410 }, { "grad_norm": 0.4885239601135254, "learning_rate": 2.928272579441781e-05, "loss": 0.0235, "step": 65420 }, { "grad_norm": 0.23662063479423523, "learning_rate": 2.9267678373393676e-05, "loss": 0.0246, "step": 65430 }, { "grad_norm": 0.43792369961738586, "learning_rate": 2.925263321962507e-05, "loss": 0.0196, "step": 65440 }, { "grad_norm": 0.3495957851409912, "learning_rate": 2.9237590334757292e-05, "loss": 0.027, "step": 65450 }, { "grad_norm": 0.5478291511535645, "learning_rate": 2.922254972043543e-05, "loss": 0.0224, "step": 65460 }, { "grad_norm": 0.5882574319839478, "learning_rate": 2.9207511378304265e-05, "loss": 0.0295, "step": 65470 }, { "grad_norm": 0.2832655608654022, "learning_rate": 2.9192475310008428e-05, "loss": 0.0208, "step": 65480 }, { "grad_norm": 0.26991695165634155, "learning_rate": 2.917744151719219e-05, "loss": 0.0177, "step": 65490 }, { "grad_norm": 0.2593878209590912, "learning_rate": 2.916241000149964e-05, "loss": 0.0274, "step": 65500 }, { "grad_norm": 0.28969264030456543, "learning_rate": 2.9147380764574622e-05, "loss": 0.0167, "step": 65510 }, { "grad_norm": 0.3509660065174103, "learning_rate": 2.9132353808060674e-05, "loss": 0.0215, "step": 65520 }, { "grad_norm": 0.5808952450752258, "learning_rate": 2.9117329133601134e-05, "loss": 0.0237, "step": 65530 }, { "grad_norm": 0.35025107860565186, "learning_rate": 2.9102306742839093e-05, "loss": 0.0231, "step": 65540 }, { "grad_norm": 0.357330858707428, "learning_rate": 2.9087286637417375e-05, "loss": 0.0176, "step": 65550 }, { "grad_norm": 0.4128030240535736, "learning_rate": 2.9072268818978537e-05, "loss": 0.0228, "step": 65560 }, { "grad_norm": 0.4091929495334625, "learning_rate": 2.905725328916493e-05, "loss": 0.0197, "step": 65570 }, { "grad_norm": 0.20827113091945648, "learning_rate": 2.9042240049618608e-05, "loss": 0.0257, "step": 65580 }, { "grad_norm": 0.7709453105926514, "learning_rate": 2.9027229101981413e-05, "loss": 0.0258, "step": 65590 }, { "grad_norm": 0.17008651793003082, "learning_rate": 2.901222044789488e-05, "loss": 0.0242, "step": 65600 }, { "grad_norm": 0.3418312966823578, "learning_rate": 2.89972140890004e-05, "loss": 0.0218, "step": 65610 }, { "grad_norm": 0.5401558876037598, "learning_rate": 2.8982210026938988e-05, "loss": 0.0295, "step": 65620 }, { "grad_norm": 0.18968744575977325, "learning_rate": 2.896720826335151e-05, "loss": 0.0341, "step": 65630 }, { "grad_norm": 0.41033220291137695, "learning_rate": 2.8952208799878493e-05, "loss": 0.0273, "step": 65640 }, { "grad_norm": 0.2922942340373993, "learning_rate": 2.89372116381603e-05, "loss": 0.0187, "step": 65650 }, { "grad_norm": 0.32849663496017456, "learning_rate": 2.8922216779836932e-05, "loss": 0.0193, "step": 65660 }, { "grad_norm": 1.2225492000579834, "learning_rate": 2.8907224226548285e-05, "loss": 0.0208, "step": 65670 }, { "grad_norm": 0.24541088938713074, "learning_rate": 2.8892233979933868e-05, "loss": 0.0204, "step": 65680 }, { "grad_norm": 0.2897505760192871, "learning_rate": 2.8877246041633026e-05, "loss": 0.0216, "step": 65690 }, { "grad_norm": 0.37063929438591003, "learning_rate": 2.886226041328478e-05, "loss": 0.0182, "step": 65700 }, { "grad_norm": 0.3849867582321167, "learning_rate": 2.8847277096527957e-05, "loss": 0.0292, "step": 65710 }, { "grad_norm": 0.31030306220054626, "learning_rate": 2.8832296093001105e-05, "loss": 0.0257, "step": 65720 }, { "grad_norm": 0.4066809117794037, "learning_rate": 2.8817317404342547e-05, "loss": 0.0216, "step": 65730 }, { "grad_norm": 0.29469403624534607, "learning_rate": 2.880234103219029e-05, "loss": 0.021, "step": 65740 }, { "grad_norm": 0.45495086908340454, "learning_rate": 2.878736697818215e-05, "loss": 0.0188, "step": 65750 }, { "grad_norm": 0.1830746978521347, "learning_rate": 2.877239524395568e-05, "loss": 0.0207, "step": 65760 }, { "grad_norm": 0.593072235584259, "learning_rate": 2.8757425831148134e-05, "loss": 0.0304, "step": 65770 }, { "grad_norm": 0.2708989679813385, "learning_rate": 2.874245874139658e-05, "loss": 0.0296, "step": 65780 }, { "grad_norm": 0.2669605314731598, "learning_rate": 2.8727493976337737e-05, "loss": 0.0243, "step": 65790 }, { "grad_norm": 0.45941129326820374, "learning_rate": 2.8712531537608218e-05, "loss": 0.0187, "step": 65800 }, { "grad_norm": 0.3634764850139618, "learning_rate": 2.8697571426844217e-05, "loss": 0.0201, "step": 65810 }, { "grad_norm": 0.38985341787338257, "learning_rate": 2.86826136456818e-05, "loss": 0.0222, "step": 65820 }, { "grad_norm": 0.3640132546424866, "learning_rate": 2.8667658195756685e-05, "loss": 0.0204, "step": 65830 }, { "grad_norm": 0.49080073833465576, "learning_rate": 2.8652705078704412e-05, "loss": 0.0264, "step": 65840 }, { "grad_norm": 0.5588036179542542, "learning_rate": 2.863775429616018e-05, "loss": 0.0272, "step": 65850 }, { "grad_norm": 0.49477970600128174, "learning_rate": 2.862280584975906e-05, "loss": 0.0242, "step": 65860 }, { "grad_norm": 0.1772465705871582, "learning_rate": 2.8607859741135735e-05, "loss": 0.0198, "step": 65870 }, { "grad_norm": 0.2568628787994385, "learning_rate": 2.8592915971924723e-05, "loss": 0.0233, "step": 65880 }, { "grad_norm": 0.5290737748146057, "learning_rate": 2.8577974543760212e-05, "loss": 0.0256, "step": 65890 }, { "grad_norm": 0.244681254029274, "learning_rate": 2.8563035458276222e-05, "loss": 0.0171, "step": 65900 }, { "grad_norm": 0.8353995084762573, "learning_rate": 2.8548098717106424e-05, "loss": 0.0246, "step": 65910 }, { "grad_norm": 0.4733603894710541, "learning_rate": 2.85331643218843e-05, "loss": 0.0245, "step": 65920 }, { "grad_norm": 0.315101683139801, "learning_rate": 2.8518232274243046e-05, "loss": 0.0236, "step": 65930 }, { "grad_norm": 0.4115816056728363, "learning_rate": 2.8503302575815638e-05, "loss": 0.0248, "step": 65940 }, { "grad_norm": 0.4478066861629486, "learning_rate": 2.8488375228234716e-05, "loss": 0.023, "step": 65950 }, { "grad_norm": 0.3330982029438019, "learning_rate": 2.8473450233132738e-05, "loss": 0.021, "step": 65960 }, { "grad_norm": 0.7103630304336548, "learning_rate": 2.8458527592141898e-05, "loss": 0.0225, "step": 65970 }, { "grad_norm": 0.3723416030406952, "learning_rate": 2.8443607306894073e-05, "loss": 0.02, "step": 65980 }, { "grad_norm": 0.1827624887228012, "learning_rate": 2.8428689379020946e-05, "loss": 0.0192, "step": 65990 }, { "grad_norm": 1.243536114692688, "learning_rate": 2.841377381015391e-05, "loss": 0.0273, "step": 66000 }, { "grad_norm": 0.31056132912635803, "learning_rate": 2.8398860601924144e-05, "loss": 0.0172, "step": 66010 }, { "grad_norm": 0.8124236464500427, "learning_rate": 2.838394975596248e-05, "loss": 0.0169, "step": 66020 }, { "grad_norm": 0.4762059450149536, "learning_rate": 2.83690412738996e-05, "loss": 0.0267, "step": 66030 }, { "grad_norm": 0.40545395016670227, "learning_rate": 2.835413515736579e-05, "loss": 0.0295, "step": 66040 }, { "grad_norm": 0.5340514183044434, "learning_rate": 2.8339231407991267e-05, "loss": 0.0243, "step": 66050 }, { "grad_norm": 0.4479151964187622, "learning_rate": 2.832433002740581e-05, "loss": 0.0251, "step": 66060 }, { "grad_norm": 0.2485704869031906, "learning_rate": 2.8309431017239047e-05, "loss": 0.0167, "step": 66070 }, { "grad_norm": 0.16202090680599213, "learning_rate": 2.829453437912028e-05, "loss": 0.0194, "step": 66080 }, { "grad_norm": 0.5130541324615479, "learning_rate": 2.8279640114678617e-05, "loss": 0.0279, "step": 66090 }, { "grad_norm": 0.6340291500091553, "learning_rate": 2.8264748225542836e-05, "loss": 0.0347, "step": 66100 }, { "grad_norm": 0.3048989474773407, "learning_rate": 2.8249858713341516e-05, "loss": 0.0194, "step": 66110 }, { "grad_norm": 0.305780291557312, "learning_rate": 2.823497157970294e-05, "loss": 0.0213, "step": 66120 }, { "grad_norm": 0.29614490270614624, "learning_rate": 2.8220086826255164e-05, "loss": 0.0192, "step": 66130 }, { "grad_norm": 0.23553436994552612, "learning_rate": 2.8205204454625934e-05, "loss": 0.0241, "step": 66140 }, { "grad_norm": 0.29919230937957764, "learning_rate": 2.8190324466442798e-05, "loss": 0.0239, "step": 66150 }, { "grad_norm": 0.6005182266235352, "learning_rate": 2.8175446863332965e-05, "loss": 0.0293, "step": 66160 }, { "grad_norm": 0.6889026761054993, "learning_rate": 2.816057164692345e-05, "loss": 0.0309, "step": 66170 }, { "grad_norm": 0.338221937417984, "learning_rate": 2.8145698818840992e-05, "loss": 0.0243, "step": 66180 }, { "grad_norm": 0.6168515086174011, "learning_rate": 2.8130828380712065e-05, "loss": 0.0216, "step": 66190 }, { "grad_norm": 0.40330323576927185, "learning_rate": 2.811596033416285e-05, "loss": 0.0195, "step": 66200 }, { "grad_norm": 0.1710941046476364, "learning_rate": 2.8101094680819307e-05, "loss": 0.0169, "step": 66210 }, { "grad_norm": 1.2577979564666748, "learning_rate": 2.808623142230715e-05, "loss": 0.0266, "step": 66220 }, { "grad_norm": 0.4064485728740692, "learning_rate": 2.8071370560251757e-05, "loss": 0.0242, "step": 66230 }, { "grad_norm": 0.47203701734542847, "learning_rate": 2.805651209627831e-05, "loss": 0.0224, "step": 66240 }, { "grad_norm": 0.30977898836135864, "learning_rate": 2.804165603201171e-05, "loss": 0.017, "step": 66250 }, { "grad_norm": 0.30653122067451477, "learning_rate": 2.802680236907661e-05, "loss": 0.0234, "step": 66260 }, { "grad_norm": 0.40179508924484253, "learning_rate": 2.8011951109097344e-05, "loss": 0.0204, "step": 66270 }, { "grad_norm": 0.4200505018234253, "learning_rate": 2.7997102253698072e-05, "loss": 0.0247, "step": 66280 }, { "grad_norm": 0.4301222562789917, "learning_rate": 2.7982255804502588e-05, "loss": 0.0222, "step": 66290 }, { "grad_norm": 0.3048946261405945, "learning_rate": 2.7967411763134516e-05, "loss": 0.0252, "step": 66300 }, { "grad_norm": 0.3225039839744568, "learning_rate": 2.7952570131217166e-05, "loss": 0.0276, "step": 66310 }, { "grad_norm": 0.3759532868862152, "learning_rate": 2.7937730910373618e-05, "loss": 0.0275, "step": 66320 }, { "grad_norm": 0.5356106758117676, "learning_rate": 2.7922894102226628e-05, "loss": 0.0259, "step": 66330 }, { "grad_norm": 1.2640196084976196, "learning_rate": 2.790805970839876e-05, "loss": 0.028, "step": 66340 }, { "grad_norm": 0.2940810024738312, "learning_rate": 2.789322773051225e-05, "loss": 0.015, "step": 66350 }, { "grad_norm": 0.6159776449203491, "learning_rate": 2.7878398170189124e-05, "loss": 0.0198, "step": 66360 }, { "grad_norm": 0.3986525535583496, "learning_rate": 2.7863571029051112e-05, "loss": 0.0157, "step": 66370 }, { "grad_norm": 0.4418579041957855, "learning_rate": 2.784874630871971e-05, "loss": 0.0232, "step": 66380 }, { "grad_norm": 0.25563251972198486, "learning_rate": 2.7833924010816086e-05, "loss": 0.0174, "step": 66390 }, { "grad_norm": 0.42886635661125183, "learning_rate": 2.7819104136961226e-05, "loss": 0.0195, "step": 66400 }, { "grad_norm": 0.24380874633789062, "learning_rate": 2.780428668877577e-05, "loss": 0.023, "step": 66410 }, { "grad_norm": 0.331642746925354, "learning_rate": 2.7789471667880152e-05, "loss": 0.0221, "step": 66420 }, { "grad_norm": 0.38600635528564453, "learning_rate": 2.7774659075894523e-05, "loss": 0.0265, "step": 66430 }, { "grad_norm": 0.42552003264427185, "learning_rate": 2.7759848914438778e-05, "loss": 0.0302, "step": 66440 }, { "grad_norm": 0.33380794525146484, "learning_rate": 2.7745041185132496e-05, "loss": 0.0198, "step": 66450 }, { "grad_norm": 0.3050999641418457, "learning_rate": 2.773023588959506e-05, "loss": 0.0213, "step": 66460 }, { "grad_norm": 0.4209077060222626, "learning_rate": 2.7715433029445555e-05, "loss": 0.0207, "step": 66470 }, { "grad_norm": 0.23473677039146423, "learning_rate": 2.7700632606302766e-05, "loss": 0.0201, "step": 66480 }, { "grad_norm": 0.5114094614982605, "learning_rate": 2.768583462178528e-05, "loss": 0.0232, "step": 66490 }, { "grad_norm": 0.1918913871049881, "learning_rate": 2.767103907751136e-05, "loss": 0.0151, "step": 66500 }, { "grad_norm": 0.27554377913475037, "learning_rate": 2.7656245975099065e-05, "loss": 0.0158, "step": 66510 }, { "grad_norm": 0.23888127505779266, "learning_rate": 2.7641455316166093e-05, "loss": 0.0271, "step": 66520 }, { "grad_norm": 0.30237478017807007, "learning_rate": 2.7626667102329973e-05, "loss": 0.0249, "step": 66530 }, { "grad_norm": 0.34604915976524353, "learning_rate": 2.761188133520788e-05, "loss": 0.0197, "step": 66540 }, { "grad_norm": 0.17473585903644562, "learning_rate": 2.7597098016416784e-05, "loss": 0.0171, "step": 66550 }, { "grad_norm": 0.29232004284858704, "learning_rate": 2.7582317147573368e-05, "loss": 0.0183, "step": 66560 }, { "grad_norm": 0.23107124865055084, "learning_rate": 2.7567538730294064e-05, "loss": 0.0223, "step": 66570 }, { "grad_norm": 0.31965598464012146, "learning_rate": 2.7552762766194973e-05, "loss": 0.024, "step": 66580 }, { "grad_norm": 0.4418955147266388, "learning_rate": 2.753798925689202e-05, "loss": 0.0325, "step": 66590 }, { "grad_norm": 0.4755910038948059, "learning_rate": 2.7523218204000774e-05, "loss": 0.0255, "step": 66600 }, { "grad_norm": 0.28705278038978577, "learning_rate": 2.7508449609136612e-05, "loss": 0.0263, "step": 66610 }, { "grad_norm": 0.366885781288147, "learning_rate": 2.749368347391454e-05, "loss": 0.0236, "step": 66620 }, { "grad_norm": 0.5597284436225891, "learning_rate": 2.7478919799949443e-05, "loss": 0.0164, "step": 66630 }, { "grad_norm": 0.2519434988498688, "learning_rate": 2.7464158588855803e-05, "loss": 0.0152, "step": 66640 }, { "grad_norm": 0.3835611343383789, "learning_rate": 2.7449399842247918e-05, "loss": 0.0209, "step": 66650 }, { "grad_norm": 0.24046826362609863, "learning_rate": 2.7434643561739736e-05, "loss": 0.0205, "step": 66660 }, { "grad_norm": 0.4257485270500183, "learning_rate": 2.7419889748945016e-05, "loss": 0.0224, "step": 66670 }, { "grad_norm": 0.2948072850704193, "learning_rate": 2.7405138405477203e-05, "loss": 0.0228, "step": 66680 }, { "grad_norm": 0.23552459478378296, "learning_rate": 2.73903895329495e-05, "loss": 0.0218, "step": 66690 }, { "grad_norm": 0.2924462556838989, "learning_rate": 2.737564313297478e-05, "loss": 0.0265, "step": 66700 }, { "grad_norm": 0.25329098105430603, "learning_rate": 2.736089920716572e-05, "loss": 0.0196, "step": 66710 }, { "grad_norm": 0.3378210961818695, "learning_rate": 2.7346157757134706e-05, "loss": 0.0211, "step": 66720 }, { "grad_norm": 0.29111024737358093, "learning_rate": 2.733141878449379e-05, "loss": 0.0152, "step": 66730 }, { "grad_norm": 0.30730685591697693, "learning_rate": 2.7316682290854844e-05, "loss": 0.0226, "step": 66740 }, { "grad_norm": 0.6010767817497253, "learning_rate": 2.7301948277829414e-05, "loss": 0.0218, "step": 66750 }, { "grad_norm": 0.4660074710845947, "learning_rate": 2.7287216747028814e-05, "loss": 0.0237, "step": 66760 }, { "grad_norm": 0.18435199558734894, "learning_rate": 2.7272487700064025e-05, "loss": 0.0227, "step": 66770 }, { "grad_norm": 0.2528565526008606, "learning_rate": 2.725776113854583e-05, "loss": 0.0261, "step": 66780 }, { "grad_norm": 0.6708462238311768, "learning_rate": 2.7243037064084664e-05, "loss": 0.0208, "step": 66790 }, { "grad_norm": 0.7860748171806335, "learning_rate": 2.7228315478290766e-05, "loss": 0.0334, "step": 66800 }, { "grad_norm": 0.2103845328092575, "learning_rate": 2.721359638277402e-05, "loss": 0.0218, "step": 66810 }, { "grad_norm": 0.298858106136322, "learning_rate": 2.7198879779144148e-05, "loss": 0.0256, "step": 66820 }, { "grad_norm": 0.32365068793296814, "learning_rate": 2.7184165669010485e-05, "loss": 0.0224, "step": 66830 }, { "grad_norm": 0.8783653974533081, "learning_rate": 2.716945405398218e-05, "loss": 0.0212, "step": 66840 }, { "grad_norm": 0.5404054522514343, "learning_rate": 2.715474493566803e-05, "loss": 0.0179, "step": 66850 }, { "grad_norm": 0.2238602489233017, "learning_rate": 2.7140038315676652e-05, "loss": 0.0201, "step": 66860 }, { "grad_norm": 0.2530215382575989, "learning_rate": 2.7125334195616275e-05, "loss": 0.0302, "step": 66870 }, { "grad_norm": 0.36864691972732544, "learning_rate": 2.7110632577094997e-05, "loss": 0.0211, "step": 66880 }, { "grad_norm": 0.2631300985813141, "learning_rate": 2.709593346172051e-05, "loss": 0.0255, "step": 66890 }, { "grad_norm": 0.2783445119857788, "learning_rate": 2.708123685110032e-05, "loss": 0.0191, "step": 66900 }, { "grad_norm": 0.3545617163181305, "learning_rate": 2.7066542746841588e-05, "loss": 0.0236, "step": 66910 }, { "grad_norm": 0.24633696675300598, "learning_rate": 2.7051851150551267e-05, "loss": 0.0191, "step": 66920 }, { "grad_norm": 0.23220664262771606, "learning_rate": 2.7037162063835996e-05, "loss": 0.0258, "step": 66930 }, { "grad_norm": 0.3994544744491577, "learning_rate": 2.702247548830218e-05, "loss": 0.0258, "step": 66940 }, { "grad_norm": 0.4797363579273224, "learning_rate": 2.7007791425555883e-05, "loss": 0.0178, "step": 66950 }, { "grad_norm": 0.1604425311088562, "learning_rate": 2.6993109877202945e-05, "loss": 0.0184, "step": 66960 }, { "grad_norm": 0.39663511514663696, "learning_rate": 2.6978430844848935e-05, "loss": 0.0243, "step": 66970 }, { "grad_norm": 0.31282976269721985, "learning_rate": 2.69637543300991e-05, "loss": 0.0188, "step": 66980 }, { "grad_norm": 0.3416732847690582, "learning_rate": 2.694908033455848e-05, "loss": 0.017, "step": 66990 }, { "grad_norm": 0.1994478404521942, "learning_rate": 2.6934408859831733e-05, "loss": 0.022, "step": 67000 }, { "grad_norm": 0.42755720019340515, "learning_rate": 2.69197399075234e-05, "loss": 0.0276, "step": 67010 }, { "grad_norm": 0.2853412330150604, "learning_rate": 2.6905073479237584e-05, "loss": 0.0162, "step": 67020 }, { "grad_norm": 0.3651603162288666, "learning_rate": 2.6890409576578235e-05, "loss": 0.0329, "step": 67030 }, { "grad_norm": 0.3704422414302826, "learning_rate": 2.687574820114892e-05, "loss": 0.0225, "step": 67040 }, { "grad_norm": 0.5270281434059143, "learning_rate": 2.686108935455305e-05, "loss": 0.0271, "step": 67050 }, { "grad_norm": 0.41745588183403015, "learning_rate": 2.6846433038393603e-05, "loss": 0.0298, "step": 67060 }, { "grad_norm": 0.2752339243888855, "learning_rate": 2.6831779254273482e-05, "loss": 0.0242, "step": 67070 }, { "grad_norm": 0.2859242558479309, "learning_rate": 2.6817128003795122e-05, "loss": 0.0168, "step": 67080 }, { "grad_norm": 0.24755708873271942, "learning_rate": 2.680247928856081e-05, "loss": 0.0219, "step": 67090 }, { "grad_norm": 0.5526158213615417, "learning_rate": 2.6787833110172466e-05, "loss": 0.0229, "step": 67100 }, { "grad_norm": 0.6020517945289612, "learning_rate": 2.6773189470231807e-05, "loss": 0.0236, "step": 67110 }, { "grad_norm": 0.3437283933162689, "learning_rate": 2.675854837034022e-05, "loss": 0.0252, "step": 67120 }, { "grad_norm": 0.31787797808647156, "learning_rate": 2.6743909812098833e-05, "loss": 0.0189, "step": 67130 }, { "grad_norm": 0.6613156199455261, "learning_rate": 2.67292737971085e-05, "loss": 0.03, "step": 67140 }, { "grad_norm": 0.33738192915916443, "learning_rate": 2.671464032696982e-05, "loss": 0.0222, "step": 67150 }, { "grad_norm": 0.3283948004245758, "learning_rate": 2.6700009403283055e-05, "loss": 0.0218, "step": 67160 }, { "grad_norm": 0.32008424401283264, "learning_rate": 2.6685381027648214e-05, "loss": 0.0177, "step": 67170 }, { "grad_norm": 0.3623867630958557, "learning_rate": 2.6670755201665086e-05, "loss": 0.0305, "step": 67180 }, { "grad_norm": 0.36602431535720825, "learning_rate": 2.6656131926933066e-05, "loss": 0.0218, "step": 67190 }, { "grad_norm": 1.9260143041610718, "learning_rate": 2.664151120505136e-05, "loss": 0.0143, "step": 67200 }, { "grad_norm": 0.34202155470848083, "learning_rate": 2.6626893037618873e-05, "loss": 0.0179, "step": 67210 }, { "grad_norm": 0.5975102782249451, "learning_rate": 2.6612277426234243e-05, "loss": 0.0262, "step": 67220 }, { "grad_norm": 0.280977725982666, "learning_rate": 2.659766437249577e-05, "loss": 0.0294, "step": 67230 }, { "grad_norm": 0.19052454829216003, "learning_rate": 2.6583053878001557e-05, "loss": 0.0177, "step": 67240 }, { "grad_norm": 0.2909024953842163, "learning_rate": 2.6568445944349323e-05, "loss": 0.0209, "step": 67250 }, { "grad_norm": 0.3421742916107178, "learning_rate": 2.6553840573136657e-05, "loss": 0.0175, "step": 67260 }, { "grad_norm": 0.440962016582489, "learning_rate": 2.6539237765960706e-05, "loss": 0.0196, "step": 67270 }, { "grad_norm": 0.2554835379123688, "learning_rate": 2.6524637524418465e-05, "loss": 0.0247, "step": 67280 }, { "grad_norm": 0.6758092045783997, "learning_rate": 2.651003985010655e-05, "loss": 0.0246, "step": 67290 }, { "grad_norm": 0.30742186307907104, "learning_rate": 2.649544474462138e-05, "loss": 0.0165, "step": 67300 }, { "grad_norm": 0.24209965765476227, "learning_rate": 2.6480852209559015e-05, "loss": 0.0279, "step": 67310 }, { "grad_norm": 0.1835554987192154, "learning_rate": 2.6466262246515296e-05, "loss": 0.0233, "step": 67320 }, { "grad_norm": 0.2277344912290573, "learning_rate": 2.6451674857085746e-05, "loss": 0.0204, "step": 67330 }, { "grad_norm": 0.6886888742446899, "learning_rate": 2.6437090042865655e-05, "loss": 0.0213, "step": 67340 }, { "grad_norm": 0.38427963852882385, "learning_rate": 2.6422507805449947e-05, "loss": 0.0201, "step": 67350 }, { "grad_norm": 0.6743014454841614, "learning_rate": 2.6407928146433357e-05, "loss": 0.0219, "step": 67360 }, { "grad_norm": 0.2858685851097107, "learning_rate": 2.639335106741026e-05, "loss": 0.0184, "step": 67370 }, { "grad_norm": 0.4514784514904022, "learning_rate": 2.6378776569974796e-05, "loss": 0.0254, "step": 67380 }, { "grad_norm": 0.4337993562221527, "learning_rate": 2.6364204655720814e-05, "loss": 0.0233, "step": 67390 }, { "grad_norm": 0.35463863611221313, "learning_rate": 2.6349635326241894e-05, "loss": 0.0292, "step": 67400 }, { "grad_norm": 0.4144255220890045, "learning_rate": 2.633506858313127e-05, "loss": 0.0157, "step": 67410 }, { "grad_norm": 0.325013130903244, "learning_rate": 2.6320504427981975e-05, "loss": 0.0233, "step": 67420 }, { "grad_norm": 0.4831215441226959, "learning_rate": 2.6305942862386734e-05, "loss": 0.027, "step": 67430 }, { "grad_norm": 0.6375423669815063, "learning_rate": 2.629138388793793e-05, "loss": 0.0286, "step": 67440 }, { "grad_norm": 0.49869275093078613, "learning_rate": 2.6276827506227743e-05, "loss": 0.0159, "step": 67450 }, { "grad_norm": 0.3404334783554077, "learning_rate": 2.6262273718848037e-05, "loss": 0.0154, "step": 67460 }, { "grad_norm": 0.36310768127441406, "learning_rate": 2.6247722527390407e-05, "loss": 0.018, "step": 67470 }, { "grad_norm": 0.2722132205963135, "learning_rate": 2.6233173933446104e-05, "loss": 0.027, "step": 67480 }, { "grad_norm": 0.3397234380245209, "learning_rate": 2.6218627938606188e-05, "loss": 0.0291, "step": 67490 }, { "grad_norm": 0.24649493396282196, "learning_rate": 2.620408454446135e-05, "loss": 0.0228, "step": 67500 }, { "grad_norm": 0.424871563911438, "learning_rate": 2.6189543752602046e-05, "loss": 0.0249, "step": 67510 }, { "grad_norm": 0.4099281132221222, "learning_rate": 2.6175005564618444e-05, "loss": 0.0242, "step": 67520 }, { "grad_norm": 0.347419798374176, "learning_rate": 2.6160469982100428e-05, "loss": 0.0242, "step": 67530 }, { "grad_norm": 0.40047869086265564, "learning_rate": 2.6145937006637555e-05, "loss": 0.0219, "step": 67540 }, { "grad_norm": 0.25233083963394165, "learning_rate": 2.6131406639819166e-05, "loss": 0.0202, "step": 67550 }, { "grad_norm": 0.2019755244255066, "learning_rate": 2.611687888323424e-05, "loss": 0.0281, "step": 67560 }, { "grad_norm": 0.2746327519416809, "learning_rate": 2.610235373847153e-05, "loss": 0.0257, "step": 67570 }, { "grad_norm": 0.6925652027130127, "learning_rate": 2.6087831207119494e-05, "loss": 0.0338, "step": 67580 }, { "grad_norm": 0.18099983036518097, "learning_rate": 2.60733112907663e-05, "loss": 0.0251, "step": 67590 }, { "grad_norm": 1.1908146142959595, "learning_rate": 2.6058793990999786e-05, "loss": 0.0338, "step": 67600 }, { "grad_norm": 0.2927172780036926, "learning_rate": 2.6044279309407593e-05, "loss": 0.0209, "step": 67610 }, { "grad_norm": 0.3696913719177246, "learning_rate": 2.6029767247576976e-05, "loss": 0.0263, "step": 67620 }, { "grad_norm": 0.36516717076301575, "learning_rate": 2.601525780709497e-05, "loss": 0.0204, "step": 67630 }, { "grad_norm": 0.25860491394996643, "learning_rate": 2.6000750989548318e-05, "loss": 0.0217, "step": 67640 }, { "grad_norm": 0.2005375176668167, "learning_rate": 2.5986246796523462e-05, "loss": 0.0158, "step": 67650 }, { "grad_norm": 0.19987080991268158, "learning_rate": 2.5971745229606546e-05, "loss": 0.0267, "step": 67660 }, { "grad_norm": 0.7364695072174072, "learning_rate": 2.5957246290383442e-05, "loss": 0.025, "step": 67670 }, { "grad_norm": 0.3060331344604492, "learning_rate": 2.594274998043975e-05, "loss": 0.0218, "step": 67680 }, { "grad_norm": 0.22139590978622437, "learning_rate": 2.5928256301360733e-05, "loss": 0.0191, "step": 67690 }, { "grad_norm": 0.28351736068725586, "learning_rate": 2.591376525473142e-05, "loss": 0.019, "step": 67700 }, { "grad_norm": 0.31267061829566956, "learning_rate": 2.5899276842136523e-05, "loss": 0.0226, "step": 67710 }, { "grad_norm": 0.26475995779037476, "learning_rate": 2.5884791065160495e-05, "loss": 0.0259, "step": 67720 }, { "grad_norm": 0.6332029700279236, "learning_rate": 2.5870307925387436e-05, "loss": 0.0201, "step": 67730 }, { "grad_norm": 0.34670817852020264, "learning_rate": 2.585582742440125e-05, "loss": 0.0224, "step": 67740 }, { "grad_norm": 0.3007964789867401, "learning_rate": 2.5841349563785455e-05, "loss": 0.0231, "step": 67750 }, { "grad_norm": 0.1409638673067093, "learning_rate": 2.5826874345123355e-05, "loss": 0.0182, "step": 67760 }, { "grad_norm": 0.2858541011810303, "learning_rate": 2.581240176999792e-05, "loss": 0.0183, "step": 67770 }, { "grad_norm": 0.27794504165649414, "learning_rate": 2.579793183999189e-05, "loss": 0.0184, "step": 67780 }, { "grad_norm": 0.32761481404304504, "learning_rate": 2.5783464556687627e-05, "loss": 0.0289, "step": 67790 }, { "grad_norm": 0.23955047130584717, "learning_rate": 2.576899992166729e-05, "loss": 0.0208, "step": 67800 }, { "grad_norm": 0.3160988986492157, "learning_rate": 2.5754537936512668e-05, "loss": 0.0236, "step": 67810 }, { "grad_norm": 0.48671257495880127, "learning_rate": 2.574007860280535e-05, "loss": 0.0268, "step": 67820 }, { "grad_norm": 0.15869522094726562, "learning_rate": 2.5725621922126518e-05, "loss": 0.0177, "step": 67830 }, { "grad_norm": 0.30770719051361084, "learning_rate": 2.571116789605722e-05, "loss": 0.0216, "step": 67840 }, { "grad_norm": 0.2506125569343567, "learning_rate": 2.5696716526178054e-05, "loss": 0.0275, "step": 67850 }, { "grad_norm": 0.5434340238571167, "learning_rate": 2.568226781406946e-05, "loss": 0.0225, "step": 67860 }, { "grad_norm": 0.34488555788993835, "learning_rate": 2.5667821761311474e-05, "loss": 0.0241, "step": 67870 }, { "grad_norm": 0.24012336134910583, "learning_rate": 2.5653378369483917e-05, "loss": 0.0166, "step": 67880 }, { "grad_norm": 0.3739297091960907, "learning_rate": 2.5638937640166292e-05, "loss": 0.0173, "step": 67890 }, { "grad_norm": 0.2187536656856537, "learning_rate": 2.5624499574937842e-05, "loss": 0.0214, "step": 67900 }, { "grad_norm": 0.4972040355205536, "learning_rate": 2.5610064175377456e-05, "loss": 0.0224, "step": 67910 }, { "grad_norm": 0.23367387056350708, "learning_rate": 2.5595631443063782e-05, "loss": 0.0199, "step": 67920 }, { "grad_norm": 0.3033953607082367, "learning_rate": 2.5581201379575177e-05, "loss": 0.0176, "step": 67930 }, { "grad_norm": 0.23875151574611664, "learning_rate": 2.5566773986489655e-05, "loss": 0.0228, "step": 67940 }, { "grad_norm": 0.3159544765949249, "learning_rate": 2.5552349265385e-05, "loss": 0.0271, "step": 67950 }, { "grad_norm": 0.43332090973854065, "learning_rate": 2.5537927217838675e-05, "loss": 0.0174, "step": 67960 }, { "grad_norm": 0.2615986168384552, "learning_rate": 2.5523507845427864e-05, "loss": 0.0253, "step": 67970 }, { "grad_norm": 0.3323262631893158, "learning_rate": 2.5509091149729413e-05, "loss": 0.0235, "step": 67980 }, { "grad_norm": 0.5720653533935547, "learning_rate": 2.5494677132319962e-05, "loss": 0.0318, "step": 67990 }, { "grad_norm": 0.45442336797714233, "learning_rate": 2.548026579477575e-05, "loss": 0.0216, "step": 68000 }, { "grad_norm": 0.2909128665924072, "learning_rate": 2.5465857138672834e-05, "loss": 0.0194, "step": 68010 }, { "grad_norm": 0.26351049542427063, "learning_rate": 2.5451451165586848e-05, "loss": 0.0261, "step": 68020 }, { "grad_norm": 0.507085919380188, "learning_rate": 2.543704787709329e-05, "loss": 0.0174, "step": 68030 }, { "grad_norm": 0.2555038034915924, "learning_rate": 2.542264727476723e-05, "loss": 0.0204, "step": 68040 }, { "grad_norm": 0.25692784786224365, "learning_rate": 2.5408249360183535e-05, "loss": 0.0293, "step": 68050 }, { "grad_norm": 0.3483445346355438, "learning_rate": 2.5393854134916694e-05, "loss": 0.0203, "step": 68060 }, { "grad_norm": 0.2972394526004791, "learning_rate": 2.5379461600540993e-05, "loss": 0.0171, "step": 68070 }, { "grad_norm": 0.37727949023246765, "learning_rate": 2.5365071758630317e-05, "loss": 0.0252, "step": 68080 }, { "grad_norm": 0.3873872756958008, "learning_rate": 2.535068461075839e-05, "loss": 0.0244, "step": 68090 }, { "grad_norm": 0.21825310587882996, "learning_rate": 2.5336300158498516e-05, "loss": 0.0289, "step": 68100 }, { "grad_norm": 0.31681761145591736, "learning_rate": 2.5321918403423793e-05, "loss": 0.0185, "step": 68110 }, { "grad_norm": 0.6808179616928101, "learning_rate": 2.5307539347106957e-05, "loss": 0.0234, "step": 68120 }, { "grad_norm": 0.2833694815635681, "learning_rate": 2.529316299112049e-05, "loss": 0.0227, "step": 68130 }, { "grad_norm": 0.3594220280647278, "learning_rate": 2.5278789337036567e-05, "loss": 0.0253, "step": 68140 }, { "grad_norm": 0.6901317238807678, "learning_rate": 2.52644183864271e-05, "loss": 0.023, "step": 68150 }, { "grad_norm": 0.21776600182056427, "learning_rate": 2.525005014086363e-05, "loss": 0.022, "step": 68160 }, { "grad_norm": 0.4087029695510864, "learning_rate": 2.5235684601917465e-05, "loss": 0.0196, "step": 68170 }, { "grad_norm": 0.28806960582733154, "learning_rate": 2.5221321771159618e-05, "loss": 0.015, "step": 68180 }, { "grad_norm": 0.37866246700286865, "learning_rate": 2.5206961650160743e-05, "loss": 0.0191, "step": 68190 }, { "grad_norm": 0.3088172972202301, "learning_rate": 2.519260424049129e-05, "loss": 0.025, "step": 68200 }, { "grad_norm": 0.29789185523986816, "learning_rate": 2.5178249543721304e-05, "loss": 0.0176, "step": 68210 }, { "grad_norm": 0.19493158161640167, "learning_rate": 2.5163897561420668e-05, "loss": 0.0234, "step": 68220 }, { "grad_norm": 0.19049382209777832, "learning_rate": 2.514954829515883e-05, "loss": 0.0187, "step": 68230 }, { "grad_norm": 0.7796042561531067, "learning_rate": 2.513520174650505e-05, "loss": 0.0211, "step": 68240 }, { "grad_norm": 0.3193778395652771, "learning_rate": 2.5120857917028207e-05, "loss": 0.0195, "step": 68250 }, { "grad_norm": 0.2262725681066513, "learning_rate": 2.5106516808296953e-05, "loss": 0.0167, "step": 68260 }, { "grad_norm": 0.3347967267036438, "learning_rate": 2.5092178421879552e-05, "loss": 0.0198, "step": 68270 }, { "grad_norm": 0.4483069181442261, "learning_rate": 2.5077842759344116e-05, "loss": 0.037, "step": 68280 }, { "grad_norm": 0.30515342950820923, "learning_rate": 2.5063509822258304e-05, "loss": 0.0245, "step": 68290 }, { "grad_norm": 0.5824438333511353, "learning_rate": 2.5049179612189576e-05, "loss": 0.0293, "step": 68300 }, { "grad_norm": 0.26298362016677856, "learning_rate": 2.5034852130705043e-05, "loss": 0.0178, "step": 68310 }, { "grad_norm": 0.2336934357881546, "learning_rate": 2.5020527379371572e-05, "loss": 0.0209, "step": 68320 }, { "grad_norm": 0.3036539852619171, "learning_rate": 2.500620535975564e-05, "loss": 0.02, "step": 68330 }, { "grad_norm": 0.4879804253578186, "learning_rate": 2.4991886073423515e-05, "loss": 0.0176, "step": 68340 }, { "grad_norm": 0.24638617038726807, "learning_rate": 2.4977569521941134e-05, "loss": 0.0172, "step": 68350 }, { "grad_norm": 0.42042118310928345, "learning_rate": 2.4963255706874143e-05, "loss": 0.0242, "step": 68360 }, { "grad_norm": 0.28363075852394104, "learning_rate": 2.4948944629787853e-05, "loss": 0.0279, "step": 68370 }, { "grad_norm": 0.4060876965522766, "learning_rate": 2.493463629224731e-05, "loss": 0.0279, "step": 68380 }, { "grad_norm": 0.26017752289772034, "learning_rate": 2.492033069581728e-05, "loss": 0.02, "step": 68390 }, { "grad_norm": 0.37456637620925903, "learning_rate": 2.4906027842062137e-05, "loss": 0.0219, "step": 68400 }, { "grad_norm": 0.18726846575737, "learning_rate": 2.48917277325461e-05, "loss": 0.0207, "step": 68410 }, { "grad_norm": 0.4737005829811096, "learning_rate": 2.487743036883295e-05, "loss": 0.0307, "step": 68420 }, { "grad_norm": 2.396160125732422, "learning_rate": 2.486313575248626e-05, "loss": 0.0213, "step": 68430 }, { "grad_norm": 0.6358768939971924, "learning_rate": 2.4848843885069235e-05, "loss": 0.0328, "step": 68440 }, { "grad_norm": 0.2908739745616913, "learning_rate": 2.483455476814484e-05, "loss": 0.0208, "step": 68450 }, { "grad_norm": 0.3232840895652771, "learning_rate": 2.4820268403275672e-05, "loss": 0.0232, "step": 68460 }, { "grad_norm": 0.2687695324420929, "learning_rate": 2.480598479202413e-05, "loss": 0.0241, "step": 68470 }, { "grad_norm": 0.3960682153701782, "learning_rate": 2.4791703935952193e-05, "loss": 0.0181, "step": 68480 }, { "grad_norm": 0.5549555420875549, "learning_rate": 2.477742583662163e-05, "loss": 0.0271, "step": 68490 }, { "grad_norm": 0.28282907605171204, "learning_rate": 2.476315049559384e-05, "loss": 0.0271, "step": 68500 }, { "grad_norm": 0.31675514578819275, "learning_rate": 2.4748877914429987e-05, "loss": 0.0361, "step": 68510 }, { "grad_norm": 0.30074283480644226, "learning_rate": 2.4734608094690857e-05, "loss": 0.0278, "step": 68520 }, { "grad_norm": 0.3847809135913849, "learning_rate": 2.4720341037937006e-05, "loss": 0.0257, "step": 68530 }, { "grad_norm": 0.693817138671875, "learning_rate": 2.4706076745728645e-05, "loss": 0.0345, "step": 68540 }, { "grad_norm": 0.3921223282814026, "learning_rate": 2.469181521962573e-05, "loss": 0.0336, "step": 68550 }, { "grad_norm": 0.6900705695152283, "learning_rate": 2.467755646118783e-05, "loss": 0.0264, "step": 68560 }, { "grad_norm": 0.41837790608406067, "learning_rate": 2.46633004719743e-05, "loss": 0.0181, "step": 68570 }, { "grad_norm": 0.4354334771633148, "learning_rate": 2.4649047253544123e-05, "loss": 0.0257, "step": 68580 }, { "grad_norm": 0.3002150058746338, "learning_rate": 2.4634796807456016e-05, "loss": 0.0216, "step": 68590 }, { "grad_norm": 0.256908655166626, "learning_rate": 2.46205491352684e-05, "loss": 0.0216, "step": 68600 }, { "grad_norm": 0.49601176381111145, "learning_rate": 2.4606304238539375e-05, "loss": 0.0215, "step": 68610 }, { "grad_norm": 0.2307058572769165, "learning_rate": 2.4592062118826753e-05, "loss": 0.0163, "step": 68620 }, { "grad_norm": 0.2936265170574188, "learning_rate": 2.4577822777688004e-05, "loss": 0.0248, "step": 68630 }, { "grad_norm": 0.4690471589565277, "learning_rate": 2.4563586216680347e-05, "loss": 0.0206, "step": 68640 }, { "grad_norm": 0.1887657642364502, "learning_rate": 2.4549352437360622e-05, "loss": 0.0193, "step": 68650 }, { "grad_norm": 0.2339869886636734, "learning_rate": 2.4535121441285493e-05, "loss": 0.0227, "step": 68660 }, { "grad_norm": 0.2652286887168884, "learning_rate": 2.4520893230011172e-05, "loss": 0.0191, "step": 68670 }, { "grad_norm": 0.5612845420837402, "learning_rate": 2.450666780509368e-05, "loss": 0.0242, "step": 68680 }, { "grad_norm": 0.358117014169693, "learning_rate": 2.4492445168088656e-05, "loss": 0.0196, "step": 68690 }, { "grad_norm": 0.51423579454422, "learning_rate": 2.447822532055149e-05, "loss": 0.0235, "step": 68700 }, { "grad_norm": 0.5094242095947266, "learning_rate": 2.4464008264037212e-05, "loss": 0.0224, "step": 68710 }, { "grad_norm": 0.1331043541431427, "learning_rate": 2.4449794000100605e-05, "loss": 0.0187, "step": 68720 }, { "grad_norm": 0.5224749445915222, "learning_rate": 2.4435582530296108e-05, "loss": 0.0158, "step": 68730 }, { "grad_norm": 0.3269572854042053, "learning_rate": 2.4421373856177887e-05, "loss": 0.0245, "step": 68740 }, { "grad_norm": 0.5020504593849182, "learning_rate": 2.440716797929975e-05, "loss": 0.0173, "step": 68750 }, { "grad_norm": 0.2904170751571655, "learning_rate": 2.439296490121526e-05, "loss": 0.0129, "step": 68760 }, { "grad_norm": 0.2440018355846405, "learning_rate": 2.4378764623477608e-05, "loss": 0.0242, "step": 68770 }, { "grad_norm": 0.2936687767505646, "learning_rate": 2.436456714763974e-05, "loss": 0.0233, "step": 68780 }, { "grad_norm": 0.37827205657958984, "learning_rate": 2.4350372475254262e-05, "loss": 0.0214, "step": 68790 }, { "grad_norm": 0.40508270263671875, "learning_rate": 2.433618060787351e-05, "loss": 0.0226, "step": 68800 }, { "grad_norm": 0.4610068202018738, "learning_rate": 2.432199154704944e-05, "loss": 0.0321, "step": 68810 }, { "grad_norm": 0.5439140796661377, "learning_rate": 2.430780529433377e-05, "loss": 0.0231, "step": 68820 }, { "grad_norm": 0.1754978597164154, "learning_rate": 2.42936218512779e-05, "loss": 0.0288, "step": 68830 }, { "grad_norm": 0.4011913239955902, "learning_rate": 2.427944121943288e-05, "loss": 0.017, "step": 68840 }, { "grad_norm": 0.5743012428283691, "learning_rate": 2.4265263400349497e-05, "loss": 0.0176, "step": 68850 }, { "grad_norm": 0.39246559143066406, "learning_rate": 2.4251088395578213e-05, "loss": 0.0231, "step": 68860 }, { "grad_norm": 0.20559699833393097, "learning_rate": 2.4236916206669214e-05, "loss": 0.025, "step": 68870 }, { "grad_norm": 0.5405351519584656, "learning_rate": 2.4222746835172304e-05, "loss": 0.0213, "step": 68880 }, { "grad_norm": 0.4152268171310425, "learning_rate": 2.4208580282637065e-05, "loss": 0.0279, "step": 68890 }, { "grad_norm": 0.6103248596191406, "learning_rate": 2.4194416550612687e-05, "loss": 0.0236, "step": 68900 }, { "grad_norm": 0.5828317999839783, "learning_rate": 2.418025564064812e-05, "loss": 0.0276, "step": 68910 }, { "grad_norm": 0.749570906162262, "learning_rate": 2.4166097554291978e-05, "loss": 0.0248, "step": 68920 }, { "grad_norm": 0.2999769449234009, "learning_rate": 2.4151942293092583e-05, "loss": 0.0188, "step": 68930 }, { "grad_norm": 0.8593279123306274, "learning_rate": 2.41377898585979e-05, "loss": 0.0223, "step": 68940 }, { "grad_norm": 0.5151987671852112, "learning_rate": 2.4123640252355652e-05, "loss": 0.03, "step": 68950 }, { "grad_norm": 0.20385032892227173, "learning_rate": 2.410949347591318e-05, "loss": 0.0164, "step": 68960 }, { "grad_norm": 0.49977898597717285, "learning_rate": 2.4095349530817578e-05, "loss": 0.0347, "step": 68970 }, { "grad_norm": 0.23654824495315552, "learning_rate": 2.4081208418615603e-05, "loss": 0.0173, "step": 68980 }, { "grad_norm": 0.2443629801273346, "learning_rate": 2.4067070140853725e-05, "loss": 0.0211, "step": 68990 }, { "grad_norm": 0.46865054965019226, "learning_rate": 2.405293469907805e-05, "loss": 0.0208, "step": 69000 }, { "grad_norm": 0.7963263392448425, "learning_rate": 2.4038802094834446e-05, "loss": 0.0237, "step": 69010 }, { "grad_norm": 0.21877779066562653, "learning_rate": 2.4024672329668395e-05, "loss": 0.021, "step": 69020 }, { "grad_norm": 0.3301939070224762, "learning_rate": 2.4010545405125125e-05, "loss": 0.0164, "step": 69030 }, { "grad_norm": 0.6709162592887878, "learning_rate": 2.3996421322749528e-05, "loss": 0.0231, "step": 69040 }, { "grad_norm": 0.6187174916267395, "learning_rate": 2.3982300084086224e-05, "loss": 0.023, "step": 69050 }, { "grad_norm": 0.5577159523963928, "learning_rate": 2.396818169067944e-05, "loss": 0.0205, "step": 69060 }, { "grad_norm": 0.5517290234565735, "learning_rate": 2.3954066144073178e-05, "loss": 0.0196, "step": 69070 }, { "grad_norm": 1.0131999254226685, "learning_rate": 2.39399534458111e-05, "loss": 0.0216, "step": 69080 }, { "grad_norm": 0.2720736563205719, "learning_rate": 2.392584359743651e-05, "loss": 0.0173, "step": 69090 }, { "grad_norm": 0.4528597891330719, "learning_rate": 2.3911736600492463e-05, "loss": 0.0235, "step": 69100 }, { "grad_norm": 0.4294593334197998, "learning_rate": 2.3897632456521672e-05, "loss": 0.0265, "step": 69110 }, { "grad_norm": 0.23740634322166443, "learning_rate": 2.388353116706658e-05, "loss": 0.0258, "step": 69120 }, { "grad_norm": 0.20044629275798798, "learning_rate": 2.3869432733669232e-05, "loss": 0.0211, "step": 69130 }, { "grad_norm": 0.13185471296310425, "learning_rate": 2.385533715787145e-05, "loss": 0.0208, "step": 69140 }, { "grad_norm": 0.44454994797706604, "learning_rate": 2.3841244441214673e-05, "loss": 0.0205, "step": 69150 }, { "grad_norm": 0.5206015110015869, "learning_rate": 2.3827154585240073e-05, "loss": 0.0169, "step": 69160 }, { "grad_norm": 0.8773677945137024, "learning_rate": 2.3813067591488498e-05, "loss": 0.0262, "step": 69170 }, { "grad_norm": 0.3899335563182831, "learning_rate": 2.3798983461500494e-05, "loss": 0.0287, "step": 69180 }, { "grad_norm": 0.2955722212791443, "learning_rate": 2.3784902196816256e-05, "loss": 0.0201, "step": 69190 }, { "grad_norm": 0.46691590547561646, "learning_rate": 2.3770823798975716e-05, "loss": 0.0237, "step": 69200 }, { "grad_norm": 0.4112744629383087, "learning_rate": 2.3756748269518424e-05, "loss": 0.0197, "step": 69210 }, { "grad_norm": 0.3413292467594147, "learning_rate": 2.374267560998371e-05, "loss": 0.0314, "step": 69220 }, { "grad_norm": 0.42209765315055847, "learning_rate": 2.3728605821910477e-05, "loss": 0.024, "step": 69230 }, { "grad_norm": 0.5313243269920349, "learning_rate": 2.3714538906837453e-05, "loss": 0.0303, "step": 69240 }, { "grad_norm": 0.27510368824005127, "learning_rate": 2.3700474866302913e-05, "loss": 0.0194, "step": 69250 }, { "grad_norm": 0.5061184167861938, "learning_rate": 2.3686413701844917e-05, "loss": 0.0185, "step": 69260 }, { "grad_norm": 0.3450395464897156, "learning_rate": 2.367235541500114e-05, "loss": 0.0171, "step": 69270 }, { "grad_norm": 0.3211096525192261, "learning_rate": 2.3658300007308993e-05, "loss": 0.0194, "step": 69280 }, { "grad_norm": 0.30105721950531006, "learning_rate": 2.364424748030555e-05, "loss": 0.0238, "step": 69290 }, { "grad_norm": 0.2569696605205536, "learning_rate": 2.3630197835527595e-05, "loss": 0.0192, "step": 69300 }, { "grad_norm": 0.31229859590530396, "learning_rate": 2.3616151074511546e-05, "loss": 0.0209, "step": 69310 }, { "grad_norm": 0.2834807336330414, "learning_rate": 2.3602107198793545e-05, "loss": 0.0215, "step": 69320 }, { "grad_norm": 0.2000436782836914, "learning_rate": 2.3588066209909427e-05, "loss": 0.0217, "step": 69330 }, { "grad_norm": 0.5970259308815002, "learning_rate": 2.3574028109394662e-05, "loss": 0.0176, "step": 69340 }, { "grad_norm": 0.35506024956703186, "learning_rate": 2.3559992898784454e-05, "loss": 0.0181, "step": 69350 }, { "grad_norm": 0.10141677409410477, "learning_rate": 2.354596057961367e-05, "loss": 0.0163, "step": 69360 }, { "grad_norm": 0.3819429576396942, "learning_rate": 2.353193115341688e-05, "loss": 0.0256, "step": 69370 }, { "grad_norm": 0.5049926042556763, "learning_rate": 2.3517904621728294e-05, "loss": 0.0208, "step": 69380 }, { "grad_norm": 0.1716334968805313, "learning_rate": 2.3503880986081855e-05, "loss": 0.0202, "step": 69390 }, { "grad_norm": 0.47979551553726196, "learning_rate": 2.3489860248011136e-05, "loss": 0.0201, "step": 69400 }, { "grad_norm": 0.22866366803646088, "learning_rate": 2.3475842409049464e-05, "loss": 0.0203, "step": 69410 }, { "grad_norm": 0.3500681221485138, "learning_rate": 2.346182747072975e-05, "loss": 0.024, "step": 69420 }, { "grad_norm": 0.4675496220588684, "learning_rate": 2.3447815434584718e-05, "loss": 0.0238, "step": 69430 }, { "grad_norm": 0.6574836373329163, "learning_rate": 2.3433806302146656e-05, "loss": 0.0211, "step": 69440 }, { "grad_norm": 0.33692431449890137, "learning_rate": 2.3419800074947612e-05, "loss": 0.0163, "step": 69450 }, { "grad_norm": 0.2847548723220825, "learning_rate": 2.3405796754519248e-05, "loss": 0.0145, "step": 69460 }, { "grad_norm": 0.32546305656433105, "learning_rate": 2.3391796342392992e-05, "loss": 0.0181, "step": 69470 }, { "grad_norm": 0.4693816900253296, "learning_rate": 2.337779884009984e-05, "loss": 0.0266, "step": 69480 }, { "grad_norm": 0.5772176384925842, "learning_rate": 2.3363804249170624e-05, "loss": 0.0327, "step": 69490 }, { "grad_norm": 0.310120552778244, "learning_rate": 2.3349812571135703e-05, "loss": 0.018, "step": 69500 }, { "grad_norm": 0.3348675072193146, "learning_rate": 2.333582380752523e-05, "loss": 0.0262, "step": 69510 }, { "grad_norm": 0.16048377752304077, "learning_rate": 2.3321837959868963e-05, "loss": 0.0171, "step": 69520 }, { "grad_norm": 0.2604922652244568, "learning_rate": 2.3307855029696383e-05, "loss": 0.0235, "step": 69530 }, { "grad_norm": 0.2866579294204712, "learning_rate": 2.3293875018536648e-05, "loss": 0.0218, "step": 69540 }, { "grad_norm": 0.37043216824531555, "learning_rate": 2.327989792791861e-05, "loss": 0.0225, "step": 69550 }, { "grad_norm": 0.35497909784317017, "learning_rate": 2.326592375937074e-05, "loss": 0.0211, "step": 69560 }, { "grad_norm": 0.40817680954933167, "learning_rate": 2.3251952514421248e-05, "loss": 0.0234, "step": 69570 }, { "grad_norm": 0.641623854637146, "learning_rate": 2.323798419459804e-05, "loss": 0.0235, "step": 69580 }, { "grad_norm": 0.25586146116256714, "learning_rate": 2.3224018801428625e-05, "loss": 0.0174, "step": 69590 }, { "grad_norm": 0.5950377583503723, "learning_rate": 2.3210056336440273e-05, "loss": 0.0233, "step": 69600 }, { "grad_norm": 0.32238346338272095, "learning_rate": 2.3196096801159844e-05, "loss": 0.0163, "step": 69610 }, { "grad_norm": 0.3422437608242035, "learning_rate": 2.3182140197114012e-05, "loss": 0.024, "step": 69620 }, { "grad_norm": 0.30009719729423523, "learning_rate": 2.3168186525828985e-05, "loss": 0.0213, "step": 69630 }, { "grad_norm": 0.35470253229141235, "learning_rate": 2.3154235788830757e-05, "loss": 0.0198, "step": 69640 }, { "grad_norm": 0.3356187641620636, "learning_rate": 2.3140287987644927e-05, "loss": 0.0242, "step": 69650 }, { "grad_norm": 0.7153302431106567, "learning_rate": 2.3126343123796833e-05, "loss": 0.0276, "step": 69660 }, { "grad_norm": 0.2434575855731964, "learning_rate": 2.3112401198811418e-05, "loss": 0.0223, "step": 69670 }, { "grad_norm": 0.3687117099761963, "learning_rate": 2.3098462214213418e-05, "loss": 0.0239, "step": 69680 }, { "grad_norm": 0.5452014207839966, "learning_rate": 2.308452617152712e-05, "loss": 0.0164, "step": 69690 }, { "grad_norm": 0.2767243981361389, "learning_rate": 2.3070593072276603e-05, "loss": 0.0201, "step": 69700 }, { "grad_norm": 0.20648519694805145, "learning_rate": 2.3056662917985518e-05, "loss": 0.0155, "step": 69710 }, { "grad_norm": 0.2458326667547226, "learning_rate": 2.3042735710177283e-05, "loss": 0.0336, "step": 69720 }, { "grad_norm": 0.3363683819770813, "learning_rate": 2.302881145037492e-05, "loss": 0.0188, "step": 69730 }, { "grad_norm": 0.30561235547065735, "learning_rate": 2.3014890140101192e-05, "loss": 0.017, "step": 69740 }, { "grad_norm": 0.37850382924079895, "learning_rate": 2.30009717808785e-05, "loss": 0.0223, "step": 69750 }, { "grad_norm": 0.2573617994785309, "learning_rate": 2.2987056374228967e-05, "loss": 0.0254, "step": 69760 }, { "grad_norm": 0.3010709881782532, "learning_rate": 2.297314392167431e-05, "loss": 0.0153, "step": 69770 }, { "grad_norm": 0.41398388147354126, "learning_rate": 2.295923442473601e-05, "loss": 0.0262, "step": 69780 }, { "grad_norm": 0.4083712697029114, "learning_rate": 2.294532788493519e-05, "loss": 0.0212, "step": 69790 }, { "grad_norm": 0.24291816353797913, "learning_rate": 2.293142430379262e-05, "loss": 0.0157, "step": 69800 }, { "grad_norm": 0.5105418562889099, "learning_rate": 2.291752368282879e-05, "loss": 0.0186, "step": 69810 }, { "grad_norm": 0.2006160467863083, "learning_rate": 2.290362602356385e-05, "loss": 0.0182, "step": 69820 }, { "grad_norm": 0.35694071650505066, "learning_rate": 2.288973132751765e-05, "loss": 0.0164, "step": 69830 }, { "grad_norm": 0.26385340094566345, "learning_rate": 2.287583959620965e-05, "loss": 0.0298, "step": 69840 }, { "grad_norm": 0.5586401224136353, "learning_rate": 2.2861950831159074e-05, "loss": 0.0233, "step": 69850 }, { "grad_norm": 0.2036193460226059, "learning_rate": 2.284806503388472e-05, "loss": 0.0186, "step": 69860 }, { "grad_norm": 0.38642504811286926, "learning_rate": 2.283418220590518e-05, "loss": 0.022, "step": 69870 }, { "grad_norm": 0.4959389865398407, "learning_rate": 2.282030234873862e-05, "loss": 0.0235, "step": 69880 }, { "grad_norm": 0.26224255561828613, "learning_rate": 2.2806425463902948e-05, "loss": 0.0153, "step": 69890 }, { "grad_norm": 0.20474332571029663, "learning_rate": 2.279255155291568e-05, "loss": 0.0196, "step": 69900 }, { "grad_norm": 0.31882187724113464, "learning_rate": 2.277868061729409e-05, "loss": 0.0238, "step": 69910 }, { "grad_norm": 0.29078957438468933, "learning_rate": 2.276481265855504e-05, "loss": 0.026, "step": 69920 }, { "grad_norm": 0.3605581820011139, "learning_rate": 2.2750947678215128e-05, "loss": 0.0196, "step": 69930 }, { "grad_norm": 0.3159201741218567, "learning_rate": 2.2737085677790616e-05, "loss": 0.0167, "step": 69940 }, { "grad_norm": 0.40978360176086426, "learning_rate": 2.2723226658797443e-05, "loss": 0.0193, "step": 69950 }, { "grad_norm": 0.2833908796310425, "learning_rate": 2.270937062275117e-05, "loss": 0.0202, "step": 69960 }, { "grad_norm": 0.3169226348400116, "learning_rate": 2.2695517571167113e-05, "loss": 0.022, "step": 69970 }, { "grad_norm": 0.4346710741519928, "learning_rate": 2.268166750556019e-05, "loss": 0.0164, "step": 69980 }, { "grad_norm": 0.35336387157440186, "learning_rate": 2.266782042744503e-05, "loss": 0.0202, "step": 69990 }, { "grad_norm": 0.33038651943206787, "learning_rate": 2.2653976338335935e-05, "loss": 0.0215, "step": 70000 }, { "grad_norm": 0.7669485807418823, "learning_rate": 2.2640135239746895e-05, "loss": 0.023, "step": 70010 }, { "grad_norm": 0.21275737881660461, "learning_rate": 2.2626297133191515e-05, "loss": 0.0128, "step": 70020 }, { "grad_norm": 0.18211200833320618, "learning_rate": 2.261246202018312e-05, "loss": 0.02, "step": 70030 }, { "grad_norm": 0.2772148847579956, "learning_rate": 2.2598629902234718e-05, "loss": 0.0181, "step": 70040 }, { "grad_norm": 0.3113843500614166, "learning_rate": 2.258480078085894e-05, "loss": 0.0251, "step": 70050 }, { "grad_norm": 0.46955037117004395, "learning_rate": 2.2570974657568126e-05, "loss": 0.0192, "step": 70060 }, { "grad_norm": 0.42846328020095825, "learning_rate": 2.2557151533874286e-05, "loss": 0.0221, "step": 70070 }, { "grad_norm": 0.2193511724472046, "learning_rate": 2.2543331411289116e-05, "loss": 0.0168, "step": 70080 }, { "grad_norm": 0.25463244318962097, "learning_rate": 2.252951429132392e-05, "loss": 0.0231, "step": 70090 }, { "grad_norm": 0.44350114464759827, "learning_rate": 2.2515700175489762e-05, "loss": 0.0185, "step": 70100 }, { "grad_norm": 0.20161662995815277, "learning_rate": 2.2501889065297288e-05, "loss": 0.0238, "step": 70110 }, { "grad_norm": 0.2920110523700714, "learning_rate": 2.2488080962256887e-05, "loss": 0.0231, "step": 70120 }, { "grad_norm": 0.8908050656318665, "learning_rate": 2.2474275867878585e-05, "loss": 0.0267, "step": 70130 }, { "grad_norm": 0.5185974836349487, "learning_rate": 2.246047378367211e-05, "loss": 0.0298, "step": 70140 }, { "grad_norm": 0.26844918727874756, "learning_rate": 2.2446674711146798e-05, "loss": 0.0199, "step": 70150 }, { "grad_norm": 0.22414016723632812, "learning_rate": 2.2432878651811734e-05, "loss": 0.0178, "step": 70160 }, { "grad_norm": 0.2671329975128174, "learning_rate": 2.2419085607175594e-05, "loss": 0.0259, "step": 70170 }, { "grad_norm": 0.11942117661237717, "learning_rate": 2.240529557874678e-05, "loss": 0.0242, "step": 70180 }, { "grad_norm": 0.332787424325943, "learning_rate": 2.2391508568033358e-05, "loss": 0.0193, "step": 70190 }, { "grad_norm": 0.3801867961883545, "learning_rate": 2.237772457654307e-05, "loss": 0.0178, "step": 70200 }, { "grad_norm": 0.44305989146232605, "learning_rate": 2.236394360578327e-05, "loss": 0.0205, "step": 70210 }, { "grad_norm": 0.38268136978149414, "learning_rate": 2.2350165657261073e-05, "loss": 0.0168, "step": 70220 }, { "grad_norm": 0.30178073048591614, "learning_rate": 2.2336390732483164e-05, "loss": 0.0204, "step": 70230 }, { "grad_norm": 0.34925809502601624, "learning_rate": 2.2322618832955976e-05, "loss": 0.0234, "step": 70240 }, { "grad_norm": 0.2891852557659149, "learning_rate": 2.2308849960185578e-05, "loss": 0.0244, "step": 70250 }, { "grad_norm": 0.2990212142467499, "learning_rate": 2.229508411567774e-05, "loss": 0.0256, "step": 70260 }, { "grad_norm": 0.21804961562156677, "learning_rate": 2.2281321300937823e-05, "loss": 0.026, "step": 70270 }, { "grad_norm": 0.1809977889060974, "learning_rate": 2.2267561517470937e-05, "loss": 0.0188, "step": 70280 }, { "grad_norm": 0.29055482149124146, "learning_rate": 2.2253804766781845e-05, "loss": 0.0216, "step": 70290 }, { "grad_norm": 0.5349477529525757, "learning_rate": 2.2240051050374933e-05, "loss": 0.022, "step": 70300 }, { "grad_norm": 0.4225695729255676, "learning_rate": 2.22263003697543e-05, "loss": 0.0231, "step": 70310 }, { "grad_norm": 0.30684754252433777, "learning_rate": 2.2212552726423702e-05, "loss": 0.0231, "step": 70320 }, { "grad_norm": 0.3329806923866272, "learning_rate": 2.2198808121886578e-05, "loss": 0.0277, "step": 70330 }, { "grad_norm": 0.28870195150375366, "learning_rate": 2.218506655764598e-05, "loss": 0.0176, "step": 70340 }, { "grad_norm": 0.2292158007621765, "learning_rate": 2.2171328035204704e-05, "loss": 0.0223, "step": 70350 }, { "grad_norm": 0.2949475049972534, "learning_rate": 2.2157592556065133e-05, "loss": 0.0208, "step": 70360 }, { "grad_norm": 0.1608022302389145, "learning_rate": 2.2143860121729382e-05, "loss": 0.0228, "step": 70370 }, { "grad_norm": 0.2398661971092224, "learning_rate": 2.2130130733699206e-05, "loss": 0.0184, "step": 70380 }, { "grad_norm": 0.27922171354293823, "learning_rate": 2.2116404393476052e-05, "loss": 0.022, "step": 70390 }, { "grad_norm": 0.6405242085456848, "learning_rate": 2.2102681102560967e-05, "loss": 0.0237, "step": 70400 }, { "grad_norm": 0.10675196349620819, "learning_rate": 2.208896086245476e-05, "loss": 0.0123, "step": 70410 }, { "grad_norm": 0.6124237179756165, "learning_rate": 2.2075243674657808e-05, "loss": 0.0172, "step": 70420 }, { "grad_norm": 0.21229587495326996, "learning_rate": 2.206152954067024e-05, "loss": 0.0125, "step": 70430 }, { "grad_norm": 0.46959149837493896, "learning_rate": 2.2047818461991758e-05, "loss": 0.0342, "step": 70440 }, { "grad_norm": 0.4392983019351959, "learning_rate": 2.2034110440121864e-05, "loss": 0.0227, "step": 70450 }, { "grad_norm": 0.27287134528160095, "learning_rate": 2.2020405476559585e-05, "loss": 0.0195, "step": 70460 }, { "grad_norm": 0.5923529267311096, "learning_rate": 2.2006703572803722e-05, "loss": 0.0179, "step": 70470 }, { "grad_norm": 0.34213221073150635, "learning_rate": 2.1993004730352646e-05, "loss": 0.021, "step": 70480 }, { "grad_norm": 0.37776729464530945, "learning_rate": 2.1979308950704464e-05, "loss": 0.018, "step": 70490 }, { "grad_norm": 0.40203505754470825, "learning_rate": 2.1965616235356923e-05, "loss": 0.0215, "step": 70500 }, { "grad_norm": 0.5429772138595581, "learning_rate": 2.1951926585807465e-05, "loss": 0.0177, "step": 70510 }, { "grad_norm": 0.22037933766841888, "learning_rate": 2.193824000355313e-05, "loss": 0.0174, "step": 70520 }, { "grad_norm": 0.35902971029281616, "learning_rate": 2.192455649009067e-05, "loss": 0.0186, "step": 70530 }, { "grad_norm": 0.45596277713775635, "learning_rate": 2.1910876046916523e-05, "loss": 0.0239, "step": 70540 }, { "grad_norm": 0.224399596452713, "learning_rate": 2.1897198675526725e-05, "loss": 0.0213, "step": 70550 }, { "grad_norm": 0.2467823177576065, "learning_rate": 2.1883524377417024e-05, "loss": 0.0232, "step": 70560 }, { "grad_norm": 0.3746330440044403, "learning_rate": 2.1869853154082825e-05, "loss": 0.0138, "step": 70570 }, { "grad_norm": 0.4171083867549896, "learning_rate": 2.1856185007019215e-05, "loss": 0.0201, "step": 70580 }, { "grad_norm": 0.4260288178920746, "learning_rate": 2.1842519937720874e-05, "loss": 0.0199, "step": 70590 }, { "grad_norm": 0.44826826453208923, "learning_rate": 2.182885794768224e-05, "loss": 0.0203, "step": 70600 }, { "grad_norm": 0.1947047859430313, "learning_rate": 2.1815199038397326e-05, "loss": 0.0133, "step": 70610 }, { "grad_norm": 0.5989816784858704, "learning_rate": 2.1801543211359893e-05, "loss": 0.0232, "step": 70620 }, { "grad_norm": 0.3105093836784363, "learning_rate": 2.178789046806326e-05, "loss": 0.0185, "step": 70630 }, { "grad_norm": 0.44295114278793335, "learning_rate": 2.1774240810000547e-05, "loss": 0.0285, "step": 70640 }, { "grad_norm": 0.3510652780532837, "learning_rate": 2.17605942386644e-05, "loss": 0.0204, "step": 70650 }, { "grad_norm": 0.45253100991249084, "learning_rate": 2.174695075554723e-05, "loss": 0.0328, "step": 70660 }, { "grad_norm": 0.44739025831222534, "learning_rate": 2.173331036214103e-05, "loss": 0.0308, "step": 70670 }, { "grad_norm": 0.4465635120868683, "learning_rate": 2.171967305993753e-05, "loss": 0.0205, "step": 70680 }, { "grad_norm": 0.21528153121471405, "learning_rate": 2.1706038850428024e-05, "loss": 0.0161, "step": 70690 }, { "grad_norm": 0.8533743619918823, "learning_rate": 2.169240773510361e-05, "loss": 0.0209, "step": 70700 }, { "grad_norm": 0.36024653911590576, "learning_rate": 2.1678779715454906e-05, "loss": 0.016, "step": 70710 }, { "grad_norm": 1.4609967470169067, "learning_rate": 2.166515479297229e-05, "loss": 0.0231, "step": 70720 }, { "grad_norm": 0.4960646331310272, "learning_rate": 2.1651532969145732e-05, "loss": 0.0128, "step": 70730 }, { "grad_norm": 0.26675042510032654, "learning_rate": 2.1637914245464903e-05, "loss": 0.02, "step": 70740 }, { "grad_norm": 0.5942783355712891, "learning_rate": 2.1624298623419133e-05, "loss": 0.025, "step": 70750 }, { "grad_norm": 0.604761004447937, "learning_rate": 2.1610686104497413e-05, "loss": 0.0216, "step": 70760 }, { "grad_norm": 0.6276088953018188, "learning_rate": 2.1597076690188366e-05, "loss": 0.0198, "step": 70770 }, { "grad_norm": 0.4225301742553711, "learning_rate": 2.1583470381980315e-05, "loss": 0.0199, "step": 70780 }, { "grad_norm": 0.2900857627391815, "learning_rate": 2.156986718136123e-05, "loss": 0.0304, "step": 70790 }, { "grad_norm": 0.35116440057754517, "learning_rate": 2.155626708981871e-05, "loss": 0.0214, "step": 70800 }, { "grad_norm": 0.49903422594070435, "learning_rate": 2.1542670108840075e-05, "loss": 0.0235, "step": 70810 }, { "grad_norm": 0.4152320623397827, "learning_rate": 2.152907623991221e-05, "loss": 0.029, "step": 70820 }, { "grad_norm": 0.39205074310302734, "learning_rate": 2.1515485484521802e-05, "loss": 0.0181, "step": 70830 }, { "grad_norm": 0.3077975809574127, "learning_rate": 2.150189784415506e-05, "loss": 0.0198, "step": 70840 }, { "grad_norm": 0.45139992237091064, "learning_rate": 2.148831332029794e-05, "loss": 0.0137, "step": 70850 }, { "grad_norm": 0.39779263734817505, "learning_rate": 2.147473191443599e-05, "loss": 0.0206, "step": 70860 }, { "grad_norm": 0.27608412504196167, "learning_rate": 2.1461153628054492e-05, "loss": 0.0196, "step": 70870 }, { "grad_norm": 0.22379547357559204, "learning_rate": 2.1447578462638284e-05, "loss": 0.0164, "step": 70880 }, { "grad_norm": 0.34732499718666077, "learning_rate": 2.143400641967201e-05, "loss": 0.0182, "step": 70890 }, { "grad_norm": 0.25899145007133484, "learning_rate": 2.1420437500639823e-05, "loss": 0.0183, "step": 70900 }, { "grad_norm": 0.3296360373497009, "learning_rate": 2.1406871707025646e-05, "loss": 0.0238, "step": 70910 }, { "grad_norm": 0.2188531458377838, "learning_rate": 2.1393309040312963e-05, "loss": 0.0262, "step": 70920 }, { "grad_norm": 0.29769375920295715, "learning_rate": 2.137974950198502e-05, "loss": 0.0191, "step": 70930 }, { "grad_norm": 0.3528154790401459, "learning_rate": 2.1366193093524617e-05, "loss": 0.0143, "step": 70940 }, { "grad_norm": 0.4079482853412628, "learning_rate": 2.1352639816414288e-05, "loss": 0.016, "step": 70950 }, { "grad_norm": 0.329899400472641, "learning_rate": 2.1339089672136202e-05, "loss": 0.0204, "step": 70960 }, { "grad_norm": 0.2682722210884094, "learning_rate": 2.1325542662172193e-05, "loss": 0.0177, "step": 70970 }, { "grad_norm": 0.2532202899456024, "learning_rate": 2.1311998788003712e-05, "loss": 0.0185, "step": 70980 }, { "grad_norm": 0.6472308039665222, "learning_rate": 2.1298458051111907e-05, "loss": 0.0246, "step": 70990 }, { "grad_norm": 0.37371841073036194, "learning_rate": 2.1284920452977602e-05, "loss": 0.0191, "step": 71000 }, { "grad_norm": 0.280373215675354, "learning_rate": 2.12713859950812e-05, "loss": 0.0173, "step": 71010 }, { "grad_norm": 0.33696046471595764, "learning_rate": 2.1257854678902832e-05, "loss": 0.0194, "step": 71020 }, { "grad_norm": 0.26741287112236023, "learning_rate": 2.124432650592227e-05, "loss": 0.025, "step": 71030 }, { "grad_norm": 0.35890263319015503, "learning_rate": 2.1230801477618937e-05, "loss": 0.029, "step": 71040 }, { "grad_norm": 0.43623462319374084, "learning_rate": 2.121727959547189e-05, "loss": 0.0274, "step": 71050 }, { "grad_norm": 0.20521390438079834, "learning_rate": 2.1203760860959886e-05, "loss": 0.0268, "step": 71060 }, { "grad_norm": 0.2592829763889313, "learning_rate": 2.119024527556127e-05, "loss": 0.0244, "step": 71070 }, { "grad_norm": 0.43477827310562134, "learning_rate": 2.117673284075415e-05, "loss": 0.0174, "step": 71080 }, { "grad_norm": 0.8536200523376465, "learning_rate": 2.1163223558016165e-05, "loss": 0.0208, "step": 71090 }, { "grad_norm": 0.5481218099594116, "learning_rate": 2.1149717428824722e-05, "loss": 0.0213, "step": 71100 }, { "grad_norm": 0.35822153091430664, "learning_rate": 2.1136214454656784e-05, "loss": 0.0248, "step": 71110 }, { "grad_norm": 0.24054643511772156, "learning_rate": 2.1122714636989056e-05, "loss": 0.0239, "step": 71120 }, { "grad_norm": 0.23916593194007874, "learning_rate": 2.1109217977297823e-05, "loss": 0.0263, "step": 71130 }, { "grad_norm": 0.14736062288284302, "learning_rate": 2.1095724477059077e-05, "loss": 0.0208, "step": 71140 }, { "grad_norm": 0.49969586730003357, "learning_rate": 2.1082234137748434e-05, "loss": 0.0285, "step": 71150 }, { "grad_norm": 0.5530140995979309, "learning_rate": 2.106874696084122e-05, "loss": 0.0165, "step": 71160 }, { "grad_norm": 0.35267895460128784, "learning_rate": 2.1055262947812315e-05, "loss": 0.0218, "step": 71170 }, { "grad_norm": 0.2687842547893524, "learning_rate": 2.104178210013636e-05, "loss": 0.0226, "step": 71180 }, { "grad_norm": 0.6230942010879517, "learning_rate": 2.1028304419287554e-05, "loss": 0.0258, "step": 71190 }, { "grad_norm": 0.3934032917022705, "learning_rate": 2.101482990673982e-05, "loss": 0.0217, "step": 71200 }, { "grad_norm": 0.4623645842075348, "learning_rate": 2.1001358563966707e-05, "loss": 0.0222, "step": 71210 }, { "grad_norm": 0.23270201683044434, "learning_rate": 2.098789039244144e-05, "loss": 0.0192, "step": 71220 }, { "grad_norm": 0.397885799407959, "learning_rate": 2.097442539363685e-05, "loss": 0.0275, "step": 71230 }, { "grad_norm": 0.37808266282081604, "learning_rate": 2.0960963569025448e-05, "loss": 0.0182, "step": 71240 }, { "grad_norm": 0.2626548707485199, "learning_rate": 2.0947504920079435e-05, "loss": 0.0169, "step": 71250 }, { "grad_norm": 0.11314579099416733, "learning_rate": 2.093404944827059e-05, "loss": 0.0182, "step": 71260 }, { "grad_norm": 0.21494613587856293, "learning_rate": 2.0920597155070394e-05, "loss": 0.015, "step": 71270 }, { "grad_norm": 0.254638135433197, "learning_rate": 2.090714804194997e-05, "loss": 0.0233, "step": 71280 }, { "grad_norm": 0.33231183886528015, "learning_rate": 2.0893702110380114e-05, "loss": 0.0176, "step": 71290 }, { "grad_norm": 0.5576093196868896, "learning_rate": 2.0880259361831216e-05, "loss": 0.0205, "step": 71300 }, { "grad_norm": 0.23664923012256622, "learning_rate": 2.0866819797773397e-05, "loss": 0.0184, "step": 71310 }, { "grad_norm": 0.4061732590198517, "learning_rate": 2.085338341967634e-05, "loss": 0.0221, "step": 71320 }, { "grad_norm": 0.5102618932723999, "learning_rate": 2.083995022900946e-05, "loss": 0.022, "step": 71330 }, { "grad_norm": 0.24961890280246735, "learning_rate": 2.0826520227241768e-05, "loss": 0.0237, "step": 71340 }, { "grad_norm": 0.2812318205833435, "learning_rate": 2.081309341584199e-05, "loss": 0.025, "step": 71350 }, { "grad_norm": 0.4274085462093353, "learning_rate": 2.0799669796278413e-05, "loss": 0.0248, "step": 71360 }, { "grad_norm": 0.2185598462820053, "learning_rate": 2.078624937001906e-05, "loss": 0.0204, "step": 71370 }, { "grad_norm": 0.538651704788208, "learning_rate": 2.0772832138531544e-05, "loss": 0.0433, "step": 71380 }, { "grad_norm": 0.36186152696609497, "learning_rate": 2.0759418103283158e-05, "loss": 0.0198, "step": 71390 }, { "grad_norm": 0.44551756978034973, "learning_rate": 2.0746007265740846e-05, "loss": 0.0251, "step": 71400 }, { "grad_norm": 0.29921603202819824, "learning_rate": 2.0732599627371215e-05, "loss": 0.0194, "step": 71410 }, { "grad_norm": 0.3916850686073303, "learning_rate": 2.0719195189640468e-05, "loss": 0.0263, "step": 71420 }, { "grad_norm": 0.6907780170440674, "learning_rate": 2.0705793954014528e-05, "loss": 0.0238, "step": 71430 }, { "grad_norm": 0.38406774401664734, "learning_rate": 2.06923959219589e-05, "loss": 0.019, "step": 71440 }, { "grad_norm": 0.5390868186950684, "learning_rate": 2.0679001094938793e-05, "loss": 0.0191, "step": 71450 }, { "grad_norm": 0.44982510805130005, "learning_rate": 2.066560947441904e-05, "loss": 0.0231, "step": 71460 }, { "grad_norm": 0.5363913178443909, "learning_rate": 2.0652221061864152e-05, "loss": 0.0157, "step": 71470 }, { "grad_norm": 0.9787425398826599, "learning_rate": 2.0638835858738226e-05, "loss": 0.0166, "step": 71480 }, { "grad_norm": 0.7732570767402649, "learning_rate": 2.0625453866505068e-05, "loss": 0.0199, "step": 71490 }, { "grad_norm": 0.1645246148109436, "learning_rate": 2.0612075086628123e-05, "loss": 0.0169, "step": 71500 }, { "grad_norm": 0.3525882363319397, "learning_rate": 2.059869952057045e-05, "loss": 0.022, "step": 71510 }, { "grad_norm": 0.727439284324646, "learning_rate": 2.0585327169794793e-05, "loss": 0.0253, "step": 71520 }, { "grad_norm": 0.3307195007801056, "learning_rate": 2.0571958035763534e-05, "loss": 0.0178, "step": 71530 }, { "grad_norm": 0.4613391160964966, "learning_rate": 2.0558592119938715e-05, "loss": 0.0199, "step": 71540 }, { "grad_norm": 0.4363623261451721, "learning_rate": 2.054522942378198e-05, "loss": 0.0242, "step": 71550 }, { "grad_norm": 0.23246023058891296, "learning_rate": 2.053186994875469e-05, "loss": 0.0197, "step": 71560 }, { "grad_norm": 0.4852232336997986, "learning_rate": 2.0518513696317775e-05, "loss": 0.0218, "step": 71570 }, { "grad_norm": 0.48416054248809814, "learning_rate": 2.050516066793188e-05, "loss": 0.0249, "step": 71580 }, { "grad_norm": 0.31153637170791626, "learning_rate": 2.0491810865057276e-05, "loss": 0.0167, "step": 71590 }, { "grad_norm": 0.2683207392692566, "learning_rate": 2.0478464289153882e-05, "loss": 0.016, "step": 71600 }, { "grad_norm": 0.295257568359375, "learning_rate": 2.0465120941681232e-05, "loss": 0.0208, "step": 71610 }, { "grad_norm": 0.28373390436172485, "learning_rate": 2.0451780824098566e-05, "loss": 0.0245, "step": 71620 }, { "grad_norm": 0.4061299264431, "learning_rate": 2.043844393786471e-05, "loss": 0.0219, "step": 71630 }, { "grad_norm": 0.5338003039360046, "learning_rate": 2.0425110284438193e-05, "loss": 0.015, "step": 71640 }, { "grad_norm": 0.23942823708057404, "learning_rate": 2.041177986527712e-05, "loss": 0.0223, "step": 71650 }, { "grad_norm": 0.49344339966773987, "learning_rate": 2.0398452681839347e-05, "loss": 0.0185, "step": 71660 }, { "grad_norm": 0.41578924655914307, "learning_rate": 2.0385128735582272e-05, "loss": 0.0382, "step": 71670 }, { "grad_norm": 0.4269852638244629, "learning_rate": 2.0371808027963003e-05, "loss": 0.027, "step": 71680 }, { "grad_norm": 0.6926394104957581, "learning_rate": 2.035849056043825e-05, "loss": 0.0209, "step": 71690 }, { "grad_norm": 0.46215420961380005, "learning_rate": 2.03451763344644e-05, "loss": 0.0241, "step": 71700 }, { "grad_norm": 0.17673909664154053, "learning_rate": 2.0331865351497482e-05, "loss": 0.017, "step": 71710 }, { "grad_norm": 0.3607116937637329, "learning_rate": 2.0318557612993185e-05, "loss": 0.0195, "step": 71720 }, { "grad_norm": 0.3284133970737457, "learning_rate": 2.0305253120406782e-05, "loss": 0.0253, "step": 71730 }, { "grad_norm": 0.1637333631515503, "learning_rate": 2.029195187519325e-05, "loss": 0.0179, "step": 71740 }, { "grad_norm": 0.66309654712677, "learning_rate": 2.0278653878807217e-05, "loss": 0.0233, "step": 71750 }, { "grad_norm": 0.36760273575782776, "learning_rate": 2.0265359132702892e-05, "loss": 0.0269, "step": 71760 }, { "grad_norm": 0.3585987687110901, "learning_rate": 2.0252067638334182e-05, "loss": 0.023, "step": 71770 }, { "grad_norm": 0.6469162702560425, "learning_rate": 2.023877939715464e-05, "loss": 0.0224, "step": 71780 }, { "grad_norm": 0.41468265652656555, "learning_rate": 2.0225494410617448e-05, "loss": 0.0198, "step": 71790 }, { "grad_norm": 0.501803994178772, "learning_rate": 2.02122126801754e-05, "loss": 0.0186, "step": 71800 }, { "grad_norm": 0.3061971068382263, "learning_rate": 2.0198934207281018e-05, "loss": 0.0247, "step": 71810 }, { "grad_norm": 0.495207279920578, "learning_rate": 2.0185658993386357e-05, "loss": 0.0269, "step": 71820 }, { "grad_norm": 0.26456719636917114, "learning_rate": 2.0172387039943226e-05, "loss": 0.0175, "step": 71830 }, { "grad_norm": 1.1535471677780151, "learning_rate": 2.0159118348402968e-05, "loss": 0.0242, "step": 71840 }, { "grad_norm": 0.34502720832824707, "learning_rate": 2.0145852920216697e-05, "loss": 0.0228, "step": 71850 }, { "grad_norm": 0.47307494282722473, "learning_rate": 2.0132590756835045e-05, "loss": 0.0247, "step": 71860 }, { "grad_norm": 0.416145384311676, "learning_rate": 2.0119331859708386e-05, "loss": 0.038, "step": 71870 }, { "grad_norm": 0.39570268988609314, "learning_rate": 2.0106076230286654e-05, "loss": 0.0255, "step": 71880 }, { "grad_norm": 0.2466677874326706, "learning_rate": 2.0092823870019488e-05, "loss": 0.0179, "step": 71890 }, { "grad_norm": 0.4559934437274933, "learning_rate": 2.0079574780356115e-05, "loss": 0.0222, "step": 71900 }, { "grad_norm": 0.30236491560935974, "learning_rate": 2.0066328962745486e-05, "loss": 0.0233, "step": 71910 }, { "grad_norm": 0.28481900691986084, "learning_rate": 2.00530864186361e-05, "loss": 0.017, "step": 71920 }, { "grad_norm": 0.36396467685699463, "learning_rate": 2.003984714947618e-05, "loss": 0.0216, "step": 71930 }, { "grad_norm": 0.15268689393997192, "learning_rate": 2.0026611156713515e-05, "loss": 0.0237, "step": 71940 }, { "grad_norm": 0.4437802731990814, "learning_rate": 2.0013378441795584e-05, "loss": 0.0169, "step": 71950 }, { "grad_norm": 0.5195471048355103, "learning_rate": 2.0000149006169506e-05, "loss": 0.0237, "step": 71960 }, { "grad_norm": 0.30484485626220703, "learning_rate": 1.9986922851282048e-05, "loss": 0.0221, "step": 71970 }, { "grad_norm": 1.124250888824463, "learning_rate": 1.997369997857956e-05, "loss": 0.0183, "step": 71980 }, { "grad_norm": 0.27965047955513, "learning_rate": 1.9960480389508103e-05, "loss": 0.0191, "step": 71990 }, { "grad_norm": 0.17727990448474884, "learning_rate": 1.9947264085513363e-05, "loss": 0.0265, "step": 72000 }, { "grad_norm": 0.2133469581604004, "learning_rate": 1.993405106804062e-05, "loss": 0.0169, "step": 72010 }, { "grad_norm": 0.24843263626098633, "learning_rate": 1.9920841338534867e-05, "loss": 0.0257, "step": 72020 }, { "grad_norm": 0.41261330246925354, "learning_rate": 1.9907634898440646e-05, "loss": 0.0281, "step": 72030 }, { "grad_norm": 0.29030343890190125, "learning_rate": 1.989443174920227e-05, "loss": 0.0211, "step": 72040 }, { "grad_norm": 1.0736854076385498, "learning_rate": 1.988123189226355e-05, "loss": 0.0302, "step": 72050 }, { "grad_norm": 0.252208411693573, "learning_rate": 1.986803532906805e-05, "loss": 0.025, "step": 72060 }, { "grad_norm": 0.23177969455718994, "learning_rate": 1.985484206105889e-05, "loss": 0.016, "step": 72070 }, { "grad_norm": 0.41150596737861633, "learning_rate": 1.9841652089678887e-05, "loss": 0.0263, "step": 72080 }, { "grad_norm": 0.16538839042186737, "learning_rate": 1.9828465416370435e-05, "loss": 0.0214, "step": 72090 }, { "grad_norm": 0.5110529065132141, "learning_rate": 1.9815282042575685e-05, "loss": 0.022, "step": 72100 }, { "grad_norm": 0.2602306306362152, "learning_rate": 1.9802101969736285e-05, "loss": 0.0168, "step": 72110 }, { "grad_norm": 0.29464346170425415, "learning_rate": 1.9788925199293633e-05, "loss": 0.0283, "step": 72120 }, { "grad_norm": 0.24133744835853577, "learning_rate": 1.977575173268868e-05, "loss": 0.0172, "step": 72130 }, { "grad_norm": 0.3604176938533783, "learning_rate": 1.9762581571362094e-05, "loss": 0.0185, "step": 72140 }, { "grad_norm": 0.6978628635406494, "learning_rate": 1.9749414716754112e-05, "loss": 0.0275, "step": 72150 }, { "grad_norm": 0.2612147331237793, "learning_rate": 1.973625117030465e-05, "loss": 0.0266, "step": 72160 }, { "grad_norm": 0.34684479236602783, "learning_rate": 1.972309093345326e-05, "loss": 0.0216, "step": 72170 }, { "grad_norm": 0.6493757963180542, "learning_rate": 1.970993400763915e-05, "loss": 0.0231, "step": 72180 }, { "grad_norm": 0.21415236592292786, "learning_rate": 1.969678039430109e-05, "loss": 0.0237, "step": 72190 }, { "grad_norm": 0.34544917941093445, "learning_rate": 1.9683630094877564e-05, "loss": 0.0194, "step": 72200 }, { "grad_norm": 0.20897983014583588, "learning_rate": 1.9670483110806696e-05, "loss": 0.0181, "step": 72210 }, { "grad_norm": 0.45331430435180664, "learning_rate": 1.9657339443526172e-05, "loss": 0.0262, "step": 72220 }, { "grad_norm": 0.2615950405597687, "learning_rate": 1.964419909447339e-05, "loss": 0.017, "step": 72230 }, { "grad_norm": 0.26109054684638977, "learning_rate": 1.963106206508536e-05, "loss": 0.0163, "step": 72240 }, { "grad_norm": 0.19093915820121765, "learning_rate": 1.9617928356798738e-05, "loss": 0.0155, "step": 72250 }, { "grad_norm": 0.18511231243610382, "learning_rate": 1.960479797104977e-05, "loss": 0.0188, "step": 72260 }, { "grad_norm": 0.28942570090293884, "learning_rate": 1.9591670909274418e-05, "loss": 0.0137, "step": 72270 }, { "grad_norm": 0.22256824374198914, "learning_rate": 1.9578547172908184e-05, "loss": 0.0204, "step": 72280 }, { "grad_norm": 0.18745911121368408, "learning_rate": 1.9565426763386325e-05, "loss": 0.0143, "step": 72290 }, { "grad_norm": 0.263032466173172, "learning_rate": 1.9552309682143622e-05, "loss": 0.0201, "step": 72300 }, { "grad_norm": 0.5327269434928894, "learning_rate": 1.9539195930614574e-05, "loss": 0.0202, "step": 72310 }, { "grad_norm": 0.23504355549812317, "learning_rate": 1.952608551023324e-05, "loss": 0.0243, "step": 72320 }, { "grad_norm": 0.1621466875076294, "learning_rate": 1.9512978422433397e-05, "loss": 0.0218, "step": 72330 }, { "grad_norm": 0.2165602594614029, "learning_rate": 1.9499874668648376e-05, "loss": 0.0167, "step": 72340 }, { "grad_norm": 3.8573765754699707, "learning_rate": 1.9486774250311213e-05, "loss": 0.0271, "step": 72350 }, { "grad_norm": 0.37072622776031494, "learning_rate": 1.9473677168854537e-05, "loss": 0.0194, "step": 72360 }, { "grad_norm": 7.642291069030762, "learning_rate": 1.946058342571065e-05, "loss": 0.027, "step": 72370 }, { "grad_norm": 0.2799839973449707, "learning_rate": 1.9447493022311424e-05, "loss": 0.022, "step": 72380 }, { "grad_norm": 0.31406137347221375, "learning_rate": 1.9434405960088442e-05, "loss": 0.0183, "step": 72390 }, { "grad_norm": 0.32650214433670044, "learning_rate": 1.9421322240472857e-05, "loss": 0.0201, "step": 72400 }, { "grad_norm": 1.6472845077514648, "learning_rate": 1.9408241864895494e-05, "loss": 0.0204, "step": 72410 }, { "grad_norm": 0.2094140201807022, "learning_rate": 1.939516483478681e-05, "loss": 0.0139, "step": 72420 }, { "grad_norm": 0.3247164785861969, "learning_rate": 1.9382091151576902e-05, "loss": 0.0165, "step": 72430 }, { "grad_norm": 0.25492170453071594, "learning_rate": 1.9369020816695454e-05, "loss": 0.0236, "step": 72440 }, { "grad_norm": 0.7065551280975342, "learning_rate": 1.935595383157184e-05, "loss": 0.0183, "step": 72450 }, { "grad_norm": 0.3077194392681122, "learning_rate": 1.9342890197635057e-05, "loss": 0.0168, "step": 72460 }, { "grad_norm": 0.2675485610961914, "learning_rate": 1.9329829916313685e-05, "loss": 0.0151, "step": 72470 }, { "grad_norm": 0.45552828907966614, "learning_rate": 1.9316772989036013e-05, "loss": 0.0243, "step": 72480 }, { "grad_norm": 0.2780449390411377, "learning_rate": 1.930371941722991e-05, "loss": 0.0141, "step": 72490 }, { "grad_norm": 0.178953155875206, "learning_rate": 1.929066920232292e-05, "loss": 0.019, "step": 72500 }, { "grad_norm": 0.1971750557422638, "learning_rate": 1.9277622345742152e-05, "loss": 0.0209, "step": 72510 }, { "grad_norm": 0.29484301805496216, "learning_rate": 1.926457884891444e-05, "loss": 0.0172, "step": 72520 }, { "grad_norm": 0.5316813588142395, "learning_rate": 1.9251538713266147e-05, "loss": 0.0164, "step": 72530 }, { "grad_norm": 0.30833661556243896, "learning_rate": 1.9238501940223358e-05, "loss": 0.0195, "step": 72540 }, { "grad_norm": 0.23170587420463562, "learning_rate": 1.922546853121175e-05, "loss": 0.0191, "step": 72550 }, { "grad_norm": 0.26149117946624756, "learning_rate": 1.9212438487656647e-05, "loss": 0.0219, "step": 72560 }, { "grad_norm": 0.38112398982048035, "learning_rate": 1.9199411810982965e-05, "loss": 0.0175, "step": 72570 }, { "grad_norm": 0.3849884569644928, "learning_rate": 1.918638850261532e-05, "loss": 0.0198, "step": 72580 }, { "grad_norm": 0.15308967232704163, "learning_rate": 1.9173368563977885e-05, "loss": 0.0192, "step": 72590 }, { "grad_norm": 0.11501148343086243, "learning_rate": 1.9160351996494513e-05, "loss": 0.0174, "step": 72600 }, { "grad_norm": 0.29293182492256165, "learning_rate": 1.9147338801588677e-05, "loss": 0.0226, "step": 72610 }, { "grad_norm": 0.43591994047164917, "learning_rate": 1.913432898068351e-05, "loss": 0.0229, "step": 72620 }, { "grad_norm": 0.32335716485977173, "learning_rate": 1.912132253520169e-05, "loss": 0.0224, "step": 72630 }, { "grad_norm": 0.35768958926200867, "learning_rate": 1.910831946656564e-05, "loss": 0.0268, "step": 72640 }, { "grad_norm": 0.24730493128299713, "learning_rate": 1.909531977619731e-05, "loss": 0.0145, "step": 72650 }, { "grad_norm": 0.15417568385601044, "learning_rate": 1.908232346551834e-05, "loss": 0.0155, "step": 72660 }, { "grad_norm": 0.4413178563117981, "learning_rate": 1.9069330535949993e-05, "loss": 0.0248, "step": 72670 }, { "grad_norm": 0.3638072907924652, "learning_rate": 1.9056340988913167e-05, "loss": 0.0167, "step": 72680 }, { "grad_norm": 1.0404044389724731, "learning_rate": 1.9043354825828357e-05, "loss": 0.0182, "step": 72690 }, { "grad_norm": 0.2450741082429886, "learning_rate": 1.9030372048115724e-05, "loss": 0.0196, "step": 72700 }, { "grad_norm": 0.29993394017219543, "learning_rate": 1.9017392657195056e-05, "loss": 0.0346, "step": 72710 }, { "grad_norm": 0.15831425786018372, "learning_rate": 1.9004416654485724e-05, "loss": 0.0207, "step": 72720 }, { "grad_norm": 0.44959887862205505, "learning_rate": 1.8991444041406785e-05, "loss": 0.0206, "step": 72730 }, { "grad_norm": 0.2869938910007477, "learning_rate": 1.8978474819376912e-05, "loss": 0.0216, "step": 72740 }, { "grad_norm": 0.34471651911735535, "learning_rate": 1.8965508989814406e-05, "loss": 0.0211, "step": 72750 }, { "grad_norm": 0.266791433095932, "learning_rate": 1.8952546554137158e-05, "loss": 0.0248, "step": 72760 }, { "grad_norm": 0.2489321082830429, "learning_rate": 1.8939587513762757e-05, "loss": 0.0156, "step": 72770 }, { "grad_norm": 0.1715446263551712, "learning_rate": 1.892663187010835e-05, "loss": 0.0185, "step": 72780 }, { "grad_norm": 0.5887241959571838, "learning_rate": 1.891367962459077e-05, "loss": 0.0225, "step": 72790 }, { "grad_norm": 0.13824521005153656, "learning_rate": 1.890073077862644e-05, "loss": 0.0215, "step": 72800 }, { "grad_norm": 0.16702023148536682, "learning_rate": 1.888778533363146e-05, "loss": 0.0184, "step": 72810 }, { "grad_norm": 0.3903496265411377, "learning_rate": 1.8874843291021478e-05, "loss": 0.0169, "step": 72820 }, { "grad_norm": 0.2889997959136963, "learning_rate": 1.8861904652211853e-05, "loss": 0.0229, "step": 72830 }, { "grad_norm": 0.9474822282791138, "learning_rate": 1.8848969418617507e-05, "loss": 0.0203, "step": 72840 }, { "grad_norm": 0.27246901392936707, "learning_rate": 1.8836037591653044e-05, "loss": 0.0212, "step": 72850 }, { "grad_norm": 0.2650180459022522, "learning_rate": 1.882310917273262e-05, "loss": 0.0205, "step": 72860 }, { "grad_norm": 0.24486032128334045, "learning_rate": 1.881018416327014e-05, "loss": 0.016, "step": 72870 }, { "grad_norm": 0.3474883735179901, "learning_rate": 1.8797262564679014e-05, "loss": 0.0228, "step": 72880 }, { "grad_norm": 0.5757020711898804, "learning_rate": 1.8784344378372355e-05, "loss": 0.0152, "step": 72890 }, { "grad_norm": 0.2739773392677307, "learning_rate": 1.8771429605762843e-05, "loss": 0.0235, "step": 72900 }, { "grad_norm": 0.2635950744152069, "learning_rate": 1.8758518248262835e-05, "loss": 0.0172, "step": 72910 }, { "grad_norm": 0.6969961524009705, "learning_rate": 1.8745610307284305e-05, "loss": 0.0284, "step": 72920 }, { "grad_norm": 0.2631533741950989, "learning_rate": 1.8732705784238862e-05, "loss": 0.0182, "step": 72930 }, { "grad_norm": 0.1752411276102066, "learning_rate": 1.8719804680537682e-05, "loss": 0.0191, "step": 72940 }, { "grad_norm": 0.29512014985084534, "learning_rate": 1.870690699759164e-05, "loss": 0.0226, "step": 72950 }, { "grad_norm": 0.27154386043548584, "learning_rate": 1.8694012736811213e-05, "loss": 0.0241, "step": 72960 }, { "grad_norm": 0.6397536993026733, "learning_rate": 1.8681121899606463e-05, "loss": 0.0208, "step": 72970 }, { "grad_norm": 0.5761160850524902, "learning_rate": 1.8668234487387144e-05, "loss": 0.0242, "step": 72980 }, { "grad_norm": 0.5964776873588562, "learning_rate": 1.8655350501562592e-05, "loss": 0.0182, "step": 72990 }, { "grad_norm": 0.25486820936203003, "learning_rate": 1.8642469943541796e-05, "loss": 0.0233, "step": 73000 }, { "grad_norm": 0.580599844455719, "learning_rate": 1.8629592814733328e-05, "loss": 0.0201, "step": 73010 }, { "grad_norm": 0.19706347584724426, "learning_rate": 1.8616719116545438e-05, "loss": 0.0159, "step": 73020 }, { "grad_norm": 1.1414663791656494, "learning_rate": 1.8603848850385952e-05, "loss": 0.0189, "step": 73030 }, { "grad_norm": 0.20875342190265656, "learning_rate": 1.859098201766236e-05, "loss": 0.0168, "step": 73040 }, { "grad_norm": 0.6032328605651855, "learning_rate": 1.8578118619781726e-05, "loss": 0.0208, "step": 73050 }, { "grad_norm": 0.48670661449432373, "learning_rate": 1.856525865815083e-05, "loss": 0.0207, "step": 73060 }, { "grad_norm": 0.5679488182067871, "learning_rate": 1.8552402134175966e-05, "loss": 0.0166, "step": 73070 }, { "grad_norm": 0.3242480456829071, "learning_rate": 1.8539549049263143e-05, "loss": 0.0173, "step": 73080 }, { "grad_norm": 0.3298462927341461, "learning_rate": 1.8526699404817917e-05, "loss": 0.0162, "step": 73090 }, { "grad_norm": 0.47838616371154785, "learning_rate": 1.8513853202245546e-05, "loss": 0.0252, "step": 73100 }, { "grad_norm": 0.3204640746116638, "learning_rate": 1.850101044295081e-05, "loss": 0.0144, "step": 73110 }, { "grad_norm": 0.29324519634246826, "learning_rate": 1.848817112833825e-05, "loss": 0.0181, "step": 73120 }, { "grad_norm": 0.867193877696991, "learning_rate": 1.8475335259811905e-05, "loss": 0.0204, "step": 73130 }, { "grad_norm": 0.3319356143474579, "learning_rate": 1.8462502838775508e-05, "loss": 0.0259, "step": 73140 }, { "grad_norm": 0.3040696978569031, "learning_rate": 1.8449673866632378e-05, "loss": 0.0264, "step": 73150 }, { "grad_norm": 0.14720940589904785, "learning_rate": 1.8436848344785473e-05, "loss": 0.0245, "step": 73160 }, { "grad_norm": 0.2508912682533264, "learning_rate": 1.8424026274637378e-05, "loss": 0.0163, "step": 73170 }, { "grad_norm": 0.19124239683151245, "learning_rate": 1.8411207657590312e-05, "loss": 0.0182, "step": 73180 }, { "grad_norm": 0.19129836559295654, "learning_rate": 1.839839249504607e-05, "loss": 0.0169, "step": 73190 }, { "grad_norm": 0.22320999205112457, "learning_rate": 1.8385580788406108e-05, "loss": 0.0194, "step": 73200 }, { "grad_norm": 0.35436108708381653, "learning_rate": 1.837277253907152e-05, "loss": 0.0152, "step": 73210 }, { "grad_norm": 0.28889384865760803, "learning_rate": 1.8359967748442957e-05, "loss": 0.0168, "step": 73220 }, { "grad_norm": 0.21243678033351898, "learning_rate": 1.8347166417920774e-05, "loss": 0.0148, "step": 73230 }, { "grad_norm": 0.22780504822731018, "learning_rate": 1.833436854890484e-05, "loss": 0.0241, "step": 73240 }, { "grad_norm": 0.2044794112443924, "learning_rate": 1.8321574142794796e-05, "loss": 0.0179, "step": 73250 }, { "grad_norm": 0.3076384961605072, "learning_rate": 1.8308783200989748e-05, "loss": 0.0149, "step": 73260 }, { "grad_norm": 0.3982655704021454, "learning_rate": 1.829599572488855e-05, "loss": 0.0202, "step": 73270 }, { "grad_norm": 0.6114745736122131, "learning_rate": 1.828321171588957e-05, "loss": 0.017, "step": 73280 }, { "grad_norm": 0.41544467210769653, "learning_rate": 1.827043117539089e-05, "loss": 0.0212, "step": 73290 }, { "grad_norm": 0.5778636336326599, "learning_rate": 1.8257654104790116e-05, "loss": 0.0204, "step": 73300 }, { "grad_norm": 0.31731903553009033, "learning_rate": 1.82448805054846e-05, "loss": 0.0147, "step": 73310 }, { "grad_norm": 0.2818782925605774, "learning_rate": 1.823211037887119e-05, "loss": 0.0222, "step": 73320 }, { "grad_norm": 0.4861329197883606, "learning_rate": 1.821934372634645e-05, "loss": 0.0143, "step": 73330 }, { "grad_norm": 0.7507821321487427, "learning_rate": 1.8206580549306472e-05, "loss": 0.0184, "step": 73340 }, { "grad_norm": 0.4312303960323334, "learning_rate": 1.819382084914707e-05, "loss": 0.03, "step": 73350 }, { "grad_norm": 0.2605774402618408, "learning_rate": 1.8181064627263577e-05, "loss": 0.0171, "step": 73360 }, { "grad_norm": 0.3230723440647125, "learning_rate": 1.8168311885051022e-05, "loss": 0.0186, "step": 73370 }, { "grad_norm": 0.20630477368831635, "learning_rate": 1.815556262390402e-05, "loss": 0.0237, "step": 73380 }, { "grad_norm": 0.7143170833587646, "learning_rate": 1.8142816845216826e-05, "loss": 0.0178, "step": 73390 }, { "grad_norm": 0.23860052227973938, "learning_rate": 1.8130074550383276e-05, "loss": 0.0201, "step": 73400 }, { "grad_norm": 0.2645273506641388, "learning_rate": 1.8117335740796848e-05, "loss": 0.016, "step": 73410 }, { "grad_norm": 0.3093896210193634, "learning_rate": 1.810460041785067e-05, "loss": 0.0272, "step": 73420 }, { "grad_norm": 0.21098381280899048, "learning_rate": 1.809186858293742e-05, "loss": 0.024, "step": 73430 }, { "grad_norm": 0.22512821853160858, "learning_rate": 1.8079140237449448e-05, "loss": 0.0221, "step": 73440 }, { "grad_norm": 0.7143855690956116, "learning_rate": 1.8066415382778707e-05, "loss": 0.0302, "step": 73450 }, { "grad_norm": 0.30130866169929504, "learning_rate": 1.8053694020316786e-05, "loss": 0.0151, "step": 73460 }, { "grad_norm": 0.4424453377723694, "learning_rate": 1.8040976151454837e-05, "loss": 0.0258, "step": 73470 }, { "grad_norm": 0.457764595746994, "learning_rate": 1.8028261777583704e-05, "loss": 0.0256, "step": 73480 }, { "grad_norm": 0.15614856779575348, "learning_rate": 1.8015550900093758e-05, "loss": 0.0153, "step": 73490 }, { "grad_norm": 0.3637429177761078, "learning_rate": 1.800284352037512e-05, "loss": 0.0211, "step": 73500 }, { "grad_norm": 0.2701592743396759, "learning_rate": 1.799013963981738e-05, "loss": 0.0181, "step": 73510 }, { "grad_norm": 0.3187335729598999, "learning_rate": 1.797743925980987e-05, "loss": 0.0186, "step": 73520 }, { "grad_norm": 0.15429188311100006, "learning_rate": 1.7964742381741435e-05, "loss": 0.0114, "step": 73530 }, { "grad_norm": 0.36007025837898254, "learning_rate": 1.7952049007000627e-05, "loss": 0.0176, "step": 73540 }, { "grad_norm": 0.5790529251098633, "learning_rate": 1.7939359136975535e-05, "loss": 0.0263, "step": 73550 }, { "grad_norm": 0.3082820177078247, "learning_rate": 1.7926672773053932e-05, "loss": 0.0196, "step": 73560 }, { "grad_norm": 0.7478104829788208, "learning_rate": 1.791398991662317e-05, "loss": 0.0283, "step": 73570 }, { "grad_norm": 0.2769967317581177, "learning_rate": 1.7901310569070244e-05, "loss": 0.0286, "step": 73580 }, { "grad_norm": 0.3923510015010834, "learning_rate": 1.7888634731781724e-05, "loss": 0.0191, "step": 73590 }, { "grad_norm": 0.27687105536460876, "learning_rate": 1.7875962406143844e-05, "loss": 0.0149, "step": 73600 }, { "grad_norm": 0.5602210164070129, "learning_rate": 1.7863293593542403e-05, "loss": 0.0321, "step": 73610 }, { "grad_norm": 0.20559926331043243, "learning_rate": 1.7850628295362853e-05, "loss": 0.0202, "step": 73620 }, { "grad_norm": 0.2507826089859009, "learning_rate": 1.7837966512990263e-05, "loss": 0.0223, "step": 73630 }, { "grad_norm": 0.5713928937911987, "learning_rate": 1.782530824780931e-05, "loss": 0.0212, "step": 73640 }, { "grad_norm": 0.17355290055274963, "learning_rate": 1.7812653501204263e-05, "loss": 0.0185, "step": 73650 }, { "grad_norm": 0.31382498145103455, "learning_rate": 1.780000227455903e-05, "loss": 0.0233, "step": 73660 }, { "grad_norm": 0.254219114780426, "learning_rate": 1.778735456925716e-05, "loss": 0.0178, "step": 73670 }, { "grad_norm": 0.38301658630371094, "learning_rate": 1.7774710386681737e-05, "loss": 0.0144, "step": 73680 }, { "grad_norm": 0.2648943066596985, "learning_rate": 1.7762069728215545e-05, "loss": 0.0169, "step": 73690 }, { "grad_norm": 0.3322746753692627, "learning_rate": 1.7749432595240932e-05, "loss": 0.0192, "step": 73700 }, { "grad_norm": 0.18647508323192596, "learning_rate": 1.773679898913991e-05, "loss": 0.0202, "step": 73710 }, { "grad_norm": 0.7490589022636414, "learning_rate": 1.7724168911294015e-05, "loss": 0.017, "step": 73720 }, { "grad_norm": 0.1785900592803955, "learning_rate": 1.7711542363084504e-05, "loss": 0.0163, "step": 73730 }, { "grad_norm": 0.326707124710083, "learning_rate": 1.7698919345892158e-05, "loss": 0.0217, "step": 73740 }, { "grad_norm": 0.6004898548126221, "learning_rate": 1.768629986109743e-05, "loss": 0.0262, "step": 73750 }, { "grad_norm": 0.4718066155910492, "learning_rate": 1.7673683910080357e-05, "loss": 0.0205, "step": 73760 }, { "grad_norm": 1.7382138967514038, "learning_rate": 1.7661071494220638e-05, "loss": 0.0213, "step": 73770 }, { "grad_norm": 0.16030579805374146, "learning_rate": 1.7648462614897498e-05, "loss": 0.0195, "step": 73780 }, { "grad_norm": 0.241336852312088, "learning_rate": 1.7635857273489863e-05, "loss": 0.0148, "step": 73790 }, { "grad_norm": 0.4809955358505249, "learning_rate": 1.7623255471376198e-05, "loss": 0.0172, "step": 73800 }, { "grad_norm": 0.4596666097640991, "learning_rate": 1.7610657209934646e-05, "loss": 0.0184, "step": 73810 }, { "grad_norm": 0.5142495632171631, "learning_rate": 1.7598062490542915e-05, "loss": 0.0241, "step": 73820 }, { "grad_norm": 0.3591676950454712, "learning_rate": 1.7585471314578377e-05, "loss": 0.0189, "step": 73830 }, { "grad_norm": 0.3699790835380554, "learning_rate": 1.7572883683417945e-05, "loss": 0.0157, "step": 73840 }, { "grad_norm": 0.37570708990097046, "learning_rate": 1.7560299598438212e-05, "loss": 0.0208, "step": 73850 }, { "grad_norm": 0.31555846333503723, "learning_rate": 1.7547719061015334e-05, "loss": 0.0176, "step": 73860 }, { "grad_norm": 0.35747987031936646, "learning_rate": 1.75351420725251e-05, "loss": 0.0169, "step": 73870 }, { "grad_norm": 0.20640797913074493, "learning_rate": 1.7522568634342923e-05, "loss": 0.0183, "step": 73880 }, { "grad_norm": 0.260517954826355, "learning_rate": 1.7509998747843825e-05, "loss": 0.0125, "step": 73890 }, { "grad_norm": 0.23716872930526733, "learning_rate": 1.7497432414402403e-05, "loss": 0.0153, "step": 73900 }, { "grad_norm": 0.4353656470775604, "learning_rate": 1.7484869635392903e-05, "loss": 0.0165, "step": 73910 }, { "grad_norm": 0.35588356852531433, "learning_rate": 1.747231041218919e-05, "loss": 0.0208, "step": 73920 }, { "grad_norm": 0.23914404213428497, "learning_rate": 1.7459754746164692e-05, "loss": 0.0178, "step": 73930 }, { "grad_norm": 0.19046004116535187, "learning_rate": 1.744720263869249e-05, "loss": 0.0199, "step": 73940 }, { "grad_norm": 0.2557552456855774, "learning_rate": 1.7434654091145263e-05, "loss": 0.0156, "step": 73950 }, { "grad_norm": 0.3653257489204407, "learning_rate": 1.742210910489532e-05, "loss": 0.0171, "step": 73960 }, { "grad_norm": 0.3688117563724518, "learning_rate": 1.7409567681314532e-05, "loss": 0.0237, "step": 73970 }, { "grad_norm": 0.3109109103679657, "learning_rate": 1.7397029821774434e-05, "loss": 0.0201, "step": 73980 }, { "grad_norm": 0.22698535025119781, "learning_rate": 1.7384495527646126e-05, "loss": 0.0226, "step": 73990 }, { "grad_norm": 1.02814519405365, "learning_rate": 1.737196480030035e-05, "loss": 0.025, "step": 74000 }, { "grad_norm": 0.3361912965774536, "learning_rate": 1.7359437641107447e-05, "loss": 0.0223, "step": 74010 }, { "grad_norm": 0.723673403263092, "learning_rate": 1.7346914051437387e-05, "loss": 0.0189, "step": 74020 }, { "grad_norm": 0.2882150113582611, "learning_rate": 1.733439403265969e-05, "loss": 0.0212, "step": 74030 }, { "grad_norm": 0.3181379735469818, "learning_rate": 1.732187758614357e-05, "loss": 0.0292, "step": 74040 }, { "grad_norm": 0.4311082661151886, "learning_rate": 1.730936471325777e-05, "loss": 0.0214, "step": 74050 }, { "grad_norm": 0.3660929501056671, "learning_rate": 1.7296855415370718e-05, "loss": 0.0198, "step": 74060 }, { "grad_norm": 0.3792535364627838, "learning_rate": 1.728434969385035e-05, "loss": 0.0238, "step": 74070 }, { "grad_norm": 0.271663099527359, "learning_rate": 1.7271847550064346e-05, "loss": 0.0233, "step": 74080 }, { "grad_norm": 0.2113189399242401, "learning_rate": 1.725934898537987e-05, "loss": 0.0199, "step": 74090 }, { "grad_norm": 1.1092405319213867, "learning_rate": 1.7246854001163783e-05, "loss": 0.0152, "step": 74100 }, { "grad_norm": 0.6289981007575989, "learning_rate": 1.7234362598782476e-05, "loss": 0.0299, "step": 74110 }, { "grad_norm": 0.2785979211330414, "learning_rate": 1.722187477960202e-05, "loss": 0.022, "step": 74120 }, { "grad_norm": 0.5019844770431519, "learning_rate": 1.7209390544988057e-05, "loss": 0.0211, "step": 74130 }, { "grad_norm": 0.31838616728782654, "learning_rate": 1.7196909896305856e-05, "loss": 0.025, "step": 74140 }, { "grad_norm": 0.5792514681816101, "learning_rate": 1.718443283492026e-05, "loss": 0.0228, "step": 74150 }, { "grad_norm": 0.43636995553970337, "learning_rate": 1.7171959362195746e-05, "loss": 0.0295, "step": 74160 }, { "grad_norm": 0.4809197783470154, "learning_rate": 1.715948947949642e-05, "loss": 0.021, "step": 74170 }, { "grad_norm": 0.40981781482696533, "learning_rate": 1.714702318818593e-05, "loss": 0.0215, "step": 74180 }, { "grad_norm": 0.3649711310863495, "learning_rate": 1.7134560489627592e-05, "loss": 0.0209, "step": 74190 }, { "grad_norm": 0.4053644835948944, "learning_rate": 1.712210138518431e-05, "loss": 0.0152, "step": 74200 }, { "grad_norm": 0.322998046875, "learning_rate": 1.71096458762186e-05, "loss": 0.0294, "step": 74210 }, { "grad_norm": 0.22199919819831848, "learning_rate": 1.7097193964092552e-05, "loss": 0.0281, "step": 74220 }, { "grad_norm": 0.5260500907897949, "learning_rate": 1.708474565016792e-05, "loss": 0.0294, "step": 74230 }, { "grad_norm": 0.23202180862426758, "learning_rate": 1.7072300935806e-05, "loss": 0.0187, "step": 74240 }, { "grad_norm": 0.45802927017211914, "learning_rate": 1.705985982236776e-05, "loss": 0.0197, "step": 74250 }, { "grad_norm": 0.2063295990228653, "learning_rate": 1.7047422311213697e-05, "loss": 0.0189, "step": 74260 }, { "grad_norm": 0.3222154378890991, "learning_rate": 1.703498840370401e-05, "loss": 0.0252, "step": 74270 }, { "grad_norm": 0.26766031980514526, "learning_rate": 1.702255810119842e-05, "loss": 0.0146, "step": 74280 }, { "grad_norm": 0.23172564804553986, "learning_rate": 1.7010131405056305e-05, "loss": 0.015, "step": 74290 }, { "grad_norm": 0.33595606684684753, "learning_rate": 1.6997708316636613e-05, "loss": 0.0224, "step": 74300 }, { "grad_norm": 0.5021001696586609, "learning_rate": 1.6985288837297932e-05, "loss": 0.0228, "step": 74310 }, { "grad_norm": 0.3086634576320648, "learning_rate": 1.6972872968398394e-05, "loss": 0.0174, "step": 74320 }, { "grad_norm": 0.39447954297065735, "learning_rate": 1.6960460711295845e-05, "loss": 0.0177, "step": 74330 }, { "grad_norm": 0.21126879751682281, "learning_rate": 1.6948052067347623e-05, "loss": 0.0182, "step": 74340 }, { "grad_norm": 0.2116389125585556, "learning_rate": 1.693564703791075e-05, "loss": 0.024, "step": 74350 }, { "grad_norm": 0.2953614890575409, "learning_rate": 1.6923245624341793e-05, "loss": 0.0223, "step": 74360 }, { "grad_norm": 1.2220685482025146, "learning_rate": 1.6910847827996962e-05, "loss": 0.0247, "step": 74370 }, { "grad_norm": 0.18702910840511322, "learning_rate": 1.689845365023206e-05, "loss": 0.0241, "step": 74380 }, { "grad_norm": 0.6663099527359009, "learning_rate": 1.6886063092402505e-05, "loss": 0.0265, "step": 74390 }, { "grad_norm": 0.7332571148872375, "learning_rate": 1.6873676155863317e-05, "loss": 0.0226, "step": 74400 }, { "grad_norm": 0.320229709148407, "learning_rate": 1.6861292841969085e-05, "loss": 0.0195, "step": 74410 }, { "grad_norm": 0.19626672565937042, "learning_rate": 1.684891315207406e-05, "loss": 0.0226, "step": 74420 }, { "grad_norm": 0.3119548261165619, "learning_rate": 1.6836537087532034e-05, "loss": 0.0153, "step": 74430 }, { "grad_norm": 0.6895332336425781, "learning_rate": 1.6824164649696465e-05, "loss": 0.0241, "step": 74440 }, { "grad_norm": 0.22637991607189178, "learning_rate": 1.681179583992034e-05, "loss": 0.0176, "step": 74450 }, { "grad_norm": 0.28525227308273315, "learning_rate": 1.679943065955636e-05, "loss": 0.0215, "step": 74460 }, { "grad_norm": 0.11471843719482422, "learning_rate": 1.678706910995671e-05, "loss": 0.0161, "step": 74470 }, { "grad_norm": 0.23093323409557343, "learning_rate": 1.6774711192473262e-05, "loss": 0.0213, "step": 74480 }, { "grad_norm": 0.26480525732040405, "learning_rate": 1.676235690845743e-05, "loss": 0.016, "step": 74490 }, { "grad_norm": 0.45838358998298645, "learning_rate": 1.6750006259260285e-05, "loss": 0.0166, "step": 74500 }, { "grad_norm": 0.25704285502433777, "learning_rate": 1.673765924623244e-05, "loss": 0.0192, "step": 74510 }, { "grad_norm": 0.4425073564052582, "learning_rate": 1.67253158707242e-05, "loss": 0.0178, "step": 74520 }, { "grad_norm": 0.2763408124446869, "learning_rate": 1.671297613408537e-05, "loss": 0.0189, "step": 74530 }, { "grad_norm": 0.18207642436027527, "learning_rate": 1.6700640037665444e-05, "loss": 0.0198, "step": 74540 }, { "grad_norm": 0.17145556211471558, "learning_rate": 1.668830758281344e-05, "loss": 0.0175, "step": 74550 }, { "grad_norm": 0.8339816331863403, "learning_rate": 1.6675978770878052e-05, "loss": 0.0219, "step": 74560 }, { "grad_norm": 0.3568556308746338, "learning_rate": 1.6663653603207506e-05, "loss": 0.0186, "step": 74570 }, { "grad_norm": 0.6636340022087097, "learning_rate": 1.6651332081149685e-05, "loss": 0.0212, "step": 74580 }, { "grad_norm": 0.3178842067718506, "learning_rate": 1.6639014206052046e-05, "loss": 0.0212, "step": 74590 }, { "grad_norm": 0.6169065833091736, "learning_rate": 1.6626699979261657e-05, "loss": 0.0207, "step": 74600 }, { "grad_norm": 0.2959808111190796, "learning_rate": 1.6614389402125197e-05, "loss": 0.0171, "step": 74610 }, { "grad_norm": 0.7121936678886414, "learning_rate": 1.66020824759889e-05, "loss": 0.0249, "step": 74620 }, { "grad_norm": 0.7174150943756104, "learning_rate": 1.6589779202198675e-05, "loss": 0.0167, "step": 74630 }, { "grad_norm": 0.20617040991783142, "learning_rate": 1.6577479582099924e-05, "loss": 0.0197, "step": 74640 }, { "grad_norm": 0.5622519254684448, "learning_rate": 1.6565183617037788e-05, "loss": 0.0263, "step": 74650 }, { "grad_norm": 0.4612925052642822, "learning_rate": 1.6552891308356894e-05, "loss": 0.0201, "step": 74660 }, { "grad_norm": 0.3070487976074219, "learning_rate": 1.6540602657401526e-05, "loss": 0.026, "step": 74670 }, { "grad_norm": 0.3156411945819855, "learning_rate": 1.6528317665515537e-05, "loss": 0.0312, "step": 74680 }, { "grad_norm": 0.20396016538143158, "learning_rate": 1.6516036334042416e-05, "loss": 0.0183, "step": 74690 }, { "grad_norm": 0.217291921377182, "learning_rate": 1.650375866432519e-05, "loss": 0.0205, "step": 74700 }, { "grad_norm": 0.35236167907714844, "learning_rate": 1.649148465770659e-05, "loss": 0.0196, "step": 74710 }, { "grad_norm": 0.2532925307750702, "learning_rate": 1.647921431552883e-05, "loss": 0.0196, "step": 74720 }, { "grad_norm": 0.2563595473766327, "learning_rate": 1.646694763913381e-05, "loss": 0.0197, "step": 74730 }, { "grad_norm": 0.5396014451980591, "learning_rate": 1.6454684629862966e-05, "loss": 0.0163, "step": 74740 }, { "grad_norm": 0.8097100853919983, "learning_rate": 1.6442425289057388e-05, "loss": 0.0167, "step": 74750 }, { "grad_norm": 0.3371105194091797, "learning_rate": 1.6430169618057718e-05, "loss": 0.0184, "step": 74760 }, { "grad_norm": 0.28106752038002014, "learning_rate": 1.641791761820422e-05, "loss": 0.0218, "step": 74770 }, { "grad_norm": 0.3094986379146576, "learning_rate": 1.640566929083676e-05, "loss": 0.0221, "step": 74780 }, { "grad_norm": 0.25498083233833313, "learning_rate": 1.6393424637294823e-05, "loss": 0.0267, "step": 74790 }, { "grad_norm": 0.3831791281700134, "learning_rate": 1.638118365891742e-05, "loss": 0.0232, "step": 74800 }, { "grad_norm": 0.4529593884944916, "learning_rate": 1.636894635704323e-05, "loss": 0.0253, "step": 74810 }, { "grad_norm": 0.9376895427703857, "learning_rate": 1.6356712733010517e-05, "loss": 0.0259, "step": 74820 }, { "grad_norm": 0.3031677007675171, "learning_rate": 1.6344482788157096e-05, "loss": 0.0131, "step": 74830 }, { "grad_norm": 0.4407421946525574, "learning_rate": 1.6332256523820444e-05, "loss": 0.014, "step": 74840 }, { "grad_norm": 0.2636311948299408, "learning_rate": 1.6320033941337597e-05, "loss": 0.0224, "step": 74850 }, { "grad_norm": 0.4172045588493347, "learning_rate": 1.6307815042045216e-05, "loss": 0.0196, "step": 74860 }, { "grad_norm": 0.7522392272949219, "learning_rate": 1.6295599827279513e-05, "loss": 0.0166, "step": 74870 }, { "grad_norm": 0.6566545367240906, "learning_rate": 1.6283388298376356e-05, "loss": 0.0251, "step": 74880 }, { "grad_norm": 0.5813511610031128, "learning_rate": 1.6271180456671125e-05, "loss": 0.0186, "step": 74890 }, { "grad_norm": 0.355239599943161, "learning_rate": 1.625897630349893e-05, "loss": 0.0169, "step": 74900 }, { "grad_norm": 0.38313519954681396, "learning_rate": 1.6246775840194335e-05, "loss": 0.0173, "step": 74910 }, { "grad_norm": 0.3878963589668274, "learning_rate": 1.6234579068091604e-05, "loss": 0.0183, "step": 74920 }, { "grad_norm": 0.41001152992248535, "learning_rate": 1.6222385988524524e-05, "loss": 0.016, "step": 74930 }, { "grad_norm": 0.2595476508140564, "learning_rate": 1.6210196602826544e-05, "loss": 0.0214, "step": 74940 }, { "grad_norm": 0.7649931311607361, "learning_rate": 1.6198010912330646e-05, "loss": 0.0217, "step": 74950 }, { "grad_norm": 0.4159626066684723, "learning_rate": 1.6185828918369456e-05, "loss": 0.0193, "step": 74960 }, { "grad_norm": 0.3041721284389496, "learning_rate": 1.617365062227517e-05, "loss": 0.0157, "step": 74970 }, { "grad_norm": 0.6341918706893921, "learning_rate": 1.616147602537962e-05, "loss": 0.0175, "step": 74980 }, { "grad_norm": 0.21458035707473755, "learning_rate": 1.614930512901415e-05, "loss": 0.0204, "step": 74990 }, { "grad_norm": 0.6869177222251892, "learning_rate": 1.6137137934509804e-05, "loss": 0.0212, "step": 75000 }, { "grad_norm": 0.23510338366031647, "learning_rate": 1.6124974443197123e-05, "loss": 0.0205, "step": 75010 }, { "grad_norm": 0.4923035502433777, "learning_rate": 1.6112814656406306e-05, "loss": 0.0189, "step": 75020 }, { "grad_norm": 0.43793636560440063, "learning_rate": 1.6100658575467125e-05, "loss": 0.0201, "step": 75030 }, { "grad_norm": 0.2083815336227417, "learning_rate": 1.6088506201708974e-05, "loss": 0.0179, "step": 75040 }, { "grad_norm": 0.3900327980518341, "learning_rate": 1.6076357536460785e-05, "loss": 0.0173, "step": 75050 }, { "grad_norm": 0.23074600100517273, "learning_rate": 1.606421258105113e-05, "loss": 0.021, "step": 75060 }, { "grad_norm": 0.37898895144462585, "learning_rate": 1.6052071336808183e-05, "loss": 0.0188, "step": 75070 }, { "grad_norm": 0.3277799189090729, "learning_rate": 1.6039933805059658e-05, "loss": 0.0193, "step": 75080 }, { "grad_norm": 0.14159350097179413, "learning_rate": 1.6027799987132915e-05, "loss": 0.0163, "step": 75090 }, { "grad_norm": 0.4914592206478119, "learning_rate": 1.6015669884354888e-05, "loss": 0.0245, "step": 75100 }, { "grad_norm": 0.34797021746635437, "learning_rate": 1.6003543498052125e-05, "loss": 0.0199, "step": 75110 }, { "grad_norm": 0.3382452726364136, "learning_rate": 1.5991420829550715e-05, "loss": 0.0201, "step": 75120 }, { "grad_norm": 0.2816982567310333, "learning_rate": 1.5979301880176405e-05, "loss": 0.0214, "step": 75130 }, { "grad_norm": 0.27538707852363586, "learning_rate": 1.5967186651254485e-05, "loss": 0.022, "step": 75140 }, { "grad_norm": 0.4427894651889801, "learning_rate": 1.5955075144109853e-05, "loss": 0.0155, "step": 75150 }, { "grad_norm": 0.3325657844543457, "learning_rate": 1.594296736006703e-05, "loss": 0.028, "step": 75160 }, { "grad_norm": 0.37033167481422424, "learning_rate": 1.59308633004501e-05, "loss": 0.0226, "step": 75170 }, { "grad_norm": 0.9844315052032471, "learning_rate": 1.5918762966582724e-05, "loss": 0.0162, "step": 75180 }, { "grad_norm": 0.2599387764930725, "learning_rate": 1.5906666359788203e-05, "loss": 0.0158, "step": 75190 }, { "grad_norm": 0.3273201286792755, "learning_rate": 1.589457348138938e-05, "loss": 0.0153, "step": 75200 }, { "grad_norm": 0.39787426590919495, "learning_rate": 1.588248433270872e-05, "loss": 0.0222, "step": 75210 }, { "grad_norm": 0.30794858932495117, "learning_rate": 1.587039891506828e-05, "loss": 0.0178, "step": 75220 }, { "grad_norm": 0.7556334733963013, "learning_rate": 1.5858317229789715e-05, "loss": 0.0244, "step": 75230 }, { "grad_norm": 0.4132966995239258, "learning_rate": 1.584623927819423e-05, "loss": 0.0206, "step": 75240 }, { "grad_norm": 0.4186529815196991, "learning_rate": 1.583416506160269e-05, "loss": 0.0147, "step": 75250 }, { "grad_norm": 0.4771162271499634, "learning_rate": 1.5822094581335472e-05, "loss": 0.0258, "step": 75260 }, { "grad_norm": 0.1781434565782547, "learning_rate": 1.5810027838712604e-05, "loss": 0.0199, "step": 75270 }, { "grad_norm": 0.22039538621902466, "learning_rate": 1.5797964835053687e-05, "loss": 0.0218, "step": 75280 }, { "grad_norm": 0.3838481605052948, "learning_rate": 1.5785905571677933e-05, "loss": 0.016, "step": 75290 }, { "grad_norm": 0.433345228433609, "learning_rate": 1.5773850049904086e-05, "loss": 0.0196, "step": 75300 }, { "grad_norm": 0.3127245306968689, "learning_rate": 1.5761798271050543e-05, "loss": 0.0195, "step": 75310 }, { "grad_norm": 0.17350825667381287, "learning_rate": 1.5749750236435277e-05, "loss": 0.0121, "step": 75320 }, { "grad_norm": 0.2626718282699585, "learning_rate": 1.5737705947375815e-05, "loss": 0.0195, "step": 75330 }, { "grad_norm": 0.3333081007003784, "learning_rate": 1.5725665405189322e-05, "loss": 0.0194, "step": 75340 }, { "grad_norm": 0.2555007040500641, "learning_rate": 1.571362861119253e-05, "loss": 0.0175, "step": 75350 }, { "grad_norm": 0.7263197898864746, "learning_rate": 1.570159556670177e-05, "loss": 0.0177, "step": 75360 }, { "grad_norm": 0.39267125725746155, "learning_rate": 1.5689566273032946e-05, "loss": 0.0218, "step": 75370 }, { "grad_norm": 0.8459703326225281, "learning_rate": 1.5677540731501583e-05, "loss": 0.0212, "step": 75380 }, { "grad_norm": 0.37569865584373474, "learning_rate": 1.5665518943422746e-05, "loss": 0.018, "step": 75390 }, { "grad_norm": 0.6507667303085327, "learning_rate": 1.5653500910111134e-05, "loss": 0.0204, "step": 75400 }, { "grad_norm": 0.33227771520614624, "learning_rate": 1.5641486632881023e-05, "loss": 0.0239, "step": 75410 }, { "grad_norm": 0.32029449939727783, "learning_rate": 1.56294761130463e-05, "loss": 0.0226, "step": 75420 }, { "grad_norm": 0.3724157214164734, "learning_rate": 1.5617469351920373e-05, "loss": 0.0238, "step": 75430 }, { "grad_norm": 0.6255595684051514, "learning_rate": 1.5605466350816315e-05, "loss": 0.0227, "step": 75440 }, { "grad_norm": 0.3427965044975281, "learning_rate": 1.559346711104674e-05, "loss": 0.0239, "step": 75450 }, { "grad_norm": 0.6727461218833923, "learning_rate": 1.5581471633923883e-05, "loss": 0.0191, "step": 75460 }, { "grad_norm": 0.34136539697647095, "learning_rate": 1.5569479920759505e-05, "loss": 0.0191, "step": 75470 }, { "grad_norm": 0.6082160472869873, "learning_rate": 1.5557491972865072e-05, "loss": 0.0191, "step": 75480 }, { "grad_norm": 0.4452208876609802, "learning_rate": 1.554550779155152e-05, "loss": 0.0254, "step": 75490 }, { "grad_norm": 0.3411501944065094, "learning_rate": 1.5533527378129453e-05, "loss": 0.0176, "step": 75500 }, { "grad_norm": 0.288322776556015, "learning_rate": 1.552155073390899e-05, "loss": 0.0147, "step": 75510 }, { "grad_norm": 0.4612552225589752, "learning_rate": 1.5509577860199903e-05, "loss": 0.0202, "step": 75520 }, { "grad_norm": 0.48692387342453003, "learning_rate": 1.5497608758311525e-05, "loss": 0.029, "step": 75530 }, { "grad_norm": 0.30421409010887146, "learning_rate": 1.5485643429552797e-05, "loss": 0.014, "step": 75540 }, { "grad_norm": 0.3447670042514801, "learning_rate": 1.5473681875232197e-05, "loss": 0.0186, "step": 75550 }, { "grad_norm": 0.31364819407463074, "learning_rate": 1.5461724096657837e-05, "loss": 0.0138, "step": 75560 }, { "grad_norm": 0.24235956370830536, "learning_rate": 1.5449770095137416e-05, "loss": 0.0162, "step": 75570 }, { "grad_norm": 0.18391671776771545, "learning_rate": 1.5437819871978178e-05, "loss": 0.0176, "step": 75580 }, { "grad_norm": 0.5149939656257629, "learning_rate": 1.5425873428486992e-05, "loss": 0.0283, "step": 75590 }, { "grad_norm": 0.28752124309539795, "learning_rate": 1.5413930765970312e-05, "loss": 0.0179, "step": 75600 }, { "grad_norm": 0.21757328510284424, "learning_rate": 1.5401991885734175e-05, "loss": 0.0178, "step": 75610 }, { "grad_norm": 0.2614951431751251, "learning_rate": 1.5390056789084174e-05, "loss": 0.0149, "step": 75620 }, { "grad_norm": 0.25814008712768555, "learning_rate": 1.5378125477325545e-05, "loss": 0.0192, "step": 75630 }, { "grad_norm": 0.15801967680454254, "learning_rate": 1.5366197951763046e-05, "loss": 0.0179, "step": 75640 }, { "grad_norm": 0.6531274914741516, "learning_rate": 1.5354274213701075e-05, "loss": 0.0324, "step": 75650 }, { "grad_norm": 0.2704862058162689, "learning_rate": 1.5342354264443566e-05, "loss": 0.0184, "step": 75660 }, { "grad_norm": 0.3248012661933899, "learning_rate": 1.533043810529411e-05, "loss": 0.0202, "step": 75670 }, { "grad_norm": 0.31183531880378723, "learning_rate": 1.5318525737555807e-05, "loss": 0.0193, "step": 75680 }, { "grad_norm": 0.3084973692893982, "learning_rate": 1.53066171625314e-05, "loss": 0.0227, "step": 75690 }, { "grad_norm": 0.2927095592021942, "learning_rate": 1.529471238152317e-05, "loss": 0.0239, "step": 75700 }, { "grad_norm": 0.3450779914855957, "learning_rate": 1.5282811395833023e-05, "loss": 0.0145, "step": 75710 }, { "grad_norm": 0.19832228124141693, "learning_rate": 1.5270914206762397e-05, "loss": 0.0178, "step": 75720 }, { "grad_norm": 0.35761988162994385, "learning_rate": 1.525902081561241e-05, "loss": 0.0223, "step": 75730 }, { "grad_norm": 0.6666639447212219, "learning_rate": 1.5247131223683664e-05, "loss": 0.0184, "step": 75740 }, { "grad_norm": 0.32190293073654175, "learning_rate": 1.5235245432276408e-05, "loss": 0.0197, "step": 75750 }, { "grad_norm": 0.26395976543426514, "learning_rate": 1.5223363442690424e-05, "loss": 0.0231, "step": 75760 }, { "grad_norm": 0.2601076364517212, "learning_rate": 1.5211485256225133e-05, "loss": 0.0184, "step": 75770 }, { "grad_norm": 0.40937668085098267, "learning_rate": 1.519961087417951e-05, "loss": 0.0211, "step": 75780 }, { "grad_norm": 0.3036107122898102, "learning_rate": 1.5187740297852137e-05, "loss": 0.0222, "step": 75790 }, { "grad_norm": 0.16063672304153442, "learning_rate": 1.517587352854113e-05, "loss": 0.0107, "step": 75800 }, { "grad_norm": 0.3897559940814972, "learning_rate": 1.516401056754423e-05, "loss": 0.0249, "step": 75810 }, { "grad_norm": 0.32113659381866455, "learning_rate": 1.5152151416158783e-05, "loss": 0.0215, "step": 75820 }, { "grad_norm": 0.7349326610565186, "learning_rate": 1.5140296075681642e-05, "loss": 0.0211, "step": 75830 }, { "grad_norm": 0.2779591679573059, "learning_rate": 1.5128444547409331e-05, "loss": 0.0139, "step": 75840 }, { "grad_norm": 0.3120582103729248, "learning_rate": 1.5116596832637864e-05, "loss": 0.023, "step": 75850 }, { "grad_norm": 0.28032687306404114, "learning_rate": 1.5104752932662952e-05, "loss": 0.0166, "step": 75860 }, { "grad_norm": 0.39262810349464417, "learning_rate": 1.5092912848779778e-05, "loss": 0.0201, "step": 75870 }, { "grad_norm": 0.17326126992702484, "learning_rate": 1.5081076582283194e-05, "loss": 0.0136, "step": 75880 }, { "grad_norm": 0.3905103802680969, "learning_rate": 1.5069244134467553e-05, "loss": 0.0182, "step": 75890 }, { "grad_norm": 0.43091103434562683, "learning_rate": 1.5057415506626876e-05, "loss": 0.0242, "step": 75900 }, { "grad_norm": 0.22290773689746857, "learning_rate": 1.5045590700054668e-05, "loss": 0.0144, "step": 75910 }, { "grad_norm": 0.6032525300979614, "learning_rate": 1.5033769716044144e-05, "loss": 0.0151, "step": 75920 }, { "grad_norm": 0.26585957407951355, "learning_rate": 1.5021952555887975e-05, "loss": 0.0146, "step": 75930 }, { "grad_norm": 0.5604678392410278, "learning_rate": 1.50101392208785e-05, "loss": 0.014, "step": 75940 }, { "grad_norm": 0.1286585032939911, "learning_rate": 1.4998329712307579e-05, "loss": 0.0208, "step": 75950 }, { "grad_norm": 0.3173992931842804, "learning_rate": 1.4986524031466708e-05, "loss": 0.0124, "step": 75960 }, { "grad_norm": 0.3000311255455017, "learning_rate": 1.4974722179646916e-05, "loss": 0.0146, "step": 75970 }, { "grad_norm": 0.26831817626953125, "learning_rate": 1.4962924158138847e-05, "loss": 0.0203, "step": 75980 }, { "grad_norm": 0.16582469642162323, "learning_rate": 1.4951129968232714e-05, "loss": 0.0158, "step": 75990 }, { "grad_norm": 0.4287616014480591, "learning_rate": 1.4939339611218329e-05, "loss": 0.0175, "step": 76000 }, { "grad_norm": 0.3618986904621124, "learning_rate": 1.4927553088385032e-05, "loss": 0.0175, "step": 76010 }, { "grad_norm": 0.29699188470840454, "learning_rate": 1.4915770401021801e-05, "loss": 0.0152, "step": 76020 }, { "grad_norm": 0.3252624571323395, "learning_rate": 1.4903991550417185e-05, "loss": 0.0167, "step": 76030 }, { "grad_norm": 0.34060904383659363, "learning_rate": 1.4892216537859277e-05, "loss": 0.0153, "step": 76040 }, { "grad_norm": 0.4754476249217987, "learning_rate": 1.4880445364635775e-05, "loss": 0.022, "step": 76050 }, { "grad_norm": 0.3246147334575653, "learning_rate": 1.4868678032033973e-05, "loss": 0.0201, "step": 76060 }, { "grad_norm": 0.4160428047180176, "learning_rate": 1.4856914541340739e-05, "loss": 0.0195, "step": 76070 }, { "grad_norm": 0.17465247213840485, "learning_rate": 1.4845154893842472e-05, "loss": 0.0207, "step": 76080 }, { "grad_norm": 0.947066605091095, "learning_rate": 1.4833399090825234e-05, "loss": 0.0194, "step": 76090 }, { "grad_norm": 0.3408980369567871, "learning_rate": 1.4821647133574562e-05, "loss": 0.0248, "step": 76100 }, { "grad_norm": 0.2558104395866394, "learning_rate": 1.4809899023375706e-05, "loss": 0.0146, "step": 76110 }, { "grad_norm": 0.3654535114765167, "learning_rate": 1.4798154761513366e-05, "loss": 0.0179, "step": 76120 }, { "grad_norm": 0.3085830509662628, "learning_rate": 1.4786414349271916e-05, "loss": 0.0222, "step": 76130 }, { "grad_norm": 0.27719518542289734, "learning_rate": 1.4774677787935226e-05, "loss": 0.0153, "step": 76140 }, { "grad_norm": 0.4129945635795593, "learning_rate": 1.4762945078786838e-05, "loss": 0.0163, "step": 76150 }, { "grad_norm": 0.18875236809253693, "learning_rate": 1.4751216223109777e-05, "loss": 0.0183, "step": 76160 }, { "grad_norm": 0.35544833540916443, "learning_rate": 1.4739491222186714e-05, "loss": 0.028, "step": 76170 }, { "grad_norm": 0.4477640390396118, "learning_rate": 1.4727770077299879e-05, "loss": 0.0169, "step": 76180 }, { "grad_norm": 1.1298695802688599, "learning_rate": 1.4716052789731094e-05, "loss": 0.0183, "step": 76190 }, { "grad_norm": 0.25856509804725647, "learning_rate": 1.470433936076171e-05, "loss": 0.0133, "step": 76200 }, { "grad_norm": 0.7799795269966125, "learning_rate": 1.4692629791672719e-05, "loss": 0.0174, "step": 76210 }, { "grad_norm": 0.26239001750946045, "learning_rate": 1.468092408374464e-05, "loss": 0.0158, "step": 76220 }, { "grad_norm": 0.3760702311992645, "learning_rate": 1.4669222238257602e-05, "loss": 0.017, "step": 76230 }, { "grad_norm": 0.2382403165102005, "learning_rate": 1.4657524256491301e-05, "loss": 0.0151, "step": 76240 }, { "grad_norm": 0.26076221466064453, "learning_rate": 1.464583013972502e-05, "loss": 0.0177, "step": 76250 }, { "grad_norm": 0.21493524312973022, "learning_rate": 1.4634139889237591e-05, "loss": 0.0174, "step": 76260 }, { "grad_norm": 0.2742161452770233, "learning_rate": 1.4622453506307448e-05, "loss": 0.0176, "step": 76270 }, { "grad_norm": 0.39064598083496094, "learning_rate": 1.461077099221262e-05, "loss": 0.0171, "step": 76280 }, { "grad_norm": 0.19979725778102875, "learning_rate": 1.4599092348230652e-05, "loss": 0.0181, "step": 76290 }, { "grad_norm": 0.47951287031173706, "learning_rate": 1.4587417575638724e-05, "loss": 0.0184, "step": 76300 }, { "grad_norm": 0.2817670404911041, "learning_rate": 1.4575746675713563e-05, "loss": 0.014, "step": 76310 }, { "grad_norm": 0.23601770401000977, "learning_rate": 1.4564079649731504e-05, "loss": 0.0173, "step": 76320 }, { "grad_norm": 0.26318681240081787, "learning_rate": 1.4552416498968402e-05, "loss": 0.0166, "step": 76330 }, { "grad_norm": 0.2373444139957428, "learning_rate": 1.4540757224699747e-05, "loss": 0.0173, "step": 76340 }, { "grad_norm": 0.37415775656700134, "learning_rate": 1.452910182820056e-05, "loss": 0.0259, "step": 76350 }, { "grad_norm": 0.45069420337677, "learning_rate": 1.4517450310745468e-05, "loss": 0.019, "step": 76360 }, { "grad_norm": 1.0405892133712769, "learning_rate": 1.450580267360866e-05, "loss": 0.0235, "step": 76370 }, { "grad_norm": 0.18048465251922607, "learning_rate": 1.4494158918063922e-05, "loss": 0.0134, "step": 76380 }, { "grad_norm": 0.22683213651180267, "learning_rate": 1.4482519045384568e-05, "loss": 0.0128, "step": 76390 }, { "grad_norm": 0.42859333753585815, "learning_rate": 1.4470883056843548e-05, "loss": 0.0211, "step": 76400 }, { "grad_norm": 0.34103477001190186, "learning_rate": 1.4459250953713322e-05, "loss": 0.021, "step": 76410 }, { "grad_norm": 0.3637416660785675, "learning_rate": 1.444762273726598e-05, "loss": 0.0144, "step": 76420 }, { "grad_norm": 0.23572391271591187, "learning_rate": 1.443599840877316e-05, "loss": 0.0187, "step": 76430 }, { "grad_norm": 0.40047693252563477, "learning_rate": 1.4424377969506103e-05, "loss": 0.021, "step": 76440 }, { "grad_norm": 0.25478631258010864, "learning_rate": 1.4412761420735565e-05, "loss": 0.0185, "step": 76450 }, { "grad_norm": 0.18533562123775482, "learning_rate": 1.4401148763731953e-05, "loss": 0.0214, "step": 76460 }, { "grad_norm": 0.412793904542923, "learning_rate": 1.4389539999765173e-05, "loss": 0.0153, "step": 76470 }, { "grad_norm": 0.2928750813007355, "learning_rate": 1.4377935130104758e-05, "loss": 0.0168, "step": 76480 }, { "grad_norm": 0.1691671460866928, "learning_rate": 1.4366334156019801e-05, "loss": 0.0223, "step": 76490 }, { "grad_norm": 0.2848648428916931, "learning_rate": 1.435473707877898e-05, "loss": 0.0212, "step": 76500 }, { "grad_norm": 0.515897274017334, "learning_rate": 1.4343143899650508e-05, "loss": 0.0195, "step": 76510 }, { "grad_norm": 0.4061746895313263, "learning_rate": 1.4331554619902204e-05, "loss": 0.0243, "step": 76520 }, { "grad_norm": 0.2557639181613922, "learning_rate": 1.4319969240801484e-05, "loss": 0.0209, "step": 76530 }, { "grad_norm": 0.8667983412742615, "learning_rate": 1.430838776361526e-05, "loss": 0.0258, "step": 76540 }, { "grad_norm": 0.3823746144771576, "learning_rate": 1.4296810189610093e-05, "loss": 0.0212, "step": 76550 }, { "grad_norm": 0.28575950860977173, "learning_rate": 1.4285236520052087e-05, "loss": 0.019, "step": 76560 }, { "grad_norm": 0.3580872714519501, "learning_rate": 1.4273666756206933e-05, "loss": 0.0182, "step": 76570 }, { "grad_norm": 0.31783485412597656, "learning_rate": 1.4262100899339848e-05, "loss": 0.0205, "step": 76580 }, { "grad_norm": 0.3012916147708893, "learning_rate": 1.4250538950715696e-05, "loss": 0.0233, "step": 76590 }, { "grad_norm": 0.5246742367744446, "learning_rate": 1.4238980911598842e-05, "loss": 0.0192, "step": 76600 }, { "grad_norm": 0.6084622740745544, "learning_rate": 1.4227426783253267e-05, "loss": 0.0232, "step": 76610 }, { "grad_norm": 0.2394709587097168, "learning_rate": 1.4215876566942521e-05, "loss": 0.0206, "step": 76620 }, { "grad_norm": 0.16323167085647583, "learning_rate": 1.4204330263929721e-05, "loss": 0.0166, "step": 76630 }, { "grad_norm": 0.20010140538215637, "learning_rate": 1.4192787875477537e-05, "loss": 0.0251, "step": 76640 }, { "grad_norm": 0.2116544246673584, "learning_rate": 1.4181249402848246e-05, "loss": 0.0185, "step": 76650 }, { "grad_norm": 0.47961416840553284, "learning_rate": 1.4169714847303656e-05, "loss": 0.0253, "step": 76660 }, { "grad_norm": 0.3339291512966156, "learning_rate": 1.4158184210105197e-05, "loss": 0.0236, "step": 76670 }, { "grad_norm": 0.41373810172080994, "learning_rate": 1.4146657492513787e-05, "loss": 0.0198, "step": 76680 }, { "grad_norm": 0.6809480786323547, "learning_rate": 1.4135134695790047e-05, "loss": 0.0184, "step": 76690 }, { "grad_norm": 0.643168568611145, "learning_rate": 1.4123615821194037e-05, "loss": 0.0224, "step": 76700 }, { "grad_norm": 0.21349544823169708, "learning_rate": 1.4112100869985473e-05, "loss": 0.0145, "step": 76710 }, { "grad_norm": 0.23385149240493774, "learning_rate": 1.4100589843423589e-05, "loss": 0.0147, "step": 76720 }, { "grad_norm": 0.5329350233078003, "learning_rate": 1.408908274276723e-05, "loss": 0.0152, "step": 76730 }, { "grad_norm": 0.5470156669616699, "learning_rate": 1.4077579569274786e-05, "loss": 0.0213, "step": 76740 }, { "grad_norm": 0.36662179231643677, "learning_rate": 1.4066080324204245e-05, "loss": 0.0176, "step": 76750 }, { "grad_norm": 0.9066492319107056, "learning_rate": 1.405458500881312e-05, "loss": 0.0409, "step": 76760 }, { "grad_norm": 0.42467865347862244, "learning_rate": 1.4043093624358538e-05, "loss": 0.0186, "step": 76770 }, { "grad_norm": 0.3607787489891052, "learning_rate": 1.4031606172097189e-05, "loss": 0.0193, "step": 76780 }, { "grad_norm": 0.44371744990348816, "learning_rate": 1.40201226532853e-05, "loss": 0.0203, "step": 76790 }, { "grad_norm": 0.15660612285137177, "learning_rate": 1.4008643069178701e-05, "loss": 0.0125, "step": 76800 }, { "grad_norm": 0.24097633361816406, "learning_rate": 1.399716742103278e-05, "loss": 0.0221, "step": 76810 }, { "grad_norm": 0.2698000371456146, "learning_rate": 1.398569571010252e-05, "loss": 0.0156, "step": 76820 }, { "grad_norm": 0.4435752332210541, "learning_rate": 1.3974227937642414e-05, "loss": 0.0204, "step": 76830 }, { "grad_norm": 0.3141614496707916, "learning_rate": 1.3962764104906595e-05, "loss": 0.0187, "step": 76840 }, { "grad_norm": 0.24712684750556946, "learning_rate": 1.3951304213148696e-05, "loss": 0.0211, "step": 76850 }, { "grad_norm": 0.3342024087905884, "learning_rate": 1.3939848263621979e-05, "loss": 0.0159, "step": 76860 }, { "grad_norm": 0.25445103645324707, "learning_rate": 1.3928396257579219e-05, "loss": 0.021, "step": 76870 }, { "grad_norm": 0.5929247736930847, "learning_rate": 1.3916948196272833e-05, "loss": 0.0176, "step": 76880 }, { "grad_norm": 0.43441566824913025, "learning_rate": 1.3905504080954734e-05, "loss": 0.0273, "step": 76890 }, { "grad_norm": 0.3038930892944336, "learning_rate": 1.3894063912876454e-05, "loss": 0.0139, "step": 76900 }, { "grad_norm": 0.42497146129608154, "learning_rate": 1.3882627693289046e-05, "loss": 0.0208, "step": 76910 }, { "grad_norm": 0.49830955266952515, "learning_rate": 1.3871195423443183e-05, "loss": 0.021, "step": 76920 }, { "grad_norm": 0.3531055748462677, "learning_rate": 1.3859767104589044e-05, "loss": 0.0203, "step": 76930 }, { "grad_norm": 0.28429916501045227, "learning_rate": 1.3848342737976467e-05, "loss": 0.0175, "step": 76940 }, { "grad_norm": 0.37735915184020996, "learning_rate": 1.3836922324854757e-05, "loss": 0.0164, "step": 76950 }, { "grad_norm": 0.5885389447212219, "learning_rate": 1.382550586647287e-05, "loss": 0.0243, "step": 76960 }, { "grad_norm": 0.20392227172851562, "learning_rate": 1.3814093364079257e-05, "loss": 0.0176, "step": 76970 }, { "grad_norm": 0.49537771940231323, "learning_rate": 1.3802684818921991e-05, "loss": 0.0206, "step": 76980 }, { "grad_norm": 0.13141407072544098, "learning_rate": 1.3791280232248687e-05, "loss": 0.021, "step": 76990 }, { "grad_norm": 0.2690834105014801, "learning_rate": 1.3779879605306555e-05, "loss": 0.0215, "step": 77000 }, { "grad_norm": 0.4737410843372345, "learning_rate": 1.376848293934232e-05, "loss": 0.0166, "step": 77010 }, { "grad_norm": 0.2888180911540985, "learning_rate": 1.3757090235602316e-05, "loss": 0.0187, "step": 77020 }, { "grad_norm": 0.19318875670433044, "learning_rate": 1.3745701495332447e-05, "loss": 0.0179, "step": 77030 }, { "grad_norm": 1.0296884775161743, "learning_rate": 1.3734316719778145e-05, "loss": 0.0181, "step": 77040 }, { "grad_norm": 0.24455223977565765, "learning_rate": 1.3722935910184454e-05, "loss": 0.0195, "step": 77050 }, { "grad_norm": 0.24415414035320282, "learning_rate": 1.3711559067795921e-05, "loss": 0.0198, "step": 77060 }, { "grad_norm": 0.7541759014129639, "learning_rate": 1.3700186193856756e-05, "loss": 0.0184, "step": 77070 }, { "grad_norm": 0.4496411383152008, "learning_rate": 1.368881728961064e-05, "loss": 0.0156, "step": 77080 }, { "grad_norm": 0.268878310918808, "learning_rate": 1.3677452356300885e-05, "loss": 0.0209, "step": 77090 }, { "grad_norm": 0.327144593000412, "learning_rate": 1.3666091395170321e-05, "loss": 0.0239, "step": 77100 }, { "grad_norm": 0.2604740858078003, "learning_rate": 1.3654734407461384e-05, "loss": 0.0168, "step": 77110 }, { "grad_norm": 0.5812186002731323, "learning_rate": 1.3643381394416022e-05, "loss": 0.0209, "step": 77120 }, { "grad_norm": 0.7221922278404236, "learning_rate": 1.3632032357275838e-05, "loss": 0.0241, "step": 77130 }, { "grad_norm": 0.38800740242004395, "learning_rate": 1.3620687297281898e-05, "loss": 0.0248, "step": 77140 }, { "grad_norm": 0.6949244141578674, "learning_rate": 1.3609346215674917e-05, "loss": 0.0204, "step": 77150 }, { "grad_norm": 0.3060310184955597, "learning_rate": 1.3598009113695109e-05, "loss": 0.0221, "step": 77160 }, { "grad_norm": 0.20879803597927094, "learning_rate": 1.358667599258231e-05, "loss": 0.0208, "step": 77170 }, { "grad_norm": 0.22240133583545685, "learning_rate": 1.357534685357586e-05, "loss": 0.015, "step": 77180 }, { "grad_norm": 0.2338256537914276, "learning_rate": 1.3564021697914715e-05, "loss": 0.0192, "step": 77190 }, { "grad_norm": 0.2179521769285202, "learning_rate": 1.3552700526837381e-05, "loss": 0.0203, "step": 77200 }, { "grad_norm": 0.25692564249038696, "learning_rate": 1.3541383341581925e-05, "loss": 0.0115, "step": 77210 }, { "grad_norm": 0.2617506980895996, "learning_rate": 1.3530070143385965e-05, "loss": 0.0165, "step": 77220 }, { "grad_norm": 0.3532789945602417, "learning_rate": 1.3518760933486702e-05, "loss": 0.0211, "step": 77230 }, { "grad_norm": 0.2541356384754181, "learning_rate": 1.3507455713120908e-05, "loss": 0.0252, "step": 77240 }, { "grad_norm": 0.13108591735363007, "learning_rate": 1.3496154483524876e-05, "loss": 0.0143, "step": 77250 }, { "grad_norm": 0.22429770231246948, "learning_rate": 1.3484857245934506e-05, "loss": 0.0153, "step": 77260 }, { "grad_norm": 0.3099835515022278, "learning_rate": 1.3473564001585248e-05, "loss": 0.0161, "step": 77270 }, { "grad_norm": 0.27089235186576843, "learning_rate": 1.3462274751712129e-05, "loss": 0.016, "step": 77280 }, { "grad_norm": 0.7290263772010803, "learning_rate": 1.3450989497549698e-05, "loss": 0.0181, "step": 77290 }, { "grad_norm": 0.4105457365512848, "learning_rate": 1.3439708240332115e-05, "loss": 0.0219, "step": 77300 }, { "grad_norm": 0.3202468156814575, "learning_rate": 1.3428430981293043e-05, "loss": 0.0192, "step": 77310 }, { "grad_norm": 0.646136999130249, "learning_rate": 1.3417157721665808e-05, "loss": 0.0231, "step": 77320 }, { "grad_norm": 0.39196744561195374, "learning_rate": 1.3405888462683186e-05, "loss": 0.0193, "step": 77330 }, { "grad_norm": 0.2703380584716797, "learning_rate": 1.3394623205577595e-05, "loss": 0.0238, "step": 77340 }, { "grad_norm": 0.479987233877182, "learning_rate": 1.3383361951580963e-05, "loss": 0.0215, "step": 77350 }, { "grad_norm": 0.41507452726364136, "learning_rate": 1.337210470192483e-05, "loss": 0.0207, "step": 77360 }, { "grad_norm": 0.10394228994846344, "learning_rate": 1.3360851457840245e-05, "loss": 0.0143, "step": 77370 }, { "grad_norm": 0.2136552780866623, "learning_rate": 1.3349602220557856e-05, "loss": 0.0182, "step": 77380 }, { "grad_norm": 0.4393501877784729, "learning_rate": 1.3338356991307865e-05, "loss": 0.0205, "step": 77390 }, { "grad_norm": 0.5039065480232239, "learning_rate": 1.3327115771320048e-05, "loss": 0.0226, "step": 77400 }, { "grad_norm": 0.255069762468338, "learning_rate": 1.3315878561823697e-05, "loss": 0.0182, "step": 77410 }, { "grad_norm": 0.22574032843112946, "learning_rate": 1.330464536404773e-05, "loss": 0.0153, "step": 77420 }, { "grad_norm": 0.26758432388305664, "learning_rate": 1.3293416179220563e-05, "loss": 0.0173, "step": 77430 }, { "grad_norm": 0.28269970417022705, "learning_rate": 1.328219100857021e-05, "loss": 0.014, "step": 77440 }, { "grad_norm": 0.4045643210411072, "learning_rate": 1.327096985332425e-05, "loss": 0.013, "step": 77450 }, { "grad_norm": 0.44465500116348267, "learning_rate": 1.3259752714709816e-05, "loss": 0.0132, "step": 77460 }, { "grad_norm": 1.02347731590271, "learning_rate": 1.3248539593953569e-05, "loss": 0.0151, "step": 77470 }, { "grad_norm": 0.41476163268089294, "learning_rate": 1.323733049228178e-05, "loss": 0.0147, "step": 77480 }, { "grad_norm": 0.4198172688484192, "learning_rate": 1.3226125410920276e-05, "loss": 0.0168, "step": 77490 }, { "grad_norm": 0.5289301872253418, "learning_rate": 1.3214924351094388e-05, "loss": 0.0196, "step": 77500 }, { "grad_norm": 0.20238429307937622, "learning_rate": 1.3203727314029069e-05, "loss": 0.0129, "step": 77510 }, { "grad_norm": 0.23612473905086517, "learning_rate": 1.3192534300948811e-05, "loss": 0.0137, "step": 77520 }, { "grad_norm": 0.28951495885849, "learning_rate": 1.3181345313077675e-05, "loss": 0.0154, "step": 77530 }, { "grad_norm": 0.4764089286327362, "learning_rate": 1.3170160351639243e-05, "loss": 0.0194, "step": 77540 }, { "grad_norm": 0.3839547336101532, "learning_rate": 1.3158979417856721e-05, "loss": 0.0327, "step": 77550 }, { "grad_norm": 0.24967321753501892, "learning_rate": 1.31478025129528e-05, "loss": 0.0171, "step": 77560 }, { "grad_norm": 0.3696317970752716, "learning_rate": 1.3136629638149794e-05, "loss": 0.0123, "step": 77570 }, { "grad_norm": 0.9831200242042542, "learning_rate": 1.3125460794669547e-05, "loss": 0.0149, "step": 77580 }, { "grad_norm": 0.22692127525806427, "learning_rate": 1.3114295983733483e-05, "loss": 0.02, "step": 77590 }, { "grad_norm": 0.5829768180847168, "learning_rate": 1.3103135206562534e-05, "loss": 0.0188, "step": 77600 }, { "grad_norm": 0.28922054171562195, "learning_rate": 1.3091978464377264e-05, "loss": 0.0161, "step": 77610 }, { "grad_norm": 0.23518788814544678, "learning_rate": 1.3080825758397724e-05, "loss": 0.0128, "step": 77620 }, { "grad_norm": 0.31746622920036316, "learning_rate": 1.3069677089843568e-05, "loss": 0.0155, "step": 77630 }, { "grad_norm": 0.2717054486274719, "learning_rate": 1.305853245993401e-05, "loss": 0.0235, "step": 77640 }, { "grad_norm": 0.3336394131183624, "learning_rate": 1.3047391869887809e-05, "loss": 0.0166, "step": 77650 }, { "grad_norm": 0.2600366175174713, "learning_rate": 1.3036255320923263e-05, "loss": 0.0207, "step": 77660 }, { "grad_norm": 0.2808679938316345, "learning_rate": 1.3025122814258273e-05, "loss": 0.0113, "step": 77670 }, { "grad_norm": 0.3373737037181854, "learning_rate": 1.3013994351110249e-05, "loss": 0.0184, "step": 77680 }, { "grad_norm": 0.3791798949241638, "learning_rate": 1.3002869932696194e-05, "loss": 0.0189, "step": 77690 }, { "grad_norm": 0.4155254662036896, "learning_rate": 1.2991749560232664e-05, "loss": 0.0186, "step": 77700 }, { "grad_norm": 0.35450419783592224, "learning_rate": 1.2980633234935768e-05, "loss": 0.0209, "step": 77710 }, { "grad_norm": 0.6653252243995667, "learning_rate": 1.2969520958021153e-05, "loss": 0.0161, "step": 77720 }, { "grad_norm": 0.20453430712223053, "learning_rate": 1.295841273070405e-05, "loss": 0.0139, "step": 77730 }, { "grad_norm": 0.23885488510131836, "learning_rate": 1.2947308554199255e-05, "loss": 0.0156, "step": 77740 }, { "grad_norm": 0.33411821722984314, "learning_rate": 1.2936208429721075e-05, "loss": 0.0196, "step": 77750 }, { "grad_norm": 0.32901668548583984, "learning_rate": 1.2925112358483416e-05, "loss": 0.015, "step": 77760 }, { "grad_norm": 0.9636901021003723, "learning_rate": 1.2914020341699728e-05, "loss": 0.0243, "step": 77770 }, { "grad_norm": 0.30231234431266785, "learning_rate": 1.2902932380583032e-05, "loss": 0.0246, "step": 77780 }, { "grad_norm": 0.42653611302375793, "learning_rate": 1.2891848476345863e-05, "loss": 0.0195, "step": 77790 }, { "grad_norm": 0.20985381305217743, "learning_rate": 1.2880768630200368e-05, "loss": 0.017, "step": 77800 }, { "grad_norm": 0.39631080627441406, "learning_rate": 1.286969284335819e-05, "loss": 0.0233, "step": 77810 }, { "grad_norm": 0.3616993725299835, "learning_rate": 1.2858621117030578e-05, "loss": 0.015, "step": 77820 }, { "grad_norm": 0.27218031883239746, "learning_rate": 1.2847553452428324e-05, "loss": 0.0152, "step": 77830 }, { "grad_norm": 0.2215292602777481, "learning_rate": 1.283648985076178e-05, "loss": 0.0197, "step": 77840 }, { "grad_norm": 0.2981621325016022, "learning_rate": 1.2825430313240817e-05, "loss": 0.0183, "step": 77850 }, { "grad_norm": 0.5223944783210754, "learning_rate": 1.281437484107491e-05, "loss": 0.0188, "step": 77860 }, { "grad_norm": 0.41903913021087646, "learning_rate": 1.2803323435473058e-05, "loss": 0.0186, "step": 77870 }, { "grad_norm": 0.17210379242897034, "learning_rate": 1.2792276097643841e-05, "loss": 0.016, "step": 77880 }, { "grad_norm": 0.508952796459198, "learning_rate": 1.2781232828795342e-05, "loss": 0.0176, "step": 77890 }, { "grad_norm": 0.26049789786338806, "learning_rate": 1.2770193630135292e-05, "loss": 0.0232, "step": 77900 }, { "grad_norm": 0.2756187617778778, "learning_rate": 1.2759158502870877e-05, "loss": 0.0268, "step": 77910 }, { "grad_norm": 0.14299602806568146, "learning_rate": 1.2748127448208913e-05, "loss": 0.0141, "step": 77920 }, { "grad_norm": 0.29944857954978943, "learning_rate": 1.2737100467355706e-05, "loss": 0.0238, "step": 77930 }, { "grad_norm": 0.2513602674007416, "learning_rate": 1.2726077561517169e-05, "loss": 0.0264, "step": 77940 }, { "grad_norm": 0.20254339277744293, "learning_rate": 1.2715058731898755e-05, "loss": 0.0149, "step": 77950 }, { "grad_norm": 0.36090293526649475, "learning_rate": 1.2704043979705471e-05, "loss": 0.0207, "step": 77960 }, { "grad_norm": 0.37987858057022095, "learning_rate": 1.269303330614185e-05, "loss": 0.0192, "step": 77970 }, { "grad_norm": 1.9559314250946045, "learning_rate": 1.2682026712412015e-05, "loss": 0.0168, "step": 77980 }, { "grad_norm": 0.1346108317375183, "learning_rate": 1.2671024199719644e-05, "loss": 0.0144, "step": 77990 }, { "grad_norm": 0.3971388638019562, "learning_rate": 1.2660025769267936e-05, "loss": 0.0224, "step": 78000 }, { "grad_norm": 0.24877481162548065, "learning_rate": 1.2649031422259667e-05, "loss": 0.0176, "step": 78010 }, { "grad_norm": 0.3457472026348114, "learning_rate": 1.2638041159897163e-05, "loss": 0.0157, "step": 78020 }, { "grad_norm": 0.842107355594635, "learning_rate": 1.262705498338232e-05, "loss": 0.0192, "step": 78030 }, { "grad_norm": 0.5531859397888184, "learning_rate": 1.2616072893916542e-05, "loss": 0.0291, "step": 78040 }, { "grad_norm": 0.2685354948043823, "learning_rate": 1.2605094892700836e-05, "loss": 0.0152, "step": 78050 }, { "grad_norm": 0.4336318373680115, "learning_rate": 1.259412098093572e-05, "loss": 0.0175, "step": 78060 }, { "grad_norm": 0.8812552690505981, "learning_rate": 1.258315115982131e-05, "loss": 0.0218, "step": 78070 }, { "grad_norm": 0.5353955030441284, "learning_rate": 1.2572185430557203e-05, "loss": 0.0177, "step": 78080 }, { "grad_norm": 0.27497953176498413, "learning_rate": 1.2561223794342659e-05, "loss": 0.0204, "step": 78090 }, { "grad_norm": 0.692061722278595, "learning_rate": 1.255026625237637e-05, "loss": 0.0195, "step": 78100 }, { "grad_norm": 0.2640630006790161, "learning_rate": 1.2539312805856678e-05, "loss": 0.021, "step": 78110 }, { "grad_norm": 0.5235506296157837, "learning_rate": 1.25283634559814e-05, "loss": 0.0199, "step": 78120 }, { "grad_norm": 0.3494451940059662, "learning_rate": 1.2517418203947972e-05, "loss": 0.0233, "step": 78130 }, { "grad_norm": 0.3133784234523773, "learning_rate": 1.2506477050953308e-05, "loss": 0.0196, "step": 78140 }, { "grad_norm": 0.40871044993400574, "learning_rate": 1.2495539998193972e-05, "loss": 0.017, "step": 78150 }, { "grad_norm": 0.3133503198623657, "learning_rate": 1.2484607046865975e-05, "loss": 0.0184, "step": 78160 }, { "grad_norm": 0.23078352212905884, "learning_rate": 1.2473678198164968e-05, "loss": 0.016, "step": 78170 }, { "grad_norm": 0.7152430415153503, "learning_rate": 1.2462753453286075e-05, "loss": 0.0318, "step": 78180 }, { "grad_norm": 0.24097298085689545, "learning_rate": 1.2451832813424031e-05, "loss": 0.0284, "step": 78190 }, { "grad_norm": 0.34607040882110596, "learning_rate": 1.2440916279773095e-05, "loss": 0.0289, "step": 78200 }, { "grad_norm": 0.38752830028533936, "learning_rate": 1.2430003853527101e-05, "loss": 0.0156, "step": 78210 }, { "grad_norm": 0.5958078503608704, "learning_rate": 1.2419095535879383e-05, "loss": 0.0293, "step": 78220 }, { "grad_norm": 0.1602204144001007, "learning_rate": 1.2408191328022878e-05, "loss": 0.0185, "step": 78230 }, { "grad_norm": 0.23909235000610352, "learning_rate": 1.2397291231150066e-05, "loss": 0.0123, "step": 78240 }, { "grad_norm": 0.31890472769737244, "learning_rate": 1.2386395246452937e-05, "loss": 0.018, "step": 78250 }, { "grad_norm": 0.16481564939022064, "learning_rate": 1.2375503375123083e-05, "loss": 0.0149, "step": 78260 }, { "grad_norm": 0.21498124301433563, "learning_rate": 1.2364615618351583e-05, "loss": 0.0286, "step": 78270 }, { "grad_norm": 0.2913263142108917, "learning_rate": 1.2353731977329169e-05, "loss": 0.0204, "step": 78280 }, { "grad_norm": 0.3620489239692688, "learning_rate": 1.2342852453246007e-05, "loss": 0.0133, "step": 78290 }, { "grad_norm": 0.37351879477500916, "learning_rate": 1.2331977047291898e-05, "loss": 0.0163, "step": 78300 }, { "grad_norm": 0.3667180836200714, "learning_rate": 1.2321105760656137e-05, "loss": 0.0175, "step": 78310 }, { "grad_norm": 0.2713157832622528, "learning_rate": 1.2310238594527613e-05, "loss": 0.0156, "step": 78320 }, { "grad_norm": 0.4425238072872162, "learning_rate": 1.2299375550094693e-05, "loss": 0.0163, "step": 78330 }, { "grad_norm": 0.17526501417160034, "learning_rate": 1.2288516628545421e-05, "loss": 0.0153, "step": 78340 }, { "grad_norm": 0.6556239128112793, "learning_rate": 1.2277661831067255e-05, "loss": 0.0191, "step": 78350 }, { "grad_norm": 0.24864408373832703, "learning_rate": 1.2266811158847286e-05, "loss": 0.0227, "step": 78360 }, { "grad_norm": 0.862169623374939, "learning_rate": 1.2255964613072107e-05, "loss": 0.019, "step": 78370 }, { "grad_norm": 0.3497917950153351, "learning_rate": 1.2245122194927905e-05, "loss": 0.0158, "step": 78380 }, { "grad_norm": 0.47525981068611145, "learning_rate": 1.2234283905600364e-05, "loss": 0.0182, "step": 78390 }, { "grad_norm": 0.20499326288700104, "learning_rate": 1.222344974627475e-05, "loss": 0.0196, "step": 78400 }, { "grad_norm": 0.7912365198135376, "learning_rate": 1.2212619718135875e-05, "loss": 0.0221, "step": 78410 }, { "grad_norm": 0.328848659992218, "learning_rate": 1.2201793822368118e-05, "loss": 0.0136, "step": 78420 }, { "grad_norm": 0.16918207705020905, "learning_rate": 1.2190972060155337e-05, "loss": 0.0165, "step": 78430 }, { "grad_norm": 0.4465782642364502, "learning_rate": 1.2180154432681013e-05, "loss": 0.0241, "step": 78440 }, { "grad_norm": 0.42039382457733154, "learning_rate": 1.2169340941128148e-05, "loss": 0.0194, "step": 78450 }, { "grad_norm": 0.20538151264190674, "learning_rate": 1.2158531586679267e-05, "loss": 0.0185, "step": 78460 }, { "grad_norm": 0.2794654369354248, "learning_rate": 1.2147726370516477e-05, "loss": 0.0192, "step": 78470 }, { "grad_norm": 0.36990001797676086, "learning_rate": 1.2136925293821422e-05, "loss": 0.0179, "step": 78480 }, { "grad_norm": 0.3753299415111542, "learning_rate": 1.2126128357775306e-05, "loss": 0.0162, "step": 78490 }, { "grad_norm": 0.21261703968048096, "learning_rate": 1.2115335563558838e-05, "loss": 0.0156, "step": 78500 }, { "grad_norm": 0.3352307379245758, "learning_rate": 1.2104546912352327e-05, "loss": 0.0184, "step": 78510 }, { "grad_norm": 0.4211600422859192, "learning_rate": 1.2093762405335557e-05, "loss": 0.014, "step": 78520 }, { "grad_norm": 0.35493165254592896, "learning_rate": 1.2082982043687974e-05, "loss": 0.0154, "step": 78530 }, { "grad_norm": 0.2821463644504547, "learning_rate": 1.207220582858845e-05, "loss": 0.0169, "step": 78540 }, { "grad_norm": 0.30767759680747986, "learning_rate": 1.206143376121549e-05, "loss": 0.0217, "step": 78550 }, { "grad_norm": 0.3283371329307556, "learning_rate": 1.2050665842747078e-05, "loss": 0.0147, "step": 78560 }, { "grad_norm": 0.3951955735683441, "learning_rate": 1.2039902074360804e-05, "loss": 0.0202, "step": 78570 }, { "grad_norm": 0.18820108473300934, "learning_rate": 1.2029142457233755e-05, "loss": 0.0121, "step": 78580 }, { "grad_norm": 0.3229849934577942, "learning_rate": 1.2018386992542601e-05, "loss": 0.0186, "step": 78590 }, { "grad_norm": 0.4201228618621826, "learning_rate": 1.2007635681463542e-05, "loss": 0.0174, "step": 78600 }, { "grad_norm": 0.3071897625923157, "learning_rate": 1.1996888525172334e-05, "loss": 0.0195, "step": 78610 }, { "grad_norm": 0.6821327805519104, "learning_rate": 1.198614552484425e-05, "loss": 0.0161, "step": 78620 }, { "grad_norm": 0.36951714754104614, "learning_rate": 1.197540668165415e-05, "loss": 0.0174, "step": 78630 }, { "grad_norm": 0.36974576115608215, "learning_rate": 1.1964671996776395e-05, "loss": 0.0226, "step": 78640 }, { "grad_norm": 0.5451748967170715, "learning_rate": 1.1953941471384922e-05, "loss": 0.0214, "step": 78650 }, { "grad_norm": 0.23244249820709229, "learning_rate": 1.1943215106653205e-05, "loss": 0.0198, "step": 78660 }, { "grad_norm": 0.2067454606294632, "learning_rate": 1.193249290375429e-05, "loss": 0.0216, "step": 78670 }, { "grad_norm": 0.34012967348098755, "learning_rate": 1.1921774863860696e-05, "loss": 0.0191, "step": 78680 }, { "grad_norm": 0.20643842220306396, "learning_rate": 1.1911060988144556e-05, "loss": 0.0124, "step": 78690 }, { "grad_norm": 0.25735220313072205, "learning_rate": 1.1900351277777533e-05, "loss": 0.0116, "step": 78700 }, { "grad_norm": 0.440153568983078, "learning_rate": 1.1889645733930798e-05, "loss": 0.0127, "step": 78710 }, { "grad_norm": 0.21708731353282928, "learning_rate": 1.1878944357775107e-05, "loss": 0.0129, "step": 78720 }, { "grad_norm": 0.12238679081201553, "learning_rate": 1.1868247150480749e-05, "loss": 0.0171, "step": 78730 }, { "grad_norm": 0.3014225363731384, "learning_rate": 1.1857554113217567e-05, "loss": 0.021, "step": 78740 }, { "grad_norm": 0.3083380162715912, "learning_rate": 1.1846865247154903e-05, "loss": 0.0181, "step": 78750 }, { "grad_norm": 0.2530538737773895, "learning_rate": 1.1836180553461711e-05, "loss": 0.0224, "step": 78760 }, { "grad_norm": 0.2744722366333008, "learning_rate": 1.1825500033306419e-05, "loss": 0.0236, "step": 78770 }, { "grad_norm": 0.31327515840530396, "learning_rate": 1.1814823687857052e-05, "loss": 0.0189, "step": 78780 }, { "grad_norm": 0.5657356977462769, "learning_rate": 1.1804151518281153e-05, "loss": 0.0147, "step": 78790 }, { "grad_norm": 0.3473825752735138, "learning_rate": 1.179348352574583e-05, "loss": 0.0271, "step": 78800 }, { "grad_norm": 0.5044610500335693, "learning_rate": 1.1782819711417697e-05, "loss": 0.0254, "step": 78810 }, { "grad_norm": 0.4086556136608124, "learning_rate": 1.1772160076462951e-05, "loss": 0.0162, "step": 78820 }, { "grad_norm": 0.4542236328125, "learning_rate": 1.1761504622047293e-05, "loss": 0.0168, "step": 78830 }, { "grad_norm": 0.3411640524864197, "learning_rate": 1.1750853349335999e-05, "loss": 0.0167, "step": 78840 }, { "grad_norm": 0.34316033124923706, "learning_rate": 1.1740206259493874e-05, "loss": 0.0173, "step": 78850 }, { "grad_norm": 0.2793810963630676, "learning_rate": 1.1729563353685286e-05, "loss": 0.026, "step": 78860 }, { "grad_norm": 0.437379390001297, "learning_rate": 1.1718924633074091e-05, "loss": 0.013, "step": 78870 }, { "grad_norm": 0.2570386230945587, "learning_rate": 1.1708290098823765e-05, "loss": 0.0217, "step": 78880 }, { "grad_norm": 0.20961938798427582, "learning_rate": 1.1697659752097245e-05, "loss": 0.0151, "step": 78890 }, { "grad_norm": 0.5610510110855103, "learning_rate": 1.1687033594057068e-05, "loss": 0.031, "step": 78900 }, { "grad_norm": 0.27258774638175964, "learning_rate": 1.1676411625865297e-05, "loss": 0.0166, "step": 78910 }, { "grad_norm": 0.2946457862854004, "learning_rate": 1.1665793848683542e-05, "loss": 0.017, "step": 78920 }, { "grad_norm": 0.6378100514411926, "learning_rate": 1.1655180263672927e-05, "loss": 0.0252, "step": 78930 }, { "grad_norm": 0.4102928638458252, "learning_rate": 1.1644570871994142e-05, "loss": 0.0154, "step": 78940 }, { "grad_norm": 0.2133931666612625, "learning_rate": 1.163396567480744e-05, "loss": 0.0166, "step": 78950 }, { "grad_norm": 0.4134303033351898, "learning_rate": 1.1623364673272552e-05, "loss": 0.0166, "step": 78960 }, { "grad_norm": 0.331253319978714, "learning_rate": 1.1612767868548807e-05, "loss": 0.0216, "step": 78970 }, { "grad_norm": 0.48123037815093994, "learning_rate": 1.1602175261795057e-05, "loss": 0.0222, "step": 78980 }, { "grad_norm": 0.33459439873695374, "learning_rate": 1.1591586854169696e-05, "loss": 0.018, "step": 78990 }, { "grad_norm": 0.27727338671684265, "learning_rate": 1.158100264683064e-05, "loss": 0.0128, "step": 79000 }, { "grad_norm": 0.37730562686920166, "learning_rate": 1.1570422640935386e-05, "loss": 0.0161, "step": 79010 }, { "grad_norm": 0.3101826608181, "learning_rate": 1.155984683764092e-05, "loss": 0.016, "step": 79020 }, { "grad_norm": 0.4073392152786255, "learning_rate": 1.1549275238103812e-05, "loss": 0.0165, "step": 79030 }, { "grad_norm": 0.34351876378059387, "learning_rate": 1.1538707843480152e-05, "loss": 0.0157, "step": 79040 }, { "grad_norm": 0.44783034920692444, "learning_rate": 1.1528144654925593e-05, "loss": 0.0239, "step": 79050 }, { "grad_norm": 0.19765986502170563, "learning_rate": 1.1517585673595277e-05, "loss": 0.0204, "step": 79060 }, { "grad_norm": 0.24565894901752472, "learning_rate": 1.150703090064395e-05, "loss": 0.0183, "step": 79070 }, { "grad_norm": 0.1175655797123909, "learning_rate": 1.1496480337225835e-05, "loss": 0.0183, "step": 79080 }, { "grad_norm": 0.4210403561592102, "learning_rate": 1.1485933984494751e-05, "loss": 0.0172, "step": 79090 }, { "grad_norm": 0.20783883333206177, "learning_rate": 1.1475391843604e-05, "loss": 0.0183, "step": 79100 }, { "grad_norm": 0.3576379418373108, "learning_rate": 1.1464853915706503e-05, "loss": 0.022, "step": 79110 }, { "grad_norm": 0.4220728874206543, "learning_rate": 1.1454320201954626e-05, "loss": 0.0118, "step": 79120 }, { "grad_norm": 0.619221568107605, "learning_rate": 1.1443790703500356e-05, "loss": 0.0181, "step": 79130 }, { "grad_norm": 0.4392869174480438, "learning_rate": 1.1433265421495154e-05, "loss": 0.0272, "step": 79140 }, { "grad_norm": 0.1520264744758606, "learning_rate": 1.1422744357090065e-05, "loss": 0.0281, "step": 79150 }, { "grad_norm": 0.31545206904411316, "learning_rate": 1.1412227511435647e-05, "loss": 0.0191, "step": 79160 }, { "grad_norm": 0.358510822057724, "learning_rate": 1.1401714885682025e-05, "loss": 0.0224, "step": 79170 }, { "grad_norm": 0.31869304180145264, "learning_rate": 1.1391206480978823e-05, "loss": 0.018, "step": 79180 }, { "grad_norm": 0.7923129796981812, "learning_rate": 1.1380702298475227e-05, "loss": 0.0215, "step": 79190 }, { "grad_norm": 0.4797472059726715, "learning_rate": 1.1370202339319985e-05, "loss": 0.0172, "step": 79200 }, { "grad_norm": 0.2195705771446228, "learning_rate": 1.1359706604661319e-05, "loss": 0.0249, "step": 79210 }, { "grad_norm": 0.5852038860321045, "learning_rate": 1.1349215095647042e-05, "loss": 0.0161, "step": 79220 }, { "grad_norm": 0.1688014715909958, "learning_rate": 1.1338727813424493e-05, "loss": 0.0215, "step": 79230 }, { "grad_norm": 0.4907076954841614, "learning_rate": 1.132824475914056e-05, "loss": 0.0211, "step": 79240 }, { "grad_norm": 0.24277140200138092, "learning_rate": 1.131776593394162e-05, "loss": 0.0166, "step": 79250 }, { "grad_norm": 0.33428165316581726, "learning_rate": 1.1307291338973657e-05, "loss": 0.0192, "step": 79260 }, { "grad_norm": 0.278432697057724, "learning_rate": 1.1296820975382121e-05, "loss": 0.0138, "step": 79270 }, { "grad_norm": 0.46295320987701416, "learning_rate": 1.1286354844312074e-05, "loss": 0.0179, "step": 79280 }, { "grad_norm": 1.460654854774475, "learning_rate": 1.1275892946908023e-05, "loss": 0.0253, "step": 79290 }, { "grad_norm": 0.26802971959114075, "learning_rate": 1.126543528431413e-05, "loss": 0.0193, "step": 79300 }, { "grad_norm": 0.4574563801288605, "learning_rate": 1.125498185767398e-05, "loss": 0.0147, "step": 79310 }, { "grad_norm": 0.16811825335025787, "learning_rate": 1.1244532668130781e-05, "loss": 0.0204, "step": 79320 }, { "grad_norm": 0.20444394648075104, "learning_rate": 1.1234087716827203e-05, "loss": 0.0242, "step": 79330 }, { "grad_norm": 0.5023881793022156, "learning_rate": 1.1223647004905524e-05, "loss": 0.0131, "step": 79340 }, { "grad_norm": 0.23122873902320862, "learning_rate": 1.1213210533507485e-05, "loss": 0.0311, "step": 79350 }, { "grad_norm": 0.2899000942707062, "learning_rate": 1.1202778303774447e-05, "loss": 0.0138, "step": 79360 }, { "grad_norm": 0.22284352779388428, "learning_rate": 1.1192350316847234e-05, "loss": 0.0154, "step": 79370 }, { "grad_norm": 0.44345682859420776, "learning_rate": 1.1181926573866258e-05, "loss": 0.0205, "step": 79380 }, { "grad_norm": 0.3083849549293518, "learning_rate": 1.1171507075971416e-05, "loss": 0.0194, "step": 79390 }, { "grad_norm": 0.21412238478660583, "learning_rate": 1.1161091824302177e-05, "loss": 0.0186, "step": 79400 }, { "grad_norm": 0.3063602149486542, "learning_rate": 1.115068081999755e-05, "loss": 0.0206, "step": 79410 }, { "grad_norm": 0.31419557332992554, "learning_rate": 1.1140274064196071e-05, "loss": 0.0246, "step": 79420 }, { "grad_norm": 0.256477415561676, "learning_rate": 1.112987155803578e-05, "loss": 0.0164, "step": 79430 }, { "grad_norm": 0.6507827639579773, "learning_rate": 1.1119473302654298e-05, "loss": 0.0205, "step": 79440 }, { "grad_norm": 0.3213677406311035, "learning_rate": 1.1109079299188769e-05, "loss": 0.0187, "step": 79450 }, { "grad_norm": 0.5671887397766113, "learning_rate": 1.1098689548775847e-05, "loss": 0.0208, "step": 79460 }, { "grad_norm": 0.32786059379577637, "learning_rate": 1.1088304052551762e-05, "loss": 0.0262, "step": 79470 }, { "grad_norm": 0.5699599981307983, "learning_rate": 1.1077922811652215e-05, "loss": 0.0255, "step": 79480 }, { "grad_norm": 0.4441964626312256, "learning_rate": 1.1067545827212534e-05, "loss": 0.0188, "step": 79490 }, { "grad_norm": 0.25697270035743713, "learning_rate": 1.1057173100367496e-05, "loss": 0.0199, "step": 79500 }, { "grad_norm": 0.2707047462463379, "learning_rate": 1.1046804632251474e-05, "loss": 0.0139, "step": 79510 }, { "grad_norm": 0.5807638764381409, "learning_rate": 1.1036440423998318e-05, "loss": 0.0256, "step": 79520 }, { "grad_norm": 0.2950270473957062, "learning_rate": 1.1026080476741469e-05, "loss": 0.0191, "step": 79530 }, { "grad_norm": 0.23796464502811432, "learning_rate": 1.1015724791613836e-05, "loss": 0.0147, "step": 79540 }, { "grad_norm": 0.3951849937438965, "learning_rate": 1.1005373369747951e-05, "loss": 0.0168, "step": 79550 }, { "grad_norm": 0.34399914741516113, "learning_rate": 1.09950262122758e-05, "loss": 0.0203, "step": 79560 }, { "grad_norm": 0.8424436450004578, "learning_rate": 1.0984683320328948e-05, "loss": 0.0182, "step": 79570 }, { "grad_norm": 0.1845405250787735, "learning_rate": 1.0974344695038458e-05, "loss": 0.02, "step": 79580 }, { "grad_norm": 0.41414111852645874, "learning_rate": 1.0964010337534968e-05, "loss": 0.0198, "step": 79590 }, { "grad_norm": 0.2493552714586258, "learning_rate": 1.0953680248948611e-05, "loss": 0.0186, "step": 79600 }, { "grad_norm": 0.21651187539100647, "learning_rate": 1.0943354430409069e-05, "loss": 0.0181, "step": 79610 }, { "grad_norm": 0.15799330174922943, "learning_rate": 1.0933032883045574e-05, "loss": 0.0204, "step": 79620 }, { "grad_norm": 0.7014970183372498, "learning_rate": 1.0922715607986872e-05, "loss": 0.0201, "step": 79630 }, { "grad_norm": 0.45070680975914, "learning_rate": 1.0912402606361234e-05, "loss": 0.0182, "step": 79640 }, { "grad_norm": 0.35757797956466675, "learning_rate": 1.090209387929647e-05, "loss": 0.0157, "step": 79650 }, { "grad_norm": 0.3877074122428894, "learning_rate": 1.0891789427919957e-05, "loss": 0.0179, "step": 79660 }, { "grad_norm": 0.24683493375778198, "learning_rate": 1.088148925335854e-05, "loss": 0.0163, "step": 79670 }, { "grad_norm": 0.4437451660633087, "learning_rate": 1.087119335673864e-05, "loss": 0.0231, "step": 79680 }, { "grad_norm": 0.35722967982292175, "learning_rate": 1.0860901739186208e-05, "loss": 0.0156, "step": 79690 }, { "grad_norm": 0.37715718150138855, "learning_rate": 1.085061440182673e-05, "loss": 0.0124, "step": 79700 }, { "grad_norm": 0.3404620885848999, "learning_rate": 1.0840331345785193e-05, "loss": 0.0206, "step": 79710 }, { "grad_norm": 0.38459891080856323, "learning_rate": 1.083005257218615e-05, "loss": 0.0211, "step": 79720 }, { "grad_norm": 0.36531686782836914, "learning_rate": 1.0819778082153642e-05, "loss": 0.0164, "step": 79730 }, { "grad_norm": 0.6205244064331055, "learning_rate": 1.0809507876811325e-05, "loss": 0.0212, "step": 79740 }, { "grad_norm": 0.19947658479213715, "learning_rate": 1.0799241957282292e-05, "loss": 0.0167, "step": 79750 }, { "grad_norm": 0.4071483910083771, "learning_rate": 1.078898032468923e-05, "loss": 0.0276, "step": 79760 }, { "grad_norm": 0.2531960904598236, "learning_rate": 1.0778722980154315e-05, "loss": 0.0186, "step": 79770 }, { "grad_norm": 0.36425817012786865, "learning_rate": 1.0768469924799302e-05, "loss": 0.0205, "step": 79780 }, { "grad_norm": 0.6406261920928955, "learning_rate": 1.0758221159745419e-05, "loss": 0.0221, "step": 79790 }, { "grad_norm": 0.3805919885635376, "learning_rate": 1.074797668611347e-05, "loss": 0.0231, "step": 79800 }, { "grad_norm": 0.19356770813465118, "learning_rate": 1.0737736505023777e-05, "loss": 0.0177, "step": 79810 }, { "grad_norm": 0.18588358163833618, "learning_rate": 1.0727500617596203e-05, "loss": 0.0151, "step": 79820 }, { "grad_norm": 0.33808639645576477, "learning_rate": 1.0717269024950104e-05, "loss": 0.0238, "step": 79830 }, { "grad_norm": 0.5829512476921082, "learning_rate": 1.070704172820442e-05, "loss": 0.0213, "step": 79840 }, { "grad_norm": 0.47092992067337036, "learning_rate": 1.0696818728477559e-05, "loss": 0.0189, "step": 79850 }, { "grad_norm": 0.22806495428085327, "learning_rate": 1.0686600026887511e-05, "loss": 0.0174, "step": 79860 }, { "grad_norm": 0.3501642048358917, "learning_rate": 1.0676385624551777e-05, "loss": 0.0167, "step": 79870 }, { "grad_norm": 0.3614404797554016, "learning_rate": 1.0666175522587402e-05, "loss": 0.0126, "step": 79880 }, { "grad_norm": 0.6261870265007019, "learning_rate": 1.0655969722110925e-05, "loss": 0.0207, "step": 79890 }, { "grad_norm": 0.4809068441390991, "learning_rate": 1.0645768224238439e-05, "loss": 0.021, "step": 79900 }, { "grad_norm": 0.23280231654644012, "learning_rate": 1.063557103008559e-05, "loss": 0.0212, "step": 79910 }, { "grad_norm": 0.21361981332302094, "learning_rate": 1.0625378140767494e-05, "loss": 0.0134, "step": 79920 }, { "grad_norm": 0.22704054415225983, "learning_rate": 1.0615189557398846e-05, "loss": 0.0155, "step": 79930 }, { "grad_norm": 0.4219742715358734, "learning_rate": 1.060500528109386e-05, "loss": 0.0168, "step": 79940 }, { "grad_norm": 0.30501917004585266, "learning_rate": 1.0594825312966272e-05, "loss": 0.016, "step": 79950 }, { "grad_norm": 0.32090243697166443, "learning_rate": 1.0584649654129336e-05, "loss": 0.0191, "step": 79960 }, { "grad_norm": 0.3389171361923218, "learning_rate": 1.0574478305695863e-05, "loss": 0.0195, "step": 79970 }, { "grad_norm": 0.30112069845199585, "learning_rate": 1.0564311268778154e-05, "loss": 0.0195, "step": 79980 }, { "grad_norm": 0.28071388602256775, "learning_rate": 1.0554148544488073e-05, "loss": 0.016, "step": 79990 }, { "grad_norm": 0.45518919825553894, "learning_rate": 1.0543990133937004e-05, "loss": 0.0114, "step": 80000 }, { "grad_norm": 0.3417305052280426, "learning_rate": 1.0533836038235868e-05, "loss": 0.0173, "step": 80010 }, { "grad_norm": 0.2561146020889282, "learning_rate": 1.052368625849507e-05, "loss": 0.0178, "step": 80020 }, { "grad_norm": 0.3586696684360504, "learning_rate": 1.05135407958246e-05, "loss": 0.0159, "step": 80030 }, { "grad_norm": 0.17827555537223816, "learning_rate": 1.0503399651333934e-05, "loss": 0.0167, "step": 80040 }, { "grad_norm": 0.3710705637931824, "learning_rate": 1.04932628261321e-05, "loss": 0.0193, "step": 80050 }, { "grad_norm": 0.2785443961620331, "learning_rate": 1.0483130321327644e-05, "loss": 0.0191, "step": 80060 }, { "grad_norm": 0.27590709924697876, "learning_rate": 1.0473002138028654e-05, "loss": 0.017, "step": 80070 }, { "grad_norm": 0.4464155435562134, "learning_rate": 1.0462878277342713e-05, "loss": 0.0144, "step": 80080 }, { "grad_norm": 0.21715833246707916, "learning_rate": 1.0452758740376972e-05, "loss": 0.0188, "step": 80090 }, { "grad_norm": 0.304135799407959, "learning_rate": 1.0442643528238067e-05, "loss": 0.0175, "step": 80100 }, { "grad_norm": 0.20921649038791656, "learning_rate": 1.0432532642032188e-05, "loss": 0.018, "step": 80110 }, { "grad_norm": 0.2691875100135803, "learning_rate": 1.042242608286505e-05, "loss": 0.0205, "step": 80120 }, { "grad_norm": 0.48575538396835327, "learning_rate": 1.041232385184191e-05, "loss": 0.0158, "step": 80130 }, { "grad_norm": 0.40156319737434387, "learning_rate": 1.0402225950067495e-05, "loss": 0.0131, "step": 80140 }, { "grad_norm": 0.2586619257926941, "learning_rate": 1.0392132378646119e-05, "loss": 0.018, "step": 80150 }, { "grad_norm": 0.4962122440338135, "learning_rate": 1.0382043138681607e-05, "loss": 0.0239, "step": 80160 }, { "grad_norm": 0.45063161849975586, "learning_rate": 1.0371958231277284e-05, "loss": 0.0181, "step": 80170 }, { "grad_norm": 0.14517173171043396, "learning_rate": 1.0361877657536024e-05, "loss": 0.0118, "step": 80180 }, { "grad_norm": 0.21299812197685242, "learning_rate": 1.035180141856023e-05, "loss": 0.0222, "step": 80190 }, { "grad_norm": 0.2435433268547058, "learning_rate": 1.0341729515451836e-05, "loss": 0.015, "step": 80200 }, { "grad_norm": 0.31916728615760803, "learning_rate": 1.0331661949312265e-05, "loss": 0.0231, "step": 80210 }, { "grad_norm": 0.20077162981033325, "learning_rate": 1.0321598721242514e-05, "loss": 0.0265, "step": 80220 }, { "grad_norm": 0.40198859572410583, "learning_rate": 1.0311539832343064e-05, "loss": 0.0144, "step": 80230 }, { "grad_norm": 0.7461147308349609, "learning_rate": 1.0301485283713942e-05, "loss": 0.0147, "step": 80240 }, { "grad_norm": 0.3025955855846405, "learning_rate": 1.0291435076454703e-05, "loss": 0.0147, "step": 80250 }, { "grad_norm": 0.2783258557319641, "learning_rate": 1.028138921166444e-05, "loss": 0.0185, "step": 80260 }, { "grad_norm": 0.39223793148994446, "learning_rate": 1.0271347690441719e-05, "loss": 0.0135, "step": 80270 }, { "grad_norm": 0.5073902606964111, "learning_rate": 1.0261310513884698e-05, "loss": 0.0186, "step": 80280 }, { "grad_norm": 0.21888592839241028, "learning_rate": 1.0251277683090998e-05, "loss": 0.0142, "step": 80290 }, { "grad_norm": 0.18632429838180542, "learning_rate": 1.0241249199157826e-05, "loss": 0.0175, "step": 80300 }, { "grad_norm": 0.28844013810157776, "learning_rate": 1.0231225063181837e-05, "loss": 0.015, "step": 80310 }, { "grad_norm": 1.0159176588058472, "learning_rate": 1.0221205276259316e-05, "loss": 0.0221, "step": 80320 }, { "grad_norm": 0.3875192403793335, "learning_rate": 1.0211189839485958e-05, "loss": 0.0229, "step": 80330 }, { "grad_norm": 0.2829366624355316, "learning_rate": 1.0201178753957075e-05, "loss": 0.0116, "step": 80340 }, { "grad_norm": 0.26789286732673645, "learning_rate": 1.0191172020767432e-05, "loss": 0.0148, "step": 80350 }, { "grad_norm": 0.2233956903219223, "learning_rate": 1.0181169641011362e-05, "loss": 0.0174, "step": 80360 }, { "grad_norm": 0.44761529564857483, "learning_rate": 1.0171171615782721e-05, "loss": 0.017, "step": 80370 }, { "grad_norm": 0.2962493598461151, "learning_rate": 1.0161177946174866e-05, "loss": 0.0171, "step": 80380 }, { "grad_norm": 0.2189234048128128, "learning_rate": 1.0151188633280706e-05, "loss": 0.0154, "step": 80390 }, { "grad_norm": 0.43402111530303955, "learning_rate": 1.0141203678192635e-05, "loss": 0.0188, "step": 80400 }, { "grad_norm": 0.3438988924026489, "learning_rate": 1.0131223082002612e-05, "loss": 0.0253, "step": 80410 }, { "grad_norm": 0.2707039415836334, "learning_rate": 1.012124684580208e-05, "loss": 0.0167, "step": 80420 }, { "grad_norm": 0.5263838768005371, "learning_rate": 1.0111274970682039e-05, "loss": 0.016, "step": 80430 }, { "grad_norm": 0.15265028178691864, "learning_rate": 1.010130745773299e-05, "loss": 0.0184, "step": 80440 }, { "grad_norm": 0.24767430126667023, "learning_rate": 1.0091344308044986e-05, "loss": 0.0133, "step": 80450 }, { "grad_norm": 0.5303127765655518, "learning_rate": 1.0081385522707554e-05, "loss": 0.0191, "step": 80460 }, { "grad_norm": 0.25914883613586426, "learning_rate": 1.0071431102809803e-05, "loss": 0.0175, "step": 80470 }, { "grad_norm": 0.6608696579933167, "learning_rate": 1.0061481049440297e-05, "loss": 0.0158, "step": 80480 }, { "grad_norm": 0.34888896346092224, "learning_rate": 1.0051535363687187e-05, "loss": 0.0177, "step": 80490 }, { "grad_norm": 0.2615068554878235, "learning_rate": 1.0041594046638087e-05, "loss": 0.0159, "step": 80500 }, { "grad_norm": 0.27799850702285767, "learning_rate": 1.0031657099380204e-05, "loss": 0.0173, "step": 80510 }, { "grad_norm": 0.26284852623939514, "learning_rate": 1.0021724523000204e-05, "loss": 0.0162, "step": 80520 }, { "grad_norm": 0.41954323649406433, "learning_rate": 1.001179631858431e-05, "loss": 0.0198, "step": 80530 }, { "grad_norm": 0.19717656075954437, "learning_rate": 1.0001872487218245e-05, "loss": 0.0168, "step": 80540 }, { "grad_norm": 0.36201223731040955, "learning_rate": 9.991953029987272e-06, "loss": 0.0247, "step": 80550 }, { "grad_norm": 0.37787044048309326, "learning_rate": 9.982037947976147e-06, "loss": 0.0188, "step": 80560 }, { "grad_norm": 0.30592140555381775, "learning_rate": 9.972127242269208e-06, "loss": 0.0204, "step": 80570 }, { "grad_norm": 0.4690265357494354, "learning_rate": 9.962220913950238e-06, "loss": 0.0197, "step": 80580 }, { "grad_norm": 0.374515563249588, "learning_rate": 9.952318964102591e-06, "loss": 0.03, "step": 80590 }, { "grad_norm": 0.13206902146339417, "learning_rate": 9.942421393809148e-06, "loss": 0.0135, "step": 80600 }, { "grad_norm": 0.6722331643104553, "learning_rate": 9.932528204152264e-06, "loss": 0.0146, "step": 80610 }, { "grad_norm": 0.7698978185653687, "learning_rate": 9.922639396213856e-06, "loss": 0.0293, "step": 80620 }, { "grad_norm": 0.3579818308353424, "learning_rate": 9.912754971075344e-06, "loss": 0.0161, "step": 80630 }, { "grad_norm": 0.16716329753398895, "learning_rate": 9.902874929817696e-06, "loss": 0.0139, "step": 80640 }, { "grad_norm": 0.27762144804000854, "learning_rate": 9.892999273521342e-06, "loss": 0.014, "step": 80650 }, { "grad_norm": 0.32979321479797363, "learning_rate": 9.883128003266302e-06, "loss": 0.0204, "step": 80660 }, { "grad_norm": 0.16839733719825745, "learning_rate": 9.87326112013206e-06, "loss": 0.022, "step": 80670 }, { "grad_norm": 0.5343306064605713, "learning_rate": 9.863398625197662e-06, "loss": 0.0126, "step": 80680 }, { "grad_norm": 0.44915977120399475, "learning_rate": 9.853540519541615e-06, "loss": 0.0207, "step": 80690 }, { "grad_norm": 0.42228278517723083, "learning_rate": 9.843686804242053e-06, "loss": 0.0188, "step": 80700 }, { "grad_norm": 0.26570871472358704, "learning_rate": 9.833837480376506e-06, "loss": 0.0251, "step": 80710 }, { "grad_norm": 0.5330974459648132, "learning_rate": 9.82399254902212e-06, "loss": 0.0216, "step": 80720 }, { "grad_norm": 0.4904211759567261, "learning_rate": 9.814152011255494e-06, "loss": 0.0135, "step": 80730 }, { "grad_norm": 0.47815531492233276, "learning_rate": 9.804315868152797e-06, "loss": 0.0225, "step": 80740 }, { "grad_norm": 0.3264099657535553, "learning_rate": 9.79448412078966e-06, "loss": 0.0182, "step": 80750 }, { "grad_norm": 0.3135107159614563, "learning_rate": 9.784656770241312e-06, "loss": 0.0183, "step": 80760 }, { "grad_norm": 0.6116104125976562, "learning_rate": 9.77483381758243e-06, "loss": 0.0148, "step": 80770 }, { "grad_norm": 0.3135100305080414, "learning_rate": 9.76501526388725e-06, "loss": 0.0163, "step": 80780 }, { "grad_norm": 0.4794136881828308, "learning_rate": 9.7552011102295e-06, "loss": 0.0394, "step": 80790 }, { "grad_norm": 0.1602679342031479, "learning_rate": 9.745391357682449e-06, "loss": 0.0218, "step": 80800 }, { "grad_norm": 0.38312163949012756, "learning_rate": 9.735586007318875e-06, "loss": 0.0256, "step": 80810 }, { "grad_norm": 2.4626922607421875, "learning_rate": 9.725785060211096e-06, "loss": 0.0165, "step": 80820 }, { "grad_norm": 0.29218822717666626, "learning_rate": 9.715988517430896e-06, "loss": 0.0139, "step": 80830 }, { "grad_norm": 0.5045432448387146, "learning_rate": 9.706196380049625e-06, "loss": 0.0131, "step": 80840 }, { "grad_norm": 0.4230516254901886, "learning_rate": 9.69640864913815e-06, "loss": 0.0137, "step": 80850 }, { "grad_norm": 0.32897645235061646, "learning_rate": 9.686625325766818e-06, "loss": 0.0155, "step": 80860 }, { "grad_norm": 0.16544552147388458, "learning_rate": 9.67684641100554e-06, "loss": 0.0214, "step": 80870 }, { "grad_norm": 0.256320983171463, "learning_rate": 9.667071905923692e-06, "loss": 0.0147, "step": 80880 }, { "grad_norm": 0.39818066358566284, "learning_rate": 9.657301811590242e-06, "loss": 0.0171, "step": 80890 }, { "grad_norm": 0.41207239031791687, "learning_rate": 9.647536129073603e-06, "loss": 0.0175, "step": 80900 }, { "grad_norm": 0.1751694679260254, "learning_rate": 9.637774859441751e-06, "loss": 0.0159, "step": 80910 }, { "grad_norm": 0.23192089796066284, "learning_rate": 9.628018003762152e-06, "loss": 0.0171, "step": 80920 }, { "grad_norm": 0.730080783367157, "learning_rate": 9.618265563101813e-06, "loss": 0.0129, "step": 80930 }, { "grad_norm": 0.24984721839427948, "learning_rate": 9.608517538527211e-06, "loss": 0.018, "step": 80940 }, { "grad_norm": 0.21619912981987, "learning_rate": 9.598773931104431e-06, "loss": 0.0204, "step": 80950 }, { "grad_norm": 0.41227132081985474, "learning_rate": 9.589034741898983e-06, "loss": 0.0178, "step": 80960 }, { "grad_norm": 0.36028480529785156, "learning_rate": 9.579299971975947e-06, "loss": 0.0221, "step": 80970 }, { "grad_norm": 0.27817943692207336, "learning_rate": 9.569569622399882e-06, "loss": 0.0158, "step": 80980 }, { "grad_norm": 0.3891325891017914, "learning_rate": 9.559843694234916e-06, "loss": 0.0149, "step": 80990 }, { "grad_norm": 0.18519936501979828, "learning_rate": 9.550122188544625e-06, "loss": 0.013, "step": 81000 }, { "grad_norm": 0.3182312548160553, "learning_rate": 9.540405106392152e-06, "loss": 0.0163, "step": 81010 }, { "grad_norm": 0.33034461736679077, "learning_rate": 9.53069244884015e-06, "loss": 0.0212, "step": 81020 }, { "grad_norm": 0.22711288928985596, "learning_rate": 9.52098421695079e-06, "loss": 0.0199, "step": 81030 }, { "grad_norm": 0.39743801951408386, "learning_rate": 9.511280411785728e-06, "loss": 0.0201, "step": 81040 }, { "grad_norm": 0.23568002879619598, "learning_rate": 9.50158103440616e-06, "loss": 0.0188, "step": 81050 }, { "grad_norm": 0.15339063107967377, "learning_rate": 9.491886085872815e-06, "loss": 0.0193, "step": 81060 }, { "grad_norm": 0.4287893772125244, "learning_rate": 9.482195567245895e-06, "loss": 0.0148, "step": 81070 }, { "grad_norm": 0.2922060489654541, "learning_rate": 9.472509479585151e-06, "loss": 0.0163, "step": 81080 }, { "grad_norm": 0.17193831503391266, "learning_rate": 9.462827823949833e-06, "loss": 0.0162, "step": 81090 }, { "grad_norm": 0.8331838250160217, "learning_rate": 9.453150601398735e-06, "loss": 0.012, "step": 81100 }, { "grad_norm": 0.30659207701683044, "learning_rate": 9.44347781299011e-06, "loss": 0.0129, "step": 81110 }, { "grad_norm": 0.32454949617385864, "learning_rate": 9.43380945978179e-06, "loss": 0.0198, "step": 81120 }, { "grad_norm": 0.12922674417495728, "learning_rate": 9.424145542831047e-06, "loss": 0.0165, "step": 81130 }, { "grad_norm": 0.17769800126552582, "learning_rate": 9.414486063194778e-06, "loss": 0.015, "step": 81140 }, { "grad_norm": 0.35331279039382935, "learning_rate": 9.404831021929273e-06, "loss": 0.0112, "step": 81150 }, { "grad_norm": 0.21463565528392792, "learning_rate": 9.395180420090427e-06, "loss": 0.0115, "step": 81160 }, { "grad_norm": 1.1117510795593262, "learning_rate": 9.385534258733591e-06, "loss": 0.0153, "step": 81170 }, { "grad_norm": 0.7006537914276123, "learning_rate": 9.375892538913677e-06, "loss": 0.0175, "step": 81180 }, { "grad_norm": 0.22386786341667175, "learning_rate": 9.36625526168507e-06, "loss": 0.0191, "step": 81190 }, { "grad_norm": 0.20481382310390472, "learning_rate": 9.356622428101686e-06, "loss": 0.0102, "step": 81200 }, { "grad_norm": 0.17049135267734528, "learning_rate": 9.346994039216972e-06, "loss": 0.0188, "step": 81210 }, { "grad_norm": 1.3857053518295288, "learning_rate": 9.337370096083886e-06, "loss": 0.0221, "step": 81220 }, { "grad_norm": 0.18403738737106323, "learning_rate": 9.327750599754854e-06, "loss": 0.0151, "step": 81230 }, { "grad_norm": 0.4858818054199219, "learning_rate": 9.318135551281875e-06, "loss": 0.0179, "step": 81240 }, { "grad_norm": 0.12131151556968689, "learning_rate": 9.308524951716424e-06, "loss": 0.0172, "step": 81250 }, { "grad_norm": 0.15620262920856476, "learning_rate": 9.298918802109497e-06, "loss": 0.015, "step": 81260 }, { "grad_norm": 0.5777589082717896, "learning_rate": 9.289317103511625e-06, "loss": 0.011, "step": 81270 }, { "grad_norm": 0.3637124300003052, "learning_rate": 9.279719856972835e-06, "loss": 0.019, "step": 81280 }, { "grad_norm": 0.4277997612953186, "learning_rate": 9.270127063542645e-06, "loss": 0.0154, "step": 81290 }, { "grad_norm": 0.3439311683177948, "learning_rate": 9.260538724270124e-06, "loss": 0.0223, "step": 81300 }, { "grad_norm": 0.4358912706375122, "learning_rate": 9.250954840203852e-06, "loss": 0.0186, "step": 81310 }, { "grad_norm": 1.2039505243301392, "learning_rate": 9.241375412391878e-06, "loss": 0.0231, "step": 81320 }, { "grad_norm": 0.2688659429550171, "learning_rate": 9.23180044188181e-06, "loss": 0.0237, "step": 81330 }, { "grad_norm": 0.30632954835891724, "learning_rate": 9.222229929720755e-06, "loss": 0.0171, "step": 81340 }, { "grad_norm": 0.39118197560310364, "learning_rate": 9.212663876955336e-06, "loss": 0.022, "step": 81350 }, { "grad_norm": 0.4409533739089966, "learning_rate": 9.20310228463166e-06, "loss": 0.0161, "step": 81360 }, { "grad_norm": 0.40159741044044495, "learning_rate": 9.19354515379539e-06, "loss": 0.0218, "step": 81370 }, { "grad_norm": 0.4669448137283325, "learning_rate": 9.183992485491666e-06, "loss": 0.0156, "step": 81380 }, { "grad_norm": 0.1499747633934021, "learning_rate": 9.174444280765154e-06, "loss": 0.0182, "step": 81390 }, { "grad_norm": 0.39244315028190613, "learning_rate": 9.164900540660032e-06, "loss": 0.0145, "step": 81400 }, { "grad_norm": 0.27978581190109253, "learning_rate": 9.155361266220008e-06, "loss": 0.0158, "step": 81410 }, { "grad_norm": 0.433290034532547, "learning_rate": 9.145826458488254e-06, "loss": 0.0202, "step": 81420 }, { "grad_norm": 0.292113721370697, "learning_rate": 9.136296118507509e-06, "loss": 0.0188, "step": 81430 }, { "grad_norm": 0.8467265963554382, "learning_rate": 9.126770247319966e-06, "loss": 0.0169, "step": 81440 }, { "grad_norm": 0.2856106758117676, "learning_rate": 9.11724884596738e-06, "loss": 0.0178, "step": 81450 }, { "grad_norm": 0.1539212316274643, "learning_rate": 9.107731915490996e-06, "loss": 0.0208, "step": 81460 }, { "grad_norm": 0.37769368290901184, "learning_rate": 9.098219456931579e-06, "loss": 0.0197, "step": 81470 }, { "grad_norm": 0.49999818205833435, "learning_rate": 9.088711471329376e-06, "loss": 0.0227, "step": 81480 }, { "grad_norm": 0.19763004779815674, "learning_rate": 9.079207959724184e-06, "loss": 0.0123, "step": 81490 }, { "grad_norm": 0.571714460849762, "learning_rate": 9.069708923155279e-06, "loss": 0.019, "step": 81500 }, { "grad_norm": 0.1098313257098198, "learning_rate": 9.060214362661468e-06, "loss": 0.0184, "step": 81510 }, { "grad_norm": 0.42245134711265564, "learning_rate": 9.050724279281064e-06, "loss": 0.0206, "step": 81520 }, { "grad_norm": 0.27087318897247314, "learning_rate": 9.041238674051894e-06, "loss": 0.0137, "step": 81530 }, { "grad_norm": 0.4190484881401062, "learning_rate": 9.031757548011277e-06, "loss": 0.019, "step": 81540 }, { "grad_norm": 0.2238367795944214, "learning_rate": 9.022280902196051e-06, "loss": 0.0181, "step": 81550 }, { "grad_norm": 0.3279605805873871, "learning_rate": 9.012808737642592e-06, "loss": 0.0161, "step": 81560 }, { "grad_norm": 0.2610987424850464, "learning_rate": 9.003341055386739e-06, "loss": 0.0187, "step": 81570 }, { "grad_norm": 0.37344861030578613, "learning_rate": 8.993877856463862e-06, "loss": 0.0171, "step": 81580 }, { "grad_norm": 0.5454393029212952, "learning_rate": 8.984419141908856e-06, "loss": 0.017, "step": 81590 }, { "grad_norm": 0.15695765614509583, "learning_rate": 8.974964912756117e-06, "loss": 0.015, "step": 81600 }, { "grad_norm": 0.28862839937210083, "learning_rate": 8.965515170039523e-06, "loss": 0.0143, "step": 81610 }, { "grad_norm": 0.721592903137207, "learning_rate": 8.956069914792514e-06, "loss": 0.0188, "step": 81620 }, { "grad_norm": 0.25426626205444336, "learning_rate": 8.946629148047969e-06, "loss": 0.0187, "step": 81630 }, { "grad_norm": 0.26959022879600525, "learning_rate": 8.93719287083834e-06, "loss": 0.03, "step": 81640 }, { "grad_norm": 0.46962565183639526, "learning_rate": 8.927761084195569e-06, "loss": 0.0156, "step": 81650 }, { "grad_norm": 0.27905499935150146, "learning_rate": 8.918333789151096e-06, "loss": 0.0154, "step": 81660 }, { "grad_norm": 0.31835296750068665, "learning_rate": 8.90891098673587e-06, "loss": 0.0138, "step": 81670 }, { "grad_norm": 0.2026922106742859, "learning_rate": 8.89949267798037e-06, "loss": 0.0105, "step": 81680 }, { "grad_norm": 0.40375784039497375, "learning_rate": 8.890078863914542e-06, "loss": 0.0227, "step": 81690 }, { "grad_norm": 0.22421644628047943, "learning_rate": 8.880669545567894e-06, "loss": 0.0147, "step": 81700 }, { "grad_norm": 0.18890926241874695, "learning_rate": 8.871264723969381e-06, "loss": 0.013, "step": 81710 }, { "grad_norm": 0.7503423094749451, "learning_rate": 8.86186440014754e-06, "loss": 0.0154, "step": 81720 }, { "grad_norm": 1.1755940914154053, "learning_rate": 8.852468575130351e-06, "loss": 0.0267, "step": 81730 }, { "grad_norm": 0.21777978539466858, "learning_rate": 8.843077249945347e-06, "loss": 0.0206, "step": 81740 }, { "grad_norm": 0.5775584578514099, "learning_rate": 8.833690425619523e-06, "loss": 0.0148, "step": 81750 }, { "grad_norm": 0.3710049092769623, "learning_rate": 8.82430810317942e-06, "loss": 0.0215, "step": 81760 }, { "grad_norm": 0.3612973392009735, "learning_rate": 8.814930283651074e-06, "loss": 0.0157, "step": 81770 }, { "grad_norm": 0.45244842767715454, "learning_rate": 8.805556968060048e-06, "loss": 0.0166, "step": 81780 }, { "grad_norm": 0.38065776228904724, "learning_rate": 8.796188157431362e-06, "loss": 0.0179, "step": 81790 }, { "grad_norm": 0.46801286935806274, "learning_rate": 8.786823852789595e-06, "loss": 0.0157, "step": 81800 }, { "grad_norm": 0.4956039488315582, "learning_rate": 8.777464055158813e-06, "loss": 0.0184, "step": 81810 }, { "grad_norm": 1.477239727973938, "learning_rate": 8.768108765562578e-06, "loss": 0.014, "step": 81820 }, { "grad_norm": 0.12598827481269836, "learning_rate": 8.758757985023975e-06, "loss": 0.0143, "step": 81830 }, { "grad_norm": 0.3375074565410614, "learning_rate": 8.749411714565598e-06, "loss": 0.0136, "step": 81840 }, { "grad_norm": 0.24700964987277985, "learning_rate": 8.74006995520954e-06, "loss": 0.0279, "step": 81850 }, { "grad_norm": 0.5524869561195374, "learning_rate": 8.730732707977397e-06, "loss": 0.0183, "step": 81860 }, { "grad_norm": 0.5298004150390625, "learning_rate": 8.721399973890281e-06, "loss": 0.0148, "step": 81870 }, { "grad_norm": 0.1582614928483963, "learning_rate": 8.712071753968792e-06, "loss": 0.018, "step": 81880 }, { "grad_norm": 0.2592598795890808, "learning_rate": 8.702748049233072e-06, "loss": 0.0175, "step": 81890 }, { "grad_norm": 0.18769003450870514, "learning_rate": 8.693428860702718e-06, "loss": 0.0135, "step": 81900 }, { "grad_norm": 0.3314589560031891, "learning_rate": 8.684114189396898e-06, "loss": 0.0212, "step": 81910 }, { "grad_norm": 0.19236695766448975, "learning_rate": 8.674804036334222e-06, "loss": 0.0143, "step": 81920 }, { "grad_norm": 0.3025909960269928, "learning_rate": 8.665498402532857e-06, "loss": 0.0157, "step": 81930 }, { "grad_norm": 0.1657467782497406, "learning_rate": 8.656197289010432e-06, "loss": 0.0141, "step": 81940 }, { "grad_norm": 0.2130810171365738, "learning_rate": 8.646900696784122e-06, "loss": 0.0134, "step": 81950 }, { "grad_norm": 0.7374669313430786, "learning_rate": 8.637608626870553e-06, "loss": 0.0175, "step": 81960 }, { "grad_norm": 0.18160606920719147, "learning_rate": 8.628321080285944e-06, "loss": 0.0187, "step": 81970 }, { "grad_norm": 0.3498727083206177, "learning_rate": 8.619038058045925e-06, "loss": 0.0166, "step": 81980 }, { "grad_norm": 0.441884309053421, "learning_rate": 8.609759561165704e-06, "loss": 0.0169, "step": 81990 }, { "grad_norm": 0.20262302458286285, "learning_rate": 8.600485590659934e-06, "loss": 0.0162, "step": 82000 }, { "grad_norm": 0.4156317710876465, "learning_rate": 8.591216147542813e-06, "loss": 0.019, "step": 82010 }, { "grad_norm": 0.31454402208328247, "learning_rate": 8.581951232828045e-06, "loss": 0.026, "step": 82020 }, { "grad_norm": 0.35292279720306396, "learning_rate": 8.572690847528825e-06, "loss": 0.018, "step": 82030 }, { "grad_norm": 0.5901070237159729, "learning_rate": 8.563434992657838e-06, "loss": 0.0138, "step": 82040 }, { "grad_norm": 0.17860199511051178, "learning_rate": 8.554183669227307e-06, "loss": 0.0192, "step": 82050 }, { "grad_norm": 0.328905314207077, "learning_rate": 8.544936878248937e-06, "loss": 0.0134, "step": 82060 }, { "grad_norm": 0.705889880657196, "learning_rate": 8.535694620733942e-06, "loss": 0.0189, "step": 82070 }, { "grad_norm": 0.22747528553009033, "learning_rate": 8.526456897693053e-06, "loss": 0.0145, "step": 82080 }, { "grad_norm": 0.5063278079032898, "learning_rate": 8.517223710136456e-06, "loss": 0.0203, "step": 82090 }, { "grad_norm": 0.16968634724617004, "learning_rate": 8.507995059073931e-06, "loss": 0.0152, "step": 82100 }, { "grad_norm": 0.3439812660217285, "learning_rate": 8.498770945514673e-06, "loss": 0.0147, "step": 82110 }, { "grad_norm": 0.2730960547924042, "learning_rate": 8.489551370467436e-06, "loss": 0.0138, "step": 82120 }, { "grad_norm": 0.24223004281520844, "learning_rate": 8.480336334940442e-06, "loss": 0.0238, "step": 82130 }, { "grad_norm": 0.31609877943992615, "learning_rate": 8.471125839941458e-06, "loss": 0.0116, "step": 82140 }, { "grad_norm": 0.2600330412387848, "learning_rate": 8.461919886477682e-06, "loss": 0.0177, "step": 82150 }, { "grad_norm": 0.3776099979877472, "learning_rate": 8.452718475555927e-06, "loss": 0.0169, "step": 82160 }, { "grad_norm": 0.2551383674144745, "learning_rate": 8.443521608182398e-06, "loss": 0.013, "step": 82170 }, { "grad_norm": 0.3521774709224701, "learning_rate": 8.43432928536288e-06, "loss": 0.0158, "step": 82180 }, { "grad_norm": 0.4287693500518799, "learning_rate": 8.425141508102607e-06, "loss": 0.0166, "step": 82190 }, { "grad_norm": 0.17272168397903442, "learning_rate": 8.415958277406365e-06, "loss": 0.0172, "step": 82200 }, { "grad_norm": 0.2594781219959259, "learning_rate": 8.406779594278392e-06, "loss": 0.0157, "step": 82210 }, { "grad_norm": 0.3009469211101532, "learning_rate": 8.397605459722468e-06, "loss": 0.0248, "step": 82220 }, { "grad_norm": 0.2859381139278412, "learning_rate": 8.388435874741863e-06, "loss": 0.0257, "step": 82230 }, { "grad_norm": 0.18528436124324799, "learning_rate": 8.37927084033936e-06, "loss": 0.0209, "step": 82240 }, { "grad_norm": 0.6344321966171265, "learning_rate": 8.370110357517208e-06, "loss": 0.0197, "step": 82250 }, { "grad_norm": 0.5018242001533508, "learning_rate": 8.360954427277201e-06, "loss": 0.0112, "step": 82260 }, { "grad_norm": 0.3085756003856659, "learning_rate": 8.351803050620627e-06, "loss": 0.019, "step": 82270 }, { "grad_norm": 0.4027242064476013, "learning_rate": 8.34265622854824e-06, "loss": 0.0202, "step": 82280 }, { "grad_norm": 0.37463057041168213, "learning_rate": 8.333513962060335e-06, "loss": 0.0274, "step": 82290 }, { "grad_norm": 0.38712260127067566, "learning_rate": 8.324376252156701e-06, "loss": 0.0113, "step": 82300 }, { "grad_norm": 0.24425092339515686, "learning_rate": 8.315243099836633e-06, "loss": 0.0245, "step": 82310 }, { "grad_norm": 0.19515861570835114, "learning_rate": 8.3061145060989e-06, "loss": 0.0192, "step": 82320 }, { "grad_norm": 0.15332676470279694, "learning_rate": 8.296990471941806e-06, "loss": 0.0213, "step": 82330 }, { "grad_norm": 0.1744990199804306, "learning_rate": 8.287870998363112e-06, "loss": 0.0138, "step": 82340 }, { "grad_norm": 0.32064956426620483, "learning_rate": 8.278756086360156e-06, "loss": 0.0156, "step": 82350 }, { "grad_norm": 0.2887340188026428, "learning_rate": 8.269645736929693e-06, "loss": 0.0163, "step": 82360 }, { "grad_norm": 0.1378573626279831, "learning_rate": 8.260539951068048e-06, "loss": 0.0147, "step": 82370 }, { "grad_norm": 0.4598594903945923, "learning_rate": 8.251438729770983e-06, "loss": 0.0222, "step": 82380 }, { "grad_norm": 0.19937646389007568, "learning_rate": 8.242342074033826e-06, "loss": 0.0192, "step": 82390 }, { "grad_norm": 0.20502251386642456, "learning_rate": 8.233249984851344e-06, "loss": 0.0186, "step": 82400 }, { "grad_norm": 0.3890814483165741, "learning_rate": 8.224162463217849e-06, "loss": 0.022, "step": 82410 }, { "grad_norm": 0.45226553082466125, "learning_rate": 8.215079510127139e-06, "loss": 0.0165, "step": 82420 }, { "grad_norm": 0.27177944779396057, "learning_rate": 8.206001126572521e-06, "loss": 0.0174, "step": 82430 }, { "grad_norm": 0.21393457055091858, "learning_rate": 8.196927313546776e-06, "loss": 0.0183, "step": 82440 }, { "grad_norm": 0.22347061336040497, "learning_rate": 8.18785807204222e-06, "loss": 0.0195, "step": 82450 }, { "grad_norm": 0.6346989274024963, "learning_rate": 8.178793403050627e-06, "loss": 0.0183, "step": 82460 }, { "grad_norm": 0.35450977087020874, "learning_rate": 8.16973330756331e-06, "loss": 0.0221, "step": 82470 }, { "grad_norm": 0.44058024883270264, "learning_rate": 8.16067778657107e-06, "loss": 0.0194, "step": 82480 }, { "grad_norm": 1.2470674514770508, "learning_rate": 8.151626841064214e-06, "loss": 0.0161, "step": 82490 }, { "grad_norm": 0.3999944031238556, "learning_rate": 8.142580472032518e-06, "loss": 0.0135, "step": 82500 }, { "grad_norm": 2.09635066986084, "learning_rate": 8.133538680465285e-06, "loss": 0.0173, "step": 82510 }, { "grad_norm": 0.5151164531707764, "learning_rate": 8.124501467351336e-06, "loss": 0.0164, "step": 82520 }, { "grad_norm": 0.9022122025489807, "learning_rate": 8.11546883367893e-06, "loss": 0.0226, "step": 82530 }, { "grad_norm": 0.5317851305007935, "learning_rate": 8.106440780435882e-06, "loss": 0.0218, "step": 82540 }, { "grad_norm": 0.148952454328537, "learning_rate": 8.097417308609485e-06, "loss": 0.0147, "step": 82550 }, { "grad_norm": 0.16600368916988373, "learning_rate": 8.088398419186544e-06, "loss": 0.0183, "step": 82560 }, { "grad_norm": 0.1961141675710678, "learning_rate": 8.079384113153321e-06, "loss": 0.0167, "step": 82570 }, { "grad_norm": 0.25222891569137573, "learning_rate": 8.070374391495644e-06, "loss": 0.0132, "step": 82580 }, { "grad_norm": 0.5161973834037781, "learning_rate": 8.061369255198764e-06, "loss": 0.0167, "step": 82590 }, { "grad_norm": 0.12739573419094086, "learning_rate": 8.052368705247493e-06, "loss": 0.0142, "step": 82600 }, { "grad_norm": 0.2127770334482193, "learning_rate": 8.043372742626114e-06, "loss": 0.0189, "step": 82610 }, { "grad_norm": 0.187111958861351, "learning_rate": 8.034381368318422e-06, "loss": 0.016, "step": 82620 }, { "grad_norm": 0.21874211728572845, "learning_rate": 8.025394583307677e-06, "loss": 0.0177, "step": 82630 }, { "grad_norm": 0.2876918315887451, "learning_rate": 8.016412388576683e-06, "loss": 0.0195, "step": 82640 }, { "grad_norm": 0.10586809366941452, "learning_rate": 8.007434785107699e-06, "loss": 0.0145, "step": 82650 }, { "grad_norm": 0.5462034940719604, "learning_rate": 7.998461773882509e-06, "loss": 0.0206, "step": 82660 }, { "grad_norm": 0.2747938334941864, "learning_rate": 7.989493355882393e-06, "loss": 0.0229, "step": 82670 }, { "grad_norm": 0.3582606315612793, "learning_rate": 7.980529532088137e-06, "loss": 0.016, "step": 82680 }, { "grad_norm": 0.2890263497829437, "learning_rate": 7.97157030347998e-06, "loss": 0.0181, "step": 82690 }, { "grad_norm": 0.296083003282547, "learning_rate": 7.96261567103772e-06, "loss": 0.015, "step": 82700 }, { "grad_norm": 1.4496185779571533, "learning_rate": 7.953665635740593e-06, "loss": 0.021, "step": 82710 }, { "grad_norm": 0.23044362664222717, "learning_rate": 7.94472019856738e-06, "loss": 0.018, "step": 82720 }, { "grad_norm": 0.34423044323921204, "learning_rate": 7.935779360496337e-06, "loss": 0.023, "step": 82730 }, { "grad_norm": 0.4363686442375183, "learning_rate": 7.92684312250523e-06, "loss": 0.0161, "step": 82740 }, { "grad_norm": 0.30094391107559204, "learning_rate": 7.917911485571289e-06, "loss": 0.0156, "step": 82750 }, { "grad_norm": 0.33053281903266907, "learning_rate": 7.90898445067128e-06, "loss": 0.02, "step": 82760 }, { "grad_norm": 0.2724505662918091, "learning_rate": 7.900062018781462e-06, "loss": 0.0182, "step": 82770 }, { "grad_norm": 0.516605794429779, "learning_rate": 7.891144190877553e-06, "loss": 0.015, "step": 82780 }, { "grad_norm": 0.18204405903816223, "learning_rate": 7.882230967934806e-06, "loss": 0.0157, "step": 82790 }, { "grad_norm": 0.3640202283859253, "learning_rate": 7.873322350927959e-06, "loss": 0.0147, "step": 82800 }, { "grad_norm": 0.2676284909248352, "learning_rate": 7.864418340831253e-06, "loss": 0.014, "step": 82810 }, { "grad_norm": 0.47884970903396606, "learning_rate": 7.855518938618395e-06, "loss": 0.0147, "step": 82820 }, { "grad_norm": 0.6728267669677734, "learning_rate": 7.846624145262633e-06, "loss": 0.0183, "step": 82830 }, { "grad_norm": 0.1478470265865326, "learning_rate": 7.837733961736672e-06, "loss": 0.0199, "step": 82840 }, { "grad_norm": 0.3328666388988495, "learning_rate": 7.828848389012727e-06, "loss": 0.0182, "step": 82850 }, { "grad_norm": 0.13186918199062347, "learning_rate": 7.819967428062524e-06, "loss": 0.0139, "step": 82860 }, { "grad_norm": 0.2862522006034851, "learning_rate": 7.811091079857274e-06, "loss": 0.0192, "step": 82870 }, { "grad_norm": 0.3713325262069702, "learning_rate": 7.802219345367667e-06, "loss": 0.0184, "step": 82880 }, { "grad_norm": 0.5512257218360901, "learning_rate": 7.793352225563916e-06, "loss": 0.0141, "step": 82890 }, { "grad_norm": 0.4209160804748535, "learning_rate": 7.7844897214157e-06, "loss": 0.0175, "step": 82900 }, { "grad_norm": 0.16036683320999146, "learning_rate": 7.775631833892228e-06, "loss": 0.015, "step": 82910 }, { "grad_norm": 0.19021467864513397, "learning_rate": 7.76677856396215e-06, "loss": 0.0108, "step": 82920 }, { "grad_norm": 0.30912351608276367, "learning_rate": 7.757929912593697e-06, "loss": 0.0141, "step": 82930 }, { "grad_norm": 0.17955026030540466, "learning_rate": 7.749085880754509e-06, "loss": 0.0214, "step": 82940 }, { "grad_norm": 0.3785037696361542, "learning_rate": 7.740246469411777e-06, "loss": 0.0186, "step": 82950 }, { "grad_norm": 0.13596326112747192, "learning_rate": 7.731411679532142e-06, "loss": 0.0112, "step": 82960 }, { "grad_norm": 0.3621261715888977, "learning_rate": 7.722581512081779e-06, "loss": 0.019, "step": 82970 }, { "grad_norm": 0.568257212638855, "learning_rate": 7.713755968026331e-06, "loss": 0.02, "step": 82980 }, { "grad_norm": 0.21763409674167633, "learning_rate": 7.704935048330969e-06, "loss": 0.0168, "step": 82990 }, { "grad_norm": 0.18252632021903992, "learning_rate": 7.696118753960313e-06, "loss": 0.0125, "step": 83000 }, { "grad_norm": 0.2999555468559265, "learning_rate": 7.687307085878503e-06, "loss": 0.0174, "step": 83010 }, { "grad_norm": 0.10930138826370239, "learning_rate": 7.678500045049186e-06, "loss": 0.0143, "step": 83020 }, { "grad_norm": 0.48217833042144775, "learning_rate": 7.669697632435463e-06, "loss": 0.0167, "step": 83030 }, { "grad_norm": 0.239618182182312, "learning_rate": 7.660899848999963e-06, "loss": 0.0125, "step": 83040 }, { "grad_norm": 0.3573301136493683, "learning_rate": 7.652106695704803e-06, "loss": 0.0202, "step": 83050 }, { "grad_norm": 0.2525336742401123, "learning_rate": 7.643318173511598e-06, "loss": 0.0119, "step": 83060 }, { "grad_norm": 0.4276452958583832, "learning_rate": 7.634534283381423e-06, "loss": 0.0167, "step": 83070 }, { "grad_norm": 0.2876846194267273, "learning_rate": 7.625755026274889e-06, "loss": 0.0193, "step": 83080 }, { "grad_norm": 0.36408329010009766, "learning_rate": 7.616980403152069e-06, "loss": 0.013, "step": 83090 }, { "grad_norm": 0.20587649941444397, "learning_rate": 7.608210414972561e-06, "loss": 0.0156, "step": 83100 }, { "grad_norm": 0.7385250926017761, "learning_rate": 7.599445062695404e-06, "loss": 0.0191, "step": 83110 }, { "grad_norm": 0.3781244456768036, "learning_rate": 7.590684347279214e-06, "loss": 0.0177, "step": 83120 }, { "grad_norm": 0.28932619094848633, "learning_rate": 7.581928269682004e-06, "loss": 0.0138, "step": 83130 }, { "grad_norm": 0.444862425327301, "learning_rate": 7.573176830861361e-06, "loss": 0.0157, "step": 83140 }, { "grad_norm": 1.1786926984786987, "learning_rate": 7.5644300317743e-06, "loss": 0.0242, "step": 83150 }, { "grad_norm": 0.2658219337463379, "learning_rate": 7.555687873377376e-06, "loss": 0.0144, "step": 83160 }, { "grad_norm": 0.2568506598472595, "learning_rate": 7.546950356626592e-06, "loss": 0.0206, "step": 83170 }, { "grad_norm": 0.1632392406463623, "learning_rate": 7.538217482477516e-06, "loss": 0.0119, "step": 83180 }, { "grad_norm": 0.369233638048172, "learning_rate": 7.529489251885119e-06, "loss": 0.0169, "step": 83190 }, { "grad_norm": 0.4484337866306305, "learning_rate": 7.520765665803941e-06, "loss": 0.0173, "step": 83200 }, { "grad_norm": 0.4326188862323761, "learning_rate": 7.512046725187949e-06, "loss": 0.02, "step": 83210 }, { "grad_norm": 0.23873159289360046, "learning_rate": 7.503332430990645e-06, "loss": 0.0177, "step": 83220 }, { "grad_norm": 0.6232613325119019, "learning_rate": 7.494622784165017e-06, "loss": 0.0234, "step": 83230 }, { "grad_norm": 1.0853184461593628, "learning_rate": 7.485917785663543e-06, "loss": 0.0201, "step": 83240 }, { "grad_norm": 0.5616881847381592, "learning_rate": 7.477217436438172e-06, "loss": 0.0223, "step": 83250 }, { "grad_norm": 0.16472779214382172, "learning_rate": 7.46852173744037e-06, "loss": 0.0171, "step": 83260 }, { "grad_norm": 0.3095816969871521, "learning_rate": 7.4598306896211004e-06, "loss": 0.0187, "step": 83270 }, { "grad_norm": 1.0360876321792603, "learning_rate": 7.451144293930773e-06, "loss": 0.0176, "step": 83280 }, { "grad_norm": 0.14277231693267822, "learning_rate": 7.4424625513193455e-06, "loss": 0.0155, "step": 83290 }, { "grad_norm": 0.4573306441307068, "learning_rate": 7.433785462736209e-06, "loss": 0.017, "step": 83300 }, { "grad_norm": 0.33280661702156067, "learning_rate": 7.425113029130315e-06, "loss": 0.0175, "step": 83310 }, { "grad_norm": 0.4619716703891754, "learning_rate": 7.416445251450044e-06, "loss": 0.0267, "step": 83320 }, { "grad_norm": 0.28926554322242737, "learning_rate": 7.4077821306433035e-06, "loss": 0.0134, "step": 83330 }, { "grad_norm": 0.6691237092018127, "learning_rate": 7.399123667657465e-06, "loss": 0.0143, "step": 83340 }, { "grad_norm": 0.7641739249229431, "learning_rate": 7.390469863439426e-06, "loss": 0.0192, "step": 83350 }, { "grad_norm": 0.3710010349750519, "learning_rate": 7.381820718935512e-06, "loss": 0.0196, "step": 83360 }, { "grad_norm": 0.3318849802017212, "learning_rate": 7.373176235091633e-06, "loss": 0.0188, "step": 83370 }, { "grad_norm": 0.18732571601867676, "learning_rate": 7.364536412853101e-06, "loss": 0.0142, "step": 83380 }, { "grad_norm": 0.17564301192760468, "learning_rate": 7.355901253164776e-06, "loss": 0.013, "step": 83390 }, { "grad_norm": 0.1750538945198059, "learning_rate": 7.34727075697097e-06, "loss": 0.0176, "step": 83400 }, { "grad_norm": 0.3966856598854065, "learning_rate": 7.3386449252155095e-06, "loss": 0.0124, "step": 83410 }, { "grad_norm": 0.36792609095573425, "learning_rate": 7.330023758841692e-06, "loss": 0.016, "step": 83420 }, { "grad_norm": 0.3095524311065674, "learning_rate": 7.321407258792323e-06, "loss": 0.0181, "step": 83430 }, { "grad_norm": 0.3626475930213928, "learning_rate": 7.312795426009694e-06, "loss": 0.0128, "step": 83440 }, { "grad_norm": 0.19243460893630981, "learning_rate": 7.304188261435585e-06, "loss": 0.0186, "step": 83450 }, { "grad_norm": 0.7805638313293457, "learning_rate": 7.295585766011242e-06, "loss": 0.0183, "step": 83460 }, { "grad_norm": 0.2705492675304413, "learning_rate": 7.286987940677436e-06, "loss": 0.0174, "step": 83470 }, { "grad_norm": 0.4132890999317169, "learning_rate": 7.278394786374426e-06, "loss": 0.0135, "step": 83480 }, { "grad_norm": 0.29943764209747314, "learning_rate": 7.269806304041915e-06, "loss": 0.0143, "step": 83490 }, { "grad_norm": 0.31175294518470764, "learning_rate": 7.261222494619147e-06, "loss": 0.0145, "step": 83500 }, { "grad_norm": 0.7105796933174133, "learning_rate": 7.252643359044825e-06, "loss": 0.03, "step": 83510 }, { "grad_norm": 0.4723714590072632, "learning_rate": 7.244068898257172e-06, "loss": 0.0142, "step": 83520 }, { "grad_norm": 0.26889941096305847, "learning_rate": 7.2354991131938495e-06, "loss": 0.0233, "step": 83530 }, { "grad_norm": 0.33960142731666565, "learning_rate": 7.2269340047920574e-06, "loss": 0.0103, "step": 83540 }, { "grad_norm": 0.2965891659259796, "learning_rate": 7.218373573988435e-06, "loss": 0.0127, "step": 83550 }, { "grad_norm": 0.2868518829345703, "learning_rate": 7.209817821719178e-06, "loss": 0.021, "step": 83560 }, { "grad_norm": 0.2695451080799103, "learning_rate": 7.201266748919894e-06, "loss": 0.0153, "step": 83570 }, { "grad_norm": 0.5308946371078491, "learning_rate": 7.19272035652575e-06, "loss": 0.0143, "step": 83580 }, { "grad_norm": 0.14762713015079498, "learning_rate": 7.184178645471334e-06, "loss": 0.0149, "step": 83590 }, { "grad_norm": 1.0234249830245972, "learning_rate": 7.175641616690775e-06, "loss": 0.0181, "step": 83600 }, { "grad_norm": 0.533781111240387, "learning_rate": 7.167109271117661e-06, "loss": 0.0153, "step": 83610 }, { "grad_norm": 0.45453208684921265, "learning_rate": 7.158581609685072e-06, "loss": 0.0192, "step": 83620 }, { "grad_norm": 0.20329268276691437, "learning_rate": 7.150058633325591e-06, "loss": 0.0129, "step": 83630 }, { "grad_norm": 0.2610153555870056, "learning_rate": 7.141540342971287e-06, "loss": 0.0152, "step": 83640 }, { "grad_norm": 0.3189498484134674, "learning_rate": 7.133026739553677e-06, "loss": 0.0179, "step": 83650 }, { "grad_norm": 0.2861562669277191, "learning_rate": 7.124517824003829e-06, "loss": 0.0185, "step": 83660 }, { "grad_norm": 0.23678092658519745, "learning_rate": 7.116013597252236e-06, "loss": 0.0183, "step": 83670 }, { "grad_norm": 0.19079621136188507, "learning_rate": 7.107514060228921e-06, "loss": 0.0205, "step": 83680 }, { "grad_norm": 0.40065187215805054, "learning_rate": 7.099019213863378e-06, "loss": 0.0205, "step": 83690 }, { "grad_norm": 0.40440645813941956, "learning_rate": 7.090529059084611e-06, "loss": 0.0184, "step": 83700 }, { "grad_norm": 0.3630419671535492, "learning_rate": 7.082043596821058e-06, "loss": 0.0168, "step": 83710 }, { "grad_norm": 0.1189693883061409, "learning_rate": 7.07356282800069e-06, "loss": 0.0251, "step": 83720 }, { "grad_norm": 0.3260253667831421, "learning_rate": 7.0650867535509645e-06, "loss": 0.0127, "step": 83730 }, { "grad_norm": 0.5509696006774902, "learning_rate": 7.0566153743987895e-06, "loss": 0.0203, "step": 83740 }, { "grad_norm": 0.3942791819572449, "learning_rate": 7.04814869147059e-06, "loss": 0.0201, "step": 83750 }, { "grad_norm": 0.3158680498600006, "learning_rate": 7.0396867056922725e-06, "loss": 0.0214, "step": 83760 }, { "grad_norm": 0.3502140939235687, "learning_rate": 7.031229417989244e-06, "loss": 0.0194, "step": 83770 }, { "grad_norm": 0.41014260053634644, "learning_rate": 7.022776829286348e-06, "loss": 0.0192, "step": 83780 }, { "grad_norm": 0.21663345396518707, "learning_rate": 7.014328940507975e-06, "loss": 0.0182, "step": 83790 }, { "grad_norm": 0.3013513684272766, "learning_rate": 7.005885752577945e-06, "loss": 0.0127, "step": 83800 }, { "grad_norm": 0.36962735652923584, "learning_rate": 6.997447266419615e-06, "loss": 0.0205, "step": 83810 }, { "grad_norm": 0.7518115639686584, "learning_rate": 6.989013482955797e-06, "loss": 0.0225, "step": 83820 }, { "grad_norm": 0.29048413038253784, "learning_rate": 6.980584403108808e-06, "loss": 0.0136, "step": 83830 }, { "grad_norm": 0.8822706341743469, "learning_rate": 6.972160027800417e-06, "loss": 0.0164, "step": 83840 }, { "grad_norm": 0.3443428575992584, "learning_rate": 6.963740357951931e-06, "loss": 0.0221, "step": 83850 }, { "grad_norm": 0.35376331210136414, "learning_rate": 6.95532539448408e-06, "loss": 0.0222, "step": 83860 }, { "grad_norm": 0.4715081453323364, "learning_rate": 6.946915138317128e-06, "loss": 0.0183, "step": 83870 }, { "grad_norm": 0.3546229898929596, "learning_rate": 6.9385095903708065e-06, "loss": 0.0151, "step": 83880 }, { "grad_norm": 0.41044941544532776, "learning_rate": 6.930108751564346e-06, "loss": 0.0181, "step": 83890 }, { "grad_norm": 0.3007082939147949, "learning_rate": 6.921712622816428e-06, "loss": 0.0158, "step": 83900 }, { "grad_norm": 0.537632405757904, "learning_rate": 6.913321205045259e-06, "loss": 0.0127, "step": 83910 }, { "grad_norm": 0.35134226083755493, "learning_rate": 6.904934499168491e-06, "loss": 0.0144, "step": 83920 }, { "grad_norm": 0.32204902172088623, "learning_rate": 6.896552506103299e-06, "loss": 0.0187, "step": 83930 }, { "grad_norm": 0.1633823662996292, "learning_rate": 6.888175226766313e-06, "loss": 0.0207, "step": 83940 }, { "grad_norm": 0.32869526743888855, "learning_rate": 6.879802662073686e-06, "loss": 0.017, "step": 83950 }, { "grad_norm": 0.308162659406662, "learning_rate": 6.871434812940991e-06, "loss": 0.021, "step": 83960 }, { "grad_norm": 0.31581535935401917, "learning_rate": 6.863071680283345e-06, "loss": 0.0208, "step": 83970 }, { "grad_norm": 0.18533489108085632, "learning_rate": 6.854713265015339e-06, "loss": 0.0291, "step": 83980 }, { "grad_norm": 0.6208977103233337, "learning_rate": 6.8463595680510025e-06, "loss": 0.0204, "step": 83990 }, { "grad_norm": 0.28038671612739563, "learning_rate": 6.838010590303911e-06, "loss": 0.0148, "step": 84000 }, { "grad_norm": 0.16054551303386688, "learning_rate": 6.829666332687085e-06, "loss": 0.0139, "step": 84010 }, { "grad_norm": 0.38169747591018677, "learning_rate": 6.821326796113059e-06, "loss": 0.013, "step": 84020 }, { "grad_norm": 0.17464573681354523, "learning_rate": 6.812991981493799e-06, "loss": 0.0145, "step": 84030 }, { "grad_norm": 0.4019191563129425, "learning_rate": 6.804661889740821e-06, "loss": 0.0177, "step": 84040 }, { "grad_norm": 0.26628178358078003, "learning_rate": 6.7963365217650625e-06, "loss": 0.0133, "step": 84050 }, { "grad_norm": 0.21477776765823364, "learning_rate": 6.788015878476983e-06, "loss": 0.0169, "step": 84060 }, { "grad_norm": 0.17091624438762665, "learning_rate": 6.779699960786523e-06, "loss": 0.0179, "step": 84070 }, { "grad_norm": 0.4650917649269104, "learning_rate": 6.771388769603104e-06, "loss": 0.0162, "step": 84080 }, { "grad_norm": 0.2390110343694687, "learning_rate": 6.763082305835605e-06, "loss": 0.0234, "step": 84090 }, { "grad_norm": 0.21424894034862518, "learning_rate": 6.754780570392433e-06, "loss": 0.0139, "step": 84100 }, { "grad_norm": 1.131618618965149, "learning_rate": 6.7464835641814215e-06, "loss": 0.0181, "step": 84110 }, { "grad_norm": 0.2082808017730713, "learning_rate": 6.7381912881099515e-06, "loss": 0.0167, "step": 84120 }, { "grad_norm": 0.2832790017127991, "learning_rate": 6.72990374308482e-06, "loss": 0.0128, "step": 84130 }, { "grad_norm": 0.2769516706466675, "learning_rate": 6.721620930012379e-06, "loss": 0.0122, "step": 84140 }, { "grad_norm": 0.48765119910240173, "learning_rate": 6.713342849798393e-06, "loss": 0.0245, "step": 84150 }, { "grad_norm": 0.27001190185546875, "learning_rate": 6.705069503348166e-06, "loss": 0.0161, "step": 84160 }, { "grad_norm": 0.3422119915485382, "learning_rate": 6.696800891566435e-06, "loss": 0.013, "step": 84170 }, { "grad_norm": 0.2386988401412964, "learning_rate": 6.688537015357449e-06, "loss": 0.0165, "step": 84180 }, { "grad_norm": 0.38388893008232117, "learning_rate": 6.680277875624941e-06, "loss": 0.0177, "step": 84190 }, { "grad_norm": 0.3125931918621063, "learning_rate": 6.672023473272121e-06, "loss": 0.0181, "step": 84200 }, { "grad_norm": 0.2675476372241974, "learning_rate": 6.663773809201667e-06, "loss": 0.0149, "step": 84210 }, { "grad_norm": 1.3914434909820557, "learning_rate": 6.655528884315748e-06, "loss": 0.0229, "step": 84220 }, { "grad_norm": 0.22669866681098938, "learning_rate": 6.647288699516036e-06, "loss": 0.0132, "step": 84230 }, { "grad_norm": 0.3890441656112671, "learning_rate": 6.639053255703642e-06, "loss": 0.0219, "step": 84240 }, { "grad_norm": 0.2153523862361908, "learning_rate": 6.6308225537791925e-06, "loss": 0.014, "step": 84250 }, { "grad_norm": 0.24248577654361725, "learning_rate": 6.622596594642777e-06, "loss": 0.0207, "step": 84260 }, { "grad_norm": 0.42940130829811096, "learning_rate": 6.614375379193999e-06, "loss": 0.0142, "step": 84270 }, { "grad_norm": 0.5725437998771667, "learning_rate": 6.606158908331889e-06, "loss": 0.0173, "step": 84280 }, { "grad_norm": 0.33688104152679443, "learning_rate": 6.597947182955005e-06, "loss": 0.0238, "step": 84290 }, { "grad_norm": 0.19812165200710297, "learning_rate": 6.589740203961359e-06, "loss": 0.0199, "step": 84300 }, { "grad_norm": 0.902482271194458, "learning_rate": 6.581537972248464e-06, "loss": 0.0202, "step": 84310 }, { "grad_norm": 0.3324979543685913, "learning_rate": 6.573340488713275e-06, "loss": 0.0161, "step": 84320 }, { "grad_norm": 0.16527825593948364, "learning_rate": 6.565147754252304e-06, "loss": 0.0175, "step": 84330 }, { "grad_norm": 0.17989811301231384, "learning_rate": 6.556959769761462e-06, "loss": 0.0171, "step": 84340 }, { "grad_norm": 0.38880565762519836, "learning_rate": 6.548776536136192e-06, "loss": 0.0178, "step": 84350 }, { "grad_norm": 0.2177702635526657, "learning_rate": 6.54059805427138e-06, "loss": 0.0177, "step": 84360 }, { "grad_norm": 0.4724600315093994, "learning_rate": 6.532424325061437e-06, "loss": 0.0196, "step": 84370 }, { "grad_norm": 0.4150477647781372, "learning_rate": 6.5242553494001935e-06, "loss": 0.013, "step": 84380 }, { "grad_norm": 0.23609554767608643, "learning_rate": 6.516091128181046e-06, "loss": 0.0137, "step": 84390 }, { "grad_norm": 0.49439993500709534, "learning_rate": 6.507931662296785e-06, "loss": 0.0183, "step": 84400 }, { "grad_norm": 0.1429225206375122, "learning_rate": 6.4997769526397354e-06, "loss": 0.0133, "step": 84410 }, { "grad_norm": 0.48903489112854004, "learning_rate": 6.491627000101663e-06, "loss": 0.0198, "step": 84420 }, { "grad_norm": 0.21694625914096832, "learning_rate": 6.4834818055738545e-06, "loss": 0.0154, "step": 84430 }, { "grad_norm": 0.19313868880271912, "learning_rate": 6.4753413699470465e-06, "loss": 0.0145, "step": 84440 }, { "grad_norm": 0.2942962944507599, "learning_rate": 6.467205694111478e-06, "loss": 0.0179, "step": 84450 }, { "grad_norm": 0.5758455991744995, "learning_rate": 6.459074778956836e-06, "loss": 0.0199, "step": 84460 }, { "grad_norm": 0.20849797129631042, "learning_rate": 6.450948625372311e-06, "loss": 0.0208, "step": 84470 }, { "grad_norm": 0.20844098925590515, "learning_rate": 6.442827234246585e-06, "loss": 0.0124, "step": 84480 }, { "grad_norm": 0.28192436695098877, "learning_rate": 6.43471060646777e-06, "loss": 0.0149, "step": 84490 }, { "grad_norm": 0.33126139640808105, "learning_rate": 6.426598742923512e-06, "loss": 0.0165, "step": 84500 }, { "grad_norm": 0.6442299485206604, "learning_rate": 6.418491644500885e-06, "loss": 0.0166, "step": 84510 }, { "grad_norm": 0.37052634358406067, "learning_rate": 6.410389312086512e-06, "loss": 0.0151, "step": 84520 }, { "grad_norm": 0.724182665348053, "learning_rate": 6.4022917465664135e-06, "loss": 0.0182, "step": 84530 }, { "grad_norm": 0.43717893958091736, "learning_rate": 6.3941989488261575e-06, "loss": 0.0234, "step": 84540 }, { "grad_norm": 0.26750293374061584, "learning_rate": 6.386110919750726e-06, "loss": 0.0152, "step": 84550 }, { "grad_norm": 0.24139763414859772, "learning_rate": 6.378027660224651e-06, "loss": 0.0125, "step": 84560 }, { "grad_norm": 0.3356485366821289, "learning_rate": 6.369949171131856e-06, "loss": 0.0144, "step": 84570 }, { "grad_norm": 0.24739518761634827, "learning_rate": 6.361875453355848e-06, "loss": 0.0129, "step": 84580 }, { "grad_norm": 0.23798267543315887, "learning_rate": 6.353806507779525e-06, "loss": 0.0138, "step": 84590 }, { "grad_norm": 0.33067622780799866, "learning_rate": 6.3457423352853044e-06, "loss": 0.0159, "step": 84600 }, { "grad_norm": 0.21747110784053802, "learning_rate": 6.337682936755062e-06, "loss": 0.0162, "step": 84610 }, { "grad_norm": 0.25239297747612, "learning_rate": 6.329628313070179e-06, "loss": 0.0165, "step": 84620 }, { "grad_norm": 0.4183598756790161, "learning_rate": 6.321578465111477e-06, "loss": 0.0132, "step": 84630 }, { "grad_norm": 0.4810405969619751, "learning_rate": 6.31353339375928e-06, "loss": 0.0165, "step": 84640 }, { "grad_norm": 0.3874051868915558, "learning_rate": 6.305493099893395e-06, "loss": 0.0134, "step": 84650 }, { "grad_norm": 0.3836261034011841, "learning_rate": 6.297457584393096e-06, "loss": 0.0164, "step": 84660 }, { "grad_norm": 0.3399176597595215, "learning_rate": 6.289426848137126e-06, "loss": 0.0219, "step": 84670 }, { "grad_norm": 0.3672768473625183, "learning_rate": 6.281400892003713e-06, "loss": 0.0122, "step": 84680 }, { "grad_norm": 0.23024632036685944, "learning_rate": 6.273379716870581e-06, "loss": 0.0164, "step": 84690 }, { "grad_norm": 0.37584713101387024, "learning_rate": 6.265363323614887e-06, "loss": 0.0151, "step": 84700 }, { "grad_norm": 0.14825181663036346, "learning_rate": 6.257351713113308e-06, "loss": 0.0161, "step": 84710 }, { "grad_norm": 0.5909973978996277, "learning_rate": 6.249344886241981e-06, "loss": 0.0121, "step": 84720 }, { "grad_norm": 0.40802276134490967, "learning_rate": 6.2413428438765295e-06, "loss": 0.0171, "step": 84730 }, { "grad_norm": 0.2654789090156555, "learning_rate": 6.233345586892026e-06, "loss": 0.017, "step": 84740 }, { "grad_norm": 0.40658754110336304, "learning_rate": 6.225353116163052e-06, "loss": 0.0256, "step": 84750 }, { "grad_norm": 0.43035954236984253, "learning_rate": 6.217365432563627e-06, "loss": 0.0267, "step": 84760 }, { "grad_norm": 0.28294631838798523, "learning_rate": 6.209382536967318e-06, "loss": 0.013, "step": 84770 }, { "grad_norm": 0.2890709936618805, "learning_rate": 6.201404430247082e-06, "loss": 0.0204, "step": 84780 }, { "grad_norm": 0.3034762144088745, "learning_rate": 6.1934311132754165e-06, "loss": 0.0128, "step": 84790 }, { "grad_norm": 0.48233258724212646, "learning_rate": 6.185462586924257e-06, "loss": 0.0179, "step": 84800 }, { "grad_norm": 0.18248943984508514, "learning_rate": 6.177498852065039e-06, "loss": 0.0133, "step": 84810 }, { "grad_norm": 0.2606765925884247, "learning_rate": 6.169539909568655e-06, "loss": 0.0199, "step": 84820 }, { "grad_norm": 0.4703151285648346, "learning_rate": 6.161585760305488e-06, "loss": 0.0151, "step": 84830 }, { "grad_norm": 0.3224543631076813, "learning_rate": 6.153636405145391e-06, "loss": 0.0177, "step": 84840 }, { "grad_norm": 1.542529582977295, "learning_rate": 6.145691844957702e-06, "loss": 0.0147, "step": 84850 }, { "grad_norm": 0.2817152142524719, "learning_rate": 6.1377520806112145e-06, "loss": 0.0251, "step": 84860 }, { "grad_norm": 0.14211587607860565, "learning_rate": 6.129817112974218e-06, "loss": 0.0197, "step": 84870 }, { "grad_norm": 0.25226029753685, "learning_rate": 6.121886942914457e-06, "loss": 0.015, "step": 84880 }, { "grad_norm": 0.32361844182014465, "learning_rate": 6.11396157129917e-06, "loss": 0.015, "step": 84890 }, { "grad_norm": 0.29363980889320374, "learning_rate": 6.106040998995066e-06, "loss": 0.02, "step": 84900 }, { "grad_norm": 0.224960058927536, "learning_rate": 6.098125226868329e-06, "loss": 0.0163, "step": 84910 }, { "grad_norm": 0.22053125500679016, "learning_rate": 6.090214255784604e-06, "loss": 0.0148, "step": 84920 }, { "grad_norm": 0.17174243927001953, "learning_rate": 6.0823080866090346e-06, "loss": 0.0253, "step": 84930 }, { "grad_norm": 0.28419196605682373, "learning_rate": 6.074406720206233e-06, "loss": 0.0153, "step": 84940 }, { "grad_norm": 0.42721790075302124, "learning_rate": 6.066510157440259e-06, "loss": 0.0151, "step": 84950 }, { "grad_norm": 0.3787602484226227, "learning_rate": 6.058618399174676e-06, "loss": 0.0176, "step": 84960 }, { "grad_norm": 1.177001953125, "learning_rate": 6.050731446272528e-06, "loss": 0.0211, "step": 84970 }, { "grad_norm": 0.3551856279373169, "learning_rate": 6.042849299596321e-06, "loss": 0.0115, "step": 84980 }, { "grad_norm": 0.5889699459075928, "learning_rate": 6.034971960008013e-06, "loss": 0.019, "step": 84990 }, { "grad_norm": 0.2565898299217224, "learning_rate": 6.027099428369082e-06, "loss": 0.014, "step": 85000 }, { "grad_norm": 0.10837864130735397, "learning_rate": 6.019231705540435e-06, "loss": 0.0116, "step": 85010 }, { "grad_norm": 0.17614230513572693, "learning_rate": 6.011368792382488e-06, "loss": 0.0215, "step": 85020 }, { "grad_norm": 0.22392894327640533, "learning_rate": 6.00351068975511e-06, "loss": 0.0199, "step": 85030 }, { "grad_norm": 0.20076240599155426, "learning_rate": 5.9956573985176674e-06, "loss": 0.0169, "step": 85040 }, { "grad_norm": 0.4069081246852875, "learning_rate": 5.987808919528959e-06, "loss": 0.0173, "step": 85050 }, { "grad_norm": 0.35337767004966736, "learning_rate": 5.979965253647307e-06, "loss": 0.0271, "step": 85060 }, { "grad_norm": 0.19241397082805634, "learning_rate": 5.9721264017304655e-06, "loss": 0.0133, "step": 85070 }, { "grad_norm": 0.36696094274520874, "learning_rate": 5.964292364635682e-06, "loss": 0.0203, "step": 85080 }, { "grad_norm": 0.14436770975589752, "learning_rate": 5.956463143219682e-06, "loss": 0.0153, "step": 85090 }, { "grad_norm": 0.2386423945426941, "learning_rate": 5.948638738338663e-06, "loss": 0.0176, "step": 85100 }, { "grad_norm": 0.2955390512943268, "learning_rate": 5.940819150848271e-06, "loss": 0.017, "step": 85110 }, { "grad_norm": 0.3620980381965637, "learning_rate": 5.933004381603663e-06, "loss": 0.0151, "step": 85120 }, { "grad_norm": 0.24644404649734497, "learning_rate": 5.92519443145943e-06, "loss": 0.0163, "step": 85130 }, { "grad_norm": 0.1684572845697403, "learning_rate": 5.917389301269671e-06, "loss": 0.0137, "step": 85140 }, { "grad_norm": 0.4114637076854706, "learning_rate": 5.9095889918879435e-06, "loss": 0.0169, "step": 85150 }, { "grad_norm": 0.6423979997634888, "learning_rate": 5.901793504167286e-06, "loss": 0.019, "step": 85160 }, { "grad_norm": 0.1283951997756958, "learning_rate": 5.894002838960172e-06, "loss": 0.0152, "step": 85170 }, { "grad_norm": 0.2656072676181793, "learning_rate": 5.886216997118604e-06, "loss": 0.0166, "step": 85180 }, { "grad_norm": 0.7441432476043701, "learning_rate": 5.878435979494023e-06, "loss": 0.0217, "step": 85190 }, { "grad_norm": 0.2335137128829956, "learning_rate": 5.870659786937344e-06, "loss": 0.0134, "step": 85200 }, { "grad_norm": 0.40316009521484375, "learning_rate": 5.862888420298962e-06, "loss": 0.0092, "step": 85210 }, { "grad_norm": 0.271150678396225, "learning_rate": 5.855121880428743e-06, "loss": 0.0106, "step": 85220 }, { "grad_norm": 0.39440682530403137, "learning_rate": 5.84736016817603e-06, "loss": 0.0167, "step": 85230 }, { "grad_norm": 0.23608455061912537, "learning_rate": 5.839603284389622e-06, "loss": 0.0179, "step": 85240 }, { "grad_norm": 0.4789820909500122, "learning_rate": 5.831851229917812e-06, "loss": 0.0132, "step": 85250 }, { "grad_norm": 0.5359089970588684, "learning_rate": 5.824104005608338e-06, "loss": 0.0118, "step": 85260 }, { "grad_norm": 0.4575055241584778, "learning_rate": 5.81636161230843e-06, "loss": 0.0157, "step": 85270 }, { "grad_norm": 0.308137983083725, "learning_rate": 5.808624050864791e-06, "loss": 0.0173, "step": 85280 }, { "grad_norm": 0.5549220442771912, "learning_rate": 5.800891322123592e-06, "loss": 0.0248, "step": 85290 }, { "grad_norm": 0.36787164211273193, "learning_rate": 5.793163426930454e-06, "loss": 0.0111, "step": 85300 }, { "grad_norm": 0.2150031477212906, "learning_rate": 5.785440366130512e-06, "loss": 0.0151, "step": 85310 }, { "grad_norm": 0.5010440349578857, "learning_rate": 5.777722140568326e-06, "loss": 0.0144, "step": 85320 }, { "grad_norm": 0.3008991777896881, "learning_rate": 5.7700087510879685e-06, "loss": 0.0148, "step": 85330 }, { "grad_norm": 0.16826115548610687, "learning_rate": 5.762300198532933e-06, "loss": 0.0115, "step": 85340 }, { "grad_norm": 0.4192766845226288, "learning_rate": 5.7545964837462606e-06, "loss": 0.0142, "step": 85350 }, { "grad_norm": 0.28187480568885803, "learning_rate": 5.746897607570384e-06, "loss": 0.0143, "step": 85360 }, { "grad_norm": 0.321345716714859, "learning_rate": 5.7392035708472626e-06, "loss": 0.0124, "step": 85370 }, { "grad_norm": 0.17373889684677124, "learning_rate": 5.731514374418278e-06, "loss": 0.0144, "step": 85380 }, { "grad_norm": 0.12628434598445892, "learning_rate": 5.7238300191243325e-06, "loss": 0.0128, "step": 85390 }, { "grad_norm": 0.944804310798645, "learning_rate": 5.716150505805762e-06, "loss": 0.0158, "step": 85400 }, { "grad_norm": 0.3344131112098694, "learning_rate": 5.708475835302407e-06, "loss": 0.0214, "step": 85410 }, { "grad_norm": 0.4199320673942566, "learning_rate": 5.70080600845353e-06, "loss": 0.033, "step": 85420 }, { "grad_norm": 0.18403033912181854, "learning_rate": 5.693141026097909e-06, "loss": 0.0214, "step": 85430 }, { "grad_norm": 0.5742470622062683, "learning_rate": 5.6854808890737745e-06, "loss": 0.015, "step": 85440 }, { "grad_norm": 0.28150832653045654, "learning_rate": 5.6778255982188225e-06, "loss": 0.0211, "step": 85450 }, { "grad_norm": 0.20583058893680573, "learning_rate": 5.670175154370217e-06, "loss": 0.0137, "step": 85460 }, { "grad_norm": 0.5186055898666382, "learning_rate": 5.662529558364616e-06, "loss": 0.0174, "step": 85470 }, { "grad_norm": 0.19387289881706238, "learning_rate": 5.654888811038128e-06, "loss": 0.0162, "step": 85480 }, { "grad_norm": 0.27747225761413574, "learning_rate": 5.647252913226314e-06, "loss": 0.0144, "step": 85490 }, { "grad_norm": 0.28459179401397705, "learning_rate": 5.639621865764255e-06, "loss": 0.0164, "step": 85500 }, { "grad_norm": 0.4271419048309326, "learning_rate": 5.63199566948644e-06, "loss": 0.0172, "step": 85510 }, { "grad_norm": 0.40363001823425293, "learning_rate": 5.6243743252268835e-06, "loss": 0.0182, "step": 85520 }, { "grad_norm": 0.3651552200317383, "learning_rate": 5.616757833819009e-06, "loss": 0.0137, "step": 85530 }, { "grad_norm": 0.5731166005134583, "learning_rate": 5.609146196095794e-06, "loss": 0.0179, "step": 85540 }, { "grad_norm": 0.5009630918502808, "learning_rate": 5.6015394128896005e-06, "loss": 0.0163, "step": 85550 }, { "grad_norm": 0.28715115785598755, "learning_rate": 5.593937485032308e-06, "loss": 0.0135, "step": 85560 }, { "grad_norm": 0.6829034090042114, "learning_rate": 5.586340413355246e-06, "loss": 0.0268, "step": 85570 }, { "grad_norm": 0.23150570690631866, "learning_rate": 5.578748198689227e-06, "loss": 0.0134, "step": 85580 }, { "grad_norm": 0.23265540599822998, "learning_rate": 5.571160841864504e-06, "loss": 0.021, "step": 85590 }, { "grad_norm": 0.2664515972137451, "learning_rate": 5.563578343710846e-06, "loss": 0.0125, "step": 85600 }, { "grad_norm": 0.2669360339641571, "learning_rate": 5.556000705057446e-06, "loss": 0.015, "step": 85610 }, { "grad_norm": 0.42122334241867065, "learning_rate": 5.548427926732996e-06, "loss": 0.0167, "step": 85620 }, { "grad_norm": 0.2972952127456665, "learning_rate": 5.540860009565629e-06, "loss": 0.02, "step": 85630 }, { "grad_norm": 0.20262394845485687, "learning_rate": 5.533296954382966e-06, "loss": 0.0168, "step": 85640 }, { "grad_norm": 0.15286670625209808, "learning_rate": 5.525738762012095e-06, "loss": 0.0132, "step": 85650 }, { "grad_norm": 0.28052520751953125, "learning_rate": 5.518185433279577e-06, "loss": 0.0157, "step": 85660 }, { "grad_norm": 0.2464313954114914, "learning_rate": 5.510636969011407e-06, "loss": 0.0143, "step": 85670 }, { "grad_norm": 0.14175951480865479, "learning_rate": 5.503093370033096e-06, "loss": 0.0156, "step": 85680 }, { "grad_norm": 0.2038637399673462, "learning_rate": 5.495554637169603e-06, "loss": 0.0119, "step": 85690 }, { "grad_norm": 0.3556685149669647, "learning_rate": 5.488020771245328e-06, "loss": 0.0199, "step": 85700 }, { "grad_norm": 0.24801743030548096, "learning_rate": 5.480491773084195e-06, "loss": 0.0174, "step": 85710 }, { "grad_norm": 0.32289883494377136, "learning_rate": 5.472967643509525e-06, "loss": 0.0131, "step": 85720 }, { "grad_norm": 0.23115549981594086, "learning_rate": 5.46544838334419e-06, "loss": 0.0137, "step": 85730 }, { "grad_norm": 0.25304025411605835, "learning_rate": 5.4579339934104525e-06, "loss": 0.0157, "step": 85740 }, { "grad_norm": 0.2494695633649826, "learning_rate": 5.450424474530091e-06, "loss": 0.0196, "step": 85750 }, { "grad_norm": 0.2909232974052429, "learning_rate": 5.442919827524323e-06, "loss": 0.0191, "step": 85760 }, { "grad_norm": 0.24977834522724152, "learning_rate": 5.435420053213863e-06, "loss": 0.0147, "step": 85770 }, { "grad_norm": 0.7132022380828857, "learning_rate": 5.427925152418845e-06, "loss": 0.026, "step": 85780 }, { "grad_norm": 0.39210066199302673, "learning_rate": 5.420435125958934e-06, "loss": 0.0154, "step": 85790 }, { "grad_norm": 0.4169138967990875, "learning_rate": 5.412949974653209e-06, "loss": 0.0189, "step": 85800 }, { "grad_norm": 0.8125894069671631, "learning_rate": 5.405469699320248e-06, "loss": 0.0133, "step": 85810 }, { "grad_norm": 0.2708275020122528, "learning_rate": 5.397994300778064e-06, "loss": 0.0116, "step": 85820 }, { "grad_norm": 0.44243988394737244, "learning_rate": 5.39052377984417e-06, "loss": 0.0155, "step": 85830 }, { "grad_norm": 0.40357714891433716, "learning_rate": 5.383058137335523e-06, "loss": 0.017, "step": 85840 }, { "grad_norm": 0.2019617110490799, "learning_rate": 5.375597374068553e-06, "loss": 0.0219, "step": 85850 }, { "grad_norm": 0.18780788779258728, "learning_rate": 5.368141490859163e-06, "loss": 0.0124, "step": 85860 }, { "grad_norm": 0.8559310436248779, "learning_rate": 5.360690488522729e-06, "loss": 0.0164, "step": 85870 }, { "grad_norm": 0.36111971735954285, "learning_rate": 5.353244367874055e-06, "loss": 0.0136, "step": 85880 }, { "grad_norm": 0.8431598544120789, "learning_rate": 5.345803129727456e-06, "loss": 0.0148, "step": 85890 }, { "grad_norm": 0.47921058535575867, "learning_rate": 5.338366774896697e-06, "loss": 0.0194, "step": 85900 }, { "grad_norm": 0.36995118856430054, "learning_rate": 5.330935304194995e-06, "loss": 0.0201, "step": 85910 }, { "grad_norm": 0.1673852503299713, "learning_rate": 5.323508718435044e-06, "loss": 0.0135, "step": 85920 }, { "grad_norm": 0.38061100244522095, "learning_rate": 5.316087018429011e-06, "loss": 0.0133, "step": 85930 }, { "grad_norm": 0.3419508635997772, "learning_rate": 5.30867020498853e-06, "loss": 0.0153, "step": 85940 }, { "grad_norm": 0.5525317192077637, "learning_rate": 5.30125827892467e-06, "loss": 0.0212, "step": 85950 }, { "grad_norm": 4.135205268859863, "learning_rate": 5.293851241048015e-06, "loss": 0.0223, "step": 85960 }, { "grad_norm": 0.3396958112716675, "learning_rate": 5.286449092168555e-06, "loss": 0.0141, "step": 85970 }, { "grad_norm": 0.42298537492752075, "learning_rate": 5.2790518330958105e-06, "loss": 0.0161, "step": 85980 }, { "grad_norm": 0.5461321473121643, "learning_rate": 5.271659464638717e-06, "loss": 0.0127, "step": 85990 }, { "grad_norm": 0.20171718299388885, "learning_rate": 5.264271987605701e-06, "loss": 0.0135, "step": 86000 }, { "grad_norm": 0.16525337100028992, "learning_rate": 5.256889402804632e-06, "loss": 0.016, "step": 86010 }, { "grad_norm": 0.5153716206550598, "learning_rate": 5.249511711042881e-06, "loss": 0.023, "step": 86020 }, { "grad_norm": 0.316781610250473, "learning_rate": 5.242138913127237e-06, "loss": 0.0121, "step": 86030 }, { "grad_norm": 0.4072447717189789, "learning_rate": 5.234771009863987e-06, "loss": 0.0154, "step": 86040 }, { "grad_norm": 0.27972692251205444, "learning_rate": 5.227408002058881e-06, "loss": 0.0205, "step": 86050 }, { "grad_norm": 0.2651077210903168, "learning_rate": 5.220049890517126e-06, "loss": 0.0171, "step": 86060 }, { "grad_norm": 0.526052713394165, "learning_rate": 5.212696676043377e-06, "loss": 0.0148, "step": 86070 }, { "grad_norm": 0.27826616168022156, "learning_rate": 5.205348359441798e-06, "loss": 0.0213, "step": 86080 }, { "grad_norm": 0.25158214569091797, "learning_rate": 5.198004941515966e-06, "loss": 0.0139, "step": 86090 }, { "grad_norm": 0.9192835092544556, "learning_rate": 5.190666423068952e-06, "loss": 0.0196, "step": 86100 }, { "grad_norm": 0.177939772605896, "learning_rate": 5.183332804903284e-06, "loss": 0.0217, "step": 86110 }, { "grad_norm": 0.3294724225997925, "learning_rate": 5.176004087820979e-06, "loss": 0.0159, "step": 86120 }, { "grad_norm": 0.21363572776317596, "learning_rate": 5.168680272623455e-06, "loss": 0.0149, "step": 86130 }, { "grad_norm": 0.43562212586402893, "learning_rate": 5.161361360111661e-06, "loss": 0.0201, "step": 86140 }, { "grad_norm": 0.3323935866355896, "learning_rate": 5.154047351085984e-06, "loss": 0.0231, "step": 86150 }, { "grad_norm": 0.23911169171333313, "learning_rate": 5.1467382463462385e-06, "loss": 0.0167, "step": 86160 }, { "grad_norm": 0.5836054086685181, "learning_rate": 5.139434046691788e-06, "loss": 0.018, "step": 86170 }, { "grad_norm": 0.29285186529159546, "learning_rate": 5.13213475292137e-06, "loss": 0.0155, "step": 86180 }, { "grad_norm": 0.34074488282203674, "learning_rate": 5.12484036583325e-06, "loss": 0.0163, "step": 86190 }, { "grad_norm": 0.3508915305137634, "learning_rate": 5.117550886225109e-06, "loss": 0.0147, "step": 86200 }, { "grad_norm": 0.2766612470149994, "learning_rate": 5.11026631489413e-06, "loss": 0.017, "step": 86210 }, { "grad_norm": 0.33757710456848145, "learning_rate": 5.102986652636926e-06, "loss": 0.0179, "step": 86220 }, { "grad_norm": 0.26232439279556274, "learning_rate": 5.095711900249606e-06, "loss": 0.0159, "step": 86230 }, { "grad_norm": 0.2704542279243469, "learning_rate": 5.088442058527715e-06, "loss": 0.0121, "step": 86240 }, { "grad_norm": 0.3594188988208771, "learning_rate": 5.08117712826629e-06, "loss": 0.02, "step": 86250 }, { "grad_norm": 0.121707484126091, "learning_rate": 5.073917110259785e-06, "loss": 0.0167, "step": 86260 }, { "grad_norm": 0.24453549087047577, "learning_rate": 5.066662005302175e-06, "loss": 0.0139, "step": 86270 }, { "grad_norm": 0.4918479025363922, "learning_rate": 5.059411814186838e-06, "loss": 0.0158, "step": 86280 }, { "grad_norm": 0.26266634464263916, "learning_rate": 5.05216653770666e-06, "loss": 0.0236, "step": 86290 }, { "grad_norm": 0.3966384828090668, "learning_rate": 5.044926176653969e-06, "loss": 0.0105, "step": 86300 }, { "grad_norm": 1.1913055181503296, "learning_rate": 5.037690731820571e-06, "loss": 0.0173, "step": 86310 }, { "grad_norm": 0.3022376000881195, "learning_rate": 5.030460203997711e-06, "loss": 0.0154, "step": 86320 }, { "grad_norm": 0.2889682650566101, "learning_rate": 5.0232345939761174e-06, "loss": 0.0131, "step": 86330 }, { "grad_norm": 0.31930917501449585, "learning_rate": 5.016013902545957e-06, "loss": 0.02, "step": 86340 }, { "grad_norm": 0.1556847095489502, "learning_rate": 5.008798130496889e-06, "loss": 0.0179, "step": 86350 }, { "grad_norm": 0.6967374086380005, "learning_rate": 5.001587278618008e-06, "loss": 0.0174, "step": 86360 }, { "grad_norm": 1.063456416130066, "learning_rate": 4.994381347697891e-06, "loss": 0.0129, "step": 86370 }, { "grad_norm": 0.31283193826675415, "learning_rate": 4.987180338524572e-06, "loss": 0.0112, "step": 86380 }, { "grad_norm": 0.2255260944366455, "learning_rate": 4.9799842518855275e-06, "loss": 0.0164, "step": 86390 }, { "grad_norm": 0.37540683150291443, "learning_rate": 4.9727930885677285e-06, "loss": 0.0184, "step": 86400 }, { "grad_norm": 0.2374560385942459, "learning_rate": 4.965606849357573e-06, "loss": 0.0168, "step": 86410 }, { "grad_norm": 0.767571747303009, "learning_rate": 4.9584255350409445e-06, "loss": 0.015, "step": 86420 }, { "grad_norm": 0.24924373626708984, "learning_rate": 4.951249146403175e-06, "loss": 0.0146, "step": 86430 }, { "grad_norm": 0.25125372409820557, "learning_rate": 4.9440776842290815e-06, "loss": 0.0155, "step": 86440 }, { "grad_norm": 0.21288231015205383, "learning_rate": 4.936911149302903e-06, "loss": 0.019, "step": 86450 }, { "grad_norm": 0.505344808101654, "learning_rate": 4.929749542408379e-06, "loss": 0.0211, "step": 86460 }, { "grad_norm": 0.29283612966537476, "learning_rate": 4.922592864328673e-06, "loss": 0.0136, "step": 86470 }, { "grad_norm": 0.17338474094867706, "learning_rate": 4.915441115846436e-06, "loss": 0.0144, "step": 86480 }, { "grad_norm": 0.1365707367658615, "learning_rate": 4.908294297743782e-06, "loss": 0.0253, "step": 86490 }, { "grad_norm": 0.2588580846786499, "learning_rate": 4.901152410802273e-06, "loss": 0.0174, "step": 86500 }, { "grad_norm": 0.2508159577846527, "learning_rate": 4.894015455802919e-06, "loss": 0.0132, "step": 86510 }, { "grad_norm": 0.2871115505695343, "learning_rate": 4.8868834335262346e-06, "loss": 0.0155, "step": 86520 }, { "grad_norm": 0.4242245554924011, "learning_rate": 4.87975634475214e-06, "loss": 0.0184, "step": 86530 }, { "grad_norm": 0.5941216349601746, "learning_rate": 4.8726341902600614e-06, "loss": 0.015, "step": 86540 }, { "grad_norm": 0.242378830909729, "learning_rate": 4.865516970828843e-06, "loss": 0.0179, "step": 86550 }, { "grad_norm": 0.3771042227745056, "learning_rate": 4.85840468723685e-06, "loss": 0.0174, "step": 86560 }, { "grad_norm": 0.5062342286109924, "learning_rate": 4.85129734026184e-06, "loss": 0.0128, "step": 86570 }, { "grad_norm": 0.34537339210510254, "learning_rate": 4.844194930681073e-06, "loss": 0.0128, "step": 86580 }, { "grad_norm": 0.28836652636528015, "learning_rate": 4.837097459271267e-06, "loss": 0.0143, "step": 86590 }, { "grad_norm": 0.2917678654193878, "learning_rate": 4.830004926808573e-06, "loss": 0.013, "step": 86600 }, { "grad_norm": 0.32467448711395264, "learning_rate": 4.822917334068627e-06, "loss": 0.018, "step": 86610 }, { "grad_norm": 0.34060442447662354, "learning_rate": 4.81583468182652e-06, "loss": 0.014, "step": 86620 }, { "grad_norm": 0.208634614944458, "learning_rate": 4.808756970856809e-06, "loss": 0.016, "step": 86630 }, { "grad_norm": 0.45672979950904846, "learning_rate": 4.801684201933476e-06, "loss": 0.0163, "step": 86640 }, { "grad_norm": 0.4301897883415222, "learning_rate": 4.794616375830019e-06, "loss": 0.0117, "step": 86650 }, { "grad_norm": 0.3812781870365143, "learning_rate": 4.787553493319341e-06, "loss": 0.0124, "step": 86660 }, { "grad_norm": 0.292897492647171, "learning_rate": 4.780495555173836e-06, "loss": 0.0152, "step": 86670 }, { "grad_norm": 0.2602386772632599, "learning_rate": 4.7734425621653475e-06, "loss": 0.02, "step": 86680 }, { "grad_norm": 0.13923023641109467, "learning_rate": 4.766394515065198e-06, "loss": 0.0156, "step": 86690 }, { "grad_norm": 0.4532676339149475, "learning_rate": 4.75935141464412e-06, "loss": 0.0195, "step": 86700 }, { "grad_norm": 0.7204634547233582, "learning_rate": 4.75231326167237e-06, "loss": 0.0154, "step": 86710 }, { "grad_norm": 0.32046326994895935, "learning_rate": 4.745280056919599e-06, "loss": 0.0135, "step": 86720 }, { "grad_norm": 0.5108964443206787, "learning_rate": 4.738251801154975e-06, "loss": 0.0167, "step": 86730 }, { "grad_norm": 0.28654325008392334, "learning_rate": 4.731228495147061e-06, "loss": 0.0217, "step": 86740 }, { "grad_norm": 0.4255617558956146, "learning_rate": 4.7242101396639595e-06, "loss": 0.0152, "step": 86750 }, { "grad_norm": 0.18612030148506165, "learning_rate": 4.717196735473151e-06, "loss": 0.0137, "step": 86760 }, { "grad_norm": 0.27564650774002075, "learning_rate": 4.710188283341643e-06, "loss": 0.016, "step": 86770 }, { "grad_norm": 0.6216149926185608, "learning_rate": 4.703184784035835e-06, "loss": 0.0191, "step": 86780 }, { "grad_norm": 0.15405908226966858, "learning_rate": 4.696186238321642e-06, "loss": 0.0184, "step": 86790 }, { "grad_norm": 0.13857373595237732, "learning_rate": 4.689192646964408e-06, "loss": 0.0136, "step": 86800 }, { "grad_norm": 0.3883258104324341, "learning_rate": 4.682204010728952e-06, "loss": 0.0195, "step": 86810 }, { "grad_norm": 0.7082752585411072, "learning_rate": 4.67522033037952e-06, "loss": 0.0179, "step": 86820 }, { "grad_norm": 0.29064393043518066, "learning_rate": 4.668241606679846e-06, "loss": 0.0167, "step": 86830 }, { "grad_norm": 0.24795584380626678, "learning_rate": 4.66126784039313e-06, "loss": 0.0119, "step": 86840 }, { "grad_norm": 0.40900900959968567, "learning_rate": 4.6542990322819836e-06, "loss": 0.0175, "step": 86850 }, { "grad_norm": 0.5621054768562317, "learning_rate": 4.64733518310852e-06, "loss": 0.0157, "step": 86860 }, { "grad_norm": 0.253322958946228, "learning_rate": 4.640376293634291e-06, "loss": 0.0112, "step": 86870 }, { "grad_norm": 0.3054920732975006, "learning_rate": 4.633422364620327e-06, "loss": 0.0147, "step": 86880 }, { "grad_norm": 0.21318738162517548, "learning_rate": 4.6264733968270745e-06, "loss": 0.0183, "step": 86890 }, { "grad_norm": 0.38710251450538635, "learning_rate": 4.619529391014482e-06, "loss": 0.0211, "step": 86900 }, { "grad_norm": 0.41362205147743225, "learning_rate": 4.61259034794192e-06, "loss": 0.0138, "step": 86910 }, { "grad_norm": 0.6118524074554443, "learning_rate": 4.605656268368247e-06, "loss": 0.0193, "step": 86920 }, { "grad_norm": 0.31642821431159973, "learning_rate": 4.598727153051735e-06, "loss": 0.0202, "step": 86930 }, { "grad_norm": 0.19176121056079865, "learning_rate": 4.591803002750178e-06, "loss": 0.0173, "step": 86940 }, { "grad_norm": 0.34126999974250793, "learning_rate": 4.58488381822077e-06, "loss": 0.015, "step": 86950 }, { "grad_norm": 0.3121556043624878, "learning_rate": 4.5779696002201946e-06, "loss": 0.014, "step": 86960 }, { "grad_norm": 1.1739078760147095, "learning_rate": 4.571060349504563e-06, "loss": 0.0144, "step": 86970 }, { "grad_norm": 0.524107813835144, "learning_rate": 4.564156066829478e-06, "loss": 0.0174, "step": 86980 }, { "grad_norm": 0.5137861967086792, "learning_rate": 4.557256752949957e-06, "loss": 0.0133, "step": 86990 }, { "grad_norm": 0.34092631936073303, "learning_rate": 4.550362408620534e-06, "loss": 0.0162, "step": 87000 }, { "grad_norm": 0.10902556031942368, "learning_rate": 4.543473034595136e-06, "loss": 0.0217, "step": 87010 }, { "grad_norm": 0.2708836495876312, "learning_rate": 4.536588631627192e-06, "loss": 0.0186, "step": 87020 }, { "grad_norm": 1.086566686630249, "learning_rate": 4.5297092004695565e-06, "loss": 0.0183, "step": 87030 }, { "grad_norm": 0.3132206201553345, "learning_rate": 4.522834741874555e-06, "loss": 0.0135, "step": 87040 }, { "grad_norm": 0.21683119237422943, "learning_rate": 4.515965256593974e-06, "loss": 0.014, "step": 87050 }, { "grad_norm": 0.19532649219036102, "learning_rate": 4.509100745379058e-06, "loss": 0.0163, "step": 87060 }, { "grad_norm": 0.15222448110580444, "learning_rate": 4.502241208980479e-06, "loss": 0.0109, "step": 87070 }, { "grad_norm": 0.2924923300743103, "learning_rate": 4.495386648148398e-06, "loss": 0.0196, "step": 87080 }, { "grad_norm": 0.38203874230384827, "learning_rate": 4.488537063632431e-06, "loss": 0.0158, "step": 87090 }, { "grad_norm": 0.34188467264175415, "learning_rate": 4.481692456181608e-06, "loss": 0.0152, "step": 87100 }, { "grad_norm": 0.30338719487190247, "learning_rate": 4.474852826544479e-06, "loss": 0.0155, "step": 87110 }, { "grad_norm": 0.21156206727027893, "learning_rate": 4.468018175468974e-06, "loss": 0.0167, "step": 87120 }, { "grad_norm": 0.3115413784980774, "learning_rate": 4.461188503702568e-06, "loss": 0.0201, "step": 87130 }, { "grad_norm": 0.29641929268836975, "learning_rate": 4.454363811992107e-06, "loss": 0.022, "step": 87140 }, { "grad_norm": 0.2405606210231781, "learning_rate": 4.447544101083956e-06, "loss": 0.0119, "step": 87150 }, { "grad_norm": 1.0993961095809937, "learning_rate": 4.440729371723884e-06, "loss": 0.0135, "step": 87160 }, { "grad_norm": 0.2890876829624176, "learning_rate": 4.433919624657162e-06, "loss": 0.0141, "step": 87170 }, { "grad_norm": 0.26052114367485046, "learning_rate": 4.427114860628462e-06, "loss": 0.0153, "step": 87180 }, { "grad_norm": 0.9155621528625488, "learning_rate": 4.420315080381982e-06, "loss": 0.0133, "step": 87190 }, { "grad_norm": 0.12243025004863739, "learning_rate": 4.4135202846613045e-06, "loss": 0.0189, "step": 87200 }, { "grad_norm": 0.43494826555252075, "learning_rate": 4.406730474209525e-06, "loss": 0.021, "step": 87210 }, { "grad_norm": 0.26812922954559326, "learning_rate": 4.399945649769138e-06, "loss": 0.0192, "step": 87220 }, { "grad_norm": 0.514728844165802, "learning_rate": 4.393165812082151e-06, "loss": 0.0199, "step": 87230 }, { "grad_norm": 0.49279850721359253, "learning_rate": 4.386390961889969e-06, "loss": 0.0184, "step": 87240 }, { "grad_norm": 0.7967134714126587, "learning_rate": 4.3796210999334895e-06, "loss": 0.0211, "step": 87250 }, { "grad_norm": 0.601273775100708, "learning_rate": 4.372856226953065e-06, "loss": 0.015, "step": 87260 }, { "grad_norm": 0.22706708312034607, "learning_rate": 4.366096343688486e-06, "loss": 0.0122, "step": 87270 }, { "grad_norm": 0.4397861957550049, "learning_rate": 4.359341450878995e-06, "loss": 0.0165, "step": 87280 }, { "grad_norm": 0.38174521923065186, "learning_rate": 4.352591549263302e-06, "loss": 0.0198, "step": 87290 }, { "grad_norm": 0.36158204078674316, "learning_rate": 4.345846639579582e-06, "loss": 0.0132, "step": 87300 }, { "grad_norm": 0.265960693359375, "learning_rate": 4.339106722565417e-06, "loss": 0.0226, "step": 87310 }, { "grad_norm": 0.41402915120124817, "learning_rate": 4.332371798957896e-06, "loss": 0.0185, "step": 87320 }, { "grad_norm": 0.314003050327301, "learning_rate": 4.3256418694935295e-06, "loss": 0.0184, "step": 87330 }, { "grad_norm": 0.6377984881401062, "learning_rate": 4.3189169349083125e-06, "loss": 0.0152, "step": 87340 }, { "grad_norm": 0.2943277060985565, "learning_rate": 4.312196995937645e-06, "loss": 0.0146, "step": 87350 }, { "grad_norm": 0.7674372792243958, "learning_rate": 4.305482053316434e-06, "loss": 0.0168, "step": 87360 }, { "grad_norm": 0.32715779542922974, "learning_rate": 4.2987721077789925e-06, "loss": 0.0188, "step": 87370 }, { "grad_norm": 0.1700495332479477, "learning_rate": 4.2920671600591325e-06, "loss": 0.02, "step": 87380 }, { "grad_norm": 0.27380579710006714, "learning_rate": 4.285367210890084e-06, "loss": 0.0165, "step": 87390 }, { "grad_norm": 0.3544623553752899, "learning_rate": 4.278672261004551e-06, "loss": 0.0178, "step": 87400 }, { "grad_norm": 0.30675122141838074, "learning_rate": 4.2719823111346704e-06, "loss": 0.0105, "step": 87410 }, { "grad_norm": 0.5730234980583191, "learning_rate": 4.265297362012061e-06, "loss": 0.0183, "step": 87420 }, { "grad_norm": 1.0875442028045654, "learning_rate": 4.258617414367766e-06, "loss": 0.0173, "step": 87430 }, { "grad_norm": 0.1956472098827362, "learning_rate": 4.251942468932296e-06, "loss": 0.0151, "step": 87440 }, { "grad_norm": 0.21854327619075775, "learning_rate": 4.2452725264356175e-06, "loss": 0.0228, "step": 87450 }, { "grad_norm": 0.4278939664363861, "learning_rate": 4.2386075876071575e-06, "loss": 0.0166, "step": 87460 }, { "grad_norm": 0.2798621654510498, "learning_rate": 4.231947653175755e-06, "loss": 0.0158, "step": 87470 }, { "grad_norm": 0.4354178011417389, "learning_rate": 4.225292723869761e-06, "loss": 0.0176, "step": 87480 }, { "grad_norm": 0.452776700258255, "learning_rate": 4.218642800416916e-06, "loss": 0.0132, "step": 87490 }, { "grad_norm": 0.31128746271133423, "learning_rate": 4.211997883544472e-06, "loss": 0.0104, "step": 87500 }, { "grad_norm": 0.26268261671066284, "learning_rate": 4.205357973979096e-06, "loss": 0.0185, "step": 87510 }, { "grad_norm": 0.1512695550918579, "learning_rate": 4.198723072446925e-06, "loss": 0.0198, "step": 87520 }, { "grad_norm": 0.4995768964290619, "learning_rate": 4.192093179673534e-06, "loss": 0.0192, "step": 87530 }, { "grad_norm": 0.294399619102478, "learning_rate": 4.185468296383959e-06, "loss": 0.016, "step": 87540 }, { "grad_norm": 0.8018308281898499, "learning_rate": 4.1788484233027005e-06, "loss": 0.0125, "step": 87550 }, { "grad_norm": 0.5215169787406921, "learning_rate": 4.172233561153677e-06, "loss": 0.0175, "step": 87560 }, { "grad_norm": 1.0773404836654663, "learning_rate": 4.165623710660288e-06, "loss": 0.0176, "step": 87570 }, { "grad_norm": 0.37513571977615356, "learning_rate": 4.1590188725453835e-06, "loss": 0.0135, "step": 87580 }, { "grad_norm": 0.2737021744251251, "learning_rate": 4.152419047531259e-06, "loss": 0.0154, "step": 87590 }, { "grad_norm": 0.23569215834140778, "learning_rate": 4.145824236339651e-06, "loss": 0.011, "step": 87600 }, { "grad_norm": 0.18094022572040558, "learning_rate": 4.1392344396917695e-06, "loss": 0.0171, "step": 87610 }, { "grad_norm": 0.42586925625801086, "learning_rate": 4.132649658308252e-06, "loss": 0.015, "step": 87620 }, { "grad_norm": 0.24551045894622803, "learning_rate": 4.126069892909213e-06, "loss": 0.0146, "step": 87630 }, { "grad_norm": 0.16210748255252838, "learning_rate": 4.119495144214191e-06, "loss": 0.0189, "step": 87640 }, { "grad_norm": 0.33875808119773865, "learning_rate": 4.112925412942214e-06, "loss": 0.0161, "step": 87650 }, { "grad_norm": 0.24523001909255981, "learning_rate": 4.106360699811712e-06, "loss": 0.0167, "step": 87660 }, { "grad_norm": 0.3053230345249176, "learning_rate": 4.099801005540616e-06, "loss": 0.0138, "step": 87670 }, { "grad_norm": 0.26519057154655457, "learning_rate": 4.093246330846256e-06, "loss": 0.0123, "step": 87680 }, { "grad_norm": 0.3253574073314667, "learning_rate": 4.086696676445467e-06, "loss": 0.0166, "step": 87690 }, { "grad_norm": 0.31821730732917786, "learning_rate": 4.0801520430544914e-06, "loss": 0.017, "step": 87700 }, { "grad_norm": 0.18287768959999084, "learning_rate": 4.073612431389062e-06, "loss": 0.0179, "step": 87710 }, { "grad_norm": 0.26456791162490845, "learning_rate": 4.067077842164319e-06, "loss": 0.0163, "step": 87720 }, { "grad_norm": 0.35605713725090027, "learning_rate": 4.0605482760948945e-06, "loss": 0.0178, "step": 87730 }, { "grad_norm": 0.24452602863311768, "learning_rate": 4.054023733894829e-06, "loss": 0.0137, "step": 87740 }, { "grad_norm": 0.34681200981140137, "learning_rate": 4.047504216277653e-06, "loss": 0.0179, "step": 87750 }, { "grad_norm": 0.5697442293167114, "learning_rate": 4.040989723956323e-06, "loss": 0.0184, "step": 87760 }, { "grad_norm": 0.39899522066116333, "learning_rate": 4.03448025764327e-06, "loss": 0.0163, "step": 87770 }, { "grad_norm": 0.8574244976043701, "learning_rate": 4.027975818050339e-06, "loss": 0.0159, "step": 87780 }, { "grad_norm": 0.3414863348007202, "learning_rate": 4.021476405888858e-06, "loss": 0.0191, "step": 87790 }, { "grad_norm": 0.28804680705070496, "learning_rate": 4.01498202186959e-06, "loss": 0.0197, "step": 87800 }, { "grad_norm": 0.3487674593925476, "learning_rate": 4.008492666702751e-06, "loss": 0.0148, "step": 87810 }, { "grad_norm": 0.26403793692588806, "learning_rate": 4.002008341098001e-06, "loss": 0.0179, "step": 87820 }, { "grad_norm": 0.201902374625206, "learning_rate": 3.995529045764462e-06, "loss": 0.0161, "step": 87830 }, { "grad_norm": 0.3383968770503998, "learning_rate": 3.989054781410712e-06, "loss": 0.0141, "step": 87840 }, { "grad_norm": 0.16148409247398376, "learning_rate": 3.982585548744744e-06, "loss": 0.0131, "step": 87850 }, { "grad_norm": 0.3535839319229126, "learning_rate": 3.9761213484740435e-06, "loss": 0.01, "step": 87860 }, { "grad_norm": 0.2881558835506439, "learning_rate": 3.9696621813055055e-06, "loss": 0.0159, "step": 87870 }, { "grad_norm": 0.4514019787311554, "learning_rate": 3.9632080479455044e-06, "loss": 0.0204, "step": 87880 }, { "grad_norm": 0.40669703483581543, "learning_rate": 3.956758949099859e-06, "loss": 0.0178, "step": 87890 }, { "grad_norm": 0.5531805157661438, "learning_rate": 3.9503148854738385e-06, "loss": 0.0163, "step": 87900 }, { "grad_norm": 0.36134371161460876, "learning_rate": 3.943875857772139e-06, "loss": 0.0329, "step": 87910 }, { "grad_norm": 0.34414437413215637, "learning_rate": 3.9374418666989375e-06, "loss": 0.0152, "step": 87920 }, { "grad_norm": 0.14111098647117615, "learning_rate": 3.931012912957838e-06, "loss": 0.0197, "step": 87930 }, { "grad_norm": 0.2784915864467621, "learning_rate": 3.9245889972519045e-06, "loss": 0.0209, "step": 87940 }, { "grad_norm": 0.2140057384967804, "learning_rate": 3.9181701202836265e-06, "loss": 0.014, "step": 87950 }, { "grad_norm": 0.20495203137397766, "learning_rate": 3.911756282755003e-06, "loss": 0.019, "step": 87960 }, { "grad_norm": 0.24839143455028534, "learning_rate": 3.905347485367411e-06, "loss": 0.0119, "step": 87970 }, { "grad_norm": 0.5739784240722656, "learning_rate": 3.898943728821725e-06, "loss": 0.0155, "step": 87980 }, { "grad_norm": 0.2777082324028015, "learning_rate": 3.892545013818227e-06, "loss": 0.015, "step": 87990 }, { "grad_norm": 0.316536545753479, "learning_rate": 3.886151341056687e-06, "loss": 0.0115, "step": 88000 }, { "grad_norm": 0.2966426908969879, "learning_rate": 3.8797627112363066e-06, "loss": 0.0204, "step": 88010 }, { "grad_norm": 0.1047966331243515, "learning_rate": 3.873379125055743e-06, "loss": 0.0141, "step": 88020 }, { "grad_norm": 0.2724635899066925, "learning_rate": 3.867000583213082e-06, "loss": 0.0159, "step": 88030 }, { "grad_norm": 0.42089658975601196, "learning_rate": 3.860627086405871e-06, "loss": 0.0157, "step": 88040 }, { "grad_norm": 0.5207192301750183, "learning_rate": 3.854258635331126e-06, "loss": 0.0144, "step": 88050 }, { "grad_norm": 0.28467658162117004, "learning_rate": 3.8478952306852665e-06, "loss": 0.0192, "step": 88060 }, { "grad_norm": 0.3210878074169159, "learning_rate": 3.841536873164198e-06, "loss": 0.0147, "step": 88070 }, { "grad_norm": 0.624985933303833, "learning_rate": 3.8351835634632626e-06, "loss": 0.0186, "step": 88080 }, { "grad_norm": 0.23494373261928558, "learning_rate": 3.828835302277245e-06, "loss": 0.0218, "step": 88090 }, { "grad_norm": 0.25528207421302795, "learning_rate": 3.822492090300378e-06, "loss": 0.0133, "step": 88100 }, { "grad_norm": 0.3245616853237152, "learning_rate": 3.8161539282263566e-06, "loss": 0.0143, "step": 88110 }, { "grad_norm": 0.3291005492210388, "learning_rate": 3.8098208167482995e-06, "loss": 0.013, "step": 88120 }, { "grad_norm": 0.38336417078971863, "learning_rate": 3.803492756558796e-06, "loss": 0.0108, "step": 88130 }, { "grad_norm": 0.3132772743701935, "learning_rate": 3.797169748349855e-06, "loss": 0.0127, "step": 88140 }, { "grad_norm": 0.27069029211997986, "learning_rate": 3.790851792812977e-06, "loss": 0.0152, "step": 88150 }, { "grad_norm": 0.29030555486679077, "learning_rate": 3.784538890639072e-06, "loss": 0.0174, "step": 88160 }, { "grad_norm": 0.2838139235973358, "learning_rate": 3.7782310425185153e-06, "loss": 0.0174, "step": 88170 }, { "grad_norm": 0.9331992864608765, "learning_rate": 3.77192824914111e-06, "loss": 0.0145, "step": 88180 }, { "grad_norm": 0.37853115797042847, "learning_rate": 3.7656305111961333e-06, "loss": 0.0149, "step": 88190 }, { "grad_norm": 0.723330557346344, "learning_rate": 3.7593378293722782e-06, "loss": 0.0216, "step": 88200 }, { "grad_norm": 0.5289313197135925, "learning_rate": 3.7530502043577333e-06, "loss": 0.0194, "step": 88210 }, { "grad_norm": 0.34435224533081055, "learning_rate": 3.746767636840076e-06, "loss": 0.0189, "step": 88220 }, { "grad_norm": 0.27747079730033875, "learning_rate": 3.74049012750638e-06, "loss": 0.0144, "step": 88230 }, { "grad_norm": 0.5245542526245117, "learning_rate": 3.734217677043128e-06, "loss": 0.0145, "step": 88240 }, { "grad_norm": 0.31895795464515686, "learning_rate": 3.727950286136267e-06, "loss": 0.0121, "step": 88250 }, { "grad_norm": 0.9615259766578674, "learning_rate": 3.7216879554711925e-06, "loss": 0.0173, "step": 88260 }, { "grad_norm": 0.2799050509929657, "learning_rate": 3.715430685732757e-06, "loss": 0.0163, "step": 88270 }, { "grad_norm": 0.14383062720298767, "learning_rate": 3.7091784776052243e-06, "loss": 0.0128, "step": 88280 }, { "grad_norm": 0.530828595161438, "learning_rate": 3.7029313317723356e-06, "loss": 0.0188, "step": 88290 }, { "grad_norm": 0.435851126909256, "learning_rate": 3.6966892489172787e-06, "loss": 0.0197, "step": 88300 }, { "grad_norm": 0.17902877926826477, "learning_rate": 3.6904522297226575e-06, "loss": 0.0152, "step": 88310 }, { "grad_norm": 0.519771933555603, "learning_rate": 3.6842202748705647e-06, "loss": 0.0172, "step": 88320 }, { "grad_norm": 0.20967665314674377, "learning_rate": 3.677993385042494e-06, "loss": 0.0166, "step": 88330 }, { "grad_norm": 0.23018601536750793, "learning_rate": 3.67177156091944e-06, "loss": 0.0169, "step": 88340 }, { "grad_norm": 0.12534485757350922, "learning_rate": 3.6655548031817866e-06, "loss": 0.0176, "step": 88350 }, { "grad_norm": 0.18040306866168976, "learning_rate": 3.6593431125094e-06, "loss": 0.0144, "step": 88360 }, { "grad_norm": 0.18173573911190033, "learning_rate": 3.6531364895815766e-06, "loss": 0.015, "step": 88370 }, { "grad_norm": 0.17187079787254333, "learning_rate": 3.6469349350770677e-06, "loss": 0.0129, "step": 88380 }, { "grad_norm": 0.29854199290275574, "learning_rate": 3.6407384496740525e-06, "loss": 0.014, "step": 88390 }, { "grad_norm": 0.4643423855304718, "learning_rate": 3.6345470340501954e-06, "loss": 0.0109, "step": 88400 }, { "grad_norm": 0.4632202684879303, "learning_rate": 3.6283606888825537e-06, "loss": 0.0153, "step": 88410 }, { "grad_norm": 0.16389437019824982, "learning_rate": 3.622179414847676e-06, "loss": 0.0119, "step": 88420 }, { "grad_norm": 0.48453861474990845, "learning_rate": 3.616003212621527e-06, "loss": 0.0196, "step": 88430 }, { "grad_norm": 0.7012900114059448, "learning_rate": 3.609832082879533e-06, "loss": 0.0152, "step": 88440 }, { "grad_norm": 0.27017492055892944, "learning_rate": 3.6036660262965485e-06, "loss": 0.0137, "step": 88450 }, { "grad_norm": 0.2726426124572754, "learning_rate": 3.597505043546895e-06, "loss": 0.0124, "step": 88460 }, { "grad_norm": 0.3457779586315155, "learning_rate": 3.5913491353043227e-06, "loss": 0.0156, "step": 88470 }, { "grad_norm": 0.43187761306762695, "learning_rate": 3.5851983022420375e-06, "loss": 0.0148, "step": 88480 }, { "grad_norm": 0.3189998269081116, "learning_rate": 3.5790525450326797e-06, "loss": 0.0166, "step": 88490 }, { "grad_norm": 0.22962602972984314, "learning_rate": 3.572911864348344e-06, "loss": 0.016, "step": 88500 }, { "grad_norm": 0.3459840416908264, "learning_rate": 3.5667762608605717e-06, "loss": 0.0107, "step": 88510 }, { "grad_norm": 0.2655686140060425, "learning_rate": 3.5606457352403255e-06, "loss": 0.0129, "step": 88520 }, { "grad_norm": 0.17387451231479645, "learning_rate": 3.554520288158042e-06, "loss": 0.0155, "step": 88530 }, { "grad_norm": 0.18890386819839478, "learning_rate": 3.5483999202835903e-06, "loss": 0.0134, "step": 88540 }, { "grad_norm": 0.2037760317325592, "learning_rate": 3.5422846322862913e-06, "loss": 0.0152, "step": 88550 }, { "grad_norm": 0.3307599127292633, "learning_rate": 3.5361744248348925e-06, "loss": 0.0158, "step": 88560 }, { "grad_norm": 0.41990476846694946, "learning_rate": 3.5300692985976047e-06, "loss": 0.0189, "step": 88570 }, { "grad_norm": 0.5845560431480408, "learning_rate": 3.52396925424206e-06, "loss": 0.0109, "step": 88580 }, { "grad_norm": 0.18997453153133392, "learning_rate": 3.5178742924353747e-06, "loss": 0.0165, "step": 88590 }, { "grad_norm": 0.2714370787143707, "learning_rate": 3.511784413844066e-06, "loss": 0.0192, "step": 88600 }, { "grad_norm": 0.1385234296321869, "learning_rate": 3.5056996191341287e-06, "loss": 0.0124, "step": 88610 }, { "grad_norm": 1.300540804862976, "learning_rate": 3.4996199089709692e-06, "loss": 0.0165, "step": 88620 }, { "grad_norm": 0.20471073687076569, "learning_rate": 3.4935452840194725e-06, "loss": 0.0148, "step": 88630 }, { "grad_norm": 0.8906956315040588, "learning_rate": 3.48747574494393e-06, "loss": 0.0182, "step": 88640 }, { "grad_norm": 0.37248679995536804, "learning_rate": 3.4814112924081154e-06, "loss": 0.0156, "step": 88650 }, { "grad_norm": 0.24615170061588287, "learning_rate": 3.475351927075221e-06, "loss": 0.0157, "step": 88660 }, { "grad_norm": 0.4695296883583069, "learning_rate": 3.469297649607894e-06, "loss": 0.0196, "step": 88670 }, { "grad_norm": 0.17641255259513855, "learning_rate": 3.463248460668217e-06, "loss": 0.011, "step": 88680 }, { "grad_norm": 0.3131196200847626, "learning_rate": 3.457204360917726e-06, "loss": 0.0159, "step": 88690 }, { "grad_norm": 0.2083476036787033, "learning_rate": 3.4511653510173827e-06, "loss": 0.0132, "step": 88700 }, { "grad_norm": 0.35296934843063354, "learning_rate": 3.4451314316276074e-06, "loss": 0.0171, "step": 88710 }, { "grad_norm": 0.6616559624671936, "learning_rate": 3.4391026034082674e-06, "loss": 0.0192, "step": 88720 }, { "grad_norm": 0.3781808018684387, "learning_rate": 3.433078867018674e-06, "loss": 0.0121, "step": 88730 }, { "grad_norm": 0.3646094501018524, "learning_rate": 3.427060223117551e-06, "loss": 0.0151, "step": 88740 }, { "grad_norm": 0.5573969483375549, "learning_rate": 3.4210466723631042e-06, "loss": 0.0133, "step": 88750 }, { "grad_norm": 0.21167445182800293, "learning_rate": 3.415038215412969e-06, "loss": 0.0134, "step": 88760 }, { "grad_norm": 0.38951313495635986, "learning_rate": 3.4090348529242088e-06, "loss": 0.0175, "step": 88770 }, { "grad_norm": 0.28260520100593567, "learning_rate": 3.4030365855533487e-06, "loss": 0.0189, "step": 88780 }, { "grad_norm": 0.4726153612136841, "learning_rate": 3.397043413956347e-06, "loss": 0.0199, "step": 88790 }, { "grad_norm": 0.3680718243122101, "learning_rate": 3.3910553387886237e-06, "loss": 0.0148, "step": 88800 }, { "grad_norm": 0.4170437753200531, "learning_rate": 3.3850723607049996e-06, "loss": 0.0116, "step": 88810 }, { "grad_norm": 0.5560464262962341, "learning_rate": 3.3790944803597902e-06, "loss": 0.0157, "step": 88820 }, { "grad_norm": 1.0916485786437988, "learning_rate": 3.373121698406706e-06, "loss": 0.018, "step": 88830 }, { "grad_norm": 1.0585497617721558, "learning_rate": 3.3671540154989302e-06, "loss": 0.0163, "step": 88840 }, { "grad_norm": 0.20584890246391296, "learning_rate": 3.361191432289079e-06, "loss": 0.0182, "step": 88850 }, { "grad_norm": 0.15885071456432343, "learning_rate": 3.3552339494292207e-06, "loss": 0.0175, "step": 88860 }, { "grad_norm": 0.2501429319381714, "learning_rate": 3.349281567570839e-06, "loss": 0.0135, "step": 88870 }, { "grad_norm": 0.24333658814430237, "learning_rate": 3.3433342873648965e-06, "loss": 0.0177, "step": 88880 }, { "grad_norm": 0.3639492392539978, "learning_rate": 3.337392109461762e-06, "loss": 0.011, "step": 88890 }, { "grad_norm": 0.5018302202224731, "learning_rate": 3.331455034511266e-06, "loss": 0.0166, "step": 88900 }, { "grad_norm": 0.8892924785614014, "learning_rate": 3.325523063162689e-06, "loss": 0.0138, "step": 88910 }, { "grad_norm": 0.3846985399723053, "learning_rate": 3.3195961960647393e-06, "loss": 0.0114, "step": 88920 }, { "grad_norm": 0.323164165019989, "learning_rate": 3.31367443386556e-06, "loss": 0.0161, "step": 88930 }, { "grad_norm": 0.22604624927043915, "learning_rate": 3.3077577772127556e-06, "loss": 0.0112, "step": 88940 }, { "grad_norm": 0.23977775871753693, "learning_rate": 3.301846226753358e-06, "loss": 0.021, "step": 88950 }, { "grad_norm": 0.35843634605407715, "learning_rate": 3.2959397831338444e-06, "loss": 0.0169, "step": 88960 }, { "grad_norm": 0.4927234649658203, "learning_rate": 3.2900384470001432e-06, "loss": 0.0155, "step": 88970 }, { "grad_norm": 0.2111567109823227, "learning_rate": 3.28414221899761e-06, "loss": 0.0111, "step": 88980 }, { "grad_norm": 0.4037220776081085, "learning_rate": 3.278251099771046e-06, "loss": 0.0174, "step": 88990 }, { "grad_norm": 0.1572311967611313, "learning_rate": 3.2723650899646906e-06, "loss": 0.0204, "step": 89000 }, { "grad_norm": 0.24367235600948334, "learning_rate": 3.2664841902222466e-06, "loss": 0.0152, "step": 89010 }, { "grad_norm": 0.39980828762054443, "learning_rate": 3.2606084011868208e-06, "loss": 0.0157, "step": 89020 }, { "grad_norm": 0.25525039434432983, "learning_rate": 3.254737723500989e-06, "loss": 0.0155, "step": 89030 }, { "grad_norm": 0.6648294925689697, "learning_rate": 3.2488721578067595e-06, "loss": 0.0184, "step": 89040 }, { "grad_norm": 0.2817060947418213, "learning_rate": 3.243011704745591e-06, "loss": 0.0201, "step": 89050 }, { "grad_norm": 0.571397066116333, "learning_rate": 3.237156364958355e-06, "loss": 0.0159, "step": 89060 }, { "grad_norm": 0.38342171907424927, "learning_rate": 3.231306139085394e-06, "loss": 0.0166, "step": 89070 }, { "grad_norm": 0.18648888170719147, "learning_rate": 3.2254610277664797e-06, "loss": 0.015, "step": 89080 }, { "grad_norm": 0.48886510729789734, "learning_rate": 3.2196210316408183e-06, "loss": 0.0153, "step": 89090 }, { "grad_norm": 0.4248339831829071, "learning_rate": 3.2137861513470644e-06, "loss": 0.0224, "step": 89100 }, { "grad_norm": 0.46262454986572266, "learning_rate": 3.2079563875233252e-06, "loss": 0.0158, "step": 89110 }, { "grad_norm": 0.20013689994812012, "learning_rate": 3.2021317408071184e-06, "loss": 0.0142, "step": 89120 }, { "grad_norm": 0.2972167134284973, "learning_rate": 3.196312211835434e-06, "loss": 0.0156, "step": 89130 }, { "grad_norm": 0.22818440198898315, "learning_rate": 3.1904978012446686e-06, "loss": 0.0173, "step": 89140 }, { "grad_norm": 0.3026304543018341, "learning_rate": 3.184688509670691e-06, "loss": 0.0172, "step": 89150 }, { "grad_norm": 0.25503239035606384, "learning_rate": 3.178884337748783e-06, "loss": 0.0126, "step": 89160 }, { "grad_norm": 0.2489001303911209, "learning_rate": 3.173085286113697e-06, "loss": 0.0183, "step": 89170 }, { "grad_norm": 0.5291799902915955, "learning_rate": 3.1672913553995986e-06, "loss": 0.0131, "step": 89180 }, { "grad_norm": 0.14647497236728668, "learning_rate": 3.1615025462401138e-06, "loss": 0.0102, "step": 89190 }, { "grad_norm": 0.2557586133480072, "learning_rate": 3.1557188592682808e-06, "loss": 0.0164, "step": 89200 }, { "grad_norm": 0.2963028848171234, "learning_rate": 3.1499402951166045e-06, "loss": 0.0154, "step": 89210 }, { "grad_norm": 0.32369786500930786, "learning_rate": 3.144166854417019e-06, "loss": 0.0168, "step": 89220 }, { "grad_norm": 0.32347795367240906, "learning_rate": 3.1383985378009073e-06, "loss": 0.0141, "step": 89230 }, { "grad_norm": 0.43254417181015015, "learning_rate": 3.1326353458990654e-06, "loss": 0.0122, "step": 89240 }, { "grad_norm": 1.4430824518203735, "learning_rate": 3.126877279341761e-06, "loss": 0.0136, "step": 89250 }, { "grad_norm": 0.4761173129081726, "learning_rate": 3.1211243387586964e-06, "loss": 0.0195, "step": 89260 }, { "grad_norm": 0.5637298822402954, "learning_rate": 3.115376524778979e-06, "loss": 0.0155, "step": 89270 }, { "grad_norm": 0.1795882135629654, "learning_rate": 3.109633838031195e-06, "loss": 0.0222, "step": 89280 }, { "grad_norm": 0.7436407804489136, "learning_rate": 3.103896279143359e-06, "loss": 0.0177, "step": 89290 }, { "grad_norm": 0.5527214407920837, "learning_rate": 3.098163848742924e-06, "loss": 0.0186, "step": 89300 }, { "grad_norm": 0.23344330489635468, "learning_rate": 3.0924365474567675e-06, "loss": 0.0192, "step": 89310 }, { "grad_norm": 0.3091360926628113, "learning_rate": 3.086714375911237e-06, "loss": 0.0174, "step": 89320 }, { "grad_norm": 0.6113858819007874, "learning_rate": 3.0809973347320775e-06, "loss": 0.0293, "step": 89330 }, { "grad_norm": 0.5432116389274597, "learning_rate": 3.075285424544516e-06, "loss": 0.017, "step": 89340 }, { "grad_norm": 0.22216202318668365, "learning_rate": 3.069578645973176e-06, "loss": 0.0121, "step": 89350 }, { "grad_norm": 0.24676550924777985, "learning_rate": 3.0638769996421735e-06, "loss": 0.0135, "step": 89360 }, { "grad_norm": 0.5510739088058472, "learning_rate": 3.058180486175005e-06, "loss": 0.0184, "step": 89370 }, { "grad_norm": 0.23877862095832825, "learning_rate": 3.052489106194645e-06, "loss": 0.0112, "step": 89380 }, { "grad_norm": 0.33920544385910034, "learning_rate": 3.0468028603234946e-06, "loss": 0.0144, "step": 89390 }, { "grad_norm": 0.11900807917118073, "learning_rate": 3.04112174918339e-06, "loss": 0.0143, "step": 89400 }, { "grad_norm": 0.31658366322517395, "learning_rate": 3.0354457733955954e-06, "loss": 0.0126, "step": 89410 }, { "grad_norm": 0.24051812291145325, "learning_rate": 3.029774933580859e-06, "loss": 0.0107, "step": 89420 }, { "grad_norm": 0.4096915125846863, "learning_rate": 3.0241092303593065e-06, "loss": 0.0118, "step": 89430 }, { "grad_norm": 0.3885928988456726, "learning_rate": 3.018448664350554e-06, "loss": 0.0203, "step": 89440 }, { "grad_norm": 0.21067841351032257, "learning_rate": 3.0127932361736055e-06, "loss": 0.013, "step": 89450 }, { "grad_norm": 1.558313012123108, "learning_rate": 3.00714294644695e-06, "loss": 0.0121, "step": 89460 }, { "grad_norm": 0.3260594606399536, "learning_rate": 3.0014977957884828e-06, "loss": 0.018, "step": 89470 }, { "grad_norm": 0.3067532777786255, "learning_rate": 2.99585778481557e-06, "loss": 0.016, "step": 89480 }, { "grad_norm": 0.26145562529563904, "learning_rate": 2.990222914144969e-06, "loss": 0.012, "step": 89490 }, { "grad_norm": 0.35856732726097107, "learning_rate": 2.984593184392914e-06, "loss": 0.0178, "step": 89500 }, { "grad_norm": 0.2912764251232147, "learning_rate": 2.978968596175069e-06, "loss": 0.0174, "step": 89510 }, { "grad_norm": 0.3190471827983856, "learning_rate": 2.9733491501065137e-06, "loss": 0.0132, "step": 89520 }, { "grad_norm": 0.22717924416065216, "learning_rate": 2.9677348468018005e-06, "loss": 0.0199, "step": 89530 }, { "grad_norm": 0.36676958203315735, "learning_rate": 2.9621256868748726e-06, "loss": 0.0117, "step": 89540 }, { "grad_norm": 0.12356717884540558, "learning_rate": 2.956521670939183e-06, "loss": 0.0126, "step": 89550 }, { "grad_norm": 0.24448269605636597, "learning_rate": 2.950922799607542e-06, "loss": 0.0199, "step": 89560 }, { "grad_norm": 0.4158083498477936, "learning_rate": 2.9453290734922535e-06, "loss": 0.0112, "step": 89570 }, { "grad_norm": 0.32132041454315186, "learning_rate": 2.939740493205029e-06, "loss": 0.025, "step": 89580 }, { "grad_norm": 0.8069343566894531, "learning_rate": 2.9341570593570345e-06, "loss": 0.0129, "step": 89590 }, { "grad_norm": 0.36363399028778076, "learning_rate": 2.928578772558843e-06, "loss": 0.018, "step": 89600 }, { "grad_norm": 0.34303387999534607, "learning_rate": 2.923005633420528e-06, "loss": 0.0161, "step": 89610 }, { "grad_norm": 0.21090364456176758, "learning_rate": 2.9174376425515237e-06, "loss": 0.0194, "step": 89620 }, { "grad_norm": 0.4058648943901062, "learning_rate": 2.91187480056076e-06, "loss": 0.0146, "step": 89630 }, { "grad_norm": 2.1132967472076416, "learning_rate": 2.9063171080565664e-06, "loss": 0.0171, "step": 89640 }, { "grad_norm": 0.15278851985931396, "learning_rate": 2.900764565646741e-06, "loss": 0.0152, "step": 89650 }, { "grad_norm": 0.7078412175178528, "learning_rate": 2.895217173938475e-06, "loss": 0.0201, "step": 89660 }, { "grad_norm": 0.17365552484989166, "learning_rate": 2.8896749335384445e-06, "loss": 0.0146, "step": 89670 }, { "grad_norm": 0.28249359130859375, "learning_rate": 2.884137845052731e-06, "loss": 0.0139, "step": 89680 }, { "grad_norm": 0.24721123278141022, "learning_rate": 2.8786059090868734e-06, "loss": 0.0127, "step": 89690 }, { "grad_norm": 0.3384774625301361, "learning_rate": 2.87307912624582e-06, "loss": 0.0181, "step": 89700 }, { "grad_norm": 0.33520743250846863, "learning_rate": 2.867557497133977e-06, "loss": 0.0157, "step": 89710 }, { "grad_norm": 0.24510106444358826, "learning_rate": 2.862041022355194e-06, "loss": 0.0214, "step": 89720 }, { "grad_norm": 0.28484514355659485, "learning_rate": 2.8565297025127226e-06, "loss": 0.0152, "step": 89730 }, { "grad_norm": 0.46176639199256897, "learning_rate": 2.8510235382092853e-06, "loss": 0.0127, "step": 89740 }, { "grad_norm": 0.3576415479183197, "learning_rate": 2.845522530047029e-06, "loss": 0.0144, "step": 89750 }, { "grad_norm": 1.8298639059066772, "learning_rate": 2.8400266786275387e-06, "loss": 0.0202, "step": 89760 }, { "grad_norm": 0.26284268498420715, "learning_rate": 2.834535984551817e-06, "loss": 0.0124, "step": 89770 }, { "grad_norm": 0.20697613060474396, "learning_rate": 2.8290504484203397e-06, "loss": 0.0339, "step": 89780 }, { "grad_norm": 0.5035728216171265, "learning_rate": 2.8235700708329705e-06, "loss": 0.0198, "step": 89790 }, { "grad_norm": 0.32680320739746094, "learning_rate": 2.818094852389058e-06, "loss": 0.0156, "step": 89800 }, { "grad_norm": 0.20386648178100586, "learning_rate": 2.8126247936873516e-06, "loss": 0.0186, "step": 89810 }, { "grad_norm": 0.4003221094608307, "learning_rate": 2.8071598953260614e-06, "loss": 0.0166, "step": 89820 }, { "grad_norm": 0.5102980136871338, "learning_rate": 2.8017001579028035e-06, "loss": 0.0187, "step": 89830 }, { "grad_norm": 0.3187549114227295, "learning_rate": 2.7962455820146617e-06, "loss": 0.0198, "step": 89840 }, { "grad_norm": 0.4889628291130066, "learning_rate": 2.7907961682581253e-06, "loss": 0.0142, "step": 89850 }, { "grad_norm": 0.381984680891037, "learning_rate": 2.7853519172291453e-06, "loss": 0.0149, "step": 89860 }, { "grad_norm": 0.5129235982894897, "learning_rate": 2.7799128295230904e-06, "loss": 0.0147, "step": 89870 }, { "grad_norm": 0.5578693747520447, "learning_rate": 2.774478905734784e-06, "loss": 0.0185, "step": 89880 }, { "grad_norm": 0.40395164489746094, "learning_rate": 2.7690501464584516e-06, "loss": 0.0151, "step": 89890 }, { "grad_norm": 0.4541506767272949, "learning_rate": 2.763626552287796e-06, "loss": 0.0159, "step": 89900 }, { "grad_norm": 1.374313235282898, "learning_rate": 2.7582081238159083e-06, "loss": 0.0171, "step": 89910 }, { "grad_norm": 0.7407733798027039, "learning_rate": 2.7527948616353548e-06, "loss": 0.0191, "step": 89920 }, { "grad_norm": 0.6808772087097168, "learning_rate": 2.7473867663381226e-06, "loss": 0.0122, "step": 89930 }, { "grad_norm": 0.2105264663696289, "learning_rate": 2.7419838385156383e-06, "loss": 0.012, "step": 89940 }, { "grad_norm": 0.33756566047668457, "learning_rate": 2.7365860787587407e-06, "loss": 0.0163, "step": 89950 }, { "grad_norm": 0.4551224112510681, "learning_rate": 2.7311934876577295e-06, "loss": 0.0179, "step": 89960 }, { "grad_norm": 0.3503491282463074, "learning_rate": 2.725806065802339e-06, "loss": 0.018, "step": 89970 }, { "grad_norm": 0.6858574748039246, "learning_rate": 2.720423813781714e-06, "loss": 0.0115, "step": 89980 }, { "grad_norm": 0.1617884635925293, "learning_rate": 2.715046732184462e-06, "loss": 0.0157, "step": 89990 }, { "grad_norm": 0.28924989700317383, "learning_rate": 2.7096748215986013e-06, "loss": 0.0176, "step": 90000 }, { "grad_norm": 0.13573147356510162, "learning_rate": 2.704308082611612e-06, "loss": 0.0089, "step": 90010 }, { "grad_norm": 0.40894341468811035, "learning_rate": 2.698946515810374e-06, "loss": 0.0151, "step": 90020 }, { "grad_norm": 0.5660732984542847, "learning_rate": 2.6935901217812367e-06, "loss": 0.0143, "step": 90030 }, { "grad_norm": 0.1018032431602478, "learning_rate": 2.688238901109952e-06, "loss": 0.0211, "step": 90040 }, { "grad_norm": 0.4457624852657318, "learning_rate": 2.68289285438173e-06, "loss": 0.0114, "step": 90050 }, { "grad_norm": 0.20799849927425385, "learning_rate": 2.6775519821812033e-06, "loss": 0.0188, "step": 90060 }, { "grad_norm": 0.31790563464164734, "learning_rate": 2.672216285092449e-06, "loss": 0.0174, "step": 90070 }, { "grad_norm": 0.9154787063598633, "learning_rate": 2.666885763698962e-06, "loss": 0.0171, "step": 90080 }, { "grad_norm": 0.24884513020515442, "learning_rate": 2.661560418583692e-06, "loss": 0.0151, "step": 90090 }, { "grad_norm": 0.26556825637817383, "learning_rate": 2.656240250328995e-06, "loss": 0.0135, "step": 90100 }, { "grad_norm": 0.2597903907299042, "learning_rate": 2.650925259516679e-06, "loss": 0.0176, "step": 90110 }, { "grad_norm": 0.33312150835990906, "learning_rate": 2.645615446727995e-06, "loss": 0.0192, "step": 90120 }, { "grad_norm": 0.26762834191322327, "learning_rate": 2.6403108125436115e-06, "loss": 0.0156, "step": 90130 }, { "grad_norm": 0.3723401129245758, "learning_rate": 2.6350113575436264e-06, "loss": 0.0127, "step": 90140 }, { "grad_norm": 0.26056161522865295, "learning_rate": 2.629717082307598e-06, "loss": 0.0122, "step": 90150 }, { "grad_norm": 0.32843631505966187, "learning_rate": 2.6244279874144793e-06, "loss": 0.0136, "step": 90160 }, { "grad_norm": 1.2872364521026611, "learning_rate": 2.6191440734426862e-06, "loss": 0.012, "step": 90170 }, { "grad_norm": 0.17754614353179932, "learning_rate": 2.613865340970062e-06, "loss": 0.0167, "step": 90180 }, { "grad_norm": 0.2771918475627899, "learning_rate": 2.608591790573889e-06, "loss": 0.0173, "step": 90190 }, { "grad_norm": 0.19782336056232452, "learning_rate": 2.603323422830861e-06, "loss": 0.0125, "step": 90200 }, { "grad_norm": 0.6656505465507507, "learning_rate": 2.5980602383171183e-06, "loss": 0.0166, "step": 90210 }, { "grad_norm": 0.4352622628211975, "learning_rate": 2.5928022376082496e-06, "loss": 0.0147, "step": 90220 }, { "grad_norm": 0.16175523400306702, "learning_rate": 2.5875494212792505e-06, "loss": 0.0134, "step": 90230 }, { "grad_norm": 0.34535038471221924, "learning_rate": 2.5823017899045565e-06, "loss": 0.0119, "step": 90240 }, { "grad_norm": 0.5775004029273987, "learning_rate": 2.577059344058047e-06, "loss": 0.008, "step": 90250 }, { "grad_norm": 0.4512802064418793, "learning_rate": 2.5718220843130415e-06, "loss": 0.0157, "step": 90260 }, { "grad_norm": 0.5155020952224731, "learning_rate": 2.5665900112422535e-06, "loss": 0.0138, "step": 90270 }, { "grad_norm": 0.18880312144756317, "learning_rate": 2.561363125417876e-06, "loss": 0.0243, "step": 90280 }, { "grad_norm": 0.28627312183380127, "learning_rate": 2.556141427411496e-06, "loss": 0.0144, "step": 90290 }, { "grad_norm": 0.37078413367271423, "learning_rate": 2.550924917794156e-06, "loss": 0.0143, "step": 90300 }, { "grad_norm": 0.45482301712036133, "learning_rate": 2.545713597136329e-06, "loss": 0.0215, "step": 90310 }, { "grad_norm": 0.364423930644989, "learning_rate": 2.5405074660079298e-06, "loss": 0.0138, "step": 90320 }, { "grad_norm": 0.5892202854156494, "learning_rate": 2.5353065249782647e-06, "loss": 0.0144, "step": 90330 }, { "grad_norm": 0.7684153318405151, "learning_rate": 2.530110774616129e-06, "loss": 0.0193, "step": 90340 }, { "grad_norm": 0.31090232729911804, "learning_rate": 2.524920215489701e-06, "loss": 0.0156, "step": 90350 }, { "grad_norm": 0.49803876876831055, "learning_rate": 2.519734848166627e-06, "loss": 0.0187, "step": 90360 }, { "grad_norm": 0.5296462774276733, "learning_rate": 2.514554673213948e-06, "loss": 0.0206, "step": 90370 }, { "grad_norm": 0.10698879510164261, "learning_rate": 2.5093796911981936e-06, "loss": 0.0104, "step": 90380 }, { "grad_norm": 0.15438447892665863, "learning_rate": 2.5042099026852672e-06, "loss": 0.0109, "step": 90390 }, { "grad_norm": 0.14640969038009644, "learning_rate": 2.499045308240544e-06, "loss": 0.0131, "step": 90400 }, { "grad_norm": 0.2743660807609558, "learning_rate": 2.4938859084288003e-06, "loss": 0.0133, "step": 90410 }, { "grad_norm": 0.22302189469337463, "learning_rate": 2.4887317038142734e-06, "loss": 0.0132, "step": 90420 }, { "grad_norm": 0.3503021001815796, "learning_rate": 2.4835826949606124e-06, "loss": 0.0198, "step": 90430 }, { "grad_norm": 0.3528991937637329, "learning_rate": 2.4784388824309167e-06, "loss": 0.017, "step": 90440 }, { "grad_norm": 0.33991357684135437, "learning_rate": 2.473300266787687e-06, "loss": 0.0135, "step": 90450 }, { "grad_norm": 0.5508308410644531, "learning_rate": 2.46816684859289e-06, "loss": 0.0195, "step": 90460 }, { "grad_norm": 0.13843880593776703, "learning_rate": 2.4630386284079043e-06, "loss": 0.0138, "step": 90470 }, { "grad_norm": 0.39354199171066284, "learning_rate": 2.4579156067935427e-06, "loss": 0.0173, "step": 90480 }, { "grad_norm": 0.16274519264698029, "learning_rate": 2.452797784310046e-06, "loss": 0.013, "step": 90490 }, { "grad_norm": 0.40502747893333435, "learning_rate": 2.447685161517105e-06, "loss": 0.0181, "step": 90500 }, { "grad_norm": 0.27160197496414185, "learning_rate": 2.4425777389738224e-06, "loss": 0.0114, "step": 90510 }, { "grad_norm": 0.36323681473731995, "learning_rate": 2.43747551723873e-06, "loss": 0.0146, "step": 90520 }, { "grad_norm": 0.1659500151872635, "learning_rate": 2.4323784968698136e-06, "loss": 0.0135, "step": 90530 }, { "grad_norm": 1.9293713569641113, "learning_rate": 2.4272866784244565e-06, "loss": 0.0182, "step": 90540 }, { "grad_norm": 0.3458572030067444, "learning_rate": 2.422200062459518e-06, "loss": 0.019, "step": 90550 }, { "grad_norm": 0.3774312138557434, "learning_rate": 2.41711864953123e-06, "loss": 0.0145, "step": 90560 }, { "grad_norm": 0.35338646173477173, "learning_rate": 2.412042440195322e-06, "loss": 0.017, "step": 90570 }, { "grad_norm": 0.38518768548965454, "learning_rate": 2.4069714350068993e-06, "loss": 0.0138, "step": 90580 }, { "grad_norm": 0.317741721868515, "learning_rate": 2.4019056345205294e-06, "loss": 0.0137, "step": 90590 }, { "grad_norm": 0.7196796536445618, "learning_rate": 2.396845039290191e-06, "loss": 0.0158, "step": 90600 }, { "grad_norm": 0.38061076402664185, "learning_rate": 2.3917896498693194e-06, "loss": 0.0183, "step": 90610 }, { "grad_norm": 0.5959421396255493, "learning_rate": 2.3867394668107335e-06, "loss": 0.0162, "step": 90620 }, { "grad_norm": 0.17211098968982697, "learning_rate": 2.3816944906667527e-06, "loss": 0.0153, "step": 90630 }, { "grad_norm": 0.4051496386528015, "learning_rate": 2.3766547219890634e-06, "loss": 0.0202, "step": 90640 }, { "grad_norm": 0.5553843975067139, "learning_rate": 2.371620161328819e-06, "loss": 0.0205, "step": 90650 }, { "grad_norm": 0.3504946529865265, "learning_rate": 2.366590809236574e-06, "loss": 0.0162, "step": 90660 }, { "grad_norm": 0.39727118611335754, "learning_rate": 2.3615666662623494e-06, "loss": 0.0199, "step": 90670 }, { "grad_norm": 0.5893981456756592, "learning_rate": 2.3565477329555664e-06, "loss": 0.0184, "step": 90680 }, { "grad_norm": 0.37194007635116577, "learning_rate": 2.3515340098650975e-06, "loss": 0.0115, "step": 90690 }, { "grad_norm": 0.27715742588043213, "learning_rate": 2.3465254975392258e-06, "loss": 0.0155, "step": 90700 }, { "grad_norm": 0.217514306306839, "learning_rate": 2.3415221965256805e-06, "loss": 0.0222, "step": 90710 }, { "grad_norm": 0.3249267637729645, "learning_rate": 2.3365241073716237e-06, "loss": 0.0147, "step": 90720 }, { "grad_norm": 0.24574607610702515, "learning_rate": 2.3315312306236183e-06, "loss": 0.0144, "step": 90730 }, { "grad_norm": 0.31489264965057373, "learning_rate": 2.326543566827699e-06, "loss": 0.0166, "step": 90740 }, { "grad_norm": 0.20157894492149353, "learning_rate": 2.3215611165292862e-06, "loss": 0.0148, "step": 90750 }, { "grad_norm": 0.18827761709690094, "learning_rate": 2.3165838802732765e-06, "loss": 0.0165, "step": 90760 }, { "grad_norm": 0.3962831497192383, "learning_rate": 2.311611858603957e-06, "loss": 0.0158, "step": 90770 }, { "grad_norm": 0.45109185576438904, "learning_rate": 2.3066450520650763e-06, "loss": 0.0147, "step": 90780 }, { "grad_norm": 0.4323773980140686, "learning_rate": 2.3016834611997773e-06, "loss": 0.0242, "step": 90790 }, { "grad_norm": 0.3004375100135803, "learning_rate": 2.296727086550665e-06, "loss": 0.0221, "step": 90800 }, { "grad_norm": 0.4112962782382965, "learning_rate": 2.2917759286597507e-06, "loss": 0.0148, "step": 90810 }, { "grad_norm": 0.16631579399108887, "learning_rate": 2.2868299880685006e-06, "loss": 0.0133, "step": 90820 }, { "grad_norm": 1.804803729057312, "learning_rate": 2.2818892653177827e-06, "loss": 0.0247, "step": 90830 }, { "grad_norm": 1.036082148551941, "learning_rate": 2.2769537609479196e-06, "loss": 0.0202, "step": 90840 }, { "grad_norm": 0.737119197845459, "learning_rate": 2.2720234754986357e-06, "loss": 0.0139, "step": 90850 }, { "grad_norm": 0.7683370113372803, "learning_rate": 2.2670984095091108e-06, "loss": 0.0142, "step": 90860 }, { "grad_norm": 0.6309463977813721, "learning_rate": 2.262178563517936e-06, "loss": 0.0234, "step": 90870 }, { "grad_norm": 0.23426631093025208, "learning_rate": 2.257263938063137e-06, "loss": 0.0145, "step": 90880 }, { "grad_norm": 0.437701940536499, "learning_rate": 2.2523545336821715e-06, "loss": 0.0158, "step": 90890 }, { "grad_norm": 0.31270918250083923, "learning_rate": 2.247450350911939e-06, "loss": 0.0207, "step": 90900 }, { "grad_norm": 0.3788050711154938, "learning_rate": 2.2425513902887328e-06, "loss": 0.0108, "step": 90910 }, { "grad_norm": 0.6095259785652161, "learning_rate": 2.2376576523483015e-06, "loss": 0.0191, "step": 90920 }, { "grad_norm": 0.3101903200149536, "learning_rate": 2.232769137625829e-06, "loss": 0.0193, "step": 90930 }, { "grad_norm": 0.3171960711479187, "learning_rate": 2.227885846655903e-06, "loss": 0.0205, "step": 90940 }, { "grad_norm": 0.255807489156723, "learning_rate": 2.2230077799725523e-06, "loss": 0.0117, "step": 90950 }, { "grad_norm": 0.40918004512786865, "learning_rate": 2.2181349381092396e-06, "loss": 0.0152, "step": 90960 }, { "grad_norm": 0.287702739238739, "learning_rate": 2.2132673215988552e-06, "loss": 0.0129, "step": 90970 }, { "grad_norm": 0.19078725576400757, "learning_rate": 2.2084049309737065e-06, "loss": 0.0131, "step": 90980 }, { "grad_norm": 0.7615818977355957, "learning_rate": 2.2035477667655513e-06, "loss": 0.0149, "step": 90990 }, { "grad_norm": 0.4907931387424469, "learning_rate": 2.1986958295055315e-06, "loss": 0.0137, "step": 91000 }, { "grad_norm": 0.33482590317726135, "learning_rate": 2.193849119724284e-06, "loss": 0.0228, "step": 91010 }, { "grad_norm": 0.303234726190567, "learning_rate": 2.1890076379518175e-06, "loss": 0.0137, "step": 91020 }, { "grad_norm": 0.40323305130004883, "learning_rate": 2.1841713847175925e-06, "loss": 0.0115, "step": 91030 }, { "grad_norm": 0.4144199788570404, "learning_rate": 2.1793403605504968e-06, "loss": 0.0194, "step": 91040 }, { "grad_norm": 0.25817275047302246, "learning_rate": 2.1745145659788413e-06, "loss": 0.022, "step": 91050 }, { "grad_norm": 0.47844958305358887, "learning_rate": 2.1696940015303645e-06, "loss": 0.0138, "step": 91060 }, { "grad_norm": 0.4375285804271698, "learning_rate": 2.1648786677322397e-06, "loss": 0.0185, "step": 91070 }, { "grad_norm": 0.11529120802879333, "learning_rate": 2.1600685651110673e-06, "loss": 0.0116, "step": 91080 }, { "grad_norm": 0.3128843307495117, "learning_rate": 2.1552636941928717e-06, "loss": 0.0206, "step": 91090 }, { "grad_norm": 0.28518256545066833, "learning_rate": 2.1504640555031043e-06, "loss": 0.0126, "step": 91100 }, { "grad_norm": 0.17802414298057556, "learning_rate": 2.1456696495666506e-06, "loss": 0.0112, "step": 91110 }, { "grad_norm": 0.5212182402610779, "learning_rate": 2.1408804769078084e-06, "loss": 0.023, "step": 91120 }, { "grad_norm": 0.3489147424697876, "learning_rate": 2.136096538050325e-06, "loss": 0.017, "step": 91130 }, { "grad_norm": 0.33012065291404724, "learning_rate": 2.13131783351736e-06, "loss": 0.0196, "step": 91140 }, { "grad_norm": 0.3318319618701935, "learning_rate": 2.126544363831512e-06, "loss": 0.0216, "step": 91150 }, { "grad_norm": 0.2601424753665924, "learning_rate": 2.121776129514791e-06, "loss": 0.016, "step": 91160 }, { "grad_norm": 0.7599613070487976, "learning_rate": 2.1170131310886523e-06, "loss": 0.0151, "step": 91170 }, { "grad_norm": 0.2917466461658478, "learning_rate": 2.112255369073968e-06, "loss": 0.02, "step": 91180 }, { "grad_norm": 0.8254636526107788, "learning_rate": 2.1075028439910328e-06, "loss": 0.0192, "step": 91190 }, { "grad_norm": 0.4960317611694336, "learning_rate": 2.1027555563595813e-06, "loss": 0.0157, "step": 91200 }, { "grad_norm": 0.4220905601978302, "learning_rate": 2.0980135066987706e-06, "loss": 0.0157, "step": 91210 }, { "grad_norm": 0.5022881627082825, "learning_rate": 2.093276695527191e-06, "loss": 0.0137, "step": 91220 }, { "grad_norm": 0.4103667736053467, "learning_rate": 2.0885451233628396e-06, "loss": 0.0134, "step": 91230 }, { "grad_norm": 0.2159319669008255, "learning_rate": 2.0838187907231633e-06, "loss": 0.0175, "step": 91240 }, { "grad_norm": 0.50027996301651, "learning_rate": 2.0790976981250153e-06, "loss": 0.0135, "step": 91250 }, { "grad_norm": 0.38938429951667786, "learning_rate": 2.0743818460847e-06, "loss": 0.0155, "step": 91260 }, { "grad_norm": 0.06712600588798523, "learning_rate": 2.069671235117926e-06, "loss": 0.0125, "step": 91270 }, { "grad_norm": 0.30802375078201294, "learning_rate": 2.0649658657398484e-06, "loss": 0.0224, "step": 91280 }, { "grad_norm": 0.21443195641040802, "learning_rate": 2.060265738465034e-06, "loss": 0.0098, "step": 91290 }, { "grad_norm": 0.1545211374759674, "learning_rate": 2.055570853807487e-06, "loss": 0.0116, "step": 91300 }, { "grad_norm": 0.24112556874752045, "learning_rate": 2.0508812122806196e-06, "loss": 0.015, "step": 91310 }, { "grad_norm": 0.27144473791122437, "learning_rate": 2.046196814397294e-06, "loss": 0.0145, "step": 91320 }, { "grad_norm": 1.1970164775848389, "learning_rate": 2.0415176606697896e-06, "loss": 0.0153, "step": 91330 }, { "grad_norm": 0.28812506794929504, "learning_rate": 2.0368437516098126e-06, "loss": 0.0131, "step": 91340 }, { "grad_norm": 0.42639386653900146, "learning_rate": 2.032175087728483e-06, "loss": 0.0177, "step": 91350 }, { "grad_norm": 0.3202979862689972, "learning_rate": 2.027511669536375e-06, "loss": 0.019, "step": 91360 }, { "grad_norm": 0.22362154722213745, "learning_rate": 2.0228534975434643e-06, "loss": 0.0179, "step": 91370 }, { "grad_norm": 0.41389650106430054, "learning_rate": 2.0182005722591548e-06, "loss": 0.0106, "step": 91380 }, { "grad_norm": 0.5928874611854553, "learning_rate": 2.0135528941922944e-06, "loss": 0.018, "step": 91390 }, { "grad_norm": 0.30463671684265137, "learning_rate": 2.008910463851149e-06, "loss": 0.0137, "step": 91400 }, { "grad_norm": 0.2691117227077484, "learning_rate": 2.0042732817433963e-06, "loss": 0.0123, "step": 91410 }, { "grad_norm": 0.42927995324134827, "learning_rate": 1.999641348376158e-06, "loss": 0.0154, "step": 91420 }, { "grad_norm": 0.348832905292511, "learning_rate": 1.9950146642559787e-06, "loss": 0.0232, "step": 91430 }, { "grad_norm": 0.5121863484382629, "learning_rate": 1.990393229888815e-06, "loss": 0.0149, "step": 91440 }, { "grad_norm": 0.3767237067222595, "learning_rate": 1.9857770457800684e-06, "loss": 0.0138, "step": 91450 }, { "grad_norm": 0.2853717505931854, "learning_rate": 1.9811661124345517e-06, "loss": 0.0169, "step": 91460 }, { "grad_norm": 0.39215371012687683, "learning_rate": 1.976560430356522e-06, "loss": 0.0156, "step": 91470 }, { "grad_norm": 0.24403686821460724, "learning_rate": 1.9719600000496383e-06, "loss": 0.0182, "step": 91480 }, { "grad_norm": 0.818845272064209, "learning_rate": 1.967364822017004e-06, "loss": 0.0173, "step": 91490 }, { "grad_norm": 0.08358437567949295, "learning_rate": 1.962774896761127e-06, "loss": 0.0121, "step": 91500 }, { "grad_norm": 0.19531308114528656, "learning_rate": 1.958190224783968e-06, "loss": 0.0164, "step": 91510 }, { "grad_norm": 0.26942482590675354, "learning_rate": 1.9536108065868975e-06, "loss": 0.015, "step": 91520 }, { "grad_norm": 0.3030293881893158, "learning_rate": 1.949036642670715e-06, "loss": 0.0122, "step": 91530 }, { "grad_norm": 0.4387955069541931, "learning_rate": 1.9444677335356377e-06, "loss": 0.0186, "step": 91540 }, { "grad_norm": 0.21618996560573578, "learning_rate": 1.939904079681326e-06, "loss": 0.0248, "step": 91550 }, { "grad_norm": 0.8112868070602417, "learning_rate": 1.935345681606837e-06, "loss": 0.0142, "step": 91560 }, { "grad_norm": 1.408868432044983, "learning_rate": 1.930792539810694e-06, "loss": 0.0148, "step": 91570 }, { "grad_norm": 0.19129349291324615, "learning_rate": 1.9262446547907877e-06, "loss": 0.0172, "step": 91580 }, { "grad_norm": 0.25618019700050354, "learning_rate": 1.921702027044503e-06, "loss": 0.0126, "step": 91590 }, { "grad_norm": 0.7251725196838379, "learning_rate": 1.917164657068593e-06, "loss": 0.0153, "step": 91600 }, { "grad_norm": 0.23039744794368744, "learning_rate": 1.912632545359272e-06, "loss": 0.0218, "step": 91610 }, { "grad_norm": 0.3123187720775604, "learning_rate": 1.9081056924121488e-06, "loss": 0.0139, "step": 91620 }, { "grad_norm": 0.18527445197105408, "learning_rate": 1.9035840987222887e-06, "loss": 0.0158, "step": 91630 }, { "grad_norm": 0.27719491720199585, "learning_rate": 1.8990677647841515e-06, "loss": 0.0223, "step": 91640 }, { "grad_norm": 0.7167475819587708, "learning_rate": 1.8945566910916534e-06, "loss": 0.0176, "step": 91650 }, { "grad_norm": 0.40561947226524353, "learning_rate": 1.8900508781381054e-06, "loss": 0.0164, "step": 91660 }, { "grad_norm": 0.1678977906703949, "learning_rate": 1.8855503264162633e-06, "loss": 0.0127, "step": 91670 }, { "grad_norm": 0.39586105942726135, "learning_rate": 1.8810550364182998e-06, "loss": 0.0183, "step": 91680 }, { "grad_norm": 0.5004842877388, "learning_rate": 1.8765650086358112e-06, "loss": 0.0123, "step": 91690 }, { "grad_norm": 0.3511539399623871, "learning_rate": 1.8720802435598151e-06, "loss": 0.0162, "step": 91700 }, { "grad_norm": 0.2436862736940384, "learning_rate": 1.8676007416807694e-06, "loss": 0.0117, "step": 91710 }, { "grad_norm": 0.29454585909843445, "learning_rate": 1.8631265034885437e-06, "loss": 0.0129, "step": 91720 }, { "grad_norm": 0.727126955986023, "learning_rate": 1.858657529472424e-06, "loss": 0.0133, "step": 91730 }, { "grad_norm": 0.22382709383964539, "learning_rate": 1.8541938201211473e-06, "loss": 0.019, "step": 91740 }, { "grad_norm": 0.43950337171554565, "learning_rate": 1.8497353759228398e-06, "loss": 0.0092, "step": 91750 }, { "grad_norm": 0.3978107273578644, "learning_rate": 1.8452821973650835e-06, "loss": 0.0194, "step": 91760 }, { "grad_norm": 0.3208135664463043, "learning_rate": 1.840834284934856e-06, "loss": 0.0171, "step": 91770 }, { "grad_norm": 0.3131520450115204, "learning_rate": 1.8363916391185954e-06, "loss": 0.0167, "step": 91780 }, { "grad_norm": 0.7937883734703064, "learning_rate": 1.8319542604021245e-06, "loss": 0.0156, "step": 91790 }, { "grad_norm": 0.11363314092159271, "learning_rate": 1.8275221492707273e-06, "loss": 0.0136, "step": 91800 }, { "grad_norm": 0.5420602560043335, "learning_rate": 1.8230953062090717e-06, "loss": 0.0223, "step": 91810 }, { "grad_norm": 0.13822400569915771, "learning_rate": 1.8186737317012926e-06, "loss": 0.0148, "step": 91820 }, { "grad_norm": 0.9237740635871887, "learning_rate": 1.8142574262308977e-06, "loss": 0.0185, "step": 91830 }, { "grad_norm": 0.2549940347671509, "learning_rate": 1.8098463902808727e-06, "loss": 0.0198, "step": 91840 }, { "grad_norm": 0.30017781257629395, "learning_rate": 1.805440624333593e-06, "loss": 0.012, "step": 91850 }, { "grad_norm": 0.5808184146881104, "learning_rate": 1.8010401288708734e-06, "loss": 0.0196, "step": 91860 }, { "grad_norm": 0.3734737038612366, "learning_rate": 1.7966449043739343e-06, "loss": 0.0116, "step": 91870 }, { "grad_norm": 0.8019247055053711, "learning_rate": 1.79225495132343e-06, "loss": 0.013, "step": 91880 }, { "grad_norm": 0.6623746156692505, "learning_rate": 1.7878702701994487e-06, "loss": 0.0173, "step": 91890 }, { "grad_norm": 0.37052035331726074, "learning_rate": 1.783490861481496e-06, "loss": 0.0181, "step": 91900 }, { "grad_norm": 0.5796210765838623, "learning_rate": 1.7791167256484832e-06, "loss": 0.0225, "step": 91910 }, { "grad_norm": 0.5857192277908325, "learning_rate": 1.774747863178766e-06, "loss": 0.0187, "step": 91920 }, { "grad_norm": 0.2452685832977295, "learning_rate": 1.7703842745501186e-06, "loss": 0.0151, "step": 91930 }, { "grad_norm": 0.40933603048324585, "learning_rate": 1.7660259602397366e-06, "loss": 0.0139, "step": 91940 }, { "grad_norm": 0.49903467297554016, "learning_rate": 1.7616729207242388e-06, "loss": 0.0213, "step": 91950 }, { "grad_norm": 0.5472672581672668, "learning_rate": 1.7573251564796556e-06, "loss": 0.0126, "step": 91960 }, { "grad_norm": 0.4434465765953064, "learning_rate": 1.7529826679814676e-06, "loss": 0.0154, "step": 91970 }, { "grad_norm": 0.3311399519443512, "learning_rate": 1.7486454557045562e-06, "loss": 0.0146, "step": 91980 }, { "grad_norm": 0.761431872844696, "learning_rate": 1.7443135201232364e-06, "loss": 0.0141, "step": 91990 }, { "grad_norm": 0.7540485262870789, "learning_rate": 1.7399868617112347e-06, "loss": 0.0229, "step": 92000 }, { "grad_norm": 0.2510911226272583, "learning_rate": 1.7356654809417171e-06, "loss": 0.0122, "step": 92010 }, { "grad_norm": 0.2590807378292084, "learning_rate": 1.7313493782872448e-06, "loss": 0.0183, "step": 92020 }, { "grad_norm": 0.4026450216770172, "learning_rate": 1.7270385542198453e-06, "loss": 0.0199, "step": 92030 }, { "grad_norm": 0.28578823804855347, "learning_rate": 1.7227330092109307e-06, "loss": 0.0171, "step": 92040 }, { "grad_norm": 0.6201162934303284, "learning_rate": 1.718432743731352e-06, "loss": 0.0169, "step": 92050 }, { "grad_norm": 0.3466912806034088, "learning_rate": 1.7141377582513718e-06, "loss": 0.0173, "step": 92060 }, { "grad_norm": 0.17872728407382965, "learning_rate": 1.709848053240698e-06, "loss": 0.0226, "step": 92070 }, { "grad_norm": 0.3945762515068054, "learning_rate": 1.7055636291684328e-06, "loss": 0.0136, "step": 92080 }, { "grad_norm": 0.2759719491004944, "learning_rate": 1.7012844865031185e-06, "loss": 0.0128, "step": 92090 }, { "grad_norm": 0.167727530002594, "learning_rate": 1.6970106257127138e-06, "loss": 0.0177, "step": 92100 }, { "grad_norm": 0.5046120285987854, "learning_rate": 1.6927420472646117e-06, "loss": 0.0209, "step": 92110 }, { "grad_norm": 0.33306220173835754, "learning_rate": 1.6884787516256106e-06, "loss": 0.0145, "step": 92120 }, { "grad_norm": 0.4780898988246918, "learning_rate": 1.6842207392619325e-06, "loss": 0.0097, "step": 92130 }, { "grad_norm": 0.2866019606590271, "learning_rate": 1.679968010639238e-06, "loss": 0.0212, "step": 92140 }, { "grad_norm": 0.25291362404823303, "learning_rate": 1.6757205662225829e-06, "loss": 0.0203, "step": 92150 }, { "grad_norm": 0.3839866518974304, "learning_rate": 1.6714784064764844e-06, "loss": 0.0178, "step": 92160 }, { "grad_norm": 0.2921735644340515, "learning_rate": 1.6672415318648383e-06, "loss": 0.0132, "step": 92170 }, { "grad_norm": 0.45314130187034607, "learning_rate": 1.6630099428510016e-06, "loss": 0.0199, "step": 92180 }, { "grad_norm": 0.282520055770874, "learning_rate": 1.6587836398977153e-06, "loss": 0.0213, "step": 92190 }, { "grad_norm": 0.3363208770751953, "learning_rate": 1.6545626234671762e-06, "loss": 0.0195, "step": 92200 }, { "grad_norm": 0.38979676365852356, "learning_rate": 1.650346894020971e-06, "loss": 0.0165, "step": 92210 }, { "grad_norm": 0.18737433850765228, "learning_rate": 1.6461364520201472e-06, "loss": 0.0147, "step": 92220 }, { "grad_norm": 0.40187472105026245, "learning_rate": 1.6419312979251366e-06, "loss": 0.0173, "step": 92230 }, { "grad_norm": 0.18474188446998596, "learning_rate": 1.6377314321958214e-06, "loss": 0.0121, "step": 92240 }, { "grad_norm": 0.2115594446659088, "learning_rate": 1.6335368552914787e-06, "loss": 0.0131, "step": 92250 }, { "grad_norm": 0.31350648403167725, "learning_rate": 1.6293475676708358e-06, "loss": 0.0166, "step": 92260 }, { "grad_norm": 0.3456001877784729, "learning_rate": 1.6251635697920097e-06, "loss": 0.0186, "step": 92270 }, { "grad_norm": 0.9560180306434631, "learning_rate": 1.6209848621125624e-06, "loss": 0.0204, "step": 92280 }, { "grad_norm": 0.4504183828830719, "learning_rate": 1.6168114450894778e-06, "loss": 0.0144, "step": 92290 }, { "grad_norm": 0.22603200376033783, "learning_rate": 1.612643319179158e-06, "loss": 0.0105, "step": 92300 }, { "grad_norm": 0.37299099564552307, "learning_rate": 1.6084804848374046e-06, "loss": 0.0234, "step": 92310 }, { "grad_norm": 0.5281956791877747, "learning_rate": 1.6043229425194817e-06, "loss": 0.0213, "step": 92320 }, { "grad_norm": 0.39103275537490845, "learning_rate": 1.6001706926800308e-06, "loss": 0.0161, "step": 92330 }, { "grad_norm": 0.22656789422035217, "learning_rate": 1.5960237357731444e-06, "loss": 0.0145, "step": 92340 }, { "grad_norm": 0.22609096765518188, "learning_rate": 1.5918820722523264e-06, "loss": 0.018, "step": 92350 }, { "grad_norm": 0.27057841420173645, "learning_rate": 1.587745702570509e-06, "loss": 0.0171, "step": 92360 }, { "grad_norm": 0.3220551311969757, "learning_rate": 1.5836146271800356e-06, "loss": 0.0139, "step": 92370 }, { "grad_norm": 0.280351459980011, "learning_rate": 1.5794888465326728e-06, "loss": 0.0103, "step": 92380 }, { "grad_norm": 0.3164025843143463, "learning_rate": 1.5753683610796155e-06, "loss": 0.0243, "step": 92390 }, { "grad_norm": 1.067918300628662, "learning_rate": 1.5712531712714528e-06, "loss": 0.0107, "step": 92400 }, { "grad_norm": 0.2027358114719391, "learning_rate": 1.5671432775582474e-06, "loss": 0.0219, "step": 92410 }, { "grad_norm": 0.28415483236312866, "learning_rate": 1.563038680389428e-06, "loss": 0.013, "step": 92420 }, { "grad_norm": 0.11236817389726639, "learning_rate": 1.5589393802138808e-06, "loss": 0.0147, "step": 92430 }, { "grad_norm": 0.2570495009422302, "learning_rate": 1.554845377479891e-06, "loss": 0.013, "step": 92440 }, { "grad_norm": 0.22418788075447083, "learning_rate": 1.550756672635173e-06, "loss": 0.0175, "step": 92450 }, { "grad_norm": 0.3669622540473938, "learning_rate": 1.5466732661268635e-06, "loss": 0.0157, "step": 92460 }, { "grad_norm": 0.17091026902198792, "learning_rate": 1.5425951584015163e-06, "loss": 0.0184, "step": 92470 }, { "grad_norm": 0.5942878723144531, "learning_rate": 1.5385223499051026e-06, "loss": 0.0193, "step": 92480 }, { "grad_norm": 0.5535098314285278, "learning_rate": 1.5344548410830328e-06, "loss": 0.0163, "step": 92490 }, { "grad_norm": 0.7518748641014099, "learning_rate": 1.5303926323801121e-06, "loss": 0.0147, "step": 92500 }, { "grad_norm": 0.27439191937446594, "learning_rate": 1.5263357242405795e-06, "loss": 0.0128, "step": 92510 }, { "grad_norm": 0.4294220805168152, "learning_rate": 1.5222841171080915e-06, "loss": 0.0151, "step": 92520 }, { "grad_norm": 0.42003437876701355, "learning_rate": 1.5182378114257268e-06, "loss": 0.015, "step": 92530 }, { "grad_norm": 0.19160601496696472, "learning_rate": 1.5141968076359869e-06, "loss": 0.0141, "step": 92540 }, { "grad_norm": 0.2464611828327179, "learning_rate": 1.5101611061807908e-06, "loss": 0.0144, "step": 92550 }, { "grad_norm": 0.6013087630271912, "learning_rate": 1.506130707501463e-06, "loss": 0.0161, "step": 92560 }, { "grad_norm": 1.0749012231826782, "learning_rate": 1.5021056120387734e-06, "loss": 0.0229, "step": 92570 }, { "grad_norm": 0.14633619785308838, "learning_rate": 1.4980858202329084e-06, "loss": 0.0115, "step": 92580 }, { "grad_norm": 0.3372955620288849, "learning_rate": 1.4940713325234445e-06, "loss": 0.0207, "step": 92590 }, { "grad_norm": 0.13511140644550323, "learning_rate": 1.4900621493494137e-06, "loss": 0.012, "step": 92600 }, { "grad_norm": 0.25710541009902954, "learning_rate": 1.4860582711492544e-06, "loss": 0.0213, "step": 92610 }, { "grad_norm": 0.3214583396911621, "learning_rate": 1.4820596983608325e-06, "loss": 0.0198, "step": 92620 }, { "grad_norm": 0.3270951807498932, "learning_rate": 1.4780664314214044e-06, "loss": 0.0137, "step": 92630 }, { "grad_norm": 0.5441331267356873, "learning_rate": 1.474078470767687e-06, "loss": 0.0165, "step": 92640 }, { "grad_norm": 0.44429972767829895, "learning_rate": 1.470095816835787e-06, "loss": 0.0153, "step": 92650 }, { "grad_norm": 0.34764334559440613, "learning_rate": 1.4661184700612395e-06, "loss": 0.0157, "step": 92660 }, { "grad_norm": 0.2818502187728882, "learning_rate": 1.4621464308790133e-06, "loss": 0.0149, "step": 92670 }, { "grad_norm": 0.2278207391500473, "learning_rate": 1.4581796997234775e-06, "loss": 0.0147, "step": 92680 }, { "grad_norm": 0.4214705526828766, "learning_rate": 1.4542182770284297e-06, "loss": 0.016, "step": 92690 }, { "grad_norm": 0.31956717371940613, "learning_rate": 1.450262163227084e-06, "loss": 0.014, "step": 92700 }, { "grad_norm": 0.5128792524337769, "learning_rate": 1.4463113587520671e-06, "loss": 0.0194, "step": 92710 }, { "grad_norm": 0.21870094537734985, "learning_rate": 1.4423658640354443e-06, "loss": 0.0133, "step": 92720 }, { "grad_norm": 0.22188328206539154, "learning_rate": 1.438425679508687e-06, "loss": 0.0177, "step": 92730 }, { "grad_norm": 0.16182637214660645, "learning_rate": 1.4344908056026895e-06, "loss": 0.0156, "step": 92740 }, { "grad_norm": 0.22655776143074036, "learning_rate": 1.4305612427477521e-06, "loss": 0.0111, "step": 92750 }, { "grad_norm": 0.23978722095489502, "learning_rate": 1.42663699137362e-06, "loss": 0.0104, "step": 92760 }, { "grad_norm": 0.2714650630950928, "learning_rate": 1.4227180519094386e-06, "loss": 0.0153, "step": 92770 }, { "grad_norm": 0.3167422115802765, "learning_rate": 1.4188044247837706e-06, "loss": 0.021, "step": 92780 }, { "grad_norm": 0.5448386669158936, "learning_rate": 1.4148961104246128e-06, "loss": 0.0159, "step": 92790 }, { "grad_norm": 0.8956809639930725, "learning_rate": 1.4109931092593731e-06, "loss": 0.0182, "step": 92800 }, { "grad_norm": 0.5827597379684448, "learning_rate": 1.4070954217148658e-06, "loss": 0.0163, "step": 92810 }, { "grad_norm": 0.25777679681777954, "learning_rate": 1.4032030482173497e-06, "loss": 0.0207, "step": 92820 }, { "grad_norm": 0.7476687431335449, "learning_rate": 1.3993159891924901e-06, "loss": 0.0188, "step": 92830 }, { "grad_norm": 0.3734751343727112, "learning_rate": 1.3954342450653524e-06, "loss": 0.0193, "step": 92840 }, { "grad_norm": 0.2610456347465515, "learning_rate": 1.3915578162604582e-06, "loss": 0.0156, "step": 92850 }, { "grad_norm": 0.47911250591278076, "learning_rate": 1.3876867032017126e-06, "loss": 0.02, "step": 92860 }, { "grad_norm": 0.1311025172472, "learning_rate": 1.383820906312472e-06, "loss": 0.0154, "step": 92870 }, { "grad_norm": 0.3036760985851288, "learning_rate": 1.3799604260154752e-06, "loss": 0.0186, "step": 92880 }, { "grad_norm": 0.27530214190483093, "learning_rate": 1.376105262732913e-06, "loss": 0.0257, "step": 92890 }, { "grad_norm": 0.2810943126678467, "learning_rate": 1.37225541688637e-06, "loss": 0.0156, "step": 92900 }, { "grad_norm": 0.24339015781879425, "learning_rate": 1.3684108888968594e-06, "loss": 0.0163, "step": 92910 }, { "grad_norm": 0.1604483723640442, "learning_rate": 1.364571679184823e-06, "loss": 0.0175, "step": 92920 }, { "grad_norm": 0.34074339270591736, "learning_rate": 1.3607377881701077e-06, "loss": 0.0229, "step": 92930 }, { "grad_norm": 0.20672369003295898, "learning_rate": 1.3569092162719787e-06, "loss": 0.0127, "step": 92940 }, { "grad_norm": 0.2081194967031479, "learning_rate": 1.353085963909123e-06, "loss": 0.01, "step": 92950 }, { "grad_norm": 0.3845423460006714, "learning_rate": 1.3492680314996454e-06, "loss": 0.0193, "step": 92960 }, { "grad_norm": 0.3831964135169983, "learning_rate": 1.345455419461078e-06, "loss": 0.0217, "step": 92970 }, { "grad_norm": 0.4485606849193573, "learning_rate": 1.3416481282103431e-06, "loss": 0.0158, "step": 92980 }, { "grad_norm": 0.6126697659492493, "learning_rate": 1.337846158163819e-06, "loss": 0.0176, "step": 92990 }, { "grad_norm": 0.19068516790866852, "learning_rate": 1.334049509737273e-06, "loss": 0.0165, "step": 93000 }, { "grad_norm": 0.26528483629226685, "learning_rate": 1.3302581833459116e-06, "loss": 0.0114, "step": 93010 }, { "grad_norm": 0.31030434370040894, "learning_rate": 1.3264721794043366e-06, "loss": 0.0125, "step": 93020 }, { "grad_norm": 0.5712450742721558, "learning_rate": 1.3226914983265836e-06, "loss": 0.0176, "step": 93030 }, { "grad_norm": 0.8481955528259277, "learning_rate": 1.3189161405260997e-06, "loss": 0.0247, "step": 93040 }, { "grad_norm": 0.29341500997543335, "learning_rate": 1.3151461064157656e-06, "loss": 0.0128, "step": 93050 }, { "grad_norm": 0.28537124395370483, "learning_rate": 1.3113813964078459e-06, "loss": 0.0102, "step": 93060 }, { "grad_norm": 0.19999413192272186, "learning_rate": 1.3076220109140614e-06, "loss": 0.0137, "step": 93070 }, { "grad_norm": 0.22007234394550323, "learning_rate": 1.3038679503455275e-06, "loss": 0.0136, "step": 93080 }, { "grad_norm": 0.32614225149154663, "learning_rate": 1.3001192151127773e-06, "loss": 0.0172, "step": 93090 }, { "grad_norm": 0.5179650783538818, "learning_rate": 1.2963758056257657e-06, "loss": 0.02, "step": 93100 }, { "grad_norm": 0.348177045583725, "learning_rate": 1.2926377222938768e-06, "loss": 0.0158, "step": 93110 }, { "grad_norm": 0.33933594822883606, "learning_rate": 1.2889049655259e-06, "loss": 0.0179, "step": 93120 }, { "grad_norm": 0.4141010046005249, "learning_rate": 1.2851775357300312e-06, "loss": 0.0225, "step": 93130 }, { "grad_norm": 0.5137236714363098, "learning_rate": 1.281455433313916e-06, "loss": 0.0167, "step": 93140 }, { "grad_norm": 0.14880014955997467, "learning_rate": 1.2777386586845796e-06, "loss": 0.011, "step": 93150 }, { "grad_norm": 0.26535847783088684, "learning_rate": 1.2740272122484964e-06, "loss": 0.0164, "step": 93160 }, { "grad_norm": 0.1283082813024521, "learning_rate": 1.2703210944115308e-06, "loss": 0.0116, "step": 93170 }, { "grad_norm": 0.13009391725063324, "learning_rate": 1.2666203055789916e-06, "loss": 0.0111, "step": 93180 }, { "grad_norm": 0.35211554169654846, "learning_rate": 1.2629248461555832e-06, "loss": 0.0125, "step": 93190 }, { "grad_norm": 0.33441784977912903, "learning_rate": 1.2592347165454431e-06, "loss": 0.0156, "step": 93200 }, { "grad_norm": 0.39901989698410034, "learning_rate": 1.2555499171521102e-06, "loss": 0.014, "step": 93210 }, { "grad_norm": 0.23940229415893555, "learning_rate": 1.251870448378556e-06, "loss": 0.0147, "step": 93220 }, { "grad_norm": 0.2617635428905487, "learning_rate": 1.2481963106271478e-06, "loss": 0.0157, "step": 93230 }, { "grad_norm": 0.3236062228679657, "learning_rate": 1.2445275042997084e-06, "loss": 0.0184, "step": 93240 }, { "grad_norm": 0.22934067249298096, "learning_rate": 1.2408640297974283e-06, "loss": 0.0139, "step": 93250 }, { "grad_norm": 0.4356459677219391, "learning_rate": 1.2372058875209535e-06, "loss": 0.0189, "step": 93260 }, { "grad_norm": 0.27049708366394043, "learning_rate": 1.2335530778703252e-06, "loss": 0.0147, "step": 93270 }, { "grad_norm": 0.1188993975520134, "learning_rate": 1.2299056012450182e-06, "loss": 0.0132, "step": 93280 }, { "grad_norm": 0.43316274881362915, "learning_rate": 1.2262634580439026e-06, "loss": 0.0137, "step": 93290 }, { "grad_norm": 1.4761557579040527, "learning_rate": 1.2226266486652926e-06, "loss": 0.0162, "step": 93300 }, { "grad_norm": 0.2547127306461334, "learning_rate": 1.2189951735068927e-06, "loss": 0.0164, "step": 93310 }, { "grad_norm": 0.2947016954421997, "learning_rate": 1.2153690329658407e-06, "loss": 0.0142, "step": 93320 }, { "grad_norm": 0.3337143659591675, "learning_rate": 1.2117482274386859e-06, "loss": 0.0138, "step": 93330 }, { "grad_norm": 0.3557473421096802, "learning_rate": 1.2081327573213953e-06, "loss": 0.0174, "step": 93340 }, { "grad_norm": 0.4167447090148926, "learning_rate": 1.2045226230093465e-06, "loss": 0.0137, "step": 93350 }, { "grad_norm": 0.206534743309021, "learning_rate": 1.2009178248973296e-06, "loss": 0.0165, "step": 93360 }, { "grad_norm": 0.530953586101532, "learning_rate": 1.1973183633795849e-06, "loss": 0.018, "step": 93370 }, { "grad_norm": 0.22651009261608124, "learning_rate": 1.193724238849725e-06, "loss": 0.0131, "step": 93380 }, { "grad_norm": 0.15030552446842194, "learning_rate": 1.190135451700808e-06, "loss": 0.0157, "step": 93390 }, { "grad_norm": 0.6872339248657227, "learning_rate": 1.186552002325292e-06, "loss": 0.0144, "step": 93400 }, { "grad_norm": 0.24736569821834564, "learning_rate": 1.1829738911150578e-06, "loss": 0.0126, "step": 93410 }, { "grad_norm": 0.209729865193367, "learning_rate": 1.1794011184613985e-06, "loss": 0.0132, "step": 93420 }, { "grad_norm": 0.12897811830043793, "learning_rate": 1.17583368475504e-06, "loss": 0.0131, "step": 93430 }, { "grad_norm": 0.3299591541290283, "learning_rate": 1.1722715903860982e-06, "loss": 0.011, "step": 93440 }, { "grad_norm": 0.18963301181793213, "learning_rate": 1.168714835744128e-06, "loss": 0.0186, "step": 93450 }, { "grad_norm": 0.23661433160305023, "learning_rate": 1.165163421218085e-06, "loss": 0.0127, "step": 93460 }, { "grad_norm": 0.4035247266292572, "learning_rate": 1.1616173471963477e-06, "loss": 0.0176, "step": 93470 }, { "grad_norm": 0.2911555767059326, "learning_rate": 1.1580766140667109e-06, "loss": 0.0136, "step": 93480 }, { "grad_norm": 0.4897574186325073, "learning_rate": 1.1545412222163764e-06, "loss": 0.0181, "step": 93490 }, { "grad_norm": 0.15795941650867462, "learning_rate": 1.1510111720319795e-06, "loss": 0.0114, "step": 93500 }, { "grad_norm": 0.2737465500831604, "learning_rate": 1.1474864638995663e-06, "loss": 0.0177, "step": 93510 }, { "grad_norm": 0.41463345289230347, "learning_rate": 1.1439670982045735e-06, "loss": 0.0155, "step": 93520 }, { "grad_norm": 0.39575526118278503, "learning_rate": 1.1404530753318877e-06, "loss": 0.0132, "step": 93530 }, { "grad_norm": 0.16741935908794403, "learning_rate": 1.136944395665801e-06, "loss": 0.0127, "step": 93540 }, { "grad_norm": 0.31853044033050537, "learning_rate": 1.1334410595900014e-06, "loss": 0.014, "step": 93550 }, { "grad_norm": 0.20429816842079163, "learning_rate": 1.129943067487621e-06, "loss": 0.0151, "step": 93560 }, { "grad_norm": 0.29513582587242126, "learning_rate": 1.126450419741193e-06, "loss": 0.0205, "step": 93570 }, { "grad_norm": 0.29233092069625854, "learning_rate": 1.1229631167326671e-06, "loss": 0.0119, "step": 93580 }, { "grad_norm": 0.2423999309539795, "learning_rate": 1.1194811588434052e-06, "loss": 0.0159, "step": 93590 }, { "grad_norm": 0.25253430008888245, "learning_rate": 1.1160045464542023e-06, "loss": 0.0103, "step": 93600 }, { "grad_norm": 0.30578216910362244, "learning_rate": 1.1125332799452325e-06, "loss": 0.0198, "step": 93610 }, { "grad_norm": 0.14727085828781128, "learning_rate": 1.109067359696131e-06, "loss": 0.0166, "step": 93620 }, { "grad_norm": 0.5631014704704285, "learning_rate": 1.1056067860859166e-06, "loss": 0.0112, "step": 93630 }, { "grad_norm": 0.19533532857894897, "learning_rate": 1.102151559493031e-06, "loss": 0.0105, "step": 93640 }, { "grad_norm": 0.26998838782310486, "learning_rate": 1.0987016802953332e-06, "loss": 0.0173, "step": 93650 }, { "grad_norm": 0.26306095719337463, "learning_rate": 1.0952571488701048e-06, "loss": 0.0186, "step": 93660 }, { "grad_norm": 0.2694999575614929, "learning_rate": 1.0918179655940163e-06, "loss": 0.0137, "step": 93670 }, { "grad_norm": 0.2018403559923172, "learning_rate": 1.088384130843184e-06, "loss": 0.0165, "step": 93680 }, { "grad_norm": 0.4448162317276001, "learning_rate": 1.0849556449931296e-06, "loss": 0.0198, "step": 93690 }, { "grad_norm": 0.5691652894020081, "learning_rate": 1.0815325084187866e-06, "loss": 0.0173, "step": 93700 }, { "grad_norm": 0.2837650775909424, "learning_rate": 1.0781147214944943e-06, "loss": 0.0129, "step": 93710 }, { "grad_norm": 0.571713387966156, "learning_rate": 1.0747022845940313e-06, "loss": 0.0146, "step": 93720 }, { "grad_norm": 0.4652059078216553, "learning_rate": 1.0712951980905605e-06, "loss": 0.0113, "step": 93730 }, { "grad_norm": 0.2874923050403595, "learning_rate": 1.0678934623566839e-06, "loss": 0.0207, "step": 93740 }, { "grad_norm": 0.5149854421615601, "learning_rate": 1.0644970777644091e-06, "loss": 0.0153, "step": 93750 }, { "grad_norm": 0.25114908814430237, "learning_rate": 1.0611060446851673e-06, "loss": 0.0121, "step": 93760 }, { "grad_norm": 0.19428254663944244, "learning_rate": 1.0577203634897836e-06, "loss": 0.0153, "step": 93770 }, { "grad_norm": 0.1760370135307312, "learning_rate": 1.0543400345485178e-06, "loss": 0.0119, "step": 93780 }, { "grad_norm": 0.09803730249404907, "learning_rate": 1.0509650582310404e-06, "loss": 0.0141, "step": 93790 }, { "grad_norm": 0.4239036738872528, "learning_rate": 1.0475954349064287e-06, "loss": 0.0158, "step": 93800 }, { "grad_norm": 0.29811590909957886, "learning_rate": 1.0442311649431825e-06, "loss": 0.0118, "step": 93810 }, { "grad_norm": 0.32845863699913025, "learning_rate": 1.0408722487092071e-06, "loss": 0.0213, "step": 93820 }, { "grad_norm": 0.27462974190711975, "learning_rate": 1.0375186865718424e-06, "loss": 0.0179, "step": 93830 }, { "grad_norm": 0.3911924362182617, "learning_rate": 1.0341704788978113e-06, "loss": 0.0195, "step": 93840 }, { "grad_norm": 0.742936909198761, "learning_rate": 1.0308276260532879e-06, "loss": 0.0213, "step": 93850 }, { "grad_norm": 0.46909862756729126, "learning_rate": 1.0274901284038241e-06, "loss": 0.0142, "step": 93860 }, { "grad_norm": 0.26278719305992126, "learning_rate": 1.0241579863144114e-06, "loss": 0.0157, "step": 93870 }, { "grad_norm": 0.262990266084671, "learning_rate": 1.0208312001494413e-06, "loss": 0.0128, "step": 93880 }, { "grad_norm": 0.3322804272174835, "learning_rate": 1.0175097702727453e-06, "loss": 0.0239, "step": 93890 }, { "grad_norm": 0.6647081971168518, "learning_rate": 1.0141936970475275e-06, "loss": 0.0242, "step": 93900 }, { "grad_norm": 0.33589011430740356, "learning_rate": 1.0108829808364417e-06, "loss": 0.0171, "step": 93910 }, { "grad_norm": 0.23441924154758453, "learning_rate": 1.0075776220015376e-06, "loss": 0.0189, "step": 93920 }, { "grad_norm": 0.15291675925254822, "learning_rate": 1.0042776209042816e-06, "loss": 0.0128, "step": 93930 }, { "grad_norm": 0.25522616505622864, "learning_rate": 1.000982977905568e-06, "loss": 0.0108, "step": 93940 }, { "grad_norm": 0.18839767575263977, "learning_rate": 9.976936933656867e-07, "loss": 0.0121, "step": 93950 }, { "grad_norm": 0.3961774706840515, "learning_rate": 9.944097676443442e-07, "loss": 0.0155, "step": 93960 }, { "grad_norm": 0.3820531964302063, "learning_rate": 9.911312011006757e-07, "loss": 0.0132, "step": 93970 }, { "grad_norm": 0.2712325155735016, "learning_rate": 9.878579940932109e-07, "loss": 0.0144, "step": 93980 }, { "grad_norm": 0.28888773918151855, "learning_rate": 9.84590146979908e-07, "loss": 0.0108, "step": 93990 }, { "grad_norm": 0.19526156783103943, "learning_rate": 9.813276601181365e-07, "loss": 0.0202, "step": 94000 }, { "grad_norm": 0.37270334362983704, "learning_rate": 9.780705338646724e-07, "loss": 0.0101, "step": 94010 }, { "grad_norm": 0.5240798592567444, "learning_rate": 9.748187685757083e-07, "loss": 0.017, "step": 94020 }, { "grad_norm": 0.18024463951587677, "learning_rate": 9.715723646068543e-07, "loss": 0.0136, "step": 94030 }, { "grad_norm": 0.906746506690979, "learning_rate": 9.68331322313143e-07, "loss": 0.011, "step": 94040 }, { "grad_norm": 0.3897082507610321, "learning_rate": 9.650956420489909e-07, "loss": 0.0166, "step": 94050 }, { "grad_norm": 0.26957979798316956, "learning_rate": 9.618653241682595e-07, "loss": 0.021, "step": 94060 }, { "grad_norm": 0.3099217414855957, "learning_rate": 9.586403690242053e-07, "loss": 0.0198, "step": 94070 }, { "grad_norm": 0.13270841538906097, "learning_rate": 9.554207769695122e-07, "loss": 0.0179, "step": 94080 }, { "grad_norm": 0.2408638447523117, "learning_rate": 9.522065483562659e-07, "loss": 0.0128, "step": 94090 }, { "grad_norm": 0.3110752999782562, "learning_rate": 9.489976835359681e-07, "loss": 0.0142, "step": 94100 }, { "grad_norm": 0.2921414077281952, "learning_rate": 9.457941828595329e-07, "loss": 0.0119, "step": 94110 }, { "grad_norm": 0.3606000244617462, "learning_rate": 9.425960466772965e-07, "loss": 0.0168, "step": 94120 }, { "grad_norm": 0.41808903217315674, "learning_rate": 9.394032753390014e-07, "loss": 0.0152, "step": 94130 }, { "grad_norm": 0.33420494198799133, "learning_rate": 9.362158691938017e-07, "loss": 0.0153, "step": 94140 }, { "grad_norm": 0.3705632984638214, "learning_rate": 9.330338285902685e-07, "loss": 0.0125, "step": 94150 }, { "grad_norm": 0.26559022068977356, "learning_rate": 9.298571538763845e-07, "loss": 0.0166, "step": 94160 }, { "grad_norm": 0.28122371435165405, "learning_rate": 9.266858453995441e-07, "loss": 0.0078, "step": 94170 }, { "grad_norm": 0.30711182951927185, "learning_rate": 9.235199035065645e-07, "loss": 0.0128, "step": 94180 }, { "grad_norm": 0.5170629620552063, "learning_rate": 9.203593285436574e-07, "loss": 0.0126, "step": 94190 }, { "grad_norm": 0.2980400621891022, "learning_rate": 9.172041208564686e-07, "loss": 0.0144, "step": 94200 }, { "grad_norm": 0.24875682592391968, "learning_rate": 9.140542807900332e-07, "loss": 0.0161, "step": 94210 }, { "grad_norm": 0.6562859416007996, "learning_rate": 9.109098086888313e-07, "loss": 0.0106, "step": 94220 }, { "grad_norm": 0.6142733693122864, "learning_rate": 9.077707048967265e-07, "loss": 0.0122, "step": 94230 }, { "grad_norm": 0.11462974548339844, "learning_rate": 9.046369697570057e-07, "loss": 0.0129, "step": 94240 }, { "grad_norm": 0.21175624430179596, "learning_rate": 9.015086036123721e-07, "loss": 0.0147, "step": 94250 }, { "grad_norm": 0.358916312456131, "learning_rate": 8.983856068049413e-07, "loss": 0.0178, "step": 94260 }, { "grad_norm": 0.4383171796798706, "learning_rate": 8.952679796762398e-07, "loss": 0.0122, "step": 94270 }, { "grad_norm": 0.6652634739875793, "learning_rate": 8.921557225671951e-07, "loss": 0.0191, "step": 94280 }, { "grad_norm": 0.12359678745269775, "learning_rate": 8.890488358181792e-07, "loss": 0.016, "step": 94290 }, { "grad_norm": 0.5410719513893127, "learning_rate": 8.859473197689372e-07, "loss": 0.0158, "step": 94300 }, { "grad_norm": 0.3957628011703491, "learning_rate": 8.828511747586588e-07, "loss": 0.0119, "step": 94310 }, { "grad_norm": 0.24884670972824097, "learning_rate": 8.797604011259287e-07, "loss": 0.0212, "step": 94320 }, { "grad_norm": 0.3177802264690399, "learning_rate": 8.766749992087487e-07, "loss": 0.0132, "step": 94330 }, { "grad_norm": 0.2928844690322876, "learning_rate": 8.735949693445378e-07, "loss": 0.0164, "step": 94340 }, { "grad_norm": 0.4171878397464752, "learning_rate": 8.705203118701211e-07, "loss": 0.0151, "step": 94350 }, { "grad_norm": 0.3591962456703186, "learning_rate": 8.674510271217295e-07, "loss": 0.0168, "step": 94360 }, { "grad_norm": 0.4144989848136902, "learning_rate": 8.643871154350336e-07, "loss": 0.0196, "step": 94370 }, { "grad_norm": 0.1785341501235962, "learning_rate": 8.613285771450818e-07, "loss": 0.0202, "step": 94380 }, { "grad_norm": 0.18195898830890656, "learning_rate": 8.582754125863623e-07, "loss": 0.0165, "step": 94390 }, { "grad_norm": 0.8549227118492126, "learning_rate": 8.55227622092758e-07, "loss": 0.0207, "step": 94400 }, { "grad_norm": 0.5282251238822937, "learning_rate": 8.521852059975743e-07, "loss": 0.0165, "step": 94410 }, { "grad_norm": 0.4591365456581116, "learning_rate": 8.491481646335231e-07, "loss": 0.0143, "step": 94420 }, { "grad_norm": 0.40766075253486633, "learning_rate": 8.46116498332733e-07, "loss": 0.0156, "step": 94430 }, { "grad_norm": 0.19426089525222778, "learning_rate": 8.430902074267333e-07, "loss": 0.0165, "step": 94440 }, { "grad_norm": 0.39440274238586426, "learning_rate": 8.400692922464925e-07, "loss": 0.0127, "step": 94450 }, { "grad_norm": 0.4713221490383148, "learning_rate": 8.370537531223577e-07, "loss": 0.0144, "step": 94460 }, { "grad_norm": 0.11045905202627182, "learning_rate": 8.340435903841092e-07, "loss": 0.017, "step": 94470 }, { "grad_norm": 0.22422583401203156, "learning_rate": 8.310388043609341e-07, "loss": 0.0172, "step": 94480 }, { "grad_norm": 0.21405139565467834, "learning_rate": 8.280393953814303e-07, "loss": 0.017, "step": 94490 }, { "grad_norm": 0.7723835706710815, "learning_rate": 8.250453637736077e-07, "loss": 0.0133, "step": 94500 }, { "grad_norm": 0.3844074308872223, "learning_rate": 8.22056709864899e-07, "loss": 0.0114, "step": 94510 }, { "grad_norm": 0.3343057930469513, "learning_rate": 8.190734339821204e-07, "loss": 0.018, "step": 94520 }, { "grad_norm": 0.3017273545265198, "learning_rate": 8.160955364515333e-07, "loss": 0.0081, "step": 94530 }, { "grad_norm": 0.27788954973220825, "learning_rate": 8.131230175987936e-07, "loss": 0.0137, "step": 94540 }, { "grad_norm": 0.14780430495738983, "learning_rate": 8.101558777489693e-07, "loss": 0.0175, "step": 94550 }, { "grad_norm": 0.2532782256603241, "learning_rate": 8.071941172265396e-07, "loss": 0.0227, "step": 94560 }, { "grad_norm": 0.4841706454753876, "learning_rate": 8.042377363553955e-07, "loss": 0.0162, "step": 94570 }, { "grad_norm": 0.3797214925289154, "learning_rate": 8.01286735458856e-07, "loss": 0.0165, "step": 94580 }, { "grad_norm": 0.4740033745765686, "learning_rate": 7.983411148596298e-07, "loss": 0.0195, "step": 94590 }, { "grad_norm": 0.1997562050819397, "learning_rate": 7.954008748798425e-07, "loss": 0.0121, "step": 94600 }, { "grad_norm": 0.3379037082195282, "learning_rate": 7.924660158410369e-07, "loss": 0.0154, "step": 94610 }, { "grad_norm": 0.20054756104946136, "learning_rate": 7.895365380641728e-07, "loss": 0.0268, "step": 94620 }, { "grad_norm": 0.3887692093849182, "learning_rate": 7.866124418695941e-07, "loss": 0.0195, "step": 94630 }, { "grad_norm": 0.19900445640087128, "learning_rate": 7.83693727577095e-07, "loss": 0.0103, "step": 94640 }, { "grad_norm": 0.16042213141918182, "learning_rate": 7.807803955058535e-07, "loss": 0.0126, "step": 94650 }, { "grad_norm": 0.45641857385635376, "learning_rate": 7.778724459744702e-07, "loss": 0.0179, "step": 94660 }, { "grad_norm": 0.6389561295509338, "learning_rate": 7.749698793009464e-07, "loss": 0.0217, "step": 94670 }, { "grad_norm": 0.3446081280708313, "learning_rate": 7.720726958027169e-07, "loss": 0.0133, "step": 94680 }, { "grad_norm": 0.24130496382713318, "learning_rate": 7.691808957965951e-07, "loss": 0.0138, "step": 94690 }, { "grad_norm": 0.12479418516159058, "learning_rate": 7.662944795988336e-07, "loss": 0.0158, "step": 94700 }, { "grad_norm": 0.24610207974910736, "learning_rate": 7.634134475250854e-07, "loss": 0.0129, "step": 94710 }, { "grad_norm": 0.3295854330062866, "learning_rate": 7.605377998904262e-07, "loss": 0.0162, "step": 94720 }, { "grad_norm": 0.15150286257266998, "learning_rate": 7.5766753700931e-07, "loss": 0.0102, "step": 94730 }, { "grad_norm": 0.26711079478263855, "learning_rate": 7.54802659195647e-07, "loss": 0.0132, "step": 94740 }, { "grad_norm": 0.1583961397409439, "learning_rate": 7.519431667627253e-07, "loss": 0.0134, "step": 94750 }, { "grad_norm": 0.1622675657272339, "learning_rate": 7.490890600232503e-07, "loss": 0.0157, "step": 94760 }, { "grad_norm": 0.2235860675573349, "learning_rate": 7.462403392893502e-07, "loss": 0.0275, "step": 94770 }, { "grad_norm": 0.2539883553981781, "learning_rate": 7.433970048725536e-07, "loss": 0.0164, "step": 94780 }, { "grad_norm": 0.9608976244926453, "learning_rate": 7.405590570838062e-07, "loss": 0.0137, "step": 94790 }, { "grad_norm": 0.5353975892066956, "learning_rate": 7.377264962334596e-07, "loss": 0.0161, "step": 94800 }, { "grad_norm": 0.7601494789123535, "learning_rate": 7.348993226312773e-07, "loss": 0.0199, "step": 94810 }, { "grad_norm": 0.23441195487976074, "learning_rate": 7.32077536586434e-07, "loss": 0.0182, "step": 94820 }, { "grad_norm": 0.1771000325679779, "learning_rate": 7.292611384075221e-07, "loss": 0.0155, "step": 94830 }, { "grad_norm": 0.4197017252445221, "learning_rate": 7.264501284025338e-07, "loss": 0.0146, "step": 94840 }, { "grad_norm": 0.21115076541900635, "learning_rate": 7.236445068788789e-07, "loss": 0.0171, "step": 94850 }, { "grad_norm": 0.20859552919864655, "learning_rate": 7.20844274143373e-07, "loss": 0.012, "step": 94860 }, { "grad_norm": 0.26224493980407715, "learning_rate": 7.180494305022489e-07, "loss": 0.0138, "step": 94870 }, { "grad_norm": 0.21227766573429108, "learning_rate": 7.152599762611456e-07, "loss": 0.0116, "step": 94880 }, { "grad_norm": 0.46788379549980164, "learning_rate": 7.124759117251079e-07, "loss": 0.0379, "step": 94890 }, { "grad_norm": 0.3309459388256073, "learning_rate": 7.096972371986033e-07, "loss": 0.0137, "step": 94900 }, { "grad_norm": 0.2632904052734375, "learning_rate": 7.069239529855054e-07, "loss": 0.0158, "step": 94910 }, { "grad_norm": 0.46371757984161377, "learning_rate": 7.041560593890939e-07, "loss": 0.0169, "step": 94920 }, { "grad_norm": 0.8067601919174194, "learning_rate": 7.013935567120599e-07, "loss": 0.0195, "step": 94930 }, { "grad_norm": 0.3065146803855896, "learning_rate": 6.986364452565009e-07, "loss": 0.0141, "step": 94940 }, { "grad_norm": 0.22205844521522522, "learning_rate": 6.95884725323942e-07, "loss": 0.0169, "step": 94950 }, { "grad_norm": 0.11605223268270493, "learning_rate": 6.931383972152982e-07, "loss": 0.0109, "step": 94960 }, { "grad_norm": 0.35551103949546814, "learning_rate": 6.903974612309127e-07, "loss": 0.0149, "step": 94970 }, { "grad_norm": 0.22844094038009644, "learning_rate": 6.876619176705179e-07, "loss": 0.0123, "step": 94980 }, { "grad_norm": 0.1927015632390976, "learning_rate": 6.8493176683328e-07, "loss": 0.0227, "step": 94990 }, { "grad_norm": 0.18020394444465637, "learning_rate": 6.822070090177657e-07, "loss": 0.0162, "step": 95000 }, { "grad_norm": 0.6105349063873291, "learning_rate": 6.794876445219367e-07, "loss": 0.0171, "step": 95010 }, { "grad_norm": 0.432445228099823, "learning_rate": 6.767736736431885e-07, "loss": 0.0235, "step": 95020 }, { "grad_norm": 0.4554072916507721, "learning_rate": 6.740650966783113e-07, "loss": 0.0169, "step": 95030 }, { "grad_norm": 0.368974894285202, "learning_rate": 6.713619139235239e-07, "loss": 0.0196, "step": 95040 }, { "grad_norm": 0.2855934798717499, "learning_rate": 6.686641256744286e-07, "loss": 0.0182, "step": 95050 }, { "grad_norm": 1.2135154008865356, "learning_rate": 6.659717322260562e-07, "loss": 0.0246, "step": 95060 }, { "grad_norm": 0.3623071014881134, "learning_rate": 6.632847338728377e-07, "loss": 0.0202, "step": 95070 }, { "grad_norm": 0.22050325572490692, "learning_rate": 6.60603130908627e-07, "loss": 0.0102, "step": 95080 }, { "grad_norm": 0.2557685673236847, "learning_rate": 6.579269236266783e-07, "loss": 0.0142, "step": 95090 }, { "grad_norm": 0.42096972465515137, "learning_rate": 6.552561123196632e-07, "loss": 0.0146, "step": 95100 }, { "grad_norm": 0.18676114082336426, "learning_rate": 6.525906972796425e-07, "loss": 0.018, "step": 95110 }, { "grad_norm": 0.32093653082847595, "learning_rate": 6.499306787981108e-07, "loss": 0.0187, "step": 95120 }, { "grad_norm": 0.27128854393959045, "learning_rate": 6.472760571659631e-07, "loss": 0.0119, "step": 95130 }, { "grad_norm": 0.4924694299697876, "learning_rate": 6.446268326735061e-07, "loss": 0.0168, "step": 95140 }, { "grad_norm": 0.3374573588371277, "learning_rate": 6.41983005610458e-07, "loss": 0.0127, "step": 95150 }, { "grad_norm": 0.2081744521856308, "learning_rate": 6.393445762659378e-07, "loss": 0.0228, "step": 95160 }, { "grad_norm": 0.3409888744354248, "learning_rate": 6.36711544928481e-07, "loss": 0.0129, "step": 95170 }, { "grad_norm": 0.44542255997657776, "learning_rate": 6.340839118860298e-07, "loss": 0.0211, "step": 95180 }, { "grad_norm": 0.26609349250793457, "learning_rate": 6.314616774259485e-07, "loss": 0.0121, "step": 95190 }, { "grad_norm": 0.6333463788032532, "learning_rate": 6.288448418349858e-07, "loss": 0.016, "step": 95200 }, { "grad_norm": 0.16896314918994904, "learning_rate": 6.262334053993235e-07, "loss": 0.0175, "step": 95210 }, { "grad_norm": 0.4257189631462097, "learning_rate": 6.236273684045501e-07, "loss": 0.0148, "step": 95220 }, { "grad_norm": 0.3652561604976654, "learning_rate": 6.210267311356488e-07, "loss": 0.0176, "step": 95230 }, { "grad_norm": 0.3336540758609772, "learning_rate": 6.184314938770197e-07, "loss": 0.0165, "step": 95240 }, { "grad_norm": 0.20356833934783936, "learning_rate": 6.158416569124858e-07, "loss": 0.0126, "step": 95250 }, { "grad_norm": 0.2949802875518799, "learning_rate": 6.132572205252541e-07, "loss": 0.0174, "step": 95260 }, { "grad_norm": 0.08396972715854645, "learning_rate": 6.106781849979648e-07, "loss": 0.0191, "step": 95270 }, { "grad_norm": 0.1640842854976654, "learning_rate": 6.081045506126482e-07, "loss": 0.0169, "step": 95280 }, { "grad_norm": 0.3758106827735901, "learning_rate": 6.055363176507678e-07, "loss": 0.0113, "step": 95290 }, { "grad_norm": 0.33716195821762085, "learning_rate": 6.029734863931602e-07, "loss": 0.0175, "step": 95300 }, { "grad_norm": 0.564011812210083, "learning_rate": 6.004160571201179e-07, "loss": 0.0185, "step": 95310 }, { "grad_norm": 0.11077670007944107, "learning_rate": 5.978640301112947e-07, "loss": 0.0124, "step": 95320 }, { "grad_norm": 0.3625495731830597, "learning_rate": 5.9531740564579e-07, "loss": 0.0162, "step": 95330 }, { "grad_norm": 0.19052855670452118, "learning_rate": 5.927761840020918e-07, "loss": 0.0106, "step": 95340 }, { "grad_norm": 0.3350904881954193, "learning_rate": 5.902403654581113e-07, "loss": 0.0141, "step": 95350 }, { "grad_norm": 0.37337246537208557, "learning_rate": 5.877099502911542e-07, "loss": 0.022, "step": 95360 }, { "grad_norm": 0.21677890419960022, "learning_rate": 5.851849387779496e-07, "loss": 0.0175, "step": 95370 }, { "grad_norm": 0.94814133644104, "learning_rate": 5.826653311946262e-07, "loss": 0.0336, "step": 95380 }, { "grad_norm": 0.1412220299243927, "learning_rate": 5.801511278167249e-07, "loss": 0.0144, "step": 95390 }, { "grad_norm": 0.2540881037712097, "learning_rate": 5.776423289191868e-07, "loss": 0.016, "step": 95400 }, { "grad_norm": 0.3867337107658386, "learning_rate": 5.751389347763869e-07, "loss": 0.0135, "step": 95410 }, { "grad_norm": 0.32731175422668457, "learning_rate": 5.726409456620841e-07, "loss": 0.0142, "step": 95420 }, { "grad_norm": 0.337197482585907, "learning_rate": 5.701483618494541e-07, "loss": 0.0125, "step": 95430 }, { "grad_norm": 0.296052485704422, "learning_rate": 5.67661183611079e-07, "loss": 0.0177, "step": 95440 }, { "grad_norm": 0.6298398971557617, "learning_rate": 5.651794112189579e-07, "loss": 0.0175, "step": 95450 }, { "grad_norm": 0.21363961696624756, "learning_rate": 5.627030449444903e-07, "loss": 0.0163, "step": 95460 }, { "grad_norm": 0.5307371020317078, "learning_rate": 5.60232085058493e-07, "loss": 0.0226, "step": 95470 }, { "grad_norm": 0.19667917490005493, "learning_rate": 5.577665318311776e-07, "loss": 0.017, "step": 95480 }, { "grad_norm": 0.16680195927619934, "learning_rate": 5.553063855321838e-07, "loss": 0.0206, "step": 95490 }, { "grad_norm": 0.23265819251537323, "learning_rate": 5.528516464305467e-07, "loss": 0.0213, "step": 95500 }, { "grad_norm": 0.5301753878593445, "learning_rate": 5.504023147947068e-07, "loss": 0.0198, "step": 95510 }, { "grad_norm": 0.5618141889572144, "learning_rate": 5.479583908925279e-07, "loss": 0.0251, "step": 95520 }, { "grad_norm": 0.23423156142234802, "learning_rate": 5.455198749912626e-07, "loss": 0.0139, "step": 95530 }, { "grad_norm": 0.4860513210296631, "learning_rate": 5.430867673575979e-07, "loss": 0.0248, "step": 95540 }, { "grad_norm": 0.23120811581611633, "learning_rate": 5.406590682575985e-07, "loss": 0.013, "step": 95550 }, { "grad_norm": 0.18971394002437592, "learning_rate": 5.382367779567687e-07, "loss": 0.0152, "step": 95560 }, { "grad_norm": 0.35308656096458435, "learning_rate": 5.358198967200023e-07, "loss": 0.0147, "step": 95570 }, { "grad_norm": 0.14200636744499207, "learning_rate": 5.334084248115989e-07, "loss": 0.0165, "step": 95580 }, { "grad_norm": 0.44144245982170105, "learning_rate": 5.310023624952754e-07, "loss": 0.0135, "step": 95590 }, { "grad_norm": 0.4630599915981293, "learning_rate": 5.286017100341656e-07, "loss": 0.0114, "step": 95600 }, { "grad_norm": 0.07169435918331146, "learning_rate": 5.262064676907874e-07, "loss": 0.0119, "step": 95610 }, { "grad_norm": 0.47979608178138733, "learning_rate": 5.238166357270924e-07, "loss": 0.0117, "step": 95620 }, { "grad_norm": 0.36257097125053406, "learning_rate": 5.214322144044215e-07, "loss": 0.0133, "step": 95630 }, { "grad_norm": 0.18362535536289215, "learning_rate": 5.190532039835327e-07, "loss": 0.0158, "step": 95640 }, { "grad_norm": 0.6004690527915955, "learning_rate": 5.166796047245903e-07, "loss": 0.0137, "step": 95650 }, { "grad_norm": 0.19964611530303955, "learning_rate": 5.143114168871699e-07, "loss": 0.014, "step": 95660 }, { "grad_norm": 0.5034099817276001, "learning_rate": 5.119486407302532e-07, "loss": 0.0208, "step": 95670 }, { "grad_norm": 0.43302401900291443, "learning_rate": 5.095912765122335e-07, "loss": 0.0254, "step": 95680 }, { "grad_norm": 0.56749027967453, "learning_rate": 5.07239324490899e-07, "loss": 0.0132, "step": 95690 }, { "grad_norm": 0.21495172381401062, "learning_rate": 5.048927849234608e-07, "loss": 0.021, "step": 95700 }, { "grad_norm": 0.5863638520240784, "learning_rate": 5.025516580665358e-07, "loss": 0.014, "step": 95710 }, { "grad_norm": 0.432475745677948, "learning_rate": 5.002159441761412e-07, "loss": 0.012, "step": 95720 }, { "grad_norm": 0.21304692327976227, "learning_rate": 4.978856435077062e-07, "loss": 0.0124, "step": 95730 }, { "grad_norm": 0.42888399958610535, "learning_rate": 4.955607563160769e-07, "loss": 0.012, "step": 95740 }, { "grad_norm": 0.22245387732982635, "learning_rate": 4.932412828554944e-07, "loss": 0.0193, "step": 95750 }, { "grad_norm": 0.2872715890407562, "learning_rate": 4.909272233796114e-07, "loss": 0.0171, "step": 95760 }, { "grad_norm": 0.3458092212677002, "learning_rate": 4.88618578141492e-07, "loss": 0.0135, "step": 95770 }, { "grad_norm": 0.12423616647720337, "learning_rate": 4.863153473936011e-07, "loss": 0.018, "step": 95780 }, { "grad_norm": 0.3126303255558014, "learning_rate": 4.840175313878259e-07, "loss": 0.0216, "step": 95790 }, { "grad_norm": 0.2611199915409088, "learning_rate": 4.817251303754433e-07, "loss": 0.0183, "step": 95800 }, { "grad_norm": 0.4110400378704071, "learning_rate": 4.794381446071583e-07, "loss": 0.0174, "step": 95810 }, { "grad_norm": 0.5577459931373596, "learning_rate": 4.771565743330597e-07, "loss": 0.0093, "step": 95820 }, { "grad_norm": 0.3866958022117615, "learning_rate": 4.7488041980266463e-07, "loss": 0.0162, "step": 95830 }, { "grad_norm": 0.20593932271003723, "learning_rate": 4.726096812648795e-07, "loss": 0.0177, "step": 95840 }, { "grad_norm": 0.16545969247817993, "learning_rate": 4.703443589680445e-07, "loss": 0.0149, "step": 95850 }, { "grad_norm": 0.6021845936775208, "learning_rate": 4.6808445315987826e-07, "loss": 0.02, "step": 95860 }, { "grad_norm": 0.3419414758682251, "learning_rate": 4.658299640875274e-07, "loss": 0.0125, "step": 95870 }, { "grad_norm": 0.13028965890407562, "learning_rate": 4.635808919975393e-07, "loss": 0.015, "step": 95880 }, { "grad_norm": 0.45395028591156006, "learning_rate": 4.6133723713586706e-07, "loss": 0.0203, "step": 95890 }, { "grad_norm": 0.2971809208393097, "learning_rate": 4.590989997478756e-07, "loss": 0.015, "step": 95900 }, { "grad_norm": 0.3015558421611786, "learning_rate": 4.5686618007833027e-07, "loss": 0.0204, "step": 95910 }, { "grad_norm": 0.389627069234848, "learning_rate": 4.54638778371419e-07, "loss": 0.0119, "step": 95920 }, { "grad_norm": 0.27104607224464417, "learning_rate": 4.524167948707192e-07, "loss": 0.0196, "step": 95930 }, { "grad_norm": 0.3622564375400543, "learning_rate": 4.5020022981921986e-07, "loss": 0.0128, "step": 95940 }, { "grad_norm": 0.43677958846092224, "learning_rate": 4.47989083459327e-07, "loss": 0.026, "step": 95950 }, { "grad_norm": 0.3260565400123596, "learning_rate": 4.457833560328528e-07, "loss": 0.0179, "step": 95960 }, { "grad_norm": 0.46049201488494873, "learning_rate": 4.4358304778100434e-07, "loss": 0.0168, "step": 95970 }, { "grad_norm": 0.27413588762283325, "learning_rate": 4.413881589444058e-07, "loss": 0.0187, "step": 95980 }, { "grad_norm": 0.21953988075256348, "learning_rate": 4.391986897630873e-07, "loss": 0.0106, "step": 95990 }, { "grad_norm": 0.2325897514820099, "learning_rate": 4.370146404764908e-07, "loss": 0.0271, "step": 96000 }, { "grad_norm": 0.3615853190422058, "learning_rate": 4.3483601132345285e-07, "loss": 0.0156, "step": 96010 }, { "grad_norm": 0.14687545597553253, "learning_rate": 4.32662802542233e-07, "loss": 0.0161, "step": 96020 }, { "grad_norm": 0.3136843740940094, "learning_rate": 4.3049501437047444e-07, "loss": 0.0121, "step": 96030 }, { "grad_norm": 0.20049870014190674, "learning_rate": 4.283326470452653e-07, "loss": 0.017, "step": 96040 }, { "grad_norm": 0.2860157787799835, "learning_rate": 4.2617570080306647e-07, "loss": 0.014, "step": 96050 }, { "grad_norm": 0.13962018489837646, "learning_rate": 4.2402417587975585e-07, "loss": 0.0154, "step": 96060 }, { "grad_norm": 0.40523967146873474, "learning_rate": 4.218780725106286e-07, "loss": 0.0165, "step": 96070 }, { "grad_norm": 0.4650585651397705, "learning_rate": 4.197373909303748e-07, "loss": 0.0215, "step": 96080 }, { "grad_norm": 0.3429529368877411, "learning_rate": 4.17602131373096e-07, "loss": 0.0123, "step": 96090 }, { "grad_norm": 0.3579123914241791, "learning_rate": 4.154722940723055e-07, "loss": 0.024, "step": 96100 }, { "grad_norm": 1.4130744934082031, "learning_rate": 4.1334787926091136e-07, "loss": 0.0128, "step": 96110 }, { "grad_norm": 0.9804251790046692, "learning_rate": 4.1122888717124444e-07, "loss": 0.0157, "step": 96120 }, { "grad_norm": 0.08668915927410126, "learning_rate": 4.091153180350249e-07, "loss": 0.0121, "step": 96130 }, { "grad_norm": 0.6738415956497192, "learning_rate": 4.0700717208340125e-07, "loss": 0.0144, "step": 96140 }, { "grad_norm": 0.41807928681373596, "learning_rate": 4.049044495469112e-07, "loss": 0.0145, "step": 96150 }, { "grad_norm": 0.24074077606201172, "learning_rate": 4.0280715065549866e-07, "loss": 0.0138, "step": 96160 }, { "grad_norm": 1.2003039121627808, "learning_rate": 4.0071527563853575e-07, "loss": 0.023, "step": 96170 }, { "grad_norm": 0.2919219136238098, "learning_rate": 3.9862882472477823e-07, "loss": 0.0137, "step": 96180 }, { "grad_norm": 0.28106623888015747, "learning_rate": 3.9654779814239375e-07, "loss": 0.0176, "step": 96190 }, { "grad_norm": 0.2541813552379608, "learning_rate": 3.944721961189668e-07, "loss": 0.013, "step": 96200 }, { "grad_norm": 0.1590230017900467, "learning_rate": 3.9240201888147697e-07, "loss": 0.0115, "step": 96210 }, { "grad_norm": 0.21584025025367737, "learning_rate": 3.9033726665632096e-07, "loss": 0.0154, "step": 96220 }, { "grad_norm": 0.25210148096084595, "learning_rate": 3.882779396692959e-07, "loss": 0.0136, "step": 96230 }, { "grad_norm": 0.18225717544555664, "learning_rate": 3.8622403814559947e-07, "loss": 0.0113, "step": 96240 }, { "grad_norm": 0.46198567748069763, "learning_rate": 3.841755623098575e-07, "loss": 0.0196, "step": 96250 }, { "grad_norm": 1.0039830207824707, "learning_rate": 3.821325123860742e-07, "loss": 0.0165, "step": 96260 }, { "grad_norm": 0.5573625564575195, "learning_rate": 3.800948885976818e-07, "loss": 0.0173, "step": 96270 }, { "grad_norm": 0.4057208299636841, "learning_rate": 3.780626911675078e-07, "loss": 0.0125, "step": 96280 }, { "grad_norm": 0.41053515672683716, "learning_rate": 3.76035920317791e-07, "loss": 0.015, "step": 96290 }, { "grad_norm": 0.2238132506608963, "learning_rate": 3.740145762701819e-07, "loss": 0.0143, "step": 96300 }, { "grad_norm": 0.6275013089179993, "learning_rate": 3.7199865924572585e-07, "loss": 0.013, "step": 96310 }, { "grad_norm": 0.33486178517341614, "learning_rate": 3.699881694648799e-07, "loss": 0.0117, "step": 96320 }, { "grad_norm": 0.2915463447570801, "learning_rate": 3.679831071475126e-07, "loss": 0.0199, "step": 96330 }, { "grad_norm": 0.19244584441184998, "learning_rate": 3.65983472512893e-07, "loss": 0.0172, "step": 96340 }, { "grad_norm": 0.43735358119010925, "learning_rate": 3.6398926577969616e-07, "loss": 0.0121, "step": 96350 }, { "grad_norm": 0.15810802578926086, "learning_rate": 3.620004871660032e-07, "loss": 0.015, "step": 96360 }, { "grad_norm": 0.1672372668981552, "learning_rate": 3.600171368893124e-07, "loss": 0.0149, "step": 96370 }, { "grad_norm": 0.1524013876914978, "learning_rate": 3.580392151665113e-07, "loss": 0.0142, "step": 96380 }, { "grad_norm": 0.1463165283203125, "learning_rate": 3.5606672221391024e-07, "loss": 0.0137, "step": 96390 }, { "grad_norm": 0.25019174814224243, "learning_rate": 3.540996582472089e-07, "loss": 0.013, "step": 96400 }, { "grad_norm": 0.2283647656440735, "learning_rate": 3.521380234815297e-07, "loss": 0.0126, "step": 96410 }, { "grad_norm": 0.39063549041748047, "learning_rate": 3.5018181813138985e-07, "loss": 0.0161, "step": 96420 }, { "grad_norm": 0.7390894889831543, "learning_rate": 3.482310424107238e-07, "loss": 0.0155, "step": 96430 }, { "grad_norm": 0.30785301327705383, "learning_rate": 3.4628569653285537e-07, "loss": 0.0188, "step": 96440 }, { "grad_norm": 0.16885408759117126, "learning_rate": 3.44345780710531e-07, "loss": 0.0172, "step": 96450 }, { "grad_norm": 0.6952768564224243, "learning_rate": 3.424112951558978e-07, "loss": 0.0176, "step": 96460 }, { "grad_norm": 0.2047518789768219, "learning_rate": 3.40482240080503e-07, "loss": 0.0143, "step": 96470 }, { "grad_norm": 0.256451815366745, "learning_rate": 3.3855861569531133e-07, "loss": 0.0128, "step": 96480 }, { "grad_norm": 0.4685869514942169, "learning_rate": 3.3664042221068226e-07, "loss": 0.0203, "step": 96490 }, { "grad_norm": 0.8866220116615295, "learning_rate": 3.347276598363924e-07, "loss": 0.0159, "step": 96500 }, { "grad_norm": 0.7308290004730225, "learning_rate": 3.328203287816134e-07, "loss": 0.0121, "step": 96510 }, { "grad_norm": 0.2525496482849121, "learning_rate": 3.3091842925493386e-07, "loss": 0.0176, "step": 96520 }, { "grad_norm": 0.6305260062217712, "learning_rate": 3.2902196146433196e-07, "loss": 0.0133, "step": 96530 }, { "grad_norm": 0.2938263714313507, "learning_rate": 3.2713092561720836e-07, "loss": 0.0136, "step": 96540 }, { "grad_norm": 0.47434699535369873, "learning_rate": 3.2524532192036996e-07, "loss": 0.0152, "step": 96550 }, { "grad_norm": 0.8741537928581238, "learning_rate": 3.2336515058001836e-07, "loss": 0.0142, "step": 96560 }, { "grad_norm": 1.1705132722854614, "learning_rate": 3.2149041180176143e-07, "loss": 0.0148, "step": 96570 }, { "grad_norm": 0.5130046606063843, "learning_rate": 3.1962110579062954e-07, "loss": 0.0195, "step": 96580 }, { "grad_norm": 0.10434477776288986, "learning_rate": 3.177572327510314e-07, "loss": 0.0095, "step": 96590 }, { "grad_norm": 0.14237648248672485, "learning_rate": 3.158987928868151e-07, "loss": 0.0159, "step": 96600 }, { "grad_norm": 1.4965095520019531, "learning_rate": 3.1404578640120144e-07, "loss": 0.0148, "step": 96610 }, { "grad_norm": 0.17297373712062836, "learning_rate": 3.1219821349683933e-07, "loss": 0.0096, "step": 96620 }, { "grad_norm": 0.3978443145751953, "learning_rate": 3.103560743757783e-07, "loss": 0.0194, "step": 96630 }, { "grad_norm": 0.27897754311561584, "learning_rate": 3.0851936923946834e-07, "loss": 0.0176, "step": 96640 }, { "grad_norm": 0.41679811477661133, "learning_rate": 3.06688098288771e-07, "loss": 0.0121, "step": 96650 }, { "grad_norm": 0.2173035889863968, "learning_rate": 3.0486226172394825e-07, "loss": 0.0167, "step": 96660 }, { "grad_norm": 0.12139467895030975, "learning_rate": 3.030418597446738e-07, "loss": 0.0155, "step": 96670 }, { "grad_norm": 0.19496898353099823, "learning_rate": 3.012268925500272e-07, "loss": 0.0181, "step": 96680 }, { "grad_norm": 0.5927493572235107, "learning_rate": 2.994173603384831e-07, "loss": 0.0163, "step": 96690 }, { "grad_norm": 0.07969536632299423, "learning_rate": 2.9761326330793317e-07, "loss": 0.0113, "step": 96700 }, { "grad_norm": 0.1984352469444275, "learning_rate": 2.958146016556751e-07, "loss": 0.0112, "step": 96710 }, { "grad_norm": 0.17371873557567596, "learning_rate": 2.94021375578396e-07, "loss": 0.0139, "step": 96720 }, { "grad_norm": 0.4087601602077484, "learning_rate": 2.9223358527221135e-07, "loss": 0.0178, "step": 96730 }, { "grad_norm": 0.7816320061683655, "learning_rate": 2.904512309326257e-07, "loss": 0.0142, "step": 96740 }, { "grad_norm": 0.2924428880214691, "learning_rate": 2.8867431275455545e-07, "loss": 0.0167, "step": 96750 }, { "grad_norm": 0.2019212394952774, "learning_rate": 2.86902830932323e-07, "loss": 0.0168, "step": 96760 }, { "grad_norm": 0.3774237334728241, "learning_rate": 2.851367856596565e-07, "loss": 0.0123, "step": 96770 }, { "grad_norm": 0.32539790868759155, "learning_rate": 2.83376177129685e-07, "loss": 0.0191, "step": 96780 }, { "grad_norm": 0.4625507891178131, "learning_rate": 2.8162100553494884e-07, "loss": 0.0139, "step": 96790 }, { "grad_norm": 0.24837292730808258, "learning_rate": 2.798712710673834e-07, "loss": 0.0144, "step": 96800 }, { "grad_norm": 0.22696326673030853, "learning_rate": 2.7812697391834676e-07, "loss": 0.0188, "step": 96810 }, { "grad_norm": 0.28615444898605347, "learning_rate": 2.7638811427858627e-07, "loss": 0.0131, "step": 96820 }, { "grad_norm": 0.39486604928970337, "learning_rate": 2.746546923382665e-07, "loss": 0.0169, "step": 96830 }, { "grad_norm": 0.39491504430770874, "learning_rate": 2.729267082869469e-07, "loss": 0.0149, "step": 96840 }, { "grad_norm": 0.2698011100292206, "learning_rate": 2.712041623135986e-07, "loss": 0.014, "step": 96850 }, { "grad_norm": 0.2644791007041931, "learning_rate": 2.694870546065931e-07, "loss": 0.0147, "step": 96860 }, { "grad_norm": 0.3236623704433441, "learning_rate": 2.677753853537246e-07, "loss": 0.0123, "step": 96870 }, { "grad_norm": 0.3404335081577301, "learning_rate": 2.660691547421601e-07, "loss": 0.016, "step": 96880 }, { "grad_norm": 0.31750452518463135, "learning_rate": 2.6436836295850033e-07, "loss": 0.0152, "step": 96890 }, { "grad_norm": 0.4445263147354126, "learning_rate": 2.6267301018874093e-07, "loss": 0.015, "step": 96900 }, { "grad_norm": 0.5730206370353699, "learning_rate": 2.6098309661828355e-07, "loss": 0.0173, "step": 96910 }, { "grad_norm": 0.26458483934402466, "learning_rate": 2.592986224319305e-07, "loss": 0.0106, "step": 96920 }, { "grad_norm": 0.7432689070701599, "learning_rate": 2.576195878139009e-07, "loss": 0.0203, "step": 96930 }, { "grad_norm": 0.16305766999721527, "learning_rate": 2.5594599294780364e-07, "loss": 0.0176, "step": 96940 }, { "grad_norm": 0.4434194266796112, "learning_rate": 2.542778380166644e-07, "loss": 0.0174, "step": 96950 }, { "grad_norm": 0.2821696102619171, "learning_rate": 2.526151232029095e-07, "loss": 0.0175, "step": 96960 }, { "grad_norm": 0.12031514197587967, "learning_rate": 2.5095784868836567e-07, "loss": 0.0142, "step": 96970 }, { "grad_norm": 0.3450433015823364, "learning_rate": 2.493060146542825e-07, "loss": 0.0142, "step": 96980 }, { "grad_norm": 0.35921257734298706, "learning_rate": 2.476596212812876e-07, "loss": 0.0108, "step": 96990 }, { "grad_norm": 0.4720989167690277, "learning_rate": 2.4601866874943703e-07, "loss": 0.0143, "step": 97000 }, { "grad_norm": 0.32436656951904297, "learning_rate": 2.4438315723818163e-07, "loss": 0.0137, "step": 97010 }, { "grad_norm": 0.15500469505786896, "learning_rate": 2.4275308692637833e-07, "loss": 0.0139, "step": 97020 }, { "grad_norm": 0.2687017619609833, "learning_rate": 2.411284579922901e-07, "loss": 0.0102, "step": 97030 }, { "grad_norm": 0.47142648696899414, "learning_rate": 2.395092706135804e-07, "loss": 0.0252, "step": 97040 }, { "grad_norm": 0.20530648529529572, "learning_rate": 2.3789552496731872e-07, "loss": 0.0152, "step": 97050 }, { "grad_norm": 0.3728860914707184, "learning_rate": 2.3628722122999158e-07, "loss": 0.0151, "step": 97060 }, { "grad_norm": 0.7616780400276184, "learning_rate": 2.3468435957747503e-07, "loss": 0.0121, "step": 97070 }, { "grad_norm": 0.2699974477291107, "learning_rate": 2.330869401850566e-07, "loss": 0.0143, "step": 97080 }, { "grad_norm": 0.44708624482154846, "learning_rate": 2.3149496322742437e-07, "loss": 0.0118, "step": 97090 }, { "grad_norm": 0.4387747049331665, "learning_rate": 2.2990842887868346e-07, "loss": 0.0199, "step": 97100 }, { "grad_norm": 0.47290998697280884, "learning_rate": 2.2832733731232292e-07, "loss": 0.0175, "step": 97110 }, { "grad_norm": 0.31839314103126526, "learning_rate": 2.2675168870124886e-07, "loss": 0.0166, "step": 97120 }, { "grad_norm": 0.22449736297130585, "learning_rate": 2.2518148321778455e-07, "loss": 0.0104, "step": 97130 }, { "grad_norm": 0.21656756103038788, "learning_rate": 2.2361672103363706e-07, "loss": 0.0187, "step": 97140 }, { "grad_norm": 0.3490493893623352, "learning_rate": 2.220574023199251e-07, "loss": 0.0143, "step": 97150 }, { "grad_norm": 0.14805898070335388, "learning_rate": 2.2050352724717894e-07, "loss": 0.0187, "step": 97160 }, { "grad_norm": 0.42315995693206787, "learning_rate": 2.1895509598532372e-07, "loss": 0.0135, "step": 97170 }, { "grad_norm": 0.3305777907371521, "learning_rate": 2.1741210870369066e-07, "loss": 0.0124, "step": 97180 }, { "grad_norm": 0.17857074737548828, "learning_rate": 2.1587456557102814e-07, "loss": 0.014, "step": 97190 }, { "grad_norm": 0.16331583261489868, "learning_rate": 2.143424667554683e-07, "loss": 0.0155, "step": 97200 }, { "grad_norm": 0.1542612463235855, "learning_rate": 2.128158124245716e-07, "loss": 0.0138, "step": 97210 }, { "grad_norm": 0.19006706774234772, "learning_rate": 2.1129460274527668e-07, "loss": 0.0105, "step": 97220 }, { "grad_norm": 0.4933798015117645, "learning_rate": 2.0977883788395046e-07, "loss": 0.0121, "step": 97230 }, { "grad_norm": 0.44861045479774475, "learning_rate": 2.0826851800634928e-07, "loss": 0.0142, "step": 97240 }, { "grad_norm": 0.27298635244369507, "learning_rate": 2.0676364327764653e-07, "loss": 0.0167, "step": 97250 }, { "grad_norm": 2.3056976795196533, "learning_rate": 2.0526421386240502e-07, "loss": 0.0163, "step": 97260 }, { "grad_norm": 0.46549105644226074, "learning_rate": 2.0377022992461025e-07, "loss": 0.0205, "step": 97270 }, { "grad_norm": 0.4577309191226959, "learning_rate": 2.0228169162763156e-07, "loss": 0.0127, "step": 97280 }, { "grad_norm": 0.8144041299819946, "learning_rate": 2.0079859913425536e-07, "loss": 0.0133, "step": 97290 }, { "grad_norm": 0.21364229917526245, "learning_rate": 1.9932095260667417e-07, "loss": 0.0153, "step": 97300 }, { "grad_norm": 0.3188936114311218, "learning_rate": 1.9784875220648092e-07, "loss": 0.0148, "step": 97310 }, { "grad_norm": 0.5808101892471313, "learning_rate": 1.9638199809466907e-07, "loss": 0.016, "step": 97320 }, { "grad_norm": 0.9468380212783813, "learning_rate": 1.9492069043164364e-07, "loss": 0.0184, "step": 97330 }, { "grad_norm": 0.3033680319786072, "learning_rate": 1.9346482937721567e-07, "loss": 0.0113, "step": 97340 }, { "grad_norm": 0.33420059084892273, "learning_rate": 1.920144150905856e-07, "loss": 0.0141, "step": 97350 }, { "grad_norm": 0.22299489378929138, "learning_rate": 1.9056944773037655e-07, "loss": 0.0138, "step": 97360 }, { "grad_norm": 0.2500461935997009, "learning_rate": 1.8912992745460102e-07, "loss": 0.013, "step": 97370 }, { "grad_norm": 1.2670691013336182, "learning_rate": 1.876958544206886e-07, "loss": 0.0144, "step": 97380 }, { "grad_norm": 0.19320261478424072, "learning_rate": 1.8626722878546942e-07, "loss": 0.0195, "step": 97390 }, { "grad_norm": 0.30269402265548706, "learning_rate": 1.8484405070516851e-07, "loss": 0.0133, "step": 97400 }, { "grad_norm": 0.5865426659584045, "learning_rate": 1.8342632033542805e-07, "loss": 0.0172, "step": 97410 }, { "grad_norm": 1.6026524305343628, "learning_rate": 1.8201403783129066e-07, "loss": 0.0179, "step": 97420 }, { "grad_norm": 0.20357844233512878, "learning_rate": 1.806072033471884e-07, "loss": 0.0134, "step": 97430 }, { "grad_norm": 0.36015453934669495, "learning_rate": 1.79205817036987e-07, "loss": 0.0119, "step": 97440 }, { "grad_norm": 0.5066429972648621, "learning_rate": 1.7780987905393065e-07, "loss": 0.0176, "step": 97450 }, { "grad_norm": 0.45509788393974304, "learning_rate": 1.764193895506805e-07, "loss": 0.0147, "step": 97460 }, { "grad_norm": 0.15908244252204895, "learning_rate": 1.7503434867929824e-07, "loss": 0.0139, "step": 97470 }, { "grad_norm": 0.45455971360206604, "learning_rate": 1.7365475659125165e-07, "loss": 0.0166, "step": 97480 }, { "grad_norm": 0.12004607915878296, "learning_rate": 1.7228061343740332e-07, "loss": 0.016, "step": 97490 }, { "grad_norm": 0.19289255142211914, "learning_rate": 1.7091191936803308e-07, "loss": 0.008, "step": 97500 }, { "grad_norm": 0.37219539284706116, "learning_rate": 1.6954867453281563e-07, "loss": 0.0127, "step": 97510 }, { "grad_norm": 0.12034860998392105, "learning_rate": 1.6819087908084285e-07, "loss": 0.0144, "step": 97520 }, { "grad_norm": 0.18409965932369232, "learning_rate": 1.6683853316058485e-07, "loss": 0.0127, "step": 97530 }, { "grad_norm": 1.0946558713912964, "learning_rate": 1.6549163691995106e-07, "loss": 0.0184, "step": 97540 }, { "grad_norm": 0.3162766993045807, "learning_rate": 1.641501905062237e-07, "loss": 0.0217, "step": 97550 }, { "grad_norm": 0.1928238570690155, "learning_rate": 1.628141940661021e-07, "loss": 0.0195, "step": 97560 }, { "grad_norm": 0.14399002492427826, "learning_rate": 1.6148364774569158e-07, "loss": 0.0113, "step": 97570 }, { "grad_norm": 0.3540320098400116, "learning_rate": 1.6015855169050355e-07, "loss": 0.0219, "step": 97580 }, { "grad_norm": 0.3273475468158722, "learning_rate": 1.5883890604543873e-07, "loss": 0.0144, "step": 97590 }, { "grad_norm": 0.659442663192749, "learning_rate": 1.5752471095482057e-07, "loss": 0.0096, "step": 97600 }, { "grad_norm": 0.2983357608318329, "learning_rate": 1.5621596656235638e-07, "loss": 0.012, "step": 97610 }, { "grad_norm": 0.39984768629074097, "learning_rate": 1.5491267301117608e-07, "loss": 0.0189, "step": 97620 }, { "grad_norm": 0.3558887243270874, "learning_rate": 1.536148304438101e-07, "loss": 0.0157, "step": 97630 }, { "grad_norm": 0.2073022872209549, "learning_rate": 1.5232243900217823e-07, "loss": 0.0167, "step": 97640 }, { "grad_norm": 0.4754290282726288, "learning_rate": 1.51035498827623e-07, "loss": 0.0166, "step": 97650 }, { "grad_norm": 0.35873880982398987, "learning_rate": 1.497540100608763e-07, "loss": 0.0175, "step": 97660 }, { "grad_norm": 0.39466676115989685, "learning_rate": 1.4847797284208153e-07, "loss": 0.0239, "step": 97670 }, { "grad_norm": 0.40577593445777893, "learning_rate": 1.472073873107882e-07, "loss": 0.0252, "step": 97680 }, { "grad_norm": 0.6960476636886597, "learning_rate": 1.459422536059407e-07, "loss": 0.0195, "step": 97690 }, { "grad_norm": 0.24550803005695343, "learning_rate": 1.4468257186589508e-07, "loss": 0.0075, "step": 97700 }, { "grad_norm": 0.4776536226272583, "learning_rate": 1.4342834222840217e-07, "loss": 0.014, "step": 97710 }, { "grad_norm": 0.4286419451236725, "learning_rate": 1.4217956483063566e-07, "loss": 0.0101, "step": 97720 }, { "grad_norm": 0.14251722395420074, "learning_rate": 1.409362398091474e-07, "loss": 0.0145, "step": 97730 }, { "grad_norm": 0.9132192730903625, "learning_rate": 1.3969836729990638e-07, "loss": 0.0147, "step": 97740 }, { "grad_norm": 0.20743022859096527, "learning_rate": 1.384659474382932e-07, "loss": 0.0145, "step": 97750 }, { "grad_norm": 0.2831834852695465, "learning_rate": 1.372389803590779e-07, "loss": 0.0134, "step": 97760 }, { "grad_norm": 0.32042327523231506, "learning_rate": 1.360174661964364e-07, "loss": 0.0175, "step": 97770 }, { "grad_norm": 0.42198047041893005, "learning_rate": 1.3480140508396188e-07, "loss": 0.0145, "step": 97780 }, { "grad_norm": 0.32358551025390625, "learning_rate": 1.335907971546313e-07, "loss": 0.0167, "step": 97790 }, { "grad_norm": 0.4198024272918701, "learning_rate": 1.3238564254083875e-07, "loss": 0.0217, "step": 97800 }, { "grad_norm": 0.915671169757843, "learning_rate": 1.3118594137437323e-07, "loss": 0.0209, "step": 97810 }, { "grad_norm": 0.40392881631851196, "learning_rate": 1.2999169378644094e-07, "loss": 0.0164, "step": 97820 }, { "grad_norm": 0.4138081669807434, "learning_rate": 1.2880289990763738e-07, "loss": 0.0206, "step": 97830 }, { "grad_norm": 3.200657606124878, "learning_rate": 1.2761955986796968e-07, "loss": 0.0167, "step": 97840 }, { "grad_norm": 0.36041495203971863, "learning_rate": 1.264416737968399e-07, "loss": 0.0212, "step": 97850 }, { "grad_norm": 0.39213263988494873, "learning_rate": 1.2526924182307275e-07, "loss": 0.0132, "step": 97860 }, { "grad_norm": 0.19518499076366425, "learning_rate": 1.2410226407487124e-07, "loss": 0.0137, "step": 97870 }, { "grad_norm": 0.13895267248153687, "learning_rate": 1.229407406798555e-07, "loss": 0.0129, "step": 97880 }, { "grad_norm": 0.19240213930606842, "learning_rate": 1.2178467176505726e-07, "loss": 0.0158, "step": 97890 }, { "grad_norm": 0.6125602722167969, "learning_rate": 1.2063405745689205e-07, "loss": 0.013, "step": 97900 }, { "grad_norm": 0.5558791756629944, "learning_rate": 1.1948889788119812e-07, "loss": 0.0149, "step": 97910 }, { "grad_norm": 0.21131667494773865, "learning_rate": 1.1834919316320303e-07, "loss": 0.0154, "step": 97920 }, { "grad_norm": 0.11271990090608597, "learning_rate": 1.1721494342754048e-07, "loss": 0.0141, "step": 97930 }, { "grad_norm": 0.4305092990398407, "learning_rate": 1.1608614879825563e-07, "loss": 0.0142, "step": 97940 }, { "grad_norm": 0.3519280254840851, "learning_rate": 1.1496280939879422e-07, "loss": 0.0187, "step": 97950 }, { "grad_norm": 0.5067976117134094, "learning_rate": 1.1384492535199687e-07, "loss": 0.0144, "step": 97960 }, { "grad_norm": 0.4114227294921875, "learning_rate": 1.1273249678011578e-07, "loss": 0.0165, "step": 97970 }, { "grad_norm": 0.8654709458351135, "learning_rate": 1.1162552380480362e-07, "loss": 0.0112, "step": 97980 }, { "grad_norm": 0.3444584906101227, "learning_rate": 1.1052400654711359e-07, "loss": 0.0142, "step": 97990 }, { "grad_norm": 0.1475488245487213, "learning_rate": 1.0942794512751597e-07, "loss": 0.014, "step": 98000 }, { "grad_norm": 0.2223569005727768, "learning_rate": 1.0833733966587045e-07, "loss": 0.0118, "step": 98010 }, { "grad_norm": 0.3326061964035034, "learning_rate": 1.0725219028143718e-07, "loss": 0.0132, "step": 98020 }, { "grad_norm": 0.2289688140153885, "learning_rate": 1.0617249709289345e-07, "loss": 0.0124, "step": 98030 }, { "grad_norm": 0.16993702948093414, "learning_rate": 1.0509826021831703e-07, "loss": 0.0113, "step": 98040 }, { "grad_norm": 1.0465507507324219, "learning_rate": 1.0402947977517508e-07, "loss": 0.0152, "step": 98050 }, { "grad_norm": 0.33440446853637695, "learning_rate": 1.0296615588035185e-07, "loss": 0.0153, "step": 98060 }, { "grad_norm": 0.3455023765563965, "learning_rate": 1.0190828865012659e-07, "loss": 0.0124, "step": 98070 }, { "grad_norm": 0.5409400463104248, "learning_rate": 1.008558782001956e-07, "loss": 0.0183, "step": 98080 }, { "grad_norm": 0.3810484707355499, "learning_rate": 9.980892464563906e-08, "loss": 0.0177, "step": 98090 }, { "grad_norm": 0.35610681772232056, "learning_rate": 9.876742810095985e-08, "loss": 0.0138, "step": 98100 }, { "grad_norm": 0.2878330945968628, "learning_rate": 9.77313886800446e-08, "loss": 0.0165, "step": 98110 }, { "grad_norm": 0.28078168630599976, "learning_rate": 9.670080649619717e-08, "loss": 0.0205, "step": 98120 }, { "grad_norm": 0.2939000129699707, "learning_rate": 9.567568166212737e-08, "loss": 0.0139, "step": 98130 }, { "grad_norm": 0.6829323172569275, "learning_rate": 9.465601428992888e-08, "loss": 0.0159, "step": 98140 }, { "grad_norm": 0.47151979804039, "learning_rate": 9.364180449111803e-08, "loss": 0.0162, "step": 98150 }, { "grad_norm": 0.22918497025966644, "learning_rate": 9.263305237661169e-08, "loss": 0.0158, "step": 98160 }, { "grad_norm": 0.30858343839645386, "learning_rate": 9.162975805671603e-08, "loss": 0.0117, "step": 98170 }, { "grad_norm": 0.33432701230049133, "learning_rate": 9.063192164115442e-08, "loss": 0.0124, "step": 98180 }, { "grad_norm": 0.6258598566055298, "learning_rate": 8.963954323904511e-08, "loss": 0.0161, "step": 98190 }, { "grad_norm": 0.6384932994842529, "learning_rate": 8.865262295891796e-08, "loss": 0.0149, "step": 98200 }, { "grad_norm": 0.2041955441236496, "learning_rate": 8.767116090870331e-08, "loss": 0.0112, "step": 98210 }, { "grad_norm": 0.19050414860248566, "learning_rate": 8.669515719572086e-08, "loss": 0.0154, "step": 98220 }, { "grad_norm": 0.530065655708313, "learning_rate": 8.572461192671855e-08, "loss": 0.0157, "step": 98230 }, { "grad_norm": 0.30809009075164795, "learning_rate": 8.475952520782815e-08, "loss": 0.0207, "step": 98240 }, { "grad_norm": 0.3291965425014496, "learning_rate": 8.379989714458747e-08, "loss": 0.0123, "step": 98250 }, { "grad_norm": 0.2652263641357422, "learning_rate": 8.284572784194034e-08, "loss": 0.0128, "step": 98260 }, { "grad_norm": 0.20299196243286133, "learning_rate": 8.189701740424216e-08, "loss": 0.0107, "step": 98270 }, { "grad_norm": 0.37625786662101746, "learning_rate": 8.095376593522663e-08, "loss": 0.0166, "step": 98280 }, { "grad_norm": 0.17939041554927826, "learning_rate": 8.001597353806123e-08, "loss": 0.0164, "step": 98290 }, { "grad_norm": 0.20584997534751892, "learning_rate": 7.90836403152917e-08, "loss": 0.0151, "step": 98300 }, { "grad_norm": 0.35080358386039734, "learning_rate": 7.815676636888092e-08, "loss": 0.0118, "step": 98310 }, { "grad_norm": 0.8087042570114136, "learning_rate": 7.723535180019226e-08, "loss": 0.0125, "step": 98320 }, { "grad_norm": 0.17975357174873352, "learning_rate": 7.6319396709984e-08, "loss": 0.021, "step": 98330 }, { "grad_norm": 0.16338545083999634, "learning_rate": 7.540890119843158e-08, "loss": 0.0165, "step": 98340 }, { "grad_norm": 0.3168964684009552, "learning_rate": 7.450386536509979e-08, "loss": 0.0144, "step": 98350 }, { "grad_norm": 0.22154207527637482, "learning_rate": 7.360428930895947e-08, "loss": 0.024, "step": 98360 }, { "grad_norm": 0.13114005327224731, "learning_rate": 7.271017312839302e-08, "loss": 0.0221, "step": 98370 }, { "grad_norm": 0.3118182420730591, "learning_rate": 7.182151692118333e-08, "loss": 0.0157, "step": 98380 }, { "grad_norm": 0.4468477666378021, "learning_rate": 7.093832078449713e-08, "loss": 0.0147, "step": 98390 }, { "grad_norm": 0.2707465589046478, "learning_rate": 7.006058481493493e-08, "loss": 0.0164, "step": 98400 }, { "grad_norm": 0.6185228824615479, "learning_rate": 6.918830910847552e-08, "loss": 0.019, "step": 98410 }, { "grad_norm": 0.20988821983337402, "learning_rate": 6.832149376052033e-08, "loss": 0.0133, "step": 98420 }, { "grad_norm": 0.3234688937664032, "learning_rate": 6.746013886584912e-08, "loss": 0.0238, "step": 98430 }, { "grad_norm": 0.2984263300895691, "learning_rate": 6.660424451866431e-08, "loss": 0.0168, "step": 98440 }, { "grad_norm": 0.1771582067012787, "learning_rate": 6.575381081256327e-08, "loss": 0.0161, "step": 98450 }, { "grad_norm": 0.2786928713321686, "learning_rate": 6.49088378405549e-08, "loss": 0.0191, "step": 98460 }, { "grad_norm": 0.1562507450580597, "learning_rate": 6.406932569503755e-08, "loss": 0.0191, "step": 98470 }, { "grad_norm": 0.6413886547088623, "learning_rate": 6.323527446782662e-08, "loss": 0.0176, "step": 98480 }, { "grad_norm": 0.6371320486068726, "learning_rate": 6.240668425012142e-08, "loss": 0.0211, "step": 98490 }, { "grad_norm": 0.46048733592033386, "learning_rate": 6.158355513254388e-08, "loss": 0.0132, "step": 98500 }, { "grad_norm": 0.30024728178977966, "learning_rate": 6.076588720510534e-08, "loss": 0.0158, "step": 98510 }, { "grad_norm": 0.17738202214241028, "learning_rate": 5.995368055722316e-08, "loss": 0.0109, "step": 98520 }, { "grad_norm": 0.35339924693107605, "learning_rate": 5.914693527773185e-08, "loss": 0.0117, "step": 98530 }, { "grad_norm": 0.24893109500408173, "learning_rate": 5.834565145483861e-08, "loss": 0.013, "step": 98540 }, { "grad_norm": 0.1463686227798462, "learning_rate": 5.754982917618446e-08, "loss": 0.0113, "step": 98550 }, { "grad_norm": 0.8346195816993713, "learning_rate": 5.675946852878866e-08, "loss": 0.0113, "step": 98560 }, { "grad_norm": 0.35679855942726135, "learning_rate": 5.59745695990932e-08, "loss": 0.0169, "step": 98570 }, { "grad_norm": 0.20409877598285675, "learning_rate": 5.51951324729294e-08, "loss": 0.0183, "step": 98580 }, { "grad_norm": 0.3763262927532196, "learning_rate": 5.4421157235529093e-08, "loss": 0.0122, "step": 98590 }, { "grad_norm": 0.3465600311756134, "learning_rate": 5.365264397154679e-08, "loss": 0.017, "step": 98600 }, { "grad_norm": 0.2526986598968506, "learning_rate": 5.288959276501526e-08, "loss": 0.0133, "step": 98610 }, { "grad_norm": 1.6555817127227783, "learning_rate": 5.2132003699378876e-08, "loss": 0.0135, "step": 98620 }, { "grad_norm": 0.4622611999511719, "learning_rate": 5.137987685749912e-08, "loss": 0.0177, "step": 98630 }, { "grad_norm": 0.2686786949634552, "learning_rate": 5.0633212321610227e-08, "loss": 0.0147, "step": 98640 }, { "grad_norm": 0.48906001448631287, "learning_rate": 4.989201017338019e-08, "loss": 0.0137, "step": 98650 }, { "grad_norm": 0.33192744851112366, "learning_rate": 4.9156270493860846e-08, "loss": 0.0189, "step": 98660 }, { "grad_norm": 0.21802930533885956, "learning_rate": 4.842599336351561e-08, "loss": 0.0133, "step": 98670 }, { "grad_norm": 0.6815438270568848, "learning_rate": 4.7701178862197274e-08, "loss": 0.0126, "step": 98680 }, { "grad_norm": 0.36007145047187805, "learning_rate": 4.698182706918131e-08, "loss": 0.0151, "step": 98690 }, { "grad_norm": 0.21551063656806946, "learning_rate": 4.6267938063121466e-08, "loss": 0.0181, "step": 98700 }, { "grad_norm": 0.14905554056167603, "learning_rate": 4.555951192209973e-08, "loss": 0.0144, "step": 98710 }, { "grad_norm": 0.2325766682624817, "learning_rate": 4.485654872358747e-08, "loss": 0.0102, "step": 98720 }, { "grad_norm": 0.5585840344429016, "learning_rate": 4.415904854445097e-08, "loss": 0.0167, "step": 98730 }, { "grad_norm": 0.9609202742576599, "learning_rate": 4.34670114609792e-08, "loss": 0.0142, "step": 98740 }, { "grad_norm": 0.4411975145339966, "learning_rate": 4.278043754884498e-08, "loss": 0.0177, "step": 98750 }, { "grad_norm": 0.3495630919933319, "learning_rate": 4.209932688312712e-08, "loss": 0.0213, "step": 98760 }, { "grad_norm": 0.42373794317245483, "learning_rate": 4.14236795383216e-08, "loss": 0.0193, "step": 98770 }, { "grad_norm": 0.3816772699356079, "learning_rate": 4.075349558830821e-08, "loss": 0.0151, "step": 98780 }, { "grad_norm": 0.25254711508750916, "learning_rate": 4.008877510638387e-08, "loss": 0.0171, "step": 98790 }, { "grad_norm": 0.6661978363990784, "learning_rate": 3.942951816523488e-08, "loss": 0.0136, "step": 98800 }, { "grad_norm": 0.3836176097393036, "learning_rate": 3.8775724836959125e-08, "loss": 0.0147, "step": 98810 }, { "grad_norm": 0.2752208411693573, "learning_rate": 3.812739519305497e-08, "loss": 0.0141, "step": 98820 }, { "grad_norm": 0.3085440993309021, "learning_rate": 3.748452930442126e-08, "loss": 0.0147, "step": 98830 }, { "grad_norm": 0.44936591386795044, "learning_rate": 3.6847127241362855e-08, "loss": 0.0195, "step": 98840 }, { "grad_norm": 0.1943357139825821, "learning_rate": 3.621518907358512e-08, "loss": 0.0206, "step": 98850 }, { "grad_norm": 0.4660629332065582, "learning_rate": 3.558871487019388e-08, "loss": 0.0135, "step": 98860 }, { "grad_norm": 0.3762430250644684, "learning_rate": 3.4967704699701007e-08, "loss": 0.0163, "step": 98870 }, { "grad_norm": 0.26988229155540466, "learning_rate": 3.435215863001884e-08, "loss": 0.0157, "step": 98880 }, { "grad_norm": 0.20270168781280518, "learning_rate": 3.374207672846019e-08, "loss": 0.0176, "step": 98890 }, { "grad_norm": 0.290383517742157, "learning_rate": 3.313745906174392e-08, "loss": 0.0157, "step": 98900 }, { "grad_norm": 0.23588506877422333, "learning_rate": 3.253830569599492e-08, "loss": 0.0096, "step": 98910 }, { "grad_norm": 0.2210855484008789, "learning_rate": 3.1944616696727436e-08, "loss": 0.0127, "step": 98920 }, { "grad_norm": 0.24882295727729797, "learning_rate": 3.135639212887287e-08, "loss": 0.015, "step": 98930 }, { "grad_norm": 0.22041930258274078, "learning_rate": 3.077363205675754e-08, "loss": 0.0138, "step": 98940 }, { "grad_norm": 0.6228664517402649, "learning_rate": 3.0196336544113804e-08, "loss": 0.0159, "step": 98950 }, { "grad_norm": 0.35254958271980286, "learning_rate": 2.9624505654063383e-08, "loss": 0.0112, "step": 98960 }, { "grad_norm": 0.3198666274547577, "learning_rate": 2.90581394491507e-08, "loss": 0.0194, "step": 98970 }, { "grad_norm": 0.32124099135398865, "learning_rate": 2.8497237991309545e-08, "loss": 0.0106, "step": 98980 }, { "grad_norm": 0.3246800899505615, "learning_rate": 2.7941801341879735e-08, "loss": 0.0153, "step": 98990 }, { "grad_norm": 0.24242694675922394, "learning_rate": 2.7391829561601578e-08, "loss": 0.0146, "step": 99000 }, { "grad_norm": 0.31045371294021606, "learning_rate": 2.6847322710621403e-08, "loss": 0.0126, "step": 99010 }, { "grad_norm": 0.2611887753009796, "learning_rate": 2.6308280848486022e-08, "loss": 0.0138, "step": 99020 }, { "grad_norm": 0.20515544712543488, "learning_rate": 2.5774704034137177e-08, "loss": 0.0146, "step": 99030 }, { "grad_norm": 0.28699421882629395, "learning_rate": 2.524659232593929e-08, "loss": 0.0145, "step": 99040 }, { "grad_norm": 0.4937914311885834, "learning_rate": 2.472394578163506e-08, "loss": 0.018, "step": 99050 }, { "grad_norm": 0.5241883397102356, "learning_rate": 2.4206764458378772e-08, "loss": 0.0179, "step": 99060 }, { "grad_norm": 0.5932901501655579, "learning_rate": 2.3695048412736285e-08, "loss": 0.0186, "step": 99070 }, { "grad_norm": 0.3689152002334595, "learning_rate": 2.3188797700668395e-08, "loss": 0.0194, "step": 99080 }, { "grad_norm": 0.18424001336097717, "learning_rate": 2.268801237753082e-08, "loss": 0.0121, "step": 99090 }, { "grad_norm": 0.11286463588476181, "learning_rate": 2.2192692498090862e-08, "loss": 0.0159, "step": 99100 }, { "grad_norm": 0.2882486879825592, "learning_rate": 2.170283811652185e-08, "loss": 0.0244, "step": 99110 }, { "grad_norm": 0.6118146777153015, "learning_rate": 2.12184492863865e-08, "loss": 0.0136, "step": 99120 }, { "grad_norm": 0.33049628138542175, "learning_rate": 2.073952606066465e-08, "loss": 0.0165, "step": 99130 }, { "grad_norm": 0.22107017040252686, "learning_rate": 2.026606849171997e-08, "loss": 0.0203, "step": 99140 }, { "grad_norm": 0.25210946798324585, "learning_rate": 1.9798076631333262e-08, "loss": 0.0149, "step": 99150 }, { "grad_norm": 0.7179688811302185, "learning_rate": 1.933555053069136e-08, "loss": 0.011, "step": 99160 }, { "grad_norm": 0.29012125730514526, "learning_rate": 1.887849024036492e-08, "loss": 0.0139, "step": 99170 }, { "grad_norm": 0.18990378081798553, "learning_rate": 1.8426895810341736e-08, "loss": 0.0113, "step": 99180 }, { "grad_norm": 0.5329614281654358, "learning_rate": 1.798076729000453e-08, "loss": 0.0153, "step": 99190 }, { "grad_norm": 0.2549789547920227, "learning_rate": 1.75401047281476e-08, "loss": 0.0194, "step": 99200 }, { "grad_norm": 0.22457440197467804, "learning_rate": 1.7104908172954626e-08, "loss": 0.0144, "step": 99210 }, { "grad_norm": 0.34666579961776733, "learning_rate": 1.6675177672020868e-08, "loss": 0.0094, "step": 99220 }, { "grad_norm": 0.5495458245277405, "learning_rate": 1.6250913272342073e-08, "loss": 0.017, "step": 99230 }, { "grad_norm": 0.2497594952583313, "learning_rate": 1.5832115020314453e-08, "loss": 0.0111, "step": 99240 }, { "grad_norm": 0.2048935443162918, "learning_rate": 1.5418782961734712e-08, "loss": 0.0145, "step": 99250 }, { "grad_norm": 1.2080563306808472, "learning_rate": 1.501091714181113e-08, "loss": 0.0207, "step": 99260 }, { "grad_norm": 0.2612110376358032, "learning_rate": 1.460851760513582e-08, "loss": 0.0163, "step": 99270 }, { "grad_norm": 0.1742517352104187, "learning_rate": 1.4211584395723566e-08, "loss": 0.0157, "step": 99280 }, { "grad_norm": 0.2093506008386612, "learning_rate": 1.382011755697854e-08, "loss": 0.02, "step": 99290 }, { "grad_norm": 0.2695150375366211, "learning_rate": 1.3434117131716494e-08, "loss": 0.0151, "step": 99300 }, { "grad_norm": 0.17909494042396545, "learning_rate": 1.3053583162142557e-08, "loss": 0.0165, "step": 99310 }, { "grad_norm": 0.4120214879512787, "learning_rate": 1.2678515689873438e-08, "loss": 0.0163, "step": 99320 }, { "grad_norm": 0.3288503587245941, "learning_rate": 1.2308914755931877e-08, "loss": 0.0127, "step": 99330 }, { "grad_norm": 0.2318258434534073, "learning_rate": 1.1944780400729993e-08, "loss": 0.0152, "step": 99340 }, { "grad_norm": 0.26606670022010803, "learning_rate": 1.1586112664085935e-08, "loss": 0.0144, "step": 99350 }, { "grad_norm": 0.19874921441078186, "learning_rate": 1.1232911585234984e-08, "loss": 0.0187, "step": 99360 }, { "grad_norm": 0.2819049656391144, "learning_rate": 1.0885177202790697e-08, "loss": 0.0173, "step": 99370 }, { "grad_norm": 0.28535547852516174, "learning_rate": 1.0542909554789315e-08, "loss": 0.0106, "step": 99380 }, { "grad_norm": 0.3259691298007965, "learning_rate": 1.0206108678656456e-08, "loss": 0.0194, "step": 99390 }, { "grad_norm": 0.09693332761526108, "learning_rate": 9.874774611223769e-09, "loss": 0.0187, "step": 99400 }, { "grad_norm": 0.521572470664978, "learning_rate": 9.548907388728933e-09, "loss": 0.0203, "step": 99410 }, { "grad_norm": 0.2101280689239502, "learning_rate": 9.228507046804557e-09, "loss": 0.0153, "step": 99420 }, { "grad_norm": 0.40070977807044983, "learning_rate": 8.91357362048928e-09, "loss": 0.0183, "step": 99430 }, { "grad_norm": 0.09437236189842224, "learning_rate": 8.604107144227769e-09, "loss": 0.0121, "step": 99440 }, { "grad_norm": 0.3183946907520294, "learning_rate": 8.300107651859623e-09, "loss": 0.012, "step": 99450 }, { "grad_norm": 0.3271963894367218, "learning_rate": 8.001575176630472e-09, "loss": 0.0137, "step": 99460 }, { "grad_norm": 0.20227831602096558, "learning_rate": 7.708509751186422e-09, "loss": 0.0136, "step": 99470 }, { "grad_norm": 0.36893972754478455, "learning_rate": 7.420911407579611e-09, "loss": 0.0151, "step": 99480 }, { "grad_norm": 0.8327710628509521, "learning_rate": 7.13878017725711e-09, "loss": 0.0168, "step": 99490 }, { "grad_norm": 0.40502604842185974, "learning_rate": 6.8621160910720125e-09, "loss": 0.019, "step": 99500 }, { "grad_norm": 1.3597346544265747, "learning_rate": 6.59091917928345e-09, "loss": 0.0171, "step": 99510 }, { "grad_norm": 0.4599474370479584, "learning_rate": 6.32518947155103e-09, "loss": 0.0135, "step": 99520 }, { "grad_norm": 0.5112684369087219, "learning_rate": 6.064926996929288e-09, "loss": 0.0162, "step": 99530 }, { "grad_norm": 0.3353360593318939, "learning_rate": 5.810131783884343e-09, "loss": 0.0126, "step": 99540 }, { "grad_norm": 0.14553654193878174, "learning_rate": 5.5608038602772415e-09, "loss": 0.0162, "step": 99550 }, { "grad_norm": 0.20843854546546936, "learning_rate": 5.316943253375062e-09, "loss": 0.0145, "step": 99560 }, { "grad_norm": 0.3547070622444153, "learning_rate": 5.07854998984536e-09, "loss": 0.0169, "step": 99570 }, { "grad_norm": 0.22549627721309662, "learning_rate": 4.845624095756174e-09, "loss": 0.0128, "step": 99580 }, { "grad_norm": 0.2015678733587265, "learning_rate": 4.618165596587121e-09, "loss": 0.0134, "step": 99590 }, { "grad_norm": 0.11018984019756317, "learning_rate": 4.396174517207197e-09, "loss": 0.016, "step": 99600 }, { "grad_norm": 0.22967903316020966, "learning_rate": 4.179650881896979e-09, "loss": 0.0162, "step": 99610 }, { "grad_norm": 0.6449858546257019, "learning_rate": 3.968594714331975e-09, "loss": 0.0108, "step": 99620 }, { "grad_norm": 0.509052574634552, "learning_rate": 3.763006037593719e-09, "loss": 0.019, "step": 99630 }, { "grad_norm": 0.1854086071252823, "learning_rate": 3.5628848741586786e-09, "loss": 0.0116, "step": 99640 }, { "grad_norm": 0.5963792204856873, "learning_rate": 3.368231245926001e-09, "loss": 0.0209, "step": 99650 }, { "grad_norm": 0.18452291190624237, "learning_rate": 3.179045174167561e-09, "loss": 0.0149, "step": 99660 }, { "grad_norm": 0.27776211500167847, "learning_rate": 2.9953266795834657e-09, "loss": 0.0191, "step": 99670 }, { "grad_norm": 0.4133114814758301, "learning_rate": 2.817075782263201e-09, "loss": 0.0152, "step": 99680 }, { "grad_norm": 0.283087819814682, "learning_rate": 2.6442925016967325e-09, "loss": 0.0114, "step": 99690 }, { "grad_norm": 6.013871192932129, "learning_rate": 2.4769768567800555e-09, "loss": 0.0184, "step": 99700 }, { "grad_norm": 0.405626118183136, "learning_rate": 2.315128865809646e-09, "loss": 0.0167, "step": 99710 }, { "grad_norm": 0.20045256614685059, "learning_rate": 2.1587485464880097e-09, "loss": 0.013, "step": 99720 }, { "grad_norm": 0.5246824622154236, "learning_rate": 2.0078359159125813e-09, "loss": 0.019, "step": 99730 }, { "grad_norm": 0.33645787835121155, "learning_rate": 1.8623909905923776e-09, "loss": 0.0217, "step": 99740 }, { "grad_norm": 0.23661044239997864, "learning_rate": 1.7224137864257916e-09, "loss": 0.0126, "step": 99750 }, { "grad_norm": 0.35067519545555115, "learning_rate": 1.5879043187283505e-09, "loss": 0.015, "step": 99760 }, { "grad_norm": 0.6287822723388672, "learning_rate": 1.458862602204958e-09, "loss": 0.0164, "step": 99770 }, { "grad_norm": 0.19927650690078735, "learning_rate": 1.3352886509720996e-09, "loss": 0.0179, "step": 99780 }, { "grad_norm": 0.37283748388290405, "learning_rate": 1.2171824785356389e-09, "loss": 0.0156, "step": 99790 }, { "grad_norm": 0.29393258690834045, "learning_rate": 1.1045440978185718e-09, "loss": 0.0121, "step": 99800 }, { "grad_norm": 0.8495547771453857, "learning_rate": 9.973735211332714e-10, "loss": 0.0163, "step": 99810 }, { "grad_norm": 0.519535481929779, "learning_rate": 8.956707602036929e-10, "loss": 0.0186, "step": 99820 }, { "grad_norm": 0.5969672799110413, "learning_rate": 7.994358261542712e-10, "loss": 0.0123, "step": 99830 }, { "grad_norm": 0.27104121446609497, "learning_rate": 7.086687295043692e-10, "loss": 0.016, "step": 99840 }, { "grad_norm": 0.4606420397758484, "learning_rate": 6.23369480179381e-10, "loss": 0.0157, "step": 99850 }, { "grad_norm": 0.17221368849277496, "learning_rate": 5.43538087510731e-10, "loss": 0.0193, "step": 99860 }, { "grad_norm": 0.5221474766731262, "learning_rate": 4.691745602303233e-10, "loss": 0.0326, "step": 99870 }, { "grad_norm": 0.42880645394325256, "learning_rate": 4.0027890646499034e-10, "loss": 0.0269, "step": 99880 }, { "grad_norm": 0.19861942529678345, "learning_rate": 3.368511337531466e-10, "loss": 0.0118, "step": 99890 }, { "grad_norm": 0.45664411783218384, "learning_rate": 2.788912490281348e-10, "loss": 0.0213, "step": 99900 }, { "grad_norm": 0.21175523102283478, "learning_rate": 2.2639925863487954e-10, "loss": 0.016, "step": 99910 }, { "grad_norm": 0.40152108669281006, "learning_rate": 1.7937516830213163e-10, "loss": 0.0116, "step": 99920 }, { "grad_norm": 0.39912018179893494, "learning_rate": 1.3781898318687702e-10, "loss": 0.0162, "step": 99930 }, { "grad_norm": 0.1768072247505188, "learning_rate": 1.0173070781882566e-10, "loss": 0.0121, "step": 99940 }, { "grad_norm": 0.4862559735774994, "learning_rate": 7.111034615592261e-11, "loss": 0.0148, "step": 99950 }, { "grad_norm": 0.32377785444259644, "learning_rate": 4.595790153993918e-11, "loss": 0.0116, "step": 99960 }, { "grad_norm": 0.3182258605957031, "learning_rate": 2.6273376729779585e-11, "loss": 0.0142, "step": 99970 }, { "grad_norm": 0.2530464828014374, "learning_rate": 1.2056773868174276e-11, "loss": 0.0099, "step": 99980 }, { "grad_norm": 0.42409440875053406, "learning_rate": 3.3080945149865926e-12, "loss": 0.0162, "step": 99990 }, { "grad_norm": 0.20356665551662445, "learning_rate": 2.7339625008337977e-14, "loss": 0.0111, "step": 100000 } ], "logging_steps": 10, "max_steps": 100000, "num_input_tokens_seen": 0, "num_train_epochs": 9223372036854775807, "save_steps": 20000, "stateful_callbacks": { "TrainerControl": { "args": { "should_epoch_stop": false, "should_evaluate": false, "should_log": false, "should_save": true, "should_training_stop": true }, "attributes": {} } }, "total_flos": 0.0, "train_batch_size": 8, "trial_name": null, "trial_params": null }